{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.4999945685825087, "eval_steps": 500, "global_step": 11507, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00043451339930695114, "grad_norm": 1.8335795402526855, "learning_rate": 5.780346820809249e-07, "loss": 0.1919, "step": 10 }, { "epoch": 0.0008690267986139023, "grad_norm": 1.1079758405685425, "learning_rate": 1.1560693641618499e-06, "loss": 0.1616, "step": 20 }, { "epoch": 0.0013035401979208535, "grad_norm": 0.7856003046035767, "learning_rate": 1.7341040462427746e-06, "loss": 0.1325, "step": 30 }, { "epoch": 0.0017380535972278046, "grad_norm": 0.47701773047447205, "learning_rate": 2.3121387283236997e-06, "loss": 0.0861, "step": 40 }, { "epoch": 0.0021725669965347557, "grad_norm": 0.4777753949165344, "learning_rate": 2.890173410404625e-06, "loss": 0.0734, "step": 50 }, { "epoch": 0.002607080395841707, "grad_norm": 0.44196391105651855, "learning_rate": 3.468208092485549e-06, "loss": 0.0565, "step": 60 }, { "epoch": 0.003041593795148658, "grad_norm": 0.3910512924194336, "learning_rate": 4.046242774566474e-06, "loss": 0.0524, "step": 70 }, { "epoch": 0.003476107194455609, "grad_norm": 0.43339353799819946, "learning_rate": 4.6242774566473994e-06, "loss": 0.0534, "step": 80 }, { "epoch": 0.00391062059376256, "grad_norm": 0.44343557953834534, "learning_rate": 5.202312138728324e-06, "loss": 0.0432, "step": 90 }, { "epoch": 0.004345133993069511, "grad_norm": 0.37480905652046204, "learning_rate": 5.78034682080925e-06, "loss": 0.0476, "step": 100 }, { "epoch": 0.004779647392376463, "grad_norm": 0.32702726125717163, "learning_rate": 6.358381502890174e-06, "loss": 0.0326, "step": 110 }, { "epoch": 0.005214160791683414, "grad_norm": 0.40295419096946716, "learning_rate": 6.936416184971098e-06, "loss": 0.0374, "step": 120 }, { "epoch": 0.005648674190990364, "grad_norm": 0.3500003516674042, "learning_rate": 7.5144508670520235e-06, "loss": 0.031, "step": 130 }, { "epoch": 0.006083187590297316, "grad_norm": 0.38030797243118286, "learning_rate": 8.092485549132949e-06, "loss": 0.0368, "step": 140 }, { "epoch": 0.006517700989604267, "grad_norm": 0.3058304190635681, "learning_rate": 8.670520231213873e-06, "loss": 0.0342, "step": 150 }, { "epoch": 0.006952214388911218, "grad_norm": 0.3178086280822754, "learning_rate": 9.248554913294799e-06, "loss": 0.0328, "step": 160 }, { "epoch": 0.00738672778821817, "grad_norm": 0.416713148355484, "learning_rate": 9.826589595375723e-06, "loss": 0.037, "step": 170 }, { "epoch": 0.00782124118752512, "grad_norm": 0.24077042937278748, "learning_rate": 1.0404624277456647e-05, "loss": 0.0313, "step": 180 }, { "epoch": 0.008255754586832071, "grad_norm": 0.27852290868759155, "learning_rate": 1.0982658959537573e-05, "loss": 0.0266, "step": 190 }, { "epoch": 0.008690267986139023, "grad_norm": 0.24909931421279907, "learning_rate": 1.15606936416185e-05, "loss": 0.0365, "step": 200 }, { "epoch": 0.009124781385445974, "grad_norm": 0.24574002623558044, "learning_rate": 1.2138728323699422e-05, "loss": 0.03, "step": 210 }, { "epoch": 0.009559294784752925, "grad_norm": 0.2927786111831665, "learning_rate": 1.2716763005780348e-05, "loss": 0.0286, "step": 220 }, { "epoch": 0.009993808184059877, "grad_norm": 0.31231585144996643, "learning_rate": 1.3294797687861272e-05, "loss": 0.0305, "step": 230 }, { "epoch": 0.010428321583366828, "grad_norm": 0.3059456944465637, "learning_rate": 1.3872832369942197e-05, "loss": 0.0313, "step": 240 }, { "epoch": 0.010862834982673777, "grad_norm": 0.23252910375595093, "learning_rate": 1.4450867052023123e-05, "loss": 0.0269, "step": 250 }, { "epoch": 0.011297348381980729, "grad_norm": 0.430499792098999, "learning_rate": 1.5028901734104047e-05, "loss": 0.0339, "step": 260 }, { "epoch": 0.01173186178128768, "grad_norm": 0.21920466423034668, "learning_rate": 1.5606936416184973e-05, "loss": 0.0233, "step": 270 }, { "epoch": 0.012166375180594631, "grad_norm": 0.2036123424768448, "learning_rate": 1.6184971098265897e-05, "loss": 0.0252, "step": 280 }, { "epoch": 0.012600888579901583, "grad_norm": 0.3722895681858063, "learning_rate": 1.676300578034682e-05, "loss": 0.027, "step": 290 }, { "epoch": 0.013035401979208534, "grad_norm": 0.33059456944465637, "learning_rate": 1.7341040462427746e-05, "loss": 0.0311, "step": 300 }, { "epoch": 0.013469915378515485, "grad_norm": 0.2505900263786316, "learning_rate": 1.7919075144508673e-05, "loss": 0.022, "step": 310 }, { "epoch": 0.013904428777822437, "grad_norm": 0.2660132348537445, "learning_rate": 1.8497109826589598e-05, "loss": 0.0282, "step": 320 }, { "epoch": 0.014338942177129388, "grad_norm": 0.22333598136901855, "learning_rate": 1.9075144508670522e-05, "loss": 0.0245, "step": 330 }, { "epoch": 0.01477345557643634, "grad_norm": 0.2619096338748932, "learning_rate": 1.9653179190751446e-05, "loss": 0.0235, "step": 340 }, { "epoch": 0.015207968975743289, "grad_norm": 0.26429101824760437, "learning_rate": 1.9999993661544142e-05, "loss": 0.0221, "step": 350 }, { "epoch": 0.01564248237505024, "grad_norm": 0.16357919573783875, "learning_rate": 1.999992235400801e-05, "loss": 0.0232, "step": 360 }, { "epoch": 0.016076995774357193, "grad_norm": 0.181153804063797, "learning_rate": 1.9999771816432783e-05, "loss": 0.0259, "step": 370 }, { "epoch": 0.016511509173664143, "grad_norm": 0.2861078679561615, "learning_rate": 1.9999542050011175e-05, "loss": 0.0215, "step": 380 }, { "epoch": 0.016946022572971092, "grad_norm": 0.19298051297664642, "learning_rate": 1.999923305656364e-05, "loss": 0.0205, "step": 390 }, { "epoch": 0.017380535972278045, "grad_norm": 0.16827619075775146, "learning_rate": 1.999884483853836e-05, "loss": 0.0251, "step": 400 }, { "epoch": 0.017815049371584995, "grad_norm": 0.16084188222885132, "learning_rate": 1.9998377399011203e-05, "loss": 0.0253, "step": 410 }, { "epoch": 0.018249562770891948, "grad_norm": 0.2285071611404419, "learning_rate": 1.9997830741685734e-05, "loss": 0.0233, "step": 420 }, { "epoch": 0.018684076170198897, "grad_norm": 0.1760404258966446, "learning_rate": 1.9997204870893147e-05, "loss": 0.0255, "step": 430 }, { "epoch": 0.01911858956950585, "grad_norm": 0.17383626103401184, "learning_rate": 1.9996499791592257e-05, "loss": 0.0224, "step": 440 }, { "epoch": 0.0195531029688128, "grad_norm": 0.182758167386055, "learning_rate": 1.9995715509369456e-05, "loss": 0.0242, "step": 450 }, { "epoch": 0.019987616368119753, "grad_norm": 0.22543829679489136, "learning_rate": 1.999485203043866e-05, "loss": 0.0204, "step": 460 }, { "epoch": 0.020422129767426703, "grad_norm": 0.19388648867607117, "learning_rate": 1.9993909361641272e-05, "loss": 0.0221, "step": 470 }, { "epoch": 0.020856643166733656, "grad_norm": 0.2187209576368332, "learning_rate": 1.999288751044612e-05, "loss": 0.0214, "step": 480 }, { "epoch": 0.021291156566040605, "grad_norm": 0.27493494749069214, "learning_rate": 1.9991786484949397e-05, "loss": 0.0247, "step": 490 }, { "epoch": 0.021725669965347555, "grad_norm": 0.17861245572566986, "learning_rate": 1.99906062938746e-05, "loss": 0.026, "step": 500 }, { "epoch": 0.022160183364654508, "grad_norm": 0.16928374767303467, "learning_rate": 1.9989346946572455e-05, "loss": 0.0244, "step": 510 }, { "epoch": 0.022594696763961458, "grad_norm": 0.5377978086471558, "learning_rate": 1.998800845302086e-05, "loss": 0.0226, "step": 520 }, { "epoch": 0.02302921016326841, "grad_norm": 0.15883971750736237, "learning_rate": 1.9986590823824785e-05, "loss": 0.0216, "step": 530 }, { "epoch": 0.02346372356257536, "grad_norm": 0.16129736602306366, "learning_rate": 1.99850940702162e-05, "loss": 0.0194, "step": 540 }, { "epoch": 0.023898236961882313, "grad_norm": 0.1318095475435257, "learning_rate": 1.9983518204053976e-05, "loss": 0.0179, "step": 550 }, { "epoch": 0.024332750361189263, "grad_norm": 0.13563190400600433, "learning_rate": 1.9981863237823807e-05, "loss": 0.0192, "step": 560 }, { "epoch": 0.024767263760496216, "grad_norm": 0.19693246483802795, "learning_rate": 1.9980129184638103e-05, "loss": 0.0192, "step": 570 }, { "epoch": 0.025201777159803165, "grad_norm": 0.14829251170158386, "learning_rate": 1.9978316058235875e-05, "loss": 0.0207, "step": 580 }, { "epoch": 0.025636290559110115, "grad_norm": 0.19513817131519318, "learning_rate": 1.9976423872982646e-05, "loss": 0.0167, "step": 590 }, { "epoch": 0.026070803958417068, "grad_norm": 0.20041579008102417, "learning_rate": 1.9974452643870318e-05, "loss": 0.0236, "step": 600 }, { "epoch": 0.026505317357724018, "grad_norm": 0.1073588877916336, "learning_rate": 1.9972402386517076e-05, "loss": 0.0136, "step": 610 }, { "epoch": 0.02693983075703097, "grad_norm": 0.1381412297487259, "learning_rate": 1.997027311716724e-05, "loss": 0.0177, "step": 620 }, { "epoch": 0.02737434415633792, "grad_norm": 0.1919044554233551, "learning_rate": 1.9968064852691145e-05, "loss": 0.0173, "step": 630 }, { "epoch": 0.027808857555644873, "grad_norm": 0.14668576419353485, "learning_rate": 1.9965777610585023e-05, "loss": 0.0214, "step": 640 }, { "epoch": 0.028243370954951823, "grad_norm": 0.13156649470329285, "learning_rate": 1.9963411408970837e-05, "loss": 0.0194, "step": 650 }, { "epoch": 0.028677884354258776, "grad_norm": 0.17079275846481323, "learning_rate": 1.9960966266596164e-05, "loss": 0.0169, "step": 660 }, { "epoch": 0.029112397753565725, "grad_norm": 0.09269203990697861, "learning_rate": 1.995844220283402e-05, "loss": 0.019, "step": 670 }, { "epoch": 0.02954691115287268, "grad_norm": 0.21620836853981018, "learning_rate": 1.995583923768273e-05, "loss": 0.0192, "step": 680 }, { "epoch": 0.029981424552179628, "grad_norm": 0.15077561140060425, "learning_rate": 1.995315739176576e-05, "loss": 0.0216, "step": 690 }, { "epoch": 0.030415937951486578, "grad_norm": 0.21490907669067383, "learning_rate": 1.9950396686331543e-05, "loss": 0.0182, "step": 700 }, { "epoch": 0.03085045135079353, "grad_norm": 0.1409023404121399, "learning_rate": 1.9947557143253337e-05, "loss": 0.0174, "step": 710 }, { "epoch": 0.03128496475010048, "grad_norm": 0.11508463323116302, "learning_rate": 1.9944638785029024e-05, "loss": 0.0155, "step": 720 }, { "epoch": 0.03171947814940743, "grad_norm": 0.10164955258369446, "learning_rate": 1.9941641634780942e-05, "loss": 0.0148, "step": 730 }, { "epoch": 0.032153991548714386, "grad_norm": 0.18500874936580658, "learning_rate": 1.993856571625572e-05, "loss": 0.0157, "step": 740 }, { "epoch": 0.03258850494802133, "grad_norm": 0.3012203872203827, "learning_rate": 1.9935411053824057e-05, "loss": 0.0161, "step": 750 }, { "epoch": 0.033023018347328285, "grad_norm": 0.1304362714290619, "learning_rate": 1.9932177672480544e-05, "loss": 0.0177, "step": 760 }, { "epoch": 0.03345753174663524, "grad_norm": 0.20605377852916718, "learning_rate": 1.992886559784348e-05, "loss": 0.0188, "step": 770 }, { "epoch": 0.033892045145942185, "grad_norm": 0.14996512234210968, "learning_rate": 1.9925474856154645e-05, "loss": 0.014, "step": 780 }, { "epoch": 0.03432655854524914, "grad_norm": 0.2049788534641266, "learning_rate": 1.992200547427911e-05, "loss": 0.0206, "step": 790 }, { "epoch": 0.03476107194455609, "grad_norm": 0.16461384296417236, "learning_rate": 1.9918457479705012e-05, "loss": 0.0153, "step": 800 }, { "epoch": 0.035195585343863044, "grad_norm": 0.13572026789188385, "learning_rate": 1.9914830900543346e-05, "loss": 0.02, "step": 810 }, { "epoch": 0.03563009874316999, "grad_norm": 0.1633780300617218, "learning_rate": 1.991112576552774e-05, "loss": 0.0203, "step": 820 }, { "epoch": 0.03606461214247694, "grad_norm": 0.14304646849632263, "learning_rate": 1.9907342104014213e-05, "loss": 0.0166, "step": 830 }, { "epoch": 0.036499125541783896, "grad_norm": 0.09156439453363419, "learning_rate": 1.9903479945980964e-05, "loss": 0.0144, "step": 840 }, { "epoch": 0.03693363894109085, "grad_norm": 0.103766530752182, "learning_rate": 1.9899539322028128e-05, "loss": 0.0169, "step": 850 }, { "epoch": 0.037368152340397795, "grad_norm": 0.18641023337841034, "learning_rate": 1.9895520263377523e-05, "loss": 0.0201, "step": 860 }, { "epoch": 0.03780266573970475, "grad_norm": 0.156049445271492, "learning_rate": 1.989142280187242e-05, "loss": 0.0191, "step": 870 }, { "epoch": 0.0382371791390117, "grad_norm": 0.19983117282390594, "learning_rate": 1.9887246969977266e-05, "loss": 0.0145, "step": 880 }, { "epoch": 0.03867169253831865, "grad_norm": 0.14561262726783752, "learning_rate": 1.9882992800777458e-05, "loss": 0.0132, "step": 890 }, { "epoch": 0.0391062059376256, "grad_norm": 0.13224513828754425, "learning_rate": 1.987866032797905e-05, "loss": 0.0144, "step": 900 }, { "epoch": 0.03954071933693255, "grad_norm": 0.1451365202665329, "learning_rate": 1.987424958590851e-05, "loss": 0.0156, "step": 910 }, { "epoch": 0.039975232736239506, "grad_norm": 0.18520617485046387, "learning_rate": 1.9869760609512435e-05, "loss": 0.0159, "step": 920 }, { "epoch": 0.04040974613554645, "grad_norm": 0.15158046782016754, "learning_rate": 1.9865193434357284e-05, "loss": 0.0169, "step": 930 }, { "epoch": 0.040844259534853405, "grad_norm": 0.18894895911216736, "learning_rate": 1.9860548096629084e-05, "loss": 0.015, "step": 940 }, { "epoch": 0.04127877293416036, "grad_norm": 0.15075677633285522, "learning_rate": 1.985582463313314e-05, "loss": 0.0186, "step": 950 }, { "epoch": 0.04171328633346731, "grad_norm": 0.14799635112285614, "learning_rate": 1.985102308129377e-05, "loss": 0.0172, "step": 960 }, { "epoch": 0.04214779973277426, "grad_norm": 0.14480791985988617, "learning_rate": 1.9846143479153986e-05, "loss": 0.0181, "step": 970 }, { "epoch": 0.04258231313208121, "grad_norm": 0.15664492547512054, "learning_rate": 1.9841185865375186e-05, "loss": 0.0206, "step": 980 }, { "epoch": 0.043016826531388164, "grad_norm": 0.1123814582824707, "learning_rate": 1.9836150279236878e-05, "loss": 0.0146, "step": 990 }, { "epoch": 0.04345133993069511, "grad_norm": 0.16255418956279755, "learning_rate": 1.9831036760636334e-05, "loss": 0.0176, "step": 1000 }, { "epoch": 0.04388585333000206, "grad_norm": 0.1457909196615219, "learning_rate": 1.98258453500883e-05, "loss": 0.0183, "step": 1010 }, { "epoch": 0.044320366729309016, "grad_norm": 0.15886881947517395, "learning_rate": 1.982057608872466e-05, "loss": 0.0221, "step": 1020 }, { "epoch": 0.04475488012861597, "grad_norm": 0.1742968112230301, "learning_rate": 1.981522901829411e-05, "loss": 0.0181, "step": 1030 }, { "epoch": 0.045189393527922915, "grad_norm": 0.12152547389268875, "learning_rate": 1.9809804181161857e-05, "loss": 0.019, "step": 1040 }, { "epoch": 0.04562390692722987, "grad_norm": 0.16744455695152283, "learning_rate": 1.9804301620309223e-05, "loss": 0.0175, "step": 1050 }, { "epoch": 0.04605842032653682, "grad_norm": 0.21038441359996796, "learning_rate": 1.9798721379333363e-05, "loss": 0.0171, "step": 1060 }, { "epoch": 0.046492933725843774, "grad_norm": 0.17659246921539307, "learning_rate": 1.9793063502446894e-05, "loss": 0.0149, "step": 1070 }, { "epoch": 0.04692744712515072, "grad_norm": 0.10321566462516785, "learning_rate": 1.978732803447754e-05, "loss": 0.0161, "step": 1080 }, { "epoch": 0.04736196052445767, "grad_norm": 0.16017237305641174, "learning_rate": 1.9781515020867793e-05, "loss": 0.0161, "step": 1090 }, { "epoch": 0.047796473923764626, "grad_norm": 0.13999983668327332, "learning_rate": 1.9775624507674543e-05, "loss": 0.0126, "step": 1100 }, { "epoch": 0.04823098732307157, "grad_norm": 0.08775936812162399, "learning_rate": 1.9769656541568703e-05, "loss": 0.0125, "step": 1110 }, { "epoch": 0.048665500722378525, "grad_norm": 0.11605579406023026, "learning_rate": 1.9763611169834865e-05, "loss": 0.0161, "step": 1120 }, { "epoch": 0.04910001412168548, "grad_norm": 0.16576404869556427, "learning_rate": 1.9757488440370904e-05, "loss": 0.0137, "step": 1130 }, { "epoch": 0.04953452752099243, "grad_norm": 0.11074787378311157, "learning_rate": 1.9751288401687603e-05, "loss": 0.0124, "step": 1140 }, { "epoch": 0.04996904092029938, "grad_norm": 0.11228064447641373, "learning_rate": 1.9745011102908277e-05, "loss": 0.0179, "step": 1150 }, { "epoch": 0.05040355431960633, "grad_norm": 0.13807277381420135, "learning_rate": 1.9738656593768372e-05, "loss": 0.0142, "step": 1160 }, { "epoch": 0.050838067718913284, "grad_norm": 0.13158458471298218, "learning_rate": 1.9732224924615083e-05, "loss": 0.0165, "step": 1170 }, { "epoch": 0.05127258111822023, "grad_norm": 0.13589569926261902, "learning_rate": 1.9725716146406948e-05, "loss": 0.0196, "step": 1180 }, { "epoch": 0.05170709451752718, "grad_norm": 0.08806575834751129, "learning_rate": 1.9719130310713438e-05, "loss": 0.0123, "step": 1190 }, { "epoch": 0.052141607916834136, "grad_norm": 0.1415325552225113, "learning_rate": 1.971246746971456e-05, "loss": 0.0177, "step": 1200 }, { "epoch": 0.05257612131614109, "grad_norm": 0.10823646932840347, "learning_rate": 1.9705727676200443e-05, "loss": 0.0125, "step": 1210 }, { "epoch": 0.053010634715448035, "grad_norm": 0.15576744079589844, "learning_rate": 1.9698910983570907e-05, "loss": 0.0167, "step": 1220 }, { "epoch": 0.05344514811475499, "grad_norm": 0.19083163142204285, "learning_rate": 1.9692017445835057e-05, "loss": 0.016, "step": 1230 }, { "epoch": 0.05387966151406194, "grad_norm": 0.17201757431030273, "learning_rate": 1.968504711761084e-05, "loss": 0.0199, "step": 1240 }, { "epoch": 0.054314174913368894, "grad_norm": 0.129535973072052, "learning_rate": 1.9678000054124626e-05, "loss": 0.0175, "step": 1250 }, { "epoch": 0.05474868831267584, "grad_norm": 0.1300555020570755, "learning_rate": 1.9670876311210763e-05, "loss": 0.0159, "step": 1260 }, { "epoch": 0.05518320171198279, "grad_norm": 0.17606855928897858, "learning_rate": 1.9663675945311125e-05, "loss": 0.0163, "step": 1270 }, { "epoch": 0.055617715111289746, "grad_norm": 0.12871307134628296, "learning_rate": 1.9656399013474686e-05, "loss": 0.0153, "step": 1280 }, { "epoch": 0.05605222851059669, "grad_norm": 0.12462568283081055, "learning_rate": 1.9649045573357053e-05, "loss": 0.0155, "step": 1290 }, { "epoch": 0.056486741909903646, "grad_norm": 0.10148585587739944, "learning_rate": 1.9641615683220017e-05, "loss": 0.0173, "step": 1300 }, { "epoch": 0.0569212553092106, "grad_norm": 0.1246766597032547, "learning_rate": 1.9634109401931076e-05, "loss": 0.0146, "step": 1310 }, { "epoch": 0.05735576870851755, "grad_norm": 0.10318879783153534, "learning_rate": 1.962652678896299e-05, "loss": 0.0161, "step": 1320 }, { "epoch": 0.0577902821078245, "grad_norm": 0.14044098556041718, "learning_rate": 1.9618867904393303e-05, "loss": 0.0134, "step": 1330 }, { "epoch": 0.05822479550713145, "grad_norm": 0.1771666258573532, "learning_rate": 1.9611132808903854e-05, "loss": 0.014, "step": 1340 }, { "epoch": 0.058659308906438404, "grad_norm": 0.13592560589313507, "learning_rate": 1.960332156378031e-05, "loss": 0.0154, "step": 1350 }, { "epoch": 0.05909382230574536, "grad_norm": 0.1836591511964798, "learning_rate": 1.9595434230911676e-05, "loss": 0.015, "step": 1360 }, { "epoch": 0.0595283357050523, "grad_norm": 0.1101338267326355, "learning_rate": 1.9587470872789813e-05, "loss": 0.0137, "step": 1370 }, { "epoch": 0.059962849104359256, "grad_norm": 0.11457395553588867, "learning_rate": 1.957943155250892e-05, "loss": 0.0139, "step": 1380 }, { "epoch": 0.06039736250366621, "grad_norm": 0.11563432216644287, "learning_rate": 1.9571316333765066e-05, "loss": 0.0141, "step": 1390 }, { "epoch": 0.060831875902973155, "grad_norm": 0.21389071643352509, "learning_rate": 1.9563125280855655e-05, "loss": 0.0148, "step": 1400 }, { "epoch": 0.06126638930228011, "grad_norm": 0.12076102942228317, "learning_rate": 1.9554858458678935e-05, "loss": 0.0136, "step": 1410 }, { "epoch": 0.06170090270158706, "grad_norm": 0.14556382596492767, "learning_rate": 1.9546515932733482e-05, "loss": 0.0134, "step": 1420 }, { "epoch": 0.062135416100894014, "grad_norm": 0.10479048639535904, "learning_rate": 1.9538097769117673e-05, "loss": 0.0159, "step": 1430 }, { "epoch": 0.06256992950020096, "grad_norm": 0.1761564463376999, "learning_rate": 1.9529604034529167e-05, "loss": 0.0169, "step": 1440 }, { "epoch": 0.06300444289950792, "grad_norm": 0.11594516783952713, "learning_rate": 1.952103479626438e-05, "loss": 0.014, "step": 1450 }, { "epoch": 0.06343895629881487, "grad_norm": 0.16869281232357025, "learning_rate": 1.951239012221795e-05, "loss": 0.0132, "step": 1460 }, { "epoch": 0.06387346969812181, "grad_norm": 0.12788125872612, "learning_rate": 1.9503670080882196e-05, "loss": 0.0166, "step": 1470 }, { "epoch": 0.06430798309742877, "grad_norm": 0.14619000256061554, "learning_rate": 1.949487474134657e-05, "loss": 0.0163, "step": 1480 }, { "epoch": 0.06474249649673572, "grad_norm": 0.12555566430091858, "learning_rate": 1.9486004173297127e-05, "loss": 0.0133, "step": 1490 }, { "epoch": 0.06517700989604266, "grad_norm": 0.17209427058696747, "learning_rate": 1.9477058447015958e-05, "loss": 0.0173, "step": 1500 }, { "epoch": 0.06561152329534962, "grad_norm": 0.11963009834289551, "learning_rate": 1.9468037633380638e-05, "loss": 0.0142, "step": 1510 }, { "epoch": 0.06604603669465657, "grad_norm": 0.21612398326396942, "learning_rate": 1.9458941803863662e-05, "loss": 0.0151, "step": 1520 }, { "epoch": 0.06648055009396352, "grad_norm": 0.10102367401123047, "learning_rate": 1.9449771030531884e-05, "loss": 0.0135, "step": 1530 }, { "epoch": 0.06691506349327048, "grad_norm": 0.12625013291835785, "learning_rate": 1.9440525386045938e-05, "loss": 0.0166, "step": 1540 }, { "epoch": 0.06734957689257742, "grad_norm": 0.14682821929454803, "learning_rate": 1.9431204943659673e-05, "loss": 0.014, "step": 1550 }, { "epoch": 0.06778409029188437, "grad_norm": 0.07684620469808578, "learning_rate": 1.9421809777219566e-05, "loss": 0.0124, "step": 1560 }, { "epoch": 0.06821860369119133, "grad_norm": 0.17129260301589966, "learning_rate": 1.9412339961164125e-05, "loss": 0.0198, "step": 1570 }, { "epoch": 0.06865311709049828, "grad_norm": 0.1657508760690689, "learning_rate": 1.9402795570523337e-05, "loss": 0.0153, "step": 1580 }, { "epoch": 0.06908763048980524, "grad_norm": 0.1310373842716217, "learning_rate": 1.9393176680918023e-05, "loss": 0.0143, "step": 1590 }, { "epoch": 0.06952214388911218, "grad_norm": 0.11203637719154358, "learning_rate": 1.938348336855928e-05, "loss": 0.0161, "step": 1600 }, { "epoch": 0.06995665728841913, "grad_norm": 0.10988133400678635, "learning_rate": 1.9373715710247855e-05, "loss": 0.0157, "step": 1610 }, { "epoch": 0.07039117068772609, "grad_norm": 0.12275862693786621, "learning_rate": 1.9363873783373538e-05, "loss": 0.0145, "step": 1620 }, { "epoch": 0.07082568408703303, "grad_norm": 0.0957581177353859, "learning_rate": 1.935395766591457e-05, "loss": 0.0122, "step": 1630 }, { "epoch": 0.07126019748633998, "grad_norm": 0.1373092085123062, "learning_rate": 1.934396743643699e-05, "loss": 0.0104, "step": 1640 }, { "epoch": 0.07169471088564694, "grad_norm": 0.11010999977588654, "learning_rate": 1.9333903174094042e-05, "loss": 0.0129, "step": 1650 }, { "epoch": 0.07212922428495389, "grad_norm": 0.1349271684885025, "learning_rate": 1.9323764958625538e-05, "loss": 0.0146, "step": 1660 }, { "epoch": 0.07256373768426083, "grad_norm": 0.12207391858100891, "learning_rate": 1.931355287035722e-05, "loss": 0.0146, "step": 1670 }, { "epoch": 0.07299825108356779, "grad_norm": 0.1334318071603775, "learning_rate": 1.9303266990200132e-05, "loss": 0.0143, "step": 1680 }, { "epoch": 0.07343276448287474, "grad_norm": 0.151198148727417, "learning_rate": 1.9292907399649974e-05, "loss": 0.0108, "step": 1690 }, { "epoch": 0.0738672778821817, "grad_norm": 0.14836585521697998, "learning_rate": 1.928247418078646e-05, "loss": 0.0158, "step": 1700 }, { "epoch": 0.07430179128148864, "grad_norm": 0.11701719462871552, "learning_rate": 1.927196741627267e-05, "loss": 0.0165, "step": 1710 }, { "epoch": 0.07473630468079559, "grad_norm": 0.14887355268001556, "learning_rate": 1.926138718935438e-05, "loss": 0.0125, "step": 1720 }, { "epoch": 0.07517081808010255, "grad_norm": 0.09698404371738434, "learning_rate": 1.9250733583859426e-05, "loss": 0.0143, "step": 1730 }, { "epoch": 0.0756053314794095, "grad_norm": 0.12021283805370331, "learning_rate": 1.9240006684197018e-05, "loss": 0.016, "step": 1740 }, { "epoch": 0.07603984487871644, "grad_norm": 0.1289375126361847, "learning_rate": 1.9229206575357086e-05, "loss": 0.0196, "step": 1750 }, { "epoch": 0.0764743582780234, "grad_norm": 0.09976855665445328, "learning_rate": 1.9218333342909595e-05, "loss": 0.0137, "step": 1760 }, { "epoch": 0.07690887167733035, "grad_norm": 0.14193719625473022, "learning_rate": 1.9207387073003882e-05, "loss": 0.0125, "step": 1770 }, { "epoch": 0.0773433850766373, "grad_norm": 0.11004018038511276, "learning_rate": 1.9196367852367958e-05, "loss": 0.0132, "step": 1780 }, { "epoch": 0.07777789847594425, "grad_norm": 0.08808097243309021, "learning_rate": 1.918527576830783e-05, "loss": 0.0159, "step": 1790 }, { "epoch": 0.0782124118752512, "grad_norm": 0.10530976951122284, "learning_rate": 1.9174110908706803e-05, "loss": 0.0152, "step": 1800 }, { "epoch": 0.07864692527455816, "grad_norm": 0.1097542867064476, "learning_rate": 1.916287336202479e-05, "loss": 0.0119, "step": 1810 }, { "epoch": 0.0790814386738651, "grad_norm": 0.09065093100070953, "learning_rate": 1.9151563217297612e-05, "loss": 0.0124, "step": 1820 }, { "epoch": 0.07951595207317205, "grad_norm": 0.1255994588136673, "learning_rate": 1.9140180564136285e-05, "loss": 0.0136, "step": 1830 }, { "epoch": 0.07995046547247901, "grad_norm": 0.14145658910274506, "learning_rate": 1.912872549272631e-05, "loss": 0.0157, "step": 1840 }, { "epoch": 0.08038497887178596, "grad_norm": 0.12845374643802643, "learning_rate": 1.9117198093826973e-05, "loss": 0.0139, "step": 1850 }, { "epoch": 0.0808194922710929, "grad_norm": 0.13135316967964172, "learning_rate": 1.910559845877061e-05, "loss": 0.0126, "step": 1860 }, { "epoch": 0.08125400567039986, "grad_norm": 0.09882261604070663, "learning_rate": 1.9093926679461883e-05, "loss": 0.012, "step": 1870 }, { "epoch": 0.08168851906970681, "grad_norm": 0.12513484060764313, "learning_rate": 1.9082182848377066e-05, "loss": 0.012, "step": 1880 }, { "epoch": 0.08212303246901376, "grad_norm": 0.11127033084630966, "learning_rate": 1.90703670585633e-05, "loss": 0.0138, "step": 1890 }, { "epoch": 0.08255754586832072, "grad_norm": 0.09939371049404144, "learning_rate": 1.9058479403637867e-05, "loss": 0.014, "step": 1900 }, { "epoch": 0.08299205926762766, "grad_norm": 0.15272824466228485, "learning_rate": 1.9046519977787424e-05, "loss": 0.0116, "step": 1910 }, { "epoch": 0.08342657266693462, "grad_norm": 0.11801430583000183, "learning_rate": 1.9034488875767296e-05, "loss": 0.0117, "step": 1920 }, { "epoch": 0.08386108606624157, "grad_norm": 0.12052637338638306, "learning_rate": 1.9022386192900682e-05, "loss": 0.0169, "step": 1930 }, { "epoch": 0.08429559946554852, "grad_norm": 0.13256338238716125, "learning_rate": 1.9010212025077938e-05, "loss": 0.0125, "step": 1940 }, { "epoch": 0.08473011286485548, "grad_norm": 0.09443484246730804, "learning_rate": 1.8997966468755785e-05, "loss": 0.014, "step": 1950 }, { "epoch": 0.08516462626416242, "grad_norm": 0.10231075435876846, "learning_rate": 1.898564962095657e-05, "loss": 0.0107, "step": 1960 }, { "epoch": 0.08559913966346937, "grad_norm": 0.10564391314983368, "learning_rate": 1.8973261579267486e-05, "loss": 0.0129, "step": 1970 }, { "epoch": 0.08603365306277633, "grad_norm": 0.14694425463676453, "learning_rate": 1.8960802441839794e-05, "loss": 0.0118, "step": 1980 }, { "epoch": 0.08646816646208327, "grad_norm": 0.14167533814907074, "learning_rate": 1.894827230738806e-05, "loss": 0.0136, "step": 1990 }, { "epoch": 0.08690267986139022, "grad_norm": 0.11116527765989304, "learning_rate": 1.8935671275189356e-05, "loss": 0.0136, "step": 2000 }, { "epoch": 0.08733719326069718, "grad_norm": 0.10359754413366318, "learning_rate": 1.8922999445082484e-05, "loss": 0.0115, "step": 2010 }, { "epoch": 0.08777170666000413, "grad_norm": 0.10434489697217941, "learning_rate": 1.8910256917467183e-05, "loss": 0.0118, "step": 2020 }, { "epoch": 0.08820622005931109, "grad_norm": 0.1596977263689041, "learning_rate": 1.8897443793303334e-05, "loss": 0.0119, "step": 2030 }, { "epoch": 0.08864073345861803, "grad_norm": 0.11361852288246155, "learning_rate": 1.888456017411016e-05, "loss": 0.013, "step": 2040 }, { "epoch": 0.08907524685792498, "grad_norm": 0.10693772882223129, "learning_rate": 1.8871606161965416e-05, "loss": 0.0142, "step": 2050 }, { "epoch": 0.08950976025723194, "grad_norm": 0.11561683565378189, "learning_rate": 1.8858581859504587e-05, "loss": 0.0297, "step": 2060 }, { "epoch": 0.08994427365653888, "grad_norm": 0.2715816795825958, "learning_rate": 1.8845487369920076e-05, "loss": 0.0105, "step": 2070 }, { "epoch": 0.09037878705584583, "grad_norm": 0.08823924511671066, "learning_rate": 1.883232279696038e-05, "loss": 0.014, "step": 2080 }, { "epoch": 0.09081330045515279, "grad_norm": 0.2058764398097992, "learning_rate": 1.8819088244929275e-05, "loss": 0.0129, "step": 2090 }, { "epoch": 0.09124781385445974, "grad_norm": 0.11656977981328964, "learning_rate": 1.8805783818684976e-05, "loss": 0.0127, "step": 2100 }, { "epoch": 0.09168232725376668, "grad_norm": 0.14761896431446075, "learning_rate": 1.8792409623639325e-05, "loss": 0.0149, "step": 2110 }, { "epoch": 0.09211684065307364, "grad_norm": 0.18636038899421692, "learning_rate": 1.8778965765756946e-05, "loss": 0.0152, "step": 2120 }, { "epoch": 0.09255135405238059, "grad_norm": 0.11357667297124863, "learning_rate": 1.8765452351554408e-05, "loss": 0.0134, "step": 2130 }, { "epoch": 0.09298586745168755, "grad_norm": 0.26886075735092163, "learning_rate": 1.875186948809937e-05, "loss": 0.0108, "step": 2140 }, { "epoch": 0.0934203808509945, "grad_norm": 0.1092400997877121, "learning_rate": 1.8738217283009747e-05, "loss": 0.0135, "step": 2150 }, { "epoch": 0.09385489425030144, "grad_norm": 0.138901486992836, "learning_rate": 1.872449584445286e-05, "loss": 0.0116, "step": 2160 }, { "epoch": 0.0942894076496084, "grad_norm": 0.0775412917137146, "learning_rate": 1.8710705281144557e-05, "loss": 0.015, "step": 2170 }, { "epoch": 0.09472392104891535, "grad_norm": 0.12415778636932373, "learning_rate": 1.869684570234838e-05, "loss": 0.0135, "step": 2180 }, { "epoch": 0.09515843444822229, "grad_norm": 0.15818843245506287, "learning_rate": 1.8682917217874675e-05, "loss": 0.0153, "step": 2190 }, { "epoch": 0.09559294784752925, "grad_norm": 0.15443842113018036, "learning_rate": 1.8668919938079738e-05, "loss": 0.0146, "step": 2200 }, { "epoch": 0.0960274612468362, "grad_norm": 0.14416638016700745, "learning_rate": 1.8654853973864928e-05, "loss": 0.0144, "step": 2210 }, { "epoch": 0.09646197464614314, "grad_norm": 0.14622808992862701, "learning_rate": 1.8640719436675806e-05, "loss": 0.0127, "step": 2220 }, { "epoch": 0.0968964880454501, "grad_norm": 0.10530474781990051, "learning_rate": 1.862651643850123e-05, "loss": 0.0132, "step": 2230 }, { "epoch": 0.09733100144475705, "grad_norm": 0.13297906517982483, "learning_rate": 1.8612245091872484e-05, "loss": 0.0096, "step": 2240 }, { "epoch": 0.097765514844064, "grad_norm": 0.09163651615381241, "learning_rate": 1.8597905509862386e-05, "loss": 0.0142, "step": 2250 }, { "epoch": 0.09820002824337096, "grad_norm": 0.09490437805652618, "learning_rate": 1.8583497806084377e-05, "loss": 0.0104, "step": 2260 }, { "epoch": 0.0986345416426779, "grad_norm": 0.12775211036205292, "learning_rate": 1.856902209469164e-05, "loss": 0.0182, "step": 2270 }, { "epoch": 0.09906905504198486, "grad_norm": 0.14316552877426147, "learning_rate": 1.8554478490376186e-05, "loss": 0.0141, "step": 2280 }, { "epoch": 0.09950356844129181, "grad_norm": 0.4003185033798218, "learning_rate": 1.8539867108367937e-05, "loss": 0.01, "step": 2290 }, { "epoch": 0.09993808184059876, "grad_norm": 0.14583055675029755, "learning_rate": 1.8525188064433827e-05, "loss": 0.0143, "step": 2300 }, { "epoch": 0.10037259523990572, "grad_norm": 0.18036730587482452, "learning_rate": 1.8510441474876893e-05, "loss": 0.0136, "step": 2310 }, { "epoch": 0.10080710863921266, "grad_norm": 0.17284591495990753, "learning_rate": 1.8495627456535316e-05, "loss": 0.0127, "step": 2320 }, { "epoch": 0.10124162203851961, "grad_norm": 0.11936520785093307, "learning_rate": 1.8480746126781544e-05, "loss": 0.0123, "step": 2330 }, { "epoch": 0.10167613543782657, "grad_norm": 0.11931222677230835, "learning_rate": 1.846579760352132e-05, "loss": 0.0125, "step": 2340 }, { "epoch": 0.10211064883713351, "grad_norm": 0.10272102057933807, "learning_rate": 1.845078200519277e-05, "loss": 0.0176, "step": 2350 }, { "epoch": 0.10254516223644046, "grad_norm": 0.2629345655441284, "learning_rate": 1.8435699450765467e-05, "loss": 0.0127, "step": 2360 }, { "epoch": 0.10297967563574742, "grad_norm": 0.1450505554676056, "learning_rate": 1.8420550059739476e-05, "loss": 0.0121, "step": 2370 }, { "epoch": 0.10341418903505437, "grad_norm": 0.1271376609802246, "learning_rate": 1.840533395214441e-05, "loss": 0.012, "step": 2380 }, { "epoch": 0.10384870243436133, "grad_norm": 0.11844556778669357, "learning_rate": 1.839005124853849e-05, "loss": 0.0166, "step": 2390 }, { "epoch": 0.10428321583366827, "grad_norm": 0.16082872450351715, "learning_rate": 1.837470207000757e-05, "loss": 0.0122, "step": 2400 }, { "epoch": 0.10471772923297522, "grad_norm": 0.12687557935714722, "learning_rate": 1.83592865381642e-05, "loss": 0.0118, "step": 2410 }, { "epoch": 0.10515224263228218, "grad_norm": 0.17328034341335297, "learning_rate": 1.8343804775146646e-05, "loss": 0.0142, "step": 2420 }, { "epoch": 0.10558675603158912, "grad_norm": 0.09038952738046646, "learning_rate": 1.8328256903617928e-05, "loss": 0.0156, "step": 2430 }, { "epoch": 0.10602126943089607, "grad_norm": 0.096098393201828, "learning_rate": 1.8312643046764854e-05, "loss": 0.0151, "step": 2440 }, { "epoch": 0.10645578283020303, "grad_norm": 0.10437561571598053, "learning_rate": 1.829696332829703e-05, "loss": 0.0129, "step": 2450 }, { "epoch": 0.10689029622950998, "grad_norm": 0.13580961525440216, "learning_rate": 1.8281217872445894e-05, "loss": 0.0114, "step": 2460 }, { "epoch": 0.10732480962881692, "grad_norm": 0.1774880290031433, "learning_rate": 1.8265406803963723e-05, "loss": 0.0135, "step": 2470 }, { "epoch": 0.10775932302812388, "grad_norm": 0.12563155591487885, "learning_rate": 1.8249530248122643e-05, "loss": 0.0118, "step": 2480 }, { "epoch": 0.10819383642743083, "grad_norm": 0.0847230926156044, "learning_rate": 1.8233588330713648e-05, "loss": 0.0141, "step": 2490 }, { "epoch": 0.10862834982673779, "grad_norm": 0.11921481043100357, "learning_rate": 1.8217581178045588e-05, "loss": 0.0097, "step": 2500 }, { "epoch": 0.10906286322604473, "grad_norm": 0.14641766250133514, "learning_rate": 1.8201508916944187e-05, "loss": 0.0134, "step": 2510 }, { "epoch": 0.10949737662535168, "grad_norm": 0.10427770018577576, "learning_rate": 1.818537167475102e-05, "loss": 0.012, "step": 2520 }, { "epoch": 0.10993189002465864, "grad_norm": 0.09014574438333511, "learning_rate": 1.816916957932251e-05, "loss": 0.0111, "step": 2530 }, { "epoch": 0.11036640342396559, "grad_norm": 0.221757173538208, "learning_rate": 1.815290275902892e-05, "loss": 0.0135, "step": 2540 }, { "epoch": 0.11080091682327253, "grad_norm": 0.1362486183643341, "learning_rate": 1.813657134275333e-05, "loss": 0.0112, "step": 2550 }, { "epoch": 0.11123543022257949, "grad_norm": 0.08970905840396881, "learning_rate": 1.812017545989063e-05, "loss": 0.013, "step": 2560 }, { "epoch": 0.11166994362188644, "grad_norm": 0.122174933552742, "learning_rate": 1.810371524034646e-05, "loss": 0.0127, "step": 2570 }, { "epoch": 0.11210445702119338, "grad_norm": 0.10636741667985916, "learning_rate": 1.808719081453622e-05, "loss": 0.0113, "step": 2580 }, { "epoch": 0.11253897042050034, "grad_norm": 0.08577335625886917, "learning_rate": 1.8070602313384018e-05, "loss": 0.0121, "step": 2590 }, { "epoch": 0.11297348381980729, "grad_norm": 0.11767273396253586, "learning_rate": 1.8053949868321637e-05, "loss": 0.0111, "step": 2600 }, { "epoch": 0.11340799721911425, "grad_norm": 0.1318366676568985, "learning_rate": 1.803723361128748e-05, "loss": 0.013, "step": 2610 }, { "epoch": 0.1138425106184212, "grad_norm": 0.08728184551000595, "learning_rate": 1.8020453674725557e-05, "loss": 0.0128, "step": 2620 }, { "epoch": 0.11427702401772814, "grad_norm": 0.07697049528360367, "learning_rate": 1.8003610191584394e-05, "loss": 0.0094, "step": 2630 }, { "epoch": 0.1147115374170351, "grad_norm": 0.07770522683858871, "learning_rate": 1.7986703295316018e-05, "loss": 0.011, "step": 2640 }, { "epoch": 0.11514605081634205, "grad_norm": 0.13932470977306366, "learning_rate": 1.7969733119874866e-05, "loss": 0.0131, "step": 2650 }, { "epoch": 0.115580564215649, "grad_norm": 0.09085828810930252, "learning_rate": 1.795269979971675e-05, "loss": 0.0113, "step": 2660 }, { "epoch": 0.11601507761495596, "grad_norm": 0.1439436376094818, "learning_rate": 1.7935603469797784e-05, "loss": 0.0119, "step": 2670 }, { "epoch": 0.1164495910142629, "grad_norm": 0.08274544030427933, "learning_rate": 1.7918444265573308e-05, "loss": 0.0149, "step": 2680 }, { "epoch": 0.11688410441356985, "grad_norm": 0.15263070166110992, "learning_rate": 1.7901222322996815e-05, "loss": 0.0124, "step": 2690 }, { "epoch": 0.11731861781287681, "grad_norm": 0.059803709387779236, "learning_rate": 1.788393777851889e-05, "loss": 0.0117, "step": 2700 }, { "epoch": 0.11775313121218375, "grad_norm": 0.10974439978599548, "learning_rate": 1.7866590769086104e-05, "loss": 0.0146, "step": 2710 }, { "epoch": 0.11818764461149071, "grad_norm": 0.11613568663597107, "learning_rate": 1.7849181432139946e-05, "loss": 0.0134, "step": 2720 }, { "epoch": 0.11862215801079766, "grad_norm": 0.0861983597278595, "learning_rate": 1.7831709905615744e-05, "loss": 0.0123, "step": 2730 }, { "epoch": 0.1190566714101046, "grad_norm": 0.0727258250117302, "learning_rate": 1.781417632794153e-05, "loss": 0.0106, "step": 2740 }, { "epoch": 0.11949118480941157, "grad_norm": 0.09521336853504181, "learning_rate": 1.7796580838037e-05, "loss": 0.0108, "step": 2750 }, { "epoch": 0.11992569820871851, "grad_norm": 0.08138874173164368, "learning_rate": 1.777892357531236e-05, "loss": 0.011, "step": 2760 }, { "epoch": 0.12036021160802546, "grad_norm": 0.10241363942623138, "learning_rate": 1.776120467966727e-05, "loss": 0.0131, "step": 2770 }, { "epoch": 0.12079472500733242, "grad_norm": 0.11157892644405365, "learning_rate": 1.774342429148969e-05, "loss": 0.0107, "step": 2780 }, { "epoch": 0.12122923840663936, "grad_norm": 0.08816739171743393, "learning_rate": 1.7725582551654804e-05, "loss": 0.0107, "step": 2790 }, { "epoch": 0.12166375180594631, "grad_norm": 0.0863705575466156, "learning_rate": 1.7707679601523882e-05, "loss": 0.0111, "step": 2800 }, { "epoch": 0.12209826520525327, "grad_norm": 0.12580861151218414, "learning_rate": 1.7689715582943167e-05, "loss": 0.0118, "step": 2810 }, { "epoch": 0.12253277860456022, "grad_norm": 0.12607578933238983, "learning_rate": 1.7671690638242763e-05, "loss": 0.0136, "step": 2820 }, { "epoch": 0.12296729200386716, "grad_norm": 0.12284299731254578, "learning_rate": 1.7653604910235474e-05, "loss": 0.0123, "step": 2830 }, { "epoch": 0.12340180540317412, "grad_norm": 0.10298939794301987, "learning_rate": 1.763545854221571e-05, "loss": 0.0097, "step": 2840 }, { "epoch": 0.12383631880248107, "grad_norm": 0.0938117504119873, "learning_rate": 1.761725167795834e-05, "loss": 0.0117, "step": 2850 }, { "epoch": 0.12427083220178803, "grad_norm": 0.1595860719680786, "learning_rate": 1.7598984461717532e-05, "loss": 0.015, "step": 2860 }, { "epoch": 0.12470534560109497, "grad_norm": 0.1603042334318161, "learning_rate": 1.758065703822564e-05, "loss": 0.0106, "step": 2870 }, { "epoch": 0.12513985900040192, "grad_norm": 0.10560432821512222, "learning_rate": 1.756226955269204e-05, "loss": 0.011, "step": 2880 }, { "epoch": 0.12557437239970887, "grad_norm": 0.1174912303686142, "learning_rate": 1.7543822150801975e-05, "loss": 0.0126, "step": 2890 }, { "epoch": 0.12600888579901584, "grad_norm": 0.09937790781259537, "learning_rate": 1.7525314978715425e-05, "loss": 0.0129, "step": 2900 }, { "epoch": 0.1264433991983228, "grad_norm": 0.10457811504602432, "learning_rate": 1.7506748183065925e-05, "loss": 0.0137, "step": 2910 }, { "epoch": 0.12687791259762973, "grad_norm": 0.15306922793388367, "learning_rate": 1.7488121910959405e-05, "loss": 0.0116, "step": 2920 }, { "epoch": 0.12731242599693668, "grad_norm": 0.0991855189204216, "learning_rate": 1.7469436309973046e-05, "loss": 0.0125, "step": 2930 }, { "epoch": 0.12774693939624363, "grad_norm": 0.20841960608959198, "learning_rate": 1.7450691528154087e-05, "loss": 0.0134, "step": 2940 }, { "epoch": 0.12818145279555057, "grad_norm": 0.09408193081617355, "learning_rate": 1.7431887714018653e-05, "loss": 0.0153, "step": 2950 }, { "epoch": 0.12861596619485755, "grad_norm": 0.1006455197930336, "learning_rate": 1.74130250165506e-05, "loss": 0.0105, "step": 2960 }, { "epoch": 0.1290504795941645, "grad_norm": 0.0960792750120163, "learning_rate": 1.7394103585200316e-05, "loss": 0.0128, "step": 2970 }, { "epoch": 0.12948499299347144, "grad_norm": 0.15480937063694, "learning_rate": 1.737512356988353e-05, "loss": 0.0136, "step": 2980 }, { "epoch": 0.12991950639277838, "grad_norm": 0.07801657170057297, "learning_rate": 1.7356085120980154e-05, "loss": 0.0105, "step": 2990 }, { "epoch": 0.13035401979208533, "grad_norm": 0.0841207206249237, "learning_rate": 1.7336988389333064e-05, "loss": 0.0133, "step": 3000 }, { "epoch": 0.1307885331913923, "grad_norm": 0.08529253304004669, "learning_rate": 1.731783352624691e-05, "loss": 0.0118, "step": 3010 }, { "epoch": 0.13122304659069925, "grad_norm": 0.09804724156856537, "learning_rate": 1.7298620683486927e-05, "loss": 0.0117, "step": 3020 }, { "epoch": 0.1316575599900062, "grad_norm": 0.09919776022434235, "learning_rate": 1.7279350013277725e-05, "loss": 0.0126, "step": 3030 }, { "epoch": 0.13209207338931314, "grad_norm": 0.11483841389417648, "learning_rate": 1.726002166830209e-05, "loss": 0.011, "step": 3040 }, { "epoch": 0.1325265867886201, "grad_norm": 0.12246759980916977, "learning_rate": 1.7240635801699755e-05, "loss": 0.0131, "step": 3050 }, { "epoch": 0.13296110018792703, "grad_norm": 0.13971829414367676, "learning_rate": 1.7221192567066215e-05, "loss": 0.0103, "step": 3060 }, { "epoch": 0.133395613587234, "grad_norm": 0.14394456148147583, "learning_rate": 1.720169211845149e-05, "loss": 0.0131, "step": 3070 }, { "epoch": 0.13383012698654095, "grad_norm": 0.13212616741657257, "learning_rate": 1.718213461035891e-05, "loss": 0.0127, "step": 3080 }, { "epoch": 0.1342646403858479, "grad_norm": 0.10326551645994186, "learning_rate": 1.716252019774389e-05, "loss": 0.0104, "step": 3090 }, { "epoch": 0.13469915378515485, "grad_norm": 0.11226284503936768, "learning_rate": 1.7142849036012706e-05, "loss": 0.0116, "step": 3100 }, { "epoch": 0.1351336671844618, "grad_norm": 0.10741928964853287, "learning_rate": 1.712312128102126e-05, "loss": 0.0107, "step": 3110 }, { "epoch": 0.13556818058376874, "grad_norm": 0.3138485550880432, "learning_rate": 1.710333708907384e-05, "loss": 0.015, "step": 3120 }, { "epoch": 0.1360026939830757, "grad_norm": 0.13233919441699982, "learning_rate": 1.70834966169219e-05, "loss": 0.0104, "step": 3130 }, { "epoch": 0.13643720738238266, "grad_norm": 0.09749718010425568, "learning_rate": 1.7063600021762798e-05, "loss": 0.0112, "step": 3140 }, { "epoch": 0.1368717207816896, "grad_norm": 0.11438994854688644, "learning_rate": 1.7043647461238557e-05, "loss": 0.0154, "step": 3150 }, { "epoch": 0.13730623418099655, "grad_norm": 0.11394909024238586, "learning_rate": 1.702363909343462e-05, "loss": 0.0132, "step": 3160 }, { "epoch": 0.1377407475803035, "grad_norm": 0.07735557854175568, "learning_rate": 1.7003575076878593e-05, "loss": 0.0119, "step": 3170 }, { "epoch": 0.13817526097961047, "grad_norm": 0.08114404231309891, "learning_rate": 1.6983455570538996e-05, "loss": 0.0106, "step": 3180 }, { "epoch": 0.13860977437891742, "grad_norm": 0.08775005489587784, "learning_rate": 1.696328073382399e-05, "loss": 0.0108, "step": 3190 }, { "epoch": 0.13904428777822436, "grad_norm": 0.10539509356021881, "learning_rate": 1.694305072658013e-05, "loss": 0.0142, "step": 3200 }, { "epoch": 0.1394788011775313, "grad_norm": 0.11596210300922394, "learning_rate": 1.6922765709091085e-05, "loss": 0.0116, "step": 3210 }, { "epoch": 0.13991331457683825, "grad_norm": 0.11444628238677979, "learning_rate": 1.6902425842076372e-05, "loss": 0.0099, "step": 3220 }, { "epoch": 0.1403478279761452, "grad_norm": 0.11727698147296906, "learning_rate": 1.6882031286690095e-05, "loss": 0.0099, "step": 3230 }, { "epoch": 0.14078234137545217, "grad_norm": 0.06600234657526016, "learning_rate": 1.6861582204519648e-05, "loss": 0.0113, "step": 3240 }, { "epoch": 0.14121685477475912, "grad_norm": 0.1237931102514267, "learning_rate": 1.6841078757584445e-05, "loss": 0.0109, "step": 3250 }, { "epoch": 0.14165136817406607, "grad_norm": 0.12242930382490158, "learning_rate": 1.6820521108334643e-05, "loss": 0.0135, "step": 3260 }, { "epoch": 0.142085881573373, "grad_norm": 0.1355576366186142, "learning_rate": 1.6799909419649835e-05, "loss": 0.0118, "step": 3270 }, { "epoch": 0.14252039497267996, "grad_norm": 0.10697627067565918, "learning_rate": 1.6779243854837784e-05, "loss": 0.0113, "step": 3280 }, { "epoch": 0.14295490837198693, "grad_norm": 0.12463831156492233, "learning_rate": 1.675852457763311e-05, "loss": 0.0106, "step": 3290 }, { "epoch": 0.14338942177129388, "grad_norm": 0.1064586415886879, "learning_rate": 1.6737751752196005e-05, "loss": 0.0089, "step": 3300 }, { "epoch": 0.14382393517060083, "grad_norm": 0.1283196061849594, "learning_rate": 1.6716925543110916e-05, "loss": 0.0105, "step": 3310 }, { "epoch": 0.14425844856990777, "grad_norm": 0.09701595455408096, "learning_rate": 1.669604611538527e-05, "loss": 0.0112, "step": 3320 }, { "epoch": 0.14469296196921472, "grad_norm": 0.06893642246723175, "learning_rate": 1.6675113634448136e-05, "loss": 0.0105, "step": 3330 }, { "epoch": 0.14512747536852166, "grad_norm": 0.13216637074947357, "learning_rate": 1.6654128266148926e-05, "loss": 0.0135, "step": 3340 }, { "epoch": 0.14556198876782864, "grad_norm": 0.09184638410806656, "learning_rate": 1.6633090176756092e-05, "loss": 0.0103, "step": 3350 }, { "epoch": 0.14599650216713558, "grad_norm": 0.07556001096963882, "learning_rate": 1.6611999532955783e-05, "loss": 0.011, "step": 3360 }, { "epoch": 0.14643101556644253, "grad_norm": 0.045008886605501175, "learning_rate": 1.6590856501850562e-05, "loss": 0.0093, "step": 3370 }, { "epoch": 0.14686552896574948, "grad_norm": 0.07397887110710144, "learning_rate": 1.6569661250958042e-05, "loss": 0.0083, "step": 3380 }, { "epoch": 0.14730004236505642, "grad_norm": 0.09333399683237076, "learning_rate": 1.6548413948209584e-05, "loss": 0.0116, "step": 3390 }, { "epoch": 0.1477345557643634, "grad_norm": 0.09300416707992554, "learning_rate": 1.6527114761948957e-05, "loss": 0.0124, "step": 3400 }, { "epoch": 0.14816906916367034, "grad_norm": 0.0548076257109642, "learning_rate": 1.6505763860931002e-05, "loss": 0.0104, "step": 3410 }, { "epoch": 0.1486035825629773, "grad_norm": 0.09666861593723297, "learning_rate": 1.6484361414320312e-05, "loss": 0.0119, "step": 3420 }, { "epoch": 0.14903809596228423, "grad_norm": 0.1338212788105011, "learning_rate": 1.6462907591689875e-05, "loss": 0.0128, "step": 3430 }, { "epoch": 0.14947260936159118, "grad_norm": 0.11156731843948364, "learning_rate": 1.644140256301972e-05, "loss": 0.0108, "step": 3440 }, { "epoch": 0.14990712276089813, "grad_norm": 0.09104102104902267, "learning_rate": 1.6419846498695605e-05, "loss": 0.0124, "step": 3450 }, { "epoch": 0.1503416361602051, "grad_norm": 0.13596515357494354, "learning_rate": 1.639823956950764e-05, "loss": 0.0157, "step": 3460 }, { "epoch": 0.15077614955951205, "grad_norm": 0.13181833922863007, "learning_rate": 1.6376581946648928e-05, "loss": 0.0111, "step": 3470 }, { "epoch": 0.151210662958819, "grad_norm": 0.08273448050022125, "learning_rate": 1.6354873801714236e-05, "loss": 0.0089, "step": 3480 }, { "epoch": 0.15164517635812594, "grad_norm": 0.17430652678012848, "learning_rate": 1.6333115306698625e-05, "loss": 0.0092, "step": 3490 }, { "epoch": 0.15207968975743288, "grad_norm": 0.10232073068618774, "learning_rate": 1.6311306633996064e-05, "loss": 0.0094, "step": 3500 }, { "epoch": 0.15251420315673986, "grad_norm": 0.11144959181547165, "learning_rate": 1.62894479563981e-05, "loss": 0.0114, "step": 3510 }, { "epoch": 0.1529487165560468, "grad_norm": 0.0800149068236351, "learning_rate": 1.6267539447092463e-05, "loss": 0.0099, "step": 3520 }, { "epoch": 0.15338322995535375, "grad_norm": 0.09478965401649475, "learning_rate": 1.6245581279661708e-05, "loss": 0.0127, "step": 3530 }, { "epoch": 0.1538177433546607, "grad_norm": 0.10897224396467209, "learning_rate": 1.6223573628081826e-05, "loss": 0.0101, "step": 3540 }, { "epoch": 0.15425225675396764, "grad_norm": 0.06671247631311417, "learning_rate": 1.620151666672089e-05, "loss": 0.0091, "step": 3550 }, { "epoch": 0.1546867701532746, "grad_norm": 0.10703366249799728, "learning_rate": 1.617941057033764e-05, "loss": 0.0139, "step": 3560 }, { "epoch": 0.15512128355258156, "grad_norm": 0.09891420602798462, "learning_rate": 1.6157255514080134e-05, "loss": 0.0119, "step": 3570 }, { "epoch": 0.1555557969518885, "grad_norm": 0.10249175131320953, "learning_rate": 1.6135051673484323e-05, "loss": 0.0123, "step": 3580 }, { "epoch": 0.15599031035119545, "grad_norm": 0.12472105771303177, "learning_rate": 1.6112799224472686e-05, "loss": 0.0102, "step": 3590 }, { "epoch": 0.1564248237505024, "grad_norm": 0.048791538923978806, "learning_rate": 1.6090498343352844e-05, "loss": 0.0106, "step": 3600 }, { "epoch": 0.15685933714980935, "grad_norm": 0.08500496298074722, "learning_rate": 1.606814920681613e-05, "loss": 0.0109, "step": 3610 }, { "epoch": 0.15729385054911632, "grad_norm": 0.07313550263643265, "learning_rate": 1.6045751991936213e-05, "loss": 0.0099, "step": 3620 }, { "epoch": 0.15772836394842327, "grad_norm": 0.14947198331356049, "learning_rate": 1.60233068761677e-05, "loss": 0.0122, "step": 3630 }, { "epoch": 0.1581628773477302, "grad_norm": 0.17013369500637054, "learning_rate": 1.60008140373447e-05, "loss": 0.0089, "step": 3640 }, { "epoch": 0.15859739074703716, "grad_norm": 0.07465286552906036, "learning_rate": 1.5978273653679458e-05, "loss": 0.0088, "step": 3650 }, { "epoch": 0.1590319041463441, "grad_norm": 0.0930761843919754, "learning_rate": 1.5955685903760905e-05, "loss": 0.0101, "step": 3660 }, { "epoch": 0.15946641754565105, "grad_norm": 0.1549626737833023, "learning_rate": 1.593305096655326e-05, "loss": 0.0103, "step": 3670 }, { "epoch": 0.15990093094495803, "grad_norm": 0.07674683630466461, "learning_rate": 1.591036902139461e-05, "loss": 0.0113, "step": 3680 }, { "epoch": 0.16033544434426497, "grad_norm": 0.06362561136484146, "learning_rate": 1.5887640247995495e-05, "loss": 0.0122, "step": 3690 }, { "epoch": 0.16076995774357192, "grad_norm": 0.09524749219417572, "learning_rate": 1.5864864826437473e-05, "loss": 0.0093, "step": 3700 }, { "epoch": 0.16120447114287886, "grad_norm": 0.10837098211050034, "learning_rate": 1.5842042937171696e-05, "loss": 0.0099, "step": 3710 }, { "epoch": 0.1616389845421858, "grad_norm": 0.1316443234682083, "learning_rate": 1.5819174761017485e-05, "loss": 0.0115, "step": 3720 }, { "epoch": 0.16207349794149278, "grad_norm": 0.08324498683214188, "learning_rate": 1.57962604791609e-05, "loss": 0.0101, "step": 3730 }, { "epoch": 0.16250801134079973, "grad_norm": 0.12263214588165283, "learning_rate": 1.5773300273153296e-05, "loss": 0.0107, "step": 3740 }, { "epoch": 0.16294252474010668, "grad_norm": 0.2212323546409607, "learning_rate": 1.5750294324909886e-05, "loss": 0.0111, "step": 3750 }, { "epoch": 0.16337703813941362, "grad_norm": 0.07996786385774612, "learning_rate": 1.57272428167083e-05, "loss": 0.0113, "step": 3760 }, { "epoch": 0.16381155153872057, "grad_norm": 0.10634341835975647, "learning_rate": 1.570414593118715e-05, "loss": 0.0115, "step": 3770 }, { "epoch": 0.16424606493802751, "grad_norm": 0.1128893718123436, "learning_rate": 1.5681003851344568e-05, "loss": 0.0105, "step": 3780 }, { "epoch": 0.1646805783373345, "grad_norm": 0.07602597773075104, "learning_rate": 1.5657816760536767e-05, "loss": 0.0096, "step": 3790 }, { "epoch": 0.16511509173664143, "grad_norm": 0.12339795380830765, "learning_rate": 1.5634584842476588e-05, "loss": 0.0107, "step": 3800 }, { "epoch": 0.16554960513594838, "grad_norm": 0.21459102630615234, "learning_rate": 1.5611308281232038e-05, "loss": 0.0121, "step": 3810 }, { "epoch": 0.16598411853525533, "grad_norm": 0.07586748898029327, "learning_rate": 1.5587987261224827e-05, "loss": 0.0094, "step": 3820 }, { "epoch": 0.16641863193456227, "grad_norm": 0.119880810379982, "learning_rate": 1.556462196722893e-05, "loss": 0.0132, "step": 3830 }, { "epoch": 0.16685314533386925, "grad_norm": 0.10775720328092575, "learning_rate": 1.55412125843691e-05, "loss": 0.0091, "step": 3840 }, { "epoch": 0.1672876587331762, "grad_norm": 0.1075601577758789, "learning_rate": 1.5517759298119406e-05, "loss": 0.0108, "step": 3850 }, { "epoch": 0.16772217213248314, "grad_norm": 0.10800988227128983, "learning_rate": 1.5494262294301768e-05, "loss": 0.012, "step": 3860 }, { "epoch": 0.16815668553179008, "grad_norm": 0.08098742365837097, "learning_rate": 1.547072175908449e-05, "loss": 0.0106, "step": 3870 }, { "epoch": 0.16859119893109703, "grad_norm": 0.11291100829839706, "learning_rate": 1.5447137878980768e-05, "loss": 0.0074, "step": 3880 }, { "epoch": 0.16902571233040398, "grad_norm": 0.13343898952007294, "learning_rate": 1.5423510840847228e-05, "loss": 0.0123, "step": 3890 }, { "epoch": 0.16946022572971095, "grad_norm": 0.11526608467102051, "learning_rate": 1.5399840831882442e-05, "loss": 0.0139, "step": 3900 }, { "epoch": 0.1698947391290179, "grad_norm": 0.07974295318126678, "learning_rate": 1.5376128039625438e-05, "loss": 0.0106, "step": 3910 }, { "epoch": 0.17032925252832484, "grad_norm": 0.08782044053077698, "learning_rate": 1.535237265195422e-05, "loss": 0.0108, "step": 3920 }, { "epoch": 0.1707637659276318, "grad_norm": 0.22049036622047424, "learning_rate": 1.5328574857084277e-05, "loss": 0.0164, "step": 3930 }, { "epoch": 0.17119827932693873, "grad_norm": 0.11717794090509415, "learning_rate": 1.53047348435671e-05, "loss": 0.009, "step": 3940 }, { "epoch": 0.1716327927262457, "grad_norm": 0.0919695645570755, "learning_rate": 1.5280852800288672e-05, "loss": 0.0093, "step": 3950 }, { "epoch": 0.17206730612555265, "grad_norm": 0.11691190302371979, "learning_rate": 1.5256928916467986e-05, "loss": 0.0078, "step": 3960 }, { "epoch": 0.1725018195248596, "grad_norm": 0.07742350548505783, "learning_rate": 1.5232963381655536e-05, "loss": 0.0108, "step": 3970 }, { "epoch": 0.17293633292416655, "grad_norm": 0.07558136433362961, "learning_rate": 1.5208956385731824e-05, "loss": 0.008, "step": 3980 }, { "epoch": 0.1733708463234735, "grad_norm": 0.10943376272916794, "learning_rate": 1.5184908118905853e-05, "loss": 0.0141, "step": 3990 }, { "epoch": 0.17380535972278044, "grad_norm": 0.05014868825674057, "learning_rate": 1.5160818771713609e-05, "loss": 0.0082, "step": 4000 }, { "epoch": 0.1742398731220874, "grad_norm": 0.09473850578069687, "learning_rate": 1.5136688535016571e-05, "loss": 0.0098, "step": 4010 }, { "epoch": 0.17467438652139436, "grad_norm": 0.08004864305257797, "learning_rate": 1.5112517600000179e-05, "loss": 0.0132, "step": 4020 }, { "epoch": 0.1751088999207013, "grad_norm": 0.07450365275144577, "learning_rate": 1.5088306158172334e-05, "loss": 0.0107, "step": 4030 }, { "epoch": 0.17554341332000825, "grad_norm": 0.18986289203166962, "learning_rate": 1.5064054401361872e-05, "loss": 0.0098, "step": 4040 }, { "epoch": 0.1759779267193152, "grad_norm": 0.08691735565662384, "learning_rate": 1.5039762521717054e-05, "loss": 0.0099, "step": 4050 }, { "epoch": 0.17641244011862217, "grad_norm": 0.07385091483592987, "learning_rate": 1.5015430711704027e-05, "loss": 0.0109, "step": 4060 }, { "epoch": 0.17684695351792912, "grad_norm": 0.141652449965477, "learning_rate": 1.4991059164105318e-05, "loss": 0.0113, "step": 4070 }, { "epoch": 0.17728146691723606, "grad_norm": 0.08523985743522644, "learning_rate": 1.496664807201829e-05, "loss": 0.0098, "step": 4080 }, { "epoch": 0.177715980316543, "grad_norm": 0.09805174171924591, "learning_rate": 1.494219762885362e-05, "loss": 0.01, "step": 4090 }, { "epoch": 0.17815049371584996, "grad_norm": 0.08430161327123642, "learning_rate": 1.4917708028333779e-05, "loss": 0.0134, "step": 4100 }, { "epoch": 0.1785850071151569, "grad_norm": 0.08099672198295593, "learning_rate": 1.4893179464491461e-05, "loss": 0.0098, "step": 4110 }, { "epoch": 0.17901952051446388, "grad_norm": 0.11928766220808029, "learning_rate": 1.4868612131668095e-05, "loss": 0.0094, "step": 4120 }, { "epoch": 0.17945403391377082, "grad_norm": 0.11964277923107147, "learning_rate": 1.4844006224512254e-05, "loss": 0.0108, "step": 4130 }, { "epoch": 0.17988854731307777, "grad_norm": 0.077316053211689, "learning_rate": 1.4819361937978162e-05, "loss": 0.009, "step": 4140 }, { "epoch": 0.18032306071238471, "grad_norm": 0.11310229450464249, "learning_rate": 1.4794679467324106e-05, "loss": 0.0083, "step": 4150 }, { "epoch": 0.18075757411169166, "grad_norm": 0.09045251458883286, "learning_rate": 1.4769959008110922e-05, "loss": 0.0095, "step": 4160 }, { "epoch": 0.18119208751099863, "grad_norm": 0.09859993308782578, "learning_rate": 1.4745200756200418e-05, "loss": 0.0085, "step": 4170 }, { "epoch": 0.18162660091030558, "grad_norm": 0.09493473172187805, "learning_rate": 1.4720404907753849e-05, "loss": 0.0085, "step": 4180 }, { "epoch": 0.18206111430961253, "grad_norm": 0.10200127214193344, "learning_rate": 1.4695571659230343e-05, "loss": 0.0108, "step": 4190 }, { "epoch": 0.18249562770891947, "grad_norm": 0.10617578774690628, "learning_rate": 1.4670701207385354e-05, "loss": 0.0113, "step": 4200 }, { "epoch": 0.18293014110822642, "grad_norm": 0.0633573904633522, "learning_rate": 1.46457937492691e-05, "loss": 0.0104, "step": 4210 }, { "epoch": 0.18336465450753336, "grad_norm": 0.17527233064174652, "learning_rate": 1.4620849482224996e-05, "loss": 0.0103, "step": 4220 }, { "epoch": 0.18379916790684034, "grad_norm": 0.11846215277910233, "learning_rate": 1.459586860388811e-05, "loss": 0.0107, "step": 4230 }, { "epoch": 0.18423368130614728, "grad_norm": 0.06739620864391327, "learning_rate": 1.4570851312183572e-05, "loss": 0.0081, "step": 4240 }, { "epoch": 0.18466819470545423, "grad_norm": 0.07695576548576355, "learning_rate": 1.4545797805325017e-05, "loss": 0.0087, "step": 4250 }, { "epoch": 0.18510270810476118, "grad_norm": 0.08914867043495178, "learning_rate": 1.4520708281813023e-05, "loss": 0.0113, "step": 4260 }, { "epoch": 0.18553722150406812, "grad_norm": 0.10220155119895935, "learning_rate": 1.4495582940433525e-05, "loss": 0.0086, "step": 4270 }, { "epoch": 0.1859717349033751, "grad_norm": 0.12287895381450653, "learning_rate": 1.4470421980256253e-05, "loss": 0.0121, "step": 4280 }, { "epoch": 0.18640624830268204, "grad_norm": 0.11932995170354843, "learning_rate": 1.4445225600633128e-05, "loss": 0.0088, "step": 4290 }, { "epoch": 0.186840761701989, "grad_norm": 0.08786104619503021, "learning_rate": 1.4419994001196727e-05, "loss": 0.0096, "step": 4300 }, { "epoch": 0.18727527510129593, "grad_norm": 0.13282403349876404, "learning_rate": 1.4394727381858655e-05, "loss": 0.0079, "step": 4310 }, { "epoch": 0.18770978850060288, "grad_norm": 0.10811638087034225, "learning_rate": 1.4369425942807994e-05, "loss": 0.009, "step": 4320 }, { "epoch": 0.18814430189990983, "grad_norm": 0.16801166534423828, "learning_rate": 1.4344089884509702e-05, "loss": 0.0106, "step": 4330 }, { "epoch": 0.1885788152992168, "grad_norm": 0.0781073272228241, "learning_rate": 1.4318719407703022e-05, "loss": 0.0081, "step": 4340 }, { "epoch": 0.18901332869852375, "grad_norm": 0.08049961924552917, "learning_rate": 1.4293314713399904e-05, "loss": 0.009, "step": 4350 }, { "epoch": 0.1894478420978307, "grad_norm": 0.1333671361207962, "learning_rate": 1.4267876002883406e-05, "loss": 0.0134, "step": 4360 }, { "epoch": 0.18988235549713764, "grad_norm": 0.129620760679245, "learning_rate": 1.424240347770609e-05, "loss": 0.0118, "step": 4370 }, { "epoch": 0.19031686889644459, "grad_norm": 0.09441433101892471, "learning_rate": 1.4216897339688446e-05, "loss": 0.0111, "step": 4380 }, { "epoch": 0.19075138229575153, "grad_norm": 0.08081834018230438, "learning_rate": 1.419135779091727e-05, "loss": 0.0097, "step": 4390 }, { "epoch": 0.1911858956950585, "grad_norm": 0.06669897586107254, "learning_rate": 1.4165785033744081e-05, "loss": 0.0124, "step": 4400 }, { "epoch": 0.19162040909436545, "grad_norm": 0.09304026514291763, "learning_rate": 1.4140179270783506e-05, "loss": 0.0106, "step": 4410 }, { "epoch": 0.1920549224936724, "grad_norm": 0.09239599853754044, "learning_rate": 1.4114540704911679e-05, "loss": 0.0098, "step": 4420 }, { "epoch": 0.19248943589297934, "grad_norm": 0.09333720058202744, "learning_rate": 1.4088869539264636e-05, "loss": 0.0117, "step": 4430 }, { "epoch": 0.1929239492922863, "grad_norm": 0.10393518954515457, "learning_rate": 1.40631659772367e-05, "loss": 0.0102, "step": 4440 }, { "epoch": 0.19335846269159326, "grad_norm": 0.059147026389837265, "learning_rate": 1.4037430222478876e-05, "loss": 0.0097, "step": 4450 }, { "epoch": 0.1937929760909002, "grad_norm": 0.09978174418210983, "learning_rate": 1.4011662478897239e-05, "loss": 0.0109, "step": 4460 }, { "epoch": 0.19422748949020716, "grad_norm": 0.08174576610326767, "learning_rate": 1.3985862950651296e-05, "loss": 0.0102, "step": 4470 }, { "epoch": 0.1946620028895141, "grad_norm": 0.05495322868227959, "learning_rate": 1.3960031842152404e-05, "loss": 0.0112, "step": 4480 }, { "epoch": 0.19509651628882105, "grad_norm": 0.10337530076503754, "learning_rate": 1.3934169358062128e-05, "loss": 0.0101, "step": 4490 }, { "epoch": 0.195531029688128, "grad_norm": 0.08432216197252274, "learning_rate": 1.3908275703290616e-05, "loss": 0.01, "step": 4500 }, { "epoch": 0.19596554308743497, "grad_norm": 0.10446428507566452, "learning_rate": 1.3882351082994996e-05, "loss": 0.0078, "step": 4510 }, { "epoch": 0.19640005648674191, "grad_norm": 0.1255977302789688, "learning_rate": 1.385639570257772e-05, "loss": 0.0091, "step": 4520 }, { "epoch": 0.19683456988604886, "grad_norm": 0.19239485263824463, "learning_rate": 1.3830409767684976e-05, "loss": 0.0082, "step": 4530 }, { "epoch": 0.1972690832853558, "grad_norm": 0.0951613038778305, "learning_rate": 1.380439348420502e-05, "loss": 0.0083, "step": 4540 }, { "epoch": 0.19770359668466275, "grad_norm": 0.10827091336250305, "learning_rate": 1.377834705826657e-05, "loss": 0.0084, "step": 4550 }, { "epoch": 0.19813811008396973, "grad_norm": 0.0801621600985527, "learning_rate": 1.3752270696237164e-05, "loss": 0.0072, "step": 4560 }, { "epoch": 0.19857262348327667, "grad_norm": 0.08785197138786316, "learning_rate": 1.3726164604721523e-05, "loss": 0.0094, "step": 4570 }, { "epoch": 0.19900713688258362, "grad_norm": 0.07245610654354095, "learning_rate": 1.370002899055992e-05, "loss": 0.0098, "step": 4580 }, { "epoch": 0.19944165028189056, "grad_norm": 0.12170054018497467, "learning_rate": 1.3673864060826531e-05, "loss": 0.0097, "step": 4590 }, { "epoch": 0.1998761636811975, "grad_norm": 0.09241359680891037, "learning_rate": 1.3647670022827815e-05, "loss": 0.0079, "step": 4600 }, { "epoch": 0.20031067708050446, "grad_norm": 0.08759228140115738, "learning_rate": 1.3621447084100843e-05, "loss": 0.0115, "step": 4610 }, { "epoch": 0.20074519047981143, "grad_norm": 0.09032405912876129, "learning_rate": 1.3595195452411674e-05, "loss": 0.0122, "step": 4620 }, { "epoch": 0.20117970387911838, "grad_norm": 0.05508997291326523, "learning_rate": 1.3568915335753704e-05, "loss": 0.0094, "step": 4630 }, { "epoch": 0.20161421727842532, "grad_norm": 0.05879232659935951, "learning_rate": 1.3542606942346019e-05, "loss": 0.0095, "step": 4640 }, { "epoch": 0.20204873067773227, "grad_norm": 0.09512756764888763, "learning_rate": 1.3516270480631738e-05, "loss": 0.0104, "step": 4650 }, { "epoch": 0.20248324407703922, "grad_norm": 0.06975122541189194, "learning_rate": 1.3489906159276374e-05, "loss": 0.0095, "step": 4660 }, { "epoch": 0.2029177574763462, "grad_norm": 0.08634886890649796, "learning_rate": 1.3463514187166169e-05, "loss": 0.0102, "step": 4670 }, { "epoch": 0.20335227087565313, "grad_norm": 0.1217721700668335, "learning_rate": 1.343709477340644e-05, "loss": 0.0107, "step": 4680 }, { "epoch": 0.20378678427496008, "grad_norm": 0.08943139016628265, "learning_rate": 1.3410648127319941e-05, "loss": 0.0095, "step": 4690 }, { "epoch": 0.20422129767426703, "grad_norm": 0.0762837752699852, "learning_rate": 1.3384174458445167e-05, "loss": 0.0069, "step": 4700 }, { "epoch": 0.20465581107357397, "grad_norm": 0.3115338385105133, "learning_rate": 1.335767397653474e-05, "loss": 0.0076, "step": 4710 }, { "epoch": 0.20509032447288092, "grad_norm": 0.07176689803600311, "learning_rate": 1.3331146891553708e-05, "loss": 0.0082, "step": 4720 }, { "epoch": 0.2055248378721879, "grad_norm": 0.05090424790978432, "learning_rate": 1.3304593413677893e-05, "loss": 0.0101, "step": 4730 }, { "epoch": 0.20595935127149484, "grad_norm": 0.1412171572446823, "learning_rate": 1.327801375329225e-05, "loss": 0.0078, "step": 4740 }, { "epoch": 0.20639386467080179, "grad_norm": 0.09543070197105408, "learning_rate": 1.325140812098916e-05, "loss": 0.0126, "step": 4750 }, { "epoch": 0.20682837807010873, "grad_norm": 0.1111612468957901, "learning_rate": 1.322477672756679e-05, "loss": 0.011, "step": 4760 }, { "epoch": 0.20726289146941568, "grad_norm": 0.09902411699295044, "learning_rate": 1.3198119784027415e-05, "loss": 0.0097, "step": 4770 }, { "epoch": 0.20769740486872265, "grad_norm": 0.11887188255786896, "learning_rate": 1.3171437501575739e-05, "loss": 0.0111, "step": 4780 }, { "epoch": 0.2081319182680296, "grad_norm": 0.07737130671739578, "learning_rate": 1.3144730091617235e-05, "loss": 0.0098, "step": 4790 }, { "epoch": 0.20856643166733654, "grad_norm": 0.09391709417104721, "learning_rate": 1.3117997765756455e-05, "loss": 0.0097, "step": 4800 }, { "epoch": 0.2090009450666435, "grad_norm": 0.1096915453672409, "learning_rate": 1.3091240735795372e-05, "loss": 0.0086, "step": 4810 }, { "epoch": 0.20943545846595044, "grad_norm": 0.0931084156036377, "learning_rate": 1.3064459213731679e-05, "loss": 0.008, "step": 4820 }, { "epoch": 0.20986997186525738, "grad_norm": 0.08801384270191193, "learning_rate": 1.3037653411757134e-05, "loss": 0.0096, "step": 4830 }, { "epoch": 0.21030448526456436, "grad_norm": 0.09326059371232986, "learning_rate": 1.301082354225585e-05, "loss": 0.0089, "step": 4840 }, { "epoch": 0.2107389986638713, "grad_norm": 0.08902314305305481, "learning_rate": 1.2983969817802653e-05, "loss": 0.01, "step": 4850 }, { "epoch": 0.21117351206317825, "grad_norm": 0.06725908070802689, "learning_rate": 1.2957092451161344e-05, "loss": 0.0076, "step": 4860 }, { "epoch": 0.2116080254624852, "grad_norm": 0.10939764231443405, "learning_rate": 1.293019165528307e-05, "loss": 0.0094, "step": 4870 }, { "epoch": 0.21204253886179214, "grad_norm": 0.0729513093829155, "learning_rate": 1.2903267643304588e-05, "loss": 0.0101, "step": 4880 }, { "epoch": 0.21247705226109911, "grad_norm": 0.09630395472049713, "learning_rate": 1.2876320628546608e-05, "loss": 0.0087, "step": 4890 }, { "epoch": 0.21291156566040606, "grad_norm": 0.0898730531334877, "learning_rate": 1.2849350824512097e-05, "loss": 0.0091, "step": 4900 }, { "epoch": 0.213346079059713, "grad_norm": 0.07306133955717087, "learning_rate": 1.2822358444884568e-05, "loss": 0.0083, "step": 4910 }, { "epoch": 0.21378059245901995, "grad_norm": 0.04950597882270813, "learning_rate": 1.279534370352642e-05, "loss": 0.0081, "step": 4920 }, { "epoch": 0.2142151058583269, "grad_norm": 0.06085862219333649, "learning_rate": 1.276830681447721e-05, "loss": 0.0076, "step": 4930 }, { "epoch": 0.21464961925763384, "grad_norm": 0.1412172168493271, "learning_rate": 1.2741247991951976e-05, "loss": 0.0093, "step": 4940 }, { "epoch": 0.21508413265694082, "grad_norm": 0.09206506609916687, "learning_rate": 1.2714167450339551e-05, "loss": 0.0107, "step": 4950 }, { "epoch": 0.21551864605624776, "grad_norm": 0.10359804332256317, "learning_rate": 1.268706540420083e-05, "loss": 0.0087, "step": 4960 }, { "epoch": 0.2159531594555547, "grad_norm": 0.08671776205301285, "learning_rate": 1.2659942068267097e-05, "loss": 0.0067, "step": 4970 }, { "epoch": 0.21638767285486166, "grad_norm": 0.0956655815243721, "learning_rate": 1.2632797657438317e-05, "loss": 0.0087, "step": 4980 }, { "epoch": 0.2168221862541686, "grad_norm": 0.08012635260820389, "learning_rate": 1.2605632386781442e-05, "loss": 0.0093, "step": 4990 }, { "epoch": 0.21725669965347558, "grad_norm": 0.4396616220474243, "learning_rate": 1.2578446471528678e-05, "loss": 0.01, "step": 5000 }, { "epoch": 0.21769121305278252, "grad_norm": 0.0880979672074318, "learning_rate": 1.2551240127075815e-05, "loss": 0.01, "step": 5010 }, { "epoch": 0.21812572645208947, "grad_norm": 0.1420845091342926, "learning_rate": 1.2524013568980496e-05, "loss": 0.0111, "step": 5020 }, { "epoch": 0.21856023985139642, "grad_norm": 0.08724194765090942, "learning_rate": 1.249676701296053e-05, "loss": 0.009, "step": 5030 }, { "epoch": 0.21899475325070336, "grad_norm": 0.11142345517873764, "learning_rate": 1.2469500674892159e-05, "loss": 0.0082, "step": 5040 }, { "epoch": 0.2194292666500103, "grad_norm": 0.0890979990363121, "learning_rate": 1.244221477080836e-05, "loss": 0.0091, "step": 5050 }, { "epoch": 0.21986378004931728, "grad_norm": 0.05859982222318649, "learning_rate": 1.2414909516897145e-05, "loss": 0.0073, "step": 5060 }, { "epoch": 0.22029829344862423, "grad_norm": 0.19758309423923492, "learning_rate": 1.2387585129499815e-05, "loss": 0.0092, "step": 5070 }, { "epoch": 0.22073280684793117, "grad_norm": 0.0901278406381607, "learning_rate": 1.2360241825109293e-05, "loss": 0.0092, "step": 5080 }, { "epoch": 0.22116732024723812, "grad_norm": 0.09502365440130234, "learning_rate": 1.2332879820368358e-05, "loss": 0.0098, "step": 5090 }, { "epoch": 0.22160183364654507, "grad_norm": 0.0769268274307251, "learning_rate": 1.2305499332067967e-05, "loss": 0.0079, "step": 5100 }, { "epoch": 0.22203634704585204, "grad_norm": 0.09428519010543823, "learning_rate": 1.2278100577145526e-05, "loss": 0.009, "step": 5110 }, { "epoch": 0.22247086044515899, "grad_norm": 0.07567259669303894, "learning_rate": 1.2250683772683151e-05, "loss": 0.0105, "step": 5120 }, { "epoch": 0.22290537384446593, "grad_norm": 0.095923513174057, "learning_rate": 1.222324913590599e-05, "loss": 0.013, "step": 5130 }, { "epoch": 0.22333988724377288, "grad_norm": 0.09458699822425842, "learning_rate": 1.2195796884180458e-05, "loss": 0.0092, "step": 5140 }, { "epoch": 0.22377440064307982, "grad_norm": 0.09393813461065292, "learning_rate": 1.2168327235012544e-05, "loss": 0.0078, "step": 5150 }, { "epoch": 0.22420891404238677, "grad_norm": 0.12188570201396942, "learning_rate": 1.2140840406046075e-05, "loss": 0.0089, "step": 5160 }, { "epoch": 0.22464342744169374, "grad_norm": 0.08482064306735992, "learning_rate": 1.2113336615060996e-05, "loss": 0.0076, "step": 5170 }, { "epoch": 0.2250779408410007, "grad_norm": 0.1528569608926773, "learning_rate": 1.2085816079971639e-05, "loss": 0.0092, "step": 5180 }, { "epoch": 0.22551245424030764, "grad_norm": 0.07889417558908463, "learning_rate": 1.205827901882501e-05, "loss": 0.008, "step": 5190 }, { "epoch": 0.22594696763961458, "grad_norm": 0.10469748824834824, "learning_rate": 1.2030725649799043e-05, "loss": 0.01, "step": 5200 }, { "epoch": 0.22638148103892153, "grad_norm": 0.13136431574821472, "learning_rate": 1.2003156191200885e-05, "loss": 0.0078, "step": 5210 }, { "epoch": 0.2268159944382285, "grad_norm": 0.1373455971479416, "learning_rate": 1.1975570861465156e-05, "loss": 0.0108, "step": 5220 }, { "epoch": 0.22725050783753545, "grad_norm": 0.051384877413511276, "learning_rate": 1.194796987915223e-05, "loss": 0.01, "step": 5230 }, { "epoch": 0.2276850212368424, "grad_norm": 0.046507228165864944, "learning_rate": 1.1920353462946503e-05, "loss": 0.0109, "step": 5240 }, { "epoch": 0.22811953463614934, "grad_norm": 0.1168011799454689, "learning_rate": 1.1892721831654638e-05, "loss": 0.0082, "step": 5250 }, { "epoch": 0.2285540480354563, "grad_norm": 0.06604877859354019, "learning_rate": 1.1865075204203866e-05, "loss": 0.007, "step": 5260 }, { "epoch": 0.22898856143476323, "grad_norm": 0.09276457130908966, "learning_rate": 1.1837413799640216e-05, "loss": 0.0083, "step": 5270 }, { "epoch": 0.2294230748340702, "grad_norm": 0.11472596973180771, "learning_rate": 1.1809737837126812e-05, "loss": 0.0091, "step": 5280 }, { "epoch": 0.22985758823337715, "grad_norm": 0.0845196470618248, "learning_rate": 1.1782047535942117e-05, "loss": 0.0093, "step": 5290 }, { "epoch": 0.2302921016326841, "grad_norm": 0.11708565801382065, "learning_rate": 1.1754343115478193e-05, "loss": 0.0088, "step": 5300 }, { "epoch": 0.23072661503199104, "grad_norm": 0.12564682960510254, "learning_rate": 1.1726624795238981e-05, "loss": 0.009, "step": 5310 }, { "epoch": 0.231161128431298, "grad_norm": 0.08443303406238556, "learning_rate": 1.1698892794838546e-05, "loss": 0.0087, "step": 5320 }, { "epoch": 0.23159564183060496, "grad_norm": 0.08368808776140213, "learning_rate": 1.167114733399934e-05, "loss": 0.0087, "step": 5330 }, { "epoch": 0.2320301552299119, "grad_norm": 0.06150970607995987, "learning_rate": 1.1643388632550468e-05, "loss": 0.0083, "step": 5340 }, { "epoch": 0.23246466862921886, "grad_norm": 0.06667175889015198, "learning_rate": 1.1615616910425928e-05, "loss": 0.0084, "step": 5350 }, { "epoch": 0.2328991820285258, "grad_norm": 0.10415839403867722, "learning_rate": 1.15878323876629e-05, "loss": 0.0084, "step": 5360 }, { "epoch": 0.23333369542783275, "grad_norm": 0.06974676251411438, "learning_rate": 1.1560035284399977e-05, "loss": 0.0101, "step": 5370 }, { "epoch": 0.2337682088271397, "grad_norm": 0.07754170894622803, "learning_rate": 1.1532225820875422e-05, "loss": 0.009, "step": 5380 }, { "epoch": 0.23420272222644667, "grad_norm": 0.09246724098920822, "learning_rate": 1.1504404217425438e-05, "loss": 0.0104, "step": 5390 }, { "epoch": 0.23463723562575362, "grad_norm": 0.11340566724538803, "learning_rate": 1.1476570694482406e-05, "loss": 0.0099, "step": 5400 }, { "epoch": 0.23507174902506056, "grad_norm": 0.05798724293708801, "learning_rate": 1.1448725472573145e-05, "loss": 0.0104, "step": 5410 }, { "epoch": 0.2355062624243675, "grad_norm": 0.07213063538074493, "learning_rate": 1.1420868772317184e-05, "loss": 0.0101, "step": 5420 }, { "epoch": 0.23594077582367445, "grad_norm": 0.10362295061349869, "learning_rate": 1.1393000814424973e-05, "loss": 0.0098, "step": 5430 }, { "epoch": 0.23637528922298143, "grad_norm": 0.07319142669439316, "learning_rate": 1.1365121819696163e-05, "loss": 0.0097, "step": 5440 }, { "epoch": 0.23680980262228837, "grad_norm": 0.13031233847141266, "learning_rate": 1.1337232009017858e-05, "loss": 0.0107, "step": 5450 }, { "epoch": 0.23724431602159532, "grad_norm": 0.10126952081918716, "learning_rate": 1.130933160336285e-05, "loss": 0.0107, "step": 5460 }, { "epoch": 0.23767882942090227, "grad_norm": 0.13342677056789398, "learning_rate": 1.1281420823787883e-05, "loss": 0.0098, "step": 5470 }, { "epoch": 0.2381133428202092, "grad_norm": 0.257994145154953, "learning_rate": 1.1253499891431882e-05, "loss": 0.0073, "step": 5480 }, { "epoch": 0.23854785621951616, "grad_norm": 0.12316708266735077, "learning_rate": 1.1225569027514229e-05, "loss": 0.0071, "step": 5490 }, { "epoch": 0.23898236961882313, "grad_norm": 0.12356661260128021, "learning_rate": 1.1197628453332986e-05, "loss": 0.0092, "step": 5500 }, { "epoch": 0.23941688301813008, "grad_norm": 0.05779719725251198, "learning_rate": 1.1169678390263143e-05, "loss": 0.0103, "step": 5510 }, { "epoch": 0.23985139641743702, "grad_norm": 0.08832740038633347, "learning_rate": 1.1141719059754884e-05, "loss": 0.0083, "step": 5520 }, { "epoch": 0.24028590981674397, "grad_norm": 0.11366628110408783, "learning_rate": 1.1113750683331813e-05, "loss": 0.0107, "step": 5530 }, { "epoch": 0.24072042321605092, "grad_norm": 0.08835666626691818, "learning_rate": 1.1085773482589206e-05, "loss": 0.0091, "step": 5540 }, { "epoch": 0.2411549366153579, "grad_norm": 0.04762517660856247, "learning_rate": 1.1057787679192256e-05, "loss": 0.0058, "step": 5550 }, { "epoch": 0.24158945001466484, "grad_norm": 0.1362943947315216, "learning_rate": 1.1029793494874312e-05, "loss": 0.0107, "step": 5560 }, { "epoch": 0.24202396341397178, "grad_norm": 0.0849776417016983, "learning_rate": 1.1001791151435131e-05, "loss": 0.0093, "step": 5570 }, { "epoch": 0.24245847681327873, "grad_norm": 0.047240015119314194, "learning_rate": 1.0973780870739111e-05, "loss": 0.0068, "step": 5580 }, { "epoch": 0.24289299021258567, "grad_norm": 0.07293111085891724, "learning_rate": 1.0945762874713537e-05, "loss": 0.0075, "step": 5590 }, { "epoch": 0.24332750361189262, "grad_norm": 0.08692555129528046, "learning_rate": 1.0917737385346828e-05, "loss": 0.006, "step": 5600 }, { "epoch": 0.2437620170111996, "grad_norm": 0.13119547069072723, "learning_rate": 1.0889704624686766e-05, "loss": 0.0098, "step": 5610 }, { "epoch": 0.24419653041050654, "grad_norm": 0.14227084815502167, "learning_rate": 1.0861664814838747e-05, "loss": 0.008, "step": 5620 }, { "epoch": 0.2446310438098135, "grad_norm": 0.06881032884120941, "learning_rate": 1.083361817796403e-05, "loss": 0.0113, "step": 5630 }, { "epoch": 0.24506555720912043, "grad_norm": 0.12126508355140686, "learning_rate": 1.0805564936277936e-05, "loss": 0.0078, "step": 5640 }, { "epoch": 0.24550007060842738, "grad_norm": 0.10700210928916931, "learning_rate": 1.0777505312048152e-05, "loss": 0.0101, "step": 5650 }, { "epoch": 0.24593458400773432, "grad_norm": 0.08859797567129135, "learning_rate": 1.0749439527592909e-05, "loss": 0.0115, "step": 5660 }, { "epoch": 0.2463690974070413, "grad_norm": 0.10005716234445572, "learning_rate": 1.0721367805279251e-05, "loss": 0.0087, "step": 5670 }, { "epoch": 0.24680361080634824, "grad_norm": 0.06850054860115051, "learning_rate": 1.0693290367521276e-05, "loss": 0.0071, "step": 5680 }, { "epoch": 0.2472381242056552, "grad_norm": 0.06010816991329193, "learning_rate": 1.0665207436778353e-05, "loss": 0.008, "step": 5690 }, { "epoch": 0.24767263760496214, "grad_norm": 0.11750059574842453, "learning_rate": 1.0637119235553388e-05, "loss": 0.0088, "step": 5700 }, { "epoch": 0.24810715100426908, "grad_norm": 0.09331734478473663, "learning_rate": 1.0609025986391032e-05, "loss": 0.0093, "step": 5710 }, { "epoch": 0.24854166440357606, "grad_norm": 0.06403643637895584, "learning_rate": 1.0580927911875938e-05, "loss": 0.0066, "step": 5720 }, { "epoch": 0.248976177802883, "grad_norm": 0.07363591343164444, "learning_rate": 1.055282523463099e-05, "loss": 0.0084, "step": 5730 }, { "epoch": 0.24941069120218995, "grad_norm": 0.08989038318395615, "learning_rate": 1.0524718177315536e-05, "loss": 0.0064, "step": 5740 }, { "epoch": 0.2498452046014969, "grad_norm": 0.12647745013237, "learning_rate": 1.0496606962623632e-05, "loss": 0.0088, "step": 5750 }, { "epoch": 0.25027971800080384, "grad_norm": 0.07326851785182953, "learning_rate": 1.0468491813282269e-05, "loss": 0.0079, "step": 5760 }, { "epoch": 0.2507142314001108, "grad_norm": 0.0844263806939125, "learning_rate": 1.0440372952049618e-05, "loss": 0.0078, "step": 5770 }, { "epoch": 0.25114874479941773, "grad_norm": 0.0971388965845108, "learning_rate": 1.0412250601713254e-05, "loss": 0.007, "step": 5780 }, { "epoch": 0.2515832581987247, "grad_norm": 0.10897351801395416, "learning_rate": 1.03841249850884e-05, "loss": 0.01, "step": 5790 }, { "epoch": 0.2520177715980317, "grad_norm": 0.07884043455123901, "learning_rate": 1.0355996325016152e-05, "loss": 0.0112, "step": 5800 }, { "epoch": 0.2524522849973386, "grad_norm": 0.08184106647968292, "learning_rate": 1.0327864844361735e-05, "loss": 0.0064, "step": 5810 }, { "epoch": 0.2528867983966456, "grad_norm": 0.10301776975393295, "learning_rate": 1.02997307660127e-05, "loss": 0.0081, "step": 5820 }, { "epoch": 0.2533213117959525, "grad_norm": 0.1618221551179886, "learning_rate": 1.0271594312877196e-05, "loss": 0.0091, "step": 5830 }, { "epoch": 0.25375582519525947, "grad_norm": 0.07037736475467682, "learning_rate": 1.024345570788218e-05, "loss": 0.0082, "step": 5840 }, { "epoch": 0.2541903385945664, "grad_norm": 0.13534654676914215, "learning_rate": 1.0215315173971661e-05, "loss": 0.0097, "step": 5850 }, { "epoch": 0.25462485199387336, "grad_norm": 0.11499059200286865, "learning_rate": 1.0187172934104934e-05, "loss": 0.0089, "step": 5860 }, { "epoch": 0.2550593653931803, "grad_norm": 0.09554502367973328, "learning_rate": 1.01590292112548e-05, "loss": 0.0101, "step": 5870 }, { "epoch": 0.25549387879248725, "grad_norm": 0.07305638492107391, "learning_rate": 1.013088422840582e-05, "loss": 0.0071, "step": 5880 }, { "epoch": 0.2559283921917942, "grad_norm": 0.06403714418411255, "learning_rate": 1.0102738208552535e-05, "loss": 0.0089, "step": 5890 }, { "epoch": 0.25636290559110114, "grad_norm": 0.06630440801382065, "learning_rate": 1.0074591374697701e-05, "loss": 0.0073, "step": 5900 }, { "epoch": 0.25679741899040814, "grad_norm": 0.05963248759508133, "learning_rate": 1.0046443949850531e-05, "loss": 0.0101, "step": 5910 }, { "epoch": 0.2572319323897151, "grad_norm": 0.06275281310081482, "learning_rate": 1.00182961570249e-05, "loss": 0.0078, "step": 5920 }, { "epoch": 0.25766644578902204, "grad_norm": 0.0786055400967598, "learning_rate": 9.990148219237623e-06, "loss": 0.0078, "step": 5930 }, { "epoch": 0.258100959188329, "grad_norm": 0.09199898689985275, "learning_rate": 9.96200035950665e-06, "loss": 0.0071, "step": 5940 }, { "epoch": 0.25853547258763593, "grad_norm": 0.06566759198904037, "learning_rate": 9.933852800849311e-06, "loss": 0.0073, "step": 5950 }, { "epoch": 0.2589699859869429, "grad_norm": 0.1568640023469925, "learning_rate": 9.905705766280564e-06, "loss": 0.008, "step": 5960 }, { "epoch": 0.2594044993862498, "grad_norm": 0.06404092907905579, "learning_rate": 9.877559478811199e-06, "loss": 0.0066, "step": 5970 }, { "epoch": 0.25983901278555677, "grad_norm": 0.08680085837841034, "learning_rate": 9.849414161446093e-06, "loss": 0.0074, "step": 5980 }, { "epoch": 0.2602735261848637, "grad_norm": 0.11825085431337357, "learning_rate": 9.821270037182442e-06, "loss": 0.0089, "step": 5990 }, { "epoch": 0.26070803958417066, "grad_norm": 0.09618745744228363, "learning_rate": 9.793127329007973e-06, "loss": 0.0087, "step": 6000 }, { "epoch": 0.2611425529834776, "grad_norm": 0.05515113100409508, "learning_rate": 9.76498625989922e-06, "loss": 0.0061, "step": 6010 }, { "epoch": 0.2615770663827846, "grad_norm": 0.10781148821115494, "learning_rate": 9.736847052819704e-06, "loss": 0.0094, "step": 6020 }, { "epoch": 0.26201157978209155, "grad_norm": 0.0913834422826767, "learning_rate": 9.708709930718204e-06, "loss": 0.0087, "step": 6030 }, { "epoch": 0.2624460931813985, "grad_norm": 0.07052706182003021, "learning_rate": 9.680575116526982e-06, "loss": 0.0081, "step": 6040 }, { "epoch": 0.26288060658070544, "grad_norm": 0.08302155137062073, "learning_rate": 9.652442833160012e-06, "loss": 0.0078, "step": 6050 }, { "epoch": 0.2633151199800124, "grad_norm": 0.157508984208107, "learning_rate": 9.624313303511218e-06, "loss": 0.0091, "step": 6060 }, { "epoch": 0.26374963337931934, "grad_norm": 0.07369590550661087, "learning_rate": 9.5961867504527e-06, "loss": 0.0069, "step": 6070 }, { "epoch": 0.2641841467786263, "grad_norm": 0.07717025279998779, "learning_rate": 9.568063396832979e-06, "loss": 0.0073, "step": 6080 }, { "epoch": 0.26461866017793323, "grad_norm": 0.07593873143196106, "learning_rate": 9.539943465475224e-06, "loss": 0.0089, "step": 6090 }, { "epoch": 0.2650531735772402, "grad_norm": 0.10739245265722275, "learning_rate": 9.511827179175496e-06, "loss": 0.0082, "step": 6100 }, { "epoch": 0.2654876869765471, "grad_norm": 0.07673447579145432, "learning_rate": 9.483714760700968e-06, "loss": 0.0094, "step": 6110 }, { "epoch": 0.26592220037585407, "grad_norm": 0.08048778027296066, "learning_rate": 9.455606432788172e-06, "loss": 0.0073, "step": 6120 }, { "epoch": 0.26635671377516107, "grad_norm": 0.06304773688316345, "learning_rate": 9.427502418141228e-06, "loss": 0.0069, "step": 6130 }, { "epoch": 0.266791227174468, "grad_norm": 0.10460677742958069, "learning_rate": 9.399402939430078e-06, "loss": 0.0094, "step": 6140 }, { "epoch": 0.26722574057377496, "grad_norm": 0.09594422578811646, "learning_rate": 9.371308219288739e-06, "loss": 0.0086, "step": 6150 }, { "epoch": 0.2676602539730819, "grad_norm": 0.07144231349229813, "learning_rate": 9.343218480313514e-06, "loss": 0.0084, "step": 6160 }, { "epoch": 0.26809476737238885, "grad_norm": 0.05371937155723572, "learning_rate": 9.315133945061243e-06, "loss": 0.009, "step": 6170 }, { "epoch": 0.2685292807716958, "grad_norm": 0.14217378199100494, "learning_rate": 9.287054836047532e-06, "loss": 0.0105, "step": 6180 }, { "epoch": 0.26896379417100275, "grad_norm": 0.09040679782629013, "learning_rate": 9.258981375745005e-06, "loss": 0.0092, "step": 6190 }, { "epoch": 0.2693983075703097, "grad_norm": 0.11461376398801804, "learning_rate": 9.230913786581523e-06, "loss": 0.0066, "step": 6200 }, { "epoch": 0.26983282096961664, "grad_norm": 0.09241148829460144, "learning_rate": 9.20285229093843e-06, "loss": 0.0091, "step": 6210 }, { "epoch": 0.2702673343689236, "grad_norm": 0.1639711558818817, "learning_rate": 9.174797111148792e-06, "loss": 0.0078, "step": 6220 }, { "epoch": 0.27070184776823053, "grad_norm": 0.10276318341493607, "learning_rate": 9.146748469495632e-06, "loss": 0.0078, "step": 6230 }, { "epoch": 0.2711363611675375, "grad_norm": 0.0878051146864891, "learning_rate": 9.11870658821018e-06, "loss": 0.008, "step": 6240 }, { "epoch": 0.2715708745668445, "grad_norm": 0.12146293371915817, "learning_rate": 9.090671689470092e-06, "loss": 0.0071, "step": 6250 }, { "epoch": 0.2720053879661514, "grad_norm": 0.07059358060359955, "learning_rate": 9.062643995397705e-06, "loss": 0.0071, "step": 6260 }, { "epoch": 0.27243990136545837, "grad_norm": 0.11675061285495758, "learning_rate": 9.034623728058269e-06, "loss": 0.0062, "step": 6270 }, { "epoch": 0.2728744147647653, "grad_norm": 0.07080376893281937, "learning_rate": 9.006611109458201e-06, "loss": 0.0084, "step": 6280 }, { "epoch": 0.27330892816407226, "grad_norm": 0.10039542615413666, "learning_rate": 8.97860636154331e-06, "loss": 0.0091, "step": 6290 }, { "epoch": 0.2737434415633792, "grad_norm": 0.1331404149532318, "learning_rate": 8.950609706197048e-06, "loss": 0.0101, "step": 6300 }, { "epoch": 0.27417795496268615, "grad_norm": 0.09555957466363907, "learning_rate": 8.922621365238742e-06, "loss": 0.0093, "step": 6310 }, { "epoch": 0.2746124683619931, "grad_norm": 0.10700193792581558, "learning_rate": 8.89464156042185e-06, "loss": 0.0081, "step": 6320 }, { "epoch": 0.27504698176130005, "grad_norm": 0.10304276645183563, "learning_rate": 8.8666705134322e-06, "loss": 0.0111, "step": 6330 }, { "epoch": 0.275481495160607, "grad_norm": 0.07287819683551788, "learning_rate": 8.838708445886223e-06, "loss": 0.0089, "step": 6340 }, { "epoch": 0.27591600855991394, "grad_norm": 0.06307504326105118, "learning_rate": 8.810755579329213e-06, "loss": 0.0065, "step": 6350 }, { "epoch": 0.27635052195922094, "grad_norm": 0.06355910003185272, "learning_rate": 8.782812135233556e-06, "loss": 0.0065, "step": 6360 }, { "epoch": 0.2767850353585279, "grad_norm": 0.06679730117321014, "learning_rate": 8.754878334996995e-06, "loss": 0.0081, "step": 6370 }, { "epoch": 0.27721954875783483, "grad_norm": 0.07842592149972916, "learning_rate": 8.726954399940855e-06, "loss": 0.0088, "step": 6380 }, { "epoch": 0.2776540621571418, "grad_norm": 0.048519112169742584, "learning_rate": 8.699040551308296e-06, "loss": 0.0109, "step": 6390 }, { "epoch": 0.2780885755564487, "grad_norm": 0.07662378251552582, "learning_rate": 8.671137010262568e-06, "loss": 0.007, "step": 6400 }, { "epoch": 0.27852308895575567, "grad_norm": 0.07653367519378662, "learning_rate": 8.643243997885253e-06, "loss": 0.0075, "step": 6410 }, { "epoch": 0.2789576023550626, "grad_norm": 0.050618138164281845, "learning_rate": 8.615361735174517e-06, "loss": 0.0087, "step": 6420 }, { "epoch": 0.27939211575436956, "grad_norm": 0.0792509987950325, "learning_rate": 8.58749044304335e-06, "loss": 0.0084, "step": 6430 }, { "epoch": 0.2798266291536765, "grad_norm": 0.05023474618792534, "learning_rate": 8.559630342317822e-06, "loss": 0.0058, "step": 6440 }, { "epoch": 0.28026114255298346, "grad_norm": 0.08565095067024231, "learning_rate": 8.531781653735334e-06, "loss": 0.0117, "step": 6450 }, { "epoch": 0.2806956559522904, "grad_norm": 0.09438503533601761, "learning_rate": 8.503944597942865e-06, "loss": 0.0088, "step": 6460 }, { "epoch": 0.2811301693515974, "grad_norm": 0.12111536413431168, "learning_rate": 8.476119395495235e-06, "loss": 0.0094, "step": 6470 }, { "epoch": 0.28156468275090435, "grad_norm": 0.0618109367787838, "learning_rate": 8.44830626685334e-06, "loss": 0.0079, "step": 6480 }, { "epoch": 0.2819991961502113, "grad_norm": 0.07699032127857208, "learning_rate": 8.42050543238242e-06, "loss": 0.0087, "step": 6490 }, { "epoch": 0.28243370954951824, "grad_norm": 0.1489276885986328, "learning_rate": 8.392717112350301e-06, "loss": 0.0108, "step": 6500 }, { "epoch": 0.2828682229488252, "grad_norm": 0.1276170313358307, "learning_rate": 8.364941526925667e-06, "loss": 0.0082, "step": 6510 }, { "epoch": 0.28330273634813213, "grad_norm": 0.11442344635725021, "learning_rate": 8.337178896176295e-06, "loss": 0.0071, "step": 6520 }, { "epoch": 0.2837372497474391, "grad_norm": 0.07063666731119156, "learning_rate": 8.309429440067324e-06, "loss": 0.0094, "step": 6530 }, { "epoch": 0.284171763146746, "grad_norm": 0.06442690640687943, "learning_rate": 8.281693378459516e-06, "loss": 0.0072, "step": 6540 }, { "epoch": 0.28460627654605297, "grad_norm": 0.09958231449127197, "learning_rate": 8.253970931107492e-06, "loss": 0.0089, "step": 6550 }, { "epoch": 0.2850407899453599, "grad_norm": 0.06900348514318466, "learning_rate": 8.226262317658027e-06, "loss": 0.0093, "step": 6560 }, { "epoch": 0.28547530334466686, "grad_norm": 0.06914185732603073, "learning_rate": 8.198567757648272e-06, "loss": 0.0084, "step": 6570 }, { "epoch": 0.28590981674397387, "grad_norm": 0.08540336787700653, "learning_rate": 8.170887470504038e-06, "loss": 0.0092, "step": 6580 }, { "epoch": 0.2863443301432808, "grad_norm": 0.05744350329041481, "learning_rate": 8.143221675538053e-06, "loss": 0.0084, "step": 6590 }, { "epoch": 0.28677884354258776, "grad_norm": 0.0658305436372757, "learning_rate": 8.115570591948222e-06, "loss": 0.0081, "step": 6600 }, { "epoch": 0.2872133569418947, "grad_norm": 0.06954361498355865, "learning_rate": 8.087934438815888e-06, "loss": 0.0079, "step": 6610 }, { "epoch": 0.28764787034120165, "grad_norm": 0.0778515487909317, "learning_rate": 8.0603134351041e-06, "loss": 0.0085, "step": 6620 }, { "epoch": 0.2880823837405086, "grad_norm": 0.030152561143040657, "learning_rate": 8.032707799655876e-06, "loss": 0.0075, "step": 6630 }, { "epoch": 0.28851689713981554, "grad_norm": 0.09243931621313095, "learning_rate": 8.005117751192472e-06, "loss": 0.0069, "step": 6640 }, { "epoch": 0.2889514105391225, "grad_norm": 0.05257497727870941, "learning_rate": 7.977543508311653e-06, "loss": 0.0069, "step": 6650 }, { "epoch": 0.28938592393842943, "grad_norm": 0.1092933937907219, "learning_rate": 7.949985289485945e-06, "loss": 0.0089, "step": 6660 }, { "epoch": 0.2898204373377364, "grad_norm": 0.08264698833227158, "learning_rate": 7.922443313060919e-06, "loss": 0.0078, "step": 6670 }, { "epoch": 0.2902549507370433, "grad_norm": 0.07861427962779999, "learning_rate": 7.894917797253452e-06, "loss": 0.0088, "step": 6680 }, { "epoch": 0.29068946413635033, "grad_norm": 0.08720831573009491, "learning_rate": 7.867408960150015e-06, "loss": 0.0092, "step": 6690 }, { "epoch": 0.2911239775356573, "grad_norm": 0.10889329016208649, "learning_rate": 7.839917019704921e-06, "loss": 0.0089, "step": 6700 }, { "epoch": 0.2915584909349642, "grad_norm": 0.10911451280117035, "learning_rate": 7.812442193738612e-06, "loss": 0.0077, "step": 6710 }, { "epoch": 0.29199300433427117, "grad_norm": 0.09256261587142944, "learning_rate": 7.784984699935934e-06, "loss": 0.0065, "step": 6720 }, { "epoch": 0.2924275177335781, "grad_norm": 0.06900645047426224, "learning_rate": 7.7575447558444e-06, "loss": 0.0082, "step": 6730 }, { "epoch": 0.29286203113288506, "grad_norm": 0.08416484296321869, "learning_rate": 7.730122578872492e-06, "loss": 0.0087, "step": 6740 }, { "epoch": 0.293296544532192, "grad_norm": 0.0650983676314354, "learning_rate": 7.702718386287904e-06, "loss": 0.0064, "step": 6750 }, { "epoch": 0.29373105793149895, "grad_norm": 0.05348186567425728, "learning_rate": 7.675332395215853e-06, "loss": 0.0089, "step": 6760 }, { "epoch": 0.2941655713308059, "grad_norm": 0.12326699495315552, "learning_rate": 7.64796482263734e-06, "loss": 0.0087, "step": 6770 }, { "epoch": 0.29460008473011284, "grad_norm": 0.07229342311620712, "learning_rate": 7.620615885387419e-06, "loss": 0.006, "step": 6780 }, { "epoch": 0.2950345981294198, "grad_norm": 0.11028210073709488, "learning_rate": 7.593285800153527e-06, "loss": 0.0069, "step": 6790 }, { "epoch": 0.2954691115287268, "grad_norm": 0.08311082422733307, "learning_rate": 7.565974783473709e-06, "loss": 0.0094, "step": 6800 }, { "epoch": 0.29590362492803374, "grad_norm": 0.06337704509496689, "learning_rate": 7.5386830517349366e-06, "loss": 0.0075, "step": 6810 }, { "epoch": 0.2963381383273407, "grad_norm": 0.09829442203044891, "learning_rate": 7.511410821171385e-06, "loss": 0.0081, "step": 6820 }, { "epoch": 0.29677265172664763, "grad_norm": 0.06810601055622101, "learning_rate": 7.484158307862726e-06, "loss": 0.008, "step": 6830 }, { "epoch": 0.2972071651259546, "grad_norm": 0.11772483587265015, "learning_rate": 7.4569257277324035e-06, "loss": 0.0087, "step": 6840 }, { "epoch": 0.2976416785252615, "grad_norm": 0.08689630776643753, "learning_rate": 7.429713296545934e-06, "loss": 0.0072, "step": 6850 }, { "epoch": 0.29807619192456847, "grad_norm": 0.06928461790084839, "learning_rate": 7.402521229909185e-06, "loss": 0.01, "step": 6860 }, { "epoch": 0.2985107053238754, "grad_norm": 0.06894993036985397, "learning_rate": 7.37534974326668e-06, "loss": 0.0061, "step": 6870 }, { "epoch": 0.29894521872318236, "grad_norm": 0.06853579729795456, "learning_rate": 7.3481990518998915e-06, "loss": 0.0064, "step": 6880 }, { "epoch": 0.2993797321224893, "grad_norm": 0.11018829047679901, "learning_rate": 7.321069370925519e-06, "loss": 0.009, "step": 6890 }, { "epoch": 0.29981424552179625, "grad_norm": 0.0806964859366417, "learning_rate": 7.293960915293803e-06, "loss": 0.0092, "step": 6900 }, { "epoch": 0.30024875892110325, "grad_norm": 0.08158484101295471, "learning_rate": 7.266873899786803e-06, "loss": 0.007, "step": 6910 }, { "epoch": 0.3006832723204102, "grad_norm": 0.04432486370205879, "learning_rate": 7.2398085390167275e-06, "loss": 0.0063, "step": 6920 }, { "epoch": 0.30111778571971715, "grad_norm": 0.08704327791929245, "learning_rate": 7.212765047424191e-06, "loss": 0.0075, "step": 6930 }, { "epoch": 0.3015522991190241, "grad_norm": 0.10205797106027603, "learning_rate": 7.185743639276552e-06, "loss": 0.0089, "step": 6940 }, { "epoch": 0.30198681251833104, "grad_norm": 0.08234458416700363, "learning_rate": 7.158744528666196e-06, "loss": 0.0066, "step": 6950 }, { "epoch": 0.302421325917638, "grad_norm": 0.14198225736618042, "learning_rate": 7.131767929508833e-06, "loss": 0.0082, "step": 6960 }, { "epoch": 0.30285583931694493, "grad_norm": 0.06368236243724823, "learning_rate": 7.104814055541838e-06, "loss": 0.0071, "step": 6970 }, { "epoch": 0.3032903527162519, "grad_norm": 0.12608693540096283, "learning_rate": 7.077883120322507e-06, "loss": 0.0091, "step": 6980 }, { "epoch": 0.3037248661155588, "grad_norm": 0.07397144287824631, "learning_rate": 7.0509753372264065e-06, "loss": 0.0083, "step": 6990 }, { "epoch": 0.30415937951486577, "grad_norm": 0.07896910607814789, "learning_rate": 7.024090919445662e-06, "loss": 0.0086, "step": 7000 }, { "epoch": 0.3045938929141727, "grad_norm": 0.06012195348739624, "learning_rate": 6.997230079987272e-06, "loss": 0.0115, "step": 7010 }, { "epoch": 0.3050284063134797, "grad_norm": 0.12889036536216736, "learning_rate": 6.970393031671428e-06, "loss": 0.0077, "step": 7020 }, { "epoch": 0.30546291971278666, "grad_norm": 0.08058907091617584, "learning_rate": 6.943579987129822e-06, "loss": 0.009, "step": 7030 }, { "epoch": 0.3058974331120936, "grad_norm": 0.10182671993970871, "learning_rate": 6.916791158803954e-06, "loss": 0.0072, "step": 7040 }, { "epoch": 0.30633194651140055, "grad_norm": 0.08950523287057877, "learning_rate": 6.890026758943464e-06, "loss": 0.0074, "step": 7050 }, { "epoch": 0.3067664599107075, "grad_norm": 0.05128556489944458, "learning_rate": 6.86328699960445e-06, "loss": 0.0091, "step": 7060 }, { "epoch": 0.30720097331001445, "grad_norm": 0.11177459359169006, "learning_rate": 6.83657209264777e-06, "loss": 0.0067, "step": 7070 }, { "epoch": 0.3076354867093214, "grad_norm": 0.06378534436225891, "learning_rate": 6.809882249737383e-06, "loss": 0.0062, "step": 7080 }, { "epoch": 0.30807000010862834, "grad_norm": 0.0639062151312828, "learning_rate": 6.783217682338655e-06, "loss": 0.0065, "step": 7090 }, { "epoch": 0.3085045135079353, "grad_norm": 0.06348402798175812, "learning_rate": 6.7565786017167004e-06, "loss": 0.0084, "step": 7100 }, { "epoch": 0.30893902690724223, "grad_norm": 0.08029096573591232, "learning_rate": 6.7299652189347e-06, "loss": 0.0103, "step": 7110 }, { "epoch": 0.3093735403065492, "grad_norm": 0.09063462167978287, "learning_rate": 6.703377744852227e-06, "loss": 0.0087, "step": 7120 }, { "epoch": 0.3098080537058562, "grad_norm": 0.09227754920721054, "learning_rate": 6.6768163901235776e-06, "loss": 0.0059, "step": 7130 }, { "epoch": 0.3102425671051631, "grad_norm": 0.09460172802209854, "learning_rate": 6.650281365196096e-06, "loss": 0.0071, "step": 7140 }, { "epoch": 0.31067708050447007, "grad_norm": 0.08981993049383163, "learning_rate": 6.623772880308534e-06, "loss": 0.0082, "step": 7150 }, { "epoch": 0.311111593903777, "grad_norm": 0.1009664312005043, "learning_rate": 6.597291145489344e-06, "loss": 0.0079, "step": 7160 }, { "epoch": 0.31154610730308396, "grad_norm": 0.07889096438884735, "learning_rate": 6.570836370555045e-06, "loss": 0.0069, "step": 7170 }, { "epoch": 0.3119806207023909, "grad_norm": 0.08060748875141144, "learning_rate": 6.544408765108549e-06, "loss": 0.0065, "step": 7180 }, { "epoch": 0.31241513410169786, "grad_norm": 0.047370001673698425, "learning_rate": 6.518008538537501e-06, "loss": 0.0085, "step": 7190 }, { "epoch": 0.3128496475010048, "grad_norm": 0.0684918537735939, "learning_rate": 6.4916359000126284e-06, "loss": 0.0089, "step": 7200 }, { "epoch": 0.31328416090031175, "grad_norm": 0.08644723892211914, "learning_rate": 6.465291058486072e-06, "loss": 0.0091, "step": 7210 }, { "epoch": 0.3137186742996187, "grad_norm": 0.06115137040615082, "learning_rate": 6.438974222689729e-06, "loss": 0.009, "step": 7220 }, { "epoch": 0.31415318769892564, "grad_norm": 0.09270749986171722, "learning_rate": 6.4126856011336146e-06, "loss": 0.0094, "step": 7230 }, { "epoch": 0.31458770109823264, "grad_norm": 0.06477729976177216, "learning_rate": 6.386425402104199e-06, "loss": 0.0079, "step": 7240 }, { "epoch": 0.3150222144975396, "grad_norm": 0.11577966809272766, "learning_rate": 6.3601938336627555e-06, "loss": 0.0079, "step": 7250 }, { "epoch": 0.31545672789684653, "grad_norm": 0.058794233947992325, "learning_rate": 6.33399110364372e-06, "loss": 0.0055, "step": 7260 }, { "epoch": 0.3158912412961535, "grad_norm": 0.09309697151184082, "learning_rate": 6.307817419653031e-06, "loss": 0.0076, "step": 7270 }, { "epoch": 0.3163257546954604, "grad_norm": 0.09355561435222626, "learning_rate": 6.281672989066501e-06, "loss": 0.0083, "step": 7280 }, { "epoch": 0.31676026809476737, "grad_norm": 0.05001560226082802, "learning_rate": 6.255558019028168e-06, "loss": 0.0086, "step": 7290 }, { "epoch": 0.3171947814940743, "grad_norm": 0.07396744191646576, "learning_rate": 6.229472716448647e-06, "loss": 0.0074, "step": 7300 }, { "epoch": 0.31762929489338126, "grad_norm": 0.05909964442253113, "learning_rate": 6.203417288003497e-06, "loss": 0.006, "step": 7310 }, { "epoch": 0.3180638082926882, "grad_norm": 0.08963511139154434, "learning_rate": 6.177391940131581e-06, "loss": 0.0077, "step": 7320 }, { "epoch": 0.31849832169199516, "grad_norm": 0.06426141411066055, "learning_rate": 6.15139687903343e-06, "loss": 0.0067, "step": 7330 }, { "epoch": 0.3189328350913021, "grad_norm": 0.09945748746395111, "learning_rate": 6.12543231066962e-06, "loss": 0.0082, "step": 7340 }, { "epoch": 0.3193673484906091, "grad_norm": 0.06709706038236618, "learning_rate": 6.099498440759123e-06, "loss": 0.0074, "step": 7350 }, { "epoch": 0.31980186188991605, "grad_norm": 0.07290440052747726, "learning_rate": 6.0735954747776856e-06, "loss": 0.0071, "step": 7360 }, { "epoch": 0.320236375289223, "grad_norm": 0.06449789553880692, "learning_rate": 6.047723617956201e-06, "loss": 0.0081, "step": 7370 }, { "epoch": 0.32067088868852994, "grad_norm": 0.08315026760101318, "learning_rate": 6.021883075279089e-06, "loss": 0.0085, "step": 7380 }, { "epoch": 0.3211054020878369, "grad_norm": 0.06351494789123535, "learning_rate": 5.996074051482657e-06, "loss": 0.0084, "step": 7390 }, { "epoch": 0.32153991548714383, "grad_norm": 0.0669468492269516, "learning_rate": 5.9702967510534884e-06, "loss": 0.0073, "step": 7400 }, { "epoch": 0.3219744288864508, "grad_norm": 0.060450613498687744, "learning_rate": 5.94455137822682e-06, "loss": 0.0064, "step": 7410 }, { "epoch": 0.3224089422857577, "grad_norm": 0.04701624810695648, "learning_rate": 5.918838136984926e-06, "loss": 0.0062, "step": 7420 }, { "epoch": 0.3228434556850647, "grad_norm": 0.0940837636590004, "learning_rate": 5.893157231055501e-06, "loss": 0.007, "step": 7430 }, { "epoch": 0.3232779690843716, "grad_norm": 0.10061509162187576, "learning_rate": 5.867508863910043e-06, "loss": 0.0079, "step": 7440 }, { "epoch": 0.32371248248367857, "grad_norm": 0.07226772606372833, "learning_rate": 5.841893238762242e-06, "loss": 0.0072, "step": 7450 }, { "epoch": 0.32414699588298557, "grad_norm": 0.10662239789962769, "learning_rate": 5.816310558566367e-06, "loss": 0.0061, "step": 7460 }, { "epoch": 0.3245815092822925, "grad_norm": 0.12014666199684143, "learning_rate": 5.790761026015675e-06, "loss": 0.0091, "step": 7470 }, { "epoch": 0.32501602268159946, "grad_norm": 0.10308293253183365, "learning_rate": 5.765244843540783e-06, "loss": 0.0096, "step": 7480 }, { "epoch": 0.3254505360809064, "grad_norm": 0.15284277498722076, "learning_rate": 5.739762213308073e-06, "loss": 0.007, "step": 7490 }, { "epoch": 0.32588504948021335, "grad_norm": 0.06631079316139221, "learning_rate": 5.714313337218087e-06, "loss": 0.0061, "step": 7500 }, { "epoch": 0.3263195628795203, "grad_norm": 0.13062670826911926, "learning_rate": 5.688898416903938e-06, "loss": 0.0084, "step": 7510 }, { "epoch": 0.32675407627882724, "grad_norm": 0.07424893230199814, "learning_rate": 5.663517653729708e-06, "loss": 0.0054, "step": 7520 }, { "epoch": 0.3271885896781342, "grad_norm": 0.054448146373033524, "learning_rate": 5.638171248788842e-06, "loss": 0.0068, "step": 7530 }, { "epoch": 0.32762310307744114, "grad_norm": 0.042936768382787704, "learning_rate": 5.6128594029025585e-06, "loss": 0.0057, "step": 7540 }, { "epoch": 0.3280576164767481, "grad_norm": 0.19675545394420624, "learning_rate": 5.587582316618276e-06, "loss": 0.0111, "step": 7550 }, { "epoch": 0.32849212987605503, "grad_norm": 0.060188472270965576, "learning_rate": 5.562340190207995e-06, "loss": 0.0083, "step": 7560 }, { "epoch": 0.32892664327536203, "grad_norm": 0.07820606976747513, "learning_rate": 5.537133223666742e-06, "loss": 0.0076, "step": 7570 }, { "epoch": 0.329361156674669, "grad_norm": 0.06374085694551468, "learning_rate": 5.511961616710961e-06, "loss": 0.0065, "step": 7580 }, { "epoch": 0.3297956700739759, "grad_norm": 0.05585566163063049, "learning_rate": 5.486825568776941e-06, "loss": 0.0074, "step": 7590 }, { "epoch": 0.33023018347328287, "grad_norm": 0.04485785588622093, "learning_rate": 5.461725279019226e-06, "loss": 0.0077, "step": 7600 }, { "epoch": 0.3306646968725898, "grad_norm": 0.08395210653543472, "learning_rate": 5.436660946309067e-06, "loss": 0.0082, "step": 7610 }, { "epoch": 0.33109921027189676, "grad_norm": 0.08163761347532272, "learning_rate": 5.411632769232808e-06, "loss": 0.0086, "step": 7620 }, { "epoch": 0.3315337236712037, "grad_norm": 0.09057643264532089, "learning_rate": 5.386640946090325e-06, "loss": 0.0084, "step": 7630 }, { "epoch": 0.33196823707051065, "grad_norm": 0.10619207471609116, "learning_rate": 5.361685674893481e-06, "loss": 0.007, "step": 7640 }, { "epoch": 0.3324027504698176, "grad_norm": 0.11618001759052277, "learning_rate": 5.3367671533645105e-06, "loss": 0.0071, "step": 7650 }, { "epoch": 0.33283726386912454, "grad_norm": 0.05917714163661003, "learning_rate": 5.3118855789345e-06, "loss": 0.0064, "step": 7660 }, { "epoch": 0.3332717772684315, "grad_norm": 0.04789431020617485, "learning_rate": 5.2870411487417825e-06, "loss": 0.006, "step": 7670 }, { "epoch": 0.3337062906677385, "grad_norm": 0.11848578602075577, "learning_rate": 5.262234059630415e-06, "loss": 0.0068, "step": 7680 }, { "epoch": 0.33414080406704544, "grad_norm": 0.06466028839349747, "learning_rate": 5.237464508148575e-06, "loss": 0.0074, "step": 7690 }, { "epoch": 0.3345753174663524, "grad_norm": 0.10490430891513824, "learning_rate": 5.212732690547047e-06, "loss": 0.0085, "step": 7700 }, { "epoch": 0.33500983086565933, "grad_norm": 0.0808047205209732, "learning_rate": 5.1880388027776415e-06, "loss": 0.0076, "step": 7710 }, { "epoch": 0.3354443442649663, "grad_norm": 0.08687549829483032, "learning_rate": 5.163383040491645e-06, "loss": 0.0086, "step": 7720 }, { "epoch": 0.3358788576642732, "grad_norm": 0.07844628393650055, "learning_rate": 5.1387655990382716e-06, "loss": 0.0078, "step": 7730 }, { "epoch": 0.33631337106358017, "grad_norm": 0.07566333562135696, "learning_rate": 5.114186673463123e-06, "loss": 0.0085, "step": 7740 }, { "epoch": 0.3367478844628871, "grad_norm": 0.042920857667922974, "learning_rate": 5.089646458506639e-06, "loss": 0.0081, "step": 7750 }, { "epoch": 0.33718239786219406, "grad_norm": 0.05019129067659378, "learning_rate": 5.065145148602542e-06, "loss": 0.0068, "step": 7760 }, { "epoch": 0.337616911261501, "grad_norm": 0.12619146704673767, "learning_rate": 5.040682937876319e-06, "loss": 0.0073, "step": 7770 }, { "epoch": 0.33805142466080795, "grad_norm": 0.10307935625314713, "learning_rate": 5.016260020143659e-06, "loss": 0.0089, "step": 7780 }, { "epoch": 0.33848593806011495, "grad_norm": 0.06839073449373245, "learning_rate": 4.9918765889089475e-06, "loss": 0.0064, "step": 7790 }, { "epoch": 0.3389204514594219, "grad_norm": 0.07737269252538681, "learning_rate": 4.967532837363695e-06, "loss": 0.0066, "step": 7800 }, { "epoch": 0.33935496485872885, "grad_norm": 0.07342928647994995, "learning_rate": 4.943228958385045e-06, "loss": 0.0073, "step": 7810 }, { "epoch": 0.3397894782580358, "grad_norm": 0.09632057696580887, "learning_rate": 4.918965144534219e-06, "loss": 0.008, "step": 7820 }, { "epoch": 0.34022399165734274, "grad_norm": 0.07035823166370392, "learning_rate": 4.894741588054993e-06, "loss": 0.0057, "step": 7830 }, { "epoch": 0.3406585050566497, "grad_norm": 0.059140220284461975, "learning_rate": 4.8705584808722065e-06, "loss": 0.0057, "step": 7840 }, { "epoch": 0.34109301845595663, "grad_norm": 0.10383981466293335, "learning_rate": 4.8464160145901894e-06, "loss": 0.0087, "step": 7850 }, { "epoch": 0.3415275318552636, "grad_norm": 0.08844143152236938, "learning_rate": 4.822314380491281e-06, "loss": 0.0083, "step": 7860 }, { "epoch": 0.3419620452545705, "grad_norm": 0.04909001663327217, "learning_rate": 4.7982537695343115e-06, "loss": 0.0079, "step": 7870 }, { "epoch": 0.34239655865387747, "grad_norm": 0.07348047941923141, "learning_rate": 4.7742343723530685e-06, "loss": 0.0083, "step": 7880 }, { "epoch": 0.3428310720531844, "grad_norm": 0.060402486473321915, "learning_rate": 4.750256379254814e-06, "loss": 0.0066, "step": 7890 }, { "epoch": 0.3432655854524914, "grad_norm": 0.12773926556110382, "learning_rate": 4.72631998021875e-06, "loss": 0.0077, "step": 7900 }, { "epoch": 0.34370009885179836, "grad_norm": 0.048774924129247665, "learning_rate": 4.70242536489454e-06, "loss": 0.007, "step": 7910 }, { "epoch": 0.3441346122511053, "grad_norm": 0.09722929447889328, "learning_rate": 4.6785727226007746e-06, "loss": 0.006, "step": 7920 }, { "epoch": 0.34456912565041226, "grad_norm": 0.08816218376159668, "learning_rate": 4.654762242323506e-06, "loss": 0.0066, "step": 7930 }, { "epoch": 0.3450036390497192, "grad_norm": 0.046284399926662445, "learning_rate": 4.63099411271473e-06, "loss": 0.0089, "step": 7940 }, { "epoch": 0.34543815244902615, "grad_norm": 0.06123678386211395, "learning_rate": 4.607268522090887e-06, "loss": 0.0088, "step": 7950 }, { "epoch": 0.3458726658483331, "grad_norm": 0.08266434818506241, "learning_rate": 4.583585658431383e-06, "loss": 0.0064, "step": 7960 }, { "epoch": 0.34630717924764004, "grad_norm": 0.061687689274549484, "learning_rate": 4.5599457093771e-06, "loss": 0.0085, "step": 7970 }, { "epoch": 0.346741692646947, "grad_norm": 0.065816231071949, "learning_rate": 4.536348862228902e-06, "loss": 0.0075, "step": 7980 }, { "epoch": 0.34717620604625393, "grad_norm": 0.045334603637456894, "learning_rate": 4.512795303946148e-06, "loss": 0.0052, "step": 7990 }, { "epoch": 0.3476107194455609, "grad_norm": 0.08470525592565536, "learning_rate": 4.489285221145227e-06, "loss": 0.0083, "step": 8000 }, { "epoch": 0.3480452328448679, "grad_norm": 0.06111738085746765, "learning_rate": 4.4658188000980586e-06, "loss": 0.0055, "step": 8010 }, { "epoch": 0.3484797462441748, "grad_norm": 0.08732720464468002, "learning_rate": 4.442396226730637e-06, "loss": 0.0082, "step": 8020 }, { "epoch": 0.34891425964348177, "grad_norm": 0.09063022583723068, "learning_rate": 4.419017686621536e-06, "loss": 0.006, "step": 8030 }, { "epoch": 0.3493487730427887, "grad_norm": 0.09084238857030869, "learning_rate": 4.395683365000468e-06, "loss": 0.0069, "step": 8040 }, { "epoch": 0.34978328644209566, "grad_norm": 0.06919077038764954, "learning_rate": 4.372393446746781e-06, "loss": 0.0078, "step": 8050 }, { "epoch": 0.3502177998414026, "grad_norm": 0.06786080449819565, "learning_rate": 4.349148116388026e-06, "loss": 0.0049, "step": 8060 }, { "epoch": 0.35065231324070956, "grad_norm": 0.12213651090860367, "learning_rate": 4.325947558098478e-06, "loss": 0.007, "step": 8070 }, { "epoch": 0.3510868266400165, "grad_norm": 0.04927683249115944, "learning_rate": 4.302791955697676e-06, "loss": 0.0061, "step": 8080 }, { "epoch": 0.35152134003932345, "grad_norm": 0.03858277201652527, "learning_rate": 4.27968149264897e-06, "loss": 0.006, "step": 8090 }, { "epoch": 0.3519558534386304, "grad_norm": 0.052387017756700516, "learning_rate": 4.256616352058073e-06, "loss": 0.0085, "step": 8100 }, { "epoch": 0.35239036683793734, "grad_norm": 0.07891060411930084, "learning_rate": 4.2335967166716064e-06, "loss": 0.0059, "step": 8110 }, { "epoch": 0.35282488023724434, "grad_norm": 0.06736122816801071, "learning_rate": 4.210622768875643e-06, "loss": 0.0077, "step": 8120 }, { "epoch": 0.3532593936365513, "grad_norm": 0.0961538702249527, "learning_rate": 4.187694690694279e-06, "loss": 0.0109, "step": 8130 }, { "epoch": 0.35369390703585823, "grad_norm": 0.0797416940331459, "learning_rate": 4.1648126637881745e-06, "loss": 0.008, "step": 8140 }, { "epoch": 0.3541284204351652, "grad_norm": 0.07754529267549515, "learning_rate": 4.141976869453123e-06, "loss": 0.0064, "step": 8150 }, { "epoch": 0.3545629338344721, "grad_norm": 0.110176220536232, "learning_rate": 4.119187488618621e-06, "loss": 0.0068, "step": 8160 }, { "epoch": 0.3549974472337791, "grad_norm": 0.09413070976734161, "learning_rate": 4.096444701846427e-06, "loss": 0.0078, "step": 8170 }, { "epoch": 0.355431960633086, "grad_norm": 0.09964610636234283, "learning_rate": 4.073748689329125e-06, "loss": 0.007, "step": 8180 }, { "epoch": 0.35586647403239297, "grad_norm": 0.14618416130542755, "learning_rate": 4.051099630888704e-06, "loss": 0.0072, "step": 8190 }, { "epoch": 0.3563009874316999, "grad_norm": 0.07930748909711838, "learning_rate": 4.028497705975139e-06, "loss": 0.0062, "step": 8200 }, { "epoch": 0.35673550083100686, "grad_norm": 0.0969531461596489, "learning_rate": 4.0059430936649645e-06, "loss": 0.0065, "step": 8210 }, { "epoch": 0.3571700142303138, "grad_norm": 0.06370487064123154, "learning_rate": 3.9834359726598415e-06, "loss": 0.008, "step": 8220 }, { "epoch": 0.3576045276296208, "grad_norm": 0.06438025087118149, "learning_rate": 3.9609765212851694e-06, "loss": 0.0062, "step": 8230 }, { "epoch": 0.35803904102892775, "grad_norm": 0.07360579073429108, "learning_rate": 3.938564917488644e-06, "loss": 0.0067, "step": 8240 }, { "epoch": 0.3584735544282347, "grad_norm": 0.10698428750038147, "learning_rate": 3.916201338838872e-06, "loss": 0.006, "step": 8250 }, { "epoch": 0.35890806782754164, "grad_norm": 0.07496725022792816, "learning_rate": 3.893885962523954e-06, "loss": 0.0065, "step": 8260 }, { "epoch": 0.3593425812268486, "grad_norm": 0.0918489471077919, "learning_rate": 3.871618965350075e-06, "loss": 0.0063, "step": 8270 }, { "epoch": 0.35977709462615554, "grad_norm": 0.06308259069919586, "learning_rate": 3.849400523740102e-06, "loss": 0.0072, "step": 8280 }, { "epoch": 0.3602116080254625, "grad_norm": 0.058937299996614456, "learning_rate": 3.82723081373221e-06, "loss": 0.0053, "step": 8290 }, { "epoch": 0.36064612142476943, "grad_norm": 0.0596042238175869, "learning_rate": 3.805110010978463e-06, "loss": 0.0073, "step": 8300 }, { "epoch": 0.3610806348240764, "grad_norm": 0.07679947465658188, "learning_rate": 3.783038290743427e-06, "loss": 0.0068, "step": 8310 }, { "epoch": 0.3615151482233833, "grad_norm": 0.04791805148124695, "learning_rate": 3.7610158279027796e-06, "loss": 0.0075, "step": 8320 }, { "epoch": 0.36194966162269027, "grad_norm": 0.06380537897348404, "learning_rate": 3.7390427969419395e-06, "loss": 0.0076, "step": 8330 }, { "epoch": 0.36238417502199727, "grad_norm": 0.06472238153219223, "learning_rate": 3.7171193719546726e-06, "loss": 0.0062, "step": 8340 }, { "epoch": 0.3628186884213042, "grad_norm": 0.09685289859771729, "learning_rate": 3.695245726641702e-06, "loss": 0.0066, "step": 8350 }, { "epoch": 0.36325320182061116, "grad_norm": 0.09350993484258652, "learning_rate": 3.6734220343093575e-06, "loss": 0.0053, "step": 8360 }, { "epoch": 0.3636877152199181, "grad_norm": 0.1431475281715393, "learning_rate": 3.6516484678681783e-06, "loss": 0.0102, "step": 8370 }, { "epoch": 0.36412222861922505, "grad_norm": 0.06581385433673859, "learning_rate": 3.629925199831552e-06, "loss": 0.0067, "step": 8380 }, { "epoch": 0.364556742018532, "grad_norm": 0.05527684837579727, "learning_rate": 3.6082524023143574e-06, "loss": 0.0059, "step": 8390 }, { "epoch": 0.36499125541783894, "grad_norm": 0.0786575973033905, "learning_rate": 3.5866302470315882e-06, "loss": 0.0069, "step": 8400 }, { "epoch": 0.3654257688171459, "grad_norm": 0.08859694749116898, "learning_rate": 3.565058905296991e-06, "loss": 0.0063, "step": 8410 }, { "epoch": 0.36586028221645284, "grad_norm": 0.10222943872213364, "learning_rate": 3.543538548021723e-06, "loss": 0.0079, "step": 8420 }, { "epoch": 0.3662947956157598, "grad_norm": 0.07007434964179993, "learning_rate": 3.5220693457129775e-06, "loss": 0.0075, "step": 8430 }, { "epoch": 0.36672930901506673, "grad_norm": 0.06237521767616272, "learning_rate": 3.5006514684726545e-06, "loss": 0.0082, "step": 8440 }, { "epoch": 0.36716382241437373, "grad_norm": 0.04917728528380394, "learning_rate": 3.4792850859959903e-06, "loss": 0.007, "step": 8450 }, { "epoch": 0.3675983358136807, "grad_norm": 0.08655932545661926, "learning_rate": 3.457970367570239e-06, "loss": 0.0088, "step": 8460 }, { "epoch": 0.3680328492129876, "grad_norm": 0.08564922958612442, "learning_rate": 3.4367074820733017e-06, "loss": 0.0053, "step": 8470 }, { "epoch": 0.36846736261229457, "grad_norm": 0.067174032330513, "learning_rate": 3.415496597972414e-06, "loss": 0.0062, "step": 8480 }, { "epoch": 0.3689018760116015, "grad_norm": 0.08985011279582977, "learning_rate": 3.394337883322805e-06, "loss": 0.0078, "step": 8490 }, { "epoch": 0.36933638941090846, "grad_norm": 0.0990537703037262, "learning_rate": 3.373231505766348e-06, "loss": 0.0075, "step": 8500 }, { "epoch": 0.3697709028102154, "grad_norm": 0.0639311894774437, "learning_rate": 3.352177632530251e-06, "loss": 0.0076, "step": 8510 }, { "epoch": 0.37020541620952235, "grad_norm": 0.03501976281404495, "learning_rate": 3.3311764304257342e-06, "loss": 0.0063, "step": 8520 }, { "epoch": 0.3706399296088293, "grad_norm": 0.06429938226938248, "learning_rate": 3.3102280658466977e-06, "loss": 0.0069, "step": 8530 }, { "epoch": 0.37107444300813625, "grad_norm": 0.0672391802072525, "learning_rate": 3.2893327047684034e-06, "loss": 0.0069, "step": 8540 }, { "epoch": 0.3715089564074432, "grad_norm": 0.10419747233390808, "learning_rate": 3.2684905127461573e-06, "loss": 0.0071, "step": 8550 }, { "epoch": 0.3719434698067502, "grad_norm": 0.08025926351547241, "learning_rate": 3.2477016549140173e-06, "loss": 0.0068, "step": 8560 }, { "epoch": 0.37237798320605714, "grad_norm": 0.05993766710162163, "learning_rate": 3.226966295983466e-06, "loss": 0.0086, "step": 8570 }, { "epoch": 0.3728124966053641, "grad_norm": 0.08144347369670868, "learning_rate": 3.206284600242102e-06, "loss": 0.006, "step": 8580 }, { "epoch": 0.37324701000467103, "grad_norm": 0.0695662721991539, "learning_rate": 3.185656731552362e-06, "loss": 0.0052, "step": 8590 }, { "epoch": 0.373681523403978, "grad_norm": 0.07274291664361954, "learning_rate": 3.1650828533501943e-06, "loss": 0.0058, "step": 8600 }, { "epoch": 0.3741160368032849, "grad_norm": 0.06351452320814133, "learning_rate": 3.144563128643776e-06, "loss": 0.0052, "step": 8610 }, { "epoch": 0.37455055020259187, "grad_norm": 0.07425472140312195, "learning_rate": 3.1240977200122422e-06, "loss": 0.0082, "step": 8620 }, { "epoch": 0.3749850636018988, "grad_norm": 0.06502246856689453, "learning_rate": 3.1036867896043574e-06, "loss": 0.0057, "step": 8630 }, { "epoch": 0.37541957700120576, "grad_norm": 0.03994821384549141, "learning_rate": 3.0833304991372557e-06, "loss": 0.0066, "step": 8640 }, { "epoch": 0.3758540904005127, "grad_norm": 0.07854374498128891, "learning_rate": 3.063029009895162e-06, "loss": 0.0063, "step": 8650 }, { "epoch": 0.37628860379981965, "grad_norm": 0.05410481616854668, "learning_rate": 3.0427824827281062e-06, "loss": 0.0077, "step": 8660 }, { "epoch": 0.3767231171991266, "grad_norm": 0.08873879164457321, "learning_rate": 3.022591078050644e-06, "loss": 0.0077, "step": 8670 }, { "epoch": 0.3771576305984336, "grad_norm": 0.062346357852220535, "learning_rate": 3.0024549558405945e-06, "loss": 0.0088, "step": 8680 }, { "epoch": 0.37759214399774055, "grad_norm": 0.06358446925878525, "learning_rate": 2.982374275637776e-06, "loss": 0.0082, "step": 8690 }, { "epoch": 0.3780266573970475, "grad_norm": 0.05350648984313011, "learning_rate": 2.9623491965427264e-06, "loss": 0.004, "step": 8700 }, { "epoch": 0.37846117079635444, "grad_norm": 0.08188925683498383, "learning_rate": 2.942379877215461e-06, "loss": 0.0068, "step": 8710 }, { "epoch": 0.3788956841956614, "grad_norm": 0.07393115013837814, "learning_rate": 2.922466475874206e-06, "loss": 0.0081, "step": 8720 }, { "epoch": 0.37933019759496833, "grad_norm": 0.08056262135505676, "learning_rate": 2.90260915029414e-06, "loss": 0.0068, "step": 8730 }, { "epoch": 0.3797647109942753, "grad_norm": 0.1089349240064621, "learning_rate": 2.882808057806149e-06, "loss": 0.0082, "step": 8740 }, { "epoch": 0.3801992243935822, "grad_norm": 0.0753493532538414, "learning_rate": 2.863063355295589e-06, "loss": 0.0055, "step": 8750 }, { "epoch": 0.38063373779288917, "grad_norm": 0.06075584515929222, "learning_rate": 2.8433751992010315e-06, "loss": 0.0064, "step": 8760 }, { "epoch": 0.3810682511921961, "grad_norm": 0.08952397853136063, "learning_rate": 2.8237437455130203e-06, "loss": 0.0058, "step": 8770 }, { "epoch": 0.38150276459150306, "grad_norm": 0.1460278481245041, "learning_rate": 2.8041691497728527e-06, "loss": 0.0095, "step": 8780 }, { "epoch": 0.38193727799081006, "grad_norm": 0.05322817713022232, "learning_rate": 2.784651567071327e-06, "loss": 0.0061, "step": 8790 }, { "epoch": 0.382371791390117, "grad_norm": 0.13151440024375916, "learning_rate": 2.7651911520475316e-06, "loss": 0.0074, "step": 8800 }, { "epoch": 0.38280630478942396, "grad_norm": 0.0687444806098938, "learning_rate": 2.745788058887604e-06, "loss": 0.0064, "step": 8810 }, { "epoch": 0.3832408181887309, "grad_norm": 0.05780453979969025, "learning_rate": 2.7264424413235267e-06, "loss": 0.0056, "step": 8820 }, { "epoch": 0.38367533158803785, "grad_norm": 0.06033443659543991, "learning_rate": 2.707154452631889e-06, "loss": 0.0045, "step": 8830 }, { "epoch": 0.3841098449873448, "grad_norm": 0.065843366086483, "learning_rate": 2.6879242456326827e-06, "loss": 0.005, "step": 8840 }, { "epoch": 0.38454435838665174, "grad_norm": 0.07421589642763138, "learning_rate": 2.6687519726881063e-06, "loss": 0.006, "step": 8850 }, { "epoch": 0.3849788717859587, "grad_norm": 0.09318463504314423, "learning_rate": 2.649637785701329e-06, "loss": 0.0081, "step": 8860 }, { "epoch": 0.38541338518526563, "grad_norm": 0.04374096915125847, "learning_rate": 2.630581836115301e-06, "loss": 0.0059, "step": 8870 }, { "epoch": 0.3858478985845726, "grad_norm": 0.1101013720035553, "learning_rate": 2.6115842749115604e-06, "loss": 0.0082, "step": 8880 }, { "epoch": 0.3862824119838795, "grad_norm": 0.08844293653964996, "learning_rate": 2.5926452526090305e-06, "loss": 0.007, "step": 8890 }, { "epoch": 0.3867169253831865, "grad_norm": 0.05187362805008888, "learning_rate": 2.573764919262819e-06, "loss": 0.0056, "step": 8900 }, { "epoch": 0.3871514387824935, "grad_norm": 0.06577293574810028, "learning_rate": 2.5549434244630478e-06, "loss": 0.0064, "step": 8910 }, { "epoch": 0.3875859521818004, "grad_norm": 0.08939126133918762, "learning_rate": 2.536180917333648e-06, "loss": 0.0069, "step": 8920 }, { "epoch": 0.38802046558110737, "grad_norm": 0.05213819071650505, "learning_rate": 2.5174775465311897e-06, "loss": 0.0074, "step": 8930 }, { "epoch": 0.3884549789804143, "grad_norm": 0.06323190778493881, "learning_rate": 2.4988334602437057e-06, "loss": 0.007, "step": 8940 }, { "epoch": 0.38888949237972126, "grad_norm": 0.05582160875201225, "learning_rate": 2.4802488061895137e-06, "loss": 0.0063, "step": 8950 }, { "epoch": 0.3893240057790282, "grad_norm": 0.07055678218603134, "learning_rate": 2.4617237316160427e-06, "loss": 0.0067, "step": 8960 }, { "epoch": 0.38975851917833515, "grad_norm": 0.04307841509580612, "learning_rate": 2.4432583832986633e-06, "loss": 0.0059, "step": 8970 }, { "epoch": 0.3901930325776421, "grad_norm": 0.08723011612892151, "learning_rate": 2.42485290753955e-06, "loss": 0.007, "step": 8980 }, { "epoch": 0.39062754597694904, "grad_norm": 0.08264581859111786, "learning_rate": 2.4065074501664863e-06, "loss": 0.007, "step": 8990 }, { "epoch": 0.391062059376256, "grad_norm": 0.10687217116355896, "learning_rate": 2.3882221565317277e-06, "loss": 0.0051, "step": 9000 }, { "epoch": 0.391496572775563, "grad_norm": 0.05457941070199013, "learning_rate": 2.3699971715108593e-06, "loss": 0.0053, "step": 9010 }, { "epoch": 0.39193108617486994, "grad_norm": 0.04098676145076752, "learning_rate": 2.3518326395016222e-06, "loss": 0.0055, "step": 9020 }, { "epoch": 0.3923655995741769, "grad_norm": 0.047154199331998825, "learning_rate": 2.3337287044227996e-06, "loss": 0.0068, "step": 9030 }, { "epoch": 0.39280011297348383, "grad_norm": 0.05270608887076378, "learning_rate": 2.315685509713046e-06, "loss": 0.0062, "step": 9040 }, { "epoch": 0.3932346263727908, "grad_norm": 0.06262046098709106, "learning_rate": 2.2977031983297817e-06, "loss": 0.0069, "step": 9050 }, { "epoch": 0.3936691397720977, "grad_norm": 0.07482646405696869, "learning_rate": 2.279781912748028e-06, "loss": 0.0067, "step": 9060 }, { "epoch": 0.39410365317140467, "grad_norm": 0.06200959533452988, "learning_rate": 2.2619217949593076e-06, "loss": 0.0074, "step": 9070 }, { "epoch": 0.3945381665707116, "grad_norm": 0.09040582180023193, "learning_rate": 2.2441229864705048e-06, "loss": 0.0074, "step": 9080 }, { "epoch": 0.39497267997001856, "grad_norm": 0.07866061478853226, "learning_rate": 2.226385628302742e-06, "loss": 0.0073, "step": 9090 }, { "epoch": 0.3954071933693255, "grad_norm": 0.07364839315414429, "learning_rate": 2.2087098609902636e-06, "loss": 0.0083, "step": 9100 }, { "epoch": 0.39584170676863245, "grad_norm": 0.07077943533658981, "learning_rate": 2.1910958245793347e-06, "loss": 0.0079, "step": 9110 }, { "epoch": 0.39627622016793945, "grad_norm": 0.09263047575950623, "learning_rate": 2.173543658627121e-06, "loss": 0.0063, "step": 9120 }, { "epoch": 0.3967107335672464, "grad_norm": 0.05478379502892494, "learning_rate": 2.1560535022005766e-06, "loss": 0.0075, "step": 9130 }, { "epoch": 0.39714524696655334, "grad_norm": 0.037011146545410156, "learning_rate": 2.138625493875359e-06, "loss": 0.0064, "step": 9140 }, { "epoch": 0.3975797603658603, "grad_norm": 0.07936471700668335, "learning_rate": 2.1212597717347183e-06, "loss": 0.0081, "step": 9150 }, { "epoch": 0.39801427376516724, "grad_norm": 0.1095275804400444, "learning_rate": 2.1039564733684014e-06, "loss": 0.0073, "step": 9160 }, { "epoch": 0.3984487871644742, "grad_norm": 0.05063824728131294, "learning_rate": 2.0867157358715794e-06, "loss": 0.0072, "step": 9170 }, { "epoch": 0.39888330056378113, "grad_norm": 0.05019955709576607, "learning_rate": 2.0695376958437442e-06, "loss": 0.0055, "step": 9180 }, { "epoch": 0.3993178139630881, "grad_norm": 0.06541518121957779, "learning_rate": 2.0524224893876253e-06, "loss": 0.0071, "step": 9190 }, { "epoch": 0.399752327362395, "grad_norm": 0.07469270378351212, "learning_rate": 2.0353702521081277e-06, "loss": 0.0052, "step": 9200 }, { "epoch": 0.40018684076170197, "grad_norm": 0.08644366264343262, "learning_rate": 2.0183811191112436e-06, "loss": 0.0062, "step": 9210 }, { "epoch": 0.4006213541610089, "grad_norm": 0.06640897691249847, "learning_rate": 2.001455225002984e-06, "loss": 0.0054, "step": 9220 }, { "epoch": 0.4010558675603159, "grad_norm": 0.07081051170825958, "learning_rate": 1.984592703888313e-06, "loss": 0.0057, "step": 9230 }, { "epoch": 0.40149038095962286, "grad_norm": 0.053732264786958694, "learning_rate": 1.967793689370093e-06, "loss": 0.0052, "step": 9240 }, { "epoch": 0.4019248943589298, "grad_norm": 0.060937847942113876, "learning_rate": 1.95105831454801e-06, "loss": 0.0065, "step": 9250 }, { "epoch": 0.40235940775823675, "grad_norm": 0.06776256114244461, "learning_rate": 1.9343867120175375e-06, "loss": 0.0072, "step": 9260 }, { "epoch": 0.4027939211575437, "grad_norm": 0.10395505279302597, "learning_rate": 1.9177790138688746e-06, "loss": 0.0068, "step": 9270 }, { "epoch": 0.40322843455685065, "grad_norm": 0.058665502816438675, "learning_rate": 1.9012353516858984e-06, "loss": 0.0055, "step": 9280 }, { "epoch": 0.4036629479561576, "grad_norm": 0.08694977313280106, "learning_rate": 1.884755856545123e-06, "loss": 0.0055, "step": 9290 }, { "epoch": 0.40409746135546454, "grad_norm": 0.05163672938942909, "learning_rate": 1.8683406590146714e-06, "loss": 0.0051, "step": 9300 }, { "epoch": 0.4045319747547715, "grad_norm": 0.10258106887340546, "learning_rate": 1.8519898891532273e-06, "loss": 0.0054, "step": 9310 }, { "epoch": 0.40496648815407843, "grad_norm": 0.0607648529112339, "learning_rate": 1.8357036765090107e-06, "loss": 0.0054, "step": 9320 }, { "epoch": 0.4054010015533854, "grad_norm": 0.08496219664812088, "learning_rate": 1.8194821501187455e-06, "loss": 0.0068, "step": 9330 }, { "epoch": 0.4058355149526924, "grad_norm": 0.0701276883482933, "learning_rate": 1.8033254385066501e-06, "loss": 0.0047, "step": 9340 }, { "epoch": 0.4062700283519993, "grad_norm": 0.043031346052885056, "learning_rate": 1.7872336696834091e-06, "loss": 0.0069, "step": 9350 }, { "epoch": 0.40670454175130627, "grad_norm": 0.09492848813533783, "learning_rate": 1.7712069711451553e-06, "loss": 0.0067, "step": 9360 }, { "epoch": 0.4071390551506132, "grad_norm": 0.23138777911663055, "learning_rate": 1.7552454698724753e-06, "loss": 0.0063, "step": 9370 }, { "epoch": 0.40757356854992016, "grad_norm": 0.07827364653348923, "learning_rate": 1.7393492923293854e-06, "loss": 0.0052, "step": 9380 }, { "epoch": 0.4080080819492271, "grad_norm": 0.07025067508220673, "learning_rate": 1.7235185644623352e-06, "loss": 0.0072, "step": 9390 }, { "epoch": 0.40844259534853405, "grad_norm": 0.0790853276848793, "learning_rate": 1.7077534116992266e-06, "loss": 0.0066, "step": 9400 }, { "epoch": 0.408877108747841, "grad_norm": 0.04930580407381058, "learning_rate": 1.692053958948393e-06, "loss": 0.0082, "step": 9410 }, { "epoch": 0.40931162214714795, "grad_norm": 0.049826521426439285, "learning_rate": 1.6764203305976224e-06, "loss": 0.0068, "step": 9420 }, { "epoch": 0.4097461355464549, "grad_norm": 0.038873229175806046, "learning_rate": 1.6608526505131773e-06, "loss": 0.007, "step": 9430 }, { "epoch": 0.41018064894576184, "grad_norm": 0.07032894343137741, "learning_rate": 1.6453510420388085e-06, "loss": 0.0049, "step": 9440 }, { "epoch": 0.41061516234506884, "grad_norm": 0.057884182780981064, "learning_rate": 1.6299156279947725e-06, "loss": 0.005, "step": 9450 }, { "epoch": 0.4110496757443758, "grad_norm": 0.06757012754678726, "learning_rate": 1.6145465306768604e-06, "loss": 0.0067, "step": 9460 }, { "epoch": 0.41148418914368273, "grad_norm": 0.07038474082946777, "learning_rate": 1.5992438718554415e-06, "loss": 0.0065, "step": 9470 }, { "epoch": 0.4119187025429897, "grad_norm": 0.06230664253234863, "learning_rate": 1.5840077727744785e-06, "loss": 0.0059, "step": 9480 }, { "epoch": 0.4123532159422966, "grad_norm": 0.06649099290370941, "learning_rate": 1.5688383541505835e-06, "loss": 0.0061, "step": 9490 }, { "epoch": 0.41278772934160357, "grad_norm": 0.10095405578613281, "learning_rate": 1.5537357361720551e-06, "loss": 0.0065, "step": 9500 }, { "epoch": 0.4132222427409105, "grad_norm": 0.056790791451931, "learning_rate": 1.5387000384979223e-06, "loss": 0.0052, "step": 9510 }, { "epoch": 0.41365675614021746, "grad_norm": 0.04234682396054268, "learning_rate": 1.5237313802569974e-06, "loss": 0.0069, "step": 9520 }, { "epoch": 0.4140912695395244, "grad_norm": 0.057066041976213455, "learning_rate": 1.5088298800469413e-06, "loss": 0.0074, "step": 9530 }, { "epoch": 0.41452578293883136, "grad_norm": 0.11709760129451752, "learning_rate": 1.4939956559333202e-06, "loss": 0.0068, "step": 9540 }, { "epoch": 0.4149602963381383, "grad_norm": 0.11893763393163681, "learning_rate": 1.479228825448654e-06, "loss": 0.0073, "step": 9550 }, { "epoch": 0.4153948097374453, "grad_norm": 0.05574529990553856, "learning_rate": 1.4645295055915154e-06, "loss": 0.0056, "step": 9560 }, { "epoch": 0.41582932313675225, "grad_norm": 0.07223575562238693, "learning_rate": 1.4498978128255691e-06, "loss": 0.0061, "step": 9570 }, { "epoch": 0.4162638365360592, "grad_norm": 0.04353675991296768, "learning_rate": 1.4353338630786817e-06, "loss": 0.0057, "step": 9580 }, { "epoch": 0.41669834993536614, "grad_norm": 0.08113046735525131, "learning_rate": 1.420837771741973e-06, "loss": 0.007, "step": 9590 }, { "epoch": 0.4171328633346731, "grad_norm": 0.06625610589981079, "learning_rate": 1.4064096536689298e-06, "loss": 0.0074, "step": 9600 }, { "epoch": 0.41756737673398003, "grad_norm": 0.06211593374609947, "learning_rate": 1.3920496231744717e-06, "loss": 0.0056, "step": 9610 }, { "epoch": 0.418001890133287, "grad_norm": 0.150951087474823, "learning_rate": 1.3777577940340558e-06, "loss": 0.0075, "step": 9620 }, { "epoch": 0.4184364035325939, "grad_norm": 0.1257888674736023, "learning_rate": 1.3635342794827888e-06, "loss": 0.0071, "step": 9630 }, { "epoch": 0.41887091693190087, "grad_norm": 0.0830865353345871, "learning_rate": 1.3493791922145027e-06, "loss": 0.0061, "step": 9640 }, { "epoch": 0.4193054303312078, "grad_norm": 0.07638338953256607, "learning_rate": 1.3352926443808778e-06, "loss": 0.0067, "step": 9650 }, { "epoch": 0.41973994373051476, "grad_norm": 0.09588000923395157, "learning_rate": 1.3212747475905564e-06, "loss": 0.0086, "step": 9660 }, { "epoch": 0.42017445712982177, "grad_norm": 0.0916999876499176, "learning_rate": 1.3073256129082534e-06, "loss": 0.0057, "step": 9670 }, { "epoch": 0.4206089705291287, "grad_norm": 0.0716477781534195, "learning_rate": 1.2934453508538746e-06, "loss": 0.0042, "step": 9680 }, { "epoch": 0.42104348392843566, "grad_norm": 0.060920000076293945, "learning_rate": 1.2796340714016419e-06, "loss": 0.0046, "step": 9690 }, { "epoch": 0.4214779973277426, "grad_norm": 0.116277314722538, "learning_rate": 1.26589188397923e-06, "loss": 0.0062, "step": 9700 }, { "epoch": 0.42191251072704955, "grad_norm": 0.09452416747808456, "learning_rate": 1.2522188974668847e-06, "loss": 0.006, "step": 9710 }, { "epoch": 0.4223470241263565, "grad_norm": 0.0834532305598259, "learning_rate": 1.2386152201965763e-06, "loss": 0.0066, "step": 9720 }, { "epoch": 0.42278153752566344, "grad_norm": 0.09147264808416367, "learning_rate": 1.2250809599511293e-06, "loss": 0.0064, "step": 9730 }, { "epoch": 0.4232160509249704, "grad_norm": 0.058695271611213684, "learning_rate": 1.2116162239633734e-06, "loss": 0.0058, "step": 9740 }, { "epoch": 0.42365056432427733, "grad_norm": 0.10094515234231949, "learning_rate": 1.198221118915287e-06, "loss": 0.0059, "step": 9750 }, { "epoch": 0.4240850777235843, "grad_norm": 0.11626178026199341, "learning_rate": 1.1848957509371739e-06, "loss": 0.0078, "step": 9760 }, { "epoch": 0.4245195911228912, "grad_norm": 0.09096948802471161, "learning_rate": 1.1716402256067905e-06, "loss": 0.0071, "step": 9770 }, { "epoch": 0.42495410452219823, "grad_norm": 0.10112593322992325, "learning_rate": 1.1584546479485316e-06, "loss": 0.0068, "step": 9780 }, { "epoch": 0.4253886179215052, "grad_norm": 0.08292309194803238, "learning_rate": 1.1453391224325928e-06, "loss": 0.0059, "step": 9790 }, { "epoch": 0.4258231313208121, "grad_norm": 0.12202641367912292, "learning_rate": 1.1322937529741384e-06, "loss": 0.0062, "step": 9800 }, { "epoch": 0.42625764472011907, "grad_norm": 0.12251067161560059, "learning_rate": 1.1193186429324887e-06, "loss": 0.008, "step": 9810 }, { "epoch": 0.426692158119426, "grad_norm": 0.14434179663658142, "learning_rate": 1.1064138951102843e-06, "loss": 0.0093, "step": 9820 }, { "epoch": 0.42712667151873296, "grad_norm": 0.08781075477600098, "learning_rate": 1.093579611752692e-06, "loss": 0.0067, "step": 9830 }, { "epoch": 0.4275611849180399, "grad_norm": 0.08094519376754761, "learning_rate": 1.080815894546574e-06, "loss": 0.0048, "step": 9840 }, { "epoch": 0.42799569831734685, "grad_norm": 0.0896298959851265, "learning_rate": 1.0681228446196978e-06, "loss": 0.0054, "step": 9850 }, { "epoch": 0.4284302117166538, "grad_norm": 0.07688537985086441, "learning_rate": 1.0555005625399316e-06, "loss": 0.0067, "step": 9860 }, { "epoch": 0.42886472511596074, "grad_norm": 0.05957145243883133, "learning_rate": 1.0429491483144394e-06, "loss": 0.0063, "step": 9870 }, { "epoch": 0.4292992385152677, "grad_norm": 0.062143776565790176, "learning_rate": 1.030468701388896e-06, "loss": 0.0055, "step": 9880 }, { "epoch": 0.4297337519145747, "grad_norm": 0.06420819461345673, "learning_rate": 1.0180593206467015e-06, "loss": 0.0071, "step": 9890 }, { "epoch": 0.43016826531388164, "grad_norm": 0.0829436257481575, "learning_rate": 1.0057211044081916e-06, "loss": 0.0053, "step": 9900 }, { "epoch": 0.4306027787131886, "grad_norm": 0.038479071110486984, "learning_rate": 9.934541504298589e-07, "loss": 0.008, "step": 9910 }, { "epoch": 0.43103729211249553, "grad_norm": 0.057884518057107925, "learning_rate": 9.812585559035848e-07, "loss": 0.007, "step": 9920 }, { "epoch": 0.4314718055118025, "grad_norm": 0.0618940070271492, "learning_rate": 9.691344174558615e-07, "loss": 0.0072, "step": 9930 }, { "epoch": 0.4319063189111094, "grad_norm": 0.06519508361816406, "learning_rate": 9.570818311470298e-07, "loss": 0.0075, "step": 9940 }, { "epoch": 0.43234083231041637, "grad_norm": 0.08697501569986343, "learning_rate": 9.451008924705196e-07, "loss": 0.0067, "step": 9950 }, { "epoch": 0.4327753457097233, "grad_norm": 0.07981958240270615, "learning_rate": 9.331916963520959e-07, "loss": 0.0064, "step": 9960 }, { "epoch": 0.43320985910903026, "grad_norm": 0.06097478047013283, "learning_rate": 9.213543371490963e-07, "loss": 0.0062, "step": 9970 }, { "epoch": 0.4336443725083372, "grad_norm": 0.055920518934726715, "learning_rate": 9.095889086496867e-07, "loss": 0.0064, "step": 9980 }, { "epoch": 0.43407888590764415, "grad_norm": 0.09264001250267029, "learning_rate": 8.978955040721371e-07, "loss": 0.0068, "step": 9990 }, { "epoch": 0.43451339930695115, "grad_norm": 0.07896417379379272, "learning_rate": 8.862742160640525e-07, "loss": 0.0064, "step": 10000 }, { "epoch": 0.4349479127062581, "grad_norm": 0.053227316588163376, "learning_rate": 8.747251367016552e-07, "loss": 0.0073, "step": 10010 }, { "epoch": 0.43538242610556505, "grad_norm": 0.04080301523208618, "learning_rate": 8.632483574890615e-07, "loss": 0.006, "step": 10020 }, { "epoch": 0.435816939504872, "grad_norm": 0.030272576957941055, "learning_rate": 8.518439693575408e-07, "loss": 0.0064, "step": 10030 }, { "epoch": 0.43625145290417894, "grad_norm": 0.06880766153335571, "learning_rate": 8.405120626648067e-07, "loss": 0.0074, "step": 10040 }, { "epoch": 0.4366859663034859, "grad_norm": 0.1094704121351242, "learning_rate": 8.292527271942996e-07, "loss": 0.0057, "step": 10050 }, { "epoch": 0.43712047970279283, "grad_norm": 0.08783411979675293, "learning_rate": 8.180660521544692e-07, "loss": 0.0094, "step": 10060 }, { "epoch": 0.4375549931020998, "grad_norm": 0.08887655287981033, "learning_rate": 8.069521261780733e-07, "loss": 0.0063, "step": 10070 }, { "epoch": 0.4379895065014067, "grad_norm": 0.07971230894327164, "learning_rate": 7.959110373214751e-07, "loss": 0.0068, "step": 10080 }, { "epoch": 0.43842401990071367, "grad_norm": 0.051013197749853134, "learning_rate": 7.849428730639463e-07, "loss": 0.0051, "step": 10090 }, { "epoch": 0.4388585333000206, "grad_norm": 0.08994479477405548, "learning_rate": 7.740477203069674e-07, "loss": 0.0058, "step": 10100 }, { "epoch": 0.4392930466993276, "grad_norm": 0.03989997133612633, "learning_rate": 7.63225665373546e-07, "loss": 0.0066, "step": 10110 }, { "epoch": 0.43972756009863456, "grad_norm": 0.058878444135189056, "learning_rate": 7.524767940075329e-07, "loss": 0.007, "step": 10120 }, { "epoch": 0.4401620734979415, "grad_norm": 0.052942000329494476, "learning_rate": 7.418011913729406e-07, "loss": 0.0061, "step": 10130 }, { "epoch": 0.44059658689724845, "grad_norm": 0.09582941979169846, "learning_rate": 7.311989420532639e-07, "loss": 0.0066, "step": 10140 }, { "epoch": 0.4410311002965554, "grad_norm": 0.07386282831430435, "learning_rate": 7.206701300508212e-07, "loss": 0.0059, "step": 10150 }, { "epoch": 0.44146561369586235, "grad_norm": 0.04970019310712814, "learning_rate": 7.102148387860764e-07, "loss": 0.0058, "step": 10160 }, { "epoch": 0.4419001270951693, "grad_norm": 0.06993437558412552, "learning_rate": 6.998331510969869e-07, "loss": 0.0067, "step": 10170 }, { "epoch": 0.44233464049447624, "grad_norm": 0.06462310254573822, "learning_rate": 6.895251492383426e-07, "loss": 0.0062, "step": 10180 }, { "epoch": 0.4427691538937832, "grad_norm": 0.06221221759915352, "learning_rate": 6.7929091488112e-07, "loss": 0.0053, "step": 10190 }, { "epoch": 0.44320366729309013, "grad_norm": 0.055420126765966415, "learning_rate": 6.691305291118234e-07, "loss": 0.0049, "step": 10200 }, { "epoch": 0.4436381806923971, "grad_norm": 0.08335283398628235, "learning_rate": 6.59044072431857e-07, "loss": 0.0055, "step": 10210 }, { "epoch": 0.4440726940917041, "grad_norm": 0.05226948857307434, "learning_rate": 6.490316247568762e-07, "loss": 0.0055, "step": 10220 }, { "epoch": 0.444507207491011, "grad_norm": 0.0770556852221489, "learning_rate": 6.390932654161596e-07, "loss": 0.0071, "step": 10230 }, { "epoch": 0.44494172089031797, "grad_norm": 0.04744845628738403, "learning_rate": 6.292290731519757e-07, "loss": 0.005, "step": 10240 }, { "epoch": 0.4453762342896249, "grad_norm": 0.07813084870576859, "learning_rate": 6.194391261189703e-07, "loss": 0.0051, "step": 10250 }, { "epoch": 0.44581074768893186, "grad_norm": 0.11257217824459076, "learning_rate": 6.097235018835279e-07, "loss": 0.0061, "step": 10260 }, { "epoch": 0.4462452610882388, "grad_norm": 0.10771900415420532, "learning_rate": 6.000822774231796e-07, "loss": 0.0082, "step": 10270 }, { "epoch": 0.44667977448754576, "grad_norm": 0.06365276128053665, "learning_rate": 5.905155291259768e-07, "loss": 0.0053, "step": 10280 }, { "epoch": 0.4471142878868527, "grad_norm": 0.04938112571835518, "learning_rate": 5.810233327898929e-07, "loss": 0.0058, "step": 10290 }, { "epoch": 0.44754880128615965, "grad_norm": 0.10017894953489304, "learning_rate": 5.716057636222172e-07, "loss": 0.006, "step": 10300 }, { "epoch": 0.4479833146854666, "grad_norm": 0.06827139854431152, "learning_rate": 5.622628962389687e-07, "loss": 0.0053, "step": 10310 }, { "epoch": 0.44841782808477354, "grad_norm": 0.09991230070590973, "learning_rate": 5.529948046642985e-07, "loss": 0.0072, "step": 10320 }, { "epoch": 0.44885234148408054, "grad_norm": 0.07327882200479507, "learning_rate": 5.43801562329902e-07, "loss": 0.0057, "step": 10330 }, { "epoch": 0.4492868548833875, "grad_norm": 0.0757019892334938, "learning_rate": 5.346832420744363e-07, "loss": 0.0069, "step": 10340 }, { "epoch": 0.44972136828269443, "grad_norm": 0.05079514533281326, "learning_rate": 5.256399161429515e-07, "loss": 0.0068, "step": 10350 }, { "epoch": 0.4501558816820014, "grad_norm": 0.06105861812829971, "learning_rate": 5.166716561863128e-07, "loss": 0.0067, "step": 10360 }, { "epoch": 0.4505903950813083, "grad_norm": 0.08541977405548096, "learning_rate": 5.077785332606266e-07, "loss": 0.0044, "step": 10370 }, { "epoch": 0.45102490848061527, "grad_norm": 0.1724163442850113, "learning_rate": 4.989606178266914e-07, "loss": 0.0063, "step": 10380 }, { "epoch": 0.4514594218799222, "grad_norm": 0.07179833203554153, "learning_rate": 4.902179797494255e-07, "loss": 0.0053, "step": 10390 }, { "epoch": 0.45189393527922916, "grad_norm": 0.11870724707841873, "learning_rate": 4.815506882973242e-07, "loss": 0.006, "step": 10400 }, { "epoch": 0.4523284486785361, "grad_norm": 0.04006374999880791, "learning_rate": 4.7295881214190486e-07, "loss": 0.0091, "step": 10410 }, { "epoch": 0.45276296207784306, "grad_norm": 0.0644485279917717, "learning_rate": 4.644424193571628e-07, "loss": 0.0058, "step": 10420 }, { "epoch": 0.45319747547715, "grad_norm": 0.05453095585107803, "learning_rate": 4.5600157741903626e-07, "loss": 0.0056, "step": 10430 }, { "epoch": 0.453631988876457, "grad_norm": 0.09095140546560287, "learning_rate": 4.4763635320486663e-07, "loss": 0.0082, "step": 10440 }, { "epoch": 0.45406650227576395, "grad_norm": 0.08895137161016464, "learning_rate": 4.3934681299287683e-07, "loss": 0.0061, "step": 10450 }, { "epoch": 0.4545010156750709, "grad_norm": 0.08378702402114868, "learning_rate": 4.311330224616328e-07, "loss": 0.0061, "step": 10460 }, { "epoch": 0.45493552907437784, "grad_norm": 0.06177133694291115, "learning_rate": 4.2299504668953383e-07, "loss": 0.0055, "step": 10470 }, { "epoch": 0.4553700424736848, "grad_norm": 0.053938720375299454, "learning_rate": 4.1493295015429645e-07, "loss": 0.0063, "step": 10480 }, { "epoch": 0.45580455587299173, "grad_norm": 0.05690125748515129, "learning_rate": 4.0694679673243807e-07, "loss": 0.0052, "step": 10490 }, { "epoch": 0.4562390692722987, "grad_norm": 0.055636215955019, "learning_rate": 3.990366496987741e-07, "loss": 0.006, "step": 10500 }, { "epoch": 0.4566735826716056, "grad_norm": 0.057698529213666916, "learning_rate": 3.912025717259194e-07, "loss": 0.006, "step": 10510 }, { "epoch": 0.4571080960709126, "grad_norm": 0.06959450244903564, "learning_rate": 3.834446248837853e-07, "loss": 0.0059, "step": 10520 }, { "epoch": 0.4575426094702195, "grad_norm": 0.07037419080734253, "learning_rate": 3.7576287063909034e-07, "loss": 0.0092, "step": 10530 }, { "epoch": 0.45797712286952647, "grad_norm": 0.0696893259882927, "learning_rate": 3.681573698548779e-07, "loss": 0.0064, "step": 10540 }, { "epoch": 0.45841163626883347, "grad_norm": 0.09161542356014252, "learning_rate": 3.606281827900282e-07, "loss": 0.0067, "step": 10550 }, { "epoch": 0.4588461496681404, "grad_norm": 0.04898115620017052, "learning_rate": 3.531753690987816e-07, "loss": 0.0074, "step": 10560 }, { "epoch": 0.45928066306744736, "grad_norm": 0.06962580233812332, "learning_rate": 3.4579898783027145e-07, "loss": 0.0075, "step": 10570 }, { "epoch": 0.4597151764667543, "grad_norm": 0.07286280393600464, "learning_rate": 3.3849909742804553e-07, "loss": 0.0052, "step": 10580 }, { "epoch": 0.46014968986606125, "grad_norm": 0.0685592070221901, "learning_rate": 3.3127575572961755e-07, "loss": 0.0081, "step": 10590 }, { "epoch": 0.4605842032653682, "grad_norm": 0.07501457631587982, "learning_rate": 3.2412901996599075e-07, "loss": 0.0059, "step": 10600 }, { "epoch": 0.46101871666467514, "grad_norm": 0.055459920316934586, "learning_rate": 3.170589467612262e-07, "loss": 0.0053, "step": 10610 }, { "epoch": 0.4614532300639821, "grad_norm": 0.07159551233053207, "learning_rate": 3.100655921319706e-07, "loss": 0.0074, "step": 10620 }, { "epoch": 0.46188774346328904, "grad_norm": 0.07556604593992233, "learning_rate": 3.03149011487035e-07, "loss": 0.0051, "step": 10630 }, { "epoch": 0.462322256862596, "grad_norm": 0.05757951736450195, "learning_rate": 2.96309259626939e-07, "loss": 0.0097, "step": 10640 }, { "epoch": 0.4627567702619029, "grad_norm": 0.07891504466533661, "learning_rate": 2.895463907434837e-07, "loss": 0.007, "step": 10650 }, { "epoch": 0.46319128366120993, "grad_norm": 0.04868793115019798, "learning_rate": 2.8286045841932064e-07, "loss": 0.006, "step": 10660 }, { "epoch": 0.4636257970605169, "grad_norm": 0.11161201447248459, "learning_rate": 2.762515156275303e-07, "loss": 0.0052, "step": 10670 }, { "epoch": 0.4640603104598238, "grad_norm": 0.07221969217061996, "learning_rate": 2.697196147311987e-07, "loss": 0.0079, "step": 10680 }, { "epoch": 0.46449482385913077, "grad_norm": 0.07179443538188934, "learning_rate": 2.6326480748300467e-07, "loss": 0.0069, "step": 10690 }, { "epoch": 0.4649293372584377, "grad_norm": 0.10213626176118851, "learning_rate": 2.5688714502480783e-07, "loss": 0.0069, "step": 10700 }, { "epoch": 0.46536385065774466, "grad_norm": 0.05147803574800491, "learning_rate": 2.5058667788724566e-07, "loss": 0.0065, "step": 10710 }, { "epoch": 0.4657983640570516, "grad_norm": 0.03046802245080471, "learning_rate": 2.4436345598932933e-07, "loss": 0.0073, "step": 10720 }, { "epoch": 0.46623287745635855, "grad_norm": 0.045375872403383255, "learning_rate": 2.3821752863805502e-07, "loss": 0.0059, "step": 10730 }, { "epoch": 0.4666673908556655, "grad_norm": 0.11396394670009613, "learning_rate": 2.3214894452800784e-07, "loss": 0.0057, "step": 10740 }, { "epoch": 0.46710190425497244, "grad_norm": 0.05735607445240021, "learning_rate": 2.2615775174097633e-07, "loss": 0.0065, "step": 10750 }, { "epoch": 0.4675364176542794, "grad_norm": 0.05877210199832916, "learning_rate": 2.2024399774556948e-07, "loss": 0.0037, "step": 10760 }, { "epoch": 0.4679709310535864, "grad_norm": 0.09503768384456635, "learning_rate": 2.1440772939685272e-07, "loss": 0.0057, "step": 10770 }, { "epoch": 0.46840544445289334, "grad_norm": 0.08510304242372513, "learning_rate": 2.086489929359603e-07, "loss": 0.0046, "step": 10780 }, { "epoch": 0.4688399578522003, "grad_norm": 0.10792459547519684, "learning_rate": 2.0296783398973452e-07, "loss": 0.0064, "step": 10790 }, { "epoch": 0.46927447125150723, "grad_norm": 0.05590936169028282, "learning_rate": 1.973642975703738e-07, "loss": 0.0055, "step": 10800 }, { "epoch": 0.4697089846508142, "grad_norm": 0.04974950850009918, "learning_rate": 1.918384280750618e-07, "loss": 0.0043, "step": 10810 }, { "epoch": 0.4701434980501211, "grad_norm": 0.14148883521556854, "learning_rate": 1.8639026928562453e-07, "loss": 0.0074, "step": 10820 }, { "epoch": 0.47057801144942807, "grad_norm": 0.08046723902225494, "learning_rate": 1.810198643681793e-07, "loss": 0.0053, "step": 10830 }, { "epoch": 0.471012524848735, "grad_norm": 0.06864185631275177, "learning_rate": 1.7572725587279738e-07, "loss": 0.007, "step": 10840 }, { "epoch": 0.47144703824804196, "grad_norm": 0.1145964041352272, "learning_rate": 1.7051248573316083e-07, "loss": 0.0063, "step": 10850 }, { "epoch": 0.4718815516473489, "grad_norm": 0.07953092455863953, "learning_rate": 1.6537559526623614e-07, "loss": 0.0063, "step": 10860 }, { "epoch": 0.47231606504665585, "grad_norm": 0.04491928219795227, "learning_rate": 1.6031662517194235e-07, "loss": 0.0059, "step": 10870 }, { "epoch": 0.47275057844596285, "grad_norm": 0.026222513988614082, "learning_rate": 1.5533561553282895e-07, "loss": 0.0042, "step": 10880 }, { "epoch": 0.4731850918452698, "grad_norm": 0.08618629723787308, "learning_rate": 1.5043260581376285e-07, "loss": 0.0058, "step": 10890 }, { "epoch": 0.47361960524457675, "grad_norm": 0.07844515144824982, "learning_rate": 1.4560763486160868e-07, "loss": 0.0068, "step": 10900 }, { "epoch": 0.4740541186438837, "grad_norm": 0.06669094413518906, "learning_rate": 1.4086074090493007e-07, "loss": 0.0058, "step": 10910 }, { "epoch": 0.47448863204319064, "grad_norm": 0.06707707047462463, "learning_rate": 1.3619196155367664e-07, "loss": 0.0055, "step": 10920 }, { "epoch": 0.4749231454424976, "grad_norm": 0.08793602883815765, "learning_rate": 1.3160133379889305e-07, "loss": 0.0068, "step": 10930 }, { "epoch": 0.47535765884180453, "grad_norm": 0.05631832033395767, "learning_rate": 1.2708889401242263e-07, "loss": 0.0064, "step": 10940 }, { "epoch": 0.4757921722411115, "grad_norm": 0.09073697775602341, "learning_rate": 1.22654677946622e-07, "loss": 0.0082, "step": 10950 }, { "epoch": 0.4762266856404184, "grad_norm": 0.08381710201501846, "learning_rate": 1.1829872073407467e-07, "loss": 0.0053, "step": 10960 }, { "epoch": 0.47666119903972537, "grad_norm": 0.044089365750551224, "learning_rate": 1.1402105688731568e-07, "loss": 0.0061, "step": 10970 }, { "epoch": 0.4770957124390323, "grad_norm": 0.06977023184299469, "learning_rate": 1.0982172029855409e-07, "loss": 0.0077, "step": 10980 }, { "epoch": 0.4775302258383393, "grad_norm": 0.07421712577342987, "learning_rate": 1.0570074423940758e-07, "loss": 0.0049, "step": 10990 }, { "epoch": 0.47796473923764626, "grad_norm": 0.0687830001115799, "learning_rate": 1.0165816136064266e-07, "loss": 0.0068, "step": 11000 }, { "epoch": 0.4783992526369532, "grad_norm": 0.05829397588968277, "learning_rate": 9.769400369190496e-08, "loss": 0.0062, "step": 11010 }, { "epoch": 0.47883376603626016, "grad_norm": 0.07584609091281891, "learning_rate": 9.3808302641476e-08, "loss": 0.0064, "step": 11020 }, { "epoch": 0.4792682794355671, "grad_norm": 0.0651635229587555, "learning_rate": 9.000108899602011e-08, "loss": 0.0065, "step": 11030 }, { "epoch": 0.47970279283487405, "grad_norm": 0.06295653432607651, "learning_rate": 8.627239292033907e-08, "loss": 0.0056, "step": 11040 }, { "epoch": 0.480137306234181, "grad_norm": 0.10909046232700348, "learning_rate": 8.262224395713559e-08, "loss": 0.0077, "step": 11050 }, { "epoch": 0.48057181963348794, "grad_norm": 0.05574299395084381, "learning_rate": 7.905067102678021e-08, "loss": 0.0051, "step": 11060 }, { "epoch": 0.4810063330327949, "grad_norm": 0.0833418220281601, "learning_rate": 7.555770242707705e-08, "loss": 0.0051, "step": 11070 }, { "epoch": 0.48144084643210183, "grad_norm": 0.07659269869327545, "learning_rate": 7.214336583304616e-08, "loss": 0.0058, "step": 11080 }, { "epoch": 0.4818753598314088, "grad_norm": 0.07921771705150604, "learning_rate": 6.880768829670036e-08, "loss": 0.006, "step": 11090 }, { "epoch": 0.4823098732307158, "grad_norm": 0.06295815110206604, "learning_rate": 6.555069624682997e-08, "loss": 0.007, "step": 11100 }, { "epoch": 0.4827443866300227, "grad_norm": 0.06317449361085892, "learning_rate": 6.237241548879613e-08, "loss": 0.0055, "step": 11110 }, { "epoch": 0.48317890002932967, "grad_norm": 0.09486140310764313, "learning_rate": 5.9272871204324457e-08, "loss": 0.0069, "step": 11120 }, { "epoch": 0.4836134134286366, "grad_norm": 0.10850790143013, "learning_rate": 5.625208795130954e-08, "loss": 0.0061, "step": 11130 }, { "epoch": 0.48404792682794356, "grad_norm": 0.06827063858509064, "learning_rate": 5.3310089663611844e-08, "loss": 0.0057, "step": 11140 }, { "epoch": 0.4844824402272505, "grad_norm": 0.055239662528038025, "learning_rate": 5.04468996508789e-08, "loss": 0.0083, "step": 11150 }, { "epoch": 0.48491695362655746, "grad_norm": 0.08159584552049637, "learning_rate": 4.766254059835107e-08, "loss": 0.0055, "step": 11160 }, { "epoch": 0.4853514670258644, "grad_norm": 0.06158966198563576, "learning_rate": 4.4957034566687205e-08, "loss": 0.0057, "step": 11170 }, { "epoch": 0.48578598042517135, "grad_norm": 0.03253835067152977, "learning_rate": 4.2330402991789255e-08, "loss": 0.0053, "step": 11180 }, { "epoch": 0.4862204938244783, "grad_norm": 0.08656269311904907, "learning_rate": 3.9782666684631266e-08, "loss": 0.0066, "step": 11190 }, { "epoch": 0.48665500722378524, "grad_norm": 0.08442550897598267, "learning_rate": 3.7313845831093984e-08, "loss": 0.0066, "step": 11200 }, { "epoch": 0.4870895206230922, "grad_norm": 0.06725499778985977, "learning_rate": 3.492395999180609e-08, "loss": 0.0061, "step": 11210 }, { "epoch": 0.4875240340223992, "grad_norm": 0.09642699360847473, "learning_rate": 3.261302810198985e-08, "loss": 0.007, "step": 11220 }, { "epoch": 0.48795854742170613, "grad_norm": 0.05808092653751373, "learning_rate": 3.038106847131128e-08, "loss": 0.0058, "step": 11230 }, { "epoch": 0.4883930608210131, "grad_norm": 0.07764837890863419, "learning_rate": 2.822809878373134e-08, "loss": 0.0076, "step": 11240 }, { "epoch": 0.48882757422032, "grad_norm": 0.047388892620801926, "learning_rate": 2.6154136097369386e-08, "loss": 0.0051, "step": 11250 }, { "epoch": 0.489262087619627, "grad_norm": 0.08450010418891907, "learning_rate": 2.415919684436774e-08, "loss": 0.0073, "step": 11260 }, { "epoch": 0.4896966010189339, "grad_norm": 0.09996630996465683, "learning_rate": 2.224329683076065e-08, "loss": 0.0078, "step": 11270 }, { "epoch": 0.49013111441824087, "grad_norm": 0.08762189745903015, "learning_rate": 2.0406451236349988e-08, "loss": 0.0059, "step": 11280 }, { "epoch": 0.4905656278175478, "grad_norm": 0.06317150592803955, "learning_rate": 1.8648674614583084e-08, "loss": 0.0062, "step": 11290 }, { "epoch": 0.49100014121685476, "grad_norm": 0.06011360511183739, "learning_rate": 1.6969980892439508e-08, "loss": 0.0057, "step": 11300 }, { "epoch": 0.4914346546161617, "grad_norm": 0.06316263228654861, "learning_rate": 1.537038337031782e-08, "loss": 0.0067, "step": 11310 }, { "epoch": 0.49186916801546865, "grad_norm": 0.10531795769929886, "learning_rate": 1.3849894721935653e-08, "loss": 0.0064, "step": 11320 }, { "epoch": 0.49230368141477565, "grad_norm": 0.09684789925813675, "learning_rate": 1.2408526994223125e-08, "loss": 0.0065, "step": 11330 }, { "epoch": 0.4927381948140826, "grad_norm": 0.06766608357429504, "learning_rate": 1.1046291607231807e-08, "loss": 0.0062, "step": 11340 }, { "epoch": 0.49317270821338954, "grad_norm": 0.04883148521184921, "learning_rate": 9.763199354041463e-09, "loss": 0.0067, "step": 11350 }, { "epoch": 0.4936072216126965, "grad_norm": 0.057381924241781235, "learning_rate": 8.559260400674564e-09, "loss": 0.0062, "step": 11360 }, { "epoch": 0.49404173501200344, "grad_norm": 0.07570614665746689, "learning_rate": 7.434484286020782e-09, "loss": 0.0075, "step": 11370 }, { "epoch": 0.4944762484113104, "grad_norm": 0.03946550935506821, "learning_rate": 6.388879921753743e-09, "loss": 0.0062, "step": 11380 }, { "epoch": 0.49491076181061733, "grad_norm": 0.059903956949710846, "learning_rate": 5.422455592264397e-09, "loss": 0.0065, "step": 11390 }, { "epoch": 0.4953452752099243, "grad_norm": 0.08148939162492752, "learning_rate": 4.535218954596632e-09, "loss": 0.0075, "step": 11400 }, { "epoch": 0.4957797886092312, "grad_norm": 0.12582159042358398, "learning_rate": 3.727177038385099e-09, "loss": 0.0057, "step": 11410 }, { "epoch": 0.49621430200853817, "grad_norm": 0.11304333060979843, "learning_rate": 2.998336245797484e-09, "loss": 0.0066, "step": 11420 }, { "epoch": 0.4966488154078451, "grad_norm": 0.08336801081895828, "learning_rate": 2.348702351487875e-09, "loss": 0.0057, "step": 11430 }, { "epoch": 0.4970833288071521, "grad_norm": 0.05257615074515343, "learning_rate": 1.778280502546803e-09, "loss": 0.0067, "step": 11440 }, { "epoch": 0.49751784220645906, "grad_norm": 0.0752919614315033, "learning_rate": 1.2870752184657165e-09, "loss": 0.0049, "step": 11450 }, { "epoch": 0.497952355605766, "grad_norm": 0.07169657945632935, "learning_rate": 8.750903910959008e-10, "loss": 0.0068, "step": 11460 }, { "epoch": 0.49838686900507295, "grad_norm": 0.07614590227603912, "learning_rate": 5.423292846196138e-10, "loss": 0.007, "step": 11470 }, { "epoch": 0.4988213824043799, "grad_norm": 0.09255485981702805, "learning_rate": 2.8879453552455026e-10, "loss": 0.0066, "step": 11480 }, { "epoch": 0.49925589580368684, "grad_norm": 0.05579246208071709, "learning_rate": 1.1448815258385815e-10, "loss": 0.0056, "step": 11490 }, { "epoch": 0.4996904092029938, "grad_norm": 0.09784293174743652, "learning_rate": 1.9411516838374612e-11, "loss": 0.0058, "step": 11500 } ], "logging_steps": 10, "max_steps": 11507, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.2555120144411525e+19, "train_batch_size": 1, "trial_name": null, "trial_params": null }