diff --git "a/trainer_state.json" "b/trainer_state.json" --- "a/trainer_state.json" +++ "b/trainer_state.json" @@ -1,6 +1,6 @@ { - "best_metric": 0.38733184337615967, - "best_model_checkpoint": "./food101_outputs/checkpoint-66283", + "best_metric": 0.6538042426109314, + "best_model_checkpoint": "./food101_outputs/checkpoint-94690", "epoch": 10.0, "eval_steps": 500, "global_step": 94690, @@ -10,66385 +10,66385 @@ "log_history": [ { "epoch": 0.0, - "grad_norm": 2.18501615524292, + "grad_norm": 14.387511253356934, "learning_rate": 1.9997887844545362e-05, - "loss": 4.6173, + "loss": 4.7177, "step": 10 }, { "epoch": 0.0, - "grad_norm": 2.0312955379486084, + "grad_norm": 18.5462703704834, "learning_rate": 1.999577568909072e-05, - "loss": 4.5932, + "loss": 4.7552, "step": 20 }, { "epoch": 0.0, - "grad_norm": 2.441335916519165, + "grad_norm": 15.047163009643555, "learning_rate": 1.9993663533636076e-05, - "loss": 4.6191, + "loss": 4.6636, "step": 30 }, { "epoch": 0.0, - "grad_norm": 2.273250102996826, + "grad_norm": 15.074764251708984, "learning_rate": 1.9991551378181437e-05, - "loss": 4.5727, + "loss": 4.6654, "step": 40 }, { "epoch": 0.01, - "grad_norm": 2.0238873958587646, + "grad_norm": 17.16153335571289, "learning_rate": 1.9989439222726794e-05, - "loss": 4.5787, + "loss": 4.5918, "step": 50 }, { "epoch": 0.01, - "grad_norm": 2.006110191345215, + "grad_norm": 13.39467716217041, "learning_rate": 1.9987327067272154e-05, - "loss": 4.5836, + "loss": 4.65, "step": 60 }, { "epoch": 0.01, - "grad_norm": 2.1601877212524414, + "grad_norm": 12.605389595031738, "learning_rate": 1.998521491181751e-05, - "loss": 4.5708, + "loss": 4.6137, "step": 70 }, { "epoch": 0.01, - "grad_norm": 2.2970635890960693, + "grad_norm": 13.440823554992676, "learning_rate": 1.9983102756362872e-05, - "loss": 4.5529, + "loss": 4.6253, "step": 80 }, { "epoch": 0.01, - "grad_norm": 2.2946841716766357, + "grad_norm": 9.523697853088379, "learning_rate": 1.998099060090823e-05, - "loss": 4.5443, + "loss": 4.5842, "step": 90 }, { "epoch": 0.01, - "grad_norm": 2.438825845718384, + "grad_norm": 10.992704391479492, "learning_rate": 1.9978878445453586e-05, - "loss": 4.5484, + "loss": 4.6164, "step": 100 }, { "epoch": 0.01, - "grad_norm": 2.2728381156921387, + "grad_norm": 10.147563934326172, "learning_rate": 1.9976766289998947e-05, - "loss": 4.5282, + "loss": 4.5307, "step": 110 }, { "epoch": 0.01, - "grad_norm": 2.279215097427368, + "grad_norm": 12.568976402282715, "learning_rate": 1.9974654134544304e-05, - "loss": 4.5248, + "loss": 4.6222, "step": 120 }, { "epoch": 0.01, - "grad_norm": 2.3124160766601562, + "grad_norm": 10.544739723205566, "learning_rate": 1.9972541979089664e-05, - "loss": 4.4828, + "loss": 4.5836, "step": 130 }, { "epoch": 0.01, - "grad_norm": 2.606757640838623, + "grad_norm": 10.174971580505371, "learning_rate": 1.997042982363502e-05, - "loss": 4.4937, + "loss": 4.5398, "step": 140 }, { "epoch": 0.02, - "grad_norm": 2.3704819679260254, + "grad_norm": 12.241962432861328, "learning_rate": 1.9968317668180378e-05, - "loss": 4.507, + "loss": 4.5444, "step": 150 }, { "epoch": 0.02, - "grad_norm": 2.567347526550293, + "grad_norm": 18.85626983642578, "learning_rate": 1.996620551272574e-05, - "loss": 4.4814, + "loss": 4.5444, "step": 160 }, { "epoch": 0.02, - "grad_norm": 3.3169007301330566, + "grad_norm": 15.641300201416016, "learning_rate": 1.9964093357271096e-05, - "loss": 4.4447, + "loss": 4.4834, "step": 170 }, { "epoch": 0.02, - "grad_norm": 3.0721073150634766, + "grad_norm": 18.38041114807129, "learning_rate": 1.9961981201816456e-05, - "loss": 4.4437, + "loss": 4.6295, "step": 180 }, { "epoch": 0.02, - "grad_norm": 2.5403361320495605, + "grad_norm": 18.390491485595703, "learning_rate": 1.9959869046361813e-05, - "loss": 4.4361, + "loss": 4.5353, "step": 190 }, { "epoch": 0.02, - "grad_norm": 2.92630934715271, + "grad_norm": 16.406524658203125, "learning_rate": 1.995775689090717e-05, - "loss": 4.4753, + "loss": 4.6098, "step": 200 }, { "epoch": 0.02, - "grad_norm": 2.492459774017334, + "grad_norm": 16.564170837402344, "learning_rate": 1.995564473545253e-05, - "loss": 4.4505, + "loss": 4.5655, "step": 210 }, { "epoch": 0.02, - "grad_norm": 2.675485849380493, + "grad_norm": 16.3575439453125, "learning_rate": 1.9953532579997888e-05, - "loss": 4.4798, + "loss": 4.5906, "step": 220 }, { "epoch": 0.02, - "grad_norm": 2.9208128452301025, + "grad_norm": 18.070573806762695, "learning_rate": 1.995142042454325e-05, - "loss": 4.4187, + "loss": 4.4757, "step": 230 }, { "epoch": 0.03, - "grad_norm": 2.775249481201172, + "grad_norm": 17.448928833007812, "learning_rate": 1.994930826908861e-05, - "loss": 4.451, + "loss": 4.5288, "step": 240 }, { "epoch": 0.03, - "grad_norm": 2.7175989151000977, + "grad_norm": 14.745076179504395, "learning_rate": 1.9947196113633963e-05, - "loss": 4.4468, + "loss": 4.4899, "step": 250 }, { "epoch": 0.03, - "grad_norm": 2.869677782058716, + "grad_norm": 14.046401023864746, "learning_rate": 1.9945083958179323e-05, - "loss": 4.3973, + "loss": 4.4295, "step": 260 }, { "epoch": 0.03, - "grad_norm": 2.9036285877227783, + "grad_norm": 17.376062393188477, "learning_rate": 1.9942971802724684e-05, - "loss": 4.4145, + "loss": 4.4818, "step": 270 }, { "epoch": 0.03, - "grad_norm": 4.500214099884033, + "grad_norm": 19.091629028320312, "learning_rate": 1.994085964727004e-05, - "loss": 4.3788, + "loss": 4.4241, "step": 280 }, { "epoch": 0.03, - "grad_norm": 2.8079919815063477, + "grad_norm": 14.785734176635742, "learning_rate": 1.99387474918154e-05, - "loss": 4.4047, + "loss": 4.4392, "step": 290 }, { "epoch": 0.03, - "grad_norm": 3.453835964202881, + "grad_norm": 18.16326332092285, "learning_rate": 1.9936635336360758e-05, - "loss": 4.3612, + "loss": 4.4116, "step": 300 }, { "epoch": 0.03, - "grad_norm": 2.865516185760498, + "grad_norm": 15.415942192077637, "learning_rate": 1.9934523180906115e-05, - "loss": 4.3835, + "loss": 4.3541, "step": 310 }, { "epoch": 0.03, - "grad_norm": 2.6516056060791016, + "grad_norm": 17.445484161376953, "learning_rate": 1.9932411025451476e-05, - "loss": 4.3501, + "loss": 4.4132, "step": 320 }, { "epoch": 0.03, - "grad_norm": 2.6655831336975098, + "grad_norm": 17.935531616210938, "learning_rate": 1.9930298869996833e-05, - "loss": 4.3721, + "loss": 4.3646, "step": 330 }, { "epoch": 0.04, - "grad_norm": 2.737273931503296, + "grad_norm": 16.038658142089844, "learning_rate": 1.9928186714542193e-05, - "loss": 4.3322, + "loss": 4.2896, "step": 340 }, { "epoch": 0.04, - "grad_norm": 2.9579668045043945, + "grad_norm": 18.92913246154785, "learning_rate": 1.992607455908755e-05, - "loss": 4.3187, + "loss": 4.2277, "step": 350 }, { "epoch": 0.04, - "grad_norm": 2.6012980937957764, + "grad_norm": 16.960397720336914, "learning_rate": 1.9923962403632908e-05, - "loss": 4.3441, + "loss": 4.4276, "step": 360 }, { "epoch": 0.04, - "grad_norm": 2.883591413497925, + "grad_norm": 19.60896873474121, "learning_rate": 1.9921850248178268e-05, - "loss": 4.3826, + "loss": 4.301, "step": 370 }, { "epoch": 0.04, - "grad_norm": 2.749418258666992, + "grad_norm": 17.474605560302734, "learning_rate": 1.9919738092723625e-05, - "loss": 4.3174, + "loss": 4.2569, "step": 380 }, { "epoch": 0.04, - "grad_norm": 2.8037569522857666, + "grad_norm": 17.65813636779785, "learning_rate": 1.9917625937268986e-05, - "loss": 4.2993, + "loss": 4.2951, "step": 390 }, { "epoch": 0.04, - "grad_norm": 2.918322801589966, + "grad_norm": 17.197153091430664, "learning_rate": 1.9915513781814343e-05, - "loss": 4.3125, + "loss": 4.204, "step": 400 }, { "epoch": 0.04, - "grad_norm": 2.616011142730713, + "grad_norm": 18.942302703857422, "learning_rate": 1.99134016263597e-05, - "loss": 4.294, + "loss": 4.2524, "step": 410 }, { "epoch": 0.04, - "grad_norm": 2.808267831802368, + "grad_norm": 18.96645164489746, "learning_rate": 1.991128947090506e-05, - "loss": 4.3075, + "loss": 4.329, "step": 420 }, { "epoch": 0.05, - "grad_norm": 2.9122743606567383, + "grad_norm": 21.482107162475586, "learning_rate": 1.9909177315450417e-05, - "loss": 4.28, + "loss": 4.1922, "step": 430 }, { "epoch": 0.05, - "grad_norm": 3.018282651901245, + "grad_norm": 18.220125198364258, "learning_rate": 1.9907065159995778e-05, - "loss": 4.2619, + "loss": 4.1938, "step": 440 }, { "epoch": 0.05, - "grad_norm": 2.6209874153137207, + "grad_norm": 19.702587127685547, "learning_rate": 1.9904953004541138e-05, - "loss": 4.2699, + "loss": 4.2504, "step": 450 }, { "epoch": 0.05, - "grad_norm": 2.8852124214172363, + "grad_norm": 20.442445755004883, "learning_rate": 1.9902840849086495e-05, - "loss": 4.2306, + "loss": 4.1665, "step": 460 }, { "epoch": 0.05, - "grad_norm": 3.15557861328125, + "grad_norm": 21.34444808959961, "learning_rate": 1.9900728693631852e-05, - "loss": 4.1786, + "loss": 4.0923, "step": 470 }, { "epoch": 0.05, - "grad_norm": 2.798964500427246, + "grad_norm": 20.636573791503906, "learning_rate": 1.9898616538177213e-05, - "loss": 4.2131, + "loss": 4.2182, "step": 480 }, { "epoch": 0.05, - "grad_norm": 2.867426872253418, + "grad_norm": 18.952808380126953, "learning_rate": 1.989650438272257e-05, - "loss": 4.2367, + "loss": 4.1409, "step": 490 }, { "epoch": 0.05, - "grad_norm": 3.0130295753479004, + "grad_norm": 18.99601936340332, "learning_rate": 1.989439222726793e-05, - "loss": 4.1569, + "loss": 4.1351, "step": 500 }, { "epoch": 0.05, - "grad_norm": 3.0225939750671387, + "grad_norm": 26.339059829711914, "learning_rate": 1.9892280071813287e-05, - "loss": 4.1888, + "loss": 3.8415, "step": 510 }, { "epoch": 0.05, - "grad_norm": 2.9687769412994385, + "grad_norm": 22.5743350982666, "learning_rate": 1.9890167916358645e-05, - "loss": 4.1683, + "loss": 4.0262, "step": 520 }, { "epoch": 0.06, - "grad_norm": 2.9489035606384277, + "grad_norm": 19.81697654724121, "learning_rate": 1.9888055760904005e-05, - "loss": 4.1844, + "loss": 4.0318, "step": 530 }, { "epoch": 0.06, - "grad_norm": 2.8630318641662598, + "grad_norm": 19.67277717590332, "learning_rate": 1.9885943605449362e-05, - "loss": 4.1701, + "loss": 4.0427, "step": 540 }, { "epoch": 0.06, - "grad_norm": 2.792797565460205, + "grad_norm": 23.035232543945312, "learning_rate": 1.9883831449994723e-05, - "loss": 4.1471, + "loss": 3.9625, "step": 550 }, { "epoch": 0.06, - "grad_norm": 3.493060827255249, + "grad_norm": 23.762420654296875, "learning_rate": 1.988171929454008e-05, - "loss": 4.1198, + "loss": 4.1049, "step": 560 }, { "epoch": 0.06, - "grad_norm": 2.8298556804656982, + "grad_norm": 20.214244842529297, "learning_rate": 1.9879607139085437e-05, - "loss": 4.1009, + "loss": 3.891, "step": 570 }, { "epoch": 0.06, - "grad_norm": 3.280174970626831, + "grad_norm": 20.340961456298828, "learning_rate": 1.9877494983630797e-05, - "loss": 4.0668, + "loss": 3.8393, "step": 580 }, { "epoch": 0.06, - "grad_norm": 3.072111129760742, + "grad_norm": 20.29729652404785, "learning_rate": 1.9875382828176154e-05, - "loss": 4.1237, + "loss": 4.0014, "step": 590 }, { "epoch": 0.06, - "grad_norm": 3.567610502243042, + "grad_norm": 24.795894622802734, "learning_rate": 1.9873270672721515e-05, - "loss": 4.0684, + "loss": 4.0228, "step": 600 }, { "epoch": 0.06, - "grad_norm": 3.1131155490875244, + "grad_norm": 26.627357482910156, "learning_rate": 1.9871158517266872e-05, - "loss": 4.0283, + "loss": 3.891, "step": 610 }, { "epoch": 0.07, - "grad_norm": 2.8927228450775146, + "grad_norm": 21.620800018310547, "learning_rate": 1.9869046361812232e-05, - "loss": 4.0787, + "loss": 3.8527, "step": 620 }, { "epoch": 0.07, - "grad_norm": 3.208298921585083, + "grad_norm": 19.732690811157227, "learning_rate": 1.986693420635759e-05, - "loss": 4.0938, + "loss": 3.9457, "step": 630 }, { "epoch": 0.07, - "grad_norm": 3.477884531021118, + "grad_norm": 20.31859588623047, "learning_rate": 1.9864822050902947e-05, - "loss": 4.0434, + "loss": 3.9541, "step": 640 }, { "epoch": 0.07, - "grad_norm": 3.1580379009246826, + "grad_norm": 19.397991180419922, "learning_rate": 1.9862709895448307e-05, - "loss": 4.0496, + "loss": 4.0279, "step": 650 }, { "epoch": 0.07, - "grad_norm": 3.322648525238037, + "grad_norm": 23.20944595336914, "learning_rate": 1.9860597739993664e-05, - "loss": 4.0754, + "loss": 4.1191, "step": 660 }, { "epoch": 0.07, - "grad_norm": 2.7844114303588867, + "grad_norm": 23.321985244750977, "learning_rate": 1.9858485584539025e-05, - "loss": 4.061, + "loss": 3.8562, "step": 670 }, { "epoch": 0.07, - "grad_norm": 3.8404934406280518, + "grad_norm": 22.900121688842773, "learning_rate": 1.985637342908438e-05, - "loss": 4.0777, + "loss": 4.0095, "step": 680 }, { "epoch": 0.07, - "grad_norm": 3.234201431274414, + "grad_norm": 21.81748390197754, "learning_rate": 1.985426127362974e-05, - "loss": 4.0201, + "loss": 3.9263, "step": 690 }, { "epoch": 0.07, - "grad_norm": 3.2527825832366943, + "grad_norm": 23.724430084228516, "learning_rate": 1.98521491181751e-05, - "loss": 4.057, + "loss": 3.8656, "step": 700 }, { "epoch": 0.07, - "grad_norm": 3.9236879348754883, + "grad_norm": 20.777790069580078, "learning_rate": 1.985003696272046e-05, - "loss": 3.9752, + "loss": 3.8204, "step": 710 }, { "epoch": 0.08, - "grad_norm": 2.9799721240997314, + "grad_norm": 25.86058235168457, "learning_rate": 1.9847924807265817e-05, - "loss": 3.9927, + "loss": 3.85, "step": 720 }, { "epoch": 0.08, - "grad_norm": 2.7905828952789307, + "grad_norm": 21.785375595092773, "learning_rate": 1.9845812651811177e-05, - "loss": 3.9634, + "loss": 3.8238, "step": 730 }, { "epoch": 0.08, - "grad_norm": 3.193319082260132, + "grad_norm": 22.77192497253418, "learning_rate": 1.9843700496356534e-05, - "loss": 3.9074, + "loss": 3.6807, "step": 740 }, { "epoch": 0.08, - "grad_norm": 2.7580745220184326, + "grad_norm": 26.122631072998047, "learning_rate": 1.984158834090189e-05, - "loss": 3.951, + "loss": 3.822, "step": 750 }, { "epoch": 0.08, - "grad_norm": 3.7128849029541016, + "grad_norm": 22.86161231994629, "learning_rate": 1.9839476185447252e-05, - "loss": 3.9795, + "loss": 3.8763, "step": 760 }, { "epoch": 0.08, - "grad_norm": 3.1459643840789795, + "grad_norm": 24.944202423095703, "learning_rate": 1.983736402999261e-05, - "loss": 3.9472, + "loss": 3.8318, "step": 770 }, { "epoch": 0.08, - "grad_norm": 3.0623414516448975, + "grad_norm": 25.475622177124023, "learning_rate": 1.983525187453797e-05, - "loss": 3.9899, + "loss": 4.0321, "step": 780 }, { "epoch": 0.08, - "grad_norm": 2.7380692958831787, + "grad_norm": 23.278427124023438, "learning_rate": 1.9833139719083327e-05, - "loss": 3.9399, + "loss": 3.8178, "step": 790 }, { "epoch": 0.08, - "grad_norm": 2.9456846714019775, + "grad_norm": 20.4398193359375, "learning_rate": 1.9831027563628684e-05, - "loss": 3.9294, + "loss": 3.8706, "step": 800 }, { "epoch": 0.09, - "grad_norm": 3.2389512062072754, + "grad_norm": 20.128019332885742, "learning_rate": 1.9828915408174044e-05, - "loss": 3.8768, + "loss": 3.6419, "step": 810 }, { "epoch": 0.09, - "grad_norm": 3.018588066101074, + "grad_norm": 19.718257904052734, "learning_rate": 1.98268032527194e-05, - "loss": 3.8985, + "loss": 3.9734, "step": 820 }, { "epoch": 0.09, - "grad_norm": 2.9473302364349365, + "grad_norm": 19.32143783569336, "learning_rate": 1.982469109726476e-05, - "loss": 3.8458, + "loss": 3.6771, "step": 830 }, { "epoch": 0.09, - "grad_norm": 4.163280487060547, + "grad_norm": 22.23832893371582, "learning_rate": 1.982257894181012e-05, - "loss": 3.8561, + "loss": 3.6317, "step": 840 }, { "epoch": 0.09, - "grad_norm": 3.241572856903076, + "grad_norm": 20.606582641601562, "learning_rate": 1.9820466786355476e-05, - "loss": 3.8584, + "loss": 3.5834, "step": 850 }, { "epoch": 0.09, - "grad_norm": 3.2776529788970947, + "grad_norm": 21.469785690307617, "learning_rate": 1.9818354630900836e-05, - "loss": 3.8499, + "loss": 3.7126, "step": 860 }, { "epoch": 0.09, - "grad_norm": 3.319232702255249, + "grad_norm": 20.262170791625977, "learning_rate": 1.9816242475446193e-05, - "loss": 3.8119, + "loss": 3.6849, "step": 870 }, { "epoch": 0.09, - "grad_norm": 3.139664649963379, + "grad_norm": 20.275297164916992, "learning_rate": 1.9814130319991554e-05, - "loss": 3.9217, + "loss": 3.8703, "step": 880 }, { "epoch": 0.09, - "grad_norm": 3.468867063522339, + "grad_norm": 20.595355987548828, "learning_rate": 1.981201816453691e-05, - "loss": 3.7769, + "loss": 3.6329, "step": 890 }, { "epoch": 0.1, - "grad_norm": 2.87005877494812, + "grad_norm": 24.25246810913086, "learning_rate": 1.9809906009082268e-05, - "loss": 3.796, + "loss": 3.6767, "step": 900 }, { "epoch": 0.1, - "grad_norm": 2.79636549949646, + "grad_norm": 19.541168212890625, "learning_rate": 1.980779385362763e-05, - "loss": 3.7419, + "loss": 3.6868, "step": 910 }, { "epoch": 0.1, - "grad_norm": 3.276163101196289, + "grad_norm": 21.042734146118164, "learning_rate": 1.9805681698172986e-05, - "loss": 3.7527, + "loss": 3.5782, "step": 920 }, { "epoch": 0.1, - "grad_norm": 4.278825759887695, + "grad_norm": 24.901121139526367, "learning_rate": 1.9803569542718346e-05, - "loss": 3.8464, + "loss": 3.7198, "step": 930 }, { "epoch": 0.1, - "grad_norm": 4.053592681884766, + "grad_norm": 23.75690460205078, "learning_rate": 1.9801457387263707e-05, - "loss": 3.7959, + "loss": 3.764, "step": 940 }, { "epoch": 0.1, - "grad_norm": 3.3290977478027344, + "grad_norm": 24.520280838012695, "learning_rate": 1.979934523180906e-05, - "loss": 3.8644, + "loss": 3.7773, "step": 950 }, { "epoch": 0.1, - "grad_norm": 3.130943775177002, + "grad_norm": 21.464502334594727, "learning_rate": 1.979723307635442e-05, - "loss": 3.8571, + "loss": 3.6058, "step": 960 }, { "epoch": 0.1, - "grad_norm": 3.8699238300323486, + "grad_norm": 25.217031478881836, "learning_rate": 1.979512092089978e-05, - "loss": 3.8675, + "loss": 3.6597, "step": 970 }, { "epoch": 0.1, - "grad_norm": 3.989039182662964, + "grad_norm": 29.299480438232422, "learning_rate": 1.9793008765445138e-05, - "loss": 3.7582, + "loss": 3.6392, "step": 980 }, { "epoch": 0.1, - "grad_norm": 4.00246524810791, + "grad_norm": 23.91437530517578, "learning_rate": 1.97908966099905e-05, - "loss": 3.7532, + "loss": 3.5512, "step": 990 }, { "epoch": 0.11, - "grad_norm": 3.0782368183135986, + "grad_norm": 21.430095672607422, "learning_rate": 1.9788784454535856e-05, - "loss": 3.783, + "loss": 3.6726, "step": 1000 }, { "epoch": 0.11, - "grad_norm": 3.9442250728607178, + "grad_norm": 19.920562744140625, "learning_rate": 1.9786672299081213e-05, - "loss": 3.805, + "loss": 3.5859, "step": 1010 }, { "epoch": 0.11, - "grad_norm": 3.447726249694824, + "grad_norm": 23.693559646606445, "learning_rate": 1.9784560143626573e-05, - "loss": 3.7921, + "loss": 3.7059, "step": 1020 }, { "epoch": 0.11, - "grad_norm": 4.929808139801025, + "grad_norm": 25.999406814575195, "learning_rate": 1.978244798817193e-05, - "loss": 3.6546, + "loss": 3.383, "step": 1030 }, { "epoch": 0.11, - "grad_norm": 3.22365665435791, + "grad_norm": 20.247129440307617, "learning_rate": 1.978033583271729e-05, - "loss": 3.7709, + "loss": 3.7147, "step": 1040 }, { "epoch": 0.11, - "grad_norm": 3.707246780395508, + "grad_norm": 27.223190307617188, "learning_rate": 1.9778223677262648e-05, - "loss": 3.7381, + "loss": 3.565, "step": 1050 }, { "epoch": 0.11, - "grad_norm": 3.2182254791259766, + "grad_norm": 23.104005813598633, "learning_rate": 1.9776111521808005e-05, - "loss": 3.6447, + "loss": 3.415, "step": 1060 }, { "epoch": 0.11, - "grad_norm": 3.872714042663574, + "grad_norm": 25.900564193725586, "learning_rate": 1.9773999366353366e-05, - "loss": 3.6921, + "loss": 3.6855, "step": 1070 }, { "epoch": 0.11, - "grad_norm": 4.200910568237305, + "grad_norm": 24.072525024414062, "learning_rate": 1.9771887210898723e-05, - "loss": 3.6924, + "loss": 3.4796, "step": 1080 }, { "epoch": 0.12, - "grad_norm": 3.52093505859375, + "grad_norm": 23.994529724121094, "learning_rate": 1.9769775055444083e-05, - "loss": 3.6986, + "loss": 3.5281, "step": 1090 }, { "epoch": 0.12, - "grad_norm": 3.1279048919677734, + "grad_norm": 25.851594924926758, "learning_rate": 1.976766289998944e-05, - "loss": 3.661, + "loss": 3.5006, "step": 1100 }, { "epoch": 0.12, - "grad_norm": 3.071439504623413, + "grad_norm": 24.81606101989746, "learning_rate": 1.9765550744534797e-05, - "loss": 3.6896, + "loss": 3.5787, "step": 1110 }, { "epoch": 0.12, - "grad_norm": 3.7281079292297363, + "grad_norm": 21.884885787963867, "learning_rate": 1.9763438589080158e-05, - "loss": 3.6118, + "loss": 3.3893, "step": 1120 }, { "epoch": 0.12, - "grad_norm": 3.7188735008239746, + "grad_norm": 21.61090087890625, "learning_rate": 1.9761326433625515e-05, - "loss": 3.5433, + "loss": 3.4158, "step": 1130 }, { "epoch": 0.12, - "grad_norm": 3.5627167224884033, + "grad_norm": 26.439529418945312, "learning_rate": 1.9759214278170875e-05, - "loss": 3.6673, + "loss": 3.6164, "step": 1140 }, { "epoch": 0.12, - "grad_norm": 3.4530186653137207, + "grad_norm": 23.802339553833008, "learning_rate": 1.9757102122716232e-05, - "loss": 3.5767, + "loss": 3.2462, "step": 1150 }, { "epoch": 0.12, - "grad_norm": 3.1565701961517334, + "grad_norm": 23.583215713500977, "learning_rate": 1.9754989967261593e-05, - "loss": 3.6196, + "loss": 3.3751, "step": 1160 }, { "epoch": 0.12, - "grad_norm": 3.567331314086914, + "grad_norm": 24.366682052612305, "learning_rate": 1.975287781180695e-05, - "loss": 3.6096, + "loss": 3.4347, "step": 1170 }, { "epoch": 0.12, - "grad_norm": 3.802793025970459, + "grad_norm": 21.2420711517334, "learning_rate": 1.9750765656352307e-05, - "loss": 3.6154, + "loss": 3.4069, "step": 1180 }, { "epoch": 0.13, - "grad_norm": 3.9474780559539795, + "grad_norm": 31.975269317626953, "learning_rate": 1.9748653500897668e-05, - "loss": 3.5395, + "loss": 3.4701, "step": 1190 }, { "epoch": 0.13, - "grad_norm": 3.7360029220581055, + "grad_norm": 21.66954231262207, "learning_rate": 1.9746541345443028e-05, - "loss": 3.5538, + "loss": 3.3577, "step": 1200 }, { "epoch": 0.13, - "grad_norm": 3.7871196269989014, + "grad_norm": 24.07439422607422, "learning_rate": 1.9744429189988385e-05, - "loss": 3.6876, + "loss": 3.6055, "step": 1210 }, { "epoch": 0.13, - "grad_norm": 5.265528202056885, + "grad_norm": 21.495677947998047, "learning_rate": 1.9742317034533742e-05, - "loss": 3.4971, + "loss": 3.2767, "step": 1220 }, { "epoch": 0.13, - "grad_norm": 4.033239841461182, + "grad_norm": 30.22323989868164, "learning_rate": 1.9740204879079103e-05, - "loss": 3.5896, + "loss": 3.3617, "step": 1230 }, { "epoch": 0.13, - "grad_norm": 8.125322341918945, + "grad_norm": 25.559993743896484, "learning_rate": 1.973809272362446e-05, - "loss": 3.5193, + "loss": 3.3466, "step": 1240 }, { "epoch": 0.13, - "grad_norm": 3.6840741634368896, + "grad_norm": 23.625041961669922, "learning_rate": 1.973598056816982e-05, - "loss": 3.6259, + "loss": 3.5106, "step": 1250 }, { "epoch": 0.13, - "grad_norm": 3.8130433559417725, + "grad_norm": 24.216169357299805, "learning_rate": 1.9733868412715177e-05, - "loss": 3.5738, + "loss": 3.4674, "step": 1260 }, { "epoch": 0.13, - "grad_norm": 3.699700355529785, + "grad_norm": 18.935136795043945, "learning_rate": 1.9731756257260538e-05, - "loss": 3.5909, + "loss": 3.4708, "step": 1270 }, { "epoch": 0.14, - "grad_norm": 3.13495135307312, + "grad_norm": 21.217662811279297, "learning_rate": 1.9729644101805895e-05, - "loss": 3.5567, + "loss": 3.4272, "step": 1280 }, { "epoch": 0.14, - "grad_norm": 3.8312883377075195, + "grad_norm": 21.96112060546875, "learning_rate": 1.9727531946351252e-05, - "loss": 3.6393, + "loss": 3.5012, "step": 1290 }, { "epoch": 0.14, - "grad_norm": 3.9041221141815186, + "grad_norm": 23.618160247802734, "learning_rate": 1.9725419790896612e-05, - "loss": 3.5658, + "loss": 3.4636, "step": 1300 }, { "epoch": 0.14, - "grad_norm": 3.4351096153259277, + "grad_norm": 20.6265869140625, "learning_rate": 1.972330763544197e-05, - "loss": 3.6088, + "loss": 3.5193, "step": 1310 }, { "epoch": 0.14, - "grad_norm": 4.054837703704834, + "grad_norm": 20.633275985717773, "learning_rate": 1.972119547998733e-05, - "loss": 3.5427, + "loss": 3.4483, "step": 1320 }, { "epoch": 0.14, - "grad_norm": 3.594484567642212, + "grad_norm": 25.025358200073242, "learning_rate": 1.9719083324532687e-05, - "loss": 3.6087, + "loss": 3.7124, "step": 1330 }, { "epoch": 0.14, - "grad_norm": 3.531461238861084, + "grad_norm": 18.0478515625, "learning_rate": 1.9716971169078044e-05, - "loss": 3.4497, + "loss": 3.3036, "step": 1340 }, { "epoch": 0.14, - "grad_norm": 3.524627208709717, + "grad_norm": 24.26726531982422, "learning_rate": 1.9714859013623405e-05, - "loss": 3.4646, + "loss": 3.2914, "step": 1350 }, { "epoch": 0.14, - "grad_norm": 3.2439498901367188, + "grad_norm": 24.585289001464844, "learning_rate": 1.971274685816876e-05, - "loss": 3.5603, + "loss": 3.3202, "step": 1360 }, { "epoch": 0.14, - "grad_norm": 4.7615861892700195, + "grad_norm": 23.008193969726562, "learning_rate": 1.9710634702714122e-05, - "loss": 3.4743, + "loss": 3.426, "step": 1370 }, { "epoch": 0.15, - "grad_norm": 3.9091198444366455, + "grad_norm": 23.345766067504883, "learning_rate": 1.9708522547259483e-05, - "loss": 3.5223, + "loss": 3.2538, "step": 1380 }, { "epoch": 0.15, - "grad_norm": 8.021449089050293, + "grad_norm": 23.57155418395996, "learning_rate": 1.9706410391804836e-05, - "loss": 3.5392, + "loss": 3.2771, "step": 1390 }, { "epoch": 0.15, - "grad_norm": 4.490795612335205, + "grad_norm": 26.990680694580078, "learning_rate": 1.9704298236350197e-05, - "loss": 3.5646, + "loss": 3.355, "step": 1400 }, { "epoch": 0.15, - "grad_norm": 3.704806327819824, + "grad_norm": 23.68392562866211, "learning_rate": 1.9702186080895557e-05, - "loss": 3.4494, + "loss": 3.2934, "step": 1410 }, { "epoch": 0.15, - "grad_norm": 3.705970287322998, + "grad_norm": 24.421947479248047, "learning_rate": 1.9700073925440914e-05, - "loss": 3.4086, + "loss": 3.4418, "step": 1420 }, { "epoch": 0.15, - "grad_norm": 4.386157512664795, + "grad_norm": 22.3470458984375, "learning_rate": 1.9697961769986275e-05, - "loss": 3.3838, + "loss": 3.3423, "step": 1430 }, { "epoch": 0.15, - "grad_norm": 4.386314868927002, + "grad_norm": 26.182706832885742, "learning_rate": 1.9695849614531632e-05, - "loss": 3.4436, + "loss": 3.3827, "step": 1440 }, { "epoch": 0.15, - "grad_norm": 3.366546392440796, + "grad_norm": 25.67595863342285, "learning_rate": 1.969373745907699e-05, - "loss": 3.4374, + "loss": 3.3442, "step": 1450 }, { "epoch": 0.15, - "grad_norm": 3.5111939907073975, + "grad_norm": 24.818788528442383, "learning_rate": 1.969162530362235e-05, - "loss": 3.4535, + "loss": 3.1827, "step": 1460 }, { "epoch": 0.16, - "grad_norm": 3.9938416481018066, + "grad_norm": 22.347183227539062, "learning_rate": 1.9689513148167707e-05, - "loss": 3.4619, + "loss": 3.3186, "step": 1470 }, { "epoch": 0.16, - "grad_norm": 4.054739952087402, + "grad_norm": 24.535755157470703, "learning_rate": 1.9687400992713067e-05, - "loss": 3.3816, + "loss": 3.1999, "step": 1480 }, { "epoch": 0.16, - "grad_norm": 3.406235933303833, + "grad_norm": 20.85918617248535, "learning_rate": 1.9685288837258424e-05, - "loss": 3.2236, + "loss": 3.0601, "step": 1490 }, { "epoch": 0.16, - "grad_norm": 4.60813570022583, + "grad_norm": 24.30315399169922, "learning_rate": 1.968317668180378e-05, - "loss": 3.3998, + "loss": 3.2628, "step": 1500 }, { "epoch": 0.16, - "grad_norm": 4.607341766357422, + "grad_norm": 22.483661651611328, "learning_rate": 1.968106452634914e-05, - "loss": 3.578, + "loss": 3.4575, "step": 1510 }, { "epoch": 0.16, - "grad_norm": 3.6629652976989746, + "grad_norm": 22.881254196166992, "learning_rate": 1.96789523708945e-05, - "loss": 3.3465, + "loss": 3.1173, "step": 1520 }, { "epoch": 0.16, - "grad_norm": 4.029366970062256, + "grad_norm": 23.10747528076172, "learning_rate": 1.967684021543986e-05, - "loss": 3.4027, + "loss": 3.3184, "step": 1530 }, { "epoch": 0.16, - "grad_norm": 4.471227645874023, + "grad_norm": 23.063983917236328, "learning_rate": 1.9674728059985216e-05, - "loss": 3.5538, + "loss": 3.4355, "step": 1540 }, { "epoch": 0.16, - "grad_norm": 3.8424534797668457, + "grad_norm": 21.188846588134766, "learning_rate": 1.9672615904530573e-05, - "loss": 3.339, + "loss": 3.3662, "step": 1550 }, { "epoch": 0.16, - "grad_norm": 3.5136735439300537, + "grad_norm": 24.590206146240234, "learning_rate": 1.9670503749075934e-05, - "loss": 3.3722, + "loss": 3.3473, "step": 1560 }, { "epoch": 0.17, - "grad_norm": 3.7872393131256104, + "grad_norm": 22.647876739501953, "learning_rate": 1.966839159362129e-05, - "loss": 3.2436, + "loss": 3.2377, "step": 1570 }, { "epoch": 0.17, - "grad_norm": 4.37291955947876, + "grad_norm": 23.447053909301758, "learning_rate": 1.966627943816665e-05, - "loss": 3.4222, + "loss": 3.4135, "step": 1580 }, { "epoch": 0.17, - "grad_norm": 3.4078798294067383, + "grad_norm": 23.366701126098633, "learning_rate": 1.966416728271201e-05, - "loss": 3.3011, + "loss": 3.2933, "step": 1590 }, { "epoch": 0.17, - "grad_norm": 3.728895902633667, + "grad_norm": 27.811948776245117, "learning_rate": 1.9662055127257366e-05, - "loss": 3.263, + "loss": 3.0247, "step": 1600 }, { "epoch": 0.17, - "grad_norm": 4.061984062194824, + "grad_norm": 22.55267906188965, "learning_rate": 1.9659942971802726e-05, - "loss": 3.34, + "loss": 3.1424, "step": 1610 }, { "epoch": 0.17, - "grad_norm": 4.826902866363525, + "grad_norm": 25.1356258392334, "learning_rate": 1.9657830816348083e-05, - "loss": 3.3219, + "loss": 3.1434, "step": 1620 }, { "epoch": 0.17, - "grad_norm": 4.131901741027832, + "grad_norm": 23.032718658447266, "learning_rate": 1.9655718660893444e-05, - "loss": 3.3285, + "loss": 3.2538, "step": 1630 }, { "epoch": 0.17, - "grad_norm": 4.223285675048828, + "grad_norm": 21.933040618896484, "learning_rate": 1.9653606505438804e-05, - "loss": 3.1889, + "loss": 3.1622, "step": 1640 }, { "epoch": 0.17, - "grad_norm": 3.6805264949798584, + "grad_norm": 20.415634155273438, "learning_rate": 1.965149434998416e-05, - "loss": 3.3265, + "loss": 3.2132, "step": 1650 }, { "epoch": 0.18, - "grad_norm": 3.465146541595459, + "grad_norm": 23.522428512573242, "learning_rate": 1.9649382194529518e-05, - "loss": 3.2484, + "loss": 3.0452, "step": 1660 }, { "epoch": 0.18, - "grad_norm": 5.077834129333496, + "grad_norm": 22.7681827545166, "learning_rate": 1.964727003907488e-05, - "loss": 3.4329, + "loss": 3.1764, "step": 1670 }, { "epoch": 0.18, - "grad_norm": 4.939765930175781, + "grad_norm": 25.001937866210938, "learning_rate": 1.9645157883620236e-05, - "loss": 3.4118, + "loss": 3.2699, "step": 1680 }, { "epoch": 0.18, - "grad_norm": 4.67460298538208, + "grad_norm": 29.0770206451416, "learning_rate": 1.9643045728165596e-05, - "loss": 3.249, + "loss": 3.2895, "step": 1690 }, { "epoch": 0.18, - "grad_norm": 3.849221706390381, + "grad_norm": 28.764110565185547, "learning_rate": 1.9640933572710953e-05, - "loss": 3.3058, + "loss": 3.2133, "step": 1700 }, { "epoch": 0.18, - "grad_norm": 3.646660089492798, + "grad_norm": 24.286548614501953, "learning_rate": 1.963882141725631e-05, - "loss": 3.1185, + "loss": 2.8323, "step": 1710 }, { "epoch": 0.18, - "grad_norm": 3.4824769496917725, + "grad_norm": 23.536100387573242, "learning_rate": 1.963670926180167e-05, - "loss": 3.2542, + "loss": 3.1191, "step": 1720 }, { "epoch": 0.18, - "grad_norm": 4.08527135848999, + "grad_norm": 25.168909072875977, "learning_rate": 1.9634597106347028e-05, - "loss": 3.2224, + "loss": 3.1448, "step": 1730 }, { "epoch": 0.18, - "grad_norm": 5.483465194702148, + "grad_norm": 25.2197208404541, "learning_rate": 1.963248495089239e-05, - "loss": 3.2481, + "loss": 3.3636, "step": 1740 }, { "epoch": 0.18, - "grad_norm": 3.4944567680358887, + "grad_norm": 22.630273818969727, "learning_rate": 1.9630372795437746e-05, - "loss": 3.1345, + "loss": 3.087, "step": 1750 }, { "epoch": 0.19, - "grad_norm": 4.70933198928833, + "grad_norm": 21.062307357788086, "learning_rate": 1.9628260639983103e-05, - "loss": 3.1668, + "loss": 3.1455, "step": 1760 }, { "epoch": 0.19, - "grad_norm": 4.7858805656433105, + "grad_norm": 29.949323654174805, "learning_rate": 1.9626148484528463e-05, - "loss": 3.2434, + "loss": 3.1407, "step": 1770 }, { "epoch": 0.19, - "grad_norm": 3.686572313308716, + "grad_norm": 24.480104446411133, "learning_rate": 1.962403632907382e-05, - "loss": 3.2327, + "loss": 3.0308, "step": 1780 }, { "epoch": 0.19, - "grad_norm": 3.8293943405151367, + "grad_norm": 23.0338077545166, "learning_rate": 1.962192417361918e-05, - "loss": 3.216, + "loss": 3.019, "step": 1790 }, { "epoch": 0.19, - "grad_norm": 4.163934230804443, + "grad_norm": 25.372461318969727, "learning_rate": 1.9619812018164538e-05, - "loss": 3.1366, + "loss": 3.0831, "step": 1800 }, { "epoch": 0.19, - "grad_norm": 3.9407880306243896, + "grad_norm": 26.883583068847656, "learning_rate": 1.9617699862709898e-05, - "loss": 3.2906, + "loss": 3.3081, "step": 1810 }, { "epoch": 0.19, - "grad_norm": 3.497360944747925, + "grad_norm": 22.77454376220703, "learning_rate": 1.9615587707255255e-05, - "loss": 3.2456, + "loss": 3.2617, "step": 1820 }, { "epoch": 0.19, - "grad_norm": 3.7981009483337402, + "grad_norm": 23.09379005432129, "learning_rate": 1.9613475551800612e-05, - "loss": 3.1809, + "loss": 3.1652, "step": 1830 }, { "epoch": 0.19, - "grad_norm": 4.289743900299072, + "grad_norm": 22.99510955810547, "learning_rate": 1.9611363396345973e-05, - "loss": 3.22, + "loss": 3.0605, "step": 1840 }, { "epoch": 0.2, - "grad_norm": 4.035226821899414, + "grad_norm": 23.03133201599121, "learning_rate": 1.960925124089133e-05, - "loss": 3.1941, + "loss": 3.2544, "step": 1850 }, { "epoch": 0.2, - "grad_norm": 5.164801597595215, + "grad_norm": 29.613941192626953, "learning_rate": 1.960713908543669e-05, - "loss": 3.1787, + "loss": 3.0559, "step": 1860 }, { "epoch": 0.2, - "grad_norm": 5.372908115386963, + "grad_norm": 27.462434768676758, "learning_rate": 1.9605026929982048e-05, - "loss": 3.0357, + "loss": 2.9359, "step": 1870 }, { "epoch": 0.2, - "grad_norm": 4.5744476318359375, + "grad_norm": 23.2030086517334, "learning_rate": 1.9602914774527405e-05, - "loss": 3.1259, + "loss": 3.2364, "step": 1880 }, { "epoch": 0.2, - "grad_norm": 5.122563362121582, + "grad_norm": 27.298410415649414, "learning_rate": 1.9600802619072765e-05, - "loss": 3.3255, + "loss": 3.3701, "step": 1890 }, { "epoch": 0.2, - "grad_norm": 4.811398506164551, + "grad_norm": 24.173133850097656, "learning_rate": 1.9598690463618126e-05, - "loss": 2.9727, + "loss": 2.8062, "step": 1900 }, { "epoch": 0.2, - "grad_norm": 5.184267044067383, + "grad_norm": 24.09808349609375, "learning_rate": 1.9596578308163483e-05, - "loss": 3.1784, + "loss": 3.2197, "step": 1910 }, { "epoch": 0.2, - "grad_norm": 5.715729236602783, + "grad_norm": 21.611618041992188, "learning_rate": 1.9594466152708843e-05, - "loss": 3.0974, + "loss": 3.0065, "step": 1920 }, { "epoch": 0.2, - "grad_norm": 5.522714614868164, + "grad_norm": 26.356563568115234, "learning_rate": 1.95923539972542e-05, - "loss": 3.2673, + "loss": 3.301, "step": 1930 }, { "epoch": 0.2, - "grad_norm": 4.847217559814453, + "grad_norm": 27.362728118896484, "learning_rate": 1.9590241841799557e-05, - "loss": 3.1419, + "loss": 3.281, "step": 1940 }, { "epoch": 0.21, - "grad_norm": 5.142326831817627, + "grad_norm": 28.08381462097168, "learning_rate": 1.9588129686344918e-05, - "loss": 3.2016, + "loss": 3.0013, "step": 1950 }, { "epoch": 0.21, - "grad_norm": 4.257440090179443, + "grad_norm": 21.970808029174805, "learning_rate": 1.9586017530890275e-05, - "loss": 3.1763, + "loss": 2.9383, "step": 1960 }, { "epoch": 0.21, - "grad_norm": 4.444497108459473, + "grad_norm": 24.749847412109375, "learning_rate": 1.9583905375435635e-05, - "loss": 3.1248, + "loss": 2.9882, "step": 1970 }, { "epoch": 0.21, - "grad_norm": 5.794654846191406, + "grad_norm": 28.738670349121094, "learning_rate": 1.9581793219980992e-05, - "loss": 3.2073, + "loss": 3.0797, "step": 1980 }, { "epoch": 0.21, - "grad_norm": 4.9274187088012695, + "grad_norm": 30.094505310058594, "learning_rate": 1.957968106452635e-05, - "loss": 3.0357, + "loss": 3.0918, "step": 1990 }, { "epoch": 0.21, - "grad_norm": 5.001906394958496, + "grad_norm": 26.640361785888672, "learning_rate": 1.957756890907171e-05, - "loss": 3.3313, + "loss": 3.224, "step": 2000 }, { "epoch": 0.21, - "grad_norm": 3.954094171524048, + "grad_norm": 26.76801300048828, "learning_rate": 1.9575456753617067e-05, - "loss": 3.0777, + "loss": 2.9837, "step": 2010 }, { "epoch": 0.21, - "grad_norm": 3.109386920928955, + "grad_norm": 21.326976776123047, "learning_rate": 1.9573344598162428e-05, - "loss": 3.0158, + "loss": 2.9424, "step": 2020 }, { "epoch": 0.21, - "grad_norm": 3.743936538696289, + "grad_norm": 24.486825942993164, "learning_rate": 1.9571232442707785e-05, - "loss": 3.013, + "loss": 3.0238, "step": 2030 }, { "epoch": 0.22, - "grad_norm": 5.1281819343566895, + "grad_norm": 21.025733947753906, "learning_rate": 1.9569120287253142e-05, - "loss": 3.2747, + "loss": 3.0635, "step": 2040 }, { "epoch": 0.22, - "grad_norm": 3.913463592529297, + "grad_norm": 25.33523178100586, "learning_rate": 1.9567008131798502e-05, - "loss": 3.0745, + "loss": 3.0711, "step": 2050 }, { "epoch": 0.22, - "grad_norm": 4.072035789489746, + "grad_norm": 26.079309463500977, "learning_rate": 1.956489597634386e-05, - "loss": 3.1604, + "loss": 2.8526, "step": 2060 }, { "epoch": 0.22, - "grad_norm": 4.895788192749023, + "grad_norm": 26.55131721496582, "learning_rate": 1.956278382088922e-05, - "loss": 2.8837, + "loss": 2.7081, "step": 2070 }, { "epoch": 0.22, - "grad_norm": 4.521069049835205, + "grad_norm": 33.120941162109375, "learning_rate": 1.956067166543458e-05, - "loss": 3.0212, + "loss": 3.07, "step": 2080 }, { "epoch": 0.22, - "grad_norm": 4.664191722869873, + "grad_norm": 24.445825576782227, "learning_rate": 1.9558559509979934e-05, - "loss": 3.0128, + "loss": 2.9877, "step": 2090 }, { "epoch": 0.22, - "grad_norm": 4.5710978507995605, + "grad_norm": 26.28845977783203, "learning_rate": 1.9556447354525294e-05, - "loss": 3.0129, + "loss": 2.8826, "step": 2100 }, { "epoch": 0.22, - "grad_norm": 4.403031349182129, + "grad_norm": 25.278268814086914, "learning_rate": 1.955433519907065e-05, - "loss": 3.1224, + "loss": 3.1773, "step": 2110 }, { "epoch": 0.22, - "grad_norm": 6.1742424964904785, + "grad_norm": 24.85350799560547, "learning_rate": 1.9552223043616012e-05, - "loss": 3.106, + "loss": 3.1822, "step": 2120 }, { "epoch": 0.22, - "grad_norm": 6.39127254486084, + "grad_norm": 28.381601333618164, "learning_rate": 1.9550110888161372e-05, - "loss": 2.8446, + "loss": 2.969, "step": 2130 }, { "epoch": 0.23, - "grad_norm": 4.200937747955322, + "grad_norm": 23.547828674316406, "learning_rate": 1.9547998732706726e-05, - "loss": 2.9985, + "loss": 2.9098, "step": 2140 }, { "epoch": 0.23, - "grad_norm": 5.786072254180908, + "grad_norm": 24.182050704956055, "learning_rate": 1.9545886577252087e-05, - "loss": 2.878, + "loss": 3.0793, "step": 2150 }, { "epoch": 0.23, - "grad_norm": 7.9164652824401855, + "grad_norm": 26.469045639038086, "learning_rate": 1.9543774421797447e-05, - "loss": 3.0469, + "loss": 2.98, "step": 2160 }, { "epoch": 0.23, - "grad_norm": 5.078585624694824, + "grad_norm": 27.616029739379883, "learning_rate": 1.9541662266342804e-05, - "loss": 3.0067, + "loss": 3.0167, "step": 2170 }, { "epoch": 0.23, - "grad_norm": 3.7193825244903564, + "grad_norm": 23.53097152709961, "learning_rate": 1.9539550110888165e-05, - "loss": 3.0468, + "loss": 2.8928, "step": 2180 }, { "epoch": 0.23, - "grad_norm": 6.0521979331970215, + "grad_norm": 27.990325927734375, "learning_rate": 1.953743795543352e-05, - "loss": 2.9561, + "loss": 2.8017, "step": 2190 }, { "epoch": 0.23, - "grad_norm": 5.472603797912598, + "grad_norm": 24.527050018310547, "learning_rate": 1.953532579997888e-05, - "loss": 2.8817, + "loss": 3.004, "step": 2200 }, { "epoch": 0.23, - "grad_norm": 4.855314254760742, + "grad_norm": 25.866886138916016, "learning_rate": 1.953321364452424e-05, - "loss": 2.913, + "loss": 2.8555, "step": 2210 }, { "epoch": 0.23, - "grad_norm": 5.176687240600586, + "grad_norm": 25.680828094482422, "learning_rate": 1.9531101489069596e-05, - "loss": 2.9994, + "loss": 2.9434, "step": 2220 }, { "epoch": 0.24, - "grad_norm": 3.652833938598633, + "grad_norm": 24.933086395263672, "learning_rate": 1.9528989333614957e-05, - "loss": 2.9823, + "loss": 3.0253, "step": 2230 }, { "epoch": 0.24, - "grad_norm": 6.007449150085449, + "grad_norm": 29.750579833984375, "learning_rate": 1.9526877178160314e-05, - "loss": 2.9488, + "loss": 2.8351, "step": 2240 }, { "epoch": 0.24, - "grad_norm": 4.300302505493164, + "grad_norm": 25.69093132019043, "learning_rate": 1.952476502270567e-05, - "loss": 2.8554, + "loss": 2.932, "step": 2250 }, { "epoch": 0.24, - "grad_norm": 3.3362135887145996, + "grad_norm": 20.288433074951172, "learning_rate": 1.952265286725103e-05, - "loss": 2.9066, + "loss": 2.7994, "step": 2260 }, { "epoch": 0.24, - "grad_norm": 5.818084239959717, + "grad_norm": 29.465280532836914, "learning_rate": 1.952054071179639e-05, - "loss": 2.8469, + "loss": 2.9288, "step": 2270 }, { "epoch": 0.24, - "grad_norm": 6.484394550323486, + "grad_norm": 22.6265926361084, "learning_rate": 1.951842855634175e-05, - "loss": 3.0636, + "loss": 3.0129, "step": 2280 }, { "epoch": 0.24, - "grad_norm": 2.9820761680603027, + "grad_norm": 26.969396591186523, "learning_rate": 1.9516316400887106e-05, - "loss": 2.994, + "loss": 2.9417, "step": 2290 }, { "epoch": 0.24, - "grad_norm": 5.652590274810791, + "grad_norm": 25.003822326660156, "learning_rate": 1.9514204245432467e-05, - "loss": 2.8655, + "loss": 2.8931, "step": 2300 }, { "epoch": 0.24, - "grad_norm": 4.712829113006592, + "grad_norm": 20.82132339477539, "learning_rate": 1.9512092089977824e-05, - "loss": 2.7257, + "loss": 2.651, "step": 2310 }, { "epoch": 0.25, - "grad_norm": 4.547958850860596, + "grad_norm": 26.930246353149414, "learning_rate": 1.950997993452318e-05, - "loss": 3.0164, + "loss": 3.1156, "step": 2320 }, { "epoch": 0.25, - "grad_norm": 3.891263961791992, + "grad_norm": 23.880403518676758, "learning_rate": 1.950786777906854e-05, - "loss": 2.89, + "loss": 2.8543, "step": 2330 }, { "epoch": 0.25, - "grad_norm": 3.8784871101379395, + "grad_norm": 20.525766372680664, "learning_rate": 1.95057556236139e-05, - "loss": 2.9746, + "loss": 3.0052, "step": 2340 }, { "epoch": 0.25, - "grad_norm": 4.459745407104492, + "grad_norm": 30.73211669921875, "learning_rate": 1.950364346815926e-05, - "loss": 2.8781, + "loss": 2.9273, "step": 2350 }, { "epoch": 0.25, - "grad_norm": 5.680963516235352, + "grad_norm": 22.85218048095703, "learning_rate": 1.9501531312704616e-05, - "loss": 2.9625, + "loss": 2.9663, "step": 2360 }, { "epoch": 0.25, - "grad_norm": 5.870044231414795, + "grad_norm": 19.72412872314453, "learning_rate": 1.9499419157249976e-05, - "loss": 2.8782, + "loss": 2.8127, "step": 2370 }, { "epoch": 0.25, - "grad_norm": 3.9679832458496094, + "grad_norm": 22.539670944213867, "learning_rate": 1.9497307001795333e-05, - "loss": 2.7614, + "loss": 2.7526, "step": 2380 }, { "epoch": 0.25, - "grad_norm": 4.740108966827393, + "grad_norm": 25.69571304321289, "learning_rate": 1.9495194846340694e-05, - "loss": 2.7904, + "loss": 2.7771, "step": 2390 }, { "epoch": 0.25, - "grad_norm": 5.298591136932373, + "grad_norm": 28.643152236938477, "learning_rate": 1.949308269088605e-05, - "loss": 2.9306, + "loss": 2.9427, "step": 2400 }, { "epoch": 0.25, - "grad_norm": 6.229260444641113, + "grad_norm": 26.257963180541992, "learning_rate": 1.9490970535431408e-05, - "loss": 2.9325, + "loss": 2.9674, "step": 2410 }, { "epoch": 0.26, - "grad_norm": 5.550780773162842, + "grad_norm": 26.103971481323242, "learning_rate": 1.948885837997677e-05, - "loss": 2.752, + "loss": 2.7164, "step": 2420 }, { "epoch": 0.26, - "grad_norm": 3.439368724822998, + "grad_norm": 23.683074951171875, "learning_rate": 1.9486746224522126e-05, - "loss": 2.8116, + "loss": 2.978, "step": 2430 }, { "epoch": 0.26, - "grad_norm": 5.177046775817871, + "grad_norm": 27.84854507446289, "learning_rate": 1.9484634069067486e-05, - "loss": 2.8068, + "loss": 2.7466, "step": 2440 }, { "epoch": 0.26, - "grad_norm": 4.153404712677002, + "grad_norm": 28.63791275024414, "learning_rate": 1.9482521913612843e-05, - "loss": 2.9056, + "loss": 3.0265, "step": 2450 }, { "epoch": 0.26, - "grad_norm": 4.520565986633301, + "grad_norm": 30.71712303161621, "learning_rate": 1.9480409758158204e-05, - "loss": 2.8715, + "loss": 2.8271, "step": 2460 }, { "epoch": 0.26, - "grad_norm": 6.328339099884033, + "grad_norm": 31.8747615814209, "learning_rate": 1.947829760270356e-05, - "loss": 2.7659, + "loss": 2.7443, "step": 2470 }, { "epoch": 0.26, - "grad_norm": 4.764604091644287, + "grad_norm": 29.52434730529785, "learning_rate": 1.9476185447248918e-05, - "loss": 2.8449, + "loss": 2.8806, "step": 2480 }, { "epoch": 0.26, - "grad_norm": 4.223827838897705, + "grad_norm": 25.367042541503906, "learning_rate": 1.9474073291794278e-05, - "loss": 2.881, + "loss": 2.9733, "step": 2490 }, { "epoch": 0.26, - "grad_norm": 5.409002780914307, + "grad_norm": 23.36689567565918, "learning_rate": 1.9471961136339635e-05, - "loss": 2.7226, + "loss": 2.7272, "step": 2500 }, { "epoch": 0.27, - "grad_norm": 4.332034587860107, + "grad_norm": 23.386768341064453, "learning_rate": 1.9469848980884996e-05, - "loss": 2.9311, + "loss": 2.9493, "step": 2510 }, { "epoch": 0.27, - "grad_norm": 5.562424659729004, + "grad_norm": 24.231420516967773, "learning_rate": 1.9467736825430353e-05, - "loss": 2.7992, + "loss": 2.9272, "step": 2520 }, { "epoch": 0.27, - "grad_norm": 4.402515888214111, + "grad_norm": 25.985071182250977, "learning_rate": 1.946562466997571e-05, - "loss": 2.713, + "loss": 2.8116, "step": 2530 }, { "epoch": 0.27, - "grad_norm": 4.798776149749756, + "grad_norm": 30.38374137878418, "learning_rate": 1.946351251452107e-05, - "loss": 2.5665, + "loss": 2.7382, "step": 2540 }, { "epoch": 0.27, - "grad_norm": 4.338649749755859, + "grad_norm": 25.97186851501465, "learning_rate": 1.9461400359066428e-05, - "loss": 2.6101, + "loss": 2.7377, "step": 2550 }, { "epoch": 0.27, - "grad_norm": 8.025331497192383, + "grad_norm": 22.865455627441406, "learning_rate": 1.9459288203611788e-05, - "loss": 2.5278, + "loss": 2.4615, "step": 2560 }, { "epoch": 0.27, - "grad_norm": 4.286011695861816, + "grad_norm": 37.63797378540039, "learning_rate": 1.945717604815715e-05, - "loss": 2.7463, + "loss": 2.7631, "step": 2570 }, { "epoch": 0.27, - "grad_norm": 6.171982765197754, + "grad_norm": 26.97764778137207, "learning_rate": 1.9455063892702502e-05, - "loss": 2.737, + "loss": 2.9278, "step": 2580 }, { "epoch": 0.27, - "grad_norm": 7.095573425292969, + "grad_norm": 26.37474250793457, "learning_rate": 1.9452951737247863e-05, - "loss": 2.6594, + "loss": 2.5898, "step": 2590 }, { "epoch": 0.27, - "grad_norm": 9.668744087219238, + "grad_norm": 30.51555633544922, "learning_rate": 1.9450839581793223e-05, - "loss": 2.8846, + "loss": 2.93, "step": 2600 }, { "epoch": 0.28, - "grad_norm": 7.699946403503418, + "grad_norm": 28.00543212890625, "learning_rate": 1.944872742633858e-05, - "loss": 2.8614, + "loss": 2.963, "step": 2610 }, { "epoch": 0.28, - "grad_norm": 4.679802417755127, + "grad_norm": 29.064790725708008, "learning_rate": 1.944661527088394e-05, - "loss": 2.8503, + "loss": 3.0481, "step": 2620 }, { "epoch": 0.28, - "grad_norm": 3.70316743850708, + "grad_norm": 31.546772003173828, "learning_rate": 1.9444503115429298e-05, - "loss": 2.5415, + "loss": 2.732, "step": 2630 }, { "epoch": 0.28, - "grad_norm": 6.045497417449951, + "grad_norm": 27.920631408691406, "learning_rate": 1.9442390959974655e-05, - "loss": 2.783, + "loss": 2.758, "step": 2640 }, { "epoch": 0.28, - "grad_norm": 5.22008752822876, + "grad_norm": 25.686893463134766, "learning_rate": 1.9440278804520015e-05, - "loss": 2.7466, + "loss": 2.7737, "step": 2650 }, { "epoch": 0.28, - "grad_norm": 6.39526891708374, + "grad_norm": 27.251657485961914, "learning_rate": 1.9438166649065372e-05, - "loss": 2.6195, + "loss": 2.55, "step": 2660 }, { "epoch": 0.28, - "grad_norm": 3.572736978530884, + "grad_norm": 27.634408950805664, "learning_rate": 1.9436054493610733e-05, - "loss": 2.7907, + "loss": 2.7678, "step": 2670 }, { "epoch": 0.28, - "grad_norm": 5.752501010894775, + "grad_norm": 25.403697967529297, "learning_rate": 1.943394233815609e-05, - "loss": 2.6171, + "loss": 2.6719, "step": 2680 }, { "epoch": 0.28, - "grad_norm": 4.677318096160889, + "grad_norm": 23.756832122802734, "learning_rate": 1.9431830182701447e-05, - "loss": 2.5986, + "loss": 2.7726, "step": 2690 }, { "epoch": 0.29, - "grad_norm": 6.603448390960693, + "grad_norm": 34.418697357177734, "learning_rate": 1.9429718027246808e-05, - "loss": 2.7656, + "loss": 2.873, "step": 2700 }, { "epoch": 0.29, - "grad_norm": 9.533599853515625, + "grad_norm": 27.89141845703125, "learning_rate": 1.9427605871792165e-05, - "loss": 2.7805, + "loss": 2.9126, "step": 2710 }, { "epoch": 0.29, - "grad_norm": 6.755883693695068, + "grad_norm": 24.525121688842773, "learning_rate": 1.9425493716337525e-05, - "loss": 2.6729, + "loss": 2.7732, "step": 2720 }, { "epoch": 0.29, - "grad_norm": 4.086599349975586, + "grad_norm": 28.72942352294922, "learning_rate": 1.9423381560882882e-05, - "loss": 2.6068, + "loss": 2.7645, "step": 2730 }, { "epoch": 0.29, - "grad_norm": 5.224214553833008, + "grad_norm": 26.070268630981445, "learning_rate": 1.942126940542824e-05, - "loss": 2.751, + "loss": 2.8361, "step": 2740 }, { "epoch": 0.29, - "grad_norm": 4.8430657386779785, + "grad_norm": 21.08803939819336, "learning_rate": 1.94191572499736e-05, - "loss": 2.6917, + "loss": 2.5611, "step": 2750 }, { "epoch": 0.29, - "grad_norm": 6.607713222503662, + "grad_norm": 22.418764114379883, "learning_rate": 1.9417045094518957e-05, - "loss": 2.7998, + "loss": 2.8652, "step": 2760 }, { "epoch": 0.29, - "grad_norm": 7.5369439125061035, + "grad_norm": 32.79839324951172, "learning_rate": 1.9414932939064317e-05, - "loss": 2.7309, + "loss": 2.964, "step": 2770 }, { "epoch": 0.29, - "grad_norm": 5.395843982696533, + "grad_norm": 26.272375106811523, "learning_rate": 1.9412820783609674e-05, - "loss": 2.7539, + "loss": 3.0982, "step": 2780 }, { "epoch": 0.29, - "grad_norm": 4.758197784423828, + "grad_norm": 25.197919845581055, "learning_rate": 1.941070862815503e-05, - "loss": 2.634, + "loss": 2.7154, "step": 2790 }, { "epoch": 0.3, - "grad_norm": 4.36293363571167, + "grad_norm": 30.387929916381836, "learning_rate": 1.9408596472700392e-05, - "loss": 2.6656, + "loss": 2.8524, "step": 2800 }, { "epoch": 0.3, - "grad_norm": 3.870196580886841, + "grad_norm": 26.983081817626953, "learning_rate": 1.940648431724575e-05, - "loss": 2.4362, + "loss": 2.5135, "step": 2810 }, { "epoch": 0.3, - "grad_norm": 7.093181133270264, + "grad_norm": 28.339792251586914, "learning_rate": 1.940437216179111e-05, - "loss": 2.697, + "loss": 2.7841, "step": 2820 }, { "epoch": 0.3, - "grad_norm": 6.077198505401611, + "grad_norm": 27.206825256347656, "learning_rate": 1.940226000633647e-05, - "loss": 2.619, + "loss": 2.8937, "step": 2830 }, { "epoch": 0.3, - "grad_norm": 4.49916934967041, + "grad_norm": 25.56400489807129, "learning_rate": 1.9400147850881827e-05, - "loss": 2.606, + "loss": 2.8945, "step": 2840 }, { "epoch": 0.3, - "grad_norm": 3.1435861587524414, + "grad_norm": 19.184999465942383, "learning_rate": 1.9398035695427184e-05, - "loss": 2.3294, + "loss": 2.4283, "step": 2850 }, { "epoch": 0.3, - "grad_norm": 8.014472961425781, + "grad_norm": 28.044363021850586, "learning_rate": 1.9395923539972545e-05, - "loss": 2.4942, + "loss": 2.8557, "step": 2860 }, { "epoch": 0.3, - "grad_norm": 5.1639838218688965, + "grad_norm": 24.07820701599121, "learning_rate": 1.9393811384517902e-05, - "loss": 2.5715, + "loss": 2.5789, "step": 2870 }, { "epoch": 0.3, - "grad_norm": 5.906452178955078, + "grad_norm": 28.872379302978516, "learning_rate": 1.9391699229063262e-05, - "loss": 2.4974, + "loss": 2.939, "step": 2880 }, { "epoch": 0.31, - "grad_norm": 5.504446983337402, + "grad_norm": 26.065088272094727, "learning_rate": 1.938958707360862e-05, - "loss": 2.623, + "loss": 2.8831, "step": 2890 }, { "epoch": 0.31, - "grad_norm": 3.6827456951141357, + "grad_norm": 23.562353134155273, "learning_rate": 1.9387474918153976e-05, - "loss": 2.4611, + "loss": 2.5463, "step": 2900 }, { "epoch": 0.31, - "grad_norm": 5.457493305206299, + "grad_norm": 25.539379119873047, "learning_rate": 1.9385362762699337e-05, - "loss": 2.6312, + "loss": 2.905, "step": 2910 }, { "epoch": 0.31, - "grad_norm": 4.52791166305542, + "grad_norm": 25.513336181640625, "learning_rate": 1.9383250607244694e-05, - "loss": 2.6083, + "loss": 2.8167, "step": 2920 }, { "epoch": 0.31, - "grad_norm": 6.674107551574707, + "grad_norm": 29.299272537231445, "learning_rate": 1.9381138451790054e-05, - "loss": 2.6579, + "loss": 2.7506, "step": 2930 }, { "epoch": 0.31, - "grad_norm": 7.559858798980713, + "grad_norm": 33.29500198364258, "learning_rate": 1.937902629633541e-05, - "loss": 2.6284, + "loss": 2.7814, "step": 2940 }, { "epoch": 0.31, - "grad_norm": 6.225968360900879, + "grad_norm": 24.499834060668945, "learning_rate": 1.9376914140880772e-05, - "loss": 2.5785, + "loss": 2.5442, "step": 2950 }, { "epoch": 0.31, - "grad_norm": 7.580895900726318, + "grad_norm": 34.200191497802734, "learning_rate": 1.937480198542613e-05, - "loss": 2.5215, + "loss": 2.633, "step": 2960 }, { "epoch": 0.31, - "grad_norm": 7.3528828620910645, + "grad_norm": 28.674549102783203, "learning_rate": 1.9372689829971486e-05, - "loss": 2.5657, + "loss": 2.6969, "step": 2970 }, { "epoch": 0.31, - "grad_norm": 5.506893157958984, + "grad_norm": 29.991575241088867, "learning_rate": 1.9370577674516847e-05, - "loss": 2.4385, + "loss": 2.5741, "step": 2980 }, { "epoch": 0.32, - "grad_norm": 6.24576997756958, + "grad_norm": 31.125850677490234, "learning_rate": 1.9368465519062204e-05, - "loss": 2.5891, + "loss": 2.6995, "step": 2990 }, { "epoch": 0.32, - "grad_norm": 5.258028030395508, + "grad_norm": 24.432714462280273, "learning_rate": 1.9366353363607564e-05, - "loss": 2.5968, + "loss": 2.6727, "step": 3000 }, { "epoch": 0.32, - "grad_norm": 8.957871437072754, + "grad_norm": 24.16388511657715, "learning_rate": 1.936424120815292e-05, - "loss": 2.5052, + "loss": 2.4474, "step": 3010 }, { "epoch": 0.32, - "grad_norm": 6.982890605926514, + "grad_norm": 30.644939422607422, "learning_rate": 1.936212905269828e-05, - "loss": 2.4994, + "loss": 2.7787, "step": 3020 }, { "epoch": 0.32, - "grad_norm": 5.451074123382568, + "grad_norm": 24.37651824951172, "learning_rate": 1.936001689724364e-05, - "loss": 2.6022, + "loss": 2.6902, "step": 3030 }, { "epoch": 0.32, - "grad_norm": 5.966454982757568, + "grad_norm": 29.367937088012695, "learning_rate": 1.9357904741789e-05, - "loss": 2.6259, + "loss": 2.8191, "step": 3040 }, { "epoch": 0.32, - "grad_norm": 5.691244125366211, + "grad_norm": 27.712074279785156, "learning_rate": 1.9355792586334356e-05, - "loss": 2.3706, + "loss": 2.6193, "step": 3050 }, { "epoch": 0.32, - "grad_norm": 7.9246673583984375, + "grad_norm": 23.955646514892578, "learning_rate": 1.9353680430879713e-05, - "loss": 2.6933, + "loss": 2.8812, "step": 3060 }, { "epoch": 0.32, - "grad_norm": 6.11812162399292, + "grad_norm": 30.951099395751953, "learning_rate": 1.9351568275425074e-05, - "loss": 2.4404, + "loss": 2.6805, "step": 3070 }, { "epoch": 0.33, - "grad_norm": 6.285652160644531, + "grad_norm": 27.928335189819336, "learning_rate": 1.934945611997043e-05, - "loss": 2.421, + "loss": 2.7529, "step": 3080 }, { "epoch": 0.33, - "grad_norm": 5.921961784362793, + "grad_norm": 24.58074951171875, "learning_rate": 1.934734396451579e-05, - "loss": 2.4235, + "loss": 2.4482, "step": 3090 }, { "epoch": 0.33, - "grad_norm": 4.12562894821167, + "grad_norm": 26.643075942993164, "learning_rate": 1.934523180906115e-05, - "loss": 2.5525, + "loss": 2.7492, "step": 3100 }, { "epoch": 0.33, - "grad_norm": 6.2871246337890625, + "grad_norm": 22.69312286376953, "learning_rate": 1.934311965360651e-05, - "loss": 2.3694, + "loss": 2.4701, "step": 3110 }, { "epoch": 0.33, - "grad_norm": 5.6484527587890625, + "grad_norm": 26.3702449798584, "learning_rate": 1.9341007498151866e-05, - "loss": 2.4381, + "loss": 2.7203, "step": 3120 }, { "epoch": 0.33, - "grad_norm": 4.3793230056762695, + "grad_norm": 29.325719833374023, "learning_rate": 1.9338895342697223e-05, - "loss": 2.4049, + "loss": 2.8317, "step": 3130 }, { "epoch": 0.33, - "grad_norm": 5.636318206787109, + "grad_norm": 25.313278198242188, "learning_rate": 1.9336783187242584e-05, - "loss": 2.4396, + "loss": 2.5672, "step": 3140 }, { "epoch": 0.33, - "grad_norm": 9.513223648071289, + "grad_norm": 27.127655029296875, "learning_rate": 1.933467103178794e-05, - "loss": 2.2565, + "loss": 2.4747, "step": 3150 }, { "epoch": 0.33, - "grad_norm": 7.083193778991699, + "grad_norm": 33.318138122558594, "learning_rate": 1.93325588763333e-05, - "loss": 2.4965, + "loss": 2.7326, "step": 3160 }, { "epoch": 0.33, - "grad_norm": 4.512002468109131, + "grad_norm": 27.136938095092773, "learning_rate": 1.9330446720878658e-05, - "loss": 2.5355, + "loss": 2.9595, "step": 3170 }, { "epoch": 0.34, - "grad_norm": 3.6300320625305176, + "grad_norm": 24.297178268432617, "learning_rate": 1.9328334565424015e-05, - "loss": 2.2292, + "loss": 2.4867, "step": 3180 }, { "epoch": 0.34, - "grad_norm": 4.826886177062988, + "grad_norm": 29.28363800048828, "learning_rate": 1.9326222409969376e-05, - "loss": 2.5045, + "loss": 2.5933, "step": 3190 }, { "epoch": 0.34, - "grad_norm": 4.924584865570068, + "grad_norm": 29.27230453491211, "learning_rate": 1.9324110254514733e-05, - "loss": 2.3346, + "loss": 2.7128, "step": 3200 }, { "epoch": 0.34, - "grad_norm": 7.137544631958008, + "grad_norm": 26.878517150878906, "learning_rate": 1.9321998099060093e-05, - "loss": 2.5685, + "loss": 2.8205, "step": 3210 }, { "epoch": 0.34, - "grad_norm": 5.48757266998291, + "grad_norm": 21.848495483398438, "learning_rate": 1.931988594360545e-05, - "loss": 2.4144, + "loss": 2.5891, "step": 3220 }, { "epoch": 0.34, - "grad_norm": 5.578709125518799, + "grad_norm": 28.93549346923828, "learning_rate": 1.9317773788150808e-05, - "loss": 2.5125, + "loss": 2.8435, "step": 3230 }, { "epoch": 0.34, - "grad_norm": 9.157731056213379, + "grad_norm": 25.873096466064453, "learning_rate": 1.9315661632696168e-05, - "loss": 2.376, + "loss": 2.6267, "step": 3240 }, { "epoch": 0.34, - "grad_norm": 4.295504093170166, + "grad_norm": 24.10552215576172, "learning_rate": 1.9313549477241525e-05, - "loss": 2.3568, + "loss": 2.4082, "step": 3250 }, { "epoch": 0.34, - "grad_norm": 4.8656792640686035, + "grad_norm": 27.333454132080078, "learning_rate": 1.9311437321786886e-05, - "loss": 2.2896, + "loss": 2.4715, "step": 3260 }, { "epoch": 0.35, - "grad_norm": 5.119721412658691, + "grad_norm": 22.286895751953125, "learning_rate": 1.9309325166332246e-05, - "loss": 2.3045, + "loss": 2.5393, "step": 3270 }, { "epoch": 0.35, - "grad_norm": 5.220632076263428, + "grad_norm": 26.58277702331543, "learning_rate": 1.93072130108776e-05, - "loss": 2.3933, + "loss": 2.736, "step": 3280 }, { "epoch": 0.35, - "grad_norm": 7.606267929077148, + "grad_norm": 24.034582138061523, "learning_rate": 1.930510085542296e-05, - "loss": 2.3994, + "loss": 2.5014, "step": 3290 }, { "epoch": 0.35, - "grad_norm": 5.019437789916992, + "grad_norm": 30.65144157409668, "learning_rate": 1.930298869996832e-05, - "loss": 2.3384, + "loss": 2.644, "step": 3300 }, { "epoch": 0.35, - "grad_norm": 11.224194526672363, + "grad_norm": 32.762386322021484, "learning_rate": 1.9300876544513678e-05, - "loss": 2.4528, + "loss": 2.7809, "step": 3310 }, { "epoch": 0.35, - "grad_norm": 7.513612270355225, + "grad_norm": 29.77081871032715, "learning_rate": 1.9298764389059038e-05, - "loss": 2.4715, + "loss": 2.8078, "step": 3320 }, { "epoch": 0.35, - "grad_norm": 6.775038719177246, + "grad_norm": 27.09537124633789, "learning_rate": 1.9296652233604395e-05, - "loss": 2.3335, + "loss": 2.5293, "step": 3330 }, { "epoch": 0.35, - "grad_norm": 7.539262771606445, + "grad_norm": 30.0086669921875, "learning_rate": 1.9294540078149752e-05, - "loss": 2.4369, + "loss": 2.7537, "step": 3340 }, { "epoch": 0.35, - "grad_norm": 5.811434268951416, + "grad_norm": 28.00872802734375, "learning_rate": 1.9292427922695113e-05, - "loss": 2.5424, + "loss": 2.7977, "step": 3350 }, { "epoch": 0.35, - "grad_norm": 5.065423011779785, + "grad_norm": 33.64259719848633, "learning_rate": 1.929031576724047e-05, - "loss": 2.178, + "loss": 2.3885, "step": 3360 }, { "epoch": 0.36, - "grad_norm": 4.123407363891602, + "grad_norm": 30.414173126220703, "learning_rate": 1.928820361178583e-05, - "loss": 2.3525, + "loss": 2.564, "step": 3370 }, { "epoch": 0.36, - "grad_norm": 5.1820478439331055, + "grad_norm": 27.63892364501953, "learning_rate": 1.9286091456331188e-05, - "loss": 2.2527, + "loss": 2.7315, "step": 3380 }, { "epoch": 0.36, - "grad_norm": 5.239874839782715, + "grad_norm": 24.281713485717773, "learning_rate": 1.9283979300876545e-05, - "loss": 2.1522, + "loss": 2.3184, "step": 3390 }, { "epoch": 0.36, - "grad_norm": 6.222786903381348, + "grad_norm": 26.023122787475586, "learning_rate": 1.9281867145421905e-05, - "loss": 2.5187, + "loss": 2.7846, "step": 3400 }, { "epoch": 0.36, - "grad_norm": 4.797713279724121, + "grad_norm": 22.43453025817871, "learning_rate": 1.9279754989967262e-05, - "loss": 2.2595, + "loss": 2.4611, "step": 3410 }, { "epoch": 0.36, - "grad_norm": 4.591484069824219, + "grad_norm": 25.51792335510254, "learning_rate": 1.9277642834512623e-05, - "loss": 2.1416, + "loss": 2.5799, "step": 3420 }, { "epoch": 0.36, - "grad_norm": 9.847600936889648, + "grad_norm": 27.628002166748047, "learning_rate": 1.927553067905798e-05, - "loss": 2.6109, + "loss": 2.8941, "step": 3430 }, { "epoch": 0.36, - "grad_norm": 3.759247303009033, + "grad_norm": 25.622026443481445, "learning_rate": 1.9273418523603337e-05, - "loss": 2.199, + "loss": 2.2651, "step": 3440 }, { "epoch": 0.36, - "grad_norm": 7.725421905517578, + "grad_norm": 33.30426025390625, "learning_rate": 1.9271306368148697e-05, - "loss": 2.2706, + "loss": 2.7614, "step": 3450 }, { "epoch": 0.37, - "grad_norm": 6.968447208404541, + "grad_norm": 21.046192169189453, "learning_rate": 1.9269194212694054e-05, - "loss": 2.2169, + "loss": 2.3344, "step": 3460 }, { "epoch": 0.37, - "grad_norm": 8.801129341125488, + "grad_norm": 31.95816993713379, "learning_rate": 1.9267082057239415e-05, - "loss": 2.2219, + "loss": 2.3136, "step": 3470 }, { "epoch": 0.37, - "grad_norm": 9.27863883972168, + "grad_norm": 27.277236938476562, "learning_rate": 1.9264969901784772e-05, - "loss": 2.3425, + "loss": 2.6167, "step": 3480 }, { "epoch": 0.37, - "grad_norm": 6.862630367279053, + "grad_norm": 21.03985595703125, "learning_rate": 1.9262857746330132e-05, - "loss": 2.1677, + "loss": 2.282, "step": 3490 }, { "epoch": 0.37, - "grad_norm": 6.714095115661621, + "grad_norm": 27.89251708984375, "learning_rate": 1.926074559087549e-05, - "loss": 2.473, + "loss": 2.7821, "step": 3500 }, { "epoch": 0.37, - "grad_norm": 5.2625226974487305, + "grad_norm": 30.769426345825195, "learning_rate": 1.9258633435420847e-05, - "loss": 2.3074, + "loss": 2.5955, "step": 3510 }, { "epoch": 0.37, - "grad_norm": 5.901721477508545, + "grad_norm": 23.69078826904297, "learning_rate": 1.9256521279966207e-05, - "loss": 2.4102, + "loss": 2.5016, "step": 3520 }, { "epoch": 0.37, - "grad_norm": 5.631887435913086, + "grad_norm": 23.94914436340332, "learning_rate": 1.9254409124511568e-05, - "loss": 2.1562, + "loss": 2.3756, "step": 3530 }, { "epoch": 0.37, - "grad_norm": 5.705519199371338, + "grad_norm": 23.33180809020996, "learning_rate": 1.9252296969056925e-05, - "loss": 2.1959, + "loss": 2.2586, "step": 3540 }, { "epoch": 0.37, - "grad_norm": 4.739804267883301, + "grad_norm": 28.994972229003906, "learning_rate": 1.9250184813602282e-05, - "loss": 2.3567, + "loss": 2.4819, "step": 3550 }, { "epoch": 0.38, - "grad_norm": 5.404460430145264, + "grad_norm": 35.72193145751953, "learning_rate": 1.9248072658147642e-05, - "loss": 2.1272, + "loss": 2.5605, "step": 3560 }, { "epoch": 0.38, - "grad_norm": 7.483436584472656, + "grad_norm": 28.35190773010254, "learning_rate": 1.9245960502693e-05, - "loss": 2.4076, + "loss": 2.6154, "step": 3570 }, { "epoch": 0.38, - "grad_norm": 6.942784786224365, + "grad_norm": 29.706205368041992, "learning_rate": 1.924384834723836e-05, - "loss": 2.2358, + "loss": 2.4051, "step": 3580 }, { "epoch": 0.38, - "grad_norm": 8.21924877166748, + "grad_norm": 29.800739288330078, "learning_rate": 1.9241736191783717e-05, - "loss": 2.2517, + "loss": 2.5257, "step": 3590 }, { "epoch": 0.38, - "grad_norm": 9.679084777832031, + "grad_norm": 25.12029457092285, "learning_rate": 1.9239624036329077e-05, - "loss": 2.2801, + "loss": 2.6129, "step": 3600 }, { "epoch": 0.38, - "grad_norm": 3.6439995765686035, + "grad_norm": 23.485294342041016, "learning_rate": 1.9237511880874434e-05, - "loss": 2.3082, + "loss": 2.6027, "step": 3610 }, { "epoch": 0.38, - "grad_norm": 3.1781134605407715, + "grad_norm": 32.83793258666992, "learning_rate": 1.923539972541979e-05, - "loss": 2.3307, + "loss": 2.6209, "step": 3620 }, { "epoch": 0.38, - "grad_norm": 7.436091423034668, + "grad_norm": 30.307437896728516, "learning_rate": 1.9233287569965152e-05, - "loss": 2.1094, + "loss": 2.4552, "step": 3630 }, { "epoch": 0.38, - "grad_norm": 18.937091827392578, + "grad_norm": 30.096006393432617, "learning_rate": 1.923117541451051e-05, - "loss": 2.1518, + "loss": 2.5375, "step": 3640 }, { "epoch": 0.39, - "grad_norm": 5.316178321838379, + "grad_norm": 31.441408157348633, "learning_rate": 1.922906325905587e-05, - "loss": 2.0638, + "loss": 2.5068, "step": 3650 }, { "epoch": 0.39, - "grad_norm": 7.4336700439453125, + "grad_norm": 27.81244659423828, "learning_rate": 1.9226951103601227e-05, - "loss": 2.3249, + "loss": 2.6417, "step": 3660 }, { "epoch": 0.39, - "grad_norm": 7.045809745788574, + "grad_norm": 36.802215576171875, "learning_rate": 1.9224838948146584e-05, - "loss": 2.2652, + "loss": 2.5133, "step": 3670 }, { "epoch": 0.39, - "grad_norm": 5.593904972076416, + "grad_norm": 22.174861907958984, "learning_rate": 1.9222726792691944e-05, - "loss": 2.1174, + "loss": 2.4814, "step": 3680 }, { "epoch": 0.39, - "grad_norm": 5.825852394104004, + "grad_norm": 26.21978187561035, "learning_rate": 1.92206146372373e-05, - "loss": 2.1879, + "loss": 2.3771, "step": 3690 }, { "epoch": 0.39, - "grad_norm": 5.812777042388916, + "grad_norm": 22.302827835083008, "learning_rate": 1.9218502481782662e-05, - "loss": 2.0999, + "loss": 2.5071, "step": 3700 }, { "epoch": 0.39, - "grad_norm": 5.4384965896606445, + "grad_norm": 25.120237350463867, "learning_rate": 1.921639032632802e-05, - "loss": 2.3798, + "loss": 2.4767, "step": 3710 }, { "epoch": 0.39, - "grad_norm": 6.923451900482178, + "grad_norm": 33.724918365478516, "learning_rate": 1.9214278170873376e-05, - "loss": 2.3407, + "loss": 2.7187, "step": 3720 }, { "epoch": 0.39, - "grad_norm": 7.181000232696533, + "grad_norm": 26.726594924926758, "learning_rate": 1.9212166015418736e-05, - "loss": 2.0428, + "loss": 2.5466, "step": 3730 }, { "epoch": 0.39, - "grad_norm": 6.269800186157227, + "grad_norm": 24.898719787597656, "learning_rate": 1.9210053859964093e-05, - "loss": 2.0233, + "loss": 2.437, "step": 3740 }, { "epoch": 0.4, - "grad_norm": 7.2651753425598145, + "grad_norm": 24.882509231567383, "learning_rate": 1.9207941704509454e-05, - "loss": 2.3031, + "loss": 2.5057, "step": 3750 }, { "epoch": 0.4, - "grad_norm": 12.752398490905762, + "grad_norm": 33.09433364868164, "learning_rate": 1.9205829549054814e-05, - "loss": 2.3358, + "loss": 2.7428, "step": 3760 }, { "epoch": 0.4, - "grad_norm": 8.212806701660156, + "grad_norm": 26.47796058654785, "learning_rate": 1.9203717393600168e-05, - "loss": 2.0839, + "loss": 2.399, "step": 3770 }, { "epoch": 0.4, - "grad_norm": 7.831247806549072, + "grad_norm": 26.91855812072754, "learning_rate": 1.920160523814553e-05, - "loss": 2.29, + "loss": 2.4939, "step": 3780 }, { "epoch": 0.4, - "grad_norm": 3.8878300189971924, + "grad_norm": 29.641740798950195, "learning_rate": 1.919949308269089e-05, - "loss": 2.1777, + "loss": 2.4715, "step": 3790 }, { "epoch": 0.4, - "grad_norm": 6.761434078216553, + "grad_norm": 26.696149826049805, "learning_rate": 1.9197380927236246e-05, - "loss": 2.2302, + "loss": 2.2856, "step": 3800 }, { "epoch": 0.4, - "grad_norm": 7.88983678817749, + "grad_norm": 27.491546630859375, "learning_rate": 1.9195268771781607e-05, - "loss": 2.234, + "loss": 2.5435, "step": 3810 }, { "epoch": 0.4, - "grad_norm": 7.8639397621154785, + "grad_norm": 33.638145446777344, "learning_rate": 1.9193156616326964e-05, - "loss": 2.1034, + "loss": 2.4337, "step": 3820 }, { "epoch": 0.4, - "grad_norm": 7.9432902336120605, + "grad_norm": 31.342668533325195, "learning_rate": 1.919104446087232e-05, - "loss": 2.3207, + "loss": 2.7323, "step": 3830 }, { "epoch": 0.41, - "grad_norm": 6.308475017547607, + "grad_norm": 22.77524757385254, "learning_rate": 1.918893230541768e-05, - "loss": 2.3649, + "loss": 2.7671, "step": 3840 }, { "epoch": 0.41, - "grad_norm": 4.932580471038818, + "grad_norm": 28.22163963317871, "learning_rate": 1.918682014996304e-05, - "loss": 2.2372, + "loss": 2.569, "step": 3850 }, { "epoch": 0.41, - "grad_norm": 7.332131385803223, + "grad_norm": 27.287769317626953, "learning_rate": 1.91847079945084e-05, - "loss": 2.1307, + "loss": 2.643, "step": 3860 }, { "epoch": 0.41, - "grad_norm": 4.123221397399902, + "grad_norm": 31.497875213623047, "learning_rate": 1.9182595839053756e-05, - "loss": 2.0696, + "loss": 2.3988, "step": 3870 }, { "epoch": 0.41, - "grad_norm": 4.999847412109375, + "grad_norm": 25.2366943359375, "learning_rate": 1.9180483683599113e-05, - "loss": 2.2464, + "loss": 2.4178, "step": 3880 }, { "epoch": 0.41, - "grad_norm": 5.174808502197266, + "grad_norm": 26.475698471069336, "learning_rate": 1.9178371528144473e-05, - "loss": 1.9537, + "loss": 2.3875, "step": 3890 }, { "epoch": 0.41, - "grad_norm": 8.811868667602539, + "grad_norm": 26.35909652709961, "learning_rate": 1.917625937268983e-05, - "loss": 2.0413, + "loss": 2.6999, "step": 3900 }, { "epoch": 0.41, - "grad_norm": 13.762826919555664, + "grad_norm": 27.042919158935547, "learning_rate": 1.917414721723519e-05, - "loss": 2.0425, + "loss": 2.7185, "step": 3910 }, { "epoch": 0.41, - "grad_norm": 8.897381782531738, + "grad_norm": 29.32121467590332, "learning_rate": 1.9172035061780548e-05, - "loss": 1.8736, + "loss": 2.2478, "step": 3920 }, { "epoch": 0.42, - "grad_norm": 9.371459007263184, + "grad_norm": 33.55501174926758, "learning_rate": 1.9169922906325905e-05, - "loss": 2.1894, + "loss": 2.514, "step": 3930 }, { "epoch": 0.42, - "grad_norm": 9.8627290725708, + "grad_norm": 24.630163192749023, "learning_rate": 1.9167810750871266e-05, - "loss": 2.1999, + "loss": 2.5009, "step": 3940 }, { "epoch": 0.42, - "grad_norm": 6.532742977142334, + "grad_norm": 32.19506072998047, "learning_rate": 1.9165698595416623e-05, - "loss": 2.1232, + "loss": 2.5945, "step": 3950 }, { "epoch": 0.42, - "grad_norm": 6.146297454833984, + "grad_norm": 23.303617477416992, "learning_rate": 1.9163586439961983e-05, - "loss": 1.7468, + "loss": 2.1306, "step": 3960 }, { "epoch": 0.42, - "grad_norm": 5.4458513259887695, + "grad_norm": 29.1339054107666, "learning_rate": 1.9161474284507344e-05, - "loss": 2.1433, + "loss": 2.3663, "step": 3970 }, { "epoch": 0.42, - "grad_norm": 4.583361625671387, + "grad_norm": 20.889320373535156, "learning_rate": 1.9159362129052697e-05, - "loss": 2.0148, + "loss": 2.3428, "step": 3980 }, { "epoch": 0.42, - "grad_norm": 5.6412672996521, + "grad_norm": 27.47104835510254, "learning_rate": 1.9157249973598058e-05, - "loss": 2.1523, + "loss": 2.5636, "step": 3990 }, { "epoch": 0.42, - "grad_norm": 7.54428243637085, + "grad_norm": 32.191226959228516, "learning_rate": 1.915513781814342e-05, - "loss": 1.9952, + "loss": 2.3133, "step": 4000 }, { "epoch": 0.42, - "grad_norm": 5.35874605178833, + "grad_norm": 23.93500328063965, "learning_rate": 1.9153025662688775e-05, - "loss": 2.2928, + "loss": 2.5868, "step": 4010 }, { "epoch": 0.42, - "grad_norm": 6.038791656494141, + "grad_norm": 29.406118392944336, "learning_rate": 1.9150913507234136e-05, - "loss": 2.0422, + "loss": 2.6351, "step": 4020 }, { "epoch": 0.43, - "grad_norm": 7.51292610168457, + "grad_norm": 30.397178649902344, "learning_rate": 1.9148801351779493e-05, - "loss": 2.2304, + "loss": 2.4912, "step": 4030 }, { "epoch": 0.43, - "grad_norm": 6.6638078689575195, + "grad_norm": 28.692279815673828, "learning_rate": 1.914668919632485e-05, - "loss": 2.2487, + "loss": 2.5579, "step": 4040 }, { "epoch": 0.43, - "grad_norm": 6.3260345458984375, + "grad_norm": 30.880273818969727, "learning_rate": 1.914457704087021e-05, - "loss": 2.286, + "loss": 2.6082, "step": 4050 }, { "epoch": 0.43, - "grad_norm": 11.657694816589355, + "grad_norm": 32.8739128112793, "learning_rate": 1.9142464885415568e-05, - "loss": 2.0576, + "loss": 2.4275, "step": 4060 }, { "epoch": 0.43, - "grad_norm": 6.073609352111816, + "grad_norm": 32.36878204345703, "learning_rate": 1.9140352729960928e-05, - "loss": 2.0916, + "loss": 2.518, "step": 4070 }, { "epoch": 0.43, - "grad_norm": 4.675407409667969, + "grad_norm": 22.36184310913086, "learning_rate": 1.9138240574506285e-05, - "loss": 2.0644, + "loss": 2.2239, "step": 4080 }, { "epoch": 0.43, - "grad_norm": 6.291007995605469, + "grad_norm": 31.08745002746582, "learning_rate": 1.9136128419051642e-05, - "loss": 1.8922, + "loss": 2.42, "step": 4090 }, { "epoch": 0.43, - "grad_norm": 9.073143005371094, + "grad_norm": 30.957063674926758, "learning_rate": 1.9134016263597003e-05, - "loss": 2.1176, + "loss": 2.631, "step": 4100 }, { "epoch": 0.43, - "grad_norm": 6.054325103759766, + "grad_norm": 30.166912078857422, "learning_rate": 1.913190410814236e-05, - "loss": 2.0279, + "loss": 2.2521, "step": 4110 }, { "epoch": 0.44, - "grad_norm": 6.587664604187012, + "grad_norm": 24.345975875854492, "learning_rate": 1.912979195268772e-05, - "loss": 2.0323, + "loss": 2.1223, "step": 4120 }, { "epoch": 0.44, - "grad_norm": 6.459173679351807, + "grad_norm": 28.648488998413086, "learning_rate": 1.9127679797233077e-05, - "loss": 1.9092, + "loss": 2.5278, "step": 4130 }, { "epoch": 0.44, - "grad_norm": 7.223755836486816, + "grad_norm": 43.785247802734375, "learning_rate": 1.9125567641778438e-05, - "loss": 1.9494, + "loss": 2.3592, "step": 4140 }, { "epoch": 0.44, - "grad_norm": 6.769535541534424, + "grad_norm": 35.41561508178711, "learning_rate": 1.9123455486323795e-05, - "loss": 1.8706, + "loss": 2.3971, "step": 4150 }, { "epoch": 0.44, - "grad_norm": 5.110337257385254, + "grad_norm": 27.789094924926758, "learning_rate": 1.9121343330869152e-05, - "loss": 2.0168, + "loss": 2.5946, "step": 4160 }, { "epoch": 0.44, - "grad_norm": 5.116240501403809, + "grad_norm": 26.909074783325195, "learning_rate": 1.9119231175414512e-05, - "loss": 1.8819, + "loss": 2.2381, "step": 4170 }, { "epoch": 0.44, - "grad_norm": 7.7864484786987305, + "grad_norm": 27.876089096069336, "learning_rate": 1.911711901995987e-05, - "loss": 2.2111, + "loss": 2.7411, "step": 4180 }, { "epoch": 0.44, - "grad_norm": 4.523825168609619, + "grad_norm": 25.05436134338379, "learning_rate": 1.911500686450523e-05, - "loss": 1.9263, + "loss": 2.283, "step": 4190 }, { "epoch": 0.44, - "grad_norm": 4.5936174392700195, + "grad_norm": 28.68523597717285, "learning_rate": 1.9112894709050587e-05, - "loss": 1.8345, + "loss": 2.3182, "step": 4200 }, { "epoch": 0.44, - "grad_norm": 2.616642713546753, + "grad_norm": 18.335264205932617, "learning_rate": 1.9110782553595944e-05, - "loss": 1.8214, + "loss": 2.1869, "step": 4210 }, { "epoch": 0.45, - "grad_norm": 7.44563102722168, + "grad_norm": 32.738739013671875, "learning_rate": 1.9108670398141305e-05, - "loss": 2.0498, + "loss": 2.4931, "step": 4220 }, { "epoch": 0.45, - "grad_norm": 7.209227085113525, + "grad_norm": 23.99049186706543, "learning_rate": 1.9106558242686665e-05, - "loss": 1.9749, + "loss": 2.3356, "step": 4230 }, { "epoch": 0.45, - "grad_norm": 10.636545181274414, + "grad_norm": 30.632375717163086, "learning_rate": 1.9104446087232022e-05, - "loss": 1.9405, + "loss": 2.3786, "step": 4240 }, { "epoch": 0.45, - "grad_norm": 10.292664527893066, + "grad_norm": 36.41792678833008, "learning_rate": 1.910233393177738e-05, - "loss": 1.9934, + "loss": 2.4983, "step": 4250 }, { "epoch": 0.45, - "grad_norm": 5.495169639587402, + "grad_norm": 33.969940185546875, "learning_rate": 1.910022177632274e-05, - "loss": 1.9072, + "loss": 2.4271, "step": 4260 }, { "epoch": 0.45, - "grad_norm": 6.346231937408447, + "grad_norm": 28.5601749420166, "learning_rate": 1.9098109620868097e-05, - "loss": 1.9717, + "loss": 2.6351, "step": 4270 }, { "epoch": 0.45, - "grad_norm": 5.661221504211426, + "grad_norm": 24.15825843811035, "learning_rate": 1.9095997465413457e-05, - "loss": 1.9785, + "loss": 2.4263, "step": 4280 }, { "epoch": 0.45, - "grad_norm": 4.9550628662109375, + "grad_norm": 36.992225646972656, "learning_rate": 1.9093885309958814e-05, - "loss": 2.0354, + "loss": 2.4229, "step": 4290 }, { "epoch": 0.45, - "grad_norm": 12.926891326904297, + "grad_norm": 30.73526954650879, "learning_rate": 1.9091773154504175e-05, - "loss": 1.7901, + "loss": 2.3359, "step": 4300 }, { "epoch": 0.46, - "grad_norm": 5.428956985473633, + "grad_norm": 29.080902099609375, "learning_rate": 1.9089660999049532e-05, - "loss": 1.7351, + "loss": 2.1799, "step": 4310 }, { "epoch": 0.46, - "grad_norm": 4.773605823516846, + "grad_norm": 25.218114852905273, "learning_rate": 1.908754884359489e-05, - "loss": 1.9385, + "loss": 2.4503, "step": 4320 }, { "epoch": 0.46, - "grad_norm": 7.30823278427124, + "grad_norm": 30.394649505615234, "learning_rate": 1.908543668814025e-05, - "loss": 1.8742, + "loss": 2.1964, "step": 4330 }, { "epoch": 0.46, - "grad_norm": 4.8741044998168945, + "grad_norm": 25.044872283935547, "learning_rate": 1.9083324532685607e-05, - "loss": 1.9713, + "loss": 2.318, "step": 4340 }, { "epoch": 0.46, - "grad_norm": 7.783595085144043, + "grad_norm": 26.89373207092285, "learning_rate": 1.9081212377230967e-05, - "loss": 1.8737, + "loss": 2.2495, "step": 4350 }, { "epoch": 0.46, - "grad_norm": 9.741421699523926, + "grad_norm": 30.206884384155273, "learning_rate": 1.9079100221776324e-05, - "loss": 1.9093, + "loss": 2.2301, "step": 4360 }, { "epoch": 0.46, - "grad_norm": 11.92908763885498, + "grad_norm": 28.77404022216797, "learning_rate": 1.907698806632168e-05, - "loss": 2.2891, + "loss": 2.5314, "step": 4370 }, { "epoch": 0.46, - "grad_norm": 8.583222389221191, + "grad_norm": 26.903778076171875, "learning_rate": 1.9074875910867042e-05, - "loss": 1.9955, + "loss": 2.2764, "step": 4380 }, { "epoch": 0.46, - "grad_norm": 5.007065296173096, + "grad_norm": 25.390682220458984, "learning_rate": 1.90727637554124e-05, - "loss": 2.0473, + "loss": 2.4411, "step": 4390 }, { "epoch": 0.46, - "grad_norm": 7.649864673614502, + "grad_norm": 26.4592342376709, "learning_rate": 1.907065159995776e-05, - "loss": 1.7022, + "loss": 2.0282, "step": 4400 }, { "epoch": 0.47, - "grad_norm": 6.012762069702148, + "grad_norm": 29.557641983032227, "learning_rate": 1.9068539444503116e-05, - "loss": 1.7775, + "loss": 2.08, "step": 4410 }, { "epoch": 0.47, - "grad_norm": 6.9586920738220215, + "grad_norm": 31.89646339416504, "learning_rate": 1.9066427289048473e-05, - "loss": 1.6957, + "loss": 2.2938, "step": 4420 }, { "epoch": 0.47, - "grad_norm": 10.563836097717285, + "grad_norm": 28.309213638305664, "learning_rate": 1.9064315133593834e-05, - "loss": 1.8172, + "loss": 2.2516, "step": 4430 }, { "epoch": 0.47, - "grad_norm": 9.8902587890625, + "grad_norm": 32.20077133178711, "learning_rate": 1.906220297813919e-05, - "loss": 1.9712, + "loss": 2.3992, "step": 4440 }, { "epoch": 0.47, - "grad_norm": 5.2699432373046875, + "grad_norm": 35.725257873535156, "learning_rate": 1.906009082268455e-05, - "loss": 1.8506, + "loss": 2.2516, "step": 4450 }, { "epoch": 0.47, - "grad_norm": 9.140056610107422, + "grad_norm": 29.70552635192871, "learning_rate": 1.9057978667229912e-05, - "loss": 1.8569, + "loss": 2.2835, "step": 4460 }, { "epoch": 0.47, - "grad_norm": 8.201128005981445, + "grad_norm": 26.003440856933594, "learning_rate": 1.9055866511775266e-05, - "loss": 2.0567, + "loss": 2.5418, "step": 4470 }, { "epoch": 0.47, - "grad_norm": 10.75178337097168, + "grad_norm": 32.019073486328125, "learning_rate": 1.9053754356320626e-05, - "loss": 1.9161, + "loss": 2.0982, "step": 4480 }, { "epoch": 0.47, - "grad_norm": 4.999585151672363, + "grad_norm": 23.783775329589844, "learning_rate": 1.9051642200865987e-05, - "loss": 1.8881, + "loss": 2.309, "step": 4490 }, { "epoch": 0.48, - "grad_norm": 10.223498344421387, + "grad_norm": 36.36665725708008, "learning_rate": 1.9049530045411344e-05, - "loss": 1.9182, + "loss": 2.4244, "step": 4500 }, { "epoch": 0.48, - "grad_norm": 7.749310493469238, + "grad_norm": 28.852813720703125, "learning_rate": 1.9047417889956704e-05, - "loss": 1.8959, + "loss": 2.574, "step": 4510 }, { "epoch": 0.48, - "grad_norm": 7.49983024597168, + "grad_norm": 35.975154876708984, "learning_rate": 1.904530573450206e-05, - "loss": 1.7116, + "loss": 2.3837, "step": 4520 }, { "epoch": 0.48, - "grad_norm": 12.782102584838867, + "grad_norm": 34.62568664550781, "learning_rate": 1.904319357904742e-05, - "loss": 1.7827, + "loss": 2.4371, "step": 4530 }, { "epoch": 0.48, - "grad_norm": 5.885812282562256, + "grad_norm": 39.02933120727539, "learning_rate": 1.904108142359278e-05, - "loss": 1.9049, + "loss": 2.4688, "step": 4540 }, { "epoch": 0.48, - "grad_norm": 5.364223957061768, + "grad_norm": 26.52240753173828, "learning_rate": 1.9038969268138136e-05, - "loss": 1.91, + "loss": 2.1797, "step": 4550 }, { "epoch": 0.48, - "grad_norm": 4.863171577453613, + "grad_norm": 27.07676887512207, "learning_rate": 1.9036857112683496e-05, - "loss": 1.7886, + "loss": 2.4005, "step": 4560 }, { "epoch": 0.48, - "grad_norm": 6.180538177490234, + "grad_norm": 28.290828704833984, "learning_rate": 1.9034744957228853e-05, - "loss": 1.734, + "loss": 2.249, "step": 4570 }, { "epoch": 0.48, - "grad_norm": 8.207633018493652, + "grad_norm": 36.4083366394043, "learning_rate": 1.903263280177421e-05, - "loss": 1.707, + "loss": 2.243, "step": 4580 }, { "epoch": 0.48, - "grad_norm": 10.918370246887207, + "grad_norm": 31.49825096130371, "learning_rate": 1.903052064631957e-05, - "loss": 1.8515, + "loss": 2.2736, "step": 4590 }, { "epoch": 0.49, - "grad_norm": 5.6864213943481445, + "grad_norm": 22.42896842956543, "learning_rate": 1.9028408490864928e-05, - "loss": 1.7327, + "loss": 1.9745, "step": 4600 }, { "epoch": 0.49, - "grad_norm": 7.199208736419678, + "grad_norm": 31.786800384521484, "learning_rate": 1.902629633541029e-05, - "loss": 1.7575, + "loss": 2.3451, "step": 4610 }, { "epoch": 0.49, - "grad_norm": 6.165760040283203, + "grad_norm": 28.464506149291992, "learning_rate": 1.9024184179955646e-05, - "loss": 1.7969, + "loss": 2.1222, "step": 4620 }, { "epoch": 0.49, - "grad_norm": 8.26678466796875, + "grad_norm": 36.329219818115234, "learning_rate": 1.9022072024501003e-05, - "loss": 1.9846, + "loss": 2.496, "step": 4630 }, { "epoch": 0.49, - "grad_norm": 5.387720584869385, + "grad_norm": 20.43042755126953, "learning_rate": 1.9019959869046363e-05, - "loss": 1.8102, + "loss": 2.4145, "step": 4640 }, { "epoch": 0.49, - "grad_norm": 8.084117889404297, + "grad_norm": 27.89584732055664, "learning_rate": 1.901784771359172e-05, - "loss": 1.9999, + "loss": 2.4882, "step": 4650 }, { "epoch": 0.49, - "grad_norm": 6.777525424957275, + "grad_norm": 28.992870330810547, "learning_rate": 1.901573555813708e-05, - "loss": 1.683, + "loss": 2.3018, "step": 4660 }, { "epoch": 0.49, - "grad_norm": 6.5585036277771, + "grad_norm": 33.747982025146484, "learning_rate": 1.901362340268244e-05, - "loss": 1.7327, + "loss": 2.5095, "step": 4670 }, { "epoch": 0.49, - "grad_norm": 10.273542404174805, + "grad_norm": 32.351654052734375, "learning_rate": 1.90115112472278e-05, - "loss": 1.8995, + "loss": 2.2846, "step": 4680 }, { "epoch": 0.5, - "grad_norm": 9.794602394104004, + "grad_norm": 27.323238372802734, "learning_rate": 1.9009399091773155e-05, - "loss": 1.6766, + "loss": 2.1481, "step": 4690 }, { "epoch": 0.5, - "grad_norm": 20.568296432495117, + "grad_norm": 37.053462982177734, "learning_rate": 1.9007286936318512e-05, - "loss": 1.7216, + "loss": 2.3974, "step": 4700 }, { "epoch": 0.5, - "grad_norm": 6.897904872894287, + "grad_norm": 27.54892349243164, "learning_rate": 1.9005174780863873e-05, - "loss": 1.8122, + "loss": 2.4579, "step": 4710 }, { "epoch": 0.5, - "grad_norm": 6.6557416915893555, + "grad_norm": 26.86530303955078, "learning_rate": 1.9003062625409233e-05, - "loss": 1.8922, + "loss": 2.5643, "step": 4720 }, { "epoch": 0.5, - "grad_norm": 7.73758602142334, + "grad_norm": 26.774612426757812, "learning_rate": 1.900095046995459e-05, - "loss": 1.6163, + "loss": 1.8813, "step": 4730 }, { "epoch": 0.5, - "grad_norm": 4.540678024291992, + "grad_norm": 22.42378044128418, "learning_rate": 1.8998838314499948e-05, - "loss": 1.6699, + "loss": 2.0807, "step": 4740 }, { "epoch": 0.5, - "grad_norm": 11.03013801574707, + "grad_norm": 31.072999954223633, "learning_rate": 1.8996726159045308e-05, - "loss": 1.7233, + "loss": 2.1928, "step": 4750 }, { "epoch": 0.5, - "grad_norm": 6.86831521987915, + "grad_norm": 32.838043212890625, "learning_rate": 1.8994614003590665e-05, - "loss": 1.4966, + "loss": 2.3507, "step": 4760 }, { "epoch": 0.5, - "grad_norm": 8.519844055175781, + "grad_norm": 30.594219207763672, "learning_rate": 1.8992501848136026e-05, - "loss": 1.9758, + "loss": 2.6246, "step": 4770 }, { "epoch": 0.5, - "grad_norm": 8.356870651245117, + "grad_norm": 23.105655670166016, "learning_rate": 1.8990389692681383e-05, - "loss": 1.7228, + "loss": 2.2012, "step": 4780 }, { "epoch": 0.51, - "grad_norm": 6.323134899139404, + "grad_norm": 26.760269165039062, "learning_rate": 1.8988277537226743e-05, - "loss": 1.8857, + "loss": 2.3501, "step": 4790 }, { "epoch": 0.51, - "grad_norm": 11.9068021774292, + "grad_norm": 31.077037811279297, "learning_rate": 1.89861653817721e-05, - "loss": 1.828, + "loss": 2.4081, "step": 4800 }, { "epoch": 0.51, - "grad_norm": 7.867719650268555, + "grad_norm": 33.67192459106445, "learning_rate": 1.8984053226317457e-05, - "loss": 1.7139, + "loss": 2.1299, "step": 4810 }, { "epoch": 0.51, - "grad_norm": 18.043161392211914, + "grad_norm": 41.59432601928711, "learning_rate": 1.8981941070862818e-05, - "loss": 1.7986, + "loss": 2.4359, "step": 4820 }, { "epoch": 0.51, - "grad_norm": 6.26272439956665, + "grad_norm": 24.75265121459961, "learning_rate": 1.8979828915408175e-05, - "loss": 1.8247, + "loss": 2.4082, "step": 4830 }, { "epoch": 0.51, - "grad_norm": 4.842712879180908, + "grad_norm": 28.401010513305664, "learning_rate": 1.8977716759953535e-05, - "loss": 1.6319, + "loss": 2.182, "step": 4840 }, { "epoch": 0.51, - "grad_norm": 7.895698547363281, + "grad_norm": 30.61347007751465, "learning_rate": 1.8975604604498892e-05, - "loss": 1.7171, + "loss": 2.2159, "step": 4850 }, { "epoch": 0.51, - "grad_norm": 6.009286880493164, + "grad_norm": 34.471248626708984, "learning_rate": 1.897349244904425e-05, - "loss": 1.8495, + "loss": 2.3167, "step": 4860 }, { "epoch": 0.51, - "grad_norm": 10.742098808288574, + "grad_norm": 34.009761810302734, "learning_rate": 1.897138029358961e-05, - "loss": 1.941, + "loss": 2.4821, "step": 4870 }, { "epoch": 0.52, - "grad_norm": 8.006036758422852, + "grad_norm": 35.433841705322266, "learning_rate": 1.8969268138134967e-05, - "loss": 1.9007, + "loss": 2.2596, "step": 4880 }, { "epoch": 0.52, - "grad_norm": 4.9336347579956055, + "grad_norm": 30.767873764038086, "learning_rate": 1.8967155982680328e-05, - "loss": 1.768, + "loss": 2.5695, "step": 4890 }, { "epoch": 0.52, - "grad_norm": 10.695127487182617, + "grad_norm": 36.60279083251953, "learning_rate": 1.8965043827225685e-05, - "loss": 1.5979, + "loss": 2.1952, "step": 4900 }, { "epoch": 0.52, - "grad_norm": 7.6003289222717285, + "grad_norm": 27.949661254882812, "learning_rate": 1.8962931671771042e-05, - "loss": 1.6167, + "loss": 2.2729, "step": 4910 }, { "epoch": 0.52, - "grad_norm": 7.659635543823242, + "grad_norm": 26.903106689453125, "learning_rate": 1.8960819516316402e-05, - "loss": 1.6846, + "loss": 2.3155, "step": 4920 }, { "epoch": 0.52, - "grad_norm": 12.417684555053711, + "grad_norm": 37.89351272583008, "learning_rate": 1.8958707360861763e-05, - "loss": 1.8879, + "loss": 2.4589, "step": 4930 }, { "epoch": 0.52, - "grad_norm": 9.979665756225586, + "grad_norm": 34.92732620239258, "learning_rate": 1.895659520540712e-05, - "loss": 1.8637, + "loss": 2.5956, "step": 4940 }, { "epoch": 0.52, - "grad_norm": 9.8301362991333, + "grad_norm": 33.411231994628906, "learning_rate": 1.895448304995248e-05, - "loss": 1.7119, + "loss": 2.2169, "step": 4950 }, { "epoch": 0.52, - "grad_norm": 9.25526237487793, + "grad_norm": 29.412246704101562, "learning_rate": 1.8952370894497837e-05, - "loss": 1.633, + "loss": 2.3855, "step": 4960 }, { "epoch": 0.52, - "grad_norm": 6.250025749206543, + "grad_norm": 25.789960861206055, "learning_rate": 1.8950258739043194e-05, - "loss": 1.5817, + "loss": 2.1853, "step": 4970 }, { "epoch": 0.53, - "grad_norm": 10.888240814208984, + "grad_norm": 28.1785945892334, "learning_rate": 1.8948146583588555e-05, - "loss": 1.7844, + "loss": 2.3762, "step": 4980 }, { "epoch": 0.53, - "grad_norm": 11.427401542663574, + "grad_norm": 23.46927261352539, "learning_rate": 1.8946034428133912e-05, - "loss": 1.6835, + "loss": 2.1641, "step": 4990 }, { "epoch": 0.53, - "grad_norm": 15.932563781738281, + "grad_norm": 36.143463134765625, "learning_rate": 1.8943922272679272e-05, - "loss": 2.0869, + "loss": 2.4334, "step": 5000 }, { "epoch": 0.53, - "grad_norm": 8.511100769042969, + "grad_norm": 43.79252243041992, "learning_rate": 1.894181011722463e-05, - "loss": 1.5805, + "loss": 2.0669, "step": 5010 }, { "epoch": 0.53, - "grad_norm": 5.99853515625, + "grad_norm": 31.99622917175293, "learning_rate": 1.8939697961769987e-05, - "loss": 1.792, + "loss": 2.3703, "step": 5020 }, { "epoch": 0.53, - "grad_norm": 8.792450904846191, + "grad_norm": 31.545543670654297, "learning_rate": 1.8937585806315347e-05, - "loss": 1.6828, + "loss": 2.3121, "step": 5030 }, { "epoch": 0.53, - "grad_norm": 11.103832244873047, + "grad_norm": 30.524154663085938, "learning_rate": 1.8935473650860704e-05, - "loss": 1.817, + "loss": 2.2982, "step": 5040 }, { "epoch": 0.53, - "grad_norm": 9.042688369750977, + "grad_norm": 22.724952697753906, "learning_rate": 1.8933361495406065e-05, - "loss": 1.6575, + "loss": 2.2356, "step": 5050 }, { "epoch": 0.53, - "grad_norm": 8.62700366973877, + "grad_norm": 33.06990051269531, "learning_rate": 1.8931249339951422e-05, - "loss": 1.7974, + "loss": 2.7267, "step": 5060 }, { "epoch": 0.54, - "grad_norm": 17.722253799438477, + "grad_norm": 42.180938720703125, "learning_rate": 1.892913718449678e-05, - "loss": 1.6332, + "loss": 2.1073, "step": 5070 }, { "epoch": 0.54, - "grad_norm": 13.66730785369873, + "grad_norm": 40.77030944824219, "learning_rate": 1.892702502904214e-05, - "loss": 1.6113, + "loss": 2.2094, "step": 5080 }, { "epoch": 0.54, - "grad_norm": 9.247448921203613, + "grad_norm": 25.67778968811035, "learning_rate": 1.8924912873587496e-05, - "loss": 1.6752, + "loss": 2.1171, "step": 5090 }, { "epoch": 0.54, - "grad_norm": 7.664931297302246, + "grad_norm": 31.92324447631836, "learning_rate": 1.8922800718132857e-05, - "loss": 1.781, + "loss": 2.2865, "step": 5100 }, { "epoch": 0.54, - "grad_norm": 9.330229759216309, + "grad_norm": 33.201576232910156, "learning_rate": 1.8920688562678214e-05, - "loss": 1.8335, + "loss": 2.3766, "step": 5110 }, { "epoch": 0.54, - "grad_norm": 11.128196716308594, + "grad_norm": 35.265342712402344, "learning_rate": 1.891857640722357e-05, - "loss": 1.6371, + "loss": 2.0106, "step": 5120 }, { "epoch": 0.54, - "grad_norm": 8.274940490722656, + "grad_norm": 22.429931640625, "learning_rate": 1.891646425176893e-05, - "loss": 1.7979, + "loss": 2.1841, "step": 5130 }, { "epoch": 0.54, - "grad_norm": 12.474943161010742, + "grad_norm": 43.597145080566406, "learning_rate": 1.891435209631429e-05, - "loss": 1.8535, + "loss": 2.2811, "step": 5140 }, { "epoch": 0.54, - "grad_norm": 9.798068046569824, + "grad_norm": 27.73024559020996, "learning_rate": 1.891223994085965e-05, - "loss": 1.6029, + "loss": 2.2848, "step": 5150 }, { "epoch": 0.54, - "grad_norm": 7.719514846801758, + "grad_norm": 35.52192306518555, "learning_rate": 1.891012778540501e-05, - "loss": 1.4642, + "loss": 2.1609, "step": 5160 }, { "epoch": 0.55, - "grad_norm": 7.056937217712402, + "grad_norm": 28.694543838500977, "learning_rate": 1.8908015629950367e-05, - "loss": 1.4929, + "loss": 2.2503, "step": 5170 }, { "epoch": 0.55, - "grad_norm": 8.52943229675293, + "grad_norm": 23.1101016998291, "learning_rate": 1.8905903474495724e-05, - "loss": 1.6688, + "loss": 2.191, "step": 5180 }, { "epoch": 0.55, - "grad_norm": 12.78560733795166, + "grad_norm": 33.67193603515625, "learning_rate": 1.8903791319041084e-05, - "loss": 1.7534, + "loss": 2.2246, "step": 5190 }, { "epoch": 0.55, - "grad_norm": 8.41891098022461, + "grad_norm": 40.76647186279297, "learning_rate": 1.890167916358644e-05, - "loss": 1.6764, + "loss": 2.3392, "step": 5200 }, { "epoch": 0.55, - "grad_norm": 9.16325855255127, + "grad_norm": 26.66714859008789, "learning_rate": 1.8899567008131802e-05, - "loss": 1.9362, + "loss": 2.5207, "step": 5210 }, { "epoch": 0.55, - "grad_norm": 9.214112281799316, + "grad_norm": 36.18425369262695, "learning_rate": 1.889745485267716e-05, - "loss": 1.7704, + "loss": 2.2335, "step": 5220 }, { "epoch": 0.55, - "grad_norm": 3.2266554832458496, + "grad_norm": 30.565229415893555, "learning_rate": 1.8895342697222516e-05, - "loss": 1.5465, + "loss": 2.0401, "step": 5230 }, { "epoch": 0.55, - "grad_norm": 17.84965705871582, + "grad_norm": 32.192054748535156, "learning_rate": 1.8893230541767876e-05, - "loss": 1.7516, + "loss": 2.2798, "step": 5240 }, { "epoch": 0.55, - "grad_norm": 7.141803741455078, + "grad_norm": 32.637691497802734, "learning_rate": 1.8891118386313233e-05, - "loss": 1.7254, + "loss": 2.11, "step": 5250 }, { "epoch": 0.56, - "grad_norm": 7.8065385818481445, + "grad_norm": 27.611331939697266, "learning_rate": 1.8889006230858594e-05, - "loss": 1.6126, + "loss": 2.3985, "step": 5260 }, { "epoch": 0.56, - "grad_norm": 12.620861053466797, + "grad_norm": 42.55600357055664, "learning_rate": 1.888689407540395e-05, - "loss": 1.8728, + "loss": 2.3482, "step": 5270 }, { "epoch": 0.56, - "grad_norm": 6.416200637817383, + "grad_norm": 30.54423713684082, "learning_rate": 1.8884781919949308e-05, - "loss": 1.7353, + "loss": 2.1372, "step": 5280 }, { "epoch": 0.56, - "grad_norm": 12.291290283203125, + "grad_norm": 41.01797103881836, "learning_rate": 1.888266976449467e-05, - "loss": 1.76, + "loss": 2.356, "step": 5290 }, { "epoch": 0.56, - "grad_norm": 5.090100288391113, + "grad_norm": 26.29137420654297, "learning_rate": 1.8880557609040026e-05, - "loss": 1.3995, + "loss": 2.1066, "step": 5300 }, { "epoch": 0.56, - "grad_norm": 6.316115379333496, + "grad_norm": 27.192663192749023, "learning_rate": 1.8878445453585386e-05, - "loss": 1.3924, + "loss": 2.2677, "step": 5310 }, { "epoch": 0.56, - "grad_norm": 8.684137344360352, + "grad_norm": 31.74489974975586, "learning_rate": 1.8876333298130743e-05, - "loss": 1.7758, + "loss": 1.9899, "step": 5320 }, { "epoch": 0.56, - "grad_norm": 11.667380332946777, + "grad_norm": 36.99678039550781, "learning_rate": 1.8874221142676104e-05, - "loss": 1.5014, + "loss": 2.4297, "step": 5330 }, { "epoch": 0.56, - "grad_norm": 16.806978225708008, + "grad_norm": 24.99443244934082, "learning_rate": 1.887210898722146e-05, - "loss": 1.416, + "loss": 2.179, "step": 5340 }, { "epoch": 0.57, - "grad_norm": 15.921977043151855, + "grad_norm": 33.653709411621094, "learning_rate": 1.8869996831766818e-05, - "loss": 1.4961, + "loss": 2.094, "step": 5350 }, { "epoch": 0.57, - "grad_norm": 9.867745399475098, + "grad_norm": 35.671058654785156, "learning_rate": 1.886788467631218e-05, - "loss": 1.555, + "loss": 2.3282, "step": 5360 }, { "epoch": 0.57, - "grad_norm": 4.019556522369385, + "grad_norm": 26.314132690429688, "learning_rate": 1.8865772520857535e-05, - "loss": 1.6962, + "loss": 2.0589, "step": 5370 }, { "epoch": 0.57, - "grad_norm": 11.986268043518066, + "grad_norm": 30.01563262939453, "learning_rate": 1.8863660365402896e-05, - "loss": 1.6863, + "loss": 2.2671, "step": 5380 }, { "epoch": 0.57, - "grad_norm": 13.122934341430664, + "grad_norm": 34.180091857910156, "learning_rate": 1.8861548209948253e-05, - "loss": 1.4985, + "loss": 2.2004, "step": 5390 }, { "epoch": 0.57, - "grad_norm": 8.458944320678711, + "grad_norm": 34.17359924316406, "learning_rate": 1.885943605449361e-05, - "loss": 1.5356, + "loss": 2.2402, "step": 5400 }, { "epoch": 0.57, - "grad_norm": 14.43547248840332, + "grad_norm": 31.47393798828125, "learning_rate": 1.885732389903897e-05, - "loss": 1.5152, + "loss": 2.1169, "step": 5410 }, { "epoch": 0.57, - "grad_norm": 3.925853967666626, + "grad_norm": 21.954078674316406, "learning_rate": 1.885521174358433e-05, - "loss": 1.89, + "loss": 2.387, "step": 5420 }, { "epoch": 0.57, - "grad_norm": 14.082711219787598, + "grad_norm": 37.841732025146484, "learning_rate": 1.8853099588129688e-05, - "loss": 1.4776, + "loss": 2.1452, "step": 5430 }, { "epoch": 0.57, - "grad_norm": 7.235020637512207, + "grad_norm": 28.29035186767578, "learning_rate": 1.885098743267505e-05, - "loss": 1.4604, + "loss": 2.146, "step": 5440 }, { "epoch": 0.58, - "grad_norm": 5.0166425704956055, + "grad_norm": 24.07041358947754, "learning_rate": 1.8848875277220406e-05, - "loss": 1.527, + "loss": 1.9878, "step": 5450 }, { "epoch": 0.58, - "grad_norm": 9.251256942749023, + "grad_norm": 35.24446487426758, "learning_rate": 1.8846763121765763e-05, - "loss": 1.5969, + "loss": 2.2523, "step": 5460 }, { "epoch": 0.58, - "grad_norm": 8.953072547912598, + "grad_norm": 39.25656509399414, "learning_rate": 1.8844650966311123e-05, - "loss": 2.1507, + "loss": 2.7441, "step": 5470 }, { "epoch": 0.58, - "grad_norm": 7.995614528656006, + "grad_norm": 37.85302734375, "learning_rate": 1.884253881085648e-05, - "loss": 1.8131, + "loss": 2.3609, "step": 5480 }, { "epoch": 0.58, - "grad_norm": 13.863811492919922, + "grad_norm": 40.23750686645508, "learning_rate": 1.884042665540184e-05, - "loss": 1.6521, + "loss": 2.3147, "step": 5490 }, { "epoch": 0.58, - "grad_norm": 10.26335334777832, + "grad_norm": 35.643455505371094, "learning_rate": 1.8838314499947198e-05, - "loss": 1.4931, + "loss": 2.289, "step": 5500 }, { "epoch": 0.58, - "grad_norm": 4.849292278289795, + "grad_norm": 26.281829833984375, "learning_rate": 1.8836202344492555e-05, - "loss": 1.4809, + "loss": 2.168, "step": 5510 }, { "epoch": 0.58, - "grad_norm": 25.918384552001953, + "grad_norm": 40.54023742675781, "learning_rate": 1.8834090189037915e-05, - "loss": 1.6356, + "loss": 2.3312, "step": 5520 }, { "epoch": 0.58, - "grad_norm": 9.258820533752441, + "grad_norm": 25.476072311401367, "learning_rate": 1.8831978033583273e-05, - "loss": 1.9428, + "loss": 2.3487, "step": 5530 }, { "epoch": 0.59, - "grad_norm": 8.385503768920898, + "grad_norm": 26.384384155273438, "learning_rate": 1.8829865878128633e-05, - "loss": 1.3828, + "loss": 1.802, "step": 5540 }, { "epoch": 0.59, - "grad_norm": 7.919586181640625, + "grad_norm": 33.947471618652344, "learning_rate": 1.882775372267399e-05, - "loss": 1.3588, + "loss": 1.9485, "step": 5550 }, { "epoch": 0.59, - "grad_norm": 10.231463432312012, + "grad_norm": 33.614200592041016, "learning_rate": 1.8825641567219347e-05, - "loss": 1.7161, + "loss": 2.1798, "step": 5560 }, { "epoch": 0.59, - "grad_norm": 11.25167179107666, + "grad_norm": 37.97392272949219, "learning_rate": 1.8823529411764708e-05, - "loss": 1.67, + "loss": 2.236, "step": 5570 }, { "epoch": 0.59, - "grad_norm": 7.77572774887085, + "grad_norm": 32.49007034301758, "learning_rate": 1.8821417256310065e-05, - "loss": 1.5247, + "loss": 2.193, "step": 5580 }, { "epoch": 0.59, - "grad_norm": 12.246149063110352, + "grad_norm": 26.953819274902344, "learning_rate": 1.8819305100855425e-05, - "loss": 1.7022, + "loss": 1.9953, "step": 5590 }, { "epoch": 0.59, - "grad_norm": 8.44796371459961, + "grad_norm": 36.37388229370117, "learning_rate": 1.8817192945400786e-05, - "loss": 1.6944, + "loss": 2.4743, "step": 5600 }, { "epoch": 0.59, - "grad_norm": 6.304355144500732, + "grad_norm": 29.802799224853516, "learning_rate": 1.881508078994614e-05, - "loss": 1.6939, + "loss": 2.4942, "step": 5610 }, { "epoch": 0.59, - "grad_norm": 12.503573417663574, + "grad_norm": 32.10693359375, "learning_rate": 1.88129686344915e-05, - "loss": 1.7702, + "loss": 2.2652, "step": 5620 }, { "epoch": 0.59, - "grad_norm": 3.578159809112549, + "grad_norm": 20.761152267456055, "learning_rate": 1.881085647903686e-05, - "loss": 1.5089, + "loss": 2.0949, "step": 5630 }, { "epoch": 0.6, - "grad_norm": 16.137832641601562, + "grad_norm": 36.99005889892578, "learning_rate": 1.8808744323582217e-05, - "loss": 1.7962, + "loss": 2.4308, "step": 5640 }, { "epoch": 0.6, - "grad_norm": 9.573020935058594, + "grad_norm": 28.847166061401367, "learning_rate": 1.8806632168127578e-05, - "loss": 1.509, + "loss": 1.8587, "step": 5650 }, { "epoch": 0.6, - "grad_norm": 4.431938648223877, + "grad_norm": 30.45964241027832, "learning_rate": 1.8804520012672935e-05, - "loss": 1.4992, + "loss": 2.1965, "step": 5660 }, { "epoch": 0.6, - "grad_norm": 9.83855152130127, + "grad_norm": 28.12378692626953, "learning_rate": 1.8802407857218292e-05, - "loss": 1.5718, + "loss": 2.4076, "step": 5670 }, { "epoch": 0.6, - "grad_norm": 8.546374320983887, + "grad_norm": 39.90840530395508, "learning_rate": 1.8800295701763652e-05, - "loss": 1.4802, + "loss": 1.9974, "step": 5680 }, { "epoch": 0.6, - "grad_norm": 5.926705837249756, + "grad_norm": 36.8283576965332, "learning_rate": 1.879818354630901e-05, - "loss": 1.494, + "loss": 2.0005, "step": 5690 }, { "epoch": 0.6, - "grad_norm": 28.776119232177734, + "grad_norm": 31.60143280029297, "learning_rate": 1.879607139085437e-05, - "loss": 1.4433, + "loss": 2.2614, "step": 5700 }, { "epoch": 0.6, - "grad_norm": 10.105273246765137, + "grad_norm": 28.023279190063477, "learning_rate": 1.8793959235399727e-05, - "loss": 1.5333, + "loss": 2.2666, "step": 5710 }, { "epoch": 0.6, - "grad_norm": 12.138835906982422, + "grad_norm": 41.4437255859375, "learning_rate": 1.8791847079945084e-05, - "loss": 1.1391, + "loss": 2.0097, "step": 5720 }, { "epoch": 0.61, - "grad_norm": 3.2980639934539795, + "grad_norm": 36.47161102294922, "learning_rate": 1.8789734924490445e-05, - "loss": 1.4513, + "loss": 2.1446, "step": 5730 }, { "epoch": 0.61, - "grad_norm": 7.3011908531188965, + "grad_norm": 26.596790313720703, "learning_rate": 1.8787622769035802e-05, - "loss": 1.5519, + "loss": 2.3163, "step": 5740 }, { "epoch": 0.61, - "grad_norm": 5.015331268310547, + "grad_norm": 35.718162536621094, "learning_rate": 1.8785510613581162e-05, - "loss": 1.5944, + "loss": 2.0863, "step": 5750 }, { "epoch": 0.61, - "grad_norm": 7.82149600982666, + "grad_norm": 31.340463638305664, "learning_rate": 1.878339845812652e-05, - "loss": 2.0472, + "loss": 2.2733, "step": 5760 }, { "epoch": 0.61, - "grad_norm": 11.436149597167969, + "grad_norm": 28.97592544555664, "learning_rate": 1.8781286302671876e-05, - "loss": 1.4955, + "loss": 2.1614, "step": 5770 }, { "epoch": 0.61, - "grad_norm": 22.05530548095703, + "grad_norm": 31.454214096069336, "learning_rate": 1.8779174147217237e-05, - "loss": 1.5571, + "loss": 2.2222, "step": 5780 }, { "epoch": 0.61, - "grad_norm": 9.871994018554688, + "grad_norm": 28.794015884399414, "learning_rate": 1.8777061991762594e-05, - "loss": 1.3133, + "loss": 2.2425, "step": 5790 }, { "epoch": 0.61, - "grad_norm": 5.845597743988037, + "grad_norm": 40.129432678222656, "learning_rate": 1.8774949836307954e-05, - "loss": 1.3585, + "loss": 2.1924, "step": 5800 }, { "epoch": 0.61, - "grad_norm": 9.46414852142334, + "grad_norm": 24.468900680541992, "learning_rate": 1.877283768085331e-05, - "loss": 1.5461, + "loss": 1.9027, "step": 5810 }, { "epoch": 0.61, - "grad_norm": 7.776662826538086, + "grad_norm": 29.873882293701172, "learning_rate": 1.877072552539867e-05, - "loss": 1.5414, + "loss": 2.2143, "step": 5820 }, { "epoch": 0.62, - "grad_norm": 11.2785062789917, + "grad_norm": 31.30638313293457, "learning_rate": 1.876861336994403e-05, - "loss": 1.3917, + "loss": 1.9857, "step": 5830 }, { "epoch": 0.62, - "grad_norm": 10.41724967956543, + "grad_norm": 41.52237319946289, "learning_rate": 1.8766501214489386e-05, - "loss": 1.4164, + "loss": 2.1074, "step": 5840 }, { "epoch": 0.62, - "grad_norm": 10.537546157836914, + "grad_norm": 33.70042419433594, "learning_rate": 1.8764389059034747e-05, - "loss": 1.5224, + "loss": 2.1412, "step": 5850 }, { "epoch": 0.62, - "grad_norm": 11.026823043823242, + "grad_norm": 33.13836669921875, "learning_rate": 1.8762276903580107e-05, - "loss": 1.6668, + "loss": 2.2284, "step": 5860 }, { "epoch": 0.62, - "grad_norm": 8.87667465209961, + "grad_norm": 35.02610778808594, "learning_rate": 1.8760164748125464e-05, - "loss": 1.2483, + "loss": 1.9808, "step": 5870 }, { "epoch": 0.62, - "grad_norm": 11.31600284576416, + "grad_norm": 36.33309555053711, "learning_rate": 1.875805259267082e-05, - "loss": 1.5979, + "loss": 2.1058, "step": 5880 }, { "epoch": 0.62, - "grad_norm": 6.1131272315979, + "grad_norm": 18.363672256469727, "learning_rate": 1.8755940437216182e-05, - "loss": 1.2737, + "loss": 1.7946, "step": 5890 }, { "epoch": 0.62, - "grad_norm": 11.899662971496582, + "grad_norm": 33.50032424926758, "learning_rate": 1.875382828176154e-05, - "loss": 1.5614, + "loss": 2.2501, "step": 5900 }, { "epoch": 0.62, - "grad_norm": 7.871194362640381, + "grad_norm": 25.399208068847656, "learning_rate": 1.87517161263069e-05, - "loss": 1.2308, + "loss": 1.7025, "step": 5910 }, { "epoch": 0.63, - "grad_norm": 6.368232250213623, + "grad_norm": 26.40305519104004, "learning_rate": 1.8749603970852256e-05, - "loss": 1.4417, + "loss": 2.3547, "step": 5920 }, { "epoch": 0.63, - "grad_norm": 5.747691631317139, + "grad_norm": 34.04867172241211, "learning_rate": 1.8747491815397613e-05, - "loss": 1.5247, + "loss": 2.0957, "step": 5930 }, { "epoch": 0.63, - "grad_norm": 9.642438888549805, + "grad_norm": 22.28134536743164, "learning_rate": 1.8745379659942974e-05, - "loss": 1.5184, + "loss": 2.3093, "step": 5940 }, { "epoch": 0.63, - "grad_norm": 7.766819953918457, + "grad_norm": 25.88188362121582, "learning_rate": 1.874326750448833e-05, - "loss": 1.5579, + "loss": 2.2345, "step": 5950 }, { "epoch": 0.63, - "grad_norm": 7.099789142608643, + "grad_norm": 31.488771438598633, "learning_rate": 1.874115534903369e-05, - "loss": 1.3767, + "loss": 1.9528, "step": 5960 }, { "epoch": 0.63, - "grad_norm": 3.685720682144165, + "grad_norm": 34.183170318603516, "learning_rate": 1.873904319357905e-05, - "loss": 1.4803, + "loss": 2.4107, "step": 5970 }, { "epoch": 0.63, - "grad_norm": 24.545074462890625, + "grad_norm": 38.10966491699219, "learning_rate": 1.873693103812441e-05, - "loss": 1.6568, + "loss": 2.4124, "step": 5980 }, { "epoch": 0.63, - "grad_norm": 7.690281867980957, + "grad_norm": 33.42572784423828, "learning_rate": 1.8734818882669766e-05, - "loss": 1.1362, + "loss": 1.8423, "step": 5990 }, { "epoch": 0.63, - "grad_norm": 8.865501403808594, + "grad_norm": 44.47340774536133, "learning_rate": 1.8732706727215123e-05, - "loss": 1.6036, + "loss": 2.0334, "step": 6000 }, { "epoch": 0.63, - "grad_norm": 16.08588218688965, + "grad_norm": 33.61896896362305, "learning_rate": 1.8730594571760484e-05, - "loss": 1.3567, + "loss": 1.8963, "step": 6010 }, { "epoch": 0.64, - "grad_norm": 7.765570163726807, + "grad_norm": 30.239404678344727, "learning_rate": 1.872848241630584e-05, - "loss": 1.4688, + "loss": 2.1427, "step": 6020 }, { "epoch": 0.64, - "grad_norm": 7.537191390991211, + "grad_norm": 31.843130111694336, "learning_rate": 1.87263702608512e-05, - "loss": 1.5052, + "loss": 2.0615, "step": 6030 }, { "epoch": 0.64, - "grad_norm": 8.725358009338379, + "grad_norm": 24.0789737701416, "learning_rate": 1.872425810539656e-05, - "loss": 1.5197, + "loss": 2.1989, "step": 6040 }, { "epoch": 0.64, - "grad_norm": 18.452238082885742, + "grad_norm": 30.554279327392578, "learning_rate": 1.8722145949941915e-05, - "loss": 1.2727, + "loss": 2.1642, "step": 6050 }, { "epoch": 0.64, - "grad_norm": 8.405889511108398, + "grad_norm": 37.51786422729492, "learning_rate": 1.8720033794487276e-05, - "loss": 1.6766, + "loss": 2.2045, "step": 6060 }, { "epoch": 0.64, - "grad_norm": 8.902116775512695, + "grad_norm": 31.557382583618164, "learning_rate": 1.8717921639032633e-05, - "loss": 1.5626, + "loss": 2.3207, "step": 6070 }, { "epoch": 0.64, - "grad_norm": 10.424487113952637, + "grad_norm": 34.77410125732422, "learning_rate": 1.8715809483577993e-05, - "loss": 1.6368, + "loss": 2.205, "step": 6080 }, { "epoch": 0.64, - "grad_norm": 22.687944412231445, + "grad_norm": 35.61802291870117, "learning_rate": 1.8713697328123354e-05, - "loss": 1.5266, + "loss": 2.1883, "step": 6090 }, { "epoch": 0.64, - "grad_norm": 8.178642272949219, + "grad_norm": 32.93732833862305, "learning_rate": 1.8711585172668708e-05, - "loss": 1.4062, + "loss": 1.9594, "step": 6100 }, { "epoch": 0.65, - "grad_norm": 9.293828010559082, + "grad_norm": 24.29566764831543, "learning_rate": 1.8709473017214068e-05, - "loss": 1.343, + "loss": 2.4519, "step": 6110 }, { "epoch": 0.65, - "grad_norm": 4.483529567718506, + "grad_norm": 28.666702270507812, "learning_rate": 1.870736086175943e-05, - "loss": 1.5108, + "loss": 2.3185, "step": 6120 }, { "epoch": 0.65, - "grad_norm": 10.687625885009766, + "grad_norm": 30.56096649169922, "learning_rate": 1.8705248706304786e-05, - "loss": 1.3257, + "loss": 2.0716, "step": 6130 }, { "epoch": 0.65, - "grad_norm": 6.250577449798584, + "grad_norm": 30.971586227416992, "learning_rate": 1.8703136550850146e-05, - "loss": 1.4622, + "loss": 2.1094, "step": 6140 }, { "epoch": 0.65, - "grad_norm": 7.194085597991943, + "grad_norm": 20.084917068481445, "learning_rate": 1.8701024395395503e-05, - "loss": 1.4738, + "loss": 2.0527, "step": 6150 }, { "epoch": 0.65, - "grad_norm": 8.95712661743164, + "grad_norm": 31.739967346191406, "learning_rate": 1.869891223994086e-05, - "loss": 1.4522, + "loss": 2.0494, "step": 6160 }, { "epoch": 0.65, - "grad_norm": 5.8951416015625, + "grad_norm": 31.648813247680664, "learning_rate": 1.869680008448622e-05, - "loss": 1.5051, + "loss": 2.354, "step": 6170 }, { "epoch": 0.65, - "grad_norm": 8.281917572021484, + "grad_norm": 32.14897537231445, "learning_rate": 1.8694687929031578e-05, - "loss": 1.661, + "loss": 2.436, "step": 6180 }, { "epoch": 0.65, - "grad_norm": 11.601783752441406, + "grad_norm": 25.253480911254883, "learning_rate": 1.869257577357694e-05, - "loss": 1.4135, + "loss": 2.0016, "step": 6190 }, { "epoch": 0.65, - "grad_norm": 3.632052421569824, + "grad_norm": 29.37637710571289, "learning_rate": 1.8690463618122295e-05, - "loss": 1.1595, + "loss": 1.8768, "step": 6200 }, { "epoch": 0.66, - "grad_norm": 3.8255083560943604, + "grad_norm": 30.526411056518555, "learning_rate": 1.8688351462667653e-05, - "loss": 1.2236, + "loss": 1.9599, "step": 6210 }, { "epoch": 0.66, - "grad_norm": 6.999131202697754, + "grad_norm": 29.423065185546875, "learning_rate": 1.8686239307213013e-05, - "loss": 1.5476, + "loss": 2.4284, "step": 6220 }, { "epoch": 0.66, - "grad_norm": 8.969925880432129, + "grad_norm": 29.271955490112305, "learning_rate": 1.868412715175837e-05, - "loss": 1.4111, + "loss": 1.9615, "step": 6230 }, { "epoch": 0.66, - "grad_norm": 11.696625709533691, + "grad_norm": 34.233131408691406, "learning_rate": 1.868201499630373e-05, - "loss": 1.3737, + "loss": 2.2783, "step": 6240 }, { "epoch": 0.66, - "grad_norm": 8.655598640441895, + "grad_norm": 40.61522674560547, "learning_rate": 1.8679902840849088e-05, - "loss": 1.368, + "loss": 2.0829, "step": 6250 }, { "epoch": 0.66, - "grad_norm": 11.08277416229248, + "grad_norm": 29.77021026611328, "learning_rate": 1.8677790685394445e-05, - "loss": 1.3096, + "loss": 2.0179, "step": 6260 }, { "epoch": 0.66, - "grad_norm": 8.886062622070312, + "grad_norm": 38.593013763427734, "learning_rate": 1.8675678529939805e-05, - "loss": 1.4219, + "loss": 2.0182, "step": 6270 }, { "epoch": 0.66, - "grad_norm": 16.44879150390625, + "grad_norm": 35.987125396728516, "learning_rate": 1.8673566374485162e-05, - "loss": 1.3114, + "loss": 2.0263, "step": 6280 }, { "epoch": 0.66, - "grad_norm": 4.634844779968262, + "grad_norm": 35.6682243347168, "learning_rate": 1.8671454219030523e-05, - "loss": 1.2954, + "loss": 2.4343, "step": 6290 }, { "epoch": 0.67, - "grad_norm": 12.2855863571167, + "grad_norm": 36.548973083496094, "learning_rate": 1.866934206357588e-05, - "loss": 1.3122, + "loss": 1.9774, "step": 6300 }, { "epoch": 0.67, - "grad_norm": 13.549945831298828, + "grad_norm": 32.951271057128906, "learning_rate": 1.8667229908121237e-05, - "loss": 1.42, + "loss": 2.2046, "step": 6310 }, { "epoch": 0.67, - "grad_norm": 8.36904239654541, + "grad_norm": 29.28367042541504, "learning_rate": 1.8665117752666597e-05, - "loss": 1.33, + "loss": 2.2658, "step": 6320 }, { "epoch": 0.67, - "grad_norm": 4.921930313110352, + "grad_norm": 30.707307815551758, "learning_rate": 1.8663005597211954e-05, - "loss": 1.3784, + "loss": 2.1415, "step": 6330 }, { "epoch": 0.67, - "grad_norm": 13.757863998413086, + "grad_norm": 39.45580291748047, "learning_rate": 1.8660893441757315e-05, - "loss": 1.734, + "loss": 2.2723, "step": 6340 }, { "epoch": 0.67, - "grad_norm": 10.491952896118164, + "grad_norm": 29.175912857055664, "learning_rate": 1.8658781286302675e-05, - "loss": 1.1117, + "loss": 2.004, "step": 6350 }, { "epoch": 0.67, - "grad_norm": 10.953246116638184, + "grad_norm": 31.869464874267578, "learning_rate": 1.8656669130848033e-05, - "loss": 1.2253, + "loss": 2.0991, "step": 6360 }, { "epoch": 0.67, - "grad_norm": 12.927535057067871, + "grad_norm": 31.840133666992188, "learning_rate": 1.865455697539339e-05, - "loss": 1.3108, + "loss": 2.0608, "step": 6370 }, { "epoch": 0.67, - "grad_norm": 8.346553802490234, + "grad_norm": 42.07888412475586, "learning_rate": 1.865244481993875e-05, - "loss": 1.2957, + "loss": 1.9745, "step": 6380 }, { "epoch": 0.67, - "grad_norm": 9.071084022521973, + "grad_norm": 34.950069427490234, "learning_rate": 1.8650332664484107e-05, - "loss": 1.3334, + "loss": 2.3188, "step": 6390 }, { "epoch": 0.68, - "grad_norm": 10.963180541992188, + "grad_norm": 35.35287094116211, "learning_rate": 1.8648220509029468e-05, - "loss": 1.6256, + "loss": 2.0932, "step": 6400 }, { "epoch": 0.68, - "grad_norm": 7.761133193969727, + "grad_norm": 31.072500228881836, "learning_rate": 1.8646108353574825e-05, - "loss": 1.4728, + "loss": 2.1664, "step": 6410 }, { "epoch": 0.68, - "grad_norm": 12.719941139221191, + "grad_norm": 31.65738296508789, "learning_rate": 1.8643996198120182e-05, - "loss": 1.6114, + "loss": 2.157, "step": 6420 }, { "epoch": 0.68, - "grad_norm": 19.55401039123535, + "grad_norm": 27.858642578125, "learning_rate": 1.8641884042665542e-05, - "loss": 1.4932, + "loss": 2.194, "step": 6430 }, { "epoch": 0.68, - "grad_norm": 10.35053539276123, + "grad_norm": 28.47040367126465, "learning_rate": 1.86397718872109e-05, - "loss": 1.1528, + "loss": 2.0951, "step": 6440 }, { "epoch": 0.68, - "grad_norm": 3.3290634155273438, + "grad_norm": 28.262086868286133, "learning_rate": 1.863765973175626e-05, - "loss": 1.2593, + "loss": 2.0183, "step": 6450 }, { "epoch": 0.68, - "grad_norm": 8.643356323242188, + "grad_norm": 30.78612518310547, "learning_rate": 1.8635547576301617e-05, - "loss": 1.3923, + "loss": 1.9215, "step": 6460 }, { "epoch": 0.68, - "grad_norm": 15.729681015014648, + "grad_norm": 32.046424865722656, "learning_rate": 1.8633435420846974e-05, - "loss": 1.5202, + "loss": 2.1191, "step": 6470 }, { "epoch": 0.68, - "grad_norm": 7.26784086227417, + "grad_norm": 22.082664489746094, "learning_rate": 1.8631323265392334e-05, - "loss": 1.5445, + "loss": 1.9333, "step": 6480 }, { "epoch": 0.69, - "grad_norm": 8.714397430419922, + "grad_norm": 22.621196746826172, "learning_rate": 1.862921110993769e-05, - "loss": 1.4298, + "loss": 1.7746, "step": 6490 }, { "epoch": 0.69, - "grad_norm": 4.198194980621338, + "grad_norm": 36.803794860839844, "learning_rate": 1.8627098954483052e-05, - "loss": 1.2448, + "loss": 1.978, "step": 6500 }, { "epoch": 0.69, - "grad_norm": 8.385896682739258, + "grad_norm": 29.25188446044922, "learning_rate": 1.862498679902841e-05, - "loss": 1.4075, + "loss": 2.3534, "step": 6510 }, { "epoch": 0.69, - "grad_norm": 12.217638969421387, + "grad_norm": 26.823362350463867, "learning_rate": 1.862287464357377e-05, - "loss": 1.3037, + "loss": 2.0715, "step": 6520 }, { "epoch": 0.69, - "grad_norm": 11.871942520141602, + "grad_norm": 21.64516258239746, "learning_rate": 1.8620762488119127e-05, - "loss": 1.0957, + "loss": 1.9563, "step": 6530 }, { "epoch": 0.69, - "grad_norm": 7.420126914978027, + "grad_norm": 30.543567657470703, "learning_rate": 1.8618650332664484e-05, - "loss": 1.1717, + "loss": 2.149, "step": 6540 }, { "epoch": 0.69, - "grad_norm": 6.9478759765625, + "grad_norm": 31.988901138305664, "learning_rate": 1.8616538177209844e-05, - "loss": 1.3461, + "loss": 2.1788, "step": 6550 }, { "epoch": 0.69, - "grad_norm": 14.59031867980957, + "grad_norm": 32.18321990966797, "learning_rate": 1.8614426021755205e-05, - "loss": 1.5857, + "loss": 2.1223, "step": 6560 }, { "epoch": 0.69, - "grad_norm": 7.038922309875488, + "grad_norm": 36.934513092041016, "learning_rate": 1.8612313866300562e-05, - "loss": 1.2807, + "loss": 1.6964, "step": 6570 }, { "epoch": 0.69, - "grad_norm": 2.813819408416748, + "grad_norm": 25.787626266479492, "learning_rate": 1.861020171084592e-05, - "loss": 1.1948, + "loss": 2.0921, "step": 6580 }, { "epoch": 0.7, - "grad_norm": 8.099096298217773, + "grad_norm": 41.996856689453125, "learning_rate": 1.860808955539128e-05, - "loss": 1.4085, + "loss": 2.0952, "step": 6590 }, { "epoch": 0.7, - "grad_norm": 8.44278621673584, + "grad_norm": 25.088716506958008, "learning_rate": 1.8605977399936636e-05, - "loss": 1.484, + "loss": 2.0312, "step": 6600 }, { "epoch": 0.7, - "grad_norm": 13.267060279846191, + "grad_norm": 27.964380264282227, "learning_rate": 1.8603865244481997e-05, - "loss": 1.3108, + "loss": 1.9373, "step": 6610 }, { "epoch": 0.7, - "grad_norm": 14.009649276733398, + "grad_norm": 37.6336784362793, "learning_rate": 1.8601753089027354e-05, - "loss": 1.2642, + "loss": 2.1446, "step": 6620 }, { "epoch": 0.7, - "grad_norm": 14.777166366577148, + "grad_norm": 30.831531524658203, "learning_rate": 1.8599640933572714e-05, - "loss": 1.3427, + "loss": 2.1256, "step": 6630 }, { "epoch": 0.7, - "grad_norm": 8.386063575744629, + "grad_norm": 32.44020462036133, "learning_rate": 1.859752877811807e-05, - "loss": 1.0496, + "loss": 1.8309, "step": 6640 }, { "epoch": 0.7, - "grad_norm": 13.817797660827637, + "grad_norm": 33.515167236328125, "learning_rate": 1.859541662266343e-05, - "loss": 1.5073, + "loss": 2.0526, "step": 6650 }, { "epoch": 0.7, - "grad_norm": 9.927387237548828, + "grad_norm": 37.440853118896484, "learning_rate": 1.859330446720879e-05, - "loss": 1.2828, + "loss": 2.26, "step": 6660 }, { "epoch": 0.7, - "grad_norm": 10.81849479675293, + "grad_norm": 39.484580993652344, "learning_rate": 1.8591192311754146e-05, - "loss": 1.4441, + "loss": 2.0412, "step": 6670 }, { "epoch": 0.71, - "grad_norm": 4.754756450653076, + "grad_norm": 32.95443344116211, "learning_rate": 1.8589080156299507e-05, - "loss": 1.4203, + "loss": 2.2796, "step": 6680 }, { "epoch": 0.71, - "grad_norm": 7.900831699371338, + "grad_norm": 33.27894592285156, "learning_rate": 1.8586968000844864e-05, - "loss": 1.1603, + "loss": 2.2452, "step": 6690 }, { "epoch": 0.71, - "grad_norm": 6.3803558349609375, + "grad_norm": 34.50008010864258, "learning_rate": 1.858485584539022e-05, - "loss": 1.3117, + "loss": 2.0539, "step": 6700 }, { "epoch": 0.71, - "grad_norm": 7.269604206085205, + "grad_norm": 34.67900085449219, "learning_rate": 1.858274368993558e-05, - "loss": 1.2656, + "loss": 1.6957, "step": 6710 }, { "epoch": 0.71, - "grad_norm": 6.055761337280273, + "grad_norm": 36.68742370605469, "learning_rate": 1.858063153448094e-05, - "loss": 1.1542, + "loss": 1.9733, "step": 6720 }, { "epoch": 0.71, - "grad_norm": 13.706278800964355, + "grad_norm": 27.756471633911133, "learning_rate": 1.85785193790263e-05, - "loss": 1.2023, + "loss": 1.8199, "step": 6730 }, { "epoch": 0.71, - "grad_norm": 13.403399467468262, + "grad_norm": 34.999141693115234, "learning_rate": 1.8576407223571656e-05, - "loss": 1.205, + "loss": 2.1255, "step": 6740 }, { "epoch": 0.71, - "grad_norm": 12.264497756958008, + "grad_norm": 40.866397857666016, "learning_rate": 1.8574295068117013e-05, - "loss": 1.3278, + "loss": 2.491, "step": 6750 }, { "epoch": 0.71, - "grad_norm": 10.726388931274414, + "grad_norm": 28.621387481689453, "learning_rate": 1.8572182912662374e-05, - "loss": 1.402, + "loss": 2.0412, "step": 6760 }, { "epoch": 0.71, - "grad_norm": 5.774784088134766, + "grad_norm": 29.822046279907227, "learning_rate": 1.857007075720773e-05, - "loss": 1.3477, + "loss": 1.9522, "step": 6770 }, { "epoch": 0.72, - "grad_norm": 16.041484832763672, + "grad_norm": 27.456199645996094, "learning_rate": 1.856795860175309e-05, - "loss": 1.1509, + "loss": 2.1436, "step": 6780 }, { "epoch": 0.72, - "grad_norm": 7.96815299987793, + "grad_norm": 31.37904930114746, "learning_rate": 1.856584644629845e-05, - "loss": 1.1373, + "loss": 1.9034, "step": 6790 }, { "epoch": 0.72, - "grad_norm": 7.112144947052002, + "grad_norm": 35.18140411376953, "learning_rate": 1.8563734290843805e-05, - "loss": 1.1102, + "loss": 2.0383, "step": 6800 }, { "epoch": 0.72, - "grad_norm": 8.722577095031738, + "grad_norm": 19.225889205932617, "learning_rate": 1.8561622135389166e-05, - "loss": 1.0768, + "loss": 1.9168, "step": 6810 }, { "epoch": 0.72, - "grad_norm": 9.814340591430664, + "grad_norm": 34.393497467041016, "learning_rate": 1.8559509979934526e-05, - "loss": 1.3217, + "loss": 1.9842, "step": 6820 }, { "epoch": 0.72, - "grad_norm": 10.163887977600098, + "grad_norm": 30.105398178100586, "learning_rate": 1.8557397824479883e-05, - "loss": 1.4891, + "loss": 2.2066, "step": 6830 }, { "epoch": 0.72, - "grad_norm": 9.801732063293457, + "grad_norm": 35.22687530517578, "learning_rate": 1.8555285669025244e-05, - "loss": 1.28, + "loss": 2.0751, "step": 6840 }, { "epoch": 0.72, - "grad_norm": 9.17847728729248, + "grad_norm": 38.2816276550293, "learning_rate": 1.85531735135706e-05, - "loss": 1.156, + "loss": 2.0814, "step": 6850 }, { "epoch": 0.72, - "grad_norm": 14.413873672485352, + "grad_norm": 35.55842590332031, "learning_rate": 1.8551061358115958e-05, - "loss": 1.2961, + "loss": 2.009, "step": 6860 }, { "epoch": 0.73, - "grad_norm": 8.71690845489502, + "grad_norm": 29.256244659423828, "learning_rate": 1.854894920266132e-05, - "loss": 1.2921, + "loss": 2.1135, "step": 6870 }, { "epoch": 0.73, - "grad_norm": 12.123479843139648, + "grad_norm": 29.765527725219727, "learning_rate": 1.8546837047206675e-05, - "loss": 1.293, + "loss": 2.0951, "step": 6880 }, { "epoch": 0.73, - "grad_norm": 7.226715087890625, + "grad_norm": 32.80646514892578, "learning_rate": 1.8544724891752036e-05, - "loss": 1.1995, + "loss": 2.0006, "step": 6890 }, { "epoch": 0.73, - "grad_norm": 7.362088680267334, + "grad_norm": 36.698814392089844, "learning_rate": 1.8542612736297393e-05, - "loss": 1.2705, + "loss": 2.1387, "step": 6900 }, { "epoch": 0.73, - "grad_norm": 11.558103561401367, + "grad_norm": 25.46678924560547, "learning_rate": 1.854050058084275e-05, - "loss": 1.3206, + "loss": 2.1153, "step": 6910 }, { "epoch": 0.73, - "grad_norm": 13.46595287322998, + "grad_norm": 32.65917205810547, "learning_rate": 1.853838842538811e-05, - "loss": 1.3516, + "loss": 2.2037, "step": 6920 }, { "epoch": 0.73, - "grad_norm": 10.411704063415527, + "grad_norm": 35.410282135009766, "learning_rate": 1.8536276269933468e-05, - "loss": 1.4189, + "loss": 2.1412, "step": 6930 }, { "epoch": 0.73, - "grad_norm": 9.109715461730957, + "grad_norm": 27.354938507080078, "learning_rate": 1.8534164114478828e-05, - "loss": 1.1043, + "loss": 1.5508, "step": 6940 }, { "epoch": 0.73, - "grad_norm": 11.990567207336426, + "grad_norm": 39.78453826904297, "learning_rate": 1.8532051959024185e-05, - "loss": 1.4947, + "loss": 2.1046, "step": 6950 }, { "epoch": 0.74, - "grad_norm": 5.6650567054748535, + "grad_norm": 31.698274612426758, "learning_rate": 1.8529939803569542e-05, - "loss": 1.2142, + "loss": 1.9733, "step": 6960 }, { "epoch": 0.74, - "grad_norm": 7.559448719024658, + "grad_norm": 22.707645416259766, "learning_rate": 1.8527827648114903e-05, - "loss": 1.262, + "loss": 1.8669, "step": 6970 }, { "epoch": 0.74, - "grad_norm": 11.341790199279785, + "grad_norm": 30.391695022583008, "learning_rate": 1.852571549266026e-05, - "loss": 1.1049, + "loss": 1.7771, "step": 6980 }, { "epoch": 0.74, - "grad_norm": 14.587282180786133, + "grad_norm": 38.77577209472656, "learning_rate": 1.852360333720562e-05, - "loss": 1.392, + "loss": 2.0956, "step": 6990 }, { "epoch": 0.74, - "grad_norm": 6.731135845184326, + "grad_norm": 29.62995719909668, "learning_rate": 1.8521491181750977e-05, - "loss": 1.225, + "loss": 2.0615, "step": 7000 }, { "epoch": 0.74, - "grad_norm": 3.9373087882995605, + "grad_norm": 25.75437355041504, "learning_rate": 1.8519379026296338e-05, - "loss": 1.2123, + "loss": 1.9599, "step": 7010 }, { "epoch": 0.74, - "grad_norm": 4.037051677703857, + "grad_norm": 37.04179382324219, "learning_rate": 1.8517266870841695e-05, - "loss": 1.1646, + "loss": 1.8572, "step": 7020 }, { "epoch": 0.74, - "grad_norm": 8.981987953186035, + "grad_norm": 23.634994506835938, "learning_rate": 1.8515154715387052e-05, - "loss": 1.3289, + "loss": 2.5132, "step": 7030 }, { "epoch": 0.74, - "grad_norm": 14.0784330368042, + "grad_norm": 37.02146911621094, "learning_rate": 1.8513042559932413e-05, - "loss": 1.3547, + "loss": 2.142, "step": 7040 }, { "epoch": 0.74, - "grad_norm": 13.212225914001465, + "grad_norm": 43.559814453125, "learning_rate": 1.8510930404477773e-05, - "loss": 1.5344, + "loss": 2.0189, "step": 7050 }, { "epoch": 0.75, - "grad_norm": 6.992954730987549, + "grad_norm": 30.5493221282959, "learning_rate": 1.850881824902313e-05, - "loss": 1.2934, + "loss": 1.8845, "step": 7060 }, { "epoch": 0.75, - "grad_norm": 9.832418441772461, + "grad_norm": 33.368873596191406, "learning_rate": 1.8506706093568487e-05, - "loss": 1.3702, + "loss": 2.0826, "step": 7070 }, { "epoch": 0.75, - "grad_norm": 8.830916404724121, + "grad_norm": 31.309202194213867, "learning_rate": 1.8504593938113848e-05, - "loss": 1.2168, + "loss": 1.8581, "step": 7080 }, { "epoch": 0.75, - "grad_norm": 14.501062393188477, + "grad_norm": 29.995098114013672, "learning_rate": 1.8502481782659205e-05, - "loss": 1.2508, + "loss": 1.9471, "step": 7090 }, { "epoch": 0.75, - "grad_norm": 8.927953720092773, + "grad_norm": 28.962230682373047, "learning_rate": 1.8500369627204565e-05, - "loss": 1.1923, + "loss": 1.9477, "step": 7100 }, { "epoch": 0.75, - "grad_norm": 12.62042236328125, + "grad_norm": 32.511268615722656, "learning_rate": 1.8498257471749922e-05, - "loss": 1.0424, + "loss": 1.9588, "step": 7110 }, { "epoch": 0.75, - "grad_norm": 9.854493141174316, + "grad_norm": 35.78900909423828, "learning_rate": 1.849614531629528e-05, - "loss": 1.1277, + "loss": 1.8599, "step": 7120 }, { "epoch": 0.75, - "grad_norm": 11.561849594116211, + "grad_norm": 33.022891998291016, "learning_rate": 1.849403316084064e-05, - "loss": 1.4905, + "loss": 2.277, "step": 7130 }, { "epoch": 0.75, - "grad_norm": 6.862412452697754, + "grad_norm": 36.131553649902344, "learning_rate": 1.8491921005385997e-05, - "loss": 1.1322, + "loss": 1.8346, "step": 7140 }, { "epoch": 0.76, - "grad_norm": 7.934305191040039, + "grad_norm": 35.74323654174805, "learning_rate": 1.8489808849931357e-05, - "loss": 1.2991, + "loss": 2.0028, "step": 7150 }, { "epoch": 0.76, - "grad_norm": 7.644724369049072, + "grad_norm": 35.12614059448242, "learning_rate": 1.8487696694476714e-05, - "loss": 0.9743, + "loss": 1.8822, "step": 7160 }, { "epoch": 0.76, - "grad_norm": 14.359984397888184, + "grad_norm": 36.2025146484375, "learning_rate": 1.8485584539022075e-05, - "loss": 1.1338, + "loss": 2.086, "step": 7170 }, { "epoch": 0.76, - "grad_norm": 16.900066375732422, + "grad_norm": 32.33845520019531, "learning_rate": 1.8483472383567432e-05, - "loss": 1.2954, + "loss": 2.1825, "step": 7180 }, { "epoch": 0.76, - "grad_norm": 4.8188090324401855, + "grad_norm": 30.182844161987305, "learning_rate": 1.848136022811279e-05, - "loss": 0.774, + "loss": 1.7787, "step": 7190 }, { "epoch": 0.76, - "grad_norm": 7.282498359680176, + "grad_norm": 35.245121002197266, "learning_rate": 1.847924807265815e-05, - "loss": 1.2139, + "loss": 2.1961, "step": 7200 }, { "epoch": 0.76, - "grad_norm": 4.970002174377441, + "grad_norm": 28.705400466918945, "learning_rate": 1.8477135917203507e-05, - "loss": 1.0327, + "loss": 1.9368, "step": 7210 }, { "epoch": 0.76, - "grad_norm": 2.017434597015381, + "grad_norm": 30.084136962890625, "learning_rate": 1.8475023761748867e-05, - "loss": 1.2113, + "loss": 2.084, "step": 7220 }, { "epoch": 0.76, - "grad_norm": 8.364070892333984, + "grad_norm": 37.966304779052734, "learning_rate": 1.8472911606294224e-05, - "loss": 1.1203, + "loss": 1.8938, "step": 7230 }, { "epoch": 0.76, - "grad_norm": 8.727665901184082, + "grad_norm": 35.56214904785156, "learning_rate": 1.847079945083958e-05, - "loss": 1.1061, + "loss": 1.6398, "step": 7240 }, { "epoch": 0.77, - "grad_norm": 10.717944145202637, + "grad_norm": 27.417112350463867, "learning_rate": 1.8468687295384942e-05, - "loss": 1.2385, + "loss": 2.1119, "step": 7250 }, { "epoch": 0.77, - "grad_norm": 6.569893836975098, + "grad_norm": 31.365568161010742, "learning_rate": 1.8466575139930302e-05, - "loss": 1.2221, + "loss": 1.8825, "step": 7260 }, { "epoch": 0.77, - "grad_norm": 25.8283748626709, + "grad_norm": 44.90843200683594, "learning_rate": 1.846446298447566e-05, - "loss": 1.6143, + "loss": 2.04, "step": 7270 }, { "epoch": 0.77, - "grad_norm": 8.119101524353027, + "grad_norm": 31.055349349975586, "learning_rate": 1.846235082902102e-05, - "loss": 1.2241, + "loss": 1.7084, "step": 7280 }, { "epoch": 0.77, - "grad_norm": 10.044090270996094, + "grad_norm": 37.024574279785156, "learning_rate": 1.8460238673566374e-05, - "loss": 1.2659, + "loss": 2.0641, "step": 7290 }, { "epoch": 0.77, - "grad_norm": 8.283089637756348, + "grad_norm": 27.673830032348633, "learning_rate": 1.8458126518111734e-05, - "loss": 1.1296, + "loss": 1.9597, "step": 7300 }, { "epoch": 0.77, - "grad_norm": 2.6380081176757812, + "grad_norm": 27.421850204467773, "learning_rate": 1.8456014362657094e-05, - "loss": 1.3447, + "loss": 2.1005, "step": 7310 }, { "epoch": 0.77, - "grad_norm": 1.9262973070144653, + "grad_norm": 24.82141876220703, "learning_rate": 1.845390220720245e-05, - "loss": 1.2398, + "loss": 2.142, "step": 7320 }, { "epoch": 0.77, - "grad_norm": 12.756826400756836, + "grad_norm": 29.709720611572266, "learning_rate": 1.8451790051747812e-05, - "loss": 1.3954, + "loss": 1.908, "step": 7330 }, { "epoch": 0.78, - "grad_norm": 13.62343692779541, + "grad_norm": 26.14937400817871, "learning_rate": 1.844967789629317e-05, - "loss": 1.076, + "loss": 1.9677, "step": 7340 }, { "epoch": 0.78, - "grad_norm": 10.456056594848633, + "grad_norm": 34.78101348876953, "learning_rate": 1.8447565740838526e-05, - "loss": 1.5476, + "loss": 2.4481, "step": 7350 }, { "epoch": 0.78, - "grad_norm": 13.494622230529785, + "grad_norm": 25.06939125061035, "learning_rate": 1.8445453585383887e-05, - "loss": 1.1255, + "loss": 1.9219, "step": 7360 }, { "epoch": 0.78, - "grad_norm": 4.698188304901123, + "grad_norm": 23.133188247680664, "learning_rate": 1.8443341429929244e-05, - "loss": 1.0232, + "loss": 1.9904, "step": 7370 }, { "epoch": 0.78, - "grad_norm": 5.755089282989502, + "grad_norm": 32.1591682434082, "learning_rate": 1.8441229274474604e-05, - "loss": 1.3382, + "loss": 1.9734, "step": 7380 }, { "epoch": 0.78, - "grad_norm": 5.013108253479004, + "grad_norm": 37.96744918823242, "learning_rate": 1.843911711901996e-05, - "loss": 0.9817, + "loss": 1.9172, "step": 7390 }, { "epoch": 0.78, - "grad_norm": 9.106605529785156, + "grad_norm": 28.81180763244629, "learning_rate": 1.843700496356532e-05, - "loss": 1.0572, + "loss": 1.9543, "step": 7400 }, { "epoch": 0.78, - "grad_norm": 11.936727523803711, + "grad_norm": 38.932342529296875, "learning_rate": 1.843489280811068e-05, - "loss": 1.2984, + "loss": 2.2254, "step": 7410 }, { "epoch": 0.78, - "grad_norm": 7.5777268409729, + "grad_norm": 50.75, "learning_rate": 1.8432780652656036e-05, - "loss": 1.01, + "loss": 1.9536, "step": 7420 }, { "epoch": 0.78, - "grad_norm": 17.742109298706055, + "grad_norm": 35.167274475097656, "learning_rate": 1.8430668497201396e-05, - "loss": 1.1793, + "loss": 2.1387, "step": 7430 }, { "epoch": 0.79, - "grad_norm": 12.76982307434082, + "grad_norm": 35.508201599121094, "learning_rate": 1.8428556341746754e-05, - "loss": 1.0138, + "loss": 1.7561, "step": 7440 }, { "epoch": 0.79, - "grad_norm": 3.536748170852661, + "grad_norm": 24.41898536682129, "learning_rate": 1.842644418629211e-05, - "loss": 1.3183, + "loss": 1.5902, "step": 7450 }, { "epoch": 0.79, - "grad_norm": 12.267800331115723, + "grad_norm": 43.08258819580078, "learning_rate": 1.842433203083747e-05, - "loss": 1.2827, + "loss": 1.8683, "step": 7460 }, { "epoch": 0.79, - "grad_norm": 10.696211814880371, + "grad_norm": 45.06007766723633, "learning_rate": 1.8422219875382828e-05, - "loss": 1.0974, + "loss": 2.0891, "step": 7470 }, { "epoch": 0.79, - "grad_norm": 13.80179214477539, + "grad_norm": 33.48527145385742, "learning_rate": 1.842010771992819e-05, - "loss": 1.347, + "loss": 2.334, "step": 7480 }, { "epoch": 0.79, - "grad_norm": 38.90763854980469, + "grad_norm": 46.07976531982422, "learning_rate": 1.841799556447355e-05, - "loss": 1.0265, + "loss": 1.6093, "step": 7490 }, { "epoch": 0.79, - "grad_norm": 8.661056518554688, + "grad_norm": 29.43234634399414, "learning_rate": 1.8415883409018903e-05, - "loss": 0.9945, + "loss": 1.8475, "step": 7500 }, { "epoch": 0.79, - "grad_norm": 9.691649436950684, + "grad_norm": 33.19222640991211, "learning_rate": 1.8413771253564263e-05, - "loss": 0.9958, + "loss": 1.9617, "step": 7510 }, { "epoch": 0.79, - "grad_norm": 21.571884155273438, + "grad_norm": 36.641326904296875, "learning_rate": 1.8411659098109624e-05, - "loss": 1.2941, + "loss": 1.9612, "step": 7520 }, { "epoch": 0.8, - "grad_norm": 6.1519975662231445, + "grad_norm": 34.78205871582031, "learning_rate": 1.840954694265498e-05, - "loss": 1.2288, + "loss": 1.8551, "step": 7530 }, { "epoch": 0.8, - "grad_norm": 9.212127685546875, + "grad_norm": 25.921249389648438, "learning_rate": 1.840743478720034e-05, - "loss": 1.0942, + "loss": 1.7295, "step": 7540 }, { "epoch": 0.8, - "grad_norm": 4.905063152313232, + "grad_norm": 29.655881881713867, "learning_rate": 1.84053226317457e-05, - "loss": 1.0956, + "loss": 1.5903, "step": 7550 }, { "epoch": 0.8, - "grad_norm": 15.089709281921387, + "grad_norm": 34.18484878540039, "learning_rate": 1.8403210476291055e-05, - "loss": 1.2481, + "loss": 2.0081, "step": 7560 }, { "epoch": 0.8, - "grad_norm": 6.2300944328308105, + "grad_norm": 22.45371437072754, "learning_rate": 1.8401098320836416e-05, - "loss": 1.3062, + "loss": 2.0251, "step": 7570 }, { "epoch": 0.8, - "grad_norm": 12.377196311950684, + "grad_norm": 28.114816665649414, "learning_rate": 1.8398986165381773e-05, - "loss": 1.0669, + "loss": 2.0058, "step": 7580 }, { "epoch": 0.8, - "grad_norm": 4.9585676193237305, + "grad_norm": 22.138607025146484, "learning_rate": 1.8396874009927134e-05, - "loss": 0.9731, + "loss": 1.8696, "step": 7590 }, { "epoch": 0.8, - "grad_norm": 8.755257606506348, + "grad_norm": 33.922218322753906, "learning_rate": 1.839476185447249e-05, - "loss": 1.1411, + "loss": 2.0964, "step": 7600 }, { "epoch": 0.8, - "grad_norm": 9.328500747680664, + "grad_norm": 34.3865966796875, "learning_rate": 1.8392649699017848e-05, - "loss": 1.2918, + "loss": 2.2832, "step": 7610 }, { "epoch": 0.8, - "grad_norm": 24.23015022277832, + "grad_norm": 33.95296859741211, "learning_rate": 1.8390537543563208e-05, - "loss": 1.3502, + "loss": 1.9381, "step": 7620 }, { "epoch": 0.81, - "grad_norm": 11.302351951599121, + "grad_norm": 50.483375549316406, "learning_rate": 1.8388425388108565e-05, - "loss": 1.275, + "loss": 2.1434, "step": 7630 }, { "epoch": 0.81, - "grad_norm": 4.829554080963135, + "grad_norm": 33.51721954345703, "learning_rate": 1.8386313232653926e-05, - "loss": 1.0375, + "loss": 1.7756, "step": 7640 }, { "epoch": 0.81, - "grad_norm": 4.567259311676025, + "grad_norm": 22.421119689941406, "learning_rate": 1.8384201077199283e-05, - "loss": 0.9553, + "loss": 1.9686, "step": 7650 }, { "epoch": 0.81, - "grad_norm": 12.602558135986328, + "grad_norm": 32.200191497802734, "learning_rate": 1.8382088921744643e-05, - "loss": 1.2054, + "loss": 1.9238, "step": 7660 }, { "epoch": 0.81, - "grad_norm": 13.721440315246582, + "grad_norm": 26.266874313354492, "learning_rate": 1.837997676629e-05, - "loss": 1.0701, + "loss": 1.8243, "step": 7670 }, { "epoch": 0.81, - "grad_norm": 8.3406982421875, + "grad_norm": 31.513620376586914, "learning_rate": 1.8377864610835357e-05, - "loss": 0.9432, + "loss": 1.6268, "step": 7680 }, { "epoch": 0.81, - "grad_norm": 10.94404125213623, + "grad_norm": 27.950664520263672, "learning_rate": 1.8375752455380718e-05, - "loss": 1.7777, + "loss": 2.4137, "step": 7690 }, { "epoch": 0.81, - "grad_norm": 15.533305168151855, + "grad_norm": 35.261268615722656, "learning_rate": 1.8373640299926075e-05, - "loss": 1.3733, + "loss": 2.1298, "step": 7700 }, { "epoch": 0.81, - "grad_norm": 15.45605182647705, + "grad_norm": 25.945266723632812, "learning_rate": 1.8371528144471435e-05, - "loss": 1.3069, + "loss": 1.8196, "step": 7710 }, { "epoch": 0.82, - "grad_norm": 15.427933692932129, + "grad_norm": 28.26560401916504, "learning_rate": 1.8369415989016793e-05, - "loss": 1.2179, + "loss": 2.0909, "step": 7720 }, { "epoch": 0.82, - "grad_norm": 7.941003799438477, + "grad_norm": 47.56690979003906, "learning_rate": 1.836730383356215e-05, - "loss": 1.3548, + "loss": 1.963, "step": 7730 }, { "epoch": 0.82, - "grad_norm": 13.818915367126465, + "grad_norm": 37.06293869018555, "learning_rate": 1.836519167810751e-05, - "loss": 1.1824, + "loss": 1.7753, "step": 7740 }, { "epoch": 0.82, - "grad_norm": 4.374635219573975, + "grad_norm": 30.814682006835938, "learning_rate": 1.836307952265287e-05, - "loss": 1.1431, + "loss": 2.2807, "step": 7750 }, { "epoch": 0.82, - "grad_norm": 10.103631019592285, + "grad_norm": 51.307552337646484, "learning_rate": 1.8360967367198228e-05, - "loss": 1.2367, + "loss": 1.9868, "step": 7760 }, { "epoch": 0.82, - "grad_norm": 14.501548767089844, + "grad_norm": 36.86030578613281, "learning_rate": 1.8358855211743585e-05, - "loss": 1.0736, + "loss": 1.9974, "step": 7770 }, { "epoch": 0.82, - "grad_norm": 6.724035739898682, + "grad_norm": 32.93803787231445, "learning_rate": 1.8356743056288945e-05, - "loss": 1.2658, + "loss": 1.9828, "step": 7780 }, { "epoch": 0.82, - "grad_norm": 7.853983402252197, + "grad_norm": 37.723018646240234, "learning_rate": 1.8354630900834302e-05, - "loss": 1.3818, + "loss": 2.1318, "step": 7790 }, { "epoch": 0.82, - "grad_norm": 10.342225074768066, + "grad_norm": 35.1579704284668, "learning_rate": 1.8352518745379663e-05, - "loss": 1.0687, + "loss": 2.0778, "step": 7800 }, { "epoch": 0.82, - "grad_norm": 6.679742336273193, + "grad_norm": 41.96892166137695, "learning_rate": 1.835040658992502e-05, - "loss": 1.4182, + "loss": 1.9688, "step": 7810 }, { "epoch": 0.83, - "grad_norm": 20.082632064819336, + "grad_norm": 40.78770446777344, "learning_rate": 1.834829443447038e-05, - "loss": 1.3928, + "loss": 2.2528, "step": 7820 }, { "epoch": 0.83, - "grad_norm": 11.251479148864746, + "grad_norm": 24.561710357666016, "learning_rate": 1.8346182279015737e-05, - "loss": 1.0916, + "loss": 1.8161, "step": 7830 }, { "epoch": 0.83, - "grad_norm": 6.726846694946289, + "grad_norm": 48.103172302246094, "learning_rate": 1.8344070123561095e-05, - "loss": 1.2678, + "loss": 2.0332, "step": 7840 }, { "epoch": 0.83, - "grad_norm": 6.8422770500183105, + "grad_norm": 41.140899658203125, "learning_rate": 1.8341957968106455e-05, - "loss": 1.0111, + "loss": 2.023, "step": 7850 }, { "epoch": 0.83, - "grad_norm": 9.898163795471191, + "grad_norm": 54.612709045410156, "learning_rate": 1.8339845812651812e-05, - "loss": 0.9564, + "loss": 1.6406, "step": 7860 }, { "epoch": 0.83, - "grad_norm": 10.847105026245117, + "grad_norm": 43.91957473754883, "learning_rate": 1.8337733657197173e-05, - "loss": 1.0207, + "loss": 2.0418, "step": 7870 }, { "epoch": 0.83, - "grad_norm": 2.325425386428833, + "grad_norm": 34.449195861816406, "learning_rate": 1.833562150174253e-05, - "loss": 1.0375, + "loss": 1.6995, "step": 7880 }, { "epoch": 0.83, - "grad_norm": 15.072937965393066, + "grad_norm": 36.76914596557617, "learning_rate": 1.8333509346287887e-05, - "loss": 1.2212, + "loss": 2.2104, "step": 7890 }, { "epoch": 0.83, - "grad_norm": 8.44444751739502, + "grad_norm": 29.128965377807617, "learning_rate": 1.8331397190833247e-05, - "loss": 1.25, + "loss": 1.9063, "step": 7900 }, { "epoch": 0.84, - "grad_norm": 2.2279345989227295, + "grad_norm": 24.172090530395508, "learning_rate": 1.8329285035378604e-05, - "loss": 1.2133, + "loss": 2.0037, "step": 7910 }, { "epoch": 0.84, - "grad_norm": 3.449772596359253, + "grad_norm": 43.563297271728516, "learning_rate": 1.8327172879923965e-05, - "loss": 1.3727, + "loss": 2.1452, "step": 7920 }, { "epoch": 0.84, - "grad_norm": 5.946283340454102, + "grad_norm": 35.62562561035156, "learning_rate": 1.8325060724469322e-05, - "loss": 1.0653, + "loss": 1.9728, "step": 7930 }, { "epoch": 0.84, - "grad_norm": 4.145539283752441, + "grad_norm": 26.951725006103516, "learning_rate": 1.832294856901468e-05, - "loss": 0.989, + "loss": 1.8833, "step": 7940 }, { "epoch": 0.84, - "grad_norm": 4.917052745819092, + "grad_norm": 36.30305099487305, "learning_rate": 1.832083641356004e-05, - "loss": 1.1583, + "loss": 1.8967, "step": 7950 }, { "epoch": 0.84, - "grad_norm": 9.028826713562012, + "grad_norm": 35.85757827758789, "learning_rate": 1.8318724258105396e-05, - "loss": 1.1965, + "loss": 2.0127, "step": 7960 }, { "epoch": 0.84, - "grad_norm": 13.556591987609863, + "grad_norm": 36.283695220947266, "learning_rate": 1.8316612102650757e-05, - "loss": 1.1176, + "loss": 1.751, "step": 7970 }, { "epoch": 0.84, - "grad_norm": 10.197393417358398, + "grad_norm": 36.71406936645508, "learning_rate": 1.8314499947196117e-05, - "loss": 1.051, + "loss": 1.6313, "step": 7980 }, { "epoch": 0.84, - "grad_norm": 14.466920852661133, + "grad_norm": 43.42887878417969, "learning_rate": 1.831238779174147e-05, - "loss": 0.8648, + "loss": 2.0775, "step": 7990 }, { "epoch": 0.84, - "grad_norm": 8.651012420654297, + "grad_norm": 22.672208786010742, "learning_rate": 1.831027563628683e-05, - "loss": 1.04, + "loss": 1.6654, "step": 8000 }, { "epoch": 0.85, - "grad_norm": 18.27894401550293, + "grad_norm": 41.7623176574707, "learning_rate": 1.8308163480832192e-05, - "loss": 1.2641, + "loss": 2.3069, "step": 8010 }, { "epoch": 0.85, - "grad_norm": 18.72589683532715, + "grad_norm": 32.01502990722656, "learning_rate": 1.830605132537755e-05, - "loss": 1.0433, + "loss": 1.755, "step": 8020 }, { "epoch": 0.85, - "grad_norm": 15.030464172363281, + "grad_norm": 29.79570960998535, "learning_rate": 1.830393916992291e-05, - "loss": 0.9419, + "loss": 1.9371, "step": 8030 }, { "epoch": 0.85, - "grad_norm": 8.287261962890625, + "grad_norm": 34.83698272705078, "learning_rate": 1.8301827014468267e-05, - "loss": 1.1986, + "loss": 1.9525, "step": 8040 }, { "epoch": 0.85, - "grad_norm": 14.850321769714355, + "grad_norm": 21.18808937072754, "learning_rate": 1.8299714859013624e-05, - "loss": 0.8716, + "loss": 1.6602, "step": 8050 }, { "epoch": 0.85, - "grad_norm": 20.98737907409668, + "grad_norm": 41.08024215698242, "learning_rate": 1.8297602703558984e-05, - "loss": 0.8848, + "loss": 1.6801, "step": 8060 }, { "epoch": 0.85, - "grad_norm": 13.392731666564941, + "grad_norm": 29.642536163330078, "learning_rate": 1.829549054810434e-05, - "loss": 1.132, + "loss": 1.935, "step": 8070 }, { "epoch": 0.85, - "grad_norm": 17.77458381652832, + "grad_norm": 34.46763229370117, "learning_rate": 1.8293378392649702e-05, - "loss": 1.157, + "loss": 1.8325, "step": 8080 }, { "epoch": 0.85, - "grad_norm": 10.836054801940918, + "grad_norm": 34.831275939941406, "learning_rate": 1.829126623719506e-05, - "loss": 1.2009, + "loss": 2.0091, "step": 8090 }, { "epoch": 0.86, - "grad_norm": 16.844730377197266, + "grad_norm": 35.46916961669922, "learning_rate": 1.8289154081740416e-05, - "loss": 1.4949, + "loss": 2.2936, "step": 8100 }, { "epoch": 0.86, - "grad_norm": 16.71429443359375, + "grad_norm": 44.524314880371094, "learning_rate": 1.8287041926285776e-05, - "loss": 1.3342, + "loss": 2.149, "step": 8110 }, { "epoch": 0.86, - "grad_norm": 3.414501667022705, + "grad_norm": 40.089263916015625, "learning_rate": 1.8284929770831134e-05, - "loss": 1.031, + "loss": 1.7697, "step": 8120 }, { "epoch": 0.86, - "grad_norm": 7.506906032562256, + "grad_norm": 36.7569580078125, "learning_rate": 1.8282817615376494e-05, - "loss": 1.1584, + "loss": 1.9418, "step": 8130 }, { "epoch": 0.86, - "grad_norm": 9.24974536895752, + "grad_norm": 40.43549346923828, "learning_rate": 1.828070545992185e-05, - "loss": 1.0379, + "loss": 1.9952, "step": 8140 }, { "epoch": 0.86, - "grad_norm": 9.322461128234863, + "grad_norm": 36.92317581176758, "learning_rate": 1.8278593304467208e-05, - "loss": 1.3773, + "loss": 1.9433, "step": 8150 }, { "epoch": 0.86, - "grad_norm": 8.157588958740234, + "grad_norm": 38.00996017456055, "learning_rate": 1.827648114901257e-05, - "loss": 0.8251, + "loss": 1.84, "step": 8160 }, { "epoch": 0.86, - "grad_norm": 15.220280647277832, + "grad_norm": 38.39075469970703, "learning_rate": 1.8274368993557926e-05, - "loss": 1.1546, + "loss": 2.1213, "step": 8170 }, { "epoch": 0.86, - "grad_norm": 16.918134689331055, + "grad_norm": 32.21208572387695, "learning_rate": 1.8272256838103286e-05, - "loss": 1.0976, + "loss": 2.2226, "step": 8180 }, { "epoch": 0.86, - "grad_norm": 4.283835411071777, + "grad_norm": 36.365482330322266, "learning_rate": 1.8270144682648647e-05, - "loss": 1.0336, + "loss": 1.852, "step": 8190 }, { "epoch": 0.87, - "grad_norm": 8.154387474060059, + "grad_norm": 31.608808517456055, "learning_rate": 1.8268032527194004e-05, - "loss": 1.3617, + "loss": 2.2223, "step": 8200 }, { "epoch": 0.87, - "grad_norm": 7.940045356750488, + "grad_norm": 37.90910720825195, "learning_rate": 1.826592037173936e-05, - "loss": 1.1156, + "loss": 2.1112, "step": 8210 }, { "epoch": 0.87, - "grad_norm": 8.040996551513672, + "grad_norm": 36.4744758605957, "learning_rate": 1.826380821628472e-05, - "loss": 1.1237, + "loss": 2.1426, "step": 8220 }, { "epoch": 0.87, - "grad_norm": 11.65139102935791, + "grad_norm": 24.88102149963379, "learning_rate": 1.826169606083008e-05, - "loss": 1.0959, + "loss": 1.9337, "step": 8230 }, { "epoch": 0.87, - "grad_norm": 11.080164909362793, + "grad_norm": 36.81837844848633, "learning_rate": 1.825958390537544e-05, - "loss": 1.1874, + "loss": 2.04, "step": 8240 }, { "epoch": 0.87, - "grad_norm": 16.9500789642334, + "grad_norm": 31.4532527923584, "learning_rate": 1.8257471749920796e-05, - "loss": 1.0976, + "loss": 1.6629, "step": 8250 }, { "epoch": 0.87, - "grad_norm": 6.0843377113342285, + "grad_norm": 26.379745483398438, "learning_rate": 1.8255359594466153e-05, - "loss": 0.8384, + "loss": 1.9068, "step": 8260 }, { "epoch": 0.87, - "grad_norm": 6.501587390899658, + "grad_norm": 38.392181396484375, "learning_rate": 1.8253247439011514e-05, - "loss": 1.1111, + "loss": 2.0009, "step": 8270 }, { "epoch": 0.87, - "grad_norm": 1.593220829963684, + "grad_norm": 30.92724609375, "learning_rate": 1.825113528355687e-05, - "loss": 1.1617, + "loss": 1.8866, "step": 8280 }, { "epoch": 0.88, - "grad_norm": 6.7084059715271, + "grad_norm": 44.44887924194336, "learning_rate": 1.824902312810223e-05, - "loss": 1.0668, + "loss": 1.6017, "step": 8290 }, { "epoch": 0.88, - "grad_norm": 19.22061538696289, + "grad_norm": 49.37348556518555, "learning_rate": 1.8246910972647588e-05, - "loss": 0.9455, + "loss": 1.7612, "step": 8300 }, { "epoch": 0.88, - "grad_norm": 4.384603023529053, + "grad_norm": 27.587724685668945, "learning_rate": 1.824479881719295e-05, - "loss": 1.019, + "loss": 1.7116, "step": 8310 }, { "epoch": 0.88, - "grad_norm": 18.05158042907715, + "grad_norm": 44.49518585205078, "learning_rate": 1.8242686661738306e-05, - "loss": 0.9827, + "loss": 2.0204, "step": 8320 }, { "epoch": 0.88, - "grad_norm": 123.507080078125, + "grad_norm": 39.1076545715332, "learning_rate": 1.8240574506283663e-05, - "loss": 1.0934, + "loss": 1.7523, "step": 8330 }, { "epoch": 0.88, - "grad_norm": 16.06003761291504, + "grad_norm": 32.47861862182617, "learning_rate": 1.8238462350829023e-05, - "loss": 1.3577, + "loss": 2.0804, "step": 8340 }, { "epoch": 0.88, - "grad_norm": 4.878872394561768, + "grad_norm": 34.698917388916016, "learning_rate": 1.823635019537438e-05, - "loss": 0.8514, + "loss": 1.8546, "step": 8350 }, { "epoch": 0.88, - "grad_norm": 17.248291015625, + "grad_norm": 36.107696533203125, "learning_rate": 1.823423803991974e-05, - "loss": 1.1545, + "loss": 1.9841, "step": 8360 }, { "epoch": 0.88, - "grad_norm": 12.62728500366211, + "grad_norm": 44.70740509033203, "learning_rate": 1.8232125884465098e-05, - "loss": 1.063, + "loss": 1.8434, "step": 8370 }, { "epoch": 0.88, - "grad_norm": 15.633353233337402, + "grad_norm": 36.1932373046875, "learning_rate": 1.8230013729010455e-05, - "loss": 1.284, + "loss": 2.1472, "step": 8380 }, { "epoch": 0.89, - "grad_norm": 13.022372245788574, + "grad_norm": 31.487503051757812, "learning_rate": 1.8227901573555815e-05, - "loss": 1.0486, + "loss": 1.6798, "step": 8390 }, { "epoch": 0.89, - "grad_norm": 14.99155330657959, + "grad_norm": 28.080480575561523, "learning_rate": 1.8225789418101173e-05, - "loss": 1.0251, + "loss": 1.7572, "step": 8400 }, { "epoch": 0.89, - "grad_norm": 8.232471466064453, + "grad_norm": 33.2248649597168, "learning_rate": 1.8223677262646533e-05, - "loss": 1.2839, + "loss": 2.1979, "step": 8410 }, { "epoch": 0.89, - "grad_norm": 9.818863868713379, + "grad_norm": 39.34006118774414, "learning_rate": 1.822156510719189e-05, - "loss": 0.8293, + "loss": 1.6996, "step": 8420 }, { "epoch": 0.89, - "grad_norm": 14.735958099365234, + "grad_norm": 38.50984573364258, "learning_rate": 1.8219452951737247e-05, - "loss": 1.1606, + "loss": 1.8913, "step": 8430 }, { "epoch": 0.89, - "grad_norm": 9.972173690795898, + "grad_norm": 44.67124557495117, "learning_rate": 1.8217340796282608e-05, - "loss": 1.0234, + "loss": 2.0983, "step": 8440 }, { "epoch": 0.89, - "grad_norm": 11.392930030822754, + "grad_norm": 41.231727600097656, "learning_rate": 1.8215228640827968e-05, - "loss": 1.2103, + "loss": 1.8405, "step": 8450 }, { "epoch": 0.89, - "grad_norm": 13.492349624633789, + "grad_norm": 33.08424377441406, "learning_rate": 1.8213116485373325e-05, - "loss": 0.8947, + "loss": 1.7023, "step": 8460 }, { "epoch": 0.89, - "grad_norm": 5.033742427825928, + "grad_norm": 38.3984260559082, "learning_rate": 1.8211004329918686e-05, - "loss": 1.0596, + "loss": 1.9988, "step": 8470 }, { "epoch": 0.9, - "grad_norm": 15.48432445526123, + "grad_norm": 42.31736755371094, "learning_rate": 1.8208892174464043e-05, - "loss": 1.2274, + "loss": 2.0666, "step": 8480 }, { "epoch": 0.9, - "grad_norm": 7.0199127197265625, + "grad_norm": 23.27914810180664, "learning_rate": 1.82067800190094e-05, - "loss": 0.8533, + "loss": 1.8482, "step": 8490 }, { "epoch": 0.9, - "grad_norm": 10.697369575500488, + "grad_norm": 30.171268463134766, "learning_rate": 1.820466786355476e-05, - "loss": 1.249, + "loss": 2.1727, "step": 8500 }, { "epoch": 0.9, - "grad_norm": 10.533666610717773, + "grad_norm": 41.59688186645508, "learning_rate": 1.8202555708100117e-05, - "loss": 0.929, + "loss": 1.9506, "step": 8510 }, { "epoch": 0.9, - "grad_norm": 15.10594654083252, + "grad_norm": 39.62214279174805, "learning_rate": 1.8200443552645478e-05, - "loss": 1.0849, + "loss": 2.0035, "step": 8520 }, { "epoch": 0.9, - "grad_norm": 8.68903636932373, + "grad_norm": 40.74443435668945, "learning_rate": 1.8198331397190835e-05, - "loss": 0.8936, + "loss": 1.8727, "step": 8530 }, { "epoch": 0.9, - "grad_norm": 16.840055465698242, + "grad_norm": 39.164207458496094, "learning_rate": 1.8196219241736192e-05, - "loss": 0.9769, + "loss": 1.7777, "step": 8540 }, { "epoch": 0.9, - "grad_norm": 3.4210445880889893, + "grad_norm": 29.553573608398438, "learning_rate": 1.8194107086281553e-05, - "loss": 1.019, + "loss": 1.7494, "step": 8550 }, { "epoch": 0.9, - "grad_norm": 16.629032135009766, + "grad_norm": 40.14094924926758, "learning_rate": 1.819199493082691e-05, - "loss": 0.9456, + "loss": 1.9804, "step": 8560 }, { "epoch": 0.91, - "grad_norm": 16.226999282836914, + "grad_norm": 41.73072052001953, "learning_rate": 1.818988277537227e-05, - "loss": 1.1678, + "loss": 1.8636, "step": 8570 }, { "epoch": 0.91, - "grad_norm": 31.28521728515625, + "grad_norm": 43.81443405151367, "learning_rate": 1.8187770619917627e-05, - "loss": 1.0854, + "loss": 2.1283, "step": 8580 }, { "epoch": 0.91, - "grad_norm": 5.128644943237305, + "grad_norm": 33.16176223754883, "learning_rate": 1.8185658464462984e-05, - "loss": 1.2482, + "loss": 1.9756, "step": 8590 }, { "epoch": 0.91, - "grad_norm": 25.05236053466797, + "grad_norm": 32.219303131103516, "learning_rate": 1.8183546309008345e-05, - "loss": 1.1421, + "loss": 2.0102, "step": 8600 }, { "epoch": 0.91, - "grad_norm": 2.7238047122955322, + "grad_norm": 34.06946563720703, "learning_rate": 1.8181434153553702e-05, - "loss": 0.9756, + "loss": 1.922, "step": 8610 }, { "epoch": 0.91, - "grad_norm": 5.601781368255615, + "grad_norm": 28.415571212768555, "learning_rate": 1.8179321998099062e-05, - "loss": 1.0356, + "loss": 1.809, "step": 8620 }, { "epoch": 0.91, - "grad_norm": 4.290815830230713, + "grad_norm": 24.24643898010254, "learning_rate": 1.817720984264442e-05, - "loss": 1.0228, + "loss": 1.7793, "step": 8630 }, { "epoch": 0.91, - "grad_norm": 21.64484214782715, + "grad_norm": 29.14616584777832, "learning_rate": 1.8175097687189776e-05, - "loss": 1.2587, + "loss": 2.1262, "step": 8640 }, { "epoch": 0.91, - "grad_norm": 10.753555297851562, + "grad_norm": 28.68203353881836, "learning_rate": 1.8172985531735137e-05, - "loss": 1.0479, + "loss": 1.9365, "step": 8650 }, { "epoch": 0.91, - "grad_norm": 22.790409088134766, + "grad_norm": 34.11137390136719, "learning_rate": 1.8170873376280494e-05, - "loss": 0.8303, + "loss": 1.8576, "step": 8660 }, { "epoch": 0.92, - "grad_norm": 3.544942855834961, + "grad_norm": 34.68653106689453, "learning_rate": 1.8168761220825855e-05, - "loss": 0.8951, + "loss": 2.0194, "step": 8670 }, { "epoch": 0.92, - "grad_norm": 10.641388893127441, + "grad_norm": 32.955055236816406, "learning_rate": 1.8166649065371215e-05, - "loss": 0.956, + "loss": 1.7931, "step": 8680 }, { "epoch": 0.92, - "grad_norm": 14.052702903747559, + "grad_norm": 38.22832107543945, "learning_rate": 1.816453690991657e-05, - "loss": 1.196, + "loss": 1.8817, "step": 8690 }, { "epoch": 0.92, - "grad_norm": 7.209864139556885, + "grad_norm": 37.79623794555664, "learning_rate": 1.816242475446193e-05, - "loss": 1.2008, + "loss": 1.8164, "step": 8700 }, { "epoch": 0.92, - "grad_norm": 7.613988876342773, + "grad_norm": 35.16669845581055, "learning_rate": 1.816031259900729e-05, - "loss": 1.3094, + "loss": 2.1042, "step": 8710 }, { "epoch": 0.92, - "grad_norm": 9.651299476623535, + "grad_norm": 34.19347381591797, "learning_rate": 1.8158200443552647e-05, - "loss": 0.8372, + "loss": 1.8695, "step": 8720 }, { "epoch": 0.92, - "grad_norm": 12.507035255432129, + "grad_norm": 40.797996520996094, "learning_rate": 1.8156088288098007e-05, - "loss": 0.7573, + "loss": 1.6447, "step": 8730 }, { "epoch": 0.92, - "grad_norm": 17.402873992919922, + "grad_norm": 29.238388061523438, "learning_rate": 1.8153976132643364e-05, - "loss": 1.0506, + "loss": 1.9742, "step": 8740 }, { "epoch": 0.92, - "grad_norm": 10.90021800994873, + "grad_norm": 35.680419921875, "learning_rate": 1.815186397718872e-05, - "loss": 1.0727, + "loss": 1.7938, "step": 8750 }, { "epoch": 0.93, - "grad_norm": 15.169095993041992, + "grad_norm": 27.588232040405273, "learning_rate": 1.8149751821734082e-05, - "loss": 1.2162, + "loss": 1.9942, "step": 8760 }, { "epoch": 0.93, - "grad_norm": 15.349626541137695, + "grad_norm": 36.73896408081055, "learning_rate": 1.814763966627944e-05, - "loss": 0.9919, + "loss": 2.0655, "step": 8770 }, { "epoch": 0.93, - "grad_norm": 11.213387489318848, + "grad_norm": 50.75896072387695, "learning_rate": 1.81455275108248e-05, - "loss": 1.0656, + "loss": 2.1662, "step": 8780 }, { "epoch": 0.93, - "grad_norm": 16.506011962890625, + "grad_norm": 39.49610137939453, "learning_rate": 1.8143415355370156e-05, - "loss": 1.1361, + "loss": 1.8451, "step": 8790 }, { "epoch": 0.93, - "grad_norm": 16.609285354614258, + "grad_norm": 36.41221618652344, "learning_rate": 1.8141303199915514e-05, - "loss": 1.0709, + "loss": 2.1275, "step": 8800 }, { "epoch": 0.93, - "grad_norm": 10.91037368774414, + "grad_norm": 31.62007713317871, "learning_rate": 1.8139191044460874e-05, - "loss": 0.8814, + "loss": 1.6726, "step": 8810 }, { "epoch": 0.93, - "grad_norm": 8.451621055603027, + "grad_norm": 35.86500930786133, "learning_rate": 1.813707888900623e-05, - "loss": 0.8373, + "loss": 1.6211, "step": 8820 }, { "epoch": 0.93, - "grad_norm": 7.39933967590332, + "grad_norm": 39.96851348876953, "learning_rate": 1.813496673355159e-05, - "loss": 1.3061, + "loss": 2.0158, "step": 8830 }, { "epoch": 0.93, - "grad_norm": 13.297159194946289, + "grad_norm": 45.62693786621094, "learning_rate": 1.813285457809695e-05, - "loss": 0.857, + "loss": 1.7271, "step": 8840 }, { "epoch": 0.93, - "grad_norm": 28.40963363647461, + "grad_norm": 21.833742141723633, "learning_rate": 1.813074242264231e-05, - "loss": 1.1364, + "loss": 1.7488, "step": 8850 }, { "epoch": 0.94, - "grad_norm": 12.515909194946289, + "grad_norm": 37.59640884399414, "learning_rate": 1.8128630267187666e-05, - "loss": 0.9211, + "loss": 1.7507, "step": 8860 }, { "epoch": 0.94, - "grad_norm": 13.278634071350098, + "grad_norm": 45.344234466552734, "learning_rate": 1.8126518111733023e-05, - "loss": 0.9068, + "loss": 1.7336, "step": 8870 }, { "epoch": 0.94, - "grad_norm": 7.6605072021484375, + "grad_norm": 29.17068099975586, "learning_rate": 1.8124405956278384e-05, - "loss": 0.987, + "loss": 1.9433, "step": 8880 }, { "epoch": 0.94, - "grad_norm": 6.391233444213867, + "grad_norm": 32.12074279785156, "learning_rate": 1.812229380082374e-05, - "loss": 1.1507, + "loss": 1.7822, "step": 8890 }, { "epoch": 0.94, - "grad_norm": 11.281806945800781, + "grad_norm": 28.935245513916016, "learning_rate": 1.81201816453691e-05, - "loss": 1.2902, + "loss": 2.1403, "step": 8900 }, { "epoch": 0.94, - "grad_norm": 8.359882354736328, + "grad_norm": 26.783170700073242, "learning_rate": 1.811806948991446e-05, - "loss": 0.8335, + "loss": 1.7006, "step": 8910 }, { "epoch": 0.94, - "grad_norm": 12.315272331237793, + "grad_norm": 38.7705078125, "learning_rate": 1.8115957334459816e-05, - "loss": 1.233, + "loss": 2.0558, "step": 8920 }, { "epoch": 0.94, - "grad_norm": 12.080918312072754, + "grad_norm": 42.22845458984375, "learning_rate": 1.8113845179005176e-05, - "loss": 1.2313, + "loss": 2.2094, "step": 8930 }, { "epoch": 0.94, - "grad_norm": 6.542693614959717, + "grad_norm": 28.66451644897461, "learning_rate": 1.8111733023550536e-05, - "loss": 1.15, + "loss": 1.9692, "step": 8940 }, { "epoch": 0.95, - "grad_norm": 17.1704158782959, + "grad_norm": 31.40700340270996, "learning_rate": 1.8109620868095894e-05, - "loss": 0.8737, + "loss": 1.8259, "step": 8950 }, { "epoch": 0.95, - "grad_norm": 10.47104549407959, + "grad_norm": 46.37440872192383, "learning_rate": 1.8107508712641254e-05, - "loss": 0.7997, + "loss": 2.1634, "step": 8960 }, { "epoch": 0.95, - "grad_norm": 12.911360740661621, + "grad_norm": 39.23902130126953, "learning_rate": 1.810539655718661e-05, - "loss": 1.179, + "loss": 1.93, "step": 8970 }, { "epoch": 0.95, - "grad_norm": 7.193451881408691, + "grad_norm": 31.852327346801758, "learning_rate": 1.8103284401731968e-05, - "loss": 1.1081, + "loss": 1.8747, "step": 8980 }, { "epoch": 0.95, - "grad_norm": 9.427313804626465, + "grad_norm": 33.352691650390625, "learning_rate": 1.810117224627733e-05, - "loss": 1.2346, + "loss": 2.0373, "step": 8990 }, { "epoch": 0.95, - "grad_norm": 12.715041160583496, + "grad_norm": 34.527305603027344, "learning_rate": 1.8099060090822686e-05, - "loss": 1.3062, + "loss": 1.8524, "step": 9000 }, { "epoch": 0.95, - "grad_norm": 11.853886604309082, + "grad_norm": 34.40365982055664, "learning_rate": 1.8096947935368046e-05, - "loss": 0.7089, + "loss": 1.5824, "step": 9010 }, { "epoch": 0.95, - "grad_norm": 21.77123260498047, + "grad_norm": 40.629493713378906, "learning_rate": 1.8094835779913403e-05, - "loss": 1.0178, + "loss": 1.8615, "step": 9020 }, { "epoch": 0.95, - "grad_norm": 12.080143928527832, + "grad_norm": 35.70902633666992, "learning_rate": 1.809272362445876e-05, - "loss": 0.8346, + "loss": 1.8083, "step": 9030 }, { "epoch": 0.95, - "grad_norm": 24.165618896484375, + "grad_norm": 31.998449325561523, "learning_rate": 1.809061146900412e-05, - "loss": 1.2692, + "loss": 2.4729, "step": 9040 }, { "epoch": 0.96, - "grad_norm": 17.731882095336914, + "grad_norm": 35.29195022583008, "learning_rate": 1.8088499313549478e-05, - "loss": 1.3324, + "loss": 2.0817, "step": 9050 }, { "epoch": 0.96, - "grad_norm": 14.011383056640625, + "grad_norm": 34.03513717651367, "learning_rate": 1.808638715809484e-05, - "loss": 1.0925, + "loss": 1.6957, "step": 9060 }, { "epoch": 0.96, - "grad_norm": 18.377748489379883, + "grad_norm": 43.619667053222656, "learning_rate": 1.8084275002640196e-05, - "loss": 1.2353, + "loss": 2.6337, "step": 9070 }, { "epoch": 0.96, - "grad_norm": 12.366424560546875, + "grad_norm": 32.684871673583984, "learning_rate": 1.8082162847185553e-05, - "loss": 1.1217, + "loss": 2.0177, "step": 9080 }, { "epoch": 0.96, - "grad_norm": 7.097836494445801, + "grad_norm": 32.4612922668457, "learning_rate": 1.8080050691730913e-05, - "loss": 0.9792, + "loss": 1.8708, "step": 9090 }, { "epoch": 0.96, - "grad_norm": 9.577353477478027, + "grad_norm": 43.49502182006836, "learning_rate": 1.807793853627627e-05, - "loss": 0.7545, + "loss": 1.5843, "step": 9100 }, { "epoch": 0.96, - "grad_norm": 15.491479873657227, + "grad_norm": 51.47791290283203, "learning_rate": 1.807582638082163e-05, - "loss": 1.0262, + "loss": 1.928, "step": 9110 }, { "epoch": 0.96, - "grad_norm": 17.45625114440918, + "grad_norm": 33.489280700683594, "learning_rate": 1.807371422536699e-05, - "loss": 1.3578, + "loss": 2.1143, "step": 9120 }, { "epoch": 0.96, - "grad_norm": 8.056740760803223, + "grad_norm": 30.118656158447266, "learning_rate": 1.8071602069912345e-05, - "loss": 1.123, + "loss": 1.6712, "step": 9130 }, { "epoch": 0.97, - "grad_norm": 10.112563133239746, + "grad_norm": 43.68329620361328, "learning_rate": 1.8069489914457705e-05, - "loss": 0.8611, + "loss": 2.2669, "step": 9140 }, { "epoch": 0.97, - "grad_norm": 28.17940330505371, + "grad_norm": 36.11957931518555, "learning_rate": 1.8067377759003066e-05, - "loss": 0.8601, + "loss": 1.8187, "step": 9150 }, { "epoch": 0.97, - "grad_norm": 7.508697032928467, + "grad_norm": 23.677852630615234, "learning_rate": 1.8065265603548423e-05, - "loss": 1.1482, + "loss": 1.9145, "step": 9160 }, { "epoch": 0.97, - "grad_norm": 15.073700904846191, + "grad_norm": 30.338218688964844, "learning_rate": 1.8063153448093783e-05, - "loss": 1.2125, + "loss": 1.9329, "step": 9170 }, { "epoch": 0.97, - "grad_norm": 10.337013244628906, + "grad_norm": 44.632171630859375, "learning_rate": 1.806104129263914e-05, - "loss": 0.9872, + "loss": 1.8053, "step": 9180 }, { "epoch": 0.97, - "grad_norm": 6.9291791915893555, + "grad_norm": 44.846290588378906, "learning_rate": 1.8058929137184497e-05, - "loss": 1.0145, + "loss": 1.7697, "step": 9190 }, { "epoch": 0.97, - "grad_norm": 10.183677673339844, + "grad_norm": 36.108428955078125, "learning_rate": 1.8056816981729858e-05, - "loss": 0.8717, + "loss": 1.8388, "step": 9200 }, { "epoch": 0.97, - "grad_norm": 11.702545166015625, + "grad_norm": 34.38186264038086, "learning_rate": 1.8054704826275215e-05, - "loss": 1.2656, + "loss": 2.1046, "step": 9210 }, { "epoch": 0.97, - "grad_norm": 13.695345878601074, + "grad_norm": 45.159088134765625, "learning_rate": 1.8052592670820576e-05, - "loss": 1.0701, + "loss": 1.8907, "step": 9220 }, { "epoch": 0.97, - "grad_norm": 6.221740245819092, + "grad_norm": 34.815208435058594, "learning_rate": 1.8050480515365933e-05, - "loss": 0.8572, + "loss": 1.6192, "step": 9230 }, { "epoch": 0.98, - "grad_norm": 25.919979095458984, + "grad_norm": 49.51512145996094, "learning_rate": 1.804836835991129e-05, - "loss": 1.026, + "loss": 2.0322, "step": 9240 }, { "epoch": 0.98, - "grad_norm": 14.497511863708496, + "grad_norm": 35.61814880371094, "learning_rate": 1.804625620445665e-05, - "loss": 0.9616, + "loss": 2.0159, "step": 9250 }, { "epoch": 0.98, - "grad_norm": 6.142013072967529, + "grad_norm": 30.701446533203125, "learning_rate": 1.8044144049002007e-05, - "loss": 1.0134, + "loss": 1.8119, "step": 9260 }, { "epoch": 0.98, - "grad_norm": 16.829177856445312, + "grad_norm": 26.023046493530273, "learning_rate": 1.8042031893547368e-05, - "loss": 0.9181, + "loss": 1.9671, "step": 9270 }, { "epoch": 0.98, - "grad_norm": 11.093836784362793, + "grad_norm": 37.29668045043945, "learning_rate": 1.8039919738092725e-05, - "loss": 1.3053, + "loss": 2.0637, "step": 9280 }, { "epoch": 0.98, - "grad_norm": 15.339434623718262, + "grad_norm": 33.529415130615234, "learning_rate": 1.8037807582638082e-05, - "loss": 1.1193, + "loss": 1.9621, "step": 9290 }, { "epoch": 0.98, - "grad_norm": 14.055992126464844, + "grad_norm": 37.246185302734375, "learning_rate": 1.8035695427183442e-05, - "loss": 1.169, + "loss": 1.715, "step": 9300 }, { "epoch": 0.98, - "grad_norm": 14.46906852722168, + "grad_norm": 36.40855407714844, "learning_rate": 1.80335832717288e-05, - "loss": 1.246, + "loss": 1.8911, "step": 9310 }, { "epoch": 0.98, - "grad_norm": 20.36118507385254, + "grad_norm": 31.27142906188965, "learning_rate": 1.803147111627416e-05, - "loss": 0.8299, + "loss": 1.7333, "step": 9320 }, { "epoch": 0.99, - "grad_norm": 10.051403045654297, + "grad_norm": 30.15191650390625, "learning_rate": 1.8029358960819517e-05, - "loss": 0.9396, + "loss": 1.8791, "step": 9330 }, { "epoch": 0.99, - "grad_norm": 18.68474578857422, + "grad_norm": 46.734344482421875, "learning_rate": 1.8027246805364874e-05, - "loss": 1.0374, + "loss": 1.7166, "step": 9340 }, { "epoch": 0.99, - "grad_norm": 16.829195022583008, + "grad_norm": 31.71116828918457, "learning_rate": 1.8025134649910235e-05, - "loss": 0.7255, + "loss": 1.7389, "step": 9350 }, { "epoch": 0.99, - "grad_norm": 11.670442581176758, + "grad_norm": 36.21526336669922, "learning_rate": 1.802302249445559e-05, - "loss": 1.1839, + "loss": 1.8965, "step": 9360 }, { "epoch": 0.99, - "grad_norm": 23.307655334472656, + "grad_norm": 27.962390899658203, "learning_rate": 1.8020910339000952e-05, - "loss": 1.1347, + "loss": 1.6961, "step": 9370 }, { "epoch": 0.99, - "grad_norm": 10.857260704040527, + "grad_norm": 30.440322875976562, "learning_rate": 1.8018798183546313e-05, - "loss": 1.0671, + "loss": 1.7487, "step": 9380 }, { "epoch": 0.99, - "grad_norm": 7.48701286315918, + "grad_norm": 25.43843650817871, "learning_rate": 1.801668602809167e-05, - "loss": 1.1456, + "loss": 1.7507, "step": 9390 }, { "epoch": 0.99, - "grad_norm": 11.284430503845215, + "grad_norm": 27.766586303710938, "learning_rate": 1.8014573872637027e-05, - "loss": 1.039, + "loss": 1.8011, "step": 9400 }, { "epoch": 0.99, - "grad_norm": 11.45370101928711, + "grad_norm": 28.31987953186035, "learning_rate": 1.8012461717182387e-05, - "loss": 1.1413, + "loss": 1.8032, "step": 9410 }, { "epoch": 0.99, - "grad_norm": 25.074888229370117, + "grad_norm": 41.11534118652344, "learning_rate": 1.8010349561727744e-05, - "loss": 0.9454, + "loss": 1.9221, "step": 9420 }, { "epoch": 1.0, - "grad_norm": 8.38526439666748, + "grad_norm": 37.566314697265625, "learning_rate": 1.8008237406273105e-05, - "loss": 1.0903, + "loss": 1.9082, "step": 9430 }, { "epoch": 1.0, - "grad_norm": 12.83663558959961, + "grad_norm": 45.9675407409668, "learning_rate": 1.8006125250818462e-05, - "loss": 0.9865, + "loss": 1.8148, "step": 9440 }, { "epoch": 1.0, - "grad_norm": 17.970874786376953, + "grad_norm": 29.065797805786133, "learning_rate": 1.800401309536382e-05, - "loss": 0.913, + "loss": 1.6811, "step": 9450 }, { "epoch": 1.0, - "grad_norm": 6.146689414978027, + "grad_norm": 20.526782989501953, "learning_rate": 1.800190093990918e-05, - "loss": 0.9431, + "loss": 1.7988, "step": 9460 }, { "epoch": 1.0, - "eval_accuracy": 0.8592871287128713, - "eval_loss": 0.6776774525642395, - "eval_runtime": 168.454, - "eval_samples_per_second": 149.893, - "eval_steps_per_second": 18.741, + "eval_accuracy": 0.6554851485148515, + "eval_loss": 1.4391714334487915, + "eval_runtime": 36.9818, + "eval_samples_per_second": 682.769, + "eval_steps_per_second": 85.366, "step": 9469 }, { "epoch": 1.0, - "grad_norm": 5.682332992553711, + "grad_norm": 30.146974563598633, "learning_rate": 1.7999788784454537e-05, - "loss": 1.1704, + "loss": 2.058, "step": 9470 }, { "epoch": 1.0, - "grad_norm": 14.054755210876465, + "grad_norm": 49.60479736328125, "learning_rate": 1.7997676628999897e-05, - "loss": 0.8697, + "loss": 1.9417, "step": 9480 }, { "epoch": 1.0, - "grad_norm": 8.886983871459961, + "grad_norm": 31.216285705566406, "learning_rate": 1.7995564473545254e-05, - "loss": 1.073, + "loss": 1.8913, "step": 9490 }, { "epoch": 1.0, - "grad_norm": 4.791759490966797, + "grad_norm": 31.833906173706055, "learning_rate": 1.7993452318090615e-05, - "loss": 0.728, + "loss": 1.8468, "step": 9500 }, { "epoch": 1.0, - "grad_norm": 12.895437240600586, + "grad_norm": 32.80206298828125, "learning_rate": 1.799134016263597e-05, - "loss": 0.8244, + "loss": 1.6623, "step": 9510 }, { "epoch": 1.01, - "grad_norm": 17.087129592895508, + "grad_norm": 25.366172790527344, "learning_rate": 1.798922800718133e-05, - "loss": 0.8041, + "loss": 1.6633, "step": 9520 }, { "epoch": 1.01, - "grad_norm": 11.150397300720215, + "grad_norm": 35.958290100097656, "learning_rate": 1.798711585172669e-05, - "loss": 1.0592, + "loss": 2.0419, "step": 9530 }, { "epoch": 1.01, - "grad_norm": 3.3145618438720703, + "grad_norm": 29.632949829101562, "learning_rate": 1.7985003696272046e-05, - "loss": 0.5994, + "loss": 1.4094, "step": 9540 }, { "epoch": 1.01, - "grad_norm": 22.388652801513672, + "grad_norm": 35.672359466552734, "learning_rate": 1.7982891540817407e-05, - "loss": 1.0692, + "loss": 2.1361, "step": 9550 }, { "epoch": 1.01, - "grad_norm": 3.206340789794922, + "grad_norm": 23.42523956298828, "learning_rate": 1.7980779385362764e-05, - "loss": 0.7508, + "loss": 1.7132, "step": 9560 }, { "epoch": 1.01, - "grad_norm": 3.9490954875946045, + "grad_norm": 32.817623138427734, "learning_rate": 1.797866722990812e-05, - "loss": 0.8654, + "loss": 1.8799, "step": 9570 }, { "epoch": 1.01, - "grad_norm": 18.132949829101562, + "grad_norm": 34.668521881103516, "learning_rate": 1.797655507445348e-05, - "loss": 0.9038, + "loss": 2.0903, "step": 9580 }, { "epoch": 1.01, - "grad_norm": 9.244780540466309, + "grad_norm": 29.469491958618164, "learning_rate": 1.797444291899884e-05, - "loss": 1.0191, + "loss": 1.6779, "step": 9590 }, { "epoch": 1.01, - "grad_norm": 16.98570442199707, + "grad_norm": 38.22907638549805, "learning_rate": 1.79723307635442e-05, - "loss": 1.1825, + "loss": 2.049, "step": 9600 }, { "epoch": 1.01, - "grad_norm": 12.585511207580566, + "grad_norm": 26.687732696533203, "learning_rate": 1.7970218608089556e-05, - "loss": 1.0487, + "loss": 1.7162, "step": 9610 }, { "epoch": 1.02, - "grad_norm": 5.056178569793701, + "grad_norm": 34.050662994384766, "learning_rate": 1.7968106452634913e-05, - "loss": 0.8545, + "loss": 1.6091, "step": 9620 }, { "epoch": 1.02, - "grad_norm": 4.910436153411865, + "grad_norm": 34.26771926879883, "learning_rate": 1.7965994297180274e-05, - "loss": 1.126, + "loss": 1.6892, "step": 9630 }, { "epoch": 1.02, - "grad_norm": 20.429964065551758, + "grad_norm": 38.551448822021484, "learning_rate": 1.7963882141725634e-05, - "loss": 0.8996, + "loss": 2.0819, "step": 9640 }, { "epoch": 1.02, - "grad_norm": 9.455340385437012, + "grad_norm": 33.288143157958984, "learning_rate": 1.796176998627099e-05, - "loss": 1.0773, + "loss": 1.9878, "step": 9650 }, { "epoch": 1.02, - "grad_norm": 12.836523056030273, + "grad_norm": 25.299135208129883, "learning_rate": 1.795965783081635e-05, - "loss": 0.9379, + "loss": 1.6605, "step": 9660 }, { "epoch": 1.02, - "grad_norm": 12.553736686706543, + "grad_norm": 36.506710052490234, "learning_rate": 1.795754567536171e-05, - "loss": 1.0303, + "loss": 1.6583, "step": 9670 }, { "epoch": 1.02, - "grad_norm": 18.061105728149414, + "grad_norm": 32.633148193359375, "learning_rate": 1.7955433519907066e-05, - "loss": 0.7226, + "loss": 1.9096, "step": 9680 }, { "epoch": 1.02, - "grad_norm": 4.802818298339844, + "grad_norm": 27.228797912597656, "learning_rate": 1.7953321364452426e-05, - "loss": 1.0726, + "loss": 1.8669, "step": 9690 }, { "epoch": 1.02, - "grad_norm": 7.307675361633301, + "grad_norm": 25.534696578979492, "learning_rate": 1.7951209208997783e-05, - "loss": 0.7618, + "loss": 1.6073, "step": 9700 }, { "epoch": 1.03, - "grad_norm": 9.695917129516602, + "grad_norm": 29.90326690673828, "learning_rate": 1.7949097053543144e-05, - "loss": 0.9062, + "loss": 1.9962, "step": 9710 }, { "epoch": 1.03, - "grad_norm": 19.77643585205078, + "grad_norm": 23.749553680419922, "learning_rate": 1.79469848980885e-05, - "loss": 1.0099, + "loss": 1.4861, "step": 9720 }, { "epoch": 1.03, - "grad_norm": 14.523262977600098, + "grad_norm": 31.187156677246094, "learning_rate": 1.7944872742633858e-05, - "loss": 0.9203, + "loss": 1.6974, "step": 9730 }, { "epoch": 1.03, - "grad_norm": 8.502359390258789, + "grad_norm": 31.348939895629883, "learning_rate": 1.794276058717922e-05, - "loss": 0.8813, + "loss": 1.599, "step": 9740 }, { "epoch": 1.03, - "grad_norm": 4.14495849609375, + "grad_norm": 27.27236557006836, "learning_rate": 1.7940648431724576e-05, - "loss": 1.0661, + "loss": 1.9442, "step": 9750 }, { "epoch": 1.03, - "grad_norm": 14.408905982971191, + "grad_norm": 30.385217666625977, "learning_rate": 1.7938536276269936e-05, - "loss": 1.1265, + "loss": 1.8327, "step": 9760 }, { "epoch": 1.03, - "grad_norm": 7.805471420288086, + "grad_norm": 38.395606994628906, "learning_rate": 1.7936424120815293e-05, - "loss": 0.8658, + "loss": 1.8152, "step": 9770 }, { "epoch": 1.03, - "grad_norm": 13.215538024902344, + "grad_norm": 35.60492706298828, "learning_rate": 1.793431196536065e-05, - "loss": 0.9041, + "loss": 1.5097, "step": 9780 }, { "epoch": 1.03, - "grad_norm": 11.225916862487793, + "grad_norm": 34.61677551269531, "learning_rate": 1.793219980990601e-05, - "loss": 0.7228, + "loss": 1.5195, "step": 9790 }, { "epoch": 1.03, - "grad_norm": 7.431766986846924, + "grad_norm": 26.265073776245117, "learning_rate": 1.7930087654451368e-05, - "loss": 0.8274, + "loss": 1.586, "step": 9800 }, { "epoch": 1.04, - "grad_norm": 13.996476173400879, + "grad_norm": 25.490629196166992, "learning_rate": 1.7927975498996728e-05, - "loss": 0.9482, + "loss": 1.6697, "step": 9810 }, { "epoch": 1.04, - "grad_norm": 12.636820793151855, + "grad_norm": 31.024368286132812, "learning_rate": 1.792586334354209e-05, - "loss": 1.0444, + "loss": 1.6998, "step": 9820 }, { "epoch": 1.04, - "grad_norm": 20.50494956970215, + "grad_norm": 43.16365051269531, "learning_rate": 1.7923751188087442e-05, - "loss": 0.7523, + "loss": 1.737, "step": 9830 }, { "epoch": 1.04, - "grad_norm": 13.294538497924805, + "grad_norm": 42.42915725708008, "learning_rate": 1.7921639032632803e-05, - "loss": 1.0105, + "loss": 1.7315, "step": 9840 }, { "epoch": 1.04, - "grad_norm": 6.7240519523620605, + "grad_norm": 41.62624740600586, "learning_rate": 1.7919526877178163e-05, - "loss": 0.9874, + "loss": 1.8475, "step": 9850 }, { "epoch": 1.04, - "grad_norm": 5.973996162414551, + "grad_norm": 31.1720027923584, "learning_rate": 1.791741472172352e-05, - "loss": 0.8841, + "loss": 1.5869, "step": 9860 }, { "epoch": 1.04, - "grad_norm": 15.215359687805176, + "grad_norm": 31.406003952026367, "learning_rate": 1.791530256626888e-05, - "loss": 0.7664, + "loss": 1.8904, "step": 9870 }, { "epoch": 1.04, - "grad_norm": 12.936103820800781, + "grad_norm": 45.40945816040039, "learning_rate": 1.7913190410814238e-05, - "loss": 0.9939, + "loss": 2.1682, "step": 9880 }, { "epoch": 1.04, - "grad_norm": 10.085347175598145, + "grad_norm": 33.55061340332031, "learning_rate": 1.7911078255359595e-05, - "loss": 0.999, + "loss": 1.7616, "step": 9890 }, { "epoch": 1.05, - "grad_norm": 14.168978691101074, + "grad_norm": 28.55853843688965, "learning_rate": 1.7908966099904956e-05, - "loss": 1.0157, + "loss": 1.5314, "step": 9900 }, { "epoch": 1.05, - "grad_norm": 5.853646278381348, + "grad_norm": 28.396583557128906, "learning_rate": 1.7906853944450313e-05, - "loss": 0.9199, + "loss": 2.0015, "step": 9910 }, { "epoch": 1.05, - "grad_norm": 17.583904266357422, + "grad_norm": 19.18535041809082, "learning_rate": 1.7904741788995673e-05, - "loss": 0.8778, + "loss": 1.6423, "step": 9920 }, { "epoch": 1.05, - "grad_norm": 6.312596321105957, + "grad_norm": 40.87828063964844, "learning_rate": 1.790262963354103e-05, - "loss": 0.9612, + "loss": 1.7613, "step": 9930 }, { "epoch": 1.05, - "grad_norm": 16.138517379760742, + "grad_norm": 43.62138366699219, "learning_rate": 1.7900517478086387e-05, - "loss": 1.0345, + "loss": 1.7285, "step": 9940 }, { "epoch": 1.05, - "grad_norm": 6.56013822555542, + "grad_norm": 32.0494384765625, "learning_rate": 1.7898405322631748e-05, - "loss": 0.8343, + "loss": 1.7621, "step": 9950 }, { "epoch": 1.05, - "grad_norm": 8.972660064697266, + "grad_norm": 31.716796875, "learning_rate": 1.7896293167177105e-05, - "loss": 0.8363, + "loss": 1.5745, "step": 9960 }, { "epoch": 1.05, - "grad_norm": 29.742406845092773, + "grad_norm": 42.49932098388672, "learning_rate": 1.7894181011722465e-05, - "loss": 1.2424, + "loss": 2.0269, "step": 9970 }, { "epoch": 1.05, - "grad_norm": 19.099777221679688, + "grad_norm": 37.230289459228516, "learning_rate": 1.7892068856267822e-05, - "loss": 0.707, + "loss": 1.9201, "step": 9980 }, { "epoch": 1.06, - "grad_norm": 11.020627975463867, + "grad_norm": 25.755586624145508, "learning_rate": 1.788995670081318e-05, - "loss": 0.9854, + "loss": 1.7296, "step": 9990 }, { "epoch": 1.06, - "grad_norm": 15.445625305175781, + "grad_norm": 30.547876358032227, "learning_rate": 1.788784454535854e-05, - "loss": 0.9246, + "loss": 1.6697, "step": 10000 }, { "epoch": 1.06, - "grad_norm": 4.149037837982178, + "grad_norm": 57.15998840332031, "learning_rate": 1.7885732389903897e-05, - "loss": 1.1823, + "loss": 2.0389, "step": 10010 }, { "epoch": 1.06, - "grad_norm": 2.9393999576568604, + "grad_norm": 37.91709899902344, "learning_rate": 1.7883620234449257e-05, - "loss": 0.9963, + "loss": 1.9657, "step": 10020 }, { "epoch": 1.06, - "grad_norm": 2.20711088180542, + "grad_norm": 28.677793502807617, "learning_rate": 1.7881508078994615e-05, - "loss": 0.8939, + "loss": 1.6377, "step": 10030 }, { "epoch": 1.06, - "grad_norm": 12.63458251953125, + "grad_norm": 29.014482498168945, "learning_rate": 1.7879395923539975e-05, - "loss": 1.0167, + "loss": 1.9201, "step": 10040 }, { "epoch": 1.06, - "grad_norm": 17.2911319732666, + "grad_norm": 45.60978317260742, "learning_rate": 1.7877283768085332e-05, - "loss": 0.9843, + "loss": 1.9095, "step": 10050 }, { "epoch": 1.06, - "grad_norm": 9.63071060180664, + "grad_norm": 21.821693420410156, "learning_rate": 1.787517161263069e-05, - "loss": 0.5248, + "loss": 1.2154, "step": 10060 }, { "epoch": 1.06, - "grad_norm": 11.95844554901123, + "grad_norm": 30.543811798095703, "learning_rate": 1.787305945717605e-05, - "loss": 1.0256, + "loss": 1.7874, "step": 10070 }, { "epoch": 1.06, - "grad_norm": 6.743292331695557, + "grad_norm": 27.19520378112793, "learning_rate": 1.787094730172141e-05, - "loss": 0.9484, + "loss": 1.5305, "step": 10080 }, { "epoch": 1.07, - "grad_norm": 10.722975730895996, + "grad_norm": 29.18716049194336, "learning_rate": 1.7868835146266767e-05, - "loss": 1.2498, + "loss": 1.7696, "step": 10090 }, { "epoch": 1.07, - "grad_norm": 15.991499900817871, + "grad_norm": 54.71036911010742, "learning_rate": 1.7866722990812124e-05, - "loss": 0.9731, + "loss": 2.0133, "step": 10100 }, { "epoch": 1.07, - "grad_norm": 29.081785202026367, + "grad_norm": 31.460838317871094, "learning_rate": 1.7864610835357485e-05, - "loss": 0.841, + "loss": 1.5665, "step": 10110 }, { "epoch": 1.07, - "grad_norm": 1.7848925590515137, + "grad_norm": 45.16185760498047, "learning_rate": 1.7862498679902842e-05, - "loss": 0.6942, + "loss": 1.8944, "step": 10120 }, { "epoch": 1.07, - "grad_norm": 8.433070182800293, + "grad_norm": 32.68368911743164, "learning_rate": 1.7860386524448202e-05, - "loss": 1.0438, + "loss": 1.8542, "step": 10130 }, { "epoch": 1.07, - "grad_norm": 8.627711296081543, + "grad_norm": 25.0883731842041, "learning_rate": 1.785827436899356e-05, - "loss": 1.0343, + "loss": 1.6198, "step": 10140 }, { "epoch": 1.07, - "grad_norm": 13.461960792541504, + "grad_norm": 43.260738372802734, "learning_rate": 1.785616221353892e-05, - "loss": 1.0615, + "loss": 1.8682, "step": 10150 }, { "epoch": 1.07, - "grad_norm": 10.451886177062988, + "grad_norm": 36.135032653808594, "learning_rate": 1.7854050058084277e-05, - "loss": 1.0369, + "loss": 1.7558, "step": 10160 }, { "epoch": 1.07, - "grad_norm": 20.93505859375, + "grad_norm": 34.412635803222656, "learning_rate": 1.7851937902629634e-05, - "loss": 0.9366, + "loss": 1.9013, "step": 10170 }, { "epoch": 1.08, - "grad_norm": 8.457751274108887, + "grad_norm": 43.429256439208984, "learning_rate": 1.7849825747174995e-05, - "loss": 0.9843, + "loss": 2.0255, "step": 10180 }, { "epoch": 1.08, - "grad_norm": 9.388884544372559, + "grad_norm": 30.680082321166992, "learning_rate": 1.784771359172035e-05, - "loss": 0.7657, + "loss": 1.4919, "step": 10190 }, { "epoch": 1.08, - "grad_norm": 25.447233200073242, + "grad_norm": 45.227481842041016, "learning_rate": 1.7845601436265712e-05, - "loss": 0.8939, + "loss": 2.0893, "step": 10200 }, { "epoch": 1.08, - "grad_norm": 11.415900230407715, + "grad_norm": 31.463470458984375, "learning_rate": 1.784348928081107e-05, - "loss": 0.9414, + "loss": 1.7674, "step": 10210 }, { "epoch": 1.08, - "grad_norm": 13.568324089050293, + "grad_norm": 27.974571228027344, "learning_rate": 1.7841377125356426e-05, - "loss": 0.8034, + "loss": 1.5434, "step": 10220 }, { "epoch": 1.08, - "grad_norm": 17.43779182434082, + "grad_norm": 46.77962875366211, "learning_rate": 1.7839264969901787e-05, - "loss": 0.8746, + "loss": 1.8658, "step": 10230 }, { "epoch": 1.08, - "grad_norm": 13.371694564819336, + "grad_norm": 47.19597244262695, "learning_rate": 1.7837152814447144e-05, - "loss": 0.7022, + "loss": 1.965, "step": 10240 }, { "epoch": 1.08, - "grad_norm": 9.403926849365234, + "grad_norm": 35.97452163696289, "learning_rate": 1.7835040658992504e-05, - "loss": 0.7544, + "loss": 1.5681, "step": 10250 }, { "epoch": 1.08, - "grad_norm": 9.277541160583496, + "grad_norm": 31.867229461669922, "learning_rate": 1.783292850353786e-05, - "loss": 0.9152, + "loss": 1.7798, "step": 10260 }, { "epoch": 1.08, - "grad_norm": 12.650355339050293, + "grad_norm": 41.757938385009766, "learning_rate": 1.783081634808322e-05, - "loss": 0.9885, + "loss": 1.9516, "step": 10270 }, { "epoch": 1.09, - "grad_norm": 8.588929176330566, + "grad_norm": 36.46381759643555, "learning_rate": 1.782870419262858e-05, - "loss": 1.0854, + "loss": 1.9428, "step": 10280 }, { "epoch": 1.09, - "grad_norm": 10.702228546142578, + "grad_norm": 35.33331298828125, "learning_rate": 1.7826592037173936e-05, - "loss": 0.9226, + "loss": 1.7583, "step": 10290 }, { "epoch": 1.09, - "grad_norm": 12.236588478088379, + "grad_norm": 33.61581802368164, "learning_rate": 1.7824479881719297e-05, - "loss": 0.8803, + "loss": 1.6686, "step": 10300 }, { "epoch": 1.09, - "grad_norm": 9.02604866027832, + "grad_norm": 41.38566970825195, "learning_rate": 1.7822367726264657e-05, - "loss": 0.7095, + "loss": 1.4716, "step": 10310 }, { "epoch": 1.09, - "grad_norm": 14.970351219177246, + "grad_norm": 36.53626251220703, "learning_rate": 1.782025557081001e-05, - "loss": 0.7458, + "loss": 1.7811, "step": 10320 }, { "epoch": 1.09, - "grad_norm": 17.478548049926758, + "grad_norm": 31.328472137451172, "learning_rate": 1.781814341535537e-05, - "loss": 0.9433, + "loss": 1.7904, "step": 10330 }, { "epoch": 1.09, - "grad_norm": 7.954148769378662, + "grad_norm": 30.208759307861328, "learning_rate": 1.781603125990073e-05, - "loss": 0.9709, + "loss": 1.7532, "step": 10340 }, { "epoch": 1.09, - "grad_norm": 8.837742805480957, + "grad_norm": 45.4921989440918, "learning_rate": 1.781391910444609e-05, - "loss": 1.0386, + "loss": 1.9773, "step": 10350 }, { "epoch": 1.09, - "grad_norm": 10.600408554077148, + "grad_norm": 37.62623977661133, "learning_rate": 1.781180694899145e-05, - "loss": 1.1048, + "loss": 2.0046, "step": 10360 }, { "epoch": 1.1, - "grad_norm": 13.206040382385254, + "grad_norm": 37.21344757080078, "learning_rate": 1.7809694793536806e-05, - "loss": 0.7389, + "loss": 1.6158, "step": 10370 }, { "epoch": 1.1, - "grad_norm": 11.477002143859863, + "grad_norm": 53.04911422729492, "learning_rate": 1.7807582638082163e-05, - "loss": 0.7157, + "loss": 1.808, "step": 10380 }, { "epoch": 1.1, - "grad_norm": 24.74976348876953, + "grad_norm": 39.303993225097656, "learning_rate": 1.7805470482627524e-05, - "loss": 1.0701, + "loss": 1.9199, "step": 10390 }, { "epoch": 1.1, - "grad_norm": 12.859298706054688, + "grad_norm": 36.39638900756836, "learning_rate": 1.780335832717288e-05, - "loss": 0.6066, + "loss": 1.6408, "step": 10400 }, { "epoch": 1.1, - "grad_norm": 8.668624877929688, + "grad_norm": 42.700294494628906, "learning_rate": 1.780124617171824e-05, - "loss": 0.7301, + "loss": 1.6026, "step": 10410 }, { "epoch": 1.1, - "grad_norm": 8.893885612487793, + "grad_norm": 47.22438049316406, "learning_rate": 1.77991340162636e-05, - "loss": 0.7873, + "loss": 1.7543, "step": 10420 }, { "epoch": 1.1, - "grad_norm": 8.908610343933105, + "grad_norm": 31.283498764038086, "learning_rate": 1.7797021860808956e-05, - "loss": 0.8288, + "loss": 1.6061, "step": 10430 }, { "epoch": 1.1, - "grad_norm": 8.821549415588379, + "grad_norm": 30.191970825195312, "learning_rate": 1.7794909705354316e-05, - "loss": 0.9822, + "loss": 1.5465, "step": 10440 }, { "epoch": 1.1, - "grad_norm": 8.244380950927734, + "grad_norm": 23.717496871948242, "learning_rate": 1.7792797549899673e-05, - "loss": 0.6074, + "loss": 1.4472, "step": 10450 }, { "epoch": 1.1, - "grad_norm": 4.722766399383545, + "grad_norm": 39.91423416137695, "learning_rate": 1.7790685394445034e-05, - "loss": 0.8082, + "loss": 1.5745, "step": 10460 }, { "epoch": 1.11, - "grad_norm": 6.1146345138549805, + "grad_norm": 32.92343521118164, "learning_rate": 1.778857323899039e-05, - "loss": 1.0697, + "loss": 1.8987, "step": 10470 }, { "epoch": 1.11, - "grad_norm": 19.64202880859375, + "grad_norm": 43.417606353759766, "learning_rate": 1.7786461083535748e-05, - "loss": 1.0067, + "loss": 1.8419, "step": 10480 }, { "epoch": 1.11, - "grad_norm": 7.854731559753418, + "grad_norm": 23.969478607177734, "learning_rate": 1.7784348928081108e-05, - "loss": 0.9053, + "loss": 1.9595, "step": 10490 }, { "epoch": 1.11, - "grad_norm": 10.784468650817871, + "grad_norm": 31.43785285949707, "learning_rate": 1.7782236772626465e-05, - "loss": 1.1288, + "loss": 1.7506, "step": 10500 }, { "epoch": 1.11, - "grad_norm": 6.69235372543335, + "grad_norm": 27.109886169433594, "learning_rate": 1.7780124617171826e-05, - "loss": 1.136, + "loss": 1.7587, "step": 10510 }, { "epoch": 1.11, - "grad_norm": 14.174864768981934, + "grad_norm": 27.43458366394043, "learning_rate": 1.7778012461717183e-05, - "loss": 0.9671, + "loss": 1.849, "step": 10520 }, { "epoch": 1.11, - "grad_norm": 9.40821647644043, + "grad_norm": 41.62321853637695, "learning_rate": 1.7775900306262543e-05, - "loss": 0.6154, + "loss": 1.8667, "step": 10530 }, { "epoch": 1.11, - "grad_norm": 6.702163219451904, + "grad_norm": 27.792478561401367, "learning_rate": 1.77737881508079e-05, - "loss": 0.608, + "loss": 1.9996, "step": 10540 }, { "epoch": 1.11, - "grad_norm": 16.74351692199707, + "grad_norm": 36.123504638671875, "learning_rate": 1.7771675995353258e-05, - "loss": 0.7748, + "loss": 1.885, "step": 10550 }, { "epoch": 1.12, - "grad_norm": 10.859332084655762, + "grad_norm": 32.79438781738281, "learning_rate": 1.7769563839898618e-05, - "loss": 0.7845, + "loss": 1.6805, "step": 10560 }, { "epoch": 1.12, - "grad_norm": 13.512421607971191, + "grad_norm": 37.84840774536133, "learning_rate": 1.776745168444398e-05, - "loss": 0.8861, + "loss": 1.6819, "step": 10570 }, { "epoch": 1.12, - "grad_norm": 6.068662643432617, + "grad_norm": 35.951324462890625, "learning_rate": 1.7765339528989336e-05, - "loss": 0.8527, + "loss": 1.778, "step": 10580 }, { "epoch": 1.12, - "grad_norm": 18.7093563079834, + "grad_norm": 37.522613525390625, "learning_rate": 1.7763227373534693e-05, - "loss": 0.905, + "loss": 1.8544, "step": 10590 }, { "epoch": 1.12, - "grad_norm": 17.2720947265625, + "grad_norm": 31.923757553100586, "learning_rate": 1.7761115218080053e-05, - "loss": 0.9572, + "loss": 1.746, "step": 10600 }, { "epoch": 1.12, - "grad_norm": 12.49015998840332, + "grad_norm": 37.68617248535156, "learning_rate": 1.775900306262541e-05, - "loss": 0.8038, + "loss": 1.7269, "step": 10610 }, { "epoch": 1.12, - "grad_norm": 12.159725189208984, + "grad_norm": 21.199684143066406, "learning_rate": 1.775689090717077e-05, - "loss": 0.9668, + "loss": 1.8475, "step": 10620 }, { "epoch": 1.12, - "grad_norm": 15.144716262817383, + "grad_norm": 40.901458740234375, "learning_rate": 1.7754778751716128e-05, - "loss": 1.0418, + "loss": 1.7838, "step": 10630 }, { "epoch": 1.12, - "grad_norm": 11.145402908325195, + "grad_norm": 32.101226806640625, "learning_rate": 1.7752666596261485e-05, - "loss": 0.9187, + "loss": 1.5459, "step": 10640 }, { "epoch": 1.12, - "grad_norm": 7.789994716644287, + "grad_norm": 25.949634552001953, "learning_rate": 1.7750554440806845e-05, - "loss": 1.0115, + "loss": 1.5288, "step": 10650 }, { "epoch": 1.13, - "grad_norm": 13.899667739868164, + "grad_norm": 40.41289520263672, "learning_rate": 1.7748442285352202e-05, - "loss": 0.8757, + "loss": 1.7555, "step": 10660 }, { "epoch": 1.13, - "grad_norm": 11.284130096435547, + "grad_norm": 32.699241638183594, "learning_rate": 1.7746330129897563e-05, - "loss": 0.9586, + "loss": 1.763, "step": 10670 }, { "epoch": 1.13, - "grad_norm": 24.08342170715332, + "grad_norm": 44.90194320678711, "learning_rate": 1.774421797444292e-05, - "loss": 0.8967, + "loss": 1.5466, "step": 10680 }, { "epoch": 1.13, - "grad_norm": 5.209827899932861, + "grad_norm": 36.05495834350586, "learning_rate": 1.774210581898828e-05, - "loss": 0.9456, + "loss": 1.6583, "step": 10690 }, { "epoch": 1.13, - "grad_norm": 12.08736801147461, + "grad_norm": 34.9971809387207, "learning_rate": 1.7739993663533638e-05, - "loss": 0.9047, + "loss": 1.9133, "step": 10700 }, { "epoch": 1.13, - "grad_norm": 15.16235637664795, + "grad_norm": 39.19689178466797, "learning_rate": 1.7737881508078995e-05, - "loss": 0.9913, + "loss": 2.1696, "step": 10710 }, { "epoch": 1.13, - "grad_norm": 9.014914512634277, + "grad_norm": 50.41804504394531, "learning_rate": 1.7735769352624355e-05, - "loss": 0.8939, + "loss": 1.8216, "step": 10720 }, { "epoch": 1.13, - "grad_norm": 15.587324142456055, + "grad_norm": 41.048946380615234, "learning_rate": 1.7733657197169712e-05, - "loss": 0.6373, + "loss": 1.6549, "step": 10730 }, { "epoch": 1.13, - "grad_norm": 13.902644157409668, + "grad_norm": 36.040313720703125, "learning_rate": 1.7731545041715073e-05, - "loss": 0.7632, + "loss": 1.9512, "step": 10740 }, { "epoch": 1.14, - "grad_norm": 7.159825325012207, + "grad_norm": 27.6641902923584, "learning_rate": 1.772943288626043e-05, - "loss": 0.8699, + "loss": 1.5869, "step": 10750 }, { "epoch": 1.14, - "grad_norm": 13.643608093261719, + "grad_norm": 34.13793182373047, "learning_rate": 1.7727320730805787e-05, - "loss": 0.9409, + "loss": 1.7443, "step": 10760 }, { "epoch": 1.14, - "grad_norm": 9.734718322753906, + "grad_norm": 38.234439849853516, "learning_rate": 1.7725208575351147e-05, - "loss": 0.836, + "loss": 1.6491, "step": 10770 }, { "epoch": 1.14, - "grad_norm": 10.799027442932129, + "grad_norm": 23.673141479492188, "learning_rate": 1.7723096419896508e-05, - "loss": 0.7552, + "loss": 1.4445, "step": 10780 }, { "epoch": 1.14, - "grad_norm": 10.25438404083252, + "grad_norm": 37.02865219116211, "learning_rate": 1.7720984264441865e-05, - "loss": 0.9653, + "loss": 1.5295, "step": 10790 }, { "epoch": 1.14, - "grad_norm": 6.112825870513916, + "grad_norm": 17.086790084838867, "learning_rate": 1.7718872108987225e-05, - "loss": 0.923, + "loss": 1.5292, "step": 10800 }, { "epoch": 1.14, - "grad_norm": 1.7167216539382935, + "grad_norm": 23.415681838989258, "learning_rate": 1.7716759953532582e-05, - "loss": 0.675, + "loss": 1.5857, "step": 10810 }, { "epoch": 1.14, - "grad_norm": 5.603817462921143, + "grad_norm": 29.53828239440918, "learning_rate": 1.771464779807794e-05, - "loss": 0.8632, + "loss": 1.7156, "step": 10820 }, { "epoch": 1.14, - "grad_norm": 5.373841762542725, + "grad_norm": 41.36430358886719, "learning_rate": 1.77125356426233e-05, - "loss": 0.6002, + "loss": 1.4936, "step": 10830 }, { "epoch": 1.14, - "grad_norm": 21.366941452026367, + "grad_norm": 45.432647705078125, "learning_rate": 1.7710423487168657e-05, - "loss": 0.8351, + "loss": 1.619, "step": 10840 }, { "epoch": 1.15, - "grad_norm": 16.85696792602539, + "grad_norm": 32.99296569824219, "learning_rate": 1.7708311331714017e-05, - "loss": 0.8068, + "loss": 1.7581, "step": 10850 }, { "epoch": 1.15, - "grad_norm": 5.205652713775635, + "grad_norm": 36.10387420654297, "learning_rate": 1.7706199176259375e-05, - "loss": 0.9252, + "loss": 1.6536, "step": 10860 }, { "epoch": 1.15, - "grad_norm": 12.99118709564209, + "grad_norm": 33.47957992553711, "learning_rate": 1.770408702080473e-05, - "loss": 0.5203, + "loss": 1.2261, "step": 10870 }, { "epoch": 1.15, - "grad_norm": 15.069470405578613, + "grad_norm": 42.7122802734375, "learning_rate": 1.7701974865350092e-05, - "loss": 1.1772, + "loss": 1.9472, "step": 10880 }, { "epoch": 1.15, - "grad_norm": 20.668655395507812, + "grad_norm": 39.370174407958984, "learning_rate": 1.769986270989545e-05, - "loss": 1.089, + "loss": 1.5241, "step": 10890 }, { "epoch": 1.15, - "grad_norm": 19.382539749145508, + "grad_norm": 29.768047332763672, "learning_rate": 1.769775055444081e-05, - "loss": 0.7605, + "loss": 1.6559, "step": 10900 }, { "epoch": 1.15, - "grad_norm": 12.973899841308594, + "grad_norm": 57.98286437988281, "learning_rate": 1.7695638398986167e-05, - "loss": 1.217, + "loss": 1.8432, "step": 10910 }, { "epoch": 1.15, - "grad_norm": 13.610129356384277, + "grad_norm": 41.813720703125, "learning_rate": 1.7693526243531524e-05, - "loss": 1.0363, + "loss": 1.8166, "step": 10920 }, { "epoch": 1.15, - "grad_norm": 20.947099685668945, + "grad_norm": 19.824453353881836, "learning_rate": 1.7691414088076884e-05, - "loss": 0.9654, + "loss": 1.7133, "step": 10930 }, { "epoch": 1.16, - "grad_norm": 3.9849729537963867, + "grad_norm": 37.66018295288086, "learning_rate": 1.768930193262224e-05, - "loss": 1.0141, + "loss": 1.5983, "step": 10940 }, { "epoch": 1.16, - "grad_norm": 14.37062931060791, + "grad_norm": 36.84150314331055, "learning_rate": 1.7687189777167602e-05, - "loss": 0.4881, + "loss": 1.4229, "step": 10950 }, { "epoch": 1.16, - "grad_norm": 6.603664398193359, + "grad_norm": 34.74435043334961, "learning_rate": 1.768507762171296e-05, - "loss": 0.9164, + "loss": 1.765, "step": 10960 }, { "epoch": 1.16, - "grad_norm": 0.9198147058486938, + "grad_norm": 34.53093719482422, "learning_rate": 1.7682965466258316e-05, - "loss": 0.8724, + "loss": 1.9074, "step": 10970 }, { "epoch": 1.16, - "grad_norm": 17.742773056030273, + "grad_norm": 37.57868194580078, "learning_rate": 1.7680853310803677e-05, - "loss": 0.7572, + "loss": 1.6379, "step": 10980 }, { "epoch": 1.16, - "grad_norm": 12.852213859558105, + "grad_norm": 38.25456619262695, "learning_rate": 1.7678741155349034e-05, - "loss": 0.773, + "loss": 1.7274, "step": 10990 }, { "epoch": 1.16, - "grad_norm": 3.5247788429260254, + "grad_norm": 21.86815071105957, "learning_rate": 1.7676628999894394e-05, - "loss": 0.9234, + "loss": 1.9043, "step": 11000 }, { "epoch": 1.16, - "grad_norm": 5.458250522613525, + "grad_norm": 33.471839904785156, "learning_rate": 1.7674516844439755e-05, - "loss": 0.8242, + "loss": 1.6518, "step": 11010 }, { "epoch": 1.16, - "grad_norm": 3.254316568374634, + "grad_norm": 31.153526306152344, "learning_rate": 1.7672404688985108e-05, - "loss": 0.6666, + "loss": 1.5989, "step": 11020 }, { "epoch": 1.16, - "grad_norm": 11.703866004943848, + "grad_norm": 29.736919403076172, "learning_rate": 1.767029253353047e-05, - "loss": 0.5518, + "loss": 1.4515, "step": 11030 }, { "epoch": 1.17, - "grad_norm": 5.252967834472656, + "grad_norm": 33.91030502319336, "learning_rate": 1.766818037807583e-05, - "loss": 0.8239, + "loss": 1.6627, "step": 11040 }, { "epoch": 1.17, - "grad_norm": 14.303006172180176, + "grad_norm": 41.716060638427734, "learning_rate": 1.7666068222621186e-05, - "loss": 0.8912, + "loss": 1.8434, "step": 11050 }, { "epoch": 1.17, - "grad_norm": 14.982497215270996, + "grad_norm": 34.7614860534668, "learning_rate": 1.7663956067166547e-05, - "loss": 0.8097, + "loss": 1.8101, "step": 11060 }, { "epoch": 1.17, - "grad_norm": 23.23073959350586, + "grad_norm": 42.8716926574707, "learning_rate": 1.7661843911711904e-05, - "loss": 0.7857, + "loss": 1.6118, "step": 11070 }, { "epoch": 1.17, - "grad_norm": 4.882040977478027, + "grad_norm": 32.97715759277344, "learning_rate": 1.765973175625726e-05, - "loss": 0.7281, + "loss": 1.6487, "step": 11080 }, { "epoch": 1.17, - "grad_norm": 2.998955488204956, + "grad_norm": 28.740577697753906, "learning_rate": 1.765761960080262e-05, - "loss": 0.8461, + "loss": 1.7399, "step": 11090 }, { "epoch": 1.17, - "grad_norm": 13.35936450958252, + "grad_norm": 22.04177474975586, "learning_rate": 1.765550744534798e-05, - "loss": 0.9272, + "loss": 1.6268, "step": 11100 }, { "epoch": 1.17, - "grad_norm": 4.895836353302002, + "grad_norm": 25.75518798828125, "learning_rate": 1.765339528989334e-05, - "loss": 0.772, + "loss": 1.7149, "step": 11110 }, { "epoch": 1.17, - "grad_norm": 15.495667457580566, + "grad_norm": 41.659339904785156, "learning_rate": 1.7651283134438696e-05, - "loss": 0.9539, + "loss": 1.8025, "step": 11120 }, { "epoch": 1.18, - "grad_norm": 1.293872356414795, + "grad_norm": 19.680273056030273, "learning_rate": 1.7649170978984053e-05, - "loss": 0.722, + "loss": 1.6756, "step": 11130 }, { "epoch": 1.18, - "grad_norm": 17.3375244140625, + "grad_norm": 37.6896858215332, "learning_rate": 1.7647058823529414e-05, - "loss": 0.8615, + "loss": 1.4924, "step": 11140 }, { "epoch": 1.18, - "grad_norm": 5.435382843017578, + "grad_norm": 22.153963088989258, "learning_rate": 1.764494666807477e-05, - "loss": 0.8682, + "loss": 1.6537, "step": 11150 }, { "epoch": 1.18, - "grad_norm": 0.7376787066459656, + "grad_norm": 28.617815017700195, "learning_rate": 1.764283451262013e-05, - "loss": 0.8412, + "loss": 1.5045, "step": 11160 }, { "epoch": 1.18, - "grad_norm": 22.540576934814453, + "grad_norm": 34.4552001953125, "learning_rate": 1.7640722357165488e-05, - "loss": 1.0135, + "loss": 1.6575, "step": 11170 }, { "epoch": 1.18, - "grad_norm": 10.578112602233887, + "grad_norm": 37.749656677246094, "learning_rate": 1.7638610201710845e-05, - "loss": 1.0749, + "loss": 1.7709, "step": 11180 }, { "epoch": 1.18, - "grad_norm": 5.305562973022461, + "grad_norm": 25.964319229125977, "learning_rate": 1.7636498046256206e-05, - "loss": 0.7722, + "loss": 1.6173, "step": 11190 }, { "epoch": 1.18, - "grad_norm": 10.375535011291504, + "grad_norm": 29.611181259155273, "learning_rate": 1.7634385890801563e-05, - "loss": 0.9643, + "loss": 1.7603, "step": 11200 }, { "epoch": 1.18, - "grad_norm": 10.987548828125, + "grad_norm": 36.273109436035156, "learning_rate": 1.7632273735346923e-05, - "loss": 0.9481, + "loss": 1.6091, "step": 11210 }, { "epoch": 1.18, - "grad_norm": 20.597742080688477, + "grad_norm": 44.5077018737793, "learning_rate": 1.763016157989228e-05, - "loss": 0.8869, + "loss": 1.7697, "step": 11220 }, { "epoch": 1.19, - "grad_norm": 23.351516723632812, + "grad_norm": 38.619972229003906, "learning_rate": 1.762804942443764e-05, - "loss": 0.9954, + "loss": 1.6177, "step": 11230 }, { "epoch": 1.19, - "grad_norm": 9.899264335632324, + "grad_norm": 31.20242691040039, "learning_rate": 1.7625937268982998e-05, - "loss": 0.8367, + "loss": 1.8744, "step": 11240 }, { "epoch": 1.19, - "grad_norm": 17.5123291015625, + "grad_norm": 39.92828369140625, "learning_rate": 1.7623825113528355e-05, - "loss": 0.8219, + "loss": 1.8424, "step": 11250 }, { "epoch": 1.19, - "grad_norm": 3.943311929702759, + "grad_norm": 32.34309005737305, "learning_rate": 1.7621712958073716e-05, - "loss": 0.7758, + "loss": 1.695, "step": 11260 }, { "epoch": 1.19, - "grad_norm": 17.091705322265625, + "grad_norm": 36.83544158935547, "learning_rate": 1.7619600802619076e-05, - "loss": 0.9817, + "loss": 1.8689, "step": 11270 }, { "epoch": 1.19, - "grad_norm": 8.571087837219238, + "grad_norm": 28.63447380065918, "learning_rate": 1.7617488647164433e-05, - "loss": 1.1118, + "loss": 1.6805, "step": 11280 }, { "epoch": 1.19, - "grad_norm": 18.43996810913086, + "grad_norm": 34.06890869140625, "learning_rate": 1.761537649170979e-05, - "loss": 0.918, + "loss": 1.5279, "step": 11290 }, { "epoch": 1.19, - "grad_norm": 38.510318756103516, + "grad_norm": 45.67680358886719, "learning_rate": 1.761326433625515e-05, - "loss": 0.6334, + "loss": 1.6127, "step": 11300 }, { "epoch": 1.19, - "grad_norm": 16.8558292388916, + "grad_norm": 38.916072845458984, "learning_rate": 1.7611152180800508e-05, - "loss": 0.6079, + "loss": 1.7657, "step": 11310 }, { "epoch": 1.2, - "grad_norm": 10.93306827545166, + "grad_norm": 33.96638488769531, "learning_rate": 1.7609040025345868e-05, - "loss": 0.5542, + "loss": 1.7597, "step": 11320 }, { "epoch": 1.2, - "grad_norm": 8.443207740783691, + "grad_norm": 31.380767822265625, "learning_rate": 1.7606927869891225e-05, - "loss": 0.7934, + "loss": 1.4411, "step": 11330 }, { "epoch": 1.2, - "grad_norm": 5.494772434234619, + "grad_norm": 36.19474792480469, "learning_rate": 1.7604815714436586e-05, - "loss": 0.8217, + "loss": 1.5901, "step": 11340 }, { "epoch": 1.2, - "grad_norm": 24.681344985961914, + "grad_norm": 48.12162780761719, "learning_rate": 1.7602703558981943e-05, - "loss": 1.1438, + "loss": 1.9075, "step": 11350 }, { "epoch": 1.2, - "grad_norm": 11.995173454284668, + "grad_norm": 46.16938781738281, "learning_rate": 1.76005914035273e-05, - "loss": 0.7788, + "loss": 1.4791, "step": 11360 }, { "epoch": 1.2, - "grad_norm": 42.544166564941406, + "grad_norm": 41.90422439575195, "learning_rate": 1.759847924807266e-05, - "loss": 1.0743, + "loss": 1.7463, "step": 11370 }, { "epoch": 1.2, - "grad_norm": 6.826353073120117, + "grad_norm": 34.995906829833984, "learning_rate": 1.7596367092618018e-05, - "loss": 0.7143, + "loss": 1.4727, "step": 11380 }, { "epoch": 1.2, - "grad_norm": 5.307865619659424, + "grad_norm": 45.37517166137695, "learning_rate": 1.7594254937163378e-05, - "loss": 0.5352, + "loss": 1.753, "step": 11390 }, { "epoch": 1.2, - "grad_norm": 6.802292346954346, + "grad_norm": 29.309349060058594, "learning_rate": 1.7592142781708735e-05, - "loss": 0.8878, + "loss": 1.4833, "step": 11400 }, { "epoch": 1.2, - "grad_norm": 12.511460304260254, + "grad_norm": 42.33164596557617, "learning_rate": 1.7590030626254092e-05, - "loss": 1.1533, + "loss": 1.9635, "step": 11410 }, { "epoch": 1.21, - "grad_norm": 9.346562385559082, + "grad_norm": 44.442604064941406, "learning_rate": 1.7587918470799453e-05, - "loss": 0.7544, + "loss": 1.5646, "step": 11420 }, { "epoch": 1.21, - "grad_norm": 9.49743366241455, + "grad_norm": 28.842195510864258, "learning_rate": 1.758580631534481e-05, - "loss": 0.8193, + "loss": 1.722, "step": 11430 }, { "epoch": 1.21, - "grad_norm": 17.571096420288086, + "grad_norm": 54.80060577392578, "learning_rate": 1.758369415989017e-05, - "loss": 0.7264, + "loss": 1.7661, "step": 11440 }, { "epoch": 1.21, - "grad_norm": 10.239763259887695, + "grad_norm": 27.354415893554688, "learning_rate": 1.758158200443553e-05, - "loss": 0.8429, + "loss": 1.5483, "step": 11450 }, { "epoch": 1.21, - "grad_norm": 11.029680252075195, + "grad_norm": 38.30986785888672, "learning_rate": 1.7579469848980884e-05, - "loss": 0.5738, + "loss": 2.2936, "step": 11460 }, { "epoch": 1.21, - "grad_norm": 13.292211532592773, + "grad_norm": 20.297019958496094, "learning_rate": 1.7577357693526245e-05, - "loss": 1.0675, + "loss": 1.7686, "step": 11470 }, { "epoch": 1.21, - "grad_norm": 8.491296768188477, + "grad_norm": 32.9290657043457, "learning_rate": 1.7575245538071602e-05, - "loss": 0.6232, + "loss": 1.393, "step": 11480 }, { "epoch": 1.21, - "grad_norm": 17.78400230407715, + "grad_norm": 27.660398483276367, "learning_rate": 1.7573133382616962e-05, - "loss": 0.7427, + "loss": 1.5285, "step": 11490 }, { "epoch": 1.21, - "grad_norm": 6.977345943450928, + "grad_norm": 26.75872230529785, "learning_rate": 1.7571021227162323e-05, - "loss": 0.4363, + "loss": 1.3759, "step": 11500 }, { "epoch": 1.22, - "grad_norm": 9.865926742553711, + "grad_norm": 34.844696044921875, "learning_rate": 1.7568909071707677e-05, - "loss": 1.0938, + "loss": 1.6878, "step": 11510 }, { "epoch": 1.22, - "grad_norm": 21.9163761138916, + "grad_norm": 34.05473327636719, "learning_rate": 1.7566796916253037e-05, - "loss": 0.7159, + "loss": 1.6238, "step": 11520 }, { "epoch": 1.22, - "grad_norm": 12.739400863647461, + "grad_norm": 41.73101806640625, "learning_rate": 1.7564684760798398e-05, - "loss": 0.8648, + "loss": 1.7031, "step": 11530 }, { "epoch": 1.22, - "grad_norm": 14.878589630126953, + "grad_norm": 27.405136108398438, "learning_rate": 1.7562572605343755e-05, - "loss": 0.8409, + "loss": 1.6502, "step": 11540 }, { "epoch": 1.22, - "grad_norm": 8.416899681091309, + "grad_norm": 36.15884017944336, "learning_rate": 1.7560460449889115e-05, - "loss": 0.9253, + "loss": 1.7716, "step": 11550 }, { "epoch": 1.22, - "grad_norm": 6.7831711769104, + "grad_norm": 43.792118072509766, "learning_rate": 1.7558348294434472e-05, - "loss": 0.941, + "loss": 1.8557, "step": 11560 }, { "epoch": 1.22, - "grad_norm": 29.11248016357422, + "grad_norm": 41.340110778808594, "learning_rate": 1.755623613897983e-05, - "loss": 0.6829, + "loss": 1.6776, "step": 11570 }, { "epoch": 1.22, - "grad_norm": 5.324409484863281, + "grad_norm": 31.217458724975586, "learning_rate": 1.755412398352519e-05, - "loss": 0.8802, + "loss": 1.5474, "step": 11580 }, { "epoch": 1.22, - "grad_norm": 14.105462074279785, + "grad_norm": 28.267501831054688, "learning_rate": 1.7552011828070547e-05, - "loss": 0.8283, + "loss": 1.4517, "step": 11590 }, { "epoch": 1.23, - "grad_norm": 6.492913246154785, + "grad_norm": 29.466691970825195, "learning_rate": 1.7549899672615907e-05, - "loss": 0.9558, + "loss": 1.5825, "step": 11600 }, { "epoch": 1.23, - "grad_norm": 4.863266468048096, + "grad_norm": 34.268489837646484, "learning_rate": 1.7547787517161264e-05, - "loss": 0.757, + "loss": 1.587, "step": 11610 }, { "epoch": 1.23, - "grad_norm": 13.81037425994873, + "grad_norm": 48.34259033203125, "learning_rate": 1.754567536170662e-05, - "loss": 0.9314, + "loss": 1.878, "step": 11620 }, { "epoch": 1.23, - "grad_norm": 8.264856338500977, + "grad_norm": 32.55610656738281, "learning_rate": 1.7543563206251982e-05, - "loss": 0.7331, + "loss": 1.5921, "step": 11630 }, { "epoch": 1.23, - "grad_norm": 4.394309997558594, + "grad_norm": 35.185062408447266, "learning_rate": 1.754145105079734e-05, - "loss": 0.6675, + "loss": 1.5026, "step": 11640 }, { "epoch": 1.23, - "grad_norm": 21.8615779876709, + "grad_norm": 39.97303009033203, "learning_rate": 1.75393388953427e-05, - "loss": 0.7748, + "loss": 1.5264, "step": 11650 }, { "epoch": 1.23, - "grad_norm": 17.12014389038086, + "grad_norm": 50.12064743041992, "learning_rate": 1.7537226739888057e-05, - "loss": 0.8426, + "loss": 1.6469, "step": 11660 }, { "epoch": 1.23, - "grad_norm": 14.213061332702637, + "grad_norm": 39.25192642211914, "learning_rate": 1.7535114584433414e-05, - "loss": 0.721, + "loss": 1.5792, "step": 11670 }, { "epoch": 1.23, - "grad_norm": 13.928631782531738, + "grad_norm": 34.931312561035156, "learning_rate": 1.7533002428978774e-05, - "loss": 0.7495, + "loss": 1.7237, "step": 11680 }, { "epoch": 1.23, - "grad_norm": 8.583803176879883, + "grad_norm": 31.49875259399414, "learning_rate": 1.753089027352413e-05, - "loss": 0.6406, + "loss": 1.723, "step": 11690 }, { "epoch": 1.24, - "grad_norm": 19.975994110107422, + "grad_norm": 41.78831481933594, "learning_rate": 1.752877811806949e-05, - "loss": 0.792, + "loss": 1.5526, "step": 11700 }, { "epoch": 1.24, - "grad_norm": 3.834601879119873, + "grad_norm": 33.17306137084961, "learning_rate": 1.7526665962614852e-05, - "loss": 0.9384, + "loss": 1.9536, "step": 11710 }, { "epoch": 1.24, - "grad_norm": 51.4061279296875, + "grad_norm": 34.5709342956543, "learning_rate": 1.752455380716021e-05, - "loss": 0.8102, + "loss": 1.5238, "step": 11720 }, { "epoch": 1.24, - "grad_norm": 10.340113639831543, + "grad_norm": 49.21088790893555, "learning_rate": 1.7522441651705566e-05, - "loss": 0.6354, + "loss": 1.7095, "step": 11730 }, { "epoch": 1.24, - "grad_norm": 10.292010307312012, + "grad_norm": 29.864788055419922, "learning_rate": 1.7520329496250927e-05, - "loss": 0.895, + "loss": 1.4331, "step": 11740 }, { "epoch": 1.24, - "grad_norm": 22.84726905822754, + "grad_norm": 28.675729751586914, "learning_rate": 1.7518217340796284e-05, - "loss": 0.7201, + "loss": 1.4502, "step": 11750 }, { "epoch": 1.24, - "grad_norm": 9.927372932434082, + "grad_norm": 23.73443603515625, "learning_rate": 1.7516105185341644e-05, - "loss": 0.7996, + "loss": 1.6514, "step": 11760 }, { "epoch": 1.24, - "grad_norm": 5.982219219207764, + "grad_norm": 31.176603317260742, "learning_rate": 1.7513993029887e-05, - "loss": 0.8888, + "loss": 1.7449, "step": 11770 }, { "epoch": 1.24, - "grad_norm": 11.459419250488281, + "grad_norm": 34.99054718017578, "learning_rate": 1.751188087443236e-05, - "loss": 0.6495, + "loss": 1.4892, "step": 11780 }, { "epoch": 1.25, - "grad_norm": 9.186482429504395, + "grad_norm": 33.93996047973633, "learning_rate": 1.750976871897772e-05, - "loss": 0.7388, + "loss": 1.5737, "step": 11790 }, { "epoch": 1.25, - "grad_norm": 11.992612838745117, + "grad_norm": 41.1075439453125, "learning_rate": 1.7507656563523076e-05, - "loss": 0.7695, + "loss": 1.7977, "step": 11800 }, { "epoch": 1.25, - "grad_norm": 2.583444833755493, + "grad_norm": 30.792566299438477, "learning_rate": 1.7505544408068437e-05, - "loss": 0.9237, + "loss": 1.6589, "step": 11810 }, { "epoch": 1.25, - "grad_norm": 14.494974136352539, + "grad_norm": 43.00666046142578, "learning_rate": 1.7503432252613794e-05, - "loss": 0.9724, + "loss": 1.8413, "step": 11820 }, { "epoch": 1.25, - "grad_norm": 23.72702407836914, + "grad_norm": 47.22118377685547, "learning_rate": 1.750132009715915e-05, - "loss": 1.2997, + "loss": 2.0868, "step": 11830 }, { "epoch": 1.25, - "grad_norm": 2.323913812637329, + "grad_norm": 25.463665008544922, "learning_rate": 1.749920794170451e-05, - "loss": 0.7014, + "loss": 1.5103, "step": 11840 }, { "epoch": 1.25, - "grad_norm": 13.711705207824707, + "grad_norm": 36.746700286865234, "learning_rate": 1.7497095786249868e-05, - "loss": 1.0042, + "loss": 1.6188, "step": 11850 }, { "epoch": 1.25, - "grad_norm": 8.930777549743652, + "grad_norm": 23.79945182800293, "learning_rate": 1.749498363079523e-05, - "loss": 0.8083, + "loss": 1.6336, "step": 11860 }, { "epoch": 1.25, - "grad_norm": 8.424141883850098, + "grad_norm": 47.923248291015625, "learning_rate": 1.7492871475340586e-05, - "loss": 0.7386, + "loss": 1.9854, "step": 11870 }, { "epoch": 1.25, - "grad_norm": 6.8680243492126465, + "grad_norm": 31.241790771484375, "learning_rate": 1.7490759319885946e-05, - "loss": 0.6811, + "loss": 1.663, "step": 11880 }, { "epoch": 1.26, - "grad_norm": 14.261960983276367, + "grad_norm": 27.43385124206543, "learning_rate": 1.7488647164431303e-05, - "loss": 0.8874, + "loss": 1.5365, "step": 11890 }, { "epoch": 1.26, - "grad_norm": 1.4549371004104614, + "grad_norm": 36.577125549316406, "learning_rate": 1.748653500897666e-05, - "loss": 0.7464, + "loss": 1.9087, "step": 11900 }, { "epoch": 1.26, - "grad_norm": 12.98028564453125, + "grad_norm": 44.698875427246094, "learning_rate": 1.748442285352202e-05, - "loss": 0.6712, + "loss": 1.7115, "step": 11910 }, { "epoch": 1.26, - "grad_norm": 9.999544143676758, + "grad_norm": 40.0858039855957, "learning_rate": 1.7482310698067378e-05, - "loss": 0.6159, + "loss": 1.5653, "step": 11920 }, { "epoch": 1.26, - "grad_norm": 9.130315780639648, + "grad_norm": 22.04449462890625, "learning_rate": 1.748019854261274e-05, - "loss": 0.8681, + "loss": 1.6391, "step": 11930 }, { "epoch": 1.26, - "grad_norm": 8.461299896240234, + "grad_norm": 32.8699951171875, "learning_rate": 1.7478086387158096e-05, - "loss": 0.7282, + "loss": 1.7523, "step": 11940 }, { "epoch": 1.26, - "grad_norm": 14.890623092651367, + "grad_norm": 27.648807525634766, "learning_rate": 1.7475974231703453e-05, - "loss": 1.1088, + "loss": 1.6199, "step": 11950 }, { "epoch": 1.26, - "grad_norm": 29.867063522338867, + "grad_norm": 29.761878967285156, "learning_rate": 1.7473862076248813e-05, - "loss": 1.1575, + "loss": 1.7403, "step": 11960 }, { "epoch": 1.26, - "grad_norm": 17.579011917114258, + "grad_norm": 26.89078140258789, "learning_rate": 1.7471749920794174e-05, - "loss": 0.9922, + "loss": 1.9038, "step": 11970 }, { "epoch": 1.27, - "grad_norm": 14.827309608459473, + "grad_norm": 25.725711822509766, "learning_rate": 1.746963776533953e-05, - "loss": 0.8113, + "loss": 1.6783, "step": 11980 }, { "epoch": 1.27, - "grad_norm": 1.646835207939148, + "grad_norm": 40.3188591003418, "learning_rate": 1.746752560988489e-05, - "loss": 0.5192, + "loss": 1.668, "step": 11990 }, { "epoch": 1.27, - "grad_norm": 37.75879669189453, + "grad_norm": 29.828163146972656, "learning_rate": 1.7465413454430248e-05, - "loss": 0.9756, + "loss": 2.0099, "step": 12000 }, { "epoch": 1.27, - "grad_norm": 13.294543266296387, + "grad_norm": 57.48540115356445, "learning_rate": 1.7463301298975605e-05, - "loss": 0.815, + "loss": 1.6568, "step": 12010 }, { "epoch": 1.27, - "grad_norm": 10.222929000854492, + "grad_norm": 42.08393096923828, "learning_rate": 1.7461189143520966e-05, - "loss": 0.6547, + "loss": 1.6461, "step": 12020 }, { "epoch": 1.27, - "grad_norm": 12.51098918914795, + "grad_norm": 39.96010971069336, "learning_rate": 1.7459076988066323e-05, - "loss": 0.5851, + "loss": 1.9393, "step": 12030 }, { "epoch": 1.27, - "grad_norm": 18.660140991210938, + "grad_norm": 39.92519760131836, "learning_rate": 1.7456964832611683e-05, - "loss": 1.131, + "loss": 1.7972, "step": 12040 }, { "epoch": 1.27, - "grad_norm": 13.021424293518066, + "grad_norm": 22.36416244506836, "learning_rate": 1.745485267715704e-05, - "loss": 0.7094, + "loss": 1.7068, "step": 12050 }, { "epoch": 1.27, - "grad_norm": 14.964948654174805, + "grad_norm": 38.263511657714844, "learning_rate": 1.7452740521702398e-05, - "loss": 0.8213, + "loss": 1.6636, "step": 12060 }, { "epoch": 1.27, - "grad_norm": 24.345382690429688, + "grad_norm": 30.628023147583008, "learning_rate": 1.7450628366247758e-05, - "loss": 0.6792, + "loss": 1.6747, "step": 12070 }, { "epoch": 1.28, - "grad_norm": 16.84917640686035, + "grad_norm": 52.985286712646484, "learning_rate": 1.7448516210793115e-05, - "loss": 0.8776, + "loss": 1.8484, "step": 12080 }, { "epoch": 1.28, - "grad_norm": 11.64574146270752, + "grad_norm": 46.66072463989258, "learning_rate": 1.7446404055338476e-05, - "loss": 0.7726, + "loss": 1.6563, "step": 12090 }, { "epoch": 1.28, - "grad_norm": 13.613197326660156, + "grad_norm": 29.667545318603516, "learning_rate": 1.7444291899883833e-05, - "loss": 0.7295, + "loss": 1.9469, "step": 12100 }, { "epoch": 1.28, - "grad_norm": 4.039699077606201, + "grad_norm": 30.681699752807617, "learning_rate": 1.744217974442919e-05, - "loss": 1.0732, + "loss": 1.7646, "step": 12110 }, { "epoch": 1.28, - "grad_norm": 7.526183128356934, + "grad_norm": 30.98236083984375, "learning_rate": 1.744006758897455e-05, - "loss": 0.6571, + "loss": 1.5174, "step": 12120 }, { "epoch": 1.28, - "grad_norm": 17.666629791259766, + "grad_norm": 57.50918197631836, "learning_rate": 1.7437955433519907e-05, - "loss": 1.016, + "loss": 1.769, "step": 12130 }, { "epoch": 1.28, - "grad_norm": 8.458776473999023, + "grad_norm": 26.632131576538086, "learning_rate": 1.7435843278065268e-05, - "loss": 0.8942, + "loss": 1.5665, "step": 12140 }, { "epoch": 1.28, - "grad_norm": 13.89876937866211, + "grad_norm": 31.41177749633789, "learning_rate": 1.7433731122610625e-05, - "loss": 0.8047, + "loss": 1.669, "step": 12150 }, { "epoch": 1.28, - "grad_norm": 8.271944999694824, + "grad_norm": 25.62762451171875, "learning_rate": 1.7431618967155982e-05, - "loss": 0.85, + "loss": 1.5512, "step": 12160 }, { "epoch": 1.29, - "grad_norm": 15.528093338012695, + "grad_norm": 44.19351577758789, "learning_rate": 1.7429506811701342e-05, - "loss": 1.1048, + "loss": 1.6648, "step": 12170 }, { "epoch": 1.29, - "grad_norm": 4.687641620635986, + "grad_norm": 38.16117477416992, "learning_rate": 1.74273946562467e-05, - "loss": 0.7202, + "loss": 1.8527, "step": 12180 }, { "epoch": 1.29, - "grad_norm": 19.28262710571289, + "grad_norm": 43.45891189575195, "learning_rate": 1.742528250079206e-05, - "loss": 1.0381, + "loss": 1.7396, "step": 12190 }, { "epoch": 1.29, - "grad_norm": 12.150522232055664, + "grad_norm": 36.3878059387207, "learning_rate": 1.742317034533742e-05, - "loss": 0.6596, + "loss": 1.772, "step": 12200 }, { "epoch": 1.29, - "grad_norm": 12.751809120178223, + "grad_norm": 44.66831970214844, "learning_rate": 1.7421058189882774e-05, - "loss": 0.9775, + "loss": 1.8634, "step": 12210 }, { "epoch": 1.29, - "grad_norm": 18.779024124145508, + "grad_norm": 34.4548454284668, "learning_rate": 1.7418946034428135e-05, - "loss": 0.7467, + "loss": 1.5971, "step": 12220 }, { "epoch": 1.29, - "grad_norm": 22.2237548828125, + "grad_norm": 47.014869689941406, "learning_rate": 1.7416833878973495e-05, - "loss": 0.7894, + "loss": 1.861, "step": 12230 }, { "epoch": 1.29, - "grad_norm": 25.045425415039062, + "grad_norm": 29.474571228027344, "learning_rate": 1.7414721723518852e-05, - "loss": 0.9831, + "loss": 1.6705, "step": 12240 }, { "epoch": 1.29, - "grad_norm": 8.968818664550781, + "grad_norm": 30.987119674682617, "learning_rate": 1.7412609568064213e-05, - "loss": 1.1577, + "loss": 1.9014, "step": 12250 }, { "epoch": 1.29, - "grad_norm": 4.73040771484375, + "grad_norm": 40.17741394042969, "learning_rate": 1.741049741260957e-05, - "loss": 0.8252, + "loss": 2.0685, "step": 12260 }, { "epoch": 1.3, - "grad_norm": 22.371509552001953, + "grad_norm": 42.234291076660156, "learning_rate": 1.7408385257154927e-05, - "loss": 0.8471, + "loss": 1.7691, "step": 12270 }, { "epoch": 1.3, - "grad_norm": 22.708852767944336, + "grad_norm": 47.02849197387695, "learning_rate": 1.7406273101700287e-05, - "loss": 0.8058, + "loss": 1.7289, "step": 12280 }, { "epoch": 1.3, - "grad_norm": 3.019073724746704, + "grad_norm": 40.902244567871094, "learning_rate": 1.7404160946245644e-05, - "loss": 0.9731, + "loss": 1.7511, "step": 12290 }, { "epoch": 1.3, - "grad_norm": 9.177370071411133, + "grad_norm": 21.4215145111084, "learning_rate": 1.7402048790791005e-05, - "loss": 0.882, + "loss": 1.6901, "step": 12300 }, { "epoch": 1.3, - "grad_norm": 13.491042137145996, + "grad_norm": 43.552215576171875, "learning_rate": 1.7399936635336362e-05, - "loss": 0.931, + "loss": 1.6975, "step": 12310 }, { "epoch": 1.3, - "grad_norm": 10.105944633483887, + "grad_norm": 41.51939392089844, "learning_rate": 1.739782447988172e-05, - "loss": 0.9479, + "loss": 1.852, "step": 12320 }, { "epoch": 1.3, - "grad_norm": 11.222771644592285, + "grad_norm": 31.56460952758789, "learning_rate": 1.739571232442708e-05, - "loss": 0.6805, + "loss": 1.6882, "step": 12330 }, { "epoch": 1.3, - "grad_norm": 2.2398037910461426, + "grad_norm": 51.0305290222168, "learning_rate": 1.7393600168972437e-05, - "loss": 0.7537, + "loss": 1.5539, "step": 12340 }, { "epoch": 1.3, - "grad_norm": 23.796466827392578, + "grad_norm": 31.933958053588867, "learning_rate": 1.7391488013517797e-05, - "loss": 0.9401, + "loss": 1.8339, "step": 12350 }, { "epoch": 1.31, - "grad_norm": 13.870549201965332, + "grad_norm": 31.957719802856445, "learning_rate": 1.7389375858063154e-05, - "loss": 0.9345, + "loss": 1.6674, "step": 12360 }, { "epoch": 1.31, - "grad_norm": 14.090126991271973, + "grad_norm": 42.430206298828125, "learning_rate": 1.7387263702608515e-05, - "loss": 0.8189, + "loss": 1.6219, "step": 12370 }, { "epoch": 1.31, - "grad_norm": 22.894222259521484, + "grad_norm": 45.70877456665039, "learning_rate": 1.738515154715387e-05, - "loss": 0.9036, + "loss": 1.8088, "step": 12380 }, { "epoch": 1.31, - "grad_norm": 13.327934265136719, + "grad_norm": 36.1104621887207, "learning_rate": 1.738303939169923e-05, - "loss": 0.7925, + "loss": 1.6869, "step": 12390 }, { "epoch": 1.31, - "grad_norm": 17.98715591430664, + "grad_norm": 36.43191909790039, "learning_rate": 1.738092723624459e-05, - "loss": 0.9532, + "loss": 1.685, "step": 12400 }, { "epoch": 1.31, - "grad_norm": 10.24599838256836, + "grad_norm": 45.29374313354492, "learning_rate": 1.737881508078995e-05, - "loss": 0.8132, + "loss": 1.77, "step": 12410 }, { "epoch": 1.31, - "grad_norm": 32.85612487792969, + "grad_norm": 47.256126403808594, "learning_rate": 1.7376702925335307e-05, - "loss": 0.8088, + "loss": 1.6849, "step": 12420 }, { "epoch": 1.31, - "grad_norm": 15.53584098815918, + "grad_norm": 41.611358642578125, "learning_rate": 1.7374590769880664e-05, - "loss": 0.9885, + "loss": 1.7758, "step": 12430 }, { "epoch": 1.31, - "grad_norm": 14.67029857635498, + "grad_norm": 33.23782730102539, "learning_rate": 1.7372478614426024e-05, - "loss": 0.893, + "loss": 1.7175, "step": 12440 }, { "epoch": 1.31, - "grad_norm": 11.481440544128418, + "grad_norm": 33.82474136352539, "learning_rate": 1.737036645897138e-05, - "loss": 0.9467, + "loss": 2.0147, "step": 12450 }, { "epoch": 1.32, - "grad_norm": 9.53743839263916, + "grad_norm": 34.0095100402832, "learning_rate": 1.7368254303516742e-05, - "loss": 0.6808, + "loss": 1.5495, "step": 12460 }, { "epoch": 1.32, - "grad_norm": 4.2207350730896, + "grad_norm": 32.6627311706543, "learning_rate": 1.73661421480621e-05, - "loss": 0.9242, + "loss": 1.7015, "step": 12470 }, { "epoch": 1.32, - "grad_norm": 16.48021697998047, + "grad_norm": 34.22625732421875, "learning_rate": 1.7364029992607456e-05, - "loss": 0.6259, + "loss": 1.5078, "step": 12480 }, { "epoch": 1.32, - "grad_norm": 20.526124954223633, + "grad_norm": 41.27090835571289, "learning_rate": 1.7361917837152817e-05, - "loss": 0.8413, + "loss": 1.4834, "step": 12490 }, { "epoch": 1.32, - "grad_norm": 15.21764087677002, + "grad_norm": 24.140024185180664, "learning_rate": 1.7359805681698174e-05, - "loss": 0.9177, + "loss": 1.3666, "step": 12500 }, { "epoch": 1.32, - "grad_norm": 13.027366638183594, + "grad_norm": 35.880069732666016, "learning_rate": 1.7357693526243534e-05, - "loss": 0.7874, + "loss": 1.7939, "step": 12510 }, { "epoch": 1.32, - "grad_norm": 29.20305061340332, + "grad_norm": 56.71812438964844, "learning_rate": 1.735558137078889e-05, - "loss": 0.7585, + "loss": 1.4909, "step": 12520 }, { "epoch": 1.32, - "grad_norm": 2.724546194076538, + "grad_norm": 28.86180877685547, "learning_rate": 1.735346921533425e-05, - "loss": 0.8607, + "loss": 1.743, "step": 12530 }, { "epoch": 1.32, - "grad_norm": 14.099873542785645, + "grad_norm": 36.28968048095703, "learning_rate": 1.735135705987961e-05, - "loss": 0.7835, + "loss": 1.44, "step": 12540 }, { "epoch": 1.33, - "grad_norm": 11.966143608093262, + "grad_norm": 49.67554473876953, "learning_rate": 1.7349244904424966e-05, - "loss": 0.7912, + "loss": 1.5539, "step": 12550 }, { "epoch": 1.33, - "grad_norm": 7.087778091430664, + "grad_norm": 36.59000778198242, "learning_rate": 1.7347132748970326e-05, - "loss": 0.6787, + "loss": 1.5868, "step": 12560 }, { "epoch": 1.33, - "grad_norm": 25.391592025756836, + "grad_norm": 15.205747604370117, "learning_rate": 1.7345020593515683e-05, - "loss": 0.7575, + "loss": 1.3342, "step": 12570 }, { "epoch": 1.33, - "grad_norm": 14.826534271240234, + "grad_norm": 53.17247772216797, "learning_rate": 1.7342908438061044e-05, - "loss": 0.9126, + "loss": 1.4852, "step": 12580 }, { "epoch": 1.33, - "grad_norm": 7.6731181144714355, + "grad_norm": 41.7662353515625, "learning_rate": 1.73407962826064e-05, - "loss": 0.7059, + "loss": 1.752, "step": 12590 }, { "epoch": 1.33, - "grad_norm": 19.6185245513916, + "grad_norm": 34.21567916870117, "learning_rate": 1.7338684127151758e-05, - "loss": 1.0065, + "loss": 2.026, "step": 12600 }, { "epoch": 1.33, - "grad_norm": 17.95235824584961, + "grad_norm": 32.50299835205078, "learning_rate": 1.733657197169712e-05, - "loss": 1.1179, + "loss": 1.5291, "step": 12610 }, { "epoch": 1.33, - "grad_norm": 15.324780464172363, + "grad_norm": 34.99828338623047, "learning_rate": 1.7334459816242476e-05, - "loss": 0.8986, + "loss": 1.5435, "step": 12620 }, { "epoch": 1.33, - "grad_norm": 7.640527248382568, + "grad_norm": 35.5179557800293, "learning_rate": 1.7332347660787836e-05, - "loss": 0.8369, + "loss": 1.71, "step": 12630 }, { "epoch": 1.33, - "grad_norm": 7.014289855957031, + "grad_norm": 43.67693328857422, "learning_rate": 1.7330235505333197e-05, - "loss": 0.7163, + "loss": 1.7762, "step": 12640 }, { "epoch": 1.34, - "grad_norm": 9.375396728515625, + "grad_norm": 36.41702651977539, "learning_rate": 1.732812334987855e-05, - "loss": 0.9173, + "loss": 1.6419, "step": 12650 }, { "epoch": 1.34, - "grad_norm": 22.338825225830078, + "grad_norm": 33.6331901550293, "learning_rate": 1.732601119442391e-05, - "loss": 0.7435, + "loss": 1.7153, "step": 12660 }, { "epoch": 1.34, - "grad_norm": 12.136519432067871, + "grad_norm": 38.341400146484375, "learning_rate": 1.732389903896927e-05, - "loss": 0.7707, + "loss": 1.774, "step": 12670 }, { "epoch": 1.34, - "grad_norm": 13.47781753540039, + "grad_norm": 33.16124725341797, "learning_rate": 1.7321786883514628e-05, - "loss": 0.8771, + "loss": 1.419, "step": 12680 }, { "epoch": 1.34, - "grad_norm": 7.419195175170898, + "grad_norm": 32.131874084472656, "learning_rate": 1.731967472805999e-05, - "loss": 0.732, + "loss": 1.5963, "step": 12690 }, { "epoch": 1.34, - "grad_norm": 25.31893539428711, + "grad_norm": 29.66114616394043, "learning_rate": 1.7317562572605346e-05, - "loss": 0.7467, + "loss": 1.4021, "step": 12700 }, { "epoch": 1.34, - "grad_norm": 17.88739776611328, + "grad_norm": 29.854999542236328, "learning_rate": 1.7315450417150703e-05, - "loss": 0.7733, + "loss": 1.3971, "step": 12710 }, { "epoch": 1.34, - "grad_norm": 11.38181209564209, + "grad_norm": 36.62122344970703, "learning_rate": 1.7313338261696063e-05, - "loss": 0.7019, + "loss": 1.8001, "step": 12720 }, { "epoch": 1.34, - "grad_norm": 3.9816935062408447, + "grad_norm": 34.342838287353516, "learning_rate": 1.731122610624142e-05, - "loss": 0.8046, + "loss": 1.7328, "step": 12730 }, { "epoch": 1.35, - "grad_norm": 7.933648109436035, + "grad_norm": 45.85532760620117, "learning_rate": 1.730911395078678e-05, - "loss": 0.7315, + "loss": 1.2641, "step": 12740 }, { "epoch": 1.35, - "grad_norm": 20.5909423828125, + "grad_norm": 20.891084671020508, "learning_rate": 1.7307001795332138e-05, - "loss": 0.8094, + "loss": 1.6925, "step": 12750 }, { "epoch": 1.35, - "grad_norm": 7.5928778648376465, + "grad_norm": 32.68850326538086, "learning_rate": 1.7304889639877495e-05, - "loss": 1.3459, + "loss": 1.8337, "step": 12760 }, { "epoch": 1.35, - "grad_norm": 3.671916961669922, + "grad_norm": 37.43027114868164, "learning_rate": 1.7302777484422856e-05, - "loss": 0.6861, + "loss": 1.5943, "step": 12770 }, { "epoch": 1.35, - "grad_norm": 9.697502136230469, + "grad_norm": 30.749561309814453, "learning_rate": 1.7300665328968213e-05, - "loss": 0.6896, + "loss": 1.7044, "step": 12780 }, { "epoch": 1.35, - "grad_norm": 14.375003814697266, + "grad_norm": 39.02311706542969, "learning_rate": 1.7298553173513573e-05, - "loss": 0.9835, + "loss": 1.6471, "step": 12790 }, { "epoch": 1.35, - "grad_norm": 11.914655685424805, + "grad_norm": 29.9886531829834, "learning_rate": 1.729644101805893e-05, - "loss": 0.8599, + "loss": 1.3472, "step": 12800 }, { "epoch": 1.35, - "grad_norm": 14.135580062866211, + "grad_norm": 35.93680953979492, "learning_rate": 1.7294328862604287e-05, - "loss": 0.6668, + "loss": 1.3402, "step": 12810 }, { "epoch": 1.35, - "grad_norm": 9.146397590637207, + "grad_norm": 46.169681549072266, "learning_rate": 1.7292216707149648e-05, - "loss": 0.738, + "loss": 1.5795, "step": 12820 }, { "epoch": 1.35, - "grad_norm": 11.626301765441895, + "grad_norm": 31.623472213745117, "learning_rate": 1.7290104551695005e-05, - "loss": 1.0899, + "loss": 1.765, "step": 12830 }, { "epoch": 1.36, - "grad_norm": 2.3828861713409424, + "grad_norm": 33.325679779052734, "learning_rate": 1.7287992396240365e-05, - "loss": 0.6775, + "loss": 1.7305, "step": 12840 }, { "epoch": 1.36, - "grad_norm": 10.224312782287598, + "grad_norm": 35.905067443847656, "learning_rate": 1.7285880240785722e-05, - "loss": 0.7035, + "loss": 1.7544, "step": 12850 }, { "epoch": 1.36, - "grad_norm": 15.601432800292969, + "grad_norm": 28.24727439880371, "learning_rate": 1.728376808533108e-05, - "loss": 0.9786, + "loss": 1.6142, "step": 12860 }, { "epoch": 1.36, - "grad_norm": 6.416252613067627, + "grad_norm": 46.53988265991211, "learning_rate": 1.728165592987644e-05, - "loss": 0.7344, + "loss": 1.5875, "step": 12870 }, { "epoch": 1.36, - "grad_norm": 11.861733436584473, + "grad_norm": 31.38528823852539, "learning_rate": 1.7279543774421797e-05, - "loss": 0.8529, + "loss": 1.6706, "step": 12880 }, { "epoch": 1.36, - "grad_norm": 15.045185089111328, + "grad_norm": 34.04641342163086, "learning_rate": 1.7277431618967158e-05, - "loss": 0.5874, + "loss": 1.2571, "step": 12890 }, { "epoch": 1.36, - "grad_norm": 2.0113914012908936, + "grad_norm": 28.070356369018555, "learning_rate": 1.7275319463512518e-05, - "loss": 0.7937, + "loss": 1.6055, "step": 12900 }, { "epoch": 1.36, - "grad_norm": 15.264212608337402, + "grad_norm": 41.31293869018555, "learning_rate": 1.7273207308057875e-05, - "loss": 1.146, + "loss": 1.8013, "step": 12910 }, { "epoch": 1.36, - "grad_norm": 5.804940223693848, + "grad_norm": 28.81425666809082, "learning_rate": 1.7271095152603232e-05, - "loss": 0.6886, + "loss": 1.4845, "step": 12920 }, { "epoch": 1.37, - "grad_norm": 26.44223976135254, + "grad_norm": 41.06563949584961, "learning_rate": 1.7268982997148593e-05, - "loss": 0.8443, + "loss": 1.8556, "step": 12930 }, { "epoch": 1.37, - "grad_norm": 8.671557426452637, + "grad_norm": 31.412830352783203, "learning_rate": 1.726687084169395e-05, - "loss": 0.7792, + "loss": 1.7428, "step": 12940 }, { "epoch": 1.37, - "grad_norm": 14.561304092407227, + "grad_norm": 34.743465423583984, "learning_rate": 1.726475868623931e-05, - "loss": 0.6918, + "loss": 1.6069, "step": 12950 }, { "epoch": 1.37, - "grad_norm": 9.479365348815918, + "grad_norm": 39.55312728881836, "learning_rate": 1.7262646530784667e-05, - "loss": 0.87, + "loss": 1.8458, "step": 12960 }, { "epoch": 1.37, - "grad_norm": 27.061004638671875, + "grad_norm": 31.143325805664062, "learning_rate": 1.7260534375330024e-05, - "loss": 0.7538, + "loss": 1.5249, "step": 12970 }, { "epoch": 1.37, - "grad_norm": 14.995992660522461, + "grad_norm": 34.960330963134766, "learning_rate": 1.7258422219875385e-05, - "loss": 1.0868, + "loss": 1.8685, "step": 12980 }, { "epoch": 1.37, - "grad_norm": 21.961910247802734, + "grad_norm": 46.414005279541016, "learning_rate": 1.7256310064420742e-05, - "loss": 0.7336, + "loss": 1.3756, "step": 12990 }, { "epoch": 1.37, - "grad_norm": 6.903402805328369, + "grad_norm": 43.70166015625, "learning_rate": 1.7254197908966102e-05, - "loss": 0.6584, + "loss": 1.6769, "step": 13000 }, { "epoch": 1.37, - "grad_norm": 2.5328943729400635, + "grad_norm": 32.398284912109375, "learning_rate": 1.725208575351146e-05, - "loss": 0.9082, + "loss": 1.774, "step": 13010 }, { "epoch": 1.38, - "grad_norm": 22.635351181030273, + "grad_norm": 34.84547805786133, "learning_rate": 1.724997359805682e-05, - "loss": 0.9082, + "loss": 1.493, "step": 13020 }, { "epoch": 1.38, - "grad_norm": 1.9310498237609863, + "grad_norm": 35.63331604003906, "learning_rate": 1.7247861442602177e-05, - "loss": 0.7967, + "loss": 1.5668, "step": 13030 }, { "epoch": 1.38, - "grad_norm": 9.059701919555664, + "grad_norm": 32.39369583129883, "learning_rate": 1.7245749287147534e-05, - "loss": 0.8123, + "loss": 1.7498, "step": 13040 }, { "epoch": 1.38, - "grad_norm": 22.9952449798584, + "grad_norm": 36.34897994995117, "learning_rate": 1.7243637131692895e-05, - "loss": 0.7935, + "loss": 1.5844, "step": 13050 }, { "epoch": 1.38, - "grad_norm": 6.802870273590088, + "grad_norm": 51.08354187011719, "learning_rate": 1.7241524976238252e-05, - "loss": 0.8587, + "loss": 1.4703, "step": 13060 }, { "epoch": 1.38, - "grad_norm": 12.784022331237793, + "grad_norm": 40.16572952270508, "learning_rate": 1.7239412820783612e-05, - "loss": 0.4676, + "loss": 1.3852, "step": 13070 }, { "epoch": 1.38, - "grad_norm": 19.169878005981445, + "grad_norm": 33.30498123168945, "learning_rate": 1.723730066532897e-05, - "loss": 0.979, + "loss": 1.8207, "step": 13080 }, { "epoch": 1.38, - "grad_norm": 9.614853858947754, + "grad_norm": 33.07147979736328, "learning_rate": 1.7235188509874326e-05, - "loss": 0.7621, + "loss": 1.4226, "step": 13090 }, { "epoch": 1.38, - "grad_norm": 15.013684272766113, + "grad_norm": 34.94673156738281, "learning_rate": 1.7233076354419687e-05, - "loss": 0.7782, + "loss": 1.5402, "step": 13100 }, { "epoch": 1.38, - "grad_norm": 12.593729019165039, + "grad_norm": 38.320556640625, "learning_rate": 1.7230964198965044e-05, - "loss": 0.9077, + "loss": 1.6943, "step": 13110 }, { "epoch": 1.39, - "grad_norm": 5.811816215515137, + "grad_norm": 28.89352798461914, "learning_rate": 1.7228852043510404e-05, - "loss": 1.0709, + "loss": 1.7263, "step": 13120 }, { "epoch": 1.39, - "grad_norm": 13.099074363708496, + "grad_norm": 44.15498733520508, "learning_rate": 1.722673988805576e-05, - "loss": 0.6915, + "loss": 1.6893, "step": 13130 }, { "epoch": 1.39, - "grad_norm": 18.8128604888916, + "grad_norm": 34.82869338989258, "learning_rate": 1.722462773260112e-05, - "loss": 0.7197, + "loss": 1.6465, "step": 13140 }, { "epoch": 1.39, - "grad_norm": 8.073107719421387, + "grad_norm": 30.59201431274414, "learning_rate": 1.722251557714648e-05, - "loss": 0.4263, + "loss": 1.2199, "step": 13150 }, { "epoch": 1.39, - "grad_norm": 8.701998710632324, + "grad_norm": 29.67953872680664, "learning_rate": 1.722040342169184e-05, - "loss": 0.6468, + "loss": 1.6701, "step": 13160 }, { "epoch": 1.39, - "grad_norm": 10.80076789855957, + "grad_norm": 39.93280029296875, "learning_rate": 1.7218291266237197e-05, - "loss": 0.7637, + "loss": 1.6064, "step": 13170 }, { "epoch": 1.39, - "grad_norm": 12.169501304626465, + "grad_norm": 35.332942962646484, "learning_rate": 1.7216179110782557e-05, - "loss": 0.6844, + "loss": 1.6568, "step": 13180 }, { "epoch": 1.39, - "grad_norm": 10.160131454467773, + "grad_norm": 43.76956558227539, "learning_rate": 1.7214066955327914e-05, - "loss": 0.5971, + "loss": 1.6407, "step": 13190 }, { "epoch": 1.39, - "grad_norm": 15.936430931091309, + "grad_norm": 32.573143005371094, "learning_rate": 1.721195479987327e-05, - "loss": 0.8424, + "loss": 1.6693, "step": 13200 }, { "epoch": 1.4, - "grad_norm": 19.097272872924805, + "grad_norm": 44.80201721191406, "learning_rate": 1.7209842644418632e-05, - "loss": 0.8764, + "loss": 1.4211, "step": 13210 }, { "epoch": 1.4, - "grad_norm": 8.49358081817627, + "grad_norm": 37.78079605102539, "learning_rate": 1.720773048896399e-05, - "loss": 1.1758, + "loss": 1.7733, "step": 13220 }, { "epoch": 1.4, - "grad_norm": 16.64105987548828, + "grad_norm": 39.78375244140625, "learning_rate": 1.720561833350935e-05, - "loss": 0.6977, + "loss": 1.5361, "step": 13230 }, { "epoch": 1.4, - "grad_norm": 19.429153442382812, + "grad_norm": 42.6457405090332, "learning_rate": 1.7203506178054706e-05, - "loss": 0.9605, + "loss": 1.7278, "step": 13240 }, { "epoch": 1.4, - "grad_norm": 6.490841865539551, + "grad_norm": 38.13432312011719, "learning_rate": 1.7201394022600063e-05, - "loss": 0.8043, + "loss": 1.556, "step": 13250 }, { "epoch": 1.4, - "grad_norm": 14.361801147460938, + "grad_norm": 49.09416961669922, "learning_rate": 1.7199281867145424e-05, - "loss": 0.8994, + "loss": 2.0255, "step": 13260 }, { "epoch": 1.4, - "grad_norm": 14.323432922363281, + "grad_norm": 41.88057327270508, "learning_rate": 1.719716971169078e-05, - "loss": 0.7983, + "loss": 1.4293, "step": 13270 }, { "epoch": 1.4, - "grad_norm": 11.410667419433594, + "grad_norm": 43.29371643066406, "learning_rate": 1.719505755623614e-05, - "loss": 0.9206, + "loss": 1.7947, "step": 13280 }, { "epoch": 1.4, - "grad_norm": 21.6000919342041, + "grad_norm": 26.583553314208984, "learning_rate": 1.71929454007815e-05, - "loss": 0.9379, + "loss": 1.5173, "step": 13290 }, { "epoch": 1.4, - "grad_norm": 16.518810272216797, + "grad_norm": 35.753055572509766, "learning_rate": 1.7190833245326856e-05, - "loss": 0.5936, + "loss": 1.4197, "step": 13300 }, { "epoch": 1.41, - "grad_norm": 4.798384189605713, + "grad_norm": 35.137962341308594, "learning_rate": 1.7188721089872216e-05, - "loss": 0.7849, + "loss": 1.4231, "step": 13310 }, { "epoch": 1.41, - "grad_norm": 12.717670440673828, + "grad_norm": 54.49961471557617, "learning_rate": 1.7186608934417573e-05, - "loss": 0.6888, + "loss": 1.4826, "step": 13320 }, { "epoch": 1.41, - "grad_norm": 12.219707489013672, + "grad_norm": 43.69656753540039, "learning_rate": 1.7184496778962934e-05, - "loss": 0.8967, + "loss": 1.682, "step": 13330 }, { "epoch": 1.41, - "grad_norm": 6.29537296295166, + "grad_norm": 31.69919204711914, "learning_rate": 1.7182384623508294e-05, - "loss": 0.674, + "loss": 1.6533, "step": 13340 }, { "epoch": 1.41, - "grad_norm": 18.266746520996094, + "grad_norm": 36.13197708129883, "learning_rate": 1.7180272468053648e-05, - "loss": 0.9762, + "loss": 1.8803, "step": 13350 }, { "epoch": 1.41, - "grad_norm": 9.799602508544922, + "grad_norm": 18.502628326416016, "learning_rate": 1.7178160312599008e-05, - "loss": 0.6627, + "loss": 1.6846, "step": 13360 }, { "epoch": 1.41, - "grad_norm": 17.593128204345703, + "grad_norm": 29.81764030456543, "learning_rate": 1.717604815714437e-05, - "loss": 1.0068, + "loss": 1.7166, "step": 13370 }, { "epoch": 1.41, - "grad_norm": 16.350454330444336, + "grad_norm": 38.81675338745117, "learning_rate": 1.7173936001689726e-05, - "loss": 1.1438, + "loss": 1.9266, "step": 13380 }, { "epoch": 1.41, - "grad_norm": 8.622786521911621, + "grad_norm": 36.80306625366211, "learning_rate": 1.7171823846235086e-05, - "loss": 1.0412, + "loss": 1.8465, "step": 13390 }, { "epoch": 1.42, - "grad_norm": 4.097494602203369, + "grad_norm": 40.14484786987305, "learning_rate": 1.7169711690780443e-05, - "loss": 0.5123, + "loss": 1.6617, "step": 13400 }, { "epoch": 1.42, - "grad_norm": 10.309398651123047, + "grad_norm": 48.362876892089844, "learning_rate": 1.71675995353258e-05, - "loss": 0.6685, + "loss": 1.5367, "step": 13410 }, { "epoch": 1.42, - "grad_norm": 19.512771606445312, + "grad_norm": 42.46928024291992, "learning_rate": 1.716548737987116e-05, - "loss": 0.7994, + "loss": 1.6366, "step": 13420 }, { "epoch": 1.42, - "grad_norm": 3.9477903842926025, + "grad_norm": 30.251564025878906, "learning_rate": 1.7163375224416518e-05, - "loss": 0.8542, + "loss": 1.5402, "step": 13430 }, { "epoch": 1.42, - "grad_norm": 5.430290222167969, + "grad_norm": 18.08034324645996, "learning_rate": 1.716126306896188e-05, - "loss": 0.9718, + "loss": 1.2805, "step": 13440 }, { "epoch": 1.42, - "grad_norm": 4.560013771057129, + "grad_norm": 27.103580474853516, "learning_rate": 1.7159150913507236e-05, - "loss": 1.1563, + "loss": 1.5695, "step": 13450 }, { "epoch": 1.42, - "grad_norm": 16.569318771362305, + "grad_norm": 36.87199401855469, "learning_rate": 1.7157038758052593e-05, - "loss": 0.7454, + "loss": 1.535, "step": 13460 }, { "epoch": 1.42, - "grad_norm": 10.642500877380371, + "grad_norm": 34.987701416015625, "learning_rate": 1.7154926602597953e-05, - "loss": 0.9608, + "loss": 1.6104, "step": 13470 }, { "epoch": 1.42, - "grad_norm": 12.699240684509277, + "grad_norm": 23.103187561035156, "learning_rate": 1.715281444714331e-05, - "loss": 1.0235, + "loss": 1.6257, "step": 13480 }, { "epoch": 1.42, - "grad_norm": 13.155047416687012, + "grad_norm": 49.29670333862305, "learning_rate": 1.715070229168867e-05, - "loss": 1.0291, + "loss": 1.7597, "step": 13490 }, { "epoch": 1.43, - "grad_norm": 8.518393516540527, + "grad_norm": 33.70546340942383, "learning_rate": 1.7148590136234028e-05, - "loss": 0.857, + "loss": 1.4353, "step": 13500 }, { "epoch": 1.43, - "grad_norm": 11.123152732849121, + "grad_norm": 33.99032974243164, "learning_rate": 1.7146477980779385e-05, - "loss": 0.5476, + "loss": 1.4984, "step": 13510 }, { "epoch": 1.43, - "grad_norm": 18.54539680480957, + "grad_norm": 45.283939361572266, "learning_rate": 1.7144365825324745e-05, - "loss": 0.5304, + "loss": 1.5504, "step": 13520 }, { "epoch": 1.43, - "grad_norm": 3.595257043838501, + "grad_norm": 33.59208679199219, "learning_rate": 1.7142253669870102e-05, - "loss": 0.9297, + "loss": 1.625, "step": 13530 }, { "epoch": 1.43, - "grad_norm": 20.825397491455078, + "grad_norm": 36.571231842041016, "learning_rate": 1.7140141514415463e-05, - "loss": 0.8657, + "loss": 1.5179, "step": 13540 }, { "epoch": 1.43, - "grad_norm": 20.893259048461914, + "grad_norm": 48.89469909667969, "learning_rate": 1.713802935896082e-05, - "loss": 0.938, + "loss": 1.6678, "step": 13550 }, { "epoch": 1.43, - "grad_norm": 17.566465377807617, + "grad_norm": 33.4320182800293, "learning_rate": 1.713591720350618e-05, - "loss": 0.8168, + "loss": 1.6634, "step": 13560 }, { "epoch": 1.43, - "grad_norm": 3.542346239089966, + "grad_norm": 30.038867950439453, "learning_rate": 1.7133805048051538e-05, - "loss": 0.7358, + "loss": 1.4672, "step": 13570 }, { "epoch": 1.43, - "grad_norm": 12.034625053405762, + "grad_norm": 42.605743408203125, "learning_rate": 1.7131692892596895e-05, - "loss": 0.9369, + "loss": 1.8983, "step": 13580 }, { "epoch": 1.44, - "grad_norm": 6.7338786125183105, + "grad_norm": 49.400970458984375, "learning_rate": 1.7129580737142255e-05, - "loss": 0.773, + "loss": 1.8642, "step": 13590 }, { "epoch": 1.44, - "grad_norm": 4.992250442504883, + "grad_norm": 42.0948371887207, "learning_rate": 1.7127468581687616e-05, - "loss": 1.0405, + "loss": 1.6483, "step": 13600 }, { "epoch": 1.44, - "grad_norm": 18.865060806274414, + "grad_norm": 32.83050537109375, "learning_rate": 1.7125356426232973e-05, - "loss": 1.2503, + "loss": 1.7241, "step": 13610 }, { "epoch": 1.44, - "grad_norm": 12.548107147216797, + "grad_norm": 35.683895111083984, "learning_rate": 1.712324427077833e-05, - "loss": 0.8794, + "loss": 1.6887, "step": 13620 }, { "epoch": 1.44, - "grad_norm": 9.657562255859375, + "grad_norm": 39.126548767089844, "learning_rate": 1.712113211532369e-05, - "loss": 0.7686, + "loss": 1.3866, "step": 13630 }, { "epoch": 1.44, - "grad_norm": 18.640756607055664, + "grad_norm": 36.6098747253418, "learning_rate": 1.7119019959869047e-05, - "loss": 0.8852, + "loss": 1.3924, "step": 13640 }, { "epoch": 1.44, - "grad_norm": 16.998092651367188, + "grad_norm": 29.5964298248291, "learning_rate": 1.7116907804414408e-05, - "loss": 1.0579, + "loss": 1.569, "step": 13650 }, { "epoch": 1.44, - "grad_norm": 28.18645477294922, + "grad_norm": 30.15324592590332, "learning_rate": 1.7114795648959765e-05, - "loss": 0.7716, + "loss": 1.7461, "step": 13660 }, { "epoch": 1.44, - "grad_norm": 15.661962509155273, + "grad_norm": 35.84877395629883, "learning_rate": 1.7112683493505125e-05, - "loss": 0.9962, + "loss": 1.6349, "step": 13670 }, { "epoch": 1.44, - "grad_norm": 18.096078872680664, + "grad_norm": 32.87227249145508, "learning_rate": 1.7110571338050482e-05, - "loss": 0.8883, + "loss": 1.8004, "step": 13680 }, { "epoch": 1.45, - "grad_norm": 10.890569686889648, + "grad_norm": 30.233478546142578, "learning_rate": 1.710845918259584e-05, - "loss": 0.7176, + "loss": 1.446, "step": 13690 }, { "epoch": 1.45, - "grad_norm": 8.14527416229248, + "grad_norm": 51.010311126708984, "learning_rate": 1.71063470271412e-05, - "loss": 0.6613, + "loss": 1.4807, "step": 13700 }, { "epoch": 1.45, - "grad_norm": 10.124648094177246, + "grad_norm": 28.25497055053711, "learning_rate": 1.7104234871686557e-05, - "loss": 1.0137, + "loss": 1.7111, "step": 13710 }, { "epoch": 1.45, - "grad_norm": 13.704651832580566, + "grad_norm": 38.12765884399414, "learning_rate": 1.7102122716231918e-05, - "loss": 0.7654, + "loss": 1.8049, "step": 13720 }, { "epoch": 1.45, - "grad_norm": 11.199312210083008, + "grad_norm": 21.19139862060547, "learning_rate": 1.7100010560777275e-05, - "loss": 0.7583, + "loss": 1.7615, "step": 13730 }, { "epoch": 1.45, - "grad_norm": 7.359131336212158, + "grad_norm": 32.40729522705078, "learning_rate": 1.7097898405322632e-05, - "loss": 0.916, + "loss": 1.431, "step": 13740 }, { "epoch": 1.45, - "grad_norm": 22.89161491394043, + "grad_norm": 44.10797119140625, "learning_rate": 1.7095786249867992e-05, - "loss": 0.7497, + "loss": 1.7261, "step": 13750 }, { "epoch": 1.45, - "grad_norm": 7.279653072357178, + "grad_norm": 36.05775451660156, "learning_rate": 1.709367409441335e-05, - "loss": 0.5351, + "loss": 1.4558, "step": 13760 }, { "epoch": 1.45, - "grad_norm": 11.741752624511719, + "grad_norm": 30.58338165283203, "learning_rate": 1.709156193895871e-05, - "loss": 0.8741, + "loss": 1.7611, "step": 13770 }, { "epoch": 1.46, - "grad_norm": 2.7177445888519287, + "grad_norm": 35.777034759521484, "learning_rate": 1.7089449783504067e-05, - "loss": 0.5424, + "loss": 1.5498, "step": 13780 }, { "epoch": 1.46, - "grad_norm": 21.94522476196289, + "grad_norm": 37.82192611694336, "learning_rate": 1.7087337628049424e-05, - "loss": 0.9488, + "loss": 1.6008, "step": 13790 }, { "epoch": 1.46, - "grad_norm": 15.516861915588379, + "grad_norm": 30.343530654907227, "learning_rate": 1.7085225472594784e-05, - "loss": 0.6252, + "loss": 1.5293, "step": 13800 }, { "epoch": 1.46, - "grad_norm": 19.832555770874023, + "grad_norm": 32.254730224609375, "learning_rate": 1.708311331714014e-05, - "loss": 1.0826, + "loss": 1.7984, "step": 13810 }, { "epoch": 1.46, - "grad_norm": 10.65912914276123, + "grad_norm": 44.94441223144531, "learning_rate": 1.7081001161685502e-05, - "loss": 0.6508, + "loss": 1.477, "step": 13820 }, { "epoch": 1.46, - "grad_norm": 22.59879493713379, + "grad_norm": 44.29616165161133, "learning_rate": 1.7078889006230862e-05, - "loss": 1.0818, + "loss": 1.8311, "step": 13830 }, { "epoch": 1.46, - "grad_norm": 15.352065086364746, + "grad_norm": 31.6783447265625, "learning_rate": 1.7076776850776216e-05, - "loss": 0.9491, + "loss": 1.722, "step": 13840 }, { "epoch": 1.46, - "grad_norm": 13.343311309814453, + "grad_norm": 47.243865966796875, "learning_rate": 1.7074664695321577e-05, - "loss": 0.721, + "loss": 1.8004, "step": 13850 }, { "epoch": 1.46, - "grad_norm": 11.551780700683594, + "grad_norm": 30.979042053222656, "learning_rate": 1.7072552539866937e-05, - "loss": 0.7402, + "loss": 1.3619, "step": 13860 }, { "epoch": 1.46, - "grad_norm": 20.55280876159668, + "grad_norm": 31.11775016784668, "learning_rate": 1.7070440384412294e-05, - "loss": 0.9083, + "loss": 1.7284, "step": 13870 }, { "epoch": 1.47, - "grad_norm": 8.72763442993164, + "grad_norm": 37.80665588378906, "learning_rate": 1.7068328228957655e-05, - "loss": 0.578, + "loss": 1.4962, "step": 13880 }, { "epoch": 1.47, - "grad_norm": 14.84498119354248, + "grad_norm": 30.229902267456055, "learning_rate": 1.7066216073503012e-05, - "loss": 0.563, + "loss": 1.3602, "step": 13890 }, { "epoch": 1.47, - "grad_norm": 21.31743621826172, + "grad_norm": 37.30548095703125, "learning_rate": 1.706410391804837e-05, - "loss": 0.7457, + "loss": 1.639, "step": 13900 }, { "epoch": 1.47, - "grad_norm": 13.745339393615723, + "grad_norm": 50.351593017578125, "learning_rate": 1.706199176259373e-05, - "loss": 0.9362, + "loss": 1.5737, "step": 13910 }, { "epoch": 1.47, - "grad_norm": 16.84552001953125, + "grad_norm": 43.1234130859375, "learning_rate": 1.7059879607139086e-05, - "loss": 0.7647, + "loss": 1.6355, "step": 13920 }, { "epoch": 1.47, - "grad_norm": 14.323765754699707, + "grad_norm": 38.73136520385742, "learning_rate": 1.7057767451684447e-05, - "loss": 0.8423, + "loss": 1.5778, "step": 13930 }, { "epoch": 1.47, - "grad_norm": 10.392318725585938, + "grad_norm": 32.13945007324219, "learning_rate": 1.7055655296229804e-05, - "loss": 0.94, + "loss": 1.5509, "step": 13940 }, { "epoch": 1.47, - "grad_norm": 11.874152183532715, + "grad_norm": 36.11677932739258, "learning_rate": 1.705354314077516e-05, - "loss": 0.9393, + "loss": 1.4689, "step": 13950 }, { "epoch": 1.47, - "grad_norm": 15.845746040344238, + "grad_norm": 39.59710693359375, "learning_rate": 1.705143098532052e-05, - "loss": 0.9236, + "loss": 1.6277, "step": 13960 }, { "epoch": 1.48, - "grad_norm": 18.488082885742188, + "grad_norm": 38.10774230957031, "learning_rate": 1.704931882986588e-05, - "loss": 0.5533, + "loss": 1.4266, "step": 13970 }, { "epoch": 1.48, - "grad_norm": 6.750539302825928, + "grad_norm": 20.19846534729004, "learning_rate": 1.704720667441124e-05, - "loss": 1.0235, + "loss": 1.6827, "step": 13980 }, { "epoch": 1.48, - "grad_norm": 14.195852279663086, + "grad_norm": 43.645652770996094, "learning_rate": 1.7045094518956596e-05, - "loss": 0.6867, + "loss": 1.7198, "step": 13990 }, { "epoch": 1.48, - "grad_norm": 24.898860931396484, + "grad_norm": 46.57114028930664, "learning_rate": 1.7042982363501953e-05, - "loss": 0.8751, + "loss": 1.6273, "step": 14000 }, { "epoch": 1.48, - "grad_norm": 7.70442533493042, + "grad_norm": 30.7756404876709, "learning_rate": 1.7040870208047314e-05, - "loss": 0.984, + "loss": 1.9327, "step": 14010 }, { "epoch": 1.48, - "grad_norm": 4.010124683380127, + "grad_norm": 26.075653076171875, "learning_rate": 1.703875805259267e-05, - "loss": 0.7087, + "loss": 1.7628, "step": 14020 }, { "epoch": 1.48, - "grad_norm": 7.61772346496582, + "grad_norm": 18.96114730834961, "learning_rate": 1.703664589713803e-05, - "loss": 0.3172, + "loss": 1.3743, "step": 14030 }, { "epoch": 1.48, - "grad_norm": 11.366494178771973, + "grad_norm": 33.6757698059082, "learning_rate": 1.7034533741683392e-05, - "loss": 0.8069, + "loss": 1.7016, "step": 14040 }, { "epoch": 1.48, - "grad_norm": 20.429100036621094, + "grad_norm": 24.99017333984375, "learning_rate": 1.7032421586228745e-05, - "loss": 0.7227, + "loss": 1.489, "step": 14050 }, { "epoch": 1.48, - "grad_norm": 5.968420028686523, + "grad_norm": 25.413976669311523, "learning_rate": 1.7030309430774106e-05, - "loss": 0.9103, + "loss": 1.5872, "step": 14060 }, { "epoch": 1.49, - "grad_norm": 10.378107070922852, + "grad_norm": 35.785804748535156, "learning_rate": 1.7028197275319463e-05, - "loss": 0.9882, + "loss": 1.6701, "step": 14070 }, { "epoch": 1.49, - "grad_norm": 19.96439552307129, + "grad_norm": 39.611507415771484, "learning_rate": 1.7026085119864823e-05, - "loss": 0.6534, + "loss": 1.3274, "step": 14080 }, { "epoch": 1.49, - "grad_norm": 5.914409637451172, + "grad_norm": 40.24076843261719, "learning_rate": 1.7023972964410184e-05, - "loss": 0.7326, + "loss": 1.5319, "step": 14090 }, { "epoch": 1.49, - "grad_norm": 7.666558742523193, + "grad_norm": 23.31126594543457, "learning_rate": 1.702186080895554e-05, - "loss": 0.5521, + "loss": 1.4542, "step": 14100 }, { "epoch": 1.49, - "grad_norm": 12.008162498474121, + "grad_norm": 33.848876953125, "learning_rate": 1.7019748653500898e-05, - "loss": 0.9228, + "loss": 1.4796, "step": 14110 }, { "epoch": 1.49, - "grad_norm": 14.930242538452148, + "grad_norm": 41.7724723815918, "learning_rate": 1.701763649804626e-05, - "loss": 1.1473, + "loss": 1.4604, "step": 14120 }, { "epoch": 1.49, - "grad_norm": 15.568536758422852, + "grad_norm": 33.2882194519043, "learning_rate": 1.7015524342591616e-05, - "loss": 0.6398, + "loss": 1.6789, "step": 14130 }, { "epoch": 1.49, - "grad_norm": 33.693965911865234, + "grad_norm": 39.706214904785156, "learning_rate": 1.7013412187136976e-05, - "loss": 0.956, + "loss": 1.5565, "step": 14140 }, { "epoch": 1.49, - "grad_norm": 7.322058200836182, + "grad_norm": 39.13677215576172, "learning_rate": 1.7011300031682333e-05, - "loss": 0.6788, + "loss": 1.2865, "step": 14150 }, { "epoch": 1.5, - "grad_norm": 11.872466087341309, + "grad_norm": 39.644893646240234, "learning_rate": 1.700918787622769e-05, - "loss": 0.7258, + "loss": 1.6409, "step": 14160 }, { "epoch": 1.5, - "grad_norm": 11.771265983581543, + "grad_norm": 37.5308723449707, "learning_rate": 1.700707572077305e-05, - "loss": 0.8574, + "loss": 1.583, "step": 14170 }, { "epoch": 1.5, - "grad_norm": 17.212688446044922, + "grad_norm": 23.00368881225586, "learning_rate": 1.7004963565318408e-05, - "loss": 0.7201, + "loss": 1.4065, "step": 14180 }, { "epoch": 1.5, - "grad_norm": 16.530710220336914, + "grad_norm": 43.70872116088867, "learning_rate": 1.700285140986377e-05, - "loss": 0.9006, + "loss": 1.7085, "step": 14190 }, { "epoch": 1.5, - "grad_norm": 3.3210716247558594, + "grad_norm": 25.393064498901367, "learning_rate": 1.7000739254409125e-05, - "loss": 0.9894, + "loss": 1.5374, "step": 14200 }, { "epoch": 1.5, - "grad_norm": 20.44171905517578, + "grad_norm": 48.931068420410156, "learning_rate": 1.6998627098954486e-05, - "loss": 0.8734, + "loss": 1.7012, "step": 14210 }, { "epoch": 1.5, - "grad_norm": 8.120062828063965, + "grad_norm": 33.767948150634766, "learning_rate": 1.6996514943499843e-05, - "loss": 0.7926, + "loss": 1.5902, "step": 14220 }, { "epoch": 1.5, - "grad_norm": 10.59035587310791, + "grad_norm": 38.48503112792969, "learning_rate": 1.69944027880452e-05, - "loss": 0.8958, + "loss": 1.6181, "step": 14230 }, { "epoch": 1.5, - "grad_norm": 29.469425201416016, + "grad_norm": 50.36385726928711, "learning_rate": 1.699229063259056e-05, - "loss": 0.5924, + "loss": 1.4025, "step": 14240 }, { "epoch": 1.5, - "grad_norm": 6.216727256774902, + "grad_norm": 24.86573600769043, "learning_rate": 1.6990178477135918e-05, - "loss": 0.7695, + "loss": 1.3773, "step": 14250 }, { "epoch": 1.51, - "grad_norm": 7.066064834594727, + "grad_norm": 37.16029739379883, "learning_rate": 1.6988066321681278e-05, - "loss": 0.5347, + "loss": 1.2921, "step": 14260 }, { "epoch": 1.51, - "grad_norm": 10.806909561157227, + "grad_norm": 25.475784301757812, "learning_rate": 1.6985954166226635e-05, - "loss": 0.6497, + "loss": 1.5515, "step": 14270 }, { "epoch": 1.51, - "grad_norm": 13.354000091552734, + "grad_norm": 31.286388397216797, "learning_rate": 1.6983842010771992e-05, - "loss": 0.7487, + "loss": 1.3067, "step": 14280 }, { "epoch": 1.51, - "grad_norm": 14.74382495880127, + "grad_norm": 38.19669723510742, "learning_rate": 1.6981729855317353e-05, - "loss": 0.826, + "loss": 1.9956, "step": 14290 }, { "epoch": 1.51, - "grad_norm": 16.931682586669922, + "grad_norm": 40.021968841552734, "learning_rate": 1.6979617699862713e-05, - "loss": 1.1594, + "loss": 1.8552, "step": 14300 }, { "epoch": 1.51, - "grad_norm": 9.566332817077637, + "grad_norm": 35.57002258300781, "learning_rate": 1.697750554440807e-05, - "loss": 0.8368, + "loss": 1.3639, "step": 14310 }, { "epoch": 1.51, - "grad_norm": 7.008213043212891, + "grad_norm": 43.5737190246582, "learning_rate": 1.6975393388953427e-05, - "loss": 0.9177, + "loss": 1.8754, "step": 14320 }, { "epoch": 1.51, - "grad_norm": 7.501955032348633, + "grad_norm": 35.663822174072266, "learning_rate": 1.6973281233498788e-05, - "loss": 0.8823, + "loss": 1.7995, "step": 14330 }, { "epoch": 1.51, - "grad_norm": 13.416635513305664, + "grad_norm": 30.685014724731445, "learning_rate": 1.6971169078044145e-05, - "loss": 0.7347, + "loss": 1.4759, "step": 14340 }, { "epoch": 1.52, - "grad_norm": 9.645087242126465, + "grad_norm": 27.170751571655273, "learning_rate": 1.6969056922589505e-05, - "loss": 0.5611, + "loss": 1.468, "step": 14350 }, { "epoch": 1.52, - "grad_norm": 16.064117431640625, + "grad_norm": 39.66084671020508, "learning_rate": 1.6966944767134862e-05, - "loss": 0.8449, + "loss": 1.9502, "step": 14360 }, { "epoch": 1.52, - "grad_norm": 1.864617943763733, + "grad_norm": 35.20427322387695, "learning_rate": 1.6964832611680223e-05, - "loss": 0.7119, + "loss": 1.6271, "step": 14370 }, { "epoch": 1.52, - "grad_norm": 12.431674003601074, + "grad_norm": 36.514766693115234, "learning_rate": 1.696272045622558e-05, - "loss": 0.7226, + "loss": 1.7221, "step": 14380 }, { "epoch": 1.52, - "grad_norm": 12.029240608215332, + "grad_norm": 34.43841552734375, "learning_rate": 1.6960608300770937e-05, - "loss": 0.8675, + "loss": 1.4405, "step": 14390 }, { "epoch": 1.52, - "grad_norm": 17.824617385864258, + "grad_norm": 40.9288444519043, "learning_rate": 1.6958496145316298e-05, - "loss": 0.7747, + "loss": 1.7458, "step": 14400 }, { "epoch": 1.52, - "grad_norm": 11.761622428894043, + "grad_norm": 31.96359634399414, "learning_rate": 1.6956383989861655e-05, - "loss": 0.5043, + "loss": 1.6625, "step": 14410 }, { "epoch": 1.52, - "grad_norm": 10.472146987915039, + "grad_norm": 39.17658996582031, "learning_rate": 1.6954271834407015e-05, - "loss": 0.6046, + "loss": 1.3681, "step": 14420 }, { "epoch": 1.52, - "grad_norm": 13.10097599029541, + "grad_norm": 30.776416778564453, "learning_rate": 1.6952159678952372e-05, - "loss": 0.8872, + "loss": 1.8026, "step": 14430 }, { "epoch": 1.52, - "grad_norm": 1.9392876625061035, + "grad_norm": 30.53438377380371, "learning_rate": 1.695004752349773e-05, - "loss": 0.7295, + "loss": 1.4872, "step": 14440 }, { "epoch": 1.53, - "grad_norm": 1.1955256462097168, + "grad_norm": 32.01997375488281, "learning_rate": 1.694793536804309e-05, - "loss": 0.7499, + "loss": 1.4805, "step": 14450 }, { "epoch": 1.53, - "grad_norm": 11.328275680541992, + "grad_norm": 33.71052932739258, "learning_rate": 1.6945823212588447e-05, - "loss": 0.75, + "loss": 1.6404, "step": 14460 }, { "epoch": 1.53, - "grad_norm": 15.812270164489746, + "grad_norm": 36.64118576049805, "learning_rate": 1.6943711057133807e-05, - "loss": 0.8543, + "loss": 1.6424, "step": 14470 }, { "epoch": 1.53, - "grad_norm": 14.526097297668457, + "grad_norm": 36.971675872802734, "learning_rate": 1.6941598901679164e-05, - "loss": 0.8248, + "loss": 1.6978, "step": 14480 }, { "epoch": 1.53, - "grad_norm": 16.726028442382812, + "grad_norm": 40.14161682128906, "learning_rate": 1.693948674622452e-05, - "loss": 0.8789, + "loss": 1.5927, "step": 14490 }, { "epoch": 1.53, - "grad_norm": 20.59229850769043, + "grad_norm": 33.07971954345703, "learning_rate": 1.6937374590769882e-05, - "loss": 0.4948, + "loss": 1.4727, "step": 14500 }, { "epoch": 1.53, - "grad_norm": 8.52121639251709, + "grad_norm": 31.043109893798828, "learning_rate": 1.693526243531524e-05, - "loss": 0.4686, + "loss": 1.3405, "step": 14510 }, { "epoch": 1.53, - "grad_norm": 24.441606521606445, + "grad_norm": 44.13527297973633, "learning_rate": 1.69331502798606e-05, - "loss": 0.9884, + "loss": 1.7124, "step": 14520 }, { "epoch": 1.53, - "grad_norm": 8.998770713806152, + "grad_norm": 41.3371467590332, "learning_rate": 1.693103812440596e-05, - "loss": 0.8262, + "loss": 1.6648, "step": 14530 }, { "epoch": 1.54, - "grad_norm": 20.969484329223633, + "grad_norm": 30.611116409301758, "learning_rate": 1.6928925968951314e-05, - "loss": 0.9267, + "loss": 1.6392, "step": 14540 }, { "epoch": 1.54, - "grad_norm": 13.65422534942627, + "grad_norm": 35.33376693725586, "learning_rate": 1.6926813813496674e-05, - "loss": 0.734, + "loss": 1.567, "step": 14550 }, { "epoch": 1.54, - "grad_norm": 3.279635429382324, + "grad_norm": 27.19129180908203, "learning_rate": 1.6924701658042035e-05, - "loss": 0.8388, + "loss": 1.4571, "step": 14560 }, { "epoch": 1.54, - "grad_norm": 16.099090576171875, + "grad_norm": 38.15247344970703, "learning_rate": 1.6922589502587392e-05, - "loss": 0.6599, + "loss": 1.7324, "step": 14570 }, { "epoch": 1.54, - "grad_norm": 15.187966346740723, + "grad_norm": 38.365234375, "learning_rate": 1.6920477347132752e-05, - "loss": 0.8621, + "loss": 1.5449, "step": 14580 }, { "epoch": 1.54, - "grad_norm": 23.507497787475586, + "grad_norm": 35.02291488647461, "learning_rate": 1.691836519167811e-05, - "loss": 1.0858, + "loss": 1.7138, "step": 14590 }, { "epoch": 1.54, - "grad_norm": 8.812139511108398, + "grad_norm": 44.3539924621582, "learning_rate": 1.6916253036223466e-05, - "loss": 0.7517, + "loss": 1.5017, "step": 14600 }, { "epoch": 1.54, - "grad_norm": 10.521097183227539, + "grad_norm": 40.58665084838867, "learning_rate": 1.6914140880768827e-05, - "loss": 0.7672, + "loss": 2.0714, "step": 14610 }, { "epoch": 1.54, - "grad_norm": 24.300050735473633, + "grad_norm": 37.653377532958984, "learning_rate": 1.6912028725314184e-05, - "loss": 0.6835, + "loss": 1.5556, "step": 14620 }, { "epoch": 1.55, - "grad_norm": 19.346309661865234, + "grad_norm": 35.35567855834961, "learning_rate": 1.6909916569859544e-05, - "loss": 0.8906, + "loss": 1.7233, "step": 14630 }, { "epoch": 1.55, - "grad_norm": 1.5917150974273682, + "grad_norm": 31.28340721130371, "learning_rate": 1.69078044144049e-05, - "loss": 0.5913, + "loss": 1.3708, "step": 14640 }, { "epoch": 1.55, - "grad_norm": 6.812307834625244, + "grad_norm": 26.185718536376953, "learning_rate": 1.690569225895026e-05, - "loss": 1.1362, + "loss": 1.6528, "step": 14650 }, { "epoch": 1.55, - "grad_norm": 20.95819091796875, + "grad_norm": 30.77073097229004, "learning_rate": 1.690358010349562e-05, - "loss": 0.6242, + "loss": 1.2623, "step": 14660 }, { "epoch": 1.55, - "grad_norm": 11.854418754577637, + "grad_norm": 41.071048736572266, "learning_rate": 1.6901467948040976e-05, - "loss": 0.7056, + "loss": 1.4719, "step": 14670 }, { "epoch": 1.55, - "grad_norm": 20.384302139282227, + "grad_norm": 50.19513702392578, "learning_rate": 1.6899355792586337e-05, - "loss": 0.8188, + "loss": 1.9356, "step": 14680 }, { "epoch": 1.55, - "grad_norm": 15.185027122497559, + "grad_norm": 38.583953857421875, "learning_rate": 1.6897243637131694e-05, - "loss": 0.7612, + "loss": 1.6489, "step": 14690 }, { "epoch": 1.55, - "grad_norm": 8.079459190368652, + "grad_norm": 30.00789451599121, "learning_rate": 1.689513148167705e-05, - "loss": 0.7007, + "loss": 1.6451, "step": 14700 }, { "epoch": 1.55, - "grad_norm": 7.661095142364502, + "grad_norm": 21.40056800842285, "learning_rate": 1.689301932622241e-05, - "loss": 0.9381, + "loss": 1.2844, "step": 14710 }, { "epoch": 1.55, - "grad_norm": 34.16789245605469, + "grad_norm": 46.526588439941406, "learning_rate": 1.689090717076777e-05, - "loss": 0.7809, + "loss": 1.3876, "step": 14720 }, { "epoch": 1.56, - "grad_norm": 10.784968376159668, + "grad_norm": 26.89333724975586, "learning_rate": 1.688879501531313e-05, - "loss": 0.9818, + "loss": 1.7681, "step": 14730 }, { "epoch": 1.56, - "grad_norm": 11.767501831054688, + "grad_norm": 23.43968963623047, "learning_rate": 1.6886682859858486e-05, - "loss": 0.649, + "loss": 1.5241, "step": 14740 }, { "epoch": 1.56, - "grad_norm": 22.0343074798584, + "grad_norm": 41.511016845703125, "learning_rate": 1.6884570704403846e-05, - "loss": 0.6048, + "loss": 1.2592, "step": 14750 }, { "epoch": 1.56, - "grad_norm": 17.173995971679688, + "grad_norm": 26.79228401184082, "learning_rate": 1.6882458548949203e-05, - "loss": 0.5192, + "loss": 1.2699, "step": 14760 }, { "epoch": 1.56, - "grad_norm": 11.884451866149902, + "grad_norm": 28.03432273864746, "learning_rate": 1.688034639349456e-05, - "loss": 0.6562, + "loss": 1.3657, "step": 14770 }, { "epoch": 1.56, - "grad_norm": 8.703150749206543, + "grad_norm": 35.91971206665039, "learning_rate": 1.687823423803992e-05, - "loss": 0.8173, + "loss": 1.3307, "step": 14780 }, { "epoch": 1.56, - "grad_norm": 43.3363037109375, + "grad_norm": 31.232608795166016, "learning_rate": 1.687612208258528e-05, - "loss": 0.8387, + "loss": 1.405, "step": 14790 }, { "epoch": 1.56, - "grad_norm": 2.5277862548828125, + "grad_norm": 28.016958236694336, "learning_rate": 1.687400992713064e-05, - "loss": 0.5033, + "loss": 1.4492, "step": 14800 }, { "epoch": 1.56, - "grad_norm": 10.958824157714844, + "grad_norm": 40.716739654541016, "learning_rate": 1.6871897771675996e-05, - "loss": 0.8921, + "loss": 1.5587, "step": 14810 }, { "epoch": 1.57, - "grad_norm": 17.30415916442871, + "grad_norm": 29.798154830932617, "learning_rate": 1.6869785616221356e-05, - "loss": 0.8174, + "loss": 1.392, "step": 14820 }, { "epoch": 1.57, - "grad_norm": 14.102200508117676, + "grad_norm": 40.10648727416992, "learning_rate": 1.6867673460766713e-05, - "loss": 0.6635, + "loss": 1.4672, "step": 14830 }, { "epoch": 1.57, - "grad_norm": 21.074739456176758, + "grad_norm": 36.29511260986328, "learning_rate": 1.6865561305312074e-05, - "loss": 0.9586, + "loss": 1.6896, "step": 14840 }, { "epoch": 1.57, - "grad_norm": 4.625239849090576, + "grad_norm": 22.289993286132812, "learning_rate": 1.686344914985743e-05, - "loss": 0.712, + "loss": 1.4699, "step": 14850 }, { "epoch": 1.57, - "grad_norm": 6.728367328643799, + "grad_norm": 29.01774787902832, "learning_rate": 1.686133699440279e-05, - "loss": 0.8403, + "loss": 1.3294, "step": 14860 }, { "epoch": 1.57, - "grad_norm": 14.804376602172852, + "grad_norm": 35.12028121948242, "learning_rate": 1.685922483894815e-05, - "loss": 0.5788, + "loss": 1.27, "step": 14870 }, { "epoch": 1.57, - "grad_norm": 7.88848876953125, + "grad_norm": 31.062274932861328, "learning_rate": 1.6857112683493505e-05, - "loss": 0.7999, + "loss": 1.3446, "step": 14880 }, { "epoch": 1.57, - "grad_norm": 17.93151092529297, + "grad_norm": 49.83659744262695, "learning_rate": 1.6855000528038866e-05, - "loss": 1.0265, + "loss": 1.5965, "step": 14890 }, { "epoch": 1.57, - "grad_norm": 9.168699264526367, + "grad_norm": 25.471542358398438, "learning_rate": 1.6852888372584223e-05, - "loss": 0.8361, + "loss": 1.6151, "step": 14900 }, { "epoch": 1.57, - "grad_norm": 10.091904640197754, + "grad_norm": 50.69866943359375, "learning_rate": 1.6850776217129583e-05, - "loss": 0.7793, + "loss": 1.9467, "step": 14910 }, { "epoch": 1.58, - "grad_norm": 10.566734313964844, + "grad_norm": 53.44159698486328, "learning_rate": 1.684866406167494e-05, - "loss": 0.8269, + "loss": 1.7777, "step": 14920 }, { "epoch": 1.58, - "grad_norm": 13.36301326751709, + "grad_norm": 13.854411125183105, "learning_rate": 1.6846551906220298e-05, - "loss": 0.6202, + "loss": 1.1507, "step": 14930 }, { "epoch": 1.58, - "grad_norm": 12.370048522949219, + "grad_norm": 40.007957458496094, "learning_rate": 1.6844439750765658e-05, - "loss": 0.819, + "loss": 1.4692, "step": 14940 }, { "epoch": 1.58, - "grad_norm": 7.360443592071533, + "grad_norm": 45.59891891479492, "learning_rate": 1.6842327595311015e-05, - "loss": 0.6579, + "loss": 1.3855, "step": 14950 }, { "epoch": 1.58, - "grad_norm": 20.92879295349121, + "grad_norm": 33.796913146972656, "learning_rate": 1.6840215439856376e-05, - "loss": 0.8435, + "loss": 1.5348, "step": 14960 }, { "epoch": 1.58, - "grad_norm": 16.893299102783203, + "grad_norm": 43.43107604980469, "learning_rate": 1.6838103284401733e-05, - "loss": 0.7273, + "loss": 1.611, "step": 14970 }, { "epoch": 1.58, - "grad_norm": 14.355326652526855, + "grad_norm": 42.3351936340332, "learning_rate": 1.683599112894709e-05, - "loss": 0.627, + "loss": 1.3339, "step": 14980 }, { "epoch": 1.58, - "grad_norm": 10.241607666015625, + "grad_norm": 43.27644729614258, "learning_rate": 1.683387897349245e-05, - "loss": 0.8952, + "loss": 1.3881, "step": 14990 }, { "epoch": 1.58, - "grad_norm": 10.502520561218262, + "grad_norm": 18.03287696838379, "learning_rate": 1.683176681803781e-05, - "loss": 0.4189, + "loss": 1.389, "step": 15000 }, { "epoch": 1.59, - "grad_norm": 7.593134880065918, + "grad_norm": 19.853349685668945, "learning_rate": 1.6829654662583168e-05, - "loss": 0.5998, + "loss": 1.4797, "step": 15010 }, { "epoch": 1.59, - "grad_norm": 9.55740737915039, + "grad_norm": 45.75817108154297, "learning_rate": 1.682754250712853e-05, - "loss": 0.9428, + "loss": 1.6672, "step": 15020 }, { "epoch": 1.59, - "grad_norm": 15.030692100524902, + "grad_norm": 48.584625244140625, "learning_rate": 1.6825430351673885e-05, - "loss": 0.7846, + "loss": 1.637, "step": 15030 }, { "epoch": 1.59, - "grad_norm": 14.671918869018555, + "grad_norm": 38.748046875, "learning_rate": 1.6823318196219242e-05, - "loss": 0.8029, + "loss": 1.5118, "step": 15040 }, { "epoch": 1.59, - "grad_norm": 17.578670501708984, + "grad_norm": 29.509769439697266, "learning_rate": 1.6821206040764603e-05, - "loss": 0.6209, + "loss": 1.3463, "step": 15050 }, { "epoch": 1.59, - "grad_norm": 11.60715103149414, + "grad_norm": 35.1699104309082, "learning_rate": 1.681909388530996e-05, - "loss": 0.5569, + "loss": 1.3899, "step": 15060 }, { "epoch": 1.59, - "grad_norm": 9.049909591674805, + "grad_norm": 25.561155319213867, "learning_rate": 1.681698172985532e-05, - "loss": 0.7352, + "loss": 1.7222, "step": 15070 }, { "epoch": 1.59, - "grad_norm": 5.359859943389893, + "grad_norm": 15.951748847961426, "learning_rate": 1.6814869574400678e-05, - "loss": 0.9987, + "loss": 1.3206, "step": 15080 }, { "epoch": 1.59, - "grad_norm": 18.65421485900879, + "grad_norm": 35.34206771850586, "learning_rate": 1.6812757418946035e-05, - "loss": 0.7911, + "loss": 1.4889, "step": 15090 }, { "epoch": 1.59, - "grad_norm": 10.211630821228027, + "grad_norm": 29.464679718017578, "learning_rate": 1.6810645263491395e-05, - "loss": 0.7272, + "loss": 1.3564, "step": 15100 }, { "epoch": 1.6, - "grad_norm": 14.382960319519043, + "grad_norm": 34.575469970703125, "learning_rate": 1.6808533108036752e-05, - "loss": 0.9942, + "loss": 1.7552, "step": 15110 }, { "epoch": 1.6, - "grad_norm": 13.036256790161133, + "grad_norm": 26.606931686401367, "learning_rate": 1.6806420952582113e-05, - "loss": 0.7417, + "loss": 1.5281, "step": 15120 }, { "epoch": 1.6, - "grad_norm": 12.64332389831543, + "grad_norm": 32.80021667480469, "learning_rate": 1.680430879712747e-05, - "loss": 0.9597, + "loss": 1.8418, "step": 15130 }, { "epoch": 1.6, - "grad_norm": 12.417752265930176, + "grad_norm": 31.871259689331055, "learning_rate": 1.6802196641672827e-05, - "loss": 0.6977, + "loss": 1.8197, "step": 15140 }, { "epoch": 1.6, - "grad_norm": 3.5796284675598145, + "grad_norm": 46.421058654785156, "learning_rate": 1.6800084486218187e-05, - "loss": 0.8361, + "loss": 1.8671, "step": 15150 }, { "epoch": 1.6, - "grad_norm": 14.09843921661377, + "grad_norm": 51.24848937988281, "learning_rate": 1.6797972330763544e-05, - "loss": 0.6265, + "loss": 1.1959, "step": 15160 }, { "epoch": 1.6, - "grad_norm": 9.470989227294922, + "grad_norm": 33.729888916015625, "learning_rate": 1.6795860175308905e-05, - "loss": 0.8238, + "loss": 1.4143, "step": 15170 }, { "epoch": 1.6, - "grad_norm": 4.818922519683838, + "grad_norm": 48.052249908447266, "learning_rate": 1.6793748019854262e-05, - "loss": 0.6396, + "loss": 1.6094, "step": 15180 }, { "epoch": 1.6, - "grad_norm": 25.972970962524414, + "grad_norm": 41.86079025268555, "learning_rate": 1.679163586439962e-05, - "loss": 0.5592, + "loss": 1.6222, "step": 15190 }, { "epoch": 1.61, - "grad_norm": 3.943021059036255, + "grad_norm": 21.13332176208496, "learning_rate": 1.678952370894498e-05, - "loss": 0.8843, + "loss": 1.369, "step": 15200 }, { "epoch": 1.61, - "grad_norm": 8.29585075378418, + "grad_norm": 29.638145446777344, "learning_rate": 1.6787411553490337e-05, - "loss": 1.0644, + "loss": 1.3106, "step": 15210 }, { "epoch": 1.61, - "grad_norm": 8.271078109741211, + "grad_norm": 40.256099700927734, "learning_rate": 1.6785299398035697e-05, - "loss": 0.8934, + "loss": 1.5698, "step": 15220 }, { "epoch": 1.61, - "grad_norm": 11.507357597351074, + "grad_norm": 31.648191452026367, "learning_rate": 1.6783187242581058e-05, - "loss": 0.6035, + "loss": 1.3254, "step": 15230 }, { "epoch": 1.61, - "grad_norm": 11.952230453491211, + "grad_norm": 32.09305191040039, "learning_rate": 1.6781075087126415e-05, - "loss": 0.8346, + "loss": 1.4517, "step": 15240 }, { "epoch": 1.61, - "grad_norm": 22.646717071533203, + "grad_norm": 27.682449340820312, "learning_rate": 1.6778962931671772e-05, - "loss": 0.8075, + "loss": 1.3914, "step": 15250 }, { "epoch": 1.61, - "grad_norm": 19.73562240600586, + "grad_norm": 46.05158996582031, "learning_rate": 1.6776850776217132e-05, - "loss": 0.5557, + "loss": 1.3581, "step": 15260 }, { "epoch": 1.61, - "grad_norm": 10.455239295959473, + "grad_norm": 51.22355270385742, "learning_rate": 1.677473862076249e-05, - "loss": 0.6965, + "loss": 1.4623, "step": 15270 }, { "epoch": 1.61, - "grad_norm": 25.108192443847656, + "grad_norm": 54.88381576538086, "learning_rate": 1.677262646530785e-05, - "loss": 0.8913, + "loss": 1.7969, "step": 15280 }, { "epoch": 1.61, - "grad_norm": 21.94588279724121, + "grad_norm": 36.22340393066406, "learning_rate": 1.6770514309853207e-05, - "loss": 0.5619, + "loss": 1.2781, "step": 15290 }, { "epoch": 1.62, - "grad_norm": 1.9740867614746094, + "grad_norm": 26.24148941040039, "learning_rate": 1.6768402154398564e-05, - "loss": 0.9417, + "loss": 1.7024, "step": 15300 }, { "epoch": 1.62, - "grad_norm": 15.683836936950684, + "grad_norm": 41.25564193725586, "learning_rate": 1.6766289998943924e-05, - "loss": 0.5239, + "loss": 1.3256, "step": 15310 }, { "epoch": 1.62, - "grad_norm": 16.54819679260254, + "grad_norm": 36.296363830566406, "learning_rate": 1.676417784348928e-05, - "loss": 0.8304, + "loss": 1.5216, "step": 15320 }, { "epoch": 1.62, - "grad_norm": 5.66841459274292, + "grad_norm": 34.82418441772461, "learning_rate": 1.6762065688034642e-05, - "loss": 0.6235, + "loss": 1.6612, "step": 15330 }, { "epoch": 1.62, - "grad_norm": 16.760427474975586, + "grad_norm": 39.65735626220703, "learning_rate": 1.675995353258e-05, - "loss": 0.8278, + "loss": 1.5163, "step": 15340 }, { "epoch": 1.62, - "grad_norm": 34.72072982788086, + "grad_norm": 37.9449577331543, "learning_rate": 1.6757841377125356e-05, - "loss": 0.848, + "loss": 1.5947, "step": 15350 }, { "epoch": 1.62, - "grad_norm": 20.128429412841797, + "grad_norm": 43.2384147644043, "learning_rate": 1.6755729221670717e-05, - "loss": 0.6788, + "loss": 1.8292, "step": 15360 }, { "epoch": 1.62, - "grad_norm": 14.350189208984375, + "grad_norm": 42.178192138671875, "learning_rate": 1.6753617066216074e-05, - "loss": 0.6716, + "loss": 1.461, "step": 15370 }, { "epoch": 1.62, - "grad_norm": 8.900530815124512, + "grad_norm": 35.328853607177734, "learning_rate": 1.6751504910761434e-05, - "loss": 0.9835, + "loss": 1.7736, "step": 15380 }, { "epoch": 1.63, - "grad_norm": 22.87515640258789, + "grad_norm": 49.34401321411133, "learning_rate": 1.674939275530679e-05, - "loss": 0.7373, + "loss": 1.6133, "step": 15390 }, { "epoch": 1.63, - "grad_norm": 7.842971324920654, + "grad_norm": 28.975353240966797, "learning_rate": 1.6747280599852152e-05, - "loss": 1.0177, + "loss": 1.7189, "step": 15400 }, { "epoch": 1.63, - "grad_norm": 19.300390243530273, + "grad_norm": 37.506568908691406, "learning_rate": 1.674516844439751e-05, - "loss": 0.532, + "loss": 1.3846, "step": 15410 }, { "epoch": 1.63, - "grad_norm": 28.47950553894043, + "grad_norm": 47.6309928894043, "learning_rate": 1.6743056288942866e-05, - "loss": 0.7968, + "loss": 1.5309, "step": 15420 }, { "epoch": 1.63, - "grad_norm": 21.0981502532959, + "grad_norm": 37.8336067199707, "learning_rate": 1.6740944133488226e-05, - "loss": 0.7375, + "loss": 1.6448, "step": 15430 }, { "epoch": 1.63, - "grad_norm": 5.127477169036865, + "grad_norm": 27.90473175048828, "learning_rate": 1.6738831978033583e-05, - "loss": 0.5914, + "loss": 1.5067, "step": 15440 }, { "epoch": 1.63, - "grad_norm": 13.48165512084961, + "grad_norm": 34.509765625, "learning_rate": 1.6736719822578944e-05, - "loss": 0.8234, + "loss": 1.6292, "step": 15450 }, { "epoch": 1.63, - "grad_norm": 12.76811695098877, + "grad_norm": 37.706748962402344, "learning_rate": 1.67346076671243e-05, - "loss": 0.71, + "loss": 1.353, "step": 15460 }, { "epoch": 1.63, - "grad_norm": 26.80246925354004, + "grad_norm": 33.50914764404297, "learning_rate": 1.6732495511669658e-05, - "loss": 0.9649, + "loss": 1.8566, "step": 15470 }, { "epoch": 1.63, - "grad_norm": 9.753400802612305, + "grad_norm": 45.81356430053711, "learning_rate": 1.673038335621502e-05, - "loss": 0.8585, + "loss": 1.5014, "step": 15480 }, { "epoch": 1.64, - "grad_norm": 19.596309661865234, + "grad_norm": 39.88015365600586, "learning_rate": 1.672827120076038e-05, - "loss": 1.1803, + "loss": 1.8397, "step": 15490 }, { "epoch": 1.64, - "grad_norm": 7.876928329467773, + "grad_norm": 43.46583938598633, "learning_rate": 1.6726159045305736e-05, - "loss": 1.002, + "loss": 1.6354, "step": 15500 }, { "epoch": 1.64, - "grad_norm": 17.26955795288086, + "grad_norm": 36.0395393371582, "learning_rate": 1.6724046889851097e-05, - "loss": 0.7267, + "loss": 1.5825, "step": 15510 }, { "epoch": 1.64, - "grad_norm": 6.7948455810546875, + "grad_norm": 23.38069725036621, "learning_rate": 1.6721934734396454e-05, - "loss": 0.6599, + "loss": 1.4677, "step": 15520 }, { "epoch": 1.64, - "grad_norm": 23.48771095275879, + "grad_norm": 48.4090690612793, "learning_rate": 1.671982257894181e-05, - "loss": 0.6638, + "loss": 1.5154, "step": 15530 }, { "epoch": 1.64, - "grad_norm": 4.275768280029297, + "grad_norm": 37.819252014160156, "learning_rate": 1.671771042348717e-05, - "loss": 0.5465, + "loss": 1.2605, "step": 15540 }, { "epoch": 1.64, - "grad_norm": 0.7272750735282898, + "grad_norm": 25.771589279174805, "learning_rate": 1.671559826803253e-05, - "loss": 0.6682, + "loss": 1.3845, "step": 15550 }, { "epoch": 1.64, - "grad_norm": 6.598641872406006, + "grad_norm": 24.91750717163086, "learning_rate": 1.671348611257789e-05, - "loss": 1.0392, + "loss": 1.6471, "step": 15560 }, { "epoch": 1.64, - "grad_norm": 3.405308246612549, + "grad_norm": 41.077999114990234, "learning_rate": 1.6711373957123246e-05, - "loss": 0.6063, + "loss": 1.6142, "step": 15570 }, { "epoch": 1.65, - "grad_norm": 13.914081573486328, + "grad_norm": 33.45512771606445, "learning_rate": 1.6709261801668603e-05, - "loss": 0.5108, + "loss": 1.3434, "step": 15580 }, { "epoch": 1.65, - "grad_norm": 13.634401321411133, + "grad_norm": 36.53364944458008, "learning_rate": 1.6707149646213963e-05, - "loss": 0.7759, + "loss": 1.7051, "step": 15590 }, { "epoch": 1.65, - "grad_norm": 2.59197998046875, + "grad_norm": 33.342796325683594, "learning_rate": 1.670503749075932e-05, - "loss": 0.5704, + "loss": 1.4234, "step": 15600 }, { "epoch": 1.65, - "grad_norm": 7.347860813140869, + "grad_norm": 27.006595611572266, "learning_rate": 1.670292533530468e-05, - "loss": 0.6851, + "loss": 1.3003, "step": 15610 }, { "epoch": 1.65, - "grad_norm": 13.745888710021973, + "grad_norm": 23.1190242767334, "learning_rate": 1.6700813179850038e-05, - "loss": 0.8978, + "loss": 1.3206, "step": 15620 }, { "epoch": 1.65, - "grad_norm": 19.971920013427734, + "grad_norm": 34.473506927490234, "learning_rate": 1.6698701024395395e-05, - "loss": 0.6444, + "loss": 1.7075, "step": 15630 }, { "epoch": 1.65, - "grad_norm": 5.087752819061279, + "grad_norm": 28.797805786132812, "learning_rate": 1.6696588868940756e-05, - "loss": 0.7589, + "loss": 1.5107, "step": 15640 }, { "epoch": 1.65, - "grad_norm": 2.959406614303589, + "grad_norm": 26.525840759277344, "learning_rate": 1.6694476713486113e-05, - "loss": 0.777, + "loss": 1.4261, "step": 15650 }, { "epoch": 1.65, - "grad_norm": 20.121810913085938, + "grad_norm": 48.580223083496094, "learning_rate": 1.6692364558031473e-05, - "loss": 0.791, + "loss": 1.5044, "step": 15660 }, { "epoch": 1.65, - "grad_norm": 21.231311798095703, + "grad_norm": 39.405738830566406, "learning_rate": 1.669025240257683e-05, - "loss": 0.9069, + "loss": 1.4414, "step": 15670 }, { "epoch": 1.66, - "grad_norm": 12.443953514099121, + "grad_norm": 30.496826171875, "learning_rate": 1.6688140247122187e-05, - "loss": 0.7834, + "loss": 1.4178, "step": 15680 }, { "epoch": 1.66, - "grad_norm": 2.247248649597168, + "grad_norm": 32.393123626708984, "learning_rate": 1.6686028091667548e-05, - "loss": 0.5918, + "loss": 1.2281, "step": 15690 }, { "epoch": 1.66, - "grad_norm": 12.59443473815918, + "grad_norm": 32.01408004760742, "learning_rate": 1.6683915936212905e-05, - "loss": 0.7303, + "loss": 1.3986, "step": 15700 }, { "epoch": 1.66, - "grad_norm": 12.83820629119873, + "grad_norm": 35.306373596191406, "learning_rate": 1.6681803780758265e-05, - "loss": 0.9614, + "loss": 1.9643, "step": 15710 }, { "epoch": 1.66, - "grad_norm": 5.938239097595215, + "grad_norm": 35.41657638549805, "learning_rate": 1.6679691625303626e-05, - "loss": 0.9583, + "loss": 1.5255, "step": 15720 }, { "epoch": 1.66, - "grad_norm": 17.74297523498535, + "grad_norm": 38.542911529541016, "learning_rate": 1.667757946984898e-05, - "loss": 0.8358, + "loss": 1.6757, "step": 15730 }, { "epoch": 1.66, - "grad_norm": 3.6865341663360596, + "grad_norm": 33.516563415527344, "learning_rate": 1.667546731439434e-05, - "loss": 0.4464, + "loss": 1.3757, "step": 15740 }, { "epoch": 1.66, - "grad_norm": 9.151399612426758, + "grad_norm": 27.46708106994629, "learning_rate": 1.66733551589397e-05, - "loss": 0.6252, + "loss": 1.0768, "step": 15750 }, { "epoch": 1.66, - "grad_norm": 8.217973709106445, + "grad_norm": 33.00764465332031, "learning_rate": 1.6671243003485058e-05, - "loss": 0.8426, + "loss": 1.7887, "step": 15760 }, { "epoch": 1.67, - "grad_norm": 21.82723045349121, + "grad_norm": 32.22587966918945, "learning_rate": 1.6669130848030418e-05, - "loss": 0.7913, + "loss": 1.5036, "step": 15770 }, { "epoch": 1.67, - "grad_norm": 16.022058486938477, + "grad_norm": 31.036563873291016, "learning_rate": 1.6667018692575775e-05, - "loss": 0.6617, + "loss": 1.5363, "step": 15780 }, { "epoch": 1.67, - "grad_norm": 6.4283647537231445, + "grad_norm": 33.29685592651367, "learning_rate": 1.6664906537121132e-05, - "loss": 0.7211, + "loss": 1.6421, "step": 15790 }, { "epoch": 1.67, - "grad_norm": 2.2965309619903564, + "grad_norm": 43.17213821411133, "learning_rate": 1.6662794381666493e-05, - "loss": 1.0143, + "loss": 1.9905, "step": 15800 }, { "epoch": 1.67, - "grad_norm": 28.58020782470703, + "grad_norm": 32.86947250366211, "learning_rate": 1.666068222621185e-05, - "loss": 0.7415, + "loss": 1.7739, "step": 15810 }, { "epoch": 1.67, - "grad_norm": 11.262870788574219, + "grad_norm": 31.5543212890625, "learning_rate": 1.665857007075721e-05, - "loss": 0.8705, + "loss": 1.6352, "step": 15820 }, { "epoch": 1.67, - "grad_norm": 16.080305099487305, + "grad_norm": 50.537452697753906, "learning_rate": 1.6656457915302567e-05, - "loss": 1.1366, + "loss": 1.4854, "step": 15830 }, { "epoch": 1.67, - "grad_norm": 13.965875625610352, + "grad_norm": 30.902326583862305, "learning_rate": 1.6654345759847924e-05, - "loss": 0.5949, + "loss": 1.1604, "step": 15840 }, { "epoch": 1.67, - "grad_norm": 15.512334823608398, + "grad_norm": 47.46070098876953, "learning_rate": 1.6652233604393285e-05, - "loss": 0.8234, + "loss": 1.6293, "step": 15850 }, { "epoch": 1.67, - "grad_norm": 2.352945566177368, + "grad_norm": 28.844520568847656, "learning_rate": 1.6650121448938642e-05, - "loss": 0.8187, + "loss": 1.5103, "step": 15860 }, { "epoch": 1.68, - "grad_norm": 10.489765167236328, + "grad_norm": 36.16962432861328, "learning_rate": 1.6648009293484002e-05, - "loss": 0.6368, + "loss": 1.5009, "step": 15870 }, { "epoch": 1.68, - "grad_norm": 15.53989315032959, + "grad_norm": 37.554527282714844, "learning_rate": 1.664589713802936e-05, - "loss": 0.548, + "loss": 1.4395, "step": 15880 }, { "epoch": 1.68, - "grad_norm": 5.436570167541504, + "grad_norm": 38.3531608581543, "learning_rate": 1.6643784982574717e-05, - "loss": 0.8209, + "loss": 1.5604, "step": 15890 }, { "epoch": 1.68, - "grad_norm": 12.504829406738281, + "grad_norm": 45.957672119140625, "learning_rate": 1.6641672827120077e-05, - "loss": 1.1175, + "loss": 2.2594, "step": 15900 }, { "epoch": 1.68, - "grad_norm": 16.282485961914062, + "grad_norm": 36.46647644042969, "learning_rate": 1.6639560671665434e-05, - "loss": 0.5566, + "loss": 1.4509, "step": 15910 }, { "epoch": 1.68, - "grad_norm": 12.394457817077637, + "grad_norm": 42.43968963623047, "learning_rate": 1.6637448516210795e-05, - "loss": 1.1404, + "loss": 1.5049, "step": 15920 }, { "epoch": 1.68, - "grad_norm": 23.482961654663086, + "grad_norm": 47.31413650512695, "learning_rate": 1.6635336360756155e-05, - "loss": 0.9323, + "loss": 1.8449, "step": 15930 }, { "epoch": 1.68, - "grad_norm": 24.016464233398438, + "grad_norm": 43.47674560546875, "learning_rate": 1.6633224205301512e-05, - "loss": 0.6705, + "loss": 1.5926, "step": 15940 }, { "epoch": 1.68, - "grad_norm": 6.672823905944824, + "grad_norm": 26.167741775512695, "learning_rate": 1.663111204984687e-05, - "loss": 0.5404, + "loss": 1.3196, "step": 15950 }, { "epoch": 1.69, - "grad_norm": 15.378896713256836, + "grad_norm": 30.1962833404541, "learning_rate": 1.662899989439223e-05, - "loss": 0.935, + "loss": 1.6998, "step": 15960 }, { "epoch": 1.69, - "grad_norm": 21.21975326538086, + "grad_norm": 34.96417999267578, "learning_rate": 1.6626887738937587e-05, - "loss": 0.7119, + "loss": 1.4395, "step": 15970 }, { "epoch": 1.69, - "grad_norm": 1.4936856031417847, + "grad_norm": 32.97806930541992, "learning_rate": 1.6624775583482947e-05, - "loss": 0.8998, + "loss": 1.6694, "step": 15980 }, { "epoch": 1.69, - "grad_norm": 12.756623268127441, + "grad_norm": 39.983924865722656, "learning_rate": 1.6622663428028304e-05, - "loss": 0.8898, + "loss": 1.7269, "step": 15990 }, { "epoch": 1.69, - "grad_norm": 6.619269371032715, + "grad_norm": 39.45886993408203, "learning_rate": 1.662055127257366e-05, - "loss": 0.4993, + "loss": 1.3979, "step": 16000 }, { "epoch": 1.69, - "grad_norm": 11.324191093444824, + "grad_norm": 36.537837982177734, "learning_rate": 1.6618439117119022e-05, - "loss": 0.6805, + "loss": 1.4393, "step": 16010 }, { "epoch": 1.69, - "grad_norm": 6.9695539474487305, + "grad_norm": 18.65880012512207, "learning_rate": 1.661632696166438e-05, - "loss": 0.7383, + "loss": 1.466, "step": 16020 }, { "epoch": 1.69, - "grad_norm": 9.899721145629883, + "grad_norm": 32.19779586791992, "learning_rate": 1.661421480620974e-05, - "loss": 0.6013, + "loss": 1.5231, "step": 16030 }, { "epoch": 1.69, - "grad_norm": 10.918224334716797, + "grad_norm": 46.75515365600586, "learning_rate": 1.6612102650755097e-05, - "loss": 0.9268, + "loss": 1.7529, "step": 16040 }, { "epoch": 1.7, - "grad_norm": 2.6782171726226807, + "grad_norm": 45.63736343383789, "learning_rate": 1.6609990495300457e-05, - "loss": 1.08, + "loss": 1.639, "step": 16050 }, { "epoch": 1.7, - "grad_norm": 13.393369674682617, + "grad_norm": 36.85334396362305, "learning_rate": 1.6607878339845814e-05, - "loss": 0.6722, + "loss": 1.7483, "step": 16060 }, { "epoch": 1.7, - "grad_norm": 3.4703333377838135, + "grad_norm": 33.41267395019531, "learning_rate": 1.660576618439117e-05, - "loss": 0.812, + "loss": 1.3421, "step": 16070 }, { "epoch": 1.7, - "grad_norm": 13.252583503723145, + "grad_norm": 26.2052001953125, "learning_rate": 1.6603654028936532e-05, - "loss": 0.7698, + "loss": 1.3251, "step": 16080 }, { "epoch": 1.7, - "grad_norm": 19.2059383392334, + "grad_norm": 32.4752311706543, "learning_rate": 1.660154187348189e-05, - "loss": 1.1766, + "loss": 1.4459, "step": 16090 }, { "epoch": 1.7, - "grad_norm": 15.782477378845215, + "grad_norm": 36.428592681884766, "learning_rate": 1.659942971802725e-05, - "loss": 0.8299, + "loss": 1.5859, "step": 16100 }, { "epoch": 1.7, - "grad_norm": 4.403731346130371, + "grad_norm": 38.23109436035156, "learning_rate": 1.6597317562572606e-05, - "loss": 0.8551, + "loss": 1.715, "step": 16110 }, { "epoch": 1.7, - "grad_norm": 28.549814224243164, + "grad_norm": 37.99486541748047, "learning_rate": 1.6595205407117963e-05, - "loss": 0.7078, + "loss": 1.4572, "step": 16120 }, { "epoch": 1.7, - "grad_norm": 10.366186141967773, + "grad_norm": 42.31111145019531, "learning_rate": 1.6593093251663324e-05, - "loss": 0.6211, + "loss": 1.5127, "step": 16130 }, { "epoch": 1.7, - "grad_norm": 9.66995620727539, + "grad_norm": 36.992034912109375, "learning_rate": 1.659098109620868e-05, - "loss": 1.1046, + "loss": 1.576, "step": 16140 }, { "epoch": 1.71, - "grad_norm": 27.43171501159668, + "grad_norm": 36.15107727050781, "learning_rate": 1.658886894075404e-05, - "loss": 0.5919, + "loss": 1.5095, "step": 16150 }, { "epoch": 1.71, - "grad_norm": 13.849665641784668, + "grad_norm": 40.38701629638672, "learning_rate": 1.6586756785299402e-05, - "loss": 0.7499, + "loss": 1.5627, "step": 16160 }, { "epoch": 1.71, - "grad_norm": 10.104740142822266, + "grad_norm": 35.33122253417969, "learning_rate": 1.6584644629844756e-05, - "loss": 0.8473, + "loss": 1.6155, "step": 16170 }, { "epoch": 1.71, - "grad_norm": 5.654960632324219, + "grad_norm": 43.42594528198242, "learning_rate": 1.6582532474390116e-05, - "loss": 0.6456, + "loss": 1.4561, "step": 16180 }, { "epoch": 1.71, - "grad_norm": 11.21319580078125, + "grad_norm": 46.23091506958008, "learning_rate": 1.6580420318935477e-05, - "loss": 0.6382, + "loss": 1.4573, "step": 16190 }, { "epoch": 1.71, - "grad_norm": 20.215770721435547, + "grad_norm": 33.44566345214844, "learning_rate": 1.6578308163480834e-05, - "loss": 0.625, + "loss": 1.1187, "step": 16200 }, { "epoch": 1.71, - "grad_norm": 6.894721508026123, + "grad_norm": 43.09161376953125, "learning_rate": 1.6576196008026194e-05, - "loss": 0.5951, + "loss": 1.6981, "step": 16210 }, { "epoch": 1.71, - "grad_norm": 7.224364280700684, + "grad_norm": 47.16818618774414, "learning_rate": 1.657408385257155e-05, - "loss": 0.7067, + "loss": 1.2321, "step": 16220 }, { "epoch": 1.71, - "grad_norm": 10.243841171264648, + "grad_norm": 60.88861846923828, "learning_rate": 1.657197169711691e-05, - "loss": 0.7196, + "loss": 1.6115, "step": 16230 }, { "epoch": 1.72, - "grad_norm": 12.36194896697998, + "grad_norm": 50.178775787353516, "learning_rate": 1.656985954166227e-05, - "loss": 0.8436, + "loss": 1.5252, "step": 16240 }, { "epoch": 1.72, - "grad_norm": 3.7607321739196777, + "grad_norm": 50.582191467285156, "learning_rate": 1.6567747386207626e-05, - "loss": 0.9466, + "loss": 1.5705, "step": 16250 }, { "epoch": 1.72, - "grad_norm": 26.647628784179688, + "grad_norm": 44.152503967285156, "learning_rate": 1.6565635230752986e-05, - "loss": 0.8529, + "loss": 1.575, "step": 16260 }, { "epoch": 1.72, - "grad_norm": 15.69359016418457, + "grad_norm": 43.82858657836914, "learning_rate": 1.6563523075298343e-05, - "loss": 0.6814, + "loss": 1.3039, "step": 16270 }, { "epoch": 1.72, - "grad_norm": 6.070733547210693, + "grad_norm": 40.2099494934082, "learning_rate": 1.65614109198437e-05, - "loss": 0.6082, + "loss": 1.4909, "step": 16280 }, { "epoch": 1.72, - "grad_norm": 2.8573226928710938, + "grad_norm": 24.592205047607422, "learning_rate": 1.655929876438906e-05, - "loss": 0.8981, + "loss": 1.5164, "step": 16290 }, { "epoch": 1.72, - "grad_norm": 7.599496841430664, + "grad_norm": 40.48691177368164, "learning_rate": 1.6557186608934418e-05, - "loss": 0.7566, + "loss": 1.5195, "step": 16300 }, { "epoch": 1.72, - "grad_norm": 8.828875541687012, + "grad_norm": 39.8051872253418, "learning_rate": 1.655507445347978e-05, - "loss": 0.854, + "loss": 1.4807, "step": 16310 }, { "epoch": 1.72, - "grad_norm": 16.495113372802734, + "grad_norm": 35.85478973388672, "learning_rate": 1.6552962298025136e-05, - "loss": 0.4018, + "loss": 1.45, "step": 16320 }, { "epoch": 1.72, - "grad_norm": 17.484996795654297, + "grad_norm": 35.73685073852539, "learning_rate": 1.6550850142570493e-05, - "loss": 0.5189, + "loss": 1.3242, "step": 16330 }, { "epoch": 1.73, - "grad_norm": 33.639312744140625, + "grad_norm": 58.44788360595703, "learning_rate": 1.6548737987115853e-05, - "loss": 0.7212, + "loss": 1.5212, "step": 16340 }, { "epoch": 1.73, - "grad_norm": 18.288280487060547, + "grad_norm": 28.6331729888916, "learning_rate": 1.654662583166121e-05, - "loss": 1.024, + "loss": 1.4566, "step": 16350 }, { "epoch": 1.73, - "grad_norm": 35.70732498168945, + "grad_norm": 25.92508316040039, "learning_rate": 1.654451367620657e-05, - "loss": 0.7594, + "loss": 1.5098, "step": 16360 }, { "epoch": 1.73, - "grad_norm": 17.323657989501953, + "grad_norm": 40.35456848144531, "learning_rate": 1.6542401520751928e-05, - "loss": 0.5177, + "loss": 1.2567, "step": 16370 }, { "epoch": 1.73, - "grad_norm": 25.762985229492188, + "grad_norm": 26.025392532348633, "learning_rate": 1.6540289365297285e-05, - "loss": 0.8106, + "loss": 1.393, "step": 16380 }, { "epoch": 1.73, - "grad_norm": 6.512430667877197, + "grad_norm": 28.428667068481445, "learning_rate": 1.6538177209842645e-05, - "loss": 0.6623, + "loss": 1.5525, "step": 16390 }, { "epoch": 1.73, - "grad_norm": 14.185491561889648, + "grad_norm": 35.167755126953125, "learning_rate": 1.6536065054388003e-05, - "loss": 0.794, + "loss": 1.445, "step": 16400 }, { "epoch": 1.73, - "grad_norm": 24.729774475097656, + "grad_norm": 36.91502380371094, "learning_rate": 1.6533952898933363e-05, - "loss": 0.7007, + "loss": 1.4766, "step": 16410 }, { "epoch": 1.73, - "grad_norm": 39.28340530395508, + "grad_norm": 24.926624298095703, "learning_rate": 1.6531840743478723e-05, - "loss": 0.8999, + "loss": 1.5019, "step": 16420 }, { "epoch": 1.74, - "grad_norm": 12.931817054748535, + "grad_norm": 55.4326286315918, "learning_rate": 1.652972858802408e-05, - "loss": 0.759, + "loss": 1.4993, "step": 16430 }, { "epoch": 1.74, - "grad_norm": 9.298349380493164, + "grad_norm": 41.459877014160156, "learning_rate": 1.6527616432569438e-05, - "loss": 0.9328, + "loss": 1.5303, "step": 16440 }, { "epoch": 1.74, - "grad_norm": 11.054045677185059, + "grad_norm": 25.59182357788086, "learning_rate": 1.6525504277114798e-05, - "loss": 0.7465, + "loss": 1.0813, "step": 16450 }, { "epoch": 1.74, - "grad_norm": 20.14855194091797, + "grad_norm": 43.805843353271484, "learning_rate": 1.6523392121660155e-05, - "loss": 1.3357, + "loss": 1.675, "step": 16460 }, { "epoch": 1.74, - "grad_norm": 11.001468658447266, + "grad_norm": 43.366512298583984, "learning_rate": 1.6521279966205516e-05, - "loss": 0.5313, + "loss": 1.5567, "step": 16470 }, { "epoch": 1.74, - "grad_norm": 9.967066764831543, + "grad_norm": 32.34950637817383, "learning_rate": 1.6519167810750873e-05, - "loss": 0.7652, + "loss": 1.7282, "step": 16480 }, { "epoch": 1.74, - "grad_norm": 1.5760014057159424, + "grad_norm": 42.789608001708984, "learning_rate": 1.651705565529623e-05, - "loss": 0.6176, + "loss": 1.6332, "step": 16490 }, { "epoch": 1.74, - "grad_norm": 8.29334831237793, + "grad_norm": 23.650602340698242, "learning_rate": 1.651494349984159e-05, - "loss": 0.7196, + "loss": 1.2763, "step": 16500 }, { "epoch": 1.74, - "grad_norm": 29.419174194335938, + "grad_norm": 40.51198196411133, "learning_rate": 1.6512831344386947e-05, - "loss": 0.9838, + "loss": 1.5127, "step": 16510 }, { "epoch": 1.74, - "grad_norm": 32.249786376953125, + "grad_norm": 41.23781204223633, "learning_rate": 1.6510719188932308e-05, - "loss": 0.5742, + "loss": 1.5247, "step": 16520 }, { "epoch": 1.75, - "grad_norm": 0.4679573178291321, + "grad_norm": 23.93157386779785, "learning_rate": 1.6508607033477665e-05, - "loss": 0.6652, + "loss": 1.2383, "step": 16530 }, { "epoch": 1.75, - "grad_norm": 12.367975234985352, + "grad_norm": 49.98651123046875, "learning_rate": 1.6506494878023022e-05, - "loss": 0.9109, + "loss": 1.6706, "step": 16540 }, { "epoch": 1.75, - "grad_norm": 19.048782348632812, + "grad_norm": 36.44184494018555, "learning_rate": 1.6504382722568383e-05, - "loss": 0.9312, + "loss": 1.2031, "step": 16550 }, { "epoch": 1.75, - "grad_norm": 4.405359268188477, + "grad_norm": 20.356521606445312, "learning_rate": 1.650227056711374e-05, - "loss": 0.7518, + "loss": 1.7742, "step": 16560 }, { "epoch": 1.75, - "grad_norm": 44.34562301635742, + "grad_norm": 37.3394889831543, "learning_rate": 1.65001584116591e-05, - "loss": 1.0457, + "loss": 1.7585, "step": 16570 }, { "epoch": 1.75, - "grad_norm": 11.00039005279541, + "grad_norm": 37.586177825927734, "learning_rate": 1.6498046256204457e-05, - "loss": 0.5543, + "loss": 1.3219, "step": 16580 }, { "epoch": 1.75, - "grad_norm": 10.043527603149414, + "grad_norm": 42.75469970703125, "learning_rate": 1.6495934100749818e-05, - "loss": 0.8293, + "loss": 1.6014, "step": 16590 }, { "epoch": 1.75, - "grad_norm": 17.86119842529297, + "grad_norm": 20.702787399291992, "learning_rate": 1.6493821945295175e-05, - "loss": 0.5652, + "loss": 1.3312, "step": 16600 }, { "epoch": 1.75, - "grad_norm": 1.9222668409347534, + "grad_norm": 33.965126037597656, "learning_rate": 1.6491709789840532e-05, - "loss": 0.7655, + "loss": 1.6233, "step": 16610 }, { "epoch": 1.76, - "grad_norm": 11.366325378417969, + "grad_norm": 39.31795120239258, "learning_rate": 1.6489597634385892e-05, - "loss": 0.5567, + "loss": 1.1721, "step": 16620 }, { "epoch": 1.76, - "grad_norm": 2.658869981765747, + "grad_norm": 41.492454528808594, "learning_rate": 1.6487485478931253e-05, - "loss": 0.6991, + "loss": 1.5446, "step": 16630 }, { "epoch": 1.76, - "grad_norm": 0.6182997822761536, + "grad_norm": 25.701316833496094, "learning_rate": 1.648537332347661e-05, - "loss": 0.6625, + "loss": 1.3965, "step": 16640 }, { "epoch": 1.76, - "grad_norm": 16.40772819519043, + "grad_norm": 46.12345886230469, "learning_rate": 1.6483261168021967e-05, - "loss": 0.6434, + "loss": 1.4784, "step": 16650 }, { "epoch": 1.76, - "grad_norm": 10.325271606445312, + "grad_norm": 8.302391052246094, "learning_rate": 1.6481149012567324e-05, - "loss": 0.8875, + "loss": 1.4067, "step": 16660 }, { "epoch": 1.76, - "grad_norm": 22.084362030029297, + "grad_norm": 35.587188720703125, "learning_rate": 1.6479036857112684e-05, - "loss": 0.8016, + "loss": 1.5565, "step": 16670 }, { "epoch": 1.76, - "grad_norm": 24.312217712402344, + "grad_norm": 43.21440887451172, "learning_rate": 1.6476924701658045e-05, - "loss": 0.9715, + "loss": 1.4931, "step": 16680 }, { "epoch": 1.76, - "grad_norm": 5.792603015899658, + "grad_norm": 35.706512451171875, "learning_rate": 1.6474812546203402e-05, - "loss": 0.6115, + "loss": 1.6976, "step": 16690 }, { "epoch": 1.76, - "grad_norm": 10.998063087463379, + "grad_norm": 46.984649658203125, "learning_rate": 1.6472700390748763e-05, - "loss": 0.66, + "loss": 1.6055, "step": 16700 }, { "epoch": 1.76, - "grad_norm": 7.060205459594727, + "grad_norm": 40.40214538574219, "learning_rate": 1.647058823529412e-05, - "loss": 0.8456, + "loss": 1.6849, "step": 16710 }, { "epoch": 1.77, - "grad_norm": 5.907840728759766, + "grad_norm": 39.06698989868164, "learning_rate": 1.6468476079839477e-05, - "loss": 0.798, + "loss": 1.7079, "step": 16720 }, { "epoch": 1.77, - "grad_norm": 12.751337051391602, + "grad_norm": 34.74582290649414, "learning_rate": 1.6466363924384837e-05, - "loss": 0.9007, + "loss": 1.5125, "step": 16730 }, { "epoch": 1.77, - "grad_norm": 15.826787948608398, + "grad_norm": 32.063072204589844, "learning_rate": 1.6464251768930194e-05, - "loss": 0.5821, + "loss": 1.21, "step": 16740 }, { "epoch": 1.77, - "grad_norm": 16.742368698120117, + "grad_norm": 12.468092918395996, "learning_rate": 1.6462139613475555e-05, - "loss": 0.6216, + "loss": 1.5622, "step": 16750 }, { "epoch": 1.77, - "grad_norm": 2.01467227935791, + "grad_norm": 24.096202850341797, "learning_rate": 1.6460027458020912e-05, - "loss": 0.3918, + "loss": 1.1691, "step": 16760 }, { "epoch": 1.77, - "grad_norm": 12.846222877502441, + "grad_norm": 49.17999267578125, "learning_rate": 1.645791530256627e-05, - "loss": 0.8002, + "loss": 1.8075, "step": 16770 }, { "epoch": 1.77, - "grad_norm": 5.770920753479004, + "grad_norm": 28.761512756347656, "learning_rate": 1.645580314711163e-05, - "loss": 0.6514, + "loss": 1.3663, "step": 16780 }, { "epoch": 1.77, - "grad_norm": 0.4666767120361328, + "grad_norm": 26.521669387817383, "learning_rate": 1.6453690991656986e-05, - "loss": 0.5926, + "loss": 1.4386, "step": 16790 }, { "epoch": 1.77, - "grad_norm": 8.078527450561523, + "grad_norm": 43.85164260864258, "learning_rate": 1.6451578836202347e-05, - "loss": 0.7209, + "loss": 1.2133, "step": 16800 }, { "epoch": 1.78, - "grad_norm": 18.627084732055664, + "grad_norm": 38.25829315185547, "learning_rate": 1.6449466680747704e-05, - "loss": 0.6707, + "loss": 1.5973, "step": 16810 }, { "epoch": 1.78, - "grad_norm": 13.834680557250977, + "grad_norm": 25.834129333496094, "learning_rate": 1.644735452529306e-05, - "loss": 0.7757, + "loss": 1.3065, "step": 16820 }, { "epoch": 1.78, - "grad_norm": 5.381798267364502, + "grad_norm": 34.820037841796875, "learning_rate": 1.644524236983842e-05, - "loss": 0.5279, + "loss": 1.5769, "step": 16830 }, { "epoch": 1.78, - "grad_norm": 11.367650985717773, + "grad_norm": 38.783599853515625, "learning_rate": 1.644313021438378e-05, - "loss": 0.8046, + "loss": 1.78, "step": 16840 }, { "epoch": 1.78, - "grad_norm": 10.594490051269531, + "grad_norm": 46.90264892578125, "learning_rate": 1.644101805892914e-05, - "loss": 0.5108, + "loss": 1.498, "step": 16850 }, { "epoch": 1.78, - "grad_norm": 21.720705032348633, + "grad_norm": 33.489017486572266, "learning_rate": 1.64389059034745e-05, - "loss": 0.6067, + "loss": 1.366, "step": 16860 }, { "epoch": 1.78, - "grad_norm": 9.891345024108887, + "grad_norm": 37.521385192871094, "learning_rate": 1.6436793748019853e-05, - "loss": 0.8123, + "loss": 1.5053, "step": 16870 }, { "epoch": 1.78, - "grad_norm": 2.3176815509796143, + "grad_norm": 41.007991790771484, "learning_rate": 1.6434681592565214e-05, - "loss": 0.9926, + "loss": 1.837, "step": 16880 }, { "epoch": 1.78, - "grad_norm": 6.299122333526611, + "grad_norm": 25.98033905029297, "learning_rate": 1.6432569437110574e-05, - "loss": 0.9741, + "loss": 1.5674, "step": 16890 }, { "epoch": 1.78, - "grad_norm": 17.60826873779297, + "grad_norm": 16.77277374267578, "learning_rate": 1.643045728165593e-05, - "loss": 0.6079, + "loss": 1.5452, "step": 16900 }, { "epoch": 1.79, - "grad_norm": 3.8228986263275146, + "grad_norm": 27.485109329223633, "learning_rate": 1.6428345126201292e-05, - "loss": 0.9454, + "loss": 1.5135, "step": 16910 }, { "epoch": 1.79, - "grad_norm": 16.208415985107422, + "grad_norm": 48.02508544921875, "learning_rate": 1.642623297074665e-05, - "loss": 1.1153, + "loss": 1.5978, "step": 16920 }, { "epoch": 1.79, - "grad_norm": 9.8814058303833, + "grad_norm": 36.64805221557617, "learning_rate": 1.6424120815292006e-05, - "loss": 0.9222, + "loss": 1.6859, "step": 16930 }, { "epoch": 1.79, - "grad_norm": 11.066062927246094, + "grad_norm": 32.9328727722168, "learning_rate": 1.6422008659837366e-05, - "loss": 0.8315, + "loss": 1.3143, "step": 16940 }, { "epoch": 1.79, - "grad_norm": 6.591475009918213, + "grad_norm": 31.676210403442383, "learning_rate": 1.6419896504382724e-05, - "loss": 0.9139, + "loss": 1.3671, "step": 16950 }, { "epoch": 1.79, - "grad_norm": 5.541287422180176, + "grad_norm": 39.33505630493164, "learning_rate": 1.6417784348928084e-05, - "loss": 0.4985, + "loss": 1.2809, "step": 16960 }, { "epoch": 1.79, - "grad_norm": 13.867006301879883, + "grad_norm": 32.333740234375, "learning_rate": 1.641567219347344e-05, - "loss": 0.7892, + "loss": 1.4694, "step": 16970 }, { "epoch": 1.79, - "grad_norm": 9.305230140686035, + "grad_norm": 25.166128158569336, "learning_rate": 1.6413560038018798e-05, - "loss": 0.7857, + "loss": 1.3521, "step": 16980 }, { "epoch": 1.79, - "grad_norm": 14.229780197143555, + "grad_norm": 34.555233001708984, "learning_rate": 1.641144788256416e-05, - "loss": 0.5909, + "loss": 1.3167, "step": 16990 }, { "epoch": 1.8, - "grad_norm": 22.40638542175293, + "grad_norm": 32.22737503051758, "learning_rate": 1.6409335727109516e-05, - "loss": 0.6501, + "loss": 1.303, "step": 17000 }, { "epoch": 1.8, - "grad_norm": 10.190711975097656, + "grad_norm": 41.18868637084961, "learning_rate": 1.6407223571654876e-05, - "loss": 0.6761, + "loss": 1.3395, "step": 17010 }, { "epoch": 1.8, - "grad_norm": 14.288476943969727, + "grad_norm": 32.04877853393555, "learning_rate": 1.6405111416200233e-05, - "loss": 0.755, + "loss": 1.4181, "step": 17020 }, { "epoch": 1.8, - "grad_norm": 2.0422375202178955, + "grad_norm": 33.26199722290039, "learning_rate": 1.640299926074559e-05, - "loss": 0.5511, + "loss": 1.5137, "step": 17030 }, { "epoch": 1.8, - "grad_norm": 27.12386703491211, + "grad_norm": 51.23695755004883, "learning_rate": 1.640088710529095e-05, - "loss": 0.8079, + "loss": 1.4127, "step": 17040 }, { "epoch": 1.8, - "grad_norm": 22.44472312927246, + "grad_norm": 40.853275299072266, "learning_rate": 1.6398774949836308e-05, - "loss": 0.9826, + "loss": 1.9945, "step": 17050 }, { "epoch": 1.8, - "grad_norm": 3.8263089656829834, + "grad_norm": 16.350011825561523, "learning_rate": 1.639666279438167e-05, - "loss": 0.4825, + "loss": 1.3745, "step": 17060 }, { "epoch": 1.8, - "grad_norm": 11.57314395904541, + "grad_norm": 37.99031448364258, "learning_rate": 1.6394550638927025e-05, - "loss": 0.7184, + "loss": 1.7956, "step": 17070 }, { "epoch": 1.8, - "grad_norm": 8.151266098022461, + "grad_norm": 27.126575469970703, "learning_rate": 1.6392438483472386e-05, - "loss": 0.634, + "loss": 1.3509, "step": 17080 }, { "epoch": 1.8, - "grad_norm": 8.741938591003418, + "grad_norm": 38.5008430480957, "learning_rate": 1.6390326328017743e-05, - "loss": 1.0032, + "loss": 1.7895, "step": 17090 }, { "epoch": 1.81, - "grad_norm": 10.313313484191895, + "grad_norm": 55.48377227783203, "learning_rate": 1.63882141725631e-05, - "loss": 0.9045, + "loss": 1.5367, "step": 17100 }, { "epoch": 1.81, - "grad_norm": 33.50632095336914, + "grad_norm": 52.09832763671875, "learning_rate": 1.638610201710846e-05, - "loss": 0.5271, + "loss": 1.2585, "step": 17110 }, { "epoch": 1.81, - "grad_norm": 18.364887237548828, + "grad_norm": 37.18764114379883, "learning_rate": 1.638398986165382e-05, - "loss": 0.7754, + "loss": 1.4117, "step": 17120 }, { "epoch": 1.81, - "grad_norm": 7.409256935119629, + "grad_norm": 42.358360290527344, "learning_rate": 1.6381877706199178e-05, - "loss": 0.9952, + "loss": 1.8002, "step": 17130 }, { "epoch": 1.81, - "grad_norm": 19.66192054748535, + "grad_norm": 40.013484954833984, "learning_rate": 1.6379765550744535e-05, - "loss": 0.8254, + "loss": 1.3688, "step": 17140 }, { "epoch": 1.81, - "grad_norm": 12.776424407958984, + "grad_norm": 27.198802947998047, "learning_rate": 1.6377653395289896e-05, - "loss": 0.6211, + "loss": 1.3411, "step": 17150 }, { "epoch": 1.81, - "grad_norm": 6.306686878204346, + "grad_norm": 28.618165969848633, "learning_rate": 1.6375541239835253e-05, - "loss": 0.7537, + "loss": 1.5485, "step": 17160 }, { "epoch": 1.81, - "grad_norm": 8.199552536010742, + "grad_norm": 33.412017822265625, "learning_rate": 1.6373429084380613e-05, - "loss": 0.7978, + "loss": 1.7261, "step": 17170 }, { "epoch": 1.81, - "grad_norm": 5.6501617431640625, + "grad_norm": 45.299861907958984, "learning_rate": 1.637131692892597e-05, - "loss": 0.5753, + "loss": 1.5137, "step": 17180 }, { "epoch": 1.82, - "grad_norm": 11.409503936767578, + "grad_norm": 38.60923767089844, "learning_rate": 1.6369204773471327e-05, - "loss": 1.0622, + "loss": 1.5109, "step": 17190 }, { "epoch": 1.82, - "grad_norm": 5.076910018920898, + "grad_norm": 35.24686813354492, "learning_rate": 1.6367092618016688e-05, - "loss": 1.0274, + "loss": 1.7836, "step": 17200 }, { "epoch": 1.82, - "grad_norm": 12.447973251342773, + "grad_norm": 35.744266510009766, "learning_rate": 1.6364980462562045e-05, - "loss": 0.8953, + "loss": 1.3772, "step": 17210 }, { "epoch": 1.82, - "grad_norm": 24.924978256225586, + "grad_norm": 36.830753326416016, "learning_rate": 1.6362868307107405e-05, - "loss": 0.7374, + "loss": 1.4398, "step": 17220 }, { "epoch": 1.82, - "grad_norm": 7.464922904968262, + "grad_norm": 27.087427139282227, "learning_rate": 1.6360756151652763e-05, - "loss": 0.7615, + "loss": 1.3708, "step": 17230 }, { "epoch": 1.82, - "grad_norm": 9.509597778320312, + "grad_norm": 32.49143981933594, "learning_rate": 1.6358643996198123e-05, - "loss": 0.9592, + "loss": 1.5204, "step": 17240 }, { "epoch": 1.82, - "grad_norm": 2.9465973377227783, + "grad_norm": 33.0968017578125, "learning_rate": 1.635653184074348e-05, - "loss": 0.4735, + "loss": 1.2724, "step": 17250 }, { "epoch": 1.82, - "grad_norm": 27.843814849853516, + "grad_norm": 49.263397216796875, "learning_rate": 1.6354419685288837e-05, - "loss": 0.998, + "loss": 1.7775, "step": 17260 }, { "epoch": 1.82, - "grad_norm": 8.39358139038086, + "grad_norm": 23.568788528442383, "learning_rate": 1.6352307529834198e-05, - "loss": 0.4969, + "loss": 1.1805, "step": 17270 }, { "epoch": 1.82, - "grad_norm": 8.511701583862305, + "grad_norm": 32.882293701171875, "learning_rate": 1.6350195374379555e-05, - "loss": 0.6285, + "loss": 1.2236, "step": 17280 }, { "epoch": 1.83, - "grad_norm": 12.4453763961792, + "grad_norm": 35.64259719848633, "learning_rate": 1.6348083218924915e-05, - "loss": 0.6012, + "loss": 1.4526, "step": 17290 }, { "epoch": 1.83, - "grad_norm": 13.351466178894043, + "grad_norm": 43.061866760253906, "learning_rate": 1.6345971063470272e-05, - "loss": 1.0261, + "loss": 1.5731, "step": 17300 }, { "epoch": 1.83, - "grad_norm": 2.012593984603882, + "grad_norm": 26.462343215942383, "learning_rate": 1.634385890801563e-05, - "loss": 0.9434, + "loss": 1.5955, "step": 17310 }, { "epoch": 1.83, - "grad_norm": 15.275026321411133, + "grad_norm": 44.6540641784668, "learning_rate": 1.634174675256099e-05, - "loss": 0.8188, + "loss": 1.5974, "step": 17320 }, { "epoch": 1.83, - "grad_norm": 14.730469703674316, + "grad_norm": 20.809795379638672, "learning_rate": 1.6339634597106347e-05, - "loss": 0.5648, + "loss": 1.3506, "step": 17330 }, { "epoch": 1.83, - "grad_norm": 0.605424165725708, + "grad_norm": 34.266178131103516, "learning_rate": 1.6337522441651707e-05, - "loss": 0.5884, + "loss": 1.5101, "step": 17340 }, { "epoch": 1.83, - "grad_norm": 17.49559211730957, + "grad_norm": 38.628173828125, "learning_rate": 1.6335410286197068e-05, - "loss": 0.7798, + "loss": 1.7847, "step": 17350 }, { "epoch": 1.83, - "grad_norm": 31.68541145324707, + "grad_norm": 42.12553787231445, "learning_rate": 1.633329813074242e-05, - "loss": 0.6351, + "loss": 1.5022, "step": 17360 }, { "epoch": 1.83, - "grad_norm": 18.222248077392578, + "grad_norm": 41.60708236694336, "learning_rate": 1.6331185975287782e-05, - "loss": 0.622, + "loss": 1.4734, "step": 17370 }, { "epoch": 1.84, - "grad_norm": 25.582706451416016, + "grad_norm": 41.52236557006836, "learning_rate": 1.6329073819833143e-05, - "loss": 0.8067, + "loss": 1.6559, "step": 17380 }, { "epoch": 1.84, - "grad_norm": 30.64195442199707, + "grad_norm": 47.95152282714844, "learning_rate": 1.63269616643785e-05, - "loss": 0.6995, + "loss": 1.5398, "step": 17390 }, { "epoch": 1.84, - "grad_norm": 12.211787223815918, + "grad_norm": 32.819576263427734, "learning_rate": 1.632484950892386e-05, - "loss": 0.7392, + "loss": 1.5947, "step": 17400 }, { "epoch": 1.84, - "grad_norm": 22.42513656616211, + "grad_norm": 43.27428436279297, "learning_rate": 1.6322737353469217e-05, - "loss": 0.9869, + "loss": 1.6751, "step": 17410 }, { "epoch": 1.84, - "grad_norm": 11.69765853881836, + "grad_norm": 37.83638000488281, "learning_rate": 1.6320625198014574e-05, - "loss": 0.6206, + "loss": 1.4555, "step": 17420 }, { "epoch": 1.84, - "grad_norm": 19.749845504760742, + "grad_norm": 35.973854064941406, "learning_rate": 1.6318513042559935e-05, - "loss": 0.6514, + "loss": 1.3315, "step": 17430 }, { "epoch": 1.84, - "grad_norm": 12.117555618286133, + "grad_norm": 38.33342742919922, "learning_rate": 1.6316400887105292e-05, - "loss": 0.9537, + "loss": 1.6613, "step": 17440 }, { "epoch": 1.84, - "grad_norm": 9.224960327148438, + "grad_norm": 42.70329666137695, "learning_rate": 1.6314288731650652e-05, - "loss": 0.9204, + "loss": 1.7449, "step": 17450 }, { "epoch": 1.84, - "grad_norm": 9.829970359802246, + "grad_norm": 35.403446197509766, "learning_rate": 1.631217657619601e-05, - "loss": 0.6848, + "loss": 1.3882, "step": 17460 }, { "epoch": 1.84, - "grad_norm": 6.119236946105957, + "grad_norm": 40.55786895751953, "learning_rate": 1.6310064420741366e-05, - "loss": 0.6983, + "loss": 1.5887, "step": 17470 }, { "epoch": 1.85, - "grad_norm": 7.326155185699463, + "grad_norm": 43.735023498535156, "learning_rate": 1.6307952265286727e-05, - "loss": 0.6793, + "loss": 1.6478, "step": 17480 }, { "epoch": 1.85, - "grad_norm": 12.868435859680176, + "grad_norm": 25.868101119995117, "learning_rate": 1.6305840109832084e-05, - "loss": 0.7411, + "loss": 1.3575, "step": 17490 }, { "epoch": 1.85, - "grad_norm": 15.110376358032227, + "grad_norm": 41.26874542236328, "learning_rate": 1.6303727954377444e-05, - "loss": 0.5194, + "loss": 1.4841, "step": 17500 }, { "epoch": 1.85, - "grad_norm": 10.07983112335205, + "grad_norm": 39.39900588989258, "learning_rate": 1.63016157989228e-05, - "loss": 0.678, + "loss": 1.336, "step": 17510 }, { "epoch": 1.85, - "grad_norm": 23.019123077392578, + "grad_norm": 47.266693115234375, "learning_rate": 1.629950364346816e-05, - "loss": 1.0418, + "loss": 1.7299, "step": 17520 }, { "epoch": 1.85, - "grad_norm": 17.011865615844727, + "grad_norm": 49.29933547973633, "learning_rate": 1.629739148801352e-05, - "loss": 0.7682, + "loss": 1.6973, "step": 17530 }, { "epoch": 1.85, - "grad_norm": 26.467378616333008, + "grad_norm": 30.61956787109375, "learning_rate": 1.6295279332558876e-05, - "loss": 0.8593, + "loss": 1.4801, "step": 17540 }, { "epoch": 1.85, - "grad_norm": 19.255813598632812, + "grad_norm": 32.17523956298828, "learning_rate": 1.6293167177104237e-05, - "loss": 0.7197, + "loss": 1.3164, "step": 17550 }, { "epoch": 1.85, - "grad_norm": 4.893608570098877, + "grad_norm": 26.541587829589844, "learning_rate": 1.6291055021649597e-05, - "loss": 0.7824, + "loss": 1.5106, "step": 17560 }, { "epoch": 1.86, - "grad_norm": 0.5480058193206787, + "grad_norm": 38.636207580566406, "learning_rate": 1.628894286619495e-05, - "loss": 0.9435, + "loss": 1.5574, "step": 17570 }, { "epoch": 1.86, - "grad_norm": 9.179495811462402, + "grad_norm": 37.27361297607422, "learning_rate": 1.628683071074031e-05, - "loss": 0.8723, + "loss": 1.515, "step": 17580 }, { "epoch": 1.86, - "grad_norm": 12.884167671203613, + "grad_norm": 48.96399688720703, "learning_rate": 1.6284718555285672e-05, - "loss": 0.8562, + "loss": 1.2905, "step": 17590 }, { "epoch": 1.86, - "grad_norm": 7.089831352233887, + "grad_norm": 30.23381996154785, "learning_rate": 1.628260639983103e-05, - "loss": 0.5887, + "loss": 1.468, "step": 17600 }, { "epoch": 1.86, - "grad_norm": 15.21693229675293, + "grad_norm": 48.50172805786133, "learning_rate": 1.628049424437639e-05, - "loss": 0.8149, + "loss": 1.5643, "step": 17610 }, { "epoch": 1.86, - "grad_norm": 6.054338455200195, + "grad_norm": 36.94963836669922, "learning_rate": 1.6278382088921746e-05, - "loss": 0.8234, + "loss": 1.33, "step": 17620 }, { "epoch": 1.86, - "grad_norm": 15.133138656616211, + "grad_norm": 20.855506896972656, "learning_rate": 1.6276269933467104e-05, - "loss": 0.7591, + "loss": 1.4713, "step": 17630 }, { "epoch": 1.86, - "grad_norm": 27.58234214782715, + "grad_norm": 39.40144348144531, "learning_rate": 1.6274157778012464e-05, - "loss": 0.6535, + "loss": 1.1965, "step": 17640 }, { "epoch": 1.86, - "grad_norm": 21.186670303344727, + "grad_norm": 35.195701599121094, "learning_rate": 1.627204562255782e-05, - "loss": 0.8118, + "loss": 1.4057, "step": 17650 }, { "epoch": 1.87, - "grad_norm": 13.730652809143066, + "grad_norm": 29.821746826171875, "learning_rate": 1.626993346710318e-05, - "loss": 0.7648, + "loss": 1.4838, "step": 17660 }, { "epoch": 1.87, - "grad_norm": 20.385713577270508, + "grad_norm": 49.231727600097656, "learning_rate": 1.626782131164854e-05, - "loss": 1.1946, + "loss": 1.5322, "step": 17670 }, { "epoch": 1.87, - "grad_norm": 11.335777282714844, + "grad_norm": 27.725189208984375, "learning_rate": 1.6265709156193896e-05, - "loss": 0.9211, + "loss": 1.5083, "step": 17680 }, { "epoch": 1.87, - "grad_norm": 4.789702415466309, + "grad_norm": 36.52976989746094, "learning_rate": 1.6263597000739256e-05, - "loss": 0.4236, + "loss": 1.2717, "step": 17690 }, { "epoch": 1.87, - "grad_norm": 11.2709379196167, + "grad_norm": 49.08247375488281, "learning_rate": 1.6261484845284613e-05, - "loss": 0.6328, + "loss": 1.3833, "step": 17700 }, { "epoch": 1.87, - "grad_norm": 22.261934280395508, + "grad_norm": 32.949520111083984, "learning_rate": 1.6259372689829974e-05, - "loss": 1.0002, + "loss": 1.5454, "step": 17710 }, { "epoch": 1.87, - "grad_norm": 11.626084327697754, + "grad_norm": 35.81278610229492, "learning_rate": 1.625726053437533e-05, - "loss": 0.7859, + "loss": 1.4997, "step": 17720 }, { "epoch": 1.87, - "grad_norm": 5.81031608581543, + "grad_norm": 32.259620666503906, "learning_rate": 1.625514837892069e-05, - "loss": 0.5501, + "loss": 1.4483, "step": 17730 }, { "epoch": 1.87, - "grad_norm": 11.801675796508789, + "grad_norm": 34.19773864746094, "learning_rate": 1.625303622346605e-05, - "loss": 0.8058, + "loss": 1.395, "step": 17740 }, { "epoch": 1.87, - "grad_norm": 7.412685394287109, + "grad_norm": 37.76911544799805, "learning_rate": 1.6250924068011405e-05, - "loss": 0.8603, + "loss": 1.579, "step": 17750 }, { "epoch": 1.88, - "grad_norm": 15.892348289489746, + "grad_norm": 38.44866943359375, "learning_rate": 1.6248811912556766e-05, - "loss": 0.8704, + "loss": 1.3193, "step": 17760 }, { "epoch": 1.88, - "grad_norm": 25.67937660217285, + "grad_norm": 23.430585861206055, "learning_rate": 1.6246699757102123e-05, - "loss": 0.8533, + "loss": 1.395, "step": 17770 }, { "epoch": 1.88, - "grad_norm": 22.533653259277344, + "grad_norm": 48.436683654785156, "learning_rate": 1.6244587601647484e-05, - "loss": 0.8178, + "loss": 1.4097, "step": 17780 }, { "epoch": 1.88, - "grad_norm": 5.732408046722412, + "grad_norm": 38.11260986328125, "learning_rate": 1.624247544619284e-05, - "loss": 0.6895, + "loss": 1.4097, "step": 17790 }, { "epoch": 1.88, - "grad_norm": 10.401698112487793, + "grad_norm": 46.8668212890625, "learning_rate": 1.6240363290738198e-05, - "loss": 1.0488, + "loss": 1.5545, "step": 17800 }, { "epoch": 1.88, - "grad_norm": 11.11428451538086, + "grad_norm": 28.842430114746094, "learning_rate": 1.6238251135283558e-05, - "loss": 0.7477, + "loss": 1.3777, "step": 17810 }, { "epoch": 1.88, - "grad_norm": 17.090866088867188, + "grad_norm": 42.05699157714844, "learning_rate": 1.623613897982892e-05, - "loss": 0.564, + "loss": 1.4949, "step": 17820 }, { "epoch": 1.88, - "grad_norm": 9.087224960327148, + "grad_norm": 51.485572814941406, "learning_rate": 1.6234026824374276e-05, - "loss": 0.6164, + "loss": 1.4352, "step": 17830 }, { "epoch": 1.88, - "grad_norm": 20.073070526123047, + "grad_norm": 37.459869384765625, "learning_rate": 1.6231914668919633e-05, - "loss": 1.0547, + "loss": 1.7222, "step": 17840 }, { "epoch": 1.89, - "grad_norm": 10.269225120544434, + "grad_norm": 32.59825134277344, "learning_rate": 1.6229802513464993e-05, - "loss": 0.6537, + "loss": 1.4616, "step": 17850 }, { "epoch": 1.89, - "grad_norm": 14.89734172821045, + "grad_norm": 24.74863052368164, "learning_rate": 1.622769035801035e-05, - "loss": 1.0166, + "loss": 1.7019, "step": 17860 }, { "epoch": 1.89, - "grad_norm": 4.836582183837891, + "grad_norm": 36.72846221923828, "learning_rate": 1.622557820255571e-05, - "loss": 0.8812, + "loss": 1.4084, "step": 17870 }, { "epoch": 1.89, - "grad_norm": 11.207135200500488, + "grad_norm": 41.688232421875, "learning_rate": 1.6223466047101068e-05, - "loss": 0.7037, + "loss": 1.1079, "step": 17880 }, { "epoch": 1.89, - "grad_norm": 13.8482084274292, + "grad_norm": 43.524261474609375, "learning_rate": 1.622135389164643e-05, - "loss": 0.5597, + "loss": 1.5598, "step": 17890 }, { "epoch": 1.89, - "grad_norm": 8.28287124633789, + "grad_norm": 13.512421607971191, "learning_rate": 1.6219241736191785e-05, - "loss": 0.7201, + "loss": 1.2236, "step": 17900 }, { "epoch": 1.89, - "grad_norm": 14.299306869506836, + "grad_norm": 47.44649124145508, "learning_rate": 1.6217129580737143e-05, - "loss": 1.0874, + "loss": 1.759, "step": 17910 }, { "epoch": 1.89, - "grad_norm": 24.98186683654785, + "grad_norm": 41.03382110595703, "learning_rate": 1.6215017425282503e-05, - "loss": 0.7838, + "loss": 1.4725, "step": 17920 }, { "epoch": 1.89, - "grad_norm": 9.939685821533203, + "grad_norm": 33.61712646484375, "learning_rate": 1.621290526982786e-05, - "loss": 0.6725, + "loss": 1.4891, "step": 17930 }, { "epoch": 1.89, - "grad_norm": 11.763669967651367, + "grad_norm": 38.816444396972656, "learning_rate": 1.621079311437322e-05, - "loss": 0.9285, + "loss": 1.7234, "step": 17940 }, { "epoch": 1.9, - "grad_norm": 23.265291213989258, + "grad_norm": 32.1539421081543, "learning_rate": 1.6208680958918578e-05, - "loss": 1.0249, + "loss": 1.3064, "step": 17950 }, { "epoch": 1.9, - "grad_norm": 15.61098575592041, + "grad_norm": 48.80228042602539, "learning_rate": 1.6206568803463935e-05, - "loss": 0.5651, + "loss": 1.7064, "step": 17960 }, { "epoch": 1.9, - "grad_norm": 7.937928199768066, + "grad_norm": 37.729862213134766, "learning_rate": 1.6204456648009295e-05, - "loss": 0.8681, + "loss": 1.7417, "step": 17970 }, { "epoch": 1.9, - "grad_norm": 18.06601333618164, + "grad_norm": 35.42780303955078, "learning_rate": 1.6202344492554652e-05, - "loss": 0.7262, + "loss": 1.358, "step": 17980 }, { "epoch": 1.9, - "grad_norm": 30.48269271850586, + "grad_norm": 30.158796310424805, "learning_rate": 1.6200232337100013e-05, - "loss": 0.8881, + "loss": 1.6044, "step": 17990 }, { "epoch": 1.9, - "grad_norm": 17.150007247924805, + "grad_norm": 48.33888626098633, "learning_rate": 1.619812018164537e-05, - "loss": 0.637, + "loss": 1.4321, "step": 18000 }, { "epoch": 1.9, - "grad_norm": 21.43360137939453, + "grad_norm": 43.48403549194336, "learning_rate": 1.6196008026190727e-05, - "loss": 0.6901, + "loss": 1.8533, "step": 18010 }, { "epoch": 1.9, - "grad_norm": 23.511985778808594, + "grad_norm": 46.03839874267578, "learning_rate": 1.6193895870736087e-05, - "loss": 0.7301, + "loss": 1.7771, "step": 18020 }, { "epoch": 1.9, - "grad_norm": 3.382720947265625, + "grad_norm": 28.529254913330078, "learning_rate": 1.6191783715281445e-05, - "loss": 0.7017, + "loss": 1.3729, "step": 18030 }, { "epoch": 1.91, - "grad_norm": 5.396388053894043, + "grad_norm": 30.175025939941406, "learning_rate": 1.6189671559826805e-05, - "loss": 0.3949, + "loss": 1.3729, "step": 18040 }, { "epoch": 1.91, - "grad_norm": 14.358920097351074, + "grad_norm": 33.388999938964844, "learning_rate": 1.6187559404372165e-05, - "loss": 0.7527, + "loss": 1.3909, "step": 18050 }, { "epoch": 1.91, - "grad_norm": 16.340845108032227, + "grad_norm": 40.89658737182617, "learning_rate": 1.618544724891752e-05, - "loss": 0.7618, + "loss": 1.4572, "step": 18060 }, { "epoch": 1.91, - "grad_norm": 3.0293359756469727, + "grad_norm": 35.85551452636719, "learning_rate": 1.618333509346288e-05, - "loss": 0.5157, + "loss": 1.4706, "step": 18070 }, { "epoch": 1.91, - "grad_norm": 14.064010620117188, + "grad_norm": 29.91685676574707, "learning_rate": 1.618122293800824e-05, - "loss": 0.5641, + "loss": 1.3282, "step": 18080 }, { "epoch": 1.91, - "grad_norm": 7.737006664276123, + "grad_norm": 27.421031951904297, "learning_rate": 1.6179110782553597e-05, - "loss": 0.5526, + "loss": 1.3947, "step": 18090 }, { "epoch": 1.91, - "grad_norm": 21.089754104614258, + "grad_norm": 46.549381256103516, "learning_rate": 1.6176998627098958e-05, - "loss": 0.7937, + "loss": 1.3894, "step": 18100 }, { "epoch": 1.91, - "grad_norm": 18.204021453857422, + "grad_norm": 41.90758514404297, "learning_rate": 1.6174886471644315e-05, - "loss": 0.7606, + "loss": 1.5252, "step": 18110 }, { "epoch": 1.91, - "grad_norm": 20.816307067871094, + "grad_norm": 41.16197204589844, "learning_rate": 1.6172774316189672e-05, - "loss": 0.7746, + "loss": 1.2395, "step": 18120 }, { "epoch": 1.91, - "grad_norm": 12.395362854003906, + "grad_norm": 47.912559509277344, "learning_rate": 1.6170662160735032e-05, - "loss": 0.7779, + "loss": 1.5912, "step": 18130 }, { "epoch": 1.92, - "grad_norm": 11.64930248260498, + "grad_norm": 37.14018630981445, "learning_rate": 1.616855000528039e-05, - "loss": 0.6477, + "loss": 1.5371, "step": 18140 }, { "epoch": 1.92, - "grad_norm": 4.3558149337768555, + "grad_norm": 41.0162467956543, "learning_rate": 1.616643784982575e-05, - "loss": 0.6476, + "loss": 1.5587, "step": 18150 }, { "epoch": 1.92, - "grad_norm": 16.750497817993164, + "grad_norm": 30.704975128173828, "learning_rate": 1.6164325694371107e-05, - "loss": 0.9733, + "loss": 1.78, "step": 18160 }, { "epoch": 1.92, - "grad_norm": 24.893096923828125, + "grad_norm": 40.7150993347168, "learning_rate": 1.6162213538916464e-05, - "loss": 0.7687, + "loss": 1.8361, "step": 18170 }, { "epoch": 1.92, - "grad_norm": 10.435084342956543, + "grad_norm": 54.46761703491211, "learning_rate": 1.6160101383461825e-05, - "loss": 0.888, + "loss": 1.8925, "step": 18180 }, { "epoch": 1.92, - "grad_norm": 5.595534324645996, + "grad_norm": 39.63401412963867, "learning_rate": 1.615798922800718e-05, - "loss": 0.4906, + "loss": 1.3892, "step": 18190 }, { "epoch": 1.92, - "grad_norm": 6.462090015411377, + "grad_norm": 36.75773620605469, "learning_rate": 1.6155877072552542e-05, - "loss": 0.5636, + "loss": 1.2992, "step": 18200 }, { "epoch": 1.92, - "grad_norm": 5.029178142547607, + "grad_norm": 26.31647300720215, "learning_rate": 1.61537649170979e-05, - "loss": 0.7552, + "loss": 1.2974, "step": 18210 }, { "epoch": 1.92, - "grad_norm": 11.36142635345459, + "grad_norm": 36.810997009277344, "learning_rate": 1.6151652761643256e-05, - "loss": 0.9123, + "loss": 1.41, "step": 18220 }, { "epoch": 1.93, - "grad_norm": 14.44530200958252, + "grad_norm": 45.53000259399414, "learning_rate": 1.6149540606188617e-05, - "loss": 0.6762, + "loss": 1.5384, "step": 18230 }, { "epoch": 1.93, - "grad_norm": 17.993099212646484, + "grad_norm": 46.765262603759766, "learning_rate": 1.6147428450733974e-05, - "loss": 0.7306, + "loss": 1.6167, "step": 18240 }, { "epoch": 1.93, - "grad_norm": 20.69666862487793, + "grad_norm": 43.95051574707031, "learning_rate": 1.6145316295279334e-05, - "loss": 0.6424, + "loss": 1.146, "step": 18250 }, { "epoch": 1.93, - "grad_norm": 4.764277935028076, + "grad_norm": 36.1072883605957, "learning_rate": 1.614320413982469e-05, - "loss": 0.6601, + "loss": 1.4668, "step": 18260 }, { "epoch": 1.93, - "grad_norm": 17.083763122558594, + "grad_norm": 40.081478118896484, "learning_rate": 1.6141091984370052e-05, - "loss": 0.9891, + "loss": 1.4318, "step": 18270 }, { "epoch": 1.93, - "grad_norm": 20.2227725982666, + "grad_norm": 31.36958122253418, "learning_rate": 1.613897982891541e-05, - "loss": 0.5378, + "loss": 1.1345, "step": 18280 }, { "epoch": 1.93, - "grad_norm": 6.832766532897949, + "grad_norm": 33.1617431640625, "learning_rate": 1.6136867673460766e-05, - "loss": 0.864, + "loss": 1.514, "step": 18290 }, { "epoch": 1.93, - "grad_norm": 16.70721435546875, + "grad_norm": 29.906442642211914, "learning_rate": 1.6134755518006126e-05, - "loss": 0.8145, + "loss": 1.4433, "step": 18300 }, { "epoch": 1.93, - "grad_norm": 13.618019104003906, + "grad_norm": 51.3143424987793, "learning_rate": 1.6132643362551487e-05, - "loss": 0.7341, + "loss": 1.2327, "step": 18310 }, { "epoch": 1.93, - "grad_norm": 2.435948371887207, + "grad_norm": 32.2031135559082, "learning_rate": 1.6130531207096844e-05, - "loss": 0.794, + "loss": 1.3147, "step": 18320 }, { "epoch": 1.94, - "grad_norm": 9.983175277709961, + "grad_norm": 35.449012756347656, "learning_rate": 1.61284190516422e-05, - "loss": 0.6721, + "loss": 1.203, "step": 18330 }, { "epoch": 1.94, - "grad_norm": 20.570846557617188, + "grad_norm": 30.5954532623291, "learning_rate": 1.612630689618756e-05, - "loss": 1.0879, + "loss": 1.5375, "step": 18340 }, { "epoch": 1.94, - "grad_norm": 9.342812538146973, + "grad_norm": 56.24901580810547, "learning_rate": 1.612419474073292e-05, - "loss": 0.7654, + "loss": 1.444, "step": 18350 }, { "epoch": 1.94, - "grad_norm": 8.749177932739258, + "grad_norm": 33.26801681518555, "learning_rate": 1.612208258527828e-05, - "loss": 0.6205, + "loss": 1.4315, "step": 18360 }, { "epoch": 1.94, - "grad_norm": 23.80336570739746, + "grad_norm": 41.50624465942383, "learning_rate": 1.6119970429823636e-05, - "loss": 0.7386, + "loss": 1.6215, "step": 18370 }, { "epoch": 1.94, - "grad_norm": 14.0230073928833, + "grad_norm": 31.023834228515625, "learning_rate": 1.6117858274368997e-05, - "loss": 1.1128, + "loss": 1.6872, "step": 18380 }, { "epoch": 1.94, - "grad_norm": 17.835647583007812, + "grad_norm": 58.81855392456055, "learning_rate": 1.6115746118914354e-05, - "loss": 0.7011, + "loss": 1.1521, "step": 18390 }, { "epoch": 1.94, - "grad_norm": 14.240971565246582, + "grad_norm": 41.262454986572266, "learning_rate": 1.611363396345971e-05, - "loss": 0.773, + "loss": 1.3286, "step": 18400 }, { "epoch": 1.94, - "grad_norm": 4.0728020668029785, + "grad_norm": 37.472389221191406, "learning_rate": 1.611152180800507e-05, - "loss": 0.9297, + "loss": 1.5931, "step": 18410 }, { "epoch": 1.95, - "grad_norm": 14.087554931640625, + "grad_norm": 29.748180389404297, "learning_rate": 1.610940965255043e-05, - "loss": 0.8481, + "loss": 1.6229, "step": 18420 }, { "epoch": 1.95, - "grad_norm": 5.576249599456787, + "grad_norm": 40.334632873535156, "learning_rate": 1.610729749709579e-05, - "loss": 0.395, + "loss": 1.2889, "step": 18430 }, { "epoch": 1.95, - "grad_norm": 10.672992706298828, + "grad_norm": 41.732669830322266, "learning_rate": 1.6105185341641146e-05, - "loss": 0.808, + "loss": 1.3347, "step": 18440 }, { "epoch": 1.95, - "grad_norm": 12.838006019592285, + "grad_norm": 38.73309326171875, "learning_rate": 1.6103073186186503e-05, - "loss": 0.5054, + "loss": 1.3946, "step": 18450 }, { "epoch": 1.95, - "grad_norm": 9.07803726196289, + "grad_norm": 13.643328666687012, "learning_rate": 1.6100961030731864e-05, - "loss": 0.6095, + "loss": 1.3005, "step": 18460 }, { "epoch": 1.95, - "grad_norm": 19.085142135620117, + "grad_norm": 40.90454864501953, "learning_rate": 1.609884887527722e-05, - "loss": 0.9492, + "loss": 1.6505, "step": 18470 }, { "epoch": 1.95, - "grad_norm": 28.00160026550293, + "grad_norm": 60.15568923950195, "learning_rate": 1.609673671982258e-05, - "loss": 0.7518, + "loss": 1.5015, "step": 18480 }, { "epoch": 1.95, - "grad_norm": 7.690191268920898, + "grad_norm": 30.094457626342773, "learning_rate": 1.6094624564367938e-05, - "loss": 0.9787, + "loss": 1.4846, "step": 18490 }, { "epoch": 1.95, - "grad_norm": 9.237311363220215, + "grad_norm": 61.52608871459961, "learning_rate": 1.6092512408913295e-05, - "loss": 0.7653, + "loss": 1.4869, "step": 18500 }, { "epoch": 1.95, - "grad_norm": 17.413515090942383, + "grad_norm": 43.83450698852539, "learning_rate": 1.6090400253458656e-05, - "loss": 0.6914, + "loss": 1.4634, "step": 18510 }, { "epoch": 1.96, - "grad_norm": 7.412738800048828, + "grad_norm": 26.818708419799805, "learning_rate": 1.6088288098004016e-05, - "loss": 0.5039, + "loss": 1.2173, "step": 18520 }, { "epoch": 1.96, - "grad_norm": 21.01041030883789, + "grad_norm": 50.99045944213867, "learning_rate": 1.6086175942549373e-05, - "loss": 0.751, + "loss": 1.0819, "step": 18530 }, { "epoch": 1.96, - "grad_norm": 1.0411614179611206, + "grad_norm": 26.56839370727539, "learning_rate": 1.6084063787094734e-05, - "loss": 0.6429, + "loss": 1.6502, "step": 18540 }, { "epoch": 1.96, - "grad_norm": 10.485559463500977, + "grad_norm": 40.2357177734375, "learning_rate": 1.608195163164009e-05, - "loss": 0.9205, + "loss": 1.8528, "step": 18550 }, { "epoch": 1.96, - "grad_norm": 18.306453704833984, + "grad_norm": 40.210777282714844, "learning_rate": 1.6079839476185448e-05, - "loss": 0.8701, + "loss": 1.4553, "step": 18560 }, { "epoch": 1.96, - "grad_norm": 7.664942741394043, + "grad_norm": 33.161190032958984, "learning_rate": 1.607772732073081e-05, - "loss": 0.958, + "loss": 1.3363, "step": 18570 }, { "epoch": 1.96, - "grad_norm": 16.135488510131836, + "grad_norm": 39.938194274902344, "learning_rate": 1.6075615165276165e-05, - "loss": 0.6189, + "loss": 1.2822, "step": 18580 }, { "epoch": 1.96, - "grad_norm": 10.73397445678711, + "grad_norm": 47.99114227294922, "learning_rate": 1.6073503009821526e-05, - "loss": 0.6554, + "loss": 1.6186, "step": 18590 }, { "epoch": 1.96, - "grad_norm": 12.909746170043945, + "grad_norm": 40.096343994140625, "learning_rate": 1.6071390854366883e-05, - "loss": 0.7676, + "loss": 1.706, "step": 18600 }, { "epoch": 1.97, - "grad_norm": 1.4000301361083984, + "grad_norm": 37.01713562011719, "learning_rate": 1.606927869891224e-05, - "loss": 0.6269, + "loss": 1.5641, "step": 18610 }, { "epoch": 1.97, - "grad_norm": 12.796710968017578, + "grad_norm": 30.723918914794922, "learning_rate": 1.60671665434576e-05, - "loss": 0.8095, + "loss": 1.5987, "step": 18620 }, { "epoch": 1.97, - "grad_norm": 10.29948616027832, + "grad_norm": 41.81086349487305, "learning_rate": 1.6065054388002958e-05, - "loss": 1.0271, + "loss": 1.6687, "step": 18630 }, { "epoch": 1.97, - "grad_norm": 24.8094425201416, + "grad_norm": 45.0186767578125, "learning_rate": 1.6062942232548318e-05, - "loss": 0.8256, + "loss": 1.6922, "step": 18640 }, { "epoch": 1.97, - "grad_norm": 20.030855178833008, + "grad_norm": 37.64236068725586, "learning_rate": 1.6060830077093675e-05, - "loss": 0.8658, + "loss": 1.1968, "step": 18650 }, { "epoch": 1.97, - "grad_norm": 16.202350616455078, + "grad_norm": 31.15154266357422, "learning_rate": 1.6058717921639032e-05, - "loss": 0.5641, + "loss": 1.4908, "step": 18660 }, { "epoch": 1.97, - "grad_norm": 6.504457950592041, + "grad_norm": 50.48983383178711, "learning_rate": 1.6056605766184393e-05, - "loss": 0.8739, + "loss": 1.3538, "step": 18670 }, { "epoch": 1.97, - "grad_norm": 7.2836503982543945, + "grad_norm": 31.15228271484375, "learning_rate": 1.605449361072975e-05, - "loss": 0.6521, + "loss": 1.088, "step": 18680 }, { "epoch": 1.97, - "grad_norm": 4.944937705993652, + "grad_norm": 56.214813232421875, "learning_rate": 1.605238145527511e-05, - "loss": 0.8166, + "loss": 1.5736, "step": 18690 }, { "epoch": 1.97, - "grad_norm": 9.276082992553711, + "grad_norm": 43.216888427734375, "learning_rate": 1.6050269299820467e-05, - "loss": 0.8579, + "loss": 1.9391, "step": 18700 }, { "epoch": 1.98, - "grad_norm": 15.804701805114746, + "grad_norm": 54.654457092285156, "learning_rate": 1.6048157144365825e-05, - "loss": 0.4937, + "loss": 1.295, "step": 18710 }, { "epoch": 1.98, - "grad_norm": 27.6137638092041, + "grad_norm": 34.32917022705078, "learning_rate": 1.6046044988911185e-05, - "loss": 0.8126, + "loss": 1.6641, "step": 18720 }, { "epoch": 1.98, - "grad_norm": 0.9717463254928589, + "grad_norm": 19.345232009887695, "learning_rate": 1.6043932833456542e-05, - "loss": 0.5588, + "loss": 1.1478, "step": 18730 }, { "epoch": 1.98, - "grad_norm": 12.725264549255371, + "grad_norm": 45.367279052734375, "learning_rate": 1.6041820678001903e-05, - "loss": 0.7124, + "loss": 1.2794, "step": 18740 }, { "epoch": 1.98, - "grad_norm": 2.4614529609680176, + "grad_norm": 26.902393341064453, "learning_rate": 1.6039708522547263e-05, - "loss": 0.5819, + "loss": 1.2122, "step": 18750 }, { "epoch": 1.98, - "grad_norm": 6.973654270172119, + "grad_norm": 33.37576675415039, "learning_rate": 1.6037596367092617e-05, - "loss": 0.8406, + "loss": 1.409, "step": 18760 }, { "epoch": 1.98, - "grad_norm": 5.533456802368164, + "grad_norm": 38.37232971191406, "learning_rate": 1.6035484211637977e-05, - "loss": 0.5717, + "loss": 1.3022, "step": 18770 }, { "epoch": 1.98, - "grad_norm": 11.511706352233887, + "grad_norm": 42.81926345825195, "learning_rate": 1.6033372056183338e-05, - "loss": 0.9121, + "loss": 1.5169, "step": 18780 }, { "epoch": 1.98, - "grad_norm": 5.5287885665893555, + "grad_norm": 31.50518226623535, "learning_rate": 1.6031259900728695e-05, - "loss": 0.5873, + "loss": 1.4068, "step": 18790 }, { "epoch": 1.99, - "grad_norm": 15.79177474975586, + "grad_norm": 46.96403503417969, "learning_rate": 1.6029147745274055e-05, - "loss": 1.0362, + "loss": 1.8445, "step": 18800 }, { "epoch": 1.99, - "grad_norm": 30.5955753326416, + "grad_norm": 43.17903518676758, "learning_rate": 1.6027035589819412e-05, - "loss": 0.6189, + "loss": 1.5889, "step": 18810 }, { "epoch": 1.99, - "grad_norm": 6.519191741943359, + "grad_norm": 28.136402130126953, "learning_rate": 1.602492343436477e-05, - "loss": 0.7908, + "loss": 1.4468, "step": 18820 }, { "epoch": 1.99, - "grad_norm": 3.073493480682373, + "grad_norm": 39.44926071166992, "learning_rate": 1.602281127891013e-05, - "loss": 0.712, + "loss": 1.3045, "step": 18830 }, { "epoch": 1.99, - "grad_norm": 13.837777137756348, + "grad_norm": 39.2743034362793, "learning_rate": 1.6020699123455487e-05, - "loss": 0.5638, + "loss": 1.53, "step": 18840 }, { "epoch": 1.99, - "grad_norm": 11.624351501464844, + "grad_norm": 49.73870086669922, "learning_rate": 1.6018586968000847e-05, - "loss": 0.7838, + "loss": 1.8415, "step": 18850 }, { "epoch": 1.99, - "grad_norm": 7.508041858673096, + "grad_norm": 21.938997268676758, "learning_rate": 1.6016474812546205e-05, - "loss": 1.1195, + "loss": 1.5702, "step": 18860 }, { "epoch": 1.99, - "grad_norm": 18.578847885131836, + "grad_norm": 25.875896453857422, "learning_rate": 1.601436265709156e-05, - "loss": 0.897, + "loss": 1.4042, "step": 18870 }, { "epoch": 1.99, - "grad_norm": 11.741835594177246, + "grad_norm": 37.83687973022461, "learning_rate": 1.6012250501636922e-05, - "loss": 0.8912, + "loss": 1.3431, "step": 18880 }, { "epoch": 1.99, - "grad_norm": 23.71407699584961, + "grad_norm": 34.913089752197266, "learning_rate": 1.601013834618228e-05, - "loss": 1.0288, + "loss": 2.174, "step": 18890 }, { "epoch": 2.0, - "grad_norm": 12.340962409973145, + "grad_norm": 30.085819244384766, "learning_rate": 1.600802619072764e-05, - "loss": 0.7181, + "loss": 1.39, "step": 18900 }, { "epoch": 2.0, - "grad_norm": 21.339839935302734, + "grad_norm": 35.64947509765625, "learning_rate": 1.6005914035272997e-05, - "loss": 0.7832, + "loss": 1.6577, "step": 18910 }, { "epoch": 2.0, - "grad_norm": 11.278648376464844, + "grad_norm": 37.957275390625, "learning_rate": 1.6003801879818357e-05, - "loss": 0.6233, + "loss": 1.3907, "step": 18920 }, { "epoch": 2.0, - "grad_norm": 27.052751541137695, + "grad_norm": 42.50807189941406, "learning_rate": 1.6001689724363714e-05, - "loss": 0.617, + "loss": 1.5581, "step": 18930 }, { "epoch": 2.0, - "eval_accuracy": 0.8801584158415842, - "eval_loss": 0.4546138346195221, - "eval_runtime": 169.1633, - "eval_samples_per_second": 149.264, - "eval_steps_per_second": 18.662, + "eval_accuracy": 0.7144554455445544, + "eval_loss": 1.1164981126785278, + "eval_runtime": 37.3333, + "eval_samples_per_second": 676.339, + "eval_steps_per_second": 84.562, "step": 18938 }, { "epoch": 2.0, - "grad_norm": 22.161972045898438, + "grad_norm": 19.19187355041504, "learning_rate": 1.599957756890907e-05, - "loss": 0.8906, + "loss": 1.4356, "step": 18940 }, { "epoch": 2.0, - "grad_norm": 10.364137649536133, + "grad_norm": 40.68629837036133, "learning_rate": 1.5997465413454432e-05, - "loss": 0.5738, + "loss": 1.2338, "step": 18950 }, { "epoch": 2.0, - "grad_norm": 12.9293851852417, + "grad_norm": 40.20417404174805, "learning_rate": 1.599535325799979e-05, - "loss": 0.5414, + "loss": 1.2651, "step": 18960 }, { "epoch": 2.0, - "grad_norm": 0.7813870310783386, + "grad_norm": 27.416582107543945, "learning_rate": 1.599324110254515e-05, - "loss": 0.8328, + "loss": 1.5385, "step": 18970 }, { "epoch": 2.0, - "grad_norm": 15.10571575164795, + "grad_norm": 58.784385681152344, "learning_rate": 1.5991128947090506e-05, - "loss": 0.6317, + "loss": 1.4734, "step": 18980 }, { "epoch": 2.01, - "grad_norm": 16.745689392089844, + "grad_norm": 26.956619262695312, "learning_rate": 1.5989016791635864e-05, - "loss": 0.5417, + "loss": 1.2987, "step": 18990 }, { "epoch": 2.01, - "grad_norm": 24.125988006591797, + "grad_norm": 47.41199493408203, "learning_rate": 1.5986904636181224e-05, - "loss": 0.5562, + "loss": 1.3334, "step": 19000 }, { "epoch": 2.01, - "grad_norm": 3.705522298812866, + "grad_norm": 27.886882781982422, "learning_rate": 1.5984792480726585e-05, - "loss": 0.3631, + "loss": 0.9899, "step": 19010 }, { "epoch": 2.01, - "grad_norm": 4.920507431030273, + "grad_norm": 31.7951602935791, "learning_rate": 1.598268032527194e-05, - "loss": 0.7375, + "loss": 1.5146, "step": 19020 }, { "epoch": 2.01, - "grad_norm": 17.707551956176758, + "grad_norm": 43.80080032348633, "learning_rate": 1.5980568169817302e-05, - "loss": 0.6326, + "loss": 1.448, "step": 19030 }, { "epoch": 2.01, - "grad_norm": 16.247507095336914, + "grad_norm": 39.626441955566406, "learning_rate": 1.597845601436266e-05, - "loss": 0.8446, + "loss": 1.6836, "step": 19040 }, { "epoch": 2.01, - "grad_norm": 0.4880840480327606, + "grad_norm": 23.232460021972656, "learning_rate": 1.5976343858908016e-05, - "loss": 0.7141, + "loss": 1.2898, "step": 19050 }, { "epoch": 2.01, - "grad_norm": 18.573440551757812, + "grad_norm": 52.4293327331543, "learning_rate": 1.5974231703453377e-05, - "loss": 0.6971, + "loss": 1.3653, "step": 19060 }, { "epoch": 2.01, - "grad_norm": 1.4338834285736084, + "grad_norm": 22.956466674804688, "learning_rate": 1.5972119547998734e-05, - "loss": 0.6141, + "loss": 1.3858, "step": 19070 }, { "epoch": 2.01, - "grad_norm": 2.568121910095215, + "grad_norm": 43.327144622802734, "learning_rate": 1.5970007392544094e-05, - "loss": 0.4884, + "loss": 1.2965, "step": 19080 }, { "epoch": 2.02, - "grad_norm": 19.110023498535156, + "grad_norm": 28.904613494873047, "learning_rate": 1.596789523708945e-05, - "loss": 0.6676, + "loss": 1.3128, "step": 19090 }, { "epoch": 2.02, - "grad_norm": 14.740888595581055, + "grad_norm": 29.757795333862305, "learning_rate": 1.596578308163481e-05, - "loss": 0.5967, + "loss": 1.2763, "step": 19100 }, { "epoch": 2.02, - "grad_norm": 10.074928283691406, + "grad_norm": 48.19044494628906, "learning_rate": 1.596367092618017e-05, - "loss": 0.7876, + "loss": 1.5611, "step": 19110 }, { "epoch": 2.02, - "grad_norm": 3.761645555496216, + "grad_norm": 43.76167678833008, "learning_rate": 1.5961558770725526e-05, - "loss": 0.5741, + "loss": 1.1767, "step": 19120 }, { "epoch": 2.02, - "grad_norm": 17.10256004333496, + "grad_norm": 37.33721160888672, "learning_rate": 1.5959446615270886e-05, - "loss": 0.6066, + "loss": 1.6249, "step": 19130 }, { "epoch": 2.02, - "grad_norm": 9.58146858215332, + "grad_norm": 26.983949661254883, "learning_rate": 1.5957334459816244e-05, - "loss": 0.7282, + "loss": 1.4488, "step": 19140 }, { "epoch": 2.02, - "grad_norm": 18.41785430908203, + "grad_norm": 35.587318420410156, "learning_rate": 1.59552223043616e-05, - "loss": 0.5434, + "loss": 1.2853, "step": 19150 }, { "epoch": 2.02, - "grad_norm": 5.852192401885986, + "grad_norm": 24.216758728027344, "learning_rate": 1.595311014890696e-05, - "loss": 0.66, + "loss": 1.9089, "step": 19160 }, { "epoch": 2.02, - "grad_norm": 22.06049919128418, + "grad_norm": 25.47186279296875, "learning_rate": 1.5950997993452318e-05, - "loss": 0.5735, + "loss": 1.2263, "step": 19170 }, { "epoch": 2.03, - "grad_norm": 8.240063667297363, + "grad_norm": 28.464916229248047, "learning_rate": 1.594888583799768e-05, - "loss": 0.5202, + "loss": 1.4529, "step": 19180 }, { "epoch": 2.03, - "grad_norm": 6.28551721572876, + "grad_norm": 59.1042366027832, "learning_rate": 1.594677368254304e-05, - "loss": 0.5761, + "loss": 1.4247, "step": 19190 }, { "epoch": 2.03, - "grad_norm": 9.9398832321167, + "grad_norm": 39.801326751708984, "learning_rate": 1.5944661527088393e-05, - "loss": 0.5131, + "loss": 1.0393, "step": 19200 }, { "epoch": 2.03, - "grad_norm": 19.288440704345703, + "grad_norm": 28.38213539123535, "learning_rate": 1.5942549371633753e-05, - "loss": 0.6903, + "loss": 1.5032, "step": 19210 }, { "epoch": 2.03, - "grad_norm": 16.463420867919922, + "grad_norm": 61.22842025756836, "learning_rate": 1.5940437216179114e-05, - "loss": 0.7795, + "loss": 1.6871, "step": 19220 }, { "epoch": 2.03, - "grad_norm": 10.435492515563965, + "grad_norm": 24.78988265991211, "learning_rate": 1.593832506072447e-05, - "loss": 0.7559, + "loss": 1.4911, "step": 19230 }, { "epoch": 2.03, - "grad_norm": 9.52122974395752, + "grad_norm": 38.941097259521484, "learning_rate": 1.593621290526983e-05, - "loss": 0.7582, + "loss": 1.3857, "step": 19240 }, { "epoch": 2.03, - "grad_norm": 34.2132568359375, + "grad_norm": 38.11163330078125, "learning_rate": 1.5934100749815185e-05, - "loss": 0.7045, + "loss": 1.5207, "step": 19250 }, { "epoch": 2.03, - "grad_norm": 11.148307800292969, + "grad_norm": 33.02006149291992, "learning_rate": 1.5931988594360546e-05, - "loss": 0.6387, + "loss": 1.4272, "step": 19260 }, { "epoch": 2.04, - "grad_norm": 7.384551525115967, + "grad_norm": 29.325538635253906, "learning_rate": 1.5929876438905906e-05, - "loss": 0.6783, + "loss": 1.1993, "step": 19270 }, { "epoch": 2.04, - "grad_norm": 13.132396697998047, + "grad_norm": 47.79296875, "learning_rate": 1.5927764283451263e-05, - "loss": 0.4804, + "loss": 1.3303, "step": 19280 }, { "epoch": 2.04, - "grad_norm": 9.740508079528809, + "grad_norm": 26.862945556640625, "learning_rate": 1.5925652127996624e-05, - "loss": 0.5604, + "loss": 1.3059, "step": 19290 }, { "epoch": 2.04, - "grad_norm": 11.169816017150879, + "grad_norm": 25.26643180847168, "learning_rate": 1.592353997254198e-05, - "loss": 0.7022, + "loss": 1.2171, "step": 19300 }, { "epoch": 2.04, - "grad_norm": 10.609028816223145, + "grad_norm": 22.397165298461914, "learning_rate": 1.5921427817087338e-05, - "loss": 0.7239, + "loss": 1.268, "step": 19310 }, { "epoch": 2.04, - "grad_norm": 12.647075653076172, + "grad_norm": 28.18300437927246, "learning_rate": 1.5919315661632698e-05, - "loss": 0.8255, + "loss": 1.4121, "step": 19320 }, { "epoch": 2.04, - "grad_norm": 19.971820831298828, + "grad_norm": 35.86153030395508, "learning_rate": 1.5917203506178055e-05, - "loss": 0.777, + "loss": 1.2762, "step": 19330 }, { "epoch": 2.04, - "grad_norm": 21.96154022216797, + "grad_norm": 39.44927978515625, "learning_rate": 1.5915091350723416e-05, - "loss": 0.593, + "loss": 1.2585, "step": 19340 }, { "epoch": 2.04, - "grad_norm": 3.648920774459839, + "grad_norm": 44.560302734375, "learning_rate": 1.5912979195268773e-05, - "loss": 0.618, + "loss": 1.6608, "step": 19350 }, { "epoch": 2.04, - "grad_norm": 11.5049409866333, + "grad_norm": 27.333097457885742, "learning_rate": 1.591086703981413e-05, - "loss": 0.6749, + "loss": 1.38, "step": 19360 }, { "epoch": 2.05, - "grad_norm": 12.83330249786377, + "grad_norm": 33.50857925415039, "learning_rate": 1.590875488435949e-05, - "loss": 0.6487, + "loss": 1.2435, "step": 19370 }, { "epoch": 2.05, - "grad_norm": 25.259958267211914, + "grad_norm": 36.07560348510742, "learning_rate": 1.5906642728904847e-05, - "loss": 0.6379, + "loss": 1.5199, "step": 19380 }, { "epoch": 2.05, - "grad_norm": 4.177110195159912, + "grad_norm": 32.64834213256836, "learning_rate": 1.5904530573450208e-05, - "loss": 0.6603, + "loss": 1.5647, "step": 19390 }, { "epoch": 2.05, - "grad_norm": 14.748346328735352, + "grad_norm": 33.743743896484375, "learning_rate": 1.5902418417995565e-05, - "loss": 0.7137, + "loss": 1.435, "step": 19400 }, { "epoch": 2.05, - "grad_norm": 26.073692321777344, + "grad_norm": 35.66614532470703, "learning_rate": 1.5900306262540922e-05, - "loss": 0.4866, + "loss": 1.535, "step": 19410 }, { "epoch": 2.05, - "grad_norm": 8.271601676940918, + "grad_norm": 36.09792709350586, "learning_rate": 1.5898194107086283e-05, - "loss": 0.6213, + "loss": 1.3881, "step": 19420 }, { "epoch": 2.05, - "grad_norm": 6.245018005371094, + "grad_norm": 42.18521499633789, "learning_rate": 1.589608195163164e-05, - "loss": 0.6066, + "loss": 1.0356, "step": 19430 }, { "epoch": 2.05, - "grad_norm": 26.03018569946289, + "grad_norm": 27.960975646972656, "learning_rate": 1.5893969796177e-05, - "loss": 0.5089, + "loss": 1.3053, "step": 19440 }, { "epoch": 2.05, - "grad_norm": 19.497554779052734, + "grad_norm": 27.938993453979492, "learning_rate": 1.589185764072236e-05, - "loss": 0.7408, + "loss": 1.4406, "step": 19450 }, { "epoch": 2.06, - "grad_norm": 7.597128391265869, + "grad_norm": 25.945558547973633, "learning_rate": 1.5889745485267718e-05, - "loss": 0.6064, + "loss": 1.5219, "step": 19460 }, { "epoch": 2.06, - "grad_norm": 15.027640342712402, + "grad_norm": 48.018531799316406, "learning_rate": 1.5887633329813075e-05, - "loss": 0.5014, + "loss": 1.2711, "step": 19470 }, { "epoch": 2.06, - "grad_norm": 11.24142837524414, + "grad_norm": 34.37765121459961, "learning_rate": 1.5885521174358435e-05, - "loss": 0.7281, + "loss": 1.359, "step": 19480 }, { "epoch": 2.06, - "grad_norm": 5.478426933288574, + "grad_norm": 15.180276870727539, "learning_rate": 1.5883409018903792e-05, - "loss": 0.7077, + "loss": 1.4684, "step": 19490 }, { "epoch": 2.06, - "grad_norm": 16.5303897857666, + "grad_norm": 46.165679931640625, "learning_rate": 1.5881296863449153e-05, - "loss": 0.4631, + "loss": 1.294, "step": 19500 }, { "epoch": 2.06, - "grad_norm": 3.335637092590332, + "grad_norm": 45.066436767578125, "learning_rate": 1.587918470799451e-05, - "loss": 0.7381, + "loss": 1.4744, "step": 19510 }, { "epoch": 2.06, - "grad_norm": 3.8165462017059326, + "grad_norm": 30.4838809967041, "learning_rate": 1.5877072552539867e-05, - "loss": 0.4115, + "loss": 1.5677, "step": 19520 }, { "epoch": 2.06, - "grad_norm": 24.02848243713379, + "grad_norm": 37.29486083984375, "learning_rate": 1.5874960397085227e-05, - "loss": 0.4662, + "loss": 1.2292, "step": 19530 }, { "epoch": 2.06, - "grad_norm": 12.894508361816406, + "grad_norm": 35.927921295166016, "learning_rate": 1.5872848241630585e-05, - "loss": 0.889, + "loss": 1.6217, "step": 19540 }, { "epoch": 2.06, - "grad_norm": 18.948774337768555, + "grad_norm": 47.18776321411133, "learning_rate": 1.5870736086175945e-05, - "loss": 0.9091, + "loss": 1.7324, "step": 19550 }, { "epoch": 2.07, - "grad_norm": 27.89786148071289, + "grad_norm": 41.840641021728516, "learning_rate": 1.5868623930721302e-05, - "loss": 0.6702, + "loss": 1.1909, "step": 19560 }, { "epoch": 2.07, - "grad_norm": 13.14383602142334, + "grad_norm": 55.38502502441406, "learning_rate": 1.5866511775266663e-05, - "loss": 0.487, + "loss": 1.2845, "step": 19570 }, { "epoch": 2.07, - "grad_norm": 13.165166854858398, + "grad_norm": 46.068084716796875, "learning_rate": 1.586439961981202e-05, - "loss": 0.6181, + "loss": 1.6112, "step": 19580 }, { "epoch": 2.07, - "grad_norm": 19.8105411529541, + "grad_norm": 24.402286529541016, "learning_rate": 1.5862287464357377e-05, - "loss": 0.7025, + "loss": 1.5322, "step": 19590 }, { "epoch": 2.07, - "grad_norm": 17.737871170043945, + "grad_norm": 17.632568359375, "learning_rate": 1.5860175308902737e-05, - "loss": 0.8664, + "loss": 1.2331, "step": 19600 }, { "epoch": 2.07, - "grad_norm": 3.290496587753296, + "grad_norm": 38.83909225463867, "learning_rate": 1.5858063153448094e-05, - "loss": 0.6046, + "loss": 1.1355, "step": 19610 }, { "epoch": 2.07, - "grad_norm": 20.9029541015625, + "grad_norm": 45.05975341796875, "learning_rate": 1.5855950997993455e-05, - "loss": 0.6894, + "loss": 1.2069, "step": 19620 }, { "epoch": 2.07, - "grad_norm": 2.6637673377990723, + "grad_norm": 34.41039276123047, "learning_rate": 1.5853838842538812e-05, - "loss": 0.5975, + "loss": 1.0628, "step": 19630 }, { "epoch": 2.07, - "grad_norm": 14.3152437210083, + "grad_norm": 30.55727195739746, "learning_rate": 1.585172668708417e-05, - "loss": 0.6152, + "loss": 1.4973, "step": 19640 }, { "epoch": 2.08, - "grad_norm": 24.497488021850586, + "grad_norm": 25.857532501220703, "learning_rate": 1.584961453162953e-05, - "loss": 0.5984, + "loss": 1.3539, "step": 19650 }, { "epoch": 2.08, - "grad_norm": 14.136295318603516, + "grad_norm": 40.64453887939453, "learning_rate": 1.5847502376174887e-05, - "loss": 0.5843, + "loss": 1.4891, "step": 19660 }, { "epoch": 2.08, - "grad_norm": 4.807919979095459, + "grad_norm": 37.7779541015625, "learning_rate": 1.5845390220720247e-05, - "loss": 0.5101, + "loss": 1.3727, "step": 19670 }, { "epoch": 2.08, - "grad_norm": 11.136591911315918, + "grad_norm": 39.4772834777832, "learning_rate": 1.5843278065265604e-05, - "loss": 0.6622, + "loss": 1.7116, "step": 19680 }, { "epoch": 2.08, - "grad_norm": 19.936723709106445, + "grad_norm": 39.419158935546875, "learning_rate": 1.584116590981096e-05, - "loss": 0.6091, + "loss": 1.5003, "step": 19690 }, { "epoch": 2.08, - "grad_norm": 0.48270049691200256, + "grad_norm": 39.78152084350586, "learning_rate": 1.583905375435632e-05, - "loss": 0.4907, + "loss": 1.4913, "step": 19700 }, { "epoch": 2.08, - "grad_norm": 5.927338123321533, + "grad_norm": 38.70999526977539, "learning_rate": 1.5836941598901682e-05, - "loss": 0.646, + "loss": 1.3346, "step": 19710 }, { "epoch": 2.08, - "grad_norm": 13.167688369750977, + "grad_norm": 48.29328155517578, "learning_rate": 1.583482944344704e-05, - "loss": 0.7181, + "loss": 1.338, "step": 19720 }, { "epoch": 2.08, - "grad_norm": 19.7769775390625, + "grad_norm": 52.33203887939453, "learning_rate": 1.58327172879924e-05, - "loss": 0.7874, + "loss": 1.5091, "step": 19730 }, { "epoch": 2.08, - "grad_norm": 7.564192295074463, + "grad_norm": 29.12432098388672, "learning_rate": 1.5830605132537757e-05, - "loss": 0.7412, + "loss": 1.5424, "step": 19740 }, { "epoch": 2.09, - "grad_norm": 18.283662796020508, + "grad_norm": 49.79229736328125, "learning_rate": 1.5828492977083114e-05, - "loss": 0.6673, + "loss": 1.3756, "step": 19750 }, { "epoch": 2.09, - "grad_norm": 17.03302574157715, + "grad_norm": 34.85970687866211, "learning_rate": 1.5826380821628474e-05, - "loss": 0.7537, + "loss": 1.5331, "step": 19760 }, { "epoch": 2.09, - "grad_norm": 39.64720153808594, + "grad_norm": 29.031274795532227, "learning_rate": 1.582426866617383e-05, - "loss": 0.6179, + "loss": 1.205, "step": 19770 }, { "epoch": 2.09, - "grad_norm": 13.295764923095703, + "grad_norm": 35.03562927246094, "learning_rate": 1.5822156510719192e-05, - "loss": 0.5534, + "loss": 1.3338, "step": 19780 }, { "epoch": 2.09, - "grad_norm": 17.395153045654297, + "grad_norm": 48.81840515136719, "learning_rate": 1.582004435526455e-05, - "loss": 0.7096, + "loss": 1.2945, "step": 19790 }, { "epoch": 2.09, - "grad_norm": 20.03976821899414, + "grad_norm": 23.989459991455078, "learning_rate": 1.5817932199809906e-05, - "loss": 0.6488, + "loss": 1.2616, "step": 19800 }, { "epoch": 2.09, - "grad_norm": 21.574939727783203, + "grad_norm": 36.53584671020508, "learning_rate": 1.5815820044355266e-05, - "loss": 0.7572, + "loss": 1.3731, "step": 19810 }, { "epoch": 2.09, - "grad_norm": 10.643689155578613, + "grad_norm": 42.07261657714844, "learning_rate": 1.5813707888900624e-05, - "loss": 0.7495, + "loss": 1.4513, "step": 19820 }, { "epoch": 2.09, - "grad_norm": 9.975518226623535, + "grad_norm": 48.338523864746094, "learning_rate": 1.5811595733445984e-05, - "loss": 0.4974, + "loss": 1.3643, "step": 19830 }, { "epoch": 2.1, - "grad_norm": 19.891801834106445, + "grad_norm": 30.433664321899414, "learning_rate": 1.580948357799134e-05, - "loss": 0.5566, + "loss": 1.2762, "step": 19840 }, { "epoch": 2.1, - "grad_norm": 10.32098388671875, + "grad_norm": 58.54095458984375, "learning_rate": 1.5807371422536698e-05, - "loss": 0.7832, + "loss": 1.6494, "step": 19850 }, { "epoch": 2.1, - "grad_norm": 21.576885223388672, + "grad_norm": 45.009124755859375, "learning_rate": 1.580525926708206e-05, - "loss": 0.8735, + "loss": 1.3888, "step": 19860 }, { "epoch": 2.1, - "grad_norm": 13.336881637573242, + "grad_norm": 42.28459930419922, "learning_rate": 1.5803147111627416e-05, - "loss": 0.7238, + "loss": 1.276, "step": 19870 }, { "epoch": 2.1, - "grad_norm": 10.624772071838379, + "grad_norm": 37.168216705322266, "learning_rate": 1.5801034956172776e-05, - "loss": 0.8689, + "loss": 1.5276, "step": 19880 }, { "epoch": 2.1, - "grad_norm": 9.150322914123535, + "grad_norm": 39.32699203491211, "learning_rate": 1.5798922800718133e-05, - "loss": 0.7381, + "loss": 1.2617, "step": 19890 }, { "epoch": 2.1, - "grad_norm": 11.372271537780762, + "grad_norm": 49.70573425292969, "learning_rate": 1.579681064526349e-05, - "loss": 0.6086, + "loss": 1.1729, "step": 19900 }, { "epoch": 2.1, - "grad_norm": 2.6746366024017334, + "grad_norm": 29.843212127685547, "learning_rate": 1.579469848980885e-05, - "loss": 0.5083, + "loss": 1.5666, "step": 19910 }, { "epoch": 2.1, - "grad_norm": 19.30531120300293, + "grad_norm": 30.107582092285156, "learning_rate": 1.5792586334354208e-05, - "loss": 0.5636, + "loss": 1.4294, "step": 19920 }, { "epoch": 2.1, - "grad_norm": 3.177243709564209, + "grad_norm": 37.74726104736328, "learning_rate": 1.579047417889957e-05, - "loss": 0.5685, + "loss": 1.6019, "step": 19930 }, { "epoch": 2.11, - "grad_norm": 11.588187217712402, + "grad_norm": 34.4830436706543, "learning_rate": 1.578836202344493e-05, - "loss": 0.6043, + "loss": 1.3602, "step": 19940 }, { "epoch": 2.11, - "grad_norm": 30.2056941986084, + "grad_norm": 36.72800064086914, "learning_rate": 1.5786249867990286e-05, - "loss": 0.5324, + "loss": 1.1838, "step": 19950 }, { "epoch": 2.11, - "grad_norm": 0.5074492692947388, + "grad_norm": 21.4737606048584, "learning_rate": 1.5784137712535643e-05, - "loss": 0.55, + "loss": 1.2932, "step": 19960 }, { "epoch": 2.11, - "grad_norm": 18.67462730407715, + "grad_norm": 34.83669662475586, "learning_rate": 1.5782025557081004e-05, - "loss": 0.6256, + "loss": 1.2943, "step": 19970 }, { "epoch": 2.11, - "grad_norm": 8.119075775146484, + "grad_norm": 47.610164642333984, "learning_rate": 1.577991340162636e-05, - "loss": 0.7572, + "loss": 1.3794, "step": 19980 }, { "epoch": 2.11, - "grad_norm": 13.034306526184082, + "grad_norm": 33.56652069091797, "learning_rate": 1.577780124617172e-05, - "loss": 0.7807, + "loss": 1.0713, "step": 19990 }, { "epoch": 2.11, - "grad_norm": 0.7613555788993835, + "grad_norm": 60.194766998291016, "learning_rate": 1.5775689090717078e-05, - "loss": 0.5044, + "loss": 1.2933, "step": 20000 }, { "epoch": 2.11, - "grad_norm": 9.254698753356934, + "grad_norm": 42.25726318359375, "learning_rate": 1.5773576935262435e-05, - "loss": 0.849, + "loss": 1.7298, "step": 20010 }, { "epoch": 2.11, - "grad_norm": 1.5903544425964355, + "grad_norm": 54.36430740356445, "learning_rate": 1.5771464779807796e-05, - "loss": 0.4597, + "loss": 1.3662, "step": 20020 }, { "epoch": 2.12, - "grad_norm": 18.69160270690918, + "grad_norm": 29.423242568969727, "learning_rate": 1.5769352624353153e-05, - "loss": 0.9927, + "loss": 1.4955, "step": 20030 }, { "epoch": 2.12, - "grad_norm": 20.406883239746094, + "grad_norm": 28.123397827148438, "learning_rate": 1.5767240468898513e-05, - "loss": 0.5909, + "loss": 1.3774, "step": 20040 }, { "epoch": 2.12, - "grad_norm": 23.736196517944336, + "grad_norm": 49.09886932373047, "learning_rate": 1.576512831344387e-05, - "loss": 0.7215, + "loss": 1.5724, "step": 20050 }, { "epoch": 2.12, - "grad_norm": 8.888106346130371, + "grad_norm": 44.33700180053711, "learning_rate": 1.5763016157989227e-05, - "loss": 0.6225, + "loss": 1.2631, "step": 20060 }, { "epoch": 2.12, - "grad_norm": 7.375900745391846, + "grad_norm": 32.71985626220703, "learning_rate": 1.5760904002534588e-05, - "loss": 0.5927, + "loss": 1.3444, "step": 20070 }, { "epoch": 2.12, - "grad_norm": 5.772567272186279, + "grad_norm": 23.347352981567383, "learning_rate": 1.5758791847079945e-05, - "loss": 0.5321, + "loss": 1.1953, "step": 20080 }, { "epoch": 2.12, - "grad_norm": 29.128944396972656, + "grad_norm": 44.9526481628418, "learning_rate": 1.5756679691625306e-05, - "loss": 0.4025, + "loss": 1.2826, "step": 20090 }, { "epoch": 2.12, - "grad_norm": 22.656360626220703, + "grad_norm": 43.651954650878906, "learning_rate": 1.5754567536170663e-05, - "loss": 0.7011, + "loss": 1.5438, "step": 20100 }, { "epoch": 2.12, - "grad_norm": 13.708439826965332, + "grad_norm": 43.87751007080078, "learning_rate": 1.5752455380716023e-05, - "loss": 0.5082, + "loss": 1.2112, "step": 20110 }, { "epoch": 2.12, - "grad_norm": 17.927518844604492, + "grad_norm": 41.07794189453125, "learning_rate": 1.575034322526138e-05, - "loss": 0.6558, + "loss": 1.6623, "step": 20120 }, { "epoch": 2.13, - "grad_norm": 12.460665702819824, + "grad_norm": 28.74135971069336, "learning_rate": 1.5748231069806737e-05, - "loss": 0.5211, + "loss": 1.263, "step": 20130 }, { "epoch": 2.13, - "grad_norm": 13.9175443649292, + "grad_norm": 33.01364517211914, "learning_rate": 1.5746118914352098e-05, - "loss": 0.6519, + "loss": 1.2547, "step": 20140 }, { "epoch": 2.13, - "grad_norm": 4.828811168670654, + "grad_norm": 40.30125427246094, "learning_rate": 1.5744006758897458e-05, - "loss": 0.5153, + "loss": 1.3236, "step": 20150 }, { "epoch": 2.13, - "grad_norm": 17.397619247436523, + "grad_norm": 46.97905731201172, "learning_rate": 1.5741894603442815e-05, - "loss": 0.7972, + "loss": 1.6379, "step": 20160 }, { "epoch": 2.13, - "grad_norm": 7.28850793838501, + "grad_norm": 68.22132873535156, "learning_rate": 1.5739782447988172e-05, - "loss": 0.6057, + "loss": 1.5679, "step": 20170 }, { "epoch": 2.13, - "grad_norm": 28.164974212646484, + "grad_norm": 39.18962097167969, "learning_rate": 1.5737670292533533e-05, - "loss": 0.6714, + "loss": 1.4965, "step": 20180 }, { "epoch": 2.13, - "grad_norm": 7.009616851806641, + "grad_norm": 34.31479263305664, "learning_rate": 1.573555813707889e-05, - "loss": 0.6011, + "loss": 1.3149, "step": 20190 }, { "epoch": 2.13, - "grad_norm": 42.198707580566406, + "grad_norm": 50.129661560058594, "learning_rate": 1.573344598162425e-05, - "loss": 0.7319, + "loss": 1.2909, "step": 20200 }, { "epoch": 2.13, - "grad_norm": 10.300783157348633, + "grad_norm": 43.96068572998047, "learning_rate": 1.5731333826169607e-05, - "loss": 0.6295, + "loss": 1.3548, "step": 20210 }, { "epoch": 2.14, - "grad_norm": 18.88796043395996, + "grad_norm": 57.11779022216797, "learning_rate": 1.5729221670714968e-05, - "loss": 0.9661, + "loss": 1.528, "step": 20220 }, { "epoch": 2.14, - "grad_norm": 12.41268539428711, + "grad_norm": 41.857215881347656, "learning_rate": 1.5727109515260325e-05, - "loss": 0.5005, + "loss": 1.5838, "step": 20230 }, { "epoch": 2.14, - "grad_norm": 16.060094833374023, + "grad_norm": 42.70103073120117, "learning_rate": 1.5724997359805682e-05, - "loss": 0.612, + "loss": 1.2971, "step": 20240 }, { "epoch": 2.14, - "grad_norm": 19.050397872924805, + "grad_norm": 56.83784866333008, "learning_rate": 1.5722885204351043e-05, - "loss": 0.7089, + "loss": 1.3569, "step": 20250 }, { "epoch": 2.14, - "grad_norm": 19.695842742919922, + "grad_norm": 42.50000762939453, "learning_rate": 1.57207730488964e-05, - "loss": 0.6851, + "loss": 1.378, "step": 20260 }, { "epoch": 2.14, - "grad_norm": 14.072586059570312, + "grad_norm": 41.398983001708984, "learning_rate": 1.571866089344176e-05, - "loss": 0.5972, + "loss": 1.5902, "step": 20270 }, { "epoch": 2.14, - "grad_norm": 14.93826961517334, + "grad_norm": 35.50179672241211, "learning_rate": 1.5716548737987117e-05, - "loss": 0.6534, + "loss": 1.5807, "step": 20280 }, { "epoch": 2.14, - "grad_norm": 40.56083679199219, + "grad_norm": 47.62235641479492, "learning_rate": 1.5714436582532474e-05, - "loss": 0.6026, + "loss": 1.3817, "step": 20290 }, { "epoch": 2.14, - "grad_norm": 18.89978790283203, + "grad_norm": 31.575931549072266, "learning_rate": 1.5712324427077835e-05, - "loss": 0.8411, + "loss": 1.3629, "step": 20300 }, { "epoch": 2.14, - "grad_norm": 16.426225662231445, + "grad_norm": 28.286983489990234, "learning_rate": 1.5710212271623192e-05, - "loss": 0.6009, + "loss": 1.3734, "step": 20310 }, { "epoch": 2.15, - "grad_norm": 8.381046295166016, + "grad_norm": 24.900033950805664, "learning_rate": 1.5708100116168552e-05, - "loss": 0.5613, + "loss": 1.6273, "step": 20320 }, { "epoch": 2.15, - "grad_norm": 8.091031074523926, + "grad_norm": 36.217185974121094, "learning_rate": 1.570598796071391e-05, - "loss": 0.5723, + "loss": 1.2101, "step": 20330 }, { "epoch": 2.15, - "grad_norm": 5.171849727630615, + "grad_norm": 38.457359313964844, "learning_rate": 1.5703875805259267e-05, - "loss": 0.3536, + "loss": 0.9663, "step": 20340 }, { "epoch": 2.15, - "grad_norm": 7.740887641906738, + "grad_norm": 39.67443084716797, "learning_rate": 1.5701763649804627e-05, - "loss": 0.9466, + "loss": 1.7453, "step": 20350 }, { "epoch": 2.15, - "grad_norm": 7.167775630950928, + "grad_norm": 43.36320495605469, "learning_rate": 1.5699651494349984e-05, - "loss": 0.6382, + "loss": 1.4247, "step": 20360 }, { "epoch": 2.15, - "grad_norm": 7.635379314422607, + "grad_norm": 35.21637725830078, "learning_rate": 1.5697539338895345e-05, - "loss": 0.5763, + "loss": 1.5493, "step": 20370 }, { "epoch": 2.15, - "grad_norm": 6.6096930503845215, + "grad_norm": 21.040205001831055, "learning_rate": 1.5695427183440705e-05, - "loss": 0.6263, + "loss": 1.1886, "step": 20380 }, { "epoch": 2.15, - "grad_norm": 21.720970153808594, + "grad_norm": 30.603954315185547, "learning_rate": 1.569331502798606e-05, - "loss": 0.6788, + "loss": 1.236, "step": 20390 }, { "epoch": 2.15, - "grad_norm": 16.105064392089844, + "grad_norm": 34.1535530090332, "learning_rate": 1.569120287253142e-05, - "loss": 0.3667, + "loss": 1.3338, "step": 20400 }, { "epoch": 2.16, - "grad_norm": 11.28699779510498, + "grad_norm": 50.933414459228516, "learning_rate": 1.568909071707678e-05, - "loss": 0.8088, + "loss": 1.4032, "step": 20410 }, { "epoch": 2.16, - "grad_norm": 3.666145086288452, + "grad_norm": 45.58281326293945, "learning_rate": 1.5686978561622137e-05, - "loss": 0.5515, + "loss": 1.4371, "step": 20420 }, { "epoch": 2.16, - "grad_norm": 8.528877258300781, + "grad_norm": 33.25333786010742, "learning_rate": 1.5684866406167497e-05, - "loss": 0.9071, + "loss": 1.4848, "step": 20430 }, { "epoch": 2.16, - "grad_norm": 14.195512771606445, + "grad_norm": 34.1765022277832, "learning_rate": 1.5682754250712854e-05, - "loss": 0.6018, + "loss": 1.0459, "step": 20440 }, { "epoch": 2.16, - "grad_norm": 10.027097702026367, + "grad_norm": 42.10125732421875, "learning_rate": 1.568064209525821e-05, - "loss": 1.0243, + "loss": 1.5364, "step": 20450 }, { "epoch": 2.16, - "grad_norm": 16.783695220947266, + "grad_norm": 53.21305847167969, "learning_rate": 1.5678529939803572e-05, - "loss": 0.7825, + "loss": 1.5117, "step": 20460 }, { "epoch": 2.16, - "grad_norm": 5.09663724899292, + "grad_norm": 42.35445785522461, "learning_rate": 1.567641778434893e-05, - "loss": 0.5583, + "loss": 1.4962, "step": 20470 }, { "epoch": 2.16, - "grad_norm": 1.3264994621276855, + "grad_norm": 40.05968475341797, "learning_rate": 1.567430562889429e-05, - "loss": 0.7507, + "loss": 1.1817, "step": 20480 }, { "epoch": 2.16, - "grad_norm": 17.52891731262207, + "grad_norm": 34.11030960083008, "learning_rate": 1.5672193473439647e-05, - "loss": 0.5016, + "loss": 1.4758, "step": 20490 }, { "epoch": 2.16, - "grad_norm": 17.502843856811523, + "grad_norm": 33.01482009887695, "learning_rate": 1.5670081317985004e-05, - "loss": 0.6071, + "loss": 1.6403, "step": 20500 }, { "epoch": 2.17, - "grad_norm": 17.392541885375977, + "grad_norm": 49.88369369506836, "learning_rate": 1.5667969162530364e-05, - "loss": 0.6031, + "loss": 1.5158, "step": 20510 }, { "epoch": 2.17, - "grad_norm": 9.24449634552002, + "grad_norm": 23.966306686401367, "learning_rate": 1.566585700707572e-05, - "loss": 0.6815, + "loss": 1.1234, "step": 20520 }, { "epoch": 2.17, - "grad_norm": 0.5039392113685608, + "grad_norm": 28.183837890625, "learning_rate": 1.566374485162108e-05, - "loss": 0.8172, + "loss": 1.2363, "step": 20530 }, { "epoch": 2.17, - "grad_norm": 9.61462116241455, + "grad_norm": 24.35736083984375, "learning_rate": 1.566163269616644e-05, - "loss": 0.6633, + "loss": 1.5289, "step": 20540 }, { "epoch": 2.17, - "grad_norm": 3.8540539741516113, + "grad_norm": 48.95355224609375, "learning_rate": 1.5659520540711796e-05, - "loss": 0.6843, + "loss": 1.3235, "step": 20550 }, { "epoch": 2.17, - "grad_norm": 8.585871696472168, + "grad_norm": 38.96675109863281, "learning_rate": 1.5657408385257156e-05, - "loss": 0.7397, + "loss": 1.2516, "step": 20560 }, { "epoch": 2.17, - "grad_norm": 19.13616943359375, + "grad_norm": 29.123092651367188, "learning_rate": 1.5655296229802513e-05, - "loss": 0.5932, + "loss": 1.2643, "step": 20570 }, { "epoch": 2.17, - "grad_norm": 25.838733673095703, + "grad_norm": 49.709102630615234, "learning_rate": 1.5653184074347874e-05, - "loss": 0.3896, + "loss": 1.2074, "step": 20580 }, { "epoch": 2.17, - "grad_norm": 18.200727462768555, + "grad_norm": 40.788978576660156, "learning_rate": 1.565107191889323e-05, - "loss": 0.5703, + "loss": 1.3969, "step": 20590 }, { "epoch": 2.18, - "grad_norm": 21.27225112915039, + "grad_norm": 49.342105865478516, "learning_rate": 1.564895976343859e-05, - "loss": 0.4645, + "loss": 1.4606, "step": 20600 }, { "epoch": 2.18, - "grad_norm": 5.026603698730469, + "grad_norm": 29.050996780395508, "learning_rate": 1.564684760798395e-05, - "loss": 0.4996, + "loss": 1.4155, "step": 20610 }, { "epoch": 2.18, - "grad_norm": 15.741429328918457, + "grad_norm": 41.5295524597168, "learning_rate": 1.5644735452529306e-05, - "loss": 0.7345, + "loss": 1.3638, "step": 20620 }, { "epoch": 2.18, - "grad_norm": 10.916876792907715, + "grad_norm": 44.65229797363281, "learning_rate": 1.5642623297074666e-05, - "loss": 0.5682, + "loss": 1.5827, "step": 20630 }, { "epoch": 2.18, - "grad_norm": 15.471184730529785, + "grad_norm": 43.90849304199219, "learning_rate": 1.5640511141620027e-05, - "loss": 0.5366, + "loss": 1.1691, "step": 20640 }, { "epoch": 2.18, - "grad_norm": 16.639137268066406, + "grad_norm": 33.81843566894531, "learning_rate": 1.5638398986165384e-05, - "loss": 0.4488, + "loss": 1.2081, "step": 20650 }, { "epoch": 2.18, - "grad_norm": 17.60270118713379, + "grad_norm": 20.24034881591797, "learning_rate": 1.563628683071074e-05, - "loss": 0.8524, + "loss": 1.3691, "step": 20660 }, { "epoch": 2.18, - "grad_norm": 13.363823890686035, + "grad_norm": 40.450618743896484, "learning_rate": 1.56341746752561e-05, - "loss": 0.5966, + "loss": 1.0514, "step": 20670 }, { "epoch": 2.18, - "grad_norm": 4.700506210327148, + "grad_norm": 36.51478576660156, "learning_rate": 1.5632062519801458e-05, - "loss": 0.3144, + "loss": 1.43, "step": 20680 }, { "epoch": 2.19, - "grad_norm": 5.2145256996154785, + "grad_norm": 39.5124626159668, "learning_rate": 1.562995036434682e-05, - "loss": 0.6223, + "loss": 1.1403, "step": 20690 }, { "epoch": 2.19, - "grad_norm": 13.700287818908691, + "grad_norm": 34.2197151184082, "learning_rate": 1.5627838208892176e-05, - "loss": 0.7161, + "loss": 1.4646, "step": 20700 }, { "epoch": 2.19, - "grad_norm": 28.78150749206543, + "grad_norm": 49.32463455200195, "learning_rate": 1.5625726053437533e-05, - "loss": 0.7957, + "loss": 1.2833, "step": 20710 }, { "epoch": 2.19, - "grad_norm": 17.88661766052246, + "grad_norm": 35.567909240722656, "learning_rate": 1.5623613897982893e-05, - "loss": 0.6605, + "loss": 1.5539, "step": 20720 }, { "epoch": 2.19, - "grad_norm": 20.490943908691406, + "grad_norm": 49.71458053588867, "learning_rate": 1.562150174252825e-05, - "loss": 0.3918, + "loss": 1.2821, "step": 20730 }, { "epoch": 2.19, - "grad_norm": 22.24807357788086, + "grad_norm": 51.425655364990234, "learning_rate": 1.561938958707361e-05, - "loss": 0.4243, + "loss": 1.2153, "step": 20740 }, { "epoch": 2.19, - "grad_norm": 13.95917797088623, + "grad_norm": 40.986793518066406, "learning_rate": 1.5617277431618968e-05, - "loss": 0.5868, + "loss": 1.5689, "step": 20750 }, { "epoch": 2.19, - "grad_norm": 16.390676498413086, + "grad_norm": 36.08663558959961, "learning_rate": 1.561516527616433e-05, - "loss": 0.5448, + "loss": 1.2451, "step": 20760 }, { "epoch": 2.19, - "grad_norm": 1.7195466756820679, + "grad_norm": 17.859045028686523, "learning_rate": 1.5613053120709686e-05, - "loss": 0.5908, + "loss": 1.3143, "step": 20770 }, { "epoch": 2.19, - "grad_norm": 10.160543441772461, + "grad_norm": 45.815670013427734, "learning_rate": 1.5610940965255043e-05, - "loss": 0.7968, + "loss": 1.2592, "step": 20780 }, { "epoch": 2.2, - "grad_norm": 27.573171615600586, + "grad_norm": 38.07283020019531, "learning_rate": 1.5608828809800403e-05, - "loss": 0.6724, + "loss": 1.2402, "step": 20790 }, { "epoch": 2.2, - "grad_norm": 0.9164018630981445, + "grad_norm": 37.780216217041016, "learning_rate": 1.560671665434576e-05, - "loss": 0.583, + "loss": 1.3743, "step": 20800 }, { "epoch": 2.2, - "grad_norm": 10.196127891540527, + "grad_norm": 42.22161102294922, "learning_rate": 1.560460449889112e-05, - "loss": 0.26, + "loss": 1.4315, "step": 20810 }, { "epoch": 2.2, - "grad_norm": 17.544260025024414, + "grad_norm": 43.66453170776367, "learning_rate": 1.5602492343436478e-05, - "loss": 0.9014, + "loss": 1.6051, "step": 20820 }, { "epoch": 2.2, - "grad_norm": 7.323877811431885, + "grad_norm": 31.179906845092773, "learning_rate": 1.5600380187981835e-05, - "loss": 0.6443, + "loss": 1.2722, "step": 20830 }, { "epoch": 2.2, - "grad_norm": 12.24282455444336, + "grad_norm": 34.36140441894531, "learning_rate": 1.5598268032527195e-05, - "loss": 0.6689, + "loss": 1.5045, "step": 20840 }, { "epoch": 2.2, - "grad_norm": 25.293521881103516, + "grad_norm": 59.87425231933594, "learning_rate": 1.5596155877072552e-05, - "loss": 0.7705, + "loss": 1.2729, "step": 20850 }, { "epoch": 2.2, - "grad_norm": 11.7415132522583, + "grad_norm": 50.22679901123047, "learning_rate": 1.5594043721617913e-05, - "loss": 0.8215, + "loss": 1.2026, "step": 20860 }, { "epoch": 2.2, - "grad_norm": 11.887205123901367, + "grad_norm": 41.16181182861328, "learning_rate": 1.5591931566163273e-05, - "loss": 0.4931, + "loss": 1.3128, "step": 20870 }, { "epoch": 2.21, - "grad_norm": 2.9110944271087646, + "grad_norm": 40.15987014770508, "learning_rate": 1.5589819410708627e-05, - "loss": 0.5275, + "loss": 1.2948, "step": 20880 }, { "epoch": 2.21, - "grad_norm": 17.406780242919922, + "grad_norm": 58.18471145629883, "learning_rate": 1.5587707255253988e-05, - "loss": 0.7486, + "loss": 1.6571, "step": 20890 }, { "epoch": 2.21, - "grad_norm": 27.37065887451172, + "grad_norm": 49.572044372558594, "learning_rate": 1.5585595099799348e-05, - "loss": 0.6799, + "loss": 1.4848, "step": 20900 }, { "epoch": 2.21, - "grad_norm": 8.836884498596191, + "grad_norm": 37.16404724121094, "learning_rate": 1.5583482944344705e-05, - "loss": 0.6689, + "loss": 1.2172, "step": 20910 }, { "epoch": 2.21, - "grad_norm": 14.439940452575684, + "grad_norm": 38.777408599853516, "learning_rate": 1.5581370788890066e-05, - "loss": 0.7403, + "loss": 1.5141, "step": 20920 }, { "epoch": 2.21, - "grad_norm": 19.73255729675293, + "grad_norm": 24.493844985961914, "learning_rate": 1.5579258633435423e-05, - "loss": 0.4806, + "loss": 1.3169, "step": 20930 }, { "epoch": 2.21, - "grad_norm": 19.709484100341797, + "grad_norm": 28.929563522338867, "learning_rate": 1.557714647798078e-05, - "loss": 0.7441, + "loss": 1.3927, "step": 20940 }, { "epoch": 2.21, - "grad_norm": 5.496258735656738, + "grad_norm": 34.89937973022461, "learning_rate": 1.557503432252614e-05, - "loss": 0.832, + "loss": 1.6327, "step": 20950 }, { "epoch": 2.21, - "grad_norm": 17.32176399230957, + "grad_norm": 50.5921516418457, "learning_rate": 1.5572922167071497e-05, - "loss": 0.8081, + "loss": 1.3514, "step": 20960 }, { "epoch": 2.21, - "grad_norm": 5.946473121643066, + "grad_norm": 37.36489486694336, "learning_rate": 1.5570810011616858e-05, - "loss": 0.8142, + "loss": 1.6432, "step": 20970 }, { "epoch": 2.22, - "grad_norm": 25.141651153564453, + "grad_norm": 57.2741813659668, "learning_rate": 1.5568697856162215e-05, - "loss": 0.7182, + "loss": 1.2867, "step": 20980 }, { "epoch": 2.22, - "grad_norm": 5.786047458648682, + "grad_norm": 26.93438148498535, "learning_rate": 1.5566585700707572e-05, - "loss": 0.6125, + "loss": 1.3341, "step": 20990 }, { "epoch": 2.22, - "grad_norm": 14.78261947631836, + "grad_norm": 43.21870040893555, "learning_rate": 1.5564473545252932e-05, - "loss": 0.8868, + "loss": 1.4957, "step": 21000 }, { "epoch": 2.22, - "grad_norm": 19.852964401245117, + "grad_norm": 44.42601776123047, "learning_rate": 1.556236138979829e-05, - "loss": 0.6747, + "loss": 1.1897, "step": 21010 }, { "epoch": 2.22, - "grad_norm": 11.265368461608887, + "grad_norm": 37.22753143310547, "learning_rate": 1.556024923434365e-05, - "loss": 0.7878, + "loss": 1.424, "step": 21020 }, { "epoch": 2.22, - "grad_norm": 17.318052291870117, + "grad_norm": 35.439659118652344, "learning_rate": 1.5558137078889007e-05, - "loss": 0.4875, + "loss": 1.1376, "step": 21030 }, { "epoch": 2.22, - "grad_norm": 28.074193954467773, + "grad_norm": 33.489723205566406, "learning_rate": 1.5556024923434364e-05, - "loss": 0.7384, + "loss": 1.3181, "step": 21040 }, { "epoch": 2.22, - "grad_norm": 1.6252293586730957, + "grad_norm": 30.748666763305664, "learning_rate": 1.5553912767979725e-05, - "loss": 0.517, + "loss": 1.2425, "step": 21050 }, { "epoch": 2.22, - "grad_norm": 4.549213886260986, + "grad_norm": 37.95582580566406, "learning_rate": 1.555180061252508e-05, - "loss": 0.6747, + "loss": 1.1868, "step": 21060 }, { "epoch": 2.23, - "grad_norm": 0.5954540967941284, + "grad_norm": 30.742515563964844, "learning_rate": 1.5549688457070442e-05, - "loss": 0.7415, + "loss": 1.2508, "step": 21070 }, { "epoch": 2.23, - "grad_norm": 17.147483825683594, + "grad_norm": 43.48808670043945, "learning_rate": 1.5547576301615803e-05, - "loss": 0.5568, + "loss": 1.359, "step": 21080 }, { "epoch": 2.23, - "grad_norm": 10.798592567443848, + "grad_norm": 34.832557678222656, "learning_rate": 1.5545464146161156e-05, - "loss": 0.6375, + "loss": 1.5678, "step": 21090 }, { "epoch": 2.23, - "grad_norm": 10.103055000305176, + "grad_norm": 36.47542953491211, "learning_rate": 1.5543351990706517e-05, - "loss": 0.5952, + "loss": 1.0164, "step": 21100 }, { "epoch": 2.23, - "grad_norm": 27.484264373779297, + "grad_norm": 45.006988525390625, "learning_rate": 1.5541239835251877e-05, - "loss": 0.642, + "loss": 1.3028, "step": 21110 }, { "epoch": 2.23, - "grad_norm": 5.593727111816406, + "grad_norm": 53.28479766845703, "learning_rate": 1.5539127679797234e-05, - "loss": 0.6831, + "loss": 1.5964, "step": 21120 }, { "epoch": 2.23, - "grad_norm": 1.2572704553604126, + "grad_norm": 50.68624496459961, "learning_rate": 1.5537015524342595e-05, - "loss": 1.3607, + "loss": 1.7433, "step": 21130 }, { "epoch": 2.23, - "grad_norm": 13.55418586730957, + "grad_norm": 40.54065704345703, "learning_rate": 1.5534903368887952e-05, - "loss": 0.7107, + "loss": 1.3696, "step": 21140 }, { "epoch": 2.23, - "grad_norm": 13.654827117919922, + "grad_norm": 32.04253005981445, "learning_rate": 1.553279121343331e-05, - "loss": 0.819, + "loss": 1.3103, "step": 21150 }, { "epoch": 2.23, - "grad_norm": 10.482336044311523, + "grad_norm": 34.423587799072266, "learning_rate": 1.553067905797867e-05, - "loss": 0.687, + "loss": 1.3949, "step": 21160 }, { "epoch": 2.24, - "grad_norm": 14.792290687561035, + "grad_norm": 28.335403442382812, "learning_rate": 1.5528566902524027e-05, - "loss": 0.7485, + "loss": 1.1386, "step": 21170 }, { "epoch": 2.24, - "grad_norm": 9.22032642364502, + "grad_norm": 45.774662017822266, "learning_rate": 1.5526454747069387e-05, - "loss": 0.6251, + "loss": 1.6007, "step": 21180 }, { "epoch": 2.24, - "grad_norm": 9.655715942382812, + "grad_norm": 39.763675689697266, "learning_rate": 1.5524342591614744e-05, - "loss": 0.7864, + "loss": 1.4927, "step": 21190 }, { "epoch": 2.24, - "grad_norm": 9.23440170288086, + "grad_norm": 29.800655364990234, "learning_rate": 1.55222304361601e-05, - "loss": 0.7215, + "loss": 1.2343, "step": 21200 }, { "epoch": 2.24, - "grad_norm": 16.842681884765625, + "grad_norm": 32.8312873840332, "learning_rate": 1.552011828070546e-05, - "loss": 0.8621, + "loss": 1.7994, "step": 21210 }, { "epoch": 2.24, - "grad_norm": 12.829689979553223, + "grad_norm": 44.471168518066406, "learning_rate": 1.551800612525082e-05, - "loss": 0.5353, + "loss": 1.0998, "step": 21220 }, { "epoch": 2.24, - "grad_norm": 42.714420318603516, + "grad_norm": 41.56793975830078, "learning_rate": 1.551589396979618e-05, - "loss": 0.4562, + "loss": 1.1252, "step": 21230 }, { "epoch": 2.24, - "grad_norm": 5.291092395782471, + "grad_norm": 64.39690399169922, "learning_rate": 1.5513781814341536e-05, - "loss": 0.8353, + "loss": 1.5086, "step": 21240 }, { "epoch": 2.24, - "grad_norm": 8.281574249267578, + "grad_norm": 31.220596313476562, "learning_rate": 1.5511669658886893e-05, - "loss": 0.6174, + "loss": 1.4157, "step": 21250 }, { "epoch": 2.25, - "grad_norm": 3.1780948638916016, + "grad_norm": 42.6527214050293, "learning_rate": 1.5509557503432254e-05, - "loss": 0.7395, + "loss": 1.2839, "step": 21260 }, { "epoch": 2.25, - "grad_norm": 12.700603485107422, + "grad_norm": 41.30350875854492, "learning_rate": 1.550744534797761e-05, - "loss": 0.637, + "loss": 1.4582, "step": 21270 }, { "epoch": 2.25, - "grad_norm": 10.703070640563965, + "grad_norm": 26.258865356445312, "learning_rate": 1.550533319252297e-05, - "loss": 0.6373, + "loss": 1.3205, "step": 21280 }, { "epoch": 2.25, - "grad_norm": 27.108440399169922, + "grad_norm": 48.67230224609375, "learning_rate": 1.550322103706833e-05, - "loss": 1.111, + "loss": 1.5843, "step": 21290 }, { "epoch": 2.25, - "grad_norm": 19.559213638305664, + "grad_norm": 30.093469619750977, "learning_rate": 1.550110888161369e-05, - "loss": 0.6154, + "loss": 1.2687, "step": 21300 }, { "epoch": 2.25, - "grad_norm": 17.320751190185547, + "grad_norm": 34.47538375854492, "learning_rate": 1.5498996726159046e-05, - "loss": 0.6551, + "loss": 1.588, "step": 21310 }, { "epoch": 2.25, - "grad_norm": 5.853331089019775, + "grad_norm": 35.11906433105469, "learning_rate": 1.5496884570704403e-05, - "loss": 0.4979, + "loss": 1.2617, "step": 21320 }, { "epoch": 2.25, - "grad_norm": 4.55927038192749, + "grad_norm": 36.9468994140625, "learning_rate": 1.5494772415249764e-05, - "loss": 0.5186, + "loss": 1.2963, "step": 21330 }, { "epoch": 2.25, - "grad_norm": 18.19736671447754, + "grad_norm": 48.441436767578125, "learning_rate": 1.5492660259795124e-05, - "loss": 0.4868, + "loss": 1.2065, "step": 21340 }, { "epoch": 2.25, - "grad_norm": 5.360877513885498, + "grad_norm": 42.03391647338867, "learning_rate": 1.549054810434048e-05, - "loss": 0.5689, + "loss": 1.314, "step": 21350 }, { "epoch": 2.26, - "grad_norm": 14.3764009475708, + "grad_norm": 28.983388900756836, "learning_rate": 1.5488435948885838e-05, - "loss": 1.0326, + "loss": 1.7031, "step": 21360 }, { "epoch": 2.26, - "grad_norm": 4.832947254180908, + "grad_norm": 43.49089050292969, "learning_rate": 1.54863237934312e-05, - "loss": 0.7856, + "loss": 1.3699, "step": 21370 }, { "epoch": 2.26, - "grad_norm": 10.418540000915527, + "grad_norm": 46.53996276855469, "learning_rate": 1.5484211637976556e-05, - "loss": 0.4751, + "loss": 1.3063, "step": 21380 }, { "epoch": 2.26, - "grad_norm": 12.240233421325684, + "grad_norm": 20.67751693725586, "learning_rate": 1.5482099482521916e-05, - "loss": 0.5623, + "loss": 1.025, "step": 21390 }, { "epoch": 2.26, - "grad_norm": 9.067007064819336, + "grad_norm": 31.819507598876953, "learning_rate": 1.5479987327067273e-05, - "loss": 0.6646, + "loss": 1.4364, "step": 21400 }, { "epoch": 2.26, - "grad_norm": 13.978510856628418, + "grad_norm": 43.869998931884766, "learning_rate": 1.5477875171612634e-05, - "loss": 0.6388, + "loss": 1.4466, "step": 21410 }, { "epoch": 2.26, - "grad_norm": 8.453396797180176, + "grad_norm": 30.3518009185791, "learning_rate": 1.547576301615799e-05, - "loss": 0.6981, + "loss": 1.2849, "step": 21420 }, { "epoch": 2.26, - "grad_norm": 12.572386741638184, + "grad_norm": 55.262107849121094, "learning_rate": 1.5473650860703348e-05, - "loss": 0.9149, + "loss": 1.5983, "step": 21430 }, { "epoch": 2.26, - "grad_norm": 21.567840576171875, + "grad_norm": 44.6037712097168, "learning_rate": 1.547153870524871e-05, - "loss": 0.4434, + "loss": 1.252, "step": 21440 }, { "epoch": 2.27, - "grad_norm": 9.299983978271484, + "grad_norm": 42.96061706542969, "learning_rate": 1.5469426549794066e-05, - "loss": 0.9518, + "loss": 1.6493, "step": 21450 }, { "epoch": 2.27, - "grad_norm": 22.869029998779297, + "grad_norm": 37.28997802734375, "learning_rate": 1.5467314394339426e-05, - "loss": 0.8157, + "loss": 1.5137, "step": 21460 }, { "epoch": 2.27, - "grad_norm": 5.923974514007568, + "grad_norm": 34.296234130859375, "learning_rate": 1.5465202238884783e-05, - "loss": 0.4608, + "loss": 1.4235, "step": 21470 }, { "epoch": 2.27, - "grad_norm": 17.724475860595703, + "grad_norm": 39.31536102294922, "learning_rate": 1.546309008343014e-05, - "loss": 0.6558, + "loss": 1.4995, "step": 21480 }, { "epoch": 2.27, - "grad_norm": 4.647461891174316, + "grad_norm": 33.018489837646484, "learning_rate": 1.54609779279755e-05, - "loss": 0.6005, + "loss": 1.5871, "step": 21490 }, { "epoch": 2.27, - "grad_norm": 21.56606674194336, + "grad_norm": 35.29375076293945, "learning_rate": 1.5458865772520858e-05, - "loss": 0.5512, + "loss": 1.199, "step": 21500 }, { "epoch": 2.27, - "grad_norm": 6.954745769500732, + "grad_norm": 36.68441390991211, "learning_rate": 1.5456753617066218e-05, - "loss": 0.6316, + "loss": 1.2047, "step": 21510 }, { "epoch": 2.27, - "grad_norm": 9.071138381958008, + "grad_norm": 48.51303482055664, "learning_rate": 1.5454641461611575e-05, - "loss": 0.6574, + "loss": 1.6924, "step": 21520 }, { "epoch": 2.27, - "grad_norm": 9.50493335723877, + "grad_norm": 29.234710693359375, "learning_rate": 1.5452529306156932e-05, - "loss": 0.8298, + "loss": 1.4063, "step": 21530 }, { "epoch": 2.27, - "grad_norm": 17.70240592956543, + "grad_norm": 41.08154296875, "learning_rate": 1.5450417150702293e-05, - "loss": 0.5867, + "loss": 1.6373, "step": 21540 }, { "epoch": 2.28, - "grad_norm": 14.37537670135498, + "grad_norm": 24.42667007446289, "learning_rate": 1.544830499524765e-05, - "loss": 0.7063, + "loss": 1.3159, "step": 21550 }, { "epoch": 2.28, - "grad_norm": 16.28536605834961, + "grad_norm": 32.04637145996094, "learning_rate": 1.544619283979301e-05, - "loss": 0.5914, + "loss": 1.2658, "step": 21560 }, { "epoch": 2.28, - "grad_norm": 3.3853840827941895, + "grad_norm": 40.470340728759766, "learning_rate": 1.544408068433837e-05, - "loss": 0.6335, + "loss": 1.0558, "step": 21570 }, { "epoch": 2.28, - "grad_norm": 12.500550270080566, + "grad_norm": 40.920021057128906, "learning_rate": 1.5441968528883725e-05, - "loss": 0.8063, + "loss": 1.5676, "step": 21580 }, { "epoch": 2.28, - "grad_norm": 4.336775779724121, + "grad_norm": 40.70728302001953, "learning_rate": 1.5439856373429085e-05, - "loss": 0.5842, + "loss": 1.4291, "step": 21590 }, { "epoch": 2.28, - "grad_norm": 6.000080108642578, + "grad_norm": 24.124713897705078, "learning_rate": 1.5437744217974446e-05, - "loss": 0.4188, + "loss": 1.2819, "step": 21600 }, { "epoch": 2.28, - "grad_norm": 6.68609094619751, + "grad_norm": 44.487937927246094, "learning_rate": 1.5435632062519803e-05, - "loss": 0.6109, + "loss": 1.4245, "step": 21610 }, { "epoch": 2.28, - "grad_norm": 1.1625694036483765, + "grad_norm": 31.5529727935791, "learning_rate": 1.5433519907065163e-05, - "loss": 0.4391, + "loss": 1.0018, "step": 21620 }, { "epoch": 2.28, - "grad_norm": 15.839173316955566, + "grad_norm": 28.539506912231445, "learning_rate": 1.543140775161052e-05, - "loss": 0.6003, + "loss": 1.3582, "step": 21630 }, { "epoch": 2.29, - "grad_norm": 14.549817085266113, + "grad_norm": 36.42344665527344, "learning_rate": 1.5429295596155877e-05, - "loss": 0.6533, + "loss": 1.214, "step": 21640 }, { "epoch": 2.29, - "grad_norm": 16.383378982543945, + "grad_norm": 34.75198745727539, "learning_rate": 1.5427183440701238e-05, - "loss": 0.394, + "loss": 1.3326, "step": 21650 }, { "epoch": 2.29, - "grad_norm": 13.818933486938477, + "grad_norm": 54.686187744140625, "learning_rate": 1.5425071285246595e-05, - "loss": 0.772, + "loss": 1.5504, "step": 21660 }, { "epoch": 2.29, - "grad_norm": 9.618260383605957, + "grad_norm": 36.241085052490234, "learning_rate": 1.5422959129791955e-05, - "loss": 0.5997, + "loss": 1.3479, "step": 21670 }, { "epoch": 2.29, - "grad_norm": 13.01084041595459, + "grad_norm": 24.804519653320312, "learning_rate": 1.5420846974337312e-05, - "loss": 0.5856, + "loss": 1.287, "step": 21680 }, { "epoch": 2.29, - "grad_norm": 6.671026229858398, + "grad_norm": 39.667762756347656, "learning_rate": 1.541873481888267e-05, - "loss": 0.4631, + "loss": 1.1542, "step": 21690 }, { "epoch": 2.29, - "grad_norm": 6.461094379425049, + "grad_norm": 29.021732330322266, "learning_rate": 1.541662266342803e-05, - "loss": 0.6062, + "loss": 1.0972, "step": 21700 }, { "epoch": 2.29, - "grad_norm": 14.229743957519531, + "grad_norm": 30.576553344726562, "learning_rate": 1.5414510507973387e-05, - "loss": 0.9077, + "loss": 1.517, "step": 21710 }, { "epoch": 2.29, - "grad_norm": 1.8420134782791138, + "grad_norm": 17.474437713623047, "learning_rate": 1.5412398352518748e-05, - "loss": 0.6069, + "loss": 1.407, "step": 21720 }, { "epoch": 2.29, - "grad_norm": 6.204957962036133, + "grad_norm": 36.31136703491211, "learning_rate": 1.5410286197064105e-05, - "loss": 0.6072, + "loss": 1.1834, "step": 21730 }, { "epoch": 2.3, - "grad_norm": 23.223087310791016, + "grad_norm": 43.671661376953125, "learning_rate": 1.540817404160946e-05, - "loss": 0.5953, + "loss": 1.3278, "step": 21740 }, { "epoch": 2.3, - "grad_norm": 1.7526785135269165, + "grad_norm": 14.894190788269043, "learning_rate": 1.5406061886154822e-05, - "loss": 0.7397, + "loss": 1.3543, "step": 21750 }, { "epoch": 2.3, - "grad_norm": 11.710954666137695, + "grad_norm": 31.73625373840332, "learning_rate": 1.540394973070018e-05, - "loss": 0.5402, + "loss": 1.1622, "step": 21760 }, { "epoch": 2.3, - "grad_norm": 4.784595489501953, + "grad_norm": 51.463417053222656, "learning_rate": 1.540183757524554e-05, - "loss": 0.43, + "loss": 1.4345, "step": 21770 }, { "epoch": 2.3, - "grad_norm": 18.417055130004883, + "grad_norm": 22.04267692565918, "learning_rate": 1.53997254197909e-05, - "loss": 0.5232, + "loss": 1.3778, "step": 21780 }, { "epoch": 2.3, - "grad_norm": 10.382081031799316, + "grad_norm": 24.528722763061523, "learning_rate": 1.5397613264336257e-05, - "loss": 0.7127, + "loss": 1.3129, "step": 21790 }, { "epoch": 2.3, - "grad_norm": 24.962635040283203, + "grad_norm": 42.81536865234375, "learning_rate": 1.5395501108881614e-05, - "loss": 0.6644, + "loss": 1.2872, "step": 21800 }, { "epoch": 2.3, - "grad_norm": 5.968369007110596, + "grad_norm": 41.56134033203125, "learning_rate": 1.5393388953426975e-05, - "loss": 0.7983, + "loss": 1.5969, "step": 21810 }, { "epoch": 2.3, - "grad_norm": 23.44498634338379, + "grad_norm": 33.60396194458008, "learning_rate": 1.5391276797972332e-05, - "loss": 0.9181, + "loss": 1.5629, "step": 21820 }, { "epoch": 2.31, - "grad_norm": 0.9164698719978333, + "grad_norm": 32.52360153198242, "learning_rate": 1.5389164642517692e-05, - "loss": 0.6397, + "loss": 1.4556, "step": 21830 }, { "epoch": 2.31, - "grad_norm": 2.2549033164978027, + "grad_norm": 29.302644729614258, "learning_rate": 1.538705248706305e-05, - "loss": 0.7362, + "loss": 1.3395, "step": 21840 }, { "epoch": 2.31, - "grad_norm": 13.738150596618652, + "grad_norm": 35.785335540771484, "learning_rate": 1.5384940331608407e-05, - "loss": 0.7046, + "loss": 1.2948, "step": 21850 }, { "epoch": 2.31, - "grad_norm": 4.713522911071777, + "grad_norm": 33.23189163208008, "learning_rate": 1.5382828176153767e-05, - "loss": 0.5839, + "loss": 1.2427, "step": 21860 }, { "epoch": 2.31, - "grad_norm": 20.472169876098633, + "grad_norm": 38.34109878540039, "learning_rate": 1.5380716020699124e-05, - "loss": 0.7535, + "loss": 1.4177, "step": 21870 }, { "epoch": 2.31, - "grad_norm": 10.776202201843262, + "grad_norm": 36.29264831542969, "learning_rate": 1.5378603865244485e-05, - "loss": 0.7504, + "loss": 1.5319, "step": 21880 }, { "epoch": 2.31, - "grad_norm": 24.855201721191406, + "grad_norm": 32.57075881958008, "learning_rate": 1.537649170978984e-05, - "loss": 0.6586, + "loss": 1.4627, "step": 21890 }, { "epoch": 2.31, - "grad_norm": 7.121510982513428, + "grad_norm": 50.43296432495117, "learning_rate": 1.53743795543352e-05, - "loss": 0.6648, + "loss": 1.3813, "step": 21900 }, { "epoch": 2.31, - "grad_norm": 12.187426567077637, + "grad_norm": 48.08408737182617, "learning_rate": 1.537226739888056e-05, - "loss": 0.6779, + "loss": 1.4137, "step": 21910 }, { "epoch": 2.31, - "grad_norm": 0.6647688150405884, + "grad_norm": 21.59383773803711, "learning_rate": 1.5370155243425916e-05, - "loss": 0.4682, + "loss": 1.1801, "step": 21920 }, { "epoch": 2.32, - "grad_norm": 16.949668884277344, + "grad_norm": 45.33760452270508, "learning_rate": 1.5368043087971277e-05, - "loss": 0.3239, + "loss": 1.2597, "step": 21930 }, { "epoch": 2.32, - "grad_norm": 16.679271697998047, + "grad_norm": 39.430381774902344, "learning_rate": 1.5365930932516634e-05, - "loss": 0.88, + "loss": 1.6751, "step": 21940 }, { "epoch": 2.32, - "grad_norm": 23.257831573486328, + "grad_norm": 35.37932205200195, "learning_rate": 1.5363818777061994e-05, - "loss": 0.5457, + "loss": 1.2754, "step": 21950 }, { "epoch": 2.32, - "grad_norm": 8.183807373046875, + "grad_norm": 34.23918914794922, "learning_rate": 1.536170662160735e-05, - "loss": 0.7158, + "loss": 1.0302, "step": 21960 }, { "epoch": 2.32, - "grad_norm": 0.08776625990867615, + "grad_norm": 30.199758529663086, "learning_rate": 1.535959446615271e-05, - "loss": 0.7334, + "loss": 1.2985, "step": 21970 }, { "epoch": 2.32, - "grad_norm": 13.428850173950195, + "grad_norm": 23.294965744018555, "learning_rate": 1.535748231069807e-05, - "loss": 0.7655, + "loss": 1.4254, "step": 21980 }, { "epoch": 2.32, - "grad_norm": 17.07765769958496, + "grad_norm": 18.46173858642578, "learning_rate": 1.5355370155243426e-05, - "loss": 0.5551, + "loss": 1.1125, "step": 21990 }, { "epoch": 2.32, - "grad_norm": 15.143954277038574, + "grad_norm": 34.18079376220703, "learning_rate": 1.5353257999788787e-05, - "loss": 0.5866, + "loss": 1.2189, "step": 22000 }, { "epoch": 2.32, - "grad_norm": 12.231800079345703, + "grad_norm": 44.6326904296875, "learning_rate": 1.5351145844334144e-05, - "loss": 0.7086, + "loss": 1.5978, "step": 22010 }, { "epoch": 2.33, - "grad_norm": 21.683544158935547, + "grad_norm": 55.09823989868164, "learning_rate": 1.53490336888795e-05, - "loss": 0.8631, + "loss": 1.6089, "step": 22020 }, { "epoch": 2.33, - "grad_norm": 11.683777809143066, + "grad_norm": 32.353424072265625, "learning_rate": 1.534692153342486e-05, - "loss": 0.6608, + "loss": 1.1322, "step": 22030 }, { "epoch": 2.33, - "grad_norm": 0.9550610184669495, + "grad_norm": 44.6668586730957, "learning_rate": 1.534480937797022e-05, - "loss": 0.4493, + "loss": 1.2472, "step": 22040 }, { "epoch": 2.33, - "grad_norm": 2.925600051879883, + "grad_norm": 37.95564270019531, "learning_rate": 1.534269722251558e-05, - "loss": 0.8414, + "loss": 1.4985, "step": 22050 }, { "epoch": 2.33, - "grad_norm": 17.908008575439453, + "grad_norm": 36.38212585449219, "learning_rate": 1.534058506706094e-05, - "loss": 0.7154, + "loss": 1.4289, "step": 22060 }, { "epoch": 2.33, - "grad_norm": 18.7332820892334, + "grad_norm": 24.98573112487793, "learning_rate": 1.5338472911606296e-05, - "loss": 0.6667, + "loss": 1.3964, "step": 22070 }, { "epoch": 2.33, - "grad_norm": 6.4643754959106445, + "grad_norm": 25.457542419433594, "learning_rate": 1.5336360756151653e-05, - "loss": 0.6954, + "loss": 1.322, "step": 22080 }, { "epoch": 2.33, - "grad_norm": 32.54081726074219, + "grad_norm": 49.65493392944336, "learning_rate": 1.5334248600697014e-05, - "loss": 0.8116, + "loss": 1.7126, "step": 22090 }, { "epoch": 2.33, - "grad_norm": 13.665678977966309, + "grad_norm": 33.69675064086914, "learning_rate": 1.533213644524237e-05, - "loss": 0.4306, + "loss": 1.1034, "step": 22100 }, { "epoch": 2.33, - "grad_norm": 18.27852439880371, + "grad_norm": 37.874542236328125, "learning_rate": 1.533002428978773e-05, - "loss": 0.5774, + "loss": 1.1861, "step": 22110 }, { "epoch": 2.34, - "grad_norm": 2.189116954803467, + "grad_norm": 44.41117858886719, "learning_rate": 1.532791213433309e-05, - "loss": 0.793, + "loss": 1.3763, "step": 22120 }, { "epoch": 2.34, - "grad_norm": 15.513866424560547, + "grad_norm": 22.87313461303711, "learning_rate": 1.5325799978878446e-05, - "loss": 0.3439, + "loss": 0.8332, "step": 22130 }, { "epoch": 2.34, - "grad_norm": 25.298885345458984, + "grad_norm": 45.89292526245117, "learning_rate": 1.5323687823423806e-05, - "loss": 0.5917, + "loss": 1.4035, "step": 22140 }, { "epoch": 2.34, - "grad_norm": 5.430473804473877, + "grad_norm": 36.06021499633789, "learning_rate": 1.5321575667969163e-05, - "loss": 0.5731, + "loss": 1.2834, "step": 22150 }, { "epoch": 2.34, - "grad_norm": 10.665436744689941, + "grad_norm": 33.15639877319336, "learning_rate": 1.5319463512514524e-05, - "loss": 0.8428, + "loss": 1.7658, "step": 22160 }, { "epoch": 2.34, - "grad_norm": 24.175365447998047, + "grad_norm": 40.32683563232422, "learning_rate": 1.531735135705988e-05, - "loss": 1.0023, + "loss": 1.712, "step": 22170 }, { "epoch": 2.34, - "grad_norm": 15.416291236877441, + "grad_norm": 24.89621925354004, "learning_rate": 1.5315239201605238e-05, - "loss": 0.5868, + "loss": 1.2081, "step": 22180 }, { "epoch": 2.34, - "grad_norm": 30.82347869873047, + "grad_norm": 31.13288116455078, "learning_rate": 1.5313127046150598e-05, - "loss": 0.5946, + "loss": 1.2883, "step": 22190 }, { "epoch": 2.34, - "grad_norm": 20.4368839263916, + "grad_norm": 48.934017181396484, "learning_rate": 1.5311014890695955e-05, - "loss": 0.8854, + "loss": 1.4457, "step": 22200 }, { "epoch": 2.35, - "grad_norm": 22.54294776916504, + "grad_norm": 50.473453521728516, "learning_rate": 1.5308902735241316e-05, - "loss": 0.5899, + "loss": 1.4591, "step": 22210 }, { "epoch": 2.35, - "grad_norm": 13.446962356567383, + "grad_norm": 48.38077926635742, "learning_rate": 1.5306790579786673e-05, - "loss": 0.3646, + "loss": 1.2566, "step": 22220 }, { "epoch": 2.35, - "grad_norm": 7.657071590423584, + "grad_norm": 30.17050552368164, "learning_rate": 1.530467842433203e-05, - "loss": 0.6705, + "loss": 1.2981, "step": 22230 }, { "epoch": 2.35, - "grad_norm": 9.908388137817383, + "grad_norm": 49.52378845214844, "learning_rate": 1.530256626887739e-05, - "loss": 0.6966, + "loss": 1.4025, "step": 22240 }, { "epoch": 2.35, - "grad_norm": 2.4402453899383545, + "grad_norm": 47.78527069091797, "learning_rate": 1.5300454113422748e-05, - "loss": 0.7196, + "loss": 1.4544, "step": 22250 }, { "epoch": 2.35, - "grad_norm": 11.936182975769043, + "grad_norm": 39.6124382019043, "learning_rate": 1.5298341957968108e-05, - "loss": 0.6211, + "loss": 1.4906, "step": 22260 }, { "epoch": 2.35, - "grad_norm": 23.375768661499023, + "grad_norm": 32.11890411376953, "learning_rate": 1.529622980251347e-05, - "loss": 0.5063, + "loss": 0.8538, "step": 22270 }, { "epoch": 2.35, - "grad_norm": 6.355723857879639, + "grad_norm": 22.371400833129883, "learning_rate": 1.5294117647058822e-05, - "loss": 0.3531, + "loss": 1.1585, "step": 22280 }, { "epoch": 2.35, - "grad_norm": 13.79101848602295, + "grad_norm": 41.185882568359375, "learning_rate": 1.5292005491604183e-05, - "loss": 0.759, + "loss": 1.6052, "step": 22290 }, { "epoch": 2.36, - "grad_norm": 10.78840160369873, + "grad_norm": 34.57830047607422, "learning_rate": 1.5289893336149543e-05, - "loss": 0.6935, + "loss": 1.2195, "step": 22300 }, { "epoch": 2.36, - "grad_norm": 0.6839573383331299, + "grad_norm": 29.77281951904297, "learning_rate": 1.52877811806949e-05, - "loss": 0.4199, + "loss": 1.3943, "step": 22310 }, { "epoch": 2.36, - "grad_norm": 9.143683433532715, + "grad_norm": 37.21223449707031, "learning_rate": 1.528566902524026e-05, - "loss": 0.9723, + "loss": 1.5376, "step": 22320 }, { "epoch": 2.36, - "grad_norm": 9.869205474853516, + "grad_norm": 23.655975341796875, "learning_rate": 1.5283556869785618e-05, - "loss": 0.74, + "loss": 1.2094, "step": 22330 }, { "epoch": 2.36, - "grad_norm": 10.5426607131958, + "grad_norm": 34.309696197509766, "learning_rate": 1.5281444714330975e-05, - "loss": 0.8282, + "loss": 1.5739, "step": 22340 }, { "epoch": 2.36, - "grad_norm": 7.307587623596191, + "grad_norm": 29.946758270263672, "learning_rate": 1.5279332558876335e-05, - "loss": 0.8785, + "loss": 1.5545, "step": 22350 }, { "epoch": 2.36, - "grad_norm": 19.544981002807617, + "grad_norm": 36.373329162597656, "learning_rate": 1.5277220403421692e-05, - "loss": 0.7538, + "loss": 1.4528, "step": 22360 }, { "epoch": 2.36, - "grad_norm": 17.136798858642578, + "grad_norm": 29.395967483520508, "learning_rate": 1.5275108247967053e-05, - "loss": 0.7914, + "loss": 1.7407, "step": 22370 }, { "epoch": 2.36, - "grad_norm": 18.88014030456543, + "grad_norm": 33.785831451416016, "learning_rate": 1.527299609251241e-05, - "loss": 0.4446, + "loss": 1.2024, "step": 22380 }, { "epoch": 2.36, - "grad_norm": 10.769773483276367, + "grad_norm": 35.61723327636719, "learning_rate": 1.5270883937057767e-05, - "loss": 0.5203, + "loss": 1.2394, "step": 22390 }, { "epoch": 2.37, - "grad_norm": 11.3226318359375, + "grad_norm": 36.08795928955078, "learning_rate": 1.5268771781603128e-05, - "loss": 0.612, + "loss": 1.0911, "step": 22400 }, { "epoch": 2.37, - "grad_norm": 18.227338790893555, + "grad_norm": 26.53786849975586, "learning_rate": 1.5266659626148485e-05, - "loss": 0.57, + "loss": 0.8838, "step": 22410 }, { "epoch": 2.37, - "grad_norm": 11.719352722167969, + "grad_norm": 30.171493530273438, "learning_rate": 1.5264547470693845e-05, - "loss": 0.3417, + "loss": 1.1546, "step": 22420 }, { "epoch": 2.37, - "grad_norm": 14.251320838928223, + "grad_norm": 23.354373931884766, "learning_rate": 1.5262435315239202e-05, - "loss": 0.7934, + "loss": 1.479, "step": 22430 }, { "epoch": 2.37, - "grad_norm": 23.964441299438477, + "grad_norm": 46.2501335144043, "learning_rate": 1.5260323159784563e-05, - "loss": 0.7547, + "loss": 1.3569, "step": 22440 }, { "epoch": 2.37, - "grad_norm": 12.228070259094238, + "grad_norm": 56.049102783203125, "learning_rate": 1.525821100432992e-05, - "loss": 0.9064, + "loss": 1.7919, "step": 22450 }, { "epoch": 2.37, - "grad_norm": 16.60118293762207, + "grad_norm": 32.09893798828125, "learning_rate": 1.5256098848875279e-05, - "loss": 0.3871, + "loss": 1.2271, "step": 22460 }, { "epoch": 2.37, - "grad_norm": 7.1247992515563965, + "grad_norm": 31.517932891845703, "learning_rate": 1.5253986693420637e-05, - "loss": 0.6033, + "loss": 1.0995, "step": 22470 }, { "epoch": 2.37, - "grad_norm": 15.6358060836792, + "grad_norm": 33.30202102661133, "learning_rate": 1.5251874537965996e-05, - "loss": 0.7693, + "loss": 1.4184, "step": 22480 }, { "epoch": 2.38, - "grad_norm": 14.182680130004883, + "grad_norm": 45.963714599609375, "learning_rate": 1.5249762382511355e-05, - "loss": 0.7769, + "loss": 1.3326, "step": 22490 }, { "epoch": 2.38, - "grad_norm": 12.887359619140625, + "grad_norm": 57.85136795043945, "learning_rate": 1.5247650227056712e-05, - "loss": 0.8363, + "loss": 1.1987, "step": 22500 }, { "epoch": 2.38, - "grad_norm": 5.0198445320129395, + "grad_norm": 43.63020324707031, "learning_rate": 1.524553807160207e-05, - "loss": 0.5298, + "loss": 1.1383, "step": 22510 }, { "epoch": 2.38, - "grad_norm": 5.882786750793457, + "grad_norm": 55.784725189208984, "learning_rate": 1.524342591614743e-05, - "loss": 0.5655, + "loss": 1.6275, "step": 22520 }, { "epoch": 2.38, - "grad_norm": 1.2655830383300781, + "grad_norm": 45.03480529785156, "learning_rate": 1.5241313760692788e-05, - "loss": 0.4446, + "loss": 1.1564, "step": 22530 }, { "epoch": 2.38, - "grad_norm": 25.758411407470703, + "grad_norm": 52.10029220581055, "learning_rate": 1.5239201605238147e-05, - "loss": 0.89, + "loss": 1.5791, "step": 22540 }, { "epoch": 2.38, - "grad_norm": 4.968374729156494, + "grad_norm": 42.3392333984375, "learning_rate": 1.5237089449783504e-05, - "loss": 0.8238, + "loss": 1.1949, "step": 22550 }, { "epoch": 2.38, - "grad_norm": 11.663932800292969, + "grad_norm": 52.512489318847656, "learning_rate": 1.5234977294328863e-05, - "loss": 0.7187, + "loss": 1.4847, "step": 22560 }, { "epoch": 2.38, - "grad_norm": 3.0315558910369873, + "grad_norm": 32.6695556640625, "learning_rate": 1.5232865138874222e-05, - "loss": 0.7752, + "loss": 1.1607, "step": 22570 }, { "epoch": 2.38, - "grad_norm": 22.169754028320312, + "grad_norm": 29.762327194213867, "learning_rate": 1.523075298341958e-05, - "loss": 0.4752, + "loss": 1.1261, "step": 22580 }, { "epoch": 2.39, - "grad_norm": 22.349584579467773, + "grad_norm": 35.26763153076172, "learning_rate": 1.522864082796494e-05, - "loss": 0.6553, + "loss": 1.1971, "step": 22590 }, { "epoch": 2.39, - "grad_norm": 8.738900184631348, + "grad_norm": 41.33247375488281, "learning_rate": 1.52265286725103e-05, - "loss": 0.4565, + "loss": 1.1259, "step": 22600 }, { "epoch": 2.39, - "grad_norm": 21.25335121154785, + "grad_norm": 28.772382736206055, "learning_rate": 1.5224416517055655e-05, - "loss": 0.6051, + "loss": 1.3316, "step": 22610 }, { "epoch": 2.39, - "grad_norm": 6.3295063972473145, + "grad_norm": 36.99485397338867, "learning_rate": 1.5222304361601014e-05, - "loss": 0.5483, + "loss": 1.2961, "step": 22620 }, { "epoch": 2.39, - "grad_norm": 8.94655990600586, + "grad_norm": 30.131471633911133, "learning_rate": 1.5220192206146374e-05, - "loss": 0.6756, + "loss": 1.0917, "step": 22630 }, { "epoch": 2.39, - "grad_norm": 18.18517303466797, + "grad_norm": 39.88261032104492, "learning_rate": 1.5218080050691733e-05, - "loss": 0.6394, + "loss": 1.4339, "step": 22640 }, { "epoch": 2.39, - "grad_norm": 21.087966918945312, + "grad_norm": 24.361621856689453, "learning_rate": 1.5215967895237092e-05, - "loss": 0.624, + "loss": 1.1915, "step": 22650 }, { "epoch": 2.39, - "grad_norm": 31.340669631958008, + "grad_norm": 34.49489212036133, "learning_rate": 1.5213855739782449e-05, - "loss": 0.9901, + "loss": 1.5294, "step": 22660 }, { "epoch": 2.39, - "grad_norm": 15.666788101196289, + "grad_norm": 51.298946380615234, "learning_rate": 1.5211743584327808e-05, - "loss": 0.8452, + "loss": 1.5524, "step": 22670 }, { "epoch": 2.4, - "grad_norm": 8.560983657836914, + "grad_norm": 29.57882308959961, "learning_rate": 1.5209631428873167e-05, - "loss": 0.6542, + "loss": 1.639, "step": 22680 }, { "epoch": 2.4, - "grad_norm": 10.313796997070312, + "grad_norm": 39.45873260498047, "learning_rate": 1.5207519273418525e-05, - "loss": 0.5872, + "loss": 1.2308, "step": 22690 }, { "epoch": 2.4, - "grad_norm": 6.944286346435547, + "grad_norm": 30.023094177246094, "learning_rate": 1.5205407117963884e-05, - "loss": 1.0341, + "loss": 1.9366, "step": 22700 }, { "epoch": 2.4, - "grad_norm": 13.367341041564941, + "grad_norm": 62.648860931396484, "learning_rate": 1.5203294962509243e-05, - "loss": 0.5201, + "loss": 1.5555, "step": 22710 }, { "epoch": 2.4, - "grad_norm": 1.787010908126831, + "grad_norm": 38.27570343017578, "learning_rate": 1.52011828070546e-05, - "loss": 0.6445, + "loss": 1.3045, "step": 22720 }, { "epoch": 2.4, - "grad_norm": 0.2507963180541992, + "grad_norm": 27.548486709594727, "learning_rate": 1.5199070651599959e-05, - "loss": 0.512, + "loss": 1.0275, "step": 22730 }, { "epoch": 2.4, - "grad_norm": 8.530539512634277, + "grad_norm": 39.90218734741211, "learning_rate": 1.5196958496145318e-05, - "loss": 0.9637, + "loss": 1.5766, "step": 22740 }, { "epoch": 2.4, - "grad_norm": 20.63922119140625, + "grad_norm": 47.81583023071289, "learning_rate": 1.5194846340690676e-05, - "loss": 0.7337, + "loss": 1.3765, "step": 22750 }, { "epoch": 2.4, - "grad_norm": 2.31571364402771, + "grad_norm": 44.57272720336914, "learning_rate": 1.5192734185236035e-05, - "loss": 0.5999, + "loss": 1.2819, "step": 22760 }, { "epoch": 2.4, - "grad_norm": 24.87570571899414, + "grad_norm": 35.06266403198242, "learning_rate": 1.5190622029781392e-05, - "loss": 0.7605, + "loss": 1.3996, "step": 22770 }, { "epoch": 2.41, - "grad_norm": 24.579675674438477, + "grad_norm": 33.82888412475586, "learning_rate": 1.5188509874326751e-05, - "loss": 0.6992, + "loss": 1.3943, "step": 22780 }, { "epoch": 2.41, - "grad_norm": 8.108358383178711, + "grad_norm": 33.329471588134766, "learning_rate": 1.518639771887211e-05, - "loss": 0.68, + "loss": 1.5003, "step": 22790 }, { "epoch": 2.41, - "grad_norm": 9.367913246154785, + "grad_norm": 38.96764373779297, "learning_rate": 1.5184285563417469e-05, - "loss": 0.6753, + "loss": 1.5697, "step": 22800 }, { "epoch": 2.41, - "grad_norm": 12.597902297973633, + "grad_norm": 35.80365753173828, "learning_rate": 1.5182173407962827e-05, - "loss": 0.7629, + "loss": 1.2721, "step": 22810 }, { "epoch": 2.41, - "grad_norm": 12.640412330627441, + "grad_norm": 45.239768981933594, "learning_rate": 1.5180061252508184e-05, - "loss": 0.4635, + "loss": 1.3042, "step": 22820 }, { "epoch": 2.41, - "grad_norm": 9.322210311889648, + "grad_norm": 24.598390579223633, "learning_rate": 1.5177949097053543e-05, - "loss": 0.7846, + "loss": 1.2583, "step": 22830 }, { "epoch": 2.41, - "grad_norm": 17.64600944519043, + "grad_norm": 40.96147918701172, "learning_rate": 1.5175836941598902e-05, - "loss": 0.6751, + "loss": 1.6025, "step": 22840 }, { "epoch": 2.41, - "grad_norm": 3.43308162689209, + "grad_norm": 19.963285446166992, "learning_rate": 1.5173724786144262e-05, - "loss": 0.3912, + "loss": 1.0556, "step": 22850 }, { "epoch": 2.41, - "grad_norm": 18.755165100097656, + "grad_norm": 46.209495544433594, "learning_rate": 1.5171612630689621e-05, - "loss": 0.8192, + "loss": 1.4704, "step": 22860 }, { "epoch": 2.42, - "grad_norm": 3.6718218326568604, + "grad_norm": 37.566856384277344, "learning_rate": 1.516950047523498e-05, - "loss": 0.4409, + "loss": 1.4962, "step": 22870 }, { "epoch": 2.42, - "grad_norm": 3.8934147357940674, + "grad_norm": 42.33334732055664, "learning_rate": 1.5167388319780335e-05, - "loss": 0.621, + "loss": 1.3189, "step": 22880 }, { "epoch": 2.42, - "grad_norm": 10.446697235107422, + "grad_norm": 20.49213981628418, "learning_rate": 1.5165276164325696e-05, - "loss": 0.6111, + "loss": 1.2537, "step": 22890 }, { "epoch": 2.42, - "grad_norm": 6.294929027557373, + "grad_norm": 16.716636657714844, "learning_rate": 1.5163164008871055e-05, - "loss": 0.6319, + "loss": 1.4006, "step": 22900 }, { "epoch": 2.42, - "grad_norm": 15.194002151489258, + "grad_norm": 33.40438461303711, "learning_rate": 1.5161051853416413e-05, - "loss": 0.4186, + "loss": 0.994, "step": 22910 }, { "epoch": 2.42, - "grad_norm": 4.516061305999756, + "grad_norm": 36.78746795654297, "learning_rate": 1.5158939697961772e-05, - "loss": 0.5128, + "loss": 1.6571, "step": 22920 }, { "epoch": 2.42, - "grad_norm": 3.5163283348083496, + "grad_norm": 39.28904724121094, "learning_rate": 1.515682754250713e-05, - "loss": 0.7718, + "loss": 1.1699, "step": 22930 }, { "epoch": 2.42, - "grad_norm": 15.083410263061523, + "grad_norm": 43.38880157470703, "learning_rate": 1.5154715387052488e-05, - "loss": 0.6127, + "loss": 1.6618, "step": 22940 }, { "epoch": 2.42, - "grad_norm": 0.4305412173271179, + "grad_norm": 25.484272003173828, "learning_rate": 1.5152603231597847e-05, - "loss": 0.5708, + "loss": 1.1876, "step": 22950 }, { "epoch": 2.42, - "grad_norm": 32.37685775756836, + "grad_norm": 48.43402099609375, "learning_rate": 1.5150491076143206e-05, - "loss": 0.4723, + "loss": 1.2122, "step": 22960 }, { "epoch": 2.43, - "grad_norm": 19.321022033691406, + "grad_norm": 25.537107467651367, "learning_rate": 1.5148378920688564e-05, - "loss": 0.4677, + "loss": 1.0521, "step": 22970 }, { "epoch": 2.43, - "grad_norm": 8.519993782043457, + "grad_norm": 40.651187896728516, "learning_rate": 1.5146266765233923e-05, - "loss": 0.5031, + "loss": 1.165, "step": 22980 }, { "epoch": 2.43, - "grad_norm": 20.729040145874023, + "grad_norm": 44.38654708862305, "learning_rate": 1.514415460977928e-05, - "loss": 0.9515, + "loss": 1.5064, "step": 22990 }, { "epoch": 2.43, - "grad_norm": 4.841820240020752, + "grad_norm": 33.7844352722168, "learning_rate": 1.5142042454324639e-05, - "loss": 0.6423, + "loss": 1.1445, "step": 23000 }, { "epoch": 2.43, - "grad_norm": 6.246988773345947, + "grad_norm": 30.6466007232666, "learning_rate": 1.5139930298869998e-05, - "loss": 0.8582, + "loss": 1.3613, "step": 23010 }, { "epoch": 2.43, - "grad_norm": 17.485515594482422, + "grad_norm": 36.493186950683594, "learning_rate": 1.5137818143415357e-05, - "loss": 0.6811, + "loss": 1.2999, "step": 23020 }, { "epoch": 2.43, - "grad_norm": 16.079275131225586, + "grad_norm": 34.864776611328125, "learning_rate": 1.5135705987960715e-05, - "loss": 0.9419, + "loss": 1.342, "step": 23030 }, { "epoch": 2.43, - "grad_norm": 15.374100685119629, + "grad_norm": 28.123180389404297, "learning_rate": 1.5133593832506072e-05, - "loss": 0.6517, + "loss": 1.2368, "step": 23040 }, { "epoch": 2.43, - "grad_norm": 13.378589630126953, + "grad_norm": 6.719646453857422, "learning_rate": 1.5131481677051431e-05, - "loss": 0.6753, + "loss": 1.1538, "step": 23050 }, { "epoch": 2.44, - "grad_norm": 11.285813331604004, + "grad_norm": 43.141109466552734, "learning_rate": 1.512936952159679e-05, - "loss": 0.7628, + "loss": 1.3467, "step": 23060 }, { "epoch": 2.44, - "grad_norm": 2.107954502105713, + "grad_norm": 35.992881774902344, "learning_rate": 1.5127257366142149e-05, - "loss": 0.5832, + "loss": 1.3957, "step": 23070 }, { "epoch": 2.44, - "grad_norm": 17.614456176757812, + "grad_norm": 25.756790161132812, "learning_rate": 1.512514521068751e-05, - "loss": 0.5369, + "loss": 1.3809, "step": 23080 }, { "epoch": 2.44, - "grad_norm": 18.90216636657715, + "grad_norm": 49.89607238769531, "learning_rate": 1.5123033055232868e-05, - "loss": 0.6621, + "loss": 1.1688, "step": 23090 }, { "epoch": 2.44, - "grad_norm": 7.931809902191162, + "grad_norm": 27.954618453979492, "learning_rate": 1.5120920899778223e-05, - "loss": 0.6347, + "loss": 1.241, "step": 23100 }, { "epoch": 2.44, - "grad_norm": 12.385035514831543, + "grad_norm": 28.86982536315918, "learning_rate": 1.5118808744323584e-05, - "loss": 0.5613, + "loss": 1.1598, "step": 23110 }, { "epoch": 2.44, - "grad_norm": 17.66061782836914, + "grad_norm": 27.289958953857422, "learning_rate": 1.5116696588868943e-05, - "loss": 0.6973, + "loss": 1.0762, "step": 23120 }, { "epoch": 2.44, - "grad_norm": 26.619192123413086, + "grad_norm": 31.333444595336914, "learning_rate": 1.5114584433414301e-05, - "loss": 0.6432, + "loss": 1.2702, "step": 23130 }, { "epoch": 2.44, - "grad_norm": 3.9438562393188477, + "grad_norm": 37.6068229675293, "learning_rate": 1.511247227795966e-05, - "loss": 0.7237, + "loss": 1.3151, "step": 23140 }, { "epoch": 2.44, - "grad_norm": 13.911828994750977, + "grad_norm": 36.237586975097656, "learning_rate": 1.5110360122505017e-05, - "loss": 0.6988, + "loss": 1.2378, "step": 23150 }, { "epoch": 2.45, - "grad_norm": 17.548402786254883, + "grad_norm": 66.99047088623047, "learning_rate": 1.5108247967050376e-05, - "loss": 0.7412, + "loss": 1.5427, "step": 23160 }, { "epoch": 2.45, - "grad_norm": 0.6841170191764832, + "grad_norm": 30.97541618347168, "learning_rate": 1.5106135811595735e-05, - "loss": 0.6538, + "loss": 1.0811, "step": 23170 }, { "epoch": 2.45, - "grad_norm": 11.202994346618652, + "grad_norm": 36.51131057739258, "learning_rate": 1.5104023656141094e-05, - "loss": 0.8233, + "loss": 1.7663, "step": 23180 }, { "epoch": 2.45, - "grad_norm": 2.028456926345825, + "grad_norm": 38.033790588378906, "learning_rate": 1.5101911500686452e-05, - "loss": 0.4108, + "loss": 1.3246, "step": 23190 }, { "epoch": 2.45, - "grad_norm": 17.464309692382812, + "grad_norm": 39.502471923828125, "learning_rate": 1.509979934523181e-05, - "loss": 0.5908, + "loss": 1.2366, "step": 23200 }, { "epoch": 2.45, - "grad_norm": 4.917288303375244, + "grad_norm": 20.23288345336914, "learning_rate": 1.5097687189777168e-05, - "loss": 0.5287, + "loss": 1.3446, "step": 23210 }, { "epoch": 2.45, - "grad_norm": 18.816219329833984, + "grad_norm": 39.88277816772461, "learning_rate": 1.5095575034322527e-05, - "loss": 0.8245, + "loss": 1.5157, "step": 23220 }, { "epoch": 2.45, - "grad_norm": 1.7709016799926758, + "grad_norm": 39.46688461303711, "learning_rate": 1.5093462878867886e-05, - "loss": 0.6666, + "loss": 1.2719, "step": 23230 }, { "epoch": 2.45, - "grad_norm": 3.5497138500213623, + "grad_norm": 29.2645263671875, "learning_rate": 1.5091350723413245e-05, - "loss": 0.8426, + "loss": 1.2092, "step": 23240 }, { "epoch": 2.46, - "grad_norm": 0.4459506571292877, + "grad_norm": 32.53333282470703, "learning_rate": 1.5089238567958603e-05, - "loss": 0.7289, + "loss": 1.3902, "step": 23250 }, { "epoch": 2.46, - "grad_norm": 3.9968764781951904, + "grad_norm": 33.32743453979492, "learning_rate": 1.508712641250396e-05, - "loss": 0.3549, + "loss": 1.1676, "step": 23260 }, { "epoch": 2.46, - "grad_norm": 21.19850730895996, + "grad_norm": 46.76601028442383, "learning_rate": 1.508501425704932e-05, - "loss": 0.5979, + "loss": 1.2754, "step": 23270 }, { "epoch": 2.46, - "grad_norm": 15.503135681152344, + "grad_norm": 54.74008560180664, "learning_rate": 1.5082902101594678e-05, - "loss": 0.4127, + "loss": 1.2841, "step": 23280 }, { "epoch": 2.46, - "grad_norm": 2.7719767093658447, + "grad_norm": 31.88199234008789, "learning_rate": 1.5080789946140037e-05, - "loss": 0.8515, + "loss": 1.2018, "step": 23290 }, { "epoch": 2.46, - "grad_norm": 9.501763343811035, + "grad_norm": 40.59041213989258, "learning_rate": 1.5078677790685397e-05, - "loss": 0.6606, + "loss": 1.7745, "step": 23300 }, { "epoch": 2.46, - "grad_norm": 17.118879318237305, + "grad_norm": 47.92329788208008, "learning_rate": 1.5076565635230753e-05, - "loss": 0.5057, + "loss": 1.5045, "step": 23310 }, { "epoch": 2.46, - "grad_norm": 16.53435707092285, + "grad_norm": 45.55239486694336, "learning_rate": 1.5074453479776111e-05, - "loss": 0.3915, + "loss": 1.1718, "step": 23320 }, { "epoch": 2.46, - "grad_norm": 15.652314186096191, + "grad_norm": 64.67566680908203, "learning_rate": 1.5072341324321472e-05, - "loss": 0.8787, + "loss": 1.3683, "step": 23330 }, { "epoch": 2.46, - "grad_norm": 2.54408597946167, + "grad_norm": 33.53837966918945, "learning_rate": 1.507022916886683e-05, - "loss": 0.8413, + "loss": 1.4143, "step": 23340 }, { "epoch": 2.47, - "grad_norm": 13.989728927612305, + "grad_norm": 35.718780517578125, "learning_rate": 1.506811701341219e-05, - "loss": 0.4846, + "loss": 1.1464, "step": 23350 }, { "epoch": 2.47, - "grad_norm": 5.717529296875, + "grad_norm": 37.01331329345703, "learning_rate": 1.5066004857957548e-05, - "loss": 0.646, + "loss": 1.0981, "step": 23360 }, { "epoch": 2.47, - "grad_norm": 21.15603256225586, + "grad_norm": 36.55937576293945, "learning_rate": 1.5063892702502905e-05, - "loss": 0.6106, + "loss": 1.1176, "step": 23370 }, { "epoch": 2.47, - "grad_norm": 3.9289846420288086, + "grad_norm": 28.14556121826172, "learning_rate": 1.5061780547048264e-05, - "loss": 0.6686, + "loss": 1.2447, "step": 23380 }, { "epoch": 2.47, - "grad_norm": 19.378828048706055, + "grad_norm": 37.08237075805664, "learning_rate": 1.5059668391593623e-05, - "loss": 0.7889, + "loss": 1.1688, "step": 23390 }, { "epoch": 2.47, - "grad_norm": 32.95957946777344, + "grad_norm": 66.43830871582031, "learning_rate": 1.5057556236138982e-05, - "loss": 0.8177, + "loss": 1.2295, "step": 23400 }, { "epoch": 2.47, - "grad_norm": 4.776169300079346, + "grad_norm": 26.220623016357422, "learning_rate": 1.505544408068434e-05, - "loss": 0.6857, + "loss": 1.2288, "step": 23410 }, { "epoch": 2.47, - "grad_norm": 14.668253898620605, + "grad_norm": 41.779605865478516, "learning_rate": 1.5053331925229698e-05, - "loss": 0.7618, + "loss": 1.5979, "step": 23420 }, { "epoch": 2.47, - "grad_norm": 16.445924758911133, + "grad_norm": 45.27897644042969, "learning_rate": 1.5051219769775056e-05, - "loss": 0.4394, + "loss": 1.301, "step": 23430 }, { "epoch": 2.48, - "grad_norm": 3.8227810859680176, + "grad_norm": 18.910518646240234, "learning_rate": 1.5049107614320415e-05, - "loss": 0.5899, + "loss": 1.2203, "step": 23440 }, { "epoch": 2.48, - "grad_norm": 7.432600021362305, + "grad_norm": 45.0555305480957, "learning_rate": 1.5046995458865774e-05, - "loss": 0.6665, + "loss": 1.3365, "step": 23450 }, { "epoch": 2.48, - "grad_norm": 14.243486404418945, + "grad_norm": 38.645904541015625, "learning_rate": 1.5044883303411133e-05, - "loss": 0.8802, + "loss": 1.5012, "step": 23460 }, { "epoch": 2.48, - "grad_norm": 23.18899154663086, + "grad_norm": 36.011531829833984, "learning_rate": 1.504277114795649e-05, - "loss": 0.6919, + "loss": 1.2956, "step": 23470 }, { "epoch": 2.48, - "grad_norm": 24.051651000976562, + "grad_norm": 32.81689453125, "learning_rate": 1.5040658992501849e-05, - "loss": 0.8737, + "loss": 1.2841, "step": 23480 }, { "epoch": 2.48, - "grad_norm": 3.081942081451416, + "grad_norm": 29.339426040649414, "learning_rate": 1.5038546837047207e-05, - "loss": 0.6865, + "loss": 1.221, "step": 23490 }, { "epoch": 2.48, - "grad_norm": 13.114781379699707, + "grad_norm": 24.54115867614746, "learning_rate": 1.5036434681592566e-05, - "loss": 0.91, + "loss": 1.3467, "step": 23500 }, { "epoch": 2.48, - "grad_norm": 12.253195762634277, + "grad_norm": 33.82833480834961, "learning_rate": 1.5034322526137925e-05, - "loss": 0.6317, + "loss": 1.2745, "step": 23510 }, { "epoch": 2.48, - "grad_norm": 2.589900255203247, + "grad_norm": 53.73040008544922, "learning_rate": 1.5032210370683284e-05, - "loss": 0.6835, + "loss": 1.0935, "step": 23520 }, { "epoch": 2.48, - "grad_norm": 3.68949294090271, + "grad_norm": 21.1787109375, "learning_rate": 1.503009821522864e-05, - "loss": 0.6945, + "loss": 1.3787, "step": 23530 }, { "epoch": 2.49, - "grad_norm": 25.80329132080078, + "grad_norm": 33.13351058959961, "learning_rate": 1.5027986059774e-05, - "loss": 0.4527, + "loss": 1.1798, "step": 23540 }, { "epoch": 2.49, - "grad_norm": 2.7211833000183105, + "grad_norm": 43.01955032348633, "learning_rate": 1.5025873904319358e-05, - "loss": 0.8348, + "loss": 1.3958, "step": 23550 }, { "epoch": 2.49, - "grad_norm": 7.685608863830566, + "grad_norm": 36.075889587402344, "learning_rate": 1.5023761748864719e-05, - "loss": 0.5563, + "loss": 1.5074, "step": 23560 }, { "epoch": 2.49, - "grad_norm": 17.261001586914062, + "grad_norm": 19.283878326416016, "learning_rate": 1.5021649593410078e-05, - "loss": 0.7189, + "loss": 1.5012, "step": 23570 }, { "epoch": 2.49, - "grad_norm": 19.065845489501953, + "grad_norm": 38.799659729003906, "learning_rate": 1.5019537437955433e-05, - "loss": 0.486, + "loss": 1.3466, "step": 23580 }, { "epoch": 2.49, - "grad_norm": 2.9216840267181396, + "grad_norm": 43.45872497558594, "learning_rate": 1.5017425282500793e-05, - "loss": 0.782, + "loss": 1.5418, "step": 23590 }, { "epoch": 2.49, - "grad_norm": 6.580178260803223, + "grad_norm": 32.79587173461914, "learning_rate": 1.5015313127046152e-05, - "loss": 0.5096, + "loss": 1.3402, "step": 23600 }, { "epoch": 2.49, - "grad_norm": 16.149261474609375, + "grad_norm": 40.973323822021484, "learning_rate": 1.5013200971591511e-05, - "loss": 0.6753, + "loss": 1.1039, "step": 23610 }, { "epoch": 2.49, - "grad_norm": 21.928892135620117, + "grad_norm": 60.65894317626953, "learning_rate": 1.501108881613687e-05, - "loss": 0.6885, + "loss": 1.5983, "step": 23620 }, { "epoch": 2.5, - "grad_norm": 2.677056074142456, + "grad_norm": 15.117423057556152, "learning_rate": 1.5008976660682229e-05, - "loss": 0.7205, + "loss": 1.1755, "step": 23630 }, { "epoch": 2.5, - "grad_norm": 9.722269058227539, + "grad_norm": 31.318918228149414, "learning_rate": 1.5006864505227586e-05, - "loss": 0.5541, + "loss": 1.0976, "step": 23640 }, { "epoch": 2.5, - "grad_norm": 21.108728408813477, + "grad_norm": 53.03226852416992, "learning_rate": 1.5004752349772944e-05, - "loss": 0.7354, + "loss": 1.4546, "step": 23650 }, { "epoch": 2.5, - "grad_norm": 30.18218421936035, + "grad_norm": 32.93924331665039, "learning_rate": 1.5002640194318303e-05, - "loss": 0.6644, + "loss": 1.3014, "step": 23660 }, { "epoch": 2.5, - "grad_norm": 14.145009994506836, + "grad_norm": 50.15010070800781, "learning_rate": 1.5000528038863662e-05, - "loss": 0.9854, + "loss": 1.8675, "step": 23670 }, { "epoch": 2.5, - "grad_norm": 17.645483016967773, + "grad_norm": 51.94035339355469, "learning_rate": 1.499841588340902e-05, - "loss": 0.8468, + "loss": 1.2349, "step": 23680 }, { "epoch": 2.5, - "grad_norm": 3.638253927230835, + "grad_norm": 26.45070457458496, "learning_rate": 1.4996303727954378e-05, - "loss": 0.6082, + "loss": 1.1024, "step": 23690 }, { "epoch": 2.5, - "grad_norm": 9.384114265441895, + "grad_norm": 34.48675537109375, "learning_rate": 1.4994191572499737e-05, - "loss": 0.8901, + "loss": 1.4873, "step": 23700 }, { "epoch": 2.5, - "grad_norm": 5.1335930824279785, + "grad_norm": 35.962528228759766, "learning_rate": 1.4992079417045095e-05, - "loss": 0.7724, + "loss": 1.176, "step": 23710 }, { "epoch": 2.51, - "grad_norm": 17.09876251220703, + "grad_norm": 39.80942153930664, "learning_rate": 1.4989967261590454e-05, - "loss": 0.4516, + "loss": 1.3768, "step": 23720 }, { "epoch": 2.51, - "grad_norm": 13.628138542175293, + "grad_norm": 34.4463996887207, "learning_rate": 1.4987855106135813e-05, - "loss": 0.8048, + "loss": 1.3789, "step": 23730 }, { "epoch": 2.51, - "grad_norm": 7.180805683135986, + "grad_norm": 38.388389587402344, "learning_rate": 1.4985742950681172e-05, - "loss": 0.5347, + "loss": 1.3305, "step": 23740 }, { "epoch": 2.51, - "grad_norm": 28.3292236328125, + "grad_norm": 29.320444107055664, "learning_rate": 1.4983630795226529e-05, - "loss": 0.4471, + "loss": 1.0368, "step": 23750 }, { "epoch": 2.51, - "grad_norm": 18.162731170654297, + "grad_norm": 40.09894561767578, "learning_rate": 1.4981518639771888e-05, - "loss": 0.7638, + "loss": 1.3056, "step": 23760 }, { "epoch": 2.51, - "grad_norm": 7.131850242614746, + "grad_norm": 53.45796203613281, "learning_rate": 1.4979406484317246e-05, - "loss": 0.3124, + "loss": 1.1026, "step": 23770 }, { "epoch": 2.51, - "grad_norm": 21.819936752319336, + "grad_norm": 36.283233642578125, "learning_rate": 1.4977294328862607e-05, - "loss": 0.5956, + "loss": 1.2227, "step": 23780 }, { "epoch": 2.51, - "grad_norm": 19.608869552612305, + "grad_norm": 51.99909591674805, "learning_rate": 1.4975182173407966e-05, - "loss": 0.6718, + "loss": 1.2495, "step": 23790 }, { "epoch": 2.51, - "grad_norm": 10.543989181518555, + "grad_norm": 34.56658172607422, "learning_rate": 1.4973070017953321e-05, - "loss": 0.6101, + "loss": 1.4789, "step": 23800 }, { "epoch": 2.51, - "grad_norm": 13.283591270446777, + "grad_norm": 24.967601776123047, "learning_rate": 1.4970957862498681e-05, - "loss": 0.4347, + "loss": 1.2663, "step": 23810 }, { "epoch": 2.52, - "grad_norm": 11.984816551208496, + "grad_norm": 50.21664047241211, "learning_rate": 1.496884570704404e-05, - "loss": 0.8313, + "loss": 1.3711, "step": 23820 }, { "epoch": 2.52, - "grad_norm": 16.5617733001709, + "grad_norm": 32.0864143371582, "learning_rate": 1.4966733551589399e-05, - "loss": 0.5954, + "loss": 1.1734, "step": 23830 }, { "epoch": 2.52, - "grad_norm": 12.347994804382324, + "grad_norm": 33.78178024291992, "learning_rate": 1.4964621396134758e-05, - "loss": 0.5528, + "loss": 1.2585, "step": 23840 }, { "epoch": 2.52, - "grad_norm": 13.640240669250488, + "grad_norm": 36.418296813964844, "learning_rate": 1.4962509240680115e-05, - "loss": 0.7084, + "loss": 1.4456, "step": 23850 }, { "epoch": 2.52, - "grad_norm": 22.098560333251953, + "grad_norm": 51.01276397705078, "learning_rate": 1.4960397085225474e-05, - "loss": 0.5105, + "loss": 1.3891, "step": 23860 }, { "epoch": 2.52, - "grad_norm": 15.362188339233398, + "grad_norm": 42.03109359741211, "learning_rate": 1.4958284929770832e-05, - "loss": 0.4182, + "loss": 1.0347, "step": 23870 }, { "epoch": 2.52, - "grad_norm": 21.51960563659668, + "grad_norm": 36.432682037353516, "learning_rate": 1.4956172774316191e-05, - "loss": 0.6004, + "loss": 1.2346, "step": 23880 }, { "epoch": 2.52, - "grad_norm": 8.463544845581055, + "grad_norm": 49.72010803222656, "learning_rate": 1.495406061886155e-05, - "loss": 0.8138, + "loss": 1.7036, "step": 23890 }, { "epoch": 2.52, - "grad_norm": 15.716696739196777, + "grad_norm": 27.580890655517578, "learning_rate": 1.4951948463406909e-05, - "loss": 0.4203, + "loss": 1.2062, "step": 23900 }, { "epoch": 2.53, - "grad_norm": 23.614124298095703, + "grad_norm": 24.04772186279297, "learning_rate": 1.4949836307952266e-05, - "loss": 0.3227, + "loss": 0.9503, "step": 23910 }, { "epoch": 2.53, - "grad_norm": 19.844987869262695, + "grad_norm": 35.753814697265625, "learning_rate": 1.4947724152497625e-05, - "loss": 0.4505, + "loss": 1.0504, "step": 23920 }, { "epoch": 2.53, - "grad_norm": 2.2555792331695557, + "grad_norm": 31.400014877319336, "learning_rate": 1.4945611997042983e-05, - "loss": 0.5621, + "loss": 1.1353, "step": 23930 }, { "epoch": 2.53, - "grad_norm": 11.75871753692627, + "grad_norm": 42.052425384521484, "learning_rate": 1.4943499841588342e-05, - "loss": 0.6506, + "loss": 1.3491, "step": 23940 }, { "epoch": 2.53, - "grad_norm": 16.958911895751953, + "grad_norm": 49.85698699951172, "learning_rate": 1.4941387686133701e-05, - "loss": 0.4437, + "loss": 1.38, "step": 23950 }, { "epoch": 2.53, - "grad_norm": 0.8014586567878723, + "grad_norm": 38.82079315185547, "learning_rate": 1.4939275530679058e-05, - "loss": 0.4164, + "loss": 1.1003, "step": 23960 }, { "epoch": 2.53, - "grad_norm": 6.798696994781494, + "grad_norm": 48.480716705322266, "learning_rate": 1.4937163375224417e-05, - "loss": 0.5556, + "loss": 1.0941, "step": 23970 }, { "epoch": 2.53, - "grad_norm": 5.9725117683410645, + "grad_norm": 27.51656150817871, "learning_rate": 1.4935051219769776e-05, - "loss": 0.9236, + "loss": 1.1066, "step": 23980 }, { "epoch": 2.53, - "grad_norm": 30.354148864746094, + "grad_norm": 29.884326934814453, "learning_rate": 1.4932939064315134e-05, - "loss": 0.9121, + "loss": 1.5131, "step": 23990 }, { "epoch": 2.53, - "grad_norm": 5.852388381958008, + "grad_norm": 36.97277069091797, "learning_rate": 1.4930826908860493e-05, - "loss": 0.5221, + "loss": 1.3745, "step": 24000 }, { "epoch": 2.54, - "grad_norm": 16.212289810180664, + "grad_norm": 36.51731491088867, "learning_rate": 1.4928714753405854e-05, - "loss": 0.7031, + "loss": 1.6993, "step": 24010 }, { "epoch": 2.54, - "grad_norm": 16.306909561157227, + "grad_norm": 39.29604721069336, "learning_rate": 1.4926602597951209e-05, - "loss": 0.5507, + "loss": 1.2195, "step": 24020 }, { "epoch": 2.54, - "grad_norm": 11.636151313781738, + "grad_norm": 49.15299987792969, "learning_rate": 1.4924490442496568e-05, - "loss": 0.7573, + "loss": 1.4115, "step": 24030 }, { "epoch": 2.54, - "grad_norm": 2.8181538581848145, + "grad_norm": 45.842529296875, "learning_rate": 1.4922378287041928e-05, - "loss": 0.4731, + "loss": 1.4832, "step": 24040 }, { "epoch": 2.54, - "grad_norm": 8.614269256591797, + "grad_norm": 28.991668701171875, "learning_rate": 1.4920266131587287e-05, - "loss": 0.6303, + "loss": 1.3839, "step": 24050 }, { "epoch": 2.54, - "grad_norm": 15.304798126220703, + "grad_norm": 32.39551544189453, "learning_rate": 1.4918153976132646e-05, - "loss": 0.9199, + "loss": 1.4279, "step": 24060 }, { "epoch": 2.54, - "grad_norm": 31.638011932373047, + "grad_norm": 33.703060150146484, "learning_rate": 1.4916041820678003e-05, - "loss": 0.8142, + "loss": 1.5044, "step": 24070 }, { "epoch": 2.54, - "grad_norm": 5.186000823974609, + "grad_norm": 36.138916015625, "learning_rate": 1.4913929665223362e-05, - "loss": 0.8743, + "loss": 1.8044, "step": 24080 }, { "epoch": 2.54, - "grad_norm": 23.94352912902832, + "grad_norm": 44.58909225463867, "learning_rate": 1.491181750976872e-05, - "loss": 0.6822, + "loss": 1.2476, "step": 24090 }, { "epoch": 2.55, - "grad_norm": 4.665832996368408, + "grad_norm": 26.193584442138672, "learning_rate": 1.490970535431408e-05, - "loss": 0.6646, + "loss": 1.1078, "step": 24100 }, { "epoch": 2.55, - "grad_norm": 7.559071063995361, + "grad_norm": 20.084915161132812, "learning_rate": 1.4907593198859438e-05, - "loss": 0.7028, + "loss": 1.335, "step": 24110 }, { "epoch": 2.55, - "grad_norm": 2.4393956661224365, + "grad_norm": 28.900842666625977, "learning_rate": 1.4905481043404795e-05, - "loss": 0.4703, + "loss": 1.4928, "step": 24120 }, { "epoch": 2.55, - "grad_norm": 8.807246208190918, + "grad_norm": 33.33913040161133, "learning_rate": 1.4903368887950154e-05, - "loss": 0.5901, + "loss": 1.4151, "step": 24130 }, { "epoch": 2.55, - "grad_norm": 15.984193801879883, + "grad_norm": 35.128849029541016, "learning_rate": 1.4901256732495513e-05, - "loss": 0.5339, + "loss": 0.9692, "step": 24140 }, { "epoch": 2.55, - "grad_norm": 0.3053434193134308, + "grad_norm": 30.123186111450195, "learning_rate": 1.4899144577040871e-05, - "loss": 0.8488, + "loss": 1.4068, "step": 24150 }, { "epoch": 2.55, - "grad_norm": 10.944032669067383, + "grad_norm": 39.259830474853516, "learning_rate": 1.489703242158623e-05, - "loss": 0.8069, + "loss": 1.2735, "step": 24160 }, { "epoch": 2.55, - "grad_norm": 29.209674835205078, + "grad_norm": 44.726436614990234, "learning_rate": 1.4894920266131589e-05, - "loss": 0.967, + "loss": 1.2897, "step": 24170 }, { "epoch": 2.55, - "grad_norm": 6.288844108581543, + "grad_norm": 44.2175407409668, "learning_rate": 1.4892808110676946e-05, - "loss": 0.6373, + "loss": 1.0493, "step": 24180 }, { "epoch": 2.55, - "grad_norm": 9.915877342224121, + "grad_norm": 53.66617202758789, "learning_rate": 1.4890695955222305e-05, - "loss": 0.5086, + "loss": 1.5806, "step": 24190 }, { "epoch": 2.56, - "grad_norm": 14.256426811218262, + "grad_norm": 33.99443817138672, "learning_rate": 1.4888583799767664e-05, - "loss": 0.6836, + "loss": 1.5604, "step": 24200 }, { "epoch": 2.56, - "grad_norm": 22.889949798583984, + "grad_norm": 41.43342590332031, "learning_rate": 1.4886471644313022e-05, - "loss": 0.4784, + "loss": 1.1515, "step": 24210 }, { "epoch": 2.56, - "grad_norm": 13.99036693572998, + "grad_norm": 32.45558547973633, "learning_rate": 1.4884359488858381e-05, - "loss": 0.5432, + "loss": 1.2933, "step": 24220 }, { "epoch": 2.56, - "grad_norm": 0.6046372652053833, + "grad_norm": 51.926429748535156, "learning_rate": 1.4882247333403738e-05, - "loss": 0.6824, + "loss": 1.4336, "step": 24230 }, { "epoch": 2.56, - "grad_norm": 0.47681859135627747, + "grad_norm": 23.008644104003906, "learning_rate": 1.4880135177949097e-05, - "loss": 0.7134, + "loss": 1.1842, "step": 24240 }, { "epoch": 2.56, - "grad_norm": 22.639646530151367, + "grad_norm": 36.788143157958984, "learning_rate": 1.4878023022494456e-05, - "loss": 0.7903, + "loss": 1.3912, "step": 24250 }, { "epoch": 2.56, - "grad_norm": 17.630769729614258, + "grad_norm": 50.616661071777344, "learning_rate": 1.4875910867039816e-05, - "loss": 0.4052, + "loss": 1.3089, "step": 24260 }, { "epoch": 2.56, - "grad_norm": 11.076565742492676, + "grad_norm": 41.27164840698242, "learning_rate": 1.4873798711585175e-05, - "loss": 0.7574, + "loss": 1.4645, "step": 24270 }, { "epoch": 2.56, - "grad_norm": 26.88067626953125, + "grad_norm": 34.61843490600586, "learning_rate": 1.4871686556130534e-05, - "loss": 0.6225, + "loss": 1.3119, "step": 24280 }, { "epoch": 2.57, - "grad_norm": 8.765049934387207, + "grad_norm": 56.751251220703125, "learning_rate": 1.4869574400675891e-05, - "loss": 0.8033, + "loss": 1.5971, "step": 24290 }, { "epoch": 2.57, - "grad_norm": 11.438042640686035, + "grad_norm": 40.20285415649414, "learning_rate": 1.486746224522125e-05, - "loss": 0.5842, + "loss": 1.3249, "step": 24300 }, { "epoch": 2.57, - "grad_norm": 18.12037467956543, + "grad_norm": 30.57199478149414, "learning_rate": 1.4865350089766609e-05, - "loss": 0.7791, + "loss": 1.6062, "step": 24310 }, { "epoch": 2.57, - "grad_norm": 16.559932708740234, + "grad_norm": 22.989137649536133, "learning_rate": 1.4863237934311967e-05, - "loss": 0.5902, + "loss": 1.2599, "step": 24320 }, { "epoch": 2.57, - "grad_norm": 17.39402198791504, + "grad_norm": 36.78236770629883, "learning_rate": 1.4861125778857326e-05, - "loss": 0.6109, + "loss": 1.2611, "step": 24330 }, { "epoch": 2.57, - "grad_norm": 10.26382827758789, + "grad_norm": 30.47361946105957, "learning_rate": 1.4859013623402683e-05, - "loss": 0.647, + "loss": 1.095, "step": 24340 }, { "epoch": 2.57, - "grad_norm": 18.5225830078125, + "grad_norm": 41.987388610839844, "learning_rate": 1.4856901467948042e-05, - "loss": 0.7729, + "loss": 1.3429, "step": 24350 }, { "epoch": 2.57, - "grad_norm": 12.87230396270752, + "grad_norm": 35.67143249511719, "learning_rate": 1.48547893124934e-05, - "loss": 0.5773, + "loss": 1.2909, "step": 24360 }, { "epoch": 2.57, - "grad_norm": 17.26911735534668, + "grad_norm": 39.25424575805664, "learning_rate": 1.485267715703876e-05, - "loss": 0.8933, + "loss": 1.4991, "step": 24370 }, { "epoch": 2.57, - "grad_norm": 27.636024475097656, + "grad_norm": 38.31764221191406, "learning_rate": 1.4850565001584118e-05, - "loss": 0.6887, + "loss": 1.4496, "step": 24380 }, { "epoch": 2.58, - "grad_norm": 19.012195587158203, + "grad_norm": 50.357364654541016, "learning_rate": 1.4848452846129475e-05, - "loss": 0.8109, + "loss": 1.5112, "step": 24390 }, { "epoch": 2.58, - "grad_norm": 28.778249740600586, + "grad_norm": 43.21502685546875, "learning_rate": 1.4846340690674834e-05, - "loss": 0.685, + "loss": 1.213, "step": 24400 }, { "epoch": 2.58, - "grad_norm": 6.839900016784668, + "grad_norm": 26.684919357299805, "learning_rate": 1.4844228535220193e-05, - "loss": 0.4761, + "loss": 1.3043, "step": 24410 }, { "epoch": 2.58, - "grad_norm": 14.871298789978027, + "grad_norm": 23.66108512878418, "learning_rate": 1.4842116379765552e-05, - "loss": 0.5211, + "loss": 1.2327, "step": 24420 }, { "epoch": 2.58, - "grad_norm": 9.687091827392578, + "grad_norm": 23.78837013244629, "learning_rate": 1.484000422431091e-05, - "loss": 0.5061, + "loss": 1.0228, "step": 24430 }, { "epoch": 2.58, - "grad_norm": 3.4191861152648926, + "grad_norm": 33.640663146972656, "learning_rate": 1.483789206885627e-05, - "loss": 0.6988, + "loss": 1.3206, "step": 24440 }, { "epoch": 2.58, - "grad_norm": 19.068683624267578, + "grad_norm": 39.708316802978516, "learning_rate": 1.4835779913401626e-05, - "loss": 0.5729, + "loss": 0.8895, "step": 24450 }, { "epoch": 2.58, - "grad_norm": 20.238061904907227, + "grad_norm": 32.97872543334961, "learning_rate": 1.4833667757946985e-05, - "loss": 0.8842, + "loss": 1.4857, "step": 24460 }, { "epoch": 2.58, - "grad_norm": 16.91153907775879, + "grad_norm": 26.29220199584961, "learning_rate": 1.4831555602492344e-05, - "loss": 0.6858, + "loss": 1.4738, "step": 24470 }, { "epoch": 2.59, - "grad_norm": 14.929586410522461, + "grad_norm": 50.51658248901367, "learning_rate": 1.4829443447037703e-05, - "loss": 0.4492, + "loss": 1.4794, "step": 24480 }, { "epoch": 2.59, - "grad_norm": 13.602909088134766, + "grad_norm": 27.09004020690918, "learning_rate": 1.4827331291583063e-05, - "loss": 0.6971, + "loss": 1.3952, "step": 24490 }, { "epoch": 2.59, - "grad_norm": 10.34124755859375, + "grad_norm": 41.47471237182617, "learning_rate": 1.4825219136128419e-05, - "loss": 0.7523, + "loss": 1.6319, "step": 24500 }, { "epoch": 2.59, - "grad_norm": 2.199103832244873, + "grad_norm": 40.83889389038086, "learning_rate": 1.4823106980673777e-05, - "loss": 0.6773, + "loss": 1.3702, "step": 24510 }, { "epoch": 2.59, - "grad_norm": 18.768386840820312, + "grad_norm": 38.318275451660156, "learning_rate": 1.4820994825219138e-05, - "loss": 0.6528, + "loss": 1.2554, "step": 24520 }, { "epoch": 2.59, - "grad_norm": 12.458382606506348, + "grad_norm": 39.03107833862305, "learning_rate": 1.4818882669764497e-05, - "loss": 0.6579, + "loss": 1.4623, "step": 24530 }, { "epoch": 2.59, - "grad_norm": 12.95892333984375, + "grad_norm": 13.155189514160156, "learning_rate": 1.4816770514309855e-05, - "loss": 0.6654, + "loss": 1.1579, "step": 24540 }, { "epoch": 2.59, - "grad_norm": 22.648630142211914, + "grad_norm": 31.942848205566406, "learning_rate": 1.4814658358855214e-05, - "loss": 0.5313, + "loss": 0.8849, "step": 24550 }, { "epoch": 2.59, - "grad_norm": 2.795624017715454, + "grad_norm": 53.04624557495117, "learning_rate": 1.4812546203400571e-05, - "loss": 0.6958, + "loss": 1.4308, "step": 24560 }, { "epoch": 2.59, - "grad_norm": 11.143491744995117, + "grad_norm": 30.448989868164062, "learning_rate": 1.481043404794593e-05, - "loss": 0.6691, + "loss": 1.3773, "step": 24570 }, { "epoch": 2.6, - "grad_norm": 7.667626857757568, + "grad_norm": 47.1045036315918, "learning_rate": 1.4808321892491289e-05, - "loss": 0.9705, + "loss": 1.524, "step": 24580 }, { "epoch": 2.6, - "grad_norm": 15.15952205657959, + "grad_norm": 48.024112701416016, "learning_rate": 1.4806209737036648e-05, - "loss": 0.8342, + "loss": 1.4027, "step": 24590 }, { "epoch": 2.6, - "grad_norm": 14.890870094299316, + "grad_norm": 35.02663803100586, "learning_rate": 1.4804097581582006e-05, - "loss": 0.5803, + "loss": 1.4169, "step": 24600 }, { "epoch": 2.6, - "grad_norm": 2.1391496658325195, + "grad_norm": 31.85002899169922, "learning_rate": 1.4801985426127363e-05, - "loss": 0.7064, + "loss": 1.6463, "step": 24610 }, { "epoch": 2.6, - "grad_norm": 11.376192092895508, + "grad_norm": 34.672332763671875, "learning_rate": 1.4799873270672722e-05, - "loss": 0.672, + "loss": 1.3304, "step": 24620 }, { "epoch": 2.6, - "grad_norm": 18.638967514038086, + "grad_norm": 33.87903594970703, "learning_rate": 1.4797761115218081e-05, - "loss": 0.8481, + "loss": 1.5978, "step": 24630 }, { "epoch": 2.6, - "grad_norm": 5.495625019073486, + "grad_norm": 37.58328628540039, "learning_rate": 1.479564895976344e-05, - "loss": 0.5108, + "loss": 1.1738, "step": 24640 }, { "epoch": 2.6, - "grad_norm": 13.570967674255371, + "grad_norm": 40.21030807495117, "learning_rate": 1.4793536804308799e-05, - "loss": 1.1026, + "loss": 1.8189, "step": 24650 }, { "epoch": 2.6, - "grad_norm": 7.7852396965026855, + "grad_norm": 43.158409118652344, "learning_rate": 1.4791424648854157e-05, - "loss": 0.5856, + "loss": 1.5301, "step": 24660 }, { "epoch": 2.61, - "grad_norm": 21.83995246887207, + "grad_norm": 23.55770492553711, "learning_rate": 1.4789312493399514e-05, - "loss": 0.8171, + "loss": 1.1481, "step": 24670 }, { "epoch": 2.61, - "grad_norm": 12.04238224029541, + "grad_norm": 44.40102005004883, "learning_rate": 1.4787200337944873e-05, - "loss": 0.5737, + "loss": 1.1994, "step": 24680 }, { "epoch": 2.61, - "grad_norm": 14.202824592590332, + "grad_norm": 49.14374542236328, "learning_rate": 1.4785088182490232e-05, - "loss": 0.5282, + "loss": 1.6241, "step": 24690 }, { "epoch": 2.61, - "grad_norm": 9.033062934875488, + "grad_norm": 37.159339904785156, "learning_rate": 1.478297602703559e-05, - "loss": 0.8494, + "loss": 1.4704, "step": 24700 }, { "epoch": 2.61, - "grad_norm": 5.109563827514648, + "grad_norm": 40.18320083618164, "learning_rate": 1.4780863871580951e-05, - "loss": 0.8726, + "loss": 1.3867, "step": 24710 }, { "epoch": 2.61, - "grad_norm": 7.4094109535217285, + "grad_norm": 40.54684066772461, "learning_rate": 1.4778751716126307e-05, - "loss": 0.7147, + "loss": 1.3172, "step": 24720 }, { "epoch": 2.61, - "grad_norm": 29.987689971923828, + "grad_norm": 43.983154296875, "learning_rate": 1.4776639560671665e-05, - "loss": 0.7816, + "loss": 1.4027, "step": 24730 }, { "epoch": 2.61, - "grad_norm": 12.265345573425293, + "grad_norm": 25.00552749633789, "learning_rate": 1.4774527405217026e-05, - "loss": 0.8554, + "loss": 1.2811, "step": 24740 }, { "epoch": 2.61, - "grad_norm": 4.0600972175598145, + "grad_norm": 54.91749954223633, "learning_rate": 1.4772415249762385e-05, - "loss": 0.6318, + "loss": 1.3405, "step": 24750 }, { "epoch": 2.61, - "grad_norm": 2.7878060340881348, + "grad_norm": 25.354337692260742, "learning_rate": 1.4770303094307743e-05, - "loss": 0.5371, + "loss": 1.0851, "step": 24760 }, { "epoch": 2.62, - "grad_norm": 18.72551918029785, + "grad_norm": 35.745479583740234, "learning_rate": 1.47681909388531e-05, - "loss": 0.7892, + "loss": 1.2975, "step": 24770 }, { "epoch": 2.62, - "grad_norm": 10.545355796813965, + "grad_norm": 53.73564147949219, "learning_rate": 1.476607878339846e-05, - "loss": 0.7766, + "loss": 1.3031, "step": 24780 }, { "epoch": 2.62, - "grad_norm": 10.314589500427246, + "grad_norm": 55.2784423828125, "learning_rate": 1.4763966627943818e-05, - "loss": 0.6722, + "loss": 1.2686, "step": 24790 }, { "epoch": 2.62, - "grad_norm": 17.951709747314453, + "grad_norm": 42.49158477783203, "learning_rate": 1.4761854472489177e-05, - "loss": 0.6656, + "loss": 1.5212, "step": 24800 }, { "epoch": 2.62, - "grad_norm": 21.45174217224121, + "grad_norm": 45.74517059326172, "learning_rate": 1.4759742317034536e-05, - "loss": 0.7536, + "loss": 1.3518, "step": 24810 }, { "epoch": 2.62, - "grad_norm": 16.14288330078125, + "grad_norm": 59.38320541381836, "learning_rate": 1.4757630161579894e-05, - "loss": 0.7745, + "loss": 1.3428, "step": 24820 }, { "epoch": 2.62, - "grad_norm": 15.320282936096191, + "grad_norm": 48.47743606567383, "learning_rate": 1.4755518006125252e-05, - "loss": 0.474, + "loss": 1.5088, "step": 24830 }, { "epoch": 2.62, - "grad_norm": 15.086441040039062, + "grad_norm": 47.659889221191406, "learning_rate": 1.475340585067061e-05, - "loss": 0.5355, + "loss": 1.1478, "step": 24840 }, { "epoch": 2.62, - "grad_norm": 8.719254493713379, + "grad_norm": 19.809249877929688, "learning_rate": 1.4751293695215969e-05, - "loss": 0.5028, + "loss": 1.1686, "step": 24850 }, { "epoch": 2.63, - "grad_norm": 8.045601844787598, + "grad_norm": 47.032798767089844, "learning_rate": 1.4749181539761328e-05, - "loss": 0.4431, + "loss": 1.3226, "step": 24860 }, { "epoch": 2.63, - "grad_norm": 26.209049224853516, + "grad_norm": 53.9599494934082, "learning_rate": 1.4747069384306687e-05, - "loss": 0.7608, + "loss": 1.5626, "step": 24870 }, { "epoch": 2.63, - "grad_norm": 19.31146812438965, + "grad_norm": 41.972660064697266, "learning_rate": 1.4744957228852044e-05, - "loss": 0.5729, + "loss": 1.1207, "step": 24880 }, { "epoch": 2.63, - "grad_norm": 21.93781852722168, + "grad_norm": 54.26142501831055, "learning_rate": 1.4742845073397402e-05, - "loss": 0.7176, + "loss": 1.6468, "step": 24890 }, { "epoch": 2.63, - "grad_norm": 4.79754638671875, + "grad_norm": 43.75844192504883, "learning_rate": 1.4740732917942761e-05, - "loss": 0.9103, + "loss": 1.5045, "step": 24900 }, { "epoch": 2.63, - "grad_norm": 35.30495071411133, + "grad_norm": 34.59361267089844, "learning_rate": 1.473862076248812e-05, - "loss": 0.4491, + "loss": 1.0276, "step": 24910 }, { "epoch": 2.63, - "grad_norm": 8.865439414978027, + "grad_norm": 41.5710563659668, "learning_rate": 1.4736508607033479e-05, - "loss": 0.7299, + "loss": 1.0876, "step": 24920 }, { "epoch": 2.63, - "grad_norm": 27.35230255126953, + "grad_norm": 25.408506393432617, "learning_rate": 1.473439645157884e-05, - "loss": 0.6803, + "loss": 1.3103, "step": 24930 }, { "epoch": 2.63, - "grad_norm": 5.544764995574951, + "grad_norm": 38.6860466003418, "learning_rate": 1.4732284296124195e-05, - "loss": 0.6007, + "loss": 1.3043, "step": 24940 }, { "epoch": 2.63, - "grad_norm": 14.554699897766113, + "grad_norm": 23.17441177368164, "learning_rate": 1.4730172140669553e-05, - "loss": 0.8551, + "loss": 1.6357, "step": 24950 }, { "epoch": 2.64, - "grad_norm": 7.415411472320557, + "grad_norm": 36.5379524230957, "learning_rate": 1.4728059985214914e-05, - "loss": 0.6687, + "loss": 1.3214, "step": 24960 }, { "epoch": 2.64, - "grad_norm": 9.470914840698242, + "grad_norm": 29.121536254882812, "learning_rate": 1.4725947829760273e-05, - "loss": 0.6708, + "loss": 1.1653, "step": 24970 }, { "epoch": 2.64, - "grad_norm": 11.0525484085083, + "grad_norm": 44.584388732910156, "learning_rate": 1.4723835674305631e-05, - "loss": 0.3324, + "loss": 1.2178, "step": 24980 }, { "epoch": 2.64, - "grad_norm": 39.93579864501953, + "grad_norm": 43.3303108215332, "learning_rate": 1.4721723518850987e-05, - "loss": 0.3721, + "loss": 1.395, "step": 24990 }, { "epoch": 2.64, - "grad_norm": 9.474876403808594, + "grad_norm": 39.68447494506836, "learning_rate": 1.4719611363396347e-05, - "loss": 0.6364, + "loss": 1.5108, "step": 25000 }, { "epoch": 2.64, - "grad_norm": 16.554121017456055, + "grad_norm": 43.774803161621094, "learning_rate": 1.4717499207941706e-05, - "loss": 0.5595, + "loss": 1.12, "step": 25010 }, { "epoch": 2.64, - "grad_norm": 4.960517883300781, + "grad_norm": 35.37923049926758, "learning_rate": 1.4715387052487065e-05, - "loss": 0.6897, + "loss": 0.9812, "step": 25020 }, { "epoch": 2.64, - "grad_norm": 29.915416717529297, + "grad_norm": 52.88051223754883, "learning_rate": 1.4713274897032424e-05, - "loss": 0.6156, + "loss": 1.402, "step": 25030 }, { "epoch": 2.64, - "grad_norm": 11.927760124206543, + "grad_norm": 28.77840232849121, "learning_rate": 1.471116274157778e-05, - "loss": 0.614, + "loss": 0.885, "step": 25040 }, { "epoch": 2.65, - "grad_norm": 19.267498016357422, + "grad_norm": 31.035680770874023, "learning_rate": 1.470905058612314e-05, - "loss": 0.444, + "loss": 1.362, "step": 25050 }, { "epoch": 2.65, - "grad_norm": 5.5006103515625, + "grad_norm": 24.169527053833008, "learning_rate": 1.4706938430668498e-05, - "loss": 0.5964, + "loss": 1.3848, "step": 25060 }, { "epoch": 2.65, - "grad_norm": 10.230973243713379, + "grad_norm": 31.358009338378906, "learning_rate": 1.4704826275213857e-05, - "loss": 0.4662, + "loss": 1.0211, "step": 25070 }, { "epoch": 2.65, - "grad_norm": 38.31246566772461, + "grad_norm": 34.02757263183594, "learning_rate": 1.4702714119759216e-05, - "loss": 0.6353, + "loss": 1.4698, "step": 25080 }, { "epoch": 2.65, - "grad_norm": 1.93467378616333, + "grad_norm": 28.2695369720459, "learning_rate": 1.4700601964304575e-05, - "loss": 0.5638, + "loss": 1.3542, "step": 25090 }, { "epoch": 2.65, - "grad_norm": 15.624272346496582, + "grad_norm": 39.833290100097656, "learning_rate": 1.4698489808849932e-05, - "loss": 0.6671, + "loss": 1.3939, "step": 25100 }, { "epoch": 2.65, - "grad_norm": 13.270512580871582, + "grad_norm": 33.89259338378906, "learning_rate": 1.469637765339529e-05, - "loss": 0.6738, + "loss": 1.3775, "step": 25110 }, { "epoch": 2.65, - "grad_norm": 10.201784133911133, + "grad_norm": 34.53770065307617, "learning_rate": 1.469426549794065e-05, - "loss": 0.5971, + "loss": 1.3927, "step": 25120 }, { "epoch": 2.65, - "grad_norm": 20.83800506591797, + "grad_norm": 47.44340515136719, "learning_rate": 1.4692153342486008e-05, - "loss": 0.6706, + "loss": 1.2768, "step": 25130 }, { "epoch": 2.65, - "grad_norm": 17.19257354736328, + "grad_norm": 41.171844482421875, "learning_rate": 1.4690041187031367e-05, - "loss": 0.6404, + "loss": 1.3643, "step": 25140 }, { "epoch": 2.66, - "grad_norm": 4.7459282875061035, + "grad_norm": 32.04668045043945, "learning_rate": 1.4687929031576724e-05, - "loss": 0.4866, + "loss": 1.3964, "step": 25150 }, { "epoch": 2.66, - "grad_norm": 15.92348861694336, + "grad_norm": 44.34598922729492, "learning_rate": 1.4685816876122083e-05, - "loss": 0.4706, + "loss": 1.1052, "step": 25160 }, { "epoch": 2.66, - "grad_norm": 8.865230560302734, + "grad_norm": 14.079484939575195, "learning_rate": 1.4683704720667442e-05, - "loss": 0.6475, + "loss": 1.1574, "step": 25170 }, { "epoch": 2.66, - "grad_norm": 9.39933967590332, + "grad_norm": 40.206504821777344, "learning_rate": 1.46815925652128e-05, - "loss": 0.7064, + "loss": 1.2367, "step": 25180 }, { "epoch": 2.66, - "grad_norm": 16.96772575378418, + "grad_norm": 30.08822250366211, "learning_rate": 1.467948040975816e-05, - "loss": 0.5783, + "loss": 1.2834, "step": 25190 }, { "epoch": 2.66, - "grad_norm": 6.024062633514404, + "grad_norm": 48.3693733215332, "learning_rate": 1.467736825430352e-05, - "loss": 0.794, + "loss": 1.6592, "step": 25200 }, { "epoch": 2.66, - "grad_norm": 15.799002647399902, + "grad_norm": 46.76577377319336, "learning_rate": 1.4675256098848875e-05, - "loss": 0.6338, + "loss": 1.1695, "step": 25210 }, { "epoch": 2.66, - "grad_norm": 13.542553901672363, + "grad_norm": 30.629638671875, "learning_rate": 1.4673143943394235e-05, - "loss": 0.5055, + "loss": 1.4346, "step": 25220 }, { "epoch": 2.66, - "grad_norm": 16.134599685668945, + "grad_norm": 32.92951965332031, "learning_rate": 1.4671031787939594e-05, - "loss": 0.7388, + "loss": 1.4625, "step": 25230 }, { "epoch": 2.67, - "grad_norm": 9.279046058654785, + "grad_norm": 28.53566551208496, "learning_rate": 1.4668919632484953e-05, - "loss": 0.5575, + "loss": 1.0961, "step": 25240 }, { "epoch": 2.67, - "grad_norm": 15.169075012207031, + "grad_norm": 31.344661712646484, "learning_rate": 1.4666807477030312e-05, - "loss": 0.6781, + "loss": 1.1696, "step": 25250 }, { "epoch": 2.67, - "grad_norm": 17.25811767578125, + "grad_norm": 37.72333908081055, "learning_rate": 1.4664695321575669e-05, - "loss": 0.6396, + "loss": 1.0533, "step": 25260 }, { "epoch": 2.67, - "grad_norm": 14.612153053283691, + "grad_norm": 48.13593292236328, "learning_rate": 1.4662583166121028e-05, - "loss": 0.8159, + "loss": 1.3435, "step": 25270 }, { "epoch": 2.67, - "grad_norm": 22.52862548828125, + "grad_norm": 40.59902572631836, "learning_rate": 1.4660471010666386e-05, - "loss": 0.8593, + "loss": 1.6958, "step": 25280 }, { "epoch": 2.67, - "grad_norm": 12.748289108276367, + "grad_norm": 32.994239807128906, "learning_rate": 1.4658358855211745e-05, - "loss": 0.6932, + "loss": 1.5136, "step": 25290 }, { "epoch": 2.67, - "grad_norm": 6.970276355743408, + "grad_norm": 43.9852294921875, "learning_rate": 1.4656246699757104e-05, - "loss": 0.7466, + "loss": 1.4802, "step": 25300 }, { "epoch": 2.67, - "grad_norm": 13.970146179199219, + "grad_norm": 32.33823013305664, "learning_rate": 1.4654134544302463e-05, - "loss": 0.6205, + "loss": 1.5835, "step": 25310 }, { "epoch": 2.67, - "grad_norm": 7.163900852203369, + "grad_norm": 49.56713104248047, "learning_rate": 1.465202238884782e-05, - "loss": 0.4106, + "loss": 1.2545, "step": 25320 }, { "epoch": 2.68, - "grad_norm": 36.548675537109375, + "grad_norm": 32.40760040283203, "learning_rate": 1.4649910233393179e-05, - "loss": 0.8066, + "loss": 1.1725, "step": 25330 }, { "epoch": 2.68, - "grad_norm": 5.774538040161133, + "grad_norm": 36.77871322631836, "learning_rate": 1.4647798077938537e-05, - "loss": 0.6411, + "loss": 1.4452, "step": 25340 }, { "epoch": 2.68, - "grad_norm": 18.858116149902344, + "grad_norm": 41.789485931396484, "learning_rate": 1.4645685922483896e-05, - "loss": 0.7286, + "loss": 1.3789, "step": 25350 }, { "epoch": 2.68, - "grad_norm": 15.630961418151855, + "grad_norm": 39.79800796508789, "learning_rate": 1.4643573767029255e-05, - "loss": 0.7312, + "loss": 1.2496, "step": 25360 }, { "epoch": 2.68, - "grad_norm": 15.65800952911377, + "grad_norm": 36.4302864074707, "learning_rate": 1.4641461611574612e-05, - "loss": 0.5482, + "loss": 1.3333, "step": 25370 }, { "epoch": 2.68, - "grad_norm": 11.243047714233398, + "grad_norm": 47.190223693847656, "learning_rate": 1.463934945611997e-05, - "loss": 0.4038, + "loss": 1.0365, "step": 25380 }, { "epoch": 2.68, - "grad_norm": 8.442903518676758, + "grad_norm": 30.786205291748047, "learning_rate": 1.463723730066533e-05, - "loss": 0.4955, + "loss": 1.4233, "step": 25390 }, { "epoch": 2.68, - "grad_norm": 19.495441436767578, + "grad_norm": 38.396827697753906, "learning_rate": 1.4635125145210688e-05, - "loss": 0.5559, + "loss": 1.0185, "step": 25400 }, { "epoch": 2.68, - "grad_norm": 15.718722343444824, + "grad_norm": 35.189937591552734, "learning_rate": 1.4633012989756049e-05, - "loss": 0.601, + "loss": 1.2564, "step": 25410 }, { "epoch": 2.68, - "grad_norm": 17.89275550842285, + "grad_norm": 34.34286117553711, "learning_rate": 1.4630900834301404e-05, - "loss": 0.7944, + "loss": 1.4451, "step": 25420 }, { "epoch": 2.69, - "grad_norm": 24.29680633544922, + "grad_norm": 46.63692855834961, "learning_rate": 1.4628788678846763e-05, - "loss": 0.863, + "loss": 1.57, "step": 25430 }, { "epoch": 2.69, - "grad_norm": 10.502164840698242, + "grad_norm": 30.537446975708008, "learning_rate": 1.4626676523392123e-05, - "loss": 0.6186, + "loss": 1.0754, "step": 25440 }, { "epoch": 2.69, - "grad_norm": 6.862312316894531, + "grad_norm": 41.62202835083008, "learning_rate": 1.4624564367937482e-05, - "loss": 0.8552, + "loss": 1.6852, "step": 25450 }, { "epoch": 2.69, - "grad_norm": 13.493900299072266, + "grad_norm": 13.531328201293945, "learning_rate": 1.4622452212482841e-05, - "loss": 0.7444, + "loss": 1.3382, "step": 25460 }, { "epoch": 2.69, - "grad_norm": 11.863822937011719, + "grad_norm": 20.471370697021484, "learning_rate": 1.46203400570282e-05, - "loss": 0.4891, + "loss": 1.3637, "step": 25470 }, { "epoch": 2.69, - "grad_norm": 13.944780349731445, + "grad_norm": 30.627098083496094, "learning_rate": 1.4618227901573557e-05, - "loss": 0.908, + "loss": 1.7785, "step": 25480 }, { "epoch": 2.69, - "grad_norm": 23.558053970336914, + "grad_norm": 44.93746566772461, "learning_rate": 1.4616115746118916e-05, - "loss": 0.5742, + "loss": 1.2077, "step": 25490 }, { "epoch": 2.69, - "grad_norm": 9.063162803649902, + "grad_norm": 25.751916885375977, "learning_rate": 1.4614003590664274e-05, - "loss": 0.4414, + "loss": 1.166, "step": 25500 }, { "epoch": 2.69, - "grad_norm": 74.18334197998047, + "grad_norm": 27.766220092773438, "learning_rate": 1.4611891435209633e-05, - "loss": 0.8948, + "loss": 1.3254, "step": 25510 }, { "epoch": 2.7, - "grad_norm": 0.503911554813385, + "grad_norm": 24.91207504272461, "learning_rate": 1.4609779279754992e-05, - "loss": 0.7113, + "loss": 1.252, "step": 25520 }, { "epoch": 2.7, - "grad_norm": 6.852249622344971, + "grad_norm": 56.96837615966797, "learning_rate": 1.4607667124300349e-05, - "loss": 0.713, + "loss": 1.4759, "step": 25530 }, { "epoch": 2.7, - "grad_norm": 6.954075336456299, + "grad_norm": 33.853878021240234, "learning_rate": 1.4605554968845708e-05, - "loss": 0.5798, + "loss": 1.4164, "step": 25540 }, { "epoch": 2.7, - "grad_norm": 8.32942008972168, + "grad_norm": 42.14360427856445, "learning_rate": 1.4603442813391067e-05, - "loss": 0.6733, + "loss": 1.5236, "step": 25550 }, { "epoch": 2.7, - "grad_norm": 10.682077407836914, + "grad_norm": 38.71790313720703, "learning_rate": 1.4601330657936425e-05, - "loss": 0.6476, + "loss": 1.5734, "step": 25560 }, { "epoch": 2.7, - "grad_norm": 0.732639491558075, + "grad_norm": 25.1511173248291, "learning_rate": 1.4599218502481784e-05, - "loss": 0.5019, + "loss": 1.1502, "step": 25570 }, { "epoch": 2.7, - "grad_norm": 14.675652503967285, + "grad_norm": 39.18107986450195, "learning_rate": 1.4597106347027143e-05, - "loss": 0.725, + "loss": 1.4348, "step": 25580 }, { "epoch": 2.7, - "grad_norm": 20.089092254638672, + "grad_norm": 44.089969635009766, "learning_rate": 1.45949941915725e-05, - "loss": 0.579, + "loss": 1.5309, "step": 25590 }, { "epoch": 2.7, - "grad_norm": 23.22336196899414, + "grad_norm": 61.57703399658203, "learning_rate": 1.4592882036117859e-05, - "loss": 0.8016, + "loss": 1.4567, "step": 25600 }, { "epoch": 2.7, - "grad_norm": 11.251928329467773, + "grad_norm": 32.22917556762695, "learning_rate": 1.4590769880663218e-05, - "loss": 0.5057, + "loss": 1.7166, "step": 25610 }, { "epoch": 2.71, - "grad_norm": 13.706049919128418, + "grad_norm": 54.84952926635742, "learning_rate": 1.4588657725208576e-05, - "loss": 0.804, + "loss": 1.5725, "step": 25620 }, { "epoch": 2.71, - "grad_norm": 10.837653160095215, + "grad_norm": 40.46086120605469, "learning_rate": 1.4586545569753935e-05, - "loss": 0.5952, + "loss": 1.0422, "step": 25630 }, { "epoch": 2.71, - "grad_norm": 39.581634521484375, + "grad_norm": 53.32914352416992, "learning_rate": 1.4584433414299292e-05, - "loss": 0.396, + "loss": 1.295, "step": 25640 }, { "epoch": 2.71, - "grad_norm": 13.953662872314453, + "grad_norm": 31.853742599487305, "learning_rate": 1.4582321258844651e-05, - "loss": 0.7298, + "loss": 1.1782, "step": 25650 }, { "epoch": 2.71, - "grad_norm": 15.88851547241211, + "grad_norm": 48.9560432434082, "learning_rate": 1.458020910339001e-05, - "loss": 0.4006, + "loss": 1.386, "step": 25660 }, { "epoch": 2.71, - "grad_norm": 8.1371431350708, + "grad_norm": 36.9483528137207, "learning_rate": 1.457809694793537e-05, - "loss": 0.6202, + "loss": 1.4006, "step": 25670 }, { "epoch": 2.71, - "grad_norm": 3.6486945152282715, + "grad_norm": 24.537036895751953, "learning_rate": 1.4575984792480729e-05, - "loss": 0.6572, + "loss": 1.3769, "step": 25680 }, { "epoch": 2.71, - "grad_norm": 6.726473808288574, + "grad_norm": 38.20989990234375, "learning_rate": 1.4573872637026084e-05, - "loss": 0.5494, + "loss": 1.183, "step": 25690 }, { "epoch": 2.71, - "grad_norm": 5.379217624664307, + "grad_norm": 36.56157684326172, "learning_rate": 1.4571760481571445e-05, - "loss": 0.6167, + "loss": 1.3464, "step": 25700 }, { "epoch": 2.72, - "grad_norm": 17.93466567993164, + "grad_norm": 46.19194030761719, "learning_rate": 1.4569648326116804e-05, - "loss": 0.6587, + "loss": 1.39, "step": 25710 }, { "epoch": 2.72, - "grad_norm": 12.569620132446289, + "grad_norm": 38.41474533081055, "learning_rate": 1.4567536170662162e-05, - "loss": 0.4945, + "loss": 1.2968, "step": 25720 }, { "epoch": 2.72, - "grad_norm": 22.71245765686035, + "grad_norm": 44.00651550292969, "learning_rate": 1.4565424015207521e-05, - "loss": 0.4294, + "loss": 1.2675, "step": 25730 }, { "epoch": 2.72, - "grad_norm": 0.7790854573249817, + "grad_norm": 23.93886947631836, "learning_rate": 1.456331185975288e-05, - "loss": 0.6513, + "loss": 1.336, "step": 25740 }, { "epoch": 2.72, - "grad_norm": 32.523963928222656, + "grad_norm": 40.77412414550781, "learning_rate": 1.4561199704298237e-05, - "loss": 0.661, + "loss": 1.4119, "step": 25750 }, { "epoch": 2.72, - "grad_norm": 11.797358512878418, + "grad_norm": 48.259422302246094, "learning_rate": 1.4559087548843596e-05, - "loss": 0.8709, + "loss": 1.7261, "step": 25760 }, { "epoch": 2.72, - "grad_norm": 13.865509986877441, + "grad_norm": 42.197723388671875, "learning_rate": 1.4556975393388955e-05, - "loss": 0.6028, + "loss": 1.2483, "step": 25770 }, { "epoch": 2.72, - "grad_norm": 1.423370122909546, + "grad_norm": 19.5023136138916, "learning_rate": 1.4554863237934313e-05, - "loss": 0.6699, + "loss": 1.366, "step": 25780 }, { "epoch": 2.72, - "grad_norm": 27.437068939208984, + "grad_norm": 43.41440200805664, "learning_rate": 1.4552751082479672e-05, - "loss": 0.6701, + "loss": 1.2813, "step": 25790 }, { "epoch": 2.72, - "grad_norm": 8.363204002380371, + "grad_norm": 38.89332962036133, "learning_rate": 1.455063892702503e-05, - "loss": 0.3386, + "loss": 1.1296, "step": 25800 }, { "epoch": 2.73, - "grad_norm": 13.111943244934082, + "grad_norm": 22.40572166442871, "learning_rate": 1.4548526771570388e-05, - "loss": 0.6696, + "loss": 1.1738, "step": 25810 }, { "epoch": 2.73, - "grad_norm": 1.2511026859283447, + "grad_norm": 52.29423141479492, "learning_rate": 1.4546414616115747e-05, - "loss": 0.5636, + "loss": 1.4226, "step": 25820 }, { "epoch": 2.73, - "grad_norm": 29.63718032836914, + "grad_norm": 40.819583892822266, "learning_rate": 1.4544302460661106e-05, - "loss": 0.8065, + "loss": 1.5954, "step": 25830 }, { "epoch": 2.73, - "grad_norm": 20.004114151000977, + "grad_norm": 34.68144989013672, "learning_rate": 1.4542190305206464e-05, - "loss": 0.4977, + "loss": 1.1758, "step": 25840 }, { "epoch": 2.73, - "grad_norm": 4.859766006469727, + "grad_norm": 32.46920394897461, "learning_rate": 1.4540078149751823e-05, - "loss": 0.8984, + "loss": 1.2971, "step": 25850 }, { "epoch": 2.73, - "grad_norm": 19.922517776489258, + "grad_norm": 40.081966400146484, "learning_rate": 1.453796599429718e-05, - "loss": 0.4358, + "loss": 1.3805, "step": 25860 }, { "epoch": 2.73, - "grad_norm": 12.987701416015625, + "grad_norm": 40.278865814208984, "learning_rate": 1.4535853838842539e-05, - "loss": 0.9294, + "loss": 1.6212, "step": 25870 }, { "epoch": 2.73, - "grad_norm": 31.448360443115234, + "grad_norm": 44.0798225402832, "learning_rate": 1.4533741683387898e-05, - "loss": 0.9598, + "loss": 1.4457, "step": 25880 }, { "epoch": 2.73, - "grad_norm": 8.263919830322266, + "grad_norm": 46.16936111450195, "learning_rate": 1.4531629527933258e-05, - "loss": 0.5309, + "loss": 1.1065, "step": 25890 }, { "epoch": 2.74, - "grad_norm": 12.097142219543457, + "grad_norm": 28.632055282592773, "learning_rate": 1.4529517372478617e-05, - "loss": 0.5117, + "loss": 1.3504, "step": 25900 }, { "epoch": 2.74, - "grad_norm": 20.758304595947266, + "grad_norm": 40.96074676513672, "learning_rate": 1.4527405217023973e-05, - "loss": 0.9947, + "loss": 1.3776, "step": 25910 }, { "epoch": 2.74, - "grad_norm": 1.3838478326797485, + "grad_norm": 27.65357208251953, "learning_rate": 1.4525293061569333e-05, - "loss": 0.7316, + "loss": 1.6788, "step": 25920 }, { "epoch": 2.74, - "grad_norm": 0.8531906008720398, + "grad_norm": 19.60964584350586, "learning_rate": 1.4523180906114692e-05, - "loss": 0.6159, + "loss": 1.3386, "step": 25930 }, { "epoch": 2.74, - "grad_norm": 24.551250457763672, + "grad_norm": 38.117313385009766, "learning_rate": 1.452106875066005e-05, - "loss": 0.5762, + "loss": 1.379, "step": 25940 }, { "epoch": 2.74, - "grad_norm": 3.0253076553344727, + "grad_norm": 41.69289016723633, "learning_rate": 1.451895659520541e-05, - "loss": 0.5241, + "loss": 1.0043, "step": 25950 }, { "epoch": 2.74, - "grad_norm": 15.924264907836914, + "grad_norm": 36.82322311401367, "learning_rate": 1.4516844439750766e-05, - "loss": 0.6292, + "loss": 1.5198, "step": 25960 }, { "epoch": 2.74, - "grad_norm": 16.85487937927246, + "grad_norm": 46.187538146972656, "learning_rate": 1.4514732284296125e-05, - "loss": 0.6056, + "loss": 1.6249, "step": 25970 }, { "epoch": 2.74, - "grad_norm": 19.777862548828125, + "grad_norm": 35.433895111083984, "learning_rate": 1.4512620128841484e-05, - "loss": 0.6471, + "loss": 0.9865, "step": 25980 }, { "epoch": 2.74, - "grad_norm": 12.2997407913208, + "grad_norm": 42.09664535522461, "learning_rate": 1.4510507973386843e-05, - "loss": 0.5665, + "loss": 1.1796, "step": 25990 }, { "epoch": 2.75, - "grad_norm": 16.09093475341797, + "grad_norm": 30.47739601135254, "learning_rate": 1.4508395817932202e-05, - "loss": 0.3922, + "loss": 1.2038, "step": 26000 }, { "epoch": 2.75, - "grad_norm": 2.628002882003784, + "grad_norm": 22.823638916015625, "learning_rate": 1.450628366247756e-05, - "loss": 0.6885, + "loss": 1.3528, "step": 26010 }, { "epoch": 2.75, - "grad_norm": 14.670366287231445, + "grad_norm": 44.750999450683594, "learning_rate": 1.4504171507022917e-05, - "loss": 0.4078, + "loss": 1.0885, "step": 26020 }, { "epoch": 2.75, - "grad_norm": 31.483510971069336, + "grad_norm": 34.139617919921875, "learning_rate": 1.4502059351568276e-05, - "loss": 0.6128, + "loss": 1.6889, "step": 26030 }, { "epoch": 2.75, - "grad_norm": 17.51835823059082, + "grad_norm": 30.93158531188965, "learning_rate": 1.4499947196113635e-05, - "loss": 0.6727, + "loss": 1.1058, "step": 26040 }, { "epoch": 2.75, - "grad_norm": 15.397427558898926, + "grad_norm": 55.588539123535156, "learning_rate": 1.4497835040658994e-05, - "loss": 0.6305, + "loss": 1.5368, "step": 26050 }, { "epoch": 2.75, - "grad_norm": 12.595888137817383, + "grad_norm": 45.275962829589844, "learning_rate": 1.4495722885204353e-05, - "loss": 0.8558, + "loss": 1.6273, "step": 26060 }, { "epoch": 2.75, - "grad_norm": 23.2119083404541, + "grad_norm": 38.17668533325195, "learning_rate": 1.449361072974971e-05, - "loss": 0.5503, + "loss": 1.0964, "step": 26070 }, { "epoch": 2.75, - "grad_norm": 12.813423156738281, + "grad_norm": 23.380592346191406, "learning_rate": 1.4491498574295068e-05, - "loss": 0.7057, + "loss": 1.3803, "step": 26080 }, { "epoch": 2.76, - "grad_norm": 24.91280174255371, + "grad_norm": 47.08763885498047, "learning_rate": 1.4489386418840427e-05, - "loss": 0.459, + "loss": 0.9502, "step": 26090 }, { "epoch": 2.76, - "grad_norm": 3.5887017250061035, + "grad_norm": 38.84933853149414, "learning_rate": 1.4487274263385786e-05, - "loss": 0.5758, + "loss": 1.029, "step": 26100 }, { "epoch": 2.76, - "grad_norm": 20.237998962402344, + "grad_norm": 35.54845428466797, "learning_rate": 1.4485162107931145e-05, - "loss": 0.7168, + "loss": 1.4105, "step": 26110 }, { "epoch": 2.76, - "grad_norm": 34.35276412963867, + "grad_norm": 46.71614074707031, "learning_rate": 1.4483049952476505e-05, - "loss": 0.5226, + "loss": 1.3983, "step": 26120 }, { "epoch": 2.76, - "grad_norm": 20.906265258789062, + "grad_norm": 37.073333740234375, "learning_rate": 1.448093779702186e-05, - "loss": 0.7702, + "loss": 1.4254, "step": 26130 }, { "epoch": 2.76, - "grad_norm": 9.87645149230957, + "grad_norm": 31.017166137695312, "learning_rate": 1.447882564156722e-05, - "loss": 0.7811, + "loss": 1.2708, "step": 26140 }, { "epoch": 2.76, - "grad_norm": 1.8361396789550781, + "grad_norm": 47.4441032409668, "learning_rate": 1.447671348611258e-05, - "loss": 0.5643, + "loss": 1.2944, "step": 26150 }, { "epoch": 2.76, - "grad_norm": 23.571514129638672, + "grad_norm": 48.17219161987305, "learning_rate": 1.4474601330657939e-05, - "loss": 0.6746, + "loss": 1.4651, "step": 26160 }, { "epoch": 2.76, - "grad_norm": 22.4217529296875, + "grad_norm": 34.843421936035156, "learning_rate": 1.4472489175203297e-05, - "loss": 0.7816, + "loss": 1.4409, "step": 26170 }, { "epoch": 2.76, - "grad_norm": 13.009736061096191, + "grad_norm": 51.649314880371094, "learning_rate": 1.4470377019748654e-05, - "loss": 0.3628, + "loss": 1.0653, "step": 26180 }, { "epoch": 2.77, - "grad_norm": 9.70933723449707, + "grad_norm": 36.0309944152832, "learning_rate": 1.4468264864294013e-05, - "loss": 0.7191, + "loss": 1.2786, "step": 26190 }, { "epoch": 2.77, - "grad_norm": 8.503499984741211, + "grad_norm": 44.84917068481445, "learning_rate": 1.4466152708839372e-05, - "loss": 0.5423, + "loss": 1.1167, "step": 26200 }, { "epoch": 2.77, - "grad_norm": 12.314536094665527, + "grad_norm": 24.57476234436035, "learning_rate": 1.446404055338473e-05, - "loss": 0.3587, + "loss": 1.2713, "step": 26210 }, { "epoch": 2.77, - "grad_norm": 10.824721336364746, + "grad_norm": 27.484333038330078, "learning_rate": 1.446192839793009e-05, - "loss": 0.6071, + "loss": 1.3903, "step": 26220 }, { "epoch": 2.77, - "grad_norm": 28.721723556518555, + "grad_norm": 39.10836410522461, "learning_rate": 1.4459816242475448e-05, - "loss": 0.4724, + "loss": 1.4654, "step": 26230 }, { "epoch": 2.77, - "grad_norm": 7.223791599273682, + "grad_norm": 35.49933624267578, "learning_rate": 1.4457704087020805e-05, - "loss": 0.6034, + "loss": 1.4118, "step": 26240 }, { "epoch": 2.77, - "grad_norm": 21.642520904541016, + "grad_norm": 54.62269973754883, "learning_rate": 1.4455591931566164e-05, - "loss": 0.5283, + "loss": 1.2131, "step": 26250 }, { "epoch": 2.77, - "grad_norm": 26.646621704101562, + "grad_norm": 34.59545135498047, "learning_rate": 1.4453479776111523e-05, - "loss": 0.8908, + "loss": 1.2577, "step": 26260 }, { "epoch": 2.77, - "grad_norm": 14.347640991210938, + "grad_norm": 35.13551712036133, "learning_rate": 1.4451367620656882e-05, - "loss": 0.6699, + "loss": 1.3505, "step": 26270 }, { "epoch": 2.78, - "grad_norm": 8.979655265808105, + "grad_norm": 31.931808471679688, "learning_rate": 1.444925546520224e-05, - "loss": 0.7032, + "loss": 1.1431, "step": 26280 }, { "epoch": 2.78, - "grad_norm": 10.944151878356934, + "grad_norm": 31.271451950073242, "learning_rate": 1.4447143309747598e-05, - "loss": 0.5153, + "loss": 1.2199, "step": 26290 }, { "epoch": 2.78, - "grad_norm": 6.661558151245117, + "grad_norm": 39.379207611083984, "learning_rate": 1.4445031154292956e-05, - "loss": 0.4654, + "loss": 1.461, "step": 26300 }, { "epoch": 2.78, - "grad_norm": 15.000690460205078, + "grad_norm": 56.28676986694336, "learning_rate": 1.4442918998838315e-05, - "loss": 0.5618, + "loss": 1.2389, "step": 26310 }, { "epoch": 2.78, - "grad_norm": 6.858580589294434, + "grad_norm": 33.59934997558594, "learning_rate": 1.4440806843383674e-05, - "loss": 0.7798, + "loss": 1.3993, "step": 26320 }, { "epoch": 2.78, - "grad_norm": 9.612177848815918, + "grad_norm": 46.945762634277344, "learning_rate": 1.4438694687929033e-05, - "loss": 0.7787, + "loss": 1.2677, "step": 26330 }, { "epoch": 2.78, - "grad_norm": 8.423362731933594, + "grad_norm": 35.67131423950195, "learning_rate": 1.443658253247439e-05, - "loss": 0.6631, + "loss": 1.4058, "step": 26340 }, { "epoch": 2.78, - "grad_norm": 22.94274139404297, + "grad_norm": 35.8778076171875, "learning_rate": 1.4434470377019749e-05, - "loss": 0.2632, + "loss": 0.9713, "step": 26350 }, { "epoch": 2.78, - "grad_norm": 22.72514533996582, + "grad_norm": 42.32845687866211, "learning_rate": 1.4432358221565107e-05, - "loss": 0.6812, + "loss": 1.431, "step": 26360 }, { "epoch": 2.78, - "grad_norm": 33.11704635620117, + "grad_norm": 20.765541076660156, "learning_rate": 1.4430246066110468e-05, - "loss": 0.3503, + "loss": 1.0499, "step": 26370 }, { "epoch": 2.79, - "grad_norm": 15.214692115783691, + "grad_norm": 27.093162536621094, "learning_rate": 1.4428133910655827e-05, - "loss": 0.4672, + "loss": 1.0046, "step": 26380 }, { "epoch": 2.79, - "grad_norm": 25.228261947631836, + "grad_norm": 27.117517471313477, "learning_rate": 1.4426021755201185e-05, - "loss": 0.6844, + "loss": 1.4345, "step": 26390 }, { "epoch": 2.79, - "grad_norm": 4.487293720245361, + "grad_norm": 41.81623840332031, "learning_rate": 1.4423909599746543e-05, - "loss": 0.6278, + "loss": 1.4652, "step": 26400 }, { "epoch": 2.79, - "grad_norm": 14.024654388427734, + "grad_norm": 37.118648529052734, "learning_rate": 1.4421797444291901e-05, - "loss": 0.6351, + "loss": 1.1965, "step": 26410 }, { "epoch": 2.79, - "grad_norm": 4.466198444366455, + "grad_norm": 38.16646194458008, "learning_rate": 1.441968528883726e-05, - "loss": 0.793, + "loss": 1.417, "step": 26420 }, { "epoch": 2.79, - "grad_norm": 3.7916505336761475, + "grad_norm": 29.661334991455078, "learning_rate": 1.4417573133382619e-05, - "loss": 0.6593, + "loss": 1.1075, "step": 26430 }, { "epoch": 2.79, - "grad_norm": 30.369342803955078, + "grad_norm": 48.95799255371094, "learning_rate": 1.4415460977927978e-05, - "loss": 0.6912, + "loss": 1.4415, "step": 26440 }, { "epoch": 2.79, - "grad_norm": 14.65196704864502, + "grad_norm": 16.587574005126953, "learning_rate": 1.4413348822473335e-05, - "loss": 0.4332, + "loss": 1.0926, "step": 26450 }, { "epoch": 2.79, - "grad_norm": 6.413945198059082, + "grad_norm": 38.45721435546875, "learning_rate": 1.4411236667018693e-05, - "loss": 0.8878, + "loss": 1.4326, "step": 26460 }, { "epoch": 2.8, - "grad_norm": 2.615788221359253, + "grad_norm": 51.37424087524414, "learning_rate": 1.4409124511564052e-05, - "loss": 0.6755, + "loss": 1.1643, "step": 26470 }, { "epoch": 2.8, - "grad_norm": 26.761940002441406, + "grad_norm": 29.314266204833984, "learning_rate": 1.4407012356109411e-05, - "loss": 0.8282, + "loss": 1.3099, "step": 26480 }, { "epoch": 2.8, - "grad_norm": 15.858684539794922, + "grad_norm": 22.952407836914062, "learning_rate": 1.440490020065477e-05, - "loss": 0.5377, + "loss": 1.1864, "step": 26490 }, { "epoch": 2.8, - "grad_norm": 10.611167907714844, + "grad_norm": 41.41761779785156, "learning_rate": 1.4402788045200129e-05, - "loss": 0.8707, + "loss": 1.2029, "step": 26500 }, { "epoch": 2.8, - "grad_norm": 4.366647720336914, + "grad_norm": 31.733177185058594, "learning_rate": 1.4400675889745486e-05, - "loss": 0.9366, + "loss": 1.5298, "step": 26510 }, { "epoch": 2.8, - "grad_norm": 8.895903587341309, + "grad_norm": 34.74082946777344, "learning_rate": 1.4398563734290844e-05, - "loss": 0.4988, + "loss": 1.3206, "step": 26520 }, { "epoch": 2.8, - "grad_norm": 9.618026733398438, + "grad_norm": 30.198028564453125, "learning_rate": 1.4396451578836203e-05, - "loss": 0.5286, + "loss": 1.284, "step": 26530 }, { "epoch": 2.8, - "grad_norm": 0.16872495412826538, + "grad_norm": 28.238927841186523, "learning_rate": 1.4394339423381562e-05, - "loss": 0.7106, + "loss": 1.1496, "step": 26540 }, { "epoch": 2.8, - "grad_norm": 11.960671424865723, + "grad_norm": 40.24373245239258, "learning_rate": 1.439222726792692e-05, - "loss": 0.7524, + "loss": 1.3933, "step": 26550 }, { "epoch": 2.8, - "grad_norm": 13.648602485656738, + "grad_norm": 42.60723876953125, "learning_rate": 1.4390115112472278e-05, - "loss": 0.4812, + "loss": 1.0134, "step": 26560 }, { "epoch": 2.81, - "grad_norm": 12.511092185974121, + "grad_norm": 36.19515609741211, "learning_rate": 1.4388002957017637e-05, - "loss": 0.6369, + "loss": 1.6072, "step": 26570 }, { "epoch": 2.81, - "grad_norm": 40.7391471862793, + "grad_norm": 31.900100708007812, "learning_rate": 1.4385890801562995e-05, - "loss": 0.4962, + "loss": 1.2753, "step": 26580 }, { "epoch": 2.81, - "grad_norm": 14.623761177062988, + "grad_norm": 40.76873016357422, "learning_rate": 1.4383778646108354e-05, - "loss": 0.7306, + "loss": 1.2863, "step": 26590 }, { "epoch": 2.81, - "grad_norm": 6.527597427368164, + "grad_norm": 41.70841979980469, "learning_rate": 1.4381666490653715e-05, - "loss": 0.7009, + "loss": 1.0967, "step": 26600 }, { "epoch": 2.81, - "grad_norm": 9.912211418151855, + "grad_norm": 30.631498336791992, "learning_rate": 1.437955433519907e-05, - "loss": 0.7695, + "loss": 1.2676, "step": 26610 }, { "epoch": 2.81, - "grad_norm": 17.064346313476562, + "grad_norm": 41.71709060668945, "learning_rate": 1.4377442179744429e-05, - "loss": 0.6931, + "loss": 1.2157, "step": 26620 }, { "epoch": 2.81, - "grad_norm": 20.084447860717773, + "grad_norm": 31.642053604125977, "learning_rate": 1.437533002428979e-05, - "loss": 0.9124, + "loss": 1.5375, "step": 26630 }, { "epoch": 2.81, - "grad_norm": 9.099656105041504, + "grad_norm": 29.474836349487305, "learning_rate": 1.4373217868835148e-05, - "loss": 0.7173, + "loss": 1.3672, "step": 26640 }, { "epoch": 2.81, - "grad_norm": 12.037697792053223, + "grad_norm": 52.88407897949219, "learning_rate": 1.4371105713380507e-05, - "loss": 0.5078, + "loss": 0.9348, "step": 26650 }, { "epoch": 2.82, - "grad_norm": 8.707119941711426, + "grad_norm": 29.735960006713867, "learning_rate": 1.4368993557925866e-05, - "loss": 0.4441, + "loss": 1.4009, "step": 26660 }, { "epoch": 2.82, - "grad_norm": 17.447778701782227, + "grad_norm": 36.5802001953125, "learning_rate": 1.4366881402471223e-05, - "loss": 0.7373, + "loss": 1.4319, "step": 26670 }, { "epoch": 2.82, - "grad_norm": 0.15239298343658447, + "grad_norm": 40.99263000488281, "learning_rate": 1.4364769247016582e-05, - "loss": 0.6165, + "loss": 1.1244, "step": 26680 }, { "epoch": 2.82, - "grad_norm": 36.42332077026367, + "grad_norm": 30.564861297607422, "learning_rate": 1.436265709156194e-05, - "loss": 0.5786, + "loss": 1.115, "step": 26690 }, { "epoch": 2.82, - "grad_norm": 8.276172637939453, + "grad_norm": 31.35621452331543, "learning_rate": 1.4360544936107299e-05, - "loss": 0.6646, + "loss": 1.3527, "step": 26700 }, { "epoch": 2.82, - "grad_norm": 5.390921592712402, + "grad_norm": 28.79714012145996, "learning_rate": 1.4358432780652658e-05, - "loss": 0.5821, + "loss": 1.3662, "step": 26710 }, { "epoch": 2.82, - "grad_norm": 11.957677841186523, + "grad_norm": 48.615272521972656, "learning_rate": 1.4356320625198015e-05, - "loss": 0.3578, + "loss": 1.0961, "step": 26720 }, { "epoch": 2.82, - "grad_norm": 13.5248441696167, + "grad_norm": 25.451807022094727, "learning_rate": 1.4354208469743374e-05, - "loss": 0.7215, + "loss": 1.1287, "step": 26730 }, { "epoch": 2.82, - "grad_norm": 15.495743751525879, + "grad_norm": 39.69667053222656, "learning_rate": 1.4352096314288733e-05, - "loss": 0.4243, + "loss": 0.9212, "step": 26740 }, { "epoch": 2.83, - "grad_norm": 17.11842155456543, + "grad_norm": 40.73474884033203, "learning_rate": 1.4349984158834091e-05, - "loss": 0.7211, + "loss": 1.5368, "step": 26750 }, { "epoch": 2.83, - "grad_norm": 9.35068130493164, + "grad_norm": 29.992874145507812, "learning_rate": 1.434787200337945e-05, - "loss": 1.1571, + "loss": 1.3361, "step": 26760 }, { "epoch": 2.83, - "grad_norm": 17.287275314331055, + "grad_norm": 39.27516555786133, "learning_rate": 1.4345759847924809e-05, - "loss": 0.5993, + "loss": 1.0436, "step": 26770 }, { "epoch": 2.83, - "grad_norm": 26.156734466552734, + "grad_norm": 30.930727005004883, "learning_rate": 1.4343647692470166e-05, - "loss": 0.8924, + "loss": 1.4166, "step": 26780 }, { "epoch": 2.83, - "grad_norm": 11.03944206237793, + "grad_norm": 28.142675399780273, "learning_rate": 1.4341535537015525e-05, - "loss": 0.6227, + "loss": 1.233, "step": 26790 }, { "epoch": 2.83, - "grad_norm": 9.828619956970215, + "grad_norm": 31.80063819885254, "learning_rate": 1.4339423381560884e-05, - "loss": 0.6496, + "loss": 1.4195, "step": 26800 }, { "epoch": 2.83, - "grad_norm": 8.139717102050781, + "grad_norm": 22.412338256835938, "learning_rate": 1.4337311226106242e-05, - "loss": 0.6839, + "loss": 1.179, "step": 26810 }, { "epoch": 2.83, - "grad_norm": 12.41656494140625, + "grad_norm": 30.067516326904297, "learning_rate": 1.4335199070651603e-05, - "loss": 0.7778, + "loss": 1.2328, "step": 26820 }, { "epoch": 2.83, - "grad_norm": 12.551342010498047, + "grad_norm": 47.27248764038086, "learning_rate": 1.4333086915196958e-05, - "loss": 0.6597, + "loss": 1.5728, "step": 26830 }, { "epoch": 2.83, - "grad_norm": 6.898613452911377, + "grad_norm": 41.90227508544922, "learning_rate": 1.4330974759742317e-05, - "loss": 0.5712, + "loss": 1.4613, "step": 26840 }, { "epoch": 2.84, - "grad_norm": 1.6482281684875488, + "grad_norm": 25.249942779541016, "learning_rate": 1.4328862604287677e-05, - "loss": 0.3411, + "loss": 1.0076, "step": 26850 }, { "epoch": 2.84, - "grad_norm": 3.570301055908203, + "grad_norm": 28.96766471862793, "learning_rate": 1.4326750448833036e-05, - "loss": 0.7635, + "loss": 1.3577, "step": 26860 }, { "epoch": 2.84, - "grad_norm": 7.224052906036377, + "grad_norm": 36.45043182373047, "learning_rate": 1.4324638293378395e-05, - "loss": 0.4176, + "loss": 1.1103, "step": 26870 }, { "epoch": 2.84, - "grad_norm": 21.280427932739258, + "grad_norm": 52.57265090942383, "learning_rate": 1.4322526137923754e-05, - "loss": 0.7904, + "loss": 1.387, "step": 26880 }, { "epoch": 2.84, - "grad_norm": 14.400222778320312, + "grad_norm": 29.716659545898438, "learning_rate": 1.432041398246911e-05, - "loss": 0.7001, + "loss": 1.1551, "step": 26890 }, { "epoch": 2.84, - "grad_norm": 12.534754753112793, + "grad_norm": 31.40945816040039, "learning_rate": 1.431830182701447e-05, - "loss": 0.9075, + "loss": 1.1874, "step": 26900 }, { "epoch": 2.84, - "grad_norm": 21.877275466918945, + "grad_norm": 46.281494140625, "learning_rate": 1.4316189671559828e-05, - "loss": 0.9681, + "loss": 1.3922, "step": 26910 }, { "epoch": 2.84, - "grad_norm": 13.403423309326172, + "grad_norm": 40.58292770385742, "learning_rate": 1.4314077516105187e-05, - "loss": 0.6946, + "loss": 1.4632, "step": 26920 }, { "epoch": 2.84, - "grad_norm": 13.608017921447754, + "grad_norm": 26.994478225708008, "learning_rate": 1.4311965360650546e-05, - "loss": 0.8074, + "loss": 1.4742, "step": 26930 }, { "epoch": 2.85, - "grad_norm": 20.65261459350586, + "grad_norm": 33.56576156616211, "learning_rate": 1.4309853205195903e-05, - "loss": 0.5551, + "loss": 1.4526, "step": 26940 }, { "epoch": 2.85, - "grad_norm": 0.6562764048576355, + "grad_norm": 24.364789962768555, "learning_rate": 1.4307741049741262e-05, - "loss": 0.746, + "loss": 1.0796, "step": 26950 }, { "epoch": 2.85, - "grad_norm": 25.76386260986328, + "grad_norm": 49.25712585449219, "learning_rate": 1.430562889428662e-05, - "loss": 0.5059, + "loss": 1.2726, "step": 26960 }, { "epoch": 2.85, - "grad_norm": 22.936819076538086, + "grad_norm": 33.51769256591797, "learning_rate": 1.430351673883198e-05, - "loss": 0.4111, + "loss": 1.2739, "step": 26970 }, { "epoch": 2.85, - "grad_norm": 13.2286958694458, + "grad_norm": 44.788841247558594, "learning_rate": 1.4301404583377338e-05, - "loss": 0.4458, + "loss": 1.1527, "step": 26980 }, { "epoch": 2.85, - "grad_norm": 26.112083435058594, + "grad_norm": 34.522823333740234, "learning_rate": 1.4299292427922695e-05, - "loss": 0.6967, + "loss": 1.3261, "step": 26990 }, { "epoch": 2.85, - "grad_norm": 7.7767815589904785, + "grad_norm": 43.25320816040039, "learning_rate": 1.4297180272468054e-05, - "loss": 0.684, + "loss": 1.5702, "step": 27000 }, { "epoch": 2.85, - "grad_norm": 22.246570587158203, + "grad_norm": 41.8881721496582, "learning_rate": 1.4295068117013413e-05, - "loss": 0.5059, + "loss": 1.4776, "step": 27010 }, { "epoch": 2.85, - "grad_norm": 15.162459373474121, + "grad_norm": 37.20798110961914, "learning_rate": 1.4292955961558772e-05, - "loss": 0.6581, + "loss": 1.1495, "step": 27020 }, { "epoch": 2.85, - "grad_norm": 2.9692275524139404, + "grad_norm": 41.97847366333008, "learning_rate": 1.429084380610413e-05, - "loss": 0.412, + "loss": 1.4031, "step": 27030 }, { "epoch": 2.86, - "grad_norm": 23.74357795715332, + "grad_norm": 40.976295471191406, "learning_rate": 1.428873165064949e-05, - "loss": 0.4533, + "loss": 1.3538, "step": 27040 }, { "epoch": 2.86, - "grad_norm": 18.87971305847168, + "grad_norm": 33.48983383178711, "learning_rate": 1.4286619495194846e-05, - "loss": 0.5489, + "loss": 1.3666, "step": 27050 }, { "epoch": 2.86, - "grad_norm": 19.29755973815918, + "grad_norm": 32.26839065551758, "learning_rate": 1.4284507339740205e-05, - "loss": 0.6686, + "loss": 1.2379, "step": 27060 }, { "epoch": 2.86, - "grad_norm": 23.47871208190918, + "grad_norm": 30.319515228271484, "learning_rate": 1.4282395184285564e-05, - "loss": 0.7334, + "loss": 1.5314, "step": 27070 }, { "epoch": 2.86, - "grad_norm": 11.476421356201172, + "grad_norm": 41.22843933105469, "learning_rate": 1.4280283028830924e-05, - "loss": 0.7908, + "loss": 1.7318, "step": 27080 }, { "epoch": 2.86, - "grad_norm": 16.162813186645508, + "grad_norm": 34.13923263549805, "learning_rate": 1.4278170873376283e-05, - "loss": 0.528, + "loss": 1.2188, "step": 27090 }, { "epoch": 2.86, - "grad_norm": 12.587010383605957, + "grad_norm": 31.617143630981445, "learning_rate": 1.4276058717921638e-05, - "loss": 0.6443, + "loss": 1.2998, "step": 27100 }, { "epoch": 2.86, - "grad_norm": 6.446773052215576, + "grad_norm": 43.554656982421875, "learning_rate": 1.4273946562466999e-05, - "loss": 0.9222, + "loss": 1.6394, "step": 27110 }, { "epoch": 2.86, - "grad_norm": 3.007917881011963, + "grad_norm": 43.61532211303711, "learning_rate": 1.4271834407012358e-05, - "loss": 0.3493, + "loss": 1.1413, "step": 27120 }, { "epoch": 2.87, - "grad_norm": 8.0367431640625, + "grad_norm": 41.49147415161133, "learning_rate": 1.4269722251557716e-05, - "loss": 0.5716, + "loss": 1.0516, "step": 27130 }, { "epoch": 2.87, - "grad_norm": 3.407684087753296, + "grad_norm": 39.65132522583008, "learning_rate": 1.4267610096103075e-05, - "loss": 0.6349, + "loss": 1.2488, "step": 27140 }, { "epoch": 2.87, - "grad_norm": 8.968683242797852, + "grad_norm": 31.03995704650879, "learning_rate": 1.4265497940648434e-05, - "loss": 0.8114, + "loss": 1.2997, "step": 27150 }, { "epoch": 2.87, - "grad_norm": 24.571949005126953, + "grad_norm": 45.421852111816406, "learning_rate": 1.4263385785193791e-05, - "loss": 0.6098, + "loss": 1.2648, "step": 27160 }, { "epoch": 2.87, - "grad_norm": 28.6878719329834, + "grad_norm": 49.280723571777344, "learning_rate": 1.426127362973915e-05, - "loss": 0.3593, + "loss": 1.1091, "step": 27170 }, { "epoch": 2.87, - "grad_norm": 11.861822128295898, + "grad_norm": 20.03152847290039, "learning_rate": 1.4259161474284509e-05, - "loss": 0.5807, + "loss": 1.0931, "step": 27180 }, { "epoch": 2.87, - "grad_norm": 18.363723754882812, + "grad_norm": 44.31555938720703, "learning_rate": 1.4257049318829867e-05, - "loss": 0.4937, + "loss": 1.2298, "step": 27190 }, { "epoch": 2.87, - "grad_norm": 11.469681739807129, + "grad_norm": 46.8553466796875, "learning_rate": 1.4254937163375226e-05, - "loss": 0.3924, + "loss": 1.3249, "step": 27200 }, { "epoch": 2.87, - "grad_norm": 3.2752881050109863, + "grad_norm": 39.74966812133789, "learning_rate": 1.4252825007920583e-05, - "loss": 0.3997, + "loss": 1.2793, "step": 27210 }, { "epoch": 2.87, - "grad_norm": 0.44543829560279846, + "grad_norm": 37.396766662597656, "learning_rate": 1.4250712852465942e-05, - "loss": 0.7674, + "loss": 1.2797, "step": 27220 }, { "epoch": 2.88, - "grad_norm": 25.056095123291016, + "grad_norm": 34.286380767822266, "learning_rate": 1.42486006970113e-05, - "loss": 0.9257, + "loss": 1.3749, "step": 27230 }, { "epoch": 2.88, - "grad_norm": 9.235565185546875, + "grad_norm": 37.59224319458008, "learning_rate": 1.424648854155666e-05, - "loss": 0.5451, + "loss": 1.406, "step": 27240 }, { "epoch": 2.88, - "grad_norm": 25.936832427978516, + "grad_norm": 33.01359176635742, "learning_rate": 1.4244376386102018e-05, - "loss": 0.6162, + "loss": 1.7235, "step": 27250 }, { "epoch": 2.88, - "grad_norm": 11.887899398803711, + "grad_norm": 53.14383316040039, "learning_rate": 1.4242264230647375e-05, - "loss": 0.5532, + "loss": 1.4965, "step": 27260 }, { "epoch": 2.88, - "grad_norm": 6.660991191864014, + "grad_norm": 27.62831687927246, "learning_rate": 1.4240152075192734e-05, - "loss": 0.7374, + "loss": 1.5407, "step": 27270 }, { "epoch": 2.88, - "grad_norm": 13.412386894226074, + "grad_norm": 31.95365333557129, "learning_rate": 1.4238039919738093e-05, - "loss": 0.728, + "loss": 1.2301, "step": 27280 }, { "epoch": 2.88, - "grad_norm": 16.013437271118164, + "grad_norm": 35.41677474975586, "learning_rate": 1.4235927764283452e-05, - "loss": 0.536, + "loss": 1.2582, "step": 27290 }, { "epoch": 2.88, - "grad_norm": 4.194121360778809, + "grad_norm": 52.6983528137207, "learning_rate": 1.4233815608828812e-05, - "loss": 0.5498, + "loss": 1.4698, "step": 27300 }, { "epoch": 2.88, - "grad_norm": 9.040087699890137, + "grad_norm": 53.647186279296875, "learning_rate": 1.4231703453374171e-05, - "loss": 0.5505, + "loss": 1.3008, "step": 27310 }, { "epoch": 2.89, - "grad_norm": 5.730649948120117, + "grad_norm": 21.30131721496582, "learning_rate": 1.4229591297919526e-05, - "loss": 0.5021, + "loss": 1.3912, "step": 27320 }, { "epoch": 2.89, - "grad_norm": 15.045702934265137, + "grad_norm": 31.20167350769043, "learning_rate": 1.4227479142464887e-05, - "loss": 0.5909, + "loss": 1.1328, "step": 27330 }, { "epoch": 2.89, - "grad_norm": 0.7010139226913452, + "grad_norm": 12.090182304382324, "learning_rate": 1.4225366987010246e-05, - "loss": 0.6654, + "loss": 1.1031, "step": 27340 }, { "epoch": 2.89, - "grad_norm": 15.802462577819824, + "grad_norm": 32.55187225341797, "learning_rate": 1.4223254831555604e-05, - "loss": 1.0255, + "loss": 1.4395, "step": 27350 }, { "epoch": 2.89, - "grad_norm": 4.134285926818848, + "grad_norm": 26.66245460510254, "learning_rate": 1.4221142676100963e-05, - "loss": 0.842, + "loss": 1.4345, "step": 27360 }, { "epoch": 2.89, - "grad_norm": 15.98482894897461, + "grad_norm": 58.105873107910156, "learning_rate": 1.421903052064632e-05, - "loss": 0.7379, + "loss": 1.2444, "step": 27370 }, { "epoch": 2.89, - "grad_norm": 25.6693115234375, + "grad_norm": 41.150413513183594, "learning_rate": 1.4216918365191679e-05, - "loss": 0.5966, + "loss": 1.3074, "step": 27380 }, { "epoch": 2.89, - "grad_norm": 15.905277252197266, + "grad_norm": 37.49351119995117, "learning_rate": 1.4214806209737038e-05, - "loss": 0.5669, + "loss": 1.5515, "step": 27390 }, { "epoch": 2.89, - "grad_norm": 14.82565689086914, + "grad_norm": 35.19620132446289, "learning_rate": 1.4212694054282397e-05, - "loss": 0.6635, + "loss": 1.2671, "step": 27400 }, { "epoch": 2.89, - "grad_norm": 15.86871337890625, + "grad_norm": 19.867910385131836, "learning_rate": 1.4210581898827755e-05, - "loss": 0.9195, + "loss": 1.2441, "step": 27410 }, { "epoch": 2.9, - "grad_norm": 8.98967456817627, + "grad_norm": 57.04841232299805, "learning_rate": 1.4208469743373114e-05, - "loss": 0.8113, + "loss": 1.432, "step": 27420 }, { "epoch": 2.9, - "grad_norm": 5.53631591796875, + "grad_norm": 30.28963279724121, "learning_rate": 1.4206357587918471e-05, - "loss": 0.5413, + "loss": 0.9451, "step": 27430 }, { "epoch": 2.9, - "grad_norm": 11.135986328125, + "grad_norm": 37.7454833984375, "learning_rate": 1.420424543246383e-05, - "loss": 0.4891, + "loss": 1.5063, "step": 27440 }, { "epoch": 2.9, - "grad_norm": 13.9522066116333, + "grad_norm": 32.779415130615234, "learning_rate": 1.4202133277009189e-05, - "loss": 0.7096, + "loss": 1.4233, "step": 27450 }, { "epoch": 2.9, - "grad_norm": 16.523372650146484, + "grad_norm": 32.03135681152344, "learning_rate": 1.4200021121554548e-05, - "loss": 0.6059, + "loss": 1.4234, "step": 27460 }, { "epoch": 2.9, - "grad_norm": 4.2992777824401855, + "grad_norm": 46.59170150756836, "learning_rate": 1.4197908966099906e-05, - "loss": 0.6252, + "loss": 1.7478, "step": 27470 }, { "epoch": 2.9, - "grad_norm": 23.7196102142334, + "grad_norm": 46.486595153808594, "learning_rate": 1.4195796810645264e-05, - "loss": 0.8102, + "loss": 1.4043, "step": 27480 }, { "epoch": 2.9, - "grad_norm": 14.030991554260254, + "grad_norm": 42.64503479003906, "learning_rate": 1.4193684655190622e-05, - "loss": 0.4839, + "loss": 1.1297, "step": 27490 }, { "epoch": 2.9, - "grad_norm": 25.467199325561523, + "grad_norm": 38.35319137573242, "learning_rate": 1.4191572499735981e-05, - "loss": 0.5563, + "loss": 1.201, "step": 27500 }, { "epoch": 2.91, - "grad_norm": 25.02648162841797, + "grad_norm": 44.75257110595703, "learning_rate": 1.418946034428134e-05, - "loss": 0.5239, + "loss": 1.2487, "step": 27510 }, { "epoch": 2.91, - "grad_norm": 12.880396842956543, + "grad_norm": 60.873722076416016, "learning_rate": 1.41873481888267e-05, - "loss": 0.4567, + "loss": 1.1928, "step": 27520 }, { "epoch": 2.91, - "grad_norm": 11.7610445022583, + "grad_norm": 33.83588790893555, "learning_rate": 1.4185236033372056e-05, - "loss": 0.5679, + "loss": 1.4696, "step": 27530 }, { "epoch": 2.91, - "grad_norm": 23.017900466918945, + "grad_norm": 51.08100891113281, "learning_rate": 1.4183123877917415e-05, - "loss": 0.7568, + "loss": 1.4249, "step": 27540 }, { "epoch": 2.91, - "grad_norm": 9.701220512390137, + "grad_norm": 45.27206039428711, "learning_rate": 1.4181011722462775e-05, - "loss": 0.869, + "loss": 1.0321, "step": 27550 }, { "epoch": 2.91, - "grad_norm": 15.999932289123535, + "grad_norm": 42.17929458618164, "learning_rate": 1.4178899567008134e-05, - "loss": 0.4962, + "loss": 1.2319, "step": 27560 }, { "epoch": 2.91, - "grad_norm": 5.337689399719238, + "grad_norm": 27.262792587280273, "learning_rate": 1.4176787411553493e-05, - "loss": 0.7455, + "loss": 1.5078, "step": 27570 }, { "epoch": 2.91, - "grad_norm": 22.195039749145508, + "grad_norm": 35.464237213134766, "learning_rate": 1.4174675256098851e-05, - "loss": 0.5426, + "loss": 1.5177, "step": 27580 }, { "epoch": 2.91, - "grad_norm": 11.297523498535156, + "grad_norm": 48.913551330566406, "learning_rate": 1.4172563100644208e-05, - "loss": 0.7722, + "loss": 1.3185, "step": 27590 }, { "epoch": 2.91, - "grad_norm": 6.6254353523254395, + "grad_norm": 33.837860107421875, "learning_rate": 1.4170450945189567e-05, - "loss": 0.726, + "loss": 1.1583, "step": 27600 }, { "epoch": 2.92, - "grad_norm": 5.126913547515869, + "grad_norm": 37.991661071777344, "learning_rate": 1.4168338789734926e-05, - "loss": 0.6292, + "loss": 1.1481, "step": 27610 }, { "epoch": 2.92, - "grad_norm": 22.763565063476562, + "grad_norm": 23.775161743164062, "learning_rate": 1.4166226634280285e-05, - "loss": 0.4179, + "loss": 1.0994, "step": 27620 }, { "epoch": 2.92, - "grad_norm": 10.975974082946777, + "grad_norm": 34.557098388671875, "learning_rate": 1.4164114478825644e-05, - "loss": 0.4042, + "loss": 0.8045, "step": 27630 }, { "epoch": 2.92, - "grad_norm": 0.20695586502552032, + "grad_norm": 33.52455139160156, "learning_rate": 1.4162002323371e-05, - "loss": 0.3816, + "loss": 1.4929, "step": 27640 }, { "epoch": 2.92, - "grad_norm": 34.50890350341797, + "grad_norm": 39.110137939453125, "learning_rate": 1.415989016791636e-05, - "loss": 0.6468, + "loss": 1.5563, "step": 27650 }, { "epoch": 2.92, - "grad_norm": 16.380998611450195, + "grad_norm": 56.2846794128418, "learning_rate": 1.4157778012461718e-05, - "loss": 0.8128, + "loss": 1.397, "step": 27660 }, { "epoch": 2.92, - "grad_norm": 2.4381675720214844, + "grad_norm": 33.56706619262695, "learning_rate": 1.4155665857007077e-05, - "loss": 0.7738, + "loss": 1.6237, "step": 27670 }, { "epoch": 2.92, - "grad_norm": 11.62938404083252, + "grad_norm": 49.90800476074219, "learning_rate": 1.4153553701552436e-05, - "loss": 0.3834, + "loss": 1.0769, "step": 27680 }, { "epoch": 2.92, - "grad_norm": 13.366825103759766, + "grad_norm": 33.298702239990234, "learning_rate": 1.4151441546097794e-05, - "loss": 0.782, + "loss": 1.3305, "step": 27690 }, { "epoch": 2.93, - "grad_norm": 6.553989887237549, + "grad_norm": 37.03019714355469, "learning_rate": 1.4149329390643152e-05, - "loss": 0.5106, + "loss": 1.2069, "step": 27700 }, { "epoch": 2.93, - "grad_norm": 10.868496894836426, + "grad_norm": 19.628948211669922, "learning_rate": 1.414721723518851e-05, - "loss": 0.7817, + "loss": 1.3095, "step": 27710 }, { "epoch": 2.93, - "grad_norm": 4.292243003845215, + "grad_norm": 28.839839935302734, "learning_rate": 1.4145105079733869e-05, - "loss": 0.4698, + "loss": 1.2525, "step": 27720 }, { "epoch": 2.93, - "grad_norm": 8.545909881591797, + "grad_norm": 38.870182037353516, "learning_rate": 1.4142992924279228e-05, - "loss": 0.5668, + "loss": 1.2616, "step": 27730 }, { "epoch": 2.93, - "grad_norm": 24.50951385498047, + "grad_norm": 33.28901672363281, "learning_rate": 1.4140880768824587e-05, - "loss": 0.4809, + "loss": 1.1093, "step": 27740 }, { "epoch": 2.93, - "grad_norm": 4.373388767242432, + "grad_norm": 20.64056396484375, "learning_rate": 1.4138768613369944e-05, - "loss": 0.9096, + "loss": 1.2834, "step": 27750 }, { "epoch": 2.93, - "grad_norm": 17.356355667114258, + "grad_norm": 26.9033260345459, "learning_rate": 1.4136656457915303e-05, - "loss": 0.6934, + "loss": 2.0867, "step": 27760 }, { "epoch": 2.93, - "grad_norm": 18.765241622924805, + "grad_norm": 47.69428634643555, "learning_rate": 1.4134544302460661e-05, - "loss": 0.6075, + "loss": 1.2644, "step": 27770 }, { "epoch": 2.93, - "grad_norm": 12.508805274963379, + "grad_norm": 38.00539016723633, "learning_rate": 1.4132432147006022e-05, - "loss": 0.7165, + "loss": 1.1774, "step": 27780 }, { "epoch": 2.93, - "grad_norm": 3.5593550205230713, + "grad_norm": 23.559934616088867, "learning_rate": 1.413031999155138e-05, - "loss": 0.4741, + "loss": 1.2731, "step": 27790 }, { "epoch": 2.94, - "grad_norm": 14.50855541229248, + "grad_norm": 37.73366165161133, "learning_rate": 1.412820783609674e-05, - "loss": 0.5231, + "loss": 1.322, "step": 27800 }, { "epoch": 2.94, - "grad_norm": 7.7570085525512695, + "grad_norm": 28.881677627563477, "learning_rate": 1.4126095680642096e-05, - "loss": 0.8298, + "loss": 1.3049, "step": 27810 }, { "epoch": 2.94, - "grad_norm": 18.494823455810547, + "grad_norm": 31.862329483032227, "learning_rate": 1.4123983525187455e-05, - "loss": 0.6198, + "loss": 1.0815, "step": 27820 }, { "epoch": 2.94, - "grad_norm": 17.719280242919922, + "grad_norm": 52.31625747680664, "learning_rate": 1.4121871369732814e-05, - "loss": 0.4892, + "loss": 1.1034, "step": 27830 }, { "epoch": 2.94, - "grad_norm": 9.623159408569336, + "grad_norm": 31.6778621673584, "learning_rate": 1.4119759214278173e-05, - "loss": 0.5046, + "loss": 1.1865, "step": 27840 }, { "epoch": 2.94, - "grad_norm": 6.709212779998779, + "grad_norm": 43.98627471923828, "learning_rate": 1.4117647058823532e-05, - "loss": 0.4741, + "loss": 1.3882, "step": 27850 }, { "epoch": 2.94, - "grad_norm": 15.404531478881836, + "grad_norm": 38.81816864013672, "learning_rate": 1.4115534903368889e-05, - "loss": 0.8833, + "loss": 1.5129, "step": 27860 }, { "epoch": 2.94, - "grad_norm": 12.642060279846191, + "grad_norm": 20.58915138244629, "learning_rate": 1.4113422747914247e-05, - "loss": 0.8721, + "loss": 1.3916, "step": 27870 }, { "epoch": 2.94, - "grad_norm": 7.833823204040527, + "grad_norm": 36.0140380859375, "learning_rate": 1.4111310592459606e-05, - "loss": 0.4338, + "loss": 1.528, "step": 27880 }, { "epoch": 2.95, - "grad_norm": 14.449695587158203, + "grad_norm": 28.89951515197754, "learning_rate": 1.4109198437004965e-05, - "loss": 0.7349, + "loss": 1.4877, "step": 27890 }, { "epoch": 2.95, - "grad_norm": 26.395519256591797, + "grad_norm": 33.64126968383789, "learning_rate": 1.4107086281550324e-05, - "loss": 0.7418, + "loss": 1.5682, "step": 27900 }, { "epoch": 2.95, - "grad_norm": 19.181365966796875, + "grad_norm": 29.61495590209961, "learning_rate": 1.4104974126095681e-05, - "loss": 0.4656, + "loss": 1.0207, "step": 27910 }, { "epoch": 2.95, - "grad_norm": 5.10526180267334, + "grad_norm": 30.089818954467773, "learning_rate": 1.410286197064104e-05, - "loss": 0.8759, + "loss": 1.3716, "step": 27920 }, { "epoch": 2.95, - "grad_norm": 4.318934440612793, + "grad_norm": 33.55873489379883, "learning_rate": 1.4100749815186398e-05, - "loss": 0.5703, + "loss": 1.1046, "step": 27930 }, { "epoch": 2.95, - "grad_norm": 16.35566520690918, + "grad_norm": 45.635780334472656, "learning_rate": 1.4098637659731757e-05, - "loss": 0.637, + "loss": 1.1674, "step": 27940 }, { "epoch": 2.95, - "grad_norm": 9.372228622436523, + "grad_norm": 23.765968322753906, "learning_rate": 1.4096525504277116e-05, - "loss": 0.7439, + "loss": 1.0767, "step": 27950 }, { "epoch": 2.95, - "grad_norm": 13.11884593963623, + "grad_norm": 32.41028594970703, "learning_rate": 1.4094413348822475e-05, - "loss": 0.6263, + "loss": 1.3745, "step": 27960 }, { "epoch": 2.95, - "grad_norm": 5.747426509857178, + "grad_norm": 30.388538360595703, "learning_rate": 1.4092301193367832e-05, - "loss": 0.7147, + "loss": 1.1318, "step": 27970 }, { "epoch": 2.95, - "grad_norm": 29.073957443237305, + "grad_norm": 29.779460906982422, "learning_rate": 1.409018903791319e-05, - "loss": 0.5826, + "loss": 1.1294, "step": 27980 }, { "epoch": 2.96, - "grad_norm": 7.968169212341309, + "grad_norm": 47.91167068481445, "learning_rate": 1.408807688245855e-05, - "loss": 0.6547, + "loss": 1.3269, "step": 27990 }, { "epoch": 2.96, - "grad_norm": 2.6982827186584473, + "grad_norm": 35.56184768676758, "learning_rate": 1.408596472700391e-05, - "loss": 0.518, + "loss": 1.1537, "step": 28000 }, { "epoch": 2.96, - "grad_norm": 0.5216006636619568, + "grad_norm": 40.238929748535156, "learning_rate": 1.4083852571549269e-05, - "loss": 0.5952, + "loss": 1.0342, "step": 28010 }, { "epoch": 2.96, - "grad_norm": 18.820253372192383, + "grad_norm": 47.5219612121582, "learning_rate": 1.4081740416094624e-05, - "loss": 0.6842, + "loss": 1.3273, "step": 28020 }, { "epoch": 2.96, - "grad_norm": 251.0028076171875, + "grad_norm": 49.21763229370117, "learning_rate": 1.4079628260639985e-05, - "loss": 0.9577, + "loss": 1.3401, "step": 28030 }, { "epoch": 2.96, - "grad_norm": 21.198719024658203, + "grad_norm": 23.07725715637207, "learning_rate": 1.4077516105185343e-05, - "loss": 0.6298, + "loss": 1.2757, "step": 28040 }, { "epoch": 2.96, - "grad_norm": 8.617956161499023, + "grad_norm": 44.75910568237305, "learning_rate": 1.4075403949730702e-05, - "loss": 0.4841, + "loss": 1.4521, "step": 28050 }, { "epoch": 2.96, - "grad_norm": 1.1768062114715576, + "grad_norm": 35.22341537475586, "learning_rate": 1.407329179427606e-05, - "loss": 0.4249, + "loss": 1.5297, "step": 28060 }, { "epoch": 2.96, - "grad_norm": 13.877388000488281, + "grad_norm": 51.72187805175781, "learning_rate": 1.407117963882142e-05, - "loss": 0.5671, + "loss": 1.2441, "step": 28070 }, { "epoch": 2.97, - "grad_norm": 0.5208641290664673, + "grad_norm": 38.08430862426758, "learning_rate": 1.4069067483366777e-05, - "loss": 0.7293, + "loss": 1.373, "step": 28080 }, { "epoch": 2.97, - "grad_norm": 0.9069938063621521, + "grad_norm": 34.390323638916016, "learning_rate": 1.4066955327912135e-05, - "loss": 0.6834, + "loss": 1.3258, "step": 28090 }, { "epoch": 2.97, - "grad_norm": 6.293151378631592, + "grad_norm": 29.354141235351562, "learning_rate": 1.4064843172457494e-05, - "loss": 0.4394, + "loss": 1.1756, "step": 28100 }, { "epoch": 2.97, - "grad_norm": 2.9518628120422363, + "grad_norm": 35.150543212890625, "learning_rate": 1.4062731017002853e-05, - "loss": 0.8477, + "loss": 1.3413, "step": 28110 }, { "epoch": 2.97, - "grad_norm": 19.069473266601562, + "grad_norm": 18.56864356994629, "learning_rate": 1.4060618861548212e-05, - "loss": 0.4463, + "loss": 0.9678, "step": 28120 }, { "epoch": 2.97, - "grad_norm": 23.818387985229492, + "grad_norm": 38.88895034790039, "learning_rate": 1.4058506706093569e-05, - "loss": 0.6331, + "loss": 1.1802, "step": 28130 }, { "epoch": 2.97, - "grad_norm": 15.319705963134766, + "grad_norm": 33.92264938354492, "learning_rate": 1.4056394550638928e-05, - "loss": 0.7715, + "loss": 1.5339, "step": 28140 }, { "epoch": 2.97, - "grad_norm": 3.0533790588378906, + "grad_norm": 47.12361526489258, "learning_rate": 1.4054282395184286e-05, - "loss": 0.6466, + "loss": 1.3632, "step": 28150 }, { "epoch": 2.97, - "grad_norm": 19.41332244873047, + "grad_norm": 48.31201171875, "learning_rate": 1.4052170239729645e-05, - "loss": 0.8351, + "loss": 1.5427, "step": 28160 }, { "epoch": 2.97, - "grad_norm": 30.660680770874023, + "grad_norm": 41.55253601074219, "learning_rate": 1.4050058084275004e-05, - "loss": 1.0235, + "loss": 1.3829, "step": 28170 }, { "epoch": 2.98, - "grad_norm": 9.277071952819824, + "grad_norm": 29.98430061340332, "learning_rate": 1.4047945928820361e-05, - "loss": 0.8527, + "loss": 1.3155, "step": 28180 }, { "epoch": 2.98, - "grad_norm": 20.932600021362305, + "grad_norm": 32.103797912597656, "learning_rate": 1.404583377336572e-05, - "loss": 0.5345, + "loss": 0.7554, "step": 28190 }, { "epoch": 2.98, - "grad_norm": 9.142674446105957, + "grad_norm": 29.58299446105957, "learning_rate": 1.4043721617911079e-05, - "loss": 0.8505, + "loss": 1.3099, "step": 28200 }, { "epoch": 2.98, - "grad_norm": 5.1292405128479, + "grad_norm": 32.51066207885742, "learning_rate": 1.4041609462456437e-05, - "loss": 0.549, + "loss": 0.9387, "step": 28210 }, { "epoch": 2.98, - "grad_norm": 16.18271827697754, + "grad_norm": 38.09770965576172, "learning_rate": 1.4039497307001796e-05, - "loss": 0.552, + "loss": 1.2958, "step": 28220 }, { "epoch": 2.98, - "grad_norm": 11.454407691955566, + "grad_norm": 45.585391998291016, "learning_rate": 1.4037385151547157e-05, - "loss": 0.578, + "loss": 1.3695, "step": 28230 }, { "epoch": 2.98, - "grad_norm": 2.1484272480010986, + "grad_norm": 25.2840633392334, "learning_rate": 1.4035272996092512e-05, - "loss": 0.7753, + "loss": 1.3702, "step": 28240 }, { "epoch": 2.98, - "grad_norm": 9.886902809143066, + "grad_norm": 32.32411575317383, "learning_rate": 1.4033160840637871e-05, - "loss": 0.5994, + "loss": 1.0617, "step": 28250 }, { "epoch": 2.98, - "grad_norm": 22.406618118286133, + "grad_norm": 32.93859100341797, "learning_rate": 1.4031048685183231e-05, - "loss": 0.6838, + "loss": 1.3612, "step": 28260 }, { "epoch": 2.99, - "grad_norm": 15.816312789916992, + "grad_norm": 39.74430847167969, "learning_rate": 1.402893652972859e-05, - "loss": 0.6045, + "loss": 1.5008, "step": 28270 }, { "epoch": 2.99, - "grad_norm": 12.45836067199707, + "grad_norm": 48.93321228027344, "learning_rate": 1.4026824374273949e-05, - "loss": 0.439, + "loss": 1.3043, "step": 28280 }, { "epoch": 2.99, - "grad_norm": 8.958867073059082, + "grad_norm": 35.89628219604492, "learning_rate": 1.4024712218819306e-05, - "loss": 0.6128, + "loss": 1.1017, "step": 28290 }, { "epoch": 2.99, - "grad_norm": 15.045181274414062, + "grad_norm": 37.79891586303711, "learning_rate": 1.4022600063364665e-05, - "loss": 0.5841, + "loss": 1.2309, "step": 28300 }, { "epoch": 2.99, - "grad_norm": 18.612934112548828, + "grad_norm": 27.855552673339844, "learning_rate": 1.4020487907910024e-05, - "loss": 0.4761, + "loss": 1.3542, "step": 28310 }, { "epoch": 2.99, - "grad_norm": 12.081652641296387, + "grad_norm": 32.202144622802734, "learning_rate": 1.4018375752455382e-05, - "loss": 0.7081, + "loss": 1.1127, "step": 28320 }, { "epoch": 2.99, - "grad_norm": 35.69585418701172, + "grad_norm": 37.56852722167969, "learning_rate": 1.4016263597000741e-05, - "loss": 0.6444, + "loss": 1.4944, "step": 28330 }, { "epoch": 2.99, - "grad_norm": 16.041658401489258, + "grad_norm": 33.03897476196289, "learning_rate": 1.40141514415461e-05, - "loss": 0.7117, + "loss": 1.1534, "step": 28340 }, { "epoch": 2.99, - "grad_norm": 8.62686824798584, + "grad_norm": 25.525819778442383, "learning_rate": 1.4012039286091457e-05, - "loss": 0.7512, + "loss": 1.3133, "step": 28350 }, { "epoch": 3.0, - "grad_norm": 0.7184221744537354, + "grad_norm": 26.723215103149414, "learning_rate": 1.4009927130636816e-05, - "loss": 0.773, + "loss": 1.5546, "step": 28360 }, { "epoch": 3.0, - "grad_norm": 3.46923565864563, + "grad_norm": 37.178077697753906, "learning_rate": 1.4007814975182175e-05, - "loss": 0.4683, + "loss": 0.9935, "step": 28370 }, { "epoch": 3.0, - "grad_norm": 14.794042587280273, + "grad_norm": 52.97393035888672, "learning_rate": 1.4005702819727533e-05, - "loss": 0.7858, + "loss": 1.423, "step": 28380 }, { "epoch": 3.0, - "grad_norm": 22.727617263793945, + "grad_norm": 24.94605255126953, "learning_rate": 1.4003590664272892e-05, - "loss": 0.7663, + "loss": 1.1643, "step": 28390 }, { "epoch": 3.0, - "grad_norm": 12.267287254333496, + "grad_norm": 33.956787109375, "learning_rate": 1.400147850881825e-05, - "loss": 0.9109, + "loss": 1.2981, "step": 28400 }, { "epoch": 3.0, - "eval_accuracy": 0.889940594059406, - "eval_loss": 0.4085531532764435, - "eval_runtime": 168.9147, - "eval_samples_per_second": 149.484, - "eval_steps_per_second": 18.69, + "eval_accuracy": 0.7557623762376238, + "eval_loss": 0.9253702163696289, + "eval_runtime": 37.067, + "eval_samples_per_second": 681.199, + "eval_steps_per_second": 85.17, "step": 28407 }, { "epoch": 3.0, - "grad_norm": 4.225605010986328, + "grad_norm": 34.70069122314453, "learning_rate": 1.3999366353363608e-05, - "loss": 0.4985, + "loss": 1.1827, "step": 28410 }, { "epoch": 3.0, - "grad_norm": 34.81728744506836, + "grad_norm": 39.199832916259766, "learning_rate": 1.3997254197908967e-05, - "loss": 0.6031, + "loss": 1.22, "step": 28420 }, { "epoch": 3.0, - "grad_norm": 9.078131675720215, + "grad_norm": 38.37471389770508, "learning_rate": 1.3995142042454325e-05, - "loss": 0.4133, + "loss": 1.283, "step": 28430 }, { "epoch": 3.0, - "grad_norm": 5.788714408874512, + "grad_norm": 37.17246627807617, "learning_rate": 1.3993029886999684e-05, - "loss": 0.6207, + "loss": 1.1333, "step": 28440 }, { "epoch": 3.0, - "grad_norm": 17.096923828125, + "grad_norm": 27.427108764648438, "learning_rate": 1.3990917731545045e-05, - "loss": 0.4997, + "loss": 0.8775, "step": 28450 }, { "epoch": 3.01, - "grad_norm": 2.884169578552246, + "grad_norm": 15.635231018066406, "learning_rate": 1.39888055760904e-05, - "loss": 0.5584, + "loss": 1.1157, "step": 28460 }, { "epoch": 3.01, - "grad_norm": 10.577092170715332, + "grad_norm": 42.51263427734375, "learning_rate": 1.3986693420635759e-05, - "loss": 0.5367, + "loss": 1.3816, "step": 28470 }, { "epoch": 3.01, - "grad_norm": 14.263782501220703, + "grad_norm": 36.403480529785156, "learning_rate": 1.398458126518112e-05, - "loss": 0.5138, + "loss": 1.2874, "step": 28480 }, { "epoch": 3.01, - "grad_norm": 0.3830653131008148, + "grad_norm": 26.970264434814453, "learning_rate": 1.3982469109726478e-05, - "loss": 0.746, + "loss": 1.21, "step": 28490 }, { "epoch": 3.01, - "grad_norm": 13.499850273132324, + "grad_norm": 24.624658584594727, "learning_rate": 1.3980356954271837e-05, - "loss": 0.5129, + "loss": 1.1074, "step": 28500 }, { "epoch": 3.01, - "grad_norm": 3.4433224201202393, + "grad_norm": 49.473148345947266, "learning_rate": 1.3978244798817194e-05, - "loss": 0.5947, + "loss": 1.1096, "step": 28510 }, { "epoch": 3.01, - "grad_norm": 19.315622329711914, + "grad_norm": 24.017742156982422, "learning_rate": 1.3976132643362553e-05, - "loss": 0.4649, + "loss": 1.4916, "step": 28520 }, { "epoch": 3.01, - "grad_norm": 7.32052755355835, + "grad_norm": 34.26243591308594, "learning_rate": 1.3974020487907912e-05, - "loss": 0.6781, + "loss": 1.2607, "step": 28530 }, { "epoch": 3.01, - "grad_norm": 13.395041465759277, + "grad_norm": 41.48011779785156, "learning_rate": 1.397190833245327e-05, - "loss": 0.6008, + "loss": 1.1992, "step": 28540 }, { "epoch": 3.02, - "grad_norm": 15.720329284667969, + "grad_norm": 49.48031997680664, "learning_rate": 1.3969796176998629e-05, - "loss": 0.5637, + "loss": 1.369, "step": 28550 }, { "epoch": 3.02, - "grad_norm": 2.5410642623901367, + "grad_norm": 35.81778335571289, "learning_rate": 1.3967684021543986e-05, - "loss": 0.3722, + "loss": 1.2113, "step": 28560 }, { "epoch": 3.02, - "grad_norm": 10.027081489562988, + "grad_norm": 27.560359954833984, "learning_rate": 1.3965571866089345e-05, - "loss": 0.5358, + "loss": 1.2935, "step": 28570 }, { "epoch": 3.02, - "grad_norm": 1.1944111585617065, + "grad_norm": 50.90874099731445, "learning_rate": 1.3963459710634704e-05, - "loss": 0.4429, + "loss": 1.0759, "step": 28580 }, { "epoch": 3.02, - "grad_norm": 6.772549629211426, + "grad_norm": 15.149406433105469, "learning_rate": 1.3961347555180063e-05, - "loss": 0.27, + "loss": 0.935, "step": 28590 }, { "epoch": 3.02, - "grad_norm": 4.0106072425842285, + "grad_norm": 29.431350708007812, "learning_rate": 1.3959235399725421e-05, - "loss": 0.3014, + "loss": 0.9223, "step": 28600 }, { "epoch": 3.02, - "grad_norm": 10.616175651550293, + "grad_norm": 35.89412307739258, "learning_rate": 1.395712324427078e-05, - "loss": 0.6795, + "loss": 1.4905, "step": 28610 }, { "epoch": 3.02, - "grad_norm": 15.051374435424805, + "grad_norm": 35.96988296508789, "learning_rate": 1.3955011088816137e-05, - "loss": 0.3783, + "loss": 1.2303, "step": 28620 }, { "epoch": 3.02, - "grad_norm": 1.5803903341293335, + "grad_norm": 37.15188217163086, "learning_rate": 1.3952898933361496e-05, - "loss": 0.432, + "loss": 1.1643, "step": 28630 }, { "epoch": 3.02, - "grad_norm": 9.765969276428223, + "grad_norm": 34.679325103759766, "learning_rate": 1.3950786777906855e-05, - "loss": 0.4334, + "loss": 1.1982, "step": 28640 }, { "epoch": 3.03, - "grad_norm": 7.9406232833862305, + "grad_norm": 40.68422317504883, "learning_rate": 1.3948674622452214e-05, - "loss": 0.319, + "loss": 1.1823, "step": 28650 }, { "epoch": 3.03, - "grad_norm": 5.766817092895508, + "grad_norm": 40.887184143066406, "learning_rate": 1.3946562466997572e-05, - "loss": 0.7286, + "loss": 1.2615, "step": 28660 }, { "epoch": 3.03, - "grad_norm": 8.184953689575195, + "grad_norm": 30.154157638549805, "learning_rate": 1.394445031154293e-05, - "loss": 0.5417, + "loss": 1.1884, "step": 28670 }, { "epoch": 3.03, - "grad_norm": 20.049222946166992, + "grad_norm": 37.93383026123047, "learning_rate": 1.3942338156088288e-05, - "loss": 0.5069, + "loss": 1.224, "step": 28680 }, { "epoch": 3.03, - "grad_norm": 12.149192810058594, + "grad_norm": 20.151689529418945, "learning_rate": 1.3940226000633647e-05, - "loss": 0.3251, + "loss": 1.1639, "step": 28690 }, { "epoch": 3.03, - "grad_norm": 21.377567291259766, + "grad_norm": 37.10142135620117, "learning_rate": 1.3938113845179006e-05, - "loss": 0.4832, + "loss": 1.2897, "step": 28700 }, { "epoch": 3.03, - "grad_norm": 6.528010368347168, + "grad_norm": 36.35879898071289, "learning_rate": 1.3936001689724366e-05, - "loss": 0.5679, + "loss": 1.5656, "step": 28710 }, { "epoch": 3.03, - "grad_norm": 6.201622009277344, + "grad_norm": 46.58597946166992, "learning_rate": 1.3933889534269725e-05, - "loss": 0.5537, + "loss": 1.3219, "step": 28720 }, { "epoch": 3.03, - "grad_norm": 12.723419189453125, + "grad_norm": 40.32083511352539, "learning_rate": 1.393177737881508e-05, - "loss": 0.4003, + "loss": 1.0752, "step": 28730 }, { "epoch": 3.04, - "grad_norm": 8.200990676879883, + "grad_norm": 46.986412048339844, "learning_rate": 1.3929665223360441e-05, - "loss": 0.4917, + "loss": 1.2528, "step": 28740 }, { "epoch": 3.04, - "grad_norm": 12.214149475097656, + "grad_norm": 58.67802429199219, "learning_rate": 1.39275530679058e-05, - "loss": 0.6207, + "loss": 1.2844, "step": 28750 }, { "epoch": 3.04, - "grad_norm": 33.32883071899414, + "grad_norm": 37.759315490722656, "learning_rate": 1.3925440912451158e-05, - "loss": 0.6334, + "loss": 1.0382, "step": 28760 }, { "epoch": 3.04, - "grad_norm": 5.667093276977539, + "grad_norm": 49.979984283447266, "learning_rate": 1.3923328756996517e-05, - "loss": 0.4974, + "loss": 1.4068, "step": 28770 }, { "epoch": 3.04, - "grad_norm": 14.669828414916992, + "grad_norm": 34.6949462890625, "learning_rate": 1.3921216601541874e-05, - "loss": 0.538, + "loss": 1.195, "step": 28780 }, { "epoch": 3.04, - "grad_norm": 27.24262809753418, + "grad_norm": 49.54807662963867, "learning_rate": 1.3919104446087233e-05, - "loss": 0.8983, + "loss": 1.3097, "step": 28790 }, { "epoch": 3.04, - "grad_norm": 111.23234558105469, + "grad_norm": 32.40266799926758, "learning_rate": 1.3916992290632592e-05, - "loss": 0.5675, + "loss": 0.9131, "step": 28800 }, { "epoch": 3.04, - "grad_norm": 11.413455963134766, + "grad_norm": 48.635414123535156, "learning_rate": 1.391488013517795e-05, - "loss": 0.3559, + "loss": 1.3959, "step": 28810 }, { "epoch": 3.04, - "grad_norm": 2.6876308917999268, + "grad_norm": 41.354156494140625, "learning_rate": 1.391276797972331e-05, - "loss": 0.7857, + "loss": 1.3558, "step": 28820 }, { "epoch": 3.04, - "grad_norm": 5.041779518127441, + "grad_norm": 37.25138854980469, "learning_rate": 1.3910655824268666e-05, - "loss": 0.521, + "loss": 1.3319, "step": 28830 }, { "epoch": 3.05, - "grad_norm": 2.409600019454956, + "grad_norm": 27.48839569091797, "learning_rate": 1.3908543668814025e-05, - "loss": 0.7589, + "loss": 1.4238, "step": 28840 }, { "epoch": 3.05, - "grad_norm": 30.42850112915039, + "grad_norm": 41.690738677978516, "learning_rate": 1.3906431513359384e-05, - "loss": 0.6361, + "loss": 1.1417, "step": 28850 }, { "epoch": 3.05, - "grad_norm": 5.480411529541016, + "grad_norm": 33.373897552490234, "learning_rate": 1.3904319357904743e-05, - "loss": 0.9199, + "loss": 1.2451, "step": 28860 }, { "epoch": 3.05, - "grad_norm": 6.8912577629089355, + "grad_norm": 28.68582534790039, "learning_rate": 1.3902207202450102e-05, - "loss": 0.3713, + "loss": 1.0233, "step": 28870 }, { "epoch": 3.05, - "grad_norm": 9.640803337097168, + "grad_norm": 34.59016036987305, "learning_rate": 1.390009504699546e-05, - "loss": 0.3891, + "loss": 1.1874, "step": 28880 }, { "epoch": 3.05, - "grad_norm": 11.915757179260254, + "grad_norm": 65.52476501464844, "learning_rate": 1.3897982891540817e-05, - "loss": 0.5395, + "loss": 1.2054, "step": 28890 }, { "epoch": 3.05, - "grad_norm": 7.70464563369751, + "grad_norm": 38.008384704589844, "learning_rate": 1.3895870736086176e-05, - "loss": 0.6443, + "loss": 1.2473, "step": 28900 }, { "epoch": 3.05, - "grad_norm": 18.05623435974121, + "grad_norm": 53.502681732177734, "learning_rate": 1.3893758580631535e-05, - "loss": 0.4805, + "loss": 1.2054, "step": 28910 }, { "epoch": 3.05, - "grad_norm": 8.131916046142578, + "grad_norm": 34.639888763427734, "learning_rate": 1.3891646425176894e-05, - "loss": 0.4486, + "loss": 1.2675, "step": 28920 }, { "epoch": 3.06, - "grad_norm": 31.354934692382812, + "grad_norm": 36.872432708740234, "learning_rate": 1.3889534269722254e-05, - "loss": 0.3982, + "loss": 1.2165, "step": 28930 }, { "epoch": 3.06, - "grad_norm": 0.3529629111289978, + "grad_norm": 6.502475738525391, "learning_rate": 1.388742211426761e-05, - "loss": 0.5789, + "loss": 1.083, "step": 28940 }, { "epoch": 3.06, - "grad_norm": 8.449238777160645, + "grad_norm": 33.7398567199707, "learning_rate": 1.3885309958812968e-05, - "loss": 0.4012, + "loss": 1.2144, "step": 28950 }, { "epoch": 3.06, - "grad_norm": 7.275576114654541, + "grad_norm": 35.572265625, "learning_rate": 1.3883197803358329e-05, - "loss": 0.3656, + "loss": 1.1335, "step": 28960 }, { "epoch": 3.06, - "grad_norm": 4.728947162628174, + "grad_norm": 36.8114013671875, "learning_rate": 1.3881085647903688e-05, - "loss": 0.4631, + "loss": 0.8964, "step": 28970 }, { "epoch": 3.06, - "grad_norm": 19.272127151489258, + "grad_norm": 23.319303512573242, "learning_rate": 1.3878973492449046e-05, - "loss": 0.572, + "loss": 1.2563, "step": 28980 }, { "epoch": 3.06, - "grad_norm": 0.291201651096344, + "grad_norm": 22.434799194335938, "learning_rate": 1.3876861336994405e-05, - "loss": 0.488, + "loss": 1.3454, "step": 28990 }, { "epoch": 3.06, - "grad_norm": 3.122300148010254, + "grad_norm": 36.75706100463867, "learning_rate": 1.3874749181539762e-05, - "loss": 0.6753, + "loss": 1.1186, "step": 29000 }, { "epoch": 3.06, - "grad_norm": 88.0995864868164, + "grad_norm": 47.71918487548828, "learning_rate": 1.3872637026085121e-05, - "loss": 0.6127, + "loss": 1.2806, "step": 29010 }, { "epoch": 3.06, - "grad_norm": 8.85733699798584, + "grad_norm": 35.52614974975586, "learning_rate": 1.387052487063048e-05, - "loss": 0.4952, + "loss": 1.2508, "step": 29020 }, { "epoch": 3.07, - "grad_norm": 4.679677486419678, + "grad_norm": 31.3001651763916, "learning_rate": 1.3868412715175839e-05, - "loss": 0.6773, + "loss": 1.3928, "step": 29030 }, { "epoch": 3.07, - "grad_norm": 12.56849479675293, + "grad_norm": 45.398109436035156, "learning_rate": 1.3866300559721197e-05, - "loss": 0.8267, + "loss": 1.6275, "step": 29040 }, { "epoch": 3.07, - "grad_norm": 5.233509063720703, + "grad_norm": 33.166290283203125, "learning_rate": 1.3864188404266555e-05, - "loss": 0.832, + "loss": 1.408, "step": 29050 }, { "epoch": 3.07, - "grad_norm": 10.632216453552246, + "grad_norm": 24.329448699951172, "learning_rate": 1.3862076248811913e-05, - "loss": 0.5842, + "loss": 1.0115, "step": 29060 }, { "epoch": 3.07, - "grad_norm": 15.647287368774414, + "grad_norm": 51.86185073852539, "learning_rate": 1.3859964093357272e-05, - "loss": 0.9921, + "loss": 1.5498, "step": 29070 }, { "epoch": 3.07, - "grad_norm": 15.691277503967285, + "grad_norm": 29.239370346069336, "learning_rate": 1.3857851937902631e-05, - "loss": 0.7748, + "loss": 1.2512, "step": 29080 }, { "epoch": 3.07, - "grad_norm": 15.580130577087402, + "grad_norm": 52.7990608215332, "learning_rate": 1.385573978244799e-05, - "loss": 0.5696, + "loss": 1.5124, "step": 29090 }, { "epoch": 3.07, - "grad_norm": 18.200260162353516, + "grad_norm": 37.4036750793457, "learning_rate": 1.3853627626993347e-05, - "loss": 0.7805, + "loss": 1.2336, "step": 29100 }, { "epoch": 3.07, - "grad_norm": 19.52519989013672, + "grad_norm": 43.80769729614258, "learning_rate": 1.3851515471538706e-05, - "loss": 0.7446, + "loss": 1.3656, "step": 29110 }, { "epoch": 3.08, - "grad_norm": 20.357891082763672, + "grad_norm": 44.39107131958008, "learning_rate": 1.3849403316084064e-05, - "loss": 0.6018, + "loss": 1.3722, "step": 29120 }, { "epoch": 3.08, - "grad_norm": 19.408798217773438, + "grad_norm": 59.95209503173828, "learning_rate": 1.3847291160629423e-05, - "loss": 0.3534, + "loss": 0.9495, "step": 29130 }, { "epoch": 3.08, - "grad_norm": 12.919946670532227, + "grad_norm": 56.32801055908203, "learning_rate": 1.3845179005174782e-05, - "loss": 0.429, + "loss": 1.2893, "step": 29140 }, { "epoch": 3.08, - "grad_norm": 7.8954548835754395, + "grad_norm": 19.813323974609375, "learning_rate": 1.3843066849720142e-05, - "loss": 0.6502, + "loss": 1.2269, "step": 29150 }, { "epoch": 3.08, - "grad_norm": 25.78050422668457, + "grad_norm": 41.630123138427734, "learning_rate": 1.3840954694265498e-05, - "loss": 0.5939, + "loss": 1.182, "step": 29160 }, { "epoch": 3.08, - "grad_norm": 11.701249122619629, + "grad_norm": 29.65980339050293, "learning_rate": 1.3838842538810856e-05, - "loss": 0.6007, + "loss": 1.0312, "step": 29170 }, { "epoch": 3.08, - "grad_norm": 12.499687194824219, + "grad_norm": 16.703475952148438, "learning_rate": 1.3836730383356215e-05, - "loss": 0.2542, + "loss": 1.1161, "step": 29180 }, { "epoch": 3.08, - "grad_norm": 10.98568344116211, + "grad_norm": 38.66923904418945, "learning_rate": 1.3834618227901576e-05, - "loss": 0.4844, + "loss": 0.8411, "step": 29190 }, { "epoch": 3.08, - "grad_norm": 12.652403831481934, + "grad_norm": 33.04026412963867, "learning_rate": 1.3832506072446935e-05, - "loss": 0.5801, + "loss": 1.2396, "step": 29200 }, { "epoch": 3.08, - "grad_norm": 15.532081604003906, + "grad_norm": 40.11335372924805, "learning_rate": 1.383039391699229e-05, - "loss": 0.4917, + "loss": 1.1874, "step": 29210 }, { "epoch": 3.09, - "grad_norm": 8.67465591430664, + "grad_norm": 28.2982234954834, "learning_rate": 1.382828176153765e-05, - "loss": 0.6832, + "loss": 1.3016, "step": 29220 }, { "epoch": 3.09, - "grad_norm": 17.37545394897461, + "grad_norm": 40.415504455566406, "learning_rate": 1.382616960608301e-05, - "loss": 0.5418, + "loss": 1.3784, "step": 29230 }, { "epoch": 3.09, - "grad_norm": 0.4559379816055298, + "grad_norm": 29.88301658630371, "learning_rate": 1.3824057450628368e-05, - "loss": 0.6371, + "loss": 1.214, "step": 29240 }, { "epoch": 3.09, - "grad_norm": 4.56401252746582, + "grad_norm": 38.24467086791992, "learning_rate": 1.3821945295173727e-05, - "loss": 0.487, + "loss": 1.2867, "step": 29250 }, { "epoch": 3.09, - "grad_norm": 10.318737030029297, + "grad_norm": 37.964839935302734, "learning_rate": 1.3819833139719085e-05, - "loss": 0.3638, + "loss": 0.9877, "step": 29260 }, { "epoch": 3.09, - "grad_norm": 8.616753578186035, + "grad_norm": 34.18330001831055, "learning_rate": 1.3817720984264443e-05, - "loss": 0.604, + "loss": 1.2593, "step": 29270 }, { "epoch": 3.09, - "grad_norm": 21.33279037475586, + "grad_norm": 38.997169494628906, "learning_rate": 1.3815608828809801e-05, - "loss": 0.7288, + "loss": 1.4471, "step": 29280 }, { "epoch": 3.09, - "grad_norm": 9.924339294433594, + "grad_norm": 42.31978988647461, "learning_rate": 1.381349667335516e-05, - "loss": 0.6051, + "loss": 1.1679, "step": 29290 }, { "epoch": 3.09, - "grad_norm": 7.844449043273926, + "grad_norm": 19.465810775756836, "learning_rate": 1.3811384517900519e-05, - "loss": 0.4168, + "loss": 1.1806, "step": 29300 }, { "epoch": 3.1, - "grad_norm": 15.700997352600098, + "grad_norm": 38.354103088378906, "learning_rate": 1.3809272362445878e-05, - "loss": 0.5831, + "loss": 1.5269, "step": 29310 }, { "epoch": 3.1, - "grad_norm": 20.81985092163086, + "grad_norm": 28.304994583129883, "learning_rate": 1.3807160206991235e-05, - "loss": 0.6173, + "loss": 1.3499, "step": 29320 }, { "epoch": 3.1, - "grad_norm": 18.41353988647461, + "grad_norm": 37.07611846923828, "learning_rate": 1.3805048051536594e-05, - "loss": 0.6608, + "loss": 1.486, "step": 29330 }, { "epoch": 3.1, - "grad_norm": 1.0028836727142334, + "grad_norm": 39.09781265258789, "learning_rate": 1.3802935896081952e-05, - "loss": 0.5767, + "loss": 1.1462, "step": 29340 }, { "epoch": 3.1, - "grad_norm": 20.599607467651367, + "grad_norm": 42.50492858886719, "learning_rate": 1.3800823740627311e-05, - "loss": 0.6243, + "loss": 1.3483, "step": 29350 }, { "epoch": 3.1, - "grad_norm": 3.649883985519409, + "grad_norm": 53.812435150146484, "learning_rate": 1.379871158517267e-05, - "loss": 0.563, + "loss": 1.3206, "step": 29360 }, { "epoch": 3.1, - "grad_norm": 15.762778282165527, + "grad_norm": 29.418079376220703, "learning_rate": 1.3796599429718029e-05, - "loss": 0.7306, + "loss": 0.989, "step": 29370 }, { "epoch": 3.1, - "grad_norm": 14.36619758605957, + "grad_norm": 38.60914993286133, "learning_rate": 1.3794487274263386e-05, - "loss": 0.5814, + "loss": 1.2397, "step": 29380 }, { "epoch": 3.1, - "grad_norm": 9.415862083435059, + "grad_norm": 34.58033752441406, "learning_rate": 1.3792375118808745e-05, - "loss": 0.5355, + "loss": 1.1879, "step": 29390 }, { "epoch": 3.1, - "grad_norm": 11.930912971496582, + "grad_norm": 38.503684997558594, "learning_rate": 1.3790262963354103e-05, - "loss": 0.5025, + "loss": 1.3051, "step": 29400 }, { "epoch": 3.11, - "grad_norm": 0.7898351550102234, + "grad_norm": 32.70675277709961, "learning_rate": 1.3788150807899464e-05, - "loss": 0.6494, + "loss": 1.2046, "step": 29410 }, { "epoch": 3.11, - "grad_norm": 18.134401321411133, + "grad_norm": 37.64572525024414, "learning_rate": 1.3786038652444823e-05, - "loss": 0.6753, + "loss": 1.2549, "step": 29420 }, { "epoch": 3.11, - "grad_norm": 7.331478118896484, + "grad_norm": 38.92711639404297, "learning_rate": 1.3783926496990178e-05, - "loss": 0.5334, + "loss": 1.1605, "step": 29430 }, { "epoch": 3.11, - "grad_norm": 14.50686264038086, + "grad_norm": 21.48169708251953, "learning_rate": 1.3781814341535538e-05, - "loss": 0.6247, + "loss": 1.0211, "step": 29440 }, { "epoch": 3.11, - "grad_norm": 6.073133945465088, + "grad_norm": 26.45429801940918, "learning_rate": 1.3779702186080897e-05, - "loss": 0.5925, + "loss": 1.592, "step": 29450 }, { "epoch": 3.11, - "grad_norm": 16.701534271240234, + "grad_norm": 19.443544387817383, "learning_rate": 1.3777590030626256e-05, - "loss": 0.9222, + "loss": 1.4135, "step": 29460 }, { "epoch": 3.11, - "grad_norm": 13.251472473144531, + "grad_norm": 38.56719207763672, "learning_rate": 1.3775477875171615e-05, - "loss": 0.6737, + "loss": 1.3269, "step": 29470 }, { "epoch": 3.11, - "grad_norm": 1.3422198295593262, + "grad_norm": 39.73451614379883, "learning_rate": 1.3773365719716972e-05, - "loss": 0.5229, + "loss": 1.5064, "step": 29480 }, { "epoch": 3.11, - "grad_norm": 26.670019149780273, + "grad_norm": 45.837188720703125, "learning_rate": 1.377125356426233e-05, - "loss": 0.7943, + "loss": 1.3459, "step": 29490 }, { "epoch": 3.12, - "grad_norm": 7.007087707519531, + "grad_norm": 44.959564208984375, "learning_rate": 1.376914140880769e-05, - "loss": 0.5741, + "loss": 1.3243, "step": 29500 }, { "epoch": 3.12, - "grad_norm": 11.226361274719238, + "grad_norm": 35.62355422973633, "learning_rate": 1.3767029253353048e-05, - "loss": 0.5182, + "loss": 1.3363, "step": 29510 }, { "epoch": 3.12, - "grad_norm": 16.71345329284668, + "grad_norm": 39.03450393676758, "learning_rate": 1.3764917097898407e-05, - "loss": 0.8276, + "loss": 1.3655, "step": 29520 }, { "epoch": 3.12, - "grad_norm": 17.32223129272461, + "grad_norm": 32.393798828125, "learning_rate": 1.3762804942443766e-05, - "loss": 0.5493, + "loss": 1.0735, "step": 29530 }, { "epoch": 3.12, - "grad_norm": 11.516748428344727, + "grad_norm": 43.74066162109375, "learning_rate": 1.3760692786989123e-05, - "loss": 0.3788, + "loss": 1.0661, "step": 29540 }, { "epoch": 3.12, - "grad_norm": 15.206853866577148, + "grad_norm": 57.999935150146484, "learning_rate": 1.3758580631534482e-05, - "loss": 0.62, + "loss": 1.0851, "step": 29550 }, { "epoch": 3.12, - "grad_norm": 8.351865768432617, + "grad_norm": 28.749713897705078, "learning_rate": 1.375646847607984e-05, - "loss": 0.435, + "loss": 1.009, "step": 29560 }, { "epoch": 3.12, - "grad_norm": 8.653533935546875, + "grad_norm": 18.015161514282227, "learning_rate": 1.37543563206252e-05, - "loss": 0.564, + "loss": 1.3205, "step": 29570 }, { "epoch": 3.12, - "grad_norm": 0.2627885043621063, + "grad_norm": 22.712095260620117, "learning_rate": 1.3752244165170558e-05, - "loss": 0.5446, + "loss": 1.1787, "step": 29580 }, { "epoch": 3.12, - "grad_norm": 10.860404968261719, + "grad_norm": 37.11049270629883, "learning_rate": 1.3750132009715915e-05, - "loss": 0.4945, + "loss": 1.017, "step": 29590 }, { "epoch": 3.13, - "grad_norm": 21.438852310180664, + "grad_norm": 59.27043151855469, "learning_rate": 1.3748019854261274e-05, - "loss": 0.6669, + "loss": 1.2456, "step": 29600 }, { "epoch": 3.13, - "grad_norm": 1.9559638500213623, + "grad_norm": 41.00160217285156, "learning_rate": 1.3745907698806633e-05, - "loss": 0.608, + "loss": 1.2142, "step": 29610 }, { "epoch": 3.13, - "grad_norm": 0.2722570300102234, + "grad_norm": 37.72785568237305, "learning_rate": 1.3743795543351991e-05, - "loss": 0.5248, + "loss": 1.1125, "step": 29620 }, { "epoch": 3.13, - "grad_norm": 9.208821296691895, + "grad_norm": 35.45351028442383, "learning_rate": 1.3741683387897352e-05, - "loss": 0.3811, + "loss": 1.0263, "step": 29630 }, { "epoch": 3.13, - "grad_norm": 15.506443977355957, + "grad_norm": 42.341453552246094, "learning_rate": 1.373957123244271e-05, - "loss": 0.5281, + "loss": 1.2222, "step": 29640 }, { "epoch": 3.13, - "grad_norm": 8.94096851348877, + "grad_norm": 35.28118133544922, "learning_rate": 1.3737459076988066e-05, - "loss": 0.7421, + "loss": 1.1436, "step": 29650 }, { "epoch": 3.13, - "grad_norm": 0.9412774443626404, + "grad_norm": 27.22231674194336, "learning_rate": 1.3735346921533425e-05, - "loss": 0.471, + "loss": 1.1243, "step": 29660 }, { "epoch": 3.13, - "grad_norm": 8.308526992797852, + "grad_norm": 25.288768768310547, "learning_rate": 1.3733234766078785e-05, - "loss": 0.6593, + "loss": 1.1196, "step": 29670 }, { "epoch": 3.13, - "grad_norm": 13.360572814941406, + "grad_norm": 31.792753219604492, "learning_rate": 1.3731122610624144e-05, - "loss": 0.4227, + "loss": 1.1119, "step": 29680 }, { "epoch": 3.14, - "grad_norm": 17.194040298461914, + "grad_norm": 52.330196380615234, "learning_rate": 1.3729010455169503e-05, - "loss": 0.4251, + "loss": 0.9743, "step": 29690 }, { "epoch": 3.14, - "grad_norm": 8.347381591796875, + "grad_norm": 61.5836181640625, "learning_rate": 1.372689829971486e-05, - "loss": 0.6125, + "loss": 1.5725, "step": 29700 }, { "epoch": 3.14, - "grad_norm": 6.694857597351074, + "grad_norm": 37.94447326660156, "learning_rate": 1.3724786144260219e-05, - "loss": 0.2681, + "loss": 0.825, "step": 29710 }, { "epoch": 3.14, - "grad_norm": 7.873359680175781, + "grad_norm": 38.05294418334961, "learning_rate": 1.3722673988805577e-05, - "loss": 0.485, + "loss": 1.1983, "step": 29720 }, { "epoch": 3.14, - "grad_norm": 9.894848823547363, + "grad_norm": 19.595611572265625, "learning_rate": 1.3720561833350936e-05, - "loss": 0.6092, + "loss": 1.1867, "step": 29730 }, { "epoch": 3.14, - "grad_norm": 5.3491597175598145, + "grad_norm": 32.462947845458984, "learning_rate": 1.3718449677896295e-05, - "loss": 0.5229, + "loss": 1.3053, "step": 29740 }, { "epoch": 3.14, - "grad_norm": 6.49989652633667, + "grad_norm": 39.060855865478516, "learning_rate": 1.3716337522441652e-05, - "loss": 0.8416, + "loss": 1.2897, "step": 29750 }, { "epoch": 3.14, - "grad_norm": 10.551483154296875, + "grad_norm": 47.93277359008789, "learning_rate": 1.3714225366987011e-05, - "loss": 0.6065, + "loss": 1.0652, "step": 29760 }, { "epoch": 3.14, - "grad_norm": 11.280160903930664, + "grad_norm": 42.892738342285156, "learning_rate": 1.371211321153237e-05, - "loss": 0.7855, + "loss": 1.1299, "step": 29770 }, { "epoch": 3.14, - "grad_norm": 0.9085181355476379, + "grad_norm": 53.10535430908203, "learning_rate": 1.3710001056077728e-05, - "loss": 0.634, + "loss": 1.1904, "step": 29780 }, { "epoch": 3.15, - "grad_norm": 14.048929214477539, + "grad_norm": 50.00741958618164, "learning_rate": 1.3707888900623087e-05, - "loss": 0.4438, + "loss": 1.0486, "step": 29790 }, { "epoch": 3.15, - "grad_norm": 17.52055549621582, + "grad_norm": 27.10503387451172, "learning_rate": 1.3705776745168446e-05, - "loss": 0.6823, + "loss": 1.3474, "step": 29800 }, { "epoch": 3.15, - "grad_norm": 14.94727897644043, + "grad_norm": 41.306251525878906, "learning_rate": 1.3703664589713803e-05, - "loss": 0.5097, + "loss": 1.2032, "step": 29810 }, { "epoch": 3.15, - "grad_norm": 6.6156206130981445, + "grad_norm": 42.035457611083984, "learning_rate": 1.3701552434259162e-05, - "loss": 0.5464, + "loss": 1.502, "step": 29820 }, { "epoch": 3.15, - "grad_norm": 15.54917049407959, + "grad_norm": 29.919084548950195, "learning_rate": 1.369944027880452e-05, - "loss": 0.6349, + "loss": 1.2588, "step": 29830 }, { "epoch": 3.15, - "grad_norm": 15.793082237243652, + "grad_norm": 34.130306243896484, "learning_rate": 1.369732812334988e-05, - "loss": 0.6664, + "loss": 1.3683, "step": 29840 }, { "epoch": 3.15, - "grad_norm": 17.47838592529297, + "grad_norm": 33.04670333862305, "learning_rate": 1.3695215967895238e-05, - "loss": 0.6106, + "loss": 1.008, "step": 29850 }, { "epoch": 3.15, - "grad_norm": 14.592084884643555, + "grad_norm": 45.90314865112305, "learning_rate": 1.3693103812440595e-05, - "loss": 0.5032, + "loss": 1.3413, "step": 29860 }, { "epoch": 3.15, - "grad_norm": 5.749638557434082, + "grad_norm": 38.68691635131836, "learning_rate": 1.3690991656985954e-05, - "loss": 0.4454, + "loss": 1.004, "step": 29870 }, { "epoch": 3.16, - "grad_norm": 42.40886688232422, + "grad_norm": 45.462989807128906, "learning_rate": 1.3688879501531313e-05, - "loss": 1.0193, + "loss": 1.485, "step": 29880 }, { "epoch": 3.16, - "grad_norm": 14.62541389465332, + "grad_norm": 34.988521575927734, "learning_rate": 1.3686767346076673e-05, - "loss": 0.5327, + "loss": 0.8391, "step": 29890 }, { "epoch": 3.16, - "grad_norm": 39.76918411254883, + "grad_norm": 38.75834655761719, "learning_rate": 1.3684655190622032e-05, - "loss": 0.3453, + "loss": 1.0245, "step": 29900 }, { "epoch": 3.16, - "grad_norm": 1.0914645195007324, + "grad_norm": 37.47809600830078, "learning_rate": 1.3682543035167391e-05, - "loss": 0.6834, + "loss": 1.0989, "step": 29910 }, { "epoch": 3.16, - "grad_norm": 13.93621826171875, + "grad_norm": 41.93830871582031, "learning_rate": 1.3680430879712748e-05, - "loss": 0.5528, + "loss": 1.1685, "step": 29920 }, { "epoch": 3.16, - "grad_norm": 19.47856330871582, + "grad_norm": 35.35197448730469, "learning_rate": 1.3678318724258107e-05, - "loss": 0.394, + "loss": 0.9762, "step": 29930 }, { "epoch": 3.16, - "grad_norm": 11.922087669372559, + "grad_norm": 40.024452209472656, "learning_rate": 1.3676206568803466e-05, - "loss": 0.4812, + "loss": 0.8122, "step": 29940 }, { "epoch": 3.16, - "grad_norm": 4.128494739532471, + "grad_norm": 24.665132522583008, "learning_rate": 1.3674094413348824e-05, - "loss": 0.4938, + "loss": 1.1225, "step": 29950 }, { "epoch": 3.16, - "grad_norm": 9.689383506774902, + "grad_norm": 31.690336227416992, "learning_rate": 1.3671982257894183e-05, - "loss": 0.269, + "loss": 0.8274, "step": 29960 }, { "epoch": 3.17, - "grad_norm": 20.98430824279785, + "grad_norm": 45.44163131713867, "learning_rate": 1.366987010243954e-05, - "loss": 0.7949, + "loss": 1.5223, "step": 29970 }, { "epoch": 3.17, - "grad_norm": 6.230413436889648, + "grad_norm": 34.744510650634766, "learning_rate": 1.3667757946984899e-05, - "loss": 0.4252, + "loss": 1.108, "step": 29980 }, { "epoch": 3.17, - "grad_norm": 28.438983917236328, + "grad_norm": 30.9404239654541, "learning_rate": 1.3665645791530258e-05, - "loss": 0.6362, + "loss": 1.2542, "step": 29990 }, { "epoch": 3.17, - "grad_norm": 31.589263916015625, + "grad_norm": 19.96123695373535, "learning_rate": 1.3663533636075616e-05, - "loss": 0.3443, + "loss": 1.0031, "step": 30000 }, { "epoch": 3.17, - "grad_norm": 19.344970703125, + "grad_norm": 40.086822509765625, "learning_rate": 1.3661421480620975e-05, - "loss": 0.6407, + "loss": 1.3598, "step": 30010 }, { "epoch": 3.17, - "grad_norm": 14.755959510803223, + "grad_norm": 39.96134948730469, "learning_rate": 1.3659309325166334e-05, - "loss": 0.95, + "loss": 1.1379, "step": 30020 }, { "epoch": 3.17, - "grad_norm": 0.5894179344177246, + "grad_norm": 10.120686531066895, "learning_rate": 1.3657197169711691e-05, - "loss": 0.4943, + "loss": 0.798, "step": 30030 }, { "epoch": 3.17, - "grad_norm": 1.225978970527649, + "grad_norm": 29.382349014282227, "learning_rate": 1.365508501425705e-05, - "loss": 0.4134, + "loss": 1.0281, "step": 30040 }, { "epoch": 3.17, - "grad_norm": 17.14017677307129, + "grad_norm": 28.023977279663086, "learning_rate": 1.3652972858802409e-05, - "loss": 0.5291, + "loss": 1.2055, "step": 30050 }, { "epoch": 3.17, - "grad_norm": 9.919217109680176, + "grad_norm": 45.72953414916992, "learning_rate": 1.3650860703347767e-05, - "loss": 0.9823, + "loss": 1.3125, "step": 30060 }, { "epoch": 3.18, - "grad_norm": 10.449519157409668, + "grad_norm": 31.86236572265625, "learning_rate": 1.3648748547893126e-05, - "loss": 0.5512, + "loss": 1.3173, "step": 30070 }, { "epoch": 3.18, - "grad_norm": 11.489480018615723, + "grad_norm": 38.125606536865234, "learning_rate": 1.3646636392438483e-05, - "loss": 0.5192, + "loss": 1.1587, "step": 30080 }, { "epoch": 3.18, - "grad_norm": 5.20477294921875, + "grad_norm": 43.475101470947266, "learning_rate": 1.3644524236983842e-05, - "loss": 0.3295, + "loss": 1.4682, "step": 30090 }, { "epoch": 3.18, - "grad_norm": 27.76072883605957, + "grad_norm": 53.32619094848633, "learning_rate": 1.3642412081529201e-05, - "loss": 0.6472, + "loss": 1.6849, "step": 30100 }, { "epoch": 3.18, - "grad_norm": 18.04008674621582, + "grad_norm": 60.34459686279297, "learning_rate": 1.3640299926074561e-05, - "loss": 0.573, + "loss": 1.2289, "step": 30110 }, { "epoch": 3.18, - "grad_norm": 0.33164429664611816, + "grad_norm": 42.283321380615234, "learning_rate": 1.363818777061992e-05, - "loss": 0.5839, + "loss": 1.5127, "step": 30120 }, { "epoch": 3.18, - "grad_norm": 19.49367332458496, + "grad_norm": 37.493892669677734, "learning_rate": 1.3636075615165276e-05, - "loss": 0.732, + "loss": 1.254, "step": 30130 }, { "epoch": 3.18, - "grad_norm": 15.310988426208496, + "grad_norm": 42.95746612548828, "learning_rate": 1.3633963459710636e-05, - "loss": 0.5391, + "loss": 1.1826, "step": 30140 }, { "epoch": 3.18, - "grad_norm": 24.28915023803711, + "grad_norm": 25.675779342651367, "learning_rate": 1.3631851304255995e-05, - "loss": 0.3783, + "loss": 1.0613, "step": 30150 }, { "epoch": 3.19, - "grad_norm": 35.51908493041992, + "grad_norm": 41.19252395629883, "learning_rate": 1.3629739148801354e-05, - "loss": 0.7806, + "loss": 1.4671, "step": 30160 }, { "epoch": 3.19, - "grad_norm": 30.001216888427734, + "grad_norm": 50.31441116333008, "learning_rate": 1.3627626993346712e-05, - "loss": 0.3373, + "loss": 1.179, "step": 30170 }, { "epoch": 3.19, - "grad_norm": 31.133785247802734, + "grad_norm": 49.00692367553711, "learning_rate": 1.3625514837892071e-05, - "loss": 0.9022, + "loss": 1.3986, "step": 30180 }, { "epoch": 3.19, - "grad_norm": 7.312440872192383, + "grad_norm": 34.65723419189453, "learning_rate": 1.3623402682437428e-05, - "loss": 0.5194, + "loss": 1.3771, "step": 30190 }, { "epoch": 3.19, - "grad_norm": 14.83166217803955, + "grad_norm": 34.94011688232422, "learning_rate": 1.3621290526982787e-05, - "loss": 0.4628, + "loss": 1.2875, "step": 30200 }, { "epoch": 3.19, - "grad_norm": 16.609039306640625, + "grad_norm": 48.9195556640625, "learning_rate": 1.3619178371528146e-05, - "loss": 0.8265, + "loss": 1.5449, "step": 30210 }, { "epoch": 3.19, - "grad_norm": 6.388683319091797, + "grad_norm": 34.89748764038086, "learning_rate": 1.3617066216073505e-05, - "loss": 0.5695, + "loss": 1.0505, "step": 30220 }, { "epoch": 3.19, - "grad_norm": 11.495753288269043, + "grad_norm": 41.98260498046875, "learning_rate": 1.3614954060618863e-05, - "loss": 0.3658, + "loss": 1.0534, "step": 30230 }, { "epoch": 3.19, - "grad_norm": 0.22838224470615387, + "grad_norm": 37.8179931640625, "learning_rate": 1.361284190516422e-05, - "loss": 0.9581, + "loss": 1.4226, "step": 30240 }, { "epoch": 3.19, - "grad_norm": 16.308216094970703, + "grad_norm": 46.71440124511719, "learning_rate": 1.361072974970958e-05, - "loss": 0.4808, + "loss": 1.1857, "step": 30250 }, { "epoch": 3.2, - "grad_norm": 21.855131149291992, + "grad_norm": 52.57198715209961, "learning_rate": 1.3608617594254938e-05, - "loss": 0.4253, + "loss": 1.2019, "step": 30260 }, { "epoch": 3.2, - "grad_norm": 18.902019500732422, + "grad_norm": 23.533933639526367, "learning_rate": 1.3606505438800297e-05, - "loss": 0.5353, + "loss": 1.1016, "step": 30270 }, { "epoch": 3.2, - "grad_norm": 11.138187408447266, + "grad_norm": 41.97760772705078, "learning_rate": 1.3604393283345656e-05, - "loss": 0.4979, + "loss": 1.1391, "step": 30280 }, { "epoch": 3.2, - "grad_norm": 10.679903984069824, + "grad_norm": 17.186166763305664, "learning_rate": 1.3602281127891014e-05, - "loss": 0.5088, + "loss": 1.0463, "step": 30290 }, { "epoch": 3.2, - "grad_norm": 23.25828742980957, + "grad_norm": 49.099185943603516, "learning_rate": 1.3600168972436371e-05, - "loss": 0.5292, + "loss": 1.4787, "step": 30300 }, { "epoch": 3.2, - "grad_norm": 13.238201141357422, + "grad_norm": 43.41545104980469, "learning_rate": 1.359805681698173e-05, - "loss": 0.5632, + "loss": 1.2829, "step": 30310 }, { "epoch": 3.2, - "grad_norm": 0.8043118119239807, + "grad_norm": 20.46790313720703, "learning_rate": 1.3595944661527089e-05, - "loss": 0.4834, + "loss": 1.1545, "step": 30320 }, { "epoch": 3.2, - "grad_norm": 9.825379371643066, + "grad_norm": 52.33667755126953, "learning_rate": 1.3593832506072448e-05, - "loss": 0.542, + "loss": 1.0867, "step": 30330 }, { "epoch": 3.2, - "grad_norm": 3.943441152572632, + "grad_norm": 29.335420608520508, "learning_rate": 1.3591720350617808e-05, - "loss": 0.4696, + "loss": 1.2202, "step": 30340 }, { "epoch": 3.21, - "grad_norm": 11.142730712890625, + "grad_norm": 49.99092102050781, "learning_rate": 1.3589608195163164e-05, - "loss": 0.8737, + "loss": 1.144, "step": 30350 }, { "epoch": 3.21, - "grad_norm": 17.834259033203125, + "grad_norm": 40.05434799194336, "learning_rate": 1.3587496039708522e-05, - "loss": 0.6479, + "loss": 1.2279, "step": 30360 }, { "epoch": 3.21, - "grad_norm": 14.39127254486084, + "grad_norm": 33.35226058959961, "learning_rate": 1.3585383884253883e-05, - "loss": 0.4475, + "loss": 1.0778, "step": 30370 }, { "epoch": 3.21, - "grad_norm": 10.393872261047363, + "grad_norm": 43.48418426513672, "learning_rate": 1.3583271728799242e-05, - "loss": 0.4985, + "loss": 1.1668, "step": 30380 }, { "epoch": 3.21, - "grad_norm": 10.312480926513672, + "grad_norm": 40.93688201904297, "learning_rate": 1.35811595733446e-05, - "loss": 0.513, + "loss": 1.2762, "step": 30390 }, { "epoch": 3.21, - "grad_norm": 15.064556121826172, + "grad_norm": 35.23093032836914, "learning_rate": 1.3579047417889957e-05, - "loss": 0.431, + "loss": 0.923, "step": 30400 }, { "epoch": 3.21, - "grad_norm": 19.035966873168945, + "grad_norm": 46.5525016784668, "learning_rate": 1.3576935262435316e-05, - "loss": 0.3656, + "loss": 1.0173, "step": 30410 }, { "epoch": 3.21, - "grad_norm": 4.170107841491699, + "grad_norm": 33.099647521972656, "learning_rate": 1.3574823106980675e-05, - "loss": 0.4536, + "loss": 1.0374, "step": 30420 }, { "epoch": 3.21, - "grad_norm": 16.92786979675293, + "grad_norm": 39.62485885620117, "learning_rate": 1.3572710951526034e-05, - "loss": 0.8703, + "loss": 1.5861, "step": 30430 }, { "epoch": 3.21, - "grad_norm": 26.65851402282715, + "grad_norm": 31.824541091918945, "learning_rate": 1.3570598796071393e-05, - "loss": 0.5187, + "loss": 1.0498, "step": 30440 }, { "epoch": 3.22, - "grad_norm": 7.786027431488037, + "grad_norm": 25.24896812438965, "learning_rate": 1.3568486640616751e-05, - "loss": 0.7678, + "loss": 1.9919, "step": 30450 }, { "epoch": 3.22, - "grad_norm": 7.566701412200928, + "grad_norm": 40.66785430908203, "learning_rate": 1.3566374485162108e-05, - "loss": 0.7743, + "loss": 1.5422, "step": 30460 }, { "epoch": 3.22, - "grad_norm": 9.506001472473145, + "grad_norm": 35.262596130371094, "learning_rate": 1.3564262329707467e-05, - "loss": 0.5261, + "loss": 1.2007, "step": 30470 }, { "epoch": 3.22, - "grad_norm": 14.14596939086914, + "grad_norm": 17.884431838989258, "learning_rate": 1.3562150174252826e-05, - "loss": 0.5249, + "loss": 0.8776, "step": 30480 }, { "epoch": 3.22, - "grad_norm": 11.18157958984375, + "grad_norm": 44.00569152832031, "learning_rate": 1.3560038018798185e-05, - "loss": 0.7922, + "loss": 1.3003, "step": 30490 }, { "epoch": 3.22, - "grad_norm": 33.417572021484375, + "grad_norm": 43.890357971191406, "learning_rate": 1.3557925863343544e-05, - "loss": 0.7054, + "loss": 1.1915, "step": 30500 }, { "epoch": 3.22, - "grad_norm": 31.890037536621094, + "grad_norm": 36.549495697021484, "learning_rate": 1.35558137078889e-05, - "loss": 0.5135, + "loss": 1.2943, "step": 30510 }, { "epoch": 3.22, - "grad_norm": 22.245563507080078, + "grad_norm": 37.846927642822266, "learning_rate": 1.355370155243426e-05, - "loss": 0.5695, + "loss": 1.3537, "step": 30520 }, { "epoch": 3.22, - "grad_norm": 10.238027572631836, + "grad_norm": 31.14398956298828, "learning_rate": 1.3551589396979618e-05, - "loss": 0.5059, + "loss": 1.0602, "step": 30530 }, { "epoch": 3.23, - "grad_norm": 16.298450469970703, + "grad_norm": 39.4617919921875, "learning_rate": 1.3549477241524977e-05, - "loss": 0.4594, + "loss": 1.0075, "step": 30540 }, { "epoch": 3.23, - "grad_norm": 14.593399047851562, + "grad_norm": 57.28291320800781, "learning_rate": 1.3547365086070336e-05, - "loss": 0.8942, + "loss": 1.4342, "step": 30550 }, { "epoch": 3.23, - "grad_norm": 5.526491165161133, + "grad_norm": 53.06760787963867, "learning_rate": 1.3545252930615696e-05, - "loss": 0.4925, + "loss": 1.096, "step": 30560 }, { "epoch": 3.23, - "grad_norm": 18.12894630432129, + "grad_norm": 39.04314041137695, "learning_rate": 1.3543140775161052e-05, - "loss": 0.4959, + "loss": 1.1058, "step": 30570 }, { "epoch": 3.23, - "grad_norm": 19.471569061279297, + "grad_norm": 46.92134094238281, "learning_rate": 1.354102861970641e-05, - "loss": 0.8541, + "loss": 1.5952, "step": 30580 }, { "epoch": 3.23, - "grad_norm": 9.874066352844238, + "grad_norm": 31.584653854370117, "learning_rate": 1.3538916464251771e-05, - "loss": 0.2791, + "loss": 1.2113, "step": 30590 }, { "epoch": 3.23, - "grad_norm": 7.792309284210205, + "grad_norm": 50.475746154785156, "learning_rate": 1.353680430879713e-05, - "loss": 0.4556, + "loss": 1.2945, "step": 30600 }, { "epoch": 3.23, - "grad_norm": 0.7725560069084167, + "grad_norm": 22.636985778808594, "learning_rate": 1.3534692153342488e-05, - "loss": 0.5948, + "loss": 1.4154, "step": 30610 }, { "epoch": 3.23, - "grad_norm": 37.39490509033203, + "grad_norm": 26.4505615234375, "learning_rate": 1.3532579997887846e-05, - "loss": 0.8824, + "loss": 1.39, "step": 30620 }, { "epoch": 3.23, - "grad_norm": 0.7608240246772766, + "grad_norm": 44.659759521484375, "learning_rate": 1.3530467842433204e-05, - "loss": 0.6253, + "loss": 1.4316, "step": 30630 }, { "epoch": 3.24, - "grad_norm": 8.07474136352539, + "grad_norm": 53.12766647338867, "learning_rate": 1.3528355686978563e-05, - "loss": 0.2666, + "loss": 1.1045, "step": 30640 }, { "epoch": 3.24, - "grad_norm": 14.736668586730957, + "grad_norm": 41.27557373046875, "learning_rate": 1.3526243531523922e-05, - "loss": 0.8036, + "loss": 1.2738, "step": 30650 }, { "epoch": 3.24, - "grad_norm": 10.183141708374023, + "grad_norm": 35.26319122314453, "learning_rate": 1.352413137606928e-05, - "loss": 0.5474, + "loss": 1.3452, "step": 30660 }, { "epoch": 3.24, - "grad_norm": 21.390926361083984, + "grad_norm": 36.14268112182617, "learning_rate": 1.352201922061464e-05, - "loss": 0.7192, + "loss": 1.3566, "step": 30670 }, { "epoch": 3.24, - "grad_norm": 0.806415319442749, + "grad_norm": 29.69717025756836, "learning_rate": 1.3519907065159997e-05, - "loss": 0.4718, + "loss": 1.3754, "step": 30680 }, { "epoch": 3.24, - "grad_norm": 9.994848251342773, + "grad_norm": 42.25239944458008, "learning_rate": 1.3517794909705355e-05, - "loss": 0.791, + "loss": 1.1961, "step": 30690 }, { "epoch": 3.24, - "grad_norm": 11.543426513671875, + "grad_norm": 28.71143913269043, "learning_rate": 1.3515682754250714e-05, - "loss": 0.4609, + "loss": 1.0353, "step": 30700 }, { "epoch": 3.24, - "grad_norm": 8.64100170135498, + "grad_norm": 31.336172103881836, "learning_rate": 1.3513570598796073e-05, - "loss": 0.5971, + "loss": 1.4956, "step": 30710 }, { "epoch": 3.24, - "grad_norm": 14.466665267944336, + "grad_norm": 31.615079879760742, "learning_rate": 1.3511458443341432e-05, - "loss": 0.5354, + "loss": 1.3673, "step": 30720 }, { "epoch": 3.25, - "grad_norm": 22.02457046508789, + "grad_norm": 45.632564544677734, "learning_rate": 1.3509346287886789e-05, - "loss": 0.5019, + "loss": 1.2874, "step": 30730 }, { "epoch": 3.25, - "grad_norm": 14.422812461853027, + "grad_norm": 35.02996826171875, "learning_rate": 1.3507234132432147e-05, - "loss": 0.8244, + "loss": 1.186, "step": 30740 }, { "epoch": 3.25, - "grad_norm": 15.857705116271973, + "grad_norm": 24.733501434326172, "learning_rate": 1.3505121976977506e-05, - "loss": 0.5531, + "loss": 1.2274, "step": 30750 }, { "epoch": 3.25, - "grad_norm": 4.4229021072387695, + "grad_norm": 50.363983154296875, "learning_rate": 1.3503009821522865e-05, - "loss": 0.7417, + "loss": 1.5432, "step": 30760 }, { "epoch": 3.25, - "grad_norm": 26.45069122314453, + "grad_norm": 45.30118942260742, "learning_rate": 1.3500897666068224e-05, - "loss": 0.6359, + "loss": 1.1999, "step": 30770 }, { "epoch": 3.25, - "grad_norm": 4.329192638397217, + "grad_norm": 40.416133880615234, "learning_rate": 1.3498785510613581e-05, - "loss": 0.5605, + "loss": 1.0785, "step": 30780 }, { "epoch": 3.25, - "grad_norm": 13.992464065551758, + "grad_norm": 51.689395904541016, "learning_rate": 1.349667335515894e-05, - "loss": 0.581, + "loss": 1.2077, "step": 30790 }, { "epoch": 3.25, - "grad_norm": 7.9046854972839355, + "grad_norm": 25.54421043395996, "learning_rate": 1.3494561199704298e-05, - "loss": 0.5648, + "loss": 1.0273, "step": 30800 }, { "epoch": 3.25, - "grad_norm": 0.2955275774002075, + "grad_norm": 22.64275360107422, "learning_rate": 1.3492449044249657e-05, - "loss": 0.6384, + "loss": 1.3432, "step": 30810 }, { "epoch": 3.25, - "grad_norm": 8.375872611999512, + "grad_norm": 32.517784118652344, "learning_rate": 1.3490336888795018e-05, - "loss": 0.5263, + "loss": 1.0632, "step": 30820 }, { "epoch": 3.26, - "grad_norm": 17.133487701416016, + "grad_norm": 46.465476989746094, "learning_rate": 1.3488224733340377e-05, - "loss": 0.736, + "loss": 1.3629, "step": 30830 }, { "epoch": 3.26, - "grad_norm": 8.654123306274414, + "grad_norm": 30.149545669555664, "learning_rate": 1.3486112577885732e-05, - "loss": 0.7989, + "loss": 1.2121, "step": 30840 }, { "epoch": 3.26, - "grad_norm": 6.3392791748046875, + "grad_norm": 36.05668258666992, "learning_rate": 1.3484000422431092e-05, - "loss": 0.6948, + "loss": 1.3359, "step": 30850 }, { "epoch": 3.26, - "grad_norm": 12.801078796386719, + "grad_norm": 31.471303939819336, "learning_rate": 1.3481888266976451e-05, - "loss": 0.583, + "loss": 1.204, "step": 30860 }, { "epoch": 3.26, - "grad_norm": 3.405095338821411, + "grad_norm": 42.02981948852539, "learning_rate": 1.347977611152181e-05, - "loss": 0.4476, + "loss": 1.008, "step": 30870 }, { "epoch": 3.26, - "grad_norm": 14.19818115234375, + "grad_norm": 33.50074768066406, "learning_rate": 1.3477663956067169e-05, - "loss": 0.5191, + "loss": 1.3986, "step": 30880 }, { "epoch": 3.26, - "grad_norm": 13.236477851867676, + "grad_norm": 37.86475372314453, "learning_rate": 1.3475551800612526e-05, - "loss": 0.5532, + "loss": 1.0896, "step": 30890 }, { "epoch": 3.26, - "grad_norm": 7.691550254821777, + "grad_norm": 33.12504196166992, "learning_rate": 1.3473439645157885e-05, - "loss": 0.5656, + "loss": 1.3003, "step": 30900 }, { "epoch": 3.26, - "grad_norm": 15.602487564086914, + "grad_norm": 63.25077438354492, "learning_rate": 1.3471327489703243e-05, - "loss": 0.7866, + "loss": 1.5481, "step": 30910 }, { "epoch": 3.27, - "grad_norm": 1.4603641033172607, + "grad_norm": 31.408227920532227, "learning_rate": 1.3469215334248602e-05, - "loss": 0.6241, + "loss": 1.3892, "step": 30920 }, { "epoch": 3.27, - "grad_norm": 20.79844856262207, + "grad_norm": 45.454978942871094, "learning_rate": 1.3467103178793961e-05, - "loss": 0.4706, + "loss": 1.117, "step": 30930 }, { "epoch": 3.27, - "grad_norm": 16.314205169677734, + "grad_norm": 48.14042282104492, "learning_rate": 1.346499102333932e-05, - "loss": 0.2959, + "loss": 1.2229, "step": 30940 }, { "epoch": 3.27, - "grad_norm": 7.304533958435059, + "grad_norm": 50.28839111328125, "learning_rate": 1.3462878867884677e-05, - "loss": 0.4628, + "loss": 1.2904, "step": 30950 }, { "epoch": 3.27, - "grad_norm": 7.845974445343018, + "grad_norm": 31.793079376220703, "learning_rate": 1.3460766712430036e-05, - "loss": 0.6274, + "loss": 1.2682, "step": 30960 }, { "epoch": 3.27, - "grad_norm": 18.520402908325195, + "grad_norm": 40.721351623535156, "learning_rate": 1.3458654556975394e-05, - "loss": 0.5031, + "loss": 1.5603, "step": 30970 }, { "epoch": 3.27, - "grad_norm": 16.558517456054688, + "grad_norm": 43.965858459472656, "learning_rate": 1.3456542401520753e-05, - "loss": 0.5539, + "loss": 1.1173, "step": 30980 }, { "epoch": 3.27, - "grad_norm": 25.42609977722168, + "grad_norm": 46.7830810546875, "learning_rate": 1.3454430246066112e-05, - "loss": 0.407, + "loss": 1.1274, "step": 30990 }, { "epoch": 3.27, - "grad_norm": 3.705322742462158, + "grad_norm": 31.52637481689453, "learning_rate": 1.3452318090611469e-05, - "loss": 0.6672, + "loss": 1.4081, "step": 31000 }, { "epoch": 3.27, - "grad_norm": 12.079619407653809, + "grad_norm": 43.909202575683594, "learning_rate": 1.3450205935156828e-05, - "loss": 0.8357, + "loss": 1.4877, "step": 31010 }, { "epoch": 3.28, - "grad_norm": 16.338973999023438, + "grad_norm": 34.44239044189453, "learning_rate": 1.3448093779702187e-05, - "loss": 0.5524, + "loss": 1.0838, "step": 31020 }, { "epoch": 3.28, - "grad_norm": 14.758296012878418, + "grad_norm": 31.845399856567383, "learning_rate": 1.3445981624247545e-05, - "loss": 0.4432, + "loss": 1.2696, "step": 31030 }, { "epoch": 3.28, - "grad_norm": 20.573863983154297, + "grad_norm": 35.425994873046875, "learning_rate": 1.3443869468792906e-05, - "loss": 0.3753, + "loss": 1.1574, "step": 31040 }, { "epoch": 3.28, - "grad_norm": 1.7548866271972656, + "grad_norm": 40.67919158935547, "learning_rate": 1.3441757313338261e-05, - "loss": 0.4875, + "loss": 0.9409, "step": 31050 }, { "epoch": 3.28, - "grad_norm": 3.2790403366088867, + "grad_norm": 14.309308052062988, "learning_rate": 1.343964515788362e-05, - "loss": 0.5083, + "loss": 1.2084, "step": 31060 }, { "epoch": 3.28, - "grad_norm": 10.818489074707031, + "grad_norm": 35.60758972167969, "learning_rate": 1.343753300242898e-05, - "loss": 0.2659, + "loss": 1.2051, "step": 31070 }, { "epoch": 3.28, - "grad_norm": 0.41419216990470886, + "grad_norm": 40.51900100708008, "learning_rate": 1.343542084697434e-05, - "loss": 0.2993, + "loss": 0.8633, "step": 31080 }, { "epoch": 3.28, - "grad_norm": 0.3087881803512573, + "grad_norm": 21.886533737182617, "learning_rate": 1.3433308691519698e-05, - "loss": 0.4299, + "loss": 1.4926, "step": 31090 }, { "epoch": 3.28, - "grad_norm": 0.672577977180481, + "grad_norm": 27.503599166870117, "learning_rate": 1.3431196536065057e-05, - "loss": 0.1626, + "loss": 0.8267, "step": 31100 }, { "epoch": 3.29, - "grad_norm": 26.62739372253418, + "grad_norm": 51.07789993286133, "learning_rate": 1.3429084380610414e-05, - "loss": 0.6847, + "loss": 1.3919, "step": 31110 }, { "epoch": 3.29, - "grad_norm": 19.95575523376465, + "grad_norm": 37.30805206298828, "learning_rate": 1.3426972225155773e-05, - "loss": 0.324, + "loss": 1.1862, "step": 31120 }, { "epoch": 3.29, - "grad_norm": 35.13624572753906, + "grad_norm": 38.6754035949707, "learning_rate": 1.3424860069701131e-05, - "loss": 0.5504, + "loss": 1.1345, "step": 31130 }, { "epoch": 3.29, - "grad_norm": 6.5051350593566895, + "grad_norm": 45.42714309692383, "learning_rate": 1.342274791424649e-05, - "loss": 0.332, + "loss": 0.9818, "step": 31140 }, { "epoch": 3.29, - "grad_norm": 2.644839286804199, + "grad_norm": 25.41725730895996, "learning_rate": 1.3420635758791849e-05, - "loss": 0.4928, + "loss": 1.0929, "step": 31150 }, { "epoch": 3.29, - "grad_norm": 14.810765266418457, + "grad_norm": 29.149919509887695, "learning_rate": 1.3418523603337206e-05, - "loss": 0.7385, + "loss": 1.0786, "step": 31160 }, { "epoch": 3.29, - "grad_norm": 5.170985698699951, + "grad_norm": 17.930191040039062, "learning_rate": 1.3416411447882565e-05, - "loss": 0.5712, + "loss": 1.3813, "step": 31170 }, { "epoch": 3.29, - "grad_norm": 13.197240829467773, + "grad_norm": 55.54917907714844, "learning_rate": 1.3414299292427924e-05, - "loss": 0.8053, + "loss": 1.3728, "step": 31180 }, { "epoch": 3.29, - "grad_norm": 20.007736206054688, + "grad_norm": 40.93471145629883, "learning_rate": 1.3412187136973282e-05, - "loss": 0.3888, + "loss": 1.0498, "step": 31190 }, { "epoch": 3.29, - "grad_norm": 4.697957515716553, + "grad_norm": 37.71003341674805, "learning_rate": 1.3410074981518641e-05, - "loss": 0.4478, + "loss": 1.606, "step": 31200 }, { "epoch": 3.3, - "grad_norm": 18.522886276245117, + "grad_norm": 53.31984329223633, "learning_rate": 1.3407962826064e-05, - "loss": 0.5588, + "loss": 1.232, "step": 31210 }, { "epoch": 3.3, - "grad_norm": 19.234298706054688, + "grad_norm": 22.891294479370117, "learning_rate": 1.3405850670609357e-05, - "loss": 0.4955, + "loss": 1.1089, "step": 31220 }, { "epoch": 3.3, - "grad_norm": 19.827795028686523, + "grad_norm": 36.47248458862305, "learning_rate": 1.3403738515154716e-05, - "loss": 0.7192, + "loss": 1.3372, "step": 31230 }, { "epoch": 3.3, - "grad_norm": 20.344820022583008, + "grad_norm": 43.401145935058594, "learning_rate": 1.3401626359700075e-05, - "loss": 0.3887, + "loss": 1.173, "step": 31240 }, { "epoch": 3.3, - "grad_norm": 31.619388580322266, + "grad_norm": 54.55660629272461, "learning_rate": 1.3399514204245433e-05, - "loss": 0.8182, + "loss": 1.5157, "step": 31250 }, { "epoch": 3.3, - "grad_norm": 3.2931160926818848, + "grad_norm": 40.198211669921875, "learning_rate": 1.3397402048790792e-05, - "loss": 0.576, + "loss": 1.2763, "step": 31260 }, { "epoch": 3.3, - "grad_norm": 12.318822860717773, + "grad_norm": 48.99371337890625, "learning_rate": 1.339528989333615e-05, - "loss": 0.4473, + "loss": 1.1081, "step": 31270 }, { "epoch": 3.3, - "grad_norm": 9.644145011901855, + "grad_norm": 27.0717830657959, "learning_rate": 1.3393177737881508e-05, - "loss": 0.4733, + "loss": 1.2568, "step": 31280 }, { "epoch": 3.3, - "grad_norm": 0.4942200779914856, + "grad_norm": 42.86932373046875, "learning_rate": 1.3391065582426867e-05, - "loss": 0.5639, + "loss": 1.2796, "step": 31290 }, { "epoch": 3.31, - "grad_norm": 3.6659159660339355, + "grad_norm": 33.797298431396484, "learning_rate": 1.3388953426972227e-05, - "loss": 0.6131, + "loss": 1.2457, "step": 31300 }, { "epoch": 3.31, - "grad_norm": 10.228530883789062, + "grad_norm": 30.620529174804688, "learning_rate": 1.3386841271517586e-05, - "loss": 0.8477, + "loss": 1.3118, "step": 31310 }, { "epoch": 3.31, - "grad_norm": 22.745840072631836, + "grad_norm": 52.85122299194336, "learning_rate": 1.3384729116062941e-05, - "loss": 0.4157, + "loss": 0.8642, "step": 31320 }, { "epoch": 3.31, - "grad_norm": 4.711630821228027, + "grad_norm": 44.83036422729492, "learning_rate": 1.3382616960608302e-05, - "loss": 0.3891, + "loss": 1.2572, "step": 31330 }, { "epoch": 3.31, - "grad_norm": 3.2030622959136963, + "grad_norm": 34.47856521606445, "learning_rate": 1.338050480515366e-05, - "loss": 0.4491, + "loss": 1.2762, "step": 31340 }, { "epoch": 3.31, - "grad_norm": 8.93775463104248, + "grad_norm": 44.1667366027832, "learning_rate": 1.337839264969902e-05, - "loss": 0.5706, + "loss": 1.3075, "step": 31350 }, { "epoch": 3.31, - "grad_norm": 13.533286094665527, + "grad_norm": 33.70988082885742, "learning_rate": 1.3376280494244378e-05, - "loss": 0.9031, + "loss": 1.6719, "step": 31360 }, { "epoch": 3.31, - "grad_norm": 11.19803237915039, + "grad_norm": 28.818199157714844, "learning_rate": 1.3374168338789737e-05, - "loss": 0.6423, + "loss": 1.1125, "step": 31370 }, { "epoch": 3.31, - "grad_norm": 12.689641952514648, + "grad_norm": 53.72470474243164, "learning_rate": 1.3372056183335094e-05, - "loss": 0.5252, + "loss": 1.3013, "step": 31380 }, { "epoch": 3.32, - "grad_norm": 0.3727712035179138, + "grad_norm": 26.003875732421875, "learning_rate": 1.3369944027880453e-05, - "loss": 0.4178, + "loss": 1.0206, "step": 31390 }, { "epoch": 3.32, - "grad_norm": 26.40431022644043, + "grad_norm": 33.67490768432617, "learning_rate": 1.3367831872425812e-05, - "loss": 0.4661, + "loss": 1.2404, "step": 31400 }, { "epoch": 3.32, - "grad_norm": 8.99509334564209, + "grad_norm": 44.18916320800781, "learning_rate": 1.336571971697117e-05, - "loss": 0.3509, + "loss": 1.1029, "step": 31410 }, { "epoch": 3.32, - "grad_norm": 12.176636695861816, + "grad_norm": 41.82207107543945, "learning_rate": 1.336360756151653e-05, - "loss": 0.5313, + "loss": 1.042, "step": 31420 }, { "epoch": 3.32, - "grad_norm": 14.707193374633789, + "grad_norm": 38.04001235961914, "learning_rate": 1.3361495406061886e-05, - "loss": 0.7397, + "loss": 0.9054, "step": 31430 }, { "epoch": 3.32, - "grad_norm": 19.7658634185791, + "grad_norm": 37.30780029296875, "learning_rate": 1.3359383250607245e-05, - "loss": 0.3384, + "loss": 0.9868, "step": 31440 }, { "epoch": 3.32, - "grad_norm": 10.986204147338867, + "grad_norm": 25.669687271118164, "learning_rate": 1.3357271095152604e-05, - "loss": 0.4502, + "loss": 1.1584, "step": 31450 }, { "epoch": 3.32, - "grad_norm": 8.512551307678223, + "grad_norm": 56.69029998779297, "learning_rate": 1.3355158939697963e-05, - "loss": 0.6185, + "loss": 1.0767, "step": 31460 }, { "epoch": 3.32, - "grad_norm": 26.04615020751953, + "grad_norm": 58.17729949951172, "learning_rate": 1.3353046784243321e-05, - "loss": 0.4358, + "loss": 1.063, "step": 31470 }, { "epoch": 3.32, - "grad_norm": 2.3070828914642334, + "grad_norm": 44.35186767578125, "learning_rate": 1.335093462878868e-05, - "loss": 0.4327, + "loss": 1.2058, "step": 31480 }, { "epoch": 3.33, - "grad_norm": 26.63279914855957, + "grad_norm": 46.42628479003906, "learning_rate": 1.3348822473334037e-05, - "loss": 0.5791, + "loss": 1.0038, "step": 31490 }, { "epoch": 3.33, - "grad_norm": 56.89142608642578, + "grad_norm": 43.36750793457031, "learning_rate": 1.3346710317879396e-05, - "loss": 0.2771, + "loss": 0.7397, "step": 31500 }, { "epoch": 3.33, - "grad_norm": 10.318696022033691, + "grad_norm": 45.14958190917969, "learning_rate": 1.3344598162424755e-05, - "loss": 0.45, + "loss": 1.1672, "step": 31510 }, { "epoch": 3.33, - "grad_norm": 0.6249569058418274, + "grad_norm": 11.46998405456543, "learning_rate": 1.3342486006970115e-05, - "loss": 0.6796, + "loss": 0.8994, "step": 31520 }, { "epoch": 3.33, - "grad_norm": 19.765151977539062, + "grad_norm": 38.470699310302734, "learning_rate": 1.3340373851515474e-05, - "loss": 0.6113, + "loss": 1.0943, "step": 31530 }, { "epoch": 3.33, - "grad_norm": 9.831058502197266, + "grad_norm": 47.60748291015625, "learning_rate": 1.333826169606083e-05, - "loss": 0.505, + "loss": 1.2451, "step": 31540 }, { "epoch": 3.33, - "grad_norm": 20.020511627197266, + "grad_norm": 40.4107780456543, "learning_rate": 1.333614954060619e-05, - "loss": 0.5518, + "loss": 1.6279, "step": 31550 }, { "epoch": 3.33, - "grad_norm": 10.853751182556152, + "grad_norm": 45.386131286621094, "learning_rate": 1.3334037385151549e-05, - "loss": 0.4642, + "loss": 1.228, "step": 31560 }, { "epoch": 3.33, - "grad_norm": 9.689486503601074, + "grad_norm": 27.844608306884766, "learning_rate": 1.3331925229696908e-05, - "loss": 0.6926, + "loss": 1.4659, "step": 31570 }, { "epoch": 3.34, - "grad_norm": 24.49869155883789, + "grad_norm": 32.444583892822266, "learning_rate": 1.3329813074242266e-05, - "loss": 0.9167, + "loss": 1.2274, "step": 31580 }, { "epoch": 3.34, - "grad_norm": 12.098922729492188, + "grad_norm": 34.893394470214844, "learning_rate": 1.3327700918787625e-05, - "loss": 0.4631, + "loss": 1.15, "step": 31590 }, { "epoch": 3.34, - "grad_norm": 21.13144874572754, + "grad_norm": 40.85598373413086, "learning_rate": 1.3325588763332982e-05, - "loss": 0.7135, + "loss": 1.4576, "step": 31600 }, { "epoch": 3.34, - "grad_norm": 15.467947959899902, + "grad_norm": 46.183837890625, "learning_rate": 1.3323476607878341e-05, - "loss": 0.6407, + "loss": 1.4416, "step": 31610 }, { "epoch": 3.34, - "grad_norm": 6.66822624206543, + "grad_norm": 51.852333068847656, "learning_rate": 1.33213644524237e-05, - "loss": 0.3465, + "loss": 1.0425, "step": 31620 }, { "epoch": 3.34, - "grad_norm": 6.45578670501709, + "grad_norm": 19.30647087097168, "learning_rate": 1.3319252296969058e-05, - "loss": 0.3565, + "loss": 1.0106, "step": 31630 }, { "epoch": 3.34, - "grad_norm": 13.944631576538086, + "grad_norm": 49.76246643066406, "learning_rate": 1.3317140141514417e-05, - "loss": 0.6043, + "loss": 1.0498, "step": 31640 }, { "epoch": 3.34, - "grad_norm": 10.213872909545898, + "grad_norm": 21.31632423400879, "learning_rate": 1.3315027986059774e-05, - "loss": 0.5674, + "loss": 0.9778, "step": 31650 }, { "epoch": 3.34, - "grad_norm": 0.34242627024650574, + "grad_norm": 36.620914459228516, "learning_rate": 1.3312915830605133e-05, - "loss": 0.5973, + "loss": 1.6917, "step": 31660 }, { "epoch": 3.34, - "grad_norm": 20.641889572143555, + "grad_norm": 15.115422248840332, "learning_rate": 1.3310803675150492e-05, - "loss": 0.5961, + "loss": 1.2392, "step": 31670 }, { "epoch": 3.35, - "grad_norm": 18.33555030822754, + "grad_norm": 24.697919845581055, "learning_rate": 1.330869151969585e-05, - "loss": 0.4415, + "loss": 0.9443, "step": 31680 }, { "epoch": 3.35, - "grad_norm": 0.21979859471321106, + "grad_norm": 25.230571746826172, "learning_rate": 1.330657936424121e-05, - "loss": 0.4839, + "loss": 1.0222, "step": 31690 }, { "epoch": 3.35, - "grad_norm": 16.097673416137695, + "grad_norm": 47.61923599243164, "learning_rate": 1.3304467208786567e-05, - "loss": 0.4997, + "loss": 1.0548, "step": 31700 }, { "epoch": 3.35, - "grad_norm": 1.3537517786026, + "grad_norm": 59.65892791748047, "learning_rate": 1.3302355053331925e-05, - "loss": 0.5351, + "loss": 1.1861, "step": 31710 }, { "epoch": 3.35, - "grad_norm": 7.235955715179443, + "grad_norm": 31.498598098754883, "learning_rate": 1.3300242897877284e-05, - "loss": 0.4792, + "loss": 1.2304, "step": 31720 }, { "epoch": 3.35, - "grad_norm": 9.369999885559082, + "grad_norm": 49.925872802734375, "learning_rate": 1.3298130742422643e-05, - "loss": 0.5804, + "loss": 1.1987, "step": 31730 }, { "epoch": 3.35, - "grad_norm": 1.3253053426742554, + "grad_norm": 30.151962280273438, "learning_rate": 1.3296018586968003e-05, - "loss": 0.6174, + "loss": 1.294, "step": 31740 }, { "epoch": 3.35, - "grad_norm": 23.21744155883789, + "grad_norm": 51.62853240966797, "learning_rate": 1.3293906431513362e-05, - "loss": 0.5538, + "loss": 1.1914, "step": 31750 }, { "epoch": 3.35, - "grad_norm": 16.121824264526367, + "grad_norm": 41.72119903564453, "learning_rate": 1.3291794276058718e-05, - "loss": 0.8566, + "loss": 1.2462, "step": 31760 }, { "epoch": 3.36, - "grad_norm": 2.6199984550476074, + "grad_norm": 27.834951400756836, "learning_rate": 1.3289682120604076e-05, - "loss": 0.487, + "loss": 1.0344, "step": 31770 }, { "epoch": 3.36, - "grad_norm": 15.422292709350586, + "grad_norm": 46.825321197509766, "learning_rate": 1.3287569965149437e-05, - "loss": 0.3679, + "loss": 1.1534, "step": 31780 }, { "epoch": 3.36, - "grad_norm": 6.984682559967041, + "grad_norm": 37.968414306640625, "learning_rate": 1.3285457809694796e-05, - "loss": 0.555, + "loss": 1.1391, "step": 31790 }, { "epoch": 3.36, - "grad_norm": 15.099629402160645, + "grad_norm": 30.78635597229004, "learning_rate": 1.3283345654240154e-05, - "loss": 0.6275, + "loss": 1.1838, "step": 31800 }, { "epoch": 3.36, - "grad_norm": 19.40889549255371, + "grad_norm": 35.04497146606445, "learning_rate": 1.3281233498785511e-05, - "loss": 0.6448, + "loss": 1.3828, "step": 31810 }, { "epoch": 3.36, - "grad_norm": 47.2883415222168, + "grad_norm": 32.828861236572266, "learning_rate": 1.327912134333087e-05, - "loss": 0.5559, + "loss": 1.1714, "step": 31820 }, { "epoch": 3.36, - "grad_norm": 16.96462059020996, + "grad_norm": 47.073341369628906, "learning_rate": 1.3277009187876229e-05, - "loss": 0.4437, + "loss": 0.9899, "step": 31830 }, { "epoch": 3.36, - "grad_norm": 17.734375, + "grad_norm": 29.39654541015625, "learning_rate": 1.3274897032421588e-05, - "loss": 0.5158, + "loss": 1.2816, "step": 31840 }, { "epoch": 3.36, - "grad_norm": 5.66486930847168, + "grad_norm": 38.3040771484375, "learning_rate": 1.3272784876966947e-05, - "loss": 0.8718, + "loss": 1.0995, "step": 31850 }, { "epoch": 3.36, - "grad_norm": 1.4257080554962158, + "grad_norm": 28.33880615234375, "learning_rate": 1.3270672721512305e-05, - "loss": 0.5893, + "loss": 1.4138, "step": 31860 }, { "epoch": 3.37, - "grad_norm": 11.549933433532715, + "grad_norm": 53.62016677856445, "learning_rate": 1.3268560566057662e-05, - "loss": 0.6509, + "loss": 1.4263, "step": 31870 }, { "epoch": 3.37, - "grad_norm": 17.03160858154297, + "grad_norm": 43.83967971801758, "learning_rate": 1.3266448410603021e-05, - "loss": 0.711, + "loss": 1.3102, "step": 31880 }, { "epoch": 3.37, - "grad_norm": 20.90876579284668, + "grad_norm": 60.49833679199219, "learning_rate": 1.326433625514838e-05, - "loss": 0.7058, + "loss": 1.522, "step": 31890 }, { "epoch": 3.37, - "grad_norm": 12.053631782531738, + "grad_norm": 50.432003021240234, "learning_rate": 1.3262224099693739e-05, - "loss": 0.781, + "loss": 1.2531, "step": 31900 }, { "epoch": 3.37, - "grad_norm": 22.97846221923828, + "grad_norm": 39.25982666015625, "learning_rate": 1.3260111944239098e-05, - "loss": 0.5565, + "loss": 1.4782, "step": 31910 }, { "epoch": 3.37, - "grad_norm": 6.52946138381958, + "grad_norm": 48.63429260253906, "learning_rate": 1.3257999788784455e-05, - "loss": 0.3999, + "loss": 1.2709, "step": 31920 }, { "epoch": 3.37, - "grad_norm": 4.714751720428467, + "grad_norm": 39.45363235473633, "learning_rate": 1.3255887633329813e-05, - "loss": 0.5097, + "loss": 1.2894, "step": 31930 }, { "epoch": 3.37, - "grad_norm": 10.861776351928711, + "grad_norm": 43.31864547729492, "learning_rate": 1.3253775477875172e-05, - "loss": 0.4474, + "loss": 1.0405, "step": 31940 }, { "epoch": 3.37, - "grad_norm": 31.34658432006836, + "grad_norm": 36.84076690673828, "learning_rate": 1.3251663322420531e-05, - "loss": 0.7131, + "loss": 1.1134, "step": 31950 }, { "epoch": 3.38, - "grad_norm": 23.631710052490234, + "grad_norm": 27.924699783325195, "learning_rate": 1.324955116696589e-05, - "loss": 0.6603, + "loss": 1.5285, "step": 31960 }, { "epoch": 3.38, - "grad_norm": 10.673746109008789, + "grad_norm": 37.7152214050293, "learning_rate": 1.3247439011511247e-05, - "loss": 0.4829, + "loss": 0.9369, "step": 31970 }, { "epoch": 3.38, - "grad_norm": 18.648550033569336, + "grad_norm": 50.90470886230469, "learning_rate": 1.3245326856056606e-05, - "loss": 0.4527, + "loss": 0.9945, "step": 31980 }, { "epoch": 3.38, - "grad_norm": 32.292545318603516, + "grad_norm": 45.034786224365234, "learning_rate": 1.3243214700601964e-05, - "loss": 0.6062, + "loss": 1.1593, "step": 31990 }, { "epoch": 3.38, - "grad_norm": 6.059606075286865, + "grad_norm": 45.78617858886719, "learning_rate": 1.3241102545147325e-05, - "loss": 0.3613, + "loss": 1.1268, "step": 32000 }, { "epoch": 3.38, - "grad_norm": 17.523557662963867, + "grad_norm": 43.03278350830078, "learning_rate": 1.3238990389692684e-05, - "loss": 0.7959, + "loss": 1.3939, "step": 32010 }, { "epoch": 3.38, - "grad_norm": 15.43974781036377, + "grad_norm": 49.222076416015625, "learning_rate": 1.3236878234238042e-05, - "loss": 0.5273, + "loss": 1.259, "step": 32020 }, { "epoch": 3.38, - "grad_norm": 10.460396766662598, + "grad_norm": 16.97509765625, "learning_rate": 1.32347660787834e-05, - "loss": 0.5126, + "loss": 1.1286, "step": 32030 }, { "epoch": 3.38, - "grad_norm": 8.636281967163086, + "grad_norm": 29.003271102905273, "learning_rate": 1.3232653923328758e-05, - "loss": 0.6825, + "loss": 1.1547, "step": 32040 }, { "epoch": 3.38, - "grad_norm": 13.669848442077637, + "grad_norm": 37.653411865234375, "learning_rate": 1.3230541767874117e-05, - "loss": 0.5246, + "loss": 1.1669, "step": 32050 }, { "epoch": 3.39, - "grad_norm": 6.038885116577148, + "grad_norm": 32.48966979980469, "learning_rate": 1.3228429612419476e-05, - "loss": 0.4007, + "loss": 1.3137, "step": 32060 }, { "epoch": 3.39, - "grad_norm": 19.935230255126953, + "grad_norm": 34.627540588378906, "learning_rate": 1.3226317456964835e-05, - "loss": 0.4531, + "loss": 0.8503, "step": 32070 }, { "epoch": 3.39, - "grad_norm": 11.108431816101074, + "grad_norm": 27.657835006713867, "learning_rate": 1.3224205301510192e-05, - "loss": 0.6494, + "loss": 1.2647, "step": 32080 }, { "epoch": 3.39, - "grad_norm": 7.625349998474121, + "grad_norm": 32.276390075683594, "learning_rate": 1.322209314605555e-05, - "loss": 0.4652, + "loss": 1.1098, "step": 32090 }, { "epoch": 3.39, - "grad_norm": 38.37046432495117, + "grad_norm": 51.90109634399414, "learning_rate": 1.321998099060091e-05, - "loss": 0.542, + "loss": 1.088, "step": 32100 }, { "epoch": 3.39, - "grad_norm": 7.050558090209961, + "grad_norm": 38.80809020996094, "learning_rate": 1.3217868835146268e-05, - "loss": 0.4605, + "loss": 1.3107, "step": 32110 }, { "epoch": 3.39, - "grad_norm": 2.216461181640625, + "grad_norm": 33.68888854980469, "learning_rate": 1.3215756679691627e-05, - "loss": 0.6339, + "loss": 1.3783, "step": 32120 }, { "epoch": 3.39, - "grad_norm": 6.625785827636719, + "grad_norm": 25.766267776489258, "learning_rate": 1.3213644524236986e-05, - "loss": 0.6927, + "loss": 1.1548, "step": 32130 }, { "epoch": 3.39, - "grad_norm": 0.7392148375511169, + "grad_norm": 36.6627311706543, "learning_rate": 1.3211532368782343e-05, - "loss": 0.5648, + "loss": 1.2663, "step": 32140 }, { "epoch": 3.4, - "grad_norm": 25.596637725830078, + "grad_norm": 44.47728729248047, "learning_rate": 1.3209420213327701e-05, - "loss": 0.7509, + "loss": 1.4066, "step": 32150 }, { "epoch": 3.4, - "grad_norm": 22.488216400146484, + "grad_norm": 36.60101318359375, "learning_rate": 1.320730805787306e-05, - "loss": 0.6363, + "loss": 1.1271, "step": 32160 }, { "epoch": 3.4, - "grad_norm": 6.962686061859131, + "grad_norm": 45.8122444152832, "learning_rate": 1.3205195902418419e-05, - "loss": 0.5878, + "loss": 1.4163, "step": 32170 }, { "epoch": 3.4, - "grad_norm": 11.015753746032715, + "grad_norm": 38.803436279296875, "learning_rate": 1.3203083746963778e-05, - "loss": 0.7897, + "loss": 1.1114, "step": 32180 }, { "epoch": 3.4, - "grad_norm": 0.7138781547546387, + "grad_norm": 37.807857513427734, "learning_rate": 1.3200971591509135e-05, - "loss": 0.5911, + "loss": 1.2583, "step": 32190 }, { "epoch": 3.4, - "grad_norm": 17.953611373901367, + "grad_norm": 42.784828186035156, "learning_rate": 1.3198859436054494e-05, - "loss": 0.7266, + "loss": 1.5959, "step": 32200 }, { "epoch": 3.4, - "grad_norm": 6.144376754760742, + "grad_norm": 13.086291313171387, "learning_rate": 1.3196747280599852e-05, - "loss": 0.5274, + "loss": 1.686, "step": 32210 }, { "epoch": 3.4, - "grad_norm": 1.8739901781082153, + "grad_norm": 39.406429290771484, "learning_rate": 1.3194635125145213e-05, - "loss": 0.4034, + "loss": 1.0537, "step": 32220 }, { "epoch": 3.4, - "grad_norm": 14.456005096435547, + "grad_norm": 41.73741912841797, "learning_rate": 1.3192522969690572e-05, - "loss": 0.7258, + "loss": 1.3923, "step": 32230 }, { "epoch": 3.4, - "grad_norm": 12.376498222351074, + "grad_norm": 38.60031509399414, "learning_rate": 1.319041081423593e-05, - "loss": 0.4257, + "loss": 1.1195, "step": 32240 }, { "epoch": 3.41, - "grad_norm": 11.759756088256836, + "grad_norm": 32.954345703125, "learning_rate": 1.3188298658781286e-05, - "loss": 0.5352, + "loss": 1.198, "step": 32250 }, { "epoch": 3.41, - "grad_norm": 14.5733003616333, + "grad_norm": 36.84732437133789, "learning_rate": 1.3186186503326646e-05, - "loss": 0.4326, + "loss": 0.9055, "step": 32260 }, { "epoch": 3.41, - "grad_norm": 6.774064540863037, + "grad_norm": 33.75140380859375, "learning_rate": 1.3184074347872005e-05, - "loss": 0.463, + "loss": 1.3364, "step": 32270 }, { "epoch": 3.41, - "grad_norm": 20.43936538696289, + "grad_norm": 38.6740608215332, "learning_rate": 1.3181962192417364e-05, - "loss": 0.7282, + "loss": 1.2196, "step": 32280 }, { "epoch": 3.41, - "grad_norm": 18.89690589904785, + "grad_norm": 33.508846282958984, "learning_rate": 1.3179850036962723e-05, - "loss": 0.4436, + "loss": 1.0141, "step": 32290 }, { "epoch": 3.41, - "grad_norm": 7.057826995849609, + "grad_norm": 18.305017471313477, "learning_rate": 1.317773788150808e-05, - "loss": 0.4742, + "loss": 1.0868, "step": 32300 }, { "epoch": 3.41, - "grad_norm": 16.324642181396484, + "grad_norm": 17.589597702026367, "learning_rate": 1.3175625726053439e-05, - "loss": 0.83, + "loss": 1.1574, "step": 32310 }, { "epoch": 3.41, - "grad_norm": 0.09173928946256638, + "grad_norm": 33.893043518066406, "learning_rate": 1.3173513570598797e-05, - "loss": 0.2381, + "loss": 1.0198, "step": 32320 }, { "epoch": 3.41, - "grad_norm": 1.4288077354431152, + "grad_norm": 12.722223281860352, "learning_rate": 1.3171401415144156e-05, - "loss": 0.3052, + "loss": 1.1204, "step": 32330 }, { "epoch": 3.42, - "grad_norm": 12.10988998413086, + "grad_norm": 39.265968322753906, "learning_rate": 1.3169289259689515e-05, - "loss": 0.4486, + "loss": 1.2767, "step": 32340 }, { "epoch": 3.42, - "grad_norm": 5.672297477722168, + "grad_norm": 39.928314208984375, "learning_rate": 1.3167177104234872e-05, - "loss": 0.5444, + "loss": 1.1676, "step": 32350 }, { "epoch": 3.42, - "grad_norm": 17.894989013671875, + "grad_norm": 49.49498748779297, "learning_rate": 1.316506494878023e-05, - "loss": 0.6014, + "loss": 1.2044, "step": 32360 }, { "epoch": 3.42, - "grad_norm": 13.410127639770508, + "grad_norm": 33.14033889770508, "learning_rate": 1.316295279332559e-05, - "loss": 0.5814, + "loss": 1.1418, "step": 32370 }, { "epoch": 3.42, - "grad_norm": 18.28502655029297, + "grad_norm": 33.73317337036133, "learning_rate": 1.3160840637870948e-05, - "loss": 0.6907, + "loss": 1.2388, "step": 32380 }, { "epoch": 3.42, - "grad_norm": 8.837469100952148, + "grad_norm": 26.494060516357422, "learning_rate": 1.3158728482416307e-05, - "loss": 1.0839, + "loss": 1.4988, "step": 32390 }, { "epoch": 3.42, - "grad_norm": 47.26670837402344, + "grad_norm": 32.30350112915039, "learning_rate": 1.3156616326961666e-05, - "loss": 0.626, + "loss": 1.3685, "step": 32400 }, { "epoch": 3.42, - "grad_norm": 37.611934661865234, + "grad_norm": 55.04549789428711, "learning_rate": 1.3154504171507023e-05, - "loss": 0.8628, + "loss": 1.6153, "step": 32410 }, { "epoch": 3.42, - "grad_norm": 23.09202766418457, + "grad_norm": 38.8289909362793, "learning_rate": 1.3152392016052382e-05, - "loss": 0.8661, + "loss": 1.5787, "step": 32420 }, { "epoch": 3.42, - "grad_norm": 9.052006721496582, + "grad_norm": 50.451663970947266, "learning_rate": 1.315027986059774e-05, - "loss": 0.4519, + "loss": 1.4852, "step": 32430 }, { "epoch": 3.43, - "grad_norm": 16.9249210357666, + "grad_norm": 52.046234130859375, "learning_rate": 1.31481677051431e-05, - "loss": 0.4227, + "loss": 1.1204, "step": 32440 }, { "epoch": 3.43, - "grad_norm": 27.8602294921875, + "grad_norm": 30.419069290161133, "learning_rate": 1.314605554968846e-05, - "loss": 0.8943, + "loss": 1.6039, "step": 32450 }, { "epoch": 3.43, - "grad_norm": 4.495337009429932, + "grad_norm": 28.66710662841797, "learning_rate": 1.3143943394233815e-05, - "loss": 0.4644, + "loss": 1.0164, "step": 32460 }, { "epoch": 3.43, - "grad_norm": 16.543148040771484, + "grad_norm": 37.29116439819336, "learning_rate": 1.3141831238779174e-05, - "loss": 0.4427, + "loss": 0.8374, "step": 32470 }, { "epoch": 3.43, - "grad_norm": 12.77571964263916, + "grad_norm": 28.811241149902344, "learning_rate": 1.3139719083324534e-05, - "loss": 0.7988, + "loss": 1.1574, "step": 32480 }, { "epoch": 3.43, - "grad_norm": 19.114835739135742, + "grad_norm": 39.19492721557617, "learning_rate": 1.3137606927869893e-05, - "loss": 0.5701, + "loss": 1.3259, "step": 32490 }, { "epoch": 3.43, - "grad_norm": 18.446849822998047, + "grad_norm": 41.31183624267578, "learning_rate": 1.3135494772415252e-05, - "loss": 0.7839, + "loss": 1.1575, "step": 32500 }, { "epoch": 3.43, - "grad_norm": 15.780656814575195, + "grad_norm": 30.372011184692383, "learning_rate": 1.313338261696061e-05, - "loss": 0.3685, + "loss": 1.0863, "step": 32510 }, { "epoch": 3.43, - "grad_norm": 4.595785140991211, + "grad_norm": 36.626041412353516, "learning_rate": 1.3131270461505968e-05, - "loss": 0.6545, + "loss": 1.169, "step": 32520 }, { "epoch": 3.44, - "grad_norm": 10.855877876281738, + "grad_norm": 57.485477447509766, "learning_rate": 1.3129158306051327e-05, - "loss": 0.7783, + "loss": 1.3137, "step": 32530 }, { "epoch": 3.44, - "grad_norm": 24.20875358581543, + "grad_norm": 51.35068130493164, "learning_rate": 1.3127046150596685e-05, - "loss": 0.6312, + "loss": 1.5673, "step": 32540 }, { "epoch": 3.44, - "grad_norm": 8.048144340515137, + "grad_norm": 35.82453536987305, "learning_rate": 1.3124933995142044e-05, - "loss": 0.8338, + "loss": 1.2016, "step": 32550 }, { "epoch": 3.44, - "grad_norm": 18.61056137084961, + "grad_norm": 33.51424026489258, "learning_rate": 1.3122821839687403e-05, - "loss": 0.5598, + "loss": 1.1617, "step": 32560 }, { "epoch": 3.44, - "grad_norm": 3.1902170181274414, + "grad_norm": 31.681941986083984, "learning_rate": 1.312070968423276e-05, - "loss": 0.4887, + "loss": 1.0277, "step": 32570 }, { "epoch": 3.44, - "grad_norm": 17.413726806640625, + "grad_norm": 45.55038833618164, "learning_rate": 1.3118597528778119e-05, - "loss": 0.5727, + "loss": 1.2883, "step": 32580 }, { "epoch": 3.44, - "grad_norm": 22.15818977355957, + "grad_norm": 48.24006652832031, "learning_rate": 1.3116485373323478e-05, - "loss": 0.4124, + "loss": 1.0239, "step": 32590 }, { "epoch": 3.44, - "grad_norm": 21.139326095581055, + "grad_norm": 40.30105972290039, "learning_rate": 1.3114373217868836e-05, - "loss": 0.5052, + "loss": 1.1482, "step": 32600 }, { "epoch": 3.44, - "grad_norm": 9.48902702331543, + "grad_norm": 35.09836959838867, "learning_rate": 1.3112261062414195e-05, - "loss": 0.5801, + "loss": 1.0115, "step": 32610 }, { "epoch": 3.44, - "grad_norm": 2.715860366821289, + "grad_norm": 45.1078987121582, "learning_rate": 1.3110148906959552e-05, - "loss": 0.4992, + "loss": 1.2309, "step": 32620 }, { "epoch": 3.45, - "grad_norm": 8.88925552368164, + "grad_norm": 31.586938858032227, "learning_rate": 1.3108036751504911e-05, - "loss": 0.3821, + "loss": 1.2999, "step": 32630 }, { "epoch": 3.45, - "grad_norm": 11.51905345916748, + "grad_norm": 37.008331298828125, "learning_rate": 1.310592459605027e-05, - "loss": 0.7568, + "loss": 1.1152, "step": 32640 }, { "epoch": 3.45, - "grad_norm": 5.581113338470459, + "grad_norm": 28.72132110595703, "learning_rate": 1.3103812440595629e-05, - "loss": 0.5401, + "loss": 0.9402, "step": 32650 }, { "epoch": 3.45, - "grad_norm": 23.00388526916504, + "grad_norm": 30.422273635864258, "learning_rate": 1.3101700285140987e-05, - "loss": 0.4852, + "loss": 0.9159, "step": 32660 }, { "epoch": 3.45, - "grad_norm": 28.56183624267578, + "grad_norm": 41.5989875793457, "learning_rate": 1.3099588129686348e-05, - "loss": 0.625, + "loss": 0.9249, "step": 32670 }, { "epoch": 3.45, - "grad_norm": 6.639039993286133, + "grad_norm": 47.84788131713867, "learning_rate": 1.3097475974231703e-05, - "loss": 0.7313, + "loss": 1.2882, "step": 32680 }, { "epoch": 3.45, - "grad_norm": 0.5636126399040222, + "grad_norm": 41.98460006713867, "learning_rate": 1.3095363818777062e-05, - "loss": 0.6669, + "loss": 1.0264, "step": 32690 }, { "epoch": 3.45, - "grad_norm": 13.910141944885254, + "grad_norm": 33.72138977050781, "learning_rate": 1.3093251663322422e-05, - "loss": 0.7144, + "loss": 1.5381, "step": 32700 }, { "epoch": 3.45, - "grad_norm": 7.109676837921143, + "grad_norm": 39.09609603881836, "learning_rate": 1.3091139507867781e-05, - "loss": 0.5853, + "loss": 1.256, "step": 32710 }, { "epoch": 3.46, - "grad_norm": 8.594489097595215, + "grad_norm": 24.50278091430664, "learning_rate": 1.308902735241314e-05, - "loss": 0.5574, + "loss": 1.0133, "step": 32720 }, { "epoch": 3.46, - "grad_norm": 26.194753646850586, + "grad_norm": 44.60417938232422, "learning_rate": 1.3086915196958497e-05, - "loss": 1.0225, + "loss": 1.4445, "step": 32730 }, { "epoch": 3.46, - "grad_norm": 18.531225204467773, + "grad_norm": 38.88678741455078, "learning_rate": 1.3084803041503856e-05, - "loss": 0.6372, + "loss": 1.5954, "step": 32740 }, { "epoch": 3.46, - "grad_norm": 10.665400505065918, + "grad_norm": 23.549144744873047, "learning_rate": 1.3082690886049215e-05, - "loss": 0.6783, + "loss": 1.318, "step": 32750 }, { "epoch": 3.46, - "grad_norm": 2.5757744312286377, + "grad_norm": 42.2791633605957, "learning_rate": 1.3080578730594573e-05, - "loss": 0.7816, + "loss": 1.4846, "step": 32760 }, { "epoch": 3.46, - "grad_norm": 7.783481121063232, + "grad_norm": 42.37057113647461, "learning_rate": 1.3078466575139932e-05, - "loss": 0.4852, + "loss": 1.1223, "step": 32770 }, { "epoch": 3.46, - "grad_norm": 2.858670473098755, + "grad_norm": 58.46373748779297, "learning_rate": 1.3076354419685291e-05, - "loss": 0.7645, + "loss": 1.3693, "step": 32780 }, { "epoch": 3.46, - "grad_norm": 15.440728187561035, + "grad_norm": 36.794315338134766, "learning_rate": 1.3074242264230648e-05, - "loss": 0.598, + "loss": 1.1834, "step": 32790 }, { "epoch": 3.46, - "grad_norm": 27.92516326904297, + "grad_norm": 34.97380828857422, "learning_rate": 1.3072130108776007e-05, - "loss": 0.6235, + "loss": 1.2878, "step": 32800 }, { "epoch": 3.46, - "grad_norm": 120.66822052001953, + "grad_norm": 52.735660552978516, "learning_rate": 1.3070017953321366e-05, - "loss": 0.6264, + "loss": 1.2407, "step": 32810 }, { "epoch": 3.47, - "grad_norm": 25.434234619140625, + "grad_norm": 58.1715202331543, "learning_rate": 1.3067905797866724e-05, - "loss": 0.5332, + "loss": 1.4914, "step": 32820 }, { "epoch": 3.47, - "grad_norm": 17.462064743041992, + "grad_norm": 43.32632064819336, "learning_rate": 1.3065793642412083e-05, - "loss": 0.8466, + "loss": 1.5795, "step": 32830 }, { "epoch": 3.47, - "grad_norm": 12.587057113647461, + "grad_norm": 39.349395751953125, "learning_rate": 1.306368148695744e-05, - "loss": 0.3322, + "loss": 1.224, "step": 32840 }, { "epoch": 3.47, - "grad_norm": 2.5902206897735596, + "grad_norm": 17.400266647338867, "learning_rate": 1.3061569331502799e-05, - "loss": 0.5347, + "loss": 1.2304, "step": 32850 }, { "epoch": 3.47, - "grad_norm": 10.487377166748047, + "grad_norm": 34.697425842285156, "learning_rate": 1.3059457176048158e-05, - "loss": 0.497, + "loss": 1.258, "step": 32860 }, { "epoch": 3.47, - "grad_norm": 7.633495330810547, + "grad_norm": 33.736244201660156, "learning_rate": 1.3057345020593517e-05, - "loss": 0.5201, + "loss": 1.3332, "step": 32870 }, { "epoch": 3.47, - "grad_norm": 12.87855052947998, + "grad_norm": 42.559722900390625, "learning_rate": 1.3055232865138875e-05, - "loss": 0.632, + "loss": 1.0617, "step": 32880 }, { "epoch": 3.47, - "grad_norm": 16.818191528320312, + "grad_norm": 37.66116714477539, "learning_rate": 1.3053120709684232e-05, - "loss": 0.5415, + "loss": 1.213, "step": 32890 }, { "epoch": 3.47, - "grad_norm": 21.822513580322266, + "grad_norm": 14.269923210144043, "learning_rate": 1.3051008554229591e-05, - "loss": 0.6074, + "loss": 1.3789, "step": 32900 }, { "epoch": 3.48, - "grad_norm": 24.88652801513672, + "grad_norm": 30.332263946533203, "learning_rate": 1.304889639877495e-05, - "loss": 0.3258, + "loss": 0.9593, "step": 32910 }, { "epoch": 3.48, - "grad_norm": 16.039215087890625, + "grad_norm": 48.980072021484375, "learning_rate": 1.3046784243320309e-05, - "loss": 0.692, + "loss": 1.0378, "step": 32920 }, { "epoch": 3.48, - "grad_norm": 30.190263748168945, + "grad_norm": 27.66055679321289, "learning_rate": 1.304467208786567e-05, - "loss": 0.6635, + "loss": 1.3677, "step": 32930 }, { "epoch": 3.48, - "grad_norm": 0.26632893085479736, + "grad_norm": 19.72407341003418, "learning_rate": 1.3042559932411028e-05, - "loss": 0.3633, + "loss": 1.1315, "step": 32940 }, { "epoch": 3.48, - "grad_norm": 8.409527778625488, + "grad_norm": 44.89051818847656, "learning_rate": 1.3040447776956383e-05, - "loss": 0.5939, + "loss": 1.4069, "step": 32950 }, { "epoch": 3.48, - "grad_norm": 7.220369338989258, + "grad_norm": 38.837181091308594, "learning_rate": 1.3038335621501744e-05, - "loss": 0.4956, + "loss": 1.2149, "step": 32960 }, { "epoch": 3.48, - "grad_norm": 4.717998027801514, + "grad_norm": 26.734281539916992, "learning_rate": 1.3036223466047103e-05, - "loss": 0.5965, + "loss": 1.1424, "step": 32970 }, { "epoch": 3.48, - "grad_norm": 2.0796597003936768, + "grad_norm": 35.11247253417969, "learning_rate": 1.3034111310592461e-05, - "loss": 0.5305, + "loss": 1.1438, "step": 32980 }, { "epoch": 3.48, - "grad_norm": 12.707292556762695, + "grad_norm": 26.570981979370117, "learning_rate": 1.303199915513782e-05, - "loss": 0.3457, + "loss": 0.953, "step": 32990 }, { "epoch": 3.49, - "grad_norm": 4.85061502456665, + "grad_norm": 31.380464553833008, "learning_rate": 1.3029886999683177e-05, - "loss": 0.6866, + "loss": 1.2278, "step": 33000 }, { "epoch": 3.49, - "grad_norm": 20.974594116210938, + "grad_norm": 42.143924713134766, "learning_rate": 1.3027774844228536e-05, - "loss": 0.4968, + "loss": 1.2038, "step": 33010 }, { "epoch": 3.49, - "grad_norm": 20.152143478393555, + "grad_norm": 38.33856964111328, "learning_rate": 1.3025662688773895e-05, - "loss": 0.55, + "loss": 1.3095, "step": 33020 }, { "epoch": 3.49, - "grad_norm": 28.691181182861328, + "grad_norm": 48.25766372680664, "learning_rate": 1.3023550533319254e-05, - "loss": 0.4754, + "loss": 1.1218, "step": 33030 }, { "epoch": 3.49, - "grad_norm": 33.11190414428711, + "grad_norm": 41.74365234375, "learning_rate": 1.3021438377864612e-05, - "loss": 0.5262, + "loss": 1.2572, "step": 33040 }, { "epoch": 3.49, - "grad_norm": 12.528075218200684, + "grad_norm": 42.34245681762695, "learning_rate": 1.3019326222409971e-05, - "loss": 0.4356, + "loss": 1.4251, "step": 33050 }, { "epoch": 3.49, - "grad_norm": 12.219746589660645, + "grad_norm": 25.983177185058594, "learning_rate": 1.3017214066955328e-05, - "loss": 0.5822, + "loss": 1.0969, "step": 33060 }, { "epoch": 3.49, - "grad_norm": 9.901021957397461, + "grad_norm": 38.12257766723633, "learning_rate": 1.3015101911500687e-05, - "loss": 0.2606, + "loss": 0.8787, "step": 33070 }, { "epoch": 3.49, - "grad_norm": 13.002409934997559, + "grad_norm": 38.277748107910156, "learning_rate": 1.3012989756046046e-05, - "loss": 0.7147, + "loss": 1.2232, "step": 33080 }, { "epoch": 3.49, - "grad_norm": 30.335954666137695, + "grad_norm": 23.320171356201172, "learning_rate": 1.3010877600591405e-05, - "loss": 0.5899, + "loss": 1.3045, "step": 33090 }, { "epoch": 3.5, - "grad_norm": 2.9695932865142822, + "grad_norm": 29.095863342285156, "learning_rate": 1.3008765445136763e-05, - "loss": 0.5256, + "loss": 1.214, "step": 33100 }, { "epoch": 3.5, - "grad_norm": 8.571295738220215, + "grad_norm": 18.256839752197266, "learning_rate": 1.300665328968212e-05, - "loss": 0.5473, + "loss": 0.8157, "step": 33110 }, { "epoch": 3.5, - "grad_norm": 5.557065010070801, + "grad_norm": 34.27114486694336, "learning_rate": 1.300454113422748e-05, - "loss": 0.5939, + "loss": 1.1126, "step": 33120 }, { "epoch": 3.5, - "grad_norm": 10.57597827911377, + "grad_norm": 45.51290512084961, "learning_rate": 1.3002428978772838e-05, - "loss": 0.522, + "loss": 1.2274, "step": 33130 }, { "epoch": 3.5, - "grad_norm": 24.129722595214844, + "grad_norm": 35.300758361816406, "learning_rate": 1.3000316823318197e-05, - "loss": 0.5356, + "loss": 1.0494, "step": 33140 }, { "epoch": 3.5, - "grad_norm": 17.196943283081055, + "grad_norm": 46.97726821899414, "learning_rate": 1.2998204667863557e-05, - "loss": 0.5987, + "loss": 1.473, "step": 33150 }, { "epoch": 3.5, - "grad_norm": 18.806076049804688, + "grad_norm": 30.95558738708496, "learning_rate": 1.2996092512408916e-05, - "loss": 0.4469, + "loss": 1.1614, "step": 33160 }, { "epoch": 3.5, - "grad_norm": 21.07026481628418, + "grad_norm": 42.984474182128906, "learning_rate": 1.2993980356954271e-05, - "loss": 0.6469, + "loss": 1.3229, "step": 33170 }, { "epoch": 3.5, - "grad_norm": 17.478477478027344, + "grad_norm": 31.395071029663086, "learning_rate": 1.2991868201499632e-05, - "loss": 0.8295, + "loss": 1.1808, "step": 33180 }, { "epoch": 3.51, - "grad_norm": 9.466779708862305, + "grad_norm": 50.819759368896484, "learning_rate": 1.298975604604499e-05, - "loss": 0.5561, + "loss": 1.2611, "step": 33190 }, { "epoch": 3.51, - "grad_norm": 8.192133903503418, + "grad_norm": 33.20209503173828, "learning_rate": 1.298764389059035e-05, - "loss": 0.3447, + "loss": 0.9786, "step": 33200 }, { "epoch": 3.51, - "grad_norm": 3.3420746326446533, + "grad_norm": 33.22514724731445, "learning_rate": 1.2985531735135708e-05, - "loss": 0.5961, + "loss": 1.0816, "step": 33210 }, { "epoch": 3.51, - "grad_norm": 26.446552276611328, + "grad_norm": 65.60016632080078, "learning_rate": 1.2983419579681065e-05, - "loss": 0.4896, + "loss": 1.2009, "step": 33220 }, { "epoch": 3.51, - "grad_norm": 13.110478401184082, + "grad_norm": 32.654720306396484, "learning_rate": 1.2981307424226424e-05, - "loss": 0.4789, + "loss": 0.9997, "step": 33230 }, { "epoch": 3.51, - "grad_norm": 9.63801383972168, + "grad_norm": 47.37004089355469, "learning_rate": 1.2979195268771783e-05, - "loss": 0.7944, + "loss": 1.2752, "step": 33240 }, { "epoch": 3.51, - "grad_norm": 13.174065589904785, + "grad_norm": 49.92596435546875, "learning_rate": 1.2977083113317142e-05, - "loss": 0.3597, + "loss": 0.9011, "step": 33250 }, { "epoch": 3.51, - "grad_norm": 14.17449951171875, + "grad_norm": 28.618783950805664, "learning_rate": 1.29749709578625e-05, - "loss": 0.443, + "loss": 1.0983, "step": 33260 }, { "epoch": 3.51, - "grad_norm": 6.806898593902588, + "grad_norm": 48.076988220214844, "learning_rate": 1.2972858802407858e-05, - "loss": 0.3316, + "loss": 1.1544, "step": 33270 }, { "epoch": 3.51, - "grad_norm": 1.4098093509674072, + "grad_norm": 51.36908721923828, "learning_rate": 1.2970746646953216e-05, - "loss": 0.5235, + "loss": 0.964, "step": 33280 }, { "epoch": 3.52, - "grad_norm": 12.186551094055176, + "grad_norm": 41.67329406738281, "learning_rate": 1.2968634491498575e-05, - "loss": 0.7961, + "loss": 1.5037, "step": 33290 }, { "epoch": 3.52, - "grad_norm": 1.545609951019287, + "grad_norm": 51.74220275878906, "learning_rate": 1.2966522336043934e-05, - "loss": 0.6144, + "loss": 1.1874, "step": 33300 }, { "epoch": 3.52, - "grad_norm": 12.748263359069824, + "grad_norm": 41.83989334106445, "learning_rate": 1.2964410180589293e-05, - "loss": 0.6771, + "loss": 1.3306, "step": 33310 }, { "epoch": 3.52, - "grad_norm": 18.751619338989258, + "grad_norm": 27.650863647460938, "learning_rate": 1.2962298025134651e-05, - "loss": 0.539, + "loss": 0.8244, "step": 33320 }, { "epoch": 3.52, - "grad_norm": 9.835277557373047, + "grad_norm": 33.379032135009766, "learning_rate": 1.2960185869680009e-05, - "loss": 0.9638, + "loss": 1.5442, "step": 33330 }, { "epoch": 3.52, - "grad_norm": 16.164722442626953, + "grad_norm": 19.402128219604492, "learning_rate": 1.2958073714225367e-05, - "loss": 0.6087, + "loss": 0.8497, "step": 33340 }, { "epoch": 3.52, - "grad_norm": 20.2434139251709, + "grad_norm": 42.29977798461914, "learning_rate": 1.2955961558770726e-05, - "loss": 0.4813, + "loss": 1.0507, "step": 33350 }, { "epoch": 3.52, - "grad_norm": 26.015722274780273, + "grad_norm": 47.79705047607422, "learning_rate": 1.2953849403316085e-05, - "loss": 0.7169, + "loss": 1.2726, "step": 33360 }, { "epoch": 3.52, - "grad_norm": 20.057907104492188, + "grad_norm": 47.63486862182617, "learning_rate": 1.2951737247861444e-05, - "loss": 0.6693, + "loss": 1.2324, "step": 33370 }, { "epoch": 3.53, - "grad_norm": 14.061769485473633, + "grad_norm": 29.374469757080078, "learning_rate": 1.29496250924068e-05, - "loss": 0.3063, + "loss": 1.0231, "step": 33380 }, { "epoch": 3.53, - "grad_norm": 0.1793479472398758, + "grad_norm": 32.57422637939453, "learning_rate": 1.294751293695216e-05, - "loss": 0.3849, + "loss": 1.144, "step": 33390 }, { "epoch": 3.53, - "grad_norm": 8.003020286560059, + "grad_norm": 37.946537017822266, "learning_rate": 1.2945400781497518e-05, - "loss": 0.5039, + "loss": 1.1431, "step": 33400 }, { "epoch": 3.53, - "grad_norm": 14.514626502990723, + "grad_norm": 43.841796875, "learning_rate": 1.2943288626042879e-05, - "loss": 0.5004, + "loss": 1.0997, "step": 33410 }, { "epoch": 3.53, - "grad_norm": 30.83905792236328, + "grad_norm": 45.09420394897461, "learning_rate": 1.2941176470588238e-05, - "loss": 0.6659, + "loss": 1.1284, "step": 33420 }, { "epoch": 3.53, - "grad_norm": 5.630946636199951, + "grad_norm": 36.02935791015625, "learning_rate": 1.2939064315133596e-05, - "loss": 0.385, + "loss": 1.0352, "step": 33430 }, { "epoch": 3.53, - "grad_norm": 5.796767234802246, + "grad_norm": 32.96920394897461, "learning_rate": 1.2936952159678953e-05, - "loss": 0.3803, + "loss": 1.0425, "step": 33440 }, { "epoch": 3.53, - "grad_norm": 11.753731727600098, + "grad_norm": 26.081344604492188, "learning_rate": 1.2934840004224312e-05, - "loss": 0.3913, + "loss": 1.0364, "step": 33450 }, { "epoch": 3.53, - "grad_norm": 26.447412490844727, + "grad_norm": 29.143884658813477, "learning_rate": 1.2932727848769671e-05, - "loss": 0.86, + "loss": 0.9785, "step": 33460 }, { "epoch": 3.53, - "grad_norm": 15.378864288330078, + "grad_norm": 37.4066162109375, "learning_rate": 1.293061569331503e-05, - "loss": 0.405, + "loss": 1.1591, "step": 33470 }, { "epoch": 3.54, - "grad_norm": 6.032354354858398, + "grad_norm": 24.339475631713867, "learning_rate": 1.2928503537860389e-05, - "loss": 0.2982, + "loss": 1.33, "step": 33480 }, { "epoch": 3.54, - "grad_norm": 23.645614624023438, + "grad_norm": 13.215755462646484, "learning_rate": 1.2926391382405746e-05, - "loss": 0.5617, + "loss": 1.1954, "step": 33490 }, { "epoch": 3.54, - "grad_norm": 26.8010311126709, + "grad_norm": 42.970985412597656, "learning_rate": 1.2924279226951104e-05, - "loss": 0.5792, + "loss": 1.1213, "step": 33500 }, { "epoch": 3.54, - "grad_norm": 10.09343147277832, + "grad_norm": 19.809711456298828, "learning_rate": 1.2922167071496463e-05, - "loss": 0.6401, + "loss": 1.1286, "step": 33510 }, { "epoch": 3.54, - "grad_norm": 5.9908599853515625, + "grad_norm": 53.31538391113281, "learning_rate": 1.2920054916041822e-05, - "loss": 0.4945, + "loss": 1.301, "step": 33520 }, { "epoch": 3.54, - "grad_norm": 16.287660598754883, + "grad_norm": 47.129817962646484, "learning_rate": 1.291794276058718e-05, - "loss": 0.3641, + "loss": 1.2438, "step": 33530 }, { "epoch": 3.54, - "grad_norm": 12.389265060424805, + "grad_norm": 39.348777770996094, "learning_rate": 1.2915830605132538e-05, - "loss": 1.1316, + "loss": 1.2793, "step": 33540 }, { "epoch": 3.54, - "grad_norm": 31.225046157836914, + "grad_norm": 35.54317092895508, "learning_rate": 1.2913718449677897e-05, - "loss": 1.017, + "loss": 1.4504, "step": 33550 }, { "epoch": 3.54, - "grad_norm": 21.057161331176758, + "grad_norm": 31.00103187561035, "learning_rate": 1.2911606294223255e-05, - "loss": 0.6281, + "loss": 1.2579, "step": 33560 }, { "epoch": 3.55, - "grad_norm": 12.677767753601074, + "grad_norm": 47.29740524291992, "learning_rate": 1.2909494138768614e-05, - "loss": 0.7126, + "loss": 1.2061, "step": 33570 }, { "epoch": 3.55, - "grad_norm": 26.027721405029297, + "grad_norm": 50.39152145385742, "learning_rate": 1.2907381983313973e-05, - "loss": 0.4096, + "loss": 0.8821, "step": 33580 }, { "epoch": 3.55, - "grad_norm": 7.289355754852295, + "grad_norm": 25.510583877563477, "learning_rate": 1.2905269827859332e-05, - "loss": 0.442, + "loss": 1.0679, "step": 33590 }, { "epoch": 3.55, - "grad_norm": 8.154864311218262, + "grad_norm": 40.712440490722656, "learning_rate": 1.2903157672404689e-05, - "loss": 0.5478, + "loss": 0.9339, "step": 33600 }, { "epoch": 3.55, - "grad_norm": 12.73331356048584, + "grad_norm": 42.31846237182617, "learning_rate": 1.2901045516950048e-05, - "loss": 0.5386, + "loss": 1.2485, "step": 33610 }, { "epoch": 3.55, - "grad_norm": 18.758373260498047, + "grad_norm": 41.407135009765625, "learning_rate": 1.2898933361495406e-05, - "loss": 0.3707, + "loss": 1.0341, "step": 33620 }, { "epoch": 3.55, - "grad_norm": 12.188140869140625, + "grad_norm": 38.86554718017578, "learning_rate": 1.2896821206040767e-05, - "loss": 0.5754, + "loss": 1.2257, "step": 33630 }, { "epoch": 3.55, - "grad_norm": 27.376665115356445, + "grad_norm": 20.1569881439209, "learning_rate": 1.2894709050586126e-05, - "loss": 0.6909, + "loss": 1.1771, "step": 33640 }, { "epoch": 3.55, - "grad_norm": 14.260956764221191, + "grad_norm": 32.5914421081543, "learning_rate": 1.2892596895131481e-05, - "loss": 0.7112, + "loss": 1.211, "step": 33650 }, { "epoch": 3.55, - "grad_norm": 11.293816566467285, + "grad_norm": 33.303924560546875, "learning_rate": 1.2890484739676841e-05, - "loss": 0.5049, + "loss": 1.3309, "step": 33660 }, { "epoch": 3.56, - "grad_norm": 17.928211212158203, + "grad_norm": 45.830780029296875, "learning_rate": 1.28883725842222e-05, - "loss": 0.6347, + "loss": 1.2602, "step": 33670 }, { "epoch": 3.56, - "grad_norm": 23.516021728515625, + "grad_norm": 28.781606674194336, "learning_rate": 1.2886260428767559e-05, - "loss": 0.8189, + "loss": 1.3301, "step": 33680 }, { "epoch": 3.56, - "grad_norm": 10.628744125366211, + "grad_norm": 37.9376106262207, "learning_rate": 1.2884148273312918e-05, - "loss": 0.8157, + "loss": 1.4857, "step": 33690 }, { "epoch": 3.56, - "grad_norm": 31.15522003173828, + "grad_norm": 53.69748306274414, "learning_rate": 1.2882036117858277e-05, - "loss": 0.3363, + "loss": 1.2758, "step": 33700 }, { "epoch": 3.56, - "grad_norm": 13.67067813873291, + "grad_norm": 45.00674819946289, "learning_rate": 1.2879923962403634e-05, - "loss": 0.634, + "loss": 1.084, "step": 33710 }, { "epoch": 3.56, - "grad_norm": 19.097455978393555, + "grad_norm": 51.138694763183594, "learning_rate": 1.2877811806948992e-05, - "loss": 0.5122, + "loss": 0.9333, "step": 33720 }, { "epoch": 3.56, - "grad_norm": 3.3453354835510254, + "grad_norm": 42.079872131347656, "learning_rate": 1.2875699651494351e-05, - "loss": 0.8074, + "loss": 1.5911, "step": 33730 }, { "epoch": 3.56, - "grad_norm": 20.725175857543945, + "grad_norm": 40.32914352416992, "learning_rate": 1.287358749603971e-05, - "loss": 0.3088, + "loss": 0.9803, "step": 33740 }, { "epoch": 3.56, - "grad_norm": 30.057741165161133, + "grad_norm": 52.022193908691406, "learning_rate": 1.2871475340585069e-05, - "loss": 0.7626, + "loss": 1.3014, "step": 33750 }, { "epoch": 3.57, - "grad_norm": 29.783742904663086, + "grad_norm": 43.82051086425781, "learning_rate": 1.2869363185130426e-05, - "loss": 0.5098, + "loss": 1.2956, "step": 33760 }, { "epoch": 3.57, - "grad_norm": 16.089588165283203, + "grad_norm": 37.9089241027832, "learning_rate": 1.2867251029675785e-05, - "loss": 0.8783, + "loss": 1.4145, "step": 33770 }, { "epoch": 3.57, - "grad_norm": 12.589228630065918, + "grad_norm": 31.471508026123047, "learning_rate": 1.2865138874221143e-05, - "loss": 0.5881, + "loss": 1.1564, "step": 33780 }, { "epoch": 3.57, - "grad_norm": 10.253647804260254, + "grad_norm": 49.49138641357422, "learning_rate": 1.2863026718766502e-05, - "loss": 0.5285, + "loss": 1.0274, "step": 33790 }, { "epoch": 3.57, - "grad_norm": 6.630813121795654, + "grad_norm": 32.20684051513672, "learning_rate": 1.2860914563311861e-05, - "loss": 0.5093, + "loss": 1.2566, "step": 33800 }, { "epoch": 3.57, - "grad_norm": 11.492720603942871, + "grad_norm": 29.480178833007812, "learning_rate": 1.285880240785722e-05, - "loss": 0.5138, + "loss": 1.4063, "step": 33810 }, { "epoch": 3.57, - "grad_norm": 0.5740587711334229, + "grad_norm": 27.9869384765625, "learning_rate": 1.2856690252402577e-05, - "loss": 0.7712, + "loss": 1.1418, "step": 33820 }, { "epoch": 3.57, - "grad_norm": 15.00172233581543, + "grad_norm": 52.596534729003906, "learning_rate": 1.2854578096947936e-05, - "loss": 0.3836, + "loss": 1.1954, "step": 33830 }, { "epoch": 3.57, - "grad_norm": 20.352649688720703, + "grad_norm": 36.35414505004883, "learning_rate": 1.2852465941493294e-05, - "loss": 0.7257, + "loss": 1.2044, "step": 33840 }, { "epoch": 3.57, - "grad_norm": 14.967586517333984, + "grad_norm": 35.30870819091797, "learning_rate": 1.2850353786038653e-05, - "loss": 0.441, + "loss": 1.1508, "step": 33850 }, { "epoch": 3.58, - "grad_norm": 12.148510932922363, + "grad_norm": 49.58064270019531, "learning_rate": 1.2848241630584014e-05, - "loss": 0.7738, + "loss": 1.175, "step": 33860 }, { "epoch": 3.58, - "grad_norm": 11.857065200805664, + "grad_norm": 42.1296272277832, "learning_rate": 1.2846129475129369e-05, - "loss": 0.6115, + "loss": 1.1445, "step": 33870 }, { "epoch": 3.58, - "grad_norm": 32.7825813293457, + "grad_norm": 44.05475616455078, "learning_rate": 1.2844017319674728e-05, - "loss": 0.5578, + "loss": 1.1634, "step": 33880 }, { "epoch": 3.58, - "grad_norm": 2.0032496452331543, + "grad_norm": 53.83842086791992, "learning_rate": 1.2841905164220088e-05, - "loss": 0.5129, + "loss": 1.1759, "step": 33890 }, { "epoch": 3.58, - "grad_norm": 14.655134201049805, + "grad_norm": 48.03342056274414, "learning_rate": 1.2839793008765447e-05, - "loss": 0.6254, + "loss": 1.0816, "step": 33900 }, { "epoch": 3.58, - "grad_norm": 6.054452419281006, + "grad_norm": 30.438493728637695, "learning_rate": 1.2837680853310806e-05, - "loss": 0.3092, + "loss": 1.0433, "step": 33910 }, { "epoch": 3.58, - "grad_norm": 18.852893829345703, + "grad_norm": 37.72282409667969, "learning_rate": 1.2835568697856163e-05, - "loss": 0.4225, + "loss": 1.0794, "step": 33920 }, { "epoch": 3.58, - "grad_norm": 6.3495917320251465, + "grad_norm": 40.227577209472656, "learning_rate": 1.2833456542401522e-05, - "loss": 0.5489, + "loss": 1.2054, "step": 33930 }, { "epoch": 3.58, - "grad_norm": 0.034406889230012894, + "grad_norm": 32.29343032836914, "learning_rate": 1.283134438694688e-05, - "loss": 0.3611, + "loss": 0.9833, "step": 33940 }, { "epoch": 3.59, - "grad_norm": 8.030403137207031, + "grad_norm": 6.768853664398193, "learning_rate": 1.282923223149224e-05, - "loss": 0.407, + "loss": 1.1181, "step": 33950 }, { "epoch": 3.59, - "grad_norm": 2.381828546524048, + "grad_norm": 26.169755935668945, "learning_rate": 1.2827120076037598e-05, - "loss": 0.4589, + "loss": 1.136, "step": 33960 }, { "epoch": 3.59, - "grad_norm": 9.754781723022461, + "grad_norm": 21.12628173828125, "learning_rate": 1.2825007920582957e-05, - "loss": 0.6699, + "loss": 1.0567, "step": 33970 }, { "epoch": 3.59, - "grad_norm": 8.717979431152344, + "grad_norm": 37.37989807128906, "learning_rate": 1.2822895765128314e-05, - "loss": 0.6532, + "loss": 1.1651, "step": 33980 }, { "epoch": 3.59, - "grad_norm": 7.809572219848633, + "grad_norm": 31.21914291381836, "learning_rate": 1.2820783609673673e-05, - "loss": 0.6839, + "loss": 1.3333, "step": 33990 }, { "epoch": 3.59, - "grad_norm": 18.961862564086914, + "grad_norm": 38.78913116455078, "learning_rate": 1.2818671454219031e-05, - "loss": 0.5644, + "loss": 1.2041, "step": 34000 }, { "epoch": 3.59, - "grad_norm": 15.622038841247559, + "grad_norm": 50.227970123291016, "learning_rate": 1.281655929876439e-05, - "loss": 0.4548, + "loss": 1.2485, "step": 34010 }, { "epoch": 3.59, - "grad_norm": 1.4767030477523804, + "grad_norm": 30.554622650146484, "learning_rate": 1.2814447143309749e-05, - "loss": 0.8312, + "loss": 1.2198, "step": 34020 }, { "epoch": 3.59, - "grad_norm": 8.335099220275879, + "grad_norm": 61.51444625854492, "learning_rate": 1.2812334987855106e-05, - "loss": 0.7845, + "loss": 1.6152, "step": 34030 }, { "epoch": 3.59, - "grad_norm": 27.630794525146484, + "grad_norm": 66.49610900878906, "learning_rate": 1.2810222832400465e-05, - "loss": 0.6925, + "loss": 1.2459, "step": 34040 }, { "epoch": 3.6, - "grad_norm": 11.718159675598145, + "grad_norm": 33.0103759765625, "learning_rate": 1.2808110676945824e-05, - "loss": 0.6691, + "loss": 1.1335, "step": 34050 }, { "epoch": 3.6, - "grad_norm": 6.140860557556152, + "grad_norm": 26.032011032104492, "learning_rate": 1.2805998521491182e-05, - "loss": 0.4438, + "loss": 0.9299, "step": 34060 }, { "epoch": 3.6, - "grad_norm": 10.7158784866333, + "grad_norm": 39.543785095214844, "learning_rate": 1.2803886366036541e-05, - "loss": 0.5626, + "loss": 1.2742, "step": 34070 }, { "epoch": 3.6, - "grad_norm": 13.934319496154785, + "grad_norm": 37.95667266845703, "learning_rate": 1.2801774210581902e-05, - "loss": 0.4819, + "loss": 1.216, "step": 34080 }, { "epoch": 3.6, - "grad_norm": 9.191874504089355, + "grad_norm": 45.35942077636719, "learning_rate": 1.2799662055127257e-05, - "loss": 0.753, + "loss": 1.0934, "step": 34090 }, { "epoch": 3.6, - "grad_norm": 0.11230888217687607, + "grad_norm": 38.6247444152832, "learning_rate": 1.2797549899672616e-05, - "loss": 0.3745, + "loss": 1.1297, "step": 34100 }, { "epoch": 3.6, - "grad_norm": 23.505220413208008, + "grad_norm": 37.68797302246094, "learning_rate": 1.2795437744217976e-05, - "loss": 0.5674, + "loss": 1.1868, "step": 34110 }, { "epoch": 3.6, - "grad_norm": 5.159240245819092, + "grad_norm": 51.89111328125, "learning_rate": 1.2793325588763335e-05, - "loss": 0.3148, + "loss": 1.2859, "step": 34120 }, { "epoch": 3.6, - "grad_norm": 9.542684555053711, + "grad_norm": 33.7185173034668, "learning_rate": 1.2791213433308694e-05, - "loss": 1.1383, + "loss": 1.7042, "step": 34130 }, { "epoch": 3.61, - "grad_norm": 3.3457751274108887, + "grad_norm": 19.5267391204834, "learning_rate": 1.2789101277854051e-05, - "loss": 0.5179, + "loss": 0.7716, "step": 34140 }, { "epoch": 3.61, - "grad_norm": 10.903831481933594, + "grad_norm": 58.907405853271484, "learning_rate": 1.278698912239941e-05, - "loss": 0.7364, + "loss": 1.3493, "step": 34150 }, { "epoch": 3.61, - "grad_norm": 31.543758392333984, + "grad_norm": 50.30116271972656, "learning_rate": 1.2784876966944769e-05, - "loss": 0.6438, + "loss": 1.2926, "step": 34160 }, { "epoch": 3.61, - "grad_norm": 21.8030948638916, + "grad_norm": 59.783348083496094, "learning_rate": 1.2782764811490127e-05, - "loss": 0.6544, + "loss": 1.005, "step": 34170 }, { "epoch": 3.61, - "grad_norm": 6.269408226013184, + "grad_norm": 23.7695369720459, "learning_rate": 1.2780652656035486e-05, - "loss": 0.5194, + "loss": 1.0807, "step": 34180 }, { "epoch": 3.61, - "grad_norm": 19.887561798095703, + "grad_norm": 39.71097946166992, "learning_rate": 1.2778540500580843e-05, - "loss": 0.7543, + "loss": 1.1618, "step": 34190 }, { "epoch": 3.61, - "grad_norm": 29.65489387512207, + "grad_norm": 42.37001419067383, "learning_rate": 1.2776428345126202e-05, - "loss": 0.6826, + "loss": 1.1035, "step": 34200 }, { "epoch": 3.61, - "grad_norm": 17.687971115112305, + "grad_norm": 36.102569580078125, "learning_rate": 1.277431618967156e-05, - "loss": 0.4764, + "loss": 1.2986, "step": 34210 }, { "epoch": 3.61, - "grad_norm": 11.598567962646484, + "grad_norm": 25.562156677246094, "learning_rate": 1.277220403421692e-05, - "loss": 0.4828, + "loss": 1.0684, "step": 34220 }, { "epoch": 3.61, - "grad_norm": 0.2175760269165039, + "grad_norm": 38.80317306518555, "learning_rate": 1.2770091878762278e-05, - "loss": 0.4178, + "loss": 1.1617, "step": 34230 }, { "epoch": 3.62, - "grad_norm": 19.156946182250977, + "grad_norm": 48.064762115478516, "learning_rate": 1.2767979723307637e-05, - "loss": 0.5833, + "loss": 1.4367, "step": 34240 }, { "epoch": 3.62, - "grad_norm": 0.8370733857154846, + "grad_norm": 29.243310928344727, "learning_rate": 1.2765867567852994e-05, - "loss": 0.7303, + "loss": 1.5032, "step": 34250 }, { "epoch": 3.62, - "grad_norm": 13.974937438964844, + "grad_norm": 32.30441665649414, "learning_rate": 1.2763755412398353e-05, - "loss": 0.7369, + "loss": 1.3133, "step": 34260 }, { "epoch": 3.62, - "grad_norm": 0.8834003806114197, + "grad_norm": 28.352174758911133, "learning_rate": 1.2761643256943712e-05, - "loss": 0.5614, + "loss": 1.1117, "step": 34270 }, { "epoch": 3.62, - "grad_norm": 28.234766006469727, + "grad_norm": 43.54275131225586, "learning_rate": 1.275953110148907e-05, - "loss": 0.5707, + "loss": 1.4, "step": 34280 }, { "epoch": 3.62, - "grad_norm": 22.09073257446289, + "grad_norm": 40.32093811035156, "learning_rate": 1.275741894603443e-05, - "loss": 0.4967, + "loss": 1.295, "step": 34290 }, { "epoch": 3.62, - "grad_norm": 4.636205673217773, + "grad_norm": 31.879566192626953, "learning_rate": 1.2755306790579786e-05, - "loss": 0.6271, + "loss": 1.172, "step": 34300 }, { "epoch": 3.62, - "grad_norm": 15.172365188598633, + "grad_norm": 40.059261322021484, "learning_rate": 1.2753194635125145e-05, - "loss": 0.4239, + "loss": 1.1066, "step": 34310 }, { "epoch": 3.62, - "grad_norm": 10.062700271606445, + "grad_norm": 25.68743896484375, "learning_rate": 1.2751082479670504e-05, - "loss": 0.3946, + "loss": 0.8045, "step": 34320 }, { "epoch": 3.63, - "grad_norm": 5.330890655517578, + "grad_norm": 36.08552551269531, "learning_rate": 1.2748970324215864e-05, - "loss": 0.6402, + "loss": 1.1582, "step": 34330 }, { "epoch": 3.63, - "grad_norm": 14.072773933410645, + "grad_norm": 27.094684600830078, "learning_rate": 1.2746858168761223e-05, - "loss": 0.8082, + "loss": 1.2294, "step": 34340 }, { "epoch": 3.63, - "grad_norm": 8.985556602478027, + "grad_norm": 36.69760513305664, "learning_rate": 1.2744746013306582e-05, - "loss": 0.4492, + "loss": 1.0309, "step": 34350 }, { "epoch": 3.63, - "grad_norm": 25.531160354614258, + "grad_norm": 37.44609069824219, "learning_rate": 1.2742633857851937e-05, - "loss": 0.2574, + "loss": 0.8178, "step": 34360 }, { "epoch": 3.63, - "grad_norm": 2.90766978263855, + "grad_norm": 42.205360412597656, "learning_rate": 1.2740521702397298e-05, - "loss": 0.655, + "loss": 1.2776, "step": 34370 }, { "epoch": 3.63, - "grad_norm": 6.772893905639648, + "grad_norm": 46.90452194213867, "learning_rate": 1.2738409546942657e-05, - "loss": 0.7069, + "loss": 1.2832, "step": 34380 }, { "epoch": 3.63, - "grad_norm": 19.38296127319336, + "grad_norm": 32.43914794921875, "learning_rate": 1.2736297391488015e-05, - "loss": 0.4724, + "loss": 1.3806, "step": 34390 }, { "epoch": 3.63, - "grad_norm": 18.234210968017578, + "grad_norm": 39.22903823852539, "learning_rate": 1.2734185236033374e-05, - "loss": 0.8449, + "loss": 1.3523, "step": 34400 }, { "epoch": 3.63, - "grad_norm": 15.624956130981445, + "grad_norm": 40.627601623535156, "learning_rate": 1.2732073080578731e-05, - "loss": 0.571, + "loss": 1.0866, "step": 34410 }, { "epoch": 3.64, - "grad_norm": 22.961061477661133, + "grad_norm": 32.97720718383789, "learning_rate": 1.272996092512409e-05, - "loss": 0.6739, + "loss": 1.0492, "step": 34420 }, { "epoch": 3.64, - "grad_norm": 25.532032012939453, + "grad_norm": 41.891502380371094, "learning_rate": 1.2727848769669449e-05, - "loss": 0.5291, + "loss": 0.8958, "step": 34430 }, { "epoch": 3.64, - "grad_norm": 34.47566223144531, + "grad_norm": 56.73436737060547, "learning_rate": 1.2725736614214808e-05, - "loss": 0.5477, + "loss": 1.325, "step": 34440 }, { "epoch": 3.64, - "grad_norm": 19.786516189575195, + "grad_norm": 42.177215576171875, "learning_rate": 1.2723624458760166e-05, - "loss": 0.557, + "loss": 1.379, "step": 34450 }, { "epoch": 3.64, - "grad_norm": 24.90369415283203, + "grad_norm": 31.274629592895508, "learning_rate": 1.2721512303305523e-05, - "loss": 0.508, + "loss": 1.1537, "step": 34460 }, { "epoch": 3.64, - "grad_norm": 23.053569793701172, + "grad_norm": 47.828125, "learning_rate": 1.2719400147850882e-05, - "loss": 0.668, + "loss": 1.3524, "step": 34470 }, { "epoch": 3.64, - "grad_norm": 9.629806518554688, + "grad_norm": 35.516510009765625, "learning_rate": 1.2717287992396241e-05, - "loss": 0.5768, + "loss": 1.4689, "step": 34480 }, { "epoch": 3.64, - "grad_norm": 15.604917526245117, + "grad_norm": 32.474308013916016, "learning_rate": 1.27151758369416e-05, - "loss": 0.7162, + "loss": 1.3393, "step": 34490 }, { "epoch": 3.64, - "grad_norm": 0.5820751190185547, + "grad_norm": 47.46836853027344, "learning_rate": 1.2713063681486959e-05, - "loss": 0.6661, + "loss": 1.1547, "step": 34500 }, { "epoch": 3.64, - "grad_norm": 5.122661113739014, + "grad_norm": 38.11113739013672, "learning_rate": 1.2710951526032317e-05, - "loss": 0.8058, + "loss": 1.5457, "step": 34510 }, { "epoch": 3.65, - "grad_norm": 0.24039201438426971, + "grad_norm": 48.19895553588867, "learning_rate": 1.2708839370577674e-05, - "loss": 0.3157, + "loss": 1.2867, "step": 34520 }, { "epoch": 3.65, - "grad_norm": 6.2962822914123535, + "grad_norm": 31.252464294433594, "learning_rate": 1.2706727215123033e-05, - "loss": 0.5859, + "loss": 1.0315, "step": 34530 }, { "epoch": 3.65, - "grad_norm": 7.951483726501465, + "grad_norm": 40.114524841308594, "learning_rate": 1.2704615059668392e-05, - "loss": 0.6618, + "loss": 1.027, "step": 34540 }, { "epoch": 3.65, - "grad_norm": 8.535661697387695, + "grad_norm": 38.22282028198242, "learning_rate": 1.270250290421375e-05, - "loss": 0.5228, + "loss": 1.0028, "step": 34550 }, { "epoch": 3.65, - "grad_norm": 29.6446475982666, + "grad_norm": 47.27494812011719, "learning_rate": 1.2700390748759111e-05, - "loss": 0.7174, + "loss": 1.184, "step": 34560 }, { "epoch": 3.65, - "grad_norm": 15.111905097961426, + "grad_norm": 34.152889251708984, "learning_rate": 1.2698278593304467e-05, - "loss": 0.4896, + "loss": 0.9751, "step": 34570 }, { "epoch": 3.65, - "grad_norm": 12.347963333129883, + "grad_norm": 38.74595260620117, "learning_rate": 1.2696166437849825e-05, - "loss": 0.3544, + "loss": 1.1401, "step": 34580 }, { "epoch": 3.65, - "grad_norm": 0.35410624742507935, + "grad_norm": 50.48884201049805, "learning_rate": 1.2694054282395186e-05, - "loss": 0.4343, + "loss": 1.1263, "step": 34590 }, { "epoch": 3.65, - "grad_norm": 12.996331214904785, + "grad_norm": 37.72564697265625, "learning_rate": 1.2691942126940545e-05, - "loss": 0.574, + "loss": 1.3043, "step": 34600 }, { "epoch": 3.66, - "grad_norm": 9.6981201171875, + "grad_norm": 50.68256378173828, "learning_rate": 1.2689829971485903e-05, - "loss": 0.5752, + "loss": 1.246, "step": 34610 }, { "epoch": 3.66, - "grad_norm": 19.230161666870117, + "grad_norm": 34.65721130371094, "learning_rate": 1.2687717816031262e-05, - "loss": 0.6414, + "loss": 1.0999, "step": 34620 }, { "epoch": 3.66, - "grad_norm": 6.059290885925293, + "grad_norm": 28.30260467529297, "learning_rate": 1.268560566057662e-05, - "loss": 0.3408, + "loss": 1.0047, "step": 34630 }, { "epoch": 3.66, - "grad_norm": 13.815314292907715, + "grad_norm": 50.09162139892578, "learning_rate": 1.2683493505121978e-05, - "loss": 0.5195, + "loss": 1.3952, "step": 34640 }, { "epoch": 3.66, - "grad_norm": 30.517290115356445, + "grad_norm": 42.94657897949219, "learning_rate": 1.2681381349667337e-05, - "loss": 0.6729, + "loss": 1.2408, "step": 34650 }, { "epoch": 3.66, - "grad_norm": 8.462276458740234, + "grad_norm": 32.47108459472656, "learning_rate": 1.2679269194212696e-05, - "loss": 0.4452, + "loss": 1.0372, "step": 34660 }, { "epoch": 3.66, - "grad_norm": 3.684953451156616, + "grad_norm": 26.749818801879883, "learning_rate": 1.2677157038758054e-05, - "loss": 0.3369, + "loss": 1.1189, "step": 34670 }, { "epoch": 3.66, - "grad_norm": 0.284588485956192, + "grad_norm": 50.173622131347656, "learning_rate": 1.2675044883303411e-05, - "loss": 0.8727, + "loss": 1.0192, "step": 34680 }, { "epoch": 3.66, - "grad_norm": 17.580965042114258, + "grad_norm": 34.23025894165039, "learning_rate": 1.267293272784877e-05, - "loss": 0.2371, + "loss": 1.0137, "step": 34690 }, { "epoch": 3.66, - "grad_norm": 9.289464950561523, + "grad_norm": 39.27052688598633, "learning_rate": 1.2670820572394129e-05, - "loss": 0.9058, + "loss": 1.3044, "step": 34700 }, { "epoch": 3.67, - "grad_norm": 3.289062738418579, + "grad_norm": 45.64544677734375, "learning_rate": 1.2668708416939488e-05, - "loss": 0.5608, + "loss": 1.2305, "step": 34710 }, { "epoch": 3.67, - "grad_norm": 10.29616928100586, + "grad_norm": 38.68323516845703, "learning_rate": 1.2666596261484847e-05, - "loss": 0.394, + "loss": 0.8645, "step": 34720 }, { "epoch": 3.67, - "grad_norm": 12.221430778503418, + "grad_norm": 38.037818908691406, "learning_rate": 1.2664484106030205e-05, - "loss": 0.5634, + "loss": 1.0395, "step": 34730 }, { "epoch": 3.67, - "grad_norm": 2.7736456394195557, + "grad_norm": 33.866493225097656, "learning_rate": 1.2662371950575562e-05, - "loss": 0.3217, + "loss": 1.1912, "step": 34740 }, { "epoch": 3.67, - "grad_norm": 11.144701957702637, + "grad_norm": 37.13053512573242, "learning_rate": 1.2660259795120921e-05, - "loss": 0.2692, + "loss": 1.255, "step": 34750 }, { "epoch": 3.67, - "grad_norm": 9.144880294799805, + "grad_norm": 42.544281005859375, "learning_rate": 1.265814763966628e-05, - "loss": 0.5013, + "loss": 1.2645, "step": 34760 }, { "epoch": 3.67, - "grad_norm": 4.117453575134277, + "grad_norm": 27.696495056152344, "learning_rate": 1.2656035484211639e-05, - "loss": 0.6635, + "loss": 1.2641, "step": 34770 }, { "epoch": 3.67, - "grad_norm": 16.747146606445312, + "grad_norm": 48.3314208984375, "learning_rate": 1.2653923328757e-05, - "loss": 0.393, + "loss": 1.0316, "step": 34780 }, { "epoch": 3.67, - "grad_norm": 12.242300033569336, + "grad_norm": 24.45708656311035, "learning_rate": 1.2651811173302355e-05, - "loss": 0.577, + "loss": 1.1977, "step": 34790 }, { "epoch": 3.68, - "grad_norm": 8.909322738647461, + "grad_norm": 42.54243850708008, "learning_rate": 1.2649699017847713e-05, - "loss": 0.5851, + "loss": 1.2753, "step": 34800 }, { "epoch": 3.68, - "grad_norm": 13.398371696472168, + "grad_norm": 43.836490631103516, "learning_rate": 1.2647586862393074e-05, - "loss": 0.595, + "loss": 1.2334, "step": 34810 }, { "epoch": 3.68, - "grad_norm": 24.168203353881836, + "grad_norm": 43.840946197509766, "learning_rate": 1.2645474706938433e-05, - "loss": 0.3095, + "loss": 1.0627, "step": 34820 }, { "epoch": 3.68, - "grad_norm": 12.93101978302002, + "grad_norm": 26.92850685119629, "learning_rate": 1.2643362551483791e-05, - "loss": 0.5664, + "loss": 1.1869, "step": 34830 }, { "epoch": 3.68, - "grad_norm": 15.461198806762695, + "grad_norm": 41.91872024536133, "learning_rate": 1.2641250396029147e-05, - "loss": 0.7705, + "loss": 1.0411, "step": 34840 }, { "epoch": 3.68, - "grad_norm": 8.655065536499023, + "grad_norm": 16.498836517333984, "learning_rate": 1.2639138240574507e-05, - "loss": 0.7859, + "loss": 1.2107, "step": 34850 }, { "epoch": 3.68, - "grad_norm": 2.047729730606079, + "grad_norm": 39.75749206542969, "learning_rate": 1.2637026085119866e-05, - "loss": 0.2606, + "loss": 1.1959, "step": 34860 }, { "epoch": 3.68, - "grad_norm": 18.375789642333984, + "grad_norm": 50.461429595947266, "learning_rate": 1.2634913929665225e-05, - "loss": 0.4992, + "loss": 1.0604, "step": 34870 }, { "epoch": 3.68, - "grad_norm": 1.784454345703125, + "grad_norm": 42.374935150146484, "learning_rate": 1.2632801774210584e-05, - "loss": 0.4377, + "loss": 1.2398, "step": 34880 }, { "epoch": 3.68, - "grad_norm": 14.348886489868164, + "grad_norm": 25.806346893310547, "learning_rate": 1.2630689618755942e-05, - "loss": 0.5094, + "loss": 1.2192, "step": 34890 }, { "epoch": 3.69, - "grad_norm": 0.7651939988136292, + "grad_norm": 52.951904296875, "learning_rate": 1.26285774633013e-05, - "loss": 0.564, + "loss": 1.3409, "step": 34900 }, { "epoch": 3.69, - "grad_norm": 11.130791664123535, + "grad_norm": 22.556114196777344, "learning_rate": 1.2626465307846658e-05, - "loss": 0.795, + "loss": 1.2324, "step": 34910 }, { "epoch": 3.69, - "grad_norm": 13.660028457641602, + "grad_norm": 27.922489166259766, "learning_rate": 1.2624353152392017e-05, - "loss": 0.3855, + "loss": 0.9758, "step": 34920 }, { "epoch": 3.69, - "grad_norm": 17.197528839111328, + "grad_norm": 50.86097717285156, "learning_rate": 1.2622240996937376e-05, - "loss": 0.4448, + "loss": 1.0956, "step": 34930 }, { "epoch": 3.69, - "grad_norm": 24.527679443359375, + "grad_norm": 38.226688385009766, "learning_rate": 1.2620128841482735e-05, - "loss": 0.4881, + "loss": 1.0259, "step": 34940 }, { "epoch": 3.69, - "grad_norm": 24.011459350585938, + "grad_norm": 50.2130241394043, "learning_rate": 1.2618016686028092e-05, - "loss": 0.643, + "loss": 1.1878, "step": 34950 }, { "epoch": 3.69, - "grad_norm": 8.819382667541504, + "grad_norm": 45.15794372558594, "learning_rate": 1.261590453057345e-05, - "loss": 0.5222, + "loss": 1.3886, "step": 34960 }, { "epoch": 3.69, - "grad_norm": 12.050969123840332, + "grad_norm": 55.69698715209961, "learning_rate": 1.261379237511881e-05, - "loss": 0.3443, + "loss": 1.1241, "step": 34970 }, { "epoch": 3.69, - "grad_norm": 11.602999687194824, + "grad_norm": 51.9925537109375, "learning_rate": 1.2611680219664168e-05, - "loss": 0.7377, + "loss": 1.0904, "step": 34980 }, { "epoch": 3.7, - "grad_norm": 13.812212944030762, + "grad_norm": 29.538272857666016, "learning_rate": 1.2609568064209527e-05, - "loss": 0.3976, + "loss": 1.0412, "step": 34990 }, { "epoch": 3.7, - "grad_norm": 13.151237487792969, + "grad_norm": 45.400394439697266, "learning_rate": 1.2607455908754886e-05, - "loss": 0.3714, + "loss": 1.2384, "step": 35000 }, { "epoch": 3.7, - "grad_norm": 5.4023518562316895, + "grad_norm": 21.848644256591797, "learning_rate": 1.2605343753300243e-05, - "loss": 0.7279, + "loss": 1.4516, "step": 35010 }, { "epoch": 3.7, - "grad_norm": 6.521059036254883, + "grad_norm": 43.03287887573242, "learning_rate": 1.2603231597845602e-05, - "loss": 0.3216, + "loss": 1.0841, "step": 35020 }, { "epoch": 3.7, - "grad_norm": 13.66526985168457, + "grad_norm": 50.12517547607422, "learning_rate": 1.260111944239096e-05, - "loss": 0.8298, + "loss": 1.4802, "step": 35030 }, { "epoch": 3.7, - "grad_norm": 16.92237091064453, + "grad_norm": 23.38997459411621, "learning_rate": 1.259900728693632e-05, - "loss": 0.5245, + "loss": 0.9473, "step": 35040 }, { "epoch": 3.7, - "grad_norm": 9.994233131408691, + "grad_norm": 46.94938278198242, "learning_rate": 1.259689513148168e-05, - "loss": 0.4104, + "loss": 1.2768, "step": 35050 }, { "epoch": 3.7, - "grad_norm": 8.941736221313477, + "grad_norm": 45.436119079589844, "learning_rate": 1.2594782976027035e-05, - "loss": 0.4795, + "loss": 1.2123, "step": 35060 }, { "epoch": 3.7, - "grad_norm": 14.299365997314453, + "grad_norm": 26.693424224853516, "learning_rate": 1.2592670820572395e-05, - "loss": 0.5332, + "loss": 0.8828, "step": 35070 }, { "epoch": 3.7, - "grad_norm": 12.104814529418945, + "grad_norm": 40.08494567871094, "learning_rate": 1.2590558665117754e-05, - "loss": 0.8178, + "loss": 1.6055, "step": 35080 }, { "epoch": 3.71, - "grad_norm": 6.414833068847656, + "grad_norm": 46.75307083129883, "learning_rate": 1.2588446509663113e-05, - "loss": 0.4776, + "loss": 0.9207, "step": 35090 }, { "epoch": 3.71, - "grad_norm": 6.335198879241943, + "grad_norm": 39.130001068115234, "learning_rate": 1.2586334354208472e-05, - "loss": 0.4929, + "loss": 1.1365, "step": 35100 }, { "epoch": 3.71, - "grad_norm": 0.12554433941841125, + "grad_norm": 34.54237747192383, "learning_rate": 1.2584222198753829e-05, - "loss": 0.1989, + "loss": 0.948, "step": 35110 }, { "epoch": 3.71, - "grad_norm": 10.679957389831543, + "grad_norm": 24.240154266357422, "learning_rate": 1.2582110043299188e-05, - "loss": 0.4968, + "loss": 1.2353, "step": 35120 }, { "epoch": 3.71, - "grad_norm": 13.020296096801758, + "grad_norm": 37.47207260131836, "learning_rate": 1.2579997887844546e-05, - "loss": 0.7031, + "loss": 1.2755, "step": 35130 }, { "epoch": 3.71, - "grad_norm": 2.684425115585327, + "grad_norm": 30.103145599365234, "learning_rate": 1.2577885732389905e-05, - "loss": 0.5113, + "loss": 1.3615, "step": 35140 }, { "epoch": 3.71, - "grad_norm": 7.097226142883301, + "grad_norm": 19.77512550354004, "learning_rate": 1.2575773576935264e-05, - "loss": 0.5983, + "loss": 1.1029, "step": 35150 }, { "epoch": 3.71, - "grad_norm": 20.09187126159668, + "grad_norm": 45.683570861816406, "learning_rate": 1.2573661421480623e-05, - "loss": 0.499, + "loss": 0.9248, "step": 35160 }, { "epoch": 3.71, - "grad_norm": 5.622162342071533, + "grad_norm": 17.123794555664062, "learning_rate": 1.257154926602598e-05, - "loss": 0.7266, + "loss": 1.1219, "step": 35170 }, { "epoch": 3.72, - "grad_norm": 7.358402729034424, + "grad_norm": 24.036808013916016, "learning_rate": 1.2569437110571339e-05, - "loss": 0.6056, + "loss": 1.1854, "step": 35180 }, { "epoch": 3.72, - "grad_norm": 15.294683456420898, + "grad_norm": 34.63641357421875, "learning_rate": 1.2567324955116697e-05, - "loss": 0.502, + "loss": 1.1595, "step": 35190 }, { "epoch": 3.72, - "grad_norm": 9.911443710327148, + "grad_norm": 22.499755859375, "learning_rate": 1.2565212799662056e-05, - "loss": 0.5314, + "loss": 1.0079, "step": 35200 }, { "epoch": 3.72, - "grad_norm": 7.366403102874756, + "grad_norm": 30.22572135925293, "learning_rate": 1.2563100644207415e-05, - "loss": 0.5306, + "loss": 0.9869, "step": 35210 }, { "epoch": 3.72, - "grad_norm": 14.292398452758789, + "grad_norm": 30.928279876708984, "learning_rate": 1.2560988488752772e-05, - "loss": 0.7114, + "loss": 1.2464, "step": 35220 }, { "epoch": 3.72, - "grad_norm": 2.7197465896606445, + "grad_norm": 36.50362014770508, "learning_rate": 1.255887633329813e-05, - "loss": 0.5302, + "loss": 1.1943, "step": 35230 }, { "epoch": 3.72, - "grad_norm": 3.0763537883758545, + "grad_norm": 46.37535095214844, "learning_rate": 1.255676417784349e-05, - "loss": 0.6871, + "loss": 1.3821, "step": 35240 }, { "epoch": 3.72, - "grad_norm": 25.98747444152832, + "grad_norm": 60.2645378112793, "learning_rate": 1.2554652022388848e-05, - "loss": 0.8204, + "loss": 1.4331, "step": 35250 }, { "epoch": 3.72, - "grad_norm": 7.344202995300293, + "grad_norm": 40.25883865356445, "learning_rate": 1.2552539866934209e-05, - "loss": 0.5084, + "loss": 1.4156, "step": 35260 }, { "epoch": 3.72, - "grad_norm": 8.112126350402832, + "grad_norm": 56.45186996459961, "learning_rate": 1.2550427711479568e-05, - "loss": 0.7829, + "loss": 1.3819, "step": 35270 }, { "epoch": 3.73, - "grad_norm": 27.84942626953125, + "grad_norm": 45.57220458984375, "learning_rate": 1.2548315556024923e-05, - "loss": 0.5925, + "loss": 1.043, "step": 35280 }, { "epoch": 3.73, - "grad_norm": 9.051929473876953, + "grad_norm": 11.140972137451172, "learning_rate": 1.2546203400570283e-05, - "loss": 0.6256, + "loss": 0.9535, "step": 35290 }, { "epoch": 3.73, - "grad_norm": 11.429824829101562, + "grad_norm": 35.039695739746094, "learning_rate": 1.2544091245115642e-05, - "loss": 0.5527, + "loss": 1.2463, "step": 35300 }, { "epoch": 3.73, - "grad_norm": 31.33831787109375, + "grad_norm": 23.66414451599121, "learning_rate": 1.2541979089661001e-05, - "loss": 0.5567, + "loss": 1.1658, "step": 35310 }, { "epoch": 3.73, - "grad_norm": 1.085180401802063, + "grad_norm": 59.235145568847656, "learning_rate": 1.253986693420636e-05, - "loss": 0.3679, + "loss": 1.6313, "step": 35320 }, { "epoch": 3.73, - "grad_norm": 4.722566604614258, + "grad_norm": 42.137596130371094, "learning_rate": 1.2537754778751717e-05, - "loss": 0.4774, + "loss": 1.4918, "step": 35330 }, { "epoch": 3.73, - "grad_norm": 27.700368881225586, + "grad_norm": 36.874122619628906, "learning_rate": 1.2535642623297076e-05, - "loss": 0.6623, + "loss": 1.168, "step": 35340 }, { "epoch": 3.73, - "grad_norm": 10.180450439453125, + "grad_norm": 26.70676040649414, "learning_rate": 1.2533530467842434e-05, - "loss": 0.6567, + "loss": 1.0663, "step": 35350 }, { "epoch": 3.73, - "grad_norm": 2.9809672832489014, + "grad_norm": 38.685245513916016, "learning_rate": 1.2531418312387793e-05, - "loss": 0.636, + "loss": 1.1694, "step": 35360 }, { "epoch": 3.74, - "grad_norm": 16.78776741027832, + "grad_norm": 44.9119873046875, "learning_rate": 1.2529306156933152e-05, - "loss": 0.7351, + "loss": 1.0653, "step": 35370 }, { "epoch": 3.74, - "grad_norm": 1.313507318496704, + "grad_norm": 47.179569244384766, "learning_rate": 1.252719400147851e-05, - "loss": 0.5396, + "loss": 1.1244, "step": 35380 }, { "epoch": 3.74, - "grad_norm": 20.389366149902344, + "grad_norm": 32.61088562011719, "learning_rate": 1.2525081846023868e-05, - "loss": 0.7195, + "loss": 1.2656, "step": 35390 }, { "epoch": 3.74, - "grad_norm": 27.135875701904297, + "grad_norm": 52.09065246582031, "learning_rate": 1.2522969690569227e-05, - "loss": 0.6444, + "loss": 1.6733, "step": 35400 }, { "epoch": 3.74, - "grad_norm": 10.650126457214355, + "grad_norm": 58.60667037963867, "learning_rate": 1.2520857535114585e-05, - "loss": 0.7092, + "loss": 1.4096, "step": 35410 }, { "epoch": 3.74, - "grad_norm": 16.27841567993164, + "grad_norm": 37.416202545166016, "learning_rate": 1.2518745379659944e-05, - "loss": 0.6692, + "loss": 1.2758, "step": 35420 }, { "epoch": 3.74, - "grad_norm": 1.7088669538497925, + "grad_norm": 37.92767333984375, "learning_rate": 1.2516633224205303e-05, - "loss": 0.3892, + "loss": 1.1544, "step": 35430 }, { "epoch": 3.74, - "grad_norm": 13.676712036132812, + "grad_norm": 51.876895904541016, "learning_rate": 1.251452106875066e-05, - "loss": 0.5274, + "loss": 1.0039, "step": 35440 }, { "epoch": 3.74, - "grad_norm": 25.074108123779297, + "grad_norm": 36.604610443115234, "learning_rate": 1.2512408913296019e-05, - "loss": 0.6831, + "loss": 1.3586, "step": 35450 }, { "epoch": 3.74, - "grad_norm": 2.707345485687256, + "grad_norm": 28.938892364501953, "learning_rate": 1.2510296757841378e-05, - "loss": 0.5529, + "loss": 0.9549, "step": 35460 }, { "epoch": 3.75, - "grad_norm": 13.288298606872559, + "grad_norm": 32.22504806518555, "learning_rate": 1.2508184602386736e-05, - "loss": 0.5808, + "loss": 1.2166, "step": 35470 }, { "epoch": 3.75, - "grad_norm": 9.224997520446777, + "grad_norm": 39.31235885620117, "learning_rate": 1.2506072446932095e-05, - "loss": 0.4709, + "loss": 1.3433, "step": 35480 }, { "epoch": 3.75, - "grad_norm": 24.732404708862305, + "grad_norm": 60.60648727416992, "learning_rate": 1.2503960291477452e-05, - "loss": 0.9029, + "loss": 1.5052, "step": 35490 }, { "epoch": 3.75, - "grad_norm": 16.855945587158203, + "grad_norm": 18.6870059967041, "learning_rate": 1.2501848136022811e-05, - "loss": 0.4664, + "loss": 0.9265, "step": 35500 }, { "epoch": 3.75, - "grad_norm": 1.5923134088516235, + "grad_norm": 32.32080841064453, "learning_rate": 1.249973598056817e-05, - "loss": 0.4672, + "loss": 1.1418, "step": 35510 }, { "epoch": 3.75, - "grad_norm": 1.2381259202957153, + "grad_norm": 18.77983856201172, "learning_rate": 1.249762382511353e-05, - "loss": 0.4604, + "loss": 1.1361, "step": 35520 }, { "epoch": 3.75, - "grad_norm": 19.608430862426758, + "grad_norm": 47.364776611328125, "learning_rate": 1.2495511669658889e-05, - "loss": 0.8262, + "loss": 1.1584, "step": 35530 }, { "epoch": 3.75, - "grad_norm": 9.276997566223145, + "grad_norm": 47.8478889465332, "learning_rate": 1.2493399514204248e-05, - "loss": 0.4289, + "loss": 1.1055, "step": 35540 }, { "epoch": 3.75, - "grad_norm": 4.454230785369873, + "grad_norm": 41.56999969482422, "learning_rate": 1.2491287358749605e-05, - "loss": 0.4903, + "loss": 0.9357, "step": 35550 }, { "epoch": 3.76, - "grad_norm": 1.0341373682022095, + "grad_norm": 34.97898864746094, "learning_rate": 1.2489175203294964e-05, - "loss": 0.5306, + "loss": 1.3531, "step": 35560 }, { "epoch": 3.76, - "grad_norm": 6.249757766723633, + "grad_norm": 48.38812255859375, "learning_rate": 1.2487063047840322e-05, - "loss": 0.5472, + "loss": 1.2718, "step": 35570 }, { "epoch": 3.76, - "grad_norm": 10.709190368652344, + "grad_norm": 35.95254898071289, "learning_rate": 1.2484950892385681e-05, - "loss": 0.4074, + "loss": 1.031, "step": 35580 }, { "epoch": 3.76, - "grad_norm": 10.328039169311523, + "grad_norm": 64.69647979736328, "learning_rate": 1.248283873693104e-05, - "loss": 0.4922, + "loss": 1.4275, "step": 35590 }, { "epoch": 3.76, - "grad_norm": 5.5934553146362305, + "grad_norm": 26.630308151245117, "learning_rate": 1.2480726581476397e-05, - "loss": 0.69, + "loss": 1.178, "step": 35600 }, { "epoch": 3.76, - "grad_norm": 26.21662712097168, + "grad_norm": 25.76730728149414, "learning_rate": 1.2478614426021756e-05, - "loss": 0.6632, + "loss": 1.1556, "step": 35610 }, { "epoch": 3.76, - "grad_norm": 18.91297721862793, + "grad_norm": 38.43231201171875, "learning_rate": 1.2476502270567115e-05, - "loss": 0.5788, + "loss": 1.0818, "step": 35620 }, { "epoch": 3.76, - "grad_norm": 6.050042152404785, + "grad_norm": 47.11017608642578, "learning_rate": 1.2474390115112473e-05, - "loss": 0.6456, + "loss": 1.0892, "step": 35630 }, { "epoch": 3.76, - "grad_norm": 3.8670272827148438, + "grad_norm": 43.257564544677734, "learning_rate": 1.2472277959657832e-05, - "loss": 0.5137, + "loss": 1.1719, "step": 35640 }, { "epoch": 3.76, - "grad_norm": 7.5877604484558105, + "grad_norm": 39.748374938964844, "learning_rate": 1.2470165804203191e-05, - "loss": 0.5, + "loss": 1.0125, "step": 35650 }, { "epoch": 3.77, - "grad_norm": 10.282142639160156, + "grad_norm": 25.836326599121094, "learning_rate": 1.2468053648748548e-05, - "loss": 0.5429, + "loss": 1.2122, "step": 35660 }, { "epoch": 3.77, - "grad_norm": 53.79902267456055, + "grad_norm": 65.39049530029297, "learning_rate": 1.2465941493293907e-05, - "loss": 0.6026, + "loss": 1.1495, "step": 35670 }, { "epoch": 3.77, - "grad_norm": 18.416378021240234, + "grad_norm": 33.259708404541016, "learning_rate": 1.2463829337839266e-05, - "loss": 0.5773, + "loss": 1.1948, "step": 35680 }, { "epoch": 3.77, - "grad_norm": 2.353283166885376, + "grad_norm": 23.933975219726562, "learning_rate": 1.2461717182384624e-05, - "loss": 0.7454, + "loss": 1.2965, "step": 35690 }, { "epoch": 3.77, - "grad_norm": 0.06543005257844925, + "grad_norm": 43.34613800048828, "learning_rate": 1.2459605026929983e-05, - "loss": 0.6855, + "loss": 1.1902, "step": 35700 }, { "epoch": 3.77, - "grad_norm": 11.60256290435791, + "grad_norm": 35.814781188964844, "learning_rate": 1.245749287147534e-05, - "loss": 0.9053, + "loss": 1.6115, "step": 35710 }, { "epoch": 3.77, - "grad_norm": 1.3547701835632324, + "grad_norm": 16.589454650878906, "learning_rate": 1.2455380716020699e-05, - "loss": 0.5426, + "loss": 1.0546, "step": 35720 }, { "epoch": 3.77, - "grad_norm": 24.908525466918945, + "grad_norm": 46.25419235229492, "learning_rate": 1.2453268560566058e-05, - "loss": 0.5531, + "loss": 1.0349, "step": 35730 }, { "epoch": 3.77, - "grad_norm": 15.627737045288086, + "grad_norm": 29.872928619384766, "learning_rate": 1.2451156405111418e-05, - "loss": 0.3283, + "loss": 0.9388, "step": 35740 }, { "epoch": 3.78, - "grad_norm": 20.36515998840332, + "grad_norm": 38.092899322509766, "learning_rate": 1.2449044249656777e-05, - "loss": 0.4615, + "loss": 1.2937, "step": 35750 }, { "epoch": 3.78, - "grad_norm": 36.02442932128906, + "grad_norm": 57.93838882446289, "learning_rate": 1.2446932094202133e-05, - "loss": 0.7009, + "loss": 1.2339, "step": 35760 }, { "epoch": 3.78, - "grad_norm": 2.4872665405273438, + "grad_norm": 27.538537979125977, "learning_rate": 1.2444819938747493e-05, - "loss": 0.4282, + "loss": 1.2563, "step": 35770 }, { "epoch": 3.78, - "grad_norm": 0.7555986046791077, + "grad_norm": 18.609567642211914, "learning_rate": 1.2442707783292852e-05, - "loss": 0.5835, + "loss": 1.0922, "step": 35780 }, { "epoch": 3.78, - "grad_norm": 18.185483932495117, + "grad_norm": 31.973554611206055, "learning_rate": 1.244059562783821e-05, - "loss": 0.6209, + "loss": 1.2251, "step": 35790 }, { "epoch": 3.78, - "grad_norm": 13.30778980255127, + "grad_norm": 37.37803268432617, "learning_rate": 1.243848347238357e-05, - "loss": 0.5765, + "loss": 1.3963, "step": 35800 }, { "epoch": 3.78, - "grad_norm": 7.466805934906006, + "grad_norm": 42.020668029785156, "learning_rate": 1.2436371316928928e-05, - "loss": 0.7511, + "loss": 1.0035, "step": 35810 }, { "epoch": 3.78, - "grad_norm": 5.7568206787109375, + "grad_norm": 28.278886795043945, "learning_rate": 1.2434259161474285e-05, - "loss": 0.5928, + "loss": 1.1165, "step": 35820 }, { "epoch": 3.78, - "grad_norm": 11.783829689025879, + "grad_norm": 49.45875549316406, "learning_rate": 1.2432147006019644e-05, - "loss": 0.5555, + "loss": 1.4287, "step": 35830 }, { "epoch": 3.78, - "grad_norm": 11.088828086853027, + "grad_norm": 30.882734298706055, "learning_rate": 1.2430034850565003e-05, - "loss": 0.4654, + "loss": 1.0087, "step": 35840 }, { "epoch": 3.79, - "grad_norm": 1.4434269666671753, + "grad_norm": 39.198402404785156, "learning_rate": 1.2427922695110362e-05, - "loss": 0.4858, + "loss": 1.0095, "step": 35850 }, { "epoch": 3.79, - "grad_norm": 22.87458038330078, + "grad_norm": 47.644737243652344, "learning_rate": 1.242581053965572e-05, - "loss": 0.3232, + "loss": 0.9441, "step": 35860 }, { "epoch": 3.79, - "grad_norm": 19.404897689819336, + "grad_norm": 19.510507583618164, "learning_rate": 1.2423698384201077e-05, - "loss": 0.7286, + "loss": 1.1826, "step": 35870 }, { "epoch": 3.79, - "grad_norm": 9.356703758239746, + "grad_norm": 37.69325256347656, "learning_rate": 1.2421586228746436e-05, - "loss": 0.6593, + "loss": 1.4421, "step": 35880 }, { "epoch": 3.79, - "grad_norm": 46.2047233581543, + "grad_norm": 49.01369857788086, "learning_rate": 1.2419474073291795e-05, - "loss": 0.4594, + "loss": 1.1196, "step": 35890 }, { "epoch": 3.79, - "grad_norm": 9.262524604797363, + "grad_norm": 34.933937072753906, "learning_rate": 1.2417361917837154e-05, - "loss": 0.4849, + "loss": 1.1985, "step": 35900 }, { "epoch": 3.79, - "grad_norm": 21.05991554260254, + "grad_norm": 58.37113571166992, "learning_rate": 1.2415249762382512e-05, - "loss": 0.8734, + "loss": 1.2527, "step": 35910 }, { "epoch": 3.79, - "grad_norm": 4.941697120666504, + "grad_norm": 38.217227935791016, "learning_rate": 1.2413137606927871e-05, - "loss": 0.3616, + "loss": 1.1913, "step": 35920 }, { "epoch": 3.79, - "grad_norm": 32.28709030151367, + "grad_norm": 46.52936935424805, "learning_rate": 1.2411025451473228e-05, - "loss": 0.5093, + "loss": 0.9082, "step": 35930 }, { "epoch": 3.8, - "grad_norm": 23.198205947875977, + "grad_norm": 52.012393951416016, "learning_rate": 1.2408913296018587e-05, - "loss": 0.5696, + "loss": 1.2218, "step": 35940 }, { "epoch": 3.8, - "grad_norm": 6.908426284790039, + "grad_norm": 24.183948516845703, "learning_rate": 1.2406801140563946e-05, - "loss": 0.481, + "loss": 1.1306, "step": 35950 }, { "epoch": 3.8, - "grad_norm": 11.493569374084473, + "grad_norm": 34.908016204833984, "learning_rate": 1.2404688985109305e-05, - "loss": 0.4774, + "loss": 1.2424, "step": 35960 }, { "epoch": 3.8, - "grad_norm": 8.239828109741211, + "grad_norm": 22.347679138183594, "learning_rate": 1.2402576829654665e-05, - "loss": 0.3422, + "loss": 1.1647, "step": 35970 }, { "epoch": 3.8, - "grad_norm": 17.087047576904297, + "grad_norm": 44.26072311401367, "learning_rate": 1.240046467420002e-05, - "loss": 0.6751, + "loss": 1.3132, "step": 35980 }, { "epoch": 3.8, - "grad_norm": 8.649214744567871, + "grad_norm": 35.269920349121094, "learning_rate": 1.239835251874538e-05, - "loss": 0.5097, + "loss": 0.849, "step": 35990 }, { "epoch": 3.8, - "grad_norm": 11.98400592803955, + "grad_norm": 33.294189453125, "learning_rate": 1.239624036329074e-05, - "loss": 0.5668, + "loss": 1.1451, "step": 36000 }, { "epoch": 3.8, - "grad_norm": 27.408823013305664, + "grad_norm": 29.579721450805664, "learning_rate": 1.2394128207836099e-05, - "loss": 0.6993, + "loss": 1.1103, "step": 36010 }, { "epoch": 3.8, - "grad_norm": 14.603255271911621, + "grad_norm": 52.9396858215332, "learning_rate": 1.2392016052381457e-05, - "loss": 0.4843, + "loss": 1.4614, "step": 36020 }, { "epoch": 3.81, - "grad_norm": 14.238158226013184, + "grad_norm": 27.222593307495117, "learning_rate": 1.2389903896926814e-05, - "loss": 0.612, + "loss": 1.2592, "step": 36030 }, { "epoch": 3.81, - "grad_norm": 8.254240989685059, + "grad_norm": 42.199649810791016, "learning_rate": 1.2387791741472173e-05, - "loss": 0.4843, + "loss": 1.0104, "step": 36040 }, { "epoch": 3.81, - "grad_norm": 16.873394012451172, + "grad_norm": 44.56316375732422, "learning_rate": 1.2385679586017532e-05, - "loss": 0.4309, + "loss": 1.0441, "step": 36050 }, { "epoch": 3.81, - "grad_norm": 15.081707000732422, + "grad_norm": 67.80950927734375, "learning_rate": 1.238356743056289e-05, - "loss": 0.5424, + "loss": 1.2201, "step": 36060 }, { "epoch": 3.81, - "grad_norm": 3.48614501953125, + "grad_norm": 26.19717788696289, "learning_rate": 1.238145527510825e-05, - "loss": 0.4936, + "loss": 1.0713, "step": 36070 }, { "epoch": 3.81, - "grad_norm": 16.37611961364746, + "grad_norm": 62.283287048339844, "learning_rate": 1.2379343119653608e-05, - "loss": 0.6498, + "loss": 1.4953, "step": 36080 }, { "epoch": 3.81, - "grad_norm": 11.636210441589355, + "grad_norm": 41.05883026123047, "learning_rate": 1.2377230964198965e-05, - "loss": 0.4999, + "loss": 1.2368, "step": 36090 }, { "epoch": 3.81, - "grad_norm": 10.419869422912598, + "grad_norm": 23.975982666015625, "learning_rate": 1.2375118808744324e-05, - "loss": 0.4327, + "loss": 1.2803, "step": 36100 }, { "epoch": 3.81, - "grad_norm": 0.524036169052124, + "grad_norm": 28.892013549804688, "learning_rate": 1.2373006653289683e-05, - "loss": 0.6301, + "loss": 1.2358, "step": 36110 }, { "epoch": 3.81, - "grad_norm": 2.1700785160064697, + "grad_norm": 19.9884090423584, "learning_rate": 1.2370894497835042e-05, - "loss": 0.465, + "loss": 1.0076, "step": 36120 }, { "epoch": 3.82, - "grad_norm": 14.206892013549805, + "grad_norm": 46.10129928588867, "learning_rate": 1.23687823423804e-05, - "loss": 0.497, + "loss": 0.9948, "step": 36130 }, { "epoch": 3.82, - "grad_norm": 6.070621967315674, + "grad_norm": 26.0598201751709, "learning_rate": 1.2366670186925758e-05, - "loss": 0.2417, + "loss": 0.9634, "step": 36140 }, { "epoch": 3.82, - "grad_norm": 9.289299964904785, + "grad_norm": 45.43126678466797, "learning_rate": 1.2364558031471116e-05, - "loss": 0.361, + "loss": 1.2419, "step": 36150 }, { "epoch": 3.82, - "grad_norm": 7.829161643981934, + "grad_norm": 41.482269287109375, "learning_rate": 1.2362445876016475e-05, - "loss": 0.9193, + "loss": 1.5533, "step": 36160 }, { "epoch": 3.82, - "grad_norm": 0.2702881693840027, + "grad_norm": 36.62955093383789, "learning_rate": 1.2360333720561834e-05, - "loss": 0.4916, + "loss": 1.0769, "step": 36170 }, { "epoch": 3.82, - "grad_norm": 13.589778900146484, + "grad_norm": 34.88429260253906, "learning_rate": 1.2358221565107193e-05, - "loss": 0.3717, + "loss": 1.0938, "step": 36180 }, { "epoch": 3.82, - "grad_norm": 10.46250057220459, + "grad_norm": 37.82277297973633, "learning_rate": 1.2356109409652553e-05, - "loss": 0.4582, + "loss": 1.3653, "step": 36190 }, { "epoch": 3.82, - "grad_norm": 22.163211822509766, + "grad_norm": 56.21626663208008, "learning_rate": 1.2353997254197909e-05, - "loss": 0.776, + "loss": 1.5745, "step": 36200 }, { "epoch": 3.82, - "grad_norm": 11.601105690002441, + "grad_norm": 26.569883346557617, "learning_rate": 1.2351885098743267e-05, - "loss": 0.3564, + "loss": 1.115, "step": 36210 }, { "epoch": 3.83, - "grad_norm": 0.595083475112915, + "grad_norm": 45.13542175292969, "learning_rate": 1.2349772943288628e-05, - "loss": 0.4478, + "loss": 1.1831, "step": 36220 }, { "epoch": 3.83, - "grad_norm": 34.29808044433594, + "grad_norm": 29.009111404418945, "learning_rate": 1.2347660787833987e-05, - "loss": 0.6793, + "loss": 1.1011, "step": 36230 }, { "epoch": 3.83, - "grad_norm": 8.87293815612793, + "grad_norm": 41.49773406982422, "learning_rate": 1.2345548632379345e-05, - "loss": 0.4932, + "loss": 1.2445, "step": 36240 }, { "epoch": 3.83, - "grad_norm": 2.877624034881592, + "grad_norm": 44.8901252746582, "learning_rate": 1.2343436476924703e-05, - "loss": 0.7348, + "loss": 1.3669, "step": 36250 }, { "epoch": 3.83, - "grad_norm": 10.336973190307617, + "grad_norm": 41.58526611328125, "learning_rate": 1.2341324321470061e-05, - "loss": 0.5452, + "loss": 0.8278, "step": 36260 }, { "epoch": 3.83, - "grad_norm": 11.417346000671387, + "grad_norm": 60.493614196777344, "learning_rate": 1.233921216601542e-05, - "loss": 0.7153, + "loss": 1.2688, "step": 36270 }, { "epoch": 3.83, - "grad_norm": 14.942082405090332, + "grad_norm": 44.41202926635742, "learning_rate": 1.2337100010560779e-05, - "loss": 0.5152, + "loss": 1.1218, "step": 36280 }, { "epoch": 3.83, - "grad_norm": 8.582841873168945, + "grad_norm": 43.40708923339844, "learning_rate": 1.2334987855106138e-05, - "loss": 0.5808, + "loss": 1.3144, "step": 36290 }, { "epoch": 3.83, - "grad_norm": 12.569143295288086, + "grad_norm": 37.35320281982422, "learning_rate": 1.2332875699651496e-05, - "loss": 0.7445, + "loss": 1.2481, "step": 36300 }, { "epoch": 3.83, - "grad_norm": 31.794267654418945, + "grad_norm": 49.497406005859375, "learning_rate": 1.2330763544196853e-05, - "loss": 0.5789, + "loss": 1.1903, "step": 36310 }, { "epoch": 3.84, - "grad_norm": 14.553699493408203, + "grad_norm": 43.315643310546875, "learning_rate": 1.2328651388742212e-05, - "loss": 0.4982, + "loss": 1.1061, "step": 36320 }, { "epoch": 3.84, - "grad_norm": 3.0201070308685303, + "grad_norm": 33.5101432800293, "learning_rate": 1.2326539233287571e-05, - "loss": 0.4843, + "loss": 1.3568, "step": 36330 }, { "epoch": 3.84, - "grad_norm": 35.78764724731445, + "grad_norm": 41.629703521728516, "learning_rate": 1.232442707783293e-05, - "loss": 0.5507, + "loss": 0.9341, "step": 36340 }, { "epoch": 3.84, - "grad_norm": 14.170534133911133, + "grad_norm": 44.8905029296875, "learning_rate": 1.2322314922378289e-05, - "loss": 0.7143, + "loss": 1.2764, "step": 36350 }, { "epoch": 3.84, - "grad_norm": 7.772879600524902, + "grad_norm": 42.928627014160156, "learning_rate": 1.2320202766923646e-05, - "loss": 0.5123, + "loss": 0.9185, "step": 36360 }, { "epoch": 3.84, - "grad_norm": 20.796855926513672, + "grad_norm": 39.2603645324707, "learning_rate": 1.2318090611469004e-05, - "loss": 0.7688, + "loss": 1.226, "step": 36370 }, { "epoch": 3.84, - "grad_norm": 27.856178283691406, + "grad_norm": 40.90132141113281, "learning_rate": 1.2315978456014363e-05, - "loss": 0.6967, + "loss": 1.3031, "step": 36380 }, { "epoch": 3.84, - "grad_norm": 12.280732154846191, + "grad_norm": 50.69279861450195, "learning_rate": 1.2313866300559722e-05, - "loss": 0.6944, + "loss": 1.0295, "step": 36390 }, { "epoch": 3.84, - "grad_norm": 5.253574848175049, + "grad_norm": 42.00298309326172, "learning_rate": 1.231175414510508e-05, - "loss": 0.4166, + "loss": 1.0098, "step": 36400 }, { "epoch": 3.85, - "grad_norm": 0.9896496534347534, + "grad_norm": 10.594656944274902, "learning_rate": 1.2309641989650438e-05, - "loss": 0.5903, + "loss": 1.2217, "step": 36410 }, { "epoch": 3.85, - "grad_norm": 21.374919891357422, + "grad_norm": 37.95171356201172, "learning_rate": 1.2307529834195797e-05, - "loss": 0.6633, + "loss": 1.1212, "step": 36420 }, { "epoch": 3.85, - "grad_norm": 25.6928768157959, + "grad_norm": 42.53965759277344, "learning_rate": 1.2305417678741155e-05, - "loss": 0.7392, + "loss": 1.5706, "step": 36430 }, { "epoch": 3.85, - "grad_norm": 9.020646095275879, + "grad_norm": 48.0261344909668, "learning_rate": 1.2303305523286514e-05, - "loss": 0.5274, + "loss": 0.9328, "step": 36440 }, { "epoch": 3.85, - "grad_norm": 1.9587074518203735, + "grad_norm": 52.7554931640625, "learning_rate": 1.2301193367831875e-05, - "loss": 0.2666, + "loss": 0.9759, "step": 36450 }, { "epoch": 3.85, - "grad_norm": 11.777399063110352, + "grad_norm": 39.24583053588867, "learning_rate": 1.2299081212377233e-05, - "loss": 0.7298, + "loss": 0.9658, "step": 36460 }, { "epoch": 3.85, - "grad_norm": 8.422381401062012, + "grad_norm": 53.875919342041016, "learning_rate": 1.2296969056922589e-05, - "loss": 0.4574, + "loss": 1.1993, "step": 36470 }, { "epoch": 3.85, - "grad_norm": 4.442147254943848, + "grad_norm": 42.656646728515625, "learning_rate": 1.229485690146795e-05, - "loss": 0.3884, + "loss": 1.0016, "step": 36480 }, { "epoch": 3.85, - "grad_norm": 13.240493774414062, + "grad_norm": 43.03586959838867, "learning_rate": 1.2292744746013308e-05, - "loss": 0.4817, + "loss": 0.8377, "step": 36490 }, { "epoch": 3.85, - "grad_norm": 10.18341064453125, + "grad_norm": 38.3391227722168, "learning_rate": 1.2290632590558667e-05, - "loss": 0.6846, + "loss": 1.1862, "step": 36500 }, { "epoch": 3.86, - "grad_norm": 13.855112075805664, + "grad_norm": 34.71149444580078, "learning_rate": 1.2288520435104026e-05, - "loss": 0.3367, + "loss": 1.0946, "step": 36510 }, { "epoch": 3.86, - "grad_norm": 12.240659713745117, + "grad_norm": 42.526344299316406, "learning_rate": 1.2286408279649383e-05, - "loss": 0.6096, + "loss": 0.8794, "step": 36520 }, { "epoch": 3.86, - "grad_norm": 10.14185619354248, + "grad_norm": 28.780885696411133, "learning_rate": 1.2284296124194742e-05, - "loss": 0.4492, + "loss": 1.0714, "step": 36530 }, { "epoch": 3.86, - "grad_norm": 12.970759391784668, + "grad_norm": 41.30292510986328, "learning_rate": 1.22821839687401e-05, - "loss": 0.4749, + "loss": 1.1233, "step": 36540 }, { "epoch": 3.86, - "grad_norm": 13.056188583374023, + "grad_norm": 35.53240203857422, "learning_rate": 1.2280071813285459e-05, - "loss": 0.4638, + "loss": 1.155, "step": 36550 }, { "epoch": 3.86, - "grad_norm": 14.348311424255371, + "grad_norm": 25.415706634521484, "learning_rate": 1.2277959657830818e-05, - "loss": 0.5867, + "loss": 1.4329, "step": 36560 }, { "epoch": 3.86, - "grad_norm": 18.187227249145508, + "grad_norm": 17.258962631225586, "learning_rate": 1.2275847502376177e-05, - "loss": 0.3213, + "loss": 0.8225, "step": 36570 }, { "epoch": 3.86, - "grad_norm": 15.629640579223633, + "grad_norm": 41.022605895996094, "learning_rate": 1.2273735346921534e-05, - "loss": 0.5416, + "loss": 1.3959, "step": 36580 }, { "epoch": 3.86, - "grad_norm": 12.483195304870605, + "grad_norm": 46.341949462890625, "learning_rate": 1.2271623191466893e-05, - "loss": 0.6217, + "loss": 1.5286, "step": 36590 }, { "epoch": 3.87, - "grad_norm": 0.45646610856056213, + "grad_norm": 24.69019889831543, "learning_rate": 1.2269511036012251e-05, - "loss": 0.628, + "loss": 1.202, "step": 36600 }, { "epoch": 3.87, - "grad_norm": 30.91944694519043, + "grad_norm": 44.56315231323242, "learning_rate": 1.226739888055761e-05, - "loss": 0.6505, + "loss": 1.3574, "step": 36610 }, { "epoch": 3.87, - "grad_norm": 35.12753677368164, + "grad_norm": 18.85830307006836, "learning_rate": 1.2265286725102969e-05, - "loss": 0.6379, + "loss": 0.9446, "step": 36620 }, { "epoch": 3.87, - "grad_norm": 9.208316802978516, + "grad_norm": 41.93849563598633, "learning_rate": 1.2263174569648326e-05, - "loss": 0.3407, + "loss": 0.8269, "step": 36630 }, { "epoch": 3.87, - "grad_norm": 22.93569564819336, + "grad_norm": 50.081050872802734, "learning_rate": 1.2261062414193685e-05, - "loss": 0.6685, + "loss": 1.378, "step": 36640 }, { "epoch": 3.87, - "grad_norm": 7.513057708740234, + "grad_norm": 50.53166961669922, "learning_rate": 1.2258950258739043e-05, - "loss": 0.525, + "loss": 1.1399, "step": 36650 }, { "epoch": 3.87, - "grad_norm": 3.9269490242004395, + "grad_norm": 41.40780258178711, "learning_rate": 1.2256838103284402e-05, - "loss": 0.5628, + "loss": 1.6104, "step": 36660 }, { "epoch": 3.87, - "grad_norm": 0.3849836587905884, + "grad_norm": 28.82883644104004, "learning_rate": 1.2254725947829763e-05, - "loss": 0.5989, + "loss": 1.0639, "step": 36670 }, { "epoch": 3.87, - "grad_norm": 1.2249386310577393, + "grad_norm": 44.600791931152344, "learning_rate": 1.2252613792375118e-05, - "loss": 0.4479, + "loss": 1.2069, "step": 36680 }, { "epoch": 3.87, - "grad_norm": 10.963091850280762, + "grad_norm": 36.72016143798828, "learning_rate": 1.2250501636920477e-05, - "loss": 0.7169, + "loss": 1.0341, "step": 36690 }, { "epoch": 3.88, - "grad_norm": 8.647217750549316, + "grad_norm": 35.413246154785156, "learning_rate": 1.2248389481465837e-05, - "loss": 0.5606, + "loss": 1.2299, "step": 36700 }, { "epoch": 3.88, - "grad_norm": 11.774408340454102, + "grad_norm": 34.67180252075195, "learning_rate": 1.2246277326011196e-05, - "loss": 0.5541, + "loss": 1.0085, "step": 36710 }, { "epoch": 3.88, - "grad_norm": 20.219934463500977, + "grad_norm": 34.09675216674805, "learning_rate": 1.2244165170556555e-05, - "loss": 0.481, + "loss": 1.2361, "step": 36720 }, { "epoch": 3.88, - "grad_norm": 19.516202926635742, + "grad_norm": 46.75685119628906, "learning_rate": 1.2242053015101914e-05, - "loss": 0.5764, + "loss": 1.1001, "step": 36730 }, { "epoch": 3.88, - "grad_norm": 18.411054611206055, + "grad_norm": 41.12613296508789, "learning_rate": 1.223994085964727e-05, - "loss": 0.4761, + "loss": 1.2528, "step": 36740 }, { "epoch": 3.88, - "grad_norm": 6.3355913162231445, + "grad_norm": 38.99103546142578, "learning_rate": 1.223782870419263e-05, - "loss": 0.561, + "loss": 1.2157, "step": 36750 }, { "epoch": 3.88, - "grad_norm": 8.16286849975586, + "grad_norm": 22.559328079223633, "learning_rate": 1.2235716548737988e-05, - "loss": 0.6843, + "loss": 1.1884, "step": 36760 }, { "epoch": 3.88, - "grad_norm": 6.573578834533691, + "grad_norm": 26.571125030517578, "learning_rate": 1.2233604393283347e-05, - "loss": 0.7416, + "loss": 1.2411, "step": 36770 }, { "epoch": 3.88, - "grad_norm": 0.42233264446258545, + "grad_norm": 17.72969627380371, "learning_rate": 1.2231492237828706e-05, - "loss": 0.4614, + "loss": 1.0915, "step": 36780 }, { "epoch": 3.89, - "grad_norm": 14.061697006225586, + "grad_norm": 48.99596405029297, "learning_rate": 1.2229380082374063e-05, - "loss": 0.7128, + "loss": 1.2742, "step": 36790 }, { "epoch": 3.89, - "grad_norm": 10.080499649047852, + "grad_norm": 39.162960052490234, "learning_rate": 1.2227267926919422e-05, - "loss": 0.8278, + "loss": 1.4622, "step": 36800 }, { "epoch": 3.89, - "grad_norm": 12.145063400268555, + "grad_norm": 34.2200813293457, "learning_rate": 1.222515577146478e-05, - "loss": 0.3946, + "loss": 1.2736, "step": 36810 }, { "epoch": 3.89, - "grad_norm": 16.680635452270508, + "grad_norm": 46.015098571777344, "learning_rate": 1.222304361601014e-05, - "loss": 0.6834, + "loss": 1.0996, "step": 36820 }, { "epoch": 3.89, - "grad_norm": 28.694082260131836, + "grad_norm": 44.08739471435547, "learning_rate": 1.2220931460555498e-05, - "loss": 0.6427, + "loss": 1.1592, "step": 36830 }, { "epoch": 3.89, - "grad_norm": 0.1001160517334938, + "grad_norm": 22.39324378967285, "learning_rate": 1.2218819305100857e-05, - "loss": 0.4767, + "loss": 0.967, "step": 36840 }, { "epoch": 3.89, - "grad_norm": 24.208362579345703, + "grad_norm": 50.17094802856445, "learning_rate": 1.2216707149646214e-05, - "loss": 0.513, + "loss": 1.4094, "step": 36850 }, { "epoch": 3.89, - "grad_norm": 11.386198997497559, + "grad_norm": 48.000057220458984, "learning_rate": 1.2214594994191573e-05, - "loss": 0.7641, + "loss": 1.2665, "step": 36860 }, { "epoch": 3.89, - "grad_norm": 14.479930877685547, + "grad_norm": 43.98944091796875, "learning_rate": 1.2212482838736932e-05, - "loss": 0.7539, + "loss": 1.5694, "step": 36870 }, { "epoch": 3.89, - "grad_norm": 16.989944458007812, + "grad_norm": 48.61083221435547, "learning_rate": 1.221037068328229e-05, - "loss": 0.5598, + "loss": 1.2106, "step": 36880 }, { "epoch": 3.9, - "grad_norm": 14.342272758483887, + "grad_norm": 32.80451202392578, "learning_rate": 1.220825852782765e-05, - "loss": 0.6315, + "loss": 1.3378, "step": 36890 }, { "epoch": 3.9, - "grad_norm": 54.03425598144531, + "grad_norm": 45.393062591552734, "learning_rate": 1.2206146372373006e-05, - "loss": 0.6216, + "loss": 1.1646, "step": 36900 }, { "epoch": 3.9, - "grad_norm": 21.73293685913086, + "grad_norm": 49.062095642089844, "learning_rate": 1.2204034216918365e-05, - "loss": 0.6961, + "loss": 1.22, "step": 36910 }, { "epoch": 3.9, - "grad_norm": 21.410194396972656, + "grad_norm": 52.686912536621094, "learning_rate": 1.2201922061463725e-05, - "loss": 0.4673, + "loss": 1.2082, "step": 36920 }, { "epoch": 3.9, - "grad_norm": 14.876522064208984, + "grad_norm": 27.863483428955078, "learning_rate": 1.2199809906009084e-05, - "loss": 0.3378, + "loss": 1.1572, "step": 36930 }, { "epoch": 3.9, - "grad_norm": 10.359058380126953, + "grad_norm": 35.75466537475586, "learning_rate": 1.2197697750554443e-05, - "loss": 0.4109, + "loss": 1.0222, "step": 36940 }, { "epoch": 3.9, - "grad_norm": 6.472219944000244, + "grad_norm": 32.739627838134766, "learning_rate": 1.2195585595099802e-05, - "loss": 0.845, + "loss": 1.1768, "step": 36950 }, { "epoch": 3.9, - "grad_norm": 9.450103759765625, + "grad_norm": 34.081825256347656, "learning_rate": 1.2193473439645159e-05, - "loss": 0.2726, + "loss": 0.9294, "step": 36960 }, { "epoch": 3.9, - "grad_norm": 9.127897262573242, + "grad_norm": 35.285335540771484, "learning_rate": 1.2191361284190518e-05, - "loss": 0.3868, + "loss": 1.1006, "step": 36970 }, { "epoch": 3.91, - "grad_norm": 24.474018096923828, + "grad_norm": 50.88041687011719, "learning_rate": 1.2189249128735876e-05, - "loss": 0.4481, + "loss": 1.0696, "step": 36980 }, { "epoch": 3.91, - "grad_norm": 20.009984970092773, + "grad_norm": 45.197479248046875, "learning_rate": 1.2187136973281235e-05, - "loss": 0.552, + "loss": 1.0265, "step": 36990 }, { "epoch": 3.91, - "grad_norm": 11.856674194335938, + "grad_norm": 44.80147933959961, "learning_rate": 1.2185024817826594e-05, - "loss": 0.6381, + "loss": 1.1978, "step": 37000 }, { "epoch": 3.91, - "grad_norm": 11.96214485168457, + "grad_norm": 37.55168151855469, "learning_rate": 1.2182912662371951e-05, - "loss": 0.5151, + "loss": 0.7675, "step": 37010 }, { "epoch": 3.91, - "grad_norm": 14.015225410461426, + "grad_norm": 29.529739379882812, "learning_rate": 1.218080050691731e-05, - "loss": 0.5385, + "loss": 1.0263, "step": 37020 }, { "epoch": 3.91, - "grad_norm": 10.729076385498047, + "grad_norm": 48.438106536865234, "learning_rate": 1.2178688351462669e-05, - "loss": 0.3514, + "loss": 1.152, "step": 37030 }, { "epoch": 3.91, - "grad_norm": 9.551197052001953, + "grad_norm": 47.69402313232422, "learning_rate": 1.2176576196008027e-05, - "loss": 0.3714, + "loss": 1.1411, "step": 37040 }, { "epoch": 3.91, - "grad_norm": 14.848993301391602, + "grad_norm": 40.087955474853516, "learning_rate": 1.2174464040553386e-05, - "loss": 0.6588, + "loss": 1.5217, "step": 37050 }, { "epoch": 3.91, - "grad_norm": 17.153167724609375, + "grad_norm": 51.048118591308594, "learning_rate": 1.2172351885098743e-05, - "loss": 0.4235, + "loss": 0.9209, "step": 37060 }, { "epoch": 3.91, - "grad_norm": 15.403038024902344, + "grad_norm": 30.422170639038086, "learning_rate": 1.2170239729644102e-05, - "loss": 0.6401, + "loss": 1.0493, "step": 37070 }, { "epoch": 3.92, - "grad_norm": 27.97184181213379, + "grad_norm": 17.786834716796875, "learning_rate": 1.216812757418946e-05, - "loss": 0.5281, + "loss": 1.2192, "step": 37080 }, { "epoch": 3.92, - "grad_norm": 13.274991989135742, + "grad_norm": 43.772735595703125, "learning_rate": 1.216601541873482e-05, - "loss": 0.2947, + "loss": 0.9212, "step": 37090 }, { "epoch": 3.92, - "grad_norm": 12.121756553649902, + "grad_norm": 38.943965911865234, "learning_rate": 1.2163903263280178e-05, - "loss": 0.8331, + "loss": 1.4487, "step": 37100 }, { "epoch": 3.92, - "grad_norm": 10.136590003967285, + "grad_norm": 50.63135528564453, "learning_rate": 1.2161791107825537e-05, - "loss": 0.6612, + "loss": 1.3291, "step": 37110 }, { "epoch": 3.92, - "grad_norm": 8.779237747192383, + "grad_norm": 56.01612854003906, "learning_rate": 1.2159678952370894e-05, - "loss": 0.3611, + "loss": 1.0277, "step": 37120 }, { "epoch": 3.92, - "grad_norm": 5.488884925842285, + "grad_norm": 36.53557205200195, "learning_rate": 1.2157566796916253e-05, - "loss": 0.4273, + "loss": 1.1915, "step": 37130 }, { "epoch": 3.92, - "grad_norm": 9.856036186218262, + "grad_norm": 27.809429168701172, "learning_rate": 1.2155454641461612e-05, - "loss": 0.4989, + "loss": 1.0831, "step": 37140 }, { "epoch": 3.92, - "grad_norm": 9.928109169006348, + "grad_norm": 35.24599838256836, "learning_rate": 1.2153342486006972e-05, - "loss": 0.4493, + "loss": 1.4093, "step": 37150 }, { "epoch": 3.92, - "grad_norm": 13.586524963378906, + "grad_norm": 41.696292877197266, "learning_rate": 1.2151230330552331e-05, - "loss": 0.4812, + "loss": 1.1891, "step": 37160 }, { "epoch": 3.93, - "grad_norm": 5.736859321594238, + "grad_norm": 43.042572021484375, "learning_rate": 1.2149118175097686e-05, - "loss": 0.5006, + "loss": 1.3798, "step": 37170 }, { "epoch": 3.93, - "grad_norm": 16.114694595336914, + "grad_norm": 42.029884338378906, "learning_rate": 1.2147006019643047e-05, - "loss": 0.8343, + "loss": 1.4513, "step": 37180 }, { "epoch": 3.93, - "grad_norm": 16.367473602294922, + "grad_norm": 45.96116256713867, "learning_rate": 1.2144893864188406e-05, - "loss": 0.6246, + "loss": 0.9643, "step": 37190 }, { "epoch": 3.93, - "grad_norm": 16.371931076049805, + "grad_norm": 17.91334342956543, "learning_rate": 1.2142781708733764e-05, - "loss": 0.569, + "loss": 1.2376, "step": 37200 }, { "epoch": 3.93, - "grad_norm": 16.300134658813477, + "grad_norm": 36.92478561401367, "learning_rate": 1.2140669553279123e-05, - "loss": 0.4741, + "loss": 1.35, "step": 37210 }, { "epoch": 3.93, - "grad_norm": 15.024394035339355, + "grad_norm": 48.408424377441406, "learning_rate": 1.2138557397824482e-05, - "loss": 0.6708, + "loss": 1.4893, "step": 37220 }, { "epoch": 3.93, - "grad_norm": 0.9203153848648071, + "grad_norm": 22.17268943786621, "learning_rate": 1.2136445242369839e-05, - "loss": 0.4585, + "loss": 0.904, "step": 37230 }, { "epoch": 3.93, - "grad_norm": 14.672334671020508, + "grad_norm": 39.846038818359375, "learning_rate": 1.2134333086915198e-05, - "loss": 0.4084, + "loss": 0.8302, "step": 37240 }, { "epoch": 3.93, - "grad_norm": 24.848175048828125, + "grad_norm": 37.489837646484375, "learning_rate": 1.2132220931460557e-05, - "loss": 0.4352, + "loss": 1.0483, "step": 37250 }, { "epoch": 3.93, - "grad_norm": 9.56224250793457, + "grad_norm": 41.717647552490234, "learning_rate": 1.2130108776005915e-05, - "loss": 0.4065, + "loss": 1.0416, "step": 37260 }, { "epoch": 3.94, - "grad_norm": 19.505859375, + "grad_norm": 46.459415435791016, "learning_rate": 1.2127996620551274e-05, - "loss": 0.825, + "loss": 1.8253, "step": 37270 }, { "epoch": 3.94, - "grad_norm": 17.494518280029297, + "grad_norm": 33.63521194458008, "learning_rate": 1.2125884465096631e-05, - "loss": 0.5716, + "loss": 0.8588, "step": 37280 }, { "epoch": 3.94, - "grad_norm": 10.472118377685547, + "grad_norm": 27.944656372070312, "learning_rate": 1.212377230964199e-05, - "loss": 0.3058, + "loss": 0.9887, "step": 37290 }, { "epoch": 3.94, - "grad_norm": 0.18710727989673615, + "grad_norm": 36.154380798339844, "learning_rate": 1.2121660154187349e-05, - "loss": 0.6711, + "loss": 0.9736, "step": 37300 }, { "epoch": 3.94, - "grad_norm": 14.618348121643066, + "grad_norm": 38.372127532958984, "learning_rate": 1.2119547998732708e-05, - "loss": 0.3902, + "loss": 1.2309, "step": 37310 }, { "epoch": 3.94, - "grad_norm": 10.199435234069824, + "grad_norm": 23.326946258544922, "learning_rate": 1.2117435843278066e-05, - "loss": 0.5569, + "loss": 1.4798, "step": 37320 }, { "epoch": 3.94, - "grad_norm": 6.815585613250732, + "grad_norm": 44.36894607543945, "learning_rate": 1.2115323687823424e-05, - "loss": 0.1915, + "loss": 1.2569, "step": 37330 }, { "epoch": 3.94, - "grad_norm": 17.309499740600586, + "grad_norm": 31.56819725036621, "learning_rate": 1.2113211532368782e-05, - "loss": 0.3698, + "loss": 0.997, "step": 37340 }, { "epoch": 3.94, - "grad_norm": 8.76121997833252, + "grad_norm": 35.2846794128418, "learning_rate": 1.2111099376914141e-05, - "loss": 0.579, + "loss": 1.346, "step": 37350 }, { "epoch": 3.95, - "grad_norm": 5.282585144042969, + "grad_norm": 31.973644256591797, "learning_rate": 1.21089872214595e-05, - "loss": 0.3304, + "loss": 0.8373, "step": 37360 }, { "epoch": 3.95, - "grad_norm": 12.851136207580566, + "grad_norm": 35.87435531616211, "learning_rate": 1.210687506600486e-05, - "loss": 0.6968, + "loss": 1.0826, "step": 37370 }, { "epoch": 3.95, - "grad_norm": 0.12977094948291779, + "grad_norm": 32.86164855957031, "learning_rate": 1.2104762910550219e-05, - "loss": 0.6467, + "loss": 1.4686, "step": 37380 }, { "epoch": 3.95, - "grad_norm": 13.381671905517578, + "grad_norm": 38.39122009277344, "learning_rate": 1.2102650755095574e-05, - "loss": 0.5906, + "loss": 1.2382, "step": 37390 }, { "epoch": 3.95, - "grad_norm": 39.327335357666016, + "grad_norm": 39.01300811767578, "learning_rate": 1.2100538599640935e-05, - "loss": 0.5348, + "loss": 1.2553, "step": 37400 }, { "epoch": 3.95, - "grad_norm": 24.804922103881836, + "grad_norm": 53.26844787597656, "learning_rate": 1.2098426444186294e-05, - "loss": 0.8136, + "loss": 1.3033, "step": 37410 }, { "epoch": 3.95, - "grad_norm": 14.67399787902832, + "grad_norm": 43.718563079833984, "learning_rate": 1.2096314288731653e-05, - "loss": 0.665, + "loss": 1.2374, "step": 37420 }, { "epoch": 3.95, - "grad_norm": 9.10966968536377, + "grad_norm": 30.518470764160156, "learning_rate": 1.2094202133277011e-05, - "loss": 0.4882, + "loss": 0.9742, "step": 37430 }, { "epoch": 3.95, - "grad_norm": 14.52433967590332, + "grad_norm": 45.54966354370117, "learning_rate": 1.2092089977822368e-05, - "loss": 0.5484, + "loss": 1.3286, "step": 37440 }, { "epoch": 3.96, - "grad_norm": 13.9821138381958, + "grad_norm": 35.97121047973633, "learning_rate": 1.2089977822367727e-05, - "loss": 0.3258, + "loss": 1.0322, "step": 37450 }, { "epoch": 3.96, - "grad_norm": 27.79945945739746, + "grad_norm": 49.52111053466797, "learning_rate": 1.2087865666913086e-05, - "loss": 0.4588, + "loss": 1.0607, "step": 37460 }, { "epoch": 3.96, - "grad_norm": 11.942042350769043, + "grad_norm": 46.61147689819336, "learning_rate": 1.2085753511458445e-05, - "loss": 0.6488, + "loss": 1.0698, "step": 37470 }, { "epoch": 3.96, - "grad_norm": 12.6549711227417, + "grad_norm": 44.62751770019531, "learning_rate": 1.2083641356003804e-05, - "loss": 0.8894, + "loss": 1.3526, "step": 37480 }, { "epoch": 3.96, - "grad_norm": 21.12003517150879, + "grad_norm": 28.10324478149414, "learning_rate": 1.2081529200549162e-05, - "loss": 0.552, + "loss": 1.101, "step": 37490 }, { "epoch": 3.96, - "grad_norm": 0.9542860388755798, + "grad_norm": 43.864906311035156, "learning_rate": 1.207941704509452e-05, - "loss": 0.3943, + "loss": 1.1432, "step": 37500 }, { "epoch": 3.96, - "grad_norm": 27.85733985900879, + "grad_norm": 46.93824005126953, "learning_rate": 1.2077304889639878e-05, - "loss": 0.4314, + "loss": 1.0821, "step": 37510 }, { "epoch": 3.96, - "grad_norm": 9.392875671386719, + "grad_norm": 54.22712326049805, "learning_rate": 1.2075192734185237e-05, - "loss": 0.4441, + "loss": 1.1834, "step": 37520 }, { "epoch": 3.96, - "grad_norm": 11.39575481414795, + "grad_norm": 36.003482818603516, "learning_rate": 1.2073080578730596e-05, - "loss": 0.6265, + "loss": 1.3972, "step": 37530 }, { "epoch": 3.96, - "grad_norm": 19.90989112854004, + "grad_norm": 38.82493209838867, "learning_rate": 1.2070968423275954e-05, - "loss": 0.3867, + "loss": 0.9988, "step": 37540 }, { "epoch": 3.97, - "grad_norm": 7.180244445800781, + "grad_norm": 23.823863983154297, "learning_rate": 1.2068856267821312e-05, - "loss": 0.4743, + "loss": 1.2427, "step": 37550 }, { "epoch": 3.97, - "grad_norm": 0.4664573669433594, + "grad_norm": 27.175464630126953, "learning_rate": 1.206674411236667e-05, - "loss": 0.5517, + "loss": 0.8475, "step": 37560 }, { "epoch": 3.97, - "grad_norm": 10.005620956420898, + "grad_norm": 32.72175598144531, "learning_rate": 1.2064631956912029e-05, - "loss": 0.8451, + "loss": 1.2432, "step": 37570 }, { "epoch": 3.97, - "grad_norm": 14.889140129089355, + "grad_norm": 29.681827545166016, "learning_rate": 1.2062519801457388e-05, - "loss": 0.7142, + "loss": 1.1433, "step": 37580 }, { "epoch": 3.97, - "grad_norm": 1.541751742362976, + "grad_norm": 49.3849983215332, "learning_rate": 1.2060407646002747e-05, - "loss": 0.5449, + "loss": 1.2479, "step": 37590 }, { "epoch": 3.97, - "grad_norm": 11.046995162963867, + "grad_norm": 43.7578010559082, "learning_rate": 1.2058295490548104e-05, - "loss": 0.4692, + "loss": 1.053, "step": 37600 }, { "epoch": 3.97, - "grad_norm": 9.561458587646484, + "grad_norm": 42.983184814453125, "learning_rate": 1.2056183335093463e-05, - "loss": 0.395, + "loss": 0.9257, "step": 37610 }, { "epoch": 3.97, - "grad_norm": 19.002710342407227, + "grad_norm": 60.06571578979492, "learning_rate": 1.2054071179638821e-05, - "loss": 0.6169, + "loss": 1.0672, "step": 37620 }, { "epoch": 3.97, - "grad_norm": 0.4325624108314514, + "grad_norm": 31.222766876220703, "learning_rate": 1.2051959024184182e-05, - "loss": 0.9579, + "loss": 1.4439, "step": 37630 }, { "epoch": 3.98, - "grad_norm": 8.356069564819336, + "grad_norm": 37.3191032409668, "learning_rate": 1.204984686872954e-05, - "loss": 0.5658, + "loss": 1.043, "step": 37640 }, { "epoch": 3.98, - "grad_norm": 10.222809791564941, + "grad_norm": 50.27363967895508, "learning_rate": 1.20477347132749e-05, - "loss": 0.6939, + "loss": 1.2107, "step": 37650 }, { "epoch": 3.98, - "grad_norm": 15.852988243103027, + "grad_norm": 47.16386032104492, "learning_rate": 1.2045622557820256e-05, - "loss": 0.4297, + "loss": 1.2087, "step": 37660 }, { "epoch": 3.98, - "grad_norm": 16.53685760498047, + "grad_norm": 46.814430236816406, "learning_rate": 1.2043510402365615e-05, - "loss": 0.917, + "loss": 1.6226, "step": 37670 }, { "epoch": 3.98, - "grad_norm": 7.956087112426758, + "grad_norm": 37.556339263916016, "learning_rate": 1.2041398246910974e-05, - "loss": 0.393, + "loss": 1.064, "step": 37680 }, { "epoch": 3.98, - "grad_norm": 20.963626861572266, + "grad_norm": 60.29244613647461, "learning_rate": 1.2039286091456333e-05, - "loss": 0.6281, + "loss": 1.3393, "step": 37690 }, { "epoch": 3.98, - "grad_norm": 20.184110641479492, + "grad_norm": 37.6685905456543, "learning_rate": 1.2037173936001692e-05, - "loss": 0.5441, + "loss": 1.3463, "step": 37700 }, { "epoch": 3.98, - "grad_norm": 8.726158142089844, + "grad_norm": 44.92082214355469, "learning_rate": 1.2035061780547049e-05, - "loss": 0.4988, + "loss": 1.1559, "step": 37710 }, { "epoch": 3.98, - "grad_norm": 13.140741348266602, + "grad_norm": 39.05729675292969, "learning_rate": 1.2032949625092407e-05, - "loss": 0.4543, + "loss": 1.2248, "step": 37720 }, { "epoch": 3.98, - "grad_norm": 4.478666305541992, + "grad_norm": 36.6021728515625, "learning_rate": 1.2030837469637766e-05, - "loss": 0.4957, + "loss": 1.0904, "step": 37730 }, { "epoch": 3.99, - "grad_norm": 7.4149556159973145, + "grad_norm": 17.716514587402344, "learning_rate": 1.2028725314183125e-05, - "loss": 0.4794, + "loss": 0.8725, "step": 37740 }, { "epoch": 3.99, - "grad_norm": 4.327811241149902, + "grad_norm": 41.723182678222656, "learning_rate": 1.2026613158728484e-05, - "loss": 0.63, + "loss": 1.1737, "step": 37750 }, { "epoch": 3.99, - "grad_norm": 11.568052291870117, + "grad_norm": 47.591285705566406, "learning_rate": 1.2024501003273843e-05, - "loss": 0.5779, + "loss": 1.322, "step": 37760 }, { "epoch": 3.99, - "grad_norm": 10.830099105834961, + "grad_norm": 25.026521682739258, "learning_rate": 1.20223888478192e-05, - "loss": 0.4834, + "loss": 1.1505, "step": 37770 }, { "epoch": 3.99, - "grad_norm": 6.213988780975342, + "grad_norm": 25.041515350341797, "learning_rate": 1.2020276692364558e-05, - "loss": 0.7472, + "loss": 1.0556, "step": 37780 }, { "epoch": 3.99, - "grad_norm": 17.107736587524414, + "grad_norm": 34.682437896728516, "learning_rate": 1.2018164536909917e-05, - "loss": 0.4527, + "loss": 1.082, "step": 37790 }, { "epoch": 3.99, - "grad_norm": 3.2169110774993896, + "grad_norm": 33.81075668334961, "learning_rate": 1.2016052381455276e-05, - "loss": 0.5441, + "loss": 1.0013, "step": 37800 }, { "epoch": 3.99, - "grad_norm": 0.8894926309585571, + "grad_norm": 29.20528221130371, "learning_rate": 1.2013940226000635e-05, - "loss": 0.2551, + "loss": 0.9641, "step": 37810 }, { "epoch": 3.99, - "grad_norm": 1.025410771369934, + "grad_norm": 53.497337341308594, "learning_rate": 1.2011828070545992e-05, - "loss": 0.2869, + "loss": 1.0559, "step": 37820 }, { "epoch": 4.0, - "grad_norm": 16.94969367980957, + "grad_norm": 50.99386978149414, "learning_rate": 1.200971591509135e-05, - "loss": 0.4352, + "loss": 1.0154, "step": 37830 }, { "epoch": 4.0, - "grad_norm": 8.652446746826172, + "grad_norm": 30.435405731201172, "learning_rate": 1.200760375963671e-05, - "loss": 0.3571, + "loss": 1.0603, "step": 37840 }, { "epoch": 4.0, - "grad_norm": 0.7458866238594055, + "grad_norm": 26.94050407409668, "learning_rate": 1.200549160418207e-05, - "loss": 0.5203, + "loss": 1.0533, "step": 37850 }, { "epoch": 4.0, - "grad_norm": 5.422346591949463, + "grad_norm": 21.422504425048828, "learning_rate": 1.2003379448727429e-05, - "loss": 0.5025, + "loss": 1.393, "step": 37860 }, { "epoch": 4.0, - "grad_norm": 29.6828670501709, + "grad_norm": 33.57707595825195, "learning_rate": 1.2001267293272787e-05, - "loss": 1.0346, + "loss": 1.4347, "step": 37870 }, { "epoch": 4.0, - "eval_accuracy": 0.8968316831683169, - "eval_loss": 0.39108872413635254, - "eval_runtime": 181.2711, - "eval_samples_per_second": 139.294, - "eval_steps_per_second": 17.416, + "eval_accuracy": 0.7738613861386139, + "eval_loss": 0.8343495726585388, + "eval_runtime": 36.9068, + "eval_samples_per_second": 684.156, + "eval_steps_per_second": 85.54, "step": 37876 }, { "epoch": 4.0, - "grad_norm": 5.816527366638184, + "grad_norm": 47.499122619628906, "learning_rate": 1.1999155137818144e-05, - "loss": 0.3149, + "loss": 0.8722, "step": 37880 }, { "epoch": 4.0, - "grad_norm": 34.33921432495117, + "grad_norm": 35.908287048339844, "learning_rate": 1.1997042982363503e-05, - "loss": 0.2743, + "loss": 1.3419, "step": 37890 }, { "epoch": 4.0, - "grad_norm": 4.6737895011901855, + "grad_norm": 19.142780303955078, "learning_rate": 1.1994930826908862e-05, - "loss": 0.6023, + "loss": 1.0422, "step": 37900 }, { "epoch": 4.0, - "grad_norm": 22.080183029174805, + "grad_norm": 19.17131233215332, "learning_rate": 1.199281867145422e-05, - "loss": 0.2813, + "loss": 1.0134, "step": 37910 }, { "epoch": 4.0, - "grad_norm": 2.1632699966430664, + "grad_norm": 35.00048065185547, "learning_rate": 1.199070651599958e-05, - "loss": 0.4917, + "loss": 1.4086, "step": 37920 }, { "epoch": 4.01, - "grad_norm": 7.605011940002441, + "grad_norm": 45.741390228271484, "learning_rate": 1.1988594360544937e-05, - "loss": 0.4179, + "loss": 1.307, "step": 37930 }, { "epoch": 4.01, - "grad_norm": 10.497692108154297, + "grad_norm": 44.17996597290039, "learning_rate": 1.1986482205090295e-05, - "loss": 0.2051, + "loss": 1.0603, "step": 37940 }, { "epoch": 4.01, - "grad_norm": 11.333343505859375, + "grad_norm": 25.948585510253906, "learning_rate": 1.1984370049635654e-05, - "loss": 0.5822, + "loss": 0.9739, "step": 37950 }, { "epoch": 4.01, - "grad_norm": 8.700035095214844, + "grad_norm": 40.368526458740234, "learning_rate": 1.1982257894181013e-05, - "loss": 0.3963, + "loss": 1.3409, "step": 37960 }, { "epoch": 4.01, - "grad_norm": 10.197785377502441, + "grad_norm": 52.85811233520508, "learning_rate": 1.1980145738726372e-05, - "loss": 0.6436, + "loss": 1.0852, "step": 37970 }, { "epoch": 4.01, - "grad_norm": 13.817719459533691, + "grad_norm": 28.613014221191406, "learning_rate": 1.1978033583271729e-05, - "loss": 0.3383, + "loss": 0.8334, "step": 37980 }, { "epoch": 4.01, - "grad_norm": 19.282684326171875, + "grad_norm": 46.94611358642578, "learning_rate": 1.1975921427817088e-05, - "loss": 0.476, + "loss": 1.23, "step": 37990 }, { "epoch": 4.01, - "grad_norm": 38.22368621826172, + "grad_norm": 36.44575881958008, "learning_rate": 1.1973809272362446e-05, - "loss": 0.4251, + "loss": 1.2905, "step": 38000 }, { "epoch": 4.01, - "grad_norm": 9.32701301574707, + "grad_norm": 31.055086135864258, "learning_rate": 1.1971697116907805e-05, - "loss": 0.4492, + "loss": 0.9049, "step": 38010 }, { "epoch": 4.02, - "grad_norm": 8.522049903869629, + "grad_norm": 42.30833053588867, "learning_rate": 1.1969584961453164e-05, - "loss": 0.3485, + "loss": 0.9363, "step": 38020 }, { "epoch": 4.02, - "grad_norm": 7.8781328201293945, + "grad_norm": 32.431884765625, "learning_rate": 1.1967472805998523e-05, - "loss": 0.6705, + "loss": 1.2408, "step": 38030 }, { "epoch": 4.02, - "grad_norm": 16.37264633178711, + "grad_norm": 37.16569900512695, "learning_rate": 1.196536065054388e-05, - "loss": 0.8648, + "loss": 1.1472, "step": 38040 }, { "epoch": 4.02, - "grad_norm": 0.6706315875053406, + "grad_norm": 32.19245147705078, "learning_rate": 1.1963248495089239e-05, - "loss": 0.4442, + "loss": 1.2816, "step": 38050 }, { "epoch": 4.02, - "grad_norm": 17.211435317993164, + "grad_norm": 46.33047103881836, "learning_rate": 1.1961136339634597e-05, - "loss": 0.6942, + "loss": 1.1505, "step": 38060 }, { "epoch": 4.02, - "grad_norm": 20.885934829711914, + "grad_norm": 53.347923278808594, "learning_rate": 1.1959024184179956e-05, - "loss": 0.5597, + "loss": 0.9903, "step": 38070 }, { "epoch": 4.02, - "grad_norm": 7.998570442199707, + "grad_norm": 34.77487564086914, "learning_rate": 1.1956912028725317e-05, - "loss": 0.6953, + "loss": 1.024, "step": 38080 }, { "epoch": 4.02, - "grad_norm": 11.891342163085938, + "grad_norm": 34.716365814208984, "learning_rate": 1.1954799873270672e-05, - "loss": 0.5441, + "loss": 1.1181, "step": 38090 }, { "epoch": 4.02, - "grad_norm": 8.9273099899292, + "grad_norm": 55.50782775878906, "learning_rate": 1.1952687717816031e-05, - "loss": 0.4955, + "loss": 1.1133, "step": 38100 }, { "epoch": 4.02, - "grad_norm": 31.14972686767578, + "grad_norm": 46.03562927246094, "learning_rate": 1.1950575562361391e-05, - "loss": 0.4278, + "loss": 1.014, "step": 38110 }, { "epoch": 4.03, - "grad_norm": 26.042465209960938, + "grad_norm": 39.32997512817383, "learning_rate": 1.194846340690675e-05, - "loss": 0.3773, + "loss": 0.9992, "step": 38120 }, { "epoch": 4.03, - "grad_norm": 8.782833099365234, + "grad_norm": 41.28805923461914, "learning_rate": 1.1946351251452109e-05, - "loss": 0.5512, + "loss": 1.1927, "step": 38130 }, { "epoch": 4.03, - "grad_norm": 8.338234901428223, + "grad_norm": 30.739065170288086, "learning_rate": 1.1944239095997468e-05, - "loss": 0.3037, + "loss": 0.9979, "step": 38140 }, { "epoch": 4.03, - "grad_norm": 2.366588592529297, + "grad_norm": 31.378690719604492, "learning_rate": 1.1942126940542825e-05, - "loss": 0.1863, + "loss": 1.121, "step": 38150 }, { "epoch": 4.03, - "grad_norm": 1.208191990852356, + "grad_norm": 57.154571533203125, "learning_rate": 1.1940014785088184e-05, - "loss": 0.5699, + "loss": 1.4041, "step": 38160 }, { "epoch": 4.03, - "grad_norm": 3.4297080039978027, + "grad_norm": 40.2799072265625, "learning_rate": 1.1937902629633542e-05, - "loss": 0.2604, + "loss": 1.2173, "step": 38170 }, { "epoch": 4.03, - "grad_norm": 5.970609664916992, + "grad_norm": 37.9283332824707, "learning_rate": 1.1935790474178901e-05, - "loss": 0.5973, + "loss": 1.0874, "step": 38180 }, { "epoch": 4.03, - "grad_norm": 6.883532524108887, + "grad_norm": 41.05622863769531, "learning_rate": 1.193367831872426e-05, - "loss": 0.3401, + "loss": 1.168, "step": 38190 }, { "epoch": 4.03, - "grad_norm": 8.614724159240723, + "grad_norm": 49.09443664550781, "learning_rate": 1.1931566163269617e-05, - "loss": 0.4199, + "loss": 1.277, "step": 38200 }, { "epoch": 4.04, - "grad_norm": 26.091882705688477, + "grad_norm": 32.346866607666016, "learning_rate": 1.1929454007814976e-05, - "loss": 0.6497, + "loss": 1.1077, "step": 38210 }, { "epoch": 4.04, - "grad_norm": 16.28439712524414, + "grad_norm": 30.903728485107422, "learning_rate": 1.1927341852360335e-05, - "loss": 0.6321, + "loss": 1.145, "step": 38220 }, { "epoch": 4.04, - "grad_norm": 7.051592826843262, + "grad_norm": 41.03510284423828, "learning_rate": 1.1925229696905693e-05, - "loss": 0.804, + "loss": 1.2972, "step": 38230 }, { "epoch": 4.04, - "grad_norm": 6.060262680053711, + "grad_norm": 42.22907257080078, "learning_rate": 1.1923117541451052e-05, - "loss": 0.3369, + "loss": 0.9817, "step": 38240 }, { "epoch": 4.04, - "grad_norm": 3.2858359813690186, + "grad_norm": 37.592220306396484, "learning_rate": 1.1921005385996409e-05, - "loss": 0.1885, + "loss": 0.8343, "step": 38250 }, { "epoch": 4.04, - "grad_norm": 0.23774999380111694, + "grad_norm": 26.679861068725586, "learning_rate": 1.1918893230541768e-05, - "loss": 0.2969, + "loss": 1.1806, "step": 38260 }, { "epoch": 4.04, - "grad_norm": 1.3468060493469238, + "grad_norm": 53.87498092651367, "learning_rate": 1.1916781075087127e-05, - "loss": 0.8937, + "loss": 1.37, "step": 38270 }, { "epoch": 4.04, - "grad_norm": 6.0810441970825195, + "grad_norm": 23.248279571533203, "learning_rate": 1.1914668919632485e-05, - "loss": 0.4684, + "loss": 0.9544, "step": 38280 }, { "epoch": 4.04, - "grad_norm": 22.580768585205078, + "grad_norm": 37.82069396972656, "learning_rate": 1.1912556764177844e-05, - "loss": 0.3293, + "loss": 1.1551, "step": 38290 }, { "epoch": 4.04, - "grad_norm": 4.6596174240112305, + "grad_norm": 35.652191162109375, "learning_rate": 1.1910444608723205e-05, - "loss": 0.3471, + "loss": 0.8275, "step": 38300 }, { "epoch": 4.05, - "grad_norm": 19.589885711669922, + "grad_norm": 32.95951843261719, "learning_rate": 1.190833245326856e-05, - "loss": 0.3666, + "loss": 1.1271, "step": 38310 }, { "epoch": 4.05, - "grad_norm": 12.701177597045898, + "grad_norm": 27.465139389038086, "learning_rate": 1.1906220297813919e-05, - "loss": 0.4757, + "loss": 1.1297, "step": 38320 }, { "epoch": 4.05, - "grad_norm": 32.592716217041016, + "grad_norm": 50.9995002746582, "learning_rate": 1.190410814235928e-05, - "loss": 0.6064, + "loss": 1.2706, "step": 38330 }, { "epoch": 4.05, - "grad_norm": 17.55474090576172, + "grad_norm": 42.65768814086914, "learning_rate": 1.1901995986904638e-05, - "loss": 0.3063, + "loss": 1.0475, "step": 38340 }, { "epoch": 4.05, - "grad_norm": 11.841391563415527, + "grad_norm": 39.85985565185547, "learning_rate": 1.1899883831449997e-05, - "loss": 0.5504, + "loss": 1.1114, "step": 38350 }, { "epoch": 4.05, - "grad_norm": 15.709468841552734, + "grad_norm": 34.338993072509766, "learning_rate": 1.1897771675995354e-05, - "loss": 0.3111, + "loss": 0.9892, "step": 38360 }, { "epoch": 4.05, - "grad_norm": 7.296017169952393, + "grad_norm": 28.45636749267578, "learning_rate": 1.1895659520540713e-05, - "loss": 0.4407, + "loss": 0.9361, "step": 38370 }, { "epoch": 4.05, - "grad_norm": 8.536079406738281, + "grad_norm": 26.29062843322754, "learning_rate": 1.1893547365086072e-05, - "loss": 0.5815, + "loss": 1.3147, "step": 38380 }, { "epoch": 4.05, - "grad_norm": 6.552495956420898, + "grad_norm": 41.934234619140625, "learning_rate": 1.189143520963143e-05, - "loss": 0.3314, + "loss": 1.0131, "step": 38390 }, { "epoch": 4.06, - "grad_norm": 6.223598003387451, + "grad_norm": 34.78412628173828, "learning_rate": 1.1889323054176789e-05, - "loss": 0.3208, + "loss": 1.1002, "step": 38400 }, { "epoch": 4.06, - "grad_norm": 2.5834107398986816, + "grad_norm": 40.445491790771484, "learning_rate": 1.1887210898722148e-05, - "loss": 0.5065, + "loss": 1.296, "step": 38410 }, { "epoch": 4.06, - "grad_norm": 46.206764221191406, + "grad_norm": 51.944664001464844, "learning_rate": 1.1885098743267505e-05, - "loss": 0.4384, + "loss": 1.425, "step": 38420 }, { "epoch": 4.06, - "grad_norm": 26.312646865844727, + "grad_norm": 53.99348449707031, "learning_rate": 1.1882986587812864e-05, - "loss": 0.7838, + "loss": 1.3793, "step": 38430 }, { "epoch": 4.06, - "grad_norm": 13.429438591003418, + "grad_norm": 28.74734878540039, "learning_rate": 1.1880874432358223e-05, - "loss": 0.3519, + "loss": 0.842, "step": 38440 }, { "epoch": 4.06, - "grad_norm": 13.450687408447266, + "grad_norm": 43.64649200439453, "learning_rate": 1.1878762276903581e-05, - "loss": 0.5775, + "loss": 0.9842, "step": 38450 }, { "epoch": 4.06, - "grad_norm": 3.7109968662261963, + "grad_norm": 25.019689559936523, "learning_rate": 1.187665012144894e-05, - "loss": 0.6791, + "loss": 1.1587, "step": 38460 }, { "epoch": 4.06, - "grad_norm": 10.102091789245605, + "grad_norm": 30.21944236755371, "learning_rate": 1.1874537965994297e-05, - "loss": 0.5627, + "loss": 1.0416, "step": 38470 }, { "epoch": 4.06, - "grad_norm": 17.438438415527344, + "grad_norm": 64.97341918945312, "learning_rate": 1.1872425810539656e-05, - "loss": 0.7412, + "loss": 1.2418, "step": 38480 }, { "epoch": 4.06, - "grad_norm": 3.5207371711730957, + "grad_norm": 46.51776885986328, "learning_rate": 1.1870313655085015e-05, - "loss": 0.4467, + "loss": 0.808, "step": 38490 }, { "epoch": 4.07, - "grad_norm": 5.383874416351318, + "grad_norm": 60.38923263549805, "learning_rate": 1.1868201499630374e-05, - "loss": 0.3545, + "loss": 1.1525, "step": 38500 }, { "epoch": 4.07, - "grad_norm": 32.81305694580078, + "grad_norm": 54.37565612792969, "learning_rate": 1.1866089344175732e-05, - "loss": 0.6373, + "loss": 1.4236, "step": 38510 }, { "epoch": 4.07, - "grad_norm": 13.3355131149292, + "grad_norm": 29.042198181152344, "learning_rate": 1.1863977188721093e-05, - "loss": 0.4303, + "loss": 0.9646, "step": 38520 }, { "epoch": 4.07, - "grad_norm": 4.763644218444824, + "grad_norm": 43.35923385620117, "learning_rate": 1.1861865033266448e-05, - "loss": 0.5481, + "loss": 1.3016, "step": 38530 }, { "epoch": 4.07, - "grad_norm": 8.119165420532227, + "grad_norm": 61.005435943603516, "learning_rate": 1.1859752877811807e-05, - "loss": 0.5759, + "loss": 0.86, "step": 38540 }, { "epoch": 4.07, - "grad_norm": 0.8085461854934692, + "grad_norm": 27.77308464050293, "learning_rate": 1.1857640722357166e-05, - "loss": 0.3152, + "loss": 0.8458, "step": 38550 }, { "epoch": 4.07, - "grad_norm": 0.4553145170211792, + "grad_norm": 18.51683235168457, "learning_rate": 1.1855528566902526e-05, - "loss": 0.3722, + "loss": 0.7111, "step": 38560 }, { "epoch": 4.07, - "grad_norm": 9.82752799987793, + "grad_norm": 30.131912231445312, "learning_rate": 1.1853416411447885e-05, - "loss": 0.6111, + "loss": 1.1229, "step": 38570 }, { "epoch": 4.07, - "grad_norm": 11.134026527404785, + "grad_norm": 43.9988899230957, "learning_rate": 1.185130425599324e-05, - "loss": 0.5541, + "loss": 1.0362, "step": 38580 }, { "epoch": 4.08, - "grad_norm": 3.0306553840637207, + "grad_norm": 25.825197219848633, "learning_rate": 1.1849192100538601e-05, - "loss": 0.3672, + "loss": 1.2983, "step": 38590 }, { "epoch": 4.08, - "grad_norm": 31.447599411010742, + "grad_norm": 53.3684196472168, "learning_rate": 1.184707994508396e-05, - "loss": 0.3401, + "loss": 0.7696, "step": 38600 }, { "epoch": 4.08, - "grad_norm": 35.40774917602539, + "grad_norm": 34.48853302001953, "learning_rate": 1.1844967789629318e-05, - "loss": 0.49, + "loss": 1.3729, "step": 38610 }, { "epoch": 4.08, - "grad_norm": 0.44370684027671814, + "grad_norm": 38.53873062133789, "learning_rate": 1.1842855634174677e-05, - "loss": 0.3781, + "loss": 0.9555, "step": 38620 }, { "epoch": 4.08, - "grad_norm": 8.573012351989746, + "grad_norm": 37.42805480957031, "learning_rate": 1.1840743478720034e-05, - "loss": 0.4404, + "loss": 1.1447, "step": 38630 }, { "epoch": 4.08, - "grad_norm": 30.279348373413086, + "grad_norm": 51.003665924072266, "learning_rate": 1.1838631323265393e-05, - "loss": 0.4871, + "loss": 0.9473, "step": 38640 }, { "epoch": 4.08, - "grad_norm": 5.6908793449401855, + "grad_norm": 36.25321578979492, "learning_rate": 1.1836519167810752e-05, - "loss": 0.5721, + "loss": 1.1402, "step": 38650 }, { "epoch": 4.08, - "grad_norm": 20.134912490844727, + "grad_norm": 34.28619384765625, "learning_rate": 1.183440701235611e-05, - "loss": 0.4473, + "loss": 0.9033, "step": 38660 }, { "epoch": 4.08, - "grad_norm": 2.4827351570129395, + "grad_norm": 40.334747314453125, "learning_rate": 1.183229485690147e-05, - "loss": 0.3695, + "loss": 1.0286, "step": 38670 }, { "epoch": 4.08, - "grad_norm": 15.024284362792969, + "grad_norm": 25.51503562927246, "learning_rate": 1.1830182701446828e-05, - "loss": 0.5707, + "loss": 1.3256, "step": 38680 }, { "epoch": 4.09, - "grad_norm": 6.943925857543945, + "grad_norm": 56.75960159301758, "learning_rate": 1.1828070545992185e-05, - "loss": 0.4062, + "loss": 1.2339, "step": 38690 }, { "epoch": 4.09, - "grad_norm": 6.523637294769287, + "grad_norm": 40.14506149291992, "learning_rate": 1.1825958390537544e-05, - "loss": 0.4841, + "loss": 1.0618, "step": 38700 }, { "epoch": 4.09, - "grad_norm": 11.793119430541992, + "grad_norm": 18.55246353149414, "learning_rate": 1.1823846235082903e-05, - "loss": 0.3104, + "loss": 1.0546, "step": 38710 }, { "epoch": 4.09, - "grad_norm": 21.92927360534668, + "grad_norm": 36.043392181396484, "learning_rate": 1.1821734079628262e-05, - "loss": 0.399, + "loss": 1.1845, "step": 38720 }, { "epoch": 4.09, - "grad_norm": 34.597408294677734, + "grad_norm": 47.9533576965332, "learning_rate": 1.181962192417362e-05, - "loss": 0.5039, + "loss": 1.1009, "step": 38730 }, { "epoch": 4.09, - "grad_norm": 6.298150539398193, + "grad_norm": 41.80986785888672, "learning_rate": 1.1817509768718977e-05, - "loss": 0.376, + "loss": 1.0335, "step": 38740 }, { "epoch": 4.09, - "grad_norm": 14.596674919128418, + "grad_norm": 39.94231033325195, "learning_rate": 1.1815397613264336e-05, - "loss": 0.4873, + "loss": 0.9661, "step": 38750 }, { "epoch": 4.09, - "grad_norm": 17.259328842163086, + "grad_norm": 37.06708908081055, "learning_rate": 1.1813285457809695e-05, - "loss": 0.4151, + "loss": 1.1554, "step": 38760 }, { "epoch": 4.09, - "grad_norm": 12.083000183105469, + "grad_norm": 22.18853759765625, "learning_rate": 1.1811173302355054e-05, - "loss": 0.3602, + "loss": 0.7494, "step": 38770 }, { "epoch": 4.1, - "grad_norm": 21.743793487548828, + "grad_norm": 31.190542221069336, "learning_rate": 1.1809061146900414e-05, - "loss": 0.4086, + "loss": 1.0463, "step": 38780 }, { "epoch": 4.1, - "grad_norm": 13.404767036437988, + "grad_norm": 27.04295539855957, "learning_rate": 1.1806948991445773e-05, - "loss": 0.2187, + "loss": 1.1629, "step": 38790 }, { "epoch": 4.1, - "grad_norm": 3.7288246154785156, + "grad_norm": 39.483150482177734, "learning_rate": 1.1804836835991128e-05, - "loss": 0.6736, + "loss": 1.1682, "step": 38800 }, { "epoch": 4.1, - "grad_norm": 4.787732124328613, + "grad_norm": 27.082056045532227, "learning_rate": 1.1802724680536489e-05, - "loss": 0.655, + "loss": 1.494, "step": 38810 }, { "epoch": 4.1, - "grad_norm": 0.14330285787582397, + "grad_norm": 27.15021514892578, "learning_rate": 1.1800612525081848e-05, - "loss": 0.3861, + "loss": 1.0014, "step": 38820 }, { "epoch": 4.1, - "grad_norm": 20.817216873168945, + "grad_norm": 40.07487487792969, "learning_rate": 1.1798500369627206e-05, - "loss": 0.623, + "loss": 1.0958, "step": 38830 }, { "epoch": 4.1, - "grad_norm": 10.755607604980469, + "grad_norm": 42.73426818847656, "learning_rate": 1.1796388214172565e-05, - "loss": 0.5435, + "loss": 0.9813, "step": 38840 }, { "epoch": 4.1, - "grad_norm": 11.111671447753906, + "grad_norm": 47.0295524597168, "learning_rate": 1.1794276058717922e-05, - "loss": 0.2529, + "loss": 1.0107, "step": 38850 }, { "epoch": 4.1, - "grad_norm": 8.24242115020752, + "grad_norm": 28.323516845703125, "learning_rate": 1.1792163903263281e-05, - "loss": 0.7317, + "loss": 1.1029, "step": 38860 }, { "epoch": 4.1, - "grad_norm": 5.042944431304932, + "grad_norm": 23.253074645996094, "learning_rate": 1.179005174780864e-05, - "loss": 0.6609, + "loss": 1.2246, "step": 38870 }, { "epoch": 4.11, - "grad_norm": 8.367749214172363, + "grad_norm": 39.827484130859375, "learning_rate": 1.1787939592353999e-05, - "loss": 0.3885, + "loss": 0.7027, "step": 38880 }, { "epoch": 4.11, - "grad_norm": 9.966440200805664, + "grad_norm": 49.327659606933594, "learning_rate": 1.1785827436899357e-05, - "loss": 0.3304, + "loss": 1.1137, "step": 38890 }, { "epoch": 4.11, - "grad_norm": 7.548742771148682, + "grad_norm": 20.86945152282715, "learning_rate": 1.1783715281444715e-05, - "loss": 0.438, + "loss": 1.131, "step": 38900 }, { "epoch": 4.11, - "grad_norm": 16.628015518188477, + "grad_norm": 61.72562026977539, "learning_rate": 1.1781603125990073e-05, - "loss": 0.4663, + "loss": 1.4078, "step": 38910 }, { "epoch": 4.11, - "grad_norm": 6.872645378112793, + "grad_norm": 30.413780212402344, "learning_rate": 1.1779490970535432e-05, - "loss": 0.4395, + "loss": 0.9485, "step": 38920 }, { "epoch": 4.11, - "grad_norm": 13.17353630065918, + "grad_norm": 25.07554054260254, "learning_rate": 1.1777378815080791e-05, - "loss": 0.4353, + "loss": 1.021, "step": 38930 }, { "epoch": 4.11, - "grad_norm": 13.442817687988281, + "grad_norm": 66.24908447265625, "learning_rate": 1.177526665962615e-05, - "loss": 0.3672, + "loss": 0.8811, "step": 38940 }, { "epoch": 4.11, - "grad_norm": 5.632991313934326, + "grad_norm": 42.09246063232422, "learning_rate": 1.1773154504171508e-05, - "loss": 0.4486, + "loss": 1.3097, "step": 38950 }, { "epoch": 4.11, - "grad_norm": 29.853858947753906, + "grad_norm": 41.53839874267578, "learning_rate": 1.1771042348716866e-05, - "loss": 0.6626, + "loss": 1.1661, "step": 38960 }, { "epoch": 4.12, - "grad_norm": 10.839792251586914, + "grad_norm": 33.87113952636719, "learning_rate": 1.1768930193262224e-05, - "loss": 0.7137, + "loss": 1.2201, "step": 38970 }, { "epoch": 4.12, - "grad_norm": 12.57347297668457, + "grad_norm": 24.764129638671875, "learning_rate": 1.1766818037807583e-05, - "loss": 0.2489, + "loss": 1.0855, "step": 38980 }, { "epoch": 4.12, - "grad_norm": 19.534439086914062, + "grad_norm": 38.50740432739258, "learning_rate": 1.1764705882352942e-05, - "loss": 0.8849, + "loss": 1.2378, "step": 38990 }, { "epoch": 4.12, - "grad_norm": 18.916532516479492, + "grad_norm": 38.942832946777344, "learning_rate": 1.1762593726898302e-05, - "loss": 0.4605, + "loss": 1.5864, "step": 39000 }, { "epoch": 4.12, - "grad_norm": 9.076791763305664, + "grad_norm": 31.63212776184082, "learning_rate": 1.1760481571443658e-05, - "loss": 0.2507, + "loss": 1.0761, "step": 39010 }, { "epoch": 4.12, - "grad_norm": 19.032054901123047, + "grad_norm": 53.84452819824219, "learning_rate": 1.1758369415989016e-05, - "loss": 0.6644, + "loss": 1.0869, "step": 39020 }, { "epoch": 4.12, - "grad_norm": 6.663818359375, + "grad_norm": 37.293724060058594, "learning_rate": 1.1756257260534375e-05, - "loss": 0.5567, + "loss": 1.3236, "step": 39030 }, { "epoch": 4.12, - "grad_norm": 4.493454456329346, + "grad_norm": 21.94818115234375, "learning_rate": 1.1754145105079736e-05, - "loss": 0.317, + "loss": 0.9502, "step": 39040 }, { "epoch": 4.12, - "grad_norm": 4.968428611755371, + "grad_norm": 23.747173309326172, "learning_rate": 1.1752032949625095e-05, - "loss": 0.4735, + "loss": 1.3509, "step": 39050 }, { "epoch": 4.13, - "grad_norm": 7.714921474456787, + "grad_norm": 22.633899688720703, "learning_rate": 1.1749920794170453e-05, - "loss": 0.5375, + "loss": 1.3329, "step": 39060 }, { "epoch": 4.13, - "grad_norm": 13.884658813476562, + "grad_norm": 29.88999366760254, "learning_rate": 1.174780863871581e-05, - "loss": 0.5749, + "loss": 1.0383, "step": 39070 }, { "epoch": 4.13, - "grad_norm": 3.8423080444335938, + "grad_norm": 36.09423828125, "learning_rate": 1.174569648326117e-05, - "loss": 0.128, + "loss": 1.1755, "step": 39080 }, { "epoch": 4.13, - "grad_norm": 19.552200317382812, + "grad_norm": 47.298912048339844, "learning_rate": 1.1743584327806528e-05, - "loss": 0.4627, + "loss": 1.1796, "step": 39090 }, { "epoch": 4.13, - "grad_norm": 29.85185432434082, + "grad_norm": 27.14703369140625, "learning_rate": 1.1741472172351887e-05, - "loss": 0.2802, + "loss": 0.915, "step": 39100 }, { "epoch": 4.13, - "grad_norm": 21.78863525390625, + "grad_norm": 37.39570999145508, "learning_rate": 1.1739360016897245e-05, - "loss": 0.6033, + "loss": 1.1607, "step": 39110 }, { "epoch": 4.13, - "grad_norm": 5.211025714874268, + "grad_norm": 37.11575698852539, "learning_rate": 1.1737247861442603e-05, - "loss": 0.3586, + "loss": 1.2438, "step": 39120 }, { "epoch": 4.13, - "grad_norm": 1.8885301351547241, + "grad_norm": 36.12704849243164, "learning_rate": 1.1735135705987961e-05, - "loss": 0.4084, + "loss": 0.81, "step": 39130 }, { "epoch": 4.13, - "grad_norm": 19.20270347595215, + "grad_norm": 39.44384002685547, "learning_rate": 1.173302355053332e-05, - "loss": 0.89, + "loss": 1.4221, "step": 39140 }, { "epoch": 4.13, - "grad_norm": 9.254717826843262, + "grad_norm": 31.533164978027344, "learning_rate": 1.1730911395078679e-05, - "loss": 0.3056, + "loss": 1.167, "step": 39150 }, { "epoch": 4.14, - "grad_norm": 13.465356826782227, + "grad_norm": 21.362730026245117, "learning_rate": 1.1728799239624038e-05, - "loss": 0.4731, + "loss": 1.2406, "step": 39160 }, { "epoch": 4.14, - "grad_norm": 24.777172088623047, + "grad_norm": 33.617618560791016, "learning_rate": 1.1726687084169395e-05, - "loss": 0.4631, + "loss": 1.0896, "step": 39170 }, { "epoch": 4.14, - "grad_norm": 36.193485260009766, + "grad_norm": 35.55668640136719, "learning_rate": 1.1724574928714754e-05, - "loss": 0.3793, + "loss": 1.2901, "step": 39180 }, { "epoch": 4.14, - "grad_norm": 0.3728710114955902, + "grad_norm": 42.149208068847656, "learning_rate": 1.1722462773260112e-05, - "loss": 0.4617, + "loss": 1.15, "step": 39190 }, { "epoch": 4.14, - "grad_norm": 11.882060050964355, + "grad_norm": 58.82199478149414, "learning_rate": 1.1720350617805471e-05, - "loss": 0.3153, + "loss": 0.8254, "step": 39200 }, { "epoch": 4.14, - "grad_norm": 13.002878189086914, + "grad_norm": 39.813323974609375, "learning_rate": 1.171823846235083e-05, - "loss": 0.7241, + "loss": 1.6177, "step": 39210 }, { "epoch": 4.14, - "grad_norm": 5.734757423400879, + "grad_norm": 31.389238357543945, "learning_rate": 1.1716126306896189e-05, - "loss": 0.383, + "loss": 1.3181, "step": 39220 }, { "epoch": 4.14, - "grad_norm": 2.2091403007507324, + "grad_norm": 34.59459686279297, "learning_rate": 1.1714014151441546e-05, - "loss": 0.3027, + "loss": 1.1537, "step": 39230 }, { "epoch": 4.14, - "grad_norm": 7.021590709686279, + "grad_norm": 31.070911407470703, "learning_rate": 1.1711901995986905e-05, - "loss": 0.4739, + "loss": 1.1819, "step": 39240 }, { "epoch": 4.15, - "grad_norm": 15.546038627624512, + "grad_norm": 56.931312561035156, "learning_rate": 1.1709789840532263e-05, - "loss": 0.4646, + "loss": 1.1131, "step": 39250 }, { "epoch": 4.15, - "grad_norm": 30.001300811767578, + "grad_norm": 36.62705612182617, "learning_rate": 1.1707677685077624e-05, - "loss": 0.7201, + "loss": 1.1713, "step": 39260 }, { "epoch": 4.15, - "grad_norm": 19.2120418548584, + "grad_norm": 40.026119232177734, "learning_rate": 1.1705565529622983e-05, - "loss": 0.5753, + "loss": 1.1864, "step": 39270 }, { "epoch": 4.15, - "grad_norm": 3.734344720840454, + "grad_norm": 33.72330856323242, "learning_rate": 1.1703453374168338e-05, - "loss": 0.4334, + "loss": 0.9254, "step": 39280 }, { "epoch": 4.15, - "grad_norm": 29.605897903442383, + "grad_norm": 65.68729400634766, "learning_rate": 1.1701341218713698e-05, - "loss": 0.6495, + "loss": 1.2522, "step": 39290 }, { "epoch": 4.15, - "grad_norm": 11.09194564819336, + "grad_norm": 20.852998733520508, "learning_rate": 1.1699229063259057e-05, - "loss": 0.6276, + "loss": 1.226, "step": 39300 }, { "epoch": 4.15, - "grad_norm": 1.7389072179794312, + "grad_norm": 37.27389144897461, "learning_rate": 1.1697116907804416e-05, - "loss": 0.6936, + "loss": 1.0485, "step": 39310 }, { "epoch": 4.15, - "grad_norm": 11.102743148803711, + "grad_norm": 28.04558753967285, "learning_rate": 1.1695004752349775e-05, - "loss": 0.4692, + "loss": 1.5158, "step": 39320 }, { "epoch": 4.15, - "grad_norm": 14.028716087341309, + "grad_norm": 40.37141799926758, "learning_rate": 1.1692892596895134e-05, - "loss": 0.8124, + "loss": 1.2893, "step": 39330 }, { "epoch": 4.15, - "grad_norm": 6.747591972351074, + "grad_norm": 28.37591552734375, "learning_rate": 1.169078044144049e-05, - "loss": 0.6572, + "loss": 1.3225, "step": 39340 }, { "epoch": 4.16, - "grad_norm": 18.97630500793457, + "grad_norm": 34.25246810913086, "learning_rate": 1.168866828598585e-05, - "loss": 0.4889, + "loss": 1.0238, "step": 39350 }, { "epoch": 4.16, - "grad_norm": 17.578407287597656, + "grad_norm": 50.082176208496094, "learning_rate": 1.1686556130531208e-05, - "loss": 0.3675, + "loss": 1.1579, "step": 39360 }, { "epoch": 4.16, - "grad_norm": 17.348243713378906, + "grad_norm": 48.06691360473633, "learning_rate": 1.1684443975076567e-05, - "loss": 0.5603, + "loss": 1.1156, "step": 39370 }, { "epoch": 4.16, - "grad_norm": 5.961493492126465, + "grad_norm": 55.97500991821289, "learning_rate": 1.1682331819621926e-05, - "loss": 0.6428, + "loss": 1.3122, "step": 39380 }, { "epoch": 4.16, - "grad_norm": 0.6814684867858887, + "grad_norm": 5.151342868804932, "learning_rate": 1.1680219664167283e-05, - "loss": 0.3336, + "loss": 1.0256, "step": 39390 }, { "epoch": 4.16, - "grad_norm": 15.229790687561035, + "grad_norm": 30.102067947387695, "learning_rate": 1.1678107508712642e-05, - "loss": 0.5557, + "loss": 1.1318, "step": 39400 }, { "epoch": 4.16, - "grad_norm": 30.20806312561035, + "grad_norm": 20.753938674926758, "learning_rate": 1.1675995353258e-05, - "loss": 0.6748, + "loss": 1.0164, "step": 39410 }, { "epoch": 4.16, - "grad_norm": 28.939937591552734, + "grad_norm": 51.86698532104492, "learning_rate": 1.167388319780336e-05, - "loss": 0.6127, + "loss": 1.5447, "step": 39420 }, { "epoch": 4.16, - "grad_norm": 22.565187454223633, + "grad_norm": 41.561744689941406, "learning_rate": 1.1671771042348718e-05, - "loss": 0.6762, + "loss": 1.0303, "step": 39430 }, { "epoch": 4.17, - "grad_norm": 5.162730693817139, + "grad_norm": 47.62318801879883, "learning_rate": 1.1669658886894077e-05, - "loss": 0.3794, + "loss": 1.2616, "step": 39440 }, { "epoch": 4.17, - "grad_norm": 16.129993438720703, + "grad_norm": 38.4918098449707, "learning_rate": 1.1667546731439434e-05, - "loss": 0.5029, + "loss": 1.3287, "step": 39450 }, { "epoch": 4.17, - "grad_norm": 4.573050498962402, + "grad_norm": 43.39376449584961, "learning_rate": 1.1665434575984793e-05, - "loss": 0.4634, + "loss": 0.9838, "step": 39460 }, { "epoch": 4.17, - "grad_norm": 16.49355697631836, + "grad_norm": 39.8817138671875, "learning_rate": 1.1663322420530151e-05, - "loss": 0.3673, + "loss": 1.1956, "step": 39470 }, { "epoch": 4.17, - "grad_norm": 14.921178817749023, + "grad_norm": 39.93708419799805, "learning_rate": 1.1661210265075512e-05, - "loss": 0.3408, + "loss": 1.1109, "step": 39480 }, { "epoch": 4.17, - "grad_norm": 17.3868408203125, + "grad_norm": 40.328857421875, "learning_rate": 1.165909810962087e-05, - "loss": 0.8033, + "loss": 1.1305, "step": 39490 }, { "epoch": 4.17, - "grad_norm": 11.812970161437988, + "grad_norm": 31.081045150756836, "learning_rate": 1.1656985954166226e-05, - "loss": 0.3989, + "loss": 0.8871, "step": 39500 }, { "epoch": 4.17, - "grad_norm": 6.717746257781982, + "grad_norm": 23.473546981811523, "learning_rate": 1.1654873798711586e-05, - "loss": 0.4338, + "loss": 1.266, "step": 39510 }, { "epoch": 4.17, - "grad_norm": 15.895795822143555, + "grad_norm": 35.54801940917969, "learning_rate": 1.1652761643256945e-05, - "loss": 0.5824, + "loss": 1.009, "step": 39520 }, { "epoch": 4.17, - "grad_norm": 11.90742015838623, + "grad_norm": 31.514728546142578, "learning_rate": 1.1650649487802304e-05, - "loss": 0.331, + "loss": 0.7623, "step": 39530 }, { "epoch": 4.18, - "grad_norm": 2.377594232559204, + "grad_norm": 44.11870574951172, "learning_rate": 1.1648537332347663e-05, - "loss": 0.3725, + "loss": 1.0882, "step": 39540 }, { "epoch": 4.18, - "grad_norm": 10.89584732055664, + "grad_norm": 48.03297805786133, "learning_rate": 1.164642517689302e-05, - "loss": 0.4393, + "loss": 1.1374, "step": 39550 }, { "epoch": 4.18, - "grad_norm": 5.2749481201171875, + "grad_norm": 43.777809143066406, "learning_rate": 1.1644313021438379e-05, - "loss": 0.3209, + "loss": 0.9529, "step": 39560 }, { "epoch": 4.18, - "grad_norm": 7.497899532318115, + "grad_norm": 36.791473388671875, "learning_rate": 1.1642200865983737e-05, - "loss": 0.6688, + "loss": 1.4733, "step": 39570 }, { "epoch": 4.18, - "grad_norm": 3.1134843826293945, + "grad_norm": 44.850685119628906, "learning_rate": 1.1640088710529096e-05, - "loss": 0.3637, + "loss": 0.9999, "step": 39580 }, { "epoch": 4.18, - "grad_norm": 1.2383695840835571, + "grad_norm": 23.655519485473633, "learning_rate": 1.1637976555074455e-05, - "loss": 0.2101, + "loss": 0.7292, "step": 39590 }, { "epoch": 4.18, - "grad_norm": 0.663666844367981, + "grad_norm": 18.78793716430664, "learning_rate": 1.1635864399619814e-05, - "loss": 0.5719, + "loss": 1.0373, "step": 39600 }, { "epoch": 4.18, - "grad_norm": 40.52870559692383, + "grad_norm": 34.99991989135742, "learning_rate": 1.1633752244165171e-05, - "loss": 0.6714, + "loss": 1.1536, "step": 39610 }, { "epoch": 4.18, - "grad_norm": 9.344012260437012, + "grad_norm": 41.59433364868164, "learning_rate": 1.163164008871053e-05, - "loss": 0.3987, + "loss": 1.0775, "step": 39620 }, { "epoch": 4.19, - "grad_norm": 35.11079025268555, + "grad_norm": 38.22294998168945, "learning_rate": 1.1629527933255888e-05, - "loss": 0.8485, + "loss": 1.1891, "step": 39630 }, { "epoch": 4.19, - "grad_norm": 20.28158187866211, + "grad_norm": 36.37709426879883, "learning_rate": 1.1627415777801247e-05, - "loss": 0.6537, + "loss": 1.1602, "step": 39640 }, { "epoch": 4.19, - "grad_norm": 10.750984191894531, + "grad_norm": 62.83260726928711, "learning_rate": 1.1625303622346606e-05, - "loss": 0.5341, + "loss": 1.2257, "step": 39650 }, { "epoch": 4.19, - "grad_norm": 0.87795490026474, + "grad_norm": 47.10731887817383, "learning_rate": 1.1623191466891963e-05, - "loss": 0.546, + "loss": 1.4648, "step": 39660 }, { "epoch": 4.19, - "grad_norm": 0.9263743758201599, + "grad_norm": 46.44131851196289, "learning_rate": 1.1621079311437322e-05, - "loss": 0.2853, + "loss": 0.9583, "step": 39670 }, { "epoch": 4.19, - "grad_norm": 16.564476013183594, + "grad_norm": 43.060943603515625, "learning_rate": 1.161896715598268e-05, - "loss": 0.5012, + "loss": 1.1005, "step": 39680 }, { "epoch": 4.19, - "grad_norm": 19.215717315673828, + "grad_norm": 44.61949920654297, "learning_rate": 1.161685500052804e-05, - "loss": 0.4361, + "loss": 1.0899, "step": 39690 }, { "epoch": 4.19, - "grad_norm": 0.3088436424732208, + "grad_norm": 19.537275314331055, "learning_rate": 1.1614742845073398e-05, - "loss": 0.3022, + "loss": 1.0582, "step": 39700 }, { "epoch": 4.19, - "grad_norm": 11.43241024017334, + "grad_norm": 17.8961124420166, "learning_rate": 1.1612630689618759e-05, - "loss": 0.5577, + "loss": 0.8207, "step": 39710 }, { "epoch": 4.19, - "grad_norm": 12.137612342834473, + "grad_norm": 39.35676956176758, "learning_rate": 1.1610518534164114e-05, - "loss": 0.6742, + "loss": 1.4285, "step": 39720 }, { "epoch": 4.2, - "grad_norm": 17.468420028686523, + "grad_norm": 53.47874450683594, "learning_rate": 1.1608406378709473e-05, - "loss": 0.8217, + "loss": 1.667, "step": 39730 }, { "epoch": 4.2, - "grad_norm": 19.899354934692383, + "grad_norm": 60.169979095458984, "learning_rate": 1.1606294223254833e-05, - "loss": 0.5491, + "loss": 1.1672, "step": 39740 }, { "epoch": 4.2, - "grad_norm": 20.550941467285156, + "grad_norm": 30.97609519958496, "learning_rate": 1.1604182067800192e-05, - "loss": 0.7572, + "loss": 0.9749, "step": 39750 }, { "epoch": 4.2, - "grad_norm": 7.296797275543213, + "grad_norm": 42.68992233276367, "learning_rate": 1.1602069912345551e-05, - "loss": 0.468, + "loss": 1.1896, "step": 39760 }, { "epoch": 4.2, - "grad_norm": 0.07744491845369339, + "grad_norm": 6.364953994750977, "learning_rate": 1.1599957756890908e-05, - "loss": 0.586, + "loss": 1.0357, "step": 39770 }, { "epoch": 4.2, - "grad_norm": 9.098978042602539, + "grad_norm": 32.24485778808594, "learning_rate": 1.1597845601436267e-05, - "loss": 0.6186, + "loss": 1.0707, "step": 39780 }, { "epoch": 4.2, - "grad_norm": 0.3375403583049774, + "grad_norm": 50.08983612060547, "learning_rate": 1.1595733445981626e-05, - "loss": 0.3857, + "loss": 0.881, "step": 39790 }, { "epoch": 4.2, - "grad_norm": 4.448338985443115, + "grad_norm": 34.62040328979492, "learning_rate": 1.1593621290526984e-05, - "loss": 0.7415, + "loss": 1.289, "step": 39800 }, { "epoch": 4.2, - "grad_norm": 10.54080867767334, + "grad_norm": 36.33688735961914, "learning_rate": 1.1591509135072343e-05, - "loss": 0.254, + "loss": 0.8032, "step": 39810 }, { "epoch": 4.21, - "grad_norm": 28.698392868041992, + "grad_norm": 39.16103744506836, "learning_rate": 1.15893969796177e-05, - "loss": 0.7439, + "loss": 1.4309, "step": 39820 }, { "epoch": 4.21, - "grad_norm": 14.084227561950684, + "grad_norm": 43.99022674560547, "learning_rate": 1.1587284824163059e-05, - "loss": 0.4608, + "loss": 1.1781, "step": 39830 }, { "epoch": 4.21, - "grad_norm": 5.908926486968994, + "grad_norm": 34.451988220214844, "learning_rate": 1.1585172668708418e-05, - "loss": 0.6575, + "loss": 1.0728, "step": 39840 }, { "epoch": 4.21, - "grad_norm": 12.41535472869873, + "grad_norm": 34.381263732910156, "learning_rate": 1.1583060513253776e-05, - "loss": 0.5056, + "loss": 1.0511, "step": 39850 }, { "epoch": 4.21, - "grad_norm": 20.956703186035156, + "grad_norm": 49.289878845214844, "learning_rate": 1.1580948357799135e-05, - "loss": 0.4837, + "loss": 1.1362, "step": 39860 }, { "epoch": 4.21, - "grad_norm": 17.32634162902832, + "grad_norm": 44.090904235839844, "learning_rate": 1.1578836202344494e-05, - "loss": 0.3584, + "loss": 1.0737, "step": 39870 }, { "epoch": 4.21, - "grad_norm": 25.28984832763672, + "grad_norm": 50.550018310546875, "learning_rate": 1.1576724046889851e-05, - "loss": 0.4887, + "loss": 0.9919, "step": 39880 }, { "epoch": 4.21, - "grad_norm": 15.632521629333496, + "grad_norm": 25.968130111694336, "learning_rate": 1.157461189143521e-05, - "loss": 0.506, + "loss": 1.1407, "step": 39890 }, { "epoch": 4.21, - "grad_norm": 9.084136009216309, + "grad_norm": 54.36870193481445, "learning_rate": 1.1572499735980569e-05, - "loss": 0.3203, + "loss": 1.0502, "step": 39900 }, { "epoch": 4.21, - "grad_norm": 17.2757625579834, + "grad_norm": 41.808067321777344, "learning_rate": 1.1570387580525927e-05, - "loss": 0.603, + "loss": 1.5057, "step": 39910 }, { "epoch": 4.22, - "grad_norm": 18.984455108642578, + "grad_norm": 28.86142349243164, "learning_rate": 1.1568275425071286e-05, - "loss": 0.9019, + "loss": 1.2128, "step": 39920 }, { "epoch": 4.22, - "grad_norm": 19.491483688354492, + "grad_norm": 18.27988624572754, "learning_rate": 1.1566163269616643e-05, - "loss": 0.6, + "loss": 1.139, "step": 39930 }, { "epoch": 4.22, - "grad_norm": 18.699230194091797, + "grad_norm": 48.00465774536133, "learning_rate": 1.1564051114162002e-05, - "loss": 0.4332, + "loss": 1.2571, "step": 39940 }, { "epoch": 4.22, - "grad_norm": 0.14837504923343658, + "grad_norm": 20.185688018798828, "learning_rate": 1.1561938958707361e-05, - "loss": 0.3851, + "loss": 0.8762, "step": 39950 }, { "epoch": 4.22, - "grad_norm": 8.554854393005371, + "grad_norm": 42.51039505004883, "learning_rate": 1.1559826803252721e-05, - "loss": 0.5209, + "loss": 1.0512, "step": 39960 }, { "epoch": 4.22, - "grad_norm": 15.707817077636719, + "grad_norm": 35.53881072998047, "learning_rate": 1.155771464779808e-05, - "loss": 0.5516, + "loss": 1.4106, "step": 39970 }, { "epoch": 4.22, - "grad_norm": 1.9223405122756958, + "grad_norm": 27.418258666992188, "learning_rate": 1.1555602492343439e-05, - "loss": 0.3467, + "loss": 0.9372, "step": 39980 }, { "epoch": 4.22, - "grad_norm": 26.598840713500977, + "grad_norm": 34.452850341796875, "learning_rate": 1.1553490336888796e-05, - "loss": 0.5924, + "loss": 1.0191, "step": 39990 }, { "epoch": 4.22, - "grad_norm": 5.639843940734863, + "grad_norm": 34.179019927978516, "learning_rate": 1.1551378181434155e-05, - "loss": 0.373, + "loss": 0.9417, "step": 40000 }, { "epoch": 4.23, - "grad_norm": 13.530112266540527, + "grad_norm": 36.481529235839844, "learning_rate": 1.1549266025979514e-05, - "loss": 0.2761, + "loss": 0.8553, "step": 40010 }, { "epoch": 4.23, - "grad_norm": 10.539457321166992, + "grad_norm": 30.947818756103516, "learning_rate": 1.1547153870524872e-05, - "loss": 0.3751, + "loss": 1.0655, "step": 40020 }, { "epoch": 4.23, - "grad_norm": 1.2875866889953613, + "grad_norm": 50.76640319824219, "learning_rate": 1.1545041715070231e-05, - "loss": 0.4159, + "loss": 0.9431, "step": 40030 }, { "epoch": 4.23, - "grad_norm": 3.8129475116729736, + "grad_norm": 38.90962600708008, "learning_rate": 1.1542929559615588e-05, - "loss": 0.4352, + "loss": 1.3371, "step": 40040 }, { "epoch": 4.23, - "grad_norm": 7.559730529785156, + "grad_norm": 63.108848571777344, "learning_rate": 1.1540817404160947e-05, - "loss": 0.3837, + "loss": 1.029, "step": 40050 }, { "epoch": 4.23, - "grad_norm": 17.536928176879883, + "grad_norm": 32.486122131347656, "learning_rate": 1.1538705248706306e-05, - "loss": 0.3936, + "loss": 1.012, "step": 40060 }, { "epoch": 4.23, - "grad_norm": 14.692266464233398, + "grad_norm": 38.85729217529297, "learning_rate": 1.1536593093251665e-05, - "loss": 0.5559, + "loss": 1.3041, "step": 40070 }, { "epoch": 4.23, - "grad_norm": 11.893195152282715, + "grad_norm": 30.246511459350586, "learning_rate": 1.1534480937797023e-05, - "loss": 0.3732, + "loss": 1.2179, "step": 40080 }, { "epoch": 4.23, - "grad_norm": 5.9275288581848145, + "grad_norm": 45.87162399291992, "learning_rate": 1.1532368782342382e-05, - "loss": 0.7515, + "loss": 1.2905, "step": 40090 }, { "epoch": 4.23, - "grad_norm": 1.526100993156433, + "grad_norm": 39.32083511352539, "learning_rate": 1.153025662688774e-05, - "loss": 0.5315, + "loss": 0.9602, "step": 40100 }, { "epoch": 4.24, - "grad_norm": 8.437020301818848, + "grad_norm": 28.84262466430664, "learning_rate": 1.1528144471433098e-05, - "loss": 0.5939, + "loss": 0.917, "step": 40110 }, { "epoch": 4.24, - "grad_norm": 8.112736701965332, + "grad_norm": 32.50984573364258, "learning_rate": 1.1526032315978457e-05, - "loss": 0.4581, + "loss": 0.8299, "step": 40120 }, { "epoch": 4.24, - "grad_norm": 0.1765032708644867, + "grad_norm": 27.032493591308594, "learning_rate": 1.1523920160523816e-05, - "loss": 0.4566, + "loss": 1.2715, "step": 40130 }, { "epoch": 4.24, - "grad_norm": 10.23727798461914, + "grad_norm": 32.437686920166016, "learning_rate": 1.1521808005069174e-05, - "loss": 0.5869, + "loss": 1.4504, "step": 40140 }, { "epoch": 4.24, - "grad_norm": 1.1138691902160645, + "grad_norm": 19.439729690551758, "learning_rate": 1.1519695849614531e-05, - "loss": 0.3956, + "loss": 1.1443, "step": 40150 }, { "epoch": 4.24, - "grad_norm": 15.41867446899414, + "grad_norm": 42.175907135009766, "learning_rate": 1.151758369415989e-05, - "loss": 0.5695, + "loss": 1.2921, "step": 40160 }, { "epoch": 4.24, - "grad_norm": 0.5754055976867676, + "grad_norm": 32.13591384887695, "learning_rate": 1.1515471538705249e-05, - "loss": 0.3749, + "loss": 0.9025, "step": 40170 }, { "epoch": 4.24, - "grad_norm": 3.781697988510132, + "grad_norm": 8.608071327209473, "learning_rate": 1.1513359383250608e-05, - "loss": 0.4444, + "loss": 0.8162, "step": 40180 }, { "epoch": 4.24, - "grad_norm": 7.638463020324707, + "grad_norm": 27.77899742126465, "learning_rate": 1.1511247227795968e-05, - "loss": 0.4752, + "loss": 0.9796, "step": 40190 }, { "epoch": 4.25, - "grad_norm": 0.18421334028244019, + "grad_norm": 36.184810638427734, "learning_rate": 1.1509135072341324e-05, - "loss": 0.4111, + "loss": 0.8, "step": 40200 }, { "epoch": 4.25, - "grad_norm": 23.869464874267578, + "grad_norm": 51.26912307739258, "learning_rate": 1.1507022916886682e-05, - "loss": 0.4088, + "loss": 1.0723, "step": 40210 }, { "epoch": 4.25, - "grad_norm": 0.8977985382080078, + "grad_norm": 52.47798156738281, "learning_rate": 1.1504910761432043e-05, - "loss": 0.3093, + "loss": 1.0982, "step": 40220 }, { "epoch": 4.25, - "grad_norm": 9.855627059936523, + "grad_norm": 42.24728775024414, "learning_rate": 1.1502798605977402e-05, - "loss": 0.6504, + "loss": 1.3316, "step": 40230 }, { "epoch": 4.25, - "grad_norm": 12.764152526855469, + "grad_norm": 28.769695281982422, "learning_rate": 1.150068645052276e-05, - "loss": 0.3649, + "loss": 0.8077, "step": 40240 }, { "epoch": 4.25, - "grad_norm": 0.02455197647213936, + "grad_norm": 36.59138488769531, "learning_rate": 1.149857429506812e-05, - "loss": 0.2861, + "loss": 1.3017, "step": 40250 }, { "epoch": 4.25, - "grad_norm": 24.908021926879883, + "grad_norm": 56.07361602783203, "learning_rate": 1.1496462139613476e-05, - "loss": 0.4332, + "loss": 1.1166, "step": 40260 }, { "epoch": 4.25, - "grad_norm": 28.868083953857422, + "grad_norm": 39.82126998901367, "learning_rate": 1.1494349984158835e-05, - "loss": 0.5308, + "loss": 1.2591, "step": 40270 }, { "epoch": 4.25, - "grad_norm": 22.728424072265625, + "grad_norm": 40.54254913330078, "learning_rate": 1.1492237828704194e-05, - "loss": 0.3247, + "loss": 0.9442, "step": 40280 }, { "epoch": 4.25, - "grad_norm": 11.50432300567627, + "grad_norm": 23.985652923583984, "learning_rate": 1.1490125673249553e-05, - "loss": 0.4491, + "loss": 1.6013, "step": 40290 }, { "epoch": 4.26, - "grad_norm": 12.746760368347168, + "grad_norm": 33.77611541748047, "learning_rate": 1.1488013517794911e-05, - "loss": 0.2716, + "loss": 0.9774, "step": 40300 }, { "epoch": 4.26, - "grad_norm": 12.437889099121094, + "grad_norm": 42.365753173828125, "learning_rate": 1.1485901362340268e-05, - "loss": 0.2963, + "loss": 0.8993, "step": 40310 }, { "epoch": 4.26, - "grad_norm": 3.7498528957366943, + "grad_norm": 23.59530258178711, "learning_rate": 1.1483789206885627e-05, - "loss": 0.4885, + "loss": 1.0199, "step": 40320 }, { "epoch": 4.26, - "grad_norm": 10.021551132202148, + "grad_norm": 57.974082946777344, "learning_rate": 1.1481677051430986e-05, - "loss": 0.5191, + "loss": 1.2432, "step": 40330 }, { "epoch": 4.26, - "grad_norm": 22.6623592376709, + "grad_norm": 31.36055564880371, "learning_rate": 1.1479564895976345e-05, - "loss": 0.637, + "loss": 1.3838, "step": 40340 }, { "epoch": 4.26, - "grad_norm": 30.557327270507812, + "grad_norm": 44.375389099121094, "learning_rate": 1.1477452740521704e-05, - "loss": 0.2481, + "loss": 1.1679, "step": 40350 }, { "epoch": 4.26, - "grad_norm": 17.675251007080078, + "grad_norm": 40.27906036376953, "learning_rate": 1.1475340585067062e-05, - "loss": 0.4982, + "loss": 1.1865, "step": 40360 }, { "epoch": 4.26, - "grad_norm": 19.221843719482422, + "grad_norm": 33.122962951660156, "learning_rate": 1.147322842961242e-05, - "loss": 0.5388, + "loss": 1.1667, "step": 40370 }, { "epoch": 4.26, - "grad_norm": 27.027372360229492, + "grad_norm": 24.144277572631836, "learning_rate": 1.1471116274157778e-05, - "loss": 0.4802, + "loss": 0.9304, "step": 40380 }, { "epoch": 4.27, - "grad_norm": 16.438831329345703, + "grad_norm": 45.629634857177734, "learning_rate": 1.1469004118703137e-05, - "loss": 0.339, + "loss": 1.4227, "step": 40390 }, { "epoch": 4.27, - "grad_norm": 10.362844467163086, + "grad_norm": 46.519596099853516, "learning_rate": 1.1466891963248496e-05, - "loss": 0.4899, + "loss": 1.4463, "step": 40400 }, { "epoch": 4.27, - "grad_norm": 20.4457950592041, + "grad_norm": 52.73751449584961, "learning_rate": 1.1464779807793856e-05, - "loss": 0.428, + "loss": 1.1987, "step": 40410 }, { "epoch": 4.27, - "grad_norm": 19.71921730041504, + "grad_norm": 44.329750061035156, "learning_rate": 1.1462667652339212e-05, - "loss": 0.5292, + "loss": 1.0434, "step": 40420 }, { "epoch": 4.27, - "grad_norm": 23.719966888427734, + "grad_norm": 51.17835998535156, "learning_rate": 1.146055549688457e-05, - "loss": 0.6753, + "loss": 1.3918, "step": 40430 }, { "epoch": 4.27, - "grad_norm": 17.30626106262207, + "grad_norm": 45.62296676635742, "learning_rate": 1.1458443341429931e-05, - "loss": 0.4915, + "loss": 0.9914, "step": 40440 }, { "epoch": 4.27, - "grad_norm": 15.292901039123535, + "grad_norm": 47.059017181396484, "learning_rate": 1.145633118597529e-05, - "loss": 0.4694, + "loss": 1.2451, "step": 40450 }, { "epoch": 4.27, - "grad_norm": 7.105876445770264, + "grad_norm": 44.348915100097656, "learning_rate": 1.1454219030520648e-05, - "loss": 0.6357, + "loss": 1.0194, "step": 40460 }, { "epoch": 4.27, - "grad_norm": 43.837738037109375, + "grad_norm": 40.61934280395508, "learning_rate": 1.1452106875066006e-05, - "loss": 1.0391, + "loss": 1.2836, "step": 40470 }, { "epoch": 4.28, - "grad_norm": 0.21511070430278778, + "grad_norm": 45.65546798706055, "learning_rate": 1.1449994719611364e-05, - "loss": 0.5665, + "loss": 1.4076, "step": 40480 }, { "epoch": 4.28, - "grad_norm": 27.25122833251953, + "grad_norm": 35.760860443115234, "learning_rate": 1.1447882564156723e-05, - "loss": 0.3543, + "loss": 1.0076, "step": 40490 }, { "epoch": 4.28, - "grad_norm": 1.5868022441864014, + "grad_norm": 39.6793098449707, "learning_rate": 1.1445770408702082e-05, - "loss": 0.2955, + "loss": 1.1741, "step": 40500 }, { "epoch": 4.28, - "grad_norm": 1.058039665222168, + "grad_norm": 25.374629974365234, "learning_rate": 1.144365825324744e-05, - "loss": 0.4737, + "loss": 1.1354, "step": 40510 }, { "epoch": 4.28, - "grad_norm": 1.528037428855896, + "grad_norm": 27.69355583190918, "learning_rate": 1.14415460977928e-05, - "loss": 0.4866, + "loss": 1.1048, "step": 40520 }, { "epoch": 4.28, - "grad_norm": 6.067834377288818, + "grad_norm": 54.8121223449707, "learning_rate": 1.1439433942338157e-05, - "loss": 0.6189, + "loss": 1.2236, "step": 40530 }, { "epoch": 4.28, - "grad_norm": 8.605327606201172, + "grad_norm": 42.97631072998047, "learning_rate": 1.1437321786883515e-05, - "loss": 0.5346, + "loss": 0.8737, "step": 40540 }, { "epoch": 4.28, - "grad_norm": 11.753207206726074, + "grad_norm": 39.851016998291016, "learning_rate": 1.1435209631428874e-05, - "loss": 0.7006, + "loss": 1.0429, "step": 40550 }, { "epoch": 4.28, - "grad_norm": 5.083379745483398, + "grad_norm": 22.44198226928711, "learning_rate": 1.1433097475974233e-05, - "loss": 0.5109, + "loss": 1.264, "step": 40560 }, { "epoch": 4.28, - "grad_norm": 6.595483779907227, + "grad_norm": 31.139842987060547, "learning_rate": 1.1430985320519592e-05, - "loss": 0.4112, + "loss": 1.2254, "step": 40570 }, { "epoch": 4.29, - "grad_norm": 22.91722869873047, + "grad_norm": 38.33565139770508, "learning_rate": 1.1428873165064949e-05, - "loss": 0.4608, + "loss": 1.1288, "step": 40580 }, { "epoch": 4.29, - "grad_norm": 20.53095054626465, + "grad_norm": 42.22339630126953, "learning_rate": 1.1426761009610307e-05, - "loss": 0.5462, + "loss": 0.9747, "step": 40590 }, { "epoch": 4.29, - "grad_norm": 0.4839990735054016, + "grad_norm": 21.156715393066406, "learning_rate": 1.1424648854155666e-05, - "loss": 0.5623, + "loss": 1.1916, "step": 40600 }, { "epoch": 4.29, - "grad_norm": 10.520370483398438, + "grad_norm": 39.629730224609375, "learning_rate": 1.1422536698701025e-05, - "loss": 0.1417, + "loss": 0.8232, "step": 40610 }, { "epoch": 4.29, - "grad_norm": 8.022796630859375, + "grad_norm": 43.38176345825195, "learning_rate": 1.1420424543246384e-05, - "loss": 0.4169, + "loss": 1.218, "step": 40620 }, { "epoch": 4.29, - "grad_norm": 18.445240020751953, + "grad_norm": 30.651885986328125, "learning_rate": 1.1418312387791743e-05, - "loss": 0.4054, + "loss": 0.9797, "step": 40630 }, { "epoch": 4.29, - "grad_norm": 10.403904914855957, + "grad_norm": 42.185089111328125, "learning_rate": 1.14162002323371e-05, - "loss": 0.4095, + "loss": 1.0716, "step": 40640 }, { "epoch": 4.29, - "grad_norm": 13.994674682617188, + "grad_norm": 43.84062957763672, "learning_rate": 1.1414088076882458e-05, - "loss": 0.4629, + "loss": 0.9273, "step": 40650 }, { "epoch": 4.29, - "grad_norm": 25.061214447021484, + "grad_norm": 52.053592681884766, "learning_rate": 1.1411975921427817e-05, - "loss": 0.7642, + "loss": 1.08, "step": 40660 }, { "epoch": 4.3, - "grad_norm": 9.4295654296875, + "grad_norm": 19.70350456237793, "learning_rate": 1.1409863765973178e-05, - "loss": 0.6413, + "loss": 1.0234, "step": 40670 }, { "epoch": 4.3, - "grad_norm": 3.548229932785034, + "grad_norm": 34.11163330078125, "learning_rate": 1.1407751610518537e-05, - "loss": 0.3689, + "loss": 0.8945, "step": 40680 }, { "epoch": 4.3, - "grad_norm": 14.829821586608887, + "grad_norm": 35.406463623046875, "learning_rate": 1.1405639455063892e-05, - "loss": 0.5007, + "loss": 1.0801, "step": 40690 }, { "epoch": 4.3, - "grad_norm": 3.0547139644622803, + "grad_norm": 36.576873779296875, "learning_rate": 1.1403527299609252e-05, - "loss": 0.8181, + "loss": 1.3537, "step": 40700 }, { "epoch": 4.3, - "grad_norm": 15.647976875305176, + "grad_norm": 38.267887115478516, "learning_rate": 1.1401415144154611e-05, - "loss": 0.5009, + "loss": 0.9728, "step": 40710 }, { "epoch": 4.3, - "grad_norm": 15.957510948181152, + "grad_norm": 46.7331428527832, "learning_rate": 1.139930298869997e-05, - "loss": 0.6795, + "loss": 1.081, "step": 40720 }, { "epoch": 4.3, - "grad_norm": 0.21378502249717712, + "grad_norm": 39.37126159667969, "learning_rate": 1.1397190833245329e-05, - "loss": 0.2397, + "loss": 1.0229, "step": 40730 }, { "epoch": 4.3, - "grad_norm": 18.38176727294922, + "grad_norm": 29.255990982055664, "learning_rate": 1.1395078677790686e-05, - "loss": 0.4179, + "loss": 0.783, "step": 40740 }, { "epoch": 4.3, - "grad_norm": 22.625255584716797, + "grad_norm": 42.61097717285156, "learning_rate": 1.1392966522336045e-05, - "loss": 0.6377, + "loss": 1.1692, "step": 40750 }, { "epoch": 4.3, - "grad_norm": 16.93536949157715, + "grad_norm": 49.682167053222656, "learning_rate": 1.1390854366881403e-05, - "loss": 0.4888, + "loss": 1.164, "step": 40760 }, { "epoch": 4.31, - "grad_norm": 22.12802505493164, + "grad_norm": 42.26911544799805, "learning_rate": 1.1388742211426762e-05, - "loss": 0.5728, + "loss": 0.9386, "step": 40770 }, { "epoch": 4.31, - "grad_norm": 7.72675895690918, + "grad_norm": 44.87266159057617, "learning_rate": 1.1386630055972121e-05, - "loss": 0.5189, + "loss": 1.0503, "step": 40780 }, { "epoch": 4.31, - "grad_norm": 24.61390495300293, + "grad_norm": 32.73188018798828, "learning_rate": 1.138451790051748e-05, - "loss": 0.6637, + "loss": 0.7745, "step": 40790 }, { "epoch": 4.31, - "grad_norm": 25.889141082763672, + "grad_norm": 52.9683837890625, "learning_rate": 1.1382405745062837e-05, - "loss": 0.5797, + "loss": 1.2588, "step": 40800 }, { "epoch": 4.31, - "grad_norm": 15.144042015075684, + "grad_norm": 30.085222244262695, "learning_rate": 1.1380293589608196e-05, - "loss": 0.7013, + "loss": 0.9542, "step": 40810 }, { "epoch": 4.31, - "grad_norm": 5.176293849945068, + "grad_norm": 35.78371810913086, "learning_rate": 1.1378181434153554e-05, - "loss": 0.2217, + "loss": 0.8277, "step": 40820 }, { "epoch": 4.31, - "grad_norm": 32.626686096191406, + "grad_norm": 37.128170013427734, "learning_rate": 1.1376069278698913e-05, - "loss": 0.9803, + "loss": 1.3207, "step": 40830 }, { "epoch": 4.31, - "grad_norm": 18.795974731445312, + "grad_norm": 61.296146392822266, "learning_rate": 1.1373957123244272e-05, - "loss": 0.5569, + "loss": 1.1724, "step": 40840 }, { "epoch": 4.31, - "grad_norm": 20.78373908996582, + "grad_norm": 9.921735763549805, "learning_rate": 1.1371844967789629e-05, - "loss": 0.7906, + "loss": 1.1957, "step": 40850 }, { "epoch": 4.32, - "grad_norm": 14.626008033752441, + "grad_norm": 31.28858184814453, "learning_rate": 1.1369732812334988e-05, - "loss": 0.2833, + "loss": 1.2303, "step": 40860 }, { "epoch": 4.32, - "grad_norm": 0.22764025628566742, + "grad_norm": 16.072338104248047, "learning_rate": 1.1367620656880347e-05, - "loss": 0.1854, + "loss": 0.6177, "step": 40870 }, { "epoch": 4.32, - "grad_norm": 31.33616065979004, + "grad_norm": 45.50018310546875, "learning_rate": 1.1365508501425705e-05, - "loss": 0.5159, + "loss": 1.2054, "step": 40880 }, { "epoch": 4.32, - "grad_norm": 12.258035659790039, + "grad_norm": 23.109167098999023, "learning_rate": 1.1363396345971066e-05, - "loss": 0.3209, + "loss": 0.9349, "step": 40890 }, { "epoch": 4.32, - "grad_norm": 30.553512573242188, + "grad_norm": 58.72578811645508, "learning_rate": 1.1361284190516425e-05, - "loss": 0.6367, + "loss": 1.4004, "step": 40900 }, { "epoch": 4.32, - "grad_norm": 4.652672290802002, + "grad_norm": 34.880958557128906, "learning_rate": 1.135917203506178e-05, - "loss": 0.4879, + "loss": 1.0187, "step": 40910 }, { "epoch": 4.32, - "grad_norm": 27.11185646057129, + "grad_norm": 30.690174102783203, "learning_rate": 1.135705987960714e-05, - "loss": 0.4425, + "loss": 0.7596, "step": 40920 }, { "epoch": 4.32, - "grad_norm": 4.076023578643799, + "grad_norm": 40.583587646484375, "learning_rate": 1.13549477241525e-05, - "loss": 0.3942, + "loss": 1.035, "step": 40930 }, { "epoch": 4.32, - "grad_norm": 6.287929534912109, + "grad_norm": 34.14455795288086, "learning_rate": 1.1352835568697858e-05, - "loss": 0.4531, + "loss": 1.0027, "step": 40940 }, { "epoch": 4.32, - "grad_norm": 2.88468337059021, + "grad_norm": 48.25914764404297, "learning_rate": 1.1350723413243217e-05, - "loss": 0.5817, + "loss": 0.9338, "step": 40950 }, { "epoch": 4.33, - "grad_norm": 0.20526385307312012, + "grad_norm": 34.53844451904297, "learning_rate": 1.1348611257788574e-05, - "loss": 0.4213, + "loss": 1.3532, "step": 40960 }, { "epoch": 4.33, - "grad_norm": 19.824655532836914, + "grad_norm": 37.02031707763672, "learning_rate": 1.1346499102333933e-05, - "loss": 0.782, + "loss": 1.3562, "step": 40970 }, { "epoch": 4.33, - "grad_norm": 13.710875511169434, + "grad_norm": 37.427242279052734, "learning_rate": 1.1344386946879291e-05, - "loss": 0.7081, + "loss": 1.4632, "step": 40980 }, { "epoch": 4.33, - "grad_norm": 11.855353355407715, + "grad_norm": 36.88764953613281, "learning_rate": 1.134227479142465e-05, - "loss": 0.602, + "loss": 1.2082, "step": 40990 }, { "epoch": 4.33, - "grad_norm": 7.854673862457275, + "grad_norm": 36.828392028808594, "learning_rate": 1.1340162635970009e-05, - "loss": 0.6562, + "loss": 1.0573, "step": 41000 }, { "epoch": 4.33, - "grad_norm": 8.914519309997559, + "grad_norm": 29.569271087646484, "learning_rate": 1.1338050480515368e-05, - "loss": 0.2967, + "loss": 0.7972, "step": 41010 }, { "epoch": 4.33, - "grad_norm": 13.994894027709961, + "grad_norm": 61.10160827636719, "learning_rate": 1.1335938325060725e-05, - "loss": 0.5789, + "loss": 1.3997, "step": 41020 }, { "epoch": 4.33, - "grad_norm": 12.689196586608887, + "grad_norm": 23.871597290039062, "learning_rate": 1.1333826169606084e-05, - "loss": 0.7591, + "loss": 1.1919, "step": 41030 }, { "epoch": 4.33, - "grad_norm": 1.784970760345459, + "grad_norm": 41.17179489135742, "learning_rate": 1.1331714014151442e-05, - "loss": 0.6539, + "loss": 1.2348, "step": 41040 }, { "epoch": 4.34, - "grad_norm": 14.082474708557129, + "grad_norm": 28.82813835144043, "learning_rate": 1.1329601858696801e-05, - "loss": 0.3006, + "loss": 1.3333, "step": 41050 }, { "epoch": 4.34, - "grad_norm": 13.779844284057617, + "grad_norm": 56.26686096191406, "learning_rate": 1.132748970324216e-05, - "loss": 0.4256, + "loss": 0.8215, "step": 41060 }, { "epoch": 4.34, - "grad_norm": 6.973468780517578, + "grad_norm": 36.12044143676758, "learning_rate": 1.1325377547787517e-05, - "loss": 0.4248, + "loss": 1.1398, "step": 41070 }, { "epoch": 4.34, - "grad_norm": 11.28349494934082, + "grad_norm": 40.564849853515625, "learning_rate": 1.1323265392332876e-05, - "loss": 0.5757, + "loss": 0.922, "step": 41080 }, { "epoch": 4.34, - "grad_norm": 7.725698947906494, + "grad_norm": 41.44871139526367, "learning_rate": 1.1321153236878235e-05, - "loss": 0.6432, + "loss": 1.5186, "step": 41090 }, { "epoch": 4.34, - "grad_norm": 3.679680824279785, + "grad_norm": 25.50931739807129, "learning_rate": 1.1319041081423593e-05, - "loss": 0.3042, + "loss": 1.0829, "step": 41100 }, { "epoch": 4.34, - "grad_norm": 25.560516357421875, + "grad_norm": 44.776309967041016, "learning_rate": 1.1316928925968954e-05, - "loss": 0.5804, + "loss": 1.395, "step": 41110 }, { "epoch": 4.34, - "grad_norm": 49.016510009765625, + "grad_norm": 27.519102096557617, "learning_rate": 1.131481677051431e-05, - "loss": 0.686, + "loss": 1.1847, "step": 41120 }, { "epoch": 4.34, - "grad_norm": 19.15426254272461, + "grad_norm": 29.6638240814209, "learning_rate": 1.1312704615059668e-05, - "loss": 0.4714, + "loss": 1.1803, "step": 41130 }, { "epoch": 4.34, - "grad_norm": 1.5465326309204102, + "grad_norm": 46.769954681396484, "learning_rate": 1.1310592459605027e-05, - "loss": 0.3426, + "loss": 0.8265, "step": 41140 }, { "epoch": 4.35, - "grad_norm": 17.659027099609375, + "grad_norm": 44.498512268066406, "learning_rate": 1.1308480304150387e-05, - "loss": 0.5275, + "loss": 1.1931, "step": 41150 }, { "epoch": 4.35, - "grad_norm": 1.285099983215332, + "grad_norm": 41.795379638671875, "learning_rate": 1.1306368148695746e-05, - "loss": 0.5305, + "loss": 1.0289, "step": 41160 }, { "epoch": 4.35, - "grad_norm": 12.49261474609375, + "grad_norm": 39.34114074707031, "learning_rate": 1.1304255993241105e-05, - "loss": 0.4914, + "loss": 0.8963, "step": 41170 }, { "epoch": 4.35, - "grad_norm": 8.345911026000977, + "grad_norm": 50.5380859375, "learning_rate": 1.1302143837786462e-05, - "loss": 0.6155, + "loss": 0.8847, "step": 41180 }, { "epoch": 4.35, - "grad_norm": 21.891223907470703, + "grad_norm": 37.088565826416016, "learning_rate": 1.130003168233182e-05, - "loss": 0.3163, + "loss": 1.0534, "step": 41190 }, { "epoch": 4.35, - "grad_norm": 12.112762451171875, + "grad_norm": 52.15786361694336, "learning_rate": 1.129791952687718e-05, - "loss": 0.4872, + "loss": 1.2702, "step": 41200 }, { "epoch": 4.35, - "grad_norm": 1.9419426918029785, + "grad_norm": 33.21182632446289, "learning_rate": 1.1295807371422538e-05, - "loss": 0.6143, + "loss": 1.1895, "step": 41210 }, { "epoch": 4.35, - "grad_norm": 2.583125114440918, + "grad_norm": 26.293197631835938, "learning_rate": 1.1293695215967897e-05, - "loss": 0.7465, + "loss": 1.0984, "step": 41220 }, { "epoch": 4.35, - "grad_norm": 10.761408805847168, + "grad_norm": 33.662349700927734, "learning_rate": 1.1291583060513254e-05, - "loss": 0.4167, + "loss": 0.8348, "step": 41230 }, { "epoch": 4.36, - "grad_norm": 3.661087989807129, + "grad_norm": 33.464500427246094, "learning_rate": 1.1289470905058613e-05, - "loss": 0.4082, + "loss": 1.0395, "step": 41240 }, { "epoch": 4.36, - "grad_norm": 16.02431869506836, + "grad_norm": 36.64920425415039, "learning_rate": 1.1287358749603972e-05, - "loss": 0.4861, + "loss": 1.0187, "step": 41250 }, { "epoch": 4.36, - "grad_norm": 10.548453330993652, + "grad_norm": 54.104122161865234, "learning_rate": 1.128524659414933e-05, - "loss": 0.2968, + "loss": 1.2768, "step": 41260 }, { "epoch": 4.36, - "grad_norm": 13.695853233337402, + "grad_norm": 35.63486862182617, "learning_rate": 1.128313443869469e-05, - "loss": 0.474, + "loss": 1.0219, "step": 41270 }, { "epoch": 4.36, - "grad_norm": 7.760513782501221, + "grad_norm": 36.55010986328125, "learning_rate": 1.1281022283240048e-05, - "loss": 0.5335, + "loss": 1.0663, "step": 41280 }, { "epoch": 4.36, - "grad_norm": 8.376749038696289, + "grad_norm": 58.15953826904297, "learning_rate": 1.1278910127785405e-05, - "loss": 0.6289, + "loss": 1.2738, "step": 41290 }, { "epoch": 4.36, - "grad_norm": 12.490745544433594, + "grad_norm": 37.251163482666016, "learning_rate": 1.1276797972330764e-05, - "loss": 0.3897, + "loss": 1.0502, "step": 41300 }, { "epoch": 4.36, - "grad_norm": 22.60707664489746, + "grad_norm": 15.219940185546875, "learning_rate": 1.1274685816876123e-05, - "loss": 0.463, + "loss": 1.0406, "step": 41310 }, { "epoch": 4.36, - "grad_norm": 0.20615100860595703, + "grad_norm": 31.146038055419922, "learning_rate": 1.1272573661421481e-05, - "loss": 0.5211, + "loss": 0.8827, "step": 41320 }, { "epoch": 4.36, - "grad_norm": 26.111886978149414, + "grad_norm": 42.40706253051758, "learning_rate": 1.127046150596684e-05, - "loss": 0.553, + "loss": 1.1288, "step": 41330 }, { "epoch": 4.37, - "grad_norm": 18.174386978149414, + "grad_norm": 26.129405975341797, "learning_rate": 1.1268349350512197e-05, - "loss": 0.6867, + "loss": 1.0658, "step": 41340 }, { "epoch": 4.37, - "grad_norm": 7.6533684730529785, + "grad_norm": 30.899812698364258, "learning_rate": 1.1266237195057556e-05, - "loss": 0.5499, + "loss": 1.2646, "step": 41350 }, { "epoch": 4.37, - "grad_norm": 23.221641540527344, + "grad_norm": 40.62495422363281, "learning_rate": 1.1264125039602915e-05, - "loss": 0.5013, + "loss": 1.0196, "step": 41360 }, { "epoch": 4.37, - "grad_norm": 11.6826753616333, + "grad_norm": 43.509552001953125, "learning_rate": 1.1262012884148275e-05, - "loss": 0.3859, + "loss": 1.4302, "step": 41370 }, { "epoch": 4.37, - "grad_norm": 13.368050575256348, + "grad_norm": 51.65105438232422, "learning_rate": 1.1259900728693634e-05, - "loss": 0.3117, + "loss": 1.0815, "step": 41380 }, { "epoch": 4.37, - "grad_norm": 12.438891410827637, + "grad_norm": 42.06118392944336, "learning_rate": 1.125778857323899e-05, - "loss": 0.6678, + "loss": 1.4684, "step": 41390 }, { "epoch": 4.37, - "grad_norm": 27.673015594482422, + "grad_norm": 51.26124572753906, "learning_rate": 1.125567641778435e-05, - "loss": 0.547, + "loss": 1.2897, "step": 41400 }, { "epoch": 4.37, - "grad_norm": 28.21664047241211, + "grad_norm": 40.218318939208984, "learning_rate": 1.1253564262329709e-05, - "loss": 0.3026, + "loss": 0.9612, "step": 41410 }, { "epoch": 4.37, - "grad_norm": 26.47369384765625, + "grad_norm": 47.307552337646484, "learning_rate": 1.1251452106875068e-05, - "loss": 0.4844, + "loss": 1.1298, "step": 41420 }, { "epoch": 4.38, - "grad_norm": 5.080040454864502, + "grad_norm": 45.061126708984375, "learning_rate": 1.1249339951420426e-05, - "loss": 0.4314, + "loss": 1.2226, "step": 41430 }, { "epoch": 4.38, - "grad_norm": 1.1098967790603638, + "grad_norm": 25.975168228149414, "learning_rate": 1.1247227795965785e-05, - "loss": 0.3236, + "loss": 1.0225, "step": 41440 }, { "epoch": 4.38, - "grad_norm": 22.04336166381836, + "grad_norm": 42.703224182128906, "learning_rate": 1.1245115640511142e-05, - "loss": 0.5653, + "loss": 1.0622, "step": 41450 }, { "epoch": 4.38, - "grad_norm": 9.968018531799316, + "grad_norm": 24.666532516479492, "learning_rate": 1.1243003485056501e-05, - "loss": 0.4278, + "loss": 1.4089, "step": 41460 }, { "epoch": 4.38, - "grad_norm": 0.7384079694747925, + "grad_norm": 42.11921310424805, "learning_rate": 1.124089132960186e-05, - "loss": 0.3267, + "loss": 0.9742, "step": 41470 }, { "epoch": 4.38, - "grad_norm": 22.541095733642578, + "grad_norm": 29.912330627441406, "learning_rate": 1.1238779174147218e-05, - "loss": 0.6788, + "loss": 1.388, "step": 41480 }, { "epoch": 4.38, - "grad_norm": 9.275505065917969, + "grad_norm": 28.530946731567383, "learning_rate": 1.1236667018692577e-05, - "loss": 0.6618, + "loss": 1.0496, "step": 41490 }, { "epoch": 4.38, - "grad_norm": 20.645666122436523, + "grad_norm": 40.05908203125, "learning_rate": 1.1234554863237934e-05, - "loss": 0.428, + "loss": 1.024, "step": 41500 }, { "epoch": 4.38, - "grad_norm": 15.5033597946167, + "grad_norm": 31.09604263305664, "learning_rate": 1.1232442707783293e-05, - "loss": 0.3635, + "loss": 1.0399, "step": 41510 }, { "epoch": 4.38, - "grad_norm": 36.63276672363281, + "grad_norm": 43.088050842285156, "learning_rate": 1.1230330552328652e-05, - "loss": 0.3644, + "loss": 0.9469, "step": 41520 }, { "epoch": 4.39, - "grad_norm": 15.404324531555176, + "grad_norm": 38.06031036376953, "learning_rate": 1.122821839687401e-05, - "loss": 0.6008, + "loss": 1.5947, "step": 41530 }, { "epoch": 4.39, - "grad_norm": 8.564438819885254, + "grad_norm": 52.242820739746094, "learning_rate": 1.122610624141937e-05, - "loss": 0.2262, + "loss": 1.0094, "step": 41540 }, { "epoch": 4.39, - "grad_norm": 9.380447387695312, + "grad_norm": 42.29734420776367, "learning_rate": 1.1223994085964728e-05, - "loss": 0.5064, + "loss": 1.0356, "step": 41550 }, { "epoch": 4.39, - "grad_norm": 9.95979118347168, + "grad_norm": 57.619720458984375, "learning_rate": 1.1221881930510085e-05, - "loss": 0.5004, + "loss": 1.2052, "step": 41560 }, { "epoch": 4.39, - "grad_norm": 0.1752442568540573, + "grad_norm": 19.854141235351562, "learning_rate": 1.1219769775055444e-05, - "loss": 0.3449, + "loss": 1.0558, "step": 41570 }, { "epoch": 4.39, - "grad_norm": 22.361644744873047, + "grad_norm": 49.67558288574219, "learning_rate": 1.1217657619600803e-05, - "loss": 0.5952, + "loss": 1.2756, "step": 41580 }, { "epoch": 4.39, - "grad_norm": 51.9811897277832, + "grad_norm": 54.566192626953125, "learning_rate": 1.1215545464146163e-05, - "loss": 0.551, + "loss": 1.0836, "step": 41590 }, { "epoch": 4.39, - "grad_norm": 1.5617400407791138, + "grad_norm": 32.72689437866211, "learning_rate": 1.1213433308691522e-05, - "loss": 0.4909, + "loss": 1.0026, "step": 41600 }, { "epoch": 4.39, - "grad_norm": 6.2026777267456055, + "grad_norm": 42.64364242553711, "learning_rate": 1.1211321153236878e-05, - "loss": 0.6745, + "loss": 1.1693, "step": 41610 }, { "epoch": 4.4, - "grad_norm": 1.6165006160736084, + "grad_norm": 30.08846664428711, "learning_rate": 1.1209208997782236e-05, - "loss": 0.5375, + "loss": 1.1638, "step": 41620 }, { "epoch": 4.4, - "grad_norm": 21.67116928100586, + "grad_norm": 38.81419372558594, "learning_rate": 1.1207096842327597e-05, - "loss": 0.4924, + "loss": 0.9464, "step": 41630 }, { "epoch": 4.4, - "grad_norm": 16.757802963256836, + "grad_norm": 44.95380783081055, "learning_rate": 1.1204984686872956e-05, - "loss": 0.2692, + "loss": 1.0378, "step": 41640 }, { "epoch": 4.4, - "grad_norm": 1.7380852699279785, + "grad_norm": 58.23954391479492, "learning_rate": 1.1202872531418314e-05, - "loss": 0.4871, + "loss": 1.4978, "step": 41650 }, { "epoch": 4.4, - "grad_norm": 15.427213668823242, + "grad_norm": 46.46141052246094, "learning_rate": 1.1200760375963673e-05, - "loss": 0.8047, + "loss": 1.1096, "step": 41660 }, { "epoch": 4.4, - "grad_norm": 11.985939979553223, + "grad_norm": 52.24104309082031, "learning_rate": 1.119864822050903e-05, - "loss": 0.9185, + "loss": 1.2122, "step": 41670 }, { "epoch": 4.4, - "grad_norm": 8.219449043273926, + "grad_norm": 39.35095977783203, "learning_rate": 1.1196536065054389e-05, - "loss": 0.5437, + "loss": 1.2385, "step": 41680 }, { "epoch": 4.4, - "grad_norm": 0.15382972359657288, + "grad_norm": 18.601333618164062, "learning_rate": 1.1194423909599748e-05, - "loss": 0.3926, + "loss": 0.9077, "step": 41690 }, { "epoch": 4.4, - "grad_norm": 13.84500503540039, + "grad_norm": 32.53806686401367, "learning_rate": 1.1192311754145107e-05, - "loss": 0.3964, + "loss": 0.958, "step": 41700 }, { "epoch": 4.4, - "grad_norm": 44.84720993041992, + "grad_norm": 34.74462890625, "learning_rate": 1.1190199598690465e-05, - "loss": 0.3446, + "loss": 0.9892, "step": 41710 }, { "epoch": 4.41, - "grad_norm": 15.584357261657715, + "grad_norm": 35.432430267333984, "learning_rate": 1.1188087443235822e-05, - "loss": 0.3364, + "loss": 0.8085, "step": 41720 }, { "epoch": 4.41, - "grad_norm": 11.45728588104248, + "grad_norm": 36.50278091430664, "learning_rate": 1.1185975287781181e-05, - "loss": 0.7363, + "loss": 1.0547, "step": 41730 }, { "epoch": 4.41, - "grad_norm": 4.379489898681641, + "grad_norm": 42.99654006958008, "learning_rate": 1.118386313232654e-05, - "loss": 0.4495, + "loss": 1.0428, "step": 41740 }, { "epoch": 4.41, - "grad_norm": 1.71314537525177, + "grad_norm": 39.799095153808594, "learning_rate": 1.1181750976871899e-05, - "loss": 0.3522, + "loss": 0.869, "step": 41750 }, { "epoch": 4.41, - "grad_norm": 10.528018951416016, + "grad_norm": 43.34877395629883, "learning_rate": 1.1179638821417258e-05, - "loss": 0.4532, + "loss": 1.0377, "step": 41760 }, { "epoch": 4.41, - "grad_norm": 9.885315895080566, + "grad_norm": 28.137046813964844, "learning_rate": 1.1177526665962615e-05, - "loss": 0.635, + "loss": 0.9619, "step": 41770 }, { "epoch": 4.41, - "grad_norm": 19.938919067382812, + "grad_norm": 45.54253387451172, "learning_rate": 1.1175414510507973e-05, - "loss": 0.5381, + "loss": 1.1005, "step": 41780 }, { "epoch": 4.41, - "grad_norm": 19.70885467529297, + "grad_norm": 34.589942932128906, "learning_rate": 1.1173302355053332e-05, - "loss": 0.5893, + "loss": 0.8141, "step": 41790 }, { "epoch": 4.41, - "grad_norm": 16.100553512573242, + "grad_norm": 34.8029899597168, "learning_rate": 1.1171190199598691e-05, - "loss": 0.7495, + "loss": 1.2015, "step": 41800 }, { "epoch": 4.42, - "grad_norm": 31.118579864501953, + "grad_norm": 25.517345428466797, "learning_rate": 1.116907804414405e-05, - "loss": 0.5314, + "loss": 1.1522, "step": 41810 }, { "epoch": 4.42, - "grad_norm": 4.022888660430908, + "grad_norm": 41.27442932128906, "learning_rate": 1.116696588868941e-05, - "loss": 0.4596, + "loss": 0.9013, "step": 41820 }, { "epoch": 4.42, - "grad_norm": 11.844586372375488, + "grad_norm": 20.80809211730957, "learning_rate": 1.1164853733234766e-05, - "loss": 0.6456, + "loss": 0.8979, "step": 41830 }, { "epoch": 4.42, - "grad_norm": 17.495548248291016, + "grad_norm": 43.66254806518555, "learning_rate": 1.1162741577780124e-05, - "loss": 0.5721, + "loss": 0.9851, "step": 41840 }, { "epoch": 4.42, - "grad_norm": 16.827491760253906, + "grad_norm": 42.87838363647461, "learning_rate": 1.1160629422325485e-05, - "loss": 0.368, + "loss": 1.0826, "step": 41850 }, { "epoch": 4.42, - "grad_norm": 2.1316843032836914, + "grad_norm": 26.30841064453125, "learning_rate": 1.1158517266870844e-05, - "loss": 0.2514, + "loss": 0.7266, "step": 41860 }, { "epoch": 4.42, - "grad_norm": 10.04633903503418, + "grad_norm": 59.67578125, "learning_rate": 1.1156405111416202e-05, - "loss": 0.7301, + "loss": 1.0985, "step": 41870 }, { "epoch": 4.42, - "grad_norm": 21.43711280822754, + "grad_norm": 37.390403747558594, "learning_rate": 1.115429295596156e-05, - "loss": 0.603, + "loss": 1.416, "step": 41880 }, { "epoch": 4.42, - "grad_norm": 5.244923114776611, + "grad_norm": 55.10600662231445, "learning_rate": 1.1152180800506918e-05, - "loss": 0.4925, + "loss": 1.1534, "step": 41890 }, { "epoch": 4.42, - "grad_norm": 4.525173664093018, + "grad_norm": 44.363094329833984, "learning_rate": 1.1150068645052277e-05, - "loss": 0.4728, + "loss": 0.9358, "step": 41900 }, { "epoch": 4.43, - "grad_norm": 2.1385653018951416, + "grad_norm": 16.25351905822754, "learning_rate": 1.1147956489597636e-05, - "loss": 0.5918, + "loss": 1.2041, "step": 41910 }, { "epoch": 4.43, - "grad_norm": 7.989027976989746, + "grad_norm": 35.03660202026367, "learning_rate": 1.1145844334142995e-05, - "loss": 0.5679, + "loss": 1.4245, "step": 41920 }, { "epoch": 4.43, - "grad_norm": 18.930185317993164, + "grad_norm": 28.65758514404297, "learning_rate": 1.1143732178688353e-05, - "loss": 0.4585, + "loss": 1.1312, "step": 41930 }, { "epoch": 4.43, - "grad_norm": 15.145203590393066, + "grad_norm": 45.14099884033203, "learning_rate": 1.114162002323371e-05, - "loss": 0.4558, + "loss": 1.1779, "step": 41940 }, { "epoch": 4.43, - "grad_norm": 15.772823333740234, + "grad_norm": 40.573280334472656, "learning_rate": 1.113950786777907e-05, - "loss": 0.4101, + "loss": 1.1684, "step": 41950 }, { "epoch": 4.43, - "grad_norm": 26.65981101989746, + "grad_norm": 42.90957260131836, "learning_rate": 1.1137395712324428e-05, - "loss": 0.6091, + "loss": 0.8976, "step": 41960 }, { "epoch": 4.43, - "grad_norm": 11.21568489074707, + "grad_norm": 64.14862060546875, "learning_rate": 1.1135283556869787e-05, - "loss": 0.6168, + "loss": 1.0268, "step": 41970 }, { "epoch": 4.43, - "grad_norm": 19.12699317932129, + "grad_norm": 43.89265823364258, "learning_rate": 1.1133171401415146e-05, - "loss": 0.5399, + "loss": 1.3171, "step": 41980 }, { "epoch": 4.43, - "grad_norm": 60.99302291870117, + "grad_norm": 45.24525451660156, "learning_rate": 1.1131059245960503e-05, - "loss": 0.7369, + "loss": 1.2853, "step": 41990 }, { "epoch": 4.44, - "grad_norm": 8.060282707214355, + "grad_norm": 39.09378433227539, "learning_rate": 1.1128947090505861e-05, - "loss": 0.5763, + "loss": 1.43, "step": 42000 }, { "epoch": 4.44, - "grad_norm": 6.701888084411621, + "grad_norm": 46.073307037353516, "learning_rate": 1.112683493505122e-05, - "loss": 0.8622, + "loss": 1.3955, "step": 42010 }, { "epoch": 4.44, - "grad_norm": 11.009459495544434, + "grad_norm": 28.533824920654297, "learning_rate": 1.1124722779596579e-05, - "loss": 0.3925, + "loss": 1.0569, "step": 42020 }, { "epoch": 4.44, - "grad_norm": 30.48014259338379, + "grad_norm": 44.11210250854492, "learning_rate": 1.1122610624141938e-05, - "loss": 1.0384, + "loss": 1.2787, "step": 42030 }, { "epoch": 4.44, - "grad_norm": 24.829269409179688, + "grad_norm": 57.59354019165039, "learning_rate": 1.1120498468687295e-05, - "loss": 0.4638, + "loss": 0.9059, "step": 42040 }, { "epoch": 4.44, - "grad_norm": 25.766761779785156, + "grad_norm": 73.13959503173828, "learning_rate": 1.1118386313232654e-05, - "loss": 0.491, + "loss": 1.1884, "step": 42050 }, { "epoch": 4.44, - "grad_norm": 6.629104137420654, + "grad_norm": 21.34298324584961, "learning_rate": 1.1116274157778012e-05, - "loss": 0.4997, + "loss": 0.941, "step": 42060 }, { "epoch": 4.44, - "grad_norm": 20.472862243652344, + "grad_norm": 63.521087646484375, "learning_rate": 1.1114162002323373e-05, - "loss": 0.6958, + "loss": 1.3462, "step": 42070 }, { "epoch": 4.44, - "grad_norm": 29.35385513305664, + "grad_norm": 37.42076873779297, "learning_rate": 1.1112049846868732e-05, - "loss": 0.4606, + "loss": 1.1194, "step": 42080 }, { "epoch": 4.45, - "grad_norm": 20.2569637298584, + "grad_norm": 32.53227233886719, "learning_rate": 1.110993769141409e-05, - "loss": 0.6541, + "loss": 1.0733, "step": 42090 }, { "epoch": 4.45, - "grad_norm": 14.680954933166504, + "grad_norm": 36.6758918762207, "learning_rate": 1.1107825535959448e-05, - "loss": 0.48, + "loss": 1.084, "step": 42100 }, { "epoch": 4.45, - "grad_norm": 17.501205444335938, + "grad_norm": 47.25674057006836, "learning_rate": 1.1105713380504806e-05, - "loss": 0.4576, + "loss": 0.9887, "step": 42110 }, { "epoch": 4.45, - "grad_norm": 10.396456718444824, + "grad_norm": 25.13741111755371, "learning_rate": 1.1103601225050165e-05, - "loss": 0.7335, + "loss": 1.1796, "step": 42120 }, { "epoch": 4.45, - "grad_norm": 9.310958862304688, + "grad_norm": 31.41295623779297, "learning_rate": 1.1101489069595524e-05, - "loss": 0.4455, + "loss": 1.2557, "step": 42130 }, { "epoch": 4.45, - "grad_norm": 4.332348346710205, + "grad_norm": 34.010528564453125, "learning_rate": 1.1099376914140883e-05, - "loss": 0.5128, + "loss": 1.1233, "step": 42140 }, { "epoch": 4.45, - "grad_norm": 2.919313907623291, + "grad_norm": 39.946868896484375, "learning_rate": 1.109726475868624e-05, - "loss": 0.2166, + "loss": 0.9014, "step": 42150 }, { "epoch": 4.45, - "grad_norm": 17.495689392089844, + "grad_norm": 45.892547607421875, "learning_rate": 1.1095152603231599e-05, - "loss": 0.4298, + "loss": 0.8536, "step": 42160 }, { "epoch": 4.45, - "grad_norm": 3.4215407371520996, + "grad_norm": 27.994604110717773, "learning_rate": 1.1093040447776957e-05, - "loss": 0.6487, + "loss": 1.253, "step": 42170 }, { "epoch": 4.45, - "grad_norm": 7.837536334991455, + "grad_norm": 24.071910858154297, "learning_rate": 1.1090928292322316e-05, - "loss": 0.5613, + "loss": 1.1671, "step": 42180 }, { "epoch": 4.46, - "grad_norm": 22.585758209228516, + "grad_norm": 62.378597259521484, "learning_rate": 1.1088816136867675e-05, - "loss": 0.6109, + "loss": 1.17, "step": 42190 }, { "epoch": 4.46, - "grad_norm": 38.78850173950195, + "grad_norm": 61.9403190612793, "learning_rate": 1.1086703981413034e-05, - "loss": 0.2423, + "loss": 0.9542, "step": 42200 }, { "epoch": 4.46, - "grad_norm": 29.843774795532227, + "grad_norm": 23.059066772460938, "learning_rate": 1.108459182595839e-05, - "loss": 0.2969, + "loss": 1.049, "step": 42210 }, { "epoch": 4.46, - "grad_norm": 12.043839454650879, + "grad_norm": 41.707584381103516, "learning_rate": 1.108247967050375e-05, - "loss": 0.5367, + "loss": 1.1432, "step": 42220 }, { "epoch": 4.46, - "grad_norm": 15.397261619567871, + "grad_norm": 48.32080841064453, "learning_rate": 1.1080367515049108e-05, - "loss": 0.604, + "loss": 0.9715, "step": 42230 }, { "epoch": 4.46, - "grad_norm": 11.042777061462402, + "grad_norm": 37.643123626708984, "learning_rate": 1.1078255359594467e-05, - "loss": 0.5732, + "loss": 0.9576, "step": 42240 }, { "epoch": 4.46, - "grad_norm": 2.1700704097747803, + "grad_norm": 29.69121551513672, "learning_rate": 1.1076143204139826e-05, - "loss": 0.5496, + "loss": 1.1628, "step": 42250 }, { "epoch": 4.46, - "grad_norm": 21.012840270996094, + "grad_norm": 52.069908142089844, "learning_rate": 1.1074031048685183e-05, - "loss": 0.5868, + "loss": 1.4198, "step": 42260 }, { "epoch": 4.46, - "grad_norm": 14.734301567077637, + "grad_norm": 41.68473815917969, "learning_rate": 1.1071918893230542e-05, - "loss": 0.6793, + "loss": 1.1002, "step": 42270 }, { "epoch": 4.47, - "grad_norm": 68.29378509521484, + "grad_norm": 32.18399429321289, "learning_rate": 1.10698067377759e-05, - "loss": 0.4504, + "loss": 1.2708, "step": 42280 }, { "epoch": 4.47, - "grad_norm": 14.153655052185059, + "grad_norm": 28.059555053710938, "learning_rate": 1.106769458232126e-05, - "loss": 0.4362, + "loss": 1.0265, "step": 42290 }, { "epoch": 4.47, - "grad_norm": 29.116018295288086, + "grad_norm": 41.42986297607422, "learning_rate": 1.106558242686662e-05, - "loss": 0.8954, + "loss": 1.3314, "step": 42300 }, { "epoch": 4.47, - "grad_norm": 15.20477294921875, + "grad_norm": 37.73373794555664, "learning_rate": 1.1063470271411978e-05, - "loss": 0.562, + "loss": 1.0721, "step": 42310 }, { "epoch": 4.47, - "grad_norm": 16.19892692565918, + "grad_norm": 41.03443145751953, "learning_rate": 1.1061358115957334e-05, - "loss": 0.3167, + "loss": 1.0488, "step": 42320 }, { "epoch": 4.47, - "grad_norm": 15.435072898864746, + "grad_norm": 50.828060150146484, "learning_rate": 1.1059245960502694e-05, - "loss": 0.4611, + "loss": 1.5554, "step": 42330 }, { "epoch": 4.47, - "grad_norm": 0.278535932302475, + "grad_norm": 26.465625762939453, "learning_rate": 1.1057133805048053e-05, - "loss": 0.689, + "loss": 0.9462, "step": 42340 }, { "epoch": 4.47, - "grad_norm": 25.0546932220459, + "grad_norm": 35.533653259277344, "learning_rate": 1.1055021649593412e-05, - "loss": 0.5987, + "loss": 1.2149, "step": 42350 }, { "epoch": 4.47, - "grad_norm": 16.81127166748047, + "grad_norm": 15.259601593017578, "learning_rate": 1.105290949413877e-05, - "loss": 0.3327, + "loss": 1.0707, "step": 42360 }, { "epoch": 4.47, - "grad_norm": 14.532692909240723, + "grad_norm": 33.13163757324219, "learning_rate": 1.1050797338684128e-05, - "loss": 0.4317, + "loss": 0.9047, "step": 42370 }, { "epoch": 4.48, - "grad_norm": 8.55321216583252, + "grad_norm": 46.30891799926758, "learning_rate": 1.1048685183229487e-05, - "loss": 0.2897, + "loss": 1.0031, "step": 42380 }, { "epoch": 4.48, - "grad_norm": 5.118427276611328, + "grad_norm": 34.00435256958008, "learning_rate": 1.1046573027774845e-05, - "loss": 0.2958, + "loss": 0.9264, "step": 42390 }, { "epoch": 4.48, - "grad_norm": 2.8060290813446045, + "grad_norm": 41.56221389770508, "learning_rate": 1.1044460872320204e-05, - "loss": 0.4017, + "loss": 0.7909, "step": 42400 }, { "epoch": 4.48, - "grad_norm": 4.772958755493164, + "grad_norm": 53.71671676635742, "learning_rate": 1.1042348716865563e-05, - "loss": 0.4491, + "loss": 1.1047, "step": 42410 }, { "epoch": 4.48, - "grad_norm": 8.27842903137207, + "grad_norm": 54.366981506347656, "learning_rate": 1.104023656141092e-05, - "loss": 0.2805, + "loss": 1.2101, "step": 42420 }, { "epoch": 4.48, - "grad_norm": 2.8489952087402344, + "grad_norm": 47.14077377319336, "learning_rate": 1.1038124405956279e-05, - "loss": 0.5472, + "loss": 1.2416, "step": 42430 }, { "epoch": 4.48, - "grad_norm": 16.18971824645996, + "grad_norm": 41.97098922729492, "learning_rate": 1.1036012250501638e-05, - "loss": 0.5018, + "loss": 1.071, "step": 42440 }, { "epoch": 4.48, - "grad_norm": 19.03404426574707, + "grad_norm": 33.535823822021484, "learning_rate": 1.1033900095046996e-05, - "loss": 0.7296, + "loss": 1.1344, "step": 42450 }, { "epoch": 4.48, - "grad_norm": 9.69471549987793, + "grad_norm": 35.271480560302734, "learning_rate": 1.1031787939592355e-05, - "loss": 0.4053, + "loss": 1.1577, "step": 42460 }, { "epoch": 4.49, - "grad_norm": 14.196464538574219, + "grad_norm": 22.921859741210938, "learning_rate": 1.1029675784137714e-05, - "loss": 0.4187, + "loss": 1.1702, "step": 42470 }, { "epoch": 4.49, - "grad_norm": 11.969006538391113, + "grad_norm": 37.64809036254883, "learning_rate": 1.1027563628683071e-05, - "loss": 0.5774, + "loss": 1.1918, "step": 42480 }, { "epoch": 4.49, - "grad_norm": 11.17210578918457, + "grad_norm": 38.8411865234375, "learning_rate": 1.102545147322843e-05, - "loss": 0.583, + "loss": 1.1656, "step": 42490 }, { "epoch": 4.49, - "grad_norm": 11.379728317260742, + "grad_norm": 36.83913803100586, "learning_rate": 1.1023339317773789e-05, - "loss": 0.4844, + "loss": 0.8294, "step": 42500 }, { "epoch": 4.49, - "grad_norm": 1.0755987167358398, + "grad_norm": 41.93242263793945, "learning_rate": 1.1021227162319147e-05, - "loss": 0.6789, + "loss": 0.998, "step": 42510 }, { "epoch": 4.49, - "grad_norm": 12.098112106323242, + "grad_norm": 16.704139709472656, "learning_rate": 1.1019115006864508e-05, - "loss": 0.4316, + "loss": 1.045, "step": 42520 }, { "epoch": 4.49, - "grad_norm": 10.99324893951416, + "grad_norm": 40.60194396972656, "learning_rate": 1.1017002851409863e-05, - "loss": 0.4906, + "loss": 1.0607, "step": 42530 }, { "epoch": 4.49, - "grad_norm": 19.57077980041504, + "grad_norm": 43.96855163574219, "learning_rate": 1.1014890695955222e-05, - "loss": 0.3258, + "loss": 0.9841, "step": 42540 }, { "epoch": 4.49, - "grad_norm": 19.37422752380371, + "grad_norm": 47.11967849731445, "learning_rate": 1.1012778540500582e-05, - "loss": 0.7331, + "loss": 1.0986, "step": 42550 }, { "epoch": 4.49, - "grad_norm": 0.05001607909798622, + "grad_norm": 8.514180183410645, "learning_rate": 1.1010666385045941e-05, - "loss": 0.4932, + "loss": 1.008, "step": 42560 }, { "epoch": 4.5, - "grad_norm": 25.425556182861328, + "grad_norm": 58.829856872558594, "learning_rate": 1.10085542295913e-05, - "loss": 0.5442, + "loss": 1.0803, "step": 42570 }, { "epoch": 4.5, - "grad_norm": 2.1994576454162598, + "grad_norm": 17.906536102294922, "learning_rate": 1.1006442074136659e-05, - "loss": 0.6877, + "loss": 1.5513, "step": 42580 }, { "epoch": 4.5, - "grad_norm": 18.713550567626953, + "grad_norm": 32.746978759765625, "learning_rate": 1.1004329918682016e-05, - "loss": 0.4646, + "loss": 0.9926, "step": 42590 }, { "epoch": 4.5, - "grad_norm": 18.45903968811035, + "grad_norm": 58.23380661010742, "learning_rate": 1.1002217763227375e-05, - "loss": 0.5574, + "loss": 1.1731, "step": 42600 }, { "epoch": 4.5, - "grad_norm": 19.232969284057617, + "grad_norm": 57.873931884765625, "learning_rate": 1.1000105607772733e-05, - "loss": 0.3906, + "loss": 1.0308, "step": 42610 }, { "epoch": 4.5, - "grad_norm": 15.513867378234863, + "grad_norm": 35.2074089050293, "learning_rate": 1.0997993452318092e-05, - "loss": 0.3796, + "loss": 1.1082, "step": 42620 }, { "epoch": 4.5, - "grad_norm": 17.703643798828125, + "grad_norm": 35.765769958496094, "learning_rate": 1.0995881296863451e-05, - "loss": 0.5303, + "loss": 1.0202, "step": 42630 }, { "epoch": 4.5, - "grad_norm": 21.446287155151367, + "grad_norm": 36.505374908447266, "learning_rate": 1.0993769141408808e-05, - "loss": 0.4369, + "loss": 0.9538, "step": 42640 }, { "epoch": 4.5, - "grad_norm": 3.9252421855926514, + "grad_norm": 33.29825973510742, "learning_rate": 1.0991656985954167e-05, - "loss": 0.4833, + "loss": 1.1167, "step": 42650 }, { "epoch": 4.51, - "grad_norm": 7.736690521240234, + "grad_norm": 29.32179832458496, "learning_rate": 1.0989544830499526e-05, - "loss": 0.5573, + "loss": 0.9386, "step": 42660 }, { "epoch": 4.51, - "grad_norm": 11.626920700073242, + "grad_norm": 24.5375919342041, "learning_rate": 1.0987432675044884e-05, - "loss": 0.8743, + "loss": 1.3623, "step": 42670 }, { "epoch": 4.51, - "grad_norm": 15.443472862243652, + "grad_norm": 40.12348556518555, "learning_rate": 1.0985320519590243e-05, - "loss": 0.6276, + "loss": 1.2368, "step": 42680 }, { "epoch": 4.51, - "grad_norm": 27.64293670654297, + "grad_norm": 32.47996520996094, "learning_rate": 1.09832083641356e-05, - "loss": 0.5508, + "loss": 1.1071, "step": 42690 }, { "epoch": 4.51, - "grad_norm": 30.854347229003906, + "grad_norm": 36.21152877807617, "learning_rate": 1.0981096208680959e-05, - "loss": 0.6491, + "loss": 1.5435, "step": 42700 }, { "epoch": 4.51, - "grad_norm": 8.19219970703125, + "grad_norm": 59.51308059692383, "learning_rate": 1.0978984053226318e-05, - "loss": 0.2748, + "loss": 0.9053, "step": 42710 }, { "epoch": 4.51, - "grad_norm": 0.14188340306282043, + "grad_norm": 25.218494415283203, "learning_rate": 1.0976871897771677e-05, - "loss": 0.3602, + "loss": 1.226, "step": 42720 }, { "epoch": 4.51, - "grad_norm": 4.059193134307861, + "grad_norm": 30.667226791381836, "learning_rate": 1.0974759742317035e-05, - "loss": 0.3385, + "loss": 1.1789, "step": 42730 }, { "epoch": 4.51, - "grad_norm": 17.789243698120117, + "grad_norm": 40.18147277832031, "learning_rate": 1.0972647586862394e-05, - "loss": 0.3422, + "loss": 1.2173, "step": 42740 }, { "epoch": 4.51, - "grad_norm": 7.30971097946167, + "grad_norm": 49.5537223815918, "learning_rate": 1.0970535431407751e-05, - "loss": 0.5756, + "loss": 0.9803, "step": 42750 }, { "epoch": 4.52, - "grad_norm": 0.9313506484031677, + "grad_norm": 30.732147216796875, "learning_rate": 1.096842327595311e-05, - "loss": 0.3659, + "loss": 0.9668, "step": 42760 }, { "epoch": 4.52, - "grad_norm": 21.35964584350586, + "grad_norm": 31.841981887817383, "learning_rate": 1.0966311120498469e-05, - "loss": 0.5406, + "loss": 1.0339, "step": 42770 }, { "epoch": 4.52, - "grad_norm": 13.799898147583008, + "grad_norm": 53.64693832397461, "learning_rate": 1.096419896504383e-05, - "loss": 0.5929, + "loss": 1.2088, "step": 42780 }, { "epoch": 4.52, - "grad_norm": 20.398273468017578, + "grad_norm": 23.804224014282227, "learning_rate": 1.0962086809589188e-05, - "loss": 0.5642, + "loss": 1.321, "step": 42790 }, { "epoch": 4.52, - "grad_norm": 10.227837562561035, + "grad_norm": 44.90458297729492, "learning_rate": 1.0959974654134543e-05, - "loss": 0.4434, + "loss": 1.4892, "step": 42800 }, { "epoch": 4.52, - "grad_norm": 24.199331283569336, + "grad_norm": 51.0438346862793, "learning_rate": 1.0957862498679904e-05, - "loss": 0.2653, + "loss": 0.9323, "step": 42810 }, { "epoch": 4.52, - "grad_norm": 10.065442085266113, + "grad_norm": 42.21614074707031, "learning_rate": 1.0955750343225263e-05, - "loss": 0.3573, + "loss": 1.0059, "step": 42820 }, { "epoch": 4.52, - "grad_norm": 1.4634449481964111, + "grad_norm": 4.25229024887085, "learning_rate": 1.0953638187770621e-05, - "loss": 0.3654, + "loss": 0.9287, "step": 42830 }, { "epoch": 4.52, - "grad_norm": 3.983048915863037, + "grad_norm": 55.23942947387695, "learning_rate": 1.095152603231598e-05, - "loss": 0.2526, + "loss": 1.1502, "step": 42840 }, { "epoch": 4.53, - "grad_norm": 19.711055755615234, + "grad_norm": 67.37933349609375, "learning_rate": 1.0949413876861339e-05, - "loss": 0.5607, + "loss": 1.2252, "step": 42850 }, { "epoch": 4.53, - "grad_norm": 22.0225887298584, + "grad_norm": 47.22565841674805, "learning_rate": 1.0947301721406696e-05, - "loss": 0.6097, + "loss": 1.3197, "step": 42860 }, { "epoch": 4.53, - "grad_norm": 11.085050582885742, + "grad_norm": 43.16935729980469, "learning_rate": 1.0945189565952055e-05, - "loss": 0.2834, + "loss": 1.106, "step": 42870 }, { "epoch": 4.53, - "grad_norm": 14.855319023132324, + "grad_norm": 39.64091873168945, "learning_rate": 1.0943077410497414e-05, - "loss": 0.3594, + "loss": 1.2458, "step": 42880 }, { "epoch": 4.53, - "grad_norm": 20.669342041015625, + "grad_norm": 41.434295654296875, "learning_rate": 1.0940965255042772e-05, - "loss": 0.789, + "loss": 1.3474, "step": 42890 }, { "epoch": 4.53, - "grad_norm": 6.809144973754883, + "grad_norm": 49.573280334472656, "learning_rate": 1.0938853099588131e-05, - "loss": 0.4081, + "loss": 1.3029, "step": 42900 }, { "epoch": 4.53, - "grad_norm": 9.766159057617188, + "grad_norm": 54.53289794921875, "learning_rate": 1.0936740944133488e-05, - "loss": 0.4714, + "loss": 0.8404, "step": 42910 }, { "epoch": 4.53, - "grad_norm": 13.160305976867676, + "grad_norm": 43.77409362792969, "learning_rate": 1.0934628788678847e-05, - "loss": 0.4432, + "loss": 1.3135, "step": 42920 }, { "epoch": 4.53, - "grad_norm": 11.631124496459961, + "grad_norm": 43.56836700439453, "learning_rate": 1.0932516633224206e-05, - "loss": 0.4966, + "loss": 1.1954, "step": 42930 }, { "epoch": 4.53, - "grad_norm": 2.8981053829193115, + "grad_norm": 40.16180419921875, "learning_rate": 1.0930404477769565e-05, - "loss": 0.4681, + "loss": 1.1079, "step": 42940 }, { "epoch": 4.54, - "grad_norm": 3.0212562084198, + "grad_norm": 30.026485443115234, "learning_rate": 1.0928292322314923e-05, - "loss": 0.5684, + "loss": 1.0296, "step": 42950 }, { "epoch": 4.54, - "grad_norm": 30.537954330444336, + "grad_norm": 39.36684799194336, "learning_rate": 1.092618016686028e-05, - "loss": 0.9613, + "loss": 1.4396, "step": 42960 }, { "epoch": 4.54, - "grad_norm": 21.33808708190918, + "grad_norm": 36.73345947265625, "learning_rate": 1.092406801140564e-05, - "loss": 0.4003, + "loss": 0.7692, "step": 42970 }, { "epoch": 4.54, - "grad_norm": 17.31436538696289, + "grad_norm": 57.33640670776367, "learning_rate": 1.0921955855950998e-05, - "loss": 0.5423, + "loss": 1.3316, "step": 42980 }, { "epoch": 4.54, - "grad_norm": 30.97877311706543, + "grad_norm": 32.934242248535156, "learning_rate": 1.0919843700496357e-05, - "loss": 0.4249, + "loss": 0.7861, "step": 42990 }, { "epoch": 4.54, - "grad_norm": 14.521710395812988, + "grad_norm": 37.883296966552734, "learning_rate": 1.0917731545041717e-05, - "loss": 0.3715, + "loss": 1.2995, "step": 43000 }, { "epoch": 4.54, - "grad_norm": 14.770068168640137, + "grad_norm": 53.166481018066406, "learning_rate": 1.0915619389587076e-05, - "loss": 0.6072, + "loss": 1.2526, "step": 43010 }, { "epoch": 4.54, - "grad_norm": 6.7977728843688965, + "grad_norm": 37.310237884521484, "learning_rate": 1.0913507234132431e-05, - "loss": 0.3474, + "loss": 1.0001, "step": 43020 }, { "epoch": 4.54, - "grad_norm": 47.92343521118164, + "grad_norm": 31.705982208251953, "learning_rate": 1.0911395078677792e-05, - "loss": 0.6207, + "loss": 1.3044, "step": 43030 }, { "epoch": 4.55, - "grad_norm": 9.091963768005371, + "grad_norm": 40.10728073120117, "learning_rate": 1.090928292322315e-05, - "loss": 0.4529, + "loss": 1.1026, "step": 43040 }, { "epoch": 4.55, - "grad_norm": 7.110542297363281, + "grad_norm": 23.177194595336914, "learning_rate": 1.090717076776851e-05, - "loss": 0.2102, + "loss": 0.9703, "step": 43050 }, { "epoch": 4.55, - "grad_norm": 6.799439430236816, + "grad_norm": 28.548397064208984, "learning_rate": 1.0905058612313868e-05, - "loss": 0.4392, + "loss": 1.093, "step": 43060 }, { "epoch": 4.55, - "grad_norm": 0.6805613040924072, + "grad_norm": 46.49765396118164, "learning_rate": 1.0902946456859225e-05, - "loss": 0.2854, + "loss": 1.0232, "step": 43070 }, { "epoch": 4.55, - "grad_norm": 19.168521881103516, + "grad_norm": 31.445405960083008, "learning_rate": 1.0900834301404584e-05, - "loss": 0.3539, + "loss": 0.899, "step": 43080 }, { "epoch": 4.55, - "grad_norm": 9.893290519714355, + "grad_norm": 41.27781677246094, "learning_rate": 1.0898722145949943e-05, - "loss": 0.3189, + "loss": 1.2708, "step": 43090 }, { "epoch": 4.55, - "grad_norm": 6.785530090332031, + "grad_norm": 47.54965591430664, "learning_rate": 1.0896609990495302e-05, - "loss": 0.1949, + "loss": 0.8676, "step": 43100 }, { "epoch": 4.55, - "grad_norm": 20.604228973388672, + "grad_norm": 62.988277435302734, "learning_rate": 1.089449783504066e-05, - "loss": 0.5305, + "loss": 1.0659, "step": 43110 }, { "epoch": 4.55, - "grad_norm": 25.3575439453125, + "grad_norm": 43.74455261230469, "learning_rate": 1.089238567958602e-05, - "loss": 0.6093, + "loss": 1.0437, "step": 43120 }, { "epoch": 4.55, - "grad_norm": 25.037260055541992, + "grad_norm": 47.66044616699219, "learning_rate": 1.0890273524131376e-05, - "loss": 0.5827, + "loss": 0.9374, "step": 43130 }, { "epoch": 4.56, - "grad_norm": 25.853515625, + "grad_norm": 40.604827880859375, "learning_rate": 1.0888161368676735e-05, - "loss": 0.5869, + "loss": 0.9351, "step": 43140 }, { "epoch": 4.56, - "grad_norm": 15.963333129882812, + "grad_norm": 50.706844329833984, "learning_rate": 1.0886049213222094e-05, - "loss": 0.8977, + "loss": 1.0042, "step": 43150 }, { "epoch": 4.56, - "grad_norm": 2.5603277683258057, + "grad_norm": 40.67806625366211, "learning_rate": 1.0883937057767453e-05, - "loss": 0.3909, + "loss": 1.0775, "step": 43160 }, { "epoch": 4.56, - "grad_norm": 4.446661949157715, + "grad_norm": 27.103429794311523, "learning_rate": 1.0881824902312811e-05, - "loss": 0.6285, + "loss": 1.1097, "step": 43170 }, { "epoch": 4.56, - "grad_norm": 26.289236068725586, + "grad_norm": 50.26628875732422, "learning_rate": 1.0879712746858169e-05, - "loss": 0.3339, + "loss": 1.1031, "step": 43180 }, { "epoch": 4.56, - "grad_norm": 6.1930999755859375, + "grad_norm": 45.597408294677734, "learning_rate": 1.0877600591403527e-05, - "loss": 0.5512, + "loss": 1.1705, "step": 43190 }, { "epoch": 4.56, - "grad_norm": 16.529010772705078, + "grad_norm": 38.58096694946289, "learning_rate": 1.0875488435948886e-05, - "loss": 0.4813, + "loss": 0.9733, "step": 43200 }, { "epoch": 4.56, - "grad_norm": 8.287566184997559, + "grad_norm": 27.265138626098633, "learning_rate": 1.0873376280494245e-05, - "loss": 0.7593, + "loss": 1.157, "step": 43210 }, { "epoch": 4.56, - "grad_norm": 0.042820822447538376, + "grad_norm": 42.75395202636719, "learning_rate": 1.0871264125039604e-05, - "loss": 0.3907, + "loss": 1.1413, "step": 43220 }, { "epoch": 4.57, - "grad_norm": 23.608917236328125, + "grad_norm": 46.959251403808594, "learning_rate": 1.0869151969584964e-05, - "loss": 0.4259, + "loss": 1.2364, "step": 43230 }, { "epoch": 4.57, - "grad_norm": 2.6405797004699707, + "grad_norm": 26.119585037231445, "learning_rate": 1.086703981413032e-05, - "loss": 0.5306, + "loss": 0.8633, "step": 43240 }, { "epoch": 4.57, - "grad_norm": 22.822336196899414, + "grad_norm": 46.05534362792969, "learning_rate": 1.0864927658675678e-05, - "loss": 0.8698, + "loss": 1.1249, "step": 43250 }, { "epoch": 4.57, - "grad_norm": 22.963090896606445, + "grad_norm": 29.625761032104492, "learning_rate": 1.0862815503221039e-05, - "loss": 0.404, + "loss": 0.7412, "step": 43260 }, { "epoch": 4.57, - "grad_norm": 21.461782455444336, + "grad_norm": 39.77946853637695, "learning_rate": 1.0860703347766398e-05, - "loss": 0.3993, + "loss": 1.0144, "step": 43270 }, { "epoch": 4.57, - "grad_norm": 25.567399978637695, + "grad_norm": 49.01778030395508, "learning_rate": 1.0858591192311756e-05, - "loss": 0.8842, + "loss": 0.9542, "step": 43280 }, { "epoch": 4.57, - "grad_norm": 11.810545921325684, + "grad_norm": 35.25225067138672, "learning_rate": 1.0856479036857113e-05, - "loss": 0.2828, + "loss": 1.0806, "step": 43290 }, { "epoch": 4.57, - "grad_norm": 15.621267318725586, + "grad_norm": 41.17979431152344, "learning_rate": 1.0854366881402472e-05, - "loss": 0.7026, + "loss": 1.0774, "step": 43300 }, { "epoch": 4.57, - "grad_norm": 12.864333152770996, + "grad_norm": 54.79584503173828, "learning_rate": 1.0852254725947831e-05, - "loss": 0.5557, + "loss": 1.2008, "step": 43310 }, { "epoch": 4.57, - "grad_norm": 8.644331932067871, + "grad_norm": 35.51232147216797, "learning_rate": 1.085014257049319e-05, - "loss": 0.4743, + "loss": 0.9869, "step": 43320 }, { "epoch": 4.58, - "grad_norm": 30.71000862121582, + "grad_norm": 41.09242630004883, "learning_rate": 1.0848030415038549e-05, - "loss": 0.6137, + "loss": 1.1104, "step": 43330 }, { "epoch": 4.58, - "grad_norm": 5.914241313934326, + "grad_norm": 22.88523292541504, "learning_rate": 1.0845918259583906e-05, - "loss": 0.6407, + "loss": 1.3458, "step": 43340 }, { "epoch": 4.58, - "grad_norm": 0.17840896546840668, + "grad_norm": 32.578739166259766, "learning_rate": 1.0843806104129264e-05, - "loss": 0.6869, + "loss": 1.071, "step": 43350 }, { "epoch": 4.58, - "grad_norm": 14.016462326049805, + "grad_norm": 45.7105712890625, "learning_rate": 1.0841693948674623e-05, - "loss": 0.4739, + "loss": 1.0443, "step": 43360 }, { "epoch": 4.58, - "grad_norm": 30.709348678588867, + "grad_norm": 57.94578552246094, "learning_rate": 1.0839581793219982e-05, - "loss": 0.5872, + "loss": 1.3408, "step": 43370 }, { "epoch": 4.58, - "grad_norm": 19.409942626953125, + "grad_norm": 34.53227615356445, "learning_rate": 1.083746963776534e-05, - "loss": 0.3584, + "loss": 1.2245, "step": 43380 }, { "epoch": 4.58, - "grad_norm": 9.613338470458984, + "grad_norm": 35.10444259643555, "learning_rate": 1.08353574823107e-05, - "loss": 0.3679, + "loss": 1.0097, "step": 43390 }, { "epoch": 4.58, - "grad_norm": 9.356117248535156, + "grad_norm": 49.120765686035156, "learning_rate": 1.0833245326856057e-05, - "loss": 0.5425, + "loss": 1.4285, "step": 43400 }, { "epoch": 4.58, - "grad_norm": 7.825122356414795, + "grad_norm": 43.939964294433594, "learning_rate": 1.0831133171401415e-05, - "loss": 0.5562, + "loss": 1.3652, "step": 43410 }, { "epoch": 4.59, - "grad_norm": 14.574831008911133, + "grad_norm": 52.31780242919922, "learning_rate": 1.0829021015946774e-05, - "loss": 0.621, + "loss": 1.4854, "step": 43420 }, { "epoch": 4.59, - "grad_norm": 16.062774658203125, + "grad_norm": 19.68421173095703, "learning_rate": 1.0826908860492133e-05, - "loss": 0.4402, + "loss": 1.0795, "step": 43430 }, { "epoch": 4.59, - "grad_norm": 6.547436237335205, + "grad_norm": 37.78608322143555, "learning_rate": 1.0824796705037492e-05, - "loss": 0.3768, + "loss": 1.4979, "step": 43440 }, { "epoch": 4.59, - "grad_norm": 14.712865829467773, + "grad_norm": 49.93827819824219, "learning_rate": 1.0822684549582849e-05, - "loss": 0.315, + "loss": 0.9344, "step": 43450 }, { "epoch": 4.59, - "grad_norm": 20.346750259399414, + "grad_norm": 44.17812728881836, "learning_rate": 1.0820572394128208e-05, - "loss": 0.643, + "loss": 1.3999, "step": 43460 }, { "epoch": 4.59, - "grad_norm": 1.5345914363861084, + "grad_norm": 31.999120712280273, "learning_rate": 1.0818460238673566e-05, - "loss": 0.5897, + "loss": 1.0774, "step": 43470 }, { "epoch": 4.59, - "grad_norm": 30.783056259155273, + "grad_norm": 37.2374382019043, "learning_rate": 1.0816348083218927e-05, - "loss": 0.6848, + "loss": 0.9515, "step": 43480 }, { "epoch": 4.59, - "grad_norm": 16.381732940673828, + "grad_norm": 48.21773147583008, "learning_rate": 1.0814235927764286e-05, - "loss": 0.6239, + "loss": 1.0441, "step": 43490 }, { "epoch": 4.59, - "grad_norm": 12.991364479064941, + "grad_norm": 24.101530075073242, "learning_rate": 1.0812123772309644e-05, - "loss": 0.4778, + "loss": 0.9815, "step": 43500 }, { "epoch": 4.59, - "grad_norm": 3.341855049133301, + "grad_norm": 31.456966400146484, "learning_rate": 1.0810011616855001e-05, - "loss": 0.3904, + "loss": 1.0875, "step": 43510 }, { "epoch": 4.6, - "grad_norm": 19.337919235229492, + "grad_norm": 21.335206985473633, "learning_rate": 1.080789946140036e-05, - "loss": 0.2839, + "loss": 0.8295, "step": 43520 }, { "epoch": 4.6, - "grad_norm": 18.434080123901367, + "grad_norm": 55.96531677246094, "learning_rate": 1.0805787305945719e-05, - "loss": 0.7602, + "loss": 1.2361, "step": 43530 }, { "epoch": 4.6, - "grad_norm": 1.281604290008545, + "grad_norm": 34.50859069824219, "learning_rate": 1.0803675150491078e-05, - "loss": 0.2208, + "loss": 0.8458, "step": 43540 }, { "epoch": 4.6, - "grad_norm": 17.952159881591797, + "grad_norm": 46.42178726196289, "learning_rate": 1.0801562995036437e-05, - "loss": 0.5251, + "loss": 1.0885, "step": 43550 }, { "epoch": 4.6, - "grad_norm": 18.841190338134766, + "grad_norm": 49.97473907470703, "learning_rate": 1.0799450839581794e-05, - "loss": 0.4414, + "loss": 0.7042, "step": 43560 }, { "epoch": 4.6, - "grad_norm": 3.333167552947998, + "grad_norm": 36.59846115112305, "learning_rate": 1.0797338684127152e-05, - "loss": 0.448, + "loss": 1.0692, "step": 43570 }, { "epoch": 4.6, - "grad_norm": 16.785900115966797, + "grad_norm": 57.266483306884766, "learning_rate": 1.0795226528672511e-05, - "loss": 0.6403, + "loss": 1.0648, "step": 43580 }, { "epoch": 4.6, - "grad_norm": 14.063979148864746, + "grad_norm": 54.23170471191406, "learning_rate": 1.079311437321787e-05, - "loss": 0.5515, + "loss": 1.0434, "step": 43590 }, { "epoch": 4.6, - "grad_norm": 13.346006393432617, + "grad_norm": 32.81477355957031, "learning_rate": 1.0791002217763229e-05, - "loss": 0.2149, + "loss": 1.0041, "step": 43600 }, { "epoch": 4.61, - "grad_norm": 42.39562225341797, + "grad_norm": 38.03068542480469, "learning_rate": 1.0788890062308586e-05, - "loss": 0.4002, + "loss": 1.1104, "step": 43610 }, { "epoch": 4.61, - "grad_norm": 1.0173879861831665, + "grad_norm": 44.80242156982422, "learning_rate": 1.0786777906853945e-05, - "loss": 0.3295, + "loss": 1.1619, "step": 43620 }, { "epoch": 4.61, - "grad_norm": 12.153127670288086, + "grad_norm": 44.097984313964844, "learning_rate": 1.0784665751399303e-05, - "loss": 0.5692, + "loss": 1.0396, "step": 43630 }, { "epoch": 4.61, - "grad_norm": 12.18136215209961, + "grad_norm": 19.66015625, "learning_rate": 1.0782553595944662e-05, - "loss": 0.2526, + "loss": 0.8011, "step": 43640 }, { "epoch": 4.61, - "grad_norm": 13.806772232055664, + "grad_norm": 26.648326873779297, "learning_rate": 1.0780441440490021e-05, - "loss": 0.6182, + "loss": 1.138, "step": 43650 }, { "epoch": 4.61, - "grad_norm": 0.3021998703479767, + "grad_norm": 73.46575927734375, "learning_rate": 1.077832928503538e-05, - "loss": 0.4653, + "loss": 1.1997, "step": 43660 }, { "epoch": 4.61, - "grad_norm": 15.670169830322266, + "grad_norm": 45.73893356323242, "learning_rate": 1.0776217129580737e-05, - "loss": 1.0042, + "loss": 1.291, "step": 43670 }, { "epoch": 4.61, - "grad_norm": 29.233373641967773, + "grad_norm": 55.0518798828125, "learning_rate": 1.0774104974126096e-05, - "loss": 0.4431, + "loss": 0.9979, "step": 43680 }, { "epoch": 4.61, - "grad_norm": 9.833619117736816, + "grad_norm": 14.651302337646484, "learning_rate": 1.0771992818671454e-05, - "loss": 0.4306, + "loss": 0.843, "step": 43690 }, { "epoch": 4.62, - "grad_norm": 15.852363586425781, + "grad_norm": 42.89011001586914, "learning_rate": 1.0769880663216815e-05, - "loss": 0.5247, + "loss": 1.2365, "step": 43700 }, { "epoch": 4.62, - "grad_norm": 9.986173629760742, + "grad_norm": 29.159652709960938, "learning_rate": 1.0767768507762174e-05, - "loss": 0.7542, + "loss": 1.1052, "step": 43710 }, { "epoch": 4.62, - "grad_norm": 1.5230772495269775, + "grad_norm": 59.115760803222656, "learning_rate": 1.0765656352307529e-05, - "loss": 0.467, + "loss": 0.8969, "step": 43720 }, { "epoch": 4.62, - "grad_norm": 15.10538101196289, + "grad_norm": 26.81389045715332, "learning_rate": 1.0763544196852888e-05, - "loss": 0.4625, + "loss": 0.9472, "step": 43730 }, { "epoch": 4.62, - "grad_norm": 2.0611555576324463, + "grad_norm": 60.22075653076172, "learning_rate": 1.0761432041398248e-05, - "loss": 0.3914, + "loss": 1.0662, "step": 43740 }, { "epoch": 4.62, - "grad_norm": 10.013463020324707, + "grad_norm": 41.451663970947266, "learning_rate": 1.0759319885943607e-05, - "loss": 0.4803, + "loss": 1.208, "step": 43750 }, { "epoch": 4.62, - "grad_norm": 11.187414169311523, + "grad_norm": 47.16312789916992, "learning_rate": 1.0757207730488966e-05, - "loss": 0.754, + "loss": 1.3676, "step": 43760 }, { "epoch": 4.62, - "grad_norm": 19.370079040527344, + "grad_norm": 42.56092071533203, "learning_rate": 1.0755095575034325e-05, - "loss": 0.5293, + "loss": 1.1532, "step": 43770 }, { "epoch": 4.62, - "grad_norm": 22.544965744018555, + "grad_norm": 49.010658264160156, "learning_rate": 1.0752983419579682e-05, - "loss": 0.565, + "loss": 1.1837, "step": 43780 }, { "epoch": 4.62, - "grad_norm": 8.775547981262207, + "grad_norm": 45.74249267578125, "learning_rate": 1.075087126412504e-05, - "loss": 0.3305, + "loss": 1.0963, "step": 43790 }, { "epoch": 4.63, - "grad_norm": 0.24358978867530823, + "grad_norm": 26.992530822753906, "learning_rate": 1.07487591086704e-05, - "loss": 0.3184, + "loss": 0.8268, "step": 43800 }, { "epoch": 4.63, - "grad_norm": 23.017444610595703, + "grad_norm": 46.72602844238281, "learning_rate": 1.0746646953215758e-05, - "loss": 0.27, + "loss": 1.366, "step": 43810 }, { "epoch": 4.63, - "grad_norm": 36.63536071777344, + "grad_norm": 47.482093811035156, "learning_rate": 1.0744534797761117e-05, - "loss": 0.5398, + "loss": 0.9942, "step": 43820 }, { "epoch": 4.63, - "grad_norm": 13.415773391723633, + "grad_norm": 43.35080337524414, "learning_rate": 1.0742422642306474e-05, - "loss": 0.4464, + "loss": 1.3294, "step": 43830 }, { "epoch": 4.63, - "grad_norm": 4.9954094886779785, + "grad_norm": 36.13351821899414, "learning_rate": 1.0740310486851833e-05, - "loss": 0.5309, + "loss": 1.4495, "step": 43840 }, { "epoch": 4.63, - "grad_norm": 15.234563827514648, + "grad_norm": 29.112457275390625, "learning_rate": 1.0738198331397191e-05, - "loss": 0.5543, + "loss": 1.02, "step": 43850 }, { "epoch": 4.63, - "grad_norm": 4.915770530700684, + "grad_norm": 38.405887603759766, "learning_rate": 1.073608617594255e-05, - "loss": 0.377, + "loss": 1.0318, "step": 43860 }, { "epoch": 4.63, - "grad_norm": 18.816696166992188, + "grad_norm": 39.99608612060547, "learning_rate": 1.0733974020487909e-05, - "loss": 0.838, + "loss": 1.3822, "step": 43870 }, { "epoch": 4.63, - "grad_norm": 24.461027145385742, + "grad_norm": 32.002593994140625, "learning_rate": 1.0731861865033268e-05, - "loss": 0.6059, + "loss": 1.0756, "step": 43880 }, { "epoch": 4.64, - "grad_norm": 5.115469932556152, + "grad_norm": 43.94717788696289, "learning_rate": 1.0729749709578625e-05, - "loss": 0.5607, + "loss": 1.0486, "step": 43890 }, { "epoch": 4.64, - "grad_norm": 23.276918411254883, + "grad_norm": 38.71465301513672, "learning_rate": 1.0727637554123984e-05, - "loss": 0.3038, + "loss": 1.003, "step": 43900 }, { "epoch": 4.64, - "grad_norm": 0.11028732359409332, + "grad_norm": 29.875877380371094, "learning_rate": 1.0725525398669342e-05, - "loss": 0.4914, + "loss": 1.1062, "step": 43910 }, { "epoch": 4.64, - "grad_norm": 1.5506675243377686, + "grad_norm": 19.55000114440918, "learning_rate": 1.0723413243214701e-05, - "loss": 0.5918, + "loss": 1.0106, "step": 43920 }, { "epoch": 4.64, - "grad_norm": 12.703132629394531, + "grad_norm": 60.14881896972656, "learning_rate": 1.0721301087760062e-05, - "loss": 0.3149, + "loss": 1.1371, "step": 43930 }, { "epoch": 4.64, - "grad_norm": 11.961240768432617, + "grad_norm": 28.89777946472168, "learning_rate": 1.0719188932305417e-05, - "loss": 0.344, + "loss": 0.962, "step": 43940 }, { "epoch": 4.64, - "grad_norm": 4.320766925811768, + "grad_norm": 41.116004943847656, "learning_rate": 1.0717076776850776e-05, - "loss": 0.1682, + "loss": 0.9688, "step": 43950 }, { "epoch": 4.64, - "grad_norm": 4.73954439163208, + "grad_norm": 9.917035102844238, "learning_rate": 1.0714964621396136e-05, - "loss": 0.483, + "loss": 1.1272, "step": 43960 }, { "epoch": 4.64, - "grad_norm": 2.1832592487335205, + "grad_norm": 34.782020568847656, "learning_rate": 1.0712852465941495e-05, - "loss": 0.4574, + "loss": 1.1882, "step": 43970 }, { "epoch": 4.64, - "grad_norm": 12.642548561096191, + "grad_norm": 38.6512565612793, "learning_rate": 1.0710740310486854e-05, - "loss": 0.6382, + "loss": 1.2304, "step": 43980 }, { "epoch": 4.65, - "grad_norm": 13.882780075073242, + "grad_norm": 30.826875686645508, "learning_rate": 1.0708628155032211e-05, - "loss": 0.7213, + "loss": 1.3451, "step": 43990 }, { "epoch": 4.65, - "grad_norm": 20.763761520385742, + "grad_norm": 27.46826934814453, "learning_rate": 1.070651599957757e-05, - "loss": 0.4341, + "loss": 1.1109, "step": 44000 }, { "epoch": 4.65, - "grad_norm": 5.302213191986084, + "grad_norm": 42.87327575683594, "learning_rate": 1.0704403844122929e-05, - "loss": 0.4139, + "loss": 1.1023, "step": 44010 }, { "epoch": 4.65, - "grad_norm": 2.4694113731384277, + "grad_norm": 30.953487396240234, "learning_rate": 1.0702291688668287e-05, - "loss": 0.6078, + "loss": 1.1001, "step": 44020 }, { "epoch": 4.65, - "grad_norm": 2.8729512691497803, + "grad_norm": 44.712833404541016, "learning_rate": 1.0700179533213646e-05, - "loss": 0.1688, + "loss": 1.0331, "step": 44030 }, { "epoch": 4.65, - "grad_norm": 2.385812997817993, + "grad_norm": 25.268522262573242, "learning_rate": 1.0698067377759005e-05, - "loss": 0.4276, + "loss": 0.9843, "step": 44040 }, { "epoch": 4.65, - "grad_norm": 24.640050888061523, + "grad_norm": 36.632530212402344, "learning_rate": 1.0695955222304362e-05, - "loss": 0.4347, + "loss": 1.1161, "step": 44050 }, { "epoch": 4.65, - "grad_norm": 7.612886905670166, + "grad_norm": 28.963834762573242, "learning_rate": 1.069384306684972e-05, - "loss": 0.501, + "loss": 1.1218, "step": 44060 }, { "epoch": 4.65, - "grad_norm": 30.375791549682617, + "grad_norm": 44.188758850097656, "learning_rate": 1.069173091139508e-05, - "loss": 0.41, + "loss": 0.9252, "step": 44070 }, { "epoch": 4.66, - "grad_norm": 28.64900016784668, + "grad_norm": 24.756908416748047, "learning_rate": 1.0689618755940438e-05, - "loss": 0.511, + "loss": 1.047, "step": 44080 }, { "epoch": 4.66, - "grad_norm": 0.6876987218856812, + "grad_norm": 28.188383102416992, "learning_rate": 1.0687506600485797e-05, - "loss": 0.3657, + "loss": 1.1846, "step": 44090 }, { "epoch": 4.66, - "grad_norm": 38.27344512939453, + "grad_norm": 62.840457916259766, "learning_rate": 1.0685394445031154e-05, - "loss": 0.6105, + "loss": 1.4627, "step": 44100 }, { "epoch": 4.66, - "grad_norm": 20.95911407470703, + "grad_norm": 62.88853454589844, "learning_rate": 1.0683282289576513e-05, - "loss": 0.6932, + "loss": 0.8944, "step": 44110 }, { "epoch": 4.66, - "grad_norm": 29.870454788208008, + "grad_norm": 36.627685546875, "learning_rate": 1.0681170134121872e-05, - "loss": 0.7396, + "loss": 1.1764, "step": 44120 }, { "epoch": 4.66, - "grad_norm": 6.798761367797852, + "grad_norm": 46.79564666748047, "learning_rate": 1.067905797866723e-05, - "loss": 0.5607, + "loss": 1.4474, "step": 44130 }, { "epoch": 4.66, - "grad_norm": 19.170047760009766, + "grad_norm": 39.7557373046875, "learning_rate": 1.067694582321259e-05, - "loss": 0.4111, + "loss": 1.025, "step": 44140 }, { "epoch": 4.66, - "grad_norm": 10.480202674865723, + "grad_norm": 38.83900451660156, "learning_rate": 1.067483366775795e-05, - "loss": 0.318, + "loss": 0.7401, "step": 44150 }, { "epoch": 4.66, - "grad_norm": 3.079728126525879, + "grad_norm": 47.04343032836914, "learning_rate": 1.0672721512303305e-05, - "loss": 0.4158, + "loss": 1.05, "step": 44160 }, { "epoch": 4.66, - "grad_norm": 11.886345863342285, + "grad_norm": 22.143442153930664, "learning_rate": 1.0670609356848664e-05, - "loss": 0.1654, + "loss": 1.0002, "step": 44170 }, { "epoch": 4.67, - "grad_norm": 14.379077911376953, + "grad_norm": 15.526544570922852, "learning_rate": 1.0668497201394024e-05, - "loss": 0.2974, + "loss": 0.9808, "step": 44180 }, { "epoch": 4.67, - "grad_norm": 26.261402130126953, + "grad_norm": 38.4040641784668, "learning_rate": 1.0666385045939383e-05, - "loss": 0.4574, + "loss": 1.1455, "step": 44190 }, { "epoch": 4.67, - "grad_norm": 12.704206466674805, + "grad_norm": 35.678131103515625, "learning_rate": 1.0664272890484742e-05, - "loss": 0.4655, + "loss": 0.9984, "step": 44200 }, { "epoch": 4.67, - "grad_norm": 34.75705337524414, + "grad_norm": 34.5914421081543, "learning_rate": 1.0662160735030097e-05, - "loss": 0.8027, + "loss": 1.0085, "step": 44210 }, { "epoch": 4.67, - "grad_norm": 5.271812915802002, + "grad_norm": 41.45012283325195, "learning_rate": 1.0660048579575458e-05, - "loss": 0.7101, + "loss": 1.2373, "step": 44220 }, { "epoch": 4.67, - "grad_norm": 19.014734268188477, + "grad_norm": 60.551719665527344, "learning_rate": 1.0657936424120817e-05, - "loss": 0.2971, + "loss": 1.1543, "step": 44230 }, { "epoch": 4.67, - "grad_norm": 13.674544334411621, + "grad_norm": 43.767578125, "learning_rate": 1.0655824268666175e-05, - "loss": 0.7611, + "loss": 1.147, "step": 44240 }, { "epoch": 4.67, - "grad_norm": 2.835390329360962, + "grad_norm": 44.09487533569336, "learning_rate": 1.0653712113211534e-05, - "loss": 0.4911, + "loss": 0.8189, "step": 44250 }, { "epoch": 4.67, - "grad_norm": 7.677165985107422, + "grad_norm": 41.97184753417969, "learning_rate": 1.0651599957756891e-05, - "loss": 0.487, + "loss": 1.0064, "step": 44260 }, { "epoch": 4.68, - "grad_norm": 17.33140754699707, + "grad_norm": 45.312255859375, "learning_rate": 1.064948780230225e-05, - "loss": 0.6236, + "loss": 1.5625, "step": 44270 }, { "epoch": 4.68, - "grad_norm": 22.23434829711914, + "grad_norm": 31.7836856842041, "learning_rate": 1.0647375646847609e-05, - "loss": 0.6227, + "loss": 1.2337, "step": 44280 }, { "epoch": 4.68, - "grad_norm": 13.037980079650879, + "grad_norm": 24.17837905883789, "learning_rate": 1.0645263491392968e-05, - "loss": 0.4822, + "loss": 1.3576, "step": 44290 }, { "epoch": 4.68, - "grad_norm": 12.775975227355957, + "grad_norm": 37.73645782470703, "learning_rate": 1.0643151335938326e-05, - "loss": 0.9623, + "loss": 1.4134, "step": 44300 }, { "epoch": 4.68, - "grad_norm": 15.758288383483887, + "grad_norm": 33.59849166870117, "learning_rate": 1.0641039180483685e-05, - "loss": 0.4317, + "loss": 0.9555, "step": 44310 }, { "epoch": 4.68, - "grad_norm": 19.859249114990234, + "grad_norm": 56.3745002746582, "learning_rate": 1.0638927025029042e-05, - "loss": 0.4496, + "loss": 0.9687, "step": 44320 }, { "epoch": 4.68, - "grad_norm": 22.77869987487793, + "grad_norm": 38.33436584472656, "learning_rate": 1.0636814869574401e-05, - "loss": 0.7038, + "loss": 1.2149, "step": 44330 }, { "epoch": 4.68, - "grad_norm": 10.858649253845215, + "grad_norm": 25.920185089111328, "learning_rate": 1.063470271411976e-05, - "loss": 0.6046, + "loss": 1.3683, "step": 44340 }, { "epoch": 4.68, - "grad_norm": 9.04749584197998, + "grad_norm": 55.31936264038086, "learning_rate": 1.0632590558665119e-05, - "loss": 0.785, + "loss": 1.1838, "step": 44350 }, { "epoch": 4.68, - "grad_norm": 20.150360107421875, + "grad_norm": 51.946990966796875, "learning_rate": 1.0630478403210477e-05, - "loss": 0.3463, + "loss": 0.8971, "step": 44360 }, { "epoch": 4.69, - "grad_norm": 5.0581841468811035, + "grad_norm": 27.064998626708984, "learning_rate": 1.0628366247755834e-05, - "loss": 0.7086, + "loss": 1.1122, "step": 44370 }, { "epoch": 4.69, - "grad_norm": 14.067818641662598, + "grad_norm": 29.612958908081055, "learning_rate": 1.0626254092301193e-05, - "loss": 0.3241, + "loss": 1.0912, "step": 44380 }, { "epoch": 4.69, - "grad_norm": 0.8624260425567627, + "grad_norm": 40.32683563232422, "learning_rate": 1.0624141936846552e-05, - "loss": 0.3766, + "loss": 1.028, "step": 44390 }, { "epoch": 4.69, - "grad_norm": 28.429744720458984, + "grad_norm": 12.255058288574219, "learning_rate": 1.062202978139191e-05, - "loss": 0.499, + "loss": 1.116, "step": 44400 }, { "epoch": 4.69, - "grad_norm": 22.444852828979492, + "grad_norm": 43.28523635864258, "learning_rate": 1.0619917625937271e-05, - "loss": 0.3474, + "loss": 0.9286, "step": 44410 }, { "epoch": 4.69, - "grad_norm": 13.657094955444336, + "grad_norm": 43.567325592041016, "learning_rate": 1.061780547048263e-05, - "loss": 0.4459, + "loss": 1.0919, "step": 44420 }, { "epoch": 4.69, - "grad_norm": 28.55990982055664, + "grad_norm": 68.6373291015625, "learning_rate": 1.0615693315027985e-05, - "loss": 0.5215, + "loss": 1.1729, "step": 44430 }, { "epoch": 4.69, - "grad_norm": 10.831570625305176, + "grad_norm": 17.234098434448242, "learning_rate": 1.0613581159573346e-05, - "loss": 0.6163, + "loss": 1.1406, "step": 44440 }, { "epoch": 4.69, - "grad_norm": 32.90367126464844, + "grad_norm": 44.20188522338867, "learning_rate": 1.0611469004118705e-05, - "loss": 0.5326, + "loss": 1.14, "step": 44450 }, { "epoch": 4.7, - "grad_norm": 8.806970596313477, + "grad_norm": 29.667072296142578, "learning_rate": 1.0609356848664063e-05, - "loss": 0.4145, + "loss": 0.8222, "step": 44460 }, { "epoch": 4.7, - "grad_norm": 0.12285063415765762, + "grad_norm": 23.091062545776367, "learning_rate": 1.0607244693209422e-05, - "loss": 0.6202, + "loss": 1.0357, "step": 44470 }, { "epoch": 4.7, - "grad_norm": 0.1779949814081192, + "grad_norm": 54.499019622802734, "learning_rate": 1.060513253775478e-05, - "loss": 0.269, + "loss": 1.0949, "step": 44480 }, { "epoch": 4.7, - "grad_norm": 15.493392944335938, + "grad_norm": 49.49454879760742, "learning_rate": 1.0603020382300138e-05, - "loss": 0.4991, + "loss": 1.2202, "step": 44490 }, { "epoch": 4.7, - "grad_norm": 33.20719528198242, + "grad_norm": 30.156604766845703, "learning_rate": 1.0600908226845497e-05, - "loss": 0.6235, + "loss": 1.3941, "step": 44500 }, { "epoch": 4.7, - "grad_norm": 1.08942449092865, + "grad_norm": 50.1152229309082, "learning_rate": 1.0598796071390856e-05, - "loss": 0.3764, + "loss": 1.1785, "step": 44510 }, { "epoch": 4.7, - "grad_norm": 5.87619161605835, + "grad_norm": 36.32884979248047, "learning_rate": 1.0596683915936214e-05, - "loss": 0.3923, + "loss": 1.3058, "step": 44520 }, { "epoch": 4.7, - "grad_norm": 7.339729309082031, + "grad_norm": 23.856599807739258, "learning_rate": 1.0594571760481571e-05, - "loss": 0.6522, + "loss": 1.2341, "step": 44530 }, { "epoch": 4.7, - "grad_norm": 23.734174728393555, + "grad_norm": 42.942832946777344, "learning_rate": 1.059245960502693e-05, - "loss": 0.4946, + "loss": 1.1607, "step": 44540 }, { "epoch": 4.7, - "grad_norm": 0.31580018997192383, + "grad_norm": 21.507017135620117, "learning_rate": 1.0590347449572289e-05, - "loss": 0.3349, + "loss": 0.8254, "step": 44550 }, { "epoch": 4.71, - "grad_norm": 2.7722809314727783, + "grad_norm": 35.6081657409668, "learning_rate": 1.0588235294117648e-05, - "loss": 0.5636, + "loss": 1.2923, "step": 44560 }, { "epoch": 4.71, - "grad_norm": 12.940821647644043, + "grad_norm": 46.49764633178711, "learning_rate": 1.0586123138663007e-05, - "loss": 0.6778, + "loss": 1.3903, "step": 44570 }, { "epoch": 4.71, - "grad_norm": 11.09471321105957, + "grad_norm": 44.08858108520508, "learning_rate": 1.0584010983208365e-05, - "loss": 0.5465, + "loss": 1.1424, "step": 44580 }, { "epoch": 4.71, - "grad_norm": 4.8667683601379395, + "grad_norm": 24.485034942626953, "learning_rate": 1.0581898827753722e-05, - "loss": 0.6161, + "loss": 0.9046, "step": 44590 }, { "epoch": 4.71, - "grad_norm": 15.827543258666992, + "grad_norm": 48.75469207763672, "learning_rate": 1.0579786672299081e-05, - "loss": 0.5279, + "loss": 1.4375, "step": 44600 }, { "epoch": 4.71, - "grad_norm": 0.18052874505519867, + "grad_norm": 34.38706970214844, "learning_rate": 1.057767451684444e-05, - "loss": 0.2387, + "loss": 0.6605, "step": 44610 }, { "epoch": 4.71, - "grad_norm": 11.248808860778809, + "grad_norm": 20.18427085876465, "learning_rate": 1.0575562361389799e-05, - "loss": 0.7244, + "loss": 0.92, "step": 44620 }, { "epoch": 4.71, - "grad_norm": 17.67243003845215, + "grad_norm": 50.61945724487305, "learning_rate": 1.057345020593516e-05, - "loss": 0.5423, + "loss": 0.8877, "step": 44630 }, { "epoch": 4.71, - "grad_norm": 12.241774559020996, + "grad_norm": 48.17643737792969, "learning_rate": 1.0571338050480515e-05, - "loss": 0.5654, + "loss": 1.1999, "step": 44640 }, { "epoch": 4.72, - "grad_norm": 19.60362434387207, + "grad_norm": 35.194664001464844, "learning_rate": 1.0569225895025873e-05, - "loss": 0.6755, + "loss": 1.2729, "step": 44650 }, { "epoch": 4.72, - "grad_norm": 15.968413352966309, + "grad_norm": 37.96891784667969, "learning_rate": 1.0567113739571234e-05, - "loss": 0.415, + "loss": 1.1393, "step": 44660 }, { "epoch": 4.72, - "grad_norm": 1.4153133630752563, + "grad_norm": 21.38011360168457, "learning_rate": 1.0565001584116593e-05, - "loss": 0.5546, + "loss": 1.2236, "step": 44670 }, { "epoch": 4.72, - "grad_norm": 21.345651626586914, + "grad_norm": 25.02914810180664, "learning_rate": 1.0562889428661951e-05, - "loss": 0.5184, + "loss": 1.316, "step": 44680 }, { "epoch": 4.72, - "grad_norm": 0.1706259399652481, + "grad_norm": 29.296934127807617, "learning_rate": 1.056077727320731e-05, - "loss": 0.4279, + "loss": 1.2651, "step": 44690 }, { "epoch": 4.72, - "grad_norm": 6.30550479888916, + "grad_norm": 22.895708084106445, "learning_rate": 1.0558665117752667e-05, - "loss": 0.3829, + "loss": 1.1498, "step": 44700 }, { "epoch": 4.72, - "grad_norm": 1.6139086484909058, + "grad_norm": 40.67845916748047, "learning_rate": 1.0556552962298026e-05, - "loss": 0.4192, + "loss": 1.1997, "step": 44710 }, { "epoch": 4.72, - "grad_norm": 0.3022497892379761, + "grad_norm": 23.228151321411133, "learning_rate": 1.0554440806843385e-05, - "loss": 0.4833, + "loss": 1.0208, "step": 44720 }, { "epoch": 4.72, - "grad_norm": 0.4501368999481201, + "grad_norm": 42.43389129638672, "learning_rate": 1.0552328651388744e-05, - "loss": 0.6005, + "loss": 1.0354, "step": 44730 }, { "epoch": 4.72, - "grad_norm": 9.907224655151367, + "grad_norm": 42.137107849121094, "learning_rate": 1.0550216495934102e-05, - "loss": 0.5354, + "loss": 1.1107, "step": 44740 }, { "epoch": 4.73, - "grad_norm": 33.61720275878906, + "grad_norm": 49.97984313964844, "learning_rate": 1.054810434047946e-05, - "loss": 0.4869, + "loss": 1.2324, "step": 44750 }, { "epoch": 4.73, - "grad_norm": 14.695619583129883, + "grad_norm": 32.03367233276367, "learning_rate": 1.0545992185024818e-05, - "loss": 0.4878, + "loss": 0.9704, "step": 44760 }, { "epoch": 4.73, - "grad_norm": 17.912330627441406, + "grad_norm": 41.776947021484375, "learning_rate": 1.0543880029570177e-05, - "loss": 0.4924, + "loss": 1.0538, "step": 44770 }, { "epoch": 4.73, - "grad_norm": 21.79955291748047, + "grad_norm": 35.047828674316406, "learning_rate": 1.0541767874115536e-05, - "loss": 0.5826, + "loss": 0.8414, "step": 44780 }, { "epoch": 4.73, - "grad_norm": 11.906737327575684, + "grad_norm": 32.22219467163086, "learning_rate": 1.0539655718660895e-05, - "loss": 0.5283, + "loss": 0.8709, "step": 44790 }, { "epoch": 4.73, - "grad_norm": 28.40541648864746, + "grad_norm": 46.669864654541016, "learning_rate": 1.0537543563206253e-05, - "loss": 0.3589, + "loss": 1.2086, "step": 44800 }, { "epoch": 4.73, - "grad_norm": 13.555207252502441, + "grad_norm": 35.795169830322266, "learning_rate": 1.053543140775161e-05, - "loss": 0.5575, + "loss": 1.1012, "step": 44810 }, { "epoch": 4.73, - "grad_norm": 18.971101760864258, + "grad_norm": 34.96262741088867, "learning_rate": 1.053331925229697e-05, - "loss": 0.5022, + "loss": 0.9585, "step": 44820 }, { "epoch": 4.73, - "grad_norm": 28.275760650634766, + "grad_norm": 51.50735855102539, "learning_rate": 1.0531207096842328e-05, - "loss": 0.6316, + "loss": 1.1774, "step": 44830 }, { "epoch": 4.74, - "grad_norm": 10.911355018615723, + "grad_norm": 46.47599411010742, "learning_rate": 1.0529094941387687e-05, - "loss": 0.726, + "loss": 1.2949, "step": 44840 }, { "epoch": 4.74, - "grad_norm": 5.136437892913818, + "grad_norm": 16.724014282226562, "learning_rate": 1.0526982785933046e-05, - "loss": 0.7756, + "loss": 1.2892, "step": 44850 }, { "epoch": 4.74, - "grad_norm": 11.20857048034668, + "grad_norm": 34.371952056884766, "learning_rate": 1.0524870630478403e-05, - "loss": 0.3537, + "loss": 0.9972, "step": 44860 }, { "epoch": 4.74, - "grad_norm": 10.593435287475586, + "grad_norm": 50.9582405090332, "learning_rate": 1.0522758475023761e-05, - "loss": 0.2814, + "loss": 1.2854, "step": 44870 }, { "epoch": 4.74, - "grad_norm": 21.413803100585938, + "grad_norm": 49.87091064453125, "learning_rate": 1.052064631956912e-05, - "loss": 0.318, + "loss": 0.8839, "step": 44880 }, { "epoch": 4.74, - "grad_norm": 20.13332176208496, + "grad_norm": 35.228275299072266, "learning_rate": 1.051853416411448e-05, - "loss": 0.4225, + "loss": 0.871, "step": 44890 }, { "epoch": 4.74, - "grad_norm": 20.294519424438477, + "grad_norm": 31.773374557495117, "learning_rate": 1.051642200865984e-05, - "loss": 0.9655, + "loss": 1.2982, "step": 44900 }, { "epoch": 4.74, - "grad_norm": 0.27617478370666504, + "grad_norm": 32.206050872802734, "learning_rate": 1.0514309853205195e-05, - "loss": 0.4089, + "loss": 0.81, "step": 44910 }, { "epoch": 4.74, - "grad_norm": 20.06007194519043, + "grad_norm": 36.249061584472656, "learning_rate": 1.0512197697750555e-05, - "loss": 0.8808, + "loss": 1.2663, "step": 44920 }, { "epoch": 4.74, - "grad_norm": 16.327871322631836, + "grad_norm": 35.98954772949219, "learning_rate": 1.0510085542295914e-05, - "loss": 0.437, + "loss": 0.8194, "step": 44930 }, { "epoch": 4.75, - "grad_norm": 6.650501728057861, + "grad_norm": 23.798748016357422, "learning_rate": 1.0507973386841273e-05, - "loss": 0.2993, + "loss": 0.7288, "step": 44940 }, { "epoch": 4.75, - "grad_norm": 22.2423095703125, + "grad_norm": 39.107421875, "learning_rate": 1.0505861231386632e-05, - "loss": 0.5151, + "loss": 1.0474, "step": 44950 }, { "epoch": 4.75, - "grad_norm": 16.326187133789062, + "grad_norm": 36.779212951660156, "learning_rate": 1.050374907593199e-05, - "loss": 0.6004, + "loss": 1.1335, "step": 44960 }, { "epoch": 4.75, - "grad_norm": 12.474946975708008, + "grad_norm": 31.020126342773438, "learning_rate": 1.0501636920477348e-05, - "loss": 0.6353, + "loss": 1.2482, "step": 44970 }, { "epoch": 4.75, - "grad_norm": 2.0429561138153076, + "grad_norm": 55.89236068725586, "learning_rate": 1.0499524765022706e-05, - "loss": 0.4814, + "loss": 1.042, "step": 44980 }, { "epoch": 4.75, - "grad_norm": 5.067514419555664, + "grad_norm": 40.347137451171875, "learning_rate": 1.0497412609568065e-05, - "loss": 0.4612, + "loss": 1.1864, "step": 44990 }, { "epoch": 4.75, - "grad_norm": 4.797189712524414, + "grad_norm": 22.238588333129883, "learning_rate": 1.0495300454113424e-05, - "loss": 0.2307, + "loss": 0.6654, "step": 45000 }, { "epoch": 4.75, - "grad_norm": 11.75805377960205, + "grad_norm": 29.064197540283203, "learning_rate": 1.0493188298658783e-05, - "loss": 0.32, + "loss": 0.8204, "step": 45010 }, { "epoch": 4.75, - "grad_norm": 10.362356185913086, + "grad_norm": 34.97114944458008, "learning_rate": 1.049107614320414e-05, - "loss": 0.4937, + "loss": 1.2141, "step": 45020 }, { "epoch": 4.76, - "grad_norm": 4.339316368103027, + "grad_norm": 47.85586929321289, "learning_rate": 1.0488963987749499e-05, - "loss": 0.5059, + "loss": 0.9478, "step": 45030 }, { "epoch": 4.76, - "grad_norm": 14.208968162536621, + "grad_norm": 36.47681427001953, "learning_rate": 1.0486851832294857e-05, - "loss": 0.6134, + "loss": 1.2808, "step": 45040 }, { "epoch": 4.76, - "grad_norm": 0.02412213385105133, + "grad_norm": 30.874731063842773, "learning_rate": 1.0484739676840216e-05, - "loss": 0.5123, + "loss": 1.0986, "step": 45050 }, { "epoch": 4.76, - "grad_norm": 12.651161193847656, + "grad_norm": 37.10901641845703, "learning_rate": 1.0482627521385575e-05, - "loss": 0.6432, + "loss": 1.3639, "step": 45060 }, { "epoch": 4.76, - "grad_norm": 5.947412014007568, + "grad_norm": 40.9049186706543, "learning_rate": 1.0480515365930934e-05, - "loss": 0.3868, + "loss": 0.9503, "step": 45070 }, { "epoch": 4.76, - "grad_norm": 0.12462792545557022, + "grad_norm": 55.2357063293457, "learning_rate": 1.047840321047629e-05, - "loss": 0.3571, + "loss": 1.0113, "step": 45080 }, { "epoch": 4.76, - "grad_norm": 6.756500720977783, + "grad_norm": 27.10926055908203, "learning_rate": 1.047629105502165e-05, - "loss": 0.9434, + "loss": 1.2498, "step": 45090 }, { "epoch": 4.76, - "grad_norm": 27.171995162963867, + "grad_norm": 35.00263977050781, "learning_rate": 1.0474178899567008e-05, - "loss": 0.6049, + "loss": 1.1238, "step": 45100 }, { "epoch": 4.76, - "grad_norm": 0.7559495568275452, + "grad_norm": 28.348350524902344, "learning_rate": 1.0472066744112369e-05, - "loss": 0.3611, + "loss": 0.9974, "step": 45110 }, { "epoch": 4.77, - "grad_norm": 0.3146228492259979, + "grad_norm": 35.453575134277344, "learning_rate": 1.0469954588657728e-05, - "loss": 0.2639, + "loss": 0.9669, "step": 45120 }, { "epoch": 4.77, - "grad_norm": 16.628047943115234, + "grad_norm": 46.969207763671875, "learning_rate": 1.0467842433203083e-05, - "loss": 0.5768, + "loss": 1.1215, "step": 45130 }, { "epoch": 4.77, - "grad_norm": 0.5084901452064514, + "grad_norm": 29.082599639892578, "learning_rate": 1.0465730277748443e-05, - "loss": 0.2279, + "loss": 0.8431, "step": 45140 }, { "epoch": 4.77, - "grad_norm": 29.73672866821289, + "grad_norm": 36.845664978027344, "learning_rate": 1.0463618122293802e-05, - "loss": 0.7491, + "loss": 1.1889, "step": 45150 }, { "epoch": 4.77, - "grad_norm": 15.486329078674316, + "grad_norm": 24.852333068847656, "learning_rate": 1.0461505966839161e-05, - "loss": 0.5455, + "loss": 1.1808, "step": 45160 }, { "epoch": 4.77, - "grad_norm": 22.201438903808594, + "grad_norm": 53.938480377197266, "learning_rate": 1.045939381138452e-05, - "loss": 0.422, + "loss": 1.1212, "step": 45170 }, { "epoch": 4.77, - "grad_norm": 1.8898711204528809, + "grad_norm": 40.17182159423828, "learning_rate": 1.0457281655929877e-05, - "loss": 0.3948, + "loss": 1.0121, "step": 45180 }, { "epoch": 4.77, - "grad_norm": 14.567368507385254, + "grad_norm": 33.26445388793945, "learning_rate": 1.0455169500475236e-05, - "loss": 0.5323, + "loss": 1.1035, "step": 45190 }, { "epoch": 4.77, - "grad_norm": 10.460960388183594, + "grad_norm": 34.818294525146484, "learning_rate": 1.0453057345020594e-05, - "loss": 0.4889, + "loss": 1.2498, "step": 45200 }, { "epoch": 4.77, - "grad_norm": 6.6985368728637695, + "grad_norm": 37.318115234375, "learning_rate": 1.0450945189565953e-05, - "loss": 0.3919, + "loss": 1.1842, "step": 45210 }, { "epoch": 4.78, - "grad_norm": 15.420588493347168, + "grad_norm": 51.64983367919922, "learning_rate": 1.0448833034111312e-05, - "loss": 0.2741, + "loss": 1.0383, "step": 45220 }, { "epoch": 4.78, - "grad_norm": 16.718984603881836, + "grad_norm": 27.439876556396484, "learning_rate": 1.044672087865667e-05, - "loss": 0.5023, + "loss": 1.0028, "step": 45230 }, { "epoch": 4.78, - "grad_norm": 21.6431884765625, + "grad_norm": 41.85303497314453, "learning_rate": 1.0444608723202028e-05, - "loss": 0.4469, + "loss": 0.8599, "step": 45240 }, { "epoch": 4.78, - "grad_norm": 9.990031242370605, + "grad_norm": 29.38489532470703, "learning_rate": 1.0442496567747387e-05, - "loss": 0.5495, + "loss": 1.1117, "step": 45250 }, { "epoch": 4.78, - "grad_norm": 18.217729568481445, + "grad_norm": 33.09591293334961, "learning_rate": 1.0440384412292745e-05, - "loss": 0.4455, + "loss": 1.1318, "step": 45260 }, { "epoch": 4.78, - "grad_norm": 20.038379669189453, + "grad_norm": 34.40373992919922, "learning_rate": 1.0438272256838104e-05, - "loss": 0.6753, + "loss": 1.0242, "step": 45270 }, { "epoch": 4.78, - "grad_norm": 2.560992956161499, + "grad_norm": 33.863067626953125, "learning_rate": 1.0436160101383463e-05, - "loss": 0.5192, + "loss": 1.1546, "step": 45280 }, { "epoch": 4.78, - "grad_norm": 13.270837783813477, + "grad_norm": 36.02675247192383, "learning_rate": 1.043404794592882e-05, - "loss": 0.4843, + "loss": 0.9712, "step": 45290 }, { "epoch": 4.78, - "grad_norm": 42.03147888183594, + "grad_norm": 50.527976989746094, "learning_rate": 1.0431935790474179e-05, - "loss": 0.4104, + "loss": 0.918, "step": 45300 }, { "epoch": 4.79, - "grad_norm": 3.510608673095703, + "grad_norm": 34.964839935302734, "learning_rate": 1.0429823635019538e-05, - "loss": 0.4935, + "loss": 0.8423, "step": 45310 }, { "epoch": 4.79, - "grad_norm": 21.00999641418457, + "grad_norm": 31.662893295288086, "learning_rate": 1.0427711479564896e-05, - "loss": 0.3598, + "loss": 1.0957, "step": 45320 }, { "epoch": 4.79, - "grad_norm": 12.577009201049805, + "grad_norm": 43.241031646728516, "learning_rate": 1.0425599324110255e-05, - "loss": 0.4814, + "loss": 0.9225, "step": 45330 }, { "epoch": 4.79, - "grad_norm": 0.7601166367530823, + "grad_norm": 15.204094886779785, "learning_rate": 1.0423487168655616e-05, - "loss": 0.2504, + "loss": 1.1351, "step": 45340 }, { "epoch": 4.79, - "grad_norm": 18.123619079589844, + "grad_norm": 39.35888671875, "learning_rate": 1.0421375013200971e-05, - "loss": 0.5592, + "loss": 1.1579, "step": 45350 }, { "epoch": 4.79, - "grad_norm": 14.670759201049805, + "grad_norm": 29.06584930419922, "learning_rate": 1.041926285774633e-05, - "loss": 0.6761, + "loss": 1.2488, "step": 45360 }, { "epoch": 4.79, - "grad_norm": 18.66687774658203, + "grad_norm": 40.49238204956055, "learning_rate": 1.041715070229169e-05, - "loss": 0.411, + "loss": 1.2243, "step": 45370 }, { "epoch": 4.79, - "grad_norm": 8.709541320800781, + "grad_norm": 35.75601577758789, "learning_rate": 1.0415038546837049e-05, - "loss": 0.5473, + "loss": 1.1438, "step": 45380 }, { "epoch": 4.79, - "grad_norm": 16.46306800842285, + "grad_norm": 15.028420448303223, "learning_rate": 1.0412926391382408e-05, - "loss": 0.5848, + "loss": 0.8672, "step": 45390 }, { "epoch": 4.79, - "grad_norm": 13.137858390808105, + "grad_norm": 45.79551315307617, "learning_rate": 1.0410814235927765e-05, - "loss": 0.4051, + "loss": 1.0905, "step": 45400 }, { "epoch": 4.8, - "grad_norm": 17.340879440307617, + "grad_norm": 41.18794631958008, "learning_rate": 1.0408702080473124e-05, - "loss": 0.5727, + "loss": 1.1446, "step": 45410 }, { "epoch": 4.8, - "grad_norm": 3.4869353771209717, + "grad_norm": 20.270998001098633, "learning_rate": 1.0406589925018482e-05, - "loss": 0.6191, + "loss": 1.0993, "step": 45420 }, { "epoch": 4.8, - "grad_norm": 5.41764497756958, + "grad_norm": 47.489501953125, "learning_rate": 1.0404477769563841e-05, - "loss": 0.4198, + "loss": 0.957, "step": 45430 }, { "epoch": 4.8, - "grad_norm": 21.66798210144043, + "grad_norm": 52.6865234375, "learning_rate": 1.04023656141092e-05, - "loss": 0.7191, + "loss": 1.2248, "step": 45440 }, { "epoch": 4.8, - "grad_norm": 0.2242582142353058, + "grad_norm": 29.33013153076172, "learning_rate": 1.0400253458654559e-05, - "loss": 0.7392, + "loss": 1.0965, "step": 45450 }, { "epoch": 4.8, - "grad_norm": 19.119789123535156, + "grad_norm": 41.661006927490234, "learning_rate": 1.0398141303199916e-05, - "loss": 0.5282, + "loss": 1.3596, "step": 45460 }, { "epoch": 4.8, - "grad_norm": 13.267374038696289, + "grad_norm": 49.025634765625, "learning_rate": 1.0396029147745275e-05, - "loss": 0.5147, + "loss": 0.9044, "step": 45470 }, { "epoch": 4.8, - "grad_norm": 3.0788843631744385, + "grad_norm": 29.13151741027832, "learning_rate": 1.0393916992290633e-05, - "loss": 0.1915, + "loss": 0.8399, "step": 45480 }, { "epoch": 4.8, - "grad_norm": 1.3088092803955078, + "grad_norm": 56.57205581665039, "learning_rate": 1.0391804836835992e-05, - "loss": 0.4969, + "loss": 1.4826, "step": 45490 }, { "epoch": 4.81, - "grad_norm": 17.872468948364258, + "grad_norm": 25.671567916870117, "learning_rate": 1.0389692681381351e-05, - "loss": 0.4483, + "loss": 1.0856, "step": 45500 }, { "epoch": 4.81, - "grad_norm": 4.8584699630737305, + "grad_norm": 43.50102996826172, "learning_rate": 1.0387580525926708e-05, - "loss": 0.3862, + "loss": 1.2384, "step": 45510 }, { "epoch": 4.81, - "grad_norm": 13.408103942871094, + "grad_norm": 39.38611602783203, "learning_rate": 1.0385468370472067e-05, - "loss": 0.3049, + "loss": 0.9742, "step": 45520 }, { "epoch": 4.81, - "grad_norm": 8.459942817687988, + "grad_norm": 33.810203552246094, "learning_rate": 1.0383356215017426e-05, - "loss": 0.7736, + "loss": 1.1178, "step": 45530 }, { "epoch": 4.81, - "grad_norm": 12.573123931884766, + "grad_norm": 33.398067474365234, "learning_rate": 1.0381244059562784e-05, - "loss": 0.4426, + "loss": 1.1778, "step": 45540 }, { "epoch": 4.81, - "grad_norm": 15.603165626525879, + "grad_norm": 55.78185272216797, "learning_rate": 1.0379131904108143e-05, - "loss": 0.6971, + "loss": 1.1274, "step": 45550 }, { "epoch": 4.81, - "grad_norm": 29.536651611328125, + "grad_norm": 68.94178009033203, "learning_rate": 1.03770197486535e-05, - "loss": 0.4527, + "loss": 1.0282, "step": 45560 }, { "epoch": 4.81, - "grad_norm": 10.130010604858398, + "grad_norm": 30.836780548095703, "learning_rate": 1.0374907593198859e-05, - "loss": 0.8049, + "loss": 1.4046, "step": 45570 }, { "epoch": 4.81, - "grad_norm": 0.052639372646808624, + "grad_norm": 37.042442321777344, "learning_rate": 1.0372795437744218e-05, - "loss": 0.3776, + "loss": 1.1005, "step": 45580 }, { "epoch": 4.81, - "grad_norm": 22.209747314453125, + "grad_norm": 36.21501922607422, "learning_rate": 1.0370683282289578e-05, - "loss": 0.7348, + "loss": 1.0552, "step": 45590 }, { "epoch": 4.82, - "grad_norm": 13.04053020477295, + "grad_norm": 50.686767578125, "learning_rate": 1.0368571126834937e-05, - "loss": 0.2884, + "loss": 0.7048, "step": 45600 }, { "epoch": 4.82, - "grad_norm": 16.615966796875, + "grad_norm": 38.39825439453125, "learning_rate": 1.0366458971380296e-05, - "loss": 0.6921, + "loss": 1.217, "step": 45610 }, { "epoch": 4.82, - "grad_norm": 8.141674995422363, + "grad_norm": 33.20802688598633, "learning_rate": 1.0364346815925653e-05, - "loss": 0.5877, + "loss": 0.9414, "step": 45620 }, { "epoch": 4.82, - "grad_norm": 20.94713020324707, + "grad_norm": 21.916160583496094, "learning_rate": 1.0362234660471012e-05, - "loss": 0.5855, + "loss": 1.1261, "step": 45630 }, { "epoch": 4.82, - "grad_norm": 0.25021907687187195, + "grad_norm": 25.387798309326172, "learning_rate": 1.036012250501637e-05, - "loss": 0.1605, + "loss": 0.8878, "step": 45640 }, { "epoch": 4.82, - "grad_norm": 0.3503841757774353, + "grad_norm": 52.09630584716797, "learning_rate": 1.035801034956173e-05, - "loss": 0.4467, + "loss": 1.1976, "step": 45650 }, { "epoch": 4.82, - "grad_norm": 10.269610404968262, + "grad_norm": 12.016051292419434, "learning_rate": 1.0355898194107088e-05, - "loss": 0.5167, + "loss": 1.2358, "step": 45660 }, { "epoch": 4.82, - "grad_norm": 6.962228298187256, + "grad_norm": 33.02684020996094, "learning_rate": 1.0353786038652445e-05, - "loss": 0.2326, + "loss": 1.1139, "step": 45670 }, { "epoch": 4.82, - "grad_norm": 8.10323715209961, + "grad_norm": 26.197654724121094, "learning_rate": 1.0351673883197804e-05, - "loss": 0.9468, + "loss": 1.4264, "step": 45680 }, { "epoch": 4.83, - "grad_norm": 0.32388362288475037, + "grad_norm": 9.376297950744629, "learning_rate": 1.0349561727743163e-05, - "loss": 0.3682, + "loss": 0.7943, "step": 45690 }, { "epoch": 4.83, - "grad_norm": 23.085763931274414, + "grad_norm": 45.03410339355469, "learning_rate": 1.0347449572288522e-05, - "loss": 0.5898, + "loss": 1.304, "step": 45700 }, { "epoch": 4.83, - "grad_norm": 0.0716022402048111, + "grad_norm": 18.87541961669922, "learning_rate": 1.034533741683388e-05, - "loss": 0.7238, + "loss": 1.2477, "step": 45710 }, { "epoch": 4.83, - "grad_norm": 17.41427993774414, + "grad_norm": 31.71727752685547, "learning_rate": 1.0343225261379239e-05, - "loss": 0.4858, + "loss": 0.8939, "step": 45720 }, { "epoch": 4.83, - "grad_norm": 20.89113998413086, + "grad_norm": 36.230682373046875, "learning_rate": 1.0341113105924596e-05, - "loss": 0.6719, + "loss": 0.9901, "step": 45730 }, { "epoch": 4.83, - "grad_norm": 14.589207649230957, + "grad_norm": 33.493648529052734, "learning_rate": 1.0339000950469955e-05, - "loss": 0.4766, + "loss": 1.0978, "step": 45740 }, { "epoch": 4.83, - "grad_norm": 15.420493125915527, + "grad_norm": 54.57400894165039, "learning_rate": 1.0336888795015314e-05, - "loss": 0.4773, + "loss": 1.2018, "step": 45750 }, { "epoch": 4.83, - "grad_norm": 26.346052169799805, + "grad_norm": 54.322669982910156, "learning_rate": 1.0334776639560672e-05, - "loss": 0.4476, + "loss": 1.0199, "step": 45760 }, { "epoch": 4.83, - "grad_norm": 13.88053035736084, + "grad_norm": 41.33549880981445, "learning_rate": 1.0332664484106031e-05, - "loss": 0.3296, + "loss": 1.0561, "step": 45770 }, { "epoch": 4.83, - "grad_norm": 29.665300369262695, + "grad_norm": 40.95713424682617, "learning_rate": 1.0330552328651388e-05, - "loss": 0.2691, + "loss": 0.7018, "step": 45780 }, { "epoch": 4.84, - "grad_norm": 11.60539436340332, + "grad_norm": 39.9712028503418, "learning_rate": 1.0328440173196747e-05, - "loss": 0.3233, + "loss": 1.006, "step": 45790 }, { "epoch": 4.84, - "grad_norm": 1.1865144968032837, + "grad_norm": 33.485843658447266, "learning_rate": 1.0326328017742106e-05, - "loss": 0.4122, + "loss": 1.0239, "step": 45800 }, { "epoch": 4.84, - "grad_norm": 25.84142303466797, + "grad_norm": 41.3981819152832, "learning_rate": 1.0324215862287465e-05, - "loss": 0.5343, + "loss": 0.9811, "step": 45810 }, { "epoch": 4.84, - "grad_norm": 31.51738929748535, + "grad_norm": 42.49476623535156, "learning_rate": 1.0322103706832825e-05, - "loss": 0.5573, + "loss": 1.0309, "step": 45820 }, { "epoch": 4.84, - "grad_norm": 6.42460823059082, + "grad_norm": 29.818090438842773, "learning_rate": 1.031999155137818e-05, - "loss": 0.5611, + "loss": 1.1637, "step": 45830 }, { "epoch": 4.84, - "grad_norm": 0.7049174308776855, + "grad_norm": 34.957847595214844, "learning_rate": 1.031787939592354e-05, - "loss": 0.4115, + "loss": 0.8825, "step": 45840 }, { "epoch": 4.84, - "grad_norm": 17.329421997070312, + "grad_norm": 51.271728515625, "learning_rate": 1.03157672404689e-05, - "loss": 0.7636, + "loss": 1.3509, "step": 45850 }, { "epoch": 4.84, - "grad_norm": 15.49108600616455, + "grad_norm": 35.71144485473633, "learning_rate": 1.0313655085014259e-05, - "loss": 0.5665, + "loss": 1.0222, "step": 45860 }, { "epoch": 4.84, - "grad_norm": 8.44483757019043, + "grad_norm": 49.50336456298828, "learning_rate": 1.0311542929559617e-05, - "loss": 0.5545, + "loss": 1.0105, "step": 45870 }, { "epoch": 4.85, - "grad_norm": 15.188454627990723, + "grad_norm": 36.11207962036133, "learning_rate": 1.0309430774104976e-05, - "loss": 0.387, + "loss": 1.2992, "step": 45880 }, { "epoch": 4.85, - "grad_norm": 21.257211685180664, + "grad_norm": 52.02039337158203, "learning_rate": 1.0307318618650333e-05, - "loss": 0.5457, + "loss": 0.9533, "step": 45890 }, { "epoch": 4.85, - "grad_norm": 26.088876724243164, + "grad_norm": 40.494384765625, "learning_rate": 1.0305206463195692e-05, - "loss": 0.5064, + "loss": 1.1223, "step": 45900 }, { "epoch": 4.85, - "grad_norm": 2.4410741329193115, + "grad_norm": 46.928985595703125, "learning_rate": 1.030309430774105e-05, - "loss": 0.4217, + "loss": 0.9737, "step": 45910 }, { "epoch": 4.85, - "grad_norm": 33.1077766418457, + "grad_norm": 52.19123840332031, "learning_rate": 1.030098215228641e-05, - "loss": 0.6379, + "loss": 1.2817, "step": 45920 }, { "epoch": 4.85, - "grad_norm": 37.521366119384766, + "grad_norm": 48.44425964355469, "learning_rate": 1.0298869996831768e-05, - "loss": 0.5184, + "loss": 1.0454, "step": 45930 }, { "epoch": 4.85, - "grad_norm": 10.269450187683105, + "grad_norm": 42.21971893310547, "learning_rate": 1.0296757841377125e-05, - "loss": 0.6276, + "loss": 1.0403, "step": 45940 }, { "epoch": 4.85, - "grad_norm": 7.661107063293457, + "grad_norm": 46.19554901123047, "learning_rate": 1.0294645685922484e-05, - "loss": 0.6948, + "loss": 1.146, "step": 45950 }, { "epoch": 4.85, - "grad_norm": 0.3286706507205963, + "grad_norm": 52.48881530761719, "learning_rate": 1.0292533530467843e-05, - "loss": 0.508, + "loss": 1.2792, "step": 45960 }, { "epoch": 4.85, - "grad_norm": 18.390541076660156, + "grad_norm": 42.477783203125, "learning_rate": 1.0290421375013202e-05, - "loss": 0.3903, + "loss": 1.0347, "step": 45970 }, { "epoch": 4.86, - "grad_norm": 3.3771166801452637, + "grad_norm": 51.70743942260742, "learning_rate": 1.028830921955856e-05, - "loss": 0.3175, + "loss": 0.7717, "step": 45980 }, { "epoch": 4.86, - "grad_norm": 12.904230117797852, + "grad_norm": 43.06668472290039, "learning_rate": 1.028619706410392e-05, - "loss": 0.3485, + "loss": 1.2115, "step": 45990 }, { "epoch": 4.86, - "grad_norm": 12.357331275939941, + "grad_norm": 32.69926452636719, "learning_rate": 1.0284084908649276e-05, - "loss": 0.7918, + "loss": 1.3198, "step": 46000 }, { "epoch": 4.86, - "grad_norm": 14.937165260314941, + "grad_norm": 21.5384521484375, "learning_rate": 1.0281972753194635e-05, - "loss": 0.3905, + "loss": 1.3004, "step": 46010 }, { "epoch": 4.86, - "grad_norm": 13.341933250427246, + "grad_norm": 58.68239212036133, "learning_rate": 1.0279860597739994e-05, - "loss": 0.3807, + "loss": 0.6701, "step": 46020 }, { "epoch": 4.86, - "grad_norm": 31.45918846130371, + "grad_norm": 43.36149978637695, "learning_rate": 1.0277748442285353e-05, - "loss": 0.4747, + "loss": 1.0181, "step": 46030 }, { "epoch": 4.86, - "grad_norm": 2.257810115814209, + "grad_norm": 28.249067306518555, "learning_rate": 1.0275636286830713e-05, - "loss": 0.4322, + "loss": 0.8842, "step": 46040 }, { "epoch": 4.86, - "grad_norm": 11.88900089263916, + "grad_norm": 47.740821838378906, "learning_rate": 1.0273524131376069e-05, - "loss": 0.7015, + "loss": 1.14, "step": 46050 }, { "epoch": 4.86, - "grad_norm": 12.79419231414795, + "grad_norm": 27.23114585876465, "learning_rate": 1.0271411975921427e-05, - "loss": 0.365, + "loss": 0.8679, "step": 46060 }, { "epoch": 4.87, - "grad_norm": 8.007049560546875, + "grad_norm": 28.885459899902344, "learning_rate": 1.0269299820466788e-05, - "loss": 0.4678, + "loss": 0.9149, "step": 46070 }, { "epoch": 4.87, - "grad_norm": 16.83902931213379, + "grad_norm": 36.636409759521484, "learning_rate": 1.0267187665012147e-05, - "loss": 0.3682, + "loss": 1.1095, "step": 46080 }, { "epoch": 4.87, - "grad_norm": 16.150972366333008, + "grad_norm": 42.002601623535156, "learning_rate": 1.0265075509557505e-05, - "loss": 0.7267, + "loss": 1.2664, "step": 46090 }, { "epoch": 4.87, - "grad_norm": 10.102410316467285, + "grad_norm": 34.34117126464844, "learning_rate": 1.0262963354102862e-05, - "loss": 0.6155, + "loss": 1.1454, "step": 46100 }, { "epoch": 4.87, - "grad_norm": 28.66226577758789, + "grad_norm": 58.24482345581055, "learning_rate": 1.0260851198648221e-05, - "loss": 0.3761, + "loss": 1.1768, "step": 46110 }, { "epoch": 4.87, - "grad_norm": 11.777852058410645, + "grad_norm": 32.14752960205078, "learning_rate": 1.025873904319358e-05, - "loss": 0.51, + "loss": 1.3756, "step": 46120 }, { "epoch": 4.87, - "grad_norm": 11.400463104248047, + "grad_norm": 36.648353576660156, "learning_rate": 1.0256626887738939e-05, - "loss": 0.4744, + "loss": 1.0954, "step": 46130 }, { "epoch": 4.87, - "grad_norm": 19.620534896850586, + "grad_norm": 57.20695877075195, "learning_rate": 1.0254514732284298e-05, - "loss": 0.5371, + "loss": 1.325, "step": 46140 }, { "epoch": 4.87, - "grad_norm": 6.168795585632324, + "grad_norm": 30.210559844970703, "learning_rate": 1.0252402576829656e-05, - "loss": 0.4564, + "loss": 0.8757, "step": 46150 }, { "epoch": 4.87, - "grad_norm": 6.57349157333374, + "grad_norm": 32.003963470458984, "learning_rate": 1.0250290421375013e-05, - "loss": 0.4835, + "loss": 1.201, "step": 46160 }, { "epoch": 4.88, - "grad_norm": 9.1922607421875, + "grad_norm": 49.274383544921875, "learning_rate": 1.0248178265920372e-05, - "loss": 0.4126, + "loss": 1.1123, "step": 46170 }, { "epoch": 4.88, - "grad_norm": 4.803238868713379, + "grad_norm": 38.736515045166016, "learning_rate": 1.0246066110465731e-05, - "loss": 0.5969, + "loss": 1.0357, "step": 46180 }, { "epoch": 4.88, - "grad_norm": 4.561303615570068, + "grad_norm": 33.95037078857422, "learning_rate": 1.024395395501109e-05, - "loss": 0.6391, + "loss": 1.108, "step": 46190 }, { "epoch": 4.88, - "grad_norm": 34.41997528076172, + "grad_norm": 44.163265228271484, "learning_rate": 1.0241841799556449e-05, - "loss": 0.6635, + "loss": 1.1571, "step": 46200 }, { "epoch": 4.88, - "grad_norm": 27.93501853942871, + "grad_norm": 34.36402893066406, "learning_rate": 1.0239729644101806e-05, - "loss": 0.2712, + "loss": 1.0478, "step": 46210 }, { "epoch": 4.88, - "grad_norm": 16.36056900024414, + "grad_norm": 32.55719757080078, "learning_rate": 1.0237617488647164e-05, - "loss": 0.3869, + "loss": 0.8109, "step": 46220 }, { "epoch": 4.88, - "grad_norm": 16.854692459106445, + "grad_norm": 33.08249282836914, "learning_rate": 1.0235505333192523e-05, - "loss": 0.6726, + "loss": 0.9611, "step": 46230 }, { "epoch": 4.88, - "grad_norm": 13.20047664642334, + "grad_norm": 43.2131462097168, "learning_rate": 1.0233393177737882e-05, - "loss": 0.6379, + "loss": 0.9149, "step": 46240 }, { "epoch": 4.88, - "grad_norm": 7.824563503265381, + "grad_norm": 23.93702507019043, "learning_rate": 1.023128102228324e-05, - "loss": 0.4515, + "loss": 1.1683, "step": 46250 }, { "epoch": 4.89, - "grad_norm": 10.650543212890625, + "grad_norm": 21.663524627685547, "learning_rate": 1.0229168866828601e-05, - "loss": 0.5319, + "loss": 1.2515, "step": 46260 }, { "epoch": 4.89, - "grad_norm": 8.889284133911133, + "grad_norm": 49.38043975830078, "learning_rate": 1.0227056711373957e-05, - "loss": 0.6579, + "loss": 1.3207, "step": 46270 }, { "epoch": 4.89, - "grad_norm": 0.1208840012550354, + "grad_norm": 31.98504638671875, "learning_rate": 1.0224944555919315e-05, - "loss": 0.6167, + "loss": 1.0922, "step": 46280 }, { "epoch": 4.89, - "grad_norm": 0.7360175251960754, + "grad_norm": 16.17498779296875, "learning_rate": 1.0222832400464676e-05, - "loss": 0.3982, + "loss": 0.6038, "step": 46290 }, { "epoch": 4.89, - "grad_norm": 8.34253215789795, + "grad_norm": 35.7758903503418, "learning_rate": 1.0220720245010035e-05, - "loss": 0.3576, + "loss": 0.7898, "step": 46300 }, { "epoch": 4.89, - "grad_norm": 9.894864082336426, + "grad_norm": 29.134801864624023, "learning_rate": 1.0218608089555393e-05, - "loss": 0.8123, + "loss": 1.1042, "step": 46310 }, { "epoch": 4.89, - "grad_norm": 1.674830436706543, + "grad_norm": 36.37629699707031, "learning_rate": 1.0216495934100749e-05, - "loss": 0.5059, + "loss": 0.8418, "step": 46320 }, { "epoch": 4.89, - "grad_norm": 6.011664390563965, + "grad_norm": 34.36598587036133, "learning_rate": 1.021438377864611e-05, - "loss": 0.5487, + "loss": 1.1181, "step": 46330 }, { "epoch": 4.89, - "grad_norm": 14.718297004699707, + "grad_norm": 19.524444580078125, "learning_rate": 1.0212271623191468e-05, - "loss": 0.5567, + "loss": 1.2096, "step": 46340 }, { "epoch": 4.89, - "grad_norm": 18.622623443603516, + "grad_norm": 36.64012908935547, "learning_rate": 1.0210159467736827e-05, - "loss": 0.5059, + "loss": 1.2016, "step": 46350 }, { "epoch": 4.9, - "grad_norm": 17.60944175720215, + "grad_norm": 40.80829620361328, "learning_rate": 1.0208047312282186e-05, - "loss": 0.4455, + "loss": 1.0284, "step": 46360 }, { "epoch": 4.9, - "grad_norm": 11.660246849060059, + "grad_norm": 28.07799530029297, "learning_rate": 1.0205935156827544e-05, - "loss": 0.4226, + "loss": 1.1789, "step": 46370 }, { "epoch": 4.9, - "grad_norm": 29.744766235351562, + "grad_norm": 43.708534240722656, "learning_rate": 1.0203823001372902e-05, - "loss": 0.6094, + "loss": 1.1998, "step": 46380 }, { "epoch": 4.9, - "grad_norm": 5.342610836029053, + "grad_norm": 35.12884521484375, "learning_rate": 1.020171084591826e-05, - "loss": 0.3747, + "loss": 0.9775, "step": 46390 }, { "epoch": 4.9, - "grad_norm": 1.3157031536102295, + "grad_norm": 28.013967514038086, "learning_rate": 1.0199598690463619e-05, - "loss": 0.2352, + "loss": 0.7897, "step": 46400 }, { "epoch": 4.9, - "grad_norm": 21.151588439941406, + "grad_norm": 28.777572631835938, "learning_rate": 1.0197486535008978e-05, - "loss": 0.4805, + "loss": 1.0298, "step": 46410 }, { "epoch": 4.9, - "grad_norm": 22.018346786499023, + "grad_norm": 64.46845245361328, "learning_rate": 1.0195374379554337e-05, - "loss": 0.7325, + "loss": 1.5668, "step": 46420 }, { "epoch": 4.9, - "grad_norm": 11.576430320739746, + "grad_norm": 32.479705810546875, "learning_rate": 1.0193262224099694e-05, - "loss": 0.8637, + "loss": 1.3319, "step": 46430 }, { "epoch": 4.9, - "grad_norm": 5.458976745605469, + "grad_norm": 45.81898498535156, "learning_rate": 1.0191150068645053e-05, - "loss": 0.642, + "loss": 1.1417, "step": 46440 }, { "epoch": 4.91, - "grad_norm": 12.179601669311523, + "grad_norm": 39.783477783203125, "learning_rate": 1.0189037913190411e-05, - "loss": 0.2633, + "loss": 0.8199, "step": 46450 }, { "epoch": 4.91, - "grad_norm": 23.861421585083008, + "grad_norm": 56.06922912597656, "learning_rate": 1.018692575773577e-05, - "loss": 0.4554, + "loss": 1.1798, "step": 46460 }, { "epoch": 4.91, - "grad_norm": 26.048625946044922, + "grad_norm": 38.01478958129883, "learning_rate": 1.0184813602281129e-05, - "loss": 0.3256, + "loss": 0.8364, "step": 46470 }, { "epoch": 4.91, - "grad_norm": 0.35813453793525696, + "grad_norm": 59.0849494934082, "learning_rate": 1.0182701446826486e-05, - "loss": 0.3588, + "loss": 0.881, "step": 46480 }, { "epoch": 4.91, - "grad_norm": 15.285527229309082, + "grad_norm": 42.939205169677734, "learning_rate": 1.0180589291371845e-05, - "loss": 0.5745, + "loss": 1.3113, "step": 46490 }, { "epoch": 4.91, - "grad_norm": 20.37617301940918, + "grad_norm": 36.821693420410156, "learning_rate": 1.0178477135917203e-05, - "loss": 0.493, + "loss": 0.9338, "step": 46500 }, { "epoch": 4.91, - "grad_norm": 0.3571266531944275, + "grad_norm": 33.0616569519043, "learning_rate": 1.0176364980462562e-05, - "loss": 0.31, + "loss": 0.9535, "step": 46510 }, { "epoch": 4.91, - "grad_norm": 15.033317565917969, + "grad_norm": 41.432647705078125, "learning_rate": 1.0174252825007923e-05, - "loss": 0.7371, + "loss": 1.2737, "step": 46520 }, { "epoch": 4.91, - "grad_norm": 9.977933883666992, + "grad_norm": 45.66921615600586, "learning_rate": 1.0172140669553282e-05, - "loss": 0.8478, + "loss": 1.4812, "step": 46530 }, { "epoch": 4.91, - "grad_norm": 8.60458755493164, + "grad_norm": 28.772228240966797, "learning_rate": 1.0170028514098637e-05, - "loss": 0.5319, + "loss": 1.3849, "step": 46540 }, { "epoch": 4.92, - "grad_norm": 9.2019681930542, + "grad_norm": 39.05488967895508, "learning_rate": 1.0167916358643997e-05, - "loss": 0.3411, + "loss": 1.0355, "step": 46550 }, { "epoch": 4.92, - "grad_norm": 14.532754898071289, + "grad_norm": 53.501434326171875, "learning_rate": 1.0165804203189356e-05, - "loss": 0.4997, + "loss": 1.1595, "step": 46560 }, { "epoch": 4.92, - "grad_norm": 6.566065788269043, + "grad_norm": 24.59358787536621, "learning_rate": 1.0163692047734715e-05, - "loss": 0.7651, + "loss": 1.1535, "step": 46570 }, { "epoch": 4.92, - "grad_norm": 4.9740986824035645, + "grad_norm": 27.46210289001465, "learning_rate": 1.0161579892280074e-05, - "loss": 0.4755, + "loss": 1.0674, "step": 46580 }, { "epoch": 4.92, - "grad_norm": 0.4667200446128845, + "grad_norm": 38.62702178955078, "learning_rate": 1.015946773682543e-05, - "loss": 0.3944, + "loss": 1.1553, "step": 46590 }, { "epoch": 4.92, - "grad_norm": 11.619133949279785, + "grad_norm": 28.91109848022461, "learning_rate": 1.015735558137079e-05, - "loss": 0.3776, + "loss": 1.1013, "step": 46600 }, { "epoch": 4.92, - "grad_norm": 19.69346809387207, + "grad_norm": 29.485910415649414, "learning_rate": 1.0155243425916148e-05, - "loss": 0.4419, + "loss": 1.0362, "step": 46610 }, { "epoch": 4.92, - "grad_norm": 26.76732635498047, + "grad_norm": 46.974552154541016, "learning_rate": 1.0153131270461507e-05, - "loss": 0.4808, + "loss": 1.0594, "step": 46620 }, { "epoch": 4.92, - "grad_norm": 5.615866661071777, + "grad_norm": 51.656150817871094, "learning_rate": 1.0151019115006866e-05, - "loss": 0.5296, + "loss": 1.1733, "step": 46630 }, { "epoch": 4.93, - "grad_norm": 15.4207124710083, + "grad_norm": 38.79384994506836, "learning_rate": 1.0148906959552225e-05, - "loss": 0.2349, + "loss": 0.9685, "step": 46640 }, { "epoch": 4.93, - "grad_norm": 0.5827026963233948, + "grad_norm": 22.5427188873291, "learning_rate": 1.0146794804097582e-05, - "loss": 0.3487, + "loss": 0.8402, "step": 46650 }, { "epoch": 4.93, - "grad_norm": 21.714807510375977, + "grad_norm": 27.98297882080078, "learning_rate": 1.014468264864294e-05, - "loss": 1.0039, + "loss": 1.004, "step": 46660 }, { "epoch": 4.93, - "grad_norm": 3.0530855655670166, + "grad_norm": 44.69047927856445, "learning_rate": 1.01425704931883e-05, - "loss": 0.5485, + "loss": 1.2167, "step": 46670 }, { "epoch": 4.93, - "grad_norm": 0.5006251931190491, + "grad_norm": 45.16090774536133, "learning_rate": 1.0140458337733658e-05, - "loss": 0.3961, + "loss": 1.0404, "step": 46680 }, { "epoch": 4.93, - "grad_norm": 20.168380737304688, + "grad_norm": 64.99828338623047, "learning_rate": 1.0138346182279017e-05, - "loss": 0.4612, + "loss": 1.0252, "step": 46690 }, { "epoch": 4.93, - "grad_norm": 14.389025688171387, + "grad_norm": 37.6053352355957, "learning_rate": 1.0136234026824374e-05, - "loss": 0.2899, + "loss": 0.9681, "step": 46700 }, { "epoch": 4.93, - "grad_norm": 10.036239624023438, + "grad_norm": 26.08725357055664, "learning_rate": 1.0134121871369733e-05, - "loss": 0.3661, + "loss": 1.1291, "step": 46710 }, { "epoch": 4.93, - "grad_norm": 19.760269165039062, + "grad_norm": 42.61625671386719, "learning_rate": 1.0132009715915092e-05, - "loss": 0.5239, + "loss": 1.2339, "step": 46720 }, { "epoch": 4.94, - "grad_norm": 34.94361877441406, + "grad_norm": 45.725337982177734, "learning_rate": 1.012989756046045e-05, - "loss": 0.5133, + "loss": 1.2801, "step": 46730 }, { "epoch": 4.94, - "grad_norm": 13.153271675109863, + "grad_norm": 30.054014205932617, "learning_rate": 1.012778540500581e-05, - "loss": 0.1483, + "loss": 0.9779, "step": 46740 }, { "epoch": 4.94, - "grad_norm": 19.502975463867188, + "grad_norm": 34.79908752441406, "learning_rate": 1.0125673249551166e-05, - "loss": 0.6483, + "loss": 0.9384, "step": 46750 }, { "epoch": 4.94, - "grad_norm": 4.825729846954346, + "grad_norm": 21.504064559936523, "learning_rate": 1.0123561094096525e-05, - "loss": 0.4535, + "loss": 1.0195, "step": 46760 }, { "epoch": 4.94, - "grad_norm": 23.35151481628418, + "grad_norm": 33.82893371582031, "learning_rate": 1.0121448938641885e-05, - "loss": 0.2928, + "loss": 0.8711, "step": 46770 }, { "epoch": 4.94, - "grad_norm": 24.18826675415039, + "grad_norm": 50.98058319091797, "learning_rate": 1.0119336783187244e-05, - "loss": 0.586, + "loss": 0.7989, "step": 46780 }, { "epoch": 4.94, - "grad_norm": 1.8343628644943237, + "grad_norm": 26.986618041992188, "learning_rate": 1.0117224627732603e-05, - "loss": 0.5966, + "loss": 1.181, "step": 46790 }, { "epoch": 4.94, - "grad_norm": 10.061541557312012, + "grad_norm": 43.40706253051758, "learning_rate": 1.0115112472277962e-05, - "loss": 0.5207, + "loss": 1.0768, "step": 46800 }, { "epoch": 4.94, - "grad_norm": 9.466080665588379, + "grad_norm": 35.23027038574219, "learning_rate": 1.0113000316823319e-05, - "loss": 0.4148, + "loss": 0.8858, "step": 46810 }, { "epoch": 4.94, - "grad_norm": 23.481403350830078, + "grad_norm": 55.341163635253906, "learning_rate": 1.0110888161368678e-05, - "loss": 0.6145, + "loss": 1.1199, "step": 46820 }, { "epoch": 4.95, - "grad_norm": 35.108604431152344, + "grad_norm": 34.75661087036133, "learning_rate": 1.0108776005914036e-05, - "loss": 0.5964, + "loss": 1.0126, "step": 46830 }, { "epoch": 4.95, - "grad_norm": 11.662888526916504, + "grad_norm": 44.18712615966797, "learning_rate": 1.0106663850459395e-05, - "loss": 0.5499, + "loss": 1.158, "step": 46840 }, { "epoch": 4.95, - "grad_norm": 29.217077255249023, + "grad_norm": 22.2288875579834, "learning_rate": 1.0104551695004754e-05, - "loss": 0.4976, + "loss": 1.2715, "step": 46850 }, { "epoch": 4.95, - "grad_norm": 86.43962860107422, + "grad_norm": 42.887638092041016, "learning_rate": 1.0102439539550111e-05, - "loss": 0.417, + "loss": 1.1053, "step": 46860 }, { "epoch": 4.95, - "grad_norm": 11.957364082336426, + "grad_norm": 64.29635620117188, "learning_rate": 1.010032738409547e-05, - "loss": 0.6658, + "loss": 1.3699, "step": 46870 }, { "epoch": 4.95, - "grad_norm": 22.10240364074707, + "grad_norm": 40.17837142944336, "learning_rate": 1.0098215228640829e-05, - "loss": 0.3664, + "loss": 0.9061, "step": 46880 }, { "epoch": 4.95, - "grad_norm": 10.196767807006836, + "grad_norm": 41.774139404296875, "learning_rate": 1.0096103073186187e-05, - "loss": 0.4805, + "loss": 1.0915, "step": 46890 }, { "epoch": 4.95, - "grad_norm": 25.126440048217773, + "grad_norm": 18.969377517700195, "learning_rate": 1.0093990917731546e-05, - "loss": 0.5654, + "loss": 0.9208, "step": 46900 }, { "epoch": 4.95, - "grad_norm": 21.76523780822754, + "grad_norm": 34.160770416259766, "learning_rate": 1.0091878762276905e-05, - "loss": 0.5521, + "loss": 1.1905, "step": 46910 }, { "epoch": 4.96, - "grad_norm": 18.97296142578125, + "grad_norm": 45.233829498291016, "learning_rate": 1.0089766606822262e-05, - "loss": 0.621, + "loss": 1.0723, "step": 46920 }, { "epoch": 4.96, - "grad_norm": 15.61257553100586, + "grad_norm": 50.85195541381836, "learning_rate": 1.008765445136762e-05, - "loss": 0.43, + "loss": 0.9053, "step": 46930 }, { "epoch": 4.96, - "grad_norm": 21.78519630432129, + "grad_norm": 31.60886001586914, "learning_rate": 1.008554229591298e-05, - "loss": 0.5736, + "loss": 0.991, "step": 46940 }, { "epoch": 4.96, - "grad_norm": 3.2403206825256348, + "grad_norm": 63.4926643371582, "learning_rate": 1.0083430140458338e-05, - "loss": 0.2612, + "loss": 0.8595, "step": 46950 }, { "epoch": 4.96, - "grad_norm": 0.5270980596542358, + "grad_norm": 40.59198760986328, "learning_rate": 1.0081317985003697e-05, - "loss": 0.4368, + "loss": 0.9904, "step": 46960 }, { "epoch": 4.96, - "grad_norm": 9.190906524658203, + "grad_norm": 34.713951110839844, "learning_rate": 1.0079205829549054e-05, - "loss": 0.4798, + "loss": 1.1743, "step": 46970 }, { "epoch": 4.96, - "grad_norm": 36.012969970703125, + "grad_norm": 53.61758804321289, "learning_rate": 1.0077093674094413e-05, - "loss": 0.5713, + "loss": 0.9859, "step": 46980 }, { "epoch": 4.96, - "grad_norm": 26.339799880981445, + "grad_norm": 35.254737854003906, "learning_rate": 1.0074981518639772e-05, - "loss": 0.7276, + "loss": 1.3009, "step": 46990 }, { "epoch": 4.96, - "grad_norm": 6.998162746429443, + "grad_norm": 41.74681091308594, "learning_rate": 1.0072869363185132e-05, - "loss": 0.4105, + "loss": 0.9201, "step": 47000 }, { "epoch": 4.96, - "grad_norm": 3.593379020690918, + "grad_norm": 30.000083923339844, "learning_rate": 1.0070757207730491e-05, - "loss": 0.5384, + "loss": 0.9611, "step": 47010 }, { "epoch": 4.97, - "grad_norm": 12.299609184265137, + "grad_norm": 50.855682373046875, "learning_rate": 1.006864505227585e-05, - "loss": 0.6347, + "loss": 1.0289, "step": 47020 }, { "epoch": 4.97, - "grad_norm": 10.190253257751465, + "grad_norm": 28.358762741088867, "learning_rate": 1.0066532896821207e-05, - "loss": 0.4858, + "loss": 1.2196, "step": 47030 }, { "epoch": 4.97, - "grad_norm": 20.125537872314453, + "grad_norm": 39.017662048339844, "learning_rate": 1.0064420741366566e-05, - "loss": 0.6571, + "loss": 1.0166, "step": 47040 }, { "epoch": 4.97, - "grad_norm": 10.589876174926758, + "grad_norm": 28.302669525146484, "learning_rate": 1.0062308585911924e-05, - "loss": 0.4779, + "loss": 1.2956, "step": 47050 }, { "epoch": 4.97, - "grad_norm": 2.824517011642456, + "grad_norm": 22.128032684326172, "learning_rate": 1.0060196430457283e-05, - "loss": 0.6392, + "loss": 0.9856, "step": 47060 }, { "epoch": 4.97, - "grad_norm": 27.146818161010742, + "grad_norm": 66.58255767822266, "learning_rate": 1.0058084275002642e-05, - "loss": 0.5541, + "loss": 0.8972, "step": 47070 }, { "epoch": 4.97, - "grad_norm": 29.121536254882812, + "grad_norm": 38.32679748535156, "learning_rate": 1.0055972119547999e-05, - "loss": 0.7727, + "loss": 0.8916, "step": 47080 }, { "epoch": 4.97, - "grad_norm": 0.08253349363803864, + "grad_norm": 29.478736877441406, "learning_rate": 1.0053859964093358e-05, - "loss": 0.5805, + "loss": 1.1329, "step": 47090 }, { "epoch": 4.97, - "grad_norm": 8.872997283935547, + "grad_norm": 49.40427017211914, "learning_rate": 1.0051747808638717e-05, - "loss": 0.5263, + "loss": 0.9509, "step": 47100 }, { "epoch": 4.98, - "grad_norm": 28.5885066986084, + "grad_norm": 21.907690048217773, "learning_rate": 1.0049635653184075e-05, - "loss": 0.3274, + "loss": 0.992, "step": 47110 }, { "epoch": 4.98, - "grad_norm": 12.295877456665039, + "grad_norm": 29.011005401611328, "learning_rate": 1.0047523497729434e-05, - "loss": 0.4175, + "loss": 0.9065, "step": 47120 }, { "epoch": 4.98, - "grad_norm": 8.202981948852539, + "grad_norm": 52.744346618652344, "learning_rate": 1.0045411342274791e-05, - "loss": 0.6068, + "loss": 1.1076, "step": 47130 }, { "epoch": 4.98, - "grad_norm": 12.721558570861816, + "grad_norm": 39.55147933959961, "learning_rate": 1.004329918682015e-05, - "loss": 0.5672, + "loss": 1.3922, "step": 47140 }, { "epoch": 4.98, - "grad_norm": 1.484118938446045, + "grad_norm": 30.608335494995117, "learning_rate": 1.0041187031365509e-05, - "loss": 0.481, + "loss": 0.8886, "step": 47150 }, { "epoch": 4.98, - "grad_norm": 11.538512229919434, + "grad_norm": 47.041622161865234, "learning_rate": 1.0039074875910868e-05, - "loss": 0.3791, + "loss": 0.7458, "step": 47160 }, { "epoch": 4.98, - "grad_norm": 11.695043563842773, + "grad_norm": 18.171266555786133, "learning_rate": 1.0036962720456226e-05, - "loss": 0.4991, + "loss": 1.1129, "step": 47170 }, { "epoch": 4.98, - "grad_norm": 1.5475507974624634, + "grad_norm": 15.747950553894043, "learning_rate": 1.0034850565001585e-05, - "loss": 0.4277, + "loss": 0.8888, "step": 47180 }, { "epoch": 4.98, - "grad_norm": 6.848155498504639, + "grad_norm": 46.389678955078125, "learning_rate": 1.0032738409546942e-05, - "loss": 0.4909, + "loss": 0.9008, "step": 47190 }, { "epoch": 4.98, - "grad_norm": 19.139890670776367, + "grad_norm": 55.28260803222656, "learning_rate": 1.0030626254092301e-05, - "loss": 0.5164, + "loss": 0.933, "step": 47200 }, { "epoch": 4.99, - "grad_norm": 17.301525115966797, + "grad_norm": 34.276161193847656, "learning_rate": 1.002851409863766e-05, - "loss": 0.6567, + "loss": 1.2067, "step": 47210 }, { "epoch": 4.99, - "grad_norm": 28.55348777770996, + "grad_norm": 40.91726303100586, "learning_rate": 1.002640194318302e-05, - "loss": 0.4716, + "loss": 1.3003, "step": 47220 }, { "epoch": 4.99, - "grad_norm": 11.755082130432129, + "grad_norm": 49.505977630615234, "learning_rate": 1.0024289787728379e-05, - "loss": 0.4005, + "loss": 0.7386, "step": 47230 }, { "epoch": 4.99, - "grad_norm": 21.453916549682617, + "grad_norm": 62.15826416015625, "learning_rate": 1.0022177632273734e-05, - "loss": 0.4667, + "loss": 1.0507, "step": 47240 }, { "epoch": 4.99, - "grad_norm": 2.4655332565307617, + "grad_norm": 12.737339973449707, "learning_rate": 1.0020065476819095e-05, - "loss": 0.5609, + "loss": 0.7207, "step": 47250 }, { "epoch": 4.99, - "grad_norm": 15.321693420410156, + "grad_norm": 36.19887161254883, "learning_rate": 1.0017953321364454e-05, - "loss": 0.4095, + "loss": 0.9203, "step": 47260 }, { "epoch": 4.99, - "grad_norm": 15.082256317138672, + "grad_norm": 36.365394592285156, "learning_rate": 1.0015841165909813e-05, - "loss": 0.2932, + "loss": 0.919, "step": 47270 }, { "epoch": 4.99, - "grad_norm": 1.2288562059402466, + "grad_norm": 39.48886489868164, "learning_rate": 1.0013729010455171e-05, - "loss": 0.2886, + "loss": 0.838, "step": 47280 }, { "epoch": 4.99, - "grad_norm": 0.5613579154014587, + "grad_norm": 10.611330032348633, "learning_rate": 1.001161685500053e-05, - "loss": 0.5439, + "loss": 1.0765, "step": 47290 }, { "epoch": 5.0, - "grad_norm": 1.0794559717178345, + "grad_norm": 36.14450454711914, "learning_rate": 1.0009504699545887e-05, - "loss": 0.544, + "loss": 0.9562, "step": 47300 }, { "epoch": 5.0, - "grad_norm": 15.585991859436035, + "grad_norm": 25.227622985839844, "learning_rate": 1.0007392544091246e-05, - "loss": 0.3117, + "loss": 0.7999, "step": 47310 }, { "epoch": 5.0, - "grad_norm": 10.382298469543457, + "grad_norm": 37.2191276550293, "learning_rate": 1.0005280388636605e-05, - "loss": 0.3658, + "loss": 1.1035, "step": 47320 }, { "epoch": 5.0, - "grad_norm": 17.309619903564453, + "grad_norm": 60.24983596801758, "learning_rate": 1.0003168233181963e-05, - "loss": 0.5604, + "loss": 0.925, "step": 47330 }, { "epoch": 5.0, - "grad_norm": 8.779257774353027, + "grad_norm": 37.29782485961914, "learning_rate": 1.0001056077727322e-05, - "loss": 0.2142, + "loss": 0.9679, "step": 47340 }, { "epoch": 5.0, - "eval_accuracy": 0.8912871287128713, - "eval_loss": 0.40979310870170593, - "eval_runtime": 181.7114, - "eval_samples_per_second": 138.957, - "eval_steps_per_second": 17.374, + "eval_accuracy": 0.7921584158415842, + "eval_loss": 0.7730677723884583, + "eval_runtime": 39.5373, + "eval_samples_per_second": 638.637, + "eval_steps_per_second": 79.849, "step": 47345 }, { "epoch": 5.0, - "grad_norm": 0.6483320593833923, + "grad_norm": 18.868343353271484, "learning_rate": 9.998943922272681e-06, - "loss": 0.3378, + "loss": 0.6838, "step": 47350 }, { "epoch": 5.0, - "grad_norm": 1.1385912895202637, + "grad_norm": 49.61851501464844, "learning_rate": 9.996831766818038e-06, - "loss": 0.3241, + "loss": 1.176, "step": 47360 }, { "epoch": 5.0, - "grad_norm": 5.300032138824463, + "grad_norm": 37.31669998168945, "learning_rate": 9.994719611363397e-06, - "loss": 0.4809, + "loss": 0.9426, "step": 47370 }, { "epoch": 5.0, - "grad_norm": 20.15835189819336, + "grad_norm": 44.289180755615234, "learning_rate": 9.992607455908756e-06, - "loss": 0.3692, + "loss": 1.0718, "step": 47380 }, { "epoch": 5.0, - "grad_norm": 16.65275764465332, + "grad_norm": 47.16557693481445, "learning_rate": 9.990495300454114e-06, - "loss": 0.6193, + "loss": 1.2135, "step": 47390 }, { "epoch": 5.01, - "grad_norm": 13.823222160339355, + "grad_norm": 30.365062713623047, "learning_rate": 9.988383144999473e-06, - "loss": 0.502, + "loss": 0.872, "step": 47400 }, { "epoch": 5.01, - "grad_norm": 24.58405113220215, + "grad_norm": 49.313682556152344, "learning_rate": 9.986270989544832e-06, - "loss": 0.6375, + "loss": 1.0156, "step": 47410 }, { "epoch": 5.01, - "grad_norm": 15.789724349975586, + "grad_norm": 33.74360656738281, "learning_rate": 9.984158834090189e-06, - "loss": 0.3913, + "loss": 0.9343, "step": 47420 }, { "epoch": 5.01, - "grad_norm": 12.805619239807129, + "grad_norm": 55.26321792602539, "learning_rate": 9.982046678635548e-06, - "loss": 0.3776, + "loss": 0.9662, "step": 47430 }, { "epoch": 5.01, - "grad_norm": 26.701515197753906, + "grad_norm": 34.911643981933594, "learning_rate": 9.979934523180907e-06, - "loss": 0.4296, + "loss": 1.2605, "step": 47440 }, { "epoch": 5.01, - "grad_norm": 2.1070716381073, + "grad_norm": 24.424638748168945, "learning_rate": 9.977822367726265e-06, - "loss": 0.3024, + "loss": 0.8161, "step": 47450 }, { "epoch": 5.01, - "grad_norm": 15.61855411529541, + "grad_norm": 37.4078254699707, "learning_rate": 9.975710212271624e-06, - "loss": 0.4198, + "loss": 0.8418, "step": 47460 }, { "epoch": 5.01, - "grad_norm": 18.353273391723633, + "grad_norm": 28.800750732421875, "learning_rate": 9.973598056816981e-06, - "loss": 0.453, + "loss": 0.944, "step": 47470 }, { "epoch": 5.01, - "grad_norm": 20.256328582763672, + "grad_norm": 45.67431640625, "learning_rate": 9.971485901362342e-06, - "loss": 0.7011, + "loss": 1.1, "step": 47480 }, { "epoch": 5.02, - "grad_norm": 0.13892492651939392, + "grad_norm": 13.451311111450195, "learning_rate": 9.9693737459077e-06, - "loss": 0.1323, + "loss": 0.8598, "step": 47490 }, { "epoch": 5.02, - "grad_norm": 9.259664535522461, + "grad_norm": 42.40803146362305, "learning_rate": 9.967261590453058e-06, - "loss": 0.6612, + "loss": 1.0052, "step": 47500 }, { "epoch": 5.02, - "grad_norm": 2.689586639404297, + "grad_norm": 42.13151550292969, "learning_rate": 9.965149434998416e-06, - "loss": 0.3638, + "loss": 1.2081, "step": 47510 }, { "epoch": 5.02, - "grad_norm": 10.11136245727539, + "grad_norm": 53.80872344970703, "learning_rate": 9.963037279543775e-06, - "loss": 0.4738, + "loss": 1.1909, "step": 47520 }, { "epoch": 5.02, - "grad_norm": 7.402719020843506, + "grad_norm": 30.11994171142578, "learning_rate": 9.960925124089134e-06, - "loss": 0.3601, + "loss": 0.8402, "step": 47530 }, { "epoch": 5.02, - "grad_norm": 6.939254283905029, + "grad_norm": 27.506439208984375, "learning_rate": 9.958812968634493e-06, - "loss": 0.7064, + "loss": 1.0987, "step": 47540 }, { "epoch": 5.02, - "grad_norm": 11.19847297668457, + "grad_norm": 42.511451721191406, "learning_rate": 9.95670081317985e-06, - "loss": 0.4211, + "loss": 1.081, "step": 47550 }, { "epoch": 5.02, - "grad_norm": 4.187222957611084, + "grad_norm": 46.0655403137207, "learning_rate": 9.954588657725209e-06, - "loss": 0.2144, + "loss": 1.0397, "step": 47560 }, { "epoch": 5.02, - "grad_norm": 28.300390243530273, + "grad_norm": 54.102291107177734, "learning_rate": 9.952476502270569e-06, - "loss": 0.5706, + "loss": 0.9672, "step": 47570 }, { "epoch": 5.02, - "grad_norm": 9.781431198120117, + "grad_norm": 38.848270416259766, "learning_rate": 9.950364346815926e-06, - "loss": 0.5256, + "loss": 1.3615, "step": 47580 }, { "epoch": 5.03, - "grad_norm": 23.43056869506836, + "grad_norm": 51.69754409790039, "learning_rate": 9.948252191361285e-06, - "loss": 0.5307, + "loss": 1.2614, "step": 47590 }, { "epoch": 5.03, - "grad_norm": 15.267600059509277, + "grad_norm": 35.20560836791992, "learning_rate": 9.946140035906644e-06, - "loss": 0.4383, + "loss": 0.8897, "step": 47600 }, { "epoch": 5.03, - "grad_norm": 23.297962188720703, + "grad_norm": 53.00564193725586, "learning_rate": 9.944027880452003e-06, - "loss": 0.5715, + "loss": 0.7693, "step": 47610 }, { "epoch": 5.03, - "grad_norm": 18.993270874023438, + "grad_norm": 55.10994338989258, "learning_rate": 9.941915724997361e-06, - "loss": 0.5557, + "loss": 1.2042, "step": 47620 }, { "epoch": 5.03, - "grad_norm": 0.0417741984128952, + "grad_norm": 31.66183853149414, "learning_rate": 9.939803569542718e-06, - "loss": 0.3021, + "loss": 0.9193, "step": 47630 }, { "epoch": 5.03, - "grad_norm": 0.4497681260108948, + "grad_norm": 23.676116943359375, "learning_rate": 9.937691414088077e-06, - "loss": 0.5495, + "loss": 1.1556, "step": 47640 }, { "epoch": 5.03, - "grad_norm": 38.56824493408203, + "grad_norm": 36.32182693481445, "learning_rate": 9.935579258633436e-06, - "loss": 0.6558, + "loss": 1.0754, "step": 47650 }, { "epoch": 5.03, - "grad_norm": 0.1718117892742157, + "grad_norm": 21.080810546875, "learning_rate": 9.933467103178795e-06, - "loss": 0.3841, + "loss": 1.1799, "step": 47660 }, { "epoch": 5.03, - "grad_norm": 5.761194229125977, + "grad_norm": 42.80767822265625, "learning_rate": 9.931354947724154e-06, - "loss": 0.598, + "loss": 0.9686, "step": 47670 }, { "epoch": 5.04, - "grad_norm": 22.232572555541992, + "grad_norm": 50.21839141845703, "learning_rate": 9.929242792269512e-06, - "loss": 0.6094, + "loss": 1.019, "step": 47680 }, { "epoch": 5.04, - "grad_norm": 16.191226959228516, + "grad_norm": 32.40782928466797, "learning_rate": 9.92713063681487e-06, - "loss": 0.5204, + "loss": 0.9313, "step": 47690 }, { "epoch": 5.04, - "grad_norm": 27.492752075195312, + "grad_norm": 33.28001022338867, "learning_rate": 9.92501848136023e-06, - "loss": 0.3311, + "loss": 1.1611, "step": 47700 }, { "epoch": 5.04, - "grad_norm": 26.370756149291992, + "grad_norm": 48.361724853515625, "learning_rate": 9.922906325905589e-06, - "loss": 0.434, + "loss": 0.8796, "step": 47710 }, { "epoch": 5.04, - "grad_norm": 3.035506010055542, + "grad_norm": 20.479873657226562, "learning_rate": 9.920794170450946e-06, - "loss": 0.4929, + "loss": 1.1913, "step": 47720 }, { "epoch": 5.04, - "grad_norm": 15.439291954040527, + "grad_norm": 38.800819396972656, "learning_rate": 9.918682014996304e-06, - "loss": 0.4517, + "loss": 0.7953, "step": 47730 }, { "epoch": 5.04, - "grad_norm": 5.021493434906006, + "grad_norm": 30.346572875976562, "learning_rate": 9.916569859541663e-06, - "loss": 0.4352, + "loss": 1.0543, "step": 47740 }, { "epoch": 5.04, - "grad_norm": 23.243846893310547, + "grad_norm": 46.47770309448242, "learning_rate": 9.914457704087022e-06, - "loss": 0.3856, + "loss": 1.1714, "step": 47750 }, { "epoch": 5.04, - "grad_norm": 0.15670686960220337, + "grad_norm": 40.77138137817383, "learning_rate": 9.91234554863238e-06, - "loss": 0.2083, + "loss": 0.7568, "step": 47760 }, { "epoch": 5.04, - "grad_norm": 3.0719387531280518, + "grad_norm": 44.763553619384766, "learning_rate": 9.910233393177738e-06, - "loss": 0.2995, + "loss": 0.9045, "step": 47770 }, { "epoch": 5.05, - "grad_norm": 29.398868560791016, + "grad_norm": 46.9594612121582, "learning_rate": 9.908121237723097e-06, - "loss": 0.5159, + "loss": 0.7832, "step": 47780 }, { "epoch": 5.05, - "grad_norm": 5.629676342010498, + "grad_norm": 25.3839168548584, "learning_rate": 9.906009082268455e-06, - "loss": 0.7335, + "loss": 0.8793, "step": 47790 }, { "epoch": 5.05, - "grad_norm": 2.35742449760437, + "grad_norm": 40.23030471801758, "learning_rate": 9.903896926813814e-06, - "loss": 0.5352, + "loss": 1.0475, "step": 47800 }, { "epoch": 5.05, - "grad_norm": 18.754404067993164, + "grad_norm": 40.073448181152344, "learning_rate": 9.901784771359173e-06, - "loss": 0.6501, + "loss": 1.2727, "step": 47810 }, { "epoch": 5.05, - "grad_norm": 11.131853103637695, + "grad_norm": 36.1837158203125, "learning_rate": 9.89967261590453e-06, - "loss": 0.7183, + "loss": 1.3527, "step": 47820 }, { "epoch": 5.05, - "grad_norm": 50.96561050415039, + "grad_norm": 47.21578598022461, "learning_rate": 9.89756046044989e-06, - "loss": 0.4353, + "loss": 1.0716, "step": 47830 }, { "epoch": 5.05, - "grad_norm": 7.603154182434082, + "grad_norm": 58.28582763671875, "learning_rate": 9.89544830499525e-06, - "loss": 0.3825, + "loss": 0.8113, "step": 47840 }, { "epoch": 5.05, - "grad_norm": 3.6078028678894043, + "grad_norm": 48.360595703125, "learning_rate": 9.893336149540606e-06, - "loss": 0.2608, + "loss": 0.8898, "step": 47850 }, { "epoch": 5.05, - "grad_norm": 10.427495002746582, + "grad_norm": 22.970901489257812, "learning_rate": 9.891223994085965e-06, - "loss": 0.322, + "loss": 1.027, "step": 47860 }, { "epoch": 5.06, - "grad_norm": 29.766178131103516, + "grad_norm": 29.725566864013672, "learning_rate": 9.889111838631324e-06, - "loss": 0.3722, + "loss": 0.9851, "step": 47870 }, { "epoch": 5.06, - "grad_norm": 6.356530666351318, + "grad_norm": 36.36155700683594, "learning_rate": 9.886999683176683e-06, - "loss": 0.1699, + "loss": 0.898, "step": 47880 }, { "epoch": 5.06, - "grad_norm": 22.290996551513672, + "grad_norm": 40.060237884521484, "learning_rate": 9.884887527722042e-06, - "loss": 0.3468, + "loss": 0.7197, "step": 47890 }, { "epoch": 5.06, - "grad_norm": 1.587743878364563, + "grad_norm": 35.81745910644531, "learning_rate": 9.882775372267399e-06, - "loss": 0.3858, + "loss": 1.1834, "step": 47900 }, { "epoch": 5.06, - "grad_norm": 19.09904670715332, + "grad_norm": 45.8852424621582, "learning_rate": 9.880663216812757e-06, - "loss": 0.5289, + "loss": 1.355, "step": 47910 }, { "epoch": 5.06, - "grad_norm": 7.354340553283691, + "grad_norm": 32.87152862548828, "learning_rate": 9.878551061358116e-06, - "loss": 0.365, + "loss": 1.0229, "step": 47920 }, { "epoch": 5.06, - "grad_norm": 11.568923950195312, + "grad_norm": 33.31669998168945, "learning_rate": 9.876438905903475e-06, - "loss": 0.2954, + "loss": 1.0487, "step": 47930 }, { "epoch": 5.06, - "grad_norm": 12.284823417663574, + "grad_norm": 37.27903747558594, "learning_rate": 9.874326750448834e-06, - "loss": 0.4892, + "loss": 0.719, "step": 47940 }, { "epoch": 5.06, - "grad_norm": 34.77622985839844, + "grad_norm": 28.758975982666016, "learning_rate": 9.872214594994193e-06, - "loss": 0.3789, + "loss": 1.1376, "step": 47950 }, { "epoch": 5.06, - "grad_norm": 0.9567446112632751, + "grad_norm": 49.921871185302734, "learning_rate": 9.870102439539551e-06, - "loss": 0.5283, + "loss": 1.2236, "step": 47960 }, { "epoch": 5.07, - "grad_norm": 8.812776565551758, + "grad_norm": 25.545618057250977, "learning_rate": 9.86799028408491e-06, - "loss": 0.3397, + "loss": 0.8705, "step": 47970 }, { "epoch": 5.07, - "grad_norm": 4.496835231781006, + "grad_norm": 22.394804000854492, "learning_rate": 9.865878128630269e-06, - "loss": 0.6243, + "loss": 1.086, "step": 47980 }, { "epoch": 5.07, - "grad_norm": 10.879115104675293, + "grad_norm": 43.9923210144043, "learning_rate": 9.863765973175626e-06, - "loss": 0.4454, + "loss": 0.9315, "step": 47990 }, { "epoch": 5.07, - "grad_norm": 8.339377403259277, + "grad_norm": 6.996351718902588, "learning_rate": 9.861653817720985e-06, - "loss": 0.5715, + "loss": 1.2015, "step": 48000 }, { "epoch": 5.07, - "grad_norm": 0.16320478916168213, + "grad_norm": 33.66905212402344, "learning_rate": 9.859541662266344e-06, - "loss": 0.3946, + "loss": 1.0043, "step": 48010 }, { "epoch": 5.07, - "grad_norm": 66.63186645507812, + "grad_norm": 39.992984771728516, "learning_rate": 9.857429506811702e-06, - "loss": 0.5221, + "loss": 1.0984, "step": 48020 }, { "epoch": 5.07, - "grad_norm": 28.17325210571289, + "grad_norm": 50.92475509643555, "learning_rate": 9.855317351357061e-06, - "loss": 0.3568, + "loss": 0.8732, "step": 48030 }, { "epoch": 5.07, - "grad_norm": 30.56814956665039, + "grad_norm": 65.44302368164062, "learning_rate": 9.853205195902418e-06, - "loss": 0.7519, + "loss": 1.2574, "step": 48040 }, { "epoch": 5.07, - "grad_norm": 22.775943756103516, + "grad_norm": 26.846132278442383, "learning_rate": 9.851093040447779e-06, - "loss": 0.5231, + "loss": 1.2068, "step": 48050 }, { "epoch": 5.08, - "grad_norm": 19.288341522216797, + "grad_norm": 37.105491638183594, "learning_rate": 9.848980884993137e-06, - "loss": 0.6303, + "loss": 1.002, "step": 48060 }, { "epoch": 5.08, - "grad_norm": 5.9363555908203125, + "grad_norm": 43.46920394897461, "learning_rate": 9.846868729538494e-06, - "loss": 0.4026, + "loss": 0.7416, "step": 48070 }, { "epoch": 5.08, - "grad_norm": 0.42941945791244507, + "grad_norm": 4.398128509521484, "learning_rate": 9.844756574083853e-06, - "loss": 0.4916, + "loss": 0.9667, "step": 48080 }, { "epoch": 5.08, - "grad_norm": 14.257588386535645, + "grad_norm": 36.47820281982422, "learning_rate": 9.842644418629212e-06, - "loss": 0.4465, + "loss": 0.9788, "step": 48090 }, { "epoch": 5.08, - "grad_norm": 6.852436542510986, + "grad_norm": 50.499271392822266, "learning_rate": 9.84053226317457e-06, - "loss": 0.638, + "loss": 1.3384, "step": 48100 }, { "epoch": 5.08, - "grad_norm": 19.45360565185547, + "grad_norm": 57.405967712402344, "learning_rate": 9.83842010771993e-06, - "loss": 0.3462, + "loss": 0.7742, "step": 48110 }, { "epoch": 5.08, - "grad_norm": 12.010282516479492, + "grad_norm": 33.40798568725586, "learning_rate": 9.836307952265287e-06, - "loss": 0.2549, + "loss": 0.8134, "step": 48120 }, { "epoch": 5.08, - "grad_norm": 19.235605239868164, + "grad_norm": 50.327388763427734, "learning_rate": 9.834195796810645e-06, - "loss": 0.3813, + "loss": 1.162, "step": 48130 }, { "epoch": 5.08, - "grad_norm": 6.561572074890137, + "grad_norm": 40.0759162902832, "learning_rate": 9.832083641356004e-06, - "loss": 0.3758, + "loss": 1.0087, "step": 48140 }, { "epoch": 5.09, - "grad_norm": 8.254953384399414, + "grad_norm": 27.472553253173828, "learning_rate": 9.829971485901363e-06, - "loss": 0.5478, + "loss": 0.8775, "step": 48150 }, { "epoch": 5.09, - "grad_norm": 1.8663042783737183, + "grad_norm": 24.07335090637207, "learning_rate": 9.827859330446722e-06, - "loss": 0.5064, + "loss": 0.8536, "step": 48160 }, { "epoch": 5.09, - "grad_norm": 35.38358688354492, + "grad_norm": 46.02684783935547, "learning_rate": 9.82574717499208e-06, - "loss": 0.4053, + "loss": 0.924, "step": 48170 }, { "epoch": 5.09, - "grad_norm": 25.35805320739746, + "grad_norm": 52.470558166503906, "learning_rate": 9.82363501953744e-06, - "loss": 0.5924, + "loss": 1.2311, "step": 48180 }, { "epoch": 5.09, - "grad_norm": 1.1377310752868652, + "grad_norm": 53.885799407958984, "learning_rate": 9.821522864082798e-06, - "loss": 0.4912, + "loss": 1.1066, "step": 48190 }, { "epoch": 5.09, - "grad_norm": 0.7237533330917358, + "grad_norm": 29.141321182250977, "learning_rate": 9.819410708628155e-06, - "loss": 0.3243, + "loss": 0.7925, "step": 48200 }, { "epoch": 5.09, - "grad_norm": 25.491806030273438, + "grad_norm": 55.21980285644531, "learning_rate": 9.817298553173514e-06, - "loss": 0.5584, + "loss": 1.2894, "step": 48210 }, { "epoch": 5.09, - "grad_norm": 1.3584221601486206, + "grad_norm": 22.457632064819336, "learning_rate": 9.815186397718873e-06, - "loss": 0.4176, + "loss": 0.9024, "step": 48220 }, { "epoch": 5.09, - "grad_norm": 27.422407150268555, + "grad_norm": 36.30323028564453, "learning_rate": 9.813074242264232e-06, - "loss": 0.391, + "loss": 0.9224, "step": 48230 }, { "epoch": 5.09, - "grad_norm": 26.003604888916016, + "grad_norm": 47.2932014465332, "learning_rate": 9.81096208680959e-06, - "loss": 0.6172, + "loss": 0.9739, "step": 48240 }, { "epoch": 5.1, - "grad_norm": 10.690865516662598, + "grad_norm": 44.319679260253906, "learning_rate": 9.808849931354949e-06, - "loss": 0.6271, + "loss": 1.0718, "step": 48250 }, { "epoch": 5.1, - "grad_norm": 2.2372806072235107, + "grad_norm": 41.07077407836914, "learning_rate": 9.806737775900306e-06, - "loss": 0.3359, + "loss": 1.0052, "step": 48260 }, { "epoch": 5.1, - "grad_norm": 20.638105392456055, + "grad_norm": 41.40342712402344, "learning_rate": 9.804625620445665e-06, - "loss": 0.4063, + "loss": 0.9155, "step": 48270 }, { "epoch": 5.1, - "grad_norm": 4.584375381469727, + "grad_norm": 13.394914627075195, "learning_rate": 9.802513464991024e-06, - "loss": 0.4903, + "loss": 0.7433, "step": 48280 }, { "epoch": 5.1, - "grad_norm": 7.110608100891113, + "grad_norm": 37.647640228271484, "learning_rate": 9.800401309536383e-06, - "loss": 0.5714, + "loss": 1.0635, "step": 48290 }, { "epoch": 5.1, - "grad_norm": 3.0616583824157715, + "grad_norm": 44.70704650878906, "learning_rate": 9.798289154081741e-06, - "loss": 0.5094, + "loss": 1.0942, "step": 48300 }, { "epoch": 5.1, - "grad_norm": 13.354927062988281, + "grad_norm": 42.11625671386719, "learning_rate": 9.7961769986271e-06, - "loss": 0.3635, + "loss": 0.9302, "step": 48310 }, { "epoch": 5.1, - "grad_norm": 16.283432006835938, + "grad_norm": 38.58860397338867, "learning_rate": 9.794064843172459e-06, - "loss": 0.4209, + "loss": 1.2142, "step": 48320 }, { "epoch": 5.1, - "grad_norm": 22.30356788635254, + "grad_norm": 41.76991271972656, "learning_rate": 9.791952687717818e-06, - "loss": 0.4547, + "loss": 0.9771, "step": 48330 }, { "epoch": 5.11, - "grad_norm": 26.056964874267578, + "grad_norm": 45.66868591308594, "learning_rate": 9.789840532263175e-06, - "loss": 0.6081, + "loss": 1.2157, "step": 48340 }, { "epoch": 5.11, - "grad_norm": 14.199923515319824, + "grad_norm": 38.05160903930664, "learning_rate": 9.787728376808534e-06, - "loss": 0.3476, + "loss": 1.3103, "step": 48350 }, { "epoch": 5.11, - "grad_norm": 2.8406243324279785, + "grad_norm": 38.188167572021484, "learning_rate": 9.785616221353892e-06, - "loss": 0.4505, + "loss": 1.1339, "step": 48360 }, { "epoch": 5.11, - "grad_norm": 0.27509841322898865, + "grad_norm": 53.58070755004883, "learning_rate": 9.783504065899251e-06, - "loss": 0.3655, + "loss": 0.8417, "step": 48370 }, { "epoch": 5.11, - "grad_norm": 12.255203247070312, + "grad_norm": 40.64924240112305, "learning_rate": 9.78139191044461e-06, - "loss": 0.5817, + "loss": 1.0464, "step": 48380 }, { "epoch": 5.11, - "grad_norm": 21.83768653869629, + "grad_norm": 35.447078704833984, "learning_rate": 9.779279754989967e-06, - "loss": 0.5343, + "loss": 1.0453, "step": 48390 }, { "epoch": 5.11, - "grad_norm": 19.975183486938477, + "grad_norm": 45.04167938232422, "learning_rate": 9.777167599535326e-06, - "loss": 0.8181, + "loss": 0.9144, "step": 48400 }, { "epoch": 5.11, - "grad_norm": 7.5409932136535645, + "grad_norm": 33.89945602416992, "learning_rate": 9.775055444080686e-06, - "loss": 0.6406, + "loss": 1.135, "step": 48410 }, { "epoch": 5.11, - "grad_norm": 13.235109329223633, + "grad_norm": 36.64363479614258, "learning_rate": 9.772943288626043e-06, - "loss": 0.4049, + "loss": 0.6125, "step": 48420 }, { "epoch": 5.11, - "grad_norm": 13.923023223876953, + "grad_norm": 29.160049438476562, "learning_rate": 9.770831133171402e-06, - "loss": 0.3359, + "loss": 1.013, "step": 48430 }, { "epoch": 5.12, - "grad_norm": 13.417984008789062, + "grad_norm": 59.8115234375, "learning_rate": 9.76871897771676e-06, - "loss": 0.5433, + "loss": 1.2639, "step": 48440 }, { "epoch": 5.12, - "grad_norm": 9.959576606750488, + "grad_norm": 39.03518295288086, "learning_rate": 9.76660682226212e-06, - "loss": 0.479, + "loss": 0.7781, "step": 48450 }, { "epoch": 5.12, - "grad_norm": 23.363296508789062, + "grad_norm": 30.391143798828125, "learning_rate": 9.764494666807478e-06, - "loss": 0.3772, + "loss": 1.1555, "step": 48460 }, { "epoch": 5.12, - "grad_norm": 20.115705490112305, + "grad_norm": 22.377477645874023, "learning_rate": 9.762382511352835e-06, - "loss": 0.4847, + "loss": 1.0643, "step": 48470 }, { "epoch": 5.12, - "grad_norm": 15.370817184448242, + "grad_norm": 35.17335891723633, "learning_rate": 9.760270355898194e-06, - "loss": 0.6443, + "loss": 1.4183, "step": 48480 }, { "epoch": 5.12, - "grad_norm": 10.946415901184082, + "grad_norm": 45.23221969604492, "learning_rate": 9.758158200443553e-06, - "loss": 0.4865, + "loss": 1.2595, "step": 48490 }, { "epoch": 5.12, - "grad_norm": 32.92060470581055, + "grad_norm": 62.2541389465332, "learning_rate": 9.756046044988912e-06, - "loss": 0.4585, + "loss": 1.2033, "step": 48500 }, { "epoch": 5.12, - "grad_norm": 7.591558456420898, + "grad_norm": 33.21445846557617, "learning_rate": 9.75393388953427e-06, - "loss": 0.8306, + "loss": 1.1947, "step": 48510 }, { "epoch": 5.12, - "grad_norm": 1.1570745706558228, + "grad_norm": 46.12285614013672, "learning_rate": 9.75182173407963e-06, - "loss": 0.373, + "loss": 1.0579, "step": 48520 }, { "epoch": 5.13, - "grad_norm": 15.430618286132812, + "grad_norm": 34.256011962890625, "learning_rate": 9.749709578624988e-06, - "loss": 0.2877, + "loss": 0.99, "step": 48530 }, { "epoch": 5.13, - "grad_norm": 0.3425959646701813, + "grad_norm": 22.589855194091797, "learning_rate": 9.747597423170347e-06, - "loss": 0.4109, + "loss": 1.01, "step": 48540 }, { "epoch": 5.13, - "grad_norm": 18.19405746459961, + "grad_norm": 33.38079071044922, "learning_rate": 9.745485267715704e-06, - "loss": 0.5264, + "loss": 1.3012, "step": 48550 }, { "epoch": 5.13, - "grad_norm": 9.276871681213379, + "grad_norm": 45.531166076660156, "learning_rate": 9.743373112261063e-06, - "loss": 0.2352, + "loss": 1.0248, "step": 48560 }, { "epoch": 5.13, - "grad_norm": 19.33701515197754, + "grad_norm": 43.69300079345703, "learning_rate": 9.741260956806422e-06, - "loss": 0.4982, + "loss": 0.956, "step": 48570 }, { "epoch": 5.13, - "grad_norm": 0.2601071298122406, + "grad_norm": 10.889063835144043, "learning_rate": 9.73914880135178e-06, - "loss": 0.3098, + "loss": 0.7273, "step": 48580 }, { "epoch": 5.13, - "grad_norm": 43.28247833251953, + "grad_norm": 63.27993392944336, "learning_rate": 9.737036645897139e-06, - "loss": 0.5849, + "loss": 1.2392, "step": 48590 }, { "epoch": 5.13, - "grad_norm": 0.813027560710907, + "grad_norm": 25.275346755981445, "learning_rate": 9.734924490442498e-06, - "loss": 0.4868, + "loss": 1.1921, "step": 48600 }, { "epoch": 5.13, - "grad_norm": 31.35352325439453, + "grad_norm": 20.78528594970703, "learning_rate": 9.732812334987855e-06, - "loss": 0.3021, + "loss": 0.8338, "step": 48610 }, { "epoch": 5.13, - "grad_norm": 9.447846412658691, + "grad_norm": 56.588558197021484, "learning_rate": 9.730700179533214e-06, - "loss": 0.4636, + "loss": 1.1028, "step": 48620 }, { "epoch": 5.14, - "grad_norm": 11.060932159423828, + "grad_norm": 36.63840103149414, "learning_rate": 9.728588024078574e-06, - "loss": 0.5153, + "loss": 1.0001, "step": 48630 }, { "epoch": 5.14, - "grad_norm": 3.4055628776550293, + "grad_norm": 35.21450424194336, "learning_rate": 9.726475868623931e-06, - "loss": 0.3296, + "loss": 0.8031, "step": 48640 }, { "epoch": 5.14, - "grad_norm": 6.447974681854248, + "grad_norm": 62.982418060302734, "learning_rate": 9.72436371316929e-06, - "loss": 0.6097, + "loss": 1.2871, "step": 48650 }, { "epoch": 5.14, - "grad_norm": 13.147216796875, + "grad_norm": 30.141260147094727, "learning_rate": 9.722251557714649e-06, - "loss": 0.4292, + "loss": 1.1824, "step": 48660 }, { "epoch": 5.14, - "grad_norm": 6.290570259094238, + "grad_norm": 20.917814254760742, "learning_rate": 9.720139402260008e-06, - "loss": 0.3074, + "loss": 0.8646, "step": 48670 }, { "epoch": 5.14, - "grad_norm": 13.71650505065918, + "grad_norm": 30.028871536254883, "learning_rate": 9.718027246805366e-06, - "loss": 0.546, + "loss": 1.1596, "step": 48680 }, { "epoch": 5.14, - "grad_norm": 5.49813985824585, + "grad_norm": 39.4668083190918, "learning_rate": 9.715915091350724e-06, - "loss": 0.5027, + "loss": 0.9727, "step": 48690 }, { "epoch": 5.14, - "grad_norm": 0.3760433495044708, + "grad_norm": 52.16600036621094, "learning_rate": 9.713802935896082e-06, - "loss": 0.3202, + "loss": 0.9528, "step": 48700 }, { "epoch": 5.14, - "grad_norm": 5.1690592765808105, + "grad_norm": 30.632905960083008, "learning_rate": 9.711690780441441e-06, - "loss": 0.2911, + "loss": 0.9299, "step": 48710 }, { "epoch": 5.15, - "grad_norm": 6.545773029327393, + "grad_norm": 45.549007415771484, "learning_rate": 9.7095786249868e-06, - "loss": 0.597, + "loss": 1.274, "step": 48720 }, { "epoch": 5.15, - "grad_norm": 6.910141944885254, + "grad_norm": 12.027886390686035, "learning_rate": 9.707466469532159e-06, - "loss": 0.5894, + "loss": 1.159, "step": 48730 }, { "epoch": 5.15, - "grad_norm": 18.705631256103516, + "grad_norm": 44.41301727294922, "learning_rate": 9.705354314077516e-06, - "loss": 0.4316, + "loss": 1.177, "step": 48740 }, { "epoch": 5.15, - "grad_norm": 1.8353008031845093, + "grad_norm": 54.731422424316406, "learning_rate": 9.703242158622875e-06, - "loss": 0.737, + "loss": 0.9863, "step": 48750 }, { "epoch": 5.15, - "grad_norm": 24.13262939453125, + "grad_norm": 61.25548553466797, "learning_rate": 9.701130003168235e-06, - "loss": 0.3236, + "loss": 0.8193, "step": 48760 }, { "epoch": 5.15, - "grad_norm": 5.6297526359558105, + "grad_norm": 41.938758850097656, "learning_rate": 9.699017847713592e-06, - "loss": 0.4759, + "loss": 0.9579, "step": 48770 }, { "epoch": 5.15, - "grad_norm": 1.075718879699707, + "grad_norm": 8.162074089050293, "learning_rate": 9.696905692258951e-06, - "loss": 0.4576, + "loss": 0.9488, "step": 48780 }, { "epoch": 5.15, - "grad_norm": 0.39534127712249756, + "grad_norm": 40.44190216064453, "learning_rate": 9.69479353680431e-06, - "loss": 0.2808, + "loss": 0.966, "step": 48790 }, { "epoch": 5.15, - "grad_norm": 5.559189319610596, + "grad_norm": 18.03466796875, "learning_rate": 9.692681381349668e-06, - "loss": 0.3612, + "loss": 1.1813, "step": 48800 }, { "epoch": 5.15, - "grad_norm": 10.621688842773438, + "grad_norm": 35.143985748291016, "learning_rate": 9.690569225895027e-06, - "loss": 0.5019, + "loss": 0.9798, "step": 48810 }, { "epoch": 5.16, - "grad_norm": 7.700271129608154, + "grad_norm": 39.96937561035156, "learning_rate": 9.688457070440386e-06, - "loss": 0.5948, + "loss": 1.1297, "step": 48820 }, { "epoch": 5.16, - "grad_norm": 15.019633293151855, + "grad_norm": 35.47991943359375, "learning_rate": 9.686344914985743e-06, - "loss": 0.3994, + "loss": 0.9788, "step": 48830 }, { "epoch": 5.16, - "grad_norm": 12.34622859954834, + "grad_norm": 40.440635681152344, "learning_rate": 9.684232759531102e-06, - "loss": 0.6429, + "loss": 1.0704, "step": 48840 }, { "epoch": 5.16, - "grad_norm": 7.110516548156738, + "grad_norm": 33.0858268737793, "learning_rate": 9.68212060407646e-06, - "loss": 0.4793, + "loss": 0.9195, "step": 48850 }, { "epoch": 5.16, - "grad_norm": 3.8318705558776855, + "grad_norm": 50.335933685302734, "learning_rate": 9.68000844862182e-06, - "loss": 0.1497, + "loss": 0.9809, "step": 48860 }, { "epoch": 5.16, - "grad_norm": 2.0139503479003906, + "grad_norm": 28.36315155029297, "learning_rate": 9.677896293167178e-06, - "loss": 0.471, + "loss": 0.7734, "step": 48870 }, { "epoch": 5.16, - "grad_norm": 16.737924575805664, + "grad_norm": 28.555099487304688, "learning_rate": 9.675784137712537e-06, - "loss": 0.2795, + "loss": 0.8844, "step": 48880 }, { "epoch": 5.16, - "grad_norm": 19.56946563720703, + "grad_norm": 37.8875617980957, "learning_rate": 9.673671982257896e-06, - "loss": 0.3629, + "loss": 1.0437, "step": 48890 }, { "epoch": 5.16, - "grad_norm": 7.287001132965088, + "grad_norm": 33.870750427246094, "learning_rate": 9.671559826803255e-06, - "loss": 0.6409, + "loss": 1.0351, "step": 48900 }, { "epoch": 5.17, - "grad_norm": 18.59352684020996, + "grad_norm": 51.354766845703125, "learning_rate": 9.669447671348612e-06, - "loss": 0.4906, + "loss": 1.1718, "step": 48910 }, { "epoch": 5.17, - "grad_norm": 5.464441776275635, + "grad_norm": 42.83990478515625, "learning_rate": 9.66733551589397e-06, - "loss": 0.3874, + "loss": 1.0821, "step": 48920 }, { "epoch": 5.17, - "grad_norm": 11.725661277770996, + "grad_norm": 40.20964431762695, "learning_rate": 9.665223360439329e-06, - "loss": 0.4264, + "loss": 0.8513, "step": 48930 }, { "epoch": 5.17, - "grad_norm": 2.5676541328430176, + "grad_norm": 35.12289047241211, "learning_rate": 9.663111204984688e-06, - "loss": 0.3211, + "loss": 0.8176, "step": 48940 }, { "epoch": 5.17, - "grad_norm": 12.897418975830078, + "grad_norm": 40.65340805053711, "learning_rate": 9.660999049530047e-06, - "loss": 0.4499, + "loss": 1.0383, "step": 48950 }, { "epoch": 5.17, - "grad_norm": 3.609104871749878, + "grad_norm": 22.06920051574707, "learning_rate": 9.658886894075404e-06, - "loss": 0.4876, + "loss": 0.8404, "step": 48960 }, { "epoch": 5.17, - "grad_norm": 29.97471809387207, + "grad_norm": 37.72929382324219, "learning_rate": 9.656774738620763e-06, - "loss": 0.2964, + "loss": 1.1149, "step": 48970 }, { "epoch": 5.17, - "grad_norm": 0.23184144496917725, + "grad_norm": 30.531267166137695, "learning_rate": 9.654662583166123e-06, - "loss": 0.2844, + "loss": 0.8309, "step": 48980 }, { "epoch": 5.17, - "grad_norm": 9.691534042358398, + "grad_norm": 25.413230895996094, "learning_rate": 9.65255042771148e-06, - "loss": 0.5667, + "loss": 0.8836, "step": 48990 }, { "epoch": 5.17, - "grad_norm": 16.72736358642578, + "grad_norm": 39.52100372314453, "learning_rate": 9.650438272256839e-06, - "loss": 0.5164, + "loss": 1.2159, "step": 49000 }, { "epoch": 5.18, - "grad_norm": 2.7651453018188477, + "grad_norm": 24.604204177856445, "learning_rate": 9.648326116802198e-06, - "loss": 0.5879, + "loss": 1.1126, "step": 49010 }, { "epoch": 5.18, - "grad_norm": 18.988550186157227, + "grad_norm": 37.07781219482422, "learning_rate": 9.646213961347556e-06, - "loss": 0.5671, + "loss": 0.9273, "step": 49020 }, { "epoch": 5.18, - "grad_norm": 11.964319229125977, + "grad_norm": 23.815603256225586, "learning_rate": 9.644101805892915e-06, - "loss": 0.2429, + "loss": 0.7342, "step": 49030 }, { "epoch": 5.18, - "grad_norm": 6.2081756591796875, + "grad_norm": 21.29592514038086, "learning_rate": 9.641989650438272e-06, - "loss": 0.4317, + "loss": 0.9888, "step": 49040 }, { "epoch": 5.18, - "grad_norm": 11.3425874710083, + "grad_norm": 25.59989356994629, "learning_rate": 9.639877494983631e-06, - "loss": 0.2572, + "loss": 0.7951, "step": 49050 }, { "epoch": 5.18, - "grad_norm": 48.533451080322266, + "grad_norm": 33.78629684448242, "learning_rate": 9.63776533952899e-06, - "loss": 0.6185, + "loss": 0.6835, "step": 49060 }, { "epoch": 5.18, - "grad_norm": 4.620457649230957, + "grad_norm": 41.66531753540039, "learning_rate": 9.635653184074349e-06, - "loss": 0.7182, + "loss": 1.2144, "step": 49070 }, { "epoch": 5.18, - "grad_norm": 6.226663112640381, + "grad_norm": 38.599830627441406, "learning_rate": 9.633541028619707e-06, - "loss": 0.5592, + "loss": 0.8789, "step": 49080 }, { "epoch": 5.18, - "grad_norm": 32.66249465942383, + "grad_norm": 53.48627471923828, "learning_rate": 9.631428873165066e-06, - "loss": 0.4183, + "loss": 1.135, "step": 49090 }, { "epoch": 5.19, - "grad_norm": 7.300009727478027, + "grad_norm": 29.22730827331543, "learning_rate": 9.629316717710423e-06, - "loss": 0.4135, + "loss": 0.7287, "step": 49100 }, { "epoch": 5.19, - "grad_norm": 1.852638840675354, + "grad_norm": 39.01308822631836, "learning_rate": 9.627204562255784e-06, - "loss": 0.3832, + "loss": 1.0854, "step": 49110 }, { "epoch": 5.19, - "grad_norm": 8.725873947143555, + "grad_norm": 45.472206115722656, "learning_rate": 9.625092406801141e-06, - "loss": 0.4179, + "loss": 0.951, "step": 49120 }, { "epoch": 5.19, - "grad_norm": 12.171109199523926, + "grad_norm": 46.14775848388672, "learning_rate": 9.6229802513465e-06, - "loss": 0.5131, + "loss": 1.1429, "step": 49130 }, { "epoch": 5.19, - "grad_norm": 9.387446403503418, + "grad_norm": 41.0487174987793, "learning_rate": 9.620868095891858e-06, - "loss": 0.4284, + "loss": 1.2865, "step": 49140 }, { "epoch": 5.19, - "grad_norm": 0.5039924383163452, + "grad_norm": 29.500812530517578, "learning_rate": 9.618755940437217e-06, - "loss": 0.3651, + "loss": 0.7744, "step": 49150 }, { "epoch": 5.19, - "grad_norm": 16.53580665588379, + "grad_norm": 40.75536346435547, "learning_rate": 9.616643784982576e-06, - "loss": 0.7132, + "loss": 1.3344, "step": 49160 }, { "epoch": 5.19, - "grad_norm": 16.66823387145996, + "grad_norm": 41.43025207519531, "learning_rate": 9.614531629527935e-06, - "loss": 0.4206, + "loss": 0.9717, "step": 49170 }, { "epoch": 5.19, - "grad_norm": 13.863875389099121, + "grad_norm": 29.83102798461914, "learning_rate": 9.612419474073292e-06, - "loss": 0.5732, + "loss": 1.0304, "step": 49180 }, { "epoch": 5.19, - "grad_norm": 13.038004875183105, + "grad_norm": 53.021732330322266, "learning_rate": 9.61030731861865e-06, - "loss": 0.5822, + "loss": 1.1588, "step": 49190 }, { "epoch": 5.2, - "grad_norm": 10.80629825592041, + "grad_norm": 50.43772888183594, "learning_rate": 9.60819516316401e-06, - "loss": 0.2299, + "loss": 1.1532, "step": 49200 }, { "epoch": 5.2, - "grad_norm": 14.029953002929688, + "grad_norm": 30.08051300048828, "learning_rate": 9.606083007709368e-06, - "loss": 0.3717, + "loss": 0.9079, "step": 49210 }, { "epoch": 5.2, - "grad_norm": 11.560868263244629, + "grad_norm": 47.5964469909668, "learning_rate": 9.603970852254727e-06, - "loss": 0.4516, + "loss": 1.2505, "step": 49220 }, { "epoch": 5.2, - "grad_norm": 9.41895866394043, + "grad_norm": 28.142118453979492, "learning_rate": 9.601858696800084e-06, - "loss": 0.3159, + "loss": 0.7774, "step": 49230 }, { "epoch": 5.2, - "grad_norm": 17.449861526489258, + "grad_norm": 33.69509506225586, "learning_rate": 9.599746541345445e-06, - "loss": 0.4439, + "loss": 1.1809, "step": 49240 }, { "epoch": 5.2, - "grad_norm": 22.90813446044922, + "grad_norm": 40.9803466796875, "learning_rate": 9.597634385890803e-06, - "loss": 0.6206, + "loss": 1.2545, "step": 49250 }, { "epoch": 5.2, - "grad_norm": 1.4577782154083252, + "grad_norm": 43.339080810546875, "learning_rate": 9.59552223043616e-06, - "loss": 0.2958, + "loss": 0.9747, "step": 49260 }, { "epoch": 5.2, - "grad_norm": 14.956207275390625, + "grad_norm": 41.71015167236328, "learning_rate": 9.59341007498152e-06, - "loss": 0.5518, + "loss": 1.0516, "step": 49270 }, { "epoch": 5.2, - "grad_norm": 11.488340377807617, + "grad_norm": 41.413360595703125, "learning_rate": 9.591297919526878e-06, - "loss": 0.5161, + "loss": 0.7636, "step": 49280 }, { "epoch": 5.21, - "grad_norm": 16.850473403930664, + "grad_norm": 40.67976760864258, "learning_rate": 9.589185764072237e-06, - "loss": 0.4525, + "loss": 0.9304, "step": 49290 }, { "epoch": 5.21, - "grad_norm": 0.4375789761543274, + "grad_norm": 48.7393913269043, "learning_rate": 9.587073608617595e-06, - "loss": 0.4894, + "loss": 0.8457, "step": 49300 }, { "epoch": 5.21, - "grad_norm": 24.592458724975586, + "grad_norm": 57.98183059692383, "learning_rate": 9.584961453162953e-06, - "loss": 0.472, + "loss": 1.1523, "step": 49310 }, { "epoch": 5.21, - "grad_norm": 13.384469032287598, + "grad_norm": 31.255233764648438, "learning_rate": 9.582849297708311e-06, - "loss": 0.5349, + "loss": 1.0897, "step": 49320 }, { "epoch": 5.21, - "grad_norm": 17.73847770690918, + "grad_norm": 37.79438018798828, "learning_rate": 9.580737142253672e-06, - "loss": 0.2568, + "loss": 0.8324, "step": 49330 }, { "epoch": 5.21, - "grad_norm": 79.24414825439453, + "grad_norm": 41.587581634521484, "learning_rate": 9.578624986799029e-06, - "loss": 0.6903, + "loss": 1.1818, "step": 49340 }, { "epoch": 5.21, - "grad_norm": 12.452071189880371, + "grad_norm": 39.45947265625, "learning_rate": 9.576512831344388e-06, - "loss": 0.5206, + "loss": 1.1125, "step": 49350 }, { "epoch": 5.21, - "grad_norm": 8.895042419433594, + "grad_norm": 35.78514862060547, "learning_rate": 9.574400675889746e-06, - "loss": 0.5737, + "loss": 1.1577, "step": 49360 }, { "epoch": 5.21, - "grad_norm": 5.486940383911133, + "grad_norm": 16.19192123413086, "learning_rate": 9.572288520435105e-06, - "loss": 0.3123, + "loss": 0.6998, "step": 49370 }, { "epoch": 5.21, - "grad_norm": 11.126075744628906, + "grad_norm": 60.26152801513672, "learning_rate": 9.570176364980464e-06, - "loss": 0.454, + "loss": 1.1708, "step": 49380 }, { "epoch": 5.22, - "grad_norm": 39.30353546142578, + "grad_norm": 42.28976821899414, "learning_rate": 9.568064209525821e-06, - "loss": 0.476, + "loss": 1.0206, "step": 49390 }, { "epoch": 5.22, - "grad_norm": 3.4949281215667725, + "grad_norm": 32.12023162841797, "learning_rate": 9.56595205407118e-06, - "loss": 0.3869, + "loss": 1.0065, "step": 49400 }, { "epoch": 5.22, - "grad_norm": 3.2185184955596924, + "grad_norm": 18.527099609375, "learning_rate": 9.563839898616539e-06, - "loss": 0.5792, + "loss": 1.021, "step": 49410 }, { "epoch": 5.22, - "grad_norm": 23.201148986816406, + "grad_norm": 28.017236709594727, "learning_rate": 9.561727743161897e-06, - "loss": 0.3677, + "loss": 1.0547, "step": 49420 }, { "epoch": 5.22, - "grad_norm": 0.34507906436920166, + "grad_norm": 24.22516632080078, "learning_rate": 9.559615587707256e-06, - "loss": 0.3278, + "loss": 0.9528, "step": 49430 }, { "epoch": 5.22, - "grad_norm": 12.1491060256958, + "grad_norm": 50.6638298034668, "learning_rate": 9.557503432252615e-06, - "loss": 0.5544, + "loss": 1.227, "step": 49440 }, { "epoch": 5.22, - "grad_norm": 23.27668571472168, + "grad_norm": 43.52532958984375, "learning_rate": 9.555391276797972e-06, - "loss": 0.5164, + "loss": 0.8369, "step": 49450 }, { "epoch": 5.22, - "grad_norm": 8.559365272521973, + "grad_norm": 41.281612396240234, "learning_rate": 9.553279121343333e-06, - "loss": 0.3533, + "loss": 0.9805, "step": 49460 }, { "epoch": 5.22, - "grad_norm": 18.429367065429688, + "grad_norm": 32.15427017211914, "learning_rate": 9.55116696588869e-06, - "loss": 0.7639, + "loss": 1.1674, "step": 49470 }, { "epoch": 5.23, - "grad_norm": 14.83450984954834, + "grad_norm": 26.81307601928711, "learning_rate": 9.549054810434048e-06, - "loss": 0.6714, + "loss": 1.222, "step": 49480 }, { "epoch": 5.23, - "grad_norm": 2.3259377479553223, + "grad_norm": 47.16569900512695, "learning_rate": 9.546942654979407e-06, - "loss": 0.3345, + "loss": 0.9175, "step": 49490 }, { "epoch": 5.23, - "grad_norm": 0.34767261147499084, + "grad_norm": 26.9356689453125, "learning_rate": 9.544830499524766e-06, - "loss": 0.3186, + "loss": 0.7601, "step": 49500 }, { "epoch": 5.23, - "grad_norm": 0.20075853168964386, + "grad_norm": 17.943605422973633, "learning_rate": 9.542718344070125e-06, - "loss": 0.5568, + "loss": 1.0783, "step": 49510 }, { "epoch": 5.23, - "grad_norm": 18.495397567749023, + "grad_norm": 27.30900764465332, "learning_rate": 9.540606188615484e-06, - "loss": 0.3855, + "loss": 0.9885, "step": 49520 }, { "epoch": 5.23, - "grad_norm": 2.5463309288024902, + "grad_norm": 35.41693878173828, "learning_rate": 9.53849403316084e-06, - "loss": 0.6971, + "loss": 1.0023, "step": 49530 }, { "epoch": 5.23, - "grad_norm": 11.615067481994629, + "grad_norm": 18.510150909423828, "learning_rate": 9.5363818777062e-06, - "loss": 0.2812, + "loss": 1.0513, "step": 49540 }, { "epoch": 5.23, - "grad_norm": 2.135444402694702, + "grad_norm": 37.90729522705078, "learning_rate": 9.534269722251558e-06, - "loss": 0.3447, + "loss": 0.7501, "step": 49550 }, { "epoch": 5.23, - "grad_norm": 14.717588424682617, + "grad_norm": 29.273099899291992, "learning_rate": 9.532157566796917e-06, - "loss": 0.4961, + "loss": 0.725, "step": 49560 }, { "epoch": 5.23, - "grad_norm": 12.921419143676758, + "grad_norm": 50.181190490722656, "learning_rate": 9.530045411342276e-06, - "loss": 0.3613, + "loss": 1.1026, "step": 49570 }, { "epoch": 5.24, - "grad_norm": 10.706605911254883, + "grad_norm": 38.39128875732422, "learning_rate": 9.527933255887633e-06, - "loss": 0.5326, + "loss": 1.1352, "step": 49580 }, { "epoch": 5.24, - "grad_norm": 20.264421463012695, + "grad_norm": 21.233856201171875, "learning_rate": 9.525821100432993e-06, - "loss": 0.2543, + "loss": 0.832, "step": 49590 }, { "epoch": 5.24, - "grad_norm": 6.105223655700684, + "grad_norm": 35.048866271972656, "learning_rate": 9.523708944978352e-06, - "loss": 0.2998, + "loss": 0.6503, "step": 49600 }, { "epoch": 5.24, - "grad_norm": 20.699764251708984, + "grad_norm": 53.141822814941406, "learning_rate": 9.52159678952371e-06, - "loss": 0.3451, + "loss": 1.2776, "step": 49610 }, { "epoch": 5.24, - "grad_norm": 1.1745197772979736, + "grad_norm": 33.177223205566406, "learning_rate": 9.519484634069068e-06, - "loss": 0.385, + "loss": 1.0455, "step": 49620 }, { "epoch": 5.24, - "grad_norm": 0.6033855080604553, + "grad_norm": 25.541791915893555, "learning_rate": 9.517372478614427e-06, - "loss": 0.3465, + "loss": 1.0083, "step": 49630 }, { "epoch": 5.24, - "grad_norm": 2.7211732864379883, + "grad_norm": 9.427568435668945, "learning_rate": 9.515260323159786e-06, - "loss": 0.2362, + "loss": 0.9517, "step": 49640 }, { "epoch": 5.24, - "grad_norm": 28.2486515045166, + "grad_norm": 40.902679443359375, "learning_rate": 9.513148167705144e-06, - "loss": 0.6885, + "loss": 1.3681, "step": 49650 }, { "epoch": 5.24, - "grad_norm": 21.547571182250977, + "grad_norm": 43.43300247192383, "learning_rate": 9.511036012250501e-06, - "loss": 0.4061, + "loss": 1.1105, "step": 49660 }, { "epoch": 5.25, - "grad_norm": 23.85540771484375, + "grad_norm": 65.54910278320312, "learning_rate": 9.50892385679586e-06, - "loss": 0.4191, + "loss": 1.2754, "step": 49670 }, { "epoch": 5.25, - "grad_norm": 14.858915328979492, + "grad_norm": 36.27178192138672, "learning_rate": 9.50681170134122e-06, - "loss": 0.4178, + "loss": 1.064, "step": 49680 }, { "epoch": 5.25, - "grad_norm": 0.5057363510131836, + "grad_norm": 28.418777465820312, "learning_rate": 9.504699545886578e-06, - "loss": 0.2361, + "loss": 0.9939, "step": 49690 }, { "epoch": 5.25, - "grad_norm": 0.021682744845747948, + "grad_norm": 43.21427536010742, "learning_rate": 9.502587390431936e-06, - "loss": 0.3688, + "loss": 0.7949, "step": 49700 }, { "epoch": 5.25, - "grad_norm": 0.033401187509298325, + "grad_norm": 7.813255786895752, "learning_rate": 9.500475234977295e-06, - "loss": 0.4806, + "loss": 0.9179, "step": 49710 }, { "epoch": 5.25, - "grad_norm": 0.7540371417999268, + "grad_norm": 36.176883697509766, "learning_rate": 9.498363079522654e-06, - "loss": 0.3992, + "loss": 1.0392, "step": 49720 }, { "epoch": 5.25, - "grad_norm": 0.04087066650390625, + "grad_norm": 28.41990089416504, "learning_rate": 9.496250924068013e-06, - "loss": 0.5824, + "loss": 1.0287, "step": 49730 }, { "epoch": 5.25, - "grad_norm": 18.333637237548828, + "grad_norm": 58.11787033081055, "learning_rate": 9.494138768613372e-06, - "loss": 0.5025, + "loss": 1.0785, "step": 49740 }, { "epoch": 5.25, - "grad_norm": 9.684039115905762, + "grad_norm": 40.557987213134766, "learning_rate": 9.492026613158729e-06, - "loss": 0.3282, + "loss": 0.9633, "step": 49750 }, { "epoch": 5.26, - "grad_norm": 18.01024055480957, + "grad_norm": 40.38203048706055, "learning_rate": 9.489914457704087e-06, - "loss": 0.3155, + "loss": 1.1208, "step": 49760 }, { "epoch": 5.26, - "grad_norm": 9.943675994873047, + "grad_norm": 35.17744064331055, "learning_rate": 9.487802302249446e-06, - "loss": 0.3645, + "loss": 1.0862, "step": 49770 }, { "epoch": 5.26, - "grad_norm": 6.910629749298096, + "grad_norm": 46.2150993347168, "learning_rate": 9.485690146794805e-06, - "loss": 0.4916, + "loss": 0.832, "step": 49780 }, { "epoch": 5.26, - "grad_norm": 36.3655891418457, + "grad_norm": 30.441438674926758, "learning_rate": 9.483577991340164e-06, - "loss": 0.3908, + "loss": 0.8867, "step": 49790 }, { "epoch": 5.26, - "grad_norm": 15.807403564453125, + "grad_norm": 28.686195373535156, "learning_rate": 9.481465835885521e-06, - "loss": 0.4657, + "loss": 0.994, "step": 49800 }, { "epoch": 5.26, - "grad_norm": 17.607646942138672, + "grad_norm": 38.63307571411133, "learning_rate": 9.479353680430881e-06, - "loss": 0.5305, + "loss": 0.8673, "step": 49810 }, { "epoch": 5.26, - "grad_norm": 77.84807586669922, + "grad_norm": 53.67400360107422, "learning_rate": 9.47724152497624e-06, - "loss": 0.3987, + "loss": 1.187, "step": 49820 }, { "epoch": 5.26, - "grad_norm": 7.937980651855469, + "grad_norm": 35.387760162353516, "learning_rate": 9.475129369521597e-06, - "loss": 0.4916, + "loss": 1.036, "step": 49830 }, { "epoch": 5.26, - "grad_norm": 8.227229118347168, + "grad_norm": 46.87686538696289, "learning_rate": 9.473017214066956e-06, - "loss": 0.5128, + "loss": 1.1195, "step": 49840 }, { "epoch": 5.26, - "grad_norm": 15.26123046875, + "grad_norm": 46.97945022583008, "learning_rate": 9.470905058612315e-06, - "loss": 0.3661, + "loss": 0.9619, "step": 49850 }, { "epoch": 5.27, - "grad_norm": 12.61691665649414, + "grad_norm": 52.892852783203125, "learning_rate": 9.468792903157674e-06, - "loss": 0.3804, + "loss": 1.2139, "step": 49860 }, { "epoch": 5.27, - "grad_norm": 7.622680187225342, + "grad_norm": 33.705265045166016, "learning_rate": 9.466680747703032e-06, - "loss": 0.5051, + "loss": 0.8992, "step": 49870 }, { "epoch": 5.27, - "grad_norm": 9.374452590942383, + "grad_norm": 40.22445297241211, "learning_rate": 9.46456859224839e-06, - "loss": 0.6169, + "loss": 1.2016, "step": 49880 }, { "epoch": 5.27, - "grad_norm": 6.576601028442383, + "grad_norm": 48.71103286743164, "learning_rate": 9.462456436793748e-06, - "loss": 0.3927, + "loss": 0.9083, "step": 49890 }, { "epoch": 5.27, - "grad_norm": 13.94214153289795, + "grad_norm": 35.2799072265625, "learning_rate": 9.460344281339107e-06, - "loss": 0.49, + "loss": 0.9214, "step": 49900 }, { "epoch": 5.27, - "grad_norm": 14.715486526489258, + "grad_norm": 41.99944305419922, "learning_rate": 9.458232125884466e-06, - "loss": 0.652, + "loss": 1.2202, "step": 49910 }, { "epoch": 5.27, - "grad_norm": 12.554106712341309, + "grad_norm": 56.0595703125, "learning_rate": 9.456119970429825e-06, - "loss": 0.6719, + "loss": 1.0583, "step": 49920 }, { "epoch": 5.27, - "grad_norm": 8.70584487915039, + "grad_norm": 43.2274284362793, "learning_rate": 9.454007814975183e-06, - "loss": 0.4596, + "loss": 1.077, "step": 49930 }, { "epoch": 5.27, - "grad_norm": 14.182904243469238, + "grad_norm": 38.869911193847656, "learning_rate": 9.451895659520542e-06, - "loss": 0.4351, + "loss": 1.0861, "step": 49940 }, { "epoch": 5.28, - "grad_norm": 13.028611183166504, + "grad_norm": 31.6461181640625, "learning_rate": 9.449783504065901e-06, - "loss": 0.3692, + "loss": 1.0618, "step": 49950 }, { "epoch": 5.28, - "grad_norm": 0.2030719518661499, + "grad_norm": 30.848051071166992, "learning_rate": 9.447671348611258e-06, - "loss": 0.3352, + "loss": 1.1297, "step": 49960 }, { "epoch": 5.28, - "grad_norm": 30.006855010986328, + "grad_norm": 24.874055862426758, "learning_rate": 9.445559193156617e-06, - "loss": 0.3652, + "loss": 0.8991, "step": 49970 }, { "epoch": 5.28, - "grad_norm": 18.056188583374023, + "grad_norm": 67.08466339111328, "learning_rate": 9.443447037701976e-06, - "loss": 0.366, + "loss": 1.1514, "step": 49980 }, { "epoch": 5.28, - "grad_norm": 19.411815643310547, + "grad_norm": 55.65642166137695, "learning_rate": 9.441334882247334e-06, - "loss": 0.1937, + "loss": 0.7632, "step": 49990 }, { "epoch": 5.28, - "grad_norm": 16.113346099853516, + "grad_norm": 41.04975128173828, "learning_rate": 9.439222726792693e-06, - "loss": 0.4905, + "loss": 1.0295, "step": 50000 }, { "epoch": 5.28, - "grad_norm": 9.504487037658691, + "grad_norm": 41.2843132019043, "learning_rate": 9.437110571338052e-06, - "loss": 0.703, + "loss": 1.1319, "step": 50010 }, { "epoch": 5.28, - "grad_norm": 17.418060302734375, + "grad_norm": 39.53182601928711, "learning_rate": 9.434998415883409e-06, - "loss": 0.4506, + "loss": 1.221, "step": 50020 }, { "epoch": 5.28, - "grad_norm": 11.943988800048828, + "grad_norm": 53.76824188232422, "learning_rate": 9.432886260428768e-06, - "loss": 0.8548, + "loss": 1.0083, "step": 50030 }, { "epoch": 5.28, - "grad_norm": 14.077067375183105, + "grad_norm": 52.78230667114258, "learning_rate": 9.430774104974126e-06, - "loss": 0.5685, + "loss": 1.0654, "step": 50040 }, { "epoch": 5.29, - "grad_norm": 54.3554573059082, + "grad_norm": 30.73558235168457, "learning_rate": 9.428661949519485e-06, - "loss": 0.1638, + "loss": 0.8192, "step": 50050 }, { "epoch": 5.29, - "grad_norm": 25.266950607299805, + "grad_norm": 62.43533706665039, "learning_rate": 9.426549794064844e-06, - "loss": 0.54, + "loss": 0.9984, "step": 50060 }, { "epoch": 5.29, - "grad_norm": 28.13621711730957, + "grad_norm": 28.69002342224121, "learning_rate": 9.424437638610203e-06, - "loss": 0.4201, + "loss": 0.9722, "step": 50070 }, { "epoch": 5.29, - "grad_norm": 3.0126304626464844, + "grad_norm": 39.373287200927734, "learning_rate": 9.422325483155562e-06, - "loss": 0.4159, + "loss": 1.0017, "step": 50080 }, { "epoch": 5.29, - "grad_norm": 7.283440589904785, + "grad_norm": 25.90550422668457, "learning_rate": 9.42021332770092e-06, - "loss": 0.4916, + "loss": 1.2374, "step": 50090 }, { "epoch": 5.29, - "grad_norm": 7.51474666595459, + "grad_norm": 23.04694366455078, "learning_rate": 9.418101172246277e-06, - "loss": 0.4707, + "loss": 0.9804, "step": 50100 }, { "epoch": 5.29, - "grad_norm": 13.301701545715332, + "grad_norm": 37.882713317871094, "learning_rate": 9.415989016791636e-06, - "loss": 0.589, + "loss": 1.0703, "step": 50110 }, { "epoch": 5.29, - "grad_norm": 5.761237144470215, + "grad_norm": 38.42633056640625, "learning_rate": 9.413876861336995e-06, - "loss": 0.5176, + "loss": 1.0477, "step": 50120 }, { "epoch": 5.29, - "grad_norm": 0.25549283623695374, + "grad_norm": 27.69415283203125, "learning_rate": 9.411764705882354e-06, - "loss": 0.4782, + "loss": 0.6977, "step": 50130 }, { "epoch": 5.3, - "grad_norm": 9.864827156066895, + "grad_norm": 34.585670471191406, "learning_rate": 9.409652550427713e-06, - "loss": 0.2982, + "loss": 1.0617, "step": 50140 }, { "epoch": 5.3, - "grad_norm": 5.851785659790039, + "grad_norm": 44.194602966308594, "learning_rate": 9.40754039497307e-06, - "loss": 0.4224, + "loss": 0.9788, "step": 50150 }, { "epoch": 5.3, - "grad_norm": 5.422621250152588, + "grad_norm": 23.489843368530273, "learning_rate": 9.40542823951843e-06, - "loss": 0.281, + "loss": 0.7805, "step": 50160 }, { "epoch": 5.3, - "grad_norm": 0.7913298010826111, + "grad_norm": 41.38981628417969, "learning_rate": 9.403316084063789e-06, - "loss": 0.366, + "loss": 0.8589, "step": 50170 }, { "epoch": 5.3, - "grad_norm": 19.358198165893555, + "grad_norm": 29.45360565185547, "learning_rate": 9.401203928609146e-06, - "loss": 0.473, + "loss": 1.0085, "step": 50180 }, { "epoch": 5.3, - "grad_norm": 2.104586124420166, + "grad_norm": 33.46861267089844, "learning_rate": 9.399091773154505e-06, - "loss": 0.3096, + "loss": 1.0923, "step": 50190 }, { "epoch": 5.3, - "grad_norm": 24.119308471679688, + "grad_norm": 32.22721862792969, "learning_rate": 9.396979617699864e-06, - "loss": 0.3272, + "loss": 0.7151, "step": 50200 }, { "epoch": 5.3, - "grad_norm": 1.139988660812378, + "grad_norm": 25.97492790222168, "learning_rate": 9.394867462245222e-06, - "loss": 0.2931, + "loss": 0.8637, "step": 50210 }, { "epoch": 5.3, - "grad_norm": 12.127799987792969, + "grad_norm": 30.51158905029297, "learning_rate": 9.392755306790581e-06, - "loss": 0.5339, + "loss": 1.0655, "step": 50220 }, { "epoch": 5.3, - "grad_norm": 12.426790237426758, + "grad_norm": 46.89535140991211, "learning_rate": 9.390643151335938e-06, - "loss": 0.235, + "loss": 1.0492, "step": 50230 }, { "epoch": 5.31, - "grad_norm": 21.25438117980957, + "grad_norm": 52.18736267089844, "learning_rate": 9.388530995881297e-06, - "loss": 0.7205, + "loss": 1.249, "step": 50240 }, { "epoch": 5.31, - "grad_norm": 14.171690940856934, + "grad_norm": 44.80299377441406, "learning_rate": 9.386418840426656e-06, - "loss": 0.2843, + "loss": 0.7829, "step": 50250 }, { "epoch": 5.31, - "grad_norm": 18.686279296875, + "grad_norm": 49.56471633911133, "learning_rate": 9.384306684972015e-06, - "loss": 0.3079, + "loss": 1.203, "step": 50260 }, { "epoch": 5.31, - "grad_norm": 4.982436180114746, + "grad_norm": 59.504634857177734, "learning_rate": 9.382194529517373e-06, - "loss": 0.4338, + "loss": 1.0281, "step": 50270 }, { "epoch": 5.31, - "grad_norm": 17.204465866088867, + "grad_norm": 42.2376708984375, "learning_rate": 9.380082374062732e-06, - "loss": 0.6717, + "loss": 1.03, "step": 50280 }, { "epoch": 5.31, - "grad_norm": 2.3053908348083496, + "grad_norm": 45.7606086730957, "learning_rate": 9.377970218608091e-06, - "loss": 0.447, + "loss": 0.7572, "step": 50290 }, { "epoch": 5.31, - "grad_norm": 7.8853325843811035, + "grad_norm": 20.83535385131836, "learning_rate": 9.37585806315345e-06, - "loss": 0.4395, + "loss": 1.0865, "step": 50300 }, { "epoch": 5.31, - "grad_norm": 14.453161239624023, + "grad_norm": 50.354557037353516, "learning_rate": 9.373745907698807e-06, - "loss": 0.3372, + "loss": 1.0504, "step": 50310 }, { "epoch": 5.31, - "grad_norm": 26.662662506103516, + "grad_norm": 42.359615325927734, "learning_rate": 9.371633752244166e-06, - "loss": 0.4781, + "loss": 1.0467, "step": 50320 }, { "epoch": 5.32, - "grad_norm": 12.261356353759766, + "grad_norm": 32.574214935302734, "learning_rate": 9.369521596789524e-06, - "loss": 0.3322, + "loss": 1.0593, "step": 50330 }, { "epoch": 5.32, - "grad_norm": 9.356599807739258, + "grad_norm": 47.59310531616211, "learning_rate": 9.367409441334883e-06, - "loss": 0.1811, + "loss": 1.0908, "step": 50340 }, { "epoch": 5.32, - "grad_norm": 1.2235743999481201, + "grad_norm": 41.759857177734375, "learning_rate": 9.365297285880242e-06, - "loss": 0.2355, + "loss": 0.8458, "step": 50350 }, { "epoch": 5.32, - "grad_norm": 29.827394485473633, + "grad_norm": 46.412288665771484, "learning_rate": 9.3631851304256e-06, - "loss": 0.3395, + "loss": 1.169, "step": 50360 }, { "epoch": 5.32, - "grad_norm": 1.5555604696273804, + "grad_norm": 38.251705169677734, "learning_rate": 9.361072974970958e-06, - "loss": 0.1674, + "loss": 0.6801, "step": 50370 }, { "epoch": 5.32, - "grad_norm": 1.0456409454345703, + "grad_norm": 46.11215591430664, "learning_rate": 9.358960819516317e-06, - "loss": 0.2658, + "loss": 0.8966, "step": 50380 }, { "epoch": 5.32, - "grad_norm": 0.7713815569877625, + "grad_norm": 54.247413635253906, "learning_rate": 9.356848664061677e-06, - "loss": 0.4611, + "loss": 0.9773, "step": 50390 }, { "epoch": 5.32, - "grad_norm": 0.5666186213493347, + "grad_norm": 23.746129989624023, "learning_rate": 9.354736508607034e-06, - "loss": 0.2683, + "loss": 0.776, "step": 50400 }, { "epoch": 5.32, - "grad_norm": 2.1993253231048584, + "grad_norm": 41.80250549316406, "learning_rate": 9.352624353152393e-06, - "loss": 0.4918, + "loss": 1.1806, "step": 50410 }, { "epoch": 5.32, - "grad_norm": 31.47117805480957, + "grad_norm": 31.053367614746094, "learning_rate": 9.350512197697752e-06, - "loss": 0.2936, + "loss": 1.0145, "step": 50420 }, { "epoch": 5.33, - "grad_norm": 6.276465892791748, + "grad_norm": 38.96528244018555, "learning_rate": 9.34840004224311e-06, - "loss": 0.5023, + "loss": 1.0082, "step": 50430 }, { "epoch": 5.33, - "grad_norm": 30.017925262451172, + "grad_norm": 66.17829895019531, "learning_rate": 9.34628788678847e-06, - "loss": 0.3972, + "loss": 0.7982, "step": 50440 }, { "epoch": 5.33, - "grad_norm": 0.6189675331115723, + "grad_norm": 28.719186782836914, "learning_rate": 9.344175731333826e-06, - "loss": 0.434, + "loss": 1.0589, "step": 50450 }, { "epoch": 5.33, - "grad_norm": 30.960430145263672, + "grad_norm": 52.165931701660156, "learning_rate": 9.342063575879185e-06, - "loss": 0.5781, + "loss": 0.809, "step": 50460 }, { "epoch": 5.33, - "grad_norm": 0.06474853307008743, + "grad_norm": 28.903059005737305, "learning_rate": 9.339951420424544e-06, - "loss": 0.5509, + "loss": 0.6693, "step": 50470 }, { "epoch": 5.33, - "grad_norm": 0.6519417762756348, + "grad_norm": 43.12187957763672, "learning_rate": 9.337839264969903e-06, - "loss": 0.2843, + "loss": 0.9928, "step": 50480 }, { "epoch": 5.33, - "grad_norm": 5.860268592834473, + "grad_norm": 53.06489181518555, "learning_rate": 9.335727109515261e-06, - "loss": 0.5371, + "loss": 1.0928, "step": 50490 }, { "epoch": 5.33, - "grad_norm": 19.739768981933594, + "grad_norm": 44.125457763671875, "learning_rate": 9.333614954060618e-06, - "loss": 0.1957, + "loss": 0.7814, "step": 50500 }, { "epoch": 5.33, - "grad_norm": 21.48714256286621, + "grad_norm": 62.73551940917969, "learning_rate": 9.331502798605977e-06, - "loss": 0.356, + "loss": 0.7327, "step": 50510 }, { "epoch": 5.34, - "grad_norm": 8.45692253112793, + "grad_norm": 27.25141716003418, "learning_rate": 9.329390643151338e-06, - "loss": 0.3326, + "loss": 1.0688, "step": 50520 }, { "epoch": 5.34, - "grad_norm": 39.824039459228516, + "grad_norm": 40.085289001464844, "learning_rate": 9.327278487696695e-06, - "loss": 0.4831, + "loss": 0.8669, "step": 50530 }, { "epoch": 5.34, - "grad_norm": 44.47932434082031, + "grad_norm": 44.456687927246094, "learning_rate": 9.325166332242054e-06, - "loss": 0.3679, + "loss": 0.9654, "step": 50540 }, { "epoch": 5.34, - "grad_norm": 22.952503204345703, + "grad_norm": 50.24147415161133, "learning_rate": 9.323054176787412e-06, - "loss": 0.4166, + "loss": 1.0017, "step": 50550 }, { "epoch": 5.34, - "grad_norm": 24.90715980529785, + "grad_norm": 53.64213562011719, "learning_rate": 9.320942021332771e-06, - "loss": 0.4221, + "loss": 1.245, "step": 50560 }, { "epoch": 5.34, - "grad_norm": 12.269614219665527, + "grad_norm": 27.818748474121094, "learning_rate": 9.31882986587813e-06, - "loss": 0.57, + "loss": 0.9174, "step": 50570 }, { "epoch": 5.34, - "grad_norm": 15.806278228759766, + "grad_norm": 37.27303695678711, "learning_rate": 9.316717710423487e-06, - "loss": 0.4661, + "loss": 1.3585, "step": 50580 }, { "epoch": 5.34, - "grad_norm": 23.794004440307617, + "grad_norm": 38.702938079833984, "learning_rate": 9.314605554968846e-06, - "loss": 0.358, + "loss": 0.7777, "step": 50590 }, { "epoch": 5.34, - "grad_norm": 22.739999771118164, + "grad_norm": 69.5210952758789, "learning_rate": 9.312493399514205e-06, - "loss": 0.7825, + "loss": 1.1799, "step": 50600 }, { "epoch": 5.34, - "grad_norm": 0.27022096514701843, + "grad_norm": 19.382131576538086, "learning_rate": 9.310381244059563e-06, - "loss": 0.4774, + "loss": 0.9607, "step": 50610 }, { "epoch": 5.35, - "grad_norm": 17.356151580810547, + "grad_norm": 31.578105926513672, "learning_rate": 9.308269088604922e-06, - "loss": 0.5958, + "loss": 1.4944, "step": 50620 }, { "epoch": 5.35, - "grad_norm": 17.034408569335938, + "grad_norm": 40.5156364440918, "learning_rate": 9.306156933150281e-06, - "loss": 0.4985, + "loss": 1.0183, "step": 50630 }, { "epoch": 5.35, - "grad_norm": 17.925094604492188, + "grad_norm": 36.67515182495117, "learning_rate": 9.30404477769564e-06, - "loss": 0.6068, + "loss": 1.0899, "step": 50640 }, { "epoch": 5.35, - "grad_norm": 19.141494750976562, + "grad_norm": 42.61861038208008, "learning_rate": 9.301932622240998e-06, - "loss": 0.6266, + "loss": 1.1903, "step": 50650 }, { "epoch": 5.35, - "grad_norm": 12.88898754119873, + "grad_norm": 29.473798751831055, "learning_rate": 9.299820466786357e-06, - "loss": 0.4343, + "loss": 1.1144, "step": 50660 }, { "epoch": 5.35, - "grad_norm": 17.98714828491211, + "grad_norm": 36.89498519897461, "learning_rate": 9.297708311331714e-06, - "loss": 0.6653, + "loss": 0.9901, "step": 50670 }, { "epoch": 5.35, - "grad_norm": 23.63365936279297, + "grad_norm": 43.18898391723633, "learning_rate": 9.295596155877073e-06, - "loss": 0.4047, + "loss": 1.1251, "step": 50680 }, { "epoch": 5.35, - "grad_norm": 0.5727275013923645, + "grad_norm": 41.109397888183594, "learning_rate": 9.293484000422432e-06, - "loss": 0.4247, + "loss": 1.0311, "step": 50690 }, { "epoch": 5.35, - "grad_norm": 21.029949188232422, + "grad_norm": 39.149295806884766, "learning_rate": 9.29137184496779e-06, - "loss": 0.3135, + "loss": 1.0243, "step": 50700 }, { "epoch": 5.36, - "grad_norm": 1.1029108762741089, + "grad_norm": 40.759159088134766, "learning_rate": 9.28925968951315e-06, - "loss": 0.4566, + "loss": 1.0629, "step": 50710 }, { "epoch": 5.36, - "grad_norm": 17.9107666015625, + "grad_norm": 30.967391967773438, "learning_rate": 9.287147534058507e-06, - "loss": 0.5714, + "loss": 1.1289, "step": 50720 }, { "epoch": 5.36, - "grad_norm": 5.4063262939453125, + "grad_norm": 23.252649307250977, "learning_rate": 9.285035378603865e-06, - "loss": 0.3646, + "loss": 0.8065, "step": 50730 }, { "epoch": 5.36, - "grad_norm": 38.672508239746094, + "grad_norm": 28.702417373657227, "learning_rate": 9.282923223149226e-06, - "loss": 0.1935, + "loss": 0.8191, "step": 50740 }, { "epoch": 5.36, - "grad_norm": 14.145588874816895, + "grad_norm": 37.94497299194336, "learning_rate": 9.280811067694583e-06, - "loss": 0.3398, + "loss": 0.7031, "step": 50750 }, { "epoch": 5.36, - "grad_norm": 4.372959613800049, + "grad_norm": 34.31140899658203, "learning_rate": 9.278698912239942e-06, - "loss": 0.5679, + "loss": 1.0915, "step": 50760 }, { "epoch": 5.36, - "grad_norm": 9.517971992492676, + "grad_norm": 24.22235870361328, "learning_rate": 9.2765867567853e-06, - "loss": 0.2969, + "loss": 0.9433, "step": 50770 }, { "epoch": 5.36, - "grad_norm": 0.10502912849187851, + "grad_norm": 33.51292419433594, "learning_rate": 9.27447460133066e-06, - "loss": 0.3689, + "loss": 1.0314, "step": 50780 }, { "epoch": 5.36, - "grad_norm": 0.05775517597794533, + "grad_norm": 24.790727615356445, "learning_rate": 9.272362445876018e-06, - "loss": 0.2128, + "loss": 0.8937, "step": 50790 }, { "epoch": 5.36, - "grad_norm": 17.802959442138672, + "grad_norm": 28.53718376159668, "learning_rate": 9.270250290421375e-06, - "loss": 0.4701, + "loss": 1.1313, "step": 50800 }, { "epoch": 5.37, - "grad_norm": 0.16117733716964722, + "grad_norm": 34.84919357299805, "learning_rate": 9.268138134966734e-06, - "loss": 0.3121, + "loss": 0.9001, "step": 50810 }, { "epoch": 5.37, - "grad_norm": 0.49223512411117554, + "grad_norm": 14.686257362365723, "learning_rate": 9.266025979512093e-06, - "loss": 0.4156, + "loss": 0.956, "step": 50820 }, { "epoch": 5.37, - "grad_norm": 10.669307708740234, + "grad_norm": 36.3492317199707, "learning_rate": 9.263913824057451e-06, - "loss": 0.3057, + "loss": 0.9375, "step": 50830 }, { "epoch": 5.37, - "grad_norm": 27.46751594543457, + "grad_norm": 32.24596405029297, "learning_rate": 9.26180166860281e-06, - "loss": 0.631, + "loss": 0.8305, "step": 50840 }, { "epoch": 5.37, - "grad_norm": 1.2135794162750244, + "grad_norm": 27.952823638916016, "learning_rate": 9.259689513148169e-06, - "loss": 0.6821, + "loss": 0.8814, "step": 50850 }, { "epoch": 5.37, - "grad_norm": 10.540837287902832, + "grad_norm": 36.18394470214844, "learning_rate": 9.257577357693526e-06, - "loss": 0.3437, + "loss": 0.9838, "step": 50860 }, { "epoch": 5.37, - "grad_norm": 10.503984451293945, + "grad_norm": 42.42488098144531, "learning_rate": 9.255465202238887e-06, - "loss": 0.5384, + "loss": 1.1025, "step": 50870 }, { "epoch": 5.37, - "grad_norm": 3.5529592037200928, + "grad_norm": 40.96393585205078, "learning_rate": 9.253353046784244e-06, - "loss": 0.4221, + "loss": 0.8379, "step": 50880 }, { "epoch": 5.37, - "grad_norm": 7.451441764831543, + "grad_norm": 32.606407165527344, "learning_rate": 9.251240891329602e-06, - "loss": 0.8253, + "loss": 0.9341, "step": 50890 }, { "epoch": 5.38, - "grad_norm": 19.15337562561035, + "grad_norm": 49.13193893432617, "learning_rate": 9.249128735874961e-06, - "loss": 0.49, + "loss": 1.0111, "step": 50900 }, { "epoch": 5.38, - "grad_norm": 13.85673713684082, + "grad_norm": 35.724361419677734, "learning_rate": 9.24701658042032e-06, - "loss": 0.2542, + "loss": 0.8955, "step": 50910 }, { "epoch": 5.38, - "grad_norm": 7.0247578620910645, + "grad_norm": 35.953392028808594, "learning_rate": 9.244904424965679e-06, - "loss": 0.8706, + "loss": 1.2662, "step": 50920 }, { "epoch": 5.38, - "grad_norm": 25.356294631958008, + "grad_norm": 45.05967712402344, "learning_rate": 9.242792269511037e-06, - "loss": 0.4762, + "loss": 1.3813, "step": 50930 }, { "epoch": 5.38, - "grad_norm": 7.272548198699951, + "grad_norm": 31.88141632080078, "learning_rate": 9.240680114056395e-06, - "loss": 0.6736, + "loss": 1.1566, "step": 50940 }, { "epoch": 5.38, - "grad_norm": 9.197854995727539, + "grad_norm": 40.52880859375, "learning_rate": 9.238567958601753e-06, - "loss": 0.5086, + "loss": 0.9591, "step": 50950 }, { "epoch": 5.38, - "grad_norm": 6.480856418609619, + "grad_norm": 18.490339279174805, "learning_rate": 9.236455803147112e-06, - "loss": 0.5423, + "loss": 0.8586, "step": 50960 }, { "epoch": 5.38, - "grad_norm": 15.538874626159668, + "grad_norm": 34.71186447143555, "learning_rate": 9.234343647692471e-06, - "loss": 0.5623, + "loss": 1.1884, "step": 50970 }, { "epoch": 5.38, - "grad_norm": 19.441390991210938, + "grad_norm": 38.061702728271484, "learning_rate": 9.23223149223783e-06, - "loss": 0.6627, + "loss": 1.3686, "step": 50980 }, { "epoch": 5.38, - "grad_norm": 20.035757064819336, + "grad_norm": 48.41822814941406, "learning_rate": 9.230119336783187e-06, - "loss": 0.3274, + "loss": 1.2686, "step": 50990 }, { "epoch": 5.39, - "grad_norm": 11.901333808898926, + "grad_norm": 59.787967681884766, "learning_rate": 9.228007181328547e-06, - "loss": 0.6765, + "loss": 1.4103, "step": 51000 }, { "epoch": 5.39, - "grad_norm": 23.401334762573242, + "grad_norm": 47.13447570800781, "learning_rate": 9.225895025873906e-06, - "loss": 0.3616, + "loss": 1.0271, "step": 51010 }, { "epoch": 5.39, - "grad_norm": 12.408196449279785, + "grad_norm": 33.317867279052734, "learning_rate": 9.223782870419263e-06, - "loss": 0.5808, + "loss": 0.9164, "step": 51020 }, { "epoch": 5.39, - "grad_norm": 13.208032608032227, + "grad_norm": 27.890256881713867, "learning_rate": 9.221670714964622e-06, - "loss": 0.4111, + "loss": 0.9706, "step": 51030 }, { "epoch": 5.39, - "grad_norm": 3.453955888748169, + "grad_norm": 18.406394958496094, "learning_rate": 9.21955855950998e-06, - "loss": 0.3259, + "loss": 0.6592, "step": 51040 }, { "epoch": 5.39, - "grad_norm": 0.8612927794456482, + "grad_norm": 17.846574783325195, "learning_rate": 9.21744640405534e-06, - "loss": 0.321, + "loss": 0.9917, "step": 51050 }, { "epoch": 5.39, - "grad_norm": 5.556461334228516, + "grad_norm": 26.389863967895508, "learning_rate": 9.215334248600698e-06, - "loss": 0.3791, + "loss": 1.3399, "step": 51060 }, { "epoch": 5.39, - "grad_norm": 9.723876953125, + "grad_norm": 33.263797760009766, "learning_rate": 9.213222093146055e-06, - "loss": 0.6402, + "loss": 0.8831, "step": 51070 }, { "epoch": 5.39, - "grad_norm": 14.20356559753418, + "grad_norm": 37.30786895751953, "learning_rate": 9.211109937691414e-06, - "loss": 0.4102, + "loss": 0.9436, "step": 51080 }, { "epoch": 5.4, - "grad_norm": 5.233372211456299, + "grad_norm": 33.16160583496094, "learning_rate": 9.208997782236775e-06, - "loss": 0.1799, + "loss": 0.6986, "step": 51090 }, { "epoch": 5.4, - "grad_norm": 11.89257526397705, + "grad_norm": 13.053960800170898, "learning_rate": 9.206885626782132e-06, - "loss": 0.5233, + "loss": 0.9847, "step": 51100 }, { "epoch": 5.4, - "grad_norm": 34.59709167480469, + "grad_norm": 27.292343139648438, "learning_rate": 9.20477347132749e-06, - "loss": 0.5605, + "loss": 1.1484, "step": 51110 }, { "epoch": 5.4, - "grad_norm": 16.903305053710938, + "grad_norm": 45.445858001708984, "learning_rate": 9.20266131587285e-06, - "loss": 0.6364, + "loss": 1.3436, "step": 51120 }, { "epoch": 5.4, - "grad_norm": 12.722990989685059, + "grad_norm": 44.14167022705078, "learning_rate": 9.200549160418208e-06, - "loss": 0.6632, + "loss": 1.1086, "step": 51130 }, { "epoch": 5.4, - "grad_norm": 21.988618850708008, + "grad_norm": 47.444217681884766, "learning_rate": 9.198437004963567e-06, - "loss": 0.3671, + "loss": 0.9147, "step": 51140 }, { "epoch": 5.4, - "grad_norm": 26.1848087310791, + "grad_norm": 55.48174285888672, "learning_rate": 9.196324849508924e-06, - "loss": 0.5317, + "loss": 1.3344, "step": 51150 }, { "epoch": 5.4, - "grad_norm": 0.5258842706680298, + "grad_norm": 28.78095245361328, "learning_rate": 9.194212694054283e-06, - "loss": 0.3551, + "loss": 0.7872, "step": 51160 }, { "epoch": 5.4, - "grad_norm": 18.846364974975586, + "grad_norm": 35.96540451049805, "learning_rate": 9.192100538599641e-06, - "loss": 0.705, + "loss": 0.9556, "step": 51170 }, { "epoch": 5.41, - "grad_norm": 16.004812240600586, + "grad_norm": 41.87092590332031, "learning_rate": 9.189988383145e-06, - "loss": 0.3209, + "loss": 0.8263, "step": 51180 }, { "epoch": 5.41, - "grad_norm": 3.2144250869750977, + "grad_norm": 36.414642333984375, "learning_rate": 9.187876227690359e-06, - "loss": 0.7993, + "loss": 1.3893, "step": 51190 }, { "epoch": 5.41, - "grad_norm": 9.431116104125977, + "grad_norm": 40.303131103515625, "learning_rate": 9.185764072235718e-06, - "loss": 0.6717, + "loss": 1.3731, "step": 51200 }, { "epoch": 5.41, - "grad_norm": 15.08937931060791, + "grad_norm": 31.88495635986328, "learning_rate": 9.183651916781075e-06, - "loss": 0.3515, + "loss": 0.9069, "step": 51210 }, { "epoch": 5.41, - "grad_norm": 4.138614654541016, + "grad_norm": 19.67959213256836, "learning_rate": 9.181539761326435e-06, - "loss": 0.3321, + "loss": 1.012, "step": 51220 }, { "epoch": 5.41, - "grad_norm": 0.5731115937232971, + "grad_norm": 33.625328063964844, "learning_rate": 9.179427605871792e-06, - "loss": 0.1583, + "loss": 0.8579, "step": 51230 }, { "epoch": 5.41, - "grad_norm": 18.93462562561035, + "grad_norm": 31.12316131591797, "learning_rate": 9.177315450417151e-06, - "loss": 0.3266, + "loss": 1.1464, "step": 51240 }, { "epoch": 5.41, - "grad_norm": 18.99652671813965, + "grad_norm": 25.586326599121094, "learning_rate": 9.17520329496251e-06, - "loss": 0.3826, + "loss": 0.9955, "step": 51250 }, { "epoch": 5.41, - "grad_norm": 24.709835052490234, + "grad_norm": 24.790536880493164, "learning_rate": 9.173091139507869e-06, - "loss": 0.4699, + "loss": 1.0581, "step": 51260 }, { "epoch": 5.41, - "grad_norm": 20.673250198364258, + "grad_norm": 31.475467681884766, "learning_rate": 9.170978984053227e-06, - "loss": 0.4115, + "loss": 1.0516, "step": 51270 }, { "epoch": 5.42, - "grad_norm": 10.174315452575684, + "grad_norm": 24.96847152709961, "learning_rate": 9.168866828598586e-06, - "loss": 0.3448, + "loss": 1.08, "step": 51280 }, { "epoch": 5.42, - "grad_norm": 2.93093204498291, + "grad_norm": 14.875314712524414, "learning_rate": 9.166754673143943e-06, - "loss": 0.2878, + "loss": 0.6802, "step": 51290 }, { "epoch": 5.42, - "grad_norm": 0.0593101941049099, + "grad_norm": 35.87545394897461, "learning_rate": 9.164642517689302e-06, - "loss": 0.5754, + "loss": 1.2707, "step": 51300 }, { "epoch": 5.42, - "grad_norm": 3.7029409408569336, + "grad_norm": 37.460289001464844, "learning_rate": 9.162530362234661e-06, - "loss": 0.3047, + "loss": 1.0394, "step": 51310 }, { "epoch": 5.42, - "grad_norm": 9.982538223266602, + "grad_norm": 49.65753936767578, "learning_rate": 9.16041820678002e-06, - "loss": 0.3986, + "loss": 1.1946, "step": 51320 }, { "epoch": 5.42, - "grad_norm": 6.0259246826171875, + "grad_norm": 30.619705200195312, "learning_rate": 9.158306051325378e-06, - "loss": 0.5275, + "loss": 1.1871, "step": 51330 }, { "epoch": 5.42, - "grad_norm": 15.327892303466797, + "grad_norm": 33.395687103271484, "learning_rate": 9.156193895870736e-06, - "loss": 0.2276, + "loss": 0.9947, "step": 51340 }, { "epoch": 5.42, - "grad_norm": 14.136309623718262, + "grad_norm": 64.37250518798828, "learning_rate": 9.154081740416096e-06, - "loss": 0.7427, + "loss": 0.8262, "step": 51350 }, { "epoch": 5.42, - "grad_norm": 14.922691345214844, + "grad_norm": 35.76814651489258, "learning_rate": 9.151969584961455e-06, - "loss": 0.9303, + "loss": 1.5575, "step": 51360 }, { "epoch": 5.43, - "grad_norm": 20.69582176208496, + "grad_norm": 54.306522369384766, "learning_rate": 9.149857429506812e-06, - "loss": 0.5068, + "loss": 1.1288, "step": 51370 }, { "epoch": 5.43, - "grad_norm": 17.127525329589844, + "grad_norm": 50.66888427734375, "learning_rate": 9.14774527405217e-06, - "loss": 0.4798, + "loss": 0.9368, "step": 51380 }, { "epoch": 5.43, - "grad_norm": 13.13377571105957, + "grad_norm": 54.758235931396484, "learning_rate": 9.14563311859753e-06, - "loss": 0.5411, + "loss": 1.1071, "step": 51390 }, { "epoch": 5.43, - "grad_norm": 9.339498519897461, + "grad_norm": 47.3731803894043, "learning_rate": 9.143520963142888e-06, - "loss": 0.4635, + "loss": 1.3726, "step": 51400 }, { "epoch": 5.43, - "grad_norm": 6.38276481628418, + "grad_norm": 43.444175720214844, "learning_rate": 9.141408807688247e-06, - "loss": 0.5181, + "loss": 1.4548, "step": 51410 }, { "epoch": 5.43, - "grad_norm": 0.7356536388397217, + "grad_norm": 23.508159637451172, "learning_rate": 9.139296652233604e-06, - "loss": 0.4264, + "loss": 0.948, "step": 51420 }, { "epoch": 5.43, - "grad_norm": 9.09988021850586, + "grad_norm": 36.89360046386719, "learning_rate": 9.137184496778963e-06, - "loss": 0.7998, + "loss": 1.4025, "step": 51430 }, { "epoch": 5.43, - "grad_norm": 0.8637728691101074, + "grad_norm": 13.961956024169922, "learning_rate": 9.135072341324323e-06, - "loss": 0.5, + "loss": 0.9255, "step": 51440 }, { "epoch": 5.43, - "grad_norm": 4.884982109069824, + "grad_norm": 51.281593322753906, "learning_rate": 9.13296018586968e-06, - "loss": 0.4402, + "loss": 1.0179, "step": 51450 }, { "epoch": 5.43, - "grad_norm": 35.52216339111328, + "grad_norm": 44.395782470703125, "learning_rate": 9.13084803041504e-06, - "loss": 0.3618, + "loss": 1.0051, "step": 51460 }, { "epoch": 5.44, - "grad_norm": 8.242124557495117, + "grad_norm": 41.245426177978516, "learning_rate": 9.128735874960398e-06, - "loss": 0.3423, + "loss": 0.8952, "step": 51470 }, { "epoch": 5.44, - "grad_norm": 0.07449859380722046, + "grad_norm": 44.63047790527344, "learning_rate": 9.126623719505757e-06, - "loss": 0.2657, + "loss": 0.9782, "step": 51480 }, { "epoch": 5.44, - "grad_norm": 17.75716209411621, + "grad_norm": 28.081464767456055, "learning_rate": 9.124511564051116e-06, - "loss": 0.3881, + "loss": 0.7613, "step": 51490 }, { "epoch": 5.44, - "grad_norm": 12.212241172790527, + "grad_norm": 29.60127067565918, "learning_rate": 9.122399408596474e-06, - "loss": 0.5986, + "loss": 0.9949, "step": 51500 }, { "epoch": 5.44, - "grad_norm": 0.49041691422462463, + "grad_norm": 26.622596740722656, "learning_rate": 9.120287253141831e-06, - "loss": 0.4477, + "loss": 1.1867, "step": 51510 }, { "epoch": 5.44, - "grad_norm": 34.077919006347656, + "grad_norm": 48.91542053222656, "learning_rate": 9.11817509768719e-06, - "loss": 0.5236, + "loss": 1.1671, "step": 51520 }, { "epoch": 5.44, - "grad_norm": 13.610111236572266, + "grad_norm": 33.90446853637695, "learning_rate": 9.116062942232549e-06, - "loss": 0.4001, + "loss": 0.9478, "step": 51530 }, { "epoch": 5.44, - "grad_norm": 81.70972442626953, + "grad_norm": 57.92710494995117, "learning_rate": 9.113950786777908e-06, - "loss": 0.5546, + "loss": 1.0405, "step": 51540 }, { "epoch": 5.44, - "grad_norm": 6.7381978034973145, + "grad_norm": 5.066338539123535, "learning_rate": 9.111838631323267e-06, - "loss": 0.6941, + "loss": 0.7512, "step": 51550 }, { "epoch": 5.45, - "grad_norm": 1.0646063089370728, + "grad_norm": 19.36600685119629, "learning_rate": 9.109726475868624e-06, - "loss": 0.6334, + "loss": 1.1966, "step": 51560 }, { "epoch": 5.45, - "grad_norm": 2.211228370666504, + "grad_norm": 22.973678588867188, "learning_rate": 9.107614320413984e-06, - "loss": 0.6091, + "loss": 1.2072, "step": 51570 }, { "epoch": 5.45, - "grad_norm": 20.09726333618164, + "grad_norm": 34.48703384399414, "learning_rate": 9.105502164959343e-06, - "loss": 0.5351, + "loss": 1.2163, "step": 51580 }, { "epoch": 5.45, - "grad_norm": 2.384934663772583, + "grad_norm": 5.587264537811279, "learning_rate": 9.1033900095047e-06, - "loss": 0.2207, + "loss": 0.7538, "step": 51590 }, { "epoch": 5.45, - "grad_norm": 32.97623062133789, + "grad_norm": 44.22422409057617, "learning_rate": 9.101277854050059e-06, - "loss": 0.4932, + "loss": 1.1858, "step": 51600 }, { "epoch": 5.45, - "grad_norm": 26.711124420166016, + "grad_norm": 43.934913635253906, "learning_rate": 9.099165698595418e-06, - "loss": 0.5442, + "loss": 1.0171, "step": 51610 }, { "epoch": 5.45, - "grad_norm": 22.327369689941406, + "grad_norm": 55.007957458496094, "learning_rate": 9.097053543140776e-06, - "loss": 0.3623, + "loss": 0.9388, "step": 51620 }, { "epoch": 5.45, - "grad_norm": 18.516170501708984, + "grad_norm": 41.83439636230469, "learning_rate": 9.094941387686135e-06, - "loss": 0.4866, + "loss": 1.0622, "step": 51630 }, { "epoch": 5.45, - "grad_norm": 37.07299041748047, + "grad_norm": 47.928871154785156, "learning_rate": 9.092829232231492e-06, - "loss": 0.4246, + "loss": 1.0109, "step": 51640 }, { "epoch": 5.45, - "grad_norm": 10.627015113830566, + "grad_norm": 32.12445831298828, "learning_rate": 9.090717076776851e-06, - "loss": 0.2446, + "loss": 0.7884, "step": 51650 }, { "epoch": 5.46, - "grad_norm": 5.773805618286133, + "grad_norm": 42.76057815551758, "learning_rate": 9.08860492132221e-06, - "loss": 0.4117, + "loss": 0.8777, "step": 51660 }, { "epoch": 5.46, - "grad_norm": 1.034468412399292, + "grad_norm": 48.07674026489258, "learning_rate": 9.086492765867568e-06, - "loss": 0.4364, + "loss": 0.8706, "step": 51670 }, { "epoch": 5.46, - "grad_norm": 21.83382797241211, + "grad_norm": 42.02615737915039, "learning_rate": 9.084380610412927e-06, - "loss": 0.3377, + "loss": 0.5345, "step": 51680 }, { "epoch": 5.46, - "grad_norm": 6.679325103759766, + "grad_norm": 41.81730270385742, "learning_rate": 9.082268454958284e-06, - "loss": 0.3721, + "loss": 1.0988, "step": 51690 }, { "epoch": 5.46, - "grad_norm": 15.603902816772461, + "grad_norm": 57.64115524291992, "learning_rate": 9.080156299503645e-06, - "loss": 0.3528, + "loss": 0.8802, "step": 51700 }, { "epoch": 5.46, - "grad_norm": 7.182570934295654, + "grad_norm": 27.713220596313477, "learning_rate": 9.078044144049004e-06, - "loss": 0.4334, + "loss": 0.8228, "step": 51710 }, { "epoch": 5.46, - "grad_norm": 6.182878494262695, + "grad_norm": 43.11452102661133, "learning_rate": 9.07593198859436e-06, - "loss": 0.2379, + "loss": 0.969, "step": 51720 }, { "epoch": 5.46, - "grad_norm": 20.17357635498047, + "grad_norm": 37.9946174621582, "learning_rate": 9.07381983313972e-06, - "loss": 0.3526, + "loss": 1.2782, "step": 51730 }, { "epoch": 5.46, - "grad_norm": 8.4638671875, + "grad_norm": 41.447998046875, "learning_rate": 9.071707677685078e-06, - "loss": 0.2172, + "loss": 1.0025, "step": 51740 }, { "epoch": 5.47, - "grad_norm": 0.9351205229759216, + "grad_norm": 31.193716049194336, "learning_rate": 9.069595522230437e-06, - "loss": 0.342, + "loss": 0.8314, "step": 51750 }, { "epoch": 5.47, - "grad_norm": 6.305736541748047, + "grad_norm": 30.82987403869629, "learning_rate": 9.067483366775796e-06, - "loss": 0.3946, + "loss": 1.0992, "step": 51760 }, { "epoch": 5.47, - "grad_norm": 13.239834785461426, + "grad_norm": 50.787322998046875, "learning_rate": 9.065371211321155e-06, - "loss": 0.5611, + "loss": 1.3218, "step": 51770 }, { "epoch": 5.47, - "grad_norm": 18.833606719970703, + "grad_norm": 35.10037612915039, "learning_rate": 9.063259055866512e-06, - "loss": 0.4749, + "loss": 0.9614, "step": 51780 }, { "epoch": 5.47, - "grad_norm": 12.617955207824707, + "grad_norm": 33.26079177856445, "learning_rate": 9.06114690041187e-06, - "loss": 0.4756, + "loss": 0.9962, "step": 51790 }, { "epoch": 5.47, - "grad_norm": 0.08864197880029678, + "grad_norm": 19.250934600830078, "learning_rate": 9.05903474495723e-06, - "loss": 0.5851, + "loss": 1.1484, "step": 51800 }, { "epoch": 5.47, - "grad_norm": 8.413296699523926, + "grad_norm": 47.24285125732422, "learning_rate": 9.056922589502588e-06, - "loss": 0.2545, + "loss": 0.9045, "step": 51810 }, { "epoch": 5.47, - "grad_norm": 27.631681442260742, + "grad_norm": 61.694190979003906, "learning_rate": 9.054810434047947e-06, - "loss": 0.2799, + "loss": 0.8072, "step": 51820 }, { "epoch": 5.47, - "grad_norm": 10.923826217651367, + "grad_norm": 25.11994743347168, "learning_rate": 9.052698278593306e-06, - "loss": 0.5138, + "loss": 1.0181, "step": 51830 }, { "epoch": 5.47, - "grad_norm": 6.704423904418945, + "grad_norm": 35.11958694458008, "learning_rate": 9.050586123138664e-06, - "loss": 0.0786, + "loss": 0.6019, "step": 51840 }, { "epoch": 5.48, - "grad_norm": 22.442523956298828, + "grad_norm": 58.25111770629883, "learning_rate": 9.048473967684023e-06, - "loss": 0.6959, + "loss": 1.1772, "step": 51850 }, { "epoch": 5.48, - "grad_norm": 25.33738136291504, + "grad_norm": 30.35346031188965, "learning_rate": 9.04636181222938e-06, - "loss": 0.4171, + "loss": 0.5907, "step": 51860 }, { "epoch": 5.48, - "grad_norm": 0.09756608307361603, + "grad_norm": 25.128786087036133, "learning_rate": 9.044249656774739e-06, - "loss": 0.3778, + "loss": 1.158, "step": 51870 }, { "epoch": 5.48, - "grad_norm": 15.127433776855469, + "grad_norm": 46.105770111083984, "learning_rate": 9.042137501320098e-06, - "loss": 0.5567, + "loss": 1.2731, "step": 51880 }, { "epoch": 5.48, - "grad_norm": 21.526437759399414, + "grad_norm": 40.11400604248047, "learning_rate": 9.040025345865457e-06, - "loss": 0.378, + "loss": 0.873, "step": 51890 }, { "epoch": 5.48, - "grad_norm": 55.03844451904297, + "grad_norm": 40.473182678222656, "learning_rate": 9.037913190410815e-06, - "loss": 0.391, + "loss": 0.6566, "step": 51900 }, { "epoch": 5.48, - "grad_norm": 19.557479858398438, + "grad_norm": 34.1539306640625, "learning_rate": 9.035801034956172e-06, - "loss": 0.5254, + "loss": 0.9954, "step": 51910 }, { "epoch": 5.48, - "grad_norm": 23.517250061035156, + "grad_norm": 53.66899490356445, "learning_rate": 9.033688879501533e-06, - "loss": 0.4561, + "loss": 1.3296, "step": 51920 }, { "epoch": 5.48, - "grad_norm": 12.342307090759277, + "grad_norm": 31.707948684692383, "learning_rate": 9.031576724046892e-06, - "loss": 0.3297, + "loss": 1.0901, "step": 51930 }, { "epoch": 5.49, - "grad_norm": 7.511523723602295, + "grad_norm": 37.587711334228516, "learning_rate": 9.029464568592249e-06, - "loss": 0.5397, + "loss": 1.157, "step": 51940 }, { "epoch": 5.49, - "grad_norm": 16.416196823120117, + "grad_norm": 35.266666412353516, "learning_rate": 9.027352413137608e-06, - "loss": 0.7314, + "loss": 1.1208, "step": 51950 }, { "epoch": 5.49, - "grad_norm": 9.821486473083496, + "grad_norm": 38.43747329711914, "learning_rate": 9.025240257682966e-06, - "loss": 0.3971, + "loss": 1.1706, "step": 51960 }, { "epoch": 5.49, - "grad_norm": 6.168426990509033, + "grad_norm": 63.24063491821289, "learning_rate": 9.023128102228325e-06, - "loss": 0.4047, + "loss": 1.0585, "step": 51970 }, { "epoch": 5.49, - "grad_norm": 10.954497337341309, + "grad_norm": 35.57136154174805, "learning_rate": 9.021015946773684e-06, - "loss": 0.4188, + "loss": 0.8675, "step": 51980 }, { "epoch": 5.49, - "grad_norm": 10.2535400390625, + "grad_norm": 33.54383850097656, "learning_rate": 9.018903791319041e-06, - "loss": 0.4036, + "loss": 1.1236, "step": 51990 }, { "epoch": 5.49, - "grad_norm": 31.713716506958008, + "grad_norm": 31.906482696533203, "learning_rate": 9.0167916358644e-06, - "loss": 0.2208, + "loss": 0.901, "step": 52000 }, { "epoch": 5.49, - "grad_norm": 15.282013893127441, + "grad_norm": 34.731056213378906, "learning_rate": 9.014679480409758e-06, - "loss": 0.7685, + "loss": 1.238, "step": 52010 }, { "epoch": 5.49, - "grad_norm": 0.3005354106426239, + "grad_norm": 42.52900695800781, "learning_rate": 9.012567324955117e-06, - "loss": 0.3164, + "loss": 0.6921, "step": 52020 }, { "epoch": 5.49, - "grad_norm": 19.39333152770996, + "grad_norm": 47.05219268798828, "learning_rate": 9.010455169500476e-06, - "loss": 0.6981, + "loss": 0.9778, "step": 52030 }, { "epoch": 5.5, - "grad_norm": 4.714667320251465, + "grad_norm": 32.96712112426758, "learning_rate": 9.008343014045835e-06, - "loss": 0.4739, + "loss": 1.0172, "step": 52040 }, { "epoch": 5.5, - "grad_norm": 1.3847168684005737, + "grad_norm": 28.410350799560547, "learning_rate": 9.006230858591194e-06, - "loss": 0.4159, + "loss": 1.1388, "step": 52050 }, { "epoch": 5.5, - "grad_norm": 9.92342758178711, + "grad_norm": 20.819599151611328, "learning_rate": 9.004118703136552e-06, - "loss": 0.4511, + "loss": 0.9273, "step": 52060 }, { "epoch": 5.5, - "grad_norm": 15.425379753112793, + "grad_norm": 39.054542541503906, "learning_rate": 9.00200654768191e-06, - "loss": 0.2228, + "loss": 1.0563, "step": 52070 }, { "epoch": 5.5, - "grad_norm": 20.872156143188477, + "grad_norm": 49.421363830566406, "learning_rate": 8.999894392227268e-06, - "loss": 0.3606, + "loss": 0.836, "step": 52080 }, { "epoch": 5.5, - "grad_norm": 20.21304702758789, + "grad_norm": 35.311153411865234, "learning_rate": 8.997782236772627e-06, - "loss": 0.359, + "loss": 1.0982, "step": 52090 }, { "epoch": 5.5, - "grad_norm": 20.921350479125977, + "grad_norm": 60.99721908569336, "learning_rate": 8.995670081317986e-06, - "loss": 0.6675, + "loss": 1.1836, "step": 52100 }, { "epoch": 5.5, - "grad_norm": 13.720499038696289, + "grad_norm": 35.7238655090332, "learning_rate": 8.993557925863345e-06, - "loss": 0.6077, + "loss": 1.1145, "step": 52110 }, { "epoch": 5.5, - "grad_norm": 17.864355087280273, + "grad_norm": 47.53273010253906, "learning_rate": 8.991445770408703e-06, - "loss": 0.2781, + "loss": 0.9067, "step": 52120 }, { "epoch": 5.51, - "grad_norm": 4.660913944244385, + "grad_norm": 35.12746810913086, "learning_rate": 8.98933361495406e-06, - "loss": 0.1464, + "loss": 0.8568, "step": 52130 }, { "epoch": 5.51, - "grad_norm": 23.77631950378418, + "grad_norm": 50.345252990722656, "learning_rate": 8.98722145949942e-06, - "loss": 0.3543, + "loss": 0.8284, "step": 52140 }, { "epoch": 5.51, - "grad_norm": 0.4399968683719635, + "grad_norm": 51.554683685302734, "learning_rate": 8.985109304044778e-06, - "loss": 0.2272, + "loss": 0.9792, "step": 52150 }, { "epoch": 5.51, - "grad_norm": 14.471075057983398, + "grad_norm": 46.490394592285156, "learning_rate": 8.982997148590137e-06, - "loss": 0.4013, + "loss": 1.0104, "step": 52160 }, { "epoch": 5.51, - "grad_norm": 0.10499802231788635, + "grad_norm": 28.637767791748047, "learning_rate": 8.980884993135496e-06, - "loss": 0.2846, + "loss": 1.1099, "step": 52170 }, { "epoch": 5.51, - "grad_norm": 1.0975745916366577, + "grad_norm": 23.66973114013672, "learning_rate": 8.978772837680854e-06, - "loss": 0.3439, + "loss": 0.9105, "step": 52180 }, { "epoch": 5.51, - "grad_norm": 12.057147026062012, + "grad_norm": 30.29077911376953, "learning_rate": 8.976660682226213e-06, - "loss": 0.5055, + "loss": 0.8467, "step": 52190 }, { "epoch": 5.51, - "grad_norm": 20.291336059570312, + "grad_norm": 51.958309173583984, "learning_rate": 8.974548526771572e-06, - "loss": 0.3795, + "loss": 1.1124, "step": 52200 }, { "epoch": 5.51, - "grad_norm": 12.581932067871094, + "grad_norm": 34.96770095825195, "learning_rate": 8.972436371316929e-06, - "loss": 0.6047, + "loss": 1.3564, "step": 52210 }, { "epoch": 5.51, - "grad_norm": 3.016063928604126, + "grad_norm": 31.59081268310547, "learning_rate": 8.970324215862288e-06, - "loss": 0.3145, + "loss": 0.9944, "step": 52220 }, { "epoch": 5.52, - "grad_norm": 22.220502853393555, + "grad_norm": 52.0082893371582, "learning_rate": 8.968212060407647e-06, - "loss": 0.231, + "loss": 1.2043, "step": 52230 }, { "epoch": 5.52, - "grad_norm": 1.4127941131591797, + "grad_norm": 29.56708526611328, "learning_rate": 8.966099904953005e-06, - "loss": 0.2429, + "loss": 0.8899, "step": 52240 }, { "epoch": 5.52, - "grad_norm": 21.321796417236328, + "grad_norm": 39.32154083251953, "learning_rate": 8.963987749498364e-06, - "loss": 0.275, + "loss": 1.0431, "step": 52250 }, { "epoch": 5.52, - "grad_norm": 0.7350202798843384, + "grad_norm": 28.933441162109375, "learning_rate": 8.961875594043721e-06, - "loss": 0.5049, + "loss": 1.2293, "step": 52260 }, { "epoch": 5.52, - "grad_norm": 4.386728763580322, + "grad_norm": 47.22176742553711, "learning_rate": 8.959763438589082e-06, - "loss": 0.3952, + "loss": 0.9489, "step": 52270 }, { "epoch": 5.52, - "grad_norm": 5.973857402801514, + "grad_norm": 41.50701904296875, "learning_rate": 8.95765128313444e-06, - "loss": 0.3064, + "loss": 0.7605, "step": 52280 }, { "epoch": 5.52, - "grad_norm": 19.804580688476562, + "grad_norm": 56.815826416015625, "learning_rate": 8.955539127679798e-06, - "loss": 0.6542, + "loss": 1.2725, "step": 52290 }, { "epoch": 5.52, - "grad_norm": 0.3209969401359558, + "grad_norm": 37.97354507446289, "learning_rate": 8.953426972225156e-06, - "loss": 0.4235, + "loss": 1.3077, "step": 52300 }, { "epoch": 5.52, - "grad_norm": 3.782083749771118, + "grad_norm": 39.749664306640625, "learning_rate": 8.951314816770515e-06, - "loss": 0.4513, + "loss": 0.8798, "step": 52310 }, { "epoch": 5.53, - "grad_norm": 8.41179370880127, + "grad_norm": 24.373796463012695, "learning_rate": 8.949202661315874e-06, - "loss": 0.5902, + "loss": 1.1174, "step": 52320 }, { "epoch": 5.53, - "grad_norm": 13.263738632202148, + "grad_norm": 41.079368591308594, "learning_rate": 8.947090505861233e-06, - "loss": 0.2572, + "loss": 1.0538, "step": 52330 }, { "epoch": 5.53, - "grad_norm": 21.375547409057617, + "grad_norm": 55.80338668823242, "learning_rate": 8.94497835040659e-06, - "loss": 0.3784, + "loss": 1.1402, "step": 52340 }, { "epoch": 5.53, - "grad_norm": 8.790900230407715, + "grad_norm": 48.292144775390625, "learning_rate": 8.942866194951949e-06, - "loss": 0.3873, + "loss": 1.3338, "step": 52350 }, { "epoch": 5.53, - "grad_norm": 0.15832895040512085, + "grad_norm": 31.70334243774414, "learning_rate": 8.940754039497307e-06, - "loss": 0.3729, + "loss": 0.8513, "step": 52360 }, { "epoch": 5.53, - "grad_norm": 24.614858627319336, + "grad_norm": 21.53739356994629, "learning_rate": 8.938641884042666e-06, - "loss": 0.4914, + "loss": 0.8596, "step": 52370 }, { "epoch": 5.53, - "grad_norm": 12.911977767944336, + "grad_norm": 35.26810836791992, "learning_rate": 8.936529728588025e-06, - "loss": 0.4884, + "loss": 0.8889, "step": 52380 }, { "epoch": 5.53, - "grad_norm": 6.26076602935791, + "grad_norm": 44.040489196777344, "learning_rate": 8.934417573133384e-06, - "loss": 0.3415, + "loss": 0.9328, "step": 52390 }, { "epoch": 5.53, - "grad_norm": 13.663407325744629, + "grad_norm": 45.15055465698242, "learning_rate": 8.932305417678742e-06, - "loss": 0.4662, + "loss": 1.2059, "step": 52400 }, { "epoch": 5.53, - "grad_norm": 28.34563636779785, + "grad_norm": 50.397945404052734, "learning_rate": 8.930193262224101e-06, - "loss": 0.6075, + "loss": 0.9003, "step": 52410 }, { "epoch": 5.54, - "grad_norm": 0.814755916595459, + "grad_norm": 43.68463897705078, "learning_rate": 8.92808110676946e-06, - "loss": 0.461, + "loss": 1.0713, "step": 52420 }, { "epoch": 5.54, - "grad_norm": 0.7289329767227173, + "grad_norm": 28.19077491760254, "learning_rate": 8.925968951314817e-06, - "loss": 0.2686, + "loss": 1.1833, "step": 52430 }, { "epoch": 5.54, - "grad_norm": 18.075674057006836, + "grad_norm": 37.3130989074707, "learning_rate": 8.923856795860176e-06, - "loss": 0.8112, + "loss": 1.0487, "step": 52440 }, { "epoch": 5.54, - "grad_norm": 32.668460845947266, + "grad_norm": 42.482059478759766, "learning_rate": 8.921744640405535e-06, - "loss": 0.4132, + "loss": 1.18, "step": 52450 }, { "epoch": 5.54, - "grad_norm": 6.931715965270996, + "grad_norm": 31.867794036865234, "learning_rate": 8.919632484950893e-06, - "loss": 0.3902, + "loss": 1.1148, "step": 52460 }, { "epoch": 5.54, - "grad_norm": 10.38728141784668, + "grad_norm": 45.000091552734375, "learning_rate": 8.917520329496252e-06, - "loss": 0.5143, + "loss": 1.3553, "step": 52470 }, { "epoch": 5.54, - "grad_norm": 14.02862548828125, + "grad_norm": 38.65494155883789, "learning_rate": 8.91540817404161e-06, - "loss": 0.4973, + "loss": 1.0369, "step": 52480 }, { "epoch": 5.54, - "grad_norm": 11.52883243560791, + "grad_norm": 36.325286865234375, "learning_rate": 8.913296018586968e-06, - "loss": 0.3924, + "loss": 1.1696, "step": 52490 }, { "epoch": 5.54, - "grad_norm": 44.78354263305664, + "grad_norm": 27.967267990112305, "learning_rate": 8.911183863132328e-06, - "loss": 0.377, + "loss": 1.0078, "step": 52500 }, { "epoch": 5.55, - "grad_norm": 8.735016822814941, + "grad_norm": 43.49846267700195, "learning_rate": 8.909071707677686e-06, - "loss": 0.5132, + "loss": 0.9862, "step": 52510 }, { "epoch": 5.55, - "grad_norm": 0.32949116826057434, + "grad_norm": 22.051904678344727, "learning_rate": 8.906959552223044e-06, - "loss": 0.6434, + "loss": 1.1134, "step": 52520 }, { "epoch": 5.55, - "grad_norm": 0.35424453020095825, + "grad_norm": 8.808829307556152, "learning_rate": 8.904847396768403e-06, - "loss": 0.1671, + "loss": 0.6442, "step": 52530 }, { "epoch": 5.55, - "grad_norm": 0.20067858695983887, + "grad_norm": 35.242977142333984, "learning_rate": 8.902735241313762e-06, - "loss": 0.4983, + "loss": 1.4318, "step": 52540 }, { "epoch": 5.55, - "grad_norm": 11.047354698181152, + "grad_norm": 47.30587387084961, "learning_rate": 8.90062308585912e-06, - "loss": 0.3933, + "loss": 1.0447, "step": 52550 }, { "epoch": 5.55, - "grad_norm": 19.277660369873047, + "grad_norm": 43.25763702392578, "learning_rate": 8.898510930404478e-06, - "loss": 0.4899, + "loss": 1.0538, "step": 52560 }, { "epoch": 5.55, - "grad_norm": 35.84602737426758, + "grad_norm": 20.162853240966797, "learning_rate": 8.896398774949837e-06, - "loss": 0.5181, + "loss": 0.9268, "step": 52570 }, { "epoch": 5.55, - "grad_norm": 7.272068023681641, + "grad_norm": 40.98930358886719, "learning_rate": 8.894286619495195e-06, - "loss": 0.428, + "loss": 1.2922, "step": 52580 }, { "epoch": 5.55, - "grad_norm": 2.221986770629883, + "grad_norm": 38.573768615722656, "learning_rate": 8.892174464040554e-06, - "loss": 0.3585, + "loss": 0.8715, "step": 52590 }, { "epoch": 5.55, - "grad_norm": 28.053146362304688, + "grad_norm": 58.181636810302734, "learning_rate": 8.890062308585913e-06, - "loss": 0.3905, + "loss": 1.0689, "step": 52600 }, { "epoch": 5.56, - "grad_norm": 12.071455955505371, + "grad_norm": 51.344703674316406, "learning_rate": 8.887950153131272e-06, - "loss": 0.5816, + "loss": 1.2902, "step": 52610 }, { "epoch": 5.56, - "grad_norm": 0.0701453909277916, + "grad_norm": 15.753684043884277, "learning_rate": 8.885837997676629e-06, - "loss": 0.461, + "loss": 0.8616, "step": 52620 }, { "epoch": 5.56, - "grad_norm": 15.377878189086914, + "grad_norm": 61.52045822143555, "learning_rate": 8.88372584222199e-06, - "loss": 0.713, + "loss": 1.4201, "step": 52630 }, { "epoch": 5.56, - "grad_norm": 11.605603218078613, + "grad_norm": 37.06349182128906, "learning_rate": 8.881613686767346e-06, - "loss": 0.351, + "loss": 1.0976, "step": 52640 }, { "epoch": 5.56, - "grad_norm": 8.129233360290527, + "grad_norm": 32.4588623046875, "learning_rate": 8.879501531312705e-06, - "loss": 0.388, + "loss": 0.7449, "step": 52650 }, { "epoch": 5.56, - "grad_norm": 0.26702460646629333, + "grad_norm": 24.785030364990234, "learning_rate": 8.877389375858064e-06, - "loss": 0.4292, + "loss": 0.9637, "step": 52660 }, { "epoch": 5.56, - "grad_norm": 7.507266044616699, + "grad_norm": 37.89152526855469, "learning_rate": 8.875277220403423e-06, - "loss": 0.222, + "loss": 1.0405, "step": 52670 }, { "epoch": 5.56, - "grad_norm": 24.172306060791016, + "grad_norm": 40.05105209350586, "learning_rate": 8.873165064948781e-06, - "loss": 0.4078, + "loss": 0.865, "step": 52680 }, { "epoch": 5.56, - "grad_norm": 12.598453521728516, + "grad_norm": 21.57243537902832, "learning_rate": 8.87105290949414e-06, - "loss": 0.5916, + "loss": 1.4029, "step": 52690 }, { "epoch": 5.57, - "grad_norm": 19.29047966003418, + "grad_norm": 34.1263542175293, "learning_rate": 8.868940754039497e-06, - "loss": 0.4008, + "loss": 0.7596, "step": 52700 }, { "epoch": 5.57, - "grad_norm": 0.3252561688423157, + "grad_norm": 41.755523681640625, "learning_rate": 8.866828598584856e-06, - "loss": 0.5705, + "loss": 1.1037, "step": 52710 }, { "epoch": 5.57, - "grad_norm": 0.6618145108222961, + "grad_norm": 51.09124755859375, "learning_rate": 8.864716443130215e-06, - "loss": 0.3593, + "loss": 0.9058, "step": 52720 }, { "epoch": 5.57, - "grad_norm": 10.47043228149414, + "grad_norm": 43.67372131347656, "learning_rate": 8.862604287675574e-06, - "loss": 0.4254, + "loss": 0.9173, "step": 52730 }, { "epoch": 5.57, - "grad_norm": 0.6694872975349426, + "grad_norm": 35.2297248840332, "learning_rate": 8.860492132220932e-06, - "loss": 0.3656, + "loss": 1.1061, "step": 52740 }, { "epoch": 5.57, - "grad_norm": 33.60746765136719, + "grad_norm": 43.46548843383789, "learning_rate": 8.858379976766291e-06, - "loss": 0.5096, + "loss": 1.1217, "step": 52750 }, { "epoch": 5.57, - "grad_norm": 13.315889358520508, + "grad_norm": 58.27772521972656, "learning_rate": 8.85626782131165e-06, - "loss": 0.7517, + "loss": 1.2443, "step": 52760 }, { "epoch": 5.57, - "grad_norm": 8.683465003967285, + "grad_norm": 54.903141021728516, "learning_rate": 8.854155665857009e-06, - "loss": 0.5307, + "loss": 1.2274, "step": 52770 }, { "epoch": 5.57, - "grad_norm": 0.998941957950592, + "grad_norm": 56.50126647949219, "learning_rate": 8.852043510402366e-06, - "loss": 0.3624, + "loss": 1.0045, "step": 52780 }, { "epoch": 5.58, - "grad_norm": 5.9000020027160645, + "grad_norm": 51.23265838623047, "learning_rate": 8.849931354947725e-06, - "loss": 0.2481, + "loss": 0.8379, "step": 52790 }, { "epoch": 5.58, - "grad_norm": 7.255644798278809, + "grad_norm": 22.4177188873291, "learning_rate": 8.847819199493083e-06, - "loss": 0.3008, + "loss": 0.9606, "step": 52800 }, { "epoch": 5.58, - "grad_norm": 1.7250744104385376, + "grad_norm": 24.726287841796875, "learning_rate": 8.845707044038442e-06, - "loss": 0.5911, + "loss": 0.8687, "step": 52810 }, { "epoch": 5.58, - "grad_norm": 0.6196933388710022, + "grad_norm": 2.572338581085205, "learning_rate": 8.843594888583801e-06, - "loss": 0.422, + "loss": 0.8995, "step": 52820 }, { "epoch": 5.58, - "grad_norm": 1.261627435684204, + "grad_norm": 43.05178451538086, "learning_rate": 8.841482733129158e-06, - "loss": 0.3131, + "loss": 1.0472, "step": 52830 }, { "epoch": 5.58, - "grad_norm": 25.43855857849121, + "grad_norm": 31.182470321655273, "learning_rate": 8.839370577674517e-06, - "loss": 0.4807, + "loss": 1.0189, "step": 52840 }, { "epoch": 5.58, - "grad_norm": 20.03304672241211, + "grad_norm": 51.4515266418457, "learning_rate": 8.837258422219877e-06, - "loss": 0.3129, + "loss": 0.8554, "step": 52850 }, { "epoch": 5.58, - "grad_norm": 19.412813186645508, + "grad_norm": 42.6013069152832, "learning_rate": 8.835146266765234e-06, - "loss": 0.3552, + "loss": 1.1096, "step": 52860 }, { "epoch": 5.58, - "grad_norm": 15.166790962219238, + "grad_norm": 45.31777572631836, "learning_rate": 8.833034111310593e-06, - "loss": 0.5051, + "loss": 1.3278, "step": 52870 }, { "epoch": 5.58, - "grad_norm": 7.5731401443481445, + "grad_norm": 30.40842056274414, "learning_rate": 8.830921955855952e-06, - "loss": 0.5978, + "loss": 1.1715, "step": 52880 }, { "epoch": 5.59, - "grad_norm": 6.7672271728515625, + "grad_norm": 33.023948669433594, "learning_rate": 8.82880980040131e-06, - "loss": 0.5705, + "loss": 0.9219, "step": 52890 }, { "epoch": 5.59, - "grad_norm": 14.640375137329102, + "grad_norm": 25.172042846679688, "learning_rate": 8.82669764494667e-06, - "loss": 0.3925, + "loss": 0.9754, "step": 52900 }, { "epoch": 5.59, - "grad_norm": 53.32052230834961, + "grad_norm": 44.991031646728516, "learning_rate": 8.824585489492027e-06, - "loss": 0.6014, + "loss": 1.0564, "step": 52910 }, { "epoch": 5.59, - "grad_norm": 7.6734299659729, + "grad_norm": 45.64791488647461, "learning_rate": 8.822473334037385e-06, - "loss": 0.7618, + "loss": 1.2232, "step": 52920 }, { "epoch": 5.59, - "grad_norm": 2.302793264389038, + "grad_norm": 34.98084259033203, "learning_rate": 8.820361178582744e-06, - "loss": 0.3275, + "loss": 1.2451, "step": 52930 }, { "epoch": 5.59, - "grad_norm": 1.5283594131469727, + "grad_norm": 35.30180358886719, "learning_rate": 8.818249023128103e-06, - "loss": 0.5211, + "loss": 0.8498, "step": 52940 }, { "epoch": 5.59, - "grad_norm": 10.623392105102539, + "grad_norm": 60.63284683227539, "learning_rate": 8.816136867673462e-06, - "loss": 0.5387, + "loss": 0.9303, "step": 52950 }, { "epoch": 5.59, - "grad_norm": 14.440719604492188, + "grad_norm": 54.152862548828125, "learning_rate": 8.81402471221882e-06, - "loss": 0.3846, + "loss": 1.1413, "step": 52960 }, { "epoch": 5.59, - "grad_norm": 11.474026679992676, + "grad_norm": 26.907617568969727, "learning_rate": 8.811912556764178e-06, - "loss": 0.6309, + "loss": 0.935, "step": 52970 }, { "epoch": 5.6, - "grad_norm": 38.13344955444336, + "grad_norm": 49.45388412475586, "learning_rate": 8.809800401309538e-06, - "loss": 0.544, + "loss": 1.1911, "step": 52980 }, { "epoch": 5.6, - "grad_norm": 15.846040725708008, + "grad_norm": 31.195886611938477, "learning_rate": 8.807688245854895e-06, - "loss": 0.3771, + "loss": 0.9971, "step": 52990 }, { "epoch": 5.6, - "grad_norm": 12.143750190734863, + "grad_norm": 27.627233505249023, "learning_rate": 8.805576090400254e-06, - "loss": 0.7317, + "loss": 1.3762, "step": 53000 }, { "epoch": 5.6, - "grad_norm": 3.1061365604400635, + "grad_norm": 25.064462661743164, "learning_rate": 8.803463934945613e-06, - "loss": 0.3092, + "loss": 0.8882, "step": 53010 }, { "epoch": 5.6, - "grad_norm": 10.019826889038086, + "grad_norm": 33.320308685302734, "learning_rate": 8.801351779490971e-06, - "loss": 0.4127, + "loss": 1.0465, "step": 53020 }, { "epoch": 5.6, - "grad_norm": 41.60369873046875, + "grad_norm": 46.57487106323242, "learning_rate": 8.79923962403633e-06, - "loss": 0.5361, + "loss": 1.118, "step": 53030 }, { "epoch": 5.6, - "grad_norm": 6.526575565338135, + "grad_norm": 42.31201934814453, "learning_rate": 8.797127468581689e-06, - "loss": 0.5442, + "loss": 0.7503, "step": 53040 }, { "epoch": 5.6, - "grad_norm": 12.391229629516602, + "grad_norm": 34.76127624511719, "learning_rate": 8.795015313127046e-06, - "loss": 0.5082, + "loss": 1.1094, "step": 53050 }, { "epoch": 5.6, - "grad_norm": 33.405426025390625, + "grad_norm": 37.11200714111328, "learning_rate": 8.792903157672405e-06, - "loss": 0.4968, + "loss": 0.964, "step": 53060 }, { "epoch": 5.6, - "grad_norm": 7.371888637542725, + "grad_norm": 46.13230514526367, "learning_rate": 8.790791002217765e-06, - "loss": 0.5741, + "loss": 1.3614, "step": 53070 }, { "epoch": 5.61, - "grad_norm": 18.089263916015625, + "grad_norm": 55.837154388427734, "learning_rate": 8.788678846763122e-06, - "loss": 0.4277, + "loss": 1.1802, "step": 53080 }, { "epoch": 5.61, - "grad_norm": 30.57468032836914, + "grad_norm": 23.041059494018555, "learning_rate": 8.786566691308481e-06, - "loss": 0.2638, + "loss": 0.9339, "step": 53090 }, { "epoch": 5.61, - "grad_norm": 17.00840187072754, + "grad_norm": 53.08238983154297, "learning_rate": 8.784454535853838e-06, - "loss": 0.5722, + "loss": 1.2422, "step": 53100 }, { "epoch": 5.61, - "grad_norm": 0.16009706258773804, + "grad_norm": 41.17927169799805, "learning_rate": 8.782342380399199e-06, - "loss": 0.4394, + "loss": 1.1291, "step": 53110 }, { "epoch": 5.61, - "grad_norm": 6.053420543670654, + "grad_norm": 37.23362731933594, "learning_rate": 8.780230224944558e-06, - "loss": 0.5037, + "loss": 1.1833, "step": 53120 }, { "epoch": 5.61, - "grad_norm": 11.250373840332031, + "grad_norm": 37.44936752319336, "learning_rate": 8.778118069489915e-06, - "loss": 0.2657, + "loss": 0.9443, "step": 53130 }, { "epoch": 5.61, - "grad_norm": 4.608458995819092, + "grad_norm": 59.15048599243164, "learning_rate": 8.776005914035273e-06, - "loss": 0.5196, + "loss": 0.9245, "step": 53140 }, { "epoch": 5.61, - "grad_norm": 2.015519857406616, + "grad_norm": 18.322153091430664, "learning_rate": 8.773893758580632e-06, - "loss": 0.3732, + "loss": 0.9959, "step": 53150 }, { "epoch": 5.61, - "grad_norm": 7.628974914550781, + "grad_norm": 43.82000732421875, "learning_rate": 8.771781603125991e-06, - "loss": 0.406, + "loss": 0.9538, "step": 53160 }, { "epoch": 5.62, - "grad_norm": 7.64800500869751, + "grad_norm": 44.65797805786133, "learning_rate": 8.76966944767135e-06, - "loss": 0.5388, + "loss": 1.1323, "step": 53170 }, { "epoch": 5.62, - "grad_norm": 8.273673057556152, + "grad_norm": 40.3377685546875, "learning_rate": 8.767557292216707e-06, - "loss": 0.4313, + "loss": 1.0818, "step": 53180 }, { "epoch": 5.62, - "grad_norm": 10.365989685058594, + "grad_norm": 30.63397789001465, "learning_rate": 8.765445136762066e-06, - "loss": 0.5888, + "loss": 0.9003, "step": 53190 }, { "epoch": 5.62, - "grad_norm": 17.140155792236328, + "grad_norm": 38.31643295288086, "learning_rate": 8.763332981307426e-06, - "loss": 0.6474, + "loss": 0.9689, "step": 53200 }, { "epoch": 5.62, - "grad_norm": 11.962675094604492, + "grad_norm": 35.86984634399414, "learning_rate": 8.761220825852783e-06, - "loss": 0.3514, + "loss": 1.0277, "step": 53210 }, { "epoch": 5.62, - "grad_norm": 3.4789326190948486, + "grad_norm": 39.68478775024414, "learning_rate": 8.759108670398142e-06, - "loss": 0.4203, + "loss": 1.0013, "step": 53220 }, { "epoch": 5.62, - "grad_norm": 31.944671630859375, + "grad_norm": 59.668731689453125, "learning_rate": 8.7569965149435e-06, - "loss": 0.5403, + "loss": 1.4292, "step": 53230 }, { "epoch": 5.62, - "grad_norm": 16.826311111450195, + "grad_norm": 50.86115264892578, "learning_rate": 8.75488435948886e-06, - "loss": 0.3122, + "loss": 0.9355, "step": 53240 }, { "epoch": 5.62, - "grad_norm": 4.117857456207275, + "grad_norm": 36.62916946411133, "learning_rate": 8.752772204034218e-06, - "loss": 0.4077, + "loss": 1.1809, "step": 53250 }, { "epoch": 5.62, - "grad_norm": 6.055971622467041, + "grad_norm": 46.07487487792969, "learning_rate": 8.750660048579575e-06, - "loss": 0.4286, + "loss": 0.939, "step": 53260 }, { "epoch": 5.63, - "grad_norm": 12.05627727508545, + "grad_norm": 73.73023986816406, "learning_rate": 8.748547893124934e-06, - "loss": 0.405, + "loss": 1.0401, "step": 53270 }, { "epoch": 5.63, - "grad_norm": 30.642717361450195, + "grad_norm": 29.62070083618164, "learning_rate": 8.746435737670293e-06, - "loss": 0.5746, + "loss": 1.0644, "step": 53280 }, { "epoch": 5.63, - "grad_norm": 21.605796813964844, + "grad_norm": 53.92266845703125, "learning_rate": 8.744323582215652e-06, - "loss": 0.6049, + "loss": 1.2143, "step": 53290 }, { "epoch": 5.63, - "grad_norm": 3.8045268058776855, + "grad_norm": 19.920896530151367, "learning_rate": 8.74221142676101e-06, - "loss": 0.4283, + "loss": 0.7878, "step": 53300 }, { "epoch": 5.63, - "grad_norm": 18.822357177734375, + "grad_norm": 53.32524490356445, "learning_rate": 8.74009927130637e-06, - "loss": 0.4289, + "loss": 1.0701, "step": 53310 }, { "epoch": 5.63, - "grad_norm": 6.950024604797363, + "grad_norm": 60.55337142944336, "learning_rate": 8.737987115851726e-06, - "loss": 0.4286, + "loss": 1.1218, "step": 53320 }, { "epoch": 5.63, - "grad_norm": 6.944697856903076, + "grad_norm": 38.17780303955078, "learning_rate": 8.735874960397087e-06, - "loss": 0.551, + "loss": 0.9974, "step": 53330 }, { "epoch": 5.63, - "grad_norm": 28.71761703491211, + "grad_norm": 49.44044876098633, "learning_rate": 8.733762804942446e-06, - "loss": 0.5086, + "loss": 0.9307, "step": 53340 }, { "epoch": 5.63, - "grad_norm": 22.260156631469727, + "grad_norm": 56.878013610839844, "learning_rate": 8.731650649487803e-06, - "loss": 0.5184, + "loss": 1.3022, "step": 53350 }, { "epoch": 5.64, - "grad_norm": 8.620725631713867, + "grad_norm": 68.2740249633789, "learning_rate": 8.729538494033161e-06, - "loss": 0.3871, + "loss": 1.4047, "step": 53360 }, { "epoch": 5.64, - "grad_norm": 4.30850887298584, + "grad_norm": 28.14646339416504, "learning_rate": 8.72742633857852e-06, - "loss": 0.4913, + "loss": 0.9576, "step": 53370 }, { "epoch": 5.64, - "grad_norm": 17.96673011779785, + "grad_norm": 74.47918701171875, "learning_rate": 8.725314183123879e-06, - "loss": 0.376, + "loss": 0.9473, "step": 53380 }, { "epoch": 5.64, - "grad_norm": 9.578166007995605, + "grad_norm": 33.43919372558594, "learning_rate": 8.723202027669238e-06, - "loss": 0.2242, + "loss": 0.8878, "step": 53390 }, { "epoch": 5.64, - "grad_norm": 9.654562950134277, + "grad_norm": 45.241615295410156, "learning_rate": 8.721089872214595e-06, - "loss": 0.3948, + "loss": 0.8997, "step": 53400 }, { "epoch": 5.64, - "grad_norm": 4.326845169067383, + "grad_norm": 31.962411880493164, "learning_rate": 8.718977716759954e-06, - "loss": 0.4909, + "loss": 1.0625, "step": 53410 }, { "epoch": 5.64, - "grad_norm": 4.129465579986572, + "grad_norm": 29.068769454956055, "learning_rate": 8.716865561305312e-06, - "loss": 0.4689, + "loss": 0.9023, "step": 53420 }, { "epoch": 5.64, - "grad_norm": 10.999557495117188, + "grad_norm": 37.15690231323242, "learning_rate": 8.714753405850671e-06, - "loss": 0.3515, + "loss": 0.8977, "step": 53430 }, { "epoch": 5.64, - "grad_norm": 12.335480690002441, + "grad_norm": 21.601226806640625, "learning_rate": 8.71264125039603e-06, - "loss": 0.7441, + "loss": 1.0331, "step": 53440 }, { "epoch": 5.64, - "grad_norm": 0.16733352839946747, + "grad_norm": 12.323020935058594, "learning_rate": 8.710529094941387e-06, - "loss": 0.1873, + "loss": 0.7273, "step": 53450 }, { "epoch": 5.65, - "grad_norm": 12.961870193481445, + "grad_norm": 48.62929153442383, "learning_rate": 8.708416939486748e-06, - "loss": 0.4732, + "loss": 1.2128, "step": 53460 }, { "epoch": 5.65, - "grad_norm": 27.923070907592773, + "grad_norm": 38.58031463623047, "learning_rate": 8.706304784032106e-06, - "loss": 0.6551, + "loss": 0.9845, "step": 53470 }, { "epoch": 5.65, - "grad_norm": 4.074800968170166, + "grad_norm": 50.22541809082031, "learning_rate": 8.704192628577463e-06, - "loss": 0.4761, + "loss": 0.904, "step": 53480 }, { "epoch": 5.65, - "grad_norm": 4.211489677429199, + "grad_norm": 28.236408233642578, "learning_rate": 8.702080473122822e-06, - "loss": 0.1739, + "loss": 0.9183, "step": 53490 }, { "epoch": 5.65, - "grad_norm": 15.885076522827148, + "grad_norm": 41.832820892333984, "learning_rate": 8.699968317668181e-06, - "loss": 0.2442, + "loss": 0.9308, "step": 53500 }, { "epoch": 5.65, - "grad_norm": 16.221492767333984, + "grad_norm": 26.918476104736328, "learning_rate": 8.69785616221354e-06, - "loss": 0.4778, + "loss": 0.9084, "step": 53510 }, { "epoch": 5.65, - "grad_norm": 9.006160736083984, + "grad_norm": 39.457557678222656, "learning_rate": 8.695744006758899e-06, - "loss": 0.5513, + "loss": 0.9718, "step": 53520 }, { "epoch": 5.65, - "grad_norm": 14.531543731689453, + "grad_norm": 28.894935607910156, "learning_rate": 8.693631851304257e-06, - "loss": 0.3541, + "loss": 0.8927, "step": 53530 }, { "epoch": 5.65, - "grad_norm": 71.78951263427734, + "grad_norm": 39.94257354736328, "learning_rate": 8.691519695849614e-06, - "loss": 0.6194, + "loss": 1.0934, "step": 53540 }, { "epoch": 5.66, - "grad_norm": 7.366119384765625, + "grad_norm": 56.1662712097168, "learning_rate": 8.689407540394975e-06, - "loss": 0.4463, + "loss": 0.9458, "step": 53550 }, { "epoch": 5.66, - "grad_norm": 7.057335376739502, + "grad_norm": 33.94694900512695, "learning_rate": 8.687295384940332e-06, - "loss": 0.3494, + "loss": 0.9672, "step": 53560 }, { "epoch": 5.66, - "grad_norm": 13.588067054748535, + "grad_norm": 45.2509880065918, "learning_rate": 8.68518322948569e-06, - "loss": 0.195, + "loss": 0.83, "step": 53570 }, { "epoch": 5.66, - "grad_norm": 31.521499633789062, + "grad_norm": 45.76314163208008, "learning_rate": 8.68307107403105e-06, - "loss": 0.6049, + "loss": 1.0562, "step": 53580 }, { "epoch": 5.66, - "grad_norm": 35.36443328857422, + "grad_norm": 51.8542366027832, "learning_rate": 8.680958918576408e-06, - "loss": 0.5031, + "loss": 0.9309, "step": 53590 }, { "epoch": 5.66, - "grad_norm": 18.23590660095215, + "grad_norm": 32.7678337097168, "learning_rate": 8.678846763121767e-06, - "loss": 0.4141, + "loss": 0.8409, "step": 53600 }, { "epoch": 5.66, - "grad_norm": 10.938919067382812, + "grad_norm": 52.888851165771484, "learning_rate": 8.676734607667126e-06, - "loss": 0.3146, + "loss": 1.1496, "step": 53610 }, { "epoch": 5.66, - "grad_norm": 14.884286880493164, + "grad_norm": 43.40687561035156, "learning_rate": 8.674622452212483e-06, - "loss": 0.3275, + "loss": 1.0506, "step": 53620 }, { "epoch": 5.66, - "grad_norm": 9.734366416931152, + "grad_norm": 15.486001968383789, "learning_rate": 8.672510296757842e-06, - "loss": 0.5914, + "loss": 0.8961, "step": 53630 }, { "epoch": 5.66, - "grad_norm": 15.553406715393066, + "grad_norm": 36.71137237548828, "learning_rate": 8.6703981413032e-06, - "loss": 0.4985, + "loss": 1.0161, "step": 53640 }, { "epoch": 5.67, - "grad_norm": 0.9445944428443909, + "grad_norm": 37.08993148803711, "learning_rate": 8.66828598584856e-06, - "loss": 0.3036, + "loss": 0.8205, "step": 53650 }, { "epoch": 5.67, - "grad_norm": 1.9345327615737915, + "grad_norm": 50.42688751220703, "learning_rate": 8.666173830393918e-06, - "loss": 0.3727, + "loss": 0.9104, "step": 53660 }, { "epoch": 5.67, - "grad_norm": 2.263294219970703, + "grad_norm": 46.84025955200195, "learning_rate": 8.664061674939275e-06, - "loss": 0.3144, + "loss": 1.1121, "step": 53670 }, { "epoch": 5.67, - "grad_norm": 33.9700813293457, + "grad_norm": 34.378334045410156, "learning_rate": 8.661949519484636e-06, - "loss": 0.5498, + "loss": 1.2004, "step": 53680 }, { "epoch": 5.67, - "grad_norm": 9.057967185974121, + "grad_norm": 39.50121307373047, "learning_rate": 8.659837364029994e-06, - "loss": 0.5459, + "loss": 1.2114, "step": 53690 }, { "epoch": 5.67, - "grad_norm": 0.7392295598983765, + "grad_norm": 31.61139678955078, "learning_rate": 8.657725208575351e-06, - "loss": 0.6464, + "loss": 1.1687, "step": 53700 }, { "epoch": 5.67, - "grad_norm": 11.438741683959961, + "grad_norm": 36.166744232177734, "learning_rate": 8.65561305312071e-06, - "loss": 0.5108, + "loss": 1.1613, "step": 53710 }, { "epoch": 5.67, - "grad_norm": 2.587207794189453, + "grad_norm": 29.871021270751953, "learning_rate": 8.653500897666069e-06, - "loss": 0.5116, + "loss": 1.2192, "step": 53720 }, { "epoch": 5.67, - "grad_norm": 28.202651977539062, + "grad_norm": 44.63933181762695, "learning_rate": 8.651388742211428e-06, - "loss": 0.5921, + "loss": 0.9393, "step": 53730 }, { "epoch": 5.68, - "grad_norm": 14.039959907531738, + "grad_norm": 33.136871337890625, "learning_rate": 8.649276586756787e-06, - "loss": 0.5556, + "loss": 1.1095, "step": 53740 }, { "epoch": 5.68, - "grad_norm": 19.4936580657959, + "grad_norm": 39.769569396972656, "learning_rate": 8.647164431302144e-06, - "loss": 0.6608, + "loss": 1.1646, "step": 53750 }, { "epoch": 5.68, - "grad_norm": 16.74781036376953, + "grad_norm": 22.896358489990234, "learning_rate": 8.645052275847502e-06, - "loss": 0.5541, + "loss": 0.959, "step": 53760 }, { "epoch": 5.68, - "grad_norm": 1.3817416429519653, + "grad_norm": 26.885814666748047, "learning_rate": 8.642940120392861e-06, - "loss": 0.8225, + "loss": 1.1589, "step": 53770 }, { "epoch": 5.68, - "grad_norm": 2.9683616161346436, + "grad_norm": 32.12527084350586, "learning_rate": 8.64082796493822e-06, - "loss": 0.3162, + "loss": 0.8638, "step": 53780 }, { "epoch": 5.68, - "grad_norm": 7.314965724945068, + "grad_norm": 49.21149444580078, "learning_rate": 8.638715809483579e-06, - "loss": 0.2443, + "loss": 0.8072, "step": 53790 }, { "epoch": 5.68, - "grad_norm": 0.27396515011787415, + "grad_norm": 35.66257858276367, "learning_rate": 8.636603654028938e-06, - "loss": 0.6091, + "loss": 0.8828, "step": 53800 }, { "epoch": 5.68, - "grad_norm": 7.905952453613281, + "grad_norm": 18.70488929748535, "learning_rate": 8.634491498574296e-06, - "loss": 0.1681, + "loss": 0.8253, "step": 53810 }, { "epoch": 5.68, - "grad_norm": 42.271968841552734, + "grad_norm": 43.62200164794922, "learning_rate": 8.632379343119655e-06, - "loss": 0.5387, + "loss": 1.1026, "step": 53820 }, { "epoch": 5.68, - "grad_norm": 17.209880828857422, + "grad_norm": 45.188297271728516, "learning_rate": 8.630267187665012e-06, - "loss": 0.5178, + "loss": 0.972, "step": 53830 }, { "epoch": 5.69, - "grad_norm": 11.658531188964844, + "grad_norm": 14.077997207641602, "learning_rate": 8.628155032210371e-06, - "loss": 0.299, + "loss": 1.102, "step": 53840 }, { "epoch": 5.69, - "grad_norm": 41.24199676513672, + "grad_norm": 25.203088760375977, "learning_rate": 8.62604287675573e-06, - "loss": 0.412, + "loss": 0.6488, "step": 53850 }, { "epoch": 5.69, - "grad_norm": 15.666391372680664, + "grad_norm": 28.553302764892578, "learning_rate": 8.623930721301089e-06, - "loss": 0.627, + "loss": 1.1089, "step": 53860 }, { "epoch": 5.69, - "grad_norm": 0.4031611979007721, + "grad_norm": 35.98609924316406, "learning_rate": 8.621818565846447e-06, - "loss": 0.5047, + "loss": 0.7395, "step": 53870 }, { "epoch": 5.69, - "grad_norm": 2.691329002380371, + "grad_norm": 62.349613189697266, "learning_rate": 8.619706410391806e-06, - "loss": 0.8072, + "loss": 1.194, "step": 53880 }, { "epoch": 5.69, - "grad_norm": 2.916863441467285, + "grad_norm": 25.48724365234375, "learning_rate": 8.617594254937163e-06, - "loss": 0.3933, + "loss": 0.9956, "step": 53890 }, { "epoch": 5.69, - "grad_norm": 8.361230850219727, + "grad_norm": 51.02449035644531, "learning_rate": 8.615482099482522e-06, - "loss": 0.2537, + "loss": 0.8657, "step": 53900 }, { "epoch": 5.69, - "grad_norm": 3.550452470779419, + "grad_norm": 35.48289108276367, "learning_rate": 8.61336994402788e-06, - "loss": 0.5013, + "loss": 1.1317, "step": 53910 }, { "epoch": 5.69, - "grad_norm": 13.829692840576172, + "grad_norm": 51.301963806152344, "learning_rate": 8.61125778857324e-06, - "loss": 0.4569, + "loss": 1.0873, "step": 53920 }, { "epoch": 5.7, - "grad_norm": 14.757414817810059, + "grad_norm": 44.57293701171875, "learning_rate": 8.609145633118598e-06, - "loss": 0.3573, + "loss": 0.8301, "step": 53930 }, { "epoch": 5.7, - "grad_norm": 0.13692305982112885, + "grad_norm": 19.829076766967773, "learning_rate": 8.607033477663957e-06, - "loss": 0.4246, + "loss": 0.8618, "step": 53940 }, { "epoch": 5.7, - "grad_norm": 26.466379165649414, + "grad_norm": 57.04278564453125, "learning_rate": 8.604921322209316e-06, - "loss": 0.1965, + "loss": 1.0036, "step": 53950 }, { "epoch": 5.7, - "grad_norm": 0.9790904521942139, + "grad_norm": 39.889732360839844, "learning_rate": 8.602809166754675e-06, - "loss": 0.5939, + "loss": 0.9847, "step": 53960 }, { "epoch": 5.7, - "grad_norm": 29.782312393188477, + "grad_norm": 36.47330856323242, "learning_rate": 8.600697011300032e-06, - "loss": 0.4411, + "loss": 0.7613, "step": 53970 }, { "epoch": 5.7, - "grad_norm": 17.164180755615234, + "grad_norm": 33.150150299072266, "learning_rate": 8.59858485584539e-06, - "loss": 0.3543, + "loss": 0.7891, "step": 53980 }, { "epoch": 5.7, - "grad_norm": 0.2979027032852173, + "grad_norm": 29.05154037475586, "learning_rate": 8.59647270039075e-06, - "loss": 0.6219, + "loss": 1.1073, "step": 53990 }, { "epoch": 5.7, - "grad_norm": 10.280902862548828, + "grad_norm": 52.440025329589844, "learning_rate": 8.594360544936108e-06, - "loss": 0.3943, + "loss": 1.0389, "step": 54000 }, { "epoch": 5.7, - "grad_norm": 3.8732616901397705, + "grad_norm": 57.39393997192383, "learning_rate": 8.592248389481467e-06, - "loss": 0.3198, + "loss": 0.9563, "step": 54010 }, { "epoch": 5.7, - "grad_norm": 19.008758544921875, + "grad_norm": 30.315338134765625, "learning_rate": 8.590136234026824e-06, - "loss": 0.5257, + "loss": 0.9292, "step": 54020 }, { "epoch": 5.71, - "grad_norm": 15.483039855957031, + "grad_norm": 58.663673400878906, "learning_rate": 8.588024078572184e-06, - "loss": 0.5028, + "loss": 0.9145, "step": 54030 }, { "epoch": 5.71, - "grad_norm": 20.824058532714844, + "grad_norm": 54.44446563720703, "learning_rate": 8.585911923117543e-06, - "loss": 0.6025, + "loss": 1.321, "step": 54040 }, { "epoch": 5.71, - "grad_norm": 15.884958267211914, + "grad_norm": 63.388545989990234, "learning_rate": 8.5837997676629e-06, - "loss": 0.4316, + "loss": 1.0555, "step": 54050 }, { "epoch": 5.71, - "grad_norm": 0.5178261995315552, + "grad_norm": 28.10798454284668, "learning_rate": 8.581687612208259e-06, - "loss": 0.4198, + "loss": 1.0065, "step": 54060 }, { "epoch": 5.71, - "grad_norm": 9.98302173614502, + "grad_norm": 35.089542388916016, "learning_rate": 8.579575456753618e-06, - "loss": 0.7305, + "loss": 0.8885, "step": 54070 }, { "epoch": 5.71, - "grad_norm": 12.389211654663086, + "grad_norm": 38.935462951660156, "learning_rate": 8.577463301298977e-06, - "loss": 0.6528, + "loss": 1.1294, "step": 54080 }, { "epoch": 5.71, - "grad_norm": 20.425933837890625, + "grad_norm": 57.88899230957031, "learning_rate": 8.575351145844335e-06, - "loss": 0.5589, + "loss": 1.204, "step": 54090 }, { "epoch": 5.71, - "grad_norm": 5.616264343261719, + "grad_norm": 26.81241798400879, "learning_rate": 8.573238990389692e-06, - "loss": 0.448, + "loss": 1.2231, "step": 54100 }, { "epoch": 5.71, - "grad_norm": 11.856103897094727, + "grad_norm": 40.93415832519531, "learning_rate": 8.571126834935051e-06, - "loss": 0.3015, + "loss": 0.9447, "step": 54110 }, { "epoch": 5.72, - "grad_norm": 3.4949710369110107, + "grad_norm": 38.02423095703125, "learning_rate": 8.56901467948041e-06, - "loss": 0.2658, + "loss": 0.714, "step": 54120 }, { "epoch": 5.72, - "grad_norm": 16.66756820678711, + "grad_norm": 65.21131896972656, "learning_rate": 8.566902524025769e-06, - "loss": 0.5855, + "loss": 0.9843, "step": 54130 }, { "epoch": 5.72, - "grad_norm": 5.361145973205566, + "grad_norm": 31.03339958190918, "learning_rate": 8.564790368571128e-06, - "loss": 0.228, + "loss": 0.897, "step": 54140 }, { "epoch": 5.72, - "grad_norm": 10.834304809570312, + "grad_norm": 66.06970977783203, "learning_rate": 8.562678213116486e-06, - "loss": 0.2262, + "loss": 1.1334, "step": 54150 }, { "epoch": 5.72, - "grad_norm": 18.447410583496094, + "grad_norm": 39.580379486083984, "learning_rate": 8.560566057661845e-06, - "loss": 0.7362, + "loss": 1.3916, "step": 54160 }, { "epoch": 5.72, - "grad_norm": 12.334428787231445, + "grad_norm": 63.425743103027344, "learning_rate": 8.558453902207204e-06, - "loss": 0.7213, + "loss": 0.9987, "step": 54170 }, { "epoch": 5.72, - "grad_norm": 14.144434928894043, + "grad_norm": 34.35240936279297, "learning_rate": 8.556341746752563e-06, - "loss": 0.4385, + "loss": 1.0159, "step": 54180 }, { "epoch": 5.72, - "grad_norm": 37.99615478515625, + "grad_norm": 58.15909957885742, "learning_rate": 8.55422959129792e-06, - "loss": 0.3448, + "loss": 1.0798, "step": 54190 }, { "epoch": 5.72, - "grad_norm": 18.191436767578125, + "grad_norm": 30.761634826660156, "learning_rate": 8.552117435843279e-06, - "loss": 0.5345, + "loss": 1.2556, "step": 54200 }, { "epoch": 5.72, - "grad_norm": 0.522146463394165, + "grad_norm": 41.54003143310547, "learning_rate": 8.550005280388637e-06, - "loss": 0.482, + "loss": 0.9148, "step": 54210 }, { "epoch": 5.73, - "grad_norm": 1.7854605913162231, + "grad_norm": 34.325069427490234, "learning_rate": 8.547893124933996e-06, - "loss": 0.3945, + "loss": 0.8098, "step": 54220 }, { "epoch": 5.73, - "grad_norm": 16.609893798828125, + "grad_norm": 67.38015747070312, "learning_rate": 8.545780969479355e-06, - "loss": 0.3837, + "loss": 1.1624, "step": 54230 }, { "epoch": 5.73, - "grad_norm": 14.450627326965332, + "grad_norm": 34.14960479736328, "learning_rate": 8.543668814024712e-06, - "loss": 0.5219, + "loss": 0.673, "step": 54240 }, { "epoch": 5.73, - "grad_norm": 18.07504653930664, + "grad_norm": 45.1065559387207, "learning_rate": 8.54155665857007e-06, - "loss": 0.4135, + "loss": 1.0153, "step": 54250 }, { "epoch": 5.73, - "grad_norm": 6.778407573699951, + "grad_norm": 35.84077835083008, "learning_rate": 8.539444503115431e-06, - "loss": 0.556, + "loss": 1.1136, "step": 54260 }, { "epoch": 5.73, - "grad_norm": 2.7398526668548584, + "grad_norm": 87.35639190673828, "learning_rate": 8.537332347660788e-06, - "loss": 0.5436, + "loss": 1.196, "step": 54270 }, { "epoch": 5.73, - "grad_norm": 2.7871885299682617, + "grad_norm": 35.40737533569336, "learning_rate": 8.535220192206147e-06, - "loss": 0.4491, + "loss": 1.3373, "step": 54280 }, { "epoch": 5.73, - "grad_norm": 19.284969329833984, + "grad_norm": 29.067035675048828, "learning_rate": 8.533108036751506e-06, - "loss": 0.6235, + "loss": 1.1259, "step": 54290 }, { "epoch": 5.73, - "grad_norm": 5.624741077423096, + "grad_norm": 30.376707077026367, "learning_rate": 8.530995881296865e-06, - "loss": 0.3023, + "loss": 0.6411, "step": 54300 }, { "epoch": 5.74, - "grad_norm": 213.96673583984375, + "grad_norm": 35.516719818115234, "learning_rate": 8.528883725842223e-06, - "loss": 0.3967, + "loss": 0.929, "step": 54310 }, { "epoch": 5.74, - "grad_norm": 16.381223678588867, + "grad_norm": 36.128517150878906, "learning_rate": 8.52677157038758e-06, - "loss": 0.2085, + "loss": 0.7189, "step": 54320 }, { "epoch": 5.74, - "grad_norm": 16.330982208251953, + "grad_norm": 22.23084259033203, "learning_rate": 8.52465941493294e-06, - "loss": 0.3818, + "loss": 1.0772, "step": 54330 }, { "epoch": 5.74, - "grad_norm": 12.868639945983887, + "grad_norm": 43.891658782958984, "learning_rate": 8.522547259478298e-06, - "loss": 0.4197, + "loss": 1.1521, "step": 54340 }, { "epoch": 5.74, - "grad_norm": 12.116636276245117, + "grad_norm": 29.69293212890625, "learning_rate": 8.520435104023657e-06, - "loss": 0.3872, + "loss": 0.7894, "step": 54350 }, { "epoch": 5.74, - "grad_norm": 20.52505111694336, + "grad_norm": 29.44789695739746, "learning_rate": 8.518322948569016e-06, - "loss": 0.3636, + "loss": 0.9026, "step": 54360 }, { "epoch": 5.74, - "grad_norm": 22.31536293029785, + "grad_norm": 36.378562927246094, "learning_rate": 8.516210793114373e-06, - "loss": 0.4876, + "loss": 1.2236, "step": 54370 }, { "epoch": 5.74, - "grad_norm": 19.029298782348633, + "grad_norm": 22.530004501342773, "learning_rate": 8.514098637659731e-06, - "loss": 0.1524, + "loss": 0.7855, "step": 54380 }, { "epoch": 5.74, - "grad_norm": 12.825210571289062, + "grad_norm": 25.5373592376709, "learning_rate": 8.511986482205092e-06, - "loss": 0.4212, + "loss": 0.8646, "step": 54390 }, { "epoch": 5.75, - "grad_norm": 8.829952239990234, + "grad_norm": 14.71876049041748, "learning_rate": 8.509874326750449e-06, - "loss": 0.3943, + "loss": 1.2664, "step": 54400 }, { "epoch": 5.75, - "grad_norm": 28.239059448242188, + "grad_norm": 46.972347259521484, "learning_rate": 8.507762171295808e-06, - "loss": 0.3934, + "loss": 1.0493, "step": 54410 }, { "epoch": 5.75, - "grad_norm": 19.267925262451172, + "grad_norm": 49.05939483642578, "learning_rate": 8.505650015841167e-06, - "loss": 0.5058, + "loss": 1.0069, "step": 54420 }, { "epoch": 5.75, - "grad_norm": 125.47604370117188, + "grad_norm": 24.772579193115234, "learning_rate": 8.503537860386525e-06, - "loss": 0.5844, + "loss": 0.8969, "step": 54430 }, { "epoch": 5.75, - "grad_norm": 14.013647079467773, + "grad_norm": 46.832969665527344, "learning_rate": 8.501425704931884e-06, - "loss": 0.1988, + "loss": 1.0777, "step": 54440 }, { "epoch": 5.75, - "grad_norm": 4.347684860229492, + "grad_norm": 47.998531341552734, "learning_rate": 8.499313549477243e-06, - "loss": 0.4288, + "loss": 1.0289, "step": 54450 }, { "epoch": 5.75, - "grad_norm": 12.840757369995117, + "grad_norm": 14.880317687988281, "learning_rate": 8.4972013940226e-06, - "loss": 0.3376, + "loss": 0.8563, "step": 54460 }, { "epoch": 5.75, - "grad_norm": 35.47502136230469, + "grad_norm": 46.65787124633789, "learning_rate": 8.495089238567959e-06, - "loss": 0.5521, + "loss": 1.1248, "step": 54470 }, { "epoch": 5.75, - "grad_norm": 12.306479454040527, + "grad_norm": 24.317758560180664, "learning_rate": 8.492977083113318e-06, - "loss": 0.2561, + "loss": 0.8218, "step": 54480 }, { "epoch": 5.75, - "grad_norm": 17.47209358215332, + "grad_norm": 40.16591262817383, "learning_rate": 8.490864927658676e-06, - "loss": 0.5544, + "loss": 1.0514, "step": 54490 }, { "epoch": 5.76, - "grad_norm": 11.495793342590332, + "grad_norm": 18.467464447021484, "learning_rate": 8.488752772204035e-06, - "loss": 0.3245, + "loss": 0.9983, "step": 54500 }, { "epoch": 5.76, - "grad_norm": 12.436441421508789, + "grad_norm": 40.362396240234375, "learning_rate": 8.486640616749394e-06, - "loss": 0.1909, + "loss": 0.7541, "step": 54510 }, { "epoch": 5.76, - "grad_norm": 14.908339500427246, + "grad_norm": 56.259620666503906, "learning_rate": 8.484528461294753e-06, - "loss": 0.4634, + "loss": 1.0112, "step": 54520 }, { "epoch": 5.76, - "grad_norm": 11.170731544494629, + "grad_norm": 29.31123161315918, "learning_rate": 8.482416305840111e-06, - "loss": 0.4047, + "loss": 1.1122, "step": 54530 }, { "epoch": 5.76, - "grad_norm": 5.22966194152832, + "grad_norm": 41.07419967651367, "learning_rate": 8.480304150385469e-06, - "loss": 0.4497, + "loss": 1.1557, "step": 54540 }, { "epoch": 5.76, - "grad_norm": 5.354905605316162, + "grad_norm": 55.205753326416016, "learning_rate": 8.478191994930827e-06, - "loss": 0.4623, + "loss": 0.9074, "step": 54550 }, { "epoch": 5.76, - "grad_norm": 0.4343416392803192, + "grad_norm": 44.016258239746094, "learning_rate": 8.476079839476186e-06, - "loss": 0.4144, + "loss": 1.2228, "step": 54560 }, { "epoch": 5.76, - "grad_norm": 5.867797374725342, + "grad_norm": 39.522003173828125, "learning_rate": 8.473967684021545e-06, - "loss": 0.2828, + "loss": 0.8863, "step": 54570 }, { "epoch": 5.76, - "grad_norm": 0.5041202306747437, + "grad_norm": 28.898008346557617, "learning_rate": 8.471855528566904e-06, - "loss": 0.2764, + "loss": 0.7547, "step": 54580 }, { "epoch": 5.77, - "grad_norm": 58.363407135009766, + "grad_norm": 37.27031707763672, "learning_rate": 8.46974337311226e-06, - "loss": 0.3532, + "loss": 0.881, "step": 54590 }, { "epoch": 5.77, - "grad_norm": 7.588818550109863, + "grad_norm": 29.908695220947266, "learning_rate": 8.46763121765762e-06, - "loss": 0.4065, + "loss": 1.147, "step": 54600 }, { "epoch": 5.77, - "grad_norm": 21.030244827270508, + "grad_norm": 46.55173110961914, "learning_rate": 8.46551906220298e-06, - "loss": 0.4397, + "loss": 0.8749, "step": 54610 }, { "epoch": 5.77, - "grad_norm": 5.120497226715088, + "grad_norm": 45.98313522338867, "learning_rate": 8.463406906748337e-06, - "loss": 0.3718, + "loss": 1.0421, "step": 54620 }, { "epoch": 5.77, - "grad_norm": 12.24535083770752, + "grad_norm": 46.82822799682617, "learning_rate": 8.461294751293696e-06, - "loss": 0.4895, + "loss": 0.9831, "step": 54630 }, { "epoch": 5.77, - "grad_norm": 0.397342711687088, + "grad_norm": 16.110950469970703, "learning_rate": 8.459182595839055e-06, - "loss": 0.2822, + "loss": 0.5725, "step": 54640 }, { "epoch": 5.77, - "grad_norm": 11.50898551940918, + "grad_norm": 10.996248245239258, "learning_rate": 8.457070440384413e-06, - "loss": 0.5568, + "loss": 0.946, "step": 54650 }, { "epoch": 5.77, - "grad_norm": 15.1553955078125, + "grad_norm": 40.4779167175293, "learning_rate": 8.454958284929772e-06, - "loss": 0.6642, + "loss": 1.5633, "step": 54660 }, { "epoch": 5.77, - "grad_norm": 0.20014157891273499, + "grad_norm": 60.39754867553711, "learning_rate": 8.45284612947513e-06, - "loss": 0.3313, + "loss": 0.7958, "step": 54670 }, { "epoch": 5.77, - "grad_norm": 21.773569107055664, + "grad_norm": 42.84486389160156, "learning_rate": 8.450733974020488e-06, - "loss": 0.5623, + "loss": 0.9511, "step": 54680 }, { "epoch": 5.78, - "grad_norm": 10.99415111541748, + "grad_norm": 49.91450119018555, "learning_rate": 8.448621818565847e-06, - "loss": 0.5463, + "loss": 1.1731, "step": 54690 }, { "epoch": 5.78, - "grad_norm": 13.696125030517578, + "grad_norm": 28.885217666625977, "learning_rate": 8.446509663111206e-06, - "loss": 0.4383, + "loss": 1.0055, "step": 54700 }, { "epoch": 5.78, - "grad_norm": 10.373237609863281, + "grad_norm": 57.37950897216797, "learning_rate": 8.444397507656564e-06, - "loss": 0.5059, + "loss": 1.0154, "step": 54710 }, { "epoch": 5.78, - "grad_norm": 13.398993492126465, + "grad_norm": 27.893404006958008, "learning_rate": 8.442285352201923e-06, - "loss": 0.5025, + "loss": 1.0312, "step": 54720 }, { "epoch": 5.78, - "grad_norm": 7.310949325561523, + "grad_norm": 40.976783752441406, "learning_rate": 8.44017319674728e-06, - "loss": 0.6605, + "loss": 1.1785, "step": 54730 }, { "epoch": 5.78, - "grad_norm": 12.873988151550293, + "grad_norm": 41.959049224853516, "learning_rate": 8.43806104129264e-06, - "loss": 0.5998, + "loss": 1.4401, "step": 54740 }, { "epoch": 5.78, - "grad_norm": 9.77197551727295, + "grad_norm": 40.757293701171875, "learning_rate": 8.435948885837998e-06, - "loss": 0.5929, + "loss": 0.8856, "step": 54750 }, { "epoch": 5.78, - "grad_norm": 17.760812759399414, + "grad_norm": 25.630413055419922, "learning_rate": 8.433836730383357e-06, - "loss": 0.3947, + "loss": 1.1394, "step": 54760 }, { "epoch": 5.78, - "grad_norm": 57.17326736450195, + "grad_norm": 21.48699378967285, "learning_rate": 8.431724574928715e-06, - "loss": 0.34, + "loss": 0.8355, "step": 54770 }, { "epoch": 5.79, - "grad_norm": 8.257134437561035, + "grad_norm": 35.7618408203125, "learning_rate": 8.429612419474074e-06, - "loss": 0.1102, + "loss": 0.9204, "step": 54780 }, { "epoch": 5.79, - "grad_norm": 61.7514533996582, + "grad_norm": 29.675310134887695, "learning_rate": 8.427500264019433e-06, - "loss": 0.2637, + "loss": 0.5818, "step": 54790 }, { "epoch": 5.79, - "grad_norm": 14.189140319824219, + "grad_norm": 37.00226974487305, "learning_rate": 8.425388108564792e-06, - "loss": 0.5006, + "loss": 1.0628, "step": 54800 }, { "epoch": 5.79, - "grad_norm": 6.212613582611084, + "grad_norm": 32.12263870239258, "learning_rate": 8.423275953110149e-06, - "loss": 0.2692, + "loss": 0.9266, "step": 54810 }, { "epoch": 5.79, - "grad_norm": 6.659351348876953, + "grad_norm": 38.7835693359375, "learning_rate": 8.421163797655508e-06, - "loss": 0.2997, + "loss": 0.8237, "step": 54820 }, { "epoch": 5.79, - "grad_norm": 22.881776809692383, + "grad_norm": 38.087745666503906, "learning_rate": 8.419051642200866e-06, - "loss": 0.6384, + "loss": 1.1101, "step": 54830 }, { "epoch": 5.79, - "grad_norm": 0.2645762264728546, + "grad_norm": 37.30086898803711, "learning_rate": 8.416939486746225e-06, - "loss": 0.2748, + "loss": 0.8385, "step": 54840 }, { "epoch": 5.79, - "grad_norm": 11.591886520385742, + "grad_norm": 33.225250244140625, "learning_rate": 8.414827331291584e-06, - "loss": 0.5107, + "loss": 0.9625, "step": 54850 }, { "epoch": 5.79, - "grad_norm": 14.395812034606934, + "grad_norm": 35.26639938354492, "learning_rate": 8.412715175836943e-06, - "loss": 0.3041, + "loss": 0.9195, "step": 54860 }, { "epoch": 5.79, - "grad_norm": 10.103325843811035, + "grad_norm": 47.62021255493164, "learning_rate": 8.410603020382301e-06, - "loss": 0.273, + "loss": 0.7366, "step": 54870 }, { "epoch": 5.8, - "grad_norm": 1.1887751817703247, + "grad_norm": 37.418052673339844, "learning_rate": 8.40849086492766e-06, - "loss": 0.1208, + "loss": 0.9216, "step": 54880 }, { "epoch": 5.8, - "grad_norm": 10.999914169311523, + "grad_norm": 45.48420715332031, "learning_rate": 8.406378709473017e-06, - "loss": 0.4478, + "loss": 1.1568, "step": 54890 }, { "epoch": 5.8, - "grad_norm": 34.66856384277344, + "grad_norm": 54.08367919921875, "learning_rate": 8.404266554018376e-06, - "loss": 0.3334, + "loss": 0.7636, "step": 54900 }, { "epoch": 5.8, - "grad_norm": 4.786667823791504, + "grad_norm": 33.906978607177734, "learning_rate": 8.402154398563735e-06, - "loss": 0.5351, + "loss": 1.3064, "step": 54910 }, { "epoch": 5.8, - "grad_norm": 9.951019287109375, + "grad_norm": 41.92189407348633, "learning_rate": 8.400042243109094e-06, - "loss": 0.4914, + "loss": 1.2883, "step": 54920 }, { "epoch": 5.8, - "grad_norm": 18.403371810913086, + "grad_norm": 34.561771392822266, "learning_rate": 8.397930087654452e-06, - "loss": 0.6345, + "loss": 0.9976, "step": 54930 }, { "epoch": 5.8, - "grad_norm": 0.3046974241733551, + "grad_norm": 35.65577697753906, "learning_rate": 8.39581793219981e-06, - "loss": 0.3889, + "loss": 0.9061, "step": 54940 }, { "epoch": 5.8, - "grad_norm": 5.062785625457764, + "grad_norm": 29.50633430480957, "learning_rate": 8.393705776745168e-06, - "loss": 0.3142, + "loss": 0.8436, "step": 54950 }, { "epoch": 5.8, - "grad_norm": 5.410884857177734, + "grad_norm": 58.029945373535156, "learning_rate": 8.391593621290529e-06, - "loss": 0.5895, + "loss": 1.2069, "step": 54960 }, { "epoch": 5.81, - "grad_norm": 21.155729293823242, + "grad_norm": 59.25861740112305, "learning_rate": 8.389481465835886e-06, - "loss": 0.221, + "loss": 0.9565, "step": 54970 }, { "epoch": 5.81, - "grad_norm": 8.176093101501465, + "grad_norm": 41.72263717651367, "learning_rate": 8.387369310381245e-06, - "loss": 0.4086, + "loss": 1.1389, "step": 54980 }, { "epoch": 5.81, - "grad_norm": 18.382362365722656, + "grad_norm": 32.01761245727539, "learning_rate": 8.385257154926603e-06, - "loss": 0.3387, + "loss": 0.9903, "step": 54990 }, { "epoch": 5.81, - "grad_norm": 14.280942916870117, + "grad_norm": 39.615447998046875, "learning_rate": 8.383144999471962e-06, - "loss": 0.5439, + "loss": 0.9649, "step": 55000 }, { "epoch": 5.81, - "grad_norm": 6.062483787536621, + "grad_norm": 19.160968780517578, "learning_rate": 8.381032844017321e-06, - "loss": 0.4593, + "loss": 1.138, "step": 55010 }, { "epoch": 5.81, - "grad_norm": 9.147607803344727, + "grad_norm": 74.858642578125, "learning_rate": 8.378920688562678e-06, - "loss": 0.3513, + "loss": 0.9419, "step": 55020 }, { "epoch": 5.81, - "grad_norm": 21.67814064025879, + "grad_norm": 16.80845832824707, "learning_rate": 8.376808533108037e-06, - "loss": 0.1747, + "loss": 0.911, "step": 55030 }, { "epoch": 5.81, - "grad_norm": 24.228612899780273, + "grad_norm": 56.515750885009766, "learning_rate": 8.374696377653396e-06, - "loss": 0.4851, + "loss": 1.0382, "step": 55040 }, { "epoch": 5.81, - "grad_norm": 25.77027702331543, + "grad_norm": 34.55647659301758, "learning_rate": 8.372584222198754e-06, - "loss": 0.3894, + "loss": 0.8863, "step": 55050 }, { "epoch": 5.81, - "grad_norm": 15.536456108093262, + "grad_norm": 43.54651641845703, "learning_rate": 8.370472066744113e-06, - "loss": 0.9381, + "loss": 1.2177, "step": 55060 }, { "epoch": 5.82, - "grad_norm": 5.641781806945801, + "grad_norm": 25.72386360168457, "learning_rate": 8.368359911289472e-06, - "loss": 0.3851, + "loss": 0.9074, "step": 55070 }, { "epoch": 5.82, - "grad_norm": 1.0839321613311768, + "grad_norm": 36.352012634277344, "learning_rate": 8.366247755834829e-06, - "loss": 0.3436, + "loss": 1.0053, "step": 55080 }, { "epoch": 5.82, - "grad_norm": 15.435956001281738, + "grad_norm": 46.6752815246582, "learning_rate": 8.36413560038019e-06, - "loss": 0.3771, + "loss": 1.2601, "step": 55090 }, { "epoch": 5.82, - "grad_norm": 8.814895629882812, + "grad_norm": 36.077205657958984, "learning_rate": 8.362023444925548e-06, - "loss": 0.5769, + "loss": 1.155, "step": 55100 }, { "epoch": 5.82, - "grad_norm": 7.001197338104248, + "grad_norm": 29.795251846313477, "learning_rate": 8.359911289470905e-06, - "loss": 0.3644, + "loss": 0.8101, "step": 55110 }, { "epoch": 5.82, - "grad_norm": 17.694665908813477, + "grad_norm": 45.40406036376953, "learning_rate": 8.357799134016264e-06, - "loss": 0.2242, + "loss": 0.5436, "step": 55120 }, { "epoch": 5.82, - "grad_norm": 15.622489929199219, + "grad_norm": 34.506134033203125, "learning_rate": 8.355686978561623e-06, - "loss": 0.3939, + "loss": 1.0029, "step": 55130 }, { "epoch": 5.82, - "grad_norm": 24.238859176635742, + "grad_norm": 34.01064682006836, "learning_rate": 8.353574823106982e-06, - "loss": 0.6613, + "loss": 0.9237, "step": 55140 }, { "epoch": 5.82, - "grad_norm": 38.87131881713867, + "grad_norm": 39.46268844604492, "learning_rate": 8.35146266765234e-06, - "loss": 0.379, + "loss": 0.6977, "step": 55150 }, { "epoch": 5.83, - "grad_norm": 16.314546585083008, + "grad_norm": 34.187747955322266, "learning_rate": 8.349350512197698e-06, - "loss": 0.7721, + "loss": 0.9843, "step": 55160 }, { "epoch": 5.83, - "grad_norm": 1.0225557088851929, + "grad_norm": 59.787654876708984, "learning_rate": 8.347238356743056e-06, - "loss": 0.6637, + "loss": 1.016, "step": 55170 }, { "epoch": 5.83, - "grad_norm": 25.84126853942871, + "grad_norm": 30.98504066467285, "learning_rate": 8.345126201288415e-06, - "loss": 0.3985, + "loss": 0.9748, "step": 55180 }, { "epoch": 5.83, - "grad_norm": 16.425098419189453, + "grad_norm": 51.9086799621582, "learning_rate": 8.343014045833774e-06, - "loss": 0.6058, + "loss": 1.013, "step": 55190 }, { "epoch": 5.83, - "grad_norm": 2.7808310985565186, + "grad_norm": 33.87501907348633, "learning_rate": 8.340901890379133e-06, - "loss": 0.3416, + "loss": 1.2655, "step": 55200 }, { "epoch": 5.83, - "grad_norm": 18.714923858642578, + "grad_norm": 44.202659606933594, "learning_rate": 8.33878973492449e-06, - "loss": 0.5204, + "loss": 0.9048, "step": 55210 }, { "epoch": 5.83, - "grad_norm": 0.5473718643188477, + "grad_norm": 48.17640686035156, "learning_rate": 8.33667757946985e-06, - "loss": 0.3053, + "loss": 1.0334, "step": 55220 }, { "epoch": 5.83, - "grad_norm": 8.088127136230469, + "grad_norm": 35.84490966796875, "learning_rate": 8.334565424015209e-06, - "loss": 0.2137, + "loss": 0.8143, "step": 55230 }, { "epoch": 5.83, - "grad_norm": 2.6003427505493164, + "grad_norm": 24.951810836791992, "learning_rate": 8.332453268560566e-06, - "loss": 0.3914, + "loss": 0.8931, "step": 55240 }, { "epoch": 5.83, - "grad_norm": 9.57520866394043, + "grad_norm": 36.245201110839844, "learning_rate": 8.330341113105925e-06, - "loss": 0.4541, + "loss": 0.9424, "step": 55250 }, { "epoch": 5.84, - "grad_norm": 6.633851528167725, + "grad_norm": 26.03984832763672, "learning_rate": 8.328228957651284e-06, - "loss": 0.7249, + "loss": 1.0347, "step": 55260 }, { "epoch": 5.84, - "grad_norm": 7.172203540802002, + "grad_norm": 40.384891510009766, "learning_rate": 8.326116802196642e-06, - "loss": 0.4138, + "loss": 0.679, "step": 55270 }, { "epoch": 5.84, - "grad_norm": 12.038793563842773, + "grad_norm": 57.35194778442383, "learning_rate": 8.324004646742001e-06, - "loss": 0.5963, + "loss": 1.0916, "step": 55280 }, { "epoch": 5.84, - "grad_norm": 0.45847946405410767, + "grad_norm": 53.2453727722168, "learning_rate": 8.321892491287358e-06, - "loss": 0.1419, + "loss": 0.9403, "step": 55290 }, { "epoch": 5.84, - "grad_norm": 13.21299934387207, + "grad_norm": 51.242958068847656, "learning_rate": 8.319780335832717e-06, - "loss": 0.4707, + "loss": 1.2361, "step": 55300 }, { "epoch": 5.84, - "grad_norm": 0.9306353330612183, + "grad_norm": 43.2484016418457, "learning_rate": 8.317668180378078e-06, - "loss": 0.3378, + "loss": 0.9532, "step": 55310 }, { "epoch": 5.84, - "grad_norm": 33.75779342651367, + "grad_norm": 26.97272300720215, "learning_rate": 8.315556024923435e-06, - "loss": 0.3409, + "loss": 0.9666, "step": 55320 }, { "epoch": 5.84, - "grad_norm": 15.862683296203613, + "grad_norm": 31.891876220703125, "learning_rate": 8.313443869468793e-06, - "loss": 0.6139, + "loss": 1.177, "step": 55330 }, { "epoch": 5.84, - "grad_norm": 10.047200202941895, + "grad_norm": 30.343725204467773, "learning_rate": 8.311331714014152e-06, - "loss": 0.2886, + "loss": 0.7133, "step": 55340 }, { "epoch": 5.85, - "grad_norm": 5.752560615539551, + "grad_norm": 25.39128303527832, "learning_rate": 8.309219558559511e-06, - "loss": 0.4157, + "loss": 1.2225, "step": 55350 }, { "epoch": 5.85, - "grad_norm": 8.892566680908203, + "grad_norm": 29.549684524536133, "learning_rate": 8.30710740310487e-06, - "loss": 0.3567, + "loss": 0.7739, "step": 55360 }, { "epoch": 5.85, - "grad_norm": 1.7758699655532837, + "grad_norm": 8.841996192932129, "learning_rate": 8.304995247650229e-06, - "loss": 0.4884, + "loss": 0.9379, "step": 55370 }, { "epoch": 5.85, - "grad_norm": 11.582368850708008, + "grad_norm": 27.75384521484375, "learning_rate": 8.302883092195586e-06, - "loss": 0.4156, + "loss": 1.0464, "step": 55380 }, { "epoch": 5.85, - "grad_norm": 21.879318237304688, + "grad_norm": 32.39195251464844, "learning_rate": 8.300770936740944e-06, - "loss": 0.3896, + "loss": 1.2535, "step": 55390 }, { "epoch": 5.85, - "grad_norm": 25.33013343811035, + "grad_norm": 49.455230712890625, "learning_rate": 8.298658781286303e-06, - "loss": 0.6066, + "loss": 0.9373, "step": 55400 }, { "epoch": 5.85, - "grad_norm": 23.465965270996094, + "grad_norm": 39.07929229736328, "learning_rate": 8.296546625831662e-06, - "loss": 0.3832, + "loss": 1.03, "step": 55410 }, { "epoch": 5.85, - "grad_norm": 1.4727504253387451, + "grad_norm": 30.057510375976562, "learning_rate": 8.29443447037702e-06, - "loss": 0.4594, + "loss": 0.9076, "step": 55420 }, { "epoch": 5.85, - "grad_norm": 19.271137237548828, + "grad_norm": 32.75689697265625, "learning_rate": 8.292322314922378e-06, - "loss": 0.4181, + "loss": 0.9335, "step": 55430 }, { "epoch": 5.85, - "grad_norm": 28.721891403198242, + "grad_norm": 48.075843811035156, "learning_rate": 8.290210159467738e-06, - "loss": 0.2733, + "loss": 0.9565, "step": 55440 }, { "epoch": 5.86, - "grad_norm": 22.722026824951172, + "grad_norm": 35.56626892089844, "learning_rate": 8.288098004013097e-06, - "loss": 0.2563, + "loss": 1.0384, "step": 55450 }, { "epoch": 5.86, - "grad_norm": 14.04690170288086, + "grad_norm": 23.507404327392578, "learning_rate": 8.285985848558454e-06, - "loss": 0.3946, + "loss": 1.0501, "step": 55460 }, { "epoch": 5.86, - "grad_norm": 14.291967391967773, + "grad_norm": 44.51312255859375, "learning_rate": 8.283873693103813e-06, - "loss": 0.4986, + "loss": 1.0972, "step": 55470 }, { "epoch": 5.86, - "grad_norm": 5.888954162597656, + "grad_norm": 48.959224700927734, "learning_rate": 8.281761537649172e-06, - "loss": 0.3463, + "loss": 1.0092, "step": 55480 }, { "epoch": 5.86, - "grad_norm": 23.515403747558594, + "grad_norm": 39.07961654663086, "learning_rate": 8.27964938219453e-06, - "loss": 0.3835, + "loss": 1.1616, "step": 55490 }, { "epoch": 5.86, - "grad_norm": 26.261463165283203, + "grad_norm": 36.13347625732422, "learning_rate": 8.27753722673989e-06, - "loss": 0.4392, + "loss": 1.0577, "step": 55500 }, { "epoch": 5.86, - "grad_norm": 3.5114364624023438, + "grad_norm": 29.1917724609375, "learning_rate": 8.275425071285246e-06, - "loss": 0.5053, + "loss": 0.632, "step": 55510 }, { "epoch": 5.86, - "grad_norm": 23.292909622192383, + "grad_norm": 42.41859436035156, "learning_rate": 8.273312915830605e-06, - "loss": 0.788, + "loss": 1.2884, "step": 55520 }, { "epoch": 5.86, - "grad_norm": 1.9034205675125122, + "grad_norm": 15.811999320983887, "learning_rate": 8.271200760375964e-06, - "loss": 0.3092, + "loss": 0.5114, "step": 55530 }, { "epoch": 5.87, - "grad_norm": 0.09580480307340622, + "grad_norm": 18.67198944091797, "learning_rate": 8.269088604921323e-06, - "loss": 0.3381, + "loss": 0.948, "step": 55540 }, { "epoch": 5.87, - "grad_norm": 9.7218656539917, + "grad_norm": 51.93266677856445, "learning_rate": 8.266976449466682e-06, - "loss": 0.2108, + "loss": 1.0329, "step": 55550 }, { "epoch": 5.87, - "grad_norm": 9.143009185791016, + "grad_norm": 38.14371109008789, "learning_rate": 8.26486429401204e-06, - "loss": 0.3602, + "loss": 0.8535, "step": 55560 }, { "epoch": 5.87, - "grad_norm": 3.7463579177856445, + "grad_norm": 47.44397735595703, "learning_rate": 8.262752138557399e-06, - "loss": 0.4133, + "loss": 1.0777, "step": 55570 }, { "epoch": 5.87, - "grad_norm": 11.019083023071289, + "grad_norm": 23.406206130981445, "learning_rate": 8.260639983102758e-06, - "loss": 0.4484, + "loss": 1.0257, "step": 55580 }, { "epoch": 5.87, - "grad_norm": 1.247473955154419, + "grad_norm": 42.813533782958984, "learning_rate": 8.258527827648115e-06, - "loss": 0.4699, + "loss": 1.2984, "step": 55590 }, { "epoch": 5.87, - "grad_norm": 0.07455097138881683, + "grad_norm": 25.389404296875, "learning_rate": 8.256415672193474e-06, - "loss": 0.5162, + "loss": 0.8445, "step": 55600 }, { "epoch": 5.87, - "grad_norm": 33.07672882080078, + "grad_norm": 41.20268249511719, "learning_rate": 8.254303516738832e-06, - "loss": 0.4005, + "loss": 0.9222, "step": 55610 }, { "epoch": 5.87, - "grad_norm": 26.938291549682617, + "grad_norm": 51.78032684326172, "learning_rate": 8.252191361284191e-06, - "loss": 0.5613, + "loss": 1.2291, "step": 55620 }, { "epoch": 5.87, - "grad_norm": 12.88986587524414, + "grad_norm": 44.691776275634766, "learning_rate": 8.25007920582955e-06, - "loss": 0.6041, + "loss": 1.237, "step": 55630 }, { "epoch": 5.88, - "grad_norm": 12.1398344039917, + "grad_norm": 54.813289642333984, "learning_rate": 8.247967050374909e-06, - "loss": 0.3949, + "loss": 1.1647, "step": 55640 }, { "epoch": 5.88, - "grad_norm": 34.420143127441406, + "grad_norm": 41.59181213378906, "learning_rate": 8.245854894920266e-06, - "loss": 0.5323, + "loss": 1.1064, "step": 55650 }, { "epoch": 5.88, - "grad_norm": 13.106616973876953, + "grad_norm": 33.38825988769531, "learning_rate": 8.243742739465626e-06, - "loss": 0.2791, + "loss": 0.913, "step": 55660 }, { "epoch": 5.88, - "grad_norm": 1.0134646892547607, + "grad_norm": 34.9565544128418, "learning_rate": 8.241630584010983e-06, - "loss": 0.3044, + "loss": 0.9125, "step": 55670 }, { "epoch": 5.88, - "grad_norm": 3.929783344268799, + "grad_norm": 12.296431541442871, "learning_rate": 8.239518428556342e-06, - "loss": 0.5742, + "loss": 0.9852, "step": 55680 }, { "epoch": 5.88, - "grad_norm": 10.189471244812012, + "grad_norm": 28.598562240600586, "learning_rate": 8.237406273101701e-06, - "loss": 0.5661, + "loss": 1.2808, "step": 55690 }, { "epoch": 5.88, - "grad_norm": 0.11490225046873093, + "grad_norm": 22.555644989013672, "learning_rate": 8.23529411764706e-06, - "loss": 0.1928, + "loss": 0.8464, "step": 55700 }, { "epoch": 5.88, - "grad_norm": 18.710718154907227, + "grad_norm": 20.93575668334961, "learning_rate": 8.233181962192419e-06, - "loss": 0.4229, + "loss": 0.7154, "step": 55710 }, { "epoch": 5.88, - "grad_norm": 16.989900588989258, + "grad_norm": 40.53439712524414, "learning_rate": 8.231069806737777e-06, - "loss": 0.4383, + "loss": 0.899, "step": 55720 }, { "epoch": 5.89, - "grad_norm": 0.17938628792762756, + "grad_norm": 50.51793670654297, "learning_rate": 8.228957651283134e-06, - "loss": 0.3514, + "loss": 1.0724, "step": 55730 }, { "epoch": 5.89, - "grad_norm": 11.119317054748535, + "grad_norm": 32.62100601196289, "learning_rate": 8.226845495828493e-06, - "loss": 0.7822, + "loss": 1.3532, "step": 55740 }, { "epoch": 5.89, - "grad_norm": 350.5458679199219, + "grad_norm": 45.371856689453125, "learning_rate": 8.224733340373852e-06, - "loss": 0.7701, + "loss": 1.2372, "step": 55750 }, { "epoch": 5.89, - "grad_norm": 20.172691345214844, + "grad_norm": 48.2555046081543, "learning_rate": 8.22262118491921e-06, - "loss": 0.5416, + "loss": 0.9491, "step": 55760 }, { "epoch": 5.89, - "grad_norm": 17.307313919067383, + "grad_norm": 53.07205581665039, "learning_rate": 8.22050902946457e-06, - "loss": 0.4515, + "loss": 0.9037, "step": 55770 }, { "epoch": 5.89, - "grad_norm": 3.237438678741455, + "grad_norm": 50.730411529541016, "learning_rate": 8.218396874009927e-06, - "loss": 0.2567, + "loss": 1.0664, "step": 55780 }, { "epoch": 5.89, - "grad_norm": 18.246976852416992, + "grad_norm": 50.00428771972656, "learning_rate": 8.216284718555287e-06, - "loss": 0.4548, + "loss": 1.0487, "step": 55790 }, { "epoch": 5.89, - "grad_norm": 18.928743362426758, + "grad_norm": 39.86876678466797, "learning_rate": 8.214172563100646e-06, - "loss": 0.4251, + "loss": 0.9617, "step": 55800 }, { "epoch": 5.89, - "grad_norm": 9.14599895477295, + "grad_norm": 35.40494918823242, "learning_rate": 8.212060407646003e-06, - "loss": 0.335, + "loss": 0.9504, "step": 55810 }, { "epoch": 5.9, - "grad_norm": 7.359029293060303, + "grad_norm": 38.263084411621094, "learning_rate": 8.209948252191362e-06, - "loss": 0.2114, + "loss": 0.9983, "step": 55820 }, { "epoch": 5.9, - "grad_norm": 25.394765853881836, + "grad_norm": 44.12120819091797, "learning_rate": 8.20783609673672e-06, - "loss": 0.7029, + "loss": 1.0815, "step": 55830 }, { "epoch": 5.9, - "grad_norm": 9.701255798339844, + "grad_norm": 41.394893646240234, "learning_rate": 8.20572394128208e-06, - "loss": 0.6365, + "loss": 1.1905, "step": 55840 }, { "epoch": 5.9, - "grad_norm": 23.276979446411133, + "grad_norm": 53.56425857543945, "learning_rate": 8.203611785827438e-06, - "loss": 0.4451, + "loss": 0.8484, "step": 55850 }, { "epoch": 5.9, - "grad_norm": 33.067413330078125, + "grad_norm": 40.95310974121094, "learning_rate": 8.201499630372795e-06, - "loss": 0.5934, + "loss": 1.1176, "step": 55860 }, { "epoch": 5.9, - "grad_norm": 14.642045021057129, + "grad_norm": 21.445409774780273, "learning_rate": 8.199387474918154e-06, - "loss": 0.2004, + "loss": 0.8039, "step": 55870 }, { "epoch": 5.9, - "grad_norm": 9.362894058227539, + "grad_norm": 28.205184936523438, "learning_rate": 8.197275319463513e-06, - "loss": 0.6245, + "loss": 1.034, "step": 55880 }, { "epoch": 5.9, - "grad_norm": 7.272494316101074, + "grad_norm": 43.243812561035156, "learning_rate": 8.195163164008872e-06, - "loss": 0.7017, + "loss": 0.9051, "step": 55890 }, { "epoch": 5.9, - "grad_norm": 10.833017349243164, + "grad_norm": 26.414140701293945, "learning_rate": 8.19305100855423e-06, - "loss": 0.3259, + "loss": 0.8419, "step": 55900 }, { "epoch": 5.9, - "grad_norm": 26.672462463378906, + "grad_norm": 52.0042724609375, "learning_rate": 8.190938853099589e-06, - "loss": 0.6199, + "loss": 0.9967, "step": 55910 }, { "epoch": 5.91, - "grad_norm": 6.0708327293396, + "grad_norm": 23.249935150146484, "learning_rate": 8.188826697644948e-06, - "loss": 0.5986, + "loss": 1.1869, "step": 55920 }, { "epoch": 5.91, - "grad_norm": 29.427364349365234, + "grad_norm": 40.01015853881836, "learning_rate": 8.186714542190307e-06, - "loss": 0.5489, + "loss": 0.8025, "step": 55930 }, { "epoch": 5.91, - "grad_norm": 11.459290504455566, + "grad_norm": 45.36445999145508, "learning_rate": 8.184602386735664e-06, - "loss": 0.4551, + "loss": 1.1744, "step": 55940 }, { "epoch": 5.91, - "grad_norm": 0.8486009836196899, + "grad_norm": 23.47413444519043, "learning_rate": 8.182490231281022e-06, - "loss": 0.51, + "loss": 1.0863, "step": 55950 }, { "epoch": 5.91, - "grad_norm": 24.88314437866211, + "grad_norm": 19.733692169189453, "learning_rate": 8.180378075826381e-06, - "loss": 0.476, + "loss": 0.8923, "step": 55960 }, { "epoch": 5.91, - "grad_norm": 32.037227630615234, + "grad_norm": 40.245967864990234, "learning_rate": 8.17826592037174e-06, - "loss": 0.6199, + "loss": 0.8633, "step": 55970 }, { "epoch": 5.91, - "grad_norm": 13.674631118774414, + "grad_norm": 18.605899810791016, "learning_rate": 8.176153764917099e-06, - "loss": 0.4502, + "loss": 1.0518, "step": 55980 }, { "epoch": 5.91, - "grad_norm": 3.690596103668213, + "grad_norm": 53.36069107055664, "learning_rate": 8.174041609462458e-06, - "loss": 0.5262, + "loss": 1.1905, "step": 55990 }, { "epoch": 5.91, - "grad_norm": 10.037485122680664, + "grad_norm": 63.68470001220703, "learning_rate": 8.171929454007815e-06, - "loss": 0.4896, + "loss": 1.0755, "step": 56000 }, { "epoch": 5.92, - "grad_norm": 3.098954200744629, + "grad_norm": 57.16289520263672, "learning_rate": 8.169817298553173e-06, - "loss": 0.3954, + "loss": 1.0959, "step": 56010 }, { "epoch": 5.92, - "grad_norm": 17.060697555541992, + "grad_norm": 48.205726623535156, "learning_rate": 8.167705143098534e-06, - "loss": 0.3927, + "loss": 0.9556, "step": 56020 }, { "epoch": 5.92, - "grad_norm": 0.2854969799518585, + "grad_norm": 13.824310302734375, "learning_rate": 8.165592987643891e-06, - "loss": 0.5009, + "loss": 1.2986, "step": 56030 }, { "epoch": 5.92, - "grad_norm": 16.2278995513916, + "grad_norm": 31.276254653930664, "learning_rate": 8.16348083218925e-06, - "loss": 0.5481, + "loss": 0.7779, "step": 56040 }, { "epoch": 5.92, - "grad_norm": 5.585930347442627, + "grad_norm": 49.3869514465332, "learning_rate": 8.161368676734609e-06, - "loss": 0.463, + "loss": 1.0089, "step": 56050 }, { "epoch": 5.92, - "grad_norm": 10.718828201293945, + "grad_norm": 35.705081939697266, "learning_rate": 8.159256521279967e-06, - "loss": 0.2445, + "loss": 0.8357, "step": 56060 }, { "epoch": 5.92, - "grad_norm": 23.473960876464844, + "grad_norm": 21.242469787597656, "learning_rate": 8.157144365825326e-06, - "loss": 0.6492, + "loss": 0.906, "step": 56070 }, { "epoch": 5.92, - "grad_norm": 12.88558292388916, + "grad_norm": 45.743656158447266, "learning_rate": 8.155032210370683e-06, - "loss": 0.3928, + "loss": 1.1826, "step": 56080 }, { "epoch": 5.92, - "grad_norm": 27.861404418945312, + "grad_norm": 43.08173370361328, "learning_rate": 8.152920054916042e-06, - "loss": 0.5152, + "loss": 0.9838, "step": 56090 }, { "epoch": 5.92, - "grad_norm": 7.349437713623047, + "grad_norm": 33.945377349853516, "learning_rate": 8.1508078994614e-06, - "loss": 0.3429, + "loss": 0.9689, "step": 56100 }, { "epoch": 5.93, - "grad_norm": 29.837093353271484, + "grad_norm": 58.69560241699219, "learning_rate": 8.14869574400676e-06, - "loss": 0.5202, + "loss": 1.1994, "step": 56110 }, { "epoch": 5.93, - "grad_norm": 20.10384178161621, + "grad_norm": 67.54344940185547, "learning_rate": 8.146583588552118e-06, - "loss": 0.9633, + "loss": 1.2947, "step": 56120 }, { "epoch": 5.93, - "grad_norm": 38.52759552001953, + "grad_norm": 35.09498596191406, "learning_rate": 8.144471433097475e-06, - "loss": 0.6876, + "loss": 1.1206, "step": 56130 }, { "epoch": 5.93, - "grad_norm": 13.397809982299805, + "grad_norm": 52.95780563354492, "learning_rate": 8.142359277642836e-06, - "loss": 0.7715, + "loss": 1.0283, "step": 56140 }, { "epoch": 5.93, - "grad_norm": 17.592323303222656, + "grad_norm": 45.5756950378418, "learning_rate": 8.140247122188195e-06, - "loss": 0.4516, + "loss": 1.1968, "step": 56150 }, { "epoch": 5.93, - "grad_norm": 0.11249080300331116, + "grad_norm": 41.483802795410156, "learning_rate": 8.138134966733552e-06, - "loss": 0.6164, + "loss": 1.1393, "step": 56160 }, { "epoch": 5.93, - "grad_norm": 30.023834228515625, + "grad_norm": 44.96038055419922, "learning_rate": 8.13602281127891e-06, - "loss": 0.573, + "loss": 0.9402, "step": 56170 }, { "epoch": 5.93, - "grad_norm": 23.957687377929688, + "grad_norm": 20.611595153808594, "learning_rate": 8.13391065582427e-06, - "loss": 0.3682, + "loss": 0.9777, "step": 56180 }, { "epoch": 5.93, - "grad_norm": 0.5578147768974304, + "grad_norm": 60.19355392456055, "learning_rate": 8.131798500369628e-06, - "loss": 0.5641, + "loss": 1.4781, "step": 56190 }, { "epoch": 5.94, - "grad_norm": 34.41382598876953, + "grad_norm": 45.10508346557617, "learning_rate": 8.129686344914987e-06, - "loss": 0.7505, + "loss": 1.3747, "step": 56200 }, { "epoch": 5.94, - "grad_norm": 0.14313514530658722, + "grad_norm": 5.305068016052246, "learning_rate": 8.127574189460346e-06, - "loss": 0.176, + "loss": 0.6221, "step": 56210 }, { "epoch": 5.94, - "grad_norm": 69.84308624267578, + "grad_norm": 25.78734016418457, "learning_rate": 8.125462034005703e-06, - "loss": 0.6812, + "loss": 1.2811, "step": 56220 }, { "epoch": 5.94, - "grad_norm": 10.629273414611816, + "grad_norm": 55.33103561401367, "learning_rate": 8.123349878551062e-06, - "loss": 0.3476, + "loss": 0.9492, "step": 56230 }, { "epoch": 5.94, - "grad_norm": 17.78660774230957, + "grad_norm": 34.30466842651367, "learning_rate": 8.12123772309642e-06, - "loss": 0.2893, + "loss": 0.8603, "step": 56240 }, { "epoch": 5.94, - "grad_norm": 34.344478607177734, + "grad_norm": 48.17898941040039, "learning_rate": 8.119125567641779e-06, - "loss": 0.5586, + "loss": 1.0425, "step": 56250 }, { "epoch": 5.94, - "grad_norm": 25.362821578979492, + "grad_norm": 49.14372253417969, "learning_rate": 8.117013412187138e-06, - "loss": 0.2674, + "loss": 0.9415, "step": 56260 }, { "epoch": 5.94, - "grad_norm": 19.57725715637207, + "grad_norm": 39.837921142578125, "learning_rate": 8.114901256732497e-06, - "loss": 0.6193, + "loss": 0.8849, "step": 56270 }, { "epoch": 5.94, - "grad_norm": 0.11237912625074387, + "grad_norm": 44.28290939331055, "learning_rate": 8.112789101277855e-06, - "loss": 0.4356, + "loss": 0.7043, "step": 56280 }, { "epoch": 5.94, - "grad_norm": 26.295869827270508, + "grad_norm": 50.59253692626953, "learning_rate": 8.110676945823214e-06, - "loss": 0.7508, + "loss": 1.0233, "step": 56290 }, { "epoch": 5.95, - "grad_norm": 3.105318784713745, + "grad_norm": 55.864036560058594, "learning_rate": 8.108564790368571e-06, - "loss": 0.421, + "loss": 0.8821, "step": 56300 }, { "epoch": 5.95, - "grad_norm": 0.6656286120414734, + "grad_norm": 24.033435821533203, "learning_rate": 8.10645263491393e-06, - "loss": 0.551, + "loss": 0.9849, "step": 56310 }, { "epoch": 5.95, - "grad_norm": 1.6746275424957275, + "grad_norm": 18.7882137298584, "learning_rate": 8.104340479459289e-06, - "loss": 0.3501, + "loss": 0.8426, "step": 56320 }, { "epoch": 5.95, - "grad_norm": 17.46854591369629, + "grad_norm": 29.962217330932617, "learning_rate": 8.102228324004648e-06, - "loss": 0.434, + "loss": 0.8911, "step": 56330 }, { "epoch": 5.95, - "grad_norm": 6.760842323303223, + "grad_norm": 52.252235412597656, "learning_rate": 8.100116168550006e-06, - "loss": 0.6008, + "loss": 1.4387, "step": 56340 }, { "epoch": 5.95, - "grad_norm": 1.2065945863723755, + "grad_norm": 35.226478576660156, "learning_rate": 8.098004013095363e-06, - "loss": 0.1491, + "loss": 1.0851, "step": 56350 }, { "epoch": 5.95, - "grad_norm": 35.045562744140625, + "grad_norm": 33.148773193359375, "learning_rate": 8.095891857640722e-06, - "loss": 0.6244, + "loss": 1.0438, "step": 56360 }, { "epoch": 5.95, - "grad_norm": 0.03403489664196968, + "grad_norm": 57.342063903808594, "learning_rate": 8.093779702186083e-06, - "loss": 0.4435, + "loss": 1.3326, "step": 56370 }, { "epoch": 5.95, - "grad_norm": 6.179675102233887, + "grad_norm": 30.837400436401367, "learning_rate": 8.09166754673144e-06, - "loss": 0.5179, + "loss": 0.8546, "step": 56380 }, { "epoch": 5.96, - "grad_norm": 31.189453125, + "grad_norm": 51.32255935668945, "learning_rate": 8.089555391276799e-06, - "loss": 0.4365, + "loss": 1.1622, "step": 56390 }, { "epoch": 5.96, - "grad_norm": 8.671122550964355, + "grad_norm": 27.861629486083984, "learning_rate": 8.087443235822157e-06, - "loss": 0.2583, + "loss": 0.8216, "step": 56400 }, { "epoch": 5.96, - "grad_norm": 42.89104080200195, + "grad_norm": 36.6127815246582, "learning_rate": 8.085331080367516e-06, - "loss": 0.5589, + "loss": 1.2138, "step": 56410 }, { "epoch": 5.96, - "grad_norm": 9.457782745361328, + "grad_norm": 50.108680725097656, "learning_rate": 8.083218924912875e-06, - "loss": 0.4714, + "loss": 1.1055, "step": 56420 }, { "epoch": 5.96, - "grad_norm": 15.598109245300293, + "grad_norm": 51.83686065673828, "learning_rate": 8.081106769458232e-06, - "loss": 0.5163, + "loss": 1.2543, "step": 56430 }, { "epoch": 5.96, - "grad_norm": 9.889619827270508, + "grad_norm": 20.02341079711914, "learning_rate": 8.07899461400359e-06, - "loss": 0.5638, + "loss": 0.993, "step": 56440 }, { "epoch": 5.96, - "grad_norm": 21.617834091186523, + "grad_norm": 27.517925262451172, "learning_rate": 8.07688245854895e-06, - "loss": 0.4943, + "loss": 1.2111, "step": 56450 }, { "epoch": 5.96, - "grad_norm": 11.942456245422363, + "grad_norm": 38.842674255371094, "learning_rate": 8.074770303094308e-06, - "loss": 0.1598, + "loss": 0.9311, "step": 56460 }, { "epoch": 5.96, - "grad_norm": 18.24910545349121, + "grad_norm": 56.68899917602539, "learning_rate": 8.072658147639667e-06, - "loss": 0.523, + "loss": 1.0255, "step": 56470 }, { "epoch": 5.96, - "grad_norm": 6.825323104858398, + "grad_norm": 38.13239669799805, "learning_rate": 8.070545992185026e-06, - "loss": 0.1828, + "loss": 1.1078, "step": 56480 }, { "epoch": 5.97, - "grad_norm": 0.37498801946640015, + "grad_norm": 3.8762118816375732, "learning_rate": 8.068433836730383e-06, - "loss": 0.6266, + "loss": 1.2139, "step": 56490 }, { "epoch": 5.97, - "grad_norm": 1.7789205312728882, + "grad_norm": 25.302091598510742, "learning_rate": 8.066321681275743e-06, - "loss": 0.509, + "loss": 1.2607, "step": 56500 }, { "epoch": 5.97, - "grad_norm": 10.538804054260254, + "grad_norm": 40.41101837158203, "learning_rate": 8.0642095258211e-06, - "loss": 0.7441, + "loss": 1.0066, "step": 56510 }, { "epoch": 5.97, - "grad_norm": 17.026447296142578, + "grad_norm": 28.245817184448242, "learning_rate": 8.06209737036646e-06, - "loss": 0.5812, + "loss": 0.9439, "step": 56520 }, { "epoch": 5.97, - "grad_norm": 17.405899047851562, + "grad_norm": 50.59400177001953, "learning_rate": 8.059985214911818e-06, - "loss": 0.355, + "loss": 1.1541, "step": 56530 }, { "epoch": 5.97, - "grad_norm": 0.9259071946144104, + "grad_norm": 18.746477127075195, "learning_rate": 8.057873059457177e-06, - "loss": 0.4164, + "loss": 0.9329, "step": 56540 }, { "epoch": 5.97, - "grad_norm": 14.447752952575684, + "grad_norm": 63.13563537597656, "learning_rate": 8.055760904002536e-06, - "loss": 0.71, + "loss": 1.2408, "step": 56550 }, { "epoch": 5.97, - "grad_norm": 26.51258659362793, + "grad_norm": 40.946537017822266, "learning_rate": 8.053648748547894e-06, - "loss": 0.3258, + "loss": 0.7499, "step": 56560 }, { "epoch": 5.97, - "grad_norm": 0.3825536370277405, + "grad_norm": 44.806419372558594, "learning_rate": 8.051536593093252e-06, - "loss": 0.5224, + "loss": 1.1304, "step": 56570 }, { "epoch": 5.98, - "grad_norm": 6.9266438484191895, + "grad_norm": 32.90978240966797, "learning_rate": 8.04942443763861e-06, - "loss": 0.4315, + "loss": 1.0583, "step": 56580 }, { "epoch": 5.98, - "grad_norm": 24.058988571166992, + "grad_norm": 32.005706787109375, "learning_rate": 8.047312282183969e-06, - "loss": 0.3627, + "loss": 0.921, "step": 56590 }, { "epoch": 5.98, - "grad_norm": 63.57395553588867, + "grad_norm": 54.38248062133789, "learning_rate": 8.045200126729328e-06, - "loss": 0.668, + "loss": 1.1787, "step": 56600 }, { "epoch": 5.98, - "grad_norm": 12.862578392028809, + "grad_norm": 30.565670013427734, "learning_rate": 8.043087971274687e-06, - "loss": 0.4322, + "loss": 0.8996, "step": 56610 }, { "epoch": 5.98, - "grad_norm": 1.8223471641540527, + "grad_norm": 37.498226165771484, "learning_rate": 8.040975815820045e-06, - "loss": 0.3459, + "loss": 0.6965, "step": 56620 }, { "epoch": 5.98, - "grad_norm": 10.508016586303711, + "grad_norm": 37.89647674560547, "learning_rate": 8.038863660365404e-06, - "loss": 0.5564, + "loss": 0.99, "step": 56630 }, { "epoch": 5.98, - "grad_norm": 0.409253865480423, + "grad_norm": 38.05378341674805, "learning_rate": 8.036751504910763e-06, - "loss": 0.631, + "loss": 1.2458, "step": 56640 }, { "epoch": 5.98, - "grad_norm": 15.686080932617188, + "grad_norm": 28.047197341918945, "learning_rate": 8.03463934945612e-06, - "loss": 0.3706, + "loss": 1.3749, "step": 56650 }, { "epoch": 5.98, - "grad_norm": 21.205270767211914, + "grad_norm": 21.00182342529297, "learning_rate": 8.032527194001479e-06, - "loss": 0.4477, + "loss": 0.9175, "step": 56660 }, { "epoch": 5.98, - "grad_norm": 7.290127277374268, + "grad_norm": 40.37843322753906, "learning_rate": 8.030415038546838e-06, - "loss": 0.6047, + "loss": 1.6091, "step": 56670 }, { "epoch": 5.99, - "grad_norm": 5.62230920791626, + "grad_norm": 42.3004264831543, "learning_rate": 8.028302883092196e-06, - "loss": 0.4857, + "loss": 1.2749, "step": 56680 }, { "epoch": 5.99, - "grad_norm": 29.584413528442383, + "grad_norm": 43.0697135925293, "learning_rate": 8.026190727637555e-06, - "loss": 0.509, + "loss": 0.6824, "step": 56690 }, { "epoch": 5.99, - "grad_norm": 5.697409629821777, + "grad_norm": 37.88706970214844, "learning_rate": 8.024078572182912e-06, - "loss": 0.2758, + "loss": 0.9687, "step": 56700 }, { "epoch": 5.99, - "grad_norm": 12.542513847351074, + "grad_norm": 28.18208122253418, "learning_rate": 8.021966416728271e-06, - "loss": 0.5674, + "loss": 1.0744, "step": 56710 }, { "epoch": 5.99, - "grad_norm": 15.410750389099121, + "grad_norm": 21.728063583374023, "learning_rate": 8.019854261273632e-06, - "loss": 0.3708, + "loss": 1.2104, "step": 56720 }, { "epoch": 5.99, - "grad_norm": 8.361873626708984, + "grad_norm": 41.34721755981445, "learning_rate": 8.017742105818989e-06, - "loss": 0.2118, + "loss": 0.8916, "step": 56730 }, { "epoch": 5.99, - "grad_norm": 19.148345947265625, + "grad_norm": 37.51137161254883, "learning_rate": 8.015629950364347e-06, - "loss": 0.5189, + "loss": 1.0846, "step": 56740 }, { "epoch": 5.99, - "grad_norm": 7.016278266906738, + "grad_norm": 41.121402740478516, "learning_rate": 8.013517794909706e-06, - "loss": 0.2404, + "loss": 0.8789, "step": 56750 }, { "epoch": 5.99, - "grad_norm": 12.198592185974121, + "grad_norm": 37.934574127197266, "learning_rate": 8.011405639455065e-06, - "loss": 0.2741, + "loss": 0.7953, "step": 56760 }, { "epoch": 6.0, - "grad_norm": 0.8069953322410583, + "grad_norm": 34.445289611816406, "learning_rate": 8.009293484000424e-06, - "loss": 0.4265, + "loss": 1.026, "step": 56770 }, { "epoch": 6.0, - "grad_norm": 32.77610778808594, + "grad_norm": 17.21687126159668, "learning_rate": 8.00718132854578e-06, - "loss": 0.4528, + "loss": 1.067, "step": 56780 }, { "epoch": 6.0, - "grad_norm": 9.278461456298828, + "grad_norm": 41.901092529296875, "learning_rate": 8.00506917309114e-06, - "loss": 0.3932, + "loss": 0.9117, "step": 56790 }, { "epoch": 6.0, - "grad_norm": 1.511436939239502, + "grad_norm": 42.810638427734375, "learning_rate": 8.002957017636498e-06, - "loss": 0.3695, + "loss": 0.7559, "step": 56800 }, { "epoch": 6.0, - "grad_norm": 15.290145874023438, + "grad_norm": 21.046178817749023, "learning_rate": 8.000844862181857e-06, - "loss": 0.7247, + "loss": 0.9975, "step": 56810 }, { "epoch": 6.0, - "eval_accuracy": 0.9007920792079208, - "eval_loss": 0.3903925120830536, - "eval_runtime": 181.2978, - "eval_samples_per_second": 139.274, - "eval_steps_per_second": 17.413, + "eval_accuracy": 0.7998019801980198, + "eval_loss": 0.7377786040306091, + "eval_runtime": 37.7177, + "eval_samples_per_second": 669.446, + "eval_steps_per_second": 83.701, "step": 56814 }, { "epoch": 6.0, - "grad_norm": 1.1198546886444092, + "grad_norm": 58.958370208740234, "learning_rate": 7.998732706727216e-06, - "loss": 0.517, + "loss": 1.2169, "step": 56820 }, { "epoch": 6.0, - "grad_norm": 16.286409378051758, + "grad_norm": 72.40776062011719, "learning_rate": 7.996620551272575e-06, - "loss": 0.336, + "loss": 1.0673, "step": 56830 }, { "epoch": 6.0, - "grad_norm": 6.583364963531494, + "grad_norm": 34.879730224609375, "learning_rate": 7.994508395817932e-06, - "loss": 0.3185, + "loss": 1.0655, "step": 56840 }, { "epoch": 6.0, - "grad_norm": 5.492551803588867, + "grad_norm": 39.0068359375, "learning_rate": 7.992396240363292e-06, - "loss": 0.5785, + "loss": 0.8959, "step": 56850 }, { "epoch": 6.0, - "grad_norm": 0.05169447138905525, + "grad_norm": 28.461143493652344, "learning_rate": 7.990284084908651e-06, - "loss": 0.4897, + "loss": 0.8911, "step": 56860 }, { "epoch": 6.01, - "grad_norm": 0.20837251842021942, + "grad_norm": 15.706762313842773, "learning_rate": 7.988171929454008e-06, - "loss": 0.4464, + "loss": 0.9221, "step": 56870 }, { "epoch": 6.01, - "grad_norm": 0.9712957143783569, + "grad_norm": 10.471026420593262, "learning_rate": 7.986059773999367e-06, - "loss": 0.4591, + "loss": 0.8955, "step": 56880 }, { "epoch": 6.01, - "grad_norm": 13.39068603515625, + "grad_norm": 41.20438003540039, "learning_rate": 7.983947618544726e-06, - "loss": 0.3433, + "loss": 0.9785, "step": 56890 }, { "epoch": 6.01, - "grad_norm": 22.04598045349121, + "grad_norm": 17.931232452392578, "learning_rate": 7.981835463090084e-06, - "loss": 0.4399, + "loss": 1.0085, "step": 56900 }, { "epoch": 6.01, - "grad_norm": 15.936150550842285, + "grad_norm": 56.2242317199707, "learning_rate": 7.979723307635443e-06, - "loss": 0.2085, + "loss": 0.7735, "step": 56910 }, { "epoch": 6.01, - "grad_norm": 12.144168853759766, + "grad_norm": 42.893253326416016, "learning_rate": 7.9776111521808e-06, - "loss": 0.3729, + "loss": 1.0123, "step": 56920 }, { "epoch": 6.01, - "grad_norm": 0.4029267132282257, + "grad_norm": 22.33249855041504, "learning_rate": 7.975498996726159e-06, - "loss": 0.3928, + "loss": 0.7313, "step": 56930 }, { "epoch": 6.01, - "grad_norm": 15.664204597473145, + "grad_norm": 40.31864547729492, "learning_rate": 7.97338684127152e-06, - "loss": 0.8953, + "loss": 1.0212, "step": 56940 }, { "epoch": 6.01, - "grad_norm": 8.748826026916504, + "grad_norm": 41.19463348388672, "learning_rate": 7.971274685816877e-06, - "loss": 0.3745, + "loss": 1.2127, "step": 56950 }, { "epoch": 6.02, - "grad_norm": 9.493971824645996, + "grad_norm": 36.510066986083984, "learning_rate": 7.969162530362235e-06, - "loss": 0.3947, + "loss": 1.11, "step": 56960 }, { "epoch": 6.02, - "grad_norm": 21.032854080200195, + "grad_norm": 37.67352294921875, "learning_rate": 7.967050374907593e-06, - "loss": 0.3195, + "loss": 1.0775, "step": 56970 }, { "epoch": 6.02, - "grad_norm": 1.173124074935913, + "grad_norm": 44.982975006103516, "learning_rate": 7.964938219452953e-06, - "loss": 0.2254, + "loss": 0.8736, "step": 56980 }, { "epoch": 6.02, - "grad_norm": 13.292925834655762, + "grad_norm": 27.77263832092285, "learning_rate": 7.962826063998312e-06, - "loss": 0.2924, + "loss": 0.9153, "step": 56990 }, { "epoch": 6.02, - "grad_norm": 17.11899185180664, + "grad_norm": 43.818355560302734, "learning_rate": 7.960713908543669e-06, - "loss": 0.3979, + "loss": 0.9211, "step": 57000 }, { "epoch": 6.02, - "grad_norm": 14.948124885559082, + "grad_norm": 22.447582244873047, "learning_rate": 7.958601753089028e-06, - "loss": 0.431, + "loss": 0.966, "step": 57010 }, { "epoch": 6.02, - "grad_norm": 6.597762107849121, + "grad_norm": 22.36911964416504, "learning_rate": 7.956489597634386e-06, - "loss": 0.1269, + "loss": 0.6814, "step": 57020 }, { "epoch": 6.02, - "grad_norm": 2.0976898670196533, + "grad_norm": 37.2284049987793, "learning_rate": 7.954377442179745e-06, - "loss": 0.1487, + "loss": 0.9648, "step": 57030 }, { "epoch": 6.02, - "grad_norm": 10.220635414123535, + "grad_norm": 45.296085357666016, "learning_rate": 7.952265286725104e-06, - "loss": 0.4439, + "loss": 0.7607, "step": 57040 }, { "epoch": 6.02, - "grad_norm": 10.900344848632812, + "grad_norm": 44.5556755065918, "learning_rate": 7.950153131270461e-06, - "loss": 0.3983, + "loss": 0.7813, "step": 57050 }, { "epoch": 6.03, - "grad_norm": 15.922964096069336, + "grad_norm": 32.38617706298828, "learning_rate": 7.94804097581582e-06, - "loss": 0.2156, + "loss": 0.6388, "step": 57060 }, { "epoch": 6.03, - "grad_norm": 0.39834272861480713, + "grad_norm": 24.397624969482422, "learning_rate": 7.94592882036118e-06, - "loss": 0.544, + "loss": 1.0326, "step": 57070 }, { "epoch": 6.03, - "grad_norm": 11.715853691101074, + "grad_norm": 10.713223457336426, "learning_rate": 7.943816664906537e-06, - "loss": 0.2576, + "loss": 0.5148, "step": 57080 }, { "epoch": 6.03, - "grad_norm": 11.954784393310547, + "grad_norm": 32.87857437133789, "learning_rate": 7.941704509451896e-06, - "loss": 0.211, + "loss": 0.8014, "step": 57090 }, { "epoch": 6.03, - "grad_norm": 2.713653087615967, + "grad_norm": 49.4992790222168, "learning_rate": 7.939592353997255e-06, - "loss": 0.4189, + "loss": 1.2483, "step": 57100 }, { "epoch": 6.03, - "grad_norm": 6.550325393676758, + "grad_norm": 40.68593215942383, "learning_rate": 7.937480198542614e-06, - "loss": 0.3952, + "loss": 0.8037, "step": 57110 }, { "epoch": 6.03, - "grad_norm": 6.19587516784668, + "grad_norm": 53.5941276550293, "learning_rate": 7.935368043087973e-06, - "loss": 0.3121, + "loss": 0.9201, "step": 57120 }, { "epoch": 6.03, - "grad_norm": 0.618001401424408, + "grad_norm": 46.75265121459961, "learning_rate": 7.933255887633331e-06, - "loss": 0.4925, + "loss": 1.039, "step": 57130 }, { "epoch": 6.03, - "grad_norm": 2.2990517616271973, + "grad_norm": 32.48169708251953, "learning_rate": 7.931143732178688e-06, - "loss": 0.3067, + "loss": 0.847, "step": 57140 }, { "epoch": 6.04, - "grad_norm": 3.9507248401641846, + "grad_norm": 19.96778678894043, "learning_rate": 7.929031576724047e-06, - "loss": 0.2176, + "loss": 0.8849, "step": 57150 }, { "epoch": 6.04, - "grad_norm": 11.96513843536377, + "grad_norm": 22.393062591552734, "learning_rate": 7.926919421269406e-06, - "loss": 0.4137, + "loss": 0.962, "step": 57160 }, { "epoch": 6.04, - "grad_norm": 22.41522789001465, + "grad_norm": 27.09273338317871, "learning_rate": 7.924807265814765e-06, - "loss": 0.5178, + "loss": 0.7017, "step": 57170 }, { "epoch": 6.04, - "grad_norm": 14.599998474121094, + "grad_norm": 46.00996780395508, "learning_rate": 7.922695110360123e-06, - "loss": 0.4575, + "loss": 1.0398, "step": 57180 }, { "epoch": 6.04, - "grad_norm": 21.77976417541504, + "grad_norm": 36.91779327392578, "learning_rate": 7.92058295490548e-06, - "loss": 0.5709, + "loss": 1.0276, "step": 57190 }, { "epoch": 6.04, - "grad_norm": 5.694126605987549, + "grad_norm": 44.85866165161133, "learning_rate": 7.918470799450841e-06, - "loss": 0.3953, + "loss": 0.9418, "step": 57200 }, { "epoch": 6.04, - "grad_norm": 0.018146337941288948, + "grad_norm": 45.399147033691406, "learning_rate": 7.9163586439962e-06, - "loss": 0.3923, + "loss": 0.9623, "step": 57210 }, { "epoch": 6.04, - "grad_norm": 4.22137451171875, + "grad_norm": 47.53273391723633, "learning_rate": 7.914246488541557e-06, - "loss": 0.4616, + "loss": 1.033, "step": 57220 }, { "epoch": 6.04, - "grad_norm": 29.438823699951172, + "grad_norm": 27.890243530273438, "learning_rate": 7.912134333086916e-06, - "loss": 0.5932, + "loss": 1.0422, "step": 57230 }, { "epoch": 6.04, - "grad_norm": 3.7468698024749756, + "grad_norm": 42.49101638793945, "learning_rate": 7.910022177632274e-06, - "loss": 0.3754, + "loss": 1.1035, "step": 57240 }, { "epoch": 6.05, - "grad_norm": 5.014320373535156, + "grad_norm": 33.11439514160156, "learning_rate": 7.907910022177633e-06, - "loss": 0.3454, + "loss": 0.9122, "step": 57250 }, { "epoch": 6.05, - "grad_norm": 0.4096521735191345, + "grad_norm": 51.00248718261719, "learning_rate": 7.905797866722992e-06, - "loss": 0.3287, + "loss": 0.9037, "step": 57260 }, { "epoch": 6.05, - "grad_norm": 34.82781219482422, + "grad_norm": 32.63103485107422, "learning_rate": 7.903685711268349e-06, - "loss": 0.4462, + "loss": 0.8006, "step": 57270 }, { "epoch": 6.05, - "grad_norm": 0.1646999716758728, + "grad_norm": 26.231172561645508, "learning_rate": 7.901573555813708e-06, - "loss": 0.2567, + "loss": 0.4862, "step": 57280 }, { "epoch": 6.05, - "grad_norm": 11.163738250732422, + "grad_norm": 49.575172424316406, "learning_rate": 7.899461400359067e-06, - "loss": 0.2183, + "loss": 0.9622, "step": 57290 }, { "epoch": 6.05, - "grad_norm": 7.8915114402771, + "grad_norm": 51.65349578857422, "learning_rate": 7.897349244904425e-06, - "loss": 0.2838, + "loss": 0.9178, "step": 57300 }, { "epoch": 6.05, - "grad_norm": 10.017045974731445, + "grad_norm": 42.50986099243164, "learning_rate": 7.895237089449784e-06, - "loss": 0.5337, + "loss": 0.9457, "step": 57310 }, { "epoch": 6.05, - "grad_norm": 3.624067544937134, + "grad_norm": 33.00217819213867, "learning_rate": 7.893124933995143e-06, - "loss": 0.5097, + "loss": 0.862, "step": 57320 }, { "epoch": 6.05, - "grad_norm": 15.920127868652344, + "grad_norm": 11.735758781433105, "learning_rate": 7.891012778540502e-06, - "loss": 0.3455, + "loss": 0.8935, "step": 57330 }, { "epoch": 6.06, - "grad_norm": 0.9184063076972961, + "grad_norm": 21.519729614257812, "learning_rate": 7.88890062308586e-06, - "loss": 0.5247, + "loss": 1.0147, "step": 57340 }, { "epoch": 6.06, - "grad_norm": 20.701946258544922, + "grad_norm": 19.345848083496094, "learning_rate": 7.886788467631218e-06, - "loss": 0.5945, + "loss": 0.9554, "step": 57350 }, { "epoch": 6.06, - "grad_norm": 36.464256286621094, + "grad_norm": 38.65700149536133, "learning_rate": 7.884676312176576e-06, - "loss": 0.3404, + "loss": 1.056, "step": 57360 }, { "epoch": 6.06, - "grad_norm": 2.045686721801758, + "grad_norm": 42.7396354675293, "learning_rate": 7.882564156721935e-06, - "loss": 0.4363, + "loss": 0.9627, "step": 57370 }, { "epoch": 6.06, - "grad_norm": 14.562143325805664, + "grad_norm": 39.1786994934082, "learning_rate": 7.880452001267294e-06, - "loss": 0.5025, + "loss": 0.9351, "step": 57380 }, { "epoch": 6.06, - "grad_norm": 25.11194610595703, + "grad_norm": 43.31999206542969, "learning_rate": 7.878339845812653e-06, - "loss": 0.2438, + "loss": 0.9254, "step": 57390 }, { "epoch": 6.06, - "grad_norm": 0.7821240425109863, + "grad_norm": 41.98588562011719, "learning_rate": 7.876227690358012e-06, - "loss": 0.3059, + "loss": 0.8358, "step": 57400 }, { "epoch": 6.06, - "grad_norm": 40.33612823486328, + "grad_norm": 39.74501419067383, "learning_rate": 7.874115534903369e-06, - "loss": 0.5279, + "loss": 0.8987, "step": 57410 }, { "epoch": 6.06, - "grad_norm": 2.820688247680664, + "grad_norm": 37.42196273803711, "learning_rate": 7.872003379448729e-06, - "loss": 0.3252, + "loss": 0.9918, "step": 57420 }, { "epoch": 6.07, - "grad_norm": 23.9632511138916, + "grad_norm": 25.486644744873047, "learning_rate": 7.869891223994086e-06, - "loss": 0.4733, + "loss": 0.9361, "step": 57430 }, { "epoch": 6.07, - "grad_norm": 1.198825716972351, + "grad_norm": 36.2104377746582, "learning_rate": 7.867779068539445e-06, - "loss": 0.1798, + "loss": 1.0974, "step": 57440 }, { "epoch": 6.07, - "grad_norm": 21.0228328704834, + "grad_norm": 48.1801872253418, "learning_rate": 7.865666913084804e-06, - "loss": 0.5457, + "loss": 0.8399, "step": 57450 }, { "epoch": 6.07, - "grad_norm": 1.4918615818023682, + "grad_norm": 33.62314987182617, "learning_rate": 7.863554757630163e-06, - "loss": 0.4725, + "loss": 0.8364, "step": 57460 }, { "epoch": 6.07, - "grad_norm": 10.560433387756348, + "grad_norm": 43.42533874511719, "learning_rate": 7.861442602175521e-06, - "loss": 0.0922, + "loss": 0.9711, "step": 57470 }, { "epoch": 6.07, - "grad_norm": 3.5665884017944336, + "grad_norm": 14.273967742919922, "learning_rate": 7.85933044672088e-06, - "loss": 0.2825, + "loss": 0.9917, "step": 57480 }, { "epoch": 6.07, - "grad_norm": 13.14256477355957, + "grad_norm": 42.678466796875, "learning_rate": 7.857218291266237e-06, - "loss": 0.387, + "loss": 0.9035, "step": 57490 }, { "epoch": 6.07, - "grad_norm": 8.184035301208496, + "grad_norm": 27.524465560913086, "learning_rate": 7.855106135811596e-06, - "loss": 0.5687, + "loss": 1.0044, "step": 57500 }, { "epoch": 6.07, - "grad_norm": 21.86392593383789, + "grad_norm": 33.62763214111328, "learning_rate": 7.852993980356955e-06, - "loss": 0.2589, + "loss": 0.6746, "step": 57510 }, { "epoch": 6.07, - "grad_norm": 34.4300537109375, + "grad_norm": 27.06725311279297, "learning_rate": 7.850881824902314e-06, - "loss": 0.517, + "loss": 1.1107, "step": 57520 }, { "epoch": 6.08, - "grad_norm": 13.00790786743164, + "grad_norm": 56.14871597290039, "learning_rate": 7.848769669447672e-06, - "loss": 0.3043, + "loss": 0.6564, "step": 57530 }, { "epoch": 6.08, - "grad_norm": 0.07460898160934448, + "grad_norm": 42.431114196777344, "learning_rate": 7.84665751399303e-06, - "loss": 0.3876, + "loss": 1.4646, "step": 57540 }, { "epoch": 6.08, - "grad_norm": 26.28559684753418, + "grad_norm": 45.28477096557617, "learning_rate": 7.84454535853839e-06, - "loss": 0.2162, + "loss": 0.8116, "step": 57550 }, { "epoch": 6.08, - "grad_norm": 0.3064970374107361, + "grad_norm": 39.12770462036133, "learning_rate": 7.842433203083749e-06, - "loss": 0.63, + "loss": 1.2236, "step": 57560 }, { "epoch": 6.08, - "grad_norm": 7.501888275146484, + "grad_norm": 56.64331817626953, "learning_rate": 7.840321047629106e-06, - "loss": 0.4221, + "loss": 1.028, "step": 57570 }, { "epoch": 6.08, - "grad_norm": 23.306798934936523, + "grad_norm": 39.87008285522461, "learning_rate": 7.838208892174464e-06, - "loss": 0.3341, + "loss": 1.005, "step": 57580 }, { "epoch": 6.08, - "grad_norm": 13.503120422363281, + "grad_norm": 20.12757682800293, "learning_rate": 7.836096736719823e-06, - "loss": 0.5305, + "loss": 0.7458, "step": 57590 }, { "epoch": 6.08, - "grad_norm": 11.453436851501465, + "grad_norm": 23.345626831054688, "learning_rate": 7.833984581265182e-06, - "loss": 0.2411, + "loss": 0.9714, "step": 57600 }, { "epoch": 6.08, - "grad_norm": 0.515518844127655, + "grad_norm": 41.054534912109375, "learning_rate": 7.83187242581054e-06, - "loss": 0.5614, + "loss": 0.7921, "step": 57610 }, { "epoch": 6.09, - "grad_norm": 9.959982872009277, + "grad_norm": 54.52687454223633, "learning_rate": 7.829760270355898e-06, - "loss": 0.5282, + "loss": 0.9224, "step": 57620 }, { "epoch": 6.09, - "grad_norm": 86.73982238769531, + "grad_norm": 43.8128776550293, "learning_rate": 7.827648114901257e-06, - "loss": 0.3906, + "loss": 1.0909, "step": 57630 }, { "epoch": 6.09, - "grad_norm": 20.407073974609375, + "grad_norm": 28.675880432128906, "learning_rate": 7.825535959446615e-06, - "loss": 0.6451, + "loss": 0.8692, "step": 57640 }, { "epoch": 6.09, - "grad_norm": 0.5605393648147583, + "grad_norm": 44.48482894897461, "learning_rate": 7.823423803991974e-06, - "loss": 0.3139, + "loss": 1.0924, "step": 57650 }, { "epoch": 6.09, - "grad_norm": 20.764493942260742, + "grad_norm": 49.111637115478516, "learning_rate": 7.821311648537333e-06, - "loss": 0.5872, + "loss": 0.6539, "step": 57660 }, { "epoch": 6.09, - "grad_norm": 32.3210563659668, + "grad_norm": 42.92708969116211, "learning_rate": 7.819199493082692e-06, - "loss": 0.4815, + "loss": 1.2211, "step": 57670 }, { "epoch": 6.09, - "grad_norm": 14.625517845153809, + "grad_norm": 27.323408126831055, "learning_rate": 7.81708733762805e-06, - "loss": 0.4475, + "loss": 1.1687, "step": 57680 }, { "epoch": 6.09, - "grad_norm": 30.83466339111328, + "grad_norm": 60.8870735168457, "learning_rate": 7.81497518217341e-06, - "loss": 0.4152, + "loss": 1.076, "step": 57690 }, { "epoch": 6.09, - "grad_norm": 13.038772583007812, + "grad_norm": 48.902374267578125, "learning_rate": 7.812863026718766e-06, - "loss": 0.6193, + "loss": 1.0019, "step": 57700 }, { "epoch": 6.09, - "grad_norm": 9.411053657531738, + "grad_norm": 46.09954071044922, "learning_rate": 7.810750871264125e-06, - "loss": 0.4216, + "loss": 1.1114, "step": 57710 }, { "epoch": 6.1, - "grad_norm": 0.21566756069660187, + "grad_norm": 31.259469985961914, "learning_rate": 7.808638715809484e-06, - "loss": 0.648, + "loss": 1.2735, "step": 57720 }, { "epoch": 6.1, - "grad_norm": 12.575669288635254, + "grad_norm": 38.97397232055664, "learning_rate": 7.806526560354843e-06, - "loss": 0.5902, + "loss": 1.0957, "step": 57730 }, { "epoch": 6.1, - "grad_norm": 17.39189338684082, + "grad_norm": 46.506752014160156, "learning_rate": 7.804414404900202e-06, - "loss": 0.4008, + "loss": 0.969, "step": 57740 }, { "epoch": 6.1, - "grad_norm": 14.729365348815918, + "grad_norm": 41.85993957519531, "learning_rate": 7.80230224944556e-06, - "loss": 0.3743, + "loss": 1.0719, "step": 57750 }, { "epoch": 6.1, - "grad_norm": 0.4170897901058197, + "grad_norm": 40.68301773071289, "learning_rate": 7.800190093990917e-06, - "loss": 0.2394, + "loss": 0.8429, "step": 57760 }, { "epoch": 6.1, - "grad_norm": 25.646387100219727, + "grad_norm": 34.366851806640625, "learning_rate": 7.798077938536276e-06, - "loss": 0.4257, + "loss": 0.9116, "step": 57770 }, { "epoch": 6.1, - "grad_norm": 0.4381301701068878, + "grad_norm": 46.7507209777832, "learning_rate": 7.795965783081637e-06, - "loss": 0.303, + "loss": 0.9971, "step": 57780 }, { "epoch": 6.1, - "grad_norm": 5.674243927001953, + "grad_norm": 52.85020065307617, "learning_rate": 7.793853627626994e-06, - "loss": 0.5126, + "loss": 1.0746, "step": 57790 }, { "epoch": 6.1, - "grad_norm": 5.468992233276367, + "grad_norm": 32.66221237182617, "learning_rate": 7.791741472172353e-06, - "loss": 0.2778, + "loss": 0.7819, "step": 57800 }, { "epoch": 6.11, - "grad_norm": 15.843356132507324, + "grad_norm": 29.00688362121582, "learning_rate": 7.789629316717711e-06, - "loss": 0.3294, + "loss": 0.5973, "step": 57810 }, { "epoch": 6.11, - "grad_norm": 12.155083656311035, + "grad_norm": 60.58496856689453, "learning_rate": 7.78751716126307e-06, - "loss": 0.465, + "loss": 0.9239, "step": 57820 }, { "epoch": 6.11, - "grad_norm": 2.7805609703063965, + "grad_norm": 78.60726928710938, "learning_rate": 7.785405005808429e-06, - "loss": 0.3796, + "loss": 1.1377, "step": 57830 }, { "epoch": 6.11, - "grad_norm": 23.970762252807617, + "grad_norm": 35.80804443359375, "learning_rate": 7.783292850353786e-06, - "loss": 0.1995, + "loss": 0.8228, "step": 57840 }, { "epoch": 6.11, - "grad_norm": 16.459598541259766, + "grad_norm": 58.48945617675781, "learning_rate": 7.781180694899145e-06, - "loss": 0.5612, + "loss": 1.2966, "step": 57850 }, { "epoch": 6.11, - "grad_norm": 2.2057478427886963, + "grad_norm": 30.38673210144043, "learning_rate": 7.779068539444504e-06, - "loss": 0.4034, + "loss": 1.1233, "step": 57860 }, { "epoch": 6.11, - "grad_norm": 8.614937782287598, + "grad_norm": 29.692195892333984, "learning_rate": 7.776956383989862e-06, - "loss": 0.3145, + "loss": 0.9994, "step": 57870 }, { "epoch": 6.11, - "grad_norm": 12.445632934570312, + "grad_norm": 27.565797805786133, "learning_rate": 7.774844228535221e-06, - "loss": 0.3402, + "loss": 0.8543, "step": 57880 }, { "epoch": 6.11, - "grad_norm": 18.598480224609375, + "grad_norm": 41.01338195800781, "learning_rate": 7.772732073080578e-06, - "loss": 0.4871, + "loss": 0.8591, "step": 57890 }, { "epoch": 6.11, - "grad_norm": 0.6587048768997192, + "grad_norm": 68.407470703125, "learning_rate": 7.770619917625939e-06, - "loss": 0.2459, + "loss": 0.9841, "step": 57900 }, { "epoch": 6.12, - "grad_norm": 15.781198501586914, + "grad_norm": 48.284934997558594, "learning_rate": 7.768507762171297e-06, - "loss": 0.4639, + "loss": 1.0732, "step": 57910 }, { "epoch": 6.12, - "grad_norm": 4.506821155548096, + "grad_norm": 43.874488830566406, "learning_rate": 7.766395606716654e-06, - "loss": 0.3198, + "loss": 1.072, "step": 57920 }, { "epoch": 6.12, - "grad_norm": 14.154147148132324, + "grad_norm": 35.910213470458984, "learning_rate": 7.764283451262013e-06, - "loss": 0.6667, + "loss": 1.1974, "step": 57930 }, { "epoch": 6.12, - "grad_norm": 30.26700782775879, + "grad_norm": 48.310279846191406, "learning_rate": 7.762171295807372e-06, - "loss": 0.5103, + "loss": 0.8252, "step": 57940 }, { "epoch": 6.12, - "grad_norm": 38.97446823120117, + "grad_norm": 44.879329681396484, "learning_rate": 7.76005914035273e-06, - "loss": 0.5036, + "loss": 1.2657, "step": 57950 }, { "epoch": 6.12, - "grad_norm": 11.095870018005371, + "grad_norm": 54.993736267089844, "learning_rate": 7.75794698489809e-06, - "loss": 0.3162, + "loss": 0.9472, "step": 57960 }, { "epoch": 6.12, - "grad_norm": 5.2433695793151855, + "grad_norm": 44.5841064453125, "learning_rate": 7.755834829443447e-06, - "loss": 0.5787, + "loss": 1.0262, "step": 57970 }, { "epoch": 6.12, - "grad_norm": 2.267408847808838, + "grad_norm": 36.10626983642578, "learning_rate": 7.753722673988805e-06, - "loss": 0.5395, + "loss": 1.0082, "step": 57980 }, { "epoch": 6.12, - "grad_norm": 10.429933547973633, + "grad_norm": 34.621559143066406, "learning_rate": 7.751610518534164e-06, - "loss": 0.5598, + "loss": 0.7278, "step": 57990 }, { "epoch": 6.13, - "grad_norm": 17.267990112304688, + "grad_norm": 32.617774963378906, "learning_rate": 7.749498363079523e-06, - "loss": 0.3663, + "loss": 0.9355, "step": 58000 }, { "epoch": 6.13, - "grad_norm": 0.041797149926424026, + "grad_norm": 49.44614791870117, "learning_rate": 7.747386207624882e-06, - "loss": 0.3776, + "loss": 1.0856, "step": 58010 }, { "epoch": 6.13, - "grad_norm": 0.3215635418891907, + "grad_norm": 4.764122486114502, "learning_rate": 7.74527405217024e-06, - "loss": 0.3061, + "loss": 0.7905, "step": 58020 }, { "epoch": 6.13, - "grad_norm": 13.700084686279297, + "grad_norm": 52.625694274902344, "learning_rate": 7.7431618967156e-06, - "loss": 0.6597, + "loss": 0.9, "step": 58030 }, { "epoch": 6.13, - "grad_norm": 42.388404846191406, + "grad_norm": 39.12831497192383, "learning_rate": 7.741049741260958e-06, - "loss": 0.324, + "loss": 0.7422, "step": 58040 }, { "epoch": 6.13, - "grad_norm": 9.649919509887695, + "grad_norm": 46.81498336791992, "learning_rate": 7.738937585806317e-06, - "loss": 0.5939, + "loss": 0.8638, "step": 58050 }, { "epoch": 6.13, - "grad_norm": 0.04760260879993439, + "grad_norm": 28.080936431884766, "learning_rate": 7.736825430351674e-06, - "loss": 0.372, + "loss": 0.9619, "step": 58060 }, { "epoch": 6.13, - "grad_norm": 12.115511894226074, + "grad_norm": 34.62874984741211, "learning_rate": 7.734713274897033e-06, - "loss": 0.2776, + "loss": 0.8111, "step": 58070 }, { "epoch": 6.13, - "grad_norm": 19.54997444152832, + "grad_norm": 21.980146408081055, "learning_rate": 7.732601119442392e-06, - "loss": 0.4579, + "loss": 1.2623, "step": 58080 }, { "epoch": 6.13, - "grad_norm": 8.214990615844727, + "grad_norm": 33.06687545776367, "learning_rate": 7.73048896398775e-06, - "loss": 0.1997, + "loss": 0.8325, "step": 58090 }, { "epoch": 6.14, - "grad_norm": 19.932493209838867, + "grad_norm": 43.6834831237793, "learning_rate": 7.728376808533109e-06, - "loss": 0.3331, + "loss": 1.0424, "step": 58100 }, { "epoch": 6.14, - "grad_norm": 0.20620892941951752, + "grad_norm": 31.09166145324707, "learning_rate": 7.726264653078466e-06, - "loss": 0.3281, + "loss": 1.0643, "step": 58110 }, { "epoch": 6.14, - "grad_norm": 9.656377792358398, + "grad_norm": 40.440216064453125, "learning_rate": 7.724152497623825e-06, - "loss": 0.3328, + "loss": 1.0507, "step": 58120 }, { "epoch": 6.14, - "grad_norm": 9.567946434020996, + "grad_norm": 47.98322677612305, "learning_rate": 7.722040342169185e-06, - "loss": 0.6266, + "loss": 1.0492, "step": 58130 }, { "epoch": 6.14, - "grad_norm": 7.83302640914917, + "grad_norm": 46.80917739868164, "learning_rate": 7.719928186714543e-06, - "loss": 0.3848, + "loss": 0.9842, "step": 58140 }, { "epoch": 6.14, - "grad_norm": 0.17017318308353424, + "grad_norm": 22.88640594482422, "learning_rate": 7.717816031259901e-06, - "loss": 0.2183, + "loss": 0.7004, "step": 58150 }, { "epoch": 6.14, - "grad_norm": 15.87339973449707, + "grad_norm": 35.34497833251953, "learning_rate": 7.71570387580526e-06, - "loss": 0.2885, + "loss": 0.8338, "step": 58160 }, { "epoch": 6.14, - "grad_norm": 15.463385581970215, + "grad_norm": 43.69340515136719, "learning_rate": 7.713591720350619e-06, - "loss": 0.3906, + "loss": 1.2631, "step": 58170 }, { "epoch": 6.14, - "grad_norm": 21.375288009643555, + "grad_norm": 83.56950378417969, "learning_rate": 7.711479564895978e-06, - "loss": 0.4103, + "loss": 1.0246, "step": 58180 }, { "epoch": 6.15, - "grad_norm": 0.6708627343177795, + "grad_norm": 33.71882629394531, "learning_rate": 7.709367409441335e-06, - "loss": 0.5624, + "loss": 0.9993, "step": 58190 }, { "epoch": 6.15, - "grad_norm": 7.6862945556640625, + "grad_norm": 25.758602142333984, "learning_rate": 7.707255253986694e-06, - "loss": 0.564, + "loss": 1.0663, "step": 58200 }, { "epoch": 6.15, - "grad_norm": 0.6291094422340393, + "grad_norm": 33.544002532958984, "learning_rate": 7.705143098532052e-06, - "loss": 0.141, + "loss": 0.6667, "step": 58210 }, { "epoch": 6.15, - "grad_norm": 9.232040405273438, + "grad_norm": 28.126399993896484, "learning_rate": 7.703030943077411e-06, - "loss": 0.2492, + "loss": 0.6977, "step": 58220 }, { "epoch": 6.15, - "grad_norm": 0.6905235052108765, + "grad_norm": 35.4185676574707, "learning_rate": 7.70091878762277e-06, - "loss": 0.8015, + "loss": 1.2137, "step": 58230 }, { "epoch": 6.15, - "grad_norm": 21.610706329345703, + "grad_norm": 47.24159622192383, "learning_rate": 7.698806632168129e-06, - "loss": 0.3754, + "loss": 0.8217, "step": 58240 }, { "epoch": 6.15, - "grad_norm": 15.52079963684082, + "grad_norm": 60.02870178222656, "learning_rate": 7.696694476713487e-06, - "loss": 0.291, + "loss": 0.9366, "step": 58250 }, { "epoch": 6.15, - "grad_norm": 34.42279815673828, + "grad_norm": 53.27028274536133, "learning_rate": 7.694582321258846e-06, - "loss": 0.6601, + "loss": 1.0977, "step": 58260 }, { "epoch": 6.15, - "grad_norm": 20.364070892333984, + "grad_norm": 31.05242347717285, "learning_rate": 7.692470165804203e-06, - "loss": 0.2059, + "loss": 0.9207, "step": 58270 }, { "epoch": 6.15, - "grad_norm": 4.366422176361084, + "grad_norm": 29.89213752746582, "learning_rate": 7.690358010349562e-06, - "loss": 0.4146, + "loss": 1.0117, "step": 58280 }, { "epoch": 6.16, - "grad_norm": 13.179944038391113, + "grad_norm": 33.0440559387207, "learning_rate": 7.68824585489492e-06, - "loss": 0.2699, + "loss": 0.6152, "step": 58290 }, { "epoch": 6.16, - "grad_norm": 16.294706344604492, + "grad_norm": 53.2236328125, "learning_rate": 7.68613369944028e-06, - "loss": 0.3688, + "loss": 0.8079, "step": 58300 }, { "epoch": 6.16, - "grad_norm": 5.983283519744873, + "grad_norm": 32.92738342285156, "learning_rate": 7.684021543985638e-06, - "loss": 0.4342, + "loss": 0.729, "step": 58310 }, { "epoch": 6.16, - "grad_norm": 21.87650489807129, + "grad_norm": 46.100154876708984, "learning_rate": 7.681909388530997e-06, - "loss": 0.1539, + "loss": 0.6204, "step": 58320 }, { "epoch": 6.16, - "grad_norm": 1.3997877836227417, + "grad_norm": 35.834312438964844, "learning_rate": 7.679797233076354e-06, - "loss": 0.2265, + "loss": 0.9217, "step": 58330 }, { "epoch": 6.16, - "grad_norm": 3.152129888534546, + "grad_norm": 31.64261817932129, "learning_rate": 7.677685077621713e-06, - "loss": 0.579, + "loss": 0.9278, "step": 58340 }, { "epoch": 6.16, - "grad_norm": 12.596598625183105, + "grad_norm": 35.1219482421875, "learning_rate": 7.675572922167072e-06, - "loss": 0.4075, + "loss": 1.1966, "step": 58350 }, { "epoch": 6.16, - "grad_norm": 21.975894927978516, + "grad_norm": 55.63541793823242, "learning_rate": 7.67346076671243e-06, - "loss": 0.5057, + "loss": 1.019, "step": 58360 }, { "epoch": 6.16, - "grad_norm": 0.14122119545936584, + "grad_norm": 44.640296936035156, "learning_rate": 7.67134861125779e-06, - "loss": 0.4196, + "loss": 0.8989, "step": 58370 }, { "epoch": 6.17, - "grad_norm": 27.701313018798828, + "grad_norm": 47.03463363647461, "learning_rate": 7.669236455803148e-06, - "loss": 0.4345, + "loss": 1.0577, "step": 58380 }, { "epoch": 6.17, - "grad_norm": 0.5699990391731262, + "grad_norm": 44.137088775634766, "learning_rate": 7.667124300348507e-06, - "loss": 0.3291, + "loss": 0.9961, "step": 58390 }, { "epoch": 6.17, - "grad_norm": 8.5996675491333, + "grad_norm": 31.356969833374023, "learning_rate": 7.665012144893866e-06, - "loss": 0.3822, + "loss": 0.775, "step": 58400 }, { "epoch": 6.17, - "grad_norm": 15.322367668151855, + "grad_norm": 39.389095306396484, "learning_rate": 7.662899989439223e-06, - "loss": 0.2367, + "loss": 0.7205, "step": 58410 }, { "epoch": 6.17, - "grad_norm": 12.872920036315918, + "grad_norm": 48.71113967895508, "learning_rate": 7.660787833984582e-06, - "loss": 0.4007, + "loss": 0.6738, "step": 58420 }, { "epoch": 6.17, - "grad_norm": 5.186731815338135, + "grad_norm": 32.368202209472656, "learning_rate": 7.65867567852994e-06, - "loss": 0.2769, + "loss": 0.9236, "step": 58430 }, { "epoch": 6.17, - "grad_norm": 0.127484530210495, + "grad_norm": 19.3196964263916, "learning_rate": 7.656563523075299e-06, - "loss": 0.4408, + "loss": 0.876, "step": 58440 }, { "epoch": 6.17, - "grad_norm": 7.02890157699585, + "grad_norm": 41.42845153808594, "learning_rate": 7.654451367620658e-06, - "loss": 0.2207, + "loss": 0.9695, "step": 58450 }, { "epoch": 6.17, - "grad_norm": 11.931239128112793, + "grad_norm": 41.551883697509766, "learning_rate": 7.652339212166015e-06, - "loss": 0.2987, + "loss": 0.7837, "step": 58460 }, { "epoch": 6.17, - "grad_norm": 6.491023540496826, + "grad_norm": 23.503705978393555, "learning_rate": 7.650227056711374e-06, - "loss": 0.4787, + "loss": 0.7744, "step": 58470 }, { "epoch": 6.18, - "grad_norm": 41.49319839477539, + "grad_norm": 47.650665283203125, "learning_rate": 7.648114901256734e-06, - "loss": 0.4648, + "loss": 0.847, "step": 58480 }, { "epoch": 6.18, - "grad_norm": 10.304847717285156, + "grad_norm": 53.79130554199219, "learning_rate": 7.646002745802091e-06, - "loss": 0.3823, + "loss": 0.9956, "step": 58490 }, { "epoch": 6.18, - "grad_norm": 22.76209831237793, + "grad_norm": 61.487586975097656, "learning_rate": 7.64389059034745e-06, - "loss": 0.3083, + "loss": 1.0976, "step": 58500 }, { "epoch": 6.18, - "grad_norm": 0.6958699226379395, + "grad_norm": 43.75914764404297, "learning_rate": 7.641778434892809e-06, - "loss": 0.4121, + "loss": 1.0315, "step": 58510 }, { "epoch": 6.18, - "grad_norm": 7.041163444519043, + "grad_norm": 47.684452056884766, "learning_rate": 7.639666279438168e-06, - "loss": 0.4822, + "loss": 1.018, "step": 58520 }, { "epoch": 6.18, - "grad_norm": 7.931618690490723, + "grad_norm": 35.752227783203125, "learning_rate": 7.637554123983526e-06, - "loss": 0.5038, + "loss": 0.9142, "step": 58530 }, { "epoch": 6.18, - "grad_norm": 0.2422199547290802, + "grad_norm": 38.625694274902344, "learning_rate": 7.635441968528884e-06, - "loss": 0.2362, + "loss": 0.8452, "step": 58540 }, { "epoch": 6.18, - "grad_norm": 15.832688331604004, + "grad_norm": 59.42380905151367, "learning_rate": 7.633329813074242e-06, - "loss": 0.4436, + "loss": 0.979, "step": 58550 }, { "epoch": 6.18, - "grad_norm": 27.848773956298828, + "grad_norm": 71.72573852539062, "learning_rate": 7.631217657619601e-06, - "loss": 0.5549, + "loss": 1.089, "step": 58560 }, { "epoch": 6.19, - "grad_norm": 7.674691677093506, + "grad_norm": 37.56916427612305, "learning_rate": 7.62910550216496e-06, - "loss": 0.4236, + "loss": 1.1512, "step": 58570 }, { "epoch": 6.19, - "grad_norm": 7.233977317810059, + "grad_norm": 34.347679138183594, "learning_rate": 7.626993346710319e-06, - "loss": 0.5695, + "loss": 1.0792, "step": 58580 }, { "epoch": 6.19, - "grad_norm": 8.35018539428711, + "grad_norm": 27.422700881958008, "learning_rate": 7.6248811912556774e-06, - "loss": 0.3345, + "loss": 0.9764, "step": 58590 }, { "epoch": 6.19, - "grad_norm": 33.91869354248047, + "grad_norm": 53.45159912109375, "learning_rate": 7.622769035801035e-06, - "loss": 0.7093, + "loss": 1.2591, "step": 58600 }, { "epoch": 6.19, - "grad_norm": 12.89661693572998, + "grad_norm": 12.296812057495117, "learning_rate": 7.620656880346394e-06, - "loss": 0.4028, + "loss": 1.0852, "step": 58610 }, { "epoch": 6.19, - "grad_norm": 1.1616640090942383, + "grad_norm": 33.41921615600586, "learning_rate": 7.618544724891752e-06, - "loss": 0.5547, + "loss": 1.0994, "step": 58620 }, { "epoch": 6.19, - "grad_norm": 0.4819948971271515, + "grad_norm": 22.928627014160156, "learning_rate": 7.616432569437111e-06, - "loss": 0.4243, + "loss": 1.0377, "step": 58630 }, { "epoch": 6.19, - "grad_norm": 17.407529830932617, + "grad_norm": 38.9754524230957, "learning_rate": 7.61432041398247e-06, - "loss": 0.4098, + "loss": 0.9853, "step": 58640 }, { "epoch": 6.19, - "grad_norm": 22.837827682495117, + "grad_norm": 54.10380172729492, "learning_rate": 7.6122082585278276e-06, - "loss": 0.5027, + "loss": 0.8689, "step": 58650 }, { "epoch": 6.19, - "grad_norm": 22.370540618896484, + "grad_norm": 43.266971588134766, "learning_rate": 7.610096103073187e-06, - "loss": 0.4324, + "loss": 0.8329, "step": 58660 }, { "epoch": 6.2, - "grad_norm": 28.335325241088867, + "grad_norm": 75.12837982177734, "learning_rate": 7.607983947618546e-06, - "loss": 0.2829, + "loss": 1.1855, "step": 58670 }, { "epoch": 6.2, - "grad_norm": 4.894138813018799, + "grad_norm": 41.38825988769531, "learning_rate": 7.605871792163904e-06, - "loss": 0.3401, + "loss": 0.8035, "step": 58680 }, { "epoch": 6.2, - "grad_norm": 0.17998015880584717, + "grad_norm": 29.298095703125, "learning_rate": 7.603759636709263e-06, - "loss": 0.6669, + "loss": 1.0787, "step": 58690 }, { "epoch": 6.2, - "grad_norm": 17.85565757751465, + "grad_norm": 48.774539947509766, "learning_rate": 7.6016474812546215e-06, - "loss": 0.4649, + "loss": 1.1237, "step": 58700 }, { "epoch": 6.2, - "grad_norm": 18.006975173950195, + "grad_norm": 36.97369384765625, "learning_rate": 7.599535325799979e-06, - "loss": 0.381, + "loss": 0.8395, "step": 58710 }, { "epoch": 6.2, - "grad_norm": 2.586904764175415, + "grad_norm": 20.258411407470703, "learning_rate": 7.597423170345338e-06, - "loss": 0.494, + "loss": 1.0155, "step": 58720 }, { "epoch": 6.2, - "grad_norm": 0.659320592880249, + "grad_norm": 48.705936431884766, "learning_rate": 7.595311014890696e-06, - "loss": 0.4372, + "loss": 1.0148, "step": 58730 }, { "epoch": 6.2, - "grad_norm": 4.208647727966309, + "grad_norm": 48.02593231201172, "learning_rate": 7.593198859436055e-06, - "loss": 0.5066, + "loss": 1.1705, "step": 58740 }, { "epoch": 6.2, - "grad_norm": 3.823650360107422, + "grad_norm": 53.16486740112305, "learning_rate": 7.591086703981414e-06, - "loss": 0.5865, + "loss": 1.0361, "step": 58750 }, { "epoch": 6.21, - "grad_norm": 17.420528411865234, + "grad_norm": 54.01332092285156, "learning_rate": 7.588974548526772e-06, - "loss": 0.2959, + "loss": 1.0286, "step": 58760 }, { "epoch": 6.21, - "grad_norm": 16.523658752441406, + "grad_norm": 52.76885223388672, "learning_rate": 7.586862393072131e-06, - "loss": 0.2331, + "loss": 0.8906, "step": 58770 }, { "epoch": 6.21, - "grad_norm": 32.81853485107422, + "grad_norm": 43.917449951171875, "learning_rate": 7.58475023761749e-06, - "loss": 0.3765, + "loss": 0.7758, "step": 58780 }, { "epoch": 6.21, - "grad_norm": 0.8612244725227356, + "grad_norm": 37.40299606323242, "learning_rate": 7.582638082162848e-06, - "loss": 0.4973, + "loss": 1.0327, "step": 58790 }, { "epoch": 6.21, - "grad_norm": 5.797936916351318, + "grad_norm": 42.39024353027344, "learning_rate": 7.580525926708207e-06, - "loss": 0.2883, + "loss": 0.9654, "step": 58800 }, { "epoch": 6.21, - "grad_norm": 2.409339666366577, + "grad_norm": 44.746726989746094, "learning_rate": 7.578413771253565e-06, - "loss": 0.22, + "loss": 0.6805, "step": 58810 }, { "epoch": 6.21, - "grad_norm": 7.328845024108887, + "grad_norm": 57.04275894165039, "learning_rate": 7.576301615798923e-06, - "loss": 0.4614, + "loss": 1.2822, "step": 58820 }, { "epoch": 6.21, - "grad_norm": 21.835857391357422, + "grad_norm": 31.086849212646484, "learning_rate": 7.574189460344282e-06, - "loss": 0.341, + "loss": 1.0382, "step": 58830 }, { "epoch": 6.21, - "grad_norm": 5.233737468719482, + "grad_norm": 38.679264068603516, "learning_rate": 7.57207730488964e-06, - "loss": 0.2006, + "loss": 1.0072, "step": 58840 }, { "epoch": 6.22, - "grad_norm": 0.08171138912439346, + "grad_norm": 31.405546188354492, "learning_rate": 7.569965149434999e-06, - "loss": 0.1763, + "loss": 1.0433, "step": 58850 }, { "epoch": 6.22, - "grad_norm": 6.03022575378418, + "grad_norm": 53.219730377197266, "learning_rate": 7.567852993980358e-06, - "loss": 0.3541, + "loss": 0.8149, "step": 58860 }, { "epoch": 6.22, - "grad_norm": 5.134571552276611, + "grad_norm": 19.214426040649414, "learning_rate": 7.565740838525716e-06, - "loss": 0.2809, + "loss": 0.9294, "step": 58870 }, { "epoch": 6.22, - "grad_norm": 8.128372192382812, + "grad_norm": 40.643653869628906, "learning_rate": 7.563628683071074e-06, - "loss": 0.4791, + "loss": 0.8417, "step": 58880 }, { "epoch": 6.22, - "grad_norm": 17.109996795654297, + "grad_norm": 31.92180061340332, "learning_rate": 7.561516527616434e-06, - "loss": 0.419, + "loss": 0.8045, "step": 58890 }, { "epoch": 6.22, - "grad_norm": 20.70769500732422, + "grad_norm": 32.91853713989258, "learning_rate": 7.559404372161792e-06, - "loss": 0.4153, + "loss": 1.2096, "step": 58900 }, { "epoch": 6.22, - "grad_norm": 0.1108669564127922, + "grad_norm": 22.155323028564453, "learning_rate": 7.557292216707151e-06, - "loss": 0.465, + "loss": 1.0549, "step": 58910 }, { "epoch": 6.22, - "grad_norm": 11.025904655456543, + "grad_norm": 67.38444519042969, "learning_rate": 7.555180061252509e-06, - "loss": 0.3164, + "loss": 0.8165, "step": 58920 }, { "epoch": 6.22, - "grad_norm": 19.968244552612305, + "grad_norm": 50.472232818603516, "learning_rate": 7.5530679057978674e-06, - "loss": 0.4758, + "loss": 0.8749, "step": 58930 }, { "epoch": 6.22, - "grad_norm": 0.05653376877307892, + "grad_norm": 30.735057830810547, "learning_rate": 7.550955750343226e-06, - "loss": 0.3041, + "loss": 1.0955, "step": 58940 }, { "epoch": 6.23, - "grad_norm": 11.154923439025879, + "grad_norm": 37.93045425415039, "learning_rate": 7.548843594888584e-06, - "loss": 0.2418, + "loss": 0.8037, "step": 58950 }, { "epoch": 6.23, - "grad_norm": 0.07322105765342712, + "grad_norm": 36.93925094604492, "learning_rate": 7.546731439433943e-06, - "loss": 0.2709, + "loss": 0.915, "step": 58960 }, { "epoch": 6.23, - "grad_norm": 21.329822540283203, + "grad_norm": 62.679466247558594, "learning_rate": 7.544619283979302e-06, - "loss": 0.3821, + "loss": 1.3616, "step": 58970 }, { "epoch": 6.23, - "grad_norm": 1.729499340057373, + "grad_norm": 42.423789978027344, "learning_rate": 7.54250712852466e-06, - "loss": 0.2102, + "loss": 0.9329, "step": 58980 }, { "epoch": 6.23, - "grad_norm": 5.866237640380859, + "grad_norm": 40.519412994384766, "learning_rate": 7.540394973070018e-06, - "loss": 0.3411, + "loss": 0.9906, "step": 58990 }, { "epoch": 6.23, - "grad_norm": 22.544139862060547, + "grad_norm": 34.26726150512695, "learning_rate": 7.538282817615376e-06, - "loss": 0.2606, + "loss": 0.7394, "step": 59000 }, { "epoch": 6.23, - "grad_norm": 31.28618049621582, + "grad_norm": 50.46338653564453, "learning_rate": 7.536170662160736e-06, - "loss": 0.5469, + "loss": 1.3046, "step": 59010 }, { "epoch": 6.23, - "grad_norm": 11.084274291992188, + "grad_norm": 40.93741989135742, "learning_rate": 7.534058506706095e-06, - "loss": 0.3498, + "loss": 1.0158, "step": 59020 }, { "epoch": 6.23, - "grad_norm": 23.308374404907227, + "grad_norm": 44.08562469482422, "learning_rate": 7.531946351251453e-06, - "loss": 0.5261, + "loss": 1.0429, "step": 59030 }, { "epoch": 6.24, - "grad_norm": 5.336085796356201, + "grad_norm": 29.663768768310547, "learning_rate": 7.5298341957968115e-06, - "loss": 0.3268, + "loss": 0.8698, "step": 59040 }, { "epoch": 6.24, - "grad_norm": 14.303459167480469, + "grad_norm": 26.61013412475586, "learning_rate": 7.52772204034217e-06, - "loss": 0.3846, + "loss": 1.0215, "step": 59050 }, { "epoch": 6.24, - "grad_norm": 20.938722610473633, + "grad_norm": 59.20584487915039, "learning_rate": 7.525609884887528e-06, - "loss": 0.5955, + "loss": 0.9915, "step": 59060 }, { "epoch": 6.24, - "grad_norm": 7.787276744842529, + "grad_norm": 30.40359878540039, "learning_rate": 7.523497729432887e-06, - "loss": 0.3745, + "loss": 0.8511, "step": 59070 }, { "epoch": 6.24, - "grad_norm": 13.263599395751953, + "grad_norm": 51.256431579589844, "learning_rate": 7.521385573978245e-06, - "loss": 0.4474, + "loss": 0.9573, "step": 59080 }, { "epoch": 6.24, - "grad_norm": 11.967280387878418, + "grad_norm": 36.64567565917969, "learning_rate": 7.519273418523604e-06, - "loss": 0.4109, + "loss": 0.7525, "step": 59090 }, { "epoch": 6.24, - "grad_norm": 10.078975677490234, + "grad_norm": 25.877729415893555, "learning_rate": 7.5171612630689624e-06, - "loss": 0.2846, + "loss": 0.6985, "step": 59100 }, { "epoch": 6.24, - "grad_norm": 16.65361976623535, + "grad_norm": 59.41950607299805, "learning_rate": 7.51504910761432e-06, - "loss": 0.5971, + "loss": 1.1381, "step": 59110 }, { "epoch": 6.24, - "grad_norm": 40.134910583496094, + "grad_norm": 31.98960304260254, "learning_rate": 7.512936952159679e-06, - "loss": 0.3658, + "loss": 0.7942, "step": 59120 }, { "epoch": 6.24, - "grad_norm": 3.9769537448883057, + "grad_norm": 19.894344329833984, "learning_rate": 7.510824796705039e-06, - "loss": 0.29, + "loss": 1.0884, "step": 59130 }, { "epoch": 6.25, - "grad_norm": 12.519350051879883, + "grad_norm": 32.616920471191406, "learning_rate": 7.508712641250397e-06, - "loss": 0.3402, + "loss": 1.0316, "step": 59140 }, { "epoch": 6.25, - "grad_norm": 16.811559677124023, + "grad_norm": 60.95001220703125, "learning_rate": 7.5066004857957555e-06, - "loss": 0.4752, + "loss": 0.8733, "step": 59150 }, { "epoch": 6.25, - "grad_norm": 22.07139015197754, + "grad_norm": 40.51070022583008, "learning_rate": 7.504488330341114e-06, - "loss": 0.2296, + "loss": 0.8641, "step": 59160 }, { "epoch": 6.25, - "grad_norm": 14.201269149780273, + "grad_norm": 36.63600540161133, "learning_rate": 7.502376174886472e-06, - "loss": 0.7405, + "loss": 1.1753, "step": 59170 }, { "epoch": 6.25, - "grad_norm": 1.3245900869369507, + "grad_norm": 17.974323272705078, "learning_rate": 7.500264019431831e-06, - "loss": 0.4258, + "loss": 1.0898, "step": 59180 }, { "epoch": 6.25, - "grad_norm": 2.5379691123962402, + "grad_norm": 32.54548645019531, "learning_rate": 7.498151863977189e-06, - "loss": 0.5779, + "loss": 1.1556, "step": 59190 }, { "epoch": 6.25, - "grad_norm": 14.390296936035156, + "grad_norm": 55.33428955078125, "learning_rate": 7.496039708522548e-06, - "loss": 0.3488, + "loss": 0.9335, "step": 59200 }, { "epoch": 6.25, - "grad_norm": 27.798891067504883, + "grad_norm": 35.82052993774414, "learning_rate": 7.4939275530679065e-06, - "loss": 0.4324, + "loss": 0.8393, "step": 59210 }, { "epoch": 6.25, - "grad_norm": 7.201465129852295, + "grad_norm": 15.260730743408203, "learning_rate": 7.491815397613264e-06, - "loss": 0.3141, + "loss": 0.8116, "step": 59220 }, { "epoch": 6.26, - "grad_norm": 13.626611709594727, + "grad_norm": 60.020416259765625, "learning_rate": 7.489703242158623e-06, - "loss": 0.6263, + "loss": 1.1793, "step": 59230 }, { "epoch": 6.26, - "grad_norm": 23.739877700805664, + "grad_norm": 40.985530853271484, "learning_rate": 7.487591086703983e-06, - "loss": 0.3037, + "loss": 0.911, "step": 59240 }, { "epoch": 6.26, - "grad_norm": 6.585206031799316, + "grad_norm": 35.700225830078125, "learning_rate": 7.485478931249341e-06, - "loss": 0.4291, + "loss": 1.1503, "step": 59250 }, { "epoch": 6.26, - "grad_norm": 3.307797431945801, + "grad_norm": 30.963909149169922, "learning_rate": 7.4833667757946995e-06, - "loss": 0.3511, + "loss": 1.0488, "step": 59260 }, { "epoch": 6.26, - "grad_norm": 28.457942962646484, + "grad_norm": 57.63664627075195, "learning_rate": 7.4812546203400574e-06, - "loss": 0.3054, + "loss": 1.0852, "step": 59270 }, { "epoch": 6.26, - "grad_norm": 11.116164207458496, + "grad_norm": 50.669944763183594, "learning_rate": 7.479142464885416e-06, - "loss": 0.3477, + "loss": 1.3291, "step": 59280 }, { "epoch": 6.26, - "grad_norm": 3.9269893169403076, + "grad_norm": 34.41289520263672, "learning_rate": 7.477030309430775e-06, - "loss": 0.3771, + "loss": 0.8205, "step": 59290 }, { "epoch": 6.26, - "grad_norm": 0.08143565058708191, + "grad_norm": 37.639976501464844, "learning_rate": 7.474918153976133e-06, - "loss": 0.1685, + "loss": 1.0933, "step": 59300 }, { "epoch": 6.26, - "grad_norm": 1.404262661933899, + "grad_norm": 36.18600845336914, "learning_rate": 7.472805998521492e-06, - "loss": 0.3496, + "loss": 0.5928, "step": 59310 }, { "epoch": 6.26, - "grad_norm": 39.33906173706055, + "grad_norm": 38.85681915283203, "learning_rate": 7.4706938430668505e-06, - "loss": 0.663, + "loss": 1.0639, "step": 59320 }, { "epoch": 6.27, - "grad_norm": 20.942529678344727, + "grad_norm": 47.909053802490234, "learning_rate": 7.4685816876122084e-06, - "loss": 0.4948, + "loss": 1.0101, "step": 59330 }, { "epoch": 6.27, - "grad_norm": 20.650779724121094, + "grad_norm": 51.23207092285156, "learning_rate": 7.466469532157567e-06, - "loss": 0.39, + "loss": 1.2488, "step": 59340 }, { "epoch": 6.27, - "grad_norm": 11.242096900939941, + "grad_norm": 26.030122756958008, "learning_rate": 7.464357376702927e-06, - "loss": 0.3049, + "loss": 0.8757, "step": 59350 }, { "epoch": 6.27, - "grad_norm": 17.447162628173828, + "grad_norm": 47.55458068847656, "learning_rate": 7.462245221248284e-06, - "loss": 0.4873, + "loss": 0.8573, "step": 59360 }, { "epoch": 6.27, - "grad_norm": 0.24511872231960297, + "grad_norm": 54.92838668823242, "learning_rate": 7.4601330657936435e-06, - "loss": 0.3707, + "loss": 0.764, "step": 59370 }, { "epoch": 6.27, - "grad_norm": 14.008968353271484, + "grad_norm": 24.923215866088867, "learning_rate": 7.4580209103390015e-06, - "loss": 0.4612, + "loss": 0.867, "step": 59380 }, { "epoch": 6.27, - "grad_norm": 2.0889334678649902, + "grad_norm": 24.425451278686523, "learning_rate": 7.45590875488436e-06, - "loss": 0.1272, + "loss": 0.7466, "step": 59390 }, { "epoch": 6.27, - "grad_norm": 33.40538787841797, + "grad_norm": 35.26579284667969, "learning_rate": 7.453796599429719e-06, - "loss": 0.6196, + "loss": 1.1159, "step": 59400 }, { "epoch": 6.27, - "grad_norm": 21.026363372802734, + "grad_norm": 47.77822494506836, "learning_rate": 7.451684443975077e-06, - "loss": 0.4653, + "loss": 1.099, "step": 59410 }, { "epoch": 6.28, - "grad_norm": 7.791131019592285, + "grad_norm": 32.47595977783203, "learning_rate": 7.449572288520436e-06, - "loss": 0.4813, + "loss": 0.8734, "step": 59420 }, { "epoch": 6.28, - "grad_norm": 6.001340389251709, + "grad_norm": 48.769989013671875, "learning_rate": 7.4474601330657945e-06, - "loss": 0.4419, + "loss": 0.9866, "step": 59430 }, { "epoch": 6.28, - "grad_norm": 1.9028376340866089, + "grad_norm": 52.29174041748047, "learning_rate": 7.4453479776111525e-06, - "loss": 0.1369, + "loss": 0.7786, "step": 59440 }, { "epoch": 6.28, - "grad_norm": 12.113272666931152, + "grad_norm": 30.55681037902832, "learning_rate": 7.443235822156511e-06, - "loss": 0.4283, + "loss": 1.1502, "step": 59450 }, { "epoch": 6.28, - "grad_norm": 9.421527862548828, + "grad_norm": 30.378992080688477, "learning_rate": 7.441123666701869e-06, - "loss": 0.3114, + "loss": 0.6764, "step": 59460 }, { "epoch": 6.28, - "grad_norm": 28.96178436279297, + "grad_norm": 38.65161895751953, "learning_rate": 7.439011511247228e-06, - "loss": 0.4839, + "loss": 0.8552, "step": 59470 }, { "epoch": 6.28, - "grad_norm": 4.103160381317139, + "grad_norm": 30.659944534301758, "learning_rate": 7.4368993557925876e-06, - "loss": 0.3929, + "loss": 0.9403, "step": 59480 }, { "epoch": 6.28, - "grad_norm": 0.10671932995319366, + "grad_norm": 32.05353927612305, "learning_rate": 7.4347872003379455e-06, - "loss": 0.4056, + "loss": 1.0149, "step": 59490 }, { "epoch": 6.28, - "grad_norm": 7.6984710693359375, + "grad_norm": 55.14085006713867, "learning_rate": 7.432675044883304e-06, - "loss": 0.3824, + "loss": 1.4117, "step": 59500 }, { "epoch": 6.28, - "grad_norm": 20.73331642150879, + "grad_norm": 68.6175765991211, "learning_rate": 7.430562889428663e-06, - "loss": 0.4175, + "loss": 1.1277, "step": 59510 }, { "epoch": 6.29, - "grad_norm": 17.25911521911621, + "grad_norm": 33.13150405883789, "learning_rate": 7.428450733974021e-06, - "loss": 0.2938, + "loss": 0.7345, "step": 59520 }, { "epoch": 6.29, - "grad_norm": 1.939328670501709, + "grad_norm": 32.41312789916992, "learning_rate": 7.42633857851938e-06, - "loss": 0.2027, + "loss": 0.6231, "step": 59530 }, { "epoch": 6.29, - "grad_norm": 0.9897793531417847, + "grad_norm": 27.475543975830078, "learning_rate": 7.424226423064738e-06, - "loss": 0.3554, + "loss": 0.9607, "step": 59540 }, { "epoch": 6.29, - "grad_norm": 12.171623229980469, + "grad_norm": 31.07724952697754, "learning_rate": 7.4221142676100965e-06, - "loss": 0.3851, + "loss": 0.9249, "step": 59550 }, { "epoch": 6.29, - "grad_norm": 12.282883644104004, + "grad_norm": 54.16040802001953, "learning_rate": 7.420002112155455e-06, - "loss": 0.6147, + "loss": 0.9921, "step": 59560 }, { "epoch": 6.29, - "grad_norm": 6.707889556884766, + "grad_norm": 11.75674057006836, "learning_rate": 7.417889956700813e-06, - "loss": 0.5491, + "loss": 0.9836, "step": 59570 }, { "epoch": 6.29, - "grad_norm": 1.4292542934417725, + "grad_norm": 73.63533782958984, "learning_rate": 7.415777801246172e-06, - "loss": 0.5019, + "loss": 1.2264, "step": 59580 }, { "epoch": 6.29, - "grad_norm": 5.003266334533691, + "grad_norm": 35.6644172668457, "learning_rate": 7.413665645791532e-06, - "loss": 0.4899, + "loss": 1.0381, "step": 59590 }, { "epoch": 6.29, - "grad_norm": 19.870702743530273, + "grad_norm": 47.36325454711914, "learning_rate": 7.411553490336889e-06, - "loss": 0.396, + "loss": 1.312, "step": 59600 }, { "epoch": 6.3, - "grad_norm": 9.707528114318848, + "grad_norm": 22.6315860748291, "learning_rate": 7.409441334882248e-06, - "loss": 0.2607, + "loss": 0.8908, "step": 59610 }, { "epoch": 6.3, - "grad_norm": 23.750808715820312, + "grad_norm": 45.7405891418457, "learning_rate": 7.407329179427607e-06, - "loss": 0.6565, + "loss": 1.0074, "step": 59620 }, { "epoch": 6.3, - "grad_norm": 8.938817024230957, + "grad_norm": 27.628236770629883, "learning_rate": 7.405217023972965e-06, - "loss": 0.6031, + "loss": 1.0359, "step": 59630 }, { "epoch": 6.3, - "grad_norm": 1.1852667331695557, + "grad_norm": 23.797927856445312, "learning_rate": 7.403104868518324e-06, - "loss": 0.3338, + "loss": 0.8918, "step": 59640 }, { "epoch": 6.3, - "grad_norm": 1.0824925899505615, + "grad_norm": 27.077598571777344, "learning_rate": 7.400992713063682e-06, - "loss": 0.455, + "loss": 1.2521, "step": 59650 }, { "epoch": 6.3, - "grad_norm": 4.885572910308838, + "grad_norm": 49.248260498046875, "learning_rate": 7.3988805576090405e-06, - "loss": 0.5127, + "loss": 0.8903, "step": 59660 }, { "epoch": 6.3, - "grad_norm": 4.240900993347168, + "grad_norm": 35.79132080078125, "learning_rate": 7.396768402154399e-06, - "loss": 0.3235, + "loss": 0.8689, "step": 59670 }, { "epoch": 6.3, - "grad_norm": 4.894435405731201, + "grad_norm": 43.12715530395508, "learning_rate": 7.394656246699757e-06, - "loss": 0.1324, + "loss": 0.8709, "step": 59680 }, { "epoch": 6.3, - "grad_norm": 3.3644373416900635, + "grad_norm": 51.21892547607422, "learning_rate": 7.392544091245116e-06, - "loss": 0.5244, + "loss": 1.1695, "step": 59690 }, { "epoch": 6.3, - "grad_norm": 20.1683406829834, + "grad_norm": 39.288761138916016, "learning_rate": 7.390431935790476e-06, - "loss": 0.4055, + "loss": 0.9835, "step": 59700 }, { "epoch": 6.31, - "grad_norm": 18.468175888061523, + "grad_norm": 47.121559143066406, "learning_rate": 7.388319780335833e-06, - "loss": 0.4085, + "loss": 0.7503, "step": 59710 }, { "epoch": 6.31, - "grad_norm": 16.23465919494629, + "grad_norm": 38.613285064697266, "learning_rate": 7.386207624881192e-06, - "loss": 0.4683, + "loss": 1.021, "step": 59720 }, { "epoch": 6.31, - "grad_norm": 15.544109344482422, + "grad_norm": 3.0946571826934814, "learning_rate": 7.38409546942655e-06, - "loss": 0.685, + "loss": 1.115, "step": 59730 }, { "epoch": 6.31, - "grad_norm": 18.84280014038086, + "grad_norm": 37.12638473510742, "learning_rate": 7.381983313971909e-06, - "loss": 0.3819, + "loss": 0.9009, "step": 59740 }, { "epoch": 6.31, - "grad_norm": 9.34996509552002, + "grad_norm": 36.88385009765625, "learning_rate": 7.379871158517268e-06, - "loss": 0.6514, + "loss": 0.9555, "step": 59750 }, { "epoch": 6.31, - "grad_norm": 16.59275245666504, + "grad_norm": 38.862667083740234, "learning_rate": 7.377759003062626e-06, - "loss": 0.3562, + "loss": 1.1419, "step": 59760 }, { "epoch": 6.31, - "grad_norm": 14.906723022460938, + "grad_norm": 35.20515060424805, "learning_rate": 7.3756468476079845e-06, - "loss": 0.4573, + "loss": 1.0459, "step": 59770 }, { "epoch": 6.31, - "grad_norm": 12.162001609802246, + "grad_norm": 40.05979919433594, "learning_rate": 7.373534692153343e-06, - "loss": 0.2814, + "loss": 1.0644, "step": 59780 }, { "epoch": 6.31, - "grad_norm": 31.656387329101562, + "grad_norm": 51.923641204833984, "learning_rate": 7.371422536698701e-06, - "loss": 0.2467, + "loss": 0.9782, "step": 59790 }, { "epoch": 6.32, - "grad_norm": 2.4770584106445312, + "grad_norm": 38.23823928833008, "learning_rate": 7.36931038124406e-06, - "loss": 0.3998, + "loss": 0.8381, "step": 59800 }, { "epoch": 6.32, - "grad_norm": 1.602397084236145, + "grad_norm": 36.232425689697266, "learning_rate": 7.36719822578942e-06, - "loss": 0.3786, + "loss": 0.9602, "step": 59810 }, { "epoch": 6.32, - "grad_norm": 14.620319366455078, + "grad_norm": 47.41489028930664, "learning_rate": 7.365086070334777e-06, - "loss": 0.4954, + "loss": 0.6955, "step": 59820 }, { "epoch": 6.32, - "grad_norm": 26.076173782348633, + "grad_norm": 52.320499420166016, "learning_rate": 7.362973914880136e-06, - "loss": 0.1154, + "loss": 0.924, "step": 59830 }, { "epoch": 6.32, - "grad_norm": 5.985979080200195, + "grad_norm": 61.50273895263672, "learning_rate": 7.3608617594254934e-06, - "loss": 0.3628, + "loss": 1.0607, "step": 59840 }, { "epoch": 6.32, - "grad_norm": 0.7753247618675232, + "grad_norm": 26.1118221282959, "learning_rate": 7.358749603970853e-06, - "loss": 0.2765, + "loss": 0.6055, "step": 59850 }, { "epoch": 6.32, - "grad_norm": 11.337801933288574, + "grad_norm": 31.209697723388672, "learning_rate": 7.356637448516212e-06, - "loss": 0.4914, + "loss": 0.9534, "step": 59860 }, { "epoch": 6.32, - "grad_norm": 10.445786476135254, + "grad_norm": 57.63028335571289, "learning_rate": 7.35452529306157e-06, - "loss": 0.4127, + "loss": 1.0811, "step": 59870 }, { "epoch": 6.32, - "grad_norm": 19.950565338134766, + "grad_norm": 33.659934997558594, "learning_rate": 7.3524131376069286e-06, - "loss": 0.4368, + "loss": 0.9947, "step": 59880 }, { "epoch": 6.32, - "grad_norm": 8.503373146057129, + "grad_norm": 42.9274787902832, "learning_rate": 7.350300982152287e-06, - "loss": 0.6399, + "loss": 1.184, "step": 59890 }, { "epoch": 6.33, - "grad_norm": 13.877084732055664, + "grad_norm": 30.223033905029297, "learning_rate": 7.348188826697645e-06, - "loss": 0.3661, + "loss": 0.8716, "step": 59900 }, { "epoch": 6.33, - "grad_norm": 22.415584564208984, + "grad_norm": 43.678672790527344, "learning_rate": 7.346076671243004e-06, - "loss": 0.4969, + "loss": 1.0416, "step": 59910 }, { "epoch": 6.33, - "grad_norm": 18.074508666992188, + "grad_norm": 48.95914077758789, "learning_rate": 7.343964515788362e-06, - "loss": 0.4285, + "loss": 0.9486, "step": 59920 }, { "epoch": 6.33, - "grad_norm": 7.419383525848389, + "grad_norm": 28.20637321472168, "learning_rate": 7.341852360333721e-06, - "loss": 0.498, + "loss": 0.9347, "step": 59930 }, { "epoch": 6.33, - "grad_norm": 18.77776336669922, + "grad_norm": 32.266754150390625, "learning_rate": 7.33974020487908e-06, - "loss": 0.3202, + "loss": 1.0911, "step": 59940 }, { "epoch": 6.33, - "grad_norm": 25.55537223815918, + "grad_norm": 33.64116668701172, "learning_rate": 7.3376280494244375e-06, - "loss": 0.5296, + "loss": 1.3066, "step": 59950 }, { "epoch": 6.33, - "grad_norm": 10.805418968200684, + "grad_norm": 39.43948745727539, "learning_rate": 7.335515893969797e-06, - "loss": 0.3581, + "loss": 1.1132, "step": 59960 }, { "epoch": 6.33, - "grad_norm": 11.1344575881958, + "grad_norm": 44.54194641113281, "learning_rate": 7.333403738515156e-06, - "loss": 0.2716, + "loss": 0.7508, "step": 59970 }, { "epoch": 6.33, - "grad_norm": 11.832480430603027, + "grad_norm": 49.63356018066406, "learning_rate": 7.331291583060514e-06, - "loss": 0.282, + "loss": 1.046, "step": 59980 }, { "epoch": 6.34, - "grad_norm": 2.230780601501465, + "grad_norm": 29.04340362548828, "learning_rate": 7.329179427605873e-06, - "loss": 0.4244, + "loss": 1.3193, "step": 59990 }, { "epoch": 6.34, - "grad_norm": 12.190003395080566, + "grad_norm": 35.17742156982422, "learning_rate": 7.327067272151231e-06, - "loss": 0.597, + "loss": 1.147, "step": 60000 }, { "epoch": 6.34, - "grad_norm": 3.2782232761383057, + "grad_norm": 52.15231704711914, "learning_rate": 7.324955116696589e-06, - "loss": 0.1345, + "loss": 1.0405, "step": 60010 }, { "epoch": 6.34, - "grad_norm": 6.976138114929199, + "grad_norm": 39.66686248779297, "learning_rate": 7.322842961241948e-06, - "loss": 0.4971, + "loss": 0.8049, "step": 60020 }, { "epoch": 6.34, - "grad_norm": 0.1625048816204071, + "grad_norm": 37.60445022583008, "learning_rate": 7.320730805787306e-06, - "loss": 0.29, + "loss": 0.729, "step": 60030 }, { "epoch": 6.34, - "grad_norm": 26.684507369995117, + "grad_norm": 24.305086135864258, "learning_rate": 7.318618650332665e-06, - "loss": 0.4553, + "loss": 0.658, "step": 60040 }, { "epoch": 6.34, - "grad_norm": 12.935179710388184, + "grad_norm": 26.69676399230957, "learning_rate": 7.316506494878024e-06, - "loss": 0.4799, + "loss": 0.735, "step": 60050 }, { "epoch": 6.34, - "grad_norm": 0.5345556735992432, + "grad_norm": 67.20134735107422, "learning_rate": 7.3143943394233815e-06, - "loss": 0.5801, + "loss": 1.1643, "step": 60060 }, { "epoch": 6.34, - "grad_norm": 6.840539932250977, + "grad_norm": 42.357398986816406, "learning_rate": 7.312282183968741e-06, - "loss": 0.3201, + "loss": 0.6728, "step": 60070 }, { "epoch": 6.34, - "grad_norm": 0.5232830047607422, + "grad_norm": 33.05667495727539, "learning_rate": 7.3101700285141e-06, - "loss": 0.5913, + "loss": 1.1501, "step": 60080 }, { "epoch": 6.35, - "grad_norm": 22.727649688720703, + "grad_norm": 46.61936950683594, "learning_rate": 7.308057873059458e-06, - "loss": 0.3689, + "loss": 1.0188, "step": 60090 }, { "epoch": 6.35, - "grad_norm": 24.94710922241211, + "grad_norm": 52.03192901611328, "learning_rate": 7.305945717604817e-06, - "loss": 0.4371, + "loss": 0.9118, "step": 60100 }, { "epoch": 6.35, - "grad_norm": 14.561208724975586, + "grad_norm": 43.186729431152344, "learning_rate": 7.3038335621501745e-06, - "loss": 0.5002, + "loss": 0.9971, "step": 60110 }, { "epoch": 6.35, - "grad_norm": 42.8281135559082, + "grad_norm": 48.9952507019043, "learning_rate": 7.301721406695533e-06, - "loss": 0.5868, + "loss": 0.9954, "step": 60120 }, { "epoch": 6.35, - "grad_norm": 7.820209503173828, + "grad_norm": 34.041160583496094, "learning_rate": 7.299609251240892e-06, - "loss": 0.5155, + "loss": 0.852, "step": 60130 }, { "epoch": 6.35, - "grad_norm": 5.990534782409668, + "grad_norm": 29.636920928955078, "learning_rate": 7.29749709578625e-06, - "loss": 0.5267, + "loss": 0.8243, "step": 60140 }, { "epoch": 6.35, - "grad_norm": 8.330293655395508, + "grad_norm": 48.39773178100586, "learning_rate": 7.295384940331609e-06, - "loss": 0.379, + "loss": 1.1395, "step": 60150 }, { "epoch": 6.35, - "grad_norm": 7.409394264221191, + "grad_norm": 29.20442771911621, "learning_rate": 7.293272784876968e-06, - "loss": 0.3372, + "loss": 1.0513, "step": 60160 }, { "epoch": 6.35, - "grad_norm": 2.3165853023529053, + "grad_norm": 26.10349464416504, "learning_rate": 7.2911606294223255e-06, - "loss": 0.3943, + "loss": 0.6888, "step": 60170 }, { "epoch": 6.36, - "grad_norm": 0.6842281818389893, + "grad_norm": 8.911745071411133, "learning_rate": 7.289048473967685e-06, - "loss": 0.2766, + "loss": 0.764, "step": 60180 }, { "epoch": 6.36, - "grad_norm": 15.519654273986816, + "grad_norm": 18.517791748046875, "learning_rate": 7.286936318513042e-06, - "loss": 0.5423, + "loss": 1.0092, "step": 60190 }, { "epoch": 6.36, - "grad_norm": 11.743553161621094, + "grad_norm": 59.641719818115234, "learning_rate": 7.284824163058402e-06, - "loss": 0.1673, + "loss": 0.7104, "step": 60200 }, { "epoch": 6.36, - "grad_norm": 0.8798866271972656, + "grad_norm": 70.06546020507812, "learning_rate": 7.282712007603761e-06, - "loss": 0.2426, + "loss": 0.958, "step": 60210 }, { "epoch": 6.36, - "grad_norm": 3.90859317779541, + "grad_norm": 45.784515380859375, "learning_rate": 7.2805998521491186e-06, - "loss": 0.4039, + "loss": 1.0733, "step": 60220 }, { "epoch": 6.36, - "grad_norm": 9.90040397644043, + "grad_norm": 49.79316329956055, "learning_rate": 7.278487696694477e-06, - "loss": 0.3732, + "loss": 1.1474, "step": 60230 }, { "epoch": 6.36, - "grad_norm": 20.376075744628906, + "grad_norm": 35.09388732910156, "learning_rate": 7.276375541239836e-06, - "loss": 0.3955, + "loss": 0.7457, "step": 60240 }, { "epoch": 6.36, - "grad_norm": 1.051681637763977, + "grad_norm": 39.82684326171875, "learning_rate": 7.274263385785194e-06, - "loss": 0.1743, + "loss": 0.7251, "step": 60250 }, { "epoch": 6.36, - "grad_norm": 19.535066604614258, + "grad_norm": 42.15519332885742, "learning_rate": 7.272151230330553e-06, - "loss": 0.5485, + "loss": 0.9188, "step": 60260 }, { "epoch": 6.36, - "grad_norm": 8.084059715270996, + "grad_norm": 10.536120414733887, "learning_rate": 7.270039074875912e-06, - "loss": 0.4188, + "loss": 0.8067, "step": 60270 }, { "epoch": 6.37, - "grad_norm": 6.707561492919922, + "grad_norm": 25.544841766357422, "learning_rate": 7.2679269194212695e-06, - "loss": 0.5265, + "loss": 1.0233, "step": 60280 }, { "epoch": 6.37, - "grad_norm": 13.115289688110352, + "grad_norm": 60.12702178955078, "learning_rate": 7.265814763966629e-06, - "loss": 0.35, + "loss": 1.1156, "step": 60290 }, { "epoch": 6.37, - "grad_norm": 3.038581609725952, + "grad_norm": 44.04487609863281, "learning_rate": 7.263702608511986e-06, - "loss": 0.2433, + "loss": 0.8239, "step": 60300 }, { "epoch": 6.37, - "grad_norm": 3.783695936203003, + "grad_norm": 55.9640998840332, "learning_rate": 7.261590453057346e-06, - "loss": 0.3046, + "loss": 0.8998, "step": 60310 }, { "epoch": 6.37, - "grad_norm": 35.035133361816406, + "grad_norm": 46.421966552734375, "learning_rate": 7.259478297602705e-06, - "loss": 0.5683, + "loss": 1.0603, "step": 60320 }, { "epoch": 6.37, - "grad_norm": 19.44709014892578, + "grad_norm": 37.099544525146484, "learning_rate": 7.257366142148063e-06, - "loss": 0.392, + "loss": 1.0698, "step": 60330 }, { "epoch": 6.37, - "grad_norm": 26.29820442199707, + "grad_norm": 33.81280517578125, "learning_rate": 7.255253986693421e-06, - "loss": 0.3738, + "loss": 1.0894, "step": 60340 }, { "epoch": 6.37, - "grad_norm": 49.011898040771484, + "grad_norm": 53.056915283203125, "learning_rate": 7.25314183123878e-06, - "loss": 0.4508, + "loss": 1.3045, "step": 60350 }, { "epoch": 6.37, - "grad_norm": 15.414215087890625, + "grad_norm": 23.056856155395508, "learning_rate": 7.251029675784138e-06, - "loss": 0.4357, + "loss": 1.0255, "step": 60360 }, { "epoch": 6.38, - "grad_norm": 20.99128532409668, + "grad_norm": 52.16392517089844, "learning_rate": 7.248917520329497e-06, - "loss": 0.3557, + "loss": 1.221, "step": 60370 }, { "epoch": 6.38, - "grad_norm": 7.859321594238281, + "grad_norm": 45.097999572753906, "learning_rate": 7.246805364874855e-06, - "loss": 0.3837, + "loss": 1.0033, "step": 60380 }, { "epoch": 6.38, - "grad_norm": 19.888105392456055, + "grad_norm": 36.16597366333008, "learning_rate": 7.2446932094202136e-06, - "loss": 0.6662, + "loss": 1.3207, "step": 60390 }, { "epoch": 6.38, - "grad_norm": 1.4414172172546387, + "grad_norm": 29.171724319458008, "learning_rate": 7.242581053965572e-06, - "loss": 0.5985, + "loss": 1.0011, "step": 60400 }, { "epoch": 6.38, - "grad_norm": 18.829160690307617, + "grad_norm": 69.93509674072266, "learning_rate": 7.24046889851093e-06, - "loss": 0.6088, + "loss": 0.9775, "step": 60410 }, { "epoch": 6.38, - "grad_norm": 8.045934677124023, + "grad_norm": 35.4500617980957, "learning_rate": 7.23835674305629e-06, - "loss": 0.3042, + "loss": 0.9844, "step": 60420 }, { "epoch": 6.38, - "grad_norm": 9.305889129638672, + "grad_norm": 41.743934631347656, "learning_rate": 7.236244587601649e-06, - "loss": 0.4093, + "loss": 0.9671, "step": 60430 }, { "epoch": 6.38, - "grad_norm": 23.160179138183594, + "grad_norm": 25.68147087097168, "learning_rate": 7.234132432147007e-06, - "loss": 0.2578, + "loss": 0.9245, "step": 60440 }, { "epoch": 6.38, - "grad_norm": 15.7325439453125, + "grad_norm": 56.82269287109375, "learning_rate": 7.232020276692365e-06, - "loss": 0.3529, + "loss": 0.8911, "step": 60450 }, { "epoch": 6.39, - "grad_norm": 21.701351165771484, + "grad_norm": 42.8682746887207, "learning_rate": 7.229908121237724e-06, - "loss": 0.5188, + "loss": 1.2165, "step": 60460 }, { "epoch": 6.39, - "grad_norm": 25.2921085357666, + "grad_norm": 68.98550415039062, "learning_rate": 7.227795965783082e-06, - "loss": 0.561, + "loss": 1.4281, "step": 60470 }, { "epoch": 6.39, - "grad_norm": 28.383882522583008, + "grad_norm": 47.9771614074707, "learning_rate": 7.225683810328441e-06, - "loss": 0.3217, + "loss": 0.9274, "step": 60480 }, { "epoch": 6.39, - "grad_norm": 18.78597640991211, + "grad_norm": 27.425079345703125, "learning_rate": 7.223571654873799e-06, - "loss": 0.3516, + "loss": 0.814, "step": 60490 }, { "epoch": 6.39, - "grad_norm": 0.9802560210227966, + "grad_norm": 44.81925964355469, "learning_rate": 7.221459499419158e-06, - "loss": 0.6242, + "loss": 1.1725, "step": 60500 }, { "epoch": 6.39, - "grad_norm": 3.1378774642944336, + "grad_norm": 43.00716781616211, "learning_rate": 7.219347343964516e-06, - "loss": 0.4955, + "loss": 0.934, "step": 60510 }, { "epoch": 6.39, - "grad_norm": 5.267636299133301, + "grad_norm": 13.245294570922852, "learning_rate": 7.217235188509874e-06, - "loss": 0.1179, + "loss": 0.599, "step": 60520 }, { "epoch": 6.39, - "grad_norm": 23.365373611450195, + "grad_norm": 46.82209396362305, "learning_rate": 7.215123033055234e-06, - "loss": 0.4995, + "loss": 1.1929, "step": 60530 }, { "epoch": 6.39, - "grad_norm": 22.135087966918945, + "grad_norm": 34.27598571777344, "learning_rate": 7.213010877600593e-06, - "loss": 0.2866, + "loss": 0.7264, "step": 60540 }, { "epoch": 6.39, - "grad_norm": 0.23166558146476746, + "grad_norm": 27.686174392700195, "learning_rate": 7.210898722145951e-06, - "loss": 0.7132, + "loss": 0.9135, "step": 60550 }, { "epoch": 6.4, - "grad_norm": 39.11579132080078, + "grad_norm": 62.589473724365234, "learning_rate": 7.2087865666913094e-06, - "loss": 0.573, + "loss": 1.0195, "step": 60560 }, { "epoch": 6.4, - "grad_norm": 49.49949264526367, + "grad_norm": 21.820234298706055, "learning_rate": 7.206674411236667e-06, - "loss": 0.3448, + "loss": 0.7159, "step": 60570 }, { "epoch": 6.4, - "grad_norm": 11.977889060974121, + "grad_norm": 28.512834548950195, "learning_rate": 7.204562255782026e-06, - "loss": 0.308, + "loss": 0.974, "step": 60580 }, { "epoch": 6.4, - "grad_norm": 42.44679641723633, + "grad_norm": 42.68867111206055, "learning_rate": 7.202450100327385e-06, - "loss": 0.3022, + "loss": 0.8673, "step": 60590 }, { "epoch": 6.4, - "grad_norm": 9.533109664916992, + "grad_norm": 38.50444412231445, "learning_rate": 7.200337944872743e-06, - "loss": 0.6979, + "loss": 1.0158, "step": 60600 }, { "epoch": 6.4, - "grad_norm": 18.876792907714844, + "grad_norm": 34.480159759521484, "learning_rate": 7.198225789418102e-06, - "loss": 0.2985, + "loss": 1.2118, "step": 60610 }, { "epoch": 6.4, - "grad_norm": 22.964799880981445, + "grad_norm": 52.93574905395508, "learning_rate": 7.19611363396346e-06, - "loss": 0.4292, + "loss": 0.8068, "step": 60620 }, { "epoch": 6.4, - "grad_norm": 0.13404180109500885, + "grad_norm": 20.77336883544922, "learning_rate": 7.194001478508818e-06, - "loss": 0.3127, + "loss": 0.8601, "step": 60630 }, { "epoch": 6.4, - "grad_norm": 17.95307731628418, + "grad_norm": 55.28954315185547, "learning_rate": 7.191889323054177e-06, - "loss": 0.6236, + "loss": 1.4527, "step": 60640 }, { "epoch": 6.41, - "grad_norm": 16.176982879638672, + "grad_norm": 39.32555389404297, "learning_rate": 7.189777167599535e-06, - "loss": 0.4959, + "loss": 1.014, "step": 60650 }, { "epoch": 6.41, - "grad_norm": 14.095586776733398, + "grad_norm": 32.55289840698242, "learning_rate": 7.187665012144895e-06, - "loss": 0.3573, + "loss": 0.9502, "step": 60660 }, { "epoch": 6.41, - "grad_norm": 11.196897506713867, + "grad_norm": 26.530029296875, "learning_rate": 7.1855528566902535e-06, - "loss": 0.1396, + "loss": 0.6927, "step": 60670 }, { "epoch": 6.41, - "grad_norm": 26.801368713378906, + "grad_norm": 29.923337936401367, "learning_rate": 7.183440701235611e-06, - "loss": 0.4346, + "loss": 0.7165, "step": 60680 }, { "epoch": 6.41, - "grad_norm": 0.8638777732849121, + "grad_norm": 16.969600677490234, "learning_rate": 7.18132854578097e-06, - "loss": 0.5192, + "loss": 1.2094, "step": 60690 }, { "epoch": 6.41, - "grad_norm": 3.5396244525909424, + "grad_norm": 21.1627140045166, "learning_rate": 7.179216390326329e-06, - "loss": 0.5414, + "loss": 0.8808, "step": 60700 }, { "epoch": 6.41, - "grad_norm": 10.841443061828613, + "grad_norm": 40.899024963378906, "learning_rate": 7.177104234871687e-06, - "loss": 0.5692, + "loss": 0.9623, "step": 60710 }, { "epoch": 6.41, - "grad_norm": 23.785655975341797, + "grad_norm": 29.174560546875, "learning_rate": 7.174992079417046e-06, - "loss": 0.4399, + "loss": 0.8956, "step": 60720 }, { "epoch": 6.41, - "grad_norm": 17.279294967651367, + "grad_norm": 41.67734146118164, "learning_rate": 7.1728799239624044e-06, - "loss": 0.4952, + "loss": 0.8842, "step": 60730 }, { "epoch": 6.41, - "grad_norm": 0.017870448529720306, + "grad_norm": 31.781660079956055, "learning_rate": 7.170767768507762e-06, - "loss": 0.2394, + "loss": 0.9072, "step": 60740 }, { "epoch": 6.42, - "grad_norm": 0.27424293756484985, + "grad_norm": 27.419937133789062, "learning_rate": 7.168655613053121e-06, - "loss": 0.2554, + "loss": 0.6889, "step": 60750 }, { "epoch": 6.42, - "grad_norm": 13.576187133789062, + "grad_norm": 16.3746280670166, "learning_rate": 7.166543457598479e-06, - "loss": 0.5746, + "loss": 1.0777, "step": 60760 }, { "epoch": 6.42, - "grad_norm": 10.900147438049316, + "grad_norm": 31.043670654296875, "learning_rate": 7.164431302143839e-06, - "loss": 0.6092, + "loss": 0.99, "step": 60770 }, { "epoch": 6.42, - "grad_norm": 7.245776176452637, + "grad_norm": 42.45387649536133, "learning_rate": 7.1623191466891975e-06, - "loss": 0.256, + "loss": 0.8414, "step": 60780 }, { "epoch": 6.42, - "grad_norm": 27.253692626953125, + "grad_norm": 55.61090087890625, "learning_rate": 7.160206991234555e-06, - "loss": 0.5761, + "loss": 1.3253, "step": 60790 }, { "epoch": 6.42, - "grad_norm": 22.825681686401367, + "grad_norm": 26.413589477539062, "learning_rate": 7.158094835779914e-06, - "loss": 0.3827, + "loss": 1.021, "step": 60800 }, { "epoch": 6.42, - "grad_norm": 22.721837997436523, + "grad_norm": 39.79566192626953, "learning_rate": 7.155982680325273e-06, - "loss": 0.4551, + "loss": 1.1358, "step": 60810 }, { "epoch": 6.42, - "grad_norm": 30.912395477294922, + "grad_norm": 35.57866287231445, "learning_rate": 7.153870524870631e-06, - "loss": 0.6065, + "loss": 0.9739, "step": 60820 }, { "epoch": 6.42, - "grad_norm": 2.1578927040100098, + "grad_norm": 32.486541748046875, "learning_rate": 7.15175836941599e-06, - "loss": 0.2036, + "loss": 1.0856, "step": 60830 }, { "epoch": 6.43, - "grad_norm": 0.09551671147346497, + "grad_norm": 20.019264221191406, "learning_rate": 7.149646213961348e-06, - "loss": 0.1909, + "loss": 0.7332, "step": 60840 }, { "epoch": 6.43, - "grad_norm": 7.933960914611816, + "grad_norm": 30.86153221130371, "learning_rate": 7.147534058506706e-06, - "loss": 0.5358, + "loss": 0.9398, "step": 60850 }, { "epoch": 6.43, - "grad_norm": 18.422651290893555, + "grad_norm": 36.95621871948242, "learning_rate": 7.145421903052065e-06, - "loss": 0.3791, + "loss": 0.82, "step": 60860 }, { "epoch": 6.43, - "grad_norm": 16.768583297729492, + "grad_norm": 33.62521743774414, "learning_rate": 7.143309747597423e-06, - "loss": 0.2148, + "loss": 1.1545, "step": 60870 }, { "epoch": 6.43, - "grad_norm": 40.6658821105957, + "grad_norm": 63.72465515136719, "learning_rate": 7.141197592142782e-06, - "loss": 0.253, + "loss": 0.9849, "step": 60880 }, { "epoch": 6.43, - "grad_norm": 22.872665405273438, + "grad_norm": 65.26847839355469, "learning_rate": 7.1390854366881415e-06, - "loss": 0.3711, + "loss": 1.1587, "step": 60890 }, { "epoch": 6.43, - "grad_norm": 1.936394214630127, + "grad_norm": 36.527774810791016, "learning_rate": 7.1369732812334994e-06, - "loss": 0.4134, + "loss": 1.1484, "step": 60900 }, { "epoch": 6.43, - "grad_norm": 6.630908966064453, + "grad_norm": 38.24110794067383, "learning_rate": 7.134861125778858e-06, - "loss": 0.2994, + "loss": 0.6932, "step": 60910 }, { "epoch": 6.43, - "grad_norm": 6.537262916564941, + "grad_norm": 22.15361785888672, "learning_rate": 7.132748970324217e-06, - "loss": 0.437, + "loss": 0.8305, "step": 60920 }, { "epoch": 6.43, - "grad_norm": 7.273198127746582, + "grad_norm": 42.21148681640625, "learning_rate": 7.130636814869575e-06, - "loss": 0.3034, + "loss": 0.8926, "step": 60930 }, { "epoch": 6.44, - "grad_norm": 41.56120681762695, + "grad_norm": 84.14460754394531, "learning_rate": 7.128524659414934e-06, - "loss": 0.2761, + "loss": 0.8574, "step": 60940 }, { "epoch": 6.44, - "grad_norm": 10.673771858215332, + "grad_norm": 10.068451881408691, "learning_rate": 7.126412503960292e-06, - "loss": 0.3499, + "loss": 0.7537, "step": 60950 }, { "epoch": 6.44, - "grad_norm": 1.9746747016906738, + "grad_norm": 44.52705383300781, "learning_rate": 7.12430034850565e-06, - "loss": 0.7065, + "loss": 1.1281, "step": 60960 }, { "epoch": 6.44, - "grad_norm": 26.512784957885742, + "grad_norm": 53.17909622192383, "learning_rate": 7.122188193051009e-06, - "loss": 0.5204, + "loss": 0.9786, "step": 60970 }, { "epoch": 6.44, - "grad_norm": 3.683894395828247, + "grad_norm": 27.09100914001465, "learning_rate": 7.120076037596367e-06, - "loss": 0.2553, + "loss": 0.8612, "step": 60980 }, { "epoch": 6.44, - "grad_norm": 0.9544952511787415, + "grad_norm": 41.290348052978516, "learning_rate": 7.117963882141726e-06, - "loss": 0.3747, + "loss": 1.0733, "step": 60990 }, { "epoch": 6.44, - "grad_norm": 5.92765474319458, + "grad_norm": 34.232486724853516, "learning_rate": 7.1158517266870855e-06, - "loss": 0.3875, + "loss": 0.8351, "step": 61000 }, { "epoch": 6.44, - "grad_norm": 20.321086883544922, + "grad_norm": 35.66914749145508, "learning_rate": 7.1137395712324435e-06, - "loss": 0.2304, + "loss": 0.721, "step": 61010 }, { "epoch": 6.44, - "grad_norm": 2.054307460784912, + "grad_norm": 32.33204650878906, "learning_rate": 7.111627415777802e-06, - "loss": 0.3539, + "loss": 0.998, "step": 61020 }, { "epoch": 6.45, - "grad_norm": 19.650484085083008, + "grad_norm": 42.706546783447266, "learning_rate": 7.10951526032316e-06, - "loss": 0.479, + "loss": 1.0856, "step": 61030 }, { "epoch": 6.45, - "grad_norm": 19.38808822631836, + "grad_norm": 35.21889877319336, "learning_rate": 7.107403104868519e-06, - "loss": 0.4291, + "loss": 1.2249, "step": 61040 }, { "epoch": 6.45, - "grad_norm": 10.11727237701416, + "grad_norm": 29.992063522338867, "learning_rate": 7.105290949413878e-06, - "loss": 0.4479, + "loss": 1.0132, "step": 61050 }, { "epoch": 6.45, - "grad_norm": 3.209700584411621, + "grad_norm": 48.46654510498047, "learning_rate": 7.103178793959236e-06, - "loss": 0.2823, + "loss": 1.0109, "step": 61060 }, { "epoch": 6.45, - "grad_norm": 3.936856985092163, + "grad_norm": 10.451431274414062, "learning_rate": 7.1010666385045944e-06, - "loss": 0.3736, + "loss": 0.5876, "step": 61070 }, { "epoch": 6.45, - "grad_norm": 31.75204849243164, + "grad_norm": 44.13108825683594, "learning_rate": 7.098954483049953e-06, - "loss": 0.5545, + "loss": 1.0824, "step": 61080 }, { "epoch": 6.45, - "grad_norm": 9.5033597946167, + "grad_norm": 12.733611106872559, "learning_rate": 7.096842327595311e-06, - "loss": 0.2648, + "loss": 0.79, "step": 61090 }, { "epoch": 6.45, - "grad_norm": 7.565301895141602, + "grad_norm": 51.25638198852539, "learning_rate": 7.09473017214067e-06, - "loss": 0.5888, + "loss": 0.9798, "step": 61100 }, { "epoch": 6.45, - "grad_norm": 13.636183738708496, + "grad_norm": 29.197799682617188, "learning_rate": 7.092618016686028e-06, - "loss": 0.3424, + "loss": 0.7468, "step": 61110 }, { "epoch": 6.45, - "grad_norm": 6.555069446563721, + "grad_norm": 47.047847747802734, "learning_rate": 7.0905058612313875e-06, - "loss": 0.2775, + "loss": 0.681, "step": 61120 }, { "epoch": 6.46, - "grad_norm": 37.35744857788086, + "grad_norm": 35.06285858154297, "learning_rate": 7.088393705776746e-06, - "loss": 0.3611, + "loss": 0.8448, "step": 61130 }, { "epoch": 6.46, - "grad_norm": 0.799351155757904, + "grad_norm": 29.84563636779785, "learning_rate": 7.086281550322104e-06, - "loss": 0.4983, + "loss": 0.8114, "step": 61140 }, { "epoch": 6.46, - "grad_norm": 2.8156440258026123, + "grad_norm": 21.9171085357666, "learning_rate": 7.084169394867463e-06, - "loss": 0.1969, + "loss": 0.6671, "step": 61150 }, { "epoch": 6.46, - "grad_norm": 11.849166870117188, + "grad_norm": 39.61293029785156, "learning_rate": 7.082057239412822e-06, - "loss": 0.4833, + "loss": 1.3576, "step": 61160 }, { "epoch": 6.46, - "grad_norm": 15.173765182495117, + "grad_norm": 53.181846618652344, "learning_rate": 7.07994508395818e-06, - "loss": 0.3688, + "loss": 0.9958, "step": 61170 }, { "epoch": 6.46, - "grad_norm": 5.004420280456543, + "grad_norm": 43.27952575683594, "learning_rate": 7.0778329285035385e-06, - "loss": 0.3215, + "loss": 1.2368, "step": 61180 }, { "epoch": 6.46, - "grad_norm": 0.06151815131306648, + "grad_norm": 27.887197494506836, "learning_rate": 7.075720773048897e-06, - "loss": 0.4714, + "loss": 1.1106, "step": 61190 }, { "epoch": 6.46, - "grad_norm": 0.6886494755744934, + "grad_norm": 22.6101016998291, "learning_rate": 7.073608617594255e-06, - "loss": 0.3107, + "loss": 0.8618, "step": 61200 }, { "epoch": 6.46, - "grad_norm": 1.3601773977279663, + "grad_norm": 28.99878692626953, "learning_rate": 7.071496462139614e-06, - "loss": 0.3417, + "loss": 0.981, "step": 61210 }, { "epoch": 6.47, - "grad_norm": 1.5008002519607544, + "grad_norm": 36.3713493347168, "learning_rate": 7.069384306684972e-06, - "loss": 0.3253, + "loss": 1.1121, "step": 61220 }, { "epoch": 6.47, - "grad_norm": 1.3461788892745972, + "grad_norm": 48.28014373779297, "learning_rate": 7.067272151230331e-06, - "loss": 0.3691, + "loss": 0.9114, "step": 61230 }, { "epoch": 6.47, - "grad_norm": 20.755390167236328, + "grad_norm": 46.72755813598633, "learning_rate": 7.06515999577569e-06, - "loss": 0.2634, + "loss": 0.9904, "step": 61240 }, { "epoch": 6.47, - "grad_norm": 0.10672837495803833, + "grad_norm": 40.02750778198242, "learning_rate": 7.063047840321048e-06, - "loss": 0.1335, + "loss": 0.7757, "step": 61250 }, { "epoch": 6.47, - "grad_norm": 0.15424270927906036, + "grad_norm": 44.14487075805664, "learning_rate": 7.060935684866407e-06, - "loss": 0.3814, + "loss": 1.0645, "step": 61260 }, { "epoch": 6.47, - "grad_norm": 2.7247354984283447, + "grad_norm": 45.2271842956543, "learning_rate": 7.058823529411766e-06, - "loss": 0.4315, + "loss": 1.0635, "step": 61270 }, { "epoch": 6.47, - "grad_norm": 8.814483642578125, + "grad_norm": 39.03321838378906, "learning_rate": 7.056711373957124e-06, - "loss": 0.2554, + "loss": 1.2129, "step": 61280 }, { "epoch": 6.47, - "grad_norm": 1.0148561000823975, + "grad_norm": 34.481876373291016, "learning_rate": 7.0545992185024825e-06, - "loss": 0.3473, + "loss": 0.887, "step": 61290 }, { "epoch": 6.47, - "grad_norm": 9.929059028625488, + "grad_norm": 43.02588653564453, "learning_rate": 7.0524870630478404e-06, - "loss": 0.5664, + "loss": 0.9794, "step": 61300 }, { "epoch": 6.47, - "grad_norm": 2.4358606338500977, + "grad_norm": 41.674766540527344, "learning_rate": 7.050374907593199e-06, - "loss": 0.4307, + "loss": 1.2355, "step": 61310 }, { "epoch": 6.48, - "grad_norm": 19.837427139282227, + "grad_norm": 46.37883377075195, "learning_rate": 7.048262752138558e-06, - "loss": 0.5373, + "loss": 1.2051, "step": 61320 }, { "epoch": 6.48, - "grad_norm": 22.524738311767578, + "grad_norm": 47.08023452758789, "learning_rate": 7.046150596683916e-06, - "loss": 0.2306, + "loss": 0.7771, "step": 61330 }, { "epoch": 6.48, - "grad_norm": 17.192956924438477, + "grad_norm": 31.923934936523438, "learning_rate": 7.044038441229275e-06, - "loss": 0.4759, + "loss": 1.1667, "step": 61340 }, { "epoch": 6.48, - "grad_norm": 7.2511091232299805, + "grad_norm": 14.838004112243652, "learning_rate": 7.041926285774634e-06, - "loss": 0.4903, + "loss": 1.0167, "step": 61350 }, { "epoch": 6.48, - "grad_norm": 12.599971771240234, + "grad_norm": 50.505760192871094, "learning_rate": 7.039814130319992e-06, - "loss": 0.4657, + "loss": 0.9402, "step": 61360 }, { "epoch": 6.48, - "grad_norm": 12.101052284240723, + "grad_norm": 22.797025680541992, "learning_rate": 7.037701974865351e-06, - "loss": 0.1622, + "loss": 0.8999, "step": 61370 }, { "epoch": 6.48, - "grad_norm": 0.6428192853927612, + "grad_norm": 22.93686866760254, "learning_rate": 7.03558981941071e-06, - "loss": 0.4405, + "loss": 0.7901, "step": 61380 }, { "epoch": 6.48, - "grad_norm": 20.763933181762695, + "grad_norm": 31.75774383544922, "learning_rate": 7.033477663956068e-06, - "loss": 0.3421, + "loss": 0.7621, "step": 61390 }, { "epoch": 6.48, - "grad_norm": 31.13280487060547, + "grad_norm": 55.29506301879883, "learning_rate": 7.0313655085014265e-06, - "loss": 0.6516, + "loss": 1.1619, "step": 61400 }, { "epoch": 6.49, - "grad_norm": 17.299175262451172, + "grad_norm": 35.33311462402344, "learning_rate": 7.0292533530467845e-06, - "loss": 0.3523, + "loss": 0.4313, "step": 61410 }, { "epoch": 6.49, - "grad_norm": 3.5294482707977295, + "grad_norm": 50.801029205322266, "learning_rate": 7.027141197592143e-06, - "loss": 0.2099, + "loss": 0.9414, "step": 61420 }, { "epoch": 6.49, - "grad_norm": 34.66842269897461, + "grad_norm": 46.21016311645508, "learning_rate": 7.025029042137502e-06, - "loss": 0.3698, + "loss": 0.7, "step": 61430 }, { "epoch": 6.49, - "grad_norm": 11.216038703918457, + "grad_norm": 30.954177856445312, "learning_rate": 7.02291688668286e-06, - "loss": 0.6982, + "loss": 1.023, "step": 61440 }, { "epoch": 6.49, - "grad_norm": 45.170570373535156, + "grad_norm": 31.752840042114258, "learning_rate": 7.020804731228219e-06, - "loss": 0.7796, + "loss": 1.0986, "step": 61450 }, { "epoch": 6.49, - "grad_norm": 4.458937644958496, + "grad_norm": 25.312292098999023, "learning_rate": 7.018692575773578e-06, - "loss": 0.2542, + "loss": 0.9764, "step": 61460 }, { "epoch": 6.49, - "grad_norm": 32.45859146118164, + "grad_norm": 28.834699630737305, "learning_rate": 7.0165804203189354e-06, - "loss": 0.3708, + "loss": 0.854, "step": 61470 }, { "epoch": 6.49, - "grad_norm": 13.107635498046875, + "grad_norm": 38.67496109008789, "learning_rate": 7.014468264864295e-06, - "loss": 0.291, + "loss": 1.0281, "step": 61480 }, { "epoch": 6.49, - "grad_norm": 18.556171417236328, + "grad_norm": 40.61162567138672, "learning_rate": 7.012356109409653e-06, - "loss": 0.5315, + "loss": 0.9654, "step": 61490 }, { "epoch": 6.49, - "grad_norm": 6.690639495849609, + "grad_norm": 57.46870422363281, "learning_rate": 7.010243953955012e-06, - "loss": 0.4773, + "loss": 1.0045, "step": 61500 }, { "epoch": 6.5, - "grad_norm": 17.03712272644043, + "grad_norm": 53.25534439086914, "learning_rate": 7.0081317985003705e-06, - "loss": 0.4036, + "loss": 1.0537, "step": 61510 }, { "epoch": 6.5, - "grad_norm": 9.474498748779297, + "grad_norm": 23.71471405029297, "learning_rate": 7.0060196430457285e-06, - "loss": 0.3151, + "loss": 0.8474, "step": 61520 }, { "epoch": 6.5, - "grad_norm": 13.03640365600586, + "grad_norm": 39.24014663696289, "learning_rate": 7.003907487591087e-06, - "loss": 0.3045, + "loss": 0.8152, "step": 61530 }, { "epoch": 6.5, - "grad_norm": 6.936209201812744, + "grad_norm": 38.8610725402832, "learning_rate": 7.001795332136446e-06, - "loss": 0.2505, + "loss": 0.7601, "step": 61540 }, { "epoch": 6.5, - "grad_norm": 10.287223815917969, + "grad_norm": 45.088436126708984, "learning_rate": 6.999683176681804e-06, - "loss": 0.4458, + "loss": 0.7825, "step": 61550 }, { "epoch": 6.5, - "grad_norm": 29.068004608154297, + "grad_norm": 49.07969665527344, "learning_rate": 6.997571021227163e-06, - "loss": 0.6014, + "loss": 1.1571, "step": 61560 }, { "epoch": 6.5, - "grad_norm": 27.290403366088867, + "grad_norm": 62.08811950683594, "learning_rate": 6.995458865772522e-06, - "loss": 0.4218, + "loss": 1.1875, "step": 61570 }, { "epoch": 6.5, - "grad_norm": 12.99228572845459, + "grad_norm": 24.831846237182617, "learning_rate": 6.9933467103178795e-06, - "loss": 0.4465, + "loss": 0.8837, "step": 61580 }, { "epoch": 6.5, - "grad_norm": 15.305882453918457, + "grad_norm": 30.794950485229492, "learning_rate": 6.991234554863239e-06, - "loss": 0.2179, + "loss": 0.8969, "step": 61590 }, { "epoch": 6.51, - "grad_norm": 0.18196366727352142, + "grad_norm": 44.2282829284668, "learning_rate": 6.989122399408597e-06, - "loss": 0.1978, + "loss": 0.9462, "step": 61600 }, { "epoch": 6.51, - "grad_norm": 8.548462867736816, + "grad_norm": 35.12602996826172, "learning_rate": 6.987010243953956e-06, - "loss": 0.3279, + "loss": 0.8225, "step": 61610 }, { "epoch": 6.51, - "grad_norm": 0.7193246483802795, + "grad_norm": 38.42698287963867, "learning_rate": 6.9848980884993146e-06, - "loss": 0.3465, + "loss": 0.9158, "step": 61620 }, { "epoch": 6.51, - "grad_norm": 50.08190155029297, + "grad_norm": 15.6392240524292, "learning_rate": 6.9827859330446725e-06, - "loss": 0.3409, + "loss": 0.8938, "step": 61630 }, { "epoch": 6.51, - "grad_norm": 26.13961410522461, + "grad_norm": 30.761917114257812, "learning_rate": 6.980673777590031e-06, - "loss": 0.5289, + "loss": 0.9889, "step": 61640 }, { "epoch": 6.51, - "grad_norm": 4.579929828643799, + "grad_norm": 9.949880599975586, "learning_rate": 6.97856162213539e-06, - "loss": 0.318, + "loss": 0.9191, "step": 61650 }, { "epoch": 6.51, - "grad_norm": 13.75798511505127, + "grad_norm": 30.93632698059082, "learning_rate": 6.976449466680748e-06, - "loss": 0.3548, + "loss": 0.7646, "step": 61660 }, { "epoch": 6.51, - "grad_norm": 4.211562633514404, + "grad_norm": 49.351802825927734, "learning_rate": 6.974337311226107e-06, - "loss": 0.3649, + "loss": 0.8561, "step": 61670 }, { "epoch": 6.51, - "grad_norm": 2.7848708629608154, + "grad_norm": 29.30666160583496, "learning_rate": 6.972225155771465e-06, - "loss": 0.4296, + "loss": 0.7875, "step": 61680 }, { "epoch": 6.51, - "grad_norm": 6.676815032958984, + "grad_norm": 35.72938919067383, "learning_rate": 6.9701130003168235e-06, - "loss": 0.6095, + "loss": 1.1308, "step": 61690 }, { "epoch": 6.52, - "grad_norm": 12.43720817565918, + "grad_norm": 29.061328887939453, "learning_rate": 6.968000844862183e-06, - "loss": 0.252, + "loss": 0.853, "step": 61700 }, { "epoch": 6.52, - "grad_norm": 0.4344608783721924, + "grad_norm": 30.958457946777344, "learning_rate": 6.96588868940754e-06, - "loss": 0.5917, + "loss": 0.9002, "step": 61710 }, { "epoch": 6.52, - "grad_norm": 37.19363784790039, + "grad_norm": 41.77238464355469, "learning_rate": 6.9637765339529e-06, - "loss": 0.4864, + "loss": 1.054, "step": 61720 }, { "epoch": 6.52, - "grad_norm": 32.60762405395508, + "grad_norm": 37.081512451171875, "learning_rate": 6.961664378498259e-06, - "loss": 0.5051, + "loss": 0.5888, "step": 61730 }, { "epoch": 6.52, - "grad_norm": 0.6077075004577637, + "grad_norm": 32.220088958740234, "learning_rate": 6.9595522230436165e-06, - "loss": 0.488, + "loss": 0.944, "step": 61740 }, { "epoch": 6.52, - "grad_norm": 1.3644523620605469, + "grad_norm": 10.69336986541748, "learning_rate": 6.957440067588975e-06, - "loss": 0.4152, + "loss": 1.3193, "step": 61750 }, { "epoch": 6.52, - "grad_norm": 22.378376007080078, + "grad_norm": 59.25510025024414, "learning_rate": 6.955327912134333e-06, - "loss": 0.3504, + "loss": 1.1859, "step": 61760 }, { "epoch": 6.52, - "grad_norm": 12.725074768066406, + "grad_norm": 66.11731719970703, "learning_rate": 6.953215756679692e-06, - "loss": 0.3844, + "loss": 1.2622, "step": 61770 }, { "epoch": 6.52, - "grad_norm": 19.838687896728516, + "grad_norm": 44.45241928100586, "learning_rate": 6.951103601225051e-06, - "loss": 0.6341, + "loss": 0.7292, "step": 61780 }, { "epoch": 6.53, - "grad_norm": 0.43066367506980896, + "grad_norm": 25.22803497314453, "learning_rate": 6.948991445770409e-06, - "loss": 0.3776, + "loss": 0.9206, "step": 61790 }, { "epoch": 6.53, - "grad_norm": 1.834172248840332, + "grad_norm": 46.23142623901367, "learning_rate": 6.9468792903157675e-06, - "loss": 0.5297, + "loss": 0.8416, "step": 61800 }, { "epoch": 6.53, - "grad_norm": 40.75456237792969, + "grad_norm": 49.9002571105957, "learning_rate": 6.944767134861127e-06, - "loss": 0.4167, + "loss": 0.9346, "step": 61810 }, { "epoch": 6.53, - "grad_norm": 2.3245060443878174, + "grad_norm": 46.405967712402344, "learning_rate": 6.942654979406484e-06, - "loss": 0.4953, + "loss": 1.0697, "step": 61820 }, { "epoch": 6.53, - "grad_norm": 6.072283744812012, + "grad_norm": 66.54711151123047, "learning_rate": 6.940542823951844e-06, - "loss": 0.4254, + "loss": 0.8711, "step": 61830 }, { "epoch": 6.53, - "grad_norm": 0.2238379418849945, + "grad_norm": 38.68852996826172, "learning_rate": 6.938430668497203e-06, - "loss": 0.2514, + "loss": 1.0028, "step": 61840 }, { "epoch": 6.53, - "grad_norm": 3.6943321228027344, + "grad_norm": 47.40900421142578, "learning_rate": 6.9363185130425606e-06, - "loss": 0.3299, + "loss": 1.032, "step": 61850 }, { "epoch": 6.53, - "grad_norm": 34.32553482055664, + "grad_norm": 49.202327728271484, "learning_rate": 6.934206357587919e-06, - "loss": 0.4347, + "loss": 0.8585, "step": 61860 }, { "epoch": 6.53, - "grad_norm": 19.57870864868164, + "grad_norm": 59.6739387512207, "learning_rate": 6.932094202133277e-06, - "loss": 0.7264, + "loss": 1.1917, "step": 61870 }, { "epoch": 6.54, - "grad_norm": 33.28584671020508, + "grad_norm": 41.66938400268555, "learning_rate": 6.929982046678636e-06, - "loss": 0.4057, + "loss": 0.8422, "step": 61880 }, { "epoch": 6.54, - "grad_norm": 21.140172958374023, + "grad_norm": 49.983341217041016, "learning_rate": 6.927869891223995e-06, - "loss": 0.2673, + "loss": 0.8125, "step": 61890 }, { "epoch": 6.54, - "grad_norm": 43.38174057006836, + "grad_norm": 52.162784576416016, "learning_rate": 6.925757735769353e-06, - "loss": 0.3158, + "loss": 0.8729, "step": 61900 }, { "epoch": 6.54, - "grad_norm": 14.918269157409668, + "grad_norm": 52.328548431396484, "learning_rate": 6.9236455803147115e-06, - "loss": 0.6683, + "loss": 1.3598, "step": 61910 }, { "epoch": 6.54, - "grad_norm": 20.675111770629883, + "grad_norm": 44.915836334228516, "learning_rate": 6.921533424860071e-06, - "loss": 0.5675, + "loss": 1.1058, "step": 61920 }, { "epoch": 6.54, - "grad_norm": 17.362943649291992, + "grad_norm": 41.90135192871094, "learning_rate": 6.919421269405428e-06, - "loss": 0.4167, + "loss": 1.1529, "step": 61930 }, { "epoch": 6.54, - "grad_norm": 20.848682403564453, + "grad_norm": 32.12583923339844, "learning_rate": 6.917309113950788e-06, - "loss": 0.4095, + "loss": 1.0728, "step": 61940 }, { "epoch": 6.54, - "grad_norm": 12.620323181152344, + "grad_norm": 40.42206954956055, "learning_rate": 6.915196958496145e-06, - "loss": 0.4396, + "loss": 0.8931, "step": 61950 }, { "epoch": 6.54, - "grad_norm": 17.256410598754883, + "grad_norm": 45.7088623046875, "learning_rate": 6.913084803041505e-06, - "loss": 0.4863, + "loss": 1.1456, "step": 61960 }, { "epoch": 6.54, - "grad_norm": 40.19995880126953, + "grad_norm": 26.85828971862793, "learning_rate": 6.910972647586863e-06, - "loss": 0.2868, + "loss": 0.7084, "step": 61970 }, { "epoch": 6.55, - "grad_norm": 9.065096855163574, + "grad_norm": 30.642288208007812, "learning_rate": 6.908860492132221e-06, - "loss": 0.3092, + "loss": 0.8786, "step": 61980 }, { "epoch": 6.55, - "grad_norm": 1.569353461265564, + "grad_norm": 37.257789611816406, "learning_rate": 6.90674833667758e-06, - "loss": 0.3542, + "loss": 0.9299, "step": 61990 }, { "epoch": 6.55, - "grad_norm": 11.124305725097656, + "grad_norm": 43.225730895996094, "learning_rate": 6.904636181222939e-06, - "loss": 0.6374, + "loss": 1.389, "step": 62000 }, { "epoch": 6.55, - "grad_norm": 7.853727340698242, + "grad_norm": 21.33567237854004, "learning_rate": 6.902524025768297e-06, - "loss": 0.233, + "loss": 0.658, "step": 62010 }, { "epoch": 6.55, - "grad_norm": 11.805658340454102, + "grad_norm": 15.304056167602539, "learning_rate": 6.9004118703136556e-06, - "loss": 0.4806, + "loss": 1.3249, "step": 62020 }, { "epoch": 6.55, - "grad_norm": 5.518868446350098, + "grad_norm": 30.130699157714844, "learning_rate": 6.898299714859014e-06, - "loss": 0.2323, + "loss": 1.1373, "step": 62030 }, { "epoch": 6.55, - "grad_norm": 70.20775604248047, + "grad_norm": 47.1948356628418, "learning_rate": 6.896187559404372e-06, - "loss": 0.5962, + "loss": 0.9593, "step": 62040 }, { "epoch": 6.55, - "grad_norm": 32.83335494995117, + "grad_norm": 28.091474533081055, "learning_rate": 6.894075403949732e-06, - "loss": 0.7701, + "loss": 1.4457, "step": 62050 }, { "epoch": 6.55, - "grad_norm": 15.07331371307373, + "grad_norm": 47.85617446899414, "learning_rate": 6.891963248495089e-06, - "loss": 0.3599, + "loss": 0.7774, "step": 62060 }, { "epoch": 6.56, - "grad_norm": 25.786558151245117, + "grad_norm": 43.43027114868164, "learning_rate": 6.889851093040449e-06, - "loss": 0.3872, + "loss": 1.1039, "step": 62070 }, { "epoch": 6.56, - "grad_norm": 7.375338554382324, + "grad_norm": 41.100746154785156, "learning_rate": 6.887738937585807e-06, - "loss": 0.4685, + "loss": 1.2105, "step": 62080 }, { "epoch": 6.56, - "grad_norm": 0.931303083896637, + "grad_norm": 47.619144439697266, "learning_rate": 6.885626782131165e-06, - "loss": 0.4055, + "loss": 1.1924, "step": 62090 }, { "epoch": 6.56, - "grad_norm": 0.41619718074798584, + "grad_norm": 39.338199615478516, "learning_rate": 6.883514626676524e-06, - "loss": 0.3244, + "loss": 0.9798, "step": 62100 }, { "epoch": 6.56, - "grad_norm": 16.089597702026367, + "grad_norm": 32.96352767944336, "learning_rate": 6.881402471221883e-06, - "loss": 0.2802, + "loss": 0.8279, "step": 62110 }, { "epoch": 6.56, - "grad_norm": 0.306647390127182, + "grad_norm": 17.41028594970703, "learning_rate": 6.879290315767241e-06, - "loss": 0.5769, + "loss": 1.1591, "step": 62120 }, { "epoch": 6.56, - "grad_norm": 24.252246856689453, + "grad_norm": 73.99786376953125, "learning_rate": 6.8771781603126e-06, - "loss": 0.461, + "loss": 1.1517, "step": 62130 }, { "epoch": 6.56, - "grad_norm": 9.56135368347168, + "grad_norm": 42.44074630737305, "learning_rate": 6.8750660048579575e-06, - "loss": 0.5383, + "loss": 0.9902, "step": 62140 }, { "epoch": 6.56, - "grad_norm": 2.475172281265259, + "grad_norm": 41.80317687988281, "learning_rate": 6.872953849403316e-06, - "loss": 0.2667, + "loss": 0.7865, "step": 62150 }, { "epoch": 6.56, - "grad_norm": 31.6461181640625, + "grad_norm": 27.328283309936523, "learning_rate": 6.870841693948676e-06, - "loss": 0.2291, + "loss": 0.944, "step": 62160 }, { "epoch": 6.57, - "grad_norm": 9.730640411376953, + "grad_norm": 46.36482620239258, "learning_rate": 6.868729538494033e-06, - "loss": 0.1637, + "loss": 0.6861, "step": 62170 }, { "epoch": 6.57, - "grad_norm": 3.265855073928833, + "grad_norm": 15.57198429107666, "learning_rate": 6.866617383039393e-06, - "loss": 0.2431, + "loss": 0.9681, "step": 62180 }, { "epoch": 6.57, - "grad_norm": 20.82486915588379, + "grad_norm": 44.494503021240234, "learning_rate": 6.864505227584751e-06, - "loss": 0.3953, + "loss": 0.7389, "step": 62190 }, { "epoch": 6.57, - "grad_norm": 0.651806116104126, + "grad_norm": 47.960296630859375, "learning_rate": 6.862393072130109e-06, - "loss": 0.2928, + "loss": 0.8987, "step": 62200 }, { "epoch": 6.57, - "grad_norm": 6.2716803550720215, + "grad_norm": 35.19856262207031, "learning_rate": 6.860280916675468e-06, - "loss": 0.8964, + "loss": 1.4631, "step": 62210 }, { "epoch": 6.57, - "grad_norm": 13.494510650634766, + "grad_norm": 47.3345947265625, "learning_rate": 6.858168761220826e-06, - "loss": 0.3712, + "loss": 0.9147, "step": 62220 }, { "epoch": 6.57, - "grad_norm": 1.854831337928772, + "grad_norm": 33.27113342285156, "learning_rate": 6.856056605766185e-06, - "loss": 0.5746, + "loss": 0.7972, "step": 62230 }, { "epoch": 6.57, - "grad_norm": 6.612288475036621, + "grad_norm": 38.53252029418945, "learning_rate": 6.853944450311544e-06, - "loss": 0.2882, + "loss": 1.0194, "step": 62240 }, { "epoch": 6.57, - "grad_norm": 5.480555534362793, + "grad_norm": 55.89189529418945, "learning_rate": 6.8518322948569015e-06, - "loss": 0.2305, + "loss": 1.016, "step": 62250 }, { "epoch": 6.58, - "grad_norm": 17.049762725830078, + "grad_norm": 52.354759216308594, "learning_rate": 6.84972013940226e-06, - "loss": 0.4713, + "loss": 0.8635, "step": 62260 }, { "epoch": 6.58, - "grad_norm": 16.73185157775879, + "grad_norm": 37.95572280883789, "learning_rate": 6.847607983947619e-06, - "loss": 0.6095, + "loss": 1.0626, "step": 62270 }, { "epoch": 6.58, - "grad_norm": 11.222627639770508, + "grad_norm": 51.761905670166016, "learning_rate": 6.845495828492977e-06, - "loss": 0.4986, + "loss": 1.4004, "step": 62280 }, { "epoch": 6.58, - "grad_norm": 7.933298587799072, + "grad_norm": 58.8017463684082, "learning_rate": 6.843383673038337e-06, - "loss": 0.5641, + "loss": 1.2662, "step": 62290 }, { "epoch": 6.58, - "grad_norm": 1.1194556951522827, + "grad_norm": 8.084580421447754, "learning_rate": 6.8412715175836954e-06, - "loss": 0.5501, + "loss": 1.0189, "step": 62300 }, { "epoch": 6.58, - "grad_norm": 15.706729888916016, + "grad_norm": 24.754636764526367, "learning_rate": 6.839159362129053e-06, - "loss": 0.4701, + "loss": 0.9779, "step": 62310 }, { "epoch": 6.58, - "grad_norm": 15.942269325256348, + "grad_norm": 55.87208557128906, "learning_rate": 6.837047206674412e-06, - "loss": 0.4228, + "loss": 1.0968, "step": 62320 }, { "epoch": 6.58, - "grad_norm": 33.24531555175781, + "grad_norm": 44.91211700439453, "learning_rate": 6.83493505121977e-06, - "loss": 0.2768, + "loss": 1.0715, "step": 62330 }, { "epoch": 6.58, - "grad_norm": 18.827899932861328, + "grad_norm": 49.89811706542969, "learning_rate": 6.832822895765129e-06, - "loss": 0.3012, + "loss": 1.2635, "step": 62340 }, { "epoch": 6.58, - "grad_norm": 21.695907592773438, + "grad_norm": 49.048583984375, "learning_rate": 6.830710740310488e-06, - "loss": 0.2155, + "loss": 0.7396, "step": 62350 }, { "epoch": 6.59, - "grad_norm": 0.7022707462310791, + "grad_norm": 12.932645797729492, "learning_rate": 6.8285985848558456e-06, - "loss": 0.8812, + "loss": 1.1611, "step": 62360 }, { "epoch": 6.59, - "grad_norm": 18.877042770385742, + "grad_norm": 38.05107498168945, "learning_rate": 6.826486429401204e-06, - "loss": 0.6914, + "loss": 1.0061, "step": 62370 }, { "epoch": 6.59, - "grad_norm": 0.6952618956565857, + "grad_norm": 24.623703002929688, "learning_rate": 6.824374273946563e-06, - "loss": 0.3275, + "loss": 0.6018, "step": 62380 }, { "epoch": 6.59, - "grad_norm": 9.364805221557617, + "grad_norm": 44.244224548339844, "learning_rate": 6.822262118491921e-06, - "loss": 0.3004, + "loss": 1.0504, "step": 62390 }, { "epoch": 6.59, - "grad_norm": 12.706867218017578, + "grad_norm": 59.81222915649414, "learning_rate": 6.820149963037281e-06, - "loss": 0.3408, + "loss": 0.9285, "step": 62400 }, { "epoch": 6.59, - "grad_norm": 36.3751220703125, + "grad_norm": 38.31977462768555, "learning_rate": 6.818037807582638e-06, - "loss": 0.6249, + "loss": 0.7365, "step": 62410 }, { "epoch": 6.59, - "grad_norm": 36.113121032714844, + "grad_norm": 22.123136520385742, "learning_rate": 6.815925652127997e-06, - "loss": 0.4488, + "loss": 0.8676, "step": 62420 }, { "epoch": 6.59, - "grad_norm": 19.009281158447266, + "grad_norm": 38.117698669433594, "learning_rate": 6.813813496673356e-06, - "loss": 0.405, + "loss": 1.1195, "step": 62430 }, { "epoch": 6.59, - "grad_norm": 25.924314498901367, + "grad_norm": 24.052453994750977, "learning_rate": 6.811701341218714e-06, - "loss": 0.3948, + "loss": 0.8996, "step": 62440 }, { "epoch": 6.6, - "grad_norm": 20.152692794799805, + "grad_norm": 39.448822021484375, "learning_rate": 6.809589185764073e-06, - "loss": 0.5331, + "loss": 0.9933, "step": 62450 }, { "epoch": 6.6, - "grad_norm": 9.554362297058105, + "grad_norm": 53.08499526977539, "learning_rate": 6.807477030309432e-06, - "loss": 0.2807, + "loss": 0.8123, "step": 62460 }, { "epoch": 6.6, - "grad_norm": 32.435821533203125, + "grad_norm": 37.84026336669922, "learning_rate": 6.80536487485479e-06, - "loss": 0.5119, + "loss": 1.0605, "step": 62470 }, { "epoch": 6.6, - "grad_norm": 2.754546642303467, + "grad_norm": 39.26096725463867, "learning_rate": 6.803252719400148e-06, - "loss": 0.2932, + "loss": 0.9962, "step": 62480 }, { "epoch": 6.6, - "grad_norm": 23.061555862426758, + "grad_norm": 45.30791091918945, "learning_rate": 6.801140563945507e-06, - "loss": 0.3815, + "loss": 1.0056, "step": 62490 }, { "epoch": 6.6, - "grad_norm": 6.522719860076904, + "grad_norm": 46.29961013793945, "learning_rate": 6.799028408490865e-06, - "loss": 0.5243, + "loss": 1.2039, "step": 62500 }, { "epoch": 6.6, - "grad_norm": 1.2171076536178589, + "grad_norm": 30.702238082885742, "learning_rate": 6.796916253036224e-06, - "loss": 0.5227, + "loss": 0.624, "step": 62510 }, { "epoch": 6.6, - "grad_norm": 16.32948112487793, + "grad_norm": 46.41106414794922, "learning_rate": 6.794804097581582e-06, - "loss": 0.4965, + "loss": 1.1668, "step": 62520 }, { "epoch": 6.6, - "grad_norm": 13.64924144744873, + "grad_norm": 30.240333557128906, "learning_rate": 6.792691942126941e-06, - "loss": 0.3817, + "loss": 0.7039, "step": 62530 }, { "epoch": 6.6, - "grad_norm": 21.033266067504883, + "grad_norm": 59.33973693847656, "learning_rate": 6.7905797866723e-06, - "loss": 0.3848, + "loss": 0.6633, "step": 62540 }, { "epoch": 6.61, - "grad_norm": 0.29960665106773376, + "grad_norm": 27.697477340698242, "learning_rate": 6.788467631217658e-06, - "loss": 0.3674, + "loss": 0.7669, "step": 62550 }, { "epoch": 6.61, - "grad_norm": 28.357879638671875, + "grad_norm": 46.28355407714844, "learning_rate": 6.786355475763017e-06, - "loss": 0.3798, + "loss": 0.7321, "step": 62560 }, { "epoch": 6.61, - "grad_norm": 13.459447860717773, + "grad_norm": 27.37053680419922, "learning_rate": 6.784243320308376e-06, - "loss": 0.3338, + "loss": 1.1836, "step": 62570 }, { "epoch": 6.61, - "grad_norm": 18.41570472717285, + "grad_norm": 28.11176109313965, "learning_rate": 6.782131164853734e-06, - "loss": 0.3941, + "loss": 0.8867, "step": 62580 }, { "epoch": 6.61, - "grad_norm": 10.296189308166504, + "grad_norm": 45.18759536743164, "learning_rate": 6.780019009399092e-06, - "loss": 0.2018, + "loss": 0.797, "step": 62590 }, { "epoch": 6.61, - "grad_norm": 37.742671966552734, + "grad_norm": 36.326072692871094, "learning_rate": 6.77790685394445e-06, - "loss": 0.6734, + "loss": 0.9012, "step": 62600 }, { "epoch": 6.61, - "grad_norm": 5.529124736785889, + "grad_norm": 36.084346771240234, "learning_rate": 6.775794698489809e-06, - "loss": 0.1906, + "loss": 0.7215, "step": 62610 }, { "epoch": 6.61, - "grad_norm": 7.261459827423096, + "grad_norm": 62.758949279785156, "learning_rate": 6.773682543035168e-06, - "loss": 0.5802, + "loss": 1.0798, "step": 62620 }, { "epoch": 6.61, - "grad_norm": 27.058252334594727, + "grad_norm": 49.27354431152344, "learning_rate": 6.771570387580526e-06, - "loss": 0.5618, + "loss": 1.0223, "step": 62630 }, { "epoch": 6.62, - "grad_norm": 15.319981575012207, + "grad_norm": 42.79924011230469, "learning_rate": 6.7694582321258854e-06, - "loss": 0.8218, + "loss": 1.1703, "step": 62640 }, { "epoch": 6.62, - "grad_norm": 14.407098770141602, + "grad_norm": 49.917537689208984, "learning_rate": 6.767346076671244e-06, - "loss": 0.4871, + "loss": 0.8137, "step": 62650 }, { "epoch": 6.62, - "grad_norm": 11.334388732910156, + "grad_norm": 35.651206970214844, "learning_rate": 6.765233921216602e-06, - "loss": 0.1331, + "loss": 0.8511, "step": 62660 }, { "epoch": 6.62, - "grad_norm": 0.7395680546760559, + "grad_norm": 33.49251937866211, "learning_rate": 6.763121765761961e-06, - "loss": 0.3565, + "loss": 1.0184, "step": 62670 }, { "epoch": 6.62, - "grad_norm": 7.154023170471191, + "grad_norm": 13.284966468811035, "learning_rate": 6.76100961030732e-06, - "loss": 0.2457, + "loss": 0.6587, "step": 62680 }, { "epoch": 6.62, - "grad_norm": 7.790151596069336, + "grad_norm": 31.753507614135742, "learning_rate": 6.758897454852678e-06, - "loss": 0.6823, + "loss": 1.2881, "step": 62690 }, { "epoch": 6.62, - "grad_norm": 0.1718451827764511, + "grad_norm": 16.005334854125977, "learning_rate": 6.7567852993980364e-06, - "loss": 0.2834, + "loss": 0.8012, "step": 62700 }, { "epoch": 6.62, - "grad_norm": 19.46021842956543, + "grad_norm": 45.34080505371094, "learning_rate": 6.754673143943394e-06, - "loss": 0.4691, + "loss": 0.9933, "step": 62710 }, { "epoch": 6.62, - "grad_norm": 12.017789840698242, + "grad_norm": 26.97416877746582, "learning_rate": 6.752560988488753e-06, - "loss": 0.3176, + "loss": 1.1477, "step": 62720 }, { "epoch": 6.62, - "grad_norm": 2.7917239665985107, + "grad_norm": 26.075654983520508, "learning_rate": 6.750448833034112e-06, - "loss": 0.3969, + "loss": 0.8447, "step": 62730 }, { "epoch": 6.63, - "grad_norm": 8.85714054107666, + "grad_norm": 48.98855209350586, "learning_rate": 6.74833667757947e-06, - "loss": 0.3713, + "loss": 0.8429, "step": 62740 }, { "epoch": 6.63, - "grad_norm": 19.675485610961914, + "grad_norm": 36.18333435058594, "learning_rate": 6.746224522124829e-06, - "loss": 0.2861, + "loss": 0.6561, "step": 62750 }, { "epoch": 6.63, - "grad_norm": 9.51939868927002, + "grad_norm": 37.980140686035156, "learning_rate": 6.744112366670188e-06, - "loss": 0.4246, + "loss": 0.7619, "step": 62760 }, { "epoch": 6.63, - "grad_norm": 46.412776947021484, + "grad_norm": 15.11555290222168, "learning_rate": 6.742000211215546e-06, - "loss": 0.3555, + "loss": 0.5794, "step": 62770 }, { "epoch": 6.63, - "grad_norm": 18.151796340942383, + "grad_norm": 29.9455623626709, "learning_rate": 6.739888055760905e-06, - "loss": 0.3693, + "loss": 0.8733, "step": 62780 }, { "epoch": 6.63, - "grad_norm": 20.688596725463867, + "grad_norm": 53.611351013183594, "learning_rate": 6.737775900306263e-06, - "loss": 0.3473, + "loss": 0.9354, "step": 62790 }, { "epoch": 6.63, - "grad_norm": 0.08775978535413742, + "grad_norm": 18.490652084350586, "learning_rate": 6.735663744851622e-06, - "loss": 0.3741, + "loss": 0.6869, "step": 62800 }, { "epoch": 6.63, - "grad_norm": 0.7646040320396423, + "grad_norm": 15.454607009887695, "learning_rate": 6.7335515893969805e-06, - "loss": 0.2107, + "loss": 0.8583, "step": 62810 }, { "epoch": 6.63, - "grad_norm": 27.8598575592041, + "grad_norm": 54.687530517578125, "learning_rate": 6.731439433942338e-06, - "loss": 0.3646, + "loss": 0.9502, "step": 62820 }, { "epoch": 6.64, - "grad_norm": 0.28154847025871277, + "grad_norm": 44.962059020996094, "learning_rate": 6.729327278487697e-06, - "loss": 0.37, + "loss": 0.9092, "step": 62830 }, { "epoch": 6.64, - "grad_norm": 21.03333282470703, + "grad_norm": 43.705650329589844, "learning_rate": 6.727215123033056e-06, - "loss": 0.291, + "loss": 0.7558, "step": 62840 }, { "epoch": 6.64, - "grad_norm": 13.644400596618652, + "grad_norm": 62.2644157409668, "learning_rate": 6.725102967578414e-06, - "loss": 0.2567, + "loss": 0.9174, "step": 62850 }, { "epoch": 6.64, - "grad_norm": 0.16078205406665802, + "grad_norm": 20.219614028930664, "learning_rate": 6.722990812123773e-06, - "loss": 0.4512, + "loss": 0.7878, "step": 62860 }, { "epoch": 6.64, - "grad_norm": 2.540505886077881, + "grad_norm": 26.715238571166992, "learning_rate": 6.720878656669131e-06, - "loss": 0.2724, + "loss": 0.8832, "step": 62870 }, { "epoch": 6.64, - "grad_norm": 2.803539991378784, + "grad_norm": 39.21358871459961, "learning_rate": 6.71876650121449e-06, - "loss": 0.5695, + "loss": 1.0931, "step": 62880 }, { "epoch": 6.64, - "grad_norm": 12.324688911437988, + "grad_norm": 51.34596252441406, "learning_rate": 6.716654345759849e-06, - "loss": 0.2962, + "loss": 1.0969, "step": 62890 }, { "epoch": 6.64, - "grad_norm": 16.39766502380371, + "grad_norm": 33.378562927246094, "learning_rate": 6.714542190305207e-06, - "loss": 0.1857, + "loss": 0.681, "step": 62900 }, { "epoch": 6.64, - "grad_norm": 0.48016199469566345, + "grad_norm": 15.424938201904297, "learning_rate": 6.712430034850566e-06, - "loss": 0.2076, + "loss": 1.0045, "step": 62910 }, { "epoch": 6.64, - "grad_norm": 14.632326126098633, + "grad_norm": 41.23094940185547, "learning_rate": 6.7103178793959245e-06, - "loss": 0.385, + "loss": 0.8461, "step": 62920 }, { "epoch": 6.65, - "grad_norm": 0.13910016417503357, + "grad_norm": 30.58301544189453, "learning_rate": 6.708205723941282e-06, - "loss": 0.2923, + "loss": 0.8334, "step": 62930 }, { "epoch": 6.65, - "grad_norm": 7.898691654205322, + "grad_norm": 52.23564147949219, "learning_rate": 6.706093568486641e-06, - "loss": 0.2189, + "loss": 0.8204, "step": 62940 }, { "epoch": 6.65, - "grad_norm": 10.954451560974121, + "grad_norm": 44.80339050292969, "learning_rate": 6.703981413032e-06, - "loss": 0.4326, + "loss": 0.7146, "step": 62950 }, { "epoch": 6.65, - "grad_norm": 17.370054244995117, + "grad_norm": 50.25423812866211, "learning_rate": 6.701869257577358e-06, - "loss": 0.4342, + "loss": 1.094, "step": 62960 }, { "epoch": 6.65, - "grad_norm": 23.993946075439453, + "grad_norm": 56.18830871582031, "learning_rate": 6.699757102122717e-06, - "loss": 0.4752, + "loss": 0.9828, "step": 62970 }, { "epoch": 6.65, - "grad_norm": 2.014691114425659, + "grad_norm": 18.26917839050293, "learning_rate": 6.697644946668075e-06, - "loss": 0.3261, + "loss": 1.0824, "step": 62980 }, { "epoch": 6.65, - "grad_norm": 13.266195297241211, + "grad_norm": 53.71892547607422, "learning_rate": 6.695532791213433e-06, - "loss": 0.7197, + "loss": 1.0915, "step": 62990 }, { "epoch": 6.65, - "grad_norm": 17.512922286987305, + "grad_norm": 37.84865188598633, "learning_rate": 6.693420635758793e-06, - "loss": 0.415, + "loss": 0.8708, "step": 63000 }, { "epoch": 6.65, - "grad_norm": 13.839255332946777, + "grad_norm": 36.502323150634766, "learning_rate": 6.691308480304151e-06, - "loss": 0.5929, + "loss": 0.991, "step": 63010 }, { "epoch": 6.66, - "grad_norm": 0.5215139985084534, + "grad_norm": 36.95267868041992, "learning_rate": 6.68919632484951e-06, - "loss": 0.2191, + "loss": 0.9538, "step": 63020 }, { "epoch": 6.66, - "grad_norm": 6.981299877166748, + "grad_norm": 53.2202033996582, "learning_rate": 6.6870841693948685e-06, - "loss": 0.5859, + "loss": 1.1361, "step": 63030 }, { "epoch": 6.66, - "grad_norm": 26.132944107055664, + "grad_norm": 39.24366760253906, "learning_rate": 6.6849720139402264e-06, - "loss": 0.3084, + "loss": 0.8703, "step": 63040 }, { "epoch": 6.66, - "grad_norm": 20.196603775024414, + "grad_norm": 44.896522521972656, "learning_rate": 6.682859858485585e-06, - "loss": 0.4839, + "loss": 0.9685, "step": 63050 }, { "epoch": 6.66, - "grad_norm": 3.6318681240081787, + "grad_norm": 40.32084655761719, "learning_rate": 6.680747703030943e-06, - "loss": 0.3197, + "loss": 0.8907, "step": 63060 }, { "epoch": 6.66, - "grad_norm": 0.021648608148097992, + "grad_norm": 27.44644546508789, "learning_rate": 6.678635547576302e-06, - "loss": 0.3896, + "loss": 0.9154, "step": 63070 }, { "epoch": 6.66, - "grad_norm": 0.1581491380929947, + "grad_norm": 44.73949432373047, "learning_rate": 6.676523392121661e-06, - "loss": 0.2296, + "loss": 0.9719, "step": 63080 }, { "epoch": 6.66, - "grad_norm": 1.0024337768554688, + "grad_norm": 45.815853118896484, "learning_rate": 6.674411236667019e-06, - "loss": 0.2598, + "loss": 0.9926, "step": 63090 }, { "epoch": 6.66, - "grad_norm": 5.935975551605225, + "grad_norm": 36.673614501953125, "learning_rate": 6.672299081212377e-06, - "loss": 0.3606, + "loss": 1.1539, "step": 63100 }, { "epoch": 6.66, - "grad_norm": 8.71825122833252, + "grad_norm": 31.235549926757812, "learning_rate": 6.670186925757737e-06, - "loss": 0.4058, + "loss": 0.6861, "step": 63110 }, { "epoch": 6.67, - "grad_norm": 0.1127292737364769, + "grad_norm": 24.45794105529785, "learning_rate": 6.668074770303095e-06, - "loss": 0.415, + "loss": 1.03, "step": 63120 }, { "epoch": 6.67, - "grad_norm": 12.78404426574707, + "grad_norm": 53.204341888427734, "learning_rate": 6.665962614848454e-06, - "loss": 0.5143, + "loss": 1.054, "step": 63130 }, { "epoch": 6.67, - "grad_norm": 0.30527690052986145, + "grad_norm": 37.34069061279297, "learning_rate": 6.6638504593938125e-06, - "loss": 0.3598, + "loss": 0.9563, "step": 63140 }, { "epoch": 6.67, - "grad_norm": 18.093212127685547, + "grad_norm": 43.7931022644043, "learning_rate": 6.6617383039391705e-06, - "loss": 0.635, + "loss": 0.9633, "step": 63150 }, { "epoch": 6.67, - "grad_norm": 46.77223205566406, + "grad_norm": 40.33416748046875, "learning_rate": 6.659626148484529e-06, - "loss": 0.3799, + "loss": 1.0572, "step": 63160 }, { "epoch": 6.67, - "grad_norm": 22.947891235351562, + "grad_norm": 32.610652923583984, "learning_rate": 6.657513993029887e-06, - "loss": 0.6373, + "loss": 1.0592, "step": 63170 }, { "epoch": 6.67, - "grad_norm": 1.0983197689056396, + "grad_norm": 27.655078887939453, "learning_rate": 6.655401837575246e-06, - "loss": 0.1871, + "loss": 0.8565, "step": 63180 }, { "epoch": 6.67, - "grad_norm": 49.587703704833984, + "grad_norm": 36.898887634277344, "learning_rate": 6.653289682120605e-06, - "loss": 0.4413, + "loss": 0.7925, "step": 63190 }, { "epoch": 6.67, - "grad_norm": 29.496135711669922, + "grad_norm": 7.07143497467041, "learning_rate": 6.651177526665963e-06, - "loss": 0.4932, + "loss": 0.8592, "step": 63200 }, { "epoch": 6.68, - "grad_norm": 13.110641479492188, + "grad_norm": 35.47466278076172, "learning_rate": 6.6490653712113214e-06, - "loss": 0.3781, + "loss": 0.9197, "step": 63210 }, { "epoch": 6.68, - "grad_norm": 8.135794639587402, + "grad_norm": 33.550289154052734, "learning_rate": 6.646953215756681e-06, - "loss": 0.4707, + "loss": 1.0356, "step": 63220 }, { "epoch": 6.68, - "grad_norm": 18.2056884765625, + "grad_norm": 41.67759323120117, "learning_rate": 6.644841060302038e-06, - "loss": 0.6819, + "loss": 1.1674, "step": 63230 }, { "epoch": 6.68, - "grad_norm": 27.201824188232422, + "grad_norm": 59.5691032409668, "learning_rate": 6.642728904847398e-06, - "loss": 0.3865, + "loss": 1.1638, "step": 63240 }, { "epoch": 6.68, - "grad_norm": 0.9442319273948669, + "grad_norm": 36.904296875, "learning_rate": 6.640616749392756e-06, - "loss": 0.4347, + "loss": 0.9187, "step": 63250 }, { "epoch": 6.68, - "grad_norm": 46.67665481567383, + "grad_norm": 46.66673278808594, "learning_rate": 6.6385045939381145e-06, - "loss": 0.5845, + "loss": 0.7793, "step": 63260 }, { "epoch": 6.68, - "grad_norm": 8.288643836975098, + "grad_norm": 69.13488006591797, "learning_rate": 6.636392438483473e-06, - "loss": 0.2422, + "loss": 0.953, "step": 63270 }, { "epoch": 6.68, - "grad_norm": 10.403959274291992, + "grad_norm": 69.97476196289062, "learning_rate": 6.634280283028831e-06, - "loss": 0.1964, + "loss": 0.9123, "step": 63280 }, { "epoch": 6.68, - "grad_norm": 15.27939224243164, + "grad_norm": 48.96623229980469, "learning_rate": 6.63216812757419e-06, - "loss": 0.2539, + "loss": 1.0488, "step": 63290 }, { "epoch": 6.68, - "grad_norm": 1.505622386932373, + "grad_norm": 37.06908416748047, "learning_rate": 6.630055972119549e-06, - "loss": 0.1634, + "loss": 0.9356, "step": 63300 }, { "epoch": 6.69, - "grad_norm": 5.378640651702881, + "grad_norm": 31.532995223999023, "learning_rate": 6.627943816664907e-06, - "loss": 0.3156, + "loss": 1.2885, "step": 63310 }, { "epoch": 6.69, - "grad_norm": 20.031705856323242, + "grad_norm": 29.778654098510742, "learning_rate": 6.6258316612102655e-06, - "loss": 0.4216, + "loss": 0.5399, "step": 63320 }, { "epoch": 6.69, - "grad_norm": 24.316020965576172, + "grad_norm": 38.54539489746094, "learning_rate": 6.623719505755623e-06, - "loss": 0.3732, + "loss": 0.6902, "step": 63330 }, { "epoch": 6.69, - "grad_norm": 6.633448600769043, + "grad_norm": 25.866310119628906, "learning_rate": 6.621607350300982e-06, - "loss": 0.4876, + "loss": 0.6759, "step": 63340 }, { "epoch": 6.69, - "grad_norm": 11.663553237915039, + "grad_norm": 56.09543991088867, "learning_rate": 6.619495194846342e-06, - "loss": 0.3337, + "loss": 0.8879, "step": 63350 }, { "epoch": 6.69, - "grad_norm": 27.995698928833008, + "grad_norm": 39.637245178222656, "learning_rate": 6.6173830393917e-06, - "loss": 0.4379, + "loss": 0.9533, "step": 63360 }, { "epoch": 6.69, - "grad_norm": 15.953140258789062, + "grad_norm": 53.911407470703125, "learning_rate": 6.6152708839370585e-06, - "loss": 0.3596, + "loss": 0.9595, "step": 63370 }, { "epoch": 6.69, - "grad_norm": 29.137107849121094, + "grad_norm": 21.675291061401367, "learning_rate": 6.613158728482417e-06, - "loss": 0.5268, + "loss": 1.1657, "step": 63380 }, { "epoch": 6.69, - "grad_norm": 11.078262329101562, + "grad_norm": 15.403060913085938, "learning_rate": 6.611046573027775e-06, - "loss": 0.3444, + "loss": 1.0495, "step": 63390 }, { "epoch": 6.7, - "grad_norm": 6.419632434844971, + "grad_norm": 17.30608558654785, "learning_rate": 6.608934417573134e-06, - "loss": 0.5028, + "loss": 1.0435, "step": 63400 }, { "epoch": 6.7, - "grad_norm": 17.280288696289062, + "grad_norm": 51.03022766113281, "learning_rate": 6.606822262118493e-06, - "loss": 0.2958, + "loss": 0.6216, "step": 63410 }, { "epoch": 6.7, - "grad_norm": 28.836692810058594, + "grad_norm": 37.948123931884766, "learning_rate": 6.604710106663851e-06, - "loss": 0.2491, + "loss": 0.9595, "step": 63420 }, { "epoch": 6.7, - "grad_norm": 23.721439361572266, + "grad_norm": 49.17271041870117, "learning_rate": 6.6025979512092095e-06, - "loss": 0.6115, + "loss": 0.918, "step": 63430 }, { "epoch": 6.7, - "grad_norm": 7.266289234161377, + "grad_norm": 31.61956024169922, "learning_rate": 6.6004857957545674e-06, - "loss": 0.2851, + "loss": 0.7271, "step": 63440 }, { "epoch": 6.7, - "grad_norm": 9.274873733520508, + "grad_norm": 39.58841323852539, "learning_rate": 6.598373640299926e-06, - "loss": 0.3677, + "loss": 0.9094, "step": 63450 }, { "epoch": 6.7, - "grad_norm": 3.4232594966888428, + "grad_norm": 33.56834030151367, "learning_rate": 6.596261484845286e-06, - "loss": 0.2779, + "loss": 0.5502, "step": 63460 }, { "epoch": 6.7, - "grad_norm": 28.432212829589844, + "grad_norm": 40.252647399902344, "learning_rate": 6.594149329390643e-06, - "loss": 0.4225, + "loss": 1.0354, "step": 63470 }, { "epoch": 6.7, - "grad_norm": 1.614375114440918, + "grad_norm": 30.649124145507812, "learning_rate": 6.5920371739360025e-06, - "loss": 0.3599, + "loss": 1.0073, "step": 63480 }, { "epoch": 6.71, - "grad_norm": 8.567790031433105, + "grad_norm": 24.84274673461914, "learning_rate": 6.589925018481361e-06, - "loss": 0.38, + "loss": 0.9672, "step": 63490 }, { "epoch": 6.71, - "grad_norm": 11.2389554977417, + "grad_norm": 29.33965301513672, "learning_rate": 6.587812863026719e-06, - "loss": 0.3904, + "loss": 0.9367, "step": 63500 }, { "epoch": 6.71, - "grad_norm": 9.947673797607422, + "grad_norm": 55.95411682128906, "learning_rate": 6.585700707572078e-06, - "loss": 0.2825, + "loss": 0.8298, "step": 63510 }, { "epoch": 6.71, - "grad_norm": 20.274093627929688, + "grad_norm": 62.9566764831543, "learning_rate": 6.583588552117436e-06, - "loss": 0.2851, + "loss": 0.9844, "step": 63520 }, { "epoch": 6.71, - "grad_norm": 0.29831603169441223, + "grad_norm": 48.783843994140625, "learning_rate": 6.581476396662795e-06, - "loss": 0.6606, + "loss": 1.2925, "step": 63530 }, { "epoch": 6.71, - "grad_norm": 9.604904174804688, + "grad_norm": 41.79275894165039, "learning_rate": 6.5793642412081535e-06, - "loss": 0.2858, + "loss": 1.0385, "step": 63540 }, { "epoch": 6.71, - "grad_norm": 21.51850128173828, + "grad_norm": 18.773942947387695, "learning_rate": 6.5772520857535115e-06, - "loss": 0.3614, + "loss": 0.8154, "step": 63550 }, { "epoch": 6.71, - "grad_norm": 10.953959465026855, + "grad_norm": 43.77490234375, "learning_rate": 6.57513993029887e-06, - "loss": 0.7164, + "loss": 0.9441, "step": 63560 }, { "epoch": 6.71, - "grad_norm": 20.20395278930664, + "grad_norm": 49.98035430908203, "learning_rate": 6.57302777484423e-06, - "loss": 0.1412, + "loss": 0.7559, "step": 63570 }, { "epoch": 6.71, - "grad_norm": 14.580379486083984, + "grad_norm": 47.191627502441406, "learning_rate": 6.570915619389587e-06, - "loss": 0.5266, + "loss": 1.1719, "step": 63580 }, { "epoch": 6.72, - "grad_norm": 12.509644508361816, + "grad_norm": 45.711761474609375, "learning_rate": 6.5688034639349466e-06, - "loss": 0.5336, + "loss": 1.1303, "step": 63590 }, { "epoch": 6.72, - "grad_norm": 30.18849754333496, + "grad_norm": 47.84355163574219, "learning_rate": 6.566691308480305e-06, - "loss": 0.4885, + "loss": 1.1789, "step": 63600 }, { "epoch": 6.72, - "grad_norm": 8.261421203613281, + "grad_norm": 46.202945709228516, "learning_rate": 6.564579153025663e-06, - "loss": 0.2293, + "loss": 0.7707, "step": 63610 }, { "epoch": 6.72, - "grad_norm": 31.61208152770996, + "grad_norm": 40.403568267822266, "learning_rate": 6.562466997571022e-06, - "loss": 0.448, + "loss": 0.9434, "step": 63620 }, { "epoch": 6.72, - "grad_norm": 9.557683944702148, + "grad_norm": 43.26679992675781, "learning_rate": 6.56035484211638e-06, - "loss": 0.4013, + "loss": 1.0358, "step": 63630 }, { "epoch": 6.72, - "grad_norm": 14.078304290771484, + "grad_norm": 24.77516746520996, "learning_rate": 6.558242686661739e-06, - "loss": 0.413, + "loss": 1.0233, "step": 63640 }, { "epoch": 6.72, - "grad_norm": 15.747143745422363, + "grad_norm": 63.514949798583984, "learning_rate": 6.5561305312070975e-06, - "loss": 0.3204, + "loss": 0.9576, "step": 63650 }, { "epoch": 6.72, - "grad_norm": 0.13503630459308624, + "grad_norm": 36.0797004699707, "learning_rate": 6.5540183757524555e-06, - "loss": 0.3653, + "loss": 0.6952, "step": 63660 }, { "epoch": 6.72, - "grad_norm": 16.907543182373047, + "grad_norm": 33.82392501831055, "learning_rate": 6.551906220297814e-06, - "loss": 0.4141, + "loss": 0.8815, "step": 63670 }, { "epoch": 6.73, - "grad_norm": 10.651500701904297, + "grad_norm": 49.30696105957031, "learning_rate": 6.549794064843174e-06, - "loss": 0.2676, + "loss": 0.9902, "step": 63680 }, { "epoch": 6.73, - "grad_norm": 4.71936559677124, + "grad_norm": 40.46494674682617, "learning_rate": 6.547681909388531e-06, - "loss": 0.3843, + "loss": 1.0011, "step": 63690 }, { "epoch": 6.73, - "grad_norm": 0.4369676113128662, + "grad_norm": 28.91827964782715, "learning_rate": 6.545569753933891e-06, - "loss": 0.61, + "loss": 1.2364, "step": 63700 }, { "epoch": 6.73, - "grad_norm": 23.67111587524414, + "grad_norm": 39.5838623046875, "learning_rate": 6.5434575984792485e-06, - "loss": 0.3537, + "loss": 1.2188, "step": 63710 }, { "epoch": 6.73, - "grad_norm": 15.416050910949707, + "grad_norm": 31.108041763305664, "learning_rate": 6.541345443024607e-06, - "loss": 0.3134, + "loss": 0.9629, "step": 63720 }, { "epoch": 6.73, - "grad_norm": 13.13318920135498, + "grad_norm": 36.42369842529297, "learning_rate": 6.539233287569966e-06, - "loss": 0.1874, + "loss": 0.8804, "step": 63730 }, { "epoch": 6.73, - "grad_norm": 0.20450274646282196, + "grad_norm": 18.772722244262695, "learning_rate": 6.537121132115324e-06, - "loss": 0.2701, + "loss": 1.1585, "step": 63740 }, { "epoch": 6.73, - "grad_norm": 2.092188835144043, + "grad_norm": 52.619590759277344, "learning_rate": 6.535008976660683e-06, - "loss": 0.4248, + "loss": 1.0076, "step": 63750 }, { "epoch": 6.73, - "grad_norm": 14.922277450561523, + "grad_norm": 36.68226623535156, "learning_rate": 6.5328968212060416e-06, - "loss": 0.4195, + "loss": 0.7206, "step": 63760 }, { "epoch": 6.73, - "grad_norm": 26.69235610961914, + "grad_norm": 38.29536819458008, "learning_rate": 6.5307846657513995e-06, - "loss": 0.5752, + "loss": 0.955, "step": 63770 }, { "epoch": 6.74, - "grad_norm": 19.440048217773438, + "grad_norm": 43.251914978027344, "learning_rate": 6.528672510296758e-06, - "loss": 0.3339, + "loss": 0.9229, "step": 63780 }, { "epoch": 6.74, - "grad_norm": 12.765205383300781, + "grad_norm": 27.358680725097656, "learning_rate": 6.526560354842116e-06, - "loss": 0.3264, + "loss": 0.6693, "step": 63790 }, { "epoch": 6.74, - "grad_norm": 25.22090721130371, + "grad_norm": 27.32120132446289, "learning_rate": 6.524448199387475e-06, - "loss": 0.5525, + "loss": 0.964, "step": 63800 }, { "epoch": 6.74, - "grad_norm": 6.116117477416992, + "grad_norm": 27.41160774230957, "learning_rate": 6.522336043932835e-06, - "loss": 0.3541, + "loss": 0.9099, "step": 63810 }, { "epoch": 6.74, - "grad_norm": 0.018236031755805016, + "grad_norm": 32.9157829284668, "learning_rate": 6.520223888478192e-06, - "loss": 0.4525, + "loss": 1.081, "step": 63820 }, { "epoch": 6.74, - "grad_norm": 28.22924041748047, + "grad_norm": 57.51820755004883, "learning_rate": 6.518111733023551e-06, - "loss": 0.677, + "loss": 0.9087, "step": 63830 }, { "epoch": 6.74, - "grad_norm": 0.22580543160438538, + "grad_norm": 38.30231475830078, "learning_rate": 6.51599957756891e-06, - "loss": 0.4006, + "loss": 0.5808, "step": 63840 }, { "epoch": 6.74, - "grad_norm": 4.969663143157959, + "grad_norm": 33.61245346069336, "learning_rate": 6.513887422114268e-06, - "loss": 0.5977, + "loss": 1.0628, "step": 63850 }, { "epoch": 6.74, - "grad_norm": 0.6682610511779785, + "grad_norm": 31.634241104125977, "learning_rate": 6.511775266659627e-06, - "loss": 0.3648, + "loss": 0.7473, "step": 63860 }, { "epoch": 6.75, - "grad_norm": 45.95930099487305, + "grad_norm": 60.16029357910156, "learning_rate": 6.509663111204986e-06, - "loss": 0.5925, + "loss": 1.2736, "step": 63870 }, { "epoch": 6.75, - "grad_norm": 28.791288375854492, + "grad_norm": 37.3236083984375, "learning_rate": 6.5075509557503435e-06, - "loss": 0.448, + "loss": 0.9052, "step": 63880 }, { "epoch": 6.75, - "grad_norm": 15.27834701538086, + "grad_norm": 35.42621612548828, "learning_rate": 6.505438800295702e-06, - "loss": 0.3559, + "loss": 1.0209, "step": 63890 }, { "epoch": 6.75, - "grad_norm": 2.9003469944000244, + "grad_norm": 47.11933135986328, "learning_rate": 6.50332664484106e-06, - "loss": 0.3272, + "loss": 0.8226, "step": 63900 }, { "epoch": 6.75, - "grad_norm": 18.48822021484375, + "grad_norm": 48.788726806640625, "learning_rate": 6.501214489386419e-06, - "loss": 0.5271, + "loss": 1.1842, "step": 63910 }, { "epoch": 6.75, - "grad_norm": 11.985468864440918, + "grad_norm": 26.54878807067871, "learning_rate": 6.499102333931779e-06, - "loss": 0.3365, + "loss": 0.7744, "step": 63920 }, { "epoch": 6.75, - "grad_norm": 22.039934158325195, + "grad_norm": 28.531333923339844, "learning_rate": 6.496990178477136e-06, - "loss": 0.1989, + "loss": 0.7345, "step": 63930 }, { "epoch": 6.75, - "grad_norm": 17.135169982910156, + "grad_norm": 56.47797393798828, "learning_rate": 6.494878023022495e-06, - "loss": 0.3663, + "loss": 0.7133, "step": 63940 }, { "epoch": 6.75, - "grad_norm": 8.096076011657715, + "grad_norm": 31.072383880615234, "learning_rate": 6.492765867567854e-06, - "loss": 0.3469, + "loss": 1.0947, "step": 63950 }, { "epoch": 6.75, - "grad_norm": 12.366827011108398, + "grad_norm": 35.596736907958984, "learning_rate": 6.490653712113212e-06, - "loss": 0.5023, + "loss": 1.4322, "step": 63960 }, { "epoch": 6.76, - "grad_norm": 1.273877501487732, + "grad_norm": 36.82416534423828, "learning_rate": 6.488541556658571e-06, - "loss": 0.3071, + "loss": 1.0868, "step": 63970 }, { "epoch": 6.76, - "grad_norm": 21.501821517944336, + "grad_norm": 30.015850067138672, "learning_rate": 6.486429401203929e-06, - "loss": 0.3547, + "loss": 1.1889, "step": 63980 }, { "epoch": 6.76, - "grad_norm": 4.231912136077881, + "grad_norm": 38.911983489990234, "learning_rate": 6.4843172457492876e-06, - "loss": 0.244, + "loss": 0.7709, "step": 63990 }, { "epoch": 6.76, - "grad_norm": 15.947311401367188, + "grad_norm": 42.04343032836914, "learning_rate": 6.482205090294646e-06, - "loss": 0.4113, + "loss": 0.8451, "step": 64000 }, { "epoch": 6.76, - "grad_norm": 4.177820205688477, + "grad_norm": 59.01166534423828, "learning_rate": 6.480092934840004e-06, - "loss": 0.4595, + "loss": 1.3575, "step": 64010 }, { "epoch": 6.76, - "grad_norm": 0.08074556291103363, + "grad_norm": 35.21895980834961, "learning_rate": 6.477980779385363e-06, - "loss": 0.3843, + "loss": 0.6189, "step": 64020 }, { "epoch": 6.76, - "grad_norm": 2.342316150665283, + "grad_norm": 47.03562927246094, "learning_rate": 6.475868623930722e-06, - "loss": 0.3031, + "loss": 0.8448, "step": 64030 }, { "epoch": 6.76, - "grad_norm": 15.46596908569336, + "grad_norm": 53.82585525512695, "learning_rate": 6.47375646847608e-06, - "loss": 0.4686, + "loss": 0.8631, "step": 64040 }, { "epoch": 6.76, - "grad_norm": 21.992277145385742, + "grad_norm": 57.730552673339844, "learning_rate": 6.471644313021439e-06, - "loss": 0.4491, + "loss": 0.981, "step": 64050 }, { "epoch": 6.77, - "grad_norm": 8.56957721710205, + "grad_norm": 13.65185546875, "learning_rate": 6.469532157566798e-06, - "loss": 0.5067, + "loss": 0.8202, "step": 64060 }, { "epoch": 6.77, - "grad_norm": 31.642005920410156, + "grad_norm": 63.72688293457031, "learning_rate": 6.467420002112156e-06, - "loss": 0.4296, + "loss": 0.8567, "step": 64070 }, { "epoch": 6.77, - "grad_norm": 8.507129669189453, + "grad_norm": 53.18495178222656, "learning_rate": 6.465307846657515e-06, - "loss": 0.2798, + "loss": 0.8844, "step": 64080 }, { "epoch": 6.77, - "grad_norm": 20.77555274963379, + "grad_norm": 50.041473388671875, "learning_rate": 6.463195691202873e-06, - "loss": 0.3475, + "loss": 0.8583, "step": 64090 }, { "epoch": 6.77, - "grad_norm": 15.283977508544922, + "grad_norm": 35.1582145690918, "learning_rate": 6.461083535748232e-06, - "loss": 0.3953, + "loss": 0.7948, "step": 64100 }, { "epoch": 6.77, - "grad_norm": 45.76782989501953, + "grad_norm": 55.30855941772461, "learning_rate": 6.45897138029359e-06, - "loss": 0.4899, + "loss": 0.9123, "step": 64110 }, { "epoch": 6.77, - "grad_norm": 5.905948162078857, + "grad_norm": 52.092430114746094, "learning_rate": 6.456859224838948e-06, - "loss": 0.4235, + "loss": 1.0081, "step": 64120 }, { "epoch": 6.77, - "grad_norm": 35.59563064575195, + "grad_norm": 46.54584503173828, "learning_rate": 6.454747069384307e-06, - "loss": 0.4908, + "loss": 1.1284, "step": 64130 }, { "epoch": 6.77, - "grad_norm": 16.53948402404785, + "grad_norm": 37.016273498535156, "learning_rate": 6.452634913929666e-06, - "loss": 0.3799, + "loss": 0.9947, "step": 64140 }, { "epoch": 6.77, - "grad_norm": 14.703980445861816, + "grad_norm": 28.910154342651367, "learning_rate": 6.450522758475024e-06, - "loss": 0.4051, + "loss": 1.1173, "step": 64150 }, { "epoch": 6.78, - "grad_norm": 29.382930755615234, + "grad_norm": 55.65020751953125, "learning_rate": 6.448410603020383e-06, - "loss": 0.3587, + "loss": 0.9863, "step": 64160 }, { "epoch": 6.78, - "grad_norm": 32.80571746826172, + "grad_norm": 29.226003646850586, "learning_rate": 6.4462984475657405e-06, - "loss": 0.5018, + "loss": 1.0832, "step": 64170 }, { "epoch": 6.78, - "grad_norm": 19.634870529174805, + "grad_norm": 44.29539108276367, "learning_rate": 6.4441862921111e-06, - "loss": 0.3158, + "loss": 1.1053, "step": 64180 }, { "epoch": 6.78, - "grad_norm": 18.95157814025879, + "grad_norm": 61.921104431152344, "learning_rate": 6.442074136656459e-06, - "loss": 0.5449, + "loss": 0.9762, "step": 64190 }, { "epoch": 6.78, - "grad_norm": 17.055469512939453, + "grad_norm": 32.654605865478516, "learning_rate": 6.439961981201817e-06, - "loss": 0.2751, + "loss": 0.5923, "step": 64200 }, { "epoch": 6.78, - "grad_norm": 1.7023260593414307, + "grad_norm": 29.76256561279297, "learning_rate": 6.437849825747176e-06, - "loss": 0.3384, + "loss": 0.5985, "step": 64210 }, { "epoch": 6.78, - "grad_norm": 54.2784423828125, + "grad_norm": 60.25014877319336, "learning_rate": 6.435737670292534e-06, - "loss": 0.5975, + "loss": 1.3112, "step": 64220 }, { "epoch": 6.78, - "grad_norm": 23.962785720825195, + "grad_norm": 38.92093276977539, "learning_rate": 6.433625514837892e-06, - "loss": 0.4448, + "loss": 0.9943, "step": 64230 }, { "epoch": 6.78, - "grad_norm": 3.6871564388275146, + "grad_norm": 44.69923782348633, "learning_rate": 6.431513359383251e-06, - "loss": 0.6597, + "loss": 1.2962, "step": 64240 }, { "epoch": 6.79, - "grad_norm": 0.1298469454050064, + "grad_norm": 49.620765686035156, "learning_rate": 6.42940120392861e-06, - "loss": 0.2006, + "loss": 0.8612, "step": 64250 }, { "epoch": 6.79, - "grad_norm": 1.2721490859985352, + "grad_norm": 45.01993942260742, "learning_rate": 6.427289048473968e-06, - "loss": 0.3792, + "loss": 0.737, "step": 64260 }, { "epoch": 6.79, - "grad_norm": 5.316025733947754, + "grad_norm": 31.927080154418945, "learning_rate": 6.425176893019327e-06, - "loss": 0.5177, + "loss": 0.9462, "step": 64270 }, { "epoch": 6.79, - "grad_norm": 27.621532440185547, + "grad_norm": 40.73223876953125, "learning_rate": 6.4230647375646845e-06, - "loss": 0.3856, + "loss": 0.8717, "step": 64280 }, { "epoch": 6.79, - "grad_norm": 20.89000129699707, + "grad_norm": 59.06919479370117, "learning_rate": 6.420952582110044e-06, - "loss": 0.5038, + "loss": 1.115, "step": 64290 }, { "epoch": 6.79, - "grad_norm": 0.22407983243465424, + "grad_norm": 22.22604751586914, "learning_rate": 6.418840426655403e-06, - "loss": 0.7424, + "loss": 0.8071, "step": 64300 }, { "epoch": 6.79, - "grad_norm": 25.995792388916016, + "grad_norm": 54.646148681640625, "learning_rate": 6.416728271200761e-06, - "loss": 0.3855, + "loss": 1.0379, "step": 64310 }, { "epoch": 6.79, - "grad_norm": 10.532631874084473, + "grad_norm": 60.0913200378418, "learning_rate": 6.41461611574612e-06, - "loss": 0.3647, + "loss": 1.2283, "step": 64320 }, { "epoch": 6.79, - "grad_norm": 14.075876235961914, + "grad_norm": 46.94139099121094, "learning_rate": 6.412503960291478e-06, - "loss": 0.4485, + "loss": 0.9171, "step": 64330 }, { "epoch": 6.79, - "grad_norm": 27.536043167114258, + "grad_norm": 45.384273529052734, "learning_rate": 6.410391804836836e-06, - "loss": 0.3284, + "loss": 0.8195, "step": 64340 }, { "epoch": 6.8, - "grad_norm": 0.17814897000789642, + "grad_norm": 30.77508544921875, "learning_rate": 6.408279649382195e-06, - "loss": 0.2917, + "loss": 1.0024, "step": 64350 }, { "epoch": 6.8, - "grad_norm": 14.54658317565918, + "grad_norm": 39.80907440185547, "learning_rate": 6.406167493927553e-06, - "loss": 0.2769, + "loss": 1.2022, "step": 64360 }, { "epoch": 6.8, - "grad_norm": 0.8578535914421082, + "grad_norm": 68.36702728271484, "learning_rate": 6.404055338472912e-06, - "loss": 0.4995, + "loss": 1.243, "step": 64370 }, { "epoch": 6.8, - "grad_norm": 3.2154901027679443, + "grad_norm": 31.11747169494629, "learning_rate": 6.401943183018271e-06, - "loss": 0.3784, + "loss": 1.0116, "step": 64380 }, { "epoch": 6.8, - "grad_norm": 0.2968433201313019, + "grad_norm": 51.54093933105469, "learning_rate": 6.3998310275636285e-06, - "loss": 0.2895, + "loss": 0.6628, "step": 64390 }, { "epoch": 6.8, - "grad_norm": 39.12588119506836, + "grad_norm": 54.89914321899414, "learning_rate": 6.397718872108988e-06, - "loss": 0.6582, + "loss": 0.9553, "step": 64400 }, { "epoch": 6.8, - "grad_norm": 14.082840919494629, + "grad_norm": 24.812177658081055, "learning_rate": 6.395606716654347e-06, - "loss": 0.5138, + "loss": 0.8924, "step": 64410 }, { "epoch": 6.8, - "grad_norm": 13.903997421264648, + "grad_norm": 43.709224700927734, "learning_rate": 6.393494561199705e-06, - "loss": 0.4049, + "loss": 1.0624, "step": 64420 }, { "epoch": 6.8, - "grad_norm": 19.934581756591797, + "grad_norm": 30.487335205078125, "learning_rate": 6.391382405745064e-06, - "loss": 0.7344, + "loss": 1.0774, "step": 64430 }, { "epoch": 6.81, - "grad_norm": 22.080110549926758, + "grad_norm": 33.545936584472656, "learning_rate": 6.389270250290422e-06, - "loss": 0.4503, + "loss": 1.0013, "step": 64440 }, { "epoch": 6.81, - "grad_norm": 7.619664669036865, + "grad_norm": 51.72589874267578, "learning_rate": 6.38715809483578e-06, - "loss": 0.5431, + "loss": 1.0618, "step": 64450 }, { "epoch": 6.81, - "grad_norm": 6.247415065765381, + "grad_norm": 30.550642013549805, "learning_rate": 6.385045939381139e-06, - "loss": 0.3922, + "loss": 0.8303, "step": 64460 }, { "epoch": 6.81, - "grad_norm": 0.38253507018089294, + "grad_norm": 25.35381317138672, "learning_rate": 6.382933783926497e-06, - "loss": 0.3583, + "loss": 0.656, "step": 64470 }, { "epoch": 6.81, - "grad_norm": 17.77663803100586, + "grad_norm": 28.773530960083008, "learning_rate": 6.380821628471856e-06, - "loss": 0.4679, + "loss": 0.9348, "step": 64480 }, { "epoch": 6.81, - "grad_norm": 27.736248016357422, + "grad_norm": 33.904396057128906, "learning_rate": 6.378709473017215e-06, - "loss": 0.2817, + "loss": 1.1173, "step": 64490 }, { "epoch": 6.81, - "grad_norm": 0.15018784999847412, + "grad_norm": 33.01457977294922, "learning_rate": 6.3765973175625726e-06, - "loss": 0.4423, + "loss": 0.872, "step": 64500 }, { "epoch": 6.81, - "grad_norm": 0.06539567559957504, + "grad_norm": 53.251583099365234, "learning_rate": 6.374485162107932e-06, - "loss": 0.3999, + "loss": 1.0624, "step": 64510 }, { "epoch": 6.81, - "grad_norm": 0.14009810984134674, + "grad_norm": 35.39534378051758, "learning_rate": 6.372373006653291e-06, - "loss": 0.5995, + "loss": 0.8063, "step": 64520 }, { "epoch": 6.81, - "grad_norm": 0.19708387553691864, + "grad_norm": 40.98356246948242, "learning_rate": 6.370260851198649e-06, - "loss": 0.5008, + "loss": 1.094, "step": 64530 }, { "epoch": 6.82, - "grad_norm": 42.068214416503906, + "grad_norm": 38.37364959716797, "learning_rate": 6.368148695744008e-06, - "loss": 0.4151, + "loss": 1.1845, "step": 64540 }, { "epoch": 6.82, - "grad_norm": 20.77120590209961, + "grad_norm": 62.924156188964844, "learning_rate": 6.366036540289366e-06, - "loss": 0.2942, + "loss": 1.04, "step": 64550 }, { "epoch": 6.82, - "grad_norm": 25.524354934692383, + "grad_norm": 39.97313690185547, "learning_rate": 6.363924384834724e-06, - "loss": 0.393, + "loss": 1.1107, "step": 64560 }, { "epoch": 6.82, - "grad_norm": 20.396949768066406, + "grad_norm": 39.447147369384766, "learning_rate": 6.361812229380083e-06, - "loss": 0.4153, + "loss": 0.7028, "step": 64570 }, { "epoch": 6.82, - "grad_norm": 15.88100814819336, + "grad_norm": 16.499637603759766, "learning_rate": 6.359700073925441e-06, - "loss": 0.3235, + "loss": 0.8184, "step": 64580 }, { "epoch": 6.82, - "grad_norm": 1.203887939453125, + "grad_norm": 30.9996280670166, "learning_rate": 6.3575879184708e-06, - "loss": 0.3607, + "loss": 0.9572, "step": 64590 }, { "epoch": 6.82, - "grad_norm": 0.07029815018177032, + "grad_norm": 32.78211975097656, "learning_rate": 6.355475763016159e-06, - "loss": 0.4276, + "loss": 0.8568, "step": 64600 }, { "epoch": 6.82, - "grad_norm": 0.4427160620689392, + "grad_norm": 33.14503860473633, "learning_rate": 6.353363607561517e-06, - "loss": 0.4007, + "loss": 0.8058, "step": 64610 }, { "epoch": 6.82, - "grad_norm": 5.6652140617370605, + "grad_norm": 43.09387969970703, "learning_rate": 6.351251452106875e-06, - "loss": 0.3698, + "loss": 0.7677, "step": 64620 }, { "epoch": 6.83, - "grad_norm": 8.040115356445312, + "grad_norm": 46.11549377441406, "learning_rate": 6.349139296652233e-06, - "loss": 0.414, + "loss": 0.8868, "step": 64630 }, { "epoch": 6.83, - "grad_norm": 23.099340438842773, + "grad_norm": 49.69927978515625, "learning_rate": 6.347027141197593e-06, - "loss": 0.2444, + "loss": 0.811, "step": 64640 }, { "epoch": 6.83, - "grad_norm": 22.550933837890625, + "grad_norm": 21.915477752685547, "learning_rate": 6.344914985742952e-06, - "loss": 0.3633, + "loss": 0.9024, "step": 64650 }, { "epoch": 6.83, - "grad_norm": 38.30808639526367, + "grad_norm": 50.845542907714844, "learning_rate": 6.34280283028831e-06, - "loss": 0.1867, + "loss": 0.8163, "step": 64660 }, { "epoch": 6.83, - "grad_norm": 7.679875373840332, + "grad_norm": 30.3620662689209, "learning_rate": 6.3406906748336684e-06, - "loss": 0.5581, + "loss": 0.9771, "step": 64670 }, { "epoch": 6.83, - "grad_norm": 1.8864567279815674, + "grad_norm": 23.284536361694336, "learning_rate": 6.338578519379027e-06, - "loss": 0.3342, + "loss": 0.9585, "step": 64680 }, { "epoch": 6.83, - "grad_norm": 15.396658897399902, + "grad_norm": 47.85205078125, "learning_rate": 6.336466363924385e-06, - "loss": 0.3473, + "loss": 0.7807, "step": 64690 }, { "epoch": 6.83, - "grad_norm": 3.757474899291992, + "grad_norm": 28.719444274902344, "learning_rate": 6.334354208469744e-06, - "loss": 0.3913, + "loss": 0.5564, "step": 64700 }, { "epoch": 6.83, - "grad_norm": 32.36862564086914, + "grad_norm": 44.8770866394043, "learning_rate": 6.332242053015103e-06, - "loss": 0.3619, + "loss": 1.0018, "step": 64710 }, { "epoch": 6.83, - "grad_norm": 7.028613567352295, + "grad_norm": 52.32823181152344, "learning_rate": 6.330129897560461e-06, - "loss": 0.3149, + "loss": 0.5659, "step": 64720 }, { "epoch": 6.84, - "grad_norm": 6.966856956481934, + "grad_norm": 35.954795837402344, "learning_rate": 6.328017742105819e-06, - "loss": 0.3066, + "loss": 0.673, "step": 64730 }, { "epoch": 6.84, - "grad_norm": 20.09161376953125, + "grad_norm": 39.51060104370117, "learning_rate": 6.325905586651177e-06, - "loss": 0.503, + "loss": 0.7458, "step": 64740 }, { "epoch": 6.84, - "grad_norm": 17.053516387939453, + "grad_norm": 71.29031372070312, "learning_rate": 6.323793431196537e-06, - "loss": 0.1728, + "loss": 0.9539, "step": 64750 }, { "epoch": 6.84, - "grad_norm": 19.64523696899414, + "grad_norm": 51.41810989379883, "learning_rate": 6.321681275741896e-06, - "loss": 0.7274, + "loss": 1.2092, "step": 64760 }, { "epoch": 6.84, - "grad_norm": 8.310595512390137, + "grad_norm": 43.87831497192383, "learning_rate": 6.319569120287254e-06, - "loss": 0.5405, + "loss": 1.0009, "step": 64770 }, { "epoch": 6.84, - "grad_norm": 32.337867736816406, + "grad_norm": 38.64042282104492, "learning_rate": 6.3174569648326125e-06, - "loss": 0.5044, + "loss": 0.9746, "step": 64780 }, { "epoch": 6.84, - "grad_norm": 15.761940002441406, + "grad_norm": 21.45062255859375, "learning_rate": 6.315344809377971e-06, - "loss": 0.6096, + "loss": 1.3064, "step": 64790 }, { "epoch": 6.84, - "grad_norm": 5.418169975280762, + "grad_norm": 27.475038528442383, "learning_rate": 6.313232653923329e-06, - "loss": 0.3869, + "loss": 1.0409, "step": 64800 }, { "epoch": 6.84, - "grad_norm": 2.2313215732574463, + "grad_norm": 45.83359909057617, "learning_rate": 6.311120498468688e-06, - "loss": 0.148, + "loss": 0.8441, "step": 64810 }, { "epoch": 6.85, - "grad_norm": 14.839144706726074, + "grad_norm": 41.50979232788086, "learning_rate": 6.309008343014046e-06, - "loss": 0.609, + "loss": 1.0311, "step": 64820 }, { "epoch": 6.85, - "grad_norm": 21.211027145385742, + "grad_norm": 52.89299392700195, "learning_rate": 6.306896187559405e-06, - "loss": 0.6149, + "loss": 1.3318, "step": 64830 }, { "epoch": 6.85, - "grad_norm": 11.216279029846191, + "grad_norm": 57.516822814941406, "learning_rate": 6.3047840321047634e-06, - "loss": 0.2492, + "loss": 1.0091, "step": 64840 }, { "epoch": 6.85, - "grad_norm": 1.7776063680648804, + "grad_norm": 37.0876350402832, "learning_rate": 6.302671876650121e-06, - "loss": 0.3064, + "loss": 0.9032, "step": 64850 }, { "epoch": 6.85, - "grad_norm": 30.017841339111328, + "grad_norm": 52.755836486816406, "learning_rate": 6.30055972119548e-06, - "loss": 0.3147, + "loss": 1.1279, "step": 64860 }, { "epoch": 6.85, - "grad_norm": 46.61854934692383, + "grad_norm": 25.458324432373047, "learning_rate": 6.29844756574084e-06, - "loss": 0.3894, + "loss": 0.7492, "step": 64870 }, { "epoch": 6.85, - "grad_norm": 0.07890544086694717, + "grad_norm": 27.636760711669922, "learning_rate": 6.296335410286198e-06, - "loss": 0.2282, + "loss": 0.7254, "step": 64880 }, { "epoch": 6.85, - "grad_norm": 4.304662704467773, + "grad_norm": 27.44355010986328, "learning_rate": 6.2942232548315565e-06, - "loss": 0.2036, + "loss": 0.832, "step": 64890 }, { "epoch": 6.85, - "grad_norm": 20.821510314941406, + "grad_norm": 42.52280044555664, "learning_rate": 6.292111099376914e-06, - "loss": 0.3803, + "loss": 0.8714, "step": 64900 }, { "epoch": 6.86, - "grad_norm": 2.601516008377075, + "grad_norm": 28.124250411987305, "learning_rate": 6.289998943922273e-06, - "loss": 0.2464, + "loss": 1.0013, "step": 64910 }, { "epoch": 6.86, - "grad_norm": 15.380260467529297, + "grad_norm": 50.680789947509766, "learning_rate": 6.287886788467632e-06, - "loss": 0.2481, + "loss": 0.6501, "step": 64920 }, { "epoch": 6.86, - "grad_norm": 31.21272850036621, + "grad_norm": 42.60406494140625, "learning_rate": 6.28577463301299e-06, - "loss": 0.4275, + "loss": 1.0889, "step": 64930 }, { "epoch": 6.86, - "grad_norm": 7.1965718269348145, + "grad_norm": 37.500579833984375, "learning_rate": 6.283662477558349e-06, - "loss": 0.4054, + "loss": 1.0586, "step": 64940 }, { "epoch": 6.86, - "grad_norm": 0.09129659086465836, + "grad_norm": 46.90925979614258, "learning_rate": 6.2815503221037075e-06, - "loss": 0.4074, + "loss": 0.9428, "step": 64950 }, { "epoch": 6.86, - "grad_norm": 7.3914313316345215, + "grad_norm": 56.5093994140625, "learning_rate": 6.279438166649065e-06, - "loss": 0.4803, + "loss": 0.9846, "step": 64960 }, { "epoch": 6.86, - "grad_norm": 29.857933044433594, + "grad_norm": 26.797456741333008, "learning_rate": 6.277326011194424e-06, - "loss": 0.3574, + "loss": 0.8898, "step": 64970 }, { "epoch": 6.86, - "grad_norm": 20.351892471313477, + "grad_norm": 50.034156799316406, "learning_rate": 6.275213855739784e-06, - "loss": 0.2821, + "loss": 0.9349, "step": 64980 }, { "epoch": 6.86, - "grad_norm": 0.8537487983703613, + "grad_norm": 44.20176696777344, "learning_rate": 6.273101700285142e-06, - "loss": 0.5175, + "loss": 0.9708, "step": 64990 }, { "epoch": 6.86, - "grad_norm": 15.05611515045166, + "grad_norm": 31.095661163330078, "learning_rate": 6.2709895448305005e-06, - "loss": 0.5107, + "loss": 1.1057, "step": 65000 }, { "epoch": 6.87, - "grad_norm": 6.271355628967285, + "grad_norm": 37.50794982910156, "learning_rate": 6.2688773893758584e-06, - "loss": 0.2538, + "loss": 1.0056, "step": 65010 }, { "epoch": 6.87, - "grad_norm": 0.761966347694397, + "grad_norm": 33.99724197387695, "learning_rate": 6.266765233921217e-06, - "loss": 0.3488, + "loss": 0.9336, "step": 65020 }, { "epoch": 6.87, - "grad_norm": 28.70721435546875, + "grad_norm": 19.92735481262207, "learning_rate": 6.264653078466576e-06, - "loss": 0.3029, + "loss": 0.8369, "step": 65030 }, { "epoch": 6.87, - "grad_norm": 7.955304145812988, + "grad_norm": 32.67523193359375, "learning_rate": 6.262540923011934e-06, - "loss": 0.3353, + "loss": 0.9605, "step": 65040 }, { "epoch": 6.87, - "grad_norm": 2.0460598468780518, + "grad_norm": 1.9416269063949585, "learning_rate": 6.260428767557293e-06, - "loss": 0.4968, + "loss": 0.953, "step": 65050 }, { "epoch": 6.87, - "grad_norm": 13.437950134277344, + "grad_norm": 21.333751678466797, "learning_rate": 6.2583166121026515e-06, - "loss": 0.6637, + "loss": 0.8928, "step": 65060 }, { "epoch": 6.87, - "grad_norm": 1.0317875146865845, + "grad_norm": 32.428810119628906, "learning_rate": 6.256204456648009e-06, - "loss": 0.2523, + "loss": 0.7428, "step": 65070 }, { "epoch": 6.87, - "grad_norm": 0.2507677376270294, + "grad_norm": 57.57966995239258, "learning_rate": 6.254092301193368e-06, - "loss": 0.4346, + "loss": 1.1574, "step": 65080 }, { "epoch": 6.87, - "grad_norm": 0.23211297392845154, + "grad_norm": 23.549827575683594, "learning_rate": 6.251980145738726e-06, - "loss": 0.4437, + "loss": 0.8464, "step": 65090 }, { "epoch": 6.88, - "grad_norm": 19.51643180847168, + "grad_norm": 33.83430862426758, "learning_rate": 6.249867990284085e-06, - "loss": 0.2902, + "loss": 1.0349, "step": 65100 }, { "epoch": 6.88, - "grad_norm": 8.227494239807129, + "grad_norm": 49.969181060791016, "learning_rate": 6.2477558348294445e-06, - "loss": 0.2127, + "loss": 0.808, "step": 65110 }, { "epoch": 6.88, - "grad_norm": 15.63725757598877, + "grad_norm": 38.97306823730469, "learning_rate": 6.2456436793748025e-06, - "loss": 0.4731, + "loss": 1.0806, "step": 65120 }, { "epoch": 6.88, - "grad_norm": 19.859405517578125, + "grad_norm": 61.55247116088867, "learning_rate": 6.243531523920161e-06, - "loss": 0.4899, + "loss": 1.1872, "step": 65130 }, { "epoch": 6.88, - "grad_norm": 0.19917218387126923, + "grad_norm": 23.136215209960938, "learning_rate": 6.24141936846552e-06, - "loss": 0.2749, + "loss": 0.7836, "step": 65140 }, { "epoch": 6.88, - "grad_norm": 26.041894912719727, + "grad_norm": 32.34538269042969, "learning_rate": 6.239307213010878e-06, - "loss": 0.5846, + "loss": 0.9547, "step": 65150 }, { "epoch": 6.88, - "grad_norm": 8.396547317504883, + "grad_norm": 37.44535827636719, "learning_rate": 6.237195057556237e-06, - "loss": 0.7723, + "loss": 0.9642, "step": 65160 }, { "epoch": 6.88, - "grad_norm": 40.93504333496094, + "grad_norm": 19.160015106201172, "learning_rate": 6.2350829021015955e-06, - "loss": 0.4361, + "loss": 0.8225, "step": 65170 }, { "epoch": 6.88, - "grad_norm": 7.82274055480957, + "grad_norm": 46.889957427978516, "learning_rate": 6.2329707466469534e-06, - "loss": 0.3233, + "loss": 0.9698, "step": 65180 }, { "epoch": 6.88, - "grad_norm": 1.4050843715667725, + "grad_norm": 28.286243438720703, "learning_rate": 6.230858591192312e-06, - "loss": 0.5796, + "loss": 0.9801, "step": 65190 }, { "epoch": 6.89, - "grad_norm": 7.981564044952393, + "grad_norm": 39.53450012207031, "learning_rate": 6.22874643573767e-06, - "loss": 0.4368, + "loss": 0.9591, "step": 65200 }, { "epoch": 6.89, - "grad_norm": 14.706910133361816, + "grad_norm": 32.881866455078125, "learning_rate": 6.226634280283029e-06, - "loss": 0.328, + "loss": 0.7847, "step": 65210 }, { "epoch": 6.89, - "grad_norm": 26.99664306640625, + "grad_norm": 46.91219711303711, "learning_rate": 6.2245221248283886e-06, - "loss": 0.4699, + "loss": 0.9143, "step": 65220 }, { "epoch": 6.89, - "grad_norm": 16.62078857421875, + "grad_norm": 63.21809768676758, "learning_rate": 6.2224099693737465e-06, - "loss": 0.3694, + "loss": 1.1335, "step": 65230 }, { "epoch": 6.89, - "grad_norm": 19.03057289123535, + "grad_norm": 41.45021438598633, "learning_rate": 6.220297813919105e-06, - "loss": 0.398, + "loss": 0.7479, "step": 65240 }, { "epoch": 6.89, - "grad_norm": 18.839847564697266, + "grad_norm": 48.22402572631836, "learning_rate": 6.218185658464464e-06, - "loss": 0.615, + "loss": 0.7722, "step": 65250 }, { "epoch": 6.89, - "grad_norm": 18.88266944885254, + "grad_norm": 41.85344696044922, "learning_rate": 6.216073503009822e-06, - "loss": 0.5439, + "loss": 1.2453, "step": 65260 }, { "epoch": 6.89, - "grad_norm": 8.401565551757812, + "grad_norm": 32.425533294677734, "learning_rate": 6.213961347555181e-06, - "loss": 0.2652, + "loss": 0.7528, "step": 65270 }, { "epoch": 6.89, - "grad_norm": 40.55706024169922, + "grad_norm": 36.17947769165039, "learning_rate": 6.211849192100539e-06, - "loss": 0.4947, + "loss": 1.017, "step": 65280 }, { "epoch": 6.9, - "grad_norm": 2.6610162258148193, + "grad_norm": 24.43809700012207, "learning_rate": 6.2097370366458975e-06, - "loss": 0.5418, + "loss": 1.0752, "step": 65290 }, { "epoch": 6.9, - "grad_norm": 0.060124121606349945, + "grad_norm": 2.2834885120391846, "learning_rate": 6.207624881191256e-06, - "loss": 0.5339, + "loss": 0.9007, "step": 65300 }, { "epoch": 6.9, - "grad_norm": 0.20476359128952026, + "grad_norm": 53.37857437133789, "learning_rate": 6.205512725736614e-06, - "loss": 0.2572, + "loss": 0.9966, "step": 65310 }, { "epoch": 6.9, - "grad_norm": 1.764794945716858, + "grad_norm": 32.332183837890625, "learning_rate": 6.203400570281973e-06, - "loss": 0.3109, + "loss": 0.9194, "step": 65320 }, { "epoch": 6.9, - "grad_norm": 21.92900276184082, + "grad_norm": 51.89722442626953, "learning_rate": 6.201288414827333e-06, - "loss": 0.2292, + "loss": 0.9092, "step": 65330 }, { "epoch": 6.9, - "grad_norm": 0.14187394082546234, + "grad_norm": 46.8034553527832, "learning_rate": 6.19917625937269e-06, - "loss": 0.4425, + "loss": 0.8014, "step": 65340 }, { "epoch": 6.9, - "grad_norm": 5.746006965637207, + "grad_norm": 48.79391098022461, "learning_rate": 6.197064103918049e-06, - "loss": 0.5727, + "loss": 0.8112, "step": 65350 }, { "epoch": 6.9, - "grad_norm": 27.29819107055664, + "grad_norm": 45.03374481201172, "learning_rate": 6.194951948463407e-06, - "loss": 0.3764, + "loss": 0.9206, "step": 65360 }, { "epoch": 6.9, - "grad_norm": 11.062472343444824, + "grad_norm": 34.804237365722656, "learning_rate": 6.192839793008766e-06, - "loss": 0.3304, + "loss": 0.8397, "step": 65370 }, { "epoch": 6.9, - "grad_norm": 24.853656768798828, + "grad_norm": 41.14748001098633, "learning_rate": 6.190727637554125e-06, - "loss": 0.3036, + "loss": 0.9635, "step": 65380 }, { "epoch": 6.91, - "grad_norm": 0.9043033719062805, + "grad_norm": 25.361251831054688, "learning_rate": 6.188615482099483e-06, - "loss": 0.2658, + "loss": 0.6979, "step": 65390 }, { "epoch": 6.91, - "grad_norm": 33.297584533691406, + "grad_norm": 55.77430725097656, "learning_rate": 6.1865033266448415e-06, - "loss": 0.7065, + "loss": 1.151, "step": 65400 }, { "epoch": 6.91, - "grad_norm": 9.632415771484375, + "grad_norm": 57.12754821777344, "learning_rate": 6.1843911711902e-06, - "loss": 0.2399, + "loss": 0.9062, "step": 65410 }, { "epoch": 6.91, - "grad_norm": 8.65435791015625, + "grad_norm": 59.19919967651367, "learning_rate": 6.182279015735558e-06, - "loss": 0.2903, + "loss": 1.1074, "step": 65420 }, { "epoch": 6.91, - "grad_norm": 34.22492218017578, + "grad_norm": 37.07661437988281, "learning_rate": 6.180166860280917e-06, - "loss": 0.8013, + "loss": 1.0917, "step": 65430 }, { "epoch": 6.91, - "grad_norm": 7.699493885040283, + "grad_norm": 35.77449417114258, "learning_rate": 6.178054704826277e-06, - "loss": 0.4241, + "loss": 0.8956, "step": 65440 }, { "epoch": 6.91, - "grad_norm": 21.368515014648438, + "grad_norm": 23.025524139404297, "learning_rate": 6.175942549371634e-06, - "loss": 0.8611, + "loss": 0.9512, "step": 65450 }, { "epoch": 6.91, - "grad_norm": 12.4188814163208, + "grad_norm": 28.972986221313477, "learning_rate": 6.173830393916993e-06, - "loss": 0.28, + "loss": 0.7606, "step": 65460 }, { "epoch": 6.91, - "grad_norm": 14.907275199890137, + "grad_norm": 39.938446044921875, "learning_rate": 6.171718238462351e-06, - "loss": 0.3181, + "loss": 0.9384, "step": 65470 }, { "epoch": 6.92, - "grad_norm": 2.961615800857544, + "grad_norm": 28.77251625061035, "learning_rate": 6.16960608300771e-06, - "loss": 0.3333, + "loss": 1.0277, "step": 65480 }, { "epoch": 6.92, - "grad_norm": 16.567855834960938, + "grad_norm": 44.2778205871582, "learning_rate": 6.167493927553069e-06, - "loss": 0.3538, + "loss": 0.9373, "step": 65490 }, { "epoch": 6.92, - "grad_norm": 17.649131774902344, + "grad_norm": 37.129451751708984, "learning_rate": 6.165381772098427e-06, - "loss": 0.2967, + "loss": 0.608, "step": 65500 }, { "epoch": 6.92, - "grad_norm": 29.27667236328125, + "grad_norm": 35.30207824707031, "learning_rate": 6.1632696166437855e-06, - "loss": 0.5406, + "loss": 1.0742, "step": 65510 }, { "epoch": 6.92, - "grad_norm": 12.891237258911133, + "grad_norm": 31.67316246032715, "learning_rate": 6.161157461189144e-06, - "loss": 0.6517, + "loss": 1.1394, "step": 65520 }, { "epoch": 6.92, - "grad_norm": 10.748887062072754, + "grad_norm": 48.276756286621094, "learning_rate": 6.159045305734502e-06, - "loss": 0.6586, + "loss": 1.292, "step": 65530 }, { "epoch": 6.92, - "grad_norm": 12.623597145080566, + "grad_norm": 48.261043548583984, "learning_rate": 6.156933150279861e-06, - "loss": 0.3673, + "loss": 0.9825, "step": 65540 }, { "epoch": 6.92, - "grad_norm": 13.110079765319824, + "grad_norm": 28.51682472229004, "learning_rate": 6.154820994825219e-06, - "loss": 0.5899, + "loss": 1.2355, "step": 65550 }, { "epoch": 6.92, - "grad_norm": 6.280153274536133, + "grad_norm": 34.4248046875, "learning_rate": 6.152708839370578e-06, - "loss": 0.4698, + "loss": 0.7235, "step": 65560 }, { "epoch": 6.92, - "grad_norm": 28.80436897277832, + "grad_norm": 53.94939041137695, "learning_rate": 6.150596683915937e-06, - "loss": 0.3066, + "loss": 1.135, "step": 65570 }, { "epoch": 6.93, - "grad_norm": 3.4673655033111572, + "grad_norm": 38.893348693847656, "learning_rate": 6.1484845284612944e-06, - "loss": 0.583, + "loss": 1.2404, "step": 65580 }, { "epoch": 6.93, - "grad_norm": 13.447654724121094, + "grad_norm": 44.57937240600586, "learning_rate": 6.146372373006654e-06, - "loss": 0.4899, + "loss": 0.9473, "step": 65590 }, { "epoch": 6.93, - "grad_norm": 3.040531873703003, + "grad_norm": 44.261932373046875, "learning_rate": 6.144260217552013e-06, - "loss": 0.299, + "loss": 0.6733, "step": 65600 }, { "epoch": 6.93, - "grad_norm": 26.0434627532959, + "grad_norm": 49.44343948364258, "learning_rate": 6.142148062097371e-06, - "loss": 0.3101, + "loss": 1.1382, "step": 65610 }, { "epoch": 6.93, - "grad_norm": 9.325292587280273, + "grad_norm": 45.65007781982422, "learning_rate": 6.1400359066427295e-06, - "loss": 0.6419, + "loss": 1.1795, "step": 65620 }, { "epoch": 6.93, - "grad_norm": 32.53392791748047, + "grad_norm": 50.70222091674805, "learning_rate": 6.137923751188088e-06, - "loss": 0.359, + "loss": 0.9542, "step": 65630 }, { "epoch": 6.93, - "grad_norm": 16.219314575195312, + "grad_norm": 39.79031753540039, "learning_rate": 6.135811595733446e-06, - "loss": 0.2429, + "loss": 0.9041, "step": 65640 }, { "epoch": 6.93, - "grad_norm": 0.6677464246749878, + "grad_norm": 46.0681266784668, "learning_rate": 6.133699440278805e-06, - "loss": 0.6161, + "loss": 1.1727, "step": 65650 }, { "epoch": 6.93, - "grad_norm": 7.4570536613464355, + "grad_norm": 20.49182891845703, "learning_rate": 6.131587284824163e-06, - "loss": 0.3813, + "loss": 0.7819, "step": 65660 }, { "epoch": 6.94, - "grad_norm": 0.12335769832134247, + "grad_norm": 35.86573791503906, "learning_rate": 6.129475129369522e-06, - "loss": 0.4804, + "loss": 1.0477, "step": 65670 }, { "epoch": 6.94, - "grad_norm": 39.640621185302734, + "grad_norm": 42.2222785949707, "learning_rate": 6.127362973914881e-06, - "loss": 0.4445, + "loss": 1.1005, "step": 65680 }, { "epoch": 6.94, - "grad_norm": 17.67512321472168, + "grad_norm": 28.861248016357422, "learning_rate": 6.1252508184602385e-06, - "loss": 0.3055, + "loss": 0.8906, "step": 65690 }, { "epoch": 6.94, - "grad_norm": 11.748019218444824, + "grad_norm": 39.230430603027344, "learning_rate": 6.123138663005598e-06, - "loss": 0.2527, + "loss": 0.7842, "step": 65700 }, { "epoch": 6.94, - "grad_norm": 13.812315940856934, + "grad_norm": 34.30377197265625, "learning_rate": 6.121026507550957e-06, - "loss": 0.4355, + "loss": 0.9071, "step": 65710 }, { "epoch": 6.94, - "grad_norm": 24.925447463989258, + "grad_norm": 23.56003761291504, "learning_rate": 6.118914352096315e-06, - "loss": 0.2751, + "loss": 0.8989, "step": 65720 }, { "epoch": 6.94, - "grad_norm": 32.44852828979492, + "grad_norm": 40.7226676940918, "learning_rate": 6.1168021966416736e-06, - "loss": 0.6836, + "loss": 1.3092, "step": 65730 }, { "epoch": 6.94, - "grad_norm": 7.494381427764893, + "grad_norm": 37.663883209228516, "learning_rate": 6.1146900411870315e-06, - "loss": 0.1278, + "loss": 0.8074, "step": 65740 }, { "epoch": 6.94, - "grad_norm": 10.044979095458984, + "grad_norm": 35.987918853759766, "learning_rate": 6.11257788573239e-06, - "loss": 0.4701, + "loss": 0.8162, "step": 65750 }, { "epoch": 6.94, - "grad_norm": 41.918739318847656, + "grad_norm": 56.06478500366211, "learning_rate": 6.110465730277749e-06, - "loss": 0.1381, + "loss": 0.8047, "step": 65760 }, { "epoch": 6.95, - "grad_norm": 0.02669399604201317, + "grad_norm": 17.82347869873047, "learning_rate": 6.108353574823107e-06, - "loss": 0.4134, + "loss": 1.0091, "step": 65770 }, { "epoch": 6.95, - "grad_norm": 0.5666239857673645, + "grad_norm": 67.83700561523438, "learning_rate": 6.106241419368466e-06, - "loss": 0.4817, + "loss": 0.9299, "step": 65780 }, { "epoch": 6.95, - "grad_norm": 16.934968948364258, + "grad_norm": 65.84854125976562, "learning_rate": 6.104129263913825e-06, - "loss": 0.5665, + "loss": 1.0708, "step": 65790 }, { "epoch": 6.95, - "grad_norm": 0.08877577632665634, + "grad_norm": 39.439334869384766, "learning_rate": 6.1020171084591825e-06, - "loss": 0.6332, + "loss": 0.814, "step": 65800 }, { "epoch": 6.95, - "grad_norm": 14.763039588928223, + "grad_norm": 30.98998260498047, "learning_rate": 6.099904953004542e-06, - "loss": 0.4342, + "loss": 0.9704, "step": 65810 }, { "epoch": 6.95, - "grad_norm": 25.758464813232422, + "grad_norm": 55.20170593261719, "learning_rate": 6.097792797549901e-06, - "loss": 0.5833, + "loss": 1.4758, "step": 65820 }, { "epoch": 6.95, - "grad_norm": 8.595956802368164, + "grad_norm": 54.94887161254883, "learning_rate": 6.095680642095259e-06, - "loss": 0.4492, + "loss": 0.9283, "step": 65830 }, { "epoch": 6.95, - "grad_norm": 2.8746368885040283, + "grad_norm": 28.83126449584961, "learning_rate": 6.093568486640618e-06, - "loss": 0.3757, + "loss": 0.9362, "step": 65840 }, { "epoch": 6.95, - "grad_norm": 29.951814651489258, + "grad_norm": 58.19230651855469, "learning_rate": 6.0914563311859755e-06, - "loss": 0.3629, + "loss": 1.2414, "step": 65850 }, { "epoch": 6.96, - "grad_norm": 9.767029762268066, + "grad_norm": 19.122385025024414, "learning_rate": 6.089344175731334e-06, - "loss": 0.6842, + "loss": 0.9106, "step": 65860 }, { "epoch": 6.96, - "grad_norm": 20.168529510498047, + "grad_norm": 41.383323669433594, "learning_rate": 6.087232020276693e-06, - "loss": 0.4973, + "loss": 0.8312, "step": 65870 }, { "epoch": 6.96, - "grad_norm": 15.994670867919922, + "grad_norm": 49.874847412109375, "learning_rate": 6.085119864822051e-06, - "loss": 0.5804, + "loss": 1.3616, "step": 65880 }, { "epoch": 6.96, - "grad_norm": 28.901676177978516, + "grad_norm": 44.164649963378906, "learning_rate": 6.08300770936741e-06, - "loss": 0.3578, + "loss": 0.9704, "step": 65890 }, { "epoch": 6.96, - "grad_norm": 2.4189319610595703, + "grad_norm": 35.159584045410156, "learning_rate": 6.080895553912769e-06, - "loss": 0.3444, + "loss": 1.2926, "step": 65900 }, { "epoch": 6.96, - "grad_norm": 20.864484786987305, + "grad_norm": 22.88376235961914, "learning_rate": 6.0787833984581265e-06, - "loss": 0.495, + "loss": 1.1403, "step": 65910 }, { "epoch": 6.96, - "grad_norm": 18.781625747680664, + "grad_norm": 41.069114685058594, "learning_rate": 6.076671243003486e-06, - "loss": 0.6015, + "loss": 1.1194, "step": 65920 }, { "epoch": 6.96, - "grad_norm": 16.71904754638672, + "grad_norm": 17.42886734008789, "learning_rate": 6.074559087548843e-06, - "loss": 0.332, + "loss": 0.9002, "step": 65930 }, { "epoch": 6.96, - "grad_norm": 12.359519958496094, + "grad_norm": 48.318206787109375, "learning_rate": 6.072446932094203e-06, - "loss": 0.5853, + "loss": 0.9473, "step": 65940 }, { "epoch": 6.96, - "grad_norm": 0.6520966291427612, + "grad_norm": 50.507049560546875, "learning_rate": 6.070334776639562e-06, - "loss": 0.4698, + "loss": 0.9736, "step": 65950 }, { "epoch": 6.97, - "grad_norm": 10.215459823608398, + "grad_norm": 32.20198059082031, "learning_rate": 6.0682226211849196e-06, - "loss": 0.579, + "loss": 1.2908, "step": 65960 }, { "epoch": 6.97, - "grad_norm": 19.0490665435791, + "grad_norm": 76.02461242675781, "learning_rate": 6.066110465730278e-06, - "loss": 0.3938, + "loss": 1.1046, "step": 65970 }, { "epoch": 6.97, - "grad_norm": 8.572076797485352, + "grad_norm": 30.685258865356445, "learning_rate": 6.063998310275637e-06, - "loss": 0.4071, + "loss": 0.9127, "step": 65980 }, { "epoch": 6.97, - "grad_norm": 20.37465476989746, + "grad_norm": 38.42310333251953, "learning_rate": 6.061886154820995e-06, - "loss": 0.3116, + "loss": 1.0461, "step": 65990 }, { "epoch": 6.97, - "grad_norm": 24.89883804321289, + "grad_norm": 33.601173400878906, "learning_rate": 6.059773999366354e-06, - "loss": 0.3796, + "loss": 0.5571, "step": 66000 }, { "epoch": 6.97, - "grad_norm": 18.700424194335938, + "grad_norm": 25.33232307434082, "learning_rate": 6.057661843911712e-06, - "loss": 0.5698, + "loss": 0.9591, "step": 66010 }, { "epoch": 6.97, - "grad_norm": 16.53180694580078, + "grad_norm": 56.23831558227539, "learning_rate": 6.0555496884570705e-06, - "loss": 0.2964, + "loss": 0.8255, "step": 66020 }, { "epoch": 6.97, - "grad_norm": 17.717782974243164, + "grad_norm": 49.14065933227539, "learning_rate": 6.05343753300243e-06, - "loss": 0.3657, + "loss": 0.9657, "step": 66030 }, { "epoch": 6.97, - "grad_norm": 1.380627155303955, + "grad_norm": 27.379592895507812, "learning_rate": 6.051325377547787e-06, - "loss": 0.3553, + "loss": 1.078, "step": 66040 }, { "epoch": 6.98, - "grad_norm": 6.318301677703857, + "grad_norm": 33.00341033935547, "learning_rate": 6.049213222093147e-06, - "loss": 0.5161, + "loss": 1.2431, "step": 66050 }, { "epoch": 6.98, - "grad_norm": 12.737546920776367, + "grad_norm": 31.23637580871582, "learning_rate": 6.047101066638506e-06, - "loss": 0.2764, + "loss": 0.8578, "step": 66060 }, { "epoch": 6.98, - "grad_norm": 21.497459411621094, + "grad_norm": 54.7639045715332, "learning_rate": 6.044988911183864e-06, - "loss": 0.4395, + "loss": 0.7127, "step": 66070 }, { "epoch": 6.98, - "grad_norm": 11.113000869750977, + "grad_norm": 24.57461166381836, "learning_rate": 6.042876755729222e-06, - "loss": 0.4141, + "loss": 0.7303, "step": 66080 }, { "epoch": 6.98, - "grad_norm": 18.190181732177734, + "grad_norm": 50.76005172729492, "learning_rate": 6.040764600274581e-06, - "loss": 0.3143, + "loss": 0.6472, "step": 66090 }, { "epoch": 6.98, - "grad_norm": 2.766918420791626, + "grad_norm": 46.229576110839844, "learning_rate": 6.038652444819939e-06, - "loss": 0.3654, + "loss": 0.9639, "step": 66100 }, { "epoch": 6.98, - "grad_norm": 25.652530670166016, + "grad_norm": 23.8380126953125, "learning_rate": 6.036540289365298e-06, - "loss": 0.2198, + "loss": 0.8196, "step": 66110 }, { "epoch": 6.98, - "grad_norm": 10.736431121826172, + "grad_norm": 58.22261047363281, "learning_rate": 6.034428133910656e-06, - "loss": 0.5385, + "loss": 0.8711, "step": 66120 }, { "epoch": 6.98, - "grad_norm": 33.03472900390625, + "grad_norm": 34.339073181152344, "learning_rate": 6.0323159784560146e-06, - "loss": 0.1884, + "loss": 0.9694, "step": 66130 }, { "epoch": 6.98, - "grad_norm": 29.380481719970703, + "grad_norm": 38.28270721435547, "learning_rate": 6.030203823001373e-06, - "loss": 0.4798, + "loss": 0.9942, "step": 66140 }, { "epoch": 6.99, - "grad_norm": 0.2586119771003723, + "grad_norm": 46.080989837646484, "learning_rate": 6.028091667546731e-06, - "loss": 0.4547, + "loss": 0.7331, "step": 66150 }, { "epoch": 6.99, - "grad_norm": 49.14113235473633, + "grad_norm": 39.11981201171875, "learning_rate": 6.025979512092091e-06, - "loss": 0.5064, + "loss": 1.0363, "step": 66160 }, { "epoch": 6.99, - "grad_norm": 11.304874420166016, + "grad_norm": 50.13003921508789, "learning_rate": 6.02386735663745e-06, - "loss": 0.4327, + "loss": 1.0723, "step": 66170 }, { "epoch": 6.99, - "grad_norm": 16.5433406829834, + "grad_norm": 40.56015396118164, "learning_rate": 6.021755201182808e-06, - "loss": 0.3667, + "loss": 0.967, "step": 66180 }, { "epoch": 6.99, - "grad_norm": 20.69442367553711, + "grad_norm": 42.558162689208984, "learning_rate": 6.019643045728166e-06, - "loss": 0.3515, + "loss": 0.9564, "step": 66190 }, { "epoch": 6.99, - "grad_norm": 0.4580593705177307, + "grad_norm": 58.739845275878906, "learning_rate": 6.017530890273524e-06, - "loss": 0.3118, + "loss": 0.879, "step": 66200 }, { "epoch": 6.99, - "grad_norm": 0.3912639617919922, + "grad_norm": 46.271942138671875, "learning_rate": 6.015418734818883e-06, - "loss": 0.2951, + "loss": 0.8612, "step": 66210 }, { "epoch": 6.99, - "grad_norm": 4.208095550537109, + "grad_norm": 31.703380584716797, "learning_rate": 6.013306579364242e-06, - "loss": 0.453, + "loss": 0.8908, "step": 66220 }, { "epoch": 6.99, - "grad_norm": 70.90357208251953, + "grad_norm": 44.27516174316406, "learning_rate": 6.0111944239096e-06, - "loss": 0.4421, + "loss": 0.8615, "step": 66230 }, { "epoch": 7.0, - "grad_norm": 19.75486946105957, + "grad_norm": 51.80461120605469, "learning_rate": 6.009082268454959e-06, - "loss": 0.2819, + "loss": 1.2188, "step": 66240 }, { "epoch": 7.0, - "grad_norm": 7.056675910949707, + "grad_norm": 39.91762924194336, "learning_rate": 6.006970113000317e-06, - "loss": 0.2124, + "loss": 0.981, "step": 66250 }, { "epoch": 7.0, - "grad_norm": 68.44386291503906, + "grad_norm": 62.482086181640625, "learning_rate": 6.004857957545675e-06, - "loss": 0.3307, + "loss": 0.9446, "step": 66260 }, { "epoch": 7.0, - "grad_norm": 0.8997002243995667, + "grad_norm": 32.615028381347656, "learning_rate": 6.002745802091035e-06, - "loss": 0.2695, + "loss": 0.8488, "step": 66270 }, { "epoch": 7.0, - "grad_norm": 0.049876920878887177, + "grad_norm": 30.921958923339844, "learning_rate": 6.000633646636394e-06, - "loss": 0.3935, + "loss": 1.0269, "step": 66280 }, { "epoch": 7.0, - "eval_accuracy": 0.9035247524752476, - "eval_loss": 0.38733184337615967, - "eval_runtime": 214.4943, - "eval_samples_per_second": 117.719, - "eval_steps_per_second": 14.718, + "eval_accuracy": 0.8072871287128713, + "eval_loss": 0.7022345066070557, + "eval_runtime": 39.569, + "eval_samples_per_second": 638.126, + "eval_steps_per_second": 79.785, "step": 66283 }, { "epoch": 7.0, - "grad_norm": 14.111310005187988, + "grad_norm": 58.936954498291016, "learning_rate": 5.998521491181752e-06, - "loss": 0.6076, + "loss": 1.018, "step": 66290 }, { "epoch": 7.0, - "grad_norm": 3.601174831390381, + "grad_norm": 14.170711517333984, "learning_rate": 5.99640933572711e-06, - "loss": 0.2287, + "loss": 0.5815, "step": 66300 }, { "epoch": 7.0, - "grad_norm": 18.151012420654297, + "grad_norm": 28.352882385253906, "learning_rate": 5.994297180272468e-06, - "loss": 0.374, + "loss": 1.0143, "step": 66310 }, { "epoch": 7.0, - "grad_norm": 11.81060791015625, + "grad_norm": 45.96833038330078, "learning_rate": 5.992185024817827e-06, - "loss": 0.6178, + "loss": 1.2594, "step": 66320 }, { "epoch": 7.0, - "grad_norm": 11.939032554626465, + "grad_norm": 46.95756149291992, "learning_rate": 5.990072869363186e-06, - "loss": 0.3871, + "loss": 1.043, "step": 66330 }, { "epoch": 7.01, - "grad_norm": 24.4205379486084, + "grad_norm": 65.57032775878906, "learning_rate": 5.987960713908544e-06, - "loss": 0.5144, + "loss": 0.9123, "step": 66340 }, { "epoch": 7.01, - "grad_norm": 3.0425972938537598, + "grad_norm": 49.3924674987793, "learning_rate": 5.985848558453903e-06, - "loss": 0.4097, + "loss": 0.8235, "step": 66350 }, { "epoch": 7.01, - "grad_norm": 5.195425510406494, + "grad_norm": 36.23579788208008, "learning_rate": 5.983736402999261e-06, - "loss": 0.349, + "loss": 1.0019, "step": 66360 }, { "epoch": 7.01, - "grad_norm": 3.6430366039276123, + "grad_norm": 45.85165786743164, "learning_rate": 5.981624247544619e-06, - "loss": 0.4561, + "loss": 1.0747, "step": 66370 }, { "epoch": 7.01, - "grad_norm": 0.7248649001121521, + "grad_norm": 25.886098861694336, "learning_rate": 5.979512092089978e-06, - "loss": 0.335, + "loss": 0.7706, "step": 66380 }, { "epoch": 7.01, - "grad_norm": 5.920119762420654, + "grad_norm": 12.264871597290039, "learning_rate": 5.977399936635336e-06, - "loss": 0.3367, + "loss": 0.691, "step": 66390 }, { "epoch": 7.01, - "grad_norm": 0.07643543183803558, + "grad_norm": 33.612667083740234, "learning_rate": 5.975287781180696e-06, - "loss": 0.2738, + "loss": 0.9488, "step": 66400 }, { "epoch": 7.01, - "grad_norm": 36.302947998046875, + "grad_norm": 32.59462356567383, "learning_rate": 5.9731756257260544e-06, - "loss": 0.3759, + "loss": 0.9255, "step": 66410 }, { "epoch": 7.01, - "grad_norm": 20.31311798095703, + "grad_norm": 51.02235794067383, "learning_rate": 5.971063470271412e-06, - "loss": 0.2179, + "loss": 0.7683, "step": 66420 }, { "epoch": 7.02, - "grad_norm": 1.666933298110962, + "grad_norm": 29.8393611907959, "learning_rate": 5.968951314816771e-06, - "loss": 0.2712, + "loss": 0.7837, "step": 66430 }, { "epoch": 7.02, - "grad_norm": 15.868318557739258, + "grad_norm": 38.1666259765625, "learning_rate": 5.96683915936213e-06, - "loss": 0.8118, + "loss": 0.9777, "step": 66440 }, { "epoch": 7.02, - "grad_norm": 22.48605728149414, + "grad_norm": 28.57309341430664, "learning_rate": 5.964727003907488e-06, - "loss": 0.283, + "loss": 0.5615, "step": 66450 }, { "epoch": 7.02, - "grad_norm": 29.113079071044922, + "grad_norm": 15.072223663330078, "learning_rate": 5.962614848452847e-06, - "loss": 0.306, + "loss": 0.7632, "step": 66460 }, { "epoch": 7.02, - "grad_norm": 6.970254898071289, + "grad_norm": 48.681034088134766, "learning_rate": 5.9605026929982046e-06, - "loss": 0.7137, + "loss": 1.2346, "step": 66470 }, { "epoch": 7.02, - "grad_norm": 16.069726943969727, + "grad_norm": 42.30389404296875, "learning_rate": 5.958390537543563e-06, - "loss": 0.2962, + "loss": 0.7667, "step": 66480 }, { "epoch": 7.02, - "grad_norm": 9.988998413085938, + "grad_norm": 47.25691223144531, "learning_rate": 5.956278382088922e-06, - "loss": 0.5255, + "loss": 1.1371, "step": 66490 }, { "epoch": 7.02, - "grad_norm": 0.03192014619708061, + "grad_norm": 29.449539184570312, "learning_rate": 5.95416622663428e-06, - "loss": 0.0875, + "loss": 0.8081, "step": 66500 }, { "epoch": 7.02, - "grad_norm": 3.1026058197021484, + "grad_norm": 22.534433364868164, "learning_rate": 5.95205407117964e-06, - "loss": 0.3365, + "loss": 0.9616, "step": 66510 }, { "epoch": 7.03, - "grad_norm": 9.752503395080566, + "grad_norm": 32.97432327270508, "learning_rate": 5.9499419157249985e-06, - "loss": 0.4583, + "loss": 0.9292, "step": 66520 }, { "epoch": 7.03, - "grad_norm": 10.676118850708008, + "grad_norm": 36.21198654174805, "learning_rate": 5.947829760270356e-06, - "loss": 0.1335, + "loss": 0.7351, "step": 66530 }, { "epoch": 7.03, - "grad_norm": 4.593667030334473, + "grad_norm": 24.9298152923584, "learning_rate": 5.945717604815715e-06, - "loss": 0.1763, + "loss": 0.8332, "step": 66540 }, { "epoch": 7.03, - "grad_norm": 0.025671107694506645, + "grad_norm": 21.434497833251953, "learning_rate": 5.943605449361074e-06, - "loss": 0.2909, + "loss": 0.7005, "step": 66550 }, { "epoch": 7.03, - "grad_norm": 9.96119499206543, + "grad_norm": 45.21477508544922, "learning_rate": 5.941493293906432e-06, - "loss": 0.6315, + "loss": 1.163, "step": 66560 }, { "epoch": 7.03, - "grad_norm": 0.07369338721036911, + "grad_norm": 38.974281311035156, "learning_rate": 5.939381138451791e-06, - "loss": 0.4924, + "loss": 0.9564, "step": 66570 }, { "epoch": 7.03, - "grad_norm": 34.56819534301758, + "grad_norm": 26.005096435546875, "learning_rate": 5.937268982997149e-06, - "loss": 0.4772, + "loss": 0.7742, "step": 66580 }, { "epoch": 7.03, - "grad_norm": 13.52147102355957, + "grad_norm": 46.481964111328125, "learning_rate": 5.935156827542507e-06, - "loss": 0.2664, + "loss": 0.9906, "step": 66590 }, { "epoch": 7.03, - "grad_norm": 0.025498248636722565, + "grad_norm": 25.425106048583984, "learning_rate": 5.933044672087866e-06, - "loss": 0.2671, + "loss": 0.6378, "step": 66600 }, { "epoch": 7.03, - "grad_norm": 23.684885025024414, + "grad_norm": 32.72923278808594, "learning_rate": 5.930932516633224e-06, - "loss": 0.4866, + "loss": 0.9051, "step": 66610 }, { "epoch": 7.04, - "grad_norm": 6.676453590393066, + "grad_norm": 50.25910949707031, "learning_rate": 5.928820361178583e-06, - "loss": 0.4786, + "loss": 0.735, "step": 66620 }, { "epoch": 7.04, - "grad_norm": 11.006547927856445, + "grad_norm": 43.086490631103516, "learning_rate": 5.9267082057239425e-06, - "loss": 0.4615, + "loss": 0.9189, "step": 66630 }, { "epoch": 7.04, - "grad_norm": 3.977776050567627, + "grad_norm": 36.35105514526367, "learning_rate": 5.9245960502693004e-06, - "loss": 0.51, + "loss": 1.0915, "step": 66640 }, { "epoch": 7.04, - "grad_norm": 14.886920928955078, + "grad_norm": 20.23328971862793, "learning_rate": 5.922483894814659e-06, - "loss": 0.5223, + "loss": 0.7474, "step": 66650 }, { "epoch": 7.04, - "grad_norm": 28.093544006347656, + "grad_norm": 34.20209884643555, "learning_rate": 5.920371739360017e-06, - "loss": 0.5091, + "loss": 0.8638, "step": 66660 }, { "epoch": 7.04, - "grad_norm": 21.766389846801758, + "grad_norm": 26.611581802368164, "learning_rate": 5.918259583905376e-06, - "loss": 0.6494, + "loss": 1.0509, "step": 66670 }, { "epoch": 7.04, - "grad_norm": 25.949438095092773, + "grad_norm": 55.6296272277832, "learning_rate": 5.916147428450735e-06, - "loss": 0.5215, + "loss": 0.995, "step": 66680 }, { "epoch": 7.04, - "grad_norm": 7.659582138061523, + "grad_norm": 20.56177520751953, "learning_rate": 5.914035272996093e-06, - "loss": 0.2795, + "loss": 0.779, "step": 66690 }, { "epoch": 7.04, - "grad_norm": 14.77236557006836, + "grad_norm": 28.863826751708984, "learning_rate": 5.911923117541451e-06, - "loss": 0.1186, + "loss": 0.7694, "step": 66700 }, { "epoch": 7.05, - "grad_norm": 12.06309986114502, + "grad_norm": 45.61697006225586, "learning_rate": 5.90981096208681e-06, - "loss": 0.4072, + "loss": 0.9762, "step": 66710 }, { "epoch": 7.05, - "grad_norm": 13.11417007446289, + "grad_norm": 61.164180755615234, "learning_rate": 5.907698806632168e-06, - "loss": 0.2826, + "loss": 0.9926, "step": 66720 }, { "epoch": 7.05, - "grad_norm": 8.668473243713379, + "grad_norm": 31.998653411865234, "learning_rate": 5.905586651177527e-06, - "loss": 0.3818, + "loss": 0.8547, "step": 66730 }, { "epoch": 7.05, - "grad_norm": 16.575679779052734, + "grad_norm": 45.69172286987305, "learning_rate": 5.9034744957228865e-06, - "loss": 0.4729, + "loss": 0.9466, "step": 66740 }, { "epoch": 7.05, - "grad_norm": 14.399765014648438, + "grad_norm": 20.900110244750977, "learning_rate": 5.9013623402682445e-06, - "loss": 0.5234, + "loss": 0.8029, "step": 66750 }, { "epoch": 7.05, - "grad_norm": 12.039482116699219, + "grad_norm": 35.02522659301758, "learning_rate": 5.899250184813603e-06, - "loss": 0.4601, + "loss": 1.0817, "step": 66760 }, { "epoch": 7.05, - "grad_norm": 6.177231311798096, + "grad_norm": 55.54115295410156, "learning_rate": 5.897138029358961e-06, - "loss": 0.4243, + "loss": 1.0754, "step": 66770 }, { "epoch": 7.05, - "grad_norm": 28.121366500854492, + "grad_norm": 50.95652770996094, "learning_rate": 5.89502587390432e-06, - "loss": 0.2539, + "loss": 0.775, "step": 66780 }, { "epoch": 7.05, - "grad_norm": 4.861355304718018, + "grad_norm": 66.31349182128906, "learning_rate": 5.892913718449679e-06, - "loss": 0.3186, + "loss": 1.0796, "step": 66790 }, { "epoch": 7.05, - "grad_norm": 0.940497875213623, + "grad_norm": 14.062159538269043, "learning_rate": 5.890801562995037e-06, - "loss": 0.2609, + "loss": 0.8454, "step": 66800 }, { "epoch": 7.06, - "grad_norm": 0.13149821758270264, + "grad_norm": 34.827720642089844, "learning_rate": 5.8886894075403954e-06, - "loss": 0.2031, + "loss": 1.0614, "step": 66810 }, { "epoch": 7.06, - "grad_norm": 1.6867523193359375, + "grad_norm": 14.149072647094727, "learning_rate": 5.886577252085754e-06, - "loss": 0.4799, + "loss": 0.8419, "step": 66820 }, { "epoch": 7.06, - "grad_norm": 15.123140335083008, + "grad_norm": 77.41722106933594, "learning_rate": 5.884465096631112e-06, - "loss": 0.2088, + "loss": 0.5668, "step": 66830 }, { "epoch": 7.06, - "grad_norm": 4.610142230987549, + "grad_norm": 29.77965545654297, "learning_rate": 5.882352941176471e-06, - "loss": 0.1605, + "loss": 0.7237, "step": 66840 }, { "epoch": 7.06, - "grad_norm": 3.5454959869384766, + "grad_norm": 61.06492614746094, "learning_rate": 5.880240785721829e-06, - "loss": 0.5274, + "loss": 0.9941, "step": 66850 }, { "epoch": 7.06, - "grad_norm": 18.868938446044922, + "grad_norm": 18.70921516418457, "learning_rate": 5.878128630267188e-06, - "loss": 0.5117, + "loss": 0.7338, "step": 66860 }, { "epoch": 7.06, - "grad_norm": 4.468079566955566, + "grad_norm": 55.933597564697266, "learning_rate": 5.876016474812547e-06, - "loss": 0.3888, + "loss": 1.0792, "step": 66870 }, { "epoch": 7.06, - "grad_norm": 15.38021183013916, + "grad_norm": 73.0027084350586, "learning_rate": 5.873904319357905e-06, - "loss": 0.4791, + "loss": 0.9455, "step": 66880 }, { "epoch": 7.06, - "grad_norm": 11.273504257202148, + "grad_norm": 52.422096252441406, "learning_rate": 5.871792163903264e-06, - "loss": 0.2894, + "loss": 1.0012, "step": 66890 }, { "epoch": 7.07, - "grad_norm": 0.1263681948184967, + "grad_norm": 43.61677932739258, "learning_rate": 5.869680008448623e-06, - "loss": 0.3594, + "loss": 0.8988, "step": 66900 }, { "epoch": 7.07, - "grad_norm": 18.558536529541016, + "grad_norm": 39.783935546875, "learning_rate": 5.867567852993981e-06, - "loss": 0.523, + "loss": 1.0693, "step": 66910 }, { "epoch": 7.07, - "grad_norm": 38.592430114746094, + "grad_norm": 54.72893142700195, "learning_rate": 5.8654556975393395e-06, - "loss": 0.285, + "loss": 0.9245, "step": 66920 }, { "epoch": 7.07, - "grad_norm": 6.62232780456543, + "grad_norm": 5.132851600646973, "learning_rate": 5.863343542084697e-06, - "loss": 0.5423, + "loss": 0.9392, "step": 66930 }, { "epoch": 7.07, - "grad_norm": 0.2210407555103302, + "grad_norm": 32.527618408203125, "learning_rate": 5.861231386630056e-06, - "loss": 0.5659, + "loss": 1.0935, "step": 66940 }, { "epoch": 7.07, - "grad_norm": 3.9087836742401123, + "grad_norm": 58.07794189453125, "learning_rate": 5.859119231175415e-06, - "loss": 0.3354, + "loss": 1.0631, "step": 66950 }, { "epoch": 7.07, - "grad_norm": 3.4289231300354004, + "grad_norm": 22.93640899658203, "learning_rate": 5.857007075720773e-06, - "loss": 0.311, + "loss": 0.6902, "step": 66960 }, { "epoch": 7.07, - "grad_norm": 16.55842399597168, + "grad_norm": 44.20817565917969, "learning_rate": 5.854894920266132e-06, - "loss": 0.4105, + "loss": 1.154, "step": 66970 }, { "epoch": 7.07, - "grad_norm": 0.7401496767997742, + "grad_norm": 28.381916046142578, "learning_rate": 5.852782764811491e-06, - "loss": 0.2159, + "loss": 0.7419, "step": 66980 }, { "epoch": 7.07, - "grad_norm": 3.644850969314575, + "grad_norm": 74.22602081298828, "learning_rate": 5.850670609356849e-06, - "loss": 0.2259, + "loss": 0.6828, "step": 66990 }, { "epoch": 7.08, - "grad_norm": 17.817319869995117, + "grad_norm": 31.055479049682617, "learning_rate": 5.848558453902208e-06, - "loss": 0.4379, + "loss": 0.9483, "step": 67000 }, { "epoch": 7.08, - "grad_norm": 0.8227089643478394, + "grad_norm": 39.44198226928711, "learning_rate": 5.846446298447567e-06, - "loss": 0.1447, + "loss": 0.9309, "step": 67010 }, { "epoch": 7.08, - "grad_norm": 14.639389038085938, + "grad_norm": 23.95738983154297, "learning_rate": 5.844334142992925e-06, - "loss": 0.3276, + "loss": 1.0659, "step": 67020 }, { "epoch": 7.08, - "grad_norm": 14.74952220916748, + "grad_norm": 12.955817222595215, "learning_rate": 5.8422219875382835e-06, - "loss": 0.3845, + "loss": 0.6889, "step": 67030 }, { "epoch": 7.08, - "grad_norm": 18.172733306884766, + "grad_norm": 47.527400970458984, "learning_rate": 5.840109832083641e-06, - "loss": 0.2629, + "loss": 0.8679, "step": 67040 }, { "epoch": 7.08, - "grad_norm": 8.968789100646973, + "grad_norm": 32.03734588623047, "learning_rate": 5.837997676629e-06, - "loss": 0.4536, + "loss": 1.0401, "step": 67050 }, { "epoch": 7.08, - "grad_norm": 8.03419017791748, + "grad_norm": 44.83007049560547, "learning_rate": 5.835885521174359e-06, - "loss": 0.4, + "loss": 0.9721, "step": 67060 }, { "epoch": 7.08, - "grad_norm": 4.553641319274902, + "grad_norm": 36.735958099365234, "learning_rate": 5.833773365719717e-06, - "loss": 0.2028, + "loss": 0.9705, "step": 67070 }, { "epoch": 7.08, - "grad_norm": 1.4518046379089355, + "grad_norm": 17.07835578918457, "learning_rate": 5.831661210265076e-06, - "loss": 0.133, + "loss": 0.6044, "step": 67080 }, { "epoch": 7.09, - "grad_norm": 0.18094685673713684, + "grad_norm": 27.33719253540039, "learning_rate": 5.829549054810435e-06, - "loss": 0.4161, + "loss": 0.8692, "step": 67090 }, { "epoch": 7.09, - "grad_norm": 0.26579275727272034, + "grad_norm": 27.75031852722168, "learning_rate": 5.827436899355793e-06, - "loss": 0.4307, + "loss": 1.0197, "step": 67100 }, { "epoch": 7.09, - "grad_norm": 14.57392406463623, + "grad_norm": 37.85744857788086, "learning_rate": 5.825324743901152e-06, - "loss": 0.3899, + "loss": 0.7439, "step": 67110 }, { "epoch": 7.09, - "grad_norm": 0.03423912823200226, + "grad_norm": 48.707069396972656, "learning_rate": 5.82321258844651e-06, - "loss": 0.4394, + "loss": 0.9919, "step": 67120 }, { "epoch": 7.09, - "grad_norm": 6.779207229614258, + "grad_norm": 42.8151969909668, "learning_rate": 5.821100432991869e-06, - "loss": 0.3478, + "loss": 1.1631, "step": 67130 }, { "epoch": 7.09, - "grad_norm": 7.106001853942871, + "grad_norm": 36.15458679199219, "learning_rate": 5.8189882775372275e-06, - "loss": 0.5608, + "loss": 0.8872, "step": 67140 }, { "epoch": 7.09, - "grad_norm": 10.848525047302246, + "grad_norm": 41.72346496582031, "learning_rate": 5.8168761220825854e-06, - "loss": 0.3821, + "loss": 0.9764, "step": 67150 }, { "epoch": 7.09, - "grad_norm": 5.985507011413574, + "grad_norm": 11.462959289550781, "learning_rate": 5.814763966627944e-06, - "loss": 0.3443, + "loss": 0.8361, "step": 67160 }, { "epoch": 7.09, - "grad_norm": 4.668211936950684, + "grad_norm": 19.42753791809082, "learning_rate": 5.812651811173303e-06, - "loss": 0.3703, + "loss": 0.7758, "step": 67170 }, { "epoch": 7.09, - "grad_norm": 0.7108073830604553, + "grad_norm": 40.904056549072266, "learning_rate": 5.810539655718661e-06, - "loss": 0.1318, + "loss": 0.6265, "step": 67180 }, { "epoch": 7.1, - "grad_norm": 21.530595779418945, + "grad_norm": 34.93580627441406, "learning_rate": 5.80842750026402e-06, - "loss": 0.4005, + "loss": 0.7129, "step": 67190 }, { "epoch": 7.1, - "grad_norm": 8.615262031555176, + "grad_norm": 28.589345932006836, "learning_rate": 5.806315344809379e-06, - "loss": 0.4082, + "loss": 1.0241, "step": 67200 }, { "epoch": 7.1, - "grad_norm": 6.550647258758545, + "grad_norm": 35.30396270751953, "learning_rate": 5.804203189354736e-06, - "loss": 0.3218, + "loss": 1.0489, "step": 67210 }, { "epoch": 7.1, - "grad_norm": 0.0736038014292717, + "grad_norm": 55.232688903808594, "learning_rate": 5.802091033900096e-06, - "loss": 0.3442, + "loss": 0.6746, "step": 67220 }, { "epoch": 7.1, - "grad_norm": 7.58587646484375, + "grad_norm": 33.28820037841797, "learning_rate": 5.799978878445454e-06, - "loss": 0.5555, + "loss": 1.1597, "step": 67230 }, { "epoch": 7.1, - "grad_norm": 21.349136352539062, + "grad_norm": 39.77424240112305, "learning_rate": 5.797866722990813e-06, - "loss": 0.4716, + "loss": 0.7767, "step": 67240 }, { "epoch": 7.1, - "grad_norm": 15.878456115722656, + "grad_norm": 47.53200912475586, "learning_rate": 5.7957545675361715e-06, - "loss": 0.3021, + "loss": 1.0093, "step": 67250 }, { "epoch": 7.1, - "grad_norm": 0.691482663154602, + "grad_norm": 23.18947982788086, "learning_rate": 5.7936424120815295e-06, - "loss": 0.3666, + "loss": 0.7374, "step": 67260 }, { "epoch": 7.1, - "grad_norm": 5.03704309463501, + "grad_norm": 35.279571533203125, "learning_rate": 5.791530256626888e-06, - "loss": 0.247, + "loss": 0.8753, "step": 67270 }, { "epoch": 7.11, - "grad_norm": 0.12437430769205093, + "grad_norm": 37.70076370239258, "learning_rate": 5.789418101172247e-06, - "loss": 0.1699, + "loss": 0.7414, "step": 67280 }, { "epoch": 7.11, - "grad_norm": 13.21839714050293, + "grad_norm": 25.564682006835938, "learning_rate": 5.787305945717605e-06, - "loss": 0.384, + "loss": 0.9081, "step": 67290 }, { "epoch": 7.11, - "grad_norm": 15.692770957946777, + "grad_norm": 38.13090133666992, "learning_rate": 5.785193790262964e-06, - "loss": 0.2552, + "loss": 0.775, "step": 67300 }, { "epoch": 7.11, - "grad_norm": 1.1039739847183228, + "grad_norm": 28.518388748168945, "learning_rate": 5.783081634808322e-06, - "loss": 0.533, + "loss": 0.8701, "step": 67310 }, { "epoch": 7.11, - "grad_norm": 20.12672233581543, + "grad_norm": 40.9205322265625, "learning_rate": 5.7809694793536804e-06, - "loss": 0.1691, + "loss": 0.6764, "step": 67320 }, { "epoch": 7.11, - "grad_norm": 2.2770602703094482, + "grad_norm": 35.66160202026367, "learning_rate": 5.77885732389904e-06, - "loss": 0.6006, + "loss": 0.9123, "step": 67330 }, { "epoch": 7.11, - "grad_norm": 0.01304466463625431, + "grad_norm": 2.579089403152466, "learning_rate": 5.776745168444398e-06, - "loss": 0.5624, + "loss": 0.9271, "step": 67340 }, { "epoch": 7.11, - "grad_norm": 4.772544860839844, + "grad_norm": 30.517518997192383, "learning_rate": 5.774633012989757e-06, - "loss": 0.3212, + "loss": 0.9662, "step": 67350 }, { "epoch": 7.11, - "grad_norm": 0.09061288088560104, + "grad_norm": 39.735145568847656, "learning_rate": 5.7725208575351156e-06, - "loss": 0.352, + "loss": 0.6511, "step": 67360 }, { "epoch": 7.11, - "grad_norm": 0.5869126319885254, + "grad_norm": 39.35492706298828, "learning_rate": 5.7704087020804735e-06, - "loss": 0.1934, + "loss": 0.8554, "step": 67370 }, { "epoch": 7.12, - "grad_norm": 24.72789764404297, + "grad_norm": 74.56819915771484, "learning_rate": 5.768296546625832e-06, - "loss": 0.3108, + "loss": 1.2272, "step": 67380 }, { "epoch": 7.12, - "grad_norm": 9.140989303588867, + "grad_norm": 36.50579833984375, "learning_rate": 5.766184391171191e-06, - "loss": 0.1729, + "loss": 0.8757, "step": 67390 }, { "epoch": 7.12, - "grad_norm": 0.16801948845386505, + "grad_norm": 19.709993362426758, "learning_rate": 5.764072235716549e-06, - "loss": 0.3386, + "loss": 0.7671, "step": 67400 }, { "epoch": 7.12, - "grad_norm": 3.3634793758392334, + "grad_norm": 21.445507049560547, "learning_rate": 5.761960080261908e-06, - "loss": 0.5201, + "loss": 1.1413, "step": 67410 }, { "epoch": 7.12, - "grad_norm": 0.11929149925708771, + "grad_norm": 42.07273483276367, "learning_rate": 5.759847924807266e-06, - "loss": 0.3777, + "loss": 0.9671, "step": 67420 }, { "epoch": 7.12, - "grad_norm": 15.178437232971191, + "grad_norm": 48.23757553100586, "learning_rate": 5.7577357693526245e-06, - "loss": 0.3474, + "loss": 0.8451, "step": 67430 }, { "epoch": 7.12, - "grad_norm": 0.02805936150252819, + "grad_norm": 41.31402587890625, "learning_rate": 5.755623613897984e-06, - "loss": 0.4936, + "loss": 0.9485, "step": 67440 }, { "epoch": 7.12, - "grad_norm": 12.837224006652832, + "grad_norm": 28.694889068603516, "learning_rate": 5.753511458443341e-06, - "loss": 0.2802, + "loss": 0.7668, "step": 67450 }, { "epoch": 7.12, - "grad_norm": 1.3966879844665527, + "grad_norm": 31.732681274414062, "learning_rate": 5.751399302988701e-06, - "loss": 0.2355, + "loss": 1.0094, "step": 67460 }, { "epoch": 7.13, - "grad_norm": 9.942917823791504, + "grad_norm": 46.316322326660156, "learning_rate": 5.74928714753406e-06, - "loss": 0.3961, + "loss": 1.069, "step": 67470 }, { "epoch": 7.13, - "grad_norm": 2.9798269271850586, + "grad_norm": 31.204679489135742, "learning_rate": 5.7471749920794175e-06, - "loss": 0.397, + "loss": 1.0474, "step": 67480 }, { "epoch": 7.13, - "grad_norm": 14.62868881225586, + "grad_norm": 58.86695098876953, "learning_rate": 5.745062836624776e-06, - "loss": 0.1174, + "loss": 0.7682, "step": 67490 }, { "epoch": 7.13, - "grad_norm": 0.633051335811615, + "grad_norm": 47.26137161254883, "learning_rate": 5.742950681170134e-06, - "loss": 0.232, + "loss": 1.0267, "step": 67500 }, { "epoch": 7.13, - "grad_norm": 6.949306011199951, + "grad_norm": 24.054826736450195, "learning_rate": 5.740838525715493e-06, - "loss": 0.3066, + "loss": 0.4665, "step": 67510 }, { "epoch": 7.13, - "grad_norm": 20.052291870117188, + "grad_norm": 33.537330627441406, "learning_rate": 5.738726370260852e-06, - "loss": 0.4411, + "loss": 0.859, "step": 67520 }, { "epoch": 7.13, - "grad_norm": 8.050630569458008, + "grad_norm": 57.894447326660156, "learning_rate": 5.73661421480621e-06, - "loss": 0.4863, + "loss": 0.8808, "step": 67530 }, { "epoch": 7.13, - "grad_norm": 14.620454788208008, + "grad_norm": 38.47071075439453, "learning_rate": 5.7345020593515685e-06, - "loss": 0.4781, + "loss": 1.1599, "step": 67540 }, { "epoch": 7.13, - "grad_norm": 0.5534293055534363, + "grad_norm": 22.259536743164062, "learning_rate": 5.732389903896928e-06, - "loss": 0.2918, + "loss": 0.9106, "step": 67550 }, { "epoch": 7.13, - "grad_norm": 2.27121901512146, + "grad_norm": 36.634063720703125, "learning_rate": 5.730277748442285e-06, - "loss": 0.4404, + "loss": 1.0314, "step": 67560 }, { "epoch": 7.14, - "grad_norm": 18.75628662109375, + "grad_norm": 32.062252044677734, "learning_rate": 5.728165592987645e-06, - "loss": 0.238, + "loss": 0.7319, "step": 67570 }, { "epoch": 7.14, - "grad_norm": 0.6850299835205078, + "grad_norm": 49.98591995239258, "learning_rate": 5.726053437533003e-06, - "loss": 0.2912, + "loss": 1.0143, "step": 67580 }, { "epoch": 7.14, - "grad_norm": 4.44368839263916, + "grad_norm": 59.6809196472168, "learning_rate": 5.7239412820783615e-06, - "loss": 0.5852, + "loss": 1.0772, "step": 67590 }, { "epoch": 7.14, - "grad_norm": 16.503713607788086, + "grad_norm": 43.200870513916016, "learning_rate": 5.72182912662372e-06, - "loss": 0.3994, + "loss": 0.9723, "step": 67600 }, { "epoch": 7.14, - "grad_norm": 3.742495059967041, + "grad_norm": 42.692996978759766, "learning_rate": 5.719716971169078e-06, - "loss": 0.2082, + "loss": 1.2313, "step": 67610 }, { "epoch": 7.14, - "grad_norm": 5.510141372680664, + "grad_norm": 22.475086212158203, "learning_rate": 5.717604815714437e-06, - "loss": 0.2395, + "loss": 0.6585, "step": 67620 }, { "epoch": 7.14, - "grad_norm": 0.17426970601081848, + "grad_norm": 38.7330207824707, "learning_rate": 5.715492660259796e-06, - "loss": 0.3107, + "loss": 1.2183, "step": 67630 }, { "epoch": 7.14, - "grad_norm": 14.663980484008789, + "grad_norm": 32.498966217041016, "learning_rate": 5.713380504805154e-06, - "loss": 0.603, + "loss": 0.959, "step": 67640 }, { "epoch": 7.14, - "grad_norm": 21.84946632385254, + "grad_norm": 69.11405181884766, "learning_rate": 5.7112683493505125e-06, - "loss": 0.3885, + "loss": 0.8358, "step": 67650 }, { "epoch": 7.15, - "grad_norm": 14.065421104431152, + "grad_norm": 36.69020462036133, "learning_rate": 5.709156193895871e-06, - "loss": 0.1143, + "loss": 0.8098, "step": 67660 }, { "epoch": 7.15, - "grad_norm": 7.463962554931641, + "grad_norm": 52.7716178894043, "learning_rate": 5.707044038441229e-06, - "loss": 0.4405, + "loss": 1.0702, "step": 67670 }, { "epoch": 7.15, - "grad_norm": 0.49059587717056274, + "grad_norm": 27.583209991455078, "learning_rate": 5.704931882986589e-06, - "loss": 0.2842, + "loss": 0.9947, "step": 67680 }, { "epoch": 7.15, - "grad_norm": 27.71459197998047, + "grad_norm": 42.41189956665039, "learning_rate": 5.702819727531946e-06, - "loss": 0.4008, + "loss": 1.3115, "step": 67690 }, { "epoch": 7.15, - "grad_norm": 24.062255859375, + "grad_norm": 20.55999183654785, "learning_rate": 5.7007075720773056e-06, - "loss": 0.4902, + "loss": 0.9434, "step": 67700 }, { "epoch": 7.15, - "grad_norm": 27.066537857055664, + "grad_norm": 52.36477279663086, "learning_rate": 5.698595416622664e-06, - "loss": 0.4009, + "loss": 1.143, "step": 67710 }, { "epoch": 7.15, - "grad_norm": 0.04438217729330063, + "grad_norm": 23.298389434814453, "learning_rate": 5.696483261168022e-06, - "loss": 0.339, + "loss": 0.6324, "step": 67720 }, { "epoch": 7.15, - "grad_norm": 9.145293235778809, + "grad_norm": 43.2011604309082, "learning_rate": 5.694371105713381e-06, - "loss": 0.339, + "loss": 0.9719, "step": 67730 }, { "epoch": 7.15, - "grad_norm": 1.3175045251846313, + "grad_norm": 39.331790924072266, "learning_rate": 5.69225895025874e-06, - "loss": 0.159, + "loss": 0.7588, "step": 67740 }, { "epoch": 7.15, - "grad_norm": 18.022863388061523, + "grad_norm": 23.470319747924805, "learning_rate": 5.690146794804098e-06, - "loss": 0.2865, + "loss": 0.6804, "step": 67750 }, { "epoch": 7.16, - "grad_norm": 0.026398243382573128, + "grad_norm": 36.037986755371094, "learning_rate": 5.6880346393494566e-06, - "loss": 0.2555, + "loss": 0.803, "step": 67760 }, { "epoch": 7.16, - "grad_norm": 26.75421714782715, + "grad_norm": 56.319801330566406, "learning_rate": 5.6859224838948145e-06, - "loss": 0.5513, + "loss": 1.034, "step": 67770 }, { "epoch": 7.16, - "grad_norm": 0.2043883055448532, + "grad_norm": 45.87761688232422, "learning_rate": 5.683810328440173e-06, - "loss": 0.245, + "loss": 0.6543, "step": 67780 }, { "epoch": 7.16, - "grad_norm": 31.050209045410156, + "grad_norm": 17.762319564819336, "learning_rate": 5.681698172985533e-06, - "loss": 0.2806, + "loss": 0.9632, "step": 67790 }, { "epoch": 7.16, - "grad_norm": 16.589065551757812, + "grad_norm": 23.396194458007812, "learning_rate": 5.67958601753089e-06, - "loss": 0.3781, + "loss": 1.3278, "step": 67800 }, { "epoch": 7.16, - "grad_norm": 0.11226998269557953, + "grad_norm": 23.613733291625977, "learning_rate": 5.67747386207625e-06, - "loss": 0.1605, + "loss": 0.7651, "step": 67810 }, { "epoch": 7.16, - "grad_norm": 0.6698103547096252, + "grad_norm": 48.720672607421875, "learning_rate": 5.675361706621608e-06, - "loss": 0.334, + "loss": 0.9577, "step": 67820 }, { "epoch": 7.16, - "grad_norm": 0.21228773891925812, + "grad_norm": 40.38692092895508, "learning_rate": 5.673249551166966e-06, - "loss": 0.422, + "loss": 1.2177, "step": 67830 }, { "epoch": 7.16, - "grad_norm": 9.044273376464844, + "grad_norm": 53.855735778808594, "learning_rate": 5.671137395712325e-06, - "loss": 0.3319, + "loss": 0.7714, "step": 67840 }, { "epoch": 7.17, - "grad_norm": 18.620763778686523, + "grad_norm": 57.66400146484375, "learning_rate": 5.669025240257684e-06, - "loss": 0.4007, + "loss": 0.9175, "step": 67850 }, { "epoch": 7.17, - "grad_norm": 0.09709101170301437, + "grad_norm": 51.915828704833984, "learning_rate": 5.666913084803042e-06, - "loss": 0.3057, + "loss": 0.7391, "step": 67860 }, { "epoch": 7.17, - "grad_norm": 1.4253449440002441, + "grad_norm": 46.57725143432617, "learning_rate": 5.664800929348401e-06, - "loss": 0.1883, + "loss": 0.8971, "step": 67870 }, { "epoch": 7.17, - "grad_norm": 7.55907678604126, + "grad_norm": 45.4176139831543, "learning_rate": 5.6626887738937585e-06, - "loss": 0.2149, + "loss": 1.1175, "step": 67880 }, { "epoch": 7.17, - "grad_norm": 23.86813735961914, + "grad_norm": 43.00972366333008, "learning_rate": 5.660576618439117e-06, - "loss": 0.3643, + "loss": 0.9888, "step": 67890 }, { "epoch": 7.17, - "grad_norm": 3.9185290336608887, + "grad_norm": 42.53841781616211, "learning_rate": 5.658464462984477e-06, - "loss": 0.2322, + "loss": 0.8815, "step": 67900 }, { "epoch": 7.17, - "grad_norm": 18.57627296447754, + "grad_norm": 40.42158508300781, "learning_rate": 5.656352307529834e-06, - "loss": 0.5666, + "loss": 0.9359, "step": 67910 }, { "epoch": 7.17, - "grad_norm": 21.28175163269043, + "grad_norm": 22.53380584716797, "learning_rate": 5.654240152075194e-06, - "loss": 0.2158, + "loss": 0.9748, "step": 67920 }, { "epoch": 7.17, - "grad_norm": 16.302549362182617, + "grad_norm": 30.228473663330078, "learning_rate": 5.652127996620552e-06, - "loss": 0.4488, + "loss": 0.9278, "step": 67930 }, { "epoch": 7.17, - "grad_norm": 15.022595405578613, + "grad_norm": 53.40456008911133, "learning_rate": 5.65001584116591e-06, - "loss": 0.3154, + "loss": 0.754, "step": 67940 }, { "epoch": 7.18, - "grad_norm": 18.031177520751953, + "grad_norm": 42.007469177246094, "learning_rate": 5.647903685711269e-06, - "loss": 0.4219, + "loss": 1.0772, "step": 67950 }, { "epoch": 7.18, - "grad_norm": 12.012528419494629, + "grad_norm": 21.179306030273438, "learning_rate": 5.645791530256627e-06, - "loss": 0.2961, + "loss": 0.9329, "step": 67960 }, { "epoch": 7.18, - "grad_norm": 0.1607952117919922, + "grad_norm": 50.84186935424805, "learning_rate": 5.643679374801986e-06, - "loss": 0.3328, + "loss": 0.9202, "step": 67970 }, { "epoch": 7.18, - "grad_norm": 12.061984062194824, + "grad_norm": 19.05703353881836, "learning_rate": 5.641567219347345e-06, - "loss": 0.3736, + "loss": 0.6411, "step": 67980 }, { "epoch": 7.18, - "grad_norm": 1.8624169826507568, + "grad_norm": 24.12449836730957, "learning_rate": 5.6394550638927025e-06, - "loss": 0.2251, + "loss": 0.9725, "step": 67990 }, { "epoch": 7.18, - "grad_norm": 0.9396217465400696, + "grad_norm": 28.470111846923828, "learning_rate": 5.637342908438061e-06, - "loss": 0.3371, + "loss": 0.889, "step": 68000 }, { "epoch": 7.18, - "grad_norm": 39.89564514160156, + "grad_norm": 37.951290130615234, "learning_rate": 5.63523075298342e-06, - "loss": 0.3752, + "loss": 0.6748, "step": 68010 }, { "epoch": 7.18, - "grad_norm": 0.8151869773864746, + "grad_norm": 46.605316162109375, "learning_rate": 5.633118597528778e-06, - "loss": 0.4964, + "loss": 1.0558, "step": 68020 }, { "epoch": 7.18, - "grad_norm": 22.197826385498047, + "grad_norm": 23.74216651916504, "learning_rate": 5.631006442074138e-06, - "loss": 0.319, + "loss": 0.8415, "step": 68030 }, { "epoch": 7.19, - "grad_norm": 10.205714225769043, + "grad_norm": 38.15515899658203, "learning_rate": 5.628894286619495e-06, - "loss": 0.3314, + "loss": 1.0036, "step": 68040 }, { "epoch": 7.19, - "grad_norm": 6.093006134033203, + "grad_norm": 47.50794982910156, "learning_rate": 5.626782131164854e-06, - "loss": 0.3478, + "loss": 0.8656, "step": 68050 }, { "epoch": 7.19, - "grad_norm": 10.380733489990234, + "grad_norm": 17.3072509765625, "learning_rate": 5.624669975710213e-06, - "loss": 0.1843, + "loss": 0.6895, "step": 68060 }, { "epoch": 7.19, - "grad_norm": 0.025214288383722305, + "grad_norm": 28.475221633911133, "learning_rate": 5.622557820255571e-06, - "loss": 0.3642, + "loss": 0.7808, "step": 68070 }, { "epoch": 7.19, - "grad_norm": 29.383153915405273, + "grad_norm": 41.03083038330078, "learning_rate": 5.62044566480093e-06, - "loss": 0.6093, + "loss": 0.9072, "step": 68080 }, { "epoch": 7.19, - "grad_norm": 0.23004870116710663, + "grad_norm": 51.49325180053711, "learning_rate": 5.618333509346289e-06, - "loss": 0.5528, + "loss": 1.3147, "step": 68090 }, { "epoch": 7.19, - "grad_norm": 16.77360725402832, + "grad_norm": 30.105512619018555, "learning_rate": 5.6162213538916466e-06, - "loss": 0.3277, + "loss": 0.7632, "step": 68100 }, { "epoch": 7.19, - "grad_norm": 5.452636241912842, + "grad_norm": 36.98438262939453, "learning_rate": 5.614109198437005e-06, - "loss": 0.4292, + "loss": 1.1315, "step": 68110 }, { "epoch": 7.19, - "grad_norm": 15.187247276306152, + "grad_norm": 42.9117546081543, "learning_rate": 5.611997042982364e-06, - "loss": 0.4027, + "loss": 1.0058, "step": 68120 }, { "epoch": 7.2, - "grad_norm": 16.331527709960938, + "grad_norm": 27.11249351501465, "learning_rate": 5.609884887527722e-06, - "loss": 0.2319, + "loss": 0.7174, "step": 68130 }, { "epoch": 7.2, - "grad_norm": 14.950335502624512, + "grad_norm": 60.26339340209961, "learning_rate": 5.607772732073082e-06, - "loss": 0.2623, + "loss": 0.7334, "step": 68140 }, { "epoch": 7.2, - "grad_norm": 14.992161750793457, + "grad_norm": 32.67487716674805, "learning_rate": 5.605660576618439e-06, - "loss": 0.342, + "loss": 0.7166, "step": 68150 }, { "epoch": 7.2, - "grad_norm": 10.85394287109375, + "grad_norm": 48.42707061767578, "learning_rate": 5.603548421163798e-06, - "loss": 0.3199, + "loss": 0.8263, "step": 68160 }, { "epoch": 7.2, - "grad_norm": 3.480689287185669, + "grad_norm": 52.51465606689453, "learning_rate": 5.601436265709157e-06, - "loss": 0.2483, + "loss": 1.08, "step": 68170 }, { "epoch": 7.2, - "grad_norm": 2.6095807552337646, + "grad_norm": 23.674118041992188, "learning_rate": 5.599324110254515e-06, - "loss": 0.1855, + "loss": 0.7959, "step": 68180 }, { "epoch": 7.2, - "grad_norm": 19.761760711669922, + "grad_norm": 51.01543426513672, "learning_rate": 5.597211954799874e-06, - "loss": 0.2867, + "loss": 1.1855, "step": 68190 }, { "epoch": 7.2, - "grad_norm": 20.611072540283203, + "grad_norm": 43.782249450683594, "learning_rate": 5.595099799345233e-06, - "loss": 0.2078, + "loss": 0.6123, "step": 68200 }, { "epoch": 7.2, - "grad_norm": 2.8641388416290283, + "grad_norm": 17.342960357666016, "learning_rate": 5.592987643890591e-06, - "loss": 0.4563, + "loss": 0.9776, "step": 68210 }, { "epoch": 7.2, - "grad_norm": 31.94619369506836, + "grad_norm": 49.59049987792969, "learning_rate": 5.590875488435949e-06, - "loss": 0.4667, + "loss": 1.3397, "step": 68220 }, { "epoch": 7.21, - "grad_norm": 4.525259971618652, + "grad_norm": 13.720354080200195, "learning_rate": 5.588763332981307e-06, - "loss": 0.2329, + "loss": 0.8097, "step": 68230 }, { "epoch": 7.21, - "grad_norm": 13.568734169006348, + "grad_norm": 28.757537841796875, "learning_rate": 5.586651177526666e-06, - "loss": 0.3654, + "loss": 0.9043, "step": 68240 }, { "epoch": 7.21, - "grad_norm": 4.413400650024414, + "grad_norm": 33.40190124511719, "learning_rate": 5.584539022072025e-06, - "loss": 0.3735, + "loss": 0.7898, "step": 68250 }, { "epoch": 7.21, - "grad_norm": 0.026153523474931717, + "grad_norm": 14.259047508239746, "learning_rate": 5.582426866617383e-06, - "loss": 0.2835, + "loss": 0.6112, "step": 68260 }, { "epoch": 7.21, - "grad_norm": 6.75555419921875, + "grad_norm": 49.577613830566406, "learning_rate": 5.580314711162742e-06, - "loss": 0.6069, + "loss": 0.8588, "step": 68270 }, { "epoch": 7.21, - "grad_norm": 0.11090301722288132, + "grad_norm": 23.906068801879883, "learning_rate": 5.578202555708101e-06, - "loss": 0.3321, + "loss": 1.0426, "step": 68280 }, { "epoch": 7.21, - "grad_norm": 13.063745498657227, + "grad_norm": 32.61146926879883, "learning_rate": 5.576090400253459e-06, - "loss": 0.4072, + "loss": 1.1189, "step": 68290 }, { "epoch": 7.21, - "grad_norm": 19.914588928222656, + "grad_norm": 19.9682674407959, "learning_rate": 5.573978244798818e-06, - "loss": 0.379, + "loss": 0.744, "step": 68300 }, { "epoch": 7.21, - "grad_norm": 22.883129119873047, + "grad_norm": 45.74843978881836, "learning_rate": 5.571866089344177e-06, - "loss": 0.3912, + "loss": 1.1531, "step": 68310 }, { "epoch": 7.22, - "grad_norm": 11.955366134643555, + "grad_norm": 41.5269660949707, "learning_rate": 5.569753933889535e-06, - "loss": 0.1716, + "loss": 0.7663, "step": 68320 }, { "epoch": 7.22, - "grad_norm": 3.769104480743408, + "grad_norm": 42.6995735168457, "learning_rate": 5.567641778434893e-06, - "loss": 0.3513, + "loss": 1.1975, "step": 68330 }, { "epoch": 7.22, - "grad_norm": 4.760906219482422, + "grad_norm": 36.33418273925781, "learning_rate": 5.565529622980251e-06, - "loss": 0.3308, + "loss": 0.8473, "step": 68340 }, { "epoch": 7.22, - "grad_norm": 4.231441974639893, + "grad_norm": 22.327259063720703, "learning_rate": 5.56341746752561e-06, - "loss": 0.3373, + "loss": 0.663, "step": 68350 }, { "epoch": 7.22, - "grad_norm": 0.3909547030925751, + "grad_norm": 34.027870178222656, "learning_rate": 5.561305312070969e-06, - "loss": 0.4531, + "loss": 1.0192, "step": 68360 }, { "epoch": 7.22, - "grad_norm": 29.6955623626709, + "grad_norm": 34.39958572387695, "learning_rate": 5.559193156616327e-06, - "loss": 0.4926, + "loss": 1.0719, "step": 68370 }, { "epoch": 7.22, - "grad_norm": 16.54426383972168, + "grad_norm": 44.58818817138672, "learning_rate": 5.5570810011616864e-06, - "loss": 0.494, + "loss": 1.0471, "step": 68380 }, { "epoch": 7.22, - "grad_norm": 23.566329956054688, + "grad_norm": 40.53993606567383, "learning_rate": 5.554968845707045e-06, - "loss": 0.4055, + "loss": 0.966, "step": 68390 }, { "epoch": 7.22, - "grad_norm": 6.697642803192139, + "grad_norm": 45.544593811035156, "learning_rate": 5.552856690252403e-06, - "loss": 0.2171, + "loss": 0.986, "step": 68400 }, { "epoch": 7.22, - "grad_norm": 17.136028289794922, + "grad_norm": 42.90753173828125, "learning_rate": 5.550744534797762e-06, - "loss": 0.2928, + "loss": 0.6957, "step": 68410 }, { "epoch": 7.23, - "grad_norm": 6.635745048522949, + "grad_norm": 31.511173248291016, "learning_rate": 5.54863237934312e-06, - "loss": 0.4737, + "loss": 1.0213, "step": 68420 }, { "epoch": 7.23, - "grad_norm": 11.147088050842285, + "grad_norm": 9.780102729797363, "learning_rate": 5.546520223888479e-06, - "loss": 0.4067, + "loss": 0.6151, "step": 68430 }, { "epoch": 7.23, - "grad_norm": 16.578310012817383, + "grad_norm": 63.62641906738281, "learning_rate": 5.544408068433837e-06, - "loss": 0.5657, + "loss": 0.7155, "step": 68440 }, { "epoch": 7.23, - "grad_norm": 15.375920295715332, + "grad_norm": 33.9171257019043, "learning_rate": 5.542295912979195e-06, - "loss": 0.4317, + "loss": 0.8704, "step": 68450 }, { "epoch": 7.23, - "grad_norm": 0.34485501050949097, + "grad_norm": 13.725204467773438, "learning_rate": 5.540183757524554e-06, - "loss": 0.3376, + "loss": 1.0656, "step": 68460 }, { "epoch": 7.23, - "grad_norm": 0.22617672383785248, + "grad_norm": 35.579627990722656, "learning_rate": 5.538071602069913e-06, - "loss": 0.2526, + "loss": 1.0443, "step": 68470 }, { "epoch": 7.23, - "grad_norm": 19.582609176635742, + "grad_norm": 40.47980880737305, "learning_rate": 5.535959446615271e-06, - "loss": 0.3926, + "loss": 0.9601, "step": 68480 }, { "epoch": 7.23, - "grad_norm": 0.4379195272922516, + "grad_norm": 27.676118850708008, "learning_rate": 5.53384729116063e-06, - "loss": 0.4715, + "loss": 0.7416, "step": 68490 }, { "epoch": 7.23, - "grad_norm": 0.3359871506690979, + "grad_norm": 34.6180534362793, "learning_rate": 5.531735135705989e-06, - "loss": 0.3603, + "loss": 0.9625, "step": 68500 }, { "epoch": 7.24, - "grad_norm": 0.29647552967071533, + "grad_norm": 24.712228775024414, "learning_rate": 5.529622980251347e-06, - "loss": 0.7665, + "loss": 1.2221, "step": 68510 }, { "epoch": 7.24, - "grad_norm": 0.022667285054922104, + "grad_norm": 13.448725700378418, "learning_rate": 5.527510824796706e-06, - "loss": 0.3896, + "loss": 0.9829, "step": 68520 }, { "epoch": 7.24, - "grad_norm": 8.528703689575195, + "grad_norm": 41.97696304321289, "learning_rate": 5.525398669342064e-06, - "loss": 0.4933, + "loss": 0.9609, "step": 68530 }, { "epoch": 7.24, - "grad_norm": 28.48029327392578, + "grad_norm": 18.812450408935547, "learning_rate": 5.523286513887423e-06, - "loss": 0.4787, + "loss": 1.0122, "step": 68540 }, { "epoch": 7.24, - "grad_norm": 21.13197135925293, + "grad_norm": 31.249601364135742, "learning_rate": 5.5211743584327814e-06, - "loss": 0.3102, + "loss": 0.888, "step": 68550 }, { "epoch": 7.24, - "grad_norm": 10.717240333557129, + "grad_norm": 60.16324996948242, "learning_rate": 5.519062202978139e-06, - "loss": 0.1014, + "loss": 0.8178, "step": 68560 }, { "epoch": 7.24, - "grad_norm": 46.31016540527344, + "grad_norm": 53.02743911743164, "learning_rate": 5.516950047523498e-06, - "loss": 0.4878, + "loss": 0.817, "step": 68570 }, { "epoch": 7.24, - "grad_norm": 0.62810879945755, + "grad_norm": 40.806549072265625, "learning_rate": 5.514837892068857e-06, - "loss": 0.3932, + "loss": 1.0057, "step": 68580 }, { "epoch": 7.24, - "grad_norm": 14.029850959777832, + "grad_norm": 51.38749694824219, "learning_rate": 5.512725736614215e-06, - "loss": 0.369, + "loss": 0.9856, "step": 68590 }, { "epoch": 7.24, - "grad_norm": 32.655982971191406, + "grad_norm": 46.00883102416992, "learning_rate": 5.510613581159574e-06, - "loss": 0.3546, + "loss": 0.9072, "step": 68600 }, { "epoch": 7.25, - "grad_norm": 6.6334686279296875, + "grad_norm": 33.5282096862793, "learning_rate": 5.508501425704932e-06, - "loss": 0.3075, + "loss": 0.6853, "step": 68610 }, { "epoch": 7.25, - "grad_norm": 7.246676921844482, + "grad_norm": 27.59615135192871, "learning_rate": 5.506389270250291e-06, - "loss": 0.3599, + "loss": 0.7777, "step": 68620 }, { "epoch": 7.25, - "grad_norm": 18.716800689697266, + "grad_norm": 42.363346099853516, "learning_rate": 5.50427711479565e-06, - "loss": 0.3147, + "loss": 0.5679, "step": 68630 }, { "epoch": 7.25, - "grad_norm": 2.4147558212280273, + "grad_norm": 39.56947708129883, "learning_rate": 5.502164959341008e-06, - "loss": 0.3606, + "loss": 1.1584, "step": 68640 }, { "epoch": 7.25, - "grad_norm": 5.960935592651367, + "grad_norm": 20.86467933654785, "learning_rate": 5.500052803886367e-06, - "loss": 0.4229, + "loss": 0.7597, "step": 68650 }, { "epoch": 7.25, - "grad_norm": 38.3817138671875, + "grad_norm": 36.292972564697266, "learning_rate": 5.4979406484317255e-06, - "loss": 0.3777, + "loss": 0.9719, "step": 68660 }, { "epoch": 7.25, - "grad_norm": 20.506973266601562, + "grad_norm": 41.56440353393555, "learning_rate": 5.495828492977083e-06, - "loss": 0.2346, + "loss": 0.7948, "step": 68670 }, { "epoch": 7.25, - "grad_norm": 13.490802764892578, + "grad_norm": 46.209598541259766, "learning_rate": 5.493716337522442e-06, - "loss": 0.4173, + "loss": 1.2683, "step": 68680 }, { "epoch": 7.25, - "grad_norm": 4.197686672210693, + "grad_norm": 52.30986022949219, "learning_rate": 5.4916041820678e-06, - "loss": 0.2084, + "loss": 0.7452, "step": 68690 }, { "epoch": 7.26, - "grad_norm": 8.929634094238281, + "grad_norm": 31.98242950439453, "learning_rate": 5.489492026613159e-06, - "loss": 0.2392, + "loss": 0.7944, "step": 68700 }, { "epoch": 7.26, - "grad_norm": 13.485166549682617, + "grad_norm": 70.3734359741211, "learning_rate": 5.487379871158518e-06, - "loss": 0.3747, + "loss": 1.0298, "step": 68710 }, { "epoch": 7.26, - "grad_norm": 30.975162506103516, + "grad_norm": 37.94429016113281, "learning_rate": 5.485267715703876e-06, - "loss": 0.3747, + "loss": 0.7945, "step": 68720 }, { "epoch": 7.26, - "grad_norm": 14.53195858001709, + "grad_norm": 15.840574264526367, "learning_rate": 5.483155560249234e-06, - "loss": 0.5385, + "loss": 0.8779, "step": 68730 }, { "epoch": 7.26, - "grad_norm": 9.255908966064453, + "grad_norm": 43.23701858520508, "learning_rate": 5.481043404794594e-06, - "loss": 0.282, + "loss": 0.7621, "step": 68740 }, { "epoch": 7.26, - "grad_norm": 10.497340202331543, + "grad_norm": 44.20004653930664, "learning_rate": 5.478931249339952e-06, - "loss": 0.5669, + "loss": 0.9901, "step": 68750 }, { "epoch": 7.26, - "grad_norm": 6.81334114074707, + "grad_norm": 41.08810806274414, "learning_rate": 5.476819093885311e-06, - "loss": 0.3657, + "loss": 1.0313, "step": 68760 }, { "epoch": 7.26, - "grad_norm": 4.297449588775635, + "grad_norm": 65.88702392578125, "learning_rate": 5.4747069384306695e-06, - "loss": 0.5321, + "loss": 0.914, "step": 68770 }, { "epoch": 7.26, - "grad_norm": 21.84109115600586, + "grad_norm": 40.99626922607422, "learning_rate": 5.4725947829760274e-06, - "loss": 0.5958, + "loss": 1.0476, "step": 68780 }, { "epoch": 7.26, - "grad_norm": 1.2360563278198242, + "grad_norm": 30.241750717163086, "learning_rate": 5.470482627521386e-06, - "loss": 0.2896, + "loss": 0.7574, "step": 68790 }, { "epoch": 7.27, - "grad_norm": 11.974678039550781, + "grad_norm": 33.03652572631836, "learning_rate": 5.468370472066744e-06, - "loss": 0.4355, + "loss": 0.971, "step": 68800 }, { "epoch": 7.27, - "grad_norm": 12.068493843078613, + "grad_norm": 34.185001373291016, "learning_rate": 5.466258316612103e-06, - "loss": 0.2619, + "loss": 0.6984, "step": 68810 }, { "epoch": 7.27, - "grad_norm": 19.436687469482422, + "grad_norm": 33.06683349609375, "learning_rate": 5.464146161157462e-06, - "loss": 0.3292, + "loss": 0.9861, "step": 68820 }, { "epoch": 7.27, - "grad_norm": 18.296703338623047, + "grad_norm": 35.35822677612305, "learning_rate": 5.46203400570282e-06, - "loss": 0.4746, + "loss": 0.9305, "step": 68830 }, { "epoch": 7.27, - "grad_norm": 18.19053077697754, + "grad_norm": 16.184341430664062, "learning_rate": 5.459921850248178e-06, - "loss": 0.3784, + "loss": 0.8175, "step": 68840 }, { "epoch": 7.27, - "grad_norm": 1.3257125616073608, + "grad_norm": 50.98857498168945, "learning_rate": 5.457809694793538e-06, - "loss": 0.7161, + "loss": 1.1783, "step": 68850 }, { "epoch": 7.27, - "grad_norm": 31.468463897705078, + "grad_norm": 25.46235466003418, "learning_rate": 5.455697539338896e-06, - "loss": 0.7369, + "loss": 0.8825, "step": 68860 }, { "epoch": 7.27, - "grad_norm": 17.703474044799805, + "grad_norm": 34.830223083496094, "learning_rate": 5.453585383884255e-06, - "loss": 0.5692, + "loss": 0.7867, "step": 68870 }, { "epoch": 7.27, - "grad_norm": 15.27010726928711, + "grad_norm": 42.431583404541016, "learning_rate": 5.451473228429613e-06, - "loss": 0.2734, + "loss": 0.8951, "step": 68880 }, { "epoch": 7.28, - "grad_norm": 9.086287498474121, + "grad_norm": 61.010887145996094, "learning_rate": 5.4493610729749715e-06, - "loss": 0.5793, + "loss": 1.1513, "step": 68890 }, { "epoch": 7.28, - "grad_norm": 42.68540573120117, + "grad_norm": 36.9219970703125, "learning_rate": 5.44724891752033e-06, - "loss": 0.5591, + "loss": 1.1637, "step": 68900 }, { "epoch": 7.28, - "grad_norm": 0.03378390520811081, + "grad_norm": 35.90922164916992, "learning_rate": 5.445136762065688e-06, - "loss": 0.1618, + "loss": 0.8143, "step": 68910 }, { "epoch": 7.28, - "grad_norm": 37.558013916015625, + "grad_norm": 35.59183883666992, "learning_rate": 5.443024606611047e-06, - "loss": 0.2732, + "loss": 0.8748, "step": 68920 }, { "epoch": 7.28, - "grad_norm": 0.7086567878723145, + "grad_norm": 31.927082061767578, "learning_rate": 5.440912451156406e-06, - "loss": 0.1408, + "loss": 0.8158, "step": 68930 }, { "epoch": 7.28, - "grad_norm": 27.723596572875977, + "grad_norm": 28.129098892211914, "learning_rate": 5.438800295701764e-06, - "loss": 0.665, + "loss": 0.8871, "step": 68940 }, { "epoch": 7.28, - "grad_norm": 4.5197014808654785, + "grad_norm": 31.340595245361328, "learning_rate": 5.4366881402471224e-06, - "loss": 0.7889, + "loss": 1.2237, "step": 68950 }, { "epoch": 7.28, - "grad_norm": 4.2090606689453125, + "grad_norm": 16.297664642333984, "learning_rate": 5.434575984792482e-06, - "loss": 0.2585, + "loss": 0.9142, "step": 68960 }, { "epoch": 7.28, - "grad_norm": 12.4225492477417, + "grad_norm": 51.34223937988281, "learning_rate": 5.432463829337839e-06, - "loss": 0.5044, + "loss": 0.9203, "step": 68970 }, { "epoch": 7.28, - "grad_norm": 0.01282711885869503, + "grad_norm": 32.21119689941406, "learning_rate": 5.430351673883199e-06, - "loss": 0.3907, + "loss": 0.994, "step": 68980 }, { "epoch": 7.29, - "grad_norm": 25.008277893066406, + "grad_norm": 47.315406799316406, "learning_rate": 5.428239518428557e-06, - "loss": 0.6376, + "loss": 1.3246, "step": 68990 }, { "epoch": 7.29, - "grad_norm": 42.745914459228516, + "grad_norm": 43.53023147583008, "learning_rate": 5.4261273629739155e-06, - "loss": 0.4382, + "loss": 0.7337, "step": 69000 }, { "epoch": 7.29, - "grad_norm": 22.533222198486328, + "grad_norm": 29.712692260742188, "learning_rate": 5.424015207519274e-06, - "loss": 0.4029, + "loss": 1.0325, "step": 69010 }, { "epoch": 7.29, - "grad_norm": 27.65599822998047, + "grad_norm": 14.064704895019531, "learning_rate": 5.421903052064632e-06, - "loss": 0.3566, + "loss": 0.697, "step": 69020 }, { "epoch": 7.29, - "grad_norm": 8.247182846069336, + "grad_norm": 57.55757522583008, "learning_rate": 5.419790896609991e-06, - "loss": 0.3668, + "loss": 0.8821, "step": 69030 }, { "epoch": 7.29, - "grad_norm": 14.459202766418457, + "grad_norm": 43.21059036254883, "learning_rate": 5.41767874115535e-06, - "loss": 0.3777, + "loss": 0.9516, "step": 69040 }, { "epoch": 7.29, - "grad_norm": 11.99595832824707, + "grad_norm": 7.760554313659668, "learning_rate": 5.415566585700708e-06, - "loss": 0.2896, + "loss": 1.1495, "step": 69050 }, { "epoch": 7.29, - "grad_norm": 30.008092880249023, + "grad_norm": 26.092374801635742, "learning_rate": 5.4134544302460665e-06, - "loss": 0.1682, + "loss": 0.8244, "step": 69060 }, { "epoch": 7.29, - "grad_norm": 25.388690948486328, + "grad_norm": 51.883995056152344, "learning_rate": 5.411342274791424e-06, - "loss": 0.3389, + "loss": 0.9745, "step": 69070 }, { "epoch": 7.3, - "grad_norm": 0.17411191761493683, + "grad_norm": 38.16046142578125, "learning_rate": 5.409230119336783e-06, - "loss": 0.5413, + "loss": 0.9935, "step": 69080 }, { "epoch": 7.3, - "grad_norm": 2.9086132049560547, + "grad_norm": 54.461708068847656, "learning_rate": 5.407117963882143e-06, - "loss": 0.3724, + "loss": 1.0088, "step": 69090 }, { "epoch": 7.3, - "grad_norm": 9.643516540527344, + "grad_norm": 46.855804443359375, "learning_rate": 5.405005808427501e-06, - "loss": 0.2731, + "loss": 0.6265, "step": 69100 }, { "epoch": 7.3, - "grad_norm": 10.651134490966797, + "grad_norm": 31.795860290527344, "learning_rate": 5.4028936529728595e-06, - "loss": 0.205, + "loss": 0.7986, "step": 69110 }, { "epoch": 7.3, - "grad_norm": 9.456106185913086, + "grad_norm": 25.72349739074707, "learning_rate": 5.400781497518218e-06, - "loss": 0.2364, + "loss": 0.5539, "step": 69120 }, { "epoch": 7.3, - "grad_norm": 1.6726951599121094, + "grad_norm": 32.15006637573242, "learning_rate": 5.398669342063576e-06, - "loss": 0.5025, + "loss": 0.9256, "step": 69130 }, { "epoch": 7.3, - "grad_norm": 29.944520950317383, + "grad_norm": 52.26784133911133, "learning_rate": 5.396557186608935e-06, - "loss": 0.5602, + "loss": 0.85, "step": 69140 }, { "epoch": 7.3, - "grad_norm": 19.61860466003418, + "grad_norm": 60.553775787353516, "learning_rate": 5.394445031154293e-06, - "loss": 0.3452, + "loss": 0.9726, "step": 69150 }, { "epoch": 7.3, - "grad_norm": 36.85206985473633, + "grad_norm": 38.99833297729492, "learning_rate": 5.392332875699652e-06, - "loss": 0.5198, + "loss": 0.8504, "step": 69160 }, { "epoch": 7.3, - "grad_norm": 18.82908058166504, + "grad_norm": 36.82698059082031, "learning_rate": 5.3902207202450105e-06, - "loss": 0.2295, + "loss": 0.707, "step": 69170 }, { "epoch": 7.31, - "grad_norm": 2.4184982776641846, + "grad_norm": 33.799503326416016, "learning_rate": 5.388108564790368e-06, - "loss": 0.3673, + "loss": 0.9066, "step": 69180 }, { "epoch": 7.31, - "grad_norm": 2.5573270320892334, + "grad_norm": 34.44902801513672, "learning_rate": 5.385996409335727e-06, - "loss": 0.38, + "loss": 0.8773, "step": 69190 }, { "epoch": 7.31, - "grad_norm": 20.477174758911133, + "grad_norm": 33.970863342285156, "learning_rate": 5.383884253881087e-06, - "loss": 0.3385, + "loss": 0.8194, "step": 69200 }, { "epoch": 7.31, - "grad_norm": 1.4868316650390625, + "grad_norm": 33.0562858581543, "learning_rate": 5.381772098426444e-06, - "loss": 0.4282, + "loss": 0.9925, "step": 69210 }, { "epoch": 7.31, - "grad_norm": 2.5126564502716064, + "grad_norm": 36.028018951416016, "learning_rate": 5.3796599429718035e-06, - "loss": 0.4201, + "loss": 0.7138, "step": 69220 }, { "epoch": 7.31, - "grad_norm": 1.2236346006393433, + "grad_norm": 46.69418716430664, "learning_rate": 5.377547787517162e-06, - "loss": 0.4473, + "loss": 1.0282, "step": 69230 }, { "epoch": 7.31, - "grad_norm": 6.963104724884033, + "grad_norm": 31.461009979248047, "learning_rate": 5.37543563206252e-06, - "loss": 0.3277, + "loss": 0.8607, "step": 69240 }, { "epoch": 7.31, - "grad_norm": 14.937129020690918, + "grad_norm": 67.64643859863281, "learning_rate": 5.373323476607879e-06, - "loss": 0.6181, + "loss": 1.299, "step": 69250 }, { "epoch": 7.31, - "grad_norm": 2.0638437271118164, + "grad_norm": 46.46276092529297, "learning_rate": 5.371211321153237e-06, - "loss": 0.3015, + "loss": 1.0313, "step": 69260 }, { "epoch": 7.32, - "grad_norm": 13.397916793823242, + "grad_norm": 27.71922492980957, "learning_rate": 5.369099165698596e-06, - "loss": 0.3599, + "loss": 0.8367, "step": 69270 }, { "epoch": 7.32, - "grad_norm": 0.2785298526287079, + "grad_norm": 24.10725212097168, "learning_rate": 5.3669870102439545e-06, - "loss": 0.2032, + "loss": 0.6799, "step": 69280 }, { "epoch": 7.32, - "grad_norm": 15.149609565734863, + "grad_norm": 42.91289138793945, "learning_rate": 5.3648748547893124e-06, - "loss": 0.4141, + "loss": 0.8131, "step": 69290 }, { "epoch": 7.32, - "grad_norm": 3.3991994857788086, + "grad_norm": 48.22883605957031, "learning_rate": 5.362762699334671e-06, - "loss": 0.6756, + "loss": 1.1862, "step": 69300 }, { "epoch": 7.32, - "grad_norm": 1.4791189432144165, + "grad_norm": 33.235809326171875, "learning_rate": 5.360650543880031e-06, - "loss": 0.1705, + "loss": 1.0309, "step": 69310 }, { "epoch": 7.32, - "grad_norm": 8.833660125732422, + "grad_norm": 46.822853088378906, "learning_rate": 5.358538388425388e-06, - "loss": 0.5225, + "loss": 1.2011, "step": 69320 }, { "epoch": 7.32, - "grad_norm": 6.2280144691467285, + "grad_norm": 45.163516998291016, "learning_rate": 5.3564262329707476e-06, - "loss": 0.4734, + "loss": 1.2052, "step": 69330 }, { "epoch": 7.32, - "grad_norm": 3.7695775032043457, + "grad_norm": 60.74092102050781, "learning_rate": 5.3543140775161055e-06, - "loss": 0.2281, + "loss": 1.1729, "step": 69340 }, { "epoch": 7.32, - "grad_norm": 10.514945983886719, + "grad_norm": 27.832643508911133, "learning_rate": 5.352201922061464e-06, - "loss": 0.2433, + "loss": 1.0564, "step": 69350 }, { "epoch": 7.32, - "grad_norm": 10.240422248840332, + "grad_norm": 36.79805374145508, "learning_rate": 5.350089766606823e-06, - "loss": 0.503, + "loss": 0.9162, "step": 69360 }, { "epoch": 7.33, - "grad_norm": 17.245689392089844, + "grad_norm": 25.60169792175293, "learning_rate": 5.347977611152181e-06, - "loss": 0.3009, + "loss": 0.7199, "step": 69370 }, { "epoch": 7.33, - "grad_norm": 14.493027687072754, + "grad_norm": 40.65415954589844, "learning_rate": 5.34586545569754e-06, - "loss": 0.5066, + "loss": 1.0642, "step": 69380 }, { "epoch": 7.33, - "grad_norm": 46.67549133300781, + "grad_norm": 43.0455436706543, "learning_rate": 5.3437533002428985e-06, - "loss": 0.4415, + "loss": 0.7706, "step": 69390 }, { "epoch": 7.33, - "grad_norm": 15.341392517089844, + "grad_norm": 55.32130432128906, "learning_rate": 5.3416411447882565e-06, - "loss": 0.4797, + "loss": 0.9044, "step": 69400 }, { "epoch": 7.33, - "grad_norm": 18.46896743774414, + "grad_norm": 63.1682014465332, "learning_rate": 5.339528989333615e-06, - "loss": 0.3795, + "loss": 0.9719, "step": 69410 }, { "epoch": 7.33, - "grad_norm": 1.4224448204040527, + "grad_norm": 17.854299545288086, "learning_rate": 5.337416833878975e-06, - "loss": 0.1905, + "loss": 0.7535, "step": 69420 }, { "epoch": 7.33, - "grad_norm": 15.995003700256348, + "grad_norm": 25.084091186523438, "learning_rate": 5.335304678424332e-06, - "loss": 0.2456, + "loss": 0.8553, "step": 69430 }, { "epoch": 7.33, - "grad_norm": 15.30771541595459, + "grad_norm": 44.07707214355469, "learning_rate": 5.333192522969692e-06, - "loss": 0.4034, + "loss": 0.8527, "step": 69440 }, { "epoch": 7.33, - "grad_norm": 5.324433326721191, + "grad_norm": 56.103424072265625, "learning_rate": 5.331080367515049e-06, - "loss": 0.4321, + "loss": 1.1195, "step": 69450 }, { "epoch": 7.34, - "grad_norm": 0.3809046447277069, + "grad_norm": 37.019351959228516, "learning_rate": 5.328968212060408e-06, - "loss": 0.4338, + "loss": 1.0488, "step": 69460 }, { "epoch": 7.34, - "grad_norm": 0.8851180076599121, + "grad_norm": 73.64967346191406, "learning_rate": 5.326856056605767e-06, - "loss": 0.2717, + "loss": 1.2107, "step": 69470 }, { "epoch": 7.34, - "grad_norm": 20.296661376953125, + "grad_norm": 25.656526565551758, "learning_rate": 5.324743901151125e-06, - "loss": 0.7009, + "loss": 1.1478, "step": 69480 }, { "epoch": 7.34, - "grad_norm": 2.004207134246826, + "grad_norm": 34.84573745727539, "learning_rate": 5.322631745696484e-06, - "loss": 0.6189, + "loss": 0.9656, "step": 69490 }, { "epoch": 7.34, - "grad_norm": 8.92926025390625, + "grad_norm": 66.45175170898438, "learning_rate": 5.3205195902418426e-06, - "loss": 0.3811, + "loss": 0.7269, "step": 69500 }, { "epoch": 7.34, - "grad_norm": 33.66132736206055, + "grad_norm": 59.21388244628906, "learning_rate": 5.3184074347872005e-06, - "loss": 0.6241, + "loss": 0.9895, "step": 69510 }, { "epoch": 7.34, - "grad_norm": 23.178707122802734, + "grad_norm": 42.332672119140625, "learning_rate": 5.316295279332559e-06, - "loss": 0.2113, + "loss": 0.7281, "step": 69520 }, { "epoch": 7.34, - "grad_norm": 44.798439025878906, + "grad_norm": 34.2385139465332, "learning_rate": 5.314183123877917e-06, - "loss": 0.3562, + "loss": 0.9943, "step": 69530 }, { "epoch": 7.34, - "grad_norm": 6.201103687286377, + "grad_norm": 51.74039840698242, "learning_rate": 5.312070968423276e-06, - "loss": 0.4968, + "loss": 0.6719, "step": 69540 }, { "epoch": 7.35, - "grad_norm": 19.999374389648438, + "grad_norm": 32.58816146850586, "learning_rate": 5.309958812968636e-06, - "loss": 0.5243, + "loss": 1.1341, "step": 69550 }, { "epoch": 7.35, - "grad_norm": 11.518562316894531, + "grad_norm": 46.889522552490234, "learning_rate": 5.307846657513993e-06, - "loss": 0.5989, + "loss": 1.0503, "step": 69560 }, { "epoch": 7.35, - "grad_norm": 0.20418192446231842, + "grad_norm": 37.0027961730957, "learning_rate": 5.305734502059352e-06, - "loss": 0.3333, + "loss": 0.8272, "step": 69570 }, { "epoch": 7.35, - "grad_norm": 3.6665420532226562, + "grad_norm": 50.936920166015625, "learning_rate": 5.303622346604711e-06, - "loss": 0.186, + "loss": 0.7084, "step": 69580 }, { "epoch": 7.35, - "grad_norm": 0.1262924075126648, + "grad_norm": 40.139408111572266, "learning_rate": 5.301510191150069e-06, - "loss": 0.2695, + "loss": 0.8216, "step": 69590 }, { "epoch": 7.35, - "grad_norm": 16.995391845703125, + "grad_norm": 50.69667053222656, "learning_rate": 5.299398035695428e-06, - "loss": 0.5747, + "loss": 0.997, "step": 69600 }, { "epoch": 7.35, - "grad_norm": 8.578633308410645, + "grad_norm": 21.863035202026367, "learning_rate": 5.297285880240786e-06, - "loss": 0.3619, + "loss": 0.9312, "step": 69610 }, { "epoch": 7.35, - "grad_norm": 1.8620444536209106, + "grad_norm": 44.809268951416016, "learning_rate": 5.2951737247861445e-06, - "loss": 0.2391, + "loss": 1.12, "step": 69620 }, { "epoch": 7.35, - "grad_norm": 1.123552918434143, + "grad_norm": 61.858245849609375, "learning_rate": 5.293061569331503e-06, - "loss": 0.0984, + "loss": 0.9633, "step": 69630 }, { "epoch": 7.35, - "grad_norm": 26.77765655517578, + "grad_norm": 48.246421813964844, "learning_rate": 5.290949413876861e-06, - "loss": 0.4831, + "loss": 0.9126, "step": 69640 }, { "epoch": 7.36, - "grad_norm": 0.12450088560581207, + "grad_norm": 16.893762588500977, "learning_rate": 5.28883725842222e-06, - "loss": 0.1978, + "loss": 0.7078, "step": 69650 }, { "epoch": 7.36, - "grad_norm": 34.63678741455078, + "grad_norm": 42.88203811645508, "learning_rate": 5.28672510296758e-06, - "loss": 0.2947, + "loss": 0.9476, "step": 69660 }, { "epoch": 7.36, - "grad_norm": 2.687354564666748, + "grad_norm": 18.283462524414062, "learning_rate": 5.284612947512937e-06, - "loss": 0.4369, + "loss": 0.8816, "step": 69670 }, { "epoch": 7.36, - "grad_norm": 0.38502904772758484, + "grad_norm": 37.38633728027344, "learning_rate": 5.282500792058296e-06, - "loss": 0.5153, + "loss": 0.7836, "step": 69680 }, { "epoch": 7.36, - "grad_norm": 2.2844741344451904, + "grad_norm": 37.61014938354492, "learning_rate": 5.280388636603655e-06, - "loss": 0.2754, + "loss": 0.9846, "step": 69690 }, { "epoch": 7.36, - "grad_norm": 0.41071879863739014, + "grad_norm": 9.885004043579102, "learning_rate": 5.278276481149013e-06, - "loss": 0.2307, + "loss": 0.8427, "step": 69700 }, { "epoch": 7.36, - "grad_norm": 17.70787239074707, + "grad_norm": 46.80490493774414, "learning_rate": 5.276164325694372e-06, - "loss": 0.2098, + "loss": 0.8805, "step": 69710 }, { "epoch": 7.36, - "grad_norm": 5.582122325897217, + "grad_norm": 34.62688064575195, "learning_rate": 5.27405217023973e-06, - "loss": 0.4295, + "loss": 0.9412, "step": 69720 }, { "epoch": 7.36, - "grad_norm": 12.976913452148438, + "grad_norm": 38.703975677490234, "learning_rate": 5.2719400147850885e-06, - "loss": 0.3127, + "loss": 0.8149, "step": 69730 }, { "epoch": 7.37, - "grad_norm": 0.4541861414909363, + "grad_norm": 10.665790557861328, "learning_rate": 5.269827859330447e-06, - "loss": 0.264, + "loss": 0.6249, "step": 69740 }, { "epoch": 7.37, - "grad_norm": 7.020536422729492, + "grad_norm": 27.139404296875, "learning_rate": 5.267715703875805e-06, - "loss": 0.3079, + "loss": 0.7555, "step": 69750 }, { "epoch": 7.37, - "grad_norm": 0.19495517015457153, + "grad_norm": 46.308326721191406, "learning_rate": 5.265603548421164e-06, - "loss": 0.1353, + "loss": 0.5925, "step": 69760 }, { "epoch": 7.37, - "grad_norm": 9.616376876831055, + "grad_norm": 58.545921325683594, "learning_rate": 5.263491392966523e-06, - "loss": 0.1744, + "loss": 0.9509, "step": 69770 }, { "epoch": 7.37, - "grad_norm": 0.38712602853775024, + "grad_norm": 28.64258575439453, "learning_rate": 5.261379237511881e-06, - "loss": 0.0665, + "loss": 0.9949, "step": 69780 }, { "epoch": 7.37, - "grad_norm": 21.657236099243164, + "grad_norm": 41.97322463989258, "learning_rate": 5.25926708205724e-06, - "loss": 0.3393, + "loss": 0.7451, "step": 69790 }, { "epoch": 7.37, - "grad_norm": 10.124736785888672, + "grad_norm": 42.87397766113281, "learning_rate": 5.2571549266025975e-06, - "loss": 0.3906, + "loss": 1.3553, "step": 69800 }, { "epoch": 7.37, - "grad_norm": 46.73078155517578, + "grad_norm": 58.70263671875, "learning_rate": 5.255042771147957e-06, - "loss": 0.2306, + "loss": 0.8706, "step": 69810 }, { "epoch": 7.37, - "grad_norm": 11.982359886169434, + "grad_norm": 33.247589111328125, "learning_rate": 5.252930615693316e-06, - "loss": 0.3276, + "loss": 0.6958, "step": 69820 }, { "epoch": 7.37, - "grad_norm": 15.671046257019043, + "grad_norm": 61.25054168701172, "learning_rate": 5.250818460238674e-06, - "loss": 0.202, + "loss": 1.011, "step": 69830 }, { "epoch": 7.38, - "grad_norm": 53.24976348876953, + "grad_norm": 39.8758659362793, "learning_rate": 5.2487063047840326e-06, - "loss": 0.3466, + "loss": 1.0985, "step": 69840 }, { "epoch": 7.38, - "grad_norm": 31.342893600463867, + "grad_norm": 41.062625885009766, "learning_rate": 5.246594149329391e-06, - "loss": 0.4178, + "loss": 1.0011, "step": 69850 }, { "epoch": 7.38, - "grad_norm": 37.2110710144043, + "grad_norm": 33.92134094238281, "learning_rate": 5.244481993874749e-06, - "loss": 0.2963, + "loss": 0.7681, "step": 69860 }, { "epoch": 7.38, - "grad_norm": 29.835071563720703, + "grad_norm": 54.822357177734375, "learning_rate": 5.242369838420108e-06, - "loss": 0.2621, + "loss": 0.7399, "step": 69870 }, { "epoch": 7.38, - "grad_norm": 25.78588104248047, + "grad_norm": 30.859769821166992, "learning_rate": 5.240257682965467e-06, - "loss": 0.2529, + "loss": 0.7915, "step": 69880 }, { "epoch": 7.38, - "grad_norm": 19.30722999572754, + "grad_norm": 39.796382904052734, "learning_rate": 5.238145527510825e-06, - "loss": 0.5121, + "loss": 1.0337, "step": 69890 }, { "epoch": 7.38, - "grad_norm": 0.3608201742172241, + "grad_norm": 36.39160919189453, "learning_rate": 5.236033372056184e-06, - "loss": 0.4727, + "loss": 0.9018, "step": 69900 }, { "epoch": 7.38, - "grad_norm": 22.644880294799805, + "grad_norm": 16.96605110168457, "learning_rate": 5.2339212166015415e-06, - "loss": 0.2006, + "loss": 0.8645, "step": 69910 }, { "epoch": 7.38, - "grad_norm": 7.782045841217041, + "grad_norm": 35.335350036621094, "learning_rate": 5.231809061146901e-06, - "loss": 0.2937, + "loss": 0.7624, "step": 69920 }, { "epoch": 7.39, - "grad_norm": 1.7963292598724365, + "grad_norm": 35.39553451538086, "learning_rate": 5.22969690569226e-06, - "loss": 0.351, + "loss": 0.9842, "step": 69930 }, { "epoch": 7.39, - "grad_norm": 24.776824951171875, + "grad_norm": 33.34782409667969, "learning_rate": 5.227584750237618e-06, - "loss": 0.6191, + "loss": 0.9079, "step": 69940 }, { "epoch": 7.39, - "grad_norm": 3.244147539138794, + "grad_norm": 31.694698333740234, "learning_rate": 5.225472594782977e-06, - "loss": 0.6319, + "loss": 0.9553, "step": 69950 }, { "epoch": 7.39, - "grad_norm": 8.186399459838867, + "grad_norm": 28.180421829223633, "learning_rate": 5.223360439328335e-06, - "loss": 0.1098, + "loss": 0.6634, "step": 69960 }, { "epoch": 7.39, - "grad_norm": 18.862293243408203, + "grad_norm": 21.07954216003418, "learning_rate": 5.221248283873693e-06, - "loss": 0.4984, + "loss": 1.1404, "step": 69970 }, { "epoch": 7.39, - "grad_norm": 18.735795974731445, + "grad_norm": 46.737648010253906, "learning_rate": 5.219136128419052e-06, - "loss": 0.3047, + "loss": 1.1243, "step": 69980 }, { "epoch": 7.39, - "grad_norm": 12.836591720581055, + "grad_norm": 31.079692840576172, "learning_rate": 5.21702397296441e-06, - "loss": 0.5112, + "loss": 1.0655, "step": 69990 }, { "epoch": 7.39, - "grad_norm": 25.92136001586914, + "grad_norm": 43.30638122558594, "learning_rate": 5.214911817509769e-06, - "loss": 0.5573, + "loss": 0.8721, "step": 70000 }, { "epoch": 7.39, - "grad_norm": 0.27757254242897034, + "grad_norm": 27.127681732177734, "learning_rate": 5.212799662055128e-06, - "loss": 0.3199, + "loss": 0.8783, "step": 70010 }, { "epoch": 7.39, - "grad_norm": 2.2881484031677246, + "grad_norm": 75.5394515991211, "learning_rate": 5.2106875066004855e-06, - "loss": 0.7978, + "loss": 1.1377, "step": 70020 }, { "epoch": 7.4, - "grad_norm": 0.09568437933921814, + "grad_norm": 7.433651447296143, "learning_rate": 5.208575351145845e-06, - "loss": 0.2466, + "loss": 0.8573, "step": 70030 }, { "epoch": 7.4, - "grad_norm": 0.19569598138332367, + "grad_norm": 29.166120529174805, "learning_rate": 5.206463195691204e-06, - "loss": 0.3887, + "loss": 0.7661, "step": 70040 }, { "epoch": 7.4, - "grad_norm": 3.1219048500061035, + "grad_norm": 40.5748291015625, "learning_rate": 5.204351040236562e-06, - "loss": 0.3111, + "loss": 0.715, "step": 70050 }, { "epoch": 7.4, - "grad_norm": 19.53557777404785, + "grad_norm": 34.598873138427734, "learning_rate": 5.202238884781921e-06, - "loss": 0.347, + "loss": 0.8363, "step": 70060 }, { "epoch": 7.4, - "grad_norm": 17.597078323364258, + "grad_norm": 54.920528411865234, "learning_rate": 5.200126729327279e-06, - "loss": 0.3903, + "loss": 1.1726, "step": 70070 }, { "epoch": 7.4, - "grad_norm": 8.124517440795898, + "grad_norm": 45.56358337402344, "learning_rate": 5.198014573872637e-06, - "loss": 0.1947, + "loss": 0.7691, "step": 70080 }, { "epoch": 7.4, - "grad_norm": 24.900222778320312, + "grad_norm": 38.63261795043945, "learning_rate": 5.195902418417996e-06, - "loss": 0.2237, + "loss": 0.9617, "step": 70090 }, { "epoch": 7.4, - "grad_norm": 0.987128734588623, + "grad_norm": 31.72807502746582, "learning_rate": 5.193790262963354e-06, - "loss": 0.3419, + "loss": 1.0001, "step": 70100 }, { "epoch": 7.4, - "grad_norm": 16.18916130065918, + "grad_norm": 42.10211181640625, "learning_rate": 5.191678107508713e-06, - "loss": 0.2436, + "loss": 0.8829, "step": 70110 }, { "epoch": 7.41, - "grad_norm": 0.33473679423332214, + "grad_norm": 61.95499038696289, "learning_rate": 5.189565952054072e-06, - "loss": 0.4438, + "loss": 1.1317, "step": 70120 }, { "epoch": 7.41, - "grad_norm": 19.852603912353516, + "grad_norm": 39.48551940917969, "learning_rate": 5.1874537965994295e-06, - "loss": 0.3802, + "loss": 1.1155, "step": 70130 }, { "epoch": 7.41, - "grad_norm": 2.0923473834991455, + "grad_norm": 33.78645706176758, "learning_rate": 5.185341641144789e-06, - "loss": 0.3476, + "loss": 0.6948, "step": 70140 }, { "epoch": 7.41, - "grad_norm": 8.209100723266602, + "grad_norm": 25.771865844726562, "learning_rate": 5.183229485690148e-06, - "loss": 0.3538, + "loss": 0.9031, "step": 70150 }, { "epoch": 7.41, - "grad_norm": 2.9124972820281982, + "grad_norm": 50.22251892089844, "learning_rate": 5.181117330235506e-06, - "loss": 0.222, + "loss": 0.8858, "step": 70160 }, { "epoch": 7.41, - "grad_norm": 20.706016540527344, + "grad_norm": 31.267242431640625, "learning_rate": 5.179005174780865e-06, - "loss": 0.6436, + "loss": 0.7584, "step": 70170 }, { "epoch": 7.41, - "grad_norm": 9.758909225463867, + "grad_norm": 68.85018920898438, "learning_rate": 5.176893019326223e-06, - "loss": 0.344, + "loss": 0.9731, "step": 70180 }, { "epoch": 7.41, - "grad_norm": 4.750969409942627, + "grad_norm": 29.658580780029297, "learning_rate": 5.174780863871581e-06, - "loss": 0.1174, + "loss": 0.9225, "step": 70190 }, { "epoch": 7.41, - "grad_norm": 0.38397979736328125, + "grad_norm": 40.831878662109375, "learning_rate": 5.17266870841694e-06, - "loss": 0.3733, + "loss": 0.9287, "step": 70200 }, { "epoch": 7.41, - "grad_norm": 6.8382792472839355, + "grad_norm": 0.7022567987442017, "learning_rate": 5.170556552962298e-06, - "loss": 0.2645, + "loss": 0.8659, "step": 70210 }, { "epoch": 7.42, - "grad_norm": 0.11332491040229797, + "grad_norm": 22.622085571289062, "learning_rate": 5.168444397507657e-06, - "loss": 0.3833, + "loss": 0.6982, "step": 70220 }, { "epoch": 7.42, - "grad_norm": 43.80815505981445, + "grad_norm": 33.93751525878906, "learning_rate": 5.166332242053016e-06, - "loss": 0.6814, + "loss": 1.2043, "step": 70230 }, { "epoch": 7.42, - "grad_norm": 21.38298988342285, + "grad_norm": 63.01081466674805, "learning_rate": 5.1642200865983736e-06, - "loss": 0.2275, + "loss": 0.8486, "step": 70240 }, { "epoch": 7.42, - "grad_norm": 21.05703353881836, + "grad_norm": 30.69886016845703, "learning_rate": 5.162107931143732e-06, - "loss": 0.2768, + "loss": 0.9029, "step": 70250 }, { "epoch": 7.42, - "grad_norm": 2.8022985458374023, + "grad_norm": 37.15800094604492, "learning_rate": 5.15999577568909e-06, - "loss": 0.1572, + "loss": 0.901, "step": 70260 }, { "epoch": 7.42, - "grad_norm": 0.6283069252967834, + "grad_norm": 41.825199127197266, "learning_rate": 5.15788362023445e-06, - "loss": 0.7373, + "loss": 1.2264, "step": 70270 }, { "epoch": 7.42, - "grad_norm": 14.594831466674805, + "grad_norm": 46.04741668701172, "learning_rate": 5.155771464779809e-06, - "loss": 0.2779, + "loss": 0.9282, "step": 70280 }, { "epoch": 7.42, - "grad_norm": 10.672776222229004, + "grad_norm": 42.6180305480957, "learning_rate": 5.153659309325167e-06, - "loss": 0.2798, + "loss": 1.2167, "step": 70290 }, { "epoch": 7.42, - "grad_norm": 19.898414611816406, + "grad_norm": 32.58660888671875, "learning_rate": 5.151547153870525e-06, - "loss": 0.423, + "loss": 0.9689, "step": 70300 }, { "epoch": 7.43, - "grad_norm": 1.261889100074768, + "grad_norm": 44.317535400390625, "learning_rate": 5.149434998415884e-06, - "loss": 0.6076, + "loss": 1.1972, "step": 70310 }, { "epoch": 7.43, - "grad_norm": 1.525178074836731, + "grad_norm": 46.32212448120117, "learning_rate": 5.147322842961242e-06, - "loss": 0.3859, + "loss": 0.8511, "step": 70320 }, { "epoch": 7.43, - "grad_norm": 0.025692541152238846, + "grad_norm": 28.933788299560547, "learning_rate": 5.145210687506601e-06, - "loss": 0.3701, + "loss": 1.179, "step": 70330 }, { "epoch": 7.43, - "grad_norm": 16.248756408691406, + "grad_norm": 49.747779846191406, "learning_rate": 5.14309853205196e-06, - "loss": 0.656, + "loss": 1.2556, "step": 70340 }, { "epoch": 7.43, - "grad_norm": 4.580924034118652, + "grad_norm": 24.44230842590332, "learning_rate": 5.140986376597318e-06, - "loss": 0.2975, + "loss": 0.9226, "step": 70350 }, { "epoch": 7.43, - "grad_norm": 3.3085086345672607, + "grad_norm": 50.90754318237305, "learning_rate": 5.138874221142676e-06, - "loss": 0.4336, + "loss": 0.8654, "step": 70360 }, { "epoch": 7.43, - "grad_norm": 35.80574035644531, + "grad_norm": 47.40642547607422, "learning_rate": 5.136762065688034e-06, - "loss": 0.4729, + "loss": 0.824, "step": 70370 }, { "epoch": 7.43, - "grad_norm": 15.402789115905762, + "grad_norm": 57.46095275878906, "learning_rate": 5.134649910233394e-06, - "loss": 0.2467, + "loss": 0.9519, "step": 70380 }, { "epoch": 7.43, - "grad_norm": 3.8732595443725586, + "grad_norm": 22.49155044555664, "learning_rate": 5.132537754778753e-06, - "loss": 0.287, + "loss": 0.9052, "step": 70390 }, { "epoch": 7.43, - "grad_norm": 16.787324905395508, + "grad_norm": 24.09930419921875, "learning_rate": 5.130425599324111e-06, - "loss": 0.3219, + "loss": 0.7326, "step": 70400 }, { "epoch": 7.44, - "grad_norm": 32.470645904541016, + "grad_norm": 48.66150665283203, "learning_rate": 5.128313443869469e-06, - "loss": 0.1264, + "loss": 0.6536, "step": 70410 }, { "epoch": 7.44, - "grad_norm": 9.136551856994629, + "grad_norm": 50.70203399658203, "learning_rate": 5.126201288414828e-06, - "loss": 0.1712, + "loss": 0.9195, "step": 70420 }, { "epoch": 7.44, - "grad_norm": 0.49920955300331116, + "grad_norm": 11.459980010986328, "learning_rate": 5.124089132960186e-06, - "loss": 0.4321, + "loss": 0.7511, "step": 70430 }, { "epoch": 7.44, - "grad_norm": 1.3641407489776611, + "grad_norm": 50.54892349243164, "learning_rate": 5.121976977505545e-06, - "loss": 0.4747, + "loss": 0.9577, "step": 70440 }, { "epoch": 7.44, - "grad_norm": 20.09441566467285, + "grad_norm": 30.059070587158203, "learning_rate": 5.119864822050903e-06, - "loss": 0.2772, + "loss": 0.7851, "step": 70450 }, { "epoch": 7.44, - "grad_norm": 11.090465545654297, + "grad_norm": 28.623985290527344, "learning_rate": 5.117752666596262e-06, - "loss": 0.6165, + "loss": 0.7387, "step": 70460 }, { "epoch": 7.44, - "grad_norm": 0.5036386251449585, + "grad_norm": 44.86009216308594, "learning_rate": 5.11564051114162e-06, - "loss": 0.3261, + "loss": 0.874, "step": 70470 }, { "epoch": 7.44, - "grad_norm": 20.98103141784668, + "grad_norm": 54.11541748046875, "learning_rate": 5.113528355686978e-06, - "loss": 0.7114, + "loss": 0.8945, "step": 70480 }, { "epoch": 7.44, - "grad_norm": 3.993561267852783, + "grad_norm": 43.69603729248047, "learning_rate": 5.111416200232338e-06, - "loss": 0.2818, + "loss": 0.7907, "step": 70490 }, { "epoch": 7.45, - "grad_norm": 21.657028198242188, + "grad_norm": 48.502166748046875, "learning_rate": 5.109304044777697e-06, - "loss": 0.5442, + "loss": 0.9727, "step": 70500 }, { "epoch": 7.45, - "grad_norm": 0.37285032868385315, + "grad_norm": 52.57289505004883, "learning_rate": 5.107191889323055e-06, - "loss": 0.3586, + "loss": 1.018, "step": 70510 }, { "epoch": 7.45, - "grad_norm": 11.299299240112305, + "grad_norm": 45.74303436279297, "learning_rate": 5.1050797338684134e-06, - "loss": 0.4044, + "loss": 0.9278, "step": 70520 }, { "epoch": 7.45, - "grad_norm": 3.5404441356658936, + "grad_norm": 28.38006019592285, "learning_rate": 5.102967578413772e-06, - "loss": 0.2438, + "loss": 0.7992, "step": 70530 }, { "epoch": 7.45, - "grad_norm": 3.5309126377105713, + "grad_norm": 53.21798324584961, "learning_rate": 5.10085542295913e-06, - "loss": 0.2051, + "loss": 1.1408, "step": 70540 }, { "epoch": 7.45, - "grad_norm": 20.589998245239258, + "grad_norm": 42.399593353271484, "learning_rate": 5.098743267504489e-06, - "loss": 0.2281, + "loss": 0.5209, "step": 70550 }, { "epoch": 7.45, - "grad_norm": 1.601185917854309, + "grad_norm": 36.84840393066406, "learning_rate": 5.096631112049847e-06, - "loss": 0.235, + "loss": 0.935, "step": 70560 }, { "epoch": 7.45, - "grad_norm": 10.425920486450195, + "grad_norm": 3.43458890914917, "learning_rate": 5.094518956595206e-06, - "loss": 0.4261, + "loss": 0.7756, "step": 70570 }, { "epoch": 7.45, - "grad_norm": 0.6625679135322571, + "grad_norm": 44.8615608215332, "learning_rate": 5.092406801140564e-06, - "loss": 0.4201, + "loss": 1.0602, "step": 70580 }, { "epoch": 7.45, - "grad_norm": 18.540483474731445, + "grad_norm": 18.962108612060547, "learning_rate": 5.090294645685922e-06, - "loss": 0.4016, + "loss": 0.9652, "step": 70590 }, { "epoch": 7.46, - "grad_norm": 18.40845489501953, + "grad_norm": 47.81342315673828, "learning_rate": 5.088182490231281e-06, - "loss": 0.2277, + "loss": 0.7988, "step": 70600 }, { "epoch": 7.46, - "grad_norm": 7.474123477935791, + "grad_norm": 38.439048767089844, "learning_rate": 5.086070334776641e-06, - "loss": 0.3734, + "loss": 1.0353, "step": 70610 }, { "epoch": 7.46, - "grad_norm": 1.787611961364746, + "grad_norm": 40.7661247253418, "learning_rate": 5.083958179321999e-06, - "loss": 0.5213, + "loss": 0.9853, "step": 70620 }, { "epoch": 7.46, - "grad_norm": 25.902376174926758, + "grad_norm": 44.08218002319336, "learning_rate": 5.0818460238673575e-06, - "loss": 0.3189, + "loss": 0.9767, "step": 70630 }, { "epoch": 7.46, - "grad_norm": 0.6625169515609741, + "grad_norm": 7.583834171295166, "learning_rate": 5.079733868412715e-06, - "loss": 0.5239, + "loss": 0.9667, "step": 70640 }, { "epoch": 7.46, - "grad_norm": 7.758029460906982, + "grad_norm": 64.47444915771484, "learning_rate": 5.077621712958074e-06, - "loss": 0.2617, + "loss": 0.7637, "step": 70650 }, { "epoch": 7.46, - "grad_norm": 9.62162971496582, + "grad_norm": 50.72957992553711, "learning_rate": 5.075509557503433e-06, - "loss": 0.3586, + "loss": 1.1681, "step": 70660 }, { "epoch": 7.46, - "grad_norm": 39.426753997802734, + "grad_norm": 33.388858795166016, "learning_rate": 5.073397402048791e-06, - "loss": 0.4004, + "loss": 0.7365, "step": 70670 }, { "epoch": 7.46, - "grad_norm": 18.133073806762695, + "grad_norm": 48.030338287353516, "learning_rate": 5.07128524659415e-06, - "loss": 0.4278, + "loss": 0.8853, "step": 70680 }, { "epoch": 7.47, - "grad_norm": 0.7856704592704773, + "grad_norm": 49.57230758666992, "learning_rate": 5.0691730911395084e-06, - "loss": 0.2647, + "loss": 0.6489, "step": 70690 }, { "epoch": 7.47, - "grad_norm": 22.760351181030273, + "grad_norm": 43.939208984375, "learning_rate": 5.067060935684866e-06, - "loss": 0.6671, + "loss": 1.0851, "step": 70700 }, { "epoch": 7.47, - "grad_norm": 13.941965103149414, + "grad_norm": 53.782798767089844, "learning_rate": 5.064948780230225e-06, - "loss": 0.4043, + "loss": 1.0115, "step": 70710 }, { "epoch": 7.47, - "grad_norm": 20.632413864135742, + "grad_norm": 47.467979431152344, "learning_rate": 5.062836624775583e-06, - "loss": 0.4388, + "loss": 0.9975, "step": 70720 }, { "epoch": 7.47, - "grad_norm": 2.7178449630737305, + "grad_norm": 28.379581451416016, "learning_rate": 5.060724469320943e-06, - "loss": 0.2097, + "loss": 0.7751, "step": 70730 }, { "epoch": 7.47, - "grad_norm": 10.697650909423828, + "grad_norm": 36.16436767578125, "learning_rate": 5.0586123138663015e-06, - "loss": 0.6511, + "loss": 1.0538, "step": 70740 }, { "epoch": 7.47, - "grad_norm": 10.187895774841309, + "grad_norm": 56.793949127197266, "learning_rate": 5.0565001584116594e-06, - "loss": 0.1245, + "loss": 0.7976, "step": 70750 }, { "epoch": 7.47, - "grad_norm": 38.86831283569336, + "grad_norm": 39.9622917175293, "learning_rate": 5.054388002957018e-06, - "loss": 0.5337, + "loss": 0.8083, "step": 70760 }, { "epoch": 7.47, - "grad_norm": 7.8734965324401855, + "grad_norm": 32.508094787597656, "learning_rate": 5.052275847502377e-06, - "loss": 0.4241, + "loss": 0.7993, "step": 70770 }, { "epoch": 7.47, - "grad_norm": 28.25473403930664, + "grad_norm": 40.024723052978516, "learning_rate": 5.050163692047735e-06, - "loss": 0.461, + "loss": 0.8643, "step": 70780 }, { "epoch": 7.48, - "grad_norm": 28.359975814819336, + "grad_norm": 45.92555236816406, "learning_rate": 5.048051536593094e-06, - "loss": 0.269, + "loss": 0.6668, "step": 70790 }, { "epoch": 7.48, - "grad_norm": 9.45223331451416, + "grad_norm": 33.91233825683594, "learning_rate": 5.0459393811384525e-06, - "loss": 0.4855, + "loss": 0.6309, "step": 70800 }, { "epoch": 7.48, - "grad_norm": 5.339942455291748, + "grad_norm": 45.46150207519531, "learning_rate": 5.04382722568381e-06, - "loss": 0.3168, + "loss": 0.5605, "step": 70810 }, { "epoch": 7.48, - "grad_norm": 9.038232803344727, + "grad_norm": 33.35713195800781, "learning_rate": 5.041715070229169e-06, - "loss": 0.3964, + "loss": 0.8955, "step": 70820 }, { "epoch": 7.48, - "grad_norm": 0.4682900905609131, + "grad_norm": 42.605770111083984, "learning_rate": 5.039602914774527e-06, - "loss": 0.2093, + "loss": 1.16, "step": 70830 }, { "epoch": 7.48, - "grad_norm": 16.268999099731445, + "grad_norm": 27.71285057067871, "learning_rate": 5.037490759319886e-06, - "loss": 0.3929, + "loss": 0.9195, "step": 70840 }, { "epoch": 7.48, - "grad_norm": 11.167661666870117, + "grad_norm": 29.70201873779297, "learning_rate": 5.0353786038652455e-06, - "loss": 0.4334, + "loss": 0.8958, "step": 70850 }, { "epoch": 7.48, - "grad_norm": 4.544824123382568, + "grad_norm": 21.56212043762207, "learning_rate": 5.0332664484106035e-06, - "loss": 0.2302, + "loss": 0.8275, "step": 70860 }, { "epoch": 7.48, - "grad_norm": 20.32644271850586, + "grad_norm": 62.52228927612305, "learning_rate": 5.031154292955962e-06, - "loss": 0.3776, + "loss": 0.8376, "step": 70870 }, { "epoch": 7.49, - "grad_norm": 32.62471008300781, + "grad_norm": 37.501869201660156, "learning_rate": 5.029042137501321e-06, - "loss": 0.4608, + "loss": 0.7265, "step": 70880 }, { "epoch": 7.49, - "grad_norm": 0.21067306399345398, + "grad_norm": 40.22846603393555, "learning_rate": 5.026929982046679e-06, - "loss": 0.1814, + "loss": 0.9217, "step": 70890 }, { "epoch": 7.49, - "grad_norm": 22.247737884521484, + "grad_norm": 45.38471221923828, "learning_rate": 5.024817826592038e-06, - "loss": 0.3882, + "loss": 0.8363, "step": 70900 }, { "epoch": 7.49, - "grad_norm": 2.6314949989318848, + "grad_norm": 22.914993286132812, "learning_rate": 5.022705671137396e-06, - "loss": 0.2391, + "loss": 0.5635, "step": 70910 }, { "epoch": 7.49, - "grad_norm": 10.447360038757324, + "grad_norm": 46.20355224609375, "learning_rate": 5.0205935156827544e-06, - "loss": 0.2405, + "loss": 0.8133, "step": 70920 }, { "epoch": 7.49, - "grad_norm": 9.120744705200195, + "grad_norm": 26.151718139648438, "learning_rate": 5.018481360228113e-06, - "loss": 0.2669, + "loss": 0.7792, "step": 70930 }, { "epoch": 7.49, - "grad_norm": 0.06302540004253387, + "grad_norm": 35.37704849243164, "learning_rate": 5.016369204773471e-06, - "loss": 0.2895, + "loss": 0.8286, "step": 70940 }, { "epoch": 7.49, - "grad_norm": 11.752287864685059, + "grad_norm": 43.52094268798828, "learning_rate": 5.01425704931883e-06, - "loss": 0.2532, + "loss": 0.5944, "step": 70950 }, { "epoch": 7.49, - "grad_norm": 13.245378494262695, + "grad_norm": 36.335086822509766, "learning_rate": 5.0121448938641895e-06, - "loss": 0.4483, + "loss": 1.0465, "step": 70960 }, { "epoch": 7.49, - "grad_norm": 16.84820556640625, + "grad_norm": 33.72743606567383, "learning_rate": 5.0100327384095475e-06, - "loss": 0.3003, + "loss": 0.9477, "step": 70970 }, { "epoch": 7.5, - "grad_norm": 8.827969551086426, + "grad_norm": 38.07695388793945, "learning_rate": 5.007920582954906e-06, - "loss": 0.285, + "loss": 0.7923, "step": 70980 }, { "epoch": 7.5, - "grad_norm": 27.389177322387695, + "grad_norm": 29.69480323791504, "learning_rate": 5.005808427500265e-06, - "loss": 0.3323, + "loss": 1.0068, "step": 70990 }, { "epoch": 7.5, - "grad_norm": 1.629065752029419, + "grad_norm": 42.984317779541016, "learning_rate": 5.003696272045623e-06, - "loss": 0.1659, + "loss": 0.7262, "step": 71000 }, { "epoch": 7.5, - "grad_norm": 12.827731132507324, + "grad_norm": 50.794925689697266, "learning_rate": 5.001584116590982e-06, - "loss": 0.2732, + "loss": 0.9458, "step": 71010 }, { "epoch": 7.5, - "grad_norm": 14.205699920654297, + "grad_norm": 36.03162384033203, "learning_rate": 4.9994719611363405e-06, - "loss": 0.3424, + "loss": 0.94, "step": 71020 }, { "epoch": 7.5, - "grad_norm": 0.09760904312133789, + "grad_norm": 38.27897644042969, "learning_rate": 4.9973598056816985e-06, - "loss": 0.1894, + "loss": 0.8141, "step": 71030 }, { "epoch": 7.5, - "grad_norm": 15.806410789489746, + "grad_norm": 33.71566390991211, "learning_rate": 4.995247650227057e-06, - "loss": 0.2878, + "loss": 0.7573, "step": 71040 }, { "epoch": 7.5, - "grad_norm": 5.278548717498779, + "grad_norm": 37.718685150146484, "learning_rate": 4.993135494772416e-06, - "loss": 0.4471, + "loss": 0.8591, "step": 71050 }, { "epoch": 7.5, - "grad_norm": 32.111297607421875, + "grad_norm": 33.54871368408203, "learning_rate": 4.991023339317774e-06, - "loss": 0.4708, + "loss": 0.9168, "step": 71060 }, { "epoch": 7.51, - "grad_norm": 16.10995864868164, + "grad_norm": 31.57840347290039, "learning_rate": 4.988911183863133e-06, - "loss": 0.5519, + "loss": 0.8233, "step": 71070 }, { "epoch": 7.51, - "grad_norm": 12.944966316223145, + "grad_norm": 53.6431884765625, "learning_rate": 4.986799028408491e-06, - "loss": 0.323, + "loss": 0.7401, "step": 71080 }, { "epoch": 7.51, - "grad_norm": 0.0751594677567482, + "grad_norm": 21.397790908813477, "learning_rate": 4.98468687295385e-06, - "loss": 0.4197, + "loss": 0.6773, "step": 71090 }, { "epoch": 7.51, - "grad_norm": 6.342922687530518, + "grad_norm": 49.777099609375, "learning_rate": 4.982574717499208e-06, - "loss": 0.32, + "loss": 0.6706, "step": 71100 }, { "epoch": 7.51, - "grad_norm": 0.48401743173599243, + "grad_norm": 21.204730987548828, "learning_rate": 4.980462562044567e-06, - "loss": 0.2749, + "loss": 0.8731, "step": 71110 }, { "epoch": 7.51, - "grad_norm": 22.891408920288086, + "grad_norm": 40.8104362487793, "learning_rate": 4.978350406589925e-06, - "loss": 0.4602, + "loss": 0.87, "step": 71120 }, { "epoch": 7.51, - "grad_norm": 20.911334991455078, + "grad_norm": 30.304920196533203, "learning_rate": 4.9762382511352846e-06, - "loss": 0.2739, + "loss": 0.9416, "step": 71130 }, { "epoch": 7.51, - "grad_norm": 9.694724082946777, + "grad_norm": 20.772193908691406, "learning_rate": 4.9741260956806425e-06, - "loss": 0.2807, + "loss": 0.7408, "step": 71140 }, { "epoch": 7.51, - "grad_norm": 16.230802536010742, + "grad_norm": 51.993255615234375, "learning_rate": 4.972013940226001e-06, - "loss": 0.3885, + "loss": 0.8101, "step": 71150 }, { "epoch": 7.52, - "grad_norm": 23.89365577697754, + "grad_norm": 51.39048385620117, "learning_rate": 4.969901784771359e-06, - "loss": 0.5061, + "loss": 1.1597, "step": 71160 }, { "epoch": 7.52, - "grad_norm": 16.322389602661133, + "grad_norm": 33.441532135009766, "learning_rate": 4.967789629316718e-06, - "loss": 0.533, + "loss": 0.9346, "step": 71170 }, { "epoch": 7.52, - "grad_norm": 27.9035701751709, + "grad_norm": 13.415246963500977, "learning_rate": 4.965677473862077e-06, - "loss": 0.2219, + "loss": 0.6812, "step": 71180 }, { "epoch": 7.52, - "grad_norm": 0.16653648018836975, + "grad_norm": 29.315643310546875, "learning_rate": 4.963565318407435e-06, - "loss": 0.3954, + "loss": 1.0018, "step": 71190 }, { "epoch": 7.52, - "grad_norm": 0.11949925124645233, + "grad_norm": 30.022092819213867, "learning_rate": 4.961453162952794e-06, - "loss": 0.3293, + "loss": 0.8727, "step": 71200 }, { "epoch": 7.52, - "grad_norm": 29.19469451904297, + "grad_norm": 38.483829498291016, "learning_rate": 4.959341007498152e-06, - "loss": 0.318, + "loss": 0.7757, "step": 71210 }, { "epoch": 7.52, - "grad_norm": 13.77968692779541, + "grad_norm": 41.266319274902344, "learning_rate": 4.957228852043511e-06, - "loss": 0.4259, + "loss": 0.9862, "step": 71220 }, { "epoch": 7.52, - "grad_norm": 0.16163623332977295, + "grad_norm": 27.69929313659668, "learning_rate": 4.955116696588869e-06, - "loss": 0.1217, + "loss": 0.6678, "step": 71230 }, { "epoch": 7.52, - "grad_norm": 9.1668062210083, + "grad_norm": 50.656673431396484, "learning_rate": 4.953004541134228e-06, - "loss": 0.463, + "loss": 0.9986, "step": 71240 }, { "epoch": 7.52, - "grad_norm": 12.380242347717285, + "grad_norm": 19.30685806274414, "learning_rate": 4.9508923856795865e-06, - "loss": 0.2029, + "loss": 0.7022, "step": 71250 }, { "epoch": 7.53, - "grad_norm": 10.021783828735352, + "grad_norm": 47.233638763427734, "learning_rate": 4.948780230224945e-06, - "loss": 0.4896, + "loss": 0.9688, "step": 71260 }, { "epoch": 7.53, - "grad_norm": 10.484097480773926, + "grad_norm": 47.48386764526367, "learning_rate": 4.946668074770303e-06, - "loss": 0.1888, + "loss": 0.9073, "step": 71270 }, { "epoch": 7.53, - "grad_norm": 8.89029598236084, + "grad_norm": 34.390377044677734, "learning_rate": 4.944555919315662e-06, - "loss": 0.5934, + "loss": 0.9198, "step": 71280 }, { "epoch": 7.53, - "grad_norm": 8.144593238830566, + "grad_norm": 31.38385772705078, "learning_rate": 4.942443763861021e-06, - "loss": 0.4428, + "loss": 0.7111, "step": 71290 }, { "epoch": 7.53, - "grad_norm": 6.967047691345215, + "grad_norm": 33.51997375488281, "learning_rate": 4.940331608406379e-06, - "loss": 0.336, + "loss": 0.7454, "step": 71300 }, { "epoch": 7.53, - "grad_norm": 0.947503924369812, + "grad_norm": 34.56338119506836, "learning_rate": 4.9382194529517375e-06, - "loss": 0.2552, + "loss": 0.857, "step": 71310 }, { "epoch": 7.53, - "grad_norm": 23.92059898376465, + "grad_norm": 42.595985412597656, "learning_rate": 4.936107297497096e-06, - "loss": 0.2916, + "loss": 1.0969, "step": 71320 }, { "epoch": 7.53, - "grad_norm": 22.532470703125, + "grad_norm": 55.478050231933594, "learning_rate": 4.933995142042455e-06, - "loss": 0.546, + "loss": 1.5721, "step": 71330 }, { "epoch": 7.53, - "grad_norm": 0.058498889207839966, + "grad_norm": 59.585540771484375, "learning_rate": 4.931882986587813e-06, - "loss": 0.5854, + "loss": 1.2271, "step": 71340 }, { "epoch": 7.54, - "grad_norm": 17.969820022583008, + "grad_norm": 52.16375732421875, "learning_rate": 4.929770831133172e-06, - "loss": 0.509, + "loss": 0.9847, "step": 71350 }, { "epoch": 7.54, - "grad_norm": 13.05607795715332, + "grad_norm": 32.72455596923828, "learning_rate": 4.9276586756785305e-06, - "loss": 0.7147, + "loss": 1.027, "step": 71360 }, { "epoch": 7.54, - "grad_norm": 0.07027821242809296, + "grad_norm": 27.002012252807617, "learning_rate": 4.925546520223889e-06, - "loss": 0.5476, + "loss": 1.04, "step": 71370 }, { "epoch": 7.54, - "grad_norm": 26.835567474365234, + "grad_norm": 37.5918083190918, "learning_rate": 4.923434364769247e-06, - "loss": 0.1942, + "loss": 0.8352, "step": 71380 }, { "epoch": 7.54, - "grad_norm": 33.244361877441406, + "grad_norm": 57.73597717285156, "learning_rate": 4.921322209314606e-06, - "loss": 0.438, + "loss": 0.9418, "step": 71390 }, { "epoch": 7.54, - "grad_norm": 0.2788954973220825, + "grad_norm": 39.72012710571289, "learning_rate": 4.919210053859965e-06, - "loss": 0.249, + "loss": 0.881, "step": 71400 }, { "epoch": 7.54, - "grad_norm": 1.2098673582077026, + "grad_norm": 50.61980056762695, "learning_rate": 4.917097898405323e-06, - "loss": 0.2894, + "loss": 1.1669, "step": 71410 }, { "epoch": 7.54, - "grad_norm": 0.13485944271087646, + "grad_norm": 16.0416259765625, "learning_rate": 4.9149857429506815e-06, - "loss": 0.2645, + "loss": 0.4113, "step": 71420 }, { "epoch": 7.54, - "grad_norm": 0.077328622341156, + "grad_norm": 53.11887741088867, "learning_rate": 4.91287358749604e-06, - "loss": 0.226, + "loss": 0.8587, "step": 71430 }, { "epoch": 7.54, - "grad_norm": 18.862932205200195, + "grad_norm": 39.50460433959961, "learning_rate": 4.910761432041399e-06, - "loss": 0.373, + "loss": 1.0789, "step": 71440 }, { "epoch": 7.55, - "grad_norm": 0.07970613241195679, + "grad_norm": 10.26354694366455, "learning_rate": 4.908649276586757e-06, - "loss": 0.4074, + "loss": 0.7728, "step": 71450 }, { "epoch": 7.55, - "grad_norm": 0.0973312109708786, + "grad_norm": 55.796852111816406, "learning_rate": 4.906537121132116e-06, - "loss": 0.4008, + "loss": 1.0931, "step": 71460 }, { "epoch": 7.55, - "grad_norm": 45.09787368774414, + "grad_norm": 27.834733963012695, "learning_rate": 4.9044249656774746e-06, - "loss": 0.4078, + "loss": 1.1411, "step": 71470 }, { "epoch": 7.55, - "grad_norm": 11.420838356018066, + "grad_norm": 19.19216537475586, "learning_rate": 4.9023128102228325e-06, - "loss": 0.2169, + "loss": 0.8615, "step": 71480 }, { "epoch": 7.55, - "grad_norm": 14.761588096618652, + "grad_norm": 42.72008514404297, "learning_rate": 4.900200654768191e-06, - "loss": 0.2843, + "loss": 0.8797, "step": 71490 }, { "epoch": 7.55, - "grad_norm": 3.090606689453125, + "grad_norm": 23.094350814819336, "learning_rate": 4.89808849931355e-06, - "loss": 0.5062, + "loss": 0.723, "step": 71500 }, { "epoch": 7.55, - "grad_norm": 25.221412658691406, + "grad_norm": 43.20659255981445, "learning_rate": 4.895976343858909e-06, - "loss": 0.4769, + "loss": 0.7689, "step": 71510 }, { "epoch": 7.55, - "grad_norm": 10.725829124450684, + "grad_norm": 39.791446685791016, "learning_rate": 4.893864188404267e-06, - "loss": 0.5362, + "loss": 0.8673, "step": 71520 }, { "epoch": 7.55, - "grad_norm": 24.756357192993164, + "grad_norm": 48.177528381347656, "learning_rate": 4.8917520329496255e-06, - "loss": 0.3706, + "loss": 0.907, "step": 71530 }, { "epoch": 7.56, - "grad_norm": 11.59644889831543, + "grad_norm": 27.575281143188477, "learning_rate": 4.8896398774949835e-06, - "loss": 0.4407, + "loss": 1.1233, "step": 71540 }, { "epoch": 7.56, - "grad_norm": 24.679241180419922, + "grad_norm": 20.597253799438477, "learning_rate": 4.887527722040343e-06, - "loss": 0.4612, + "loss": 0.662, "step": 71550 }, { "epoch": 7.56, - "grad_norm": 17.20545768737793, + "grad_norm": 55.452423095703125, "learning_rate": 4.885415566585701e-06, - "loss": 0.266, + "loss": 0.808, "step": 71560 }, { "epoch": 7.56, - "grad_norm": 6.331724643707275, + "grad_norm": 10.571163177490234, "learning_rate": 4.88330341113106e-06, - "loss": 0.3419, + "loss": 0.8178, "step": 71570 }, { "epoch": 7.56, - "grad_norm": 5.176941394805908, + "grad_norm": 32.6451301574707, "learning_rate": 4.881191255676418e-06, - "loss": 0.2721, + "loss": 1.0185, "step": 71580 }, { "epoch": 7.56, - "grad_norm": 13.17010498046875, + "grad_norm": 34.87348175048828, "learning_rate": 4.8790791002217765e-06, - "loss": 0.6369, + "loss": 0.9157, "step": 71590 }, { "epoch": 7.56, - "grad_norm": 6.595737457275391, + "grad_norm": 25.772558212280273, "learning_rate": 4.876966944767135e-06, - "loss": 0.2586, + "loss": 1.2881, "step": 71600 }, { "epoch": 7.56, - "grad_norm": 34.00047302246094, + "grad_norm": 35.4743766784668, "learning_rate": 4.874854789312494e-06, - "loss": 0.3584, + "loss": 0.8658, "step": 71610 }, { "epoch": 7.56, - "grad_norm": 43.2764778137207, + "grad_norm": 48.30557632446289, "learning_rate": 4.872742633857852e-06, - "loss": 0.4088, + "loss": 0.6766, "step": 71620 }, { "epoch": 7.56, - "grad_norm": 35.61331558227539, + "grad_norm": 58.17790985107422, "learning_rate": 4.870630478403211e-06, - "loss": 0.1992, + "loss": 0.9376, "step": 71630 }, { "epoch": 7.57, - "grad_norm": 0.5891434550285339, + "grad_norm": 57.53498077392578, "learning_rate": 4.8685183229485696e-06, - "loss": 0.4303, + "loss": 0.9456, "step": 71640 }, { "epoch": 7.57, - "grad_norm": 6.097708225250244, + "grad_norm": 65.0942611694336, "learning_rate": 4.8664061674939275e-06, - "loss": 0.2933, + "loss": 0.8459, "step": 71650 }, { "epoch": 7.57, - "grad_norm": 32.92715072631836, + "grad_norm": 35.854103088378906, "learning_rate": 4.864294012039287e-06, - "loss": 0.259, + "loss": 0.8869, "step": 71660 }, { "epoch": 7.57, - "grad_norm": 15.470625877380371, + "grad_norm": 44.38779830932617, "learning_rate": 4.862181856584645e-06, - "loss": 0.2801, + "loss": 0.6023, "step": 71670 }, { "epoch": 7.57, - "grad_norm": 7.429263591766357, + "grad_norm": 23.387609481811523, "learning_rate": 4.860069701130004e-06, - "loss": 0.3041, + "loss": 0.8318, "step": 71680 }, { "epoch": 7.57, - "grad_norm": 14.898618698120117, + "grad_norm": 29.772796630859375, "learning_rate": 4.857957545675362e-06, - "loss": 0.2409, + "loss": 0.5988, "step": 71690 }, { "epoch": 7.57, - "grad_norm": 46.8839225769043, + "grad_norm": 37.28224182128906, "learning_rate": 4.8558453902207205e-06, - "loss": 0.2725, + "loss": 0.9183, "step": 71700 }, { "epoch": 7.57, - "grad_norm": 28.22022247314453, + "grad_norm": 41.11168670654297, "learning_rate": 4.853733234766079e-06, - "loss": 0.5078, + "loss": 0.7715, "step": 71710 }, { "epoch": 7.57, - "grad_norm": 1.4459960460662842, + "grad_norm": 46.238407135009766, "learning_rate": 4.851621079311437e-06, - "loss": 0.2612, + "loss": 0.5795, "step": 71720 }, { "epoch": 7.58, - "grad_norm": 0.9456952810287476, + "grad_norm": 42.17549133300781, "learning_rate": 4.849508923856796e-06, - "loss": 0.2864, + "loss": 0.886, "step": 71730 }, { "epoch": 7.58, - "grad_norm": 19.75490951538086, + "grad_norm": 34.31361389160156, "learning_rate": 4.847396768402155e-06, - "loss": 0.5109, + "loss": 1.136, "step": 71740 }, { "epoch": 7.58, - "grad_norm": 16.801546096801758, + "grad_norm": 43.922115325927734, "learning_rate": 4.845284612947514e-06, - "loss": 0.4633, + "loss": 1.2467, "step": 71750 }, { "epoch": 7.58, - "grad_norm": 12.136834144592285, + "grad_norm": 15.898256301879883, "learning_rate": 4.8431724574928715e-06, - "loss": 0.2735, + "loss": 0.8361, "step": 71760 }, { "epoch": 7.58, - "grad_norm": 24.14370346069336, + "grad_norm": 54.37864303588867, "learning_rate": 4.84106030203823e-06, - "loss": 0.3004, + "loss": 0.8421, "step": 71770 }, { "epoch": 7.58, - "grad_norm": 17.296974182128906, + "grad_norm": 76.14704895019531, "learning_rate": 4.838948146583589e-06, - "loss": 0.6717, + "loss": 1.0992, "step": 71780 }, { "epoch": 7.58, - "grad_norm": 21.015010833740234, + "grad_norm": 28.301916122436523, "learning_rate": 4.836835991128948e-06, - "loss": 0.4209, + "loss": 0.694, "step": 71790 }, { "epoch": 7.58, - "grad_norm": 2.948866844177246, + "grad_norm": 25.9097957611084, "learning_rate": 4.834723835674306e-06, - "loss": 0.4299, + "loss": 0.784, "step": 71800 }, { "epoch": 7.58, - "grad_norm": 0.3226434886455536, + "grad_norm": 31.655447006225586, "learning_rate": 4.8326116802196646e-06, - "loss": 0.2497, + "loss": 0.7676, "step": 71810 }, { "epoch": 7.58, - "grad_norm": 16.2379207611084, + "grad_norm": 28.541969299316406, "learning_rate": 4.830499524765023e-06, - "loss": 0.1977, + "loss": 0.6893, "step": 71820 }, { "epoch": 7.59, - "grad_norm": 0.9164279103279114, + "grad_norm": 22.045631408691406, "learning_rate": 4.828387369310381e-06, - "loss": 0.5591, + "loss": 0.9051, "step": 71830 }, { "epoch": 7.59, - "grad_norm": 15.833215713500977, + "grad_norm": 45.95463562011719, "learning_rate": 4.82627521385574e-06, - "loss": 0.3941, + "loss": 0.8299, "step": 71840 }, { "epoch": 7.59, - "grad_norm": 0.5548765063285828, + "grad_norm": 35.33644104003906, "learning_rate": 4.824163058401099e-06, - "loss": 0.2066, + "loss": 0.7988, "step": 71850 }, { "epoch": 7.59, - "grad_norm": 23.416990280151367, + "grad_norm": 40.355926513671875, "learning_rate": 4.822050902946458e-06, - "loss": 0.4455, + "loss": 0.8721, "step": 71860 }, { "epoch": 7.59, - "grad_norm": 1.923115849494934, + "grad_norm": 45.06378173828125, "learning_rate": 4.8199387474918156e-06, - "loss": 0.212, + "loss": 0.8558, "step": 71870 }, { "epoch": 7.59, - "grad_norm": 19.65700912475586, + "grad_norm": 54.31217575073242, "learning_rate": 4.817826592037174e-06, - "loss": 0.556, + "loss": 1.1514, "step": 71880 }, { "epoch": 7.59, - "grad_norm": 27.968751907348633, + "grad_norm": 63.87968826293945, "learning_rate": 4.815714436582533e-06, - "loss": 0.6534, + "loss": 1.2109, "step": 71890 }, { "epoch": 7.59, - "grad_norm": 4.350430488586426, + "grad_norm": 22.898639678955078, "learning_rate": 4.813602281127892e-06, - "loss": 0.4323, + "loss": 0.76, "step": 71900 }, { "epoch": 7.59, - "grad_norm": 17.470998764038086, + "grad_norm": 37.75605773925781, "learning_rate": 4.81149012567325e-06, - "loss": 0.2706, + "loss": 0.9788, "step": 71910 }, { "epoch": 7.6, - "grad_norm": 25.878185272216797, + "grad_norm": 60.932804107666016, "learning_rate": 4.809377970218609e-06, - "loss": 0.4069, + "loss": 0.8097, "step": 71920 }, { "epoch": 7.6, - "grad_norm": 1.6451725959777832, + "grad_norm": 52.41938018798828, "learning_rate": 4.807265814763967e-06, - "loss": 0.3367, + "loss": 0.8796, "step": 71930 }, { "epoch": 7.6, - "grad_norm": 9.876347541809082, + "grad_norm": 26.371034622192383, "learning_rate": 4.805153659309325e-06, - "loss": 0.3634, + "loss": 0.8504, "step": 71940 }, { "epoch": 7.6, - "grad_norm": 35.35315704345703, + "grad_norm": 32.3948860168457, "learning_rate": 4.803041503854684e-06, - "loss": 0.3295, + "loss": 0.9019, "step": 71950 }, { "epoch": 7.6, - "grad_norm": 0.6201117038726807, + "grad_norm": 10.96035385131836, "learning_rate": 4.800929348400042e-06, - "loss": 0.2161, + "loss": 1.001, "step": 71960 }, { "epoch": 7.6, - "grad_norm": 113.8854751586914, + "grad_norm": 49.88179016113281, "learning_rate": 4.798817192945402e-06, - "loss": 0.4591, + "loss": 0.7575, "step": 71970 }, { "epoch": 7.6, - "grad_norm": 7.162062644958496, + "grad_norm": 42.35453796386719, "learning_rate": 4.79670503749076e-06, - "loss": 0.3483, + "loss": 1.0506, "step": 71980 }, { "epoch": 7.6, - "grad_norm": 0.22540734708309174, + "grad_norm": 21.992277145385742, "learning_rate": 4.794592882036118e-06, - "loss": 0.3, + "loss": 0.8678, "step": 71990 }, { "epoch": 7.6, - "grad_norm": 16.694679260253906, + "grad_norm": 31.643545150756836, "learning_rate": 4.792480726581476e-06, - "loss": 0.442, + "loss": 0.8211, "step": 72000 }, { "epoch": 7.6, - "grad_norm": 0.6691588163375854, + "grad_norm": 43.608863830566406, "learning_rate": 4.790368571126836e-06, - "loss": 0.2336, + "loss": 0.8551, "step": 72010 }, { "epoch": 7.61, - "grad_norm": 2.2611682415008545, + "grad_norm": 33.161521911621094, "learning_rate": 4.788256415672194e-06, - "loss": 0.322, + "loss": 0.9555, "step": 72020 }, { "epoch": 7.61, - "grad_norm": 23.328506469726562, + "grad_norm": 63.46328353881836, "learning_rate": 4.786144260217553e-06, - "loss": 0.3164, + "loss": 1.2194, "step": 72030 }, { "epoch": 7.61, - "grad_norm": 42.1422233581543, + "grad_norm": 50.093143463134766, "learning_rate": 4.7840321047629106e-06, - "loss": 0.4445, + "loss": 0.8932, "step": 72040 }, { "epoch": 7.61, - "grad_norm": 17.885009765625, + "grad_norm": 56.22908401489258, "learning_rate": 4.781919949308269e-06, - "loss": 0.4206, + "loss": 0.9442, "step": 72050 }, { "epoch": 7.61, - "grad_norm": 8.997892379760742, + "grad_norm": 63.52980041503906, "learning_rate": 4.779807793853628e-06, - "loss": 0.4184, + "loss": 0.992, "step": 72060 }, { "epoch": 7.61, - "grad_norm": 0.22156551480293274, + "grad_norm": 28.643537521362305, "learning_rate": 4.777695638398986e-06, - "loss": 0.3595, + "loss": 0.7894, "step": 72070 }, { "epoch": 7.61, - "grad_norm": 19.363157272338867, + "grad_norm": 51.766414642333984, "learning_rate": 4.775583482944345e-06, - "loss": 0.5253, + "loss": 1.0001, "step": 72080 }, { "epoch": 7.61, - "grad_norm": 8.449490547180176, + "grad_norm": 33.649147033691406, "learning_rate": 4.773471327489704e-06, - "loss": 0.4366, + "loss": 0.7052, "step": 72090 }, { "epoch": 7.61, - "grad_norm": 0.13566023111343384, + "grad_norm": 46.910675048828125, "learning_rate": 4.771359172035062e-06, - "loss": 0.255, + "loss": 0.6124, "step": 72100 }, { "epoch": 7.62, - "grad_norm": 1.1627379655838013, + "grad_norm": 38.518497467041016, "learning_rate": 4.76924701658042e-06, - "loss": 0.3531, + "loss": 0.7903, "step": 72110 }, { "epoch": 7.62, - "grad_norm": 46.36796188354492, + "grad_norm": 55.4771842956543, "learning_rate": 4.767134861125779e-06, - "loss": 0.4361, + "loss": 1.0694, "step": 72120 }, { "epoch": 7.62, - "grad_norm": 0.027526719495654106, + "grad_norm": 27.63654136657715, "learning_rate": 4.765022705671138e-06, - "loss": 0.2723, + "loss": 0.4791, "step": 72130 }, { "epoch": 7.62, - "grad_norm": 1.3267406225204468, + "grad_norm": 29.290714263916016, "learning_rate": 4.762910550216497e-06, - "loss": 0.4032, + "loss": 1.1077, "step": 72140 }, { "epoch": 7.62, - "grad_norm": 1.815149188041687, + "grad_norm": 36.261348724365234, "learning_rate": 4.760798394761855e-06, - "loss": 0.4591, + "loss": 0.9503, "step": 72150 }, { "epoch": 7.62, - "grad_norm": 2.5961856842041016, + "grad_norm": 23.085147857666016, "learning_rate": 4.758686239307213e-06, - "loss": 0.4344, + "loss": 0.9561, "step": 72160 }, { "epoch": 7.62, - "grad_norm": 11.458305358886719, + "grad_norm": 36.78815841674805, "learning_rate": 4.756574083852572e-06, - "loss": 0.4576, + "loss": 0.9219, "step": 72170 }, { "epoch": 7.62, - "grad_norm": 1.4952924251556396, + "grad_norm": 12.832230567932129, "learning_rate": 4.75446192839793e-06, - "loss": 0.4364, + "loss": 0.8321, "step": 72180 }, { "epoch": 7.62, - "grad_norm": 22.831850051879883, + "grad_norm": 34.928245544433594, "learning_rate": 4.752349772943289e-06, - "loss": 0.2133, + "loss": 0.8659, "step": 72190 }, { "epoch": 7.62, - "grad_norm": 17.87241554260254, + "grad_norm": 49.43537902832031, "learning_rate": 4.750237617488648e-06, - "loss": 0.7099, + "loss": 0.9824, "step": 72200 }, { "epoch": 7.63, - "grad_norm": 0.0646611675620079, + "grad_norm": 24.735279083251953, "learning_rate": 4.748125462034006e-06, - "loss": 0.3061, + "loss": 0.7277, "step": 72210 }, { "epoch": 7.63, - "grad_norm": 22.50699234008789, + "grad_norm": 40.004150390625, "learning_rate": 4.746013306579364e-06, - "loss": 0.5276, + "loss": 1.1419, "step": 72220 }, { "epoch": 7.63, - "grad_norm": 24.311738967895508, + "grad_norm": 45.88239669799805, "learning_rate": 4.743901151124723e-06, - "loss": 0.3229, + "loss": 0.8499, "step": 72230 }, { "epoch": 7.63, - "grad_norm": 34.20420837402344, + "grad_norm": 46.186458587646484, "learning_rate": 4.741788995670082e-06, - "loss": 0.4358, + "loss": 0.8178, "step": 72240 }, { "epoch": 7.63, - "grad_norm": 16.420379638671875, + "grad_norm": 41.09825897216797, "learning_rate": 4.739676840215441e-06, - "loss": 0.2144, + "loss": 0.6536, "step": 72250 }, { "epoch": 7.63, - "grad_norm": 0.4519244134426117, + "grad_norm": 35.13558578491211, "learning_rate": 4.737564684760799e-06, - "loss": 0.2495, + "loss": 1.1824, "step": 72260 }, { "epoch": 7.63, - "grad_norm": 20.12946128845215, + "grad_norm": 52.64337158203125, "learning_rate": 4.735452529306157e-06, - "loss": 0.3052, + "loss": 0.7594, "step": 72270 }, { "epoch": 7.63, - "grad_norm": 20.21854591369629, + "grad_norm": 18.55494499206543, "learning_rate": 4.733340373851516e-06, - "loss": 0.4373, + "loss": 0.9005, "step": 72280 }, { "epoch": 7.63, - "grad_norm": 10.499225616455078, + "grad_norm": 47.02424240112305, "learning_rate": 4.731228218396874e-06, - "loss": 0.4636, + "loss": 1.0619, "step": 72290 }, { "epoch": 7.64, - "grad_norm": 1.5741785764694214, + "grad_norm": 45.59566116333008, "learning_rate": 4.729116062942233e-06, - "loss": 0.2487, + "loss": 0.611, "step": 72300 }, { "epoch": 7.64, - "grad_norm": 4.475569248199463, + "grad_norm": 16.72843360900879, "learning_rate": 4.727003907487592e-06, - "loss": 0.4437, + "loss": 0.9791, "step": 72310 }, { "epoch": 7.64, - "grad_norm": 11.757635116577148, + "grad_norm": 43.27763366699219, "learning_rate": 4.7248917520329504e-06, - "loss": 0.5558, + "loss": 0.893, "step": 72320 }, { "epoch": 7.64, - "grad_norm": 13.071549415588379, + "grad_norm": 40.98398208618164, "learning_rate": 4.722779596578308e-06, - "loss": 0.295, + "loss": 0.7065, "step": 72330 }, { "epoch": 7.64, - "grad_norm": 0.1901068240404129, + "grad_norm": 39.370155334472656, "learning_rate": 4.720667441123667e-06, - "loss": 0.1855, + "loss": 0.791, "step": 72340 }, { "epoch": 7.64, - "grad_norm": 0.5206019878387451, + "grad_norm": 5.7890849113464355, "learning_rate": 4.718555285669026e-06, - "loss": 0.2073, + "loss": 0.4894, "step": 72350 }, { "epoch": 7.64, - "grad_norm": 25.92449378967285, + "grad_norm": 33.56106185913086, "learning_rate": 4.716443130214384e-06, - "loss": 0.55, + "loss": 0.7369, "step": 72360 }, { "epoch": 7.64, - "grad_norm": 5.529338836669922, + "grad_norm": 40.278263092041016, "learning_rate": 4.714330974759743e-06, - "loss": 0.4408, + "loss": 0.6788, "step": 72370 }, { "epoch": 7.64, - "grad_norm": 0.07130173593759537, + "grad_norm": 21.020191192626953, "learning_rate": 4.712218819305101e-06, - "loss": 0.3665, + "loss": 0.8298, "step": 72380 }, { "epoch": 7.64, - "grad_norm": 31.460220336914062, + "grad_norm": 20.05365562438965, "learning_rate": 4.71010666385046e-06, - "loss": 0.4088, + "loss": 0.8936, "step": 72390 }, { "epoch": 7.65, - "grad_norm": 3.9983346462249756, + "grad_norm": 37.31124496459961, "learning_rate": 4.707994508395818e-06, - "loss": 0.198, + "loss": 0.8772, "step": 72400 }, { "epoch": 7.65, - "grad_norm": 18.180715560913086, + "grad_norm": 38.80081558227539, "learning_rate": 4.705882352941177e-06, - "loss": 0.415, + "loss": 1.1449, "step": 72410 }, { "epoch": 7.65, - "grad_norm": 7.9448370933532715, + "grad_norm": 40.25288391113281, "learning_rate": 4.703770197486535e-06, - "loss": 0.5278, + "loss": 1.2223, "step": 72420 }, { "epoch": 7.65, - "grad_norm": 10.417877197265625, + "grad_norm": 25.298152923583984, "learning_rate": 4.7016580420318945e-06, - "loss": 0.3027, + "loss": 0.8281, "step": 72430 }, { "epoch": 7.65, - "grad_norm": 4.314939022064209, + "grad_norm": 26.676546096801758, "learning_rate": 4.699545886577252e-06, - "loss": 0.2977, + "loss": 0.8656, "step": 72440 }, { "epoch": 7.65, - "grad_norm": 5.889692783355713, + "grad_norm": 46.83574295043945, "learning_rate": 4.697433731122611e-06, - "loss": 0.1388, + "loss": 0.585, "step": 72450 }, { "epoch": 7.65, - "grad_norm": 5.583048343658447, + "grad_norm": 34.1896858215332, "learning_rate": 4.695321575667969e-06, - "loss": 0.5151, + "loss": 1.1702, "step": 72460 }, { "epoch": 7.65, - "grad_norm": 10.47220516204834, + "grad_norm": 29.80076789855957, "learning_rate": 4.693209420213328e-06, - "loss": 0.4303, + "loss": 0.7524, "step": 72470 }, { "epoch": 7.65, - "grad_norm": 0.930537760257721, + "grad_norm": 50.521976470947266, "learning_rate": 4.691097264758687e-06, - "loss": 0.4041, + "loss": 1.1834, "step": 72480 }, { "epoch": 7.66, - "grad_norm": 0.07756185531616211, + "grad_norm": 29.298494338989258, "learning_rate": 4.6889851093040454e-06, - "loss": 0.3309, + "loss": 0.8869, "step": 72490 }, { "epoch": 7.66, - "grad_norm": 22.85638427734375, + "grad_norm": 67.68445587158203, "learning_rate": 4.686872953849403e-06, - "loss": 0.5699, + "loss": 1.1272, "step": 72500 }, { "epoch": 7.66, - "grad_norm": 41.43742752075195, + "grad_norm": 51.2753791809082, "learning_rate": 4.684760798394762e-06, - "loss": 0.3332, + "loss": 0.9193, "step": 72510 }, { "epoch": 7.66, - "grad_norm": 10.605859756469727, + "grad_norm": 39.45002746582031, "learning_rate": 4.682648642940121e-06, - "loss": 0.3766, + "loss": 0.7056, "step": 72520 }, { "epoch": 7.66, - "grad_norm": 4.157266616821289, + "grad_norm": 26.599864959716797, "learning_rate": 4.680536487485479e-06, - "loss": 0.2948, + "loss": 0.7181, "step": 72530 }, { "epoch": 7.66, - "grad_norm": 4.8499531745910645, + "grad_norm": 48.50108337402344, "learning_rate": 4.6784243320308385e-06, - "loss": 0.2617, + "loss": 0.8845, "step": 72540 }, { "epoch": 7.66, - "grad_norm": 12.41739559173584, + "grad_norm": 52.785865783691406, "learning_rate": 4.676312176576196e-06, - "loss": 0.284, + "loss": 0.9335, "step": 72550 }, { "epoch": 7.66, - "grad_norm": 19.67769432067871, + "grad_norm": 44.44246292114258, "learning_rate": 4.674200021121555e-06, - "loss": 0.3209, + "loss": 0.9498, "step": 72560 }, { "epoch": 7.66, - "grad_norm": 1.0708603858947754, + "grad_norm": 34.0334358215332, "learning_rate": 4.672087865666913e-06, - "loss": 0.1419, + "loss": 0.5429, "step": 72570 }, { "epoch": 7.67, - "grad_norm": 6.050941467285156, + "grad_norm": 52.59455490112305, "learning_rate": 4.669975710212272e-06, - "loss": 0.4877, + "loss": 0.8085, "step": 72580 }, { "epoch": 7.67, - "grad_norm": 0.015787361189723015, + "grad_norm": 33.92830276489258, "learning_rate": 4.667863554757631e-06, - "loss": 0.3001, + "loss": 0.8206, "step": 72590 }, { "epoch": 7.67, - "grad_norm": 31.814556121826172, + "grad_norm": 42.940826416015625, "learning_rate": 4.665751399302989e-06, - "loss": 0.4519, + "loss": 0.6875, "step": 72600 }, { "epoch": 7.67, - "grad_norm": 17.395896911621094, + "grad_norm": 20.899030685424805, "learning_rate": 4.663639243848347e-06, - "loss": 0.316, + "loss": 0.592, "step": 72610 }, { "epoch": 7.67, - "grad_norm": 0.12986746430397034, + "grad_norm": 33.12510681152344, "learning_rate": 4.661527088393706e-06, - "loss": 0.1986, + "loss": 0.6792, "step": 72620 }, { "epoch": 7.67, - "grad_norm": 0.056006450206041336, + "grad_norm": 2.792430877685547, "learning_rate": 4.659414932939065e-06, - "loss": 0.2314, + "loss": 0.7855, "step": 72630 }, { "epoch": 7.67, - "grad_norm": 31.870180130004883, + "grad_norm": 54.71308898925781, "learning_rate": 4.657302777484423e-06, - "loss": 0.3238, + "loss": 0.953, "step": 72640 }, { "epoch": 7.67, - "grad_norm": 7.053109169006348, + "grad_norm": 25.780895233154297, "learning_rate": 4.655190622029782e-06, - "loss": 0.4588, + "loss": 0.9324, "step": 72650 }, { "epoch": 7.67, - "grad_norm": 24.466800689697266, + "grad_norm": 36.70669174194336, "learning_rate": 4.6530784665751404e-06, - "loss": 0.2984, + "loss": 0.7267, "step": 72660 }, { "epoch": 7.67, - "grad_norm": 0.10498334467411041, + "grad_norm": 32.67198944091797, "learning_rate": 4.650966311120499e-06, - "loss": 0.3039, + "loss": 0.9708, "step": 72670 }, { "epoch": 7.68, - "grad_norm": 29.1934757232666, + "grad_norm": 37.41873550415039, "learning_rate": 4.648854155665857e-06, - "loss": 0.5842, + "loss": 0.7055, "step": 72680 }, { "epoch": 7.68, - "grad_norm": 2.057347059249878, + "grad_norm": 12.316906929016113, "learning_rate": 4.646742000211216e-06, - "loss": 0.3027, + "loss": 0.8568, "step": 72690 }, { "epoch": 7.68, - "grad_norm": 19.230600357055664, + "grad_norm": 38.598236083984375, "learning_rate": 4.644629844756575e-06, - "loss": 0.633, + "loss": 0.9303, "step": 72700 }, { "epoch": 7.68, - "grad_norm": 5.637243747711182, + "grad_norm": 46.329627990722656, "learning_rate": 4.642517689301933e-06, - "loss": 0.272, + "loss": 0.6649, "step": 72710 }, { "epoch": 7.68, - "grad_norm": 2.016667366027832, + "grad_norm": 31.07103729248047, "learning_rate": 4.6404055338472914e-06, - "loss": 0.3629, + "loss": 0.8806, "step": 72720 }, { "epoch": 7.68, - "grad_norm": 12.94670581817627, + "grad_norm": 35.08353805541992, "learning_rate": 4.63829337839265e-06, - "loss": 0.4128, + "loss": 1.0027, "step": 72730 }, { "epoch": 7.68, - "grad_norm": 0.5521777868270874, + "grad_norm": 56.647178649902344, "learning_rate": 4.636181222938009e-06, - "loss": 0.3544, + "loss": 1.047, "step": 72740 }, { "epoch": 7.68, - "grad_norm": 10.271215438842773, + "grad_norm": 38.02851486206055, "learning_rate": 4.634069067483367e-06, - "loss": 0.3726, + "loss": 0.6486, "step": 72750 }, { "epoch": 7.68, - "grad_norm": 3.196169137954712, + "grad_norm": 40.669830322265625, "learning_rate": 4.631956912028726e-06, - "loss": 0.2166, + "loss": 0.8325, "step": 72760 }, { "epoch": 7.69, - "grad_norm": 30.464035034179688, + "grad_norm": 55.132843017578125, "learning_rate": 4.6298447565740845e-06, - "loss": 0.468, + "loss": 0.8951, "step": 72770 }, { "epoch": 7.69, - "grad_norm": 0.180937260389328, + "grad_norm": 14.194440841674805, "learning_rate": 4.627732601119443e-06, - "loss": 0.4409, + "loss": 0.9255, "step": 72780 }, { "epoch": 7.69, - "grad_norm": 17.226106643676758, + "grad_norm": 48.65623092651367, "learning_rate": 4.625620445664801e-06, - "loss": 0.2678, + "loss": 0.8255, "step": 72790 }, { "epoch": 7.69, - "grad_norm": 0.0567459836602211, + "grad_norm": 67.1982650756836, "learning_rate": 4.62350829021016e-06, - "loss": 0.3803, + "loss": 0.979, "step": 72800 }, { "epoch": 7.69, - "grad_norm": 4.102022171020508, + "grad_norm": 36.63792037963867, "learning_rate": 4.621396134755519e-06, - "loss": 0.5723, + "loss": 0.7314, "step": 72810 }, { "epoch": 7.69, - "grad_norm": 0.26308873295783997, + "grad_norm": 42.99497604370117, "learning_rate": 4.619283979300877e-06, - "loss": 0.2674, + "loss": 0.7519, "step": 72820 }, { "epoch": 7.69, - "grad_norm": 27.586570739746094, + "grad_norm": 42.01064682006836, "learning_rate": 4.6171718238462355e-06, - "loss": 0.3741, + "loss": 1.0337, "step": 72830 }, { "epoch": 7.69, - "grad_norm": 0.6313451528549194, + "grad_norm": 46.93286895751953, "learning_rate": 4.615059668391593e-06, - "loss": 0.3249, + "loss": 0.6717, "step": 72840 }, { "epoch": 7.69, - "grad_norm": 49.717105865478516, + "grad_norm": 41.93581771850586, "learning_rate": 4.612947512936953e-06, - "loss": 0.5819, + "loss": 0.9292, "step": 72850 }, { "epoch": 7.69, - "grad_norm": 3.1867711544036865, + "grad_norm": 34.90724563598633, "learning_rate": 4.610835357482311e-06, - "loss": 0.1632, + "loss": 1.0006, "step": 72860 }, { "epoch": 7.7, - "grad_norm": 18.67875862121582, + "grad_norm": 37.50251388549805, "learning_rate": 4.60872320202767e-06, - "loss": 0.1602, + "loss": 0.5953, "step": 72870 }, { "epoch": 7.7, - "grad_norm": 30.786088943481445, + "grad_norm": 30.02842903137207, "learning_rate": 4.606611046573028e-06, - "loss": 0.4149, + "loss": 1.0891, "step": 72880 }, { "epoch": 7.7, - "grad_norm": 17.064176559448242, + "grad_norm": 41.50509262084961, "learning_rate": 4.604498891118387e-06, - "loss": 0.5935, + "loss": 1.0371, "step": 72890 }, { "epoch": 7.7, - "grad_norm": 36.29745864868164, + "grad_norm": 81.61898803710938, "learning_rate": 4.602386735663745e-06, - "loss": 0.6447, + "loss": 0.7834, "step": 72900 }, { "epoch": 7.7, - "grad_norm": 1.4470685720443726, + "grad_norm": 40.9259033203125, "learning_rate": 4.600274580209104e-06, - "loss": 0.1759, + "loss": 0.7043, "step": 72910 }, { "epoch": 7.7, - "grad_norm": 16.60610008239746, + "grad_norm": 16.257144927978516, "learning_rate": 4.598162424754462e-06, - "loss": 0.2051, + "loss": 1.0783, "step": 72920 }, { "epoch": 7.7, - "grad_norm": 8.593973159790039, + "grad_norm": 69.70109558105469, "learning_rate": 4.596050269299821e-06, - "loss": 0.4473, + "loss": 1.1133, "step": 72930 }, { "epoch": 7.7, - "grad_norm": 20.933521270751953, + "grad_norm": 44.253055572509766, "learning_rate": 4.5939381138451795e-06, - "loss": 0.5566, + "loss": 1.0191, "step": 72940 }, { "epoch": 7.7, - "grad_norm": 5.28233528137207, + "grad_norm": 42.2840461730957, "learning_rate": 4.591825958390537e-06, - "loss": 0.5548, + "loss": 1.0, "step": 72950 }, { "epoch": 7.71, - "grad_norm": 13.740427017211914, + "grad_norm": 51.0233039855957, "learning_rate": 4.589713802935896e-06, - "loss": 0.3342, + "loss": 0.9524, "step": 72960 }, { "epoch": 7.71, - "grad_norm": 6.058391571044922, + "grad_norm": 65.60305786132812, "learning_rate": 4.587601647481255e-06, - "loss": 0.3644, + "loss": 0.7604, "step": 72970 }, { "epoch": 7.71, - "grad_norm": 0.6681767702102661, + "grad_norm": 69.6203842163086, "learning_rate": 4.585489492026614e-06, - "loss": 0.2897, + "loss": 0.9302, "step": 72980 }, { "epoch": 7.71, - "grad_norm": 0.43374884128570557, + "grad_norm": 39.41883850097656, "learning_rate": 4.583377336571972e-06, - "loss": 0.4439, + "loss": 0.764, "step": 72990 }, { "epoch": 7.71, - "grad_norm": 22.79631996154785, + "grad_norm": 36.2724609375, "learning_rate": 4.5812651811173305e-06, - "loss": 0.3956, + "loss": 0.8358, "step": 73000 }, { "epoch": 7.71, - "grad_norm": 13.797137260437012, + "grad_norm": 39.42313003540039, "learning_rate": 4.579153025662689e-06, - "loss": 0.2766, + "loss": 0.6952, "step": 73010 }, { "epoch": 7.71, - "grad_norm": 26.090192794799805, + "grad_norm": 39.383033752441406, "learning_rate": 4.577040870208048e-06, - "loss": 0.3958, + "loss": 0.9276, "step": 73020 }, { "epoch": 7.71, - "grad_norm": 0.5389100313186646, + "grad_norm": 43.12186050415039, "learning_rate": 4.574928714753406e-06, - "loss": 0.3462, + "loss": 1.0725, "step": 73030 }, { "epoch": 7.71, - "grad_norm": 11.593517303466797, + "grad_norm": 44.737945556640625, "learning_rate": 4.572816559298765e-06, - "loss": 0.4261, + "loss": 0.812, "step": 73040 }, { "epoch": 7.71, - "grad_norm": 18.337308883666992, + "grad_norm": 31.06098747253418, "learning_rate": 4.5707044038441235e-06, - "loss": 0.5917, + "loss": 0.7598, "step": 73050 }, { "epoch": 7.72, - "grad_norm": 19.03549575805664, + "grad_norm": 22.57899284362793, "learning_rate": 4.5685922483894814e-06, - "loss": 0.219, + "loss": 0.7338, "step": 73060 }, { "epoch": 7.72, - "grad_norm": 15.378865242004395, + "grad_norm": 54.46413040161133, "learning_rate": 4.56648009293484e-06, - "loss": 0.5181, + "loss": 0.9861, "step": 73070 }, { "epoch": 7.72, - "grad_norm": 25.314697265625, + "grad_norm": 35.63479232788086, "learning_rate": 4.564367937480199e-06, - "loss": 0.3242, + "loss": 0.9104, "step": 73080 }, { "epoch": 7.72, - "grad_norm": 15.96038818359375, + "grad_norm": 44.376461029052734, "learning_rate": 4.562255782025558e-06, - "loss": 0.4717, + "loss": 0.9513, "step": 73090 }, { "epoch": 7.72, - "grad_norm": 21.938467025756836, + "grad_norm": 53.26155090332031, "learning_rate": 4.560143626570916e-06, - "loss": 0.413, + "loss": 1.1533, "step": 73100 }, { "epoch": 7.72, - "grad_norm": 23.64409065246582, + "grad_norm": 28.495290756225586, "learning_rate": 4.5580314711162745e-06, - "loss": 0.2576, + "loss": 0.7829, "step": 73110 }, { "epoch": 7.72, - "grad_norm": 19.874570846557617, + "grad_norm": 49.738525390625, "learning_rate": 4.555919315661633e-06, - "loss": 0.6944, + "loss": 1.0624, "step": 73120 }, { "epoch": 7.72, - "grad_norm": 0.07371705025434494, + "grad_norm": 22.53898811340332, "learning_rate": 4.553807160206992e-06, - "loss": 0.331, + "loss": 0.6979, "step": 73130 }, { "epoch": 7.72, - "grad_norm": 11.01171588897705, + "grad_norm": 45.656436920166016, "learning_rate": 4.55169500475235e-06, - "loss": 0.3674, + "loss": 0.7936, "step": 73140 }, { "epoch": 7.73, - "grad_norm": 15.962007522583008, + "grad_norm": 39.61964416503906, "learning_rate": 4.549582849297709e-06, - "loss": 0.3619, + "loss": 1.1539, "step": 73150 }, { "epoch": 7.73, - "grad_norm": 11.737896919250488, + "grad_norm": 44.52445602416992, "learning_rate": 4.5474706938430675e-06, - "loss": 0.2215, + "loss": 0.7432, "step": 73160 }, { "epoch": 7.73, - "grad_norm": 23.783267974853516, + "grad_norm": 46.3121223449707, "learning_rate": 4.5453585383884255e-06, - "loss": 0.2648, + "loss": 0.8031, "step": 73170 }, { "epoch": 7.73, - "grad_norm": 10.937246322631836, + "grad_norm": 26.594558715820312, "learning_rate": 4.543246382933784e-06, - "loss": 0.2721, + "loss": 0.6763, "step": 73180 }, { "epoch": 7.73, - "grad_norm": 16.46481704711914, + "grad_norm": 54.01954650878906, "learning_rate": 4.541134227479142e-06, - "loss": 0.3966, + "loss": 0.7908, "step": 73190 }, { "epoch": 7.73, - "grad_norm": 14.9650297164917, + "grad_norm": 38.62772750854492, "learning_rate": 4.539022072024502e-06, - "loss": 0.3259, + "loss": 0.605, "step": 73200 }, { "epoch": 7.73, - "grad_norm": 0.030800295993685722, + "grad_norm": 35.833900451660156, "learning_rate": 4.53690991656986e-06, - "loss": 0.3275, + "loss": 0.7455, "step": 73210 }, { "epoch": 7.73, - "grad_norm": 17.381912231445312, + "grad_norm": 35.04436492919922, "learning_rate": 4.5347977611152185e-06, - "loss": 0.3442, + "loss": 0.9418, "step": 73220 }, { "epoch": 7.73, - "grad_norm": 18.032264709472656, + "grad_norm": 32.62653350830078, "learning_rate": 4.532685605660577e-06, - "loss": 0.2312, + "loss": 0.64, "step": 73230 }, { "epoch": 7.73, - "grad_norm": 3.2024407386779785, + "grad_norm": 9.228525161743164, "learning_rate": 4.530573450205935e-06, - "loss": 0.312, + "loss": 0.8913, "step": 73240 }, { "epoch": 7.74, - "grad_norm": 4.406601905822754, + "grad_norm": 45.28355407714844, "learning_rate": 4.528461294751294e-06, - "loss": 0.3202, + "loss": 0.7808, "step": 73250 }, { "epoch": 7.74, - "grad_norm": 22.607025146484375, + "grad_norm": 37.30604934692383, "learning_rate": 4.526349139296653e-06, - "loss": 0.1918, + "loss": 0.6521, "step": 73260 }, { "epoch": 7.74, - "grad_norm": 16.377683639526367, + "grad_norm": 44.7821159362793, "learning_rate": 4.5242369838420116e-06, - "loss": 0.3151, + "loss": 1.0049, "step": 73270 }, { "epoch": 7.74, - "grad_norm": 7.846593379974365, + "grad_norm": 31.488554000854492, "learning_rate": 4.5221248283873695e-06, - "loss": 0.3882, + "loss": 0.8992, "step": 73280 }, { "epoch": 7.74, - "grad_norm": 36.03651428222656, + "grad_norm": 49.18708419799805, "learning_rate": 4.520012672932728e-06, - "loss": 0.3741, + "loss": 0.8353, "step": 73290 }, { "epoch": 7.74, - "grad_norm": 5.732601165771484, + "grad_norm": 34.01760482788086, "learning_rate": 4.517900517478086e-06, - "loss": 0.2063, + "loss": 0.7811, "step": 73300 }, { "epoch": 7.74, - "grad_norm": 8.24877643585205, + "grad_norm": 35.756553649902344, "learning_rate": 4.515788362023446e-06, - "loss": 0.4011, + "loss": 0.8064, "step": 73310 }, { "epoch": 7.74, - "grad_norm": 8.338444709777832, + "grad_norm": 10.789972305297852, "learning_rate": 4.513676206568804e-06, - "loss": 0.5356, + "loss": 0.9285, "step": 73320 }, { "epoch": 7.74, - "grad_norm": 33.170406341552734, + "grad_norm": 50.11872482299805, "learning_rate": 4.5115640511141625e-06, - "loss": 0.3477, + "loss": 1.044, "step": 73330 }, { "epoch": 7.75, - "grad_norm": 8.750493049621582, + "grad_norm": 46.73779296875, "learning_rate": 4.5094518956595205e-06, - "loss": 0.2725, + "loss": 0.9133, "step": 73340 }, { "epoch": 7.75, - "grad_norm": 7.998171806335449, + "grad_norm": 48.4254150390625, "learning_rate": 4.507339740204879e-06, - "loss": 0.2523, + "loss": 0.7311, "step": 73350 }, { "epoch": 7.75, - "grad_norm": 20.53316879272461, + "grad_norm": 61.64234161376953, "learning_rate": 4.505227584750238e-06, - "loss": 0.2076, + "loss": 0.8342, "step": 73360 }, { "epoch": 7.75, - "grad_norm": 2.0292916297912598, + "grad_norm": 41.83656692504883, "learning_rate": 4.503115429295597e-06, - "loss": 0.0652, + "loss": 0.8517, "step": 73370 }, { "epoch": 7.75, - "grad_norm": 19.79364013671875, + "grad_norm": 23.063472747802734, "learning_rate": 4.501003273840955e-06, - "loss": 0.4411, + "loss": 0.9937, "step": 73380 }, { "epoch": 7.75, - "grad_norm": 0.3596009612083435, + "grad_norm": 24.9834041595459, "learning_rate": 4.4988911183863135e-06, - "loss": 0.1177, + "loss": 0.5493, "step": 73390 }, { "epoch": 7.75, - "grad_norm": 0.8260490894317627, + "grad_norm": 31.510478973388672, "learning_rate": 4.496778962931672e-06, - "loss": 0.4447, + "loss": 0.9544, "step": 73400 }, { "epoch": 7.75, - "grad_norm": 0.9646775126457214, + "grad_norm": 36.486934661865234, "learning_rate": 4.49466680747703e-06, - "loss": 0.3524, + "loss": 1.1024, "step": 73410 }, { "epoch": 7.75, - "grad_norm": 5.48388671875, + "grad_norm": 32.27798080444336, "learning_rate": 4.492554652022389e-06, - "loss": 0.5486, + "loss": 1.0348, "step": 73420 }, { "epoch": 7.75, - "grad_norm": 11.577279090881348, + "grad_norm": 48.06428146362305, "learning_rate": 4.490442496567748e-06, - "loss": 0.1862, + "loss": 1.1997, "step": 73430 }, { "epoch": 7.76, - "grad_norm": 0.3555379807949066, + "grad_norm": 38.141456604003906, "learning_rate": 4.4883303411131066e-06, - "loss": 0.2439, + "loss": 0.5883, "step": 73440 }, { "epoch": 7.76, - "grad_norm": 7.371316432952881, + "grad_norm": 29.991146087646484, "learning_rate": 4.4862181856584645e-06, - "loss": 0.2622, + "loss": 0.7626, "step": 73450 }, { "epoch": 7.76, - "grad_norm": 24.25235366821289, + "grad_norm": 46.32408142089844, "learning_rate": 4.484106030203823e-06, - "loss": 0.5803, + "loss": 1.1939, "step": 73460 }, { "epoch": 7.76, - "grad_norm": 25.66008186340332, + "grad_norm": 39.3661003112793, "learning_rate": 4.481993874749182e-06, - "loss": 0.4403, + "loss": 0.8748, "step": 73470 }, { "epoch": 7.76, - "grad_norm": 2.87286114692688, + "grad_norm": 24.404443740844727, "learning_rate": 4.479881719294541e-06, - "loss": 0.37, + "loss": 0.5219, "step": 73480 }, { "epoch": 7.76, - "grad_norm": 16.748449325561523, + "grad_norm": 64.47806549072266, "learning_rate": 4.477769563839899e-06, - "loss": 0.4223, + "loss": 0.8936, "step": 73490 }, { "epoch": 7.76, - "grad_norm": 0.38207828998565674, + "grad_norm": 38.38155746459961, "learning_rate": 4.4756574083852575e-06, - "loss": 0.3971, + "loss": 0.7702, "step": 73500 }, { "epoch": 7.76, - "grad_norm": 0.24556663632392883, + "grad_norm": 43.467742919921875, "learning_rate": 4.473545252930616e-06, - "loss": 0.3616, + "loss": 1.0256, "step": 73510 }, { "epoch": 7.76, - "grad_norm": 1.608725905418396, + "grad_norm": 27.10320281982422, "learning_rate": 4.471433097475974e-06, - "loss": 0.4334, + "loss": 0.7386, "step": 73520 }, { "epoch": 7.77, - "grad_norm": 2.898695230484009, + "grad_norm": 58.00650405883789, "learning_rate": 4.469320942021333e-06, - "loss": 0.5079, + "loss": 0.7751, "step": 73530 }, { "epoch": 7.77, - "grad_norm": 21.70574188232422, + "grad_norm": 45.19317626953125, "learning_rate": 4.467208786566692e-06, - "loss": 0.4115, + "loss": 0.9468, "step": 73540 }, { "epoch": 7.77, - "grad_norm": 0.22013981640338898, + "grad_norm": 6.1956658363342285, "learning_rate": 4.465096631112051e-06, - "loss": 0.2887, + "loss": 1.0835, "step": 73550 }, { "epoch": 7.77, - "grad_norm": 26.81087875366211, + "grad_norm": 41.082969665527344, "learning_rate": 4.4629844756574085e-06, - "loss": 0.4844, + "loss": 1.0656, "step": 73560 }, { "epoch": 7.77, - "grad_norm": 9.117496490478516, + "grad_norm": 44.46257019042969, "learning_rate": 4.460872320202767e-06, - "loss": 0.3769, + "loss": 0.878, "step": 73570 }, { "epoch": 7.77, - "grad_norm": 0.08062877506017685, + "grad_norm": 33.05458450317383, "learning_rate": 4.458760164748126e-06, - "loss": 0.2098, + "loss": 0.944, "step": 73580 }, { "epoch": 7.77, - "grad_norm": 4.193793773651123, + "grad_norm": 42.7156982421875, "learning_rate": 4.456648009293484e-06, - "loss": 0.2958, + "loss": 0.8501, "step": 73590 }, { "epoch": 7.77, - "grad_norm": 0.5785910487174988, + "grad_norm": 34.89545822143555, "learning_rate": 4.454535853838843e-06, - "loss": 0.2915, + "loss": 1.1603, "step": 73600 }, { "epoch": 7.77, - "grad_norm": 0.4385339915752411, + "grad_norm": 23.39493179321289, "learning_rate": 4.4524236983842016e-06, - "loss": 0.309, + "loss": 0.6583, "step": 73610 }, { "epoch": 7.77, - "grad_norm": 1.210656762123108, + "grad_norm": 15.527189254760742, "learning_rate": 4.45031154292956e-06, - "loss": 0.3785, + "loss": 0.9191, "step": 73620 }, { "epoch": 7.78, - "grad_norm": 10.395580291748047, + "grad_norm": 57.160186767578125, "learning_rate": 4.448199387474918e-06, - "loss": 0.3292, + "loss": 0.8145, "step": 73630 }, { "epoch": 7.78, - "grad_norm": 17.779823303222656, + "grad_norm": 60.79945373535156, "learning_rate": 4.446087232020277e-06, - "loss": 0.5047, + "loss": 0.9172, "step": 73640 }, { "epoch": 7.78, - "grad_norm": 15.668046951293945, + "grad_norm": 21.19878387451172, "learning_rate": 4.443975076565636e-06, - "loss": 0.1883, + "loss": 0.6731, "step": 73650 }, { "epoch": 7.78, - "grad_norm": 10.446332931518555, + "grad_norm": 25.076757431030273, "learning_rate": 4.441862921110995e-06, - "loss": 0.4226, + "loss": 1.0122, "step": 73660 }, { "epoch": 7.78, - "grad_norm": 46.68268966674805, + "grad_norm": 37.99405288696289, "learning_rate": 4.4397507656563525e-06, - "loss": 0.1392, + "loss": 0.5771, "step": 73670 }, { "epoch": 7.78, - "grad_norm": 24.29764175415039, + "grad_norm": 39.94744110107422, "learning_rate": 4.437638610201711e-06, - "loss": 0.4913, + "loss": 1.0213, "step": 73680 }, { "epoch": 7.78, - "grad_norm": 26.664653778076172, + "grad_norm": 55.13639450073242, "learning_rate": 4.43552645474707e-06, - "loss": 0.3253, + "loss": 0.8266, "step": 73690 }, { "epoch": 7.78, - "grad_norm": 0.4852583706378937, + "grad_norm": 48.501766204833984, "learning_rate": 4.433414299292428e-06, - "loss": 0.1083, + "loss": 0.6669, "step": 73700 }, { "epoch": 7.78, - "grad_norm": 16.786638259887695, + "grad_norm": 44.10354232788086, "learning_rate": 4.431302143837787e-06, - "loss": 0.2386, + "loss": 0.7726, "step": 73710 }, { "epoch": 7.79, - "grad_norm": 19.288713455200195, + "grad_norm": 55.24021911621094, "learning_rate": 4.429189988383146e-06, - "loss": 0.6384, + "loss": 1.1827, "step": 73720 }, { "epoch": 7.79, - "grad_norm": 2.189851760864258, + "grad_norm": 60.94464874267578, "learning_rate": 4.427077832928504e-06, - "loss": 0.2741, + "loss": 1.1471, "step": 73730 }, { "epoch": 7.79, - "grad_norm": 0.19010654091835022, + "grad_norm": 31.198762893676758, "learning_rate": 4.424965677473862e-06, - "loss": 0.4615, + "loss": 0.8911, "step": 73740 }, { "epoch": 7.79, - "grad_norm": 41.972164154052734, + "grad_norm": 39.288414001464844, "learning_rate": 4.422853522019221e-06, - "loss": 0.4022, + "loss": 1.0034, "step": 73750 }, { "epoch": 7.79, - "grad_norm": 31.153133392333984, + "grad_norm": 61.180904388427734, "learning_rate": 4.420741366564579e-06, - "loss": 0.5913, + "loss": 1.2547, "step": 73760 }, { "epoch": 7.79, - "grad_norm": 12.320566177368164, + "grad_norm": 67.87934112548828, "learning_rate": 4.418629211109939e-06, - "loss": 0.5106, + "loss": 1.0813, "step": 73770 }, { "epoch": 7.79, - "grad_norm": 14.641850471496582, + "grad_norm": 6.640219211578369, "learning_rate": 4.4165170556552966e-06, - "loss": 0.3117, + "loss": 0.9322, "step": 73780 }, { "epoch": 7.79, - "grad_norm": 19.769365310668945, + "grad_norm": 35.78032684326172, "learning_rate": 4.414404900200655e-06, - "loss": 0.1845, + "loss": 0.7671, "step": 73790 }, { "epoch": 7.79, - "grad_norm": 15.414813041687012, + "grad_norm": 34.368595123291016, "learning_rate": 4.412292744746013e-06, - "loss": 0.4196, + "loss": 0.7236, "step": 73800 }, { "epoch": 7.79, - "grad_norm": 25.292293548583984, + "grad_norm": 45.361328125, "learning_rate": 4.410180589291372e-06, - "loss": 0.468, + "loss": 1.251, "step": 73810 }, { "epoch": 7.8, - "grad_norm": 0.45697030425071716, + "grad_norm": 47.06184768676758, "learning_rate": 4.408068433836731e-06, - "loss": 0.5219, + "loss": 1.0253, "step": 73820 }, { "epoch": 7.8, - "grad_norm": 3.2428691387176514, + "grad_norm": 46.467796325683594, "learning_rate": 4.405956278382089e-06, - "loss": 0.2503, + "loss": 0.6165, "step": 73830 }, { "epoch": 7.8, - "grad_norm": 10.067049026489258, + "grad_norm": 40.17295455932617, "learning_rate": 4.4038441229274476e-06, - "loss": 0.2544, + "loss": 0.6669, "step": 73840 }, { "epoch": 7.8, - "grad_norm": 22.047828674316406, + "grad_norm": 47.375247955322266, "learning_rate": 4.401731967472806e-06, - "loss": 0.3549, + "loss": 1.187, "step": 73850 }, { "epoch": 7.8, - "grad_norm": 6.214317798614502, + "grad_norm": 46.29753494262695, "learning_rate": 4.399619812018165e-06, - "loss": 0.4861, + "loss": 1.014, "step": 73860 }, { "epoch": 7.8, - "grad_norm": 32.8359260559082, + "grad_norm": 36.924644470214844, "learning_rate": 4.397507656563523e-06, - "loss": 0.3909, + "loss": 0.7048, "step": 73870 }, { "epoch": 7.8, - "grad_norm": 0.39503100514411926, + "grad_norm": 26.70679473876953, "learning_rate": 4.395395501108883e-06, - "loss": 0.2364, + "loss": 0.7547, "step": 73880 }, { "epoch": 7.8, - "grad_norm": 11.192012786865234, + "grad_norm": 41.98200607299805, "learning_rate": 4.393283345654241e-06, - "loss": 0.2172, + "loss": 0.7698, "step": 73890 }, { "epoch": 7.8, - "grad_norm": 10.315020561218262, + "grad_norm": 12.13925838470459, "learning_rate": 4.391171190199599e-06, - "loss": 0.3448, + "loss": 0.841, "step": 73900 }, { "epoch": 7.81, - "grad_norm": 4.372607231140137, + "grad_norm": 32.5552978515625, "learning_rate": 4.389059034744957e-06, - "loss": 0.2259, + "loss": 0.7637, "step": 73910 }, { "epoch": 7.81, - "grad_norm": 1.2432966232299805, + "grad_norm": 34.395320892333984, "learning_rate": 4.386946879290316e-06, - "loss": 0.1986, + "loss": 0.6927, "step": 73920 }, { "epoch": 7.81, - "grad_norm": 14.225735664367676, + "grad_norm": 58.22545623779297, "learning_rate": 4.384834723835675e-06, - "loss": 0.4947, + "loss": 1.1922, "step": 73930 }, { "epoch": 7.81, - "grad_norm": 17.105215072631836, + "grad_norm": 66.36048889160156, "learning_rate": 4.382722568381033e-06, - "loss": 0.4997, + "loss": 0.9531, "step": 73940 }, { "epoch": 7.81, - "grad_norm": 0.6176624298095703, + "grad_norm": 4.450632572174072, "learning_rate": 4.380610412926392e-06, - "loss": 0.3977, + "loss": 0.7902, "step": 73950 }, { "epoch": 7.81, - "grad_norm": 9.064380645751953, + "grad_norm": 39.52389144897461, "learning_rate": 4.37849825747175e-06, - "loss": 0.4533, + "loss": 0.5878, "step": 73960 }, { "epoch": 7.81, - "grad_norm": 0.32042765617370605, + "grad_norm": 37.81099319458008, "learning_rate": 4.376386102017109e-06, - "loss": 0.321, + "loss": 0.9549, "step": 73970 }, { "epoch": 7.81, - "grad_norm": 23.706188201904297, + "grad_norm": 35.80106735229492, "learning_rate": 4.374273946562467e-06, - "loss": 0.7696, + "loss": 0.9701, "step": 73980 }, { "epoch": 7.81, - "grad_norm": 5.463258743286133, + "grad_norm": 38.764549255371094, "learning_rate": 4.372161791107826e-06, - "loss": 0.5032, + "loss": 0.9881, "step": 73990 }, { "epoch": 7.81, - "grad_norm": 18.493196487426758, + "grad_norm": 65.08222961425781, "learning_rate": 4.370049635653185e-06, - "loss": 0.3301, + "loss": 0.8762, "step": 74000 }, { "epoch": 7.82, - "grad_norm": 7.352053165435791, + "grad_norm": 25.844650268554688, "learning_rate": 4.367937480198543e-06, - "loss": 0.2642, + "loss": 0.516, "step": 74010 }, { "epoch": 7.82, - "grad_norm": 10.52470874786377, + "grad_norm": 17.31865882873535, "learning_rate": 4.365825324743901e-06, - "loss": 0.6397, + "loss": 0.9284, "step": 74020 }, { "epoch": 7.82, - "grad_norm": 9.033182144165039, + "grad_norm": 62.45405578613281, "learning_rate": 4.36371316928926e-06, - "loss": 0.2776, + "loss": 0.8667, "step": 74030 }, { "epoch": 7.82, - "grad_norm": 0.5244602560997009, + "grad_norm": 47.077796936035156, "learning_rate": 4.361601013834619e-06, - "loss": 0.6647, + "loss": 1.299, "step": 74040 }, { "epoch": 7.82, - "grad_norm": 14.665353775024414, + "grad_norm": 33.53680419921875, "learning_rate": 4.359488858379977e-06, - "loss": 0.3188, + "loss": 0.8335, "step": 74050 }, { "epoch": 7.82, - "grad_norm": 2.3589742183685303, + "grad_norm": 34.63634490966797, "learning_rate": 4.357376702925336e-06, - "loss": 0.3389, + "loss": 1.1278, "step": 74060 }, { "epoch": 7.82, - "grad_norm": 2.3879663944244385, + "grad_norm": 40.092952728271484, "learning_rate": 4.3552645474706935e-06, - "loss": 0.2926, + "loss": 0.8596, "step": 74070 }, { "epoch": 7.82, - "grad_norm": 7.245311260223389, + "grad_norm": 26.13100814819336, "learning_rate": 4.353152392016053e-06, - "loss": 0.3864, + "loss": 0.7825, "step": 74080 }, { "epoch": 7.82, - "grad_norm": 74.2876968383789, + "grad_norm": 43.58717346191406, "learning_rate": 4.351040236561411e-06, - "loss": 0.7249, + "loss": 1.0931, "step": 74090 }, { "epoch": 7.83, - "grad_norm": 10.531655311584473, + "grad_norm": 47.05747604370117, "learning_rate": 4.34892808110677e-06, - "loss": 0.3609, + "loss": 0.9822, "step": 74100 }, { "epoch": 7.83, - "grad_norm": 5.876444339752197, + "grad_norm": 32.69744110107422, "learning_rate": 4.346815925652129e-06, - "loss": 0.3766, + "loss": 1.1738, "step": 74110 }, { "epoch": 7.83, - "grad_norm": 2.2187955379486084, + "grad_norm": 25.696819305419922, "learning_rate": 4.3447037701974874e-06, - "loss": 0.2852, + "loss": 0.7327, "step": 74120 }, { "epoch": 7.83, - "grad_norm": 0.27003100514411926, + "grad_norm": 52.518924713134766, "learning_rate": 4.342591614742845e-06, - "loss": 0.4849, + "loss": 0.8639, "step": 74130 }, { "epoch": 7.83, - "grad_norm": 28.609628677368164, + "grad_norm": 69.92491149902344, "learning_rate": 4.340479459288204e-06, - "loss": 0.5579, + "loss": 1.1834, "step": 74140 }, { "epoch": 7.83, - "grad_norm": 0.548175036907196, + "grad_norm": 11.79142951965332, "learning_rate": 4.338367303833563e-06, - "loss": 0.278, + "loss": 0.667, "step": 74150 }, { "epoch": 7.83, - "grad_norm": 0.14970748126506805, + "grad_norm": 42.684242248535156, "learning_rate": 4.336255148378921e-06, - "loss": 0.306, + "loss": 0.6482, "step": 74160 }, { "epoch": 7.83, - "grad_norm": 9.550860404968262, + "grad_norm": 34.11663818359375, "learning_rate": 4.33414299292428e-06, - "loss": 0.4489, + "loss": 0.9589, "step": 74170 }, { "epoch": 7.83, - "grad_norm": 9.608717918395996, + "grad_norm": 45.04851531982422, "learning_rate": 4.3320308374696376e-06, - "loss": 0.9672, + "loss": 0.968, "step": 74180 }, { "epoch": 7.84, - "grad_norm": 13.936474800109863, + "grad_norm": 57.145084381103516, "learning_rate": 4.329918682014997e-06, - "loss": 0.3904, + "loss": 0.8935, "step": 74190 }, { "epoch": 7.84, - "grad_norm": 0.6951881647109985, + "grad_norm": 27.182939529418945, "learning_rate": 4.327806526560355e-06, - "loss": 0.1932, + "loss": 0.6704, "step": 74200 }, { "epoch": 7.84, - "grad_norm": 10.630921363830566, + "grad_norm": 29.78841781616211, "learning_rate": 4.325694371105714e-06, - "loss": 0.3332, + "loss": 1.2801, "step": 74210 }, { "epoch": 7.84, - "grad_norm": 14.528374671936035, + "grad_norm": 28.712800979614258, "learning_rate": 4.323582215651072e-06, - "loss": 0.3254, + "loss": 0.9673, "step": 74220 }, { "epoch": 7.84, - "grad_norm": 14.789806365966797, + "grad_norm": 65.75530242919922, "learning_rate": 4.321470060196431e-06, - "loss": 0.324, + "loss": 1.1238, "step": 74230 }, { "epoch": 7.84, - "grad_norm": 17.658063888549805, + "grad_norm": 43.05975341796875, "learning_rate": 4.319357904741789e-06, - "loss": 0.4748, + "loss": 0.9735, "step": 74240 }, { "epoch": 7.84, - "grad_norm": 3.1753406524658203, + "grad_norm": 29.62091636657715, "learning_rate": 4.317245749287148e-06, - "loss": 0.2629, + "loss": 0.8559, "step": 74250 }, { "epoch": 7.84, - "grad_norm": 0.42871785163879395, + "grad_norm": 33.775779724121094, "learning_rate": 4.315133593832506e-06, - "loss": 0.0993, + "loss": 0.7216, "step": 74260 }, { "epoch": 7.84, - "grad_norm": 0.9158266186714172, + "grad_norm": 36.40878677368164, "learning_rate": 4.313021438377865e-06, - "loss": 0.3131, + "loss": 0.8168, "step": 74270 }, { "epoch": 7.84, - "grad_norm": 0.6850767135620117, + "grad_norm": 56.92228317260742, "learning_rate": 4.310909282923224e-06, - "loss": 0.299, + "loss": 1.0511, "step": 74280 }, { "epoch": 7.85, - "grad_norm": 7.046265602111816, + "grad_norm": 40.34537124633789, "learning_rate": 4.308797127468582e-06, - "loss": 0.5166, + "loss": 1.0309, "step": 74290 }, { "epoch": 7.85, - "grad_norm": 3.265669345855713, + "grad_norm": 54.28852081298828, "learning_rate": 4.30668497201394e-06, - "loss": 0.4287, + "loss": 0.8974, "step": 74300 }, { "epoch": 7.85, - "grad_norm": 0.09087672829627991, + "grad_norm": 29.642427444458008, "learning_rate": 4.304572816559299e-06, - "loss": 0.4433, + "loss": 0.6141, "step": 74310 }, { "epoch": 7.85, - "grad_norm": 0.04179533198475838, + "grad_norm": 28.352293014526367, "learning_rate": 4.302460661104658e-06, - "loss": 0.3663, + "loss": 1.2137, "step": 74320 }, { "epoch": 7.85, - "grad_norm": 10.455039024353027, + "grad_norm": 54.60948181152344, "learning_rate": 4.300348505650016e-06, - "loss": 0.347, + "loss": 0.7447, "step": 74330 }, { "epoch": 7.85, - "grad_norm": 18.858251571655273, + "grad_norm": 55.53760528564453, "learning_rate": 4.298236350195375e-06, - "loss": 0.3504, + "loss": 0.8364, "step": 74340 }, { "epoch": 7.85, - "grad_norm": 33.223819732666016, + "grad_norm": 35.60714340209961, "learning_rate": 4.296124194740733e-06, - "loss": 0.27, + "loss": 0.8729, "step": 74350 }, { "epoch": 7.85, - "grad_norm": 40.730037689208984, + "grad_norm": 66.15250396728516, "learning_rate": 4.294012039286092e-06, - "loss": 0.4872, + "loss": 0.8285, "step": 74360 }, { "epoch": 7.85, - "grad_norm": 0.21824051439762115, + "grad_norm": 28.41696548461914, "learning_rate": 4.29189988383145e-06, - "loss": 0.3356, + "loss": 1.0209, "step": 74370 }, { "epoch": 7.86, - "grad_norm": 16.80139923095703, + "grad_norm": 45.91618347167969, "learning_rate": 4.289787728376809e-06, - "loss": 0.3392, + "loss": 0.8809, "step": 74380 }, { "epoch": 7.86, - "grad_norm": 40.0723876953125, + "grad_norm": 15.924152374267578, "learning_rate": 4.287675572922168e-06, - "loss": 0.5338, + "loss": 0.9427, "step": 74390 }, { "epoch": 7.86, - "grad_norm": 13.10727310180664, + "grad_norm": 49.99258804321289, "learning_rate": 4.285563417467526e-06, - "loss": 0.3321, + "loss": 0.9177, "step": 74400 }, { "epoch": 7.86, - "grad_norm": 15.038656234741211, + "grad_norm": 35.57398223876953, "learning_rate": 4.283451262012884e-06, - "loss": 0.3777, + "loss": 0.8972, "step": 74410 }, { "epoch": 7.86, - "grad_norm": 2.065048933029175, + "grad_norm": 42.005985260009766, "learning_rate": 4.281339106558243e-06, - "loss": 0.3698, + "loss": 1.0964, "step": 74420 }, { "epoch": 7.86, - "grad_norm": 0.28324946761131287, + "grad_norm": 36.12080001831055, "learning_rate": 4.279226951103602e-06, - "loss": 0.2683, + "loss": 0.7479, "step": 74430 }, { "epoch": 7.86, - "grad_norm": 18.9285831451416, + "grad_norm": 28.30208396911621, "learning_rate": 4.27711479564896e-06, - "loss": 0.5917, + "loss": 1.0972, "step": 74440 }, { "epoch": 7.86, - "grad_norm": 13.586396217346191, + "grad_norm": 34.47520446777344, "learning_rate": 4.275002640194319e-06, - "loss": 0.3605, + "loss": 0.783, "step": 74450 }, { "epoch": 7.86, - "grad_norm": 0.5788981914520264, + "grad_norm": 33.145606994628906, "learning_rate": 4.2728904847396774e-06, - "loss": 0.3699, + "loss": 0.7601, "step": 74460 }, { "epoch": 7.86, - "grad_norm": 3.775200605392456, + "grad_norm": 6.837944030761719, "learning_rate": 4.270778329285035e-06, - "loss": 0.2125, + "loss": 0.5588, "step": 74470 }, { "epoch": 7.87, - "grad_norm": 0.29504522681236267, + "grad_norm": 29.573305130004883, "learning_rate": 4.268666173830394e-06, - "loss": 0.2124, + "loss": 0.7197, "step": 74480 }, { "epoch": 7.87, - "grad_norm": 21.20707130432129, + "grad_norm": 60.414894104003906, "learning_rate": 4.266554018375753e-06, - "loss": 0.3192, + "loss": 1.0981, "step": 74490 }, { "epoch": 7.87, - "grad_norm": 6.754000663757324, + "grad_norm": 29.729822158813477, "learning_rate": 4.264441862921112e-06, - "loss": 0.2027, + "loss": 0.8147, "step": 74500 }, { "epoch": 7.87, - "grad_norm": 0.8465373516082764, + "grad_norm": 47.109920501708984, "learning_rate": 4.26232970746647e-06, - "loss": 0.4239, + "loss": 1.0706, "step": 74510 }, { "epoch": 7.87, - "grad_norm": 4.391017436981201, + "grad_norm": 22.95611572265625, "learning_rate": 4.260217552011828e-06, - "loss": 0.2592, + "loss": 0.7921, "step": 74520 }, { "epoch": 7.87, - "grad_norm": 0.9963490962982178, + "grad_norm": 42.87730026245117, "learning_rate": 4.258105396557186e-06, - "loss": 0.5107, + "loss": 0.9477, "step": 74530 }, { "epoch": 7.87, - "grad_norm": 7.931969165802002, + "grad_norm": 56.88566207885742, "learning_rate": 4.255993241102546e-06, - "loss": 0.3951, + "loss": 0.9751, "step": 74540 }, { "epoch": 7.87, - "grad_norm": 15.220813751220703, + "grad_norm": 40.2204704284668, "learning_rate": 4.253881085647904e-06, - "loss": 0.4825, + "loss": 0.9915, "step": 74550 }, { "epoch": 7.87, - "grad_norm": 32.97338104248047, + "grad_norm": 45.48074722290039, "learning_rate": 4.251768930193263e-06, - "loss": 0.4593, + "loss": 0.9862, "step": 74560 }, { "epoch": 7.88, - "grad_norm": 2.54538893699646, + "grad_norm": 47.4946174621582, "learning_rate": 4.2496567747386215e-06, - "loss": 0.5969, + "loss": 0.8542, "step": 74570 }, { "epoch": 7.88, - "grad_norm": 1.52864670753479, + "grad_norm": 25.106685638427734, "learning_rate": 4.247544619283979e-06, - "loss": 0.302, + "loss": 0.6631, "step": 74580 }, { "epoch": 7.88, - "grad_norm": 24.69964599609375, + "grad_norm": 16.46446990966797, "learning_rate": 4.245432463829338e-06, - "loss": 0.4921, + "loss": 1.04, "step": 74590 }, { "epoch": 7.88, - "grad_norm": 2.381385087966919, + "grad_norm": 30.21912384033203, "learning_rate": 4.243320308374697e-06, - "loss": 0.3973, + "loss": 0.8582, "step": 74600 }, { "epoch": 7.88, - "grad_norm": 26.77330780029297, + "grad_norm": 43.35137939453125, "learning_rate": 4.241208152920056e-06, - "loss": 0.3088, + "loss": 0.8535, "step": 74610 }, { "epoch": 7.88, - "grad_norm": 0.5196089148521423, + "grad_norm": 27.798004150390625, "learning_rate": 4.239095997465414e-06, - "loss": 0.2235, + "loss": 0.8395, "step": 74620 }, { "epoch": 7.88, - "grad_norm": 7.627772331237793, + "grad_norm": 11.823080062866211, "learning_rate": 4.2369838420107724e-06, - "loss": 0.4661, + "loss": 0.9769, "step": 74630 }, { "epoch": 7.88, - "grad_norm": 16.227359771728516, + "grad_norm": 40.56797790527344, "learning_rate": 4.23487168655613e-06, - "loss": 0.2675, + "loss": 0.8243, "step": 74640 }, { "epoch": 7.88, - "grad_norm": 40.13517379760742, + "grad_norm": 53.95729064941406, "learning_rate": 4.23275953110149e-06, - "loss": 0.1801, + "loss": 0.9254, "step": 74650 }, { "epoch": 7.88, - "grad_norm": 5.941359519958496, + "grad_norm": 54.00913619995117, "learning_rate": 4.230647375646848e-06, - "loss": 0.2727, + "loss": 0.6873, "step": 74660 }, { "epoch": 7.89, - "grad_norm": 11.265010833740234, + "grad_norm": 56.370609283447266, "learning_rate": 4.228535220192207e-06, - "loss": 0.3372, + "loss": 0.7456, "step": 74670 }, { "epoch": 7.89, - "grad_norm": 3.5763418674468994, + "grad_norm": 51.262969970703125, "learning_rate": 4.226423064737565e-06, - "loss": 0.2906, + "loss": 0.9792, "step": 74680 }, { "epoch": 7.89, - "grad_norm": 32.08008575439453, + "grad_norm": 45.88380813598633, "learning_rate": 4.2243109092829234e-06, - "loss": 0.6816, + "loss": 1.2164, "step": 74690 }, { "epoch": 7.89, - "grad_norm": 2.01832914352417, + "grad_norm": 39.913818359375, "learning_rate": 4.222198753828282e-06, - "loss": 0.3725, + "loss": 0.8774, "step": 74700 }, { "epoch": 7.89, - "grad_norm": 0.23675434291362762, + "grad_norm": 29.69436264038086, "learning_rate": 4.22008659837364e-06, - "loss": 0.4034, + "loss": 1.0959, "step": 74710 }, { "epoch": 7.89, - "grad_norm": 0.03801574558019638, + "grad_norm": 66.64119720458984, "learning_rate": 4.217974442918999e-06, - "loss": 0.2157, + "loss": 0.6972, "step": 74720 }, { "epoch": 7.89, - "grad_norm": 15.69558048248291, + "grad_norm": 44.65019607543945, "learning_rate": 4.215862287464358e-06, - "loss": 0.3913, + "loss": 0.6679, "step": 74730 }, { "epoch": 7.89, - "grad_norm": 10.383456230163574, + "grad_norm": 50.0501594543457, "learning_rate": 4.2137501320097165e-06, - "loss": 0.205, + "loss": 1.1562, "step": 74740 }, { "epoch": 7.89, - "grad_norm": 17.209339141845703, + "grad_norm": 51.95746994018555, "learning_rate": 4.211637976555074e-06, - "loss": 0.4431, + "loss": 1.0009, "step": 74750 }, { "epoch": 7.9, - "grad_norm": 20.794282913208008, + "grad_norm": 35.55929946899414, "learning_rate": 4.209525821100433e-06, - "loss": 0.619, + "loss": 0.7739, "step": 74760 }, { "epoch": 7.9, - "grad_norm": 14.73742961883545, + "grad_norm": 24.067960739135742, "learning_rate": 4.207413665645792e-06, - "loss": 0.4803, + "loss": 0.7763, "step": 74770 }, { "epoch": 7.9, - "grad_norm": 0.36124712228775024, + "grad_norm": 41.058128356933594, "learning_rate": 4.205301510191151e-06, - "loss": 0.2604, + "loss": 0.8583, "step": 74780 }, { "epoch": 7.9, - "grad_norm": 24.320402145385742, + "grad_norm": 46.06356430053711, "learning_rate": 4.203189354736509e-06, - "loss": 0.4079, + "loss": 0.9198, "step": 74790 }, { "epoch": 7.9, - "grad_norm": 0.63112872838974, + "grad_norm": 39.314918518066406, "learning_rate": 4.2010771992818675e-06, - "loss": 0.4767, + "loss": 0.9511, "step": 74800 }, { "epoch": 7.9, - "grad_norm": 15.829438209533691, + "grad_norm": 34.150142669677734, "learning_rate": 4.198965043827226e-06, - "loss": 0.4606, + "loss": 0.897, "step": 74810 }, { "epoch": 7.9, - "grad_norm": 11.432676315307617, + "grad_norm": 50.54690933227539, "learning_rate": 4.196852888372584e-06, - "loss": 0.1421, + "loss": 0.8868, "step": 74820 }, { "epoch": 7.9, - "grad_norm": 7.90922212600708, + "grad_norm": 47.1580810546875, "learning_rate": 4.194740732917943e-06, - "loss": 0.4665, + "loss": 0.918, "step": 74830 }, { "epoch": 7.9, - "grad_norm": 1.0630953311920166, + "grad_norm": 39.888397216796875, "learning_rate": 4.192628577463302e-06, - "loss": 0.4395, + "loss": 0.8171, "step": 74840 }, { "epoch": 7.9, - "grad_norm": 9.985666275024414, + "grad_norm": 38.58782196044922, "learning_rate": 4.1905164220086605e-06, - "loss": 0.4134, + "loss": 0.8671, "step": 74850 }, { "epoch": 7.91, - "grad_norm": 3.2522895336151123, + "grad_norm": 19.639026641845703, "learning_rate": 4.1884042665540184e-06, - "loss": 0.4438, + "loss": 0.8656, "step": 74860 }, { "epoch": 7.91, - "grad_norm": 30.838647842407227, + "grad_norm": 35.8124885559082, "learning_rate": 4.186292111099377e-06, - "loss": 0.5586, + "loss": 0.9318, "step": 74870 }, { "epoch": 7.91, - "grad_norm": 20.511730194091797, + "grad_norm": 34.33412551879883, "learning_rate": 4.184179955644736e-06, - "loss": 0.2192, + "loss": 0.6952, "step": 74880 }, { "epoch": 7.91, - "grad_norm": 37.29627990722656, + "grad_norm": 33.3859748840332, "learning_rate": 4.182067800190095e-06, - "loss": 0.4895, + "loss": 0.9759, "step": 74890 }, { "epoch": 7.91, - "grad_norm": 18.360708236694336, + "grad_norm": 33.736568450927734, "learning_rate": 4.179955644735453e-06, - "loss": 0.2081, + "loss": 0.8752, "step": 74900 }, { "epoch": 7.91, - "grad_norm": 0.9630420804023743, + "grad_norm": 55.95796585083008, "learning_rate": 4.1778434892808115e-06, - "loss": 0.5649, + "loss": 1.0165, "step": 74910 }, { "epoch": 7.91, - "grad_norm": 8.65583610534668, + "grad_norm": 41.57809066772461, "learning_rate": 4.17573133382617e-06, - "loss": 0.2009, + "loss": 0.9348, "step": 74920 }, { "epoch": 7.91, - "grad_norm": 11.775819778442383, + "grad_norm": 44.669944763183594, "learning_rate": 4.173619178371528e-06, - "loss": 0.3281, + "loss": 1.0276, "step": 74930 }, { "epoch": 7.91, - "grad_norm": 29.29155158996582, + "grad_norm": 28.98102378845215, "learning_rate": 4.171507022916887e-06, - "loss": 0.1729, + "loss": 0.7723, "step": 74940 }, { "epoch": 7.92, - "grad_norm": 20.289438247680664, + "grad_norm": 26.903886795043945, "learning_rate": 4.169394867462245e-06, - "loss": 0.1863, + "loss": 0.6652, "step": 74950 }, { "epoch": 7.92, - "grad_norm": 20.345863342285156, + "grad_norm": 30.042007446289062, "learning_rate": 4.1672827120076045e-06, - "loss": 0.4541, + "loss": 0.9399, "step": 74960 }, { "epoch": 7.92, - "grad_norm": 1.7319467067718506, + "grad_norm": 28.5837459564209, "learning_rate": 4.1651705565529625e-06, - "loss": 0.2585, + "loss": 0.7743, "step": 74970 }, { "epoch": 7.92, - "grad_norm": 4.5446577072143555, + "grad_norm": 25.373186111450195, "learning_rate": 4.163058401098321e-06, - "loss": 0.2211, + "loss": 0.8128, "step": 74980 }, { "epoch": 7.92, - "grad_norm": 0.24466173350811005, + "grad_norm": 49.61711120605469, "learning_rate": 4.160946245643679e-06, - "loss": 0.3594, + "loss": 0.9163, "step": 74990 }, { "epoch": 7.92, - "grad_norm": 52.71653747558594, + "grad_norm": 26.357421875, "learning_rate": 4.158834090189039e-06, - "loss": 0.2374, + "loss": 0.937, "step": 75000 }, { "epoch": 7.92, - "grad_norm": 1.3439245223999023, + "grad_norm": 23.586957931518555, "learning_rate": 4.156721934734397e-06, - "loss": 0.5069, + "loss": 0.9094, "step": 75010 }, { "epoch": 7.92, - "grad_norm": 9.5512056350708, + "grad_norm": 72.67071533203125, "learning_rate": 4.1546097792797555e-06, - "loss": 0.4071, + "loss": 1.0276, "step": 75020 }, { "epoch": 7.92, - "grad_norm": 0.05511058494448662, + "grad_norm": 36.287071228027344, "learning_rate": 4.152497623825114e-06, - "loss": 0.3313, + "loss": 0.9165, "step": 75030 }, { "epoch": 7.92, - "grad_norm": 2.0040149688720703, + "grad_norm": 58.72295379638672, "learning_rate": 4.150385468370472e-06, - "loss": 0.1111, + "loss": 1.3354, "step": 75040 }, { "epoch": 7.93, - "grad_norm": 16.1384220123291, + "grad_norm": 39.65127944946289, "learning_rate": 4.148273312915831e-06, - "loss": 0.2661, + "loss": 0.6189, "step": 75050 }, { "epoch": 7.93, - "grad_norm": 8.519018173217773, + "grad_norm": 31.340774536132812, "learning_rate": 4.146161157461189e-06, - "loss": 0.381, + "loss": 0.6421, "step": 75060 }, { "epoch": 7.93, - "grad_norm": 15.408010482788086, + "grad_norm": 46.03690719604492, "learning_rate": 4.1440490020065486e-06, - "loss": 0.1965, + "loss": 0.6709, "step": 75070 }, { "epoch": 7.93, - "grad_norm": 28.1136531829834, + "grad_norm": 15.376952171325684, "learning_rate": 4.1419368465519065e-06, - "loss": 0.483, + "loss": 0.7178, "step": 75080 }, { "epoch": 7.93, - "grad_norm": 23.167322158813477, + "grad_norm": 35.15133285522461, "learning_rate": 4.139824691097265e-06, - "loss": 0.3506, + "loss": 0.7205, "step": 75090 }, { "epoch": 7.93, - "grad_norm": 0.02125697024166584, + "grad_norm": 43.39228057861328, "learning_rate": 4.137712535642623e-06, - "loss": 0.4713, + "loss": 0.7809, "step": 75100 }, { "epoch": 7.93, - "grad_norm": 9.977216720581055, + "grad_norm": 29.18212127685547, "learning_rate": 4.135600380187982e-06, - "loss": 0.3219, + "loss": 0.7314, "step": 75110 }, { "epoch": 7.93, - "grad_norm": 0.23485343158245087, + "grad_norm": 33.38829803466797, "learning_rate": 4.133488224733341e-06, - "loss": 0.504, + "loss": 0.7426, "step": 75120 }, { "epoch": 7.93, - "grad_norm": 15.525080680847168, + "grad_norm": 63.1003303527832, "learning_rate": 4.1313760692786995e-06, - "loss": 0.608, + "loss": 1.3301, "step": 75130 }, { "epoch": 7.94, - "grad_norm": 18.554777145385742, + "grad_norm": 47.91437530517578, "learning_rate": 4.1292639138240575e-06, - "loss": 0.4232, + "loss": 0.7407, "step": 75140 }, { "epoch": 7.94, - "grad_norm": 14.08617877960205, + "grad_norm": 41.12310028076172, "learning_rate": 4.127151758369416e-06, - "loss": 0.5316, + "loss": 1.0058, "step": 75150 }, { "epoch": 7.94, - "grad_norm": 1.7554539442062378, + "grad_norm": 49.81254959106445, "learning_rate": 4.125039602914775e-06, - "loss": 0.2984, + "loss": 1.098, "step": 75160 }, { "epoch": 7.94, - "grad_norm": 25.056318283081055, + "grad_norm": 41.45735168457031, "learning_rate": 4.122927447460133e-06, - "loss": 0.6825, + "loss": 1.0496, "step": 75170 }, { "epoch": 7.94, - "grad_norm": 54.60165023803711, + "grad_norm": 33.981101989746094, "learning_rate": 4.120815292005492e-06, - "loss": 0.3031, + "loss": 0.9152, "step": 75180 }, { "epoch": 7.94, - "grad_norm": 1.3052012920379639, + "grad_norm": 52.21247863769531, "learning_rate": 4.1187031365508505e-06, - "loss": 0.3167, + "loss": 1.0537, "step": 75190 }, { "epoch": 7.94, - "grad_norm": 0.49126970767974854, + "grad_norm": 27.709928512573242, "learning_rate": 4.116590981096209e-06, - "loss": 0.2493, + "loss": 0.8953, "step": 75200 }, { "epoch": 7.94, - "grad_norm": 13.616484642028809, + "grad_norm": 37.657958984375, "learning_rate": 4.114478825641567e-06, - "loss": 0.5762, + "loss": 0.8055, "step": 75210 }, { "epoch": 7.94, - "grad_norm": 9.636366844177246, + "grad_norm": 49.08051681518555, "learning_rate": 4.112366670186926e-06, - "loss": 0.3682, + "loss": 1.0075, "step": 75220 }, { "epoch": 7.94, - "grad_norm": 27.629575729370117, + "grad_norm": 29.605487823486328, "learning_rate": 4.110254514732285e-06, - "loss": 0.4129, + "loss": 0.8567, "step": 75230 }, { "epoch": 7.95, - "grad_norm": 29.822162628173828, + "grad_norm": 49.63994216918945, "learning_rate": 4.1081423592776436e-06, - "loss": 0.1637, + "loss": 0.8981, "step": 75240 }, { "epoch": 7.95, - "grad_norm": 12.628683090209961, + "grad_norm": 46.16111755371094, "learning_rate": 4.1060302038230015e-06, - "loss": 0.6868, + "loss": 1.0246, "step": 75250 }, { "epoch": 7.95, - "grad_norm": 0.15593115985393524, + "grad_norm": 21.805559158325195, "learning_rate": 4.10391804836836e-06, - "loss": 0.3273, + "loss": 0.9571, "step": 75260 }, { "epoch": 7.95, - "grad_norm": 0.6657549142837524, + "grad_norm": 38.030181884765625, "learning_rate": 4.101805892913719e-06, - "loss": 0.3055, + "loss": 1.0417, "step": 75270 }, { "epoch": 7.95, - "grad_norm": 97.75869750976562, + "grad_norm": 33.23841857910156, "learning_rate": 4.099693737459077e-06, - "loss": 0.3968, + "loss": 0.5456, "step": 75280 }, { "epoch": 7.95, - "grad_norm": 11.188749313354492, + "grad_norm": 37.732421875, "learning_rate": 4.097581582004436e-06, - "loss": 0.1773, + "loss": 0.703, "step": 75290 }, { "epoch": 7.95, - "grad_norm": 18.634977340698242, + "grad_norm": 53.65509033203125, "learning_rate": 4.0954694265497945e-06, - "loss": 0.2712, + "loss": 1.2626, "step": 75300 }, { "epoch": 7.95, - "grad_norm": 2.773984909057617, + "grad_norm": 34.67169189453125, "learning_rate": 4.093357271095153e-06, - "loss": 0.5178, + "loss": 1.0658, "step": 75310 }, { "epoch": 7.95, - "grad_norm": 12.019537925720215, + "grad_norm": 36.96458053588867, "learning_rate": 4.091245115640511e-06, - "loss": 0.2325, + "loss": 0.8396, "step": 75320 }, { "epoch": 7.96, - "grad_norm": 11.386368751525879, + "grad_norm": 13.299067497253418, "learning_rate": 4.08913296018587e-06, - "loss": 0.3468, + "loss": 0.682, "step": 75330 }, { "epoch": 7.96, - "grad_norm": 9.02527141571045, + "grad_norm": 32.81097412109375, "learning_rate": 4.087020804731229e-06, - "loss": 0.4886, + "loss": 0.8271, "step": 75340 }, { "epoch": 7.96, - "grad_norm": 15.525890350341797, + "grad_norm": 59.559043884277344, "learning_rate": 4.084908649276587e-06, - "loss": 0.6869, + "loss": 1.2519, "step": 75350 }, { "epoch": 7.96, - "grad_norm": 23.637107849121094, + "grad_norm": 38.5053596496582, "learning_rate": 4.0827964938219455e-06, - "loss": 0.3483, + "loss": 1.1782, "step": 75360 }, { "epoch": 7.96, - "grad_norm": 0.16287614405155182, + "grad_norm": 22.87129020690918, "learning_rate": 4.080684338367304e-06, - "loss": 0.2055, + "loss": 0.8043, "step": 75370 }, { "epoch": 7.96, - "grad_norm": 30.3100528717041, + "grad_norm": 60.19916915893555, "learning_rate": 4.078572182912663e-06, - "loss": 0.5211, + "loss": 0.9617, "step": 75380 }, { "epoch": 7.96, - "grad_norm": 19.44643211364746, + "grad_norm": 54.11901092529297, "learning_rate": 4.076460027458021e-06, - "loss": 0.1961, + "loss": 0.8595, "step": 75390 }, { "epoch": 7.96, - "grad_norm": 12.139644622802734, + "grad_norm": 19.716869354248047, "learning_rate": 4.07434787200338e-06, - "loss": 0.3583, + "loss": 0.9765, "step": 75400 }, { "epoch": 7.96, - "grad_norm": 14.978793144226074, + "grad_norm": 37.74128341674805, "learning_rate": 4.072235716548738e-06, - "loss": 0.437, + "loss": 0.8074, "step": 75410 }, { "epoch": 7.96, - "grad_norm": 25.219120025634766, + "grad_norm": 60.28083038330078, "learning_rate": 4.070123561094097e-06, - "loss": 0.4557, + "loss": 0.6669, "step": 75420 }, { "epoch": 7.97, - "grad_norm": 14.67039680480957, + "grad_norm": 52.81187057495117, "learning_rate": 4.068011405639455e-06, - "loss": 0.4828, + "loss": 1.0653, "step": 75430 }, { "epoch": 7.97, - "grad_norm": 16.331632614135742, + "grad_norm": 51.17252731323242, "learning_rate": 4.065899250184814e-06, - "loss": 0.2613, + "loss": 1.1982, "step": 75440 }, { "epoch": 7.97, - "grad_norm": 0.5331940650939941, + "grad_norm": 43.74235153198242, "learning_rate": 4.063787094730173e-06, - "loss": 0.5499, + "loss": 0.9679, "step": 75450 }, { "epoch": 7.97, - "grad_norm": 42.312744140625, + "grad_norm": 46.59767532348633, "learning_rate": 4.061674939275531e-06, - "loss": 0.524, + "loss": 0.8549, "step": 75460 }, { "epoch": 7.97, - "grad_norm": 1.064154863357544, + "grad_norm": 13.104887962341309, "learning_rate": 4.0595627838208895e-06, - "loss": 0.261, + "loss": 1.1067, "step": 75470 }, { "epoch": 7.97, - "grad_norm": 5.593687534332275, + "grad_norm": 40.904354095458984, "learning_rate": 4.057450628366248e-06, - "loss": 0.4614, + "loss": 0.8053, "step": 75480 }, { "epoch": 7.97, - "grad_norm": 11.8132905960083, + "grad_norm": 39.87453079223633, "learning_rate": 4.055338472911607e-06, - "loss": 0.1851, + "loss": 0.7331, "step": 75490 }, { "epoch": 7.97, - "grad_norm": 4.696486949920654, + "grad_norm": 24.96514129638672, "learning_rate": 4.053226317456965e-06, - "loss": 0.6076, + "loss": 0.5879, "step": 75500 }, { "epoch": 7.97, - "grad_norm": 18.952587127685547, + "grad_norm": 38.0311164855957, "learning_rate": 4.051114162002324e-06, - "loss": 0.6673, + "loss": 1.2454, "step": 75510 }, { "epoch": 7.98, - "grad_norm": 0.15674082934856415, + "grad_norm": 48.84183120727539, "learning_rate": 4.049002006547682e-06, - "loss": 0.2469, + "loss": 0.9674, "step": 75520 }, { "epoch": 7.98, - "grad_norm": 15.579913139343262, + "grad_norm": 30.845300674438477, "learning_rate": 4.046889851093041e-06, - "loss": 0.3902, + "loss": 0.9901, "step": 75530 }, { "epoch": 7.98, - "grad_norm": 33.22214126586914, + "grad_norm": 52.61882781982422, "learning_rate": 4.044777695638399e-06, - "loss": 0.5904, + "loss": 0.9571, "step": 75540 }, { "epoch": 7.98, - "grad_norm": 16.658628463745117, + "grad_norm": 45.7989387512207, "learning_rate": 4.042665540183758e-06, - "loss": 0.5794, + "loss": 1.1787, "step": 75550 }, { "epoch": 7.98, - "grad_norm": 50.82689666748047, + "grad_norm": 32.36574935913086, "learning_rate": 4.040553384729116e-06, - "loss": 0.3253, + "loss": 0.7854, "step": 75560 }, { "epoch": 7.98, - "grad_norm": 38.84156799316406, + "grad_norm": 40.4644889831543, "learning_rate": 4.038441229274475e-06, - "loss": 0.3972, + "loss": 0.5795, "step": 75570 }, { "epoch": 7.98, - "grad_norm": 13.07491683959961, + "grad_norm": 23.32135581970215, "learning_rate": 4.0363290738198336e-06, - "loss": 0.4447, + "loss": 0.8258, "step": 75580 }, { "epoch": 7.98, - "grad_norm": 4.281503200531006, + "grad_norm": 37.804832458496094, "learning_rate": 4.0342169183651915e-06, - "loss": 0.0999, + "loss": 0.9758, "step": 75590 }, { "epoch": 7.98, - "grad_norm": 5.939892292022705, + "grad_norm": 45.48539733886719, "learning_rate": 4.03210476291055e-06, - "loss": 0.201, + "loss": 0.8175, "step": 75600 }, { "epoch": 7.99, - "grad_norm": 16.133319854736328, + "grad_norm": 24.910642623901367, "learning_rate": 4.029992607455909e-06, - "loss": 0.5385, + "loss": 0.9977, "step": 75610 }, { "epoch": 7.99, - "grad_norm": 10.509668350219727, + "grad_norm": 49.842769622802734, "learning_rate": 4.027880452001268e-06, - "loss": 0.28, + "loss": 0.8136, "step": 75620 }, { "epoch": 7.99, - "grad_norm": 1.3435183763504028, + "grad_norm": 40.00590515136719, "learning_rate": 4.025768296546626e-06, - "loss": 0.4916, + "loss": 1.0046, "step": 75630 }, { "epoch": 7.99, - "grad_norm": 32.471412658691406, + "grad_norm": 43.2944221496582, "learning_rate": 4.0236561410919845e-06, - "loss": 0.3388, + "loss": 0.8687, "step": 75640 }, { "epoch": 7.99, - "grad_norm": 18.319494247436523, + "grad_norm": 33.04887008666992, "learning_rate": 4.021543985637343e-06, - "loss": 0.5707, + "loss": 1.1553, "step": 75650 }, { "epoch": 7.99, - "grad_norm": 0.9679093956947327, + "grad_norm": 34.5339241027832, "learning_rate": 4.019431830182702e-06, - "loss": 0.1428, + "loss": 0.9019, "step": 75660 }, { "epoch": 7.99, - "grad_norm": 30.683849334716797, + "grad_norm": 64.03108215332031, "learning_rate": 4.01731967472806e-06, - "loss": 0.522, + "loss": 1.0318, "step": 75670 }, { "epoch": 7.99, - "grad_norm": 1.2637443542480469, + "grad_norm": 23.93154525756836, "learning_rate": 4.015207519273419e-06, - "loss": 0.2549, + "loss": 0.6882, "step": 75680 }, { "epoch": 7.99, - "grad_norm": 12.575556755065918, + "grad_norm": 12.200983047485352, "learning_rate": 4.013095363818778e-06, - "loss": 0.4542, + "loss": 0.8115, "step": 75690 }, { "epoch": 7.99, - "grad_norm": 12.55077075958252, + "grad_norm": 46.67206954956055, "learning_rate": 4.0109832083641355e-06, - "loss": 0.5628, + "loss": 1.2104, "step": 75700 }, { "epoch": 8.0, - "grad_norm": 11.64674186706543, + "grad_norm": 57.80583953857422, "learning_rate": 4.008871052909494e-06, - "loss": 0.1788, + "loss": 0.9073, "step": 75710 }, { "epoch": 8.0, - "grad_norm": 17.065706253051758, + "grad_norm": 23.582008361816406, "learning_rate": 4.006758897454853e-06, - "loss": 0.2626, + "loss": 1.0043, "step": 75720 }, { "epoch": 8.0, - "grad_norm": 1.4481414556503296, + "grad_norm": 43.052886962890625, "learning_rate": 4.004646742000212e-06, - "loss": 0.0908, + "loss": 0.5709, "step": 75730 }, { "epoch": 8.0, - "grad_norm": 2.6940932273864746, + "grad_norm": 39.98506164550781, "learning_rate": 4.00253458654557e-06, - "loss": 0.3565, + "loss": 1.0053, "step": 75740 }, { "epoch": 8.0, - "grad_norm": 1.4701557159423828, + "grad_norm": 26.02156639099121, "learning_rate": 4.0004224310909286e-06, - "loss": 0.552, + "loss": 0.8969, "step": 75750 }, { "epoch": 8.0, - "eval_accuracy": 0.9053069306930693, - "eval_loss": 0.3915071487426758, - "eval_runtime": 150.4006, - "eval_samples_per_second": 167.885, - "eval_steps_per_second": 20.991, + "eval_accuracy": 0.8116831683168317, + "eval_loss": 0.6865377426147461, + "eval_runtime": 37.1519, + "eval_samples_per_second": 679.643, + "eval_steps_per_second": 84.976, "step": 75752 }, { "epoch": 8.0, - "grad_norm": 19.41884422302246, + "grad_norm": 51.396697998046875, "learning_rate": 3.998310275636287e-06, - "loss": 0.3003, + "loss": 0.8005, "step": 75760 }, { "epoch": 8.0, - "grad_norm": 0.31812191009521484, + "grad_norm": 46.96210479736328, "learning_rate": 3.996198120181646e-06, - "loss": 0.2754, + "loss": 0.9083, "step": 75770 }, { "epoch": 8.0, - "grad_norm": 50.16751480102539, + "grad_norm": 62.57152557373047, "learning_rate": 3.994085964727004e-06, - "loss": 0.2973, + "loss": 0.881, "step": 75780 }, { "epoch": 8.0, - "grad_norm": 17.94882583618164, + "grad_norm": 50.014427185058594, "learning_rate": 3.991973809272363e-06, - "loss": 0.4323, + "loss": 0.89, "step": 75790 }, { "epoch": 8.01, - "grad_norm": 10.917844772338867, + "grad_norm": 28.54424285888672, "learning_rate": 3.989861653817722e-06, - "loss": 0.3077, + "loss": 0.9436, "step": 75800 }, { "epoch": 8.01, - "grad_norm": 13.581189155578613, + "grad_norm": 47.13255310058594, "learning_rate": 3.9877494983630796e-06, - "loss": 0.6535, + "loss": 1.0303, "step": 75810 }, { "epoch": 8.01, - "grad_norm": 12.885064125061035, + "grad_norm": 50.47044372558594, "learning_rate": 3.985637342908438e-06, - "loss": 0.2511, + "loss": 0.6868, "step": 75820 }, { "epoch": 8.01, - "grad_norm": 8.589248657226562, + "grad_norm": 60.15666961669922, "learning_rate": 3.983525187453796e-06, - "loss": 0.4339, + "loss": 1.3567, "step": 75830 }, { "epoch": 8.01, - "grad_norm": 15.38491153717041, + "grad_norm": 45.45216751098633, "learning_rate": 3.981413031999156e-06, - "loss": 0.2202, + "loss": 0.7539, "step": 75840 }, { "epoch": 8.01, - "grad_norm": 0.2755224108695984, + "grad_norm": 45.504913330078125, "learning_rate": 3.979300876544514e-06, - "loss": 0.3022, + "loss": 0.672, "step": 75850 }, { "epoch": 8.01, - "grad_norm": 9.498724937438965, + "grad_norm": 34.71426010131836, "learning_rate": 3.977188721089873e-06, - "loss": 0.2453, + "loss": 0.8554, "step": 75860 }, { "epoch": 8.01, - "grad_norm": 24.352373123168945, + "grad_norm": 29.127399444580078, "learning_rate": 3.9750765656352305e-06, - "loss": 0.4338, + "loss": 0.8327, "step": 75870 }, { "epoch": 8.01, - "grad_norm": 23.96543312072754, + "grad_norm": 14.885248184204102, "learning_rate": 3.97296441018059e-06, - "loss": 0.3526, + "loss": 0.7849, "step": 75880 }, { "epoch": 8.01, - "grad_norm": 17.79979133605957, + "grad_norm": 34.91730880737305, "learning_rate": 3.970852254725948e-06, - "loss": 0.1288, + "loss": 0.8851, "step": 75890 }, { "epoch": 8.02, - "grad_norm": 25.800020217895508, + "grad_norm": 30.21282958984375, "learning_rate": 3.968740099271307e-06, - "loss": 0.4411, + "loss": 1.1577, "step": 75900 }, { "epoch": 8.02, - "grad_norm": 0.22764214873313904, + "grad_norm": 45.73076248168945, "learning_rate": 3.966627943816666e-06, - "loss": 0.1851, + "loss": 1.0559, "step": 75910 }, { "epoch": 8.02, - "grad_norm": 21.472002029418945, + "grad_norm": 47.629364013671875, "learning_rate": 3.964515788362024e-06, - "loss": 0.6507, + "loss": 0.96, "step": 75920 }, { "epoch": 8.02, - "grad_norm": 26.212228775024414, + "grad_norm": 39.75006866455078, "learning_rate": 3.962403632907382e-06, - "loss": 0.3942, + "loss": 1.0243, "step": 75930 }, { "epoch": 8.02, - "grad_norm": 27.422698974609375, + "grad_norm": 45.664459228515625, "learning_rate": 3.96029147745274e-06, - "loss": 0.5619, + "loss": 0.9953, "step": 75940 }, { "epoch": 8.02, - "grad_norm": 19.20816993713379, + "grad_norm": 46.092952728271484, "learning_rate": 3.9581793219981e-06, - "loss": 0.3696, + "loss": 0.6485, "step": 75950 }, { "epoch": 8.02, - "grad_norm": 6.515803337097168, + "grad_norm": 56.05353546142578, "learning_rate": 3.956067166543458e-06, - "loss": 0.5102, + "loss": 0.8571, "step": 75960 }, { "epoch": 8.02, - "grad_norm": 10.355790138244629, + "grad_norm": 51.39031219482422, "learning_rate": 3.953955011088817e-06, - "loss": 0.3982, + "loss": 0.7997, "step": 75970 }, { "epoch": 8.02, - "grad_norm": 13.501233100891113, + "grad_norm": 35.56836700439453, "learning_rate": 3.9518428556341746e-06, - "loss": 0.2437, + "loss": 0.9232, "step": 75980 }, { "epoch": 8.03, - "grad_norm": 0.18299558758735657, + "grad_norm": 29.06903076171875, "learning_rate": 3.949730700179533e-06, - "loss": 0.491, + "loss": 0.8742, "step": 75990 }, { "epoch": 8.03, - "grad_norm": 3.602692127227783, + "grad_norm": 58.81188201904297, "learning_rate": 3.947618544724892e-06, - "loss": 0.295, + "loss": 0.7305, "step": 76000 }, { "epoch": 8.03, - "grad_norm": 12.329545974731445, + "grad_norm": 36.203311920166016, "learning_rate": 3.945506389270251e-06, - "loss": 0.5658, + "loss": 0.8218, "step": 76010 }, { "epoch": 8.03, - "grad_norm": 23.870826721191406, + "grad_norm": 42.42064666748047, "learning_rate": 3.943394233815609e-06, - "loss": 0.6159, + "loss": 1.1694, "step": 76020 }, { "epoch": 8.03, - "grad_norm": 11.138745307922363, + "grad_norm": 65.83843994140625, "learning_rate": 3.941282078360968e-06, - "loss": 0.282, + "loss": 1.0375, "step": 76030 }, { "epoch": 8.03, - "grad_norm": 5.704168319702148, + "grad_norm": 41.90058898925781, "learning_rate": 3.939169922906326e-06, - "loss": 0.5346, + "loss": 1.1503, "step": 76040 }, { "epoch": 8.03, - "grad_norm": 22.808225631713867, + "grad_norm": 40.381980895996094, "learning_rate": 3.937057767451684e-06, - "loss": 0.4845, + "loss": 0.7445, "step": 76050 }, { "epoch": 8.03, - "grad_norm": 0.06041138991713524, + "grad_norm": 38.04759216308594, "learning_rate": 3.934945611997043e-06, - "loss": 0.1315, + "loss": 0.7354, "step": 76060 }, { "epoch": 8.03, - "grad_norm": 6.407318115234375, + "grad_norm": 31.043289184570312, "learning_rate": 3.932833456542402e-06, - "loss": 0.2276, + "loss": 0.8558, "step": 76070 }, { "epoch": 8.03, - "grad_norm": 0.018613610416650772, + "grad_norm": 19.94843101501465, "learning_rate": 3.930721301087761e-06, - "loss": 0.2255, + "loss": 0.5349, "step": 76080 }, { "epoch": 8.04, - "grad_norm": 0.26306670904159546, + "grad_norm": 21.123830795288086, "learning_rate": 3.928609145633119e-06, - "loss": 0.2243, + "loss": 0.7957, "step": 76090 }, { "epoch": 8.04, - "grad_norm": 13.95859146118164, + "grad_norm": 33.508209228515625, "learning_rate": 3.926496990178477e-06, - "loss": 0.3259, + "loss": 0.8346, "step": 76100 }, { "epoch": 8.04, - "grad_norm": 8.395294189453125, + "grad_norm": 52.80196762084961, "learning_rate": 3.924384834723836e-06, - "loss": 0.2518, + "loss": 0.6413, "step": 76110 }, { "epoch": 8.04, - "grad_norm": 1.194173812866211, + "grad_norm": 57.0904655456543, "learning_rate": 3.922272679269195e-06, - "loss": 0.3378, + "loss": 0.8384, "step": 76120 }, { "epoch": 8.04, - "grad_norm": 12.38266372680664, + "grad_norm": 46.486576080322266, "learning_rate": 3.920160523814553e-06, - "loss": 0.3462, + "loss": 0.8597, "step": 76130 }, { "epoch": 8.04, - "grad_norm": 36.54191970825195, + "grad_norm": 39.407257080078125, "learning_rate": 3.918048368359912e-06, - "loss": 0.3439, + "loss": 0.6891, "step": 76140 }, { "epoch": 8.04, - "grad_norm": 45.1425666809082, + "grad_norm": 36.59585952758789, "learning_rate": 3.91593621290527e-06, - "loss": 0.389, + "loss": 0.6251, "step": 76150 }, { "epoch": 8.04, - "grad_norm": 49.564453125, + "grad_norm": 34.52202224731445, "learning_rate": 3.913824057450628e-06, - "loss": 0.2848, + "loss": 0.765, "step": 76160 }, { "epoch": 8.04, - "grad_norm": 15.484761238098145, + "grad_norm": 44.44729995727539, "learning_rate": 3.911711901995987e-06, - "loss": 0.3501, + "loss": 0.7134, "step": 76170 }, { "epoch": 8.05, - "grad_norm": 43.650428771972656, + "grad_norm": 17.29265785217285, "learning_rate": 3.909599746541346e-06, - "loss": 0.267, + "loss": 0.7753, "step": 76180 }, { "epoch": 8.05, - "grad_norm": 10.316177368164062, + "grad_norm": 36.05371856689453, "learning_rate": 3.907487591086705e-06, - "loss": 0.265, + "loss": 0.5571, "step": 76190 }, { "epoch": 8.05, - "grad_norm": 35.15428161621094, + "grad_norm": 43.255470275878906, "learning_rate": 3.905375435632063e-06, - "loss": 0.2324, + "loss": 0.9133, "step": 76200 }, { "epoch": 8.05, - "grad_norm": 14.188833236694336, + "grad_norm": 55.63020706176758, "learning_rate": 3.903263280177421e-06, - "loss": 0.3411, + "loss": 1.0659, "step": 76210 }, { "epoch": 8.05, - "grad_norm": 30.591930389404297, + "grad_norm": 51.31069564819336, "learning_rate": 3.90115112472278e-06, - "loss": 0.4437, + "loss": 0.8964, "step": 76220 }, { "epoch": 8.05, - "grad_norm": 1.032712459564209, + "grad_norm": 48.828582763671875, "learning_rate": 3.899038969268138e-06, - "loss": 0.2608, + "loss": 0.5291, "step": 76230 }, { "epoch": 8.05, - "grad_norm": 23.076974868774414, + "grad_norm": 49.043697357177734, "learning_rate": 3.896926813813497e-06, - "loss": 0.3072, + "loss": 0.7035, "step": 76240 }, { "epoch": 8.05, - "grad_norm": 29.829870223999023, + "grad_norm": 44.91276931762695, "learning_rate": 3.894814658358856e-06, - "loss": 0.2784, + "loss": 0.6354, "step": 76250 }, { "epoch": 8.05, - "grad_norm": 0.18127800524234772, + "grad_norm": 38.010066986083984, "learning_rate": 3.8927025029042144e-06, - "loss": 0.2574, + "loss": 0.7511, "step": 76260 }, { "epoch": 8.05, - "grad_norm": 3.5314133167266846, + "grad_norm": 34.418670654296875, "learning_rate": 3.890590347449572e-06, - "loss": 0.1453, + "loss": 0.6907, "step": 76270 }, { "epoch": 8.06, - "grad_norm": 18.701297760009766, + "grad_norm": 60.452354431152344, "learning_rate": 3.888478191994931e-06, - "loss": 0.3782, + "loss": 0.753, "step": 76280 }, { "epoch": 8.06, - "grad_norm": 0.1347758024930954, + "grad_norm": 25.211578369140625, "learning_rate": 3.886366036540289e-06, - "loss": 0.2595, + "loss": 0.8414, "step": 76290 }, { "epoch": 8.06, - "grad_norm": 0.3699549734592438, + "grad_norm": 22.200902938842773, "learning_rate": 3.884253881085649e-06, - "loss": 0.199, + "loss": 0.6859, "step": 76300 }, { "epoch": 8.06, - "grad_norm": 16.021669387817383, + "grad_norm": 35.46787643432617, "learning_rate": 3.882141725631007e-06, - "loss": 0.4484, + "loss": 1.3394, "step": 76310 }, { "epoch": 8.06, - "grad_norm": 16.956850051879883, + "grad_norm": 34.31077194213867, "learning_rate": 3.880029570176365e-06, - "loss": 0.2623, + "loss": 0.8159, "step": 76320 }, { "epoch": 8.06, - "grad_norm": 6.270880222320557, + "grad_norm": 54.23385238647461, "learning_rate": 3.877917414721723e-06, - "loss": 0.2061, + "loss": 1.0343, "step": 76330 }, { "epoch": 8.06, - "grad_norm": 21.483959197998047, + "grad_norm": 55.655609130859375, "learning_rate": 3.875805259267082e-06, - "loss": 0.2837, + "loss": 0.7789, "step": 76340 }, { "epoch": 8.06, - "grad_norm": 15.500407218933105, + "grad_norm": 15.139958381652832, "learning_rate": 3.873693103812441e-06, - "loss": 0.4021, + "loss": 0.7758, "step": 76350 }, { "epoch": 8.06, - "grad_norm": 7.735941410064697, + "grad_norm": 26.80805206298828, "learning_rate": 3.8715809483578e-06, - "loss": 0.3957, + "loss": 0.8493, "step": 76360 }, { "epoch": 8.07, - "grad_norm": 0.08314790576696396, + "grad_norm": 28.271446228027344, "learning_rate": 3.8694687929031585e-06, - "loss": 0.2107, + "loss": 0.8013, "step": 76370 }, { "epoch": 8.07, - "grad_norm": 10.381183624267578, + "grad_norm": 31.331430435180664, "learning_rate": 3.867356637448516e-06, - "loss": 0.2465, + "loss": 0.8526, "step": 76380 }, { "epoch": 8.07, - "grad_norm": 0.11781269311904907, + "grad_norm": 69.7100830078125, "learning_rate": 3.865244481993875e-06, - "loss": 0.474, + "loss": 1.069, "step": 76390 }, { "epoch": 8.07, - "grad_norm": 18.931489944458008, + "grad_norm": 47.345638275146484, "learning_rate": 3.863132326539233e-06, - "loss": 0.503, + "loss": 0.8419, "step": 76400 }, { "epoch": 8.07, - "grad_norm": 10.344649314880371, + "grad_norm": 49.873077392578125, "learning_rate": 3.861020171084593e-06, - "loss": 0.3966, + "loss": 0.882, "step": 76410 }, { "epoch": 8.07, - "grad_norm": 10.028229713439941, + "grad_norm": 42.47067642211914, "learning_rate": 3.858908015629951e-06, - "loss": 0.4482, + "loss": 0.9438, "step": 76420 }, { "epoch": 8.07, - "grad_norm": 0.7364844083786011, + "grad_norm": 16.2875919342041, "learning_rate": 3.8567958601753094e-06, - "loss": 0.1594, + "loss": 0.8907, "step": 76430 }, { "epoch": 8.07, - "grad_norm": 2.9318110942840576, + "grad_norm": 41.664066314697266, "learning_rate": 3.854683704720667e-06, - "loss": 0.374, + "loss": 0.652, "step": 76440 }, { "epoch": 8.07, - "grad_norm": 14.525618553161621, + "grad_norm": 38.44405746459961, "learning_rate": 3.852571549266026e-06, - "loss": 0.4622, + "loss": 0.9773, "step": 76450 }, { "epoch": 8.07, - "grad_norm": 16.010263442993164, + "grad_norm": 50.1849365234375, "learning_rate": 3.850459393811385e-06, - "loss": 0.1693, + "loss": 1.0305, "step": 76460 }, { "epoch": 8.08, - "grad_norm": 10.948208808898926, + "grad_norm": 41.472389221191406, "learning_rate": 3.848347238356744e-06, - "loss": 0.4448, + "loss": 0.8766, "step": 76470 }, { "epoch": 8.08, - "grad_norm": 7.51269006729126, + "grad_norm": 35.504295349121094, "learning_rate": 3.846235082902102e-06, - "loss": 0.3348, + "loss": 0.6644, "step": 76480 }, { "epoch": 8.08, - "grad_norm": 16.29891586303711, + "grad_norm": 53.61573028564453, "learning_rate": 3.84412292744746e-06, - "loss": 0.2476, + "loss": 0.751, "step": 76490 }, { "epoch": 8.08, - "grad_norm": 0.29972410202026367, + "grad_norm": 56.46197509765625, "learning_rate": 3.842010771992819e-06, - "loss": 0.5719, + "loss": 0.9028, "step": 76500 }, { "epoch": 8.08, - "grad_norm": 11.591474533081055, + "grad_norm": 49.0990104675293, "learning_rate": 3.839898616538177e-06, - "loss": 0.3811, + "loss": 0.7691, "step": 76510 }, { "epoch": 8.08, - "grad_norm": 2.5800235271453857, + "grad_norm": 78.4098892211914, "learning_rate": 3.837786461083536e-06, - "loss": 0.2486, + "loss": 0.9928, "step": 76520 }, { "epoch": 8.08, - "grad_norm": 24.484458923339844, + "grad_norm": 30.107892990112305, "learning_rate": 3.835674305628895e-06, - "loss": 0.4029, + "loss": 0.7588, "step": 76530 }, { "epoch": 8.08, - "grad_norm": 8.82453727722168, + "grad_norm": 31.75087547302246, "learning_rate": 3.8335621501742535e-06, - "loss": 0.524, + "loss": 0.8378, "step": 76540 }, { "epoch": 8.08, - "grad_norm": 12.238736152648926, + "grad_norm": 44.24927520751953, "learning_rate": 3.831449994719611e-06, - "loss": 0.406, + "loss": 0.959, "step": 76550 }, { "epoch": 8.09, - "grad_norm": 0.6767379641532898, + "grad_norm": 32.86566925048828, "learning_rate": 3.82933783926497e-06, - "loss": 0.1497, + "loss": 0.7186, "step": 76560 }, { "epoch": 8.09, - "grad_norm": 12.917670249938965, + "grad_norm": 38.55571746826172, "learning_rate": 3.827225683810329e-06, - "loss": 0.1908, + "loss": 0.5718, "step": 76570 }, { "epoch": 8.09, - "grad_norm": 3.5133163928985596, + "grad_norm": 22.9238338470459, "learning_rate": 3.825113528355687e-06, - "loss": 0.3823, + "loss": 0.9195, "step": 76580 }, { "epoch": 8.09, - "grad_norm": 2.1936402320861816, + "grad_norm": 37.02349853515625, "learning_rate": 3.823001372901046e-06, - "loss": 0.2487, + "loss": 0.9221, "step": 76590 }, { "epoch": 8.09, - "grad_norm": 1.8553451299667358, + "grad_norm": 17.908910751342773, "learning_rate": 3.8208892174464044e-06, - "loss": 0.2422, + "loss": 0.9447, "step": 76600 }, { "epoch": 8.09, - "grad_norm": 14.03058910369873, + "grad_norm": 34.095542907714844, "learning_rate": 3.818777061991763e-06, - "loss": 0.3505, + "loss": 0.9403, "step": 76610 }, { "epoch": 8.09, - "grad_norm": 3.647488594055176, + "grad_norm": 28.195518493652344, "learning_rate": 3.816664906537121e-06, - "loss": 0.144, + "loss": 0.6191, "step": 76620 }, { "epoch": 8.09, - "grad_norm": 20.479955673217773, + "grad_norm": 74.2109146118164, "learning_rate": 3.81455275108248e-06, - "loss": 0.3549, + "loss": 0.768, "step": 76630 }, { "epoch": 8.09, - "grad_norm": 9.187689781188965, + "grad_norm": 49.36863708496094, "learning_rate": 3.8124405956278387e-06, - "loss": 0.1189, + "loss": 1.1012, "step": 76640 }, { "epoch": 8.09, - "grad_norm": 30.256576538085938, + "grad_norm": 47.46691131591797, "learning_rate": 3.810328440173197e-06, - "loss": 0.6372, + "loss": 1.2449, "step": 76650 }, { "epoch": 8.1, - "grad_norm": 13.441614151000977, + "grad_norm": 45.24867630004883, "learning_rate": 3.8082162847185554e-06, - "loss": 0.4621, + "loss": 1.1876, "step": 76660 }, { "epoch": 8.1, - "grad_norm": 1.0047491788864136, + "grad_norm": 40.804195404052734, "learning_rate": 3.8061041292639138e-06, - "loss": 0.1899, + "loss": 0.956, "step": 76670 }, { "epoch": 8.1, - "grad_norm": 25.448091506958008, + "grad_norm": 54.419334411621094, "learning_rate": 3.803991973809273e-06, - "loss": 0.3331, + "loss": 0.7771, "step": 76680 }, { "epoch": 8.1, - "grad_norm": 7.318388938903809, + "grad_norm": 61.133358001708984, "learning_rate": 3.8018798183546313e-06, - "loss": 0.4041, + "loss": 0.9594, "step": 76690 }, { "epoch": 8.1, - "grad_norm": 0.054393135011196136, + "grad_norm": 60.04026412963867, "learning_rate": 3.7997676628999897e-06, - "loss": 0.3586, + "loss": 0.9053, "step": 76700 }, { "epoch": 8.1, - "grad_norm": 1.8305673599243164, + "grad_norm": 35.768516540527344, "learning_rate": 3.797655507445348e-06, - "loss": 0.0703, + "loss": 0.5943, "step": 76710 }, { "epoch": 8.1, - "grad_norm": 15.959915161132812, + "grad_norm": 43.62331771850586, "learning_rate": 3.795543351990707e-06, - "loss": 0.2197, + "loss": 0.8444, "step": 76720 }, { "epoch": 8.1, - "grad_norm": 3.1446335315704346, + "grad_norm": 40.18899917602539, "learning_rate": 3.7934311965360656e-06, - "loss": 0.4216, + "loss": 0.765, "step": 76730 }, { "epoch": 8.1, - "grad_norm": 10.685405731201172, + "grad_norm": 29.607765197753906, "learning_rate": 3.791319041081424e-06, - "loss": 0.2681, + "loss": 0.5645, "step": 76740 }, { "epoch": 8.11, - "grad_norm": 0.10873313993215561, + "grad_norm": 35.96965408325195, "learning_rate": 3.7892068856267823e-06, - "loss": 0.193, + "loss": 0.6536, "step": 76750 }, { "epoch": 8.11, - "grad_norm": 2.906334400177002, + "grad_norm": 37.067684173583984, "learning_rate": 3.787094730172141e-06, - "loss": 0.2757, + "loss": 0.9956, "step": 76760 }, { "epoch": 8.11, - "grad_norm": 0.1718871146440506, + "grad_norm": 14.427591323852539, "learning_rate": 3.7849825747174994e-06, - "loss": 0.1548, + "loss": 0.8631, "step": 76770 }, { "epoch": 8.11, - "grad_norm": 44.90729522705078, + "grad_norm": 59.0225715637207, "learning_rate": 3.782870419262858e-06, - "loss": 0.4305, + "loss": 0.9145, "step": 76780 }, { "epoch": 8.11, - "grad_norm": 5.067075252532959, + "grad_norm": 53.6340217590332, "learning_rate": 3.780758263808217e-06, - "loss": 0.2136, + "loss": 0.8677, "step": 76790 }, { "epoch": 8.11, - "grad_norm": 21.080570220947266, + "grad_norm": 34.00841522216797, "learning_rate": 3.7786461083535754e-06, - "loss": 0.1754, + "loss": 0.9375, "step": 76800 }, { "epoch": 8.11, - "grad_norm": 0.07509978860616684, + "grad_norm": 34.68312454223633, "learning_rate": 3.7765339528989337e-06, - "loss": 0.2433, + "loss": 0.82, "step": 76810 }, { "epoch": 8.11, - "grad_norm": 2.6106531620025635, + "grad_norm": 8.434770584106445, "learning_rate": 3.774421797444292e-06, - "loss": 0.3118, + "loss": 0.8188, "step": 76820 }, { "epoch": 8.11, - "grad_norm": 41.95492935180664, + "grad_norm": 45.70237731933594, "learning_rate": 3.772309641989651e-06, - "loss": 0.4551, + "loss": 0.9661, "step": 76830 }, { "epoch": 8.11, - "grad_norm": 0.019440921023488045, + "grad_norm": 29.85289192199707, "learning_rate": 3.770197486535009e-06, - "loss": 0.5742, + "loss": 0.9164, "step": 76840 }, { "epoch": 8.12, - "grad_norm": 5.008790016174316, + "grad_norm": 30.963647842407227, "learning_rate": 3.768085331080368e-06, - "loss": 0.2739, + "loss": 0.9968, "step": 76850 }, { "epoch": 8.12, - "grad_norm": 0.9866737127304077, + "grad_norm": 41.858394622802734, "learning_rate": 3.7659731756257263e-06, - "loss": 0.2353, + "loss": 0.9339, "step": 76860 }, { "epoch": 8.12, - "grad_norm": 0.5235031247138977, + "grad_norm": 37.18782424926758, "learning_rate": 3.763861020171085e-06, - "loss": 0.3698, + "loss": 0.8067, "step": 76870 }, { "epoch": 8.12, - "grad_norm": 1.7910089492797852, + "grad_norm": 42.75614929199219, "learning_rate": 3.7617488647164435e-06, - "loss": 0.2036, + "loss": 0.718, "step": 76880 }, { "epoch": 8.12, - "grad_norm": 1.9106016159057617, + "grad_norm": 17.23794174194336, "learning_rate": 3.759636709261802e-06, - "loss": 0.2775, + "loss": 0.6205, "step": 76890 }, { "epoch": 8.12, - "grad_norm": 18.80771255493164, + "grad_norm": 33.72243118286133, "learning_rate": 3.75752455380716e-06, - "loss": 0.3647, + "loss": 0.75, "step": 76900 }, { "epoch": 8.12, - "grad_norm": 23.74785804748535, + "grad_norm": 54.16986846923828, "learning_rate": 3.7554123983525194e-06, - "loss": 0.2062, + "loss": 0.9014, "step": 76910 }, { "epoch": 8.12, - "grad_norm": 25.44462776184082, + "grad_norm": 43.05165481567383, "learning_rate": 3.7533002428978777e-06, - "loss": 0.4439, + "loss": 1.0091, "step": 76920 }, { "epoch": 8.12, - "grad_norm": 0.5694310665130615, + "grad_norm": 14.879020690917969, "learning_rate": 3.751188087443236e-06, - "loss": 0.136, + "loss": 0.8805, "step": 76930 }, { "epoch": 8.13, - "grad_norm": 6.397307395935059, + "grad_norm": 37.8563232421875, "learning_rate": 3.7490759319885945e-06, - "loss": 0.2881, + "loss": 1.0466, "step": 76940 }, { "epoch": 8.13, - "grad_norm": 10.938243865966797, + "grad_norm": 49.041439056396484, "learning_rate": 3.7469637765339532e-06, - "loss": 0.3993, + "loss": 0.8798, "step": 76950 }, { "epoch": 8.13, - "grad_norm": 13.394207954406738, + "grad_norm": 50.04499435424805, "learning_rate": 3.7448516210793116e-06, - "loss": 0.2026, + "loss": 0.7705, "step": 76960 }, { "epoch": 8.13, - "grad_norm": 0.2509978115558624, + "grad_norm": 21.743755340576172, "learning_rate": 3.7427394656246704e-06, - "loss": 0.3271, + "loss": 0.799, "step": 76970 }, { "epoch": 8.13, - "grad_norm": 7.447263240814209, + "grad_norm": 51.715248107910156, "learning_rate": 3.7406273101700287e-06, - "loss": 0.4557, + "loss": 0.9503, "step": 76980 }, { "epoch": 8.13, - "grad_norm": 7.498091220855713, + "grad_norm": 49.198673248291016, "learning_rate": 3.7385151547153875e-06, - "loss": 0.4368, + "loss": 1.008, "step": 76990 }, { "epoch": 8.13, - "grad_norm": 30.40482521057129, + "grad_norm": 36.608978271484375, "learning_rate": 3.736402999260746e-06, - "loss": 0.1015, + "loss": 0.6802, "step": 77000 }, { "epoch": 8.13, - "grad_norm": 7.470492362976074, + "grad_norm": 19.373334884643555, "learning_rate": 3.7342908438061042e-06, - "loss": 0.2418, + "loss": 0.5637, "step": 77010 }, { "epoch": 8.13, - "grad_norm": 0.6002213954925537, + "grad_norm": 23.418622970581055, "learning_rate": 3.7321786883514634e-06, - "loss": 0.2666, + "loss": 0.7757, "step": 77020 }, { "epoch": 8.13, - "grad_norm": 2.779494524002075, + "grad_norm": 32.67375183105469, "learning_rate": 3.7300665328968218e-06, - "loss": 0.3766, + "loss": 0.8831, "step": 77030 }, { "epoch": 8.14, - "grad_norm": 13.931180000305176, + "grad_norm": 42.43761444091797, "learning_rate": 3.72795437744218e-06, - "loss": 0.3933, + "loss": 1.0145, "step": 77040 }, { "epoch": 8.14, - "grad_norm": 25.931324005126953, + "grad_norm": 41.6734733581543, "learning_rate": 3.7258422219875385e-06, - "loss": 0.2601, + "loss": 0.5219, "step": 77050 }, { "epoch": 8.14, - "grad_norm": 21.645973205566406, + "grad_norm": 26.594911575317383, "learning_rate": 3.7237300665328973e-06, - "loss": 0.3358, + "loss": 0.9735, "step": 77060 }, { "epoch": 8.14, - "grad_norm": 35.044673919677734, + "grad_norm": 39.43775939941406, "learning_rate": 3.7216179110782556e-06, - "loss": 0.5299, + "loss": 0.7312, "step": 77070 }, { "epoch": 8.14, - "grad_norm": 2.7541232109069824, + "grad_norm": 41.57223892211914, "learning_rate": 3.719505755623614e-06, - "loss": 0.4474, + "loss": 0.9431, "step": 77080 }, { "epoch": 8.14, - "grad_norm": 0.03707392141222954, + "grad_norm": 45.210262298583984, "learning_rate": 3.7173936001689727e-06, - "loss": 0.3568, + "loss": 1.0038, "step": 77090 }, { "epoch": 8.14, - "grad_norm": 30.165184020996094, + "grad_norm": 36.9918212890625, "learning_rate": 3.7152814447143315e-06, - "loss": 0.4461, + "loss": 1.1112, "step": 77100 }, { "epoch": 8.14, - "grad_norm": 8.234458923339844, + "grad_norm": 41.88995361328125, "learning_rate": 3.71316928925969e-06, - "loss": 0.4293, + "loss": 1.1688, "step": 77110 }, { "epoch": 8.14, - "grad_norm": 2.0923054218292236, + "grad_norm": 27.503597259521484, "learning_rate": 3.7110571338050482e-06, - "loss": 0.3161, + "loss": 0.9127, "step": 77120 }, { "epoch": 8.15, - "grad_norm": 32.05842590332031, + "grad_norm": 27.791757583618164, "learning_rate": 3.7089449783504066e-06, - "loss": 0.4686, + "loss": 0.921, "step": 77130 }, { "epoch": 8.15, - "grad_norm": 13.735872268676758, + "grad_norm": 36.56300735473633, "learning_rate": 3.706832822895766e-06, - "loss": 0.571, + "loss": 0.7983, "step": 77140 }, { "epoch": 8.15, - "grad_norm": 1.6168872117996216, + "grad_norm": 36.31320571899414, "learning_rate": 3.704720667441124e-06, - "loss": 0.3057, + "loss": 0.6321, "step": 77150 }, { "epoch": 8.15, - "grad_norm": 2.389066219329834, + "grad_norm": 40.447505950927734, "learning_rate": 3.7026085119864825e-06, - "loss": 0.2312, + "loss": 0.846, "step": 77160 }, { "epoch": 8.15, - "grad_norm": 7.514476776123047, + "grad_norm": 18.913942337036133, "learning_rate": 3.700496356531841e-06, - "loss": 0.2681, + "loss": 0.6933, "step": 77170 }, { "epoch": 8.15, - "grad_norm": 6.441927909851074, + "grad_norm": 59.078922271728516, "learning_rate": 3.6983842010771996e-06, - "loss": 0.2806, + "loss": 1.1485, "step": 77180 }, { "epoch": 8.15, - "grad_norm": 17.333860397338867, + "grad_norm": 26.162799835205078, "learning_rate": 3.696272045622558e-06, - "loss": 0.252, + "loss": 0.646, "step": 77190 }, { "epoch": 8.15, - "grad_norm": 1.2782955169677734, + "grad_norm": 58.87614822387695, "learning_rate": 3.6941598901679164e-06, - "loss": 0.2612, + "loss": 0.9394, "step": 77200 }, { "epoch": 8.15, - "grad_norm": 5.98482608795166, + "grad_norm": 26.147663116455078, "learning_rate": 3.692047734713275e-06, - "loss": 0.18, + "loss": 0.8244, "step": 77210 }, { "epoch": 8.16, - "grad_norm": 33.671958923339844, + "grad_norm": 43.53136444091797, "learning_rate": 3.689935579258634e-06, - "loss": 0.479, + "loss": 0.595, "step": 77220 }, { "epoch": 8.16, - "grad_norm": 29.033716201782227, + "grad_norm": 60.36729431152344, "learning_rate": 3.6878234238039923e-06, - "loss": 0.3304, + "loss": 0.9744, "step": 77230 }, { "epoch": 8.16, - "grad_norm": 12.79233169555664, + "grad_norm": 41.02323532104492, "learning_rate": 3.6857112683493506e-06, - "loss": 0.1893, + "loss": 0.7333, "step": 77240 }, { "epoch": 8.16, - "grad_norm": 12.280865669250488, + "grad_norm": 38.09739303588867, "learning_rate": 3.68359911289471e-06, - "loss": 0.2355, + "loss": 0.7757, "step": 77250 }, { "epoch": 8.16, - "grad_norm": 1.4073601961135864, + "grad_norm": 41.02257537841797, "learning_rate": 3.681486957440068e-06, - "loss": 0.2991, + "loss": 0.6492, "step": 77260 }, { "epoch": 8.16, - "grad_norm": 15.14755916595459, + "grad_norm": 43.027137756347656, "learning_rate": 3.6793748019854265e-06, - "loss": 0.4875, + "loss": 1.0214, "step": 77270 }, { "epoch": 8.16, - "grad_norm": 19.417404174804688, + "grad_norm": 28.835803985595703, "learning_rate": 3.677262646530785e-06, - "loss": 0.1605, + "loss": 0.6678, "step": 77280 }, { "epoch": 8.16, - "grad_norm": 25.726438522338867, + "grad_norm": 59.791587829589844, "learning_rate": 3.6751504910761437e-06, - "loss": 0.699, + "loss": 1.1461, "step": 77290 }, { "epoch": 8.16, - "grad_norm": 6.424754619598389, + "grad_norm": 64.96941375732422, "learning_rate": 3.673038335621502e-06, - "loss": 0.4249, + "loss": 0.9752, "step": 77300 }, { "epoch": 8.16, - "grad_norm": 21.742475509643555, + "grad_norm": 48.014408111572266, "learning_rate": 3.6709261801668604e-06, - "loss": 0.482, + "loss": 1.0623, "step": 77310 }, { "epoch": 8.17, - "grad_norm": 16.976381301879883, + "grad_norm": 44.522212982177734, "learning_rate": 3.6688140247122187e-06, - "loss": 0.4272, + "loss": 0.8598, "step": 77320 }, { "epoch": 8.17, - "grad_norm": 9.20108699798584, + "grad_norm": 50.17596435546875, "learning_rate": 3.666701869257578e-06, - "loss": 0.2434, + "loss": 1.1134, "step": 77330 }, { "epoch": 8.17, - "grad_norm": 1.3193303346633911, + "grad_norm": 25.251888275146484, "learning_rate": 3.6645897138029363e-06, - "loss": 0.382, + "loss": 1.1651, "step": 77340 }, { "epoch": 8.17, - "grad_norm": 0.09259151667356491, + "grad_norm": 39.820438385009766, "learning_rate": 3.6624775583482946e-06, - "loss": 0.2005, + "loss": 0.7947, "step": 77350 }, { "epoch": 8.17, - "grad_norm": 6.77695369720459, + "grad_norm": 36.45930480957031, "learning_rate": 3.660365402893653e-06, - "loss": 0.3345, + "loss": 0.9908, "step": 77360 }, { "epoch": 8.17, - "grad_norm": 0.18113112449645996, + "grad_norm": 44.517765045166016, "learning_rate": 3.658253247439012e-06, - "loss": 0.3897, + "loss": 0.9277, "step": 77370 }, { "epoch": 8.17, - "grad_norm": 16.893028259277344, + "grad_norm": 31.410236358642578, "learning_rate": 3.6561410919843706e-06, - "loss": 0.3052, + "loss": 0.8657, "step": 77380 }, { "epoch": 8.17, - "grad_norm": 0.20784644782543182, + "grad_norm": 31.48554229736328, "learning_rate": 3.654028936529729e-06, - "loss": 0.4123, + "loss": 0.6636, "step": 77390 }, { "epoch": 8.17, - "grad_norm": 11.348443031311035, + "grad_norm": 33.89075469970703, "learning_rate": 3.6519167810750873e-06, - "loss": 0.465, + "loss": 1.2429, "step": 77400 }, { "epoch": 8.18, - "grad_norm": 3.1734375953674316, + "grad_norm": 37.97579574584961, "learning_rate": 3.649804625620446e-06, - "loss": 0.3931, + "loss": 0.8312, "step": 77410 }, { "epoch": 8.18, - "grad_norm": 0.34331437945365906, + "grad_norm": 31.248422622680664, "learning_rate": 3.6476924701658044e-06, - "loss": 0.5802, + "loss": 0.8319, "step": 77420 }, { "epoch": 8.18, - "grad_norm": 22.494749069213867, + "grad_norm": 49.47380447387695, "learning_rate": 3.6455803147111628e-06, - "loss": 0.5198, + "loss": 0.9976, "step": 77430 }, { "epoch": 8.18, - "grad_norm": 17.867433547973633, + "grad_norm": 55.885276794433594, "learning_rate": 3.643468159256521e-06, - "loss": 0.3513, + "loss": 0.9726, "step": 77440 }, { "epoch": 8.18, - "grad_norm": 0.035877835005521774, + "grad_norm": 31.65802574157715, "learning_rate": 3.6413560038018803e-06, - "loss": 0.2177, + "loss": 0.5446, "step": 77450 }, { "epoch": 8.18, - "grad_norm": 3.158801794052124, + "grad_norm": 24.045286178588867, "learning_rate": 3.6392438483472387e-06, - "loss": 0.261, + "loss": 1.0624, "step": 77460 }, { "epoch": 8.18, - "grad_norm": 0.9709428548812866, + "grad_norm": 25.889726638793945, "learning_rate": 3.637131692892597e-06, - "loss": 0.4184, + "loss": 0.6085, "step": 77470 }, { "epoch": 8.18, - "grad_norm": 24.83090591430664, + "grad_norm": 39.74617385864258, "learning_rate": 3.635019537437956e-06, - "loss": 0.3087, + "loss": 0.7692, "step": 77480 }, { "epoch": 8.18, - "grad_norm": 34.57716751098633, + "grad_norm": 29.05797004699707, "learning_rate": 3.6329073819833146e-06, - "loss": 0.4505, + "loss": 0.7879, "step": 77490 }, { "epoch": 8.18, - "grad_norm": 0.28037774562835693, + "grad_norm": 53.69739532470703, "learning_rate": 3.630795226528673e-06, - "loss": 0.3168, + "loss": 0.8172, "step": 77500 }, { "epoch": 8.19, - "grad_norm": 21.349899291992188, + "grad_norm": 57.65095520019531, "learning_rate": 3.6286830710740313e-06, - "loss": 0.3747, + "loss": 0.8765, "step": 77510 }, { "epoch": 8.19, - "grad_norm": 5.62263822555542, + "grad_norm": 32.2320442199707, "learning_rate": 3.62657091561939e-06, - "loss": 0.2858, + "loss": 0.8522, "step": 77520 }, { "epoch": 8.19, - "grad_norm": 6.524664878845215, + "grad_norm": 33.132450103759766, "learning_rate": 3.6244587601647484e-06, - "loss": 0.4533, + "loss": 0.6524, "step": 77530 }, { "epoch": 8.19, - "grad_norm": 35.01191329956055, + "grad_norm": 36.522727966308594, "learning_rate": 3.6223466047101068e-06, - "loss": 0.2813, + "loss": 0.9016, "step": 77540 }, { "epoch": 8.19, - "grad_norm": 0.3237542510032654, + "grad_norm": 64.5953598022461, "learning_rate": 3.620234449255465e-06, - "loss": 0.4117, + "loss": 1.1872, "step": 77550 }, { "epoch": 8.19, - "grad_norm": 7.961933135986328, + "grad_norm": 43.81662368774414, "learning_rate": 3.6181222938008243e-06, - "loss": 0.2403, + "loss": 0.6832, "step": 77560 }, { "epoch": 8.19, - "grad_norm": 27.286813735961914, + "grad_norm": 62.28089904785156, "learning_rate": 3.6160101383461827e-06, - "loss": 0.3357, + "loss": 0.6831, "step": 77570 }, { "epoch": 8.19, - "grad_norm": 25.223712921142578, + "grad_norm": 46.995201110839844, "learning_rate": 3.613897982891541e-06, - "loss": 0.3432, + "loss": 1.0172, "step": 77580 }, { "epoch": 8.19, - "grad_norm": 31.37757110595703, + "grad_norm": 58.12638854980469, "learning_rate": 3.6117858274368994e-06, - "loss": 0.3685, + "loss": 0.7712, "step": 77590 }, { "epoch": 8.2, - "grad_norm": 2.0790815353393555, + "grad_norm": 23.333322525024414, "learning_rate": 3.609673671982258e-06, - "loss": 0.2865, + "loss": 0.5747, "step": 77600 }, { "epoch": 8.2, - "grad_norm": 26.70370864868164, + "grad_norm": 23.948925018310547, "learning_rate": 3.607561516527617e-06, - "loss": 0.4296, + "loss": 0.8965, "step": 77610 }, { "epoch": 8.2, - "grad_norm": 5.492709636688232, + "grad_norm": 20.810169219970703, "learning_rate": 3.6054493610729753e-06, - "loss": 0.4154, + "loss": 0.8816, "step": 77620 }, { "epoch": 8.2, - "grad_norm": 18.845806121826172, + "grad_norm": 58.49864959716797, "learning_rate": 3.6033372056183337e-06, - "loss": 0.2118, + "loss": 0.822, "step": 77630 }, { "epoch": 8.2, - "grad_norm": 22.3959903717041, + "grad_norm": 39.62188720703125, "learning_rate": 3.6012250501636925e-06, - "loss": 0.4284, + "loss": 0.8519, "step": 77640 }, { "epoch": 8.2, - "grad_norm": 14.952693939208984, + "grad_norm": 38.39491271972656, "learning_rate": 3.599112894709051e-06, - "loss": 0.5615, + "loss": 1.0379, "step": 77650 }, { "epoch": 8.2, - "grad_norm": 9.377654075622559, + "grad_norm": 6.31091833114624, "learning_rate": 3.597000739254409e-06, - "loss": 0.4482, + "loss": 0.6502, "step": 77660 }, { "epoch": 8.2, - "grad_norm": 6.049966812133789, + "grad_norm": 18.487916946411133, "learning_rate": 3.5948885837997675e-06, - "loss": 0.1979, + "loss": 0.8022, "step": 77670 }, { "epoch": 8.2, - "grad_norm": 14.652152061462402, + "grad_norm": 38.994972229003906, "learning_rate": 3.5927764283451267e-06, - "loss": 0.5301, + "loss": 0.9463, "step": 77680 }, { "epoch": 8.2, - "grad_norm": 24.775129318237305, + "grad_norm": 35.416748046875, "learning_rate": 3.590664272890485e-06, - "loss": 0.5456, + "loss": 1.0766, "step": 77690 }, { "epoch": 8.21, - "grad_norm": 0.469729483127594, + "grad_norm": 24.294008255004883, "learning_rate": 3.5885521174358434e-06, - "loss": 0.1335, + "loss": 0.8291, "step": 77700 }, { "epoch": 8.21, - "grad_norm": 10.619551658630371, + "grad_norm": 20.19132423400879, "learning_rate": 3.5864399619812022e-06, - "loss": 0.3751, + "loss": 0.7899, "step": 77710 }, { "epoch": 8.21, - "grad_norm": 7.325500965118408, + "grad_norm": 45.74686050415039, "learning_rate": 3.5843278065265606e-06, - "loss": 0.4217, + "loss": 0.811, "step": 77720 }, { "epoch": 8.21, - "grad_norm": 0.028214825317263603, + "grad_norm": 3.06451416015625, "learning_rate": 3.5822156510719193e-06, - "loss": 0.237, + "loss": 0.6156, "step": 77730 }, { "epoch": 8.21, - "grad_norm": 14.165878295898438, + "grad_norm": 20.169456481933594, "learning_rate": 3.5801034956172777e-06, - "loss": 0.4318, + "loss": 0.7765, "step": 77740 }, { "epoch": 8.21, - "grad_norm": 8.319300651550293, + "grad_norm": 22.324661254882812, "learning_rate": 3.5779913401626365e-06, - "loss": 0.4218, + "loss": 1.1144, "step": 77750 }, { "epoch": 8.21, - "grad_norm": 3.714085578918457, + "grad_norm": 45.7922248840332, "learning_rate": 3.575879184707995e-06, - "loss": 0.3184, + "loss": 0.8683, "step": 77760 }, { "epoch": 8.21, - "grad_norm": 1.0457760095596313, + "grad_norm": 33.39932632446289, "learning_rate": 3.573767029253353e-06, - "loss": 0.3001, + "loss": 0.8041, "step": 77770 }, { "epoch": 8.21, - "grad_norm": 1.566832423210144, + "grad_norm": 44.6665153503418, "learning_rate": 3.5716548737987115e-06, - "loss": 0.2554, + "loss": 0.8159, "step": 77780 }, { "epoch": 8.22, - "grad_norm": 15.563639640808105, + "grad_norm": 59.862632751464844, "learning_rate": 3.5695427183440708e-06, - "loss": 0.4511, + "loss": 0.9903, "step": 77790 }, { "epoch": 8.22, - "grad_norm": 0.8699021339416504, + "grad_norm": 24.00982666015625, "learning_rate": 3.567430562889429e-06, - "loss": 0.4369, + "loss": 1.0032, "step": 77800 }, { "epoch": 8.22, - "grad_norm": 0.05534956231713295, + "grad_norm": 12.758292198181152, "learning_rate": 3.5653184074347875e-06, - "loss": 0.3485, + "loss": 0.6416, "step": 77810 }, { "epoch": 8.22, - "grad_norm": 16.230615615844727, + "grad_norm": 34.49304962158203, "learning_rate": 3.563206251980146e-06, - "loss": 0.2492, + "loss": 0.8617, "step": 77820 }, { "epoch": 8.22, - "grad_norm": 2.959404230117798, + "grad_norm": 45.462066650390625, "learning_rate": 3.5610940965255046e-06, - "loss": 0.5356, + "loss": 1.2159, "step": 77830 }, { "epoch": 8.22, - "grad_norm": 7.4731597900390625, + "grad_norm": 43.451900482177734, "learning_rate": 3.558981941070863e-06, - "loss": 0.1796, + "loss": 0.9078, "step": 77840 }, { "epoch": 8.22, - "grad_norm": 7.943274021148682, + "grad_norm": 61.42511749267578, "learning_rate": 3.5568697856162217e-06, - "loss": 0.6202, + "loss": 0.7971, "step": 77850 }, { "epoch": 8.22, - "grad_norm": 0.2073638141155243, + "grad_norm": 61.6181640625, "learning_rate": 3.55475763016158e-06, - "loss": 0.2928, + "loss": 0.7421, "step": 77860 }, { "epoch": 8.22, - "grad_norm": 0.038703497499227524, + "grad_norm": 23.51395034790039, "learning_rate": 3.552645474706939e-06, - "loss": 0.4333, + "loss": 0.989, "step": 77870 }, { "epoch": 8.22, - "grad_norm": 17.548927307128906, + "grad_norm": 33.49885940551758, "learning_rate": 3.5505333192522972e-06, - "loss": 0.2513, + "loss": 1.2009, "step": 77880 }, { "epoch": 8.23, - "grad_norm": 36.73402404785156, + "grad_norm": 48.043704986572266, "learning_rate": 3.5484211637976556e-06, - "loss": 0.1346, + "loss": 0.7293, "step": 77890 }, { "epoch": 8.23, - "grad_norm": 7.3492045402526855, + "grad_norm": 8.014723777770996, "learning_rate": 3.546309008343014e-06, - "loss": 0.6345, + "loss": 1.0092, "step": 77900 }, { "epoch": 8.23, - "grad_norm": 0.33654677867889404, + "grad_norm": 52.28640365600586, "learning_rate": 3.544196852888373e-06, - "loss": 0.1215, + "loss": 0.4834, "step": 77910 }, { "epoch": 8.23, - "grad_norm": 0.0358554869890213, + "grad_norm": 18.313932418823242, "learning_rate": 3.5420846974337315e-06, - "loss": 0.2339, + "loss": 0.6636, "step": 77920 }, { "epoch": 8.23, - "grad_norm": 0.055990442633628845, + "grad_norm": 44.68718719482422, "learning_rate": 3.53997254197909e-06, - "loss": 0.2604, + "loss": 0.9144, "step": 77930 }, { "epoch": 8.23, - "grad_norm": 1.9440339803695679, + "grad_norm": 34.95731735229492, "learning_rate": 3.5378603865244486e-06, - "loss": 0.293, + "loss": 0.8296, "step": 77940 }, { "epoch": 8.23, - "grad_norm": 6.449297904968262, + "grad_norm": 22.955123901367188, "learning_rate": 3.535748231069807e-06, - "loss": 0.2225, + "loss": 0.796, "step": 77950 }, { "epoch": 8.23, - "grad_norm": 10.881413459777832, + "grad_norm": 42.43042755126953, "learning_rate": 3.5336360756151653e-06, - "loss": 0.2547, + "loss": 0.771, "step": 77960 }, { "epoch": 8.23, - "grad_norm": 15.10756778717041, + "grad_norm": 25.21091651916504, "learning_rate": 3.531523920160524e-06, - "loss": 0.2544, + "loss": 0.6001, "step": 77970 }, { "epoch": 8.24, - "grad_norm": 35.20139694213867, + "grad_norm": 23.839439392089844, "learning_rate": 3.529411764705883e-06, - "loss": 0.5643, + "loss": 0.6847, "step": 77980 }, { "epoch": 8.24, - "grad_norm": 21.717098236083984, + "grad_norm": 9.278824806213379, "learning_rate": 3.5272996092512412e-06, - "loss": 0.2606, + "loss": 0.8275, "step": 77990 }, { "epoch": 8.24, - "grad_norm": 24.726112365722656, + "grad_norm": 60.5943717956543, "learning_rate": 3.5251874537965996e-06, - "loss": 0.3862, + "loss": 1.068, "step": 78000 }, { "epoch": 8.24, - "grad_norm": 0.09837009012699127, + "grad_norm": 55.93135452270508, "learning_rate": 3.523075298341958e-06, - "loss": 0.3792, + "loss": 1.1988, "step": 78010 }, { "epoch": 8.24, - "grad_norm": 27.597667694091797, + "grad_norm": 36.444828033447266, "learning_rate": 3.520963142887317e-06, - "loss": 0.2592, + "loss": 0.6734, "step": 78020 }, { "epoch": 8.24, - "grad_norm": 14.339701652526855, + "grad_norm": 20.43733024597168, "learning_rate": 3.5188509874326755e-06, - "loss": 0.2665, + "loss": 0.894, "step": 78030 }, { "epoch": 8.24, - "grad_norm": 3.8075695037841797, + "grad_norm": 16.617708206176758, "learning_rate": 3.516738831978034e-06, - "loss": 0.3723, + "loss": 0.944, "step": 78040 }, { "epoch": 8.24, - "grad_norm": 0.7712281346321106, + "grad_norm": 47.877872467041016, "learning_rate": 3.5146266765233922e-06, - "loss": 0.4248, + "loss": 1.1552, "step": 78050 }, { "epoch": 8.24, - "grad_norm": 0.16534794867038727, + "grad_norm": 25.45210075378418, "learning_rate": 3.512514521068751e-06, - "loss": 0.284, + "loss": 0.7805, "step": 78060 }, { "epoch": 8.24, - "grad_norm": 7.0733466148376465, + "grad_norm": 47.485450744628906, "learning_rate": 3.5104023656141094e-06, - "loss": 0.3153, + "loss": 0.8871, "step": 78070 }, { "epoch": 8.25, - "grad_norm": 0.39071959257125854, + "grad_norm": 22.826190948486328, "learning_rate": 3.5082902101594677e-06, - "loss": 0.2414, + "loss": 0.7105, "step": 78080 }, { "epoch": 8.25, - "grad_norm": 15.431931495666504, + "grad_norm": 22.52375030517578, "learning_rate": 3.5061780547048265e-06, - "loss": 0.3121, + "loss": 0.8025, "step": 78090 }, { "epoch": 8.25, - "grad_norm": 20.3333683013916, + "grad_norm": 37.16407012939453, "learning_rate": 3.5040658992501853e-06, - "loss": 0.2747, + "loss": 0.9433, "step": 78100 }, { "epoch": 8.25, - "grad_norm": 0.03506368026137352, + "grad_norm": 14.679417610168457, "learning_rate": 3.5019537437955436e-06, - "loss": 0.2786, + "loss": 1.2153, "step": 78110 }, { "epoch": 8.25, - "grad_norm": 28.388917922973633, + "grad_norm": 59.02058029174805, "learning_rate": 3.499841588340902e-06, - "loss": 0.4666, + "loss": 1.0141, "step": 78120 }, { "epoch": 8.25, - "grad_norm": 0.021657217293977737, + "grad_norm": 53.6971549987793, "learning_rate": 3.497729432886261e-06, - "loss": 0.2139, + "loss": 0.8125, "step": 78130 }, { "epoch": 8.25, - "grad_norm": 0.008670026436448097, + "grad_norm": 44.75459671020508, "learning_rate": 3.4956172774316195e-06, - "loss": 0.3323, + "loss": 0.7333, "step": 78140 }, { "epoch": 8.25, - "grad_norm": 1.6163524389266968, + "grad_norm": 35.37485122680664, "learning_rate": 3.493505121976978e-06, - "loss": 0.3324, + "loss": 0.674, "step": 78150 }, { "epoch": 8.25, - "grad_norm": 2.086963415145874, + "grad_norm": 25.384254455566406, "learning_rate": 3.4913929665223363e-06, - "loss": 0.2074, + "loss": 1.0283, "step": 78160 }, { "epoch": 8.26, - "grad_norm": 13.199884414672852, + "grad_norm": 43.22197341918945, "learning_rate": 3.489280811067695e-06, - "loss": 0.357, + "loss": 0.8654, "step": 78170 }, { "epoch": 8.26, - "grad_norm": 9.889691352844238, + "grad_norm": 39.22345733642578, "learning_rate": 3.4871686556130534e-06, - "loss": 0.199, + "loss": 0.6639, "step": 78180 }, { "epoch": 8.26, - "grad_norm": 14.362906455993652, + "grad_norm": 22.171405792236328, "learning_rate": 3.4850565001584117e-06, - "loss": 0.3089, + "loss": 0.546, "step": 78190 }, { "epoch": 8.26, - "grad_norm": 0.33534857630729675, + "grad_norm": 31.029150009155273, "learning_rate": 3.48294434470377e-06, - "loss": 0.0846, + "loss": 0.6862, "step": 78200 }, { "epoch": 8.26, - "grad_norm": 9.646768569946289, + "grad_norm": 46.22352600097656, "learning_rate": 3.4808321892491293e-06, - "loss": 0.4914, + "loss": 0.9052, "step": 78210 }, { "epoch": 8.26, - "grad_norm": 5.346583843231201, + "grad_norm": 10.113018989562988, "learning_rate": 3.4787200337944877e-06, - "loss": 0.1125, + "loss": 0.4403, "step": 78220 }, { "epoch": 8.26, - "grad_norm": 2.6723382472991943, + "grad_norm": 50.96648025512695, "learning_rate": 3.476607878339846e-06, - "loss": 0.1276, + "loss": 0.7792, "step": 78230 }, { "epoch": 8.26, - "grad_norm": 19.839279174804688, + "grad_norm": 46.004241943359375, "learning_rate": 3.4744957228852044e-06, - "loss": 0.1844, + "loss": 0.987, "step": 78240 }, { "epoch": 8.26, - "grad_norm": 4.576242923736572, + "grad_norm": 69.16532135009766, "learning_rate": 3.4723835674305636e-06, - "loss": 0.1153, + "loss": 0.8733, "step": 78250 }, { "epoch": 8.26, - "grad_norm": 0.0543917752802372, + "grad_norm": 18.802942276000977, "learning_rate": 3.470271411975922e-06, - "loss": 0.2489, + "loss": 0.7117, "step": 78260 }, { "epoch": 8.27, - "grad_norm": 8.809123039245605, + "grad_norm": 30.254596710205078, "learning_rate": 3.4681592565212803e-06, - "loss": 0.5028, + "loss": 0.8383, "step": 78270 }, { "epoch": 8.27, - "grad_norm": 11.61505126953125, + "grad_norm": 39.77037811279297, "learning_rate": 3.4660471010666386e-06, - "loss": 0.3289, + "loss": 0.8434, "step": 78280 }, { "epoch": 8.27, - "grad_norm": 8.661539077758789, + "grad_norm": 59.28615188598633, "learning_rate": 3.4639349456119974e-06, - "loss": 0.2824, + "loss": 0.8457, "step": 78290 }, { "epoch": 8.27, - "grad_norm": 7.979767322540283, + "grad_norm": 17.85972023010254, "learning_rate": 3.4618227901573558e-06, - "loss": 0.0777, + "loss": 0.6866, "step": 78300 }, { "epoch": 8.27, - "grad_norm": 5.4941511154174805, + "grad_norm": 22.346986770629883, "learning_rate": 3.459710634702714e-06, - "loss": 0.4776, + "loss": 0.832, "step": 78310 }, { "epoch": 8.27, - "grad_norm": 18.235836029052734, + "grad_norm": 47.45930480957031, "learning_rate": 3.4575984792480725e-06, - "loss": 0.316, + "loss": 0.7455, "step": 78320 }, { "epoch": 8.27, - "grad_norm": 6.098620891571045, + "grad_norm": 57.637298583984375, "learning_rate": 3.4554863237934317e-06, - "loss": 0.3284, + "loss": 1.0852, "step": 78330 }, { "epoch": 8.27, - "grad_norm": 1.842379093170166, + "grad_norm": 47.18904113769531, "learning_rate": 3.45337416833879e-06, - "loss": 0.3924, + "loss": 0.9642, "step": 78340 }, { "epoch": 8.27, - "grad_norm": 0.9372419714927673, + "grad_norm": 47.94710159301758, "learning_rate": 3.4512620128841484e-06, - "loss": 0.4514, + "loss": 0.9298, "step": 78350 }, { "epoch": 8.28, - "grad_norm": 0.031861770898103714, + "grad_norm": 29.203838348388672, "learning_rate": 3.449149857429507e-06, - "loss": 0.2196, + "loss": 0.5801, "step": 78360 }, { "epoch": 8.28, - "grad_norm": 1.0363377332687378, + "grad_norm": 52.13480758666992, "learning_rate": 3.447037701974866e-06, - "loss": 0.3234, + "loss": 0.9401, "step": 78370 }, { "epoch": 8.28, - "grad_norm": 0.3921992778778076, + "grad_norm": 14.528130531311035, "learning_rate": 3.4449255465202243e-06, - "loss": 0.414, + "loss": 0.6798, "step": 78380 }, { "epoch": 8.28, - "grad_norm": 10.502466201782227, + "grad_norm": 54.32728576660156, "learning_rate": 3.4428133910655827e-06, - "loss": 0.398, + "loss": 0.9047, "step": 78390 }, { "epoch": 8.28, - "grad_norm": 5.783351421356201, + "grad_norm": 9.449475288391113, "learning_rate": 3.4407012356109414e-06, - "loss": 0.3561, + "loss": 0.7516, "step": 78400 }, { "epoch": 8.28, - "grad_norm": 3.989466667175293, + "grad_norm": 26.3072452545166, "learning_rate": 3.4385890801563e-06, - "loss": 0.4254, + "loss": 0.8088, "step": 78410 }, { "epoch": 8.28, - "grad_norm": 24.925464630126953, + "grad_norm": 29.899412155151367, "learning_rate": 3.436476924701658e-06, - "loss": 0.5696, + "loss": 1.0091, "step": 78420 }, { "epoch": 8.28, - "grad_norm": 28.15450668334961, + "grad_norm": 48.74740982055664, "learning_rate": 3.4343647692470165e-06, - "loss": 0.3132, + "loss": 0.8206, "step": 78430 }, { "epoch": 8.28, - "grad_norm": 18.397382736206055, + "grad_norm": 25.946998596191406, "learning_rate": 3.4322526137923757e-06, - "loss": 0.2552, + "loss": 0.9209, "step": 78440 }, { "epoch": 8.28, - "grad_norm": 0.022521328181028366, + "grad_norm": 51.89414596557617, "learning_rate": 3.430140458337734e-06, - "loss": 0.1766, + "loss": 0.4731, "step": 78450 }, { "epoch": 8.29, - "grad_norm": 22.10721778869629, + "grad_norm": 33.695350646972656, "learning_rate": 3.4280283028830924e-06, - "loss": 0.3112, + "loss": 0.7767, "step": 78460 }, { "epoch": 8.29, - "grad_norm": 0.04890979081392288, + "grad_norm": 34.94059753417969, "learning_rate": 3.4259161474284508e-06, - "loss": 0.1994, + "loss": 1.0808, "step": 78470 }, { "epoch": 8.29, - "grad_norm": 15.2684965133667, + "grad_norm": 64.25703430175781, "learning_rate": 3.4238039919738096e-06, - "loss": 0.2004, + "loss": 0.7279, "step": 78480 }, { "epoch": 8.29, - "grad_norm": 27.407562255859375, + "grad_norm": 31.7177677154541, "learning_rate": 3.4216918365191683e-06, - "loss": 0.4284, + "loss": 0.9257, "step": 78490 }, { "epoch": 8.29, - "grad_norm": 1.3551228046417236, + "grad_norm": 40.35898971557617, "learning_rate": 3.4195796810645267e-06, - "loss": 0.4571, + "loss": 1.2454, "step": 78500 }, { "epoch": 8.29, - "grad_norm": 12.51526927947998, + "grad_norm": 66.7293701171875, "learning_rate": 3.417467525609885e-06, - "loss": 0.3979, + "loss": 0.8713, "step": 78510 }, { "epoch": 8.29, - "grad_norm": 6.794954299926758, + "grad_norm": 41.83265686035156, "learning_rate": 3.415355370155244e-06, - "loss": 0.4516, + "loss": 0.753, "step": 78520 }, { "epoch": 8.29, - "grad_norm": 17.92502784729004, + "grad_norm": 63.57430648803711, "learning_rate": 3.413243214700602e-06, - "loss": 0.3304, + "loss": 1.0169, "step": 78530 }, { "epoch": 8.29, - "grad_norm": 4.98729944229126, + "grad_norm": 22.60603904724121, "learning_rate": 3.4111310592459605e-06, - "loss": 0.551, + "loss": 0.9087, "step": 78540 }, { "epoch": 8.3, - "grad_norm": 18.89369010925293, + "grad_norm": 42.84197998046875, "learning_rate": 3.409018903791319e-06, - "loss": 0.4665, + "loss": 0.6616, "step": 78550 }, { "epoch": 8.3, - "grad_norm": 0.3938539922237396, + "grad_norm": 36.27184295654297, "learning_rate": 3.406906748336678e-06, - "loss": 0.4011, + "loss": 1.0794, "step": 78560 }, { "epoch": 8.3, - "grad_norm": 12.307405471801758, + "grad_norm": 23.595985412597656, "learning_rate": 3.4047945928820364e-06, - "loss": 0.5695, + "loss": 1.0933, "step": 78570 }, { "epoch": 8.3, - "grad_norm": 12.000617027282715, + "grad_norm": 56.873661041259766, "learning_rate": 3.402682437427395e-06, - "loss": 0.4973, + "loss": 1.1345, "step": 78580 }, { "epoch": 8.3, - "grad_norm": 34.44903564453125, + "grad_norm": 29.1407470703125, "learning_rate": 3.4005702819727536e-06, - "loss": 0.3504, + "loss": 0.757, "step": 78590 }, { "epoch": 8.3, - "grad_norm": 0.20289349555969238, + "grad_norm": 22.684152603149414, "learning_rate": 3.398458126518112e-06, - "loss": 0.4272, + "loss": 0.9815, "step": 78600 }, { "epoch": 8.3, - "grad_norm": 6.162397384643555, + "grad_norm": 41.48590087890625, "learning_rate": 3.3963459710634707e-06, - "loss": 0.2808, + "loss": 0.7996, "step": 78610 }, { "epoch": 8.3, - "grad_norm": 11.583560943603516, + "grad_norm": 35.73662185668945, "learning_rate": 3.394233815608829e-06, - "loss": 0.1803, + "loss": 0.8365, "step": 78620 }, { "epoch": 8.3, - "grad_norm": 18.8724365234375, + "grad_norm": 75.16204071044922, "learning_rate": 3.392121660154188e-06, - "loss": 0.4412, + "loss": 0.6831, "step": 78630 }, { "epoch": 8.3, - "grad_norm": 7.535359859466553, + "grad_norm": 49.96588897705078, "learning_rate": 3.390009504699546e-06, - "loss": 0.4033, + "loss": 1.2882, "step": 78640 }, { "epoch": 8.31, - "grad_norm": 3.85715913772583, + "grad_norm": 62.41805648803711, "learning_rate": 3.3878973492449046e-06, - "loss": 0.4327, + "loss": 0.9103, "step": 78650 }, { "epoch": 8.31, - "grad_norm": 7.2550740242004395, + "grad_norm": 40.695640563964844, "learning_rate": 3.385785193790263e-06, - "loss": 0.5399, + "loss": 0.9007, "step": 78660 }, { "epoch": 8.31, - "grad_norm": 1.764076590538025, + "grad_norm": 23.200252532958984, "learning_rate": 3.383673038335622e-06, - "loss": 0.3253, + "loss": 0.6287, "step": 78670 }, { "epoch": 8.31, - "grad_norm": 19.423688888549805, + "grad_norm": 32.571659088134766, "learning_rate": 3.3815608828809805e-06, - "loss": 0.5152, + "loss": 1.5749, "step": 78680 }, { "epoch": 8.31, - "grad_norm": 29.488018035888672, + "grad_norm": 32.17451095581055, "learning_rate": 3.379448727426339e-06, - "loss": 0.383, + "loss": 0.8422, "step": 78690 }, { "epoch": 8.31, - "grad_norm": 0.05585739016532898, + "grad_norm": 12.506895065307617, "learning_rate": 3.377336571971697e-06, - "loss": 0.2995, + "loss": 0.9996, "step": 78700 }, { "epoch": 8.31, - "grad_norm": 0.16686537861824036, + "grad_norm": 39.47262954711914, "learning_rate": 3.375224416517056e-06, - "loss": 0.1905, + "loss": 1.041, "step": 78710 }, { "epoch": 8.31, - "grad_norm": 11.45064640045166, + "grad_norm": 43.89374542236328, "learning_rate": 3.3731122610624143e-06, - "loss": 0.1633, + "loss": 0.8078, "step": 78720 }, { "epoch": 8.31, - "grad_norm": 41.03099822998047, + "grad_norm": 48.34931564331055, "learning_rate": 3.371000105607773e-06, - "loss": 0.5404, + "loss": 0.8996, "step": 78730 }, { "epoch": 8.32, - "grad_norm": 40.99474334716797, + "grad_norm": 48.6919059753418, "learning_rate": 3.3688879501531314e-06, - "loss": 0.4788, + "loss": 1.0535, "step": 78740 }, { "epoch": 8.32, - "grad_norm": 29.361371994018555, + "grad_norm": 61.20661163330078, "learning_rate": 3.3667757946984902e-06, - "loss": 0.3366, + "loss": 1.0659, "step": 78750 }, { "epoch": 8.32, - "grad_norm": 16.347633361816406, + "grad_norm": 43.02803421020508, "learning_rate": 3.3646636392438486e-06, - "loss": 0.3699, + "loss": 0.7673, "step": 78760 }, { "epoch": 8.32, - "grad_norm": 14.209993362426758, + "grad_norm": 48.78608703613281, "learning_rate": 3.362551483789207e-06, - "loss": 0.161, + "loss": 0.7915, "step": 78770 }, { "epoch": 8.32, - "grad_norm": 3.214400291442871, + "grad_norm": 41.66514205932617, "learning_rate": 3.3604393283345653e-06, - "loss": 0.1457, + "loss": 0.7569, "step": 78780 }, { "epoch": 8.32, - "grad_norm": 19.368656158447266, + "grad_norm": 42.494476318359375, "learning_rate": 3.3583271728799245e-06, - "loss": 0.3795, + "loss": 1.0415, "step": 78790 }, { "epoch": 8.32, - "grad_norm": 6.876274585723877, + "grad_norm": 56.24967956542969, "learning_rate": 3.356215017425283e-06, - "loss": 0.3891, + "loss": 0.9982, "step": 78800 }, { "epoch": 8.32, - "grad_norm": 0.3802970349788666, + "grad_norm": 52.53630065917969, "learning_rate": 3.354102861970641e-06, - "loss": 0.3891, + "loss": 0.8278, "step": 78810 }, { "epoch": 8.32, - "grad_norm": 0.25804853439331055, + "grad_norm": 26.48975944519043, "learning_rate": 3.351990706516e-06, - "loss": 0.361, + "loss": 0.8401, "step": 78820 }, { "epoch": 8.33, - "grad_norm": 0.8657065629959106, + "grad_norm": 25.724348068237305, "learning_rate": 3.3498785510613583e-06, - "loss": 0.2769, + "loss": 1.0121, "step": 78830 }, { "epoch": 8.33, - "grad_norm": 0.3219817280769348, + "grad_norm": 16.024250030517578, "learning_rate": 3.3477663956067167e-06, - "loss": 0.4375, + "loss": 0.9188, "step": 78840 }, { "epoch": 8.33, - "grad_norm": 8.479878425598145, + "grad_norm": 45.849220275878906, "learning_rate": 3.3456542401520755e-06, - "loss": 0.4995, + "loss": 1.0427, "step": 78850 }, { "epoch": 8.33, - "grad_norm": 0.9877587556838989, + "grad_norm": 37.383689880371094, "learning_rate": 3.3435420846974343e-06, - "loss": 0.2104, + "loss": 0.9203, "step": 78860 }, { "epoch": 8.33, - "grad_norm": 2.7895867824554443, + "grad_norm": 53.381370544433594, "learning_rate": 3.3414299292427926e-06, - "loss": 0.2041, + "loss": 0.8491, "step": 78870 }, { "epoch": 8.33, - "grad_norm": 45.69293975830078, + "grad_norm": 35.098876953125, "learning_rate": 3.339317773788151e-06, - "loss": 0.2526, + "loss": 0.7882, "step": 78880 }, { "epoch": 8.33, - "grad_norm": 7.15326452255249, + "grad_norm": 55.5716667175293, "learning_rate": 3.3372056183335093e-06, - "loss": 0.0918, + "loss": 0.8633, "step": 78890 }, { "epoch": 8.33, - "grad_norm": 0.1266065388917923, + "grad_norm": 52.406982421875, "learning_rate": 3.3350934628788685e-06, - "loss": 0.3561, + "loss": 1.1477, "step": 78900 }, { "epoch": 8.33, - "grad_norm": 42.42872619628906, + "grad_norm": 49.703189849853516, "learning_rate": 3.332981307424227e-06, - "loss": 0.4432, + "loss": 0.9262, "step": 78910 }, { "epoch": 8.33, - "grad_norm": 23.901884078979492, + "grad_norm": 40.14885330200195, "learning_rate": 3.3308691519695852e-06, - "loss": 0.2653, + "loss": 0.7898, "step": 78920 }, { "epoch": 8.34, - "grad_norm": 4.329481601715088, + "grad_norm": 51.515281677246094, "learning_rate": 3.3287569965149436e-06, - "loss": 0.3822, + "loss": 0.7813, "step": 78930 }, { "epoch": 8.34, - "grad_norm": 0.9658788442611694, + "grad_norm": 32.51473617553711, "learning_rate": 3.3266448410603024e-06, - "loss": 0.2016, + "loss": 1.0676, "step": 78940 }, { "epoch": 8.34, - "grad_norm": 8.65249252319336, + "grad_norm": 29.04009246826172, "learning_rate": 3.3245326856056607e-06, - "loss": 0.2431, + "loss": 0.8335, "step": 78950 }, { "epoch": 8.34, - "grad_norm": 11.003281593322754, + "grad_norm": 36.40818786621094, "learning_rate": 3.322420530151019e-06, - "loss": 0.3907, + "loss": 0.996, "step": 78960 }, { "epoch": 8.34, - "grad_norm": 23.168012619018555, + "grad_norm": 46.347042083740234, "learning_rate": 3.320308374696378e-06, - "loss": 0.3451, + "loss": 0.8131, "step": 78970 }, { "epoch": 8.34, - "grad_norm": 12.27682113647461, + "grad_norm": 41.29164123535156, "learning_rate": 3.3181962192417366e-06, - "loss": 0.2492, + "loss": 0.8369, "step": 78980 }, { "epoch": 8.34, - "grad_norm": 0.36363208293914795, + "grad_norm": 33.50315475463867, "learning_rate": 3.316084063787095e-06, - "loss": 0.3555, + "loss": 0.7508, "step": 78990 }, { "epoch": 8.34, - "grad_norm": 26.818897247314453, + "grad_norm": 25.413509368896484, "learning_rate": 3.3139719083324533e-06, - "loss": 0.5481, + "loss": 0.7352, "step": 79000 }, { "epoch": 8.34, - "grad_norm": 13.063281059265137, + "grad_norm": 32.332706451416016, "learning_rate": 3.3118597528778117e-06, - "loss": 0.2495, + "loss": 0.9892, "step": 79010 }, { "epoch": 8.35, - "grad_norm": 31.412065505981445, + "grad_norm": 28.757312774658203, "learning_rate": 3.309747597423171e-06, - "loss": 0.3972, + "loss": 0.8786, "step": 79020 }, { "epoch": 8.35, - "grad_norm": 22.460601806640625, + "grad_norm": 59.694271087646484, "learning_rate": 3.3076354419685293e-06, - "loss": 0.402, + "loss": 0.8534, "step": 79030 }, { "epoch": 8.35, - "grad_norm": 0.6790599822998047, + "grad_norm": 52.90261459350586, "learning_rate": 3.3055232865138876e-06, - "loss": 0.4126, + "loss": 0.729, "step": 79040 }, { "epoch": 8.35, - "grad_norm": 37.94498062133789, + "grad_norm": 57.62074279785156, "learning_rate": 3.3034111310592464e-06, - "loss": 0.4577, + "loss": 0.8878, "step": 79050 }, { "epoch": 8.35, - "grad_norm": 6.4243950843811035, + "grad_norm": 30.228811264038086, "learning_rate": 3.3012989756046047e-06, - "loss": 0.3129, + "loss": 0.7811, "step": 79060 }, { "epoch": 8.35, - "grad_norm": 33.4234619140625, + "grad_norm": 55.61953353881836, "learning_rate": 3.299186820149963e-06, - "loss": 0.4192, + "loss": 0.8204, "step": 79070 }, { "epoch": 8.35, - "grad_norm": 25.442060470581055, + "grad_norm": 52.56435012817383, "learning_rate": 3.2970746646953215e-06, - "loss": 0.3577, + "loss": 0.9576, "step": 79080 }, { "epoch": 8.35, - "grad_norm": 34.71185302734375, + "grad_norm": 58.899356842041016, "learning_rate": 3.2949625092406807e-06, - "loss": 0.52, + "loss": 1.0691, "step": 79090 }, { "epoch": 8.35, - "grad_norm": 62.3248176574707, + "grad_norm": 36.682979583740234, "learning_rate": 3.292850353786039e-06, - "loss": 0.2637, + "loss": 0.468, "step": 79100 }, { "epoch": 8.35, - "grad_norm": 0.15374362468719482, + "grad_norm": 21.068134307861328, "learning_rate": 3.2907381983313974e-06, - "loss": 0.5619, + "loss": 0.9151, "step": 79110 }, { "epoch": 8.36, - "grad_norm": 8.33309268951416, + "grad_norm": 52.74040222167969, "learning_rate": 3.2886260428767557e-06, - "loss": 0.4852, + "loss": 0.5753, "step": 79120 }, { "epoch": 8.36, - "grad_norm": 0.8554682731628418, + "grad_norm": 12.958877563476562, "learning_rate": 3.286513887422115e-06, - "loss": 0.2989, + "loss": 0.8312, "step": 79130 }, { "epoch": 8.36, - "grad_norm": 0.2850620150566101, + "grad_norm": 34.36360549926758, "learning_rate": 3.2844017319674733e-06, - "loss": 0.2542, + "loss": 0.8052, "step": 79140 }, { "epoch": 8.36, - "grad_norm": 1.3779264688491821, + "grad_norm": 58.29954147338867, "learning_rate": 3.2822895765128316e-06, - "loss": 0.251, + "loss": 0.7842, "step": 79150 }, { "epoch": 8.36, - "grad_norm": 23.294593811035156, + "grad_norm": 40.56744384765625, "learning_rate": 3.28017742105819e-06, - "loss": 0.2521, + "loss": 0.8518, "step": 79160 }, { "epoch": 8.36, - "grad_norm": 0.712092936038971, + "grad_norm": 7.916815280914307, "learning_rate": 3.2780652656035488e-06, - "loss": 0.3364, + "loss": 0.6273, "step": 79170 }, { "epoch": 8.36, - "grad_norm": 19.82746124267578, + "grad_norm": 38.85192108154297, "learning_rate": 3.275953110148907e-06, - "loss": 0.5448, + "loss": 0.8784, "step": 79180 }, { "epoch": 8.36, - "grad_norm": 1.736867070198059, + "grad_norm": 7.764279365539551, "learning_rate": 3.2738409546942655e-06, - "loss": 0.2079, + "loss": 1.0226, "step": 79190 }, { "epoch": 8.36, - "grad_norm": 0.1744367927312851, + "grad_norm": 32.80656051635742, "learning_rate": 3.2717287992396243e-06, - "loss": 0.275, + "loss": 0.9004, "step": 79200 }, { "epoch": 8.37, - "grad_norm": 1.3534969091415405, + "grad_norm": 39.54289245605469, "learning_rate": 3.269616643784983e-06, - "loss": 0.3474, + "loss": 0.8542, "step": 79210 }, { "epoch": 8.37, - "grad_norm": 0.097984679043293, + "grad_norm": 19.229944229125977, "learning_rate": 3.2675044883303414e-06, - "loss": 0.2038, + "loss": 0.9688, "step": 79220 }, { "epoch": 8.37, - "grad_norm": 14.406039237976074, + "grad_norm": 22.557098388671875, "learning_rate": 3.2653923328756998e-06, - "loss": 0.3525, + "loss": 0.5922, "step": 79230 }, { "epoch": 8.37, - "grad_norm": 18.165082931518555, + "grad_norm": 29.248220443725586, "learning_rate": 3.263280177421058e-06, - "loss": 0.5353, + "loss": 1.1647, "step": 79240 }, { "epoch": 8.37, - "grad_norm": 0.9951454997062683, + "grad_norm": 43.432430267333984, "learning_rate": 3.2611680219664173e-06, - "loss": 0.3225, + "loss": 1.0347, "step": 79250 }, { "epoch": 8.37, - "grad_norm": 3.9812114238739014, + "grad_norm": 69.87034606933594, "learning_rate": 3.2590558665117757e-06, - "loss": 0.3264, + "loss": 0.7814, "step": 79260 }, { "epoch": 8.37, - "grad_norm": 4.699596405029297, + "grad_norm": 42.79716873168945, "learning_rate": 3.256943711057134e-06, - "loss": 0.6384, + "loss": 1.0408, "step": 79270 }, { "epoch": 8.37, - "grad_norm": 10.04299259185791, + "grad_norm": 37.44090270996094, "learning_rate": 3.254831555602493e-06, - "loss": 0.5098, + "loss": 0.8021, "step": 79280 }, { "epoch": 8.37, - "grad_norm": 13.666363716125488, + "grad_norm": 46.40593338012695, "learning_rate": 3.252719400147851e-06, - "loss": 0.4434, + "loss": 0.9237, "step": 79290 }, { "epoch": 8.37, - "grad_norm": 9.302613258361816, + "grad_norm": 38.11054992675781, "learning_rate": 3.2506072446932095e-06, - "loss": 0.4624, + "loss": 1.0394, "step": 79300 }, { "epoch": 8.38, - "grad_norm": 19.552915573120117, + "grad_norm": 39.13639831542969, "learning_rate": 3.248495089238568e-06, - "loss": 0.3386, + "loss": 0.8274, "step": 79310 }, { "epoch": 8.38, - "grad_norm": 8.290383338928223, + "grad_norm": 49.424522399902344, "learning_rate": 3.246382933783927e-06, - "loss": 0.2115, + "loss": 0.9241, "step": 79320 }, { "epoch": 8.38, - "grad_norm": 0.20990005135536194, + "grad_norm": 26.45357894897461, "learning_rate": 3.2442707783292854e-06, - "loss": 0.5406, + "loss": 0.7988, "step": 79330 }, { "epoch": 8.38, - "grad_norm": 6.461280822753906, + "grad_norm": 61.83951950073242, "learning_rate": 3.2421586228746438e-06, - "loss": 0.5959, + "loss": 0.9987, "step": 79340 }, { "epoch": 8.38, - "grad_norm": 19.36378288269043, + "grad_norm": 55.935218811035156, "learning_rate": 3.240046467420002e-06, - "loss": 0.4525, + "loss": 1.1904, "step": 79350 }, { "epoch": 8.38, - "grad_norm": 0.016218291595578194, + "grad_norm": 24.073959350585938, "learning_rate": 3.237934311965361e-06, - "loss": 0.3155, + "loss": 0.8315, "step": 79360 }, { "epoch": 8.38, - "grad_norm": 8.218461990356445, + "grad_norm": 21.63350486755371, "learning_rate": 3.2358221565107197e-06, - "loss": 0.3921, + "loss": 0.8101, "step": 79370 }, { "epoch": 8.38, - "grad_norm": 13.619447708129883, + "grad_norm": 56.64776611328125, "learning_rate": 3.233710001056078e-06, - "loss": 0.453, + "loss": 1.0673, "step": 79380 }, { "epoch": 8.38, - "grad_norm": 15.519943237304688, + "grad_norm": 29.559711456298828, "learning_rate": 3.2315978456014364e-06, - "loss": 0.4254, + "loss": 0.8119, "step": 79390 }, { "epoch": 8.39, - "grad_norm": 3.6093480587005615, + "grad_norm": 60.02869415283203, "learning_rate": 3.229485690146795e-06, - "loss": 0.2842, + "loss": 0.976, "step": 79400 }, { "epoch": 8.39, - "grad_norm": 7.175117015838623, + "grad_norm": 51.64737319946289, "learning_rate": 3.2273735346921535e-06, - "loss": 0.2838, + "loss": 0.9618, "step": 79410 }, { "epoch": 8.39, - "grad_norm": 20.33465003967285, + "grad_norm": 35.97504425048828, "learning_rate": 3.225261379237512e-06, - "loss": 0.2004, + "loss": 0.5113, "step": 79420 }, { "epoch": 8.39, - "grad_norm": 27.178646087646484, + "grad_norm": 60.90279769897461, "learning_rate": 3.2231492237828702e-06, - "loss": 0.3512, + "loss": 0.8681, "step": 79430 }, { "epoch": 8.39, - "grad_norm": 5.474806785583496, + "grad_norm": 47.37028121948242, "learning_rate": 3.2210370683282295e-06, - "loss": 0.2161, + "loss": 0.7059, "step": 79440 }, { "epoch": 8.39, - "grad_norm": 14.732224464416504, + "grad_norm": 29.87799644470215, "learning_rate": 3.218924912873588e-06, - "loss": 0.2134, + "loss": 0.6546, "step": 79450 }, { "epoch": 8.39, - "grad_norm": 60.89223861694336, + "grad_norm": 35.237728118896484, "learning_rate": 3.216812757418946e-06, - "loss": 0.4533, + "loss": 1.041, "step": 79460 }, { "epoch": 8.39, - "grad_norm": 13.092641830444336, + "grad_norm": 45.67866516113281, "learning_rate": 3.214700601964305e-06, - "loss": 0.3689, + "loss": 0.8661, "step": 79470 }, { "epoch": 8.39, - "grad_norm": 10.999640464782715, + "grad_norm": 61.0613899230957, "learning_rate": 3.2125884465096633e-06, - "loss": 0.5321, + "loss": 1.0106, "step": 79480 }, { "epoch": 8.39, - "grad_norm": 31.76641273498535, + "grad_norm": 58.81052017211914, "learning_rate": 3.210476291055022e-06, - "loss": 0.4557, + "loss": 0.8898, "step": 79490 }, { "epoch": 8.4, - "grad_norm": 11.879267692565918, + "grad_norm": 34.8759765625, "learning_rate": 3.2083641356003804e-06, - "loss": 0.2186, + "loss": 0.8135, "step": 79500 }, { "epoch": 8.4, - "grad_norm": 0.43350163102149963, + "grad_norm": 76.83100891113281, "learning_rate": 3.206251980145739e-06, - "loss": 0.2707, + "loss": 0.672, "step": 79510 }, { "epoch": 8.4, - "grad_norm": 3.3593671321868896, + "grad_norm": 44.75893783569336, "learning_rate": 3.2041398246910976e-06, - "loss": 0.3539, + "loss": 0.966, "step": 79520 }, { "epoch": 8.4, - "grad_norm": 2.4218571186065674, + "grad_norm": 23.93438720703125, "learning_rate": 3.202027669236456e-06, - "loss": 0.327, + "loss": 0.6826, "step": 79530 }, { "epoch": 8.4, - "grad_norm": 10.83432388305664, + "grad_norm": 25.63007354736328, "learning_rate": 3.1999155137818143e-06, - "loss": 0.2329, + "loss": 0.5934, "step": 79540 }, { "epoch": 8.4, - "grad_norm": 4.593348979949951, + "grad_norm": 53.40339660644531, "learning_rate": 3.1978033583271735e-06, - "loss": 0.2299, + "loss": 0.769, "step": 79550 }, { "epoch": 8.4, - "grad_norm": 16.10270881652832, + "grad_norm": 33.37877655029297, "learning_rate": 3.195691202872532e-06, - "loss": 0.0857, + "loss": 0.7546, "step": 79560 }, { "epoch": 8.4, - "grad_norm": 4.4603271484375, + "grad_norm": 53.41510009765625, "learning_rate": 3.19357904741789e-06, - "loss": 0.4372, + "loss": 0.8015, "step": 79570 }, { "epoch": 8.4, - "grad_norm": 19.209306716918945, + "grad_norm": 34.74237060546875, "learning_rate": 3.1914668919632485e-06, - "loss": 0.7172, + "loss": 1.0078, "step": 79580 }, { "epoch": 8.41, - "grad_norm": 2.7568271160125732, + "grad_norm": 21.462007522583008, "learning_rate": 3.1893547365086073e-06, - "loss": 0.3092, + "loss": 0.9401, "step": 79590 }, { "epoch": 8.41, - "grad_norm": 39.7327995300293, + "grad_norm": 48.74633026123047, "learning_rate": 3.187242581053966e-06, - "loss": 0.3799, + "loss": 1.0, "step": 79600 }, { "epoch": 8.41, - "grad_norm": 0.06718118488788605, + "grad_norm": 16.378704071044922, "learning_rate": 3.1851304255993245e-06, - "loss": 0.2084, + "loss": 0.5679, "step": 79610 }, { "epoch": 8.41, - "grad_norm": 8.751466751098633, + "grad_norm": 70.29997253417969, "learning_rate": 3.183018270144683e-06, - "loss": 0.2444, + "loss": 0.8476, "step": 79620 }, { "epoch": 8.41, - "grad_norm": 25.197551727294922, + "grad_norm": 73.07614135742188, "learning_rate": 3.1809061146900416e-06, - "loss": 0.3043, + "loss": 0.8417, "step": 79630 }, { "epoch": 8.41, - "grad_norm": 21.865856170654297, + "grad_norm": 33.14773941040039, "learning_rate": 3.1787939592354e-06, - "loss": 0.3127, + "loss": 0.7095, "step": 79640 }, { "epoch": 8.41, - "grad_norm": 15.502046585083008, + "grad_norm": 25.893644332885742, "learning_rate": 3.1766818037807583e-06, - "loss": 0.2718, + "loss": 0.9275, "step": 79650 }, { "epoch": 8.41, - "grad_norm": 0.26129359006881714, + "grad_norm": 48.69903564453125, "learning_rate": 3.1745696483261167e-06, - "loss": 0.4063, + "loss": 0.9915, "step": 79660 }, { "epoch": 8.41, - "grad_norm": 32.959571838378906, + "grad_norm": 53.715213775634766, "learning_rate": 3.172457492871476e-06, - "loss": 0.7304, + "loss": 0.8258, "step": 79670 }, { "epoch": 8.41, - "grad_norm": 0.4951886534690857, + "grad_norm": 45.26633834838867, "learning_rate": 3.1703453374168342e-06, - "loss": 0.2326, + "loss": 1.1601, "step": 79680 }, { "epoch": 8.42, - "grad_norm": 0.18942275643348694, + "grad_norm": 21.715656280517578, "learning_rate": 3.1682331819621926e-06, - "loss": 0.3688, + "loss": 0.9094, "step": 79690 }, { "epoch": 8.42, - "grad_norm": 18.85514259338379, + "grad_norm": 62.700870513916016, "learning_rate": 3.1661210265075513e-06, - "loss": 0.156, + "loss": 0.8379, "step": 79700 }, { "epoch": 8.42, - "grad_norm": 0.20122124254703522, + "grad_norm": 36.311729431152344, "learning_rate": 3.1640088710529097e-06, - "loss": 0.2047, + "loss": 0.9626, "step": 79710 }, { "epoch": 8.42, - "grad_norm": 2.2026429176330566, + "grad_norm": 43.56148910522461, "learning_rate": 3.1618967155982685e-06, - "loss": 0.3363, + "loss": 0.8542, "step": 79720 }, { "epoch": 8.42, - "grad_norm": 4.67410945892334, + "grad_norm": 29.0123348236084, "learning_rate": 3.159784560143627e-06, - "loss": 0.5384, + "loss": 1.109, "step": 79730 }, { "epoch": 8.42, - "grad_norm": 10.400311470031738, + "grad_norm": 47.96901321411133, "learning_rate": 3.1576724046889856e-06, - "loss": 0.3683, + "loss": 0.7012, "step": 79740 }, { "epoch": 8.42, - "grad_norm": 26.708663940429688, + "grad_norm": 46.668212890625, "learning_rate": 3.155560249234344e-06, - "loss": 0.222, + "loss": 0.7322, "step": 79750 }, { "epoch": 8.42, - "grad_norm": 39.16966247558594, + "grad_norm": 31.920198440551758, "learning_rate": 3.1534480937797023e-06, - "loss": 0.5455, + "loss": 1.0488, "step": 79760 }, { "epoch": 8.42, - "grad_norm": 0.06876721978187561, + "grad_norm": 43.67903137207031, "learning_rate": 3.1513359383250607e-06, - "loss": 0.1381, + "loss": 0.7232, "step": 79770 }, { "epoch": 8.43, - "grad_norm": 5.370824337005615, + "grad_norm": 22.100486755371094, "learning_rate": 3.14922378287042e-06, - "loss": 0.4087, + "loss": 0.9858, "step": 79780 }, { "epoch": 8.43, - "grad_norm": 21.9807071685791, + "grad_norm": 51.16670227050781, "learning_rate": 3.1471116274157782e-06, - "loss": 0.5746, + "loss": 1.1352, "step": 79790 }, { "epoch": 8.43, - "grad_norm": 16.45870590209961, + "grad_norm": 33.244972229003906, "learning_rate": 3.1449994719611366e-06, - "loss": 0.3784, + "loss": 0.8453, "step": 79800 }, { "epoch": 8.43, - "grad_norm": 12.615606307983398, + "grad_norm": 82.99275207519531, "learning_rate": 3.142887316506495e-06, - "loss": 0.3251, + "loss": 0.769, "step": 79810 }, { "epoch": 8.43, - "grad_norm": 14.356255531311035, + "grad_norm": 53.533084869384766, "learning_rate": 3.1407751610518537e-06, - "loss": 0.315, + "loss": 0.6434, "step": 79820 }, { "epoch": 8.43, - "grad_norm": 12.531862258911133, + "grad_norm": 46.07285690307617, "learning_rate": 3.138663005597212e-06, - "loss": 0.2908, + "loss": 0.9589, "step": 79830 }, { "epoch": 8.43, - "grad_norm": 27.455007553100586, + "grad_norm": 41.33053970336914, "learning_rate": 3.136550850142571e-06, - "loss": 0.4139, + "loss": 0.8248, "step": 79840 }, { "epoch": 8.43, - "grad_norm": 6.791125774383545, + "grad_norm": 55.67633819580078, "learning_rate": 3.1344386946879292e-06, - "loss": 0.267, + "loss": 0.8767, "step": 79850 }, { "epoch": 8.43, - "grad_norm": 33.315093994140625, + "grad_norm": 39.388545989990234, "learning_rate": 3.132326539233288e-06, - "loss": 0.4589, + "loss": 1.0785, "step": 79860 }, { "epoch": 8.43, - "grad_norm": 19.26753807067871, + "grad_norm": 47.67173767089844, "learning_rate": 3.1302143837786464e-06, - "loss": 0.6062, + "loss": 1.0103, "step": 79870 }, { "epoch": 8.44, - "grad_norm": 23.446041107177734, + "grad_norm": 51.67881393432617, "learning_rate": 3.1281022283240047e-06, - "loss": 0.2892, + "loss": 0.7054, "step": 79880 }, { "epoch": 8.44, - "grad_norm": 28.617740631103516, + "grad_norm": 42.2391471862793, "learning_rate": 3.125990072869363e-06, - "loss": 0.474, + "loss": 0.7586, "step": 79890 }, { "epoch": 8.44, - "grad_norm": 0.27949926257133484, + "grad_norm": 58.700401306152344, "learning_rate": 3.1238779174147223e-06, - "loss": 0.3409, + "loss": 0.9835, "step": 79900 }, { "epoch": 8.44, - "grad_norm": 17.142410278320312, + "grad_norm": 44.90876007080078, "learning_rate": 3.1217657619600806e-06, - "loss": 0.3837, + "loss": 0.77, "step": 79910 }, { "epoch": 8.44, - "grad_norm": 13.767513275146484, + "grad_norm": 51.05586624145508, "learning_rate": 3.119653606505439e-06, - "loss": 0.3204, + "loss": 0.9438, "step": 79920 }, { "epoch": 8.44, - "grad_norm": 9.05434799194336, + "grad_norm": 45.94837188720703, "learning_rate": 3.1175414510507978e-06, - "loss": 0.2164, + "loss": 0.7982, "step": 79930 }, { "epoch": 8.44, - "grad_norm": 4.41387414932251, + "grad_norm": 19.92125129699707, "learning_rate": 3.115429295596156e-06, - "loss": 0.1443, + "loss": 0.522, "step": 79940 }, { "epoch": 8.44, - "grad_norm": 10.817092895507812, + "grad_norm": 54.28466033935547, "learning_rate": 3.1133171401415145e-06, - "loss": 0.2105, + "loss": 0.9589, "step": 79950 }, { "epoch": 8.44, - "grad_norm": 7.767121315002441, + "grad_norm": 38.9552116394043, "learning_rate": 3.1112049846868732e-06, - "loss": 0.3411, + "loss": 0.7716, "step": 79960 }, { "epoch": 8.45, - "grad_norm": 32.32919692993164, + "grad_norm": 45.47779846191406, "learning_rate": 3.109092829232232e-06, - "loss": 0.5609, + "loss": 0.6894, "step": 79970 }, { "epoch": 8.45, - "grad_norm": 28.235721588134766, + "grad_norm": 36.56844711303711, "learning_rate": 3.1069806737775904e-06, - "loss": 0.4082, + "loss": 0.9412, "step": 79980 }, { "epoch": 8.45, - "grad_norm": 0.033015962690114975, + "grad_norm": 38.922550201416016, "learning_rate": 3.1048685183229487e-06, - "loss": 0.2496, + "loss": 1.0725, "step": 79990 }, { "epoch": 8.45, - "grad_norm": 20.36418914794922, + "grad_norm": 33.894283294677734, "learning_rate": 3.102756362868307e-06, - "loss": 0.3254, + "loss": 0.6538, "step": 80000 }, { "epoch": 8.45, - "grad_norm": 0.03758475184440613, + "grad_norm": 48.13404083251953, "learning_rate": 3.1006442074136663e-06, - "loss": 0.3222, + "loss": 0.8732, "step": 80010 }, { "epoch": 8.45, - "grad_norm": 7.468204498291016, + "grad_norm": 47.64911651611328, "learning_rate": 3.0985320519590246e-06, - "loss": 0.2441, + "loss": 1.1188, "step": 80020 }, { "epoch": 8.45, - "grad_norm": 35.378753662109375, + "grad_norm": 65.1678695678711, "learning_rate": 3.096419896504383e-06, - "loss": 0.5496, + "loss": 1.1355, "step": 80030 }, { "epoch": 8.45, - "grad_norm": 0.03810983896255493, + "grad_norm": 35.548675537109375, "learning_rate": 3.0943077410497414e-06, - "loss": 0.3233, + "loss": 1.1163, "step": 80040 }, { "epoch": 8.45, - "grad_norm": 13.755533218383789, + "grad_norm": 26.711933135986328, "learning_rate": 3.0921955855951e-06, - "loss": 0.3793, + "loss": 0.7801, "step": 80050 }, { "epoch": 8.45, - "grad_norm": 16.158735275268555, + "grad_norm": 24.69879913330078, "learning_rate": 3.0900834301404585e-06, - "loss": 0.3629, + "loss": 0.5759, "step": 80060 }, { "epoch": 8.46, - "grad_norm": 60.57056427001953, + "grad_norm": 59.60083770751953, "learning_rate": 3.087971274685817e-06, - "loss": 0.5213, + "loss": 0.7983, "step": 80070 }, { "epoch": 8.46, - "grad_norm": 0.06630846858024597, + "grad_norm": 62.45195770263672, "learning_rate": 3.0858591192311756e-06, - "loss": 0.3871, + "loss": 1.0497, "step": 80080 }, { "epoch": 8.46, - "grad_norm": 18.364885330200195, + "grad_norm": 33.06003952026367, "learning_rate": 3.0837469637765344e-06, - "loss": 0.4159, + "loss": 0.7062, "step": 80090 }, { "epoch": 8.46, - "grad_norm": 16.45197296142578, + "grad_norm": 37.704246520996094, "learning_rate": 3.0816348083218928e-06, - "loss": 0.2158, + "loss": 0.6159, "step": 80100 }, { "epoch": 8.46, - "grad_norm": 19.995132446289062, + "grad_norm": 17.902624130249023, "learning_rate": 3.079522652867251e-06, - "loss": 0.3137, + "loss": 0.9044, "step": 80110 }, { "epoch": 8.46, - "grad_norm": 7.832923412322998, + "grad_norm": 43.83658981323242, "learning_rate": 3.0774104974126095e-06, - "loss": 0.3453, + "loss": 0.6928, "step": 80120 }, { "epoch": 8.46, - "grad_norm": 14.837389945983887, + "grad_norm": 17.926746368408203, "learning_rate": 3.0752983419579687e-06, - "loss": 0.6092, + "loss": 0.6742, "step": 80130 }, { "epoch": 8.46, - "grad_norm": 0.023424232378602028, + "grad_norm": 34.1112174987793, "learning_rate": 3.073186186503327e-06, - "loss": 0.2749, + "loss": 0.8057, "step": 80140 }, { "epoch": 8.46, - "grad_norm": 15.802947044372559, + "grad_norm": 43.255001068115234, "learning_rate": 3.0710740310486854e-06, - "loss": 0.4501, + "loss": 0.6797, "step": 80150 }, { "epoch": 8.47, - "grad_norm": 11.11887264251709, + "grad_norm": 62.65372085571289, "learning_rate": 3.068961875594044e-06, - "loss": 0.3804, + "loss": 1.0936, "step": 80160 }, { "epoch": 8.47, - "grad_norm": 0.7625202536582947, + "grad_norm": 37.72698211669922, "learning_rate": 3.0668497201394025e-06, - "loss": 0.264, + "loss": 0.799, "step": 80170 }, { "epoch": 8.47, - "grad_norm": 23.513999938964844, + "grad_norm": 22.57618522644043, "learning_rate": 3.064737564684761e-06, - "loss": 0.2588, + "loss": 0.5945, "step": 80180 }, { "epoch": 8.47, - "grad_norm": 38.22673416137695, + "grad_norm": 53.19593048095703, "learning_rate": 3.0626254092301192e-06, - "loss": 0.432, + "loss": 0.7293, "step": 80190 }, { "epoch": 8.47, - "grad_norm": 18.492464065551758, + "grad_norm": 56.93365478515625, "learning_rate": 3.0605132537754784e-06, - "loss": 0.3984, + "loss": 0.7081, "step": 80200 }, { "epoch": 8.47, - "grad_norm": 0.24515646696090698, + "grad_norm": 68.48328399658203, "learning_rate": 3.0584010983208368e-06, - "loss": 0.431, + "loss": 1.2707, "step": 80210 }, { "epoch": 8.47, - "grad_norm": 11.931208610534668, + "grad_norm": 30.51512908935547, "learning_rate": 3.056288942866195e-06, - "loss": 0.2801, + "loss": 0.7892, "step": 80220 }, { "epoch": 8.47, - "grad_norm": 24.325809478759766, + "grad_norm": 45.07834243774414, "learning_rate": 3.0541767874115535e-06, - "loss": 0.4084, + "loss": 0.5816, "step": 80230 }, { "epoch": 8.47, - "grad_norm": 2.5159494876861572, + "grad_norm": 28.561372756958008, "learning_rate": 3.0520646319569127e-06, - "loss": 0.3046, + "loss": 0.8684, "step": 80240 }, { "epoch": 8.48, - "grad_norm": 1.2642579078674316, + "grad_norm": 52.17775344848633, "learning_rate": 3.049952476502271e-06, - "loss": 0.3486, + "loss": 0.8836, "step": 80250 }, { "epoch": 8.48, - "grad_norm": 12.5758638381958, + "grad_norm": 63.20729064941406, "learning_rate": 3.0478403210476294e-06, - "loss": 0.2609, + "loss": 0.8899, "step": 80260 }, { "epoch": 8.48, - "grad_norm": 14.442668914794922, + "grad_norm": 21.867063522338867, "learning_rate": 3.0457281655929878e-06, - "loss": 0.5241, + "loss": 0.8026, "step": 80270 }, { "epoch": 8.48, - "grad_norm": 2.2385222911834717, + "grad_norm": 27.497379302978516, "learning_rate": 3.0436160101383465e-06, - "loss": 0.1466, + "loss": 0.6238, "step": 80280 }, { "epoch": 8.48, - "grad_norm": 26.35820770263672, + "grad_norm": 37.143192291259766, "learning_rate": 3.041503854683705e-06, - "loss": 0.4248, + "loss": 1.0867, "step": 80290 }, { "epoch": 8.48, - "grad_norm": 13.801681518554688, + "grad_norm": 49.543601989746094, "learning_rate": 3.0393916992290633e-06, - "loss": 0.4165, + "loss": 0.6939, "step": 80300 }, { "epoch": 8.48, - "grad_norm": 0.07320734113454819, + "grad_norm": 21.509645462036133, "learning_rate": 3.0372795437744216e-06, - "loss": 0.4757, + "loss": 0.8571, "step": 80310 }, { "epoch": 8.48, - "grad_norm": 5.573448181152344, + "grad_norm": 37.465049743652344, "learning_rate": 3.035167388319781e-06, - "loss": 0.2505, + "loss": 0.9767, "step": 80320 }, { "epoch": 8.48, - "grad_norm": 23.229576110839844, + "grad_norm": 55.220680236816406, "learning_rate": 3.033055232865139e-06, - "loss": 0.1252, + "loss": 0.9203, "step": 80330 }, { "epoch": 8.48, - "grad_norm": 16.893821716308594, + "grad_norm": 47.14430618286133, "learning_rate": 3.0309430774104975e-06, - "loss": 0.6239, + "loss": 1.084, "step": 80340 }, { "epoch": 8.49, - "grad_norm": 8.799386024475098, + "grad_norm": 56.10096740722656, "learning_rate": 3.028830921955856e-06, - "loss": 0.7294, + "loss": 1.0827, "step": 80350 }, { "epoch": 8.49, - "grad_norm": 21.72884178161621, + "grad_norm": 66.8319091796875, "learning_rate": 3.026718766501215e-06, - "loss": 0.2717, + "loss": 0.9108, "step": 80360 }, { "epoch": 8.49, - "grad_norm": 0.38579797744750977, + "grad_norm": 38.189090728759766, "learning_rate": 3.0246066110465734e-06, - "loss": 0.5121, + "loss": 0.8773, "step": 80370 }, { "epoch": 8.49, - "grad_norm": 20.507173538208008, + "grad_norm": 39.32867431640625, "learning_rate": 3.022494455591932e-06, - "loss": 0.3451, + "loss": 0.5841, "step": 80380 }, { "epoch": 8.49, - "grad_norm": 13.132135391235352, + "grad_norm": 68.1041488647461, "learning_rate": 3.0203823001372906e-06, - "loss": 0.2293, + "loss": 1.0139, "step": 80390 }, { "epoch": 8.49, - "grad_norm": 17.778139114379883, + "grad_norm": 47.54877853393555, "learning_rate": 3.018270144682649e-06, - "loss": 0.2758, + "loss": 0.6958, "step": 80400 }, { "epoch": 8.49, - "grad_norm": 15.73603343963623, + "grad_norm": 52.6194953918457, "learning_rate": 3.0161579892280073e-06, - "loss": 0.3834, + "loss": 1.0105, "step": 80410 }, { "epoch": 8.49, - "grad_norm": 7.977198600769043, + "grad_norm": 29.31043815612793, "learning_rate": 3.0140458337733656e-06, - "loss": 0.2239, + "loss": 0.7902, "step": 80420 }, { "epoch": 8.49, - "grad_norm": 4.430424213409424, + "grad_norm": 44.80631637573242, "learning_rate": 3.011933678318725e-06, - "loss": 0.2283, + "loss": 0.814, "step": 80430 }, { "epoch": 8.5, - "grad_norm": 33.81959533691406, + "grad_norm": 33.350643157958984, "learning_rate": 3.009821522864083e-06, - "loss": 0.2681, + "loss": 1.0534, "step": 80440 }, { "epoch": 8.5, - "grad_norm": 53.13002395629883, + "grad_norm": 33.57275390625, "learning_rate": 3.0077093674094415e-06, - "loss": 0.1569, + "loss": 0.584, "step": 80450 }, { "epoch": 8.5, - "grad_norm": 47.13328170776367, + "grad_norm": 54.9034423828125, "learning_rate": 3.0055972119548e-06, - "loss": 0.4415, + "loss": 1.117, "step": 80460 }, { "epoch": 8.5, - "grad_norm": 0.023977937176823616, + "grad_norm": 44.18779373168945, "learning_rate": 3.0034850565001587e-06, - "loss": 0.2986, + "loss": 0.8769, "step": 80470 }, { "epoch": 8.5, - "grad_norm": 17.68944549560547, + "grad_norm": 51.21962356567383, "learning_rate": 3.0013729010455175e-06, - "loss": 0.3595, + "loss": 0.9388, "step": 80480 }, { "epoch": 8.5, - "grad_norm": 1.6277217864990234, + "grad_norm": 27.840166091918945, "learning_rate": 2.999260745590876e-06, - "loss": 0.3296, + "loss": 0.6786, "step": 80490 }, { "epoch": 8.5, - "grad_norm": 15.195438385009766, + "grad_norm": 22.41335678100586, "learning_rate": 2.997148590136234e-06, - "loss": 0.3619, + "loss": 1.0718, "step": 80500 }, { "epoch": 8.5, - "grad_norm": 4.216322898864746, + "grad_norm": 28.839763641357422, "learning_rate": 2.995036434681593e-06, - "loss": 0.1383, + "loss": 0.9114, "step": 80510 }, { "epoch": 8.5, - "grad_norm": 11.426739692687988, + "grad_norm": 46.46866226196289, "learning_rate": 2.9929242792269513e-06, - "loss": 0.3288, + "loss": 0.8311, "step": 80520 }, { "epoch": 8.5, - "grad_norm": 0.14315003156661987, + "grad_norm": 44.3682746887207, "learning_rate": 2.9908121237723097e-06, - "loss": 0.2947, + "loss": 1.2561, "step": 80530 }, { "epoch": 8.51, - "grad_norm": 12.036117553710938, + "grad_norm": 40.03361892700195, "learning_rate": 2.988699968317668e-06, - "loss": 0.4005, + "loss": 0.7655, "step": 80540 }, { "epoch": 8.51, - "grad_norm": 9.8843355178833, + "grad_norm": 19.683429718017578, "learning_rate": 2.9865878128630272e-06, - "loss": 0.2187, + "loss": 0.6712, "step": 80550 }, { "epoch": 8.51, - "grad_norm": 29.163667678833008, + "grad_norm": 68.28556060791016, "learning_rate": 2.9844756574083856e-06, - "loss": 0.3274, + "loss": 0.862, "step": 80560 }, { "epoch": 8.51, - "grad_norm": 26.179853439331055, + "grad_norm": 48.540340423583984, "learning_rate": 2.982363501953744e-06, - "loss": 0.4366, + "loss": 0.8931, "step": 80570 }, { "epoch": 8.51, - "grad_norm": 3.401503324508667, + "grad_norm": 27.903993606567383, "learning_rate": 2.9802513464991023e-06, - "loss": 0.2662, + "loss": 0.9204, "step": 80580 }, { "epoch": 8.51, - "grad_norm": 8.749801635742188, + "grad_norm": 24.820119857788086, "learning_rate": 2.978139191044461e-06, - "loss": 0.4464, + "loss": 0.9743, "step": 80590 }, { "epoch": 8.51, - "grad_norm": 28.206226348876953, + "grad_norm": 58.2739143371582, "learning_rate": 2.97602703558982e-06, - "loss": 0.5323, + "loss": 1.091, "step": 80600 }, { "epoch": 8.51, - "grad_norm": 27.518888473510742, + "grad_norm": 35.41468048095703, "learning_rate": 2.973914880135178e-06, - "loss": 0.1317, + "loss": 0.6782, "step": 80610 }, { "epoch": 8.51, - "grad_norm": 5.492154598236084, + "grad_norm": 44.861698150634766, "learning_rate": 2.971802724680537e-06, - "loss": 0.1764, + "loss": 0.9461, "step": 80620 }, { "epoch": 8.52, - "grad_norm": 7.263903617858887, + "grad_norm": 35.87763214111328, "learning_rate": 2.9696905692258953e-06, - "loss": 0.2691, + "loss": 0.9501, "step": 80630 }, { "epoch": 8.52, - "grad_norm": 2.57013201713562, + "grad_norm": 43.60661697387695, "learning_rate": 2.9675784137712537e-06, - "loss": 0.2052, + "loss": 0.6985, "step": 80640 }, { "epoch": 8.52, - "grad_norm": 0.058703791350126266, + "grad_norm": 19.319665908813477, "learning_rate": 2.965466258316612e-06, - "loss": 0.4201, + "loss": 0.6582, "step": 80650 }, { "epoch": 8.52, - "grad_norm": 0.03098973259329796, + "grad_norm": 35.378047943115234, "learning_rate": 2.9633541028619712e-06, - "loss": 0.4941, + "loss": 0.7083, "step": 80660 }, { "epoch": 8.52, - "grad_norm": 25.61510467529297, + "grad_norm": 31.722957611083984, "learning_rate": 2.9612419474073296e-06, - "loss": 0.1672, + "loss": 0.7045, "step": 80670 }, { "epoch": 8.52, - "grad_norm": 12.979034423828125, + "grad_norm": 29.36503791809082, "learning_rate": 2.959129791952688e-06, - "loss": 0.504, + "loss": 0.7629, "step": 80680 }, { "epoch": 8.52, - "grad_norm": 5.784929275512695, + "grad_norm": 45.516605377197266, "learning_rate": 2.9570176364980463e-06, - "loss": 0.2688, + "loss": 0.8885, "step": 80690 }, { "epoch": 8.52, - "grad_norm": 0.36384692788124084, + "grad_norm": 35.350345611572266, "learning_rate": 2.954905481043405e-06, - "loss": 0.2026, + "loss": 0.8468, "step": 80700 }, { "epoch": 8.52, - "grad_norm": 2.6797664165496826, + "grad_norm": 22.33269691467285, "learning_rate": 2.9527933255887634e-06, - "loss": 0.1952, + "loss": 0.8619, "step": 80710 }, { "epoch": 8.52, - "grad_norm": 192.5492706298828, + "grad_norm": 48.852272033691406, "learning_rate": 2.9506811701341222e-06, - "loss": 0.3898, + "loss": 0.9773, "step": 80720 }, { "epoch": 8.53, - "grad_norm": 0.019383830949664116, + "grad_norm": 40.177215576171875, "learning_rate": 2.9485690146794806e-06, - "loss": 0.1861, + "loss": 0.8101, "step": 80730 }, { "epoch": 8.53, - "grad_norm": 3.5086820125579834, + "grad_norm": 52.518272399902344, "learning_rate": 2.9464568592248394e-06, - "loss": 0.0333, + "loss": 0.699, "step": 80740 }, { "epoch": 8.53, - "grad_norm": 37.17664337158203, + "grad_norm": 48.53519821166992, "learning_rate": 2.9443447037701977e-06, - "loss": 0.4424, + "loss": 1.1009, "step": 80750 }, { "epoch": 8.53, - "grad_norm": 0.11363282054662704, + "grad_norm": 28.94816017150879, "learning_rate": 2.942232548315556e-06, - "loss": 0.323, + "loss": 0.8286, "step": 80760 }, { "epoch": 8.53, - "grad_norm": 8.301369667053223, + "grad_norm": 30.200634002685547, "learning_rate": 2.9401203928609144e-06, - "loss": 0.3816, + "loss": 0.9768, "step": 80770 }, { "epoch": 8.53, - "grad_norm": 25.062843322753906, + "grad_norm": 35.81803894042969, "learning_rate": 2.9380082374062736e-06, - "loss": 0.4709, + "loss": 0.9284, "step": 80780 }, { "epoch": 8.53, - "grad_norm": 0.45848286151885986, + "grad_norm": 10.828173637390137, "learning_rate": 2.935896081951632e-06, - "loss": 0.1722, + "loss": 0.8539, "step": 80790 }, { "epoch": 8.53, - "grad_norm": 26.009185791015625, + "grad_norm": 57.67094421386719, "learning_rate": 2.9337839264969903e-06, - "loss": 0.3575, + "loss": 0.8635, "step": 80800 }, { "epoch": 8.53, - "grad_norm": 7.156076908111572, + "grad_norm": 47.98212432861328, "learning_rate": 2.9316717710423487e-06, - "loss": 0.4443, + "loss": 0.952, "step": 80810 }, { "epoch": 8.54, - "grad_norm": 8.950041770935059, + "grad_norm": 33.4070930480957, "learning_rate": 2.9295596155877075e-06, - "loss": 0.2389, + "loss": 0.7017, "step": 80820 }, { "epoch": 8.54, - "grad_norm": 31.231292724609375, + "grad_norm": 18.48245620727539, "learning_rate": 2.927447460133066e-06, - "loss": 0.4583, + "loss": 0.8, "step": 80830 }, { "epoch": 8.54, - "grad_norm": 2.6301991939544678, + "grad_norm": 40.22176742553711, "learning_rate": 2.9253353046784246e-06, - "loss": 0.5329, + "loss": 0.9626, "step": 80840 }, { "epoch": 8.54, - "grad_norm": 0.4137554466724396, + "grad_norm": 50.110679626464844, "learning_rate": 2.9232231492237834e-06, - "loss": 0.4982, + "loss": 0.9529, "step": 80850 }, { "epoch": 8.54, - "grad_norm": 0.40836259722709656, + "grad_norm": 46.53584289550781, "learning_rate": 2.9211109937691417e-06, - "loss": 0.1836, + "loss": 0.8244, "step": 80860 }, { "epoch": 8.54, - "grad_norm": 0.6970272660255432, + "grad_norm": 45.04402542114258, "learning_rate": 2.9189988383145e-06, - "loss": 0.3172, + "loss": 0.8789, "step": 80870 }, { "epoch": 8.54, - "grad_norm": 6.028761386871338, + "grad_norm": 30.04508399963379, "learning_rate": 2.9168866828598585e-06, - "loss": 0.2261, + "loss": 1.0375, "step": 80880 }, { "epoch": 8.54, - "grad_norm": 4.1838812828063965, + "grad_norm": 53.52761459350586, "learning_rate": 2.9147745274052177e-06, - "loss": 0.2764, + "loss": 0.8171, "step": 80890 }, { "epoch": 8.54, - "grad_norm": 16.428789138793945, + "grad_norm": 26.954692840576172, "learning_rate": 2.912662371950576e-06, - "loss": 0.3286, + "loss": 0.6929, "step": 80900 }, { "epoch": 8.54, - "grad_norm": 5.733017921447754, + "grad_norm": 42.23881912231445, "learning_rate": 2.9105502164959344e-06, - "loss": 0.4348, + "loss": 0.8104, "step": 80910 }, { "epoch": 8.55, - "grad_norm": 30.976747512817383, + "grad_norm": 28.70828628540039, "learning_rate": 2.9084380610412927e-06, - "loss": 0.5219, + "loss": 1.0273, "step": 80920 }, { "epoch": 8.55, - "grad_norm": 0.03254938870668411, + "grad_norm": 28.849611282348633, "learning_rate": 2.9063259055866515e-06, - "loss": 0.3193, + "loss": 0.8398, "step": 80930 }, { "epoch": 8.55, - "grad_norm": 1.4462357759475708, + "grad_norm": 52.59336853027344, "learning_rate": 2.90421375013201e-06, - "loss": 0.3358, + "loss": 0.7561, "step": 80940 }, { "epoch": 8.55, - "grad_norm": 2.493384838104248, + "grad_norm": 25.66273307800293, "learning_rate": 2.902101594677368e-06, - "loss": 0.2441, + "loss": 0.5308, "step": 80950 }, { "epoch": 8.55, - "grad_norm": 14.298423767089844, + "grad_norm": 51.744815826416016, "learning_rate": 2.899989439222727e-06, - "loss": 0.4361, + "loss": 0.7028, "step": 80960 }, { "epoch": 8.55, - "grad_norm": 0.03451629728078842, + "grad_norm": 2.7362844944000244, "learning_rate": 2.8978772837680858e-06, - "loss": 0.3536, + "loss": 0.5492, "step": 80970 }, { "epoch": 8.55, - "grad_norm": 0.029147854074835777, + "grad_norm": 34.71867752075195, "learning_rate": 2.895765128313444e-06, - "loss": 0.327, + "loss": 0.7453, "step": 80980 }, { "epoch": 8.55, - "grad_norm": 9.265180587768555, + "grad_norm": 18.91152572631836, "learning_rate": 2.8936529728588025e-06, - "loss": 0.3063, + "loss": 0.8378, "step": 80990 }, { "epoch": 8.55, - "grad_norm": 9.608473777770996, + "grad_norm": 75.44181060791016, "learning_rate": 2.891540817404161e-06, - "loss": 0.2883, + "loss": 1.1055, "step": 81000 }, { "epoch": 8.56, - "grad_norm": 0.4733183681964874, + "grad_norm": 19.62738800048828, "learning_rate": 2.88942866194952e-06, - "loss": 0.3075, + "loss": 0.7639, "step": 81010 }, { "epoch": 8.56, - "grad_norm": 25.495027542114258, + "grad_norm": 49.17155456542969, "learning_rate": 2.8873165064948784e-06, - "loss": 0.4899, + "loss": 0.763, "step": 81020 }, { "epoch": 8.56, - "grad_norm": 12.26251220703125, + "grad_norm": 46.81233596801758, "learning_rate": 2.8852043510402367e-06, - "loss": 0.3263, + "loss": 0.8251, "step": 81030 }, { "epoch": 8.56, - "grad_norm": 0.09431151300668716, + "grad_norm": 36.43537902832031, "learning_rate": 2.8830921955855955e-06, - "loss": 0.3808, + "loss": 0.7973, "step": 81040 }, { "epoch": 8.56, - "grad_norm": 8.178723335266113, + "grad_norm": 67.75019073486328, "learning_rate": 2.880980040130954e-06, - "loss": 0.1131, + "loss": 0.8949, "step": 81050 }, { "epoch": 8.56, - "grad_norm": 13.879781723022461, + "grad_norm": 24.234560012817383, "learning_rate": 2.8788678846763122e-06, - "loss": 0.2672, + "loss": 0.7956, "step": 81060 }, { "epoch": 8.56, - "grad_norm": 2.6756436824798584, + "grad_norm": 69.85120391845703, "learning_rate": 2.8767557292216706e-06, - "loss": 0.269, + "loss": 0.7218, "step": 81070 }, { "epoch": 8.56, - "grad_norm": 1.9413800239562988, + "grad_norm": 32.34931182861328, "learning_rate": 2.87464357376703e-06, - "loss": 0.2671, + "loss": 0.871, "step": 81080 }, { "epoch": 8.56, - "grad_norm": 26.312734603881836, + "grad_norm": 16.21397590637207, "learning_rate": 2.872531418312388e-06, - "loss": 0.449, + "loss": 0.7365, "step": 81090 }, { "epoch": 8.56, - "grad_norm": 2.2664787769317627, + "grad_norm": 20.18290138244629, "learning_rate": 2.8704192628577465e-06, - "loss": 0.3455, + "loss": 0.6557, "step": 81100 }, { "epoch": 8.57, - "grad_norm": 8.211226463317871, + "grad_norm": 62.14852523803711, "learning_rate": 2.868307107403105e-06, - "loss": 0.4009, + "loss": 0.9897, "step": 81110 }, { "epoch": 8.57, - "grad_norm": 15.67139720916748, + "grad_norm": 27.576108932495117, "learning_rate": 2.866194951948464e-06, - "loss": 0.2535, + "loss": 0.7177, "step": 81120 }, { "epoch": 8.57, - "grad_norm": 29.531776428222656, + "grad_norm": 35.32375717163086, "learning_rate": 2.8640827964938224e-06, - "loss": 0.3372, + "loss": 1.1099, "step": 81130 }, { "epoch": 8.57, - "grad_norm": 22.08641242980957, + "grad_norm": 30.400774002075195, "learning_rate": 2.8619706410391808e-06, - "loss": 0.3904, + "loss": 0.5949, "step": 81140 }, { "epoch": 8.57, - "grad_norm": 15.695520401000977, + "grad_norm": 59.484867095947266, "learning_rate": 2.859858485584539e-06, - "loss": 0.5509, + "loss": 0.8949, "step": 81150 }, { "epoch": 8.57, - "grad_norm": 0.6885107159614563, + "grad_norm": 30.879182815551758, "learning_rate": 2.857746330129898e-06, - "loss": 0.1186, + "loss": 0.8339, "step": 81160 }, { "epoch": 8.57, - "grad_norm": 27.187484741210938, + "grad_norm": 49.10654830932617, "learning_rate": 2.8556341746752563e-06, - "loss": 0.3644, + "loss": 1.191, "step": 81170 }, { "epoch": 8.57, - "grad_norm": 27.261585235595703, + "grad_norm": 87.61954498291016, "learning_rate": 2.8535220192206146e-06, - "loss": 0.3633, + "loss": 0.8086, "step": 81180 }, { "epoch": 8.57, - "grad_norm": 19.927570343017578, + "grad_norm": 33.36327362060547, "learning_rate": 2.851409863765973e-06, - "loss": 0.4113, + "loss": 0.8983, "step": 81190 }, { "epoch": 8.58, - "grad_norm": 42.954837799072266, + "grad_norm": 49.13130187988281, "learning_rate": 2.849297708311332e-06, - "loss": 0.579, + "loss": 0.927, "step": 81200 }, { "epoch": 8.58, - "grad_norm": 15.652969360351562, + "grad_norm": 59.89314651489258, "learning_rate": 2.8471855528566905e-06, - "loss": 0.4775, + "loss": 1.015, "step": 81210 }, { "epoch": 8.58, - "grad_norm": 14.317572593688965, + "grad_norm": 48.04935073852539, "learning_rate": 2.845073397402049e-06, - "loss": 0.2114, + "loss": 0.9901, "step": 81220 }, { "epoch": 8.58, - "grad_norm": 0.051547691226005554, + "grad_norm": 18.740461349487305, "learning_rate": 2.8429612419474072e-06, - "loss": 0.589, + "loss": 0.7135, "step": 81230 }, { "epoch": 8.58, - "grad_norm": 11.035548210144043, + "grad_norm": 45.70999526977539, "learning_rate": 2.8408490864927664e-06, - "loss": 0.4467, + "loss": 0.9544, "step": 81240 }, { "epoch": 8.58, - "grad_norm": 29.277170181274414, + "grad_norm": 30.13225555419922, "learning_rate": 2.838736931038125e-06, - "loss": 0.2288, + "loss": 0.6391, "step": 81250 }, { "epoch": 8.58, - "grad_norm": 1.9940171241760254, + "grad_norm": 49.56454086303711, "learning_rate": 2.836624775583483e-06, - "loss": 0.3494, + "loss": 0.7116, "step": 81260 }, { "epoch": 8.58, - "grad_norm": 0.20393580198287964, + "grad_norm": 30.74262809753418, "learning_rate": 2.834512620128842e-06, - "loss": 0.2394, + "loss": 0.597, "step": 81270 }, { "epoch": 8.58, - "grad_norm": 36.974693298339844, + "grad_norm": 67.13160705566406, "learning_rate": 2.8324004646742003e-06, - "loss": 0.3692, + "loss": 1.0281, "step": 81280 }, { "epoch": 8.58, - "grad_norm": 0.13725580275058746, + "grad_norm": 62.86509323120117, "learning_rate": 2.8302883092195586e-06, - "loss": 0.1277, + "loss": 0.6759, "step": 81290 }, { "epoch": 8.59, - "grad_norm": 39.42811965942383, + "grad_norm": 43.70414733886719, "learning_rate": 2.828176153764917e-06, - "loss": 0.3556, + "loss": 0.719, "step": 81300 }, { "epoch": 8.59, - "grad_norm": 0.8655869960784912, + "grad_norm": 28.692462921142578, "learning_rate": 2.826063998310276e-06, - "loss": 0.5591, + "loss": 0.9826, "step": 81310 }, { "epoch": 8.59, - "grad_norm": 8.957398414611816, + "grad_norm": 18.509113311767578, "learning_rate": 2.8239518428556346e-06, - "loss": 0.0747, + "loss": 0.6572, "step": 81320 }, { "epoch": 8.59, - "grad_norm": 1.6600918769836426, + "grad_norm": 31.340599060058594, "learning_rate": 2.821839687400993e-06, - "loss": 0.2848, + "loss": 0.9757, "step": 81330 }, { "epoch": 8.59, - "grad_norm": 6.713106155395508, + "grad_norm": 46.52595520019531, "learning_rate": 2.8197275319463513e-06, - "loss": 0.3141, + "loss": 0.8438, "step": 81340 }, { "epoch": 8.59, - "grad_norm": 0.11742492765188217, + "grad_norm": 52.5865478515625, "learning_rate": 2.81761537649171e-06, - "loss": 0.3589, + "loss": 0.7831, "step": 81350 }, { "epoch": 8.59, - "grad_norm": 11.260807991027832, + "grad_norm": 32.44768142700195, "learning_rate": 2.815503221037069e-06, - "loss": 0.2723, + "loss": 0.86, "step": 81360 }, { "epoch": 8.59, - "grad_norm": 10.508626937866211, + "grad_norm": 22.883453369140625, "learning_rate": 2.813391065582427e-06, - "loss": 0.3452, + "loss": 0.9238, "step": 81370 }, { "epoch": 8.59, - "grad_norm": 22.935482025146484, + "grad_norm": 23.848331451416016, "learning_rate": 2.8112789101277855e-06, - "loss": 0.2336, + "loss": 0.7701, "step": 81380 }, { "epoch": 8.6, - "grad_norm": 0.012224847450852394, + "grad_norm": 20.761144638061523, "learning_rate": 2.8091667546731443e-06, - "loss": 0.5063, + "loss": 0.9731, "step": 81390 }, { "epoch": 8.6, - "grad_norm": 18.417646408081055, + "grad_norm": 36.58815383911133, "learning_rate": 2.8070545992185027e-06, - "loss": 0.4232, + "loss": 0.8028, "step": 81400 }, { "epoch": 8.6, - "grad_norm": 3.921426773071289, + "grad_norm": 10.802850723266602, "learning_rate": 2.804942443763861e-06, - "loss": 0.4089, + "loss": 0.801, "step": 81410 }, { "epoch": 8.6, - "grad_norm": 6.23084020614624, + "grad_norm": 33.32789611816406, "learning_rate": 2.8028302883092194e-06, - "loss": 0.6703, + "loss": 0.9737, "step": 81420 }, { "epoch": 8.6, - "grad_norm": 12.579513549804688, + "grad_norm": 41.57495880126953, "learning_rate": 2.8007181328545786e-06, - "loss": 0.3823, + "loss": 0.6093, "step": 81430 }, { "epoch": 8.6, - "grad_norm": 12.336804389953613, + "grad_norm": 67.91073608398438, "learning_rate": 2.798605977399937e-06, - "loss": 0.2414, + "loss": 0.7921, "step": 81440 }, { "epoch": 8.6, - "grad_norm": 0.2473171502351761, + "grad_norm": 22.082965850830078, "learning_rate": 2.7964938219452953e-06, - "loss": 0.2277, + "loss": 0.6838, "step": 81450 }, { "epoch": 8.6, - "grad_norm": 2.1629910469055176, + "grad_norm": 64.769287109375, "learning_rate": 2.7943816664906536e-06, - "loss": 0.3569, + "loss": 1.0787, "step": 81460 }, { "epoch": 8.6, - "grad_norm": 0.13074077665805817, + "grad_norm": 72.69258880615234, "learning_rate": 2.7922695110360124e-06, - "loss": 0.4566, + "loss": 0.9249, "step": 81470 }, { "epoch": 8.6, - "grad_norm": 13.852356910705566, + "grad_norm": 48.22627258300781, "learning_rate": 2.790157355581371e-06, - "loss": 0.4817, + "loss": 0.7652, "step": 81480 }, { "epoch": 8.61, - "grad_norm": 66.2428207397461, + "grad_norm": 62.69499588012695, "learning_rate": 2.7880452001267296e-06, - "loss": 0.3027, + "loss": 0.8209, "step": 81490 }, { "epoch": 8.61, - "grad_norm": 0.6229877471923828, + "grad_norm": 34.57086944580078, "learning_rate": 2.7859330446720883e-06, - "loss": 0.6459, + "loss": 0.7445, "step": 81500 }, { "epoch": 8.61, - "grad_norm": 10.180809020996094, + "grad_norm": 46.496185302734375, "learning_rate": 2.7838208892174467e-06, - "loss": 0.2133, + "loss": 0.6525, "step": 81510 }, { "epoch": 8.61, - "grad_norm": 8.558706283569336, + "grad_norm": 36.052978515625, "learning_rate": 2.781708733762805e-06, - "loss": 0.3225, + "loss": 1.1499, "step": 81520 }, { "epoch": 8.61, - "grad_norm": 10.543925285339355, + "grad_norm": 41.34526443481445, "learning_rate": 2.7795965783081634e-06, - "loss": 0.2168, + "loss": 0.7159, "step": 81530 }, { "epoch": 8.61, - "grad_norm": 17.268043518066406, + "grad_norm": 53.32085037231445, "learning_rate": 2.7774844228535226e-06, - "loss": 0.1515, + "loss": 0.8295, "step": 81540 }, { "epoch": 8.61, - "grad_norm": 5.337953567504883, + "grad_norm": 52.578086853027344, "learning_rate": 2.775372267398881e-06, - "loss": 0.331, + "loss": 0.7849, "step": 81550 }, { "epoch": 8.61, - "grad_norm": 24.673274993896484, + "grad_norm": 13.288585662841797, "learning_rate": 2.7732601119442393e-06, - "loss": 0.3361, + "loss": 0.8715, "step": 81560 }, { "epoch": 8.61, - "grad_norm": 0.05479868873953819, + "grad_norm": 28.07908058166504, "learning_rate": 2.7711479564895977e-06, - "loss": 0.4601, + "loss": 1.0475, "step": 81570 }, { "epoch": 8.62, - "grad_norm": 41.845314025878906, + "grad_norm": 47.335391998291016, "learning_rate": 2.7690358010349565e-06, - "loss": 0.1865, + "loss": 0.913, "step": 81580 }, { "epoch": 8.62, - "grad_norm": 0.8897597193717957, + "grad_norm": 58.974483489990234, "learning_rate": 2.766923645580315e-06, - "loss": 0.1448, + "loss": 0.6484, "step": 81590 }, { "epoch": 8.62, - "grad_norm": 8.558199882507324, + "grad_norm": 49.08378982543945, "learning_rate": 2.7648114901256736e-06, - "loss": 0.2562, + "loss": 0.623, "step": 81600 }, { "epoch": 8.62, - "grad_norm": 0.546162486076355, + "grad_norm": 15.560876846313477, "learning_rate": 2.762699334671032e-06, - "loss": 0.5299, + "loss": 0.9278, "step": 81610 }, { "epoch": 8.62, - "grad_norm": 0.0530855618417263, + "grad_norm": 48.661800384521484, "learning_rate": 2.7605871792163907e-06, - "loss": 0.1149, + "loss": 0.9945, "step": 81620 }, { "epoch": 8.62, - "grad_norm": 15.526700019836426, + "grad_norm": 21.55033302307129, "learning_rate": 2.758475023761749e-06, - "loss": 0.4472, + "loss": 1.0962, "step": 81630 }, { "epoch": 8.62, - "grad_norm": 3.8556764125823975, + "grad_norm": 24.742761611938477, "learning_rate": 2.7563628683071074e-06, - "loss": 0.3062, + "loss": 0.7155, "step": 81640 }, { "epoch": 8.62, - "grad_norm": 8.779288291931152, + "grad_norm": 50.68307876586914, "learning_rate": 2.754250712852466e-06, - "loss": 0.3901, + "loss": 0.9324, "step": 81650 }, { "epoch": 8.62, - "grad_norm": 14.964906692504883, + "grad_norm": 24.035947799682617, "learning_rate": 2.752138557397825e-06, - "loss": 0.7744, + "loss": 0.9031, "step": 81660 }, { "epoch": 8.62, - "grad_norm": 27.413728713989258, + "grad_norm": 55.24799346923828, "learning_rate": 2.7500264019431833e-06, - "loss": 0.3392, + "loss": 1.2171, "step": 81670 }, { "epoch": 8.63, - "grad_norm": 7.296975135803223, + "grad_norm": 23.012041091918945, "learning_rate": 2.7479142464885417e-06, - "loss": 0.3441, + "loss": 0.9172, "step": 81680 }, { "epoch": 8.63, - "grad_norm": 6.827291011810303, + "grad_norm": 64.19342803955078, "learning_rate": 2.7458020910339e-06, - "loss": 0.3176, + "loss": 1.0237, "step": 81690 }, { "epoch": 8.63, - "grad_norm": 1.2254585027694702, + "grad_norm": 44.20667266845703, "learning_rate": 2.743689935579259e-06, - "loss": 0.3504, + "loss": 1.1562, "step": 81700 }, { "epoch": 8.63, - "grad_norm": 25.402082443237305, + "grad_norm": 52.47505569458008, "learning_rate": 2.741577780124617e-06, - "loss": 0.3336, + "loss": 0.9022, "step": 81710 }, { "epoch": 8.63, - "grad_norm": 0.13340134918689728, + "grad_norm": 43.46904754638672, "learning_rate": 2.739465624669976e-06, - "loss": 0.2126, + "loss": 0.8473, "step": 81720 }, { "epoch": 8.63, - "grad_norm": 3.4232559204101562, + "grad_norm": 33.217281341552734, "learning_rate": 2.7373534692153347e-06, - "loss": 0.7036, + "loss": 0.9353, "step": 81730 }, { "epoch": 8.63, - "grad_norm": 23.148765563964844, + "grad_norm": 42.27854919433594, "learning_rate": 2.735241313760693e-06, - "loss": 0.428, + "loss": 1.0151, "step": 81740 }, { "epoch": 8.63, - "grad_norm": 11.318456649780273, + "grad_norm": 48.1832389831543, "learning_rate": 2.7331291583060515e-06, - "loss": 0.2635, + "loss": 0.7713, "step": 81750 }, { "epoch": 8.63, - "grad_norm": 12.612737655639648, + "grad_norm": 41.316280364990234, "learning_rate": 2.73101700285141e-06, - "loss": 0.398, + "loss": 0.9426, "step": 81760 }, { "epoch": 8.64, - "grad_norm": 0.029821239411830902, + "grad_norm": 23.11927604675293, "learning_rate": 2.728904847396769e-06, - "loss": 0.2263, + "loss": 0.9463, "step": 81770 }, { "epoch": 8.64, - "grad_norm": 0.4214801490306854, + "grad_norm": 23.617633819580078, "learning_rate": 2.7267926919421274e-06, - "loss": 0.2051, + "loss": 0.671, "step": 81780 }, { "epoch": 8.64, - "grad_norm": 12.967470169067383, + "grad_norm": 41.53396224975586, "learning_rate": 2.7246805364874857e-06, - "loss": 0.3424, + "loss": 1.0579, "step": 81790 }, { "epoch": 8.64, - "grad_norm": 54.920772552490234, + "grad_norm": 59.522239685058594, "learning_rate": 2.722568381032844e-06, - "loss": 0.2671, + "loss": 0.9195, "step": 81800 }, { "epoch": 8.64, - "grad_norm": 17.022857666015625, + "grad_norm": 38.834171295166016, "learning_rate": 2.720456225578203e-06, - "loss": 0.4628, + "loss": 1.271, "step": 81810 }, { "epoch": 8.64, - "grad_norm": 12.010143280029297, + "grad_norm": 30.983516693115234, "learning_rate": 2.7183440701235612e-06, - "loss": 0.3694, + "loss": 0.8771, "step": 81820 }, { "epoch": 8.64, - "grad_norm": 8.984530448913574, + "grad_norm": 39.54784393310547, "learning_rate": 2.7162319146689196e-06, - "loss": 0.2785, + "loss": 0.8919, "step": 81830 }, { "epoch": 8.64, - "grad_norm": 0.09445173293352127, + "grad_norm": 27.502920150756836, "learning_rate": 2.7141197592142784e-06, - "loss": 0.422, + "loss": 0.7834, "step": 81840 }, { "epoch": 8.64, - "grad_norm": 13.577309608459473, + "grad_norm": 38.269447326660156, "learning_rate": 2.712007603759637e-06, - "loss": 0.3674, + "loss": 0.8363, "step": 81850 }, { "epoch": 8.65, - "grad_norm": 0.08954726159572601, + "grad_norm": 5.275543212890625, "learning_rate": 2.7098954483049955e-06, - "loss": 0.0812, + "loss": 0.4605, "step": 81860 }, { "epoch": 8.65, - "grad_norm": 1.0933270454406738, + "grad_norm": 29.77422332763672, "learning_rate": 2.707783292850354e-06, - "loss": 0.4259, + "loss": 0.926, "step": 81870 }, { "epoch": 8.65, - "grad_norm": 26.77652931213379, + "grad_norm": 17.552230834960938, "learning_rate": 2.705671137395712e-06, - "loss": 0.0747, + "loss": 0.6777, "step": 81880 }, { "epoch": 8.65, - "grad_norm": 0.06734858453273773, + "grad_norm": 16.764999389648438, "learning_rate": 2.7035589819410714e-06, - "loss": 0.25, + "loss": 0.9265, "step": 81890 }, { "epoch": 8.65, - "grad_norm": 1.8032286167144775, + "grad_norm": 30.951107025146484, "learning_rate": 2.7014468264864298e-06, - "loss": 0.3597, + "loss": 0.8487, "step": 81900 }, { "epoch": 8.65, - "grad_norm": 20.52814292907715, + "grad_norm": 25.146408081054688, "learning_rate": 2.699334671031788e-06, - "loss": 0.3771, + "loss": 1.2638, "step": 81910 }, { "epoch": 8.65, - "grad_norm": 20.577455520629883, + "grad_norm": 58.22303009033203, "learning_rate": 2.6972225155771465e-06, - "loss": 0.6117, + "loss": 1.2071, "step": 81920 }, { "epoch": 8.65, - "grad_norm": 0.2619817852973938, + "grad_norm": 52.78644561767578, "learning_rate": 2.6951103601225052e-06, - "loss": 0.1592, + "loss": 0.7076, "step": 81930 }, { "epoch": 8.65, - "grad_norm": 4.982455730438232, + "grad_norm": 28.23102378845215, "learning_rate": 2.6929982046678636e-06, - "loss": 0.4936, + "loss": 1.0098, "step": 81940 }, { "epoch": 8.65, - "grad_norm": 76.38843536376953, + "grad_norm": 24.61177635192871, "learning_rate": 2.690886049213222e-06, - "loss": 0.3115, + "loss": 1.1655, "step": 81950 }, { "epoch": 8.66, - "grad_norm": 11.64350700378418, + "grad_norm": 30.6015567779541, "learning_rate": 2.688773893758581e-06, - "loss": 0.372, + "loss": 0.5899, "step": 81960 }, { "epoch": 8.66, - "grad_norm": 16.750640869140625, + "grad_norm": 39.45302963256836, "learning_rate": 2.6866617383039395e-06, - "loss": 0.3226, + "loss": 0.9206, "step": 81970 }, { "epoch": 8.66, - "grad_norm": 2.0245511531829834, + "grad_norm": 52.26906204223633, "learning_rate": 2.684549582849298e-06, - "loss": 0.5058, + "loss": 0.6543, "step": 81980 }, { "epoch": 8.66, - "grad_norm": 13.043594360351562, + "grad_norm": 44.03887176513672, "learning_rate": 2.6824374273946562e-06, - "loss": 0.2799, + "loss": 1.0834, "step": 81990 }, { "epoch": 8.66, - "grad_norm": 0.05171246454119682, + "grad_norm": 46.072181701660156, "learning_rate": 2.6803252719400154e-06, - "loss": 0.5219, + "loss": 0.8042, "step": 82000 }, { "epoch": 8.66, - "grad_norm": 13.34732437133789, + "grad_norm": 62.938716888427734, "learning_rate": 2.6782131164853738e-06, - "loss": 0.391, + "loss": 0.7509, "step": 82010 }, { "epoch": 8.66, - "grad_norm": 6.2642011642456055, + "grad_norm": 39.53767776489258, "learning_rate": 2.676100961030732e-06, - "loss": 0.165, + "loss": 0.8262, "step": 82020 }, { "epoch": 8.66, - "grad_norm": 27.274456024169922, + "grad_norm": 39.81260299682617, "learning_rate": 2.6739888055760905e-06, - "loss": 0.3861, + "loss": 1.0933, "step": 82030 }, { "epoch": 8.66, - "grad_norm": 18.039770126342773, + "grad_norm": 80.00574493408203, "learning_rate": 2.6718766501214493e-06, - "loss": 0.4794, + "loss": 0.868, "step": 82040 }, { "epoch": 8.67, - "grad_norm": 11.09554672241211, + "grad_norm": 40.13013458251953, "learning_rate": 2.6697644946668076e-06, - "loss": 0.2209, + "loss": 1.2656, "step": 82050 }, { "epoch": 8.67, - "grad_norm": 0.7991890907287598, + "grad_norm": 49.03324508666992, "learning_rate": 2.667652339212166e-06, - "loss": 0.5006, + "loss": 0.9316, "step": 82060 }, { "epoch": 8.67, - "grad_norm": 22.828615188598633, + "grad_norm": 34.17824172973633, "learning_rate": 2.6655401837575243e-06, - "loss": 0.4994, + "loss": 1.1175, "step": 82070 }, { "epoch": 8.67, - "grad_norm": 8.506583213806152, + "grad_norm": 24.575401306152344, "learning_rate": 2.6634280283028835e-06, - "loss": 0.2327, + "loss": 0.7658, "step": 82080 }, { "epoch": 8.67, - "grad_norm": 15.64758586883545, + "grad_norm": 37.342689514160156, "learning_rate": 2.661315872848242e-06, - "loss": 0.3573, + "loss": 0.8035, "step": 82090 }, { "epoch": 8.67, - "grad_norm": 9.839761734008789, + "grad_norm": 45.32169723510742, "learning_rate": 2.6592037173936002e-06, - "loss": 0.4217, + "loss": 0.7743, "step": 82100 }, { "epoch": 8.67, - "grad_norm": 30.8160343170166, + "grad_norm": 37.80854797363281, "learning_rate": 2.6570915619389586e-06, - "loss": 0.3973, + "loss": 0.6777, "step": 82110 }, { "epoch": 8.67, - "grad_norm": 0.1315319538116455, + "grad_norm": 43.956233978271484, "learning_rate": 2.654979406484318e-06, - "loss": 0.3747, + "loss": 0.8483, "step": 82120 }, { "epoch": 8.67, - "grad_norm": 7.586221694946289, + "grad_norm": 41.57847595214844, "learning_rate": 2.652867251029676e-06, - "loss": 0.2793, + "loss": 0.8193, "step": 82130 }, { "epoch": 8.67, - "grad_norm": 9.048646926879883, + "grad_norm": 57.397300720214844, "learning_rate": 2.6507550955750345e-06, - "loss": 0.4602, + "loss": 1.0367, "step": 82140 }, { "epoch": 8.68, - "grad_norm": 10.801080703735352, + "grad_norm": 22.148672103881836, "learning_rate": 2.648642940120393e-06, - "loss": 0.3802, + "loss": 0.6207, "step": 82150 }, { "epoch": 8.68, - "grad_norm": 32.06532287597656, + "grad_norm": 48.28506851196289, "learning_rate": 2.6465307846657517e-06, - "loss": 0.3284, + "loss": 0.9759, "step": 82160 }, { "epoch": 8.68, - "grad_norm": 28.240514755249023, + "grad_norm": 20.81433868408203, "learning_rate": 2.64441862921111e-06, - "loss": 0.4285, + "loss": 0.9485, "step": 82170 }, { "epoch": 8.68, - "grad_norm": 13.628557205200195, + "grad_norm": 50.335472106933594, "learning_rate": 2.6423064737564684e-06, - "loss": 0.4482, + "loss": 0.9915, "step": 82180 }, { "epoch": 8.68, - "grad_norm": 20.736217498779297, + "grad_norm": 41.7040901184082, "learning_rate": 2.6401943183018276e-06, - "loss": 0.2838, + "loss": 0.8948, "step": 82190 }, { "epoch": 8.68, - "grad_norm": 6.839109420776367, + "grad_norm": 54.87863540649414, "learning_rate": 2.638082162847186e-06, - "loss": 0.3947, + "loss": 1.1433, "step": 82200 }, { "epoch": 8.68, - "grad_norm": 10.780982971191406, + "grad_norm": 53.83258056640625, "learning_rate": 2.6359700073925443e-06, - "loss": 0.4214, + "loss": 0.7373, "step": 82210 }, { "epoch": 8.68, - "grad_norm": 3.7728328704833984, + "grad_norm": 10.813943862915039, "learning_rate": 2.6338578519379026e-06, - "loss": 0.2648, + "loss": 1.1918, "step": 82220 }, { "epoch": 8.68, - "grad_norm": 2.762573003768921, + "grad_norm": 52.88519287109375, "learning_rate": 2.6317456964832614e-06, - "loss": 0.3783, + "loss": 0.8549, "step": 82230 }, { "epoch": 8.69, - "grad_norm": 36.62150192260742, + "grad_norm": 43.724342346191406, "learning_rate": 2.62963354102862e-06, - "loss": 0.3508, + "loss": 0.8372, "step": 82240 }, { "epoch": 8.69, - "grad_norm": 17.16111946105957, + "grad_norm": 49.992218017578125, "learning_rate": 2.6275213855739785e-06, - "loss": 0.6217, + "loss": 0.9606, "step": 82250 }, { "epoch": 8.69, - "grad_norm": 0.6596161723136902, + "grad_norm": 31.2923583984375, "learning_rate": 2.625409230119337e-06, - "loss": 0.2263, + "loss": 0.7676, "step": 82260 }, { "epoch": 8.69, - "grad_norm": 5.864706039428711, + "grad_norm": 47.94187927246094, "learning_rate": 2.6232970746646957e-06, - "loss": 0.3207, + "loss": 0.9424, "step": 82270 }, { "epoch": 8.69, - "grad_norm": 5.3495917320251465, + "grad_norm": 26.859724044799805, "learning_rate": 2.621184919210054e-06, - "loss": 0.2661, + "loss": 0.6552, "step": 82280 }, { "epoch": 8.69, - "grad_norm": 12.67839527130127, + "grad_norm": 31.77425193786621, "learning_rate": 2.6190727637554124e-06, - "loss": 0.2974, + "loss": 0.6549, "step": 82290 }, { "epoch": 8.69, - "grad_norm": 29.252084732055664, + "grad_norm": 58.26593017578125, "learning_rate": 2.6169606083007707e-06, - "loss": 0.6023, + "loss": 0.9281, "step": 82300 }, { "epoch": 8.69, - "grad_norm": 0.08861314505338669, + "grad_norm": 47.469322204589844, "learning_rate": 2.61484845284613e-06, - "loss": 0.3454, + "loss": 0.7324, "step": 82310 }, { "epoch": 8.69, - "grad_norm": 6.128052711486816, + "grad_norm": 36.99312210083008, "learning_rate": 2.6127362973914883e-06, - "loss": 0.1932, + "loss": 0.9828, "step": 82320 }, { "epoch": 8.69, - "grad_norm": 9.551315307617188, + "grad_norm": 35.04888153076172, "learning_rate": 2.6106241419368467e-06, - "loss": 0.1749, + "loss": 0.7185, "step": 82330 }, { "epoch": 8.7, - "grad_norm": 17.36202049255371, + "grad_norm": 58.4378662109375, "learning_rate": 2.608511986482205e-06, - "loss": 0.3621, + "loss": 0.9753, "step": 82340 }, { "epoch": 8.7, - "grad_norm": 3.078214406967163, + "grad_norm": 45.51813888549805, "learning_rate": 2.606399831027564e-06, - "loss": 0.1509, + "loss": 0.8826, "step": 82350 }, { "epoch": 8.7, - "grad_norm": 19.265419006347656, + "grad_norm": 32.858707427978516, "learning_rate": 2.6042876755729226e-06, - "loss": 0.2605, + "loss": 0.8003, "step": 82360 }, { "epoch": 8.7, - "grad_norm": 20.873523712158203, + "grad_norm": 48.062469482421875, "learning_rate": 2.602175520118281e-06, - "loss": 0.4705, + "loss": 0.9747, "step": 82370 }, { "epoch": 8.7, - "grad_norm": 24.104400634765625, + "grad_norm": 38.1889762878418, "learning_rate": 2.6000633646636397e-06, - "loss": 0.1891, + "loss": 0.9301, "step": 82380 }, { "epoch": 8.7, - "grad_norm": 26.220170974731445, + "grad_norm": 26.52532386779785, "learning_rate": 2.597951209208998e-06, - "loss": 0.2935, + "loss": 0.8097, "step": 82390 }, { "epoch": 8.7, - "grad_norm": 20.27950668334961, + "grad_norm": 40.209537506103516, "learning_rate": 2.5958390537543564e-06, - "loss": 0.3054, + "loss": 0.6613, "step": 82400 }, { "epoch": 8.7, - "grad_norm": 12.013838768005371, + "grad_norm": 32.17326736450195, "learning_rate": 2.5937268982997148e-06, - "loss": 0.3976, + "loss": 0.6145, "step": 82410 }, { "epoch": 8.7, - "grad_norm": 1.2132481336593628, + "grad_norm": 20.750032424926758, "learning_rate": 2.591614742845074e-06, - "loss": 0.3111, + "loss": 0.6493, "step": 82420 }, { "epoch": 8.71, - "grad_norm": 0.7984685897827148, + "grad_norm": 39.666419982910156, "learning_rate": 2.5895025873904323e-06, - "loss": 0.5311, + "loss": 0.7546, "step": 82430 }, { "epoch": 8.71, - "grad_norm": 24.443065643310547, + "grad_norm": 40.35537338256836, "learning_rate": 2.5873904319357907e-06, - "loss": 0.5937, + "loss": 0.8704, "step": 82440 }, { "epoch": 8.71, - "grad_norm": 1.445335865020752, + "grad_norm": 41.06154251098633, "learning_rate": 2.585278276481149e-06, - "loss": 0.3969, + "loss": 0.786, "step": 82450 }, { "epoch": 8.71, - "grad_norm": 0.13641604781150818, + "grad_norm": 51.26339340209961, "learning_rate": 2.583166121026508e-06, - "loss": 0.1672, + "loss": 0.8667, "step": 82460 }, { "epoch": 8.71, - "grad_norm": 0.38238391280174255, + "grad_norm": 16.526552200317383, "learning_rate": 2.581053965571866e-06, - "loss": 0.118, + "loss": 0.5226, "step": 82470 }, { "epoch": 8.71, - "grad_norm": 17.978591918945312, + "grad_norm": 27.107059478759766, "learning_rate": 2.578941810117225e-06, - "loss": 0.1996, + "loss": 0.8082, "step": 82480 }, { "epoch": 8.71, - "grad_norm": 4.141030311584473, + "grad_norm": 31.83437728881836, "learning_rate": 2.5768296546625833e-06, - "loss": 0.4517, + "loss": 0.9308, "step": 82490 }, { "epoch": 8.71, - "grad_norm": 14.024886131286621, + "grad_norm": 60.57825469970703, "learning_rate": 2.574717499207942e-06, - "loss": 0.3739, + "loss": 0.932, "step": 82500 }, { "epoch": 8.71, - "grad_norm": 0.09269674867391586, + "grad_norm": 13.725010871887207, "learning_rate": 2.5726053437533004e-06, - "loss": 0.2851, + "loss": 0.7087, "step": 82510 }, { "epoch": 8.71, - "grad_norm": 15.043489456176758, + "grad_norm": 60.63288116455078, "learning_rate": 2.570493188298659e-06, - "loss": 0.3154, + "loss": 0.7351, "step": 82520 }, { "epoch": 8.72, - "grad_norm": 0.11016478389501572, + "grad_norm": 41.68309783935547, "learning_rate": 2.568381032844017e-06, - "loss": 0.4693, + "loss": 1.0353, "step": 82530 }, { "epoch": 8.72, - "grad_norm": 14.064167976379395, + "grad_norm": 27.502431869506836, "learning_rate": 2.5662688773893764e-06, - "loss": 0.3607, + "loss": 0.925, "step": 82540 }, { "epoch": 8.72, - "grad_norm": 0.19580493867397308, + "grad_norm": 42.7664680480957, "learning_rate": 2.5641567219347347e-06, - "loss": 0.5372, + "loss": 0.7947, "step": 82550 }, { "epoch": 8.72, - "grad_norm": 0.14322054386138916, + "grad_norm": 54.144981384277344, "learning_rate": 2.562044566480093e-06, - "loss": 0.228, + "loss": 0.6425, "step": 82560 }, { "epoch": 8.72, - "grad_norm": 11.111639022827148, + "grad_norm": 36.85248565673828, "learning_rate": 2.5599324110254514e-06, - "loss": 0.5465, + "loss": 0.8914, "step": 82570 }, { "epoch": 8.72, - "grad_norm": 11.19228458404541, + "grad_norm": 34.36338806152344, "learning_rate": 2.55782025557081e-06, - "loss": 0.2464, + "loss": 0.7638, "step": 82580 }, { "epoch": 8.72, - "grad_norm": 4.82037878036499, + "grad_norm": 61.05381774902344, "learning_rate": 2.555708100116169e-06, - "loss": 0.2382, + "loss": 0.8827, "step": 82590 }, { "epoch": 8.72, - "grad_norm": 6.770522594451904, + "grad_norm": 20.01201057434082, "learning_rate": 2.5535959446615273e-06, - "loss": 0.2095, + "loss": 0.8811, "step": 82600 }, { "epoch": 8.72, - "grad_norm": 6.436943054199219, + "grad_norm": 3.807539463043213, "learning_rate": 2.551483789206886e-06, - "loss": 0.4203, + "loss": 0.8817, "step": 82610 }, { "epoch": 8.73, - "grad_norm": 16.21567153930664, + "grad_norm": 52.17580032348633, "learning_rate": 2.5493716337522445e-06, - "loss": 0.2436, + "loss": 0.7592, "step": 82620 }, { "epoch": 8.73, - "grad_norm": 29.5321102142334, + "grad_norm": 20.660079956054688, "learning_rate": 2.547259478297603e-06, - "loss": 0.2808, + "loss": 0.9883, "step": 82630 }, { "epoch": 8.73, - "grad_norm": 10.297870635986328, + "grad_norm": 27.705015182495117, "learning_rate": 2.545147322842961e-06, - "loss": 0.2566, + "loss": 0.6756, "step": 82640 }, { "epoch": 8.73, - "grad_norm": 27.477216720581055, + "grad_norm": 33.683345794677734, "learning_rate": 2.5430351673883204e-06, - "loss": 0.1985, + "loss": 0.7966, "step": 82650 }, { "epoch": 8.73, - "grad_norm": 0.5590819716453552, + "grad_norm": 38.40707778930664, "learning_rate": 2.5409230119336787e-06, - "loss": 0.2181, + "loss": 0.6962, "step": 82660 }, { "epoch": 8.73, - "grad_norm": 18.282957077026367, + "grad_norm": 49.45185089111328, "learning_rate": 2.538810856479037e-06, - "loss": 0.3371, + "loss": 0.9329, "step": 82670 }, { "epoch": 8.73, - "grad_norm": 12.469778060913086, + "grad_norm": 35.52604675292969, "learning_rate": 2.5366987010243954e-06, - "loss": 0.2781, + "loss": 0.7598, "step": 82680 }, { "epoch": 8.73, - "grad_norm": 0.4322793781757355, + "grad_norm": 32.93948745727539, "learning_rate": 2.5345865455697542e-06, - "loss": 0.21, + "loss": 0.8671, "step": 82690 }, { "epoch": 8.73, - "grad_norm": 19.90980339050293, + "grad_norm": 18.616010665893555, "learning_rate": 2.5324743901151126e-06, - "loss": 0.5728, + "loss": 1.0898, "step": 82700 }, { "epoch": 8.73, - "grad_norm": 18.30820655822754, + "grad_norm": 50.44078063964844, "learning_rate": 2.5303622346604714e-06, - "loss": 0.403, + "loss": 1.2757, "step": 82710 }, { "epoch": 8.74, - "grad_norm": 25.513227462768555, + "grad_norm": 25.06107521057129, "learning_rate": 2.5282500792058297e-06, - "loss": 0.4989, + "loss": 0.9571, "step": 82720 }, { "epoch": 8.74, - "grad_norm": 7.242938995361328, + "grad_norm": 16.371286392211914, "learning_rate": 2.5261379237511885e-06, - "loss": 0.2525, + "loss": 0.7193, "step": 82730 }, { "epoch": 8.74, - "grad_norm": 0.036988429725170135, + "grad_norm": 45.93461990356445, "learning_rate": 2.524025768296547e-06, - "loss": 0.4384, + "loss": 0.6728, "step": 82740 }, { "epoch": 8.74, - "grad_norm": 0.8263336420059204, + "grad_norm": 2.3001041412353516, "learning_rate": 2.521913612841905e-06, - "loss": 0.4069, + "loss": 0.7244, "step": 82750 }, { "epoch": 8.74, - "grad_norm": 3.9332737922668457, + "grad_norm": 51.873775482177734, "learning_rate": 2.5198014573872636e-06, - "loss": 0.1351, + "loss": 0.7504, "step": 82760 }, { "epoch": 8.74, - "grad_norm": 8.235828399658203, + "grad_norm": 18.563119888305664, "learning_rate": 2.5176893019326228e-06, - "loss": 0.2412, + "loss": 0.669, "step": 82770 }, { "epoch": 8.74, - "grad_norm": 9.45411491394043, + "grad_norm": 43.54107666015625, "learning_rate": 2.515577146477981e-06, - "loss": 0.259, + "loss": 0.7973, "step": 82780 }, { "epoch": 8.74, - "grad_norm": 24.91488265991211, + "grad_norm": 24.426448822021484, "learning_rate": 2.5134649910233395e-06, - "loss": 0.2133, + "loss": 0.7847, "step": 82790 }, { "epoch": 8.74, - "grad_norm": 17.359479904174805, + "grad_norm": 41.600067138671875, "learning_rate": 2.511352835568698e-06, - "loss": 0.3745, + "loss": 0.8893, "step": 82800 }, { "epoch": 8.75, - "grad_norm": 38.4270133972168, + "grad_norm": 52.49769973754883, "learning_rate": 2.5092406801140566e-06, - "loss": 0.5054, + "loss": 1.1852, "step": 82810 }, { "epoch": 8.75, - "grad_norm": 22.438180923461914, + "grad_norm": 61.46405792236328, "learning_rate": 2.507128524659415e-06, - "loss": 0.2994, + "loss": 0.8228, "step": 82820 }, { "epoch": 8.75, - "grad_norm": 1.3325453996658325, + "grad_norm": 38.16945266723633, "learning_rate": 2.5050163692047737e-06, - "loss": 0.5281, + "loss": 0.9824, "step": 82830 }, { "epoch": 8.75, - "grad_norm": 19.432552337646484, + "grad_norm": 29.562726974487305, "learning_rate": 2.5029042137501325e-06, - "loss": 0.1764, + "loss": 0.5618, "step": 82840 }, { "epoch": 8.75, - "grad_norm": 18.649030685424805, + "grad_norm": 61.42092514038086, "learning_rate": 2.500792058295491e-06, - "loss": 0.3484, + "loss": 0.6481, "step": 82850 }, { "epoch": 8.75, - "grad_norm": 20.199678421020508, + "grad_norm": 46.416221618652344, "learning_rate": 2.4986799028408492e-06, - "loss": 0.1115, + "loss": 0.758, "step": 82860 }, { "epoch": 8.75, - "grad_norm": 37.195556640625, + "grad_norm": 65.21392059326172, "learning_rate": 2.496567747386208e-06, - "loss": 0.4063, + "loss": 1.14, "step": 82870 }, { "epoch": 8.75, - "grad_norm": 20.423458099365234, + "grad_norm": 68.66676330566406, "learning_rate": 2.4944555919315664e-06, - "loss": 0.4028, + "loss": 0.705, "step": 82880 }, { "epoch": 8.75, - "grad_norm": 9.179692268371582, + "grad_norm": 30.5376033782959, "learning_rate": 2.492343436476925e-06, - "loss": 0.2757, + "loss": 0.7679, "step": 82890 }, { "epoch": 8.75, - "grad_norm": 0.29933181405067444, + "grad_norm": 32.25856018066406, "learning_rate": 2.4902312810222835e-06, - "loss": 0.2694, + "loss": 0.9977, "step": 82900 }, { "epoch": 8.76, - "grad_norm": 0.5271009802818298, + "grad_norm": 16.64865493774414, "learning_rate": 2.4881191255676423e-06, - "loss": 0.2994, + "loss": 0.8563, "step": 82910 }, { "epoch": 8.76, - "grad_norm": 4.217239856719971, + "grad_norm": 45.1708984375, "learning_rate": 2.4860069701130006e-06, - "loss": 0.7304, + "loss": 1.1293, "step": 82920 }, { "epoch": 8.76, - "grad_norm": 7.1819000244140625, + "grad_norm": 36.67914962768555, "learning_rate": 2.483894814658359e-06, - "loss": 0.3823, + "loss": 0.809, "step": 82930 }, { "epoch": 8.76, - "grad_norm": 0.2669968605041504, + "grad_norm": 32.48982620239258, "learning_rate": 2.4817826592037173e-06, - "loss": 0.3007, + "loss": 0.6685, "step": 82940 }, { "epoch": 8.76, - "grad_norm": 15.3150634765625, + "grad_norm": 32.061744689941406, "learning_rate": 2.479670503749076e-06, - "loss": 0.147, + "loss": 0.7677, "step": 82950 }, { "epoch": 8.76, - "grad_norm": 0.02523491531610489, + "grad_norm": 29.957759857177734, "learning_rate": 2.4775583482944345e-06, - "loss": 0.2552, + "loss": 0.8367, "step": 82960 }, { "epoch": 8.76, - "grad_norm": 0.45265159010887146, + "grad_norm": 55.84589385986328, "learning_rate": 2.4754461928397933e-06, - "loss": 0.3214, + "loss": 0.9796, "step": 82970 }, { "epoch": 8.76, - "grad_norm": 0.2482157200574875, + "grad_norm": 42.09028625488281, "learning_rate": 2.4733340373851516e-06, - "loss": 0.1622, + "loss": 0.8956, "step": 82980 }, { "epoch": 8.76, - "grad_norm": 24.415071487426758, + "grad_norm": 35.053611755371094, "learning_rate": 2.4712218819305104e-06, - "loss": 0.1946, + "loss": 0.6466, "step": 82990 }, { "epoch": 8.77, - "grad_norm": 29.377355575561523, + "grad_norm": 45.85250473022461, "learning_rate": 2.4691097264758687e-06, - "loss": 0.5292, + "loss": 0.9218, "step": 83000 }, { "epoch": 8.77, - "grad_norm": 16.97059440612793, + "grad_norm": 53.93932342529297, "learning_rate": 2.4669975710212275e-06, - "loss": 0.5666, + "loss": 0.9899, "step": 83010 }, { "epoch": 8.77, - "grad_norm": 18.069734573364258, + "grad_norm": 59.494197845458984, "learning_rate": 2.464885415566586e-06, - "loss": 0.2933, + "loss": 1.1212, "step": 83020 }, { "epoch": 8.77, - "grad_norm": 27.100332260131836, + "grad_norm": 41.404483795166016, "learning_rate": 2.4627732601119447e-06, - "loss": 0.5595, + "loss": 0.8197, "step": 83030 }, { "epoch": 8.77, - "grad_norm": 20.826162338256836, + "grad_norm": 23.704530715942383, "learning_rate": 2.460661104657303e-06, - "loss": 0.3011, + "loss": 0.9273, "step": 83040 }, { "epoch": 8.77, - "grad_norm": 28.66699981689453, + "grad_norm": 24.816614151000977, "learning_rate": 2.4585489492026614e-06, - "loss": 0.2904, + "loss": 0.6339, "step": 83050 }, { "epoch": 8.77, - "grad_norm": 19.12812614440918, + "grad_norm": 58.395755767822266, "learning_rate": 2.45643679374802e-06, - "loss": 0.2775, + "loss": 0.8521, "step": 83060 }, { "epoch": 8.77, - "grad_norm": 30.15176010131836, + "grad_norm": 38.669776916503906, "learning_rate": 2.4543246382933785e-06, - "loss": 0.4748, + "loss": 0.9065, "step": 83070 }, { "epoch": 8.77, - "grad_norm": 0.4283275008201599, + "grad_norm": 52.84174728393555, "learning_rate": 2.4522124828387373e-06, - "loss": 0.2666, + "loss": 0.8418, "step": 83080 }, { "epoch": 8.77, - "grad_norm": 6.670082092285156, + "grad_norm": 44.46549606323242, "learning_rate": 2.4501003273840956e-06, - "loss": 0.104, + "loss": 0.8241, "step": 83090 }, { "epoch": 8.78, - "grad_norm": 8.76090145111084, + "grad_norm": 47.83694839477539, "learning_rate": 2.4479881719294544e-06, - "loss": 0.2538, + "loss": 0.9409, "step": 83100 }, { "epoch": 8.78, - "grad_norm": 0.11169039458036423, + "grad_norm": 34.04371643066406, "learning_rate": 2.4458760164748128e-06, - "loss": 0.067, + "loss": 0.5096, "step": 83110 }, { "epoch": 8.78, - "grad_norm": 0.24964973330497742, + "grad_norm": 29.645736694335938, "learning_rate": 2.4437638610201716e-06, - "loss": 0.3023, + "loss": 0.9416, "step": 83120 }, { "epoch": 8.78, - "grad_norm": 0.6734632253646851, + "grad_norm": 23.04947853088379, "learning_rate": 2.44165170556553e-06, - "loss": 0.2464, + "loss": 0.5809, "step": 83130 }, { "epoch": 8.78, - "grad_norm": 3.0457096099853516, + "grad_norm": 48.628021240234375, "learning_rate": 2.4395395501108883e-06, - "loss": 0.5207, + "loss": 1.0462, "step": 83140 }, { "epoch": 8.78, - "grad_norm": 15.931844711303711, + "grad_norm": 43.51254653930664, "learning_rate": 2.437427394656247e-06, - "loss": 0.6146, + "loss": 1.0869, "step": 83150 }, { "epoch": 8.78, - "grad_norm": 0.030378609895706177, + "grad_norm": 29.390918731689453, "learning_rate": 2.4353152392016054e-06, - "loss": 0.2157, + "loss": 0.9811, "step": 83160 }, { "epoch": 8.78, - "grad_norm": 24.72281265258789, + "grad_norm": 43.04023361206055, "learning_rate": 2.4332030837469638e-06, - "loss": 0.3014, + "loss": 0.7557, "step": 83170 }, { "epoch": 8.78, - "grad_norm": 16.736248016357422, + "grad_norm": 36.91523361206055, "learning_rate": 2.4310909282923225e-06, - "loss": 0.2254, + "loss": 0.971, "step": 83180 }, { "epoch": 8.79, - "grad_norm": 0.23564442992210388, + "grad_norm": 45.05168914794922, "learning_rate": 2.428978772837681e-06, - "loss": 0.2753, + "loss": 0.6157, "step": 83190 }, { "epoch": 8.79, - "grad_norm": 18.456729888916016, + "grad_norm": 44.26642608642578, "learning_rate": 2.4268666173830397e-06, - "loss": 0.4482, + "loss": 1.2041, "step": 83200 }, { "epoch": 8.79, - "grad_norm": 0.3044164478778839, + "grad_norm": 45.88425064086914, "learning_rate": 2.424754461928398e-06, - "loss": 0.2179, + "loss": 0.7295, "step": 83210 }, { "epoch": 8.79, - "grad_norm": 6.1253814697265625, + "grad_norm": 1.8997429609298706, "learning_rate": 2.422642306473757e-06, - "loss": 0.1937, + "loss": 0.6284, "step": 83220 }, { "epoch": 8.79, - "grad_norm": 12.874032020568848, + "grad_norm": 38.185977935791016, "learning_rate": 2.420530151019115e-06, - "loss": 0.3036, + "loss": 0.9478, "step": 83230 }, { "epoch": 8.79, - "grad_norm": 6.844545841217041, + "grad_norm": 19.676956176757812, "learning_rate": 2.418417995564474e-06, - "loss": 0.0661, + "loss": 0.5183, "step": 83240 }, { "epoch": 8.79, - "grad_norm": 9.527207374572754, + "grad_norm": 35.878143310546875, "learning_rate": 2.4163058401098323e-06, - "loss": 0.4759, + "loss": 0.9023, "step": 83250 }, { "epoch": 8.79, - "grad_norm": 2.4251692295074463, + "grad_norm": 9.645193099975586, "learning_rate": 2.4141936846551906e-06, - "loss": 0.3108, + "loss": 1.0035, "step": 83260 }, { "epoch": 8.79, - "grad_norm": 0.0219525545835495, + "grad_norm": 39.205299377441406, "learning_rate": 2.4120815292005494e-06, - "loss": 0.2351, + "loss": 0.6974, "step": 83270 }, { "epoch": 8.8, - "grad_norm": 29.67926025390625, + "grad_norm": 37.46156692504883, "learning_rate": 2.4099693737459078e-06, - "loss": 0.2968, + "loss": 0.9502, "step": 83280 }, { "epoch": 8.8, - "grad_norm": 12.111532211303711, + "grad_norm": 30.566335678100586, "learning_rate": 2.4078572182912666e-06, - "loss": 0.3922, + "loss": 0.8306, "step": 83290 }, { "epoch": 8.8, - "grad_norm": 0.04767506197094917, + "grad_norm": 11.616884231567383, "learning_rate": 2.405745062836625e-06, - "loss": 0.3136, + "loss": 0.8943, "step": 83300 }, { "epoch": 8.8, - "grad_norm": 0.5789583921432495, + "grad_norm": 37.2587776184082, "learning_rate": 2.4036329073819837e-06, - "loss": 0.2954, + "loss": 0.971, "step": 83310 }, { "epoch": 8.8, - "grad_norm": 0.9334201812744141, + "grad_norm": 51.11481857299805, "learning_rate": 2.401520751927342e-06, - "loss": 0.4854, + "loss": 1.2025, "step": 83320 }, { "epoch": 8.8, - "grad_norm": 27.13875389099121, + "grad_norm": 29.951444625854492, "learning_rate": 2.399408596472701e-06, - "loss": 0.4227, + "loss": 0.8808, "step": 83330 }, { "epoch": 8.8, - "grad_norm": 0.611283540725708, + "grad_norm": 34.874305725097656, "learning_rate": 2.397296441018059e-06, - "loss": 0.4801, + "loss": 1.0921, "step": 83340 }, { "epoch": 8.8, - "grad_norm": 11.564566612243652, + "grad_norm": 37.4177131652832, "learning_rate": 2.395184285563418e-06, - "loss": 0.5179, + "loss": 0.8891, "step": 83350 }, { "epoch": 8.8, - "grad_norm": 10.063833236694336, + "grad_norm": 40.853660583496094, "learning_rate": 2.3930721301087763e-06, - "loss": 0.395, + "loss": 0.934, "step": 83360 }, { "epoch": 8.8, - "grad_norm": 16.547475814819336, + "grad_norm": 51.99988555908203, "learning_rate": 2.3909599746541347e-06, - "loss": 0.2003, + "loss": 0.8982, "step": 83370 }, { "epoch": 8.81, - "grad_norm": 3.181128740310669, + "grad_norm": 43.73463439941406, "learning_rate": 2.388847819199493e-06, - "loss": 0.3292, + "loss": 1.0047, "step": 83380 }, { "epoch": 8.81, - "grad_norm": 19.8532772064209, + "grad_norm": 56.27779769897461, "learning_rate": 2.386735663744852e-06, - "loss": 0.4717, + "loss": 0.88, "step": 83390 }, { "epoch": 8.81, - "grad_norm": 2.1300668716430664, + "grad_norm": 37.221595764160156, "learning_rate": 2.38462350829021e-06, - "loss": 0.2388, + "loss": 0.7547, "step": 83400 }, { "epoch": 8.81, - "grad_norm": 31.059612274169922, + "grad_norm": 37.04024887084961, "learning_rate": 2.382511352835569e-06, - "loss": 0.3407, + "loss": 0.7503, "step": 83410 }, { "epoch": 8.81, - "grad_norm": 3.964128255844116, + "grad_norm": 37.25203323364258, "learning_rate": 2.3803991973809273e-06, - "loss": 0.3498, + "loss": 0.761, "step": 83420 }, { "epoch": 8.81, - "grad_norm": 15.198229789733887, + "grad_norm": 63.11129379272461, "learning_rate": 2.378287041926286e-06, - "loss": 0.3961, + "loss": 1.0179, "step": 83430 }, { "epoch": 8.81, - "grad_norm": 11.02817440032959, + "grad_norm": 49.35309600830078, "learning_rate": 2.3761748864716444e-06, - "loss": 0.4886, + "loss": 0.9352, "step": 83440 }, { "epoch": 8.81, - "grad_norm": 3.16357684135437, + "grad_norm": 17.868886947631836, "learning_rate": 2.374062731017003e-06, - "loss": 0.2902, + "loss": 0.5289, "step": 83450 }, { "epoch": 8.81, - "grad_norm": 0.8794878721237183, + "grad_norm": 62.16817092895508, "learning_rate": 2.3719505755623616e-06, - "loss": 0.1385, + "loss": 0.8244, "step": 83460 }, { "epoch": 8.82, - "grad_norm": 32.9735221862793, + "grad_norm": 41.83926773071289, "learning_rate": 2.3698384201077203e-06, - "loss": 0.4177, + "loss": 0.8226, "step": 83470 }, { "epoch": 8.82, - "grad_norm": 6.288939476013184, + "grad_norm": 63.05118942260742, "learning_rate": 2.3677262646530787e-06, - "loss": 0.4546, + "loss": 0.8437, "step": 83480 }, { "epoch": 8.82, - "grad_norm": 1.8413362503051758, + "grad_norm": 30.624225616455078, "learning_rate": 2.365614109198437e-06, - "loss": 0.4492, + "loss": 1.2459, "step": 83490 }, { "epoch": 8.82, - "grad_norm": 19.556119918823242, + "grad_norm": 59.718204498291016, "learning_rate": 2.363501953743796e-06, - "loss": 0.1363, + "loss": 0.7983, "step": 83500 }, { "epoch": 8.82, - "grad_norm": 13.061901092529297, + "grad_norm": 56.414207458496094, "learning_rate": 2.361389798289154e-06, - "loss": 0.3648, + "loss": 0.8027, "step": 83510 }, { "epoch": 8.82, - "grad_norm": 7.562533855438232, + "grad_norm": 46.34878158569336, "learning_rate": 2.359277642834513e-06, - "loss": 0.2482, + "loss": 1.1072, "step": 83520 }, { "epoch": 8.82, - "grad_norm": 7.413398265838623, + "grad_norm": 17.98820686340332, "learning_rate": 2.3571654873798713e-06, - "loss": 0.1858, + "loss": 0.6029, "step": 83530 }, { "epoch": 8.82, - "grad_norm": 10.735782623291016, + "grad_norm": 33.775428771972656, "learning_rate": 2.35505333192523e-06, - "loss": 0.1164, + "loss": 1.0259, "step": 83540 }, { "epoch": 8.82, - "grad_norm": 32.331851959228516, + "grad_norm": 56.59265899658203, "learning_rate": 2.3529411764705885e-06, - "loss": 0.5033, + "loss": 0.9359, "step": 83550 }, { "epoch": 8.82, - "grad_norm": 15.258628845214844, + "grad_norm": 21.08516502380371, "learning_rate": 2.3508290210159472e-06, - "loss": 0.1805, + "loss": 0.6245, "step": 83560 }, { "epoch": 8.83, - "grad_norm": 18.926597595214844, + "grad_norm": 16.393098831176758, "learning_rate": 2.3487168655613056e-06, - "loss": 0.5372, + "loss": 0.8254, "step": 83570 }, { "epoch": 8.83, - "grad_norm": 2.7127251625061035, + "grad_norm": 49.96042251586914, "learning_rate": 2.346604710106664e-06, - "loss": 0.3835, + "loss": 0.8076, "step": 83580 }, { "epoch": 8.83, - "grad_norm": 10.237618446350098, + "grad_norm": 38.891021728515625, "learning_rate": 2.3444925546520227e-06, - "loss": 0.2573, + "loss": 1.1253, "step": 83590 }, { "epoch": 8.83, - "grad_norm": 1.4284775257110596, + "grad_norm": 29.850730895996094, "learning_rate": 2.342380399197381e-06, - "loss": 0.5661, + "loss": 0.9234, "step": 83600 }, { "epoch": 8.83, - "grad_norm": 0.024252962321043015, + "grad_norm": 27.224626541137695, "learning_rate": 2.3402682437427394e-06, - "loss": 0.2851, + "loss": 0.7371, "step": 83610 }, { "epoch": 8.83, - "grad_norm": 0.05943336710333824, + "grad_norm": 66.19739532470703, "learning_rate": 2.338156088288098e-06, - "loss": 0.2528, + "loss": 0.8722, "step": 83620 }, { "epoch": 8.83, - "grad_norm": 0.24778544902801514, + "grad_norm": 10.695734977722168, "learning_rate": 2.3360439328334566e-06, - "loss": 0.4851, + "loss": 0.8931, "step": 83630 }, { "epoch": 8.83, - "grad_norm": 4.1159491539001465, + "grad_norm": 39.18965530395508, "learning_rate": 2.3339317773788153e-06, - "loss": 0.1673, + "loss": 0.9419, "step": 83640 }, { "epoch": 8.83, - "grad_norm": 10.661661148071289, + "grad_norm": 44.21870422363281, "learning_rate": 2.3318196219241737e-06, - "loss": 0.5143, + "loss": 0.7878, "step": 83650 }, { "epoch": 8.84, - "grad_norm": 22.673545837402344, + "grad_norm": 28.234342575073242, "learning_rate": 2.3297074664695325e-06, - "loss": 0.4062, + "loss": 0.5879, "step": 83660 }, { "epoch": 8.84, - "grad_norm": 23.312515258789062, + "grad_norm": 33.441978454589844, "learning_rate": 2.327595311014891e-06, - "loss": 0.4094, + "loss": 1.1011, "step": 83670 }, { "epoch": 8.84, - "grad_norm": 0.6042251586914062, + "grad_norm": 53.30372619628906, "learning_rate": 2.3254831555602496e-06, - "loss": 0.4598, + "loss": 1.0541, "step": 83680 }, { "epoch": 8.84, - "grad_norm": 0.010239911265671253, + "grad_norm": 23.99544906616211, "learning_rate": 2.323371000105608e-06, - "loss": 0.2776, + "loss": 0.9978, "step": 83690 }, { "epoch": 8.84, - "grad_norm": 15.269939422607422, + "grad_norm": 48.13832473754883, "learning_rate": 2.3212588446509663e-06, - "loss": 0.3626, + "loss": 0.6962, "step": 83700 }, { "epoch": 8.84, - "grad_norm": 6.050297737121582, + "grad_norm": 56.354286193847656, "learning_rate": 2.319146689196325e-06, - "loss": 0.2348, + "loss": 0.7549, "step": 83710 }, { "epoch": 8.84, - "grad_norm": 35.4648323059082, + "grad_norm": 41.26181411743164, "learning_rate": 2.3170345337416835e-06, - "loss": 0.374, + "loss": 0.7699, "step": 83720 }, { "epoch": 8.84, - "grad_norm": 8.087577819824219, + "grad_norm": 48.62361526489258, "learning_rate": 2.3149223782870422e-06, - "loss": 0.3896, + "loss": 1.0013, "step": 83730 }, { "epoch": 8.84, - "grad_norm": 3.2105493545532227, + "grad_norm": 46.90280532836914, "learning_rate": 2.3128102228324006e-06, - "loss": 0.5049, + "loss": 1.0988, "step": 83740 }, { "epoch": 8.84, - "grad_norm": 18.20578956604004, + "grad_norm": 58.63530731201172, "learning_rate": 2.3106980673777594e-06, - "loss": 0.4617, + "loss": 0.8736, "step": 83750 }, { "epoch": 8.85, - "grad_norm": 9.089902877807617, + "grad_norm": 46.03221130371094, "learning_rate": 2.3085859119231177e-06, - "loss": 0.4878, + "loss": 0.9996, "step": 83760 }, { "epoch": 8.85, - "grad_norm": 5.12348747253418, + "grad_norm": 60.32621383666992, "learning_rate": 2.3064737564684765e-06, - "loss": 0.3212, + "loss": 0.8968, "step": 83770 }, { "epoch": 8.85, - "grad_norm": 0.4375591278076172, + "grad_norm": 24.463891983032227, "learning_rate": 2.304361601013835e-06, - "loss": 0.2641, + "loss": 0.9581, "step": 83780 }, { "epoch": 8.85, - "grad_norm": 2.3739049434661865, + "grad_norm": 40.12660217285156, "learning_rate": 2.3022494455591936e-06, - "loss": 0.4137, + "loss": 0.8852, "step": 83790 }, { "epoch": 8.85, - "grad_norm": 15.21196174621582, + "grad_norm": 27.803367614746094, "learning_rate": 2.300137290104552e-06, - "loss": 0.3425, + "loss": 0.9731, "step": 83800 }, { "epoch": 8.85, - "grad_norm": 0.1395282745361328, + "grad_norm": 55.847808837890625, "learning_rate": 2.2980251346499103e-06, - "loss": 0.1987, + "loss": 1.0195, "step": 83810 }, { "epoch": 8.85, - "grad_norm": 27.641830444335938, + "grad_norm": 58.09909439086914, "learning_rate": 2.2959129791952687e-06, - "loss": 0.5933, + "loss": 0.7662, "step": 83820 }, { "epoch": 8.85, - "grad_norm": 16.161283493041992, + "grad_norm": 33.89020538330078, "learning_rate": 2.2938008237406275e-06, - "loss": 0.4238, + "loss": 0.9747, "step": 83830 }, { "epoch": 8.85, - "grad_norm": 21.514036178588867, + "grad_norm": 27.055635452270508, "learning_rate": 2.291688668285986e-06, - "loss": 0.1858, + "loss": 0.91, "step": 83840 }, { "epoch": 8.86, - "grad_norm": 1.2978721857070923, + "grad_norm": 27.59817123413086, "learning_rate": 2.2895765128313446e-06, - "loss": 0.1936, + "loss": 0.7401, "step": 83850 }, { "epoch": 8.86, - "grad_norm": 18.25662612915039, + "grad_norm": 50.91975021362305, "learning_rate": 2.287464357376703e-06, - "loss": 0.3188, + "loss": 0.9312, "step": 83860 }, { "epoch": 8.86, - "grad_norm": 0.559874951839447, + "grad_norm": 35.838863372802734, "learning_rate": 2.2853522019220618e-06, - "loss": 0.3138, + "loss": 0.7423, "step": 83870 }, { "epoch": 8.86, - "grad_norm": 23.072599411010742, + "grad_norm": 51.0557746887207, "learning_rate": 2.28324004646742e-06, - "loss": 0.3508, + "loss": 0.6385, "step": 83880 }, { "epoch": 8.86, - "grad_norm": 2.655043601989746, + "grad_norm": 58.871238708496094, "learning_rate": 2.281127891012779e-06, - "loss": 0.3164, + "loss": 1.0109, "step": 83890 }, { "epoch": 8.86, - "grad_norm": 1.1907140016555786, + "grad_norm": 37.473480224609375, "learning_rate": 2.2790157355581372e-06, - "loss": 0.2811, + "loss": 0.7705, "step": 83900 }, { "epoch": 8.86, - "grad_norm": 0.2564692199230194, + "grad_norm": 25.332645416259766, "learning_rate": 2.276903580103496e-06, - "loss": 0.3056, + "loss": 0.8163, "step": 83910 }, { "epoch": 8.86, - "grad_norm": 0.17902638018131256, + "grad_norm": 30.46021842956543, "learning_rate": 2.2747914246488544e-06, - "loss": 0.2118, + "loss": 0.8403, "step": 83920 }, { "epoch": 8.86, - "grad_norm": 15.445877075195312, + "grad_norm": 48.17649841308594, "learning_rate": 2.2726792691942127e-06, - "loss": 0.7453, + "loss": 0.9908, "step": 83930 }, { "epoch": 8.86, - "grad_norm": 0.03319980576634407, + "grad_norm": 4.184080600738525, "learning_rate": 2.270567113739571e-06, - "loss": 0.422, + "loss": 0.8045, "step": 83940 }, { "epoch": 8.87, - "grad_norm": 18.16193199157715, + "grad_norm": 46.94548416137695, "learning_rate": 2.26845495828493e-06, - "loss": 0.4013, + "loss": 0.7281, "step": 83950 }, { "epoch": 8.87, - "grad_norm": 11.168543815612793, + "grad_norm": 36.644371032714844, "learning_rate": 2.2663428028302886e-06, - "loss": 0.1781, + "loss": 1.0069, "step": 83960 }, { "epoch": 8.87, - "grad_norm": 28.4660701751709, + "grad_norm": 49.782466888427734, "learning_rate": 2.264230647375647e-06, - "loss": 0.2771, + "loss": 0.9975, "step": 83970 }, { "epoch": 8.87, - "grad_norm": 14.83912181854248, + "grad_norm": 59.60932159423828, "learning_rate": 2.2621184919210058e-06, - "loss": 0.6287, + "loss": 0.7853, "step": 83980 }, { "epoch": 8.87, - "grad_norm": 1.0587235689163208, + "grad_norm": 32.96002960205078, "learning_rate": 2.260006336466364e-06, - "loss": 0.2639, + "loss": 0.821, "step": 83990 }, { "epoch": 8.87, - "grad_norm": 1.518808364868164, + "grad_norm": 31.011924743652344, "learning_rate": 2.257894181011723e-06, - "loss": 0.2153, + "loss": 0.7896, "step": 84000 }, { "epoch": 8.87, - "grad_norm": 9.10312271118164, + "grad_norm": 29.845640182495117, "learning_rate": 2.2557820255570813e-06, - "loss": 0.4061, + "loss": 0.7955, "step": 84010 }, { "epoch": 8.87, - "grad_norm": 6.573938846588135, + "grad_norm": 40.334938049316406, "learning_rate": 2.2536698701024396e-06, - "loss": 0.4918, + "loss": 1.0452, "step": 84020 }, { "epoch": 8.87, - "grad_norm": 0.24016600847244263, + "grad_norm": 27.804874420166016, "learning_rate": 2.2515577146477984e-06, - "loss": 0.4252, + "loss": 0.659, "step": 84030 }, { "epoch": 8.88, - "grad_norm": 8.831706047058105, + "grad_norm": 43.03264236450195, "learning_rate": 2.2494455591931568e-06, - "loss": 0.2613, + "loss": 0.9705, "step": 84040 }, { "epoch": 8.88, - "grad_norm": 26.390605926513672, + "grad_norm": 26.01602554321289, "learning_rate": 2.247333403738515e-06, - "loss": 0.3407, + "loss": 0.8333, "step": 84050 }, { "epoch": 8.88, - "grad_norm": 10.306145668029785, + "grad_norm": 63.44452667236328, "learning_rate": 2.245221248283874e-06, - "loss": 0.6883, + "loss": 1.0727, "step": 84060 }, { "epoch": 8.88, - "grad_norm": 0.08621324598789215, + "grad_norm": 24.148361206054688, "learning_rate": 2.2431090928292322e-06, - "loss": 0.2708, + "loss": 0.9827, "step": 84070 }, { "epoch": 8.88, - "grad_norm": 0.13864077627658844, + "grad_norm": 8.607580184936523, "learning_rate": 2.240996937374591e-06, - "loss": 0.3969, + "loss": 0.9498, "step": 84080 }, { "epoch": 8.88, - "grad_norm": 5.565453052520752, + "grad_norm": 55.624778747558594, "learning_rate": 2.2388847819199494e-06, - "loss": 0.3446, + "loss": 1.0491, "step": 84090 }, { "epoch": 8.88, - "grad_norm": 10.74009895324707, + "grad_norm": 31.877704620361328, "learning_rate": 2.236772626465308e-06, - "loss": 0.1842, + "loss": 0.8159, "step": 84100 }, { "epoch": 8.88, - "grad_norm": 11.031627655029297, + "grad_norm": 69.32553100585938, "learning_rate": 2.2346604710106665e-06, - "loss": 0.3665, + "loss": 0.9563, "step": 84110 }, { "epoch": 8.88, - "grad_norm": 4.837457656860352, + "grad_norm": 18.428905487060547, "learning_rate": 2.2325483155560253e-06, - "loss": 0.2568, + "loss": 0.8174, "step": 84120 }, { "epoch": 8.88, - "grad_norm": 2.1591365337371826, + "grad_norm": 66.67303466796875, "learning_rate": 2.2304361601013836e-06, - "loss": 0.2534, + "loss": 1.0676, "step": 84130 }, { "epoch": 8.89, - "grad_norm": 14.579270362854004, + "grad_norm": 45.264102935791016, "learning_rate": 2.228324004646742e-06, - "loss": 0.1382, + "loss": 0.6037, "step": 84140 }, { "epoch": 8.89, - "grad_norm": 12.697659492492676, + "grad_norm": 27.535951614379883, "learning_rate": 2.2262118491921008e-06, - "loss": 0.4702, + "loss": 0.8337, "step": 84150 }, { "epoch": 8.89, - "grad_norm": 29.778362274169922, + "grad_norm": 49.29058074951172, "learning_rate": 2.224099693737459e-06, - "loss": 0.4301, + "loss": 0.8062, "step": 84160 }, { "epoch": 8.89, - "grad_norm": 32.382869720458984, + "grad_norm": 49.26533889770508, "learning_rate": 2.221987538282818e-06, - "loss": 0.6055, + "loss": 1.0092, "step": 84170 }, { "epoch": 8.89, - "grad_norm": 0.0649140402674675, + "grad_norm": 41.443477630615234, "learning_rate": 2.2198753828281763e-06, - "loss": 0.3534, + "loss": 1.0965, "step": 84180 }, { "epoch": 8.89, - "grad_norm": 10.892370223999023, + "grad_norm": 62.25000762939453, "learning_rate": 2.217763227373535e-06, - "loss": 0.2811, + "loss": 0.5716, "step": 84190 }, { "epoch": 8.89, - "grad_norm": 17.8467960357666, + "grad_norm": 41.37656021118164, "learning_rate": 2.2156510719188934e-06, - "loss": 0.4933, + "loss": 0.9584, "step": 84200 }, { "epoch": 8.89, - "grad_norm": 30.11771583557129, + "grad_norm": 19.966487884521484, "learning_rate": 2.213538916464252e-06, - "loss": 0.2675, + "loss": 0.989, "step": 84210 }, { "epoch": 8.89, - "grad_norm": 7.685906887054443, + "grad_norm": 63.41432571411133, "learning_rate": 2.2114267610096105e-06, - "loss": 0.414, + "loss": 0.826, "step": 84220 }, { "epoch": 8.9, - "grad_norm": 19.902822494506836, + "grad_norm": 33.27949905395508, "learning_rate": 2.2093146055549693e-06, - "loss": 0.5841, + "loss": 1.1038, "step": 84230 }, { "epoch": 8.9, - "grad_norm": 0.6987266540527344, + "grad_norm": 30.116382598876953, "learning_rate": 2.2072024501003277e-06, - "loss": 0.4015, + "loss": 1.0582, "step": 84240 }, { "epoch": 8.9, - "grad_norm": 0.03764040395617485, + "grad_norm": 32.1053352355957, "learning_rate": 2.205090294645686e-06, - "loss": 0.2733, + "loss": 0.7116, "step": 84250 }, { "epoch": 8.9, - "grad_norm": 17.89554214477539, + "grad_norm": 51.53263854980469, "learning_rate": 2.2029781391910444e-06, - "loss": 0.4153, + "loss": 0.8533, "step": 84260 }, { "epoch": 8.9, - "grad_norm": 1.5645098686218262, + "grad_norm": 41.95273208618164, "learning_rate": 2.200865983736403e-06, - "loss": 0.2491, + "loss": 0.9211, "step": 84270 }, { "epoch": 8.9, - "grad_norm": 0.11339319497346878, + "grad_norm": 14.869037628173828, "learning_rate": 2.1987538282817615e-06, - "loss": 0.3028, + "loss": 0.5385, "step": 84280 }, { "epoch": 8.9, - "grad_norm": 2.0357730388641357, + "grad_norm": 15.955339431762695, "learning_rate": 2.1966416728271203e-06, - "loss": 0.2765, + "loss": 0.6471, "step": 84290 }, { "epoch": 8.9, - "grad_norm": 0.019890213385224342, + "grad_norm": 13.872275352478027, "learning_rate": 2.1945295173724787e-06, - "loss": 0.1347, + "loss": 0.7233, "step": 84300 }, { "epoch": 8.9, - "grad_norm": 4.8608269691467285, + "grad_norm": 42.26083755493164, "learning_rate": 2.1924173619178374e-06, - "loss": 0.5425, + "loss": 0.8556, "step": 84310 }, { "epoch": 8.9, - "grad_norm": 49.93098449707031, + "grad_norm": 59.54795837402344, "learning_rate": 2.190305206463196e-06, - "loss": 0.4636, + "loss": 0.9948, "step": 84320 }, { "epoch": 8.91, - "grad_norm": 10.468873977661133, + "grad_norm": 47.51835250854492, "learning_rate": 2.1881930510085546e-06, - "loss": 0.3686, + "loss": 1.0528, "step": 84330 }, { "epoch": 8.91, - "grad_norm": 0.3838830590248108, + "grad_norm": 52.38534164428711, "learning_rate": 2.186080895553913e-06, - "loss": 0.2961, + "loss": 0.7646, "step": 84340 }, { "epoch": 8.91, - "grad_norm": 0.09858237951993942, + "grad_norm": 29.227115631103516, "learning_rate": 2.1839687400992717e-06, - "loss": 0.3122, + "loss": 0.6842, "step": 84350 }, { "epoch": 8.91, - "grad_norm": 1.7554600238800049, + "grad_norm": 27.6495418548584, "learning_rate": 2.18185658464463e-06, - "loss": 0.3559, + "loss": 0.9974, "step": 84360 }, { "epoch": 8.91, - "grad_norm": 1.577849268913269, + "grad_norm": 71.54530334472656, "learning_rate": 2.1797444291899884e-06, - "loss": 0.1149, + "loss": 0.9643, "step": 84370 }, { "epoch": 8.91, - "grad_norm": 3.7131428718566895, + "grad_norm": 45.12397766113281, "learning_rate": 2.1776322737353468e-06, - "loss": 0.7423, + "loss": 0.819, "step": 84380 }, { "epoch": 8.91, - "grad_norm": 10.05324649810791, + "grad_norm": 40.11349105834961, "learning_rate": 2.1755201182807055e-06, - "loss": 0.4383, + "loss": 0.9819, "step": 84390 }, { "epoch": 8.91, - "grad_norm": 9.770354270935059, + "grad_norm": 53.373382568359375, "learning_rate": 2.1734079628260643e-06, - "loss": 0.2109, + "loss": 0.7061, "step": 84400 }, { "epoch": 8.91, - "grad_norm": 0.23831366002559662, + "grad_norm": 16.024120330810547, "learning_rate": 2.1712958073714227e-06, - "loss": 0.2457, + "loss": 0.6541, "step": 84410 }, { "epoch": 8.92, - "grad_norm": 30.386775970458984, + "grad_norm": 38.8643913269043, "learning_rate": 2.1691836519167815e-06, - "loss": 0.6192, + "loss": 0.8893, "step": 84420 }, { "epoch": 8.92, - "grad_norm": 13.312051773071289, + "grad_norm": 6.647177219390869, "learning_rate": 2.16707149646214e-06, - "loss": 0.3606, + "loss": 0.922, "step": 84430 }, { "epoch": 8.92, - "grad_norm": 2.9715421199798584, + "grad_norm": 47.92127227783203, "learning_rate": 2.1649593410074986e-06, - "loss": 0.2261, + "loss": 0.5214, "step": 84440 }, { "epoch": 8.92, - "grad_norm": 10.63807201385498, + "grad_norm": 24.65057373046875, "learning_rate": 2.162847185552857e-06, - "loss": 0.4294, + "loss": 1.0016, "step": 84450 }, { "epoch": 8.92, - "grad_norm": 0.610733151435852, + "grad_norm": 29.03304100036621, "learning_rate": 2.1607350300982153e-06, - "loss": 0.1949, + "loss": 0.6791, "step": 84460 }, { "epoch": 8.92, - "grad_norm": 6.6884589195251465, + "grad_norm": 44.22364807128906, "learning_rate": 2.158622874643574e-06, - "loss": 0.3804, + "loss": 0.9033, "step": 84470 }, { "epoch": 8.92, - "grad_norm": 4.864074230194092, + "grad_norm": 26.84343910217285, "learning_rate": 2.1565107191889324e-06, - "loss": 0.5534, + "loss": 0.9546, "step": 84480 }, { "epoch": 8.92, - "grad_norm": 24.145902633666992, + "grad_norm": 38.11946487426758, "learning_rate": 2.154398563734291e-06, - "loss": 0.2283, + "loss": 0.7187, "step": 84490 }, { "epoch": 8.92, - "grad_norm": 27.49432945251465, + "grad_norm": 16.732563018798828, "learning_rate": 2.1522864082796496e-06, - "loss": 0.4124, + "loss": 0.8387, "step": 84500 }, { "epoch": 8.92, - "grad_norm": 34.433536529541016, + "grad_norm": 35.30216979980469, "learning_rate": 2.150174252825008e-06, - "loss": 0.3342, + "loss": 0.7229, "step": 84510 }, { "epoch": 8.93, - "grad_norm": 16.92571449279785, + "grad_norm": 18.59609603881836, "learning_rate": 2.1480620973703667e-06, - "loss": 0.606, + "loss": 0.6217, "step": 84520 }, { "epoch": 8.93, - "grad_norm": 11.835471153259277, + "grad_norm": 38.738277435302734, "learning_rate": 2.145949941915725e-06, - "loss": 0.2885, + "loss": 1.2202, "step": 84530 }, { "epoch": 8.93, - "grad_norm": 10.426703453063965, + "grad_norm": 47.80493927001953, "learning_rate": 2.143837786461084e-06, - "loss": 0.2239, + "loss": 0.7254, "step": 84540 }, { "epoch": 8.93, - "grad_norm": 25.567764282226562, + "grad_norm": 48.434391021728516, "learning_rate": 2.141725631006442e-06, - "loss": 0.2355, + "loss": 0.4902, "step": 84550 }, { "epoch": 8.93, - "grad_norm": 5.439464092254639, + "grad_norm": 44.670318603515625, "learning_rate": 2.139613475551801e-06, - "loss": 0.3613, + "loss": 1.0392, "step": 84560 }, { "epoch": 8.93, - "grad_norm": 13.390445709228516, + "grad_norm": 40.35314178466797, "learning_rate": 2.1375013200971593e-06, - "loss": 0.4748, + "loss": 0.5543, "step": 84570 }, { "epoch": 8.93, - "grad_norm": 44.14957809448242, + "grad_norm": 39.132843017578125, "learning_rate": 2.1353891646425177e-06, - "loss": 0.2215, + "loss": 0.991, "step": 84580 }, { "epoch": 8.93, - "grad_norm": 0.2539387047290802, + "grad_norm": 58.98158264160156, "learning_rate": 2.1332770091878765e-06, - "loss": 0.4284, + "loss": 0.9352, "step": 84590 }, { "epoch": 8.93, - "grad_norm": 11.375728607177734, + "grad_norm": 44.120140075683594, "learning_rate": 2.131164853733235e-06, - "loss": 0.6595, + "loss": 0.8395, "step": 84600 }, { "epoch": 8.94, - "grad_norm": 29.014619827270508, + "grad_norm": 53.253753662109375, "learning_rate": 2.129052698278593e-06, - "loss": 0.4637, + "loss": 1.2757, "step": 84610 }, { "epoch": 8.94, - "grad_norm": 9.996240615844727, + "grad_norm": 55.212799072265625, "learning_rate": 2.126940542823952e-06, - "loss": 0.1231, + "loss": 0.6609, "step": 84620 }, { "epoch": 8.94, - "grad_norm": 4.809905052185059, + "grad_norm": 14.161224365234375, "learning_rate": 2.1248283873693107e-06, - "loss": 0.4227, + "loss": 0.746, "step": 84630 }, { "epoch": 8.94, - "grad_norm": 0.06438569724559784, + "grad_norm": 21.486351013183594, "learning_rate": 2.122716231914669e-06, - "loss": 0.3267, + "loss": 0.5794, "step": 84640 }, { "epoch": 8.94, - "grad_norm": 7.49101448059082, + "grad_norm": 45.66145706176758, "learning_rate": 2.120604076460028e-06, - "loss": 0.3293, + "loss": 0.8963, "step": 84650 }, { "epoch": 8.94, - "grad_norm": 12.663186073303223, + "grad_norm": 48.938350677490234, "learning_rate": 2.1184919210053862e-06, - "loss": 0.5652, + "loss": 1.2087, "step": 84660 }, { "epoch": 8.94, - "grad_norm": 22.757991790771484, + "grad_norm": 58.17861557006836, "learning_rate": 2.116379765550745e-06, - "loss": 0.4504, + "loss": 1.0096, "step": 84670 }, { "epoch": 8.94, - "grad_norm": 0.4509471654891968, + "grad_norm": 13.14881420135498, "learning_rate": 2.1142676100961034e-06, - "loss": 0.2058, + "loss": 0.4398, "step": 84680 }, { "epoch": 8.94, - "grad_norm": 0.13681066036224365, + "grad_norm": 15.497620582580566, "learning_rate": 2.1121554546414617e-06, - "loss": 0.4377, + "loss": 0.7929, "step": 84690 }, { "epoch": 8.94, - "grad_norm": 0.8215176463127136, + "grad_norm": 56.96913528442383, "learning_rate": 2.11004329918682e-06, - "loss": 0.2034, + "loss": 0.8586, "step": 84700 }, { "epoch": 8.95, - "grad_norm": 16.549352645874023, + "grad_norm": 33.28471755981445, "learning_rate": 2.107931143732179e-06, - "loss": 0.4669, + "loss": 0.9884, "step": 84710 }, { "epoch": 8.95, - "grad_norm": 1.5604307651519775, + "grad_norm": 46.658599853515625, "learning_rate": 2.105818988277537e-06, - "loss": 0.3193, + "loss": 0.9764, "step": 84720 }, { "epoch": 8.95, - "grad_norm": 32.53676223754883, + "grad_norm": 52.80826950073242, "learning_rate": 2.103706832822896e-06, - "loss": 0.2246, + "loss": 0.8783, "step": 84730 }, { "epoch": 8.95, - "grad_norm": 47.24123764038086, + "grad_norm": 39.82020568847656, "learning_rate": 2.1015946773682543e-06, - "loss": 0.4877, + "loss": 1.1178, "step": 84740 }, { "epoch": 8.95, - "grad_norm": 11.175087928771973, + "grad_norm": 40.0911750793457, "learning_rate": 2.099482521913613e-06, - "loss": 0.2922, + "loss": 0.6357, "step": 84750 }, { "epoch": 8.95, - "grad_norm": 6.601185321807861, + "grad_norm": 51.727115631103516, "learning_rate": 2.0973703664589715e-06, - "loss": 0.4113, + "loss": 0.892, "step": 84760 }, { "epoch": 8.95, - "grad_norm": 0.5368964076042175, + "grad_norm": 38.733638763427734, "learning_rate": 2.0952582110043302e-06, - "loss": 0.3593, + "loss": 0.9316, "step": 84770 }, { "epoch": 8.95, - "grad_norm": 15.495409965515137, + "grad_norm": 29.676013946533203, "learning_rate": 2.0931460555496886e-06, - "loss": 0.4938, + "loss": 0.913, "step": 84780 }, { "epoch": 8.95, - "grad_norm": 45.553466796875, + "grad_norm": 6.7125372886657715, "learning_rate": 2.0910339000950474e-06, - "loss": 0.3626, + "loss": 0.8677, "step": 84790 }, { "epoch": 8.96, - "grad_norm": 25.563825607299805, + "grad_norm": 43.33072280883789, "learning_rate": 2.0889217446404057e-06, - "loss": 0.5151, + "loss": 0.919, "step": 84800 }, { "epoch": 8.96, - "grad_norm": 43.232879638671875, + "grad_norm": 34.0038948059082, "learning_rate": 2.086809589185764e-06, - "loss": 0.2662, + "loss": 0.9802, "step": 84810 }, { "epoch": 8.96, - "grad_norm": 0.4882924258708954, + "grad_norm": 29.61541175842285, "learning_rate": 2.0846974337311224e-06, - "loss": 0.5394, + "loss": 0.8998, "step": 84820 }, { "epoch": 8.96, - "grad_norm": 21.57935333251953, + "grad_norm": 35.30944061279297, "learning_rate": 2.0825852782764812e-06, - "loss": 0.2936, + "loss": 0.7791, "step": 84830 }, { "epoch": 8.96, - "grad_norm": 0.08716855198144913, + "grad_norm": 33.19873809814453, "learning_rate": 2.0804731228218396e-06, - "loss": 0.2939, + "loss": 1.0112, "step": 84840 }, { "epoch": 8.96, - "grad_norm": 0.18033136427402496, + "grad_norm": 29.017621994018555, "learning_rate": 2.0783609673671984e-06, - "loss": 0.2319, + "loss": 0.8052, "step": 84850 }, { "epoch": 8.96, - "grad_norm": 11.929094314575195, + "grad_norm": 44.057247161865234, "learning_rate": 2.076248811912557e-06, - "loss": 0.1668, + "loss": 0.8515, "step": 84860 }, { "epoch": 8.96, - "grad_norm": 0.023901360109448433, + "grad_norm": 34.10380935668945, "learning_rate": 2.0741366564579155e-06, - "loss": 0.3457, + "loss": 0.9091, "step": 84870 }, { "epoch": 8.96, - "grad_norm": 1.2959020137786865, + "grad_norm": 46.09735870361328, "learning_rate": 2.0720245010032743e-06, - "loss": 0.2425, + "loss": 0.956, "step": 84880 }, { "epoch": 8.97, - "grad_norm": 16.47292709350586, + "grad_norm": 25.938446044921875, "learning_rate": 2.0699123455486326e-06, - "loss": 0.3379, + "loss": 0.7041, "step": 84890 }, { "epoch": 8.97, - "grad_norm": 10.285141944885254, + "grad_norm": 39.24211120605469, "learning_rate": 2.067800190093991e-06, - "loss": 0.3411, + "loss": 0.6524, "step": 84900 }, { "epoch": 8.97, - "grad_norm": 35.92616271972656, + "grad_norm": 27.024263381958008, "learning_rate": 2.0656880346393498e-06, - "loss": 0.2393, + "loss": 0.8352, "step": 84910 }, { "epoch": 8.97, - "grad_norm": 11.521953582763672, + "grad_norm": 60.512855529785156, "learning_rate": 2.063575879184708e-06, - "loss": 0.3256, + "loss": 0.8898, "step": 84920 }, { "epoch": 8.97, - "grad_norm": 5.153337478637695, + "grad_norm": 42.67161178588867, "learning_rate": 2.0614637237300665e-06, - "loss": 0.3764, + "loss": 0.9229, "step": 84930 }, { "epoch": 8.97, - "grad_norm": 7.941612243652344, + "grad_norm": 45.121788024902344, "learning_rate": 2.0593515682754253e-06, - "loss": 0.3011, + "loss": 0.7533, "step": 84940 }, { "epoch": 8.97, - "grad_norm": 11.409013748168945, + "grad_norm": 47.56502151489258, "learning_rate": 2.0572394128207836e-06, - "loss": 0.2511, + "loss": 0.5187, "step": 84950 }, { "epoch": 8.97, - "grad_norm": 1.8828061819076538, + "grad_norm": 21.480182647705078, "learning_rate": 2.0551272573661424e-06, - "loss": 0.4428, + "loss": 0.6923, "step": 84960 }, { "epoch": 8.97, - "grad_norm": 8.837446212768555, + "grad_norm": 31.254104614257812, "learning_rate": 2.0530151019115007e-06, - "loss": 0.3486, + "loss": 1.0836, "step": 84970 }, { "epoch": 8.97, - "grad_norm": 19.783409118652344, + "grad_norm": 59.69302749633789, "learning_rate": 2.0509029464568595e-06, - "loss": 0.4596, + "loss": 0.7863, "step": 84980 }, { "epoch": 8.98, - "grad_norm": 21.10133171081543, + "grad_norm": 37.46757507324219, "learning_rate": 2.048790791002218e-06, - "loss": 0.3203, + "loss": 0.7294, "step": 84990 }, { "epoch": 8.98, - "grad_norm": 18.40833854675293, + "grad_norm": 46.36732864379883, "learning_rate": 2.0466786355475767e-06, - "loss": 0.5791, + "loss": 1.1545, "step": 85000 }, { "epoch": 8.98, - "grad_norm": 0.159646674990654, + "grad_norm": 17.172414779663086, "learning_rate": 2.044566480092935e-06, - "loss": 0.3568, + "loss": 0.7273, "step": 85010 }, { "epoch": 8.98, - "grad_norm": 18.952449798583984, + "grad_norm": 41.30519485473633, "learning_rate": 2.0424543246382934e-06, - "loss": 0.5359, + "loss": 1.0931, "step": 85020 }, { "epoch": 8.98, - "grad_norm": 4.451479434967041, + "grad_norm": 26.089462280273438, "learning_rate": 2.040342169183652e-06, - "loss": 0.6679, + "loss": 0.8126, "step": 85030 }, { "epoch": 8.98, - "grad_norm": 12.538009643554688, + "grad_norm": 38.487850189208984, "learning_rate": 2.0382300137290105e-06, - "loss": 0.6306, + "loss": 1.0378, "step": 85040 }, { "epoch": 8.98, - "grad_norm": 4.685783863067627, + "grad_norm": 67.14640808105469, "learning_rate": 2.036117858274369e-06, - "loss": 0.3513, + "loss": 0.8164, "step": 85050 }, { "epoch": 8.98, - "grad_norm": 11.686248779296875, + "grad_norm": 39.420291900634766, "learning_rate": 2.0340057028197276e-06, - "loss": 0.3873, + "loss": 1.0473, "step": 85060 }, { "epoch": 8.98, - "grad_norm": 1.8737685680389404, + "grad_norm": 52.03282165527344, "learning_rate": 2.0318935473650864e-06, - "loss": 0.3964, + "loss": 1.0351, "step": 85070 }, { "epoch": 8.99, - "grad_norm": 1.0164872407913208, + "grad_norm": 58.049678802490234, "learning_rate": 2.0297813919104448e-06, - "loss": 0.2454, + "loss": 0.9197, "step": 85080 }, { "epoch": 8.99, - "grad_norm": 7.425736427307129, + "grad_norm": 46.967472076416016, "learning_rate": 2.0276692364558035e-06, - "loss": 0.152, + "loss": 0.6994, "step": 85090 }, { "epoch": 8.99, - "grad_norm": 17.281978607177734, + "grad_norm": 28.283910751342773, "learning_rate": 2.025557081001162e-06, - "loss": 0.4926, + "loss": 0.6929, "step": 85100 }, { "epoch": 8.99, - "grad_norm": 0.12159301340579987, + "grad_norm": 12.8204345703125, "learning_rate": 2.0234449255465207e-06, - "loss": 0.3806, + "loss": 0.9565, "step": 85110 }, { "epoch": 8.99, - "grad_norm": 26.74652671813965, + "grad_norm": 18.336055755615234, "learning_rate": 2.021332770091879e-06, - "loss": 0.3357, + "loss": 0.6018, "step": 85120 }, { "epoch": 8.99, - "grad_norm": 39.84540557861328, + "grad_norm": 44.754695892333984, "learning_rate": 2.0192206146372374e-06, - "loss": 0.3223, + "loss": 0.9271, "step": 85130 }, { "epoch": 8.99, - "grad_norm": 16.785932540893555, + "grad_norm": 52.176761627197266, "learning_rate": 2.0171084591825957e-06, - "loss": 0.3784, + "loss": 0.9442, "step": 85140 }, { "epoch": 8.99, - "grad_norm": 6.684837818145752, + "grad_norm": 37.81288528442383, "learning_rate": 2.0149963037279545e-06, - "loss": 0.4205, + "loss": 0.668, "step": 85150 }, { "epoch": 8.99, - "grad_norm": 6.762535572052002, + "grad_norm": 39.04780197143555, "learning_rate": 2.012884148273313e-06, - "loss": 0.3821, + "loss": 0.6662, "step": 85160 }, { "epoch": 8.99, - "grad_norm": 28.752254486083984, + "grad_norm": 28.231380462646484, "learning_rate": 2.0107719928186717e-06, - "loss": 0.2827, + "loss": 0.784, "step": 85170 }, { "epoch": 9.0, - "grad_norm": 1.3097100257873535, + "grad_norm": 43.0155029296875, "learning_rate": 2.00865983736403e-06, - "loss": 0.2812, + "loss": 1.1915, "step": 85180 }, { "epoch": 9.0, - "grad_norm": 0.13501164317131042, + "grad_norm": 40.938045501708984, "learning_rate": 2.006547681909389e-06, - "loss": 0.1813, + "loss": 0.7668, "step": 85190 }, { "epoch": 9.0, - "grad_norm": 15.753822326660156, + "grad_norm": 43.94539260864258, "learning_rate": 2.004435526454747e-06, - "loss": 0.3526, + "loss": 0.7032, "step": 85200 }, { "epoch": 9.0, - "grad_norm": 7.648777961730957, + "grad_norm": 51.3074951171875, "learning_rate": 2.002323371000106e-06, - "loss": 0.5924, + "loss": 0.5989, "step": 85210 }, { "epoch": 9.0, - "grad_norm": 14.120691299438477, + "grad_norm": 28.253793716430664, "learning_rate": 2.0002112155454643e-06, - "loss": 0.4012, + "loss": 0.8152, "step": 85220 }, { "epoch": 9.0, - "eval_accuracy": 0.9052277227722773, - "eval_loss": 0.3955126106739044, - "eval_runtime": 228.8924, - "eval_samples_per_second": 110.314, - "eval_steps_per_second": 13.793, + "eval_accuracy": 0.8194455445544554, + "eval_loss": 0.6640510559082031, + "eval_runtime": 36.984, + "eval_samples_per_second": 682.728, + "eval_steps_per_second": 85.361, "step": 85221 }, { "epoch": 9.0, - "grad_norm": 22.09530258178711, + "grad_norm": 28.650550842285156, "learning_rate": 1.998099060090823e-06, - "loss": 0.2941, + "loss": 0.7456, "step": 85230 }, { "epoch": 9.0, - "grad_norm": 3.419462203979492, + "grad_norm": 44.29703903198242, "learning_rate": 1.9959869046361814e-06, - "loss": 0.4018, + "loss": 0.5852, "step": 85240 }, { "epoch": 9.0, - "grad_norm": 0.22318211197853088, + "grad_norm": 20.038700103759766, "learning_rate": 1.9938747491815398e-06, - "loss": 0.2363, + "loss": 0.5145, "step": 85250 }, { "epoch": 9.0, - "grad_norm": 0.062369007617235184, + "grad_norm": 37.69400405883789, "learning_rate": 1.991762593726898e-06, - "loss": 0.1508, + "loss": 0.5468, "step": 85260 }, { "epoch": 9.01, - "grad_norm": 0.05892607197165489, + "grad_norm": 35.544830322265625, "learning_rate": 1.989650438272257e-06, - "loss": 0.3683, + "loss": 1.1016, "step": 85270 }, { "epoch": 9.01, - "grad_norm": 0.2950213849544525, + "grad_norm": 52.35929870605469, "learning_rate": 1.9875382828176153e-06, - "loss": 0.3916, + "loss": 0.6727, "step": 85280 }, { "epoch": 9.01, - "grad_norm": 13.2662935256958, + "grad_norm": 39.25870895385742, "learning_rate": 1.985426127362974e-06, - "loss": 0.613, + "loss": 0.9493, "step": 85290 }, { "epoch": 9.01, - "grad_norm": 31.804704666137695, + "grad_norm": 40.531280517578125, "learning_rate": 1.983313971908333e-06, - "loss": 0.3953, + "loss": 0.9144, "step": 85300 }, { "epoch": 9.01, - "grad_norm": 40.16413879394531, + "grad_norm": 35.69968032836914, "learning_rate": 1.981201816453691e-06, - "loss": 0.4161, + "loss": 0.9191, "step": 85310 }, { "epoch": 9.01, - "grad_norm": 44.28584289550781, + "grad_norm": 22.216333389282227, "learning_rate": 1.97908966099905e-06, - "loss": 0.3481, + "loss": 0.6922, "step": 85320 }, { "epoch": 9.01, - "grad_norm": 0.21953724324703217, + "grad_norm": 26.666976928710938, "learning_rate": 1.9769775055444083e-06, - "loss": 0.241, + "loss": 1.0376, "step": 85330 }, { "epoch": 9.01, - "grad_norm": 0.03941649571061134, + "grad_norm": 17.174461364746094, "learning_rate": 1.9748653500897667e-06, - "loss": 0.171, + "loss": 0.6078, "step": 85340 }, { "epoch": 9.01, - "grad_norm": 31.070547103881836, + "grad_norm": 33.64071273803711, "learning_rate": 1.9727531946351254e-06, - "loss": 0.449, + "loss": 1.0766, "step": 85350 }, { "epoch": 9.01, - "grad_norm": 21.096641540527344, + "grad_norm": 34.36848831176758, "learning_rate": 1.970641039180484e-06, - "loss": 0.3767, + "loss": 0.9315, "step": 85360 }, { "epoch": 9.02, - "grad_norm": 28.79813575744629, + "grad_norm": 58.35131072998047, "learning_rate": 1.968528883725842e-06, - "loss": 0.2105, + "loss": 0.5634, "step": 85370 }, { "epoch": 9.02, - "grad_norm": 8.098816871643066, + "grad_norm": 44.24089813232422, "learning_rate": 1.966416728271201e-06, - "loss": 0.3678, + "loss": 0.9283, "step": 85380 }, { "epoch": 9.02, - "grad_norm": 36.69988250732422, + "grad_norm": 39.94849395751953, "learning_rate": 1.9643045728165593e-06, - "loss": 0.1496, + "loss": 0.6135, "step": 85390 }, { "epoch": 9.02, - "grad_norm": 0.8610988259315491, + "grad_norm": 31.30631446838379, "learning_rate": 1.962192417361918e-06, - "loss": 0.2334, + "loss": 0.5995, "step": 85400 }, { "epoch": 9.02, - "grad_norm": 20.96742057800293, + "grad_norm": 40.46148681640625, "learning_rate": 1.9600802619072764e-06, - "loss": 0.3072, + "loss": 0.8491, "step": 85410 }, { "epoch": 9.02, - "grad_norm": 0.20374047756195068, + "grad_norm": 34.09043884277344, "learning_rate": 1.957968106452635e-06, - "loss": 0.1204, + "loss": 0.5136, "step": 85420 }, { "epoch": 9.02, - "grad_norm": 24.915454864501953, + "grad_norm": 58.684364318847656, "learning_rate": 1.9558559509979936e-06, - "loss": 0.3618, + "loss": 0.8752, "step": 85430 }, { "epoch": 9.02, - "grad_norm": 12.930524826049805, + "grad_norm": 39.55942916870117, "learning_rate": 1.9537437955433523e-06, - "loss": 0.2756, + "loss": 0.6582, "step": 85440 }, { "epoch": 9.02, - "grad_norm": 13.150190353393555, + "grad_norm": 24.068845748901367, "learning_rate": 1.9516316400887107e-06, - "loss": 0.356, + "loss": 0.7175, "step": 85450 }, { "epoch": 9.03, - "grad_norm": 20.15390968322754, + "grad_norm": 38.5447998046875, "learning_rate": 1.949519484634069e-06, - "loss": 0.1683, + "loss": 0.7768, "step": 85460 }, { "epoch": 9.03, - "grad_norm": 10.920604705810547, + "grad_norm": 27.203445434570312, "learning_rate": 1.947407329179428e-06, - "loss": 0.2994, + "loss": 0.7909, "step": 85470 }, { "epoch": 9.03, - "grad_norm": 9.070980072021484, + "grad_norm": 31.35860252380371, "learning_rate": 1.945295173724786e-06, - "loss": 0.4448, + "loss": 0.689, "step": 85480 }, { "epoch": 9.03, - "grad_norm": 40.70277404785156, + "grad_norm": 34.459163665771484, "learning_rate": 1.9431830182701445e-06, - "loss": 0.2795, + "loss": 0.6084, "step": 85490 }, { "epoch": 9.03, - "grad_norm": 31.66405487060547, + "grad_norm": 52.543663024902344, "learning_rate": 1.9410708628155033e-06, - "loss": 0.2071, + "loss": 0.5931, "step": 85500 }, { "epoch": 9.03, - "grad_norm": 12.488062858581543, + "grad_norm": 17.330289840698242, "learning_rate": 1.9389587073608617e-06, - "loss": 0.5425, + "loss": 0.7383, "step": 85510 }, { "epoch": 9.03, - "grad_norm": 29.56043243408203, + "grad_norm": 59.3057746887207, "learning_rate": 1.9368465519062205e-06, - "loss": 0.5684, + "loss": 1.0817, "step": 85520 }, { "epoch": 9.03, - "grad_norm": 9.449468612670898, + "grad_norm": 46.190643310546875, "learning_rate": 1.9347343964515792e-06, - "loss": 0.1674, + "loss": 0.5186, "step": 85530 }, { "epoch": 9.03, - "grad_norm": 1.5214742422103882, + "grad_norm": 46.17074966430664, "learning_rate": 1.9326222409969376e-06, - "loss": 0.1294, + "loss": 0.679, "step": 85540 }, { "epoch": 9.03, - "grad_norm": 0.0377691388130188, + "grad_norm": 33.27495193481445, "learning_rate": 1.9305100855422964e-06, - "loss": 0.3307, + "loss": 0.6751, "step": 85550 }, { "epoch": 9.04, - "grad_norm": 0.2533317506313324, + "grad_norm": 62.149078369140625, "learning_rate": 1.9283979300876547e-06, - "loss": 0.3347, + "loss": 1.1085, "step": 85560 }, { "epoch": 9.04, - "grad_norm": 0.5735855102539062, + "grad_norm": 26.206485748291016, "learning_rate": 1.926285774633013e-06, - "loss": 0.3526, + "loss": 0.8711, "step": 85570 }, { "epoch": 9.04, - "grad_norm": 5.493261337280273, + "grad_norm": 40.474361419677734, "learning_rate": 1.924173619178372e-06, - "loss": 0.4207, + "loss": 0.7234, "step": 85580 }, { "epoch": 9.04, - "grad_norm": 12.39021110534668, + "grad_norm": 37.663753509521484, "learning_rate": 1.92206146372373e-06, - "loss": 0.4706, + "loss": 0.5992, "step": 85590 }, { "epoch": 9.04, - "grad_norm": 23.14971160888672, + "grad_norm": 33.273521423339844, "learning_rate": 1.9199493082690886e-06, - "loss": 0.5092, + "loss": 0.901, "step": 85600 }, { "epoch": 9.04, - "grad_norm": 12.222249031066895, + "grad_norm": 40.289554595947266, "learning_rate": 1.9178371528144473e-06, - "loss": 0.5226, + "loss": 0.883, "step": 85610 }, { "epoch": 9.04, - "grad_norm": 0.04108913987874985, + "grad_norm": 26.4814453125, "learning_rate": 1.9157249973598057e-06, - "loss": 0.1894, + "loss": 0.7968, "step": 85620 }, { "epoch": 9.04, - "grad_norm": 25.13408660888672, + "grad_norm": 41.41611099243164, "learning_rate": 1.9136128419051645e-06, - "loss": 0.4673, + "loss": 1.0602, "step": 85630 }, { "epoch": 9.04, - "grad_norm": 12.369742393493652, + "grad_norm": 50.90177917480469, "learning_rate": 1.911500686450523e-06, - "loss": 0.1671, + "loss": 0.77, "step": 85640 }, { "epoch": 9.05, - "grad_norm": 20.021760940551758, + "grad_norm": 19.776296615600586, "learning_rate": 1.9093885309958816e-06, - "loss": 0.0803, + "loss": 0.433, "step": 85650 }, { "epoch": 9.05, - "grad_norm": 7.682570934295654, + "grad_norm": 41.477210998535156, "learning_rate": 1.90727637554124e-06, - "loss": 0.1856, + "loss": 0.8282, "step": 85660 }, { "epoch": 9.05, - "grad_norm": 22.854690551757812, + "grad_norm": 41.49897766113281, "learning_rate": 1.9051642200865985e-06, - "loss": 0.2743, + "loss": 0.9313, "step": 85670 }, { "epoch": 9.05, - "grad_norm": 8.379915237426758, + "grad_norm": 35.76242446899414, "learning_rate": 1.9030520646319569e-06, - "loss": 0.4401, + "loss": 0.8477, "step": 85680 }, { "epoch": 9.05, - "grad_norm": 18.819698333740234, + "grad_norm": 26.609310150146484, "learning_rate": 1.9009399091773157e-06, - "loss": 0.4486, + "loss": 0.6963, "step": 85690 }, { "epoch": 9.05, - "grad_norm": 39.851646423339844, + "grad_norm": 90.31665802001953, "learning_rate": 1.898827753722674e-06, - "loss": 0.3374, + "loss": 1.2333, "step": 85700 }, { "epoch": 9.05, - "grad_norm": 6.122716903686523, + "grad_norm": 61.714534759521484, "learning_rate": 1.8967155982680328e-06, - "loss": 0.2705, + "loss": 0.9214, "step": 85710 }, { "epoch": 9.05, - "grad_norm": 11.135601043701172, + "grad_norm": 35.944557189941406, "learning_rate": 1.8946034428133912e-06, - "loss": 0.6846, + "loss": 0.9169, "step": 85720 }, { "epoch": 9.05, - "grad_norm": 23.73712730407715, + "grad_norm": 53.309024810791016, "learning_rate": 1.8924912873587497e-06, - "loss": 0.5094, + "loss": 0.9992, "step": 85730 }, { "epoch": 9.05, - "grad_norm": 16.12652015686035, + "grad_norm": 36.93429946899414, "learning_rate": 1.8903791319041085e-06, - "loss": 0.4331, + "loss": 1.0485, "step": 85740 }, { "epoch": 9.06, - "grad_norm": 1.8916215896606445, + "grad_norm": 28.024402618408203, "learning_rate": 1.8882669764494669e-06, - "loss": 0.1007, + "loss": 0.5774, "step": 85750 }, { "epoch": 9.06, - "grad_norm": 49.673583984375, + "grad_norm": 37.38767623901367, "learning_rate": 1.8861548209948254e-06, - "loss": 0.3841, + "loss": 1.2402, "step": 85760 }, { "epoch": 9.06, - "grad_norm": 2.7122836112976074, + "grad_norm": 25.62535858154297, "learning_rate": 1.884042665540184e-06, - "loss": 0.2077, + "loss": 1.0562, "step": 85770 }, { "epoch": 9.06, - "grad_norm": 18.829586029052734, + "grad_norm": 45.83534622192383, "learning_rate": 1.8819305100855426e-06, - "loss": 0.4582, + "loss": 1.2019, "step": 85780 }, { "epoch": 9.06, - "grad_norm": 29.307022094726562, + "grad_norm": 26.436120986938477, "learning_rate": 1.879818354630901e-06, - "loss": 0.3618, + "loss": 0.8973, "step": 85790 }, { "epoch": 9.06, - "grad_norm": 49.2685661315918, + "grad_norm": 49.11906051635742, "learning_rate": 1.8777061991762597e-06, - "loss": 0.3427, + "loss": 0.9949, "step": 85800 }, { "epoch": 9.06, - "grad_norm": 0.06752130389213562, + "grad_norm": 29.533409118652344, "learning_rate": 1.875594043721618e-06, - "loss": 0.1908, + "loss": 0.7243, "step": 85810 }, { "epoch": 9.06, - "grad_norm": 2.1450631618499756, + "grad_norm": 14.780553817749023, "learning_rate": 1.8734818882669766e-06, - "loss": 0.2855, + "loss": 0.7498, "step": 85820 }, { "epoch": 9.06, - "grad_norm": 9.940571784973145, + "grad_norm": 30.605371475219727, "learning_rate": 1.8713697328123352e-06, - "loss": 0.189, + "loss": 0.9226, "step": 85830 }, { "epoch": 9.07, - "grad_norm": 2.6305768489837646, + "grad_norm": 33.158878326416016, "learning_rate": 1.8692575773576938e-06, - "loss": 0.3697, + "loss": 0.9552, "step": 85840 }, { "epoch": 9.07, - "grad_norm": 0.024202430620789528, + "grad_norm": 36.04365158081055, "learning_rate": 1.8671454219030521e-06, - "loss": 0.2657, + "loss": 0.7758, "step": 85850 }, { "epoch": 9.07, - "grad_norm": 0.9732239842414856, + "grad_norm": 37.302608489990234, "learning_rate": 1.8650332664484109e-06, - "loss": 0.3198, + "loss": 0.698, "step": 85860 }, { "epoch": 9.07, - "grad_norm": 25.60129165649414, + "grad_norm": 52.0416145324707, "learning_rate": 1.8629211109937692e-06, - "loss": 0.2906, + "loss": 0.759, "step": 85870 }, { "epoch": 9.07, - "grad_norm": 0.0366046316921711, + "grad_norm": 36.03324508666992, "learning_rate": 1.8608089555391278e-06, - "loss": 0.2335, + "loss": 0.4823, "step": 85880 }, { "epoch": 9.07, - "grad_norm": 1.4381353855133057, + "grad_norm": 36.450408935546875, "learning_rate": 1.8586968000844864e-06, - "loss": 0.2128, + "loss": 0.8741, "step": 85890 }, { "epoch": 9.07, - "grad_norm": 35.14651870727539, + "grad_norm": 58.938148498535156, "learning_rate": 1.856584644629845e-06, - "loss": 0.3395, + "loss": 1.0625, "step": 85900 }, { "epoch": 9.07, - "grad_norm": 37.97307586669922, + "grad_norm": 56.68923568725586, "learning_rate": 1.8544724891752033e-06, - "loss": 0.4388, + "loss": 1.3608, "step": 85910 }, { "epoch": 9.07, - "grad_norm": 0.3533848226070404, + "grad_norm": 41.13768005371094, "learning_rate": 1.852360333720562e-06, - "loss": 0.7915, + "loss": 1.022, "step": 85920 }, { "epoch": 9.07, - "grad_norm": 7.565413951873779, + "grad_norm": 30.61220932006836, "learning_rate": 1.8502481782659204e-06, - "loss": 0.5078, + "loss": 0.8588, "step": 85930 }, { "epoch": 9.08, - "grad_norm": 0.6755300760269165, + "grad_norm": 33.20631790161133, "learning_rate": 1.848136022811279e-06, - "loss": 0.3007, + "loss": 0.7104, "step": 85940 }, { "epoch": 9.08, - "grad_norm": 0.3878154754638672, + "grad_norm": 26.096696853637695, "learning_rate": 1.8460238673566376e-06, - "loss": 0.2104, + "loss": 0.609, "step": 85950 }, { "epoch": 9.08, - "grad_norm": 10.639801025390625, + "grad_norm": 5.146009922027588, "learning_rate": 1.8439117119019961e-06, - "loss": 0.346, + "loss": 0.672, "step": 85960 }, { "epoch": 9.08, - "grad_norm": 0.041313499212265015, + "grad_norm": 28.1271915435791, "learning_rate": 1.841799556447355e-06, - "loss": 0.2331, + "loss": 0.7207, "step": 85970 }, { "epoch": 9.08, - "grad_norm": 4.42661190032959, + "grad_norm": 60.03956985473633, "learning_rate": 1.8396874009927133e-06, - "loss": 0.4611, + "loss": 0.6768, "step": 85980 }, { "epoch": 9.08, - "grad_norm": 23.885334014892578, + "grad_norm": 25.738752365112305, "learning_rate": 1.8375752455380718e-06, - "loss": 0.3151, + "loss": 0.988, "step": 85990 }, { "epoch": 9.08, - "grad_norm": 18.726869583129883, + "grad_norm": 49.78805160522461, "learning_rate": 1.8354630900834302e-06, - "loss": 0.2817, + "loss": 0.7848, "step": 86000 }, { "epoch": 9.08, - "grad_norm": 43.549373626708984, + "grad_norm": 30.290008544921875, "learning_rate": 1.833350934628789e-06, - "loss": 0.3467, + "loss": 0.7507, "step": 86010 }, { "epoch": 9.08, - "grad_norm": 7.818694114685059, + "grad_norm": 38.70718002319336, "learning_rate": 1.8312387791741473e-06, - "loss": 0.2262, + "loss": 0.7484, "step": 86020 }, { "epoch": 9.09, - "grad_norm": 15.61884593963623, + "grad_norm": 26.9303035736084, "learning_rate": 1.829126623719506e-06, - "loss": 0.3671, + "loss": 0.8276, "step": 86030 }, { "epoch": 9.09, - "grad_norm": 14.231195449829102, + "grad_norm": 34.928218841552734, "learning_rate": 1.8270144682648645e-06, - "loss": 0.2312, + "loss": 1.0408, "step": 86040 }, { "epoch": 9.09, - "grad_norm": 1.5766385793685913, + "grad_norm": 37.16650390625, "learning_rate": 1.824902312810223e-06, - "loss": 0.2506, + "loss": 0.7405, "step": 86050 }, { "epoch": 9.09, - "grad_norm": 0.1331988126039505, + "grad_norm": 15.334647178649902, "learning_rate": 1.8227901573555814e-06, - "loss": 0.3947, + "loss": 0.842, "step": 86060 }, { "epoch": 9.09, - "grad_norm": 31.144784927368164, + "grad_norm": 40.417301177978516, "learning_rate": 1.8206780019009402e-06, - "loss": 0.3658, + "loss": 0.841, "step": 86070 }, { "epoch": 9.09, - "grad_norm": 1.70386803150177, + "grad_norm": 33.4033088684082, "learning_rate": 1.8185658464462985e-06, - "loss": 0.4125, + "loss": 0.7356, "step": 86080 }, { "epoch": 9.09, - "grad_norm": 25.090164184570312, + "grad_norm": 51.41974639892578, "learning_rate": 1.8164536909916573e-06, - "loss": 0.3145, + "loss": 0.8609, "step": 86090 }, { "epoch": 9.09, - "grad_norm": 0.23820187151432037, + "grad_norm": 43.718017578125, "learning_rate": 1.8143415355370156e-06, - "loss": 0.3198, + "loss": 0.9662, "step": 86100 }, { "epoch": 9.09, - "grad_norm": 2.240875482559204, + "grad_norm": 36.0889778137207, "learning_rate": 1.8122293800823742e-06, - "loss": 0.3269, + "loss": 0.7449, "step": 86110 }, { "epoch": 9.09, - "grad_norm": 11.159897804260254, + "grad_norm": 30.843839645385742, "learning_rate": 1.8101172246277326e-06, - "loss": 0.4025, + "loss": 1.237, "step": 86120 }, { "epoch": 9.1, - "grad_norm": 6.668144226074219, + "grad_norm": 50.3123779296875, "learning_rate": 1.8080050691730913e-06, - "loss": 0.3158, + "loss": 0.9624, "step": 86130 }, { "epoch": 9.1, - "grad_norm": 4.067257881164551, + "grad_norm": 41.89313507080078, "learning_rate": 1.8058929137184497e-06, - "loss": 0.2347, + "loss": 0.8884, "step": 86140 }, { "epoch": 9.1, - "grad_norm": 21.310100555419922, + "grad_norm": 53.313323974609375, "learning_rate": 1.8037807582638085e-06, - "loss": 0.3916, + "loss": 0.8305, "step": 86150 }, { "epoch": 9.1, - "grad_norm": 5.963346004486084, + "grad_norm": 38.1559944152832, "learning_rate": 1.8016686028091668e-06, - "loss": 0.1559, + "loss": 0.7674, "step": 86160 }, { "epoch": 9.1, - "grad_norm": 7.630881309509277, + "grad_norm": 40.78841018676758, "learning_rate": 1.7995564473545254e-06, - "loss": 0.4847, + "loss": 0.6539, "step": 86170 }, { "epoch": 9.1, - "grad_norm": 40.85802459716797, + "grad_norm": 31.41204071044922, "learning_rate": 1.7974442918998838e-06, - "loss": 0.4464, + "loss": 1.1454, "step": 86180 }, { "epoch": 9.1, - "grad_norm": 16.095787048339844, + "grad_norm": 29.014814376831055, "learning_rate": 1.7953321364452425e-06, - "loss": 0.367, + "loss": 0.9971, "step": 86190 }, { "epoch": 9.1, - "grad_norm": 0.8330962061882019, + "grad_norm": 24.2965030670166, "learning_rate": 1.7932199809906011e-06, - "loss": 0.4447, + "loss": 0.8575, "step": 86200 }, { "epoch": 9.1, - "grad_norm": 22.460346221923828, + "grad_norm": 38.3681526184082, "learning_rate": 1.7911078255359597e-06, - "loss": 0.2532, + "loss": 0.6931, "step": 86210 }, { "epoch": 9.11, - "grad_norm": 0.693892240524292, + "grad_norm": 53.6851806640625, "learning_rate": 1.7889956700813182e-06, - "loss": 0.5167, + "loss": 1.2292, "step": 86220 }, { "epoch": 9.11, - "grad_norm": 17.0240535736084, + "grad_norm": 24.82383918762207, "learning_rate": 1.7868835146266766e-06, - "loss": 0.3168, + "loss": 0.9216, "step": 86230 }, { "epoch": 9.11, - "grad_norm": 12.667366027832031, + "grad_norm": 30.61351203918457, "learning_rate": 1.7847713591720354e-06, - "loss": 0.5308, + "loss": 0.8477, "step": 86240 }, { "epoch": 9.11, - "grad_norm": 1.8634850978851318, + "grad_norm": 22.10991668701172, "learning_rate": 1.7826592037173937e-06, - "loss": 0.204, + "loss": 0.8057, "step": 86250 }, { "epoch": 9.11, - "grad_norm": 7.811666488647461, + "grad_norm": 79.18370056152344, "learning_rate": 1.7805470482627523e-06, - "loss": 0.4032, + "loss": 0.8173, "step": 86260 }, { "epoch": 9.11, - "grad_norm": 6.114731311798096, + "grad_norm": 37.0032844543457, "learning_rate": 1.7784348928081109e-06, - "loss": 0.1857, + "loss": 0.8224, "step": 86270 }, { "epoch": 9.11, - "grad_norm": 5.660872936248779, + "grad_norm": 73.47157287597656, "learning_rate": 1.7763227373534694e-06, - "loss": 0.143, + "loss": 0.6526, "step": 86280 }, { "epoch": 9.11, - "grad_norm": 2.0063135623931885, + "grad_norm": 40.70072937011719, "learning_rate": 1.7742105818988278e-06, - "loss": 0.3005, + "loss": 0.7005, "step": 86290 }, { "epoch": 9.11, - "grad_norm": 12.135607719421387, + "grad_norm": 33.954200744628906, "learning_rate": 1.7720984264441866e-06, - "loss": 0.4528, + "loss": 0.8792, "step": 86300 }, { "epoch": 9.12, - "grad_norm": 8.875442504882812, + "grad_norm": 45.192935943603516, "learning_rate": 1.769986270989545e-06, - "loss": 0.3627, + "loss": 1.0528, "step": 86310 }, { "epoch": 9.12, - "grad_norm": 20.93657875061035, + "grad_norm": 28.580821990966797, "learning_rate": 1.7678741155349035e-06, - "loss": 0.5099, + "loss": 0.7033, "step": 86320 }, { "epoch": 9.12, - "grad_norm": 1.1902164220809937, + "grad_norm": 75.03001403808594, "learning_rate": 1.765761960080262e-06, - "loss": 0.2756, + "loss": 1.0931, "step": 86330 }, { "epoch": 9.12, - "grad_norm": 10.608695983886719, + "grad_norm": 33.00221633911133, "learning_rate": 1.7636498046256206e-06, - "loss": 0.1123, + "loss": 0.5421, "step": 86340 }, { "epoch": 9.12, - "grad_norm": 9.757363319396973, + "grad_norm": 40.48065185546875, "learning_rate": 1.761537649170979e-06, - "loss": 0.2967, + "loss": 0.7116, "step": 86350 }, { "epoch": 9.12, - "grad_norm": 31.037944793701172, + "grad_norm": 5.1185622215271, "learning_rate": 1.7594254937163378e-06, - "loss": 0.3196, + "loss": 0.4941, "step": 86360 }, { "epoch": 9.12, - "grad_norm": 17.339590072631836, + "grad_norm": 42.945491790771484, "learning_rate": 1.7573133382616961e-06, - "loss": 0.3038, + "loss": 0.8355, "step": 86370 }, { "epoch": 9.12, - "grad_norm": 6.888354301452637, + "grad_norm": 67.2388687133789, "learning_rate": 1.7552011828070547e-06, - "loss": 0.2129, + "loss": 0.9346, "step": 86380 }, { "epoch": 9.12, - "grad_norm": 6.339604377746582, + "grad_norm": 32.65815353393555, "learning_rate": 1.7530890273524132e-06, - "loss": 0.3088, + "loss": 0.738, "step": 86390 }, { "epoch": 9.12, - "grad_norm": 33.00740432739258, + "grad_norm": 47.96625900268555, "learning_rate": 1.7509768718977718e-06, - "loss": 0.4201, + "loss": 0.6824, "step": 86400 }, { "epoch": 9.13, - "grad_norm": 3.295837163925171, + "grad_norm": 36.55175018310547, "learning_rate": 1.7488647164431306e-06, - "loss": 0.1485, + "loss": 0.51, "step": 86410 }, { "epoch": 9.13, - "grad_norm": 11.87736701965332, + "grad_norm": 47.79948806762695, "learning_rate": 1.746752560988489e-06, - "loss": 0.2138, + "loss": 1.022, "step": 86420 }, { "epoch": 9.13, - "grad_norm": 28.848342895507812, + "grad_norm": 46.002777099609375, "learning_rate": 1.7446404055338475e-06, - "loss": 0.3913, + "loss": 0.9149, "step": 86430 }, { "epoch": 9.13, - "grad_norm": 41.082149505615234, + "grad_norm": 69.37648010253906, "learning_rate": 1.7425282500792059e-06, - "loss": 0.1467, + "loss": 0.8948, "step": 86440 }, { "epoch": 9.13, - "grad_norm": 15.789936065673828, + "grad_norm": 40.12652587890625, "learning_rate": 1.7404160946245646e-06, - "loss": 0.373, + "loss": 0.9691, "step": 86450 }, { "epoch": 9.13, - "grad_norm": 14.559429168701172, + "grad_norm": 48.73778533935547, "learning_rate": 1.738303939169923e-06, - "loss": 0.2717, + "loss": 0.923, "step": 86460 }, { "epoch": 9.13, - "grad_norm": 2.532136917114258, + "grad_norm": 41.52323532104492, "learning_rate": 1.7361917837152818e-06, - "loss": 0.2506, + "loss": 0.9187, "step": 86470 }, { "epoch": 9.13, - "grad_norm": 36.83451843261719, + "grad_norm": 55.42261505126953, "learning_rate": 1.7340796282606401e-06, - "loss": 0.5091, + "loss": 0.7923, "step": 86480 }, { "epoch": 9.13, - "grad_norm": 23.71908950805664, + "grad_norm": 63.92594528198242, "learning_rate": 1.7319674728059987e-06, - "loss": 0.4282, + "loss": 0.7387, "step": 86490 }, { "epoch": 9.14, - "grad_norm": 0.7385298013687134, + "grad_norm": 13.988326072692871, "learning_rate": 1.729855317351357e-06, - "loss": 0.1306, + "loss": 0.8446, "step": 86500 }, { "epoch": 9.14, - "grad_norm": 0.5954514741897583, + "grad_norm": 33.66005325317383, "learning_rate": 1.7277431618967158e-06, - "loss": 0.4706, + "loss": 0.7191, "step": 86510 }, { "epoch": 9.14, - "grad_norm": 0.01945386454463005, + "grad_norm": 27.6021785736084, "learning_rate": 1.7256310064420742e-06, - "loss": 0.28, + "loss": 0.7631, "step": 86520 }, { "epoch": 9.14, - "grad_norm": 5.8449835777282715, + "grad_norm": 39.69367218017578, "learning_rate": 1.723518850987433e-06, - "loss": 0.1996, + "loss": 0.7534, "step": 86530 }, { "epoch": 9.14, - "grad_norm": 11.920300483703613, + "grad_norm": 39.466732025146484, "learning_rate": 1.7214066955327913e-06, - "loss": 0.6303, + "loss": 1.1419, "step": 86540 }, { "epoch": 9.14, - "grad_norm": 5.958524703979492, + "grad_norm": 26.01089859008789, "learning_rate": 1.71929454007815e-06, - "loss": 0.3576, + "loss": 0.7726, "step": 86550 }, { "epoch": 9.14, - "grad_norm": 42.92369079589844, + "grad_norm": 28.668350219726562, "learning_rate": 1.7171823846235083e-06, - "loss": 0.6938, + "loss": 0.9487, "step": 86560 }, { "epoch": 9.14, - "grad_norm": 16.95383071899414, + "grad_norm": 37.308067321777344, "learning_rate": 1.715070229168867e-06, - "loss": 0.6871, + "loss": 0.9196, "step": 86570 }, { "epoch": 9.14, - "grad_norm": 7.696703910827637, + "grad_norm": 56.281864166259766, "learning_rate": 1.7129580737142254e-06, - "loss": 0.3163, + "loss": 0.9199, "step": 86580 }, { "epoch": 9.14, - "grad_norm": 21.236604690551758, + "grad_norm": 53.045166015625, "learning_rate": 1.7108459182595842e-06, - "loss": 0.6095, + "loss": 0.8332, "step": 86590 }, { "epoch": 9.15, - "grad_norm": 6.271078109741211, + "grad_norm": 41.70802307128906, "learning_rate": 1.7087337628049425e-06, - "loss": 0.4647, + "loss": 0.8639, "step": 86600 }, { "epoch": 9.15, - "grad_norm": 7.796981334686279, + "grad_norm": 54.52933120727539, "learning_rate": 1.706621607350301e-06, - "loss": 0.3026, + "loss": 0.9421, "step": 86610 }, { "epoch": 9.15, - "grad_norm": 12.642619132995605, + "grad_norm": 36.091880798339844, "learning_rate": 1.7045094518956594e-06, - "loss": 0.3075, + "loss": 1.2585, "step": 86620 }, { "epoch": 9.15, - "grad_norm": 10.473370552062988, + "grad_norm": 41.70355987548828, "learning_rate": 1.7023972964410182e-06, - "loss": 0.2489, + "loss": 1.0224, "step": 86630 }, { "epoch": 9.15, - "grad_norm": 1.6733001470565796, + "grad_norm": 57.474613189697266, "learning_rate": 1.7002851409863768e-06, - "loss": 0.2204, + "loss": 0.6061, "step": 86640 }, { "epoch": 9.15, - "grad_norm": 20.677242279052734, + "grad_norm": 46.027774810791016, "learning_rate": 1.6981729855317354e-06, - "loss": 0.6103, + "loss": 1.2386, "step": 86650 }, { "epoch": 9.15, - "grad_norm": 8.68224048614502, + "grad_norm": 53.30100631713867, "learning_rate": 1.696060830077094e-06, - "loss": 0.4294, + "loss": 0.7645, "step": 86660 }, { "epoch": 9.15, - "grad_norm": 8.917902946472168, + "grad_norm": 43.496788024902344, "learning_rate": 1.6939486746224523e-06, - "loss": 0.3404, + "loss": 0.9084, "step": 86670 }, { "epoch": 9.15, - "grad_norm": 9.046239852905273, + "grad_norm": 32.528472900390625, "learning_rate": 1.691836519167811e-06, - "loss": 0.2212, + "loss": 0.7121, "step": 86680 }, { "epoch": 9.16, - "grad_norm": 8.225378036499023, + "grad_norm": 65.04573822021484, "learning_rate": 1.6897243637131694e-06, - "loss": 0.438, + "loss": 1.0173, "step": 86690 }, { "epoch": 9.16, - "grad_norm": 0.4617297351360321, + "grad_norm": 35.45048904418945, "learning_rate": 1.687612208258528e-06, - "loss": 0.3873, + "loss": 0.6944, "step": 86700 }, { "epoch": 9.16, - "grad_norm": 21.36789894104004, + "grad_norm": 21.71835708618164, "learning_rate": 1.6855000528038865e-06, - "loss": 0.3338, + "loss": 0.6849, "step": 86710 }, { "epoch": 9.16, - "grad_norm": 0.056357648223638535, + "grad_norm": 35.747833251953125, "learning_rate": 1.6833878973492451e-06, - "loss": 0.4881, + "loss": 1.0003, "step": 86720 }, { "epoch": 9.16, - "grad_norm": 12.250275611877441, + "grad_norm": 41.21940612792969, "learning_rate": 1.6812757418946035e-06, - "loss": 0.2183, + "loss": 1.0023, "step": 86730 }, { "epoch": 9.16, - "grad_norm": 13.37425708770752, + "grad_norm": 42.69206237792969, "learning_rate": 1.6791635864399622e-06, - "loss": 0.2675, + "loss": 0.6226, "step": 86740 }, { "epoch": 9.16, - "grad_norm": 28.51614761352539, + "grad_norm": 15.072918891906738, "learning_rate": 1.6770514309853206e-06, - "loss": 0.4167, + "loss": 0.885, "step": 86750 }, { "epoch": 9.16, - "grad_norm": 14.114238739013672, + "grad_norm": 52.017940521240234, "learning_rate": 1.6749392755306792e-06, - "loss": 0.5407, + "loss": 0.9634, "step": 86760 }, { "epoch": 9.16, - "grad_norm": 19.55400848388672, + "grad_norm": 47.37761306762695, "learning_rate": 1.6728271200760377e-06, - "loss": 0.2608, + "loss": 1.1614, "step": 86770 }, { "epoch": 9.16, - "grad_norm": 9.603487968444824, + "grad_norm": 25.323177337646484, "learning_rate": 1.6707149646213963e-06, - "loss": 0.2645, + "loss": 0.6317, "step": 86780 }, { "epoch": 9.17, - "grad_norm": 6.375642776489258, + "grad_norm": 66.01636505126953, "learning_rate": 1.6686028091667547e-06, - "loss": 0.4081, + "loss": 0.9245, "step": 86790 }, { "epoch": 9.17, - "grad_norm": 9.940678596496582, + "grad_norm": 35.808536529541016, "learning_rate": 1.6664906537121134e-06, - "loss": 0.2041, + "loss": 0.7998, "step": 86800 }, { "epoch": 9.17, - "grad_norm": 9.709900856018066, + "grad_norm": 42.60888671875, "learning_rate": 1.6643784982574718e-06, - "loss": 0.1788, + "loss": 0.6737, "step": 86810 }, { "epoch": 9.17, - "grad_norm": 0.11085433512926102, + "grad_norm": 15.267931938171387, "learning_rate": 1.6622663428028304e-06, - "loss": 0.3761, + "loss": 0.6785, "step": 86820 }, { "epoch": 9.17, - "grad_norm": 6.070061206817627, + "grad_norm": 25.40886116027832, "learning_rate": 1.660154187348189e-06, - "loss": 0.2609, + "loss": 0.5745, "step": 86830 }, { "epoch": 9.17, - "grad_norm": 29.975231170654297, + "grad_norm": 53.48643112182617, "learning_rate": 1.6580420318935475e-06, - "loss": 0.6233, + "loss": 0.7933, "step": 86840 }, { "epoch": 9.17, - "grad_norm": 10.051937103271484, + "grad_norm": 9.625658988952637, "learning_rate": 1.6559298764389059e-06, - "loss": 0.1301, + "loss": 0.5274, "step": 86850 }, { "epoch": 9.17, - "grad_norm": 5.182826519012451, + "grad_norm": 30.338260650634766, "learning_rate": 1.6538177209842646e-06, - "loss": 0.415, + "loss": 0.9546, "step": 86860 }, { "epoch": 9.17, - "grad_norm": 0.06090494990348816, + "grad_norm": 39.39092254638672, "learning_rate": 1.6517055655296232e-06, - "loss": 0.3847, + "loss": 0.8743, "step": 86870 }, { "epoch": 9.18, - "grad_norm": 3.693413019180298, + "grad_norm": 26.59549331665039, "learning_rate": 1.6495934100749816e-06, - "loss": 0.3113, + "loss": 0.6691, "step": 86880 }, { "epoch": 9.18, - "grad_norm": 27.15974998474121, + "grad_norm": 47.85123062133789, "learning_rate": 1.6474812546203403e-06, - "loss": 0.4524, + "loss": 1.1606, "step": 86890 }, { "epoch": 9.18, - "grad_norm": 16.497703552246094, + "grad_norm": 34.46141052246094, "learning_rate": 1.6453690991656987e-06, - "loss": 0.1376, + "loss": 0.5571, "step": 86900 }, { "epoch": 9.18, - "grad_norm": 18.60405158996582, + "grad_norm": 13.67829418182373, "learning_rate": 1.6432569437110575e-06, - "loss": 0.2327, + "loss": 0.621, "step": 86910 }, { "epoch": 9.18, - "grad_norm": 23.446826934814453, + "grad_norm": 35.198238372802734, "learning_rate": 1.6411447882564158e-06, - "loss": 0.2623, + "loss": 0.7688, "step": 86920 }, { "epoch": 9.18, - "grad_norm": 0.8650422096252441, + "grad_norm": 8.846382141113281, "learning_rate": 1.6390326328017744e-06, - "loss": 0.3172, + "loss": 1.0769, "step": 86930 }, { "epoch": 9.18, - "grad_norm": 0.17261789739131927, + "grad_norm": 18.494726181030273, "learning_rate": 1.6369204773471327e-06, - "loss": 0.3835, + "loss": 0.4778, "step": 86940 }, { "epoch": 9.18, - "grad_norm": 10.973858833312988, + "grad_norm": 42.98992156982422, "learning_rate": 1.6348083218924915e-06, - "loss": 0.1436, + "loss": 0.7605, "step": 86950 }, { "epoch": 9.18, - "grad_norm": 28.70472526550293, + "grad_norm": 40.3013801574707, "learning_rate": 1.6326961664378499e-06, - "loss": 0.5925, + "loss": 0.9007, "step": 86960 }, { "epoch": 9.18, - "grad_norm": 27.599933624267578, + "grad_norm": 32.997283935546875, "learning_rate": 1.6305840109832087e-06, - "loss": 0.3184, + "loss": 0.8896, "step": 86970 }, { "epoch": 9.19, - "grad_norm": 0.7445393800735474, + "grad_norm": 20.007055282592773, "learning_rate": 1.628471855528567e-06, - "loss": 0.1188, + "loss": 0.9093, "step": 86980 }, { "epoch": 9.19, - "grad_norm": 3.277397871017456, + "grad_norm": 29.739145278930664, "learning_rate": 1.6263597000739256e-06, - "loss": 0.4675, + "loss": 0.9302, "step": 86990 }, { "epoch": 9.19, - "grad_norm": 2.845154285430908, + "grad_norm": 61.42828369140625, "learning_rate": 1.624247544619284e-06, - "loss": 0.2214, + "loss": 0.8544, "step": 87000 }, { "epoch": 9.19, - "grad_norm": 22.378694534301758, + "grad_norm": 55.636695861816406, "learning_rate": 1.6221353891646427e-06, - "loss": 0.2114, + "loss": 1.0141, "step": 87010 }, { "epoch": 9.19, - "grad_norm": 0.01573490910232067, + "grad_norm": 35.427268981933594, "learning_rate": 1.620023233710001e-06, - "loss": 0.1749, + "loss": 0.7536, "step": 87020 }, { "epoch": 9.19, - "grad_norm": 12.866598129272461, + "grad_norm": 29.219619750976562, "learning_rate": 1.6179110782553598e-06, - "loss": 0.1808, + "loss": 0.6233, "step": 87030 }, { "epoch": 9.19, - "grad_norm": 24.56196403503418, + "grad_norm": 39.7322883605957, "learning_rate": 1.6157989228007182e-06, - "loss": 0.3929, + "loss": 1.0708, "step": 87040 }, { "epoch": 9.19, - "grad_norm": 13.38757610321045, + "grad_norm": 29.382829666137695, "learning_rate": 1.6136867673460768e-06, - "loss": 0.2119, + "loss": 0.6584, "step": 87050 }, { "epoch": 9.19, - "grad_norm": 21.262691497802734, + "grad_norm": 19.545501708984375, "learning_rate": 1.6115746118914351e-06, - "loss": 0.1738, + "loss": 0.6574, "step": 87060 }, { "epoch": 9.2, - "grad_norm": 40.178836822509766, + "grad_norm": 35.67866516113281, "learning_rate": 1.609462456436794e-06, - "loss": 0.2531, + "loss": 0.9061, "step": 87070 }, { "epoch": 9.2, - "grad_norm": 0.06948734819889069, + "grad_norm": 27.070941925048828, "learning_rate": 1.6073503009821525e-06, - "loss": 0.3017, + "loss": 0.9105, "step": 87080 }, { "epoch": 9.2, - "grad_norm": 2.298297643661499, + "grad_norm": 50.317909240722656, "learning_rate": 1.605238145527511e-06, - "loss": 0.316, + "loss": 0.7511, "step": 87090 }, { "epoch": 9.2, - "grad_norm": 6.53287410736084, + "grad_norm": 32.199546813964844, "learning_rate": 1.6031259900728696e-06, - "loss": 0.1439, + "loss": 0.5962, "step": 87100 }, { "epoch": 9.2, - "grad_norm": 0.21779730916023254, + "grad_norm": 17.666419982910156, "learning_rate": 1.601013834618228e-06, - "loss": 0.5562, + "loss": 0.9938, "step": 87110 }, { "epoch": 9.2, - "grad_norm": 5.341863632202148, + "grad_norm": 22.515642166137695, "learning_rate": 1.5989016791635867e-06, - "loss": 0.1961, + "loss": 0.713, "step": 87120 }, { "epoch": 9.2, - "grad_norm": 16.025894165039062, + "grad_norm": 25.468320846557617, "learning_rate": 1.596789523708945e-06, - "loss": 0.2864, + "loss": 0.8505, "step": 87130 }, { "epoch": 9.2, - "grad_norm": 13.469792366027832, + "grad_norm": 25.43486976623535, "learning_rate": 1.5946773682543037e-06, - "loss": 0.1277, + "loss": 0.7425, "step": 87140 }, { "epoch": 9.2, - "grad_norm": 22.01038932800293, + "grad_norm": 41.374794006347656, "learning_rate": 1.5925652127996622e-06, - "loss": 0.321, + "loss": 0.6865, "step": 87150 }, { "epoch": 9.2, - "grad_norm": 18.965721130371094, + "grad_norm": 37.627113342285156, "learning_rate": 1.5904530573450208e-06, - "loss": 0.3965, + "loss": 0.7766, "step": 87160 }, { "epoch": 9.21, - "grad_norm": 14.411699295043945, + "grad_norm": 21.002050399780273, "learning_rate": 1.5883409018903792e-06, - "loss": 0.523, + "loss": 0.768, "step": 87170 }, { "epoch": 9.21, - "grad_norm": 6.288346767425537, + "grad_norm": 37.25080871582031, "learning_rate": 1.586228746435738e-06, - "loss": 0.5081, + "loss": 0.9428, "step": 87180 }, { "epoch": 9.21, - "grad_norm": 36.546112060546875, + "grad_norm": 58.23637008666992, "learning_rate": 1.5841165909810963e-06, - "loss": 0.5421, + "loss": 0.891, "step": 87190 }, { "epoch": 9.21, - "grad_norm": 43.564998626708984, + "grad_norm": 31.07380485534668, "learning_rate": 1.5820044355264549e-06, - "loss": 0.1494, + "loss": 0.5393, "step": 87200 }, { "epoch": 9.21, - "grad_norm": 2.119225025177002, + "grad_norm": 64.84526824951172, "learning_rate": 1.5798922800718134e-06, - "loss": 0.2034, + "loss": 0.6962, "step": 87210 }, { "epoch": 9.21, - "grad_norm": 13.889885902404785, + "grad_norm": 61.621028900146484, "learning_rate": 1.577780124617172e-06, - "loss": 0.2151, + "loss": 1.0145, "step": 87220 }, { "epoch": 9.21, - "grad_norm": 8.019906044006348, + "grad_norm": 45.848350524902344, "learning_rate": 1.5756679691625303e-06, - "loss": 0.326, + "loss": 1.1383, "step": 87230 }, { "epoch": 9.21, - "grad_norm": 20.337154388427734, + "grad_norm": 47.97071838378906, "learning_rate": 1.5735558137078891e-06, - "loss": 0.2233, + "loss": 0.9793, "step": 87240 }, { "epoch": 9.21, - "grad_norm": 0.025726741179823875, + "grad_norm": 7.602646350860596, "learning_rate": 1.5714436582532475e-06, - "loss": 0.3105, + "loss": 0.7644, "step": 87250 }, { "epoch": 9.22, - "grad_norm": 13.863059043884277, + "grad_norm": 36.96261215209961, "learning_rate": 1.569331502798606e-06, - "loss": 0.4874, + "loss": 0.9239, "step": 87260 }, { "epoch": 9.22, - "grad_norm": 5.514533996582031, + "grad_norm": 53.13560485839844, "learning_rate": 1.5672193473439646e-06, - "loss": 0.1531, + "loss": 0.6431, "step": 87270 }, { "epoch": 9.22, - "grad_norm": 33.22738265991211, + "grad_norm": 15.68697738647461, "learning_rate": 1.5651071918893232e-06, - "loss": 0.1639, + "loss": 0.5472, "step": 87280 }, { "epoch": 9.22, - "grad_norm": 1.3378472328186035, + "grad_norm": 8.370047569274902, "learning_rate": 1.5629950364346815e-06, - "loss": 0.3066, + "loss": 0.9024, "step": 87290 }, { "epoch": 9.22, - "grad_norm": 18.98434829711914, + "grad_norm": 34.33846664428711, "learning_rate": 1.5608828809800403e-06, - "loss": 0.2907, + "loss": 0.7596, "step": 87300 }, { "epoch": 9.22, - "grad_norm": 1.2351820468902588, + "grad_norm": 45.09989547729492, "learning_rate": 1.5587707255253989e-06, - "loss": 0.3143, + "loss": 0.7702, "step": 87310 }, { "epoch": 9.22, - "grad_norm": 2.3671634197235107, + "grad_norm": 42.72468948364258, "learning_rate": 1.5566585700707572e-06, - "loss": 0.1923, + "loss": 0.8201, "step": 87320 }, { "epoch": 9.22, - "grad_norm": 10.135255813598633, + "grad_norm": 76.60911560058594, "learning_rate": 1.554546414616116e-06, - "loss": 0.3361, + "loss": 0.8433, "step": 87330 }, { "epoch": 9.22, - "grad_norm": 12.006550788879395, + "grad_norm": 49.74721145629883, "learning_rate": 1.5524342591614744e-06, - "loss": 0.377, + "loss": 0.6716, "step": 87340 }, { "epoch": 9.22, - "grad_norm": 0.38597625494003296, + "grad_norm": 30.023326873779297, "learning_rate": 1.5503221037068331e-06, - "loss": 0.3034, + "loss": 0.7376, "step": 87350 }, { "epoch": 9.23, - "grad_norm": 1.2937102317810059, + "grad_norm": 42.83390808105469, "learning_rate": 1.5482099482521915e-06, - "loss": 0.2535, + "loss": 0.7951, "step": 87360 }, { "epoch": 9.23, - "grad_norm": 9.794378280639648, + "grad_norm": 37.89130401611328, "learning_rate": 1.54609779279755e-06, - "loss": 0.3808, + "loss": 0.7933, "step": 87370 }, { "epoch": 9.23, - "grad_norm": 5.3599934577941895, + "grad_norm": 27.31538963317871, "learning_rate": 1.5439856373429084e-06, - "loss": 0.2598, + "loss": 0.6753, "step": 87380 }, { "epoch": 9.23, - "grad_norm": 144.16415405273438, + "grad_norm": 20.62470054626465, "learning_rate": 1.5418734818882672e-06, - "loss": 0.2858, + "loss": 0.9364, "step": 87390 }, { "epoch": 9.23, - "grad_norm": 6.883110523223877, + "grad_norm": 57.99580383300781, "learning_rate": 1.5397613264336256e-06, - "loss": 0.1504, + "loss": 1.0543, "step": 87400 }, { "epoch": 9.23, - "grad_norm": 0.021212907508015633, + "grad_norm": 27.905426025390625, "learning_rate": 1.5376491709789843e-06, - "loss": 0.1641, + "loss": 0.6754, "step": 87410 }, { "epoch": 9.23, - "grad_norm": 0.705109715461731, + "grad_norm": 6.663238525390625, "learning_rate": 1.5355370155243427e-06, - "loss": 0.2064, + "loss": 0.6945, "step": 87420 }, { "epoch": 9.23, - "grad_norm": 0.362067312002182, + "grad_norm": 38.307132720947266, "learning_rate": 1.5334248600697013e-06, - "loss": 0.2658, + "loss": 0.6543, "step": 87430 }, { "epoch": 9.23, - "grad_norm": 0.03453684225678444, + "grad_norm": 35.140560150146484, "learning_rate": 1.5313127046150596e-06, - "loss": 0.3675, + "loss": 0.9049, "step": 87440 }, { "epoch": 9.24, - "grad_norm": 11.691078186035156, + "grad_norm": 67.44263458251953, "learning_rate": 1.5292005491604184e-06, - "loss": 0.2087, + "loss": 0.9116, "step": 87450 }, { "epoch": 9.24, - "grad_norm": 20.172298431396484, + "grad_norm": 56.12212371826172, "learning_rate": 1.5270883937057767e-06, - "loss": 0.475, + "loss": 0.753, "step": 87460 }, { "epoch": 9.24, - "grad_norm": 24.683786392211914, + "grad_norm": 73.93553924560547, "learning_rate": 1.5249762382511355e-06, - "loss": 0.5158, + "loss": 1.0915, "step": 87470 }, { "epoch": 9.24, - "grad_norm": 24.822158813476562, + "grad_norm": 39.82536697387695, "learning_rate": 1.5228640827964939e-06, - "loss": 0.4658, + "loss": 0.637, "step": 87480 }, { "epoch": 9.24, - "grad_norm": 0.35889631509780884, + "grad_norm": 25.42123794555664, "learning_rate": 1.5207519273418524e-06, - "loss": 0.3896, + "loss": 0.7276, "step": 87490 }, { "epoch": 9.24, - "grad_norm": 39.669761657714844, + "grad_norm": 58.78672790527344, "learning_rate": 1.5186397718872108e-06, - "loss": 0.4335, + "loss": 0.8385, "step": 87500 }, { "epoch": 9.24, - "grad_norm": 1.9146608114242554, + "grad_norm": 29.102764129638672, "learning_rate": 1.5165276164325696e-06, - "loss": 0.36, + "loss": 0.8183, "step": 87510 }, { "epoch": 9.24, - "grad_norm": 12.01889705657959, + "grad_norm": 40.0723876953125, "learning_rate": 1.514415460977928e-06, - "loss": 0.1342, + "loss": 0.5139, "step": 87520 }, { "epoch": 9.24, - "grad_norm": 0.08571749925613403, + "grad_norm": 38.3127555847168, "learning_rate": 1.5123033055232867e-06, - "loss": 0.1111, + "loss": 0.4962, "step": 87530 }, { "epoch": 9.24, - "grad_norm": 27.88450050354004, + "grad_norm": 38.20927429199219, "learning_rate": 1.5101911500686453e-06, - "loss": 0.4154, + "loss": 1.0581, "step": 87540 }, { "epoch": 9.25, - "grad_norm": 8.10840129852295, + "grad_norm": 39.946693420410156, "learning_rate": 1.5080789946140036e-06, - "loss": 0.3857, + "loss": 0.9222, "step": 87550 }, { "epoch": 9.25, - "grad_norm": 4.791710376739502, + "grad_norm": 62.82667541503906, "learning_rate": 1.5059668391593624e-06, - "loss": 0.3904, + "loss": 1.0457, "step": 87560 }, { "epoch": 9.25, - "grad_norm": 8.611577033996582, + "grad_norm": 43.7117805480957, "learning_rate": 1.5038546837047208e-06, - "loss": 0.2799, + "loss": 0.6078, "step": 87570 }, { "epoch": 9.25, - "grad_norm": 19.090198516845703, + "grad_norm": 34.49597930908203, "learning_rate": 1.5017425282500793e-06, - "loss": 0.4402, + "loss": 0.8944, "step": 87580 }, { "epoch": 9.25, - "grad_norm": 14.916316032409668, + "grad_norm": 51.879390716552734, "learning_rate": 1.499630372795438e-06, - "loss": 0.398, + "loss": 0.8824, "step": 87590 }, { "epoch": 9.25, - "grad_norm": 14.332602500915527, + "grad_norm": 27.988182067871094, "learning_rate": 1.4975182173407965e-06, - "loss": 0.3628, + "loss": 0.6199, "step": 87600 }, { "epoch": 9.25, - "grad_norm": 2.908928155899048, + "grad_norm": 28.777238845825195, "learning_rate": 1.4954060618861548e-06, - "loss": 0.1138, + "loss": 0.6825, "step": 87610 }, { "epoch": 9.25, - "grad_norm": 1.6275131702423096, + "grad_norm": 61.44398498535156, "learning_rate": 1.4932939064315136e-06, - "loss": 0.0823, + "loss": 0.8814, "step": 87620 }, { "epoch": 9.25, - "grad_norm": 0.02585126832127571, + "grad_norm": 51.53357696533203, "learning_rate": 1.491181750976872e-06, - "loss": 0.0608, + "loss": 0.8603, "step": 87630 }, { "epoch": 9.26, - "grad_norm": 0.7505419850349426, + "grad_norm": 22.965742111206055, "learning_rate": 1.4890695955222305e-06, - "loss": 0.2356, + "loss": 0.714, "step": 87640 }, { "epoch": 9.26, - "grad_norm": 7.419224262237549, + "grad_norm": 22.945236206054688, "learning_rate": 1.486957440067589e-06, - "loss": 0.1578, + "loss": 0.8628, "step": 87650 }, { "epoch": 9.26, - "grad_norm": 2.690068006515503, + "grad_norm": 58.03616714477539, "learning_rate": 1.4848452846129477e-06, - "loss": 0.4803, + "loss": 0.7353, "step": 87660 }, { "epoch": 9.26, - "grad_norm": 28.995391845703125, + "grad_norm": 43.442256927490234, "learning_rate": 1.482733129158306e-06, - "loss": 0.3839, + "loss": 0.9347, "step": 87670 }, { "epoch": 9.26, - "grad_norm": 1.2381808757781982, + "grad_norm": 34.502044677734375, "learning_rate": 1.4806209737036648e-06, - "loss": 0.2882, + "loss": 0.9454, "step": 87680 }, { "epoch": 9.26, - "grad_norm": 0.05159958079457283, + "grad_norm": 27.518356323242188, "learning_rate": 1.4785088182490232e-06, - "loss": 0.1483, + "loss": 0.6186, "step": 87690 }, { "epoch": 9.26, - "grad_norm": 0.06291074305772781, + "grad_norm": 41.998878479003906, "learning_rate": 1.4763966627943817e-06, - "loss": 0.2507, + "loss": 0.8729, "step": 87700 }, { "epoch": 9.26, - "grad_norm": 0.5754205584526062, + "grad_norm": 45.841156005859375, "learning_rate": 1.4742845073397403e-06, - "loss": 0.2018, + "loss": 0.7325, "step": 87710 }, { "epoch": 9.26, - "grad_norm": 2.8288705348968506, + "grad_norm": 51.979774475097656, "learning_rate": 1.4721723518850989e-06, - "loss": 0.302, + "loss": 0.9889, "step": 87720 }, { "epoch": 9.26, - "grad_norm": 0.9646590948104858, + "grad_norm": 11.91099739074707, "learning_rate": 1.4700601964304572e-06, - "loss": 0.139, + "loss": 0.6278, "step": 87730 }, { "epoch": 9.27, - "grad_norm": 0.12640413641929626, + "grad_norm": 34.36797332763672, "learning_rate": 1.467948040975816e-06, - "loss": 0.3635, + "loss": 0.9453, "step": 87740 }, { "epoch": 9.27, - "grad_norm": 0.5466936826705933, + "grad_norm": 33.52910232543945, "learning_rate": 1.4658358855211743e-06, - "loss": 0.4263, + "loss": 0.659, "step": 87750 }, { "epoch": 9.27, - "grad_norm": 22.998506546020508, + "grad_norm": 57.099613189697266, "learning_rate": 1.463723730066533e-06, - "loss": 0.3001, + "loss": 0.8998, "step": 87760 }, { "epoch": 9.27, - "grad_norm": 4.582061767578125, + "grad_norm": 12.302288055419922, "learning_rate": 1.4616115746118917e-06, - "loss": 0.2384, + "loss": 0.7446, "step": 87770 }, { "epoch": 9.27, - "grad_norm": 12.389884948730469, + "grad_norm": 51.58038330078125, "learning_rate": 1.45949941915725e-06, - "loss": 0.2501, + "loss": 1.2494, "step": 87780 }, { "epoch": 9.27, - "grad_norm": 0.032020505517721176, + "grad_norm": 12.367197036743164, "learning_rate": 1.4573872637026088e-06, - "loss": 0.4911, + "loss": 1.1834, "step": 87790 }, { "epoch": 9.27, - "grad_norm": 0.7508631348609924, + "grad_norm": 35.63593292236328, "learning_rate": 1.4552751082479672e-06, - "loss": 0.4047, + "loss": 1.0933, "step": 87800 }, { "epoch": 9.27, - "grad_norm": 19.36077880859375, + "grad_norm": 67.38157653808594, "learning_rate": 1.4531629527933257e-06, - "loss": 0.4026, + "loss": 0.731, "step": 87810 }, { "epoch": 9.27, - "grad_norm": 16.46021842956543, + "grad_norm": 10.92448902130127, "learning_rate": 1.451050797338684e-06, - "loss": 0.3337, + "loss": 0.6669, "step": 87820 }, { "epoch": 9.28, - "grad_norm": 12.408862113952637, + "grad_norm": 50.422637939453125, "learning_rate": 1.4489386418840429e-06, - "loss": 0.6727, + "loss": 1.168, "step": 87830 }, { "epoch": 9.28, - "grad_norm": 19.90821647644043, + "grad_norm": 28.581045150756836, "learning_rate": 1.4468264864294012e-06, - "loss": 0.2588, + "loss": 0.8364, "step": 87840 }, { "epoch": 9.28, - "grad_norm": 14.490365982055664, + "grad_norm": 40.384674072265625, "learning_rate": 1.44471433097476e-06, - "loss": 0.2779, + "loss": 0.7963, "step": 87850 }, { "epoch": 9.28, - "grad_norm": 5.2678046226501465, + "grad_norm": 42.51392364501953, "learning_rate": 1.4426021755201184e-06, - "loss": 0.3377, + "loss": 0.8663, "step": 87860 }, { "epoch": 9.28, - "grad_norm": 4.750284671783447, + "grad_norm": 62.68831253051758, "learning_rate": 1.440490020065477e-06, - "loss": 0.4888, + "loss": 1.0081, "step": 87870 }, { "epoch": 9.28, - "grad_norm": 16.589262008666992, + "grad_norm": 39.13676834106445, "learning_rate": 1.4383778646108353e-06, - "loss": 0.4101, + "loss": 0.903, "step": 87880 }, { "epoch": 9.28, - "grad_norm": 5.273193359375, + "grad_norm": 49.57783508300781, "learning_rate": 1.436265709156194e-06, - "loss": 0.2164, + "loss": 0.8336, "step": 87890 }, { "epoch": 9.28, - "grad_norm": 27.16796112060547, + "grad_norm": 20.85001564025879, "learning_rate": 1.4341535537015524e-06, - "loss": 0.4336, + "loss": 0.8967, "step": 87900 }, { "epoch": 9.28, - "grad_norm": 0.7785083055496216, + "grad_norm": 26.31241226196289, "learning_rate": 1.4320413982469112e-06, - "loss": 0.37, + "loss": 0.6801, "step": 87910 }, { "epoch": 9.29, - "grad_norm": 0.35634270310401917, + "grad_norm": 14.229172706604004, "learning_rate": 1.4299292427922696e-06, - "loss": 0.3866, + "loss": 1.0044, "step": 87920 }, { "epoch": 9.29, - "grad_norm": 2.719475507736206, + "grad_norm": 35.168182373046875, "learning_rate": 1.4278170873376281e-06, - "loss": 0.4462, + "loss": 0.892, "step": 87930 }, { "epoch": 9.29, - "grad_norm": 0.03168491646647453, + "grad_norm": 18.32174301147461, "learning_rate": 1.4257049318829865e-06, - "loss": 0.4022, + "loss": 1.0854, "step": 87940 }, { "epoch": 9.29, - "grad_norm": 0.6552138328552246, + "grad_norm": 52.711002349853516, "learning_rate": 1.4235927764283453e-06, - "loss": 0.2644, + "loss": 0.7681, "step": 87950 }, { "epoch": 9.29, - "grad_norm": 24.983747482299805, + "grad_norm": 43.27142333984375, "learning_rate": 1.4214806209737036e-06, - "loss": 0.5999, + "loss": 1.0387, "step": 87960 }, { "epoch": 9.29, - "grad_norm": 0.08304473012685776, + "grad_norm": 56.75383758544922, "learning_rate": 1.4193684655190624e-06, - "loss": 0.4298, + "loss": 0.9287, "step": 87970 }, { "epoch": 9.29, - "grad_norm": 27.120235443115234, + "grad_norm": 50.5504035949707, "learning_rate": 1.417256310064421e-06, - "loss": 0.2902, + "loss": 0.89, "step": 87980 }, { "epoch": 9.29, - "grad_norm": 3.3759303092956543, + "grad_norm": 45.390079498291016, "learning_rate": 1.4151441546097793e-06, - "loss": 0.3429, + "loss": 0.7871, "step": 87990 }, { "epoch": 9.29, - "grad_norm": 3.9883430004119873, + "grad_norm": 10.25801944732666, "learning_rate": 1.413031999155138e-06, - "loss": 0.2228, + "loss": 0.6656, "step": 88000 }, { "epoch": 9.29, - "grad_norm": 22.738693237304688, + "grad_norm": 38.20535659790039, "learning_rate": 1.4109198437004965e-06, - "loss": 0.4545, + "loss": 1.1566, "step": 88010 }, { "epoch": 9.3, - "grad_norm": 25.145923614501953, + "grad_norm": 29.648324966430664, "learning_rate": 1.408807688245855e-06, - "loss": 0.2586, + "loss": 1.1284, "step": 88020 }, { "epoch": 9.3, - "grad_norm": 1.418139100074768, + "grad_norm": 20.116090774536133, "learning_rate": 1.4066955327912136e-06, - "loss": 0.3838, + "loss": 0.7366, "step": 88030 }, { "epoch": 9.3, - "grad_norm": 0.04688670486211777, + "grad_norm": 26.11695098876953, "learning_rate": 1.4045833773365722e-06, - "loss": 0.2247, + "loss": 0.6971, "step": 88040 }, { "epoch": 9.3, - "grad_norm": 33.14594268798828, + "grad_norm": 57.9742317199707, "learning_rate": 1.4024712218819305e-06, - "loss": 0.1798, + "loss": 0.6961, "step": 88050 }, { "epoch": 9.3, - "grad_norm": 0.7933502197265625, + "grad_norm": 39.57172393798828, "learning_rate": 1.4003590664272893e-06, - "loss": 0.2578, + "loss": 0.8487, "step": 88060 }, { "epoch": 9.3, - "grad_norm": 57.858211517333984, + "grad_norm": 40.0565071105957, "learning_rate": 1.3982469109726476e-06, - "loss": 0.5317, + "loss": 0.6518, "step": 88070 }, { "epoch": 9.3, - "grad_norm": 7.3073954582214355, + "grad_norm": 34.56169509887695, "learning_rate": 1.3961347555180062e-06, - "loss": 0.2908, + "loss": 1.1123, "step": 88080 }, { "epoch": 9.3, - "grad_norm": 12.976879119873047, + "grad_norm": 50.90713119506836, "learning_rate": 1.3940226000633648e-06, - "loss": 0.2483, + "loss": 0.7946, "step": 88090 }, { "epoch": 9.3, - "grad_norm": 21.034238815307617, + "grad_norm": 40.05735778808594, "learning_rate": 1.3919104446087233e-06, - "loss": 0.2918, + "loss": 0.5789, "step": 88100 }, { "epoch": 9.31, - "grad_norm": 1.4282281398773193, + "grad_norm": 42.48089599609375, "learning_rate": 1.3897982891540817e-06, - "loss": 0.2457, + "loss": 0.8251, "step": 88110 }, { "epoch": 9.31, - "grad_norm": 25.21343231201172, + "grad_norm": 60.05878448486328, "learning_rate": 1.3876861336994405e-06, - "loss": 0.2801, + "loss": 0.9493, "step": 88120 }, { "epoch": 9.31, - "grad_norm": 0.05783767253160477, + "grad_norm": 54.02157211303711, "learning_rate": 1.3855739782447988e-06, - "loss": 0.3131, + "loss": 0.8595, "step": 88130 }, { "epoch": 9.31, - "grad_norm": 10.625197410583496, + "grad_norm": 47.27408218383789, "learning_rate": 1.3834618227901574e-06, - "loss": 0.127, + "loss": 0.9049, "step": 88140 }, { "epoch": 9.31, - "grad_norm": 0.055524975061416626, + "grad_norm": 49.130306243896484, "learning_rate": 1.381349667335516e-06, - "loss": 0.2561, + "loss": 0.8316, "step": 88150 }, { "epoch": 9.31, - "grad_norm": 1.1107245683670044, + "grad_norm": 22.41647720336914, "learning_rate": 1.3792375118808745e-06, - "loss": 0.256, + "loss": 0.8281, "step": 88160 }, { "epoch": 9.31, - "grad_norm": 3.2928707599639893, + "grad_norm": 38.0706672668457, "learning_rate": 1.377125356426233e-06, - "loss": 0.1924, + "loss": 0.7515, "step": 88170 }, { "epoch": 9.31, - "grad_norm": 0.49438926577568054, + "grad_norm": 27.531654357910156, "learning_rate": 1.3750132009715917e-06, - "loss": 0.1247, + "loss": 0.6773, "step": 88180 }, { "epoch": 9.31, - "grad_norm": 0.333199679851532, + "grad_norm": 26.42129898071289, "learning_rate": 1.37290104551695e-06, - "loss": 0.5803, + "loss": 0.8884, "step": 88190 }, { "epoch": 9.31, - "grad_norm": 19.78152847290039, + "grad_norm": 44.387664794921875, "learning_rate": 1.3707888900623086e-06, - "loss": 0.5774, + "loss": 1.0727, "step": 88200 }, { "epoch": 9.32, - "grad_norm": 3.6855990886688232, + "grad_norm": 20.37154197692871, "learning_rate": 1.3686767346076674e-06, - "loss": 0.3608, + "loss": 0.5749, "step": 88210 }, { "epoch": 9.32, - "grad_norm": 4.504386901855469, + "grad_norm": 28.820093154907227, "learning_rate": 1.3665645791530257e-06, - "loss": 0.4452, + "loss": 1.1919, "step": 88220 }, { "epoch": 9.32, - "grad_norm": 1.3057063817977905, + "grad_norm": 30.656274795532227, "learning_rate": 1.3644524236983845e-06, - "loss": 0.2565, + "loss": 0.7742, "step": 88230 }, { "epoch": 9.32, - "grad_norm": 5.161217212677002, + "grad_norm": 54.67034912109375, "learning_rate": 1.3623402682437429e-06, - "loss": 0.2757, + "loss": 0.7378, "step": 88240 }, { "epoch": 9.32, - "grad_norm": 0.06997466832399368, + "grad_norm": 34.144989013671875, "learning_rate": 1.3602281127891014e-06, - "loss": 0.2237, + "loss": 0.5922, "step": 88250 }, { "epoch": 9.32, - "grad_norm": 70.5326156616211, + "grad_norm": 48.50392150878906, "learning_rate": 1.3581159573344598e-06, - "loss": 0.2457, + "loss": 0.5436, "step": 88260 }, { "epoch": 9.32, - "grad_norm": 3.5650413036346436, + "grad_norm": 50.73369216918945, "learning_rate": 1.3560038018798186e-06, - "loss": 0.2444, + "loss": 0.9562, "step": 88270 }, { "epoch": 9.32, - "grad_norm": 3.5869300365448, + "grad_norm": 12.713127136230469, "learning_rate": 1.353891646425177e-06, - "loss": 0.2272, + "loss": 0.7403, "step": 88280 }, { "epoch": 9.32, - "grad_norm": 1.434387445449829, + "grad_norm": 39.449607849121094, "learning_rate": 1.3517794909705357e-06, - "loss": 0.4658, + "loss": 1.0298, "step": 88290 }, { "epoch": 9.33, - "grad_norm": 51.96986770629883, + "grad_norm": 28.333383560180664, "learning_rate": 1.349667335515894e-06, - "loss": 0.4002, + "loss": 0.9417, "step": 88300 }, { "epoch": 9.33, - "grad_norm": 0.24315738677978516, + "grad_norm": 4.131734848022461, "learning_rate": 1.3475551800612526e-06, - "loss": 0.4251, + "loss": 0.7151, "step": 88310 }, { "epoch": 9.33, - "grad_norm": 8.511996269226074, + "grad_norm": 9.765399932861328, "learning_rate": 1.345443024606611e-06, - "loss": 0.6327, + "loss": 0.6308, "step": 88320 }, { "epoch": 9.33, - "grad_norm": 0.8064764738082886, + "grad_norm": 44.57937240600586, "learning_rate": 1.3433308691519698e-06, - "loss": 0.2756, + "loss": 0.5879, "step": 88330 }, { "epoch": 9.33, - "grad_norm": 9.680606842041016, + "grad_norm": 41.51827621459961, "learning_rate": 1.3412187136973281e-06, - "loss": 0.3012, + "loss": 0.8125, "step": 88340 }, { "epoch": 9.33, - "grad_norm": 38.769691467285156, + "grad_norm": 49.74962615966797, "learning_rate": 1.3391065582426869e-06, - "loss": 0.4212, + "loss": 1.2629, "step": 88350 }, { "epoch": 9.33, - "grad_norm": 3.205209493637085, + "grad_norm": 40.059383392333984, "learning_rate": 1.3369944027880452e-06, - "loss": 0.3667, + "loss": 0.7585, "step": 88360 }, { "epoch": 9.33, - "grad_norm": 11.407551765441895, + "grad_norm": 58.48928451538086, "learning_rate": 1.3348822473334038e-06, - "loss": 0.2679, + "loss": 0.8814, "step": 88370 }, { "epoch": 9.33, - "grad_norm": 1.9095567464828491, + "grad_norm": 46.34199142456055, "learning_rate": 1.3327700918787622e-06, - "loss": 0.405, + "loss": 1.2175, "step": 88380 }, { "epoch": 9.33, - "grad_norm": 9.305392265319824, + "grad_norm": 61.22062683105469, "learning_rate": 1.330657936424121e-06, - "loss": 0.4014, + "loss": 0.8195, "step": 88390 }, { "epoch": 9.34, - "grad_norm": 14.119522094726562, + "grad_norm": 1.7503503561019897, "learning_rate": 1.3285457809694793e-06, - "loss": 0.4075, + "loss": 0.758, "step": 88400 }, { "epoch": 9.34, - "grad_norm": 18.372472763061523, + "grad_norm": 26.547033309936523, "learning_rate": 1.326433625514838e-06, - "loss": 0.195, + "loss": 0.8478, "step": 88410 }, { "epoch": 9.34, - "grad_norm": 1.3916341066360474, + "grad_norm": 18.88149070739746, "learning_rate": 1.3243214700601964e-06, - "loss": 0.2912, + "loss": 0.8845, "step": 88420 }, { "epoch": 9.34, - "grad_norm": 14.751819610595703, + "grad_norm": 38.40372848510742, "learning_rate": 1.322209314605555e-06, - "loss": 0.5166, + "loss": 0.8297, "step": 88430 }, { "epoch": 9.34, - "grad_norm": 16.095985412597656, + "grad_norm": 42.493682861328125, "learning_rate": 1.3200971591509138e-06, - "loss": 0.438, + "loss": 0.6687, "step": 88440 }, { "epoch": 9.34, - "grad_norm": 0.0550924651324749, + "grad_norm": 46.164100646972656, "learning_rate": 1.3179850036962721e-06, - "loss": 0.2362, + "loss": 0.5334, "step": 88450 }, { "epoch": 9.34, - "grad_norm": 15.45985221862793, + "grad_norm": 23.724082946777344, "learning_rate": 1.3158728482416307e-06, - "loss": 0.2942, + "loss": 0.4428, "step": 88460 }, { "epoch": 9.34, - "grad_norm": 0.20988571643829346, + "grad_norm": 42.51344680786133, "learning_rate": 1.3137606927869893e-06, - "loss": 0.3433, + "loss": 1.3731, "step": 88470 }, { "epoch": 9.34, - "grad_norm": 2.153019428253174, + "grad_norm": 25.841768264770508, "learning_rate": 1.3116485373323478e-06, - "loss": 0.1984, + "loss": 0.8376, "step": 88480 }, { "epoch": 9.35, - "grad_norm": 0.08256742358207703, + "grad_norm": 39.44149398803711, "learning_rate": 1.3095363818777062e-06, - "loss": 0.354, + "loss": 0.8276, "step": 88490 }, { "epoch": 9.35, - "grad_norm": 9.971628189086914, + "grad_norm": 39.17467498779297, "learning_rate": 1.307424226423065e-06, - "loss": 0.3225, + "loss": 1.0571, "step": 88500 }, { "epoch": 9.35, - "grad_norm": 0.5175330638885498, + "grad_norm": 46.372982025146484, "learning_rate": 1.3053120709684233e-06, - "loss": 0.3838, + "loss": 0.8433, "step": 88510 }, { "epoch": 9.35, - "grad_norm": 2.885620355606079, + "grad_norm": 31.429094314575195, "learning_rate": 1.303199915513782e-06, - "loss": 0.1695, + "loss": 0.8489, "step": 88520 }, { "epoch": 9.35, - "grad_norm": 4.6350297927856445, + "grad_norm": 45.276546478271484, "learning_rate": 1.3010877600591405e-06, - "loss": 0.1943, + "loss": 0.9903, "step": 88530 }, { "epoch": 9.35, - "grad_norm": 34.69743347167969, + "grad_norm": 28.102096557617188, "learning_rate": 1.298975604604499e-06, - "loss": 0.3052, + "loss": 0.8698, "step": 88540 }, { "epoch": 9.35, - "grad_norm": 30.908061981201172, + "grad_norm": 56.95777130126953, "learning_rate": 1.2968634491498574e-06, - "loss": 0.5318, + "loss": 0.6463, "step": 88550 }, { "epoch": 9.35, - "grad_norm": 4.2646589279174805, + "grad_norm": 37.10520935058594, "learning_rate": 1.2947512936952162e-06, - "loss": 0.3223, + "loss": 0.7963, "step": 88560 }, { "epoch": 9.35, - "grad_norm": 0.0277232825756073, + "grad_norm": 13.832846641540527, "learning_rate": 1.2926391382405745e-06, - "loss": 0.3538, + "loss": 0.8978, "step": 88570 }, { "epoch": 9.35, - "grad_norm": 0.024179764091968536, + "grad_norm": 16.36171531677246, "learning_rate": 1.290526982785933e-06, - "loss": 0.3695, + "loss": 0.4817, "step": 88580 }, { "epoch": 9.36, - "grad_norm": 20.738988876342773, + "grad_norm": 53.19978713989258, "learning_rate": 1.2884148273312917e-06, - "loss": 0.201, + "loss": 0.702, "step": 88590 }, { "epoch": 9.36, - "grad_norm": 20.80562400817871, + "grad_norm": 48.19716262817383, "learning_rate": 1.2863026718766502e-06, - "loss": 0.5262, + "loss": 0.7952, "step": 88600 }, { "epoch": 9.36, - "grad_norm": 0.3515646755695343, + "grad_norm": 45.822452545166016, "learning_rate": 1.2841905164220086e-06, - "loss": 0.4163, + "loss": 0.7401, "step": 88610 }, { "epoch": 9.36, - "grad_norm": 2.8195619583129883, + "grad_norm": 47.60471725463867, "learning_rate": 1.2820783609673674e-06, - "loss": 0.4126, + "loss": 1.0139, "step": 88620 }, { "epoch": 9.36, - "grad_norm": 1.0118829011917114, + "grad_norm": 25.95718002319336, "learning_rate": 1.2799662055127257e-06, - "loss": 0.3431, + "loss": 0.7237, "step": 88630 }, { "epoch": 9.36, - "grad_norm": 24.31789207458496, + "grad_norm": 26.557588577270508, "learning_rate": 1.2778540500580845e-06, - "loss": 0.3075, + "loss": 0.68, "step": 88640 }, { "epoch": 9.36, - "grad_norm": 20.667987823486328, + "grad_norm": 28.5255069732666, "learning_rate": 1.275741894603443e-06, - "loss": 0.4689, + "loss": 1.0513, "step": 88650 }, { "epoch": 9.36, - "grad_norm": 3.2484819889068604, + "grad_norm": 54.58644104003906, "learning_rate": 1.2736297391488014e-06, - "loss": 0.449, + "loss": 1.0616, "step": 88660 }, { "epoch": 9.36, - "grad_norm": 9.348763465881348, + "grad_norm": 53.21269989013672, "learning_rate": 1.2715175836941602e-06, - "loss": 0.4618, + "loss": 1.3062, "step": 88670 }, { "epoch": 9.37, - "grad_norm": 14.094366073608398, + "grad_norm": 31.364707946777344, "learning_rate": 1.2694054282395185e-06, - "loss": 0.651, + "loss": 0.9279, "step": 88680 }, { "epoch": 9.37, - "grad_norm": 10.240071296691895, + "grad_norm": 28.00006675720215, "learning_rate": 1.2672932727848771e-06, - "loss": 0.2562, + "loss": 0.885, "step": 88690 }, { "epoch": 9.37, - "grad_norm": 9.76449966430664, + "grad_norm": 33.76953887939453, "learning_rate": 1.2651811173302357e-06, - "loss": 0.2345, + "loss": 0.9967, "step": 88700 }, { "epoch": 9.37, - "grad_norm": 8.965413093566895, + "grad_norm": 14.288381576538086, "learning_rate": 1.2630689618755942e-06, - "loss": 0.1058, + "loss": 0.6606, "step": 88710 }, { "epoch": 9.37, - "grad_norm": 28.979848861694336, + "grad_norm": 28.635581970214844, "learning_rate": 1.2609568064209526e-06, - "loss": 0.2836, + "loss": 0.7327, "step": 88720 }, { "epoch": 9.37, - "grad_norm": 25.220458984375, + "grad_norm": 69.86980438232422, "learning_rate": 1.2588446509663114e-06, - "loss": 0.3164, + "loss": 0.9488, "step": 88730 }, { "epoch": 9.37, - "grad_norm": 9.892459869384766, + "grad_norm": 25.620718002319336, "learning_rate": 1.2567324955116697e-06, - "loss": 0.2234, + "loss": 0.5976, "step": 88740 }, { "epoch": 9.37, - "grad_norm": 0.009173806756734848, + "grad_norm": 31.226383209228516, "learning_rate": 1.2546203400570283e-06, - "loss": 0.4035, + "loss": 1.0851, "step": 88750 }, { "epoch": 9.37, - "grad_norm": 47.08222579956055, + "grad_norm": 35.143314361572266, "learning_rate": 1.2525081846023869e-06, - "loss": 0.4382, + "loss": 0.9698, "step": 88760 }, { "epoch": 9.37, - "grad_norm": 14.293161392211914, + "grad_norm": 47.9718132019043, "learning_rate": 1.2503960291477454e-06, - "loss": 0.3005, + "loss": 0.639, "step": 88770 }, { "epoch": 9.38, - "grad_norm": 13.285529136657715, + "grad_norm": 37.786705017089844, "learning_rate": 1.248283873693104e-06, - "loss": 0.2225, + "loss": 0.8968, "step": 88780 }, { "epoch": 9.38, - "grad_norm": 2.9050655364990234, + "grad_norm": 36.59130096435547, "learning_rate": 1.2461717182384626e-06, - "loss": 0.3315, + "loss": 0.5843, "step": 88790 }, { "epoch": 9.38, - "grad_norm": 13.704904556274414, + "grad_norm": 46.769447326660156, "learning_rate": 1.2440595627838211e-06, - "loss": 0.1962, + "loss": 0.5005, "step": 88800 }, { "epoch": 9.38, - "grad_norm": 14.850579261779785, + "grad_norm": 38.00374221801758, "learning_rate": 1.2419474073291795e-06, - "loss": 0.387, + "loss": 1.1902, "step": 88810 }, { "epoch": 9.38, - "grad_norm": 12.56837272644043, + "grad_norm": 63.42015075683594, "learning_rate": 1.239835251874538e-06, - "loss": 0.3275, + "loss": 0.5352, "step": 88820 }, { "epoch": 9.38, - "grad_norm": 14.117676734924316, + "grad_norm": 41.8402099609375, "learning_rate": 1.2377230964198966e-06, - "loss": 0.6657, + "loss": 0.8783, "step": 88830 }, { "epoch": 9.38, - "grad_norm": 34.14234161376953, + "grad_norm": 57.1761360168457, "learning_rate": 1.2356109409652552e-06, - "loss": 0.2202, + "loss": 0.9407, "step": 88840 }, { "epoch": 9.38, - "grad_norm": 32.537933349609375, + "grad_norm": 37.51485061645508, "learning_rate": 1.2334987855106138e-06, - "loss": 0.1319, + "loss": 0.7632, "step": 88850 }, { "epoch": 9.38, - "grad_norm": 29.660947799682617, + "grad_norm": 59.548622131347656, "learning_rate": 1.2313866300559723e-06, - "loss": 0.4358, + "loss": 0.8173, "step": 88860 }, { "epoch": 9.39, - "grad_norm": 0.41554614901542664, + "grad_norm": 64.33069610595703, "learning_rate": 1.2292744746013307e-06, - "loss": 0.2575, + "loss": 0.8364, "step": 88870 }, { "epoch": 9.39, - "grad_norm": 5.969208717346191, + "grad_norm": 33.47277069091797, "learning_rate": 1.2271623191466893e-06, - "loss": 0.2077, + "loss": 0.9726, "step": 88880 }, { "epoch": 9.39, - "grad_norm": 66.8689956665039, + "grad_norm": 37.355194091796875, "learning_rate": 1.2250501636920478e-06, - "loss": 0.4901, + "loss": 0.8398, "step": 88890 }, { "epoch": 9.39, - "grad_norm": 0.3623167872428894, + "grad_norm": 28.548105239868164, "learning_rate": 1.2229380082374064e-06, - "loss": 0.2569, + "loss": 0.5779, "step": 88900 }, { "epoch": 9.39, - "grad_norm": 14.154921531677246, + "grad_norm": 16.785442352294922, "learning_rate": 1.220825852782765e-06, - "loss": 0.3244, + "loss": 0.7693, "step": 88910 }, { "epoch": 9.39, - "grad_norm": 1.4359050989151, + "grad_norm": 50.80205535888672, "learning_rate": 1.2187136973281235e-06, - "loss": 0.5077, + "loss": 1.0824, "step": 88920 }, { "epoch": 9.39, - "grad_norm": 1.2971701622009277, + "grad_norm": 32.546268463134766, "learning_rate": 1.2166015418734819e-06, - "loss": 0.3889, + "loss": 0.8591, "step": 88930 }, { "epoch": 9.39, - "grad_norm": 21.392906188964844, + "grad_norm": 41.09899139404297, "learning_rate": 1.2144893864188404e-06, - "loss": 0.3663, + "loss": 1.0298, "step": 88940 }, { "epoch": 9.39, - "grad_norm": 0.16155102849006653, + "grad_norm": 60.11708068847656, "learning_rate": 1.212377230964199e-06, - "loss": 0.1393, + "loss": 0.7945, "step": 88950 }, { "epoch": 9.39, - "grad_norm": 45.99226760864258, + "grad_norm": 41.02069091796875, "learning_rate": 1.2102650755095576e-06, - "loss": 0.2043, + "loss": 1.1055, "step": 88960 }, { "epoch": 9.4, - "grad_norm": 0.41699254512786865, + "grad_norm": 26.58252716064453, "learning_rate": 1.2081529200549161e-06, - "loss": 0.2011, + "loss": 0.9663, "step": 88970 }, { "epoch": 9.4, - "grad_norm": 11.571470260620117, + "grad_norm": 37.374839782714844, "learning_rate": 1.2060407646002747e-06, - "loss": 0.296, + "loss": 0.9414, "step": 88980 }, { "epoch": 9.4, - "grad_norm": 3.817608594894409, + "grad_norm": 20.846588134765625, "learning_rate": 1.2039286091456333e-06, - "loss": 0.2857, + "loss": 0.8081, "step": 88990 }, { "epoch": 9.4, - "grad_norm": 0.4923166334629059, + "grad_norm": 51.195194244384766, "learning_rate": 1.2018164536909918e-06, - "loss": 0.207, + "loss": 0.9445, "step": 89000 }, { "epoch": 9.4, - "grad_norm": 27.277978897094727, + "grad_norm": 66.66936492919922, "learning_rate": 1.1997042982363504e-06, - "loss": 0.489, + "loss": 1.0258, "step": 89010 }, { "epoch": 9.4, - "grad_norm": 0.7319654226303101, + "grad_norm": 77.23774719238281, "learning_rate": 1.197592142781709e-06, - "loss": 0.4376, + "loss": 0.8618, "step": 89020 }, { "epoch": 9.4, - "grad_norm": 16.106552124023438, + "grad_norm": 52.926177978515625, "learning_rate": 1.1954799873270673e-06, - "loss": 0.2779, + "loss": 0.9705, "step": 89030 }, { "epoch": 9.4, - "grad_norm": 9.028382301330566, + "grad_norm": 34.717777252197266, "learning_rate": 1.193367831872426e-06, - "loss": 0.3457, + "loss": 0.6303, "step": 89040 }, { "epoch": 9.4, - "grad_norm": 10.349404335021973, + "grad_norm": 32.93898391723633, "learning_rate": 1.1912556764177845e-06, - "loss": 0.1429, + "loss": 0.708, "step": 89050 }, { "epoch": 9.41, - "grad_norm": 0.09518537670373917, + "grad_norm": 52.36660385131836, "learning_rate": 1.189143520963143e-06, - "loss": 0.4015, + "loss": 0.9227, "step": 89060 }, { "epoch": 9.41, - "grad_norm": 18.10097885131836, + "grad_norm": 32.84447479248047, "learning_rate": 1.1870313655085016e-06, - "loss": 0.2846, + "loss": 0.8399, "step": 89070 }, { "epoch": 9.41, - "grad_norm": 30.583740234375, + "grad_norm": 67.49385833740234, "learning_rate": 1.1849192100538602e-06, - "loss": 0.4902, + "loss": 0.8392, "step": 89080 }, { "epoch": 9.41, - "grad_norm": 20.69078826904297, + "grad_norm": 52.28436279296875, "learning_rate": 1.1828070545992185e-06, - "loss": 0.5774, + "loss": 0.971, "step": 89090 }, { "epoch": 9.41, - "grad_norm": 1.4753992557525635, + "grad_norm": 43.16322326660156, "learning_rate": 1.180694899144577e-06, - "loss": 0.2395, + "loss": 0.8577, "step": 89100 }, { "epoch": 9.41, - "grad_norm": 19.890026092529297, + "grad_norm": 14.623209953308105, "learning_rate": 1.1785827436899357e-06, - "loss": 0.1145, + "loss": 0.7037, "step": 89110 }, { "epoch": 9.41, - "grad_norm": 11.503521919250488, + "grad_norm": 35.004669189453125, "learning_rate": 1.1764705882352942e-06, - "loss": 0.3989, + "loss": 0.9401, "step": 89120 }, { "epoch": 9.41, - "grad_norm": 3.1735596656799316, + "grad_norm": 40.151817321777344, "learning_rate": 1.1743584327806528e-06, - "loss": 0.4024, + "loss": 0.945, "step": 89130 }, { "epoch": 9.41, - "grad_norm": 6.202307224273682, + "grad_norm": 37.215389251708984, "learning_rate": 1.1722462773260114e-06, - "loss": 0.1716, + "loss": 0.732, "step": 89140 }, { "epoch": 9.41, - "grad_norm": 17.394968032836914, + "grad_norm": 25.492765426635742, "learning_rate": 1.1701341218713697e-06, - "loss": 0.1607, + "loss": 1.0382, "step": 89150 }, { "epoch": 9.42, - "grad_norm": 19.814876556396484, + "grad_norm": 57.703895568847656, "learning_rate": 1.1680219664167283e-06, - "loss": 0.5657, + "loss": 0.9359, "step": 89160 }, { "epoch": 9.42, - "grad_norm": 15.86755084991455, + "grad_norm": 34.86428451538086, "learning_rate": 1.1659098109620869e-06, - "loss": 0.4086, + "loss": 0.7317, "step": 89170 }, { "epoch": 9.42, - "grad_norm": 6.464626789093018, + "grad_norm": 44.3116455078125, "learning_rate": 1.1637976555074454e-06, - "loss": 0.3094, + "loss": 0.9386, "step": 89180 }, { "epoch": 9.42, - "grad_norm": 98.2881088256836, + "grad_norm": 48.193519592285156, "learning_rate": 1.161685500052804e-06, - "loss": 0.1037, + "loss": 0.752, "step": 89190 }, { "epoch": 9.42, - "grad_norm": 0.05126018077135086, + "grad_norm": 33.50981521606445, "learning_rate": 1.1595733445981626e-06, - "loss": 0.3312, + "loss": 0.8636, "step": 89200 }, { "epoch": 9.42, - "grad_norm": 10.151254653930664, + "grad_norm": 52.33675003051758, "learning_rate": 1.1574611891435211e-06, - "loss": 0.2952, + "loss": 1.2015, "step": 89210 }, { "epoch": 9.42, - "grad_norm": 16.808141708374023, + "grad_norm": 11.145293235778809, "learning_rate": 1.1553490336888797e-06, - "loss": 0.2586, + "loss": 0.7015, "step": 89220 }, { "epoch": 9.42, - "grad_norm": 0.5189464688301086, + "grad_norm": 25.94916534423828, "learning_rate": 1.1532368782342383e-06, - "loss": 0.3023, + "loss": 0.9154, "step": 89230 }, { "epoch": 9.42, - "grad_norm": 14.654438018798828, + "grad_norm": 54.18556594848633, "learning_rate": 1.1511247227795968e-06, - "loss": 0.4369, + "loss": 1.1253, "step": 89240 }, { "epoch": 9.43, - "grad_norm": 19.01473045349121, + "grad_norm": 53.905635833740234, "learning_rate": 1.1490125673249552e-06, - "loss": 0.3729, + "loss": 0.8841, "step": 89250 }, { "epoch": 9.43, - "grad_norm": 25.892024993896484, + "grad_norm": 28.260576248168945, "learning_rate": 1.1469004118703137e-06, - "loss": 0.4228, + "loss": 0.7716, "step": 89260 }, { "epoch": 9.43, - "grad_norm": 3.271061897277832, + "grad_norm": 45.69192886352539, "learning_rate": 1.1447882564156723e-06, - "loss": 0.1197, + "loss": 0.5623, "step": 89270 }, { "epoch": 9.43, - "grad_norm": 11.901851654052734, + "grad_norm": 36.428985595703125, "learning_rate": 1.1426761009610309e-06, - "loss": 0.2396, + "loss": 0.533, "step": 89280 }, { "epoch": 9.43, - "grad_norm": 35.723445892333984, + "grad_norm": 63.172916412353516, "learning_rate": 1.1405639455063894e-06, - "loss": 0.2179, + "loss": 0.8402, "step": 89290 }, { "epoch": 9.43, - "grad_norm": 13.239872932434082, + "grad_norm": 43.89324951171875, "learning_rate": 1.138451790051748e-06, - "loss": 0.2712, + "loss": 0.7766, "step": 89300 }, { "epoch": 9.43, - "grad_norm": 6.709028244018555, + "grad_norm": 24.107460021972656, "learning_rate": 1.1363396345971064e-06, - "loss": 0.2899, + "loss": 1.0218, "step": 89310 }, { "epoch": 9.43, - "grad_norm": 16.525161743164062, + "grad_norm": 38.21196365356445, "learning_rate": 1.134227479142465e-06, - "loss": 0.3021, + "loss": 0.7769, "step": 89320 }, { "epoch": 9.43, - "grad_norm": 41.31047439575195, + "grad_norm": 40.331581115722656, "learning_rate": 1.1321153236878235e-06, - "loss": 0.2944, + "loss": 0.5704, "step": 89330 }, { "epoch": 9.43, - "grad_norm": 20.58791732788086, + "grad_norm": 47.61863708496094, "learning_rate": 1.130003168233182e-06, - "loss": 0.4732, + "loss": 0.978, "step": 89340 }, { "epoch": 9.44, - "grad_norm": 12.259522438049316, + "grad_norm": 34.00428771972656, "learning_rate": 1.1278910127785406e-06, - "loss": 0.4017, + "loss": 0.7621, "step": 89350 }, { "epoch": 9.44, - "grad_norm": 3.477875232696533, + "grad_norm": 93.24580383300781, "learning_rate": 1.1257788573238992e-06, - "loss": 0.3858, + "loss": 1.0977, "step": 89360 }, { "epoch": 9.44, - "grad_norm": 0.20910325646400452, + "grad_norm": 31.934507369995117, "learning_rate": 1.1236667018692576e-06, - "loss": 0.0813, + "loss": 0.7456, "step": 89370 }, { "epoch": 9.44, - "grad_norm": 0.042760442942380905, + "grad_norm": 13.585480690002441, "learning_rate": 1.1215545464146161e-06, - "loss": 0.3001, + "loss": 0.9019, "step": 89380 }, { "epoch": 9.44, - "grad_norm": 14.420151710510254, + "grad_norm": 42.699398040771484, "learning_rate": 1.1194423909599747e-06, - "loss": 0.4192, + "loss": 0.8963, "step": 89390 }, { "epoch": 9.44, - "grad_norm": 21.214040756225586, + "grad_norm": 63.815208435058594, "learning_rate": 1.1173302355053333e-06, - "loss": 0.4301, + "loss": 0.8677, "step": 89400 }, { "epoch": 9.44, - "grad_norm": 17.098129272460938, + "grad_norm": 46.407535552978516, "learning_rate": 1.1152180800506918e-06, - "loss": 0.5141, + "loss": 0.9511, "step": 89410 }, { "epoch": 9.44, - "grad_norm": 0.33670902252197266, + "grad_norm": 35.51681137084961, "learning_rate": 1.1131059245960504e-06, - "loss": 0.2169, + "loss": 1.0641, "step": 89420 }, { "epoch": 9.44, - "grad_norm": 39.97218322753906, + "grad_norm": 30.519306182861328, "learning_rate": 1.110993769141409e-06, - "loss": 0.3348, + "loss": 0.6227, "step": 89430 }, { "epoch": 9.45, - "grad_norm": 21.37680435180664, + "grad_norm": 41.79139709472656, "learning_rate": 1.1088816136867675e-06, - "loss": 0.4059, + "loss": 0.8138, "step": 89440 }, { "epoch": 9.45, - "grad_norm": 15.896857261657715, + "grad_norm": 33.03495407104492, "learning_rate": 1.106769458232126e-06, - "loss": 0.4284, + "loss": 1.0027, "step": 89450 }, { "epoch": 9.45, - "grad_norm": 9.069596290588379, + "grad_norm": 28.761810302734375, "learning_rate": 1.1046573027774847e-06, - "loss": 0.2793, + "loss": 0.5274, "step": 89460 }, { "epoch": 9.45, - "grad_norm": 15.621601104736328, + "grad_norm": 21.944074630737305, "learning_rate": 1.102545147322843e-06, - "loss": 0.3873, + "loss": 0.835, "step": 89470 }, { "epoch": 9.45, - "grad_norm": 0.2617167532444, + "grad_norm": 22.806838989257812, "learning_rate": 1.1004329918682016e-06, - "loss": 0.1187, + "loss": 0.6451, "step": 89480 }, { "epoch": 9.45, - "grad_norm": 0.8456488251686096, + "grad_norm": 41.42244338989258, "learning_rate": 1.0983208364135601e-06, - "loss": 0.396, + "loss": 0.7874, "step": 89490 }, { "epoch": 9.45, - "grad_norm": 18.184680938720703, + "grad_norm": 62.105838775634766, "learning_rate": 1.0962086809589187e-06, - "loss": 0.4343, + "loss": 0.9399, "step": 89500 }, { "epoch": 9.45, - "grad_norm": 6.727877616882324, + "grad_norm": 46.834903717041016, "learning_rate": 1.0940965255042773e-06, - "loss": 0.2543, + "loss": 0.7574, "step": 89510 }, { "epoch": 9.45, - "grad_norm": 25.86722183227539, + "grad_norm": 34.61198425292969, "learning_rate": 1.0919843700496359e-06, - "loss": 0.5168, + "loss": 1.0536, "step": 89520 }, { "epoch": 9.46, - "grad_norm": 1.5522427558898926, + "grad_norm": 3.252377986907959, "learning_rate": 1.0898722145949942e-06, - "loss": 0.3703, + "loss": 0.694, "step": 89530 }, { "epoch": 9.46, - "grad_norm": 11.310200691223145, + "grad_norm": 41.82362747192383, "learning_rate": 1.0877600591403528e-06, - "loss": 0.3823, + "loss": 0.9589, "step": 89540 }, { "epoch": 9.46, - "grad_norm": 11.778197288513184, + "grad_norm": 48.87786102294922, "learning_rate": 1.0856479036857113e-06, - "loss": 0.3079, + "loss": 1.3369, "step": 89550 }, { "epoch": 9.46, - "grad_norm": 0.845786988735199, + "grad_norm": 56.629478454589844, "learning_rate": 1.08353574823107e-06, - "loss": 0.4138, + "loss": 0.8061, "step": 89560 }, { "epoch": 9.46, - "grad_norm": 0.10486291348934174, + "grad_norm": 20.63834571838379, "learning_rate": 1.0814235927764285e-06, - "loss": 0.1064, + "loss": 0.893, "step": 89570 }, { "epoch": 9.46, - "grad_norm": 15.910896301269531, + "grad_norm": 15.490776062011719, "learning_rate": 1.079311437321787e-06, - "loss": 0.303, + "loss": 0.9671, "step": 89580 }, { "epoch": 9.46, - "grad_norm": 8.726815223693848, + "grad_norm": 40.76172637939453, "learning_rate": 1.0771992818671454e-06, - "loss": 0.3917, + "loss": 0.9851, "step": 89590 }, { "epoch": 9.46, - "grad_norm": 8.26207447052002, + "grad_norm": 21.26823616027832, "learning_rate": 1.075087126412504e-06, - "loss": 0.4349, + "loss": 0.9476, "step": 89600 }, { "epoch": 9.46, - "grad_norm": 38.78480529785156, + "grad_norm": 32.1103630065918, "learning_rate": 1.0729749709578625e-06, - "loss": 0.3303, + "loss": 0.7804, "step": 89610 }, { "epoch": 9.46, - "grad_norm": 11.370686531066895, + "grad_norm": 35.79645919799805, "learning_rate": 1.070862815503221e-06, - "loss": 0.3956, + "loss": 0.7392, "step": 89620 }, { "epoch": 9.47, - "grad_norm": 13.373658180236816, + "grad_norm": 28.405282974243164, "learning_rate": 1.0687506600485797e-06, - "loss": 0.1602, + "loss": 0.7317, "step": 89630 }, { "epoch": 9.47, - "grad_norm": 17.413198471069336, + "grad_norm": 30.247953414916992, "learning_rate": 1.0666385045939382e-06, - "loss": 0.3045, + "loss": 1.127, "step": 89640 }, { "epoch": 9.47, - "grad_norm": 1.082585096359253, + "grad_norm": 28.250633239746094, "learning_rate": 1.0645263491392966e-06, - "loss": 0.3787, + "loss": 1.0289, "step": 89650 }, { "epoch": 9.47, - "grad_norm": 27.655216217041016, + "grad_norm": 47.584991455078125, "learning_rate": 1.0624141936846554e-06, - "loss": 0.5061, + "loss": 0.805, "step": 89660 }, { "epoch": 9.47, - "grad_norm": 17.939083099365234, + "grad_norm": 13.761555671691895, "learning_rate": 1.060302038230014e-06, - "loss": 0.2867, + "loss": 0.7237, "step": 89670 }, { "epoch": 9.47, - "grad_norm": 35.76050567626953, + "grad_norm": 38.54896545410156, "learning_rate": 1.0581898827753725e-06, - "loss": 0.2191, + "loss": 0.7773, "step": 89680 }, { "epoch": 9.47, - "grad_norm": 33.02425765991211, + "grad_norm": 44.91580581665039, "learning_rate": 1.0560777273207309e-06, - "loss": 0.3087, + "loss": 1.138, "step": 89690 }, { "epoch": 9.47, - "grad_norm": 1.2902803421020508, + "grad_norm": 33.359867095947266, "learning_rate": 1.0539655718660894e-06, - "loss": 0.2886, + "loss": 0.9019, "step": 89700 }, { "epoch": 9.47, - "grad_norm": 50.309349060058594, + "grad_norm": 18.048128128051758, "learning_rate": 1.051853416411448e-06, - "loss": 0.4316, + "loss": 0.7427, "step": 89710 }, { "epoch": 9.48, - "grad_norm": 0.4022395610809326, + "grad_norm": 21.574377059936523, "learning_rate": 1.0497412609568066e-06, - "loss": 0.1701, + "loss": 0.5425, "step": 89720 }, { "epoch": 9.48, - "grad_norm": 4.042321681976318, + "grad_norm": 39.304473876953125, "learning_rate": 1.0476291055021651e-06, - "loss": 0.2636, + "loss": 0.793, "step": 89730 }, { "epoch": 9.48, - "grad_norm": 25.78554916381836, + "grad_norm": 35.651092529296875, "learning_rate": 1.0455169500475237e-06, - "loss": 0.421, + "loss": 0.8977, "step": 89740 }, { "epoch": 9.48, - "grad_norm": 0.12205401062965393, + "grad_norm": 40.48830032348633, "learning_rate": 1.043404794592882e-06, - "loss": 0.2048, + "loss": 1.0517, "step": 89750 }, { "epoch": 9.48, - "grad_norm": 38.52772521972656, + "grad_norm": 24.428823471069336, "learning_rate": 1.0412926391382406e-06, - "loss": 0.2554, + "loss": 0.617, "step": 89760 }, { "epoch": 9.48, - "grad_norm": 37.96556091308594, + "grad_norm": 37.705204010009766, "learning_rate": 1.0391804836835992e-06, - "loss": 0.5126, + "loss": 1.079, "step": 89770 }, { "epoch": 9.48, - "grad_norm": 15.617012977600098, + "grad_norm": 49.841304779052734, "learning_rate": 1.0370683282289577e-06, - "loss": 0.3342, + "loss": 0.8077, "step": 89780 }, { "epoch": 9.48, - "grad_norm": 0.44893476366996765, + "grad_norm": 9.602840423583984, "learning_rate": 1.0349561727743163e-06, - "loss": 0.3574, + "loss": 0.6801, "step": 89790 }, { "epoch": 9.48, - "grad_norm": 0.18821121752262115, + "grad_norm": 37.01138687133789, "learning_rate": 1.0328440173196749e-06, - "loss": 0.1833, + "loss": 0.7616, "step": 89800 }, { "epoch": 9.48, - "grad_norm": 0.18154332041740417, + "grad_norm": 22.234262466430664, "learning_rate": 1.0307318618650332e-06, - "loss": 0.474, + "loss": 0.8281, "step": 89810 }, { "epoch": 9.49, - "grad_norm": 0.7490761280059814, + "grad_norm": 32.6552734375, "learning_rate": 1.0286197064103918e-06, - "loss": 0.3049, + "loss": 0.8002, "step": 89820 }, { "epoch": 9.49, - "grad_norm": 26.753019332885742, + "grad_norm": 55.25130844116211, "learning_rate": 1.0265075509557504e-06, - "loss": 0.3888, + "loss": 0.5947, "step": 89830 }, { "epoch": 9.49, - "grad_norm": 7.125583648681641, + "grad_norm": 32.75558090209961, "learning_rate": 1.024395395501109e-06, - "loss": 0.2697, + "loss": 0.5629, "step": 89840 }, { "epoch": 9.49, - "grad_norm": 17.167621612548828, + "grad_norm": 37.08262634277344, "learning_rate": 1.0222832400464675e-06, - "loss": 0.3973, + "loss": 0.8255, "step": 89850 }, { "epoch": 9.49, - "grad_norm": 1.3694995641708374, + "grad_norm": 26.538667678833008, "learning_rate": 1.020171084591826e-06, - "loss": 0.2213, + "loss": 0.8368, "step": 89860 }, { "epoch": 9.49, - "grad_norm": 15.679282188415527, + "grad_norm": 57.51693344116211, "learning_rate": 1.0180589291371844e-06, - "loss": 0.2631, + "loss": 0.8991, "step": 89870 }, { "epoch": 9.49, - "grad_norm": 15.479473114013672, + "grad_norm": 53.30796432495117, "learning_rate": 1.0159467736825432e-06, - "loss": 0.3475, + "loss": 1.1028, "step": 89880 }, { "epoch": 9.49, - "grad_norm": 0.11599647998809814, + "grad_norm": 40.50405502319336, "learning_rate": 1.0138346182279018e-06, - "loss": 0.0971, + "loss": 0.649, "step": 89890 }, { "epoch": 9.49, - "grad_norm": 14.046319961547852, + "grad_norm": 26.685861587524414, "learning_rate": 1.0117224627732603e-06, - "loss": 0.1548, + "loss": 0.4018, "step": 89900 }, { "epoch": 9.5, - "grad_norm": 10.169732093811035, + "grad_norm": 56.457984924316406, "learning_rate": 1.0096103073186187e-06, - "loss": 0.4195, + "loss": 0.6927, "step": 89910 }, { "epoch": 9.5, - "grad_norm": 0.04205802455544472, + "grad_norm": 48.50800323486328, "learning_rate": 1.0074981518639773e-06, - "loss": 0.4025, + "loss": 0.6625, "step": 89920 }, { "epoch": 9.5, - "grad_norm": 19.490005493164062, + "grad_norm": 54.17573165893555, "learning_rate": 1.0053859964093358e-06, - "loss": 0.1446, + "loss": 0.7806, "step": 89930 }, { "epoch": 9.5, - "grad_norm": 41.50940704345703, + "grad_norm": 42.36492919921875, "learning_rate": 1.0032738409546944e-06, - "loss": 0.4989, + "loss": 1.143, "step": 89940 }, { "epoch": 9.5, - "grad_norm": 13.905963897705078, + "grad_norm": 41.182796478271484, "learning_rate": 1.001161685500053e-06, - "loss": 0.0895, + "loss": 0.7509, "step": 89950 }, { "epoch": 9.5, - "grad_norm": 14.535218238830566, + "grad_norm": 64.97492218017578, "learning_rate": 9.990495300454115e-07, - "loss": 0.1781, + "loss": 0.7582, "step": 89960 }, { "epoch": 9.5, - "grad_norm": 1.4646615982055664, + "grad_norm": 17.503263473510742, "learning_rate": 9.969373745907699e-07, - "loss": 0.3087, + "loss": 0.6335, "step": 89970 }, { "epoch": 9.5, - "grad_norm": 11.940465927124023, + "grad_norm": 39.63725280761719, "learning_rate": 9.948252191361285e-07, - "loss": 0.222, + "loss": 0.5716, "step": 89980 }, { "epoch": 9.5, - "grad_norm": 1.300437331199646, + "grad_norm": 44.977237701416016, "learning_rate": 9.92713063681487e-07, - "loss": 0.2875, + "loss": 0.8022, "step": 89990 }, { "epoch": 9.5, - "grad_norm": 0.5833389163017273, + "grad_norm": 42.30424118041992, "learning_rate": 9.906009082268456e-07, - "loss": 0.4923, + "loss": 1.1029, "step": 90000 }, { "epoch": 9.51, - "grad_norm": 6.41705322265625, + "grad_norm": 22.16236686706543, "learning_rate": 9.884887527722042e-07, - "loss": 0.4147, + "loss": 0.7932, "step": 90010 }, { "epoch": 9.51, - "grad_norm": 0.2426244169473648, + "grad_norm": 41.711238861083984, "learning_rate": 9.863765973175627e-07, - "loss": 0.4668, + "loss": 0.8057, "step": 90020 }, { "epoch": 9.51, - "grad_norm": 12.349328994750977, + "grad_norm": 15.374921798706055, "learning_rate": 9.84264441862921e-07, - "loss": 0.4861, + "loss": 0.7976, "step": 90030 }, { "epoch": 9.51, - "grad_norm": 17.007247924804688, + "grad_norm": 21.34056854248047, "learning_rate": 9.821522864082796e-07, - "loss": 0.2461, + "loss": 1.1669, "step": 90040 }, { "epoch": 9.51, - "grad_norm": 31.73910903930664, + "grad_norm": 32.211673736572266, "learning_rate": 9.800401309536382e-07, - "loss": 0.4802, + "loss": 0.7622, "step": 90050 }, { "epoch": 9.51, - "grad_norm": 0.5562722682952881, + "grad_norm": 38.2407112121582, "learning_rate": 9.779279754989968e-07, - "loss": 0.2785, + "loss": 0.8956, "step": 90060 }, { "epoch": 9.51, - "grad_norm": 5.618105411529541, + "grad_norm": 33.77889633178711, "learning_rate": 9.758158200443553e-07, - "loss": 0.3403, + "loss": 0.6468, "step": 90070 }, { "epoch": 9.51, - "grad_norm": 12.828094482421875, + "grad_norm": 43.74882125854492, "learning_rate": 9.73703664589714e-07, - "loss": 0.2026, + "loss": 0.8437, "step": 90080 }, { "epoch": 9.51, - "grad_norm": 46.02744674682617, + "grad_norm": 38.85940170288086, "learning_rate": 9.715915091350723e-07, - "loss": 0.5378, + "loss": 0.9899, "step": 90090 }, { "epoch": 9.52, - "grad_norm": 14.26287841796875, + "grad_norm": 35.39712142944336, "learning_rate": 9.694793536804308e-07, - "loss": 0.3073, + "loss": 0.5867, "step": 90100 }, { "epoch": 9.52, - "grad_norm": 5.087320804595947, + "grad_norm": 42.63869857788086, "learning_rate": 9.673671982257896e-07, - "loss": 0.3409, + "loss": 1.022, "step": 90110 }, { "epoch": 9.52, - "grad_norm": 8.152528762817383, + "grad_norm": 57.81391906738281, "learning_rate": 9.652550427711482e-07, - "loss": 0.5914, + "loss": 1.159, "step": 90120 }, { "epoch": 9.52, - "grad_norm": 10.472633361816406, + "grad_norm": 21.54958152770996, "learning_rate": 9.631428873165065e-07, - "loss": 0.5026, + "loss": 0.4709, "step": 90130 }, { "epoch": 9.52, - "grad_norm": 22.550912857055664, + "grad_norm": 58.55888366699219, "learning_rate": 9.61030731861865e-07, - "loss": 0.208, + "loss": 0.7772, "step": 90140 }, { "epoch": 9.52, - "grad_norm": 25.72873878479004, + "grad_norm": 42.3355598449707, "learning_rate": 9.589185764072237e-07, - "loss": 0.4042, + "loss": 0.758, "step": 90150 }, { "epoch": 9.52, - "grad_norm": 12.846564292907715, + "grad_norm": 35.16487121582031, "learning_rate": 9.568064209525822e-07, - "loss": 0.3021, + "loss": 0.9044, "step": 90160 }, { "epoch": 9.52, - "grad_norm": 2.779663324356079, + "grad_norm": 54.20179748535156, "learning_rate": 9.546942654979408e-07, - "loss": 0.2862, + "loss": 0.575, "step": 90170 }, { "epoch": 9.52, - "grad_norm": 9.162612915039062, + "grad_norm": 48.12510681152344, "learning_rate": 9.525821100432993e-07, - "loss": 0.1888, + "loss": 0.7261, "step": 90180 }, { "epoch": 9.52, - "grad_norm": 35.344093322753906, + "grad_norm": 63.04376983642578, "learning_rate": 9.504699545886578e-07, - "loss": 0.4429, + "loss": 0.9101, "step": 90190 }, { "epoch": 9.53, - "grad_norm": 17.011871337890625, + "grad_norm": 34.96519470214844, "learning_rate": 9.483577991340164e-07, - "loss": 0.4184, + "loss": 0.9848, "step": 90200 }, { "epoch": 9.53, - "grad_norm": 15.54268741607666, + "grad_norm": 58.79498291015625, "learning_rate": 9.462456436793749e-07, - "loss": 0.4831, + "loss": 1.1189, "step": 90210 }, { "epoch": 9.53, - "grad_norm": 12.693855285644531, + "grad_norm": 46.82606506347656, "learning_rate": 9.441334882247334e-07, - "loss": 0.341, + "loss": 0.6033, "step": 90220 }, { "epoch": 9.53, - "grad_norm": 9.793563842773438, + "grad_norm": 45.842830657958984, "learning_rate": 9.42021332770092e-07, - "loss": 0.3453, + "loss": 0.9097, "step": 90230 }, { "epoch": 9.53, - "grad_norm": 9.446759223937988, + "grad_norm": 25.18867301940918, "learning_rate": 9.399091773154505e-07, - "loss": 0.3754, + "loss": 0.6102, "step": 90240 }, { "epoch": 9.53, - "grad_norm": 20.152395248413086, + "grad_norm": 41.179237365722656, "learning_rate": 9.37797021860809e-07, - "loss": 0.315, + "loss": 0.707, "step": 90250 }, { "epoch": 9.53, - "grad_norm": 0.08902973681688309, + "grad_norm": 45.50625991821289, "learning_rate": 9.356848664061676e-07, - "loss": 0.1397, + "loss": 0.6965, "step": 90260 }, { "epoch": 9.53, - "grad_norm": 2.2376463413238525, + "grad_norm": 66.72264099121094, "learning_rate": 9.335727109515261e-07, - "loss": 0.393, + "loss": 0.8641, "step": 90270 }, { "epoch": 9.53, - "grad_norm": 20.27475929260254, + "grad_norm": 31.54636573791504, "learning_rate": 9.314605554968846e-07, - "loss": 0.7155, + "loss": 0.8068, "step": 90280 }, { "epoch": 9.54, - "grad_norm": 19.574129104614258, + "grad_norm": 42.012298583984375, "learning_rate": 9.293484000422432e-07, - "loss": 0.8556, + "loss": 1.2536, "step": 90290 }, { "epoch": 9.54, - "grad_norm": 13.125609397888184, + "grad_norm": 37.32546615600586, "learning_rate": 9.272362445876016e-07, - "loss": 0.1384, + "loss": 0.8127, "step": 90300 }, { "epoch": 9.54, - "grad_norm": 31.080642700195312, + "grad_norm": 30.544790267944336, "learning_rate": 9.251240891329602e-07, - "loss": 0.4468, + "loss": 0.7278, "step": 90310 }, { "epoch": 9.54, - "grad_norm": 0.7375661730766296, + "grad_norm": 58.0800666809082, "learning_rate": 9.230119336783188e-07, - "loss": 0.4283, + "loss": 0.9308, "step": 90320 }, { "epoch": 9.54, - "grad_norm": 12.315911293029785, + "grad_norm": 40.51138687133789, "learning_rate": 9.208997782236775e-07, - "loss": 0.503, + "loss": 0.8253, "step": 90330 }, { "epoch": 9.54, - "grad_norm": 8.955893516540527, + "grad_norm": 29.602863311767578, "learning_rate": 9.187876227690359e-07, - "loss": 0.3713, + "loss": 0.7713, "step": 90340 }, { "epoch": 9.54, - "grad_norm": 15.883452415466309, + "grad_norm": 49.65875244140625, "learning_rate": 9.166754673143945e-07, - "loss": 0.4175, + "loss": 0.8473, "step": 90350 }, { "epoch": 9.54, - "grad_norm": 12.441354751586914, + "grad_norm": 23.966514587402344, "learning_rate": 9.14563311859753e-07, - "loss": 0.2756, + "loss": 0.8984, "step": 90360 }, { "epoch": 9.54, - "grad_norm": 9.087922096252441, + "grad_norm": 35.98883819580078, "learning_rate": 9.124511564051115e-07, - "loss": 0.133, + "loss": 0.5948, "step": 90370 }, { "epoch": 9.54, - "grad_norm": 10.275541305541992, + "grad_norm": 26.083576202392578, "learning_rate": 9.103390009504701e-07, - "loss": 0.4089, + "loss": 0.7481, "step": 90380 }, { "epoch": 9.55, - "grad_norm": 0.5803676843643188, + "grad_norm": 26.53076171875, "learning_rate": 9.082268454958286e-07, - "loss": 0.3448, + "loss": 0.8612, "step": 90390 }, { "epoch": 9.55, - "grad_norm": 4.740476131439209, + "grad_norm": 40.247886657714844, "learning_rate": 9.061146900411871e-07, - "loss": 0.2091, + "loss": 1.0449, "step": 90400 }, { "epoch": 9.55, - "grad_norm": 15.06518268585205, + "grad_norm": 27.59382438659668, "learning_rate": 9.040025345865457e-07, - "loss": 0.5811, + "loss": 1.0001, "step": 90410 }, { "epoch": 9.55, - "grad_norm": 16.79167366027832, + "grad_norm": 50.94527053833008, "learning_rate": 9.018903791319042e-07, - "loss": 0.5354, + "loss": 1.0192, "step": 90420 }, { "epoch": 9.55, - "grad_norm": 0.3987780511379242, + "grad_norm": 50.30308532714844, "learning_rate": 8.997782236772627e-07, - "loss": 0.5038, + "loss": 0.9637, "step": 90430 }, { "epoch": 9.55, - "grad_norm": 9.771740913391113, + "grad_norm": 46.812103271484375, "learning_rate": 8.976660682226213e-07, - "loss": 0.7448, + "loss": 0.949, "step": 90440 }, { "epoch": 9.55, - "grad_norm": 16.129411697387695, + "grad_norm": 13.106730461120605, "learning_rate": 8.955539127679798e-07, - "loss": 0.3121, + "loss": 0.6898, "step": 90450 }, { "epoch": 9.55, - "grad_norm": 11.365227699279785, + "grad_norm": 51.36326217651367, "learning_rate": 8.934417573133383e-07, - "loss": 0.3483, + "loss": 1.0615, "step": 90460 }, { "epoch": 9.55, - "grad_norm": 27.494998931884766, + "grad_norm": 53.176124572753906, "learning_rate": 8.913296018586969e-07, - "loss": 0.1983, + "loss": 0.901, "step": 90470 }, { "epoch": 9.56, - "grad_norm": 17.62584114074707, + "grad_norm": 34.907981872558594, "learning_rate": 8.892174464040554e-07, - "loss": 0.1933, + "loss": 0.9224, "step": 90480 }, { "epoch": 9.56, - "grad_norm": 10.773497581481934, + "grad_norm": 16.488283157348633, "learning_rate": 8.871052909494139e-07, - "loss": 0.2868, + "loss": 0.8399, "step": 90490 }, { "epoch": 9.56, - "grad_norm": 21.420705795288086, + "grad_norm": 46.96750259399414, "learning_rate": 8.849931354947725e-07, - "loss": 0.3518, + "loss": 0.8801, "step": 90500 }, { "epoch": 9.56, - "grad_norm": 2.9336369037628174, + "grad_norm": 36.350521087646484, "learning_rate": 8.82880980040131e-07, - "loss": 0.2894, + "loss": 0.8589, "step": 90510 }, { "epoch": 9.56, - "grad_norm": 9.797449111938477, + "grad_norm": 45.59267807006836, "learning_rate": 8.807688245854895e-07, - "loss": 0.2079, + "loss": 0.8008, "step": 90520 }, { "epoch": 9.56, - "grad_norm": 40.99568557739258, + "grad_norm": 59.63791275024414, "learning_rate": 8.786566691308481e-07, - "loss": 0.1938, + "loss": 0.751, "step": 90530 }, { "epoch": 9.56, - "grad_norm": 31.113101959228516, + "grad_norm": 58.684776306152344, "learning_rate": 8.765445136762066e-07, - "loss": 0.2688, + "loss": 0.77, "step": 90540 }, { "epoch": 9.56, - "grad_norm": 12.311820030212402, + "grad_norm": 45.125797271728516, "learning_rate": 8.744323582215653e-07, - "loss": 0.342, + "loss": 0.7532, "step": 90550 }, { "epoch": 9.56, - "grad_norm": 11.581914901733398, + "grad_norm": 35.87857437133789, "learning_rate": 8.723202027669238e-07, - "loss": 0.3689, + "loss": 0.789, "step": 90560 }, { "epoch": 9.56, - "grad_norm": 39.06825256347656, + "grad_norm": 50.78922653198242, "learning_rate": 8.702080473122823e-07, - "loss": 0.4271, + "loss": 1.0051, "step": 90570 }, { "epoch": 9.57, - "grad_norm": 34.523651123046875, + "grad_norm": 21.892383575439453, "learning_rate": 8.680958918576409e-07, - "loss": 0.3722, + "loss": 0.6889, "step": 90580 }, { "epoch": 9.57, - "grad_norm": 11.708928108215332, + "grad_norm": 3.7694239616394043, "learning_rate": 8.659837364029994e-07, - "loss": 0.3144, + "loss": 0.4692, "step": 90590 }, { "epoch": 9.57, - "grad_norm": 0.026609256863594055, + "grad_norm": 54.201148986816406, "learning_rate": 8.638715809483579e-07, - "loss": 0.5233, + "loss": 0.9428, "step": 90600 }, { "epoch": 9.57, - "grad_norm": 16.14963150024414, + "grad_norm": 49.64582061767578, "learning_rate": 8.617594254937165e-07, - "loss": 0.3681, + "loss": 0.8024, "step": 90610 }, { "epoch": 9.57, - "grad_norm": 11.92678165435791, + "grad_norm": 45.834651947021484, "learning_rate": 8.59647270039075e-07, - "loss": 0.1924, + "loss": 0.8398, "step": 90620 }, { "epoch": 9.57, - "grad_norm": 21.259002685546875, + "grad_norm": 64.75283813476562, "learning_rate": 8.575351145844335e-07, - "loss": 0.4168, + "loss": 0.8275, "step": 90630 }, { "epoch": 9.57, - "grad_norm": 13.167372703552246, + "grad_norm": 18.579862594604492, "learning_rate": 8.554229591297921e-07, - "loss": 0.2809, + "loss": 1.015, "step": 90640 }, { "epoch": 9.57, - "grad_norm": 36.939918518066406, + "grad_norm": 65.9393310546875, "learning_rate": 8.533108036751505e-07, - "loss": 0.1213, + "loss": 0.6694, "step": 90650 }, { "epoch": 9.57, - "grad_norm": 0.07219015806913376, + "grad_norm": 43.71809387207031, "learning_rate": 8.511986482205091e-07, - "loss": 0.6056, + "loss": 0.8378, "step": 90660 }, { "epoch": 9.58, - "grad_norm": 9.839282035827637, + "grad_norm": 52.771358489990234, "learning_rate": 8.490864927658677e-07, - "loss": 0.2844, + "loss": 0.7804, "step": 90670 }, { "epoch": 9.58, - "grad_norm": 15.779833793640137, + "grad_norm": 34.50354766845703, "learning_rate": 8.469743373112261e-07, - "loss": 0.4362, + "loss": 0.8249, "step": 90680 }, { "epoch": 9.58, - "grad_norm": 0.9341706037521362, + "grad_norm": 60.80331802368164, "learning_rate": 8.448621818565847e-07, - "loss": 0.4271, + "loss": 1.0416, "step": 90690 }, { "epoch": 9.58, - "grad_norm": 6.6965155601501465, + "grad_norm": 34.44702911376953, "learning_rate": 8.427500264019433e-07, - "loss": 0.4236, + "loss": 0.774, "step": 90700 }, { "epoch": 9.58, - "grad_norm": 0.07416760921478271, + "grad_norm": 45.074745178222656, "learning_rate": 8.406378709473017e-07, - "loss": 0.0991, + "loss": 0.5949, "step": 90710 }, { "epoch": 9.58, - "grad_norm": 0.07218966633081436, + "grad_norm": 62.50275421142578, "learning_rate": 8.385257154926603e-07, - "loss": 0.4431, + "loss": 1.0964, "step": 90720 }, { "epoch": 9.58, - "grad_norm": 0.6922550797462463, + "grad_norm": 39.008880615234375, "learning_rate": 8.364135600380189e-07, - "loss": 0.2452, + "loss": 0.9118, "step": 90730 }, { "epoch": 9.58, - "grad_norm": 36.39661407470703, + "grad_norm": 49.382240295410156, "learning_rate": 8.343014045833773e-07, - "loss": 0.4034, + "loss": 0.8171, "step": 90740 }, { "epoch": 9.58, - "grad_norm": 11.347256660461426, + "grad_norm": 39.942474365234375, "learning_rate": 8.321892491287359e-07, - "loss": 0.6712, + "loss": 0.6861, "step": 90750 }, { "epoch": 9.58, - "grad_norm": 18.136816024780273, + "grad_norm": 40.973690032958984, "learning_rate": 8.300770936740945e-07, - "loss": 0.2938, + "loss": 0.7374, "step": 90760 }, { "epoch": 9.59, - "grad_norm": 8.986611366271973, + "grad_norm": 22.091005325317383, "learning_rate": 8.279649382194529e-07, - "loss": 0.4283, + "loss": 0.93, "step": 90770 }, { "epoch": 9.59, - "grad_norm": 10.298842430114746, + "grad_norm": 34.71884536743164, "learning_rate": 8.258527827648116e-07, - "loss": 0.4501, + "loss": 1.0376, "step": 90780 }, { "epoch": 9.59, - "grad_norm": 2.371955633163452, + "grad_norm": 56.566383361816406, "learning_rate": 8.237406273101702e-07, - "loss": 0.3641, + "loss": 0.9003, "step": 90790 }, { "epoch": 9.59, - "grad_norm": 0.3243175148963928, + "grad_norm": 46.417354583740234, "learning_rate": 8.216284718555287e-07, - "loss": 0.2144, + "loss": 0.8709, "step": 90800 }, { "epoch": 9.59, - "grad_norm": 1.1304117441177368, + "grad_norm": 61.34220886230469, "learning_rate": 8.195163164008872e-07, - "loss": 0.1972, + "loss": 0.8037, "step": 90810 }, { "epoch": 9.59, - "grad_norm": 1.4294089078903198, + "grad_norm": 52.81298065185547, "learning_rate": 8.174041609462458e-07, - "loss": 0.2223, + "loss": 0.817, "step": 90820 }, { "epoch": 9.59, - "grad_norm": 0.1315637230873108, + "grad_norm": 37.536048889160156, "learning_rate": 8.152920054916043e-07, - "loss": 0.2566, + "loss": 0.4739, "step": 90830 }, { "epoch": 9.59, - "grad_norm": 26.06801986694336, + "grad_norm": 59.14977264404297, "learning_rate": 8.131798500369628e-07, - "loss": 0.3458, + "loss": 0.8523, "step": 90840 }, { "epoch": 9.59, - "grad_norm": 2.1663551330566406, + "grad_norm": 65.6732406616211, "learning_rate": 8.110676945823214e-07, - "loss": 0.1087, + "loss": 0.7292, "step": 90850 }, { "epoch": 9.6, - "grad_norm": 10.098234176635742, + "grad_norm": 29.194732666015625, "learning_rate": 8.089555391276799e-07, - "loss": 0.1483, + "loss": 0.4193, "step": 90860 }, { "epoch": 9.6, - "grad_norm": 2.5219075679779053, + "grad_norm": 56.023406982421875, "learning_rate": 8.068433836730384e-07, - "loss": 0.3154, + "loss": 0.8476, "step": 90870 }, { "epoch": 9.6, - "grad_norm": 3.231099843978882, + "grad_norm": 29.98488998413086, "learning_rate": 8.04731228218397e-07, - "loss": 0.4156, + "loss": 0.7964, "step": 90880 }, { "epoch": 9.6, - "grad_norm": 0.07884915918111801, + "grad_norm": 37.246490478515625, "learning_rate": 8.026190727637555e-07, - "loss": 0.1346, + "loss": 0.8045, "step": 90890 }, { "epoch": 9.6, - "grad_norm": 17.731733322143555, + "grad_norm": 46.38047409057617, "learning_rate": 8.00506917309114e-07, - "loss": 0.3014, + "loss": 0.8752, "step": 90900 }, { "epoch": 9.6, - "grad_norm": 38.040733337402344, + "grad_norm": 46.92715835571289, "learning_rate": 7.983947618544725e-07, - "loss": 0.4901, + "loss": 0.7464, "step": 90910 }, { "epoch": 9.6, - "grad_norm": 8.744112968444824, + "grad_norm": 54.850528717041016, "learning_rate": 7.962826063998311e-07, - "loss": 0.4732, + "loss": 1.0001, "step": 90920 }, { "epoch": 9.6, - "grad_norm": 23.960098266601562, + "grad_norm": 73.52128601074219, "learning_rate": 7.941704509451896e-07, - "loss": 0.3467, + "loss": 0.8099, "step": 90930 }, { "epoch": 9.6, - "grad_norm": 17.90790367126465, + "grad_norm": 37.96084213256836, "learning_rate": 7.920582954905481e-07, - "loss": 0.2329, + "loss": 0.8644, "step": 90940 }, { "epoch": 9.61, - "grad_norm": 0.845077395439148, + "grad_norm": 55.25069808959961, "learning_rate": 7.899461400359067e-07, - "loss": 0.3366, + "loss": 0.7308, "step": 90950 }, { "epoch": 9.61, - "grad_norm": 1.1804951429367065, + "grad_norm": 27.728107452392578, "learning_rate": 7.878339845812652e-07, - "loss": 0.3851, + "loss": 0.7542, "step": 90960 }, { "epoch": 9.61, - "grad_norm": 0.8129525780677795, + "grad_norm": 42.041099548339844, "learning_rate": 7.857218291266237e-07, - "loss": 0.6283, + "loss": 0.8799, "step": 90970 }, { "epoch": 9.61, - "grad_norm": 2.5042941570281982, + "grad_norm": 46.817657470703125, "learning_rate": 7.836096736719823e-07, - "loss": 0.4009, + "loss": 0.6687, "step": 90980 }, { "epoch": 9.61, - "grad_norm": 0.14614351093769073, + "grad_norm": 56.58622741699219, "learning_rate": 7.814975182173408e-07, - "loss": 0.3091, + "loss": 0.8249, "step": 90990 }, { "epoch": 9.61, - "grad_norm": 26.14902687072754, + "grad_norm": 35.92071533203125, "learning_rate": 7.793853627626994e-07, - "loss": 0.1782, + "loss": 0.6441, "step": 91000 }, { "epoch": 9.61, - "grad_norm": 0.19956138730049133, + "grad_norm": 13.483925819396973, "learning_rate": 7.77273207308058e-07, - "loss": 0.3047, + "loss": 0.6853, "step": 91010 }, { "epoch": 9.61, - "grad_norm": 20.471378326416016, + "grad_norm": 51.49970245361328, "learning_rate": 7.751610518534166e-07, - "loss": 0.2753, + "loss": 0.7394, "step": 91020 }, { "epoch": 9.61, - "grad_norm": 2.396139144897461, + "grad_norm": 30.76927947998047, "learning_rate": 7.73048896398775e-07, - "loss": 0.2392, + "loss": 0.847, "step": 91030 }, { "epoch": 9.61, - "grad_norm": 9.084491729736328, + "grad_norm": 2.962906837463379, "learning_rate": 7.709367409441336e-07, - "loss": 0.1741, + "loss": 0.6386, "step": 91040 }, { "epoch": 9.62, - "grad_norm": 4.519883155822754, + "grad_norm": 45.48189163208008, "learning_rate": 7.688245854894922e-07, - "loss": 0.3907, + "loss": 0.7945, "step": 91050 }, { "epoch": 9.62, - "grad_norm": 0.3876182734966278, + "grad_norm": 53.32221984863281, "learning_rate": 7.667124300348506e-07, - "loss": 0.1917, + "loss": 0.7364, "step": 91060 }, { "epoch": 9.62, - "grad_norm": 24.10240936279297, + "grad_norm": 38.8040657043457, "learning_rate": 7.646002745802092e-07, - "loss": 0.3193, + "loss": 0.8491, "step": 91070 }, { "epoch": 9.62, - "grad_norm": 8.65658187866211, + "grad_norm": 35.21381759643555, "learning_rate": 7.624881191255678e-07, - "loss": 0.1962, + "loss": 0.6611, "step": 91080 }, { "epoch": 9.62, - "grad_norm": 9.607185363769531, + "grad_norm": 32.36177062988281, "learning_rate": 7.603759636709262e-07, - "loss": 0.3148, + "loss": 0.728, "step": 91090 }, { "epoch": 9.62, - "grad_norm": 0.02029714547097683, + "grad_norm": 21.013328552246094, "learning_rate": 7.582638082162848e-07, - "loss": 0.3579, + "loss": 0.6661, "step": 91100 }, { "epoch": 9.62, - "grad_norm": 25.176050186157227, + "grad_norm": 51.81443786621094, "learning_rate": 7.561516527616434e-07, - "loss": 0.405, + "loss": 0.7335, "step": 91110 }, { "epoch": 9.62, - "grad_norm": 0.25518301129341125, + "grad_norm": 49.859378814697266, "learning_rate": 7.540394973070018e-07, - "loss": 0.2002, + "loss": 0.7404, "step": 91120 }, { "epoch": 9.62, - "grad_norm": 5.165355682373047, + "grad_norm": 63.179996490478516, "learning_rate": 7.519273418523604e-07, - "loss": 0.2703, + "loss": 1.0043, "step": 91130 }, { "epoch": 9.63, - "grad_norm": 4.458087921142578, + "grad_norm": 36.612022399902344, "learning_rate": 7.49815186397719e-07, - "loss": 0.2279, + "loss": 1.0636, "step": 91140 }, { "epoch": 9.63, - "grad_norm": 22.032854080200195, + "grad_norm": 34.87803268432617, "learning_rate": 7.477030309430774e-07, - "loss": 0.3075, + "loss": 0.7974, "step": 91150 }, { "epoch": 9.63, - "grad_norm": 0.6843864917755127, + "grad_norm": 39.57902526855469, "learning_rate": 7.45590875488436e-07, - "loss": 0.1639, + "loss": 0.6833, "step": 91160 }, { "epoch": 9.63, - "grad_norm": 0.03705647587776184, + "grad_norm": 36.324058532714844, "learning_rate": 7.434787200337945e-07, - "loss": 0.3508, + "loss": 0.8156, "step": 91170 }, { "epoch": 9.63, - "grad_norm": 17.520843505859375, + "grad_norm": 45.16336441040039, "learning_rate": 7.41366564579153e-07, - "loss": 0.432, + "loss": 1.0956, "step": 91180 }, { "epoch": 9.63, - "grad_norm": 9.926667213439941, + "grad_norm": 27.30628204345703, "learning_rate": 7.392544091245116e-07, - "loss": 0.1671, + "loss": 0.6192, "step": 91190 }, { "epoch": 9.63, - "grad_norm": 29.08599090576172, + "grad_norm": 27.444011688232422, "learning_rate": 7.371422536698701e-07, - "loss": 0.3651, + "loss": 0.8814, "step": 91200 }, { "epoch": 9.63, - "grad_norm": 1.2157115936279297, + "grad_norm": 34.05739974975586, "learning_rate": 7.350300982152286e-07, - "loss": 0.3431, + "loss": 0.6889, "step": 91210 }, { "epoch": 9.63, - "grad_norm": 0.32969850301742554, + "grad_norm": 23.182296752929688, "learning_rate": 7.329179427605872e-07, - "loss": 0.1765, + "loss": 0.9327, "step": 91220 }, { "epoch": 9.63, - "grad_norm": 8.461549758911133, + "grad_norm": 30.691417694091797, "learning_rate": 7.308057873059458e-07, - "loss": 0.2474, + "loss": 0.8702, "step": 91230 }, { "epoch": 9.64, - "grad_norm": 12.642951011657715, + "grad_norm": 28.512737274169922, "learning_rate": 7.286936318513044e-07, - "loss": 0.2257, + "loss": 1.0739, "step": 91240 }, { "epoch": 9.64, - "grad_norm": 0.39546144008636475, + "grad_norm": 50.29807662963867, "learning_rate": 7.265814763966629e-07, - "loss": 0.07, + "loss": 0.9548, "step": 91250 }, { "epoch": 9.64, - "grad_norm": 6.3042097091674805, + "grad_norm": 18.05460548400879, "learning_rate": 7.244693209420214e-07, - "loss": 0.336, + "loss": 0.7591, "step": 91260 }, { "epoch": 9.64, - "grad_norm": 3.7415530681610107, + "grad_norm": 41.20867919921875, "learning_rate": 7.2235716548738e-07, - "loss": 0.2388, + "loss": 1.0245, "step": 91270 }, { "epoch": 9.64, - "grad_norm": 2.870445966720581, + "grad_norm": 9.151644706726074, "learning_rate": 7.202450100327385e-07, - "loss": 0.2924, + "loss": 0.7271, "step": 91280 }, { "epoch": 9.64, - "grad_norm": 0.06398669630289078, + "grad_norm": 39.045284271240234, "learning_rate": 7.18132854578097e-07, - "loss": 0.4031, + "loss": 0.8038, "step": 91290 }, { "epoch": 9.64, - "grad_norm": 1.6685630083084106, + "grad_norm": 44.66496658325195, "learning_rate": 7.160206991234556e-07, - "loss": 0.2556, + "loss": 0.8548, "step": 91300 }, { "epoch": 9.64, - "grad_norm": 23.642791748046875, + "grad_norm": 62.26669692993164, "learning_rate": 7.139085436688141e-07, - "loss": 0.4782, + "loss": 0.8644, "step": 91310 }, { "epoch": 9.64, - "grad_norm": 17.472881317138672, + "grad_norm": 57.927696228027344, "learning_rate": 7.117963882141726e-07, - "loss": 0.5201, + "loss": 0.7465, "step": 91320 }, { "epoch": 9.65, - "grad_norm": 2.8355295658111572, + "grad_norm": 31.907365798950195, "learning_rate": 7.096842327595312e-07, - "loss": 0.188, + "loss": 0.7379, "step": 91330 }, { "epoch": 9.65, - "grad_norm": 25.27044105529785, + "grad_norm": 46.65224075317383, "learning_rate": 7.075720773048897e-07, - "loss": 0.3503, + "loss": 1.0576, "step": 91340 }, { "epoch": 9.65, - "grad_norm": 2.5414652824401855, + "grad_norm": 29.491804122924805, "learning_rate": 7.054599218502482e-07, - "loss": 0.1744, + "loss": 0.7126, "step": 91350 }, { "epoch": 9.65, - "grad_norm": 3.344074249267578, + "grad_norm": 48.06638717651367, "learning_rate": 7.033477663956068e-07, - "loss": 0.3163, + "loss": 0.6948, "step": 91360 }, { "epoch": 9.65, - "grad_norm": 26.376638412475586, + "grad_norm": 43.3442497253418, "learning_rate": 7.012356109409653e-07, - "loss": 0.3495, + "loss": 0.7481, "step": 91370 }, { "epoch": 9.65, - "grad_norm": 2.857046127319336, + "grad_norm": 68.05363464355469, "learning_rate": 6.991234554863238e-07, - "loss": 0.3289, + "loss": 0.7821, "step": 91380 }, { "epoch": 9.65, - "grad_norm": 42.07695007324219, + "grad_norm": 52.34551239013672, "learning_rate": 6.970113000316824e-07, - "loss": 0.3481, + "loss": 0.8018, "step": 91390 }, { "epoch": 9.65, - "grad_norm": 24.6014347076416, + "grad_norm": 39.14630889892578, "learning_rate": 6.948991445770409e-07, - "loss": 0.5013, + "loss": 0.7133, "step": 91400 }, { "epoch": 9.65, - "grad_norm": 34.81342315673828, + "grad_norm": 48.81876754760742, "learning_rate": 6.927869891223994e-07, - "loss": 0.399, + "loss": 0.8279, "step": 91410 }, { "epoch": 9.65, - "grad_norm": 7.798862934112549, + "grad_norm": 57.480934143066406, "learning_rate": 6.90674833667758e-07, - "loss": 0.4143, + "loss": 1.0358, "step": 91420 }, { "epoch": 9.66, - "grad_norm": 11.401017189025879, + "grad_norm": 27.00031280517578, "learning_rate": 6.885626782131164e-07, - "loss": 0.5796, + "loss": 0.9297, "step": 91430 }, { "epoch": 9.66, - "grad_norm": 30.488983154296875, + "grad_norm": 50.152225494384766, "learning_rate": 6.86450522758475e-07, - "loss": 0.3655, + "loss": 0.7169, "step": 91440 }, { "epoch": 9.66, - "grad_norm": 0.052423398941755295, + "grad_norm": 36.47332000732422, "learning_rate": 6.843383673038337e-07, - "loss": 0.186, + "loss": 0.7774, "step": 91450 }, { "epoch": 9.66, - "grad_norm": 5.101158142089844, + "grad_norm": 40.914207458496094, "learning_rate": 6.822262118491923e-07, - "loss": 0.3203, + "loss": 0.768, "step": 91460 }, { "epoch": 9.66, - "grad_norm": 0.38649311661720276, + "grad_norm": 30.404748916625977, "learning_rate": 6.801140563945507e-07, - "loss": 0.2863, + "loss": 0.5636, "step": 91470 }, { "epoch": 9.66, - "grad_norm": 27.08209800720215, + "grad_norm": 37.9566650390625, "learning_rate": 6.780019009399093e-07, - "loss": 0.248, + "loss": 0.6871, "step": 91480 }, { "epoch": 9.66, - "grad_norm": 8.141801834106445, + "grad_norm": 52.795196533203125, "learning_rate": 6.758897454852678e-07, - "loss": 0.6341, + "loss": 1.0156, "step": 91490 }, { "epoch": 9.66, - "grad_norm": 11.998536109924316, + "grad_norm": 31.946123123168945, "learning_rate": 6.737775900306263e-07, - "loss": 0.3203, + "loss": 0.9863, "step": 91500 }, { "epoch": 9.66, - "grad_norm": 8.102821350097656, + "grad_norm": 37.825904846191406, "learning_rate": 6.716654345759849e-07, - "loss": 0.314, + "loss": 0.7576, "step": 91510 }, { "epoch": 9.67, - "grad_norm": 9.978463172912598, + "grad_norm": 47.83977127075195, "learning_rate": 6.695532791213434e-07, - "loss": 0.104, + "loss": 1.1165, "step": 91520 }, { "epoch": 9.67, - "grad_norm": 1.2661367654800415, + "grad_norm": 16.196744918823242, "learning_rate": 6.674411236667019e-07, - "loss": 0.2394, + "loss": 0.6936, "step": 91530 }, { "epoch": 9.67, - "grad_norm": 36.00040054321289, + "grad_norm": 41.56190490722656, "learning_rate": 6.653289682120605e-07, - "loss": 0.2631, + "loss": 0.7016, "step": 91540 }, { "epoch": 9.67, - "grad_norm": 30.412460327148438, + "grad_norm": 32.36183166503906, "learning_rate": 6.63216812757419e-07, - "loss": 0.3442, + "loss": 0.7148, "step": 91550 }, { "epoch": 9.67, - "grad_norm": 1.6057958602905273, + "grad_norm": 14.765108108520508, "learning_rate": 6.611046573027775e-07, - "loss": 0.2917, + "loss": 0.6343, "step": 91560 }, { "epoch": 9.67, - "grad_norm": 0.027803553268313408, + "grad_norm": 51.784603118896484, "learning_rate": 6.589925018481361e-07, - "loss": 0.4347, + "loss": 0.8925, "step": 91570 }, { "epoch": 9.67, - "grad_norm": 0.1952759176492691, + "grad_norm": 41.8728141784668, "learning_rate": 6.568803463934946e-07, - "loss": 0.4833, + "loss": 0.9056, "step": 91580 }, { "epoch": 9.67, - "grad_norm": 0.894599199295044, + "grad_norm": 18.612844467163086, "learning_rate": 6.547681909388531e-07, - "loss": 0.3253, + "loss": 0.5935, "step": 91590 }, { "epoch": 9.67, - "grad_norm": 32.85356521606445, + "grad_norm": 57.8775749206543, "learning_rate": 6.526560354842117e-07, - "loss": 0.482, + "loss": 1.1588, "step": 91600 }, { "epoch": 9.67, - "grad_norm": 1.1305097341537476, + "grad_norm": 4.49478006362915, "learning_rate": 6.505438800295702e-07, - "loss": 0.4897, + "loss": 0.9878, "step": 91610 }, { "epoch": 9.68, - "grad_norm": 30.726886749267578, + "grad_norm": 39.53236389160156, "learning_rate": 6.484317245749287e-07, - "loss": 0.2458, + "loss": 0.7826, "step": 91620 }, { "epoch": 9.68, - "grad_norm": 10.174888610839844, + "grad_norm": 39.458805084228516, "learning_rate": 6.463195691202873e-07, - "loss": 0.3288, + "loss": 0.7849, "step": 91630 }, { "epoch": 9.68, - "grad_norm": 5.310745716094971, + "grad_norm": 37.78696823120117, "learning_rate": 6.442074136656458e-07, - "loss": 0.3976, + "loss": 0.6826, "step": 91640 }, { "epoch": 9.68, - "grad_norm": 0.9752566814422607, + "grad_norm": 31.332197189331055, "learning_rate": 6.420952582110043e-07, - "loss": 0.371, + "loss": 0.805, "step": 91650 }, { "epoch": 9.68, - "grad_norm": 0.16022586822509766, + "grad_norm": 42.36564254760742, "learning_rate": 6.399831027563629e-07, - "loss": 0.4327, + "loss": 0.8876, "step": 91660 }, { "epoch": 9.68, - "grad_norm": 29.451993942260742, + "grad_norm": 55.828704833984375, "learning_rate": 6.378709473017215e-07, - "loss": 0.2548, + "loss": 0.8038, "step": 91670 }, { "epoch": 9.68, - "grad_norm": 0.8884323835372925, + "grad_norm": 37.3210563659668, "learning_rate": 6.357587918470801e-07, - "loss": 0.2906, + "loss": 1.0567, "step": 91680 }, { "epoch": 9.68, - "grad_norm": 0.881788432598114, + "grad_norm": 44.240272521972656, "learning_rate": 6.336466363924386e-07, - "loss": 0.1775, + "loss": 0.8429, "step": 91690 }, { "epoch": 9.68, - "grad_norm": 1.0384840965270996, + "grad_norm": 41.87730407714844, "learning_rate": 6.315344809377971e-07, - "loss": 0.45, + "loss": 0.9418, "step": 91700 }, { "epoch": 9.69, - "grad_norm": 32.50480270385742, + "grad_norm": 31.605974197387695, "learning_rate": 6.294223254831557e-07, - "loss": 0.3916, + "loss": 0.9173, "step": 91710 }, { "epoch": 9.69, - "grad_norm": 0.3663174510002136, + "grad_norm": 19.791439056396484, "learning_rate": 6.273101700285142e-07, - "loss": 0.1434, + "loss": 0.6497, "step": 91720 }, { "epoch": 9.69, - "grad_norm": 15.6102933883667, + "grad_norm": 10.083941459655762, "learning_rate": 6.251980145738727e-07, - "loss": 0.2595, + "loss": 0.8712, "step": 91730 }, { "epoch": 9.69, - "grad_norm": 5.036355018615723, + "grad_norm": 51.67003631591797, "learning_rate": 6.230858591192313e-07, - "loss": 0.5342, + "loss": 1.0658, "step": 91740 }, { "epoch": 9.69, - "grad_norm": 10.276276588439941, + "grad_norm": 41.824607849121094, "learning_rate": 6.209737036645897e-07, - "loss": 0.1412, + "loss": 0.8094, "step": 91750 }, { "epoch": 9.69, - "grad_norm": 1.1318987607955933, + "grad_norm": 26.974430084228516, "learning_rate": 6.188615482099483e-07, - "loss": 0.1685, + "loss": 0.4476, "step": 91760 }, { "epoch": 9.69, - "grad_norm": 0.07652599364519119, + "grad_norm": 54.03257369995117, "learning_rate": 6.167493927553069e-07, - "loss": 0.4575, + "loss": 0.7826, "step": 91770 }, { "epoch": 9.69, - "grad_norm": 52.54930114746094, + "grad_norm": 29.37063980102539, "learning_rate": 6.146372373006653e-07, - "loss": 0.4355, + "loss": 0.8286, "step": 91780 }, { "epoch": 9.69, - "grad_norm": 2.450606107711792, + "grad_norm": 37.346675872802734, "learning_rate": 6.125250818460239e-07, - "loss": 0.2518, + "loss": 0.8587, "step": 91790 }, { "epoch": 9.69, - "grad_norm": 26.95941925048828, + "grad_norm": 53.602561950683594, "learning_rate": 6.104129263913825e-07, - "loss": 0.2459, + "loss": 0.7649, "step": 91800 }, { "epoch": 9.7, - "grad_norm": 1.432450532913208, + "grad_norm": 50.561153411865234, "learning_rate": 6.083007709367409e-07, - "loss": 0.2853, + "loss": 0.9387, "step": 91810 }, { "epoch": 9.7, - "grad_norm": 0.02458677813410759, + "grad_norm": 83.54994201660156, "learning_rate": 6.061886154820995e-07, - "loss": 0.301, + "loss": 0.7681, "step": 91820 }, { "epoch": 9.7, - "grad_norm": 0.11375468969345093, + "grad_norm": 23.74069595336914, "learning_rate": 6.040764600274581e-07, - "loss": 0.2, + "loss": 0.7472, "step": 91830 }, { "epoch": 9.7, - "grad_norm": 4.159168720245361, + "grad_norm": 46.964603424072266, "learning_rate": 6.019643045728166e-07, - "loss": 0.2069, + "loss": 0.9103, "step": 91840 }, { "epoch": 9.7, - "grad_norm": 22.95741844177246, + "grad_norm": 38.34809494018555, "learning_rate": 5.998521491181752e-07, - "loss": 0.4917, + "loss": 0.9666, "step": 91850 }, { "epoch": 9.7, - "grad_norm": 40.0131950378418, + "grad_norm": 42.888893127441406, "learning_rate": 5.977399936635337e-07, - "loss": 0.3006, + "loss": 0.774, "step": 91860 }, { "epoch": 9.7, - "grad_norm": 16.094879150390625, + "grad_norm": 54.861175537109375, "learning_rate": 5.956278382088922e-07, - "loss": 0.2785, + "loss": 0.9987, "step": 91870 }, { "epoch": 9.7, - "grad_norm": 21.717119216918945, + "grad_norm": 23.43645477294922, "learning_rate": 5.935156827542508e-07, - "loss": 0.3938, + "loss": 0.7099, "step": 91880 }, { "epoch": 9.7, - "grad_norm": 9.089969635009766, + "grad_norm": 54.619667053222656, "learning_rate": 5.914035272996093e-07, - "loss": 0.0751, + "loss": 0.9985, "step": 91890 }, { "epoch": 9.71, - "grad_norm": 0.008993679657578468, + "grad_norm": 58.282901763916016, "learning_rate": 5.892913718449678e-07, - "loss": 0.4541, + "loss": 0.9538, "step": 91900 }, { "epoch": 9.71, - "grad_norm": 4.069085597991943, + "grad_norm": 7.627738952636719, "learning_rate": 5.871792163903264e-07, - "loss": 0.2879, + "loss": 0.6329, "step": 91910 }, { "epoch": 9.71, - "grad_norm": 19.74635124206543, + "grad_norm": 44.62681198120117, "learning_rate": 5.850670609356849e-07, - "loss": 0.4556, + "loss": 0.9192, "step": 91920 }, { "epoch": 9.71, - "grad_norm": 10.918760299682617, + "grad_norm": 38.360904693603516, "learning_rate": 5.829549054810434e-07, - "loss": 0.528, + "loss": 1.0532, "step": 91930 }, { "epoch": 9.71, - "grad_norm": 17.66785430908203, + "grad_norm": 73.47575378417969, "learning_rate": 5.80842750026402e-07, - "loss": 0.6049, + "loss": 1.0627, "step": 91940 }, { "epoch": 9.71, - "grad_norm": 23.80306053161621, + "grad_norm": 46.8795051574707, "learning_rate": 5.787305945717606e-07, - "loss": 0.3617, + "loss": 0.8449, "step": 91950 }, { "epoch": 9.71, - "grad_norm": 0.1892087310552597, + "grad_norm": 33.8532600402832, "learning_rate": 5.766184391171191e-07, - "loss": 0.6592, + "loss": 0.5275, "step": 91960 }, { "epoch": 9.71, - "grad_norm": 18.78036117553711, + "grad_norm": 51.30249786376953, "learning_rate": 5.745062836624776e-07, - "loss": 0.1397, + "loss": 0.7843, "step": 91970 }, { "epoch": 9.71, - "grad_norm": 22.434598922729492, + "grad_norm": 59.0237922668457, "learning_rate": 5.723941282078362e-07, - "loss": 0.3176, + "loss": 0.9946, "step": 91980 }, { "epoch": 9.71, - "grad_norm": 10.437732696533203, + "grad_norm": 47.55387496948242, "learning_rate": 5.702819727531947e-07, - "loss": 0.2605, + "loss": 0.6739, "step": 91990 }, { "epoch": 9.72, - "grad_norm": 23.79856300354004, + "grad_norm": 40.59804153442383, "learning_rate": 5.681698172985532e-07, - "loss": 0.1801, + "loss": 0.7316, "step": 92000 }, { "epoch": 9.72, - "grad_norm": 40.54852294921875, + "grad_norm": 56.01327896118164, "learning_rate": 5.660576618439117e-07, - "loss": 0.3105, + "loss": 0.7657, "step": 92010 }, { "epoch": 9.72, - "grad_norm": 13.134236335754395, + "grad_norm": 35.534461975097656, "learning_rate": 5.639455063892703e-07, - "loss": 0.3054, + "loss": 0.9599, "step": 92020 }, { "epoch": 9.72, - "grad_norm": 6.398965358734131, + "grad_norm": 16.185144424438477, "learning_rate": 5.618333509346288e-07, - "loss": 0.0975, + "loss": 0.6103, "step": 92030 }, { "epoch": 9.72, - "grad_norm": 0.6140761375427246, + "grad_norm": 56.27544403076172, "learning_rate": 5.597211954799873e-07, - "loss": 0.5942, + "loss": 0.8884, "step": 92040 }, { "epoch": 9.72, - "grad_norm": 0.16928158700466156, + "grad_norm": 3.8154890537261963, "learning_rate": 5.576090400253459e-07, - "loss": 0.3155, + "loss": 0.5937, "step": 92050 }, { "epoch": 9.72, - "grad_norm": 4.182478904724121, + "grad_norm": 49.52070999145508, "learning_rate": 5.554968845707045e-07, - "loss": 0.2525, + "loss": 1.0174, "step": 92060 }, { "epoch": 9.72, - "grad_norm": 4.598725318908691, + "grad_norm": 29.911636352539062, "learning_rate": 5.53384729116063e-07, - "loss": 0.256, + "loss": 0.6283, "step": 92070 }, { "epoch": 9.72, - "grad_norm": 0.46361032128334045, + "grad_norm": 45.743614196777344, "learning_rate": 5.512725736614215e-07, - "loss": 0.2904, + "loss": 0.8106, "step": 92080 }, { "epoch": 9.73, - "grad_norm": 5.7301836013793945, + "grad_norm": 54.57223129272461, "learning_rate": 5.491604182067801e-07, - "loss": 0.293, + "loss": 0.7297, "step": 92090 }, { "epoch": 9.73, - "grad_norm": 0.028667835518717766, + "grad_norm": 7.092637062072754, "learning_rate": 5.470482627521386e-07, - "loss": 0.4025, + "loss": 0.6189, "step": 92100 }, { "epoch": 9.73, - "grad_norm": 12.822351455688477, + "grad_norm": 56.54934310913086, "learning_rate": 5.449361072974971e-07, - "loss": 0.3009, + "loss": 0.8892, "step": 92110 }, { "epoch": 9.73, - "grad_norm": 12.77377700805664, + "grad_norm": 27.917030334472656, "learning_rate": 5.428239518428557e-07, - "loss": 0.4583, + "loss": 0.9172, "step": 92120 }, { "epoch": 9.73, - "grad_norm": 13.00837516784668, + "grad_norm": 28.03529167175293, "learning_rate": 5.407117963882142e-07, - "loss": 0.156, + "loss": 0.5498, "step": 92130 }, { "epoch": 9.73, - "grad_norm": 24.613759994506836, + "grad_norm": 33.32341384887695, "learning_rate": 5.385996409335727e-07, - "loss": 0.5337, + "loss": 0.88, "step": 92140 }, { "epoch": 9.73, - "grad_norm": 24.687522888183594, + "grad_norm": 54.703285217285156, "learning_rate": 5.364874854789313e-07, - "loss": 0.5546, + "loss": 0.9775, "step": 92150 }, { "epoch": 9.73, - "grad_norm": 0.09648080170154572, + "grad_norm": 43.81085968017578, "learning_rate": 5.343753300242898e-07, - "loss": 0.4459, + "loss": 0.8042, "step": 92160 }, { "epoch": 9.73, - "grad_norm": 14.85474967956543, + "grad_norm": 25.320537567138672, "learning_rate": 5.322631745696483e-07, - "loss": 0.2187, + "loss": 0.7999, "step": 92170 }, { "epoch": 9.73, - "grad_norm": 0.08348891884088516, + "grad_norm": 6.929249286651611, "learning_rate": 5.30151019115007e-07, - "loss": 0.5892, + "loss": 1.5358, "step": 92180 }, { "epoch": 9.74, - "grad_norm": 0.04957844689488411, + "grad_norm": 48.82494354248047, "learning_rate": 5.280388636603654e-07, - "loss": 0.2623, + "loss": 0.8952, "step": 92190 }, { "epoch": 9.74, - "grad_norm": 29.916139602661133, + "grad_norm": 52.05002975463867, "learning_rate": 5.25926708205724e-07, - "loss": 0.1843, + "loss": 0.6872, "step": 92200 }, { "epoch": 9.74, - "grad_norm": 8.878111839294434, + "grad_norm": 46.4649658203125, "learning_rate": 5.238145527510826e-07, - "loss": 0.3012, + "loss": 0.5858, "step": 92210 }, { "epoch": 9.74, - "grad_norm": 19.679065704345703, + "grad_norm": 18.103023529052734, "learning_rate": 5.21702397296441e-07, - "loss": 0.2072, + "loss": 0.5536, "step": 92220 }, { "epoch": 9.74, - "grad_norm": 0.8057219386100769, + "grad_norm": 17.0941219329834, "learning_rate": 5.195902418417996e-07, - "loss": 0.3054, + "loss": 0.7368, "step": 92230 }, { "epoch": 9.74, - "grad_norm": 16.377723693847656, + "grad_norm": 23.568029403686523, "learning_rate": 5.174780863871582e-07, - "loss": 0.3076, + "loss": 0.748, "step": 92240 }, { "epoch": 9.74, - "grad_norm": 4.651143550872803, + "grad_norm": 37.31193542480469, "learning_rate": 5.153659309325166e-07, - "loss": 0.2314, + "loss": 0.6742, "step": 92250 }, { "epoch": 9.74, - "grad_norm": 8.72877025604248, + "grad_norm": 50.64253616333008, "learning_rate": 5.132537754778752e-07, - "loss": 0.403, + "loss": 0.8861, "step": 92260 }, { "epoch": 9.74, - "grad_norm": 17.956209182739258, + "grad_norm": 48.7078742980957, "learning_rate": 5.111416200232338e-07, - "loss": 0.4327, + "loss": 1.011, "step": 92270 }, { "epoch": 9.75, - "grad_norm": 0.03723457455635071, + "grad_norm": 22.592859268188477, "learning_rate": 5.090294645685922e-07, - "loss": 0.2682, + "loss": 0.7119, "step": 92280 }, { "epoch": 9.75, - "grad_norm": 6.941770553588867, + "grad_norm": 29.654767990112305, "learning_rate": 5.069173091139509e-07, - "loss": 0.4408, + "loss": 0.7135, "step": 92290 }, { "epoch": 9.75, - "grad_norm": 12.792967796325684, + "grad_norm": 22.49986457824707, "learning_rate": 5.048051536593093e-07, - "loss": 0.6462, + "loss": 1.1726, "step": 92300 }, { "epoch": 9.75, - "grad_norm": 0.025274725630879402, + "grad_norm": 45.85993576049805, "learning_rate": 5.026929982046679e-07, - "loss": 0.3494, + "loss": 0.6756, "step": 92310 }, { "epoch": 9.75, - "grad_norm": 2.6234796047210693, + "grad_norm": 20.958194732666016, "learning_rate": 5.005808427500265e-07, - "loss": 0.2685, + "loss": 0.7136, "step": 92320 }, { "epoch": 9.75, - "grad_norm": 2.392453908920288, + "grad_norm": 47.01417922973633, "learning_rate": 4.984686872953849e-07, - "loss": 0.3971, + "loss": 0.8206, "step": 92330 }, { "epoch": 9.75, - "grad_norm": 13.28898811340332, + "grad_norm": 41.954647064208984, "learning_rate": 4.963565318407435e-07, - "loss": 0.3332, + "loss": 0.7612, "step": 92340 }, { "epoch": 9.75, - "grad_norm": 1.2095608711242676, + "grad_norm": 43.487510681152344, "learning_rate": 4.942443763861021e-07, - "loss": 0.1571, + "loss": 0.9179, "step": 92350 }, { "epoch": 9.75, - "grad_norm": 0.6927649974822998, + "grad_norm": 53.704586029052734, "learning_rate": 4.921322209314605e-07, - "loss": 0.1147, + "loss": 0.8531, "step": 92360 }, { "epoch": 9.75, - "grad_norm": 0.14353874325752258, + "grad_norm": 28.632434844970703, "learning_rate": 4.900200654768191e-07, - "loss": 0.3322, + "loss": 0.8555, "step": 92370 }, { "epoch": 9.76, - "grad_norm": 9.927435874938965, + "grad_norm": 50.07643508911133, "learning_rate": 4.879079100221777e-07, - "loss": 0.3533, + "loss": 0.7327, "step": 92380 }, { "epoch": 9.76, - "grad_norm": 21.373699188232422, + "grad_norm": 55.09672927856445, "learning_rate": 4.857957545675361e-07, - "loss": 0.3598, + "loss": 0.7772, "step": 92390 }, { "epoch": 9.76, - "grad_norm": 7.226205348968506, + "grad_norm": 43.34902572631836, "learning_rate": 4.836835991128948e-07, - "loss": 0.2741, + "loss": 0.6294, "step": 92400 }, { "epoch": 9.76, - "grad_norm": 12.742669105529785, + "grad_norm": 26.950063705444336, "learning_rate": 4.815714436582533e-07, - "loss": 0.345, + "loss": 0.655, "step": 92410 }, { "epoch": 9.76, - "grad_norm": 0.3910374939441681, + "grad_norm": 37.058677673339844, "learning_rate": 4.794592882036118e-07, - "loss": 0.2058, + "loss": 0.8409, "step": 92420 }, { "epoch": 9.76, - "grad_norm": 17.939517974853516, + "grad_norm": 41.0189323425293, "learning_rate": 4.773471327489704e-07, - "loss": 0.3412, + "loss": 0.735, "step": 92430 }, { "epoch": 9.76, - "grad_norm": 5.559237957000732, + "grad_norm": 40.27288055419922, "learning_rate": 4.752349772943289e-07, - "loss": 0.1693, + "loss": 0.6908, "step": 92440 }, { "epoch": 9.76, - "grad_norm": 2.6019020080566406, + "grad_norm": 44.6942024230957, "learning_rate": 4.7312282183968743e-07, - "loss": 0.1447, + "loss": 0.7953, "step": 92450 }, { "epoch": 9.76, - "grad_norm": 32.844783782958984, + "grad_norm": 32.78018569946289, "learning_rate": 4.71010666385046e-07, - "loss": 0.2039, + "loss": 0.7746, "step": 92460 }, { "epoch": 9.77, - "grad_norm": 40.2318229675293, + "grad_norm": 45.30517578125, "learning_rate": 4.688985109304045e-07, - "loss": 0.1785, + "loss": 0.5734, "step": 92470 }, { "epoch": 9.77, - "grad_norm": 15.186697959899902, + "grad_norm": 40.930538177490234, "learning_rate": 4.6678635547576303e-07, - "loss": 0.4686, + "loss": 0.8008, "step": 92480 }, { "epoch": 9.77, - "grad_norm": 20.979001998901367, + "grad_norm": 62.62435531616211, "learning_rate": 4.646742000211216e-07, - "loss": 0.4369, + "loss": 0.8099, "step": 92490 }, { "epoch": 9.77, - "grad_norm": 0.09654127806425095, + "grad_norm": 31.868682861328125, "learning_rate": 4.625620445664801e-07, - "loss": 0.1994, + "loss": 0.807, "step": 92500 }, { "epoch": 9.77, - "grad_norm": 45.757572174072266, + "grad_norm": 50.004581451416016, "learning_rate": 4.6044988911183873e-07, - "loss": 0.3609, + "loss": 0.7968, "step": 92510 }, { "epoch": 9.77, - "grad_norm": 15.779943466186523, + "grad_norm": 43.35755157470703, "learning_rate": 4.5833773365719724e-07, - "loss": 0.399, + "loss": 1.1185, "step": 92520 }, { "epoch": 9.77, - "grad_norm": 0.22179876267910004, + "grad_norm": 30.86554527282715, "learning_rate": 4.5622557820255576e-07, - "loss": 0.3808, + "loss": 0.646, "step": 92530 }, { "epoch": 9.77, - "grad_norm": 0.013814330101013184, + "grad_norm": 4.2680768966674805, "learning_rate": 4.541134227479143e-07, - "loss": 0.3847, + "loss": 0.7093, "step": 92540 }, { "epoch": 9.77, - "grad_norm": 26.604251861572266, + "grad_norm": 58.07631301879883, "learning_rate": 4.5200126729327284e-07, - "loss": 0.284, + "loss": 0.9342, "step": 92550 }, { "epoch": 9.78, - "grad_norm": 8.657872200012207, + "grad_norm": 33.449546813964844, "learning_rate": 4.4988911183863135e-07, - "loss": 0.258, + "loss": 0.8016, "step": 92560 }, { "epoch": 9.78, - "grad_norm": 18.976438522338867, + "grad_norm": 40.422847747802734, "learning_rate": 4.477769563839899e-07, - "loss": 0.2279, + "loss": 0.884, "step": 92570 }, { "epoch": 9.78, - "grad_norm": 5.169310569763184, + "grad_norm": 31.67963218688965, "learning_rate": 4.4566480092934843e-07, - "loss": 0.2668, + "loss": 0.8754, "step": 92580 }, { "epoch": 9.78, - "grad_norm": 10.027911186218262, + "grad_norm": 38.69776153564453, "learning_rate": 4.4355264547470695e-07, - "loss": 0.1459, + "loss": 0.6068, "step": 92590 }, { "epoch": 9.78, - "grad_norm": 0.37507644295692444, + "grad_norm": 57.04981231689453, "learning_rate": 4.414404900200655e-07, - "loss": 0.2693, + "loss": 0.9021, "step": 92600 }, { "epoch": 9.78, - "grad_norm": 0.4498390257358551, + "grad_norm": 32.30918884277344, "learning_rate": 4.3932833456542403e-07, - "loss": 0.3074, + "loss": 1.0308, "step": 92610 }, { "epoch": 9.78, - "grad_norm": 14.549542427062988, + "grad_norm": 27.982419967651367, "learning_rate": 4.3721617911078265e-07, - "loss": 0.3927, + "loss": 0.663, "step": 92620 }, { "epoch": 9.78, - "grad_norm": 1.6289970874786377, + "grad_norm": 34.9162712097168, "learning_rate": 4.3510402365614116e-07, - "loss": 0.1948, + "loss": 0.7015, "step": 92630 }, { "epoch": 9.78, - "grad_norm": 7.412614822387695, + "grad_norm": 30.208675384521484, "learning_rate": 4.329918682014997e-07, - "loss": 0.5557, + "loss": 0.94, "step": 92640 }, { "epoch": 9.78, - "grad_norm": 0.1368979811668396, + "grad_norm": 39.02113342285156, "learning_rate": 4.3087971274685824e-07, - "loss": 0.1976, + "loss": 0.9327, "step": 92650 }, { "epoch": 9.79, - "grad_norm": 5.119966983795166, + "grad_norm": 28.86403465270996, "learning_rate": 4.2876755729221676e-07, - "loss": 0.2579, + "loss": 1.2347, "step": 92660 }, { "epoch": 9.79, - "grad_norm": 17.045963287353516, + "grad_norm": 40.4588623046875, "learning_rate": 4.2665540183757527e-07, - "loss": 0.3039, + "loss": 0.8905, "step": 92670 }, { "epoch": 9.79, - "grad_norm": 12.750795364379883, + "grad_norm": 36.47873306274414, "learning_rate": 4.2454324638293384e-07, - "loss": 0.475, + "loss": 0.7899, "step": 92680 }, { "epoch": 9.79, - "grad_norm": 33.793922424316406, + "grad_norm": 38.84761047363281, "learning_rate": 4.2243109092829235e-07, - "loss": 0.2306, + "loss": 0.7588, "step": 92690 }, { "epoch": 9.79, - "grad_norm": 1.6838597059249878, + "grad_norm": 50.124141693115234, "learning_rate": 4.2031893547365087e-07, - "loss": 0.1654, + "loss": 0.6921, "step": 92700 }, { "epoch": 9.79, - "grad_norm": 3.6123385429382324, + "grad_norm": 52.02468490600586, "learning_rate": 4.1820678001900943e-07, - "loss": 0.2054, + "loss": 0.5547, "step": 92710 }, { "epoch": 9.79, - "grad_norm": 5.598217487335205, + "grad_norm": 22.08332633972168, "learning_rate": 4.1609462456436795e-07, - "loss": 0.1766, + "loss": 0.8812, "step": 92720 }, { "epoch": 9.79, - "grad_norm": 21.72490692138672, + "grad_norm": 54.23299026489258, "learning_rate": 4.1398246910972646e-07, - "loss": 0.4328, + "loss": 1.001, "step": 92730 }, { "epoch": 9.79, - "grad_norm": 0.6704733371734619, + "grad_norm": 40.64921569824219, "learning_rate": 4.118703136550851e-07, - "loss": 0.4143, + "loss": 0.9547, "step": 92740 }, { "epoch": 9.8, - "grad_norm": 10.396505355834961, + "grad_norm": 73.08451080322266, "learning_rate": 4.097581582004436e-07, - "loss": 0.4435, + "loss": 0.9813, "step": 92750 }, { "epoch": 9.8, - "grad_norm": 37.61032485961914, + "grad_norm": 32.25030517578125, "learning_rate": 4.0764600274580216e-07, - "loss": 0.561, + "loss": 0.7099, "step": 92760 }, { "epoch": 9.8, - "grad_norm": 31.4813289642334, + "grad_norm": 48.10563659667969, "learning_rate": 4.055338472911607e-07, - "loss": 0.3416, + "loss": 0.978, "step": 92770 }, { "epoch": 9.8, - "grad_norm": 0.03214171528816223, + "grad_norm": 71.95187377929688, "learning_rate": 4.034216918365192e-07, - "loss": 0.542, + "loss": 1.1332, "step": 92780 }, { "epoch": 9.8, - "grad_norm": 33.03059768676758, + "grad_norm": 42.39582824707031, "learning_rate": 4.0130953638187776e-07, - "loss": 0.4692, + "loss": 1.0173, "step": 92790 }, { "epoch": 9.8, - "grad_norm": 24.15250587463379, + "grad_norm": 51.743408203125, "learning_rate": 3.9919738092723627e-07, - "loss": 0.2229, + "loss": 0.7238, "step": 92800 }, { "epoch": 9.8, - "grad_norm": 0.03569893538951874, + "grad_norm": 25.257524490356445, "learning_rate": 3.970852254725948e-07, - "loss": 0.1633, + "loss": 0.6784, "step": 92810 }, { "epoch": 9.8, - "grad_norm": 1.2661199569702148, + "grad_norm": 34.8839225769043, "learning_rate": 3.9497307001795335e-07, - "loss": 0.251, + "loss": 0.582, "step": 92820 }, { "epoch": 9.8, - "grad_norm": 0.5058748722076416, + "grad_norm": 42.632022857666016, "learning_rate": 3.9286091456331187e-07, - "loss": 0.3705, + "loss": 0.757, "step": 92830 }, { "epoch": 9.8, - "grad_norm": 16.629878997802734, + "grad_norm": 33.78776550292969, "learning_rate": 3.907487591086704e-07, - "loss": 0.1128, + "loss": 0.6866, "step": 92840 }, { "epoch": 9.81, - "grad_norm": 11.47283935546875, + "grad_norm": 49.54443359375, "learning_rate": 3.88636603654029e-07, - "loss": 0.3236, + "loss": 0.7379, "step": 92850 }, { "epoch": 9.81, - "grad_norm": 13.610118865966797, + "grad_norm": 40.140140533447266, "learning_rate": 3.865244481993875e-07, - "loss": 0.1769, + "loss": 0.6919, "step": 92860 }, { "epoch": 9.81, - "grad_norm": 34.26518249511719, + "grad_norm": 56.32255554199219, "learning_rate": 3.844122927447461e-07, - "loss": 0.354, + "loss": 0.7124, "step": 92870 }, { "epoch": 9.81, - "grad_norm": 1.3657145500183105, + "grad_norm": 7.326492786407471, "learning_rate": 3.823001372901046e-07, - "loss": 0.2277, + "loss": 0.6064, "step": 92880 }, { "epoch": 9.81, - "grad_norm": 38.4256477355957, + "grad_norm": 68.25688171386719, "learning_rate": 3.801879818354631e-07, - "loss": 0.505, + "loss": 0.9388, "step": 92890 }, { "epoch": 9.81, - "grad_norm": 21.215425491333008, + "grad_norm": 46.75513458251953, "learning_rate": 3.780758263808217e-07, - "loss": 0.3577, + "loss": 0.8217, "step": 92900 }, { "epoch": 9.81, - "grad_norm": 3.695166826248169, + "grad_norm": 37.4159049987793, "learning_rate": 3.759636709261802e-07, - "loss": 0.2675, + "loss": 0.6552, "step": 92910 }, { "epoch": 9.81, - "grad_norm": 22.72596549987793, + "grad_norm": 66.97799682617188, "learning_rate": 3.738515154715387e-07, - "loss": 0.3315, + "loss": 0.8402, "step": 92920 }, { "epoch": 9.81, - "grad_norm": 18.648042678833008, + "grad_norm": 51.97731399536133, "learning_rate": 3.717393600168973e-07, - "loss": 0.2072, + "loss": 0.7219, "step": 92930 }, { "epoch": 9.82, - "grad_norm": 0.11061213910579681, + "grad_norm": 42.892051696777344, "learning_rate": 3.696272045622558e-07, - "loss": 0.0695, + "loss": 0.8104, "step": 92940 }, { "epoch": 9.82, - "grad_norm": 0.06645575910806656, + "grad_norm": 81.3183822631836, "learning_rate": 3.675150491076143e-07, - "loss": 0.1219, + "loss": 0.6655, "step": 92950 }, { "epoch": 9.82, - "grad_norm": 25.285892486572266, + "grad_norm": 49.49727249145508, "learning_rate": 3.654028936529729e-07, - "loss": 0.2989, + "loss": 1.0592, "step": 92960 }, { "epoch": 9.82, - "grad_norm": 0.4908931255340576, + "grad_norm": 51.38223648071289, "learning_rate": 3.6329073819833144e-07, - "loss": 0.3604, + "loss": 0.8925, "step": 92970 }, { "epoch": 9.82, - "grad_norm": 15.325206756591797, + "grad_norm": 19.787689208984375, "learning_rate": 3.6117858274369e-07, - "loss": 0.4602, + "loss": 0.8521, "step": 92980 }, { "epoch": 9.82, - "grad_norm": 0.07303480058908463, + "grad_norm": 20.647520065307617, "learning_rate": 3.590664272890485e-07, - "loss": 0.4669, + "loss": 0.9459, "step": 92990 }, { "epoch": 9.82, - "grad_norm": 30.12611961364746, + "grad_norm": 34.79795455932617, "learning_rate": 3.5695427183440703e-07, - "loss": 0.416, + "loss": 1.0929, "step": 93000 }, { "epoch": 9.82, - "grad_norm": 6.1275954246521, + "grad_norm": 36.417545318603516, "learning_rate": 3.548421163797656e-07, - "loss": 0.1933, + "loss": 0.6069, "step": 93010 }, { "epoch": 9.82, - "grad_norm": 0.22522102296352386, + "grad_norm": 40.0058479309082, "learning_rate": 3.527299609251241e-07, - "loss": 0.3616, + "loss": 0.8537, "step": 93020 }, { "epoch": 9.82, - "grad_norm": 8.735541343688965, + "grad_norm": 17.205734252929688, "learning_rate": 3.5061780547048263e-07, - "loss": 0.1836, + "loss": 0.9432, "step": 93030 }, { "epoch": 9.83, - "grad_norm": 14.465035438537598, + "grad_norm": 26.316457748413086, "learning_rate": 3.485056500158412e-07, - "loss": 0.3718, + "loss": 0.7789, "step": 93040 }, { "epoch": 9.83, - "grad_norm": 50.49906539916992, + "grad_norm": 42.34242248535156, "learning_rate": 3.463934945611997e-07, - "loss": 0.3091, + "loss": 0.5253, "step": 93050 }, { "epoch": 9.83, - "grad_norm": 3.400045394897461, + "grad_norm": 15.141295433044434, "learning_rate": 3.442813391065582e-07, - "loss": 0.1492, + "loss": 0.9042, "step": 93060 }, { "epoch": 9.83, - "grad_norm": 33.426143646240234, + "grad_norm": 39.13779830932617, "learning_rate": 3.4216918365191684e-07, - "loss": 0.2044, + "loss": 0.9424, "step": 93070 }, { "epoch": 9.83, - "grad_norm": 0.8815546631813049, + "grad_norm": 33.4013786315918, "learning_rate": 3.4005702819727536e-07, - "loss": 0.5744, + "loss": 1.2784, "step": 93080 }, { "epoch": 9.83, - "grad_norm": 0.7784902453422546, + "grad_norm": 24.536457061767578, "learning_rate": 3.379448727426339e-07, - "loss": 0.2824, + "loss": 1.0516, "step": 93090 }, { "epoch": 9.83, - "grad_norm": 27.59861946105957, + "grad_norm": 75.16217803955078, "learning_rate": 3.3583271728799244e-07, - "loss": 0.4962, + "loss": 0.8733, "step": 93100 }, { "epoch": 9.83, - "grad_norm": 0.10639912635087967, + "grad_norm": 40.324649810791016, "learning_rate": 3.3372056183335095e-07, - "loss": 0.1485, + "loss": 0.6243, "step": 93110 }, { "epoch": 9.83, - "grad_norm": 0.6303146481513977, + "grad_norm": 56.67033767700195, "learning_rate": 3.316084063787095e-07, - "loss": 0.1314, + "loss": 0.9534, "step": 93120 }, { "epoch": 9.84, - "grad_norm": 17.147459030151367, + "grad_norm": 46.5776481628418, "learning_rate": 3.2949625092406803e-07, - "loss": 0.2432, + "loss": 0.9805, "step": 93130 }, { "epoch": 9.84, - "grad_norm": 23.076881408691406, + "grad_norm": 37.745262145996094, "learning_rate": 3.2738409546942655e-07, - "loss": 0.3063, + "loss": 0.5565, "step": 93140 }, { "epoch": 9.84, - "grad_norm": 0.17130112648010254, + "grad_norm": 73.93975830078125, "learning_rate": 3.252719400147851e-07, - "loss": 0.2698, + "loss": 0.9208, "step": 93150 }, { "epoch": 9.84, - "grad_norm": 21.73076629638672, + "grad_norm": 40.42594528198242, "learning_rate": 3.2315978456014363e-07, - "loss": 0.3668, + "loss": 0.7764, "step": 93160 }, { "epoch": 9.84, - "grad_norm": 0.6155845522880554, + "grad_norm": 70.90776824951172, "learning_rate": 3.2104762910550214e-07, - "loss": 0.1933, + "loss": 0.6958, "step": 93170 }, { "epoch": 9.84, - "grad_norm": 32.67874526977539, + "grad_norm": 7.092336177825928, "learning_rate": 3.1893547365086076e-07, - "loss": 0.2408, + "loss": 0.6529, "step": 93180 }, { "epoch": 9.84, - "grad_norm": 0.8249304890632629, + "grad_norm": 49.68242263793945, "learning_rate": 3.168233181962193e-07, - "loss": 0.268, + "loss": 0.801, "step": 93190 }, { "epoch": 9.84, - "grad_norm": 10.214070320129395, + "grad_norm": 50.85994338989258, "learning_rate": 3.1471116274157785e-07, - "loss": 0.3796, + "loss": 0.8553, "step": 93200 }, { "epoch": 9.84, - "grad_norm": 31.201642990112305, + "grad_norm": 49.712337493896484, "learning_rate": 3.1259900728693636e-07, - "loss": 0.5028, + "loss": 0.8628, "step": 93210 }, { "epoch": 9.84, - "grad_norm": 0.08253432810306549, + "grad_norm": 31.256420135498047, "learning_rate": 3.1048685183229487e-07, - "loss": 0.3213, + "loss": 0.8425, "step": 93220 }, { "epoch": 9.85, - "grad_norm": 6.762928009033203, + "grad_norm": 59.29963302612305, "learning_rate": 3.0837469637765344e-07, - "loss": 0.1575, + "loss": 0.8655, "step": 93230 }, { "epoch": 9.85, - "grad_norm": 38.879791259765625, + "grad_norm": 58.52448272705078, "learning_rate": 3.0626254092301195e-07, - "loss": 0.5665, + "loss": 1.193, "step": 93240 }, { "epoch": 9.85, - "grad_norm": 0.0677885115146637, + "grad_norm": 37.19865798950195, "learning_rate": 3.0415038546837047e-07, - "loss": 0.3113, + "loss": 0.9494, "step": 93250 }, { "epoch": 9.85, - "grad_norm": 0.333970308303833, + "grad_norm": 22.112293243408203, "learning_rate": 3.0203823001372904e-07, - "loss": 0.4614, + "loss": 0.7486, "step": 93260 }, { "epoch": 9.85, - "grad_norm": 0.4106081426143646, + "grad_norm": 33.9344367980957, "learning_rate": 2.999260745590876e-07, - "loss": 0.3654, + "loss": 0.907, "step": 93270 }, { "epoch": 9.85, - "grad_norm": 26.08648681640625, + "grad_norm": 21.856971740722656, "learning_rate": 2.978139191044461e-07, - "loss": 0.2398, + "loss": 0.871, "step": 93280 }, { "epoch": 9.85, - "grad_norm": 1.0363154411315918, + "grad_norm": 56.913856506347656, "learning_rate": 2.9570176364980463e-07, - "loss": 0.4041, + "loss": 1.0357, "step": 93290 }, { "epoch": 9.85, - "grad_norm": 0.9162876605987549, + "grad_norm": 38.09214782714844, "learning_rate": 2.935896081951632e-07, - "loss": 0.2169, + "loss": 0.8686, "step": 93300 }, { "epoch": 9.85, - "grad_norm": 25.391454696655273, + "grad_norm": 35.74973678588867, "learning_rate": 2.914774527405217e-07, - "loss": 0.1989, + "loss": 0.6371, "step": 93310 }, { "epoch": 9.86, - "grad_norm": 35.28892135620117, + "grad_norm": 50.348323822021484, "learning_rate": 2.893652972858803e-07, - "loss": 0.4197, + "loss": 1.1114, "step": 93320 }, { "epoch": 9.86, - "grad_norm": 10.878348350524902, + "grad_norm": 41.07093811035156, "learning_rate": 2.872531418312388e-07, - "loss": 0.5465, + "loss": 0.9903, "step": 93330 }, { "epoch": 9.86, - "grad_norm": 21.478723526000977, + "grad_norm": 44.763851165771484, "learning_rate": 2.8514098637659736e-07, - "loss": 0.2953, + "loss": 1.0214, "step": 93340 }, { "epoch": 9.86, - "grad_norm": 3.6020331382751465, + "grad_norm": 55.046051025390625, "learning_rate": 2.830288309219559e-07, - "loss": 0.2052, + "loss": 0.7821, "step": 93350 }, { "epoch": 9.86, - "grad_norm": 0.5232782959938049, + "grad_norm": 38.454986572265625, "learning_rate": 2.809166754673144e-07, - "loss": 0.3965, + "loss": 0.8387, "step": 93360 }, { "epoch": 9.86, - "grad_norm": 8.138541221618652, + "grad_norm": 47.963233947753906, "learning_rate": 2.7880452001267296e-07, - "loss": 0.2852, + "loss": 0.9586, "step": 93370 }, { "epoch": 9.86, - "grad_norm": 10.273197174072266, + "grad_norm": 25.23090934753418, "learning_rate": 2.766923645580315e-07, - "loss": 0.3219, + "loss": 0.8982, "step": 93380 }, { "epoch": 9.86, - "grad_norm": 0.03050088882446289, + "grad_norm": 42.701515197753906, "learning_rate": 2.7458020910339004e-07, - "loss": 0.4835, + "loss": 0.9366, "step": 93390 }, { "epoch": 9.86, - "grad_norm": 8.080219268798828, + "grad_norm": 30.908241271972656, "learning_rate": 2.7246805364874855e-07, - "loss": 0.4111, + "loss": 0.8706, "step": 93400 }, { "epoch": 9.86, - "grad_norm": 17.27360725402832, + "grad_norm": 40.256263732910156, "learning_rate": 2.703558981941071e-07, - "loss": 0.4602, + "loss": 0.7657, "step": 93410 }, { "epoch": 9.87, - "grad_norm": 26.517980575561523, + "grad_norm": 33.04076385498047, "learning_rate": 2.6824374273946563e-07, - "loss": 0.3751, + "loss": 0.6851, "step": 93420 }, { "epoch": 9.87, - "grad_norm": 26.831941604614258, + "grad_norm": 51.32855224609375, "learning_rate": 2.6613158728482415e-07, - "loss": 0.455, + "loss": 0.888, "step": 93430 }, { "epoch": 9.87, - "grad_norm": 10.351744651794434, + "grad_norm": 48.8125114440918, "learning_rate": 2.640194318301827e-07, - "loss": 0.2821, + "loss": 0.6154, "step": 93440 }, { "epoch": 9.87, - "grad_norm": 11.384870529174805, + "grad_norm": 53.742340087890625, "learning_rate": 2.619072763755413e-07, - "loss": 0.3317, + "loss": 0.9369, "step": 93450 }, { "epoch": 9.87, - "grad_norm": 29.06636619567871, + "grad_norm": 52.94428634643555, "learning_rate": 2.597951209208998e-07, - "loss": 0.3845, + "loss": 1.2023, "step": 93460 }, { "epoch": 9.87, - "grad_norm": 13.550580978393555, + "grad_norm": 51.989219665527344, "learning_rate": 2.576829654662583e-07, - "loss": 0.8164, + "loss": 1.071, "step": 93470 }, { "epoch": 9.87, - "grad_norm": 25.310800552368164, + "grad_norm": 33.95663070678711, "learning_rate": 2.555708100116169e-07, - "loss": 0.3085, + "loss": 0.8688, "step": 93480 }, { "epoch": 9.87, - "grad_norm": 0.007905194535851479, + "grad_norm": 20.458511352539062, "learning_rate": 2.5345865455697544e-07, - "loss": 0.3312, + "loss": 0.6355, "step": 93490 }, { "epoch": 9.87, - "grad_norm": 8.982732772827148, + "grad_norm": 39.48476028442383, "learning_rate": 2.5134649910233396e-07, - "loss": 0.2071, + "loss": 0.754, "step": 93500 }, { "epoch": 9.88, - "grad_norm": 0.08539891242980957, + "grad_norm": 43.72175216674805, "learning_rate": 2.4923434364769247e-07, - "loss": 0.3795, + "loss": 0.7102, "step": 93510 }, { "epoch": 9.88, - "grad_norm": 5.146078109741211, + "grad_norm": 54.86013412475586, "learning_rate": 2.4712218819305104e-07, - "loss": 0.5064, + "loss": 1.0487, "step": 93520 }, { "epoch": 9.88, - "grad_norm": 0.2132498323917389, + "grad_norm": 15.800421714782715, "learning_rate": 2.4501003273840955e-07, - "loss": 0.0725, + "loss": 0.6321, "step": 93530 }, { "epoch": 9.88, - "grad_norm": 21.635574340820312, + "grad_norm": 65.92176818847656, "learning_rate": 2.4289787728376807e-07, - "loss": 0.1654, + "loss": 0.716, "step": 93540 }, { "epoch": 9.88, - "grad_norm": 18.802091598510742, + "grad_norm": 35.46726608276367, "learning_rate": 2.4078572182912663e-07, - "loss": 0.4046, + "loss": 0.867, "step": 93550 }, { "epoch": 9.88, - "grad_norm": 27.650590896606445, + "grad_norm": 40.527679443359375, "learning_rate": 2.386735663744852e-07, - "loss": 0.4519, + "loss": 0.8996, "step": 93560 }, { "epoch": 9.88, - "grad_norm": 0.02704506739974022, + "grad_norm": 17.508926391601562, "learning_rate": 2.3656141091984372e-07, - "loss": 0.3053, + "loss": 0.5821, "step": 93570 }, { "epoch": 9.88, - "grad_norm": 1.3543130159378052, + "grad_norm": 33.46646499633789, "learning_rate": 2.3444925546520226e-07, - "loss": 0.2407, + "loss": 1.034, "step": 93580 }, { "epoch": 9.88, - "grad_norm": 12.662484169006348, + "grad_norm": 72.0510025024414, "learning_rate": 2.323371000105608e-07, - "loss": 0.4331, + "loss": 1.1004, "step": 93590 }, { "epoch": 9.88, - "grad_norm": 0.5168079733848572, + "grad_norm": 9.72517204284668, "learning_rate": 2.3022494455591936e-07, - "loss": 0.1416, + "loss": 0.7093, "step": 93600 }, { "epoch": 9.89, - "grad_norm": 7.360469341278076, + "grad_norm": 52.97850799560547, "learning_rate": 2.2811278910127788e-07, - "loss": 0.2602, + "loss": 0.5816, "step": 93610 }, { "epoch": 9.89, - "grad_norm": 2.5985782146453857, + "grad_norm": 19.63921356201172, "learning_rate": 2.2600063364663642e-07, - "loss": 0.1869, + "loss": 0.8993, "step": 93620 }, { "epoch": 9.89, - "grad_norm": 0.11120697855949402, + "grad_norm": 23.060781478881836, "learning_rate": 2.2388847819199496e-07, - "loss": 0.1688, + "loss": 0.7263, "step": 93630 }, { "epoch": 9.89, - "grad_norm": 0.019960252568125725, + "grad_norm": 60.21377182006836, "learning_rate": 2.2177632273735347e-07, - "loss": 0.3254, + "loss": 0.7499, "step": 93640 }, { "epoch": 9.89, - "grad_norm": 17.153297424316406, + "grad_norm": 49.06014633178711, "learning_rate": 2.1966416728271201e-07, - "loss": 0.2563, + "loss": 0.7069, "step": 93650 }, { "epoch": 9.89, - "grad_norm": 8.80745792388916, + "grad_norm": 52.02843475341797, "learning_rate": 2.1755201182807058e-07, - "loss": 0.4301, + "loss": 0.7904, "step": 93660 }, { "epoch": 9.89, - "grad_norm": 10.147513389587402, + "grad_norm": 8.07014274597168, "learning_rate": 2.1543985637342912e-07, - "loss": 0.2892, + "loss": 0.6387, "step": 93670 }, { "epoch": 9.89, - "grad_norm": 115.73806762695312, + "grad_norm": 68.60924530029297, "learning_rate": 2.1332770091878764e-07, - "loss": 0.4763, + "loss": 0.8124, "step": 93680 }, { "epoch": 9.89, - "grad_norm": 0.19780133664608002, + "grad_norm": 19.25551986694336, "learning_rate": 2.1121554546414618e-07, - "loss": 0.099, + "loss": 0.7977, "step": 93690 }, { "epoch": 9.9, - "grad_norm": 8.174663543701172, + "grad_norm": 28.07084846496582, "learning_rate": 2.0910339000950472e-07, - "loss": 0.2979, + "loss": 0.9248, "step": 93700 }, { "epoch": 9.9, - "grad_norm": 0.9677321314811707, + "grad_norm": 35.347129821777344, "learning_rate": 2.0699123455486323e-07, - "loss": 0.389, + "loss": 0.6258, "step": 93710 }, { "epoch": 9.9, - "grad_norm": 22.479507446289062, + "grad_norm": 38.47208023071289, "learning_rate": 2.048790791002218e-07, - "loss": 0.4523, + "loss": 1.0112, "step": 93720 }, { "epoch": 9.9, - "grad_norm": 18.043590545654297, + "grad_norm": 22.000749588012695, "learning_rate": 2.0276692364558034e-07, - "loss": 0.6633, + "loss": 0.9189, "step": 93730 }, { "epoch": 9.9, - "grad_norm": 9.309534072875977, + "grad_norm": 43.73223876953125, "learning_rate": 2.0065476819093888e-07, - "loss": 0.3865, + "loss": 0.9646, "step": 93740 }, { "epoch": 9.9, - "grad_norm": 32.22866439819336, + "grad_norm": 33.749610900878906, "learning_rate": 1.985426127362974e-07, - "loss": 0.3737, + "loss": 0.7418, "step": 93750 }, { "epoch": 9.9, - "grad_norm": 8.562235832214355, + "grad_norm": 42.74787902832031, "learning_rate": 1.9643045728165593e-07, - "loss": 0.2424, + "loss": 0.8716, "step": 93760 }, { "epoch": 9.9, - "grad_norm": 11.70975399017334, + "grad_norm": 51.16815185546875, "learning_rate": 1.943183018270145e-07, - "loss": 0.2877, + "loss": 0.7987, "step": 93770 }, { "epoch": 9.9, - "grad_norm": 17.607812881469727, + "grad_norm": 36.188785552978516, "learning_rate": 1.9220614637237304e-07, - "loss": 0.3371, + "loss": 0.8389, "step": 93780 }, { "epoch": 9.9, - "grad_norm": 15.377028465270996, + "grad_norm": 58.56796646118164, "learning_rate": 1.9009399091773156e-07, - "loss": 0.2181, + "loss": 0.8799, "step": 93790 }, { "epoch": 9.91, - "grad_norm": 5.3677287101745605, + "grad_norm": 38.72188949584961, "learning_rate": 1.879818354630901e-07, - "loss": 0.4737, + "loss": 0.8235, "step": 93800 }, { "epoch": 9.91, - "grad_norm": 8.13587474822998, + "grad_norm": 49.93680191040039, "learning_rate": 1.8586968000844864e-07, - "loss": 0.3289, + "loss": 0.8359, "step": 93810 }, { "epoch": 9.91, - "grad_norm": 0.9783810973167419, + "grad_norm": 24.66501808166504, "learning_rate": 1.8375752455380715e-07, - "loss": 0.3317, + "loss": 0.7502, "step": 93820 }, { "epoch": 9.91, - "grad_norm": 26.58494758605957, + "grad_norm": 33.77790069580078, "learning_rate": 1.8164536909916572e-07, - "loss": 0.4377, + "loss": 1.0494, "step": 93830 }, { "epoch": 9.91, - "grad_norm": 2.647897958755493, + "grad_norm": 31.591190338134766, "learning_rate": 1.7953321364452426e-07, - "loss": 0.2106, + "loss": 0.8048, "step": 93840 }, { "epoch": 9.91, - "grad_norm": 5.519069194793701, + "grad_norm": 44.446449279785156, "learning_rate": 1.774210581898828e-07, - "loss": 0.1765, + "loss": 0.8951, "step": 93850 }, { "epoch": 9.91, - "grad_norm": 0.05468748137354851, + "grad_norm": 39.22816848754883, "learning_rate": 1.7530890273524131e-07, - "loss": 0.3711, + "loss": 0.7998, "step": 93860 }, { "epoch": 9.91, - "grad_norm": 23.414356231689453, + "grad_norm": 53.307010650634766, "learning_rate": 1.7319674728059985e-07, - "loss": 0.2418, + "loss": 1.1835, "step": 93870 }, { "epoch": 9.91, - "grad_norm": 0.44908764958381653, + "grad_norm": 32.06808090209961, "learning_rate": 1.7108459182595842e-07, - "loss": 0.2621, + "loss": 0.6346, "step": 93880 }, { "epoch": 9.92, - "grad_norm": 1.234279751777649, + "grad_norm": 36.85811996459961, "learning_rate": 1.6897243637131696e-07, - "loss": 0.3918, + "loss": 0.7312, "step": 93890 }, { "epoch": 9.92, - "grad_norm": 21.825580596923828, + "grad_norm": 33.166664123535156, "learning_rate": 1.6686028091667548e-07, - "loss": 0.4281, + "loss": 0.9298, "step": 93900 }, { "epoch": 9.92, - "grad_norm": 25.549957275390625, + "grad_norm": 53.202720642089844, "learning_rate": 1.6474812546203402e-07, - "loss": 0.3235, + "loss": 0.4932, "step": 93910 }, { "epoch": 9.92, - "grad_norm": 1.848997712135315, + "grad_norm": 14.954949378967285, "learning_rate": 1.6263597000739256e-07, - "loss": 0.1981, + "loss": 0.7568, "step": 93920 }, { "epoch": 9.92, - "grad_norm": 3.9947774410247803, + "grad_norm": 44.669281005859375, "learning_rate": 1.6052381455275107e-07, - "loss": 0.3548, + "loss": 0.6495, "step": 93930 }, { "epoch": 9.92, - "grad_norm": 0.8290854692459106, + "grad_norm": 57.876651763916016, "learning_rate": 1.5841165909810964e-07, - "loss": 0.3474, + "loss": 0.7567, "step": 93940 }, { "epoch": 9.92, - "grad_norm": 15.375515937805176, + "grad_norm": 48.616947174072266, "learning_rate": 1.5629950364346818e-07, - "loss": 0.2675, + "loss": 0.7575, "step": 93950 }, { "epoch": 9.92, - "grad_norm": 8.390115737915039, + "grad_norm": 24.337003707885742, "learning_rate": 1.5418734818882672e-07, - "loss": 0.1864, + "loss": 0.7438, "step": 93960 }, { "epoch": 9.92, - "grad_norm": 22.52672576904297, + "grad_norm": 51.54349899291992, "learning_rate": 1.5207519273418523e-07, - "loss": 0.2122, + "loss": 1.0068, "step": 93970 }, { "epoch": 9.93, - "grad_norm": 0.13925258815288544, + "grad_norm": 62.974300384521484, "learning_rate": 1.499630372795438e-07, - "loss": 0.1876, + "loss": 0.7297, "step": 93980 }, { "epoch": 9.93, - "grad_norm": 38.295021057128906, + "grad_norm": 33.31874465942383, "learning_rate": 1.4785088182490232e-07, - "loss": 0.3363, + "loss": 0.9261, "step": 93990 }, { "epoch": 9.93, - "grad_norm": 23.242719650268555, + "grad_norm": 58.434288024902344, "learning_rate": 1.4573872637026086e-07, - "loss": 0.2425, + "loss": 0.6771, "step": 94000 }, { "epoch": 9.93, - "grad_norm": 23.31476402282715, + "grad_norm": 44.8173713684082, "learning_rate": 1.436265709156194e-07, - "loss": 0.3933, + "loss": 0.6495, "step": 94010 }, { "epoch": 9.93, - "grad_norm": 4.398312091827393, + "grad_norm": 39.445655822753906, "learning_rate": 1.4151441546097794e-07, - "loss": 0.3062, + "loss": 0.7841, "step": 94020 }, { "epoch": 9.93, - "grad_norm": 16.463624954223633, + "grad_norm": 52.48044967651367, "learning_rate": 1.3940226000633648e-07, - "loss": 0.4015, + "loss": 0.6176, "step": 94030 }, { "epoch": 9.93, - "grad_norm": 7.495766639709473, + "grad_norm": 34.83245849609375, "learning_rate": 1.3729010455169502e-07, - "loss": 0.2183, + "loss": 0.6576, "step": 94040 }, { "epoch": 9.93, - "grad_norm": 25.269187927246094, + "grad_norm": 36.6810188293457, "learning_rate": 1.3517794909705356e-07, - "loss": 0.222, + "loss": 0.6678, "step": 94050 }, { "epoch": 9.93, - "grad_norm": 23.8918514251709, + "grad_norm": 59.48587417602539, "learning_rate": 1.3306579364241207e-07, - "loss": 0.1309, + "loss": 0.835, "step": 94060 }, { "epoch": 9.93, - "grad_norm": 0.034589733928442, + "grad_norm": 61.653480529785156, "learning_rate": 1.3095363818777064e-07, - "loss": 0.2698, + "loss": 1.0785, "step": 94070 }, { "epoch": 9.94, - "grad_norm": 4.655220985412598, + "grad_norm": 38.754554748535156, "learning_rate": 1.2884148273312915e-07, - "loss": 0.2323, + "loss": 0.5932, "step": 94080 }, { "epoch": 9.94, - "grad_norm": 0.25697973370552063, + "grad_norm": 41.941131591796875, "learning_rate": 1.2672932727848772e-07, - "loss": 0.3815, + "loss": 0.7118, "step": 94090 }, { "epoch": 9.94, - "grad_norm": 3.5991857051849365, + "grad_norm": 66.41291046142578, "learning_rate": 1.2461717182384624e-07, - "loss": 0.3107, + "loss": 0.6699, "step": 94100 }, { "epoch": 9.94, - "grad_norm": 3.142202615737915, + "grad_norm": 28.7824764251709, "learning_rate": 1.2250501636920478e-07, - "loss": 0.2316, + "loss": 0.7598, "step": 94110 }, { "epoch": 9.94, - "grad_norm": 26.116809844970703, + "grad_norm": 29.992919921875, "learning_rate": 1.2039286091456332e-07, - "loss": 0.4277, + "loss": 0.598, "step": 94120 }, { "epoch": 9.94, - "grad_norm": 0.06505630165338516, + "grad_norm": 29.20918083190918, "learning_rate": 1.1828070545992186e-07, - "loss": 0.3711, + "loss": 0.806, "step": 94130 }, { "epoch": 9.94, - "grad_norm": 18.87607765197754, + "grad_norm": 48.73361587524414, "learning_rate": 1.161685500052804e-07, - "loss": 0.4948, + "loss": 1.0094, "step": 94140 }, { "epoch": 9.94, - "grad_norm": 27.82979965209961, + "grad_norm": 74.5548324584961, "learning_rate": 1.1405639455063894e-07, - "loss": 0.273, + "loss": 0.7897, "step": 94150 }, { "epoch": 9.94, - "grad_norm": 0.06304705888032913, + "grad_norm": 23.728843688964844, "learning_rate": 1.1194423909599748e-07, - "loss": 0.2777, + "loss": 0.9866, "step": 94160 }, { "epoch": 9.95, - "grad_norm": 0.2837618589401245, + "grad_norm": 50.069705963134766, "learning_rate": 1.0983208364135601e-07, - "loss": 0.2463, + "loss": 0.6561, "step": 94170 }, { "epoch": 9.95, - "grad_norm": 12.603565216064453, + "grad_norm": 56.23488235473633, "learning_rate": 1.0771992818671456e-07, - "loss": 0.3531, + "loss": 0.6842, "step": 94180 }, { "epoch": 9.95, - "grad_norm": 4.313179016113281, + "grad_norm": 16.933359146118164, "learning_rate": 1.0560777273207309e-07, - "loss": 0.111, + "loss": 0.658, "step": 94190 }, { "epoch": 9.95, - "grad_norm": 40.60833740234375, + "grad_norm": 26.667892456054688, "learning_rate": 1.0349561727743162e-07, - "loss": 0.4447, + "loss": 0.9617, "step": 94200 }, { "epoch": 9.95, - "grad_norm": 9.594927787780762, + "grad_norm": 33.110984802246094, "learning_rate": 1.0138346182279017e-07, - "loss": 0.4402, + "loss": 0.896, "step": 94210 }, { "epoch": 9.95, - "grad_norm": 3.9277567863464355, + "grad_norm": 42.093162536621094, "learning_rate": 9.92713063681487e-08, - "loss": 0.2934, + "loss": 0.7113, "step": 94220 }, { "epoch": 9.95, - "grad_norm": 12.845545768737793, + "grad_norm": 29.182172775268555, "learning_rate": 9.715915091350725e-08, - "loss": 0.342, + "loss": 0.9301, "step": 94230 }, { "epoch": 9.95, - "grad_norm": 20.24753189086914, + "grad_norm": 41.36017608642578, "learning_rate": 9.504699545886578e-08, - "loss": 0.2015, + "loss": 0.7459, "step": 94240 }, { "epoch": 9.95, - "grad_norm": 9.363421440124512, + "grad_norm": 27.86365509033203, "learning_rate": 9.293484000422432e-08, - "loss": 0.3418, + "loss": 0.917, "step": 94250 }, { "epoch": 9.95, - "grad_norm": 15.755599975585938, + "grad_norm": 55.381507873535156, "learning_rate": 9.082268454958286e-08, - "loss": 0.4105, + "loss": 0.7827, "step": 94260 }, { "epoch": 9.96, - "grad_norm": 23.8909912109375, + "grad_norm": 66.65986633300781, "learning_rate": 8.87105290949414e-08, - "loss": 0.2631, + "loss": 0.8179, "step": 94270 }, { "epoch": 9.96, - "grad_norm": 0.014904187992215157, + "grad_norm": 24.459991455078125, "learning_rate": 8.659837364029993e-08, - "loss": 0.2216, + "loss": 0.7978, "step": 94280 }, { "epoch": 9.96, - "grad_norm": 6.4723100662231445, + "grad_norm": 37.13322067260742, "learning_rate": 8.448621818565848e-08, - "loss": 0.3547, + "loss": 0.5952, "step": 94290 }, { "epoch": 9.96, - "grad_norm": 13.555961608886719, + "grad_norm": 14.82149887084961, "learning_rate": 8.237406273101701e-08, - "loss": 0.4443, + "loss": 0.5721, "step": 94300 }, { "epoch": 9.96, - "grad_norm": 6.17055082321167, + "grad_norm": 17.552268981933594, "learning_rate": 8.026190727637554e-08, - "loss": 0.1998, + "loss": 0.6626, "step": 94310 }, { "epoch": 9.96, - "grad_norm": 0.14323875308036804, + "grad_norm": 29.518573760986328, "learning_rate": 7.814975182173409e-08, - "loss": 0.3288, + "loss": 0.6772, "step": 94320 }, { "epoch": 9.96, - "grad_norm": 7.913241863250732, + "grad_norm": 41.42709732055664, "learning_rate": 7.603759636709262e-08, - "loss": 0.2973, + "loss": 1.0002, "step": 94330 }, { "epoch": 9.96, - "grad_norm": 36.910484313964844, + "grad_norm": 49.39017868041992, "learning_rate": 7.392544091245116e-08, - "loss": 0.3286, + "loss": 0.75, "step": 94340 }, { "epoch": 9.96, - "grad_norm": 0.30807071924209595, + "grad_norm": 37.56350326538086, "learning_rate": 7.18132854578097e-08, - "loss": 0.1102, + "loss": 0.8415, "step": 94350 }, { "epoch": 9.97, - "grad_norm": 0.12505747377872467, + "grad_norm": 10.552456855773926, "learning_rate": 6.970113000316824e-08, - "loss": 0.2963, + "loss": 0.7517, "step": 94360 }, { "epoch": 9.97, - "grad_norm": 28.966279983520508, + "grad_norm": 35.80953598022461, "learning_rate": 6.758897454852678e-08, - "loss": 0.236, + "loss": 0.553, "step": 94370 }, { "epoch": 9.97, - "grad_norm": 4.6660237312316895, + "grad_norm": 27.630142211914062, "learning_rate": 6.547681909388532e-08, - "loss": 0.1451, + "loss": 0.5354, "step": 94380 }, { "epoch": 9.97, - "grad_norm": 29.031957626342773, + "grad_norm": 37.87149429321289, "learning_rate": 6.336466363924386e-08, - "loss": 0.2402, + "loss": 0.9188, "step": 94390 }, { "epoch": 9.97, - "grad_norm": 0.4105350375175476, + "grad_norm": 25.735733032226562, "learning_rate": 6.125250818460239e-08, - "loss": 0.6955, + "loss": 0.9524, "step": 94400 }, { "epoch": 9.97, - "grad_norm": 1.6894460916519165, + "grad_norm": 45.774627685546875, "learning_rate": 5.914035272996093e-08, - "loss": 0.3104, + "loss": 1.0664, "step": 94410 }, { "epoch": 9.97, - "grad_norm": 21.700204849243164, + "grad_norm": 47.61908721923828, "learning_rate": 5.702819727531947e-08, - "loss": 0.3762, + "loss": 0.9123, "step": 94420 }, { "epoch": 9.97, - "grad_norm": 7.254398822784424, + "grad_norm": 38.00491714477539, "learning_rate": 5.4916041820678004e-08, - "loss": 0.3415, + "loss": 0.5889, "step": 94430 }, { "epoch": 9.97, - "grad_norm": 4.643754005432129, + "grad_norm": 57.72441864013672, "learning_rate": 5.2803886366036544e-08, - "loss": 0.1333, + "loss": 0.7415, "step": 94440 }, { "epoch": 9.97, - "grad_norm": 13.833614349365234, + "grad_norm": 52.10158157348633, "learning_rate": 5.0691730911395085e-08, - "loss": 0.2566, + "loss": 0.9054, "step": 94450 }, { "epoch": 9.98, - "grad_norm": 1.4318572282791138, + "grad_norm": 36.97465133666992, "learning_rate": 4.8579575456753625e-08, - "loss": 0.3789, + "loss": 1.0207, "step": 94460 }, { "epoch": 9.98, - "grad_norm": 0.7431033253669739, + "grad_norm": 43.235050201416016, "learning_rate": 4.646742000211216e-08, - "loss": 0.4111, + "loss": 1.001, "step": 94470 }, { "epoch": 9.98, - "grad_norm": 2.1446645259857178, + "grad_norm": 39.54731750488281, "learning_rate": 4.43552645474707e-08, - "loss": 0.2587, + "loss": 1.0034, "step": 94480 }, { "epoch": 9.98, - "grad_norm": 2.56929087638855, + "grad_norm": 43.35914993286133, "learning_rate": 4.224310909282924e-08, - "loss": 0.1811, + "loss": 1.0778, "step": 94490 }, { "epoch": 9.98, - "grad_norm": 3.7090225219726562, + "grad_norm": 43.679561614990234, "learning_rate": 4.013095363818777e-08, - "loss": 0.3823, + "loss": 0.738, "step": 94500 }, { "epoch": 9.98, - "grad_norm": 19.517879486083984, + "grad_norm": 19.321983337402344, "learning_rate": 3.801879818354631e-08, - "loss": 0.632, + "loss": 0.8381, "step": 94510 }, { "epoch": 9.98, - "grad_norm": 16.57964515686035, + "grad_norm": 54.165977478027344, "learning_rate": 3.590664272890485e-08, - "loss": 0.2772, + "loss": 0.7901, "step": 94520 }, { "epoch": 9.98, - "grad_norm": 10.921294212341309, + "grad_norm": 54.50868606567383, "learning_rate": 3.379448727426339e-08, - "loss": 0.0742, + "loss": 0.8594, "step": 94530 }, { "epoch": 9.98, - "grad_norm": 0.09099742770195007, + "grad_norm": 43.779972076416016, "learning_rate": 3.168233181962193e-08, - "loss": 0.3104, + "loss": 0.7942, "step": 94540 }, { "epoch": 9.99, - "grad_norm": 13.063875198364258, + "grad_norm": 28.19612693786621, "learning_rate": 2.9570176364980464e-08, - "loss": 0.3505, + "loss": 0.7349, "step": 94550 }, { "epoch": 9.99, - "grad_norm": 18.445262908935547, + "grad_norm": 61.22126388549805, "learning_rate": 2.7458020910339002e-08, - "loss": 0.4989, + "loss": 0.7764, "step": 94560 }, { "epoch": 9.99, - "grad_norm": 0.4335992932319641, + "grad_norm": 84.55147552490234, "learning_rate": 2.5345865455697542e-08, - "loss": 0.1697, + "loss": 0.9886, "step": 94570 }, { "epoch": 9.99, - "grad_norm": 0.14158082008361816, + "grad_norm": 52.04364013671875, "learning_rate": 2.323371000105608e-08, - "loss": 0.2024, + "loss": 0.8929, "step": 94580 }, { "epoch": 9.99, - "grad_norm": 0.5155888199806213, + "grad_norm": 30.470081329345703, "learning_rate": 2.112155454641462e-08, - "loss": 0.2067, + "loss": 0.9814, "step": 94590 }, { "epoch": 9.99, - "grad_norm": 6.225322246551514, + "grad_norm": 29.325666427612305, "learning_rate": 1.9009399091773154e-08, - "loss": 0.3917, + "loss": 0.9854, "step": 94600 }, { "epoch": 9.99, - "grad_norm": 18.707229614257812, + "grad_norm": 36.0351676940918, "learning_rate": 1.6897243637131695e-08, - "loss": 0.2365, + "loss": 0.7083, "step": 94610 }, { "epoch": 9.99, - "grad_norm": 11.871989250183105, + "grad_norm": 34.118099212646484, "learning_rate": 1.4785088182490232e-08, - "loss": 0.2894, + "loss": 0.7237, "step": 94620 }, { "epoch": 9.99, - "grad_norm": 20.030393600463867, + "grad_norm": 28.25765609741211, "learning_rate": 1.2672932727848771e-08, - "loss": 0.2456, + "loss": 0.9192, "step": 94630 }, { "epoch": 9.99, - "grad_norm": 16.998205184936523, + "grad_norm": 63.757408142089844, "learning_rate": 1.056077727320731e-08, - "loss": 0.2924, + "loss": 0.659, "step": 94640 }, { "epoch": 10.0, - "grad_norm": 24.512149810791016, + "grad_norm": 46.26067352294922, "learning_rate": 8.448621818565847e-09, - "loss": 0.4706, + "loss": 0.7402, "step": 94650 }, { "epoch": 10.0, - "grad_norm": 0.03656615689396858, + "grad_norm": 29.647729873657227, "learning_rate": 6.336466363924386e-09, - "loss": 0.2152, + "loss": 0.8519, "step": 94660 }, { "epoch": 10.0, - "grad_norm": 29.174745559692383, + "grad_norm": 83.67559051513672, "learning_rate": 4.224310909282924e-09, - "loss": 0.2361, + "loss": 0.9337, "step": 94670 }, { "epoch": 10.0, - "grad_norm": 0.04462029039859772, + "grad_norm": 46.47532653808594, "learning_rate": 2.112155454641462e-09, - "loss": 0.2021, + "loss": 1.2293, "step": 94680 }, { "epoch": 10.0, - "grad_norm": 0.012345206923782825, + "grad_norm": 4.861952304840088, "learning_rate": 0.0, - "loss": 0.2699, + "loss": 0.7437, "step": 94690 }, { "epoch": 10.0, - "eval_accuracy": 0.9072079207920792, - "eval_loss": 0.38942983746528625, - "eval_runtime": 228.6878, - "eval_samples_per_second": 110.413, - "eval_steps_per_second": 13.805, + "eval_accuracy": 0.8204356435643564, + "eval_loss": 0.6538042426109314, + "eval_runtime": 37.9577, + "eval_samples_per_second": 665.214, + "eval_steps_per_second": 83.172, "step": 94690 }, { "epoch": 10.0, "step": 94690, - "total_flos": 5.875226713843201e+19, - "train_loss": 0.6563277260213383, - "train_runtime": 13063.3743, - "train_samples_per_second": 57.987, - "train_steps_per_second": 7.249 + "total_flos": 3.79286401181184e+18, + "train_loss": 1.2358400564811112, + "train_runtime": 3207.066, + "train_samples_per_second": 236.197, + "train_steps_per_second": 29.525 } ], "logging_steps": 10, @@ -66396,7 +66396,7 @@ "num_input_tokens_seen": 0, "num_train_epochs": 10, "save_steps": 500, - "total_flos": 5.875226713843201e+19, + "total_flos": 3.79286401181184e+18, "train_batch_size": 8, "trial_name": null, "trial_params": null