{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9999769511661064, "eval_steps": 500, "global_step": 37962, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00026341524449873725, "grad_norm": 283.7031555175781, "learning_rate": 1.7559262510974542e-07, "loss": 3.1922, "step": 10 }, { "epoch": 0.0005268304889974745, "grad_norm": 176.34739685058594, "learning_rate": 3.5118525021949084e-07, "loss": 2.8665, "step": 20 }, { "epoch": 0.0007902457334962118, "grad_norm": 77.87225341796875, "learning_rate": 5.267778753292362e-07, "loss": 2.3801, "step": 30 }, { "epoch": 0.001053660977994949, "grad_norm": 28.646303176879883, "learning_rate": 7.023705004389817e-07, "loss": 1.9282, "step": 40 }, { "epoch": 0.0013170762224936862, "grad_norm": 102.01678466796875, "learning_rate": 8.77963125548727e-07, "loss": 2.0488, "step": 50 }, { "epoch": 0.0015804914669924235, "grad_norm": 83.87590789794922, "learning_rate": 1.0535557506584724e-06, "loss": 1.9304, "step": 60 }, { "epoch": 0.0018439067114911607, "grad_norm": 46.90418243408203, "learning_rate": 1.2291483757682179e-06, "loss": 1.8107, "step": 70 }, { "epoch": 0.002107321955989898, "grad_norm": 23.468515396118164, "learning_rate": 1.4047410008779634e-06, "loss": 1.9055, "step": 80 }, { "epoch": 0.002370737200488635, "grad_norm": 83.90035247802734, "learning_rate": 1.5803336259877086e-06, "loss": 1.79, "step": 90 }, { "epoch": 0.0026341524449873723, "grad_norm": 59.483978271484375, "learning_rate": 1.755926251097454e-06, "loss": 1.7498, "step": 100 }, { "epoch": 0.00289756768948611, "grad_norm": 39.354087829589844, "learning_rate": 1.9315188762071994e-06, "loss": 1.81, "step": 110 }, { "epoch": 0.003160982933984847, "grad_norm": 56.23602294921875, "learning_rate": 2.1071115013169447e-06, "loss": 1.8988, "step": 120 }, { "epoch": 0.003424398178483584, "grad_norm": 24.504413604736328, "learning_rate": 2.2827041264266904e-06, "loss": 1.7794, "step": 130 }, { "epoch": 0.0036878134229823213, "grad_norm": 81.9885482788086, "learning_rate": 2.4582967515364357e-06, "loss": 1.7848, "step": 140 }, { "epoch": 0.0039512286674810585, "grad_norm": 17.06349754333496, "learning_rate": 2.633889376646181e-06, "loss": 1.7798, "step": 150 }, { "epoch": 0.004214643911979796, "grad_norm": 83.96887969970703, "learning_rate": 2.8094820017559267e-06, "loss": 1.8855, "step": 160 }, { "epoch": 0.004478059156478534, "grad_norm": 67.86595916748047, "learning_rate": 2.9850746268656716e-06, "loss": 1.8967, "step": 170 }, { "epoch": 0.00474147440097727, "grad_norm": 46.12517547607422, "learning_rate": 3.1606672519754173e-06, "loss": 1.802, "step": 180 }, { "epoch": 0.005004889645476008, "grad_norm": 16.377017974853516, "learning_rate": 3.336259877085163e-06, "loss": 1.7118, "step": 190 }, { "epoch": 0.005268304889974745, "grad_norm": 70.087890625, "learning_rate": 3.511852502194908e-06, "loss": 1.8184, "step": 200 }, { "epoch": 0.005531720134473482, "grad_norm": 68.34032440185547, "learning_rate": 3.6874451273046536e-06, "loss": 1.7608, "step": 210 }, { "epoch": 0.00579513537897222, "grad_norm": 158.8023223876953, "learning_rate": 3.863037752414399e-06, "loss": 1.6859, "step": 220 }, { "epoch": 0.0060585506234709565, "grad_norm": 79.01270294189453, "learning_rate": 4.038630377524144e-06, "loss": 1.7302, "step": 230 }, { "epoch": 0.006321965867969694, "grad_norm": 110.54615020751953, "learning_rate": 4.2142230026338894e-06, "loss": 1.99, "step": 240 }, { "epoch": 0.006585381112468432, "grad_norm": 22.725868225097656, "learning_rate": 4.389815627743635e-06, "loss": 1.5635, "step": 250 }, { "epoch": 0.006848796356967168, "grad_norm": 28.27669906616211, "learning_rate": 4.565408252853381e-06, "loss": 1.849, "step": 260 }, { "epoch": 0.007112211601465906, "grad_norm": 30.341520309448242, "learning_rate": 4.741000877963126e-06, "loss": 1.5261, "step": 270 }, { "epoch": 0.007375626845964643, "grad_norm": 41.23030090332031, "learning_rate": 4.916593503072871e-06, "loss": 1.7342, "step": 280 }, { "epoch": 0.00763904209046338, "grad_norm": 33.45962142944336, "learning_rate": 5.092186128182617e-06, "loss": 1.6142, "step": 290 }, { "epoch": 0.007902457334962117, "grad_norm": 54.00699234008789, "learning_rate": 5.267778753292362e-06, "loss": 1.6266, "step": 300 }, { "epoch": 0.008165872579460855, "grad_norm": 43.90302276611328, "learning_rate": 5.443371378402108e-06, "loss": 1.5956, "step": 310 }, { "epoch": 0.008429287823959592, "grad_norm": 91.89335632324219, "learning_rate": 5.618964003511853e-06, "loss": 1.7029, "step": 320 }, { "epoch": 0.008692703068458329, "grad_norm": 13.909472465515137, "learning_rate": 5.794556628621598e-06, "loss": 1.7131, "step": 330 }, { "epoch": 0.008956118312957067, "grad_norm": 44.409873962402344, "learning_rate": 5.970149253731343e-06, "loss": 1.6425, "step": 340 }, { "epoch": 0.009219533557455804, "grad_norm": 17.333356857299805, "learning_rate": 6.145741878841089e-06, "loss": 1.697, "step": 350 }, { "epoch": 0.00948294880195454, "grad_norm": 7.805108070373535, "learning_rate": 6.3213345039508346e-06, "loss": 1.6329, "step": 360 }, { "epoch": 0.00974636404645328, "grad_norm": 18.26932716369629, "learning_rate": 6.49692712906058e-06, "loss": 1.7064, "step": 370 }, { "epoch": 0.010009779290952016, "grad_norm": 76.36417388916016, "learning_rate": 6.672519754170326e-06, "loss": 1.7632, "step": 380 }, { "epoch": 0.010273194535450753, "grad_norm": 94.7707748413086, "learning_rate": 6.84811237928007e-06, "loss": 1.6917, "step": 390 }, { "epoch": 0.01053660977994949, "grad_norm": 17.537052154541016, "learning_rate": 7.023705004389816e-06, "loss": 1.6339, "step": 400 }, { "epoch": 0.010800025024448228, "grad_norm": 18.1473445892334, "learning_rate": 7.199297629499562e-06, "loss": 1.5146, "step": 410 }, { "epoch": 0.011063440268946964, "grad_norm": 28.40329933166504, "learning_rate": 7.374890254609307e-06, "loss": 1.5836, "step": 420 }, { "epoch": 0.011326855513445701, "grad_norm": 45.898616790771484, "learning_rate": 7.550482879719052e-06, "loss": 1.5443, "step": 430 }, { "epoch": 0.01159027075794444, "grad_norm": 16.250577926635742, "learning_rate": 7.726075504828798e-06, "loss": 1.6585, "step": 440 }, { "epoch": 0.011853686002443176, "grad_norm": 58.57795715332031, "learning_rate": 7.901668129938544e-06, "loss": 1.6632, "step": 450 }, { "epoch": 0.012117101246941913, "grad_norm": 50.49228286743164, "learning_rate": 8.077260755048288e-06, "loss": 1.6166, "step": 460 }, { "epoch": 0.012380516491440651, "grad_norm": 18.522960662841797, "learning_rate": 8.252853380158033e-06, "loss": 1.5089, "step": 470 }, { "epoch": 0.012643931735939388, "grad_norm": 17.64044952392578, "learning_rate": 8.428446005267779e-06, "loss": 1.7003, "step": 480 }, { "epoch": 0.012907346980438125, "grad_norm": 25.662025451660156, "learning_rate": 8.604038630377525e-06, "loss": 1.6881, "step": 490 }, { "epoch": 0.013170762224936863, "grad_norm": 58.899742126464844, "learning_rate": 8.77963125548727e-06, "loss": 1.8645, "step": 500 }, { "epoch": 0.0134341774694356, "grad_norm": 72.4648666381836, "learning_rate": 8.955223880597016e-06, "loss": 1.6712, "step": 510 }, { "epoch": 0.013697592713934337, "grad_norm": 63.722835540771484, "learning_rate": 9.130816505706762e-06, "loss": 1.6911, "step": 520 }, { "epoch": 0.013961007958433075, "grad_norm": 56.987552642822266, "learning_rate": 9.306409130816506e-06, "loss": 1.6364, "step": 530 }, { "epoch": 0.014224423202931812, "grad_norm": 63.16893768310547, "learning_rate": 9.482001755926252e-06, "loss": 1.668, "step": 540 }, { "epoch": 0.014487838447430549, "grad_norm": 14.675129890441895, "learning_rate": 9.657594381035997e-06, "loss": 1.5511, "step": 550 }, { "epoch": 0.014751253691929285, "grad_norm": 16.1970157623291, "learning_rate": 9.833187006145743e-06, "loss": 1.5133, "step": 560 }, { "epoch": 0.015014668936428024, "grad_norm": 39.011173248291016, "learning_rate": 1.0008779631255487e-05, "loss": 1.6573, "step": 570 }, { "epoch": 0.01527808418092676, "grad_norm": 55.41944885253906, "learning_rate": 1.0184372256365233e-05, "loss": 1.6567, "step": 580 }, { "epoch": 0.015541499425425497, "grad_norm": 29.878705978393555, "learning_rate": 1.035996488147498e-05, "loss": 1.5952, "step": 590 }, { "epoch": 0.015804914669924234, "grad_norm": 40.10651397705078, "learning_rate": 1.0535557506584724e-05, "loss": 1.5811, "step": 600 }, { "epoch": 0.016068329914422974, "grad_norm": 31.494121551513672, "learning_rate": 1.071115013169447e-05, "loss": 1.6497, "step": 610 }, { "epoch": 0.01633174515892171, "grad_norm": 26.42911720275879, "learning_rate": 1.0886742756804216e-05, "loss": 1.6076, "step": 620 }, { "epoch": 0.016595160403420448, "grad_norm": 66.62787628173828, "learning_rate": 1.106233538191396e-05, "loss": 1.7983, "step": 630 }, { "epoch": 0.016858575647919184, "grad_norm": 64.07002258300781, "learning_rate": 1.1237928007023707e-05, "loss": 1.5466, "step": 640 }, { "epoch": 0.01712199089241792, "grad_norm": 16.801776885986328, "learning_rate": 1.1413520632133453e-05, "loss": 1.5491, "step": 650 }, { "epoch": 0.017385406136916658, "grad_norm": 13.05986499786377, "learning_rate": 1.1589113257243196e-05, "loss": 1.6565, "step": 660 }, { "epoch": 0.017648821381415394, "grad_norm": 36.30992889404297, "learning_rate": 1.1764705882352942e-05, "loss": 1.654, "step": 670 }, { "epoch": 0.017912236625914135, "grad_norm": 6.151001930236816, "learning_rate": 1.1940298507462686e-05, "loss": 1.627, "step": 680 }, { "epoch": 0.01817565187041287, "grad_norm": 59.942779541015625, "learning_rate": 1.2115891132572432e-05, "loss": 1.7311, "step": 690 }, { "epoch": 0.018439067114911608, "grad_norm": 59.08943176269531, "learning_rate": 1.2291483757682179e-05, "loss": 1.7379, "step": 700 }, { "epoch": 0.018702482359410345, "grad_norm": 54.9960823059082, "learning_rate": 1.2467076382791923e-05, "loss": 1.5843, "step": 710 }, { "epoch": 0.01896589760390908, "grad_norm": 13.231147766113281, "learning_rate": 1.2642669007901669e-05, "loss": 1.6225, "step": 720 }, { "epoch": 0.019229312848407818, "grad_norm": 46.019798278808594, "learning_rate": 1.2818261633011415e-05, "loss": 1.6166, "step": 730 }, { "epoch": 0.01949272809290656, "grad_norm": 43.63382339477539, "learning_rate": 1.299385425812116e-05, "loss": 1.6617, "step": 740 }, { "epoch": 0.019756143337405295, "grad_norm": 57.042537689208984, "learning_rate": 1.3169446883230906e-05, "loss": 1.7987, "step": 750 }, { "epoch": 0.02001955858190403, "grad_norm": 51.18360137939453, "learning_rate": 1.3345039508340652e-05, "loss": 1.6918, "step": 760 }, { "epoch": 0.02028297382640277, "grad_norm": 48.76926803588867, "learning_rate": 1.3520632133450395e-05, "loss": 1.6811, "step": 770 }, { "epoch": 0.020546389070901505, "grad_norm": 43.75407409667969, "learning_rate": 1.369622475856014e-05, "loss": 1.6549, "step": 780 }, { "epoch": 0.020809804315400242, "grad_norm": 55.638633728027344, "learning_rate": 1.3871817383669887e-05, "loss": 1.6094, "step": 790 }, { "epoch": 0.02107321955989898, "grad_norm": 64.54676055908203, "learning_rate": 1.4047410008779631e-05, "loss": 1.9367, "step": 800 }, { "epoch": 0.02133663480439772, "grad_norm": 61.4254150390625, "learning_rate": 1.4223002633889378e-05, "loss": 1.6365, "step": 810 }, { "epoch": 0.021600050048896455, "grad_norm": 19.859241485595703, "learning_rate": 1.4398595258999124e-05, "loss": 1.5813, "step": 820 }, { "epoch": 0.021863465293395192, "grad_norm": 83.693603515625, "learning_rate": 1.4574187884108868e-05, "loss": 1.759, "step": 830 }, { "epoch": 0.02212688053789393, "grad_norm": 35.15703582763672, "learning_rate": 1.4749780509218614e-05, "loss": 1.6277, "step": 840 }, { "epoch": 0.022390295782392666, "grad_norm": 52.248008728027344, "learning_rate": 1.492537313432836e-05, "loss": 1.5686, "step": 850 }, { "epoch": 0.022653711026891402, "grad_norm": 21.968286514282227, "learning_rate": 1.5100965759438105e-05, "loss": 1.5431, "step": 860 }, { "epoch": 0.022917126271390142, "grad_norm": 76.19892120361328, "learning_rate": 1.527655838454785e-05, "loss": 1.6102, "step": 870 }, { "epoch": 0.02318054151588888, "grad_norm": 22.568992614746094, "learning_rate": 1.5452151009657595e-05, "loss": 1.6927, "step": 880 }, { "epoch": 0.023443956760387616, "grad_norm": 62.15504837036133, "learning_rate": 1.562774363476734e-05, "loss": 1.6507, "step": 890 }, { "epoch": 0.023707372004886353, "grad_norm": 25.25591278076172, "learning_rate": 1.5803336259877088e-05, "loss": 1.5483, "step": 900 }, { "epoch": 0.02397078724938509, "grad_norm": 10.233015060424805, "learning_rate": 1.5978928884986834e-05, "loss": 1.6073, "step": 910 }, { "epoch": 0.024234202493883826, "grad_norm": 38.76236343383789, "learning_rate": 1.6154521510096577e-05, "loss": 1.7568, "step": 920 }, { "epoch": 0.024497617738382566, "grad_norm": 30.897741317749023, "learning_rate": 1.6330114135206323e-05, "loss": 1.6173, "step": 930 }, { "epoch": 0.024761032982881303, "grad_norm": 37.78947830200195, "learning_rate": 1.6505706760316065e-05, "loss": 1.5767, "step": 940 }, { "epoch": 0.02502444822738004, "grad_norm": 30.57618522644043, "learning_rate": 1.668129938542581e-05, "loss": 1.4724, "step": 950 }, { "epoch": 0.025287863471878776, "grad_norm": 27.889238357543945, "learning_rate": 1.6856892010535558e-05, "loss": 1.612, "step": 960 }, { "epoch": 0.025551278716377513, "grad_norm": 9.004977226257324, "learning_rate": 1.7032484635645304e-05, "loss": 1.483, "step": 970 }, { "epoch": 0.02581469396087625, "grad_norm": 28.223093032836914, "learning_rate": 1.720807726075505e-05, "loss": 1.7255, "step": 980 }, { "epoch": 0.026078109205374986, "grad_norm": 20.885543823242188, "learning_rate": 1.7383669885864796e-05, "loss": 1.562, "step": 990 }, { "epoch": 0.026341524449873727, "grad_norm": 26.906631469726562, "learning_rate": 1.755926251097454e-05, "loss": 1.6316, "step": 1000 }, { "epoch": 0.026604939694372463, "grad_norm": 7.186174392700195, "learning_rate": 1.7734855136084285e-05, "loss": 1.5701, "step": 1010 }, { "epoch": 0.0268683549388712, "grad_norm": 17.23870086669922, "learning_rate": 1.791044776119403e-05, "loss": 1.7596, "step": 1020 }, { "epoch": 0.027131770183369937, "grad_norm": 17.744979858398438, "learning_rate": 1.8086040386303777e-05, "loss": 1.6703, "step": 1030 }, { "epoch": 0.027395185427868674, "grad_norm": 7.944180488586426, "learning_rate": 1.8261633011413523e-05, "loss": 1.7382, "step": 1040 }, { "epoch": 0.02765860067236741, "grad_norm": 17.63950538635254, "learning_rate": 1.843722563652327e-05, "loss": 1.6142, "step": 1050 }, { "epoch": 0.02792201591686615, "grad_norm": 48.01411056518555, "learning_rate": 1.8612818261633012e-05, "loss": 1.6398, "step": 1060 }, { "epoch": 0.028185431161364887, "grad_norm": 31.48131561279297, "learning_rate": 1.878841088674276e-05, "loss": 1.5782, "step": 1070 }, { "epoch": 0.028448846405863624, "grad_norm": 10.33256721496582, "learning_rate": 1.8964003511852505e-05, "loss": 1.7118, "step": 1080 }, { "epoch": 0.02871226165036236, "grad_norm": 11.275463104248047, "learning_rate": 1.9139596136962247e-05, "loss": 1.6845, "step": 1090 }, { "epoch": 0.028975676894861097, "grad_norm": 61.95518112182617, "learning_rate": 1.9315188762071993e-05, "loss": 1.6839, "step": 1100 }, { "epoch": 0.029239092139359834, "grad_norm": 69.8826904296875, "learning_rate": 1.949078138718174e-05, "loss": 1.6975, "step": 1110 }, { "epoch": 0.02950250738385857, "grad_norm": 12.102856636047363, "learning_rate": 1.9666374012291486e-05, "loss": 1.5871, "step": 1120 }, { "epoch": 0.02976592262835731, "grad_norm": 69.20232391357422, "learning_rate": 1.9841966637401232e-05, "loss": 1.8457, "step": 1130 }, { "epoch": 0.030029337872856048, "grad_norm": 36.88510513305664, "learning_rate": 1.9999999963605866e-05, "loss": 1.7612, "step": 1140 }, { "epoch": 0.030292753117354784, "grad_norm": 63.974308013916016, "learning_rate": 1.9999995596309868e-05, "loss": 1.6623, "step": 1150 }, { "epoch": 0.03055616836185352, "grad_norm": 39.72188186645508, "learning_rate": 1.999998395019031e-05, "loss": 1.5893, "step": 1160 }, { "epoch": 0.030819583606352258, "grad_norm": 38.60426712036133, "learning_rate": 1.9999965025255676e-05, "loss": 1.6189, "step": 1170 }, { "epoch": 0.031082998850850994, "grad_norm": 25.619131088256836, "learning_rate": 1.9999938821519738e-05, "loss": 1.5236, "step": 1180 }, { "epoch": 0.03134641409534973, "grad_norm": 20.88021469116211, "learning_rate": 1.9999905339001568e-05, "loss": 1.5725, "step": 1190 }, { "epoch": 0.03160982933984847, "grad_norm": 51.05508804321289, "learning_rate": 1.9999864577725537e-05, "loss": 1.4729, "step": 1200 }, { "epoch": 0.031873244584347205, "grad_norm": 16.338815689086914, "learning_rate": 1.9999816537721318e-05, "loss": 1.5339, "step": 1210 }, { "epoch": 0.03213665982884595, "grad_norm": 67.26805114746094, "learning_rate": 1.9999761219023873e-05, "loss": 1.629, "step": 1220 }, { "epoch": 0.032400075073344685, "grad_norm": 11.371353149414062, "learning_rate": 1.9999698621673474e-05, "loss": 1.533, "step": 1230 }, { "epoch": 0.03266349031784342, "grad_norm": 48.50962448120117, "learning_rate": 1.9999628745715678e-05, "loss": 1.5784, "step": 1240 }, { "epoch": 0.03292690556234216, "grad_norm": 21.110523223876953, "learning_rate": 1.999955159120135e-05, "loss": 1.5816, "step": 1250 }, { "epoch": 0.033190320806840895, "grad_norm": 15.816736221313477, "learning_rate": 1.9999467158186652e-05, "loss": 1.6949, "step": 1260 }, { "epoch": 0.03345373605133963, "grad_norm": 29.19423484802246, "learning_rate": 1.9999375446733035e-05, "loss": 1.6069, "step": 1270 }, { "epoch": 0.03371715129583837, "grad_norm": 12.36455249786377, "learning_rate": 1.9999276456907262e-05, "loss": 1.5314, "step": 1280 }, { "epoch": 0.033980566540337105, "grad_norm": 33.648460388183594, "learning_rate": 1.999917018878138e-05, "loss": 1.7558, "step": 1290 }, { "epoch": 0.03424398178483584, "grad_norm": 25.541271209716797, "learning_rate": 1.9999056642432738e-05, "loss": 1.5849, "step": 1300 }, { "epoch": 0.03450739702933458, "grad_norm": 31.491891860961914, "learning_rate": 1.999893581794399e-05, "loss": 1.5663, "step": 1310 }, { "epoch": 0.034770812273833315, "grad_norm": 27.946325302124023, "learning_rate": 1.9998807715403083e-05, "loss": 1.5944, "step": 1320 }, { "epoch": 0.03503422751833205, "grad_norm": 47.63523483276367, "learning_rate": 1.9998672334903253e-05, "loss": 1.6291, "step": 1330 }, { "epoch": 0.03529764276283079, "grad_norm": 32.599159240722656, "learning_rate": 1.9998529676543048e-05, "loss": 1.6477, "step": 1340 }, { "epoch": 0.03556105800732953, "grad_norm": 39.296539306640625, "learning_rate": 1.9998379740426303e-05, "loss": 1.5897, "step": 1350 }, { "epoch": 0.03582447325182827, "grad_norm": 17.99929428100586, "learning_rate": 1.9998222526662157e-05, "loss": 1.6937, "step": 1360 }, { "epoch": 0.036087888496327006, "grad_norm": 13.584074020385742, "learning_rate": 1.9998058035365038e-05, "loss": 1.5566, "step": 1370 }, { "epoch": 0.03635130374082574, "grad_norm": 22.093358993530273, "learning_rate": 1.999788626665468e-05, "loss": 1.5561, "step": 1380 }, { "epoch": 0.03661471898532448, "grad_norm": 11.384148597717285, "learning_rate": 1.9997707220656114e-05, "loss": 1.6367, "step": 1390 }, { "epoch": 0.036878134229823216, "grad_norm": 15.962431907653809, "learning_rate": 1.9997520897499652e-05, "loss": 1.5147, "step": 1400 }, { "epoch": 0.03714154947432195, "grad_norm": 38.767581939697266, "learning_rate": 1.999732729732093e-05, "loss": 1.5826, "step": 1410 }, { "epoch": 0.03740496471882069, "grad_norm": 6.569390296936035, "learning_rate": 1.999712642026086e-05, "loss": 1.6265, "step": 1420 }, { "epoch": 0.037668379963319426, "grad_norm": 40.68317413330078, "learning_rate": 1.999691826646566e-05, "loss": 1.5759, "step": 1430 }, { "epoch": 0.03793179520781816, "grad_norm": 27.016881942749023, "learning_rate": 1.9996702836086835e-05, "loss": 1.6018, "step": 1440 }, { "epoch": 0.0381952104523169, "grad_norm": 19.092866897583008, "learning_rate": 1.9996480129281195e-05, "loss": 1.7494, "step": 1450 }, { "epoch": 0.038458625696815636, "grad_norm": 19.89524269104004, "learning_rate": 1.999625014621085e-05, "loss": 1.5535, "step": 1460 }, { "epoch": 0.03872204094131437, "grad_norm": 37.28947830200195, "learning_rate": 1.999601288704319e-05, "loss": 1.4804, "step": 1470 }, { "epoch": 0.03898545618581312, "grad_norm": 11.237868309020996, "learning_rate": 1.9995768351950926e-05, "loss": 1.6079, "step": 1480 }, { "epoch": 0.03924887143031185, "grad_norm": 25.3029727935791, "learning_rate": 1.999551654111204e-05, "loss": 1.5832, "step": 1490 }, { "epoch": 0.03951228667481059, "grad_norm": 47.03425216674805, "learning_rate": 1.9995257454709824e-05, "loss": 1.6602, "step": 1500 }, { "epoch": 0.03977570191930933, "grad_norm": 17.505346298217773, "learning_rate": 1.9994991092932863e-05, "loss": 1.415, "step": 1510 }, { "epoch": 0.04003911716380806, "grad_norm": 20.101842880249023, "learning_rate": 1.999471745597504e-05, "loss": 1.4591, "step": 1520 }, { "epoch": 0.0403025324083068, "grad_norm": 25.03113555908203, "learning_rate": 1.999443654403552e-05, "loss": 1.4966, "step": 1530 }, { "epoch": 0.04056594765280554, "grad_norm": 66.55901336669922, "learning_rate": 1.999414835731879e-05, "loss": 1.6699, "step": 1540 }, { "epoch": 0.040829362897304274, "grad_norm": 9.774673461914062, "learning_rate": 1.9993852896034603e-05, "loss": 1.571, "step": 1550 }, { "epoch": 0.04109277814180301, "grad_norm": 11.24197769165039, "learning_rate": 1.9993550160398026e-05, "loss": 1.4993, "step": 1560 }, { "epoch": 0.04135619338630175, "grad_norm": 41.2164306640625, "learning_rate": 1.999324015062941e-05, "loss": 1.5106, "step": 1570 }, { "epoch": 0.041619608630800484, "grad_norm": 10.746949195861816, "learning_rate": 1.9992922866954413e-05, "loss": 1.5473, "step": 1580 }, { "epoch": 0.04188302387529922, "grad_norm": 19.454713821411133, "learning_rate": 1.9992598309603976e-05, "loss": 1.5379, "step": 1590 }, { "epoch": 0.04214643911979796, "grad_norm": 33.08735656738281, "learning_rate": 1.999226647881434e-05, "loss": 1.5885, "step": 1600 }, { "epoch": 0.0424098543642967, "grad_norm": 14.927067756652832, "learning_rate": 1.999192737482704e-05, "loss": 1.4829, "step": 1610 }, { "epoch": 0.04267326960879544, "grad_norm": 8.97916030883789, "learning_rate": 1.9991580997888897e-05, "loss": 1.5836, "step": 1620 }, { "epoch": 0.042936684853294174, "grad_norm": 253.7731170654297, "learning_rate": 1.999122734825204e-05, "loss": 1.5624, "step": 1630 }, { "epoch": 0.04320010009779291, "grad_norm": 5653.267578125, "learning_rate": 1.9990866426173886e-05, "loss": 22.6966, "step": 1640 }, { "epoch": 0.04346351534229165, "grad_norm": 1016.4690551757812, "learning_rate": 1.9990498231917135e-05, "loss": 4.0821, "step": 1650 }, { "epoch": 0.043726930586790384, "grad_norm": 35.97153091430664, "learning_rate": 1.9990122765749796e-05, "loss": 1.9765, "step": 1660 }, { "epoch": 0.04399034583128912, "grad_norm": 198.94776916503906, "learning_rate": 1.9989740027945164e-05, "loss": 1.8108, "step": 1670 }, { "epoch": 0.04425376107578786, "grad_norm": 40.04741668701172, "learning_rate": 1.9989350018781826e-05, "loss": 1.5637, "step": 1680 }, { "epoch": 0.044517176320286594, "grad_norm": 18.696552276611328, "learning_rate": 1.998895273854366e-05, "loss": 1.4969, "step": 1690 }, { "epoch": 0.04478059156478533, "grad_norm": 15.722201347351074, "learning_rate": 1.998854818751985e-05, "loss": 1.6165, "step": 1700 }, { "epoch": 0.04504400680928407, "grad_norm": 14.195305824279785, "learning_rate": 1.9988136366004845e-05, "loss": 1.5769, "step": 1710 }, { "epoch": 0.045307422053782805, "grad_norm": 24.94065284729004, "learning_rate": 1.9987717274298415e-05, "loss": 1.6393, "step": 1720 }, { "epoch": 0.04557083729828155, "grad_norm": 27.443214416503906, "learning_rate": 1.9987290912705607e-05, "loss": 1.68, "step": 1730 }, { "epoch": 0.045834252542780285, "grad_norm": 21.934724807739258, "learning_rate": 1.9986857281536763e-05, "loss": 1.5109, "step": 1740 }, { "epoch": 0.04609766778727902, "grad_norm": 11.89421272277832, "learning_rate": 1.9986416381107512e-05, "loss": 1.6754, "step": 1750 }, { "epoch": 0.04636108303177776, "grad_norm": 27.342628479003906, "learning_rate": 1.9985968211738783e-05, "loss": 1.6331, "step": 1760 }, { "epoch": 0.046624498276276495, "grad_norm": 3.336338520050049, "learning_rate": 1.9985512773756784e-05, "loss": 1.6099, "step": 1770 }, { "epoch": 0.04688791352077523, "grad_norm": 33.684730529785156, "learning_rate": 1.9985050067493028e-05, "loss": 1.5391, "step": 1780 }, { "epoch": 0.04715132876527397, "grad_norm": 16.756601333618164, "learning_rate": 1.9984580093284305e-05, "loss": 1.4882, "step": 1790 }, { "epoch": 0.047414744009772705, "grad_norm": 12.322381019592285, "learning_rate": 1.99841028514727e-05, "loss": 1.6272, "step": 1800 }, { "epoch": 0.04767815925427144, "grad_norm": 33.99985885620117, "learning_rate": 1.9983618342405593e-05, "loss": 1.6318, "step": 1810 }, { "epoch": 0.04794157449877018, "grad_norm": 535.0435180664062, "learning_rate": 1.9983126566435655e-05, "loss": 1.6623, "step": 1820 }, { "epoch": 0.048204989743268915, "grad_norm": 22.206634521484375, "learning_rate": 1.998262752392083e-05, "loss": 1.5772, "step": 1830 }, { "epoch": 0.04846840498776765, "grad_norm": 16.05791473388672, "learning_rate": 1.9982121215224367e-05, "loss": 1.5535, "step": 1840 }, { "epoch": 0.04873182023226639, "grad_norm": 9.760551452636719, "learning_rate": 1.9981607640714804e-05, "loss": 1.4902, "step": 1850 }, { "epoch": 0.04899523547676513, "grad_norm": 17.269702911376953, "learning_rate": 1.9981086800765957e-05, "loss": 1.4701, "step": 1860 }, { "epoch": 0.04925865072126387, "grad_norm": 44.855865478515625, "learning_rate": 1.9980558695756937e-05, "loss": 1.6399, "step": 1870 }, { "epoch": 0.049522065965762606, "grad_norm": 29.226150512695312, "learning_rate": 1.9980023326072145e-05, "loss": 1.5269, "step": 1880 }, { "epoch": 0.04978548121026134, "grad_norm": 22.832210540771484, "learning_rate": 1.9979480692101263e-05, "loss": 1.5901, "step": 1890 }, { "epoch": 0.05004889645476008, "grad_norm": 48.36526107788086, "learning_rate": 1.9978930794239272e-05, "loss": 1.5673, "step": 1900 }, { "epoch": 0.050312311699258816, "grad_norm": 21.323749542236328, "learning_rate": 1.9978373632886424e-05, "loss": 1.5328, "step": 1910 }, { "epoch": 0.05057572694375755, "grad_norm": 45.57045364379883, "learning_rate": 1.9977809208448278e-05, "loss": 1.3632, "step": 1920 }, { "epoch": 0.05083914218825629, "grad_norm": 17.159040451049805, "learning_rate": 1.997723752133566e-05, "loss": 1.6033, "step": 1930 }, { "epoch": 0.051102557432755026, "grad_norm": 26.004825592041016, "learning_rate": 1.9976658571964693e-05, "loss": 1.6035, "step": 1940 }, { "epoch": 0.05136597267725376, "grad_norm": 36.11913299560547, "learning_rate": 1.9976072360756787e-05, "loss": 1.511, "step": 1950 }, { "epoch": 0.0516293879217525, "grad_norm": 20.601947784423828, "learning_rate": 1.997547888813863e-05, "loss": 1.6204, "step": 1960 }, { "epoch": 0.051892803166251236, "grad_norm": 75.61493682861328, "learning_rate": 1.997487815454221e-05, "loss": 1.5366, "step": 1970 }, { "epoch": 0.05215621841074997, "grad_norm": 18.20416831970215, "learning_rate": 1.997427016040478e-05, "loss": 1.5607, "step": 1980 }, { "epoch": 0.05241963365524872, "grad_norm": 20.338754653930664, "learning_rate": 1.9973654906168894e-05, "loss": 1.4479, "step": 1990 }, { "epoch": 0.05268304889974745, "grad_norm": 48.998695373535156, "learning_rate": 1.9973032392282382e-05, "loss": 1.5047, "step": 2000 }, { "epoch": 0.05294646414424619, "grad_norm": 60.957462310791016, "learning_rate": 1.9972402619198365e-05, "loss": 1.6015, "step": 2010 }, { "epoch": 0.05320987938874493, "grad_norm": 50.907962799072266, "learning_rate": 1.9971765587375236e-05, "loss": 1.7273, "step": 2020 }, { "epoch": 0.05347329463324366, "grad_norm": 29.735641479492188, "learning_rate": 1.997112129727669e-05, "loss": 1.5959, "step": 2030 }, { "epoch": 0.0537367098777424, "grad_norm": 27.526500701904297, "learning_rate": 1.9970469749371688e-05, "loss": 1.5412, "step": 2040 }, { "epoch": 0.05400012512224114, "grad_norm": 10.339805603027344, "learning_rate": 1.9969810944134484e-05, "loss": 1.5584, "step": 2050 }, { "epoch": 0.054263540366739874, "grad_norm": 36.826358795166016, "learning_rate": 1.9969144882044606e-05, "loss": 1.5929, "step": 2060 }, { "epoch": 0.05452695561123861, "grad_norm": 47.29088592529297, "learning_rate": 1.9968471563586874e-05, "loss": 1.6307, "step": 2070 }, { "epoch": 0.05479037085573735, "grad_norm": 17.101884841918945, "learning_rate": 1.996779098925138e-05, "loss": 1.5343, "step": 2080 }, { "epoch": 0.055053786100236084, "grad_norm": 18.82046127319336, "learning_rate": 1.9967103159533508e-05, "loss": 1.4896, "step": 2090 }, { "epoch": 0.05531720134473482, "grad_norm": 6.147754192352295, "learning_rate": 1.9966408074933913e-05, "loss": 1.569, "step": 2100 }, { "epoch": 0.05558061658923356, "grad_norm": 6.583403587341309, "learning_rate": 1.9965705735958537e-05, "loss": 1.4459, "step": 2110 }, { "epoch": 0.0558440318337323, "grad_norm": 15.984086990356445, "learning_rate": 1.99649961431186e-05, "loss": 1.5438, "step": 2120 }, { "epoch": 0.05610744707823104, "grad_norm": 33.396324157714844, "learning_rate": 1.9964279296930602e-05, "loss": 1.5472, "step": 2130 }, { "epoch": 0.056370862322729774, "grad_norm": 28.127849578857422, "learning_rate": 1.9963555197916326e-05, "loss": 1.5113, "step": 2140 }, { "epoch": 0.05663427756722851, "grad_norm": 26.367124557495117, "learning_rate": 1.9962823846602825e-05, "loss": 1.45, "step": 2150 }, { "epoch": 0.05689769281172725, "grad_norm": 15.139095306396484, "learning_rate": 1.996208524352244e-05, "loss": 1.4468, "step": 2160 }, { "epoch": 0.057161108056225984, "grad_norm": 16.23615837097168, "learning_rate": 1.9961339389212786e-05, "loss": 1.411, "step": 2170 }, { "epoch": 0.05742452330072472, "grad_norm": 7.391385078430176, "learning_rate": 1.996058628421676e-05, "loss": 1.6963, "step": 2180 }, { "epoch": 0.05768793854522346, "grad_norm": 29.048534393310547, "learning_rate": 1.9959825929082535e-05, "loss": 1.6069, "step": 2190 }, { "epoch": 0.057951353789722194, "grad_norm": 8.025056838989258, "learning_rate": 1.9959058324363558e-05, "loss": 1.4333, "step": 2200 }, { "epoch": 0.05821476903422093, "grad_norm": 10.638434410095215, "learning_rate": 1.9958283470618555e-05, "loss": 1.5731, "step": 2210 }, { "epoch": 0.05847818427871967, "grad_norm": 10.70621395111084, "learning_rate": 1.9957501368411528e-05, "loss": 1.5306, "step": 2220 }, { "epoch": 0.058741599523218405, "grad_norm": 22.240854263305664, "learning_rate": 1.9956712018311756e-05, "loss": 1.5552, "step": 2230 }, { "epoch": 0.05900501476771714, "grad_norm": 31.46197509765625, "learning_rate": 1.99559154208938e-05, "loss": 1.5237, "step": 2240 }, { "epoch": 0.059268430012215885, "grad_norm": 25.140926361083984, "learning_rate": 1.9955111576737475e-05, "loss": 1.5934, "step": 2250 }, { "epoch": 0.05953184525671462, "grad_norm": 25.01129913330078, "learning_rate": 1.99543004864279e-05, "loss": 1.4935, "step": 2260 }, { "epoch": 0.05979526050121336, "grad_norm": 31.89696502685547, "learning_rate": 1.9953482150555446e-05, "loss": 1.4984, "step": 2270 }, { "epoch": 0.060058675745712095, "grad_norm": 11.756664276123047, "learning_rate": 1.9952656569715766e-05, "loss": 1.5848, "step": 2280 }, { "epoch": 0.06032209099021083, "grad_norm": 25.914997100830078, "learning_rate": 1.9951823744509787e-05, "loss": 1.4692, "step": 2290 }, { "epoch": 0.06058550623470957, "grad_norm": 28.28131866455078, "learning_rate": 1.995098367554371e-05, "loss": 1.4829, "step": 2300 }, { "epoch": 0.060848921479208305, "grad_norm": 10.771398544311523, "learning_rate": 1.9950136363429002e-05, "loss": 1.4676, "step": 2310 }, { "epoch": 0.06111233672370704, "grad_norm": 7.392683029174805, "learning_rate": 1.994928180878241e-05, "loss": 1.629, "step": 2320 }, { "epoch": 0.06137575196820578, "grad_norm": 15.329215049743652, "learning_rate": 1.994842001222595e-05, "loss": 1.4655, "step": 2330 }, { "epoch": 0.061639167212704515, "grad_norm": 28.639211654663086, "learning_rate": 1.9947550974386907e-05, "loss": 1.5131, "step": 2340 }, { "epoch": 0.06190258245720325, "grad_norm": 46.474666595458984, "learning_rate": 1.994667469589784e-05, "loss": 1.5858, "step": 2350 }, { "epoch": 0.06216599770170199, "grad_norm": 13.97075366973877, "learning_rate": 1.9945791177396573e-05, "loss": 1.5628, "step": 2360 }, { "epoch": 0.06242941294620073, "grad_norm": 46.04969787597656, "learning_rate": 1.994490041952621e-05, "loss": 1.5245, "step": 2370 }, { "epoch": 0.06269282819069946, "grad_norm": 17.371849060058594, "learning_rate": 1.9944002422935114e-05, "loss": 1.4347, "step": 2380 }, { "epoch": 0.0629562434351982, "grad_norm": 30.52956199645996, "learning_rate": 1.9943097188276926e-05, "loss": 1.5758, "step": 2390 }, { "epoch": 0.06321965867969694, "grad_norm": 23.86384391784668, "learning_rate": 1.994218471621054e-05, "loss": 1.4984, "step": 2400 }, { "epoch": 0.06348307392419568, "grad_norm": 14.62930965423584, "learning_rate": 1.994126500740014e-05, "loss": 1.5404, "step": 2410 }, { "epoch": 0.06374648916869441, "grad_norm": 15.240072250366211, "learning_rate": 1.9940338062515162e-05, "loss": 1.5485, "step": 2420 }, { "epoch": 0.06400990441319315, "grad_norm": 13.681964874267578, "learning_rate": 1.9939403882230313e-05, "loss": 1.4778, "step": 2430 }, { "epoch": 0.0642733196576919, "grad_norm": 20.497394561767578, "learning_rate": 1.9938462467225564e-05, "loss": 1.5996, "step": 2440 }, { "epoch": 0.06453673490219063, "grad_norm": 28.412490844726562, "learning_rate": 1.9937513818186162e-05, "loss": 1.3901, "step": 2450 }, { "epoch": 0.06480015014668937, "grad_norm": 48.34121322631836, "learning_rate": 1.9936557935802604e-05, "loss": 1.4919, "step": 2460 }, { "epoch": 0.0650635653911881, "grad_norm": 39.16920471191406, "learning_rate": 1.9935594820770663e-05, "loss": 1.5662, "step": 2470 }, { "epoch": 0.06532698063568684, "grad_norm": 8.62296199798584, "learning_rate": 1.9934624473791375e-05, "loss": 1.6143, "step": 2480 }, { "epoch": 0.06559039588018557, "grad_norm": 37.58180236816406, "learning_rate": 1.9933646895571036e-05, "loss": 1.6641, "step": 2490 }, { "epoch": 0.06585381112468432, "grad_norm": 16.27813720703125, "learning_rate": 1.9932662086821215e-05, "loss": 1.4943, "step": 2500 }, { "epoch": 0.06611722636918305, "grad_norm": 28.223989486694336, "learning_rate": 1.993167004825873e-05, "loss": 1.5566, "step": 2510 }, { "epoch": 0.06638064161368179, "grad_norm": 12.884455680847168, "learning_rate": 1.9930670780605673e-05, "loss": 1.527, "step": 2520 }, { "epoch": 0.06664405685818052, "grad_norm": 28.79819679260254, "learning_rate": 1.992966428458939e-05, "loss": 1.6302, "step": 2530 }, { "epoch": 0.06690747210267926, "grad_norm": 24.192474365234375, "learning_rate": 1.992865056094249e-05, "loss": 1.5155, "step": 2540 }, { "epoch": 0.067170887347178, "grad_norm": 7.118890762329102, "learning_rate": 1.9927629610402857e-05, "loss": 1.383, "step": 2550 }, { "epoch": 0.06743430259167674, "grad_norm": 18.268516540527344, "learning_rate": 1.992660143371361e-05, "loss": 1.6273, "step": 2560 }, { "epoch": 0.06769771783617548, "grad_norm": 31.832805633544922, "learning_rate": 1.9925566031623146e-05, "loss": 1.5442, "step": 2570 }, { "epoch": 0.06796113308067421, "grad_norm": 25.95115852355957, "learning_rate": 1.9924523404885118e-05, "loss": 1.6271, "step": 2580 }, { "epoch": 0.06822454832517295, "grad_norm": 27.0264949798584, "learning_rate": 1.9923473554258434e-05, "loss": 1.5992, "step": 2590 }, { "epoch": 0.06848796356967168, "grad_norm": 47.3375358581543, "learning_rate": 1.992241648050726e-05, "loss": 4.5666, "step": 2600 }, { "epoch": 0.06875137881417043, "grad_norm": 96.86378479003906, "learning_rate": 1.992135218440103e-05, "loss": 1.5641, "step": 2610 }, { "epoch": 0.06901479405866916, "grad_norm": 12.757888793945312, "learning_rate": 1.9920280666714413e-05, "loss": 1.5759, "step": 2620 }, { "epoch": 0.0692782093031679, "grad_norm": 18.262788772583008, "learning_rate": 1.991920192822736e-05, "loss": 1.4545, "step": 2630 }, { "epoch": 0.06954162454766663, "grad_norm": 7.8612380027771, "learning_rate": 1.991811596972506e-05, "loss": 1.5043, "step": 2640 }, { "epoch": 0.06980503979216537, "grad_norm": 9.754074096679688, "learning_rate": 1.991702279199797e-05, "loss": 1.4525, "step": 2650 }, { "epoch": 0.0700684550366641, "grad_norm": 34.565391540527344, "learning_rate": 1.9915922395841783e-05, "loss": 1.5317, "step": 2660 }, { "epoch": 0.07033187028116285, "grad_norm": 3.683598279953003, "learning_rate": 1.991481478205747e-05, "loss": 1.4435, "step": 2670 }, { "epoch": 0.07059528552566158, "grad_norm": 9.371899604797363, "learning_rate": 1.9913699951451238e-05, "loss": 1.4105, "step": 2680 }, { "epoch": 0.07085870077016032, "grad_norm": 7.9052324295043945, "learning_rate": 1.9912577904834557e-05, "loss": 1.5331, "step": 2690 }, { "epoch": 0.07112211601465906, "grad_norm": 23.81945037841797, "learning_rate": 1.9911448643024142e-05, "loss": 1.5714, "step": 2700 }, { "epoch": 0.0713855312591578, "grad_norm": 11.587237358093262, "learning_rate": 1.991031216684196e-05, "loss": 1.4574, "step": 2710 }, { "epoch": 0.07164894650365654, "grad_norm": 17.335559844970703, "learning_rate": 1.990916847711524e-05, "loss": 1.6168, "step": 2720 }, { "epoch": 0.07191236174815527, "grad_norm": 13.279732704162598, "learning_rate": 1.990801757467645e-05, "loss": 1.6199, "step": 2730 }, { "epoch": 0.07217577699265401, "grad_norm": 22.874755859375, "learning_rate": 1.9906859460363307e-05, "loss": 1.5208, "step": 2740 }, { "epoch": 0.07243919223715274, "grad_norm": 6.159449100494385, "learning_rate": 1.9905694135018795e-05, "loss": 1.4941, "step": 2750 }, { "epoch": 0.07270260748165149, "grad_norm": 14.45523738861084, "learning_rate": 1.9904521599491122e-05, "loss": 1.5377, "step": 2760 }, { "epoch": 0.07296602272615021, "grad_norm": 19.553503036499023, "learning_rate": 1.9903341854633762e-05, "loss": 1.5445, "step": 2770 }, { "epoch": 0.07322943797064896, "grad_norm": 19.57867431640625, "learning_rate": 1.990215490130543e-05, "loss": 1.4254, "step": 2780 }, { "epoch": 0.07349285321514769, "grad_norm": 13.853572845458984, "learning_rate": 1.9900960740370085e-05, "loss": 1.4432, "step": 2790 }, { "epoch": 0.07375626845964643, "grad_norm": 46.768489837646484, "learning_rate": 1.9899759372696943e-05, "loss": 1.5541, "step": 2800 }, { "epoch": 0.07401968370414516, "grad_norm": 12.32218074798584, "learning_rate": 1.9898550799160454e-05, "loss": 1.5375, "step": 2810 }, { "epoch": 0.0742830989486439, "grad_norm": 36.678226470947266, "learning_rate": 1.9897335020640322e-05, "loss": 1.5565, "step": 2820 }, { "epoch": 0.07454651419314265, "grad_norm": 10.698866844177246, "learning_rate": 1.9896112038021483e-05, "loss": 1.613, "step": 2830 }, { "epoch": 0.07480992943764138, "grad_norm": 17.686052322387695, "learning_rate": 1.9894881852194135e-05, "loss": 1.4683, "step": 2840 }, { "epoch": 0.07507334468214012, "grad_norm": 12.476663589477539, "learning_rate": 1.9893644464053698e-05, "loss": 1.5298, "step": 2850 }, { "epoch": 0.07533675992663885, "grad_norm": 39.018436431884766, "learning_rate": 1.9892399874500857e-05, "loss": 1.6027, "step": 2860 }, { "epoch": 0.0756001751711376, "grad_norm": 15.008227348327637, "learning_rate": 1.989114808444152e-05, "loss": 1.4975, "step": 2870 }, { "epoch": 0.07586359041563633, "grad_norm": 13.855270385742188, "learning_rate": 1.988988909478684e-05, "loss": 1.6336, "step": 2880 }, { "epoch": 0.07612700566013507, "grad_norm": 40.74266815185547, "learning_rate": 1.9888622906453225e-05, "loss": 1.4263, "step": 2890 }, { "epoch": 0.0763904209046338, "grad_norm": 21.152265548706055, "learning_rate": 1.98873495203623e-05, "loss": 1.2614, "step": 2900 }, { "epoch": 0.07665383614913254, "grad_norm": 5.441744327545166, "learning_rate": 1.988606893744095e-05, "loss": 1.5998, "step": 2910 }, { "epoch": 0.07691725139363127, "grad_norm": 30.66222381591797, "learning_rate": 1.9884781158621283e-05, "loss": 1.4894, "step": 2920 }, { "epoch": 0.07718066663813002, "grad_norm": 31.655927658081055, "learning_rate": 1.9883486184840653e-05, "loss": 1.3842, "step": 2930 }, { "epoch": 0.07744408188262875, "grad_norm": 27.149215698242188, "learning_rate": 1.9882184017041646e-05, "loss": 1.631, "step": 2940 }, { "epoch": 0.07770749712712749, "grad_norm": 4.17231559753418, "learning_rate": 1.9880874656172094e-05, "loss": 1.4859, "step": 2950 }, { "epoch": 0.07797091237162623, "grad_norm": 15.104559898376465, "learning_rate": 1.9879558103185053e-05, "loss": 1.5555, "step": 2960 }, { "epoch": 0.07823432761612496, "grad_norm": 8.142437934875488, "learning_rate": 1.987823435903882e-05, "loss": 1.5478, "step": 2970 }, { "epoch": 0.0784977428606237, "grad_norm": 13.499261856079102, "learning_rate": 1.9876903424696923e-05, "loss": 1.5522, "step": 2980 }, { "epoch": 0.07876115810512244, "grad_norm": 41.704341888427734, "learning_rate": 1.9875565301128135e-05, "loss": 1.4805, "step": 2990 }, { "epoch": 0.07902457334962118, "grad_norm": 15.227733612060547, "learning_rate": 1.987421998930644e-05, "loss": 1.4063, "step": 3000 }, { "epoch": 0.07928798859411991, "grad_norm": 4.233680248260498, "learning_rate": 1.987286749021108e-05, "loss": 1.6202, "step": 3010 }, { "epoch": 0.07955140383861865, "grad_norm": 13.817293167114258, "learning_rate": 1.9871507804826505e-05, "loss": 1.4142, "step": 3020 }, { "epoch": 0.07981481908311738, "grad_norm": 27.799190521240234, "learning_rate": 1.9870140934142413e-05, "loss": 1.3596, "step": 3030 }, { "epoch": 0.08007823432761613, "grad_norm": 36.75443649291992, "learning_rate": 1.9868766879153726e-05, "loss": 1.4682, "step": 3040 }, { "epoch": 0.08034164957211486, "grad_norm": 22.045642852783203, "learning_rate": 1.986738564086059e-05, "loss": 1.5231, "step": 3050 }, { "epoch": 0.0806050648166136, "grad_norm": 17.84931755065918, "learning_rate": 1.9865997220268385e-05, "loss": 1.5785, "step": 3060 }, { "epoch": 0.08086848006111233, "grad_norm": 27.923755645751953, "learning_rate": 1.986460161838772e-05, "loss": 1.4401, "step": 3070 }, { "epoch": 0.08113189530561107, "grad_norm": 23.419239044189453, "learning_rate": 1.9863198836234434e-05, "loss": 1.6572, "step": 3080 }, { "epoch": 0.08139531055010982, "grad_norm": 1.9719781875610352, "learning_rate": 1.9861788874829578e-05, "loss": 1.4697, "step": 3090 }, { "epoch": 0.08165872579460855, "grad_norm": 9.623125076293945, "learning_rate": 1.9860371735199446e-05, "loss": 1.7168, "step": 3100 }, { "epoch": 0.08192214103910729, "grad_norm": 11.467750549316406, "learning_rate": 1.985894741837555e-05, "loss": 1.4969, "step": 3110 }, { "epoch": 0.08218555628360602, "grad_norm": 16.401763916015625, "learning_rate": 1.9857515925394617e-05, "loss": 1.4742, "step": 3120 }, { "epoch": 0.08244897152810476, "grad_norm": 28.620059967041016, "learning_rate": 1.9856077257298616e-05, "loss": 1.5422, "step": 3130 }, { "epoch": 0.0827123867726035, "grad_norm": 11.462540626525879, "learning_rate": 1.9854631415134723e-05, "loss": 1.4457, "step": 3140 }, { "epoch": 0.08297580201710224, "grad_norm": 18.240942001342773, "learning_rate": 1.9853178399955342e-05, "loss": 1.5232, "step": 3150 }, { "epoch": 0.08323921726160097, "grad_norm": 28.66497230529785, "learning_rate": 1.98517182128181e-05, "loss": 1.4572, "step": 3160 }, { "epoch": 0.08350263250609971, "grad_norm": 4.384847640991211, "learning_rate": 1.985025085478584e-05, "loss": 1.356, "step": 3170 }, { "epoch": 0.08376604775059844, "grad_norm": 20.130245208740234, "learning_rate": 1.9848776326926623e-05, "loss": 1.3667, "step": 3180 }, { "epoch": 0.08402946299509718, "grad_norm": 15.157307624816895, "learning_rate": 1.984729463031374e-05, "loss": 1.4746, "step": 3190 }, { "epoch": 0.08429287823959591, "grad_norm": 17.457744598388672, "learning_rate": 1.9845805766025687e-05, "loss": 1.5988, "step": 3200 }, { "epoch": 0.08455629348409466, "grad_norm": 28.916650772094727, "learning_rate": 1.9844309735146184e-05, "loss": 1.4593, "step": 3210 }, { "epoch": 0.0848197087285934, "grad_norm": 6.115513801574707, "learning_rate": 1.9842806538764165e-05, "loss": 1.445, "step": 3220 }, { "epoch": 0.08508312397309213, "grad_norm": 20.28491973876953, "learning_rate": 1.984129617797378e-05, "loss": 1.5044, "step": 3230 }, { "epoch": 0.08534653921759087, "grad_norm": 12.839987754821777, "learning_rate": 1.98397786538744e-05, "loss": 1.4102, "step": 3240 }, { "epoch": 0.0856099544620896, "grad_norm": 32.2998161315918, "learning_rate": 1.98382539675706e-05, "loss": 1.6508, "step": 3250 }, { "epoch": 0.08587336970658835, "grad_norm": 21.531455993652344, "learning_rate": 1.9836722120172172e-05, "loss": 1.4087, "step": 3260 }, { "epoch": 0.08613678495108708, "grad_norm": 8.552406311035156, "learning_rate": 1.983518311279412e-05, "loss": 1.5002, "step": 3270 }, { "epoch": 0.08640020019558582, "grad_norm": 13.527255058288574, "learning_rate": 1.9833636946556663e-05, "loss": 1.5313, "step": 3280 }, { "epoch": 0.08666361544008455, "grad_norm": 34.39051818847656, "learning_rate": 1.983208362258523e-05, "loss": 1.524, "step": 3290 }, { "epoch": 0.0869270306845833, "grad_norm": 16.269207000732422, "learning_rate": 1.983052314201046e-05, "loss": 1.6223, "step": 3300 }, { "epoch": 0.08719044592908202, "grad_norm": 31.936786651611328, "learning_rate": 1.9828955505968195e-05, "loss": 1.4482, "step": 3310 }, { "epoch": 0.08745386117358077, "grad_norm": 61.02764129638672, "learning_rate": 1.982738071559949e-05, "loss": 1.435, "step": 3320 }, { "epoch": 0.0877172764180795, "grad_norm": 16.22569465637207, "learning_rate": 1.9825798772050616e-05, "loss": 1.5219, "step": 3330 }, { "epoch": 0.08798069166257824, "grad_norm": 12.43005084991455, "learning_rate": 1.982420967647303e-05, "loss": 1.527, "step": 3340 }, { "epoch": 0.08824410690707699, "grad_norm": 4.167381763458252, "learning_rate": 1.9822613430023416e-05, "loss": 1.5947, "step": 3350 }, { "epoch": 0.08850752215157572, "grad_norm": 8.551668167114258, "learning_rate": 1.9821010033863648e-05, "loss": 1.61, "step": 3360 }, { "epoch": 0.08877093739607446, "grad_norm": 28.23368263244629, "learning_rate": 1.9819399489160818e-05, "loss": 1.5044, "step": 3370 }, { "epoch": 0.08903435264057319, "grad_norm": 33.19887161254883, "learning_rate": 1.981778179708721e-05, "loss": 1.3536, "step": 3380 }, { "epoch": 0.08929776788507193, "grad_norm": 8.066132545471191, "learning_rate": 1.9816156958820308e-05, "loss": 1.4093, "step": 3390 }, { "epoch": 0.08956118312957066, "grad_norm": 11.486530303955078, "learning_rate": 1.981452497554281e-05, "loss": 1.4888, "step": 3400 }, { "epoch": 0.0898245983740694, "grad_norm": 26.213485717773438, "learning_rate": 1.981288584844261e-05, "loss": 1.4612, "step": 3410 }, { "epoch": 0.09008801361856814, "grad_norm": 7.543020248413086, "learning_rate": 1.98112395787128e-05, "loss": 1.4407, "step": 3420 }, { "epoch": 0.09035142886306688, "grad_norm": 15.23840045928955, "learning_rate": 1.9809586167551664e-05, "loss": 1.5133, "step": 3430 }, { "epoch": 0.09061484410756561, "grad_norm": 8.588821411132812, "learning_rate": 1.98079256161627e-05, "loss": 1.55, "step": 3440 }, { "epoch": 0.09087825935206435, "grad_norm": 27.411909103393555, "learning_rate": 1.9806257925754586e-05, "loss": 1.483, "step": 3450 }, { "epoch": 0.0911416745965631, "grad_norm": 10.379570007324219, "learning_rate": 1.980458309754121e-05, "loss": 1.4193, "step": 3460 }, { "epoch": 0.09140508984106183, "grad_norm": 16.625946044921875, "learning_rate": 1.9802901132741653e-05, "loss": 1.415, "step": 3470 }, { "epoch": 0.09166850508556057, "grad_norm": 16.713918685913086, "learning_rate": 1.9801212032580186e-05, "loss": 1.5911, "step": 3480 }, { "epoch": 0.0919319203300593, "grad_norm": 1.4548166990280151, "learning_rate": 1.9799515798286272e-05, "loss": 1.3815, "step": 3490 }, { "epoch": 0.09219533557455804, "grad_norm": 5.410892486572266, "learning_rate": 1.9797812431094574e-05, "loss": 1.5424, "step": 3500 }, { "epoch": 0.09245875081905677, "grad_norm": 13.775588035583496, "learning_rate": 1.9796101932244944e-05, "loss": 1.43, "step": 3510 }, { "epoch": 0.09272216606355552, "grad_norm": 6.4887614250183105, "learning_rate": 1.9794384302982422e-05, "loss": 1.4723, "step": 3520 }, { "epoch": 0.09298558130805425, "grad_norm": 3.845939874649048, "learning_rate": 1.9792659544557244e-05, "loss": 1.4584, "step": 3530 }, { "epoch": 0.09324899655255299, "grad_norm": 15.996095657348633, "learning_rate": 1.9790927658224825e-05, "loss": 1.5837, "step": 3540 }, { "epoch": 0.09351241179705172, "grad_norm": 12.655293464660645, "learning_rate": 1.9789188645245786e-05, "loss": 1.4761, "step": 3550 }, { "epoch": 0.09377582704155046, "grad_norm": 10.486674308776855, "learning_rate": 1.9787442506885913e-05, "loss": 1.6177, "step": 3560 }, { "epoch": 0.0940392422860492, "grad_norm": 10.727705001831055, "learning_rate": 1.9785689244416198e-05, "loss": 1.5198, "step": 3570 }, { "epoch": 0.09430265753054794, "grad_norm": 26.463388442993164, "learning_rate": 1.9783928859112808e-05, "loss": 1.4462, "step": 3580 }, { "epoch": 0.09456607277504668, "grad_norm": 15.121853828430176, "learning_rate": 1.978216135225709e-05, "loss": 1.4796, "step": 3590 }, { "epoch": 0.09482948801954541, "grad_norm": 18.63764762878418, "learning_rate": 1.9780386725135595e-05, "loss": 1.5065, "step": 3600 }, { "epoch": 0.09509290326404415, "grad_norm": 21.91177749633789, "learning_rate": 1.977860497904004e-05, "loss": 1.7058, "step": 3610 }, { "epoch": 0.09535631850854288, "grad_norm": 38.426536560058594, "learning_rate": 1.9776816115267314e-05, "loss": 1.577, "step": 3620 }, { "epoch": 0.09561973375304163, "grad_norm": 19.14558982849121, "learning_rate": 1.9775020135119513e-05, "loss": 1.6971, "step": 3630 }, { "epoch": 0.09588314899754036, "grad_norm": 39.78761672973633, "learning_rate": 1.9773217039903897e-05, "loss": 1.8455, "step": 3640 }, { "epoch": 0.0961465642420391, "grad_norm": 10.848459243774414, "learning_rate": 1.9771406830932906e-05, "loss": 1.6466, "step": 3650 }, { "epoch": 0.09640997948653783, "grad_norm": 22.35468864440918, "learning_rate": 1.9769589509524163e-05, "loss": 1.4754, "step": 3660 }, { "epoch": 0.09667339473103657, "grad_norm": 56.802330017089844, "learning_rate": 1.976776507700046e-05, "loss": 1.6114, "step": 3670 }, { "epoch": 0.0969368099755353, "grad_norm": 23.9422607421875, "learning_rate": 1.9765933534689775e-05, "loss": 1.5733, "step": 3680 }, { "epoch": 0.09720022522003405, "grad_norm": 11.309412956237793, "learning_rate": 1.976409488392525e-05, "loss": 1.5558, "step": 3690 }, { "epoch": 0.09746364046453278, "grad_norm": 9.368575096130371, "learning_rate": 1.9762249126045214e-05, "loss": 1.4379, "step": 3700 }, { "epoch": 0.09772705570903152, "grad_norm": 7.18232536315918, "learning_rate": 1.9760396262393155e-05, "loss": 1.6052, "step": 3710 }, { "epoch": 0.09799047095353026, "grad_norm": 7.639756202697754, "learning_rate": 1.9758536294317748e-05, "loss": 1.4501, "step": 3720 }, { "epoch": 0.098253886198029, "grad_norm": 24.124452590942383, "learning_rate": 1.975666922317282e-05, "loss": 1.4066, "step": 3730 }, { "epoch": 0.09851730144252774, "grad_norm": 15.980554580688477, "learning_rate": 1.9754795050317398e-05, "loss": 1.4057, "step": 3740 }, { "epoch": 0.09878071668702647, "grad_norm": 24.554969787597656, "learning_rate": 1.975291377711564e-05, "loss": 1.4632, "step": 3750 }, { "epoch": 0.09904413193152521, "grad_norm": 4.167383193969727, "learning_rate": 1.9751025404936907e-05, "loss": 1.3832, "step": 3760 }, { "epoch": 0.09930754717602394, "grad_norm": 12.544156074523926, "learning_rate": 1.9749129935155704e-05, "loss": 1.4631, "step": 3770 }, { "epoch": 0.09957096242052269, "grad_norm": 17.877382278442383, "learning_rate": 1.974722736915171e-05, "loss": 1.4958, "step": 3780 }, { "epoch": 0.09983437766502141, "grad_norm": 17.42013168334961, "learning_rate": 1.974531770830978e-05, "loss": 1.5524, "step": 3790 }, { "epoch": 0.10009779290952016, "grad_norm": 26.338783264160156, "learning_rate": 1.974340095401991e-05, "loss": 1.5187, "step": 3800 }, { "epoch": 0.10036120815401889, "grad_norm": 29.520784378051758, "learning_rate": 1.9741477107677282e-05, "loss": 1.4967, "step": 3810 }, { "epoch": 0.10062462339851763, "grad_norm": 13.96440601348877, "learning_rate": 1.973954617068223e-05, "loss": 1.4196, "step": 3820 }, { "epoch": 0.10088803864301636, "grad_norm": 12.393430709838867, "learning_rate": 1.973760814444024e-05, "loss": 1.5988, "step": 3830 }, { "epoch": 0.1011514538875151, "grad_norm": 18.336427688598633, "learning_rate": 1.9735663030361974e-05, "loss": 1.4195, "step": 3840 }, { "epoch": 0.10141486913201385, "grad_norm": 10.676070213317871, "learning_rate": 1.9733710829863252e-05, "loss": 1.4876, "step": 3850 }, { "epoch": 0.10167828437651258, "grad_norm": 20.098827362060547, "learning_rate": 1.973175154436504e-05, "loss": 1.5484, "step": 3860 }, { "epoch": 0.10194169962101132, "grad_norm": 7.938894748687744, "learning_rate": 1.972978517529347e-05, "loss": 1.5342, "step": 3870 }, { "epoch": 0.10220511486551005, "grad_norm": 14.114721298217773, "learning_rate": 1.9727811724079828e-05, "loss": 1.5417, "step": 3880 }, { "epoch": 0.1024685301100088, "grad_norm": 13.582354545593262, "learning_rate": 1.9725831192160556e-05, "loss": 1.4928, "step": 3890 }, { "epoch": 0.10273194535450753, "grad_norm": 19.619625091552734, "learning_rate": 1.9723843580977246e-05, "loss": 1.4466, "step": 3900 }, { "epoch": 0.10299536059900627, "grad_norm": 6.047610759735107, "learning_rate": 1.9721848891976644e-05, "loss": 1.5625, "step": 3910 }, { "epoch": 0.103258775843505, "grad_norm": 5.205148696899414, "learning_rate": 1.971984712661066e-05, "loss": 1.5215, "step": 3920 }, { "epoch": 0.10352219108800374, "grad_norm": 16.013450622558594, "learning_rate": 1.9717838286336334e-05, "loss": 1.4224, "step": 3930 }, { "epoch": 0.10378560633250247, "grad_norm": 11.868693351745605, "learning_rate": 1.9715822372615867e-05, "loss": 1.4691, "step": 3940 }, { "epoch": 0.10404902157700122, "grad_norm": 10.655168533325195, "learning_rate": 1.9713799386916612e-05, "loss": 1.6906, "step": 3950 }, { "epoch": 0.10431243682149995, "grad_norm": 15.245192527770996, "learning_rate": 1.9711769330711067e-05, "loss": 1.6595, "step": 3960 }, { "epoch": 0.10457585206599869, "grad_norm": 40.055274963378906, "learning_rate": 1.9709732205476867e-05, "loss": 1.5711, "step": 3970 }, { "epoch": 0.10483926731049743, "grad_norm": 37.79780960083008, "learning_rate": 1.9707688012696805e-05, "loss": 1.5253, "step": 3980 }, { "epoch": 0.10510268255499616, "grad_norm": 29.19424057006836, "learning_rate": 1.9705636753858813e-05, "loss": 1.5286, "step": 3990 }, { "epoch": 0.1053660977994949, "grad_norm": 29.660247802734375, "learning_rate": 1.9703578430455966e-05, "loss": 1.4641, "step": 4000 }, { "epoch": 0.10562951304399364, "grad_norm": 9.799918174743652, "learning_rate": 1.9701513043986482e-05, "loss": 1.5903, "step": 4010 }, { "epoch": 0.10589292828849238, "grad_norm": 11.159089088439941, "learning_rate": 1.9699440595953722e-05, "loss": 1.4987, "step": 4020 }, { "epoch": 0.10615634353299111, "grad_norm": 11.761488914489746, "learning_rate": 1.9697361087866182e-05, "loss": 1.5176, "step": 4030 }, { "epoch": 0.10641975877748985, "grad_norm": 3.4428839683532715, "learning_rate": 1.9695274521237502e-05, "loss": 1.5053, "step": 4040 }, { "epoch": 0.10668317402198858, "grad_norm": 26.12874984741211, "learning_rate": 1.9693180897586457e-05, "loss": 1.3169, "step": 4050 }, { "epoch": 0.10694658926648733, "grad_norm": 30.460630416870117, "learning_rate": 1.969108021843696e-05, "loss": 1.5412, "step": 4060 }, { "epoch": 0.10721000451098606, "grad_norm": 14.724608421325684, "learning_rate": 1.9688972485318058e-05, "loss": 1.5537, "step": 4070 }, { "epoch": 0.1074734197554848, "grad_norm": 14.210907936096191, "learning_rate": 1.9686857699763932e-05, "loss": 1.4414, "step": 4080 }, { "epoch": 0.10773683499998353, "grad_norm": 11.68695068359375, "learning_rate": 1.9684735863313904e-05, "loss": 1.4733, "step": 4090 }, { "epoch": 0.10800025024448227, "grad_norm": 12.959821701049805, "learning_rate": 1.9682606977512416e-05, "loss": 1.6705, "step": 4100 }, { "epoch": 0.10826366548898102, "grad_norm": 17.853389739990234, "learning_rate": 1.968047104390905e-05, "loss": 1.3417, "step": 4110 }, { "epoch": 0.10852708073347975, "grad_norm": 13.674656867980957, "learning_rate": 1.967832806405851e-05, "loss": 1.5744, "step": 4120 }, { "epoch": 0.10879049597797849, "grad_norm": 36.04280090332031, "learning_rate": 1.967617803952064e-05, "loss": 1.5232, "step": 4130 }, { "epoch": 0.10905391122247722, "grad_norm": 15.145776748657227, "learning_rate": 1.9674020971860402e-05, "loss": 1.5108, "step": 4140 }, { "epoch": 0.10931732646697596, "grad_norm": 12.25979995727539, "learning_rate": 1.967185686264789e-05, "loss": 1.4892, "step": 4150 }, { "epoch": 0.1095807417114747, "grad_norm": 44.1259651184082, "learning_rate": 1.9669685713458323e-05, "loss": 1.4389, "step": 4160 }, { "epoch": 0.10984415695597344, "grad_norm": 38.08648681640625, "learning_rate": 1.966750752587204e-05, "loss": 1.5078, "step": 4170 }, { "epoch": 0.11010757220047217, "grad_norm": 9.913771629333496, "learning_rate": 1.9665322301474506e-05, "loss": 1.4254, "step": 4180 }, { "epoch": 0.11037098744497091, "grad_norm": 8.879205703735352, "learning_rate": 1.966313004185631e-05, "loss": 1.5044, "step": 4190 }, { "epoch": 0.11063440268946964, "grad_norm": 5.761706352233887, "learning_rate": 1.966093074861316e-05, "loss": 1.5496, "step": 4200 }, { "epoch": 0.11089781793396838, "grad_norm": 14.587227821350098, "learning_rate": 1.965872442334588e-05, "loss": 1.4568, "step": 4210 }, { "epoch": 0.11116123317846711, "grad_norm": 5.551246166229248, "learning_rate": 1.9656511067660417e-05, "loss": 1.3505, "step": 4220 }, { "epoch": 0.11142464842296586, "grad_norm": 3.064429998397827, "learning_rate": 1.965429068316784e-05, "loss": 1.5121, "step": 4230 }, { "epoch": 0.1116880636674646, "grad_norm": 27.130346298217773, "learning_rate": 1.965206327148432e-05, "loss": 1.4931, "step": 4240 }, { "epoch": 0.11195147891196333, "grad_norm": 19.038681030273438, "learning_rate": 1.964982883423116e-05, "loss": 1.4558, "step": 4250 }, { "epoch": 0.11221489415646207, "grad_norm": 10.911199569702148, "learning_rate": 1.9647587373034756e-05, "loss": 1.4834, "step": 4260 }, { "epoch": 0.1124783094009608, "grad_norm": 13.177024841308594, "learning_rate": 1.964533888952664e-05, "loss": 1.5514, "step": 4270 }, { "epoch": 0.11274172464545955, "grad_norm": 20.846553802490234, "learning_rate": 1.964308338534344e-05, "loss": 1.5303, "step": 4280 }, { "epoch": 0.11300513988995828, "grad_norm": 27.264062881469727, "learning_rate": 1.96408208621269e-05, "loss": 1.5342, "step": 4290 }, { "epoch": 0.11326855513445702, "grad_norm": 29.97342300415039, "learning_rate": 1.963855132152387e-05, "loss": 1.3951, "step": 4300 }, { "epoch": 0.11353197037895575, "grad_norm": 7.059381008148193, "learning_rate": 1.9636274765186308e-05, "loss": 1.5764, "step": 4310 }, { "epoch": 0.1137953856234545, "grad_norm": 25.216533660888672, "learning_rate": 1.9633991194771283e-05, "loss": 1.5224, "step": 4320 }, { "epoch": 0.11405880086795323, "grad_norm": 5.253264427185059, "learning_rate": 1.9631700611940966e-05, "loss": 1.4854, "step": 4330 }, { "epoch": 0.11432221611245197, "grad_norm": 16.352340698242188, "learning_rate": 1.9629403018362628e-05, "loss": 1.538, "step": 4340 }, { "epoch": 0.1145856313569507, "grad_norm": 26.118818283081055, "learning_rate": 1.962709841570865e-05, "loss": 1.5862, "step": 4350 }, { "epoch": 0.11484904660144944, "grad_norm": 10.031881332397461, "learning_rate": 1.9624786805656517e-05, "loss": 1.5165, "step": 4360 }, { "epoch": 0.11511246184594819, "grad_norm": 4.70387077331543, "learning_rate": 1.9622468189888805e-05, "loss": 1.5422, "step": 4370 }, { "epoch": 0.11537587709044692, "grad_norm": 6.059713363647461, "learning_rate": 1.9620142570093194e-05, "loss": 1.5215, "step": 4380 }, { "epoch": 0.11563929233494566, "grad_norm": 8.691885948181152, "learning_rate": 1.9617809947962464e-05, "loss": 1.3604, "step": 4390 }, { "epoch": 0.11590270757944439, "grad_norm": 23.859235763549805, "learning_rate": 1.9615470325194495e-05, "loss": 1.4971, "step": 4400 }, { "epoch": 0.11616612282394313, "grad_norm": 29.058679580688477, "learning_rate": 1.9613123703492247e-05, "loss": 1.4667, "step": 4410 }, { "epoch": 0.11642953806844186, "grad_norm": 28.644695281982422, "learning_rate": 1.9610770084563792e-05, "loss": 1.5282, "step": 4420 }, { "epoch": 0.1166929533129406, "grad_norm": 13.615556716918945, "learning_rate": 1.9608409470122288e-05, "loss": 1.3776, "step": 4430 }, { "epoch": 0.11695636855743934, "grad_norm": 19.92193031311035, "learning_rate": 1.960604186188599e-05, "loss": 1.4184, "step": 4440 }, { "epoch": 0.11721978380193808, "grad_norm": 9.075550079345703, "learning_rate": 1.960366726157823e-05, "loss": 1.5302, "step": 4450 }, { "epoch": 0.11748319904643681, "grad_norm": 6.196775913238525, "learning_rate": 1.9601285670927442e-05, "loss": 1.5333, "step": 4460 }, { "epoch": 0.11774661429093555, "grad_norm": 14.592398643493652, "learning_rate": 1.9598897091667144e-05, "loss": 1.4136, "step": 4470 }, { "epoch": 0.11801002953543428, "grad_norm": 15.705561637878418, "learning_rate": 1.959650152553594e-05, "loss": 1.413, "step": 4480 }, { "epoch": 0.11827344477993303, "grad_norm": 15.03614330291748, "learning_rate": 1.959409897427753e-05, "loss": 1.5398, "step": 4490 }, { "epoch": 0.11853686002443177, "grad_norm": 4.167313575744629, "learning_rate": 1.9591689439640678e-05, "loss": 1.4794, "step": 4500 }, { "epoch": 0.1188002752689305, "grad_norm": 19.23577308654785, "learning_rate": 1.958927292337925e-05, "loss": 1.3969, "step": 4510 }, { "epoch": 0.11906369051342924, "grad_norm": 15.120834350585938, "learning_rate": 1.9586849427252188e-05, "loss": 1.5522, "step": 4520 }, { "epoch": 0.11932710575792797, "grad_norm": 9.899072647094727, "learning_rate": 1.9584418953023502e-05, "loss": 1.5639, "step": 4530 }, { "epoch": 0.11959052100242672, "grad_norm": 5.548903465270996, "learning_rate": 1.95819815024623e-05, "loss": 1.5244, "step": 4540 }, { "epoch": 0.11985393624692545, "grad_norm": 14.200756072998047, "learning_rate": 1.9579537077342764e-05, "loss": 1.3268, "step": 4550 }, { "epoch": 0.12011735149142419, "grad_norm": 11.563554763793945, "learning_rate": 1.957708567944414e-05, "loss": 1.5127, "step": 4560 }, { "epoch": 0.12038076673592292, "grad_norm": 4.283222198486328, "learning_rate": 1.9574627310550757e-05, "loss": 1.4877, "step": 4570 }, { "epoch": 0.12064418198042166, "grad_norm": 6.657790184020996, "learning_rate": 1.957216197245203e-05, "loss": 1.4381, "step": 4580 }, { "epoch": 0.1209075972249204, "grad_norm": 17.211997985839844, "learning_rate": 1.956968966694243e-05, "loss": 1.4244, "step": 4590 }, { "epoch": 0.12117101246941914, "grad_norm": 20.857580184936523, "learning_rate": 1.95672103958215e-05, "loss": 1.5535, "step": 4600 }, { "epoch": 0.12143442771391787, "grad_norm": 23.938709259033203, "learning_rate": 1.9564724160893866e-05, "loss": 1.5252, "step": 4610 }, { "epoch": 0.12169784295841661, "grad_norm": 20.435848236083984, "learning_rate": 1.9562230963969213e-05, "loss": 1.4612, "step": 4620 }, { "epoch": 0.12196125820291535, "grad_norm": 25.573760986328125, "learning_rate": 1.9559730806862295e-05, "loss": 1.4839, "step": 4630 }, { "epoch": 0.12222467344741408, "grad_norm": 12.994621276855469, "learning_rate": 1.9557223691392933e-05, "loss": 1.428, "step": 4640 }, { "epoch": 0.12248808869191283, "grad_norm": 23.9560604095459, "learning_rate": 1.9554709619386015e-05, "loss": 1.4504, "step": 4650 }, { "epoch": 0.12275150393641156, "grad_norm": 6.256830215454102, "learning_rate": 1.9552188592671487e-05, "loss": 1.5798, "step": 4660 }, { "epoch": 0.1230149191809103, "grad_norm": 8.829842567443848, "learning_rate": 1.9549660613084366e-05, "loss": 1.5442, "step": 4670 }, { "epoch": 0.12327833442540903, "grad_norm": 38.952789306640625, "learning_rate": 1.954712568246472e-05, "loss": 1.4395, "step": 4680 }, { "epoch": 0.12354174966990777, "grad_norm": 16.142030715942383, "learning_rate": 1.9544583802657683e-05, "loss": 1.3847, "step": 4690 }, { "epoch": 0.1238051649144065, "grad_norm": 8.3980712890625, "learning_rate": 1.9542034975513442e-05, "loss": 1.3783, "step": 4700 }, { "epoch": 0.12406858015890525, "grad_norm": 30.560993194580078, "learning_rate": 1.9539479202887248e-05, "loss": 1.5554, "step": 4710 }, { "epoch": 0.12433199540340398, "grad_norm": 23.749189376831055, "learning_rate": 1.9536916486639405e-05, "loss": 1.5529, "step": 4720 }, { "epoch": 0.12459541064790272, "grad_norm": 24.92456817626953, "learning_rate": 1.9534346828635266e-05, "loss": 1.5334, "step": 4730 }, { "epoch": 0.12485882589240146, "grad_norm": 12.458970069885254, "learning_rate": 1.953177023074524e-05, "loss": 1.5862, "step": 4740 }, { "epoch": 0.1251222411369002, "grad_norm": 16.200027465820312, "learning_rate": 1.9529186694844787e-05, "loss": 1.5699, "step": 4750 }, { "epoch": 0.12538565638139892, "grad_norm": 20.594797134399414, "learning_rate": 1.9526596222814425e-05, "loss": 1.3821, "step": 4760 }, { "epoch": 0.12564907162589767, "grad_norm": 8.395151138305664, "learning_rate": 1.9523998816539706e-05, "loss": 1.5564, "step": 4770 }, { "epoch": 0.1259124868703964, "grad_norm": 10.896271705627441, "learning_rate": 1.9521394477911245e-05, "loss": 1.5579, "step": 4780 }, { "epoch": 0.12617590211489516, "grad_norm": 6.0535993576049805, "learning_rate": 1.9518783208824688e-05, "loss": 1.5197, "step": 4790 }, { "epoch": 0.12643931735939387, "grad_norm": 7.617906093597412, "learning_rate": 1.9516165011180735e-05, "loss": 1.4492, "step": 4800 }, { "epoch": 0.12670273260389261, "grad_norm": 4.157278060913086, "learning_rate": 1.9513539886885128e-05, "loss": 1.4218, "step": 4810 }, { "epoch": 0.12696614784839136, "grad_norm": 17.745332717895508, "learning_rate": 1.9510907837848646e-05, "loss": 1.5949, "step": 4820 }, { "epoch": 0.1272295630928901, "grad_norm": 35.33827209472656, "learning_rate": 1.9508268865987112e-05, "loss": 1.5182, "step": 4830 }, { "epoch": 0.12749297833738882, "grad_norm": 32.935089111328125, "learning_rate": 1.9505622973221394e-05, "loss": 1.5437, "step": 4840 }, { "epoch": 0.12775639358188756, "grad_norm": 17.70271873474121, "learning_rate": 1.950297016147739e-05, "loss": 1.5995, "step": 4850 }, { "epoch": 0.1280198088263863, "grad_norm": 16.751222610473633, "learning_rate": 1.9500310432686026e-05, "loss": 1.4685, "step": 4860 }, { "epoch": 0.12828322407088505, "grad_norm": 11.15281867980957, "learning_rate": 1.9497643788783283e-05, "loss": 1.5412, "step": 4870 }, { "epoch": 0.1285466393153838, "grad_norm": 10.11408805847168, "learning_rate": 1.949497023171016e-05, "loss": 1.4748, "step": 4880 }, { "epoch": 0.1288100545598825, "grad_norm": 5.804219722747803, "learning_rate": 1.94922897634127e-05, "loss": 1.5481, "step": 4890 }, { "epoch": 0.12907346980438125, "grad_norm": 28.850074768066406, "learning_rate": 1.948960238584196e-05, "loss": 1.4965, "step": 4900 }, { "epoch": 0.12933688504888, "grad_norm": 35.23063278198242, "learning_rate": 1.9486908100954035e-05, "loss": 1.4768, "step": 4910 }, { "epoch": 0.12960030029337874, "grad_norm": 15.551379203796387, "learning_rate": 1.948420691071006e-05, "loss": 1.592, "step": 4920 }, { "epoch": 0.12986371553787746, "grad_norm": 12.843875885009766, "learning_rate": 1.948149881707617e-05, "loss": 1.5485, "step": 4930 }, { "epoch": 0.1301271307823762, "grad_norm": 12.876972198486328, "learning_rate": 1.9478783822023555e-05, "loss": 1.4603, "step": 4940 }, { "epoch": 0.13039054602687494, "grad_norm": 12.796365737915039, "learning_rate": 1.94760619275284e-05, "loss": 1.4661, "step": 4950 }, { "epoch": 0.1306539612713737, "grad_norm": 5.561975479125977, "learning_rate": 1.947333313557193e-05, "loss": 1.559, "step": 4960 }, { "epoch": 0.1309173765158724, "grad_norm": 20.76727867126465, "learning_rate": 1.9470597448140383e-05, "loss": 1.501, "step": 4970 }, { "epoch": 0.13118079176037115, "grad_norm": 8.94873332977295, "learning_rate": 1.946785486722502e-05, "loss": 1.5444, "step": 4980 }, { "epoch": 0.1314442070048699, "grad_norm": 13.087124824523926, "learning_rate": 1.9465105394822123e-05, "loss": 1.4132, "step": 4990 }, { "epoch": 0.13170762224936863, "grad_norm": 7.142757892608643, "learning_rate": 1.9462349032932982e-05, "loss": 1.4704, "step": 5000 }, { "epoch": 0.13197103749386738, "grad_norm": 9.076221466064453, "learning_rate": 1.94595857835639e-05, "loss": 1.4631, "step": 5010 }, { "epoch": 0.1322344527383661, "grad_norm": 6.705338954925537, "learning_rate": 1.94568156487262e-05, "loss": 1.435, "step": 5020 }, { "epoch": 0.13249786798286484, "grad_norm": 11.892135620117188, "learning_rate": 1.9454038630436222e-05, "loss": 1.4933, "step": 5030 }, { "epoch": 0.13276128322736358, "grad_norm": 7.921156406402588, "learning_rate": 1.9451254730715302e-05, "loss": 1.5221, "step": 5040 }, { "epoch": 0.13302469847186232, "grad_norm": 5.280564308166504, "learning_rate": 1.9448463951589797e-05, "loss": 1.5345, "step": 5050 }, { "epoch": 0.13328811371636104, "grad_norm": 8.546884536743164, "learning_rate": 1.9445666295091066e-05, "loss": 1.4772, "step": 5060 }, { "epoch": 0.13355152896085978, "grad_norm": 10.15125560760498, "learning_rate": 1.9442861763255473e-05, "loss": 1.4185, "step": 5070 }, { "epoch": 0.13381494420535853, "grad_norm": 25.083036422729492, "learning_rate": 1.944005035812439e-05, "loss": 1.5685, "step": 5080 }, { "epoch": 0.13407835944985727, "grad_norm": 16.347349166870117, "learning_rate": 1.9437232081744183e-05, "loss": 1.3894, "step": 5090 }, { "epoch": 0.134341774694356, "grad_norm": 25.382244110107422, "learning_rate": 1.9434406936166236e-05, "loss": 1.4386, "step": 5100 }, { "epoch": 0.13460518993885473, "grad_norm": 13.641060829162598, "learning_rate": 1.9431574923446924e-05, "loss": 1.4935, "step": 5110 }, { "epoch": 0.13486860518335347, "grad_norm": 30.79372787475586, "learning_rate": 1.9428736045647613e-05, "loss": 1.4224, "step": 5120 }, { "epoch": 0.13513202042785222, "grad_norm": 9.646574020385742, "learning_rate": 1.9425890304834675e-05, "loss": 1.3762, "step": 5130 }, { "epoch": 0.13539543567235096, "grad_norm": 10.041457176208496, "learning_rate": 1.9423037703079476e-05, "loss": 1.4583, "step": 5140 }, { "epoch": 0.13565885091684968, "grad_norm": 10.575789451599121, "learning_rate": 1.9420178242458377e-05, "loss": 1.4995, "step": 5150 }, { "epoch": 0.13592226616134842, "grad_norm": 6.112440586090088, "learning_rate": 1.941731192505273e-05, "loss": 1.4365, "step": 5160 }, { "epoch": 0.13618568140584716, "grad_norm": 13.74649715423584, "learning_rate": 1.9414438752948873e-05, "loss": 1.6049, "step": 5170 }, { "epoch": 0.1364490966503459, "grad_norm": 6.823697090148926, "learning_rate": 1.9411558728238146e-05, "loss": 1.4401, "step": 5180 }, { "epoch": 0.13671251189484462, "grad_norm": 4.455138206481934, "learning_rate": 1.9408671853016866e-05, "loss": 1.5129, "step": 5190 }, { "epoch": 0.13697592713934337, "grad_norm": 5.544444561004639, "learning_rate": 1.9405778129386332e-05, "loss": 1.485, "step": 5200 }, { "epoch": 0.1372393423838421, "grad_norm": 42.43342971801758, "learning_rate": 1.9402877559452845e-05, "loss": 1.5195, "step": 5210 }, { "epoch": 0.13750275762834085, "grad_norm": 6.364267826080322, "learning_rate": 1.9399970145327674e-05, "loss": 1.5061, "step": 5220 }, { "epoch": 0.13776617287283957, "grad_norm": 6.046536445617676, "learning_rate": 1.939705588912708e-05, "loss": 1.5812, "step": 5230 }, { "epoch": 0.13802958811733831, "grad_norm": 12.919092178344727, "learning_rate": 1.93941347929723e-05, "loss": 1.4204, "step": 5240 }, { "epoch": 0.13829300336183706, "grad_norm": 14.953060150146484, "learning_rate": 1.939120685898954e-05, "loss": 1.4664, "step": 5250 }, { "epoch": 0.1385564186063358, "grad_norm": 7.66500997543335, "learning_rate": 1.9388272089310004e-05, "loss": 1.5816, "step": 5260 }, { "epoch": 0.13881983385083455, "grad_norm": 12.277414321899414, "learning_rate": 1.9385330486069854e-05, "loss": 1.4704, "step": 5270 }, { "epoch": 0.13908324909533326, "grad_norm": 8.376077651977539, "learning_rate": 1.938238205141023e-05, "loss": 1.4266, "step": 5280 }, { "epoch": 0.139346664339832, "grad_norm": 15.680191993713379, "learning_rate": 1.9379426787477255e-05, "loss": 1.4479, "step": 5290 }, { "epoch": 0.13961007958433075, "grad_norm": 22.53719711303711, "learning_rate": 1.9376464696422003e-05, "loss": 1.4492, "step": 5300 }, { "epoch": 0.1398734948288295, "grad_norm": 9.848931312561035, "learning_rate": 1.9373495780400534e-05, "loss": 1.4858, "step": 5310 }, { "epoch": 0.1401369100733282, "grad_norm": 14.082700729370117, "learning_rate": 1.9370520041573876e-05, "loss": 1.4869, "step": 5320 }, { "epoch": 0.14040032531782695, "grad_norm": 11.256121635437012, "learning_rate": 1.936753748210801e-05, "loss": 1.5754, "step": 5330 }, { "epoch": 0.1406637405623257, "grad_norm": 14.223258972167969, "learning_rate": 1.9364548104173894e-05, "loss": 1.4487, "step": 5340 }, { "epoch": 0.14092715580682444, "grad_norm": 5.047085285186768, "learning_rate": 1.936155190994744e-05, "loss": 1.4111, "step": 5350 }, { "epoch": 0.14119057105132315, "grad_norm": 11.50040340423584, "learning_rate": 1.9358548901609532e-05, "loss": 1.5046, "step": 5360 }, { "epoch": 0.1414539862958219, "grad_norm": 11.894936561584473, "learning_rate": 1.9355539081346003e-05, "loss": 1.4808, "step": 5370 }, { "epoch": 0.14171740154032064, "grad_norm": 10.629592895507812, "learning_rate": 1.9352522451347647e-05, "loss": 1.4464, "step": 5380 }, { "epoch": 0.14198081678481939, "grad_norm": 18.75467872619629, "learning_rate": 1.9349499013810224e-05, "loss": 1.5933, "step": 5390 }, { "epoch": 0.14224423202931813, "grad_norm": 40.747154235839844, "learning_rate": 1.9346468770934438e-05, "loss": 1.5137, "step": 5400 }, { "epoch": 0.14250764727381685, "grad_norm": 42.30437088012695, "learning_rate": 1.9343431724925948e-05, "loss": 1.5434, "step": 5410 }, { "epoch": 0.1427710625183156, "grad_norm": 18.229022979736328, "learning_rate": 1.934038787799537e-05, "loss": 1.6061, "step": 5420 }, { "epoch": 0.14303447776281433, "grad_norm": 18.547706604003906, "learning_rate": 1.9337337232358268e-05, "loss": 1.5176, "step": 5430 }, { "epoch": 0.14329789300731308, "grad_norm": 14.541882514953613, "learning_rate": 1.9334279790235148e-05, "loss": 1.4847, "step": 5440 }, { "epoch": 0.1435613082518118, "grad_norm": 4.330589771270752, "learning_rate": 1.9331215553851476e-05, "loss": 1.4701, "step": 5450 }, { "epoch": 0.14382472349631054, "grad_norm": 17.344158172607422, "learning_rate": 1.9328144525437657e-05, "loss": 1.5454, "step": 5460 }, { "epoch": 0.14408813874080928, "grad_norm": 16.421100616455078, "learning_rate": 1.9325066707229037e-05, "loss": 1.5498, "step": 5470 }, { "epoch": 0.14435155398530802, "grad_norm": 9.995844841003418, "learning_rate": 1.9321982101465907e-05, "loss": 1.4948, "step": 5480 }, { "epoch": 0.14461496922980674, "grad_norm": 11.987297058105469, "learning_rate": 1.9318890710393497e-05, "loss": 1.4809, "step": 5490 }, { "epoch": 0.14487838447430548, "grad_norm": 4.856269836425781, "learning_rate": 1.9315792536261977e-05, "loss": 1.5004, "step": 5500 }, { "epoch": 0.14514179971880423, "grad_norm": 16.11746597290039, "learning_rate": 1.9312687581326454e-05, "loss": 1.6047, "step": 5510 }, { "epoch": 0.14540521496330297, "grad_norm": 10.342576026916504, "learning_rate": 1.930957584784697e-05, "loss": 1.5072, "step": 5520 }, { "epoch": 0.1456686302078017, "grad_norm": 209.44180297851562, "learning_rate": 1.930645733808851e-05, "loss": 1.5299, "step": 5530 }, { "epoch": 0.14593204545230043, "grad_norm": 77.56652069091797, "learning_rate": 1.930333205432097e-05, "loss": 1.5775, "step": 5540 }, { "epoch": 0.14619546069679917, "grad_norm": 50.099063873291016, "learning_rate": 1.93001999988192e-05, "loss": 1.4323, "step": 5550 }, { "epoch": 0.14645887594129792, "grad_norm": 43.60694885253906, "learning_rate": 1.9297061173862967e-05, "loss": 1.3863, "step": 5560 }, { "epoch": 0.14672229118579666, "grad_norm": 107.82313537597656, "learning_rate": 1.9293915581736966e-05, "loss": 1.4396, "step": 5570 }, { "epoch": 0.14698570643029538, "grad_norm": 309.54583740234375, "learning_rate": 1.9290763224730818e-05, "loss": 1.5672, "step": 5580 }, { "epoch": 0.14724912167479412, "grad_norm": 36.68277359008789, "learning_rate": 1.928760410513907e-05, "loss": 1.4473, "step": 5590 }, { "epoch": 0.14751253691929286, "grad_norm": 16.421710968017578, "learning_rate": 1.928443822526119e-05, "loss": 1.473, "step": 5600 }, { "epoch": 0.1477759521637916, "grad_norm": 46.022254943847656, "learning_rate": 1.928126558740157e-05, "loss": 1.5026, "step": 5610 }, { "epoch": 0.14803936740829032, "grad_norm": 82.41382598876953, "learning_rate": 1.9278086193869514e-05, "loss": 1.3835, "step": 5620 }, { "epoch": 0.14830278265278907, "grad_norm": 24.610567092895508, "learning_rate": 1.927490004697925e-05, "loss": 1.4417, "step": 5630 }, { "epoch": 0.1485661978972878, "grad_norm": 20.438344955444336, "learning_rate": 1.9271707149049917e-05, "loss": 1.4505, "step": 5640 }, { "epoch": 0.14882961314178655, "grad_norm": 43.40306854248047, "learning_rate": 1.9268507502405574e-05, "loss": 1.4176, "step": 5650 }, { "epoch": 0.1490930283862853, "grad_norm": 24.8510684967041, "learning_rate": 1.9265301109375185e-05, "loss": 1.6208, "step": 5660 }, { "epoch": 0.149356443630784, "grad_norm": 30.880521774291992, "learning_rate": 1.9262087972292628e-05, "loss": 1.5051, "step": 5670 }, { "epoch": 0.14961985887528276, "grad_norm": 5.721530914306641, "learning_rate": 1.925886809349669e-05, "loss": 1.4677, "step": 5680 }, { "epoch": 0.1498832741197815, "grad_norm": 5.317755699157715, "learning_rate": 1.9255641475331066e-05, "loss": 1.482, "step": 5690 }, { "epoch": 0.15014668936428024, "grad_norm": 4.381497383117676, "learning_rate": 1.9252408120144357e-05, "loss": 1.2761, "step": 5700 }, { "epoch": 0.15041010460877896, "grad_norm": 26.98045539855957, "learning_rate": 1.9249168030290062e-05, "loss": 1.4058, "step": 5710 }, { "epoch": 0.1506735198532777, "grad_norm": 31.20049285888672, "learning_rate": 1.9245921208126592e-05, "loss": 1.4197, "step": 5720 }, { "epoch": 0.15093693509777645, "grad_norm": 56.033199310302734, "learning_rate": 1.9242667656017247e-05, "loss": 1.5427, "step": 5730 }, { "epoch": 0.1512003503422752, "grad_norm": 17.526260375976562, "learning_rate": 1.9239407376330233e-05, "loss": 1.5271, "step": 5740 }, { "epoch": 0.1514637655867739, "grad_norm": 12.591251373291016, "learning_rate": 1.923614037143865e-05, "loss": 1.3915, "step": 5750 }, { "epoch": 0.15172718083127265, "grad_norm": 6.463188171386719, "learning_rate": 1.9232866643720497e-05, "loss": 1.4193, "step": 5760 }, { "epoch": 0.1519905960757714, "grad_norm": 7.712193012237549, "learning_rate": 1.9229586195558665e-05, "loss": 1.4187, "step": 5770 }, { "epoch": 0.15225401132027014, "grad_norm": 12.221109390258789, "learning_rate": 1.922629902934093e-05, "loss": 1.4172, "step": 5780 }, { "epoch": 0.15251742656476888, "grad_norm": 13.509597778320312, "learning_rate": 1.9223005147459968e-05, "loss": 1.3494, "step": 5790 }, { "epoch": 0.1527808418092676, "grad_norm": 42.09486770629883, "learning_rate": 1.9219704552313335e-05, "loss": 1.5367, "step": 5800 }, { "epoch": 0.15304425705376634, "grad_norm": 12.472993850708008, "learning_rate": 1.9216397246303482e-05, "loss": 1.4669, "step": 5810 }, { "epoch": 0.15330767229826509, "grad_norm": 6.267317771911621, "learning_rate": 1.9213083231837733e-05, "loss": 1.4118, "step": 5820 }, { "epoch": 0.15357108754276383, "grad_norm": 25.158239364624023, "learning_rate": 1.9209762511328308e-05, "loss": 1.4817, "step": 5830 }, { "epoch": 0.15383450278726254, "grad_norm": 3.0872983932495117, "learning_rate": 1.9206435087192296e-05, "loss": 1.3603, "step": 5840 }, { "epoch": 0.1540979180317613, "grad_norm": 14.03447151184082, "learning_rate": 1.9203100961851678e-05, "loss": 1.3983, "step": 5850 }, { "epoch": 0.15436133327626003, "grad_norm": 28.994007110595703, "learning_rate": 1.9199760137733303e-05, "loss": 1.4053, "step": 5860 }, { "epoch": 0.15462474852075878, "grad_norm": 13.246676445007324, "learning_rate": 1.9196412617268895e-05, "loss": 1.4518, "step": 5870 }, { "epoch": 0.1548881637652575, "grad_norm": 18.557640075683594, "learning_rate": 1.9193058402895063e-05, "loss": 1.3854, "step": 5880 }, { "epoch": 0.15515157900975624, "grad_norm": 19.95437240600586, "learning_rate": 1.9189697497053277e-05, "loss": 1.3978, "step": 5890 }, { "epoch": 0.15541499425425498, "grad_norm": 16.190902709960938, "learning_rate": 1.9186329902189885e-05, "loss": 1.4218, "step": 5900 }, { "epoch": 0.15567840949875372, "grad_norm": 12.3582763671875, "learning_rate": 1.9182955620756098e-05, "loss": 1.4526, "step": 5910 }, { "epoch": 0.15594182474325247, "grad_norm": 19.59125518798828, "learning_rate": 1.9179574655208e-05, "loss": 1.4434, "step": 5920 }, { "epoch": 0.15620523998775118, "grad_norm": 7.594522476196289, "learning_rate": 1.9176187008006534e-05, "loss": 1.5077, "step": 5930 }, { "epoch": 0.15646865523224993, "grad_norm": 13.605952262878418, "learning_rate": 1.9172792681617508e-05, "loss": 1.5387, "step": 5940 }, { "epoch": 0.15673207047674867, "grad_norm": 10.045517921447754, "learning_rate": 1.91693916785116e-05, "loss": 1.5417, "step": 5950 }, { "epoch": 0.1569954857212474, "grad_norm": 14.648824691772461, "learning_rate": 1.916598400116434e-05, "loss": 1.4175, "step": 5960 }, { "epoch": 0.15725890096574613, "grad_norm": 11.95168685913086, "learning_rate": 1.9162569652056114e-05, "loss": 1.4761, "step": 5970 }, { "epoch": 0.15752231621024487, "grad_norm": 6.80690860748291, "learning_rate": 1.915914863367217e-05, "loss": 1.4737, "step": 5980 }, { "epoch": 0.15778573145474362, "grad_norm": 19.92470359802246, "learning_rate": 1.9155720948502607e-05, "loss": 1.5266, "step": 5990 }, { "epoch": 0.15804914669924236, "grad_norm": 7.003292560577393, "learning_rate": 1.9152286599042376e-05, "loss": 1.4909, "step": 6000 }, { "epoch": 0.15831256194374108, "grad_norm": 29.643287658691406, "learning_rate": 1.9148845587791282e-05, "loss": 1.5475, "step": 6010 }, { "epoch": 0.15857597718823982, "grad_norm": 7.8319993019104, "learning_rate": 1.9145397917253977e-05, "loss": 1.5377, "step": 6020 }, { "epoch": 0.15883939243273856, "grad_norm": 7.99707555770874, "learning_rate": 1.9141943589939964e-05, "loss": 1.4419, "step": 6030 }, { "epoch": 0.1591028076772373, "grad_norm": 16.28982925415039, "learning_rate": 1.9138482608363583e-05, "loss": 1.4792, "step": 6040 }, { "epoch": 0.15936622292173605, "grad_norm": 7.003361225128174, "learning_rate": 1.9135014975044024e-05, "loss": 1.4198, "step": 6050 }, { "epoch": 0.15962963816623477, "grad_norm": 11.890801429748535, "learning_rate": 1.9131540692505316e-05, "loss": 1.457, "step": 6060 }, { "epoch": 0.1598930534107335, "grad_norm": 9.749465942382812, "learning_rate": 1.9128059763276336e-05, "loss": 1.4492, "step": 6070 }, { "epoch": 0.16015646865523225, "grad_norm": 24.678009033203125, "learning_rate": 1.9124572189890787e-05, "loss": 1.4812, "step": 6080 }, { "epoch": 0.160419883899731, "grad_norm": 24.09703826904297, "learning_rate": 1.9121077974887213e-05, "loss": 1.4545, "step": 6090 }, { "epoch": 0.1606832991442297, "grad_norm": 5.335607051849365, "learning_rate": 1.911757712080899e-05, "loss": 1.5028, "step": 6100 }, { "epoch": 0.16094671438872846, "grad_norm": 7.727433681488037, "learning_rate": 1.9114069630204333e-05, "loss": 1.5123, "step": 6110 }, { "epoch": 0.1612101296332272, "grad_norm": 16.5827693939209, "learning_rate": 1.9110555505626282e-05, "loss": 1.5112, "step": 6120 }, { "epoch": 0.16147354487772594, "grad_norm": 25.386280059814453, "learning_rate": 1.9107034749632708e-05, "loss": 1.4472, "step": 6130 }, { "epoch": 0.16173696012222466, "grad_norm": 7.038825511932373, "learning_rate": 1.9103507364786313e-05, "loss": 1.5119, "step": 6140 }, { "epoch": 0.1620003753667234, "grad_norm": 21.636554718017578, "learning_rate": 1.9099973353654612e-05, "loss": 1.3824, "step": 6150 }, { "epoch": 0.16226379061122215, "grad_norm": 16.58153533935547, "learning_rate": 1.909643271880995e-05, "loss": 1.4172, "step": 6160 }, { "epoch": 0.1625272058557209, "grad_norm": 12.029683113098145, "learning_rate": 1.90928854628295e-05, "loss": 1.4303, "step": 6170 }, { "epoch": 0.16279062110021963, "grad_norm": 15.57595443725586, "learning_rate": 1.9089331588295243e-05, "loss": 1.519, "step": 6180 }, { "epoch": 0.16305403634471835, "grad_norm": 12.059329986572266, "learning_rate": 1.9085771097793986e-05, "loss": 1.4641, "step": 6190 }, { "epoch": 0.1633174515892171, "grad_norm": 8.388490676879883, "learning_rate": 1.9082203993917344e-05, "loss": 1.5327, "step": 6200 }, { "epoch": 0.16358086683371584, "grad_norm": 6.851354122161865, "learning_rate": 1.907863027926176e-05, "loss": 1.4506, "step": 6210 }, { "epoch": 0.16384428207821458, "grad_norm": 8.893203735351562, "learning_rate": 1.907504995642847e-05, "loss": 1.5331, "step": 6220 }, { "epoch": 0.1641076973227133, "grad_norm": 5.783923149108887, "learning_rate": 1.9071463028023527e-05, "loss": 1.4895, "step": 6230 }, { "epoch": 0.16437111256721204, "grad_norm": 7.178958415985107, "learning_rate": 1.9067869496657802e-05, "loss": 1.4212, "step": 6240 }, { "epoch": 0.16463452781171078, "grad_norm": 21.878192901611328, "learning_rate": 1.906426936494696e-05, "loss": 1.503, "step": 6250 }, { "epoch": 0.16489794305620953, "grad_norm": 7.039337158203125, "learning_rate": 1.9060662635511474e-05, "loss": 1.4358, "step": 6260 }, { "epoch": 0.16516135830070824, "grad_norm": 15.890427589416504, "learning_rate": 1.9057049310976618e-05, "loss": 1.5697, "step": 6270 }, { "epoch": 0.165424773545207, "grad_norm": 7.094673156738281, "learning_rate": 1.9053429393972478e-05, "loss": 1.5479, "step": 6280 }, { "epoch": 0.16568818878970573, "grad_norm": 25.221776962280273, "learning_rate": 1.9049802887133917e-05, "loss": 1.4789, "step": 6290 }, { "epoch": 0.16595160403420448, "grad_norm": 21.335376739501953, "learning_rate": 1.9046169793100614e-05, "loss": 1.4038, "step": 6300 }, { "epoch": 0.16621501927870322, "grad_norm": 23.99860191345215, "learning_rate": 1.904253011451703e-05, "loss": 1.4767, "step": 6310 }, { "epoch": 0.16647843452320193, "grad_norm": 15.537790298461914, "learning_rate": 1.903888385403243e-05, "loss": 1.399, "step": 6320 }, { "epoch": 0.16674184976770068, "grad_norm": 12.00252914428711, "learning_rate": 1.9035231014300855e-05, "loss": 1.4695, "step": 6330 }, { "epoch": 0.16700526501219942, "grad_norm": 16.461694717407227, "learning_rate": 1.9031571597981154e-05, "loss": 1.3675, "step": 6340 }, { "epoch": 0.16726868025669817, "grad_norm": 4.490846633911133, "learning_rate": 1.902790560773695e-05, "loss": 1.4101, "step": 6350 }, { "epoch": 0.16753209550119688, "grad_norm": 12.76328182220459, "learning_rate": 1.902423304623665e-05, "loss": 1.4461, "step": 6360 }, { "epoch": 0.16779551074569563, "grad_norm": 10.525606155395508, "learning_rate": 1.9020553916153448e-05, "loss": 1.4715, "step": 6370 }, { "epoch": 0.16805892599019437, "grad_norm": 19.897045135498047, "learning_rate": 1.901686822016532e-05, "loss": 1.3916, "step": 6380 }, { "epoch": 0.1683223412346931, "grad_norm": 22.964447021484375, "learning_rate": 1.9013175960955027e-05, "loss": 1.2664, "step": 6390 }, { "epoch": 0.16858575647919183, "grad_norm": 13.278956413269043, "learning_rate": 1.9009477141210088e-05, "loss": 1.4455, "step": 6400 }, { "epoch": 0.16884917172369057, "grad_norm": 10.78231143951416, "learning_rate": 1.900577176362282e-05, "loss": 1.311, "step": 6410 }, { "epoch": 0.16911258696818932, "grad_norm": 11.10671615600586, "learning_rate": 1.9002059830890302e-05, "loss": 1.3817, "step": 6420 }, { "epoch": 0.16937600221268806, "grad_norm": 13.355144500732422, "learning_rate": 1.8998341345714376e-05, "loss": 1.5502, "step": 6430 }, { "epoch": 0.1696394174571868, "grad_norm": 15.899882316589355, "learning_rate": 1.8994616310801674e-05, "loss": 1.4006, "step": 6440 }, { "epoch": 0.16990283270168552, "grad_norm": 3.653938055038452, "learning_rate": 1.899088472886358e-05, "loss": 1.4882, "step": 6450 }, { "epoch": 0.17016624794618426, "grad_norm": 9.790179252624512, "learning_rate": 1.898714660261625e-05, "loss": 1.5547, "step": 6460 }, { "epoch": 0.170429663190683, "grad_norm": 11.300407409667969, "learning_rate": 1.8983401934780595e-05, "loss": 1.517, "step": 6470 }, { "epoch": 0.17069307843518175, "grad_norm": 28.179386138916016, "learning_rate": 1.8979650728082305e-05, "loss": 1.4514, "step": 6480 }, { "epoch": 0.17095649367968047, "grad_norm": 21.107830047607422, "learning_rate": 1.8975892985251808e-05, "loss": 1.427, "step": 6490 }, { "epoch": 0.1712199089241792, "grad_norm": 2.137543201446533, "learning_rate": 1.8972128709024305e-05, "loss": 1.4659, "step": 6500 }, { "epoch": 0.17148332416867795, "grad_norm": 31.691843032836914, "learning_rate": 1.8968357902139744e-05, "loss": 1.4279, "step": 6510 }, { "epoch": 0.1717467394131767, "grad_norm": 15.218578338623047, "learning_rate": 1.896458056734283e-05, "loss": 1.4986, "step": 6520 }, { "epoch": 0.1720101546576754, "grad_norm": 17.46087074279785, "learning_rate": 1.8960796707383027e-05, "loss": 1.5328, "step": 6530 }, { "epoch": 0.17227356990217416, "grad_norm": 10.051220893859863, "learning_rate": 1.8957006325014536e-05, "loss": 1.4926, "step": 6540 }, { "epoch": 0.1725369851466729, "grad_norm": 7.401064395904541, "learning_rate": 1.895320942299631e-05, "loss": 1.3932, "step": 6550 }, { "epoch": 0.17280040039117164, "grad_norm": 8.164896965026855, "learning_rate": 1.8949406004092045e-05, "loss": 1.6015, "step": 6560 }, { "epoch": 0.1730638156356704, "grad_norm": 11.649212837219238, "learning_rate": 1.894559607107019e-05, "loss": 1.417, "step": 6570 }, { "epoch": 0.1733272308801691, "grad_norm": 7.982486724853516, "learning_rate": 1.8941779626703928e-05, "loss": 1.4785, "step": 6580 }, { "epoch": 0.17359064612466785, "grad_norm": 16.339500427246094, "learning_rate": 1.893795667377118e-05, "loss": 1.4899, "step": 6590 }, { "epoch": 0.1738540613691666, "grad_norm": 2.8750782012939453, "learning_rate": 1.8934127215054604e-05, "loss": 1.5503, "step": 6600 }, { "epoch": 0.17411747661366533, "grad_norm": 25.59984588623047, "learning_rate": 1.8930291253341608e-05, "loss": 1.4165, "step": 6610 }, { "epoch": 0.17438089185816405, "grad_norm": 9.73617935180664, "learning_rate": 1.8926448791424313e-05, "loss": 1.4003, "step": 6620 }, { "epoch": 0.1746443071026628, "grad_norm": 6.157930374145508, "learning_rate": 1.8922599832099585e-05, "loss": 1.4631, "step": 6630 }, { "epoch": 0.17490772234716154, "grad_norm": 28.484355926513672, "learning_rate": 1.8918744378169012e-05, "loss": 1.548, "step": 6640 }, { "epoch": 0.17517113759166028, "grad_norm": 18.83116340637207, "learning_rate": 1.891488243243891e-05, "loss": 1.3877, "step": 6650 }, { "epoch": 0.175434552836159, "grad_norm": 14.265541076660156, "learning_rate": 1.8911013997720327e-05, "loss": 1.4675, "step": 6660 }, { "epoch": 0.17569796808065774, "grad_norm": 23.068174362182617, "learning_rate": 1.8907139076829026e-05, "loss": 1.4116, "step": 6670 }, { "epoch": 0.17596138332515648, "grad_norm": 4.396834850311279, "learning_rate": 1.89032576725855e-05, "loss": 1.4321, "step": 6680 }, { "epoch": 0.17622479856965523, "grad_norm": 14.342424392700195, "learning_rate": 1.8899369787814952e-05, "loss": 1.5086, "step": 6690 }, { "epoch": 0.17648821381415397, "grad_norm": 18.629901885986328, "learning_rate": 1.8895475425347305e-05, "loss": 1.4711, "step": 6700 }, { "epoch": 0.1767516290586527, "grad_norm": 21.915367126464844, "learning_rate": 1.88915745880172e-05, "loss": 1.5166, "step": 6710 }, { "epoch": 0.17701504430315143, "grad_norm": 23.514169692993164, "learning_rate": 1.888766727866399e-05, "loss": 1.3911, "step": 6720 }, { "epoch": 0.17727845954765017, "grad_norm": 21.24363899230957, "learning_rate": 1.888375350013173e-05, "loss": 1.5038, "step": 6730 }, { "epoch": 0.17754187479214892, "grad_norm": 12.076475143432617, "learning_rate": 1.8879833255269203e-05, "loss": 1.4195, "step": 6740 }, { "epoch": 0.17780529003664763, "grad_norm": 22.349382400512695, "learning_rate": 1.8875906546929882e-05, "loss": 1.5543, "step": 6750 }, { "epoch": 0.17806870528114638, "grad_norm": 27.642847061157227, "learning_rate": 1.8871973377971947e-05, "loss": 1.4779, "step": 6760 }, { "epoch": 0.17833212052564512, "grad_norm": 17.915102005004883, "learning_rate": 1.8868033751258288e-05, "loss": 1.4492, "step": 6770 }, { "epoch": 0.17859553577014387, "grad_norm": 19.781291961669922, "learning_rate": 1.886408766965649e-05, "loss": 1.4933, "step": 6780 }, { "epoch": 0.1788589510146426, "grad_norm": 17.94605827331543, "learning_rate": 1.8860135136038836e-05, "loss": 1.4874, "step": 6790 }, { "epoch": 0.17912236625914132, "grad_norm": 10.867156982421875, "learning_rate": 1.885617615328231e-05, "loss": 1.4479, "step": 6800 }, { "epoch": 0.17938578150364007, "grad_norm": 4.8006367683410645, "learning_rate": 1.8852210724268586e-05, "loss": 1.4143, "step": 6810 }, { "epoch": 0.1796491967481388, "grad_norm": 19.57815933227539, "learning_rate": 1.8848238851884028e-05, "loss": 1.4277, "step": 6820 }, { "epoch": 0.17991261199263756, "grad_norm": 11.318155288696289, "learning_rate": 1.8844260539019695e-05, "loss": 1.3884, "step": 6830 }, { "epoch": 0.18017602723713627, "grad_norm": 18.798397064208984, "learning_rate": 1.884027578857133e-05, "loss": 1.432, "step": 6840 }, { "epoch": 0.18043944248163502, "grad_norm": 22.318639755249023, "learning_rate": 1.8836284603439368e-05, "loss": 1.5442, "step": 6850 }, { "epoch": 0.18070285772613376, "grad_norm": 20.5391902923584, "learning_rate": 1.883228698652892e-05, "loss": 1.5079, "step": 6860 }, { "epoch": 0.1809662729706325, "grad_norm": 7.627932071685791, "learning_rate": 1.8828282940749785e-05, "loss": 1.4573, "step": 6870 }, { "epoch": 0.18122968821513122, "grad_norm": 2.4700329303741455, "learning_rate": 1.8824272469016436e-05, "loss": 1.3682, "step": 6880 }, { "epoch": 0.18149310345962996, "grad_norm": 13.579471588134766, "learning_rate": 1.8820255574248027e-05, "loss": 1.4452, "step": 6890 }, { "epoch": 0.1817565187041287, "grad_norm": 17.530471801757812, "learning_rate": 1.8816232259368385e-05, "loss": 1.5167, "step": 6900 }, { "epoch": 0.18201993394862745, "grad_norm": 7.288733959197998, "learning_rate": 1.8812202527306014e-05, "loss": 1.482, "step": 6910 }, { "epoch": 0.1822833491931262, "grad_norm": 12.44015121459961, "learning_rate": 1.8808166380994082e-05, "loss": 1.4378, "step": 6920 }, { "epoch": 0.1825467644376249, "grad_norm": 9.751638412475586, "learning_rate": 1.8804123823370436e-05, "loss": 1.461, "step": 6930 }, { "epoch": 0.18281017968212365, "grad_norm": 16.764854431152344, "learning_rate": 1.8800074857377576e-05, "loss": 1.4777, "step": 6940 }, { "epoch": 0.1830735949266224, "grad_norm": 12.488768577575684, "learning_rate": 1.879601948596268e-05, "loss": 1.4068, "step": 6950 }, { "epoch": 0.18333701017112114, "grad_norm": 6.873770236968994, "learning_rate": 1.879195771207758e-05, "loss": 1.4182, "step": 6960 }, { "epoch": 0.18360042541561986, "grad_norm": 19.818965911865234, "learning_rate": 1.8787889538678775e-05, "loss": 1.4373, "step": 6970 }, { "epoch": 0.1838638406601186, "grad_norm": 14.178266525268555, "learning_rate": 1.878381496872741e-05, "loss": 1.5206, "step": 6980 }, { "epoch": 0.18412725590461734, "grad_norm": 19.88665771484375, "learning_rate": 1.8779734005189302e-05, "loss": 1.4891, "step": 6990 }, { "epoch": 0.1843906711491161, "grad_norm": 4.3031206130981445, "learning_rate": 1.8775646651034906e-05, "loss": 1.463, "step": 7000 }, { "epoch": 0.1846540863936148, "grad_norm": 24.27794647216797, "learning_rate": 1.8771552909239342e-05, "loss": 1.4278, "step": 7010 }, { "epoch": 0.18491750163811355, "grad_norm": 17.979900360107422, "learning_rate": 1.8767452782782372e-05, "loss": 1.4421, "step": 7020 }, { "epoch": 0.1851809168826123, "grad_norm": 10.763005256652832, "learning_rate": 1.876334627464841e-05, "loss": 1.5131, "step": 7030 }, { "epoch": 0.18544433212711103, "grad_norm": 4.550689220428467, "learning_rate": 1.875923338782651e-05, "loss": 1.4459, "step": 7040 }, { "epoch": 0.18570774737160978, "grad_norm": 11.719257354736328, "learning_rate": 1.8755114125310364e-05, "loss": 1.3491, "step": 7050 }, { "epoch": 0.1859711626161085, "grad_norm": 8.02199649810791, "learning_rate": 1.8750988490098323e-05, "loss": 1.5198, "step": 7060 }, { "epoch": 0.18623457786060724, "grad_norm": 21.874345779418945, "learning_rate": 1.8746856485193355e-05, "loss": 1.4509, "step": 7070 }, { "epoch": 0.18649799310510598, "grad_norm": 10.484390258789062, "learning_rate": 1.8742718113603085e-05, "loss": 1.3743, "step": 7080 }, { "epoch": 0.18676140834960472, "grad_norm": 16.07930564880371, "learning_rate": 1.8738573378339758e-05, "loss": 1.5748, "step": 7090 }, { "epoch": 0.18702482359410344, "grad_norm": 20.511184692382812, "learning_rate": 1.8734422282420254e-05, "loss": 1.447, "step": 7100 }, { "epoch": 0.18728823883860218, "grad_norm": 25.905662536621094, "learning_rate": 1.8730264828866084e-05, "loss": 1.4811, "step": 7110 }, { "epoch": 0.18755165408310093, "grad_norm": 14.59699535369873, "learning_rate": 1.8726101020703386e-05, "loss": 1.7118, "step": 7120 }, { "epoch": 0.18781506932759967, "grad_norm": 18.654422760009766, "learning_rate": 1.8721930860962926e-05, "loss": 1.4987, "step": 7130 }, { "epoch": 0.1880784845720984, "grad_norm": 18.301010131835938, "learning_rate": 1.8717754352680087e-05, "loss": 1.6078, "step": 7140 }, { "epoch": 0.18834189981659713, "grad_norm": 2.747265100479126, "learning_rate": 1.8713571498894878e-05, "loss": 1.3748, "step": 7150 }, { "epoch": 0.18860531506109587, "grad_norm": 6.859374046325684, "learning_rate": 1.870938230265193e-05, "loss": 1.5463, "step": 7160 }, { "epoch": 0.18886873030559462, "grad_norm": 23.336410522460938, "learning_rate": 1.8705186767000484e-05, "loss": 1.4062, "step": 7170 }, { "epoch": 0.18913214555009336, "grad_norm": 7.404491901397705, "learning_rate": 1.8700984894994398e-05, "loss": 1.4505, "step": 7180 }, { "epoch": 0.18939556079459208, "grad_norm": 10.213303565979004, "learning_rate": 1.8696776689692137e-05, "loss": 1.5122, "step": 7190 }, { "epoch": 0.18965897603909082, "grad_norm": 9.46784782409668, "learning_rate": 1.869256215415679e-05, "loss": 1.3118, "step": 7200 }, { "epoch": 0.18992239128358956, "grad_norm": 14.267656326293945, "learning_rate": 1.8688341291456035e-05, "loss": 1.4458, "step": 7210 }, { "epoch": 0.1901858065280883, "grad_norm": 4.393117904663086, "learning_rate": 1.868411410466217e-05, "loss": 1.4858, "step": 7220 }, { "epoch": 0.19044922177258702, "grad_norm": 18.1352481842041, "learning_rate": 1.8679880596852092e-05, "loss": 1.4746, "step": 7230 }, { "epoch": 0.19071263701708577, "grad_norm": 31.056867599487305, "learning_rate": 1.8675640771107293e-05, "loss": 1.4183, "step": 7240 }, { "epoch": 0.1909760522615845, "grad_norm": 7.42111873626709, "learning_rate": 1.8671394630513876e-05, "loss": 1.4688, "step": 7250 }, { "epoch": 0.19123946750608325, "grad_norm": 21.90095329284668, "learning_rate": 1.8667142178162526e-05, "loss": 1.4876, "step": 7260 }, { "epoch": 0.19150288275058197, "grad_norm": 9.613465309143066, "learning_rate": 1.8662883417148535e-05, "loss": 1.4538, "step": 7270 }, { "epoch": 0.19176629799508071, "grad_norm": 3.7513937950134277, "learning_rate": 1.8658618350571775e-05, "loss": 1.4316, "step": 7280 }, { "epoch": 0.19202971323957946, "grad_norm": 7.581501007080078, "learning_rate": 1.8654346981536722e-05, "loss": 1.5166, "step": 7290 }, { "epoch": 0.1922931284840782, "grad_norm": 9.13746452331543, "learning_rate": 1.8650069313152426e-05, "loss": 1.4835, "step": 7300 }, { "epoch": 0.19255654372857695, "grad_norm": 9.900339126586914, "learning_rate": 1.864578534853253e-05, "loss": 1.3715, "step": 7310 }, { "epoch": 0.19281995897307566, "grad_norm": 17.260103225708008, "learning_rate": 1.8641495090795258e-05, "loss": 1.2981, "step": 7320 }, { "epoch": 0.1930833742175744, "grad_norm": 3.2433547973632812, "learning_rate": 1.8637198543063407e-05, "loss": 1.5262, "step": 7330 }, { "epoch": 0.19334678946207315, "grad_norm": 14.375350952148438, "learning_rate": 1.863289570846437e-05, "loss": 1.5059, "step": 7340 }, { "epoch": 0.1936102047065719, "grad_norm": 12.896259307861328, "learning_rate": 1.8628586590130102e-05, "loss": 1.4571, "step": 7350 }, { "epoch": 0.1938736199510706, "grad_norm": 23.574466705322266, "learning_rate": 1.8624271191197133e-05, "loss": 1.2972, "step": 7360 }, { "epoch": 0.19413703519556935, "grad_norm": 12.558173179626465, "learning_rate": 1.8619949514806573e-05, "loss": 1.4591, "step": 7370 }, { "epoch": 0.1944004504400681, "grad_norm": 12.582833290100098, "learning_rate": 1.8615621564104084e-05, "loss": 1.4461, "step": 7380 }, { "epoch": 0.19466386568456684, "grad_norm": 11.92378044128418, "learning_rate": 1.8611287342239922e-05, "loss": 1.5034, "step": 7390 }, { "epoch": 0.19492728092906556, "grad_norm": 5.182107925415039, "learning_rate": 1.8606946852368878e-05, "loss": 1.3946, "step": 7400 }, { "epoch": 0.1951906961735643, "grad_norm": 6.762869358062744, "learning_rate": 1.860260009765033e-05, "loss": 1.4579, "step": 7410 }, { "epoch": 0.19545411141806304, "grad_norm": 16.47007942199707, "learning_rate": 1.85982470812482e-05, "loss": 1.3434, "step": 7420 }, { "epoch": 0.19571752666256179, "grad_norm": 3.7040112018585205, "learning_rate": 1.859388780633097e-05, "loss": 1.5624, "step": 7430 }, { "epoch": 0.19598094190706053, "grad_norm": 10.654128074645996, "learning_rate": 1.858952227607169e-05, "loss": 1.587, "step": 7440 }, { "epoch": 0.19624435715155925, "grad_norm": 4.258488655090332, "learning_rate": 1.8585150493647945e-05, "loss": 1.4183, "step": 7450 }, { "epoch": 0.196507772396058, "grad_norm": 5.110781669616699, "learning_rate": 1.8580772462241885e-05, "loss": 1.4875, "step": 7460 }, { "epoch": 0.19677118764055673, "grad_norm": 20.08193588256836, "learning_rate": 1.85763881850402e-05, "loss": 1.5466, "step": 7470 }, { "epoch": 0.19703460288505548, "grad_norm": 11.538524627685547, "learning_rate": 1.857199766523413e-05, "loss": 1.4126, "step": 7480 }, { "epoch": 0.1972980181295542, "grad_norm": 3.3468339443206787, "learning_rate": 1.856760090601946e-05, "loss": 1.3248, "step": 7490 }, { "epoch": 0.19756143337405294, "grad_norm": 7.765769004821777, "learning_rate": 1.8563197910596516e-05, "loss": 1.5416, "step": 7500 }, { "epoch": 0.19782484861855168, "grad_norm": 11.68748950958252, "learning_rate": 1.8558788682170157e-05, "loss": 1.463, "step": 7510 }, { "epoch": 0.19808826386305042, "grad_norm": 6.1479010581970215, "learning_rate": 1.855437322394979e-05, "loss": 1.3357, "step": 7520 }, { "epoch": 0.19835167910754914, "grad_norm": 4.37684440612793, "learning_rate": 1.854995153914935e-05, "loss": 1.4246, "step": 7530 }, { "epoch": 0.19861509435204788, "grad_norm": 8.319536209106445, "learning_rate": 1.8545523630987297e-05, "loss": 1.4105, "step": 7540 }, { "epoch": 0.19887850959654663, "grad_norm": 12.579939842224121, "learning_rate": 1.8541089502686635e-05, "loss": 1.5232, "step": 7550 }, { "epoch": 0.19914192484104537, "grad_norm": 6.631847858428955, "learning_rate": 1.853664915747489e-05, "loss": 1.502, "step": 7560 }, { "epoch": 0.1994053400855441, "grad_norm": 4.5043182373046875, "learning_rate": 1.8532202598584107e-05, "loss": 1.5116, "step": 7570 }, { "epoch": 0.19966875533004283, "grad_norm": 6.453707695007324, "learning_rate": 1.852774982925087e-05, "loss": 1.4139, "step": 7580 }, { "epoch": 0.19993217057454157, "grad_norm": 7.739419460296631, "learning_rate": 1.852329085271626e-05, "loss": 1.414, "step": 7590 }, { "epoch": 0.20019558581904032, "grad_norm": 3.2093746662139893, "learning_rate": 1.8518825672225894e-05, "loss": 1.3792, "step": 7600 }, { "epoch": 0.20045900106353906, "grad_norm": 4.608486652374268, "learning_rate": 1.85143542910299e-05, "loss": 1.5945, "step": 7610 }, { "epoch": 0.20072241630803778, "grad_norm": 20.764062881469727, "learning_rate": 1.8509876712382923e-05, "loss": 1.4966, "step": 7620 }, { "epoch": 0.20098583155253652, "grad_norm": 28.894126892089844, "learning_rate": 1.8505392939544106e-05, "loss": 1.3065, "step": 7630 }, { "epoch": 0.20124924679703526, "grad_norm": 8.782061576843262, "learning_rate": 1.8500902975777113e-05, "loss": 1.3922, "step": 7640 }, { "epoch": 0.201512662041534, "grad_norm": 10.291102409362793, "learning_rate": 1.8496406824350118e-05, "loss": 1.3658, "step": 7650 }, { "epoch": 0.20177607728603272, "grad_norm": 9.185393333435059, "learning_rate": 1.8491904488535782e-05, "loss": 1.5259, "step": 7660 }, { "epoch": 0.20203949253053147, "grad_norm": 6.680044174194336, "learning_rate": 1.848739597161128e-05, "loss": 1.3788, "step": 7670 }, { "epoch": 0.2023029077750302, "grad_norm": 6.952576637268066, "learning_rate": 1.8482881276858286e-05, "loss": 1.3793, "step": 7680 }, { "epoch": 0.20256632301952895, "grad_norm": 11.261947631835938, "learning_rate": 1.8478360407562964e-05, "loss": 1.4661, "step": 7690 }, { "epoch": 0.2028297382640277, "grad_norm": 21.51414680480957, "learning_rate": 1.847383336701598e-05, "loss": 1.4252, "step": 7700 }, { "epoch": 0.20309315350852641, "grad_norm": 23.868408203125, "learning_rate": 1.8469300158512488e-05, "loss": 1.4936, "step": 7710 }, { "epoch": 0.20335656875302516, "grad_norm": 26.173173904418945, "learning_rate": 1.846476078535213e-05, "loss": 1.443, "step": 7720 }, { "epoch": 0.2036199839975239, "grad_norm": 4.012290954589844, "learning_rate": 1.8460215250839044e-05, "loss": 1.5022, "step": 7730 }, { "epoch": 0.20388339924202264, "grad_norm": 7.1301774978637695, "learning_rate": 1.8455663558281834e-05, "loss": 1.5015, "step": 7740 }, { "epoch": 0.20414681448652136, "grad_norm": 23.813858032226562, "learning_rate": 1.8451105710993605e-05, "loss": 1.4628, "step": 7750 }, { "epoch": 0.2044102297310201, "grad_norm": 10.864272117614746, "learning_rate": 1.8446541712291936e-05, "loss": 1.4534, "step": 7760 }, { "epoch": 0.20467364497551885, "grad_norm": 8.994521141052246, "learning_rate": 1.844197156549888e-05, "loss": 1.4285, "step": 7770 }, { "epoch": 0.2049370602200176, "grad_norm": 7.040415287017822, "learning_rate": 1.843739527394097e-05, "loss": 1.4768, "step": 7780 }, { "epoch": 0.2052004754645163, "grad_norm": 11.234932899475098, "learning_rate": 1.8432812840949205e-05, "loss": 1.4001, "step": 7790 }, { "epoch": 0.20546389070901505, "grad_norm": 22.936113357543945, "learning_rate": 1.8428224269859062e-05, "loss": 1.4512, "step": 7800 }, { "epoch": 0.2057273059535138, "grad_norm": 5.005373477935791, "learning_rate": 1.8423629564010485e-05, "loss": 1.4184, "step": 7810 }, { "epoch": 0.20599072119801254, "grad_norm": 23.4188232421875, "learning_rate": 1.8419028726747873e-05, "loss": 1.3978, "step": 7820 }, { "epoch": 0.20625413644251128, "grad_norm": 14.410547256469727, "learning_rate": 1.8414421761420104e-05, "loss": 1.5123, "step": 7830 }, { "epoch": 0.20651755168701, "grad_norm": 25.799779891967773, "learning_rate": 1.8409808671380506e-05, "loss": 1.3974, "step": 7840 }, { "epoch": 0.20678096693150874, "grad_norm": 7.348156929016113, "learning_rate": 1.8405189459986864e-05, "loss": 1.4017, "step": 7850 }, { "epoch": 0.20704438217600749, "grad_norm": 11.87939167022705, "learning_rate": 1.8400564130601424e-05, "loss": 1.5068, "step": 7860 }, { "epoch": 0.20730779742050623, "grad_norm": 8.974173545837402, "learning_rate": 1.8395932686590882e-05, "loss": 1.3974, "step": 7870 }, { "epoch": 0.20757121266500494, "grad_norm": 12.853569984436035, "learning_rate": 1.839129513132639e-05, "loss": 1.4999, "step": 7880 }, { "epoch": 0.2078346279095037, "grad_norm": 13.428462982177734, "learning_rate": 1.8386651468183537e-05, "loss": 1.4251, "step": 7890 }, { "epoch": 0.20809804315400243, "grad_norm": 21.609275817871094, "learning_rate": 1.838200170054237e-05, "loss": 1.3643, "step": 7900 }, { "epoch": 0.20836145839850118, "grad_norm": 6.954063415527344, "learning_rate": 1.8377345831787375e-05, "loss": 1.4328, "step": 7910 }, { "epoch": 0.2086248736429999, "grad_norm": 12.26108455657959, "learning_rate": 1.8372683865307478e-05, "loss": 1.4021, "step": 7920 }, { "epoch": 0.20888828888749864, "grad_norm": 22.034774780273438, "learning_rate": 1.8368015804496042e-05, "loss": 1.4142, "step": 7930 }, { "epoch": 0.20915170413199738, "grad_norm": 9.830514907836914, "learning_rate": 1.8363341652750867e-05, "loss": 1.4216, "step": 7940 }, { "epoch": 0.20941511937649612, "grad_norm": 17.104509353637695, "learning_rate": 1.8358661413474187e-05, "loss": 1.4748, "step": 7950 }, { "epoch": 0.20967853462099487, "grad_norm": 4.474307537078857, "learning_rate": 1.835397509007267e-05, "loss": 1.4474, "step": 7960 }, { "epoch": 0.20994194986549358, "grad_norm": 20.870840072631836, "learning_rate": 1.8349282685957406e-05, "loss": 1.3737, "step": 7970 }, { "epoch": 0.21020536510999233, "grad_norm": 3.135221242904663, "learning_rate": 1.8344584204543915e-05, "loss": 1.4003, "step": 7980 }, { "epoch": 0.21046878035449107, "grad_norm": 19.170940399169922, "learning_rate": 1.8339879649252143e-05, "loss": 1.4519, "step": 7990 }, { "epoch": 0.2107321955989898, "grad_norm": 12.897626876831055, "learning_rate": 1.8335169023506454e-05, "loss": 1.5245, "step": 8000 }, { "epoch": 0.21099561084348853, "grad_norm": 4.79339075088501, "learning_rate": 1.8330452330735628e-05, "loss": 1.4311, "step": 8010 }, { "epoch": 0.21125902608798727, "grad_norm": 18.399019241333008, "learning_rate": 1.8325729574372868e-05, "loss": 1.5134, "step": 8020 }, { "epoch": 0.21152244133248602, "grad_norm": 14.639631271362305, "learning_rate": 1.832100075785578e-05, "loss": 1.5236, "step": 8030 }, { "epoch": 0.21178585657698476, "grad_norm": 9.048181533813477, "learning_rate": 1.8316265884626398e-05, "loss": 1.4059, "step": 8040 }, { "epoch": 0.21204927182148348, "grad_norm": 10.649868965148926, "learning_rate": 1.8311524958131142e-05, "loss": 1.4439, "step": 8050 }, { "epoch": 0.21231268706598222, "grad_norm": 7.749479293823242, "learning_rate": 1.830677798182086e-05, "loss": 1.3566, "step": 8060 }, { "epoch": 0.21257610231048096, "grad_norm": 19.676347732543945, "learning_rate": 1.830202495915079e-05, "loss": 1.3192, "step": 8070 }, { "epoch": 0.2128395175549797, "grad_norm": 20.5679931640625, "learning_rate": 1.829726589358057e-05, "loss": 1.5363, "step": 8080 }, { "epoch": 0.21310293279947845, "grad_norm": 14.283836364746094, "learning_rate": 1.829250078857425e-05, "loss": 1.4819, "step": 8090 }, { "epoch": 0.21336634804397717, "grad_norm": 14.665865898132324, "learning_rate": 1.8287729647600267e-05, "loss": 1.5937, "step": 8100 }, { "epoch": 0.2136297632884759, "grad_norm": 19.260404586791992, "learning_rate": 1.8282952474131442e-05, "loss": 1.4631, "step": 8110 }, { "epoch": 0.21389317853297465, "grad_norm": 13.636507034301758, "learning_rate": 1.8278169271645007e-05, "loss": 1.3554, "step": 8120 }, { "epoch": 0.2141565937774734, "grad_norm": 11.58022403717041, "learning_rate": 1.8273380043622572e-05, "loss": 1.4486, "step": 8130 }, { "epoch": 0.2144200090219721, "grad_norm": 16.682479858398438, "learning_rate": 1.8268584793550124e-05, "loss": 1.5198, "step": 8140 }, { "epoch": 0.21468342426647086, "grad_norm": 18.060522079467773, "learning_rate": 1.8263783524918047e-05, "loss": 1.4521, "step": 8150 }, { "epoch": 0.2149468395109696, "grad_norm": 1.8772904872894287, "learning_rate": 1.825897624122111e-05, "loss": 1.4031, "step": 8160 }, { "epoch": 0.21521025475546834, "grad_norm": 10.103849411010742, "learning_rate": 1.825416294595844e-05, "loss": 1.4261, "step": 8170 }, { "epoch": 0.21547366999996706, "grad_norm": 10.959770202636719, "learning_rate": 1.8249343642633556e-05, "loss": 1.3749, "step": 8180 }, { "epoch": 0.2157370852444658, "grad_norm": 21.110109329223633, "learning_rate": 1.824451833475434e-05, "loss": 1.4826, "step": 8190 }, { "epoch": 0.21600050048896455, "grad_norm": 19.7383975982666, "learning_rate": 1.823968702583306e-05, "loss": 1.5701, "step": 8200 }, { "epoch": 0.2162639157334633, "grad_norm": 11.008354187011719, "learning_rate": 1.8234849719386333e-05, "loss": 1.4694, "step": 8210 }, { "epoch": 0.21652733097796203, "grad_norm": 6.011826038360596, "learning_rate": 1.8230006418935157e-05, "loss": 1.4046, "step": 8220 }, { "epoch": 0.21679074622246075, "grad_norm": 7.7159647941589355, "learning_rate": 1.822515712800488e-05, "loss": 1.4031, "step": 8230 }, { "epoch": 0.2170541614669595, "grad_norm": 3.4496095180511475, "learning_rate": 1.8220301850125226e-05, "loss": 1.3999, "step": 8240 }, { "epoch": 0.21731757671145824, "grad_norm": 12.347916603088379, "learning_rate": 1.821544058883026e-05, "loss": 1.3857, "step": 8250 }, { "epoch": 0.21758099195595698, "grad_norm": 6.3514838218688965, "learning_rate": 1.8210573347658413e-05, "loss": 1.5068, "step": 8260 }, { "epoch": 0.2178444072004557, "grad_norm": 2.572840690612793, "learning_rate": 1.8205700130152465e-05, "loss": 1.3398, "step": 8270 }, { "epoch": 0.21810782244495444, "grad_norm": 5.786165714263916, "learning_rate": 1.8200820939859544e-05, "loss": 1.4208, "step": 8280 }, { "epoch": 0.21837123768945318, "grad_norm": 7.4862189292907715, "learning_rate": 1.8195935780331128e-05, "loss": 1.4727, "step": 8290 }, { "epoch": 0.21863465293395193, "grad_norm": 6.145331859588623, "learning_rate": 1.819104465512305e-05, "loss": 1.3684, "step": 8300 }, { "epoch": 0.21889806817845064, "grad_norm": 7.7216057777404785, "learning_rate": 1.8186147567795462e-05, "loss": 1.3679, "step": 8310 }, { "epoch": 0.2191614834229494, "grad_norm": 7.676534175872803, "learning_rate": 1.8181244521912874e-05, "loss": 1.4197, "step": 8320 }, { "epoch": 0.21942489866744813, "grad_norm": 5.494992256164551, "learning_rate": 1.817633552104413e-05, "loss": 1.3918, "step": 8330 }, { "epoch": 0.21968831391194688, "grad_norm": 17.316965103149414, "learning_rate": 1.817142056876241e-05, "loss": 1.5457, "step": 8340 }, { "epoch": 0.21995172915644562, "grad_norm": 5.5368266105651855, "learning_rate": 1.8166499668645213e-05, "loss": 1.55, "step": 8350 }, { "epoch": 0.22021514440094433, "grad_norm": 12.39454460144043, "learning_rate": 1.8161572824274384e-05, "loss": 1.4262, "step": 8360 }, { "epoch": 0.22047855964544308, "grad_norm": 3.8247854709625244, "learning_rate": 1.8156640039236085e-05, "loss": 1.3617, "step": 8370 }, { "epoch": 0.22074197488994182, "grad_norm": 10.627710342407227, "learning_rate": 1.8151701317120808e-05, "loss": 1.4845, "step": 8380 }, { "epoch": 0.22100539013444057, "grad_norm": 4.776463508605957, "learning_rate": 1.814675666152336e-05, "loss": 1.4815, "step": 8390 }, { "epoch": 0.22126880537893928, "grad_norm": 9.285021781921387, "learning_rate": 1.814180607604287e-05, "loss": 1.3421, "step": 8400 }, { "epoch": 0.22153222062343803, "grad_norm": 9.285212516784668, "learning_rate": 1.8136849564282785e-05, "loss": 1.5045, "step": 8410 }, { "epoch": 0.22179563586793677, "grad_norm": 5.17016077041626, "learning_rate": 1.8131887129850862e-05, "loss": 1.345, "step": 8420 }, { "epoch": 0.2220590511124355, "grad_norm": 3.511632204055786, "learning_rate": 1.812691877635917e-05, "loss": 1.4535, "step": 8430 }, { "epoch": 0.22232246635693423, "grad_norm": 13.215763092041016, "learning_rate": 1.8121944507424093e-05, "loss": 1.4652, "step": 8440 }, { "epoch": 0.22258588160143297, "grad_norm": 22.14901351928711, "learning_rate": 1.8116964326666314e-05, "loss": 1.4009, "step": 8450 }, { "epoch": 0.22284929684593172, "grad_norm": 6.4849653244018555, "learning_rate": 1.8111978237710817e-05, "loss": 1.4193, "step": 8460 }, { "epoch": 0.22311271209043046, "grad_norm": 4.807112216949463, "learning_rate": 1.810698624418689e-05, "loss": 1.3861, "step": 8470 }, { "epoch": 0.2233761273349292, "grad_norm": 12.223381996154785, "learning_rate": 1.810198834972812e-05, "loss": 1.4903, "step": 8480 }, { "epoch": 0.22363954257942792, "grad_norm": 6.607932090759277, "learning_rate": 1.809698455797239e-05, "loss": 1.3941, "step": 8490 }, { "epoch": 0.22390295782392666, "grad_norm": 11.637933731079102, "learning_rate": 1.8091974872561868e-05, "loss": 1.4079, "step": 8500 }, { "epoch": 0.2241663730684254, "grad_norm": 11.12977409362793, "learning_rate": 1.808695929714302e-05, "loss": 1.4067, "step": 8510 }, { "epoch": 0.22442978831292415, "grad_norm": 8.856128692626953, "learning_rate": 1.8081937835366598e-05, "loss": 1.4737, "step": 8520 }, { "epoch": 0.22469320355742287, "grad_norm": 6.893471717834473, "learning_rate": 1.8076910490887635e-05, "loss": 1.4251, "step": 8530 }, { "epoch": 0.2249566188019216, "grad_norm": 6.610567092895508, "learning_rate": 1.8071877267365448e-05, "loss": 1.346, "step": 8540 }, { "epoch": 0.22522003404642035, "grad_norm": 5.463042736053467, "learning_rate": 1.8066838168463637e-05, "loss": 1.426, "step": 8550 }, { "epoch": 0.2254834492909191, "grad_norm": 8.718052864074707, "learning_rate": 1.806179319785007e-05, "loss": 1.4808, "step": 8560 }, { "epoch": 0.2257468645354178, "grad_norm": 2.978691577911377, "learning_rate": 1.805674235919689e-05, "loss": 1.3365, "step": 8570 }, { "epoch": 0.22601027977991656, "grad_norm": 3.123349666595459, "learning_rate": 1.8051685656180528e-05, "loss": 1.4732, "step": 8580 }, { "epoch": 0.2262736950244153, "grad_norm": 14.050543785095215, "learning_rate": 1.804662309248166e-05, "loss": 1.4053, "step": 8590 }, { "epoch": 0.22653711026891404, "grad_norm": 15.999217987060547, "learning_rate": 1.8041554671785237e-05, "loss": 1.5644, "step": 8600 }, { "epoch": 0.2268005255134128, "grad_norm": 3.995640516281128, "learning_rate": 1.8036480397780483e-05, "loss": 1.3877, "step": 8610 }, { "epoch": 0.2270639407579115, "grad_norm": 10.210978507995605, "learning_rate": 1.803140027416087e-05, "loss": 1.356, "step": 8620 }, { "epoch": 0.22732735600241025, "grad_norm": 22.377599716186523, "learning_rate": 1.8026314304624133e-05, "loss": 1.36, "step": 8630 }, { "epoch": 0.227590771246909, "grad_norm": 7.86072301864624, "learning_rate": 1.802122249287226e-05, "loss": 1.4718, "step": 8640 }, { "epoch": 0.22785418649140773, "grad_norm": 9.0934476852417, "learning_rate": 1.801612484261149e-05, "loss": 1.5526, "step": 8650 }, { "epoch": 0.22811760173590645, "grad_norm": 8.159210205078125, "learning_rate": 1.8011021357552317e-05, "loss": 1.4796, "step": 8660 }, { "epoch": 0.2283810169804052, "grad_norm": 2.4113881587982178, "learning_rate": 1.800591204140948e-05, "loss": 1.4665, "step": 8670 }, { "epoch": 0.22864443222490394, "grad_norm": 13.823946952819824, "learning_rate": 1.800079689790196e-05, "loss": 1.4429, "step": 8680 }, { "epoch": 0.22890784746940268, "grad_norm": 8.707887649536133, "learning_rate": 1.7995675930752985e-05, "loss": 1.4694, "step": 8690 }, { "epoch": 0.2291712627139014, "grad_norm": 5.312180042266846, "learning_rate": 1.7990549143690013e-05, "loss": 1.4056, "step": 8700 }, { "epoch": 0.22943467795840014, "grad_norm": 8.792194366455078, "learning_rate": 1.7985416540444747e-05, "loss": 1.4202, "step": 8710 }, { "epoch": 0.22969809320289888, "grad_norm": 8.998208999633789, "learning_rate": 1.7980278124753118e-05, "loss": 1.359, "step": 8720 }, { "epoch": 0.22996150844739763, "grad_norm": 9.951342582702637, "learning_rate": 1.797513390035529e-05, "loss": 1.4365, "step": 8730 }, { "epoch": 0.23022492369189637, "grad_norm": 4.317872524261475, "learning_rate": 1.796998387099566e-05, "loss": 1.3924, "step": 8740 }, { "epoch": 0.2304883389363951, "grad_norm": 12.699377059936523, "learning_rate": 1.7964828040422843e-05, "loss": 1.39, "step": 8750 }, { "epoch": 0.23075175418089383, "grad_norm": 3.693254232406616, "learning_rate": 1.795966641238967e-05, "loss": 1.4199, "step": 8760 }, { "epoch": 0.23101516942539257, "grad_norm": 4.422720909118652, "learning_rate": 1.7954498990653208e-05, "loss": 1.3843, "step": 8770 }, { "epoch": 0.23127858466989132, "grad_norm": 5.710796356201172, "learning_rate": 1.7949325778974737e-05, "loss": 1.3406, "step": 8780 }, { "epoch": 0.23154199991439003, "grad_norm": 15.8676118850708, "learning_rate": 1.794414678111974e-05, "loss": 1.5234, "step": 8790 }, { "epoch": 0.23180541515888878, "grad_norm": 8.596609115600586, "learning_rate": 1.7938962000857924e-05, "loss": 1.3995, "step": 8800 }, { "epoch": 0.23206883040338752, "grad_norm": 13.734110832214355, "learning_rate": 1.7933771441963203e-05, "loss": 1.679, "step": 8810 }, { "epoch": 0.23233224564788627, "grad_norm": 5.3207173347473145, "learning_rate": 1.792857510821369e-05, "loss": 1.4104, "step": 8820 }, { "epoch": 0.23259566089238498, "grad_norm": 5.227005958557129, "learning_rate": 1.7923373003391714e-05, "loss": 1.3838, "step": 8830 }, { "epoch": 0.23285907613688372, "grad_norm": 29.93716812133789, "learning_rate": 1.7918165131283786e-05, "loss": 1.3895, "step": 8840 }, { "epoch": 0.23312249138138247, "grad_norm": 10.15592098236084, "learning_rate": 1.7912951495680634e-05, "loss": 1.5087, "step": 8850 }, { "epoch": 0.2333859066258812, "grad_norm": 27.096113204956055, "learning_rate": 1.790773210037717e-05, "loss": 1.4461, "step": 8860 }, { "epoch": 0.23364932187037996, "grad_norm": 14.67995548248291, "learning_rate": 1.7902506949172503e-05, "loss": 1.4495, "step": 8870 }, { "epoch": 0.23391273711487867, "grad_norm": 24.803268432617188, "learning_rate": 1.789727604586993e-05, "loss": 1.4429, "step": 8880 }, { "epoch": 0.23417615235937742, "grad_norm": 32.97622299194336, "learning_rate": 1.7892039394276934e-05, "loss": 1.3629, "step": 8890 }, { "epoch": 0.23443956760387616, "grad_norm": 11.019721984863281, "learning_rate": 1.7886796998205183e-05, "loss": 1.5197, "step": 8900 }, { "epoch": 0.2347029828483749, "grad_norm": 15.964759826660156, "learning_rate": 1.7881548861470525e-05, "loss": 1.4597, "step": 8910 }, { "epoch": 0.23496639809287362, "grad_norm": 5.054630279541016, "learning_rate": 1.7876294987892992e-05, "loss": 1.4655, "step": 8920 }, { "epoch": 0.23522981333737236, "grad_norm": 9.918245315551758, "learning_rate": 1.7871035381296782e-05, "loss": 1.4636, "step": 8930 }, { "epoch": 0.2354932285818711, "grad_norm": 23.965805053710938, "learning_rate": 1.7865770045510276e-05, "loss": 1.3182, "step": 8940 }, { "epoch": 0.23575664382636985, "grad_norm": 7.497025012969971, "learning_rate": 1.7860498984366018e-05, "loss": 1.5125, "step": 8950 }, { "epoch": 0.23602005907086857, "grad_norm": 6.097912788391113, "learning_rate": 1.7855222201700725e-05, "loss": 1.3319, "step": 8960 }, { "epoch": 0.2362834743153673, "grad_norm": 10.548101425170898, "learning_rate": 1.784993970135527e-05, "loss": 1.4697, "step": 8970 }, { "epoch": 0.23654688955986605, "grad_norm": 24.230907440185547, "learning_rate": 1.78446514871747e-05, "loss": 1.5635, "step": 8980 }, { "epoch": 0.2368103048043648, "grad_norm": 20.92695426940918, "learning_rate": 1.7839357563008214e-05, "loss": 1.4374, "step": 8990 }, { "epoch": 0.23707372004886354, "grad_norm": 16.315082550048828, "learning_rate": 1.7834057932709163e-05, "loss": 1.3859, "step": 9000 }, { "epoch": 0.23733713529336226, "grad_norm": 6.226747989654541, "learning_rate": 1.7828752600135054e-05, "loss": 1.4249, "step": 9010 }, { "epoch": 0.237600550537861, "grad_norm": 10.492033958435059, "learning_rate": 1.7823441569147558e-05, "loss": 1.4275, "step": 9020 }, { "epoch": 0.23786396578235974, "grad_norm": 20.60093879699707, "learning_rate": 1.7818124843612473e-05, "loss": 1.5047, "step": 9030 }, { "epoch": 0.2381273810268585, "grad_norm": 8.934575080871582, "learning_rate": 1.7812802427399753e-05, "loss": 1.2648, "step": 9040 }, { "epoch": 0.2383907962713572, "grad_norm": 3.702556848526001, "learning_rate": 1.7807474324383496e-05, "loss": 1.4232, "step": 9050 }, { "epoch": 0.23865421151585595, "grad_norm": 1.5563379526138306, "learning_rate": 1.780214053844193e-05, "loss": 1.4355, "step": 9060 }, { "epoch": 0.2389176267603547, "grad_norm": 5.938529968261719, "learning_rate": 1.7796801073457436e-05, "loss": 1.4249, "step": 9070 }, { "epoch": 0.23918104200485343, "grad_norm": 23.56593894958496, "learning_rate": 1.7791455933316505e-05, "loss": 1.4843, "step": 9080 }, { "epoch": 0.23944445724935215, "grad_norm": 7.194381237030029, "learning_rate": 1.778610512190978e-05, "loss": 1.4621, "step": 9090 }, { "epoch": 0.2397078724938509, "grad_norm": 8.68651294708252, "learning_rate": 1.7780748643132017e-05, "loss": 1.4509, "step": 9100 }, { "epoch": 0.23997128773834964, "grad_norm": 6.123380661010742, "learning_rate": 1.7775386500882113e-05, "loss": 1.4456, "step": 9110 }, { "epoch": 0.24023470298284838, "grad_norm": 6.692594051361084, "learning_rate": 1.7770018699063075e-05, "loss": 1.2593, "step": 9120 }, { "epoch": 0.24049811822734712, "grad_norm": 12.641352653503418, "learning_rate": 1.776464524158203e-05, "loss": 1.4939, "step": 9130 }, { "epoch": 0.24076153347184584, "grad_norm": 8.251809120178223, "learning_rate": 1.775926613235023e-05, "loss": 1.4759, "step": 9140 }, { "epoch": 0.24102494871634458, "grad_norm": 3.0155279636383057, "learning_rate": 1.775388137528303e-05, "loss": 1.41, "step": 9150 }, { "epoch": 0.24128836396084333, "grad_norm": 13.33884334564209, "learning_rate": 1.7748490974299906e-05, "loss": 1.3403, "step": 9160 }, { "epoch": 0.24155177920534207, "grad_norm": 16.353378295898438, "learning_rate": 1.7743094933324432e-05, "loss": 1.4259, "step": 9170 }, { "epoch": 0.2418151944498408, "grad_norm": 19.454835891723633, "learning_rate": 1.7737693256284298e-05, "loss": 1.4421, "step": 9180 }, { "epoch": 0.24207860969433953, "grad_norm": 5.868311405181885, "learning_rate": 1.773228594711129e-05, "loss": 1.3206, "step": 9190 }, { "epoch": 0.24234202493883827, "grad_norm": 13.859254837036133, "learning_rate": 1.7726873009741293e-05, "loss": 1.3324, "step": 9200 }, { "epoch": 0.24260544018333702, "grad_norm": 3.8151609897613525, "learning_rate": 1.7721454448114287e-05, "loss": 1.5416, "step": 9210 }, { "epoch": 0.24286885542783573, "grad_norm": 8.186395645141602, "learning_rate": 1.7716030266174357e-05, "loss": 1.3844, "step": 9220 }, { "epoch": 0.24313227067233448, "grad_norm": 16.205904006958008, "learning_rate": 1.7710600467869666e-05, "loss": 1.5087, "step": 9230 }, { "epoch": 0.24339568591683322, "grad_norm": 12.89169692993164, "learning_rate": 1.7705165057152476e-05, "loss": 1.46, "step": 9240 }, { "epoch": 0.24365910116133196, "grad_norm": 6.3555402755737305, "learning_rate": 1.769972403797912e-05, "loss": 1.4239, "step": 9250 }, { "epoch": 0.2439225164058307, "grad_norm": 8.18881893157959, "learning_rate": 1.7694277414310027e-05, "loss": 1.4282, "step": 9260 }, { "epoch": 0.24418593165032942, "grad_norm": 2.687380790710449, "learning_rate": 1.7688825190109695e-05, "loss": 1.3653, "step": 9270 }, { "epoch": 0.24444934689482817, "grad_norm": 11.020258903503418, "learning_rate": 1.7683367369346708e-05, "loss": 1.3129, "step": 9280 }, { "epoch": 0.2447127621393269, "grad_norm": 6.068477630615234, "learning_rate": 1.767790395599372e-05, "loss": 1.4425, "step": 9290 }, { "epoch": 0.24497617738382566, "grad_norm": 17.503808975219727, "learning_rate": 1.7672434954027452e-05, "loss": 1.3462, "step": 9300 }, { "epoch": 0.24523959262832437, "grad_norm": 6.320077896118164, "learning_rate": 1.76669603674287e-05, "loss": 1.3758, "step": 9310 }, { "epoch": 0.24550300787282311, "grad_norm": 4.739846706390381, "learning_rate": 1.7661480200182314e-05, "loss": 1.3541, "step": 9320 }, { "epoch": 0.24576642311732186, "grad_norm": 5.5355095863342285, "learning_rate": 1.7655994456277213e-05, "loss": 1.3929, "step": 9330 }, { "epoch": 0.2460298383618206, "grad_norm": 8.930903434753418, "learning_rate": 1.7650503139706387e-05, "loss": 1.3842, "step": 9340 }, { "epoch": 0.24629325360631932, "grad_norm": 38.31852722167969, "learning_rate": 1.7645006254466852e-05, "loss": 1.4781, "step": 9350 }, { "epoch": 0.24655666885081806, "grad_norm": 14.614728927612305, "learning_rate": 1.763950380455971e-05, "loss": 1.3797, "step": 9360 }, { "epoch": 0.2468200840953168, "grad_norm": 9.36707878112793, "learning_rate": 1.7633995793990095e-05, "loss": 1.5341, "step": 9370 }, { "epoch": 0.24708349933981555, "grad_norm": 11.94587230682373, "learning_rate": 1.762848222676719e-05, "loss": 1.345, "step": 9380 }, { "epoch": 0.2473469145843143, "grad_norm": 5.756828784942627, "learning_rate": 1.7622963106904224e-05, "loss": 1.3378, "step": 9390 }, { "epoch": 0.247610329828813, "grad_norm": 14.116488456726074, "learning_rate": 1.7617438438418476e-05, "loss": 1.5249, "step": 9400 }, { "epoch": 0.24787374507331175, "grad_norm": 4.373140335083008, "learning_rate": 1.761190822533125e-05, "loss": 1.3007, "step": 9410 }, { "epoch": 0.2481371603178105, "grad_norm": 11.82955265045166, "learning_rate": 1.7606372471667893e-05, "loss": 1.45, "step": 9420 }, { "epoch": 0.24840057556230924, "grad_norm": 14.190711975097656, "learning_rate": 1.7600831181457785e-05, "loss": 1.4779, "step": 9430 }, { "epoch": 0.24866399080680796, "grad_norm": 8.529472351074219, "learning_rate": 1.759528435873434e-05, "loss": 1.3642, "step": 9440 }, { "epoch": 0.2489274060513067, "grad_norm": 24.774723052978516, "learning_rate": 1.7589732007534984e-05, "loss": 1.461, "step": 9450 }, { "epoch": 0.24919082129580544, "grad_norm": 2.724869728088379, "learning_rate": 1.7584174131901182e-05, "loss": 1.3526, "step": 9460 }, { "epoch": 0.2494542365403042, "grad_norm": 13.814647674560547, "learning_rate": 1.757861073587842e-05, "loss": 1.422, "step": 9470 }, { "epoch": 0.24971765178480293, "grad_norm": 11.529662132263184, "learning_rate": 1.7573041823516194e-05, "loss": 1.4093, "step": 9480 }, { "epoch": 0.24998106702930165, "grad_norm": 15.55040454864502, "learning_rate": 1.7567467398868015e-05, "loss": 1.3466, "step": 9490 }, { "epoch": 0.2502444822738004, "grad_norm": 8.819802284240723, "learning_rate": 1.7561887465991418e-05, "loss": 1.34, "step": 9500 }, { "epoch": 0.2505078975182991, "grad_norm": 11.171469688415527, "learning_rate": 1.7556302028947932e-05, "loss": 1.4147, "step": 9510 }, { "epoch": 0.25077131276279785, "grad_norm": 12.487887382507324, "learning_rate": 1.755071109180311e-05, "loss": 1.3862, "step": 9520 }, { "epoch": 0.2510347280072966, "grad_norm": 13.787839889526367, "learning_rate": 1.754511465862648e-05, "loss": 1.4067, "step": 9530 }, { "epoch": 0.25129814325179534, "grad_norm": 6.18507194519043, "learning_rate": 1.753951273349161e-05, "loss": 1.378, "step": 9540 }, { "epoch": 0.2515615584962941, "grad_norm": 4.558582305908203, "learning_rate": 1.753390532047603e-05, "loss": 1.4965, "step": 9550 }, { "epoch": 0.2518249737407928, "grad_norm": 5.872384071350098, "learning_rate": 1.7528292423661283e-05, "loss": 1.3626, "step": 9560 }, { "epoch": 0.25208838898529157, "grad_norm": 9.10808277130127, "learning_rate": 1.75226740471329e-05, "loss": 1.3496, "step": 9570 }, { "epoch": 0.2523518042297903, "grad_norm": 9.000541687011719, "learning_rate": 1.75170501949804e-05, "loss": 1.4218, "step": 9580 }, { "epoch": 0.252615219474289, "grad_norm": 8.382094383239746, "learning_rate": 1.7511420871297288e-05, "loss": 1.4401, "step": 9590 }, { "epoch": 0.25287863471878774, "grad_norm": 16.166828155517578, "learning_rate": 1.7505786080181043e-05, "loss": 1.4416, "step": 9600 }, { "epoch": 0.2531420499632865, "grad_norm": 21.841997146606445, "learning_rate": 1.7500145825733147e-05, "loss": 1.404, "step": 9610 }, { "epoch": 0.25340546520778523, "grad_norm": 10.182059288024902, "learning_rate": 1.7494500112059033e-05, "loss": 1.5644, "step": 9620 }, { "epoch": 0.253668880452284, "grad_norm": 22.30634117126465, "learning_rate": 1.7488848943268124e-05, "loss": 1.4198, "step": 9630 }, { "epoch": 0.2539322956967827, "grad_norm": 9.819772720336914, "learning_rate": 1.7483192323473802e-05, "loss": 1.4334, "step": 9640 }, { "epoch": 0.25419571094128146, "grad_norm": 9.199249267578125, "learning_rate": 1.7477530256793422e-05, "loss": 1.3408, "step": 9650 }, { "epoch": 0.2544591261857802, "grad_norm": 10.712924003601074, "learning_rate": 1.747186274734831e-05, "loss": 1.4342, "step": 9660 }, { "epoch": 0.25472254143027895, "grad_norm": 17.250553131103516, "learning_rate": 1.7466189799263747e-05, "loss": 1.4293, "step": 9670 }, { "epoch": 0.25498595667477764, "grad_norm": 5.623974323272705, "learning_rate": 1.746051141666897e-05, "loss": 1.4058, "step": 9680 }, { "epoch": 0.2552493719192764, "grad_norm": 24.104249954223633, "learning_rate": 1.745482760369718e-05, "loss": 1.5155, "step": 9690 }, { "epoch": 0.2555127871637751, "grad_norm": 9.512042045593262, "learning_rate": 1.7449138364485524e-05, "loss": 1.4951, "step": 9700 }, { "epoch": 0.25577620240827387, "grad_norm": 7.285160541534424, "learning_rate": 1.7443443703175095e-05, "loss": 1.2372, "step": 9710 }, { "epoch": 0.2560396176527726, "grad_norm": 2.373140335083008, "learning_rate": 1.743774362391095e-05, "loss": 1.5012, "step": 9720 }, { "epoch": 0.25630303289727135, "grad_norm": 16.646223068237305, "learning_rate": 1.7432038130842063e-05, "loss": 1.4207, "step": 9730 }, { "epoch": 0.2565664481417701, "grad_norm": 15.934700012207031, "learning_rate": 1.7426327228121375e-05, "loss": 1.4876, "step": 9740 }, { "epoch": 0.25682986338626884, "grad_norm": 11.875810623168945, "learning_rate": 1.742061091990575e-05, "loss": 1.2346, "step": 9750 }, { "epoch": 0.2570932786307676, "grad_norm": 10.218939781188965, "learning_rate": 1.7414889210355992e-05, "loss": 1.3497, "step": 9760 }, { "epoch": 0.2573566938752663, "grad_norm": 9.011862754821777, "learning_rate": 1.740916210363683e-05, "loss": 1.2256, "step": 9770 }, { "epoch": 0.257620109119765, "grad_norm": 5.259324073791504, "learning_rate": 1.7403429603916927e-05, "loss": 1.4216, "step": 9780 }, { "epoch": 0.25788352436426376, "grad_norm": 10.163860321044922, "learning_rate": 1.7397691715368875e-05, "loss": 1.3095, "step": 9790 }, { "epoch": 0.2581469396087625, "grad_norm": 13.664591789245605, "learning_rate": 1.7391948442169172e-05, "loss": 1.424, "step": 9800 }, { "epoch": 0.25841035485326125, "grad_norm": 15.566182136535645, "learning_rate": 1.738619978849826e-05, "loss": 1.3193, "step": 9810 }, { "epoch": 0.25867377009776, "grad_norm": 6.85333251953125, "learning_rate": 1.738044575854048e-05, "loss": 1.2749, "step": 9820 }, { "epoch": 0.25893718534225874, "grad_norm": 12.405961036682129, "learning_rate": 1.737468635648409e-05, "loss": 1.4487, "step": 9830 }, { "epoch": 0.2592006005867575, "grad_norm": 2.9343132972717285, "learning_rate": 1.7368921586521256e-05, "loss": 1.3707, "step": 9840 }, { "epoch": 0.25946401583125617, "grad_norm": 8.643600463867188, "learning_rate": 1.736315145284806e-05, "loss": 1.4033, "step": 9850 }, { "epoch": 0.2597274310757549, "grad_norm": 15.15362548828125, "learning_rate": 1.7357375959664476e-05, "loss": 1.2536, "step": 9860 }, { "epoch": 0.25999084632025365, "grad_norm": 10.478927612304688, "learning_rate": 1.7351595111174387e-05, "loss": 1.4343, "step": 9870 }, { "epoch": 0.2602542615647524, "grad_norm": 8.922406196594238, "learning_rate": 1.734580891158558e-05, "loss": 1.4897, "step": 9880 }, { "epoch": 0.26051767680925114, "grad_norm": 5.50809383392334, "learning_rate": 1.734001736510972e-05, "loss": 1.4025, "step": 9890 }, { "epoch": 0.2607810920537499, "grad_norm": 13.72231674194336, "learning_rate": 1.7334220475962377e-05, "loss": 1.4673, "step": 9900 }, { "epoch": 0.26104450729824863, "grad_norm": 7.844658851623535, "learning_rate": 1.7328418248363007e-05, "loss": 1.4346, "step": 9910 }, { "epoch": 0.2613079225427474, "grad_norm": 5.5728440284729, "learning_rate": 1.7322610686534952e-05, "loss": 1.3521, "step": 9920 }, { "epoch": 0.2615713377872461, "grad_norm": 9.070394515991211, "learning_rate": 1.7316797794705432e-05, "loss": 1.5188, "step": 9930 }, { "epoch": 0.2618347530317448, "grad_norm": 7.583191871643066, "learning_rate": 1.731097957710555e-05, "loss": 1.3805, "step": 9940 }, { "epoch": 0.26209816827624355, "grad_norm": 7.576006889343262, "learning_rate": 1.7305156037970293e-05, "loss": 1.4032, "step": 9950 }, { "epoch": 0.2623615835207423, "grad_norm": 14.709942817687988, "learning_rate": 1.7299327181538506e-05, "loss": 1.4457, "step": 9960 }, { "epoch": 0.26262499876524104, "grad_norm": 6.896272659301758, "learning_rate": 1.7293493012052917e-05, "loss": 1.3365, "step": 9970 }, { "epoch": 0.2628884140097398, "grad_norm": 15.76815414428711, "learning_rate": 1.7287653533760114e-05, "loss": 1.4193, "step": 9980 }, { "epoch": 0.2631518292542385, "grad_norm": 15.889115333557129, "learning_rate": 1.7281808750910554e-05, "loss": 1.4703, "step": 9990 }, { "epoch": 0.26341524449873727, "grad_norm": 27.858835220336914, "learning_rate": 1.7275958667758553e-05, "loss": 1.4275, "step": 10000 }, { "epoch": 0.263678659743236, "grad_norm": 10.806760787963867, "learning_rate": 1.7270103288562286e-05, "loss": 1.5337, "step": 10010 }, { "epoch": 0.26394207498773475, "grad_norm": 10.30576229095459, "learning_rate": 1.7264242617583778e-05, "loss": 1.428, "step": 10020 }, { "epoch": 0.26420549023223344, "grad_norm": 9.926338195800781, "learning_rate": 1.725837665908892e-05, "loss": 1.3752, "step": 10030 }, { "epoch": 0.2644689054767322, "grad_norm": 7.84537935256958, "learning_rate": 1.725250541734743e-05, "loss": 1.498, "step": 10040 }, { "epoch": 0.26473232072123093, "grad_norm": 9.879533767700195, "learning_rate": 1.7246628896632888e-05, "loss": 1.4404, "step": 10050 }, { "epoch": 0.2649957359657297, "grad_norm": 4.346865653991699, "learning_rate": 1.724074710122271e-05, "loss": 1.4357, "step": 10060 }, { "epoch": 0.2652591512102284, "grad_norm": 3.60611629486084, "learning_rate": 1.7234860035398155e-05, "loss": 1.3869, "step": 10070 }, { "epoch": 0.26552256645472716, "grad_norm": 8.183425903320312, "learning_rate": 1.722896770344432e-05, "loss": 1.5483, "step": 10080 }, { "epoch": 0.2657859816992259, "grad_norm": 14.513091087341309, "learning_rate": 1.722307010965012e-05, "loss": 1.4294, "step": 10090 }, { "epoch": 0.26604939694372465, "grad_norm": 4.75250244140625, "learning_rate": 1.7217167258308323e-05, "loss": 1.3358, "step": 10100 }, { "epoch": 0.26631281218822334, "grad_norm": 17.50128173828125, "learning_rate": 1.7211259153715504e-05, "loss": 1.4284, "step": 10110 }, { "epoch": 0.2665762274327221, "grad_norm": 14.623844146728516, "learning_rate": 1.720534580017208e-05, "loss": 1.4892, "step": 10120 }, { "epoch": 0.2668396426772208, "grad_norm": 15.394331932067871, "learning_rate": 1.7199427201982268e-05, "loss": 1.4276, "step": 10130 }, { "epoch": 0.26710305792171957, "grad_norm": 7.127656936645508, "learning_rate": 1.7193503363454115e-05, "loss": 1.4053, "step": 10140 }, { "epoch": 0.2673664731662183, "grad_norm": 9.381247520446777, "learning_rate": 1.7187574288899486e-05, "loss": 1.3784, "step": 10150 }, { "epoch": 0.26762988841071705, "grad_norm": 5.514267444610596, "learning_rate": 1.7181639982634047e-05, "loss": 1.3601, "step": 10160 }, { "epoch": 0.2678933036552158, "grad_norm": 6.509953498840332, "learning_rate": 1.7175700448977277e-05, "loss": 1.3722, "step": 10170 }, { "epoch": 0.26815671889971454, "grad_norm": 18.630632400512695, "learning_rate": 1.716975569225246e-05, "loss": 1.3553, "step": 10180 }, { "epoch": 0.2684201341442133, "grad_norm": 5.28150749206543, "learning_rate": 1.7163805716786687e-05, "loss": 1.3843, "step": 10190 }, { "epoch": 0.268683549388712, "grad_norm": 12.375866889953613, "learning_rate": 1.715785052691083e-05, "loss": 1.4096, "step": 10200 }, { "epoch": 0.2689469646332107, "grad_norm": 5.365541458129883, "learning_rate": 1.7151890126959583e-05, "loss": 1.4068, "step": 10210 }, { "epoch": 0.26921037987770946, "grad_norm": 22.405471801757812, "learning_rate": 1.7145924521271408e-05, "loss": 1.3548, "step": 10220 }, { "epoch": 0.2694737951222082, "grad_norm": 2.3575403690338135, "learning_rate": 1.7139953714188565e-05, "loss": 1.3288, "step": 10230 }, { "epoch": 0.26973721036670695, "grad_norm": 6.551501750946045, "learning_rate": 1.713397771005711e-05, "loss": 1.3631, "step": 10240 }, { "epoch": 0.2700006256112057, "grad_norm": 7.130479335784912, "learning_rate": 1.7127996513226863e-05, "loss": 1.3245, "step": 10250 }, { "epoch": 0.27026404085570443, "grad_norm": 5.8388352394104, "learning_rate": 1.712201012805144e-05, "loss": 1.4094, "step": 10260 }, { "epoch": 0.2705274561002032, "grad_norm": 25.963781356811523, "learning_rate": 1.7116018558888225e-05, "loss": 1.3182, "step": 10270 }, { "epoch": 0.2707908713447019, "grad_norm": 6.962029933929443, "learning_rate": 1.7110021810098382e-05, "loss": 1.3669, "step": 10280 }, { "epoch": 0.2710542865892006, "grad_norm": 9.447394371032715, "learning_rate": 1.710401988604683e-05, "loss": 1.4282, "step": 10290 }, { "epoch": 0.27131770183369935, "grad_norm": 5.864366054534912, "learning_rate": 1.7098012791102275e-05, "loss": 1.3482, "step": 10300 }, { "epoch": 0.2715811170781981, "grad_norm": 13.406704902648926, "learning_rate": 1.709200052963717e-05, "loss": 1.3161, "step": 10310 }, { "epoch": 0.27184453232269684, "grad_norm": 6.410478591918945, "learning_rate": 1.7085983106027744e-05, "loss": 1.4372, "step": 10320 }, { "epoch": 0.2721079475671956, "grad_norm": 6.863880634307861, "learning_rate": 1.7079960524653972e-05, "loss": 1.3419, "step": 10330 }, { "epoch": 0.27237136281169433, "grad_norm": 8.327899932861328, "learning_rate": 1.7073932789899582e-05, "loss": 1.3873, "step": 10340 }, { "epoch": 0.2726347780561931, "grad_norm": 3.1528546810150146, "learning_rate": 1.7067899906152066e-05, "loss": 1.2646, "step": 10350 }, { "epoch": 0.2728981933006918, "grad_norm": 13.592677116394043, "learning_rate": 1.706186187780265e-05, "loss": 1.4475, "step": 10360 }, { "epoch": 0.2731616085451905, "grad_norm": 3.0780158042907715, "learning_rate": 1.7055818709246307e-05, "loss": 1.4143, "step": 10370 }, { "epoch": 0.27342502378968925, "grad_norm": 18.138545989990234, "learning_rate": 1.7049770404881764e-05, "loss": 1.332, "step": 10380 }, { "epoch": 0.273688439034188, "grad_norm": 14.070090293884277, "learning_rate": 1.7043716969111474e-05, "loss": 1.4209, "step": 10390 }, { "epoch": 0.27395185427868674, "grad_norm": 14.453300476074219, "learning_rate": 1.7037658406341623e-05, "loss": 1.4409, "step": 10400 }, { "epoch": 0.2742152695231855, "grad_norm": 9.975614547729492, "learning_rate": 1.7031594720982134e-05, "loss": 1.4234, "step": 10410 }, { "epoch": 0.2744786847676842, "grad_norm": 23.27369499206543, "learning_rate": 1.7025525917446667e-05, "loss": 1.3255, "step": 10420 }, { "epoch": 0.27474210001218297, "grad_norm": 19.014198303222656, "learning_rate": 1.701945200015259e-05, "loss": 1.4333, "step": 10430 }, { "epoch": 0.2750055152566817, "grad_norm": 6.786633491516113, "learning_rate": 1.701337297352101e-05, "loss": 1.3894, "step": 10440 }, { "epoch": 0.27526893050118045, "grad_norm": 7.586590766906738, "learning_rate": 1.7007288841976736e-05, "loss": 1.3027, "step": 10450 }, { "epoch": 0.27553234574567914, "grad_norm": 10.669122695922852, "learning_rate": 1.700119960994831e-05, "loss": 1.4249, "step": 10460 }, { "epoch": 0.2757957609901779, "grad_norm": 5.211860179901123, "learning_rate": 1.6995105281867973e-05, "loss": 1.519, "step": 10470 }, { "epoch": 0.27605917623467663, "grad_norm": 11.10645866394043, "learning_rate": 1.698900586217169e-05, "loss": 1.4937, "step": 10480 }, { "epoch": 0.2763225914791754, "grad_norm": 31.61640739440918, "learning_rate": 1.6982901355299106e-05, "loss": 1.5356, "step": 10490 }, { "epoch": 0.2765860067236741, "grad_norm": 10.761563301086426, "learning_rate": 1.6976791765693602e-05, "loss": 1.3284, "step": 10500 }, { "epoch": 0.27684942196817286, "grad_norm": 11.6333646774292, "learning_rate": 1.6970677097802236e-05, "loss": 1.4126, "step": 10510 }, { "epoch": 0.2771128372126716, "grad_norm": 18.506364822387695, "learning_rate": 1.6964557356075767e-05, "loss": 1.4142, "step": 10520 }, { "epoch": 0.27737625245717035, "grad_norm": 5.859353065490723, "learning_rate": 1.6958432544968656e-05, "loss": 1.3807, "step": 10530 }, { "epoch": 0.2776396677016691, "grad_norm": 10.047774314880371, "learning_rate": 1.6952302668939038e-05, "loss": 1.4916, "step": 10540 }, { "epoch": 0.2779030829461678, "grad_norm": 12.138692855834961, "learning_rate": 1.694616773244875e-05, "loss": 1.4542, "step": 10550 }, { "epoch": 0.2781664981906665, "grad_norm": 13.369186401367188, "learning_rate": 1.69400277399633e-05, "loss": 1.3316, "step": 10560 }, { "epoch": 0.27842991343516527, "grad_norm": 8.96480655670166, "learning_rate": 1.693388269595189e-05, "loss": 1.441, "step": 10570 }, { "epoch": 0.278693328679664, "grad_norm": 6.4151291847229, "learning_rate": 1.6927732604887385e-05, "loss": 1.4222, "step": 10580 }, { "epoch": 0.27895674392416275, "grad_norm": 68.81958770751953, "learning_rate": 1.692157747124633e-05, "loss": 1.3837, "step": 10590 }, { "epoch": 0.2792201591686615, "grad_norm": 48.135948181152344, "learning_rate": 1.6915417299508947e-05, "loss": 1.3375, "step": 10600 }, { "epoch": 0.27948357441316024, "grad_norm": 6.188029766082764, "learning_rate": 1.6909252094159106e-05, "loss": 1.4013, "step": 10610 }, { "epoch": 0.279746989657659, "grad_norm": 16.440561294555664, "learning_rate": 1.6903081859684362e-05, "loss": 1.5047, "step": 10620 }, { "epoch": 0.2800104049021577, "grad_norm": 10.713726043701172, "learning_rate": 1.6896906600575925e-05, "loss": 1.4008, "step": 10630 }, { "epoch": 0.2802738201466564, "grad_norm": 6.877676963806152, "learning_rate": 1.6890726321328652e-05, "loss": 1.3936, "step": 10640 }, { "epoch": 0.28053723539115516, "grad_norm": 7.687427043914795, "learning_rate": 1.6884541026441058e-05, "loss": 1.4532, "step": 10650 }, { "epoch": 0.2808006506356539, "grad_norm": 7.978679180145264, "learning_rate": 1.6878350720415326e-05, "loss": 1.3041, "step": 10660 }, { "epoch": 0.28106406588015265, "grad_norm": 15.336507797241211, "learning_rate": 1.6872155407757257e-05, "loss": 1.482, "step": 10670 }, { "epoch": 0.2813274811246514, "grad_norm": 21.367015838623047, "learning_rate": 1.6865955092976324e-05, "loss": 1.4064, "step": 10680 }, { "epoch": 0.28159089636915013, "grad_norm": 23.915796279907227, "learning_rate": 1.685974978058562e-05, "loss": 1.4977, "step": 10690 }, { "epoch": 0.2818543116136489, "grad_norm": 9.73816204071045, "learning_rate": 1.685353947510189e-05, "loss": 1.3661, "step": 10700 }, { "epoch": 0.2821177268581476, "grad_norm": 5.3168816566467285, "learning_rate": 1.6847324181045503e-05, "loss": 1.4091, "step": 10710 }, { "epoch": 0.2823811421026463, "grad_norm": 3.060407876968384, "learning_rate": 1.6841103902940464e-05, "loss": 1.2937, "step": 10720 }, { "epoch": 0.28264455734714505, "grad_norm": 11.058568000793457, "learning_rate": 1.6834878645314413e-05, "loss": 1.2962, "step": 10730 }, { "epoch": 0.2829079725916438, "grad_norm": 5.227705001831055, "learning_rate": 1.68286484126986e-05, "loss": 1.4272, "step": 10740 }, { "epoch": 0.28317138783614254, "grad_norm": 12.315330505371094, "learning_rate": 1.6822413209627907e-05, "loss": 1.5406, "step": 10750 }, { "epoch": 0.2834348030806413, "grad_norm": 8.483112335205078, "learning_rate": 1.6816173040640827e-05, "loss": 1.4798, "step": 10760 }, { "epoch": 0.28369821832514003, "grad_norm": 5.636556625366211, "learning_rate": 1.6809927910279475e-05, "loss": 1.4791, "step": 10770 }, { "epoch": 0.28396163356963877, "grad_norm": 16.112831115722656, "learning_rate": 1.680367782308957e-05, "loss": 1.4543, "step": 10780 }, { "epoch": 0.2842250488141375, "grad_norm": 21.118633270263672, "learning_rate": 1.6797422783620442e-05, "loss": 1.3782, "step": 10790 }, { "epoch": 0.28448846405863626, "grad_norm": 16.596115112304688, "learning_rate": 1.6791162796425028e-05, "loss": 1.4306, "step": 10800 }, { "epoch": 0.28475187930313495, "grad_norm": 14.366751670837402, "learning_rate": 1.6784897866059863e-05, "loss": 1.4283, "step": 10810 }, { "epoch": 0.2850152945476337, "grad_norm": 12.539999008178711, "learning_rate": 1.6778627997085084e-05, "loss": 1.3525, "step": 10820 }, { "epoch": 0.28527870979213243, "grad_norm": 19.15412712097168, "learning_rate": 1.6772353194064413e-05, "loss": 1.4816, "step": 10830 }, { "epoch": 0.2855421250366312, "grad_norm": 22.995914459228516, "learning_rate": 1.676607346156518e-05, "loss": 1.3326, "step": 10840 }, { "epoch": 0.2858055402811299, "grad_norm": 8.671506881713867, "learning_rate": 1.6759788804158286e-05, "loss": 1.374, "step": 10850 }, { "epoch": 0.28606895552562867, "grad_norm": 9.748671531677246, "learning_rate": 1.6753499226418228e-05, "loss": 1.4263, "step": 10860 }, { "epoch": 0.2863323707701274, "grad_norm": 3.421365976333618, "learning_rate": 1.674720473292308e-05, "loss": 1.4203, "step": 10870 }, { "epoch": 0.28659578601462615, "grad_norm": 23.0704288482666, "learning_rate": 1.674090532825449e-05, "loss": 1.4973, "step": 10880 }, { "epoch": 0.2868592012591249, "grad_norm": 11.372081756591797, "learning_rate": 1.6734601016997695e-05, "loss": 1.4191, "step": 10890 }, { "epoch": 0.2871226165036236, "grad_norm": 2.87218976020813, "learning_rate": 1.6728291803741485e-05, "loss": 1.4, "step": 10900 }, { "epoch": 0.28738603174812233, "grad_norm": 19.180843353271484, "learning_rate": 1.672197769307823e-05, "loss": 1.4073, "step": 10910 }, { "epoch": 0.28764944699262107, "grad_norm": 2.8647992610931396, "learning_rate": 1.6715658689603866e-05, "loss": 1.4504, "step": 10920 }, { "epoch": 0.2879128622371198, "grad_norm": 13.637640953063965, "learning_rate": 1.670933479791788e-05, "loss": 1.3366, "step": 10930 }, { "epoch": 0.28817627748161856, "grad_norm": 17.84374237060547, "learning_rate": 1.6703006022623327e-05, "loss": 1.3657, "step": 10940 }, { "epoch": 0.2884396927261173, "grad_norm": 9.04928970336914, "learning_rate": 1.6696672368326814e-05, "loss": 1.2915, "step": 10950 }, { "epoch": 0.28870310797061605, "grad_norm": 16.57237434387207, "learning_rate": 1.6690333839638494e-05, "loss": 1.2984, "step": 10960 }, { "epoch": 0.2889665232151148, "grad_norm": 5.867823123931885, "learning_rate": 1.6683990441172073e-05, "loss": 1.3516, "step": 10970 }, { "epoch": 0.2892299384596135, "grad_norm": 5.3153252601623535, "learning_rate": 1.6677642177544803e-05, "loss": 1.4213, "step": 10980 }, { "epoch": 0.2894933537041122, "grad_norm": 13.790233612060547, "learning_rate": 1.6671289053377476e-05, "loss": 1.5141, "step": 10990 }, { "epoch": 0.28975676894861097, "grad_norm": 4.852967262268066, "learning_rate": 1.6664931073294413e-05, "loss": 1.5027, "step": 11000 }, { "epoch": 0.2900201841931097, "grad_norm": 14.57627010345459, "learning_rate": 1.6658568241923488e-05, "loss": 1.4078, "step": 11010 }, { "epoch": 0.29028359943760845, "grad_norm": 12.767474174499512, "learning_rate": 1.6652200563896094e-05, "loss": 1.4368, "step": 11020 }, { "epoch": 0.2905470146821072, "grad_norm": 10.942769050598145, "learning_rate": 1.6645828043847147e-05, "loss": 1.3787, "step": 11030 }, { "epoch": 0.29081042992660594, "grad_norm": 7.967550754547119, "learning_rate": 1.6639450686415103e-05, "loss": 1.5008, "step": 11040 }, { "epoch": 0.2910738451711047, "grad_norm": 22.981107711791992, "learning_rate": 1.663306849624192e-05, "loss": 1.3686, "step": 11050 }, { "epoch": 0.2913372604156034, "grad_norm": 6.2156982421875, "learning_rate": 1.6626681477973092e-05, "loss": 1.4371, "step": 11060 }, { "epoch": 0.2916006756601021, "grad_norm": 3.8957228660583496, "learning_rate": 1.662028963625761e-05, "loss": 1.4377, "step": 11070 }, { "epoch": 0.29186409090460086, "grad_norm": 12.599555015563965, "learning_rate": 1.6613892975747998e-05, "loss": 1.4782, "step": 11080 }, { "epoch": 0.2921275061490996, "grad_norm": 12.195301055908203, "learning_rate": 1.6607491501100262e-05, "loss": 1.3241, "step": 11090 }, { "epoch": 0.29239092139359835, "grad_norm": 6.27243709564209, "learning_rate": 1.6601085216973932e-05, "loss": 1.3788, "step": 11100 }, { "epoch": 0.2926543366380971, "grad_norm": 6.370360374450684, "learning_rate": 1.6594674128032026e-05, "loss": 1.4074, "step": 11110 }, { "epoch": 0.29291775188259583, "grad_norm": 8.430095672607422, "learning_rate": 1.6588258238941066e-05, "loss": 1.4595, "step": 11120 }, { "epoch": 0.2931811671270946, "grad_norm": 7.276546478271484, "learning_rate": 1.6581837554371072e-05, "loss": 1.3929, "step": 11130 }, { "epoch": 0.2934445823715933, "grad_norm": 5.935778617858887, "learning_rate": 1.657541207899554e-05, "loss": 1.4663, "step": 11140 }, { "epoch": 0.29370799761609206, "grad_norm": 11.938579559326172, "learning_rate": 1.656898181749147e-05, "loss": 1.3986, "step": 11150 }, { "epoch": 0.29397141286059075, "grad_norm": 7.343424320220947, "learning_rate": 1.6562546774539336e-05, "loss": 1.391, "step": 11160 }, { "epoch": 0.2942348281050895, "grad_norm": 17.60485076904297, "learning_rate": 1.655610695482309e-05, "loss": 1.3937, "step": 11170 }, { "epoch": 0.29449824334958824, "grad_norm": 14.37977123260498, "learning_rate": 1.6549662363030167e-05, "loss": 1.4869, "step": 11180 }, { "epoch": 0.294761658594087, "grad_norm": 12.51855754852295, "learning_rate": 1.654321300385148e-05, "loss": 1.4631, "step": 11190 }, { "epoch": 0.2950250738385857, "grad_norm": 5.2395429611206055, "learning_rate": 1.6536758881981397e-05, "loss": 1.4011, "step": 11200 }, { "epoch": 0.29528848908308447, "grad_norm": 4.581459045410156, "learning_rate": 1.653030000211777e-05, "loss": 1.444, "step": 11210 }, { "epoch": 0.2955519043275832, "grad_norm": 7.901134490966797, "learning_rate": 1.6523836368961895e-05, "loss": 1.4952, "step": 11220 }, { "epoch": 0.29581531957208196, "grad_norm": 33.4068489074707, "learning_rate": 1.6517367987218552e-05, "loss": 1.4896, "step": 11230 }, { "epoch": 0.29607873481658065, "grad_norm": 14.283897399902344, "learning_rate": 1.6510894861595953e-05, "loss": 1.5399, "step": 11240 }, { "epoch": 0.2963421500610794, "grad_norm": 19.251859664916992, "learning_rate": 1.6504416996805775e-05, "loss": 1.3279, "step": 11250 }, { "epoch": 0.29660556530557813, "grad_norm": 16.487606048583984, "learning_rate": 1.649793439756315e-05, "loss": 1.4273, "step": 11260 }, { "epoch": 0.2968689805500769, "grad_norm": 8.102100372314453, "learning_rate": 1.6491447068586644e-05, "loss": 1.4234, "step": 11270 }, { "epoch": 0.2971323957945756, "grad_norm": 8.766030311584473, "learning_rate": 1.6484955014598273e-05, "loss": 1.3522, "step": 11280 }, { "epoch": 0.29739581103907436, "grad_norm": 3.9384922981262207, "learning_rate": 1.6478458240323495e-05, "loss": 1.4133, "step": 11290 }, { "epoch": 0.2976592262835731, "grad_norm": 9.32307243347168, "learning_rate": 1.6471956750491192e-05, "loss": 1.2923, "step": 11300 }, { "epoch": 0.29792264152807185, "grad_norm": 12.604999542236328, "learning_rate": 1.6465450549833687e-05, "loss": 1.527, "step": 11310 }, { "epoch": 0.2981860567725706, "grad_norm": 15.302648544311523, "learning_rate": 1.645893964308674e-05, "loss": 1.4863, "step": 11320 }, { "epoch": 0.2984494720170693, "grad_norm": 24.4898681640625, "learning_rate": 1.6452424034989512e-05, "loss": 1.3936, "step": 11330 }, { "epoch": 0.298712887261568, "grad_norm": 7.22315788269043, "learning_rate": 1.6445903730284617e-05, "loss": 1.3531, "step": 11340 }, { "epoch": 0.29897630250606677, "grad_norm": 15.466109275817871, "learning_rate": 1.6439378733718062e-05, "loss": 1.4779, "step": 11350 }, { "epoch": 0.2992397177505655, "grad_norm": 2.7611029148101807, "learning_rate": 1.6432849050039278e-05, "loss": 1.4559, "step": 11360 }, { "epoch": 0.29950313299506426, "grad_norm": 9.261701583862305, "learning_rate": 1.6426314684001118e-05, "loss": 1.3732, "step": 11370 }, { "epoch": 0.299766548239563, "grad_norm": 7.383176803588867, "learning_rate": 1.6419775640359822e-05, "loss": 1.3382, "step": 11380 }, { "epoch": 0.30002996348406175, "grad_norm": 7.435217380523682, "learning_rate": 1.6413231923875053e-05, "loss": 1.3494, "step": 11390 }, { "epoch": 0.3002933787285605, "grad_norm": 8.045174598693848, "learning_rate": 1.6406683539309872e-05, "loss": 1.399, "step": 11400 }, { "epoch": 0.30055679397305923, "grad_norm": 6.5585150718688965, "learning_rate": 1.6400130491430726e-05, "loss": 1.3808, "step": 11410 }, { "epoch": 0.3008202092175579, "grad_norm": 10.903836250305176, "learning_rate": 1.6393572785007474e-05, "loss": 1.4771, "step": 11420 }, { "epoch": 0.30108362446205666, "grad_norm": 5.828615188598633, "learning_rate": 1.6387010424813354e-05, "loss": 1.3633, "step": 11430 }, { "epoch": 0.3013470397065554, "grad_norm": 19.3684024810791, "learning_rate": 1.638044341562499e-05, "loss": 1.4437, "step": 11440 }, { "epoch": 0.30161045495105415, "grad_norm": 13.786738395690918, "learning_rate": 1.6373871762222398e-05, "loss": 1.3755, "step": 11450 }, { "epoch": 0.3018738701955529, "grad_norm": 6.257304668426514, "learning_rate": 1.6367295469388967e-05, "loss": 1.4433, "step": 11460 }, { "epoch": 0.30213728544005164, "grad_norm": 4.551675796508789, "learning_rate": 1.636071454191148e-05, "loss": 1.5617, "step": 11470 }, { "epoch": 0.3024007006845504, "grad_norm": 12.681303977966309, "learning_rate": 1.635412898458006e-05, "loss": 1.2626, "step": 11480 }, { "epoch": 0.3026641159290491, "grad_norm": 2.4506781101226807, "learning_rate": 1.634753880218823e-05, "loss": 1.4016, "step": 11490 }, { "epoch": 0.3029275311735478, "grad_norm": 7.696333885192871, "learning_rate": 1.6340943999532873e-05, "loss": 1.4061, "step": 11500 }, { "epoch": 0.30319094641804656, "grad_norm": 16.586326599121094, "learning_rate": 1.6334344581414224e-05, "loss": 1.4813, "step": 11510 }, { "epoch": 0.3034543616625453, "grad_norm": 16.241857528686523, "learning_rate": 1.6327740552635895e-05, "loss": 1.4033, "step": 11520 }, { "epoch": 0.30371777690704405, "grad_norm": 8.393757820129395, "learning_rate": 1.6321131918004835e-05, "loss": 1.3201, "step": 11530 }, { "epoch": 0.3039811921515428, "grad_norm": 4.901735305786133, "learning_rate": 1.631451868233136e-05, "loss": 1.3524, "step": 11540 }, { "epoch": 0.30424460739604153, "grad_norm": 9.770957946777344, "learning_rate": 1.6307900850429126e-05, "loss": 1.4204, "step": 11550 }, { "epoch": 0.3045080226405403, "grad_norm": 6.626745223999023, "learning_rate": 1.630127842711514e-05, "loss": 1.4936, "step": 11560 }, { "epoch": 0.304771437885039, "grad_norm": 9.699769973754883, "learning_rate": 1.629465141720975e-05, "loss": 1.4657, "step": 11570 }, { "epoch": 0.30503485312953776, "grad_norm": 11.607088088989258, "learning_rate": 1.628801982553664e-05, "loss": 1.4673, "step": 11580 }, { "epoch": 0.30529826837403645, "grad_norm": 22.222490310668945, "learning_rate": 1.6281383656922833e-05, "loss": 1.4012, "step": 11590 }, { "epoch": 0.3055616836185352, "grad_norm": 3.7001383304595947, "learning_rate": 1.6274742916198673e-05, "loss": 1.3686, "step": 11600 }, { "epoch": 0.30582509886303394, "grad_norm": 36.090576171875, "learning_rate": 1.6268097608197855e-05, "loss": 1.3965, "step": 11610 }, { "epoch": 0.3060885141075327, "grad_norm": 5.05103063583374, "learning_rate": 1.626144773775737e-05, "loss": 1.5019, "step": 11620 }, { "epoch": 0.3063519293520314, "grad_norm": 4.36926794052124, "learning_rate": 1.6254793309717553e-05, "loss": 1.4254, "step": 11630 }, { "epoch": 0.30661534459653017, "grad_norm": 10.330196380615234, "learning_rate": 1.624813432892204e-05, "loss": 1.4926, "step": 11640 }, { "epoch": 0.3068787598410289, "grad_norm": 10.573012351989746, "learning_rate": 1.624147080021779e-05, "loss": 1.3511, "step": 11650 }, { "epoch": 0.30714217508552766, "grad_norm": 14.491007804870605, "learning_rate": 1.6234802728455075e-05, "loss": 1.42, "step": 11660 }, { "epoch": 0.3074055903300264, "grad_norm": 10.059892654418945, "learning_rate": 1.6228130118487458e-05, "loss": 1.4034, "step": 11670 }, { "epoch": 0.3076690055745251, "grad_norm": 17.897216796875, "learning_rate": 1.6221452975171826e-05, "loss": 1.359, "step": 11680 }, { "epoch": 0.30793242081902383, "grad_norm": 15.310016632080078, "learning_rate": 1.6214771303368357e-05, "loss": 1.3946, "step": 11690 }, { "epoch": 0.3081958360635226, "grad_norm": 10.818742752075195, "learning_rate": 1.6208085107940516e-05, "loss": 1.4407, "step": 11700 }, { "epoch": 0.3084592513080213, "grad_norm": 6.188978672027588, "learning_rate": 1.620139439375507e-05, "loss": 1.2963, "step": 11710 }, { "epoch": 0.30872266655252006, "grad_norm": 7.344741344451904, "learning_rate": 1.6194699165682077e-05, "loss": 1.3941, "step": 11720 }, { "epoch": 0.3089860817970188, "grad_norm": 10.95267105102539, "learning_rate": 1.6187999428594882e-05, "loss": 1.4899, "step": 11730 }, { "epoch": 0.30924949704151755, "grad_norm": 14.109932899475098, "learning_rate": 1.6181295187370095e-05, "loss": 1.3278, "step": 11740 }, { "epoch": 0.3095129122860163, "grad_norm": 7.154019832611084, "learning_rate": 1.617458644688763e-05, "loss": 1.4507, "step": 11750 }, { "epoch": 0.309776327530515, "grad_norm": 11.458073616027832, "learning_rate": 1.616787321203066e-05, "loss": 1.4337, "step": 11760 }, { "epoch": 0.3100397427750137, "grad_norm": 4.665543079376221, "learning_rate": 1.6161155487685624e-05, "loss": 1.3787, "step": 11770 }, { "epoch": 0.31030315801951247, "grad_norm": 16.608346939086914, "learning_rate": 1.615443327874225e-05, "loss": 1.3792, "step": 11780 }, { "epoch": 0.3105665732640112, "grad_norm": 5.418328762054443, "learning_rate": 1.6147706590093506e-05, "loss": 1.2397, "step": 11790 }, { "epoch": 0.31082998850850996, "grad_norm": 12.209650993347168, "learning_rate": 1.6140975426635637e-05, "loss": 1.3329, "step": 11800 }, { "epoch": 0.3110934037530087, "grad_norm": 45.66841506958008, "learning_rate": 1.613423979326814e-05, "loss": 1.454, "step": 11810 }, { "epoch": 0.31135681899750745, "grad_norm": 5.743036270141602, "learning_rate": 1.612749969489377e-05, "loss": 1.2572, "step": 11820 }, { "epoch": 0.3116202342420062, "grad_norm": 4.474371433258057, "learning_rate": 1.612075513641852e-05, "loss": 1.4759, "step": 11830 }, { "epoch": 0.31188364948650493, "grad_norm": 11.682698249816895, "learning_rate": 1.611400612275164e-05, "loss": 1.3493, "step": 11840 }, { "epoch": 0.3121470647310036, "grad_norm": 5.382815361022949, "learning_rate": 1.610725265880562e-05, "loss": 1.4582, "step": 11850 }, { "epoch": 0.31241047997550236, "grad_norm": 20.161046981811523, "learning_rate": 1.6100494749496195e-05, "loss": 1.3452, "step": 11860 }, { "epoch": 0.3126738952200011, "grad_norm": 15.565869331359863, "learning_rate": 1.6093732399742324e-05, "loss": 1.303, "step": 11870 }, { "epoch": 0.31293731046449985, "grad_norm": 14.290550231933594, "learning_rate": 1.6086965614466206e-05, "loss": 1.4329, "step": 11880 }, { "epoch": 0.3132007257089986, "grad_norm": 10.574429512023926, "learning_rate": 1.6080194398593267e-05, "loss": 1.3956, "step": 11890 }, { "epoch": 0.31346414095349734, "grad_norm": 11.35422134399414, "learning_rate": 1.6073418757052156e-05, "loss": 1.3009, "step": 11900 }, { "epoch": 0.3137275561979961, "grad_norm": 7.912886619567871, "learning_rate": 1.6066638694774747e-05, "loss": 1.3234, "step": 11910 }, { "epoch": 0.3139909714424948, "grad_norm": 6.515002250671387, "learning_rate": 1.605985421669613e-05, "loss": 1.4081, "step": 11920 }, { "epoch": 0.31425438668699357, "grad_norm": 15.646952629089355, "learning_rate": 1.605306532775461e-05, "loss": 1.38, "step": 11930 }, { "epoch": 0.31451780193149226, "grad_norm": 15.696507453918457, "learning_rate": 1.6046272032891698e-05, "loss": 1.3643, "step": 11940 }, { "epoch": 0.314781217175991, "grad_norm": 13.89185619354248, "learning_rate": 1.6039474337052116e-05, "loss": 1.4689, "step": 11950 }, { "epoch": 0.31504463242048975, "grad_norm": 6.697390079498291, "learning_rate": 1.6032672245183793e-05, "loss": 1.2771, "step": 11960 }, { "epoch": 0.3153080476649885, "grad_norm": 7.575247287750244, "learning_rate": 1.6025865762237854e-05, "loss": 1.4956, "step": 11970 }, { "epoch": 0.31557146290948723, "grad_norm": 15.326065063476562, "learning_rate": 1.6019054893168614e-05, "loss": 1.4203, "step": 11980 }, { "epoch": 0.315834878153986, "grad_norm": 16.28388214111328, "learning_rate": 1.601223964293359e-05, "loss": 1.4028, "step": 11990 }, { "epoch": 0.3160982933984847, "grad_norm": 21.975643157958984, "learning_rate": 1.6005420016493485e-05, "loss": 1.6952, "step": 12000 }, { "epoch": 0.31636170864298346, "grad_norm": 10.333736419677734, "learning_rate": 1.599859601881219e-05, "loss": 1.4561, "step": 12010 }, { "epoch": 0.31662512388748215, "grad_norm": 5.278055191040039, "learning_rate": 1.5991767654856772e-05, "loss": 1.4471, "step": 12020 }, { "epoch": 0.3168885391319809, "grad_norm": 33.18233108520508, "learning_rate": 1.5984934929597473e-05, "loss": 1.2913, "step": 12030 }, { "epoch": 0.31715195437647964, "grad_norm": 52.71499252319336, "learning_rate": 1.5978097848007723e-05, "loss": 1.4315, "step": 12040 }, { "epoch": 0.3174153696209784, "grad_norm": 10.75975227355957, "learning_rate": 1.5971256415064115e-05, "loss": 1.4687, "step": 12050 }, { "epoch": 0.3176787848654771, "grad_norm": 8.186162948608398, "learning_rate": 1.5964410635746406e-05, "loss": 1.3882, "step": 12060 }, { "epoch": 0.31794220010997587, "grad_norm": 63.93559265136719, "learning_rate": 1.5957560515037518e-05, "loss": 1.4094, "step": 12070 }, { "epoch": 0.3182056153544746, "grad_norm": 10.444463729858398, "learning_rate": 1.5950706057923544e-05, "loss": 1.5581, "step": 12080 }, { "epoch": 0.31846903059897336, "grad_norm": 13.02112102508545, "learning_rate": 1.5943847269393716e-05, "loss": 1.4806, "step": 12090 }, { "epoch": 0.3187324458434721, "grad_norm": 8.401366233825684, "learning_rate": 1.5936984154440435e-05, "loss": 1.3799, "step": 12100 }, { "epoch": 0.3189958610879708, "grad_norm": 6.218594074249268, "learning_rate": 1.5930116718059237e-05, "loss": 1.3766, "step": 12110 }, { "epoch": 0.31925927633246953, "grad_norm": 10.2295560836792, "learning_rate": 1.592324496524881e-05, "loss": 1.3985, "step": 12120 }, { "epoch": 0.3195226915769683, "grad_norm": 8.347479820251465, "learning_rate": 1.5916368901010987e-05, "loss": 1.4876, "step": 12130 }, { "epoch": 0.319786106821467, "grad_norm": 10.813390731811523, "learning_rate": 1.5909488530350735e-05, "loss": 1.3582, "step": 12140 }, { "epoch": 0.32004952206596576, "grad_norm": 37.29164505004883, "learning_rate": 1.5902603858276155e-05, "loss": 1.4061, "step": 12150 }, { "epoch": 0.3203129373104645, "grad_norm": 684.0595092773438, "learning_rate": 1.5895714889798485e-05, "loss": 1.442, "step": 12160 }, { "epoch": 0.32057635255496325, "grad_norm": 399.1373291015625, "learning_rate": 1.5888821629932084e-05, "loss": 1.488, "step": 12170 }, { "epoch": 0.320839767799462, "grad_norm": 25.75657844543457, "learning_rate": 1.5881924083694435e-05, "loss": 1.515, "step": 12180 }, { "epoch": 0.32110318304396074, "grad_norm": 4.991872310638428, "learning_rate": 1.587502225610614e-05, "loss": 1.4289, "step": 12190 }, { "epoch": 0.3213665982884594, "grad_norm": 7.99547004699707, "learning_rate": 1.5868116152190925e-05, "loss": 1.4483, "step": 12200 }, { "epoch": 0.32163001353295817, "grad_norm": 11.554422378540039, "learning_rate": 1.586120577697562e-05, "loss": 1.3018, "step": 12210 }, { "epoch": 0.3218934287774569, "grad_norm": 7.3297014236450195, "learning_rate": 1.5854291135490167e-05, "loss": 1.5076, "step": 12220 }, { "epoch": 0.32215684402195566, "grad_norm": 10.765825271606445, "learning_rate": 1.584737223276762e-05, "loss": 1.2163, "step": 12230 }, { "epoch": 0.3224202592664544, "grad_norm": 49.778411865234375, "learning_rate": 1.5840449073844118e-05, "loss": 1.4524, "step": 12240 }, { "epoch": 0.32268367451095314, "grad_norm": 8.963550567626953, "learning_rate": 1.5833521663758915e-05, "loss": 1.5056, "step": 12250 }, { "epoch": 0.3229470897554519, "grad_norm": 6.211223125457764, "learning_rate": 1.5826590007554356e-05, "loss": 1.4518, "step": 12260 }, { "epoch": 0.32321050499995063, "grad_norm": 8.3572359085083, "learning_rate": 1.581965411027586e-05, "loss": 1.2503, "step": 12270 }, { "epoch": 0.3234739202444493, "grad_norm": 8.056180000305176, "learning_rate": 1.5812713976971966e-05, "loss": 1.3305, "step": 12280 }, { "epoch": 0.32373733548894806, "grad_norm": 3.9289252758026123, "learning_rate": 1.580576961269426e-05, "loss": 1.4634, "step": 12290 }, { "epoch": 0.3240007507334468, "grad_norm": 11.075051307678223, "learning_rate": 1.5798821022497435e-05, "loss": 1.4865, "step": 12300 }, { "epoch": 0.32426416597794555, "grad_norm": 18.046520233154297, "learning_rate": 1.5791868211439244e-05, "loss": 1.4686, "step": 12310 }, { "epoch": 0.3245275812224443, "grad_norm": 25.2907772064209, "learning_rate": 1.578491118458052e-05, "loss": 1.458, "step": 12320 }, { "epoch": 0.32479099646694304, "grad_norm": 13.678783416748047, "learning_rate": 1.5777949946985157e-05, "loss": 1.3756, "step": 12330 }, { "epoch": 0.3250544117114418, "grad_norm": 11.862297058105469, "learning_rate": 1.577098450372013e-05, "loss": 1.3323, "step": 12340 }, { "epoch": 0.3253178269559405, "grad_norm": 4.0484747886657715, "learning_rate": 1.5764014859855456e-05, "loss": 1.5687, "step": 12350 }, { "epoch": 0.32558124220043927, "grad_norm": 3.4585046768188477, "learning_rate": 1.5757041020464215e-05, "loss": 1.3572, "step": 12360 }, { "epoch": 0.32584465744493796, "grad_norm": 12.318306922912598, "learning_rate": 1.5750062990622558e-05, "loss": 1.4131, "step": 12370 }, { "epoch": 0.3261080726894367, "grad_norm": 10.13508129119873, "learning_rate": 1.5743080775409654e-05, "loss": 1.4599, "step": 12380 }, { "epoch": 0.32637148793393544, "grad_norm": 7.670694828033447, "learning_rate": 1.5736094379907755e-05, "loss": 1.3552, "step": 12390 }, { "epoch": 0.3266349031784342, "grad_norm": 5.777686595916748, "learning_rate": 1.572910380920212e-05, "loss": 1.2865, "step": 12400 }, { "epoch": 0.32689831842293293, "grad_norm": 21.895729064941406, "learning_rate": 1.5722109068381076e-05, "loss": 1.5236, "step": 12410 }, { "epoch": 0.3271617336674317, "grad_norm": 5.889650821685791, "learning_rate": 1.5715110162535972e-05, "loss": 1.4118, "step": 12420 }, { "epoch": 0.3274251489119304, "grad_norm": 8.279088973999023, "learning_rate": 1.5708107096761184e-05, "loss": 1.3995, "step": 12430 }, { "epoch": 0.32768856415642916, "grad_norm": 13.801074028015137, "learning_rate": 1.570109987615413e-05, "loss": 1.4041, "step": 12440 }, { "epoch": 0.3279519794009279, "grad_norm": 9.493692398071289, "learning_rate": 1.5694088505815238e-05, "loss": 1.4303, "step": 12450 }, { "epoch": 0.3282153946454266, "grad_norm": 12.441193580627441, "learning_rate": 1.5687072990847965e-05, "loss": 1.3846, "step": 12460 }, { "epoch": 0.32847880988992534, "grad_norm": 13.371628761291504, "learning_rate": 1.568005333635879e-05, "loss": 1.2764, "step": 12470 }, { "epoch": 0.3287422251344241, "grad_norm": 16.4624080657959, "learning_rate": 1.5673029547457184e-05, "loss": 1.2931, "step": 12480 }, { "epoch": 0.3290056403789228, "grad_norm": 29.842636108398438, "learning_rate": 1.5666001629255655e-05, "loss": 1.3712, "step": 12490 }, { "epoch": 0.32926905562342157, "grad_norm": 6.33497428894043, "learning_rate": 1.5658969586869694e-05, "loss": 1.4381, "step": 12500 }, { "epoch": 0.3295324708679203, "grad_norm": 12.79800033569336, "learning_rate": 1.5651933425417804e-05, "loss": 1.3508, "step": 12510 }, { "epoch": 0.32979588611241906, "grad_norm": 7.985075950622559, "learning_rate": 1.564489315002149e-05, "loss": 1.4031, "step": 12520 }, { "epoch": 0.3300593013569178, "grad_norm": 5.1234612464904785, "learning_rate": 1.5637848765805238e-05, "loss": 1.413, "step": 12530 }, { "epoch": 0.3303227166014165, "grad_norm": 5.010728359222412, "learning_rate": 1.5630800277896544e-05, "loss": 1.4298, "step": 12540 }, { "epoch": 0.33058613184591523, "grad_norm": 7.344033241271973, "learning_rate": 1.5623747691425872e-05, "loss": 1.3648, "step": 12550 }, { "epoch": 0.330849547090414, "grad_norm": 10.601235389709473, "learning_rate": 1.561669101152668e-05, "loss": 1.3211, "step": 12560 }, { "epoch": 0.3311129623349127, "grad_norm": 13.062161445617676, "learning_rate": 1.5609630243335404e-05, "loss": 1.3769, "step": 12570 }, { "epoch": 0.33137637757941146, "grad_norm": 6.036024570465088, "learning_rate": 1.560256539199146e-05, "loss": 1.4366, "step": 12580 }, { "epoch": 0.3316397928239102, "grad_norm": 14.684640884399414, "learning_rate": 1.5595496462637215e-05, "loss": 1.3641, "step": 12590 }, { "epoch": 0.33190320806840895, "grad_norm": 7.113190650939941, "learning_rate": 1.5588423460418038e-05, "loss": 1.3579, "step": 12600 }, { "epoch": 0.3321666233129077, "grad_norm": 8.90922737121582, "learning_rate": 1.5581346390482228e-05, "loss": 1.5132, "step": 12610 }, { "epoch": 0.33243003855740644, "grad_norm": 8.317419052124023, "learning_rate": 1.5574265257981077e-05, "loss": 1.284, "step": 12620 }, { "epoch": 0.3326934538019051, "grad_norm": 5.407451629638672, "learning_rate": 1.5567180068068812e-05, "loss": 1.4225, "step": 12630 }, { "epoch": 0.33295686904640387, "grad_norm": 12.0491361618042, "learning_rate": 1.556009082590262e-05, "loss": 1.4021, "step": 12640 }, { "epoch": 0.3332202842909026, "grad_norm": 14.125913619995117, "learning_rate": 1.5552997536642637e-05, "loss": 1.4251, "step": 12650 }, { "epoch": 0.33348369953540136, "grad_norm": 6.6380510330200195, "learning_rate": 1.5545900205451947e-05, "loss": 1.3716, "step": 12660 }, { "epoch": 0.3337471147799001, "grad_norm": 18.087665557861328, "learning_rate": 1.5538798837496568e-05, "loss": 1.3759, "step": 12670 }, { "epoch": 0.33401053002439884, "grad_norm": 4.536417007446289, "learning_rate": 1.5531693437945476e-05, "loss": 1.3794, "step": 12680 }, { "epoch": 0.3342739452688976, "grad_norm": 10.937272071838379, "learning_rate": 1.5524584011970558e-05, "loss": 1.3874, "step": 12690 }, { "epoch": 0.33453736051339633, "grad_norm": 8.105043411254883, "learning_rate": 1.5517470564746644e-05, "loss": 1.284, "step": 12700 }, { "epoch": 0.3348007757578951, "grad_norm": 8.407843589782715, "learning_rate": 1.551035310145149e-05, "loss": 1.3609, "step": 12710 }, { "epoch": 0.33506419100239376, "grad_norm": 18.736358642578125, "learning_rate": 1.5503231627265772e-05, "loss": 1.4181, "step": 12720 }, { "epoch": 0.3353276062468925, "grad_norm": 4.751344203948975, "learning_rate": 1.5496106147373093e-05, "loss": 1.3641, "step": 12730 }, { "epoch": 0.33559102149139125, "grad_norm": 6.319554328918457, "learning_rate": 1.548897666695996e-05, "loss": 1.403, "step": 12740 }, { "epoch": 0.33585443673589, "grad_norm": 14.073752403259277, "learning_rate": 1.5481843191215803e-05, "loss": 1.4101, "step": 12750 }, { "epoch": 0.33611785198038874, "grad_norm": 5.474279403686523, "learning_rate": 1.5474705725332954e-05, "loss": 1.2445, "step": 12760 }, { "epoch": 0.3363812672248875, "grad_norm": 10.729292869567871, "learning_rate": 1.5467564274506647e-05, "loss": 1.4834, "step": 12770 }, { "epoch": 0.3366446824693862, "grad_norm": 15.940821647644043, "learning_rate": 1.5460418843935027e-05, "loss": 1.3707, "step": 12780 }, { "epoch": 0.33690809771388497, "grad_norm": 9.723480224609375, "learning_rate": 1.5453269438819125e-05, "loss": 1.3665, "step": 12790 }, { "epoch": 0.33717151295838366, "grad_norm": 9.150084495544434, "learning_rate": 1.5446116064362867e-05, "loss": 1.5055, "step": 12800 }, { "epoch": 0.3374349282028824, "grad_norm": 13.373900413513184, "learning_rate": 1.5438958725773077e-05, "loss": 1.3847, "step": 12810 }, { "epoch": 0.33769834344738114, "grad_norm": 11.6533784866333, "learning_rate": 1.543179742825945e-05, "loss": 1.3438, "step": 12820 }, { "epoch": 0.3379617586918799, "grad_norm": 2.24631404876709, "learning_rate": 1.5424632177034577e-05, "loss": 1.3879, "step": 12830 }, { "epoch": 0.33822517393637863, "grad_norm": 4.449690341949463, "learning_rate": 1.5417462977313918e-05, "loss": 1.3269, "step": 12840 }, { "epoch": 0.3384885891808774, "grad_norm": 3.368565559387207, "learning_rate": 1.5410289834315812e-05, "loss": 1.306, "step": 12850 }, { "epoch": 0.3387520044253761, "grad_norm": 5.641608715057373, "learning_rate": 1.5403112753261458e-05, "loss": 1.4553, "step": 12860 }, { "epoch": 0.33901541966987486, "grad_norm": 11.098250389099121, "learning_rate": 1.5395931739374935e-05, "loss": 1.3824, "step": 12870 }, { "epoch": 0.3392788349143736, "grad_norm": 3.5989723205566406, "learning_rate": 1.5388746797883172e-05, "loss": 1.4015, "step": 12880 }, { "epoch": 0.3395422501588723, "grad_norm": 2.5168631076812744, "learning_rate": 1.5381557934015975e-05, "loss": 1.3922, "step": 12890 }, { "epoch": 0.33980566540337104, "grad_norm": 10.570174217224121, "learning_rate": 1.5374365153005982e-05, "loss": 1.3922, "step": 12900 }, { "epoch": 0.3400690806478698, "grad_norm": 11.620294570922852, "learning_rate": 1.5367168460088703e-05, "loss": 1.4506, "step": 12910 }, { "epoch": 0.3403324958923685, "grad_norm": 17.21299171447754, "learning_rate": 1.5359967860502484e-05, "loss": 1.468, "step": 12920 }, { "epoch": 0.34059591113686727, "grad_norm": 6.778715133666992, "learning_rate": 1.5352763359488506e-05, "loss": 1.4487, "step": 12930 }, { "epoch": 0.340859326381366, "grad_norm": 15.515448570251465, "learning_rate": 1.5345554962290816e-05, "loss": 1.4136, "step": 12940 }, { "epoch": 0.34112274162586476, "grad_norm": 25.95926284790039, "learning_rate": 1.5338342674156274e-05, "loss": 1.3851, "step": 12950 }, { "epoch": 0.3413861568703635, "grad_norm": 2.595224380493164, "learning_rate": 1.5331126500334573e-05, "loss": 1.3657, "step": 12960 }, { "epoch": 0.34164957211486224, "grad_norm": 5.58395528793335, "learning_rate": 1.5323906446078254e-05, "loss": 1.4002, "step": 12970 }, { "epoch": 0.34191298735936093, "grad_norm": 13.379227638244629, "learning_rate": 1.5316682516642663e-05, "loss": 1.3373, "step": 12980 }, { "epoch": 0.3421764026038597, "grad_norm": 12.631510734558105, "learning_rate": 1.5309454717285972e-05, "loss": 1.3197, "step": 12990 }, { "epoch": 0.3424398178483584, "grad_norm": 5.114025592803955, "learning_rate": 1.530222305326917e-05, "loss": 1.3645, "step": 13000 }, { "epoch": 0.34270323309285716, "grad_norm": 10.444599151611328, "learning_rate": 1.5294987529856066e-05, "loss": 1.4057, "step": 13010 }, { "epoch": 0.3429666483373559, "grad_norm": 10.352643013000488, "learning_rate": 1.528774815231327e-05, "loss": 1.4123, "step": 13020 }, { "epoch": 0.34323006358185465, "grad_norm": 11.244441032409668, "learning_rate": 1.5280504925910195e-05, "loss": 1.6236, "step": 13030 }, { "epoch": 0.3434934788263534, "grad_norm": 13.948497772216797, "learning_rate": 1.5273257855919062e-05, "loss": 1.3617, "step": 13040 }, { "epoch": 0.34375689407085214, "grad_norm": 13.996700286865234, "learning_rate": 1.526600694761489e-05, "loss": 1.3615, "step": 13050 }, { "epoch": 0.3440203093153508, "grad_norm": 12.247322082519531, "learning_rate": 1.5258752206275486e-05, "loss": 1.4132, "step": 13060 }, { "epoch": 0.34428372455984957, "grad_norm": 14.586034774780273, "learning_rate": 1.5251493637181453e-05, "loss": 1.4402, "step": 13070 }, { "epoch": 0.3445471398043483, "grad_norm": 10.141186714172363, "learning_rate": 1.5244231245616182e-05, "loss": 1.4002, "step": 13080 }, { "epoch": 0.34481055504884706, "grad_norm": 6.6555070877075195, "learning_rate": 1.5236965036865834e-05, "loss": 1.3957, "step": 13090 }, { "epoch": 0.3450739702933458, "grad_norm": 2.5723233222961426, "learning_rate": 1.5229695016219358e-05, "loss": 1.3859, "step": 13100 }, { "epoch": 0.34533738553784454, "grad_norm": 18.0116024017334, "learning_rate": 1.522242118896848e-05, "loss": 1.377, "step": 13110 }, { "epoch": 0.3456008007823433, "grad_norm": 3.1253726482391357, "learning_rate": 1.5215143560407693e-05, "loss": 1.3487, "step": 13120 }, { "epoch": 0.34586421602684203, "grad_norm": 4.386977672576904, "learning_rate": 1.5207862135834252e-05, "loss": 1.4382, "step": 13130 }, { "epoch": 0.3461276312713408, "grad_norm": 5.462677001953125, "learning_rate": 1.520057692054818e-05, "loss": 1.2968, "step": 13140 }, { "epoch": 0.34639104651583946, "grad_norm": 12.846744537353516, "learning_rate": 1.5193287919852267e-05, "loss": 1.445, "step": 13150 }, { "epoch": 0.3466544617603382, "grad_norm": 7.607128620147705, "learning_rate": 1.5185995139052044e-05, "loss": 1.3689, "step": 13160 }, { "epoch": 0.34691787700483695, "grad_norm": 9.580033302307129, "learning_rate": 1.5178698583455795e-05, "loss": 1.2586, "step": 13170 }, { "epoch": 0.3471812922493357, "grad_norm": 14.428675651550293, "learning_rate": 1.5171398258374566e-05, "loss": 1.369, "step": 13180 }, { "epoch": 0.34744470749383444, "grad_norm": 14.435626983642578, "learning_rate": 1.5164094169122132e-05, "loss": 1.4269, "step": 13190 }, { "epoch": 0.3477081227383332, "grad_norm": 3.1616885662078857, "learning_rate": 1.5156786321015013e-05, "loss": 1.4491, "step": 13200 }, { "epoch": 0.3479715379828319, "grad_norm": 12.13348388671875, "learning_rate": 1.514947471937247e-05, "loss": 1.2706, "step": 13210 }, { "epoch": 0.34823495322733067, "grad_norm": 6.760093688964844, "learning_rate": 1.5142159369516483e-05, "loss": 1.3064, "step": 13220 }, { "epoch": 0.3484983684718294, "grad_norm": 6.532286167144775, "learning_rate": 1.5134840276771777e-05, "loss": 1.5262, "step": 13230 }, { "epoch": 0.3487617837163281, "grad_norm": 4.950694561004639, "learning_rate": 1.5127517446465785e-05, "loss": 1.2846, "step": 13240 }, { "epoch": 0.34902519896082684, "grad_norm": 3.5839781761169434, "learning_rate": 1.5120190883928671e-05, "loss": 1.3826, "step": 13250 }, { "epoch": 0.3492886142053256, "grad_norm": 4.57205057144165, "learning_rate": 1.511286059449332e-05, "loss": 1.3831, "step": 13260 }, { "epoch": 0.34955202944982433, "grad_norm": 7.147394180297852, "learning_rate": 1.5105526583495311e-05, "loss": 1.4039, "step": 13270 }, { "epoch": 0.3498154446943231, "grad_norm": 7.256117343902588, "learning_rate": 1.5098188856272948e-05, "loss": 1.2891, "step": 13280 }, { "epoch": 0.3500788599388218, "grad_norm": 13.992640495300293, "learning_rate": 1.509084741816724e-05, "loss": 1.5507, "step": 13290 }, { "epoch": 0.35034227518332056, "grad_norm": 14.541725158691406, "learning_rate": 1.5083502274521888e-05, "loss": 1.3419, "step": 13300 }, { "epoch": 0.3506056904278193, "grad_norm": 12.46586799621582, "learning_rate": 1.5076153430683295e-05, "loss": 1.3729, "step": 13310 }, { "epoch": 0.350869105672318, "grad_norm": 8.058752059936523, "learning_rate": 1.5068800892000561e-05, "loss": 1.3458, "step": 13320 }, { "epoch": 0.35113252091681674, "grad_norm": 18.819734573364258, "learning_rate": 1.5061444663825467e-05, "loss": 1.5226, "step": 13330 }, { "epoch": 0.3513959361613155, "grad_norm": 7.524699687957764, "learning_rate": 1.5054084751512483e-05, "loss": 1.2759, "step": 13340 }, { "epoch": 0.3516593514058142, "grad_norm": 7.355799198150635, "learning_rate": 1.5046721160418769e-05, "loss": 1.4278, "step": 13350 }, { "epoch": 0.35192276665031297, "grad_norm": 7.913285255432129, "learning_rate": 1.503935389590415e-05, "loss": 1.3536, "step": 13360 }, { "epoch": 0.3521861818948117, "grad_norm": 7.3410234451293945, "learning_rate": 1.503198296333113e-05, "loss": 1.404, "step": 13370 }, { "epoch": 0.35244959713931046, "grad_norm": 4.592426776885986, "learning_rate": 1.5024608368064883e-05, "loss": 1.5092, "step": 13380 }, { "epoch": 0.3527130123838092, "grad_norm": 8.372693061828613, "learning_rate": 1.5017230115473252e-05, "loss": 1.3978, "step": 13390 }, { "epoch": 0.35297642762830794, "grad_norm": 5.466429710388184, "learning_rate": 1.5009848210926736e-05, "loss": 1.3584, "step": 13400 }, { "epoch": 0.35323984287280663, "grad_norm": 12.244159698486328, "learning_rate": 1.5002462659798498e-05, "loss": 1.3713, "step": 13410 }, { "epoch": 0.3535032581173054, "grad_norm": 18.255769729614258, "learning_rate": 1.4995073467464352e-05, "loss": 1.4009, "step": 13420 }, { "epoch": 0.3537666733618041, "grad_norm": 10.710450172424316, "learning_rate": 1.4987680639302761e-05, "loss": 1.3289, "step": 13430 }, { "epoch": 0.35403008860630286, "grad_norm": 35.073429107666016, "learning_rate": 1.4980284180694839e-05, "loss": 1.312, "step": 13440 }, { "epoch": 0.3542935038508016, "grad_norm": 12.47243881225586, "learning_rate": 1.497288409702434e-05, "loss": 1.4485, "step": 13450 }, { "epoch": 0.35455691909530035, "grad_norm": 16.22907066345215, "learning_rate": 1.4965480393677656e-05, "loss": 1.3259, "step": 13460 }, { "epoch": 0.3548203343397991, "grad_norm": 24.07572364807129, "learning_rate": 1.4958073076043817e-05, "loss": 1.429, "step": 13470 }, { "epoch": 0.35508374958429784, "grad_norm": 7.839529991149902, "learning_rate": 1.4950662149514475e-05, "loss": 1.4567, "step": 13480 }, { "epoch": 0.3553471648287966, "grad_norm": 4.368978977203369, "learning_rate": 1.494324761948392e-05, "loss": 1.2798, "step": 13490 }, { "epoch": 0.35561058007329527, "grad_norm": 4.198204040527344, "learning_rate": 1.4935829491349059e-05, "loss": 1.263, "step": 13500 }, { "epoch": 0.355873995317794, "grad_norm": 6.534206867218018, "learning_rate": 1.4928407770509416e-05, "loss": 1.3813, "step": 13510 }, { "epoch": 0.35613741056229276, "grad_norm": 6.692596912384033, "learning_rate": 1.492098246236714e-05, "loss": 1.3221, "step": 13520 }, { "epoch": 0.3564008258067915, "grad_norm": 9.253410339355469, "learning_rate": 1.4913553572326978e-05, "loss": 1.4581, "step": 13530 }, { "epoch": 0.35666424105129024, "grad_norm": 8.166877746582031, "learning_rate": 1.4906121105796289e-05, "loss": 1.3248, "step": 13540 }, { "epoch": 0.356927656295789, "grad_norm": 4.618102073669434, "learning_rate": 1.4898685068185043e-05, "loss": 1.3628, "step": 13550 }, { "epoch": 0.35719107154028773, "grad_norm": 16.70050811767578, "learning_rate": 1.4891245464905799e-05, "loss": 1.2954, "step": 13560 }, { "epoch": 0.3574544867847865, "grad_norm": 5.224015712738037, "learning_rate": 1.488380230137372e-05, "loss": 1.2838, "step": 13570 }, { "epoch": 0.3577179020292852, "grad_norm": 2.4615893363952637, "learning_rate": 1.4876355583006546e-05, "loss": 1.3631, "step": 13580 }, { "epoch": 0.3579813172737839, "grad_norm": 10.01089096069336, "learning_rate": 1.4868905315224622e-05, "loss": 1.4275, "step": 13590 }, { "epoch": 0.35824473251828265, "grad_norm": 7.35436487197876, "learning_rate": 1.486145150345087e-05, "loss": 1.4282, "step": 13600 }, { "epoch": 0.3585081477627814, "grad_norm": 14.222929000854492, "learning_rate": 1.4853994153110784e-05, "loss": 1.3901, "step": 13610 }, { "epoch": 0.35877156300728014, "grad_norm": 19.453901290893555, "learning_rate": 1.4846533269632447e-05, "loss": 1.3364, "step": 13620 }, { "epoch": 0.3590349782517789, "grad_norm": 9.946592330932617, "learning_rate": 1.4839068858446498e-05, "loss": 1.4418, "step": 13630 }, { "epoch": 0.3592983934962776, "grad_norm": 18.906896591186523, "learning_rate": 1.4831600924986162e-05, "loss": 1.2256, "step": 13640 }, { "epoch": 0.35956180874077637, "grad_norm": 8.317885398864746, "learning_rate": 1.4824129474687215e-05, "loss": 1.2316, "step": 13650 }, { "epoch": 0.3598252239852751, "grad_norm": 11.247236251831055, "learning_rate": 1.4816654512987994e-05, "loss": 1.4655, "step": 13660 }, { "epoch": 0.3600886392297738, "grad_norm": 2.615150213241577, "learning_rate": 1.4809176045329397e-05, "loss": 1.4404, "step": 13670 }, { "epoch": 0.36035205447427254, "grad_norm": 5.09364128112793, "learning_rate": 1.480169407715487e-05, "loss": 1.4244, "step": 13680 }, { "epoch": 0.3606154697187713, "grad_norm": 12.840813636779785, "learning_rate": 1.4794208613910405e-05, "loss": 1.4407, "step": 13690 }, { "epoch": 0.36087888496327003, "grad_norm": 7.254754066467285, "learning_rate": 1.4786719661044548e-05, "loss": 1.3371, "step": 13700 }, { "epoch": 0.3611423002077688, "grad_norm": 20.475887298583984, "learning_rate": 1.4779227224008372e-05, "loss": 1.3778, "step": 13710 }, { "epoch": 0.3614057154522675, "grad_norm": 5.804864883422852, "learning_rate": 1.4771731308255492e-05, "loss": 1.4109, "step": 13720 }, { "epoch": 0.36166913069676626, "grad_norm": 7.684752941131592, "learning_rate": 1.476423191924206e-05, "loss": 1.4241, "step": 13730 }, { "epoch": 0.361932545941265, "grad_norm": 5.1446614265441895, "learning_rate": 1.4756729062426748e-05, "loss": 1.5095, "step": 13740 }, { "epoch": 0.36219596118576375, "grad_norm": 5.480652809143066, "learning_rate": 1.4749222743270757e-05, "loss": 1.3515, "step": 13750 }, { "epoch": 0.36245937643026244, "grad_norm": 9.661053657531738, "learning_rate": 1.4741712967237804e-05, "loss": 1.4088, "step": 13760 }, { "epoch": 0.3627227916747612, "grad_norm": 9.502453804016113, "learning_rate": 1.4734199739794128e-05, "loss": 1.404, "step": 13770 }, { "epoch": 0.3629862069192599, "grad_norm": 12.694108009338379, "learning_rate": 1.4726683066408478e-05, "loss": 1.4646, "step": 13780 }, { "epoch": 0.36324962216375867, "grad_norm": 17.397258758544922, "learning_rate": 1.4719162952552104e-05, "loss": 1.4172, "step": 13790 }, { "epoch": 0.3635130374082574, "grad_norm": 19.81020736694336, "learning_rate": 1.4711639403698772e-05, "loss": 1.4191, "step": 13800 }, { "epoch": 0.36377645265275615, "grad_norm": 8.570978164672852, "learning_rate": 1.4704112425324743e-05, "loss": 1.4809, "step": 13810 }, { "epoch": 0.3640398678972549, "grad_norm": 5.395107746124268, "learning_rate": 1.4696582022908772e-05, "loss": 1.4388, "step": 13820 }, { "epoch": 0.36430328314175364, "grad_norm": 11.466837882995605, "learning_rate": 1.468904820193211e-05, "loss": 1.3243, "step": 13830 }, { "epoch": 0.3645666983862524, "grad_norm": 6.657503128051758, "learning_rate": 1.4681510967878491e-05, "loss": 1.4434, "step": 13840 }, { "epoch": 0.3648301136307511, "grad_norm": 10.612898826599121, "learning_rate": 1.4673970326234143e-05, "loss": 1.4736, "step": 13850 }, { "epoch": 0.3650935288752498, "grad_norm": 13.192609786987305, "learning_rate": 1.4666426282487765e-05, "loss": 1.4404, "step": 13860 }, { "epoch": 0.36535694411974856, "grad_norm": 5.312329292297363, "learning_rate": 1.4658878842130535e-05, "loss": 1.4091, "step": 13870 }, { "epoch": 0.3656203593642473, "grad_norm": 20.19355583190918, "learning_rate": 1.4651328010656105e-05, "loss": 1.3332, "step": 13880 }, { "epoch": 0.36588377460874605, "grad_norm": 5.760142803192139, "learning_rate": 1.4643773793560599e-05, "loss": 1.3347, "step": 13890 }, { "epoch": 0.3661471898532448, "grad_norm": 5.5318827629089355, "learning_rate": 1.4636216196342592e-05, "loss": 1.2946, "step": 13900 }, { "epoch": 0.36641060509774354, "grad_norm": 6.95716667175293, "learning_rate": 1.4628655224503136e-05, "loss": 1.3397, "step": 13910 }, { "epoch": 0.3666740203422423, "grad_norm": 9.773702621459961, "learning_rate": 1.4621090883545726e-05, "loss": 1.4193, "step": 13920 }, { "epoch": 0.36693743558674097, "grad_norm": 8.297050476074219, "learning_rate": 1.4613523178976315e-05, "loss": 1.4935, "step": 13930 }, { "epoch": 0.3672008508312397, "grad_norm": 2.1775271892547607, "learning_rate": 1.460595211630331e-05, "loss": 1.4406, "step": 13940 }, { "epoch": 0.36746426607573845, "grad_norm": 14.061629295349121, "learning_rate": 1.459837770103755e-05, "loss": 1.3526, "step": 13950 }, { "epoch": 0.3677276813202372, "grad_norm": 5.127872467041016, "learning_rate": 1.4590799938692322e-05, "loss": 1.3323, "step": 13960 }, { "epoch": 0.36799109656473594, "grad_norm": 7.264883518218994, "learning_rate": 1.458321883478335e-05, "loss": 1.4106, "step": 13970 }, { "epoch": 0.3682545118092347, "grad_norm": 5.5468950271606445, "learning_rate": 1.4575634394828785e-05, "loss": 1.4023, "step": 13980 }, { "epoch": 0.36851792705373343, "grad_norm": 4.172088146209717, "learning_rate": 1.4568046624349215e-05, "loss": 1.4126, "step": 13990 }, { "epoch": 0.3687813422982322, "grad_norm": 2.0598316192626953, "learning_rate": 1.456045552886764e-05, "loss": 1.3878, "step": 14000 }, { "epoch": 0.3690447575427309, "grad_norm": 7.383871078491211, "learning_rate": 1.4552861113909491e-05, "loss": 1.5055, "step": 14010 }, { "epoch": 0.3693081727872296, "grad_norm": 3.057752847671509, "learning_rate": 1.4545263385002612e-05, "loss": 1.3048, "step": 14020 }, { "epoch": 0.36957158803172835, "grad_norm": 10.646309852600098, "learning_rate": 1.4537662347677252e-05, "loss": 1.3537, "step": 14030 }, { "epoch": 0.3698350032762271, "grad_norm": 12.763510704040527, "learning_rate": 1.4530058007466086e-05, "loss": 1.3828, "step": 14040 }, { "epoch": 0.37009841852072584, "grad_norm": 9.680891036987305, "learning_rate": 1.4522450369904169e-05, "loss": 1.5095, "step": 14050 }, { "epoch": 0.3703618337652246, "grad_norm": 12.855578422546387, "learning_rate": 1.4514839440528973e-05, "loss": 1.3467, "step": 14060 }, { "epoch": 0.3706252490097233, "grad_norm": 10.740635871887207, "learning_rate": 1.4507225224880363e-05, "loss": 1.2941, "step": 14070 }, { "epoch": 0.37088866425422207, "grad_norm": 10.499129295349121, "learning_rate": 1.4499607728500598e-05, "loss": 1.4473, "step": 14080 }, { "epoch": 0.3711520794987208, "grad_norm": 10.872820854187012, "learning_rate": 1.4491986956934315e-05, "loss": 1.3218, "step": 14090 }, { "epoch": 0.37141549474321955, "grad_norm": 8.625067710876465, "learning_rate": 1.4484362915728543e-05, "loss": 1.3825, "step": 14100 }, { "epoch": 0.37167890998771824, "grad_norm": 13.40442180633545, "learning_rate": 1.4476735610432693e-05, "loss": 1.3618, "step": 14110 }, { "epoch": 0.371942325232217, "grad_norm": 15.46483325958252, "learning_rate": 1.4469105046598546e-05, "loss": 1.4133, "step": 14120 }, { "epoch": 0.37220574047671573, "grad_norm": 10.819055557250977, "learning_rate": 1.4461471229780255e-05, "loss": 1.3227, "step": 14130 }, { "epoch": 0.3724691557212145, "grad_norm": 90.15763092041016, "learning_rate": 1.4453834165534348e-05, "loss": 1.3726, "step": 14140 }, { "epoch": 0.3727325709657132, "grad_norm": 121.56712341308594, "learning_rate": 1.444619385941971e-05, "loss": 1.3874, "step": 14150 }, { "epoch": 0.37299598621021196, "grad_norm": 9.503137588500977, "learning_rate": 1.4438550316997585e-05, "loss": 1.4003, "step": 14160 }, { "epoch": 0.3732594014547107, "grad_norm": 9.128656387329102, "learning_rate": 1.4430903543831579e-05, "loss": 1.4342, "step": 14170 }, { "epoch": 0.37352281669920945, "grad_norm": 8.004496574401855, "learning_rate": 1.442325354548764e-05, "loss": 1.4971, "step": 14180 }, { "epoch": 0.37378623194370814, "grad_norm": 18.990219116210938, "learning_rate": 1.4415600327534076e-05, "loss": 1.2692, "step": 14190 }, { "epoch": 0.3740496471882069, "grad_norm": 3.104677200317383, "learning_rate": 1.4407943895541527e-05, "loss": 1.4004, "step": 14200 }, { "epoch": 0.3743130624327056, "grad_norm": 12.727896690368652, "learning_rate": 1.4400284255082979e-05, "loss": 1.3113, "step": 14210 }, { "epoch": 0.37457647767720437, "grad_norm": 3.225799798965454, "learning_rate": 1.4392621411733753e-05, "loss": 1.4055, "step": 14220 }, { "epoch": 0.3748398929217031, "grad_norm": 8.016375541687012, "learning_rate": 1.4384955371071501e-05, "loss": 1.3864, "step": 14230 }, { "epoch": 0.37510330816620185, "grad_norm": 2.8889567852020264, "learning_rate": 1.437728613867619e-05, "loss": 1.3598, "step": 14240 }, { "epoch": 0.3753667234107006, "grad_norm": 5.612499237060547, "learning_rate": 1.4369613720130138e-05, "loss": 1.4011, "step": 14250 }, { "epoch": 0.37563013865519934, "grad_norm": 4.765489101409912, "learning_rate": 1.4361938121017953e-05, "loss": 1.3727, "step": 14260 }, { "epoch": 0.3758935538996981, "grad_norm": 6.395460605621338, "learning_rate": 1.4354259346926576e-05, "loss": 1.3924, "step": 14270 }, { "epoch": 0.3761569691441968, "grad_norm": 6.062030792236328, "learning_rate": 1.4346577403445255e-05, "loss": 1.5624, "step": 14280 }, { "epoch": 0.3764203843886955, "grad_norm": 5.744520664215088, "learning_rate": 1.4338892296165538e-05, "loss": 1.2837, "step": 14290 }, { "epoch": 0.37668379963319426, "grad_norm": 16.997982025146484, "learning_rate": 1.4331204030681285e-05, "loss": 1.4819, "step": 14300 }, { "epoch": 0.376947214877693, "grad_norm": 7.562694549560547, "learning_rate": 1.4323512612588655e-05, "loss": 1.4081, "step": 14310 }, { "epoch": 0.37721063012219175, "grad_norm": 5.358477592468262, "learning_rate": 1.431581804748609e-05, "loss": 1.3309, "step": 14320 }, { "epoch": 0.3774740453666905, "grad_norm": 3.4188365936279297, "learning_rate": 1.4308120340974339e-05, "loss": 1.205, "step": 14330 }, { "epoch": 0.37773746061118924, "grad_norm": 17.949207305908203, "learning_rate": 1.4300419498656419e-05, "loss": 1.3466, "step": 14340 }, { "epoch": 0.378000875855688, "grad_norm": 12.3236722946167, "learning_rate": 1.4292715526137646e-05, "loss": 1.4071, "step": 14350 }, { "epoch": 0.3782642911001867, "grad_norm": 9.832478523254395, "learning_rate": 1.4285008429025613e-05, "loss": 1.3305, "step": 14360 }, { "epoch": 0.3785277063446854, "grad_norm": 7.953855037689209, "learning_rate": 1.4277298212930172e-05, "loss": 1.3967, "step": 14370 }, { "epoch": 0.37879112158918415, "grad_norm": 8.537074089050293, "learning_rate": 1.4269584883463466e-05, "loss": 1.4285, "step": 14380 }, { "epoch": 0.3790545368336829, "grad_norm": 27.083322525024414, "learning_rate": 1.4261868446239886e-05, "loss": 1.3092, "step": 14390 }, { "epoch": 0.37931795207818164, "grad_norm": 20.45146942138672, "learning_rate": 1.4254148906876095e-05, "loss": 1.4678, "step": 14400 }, { "epoch": 0.3795813673226804, "grad_norm": 5.861148357391357, "learning_rate": 1.4246426270991014e-05, "loss": 1.3874, "step": 14410 }, { "epoch": 0.37984478256717913, "grad_norm": 43.94831085205078, "learning_rate": 1.4238700544205814e-05, "loss": 1.501, "step": 14420 }, { "epoch": 0.3801081978116779, "grad_norm": 94.89093780517578, "learning_rate": 1.423097173214392e-05, "loss": 1.3529, "step": 14430 }, { "epoch": 0.3803716130561766, "grad_norm": 11.575323104858398, "learning_rate": 1.4223239840430996e-05, "loss": 1.5298, "step": 14440 }, { "epoch": 0.3806350283006753, "grad_norm": 192.29547119140625, "learning_rate": 1.4215504874694957e-05, "loss": 1.3726, "step": 14450 }, { "epoch": 0.38089844354517405, "grad_norm": 28.046993255615234, "learning_rate": 1.4207766840565949e-05, "loss": 1.9216, "step": 14460 }, { "epoch": 0.3811618587896728, "grad_norm": 54.38966751098633, "learning_rate": 1.4200025743676348e-05, "loss": 1.9393, "step": 14470 }, { "epoch": 0.38142527403417154, "grad_norm": 20.182350158691406, "learning_rate": 1.4192281589660771e-05, "loss": 5.9251, "step": 14480 }, { "epoch": 0.3816886892786703, "grad_norm": 47.746761322021484, "learning_rate": 1.4184534384156053e-05, "loss": 1.5837, "step": 14490 }, { "epoch": 0.381952104523169, "grad_norm": 38.669559478759766, "learning_rate": 1.4176784132801248e-05, "loss": 1.4632, "step": 14500 }, { "epoch": 0.38221551976766777, "grad_norm": 35.13742446899414, "learning_rate": 1.416903084123763e-05, "loss": 1.3411, "step": 14510 }, { "epoch": 0.3824789350121665, "grad_norm": 26.411706924438477, "learning_rate": 1.4161274515108686e-05, "loss": 1.5661, "step": 14520 }, { "epoch": 0.38274235025666525, "grad_norm": 22.797040939331055, "learning_rate": 1.4153515160060111e-05, "loss": 1.3658, "step": 14530 }, { "epoch": 0.38300576550116394, "grad_norm": 19.911762237548828, "learning_rate": 1.4145752781739805e-05, "loss": 1.3823, "step": 14540 }, { "epoch": 0.3832691807456627, "grad_norm": 5.983964920043945, "learning_rate": 1.4137987385797871e-05, "loss": 1.5458, "step": 14550 }, { "epoch": 0.38353259599016143, "grad_norm": 9.547913551330566, "learning_rate": 1.4130218977886605e-05, "loss": 1.5966, "step": 14560 }, { "epoch": 0.3837960112346602, "grad_norm": 7.954235553741455, "learning_rate": 1.4122447563660498e-05, "loss": 1.3778, "step": 14570 }, { "epoch": 0.3840594264791589, "grad_norm": 30.239574432373047, "learning_rate": 1.4114673148776221e-05, "loss": 1.5369, "step": 14580 }, { "epoch": 0.38432284172365766, "grad_norm": 10.088793754577637, "learning_rate": 1.4106895738892647e-05, "loss": 1.4199, "step": 14590 }, { "epoch": 0.3845862569681564, "grad_norm": 14.220598220825195, "learning_rate": 1.4099115339670808e-05, "loss": 1.4314, "step": 14600 }, { "epoch": 0.38484967221265515, "grad_norm": 58.917606353759766, "learning_rate": 1.4091331956773924e-05, "loss": 1.4164, "step": 14610 }, { "epoch": 0.3851130874571539, "grad_norm": 30.70146942138672, "learning_rate": 1.408354559586739e-05, "loss": 1.4864, "step": 14620 }, { "epoch": 0.3853765027016526, "grad_norm": 45.016849517822266, "learning_rate": 1.4075756262618757e-05, "loss": 1.4774, "step": 14630 }, { "epoch": 0.3856399179461513, "grad_norm": 21.885793685913086, "learning_rate": 1.4067963962697748e-05, "loss": 1.5272, "step": 14640 }, { "epoch": 0.38590333319065007, "grad_norm": 33.49884033203125, "learning_rate": 1.4060168701776242e-05, "loss": 1.4403, "step": 14650 }, { "epoch": 0.3861667484351488, "grad_norm": 49.4317626953125, "learning_rate": 1.4052370485528277e-05, "loss": 1.4147, "step": 14660 }, { "epoch": 0.38643016367964755, "grad_norm": 18.248889923095703, "learning_rate": 1.4044569319630039e-05, "loss": 1.5087, "step": 14670 }, { "epoch": 0.3866935789241463, "grad_norm": 12.690756797790527, "learning_rate": 1.403676520975986e-05, "loss": 1.2942, "step": 14680 }, { "epoch": 0.38695699416864504, "grad_norm": 10.565882682800293, "learning_rate": 1.4028958161598217e-05, "loss": 1.3361, "step": 14690 }, { "epoch": 0.3872204094131438, "grad_norm": 11.384231567382812, "learning_rate": 1.4021148180827726e-05, "loss": 1.2748, "step": 14700 }, { "epoch": 0.3874838246576425, "grad_norm": 14.115474700927734, "learning_rate": 1.4013335273133132e-05, "loss": 1.648, "step": 14710 }, { "epoch": 0.3877472399021412, "grad_norm": 90.19509887695312, "learning_rate": 1.4005519444201323e-05, "loss": 1.425, "step": 14720 }, { "epoch": 0.38801065514663996, "grad_norm": 12794.9755859375, "learning_rate": 1.3997700699721302e-05, "loss": 1.3708, "step": 14730 }, { "epoch": 0.3882740703911387, "grad_norm": 14.234919548034668, "learning_rate": 1.3989879045384194e-05, "loss": 1.4427, "step": 14740 }, { "epoch": 0.38853748563563745, "grad_norm": 12.258428573608398, "learning_rate": 1.398205448688325e-05, "loss": 1.5527, "step": 14750 }, { "epoch": 0.3888009008801362, "grad_norm": 13.358673095703125, "learning_rate": 1.3974227029913827e-05, "loss": 1.3268, "step": 14760 }, { "epoch": 0.38906431612463493, "grad_norm": 16.93588638305664, "learning_rate": 1.3966396680173403e-05, "loss": 1.4112, "step": 14770 }, { "epoch": 0.3893277313691337, "grad_norm": 14.2131929397583, "learning_rate": 1.3958563443361543e-05, "loss": 1.3377, "step": 14780 }, { "epoch": 0.3895911466136324, "grad_norm": 7.659984588623047, "learning_rate": 1.3950727325179934e-05, "loss": 1.4165, "step": 14790 }, { "epoch": 0.3898545618581311, "grad_norm": 907.00146484375, "learning_rate": 1.3942888331332347e-05, "loss": 1.3507, "step": 14800 }, { "epoch": 0.39011797710262985, "grad_norm": 19.37653923034668, "learning_rate": 1.3935046467524647e-05, "loss": 1.4552, "step": 14810 }, { "epoch": 0.3903813923471286, "grad_norm": 339.8660888671875, "learning_rate": 1.3927201739464793e-05, "loss": 1.4162, "step": 14820 }, { "epoch": 0.39064480759162734, "grad_norm": 42.60823440551758, "learning_rate": 1.391935415286283e-05, "loss": 1.3108, "step": 14830 }, { "epoch": 0.3909082228361261, "grad_norm": 12.07181453704834, "learning_rate": 1.3911503713430875e-05, "loss": 1.3716, "step": 14840 }, { "epoch": 0.39117163808062483, "grad_norm": 10.790331840515137, "learning_rate": 1.3903650426883128e-05, "loss": 1.4417, "step": 14850 }, { "epoch": 0.39143505332512357, "grad_norm": 16.120529174804688, "learning_rate": 1.3895794298935862e-05, "loss": 1.4093, "step": 14860 }, { "epoch": 0.3916984685696223, "grad_norm": 22.032482147216797, "learning_rate": 1.3887935335307417e-05, "loss": 1.2992, "step": 14870 }, { "epoch": 0.39196188381412106, "grad_norm": 14.725638389587402, "learning_rate": 1.3880073541718197e-05, "loss": 1.3545, "step": 14880 }, { "epoch": 0.39222529905861975, "grad_norm": 60.39795684814453, "learning_rate": 1.387220892389066e-05, "loss": 1.3432, "step": 14890 }, { "epoch": 0.3924887143031185, "grad_norm": 10.220460891723633, "learning_rate": 1.3864341487549331e-05, "loss": 1.3131, "step": 14900 }, { "epoch": 0.39275212954761723, "grad_norm": 56.70800018310547, "learning_rate": 1.3856471238420778e-05, "loss": 1.5443, "step": 14910 }, { "epoch": 0.393015544792116, "grad_norm": 13.991398811340332, "learning_rate": 1.3848598182233617e-05, "loss": 1.3388, "step": 14920 }, { "epoch": 0.3932789600366147, "grad_norm": 11.70394229888916, "learning_rate": 1.3840722324718516e-05, "loss": 1.3692, "step": 14930 }, { "epoch": 0.39354237528111347, "grad_norm": 69.9419174194336, "learning_rate": 1.3832843671608167e-05, "loss": 1.348, "step": 14940 }, { "epoch": 0.3938057905256122, "grad_norm": 47.75214385986328, "learning_rate": 1.3824962228637311e-05, "loss": 1.3852, "step": 14950 }, { "epoch": 0.39406920577011095, "grad_norm": 8.550277709960938, "learning_rate": 1.3817078001542711e-05, "loss": 1.3647, "step": 14960 }, { "epoch": 0.39433262101460964, "grad_norm": 12.094663619995117, "learning_rate": 1.380919099606316e-05, "loss": 1.2506, "step": 14970 }, { "epoch": 0.3945960362591084, "grad_norm": 29.14276695251465, "learning_rate": 1.3801301217939474e-05, "loss": 1.3535, "step": 14980 }, { "epoch": 0.39485945150360713, "grad_norm": 6.793166160583496, "learning_rate": 1.379340867291448e-05, "loss": 1.4151, "step": 14990 }, { "epoch": 0.39512286674810587, "grad_norm": 8.16550064086914, "learning_rate": 1.3785513366733034e-05, "loss": 1.4206, "step": 15000 }, { "epoch": 0.3953862819926046, "grad_norm": 18.75669288635254, "learning_rate": 1.3777615305141989e-05, "loss": 1.3404, "step": 15010 }, { "epoch": 0.39564969723710336, "grad_norm": 12.622206687927246, "learning_rate": 1.3769714493890203e-05, "loss": 1.437, "step": 15020 }, { "epoch": 0.3959131124816021, "grad_norm": 9.983560562133789, "learning_rate": 1.3761810938728541e-05, "loss": 1.4743, "step": 15030 }, { "epoch": 0.39617652772610085, "grad_norm": 19.950082778930664, "learning_rate": 1.375390464540987e-05, "loss": 1.4266, "step": 15040 }, { "epoch": 0.3964399429705996, "grad_norm": 220.5886688232422, "learning_rate": 1.3745995619689036e-05, "loss": 1.3316, "step": 15050 }, { "epoch": 0.3967033582150983, "grad_norm": 54.752262115478516, "learning_rate": 1.3738083867322885e-05, "loss": 1.3816, "step": 15060 }, { "epoch": 0.396966773459597, "grad_norm": 12.10908317565918, "learning_rate": 1.373016939407025e-05, "loss": 1.2748, "step": 15070 }, { "epoch": 0.39723018870409577, "grad_norm": 25.632116317749023, "learning_rate": 1.3722252205691929e-05, "loss": 1.4362, "step": 15080 }, { "epoch": 0.3974936039485945, "grad_norm": 15.033336639404297, "learning_rate": 1.3714332307950713e-05, "loss": 1.3004, "step": 15090 }, { "epoch": 0.39775701919309325, "grad_norm": 17.965044021606445, "learning_rate": 1.3706409706611354e-05, "loss": 1.4565, "step": 15100 }, { "epoch": 0.398020434437592, "grad_norm": 10.019550323486328, "learning_rate": 1.3698484407440582e-05, "loss": 1.3403, "step": 15110 }, { "epoch": 0.39828384968209074, "grad_norm": 21.99286460876465, "learning_rate": 1.3690556416207081e-05, "loss": 1.4399, "step": 15120 }, { "epoch": 0.3985472649265895, "grad_norm": 16.343158721923828, "learning_rate": 1.3682625738681496e-05, "loss": 1.2982, "step": 15130 }, { "epoch": 0.3988106801710882, "grad_norm": 37.03549575805664, "learning_rate": 1.3674692380636438e-05, "loss": 1.3837, "step": 15140 }, { "epoch": 0.3990740954155869, "grad_norm": 84.72125244140625, "learning_rate": 1.3666756347846452e-05, "loss": 1.4369, "step": 15150 }, { "epoch": 0.39933751066008566, "grad_norm": 25.428346633911133, "learning_rate": 1.3658817646088045e-05, "loss": 1.4695, "step": 15160 }, { "epoch": 0.3996009259045844, "grad_norm": 32.85851287841797, "learning_rate": 1.3650876281139657e-05, "loss": 1.3223, "step": 15170 }, { "epoch": 0.39986434114908315, "grad_norm": 10.869043350219727, "learning_rate": 1.364293225878167e-05, "loss": 1.3116, "step": 15180 }, { "epoch": 0.4001277563935819, "grad_norm": 74.46104431152344, "learning_rate": 1.3634985584796403e-05, "loss": 1.4307, "step": 15190 }, { "epoch": 0.40039117163808063, "grad_norm": 178.43858337402344, "learning_rate": 1.3627036264968098e-05, "loss": 1.3973, "step": 15200 }, { "epoch": 0.4006545868825794, "grad_norm": 8.72211742401123, "learning_rate": 1.3619084305082932e-05, "loss": 1.3751, "step": 15210 }, { "epoch": 0.4009180021270781, "grad_norm": 14.177505493164062, "learning_rate": 1.3611129710928998e-05, "loss": 1.3297, "step": 15220 }, { "epoch": 0.4011814173715768, "grad_norm": 32.30446243286133, "learning_rate": 1.3603172488296304e-05, "loss": 1.4046, "step": 15230 }, { "epoch": 0.40144483261607555, "grad_norm": 13.893932342529297, "learning_rate": 1.3595212642976775e-05, "loss": 1.4261, "step": 15240 }, { "epoch": 0.4017082478605743, "grad_norm": 170.42160034179688, "learning_rate": 1.358725018076425e-05, "loss": 1.2977, "step": 15250 }, { "epoch": 0.40197166310507304, "grad_norm": 37.966896057128906, "learning_rate": 1.357928510745446e-05, "loss": 1.3176, "step": 15260 }, { "epoch": 0.4022350783495718, "grad_norm": 8.381155967712402, "learning_rate": 1.3571317428845048e-05, "loss": 1.3153, "step": 15270 }, { "epoch": 0.4024984935940705, "grad_norm": 10.543867111206055, "learning_rate": 1.356334715073555e-05, "loss": 1.3636, "step": 15280 }, { "epoch": 0.40276190883856927, "grad_norm": 12.562834739685059, "learning_rate": 1.3555374278927392e-05, "loss": 1.2938, "step": 15290 }, { "epoch": 0.403025324083068, "grad_norm": 21.459482192993164, "learning_rate": 1.3547398819223885e-05, "loss": 1.2472, "step": 15300 }, { "epoch": 0.40328873932756676, "grad_norm": 8.907787322998047, "learning_rate": 1.3539420777430235e-05, "loss": 1.5274, "step": 15310 }, { "epoch": 0.40355215457206545, "grad_norm": 10.799487113952637, "learning_rate": 1.3531440159353518e-05, "loss": 1.4272, "step": 15320 }, { "epoch": 0.4038155698165642, "grad_norm": 6.930497646331787, "learning_rate": 1.3523456970802684e-05, "loss": 1.4475, "step": 15330 }, { "epoch": 0.40407898506106293, "grad_norm": 7.383173942565918, "learning_rate": 1.351547121758856e-05, "loss": 1.3667, "step": 15340 }, { "epoch": 0.4043424003055617, "grad_norm": 10.1578369140625, "learning_rate": 1.3507482905523843e-05, "loss": 1.4454, "step": 15350 }, { "epoch": 0.4046058155500604, "grad_norm": 5.529361724853516, "learning_rate": 1.3499492040423082e-05, "loss": 1.4197, "step": 15360 }, { "epoch": 0.40486923079455917, "grad_norm": 10.250508308410645, "learning_rate": 1.3491498628102683e-05, "loss": 1.4365, "step": 15370 }, { "epoch": 0.4051326460390579, "grad_norm": 4.068673610687256, "learning_rate": 1.3483502674380923e-05, "loss": 1.2701, "step": 15380 }, { "epoch": 0.40539606128355665, "grad_norm": 20.9726505279541, "learning_rate": 1.3475504185077915e-05, "loss": 1.3523, "step": 15390 }, { "epoch": 0.4056594765280554, "grad_norm": 5.3065185546875, "learning_rate": 1.3467503166015616e-05, "loss": 1.3059, "step": 15400 }, { "epoch": 0.4059228917725541, "grad_norm": 5.9524664878845215, "learning_rate": 1.3459499623017835e-05, "loss": 1.4898, "step": 15410 }, { "epoch": 0.40618630701705283, "grad_norm": 5.131724834442139, "learning_rate": 1.3451493561910211e-05, "loss": 1.3894, "step": 15420 }, { "epoch": 0.40644972226155157, "grad_norm": 11.42402172088623, "learning_rate": 1.3443484988520217e-05, "loss": 1.3855, "step": 15430 }, { "epoch": 0.4067131375060503, "grad_norm": 11.634478569030762, "learning_rate": 1.343547390867715e-05, "loss": 1.3705, "step": 15440 }, { "epoch": 0.40697655275054906, "grad_norm": 7.0573248863220215, "learning_rate": 1.3427460328212143e-05, "loss": 1.4243, "step": 15450 }, { "epoch": 0.4072399679950478, "grad_norm": 19.362293243408203, "learning_rate": 1.341944425295814e-05, "loss": 1.4012, "step": 15460 }, { "epoch": 0.40750338323954655, "grad_norm": 6.649326324462891, "learning_rate": 1.3411425688749901e-05, "loss": 1.268, "step": 15470 }, { "epoch": 0.4077667984840453, "grad_norm": 8.10913372039795, "learning_rate": 1.3403404641424007e-05, "loss": 1.3115, "step": 15480 }, { "epoch": 0.408030213728544, "grad_norm": 5.259890556335449, "learning_rate": 1.339538111681883e-05, "loss": 1.3388, "step": 15490 }, { "epoch": 0.4082936289730427, "grad_norm": 9.884650230407715, "learning_rate": 1.338735512077456e-05, "loss": 1.3575, "step": 15500 }, { "epoch": 0.40855704421754147, "grad_norm": 9.518773078918457, "learning_rate": 1.3379326659133181e-05, "loss": 1.3098, "step": 15510 }, { "epoch": 0.4088204594620402, "grad_norm": 9.161428451538086, "learning_rate": 1.3371295737738468e-05, "loss": 1.3905, "step": 15520 }, { "epoch": 0.40908387470653895, "grad_norm": 12.229033470153809, "learning_rate": 1.3363262362435992e-05, "loss": 1.4098, "step": 15530 }, { "epoch": 0.4093472899510377, "grad_norm": 11.670672416687012, "learning_rate": 1.3355226539073104e-05, "loss": 1.4855, "step": 15540 }, { "epoch": 0.40961070519553644, "grad_norm": 3.850428819656372, "learning_rate": 1.3347188273498948e-05, "loss": 1.3829, "step": 15550 }, { "epoch": 0.4098741204400352, "grad_norm": 9.473344802856445, "learning_rate": 1.3339147571564432e-05, "loss": 1.3849, "step": 15560 }, { "epoch": 0.4101375356845339, "grad_norm": 1.9616492986679077, "learning_rate": 1.3331104439122244e-05, "loss": 1.3736, "step": 15570 }, { "epoch": 0.4104009509290326, "grad_norm": 5.797630310058594, "learning_rate": 1.3323058882026843e-05, "loss": 1.4675, "step": 15580 }, { "epoch": 0.41066436617353136, "grad_norm": 15.889266967773438, "learning_rate": 1.3315010906134452e-05, "loss": 1.2048, "step": 15590 }, { "epoch": 0.4109277814180301, "grad_norm": 11.765287399291992, "learning_rate": 1.3306960517303049e-05, "loss": 1.4179, "step": 15600 }, { "epoch": 0.41119119666252885, "grad_norm": 4.846033573150635, "learning_rate": 1.3298907721392374e-05, "loss": 1.4878, "step": 15610 }, { "epoch": 0.4114546119070276, "grad_norm": 6.000058650970459, "learning_rate": 1.329085252426392e-05, "loss": 1.3669, "step": 15620 }, { "epoch": 0.41171802715152633, "grad_norm": 6.277578353881836, "learning_rate": 1.3282794931780924e-05, "loss": 1.3584, "step": 15630 }, { "epoch": 0.4119814423960251, "grad_norm": 3.944087505340576, "learning_rate": 1.3274734949808366e-05, "loss": 1.3247, "step": 15640 }, { "epoch": 0.4122448576405238, "grad_norm": 12.023463249206543, "learning_rate": 1.326667258421297e-05, "loss": 1.4622, "step": 15650 }, { "epoch": 0.41250827288502256, "grad_norm": 5.0492472648620605, "learning_rate": 1.325860784086319e-05, "loss": 1.4178, "step": 15660 }, { "epoch": 0.41277168812952125, "grad_norm": 8.90405559539795, "learning_rate": 1.3250540725629217e-05, "loss": 1.3226, "step": 15670 }, { "epoch": 0.41303510337402, "grad_norm": 10.668108940124512, "learning_rate": 1.3242471244382957e-05, "loss": 1.3729, "step": 15680 }, { "epoch": 0.41329851861851874, "grad_norm": 6.2647881507873535, "learning_rate": 1.3234399402998053e-05, "loss": 1.4294, "step": 15690 }, { "epoch": 0.4135619338630175, "grad_norm": 4.72715950012207, "learning_rate": 1.3226325207349855e-05, "loss": 1.4017, "step": 15700 }, { "epoch": 0.4138253491075162, "grad_norm": 13.50994873046875, "learning_rate": 1.3218248663315429e-05, "loss": 1.2765, "step": 15710 }, { "epoch": 0.41408876435201497, "grad_norm": 5.650006294250488, "learning_rate": 1.3210169776773558e-05, "loss": 1.4294, "step": 15720 }, { "epoch": 0.4143521795965137, "grad_norm": 8.880769729614258, "learning_rate": 1.3202088553604718e-05, "loss": 1.4633, "step": 15730 }, { "epoch": 0.41461559484101246, "grad_norm": 3.0964739322662354, "learning_rate": 1.3194004999691089e-05, "loss": 1.408, "step": 15740 }, { "epoch": 0.41487901008551115, "grad_norm": 30.252710342407227, "learning_rate": 1.3185919120916554e-05, "loss": 1.3881, "step": 15750 }, { "epoch": 0.4151424253300099, "grad_norm": 16.142927169799805, "learning_rate": 1.3177830923166684e-05, "loss": 1.4627, "step": 15760 }, { "epoch": 0.41540584057450863, "grad_norm": 4.905526638031006, "learning_rate": 1.3169740412328737e-05, "loss": 1.4821, "step": 15770 }, { "epoch": 0.4156692558190074, "grad_norm": 4.244324207305908, "learning_rate": 1.3161647594291656e-05, "loss": 1.3643, "step": 15780 }, { "epoch": 0.4159326710635061, "grad_norm": 7.677330017089844, "learning_rate": 1.315355247494606e-05, "loss": 1.4528, "step": 15790 }, { "epoch": 0.41619608630800486, "grad_norm": 4.299878120422363, "learning_rate": 1.3145455060184256e-05, "loss": 1.4125, "step": 15800 }, { "epoch": 0.4164595015525036, "grad_norm": 5.14202356338501, "learning_rate": 1.31373553559002e-05, "loss": 1.3351, "step": 15810 }, { "epoch": 0.41672291679700235, "grad_norm": 13.069316864013672, "learning_rate": 1.312925336798953e-05, "loss": 1.4487, "step": 15820 }, { "epoch": 0.4169863320415011, "grad_norm": 5.252227783203125, "learning_rate": 1.3121149102349549e-05, "loss": 1.4011, "step": 15830 }, { "epoch": 0.4172497472859998, "grad_norm": 12.533105850219727, "learning_rate": 1.3113042564879201e-05, "loss": 1.4049, "step": 15840 }, { "epoch": 0.4175131625304985, "grad_norm": 9.362533569335938, "learning_rate": 1.3104933761479103e-05, "loss": 1.2646, "step": 15850 }, { "epoch": 0.41777657777499727, "grad_norm": 6.303677558898926, "learning_rate": 1.309682269805151e-05, "loss": 1.4254, "step": 15860 }, { "epoch": 0.418039993019496, "grad_norm": 12.081802368164062, "learning_rate": 1.3088709380500322e-05, "loss": 1.3672, "step": 15870 }, { "epoch": 0.41830340826399476, "grad_norm": 6.973888397216797, "learning_rate": 1.3080593814731084e-05, "loss": 1.3946, "step": 15880 }, { "epoch": 0.4185668235084935, "grad_norm": 11.671911239624023, "learning_rate": 1.3072476006650976e-05, "loss": 1.4299, "step": 15890 }, { "epoch": 0.41883023875299225, "grad_norm": 15.659972190856934, "learning_rate": 1.3064355962168813e-05, "loss": 1.447, "step": 15900 }, { "epoch": 0.419093653997491, "grad_norm": 7.726922035217285, "learning_rate": 1.305623368719503e-05, "loss": 1.369, "step": 15910 }, { "epoch": 0.41935706924198973, "grad_norm": 9.023506164550781, "learning_rate": 1.3048109187641687e-05, "loss": 1.3382, "step": 15920 }, { "epoch": 0.4196204844864884, "grad_norm": 8.783249855041504, "learning_rate": 1.3039982469422476e-05, "loss": 1.362, "step": 15930 }, { "epoch": 0.41988389973098716, "grad_norm": 3.572702407836914, "learning_rate": 1.303185353845269e-05, "loss": 1.3394, "step": 15940 }, { "epoch": 0.4201473149754859, "grad_norm": 2.8077034950256348, "learning_rate": 1.3023722400649234e-05, "loss": 1.3894, "step": 15950 }, { "epoch": 0.42041073021998465, "grad_norm": 6.189970016479492, "learning_rate": 1.3015589061930627e-05, "loss": 1.4683, "step": 15960 }, { "epoch": 0.4206741454644834, "grad_norm": 3.104602575302124, "learning_rate": 1.3007453528216981e-05, "loss": 1.4569, "step": 15970 }, { "epoch": 0.42093756070898214, "grad_norm": 3.7802796363830566, "learning_rate": 1.2999315805430015e-05, "loss": 1.407, "step": 15980 }, { "epoch": 0.4212009759534809, "grad_norm": 4.91039514541626, "learning_rate": 1.2991175899493032e-05, "loss": 1.3797, "step": 15990 }, { "epoch": 0.4214643911979796, "grad_norm": 7.9525146484375, "learning_rate": 1.298303381633093e-05, "loss": 1.3416, "step": 16000 }, { "epoch": 0.4217278064424783, "grad_norm": 9.515143394470215, "learning_rate": 1.2974889561870195e-05, "loss": 1.3937, "step": 16010 }, { "epoch": 0.42199122168697706, "grad_norm": 24.651504516601562, "learning_rate": 1.2966743142038879e-05, "loss": 1.3005, "step": 16020 }, { "epoch": 0.4222546369314758, "grad_norm": 10.607142448425293, "learning_rate": 1.2958594562766627e-05, "loss": 1.2954, "step": 16030 }, { "epoch": 0.42251805217597455, "grad_norm": 3.8237063884735107, "learning_rate": 1.2950443829984649e-05, "loss": 1.3965, "step": 16040 }, { "epoch": 0.4227814674204733, "grad_norm": 9.423677444458008, "learning_rate": 1.2942290949625717e-05, "loss": 1.4714, "step": 16050 }, { "epoch": 0.42304488266497203, "grad_norm": 13.82089614868164, "learning_rate": 1.2934135927624176e-05, "loss": 1.4917, "step": 16060 }, { "epoch": 0.4233082979094708, "grad_norm": 2.9570960998535156, "learning_rate": 1.2925978769915921e-05, "loss": 1.3696, "step": 16070 }, { "epoch": 0.4235717131539695, "grad_norm": 10.938511848449707, "learning_rate": 1.2917819482438412e-05, "loss": 1.3548, "step": 16080 }, { "epoch": 0.42383512839846826, "grad_norm": 13.101497650146484, "learning_rate": 1.2909658071130648e-05, "loss": 1.4372, "step": 16090 }, { "epoch": 0.42409854364296695, "grad_norm": 7.241787910461426, "learning_rate": 1.2901494541933181e-05, "loss": 1.3544, "step": 16100 }, { "epoch": 0.4243619588874657, "grad_norm": 3.909675121307373, "learning_rate": 1.2893328900788102e-05, "loss": 1.3894, "step": 16110 }, { "epoch": 0.42462537413196444, "grad_norm": 9.922338485717773, "learning_rate": 1.2885161153639039e-05, "loss": 1.3848, "step": 16120 }, { "epoch": 0.4248887893764632, "grad_norm": 18.903465270996094, "learning_rate": 1.2876991306431153e-05, "loss": 1.4313, "step": 16130 }, { "epoch": 0.4251522046209619, "grad_norm": 5.6014299392700195, "learning_rate": 1.2868819365111138e-05, "loss": 1.4513, "step": 16140 }, { "epoch": 0.42541561986546067, "grad_norm": 19.40557098388672, "learning_rate": 1.2860645335627207e-05, "loss": 1.3996, "step": 16150 }, { "epoch": 0.4256790351099594, "grad_norm": 6.557480335235596, "learning_rate": 1.2852469223929087e-05, "loss": 1.5025, "step": 16160 }, { "epoch": 0.42594245035445816, "grad_norm": 9.43835163116455, "learning_rate": 1.2844291035968042e-05, "loss": 1.3485, "step": 16170 }, { "epoch": 0.4262058655989569, "grad_norm": 4.642200946807861, "learning_rate": 1.2836110777696822e-05, "loss": 1.32, "step": 16180 }, { "epoch": 0.4264692808434556, "grad_norm": 5.85961389541626, "learning_rate": 1.28279284550697e-05, "loss": 1.4316, "step": 16190 }, { "epoch": 0.42673269608795433, "grad_norm": 12.451931953430176, "learning_rate": 1.2819744074042446e-05, "loss": 1.4176, "step": 16200 }, { "epoch": 0.4269961113324531, "grad_norm": 6.895419120788574, "learning_rate": 1.2811557640572329e-05, "loss": 1.4121, "step": 16210 }, { "epoch": 0.4272595265769518, "grad_norm": 7.075610637664795, "learning_rate": 1.2803369160618115e-05, "loss": 1.2888, "step": 16220 }, { "epoch": 0.42752294182145056, "grad_norm": 10.624726295471191, "learning_rate": 1.279517864014005e-05, "loss": 1.3997, "step": 16230 }, { "epoch": 0.4277863570659493, "grad_norm": 18.465593338012695, "learning_rate": 1.2786986085099879e-05, "loss": 1.4445, "step": 16240 }, { "epoch": 0.42804977231044805, "grad_norm": 3.063359260559082, "learning_rate": 1.2778791501460817e-05, "loss": 1.3344, "step": 16250 }, { "epoch": 0.4283131875549468, "grad_norm": 5.968415260314941, "learning_rate": 1.2770594895187558e-05, "loss": 1.2595, "step": 16260 }, { "epoch": 0.42857660279944554, "grad_norm": 12.125001907348633, "learning_rate": 1.2762396272246273e-05, "loss": 1.3247, "step": 16270 }, { "epoch": 0.4288400180439442, "grad_norm": 4.532658100128174, "learning_rate": 1.27541956386046e-05, "loss": 1.3405, "step": 16280 }, { "epoch": 0.42910343328844297, "grad_norm": 11.607972145080566, "learning_rate": 1.2745993000231632e-05, "loss": 1.4102, "step": 16290 }, { "epoch": 0.4293668485329417, "grad_norm": 6.874989986419678, "learning_rate": 1.273778836309793e-05, "loss": 1.3324, "step": 16300 }, { "epoch": 0.42963026377744046, "grad_norm": 12.042860984802246, "learning_rate": 1.2729581733175508e-05, "loss": 1.2742, "step": 16310 }, { "epoch": 0.4298936790219392, "grad_norm": 13.916406631469727, "learning_rate": 1.2721373116437832e-05, "loss": 1.4552, "step": 16320 }, { "epoch": 0.43015709426643794, "grad_norm": 10.46174144744873, "learning_rate": 1.2713162518859803e-05, "loss": 1.3336, "step": 16330 }, { "epoch": 0.4304205095109367, "grad_norm": 12.4313325881958, "learning_rate": 1.2704949946417781e-05, "loss": 1.3482, "step": 16340 }, { "epoch": 0.43068392475543543, "grad_norm": 18.28696632385254, "learning_rate": 1.2696735405089557e-05, "loss": 1.4838, "step": 16350 }, { "epoch": 0.4309473399999341, "grad_norm": 6.721951484680176, "learning_rate": 1.2688518900854347e-05, "loss": 1.4805, "step": 16360 }, { "epoch": 0.43121075524443286, "grad_norm": 11.45789909362793, "learning_rate": 1.2680300439692803e-05, "loss": 1.3802, "step": 16370 }, { "epoch": 0.4314741704889316, "grad_norm": 17.696706771850586, "learning_rate": 1.2672080027587003e-05, "loss": 1.4597, "step": 16380 }, { "epoch": 0.43173758573343035, "grad_norm": 8.663355827331543, "learning_rate": 1.2663857670520442e-05, "loss": 1.451, "step": 16390 }, { "epoch": 0.4320010009779291, "grad_norm": 25.208045959472656, "learning_rate": 1.2655633374478031e-05, "loss": 1.4477, "step": 16400 }, { "epoch": 0.43226441622242784, "grad_norm": 8.942983627319336, "learning_rate": 1.2647407145446092e-05, "loss": 1.2936, "step": 16410 }, { "epoch": 0.4325278314669266, "grad_norm": 6.198161602020264, "learning_rate": 1.2639178989412356e-05, "loss": 1.3445, "step": 16420 }, { "epoch": 0.4327912467114253, "grad_norm": 2.4878838062286377, "learning_rate": 1.263094891236595e-05, "loss": 1.3899, "step": 16430 }, { "epoch": 0.43305466195592407, "grad_norm": 6.758609294891357, "learning_rate": 1.2622716920297413e-05, "loss": 1.4104, "step": 16440 }, { "epoch": 0.43331807720042276, "grad_norm": 8.211750984191895, "learning_rate": 1.2614483019198667e-05, "loss": 1.3305, "step": 16450 }, { "epoch": 0.4335814924449215, "grad_norm": 18.20709991455078, "learning_rate": 1.2606247215063019e-05, "loss": 1.4101, "step": 16460 }, { "epoch": 0.43384490768942024, "grad_norm": 2.9763343334198, "learning_rate": 1.2598009513885174e-05, "loss": 1.4965, "step": 16470 }, { "epoch": 0.434108322933919, "grad_norm": 7.870472431182861, "learning_rate": 1.2589769921661215e-05, "loss": 1.2799, "step": 16480 }, { "epoch": 0.43437173817841773, "grad_norm": 7.063776016235352, "learning_rate": 1.2581528444388593e-05, "loss": 1.49, "step": 16490 }, { "epoch": 0.4346351534229165, "grad_norm": 7.303397178649902, "learning_rate": 1.2573285088066137e-05, "loss": 1.4383, "step": 16500 }, { "epoch": 0.4348985686674152, "grad_norm": 82.07682800292969, "learning_rate": 1.2565039858694044e-05, "loss": 1.3957, "step": 16510 }, { "epoch": 0.43516198391191396, "grad_norm": 2.4122869968414307, "learning_rate": 1.2556792762273875e-05, "loss": 1.3496, "step": 16520 }, { "epoch": 0.4354253991564127, "grad_norm": 5.706669330596924, "learning_rate": 1.2548543804808549e-05, "loss": 1.3585, "step": 16530 }, { "epoch": 0.4356888144009114, "grad_norm": 6.834168434143066, "learning_rate": 1.2540292992302336e-05, "loss": 1.2756, "step": 16540 }, { "epoch": 0.43595222964541014, "grad_norm": 8.142898559570312, "learning_rate": 1.2532040330760863e-05, "loss": 1.264, "step": 16550 }, { "epoch": 0.4362156448899089, "grad_norm": 3.8998258113861084, "learning_rate": 1.25237858261911e-05, "loss": 1.3304, "step": 16560 }, { "epoch": 0.4364790601344076, "grad_norm": 5.435478687286377, "learning_rate": 1.2515529484601349e-05, "loss": 1.3694, "step": 16570 }, { "epoch": 0.43674247537890637, "grad_norm": 7.5678205490112305, "learning_rate": 1.250727131200127e-05, "loss": 1.4248, "step": 16580 }, { "epoch": 0.4370058906234051, "grad_norm": 11.584285736083984, "learning_rate": 1.2499011314401839e-05, "loss": 1.4421, "step": 16590 }, { "epoch": 0.43726930586790386, "grad_norm": 28.495189666748047, "learning_rate": 1.2490749497815362e-05, "loss": 1.4728, "step": 16600 }, { "epoch": 0.4375327211124026, "grad_norm": 7.732114315032959, "learning_rate": 1.2482485868255476e-05, "loss": 1.3512, "step": 16610 }, { "epoch": 0.4377961363569013, "grad_norm": 8.742936134338379, "learning_rate": 1.2474220431737135e-05, "loss": 1.2919, "step": 16620 }, { "epoch": 0.43805955160140003, "grad_norm": 10.803218841552734, "learning_rate": 1.24659531942766e-05, "loss": 1.2377, "step": 16630 }, { "epoch": 0.4383229668458988, "grad_norm": 7.6756744384765625, "learning_rate": 1.2457684161891456e-05, "loss": 1.3961, "step": 16640 }, { "epoch": 0.4385863820903975, "grad_norm": 5.053336143493652, "learning_rate": 1.2449413340600586e-05, "loss": 1.3993, "step": 16650 }, { "epoch": 0.43884979733489626, "grad_norm": 15.580668449401855, "learning_rate": 1.2441140736424185e-05, "loss": 1.4214, "step": 16660 }, { "epoch": 0.439113212579395, "grad_norm": 10.165223121643066, "learning_rate": 1.2432866355383723e-05, "loss": 1.3514, "step": 16670 }, { "epoch": 0.43937662782389375, "grad_norm": 11.372960090637207, "learning_rate": 1.2424590203501992e-05, "loss": 1.3138, "step": 16680 }, { "epoch": 0.4396400430683925, "grad_norm": 9.802559852600098, "learning_rate": 1.2416312286803057e-05, "loss": 1.3031, "step": 16690 }, { "epoch": 0.43990345831289124, "grad_norm": 17.825084686279297, "learning_rate": 1.2408032611312267e-05, "loss": 1.2657, "step": 16700 }, { "epoch": 0.4401668735573899, "grad_norm": 12.653438568115234, "learning_rate": 1.239975118305625e-05, "loss": 1.3587, "step": 16710 }, { "epoch": 0.44043028880188867, "grad_norm": 100.38288879394531, "learning_rate": 1.2391468008062926e-05, "loss": 1.4649, "step": 16720 }, { "epoch": 0.4406937040463874, "grad_norm": 8.964167594909668, "learning_rate": 1.2383183092361466e-05, "loss": 1.3451, "step": 16730 }, { "epoch": 0.44095711929088616, "grad_norm": 9.696819305419922, "learning_rate": 1.2374896441982319e-05, "loss": 1.4573, "step": 16740 }, { "epoch": 0.4412205345353849, "grad_norm": 5.018054962158203, "learning_rate": 1.2366608062957196e-05, "loss": 1.3852, "step": 16750 }, { "epoch": 0.44148394977988364, "grad_norm": 12.65772819519043, "learning_rate": 1.2358317961319059e-05, "loss": 1.375, "step": 16760 }, { "epoch": 0.4417473650243824, "grad_norm": 139.33087158203125, "learning_rate": 1.2350026143102134e-05, "loss": 1.4334, "step": 16770 }, { "epoch": 0.44201078026888113, "grad_norm": 17.72840118408203, "learning_rate": 1.2341732614341892e-05, "loss": 1.4002, "step": 16780 }, { "epoch": 0.4422741955133799, "grad_norm": 7.262829303741455, "learning_rate": 1.2333437381075047e-05, "loss": 1.3396, "step": 16790 }, { "epoch": 0.44253761075787856, "grad_norm": 12.905970573425293, "learning_rate": 1.2325140449339558e-05, "loss": 1.412, "step": 16800 }, { "epoch": 0.4428010260023773, "grad_norm": 5.6244635581970215, "learning_rate": 1.2316841825174614e-05, "loss": 1.3665, "step": 16810 }, { "epoch": 0.44306444124687605, "grad_norm": 10.447578430175781, "learning_rate": 1.2308541514620645e-05, "loss": 1.363, "step": 16820 }, { "epoch": 0.4433278564913748, "grad_norm": 8.445951461791992, "learning_rate": 1.23002395237193e-05, "loss": 1.4304, "step": 16830 }, { "epoch": 0.44359127173587354, "grad_norm": 8.335794448852539, "learning_rate": 1.2291935858513451e-05, "loss": 1.2957, "step": 16840 }, { "epoch": 0.4438546869803723, "grad_norm": 6.921783447265625, "learning_rate": 1.2283630525047197e-05, "loss": 1.2718, "step": 16850 }, { "epoch": 0.444118102224871, "grad_norm": 11.129327774047852, "learning_rate": 1.2275323529365846e-05, "loss": 1.4568, "step": 16860 }, { "epoch": 0.44438151746936977, "grad_norm": 24.843137741088867, "learning_rate": 1.2267014877515918e-05, "loss": 1.3401, "step": 16870 }, { "epoch": 0.44464493271386846, "grad_norm": 10.04791259765625, "learning_rate": 1.2258704575545126e-05, "loss": 1.3668, "step": 16880 }, { "epoch": 0.4449083479583672, "grad_norm": 68.27047729492188, "learning_rate": 1.225039262950241e-05, "loss": 1.3379, "step": 16890 }, { "epoch": 0.44517176320286594, "grad_norm": 5.000762462615967, "learning_rate": 1.2242079045437883e-05, "loss": 1.3539, "step": 16900 }, { "epoch": 0.4454351784473647, "grad_norm": 29.6456241607666, "learning_rate": 1.2233763829402856e-05, "loss": 1.438, "step": 16910 }, { "epoch": 0.44569859369186343, "grad_norm": 7.272820949554443, "learning_rate": 1.2225446987449838e-05, "loss": 1.2982, "step": 16920 }, { "epoch": 0.4459620089363622, "grad_norm": 27.656280517578125, "learning_rate": 1.2217128525632513e-05, "loss": 1.2816, "step": 16930 }, { "epoch": 0.4462254241808609, "grad_norm": 13.898427963256836, "learning_rate": 1.2208808450005743e-05, "loss": 1.301, "step": 16940 }, { "epoch": 0.44648883942535966, "grad_norm": 21.38923454284668, "learning_rate": 1.2200486766625566e-05, "loss": 1.3609, "step": 16950 }, { "epoch": 0.4467522546698584, "grad_norm": 6.022297382354736, "learning_rate": 1.2192163481549192e-05, "loss": 1.3708, "step": 16960 }, { "epoch": 0.4470156699143571, "grad_norm": 3.8119192123413086, "learning_rate": 1.2183838600834997e-05, "loss": 1.292, "step": 16970 }, { "epoch": 0.44727908515885584, "grad_norm": 6.753296852111816, "learning_rate": 1.2175512130542516e-05, "loss": 1.3145, "step": 16980 }, { "epoch": 0.4475425004033546, "grad_norm": 10.175582885742188, "learning_rate": 1.2167184076732444e-05, "loss": 1.345, "step": 16990 }, { "epoch": 0.4478059156478533, "grad_norm": 5.681181907653809, "learning_rate": 1.2158854445466623e-05, "loss": 1.2474, "step": 17000 }, { "epoch": 0.44806933089235207, "grad_norm": 10.444528579711914, "learning_rate": 1.2150523242808057e-05, "loss": 1.2891, "step": 17010 }, { "epoch": 0.4483327461368508, "grad_norm": 8.15530776977539, "learning_rate": 1.2142190474820873e-05, "loss": 1.388, "step": 17020 }, { "epoch": 0.44859616138134956, "grad_norm": 9.996291160583496, "learning_rate": 1.2133856147570358e-05, "loss": 1.4086, "step": 17030 }, { "epoch": 0.4488595766258483, "grad_norm": 5.583968639373779, "learning_rate": 1.2125520267122915e-05, "loss": 1.3998, "step": 17040 }, { "epoch": 0.44912299187034704, "grad_norm": 9.561891555786133, "learning_rate": 1.2117182839546091e-05, "loss": 1.4876, "step": 17050 }, { "epoch": 0.44938640711484573, "grad_norm": 22.232030868530273, "learning_rate": 1.2108843870908559e-05, "loss": 1.388, "step": 17060 }, { "epoch": 0.4496498223593445, "grad_norm": 7.613271713256836, "learning_rate": 1.2100503367280108e-05, "loss": 1.2969, "step": 17070 }, { "epoch": 0.4499132376038432, "grad_norm": 16.274017333984375, "learning_rate": 1.2092161334731643e-05, "loss": 1.3251, "step": 17080 }, { "epoch": 0.45017665284834196, "grad_norm": 22.479900360107422, "learning_rate": 1.2083817779335185e-05, "loss": 1.4244, "step": 17090 }, { "epoch": 0.4504400680928407, "grad_norm": 9.764890670776367, "learning_rate": 1.2075472707163867e-05, "loss": 1.3025, "step": 17100 }, { "epoch": 0.45070348333733945, "grad_norm": 21.82045555114746, "learning_rate": 1.2067126124291918e-05, "loss": 1.4526, "step": 17110 }, { "epoch": 0.4509668985818382, "grad_norm": 8.007939338684082, "learning_rate": 1.2058778036794676e-05, "loss": 1.3843, "step": 17120 }, { "epoch": 0.45123031382633694, "grad_norm": 18.004653930664062, "learning_rate": 1.2050428450748567e-05, "loss": 1.332, "step": 17130 }, { "epoch": 0.4514937290708356, "grad_norm": 3.740792751312256, "learning_rate": 1.2042077372231109e-05, "loss": 1.2857, "step": 17140 }, { "epoch": 0.45175714431533437, "grad_norm": 5.3372321128845215, "learning_rate": 1.203372480732091e-05, "loss": 1.3336, "step": 17150 }, { "epoch": 0.4520205595598331, "grad_norm": 14.75044059753418, "learning_rate": 1.2025370762097652e-05, "loss": 1.2689, "step": 17160 }, { "epoch": 0.45228397480433186, "grad_norm": 8.255457878112793, "learning_rate": 1.2017015242642106e-05, "loss": 1.2972, "step": 17170 }, { "epoch": 0.4525473900488306, "grad_norm": 5.61617374420166, "learning_rate": 1.2008658255036107e-05, "loss": 1.2964, "step": 17180 }, { "epoch": 0.45281080529332934, "grad_norm": 16.12565803527832, "learning_rate": 1.2000299805362562e-05, "loss": 1.2318, "step": 17190 }, { "epoch": 0.4530742205378281, "grad_norm": 10.079707145690918, "learning_rate": 1.199193989970544e-05, "loss": 1.3524, "step": 17200 }, { "epoch": 0.45333763578232683, "grad_norm": 7.114057540893555, "learning_rate": 1.1983578544149776e-05, "loss": 1.3439, "step": 17210 }, { "epoch": 0.4536010510268256, "grad_norm": 9.709219932556152, "learning_rate": 1.197521574478165e-05, "loss": 1.2414, "step": 17220 }, { "epoch": 0.45386446627132426, "grad_norm": 7.369729995727539, "learning_rate": 1.1966851507688203e-05, "loss": 1.2717, "step": 17230 }, { "epoch": 0.454127881515823, "grad_norm": 12.259556770324707, "learning_rate": 1.1958485838957618e-05, "loss": 1.4083, "step": 17240 }, { "epoch": 0.45439129676032175, "grad_norm": 8.01494026184082, "learning_rate": 1.1950118744679118e-05, "loss": 1.3734, "step": 17250 }, { "epoch": 0.4546547120048205, "grad_norm": 17.103248596191406, "learning_rate": 1.1941750230942963e-05, "loss": 1.3972, "step": 17260 }, { "epoch": 0.45491812724931924, "grad_norm": 72.14289855957031, "learning_rate": 1.1933380303840461e-05, "loss": 1.4064, "step": 17270 }, { "epoch": 0.455181542493818, "grad_norm": 4.074512004852295, "learning_rate": 1.1925008969463925e-05, "loss": 1.3733, "step": 17280 }, { "epoch": 0.4554449577383167, "grad_norm": 11.39741325378418, "learning_rate": 1.1916636233906709e-05, "loss": 1.4913, "step": 17290 }, { "epoch": 0.45570837298281547, "grad_norm": 12.816291809082031, "learning_rate": 1.1908262103263184e-05, "loss": 1.496, "step": 17300 }, { "epoch": 0.4559717882273142, "grad_norm": 11.657978057861328, "learning_rate": 1.1899886583628731e-05, "loss": 1.3381, "step": 17310 }, { "epoch": 0.4562352034718129, "grad_norm": 9.631767272949219, "learning_rate": 1.1891509681099744e-05, "loss": 1.377, "step": 17320 }, { "epoch": 0.45649861871631164, "grad_norm": 4.321954250335693, "learning_rate": 1.188313140177363e-05, "loss": 1.292, "step": 17330 }, { "epoch": 0.4567620339608104, "grad_norm": 5.757198810577393, "learning_rate": 1.1874751751748794e-05, "loss": 1.2805, "step": 17340 }, { "epoch": 0.45702544920530913, "grad_norm": 12.165517807006836, "learning_rate": 1.1866370737124637e-05, "loss": 1.3795, "step": 17350 }, { "epoch": 0.4572888644498079, "grad_norm": 2.6982247829437256, "learning_rate": 1.185798836400155e-05, "loss": 1.4036, "step": 17360 }, { "epoch": 0.4575522796943066, "grad_norm": 10.534849166870117, "learning_rate": 1.184960463848092e-05, "loss": 1.3653, "step": 17370 }, { "epoch": 0.45781569493880536, "grad_norm": 9.679082870483398, "learning_rate": 1.184121956666512e-05, "loss": 1.284, "step": 17380 }, { "epoch": 0.4580791101833041, "grad_norm": 9.43470573425293, "learning_rate": 1.1832833154657488e-05, "loss": 1.3792, "step": 17390 }, { "epoch": 0.4583425254278028, "grad_norm": 5.2622551918029785, "learning_rate": 1.1824445408562357e-05, "loss": 1.3408, "step": 17400 }, { "epoch": 0.45860594067230154, "grad_norm": 7.97438383102417, "learning_rate": 1.181605633448502e-05, "loss": 1.3875, "step": 17410 }, { "epoch": 0.4588693559168003, "grad_norm": 3.620469331741333, "learning_rate": 1.180766593853174e-05, "loss": 1.3575, "step": 17420 }, { "epoch": 0.459132771161299, "grad_norm": 20.46158790588379, "learning_rate": 1.1799274226809736e-05, "loss": 1.348, "step": 17430 }, { "epoch": 0.45939618640579777, "grad_norm": 10.182454109191895, "learning_rate": 1.1790881205427193e-05, "loss": 1.3412, "step": 17440 }, { "epoch": 0.4596596016502965, "grad_norm": 17.81354331970215, "learning_rate": 1.1782486880493249e-05, "loss": 1.3644, "step": 17450 }, { "epoch": 0.45992301689479526, "grad_norm": 9.221073150634766, "learning_rate": 1.177409125811798e-05, "loss": 1.3352, "step": 17460 }, { "epoch": 0.460186432139294, "grad_norm": 6.749099254608154, "learning_rate": 1.1765694344412417e-05, "loss": 1.4912, "step": 17470 }, { "epoch": 0.46044984738379274, "grad_norm": 7.90328311920166, "learning_rate": 1.1757296145488535e-05, "loss": 1.4312, "step": 17480 }, { "epoch": 0.46071326262829143, "grad_norm": 6.748524188995361, "learning_rate": 1.174889666745923e-05, "loss": 1.2442, "step": 17490 }, { "epoch": 0.4609766778727902, "grad_norm": 17.53350067138672, "learning_rate": 1.1740495916438334e-05, "loss": 1.3163, "step": 17500 }, { "epoch": 0.4612400931172889, "grad_norm": 17.300928115844727, "learning_rate": 1.1732093898540616e-05, "loss": 1.3427, "step": 17510 }, { "epoch": 0.46150350836178766, "grad_norm": 5.9709062576293945, "learning_rate": 1.1723690619881754e-05, "loss": 1.3124, "step": 17520 }, { "epoch": 0.4617669236062864, "grad_norm": 6.436551570892334, "learning_rate": 1.1715286086578352e-05, "loss": 1.4312, "step": 17530 }, { "epoch": 0.46203033885078515, "grad_norm": 5.119504928588867, "learning_rate": 1.1706880304747923e-05, "loss": 1.3081, "step": 17540 }, { "epoch": 0.4622937540952839, "grad_norm": 9.107444763183594, "learning_rate": 1.1698473280508888e-05, "loss": 1.29, "step": 17550 }, { "epoch": 0.46255716933978264, "grad_norm": 7.0895094871521, "learning_rate": 1.1690065019980582e-05, "loss": 1.3612, "step": 17560 }, { "epoch": 0.4628205845842814, "grad_norm": 5.812620639801025, "learning_rate": 1.1681655529283218e-05, "loss": 1.361, "step": 17570 }, { "epoch": 0.46308399982878007, "grad_norm": 15.018234252929688, "learning_rate": 1.1673244814537933e-05, "loss": 1.3411, "step": 17580 }, { "epoch": 0.4633474150732788, "grad_norm": 15.493953704833984, "learning_rate": 1.1664832881866731e-05, "loss": 1.4037, "step": 17590 }, { "epoch": 0.46361083031777756, "grad_norm": 8.683887481689453, "learning_rate": 1.1656419737392514e-05, "loss": 1.2953, "step": 17600 }, { "epoch": 0.4638742455622763, "grad_norm": 4.542469501495361, "learning_rate": 1.1648005387239072e-05, "loss": 1.3635, "step": 17610 }, { "epoch": 0.46413766080677504, "grad_norm": 11.061468124389648, "learning_rate": 1.1639589837531056e-05, "loss": 1.4093, "step": 17620 }, { "epoch": 0.4644010760512738, "grad_norm": 13.695755958557129, "learning_rate": 1.1631173094394e-05, "loss": 1.2486, "step": 17630 }, { "epoch": 0.46466449129577253, "grad_norm": 7.886596202850342, "learning_rate": 1.1622755163954306e-05, "loss": 1.3665, "step": 17640 }, { "epoch": 0.4649279065402713, "grad_norm": 8.5760498046875, "learning_rate": 1.1614336052339242e-05, "loss": 1.4029, "step": 17650 }, { "epoch": 0.46519132178476996, "grad_norm": 7.391860008239746, "learning_rate": 1.1605915765676931e-05, "loss": 1.4252, "step": 17660 }, { "epoch": 0.4654547370292687, "grad_norm": 10.017147064208984, "learning_rate": 1.1597494310096354e-05, "loss": 1.3425, "step": 17670 }, { "epoch": 0.46571815227376745, "grad_norm": 9.948430061340332, "learning_rate": 1.1589071691727345e-05, "loss": 1.3727, "step": 17680 }, { "epoch": 0.4659815675182662, "grad_norm": 8.862388610839844, "learning_rate": 1.158064791670058e-05, "loss": 1.3159, "step": 17690 }, { "epoch": 0.46624498276276494, "grad_norm": 17.487131118774414, "learning_rate": 1.157222299114758e-05, "loss": 1.3681, "step": 17700 }, { "epoch": 0.4665083980072637, "grad_norm": 7.854982852935791, "learning_rate": 1.15637969212007e-05, "loss": 1.3573, "step": 17710 }, { "epoch": 0.4667718132517624, "grad_norm": 11.474493026733398, "learning_rate": 1.1555369712993132e-05, "loss": 1.2794, "step": 17720 }, { "epoch": 0.46703522849626117, "grad_norm": 15.570900917053223, "learning_rate": 1.1546941372658896e-05, "loss": 1.3269, "step": 17730 }, { "epoch": 0.4672986437407599, "grad_norm": 4.000212669372559, "learning_rate": 1.1538511906332834e-05, "loss": 1.3744, "step": 17740 }, { "epoch": 0.4675620589852586, "grad_norm": 2.809767246246338, "learning_rate": 1.1530081320150605e-05, "loss": 1.294, "step": 17750 }, { "epoch": 0.46782547422975734, "grad_norm": 10.884756088256836, "learning_rate": 1.1521649620248694e-05, "loss": 1.3266, "step": 17760 }, { "epoch": 0.4680888894742561, "grad_norm": 5.4316725730896, "learning_rate": 1.1513216812764383e-05, "loss": 1.2521, "step": 17770 }, { "epoch": 0.46835230471875483, "grad_norm": 10.219120979309082, "learning_rate": 1.1504782903835769e-05, "loss": 1.3743, "step": 17780 }, { "epoch": 0.4686157199632536, "grad_norm": 19.54829978942871, "learning_rate": 1.1496347899601752e-05, "loss": 1.3234, "step": 17790 }, { "epoch": 0.4688791352077523, "grad_norm": 13.243267059326172, "learning_rate": 1.1487911806202016e-05, "loss": 1.4258, "step": 17800 }, { "epoch": 0.46914255045225106, "grad_norm": 5.31924295425415, "learning_rate": 1.1479474629777051e-05, "loss": 1.3668, "step": 17810 }, { "epoch": 0.4694059656967498, "grad_norm": 10.356910705566406, "learning_rate": 1.1471036376468138e-05, "loss": 1.4023, "step": 17820 }, { "epoch": 0.46966938094124855, "grad_norm": 9.739974975585938, "learning_rate": 1.146259705241733e-05, "loss": 1.4303, "step": 17830 }, { "epoch": 0.46993279618574724, "grad_norm": 13.037135124206543, "learning_rate": 1.1454156663767465e-05, "loss": 1.3083, "step": 17840 }, { "epoch": 0.470196211430246, "grad_norm": 10.98000717163086, "learning_rate": 1.1445715216662155e-05, "loss": 1.3051, "step": 17850 }, { "epoch": 0.4704596266747447, "grad_norm": 2.5696017742156982, "learning_rate": 1.1437272717245782e-05, "loss": 1.3033, "step": 17860 }, { "epoch": 0.47072304191924347, "grad_norm": 7.220703125, "learning_rate": 1.1428829171663501e-05, "loss": 1.3264, "step": 17870 }, { "epoch": 0.4709864571637422, "grad_norm": 8.379532814025879, "learning_rate": 1.1420384586061215e-05, "loss": 1.336, "step": 17880 }, { "epoch": 0.47124987240824096, "grad_norm": 6.1946306228637695, "learning_rate": 1.1411938966585599e-05, "loss": 1.422, "step": 17890 }, { "epoch": 0.4715132876527397, "grad_norm": 3.783933401107788, "learning_rate": 1.140349231938407e-05, "loss": 1.3865, "step": 17900 }, { "epoch": 0.47177670289723844, "grad_norm": 5.940703392028809, "learning_rate": 1.139504465060479e-05, "loss": 1.2952, "step": 17910 }, { "epoch": 0.47204011814173713, "grad_norm": 11.068153381347656, "learning_rate": 1.1386595966396681e-05, "loss": 1.3672, "step": 17920 }, { "epoch": 0.4723035333862359, "grad_norm": 14.5176362991333, "learning_rate": 1.1378146272909387e-05, "loss": 1.3043, "step": 17930 }, { "epoch": 0.4725669486307346, "grad_norm": 13.596080780029297, "learning_rate": 1.1369695576293298e-05, "loss": 1.2776, "step": 17940 }, { "epoch": 0.47283036387523336, "grad_norm": 11.712677955627441, "learning_rate": 1.1361243882699524e-05, "loss": 1.4124, "step": 17950 }, { "epoch": 0.4730937791197321, "grad_norm": 5.001216888427734, "learning_rate": 1.1352791198279913e-05, "loss": 1.191, "step": 17960 }, { "epoch": 0.47335719436423085, "grad_norm": 3.1461751461029053, "learning_rate": 1.1344337529187022e-05, "loss": 1.3725, "step": 17970 }, { "epoch": 0.4736206096087296, "grad_norm": 13.747640609741211, "learning_rate": 1.1335882881574134e-05, "loss": 1.3915, "step": 17980 }, { "epoch": 0.47388402485322834, "grad_norm": 5.127624988555908, "learning_rate": 1.1327427261595238e-05, "loss": 1.3045, "step": 17990 }, { "epoch": 0.4741474400977271, "grad_norm": 9.164541244506836, "learning_rate": 1.1318970675405037e-05, "loss": 1.3864, "step": 18000 }, { "epoch": 0.47441085534222577, "grad_norm": 18.0811767578125, "learning_rate": 1.1310513129158926e-05, "loss": 1.4237, "step": 18010 }, { "epoch": 0.4746742705867245, "grad_norm": 11.857624053955078, "learning_rate": 1.1302054629013015e-05, "loss": 1.299, "step": 18020 }, { "epoch": 0.47493768583122326, "grad_norm": 9.494019508361816, "learning_rate": 1.1293595181124097e-05, "loss": 1.3542, "step": 18030 }, { "epoch": 0.475201101075722, "grad_norm": 3.08787202835083, "learning_rate": 1.1285134791649655e-05, "loss": 1.3622, "step": 18040 }, { "epoch": 0.47546451632022074, "grad_norm": 8.90296745300293, "learning_rate": 1.127667346674786e-05, "loss": 1.4552, "step": 18050 }, { "epoch": 0.4757279315647195, "grad_norm": 9.801215171813965, "learning_rate": 1.1268211212577567e-05, "loss": 1.3724, "step": 18060 }, { "epoch": 0.47599134680921823, "grad_norm": 9.670479774475098, "learning_rate": 1.1259748035298304e-05, "loss": 1.2491, "step": 18070 }, { "epoch": 0.476254762053717, "grad_norm": 13.312700271606445, "learning_rate": 1.1251283941070268e-05, "loss": 1.3442, "step": 18080 }, { "epoch": 0.4765181772982157, "grad_norm": 18.127330780029297, "learning_rate": 1.1242818936054327e-05, "loss": 1.3298, "step": 18090 }, { "epoch": 0.4767815925427144, "grad_norm": 7.370855808258057, "learning_rate": 1.1234353026412015e-05, "loss": 1.3974, "step": 18100 }, { "epoch": 0.47704500778721315, "grad_norm": 13.715600967407227, "learning_rate": 1.1225886218305517e-05, "loss": 1.2922, "step": 18110 }, { "epoch": 0.4773084230317119, "grad_norm": 4.454719066619873, "learning_rate": 1.121741851789768e-05, "loss": 1.3602, "step": 18120 }, { "epoch": 0.47757183827621064, "grad_norm": 5.337070465087891, "learning_rate": 1.1208949931351995e-05, "loss": 1.3888, "step": 18130 }, { "epoch": 0.4778352535207094, "grad_norm": 18.37483024597168, "learning_rate": 1.1200480464832597e-05, "loss": 1.3646, "step": 18140 }, { "epoch": 0.4780986687652081, "grad_norm": 9.429910659790039, "learning_rate": 1.1192010124504263e-05, "loss": 1.3423, "step": 18150 }, { "epoch": 0.47836208400970687, "grad_norm": 6.537646770477295, "learning_rate": 1.1183538916532413e-05, "loss": 1.3409, "step": 18160 }, { "epoch": 0.4786254992542056, "grad_norm": 8.128035545349121, "learning_rate": 1.1175066847083088e-05, "loss": 1.3802, "step": 18170 }, { "epoch": 0.4788889144987043, "grad_norm": 8.956013679504395, "learning_rate": 1.1166593922322962e-05, "loss": 1.2532, "step": 18180 }, { "epoch": 0.47915232974320304, "grad_norm": 11.01744270324707, "learning_rate": 1.115812014841933e-05, "loss": 1.3569, "step": 18190 }, { "epoch": 0.4794157449877018, "grad_norm": 22.02098274230957, "learning_rate": 1.1149645531540104e-05, "loss": 1.4022, "step": 18200 }, { "epoch": 0.47967916023220053, "grad_norm": 8.176229476928711, "learning_rate": 1.1141170077853814e-05, "loss": 1.3723, "step": 18210 }, { "epoch": 0.4799425754766993, "grad_norm": 6.5172247886657715, "learning_rate": 1.1132693793529588e-05, "loss": 1.4838, "step": 18220 }, { "epoch": 0.480205990721198, "grad_norm": 2.8326313495635986, "learning_rate": 1.1124216684737177e-05, "loss": 1.3874, "step": 18230 }, { "epoch": 0.48046940596569676, "grad_norm": 4.887580394744873, "learning_rate": 1.1115738757646917e-05, "loss": 1.3754, "step": 18240 }, { "epoch": 0.4807328212101955, "grad_norm": 7.576440811157227, "learning_rate": 1.1107260018429741e-05, "loss": 1.2877, "step": 18250 }, { "epoch": 0.48099623645469425, "grad_norm": 5.947535991668701, "learning_rate": 1.1098780473257184e-05, "loss": 1.3264, "step": 18260 }, { "epoch": 0.48125965169919294, "grad_norm": 15.813035011291504, "learning_rate": 1.1090300128301352e-05, "loss": 1.3794, "step": 18270 }, { "epoch": 0.4815230669436917, "grad_norm": 3.2813379764556885, "learning_rate": 1.1081818989734946e-05, "loss": 1.3281, "step": 18280 }, { "epoch": 0.4817864821881904, "grad_norm": 11.992558479309082, "learning_rate": 1.1073337063731239e-05, "loss": 1.3367, "step": 18290 }, { "epoch": 0.48204989743268917, "grad_norm": 6.668792247772217, "learning_rate": 1.1064854356464079e-05, "loss": 1.272, "step": 18300 }, { "epoch": 0.4823133126771879, "grad_norm": 5.277916431427002, "learning_rate": 1.1056370874107879e-05, "loss": 1.2833, "step": 18310 }, { "epoch": 0.48257672792168665, "grad_norm": 4.3763909339904785, "learning_rate": 1.1047886622837619e-05, "loss": 1.2934, "step": 18320 }, { "epoch": 0.4828401431661854, "grad_norm": 3.4746158123016357, "learning_rate": 1.103940160882884e-05, "loss": 1.3728, "step": 18330 }, { "epoch": 0.48310355841068414, "grad_norm": 5.771236896514893, "learning_rate": 1.103091583825764e-05, "loss": 1.4169, "step": 18340 }, { "epoch": 0.4833669736551829, "grad_norm": 19.925865173339844, "learning_rate": 1.1022429317300655e-05, "loss": 1.3383, "step": 18350 }, { "epoch": 0.4836303888996816, "grad_norm": 5.696928024291992, "learning_rate": 1.1013942052135082e-05, "loss": 1.3163, "step": 18360 }, { "epoch": 0.4838938041441803, "grad_norm": 5.728933334350586, "learning_rate": 1.1005454048938659e-05, "loss": 1.2002, "step": 18370 }, { "epoch": 0.48415721938867906, "grad_norm": 8.316256523132324, "learning_rate": 1.0996965313889652e-05, "loss": 1.2748, "step": 18380 }, { "epoch": 0.4844206346331778, "grad_norm": 6.635368824005127, "learning_rate": 1.098847585316686e-05, "loss": 1.3693, "step": 18390 }, { "epoch": 0.48468404987767655, "grad_norm": 10.656562805175781, "learning_rate": 1.0979985672949622e-05, "loss": 1.4099, "step": 18400 }, { "epoch": 0.4849474651221753, "grad_norm": 6.002817630767822, "learning_rate": 1.0971494779417789e-05, "loss": 1.3079, "step": 18410 }, { "epoch": 0.48521088036667404, "grad_norm": 8.822344779968262, "learning_rate": 1.0963003178751736e-05, "loss": 1.3305, "step": 18420 }, { "epoch": 0.4854742956111728, "grad_norm": 5.804863452911377, "learning_rate": 1.0954510877132352e-05, "loss": 1.3176, "step": 18430 }, { "epoch": 0.48573771085567147, "grad_norm": 18.32676124572754, "learning_rate": 1.0946017880741035e-05, "loss": 1.1872, "step": 18440 }, { "epoch": 0.4860011261001702, "grad_norm": 5.125613689422607, "learning_rate": 1.0937524195759696e-05, "loss": 1.4034, "step": 18450 }, { "epoch": 0.48626454134466895, "grad_norm": 6.892393112182617, "learning_rate": 1.0929029828370732e-05, "loss": 1.2514, "step": 18460 }, { "epoch": 0.4865279565891677, "grad_norm": 8.154943466186523, "learning_rate": 1.0920534784757053e-05, "loss": 1.3142, "step": 18470 }, { "epoch": 0.48679137183366644, "grad_norm": 14.054874420166016, "learning_rate": 1.091203907110205e-05, "loss": 1.3768, "step": 18480 }, { "epoch": 0.4870547870781652, "grad_norm": 7.000333786010742, "learning_rate": 1.0903542693589607e-05, "loss": 1.4095, "step": 18490 }, { "epoch": 0.48731820232266393, "grad_norm": 10.079047203063965, "learning_rate": 1.0895045658404088e-05, "loss": 1.4148, "step": 18500 }, { "epoch": 0.4875816175671627, "grad_norm": 8.294957160949707, "learning_rate": 1.0886547971730341e-05, "loss": 1.2723, "step": 18510 }, { "epoch": 0.4878450328116614, "grad_norm": 5.005089282989502, "learning_rate": 1.0878049639753686e-05, "loss": 1.2843, "step": 18520 }, { "epoch": 0.4881084480561601, "grad_norm": 7.013195514678955, "learning_rate": 1.0869550668659908e-05, "loss": 1.3356, "step": 18530 }, { "epoch": 0.48837186330065885, "grad_norm": 8.040790557861328, "learning_rate": 1.0861051064635262e-05, "loss": 1.3048, "step": 18540 }, { "epoch": 0.4886352785451576, "grad_norm": 8.393134117126465, "learning_rate": 1.0852550833866464e-05, "loss": 1.3405, "step": 18550 }, { "epoch": 0.48889869378965634, "grad_norm": 15.457324028015137, "learning_rate": 1.084404998254068e-05, "loss": 1.3283, "step": 18560 }, { "epoch": 0.4891621090341551, "grad_norm": 6.33058500289917, "learning_rate": 1.083554851684554e-05, "loss": 1.3588, "step": 18570 }, { "epoch": 0.4894255242786538, "grad_norm": 3.9698903560638428, "learning_rate": 1.082704644296911e-05, "loss": 1.293, "step": 18580 }, { "epoch": 0.48968893952315257, "grad_norm": 13.630831718444824, "learning_rate": 1.0818543767099899e-05, "loss": 1.3361, "step": 18590 }, { "epoch": 0.4899523547676513, "grad_norm": 6.548192501068115, "learning_rate": 1.081004049542686e-05, "loss": 1.3256, "step": 18600 }, { "epoch": 0.49021577001215005, "grad_norm": 4.088810920715332, "learning_rate": 1.0801536634139378e-05, "loss": 1.3426, "step": 18610 }, { "epoch": 0.49047918525664874, "grad_norm": 15.684075355529785, "learning_rate": 1.0793032189427266e-05, "loss": 1.3614, "step": 18620 }, { "epoch": 0.4907426005011475, "grad_norm": 5.512744903564453, "learning_rate": 1.078452716748076e-05, "loss": 1.2587, "step": 18630 }, { "epoch": 0.49100601574564623, "grad_norm": 7.062557697296143, "learning_rate": 1.077602157449052e-05, "loss": 1.2922, "step": 18640 }, { "epoch": 0.491269430990145, "grad_norm": 6.211926460266113, "learning_rate": 1.0767515416647622e-05, "loss": 1.3275, "step": 18650 }, { "epoch": 0.4915328462346437, "grad_norm": 5.405052661895752, "learning_rate": 1.0759008700143548e-05, "loss": 1.3268, "step": 18660 }, { "epoch": 0.49179626147914246, "grad_norm": 7.603171348571777, "learning_rate": 1.0750501431170188e-05, "loss": 1.3938, "step": 18670 }, { "epoch": 0.4920596767236412, "grad_norm": 15.55841064453125, "learning_rate": 1.074199361591984e-05, "loss": 1.3844, "step": 18680 }, { "epoch": 0.49232309196813995, "grad_norm": 6.230996608734131, "learning_rate": 1.0733485260585194e-05, "loss": 1.3844, "step": 18690 }, { "epoch": 0.49258650721263864, "grad_norm": 9.104446411132812, "learning_rate": 1.072497637135933e-05, "loss": 1.315, "step": 18700 }, { "epoch": 0.4928499224571374, "grad_norm": 8.425771713256836, "learning_rate": 1.0716466954435729e-05, "loss": 1.2455, "step": 18710 }, { "epoch": 0.4931133377016361, "grad_norm": 11.723322868347168, "learning_rate": 1.070795701600824e-05, "loss": 1.294, "step": 18720 }, { "epoch": 0.49337675294613487, "grad_norm": 13.477933883666992, "learning_rate": 1.0699446562271104e-05, "loss": 1.2993, "step": 18730 }, { "epoch": 0.4936401681906336, "grad_norm": 12.618258476257324, "learning_rate": 1.0690935599418935e-05, "loss": 1.3694, "step": 18740 }, { "epoch": 0.49390358343513235, "grad_norm": 6.200601100921631, "learning_rate": 1.0682424133646712e-05, "loss": 1.2826, "step": 18750 }, { "epoch": 0.4941669986796311, "grad_norm": 360.7298278808594, "learning_rate": 1.0673912171149786e-05, "loss": 1.3296, "step": 18760 }, { "epoch": 0.49443041392412984, "grad_norm": 11.332207679748535, "learning_rate": 1.0665399718123859e-05, "loss": 1.3509, "step": 18770 }, { "epoch": 0.4946938291686286, "grad_norm": 11.473334312438965, "learning_rate": 1.0656886780765007e-05, "loss": 1.4382, "step": 18780 }, { "epoch": 0.4949572444131273, "grad_norm": 6.741174221038818, "learning_rate": 1.064837336526965e-05, "loss": 1.3778, "step": 18790 }, { "epoch": 0.495220659657626, "grad_norm": 4.8239665031433105, "learning_rate": 1.0639859477834551e-05, "loss": 1.3624, "step": 18800 }, { "epoch": 0.49548407490212476, "grad_norm": 9.023909568786621, "learning_rate": 1.0631345124656825e-05, "loss": 1.4217, "step": 18810 }, { "epoch": 0.4957474901466235, "grad_norm": 6.240799903869629, "learning_rate": 1.0622830311933918e-05, "loss": 1.3897, "step": 18820 }, { "epoch": 0.49601090539112225, "grad_norm": 26.3687744140625, "learning_rate": 1.0614315045863617e-05, "loss": 1.2823, "step": 18830 }, { "epoch": 0.496274320635621, "grad_norm": 7.929842948913574, "learning_rate": 1.0605799332644036e-05, "loss": 1.3288, "step": 18840 }, { "epoch": 0.49653773588011973, "grad_norm": 15.663352966308594, "learning_rate": 1.0597283178473614e-05, "loss": 1.2215, "step": 18850 }, { "epoch": 0.4968011511246185, "grad_norm": 11.104543685913086, "learning_rate": 1.0588766589551116e-05, "loss": 1.3198, "step": 18860 }, { "epoch": 0.4970645663691172, "grad_norm": 11.175535202026367, "learning_rate": 1.0580249572075615e-05, "loss": 1.3991, "step": 18870 }, { "epoch": 0.4973279816136159, "grad_norm": 4.537485599517822, "learning_rate": 1.0571732132246504e-05, "loss": 1.4064, "step": 18880 }, { "epoch": 0.49759139685811465, "grad_norm": 3.0803215503692627, "learning_rate": 1.0563214276263481e-05, "loss": 1.3479, "step": 18890 }, { "epoch": 0.4978548121026134, "grad_norm": 6.9684295654296875, "learning_rate": 1.055469601032654e-05, "loss": 1.3053, "step": 18900 }, { "epoch": 0.49811822734711214, "grad_norm": 7.465551376342773, "learning_rate": 1.0546177340635981e-05, "loss": 1.2095, "step": 18910 }, { "epoch": 0.4983816425916109, "grad_norm": 11.208521842956543, "learning_rate": 1.0537658273392404e-05, "loss": 1.3669, "step": 18920 }, { "epoch": 0.49864505783610963, "grad_norm": 12.768182754516602, "learning_rate": 1.052913881479668e-05, "loss": 1.3737, "step": 18930 }, { "epoch": 0.4989084730806084, "grad_norm": 16.34889030456543, "learning_rate": 1.0520618971049978e-05, "loss": 1.3527, "step": 18940 }, { "epoch": 0.4991718883251071, "grad_norm": 2.9364967346191406, "learning_rate": 1.0512098748353748e-05, "loss": 1.3147, "step": 18950 }, { "epoch": 0.49943530356960586, "grad_norm": 9.796316146850586, "learning_rate": 1.050357815290971e-05, "loss": 1.3696, "step": 18960 }, { "epoch": 0.49969871881410455, "grad_norm": 4.13497257232666, "learning_rate": 1.0495057190919859e-05, "loss": 1.3308, "step": 18970 }, { "epoch": 0.4999621340586033, "grad_norm": 4.553013801574707, "learning_rate": 1.0486535868586454e-05, "loss": 1.3738, "step": 18980 }, { "epoch": 0.500225549303102, "grad_norm": 7.501701831817627, "learning_rate": 1.0478014192112021e-05, "loss": 1.3732, "step": 18990 }, { "epoch": 0.5004889645476008, "grad_norm": 6.131885528564453, "learning_rate": 1.0469492167699337e-05, "loss": 1.2931, "step": 19000 }, { "epoch": 0.5007523797920995, "grad_norm": 13.974555969238281, "learning_rate": 1.0460969801551437e-05, "loss": 1.3346, "step": 19010 }, { "epoch": 0.5010157950365982, "grad_norm": 8.073199272155762, "learning_rate": 1.0452447099871605e-05, "loss": 1.3629, "step": 19020 }, { "epoch": 0.501279210281097, "grad_norm": 7.913654804229736, "learning_rate": 1.044392406886337e-05, "loss": 1.3221, "step": 19030 }, { "epoch": 0.5015426255255957, "grad_norm": 11.810351371765137, "learning_rate": 1.0435400714730489e-05, "loss": 1.3289, "step": 19040 }, { "epoch": 0.5018060407700945, "grad_norm": 7.955169677734375, "learning_rate": 1.0426877043676975e-05, "loss": 1.3315, "step": 19050 }, { "epoch": 0.5020694560145932, "grad_norm": 12.637120246887207, "learning_rate": 1.0418353061907056e-05, "loss": 1.3215, "step": 19060 }, { "epoch": 0.502332871259092, "grad_norm": 13.099584579467773, "learning_rate": 1.040982877562519e-05, "loss": 1.4934, "step": 19070 }, { "epoch": 0.5025962865035907, "grad_norm": 13.352666854858398, "learning_rate": 1.0401304191036056e-05, "loss": 1.3403, "step": 19080 }, { "epoch": 0.5028597017480895, "grad_norm": 7.168139457702637, "learning_rate": 1.0392779314344556e-05, "loss": 1.3111, "step": 19090 }, { "epoch": 0.5031231169925882, "grad_norm": 10.92310905456543, "learning_rate": 1.0384254151755799e-05, "loss": 1.2771, "step": 19100 }, { "epoch": 0.5033865322370868, "grad_norm": 3.1108086109161377, "learning_rate": 1.03757287094751e-05, "loss": 1.3583, "step": 19110 }, { "epoch": 0.5036499474815856, "grad_norm": 9.135017395019531, "learning_rate": 1.0367202993707981e-05, "loss": 1.2637, "step": 19120 }, { "epoch": 0.5039133627260843, "grad_norm": 23.910722732543945, "learning_rate": 1.0358677010660167e-05, "loss": 1.2925, "step": 19130 }, { "epoch": 0.5041767779705831, "grad_norm": 5.586996078491211, "learning_rate": 1.0350150766537568e-05, "loss": 1.3327, "step": 19140 }, { "epoch": 0.5044401932150818, "grad_norm": 17.66317367553711, "learning_rate": 1.0341624267546293e-05, "loss": 1.4213, "step": 19150 }, { "epoch": 0.5047036084595806, "grad_norm": 6.6890177726745605, "learning_rate": 1.0333097519892632e-05, "loss": 1.2873, "step": 19160 }, { "epoch": 0.5049670237040793, "grad_norm": 8.560919761657715, "learning_rate": 1.0324570529783056e-05, "loss": 1.226, "step": 19170 }, { "epoch": 0.505230438948578, "grad_norm": 8.851001739501953, "learning_rate": 1.0316043303424215e-05, "loss": 1.3746, "step": 19180 }, { "epoch": 0.5054938541930768, "grad_norm": 7.846219062805176, "learning_rate": 1.0307515847022929e-05, "loss": 1.4172, "step": 19190 }, { "epoch": 0.5057572694375755, "grad_norm": 8.423254013061523, "learning_rate": 1.0298988166786184e-05, "loss": 1.3775, "step": 19200 }, { "epoch": 0.5060206846820743, "grad_norm": 2.2972495555877686, "learning_rate": 1.0290460268921133e-05, "loss": 1.291, "step": 19210 }, { "epoch": 0.506284099926573, "grad_norm": 11.780354499816895, "learning_rate": 1.0281932159635083e-05, "loss": 1.3582, "step": 19220 }, { "epoch": 0.5065475151710718, "grad_norm": 3.0510873794555664, "learning_rate": 1.02734038451355e-05, "loss": 1.4108, "step": 19230 }, { "epoch": 0.5068109304155705, "grad_norm": 11.734567642211914, "learning_rate": 1.0264875331629993e-05, "loss": 1.3422, "step": 19240 }, { "epoch": 0.5070743456600693, "grad_norm": 11.266237258911133, "learning_rate": 1.0256346625326322e-05, "loss": 1.3607, "step": 19250 }, { "epoch": 0.507337760904568, "grad_norm": 7.599793910980225, "learning_rate": 1.0247817732432383e-05, "loss": 1.3699, "step": 19260 }, { "epoch": 0.5076011761490666, "grad_norm": 42.46408462524414, "learning_rate": 1.0239288659156207e-05, "loss": 1.213, "step": 19270 }, { "epoch": 0.5078645913935654, "grad_norm": 7.755181789398193, "learning_rate": 1.0230759411705966e-05, "loss": 1.337, "step": 19280 }, { "epoch": 0.5081280066380641, "grad_norm": 6.284817695617676, "learning_rate": 1.0222229996289944e-05, "loss": 1.2605, "step": 19290 }, { "epoch": 0.5083914218825629, "grad_norm": 6.445411205291748, "learning_rate": 1.0213700419116556e-05, "loss": 1.3846, "step": 19300 }, { "epoch": 0.5086548371270616, "grad_norm": 5.422335624694824, "learning_rate": 1.020517068639434e-05, "loss": 1.4485, "step": 19310 }, { "epoch": 0.5089182523715604, "grad_norm": 2.1947340965270996, "learning_rate": 1.019664080433193e-05, "loss": 1.3307, "step": 19320 }, { "epoch": 0.5091816676160591, "grad_norm": 15.02253246307373, "learning_rate": 1.018811077913809e-05, "loss": 1.3519, "step": 19330 }, { "epoch": 0.5094450828605579, "grad_norm": 4.987725257873535, "learning_rate": 1.017958061702167e-05, "loss": 1.2922, "step": 19340 }, { "epoch": 0.5097084981050566, "grad_norm": 10.358150482177734, "learning_rate": 1.017105032419163e-05, "loss": 1.3173, "step": 19350 }, { "epoch": 0.5099719133495553, "grad_norm": 7.917600154876709, "learning_rate": 1.0162519906857024e-05, "loss": 1.2817, "step": 19360 }, { "epoch": 0.5102353285940541, "grad_norm": 11.015604019165039, "learning_rate": 1.0153989371226991e-05, "loss": 1.3676, "step": 19370 }, { "epoch": 0.5104987438385528, "grad_norm": 7.387674331665039, "learning_rate": 1.0145458723510764e-05, "loss": 1.2738, "step": 19380 }, { "epoch": 0.5107621590830516, "grad_norm": 9.793679237365723, "learning_rate": 1.0136927969917651e-05, "loss": 1.3773, "step": 19390 }, { "epoch": 0.5110255743275502, "grad_norm": 8.170072555541992, "learning_rate": 1.0128397116657041e-05, "loss": 1.3782, "step": 19400 }, { "epoch": 0.511288989572049, "grad_norm": 11.991969108581543, "learning_rate": 1.0119866169938395e-05, "loss": 1.2558, "step": 19410 }, { "epoch": 0.5115524048165477, "grad_norm": 12.346480369567871, "learning_rate": 1.0111335135971238e-05, "loss": 1.242, "step": 19420 }, { "epoch": 0.5118158200610465, "grad_norm": 7.305009365081787, "learning_rate": 1.0102804020965167e-05, "loss": 1.2919, "step": 19430 }, { "epoch": 0.5120792353055452, "grad_norm": 13.351061820983887, "learning_rate": 1.009427283112983e-05, "loss": 1.2355, "step": 19440 }, { "epoch": 0.5123426505500439, "grad_norm": 11.127525329589844, "learning_rate": 1.008574157267493e-05, "loss": 1.4516, "step": 19450 }, { "epoch": 0.5126060657945427, "grad_norm": 12.154691696166992, "learning_rate": 1.0077210251810225e-05, "loss": 1.3915, "step": 19460 }, { "epoch": 0.5128694810390414, "grad_norm": 7.51495885848999, "learning_rate": 1.006867887474552e-05, "loss": 1.4071, "step": 19470 }, { "epoch": 0.5131328962835402, "grad_norm": 9.554969787597656, "learning_rate": 1.0060147447690647e-05, "loss": 1.3156, "step": 19480 }, { "epoch": 0.5133963115280389, "grad_norm": 18.020851135253906, "learning_rate": 1.0051615976855488e-05, "loss": 1.231, "step": 19490 }, { "epoch": 0.5136597267725377, "grad_norm": 5.085447788238525, "learning_rate": 1.0043084468449957e-05, "loss": 1.2869, "step": 19500 }, { "epoch": 0.5139231420170364, "grad_norm": 11.203155517578125, "learning_rate": 1.0034552928683987e-05, "loss": 1.331, "step": 19510 }, { "epoch": 0.5141865572615352, "grad_norm": 8.833269119262695, "learning_rate": 1.0026021363767538e-05, "loss": 1.3495, "step": 19520 }, { "epoch": 0.5144499725060339, "grad_norm": 8.810002326965332, "learning_rate": 1.0017489779910589e-05, "loss": 1.4133, "step": 19530 }, { "epoch": 0.5147133877505325, "grad_norm": 6.781354904174805, "learning_rate": 1.0008958183323134e-05, "loss": 1.3367, "step": 19540 }, { "epoch": 0.5149768029950313, "grad_norm": 5.674666881561279, "learning_rate": 1.0000426580215173e-05, "loss": 1.3715, "step": 19550 }, { "epoch": 0.51524021823953, "grad_norm": 11.790881156921387, "learning_rate": 9.99189497679671e-06, "loss": 1.3588, "step": 19560 }, { "epoch": 0.5155036334840288, "grad_norm": 4.221320152282715, "learning_rate": 9.983363379277753e-06, "loss": 1.3065, "step": 19570 }, { "epoch": 0.5157670487285275, "grad_norm": 19.112224578857422, "learning_rate": 9.974831793868302e-06, "loss": 1.4121, "step": 19580 }, { "epoch": 0.5160304639730263, "grad_norm": 6.683890342712402, "learning_rate": 9.966300226778352e-06, "loss": 1.4038, "step": 19590 }, { "epoch": 0.516293879217525, "grad_norm": 2.4628565311431885, "learning_rate": 9.957768684217884e-06, "loss": 1.3573, "step": 19600 }, { "epoch": 0.5165572944620238, "grad_norm": 5.177793025970459, "learning_rate": 9.949237172396866e-06, "loss": 1.4354, "step": 19610 }, { "epoch": 0.5168207097065225, "grad_norm": 4.745413780212402, "learning_rate": 9.940705697525223e-06, "loss": 1.3068, "step": 19620 }, { "epoch": 0.5170841249510212, "grad_norm": 9.126092910766602, "learning_rate": 9.932174265812876e-06, "loss": 1.39, "step": 19630 }, { "epoch": 0.51734754019552, "grad_norm": 6.925912380218506, "learning_rate": 9.923642883469709e-06, "loss": 1.255, "step": 19640 }, { "epoch": 0.5176109554400187, "grad_norm": 4.835154056549072, "learning_rate": 9.915111556705562e-06, "loss": 1.3306, "step": 19650 }, { "epoch": 0.5178743706845175, "grad_norm": 9.339128494262695, "learning_rate": 9.906580291730244e-06, "loss": 1.3293, "step": 19660 }, { "epoch": 0.5181377859290162, "grad_norm": 5.046908855438232, "learning_rate": 9.898049094753513e-06, "loss": 1.4013, "step": 19670 }, { "epoch": 0.518401201173515, "grad_norm": 5.9152655601501465, "learning_rate": 9.88951797198508e-06, "loss": 1.2559, "step": 19680 }, { "epoch": 0.5186646164180136, "grad_norm": 10.951078414916992, "learning_rate": 9.8809869296346e-06, "loss": 1.2487, "step": 19690 }, { "epoch": 0.5189280316625123, "grad_norm": 7.968634128570557, "learning_rate": 9.87245597391167e-06, "loss": 1.2949, "step": 19700 }, { "epoch": 0.5191914469070111, "grad_norm": 6.710391044616699, "learning_rate": 9.86392511102583e-06, "loss": 1.3985, "step": 19710 }, { "epoch": 0.5194548621515098, "grad_norm": 17.539810180664062, "learning_rate": 9.855394347186546e-06, "loss": 1.2565, "step": 19720 }, { "epoch": 0.5197182773960086, "grad_norm": 4.416016578674316, "learning_rate": 9.84686368860321e-06, "loss": 1.2962, "step": 19730 }, { "epoch": 0.5199816926405073, "grad_norm": 13.236982345581055, "learning_rate": 9.83833314148514e-06, "loss": 1.3806, "step": 19740 }, { "epoch": 0.5202451078850061, "grad_norm": 10.089315414428711, "learning_rate": 9.829802712041577e-06, "loss": 1.4437, "step": 19750 }, { "epoch": 0.5205085231295048, "grad_norm": 11.692655563354492, "learning_rate": 9.821272406481672e-06, "loss": 1.2536, "step": 19760 }, { "epoch": 0.5207719383740036, "grad_norm": 9.34780502319336, "learning_rate": 9.812742231014489e-06, "loss": 1.2328, "step": 19770 }, { "epoch": 0.5210353536185023, "grad_norm": 7.544678688049316, "learning_rate": 9.804212191848988e-06, "loss": 1.2737, "step": 19780 }, { "epoch": 0.521298768863001, "grad_norm": 4.627278804779053, "learning_rate": 9.79568229519404e-06, "loss": 1.2825, "step": 19790 }, { "epoch": 0.5215621841074998, "grad_norm": 6.48158597946167, "learning_rate": 9.787152547258411e-06, "loss": 1.3214, "step": 19800 }, { "epoch": 0.5218255993519985, "grad_norm": 7.222095489501953, "learning_rate": 9.778622954250756e-06, "loss": 1.2974, "step": 19810 }, { "epoch": 0.5220890145964973, "grad_norm": 6.295620918273926, "learning_rate": 9.77009352237962e-06, "loss": 1.2679, "step": 19820 }, { "epoch": 0.522352429840996, "grad_norm": 8.712936401367188, "learning_rate": 9.761564257853422e-06, "loss": 1.3917, "step": 19830 }, { "epoch": 0.5226158450854947, "grad_norm": 12.404751777648926, "learning_rate": 9.753035166880472e-06, "loss": 1.2978, "step": 19840 }, { "epoch": 0.5228792603299934, "grad_norm": 3.395648717880249, "learning_rate": 9.744506255668946e-06, "loss": 1.258, "step": 19850 }, { "epoch": 0.5231426755744922, "grad_norm": 4.312768936157227, "learning_rate": 9.735977530426889e-06, "loss": 1.3101, "step": 19860 }, { "epoch": 0.5234060908189909, "grad_norm": 5.7269439697265625, "learning_rate": 9.72744899736221e-06, "loss": 1.325, "step": 19870 }, { "epoch": 0.5236695060634896, "grad_norm": 10.166214942932129, "learning_rate": 9.718920662682695e-06, "loss": 1.3277, "step": 19880 }, { "epoch": 0.5239329213079884, "grad_norm": 4.362672805786133, "learning_rate": 9.710392532595954e-06, "loss": 1.3079, "step": 19890 }, { "epoch": 0.5241963365524871, "grad_norm": 10.407474517822266, "learning_rate": 9.701864613309468e-06, "loss": 1.3062, "step": 19900 }, { "epoch": 0.5244597517969859, "grad_norm": 10.98146915435791, "learning_rate": 9.693336911030568e-06, "loss": 1.3597, "step": 19910 }, { "epoch": 0.5247231670414846, "grad_norm": 7.723073482513428, "learning_rate": 9.684809431966419e-06, "loss": 1.4415, "step": 19920 }, { "epoch": 0.5249865822859834, "grad_norm": 7.867602348327637, "learning_rate": 9.676282182324025e-06, "loss": 1.3069, "step": 19930 }, { "epoch": 0.5252499975304821, "grad_norm": 7.802581787109375, "learning_rate": 9.667755168310222e-06, "loss": 1.3097, "step": 19940 }, { "epoch": 0.5255134127749809, "grad_norm": 5.681435585021973, "learning_rate": 9.659228396131675e-06, "loss": 1.4422, "step": 19950 }, { "epoch": 0.5257768280194796, "grad_norm": 4.71826696395874, "learning_rate": 9.650701871994874e-06, "loss": 1.2761, "step": 19960 }, { "epoch": 0.5260402432639782, "grad_norm": 24.492860794067383, "learning_rate": 9.64217560210613e-06, "loss": 1.2943, "step": 19970 }, { "epoch": 0.526303658508477, "grad_norm": 13.512646675109863, "learning_rate": 9.633649592671568e-06, "loss": 1.424, "step": 19980 }, { "epoch": 0.5265670737529757, "grad_norm": 8.41482162475586, "learning_rate": 9.625123849897121e-06, "loss": 1.4138, "step": 19990 }, { "epoch": 0.5268304889974745, "grad_norm": 5.459377765655518, "learning_rate": 9.616598379988526e-06, "loss": 1.2497, "step": 20000 }, { "epoch": 0.5270939042419732, "grad_norm": 11.476192474365234, "learning_rate": 9.60807318915133e-06, "loss": 1.2553, "step": 20010 }, { "epoch": 0.527357319486472, "grad_norm": 17.393442153930664, "learning_rate": 9.599548283590869e-06, "loss": 1.3957, "step": 20020 }, { "epoch": 0.5276207347309707, "grad_norm": 8.594742774963379, "learning_rate": 9.59102366951228e-06, "loss": 1.1821, "step": 20030 }, { "epoch": 0.5278841499754695, "grad_norm": 10.096039772033691, "learning_rate": 9.582499353120472e-06, "loss": 1.431, "step": 20040 }, { "epoch": 0.5281475652199682, "grad_norm": 6.465191841125488, "learning_rate": 9.573975340620153e-06, "loss": 1.4164, "step": 20050 }, { "epoch": 0.5284109804644669, "grad_norm": 6.256454944610596, "learning_rate": 9.565451638215803e-06, "loss": 1.3888, "step": 20060 }, { "epoch": 0.5286743957089657, "grad_norm": 3.5726234912872314, "learning_rate": 9.556928252111679e-06, "loss": 1.4147, "step": 20070 }, { "epoch": 0.5289378109534644, "grad_norm": 12.142322540283203, "learning_rate": 9.548405188511806e-06, "loss": 1.3382, "step": 20080 }, { "epoch": 0.5292012261979632, "grad_norm": 9.740056037902832, "learning_rate": 9.53988245361997e-06, "loss": 1.3794, "step": 20090 }, { "epoch": 0.5294646414424619, "grad_norm": 5.354048728942871, "learning_rate": 9.531360053639724e-06, "loss": 1.3674, "step": 20100 }, { "epoch": 0.5297280566869607, "grad_norm": 4.187978744506836, "learning_rate": 9.522837994774376e-06, "loss": 1.2795, "step": 20110 }, { "epoch": 0.5299914719314593, "grad_norm": 8.987088203430176, "learning_rate": 9.514316283226987e-06, "loss": 1.3634, "step": 20120 }, { "epoch": 0.5302548871759581, "grad_norm": 15.204462051391602, "learning_rate": 9.505794925200363e-06, "loss": 1.5045, "step": 20130 }, { "epoch": 0.5305183024204568, "grad_norm": 5.814431667327881, "learning_rate": 9.497273926897052e-06, "loss": 1.2509, "step": 20140 }, { "epoch": 0.5307817176649555, "grad_norm": 4.966856479644775, "learning_rate": 9.488753294519339e-06, "loss": 1.3408, "step": 20150 }, { "epoch": 0.5310451329094543, "grad_norm": 6.494147777557373, "learning_rate": 9.480233034269245e-06, "loss": 1.3895, "step": 20160 }, { "epoch": 0.531308548153953, "grad_norm": 15.840327262878418, "learning_rate": 9.471713152348522e-06, "loss": 1.321, "step": 20170 }, { "epoch": 0.5315719633984518, "grad_norm": 5.250042915344238, "learning_rate": 9.463193654958642e-06, "loss": 1.2217, "step": 20180 }, { "epoch": 0.5318353786429505, "grad_norm": 612.7410888671875, "learning_rate": 9.454674548300805e-06, "loss": 1.2964, "step": 20190 }, { "epoch": 0.5320987938874493, "grad_norm": 16.88507080078125, "learning_rate": 9.446155838575914e-06, "loss": 1.3933, "step": 20200 }, { "epoch": 0.532362209131948, "grad_norm": 10.91889762878418, "learning_rate": 9.437637531984593e-06, "loss": 1.2719, "step": 20210 }, { "epoch": 0.5326256243764467, "grad_norm": 10.554426193237305, "learning_rate": 9.429119634727171e-06, "loss": 1.3065, "step": 20220 }, { "epoch": 0.5328890396209455, "grad_norm": 116.7707290649414, "learning_rate": 9.420602153003676e-06, "loss": 1.3798, "step": 20230 }, { "epoch": 0.5331524548654442, "grad_norm": 47.651519775390625, "learning_rate": 9.412085093013841e-06, "loss": 1.2545, "step": 20240 }, { "epoch": 0.533415870109943, "grad_norm": 9.32347297668457, "learning_rate": 9.403568460957078e-06, "loss": 1.3333, "step": 20250 }, { "epoch": 0.5336792853544416, "grad_norm": 9.226522445678711, "learning_rate": 9.3950522630325e-06, "loss": 1.4144, "step": 20260 }, { "epoch": 0.5339427005989404, "grad_norm": 20.721189498901367, "learning_rate": 9.3865365054389e-06, "loss": 1.2635, "step": 20270 }, { "epoch": 0.5342061158434391, "grad_norm": 19.534564971923828, "learning_rate": 9.378021194374747e-06, "loss": 1.3063, "step": 20280 }, { "epoch": 0.5344695310879379, "grad_norm": 80.9044189453125, "learning_rate": 9.3695063360382e-06, "loss": 1.3987, "step": 20290 }, { "epoch": 0.5347329463324366, "grad_norm": 3.3335559368133545, "learning_rate": 9.360991936627064e-06, "loss": 1.3448, "step": 20300 }, { "epoch": 0.5349963615769353, "grad_norm": 5.73553466796875, "learning_rate": 9.352478002338823e-06, "loss": 1.2929, "step": 20310 }, { "epoch": 0.5352597768214341, "grad_norm": 24.02547264099121, "learning_rate": 9.34396453937063e-06, "loss": 1.25, "step": 20320 }, { "epoch": 0.5355231920659328, "grad_norm": 8.770973205566406, "learning_rate": 9.335451553919283e-06, "loss": 1.3628, "step": 20330 }, { "epoch": 0.5357866073104316, "grad_norm": 6.259378433227539, "learning_rate": 9.326939052181238e-06, "loss": 1.3167, "step": 20340 }, { "epoch": 0.5360500225549303, "grad_norm": 13.662786483764648, "learning_rate": 9.3184270403526e-06, "loss": 1.4321, "step": 20350 }, { "epoch": 0.5363134377994291, "grad_norm": 34.336181640625, "learning_rate": 9.309915524629109e-06, "loss": 1.3401, "step": 20360 }, { "epoch": 0.5365768530439278, "grad_norm": 12.778273582458496, "learning_rate": 9.301404511206152e-06, "loss": 1.2822, "step": 20370 }, { "epoch": 0.5368402682884266, "grad_norm": 6.733778953552246, "learning_rate": 9.292894006278748e-06, "loss": 1.4229, "step": 20380 }, { "epoch": 0.5371036835329253, "grad_norm": 5.283506870269775, "learning_rate": 9.284384016041549e-06, "loss": 1.2162, "step": 20390 }, { "epoch": 0.537367098777424, "grad_norm": 9.889086723327637, "learning_rate": 9.275874546688831e-06, "loss": 1.3475, "step": 20400 }, { "epoch": 0.5376305140219227, "grad_norm": 6.22169828414917, "learning_rate": 9.26736560441448e-06, "loss": 1.3162, "step": 20410 }, { "epoch": 0.5378939292664214, "grad_norm": 6.661123752593994, "learning_rate": 9.258857195412015e-06, "loss": 1.2931, "step": 20420 }, { "epoch": 0.5381573445109202, "grad_norm": 12.40054702758789, "learning_rate": 9.250349325874558e-06, "loss": 1.3809, "step": 20430 }, { "epoch": 0.5384207597554189, "grad_norm": 8.174386024475098, "learning_rate": 9.241842001994842e-06, "loss": 1.3467, "step": 20440 }, { "epoch": 0.5386841749999177, "grad_norm": 15.328058242797852, "learning_rate": 9.233335229965199e-06, "loss": 1.353, "step": 20450 }, { "epoch": 0.5389475902444164, "grad_norm": 6.412385940551758, "learning_rate": 9.224829015977558e-06, "loss": 1.3483, "step": 20460 }, { "epoch": 0.5392110054889152, "grad_norm": 4.807486057281494, "learning_rate": 9.21632336622345e-06, "loss": 1.3242, "step": 20470 }, { "epoch": 0.5394744207334139, "grad_norm": 2.8902223110198975, "learning_rate": 9.207818286893984e-06, "loss": 1.3244, "step": 20480 }, { "epoch": 0.5397378359779126, "grad_norm": 3.055560827255249, "learning_rate": 9.199313784179863e-06, "loss": 1.2437, "step": 20490 }, { "epoch": 0.5400012512224114, "grad_norm": 11.85980224609375, "learning_rate": 9.190809864271372e-06, "loss": 1.3443, "step": 20500 }, { "epoch": 0.5402646664669101, "grad_norm": 18.41385841369629, "learning_rate": 9.182306533358358e-06, "loss": 1.3776, "step": 20510 }, { "epoch": 0.5405280817114089, "grad_norm": 7.31540060043335, "learning_rate": 9.173803797630255e-06, "loss": 1.3802, "step": 20520 }, { "epoch": 0.5407914969559076, "grad_norm": 10.61247444152832, "learning_rate": 9.165301663276056e-06, "loss": 1.3786, "step": 20530 }, { "epoch": 0.5410549122004064, "grad_norm": 11.754375457763672, "learning_rate": 9.156800136484315e-06, "loss": 1.3429, "step": 20540 }, { "epoch": 0.541318327444905, "grad_norm": 9.424421310424805, "learning_rate": 9.148299223443145e-06, "loss": 1.3342, "step": 20550 }, { "epoch": 0.5415817426894038, "grad_norm": 3.3435277938842773, "learning_rate": 9.13979893034022e-06, "loss": 1.3249, "step": 20560 }, { "epoch": 0.5418451579339025, "grad_norm": 7.513850212097168, "learning_rate": 9.13129926336275e-06, "loss": 1.2849, "step": 20570 }, { "epoch": 0.5421085731784012, "grad_norm": 11.312723159790039, "learning_rate": 9.122800228697495e-06, "loss": 1.2955, "step": 20580 }, { "epoch": 0.5423719884229, "grad_norm": 9.243160247802734, "learning_rate": 9.114301832530755e-06, "loss": 1.369, "step": 20590 }, { "epoch": 0.5426354036673987, "grad_norm": 6.749199867248535, "learning_rate": 9.105804081048372e-06, "loss": 1.3163, "step": 20600 }, { "epoch": 0.5428988189118975, "grad_norm": 4.227711200714111, "learning_rate": 9.09730698043571e-06, "loss": 1.2846, "step": 20610 }, { "epoch": 0.5431622341563962, "grad_norm": 14.402875900268555, "learning_rate": 9.088810536877652e-06, "loss": 1.2877, "step": 20620 }, { "epoch": 0.543425649400895, "grad_norm": 12.266899108886719, "learning_rate": 9.080314756558622e-06, "loss": 1.3569, "step": 20630 }, { "epoch": 0.5436890646453937, "grad_norm": 7.241451263427734, "learning_rate": 9.071819645662548e-06, "loss": 1.3087, "step": 20640 }, { "epoch": 0.5439524798898925, "grad_norm": 5.761120319366455, "learning_rate": 9.063325210372876e-06, "loss": 1.2729, "step": 20650 }, { "epoch": 0.5442158951343912, "grad_norm": 6.2109761238098145, "learning_rate": 9.054831456872556e-06, "loss": 1.3793, "step": 20660 }, { "epoch": 0.5444793103788899, "grad_norm": 10.001215934753418, "learning_rate": 9.046338391344044e-06, "loss": 1.3159, "step": 20670 }, { "epoch": 0.5447427256233887, "grad_norm": 4.342574119567871, "learning_rate": 9.037846019969296e-06, "loss": 1.3992, "step": 20680 }, { "epoch": 0.5450061408678873, "grad_norm": 12.889762878417969, "learning_rate": 9.029354348929759e-06, "loss": 1.2675, "step": 20690 }, { "epoch": 0.5452695561123861, "grad_norm": 7.007455348968506, "learning_rate": 9.020863384406378e-06, "loss": 1.3296, "step": 20700 }, { "epoch": 0.5455329713568848, "grad_norm": 22.111038208007812, "learning_rate": 9.012373132579582e-06, "loss": 1.2592, "step": 20710 }, { "epoch": 0.5457963866013836, "grad_norm": 14.419602394104004, "learning_rate": 9.003883599629266e-06, "loss": 1.3844, "step": 20720 }, { "epoch": 0.5460598018458823, "grad_norm": 2.2722208499908447, "learning_rate": 8.995394791734825e-06, "loss": 1.3155, "step": 20730 }, { "epoch": 0.546323217090381, "grad_norm": 6.060754776000977, "learning_rate": 8.98690671507511e-06, "loss": 1.2926, "step": 20740 }, { "epoch": 0.5465866323348798, "grad_norm": 9.068933486938477, "learning_rate": 8.978419375828448e-06, "loss": 1.4526, "step": 20750 }, { "epoch": 0.5468500475793785, "grad_norm": 11.597220420837402, "learning_rate": 8.969932780172623e-06, "loss": 1.2994, "step": 20760 }, { "epoch": 0.5471134628238773, "grad_norm": 8.490080833435059, "learning_rate": 8.961446934284886e-06, "loss": 1.2906, "step": 20770 }, { "epoch": 0.547376878068376, "grad_norm": 17.67237091064453, "learning_rate": 8.952961844341933e-06, "loss": 1.3979, "step": 20780 }, { "epoch": 0.5476402933128748, "grad_norm": 7.586197376251221, "learning_rate": 8.94447751651991e-06, "loss": 1.3617, "step": 20790 }, { "epoch": 0.5479037085573735, "grad_norm": 10.99543571472168, "learning_rate": 8.93599395699442e-06, "loss": 1.2362, "step": 20800 }, { "epoch": 0.5481671238018723, "grad_norm": 3.9197394847869873, "learning_rate": 8.927511171940496e-06, "loss": 1.1337, "step": 20810 }, { "epoch": 0.548430539046371, "grad_norm": 10.230262756347656, "learning_rate": 8.919029167532612e-06, "loss": 1.2125, "step": 20820 }, { "epoch": 0.5486939542908696, "grad_norm": 9.550813674926758, "learning_rate": 8.91054794994467e-06, "loss": 1.3553, "step": 20830 }, { "epoch": 0.5489573695353684, "grad_norm": 6.1724324226379395, "learning_rate": 8.90206752535e-06, "loss": 1.367, "step": 20840 }, { "epoch": 0.5492207847798671, "grad_norm": 3.3362905979156494, "learning_rate": 8.89358789992136e-06, "loss": 1.322, "step": 20850 }, { "epoch": 0.5494842000243659, "grad_norm": 9.346821784973145, "learning_rate": 8.885109079830918e-06, "loss": 1.3633, "step": 20860 }, { "epoch": 0.5497476152688646, "grad_norm": 8.440089225769043, "learning_rate": 8.876631071250266e-06, "loss": 1.2528, "step": 20870 }, { "epoch": 0.5500110305133634, "grad_norm": 12.390538215637207, "learning_rate": 8.868153880350394e-06, "loss": 1.3402, "step": 20880 }, { "epoch": 0.5502744457578621, "grad_norm": 13.170918464660645, "learning_rate": 8.859677513301705e-06, "loss": 1.1498, "step": 20890 }, { "epoch": 0.5505378610023609, "grad_norm": 9.257023811340332, "learning_rate": 8.851201976273997e-06, "loss": 1.3826, "step": 20900 }, { "epoch": 0.5508012762468596, "grad_norm": 10.819096565246582, "learning_rate": 8.84272727543647e-06, "loss": 1.2814, "step": 20910 }, { "epoch": 0.5510646914913583, "grad_norm": 9.847457885742188, "learning_rate": 8.834253416957714e-06, "loss": 1.2574, "step": 20920 }, { "epoch": 0.5513281067358571, "grad_norm": 7.28328800201416, "learning_rate": 8.825780407005697e-06, "loss": 1.1774, "step": 20930 }, { "epoch": 0.5515915219803558, "grad_norm": 3.793949604034424, "learning_rate": 8.817308251747782e-06, "loss": 1.1998, "step": 20940 }, { "epoch": 0.5518549372248546, "grad_norm": 7.004604339599609, "learning_rate": 8.8088369573507e-06, "loss": 1.245, "step": 20950 }, { "epoch": 0.5521183524693533, "grad_norm": 13.110543251037598, "learning_rate": 8.800366529980563e-06, "loss": 1.4016, "step": 20960 }, { "epoch": 0.5523817677138521, "grad_norm": 3.2150449752807617, "learning_rate": 8.791896975802843e-06, "loss": 1.2877, "step": 20970 }, { "epoch": 0.5526451829583507, "grad_norm": 13.143166542053223, "learning_rate": 8.783428300982393e-06, "loss": 1.2929, "step": 20980 }, { "epoch": 0.5529085982028495, "grad_norm": 3.1480088233947754, "learning_rate": 8.774960511683402e-06, "loss": 1.1662, "step": 20990 }, { "epoch": 0.5531720134473482, "grad_norm": 11.486973762512207, "learning_rate": 8.76649361406943e-06, "loss": 1.2509, "step": 21000 }, { "epoch": 0.5534354286918469, "grad_norm": 7.297770977020264, "learning_rate": 8.75802761430339e-06, "loss": 1.3359, "step": 21010 }, { "epoch": 0.5536988439363457, "grad_norm": 3.477478504180908, "learning_rate": 8.749562518547532e-06, "loss": 1.2418, "step": 21020 }, { "epoch": 0.5539622591808444, "grad_norm": 6.981030464172363, "learning_rate": 8.741098332963459e-06, "loss": 1.3224, "step": 21030 }, { "epoch": 0.5542256744253432, "grad_norm": 7.039678573608398, "learning_rate": 8.732635063712096e-06, "loss": 1.2956, "step": 21040 }, { "epoch": 0.5544890896698419, "grad_norm": 3.488987922668457, "learning_rate": 8.724172716953717e-06, "loss": 1.2812, "step": 21050 }, { "epoch": 0.5547525049143407, "grad_norm": 9.06523323059082, "learning_rate": 8.715711298847913e-06, "loss": 1.3684, "step": 21060 }, { "epoch": 0.5550159201588394, "grad_norm": 6.24688196182251, "learning_rate": 8.707250815553605e-06, "loss": 1.3058, "step": 21070 }, { "epoch": 0.5552793354033382, "grad_norm": 3.984243154525757, "learning_rate": 8.69879127322904e-06, "loss": 1.1723, "step": 21080 }, { "epoch": 0.5555427506478369, "grad_norm": 3.7519986629486084, "learning_rate": 8.69033267803176e-06, "loss": 1.3182, "step": 21090 }, { "epoch": 0.5558061658923356, "grad_norm": 9.409341812133789, "learning_rate": 8.681875036118632e-06, "loss": 1.3108, "step": 21100 }, { "epoch": 0.5560695811368344, "grad_norm": 8.394383430480957, "learning_rate": 8.673418353645834e-06, "loss": 1.288, "step": 21110 }, { "epoch": 0.556332996381333, "grad_norm": 9.494107246398926, "learning_rate": 8.664962636768834e-06, "loss": 1.3643, "step": 21120 }, { "epoch": 0.5565964116258318, "grad_norm": 4.449470520019531, "learning_rate": 8.656507891642405e-06, "loss": 1.4254, "step": 21130 }, { "epoch": 0.5568598268703305, "grad_norm": 5.849430561065674, "learning_rate": 8.648054124420606e-06, "loss": 1.3365, "step": 21140 }, { "epoch": 0.5571232421148293, "grad_norm": 9.65716552734375, "learning_rate": 8.639601341256788e-06, "loss": 1.2583, "step": 21150 }, { "epoch": 0.557386657359328, "grad_norm": 10.847428321838379, "learning_rate": 8.631149548303588e-06, "loss": 1.3508, "step": 21160 }, { "epoch": 0.5576500726038268, "grad_norm": 9.496957778930664, "learning_rate": 8.622698751712915e-06, "loss": 1.2268, "step": 21170 }, { "epoch": 0.5579134878483255, "grad_norm": 10.514594078063965, "learning_rate": 8.614248957635961e-06, "loss": 1.1914, "step": 21180 }, { "epoch": 0.5581769030928242, "grad_norm": 9.496868133544922, "learning_rate": 8.60580017222319e-06, "loss": 1.375, "step": 21190 }, { "epoch": 0.558440318337323, "grad_norm": 5.875173091888428, "learning_rate": 8.597352401624315e-06, "loss": 1.406, "step": 21200 }, { "epoch": 0.5587037335818217, "grad_norm": 3.749242067337036, "learning_rate": 8.588905651988328e-06, "loss": 1.2289, "step": 21210 }, { "epoch": 0.5589671488263205, "grad_norm": 12.955780982971191, "learning_rate": 8.580459929463474e-06, "loss": 1.2991, "step": 21220 }, { "epoch": 0.5592305640708192, "grad_norm": 5.630696773529053, "learning_rate": 8.572015240197244e-06, "loss": 1.3229, "step": 21230 }, { "epoch": 0.559493979315318, "grad_norm": 7.698151111602783, "learning_rate": 8.563571590336383e-06, "loss": 1.2925, "step": 21240 }, { "epoch": 0.5597573945598167, "grad_norm": 10.453741073608398, "learning_rate": 8.555128986026877e-06, "loss": 1.2078, "step": 21250 }, { "epoch": 0.5600208098043153, "grad_norm": 13.10828971862793, "learning_rate": 8.546687433413952e-06, "loss": 1.3802, "step": 21260 }, { "epoch": 0.5602842250488141, "grad_norm": 2.817138671875, "learning_rate": 8.538246938642066e-06, "loss": 1.3378, "step": 21270 }, { "epoch": 0.5605476402933128, "grad_norm": 7.206291675567627, "learning_rate": 8.529807507854907e-06, "loss": 1.2919, "step": 21280 }, { "epoch": 0.5608110555378116, "grad_norm": 5.667418956756592, "learning_rate": 8.521369147195403e-06, "loss": 1.1976, "step": 21290 }, { "epoch": 0.5610744707823103, "grad_norm": 11.784053802490234, "learning_rate": 8.512931862805674e-06, "loss": 1.2713, "step": 21300 }, { "epoch": 0.5613378860268091, "grad_norm": 4.39980411529541, "learning_rate": 8.504495660827081e-06, "loss": 1.1995, "step": 21310 }, { "epoch": 0.5616013012713078, "grad_norm": 11.938528060913086, "learning_rate": 8.49606054740019e-06, "loss": 1.2898, "step": 21320 }, { "epoch": 0.5618647165158066, "grad_norm": 4.576653957366943, "learning_rate": 8.487626528664771e-06, "loss": 1.2358, "step": 21330 }, { "epoch": 0.5621281317603053, "grad_norm": 3.9316813945770264, "learning_rate": 8.479193610759804e-06, "loss": 1.2551, "step": 21340 }, { "epoch": 0.562391547004804, "grad_norm": 4.396370887756348, "learning_rate": 8.470761799823461e-06, "loss": 1.3201, "step": 21350 }, { "epoch": 0.5626549622493028, "grad_norm": 9.86244010925293, "learning_rate": 8.462331101993111e-06, "loss": 1.274, "step": 21360 }, { "epoch": 0.5629183774938015, "grad_norm": 8.69291877746582, "learning_rate": 8.453901523405314e-06, "loss": 1.292, "step": 21370 }, { "epoch": 0.5631817927383003, "grad_norm": 15.838998794555664, "learning_rate": 8.445473070195811e-06, "loss": 1.3404, "step": 21380 }, { "epoch": 0.563445207982799, "grad_norm": 10.592498779296875, "learning_rate": 8.437045748499531e-06, "loss": 1.1974, "step": 21390 }, { "epoch": 0.5637086232272978, "grad_norm": 7.39174222946167, "learning_rate": 8.42861956445058e-06, "loss": 1.4133, "step": 21400 }, { "epoch": 0.5639720384717964, "grad_norm": 10.300704956054688, "learning_rate": 8.420194524182219e-06, "loss": 1.42, "step": 21410 }, { "epoch": 0.5642354537162952, "grad_norm": 8.795065879821777, "learning_rate": 8.411770633826898e-06, "loss": 1.3035, "step": 21420 }, { "epoch": 0.5644988689607939, "grad_norm": 7.380778789520264, "learning_rate": 8.403347899516218e-06, "loss": 1.337, "step": 21430 }, { "epoch": 0.5647622842052926, "grad_norm": 3.6607167720794678, "learning_rate": 8.394926327380942e-06, "loss": 1.3854, "step": 21440 }, { "epoch": 0.5650256994497914, "grad_norm": 8.759010314941406, "learning_rate": 8.386505923550989e-06, "loss": 1.1598, "step": 21450 }, { "epoch": 0.5652891146942901, "grad_norm": 12.915225982666016, "learning_rate": 8.37808669415542e-06, "loss": 1.4663, "step": 21460 }, { "epoch": 0.5655525299387889, "grad_norm": 3.9171478748321533, "learning_rate": 8.369668645322449e-06, "loss": 1.3152, "step": 21470 }, { "epoch": 0.5658159451832876, "grad_norm": 17.44007110595703, "learning_rate": 8.361251783179424e-06, "loss": 1.2279, "step": 21480 }, { "epoch": 0.5660793604277864, "grad_norm": 17.90239143371582, "learning_rate": 8.352836113852838e-06, "loss": 1.308, "step": 21490 }, { "epoch": 0.5663427756722851, "grad_norm": 8.316226959228516, "learning_rate": 8.344421643468316e-06, "loss": 1.2626, "step": 21500 }, { "epoch": 0.5666061909167839, "grad_norm": 11.291923522949219, "learning_rate": 8.336008378150593e-06, "loss": 1.261, "step": 21510 }, { "epoch": 0.5668696061612826, "grad_norm": 6.602450847625732, "learning_rate": 8.327596324023545e-06, "loss": 1.538, "step": 21520 }, { "epoch": 0.5671330214057813, "grad_norm": 9.672451972961426, "learning_rate": 8.319185487210162e-06, "loss": 1.267, "step": 21530 }, { "epoch": 0.5673964366502801, "grad_norm": 11.05896282196045, "learning_rate": 8.310775873832545e-06, "loss": 1.2924, "step": 21540 }, { "epoch": 0.5676598518947787, "grad_norm": 5.1027374267578125, "learning_rate": 8.302367490011904e-06, "loss": 1.3202, "step": 21550 }, { "epoch": 0.5679232671392775, "grad_norm": 6.766063690185547, "learning_rate": 8.293960341868558e-06, "loss": 1.3297, "step": 21560 }, { "epoch": 0.5681866823837762, "grad_norm": 9.693469047546387, "learning_rate": 8.285554435521926e-06, "loss": 1.2257, "step": 21570 }, { "epoch": 0.568450097628275, "grad_norm": 4.547384262084961, "learning_rate": 8.277149777090514e-06, "loss": 1.2374, "step": 21580 }, { "epoch": 0.5687135128727737, "grad_norm": 11.056100845336914, "learning_rate": 8.268746372691937e-06, "loss": 1.2828, "step": 21590 }, { "epoch": 0.5689769281172725, "grad_norm": 6.4510297775268555, "learning_rate": 8.260344228442883e-06, "loss": 1.2188, "step": 21600 }, { "epoch": 0.5692403433617712, "grad_norm": 6.645703315734863, "learning_rate": 8.25194335045913e-06, "loss": 1.2733, "step": 21610 }, { "epoch": 0.5695037586062699, "grad_norm": 5.12797212600708, "learning_rate": 8.243543744855526e-06, "loss": 1.2992, "step": 21620 }, { "epoch": 0.5697671738507687, "grad_norm": 6.631440162658691, "learning_rate": 8.235145417746004e-06, "loss": 1.3106, "step": 21630 }, { "epoch": 0.5700305890952674, "grad_norm": 9.748980522155762, "learning_rate": 8.226748375243556e-06, "loss": 1.3325, "step": 21640 }, { "epoch": 0.5702940043397662, "grad_norm": 8.703916549682617, "learning_rate": 8.218352623460247e-06, "loss": 1.2433, "step": 21650 }, { "epoch": 0.5705574195842649, "grad_norm": 6.609623908996582, "learning_rate": 8.2099581685072e-06, "loss": 1.2976, "step": 21660 }, { "epoch": 0.5708208348287637, "grad_norm": 11.439959526062012, "learning_rate": 8.201565016494592e-06, "loss": 1.3113, "step": 21670 }, { "epoch": 0.5710842500732624, "grad_norm": 17.232013702392578, "learning_rate": 8.19317317353165e-06, "loss": 1.2526, "step": 21680 }, { "epoch": 0.5713476653177612, "grad_norm": 10.785612106323242, "learning_rate": 8.184782645726654e-06, "loss": 1.3091, "step": 21690 }, { "epoch": 0.5716110805622598, "grad_norm": 18.17803192138672, "learning_rate": 8.176393439186923e-06, "loss": 1.348, "step": 21700 }, { "epoch": 0.5718744958067585, "grad_norm": 7.16942834854126, "learning_rate": 8.16800556001882e-06, "loss": 1.2567, "step": 21710 }, { "epoch": 0.5721379110512573, "grad_norm": 10.197978019714355, "learning_rate": 8.15961901432773e-06, "loss": 1.3366, "step": 21720 }, { "epoch": 0.572401326295756, "grad_norm": 8.756180763244629, "learning_rate": 8.151233808218077e-06, "loss": 1.2391, "step": 21730 }, { "epoch": 0.5726647415402548, "grad_norm": 5.928918838500977, "learning_rate": 8.142849947793308e-06, "loss": 1.2503, "step": 21740 }, { "epoch": 0.5729281567847535, "grad_norm": 3.1544532775878906, "learning_rate": 8.134467439155888e-06, "loss": 1.357, "step": 21750 }, { "epoch": 0.5731915720292523, "grad_norm": 9.561430931091309, "learning_rate": 8.126086288407304e-06, "loss": 1.3025, "step": 21760 }, { "epoch": 0.573454987273751, "grad_norm": 6.879429340362549, "learning_rate": 8.117706501648045e-06, "loss": 1.3337, "step": 21770 }, { "epoch": 0.5737184025182498, "grad_norm": 6.050500869750977, "learning_rate": 8.109328084977615e-06, "loss": 1.3466, "step": 21780 }, { "epoch": 0.5739818177627485, "grad_norm": 5.530920505523682, "learning_rate": 8.100951044494516e-06, "loss": 1.3284, "step": 21790 }, { "epoch": 0.5742452330072472, "grad_norm": 5.658797264099121, "learning_rate": 8.092575386296255e-06, "loss": 1.4156, "step": 21800 }, { "epoch": 0.574508648251746, "grad_norm": 9.025680541992188, "learning_rate": 8.08420111647933e-06, "loss": 1.3163, "step": 21810 }, { "epoch": 0.5747720634962447, "grad_norm": 9.012530326843262, "learning_rate": 8.075828241139218e-06, "loss": 1.2727, "step": 21820 }, { "epoch": 0.5750354787407435, "grad_norm": 5.21568489074707, "learning_rate": 8.067456766370398e-06, "loss": 1.3664, "step": 21830 }, { "epoch": 0.5752988939852421, "grad_norm": 7.612646102905273, "learning_rate": 8.059086698266317e-06, "loss": 1.3179, "step": 21840 }, { "epoch": 0.5755623092297409, "grad_norm": 5.655656337738037, "learning_rate": 8.050718042919405e-06, "loss": 1.3834, "step": 21850 }, { "epoch": 0.5758257244742396, "grad_norm": 10.597208023071289, "learning_rate": 8.042350806421059e-06, "loss": 1.3612, "step": 21860 }, { "epoch": 0.5760891397187383, "grad_norm": 4.16780424118042, "learning_rate": 8.033984994861653e-06, "loss": 1.235, "step": 21870 }, { "epoch": 0.5763525549632371, "grad_norm": 12.96108627319336, "learning_rate": 8.025620614330506e-06, "loss": 1.2194, "step": 21880 }, { "epoch": 0.5766159702077358, "grad_norm": 14.43004322052002, "learning_rate": 8.017257670915908e-06, "loss": 1.404, "step": 21890 }, { "epoch": 0.5768793854522346, "grad_norm": 5.399217128753662, "learning_rate": 8.008896170705106e-06, "loss": 1.2659, "step": 21900 }, { "epoch": 0.5771428006967333, "grad_norm": 6.441407203674316, "learning_rate": 8.000536119784285e-06, "loss": 1.3241, "step": 21910 }, { "epoch": 0.5774062159412321, "grad_norm": 9.41280746459961, "learning_rate": 7.992177524238588e-06, "loss": 1.299, "step": 21920 }, { "epoch": 0.5776696311857308, "grad_norm": 7.931628227233887, "learning_rate": 7.983820390152086e-06, "loss": 1.2155, "step": 21930 }, { "epoch": 0.5779330464302296, "grad_norm": 13.773015975952148, "learning_rate": 7.975464723607794e-06, "loss": 1.4778, "step": 21940 }, { "epoch": 0.5781964616747283, "grad_norm": 9.561707496643066, "learning_rate": 7.967110530687655e-06, "loss": 1.194, "step": 21950 }, { "epoch": 0.578459876919227, "grad_norm": 10.50223159790039, "learning_rate": 7.958757817472546e-06, "loss": 1.3109, "step": 21960 }, { "epoch": 0.5787232921637258, "grad_norm": 6.874154567718506, "learning_rate": 7.95040659004226e-06, "loss": 1.3248, "step": 21970 }, { "epoch": 0.5789867074082244, "grad_norm": 8.245305061340332, "learning_rate": 7.94205685447551e-06, "loss": 1.2976, "step": 21980 }, { "epoch": 0.5792501226527232, "grad_norm": 5.68314790725708, "learning_rate": 7.933708616849923e-06, "loss": 1.2739, "step": 21990 }, { "epoch": 0.5795135378972219, "grad_norm": 6.084063529968262, "learning_rate": 7.925361883242036e-06, "loss": 1.3234, "step": 22000 }, { "epoch": 0.5797769531417207, "grad_norm": 12.827095031738281, "learning_rate": 7.917016659727297e-06, "loss": 1.3888, "step": 22010 }, { "epoch": 0.5800403683862194, "grad_norm": 9.698895454406738, "learning_rate": 7.908672952380045e-06, "loss": 1.3301, "step": 22020 }, { "epoch": 0.5803037836307182, "grad_norm": 7.428830623626709, "learning_rate": 7.900330767273521e-06, "loss": 1.2687, "step": 22030 }, { "epoch": 0.5805671988752169, "grad_norm": 9.695491790771484, "learning_rate": 7.891990110479858e-06, "loss": 1.3227, "step": 22040 }, { "epoch": 0.5808306141197156, "grad_norm": 6.558418273925781, "learning_rate": 7.883650988070074e-06, "loss": 1.3641, "step": 22050 }, { "epoch": 0.5810940293642144, "grad_norm": 12.280397415161133, "learning_rate": 7.875313406114073e-06, "loss": 1.4201, "step": 22060 }, { "epoch": 0.5813574446087131, "grad_norm": 4.086822032928467, "learning_rate": 7.866977370680633e-06, "loss": 1.301, "step": 22070 }, { "epoch": 0.5816208598532119, "grad_norm": 4.18739652633667, "learning_rate": 7.85864288783742e-06, "loss": 1.3656, "step": 22080 }, { "epoch": 0.5818842750977106, "grad_norm": 4.360587120056152, "learning_rate": 7.850309963650947e-06, "loss": 1.3468, "step": 22090 }, { "epoch": 0.5821476903422094, "grad_norm": 11.275213241577148, "learning_rate": 7.84197860418661e-06, "loss": 1.3492, "step": 22100 }, { "epoch": 0.5824111055867081, "grad_norm": 3.603940963745117, "learning_rate": 7.833648815508663e-06, "loss": 1.3176, "step": 22110 }, { "epoch": 0.5826745208312069, "grad_norm": 5.707281112670898, "learning_rate": 7.825320603680213e-06, "loss": 1.3373, "step": 22120 }, { "epoch": 0.5829379360757055, "grad_norm": 2.9216716289520264, "learning_rate": 7.816993974763224e-06, "loss": 1.2925, "step": 22130 }, { "epoch": 0.5832013513202042, "grad_norm": 7.1156158447265625, "learning_rate": 7.808668934818503e-06, "loss": 1.2506, "step": 22140 }, { "epoch": 0.583464766564703, "grad_norm": 9.864612579345703, "learning_rate": 7.8003454899057e-06, "loss": 1.339, "step": 22150 }, { "epoch": 0.5837281818092017, "grad_norm": 10.434876441955566, "learning_rate": 7.79202364608331e-06, "loss": 1.2476, "step": 22160 }, { "epoch": 0.5839915970537005, "grad_norm": 7.231067180633545, "learning_rate": 7.783703409408655e-06, "loss": 1.3085, "step": 22170 }, { "epoch": 0.5842550122981992, "grad_norm": 11.487489700317383, "learning_rate": 7.775384785937899e-06, "loss": 1.235, "step": 22180 }, { "epoch": 0.584518427542698, "grad_norm": 6.1279144287109375, "learning_rate": 7.767067781726016e-06, "loss": 1.3632, "step": 22190 }, { "epoch": 0.5847818427871967, "grad_norm": 6.82244348526001, "learning_rate": 7.758752402826808e-06, "loss": 1.1362, "step": 22200 }, { "epoch": 0.5850452580316955, "grad_norm": 5.009152412414551, "learning_rate": 7.750438655292901e-06, "loss": 1.2972, "step": 22210 }, { "epoch": 0.5853086732761942, "grad_norm": 4.370853900909424, "learning_rate": 7.742126545175727e-06, "loss": 1.2455, "step": 22220 }, { "epoch": 0.5855720885206929, "grad_norm": 12.881572723388672, "learning_rate": 7.733816078525527e-06, "loss": 1.2228, "step": 22230 }, { "epoch": 0.5858355037651917, "grad_norm": 6.866861820220947, "learning_rate": 7.725507261391344e-06, "loss": 1.2054, "step": 22240 }, { "epoch": 0.5860989190096904, "grad_norm": 9.818111419677734, "learning_rate": 7.71720009982102e-06, "loss": 1.286, "step": 22250 }, { "epoch": 0.5863623342541892, "grad_norm": 8.556487083435059, "learning_rate": 7.708894599861198e-06, "loss": 1.2043, "step": 22260 }, { "epoch": 0.5866257494986878, "grad_norm": 7.827797889709473, "learning_rate": 7.700590767557302e-06, "loss": 1.2298, "step": 22270 }, { "epoch": 0.5868891647431866, "grad_norm": 4.906430721282959, "learning_rate": 7.692288608953556e-06, "loss": 1.3703, "step": 22280 }, { "epoch": 0.5871525799876853, "grad_norm": 5.548553466796875, "learning_rate": 7.683988130092958e-06, "loss": 1.3155, "step": 22290 }, { "epoch": 0.5874159952321841, "grad_norm": 9.5966796875, "learning_rate": 7.67568933701727e-06, "loss": 1.221, "step": 22300 }, { "epoch": 0.5876794104766828, "grad_norm": 14.21728515625, "learning_rate": 7.66739223576705e-06, "loss": 1.1998, "step": 22310 }, { "epoch": 0.5879428257211815, "grad_norm": 3.2985284328460693, "learning_rate": 7.659096832381613e-06, "loss": 1.3749, "step": 22320 }, { "epoch": 0.5882062409656803, "grad_norm": 9.1517972946167, "learning_rate": 7.650803132899039e-06, "loss": 1.2763, "step": 22330 }, { "epoch": 0.588469656210179, "grad_norm": 2.767810106277466, "learning_rate": 7.642511143356168e-06, "loss": 1.1611, "step": 22340 }, { "epoch": 0.5887330714546778, "grad_norm": 7.031129360198975, "learning_rate": 7.634220869788593e-06, "loss": 1.1701, "step": 22350 }, { "epoch": 0.5889964866991765, "grad_norm": 11.872005462646484, "learning_rate": 7.625932318230662e-06, "loss": 1.2993, "step": 22360 }, { "epoch": 0.5892599019436753, "grad_norm": 6.16968297958374, "learning_rate": 7.617645494715467e-06, "loss": 1.2245, "step": 22370 }, { "epoch": 0.589523317188174, "grad_norm": 8.109123229980469, "learning_rate": 7.609360405274848e-06, "loss": 1.1943, "step": 22380 }, { "epoch": 0.5897867324326727, "grad_norm": 4.905035495758057, "learning_rate": 7.6010770559393784e-06, "loss": 1.1714, "step": 22390 }, { "epoch": 0.5900501476771715, "grad_norm": 18.849170684814453, "learning_rate": 7.592795452738355e-06, "loss": 1.252, "step": 22400 }, { "epoch": 0.5903135629216701, "grad_norm": 14.629880905151367, "learning_rate": 7.584515601699822e-06, "loss": 1.3357, "step": 22410 }, { "epoch": 0.5905769781661689, "grad_norm": 8.929216384887695, "learning_rate": 7.576237508850537e-06, "loss": 1.3252, "step": 22420 }, { "epoch": 0.5908403934106676, "grad_norm": 15.515727996826172, "learning_rate": 7.56796118021598e-06, "loss": 1.2437, "step": 22430 }, { "epoch": 0.5911038086551664, "grad_norm": 7.28492546081543, "learning_rate": 7.55968662182035e-06, "loss": 1.4, "step": 22440 }, { "epoch": 0.5913672238996651, "grad_norm": 6.34182071685791, "learning_rate": 7.551413839686551e-06, "loss": 1.3336, "step": 22450 }, { "epoch": 0.5916306391441639, "grad_norm": 6.427453517913818, "learning_rate": 7.543142839836199e-06, "loss": 1.3027, "step": 22460 }, { "epoch": 0.5918940543886626, "grad_norm": 2.2451975345611572, "learning_rate": 7.534873628289612e-06, "loss": 1.3486, "step": 22470 }, { "epoch": 0.5921574696331613, "grad_norm": 11.31552791595459, "learning_rate": 7.526606211065803e-06, "loss": 1.2321, "step": 22480 }, { "epoch": 0.5924208848776601, "grad_norm": 6.722403526306152, "learning_rate": 7.5183405941824875e-06, "loss": 1.312, "step": 22490 }, { "epoch": 0.5926843001221588, "grad_norm": 5.227752208709717, "learning_rate": 7.510076783656061e-06, "loss": 1.2261, "step": 22500 }, { "epoch": 0.5929477153666576, "grad_norm": 7.010986804962158, "learning_rate": 7.50181478550161e-06, "loss": 1.2546, "step": 22510 }, { "epoch": 0.5932111306111563, "grad_norm": 8.845512390136719, "learning_rate": 7.493554605732897e-06, "loss": 1.263, "step": 22520 }, { "epoch": 0.5934745458556551, "grad_norm": 5.356457710266113, "learning_rate": 7.485296250362365e-06, "loss": 1.2805, "step": 22530 }, { "epoch": 0.5937379611001538, "grad_norm": 9.55726432800293, "learning_rate": 7.477039725401128e-06, "loss": 1.2362, "step": 22540 }, { "epoch": 0.5940013763446526, "grad_norm": 5.571455001831055, "learning_rate": 7.46878503685897e-06, "loss": 1.3074, "step": 22550 }, { "epoch": 0.5942647915891512, "grad_norm": 12.443653106689453, "learning_rate": 7.460532190744332e-06, "loss": 1.3609, "step": 22560 }, { "epoch": 0.5945282068336499, "grad_norm": 10.531793594360352, "learning_rate": 7.452281193064318e-06, "loss": 1.2594, "step": 22570 }, { "epoch": 0.5947916220781487, "grad_norm": 7.4541215896606445, "learning_rate": 7.444032049824686e-06, "loss": 1.2709, "step": 22580 }, { "epoch": 0.5950550373226474, "grad_norm": 7.200699329376221, "learning_rate": 7.435784767029848e-06, "loss": 1.3949, "step": 22590 }, { "epoch": 0.5953184525671462, "grad_norm": 8.660252571105957, "learning_rate": 7.42753935068286e-06, "loss": 1.2772, "step": 22600 }, { "epoch": 0.5955818678116449, "grad_norm": 7.5081787109375, "learning_rate": 7.419295806785407e-06, "loss": 1.2971, "step": 22610 }, { "epoch": 0.5958452830561437, "grad_norm": 8.609156608581543, "learning_rate": 7.411054141337831e-06, "loss": 1.2775, "step": 22620 }, { "epoch": 0.5961086983006424, "grad_norm": 5.893675804138184, "learning_rate": 7.4028143603390945e-06, "loss": 1.2311, "step": 22630 }, { "epoch": 0.5963721135451412, "grad_norm": 10.977543830871582, "learning_rate": 7.394576469786794e-06, "loss": 1.2673, "step": 22640 }, { "epoch": 0.5966355287896399, "grad_norm": 5.313388347625732, "learning_rate": 7.386340475677145e-06, "loss": 1.1948, "step": 22650 }, { "epoch": 0.5968989440341386, "grad_norm": 9.700843811035156, "learning_rate": 7.378106384004985e-06, "loss": 1.2582, "step": 22660 }, { "epoch": 0.5971623592786374, "grad_norm": 5.491483688354492, "learning_rate": 7.369874200763766e-06, "loss": 1.2738, "step": 22670 }, { "epoch": 0.597425774523136, "grad_norm": 9.734855651855469, "learning_rate": 7.36164393194555e-06, "loss": 1.2314, "step": 22680 }, { "epoch": 0.5976891897676349, "grad_norm": 8.397748947143555, "learning_rate": 7.353415583541013e-06, "loss": 1.2533, "step": 22690 }, { "epoch": 0.5979526050121335, "grad_norm": 16.1705265045166, "learning_rate": 7.345189161539422e-06, "loss": 1.1657, "step": 22700 }, { "epoch": 0.5982160202566323, "grad_norm": 6.640005588531494, "learning_rate": 7.336964671928653e-06, "loss": 1.3113, "step": 22710 }, { "epoch": 0.598479435501131, "grad_norm": 18.024147033691406, "learning_rate": 7.328742120695162e-06, "loss": 1.3902, "step": 22720 }, { "epoch": 0.5987428507456298, "grad_norm": 9.143914222717285, "learning_rate": 7.320521513824004e-06, "loss": 1.161, "step": 22730 }, { "epoch": 0.5990062659901285, "grad_norm": 5.367995738983154, "learning_rate": 7.312302857298819e-06, "loss": 1.1857, "step": 22740 }, { "epoch": 0.5992696812346272, "grad_norm": 4.838736057281494, "learning_rate": 7.3040861571018216e-06, "loss": 1.2639, "step": 22750 }, { "epoch": 0.599533096479126, "grad_norm": 9.055205345153809, "learning_rate": 7.29587141921381e-06, "loss": 1.3902, "step": 22760 }, { "epoch": 0.5997965117236247, "grad_norm": 7.2745137214660645, "learning_rate": 7.2876586496141446e-06, "loss": 1.2344, "step": 22770 }, { "epoch": 0.6000599269681235, "grad_norm": 6.434030055999756, "learning_rate": 7.279447854280758e-06, "loss": 1.2757, "step": 22780 }, { "epoch": 0.6003233422126222, "grad_norm": 11.080344200134277, "learning_rate": 7.27123903919015e-06, "loss": 1.3027, "step": 22790 }, { "epoch": 0.600586757457121, "grad_norm": 9.357545852661133, "learning_rate": 7.263032210317371e-06, "loss": 1.3054, "step": 22800 }, { "epoch": 0.6008501727016197, "grad_norm": 9.832756996154785, "learning_rate": 7.254827373636035e-06, "loss": 1.2021, "step": 22810 }, { "epoch": 0.6011135879461185, "grad_norm": 10.783754348754883, "learning_rate": 7.2466245351182966e-06, "loss": 1.34, "step": 22820 }, { "epoch": 0.6013770031906172, "grad_norm": 14.342830657958984, "learning_rate": 7.238423700734858e-06, "loss": 1.2609, "step": 22830 }, { "epoch": 0.6016404184351158, "grad_norm": 4.901958465576172, "learning_rate": 7.230224876454966e-06, "loss": 1.1743, "step": 22840 }, { "epoch": 0.6019038336796146, "grad_norm": 12.286541938781738, "learning_rate": 7.222028068246405e-06, "loss": 1.34, "step": 22850 }, { "epoch": 0.6021672489241133, "grad_norm": 5.526609897613525, "learning_rate": 7.213833282075489e-06, "loss": 1.2289, "step": 22860 }, { "epoch": 0.6024306641686121, "grad_norm": 11.265115737915039, "learning_rate": 7.205640523907058e-06, "loss": 1.2791, "step": 22870 }, { "epoch": 0.6026940794131108, "grad_norm": 6.980140686035156, "learning_rate": 7.197449799704481e-06, "loss": 1.3153, "step": 22880 }, { "epoch": 0.6029574946576096, "grad_norm": 3.178445816040039, "learning_rate": 7.189261115429642e-06, "loss": 1.2374, "step": 22890 }, { "epoch": 0.6032209099021083, "grad_norm": 5.948081016540527, "learning_rate": 7.1810744770429465e-06, "loss": 1.278, "step": 22900 }, { "epoch": 0.603484325146607, "grad_norm": 7.235008716583252, "learning_rate": 7.172889890503303e-06, "loss": 1.3081, "step": 22910 }, { "epoch": 0.6037477403911058, "grad_norm": 3.4584312438964844, "learning_rate": 7.164707361768136e-06, "loss": 1.2099, "step": 22920 }, { "epoch": 0.6040111556356045, "grad_norm": 8.108735084533691, "learning_rate": 7.156526896793359e-06, "loss": 1.258, "step": 22930 }, { "epoch": 0.6042745708801033, "grad_norm": 13.977660179138184, "learning_rate": 7.148348501533396e-06, "loss": 1.2228, "step": 22940 }, { "epoch": 0.604537986124602, "grad_norm": 7.721965312957764, "learning_rate": 7.140172181941154e-06, "loss": 1.2565, "step": 22950 }, { "epoch": 0.6048014013691008, "grad_norm": 2.3341660499572754, "learning_rate": 7.131997943968039e-06, "loss": 1.3464, "step": 22960 }, { "epoch": 0.6050648166135995, "grad_norm": 15.479608535766602, "learning_rate": 7.123825793563939e-06, "loss": 1.2324, "step": 22970 }, { "epoch": 0.6053282318580983, "grad_norm": 8.016624450683594, "learning_rate": 7.115655736677216e-06, "loss": 1.3919, "step": 22980 }, { "epoch": 0.6055916471025969, "grad_norm": 6.168022155761719, "learning_rate": 7.10748777925471e-06, "loss": 1.1698, "step": 22990 }, { "epoch": 0.6058550623470956, "grad_norm": 3.46389102935791, "learning_rate": 7.0993219272417444e-06, "loss": 1.1989, "step": 23000 }, { "epoch": 0.6061184775915944, "grad_norm": 5.359439373016357, "learning_rate": 7.091158186582095e-06, "loss": 1.2705, "step": 23010 }, { "epoch": 0.6063818928360931, "grad_norm": 9.111003875732422, "learning_rate": 7.082996563218013e-06, "loss": 1.3928, "step": 23020 }, { "epoch": 0.6066453080805919, "grad_norm": 11.575864791870117, "learning_rate": 7.074837063090196e-06, "loss": 1.2596, "step": 23030 }, { "epoch": 0.6069087233250906, "grad_norm": 7.501400470733643, "learning_rate": 7.066679692137808e-06, "loss": 1.2704, "step": 23040 }, { "epoch": 0.6071721385695894, "grad_norm": 14.277399063110352, "learning_rate": 7.058524456298454e-06, "loss": 1.2956, "step": 23050 }, { "epoch": 0.6074355538140881, "grad_norm": 8.56878662109375, "learning_rate": 7.0503713615081885e-06, "loss": 1.3149, "step": 23060 }, { "epoch": 0.6076989690585869, "grad_norm": 7.688422679901123, "learning_rate": 7.042220413701516e-06, "loss": 1.1996, "step": 23070 }, { "epoch": 0.6079623843030856, "grad_norm": 3.858163356781006, "learning_rate": 7.034071618811361e-06, "loss": 1.2929, "step": 23080 }, { "epoch": 0.6082257995475843, "grad_norm": 14.03501033782959, "learning_rate": 7.025924982769091e-06, "loss": 1.2255, "step": 23090 }, { "epoch": 0.6084892147920831, "grad_norm": 9.552128791809082, "learning_rate": 7.0177805115045065e-06, "loss": 1.3797, "step": 23100 }, { "epoch": 0.6087526300365818, "grad_norm": 4.1083269119262695, "learning_rate": 7.009638210945824e-06, "loss": 1.2575, "step": 23110 }, { "epoch": 0.6090160452810806, "grad_norm": 6.741252899169922, "learning_rate": 7.001498087019683e-06, "loss": 1.1832, "step": 23120 }, { "epoch": 0.6092794605255792, "grad_norm": 8.085142135620117, "learning_rate": 6.99336014565114e-06, "loss": 1.158, "step": 23130 }, { "epoch": 0.609542875770078, "grad_norm": 9.445575714111328, "learning_rate": 6.9852243927636595e-06, "loss": 1.309, "step": 23140 }, { "epoch": 0.6098062910145767, "grad_norm": 6.4371819496154785, "learning_rate": 6.977090834279117e-06, "loss": 1.431, "step": 23150 }, { "epoch": 0.6100697062590755, "grad_norm": 11.130436897277832, "learning_rate": 6.968959476117787e-06, "loss": 1.3927, "step": 23160 }, { "epoch": 0.6103331215035742, "grad_norm": 6.79194974899292, "learning_rate": 6.960830324198342e-06, "loss": 1.1526, "step": 23170 }, { "epoch": 0.6105965367480729, "grad_norm": 6.564202308654785, "learning_rate": 6.952703384437863e-06, "loss": 1.2336, "step": 23180 }, { "epoch": 0.6108599519925717, "grad_norm": 7.813756465911865, "learning_rate": 6.944578662751791e-06, "loss": 1.3147, "step": 23190 }, { "epoch": 0.6111233672370704, "grad_norm": 3.7447543144226074, "learning_rate": 6.936456165053982e-06, "loss": 1.2858, "step": 23200 }, { "epoch": 0.6113867824815692, "grad_norm": 10.101070404052734, "learning_rate": 6.928335897256657e-06, "loss": 1.2691, "step": 23210 }, { "epoch": 0.6116501977260679, "grad_norm": 6.772805213928223, "learning_rate": 6.920217865270421e-06, "loss": 1.3508, "step": 23220 }, { "epoch": 0.6119136129705667, "grad_norm": 7.834812641143799, "learning_rate": 6.912102075004249e-06, "loss": 1.2964, "step": 23230 }, { "epoch": 0.6121770282150654, "grad_norm": 7.67508602142334, "learning_rate": 6.90398853236548e-06, "loss": 1.2855, "step": 23240 }, { "epoch": 0.6124404434595642, "grad_norm": 11.748907089233398, "learning_rate": 6.895877243259823e-06, "loss": 1.2013, "step": 23250 }, { "epoch": 0.6127038587040629, "grad_norm": 4.221116065979004, "learning_rate": 6.887768213591346e-06, "loss": 1.307, "step": 23260 }, { "epoch": 0.6129672739485615, "grad_norm": 6.391451835632324, "learning_rate": 6.879661449262469e-06, "loss": 1.3123, "step": 23270 }, { "epoch": 0.6132306891930603, "grad_norm": 5.364327907562256, "learning_rate": 6.871556956173973e-06, "loss": 1.2691, "step": 23280 }, { "epoch": 0.613494104437559, "grad_norm": 10.242194175720215, "learning_rate": 6.863454740224965e-06, "loss": 1.253, "step": 23290 }, { "epoch": 0.6137575196820578, "grad_norm": 8.30744457244873, "learning_rate": 6.855354807312917e-06, "loss": 1.3808, "step": 23300 }, { "epoch": 0.6140209349265565, "grad_norm": 4.349042892456055, "learning_rate": 6.847257163333628e-06, "loss": 1.2559, "step": 23310 }, { "epoch": 0.6142843501710553, "grad_norm": 3.6985018253326416, "learning_rate": 6.839161814181232e-06, "loss": 1.2363, "step": 23320 }, { "epoch": 0.614547765415554, "grad_norm": 6.641138076782227, "learning_rate": 6.831068765748194e-06, "loss": 1.3495, "step": 23330 }, { "epoch": 0.6148111806600528, "grad_norm": 7.832810878753662, "learning_rate": 6.822978023925305e-06, "loss": 1.3818, "step": 23340 }, { "epoch": 0.6150745959045515, "grad_norm": 7.324801445007324, "learning_rate": 6.814889594601673e-06, "loss": 1.3798, "step": 23350 }, { "epoch": 0.6153380111490502, "grad_norm": 17.30986976623535, "learning_rate": 6.806803483664726e-06, "loss": 1.3829, "step": 23360 }, { "epoch": 0.615601426393549, "grad_norm": 5.366345405578613, "learning_rate": 6.798719697000202e-06, "loss": 1.3172, "step": 23370 }, { "epoch": 0.6158648416380477, "grad_norm": 8.090523719787598, "learning_rate": 6.7906382404921554e-06, "loss": 1.2709, "step": 23380 }, { "epoch": 0.6161282568825465, "grad_norm": 9.028170585632324, "learning_rate": 6.78255912002294e-06, "loss": 1.2771, "step": 23390 }, { "epoch": 0.6163916721270452, "grad_norm": 12.425600051879883, "learning_rate": 6.774482341473196e-06, "loss": 1.2674, "step": 23400 }, { "epoch": 0.616655087371544, "grad_norm": 11.2600679397583, "learning_rate": 6.76640791072188e-06, "loss": 1.2337, "step": 23410 }, { "epoch": 0.6169185026160426, "grad_norm": 9.535510063171387, "learning_rate": 6.758335833646226e-06, "loss": 1.2993, "step": 23420 }, { "epoch": 0.6171819178605413, "grad_norm": 11.369451522827148, "learning_rate": 6.750266116121762e-06, "loss": 1.34, "step": 23430 }, { "epoch": 0.6174453331050401, "grad_norm": 11.331439018249512, "learning_rate": 6.742198764022296e-06, "loss": 1.2988, "step": 23440 }, { "epoch": 0.6177087483495388, "grad_norm": 10.294694900512695, "learning_rate": 6.734133783219909e-06, "loss": 1.2382, "step": 23450 }, { "epoch": 0.6179721635940376, "grad_norm": 6.10054874420166, "learning_rate": 6.726071179584966e-06, "loss": 1.3111, "step": 23460 }, { "epoch": 0.6182355788385363, "grad_norm": 6.449000358581543, "learning_rate": 6.718010958986091e-06, "loss": 1.3063, "step": 23470 }, { "epoch": 0.6184989940830351, "grad_norm": 8.075651168823242, "learning_rate": 6.709953127290185e-06, "loss": 1.1899, "step": 23480 }, { "epoch": 0.6187624093275338, "grad_norm": 6.170894622802734, "learning_rate": 6.701897690362405e-06, "loss": 1.3529, "step": 23490 }, { "epoch": 0.6190258245720326, "grad_norm": 5.929863929748535, "learning_rate": 6.693844654066155e-06, "loss": 1.2523, "step": 23500 }, { "epoch": 0.6192892398165313, "grad_norm": 7.320000648498535, "learning_rate": 6.685794024263108e-06, "loss": 1.2907, "step": 23510 }, { "epoch": 0.61955265506103, "grad_norm": 9.73523998260498, "learning_rate": 6.677745806813176e-06, "loss": 1.3619, "step": 23520 }, { "epoch": 0.6198160703055288, "grad_norm": 4.73565673828125, "learning_rate": 6.669700007574517e-06, "loss": 1.2824, "step": 23530 }, { "epoch": 0.6200794855500275, "grad_norm": 11.902194023132324, "learning_rate": 6.6616566324035285e-06, "loss": 1.1931, "step": 23540 }, { "epoch": 0.6203429007945263, "grad_norm": 9.387982368469238, "learning_rate": 6.653615687154846e-06, "loss": 1.3449, "step": 23550 }, { "epoch": 0.6206063160390249, "grad_norm": 10.049251556396484, "learning_rate": 6.645577177681333e-06, "loss": 1.2027, "step": 23560 }, { "epoch": 0.6208697312835237, "grad_norm": 5.822140693664551, "learning_rate": 6.637541109834077e-06, "loss": 1.3259, "step": 23570 }, { "epoch": 0.6211331465280224, "grad_norm": 9.476832389831543, "learning_rate": 6.629507489462399e-06, "loss": 1.2612, "step": 23580 }, { "epoch": 0.6213965617725212, "grad_norm": 10.403910636901855, "learning_rate": 6.62147632241383e-06, "loss": 1.1804, "step": 23590 }, { "epoch": 0.6216599770170199, "grad_norm": 9.095616340637207, "learning_rate": 6.613447614534119e-06, "loss": 1.2668, "step": 23600 }, { "epoch": 0.6219233922615186, "grad_norm": 14.865009307861328, "learning_rate": 6.605421371667219e-06, "loss": 1.2621, "step": 23610 }, { "epoch": 0.6221868075060174, "grad_norm": 8.125730514526367, "learning_rate": 6.597397599655295e-06, "loss": 1.2812, "step": 23620 }, { "epoch": 0.6224502227505161, "grad_norm": 3.3977854251861572, "learning_rate": 6.589376304338713e-06, "loss": 1.2496, "step": 23630 }, { "epoch": 0.6227136379950149, "grad_norm": 5.114952087402344, "learning_rate": 6.581357491556034e-06, "loss": 1.2761, "step": 23640 }, { "epoch": 0.6229770532395136, "grad_norm": 3.1746456623077393, "learning_rate": 6.573341167144015e-06, "loss": 1.2129, "step": 23650 }, { "epoch": 0.6232404684840124, "grad_norm": 5.782456398010254, "learning_rate": 6.565327336937596e-06, "loss": 1.2632, "step": 23660 }, { "epoch": 0.6235038837285111, "grad_norm": 9.44813346862793, "learning_rate": 6.557316006769906e-06, "loss": 1.1876, "step": 23670 }, { "epoch": 0.6237672989730099, "grad_norm": 4.75942325592041, "learning_rate": 6.549307182472253e-06, "loss": 1.1504, "step": 23680 }, { "epoch": 0.6240307142175086, "grad_norm": 24.049909591674805, "learning_rate": 6.541300869874123e-06, "loss": 1.3393, "step": 23690 }, { "epoch": 0.6242941294620072, "grad_norm": 5.5013556480407715, "learning_rate": 6.533297074803175e-06, "loss": 1.1778, "step": 23700 }, { "epoch": 0.624557544706506, "grad_norm": 4.614253997802734, "learning_rate": 6.525295803085229e-06, "loss": 1.3862, "step": 23710 }, { "epoch": 0.6248209599510047, "grad_norm": 14.510383605957031, "learning_rate": 6.517297060544272e-06, "loss": 1.2778, "step": 23720 }, { "epoch": 0.6250843751955035, "grad_norm": 11.528109550476074, "learning_rate": 6.509300853002451e-06, "loss": 1.1987, "step": 23730 }, { "epoch": 0.6253477904400022, "grad_norm": 9.3656644821167, "learning_rate": 6.501307186280066e-06, "loss": 1.2856, "step": 23740 }, { "epoch": 0.625611205684501, "grad_norm": 14.30043888092041, "learning_rate": 6.493316066195572e-06, "loss": 1.3209, "step": 23750 }, { "epoch": 0.6258746209289997, "grad_norm": 6.584784984588623, "learning_rate": 6.485327498565562e-06, "loss": 1.3194, "step": 23760 }, { "epoch": 0.6261380361734985, "grad_norm": 10.42960262298584, "learning_rate": 6.4773414892047785e-06, "loss": 1.3032, "step": 23770 }, { "epoch": 0.6264014514179972, "grad_norm": 9.03256893157959, "learning_rate": 6.469358043926097e-06, "loss": 1.2705, "step": 23780 }, { "epoch": 0.6266648666624959, "grad_norm": 5.9911885261535645, "learning_rate": 6.461377168540534e-06, "loss": 1.3129, "step": 23790 }, { "epoch": 0.6269282819069947, "grad_norm": 6.663418292999268, "learning_rate": 6.453398868857226e-06, "loss": 1.2359, "step": 23800 }, { "epoch": 0.6271916971514934, "grad_norm": 5.57323694229126, "learning_rate": 6.445423150683444e-06, "loss": 1.1724, "step": 23810 }, { "epoch": 0.6274551123959922, "grad_norm": 6.364928245544434, "learning_rate": 6.4374500198245695e-06, "loss": 1.2486, "step": 23820 }, { "epoch": 0.6277185276404909, "grad_norm": 8.837688446044922, "learning_rate": 6.429479482084109e-06, "loss": 1.306, "step": 23830 }, { "epoch": 0.6279819428849897, "grad_norm": 11.236944198608398, "learning_rate": 6.4215115432636774e-06, "loss": 1.2597, "step": 23840 }, { "epoch": 0.6282453581294883, "grad_norm": 11.27929973602295, "learning_rate": 6.4135462091629995e-06, "loss": 1.2744, "step": 23850 }, { "epoch": 0.6285087733739871, "grad_norm": 18.471120834350586, "learning_rate": 6.4055834855799095e-06, "loss": 1.1776, "step": 23860 }, { "epoch": 0.6287721886184858, "grad_norm": 6.142937660217285, "learning_rate": 6.397623378310329e-06, "loss": 1.3147, "step": 23870 }, { "epoch": 0.6290356038629845, "grad_norm": 15.222465515136719, "learning_rate": 6.389665893148281e-06, "loss": 1.148, "step": 23880 }, { "epoch": 0.6292990191074833, "grad_norm": 6.507455348968506, "learning_rate": 6.381711035885887e-06, "loss": 1.2741, "step": 23890 }, { "epoch": 0.629562434351982, "grad_norm": 10.206660270690918, "learning_rate": 6.373758812313349e-06, "loss": 1.2669, "step": 23900 }, { "epoch": 0.6298258495964808, "grad_norm": 14.38365364074707, "learning_rate": 6.365809228218951e-06, "loss": 1.3955, "step": 23910 }, { "epoch": 0.6300892648409795, "grad_norm": 7.23480749130249, "learning_rate": 6.357862289389056e-06, "loss": 1.1234, "step": 23920 }, { "epoch": 0.6303526800854783, "grad_norm": 11.297053337097168, "learning_rate": 6.3499180016081065e-06, "loss": 1.3205, "step": 23930 }, { "epoch": 0.630616095329977, "grad_norm": 6.937431335449219, "learning_rate": 6.341976370658609e-06, "loss": 1.3233, "step": 23940 }, { "epoch": 0.6308795105744757, "grad_norm": 8.801004409790039, "learning_rate": 6.3340374023211405e-06, "loss": 1.3714, "step": 23950 }, { "epoch": 0.6311429258189745, "grad_norm": 10.611309051513672, "learning_rate": 6.326101102374341e-06, "loss": 1.2661, "step": 23960 }, { "epoch": 0.6314063410634732, "grad_norm": 14.521102905273438, "learning_rate": 6.318167476594901e-06, "loss": 1.2127, "step": 23970 }, { "epoch": 0.631669756307972, "grad_norm": 4.533242225646973, "learning_rate": 6.31023653075757e-06, "loss": 1.2591, "step": 23980 }, { "epoch": 0.6319331715524706, "grad_norm": 4.9792890548706055, "learning_rate": 6.302308270635147e-06, "loss": 1.2679, "step": 23990 }, { "epoch": 0.6321965867969694, "grad_norm": 8.655173301696777, "learning_rate": 6.294382701998478e-06, "loss": 1.1929, "step": 24000 }, { "epoch": 0.6324600020414681, "grad_norm": 17.959636688232422, "learning_rate": 6.2864598306164435e-06, "loss": 1.2689, "step": 24010 }, { "epoch": 0.6327234172859669, "grad_norm": 16.479385375976562, "learning_rate": 6.278539662255969e-06, "loss": 1.314, "step": 24020 }, { "epoch": 0.6329868325304656, "grad_norm": 9.303955078125, "learning_rate": 6.270622202682002e-06, "loss": 1.3056, "step": 24030 }, { "epoch": 0.6332502477749643, "grad_norm": 4.5777411460876465, "learning_rate": 6.262707457657527e-06, "loss": 1.3237, "step": 24040 }, { "epoch": 0.6335136630194631, "grad_norm": 5.1746978759765625, "learning_rate": 6.2547954329435475e-06, "loss": 1.3178, "step": 24050 }, { "epoch": 0.6337770782639618, "grad_norm": 6.559526443481445, "learning_rate": 6.246886134299091e-06, "loss": 1.3369, "step": 24060 }, { "epoch": 0.6340404935084606, "grad_norm": 5.893487930297852, "learning_rate": 6.238979567481204e-06, "loss": 1.2553, "step": 24070 }, { "epoch": 0.6343039087529593, "grad_norm": 9.396418571472168, "learning_rate": 6.231075738244931e-06, "loss": 1.1475, "step": 24080 }, { "epoch": 0.6345673239974581, "grad_norm": 18.963613510131836, "learning_rate": 6.223174652343334e-06, "loss": 1.1815, "step": 24090 }, { "epoch": 0.6348307392419568, "grad_norm": 11.102027893066406, "learning_rate": 6.215276315527481e-06, "loss": 1.2455, "step": 24100 }, { "epoch": 0.6350941544864556, "grad_norm": 11.724854469299316, "learning_rate": 6.207380733546431e-06, "loss": 1.2805, "step": 24110 }, { "epoch": 0.6353575697309543, "grad_norm": 12.711962699890137, "learning_rate": 6.199487912147245e-06, "loss": 1.2309, "step": 24120 }, { "epoch": 0.6356209849754529, "grad_norm": 11.612534523010254, "learning_rate": 6.191597857074967e-06, "loss": 1.3245, "step": 24130 }, { "epoch": 0.6358844002199517, "grad_norm": 71.12506866455078, "learning_rate": 6.1837105740726324e-06, "loss": 1.2914, "step": 24140 }, { "epoch": 0.6361478154644504, "grad_norm": 13.837928771972656, "learning_rate": 6.175826068881258e-06, "loss": 1.1951, "step": 24150 }, { "epoch": 0.6364112307089492, "grad_norm": 11.606605529785156, "learning_rate": 6.1679443472398356e-06, "loss": 1.3062, "step": 24160 }, { "epoch": 0.6366746459534479, "grad_norm": 17.915681838989258, "learning_rate": 6.160065414885346e-06, "loss": 1.3061, "step": 24170 }, { "epoch": 0.6369380611979467, "grad_norm": 27.647504806518555, "learning_rate": 6.152189277552711e-06, "loss": 1.2559, "step": 24180 }, { "epoch": 0.6372014764424454, "grad_norm": 20.00558853149414, "learning_rate": 6.144315940974843e-06, "loss": 1.2777, "step": 24190 }, { "epoch": 0.6374648916869442, "grad_norm": 7.387813091278076, "learning_rate": 6.136445410882608e-06, "loss": 1.2114, "step": 24200 }, { "epoch": 0.6377283069314429, "grad_norm": 20.884267807006836, "learning_rate": 6.128577693004826e-06, "loss": 1.2635, "step": 24210 }, { "epoch": 0.6379917221759416, "grad_norm": 433.44134521484375, "learning_rate": 6.120712793068273e-06, "loss": 1.1912, "step": 24220 }, { "epoch": 0.6382551374204404, "grad_norm": 6.703336238861084, "learning_rate": 6.112850716797678e-06, "loss": 1.3594, "step": 24230 }, { "epoch": 0.6385185526649391, "grad_norm": 104.52894592285156, "learning_rate": 6.104991469915702e-06, "loss": 1.1724, "step": 24240 }, { "epoch": 0.6387819679094379, "grad_norm": 35.40473175048828, "learning_rate": 6.09713505814296e-06, "loss": 1.268, "step": 24250 }, { "epoch": 0.6390453831539366, "grad_norm": 4.9983439445495605, "learning_rate": 6.089281487197994e-06, "loss": 1.2303, "step": 24260 }, { "epoch": 0.6393087983984354, "grad_norm": 21.454208374023438, "learning_rate": 6.081430762797287e-06, "loss": 1.299, "step": 24270 }, { "epoch": 0.639572213642934, "grad_norm": 4.333522796630859, "learning_rate": 6.073582890655246e-06, "loss": 1.306, "step": 24280 }, { "epoch": 0.6398356288874328, "grad_norm": 44.91768264770508, "learning_rate": 6.0657378764841945e-06, "loss": 1.2388, "step": 24290 }, { "epoch": 0.6400990441319315, "grad_norm": 58.400604248046875, "learning_rate": 6.057895725994387e-06, "loss": 1.257, "step": 24300 }, { "epoch": 0.6403624593764302, "grad_norm": 13.313076972961426, "learning_rate": 6.050056444893988e-06, "loss": 1.2417, "step": 24310 }, { "epoch": 0.640625874620929, "grad_norm": 20.55388832092285, "learning_rate": 6.042220038889075e-06, "loss": 1.3466, "step": 24320 }, { "epoch": 0.6408892898654277, "grad_norm": 6.284143447875977, "learning_rate": 6.034386513683634e-06, "loss": 1.1568, "step": 24330 }, { "epoch": 0.6411527051099265, "grad_norm": 17.66733741760254, "learning_rate": 6.026555874979549e-06, "loss": 1.2753, "step": 24340 }, { "epoch": 0.6414161203544252, "grad_norm": 66.92606353759766, "learning_rate": 6.018728128476606e-06, "loss": 1.3009, "step": 24350 }, { "epoch": 0.641679535598924, "grad_norm": 58.176361083984375, "learning_rate": 6.010903279872489e-06, "loss": 1.2513, "step": 24360 }, { "epoch": 0.6419429508434227, "grad_norm": 7.7237420082092285, "learning_rate": 6.0030813348627645e-06, "loss": 1.1387, "step": 24370 }, { "epoch": 0.6422063660879215, "grad_norm": 109.33341217041016, "learning_rate": 5.995262299140901e-06, "loss": 1.1578, "step": 24380 }, { "epoch": 0.6424697813324202, "grad_norm": 6.186466217041016, "learning_rate": 5.9874461783982275e-06, "loss": 1.3084, "step": 24390 }, { "epoch": 0.6427331965769189, "grad_norm": 22.49488639831543, "learning_rate": 5.979632978323969e-06, "loss": 1.2303, "step": 24400 }, { "epoch": 0.6429966118214177, "grad_norm": 10.102582931518555, "learning_rate": 5.971822704605219e-06, "loss": 1.2891, "step": 24410 }, { "epoch": 0.6432600270659163, "grad_norm": 11.14193058013916, "learning_rate": 5.964015362926937e-06, "loss": 1.1173, "step": 24420 }, { "epoch": 0.6435234423104151, "grad_norm": 14.617886543273926, "learning_rate": 5.956210958971955e-06, "loss": 1.2933, "step": 24430 }, { "epoch": 0.6437868575549138, "grad_norm": 8.63116455078125, "learning_rate": 5.9484094984209635e-06, "loss": 1.1829, "step": 24440 }, { "epoch": 0.6440502727994126, "grad_norm": 15.653342247009277, "learning_rate": 5.9406109869525065e-06, "loss": 1.2726, "step": 24450 }, { "epoch": 0.6443136880439113, "grad_norm": 10.092753410339355, "learning_rate": 5.932815430242989e-06, "loss": 1.297, "step": 24460 }, { "epoch": 0.6445771032884101, "grad_norm": 13.414009094238281, "learning_rate": 5.925022833966658e-06, "loss": 1.309, "step": 24470 }, { "epoch": 0.6448405185329088, "grad_norm": 9.195425987243652, "learning_rate": 5.917233203795616e-06, "loss": 1.2707, "step": 24480 }, { "epoch": 0.6451039337774075, "grad_norm": 19.154054641723633, "learning_rate": 5.909446545399795e-06, "loss": 1.3124, "step": 24490 }, { "epoch": 0.6453673490219063, "grad_norm": 12.192293167114258, "learning_rate": 5.901662864446969e-06, "loss": 1.3942, "step": 24500 }, { "epoch": 0.645630764266405, "grad_norm": 5.344665050506592, "learning_rate": 5.893882166602745e-06, "loss": 1.1983, "step": 24510 }, { "epoch": 0.6458941795109038, "grad_norm": 84.73094940185547, "learning_rate": 5.886104457530559e-06, "loss": 1.4042, "step": 24520 }, { "epoch": 0.6461575947554025, "grad_norm": 24.291522979736328, "learning_rate": 5.878329742891669e-06, "loss": 1.2496, "step": 24530 }, { "epoch": 0.6464210099999013, "grad_norm": 7.835628509521484, "learning_rate": 5.870558028345159e-06, "loss": 1.3037, "step": 24540 }, { "epoch": 0.6466844252444, "grad_norm": 7.984575271606445, "learning_rate": 5.862789319547919e-06, "loss": 1.2818, "step": 24550 }, { "epoch": 0.6469478404888986, "grad_norm": 3.5392191410064697, "learning_rate": 5.855023622154663e-06, "loss": 1.2166, "step": 24560 }, { "epoch": 0.6472112557333974, "grad_norm": 14.24640941619873, "learning_rate": 5.847260941817903e-06, "loss": 1.333, "step": 24570 }, { "epoch": 0.6474746709778961, "grad_norm": 5.820329666137695, "learning_rate": 5.839501284187963e-06, "loss": 1.2871, "step": 24580 }, { "epoch": 0.6477380862223949, "grad_norm": 103.34264373779297, "learning_rate": 5.831744654912965e-06, "loss": 1.2092, "step": 24590 }, { "epoch": 0.6480015014668936, "grad_norm": 8.607818603515625, "learning_rate": 5.8239910596388224e-06, "loss": 1.2191, "step": 24600 }, { "epoch": 0.6482649167113924, "grad_norm": 10.013311386108398, "learning_rate": 5.816240504009241e-06, "loss": 1.3354, "step": 24610 }, { "epoch": 0.6485283319558911, "grad_norm": 8.941361427307129, "learning_rate": 5.808492993665715e-06, "loss": 1.2892, "step": 24620 }, { "epoch": 0.6487917472003899, "grad_norm": 10.48447036743164, "learning_rate": 5.800748534247533e-06, "loss": 1.2331, "step": 24630 }, { "epoch": 0.6490551624448886, "grad_norm": 15.687726020812988, "learning_rate": 5.79300713139174e-06, "loss": 1.3333, "step": 24640 }, { "epoch": 0.6493185776893873, "grad_norm": 11.545320510864258, "learning_rate": 5.785268790733181e-06, "loss": 1.3121, "step": 24650 }, { "epoch": 0.6495819929338861, "grad_norm": 363.5295104980469, "learning_rate": 5.777533517904451e-06, "loss": 1.2488, "step": 24660 }, { "epoch": 0.6498454081783848, "grad_norm": 7.749255180358887, "learning_rate": 5.769801318535921e-06, "loss": 1.275, "step": 24670 }, { "epoch": 0.6501088234228836, "grad_norm": 28.032316207885742, "learning_rate": 5.762072198255734e-06, "loss": 1.3136, "step": 24680 }, { "epoch": 0.6503722386673823, "grad_norm": 7.7889084815979, "learning_rate": 5.754346162689772e-06, "loss": 1.1822, "step": 24690 }, { "epoch": 0.650635653911881, "grad_norm": 8.536368370056152, "learning_rate": 5.74662321746169e-06, "loss": 1.2164, "step": 24700 }, { "epoch": 0.6508990691563797, "grad_norm": 12.466568946838379, "learning_rate": 5.738903368192884e-06, "loss": 1.1822, "step": 24710 }, { "epoch": 0.6511624844008785, "grad_norm": 6.191939830780029, "learning_rate": 5.7311866205024934e-06, "loss": 1.3404, "step": 24720 }, { "epoch": 0.6514258996453772, "grad_norm": 15.161687850952148, "learning_rate": 5.72347298000741e-06, "loss": 1.2371, "step": 24730 }, { "epoch": 0.6516893148898759, "grad_norm": 8.813163757324219, "learning_rate": 5.715762452322263e-06, "loss": 1.3787, "step": 24740 }, { "epoch": 0.6519527301343747, "grad_norm": 8.468017578125, "learning_rate": 5.7080550430594085e-06, "loss": 1.2142, "step": 24750 }, { "epoch": 0.6522161453788734, "grad_norm": 6.7396626472473145, "learning_rate": 5.70035075782893e-06, "loss": 1.3829, "step": 24760 }, { "epoch": 0.6524795606233722, "grad_norm": 8.925597190856934, "learning_rate": 5.692649602238652e-06, "loss": 1.2607, "step": 24770 }, { "epoch": 0.6527429758678709, "grad_norm": 9.139082908630371, "learning_rate": 5.684951581894106e-06, "loss": 1.1922, "step": 24780 }, { "epoch": 0.6530063911123697, "grad_norm": 22.831769943237305, "learning_rate": 5.677256702398553e-06, "loss": 1.2189, "step": 24790 }, { "epoch": 0.6532698063568684, "grad_norm": 5.69135046005249, "learning_rate": 5.6695649693529605e-06, "loss": 1.2554, "step": 24800 }, { "epoch": 0.6535332216013672, "grad_norm": 19.217451095581055, "learning_rate": 5.661876388356005e-06, "loss": 1.2776, "step": 24810 }, { "epoch": 0.6537966368458659, "grad_norm": 4.157867431640625, "learning_rate": 5.654190965004076e-06, "loss": 1.4224, "step": 24820 }, { "epoch": 0.6540600520903646, "grad_norm": 6.543351173400879, "learning_rate": 5.6465087048912534e-06, "loss": 1.218, "step": 24830 }, { "epoch": 0.6543234673348634, "grad_norm": 10.412494659423828, "learning_rate": 5.6388296136093266e-06, "loss": 1.2798, "step": 24840 }, { "epoch": 0.654586882579362, "grad_norm": 5.017049312591553, "learning_rate": 5.631153696747775e-06, "loss": 1.2827, "step": 24850 }, { "epoch": 0.6548502978238608, "grad_norm": 11.597733497619629, "learning_rate": 5.623480959893763e-06, "loss": 1.2146, "step": 24860 }, { "epoch": 0.6551137130683595, "grad_norm": 5.675497055053711, "learning_rate": 5.615811408632137e-06, "loss": 1.3387, "step": 24870 }, { "epoch": 0.6553771283128583, "grad_norm": 7.64119815826416, "learning_rate": 5.6081450485454414e-06, "loss": 1.223, "step": 24880 }, { "epoch": 0.655640543557357, "grad_norm": 8.187503814697266, "learning_rate": 5.600481885213875e-06, "loss": 1.3209, "step": 24890 }, { "epoch": 0.6559039588018558, "grad_norm": 14.66977596282959, "learning_rate": 5.5928219242153306e-06, "loss": 1.3173, "step": 24900 }, { "epoch": 0.6561673740463545, "grad_norm": 7.250889778137207, "learning_rate": 5.585165171125365e-06, "loss": 1.2257, "step": 24910 }, { "epoch": 0.6564307892908532, "grad_norm": 14.946112632751465, "learning_rate": 5.577511631517183e-06, "loss": 1.2273, "step": 24920 }, { "epoch": 0.656694204535352, "grad_norm": 28.251611709594727, "learning_rate": 5.569861310961668e-06, "loss": 1.3115, "step": 24930 }, { "epoch": 0.6569576197798507, "grad_norm": 21.359912872314453, "learning_rate": 5.562214215027363e-06, "loss": 1.2561, "step": 24940 }, { "epoch": 0.6572210350243495, "grad_norm": 14.089552879333496, "learning_rate": 5.554570349280447e-06, "loss": 1.3305, "step": 24950 }, { "epoch": 0.6574844502688482, "grad_norm": 3.4369020462036133, "learning_rate": 5.546929719284766e-06, "loss": 1.2375, "step": 24960 }, { "epoch": 0.657747865513347, "grad_norm": 159.31581115722656, "learning_rate": 5.539292330601799e-06, "loss": 1.2634, "step": 24970 }, { "epoch": 0.6580112807578457, "grad_norm": 5.253540992736816, "learning_rate": 5.531658188790663e-06, "loss": 1.2459, "step": 24980 }, { "epoch": 0.6582746960023445, "grad_norm": 9.876240730285645, "learning_rate": 5.5240272994081276e-06, "loss": 1.1287, "step": 24990 }, { "epoch": 0.6585381112468431, "grad_norm": 8.079313278198242, "learning_rate": 5.5163996680085765e-06, "loss": 1.244, "step": 25000 }, { "epoch": 0.6588015264913418, "grad_norm": 7.221560478210449, "learning_rate": 5.508775300144036e-06, "loss": 1.2519, "step": 25010 }, { "epoch": 0.6590649417358406, "grad_norm": 7.600493907928467, "learning_rate": 5.501154201364151e-06, "loss": 1.2583, "step": 25020 }, { "epoch": 0.6593283569803393, "grad_norm": 6.292059421539307, "learning_rate": 5.493536377216181e-06, "loss": 1.2936, "step": 25030 }, { "epoch": 0.6595917722248381, "grad_norm": 5.1319451332092285, "learning_rate": 5.485921833245014e-06, "loss": 1.2549, "step": 25040 }, { "epoch": 0.6598551874693368, "grad_norm": 7.589489936828613, "learning_rate": 5.478310574993146e-06, "loss": 1.2017, "step": 25050 }, { "epoch": 0.6601186027138356, "grad_norm": 11.390473365783691, "learning_rate": 5.4707026080006735e-06, "loss": 1.2164, "step": 25060 }, { "epoch": 0.6603820179583343, "grad_norm": 3.9240665435791016, "learning_rate": 5.463097937805313e-06, "loss": 1.2473, "step": 25070 }, { "epoch": 0.660645433202833, "grad_norm": 6.549188613891602, "learning_rate": 5.455496569942367e-06, "loss": 1.2763, "step": 25080 }, { "epoch": 0.6609088484473318, "grad_norm": 4.976312637329102, "learning_rate": 5.447898509944736e-06, "loss": 1.2546, "step": 25090 }, { "epoch": 0.6611722636918305, "grad_norm": 18.180500030517578, "learning_rate": 5.4403037633429175e-06, "loss": 1.2405, "step": 25100 }, { "epoch": 0.6614356789363293, "grad_norm": 8.363852500915527, "learning_rate": 5.432712335665005e-06, "loss": 1.1637, "step": 25110 }, { "epoch": 0.661699094180828, "grad_norm": 8.820781707763672, "learning_rate": 5.42512423243666e-06, "loss": 1.2357, "step": 25120 }, { "epoch": 0.6619625094253268, "grad_norm": 8.861018180847168, "learning_rate": 5.417539459181127e-06, "loss": 1.3383, "step": 25130 }, { "epoch": 0.6622259246698254, "grad_norm": 7.214784622192383, "learning_rate": 5.409958021419241e-06, "loss": 1.2333, "step": 25140 }, { "epoch": 0.6624893399143242, "grad_norm": 8.073467254638672, "learning_rate": 5.402379924669389e-06, "loss": 1.2782, "step": 25150 }, { "epoch": 0.6627527551588229, "grad_norm": 8.478598594665527, "learning_rate": 5.394805174447546e-06, "loss": 1.145, "step": 25160 }, { "epoch": 0.6630161704033216, "grad_norm": 10.074535369873047, "learning_rate": 5.387233776267238e-06, "loss": 1.3947, "step": 25170 }, { "epoch": 0.6632795856478204, "grad_norm": 7.660836696624756, "learning_rate": 5.379665735639547e-06, "loss": 1.3397, "step": 25180 }, { "epoch": 0.6635430008923191, "grad_norm": 3.5005226135253906, "learning_rate": 5.372101058073131e-06, "loss": 1.0633, "step": 25190 }, { "epoch": 0.6638064161368179, "grad_norm": 16.8642578125, "learning_rate": 5.364539749074178e-06, "loss": 1.3575, "step": 25200 }, { "epoch": 0.6640698313813166, "grad_norm": 12.858173370361328, "learning_rate": 5.356981814146438e-06, "loss": 1.1864, "step": 25210 }, { "epoch": 0.6643332466258154, "grad_norm": 5.893404006958008, "learning_rate": 5.349427258791209e-06, "loss": 1.2545, "step": 25220 }, { "epoch": 0.6645966618703141, "grad_norm": 9.388603210449219, "learning_rate": 5.3418760885073026e-06, "loss": 1.3927, "step": 25230 }, { "epoch": 0.6648600771148129, "grad_norm": 7.871059894561768, "learning_rate": 5.334328308791097e-06, "loss": 1.2672, "step": 25240 }, { "epoch": 0.6651234923593116, "grad_norm": 7.373490810394287, "learning_rate": 5.326783925136491e-06, "loss": 1.2542, "step": 25250 }, { "epoch": 0.6653869076038103, "grad_norm": 16.083660125732422, "learning_rate": 5.319242943034906e-06, "loss": 1.2704, "step": 25260 }, { "epoch": 0.665650322848309, "grad_norm": 7.1136250495910645, "learning_rate": 5.311705367975296e-06, "loss": 1.2866, "step": 25270 }, { "epoch": 0.6659137380928077, "grad_norm": 5.501832008361816, "learning_rate": 5.304171205444131e-06, "loss": 1.2343, "step": 25280 }, { "epoch": 0.6661771533373065, "grad_norm": 5.678886413574219, "learning_rate": 5.29664046092539e-06, "loss": 1.2347, "step": 25290 }, { "epoch": 0.6664405685818052, "grad_norm": 10.199728965759277, "learning_rate": 5.2891131399005834e-06, "loss": 1.2253, "step": 25300 }, { "epoch": 0.666703983826304, "grad_norm": 7.2226433753967285, "learning_rate": 5.281589247848707e-06, "loss": 1.282, "step": 25310 }, { "epoch": 0.6669673990708027, "grad_norm": 8.089058876037598, "learning_rate": 5.274068790246276e-06, "loss": 1.1734, "step": 25320 }, { "epoch": 0.6672308143153015, "grad_norm": 13.787459373474121, "learning_rate": 5.2665517725673084e-06, "loss": 1.2023, "step": 25330 }, { "epoch": 0.6674942295598002, "grad_norm": 6.604301452636719, "learning_rate": 5.259038200283297e-06, "loss": 1.2364, "step": 25340 }, { "epoch": 0.6677576448042989, "grad_norm": 22.778919219970703, "learning_rate": 5.2515280788632494e-06, "loss": 1.3174, "step": 25350 }, { "epoch": 0.6680210600487977, "grad_norm": 10.793685913085938, "learning_rate": 5.244021413773654e-06, "loss": 1.1681, "step": 25360 }, { "epoch": 0.6682844752932964, "grad_norm": 8.898537635803223, "learning_rate": 5.236518210478477e-06, "loss": 1.3077, "step": 25370 }, { "epoch": 0.6685478905377952, "grad_norm": 10.747381210327148, "learning_rate": 5.229018474439176e-06, "loss": 1.2737, "step": 25380 }, { "epoch": 0.6688113057822939, "grad_norm": 8.030888557434082, "learning_rate": 5.221522211114676e-06, "loss": 1.1539, "step": 25390 }, { "epoch": 0.6690747210267927, "grad_norm": 13.118841171264648, "learning_rate": 5.214029425961373e-06, "loss": 1.252, "step": 25400 }, { "epoch": 0.6693381362712914, "grad_norm": 6.408081531524658, "learning_rate": 5.206540124433139e-06, "loss": 1.2297, "step": 25410 }, { "epoch": 0.6696015515157901, "grad_norm": 11.952485084533691, "learning_rate": 5.199054311981313e-06, "loss": 1.2744, "step": 25420 }, { "epoch": 0.6698649667602888, "grad_norm": 13.3470458984375, "learning_rate": 5.191571994054681e-06, "loss": 1.2639, "step": 25430 }, { "epoch": 0.6701283820047875, "grad_norm": 4.612419128417969, "learning_rate": 5.1840931760994915e-06, "loss": 1.2104, "step": 25440 }, { "epoch": 0.6703917972492863, "grad_norm": 7.1024088859558105, "learning_rate": 5.176617863559453e-06, "loss": 1.2359, "step": 25450 }, { "epoch": 0.670655212493785, "grad_norm": 7.2683515548706055, "learning_rate": 5.169146061875709e-06, "loss": 1.2624, "step": 25460 }, { "epoch": 0.6709186277382838, "grad_norm": 6.228296756744385, "learning_rate": 5.16167777648686e-06, "loss": 1.2968, "step": 25470 }, { "epoch": 0.6711820429827825, "grad_norm": 5.3247270584106445, "learning_rate": 5.154213012828939e-06, "loss": 1.3243, "step": 25480 }, { "epoch": 0.6714454582272813, "grad_norm": 4.917222023010254, "learning_rate": 5.1467517763354166e-06, "loss": 1.1415, "step": 25490 }, { "epoch": 0.67170887347178, "grad_norm": 9.65062141418457, "learning_rate": 5.139294072437202e-06, "loss": 1.2499, "step": 25500 }, { "epoch": 0.6719722887162788, "grad_norm": 7.8579840660095215, "learning_rate": 5.131839906562623e-06, "loss": 1.3418, "step": 25510 }, { "epoch": 0.6722357039607775, "grad_norm": 10.6325044631958, "learning_rate": 5.12438928413744e-06, "loss": 1.1668, "step": 25520 }, { "epoch": 0.6724991192052762, "grad_norm": 17.93938446044922, "learning_rate": 5.116942210584836e-06, "loss": 1.3203, "step": 25530 }, { "epoch": 0.672762534449775, "grad_norm": 14.400424003601074, "learning_rate": 5.109498691325407e-06, "loss": 1.3064, "step": 25540 }, { "epoch": 0.6730259496942737, "grad_norm": 4.8820719718933105, "learning_rate": 5.102058731777152e-06, "loss": 1.2811, "step": 25550 }, { "epoch": 0.6732893649387724, "grad_norm": 14.821746826171875, "learning_rate": 5.094622337355499e-06, "loss": 1.2357, "step": 25560 }, { "epoch": 0.6735527801832711, "grad_norm": 5.6725263595581055, "learning_rate": 5.087189513473263e-06, "loss": 1.2265, "step": 25570 }, { "epoch": 0.6738161954277699, "grad_norm": 9.489623069763184, "learning_rate": 5.079760265540677e-06, "loss": 1.181, "step": 25580 }, { "epoch": 0.6740796106722686, "grad_norm": 12.100436210632324, "learning_rate": 5.072334598965355e-06, "loss": 1.3302, "step": 25590 }, { "epoch": 0.6743430259167673, "grad_norm": 3.5319018363952637, "learning_rate": 5.064912519152308e-06, "loss": 1.187, "step": 25600 }, { "epoch": 0.6746064411612661, "grad_norm": 7.3396453857421875, "learning_rate": 5.0574940315039485e-06, "loss": 1.2882, "step": 25610 }, { "epoch": 0.6748698564057648, "grad_norm": 9.092619895935059, "learning_rate": 5.050079141420057e-06, "loss": 1.2141, "step": 25620 }, { "epoch": 0.6751332716502636, "grad_norm": 6.60184383392334, "learning_rate": 5.042667854297807e-06, "loss": 1.245, "step": 25630 }, { "epoch": 0.6753966868947623, "grad_norm": 14.064559936523438, "learning_rate": 5.035260175531749e-06, "loss": 1.2344, "step": 25640 }, { "epoch": 0.6756601021392611, "grad_norm": 4.638503551483154, "learning_rate": 5.027856110513802e-06, "loss": 1.2411, "step": 25650 }, { "epoch": 0.6759235173837598, "grad_norm": 13.772846221923828, "learning_rate": 5.02045566463325e-06, "loss": 1.1698, "step": 25660 }, { "epoch": 0.6761869326282586, "grad_norm": 4.584293842315674, "learning_rate": 5.01305884327676e-06, "loss": 1.2528, "step": 25670 }, { "epoch": 0.6764503478727573, "grad_norm": 6.553313732147217, "learning_rate": 5.005665651828342e-06, "loss": 1.3255, "step": 25680 }, { "epoch": 0.676713763117256, "grad_norm": 5.754970550537109, "learning_rate": 4.998276095669377e-06, "loss": 1.244, "step": 25690 }, { "epoch": 0.6769771783617548, "grad_norm": 4.153213024139404, "learning_rate": 4.990890180178595e-06, "loss": 1.2931, "step": 25700 }, { "epoch": 0.6772405936062534, "grad_norm": 5.867581367492676, "learning_rate": 4.983507910732069e-06, "loss": 1.2207, "step": 25710 }, { "epoch": 0.6775040088507522, "grad_norm": 7.608166694641113, "learning_rate": 4.976129292703231e-06, "loss": 1.1769, "step": 25720 }, { "epoch": 0.6777674240952509, "grad_norm": 6.769532203674316, "learning_rate": 4.968754331462852e-06, "loss": 1.2564, "step": 25730 }, { "epoch": 0.6780308393397497, "grad_norm": 7.936191558837891, "learning_rate": 4.961383032379029e-06, "loss": 1.1354, "step": 25740 }, { "epoch": 0.6782942545842484, "grad_norm": 8.05709171295166, "learning_rate": 4.954015400817215e-06, "loss": 1.2458, "step": 25750 }, { "epoch": 0.6785576698287472, "grad_norm": 9.296412467956543, "learning_rate": 4.946651442140173e-06, "loss": 1.3451, "step": 25760 }, { "epoch": 0.6788210850732459, "grad_norm": 14.730242729187012, "learning_rate": 4.939291161708001e-06, "loss": 1.3016, "step": 25770 }, { "epoch": 0.6790845003177446, "grad_norm": 3.7692058086395264, "learning_rate": 4.931934564878124e-06, "loss": 1.1254, "step": 25780 }, { "epoch": 0.6793479155622434, "grad_norm": 8.029806137084961, "learning_rate": 4.924581657005276e-06, "loss": 1.227, "step": 25790 }, { "epoch": 0.6796113308067421, "grad_norm": 5.416784763336182, "learning_rate": 4.917232443441518e-06, "loss": 1.275, "step": 25800 }, { "epoch": 0.6798747460512409, "grad_norm": 9.923258781433105, "learning_rate": 4.9098869295362105e-06, "loss": 1.1591, "step": 25810 }, { "epoch": 0.6801381612957396, "grad_norm": 17.738723754882812, "learning_rate": 4.902545120636024e-06, "loss": 1.2427, "step": 25820 }, { "epoch": 0.6804015765402384, "grad_norm": 8.052451133728027, "learning_rate": 4.895207022084937e-06, "loss": 1.2402, "step": 25830 }, { "epoch": 0.680664991784737, "grad_norm": 6.196129322052002, "learning_rate": 4.887872639224227e-06, "loss": 1.2684, "step": 25840 }, { "epoch": 0.6809284070292358, "grad_norm": 11.164350509643555, "learning_rate": 4.880541977392462e-06, "loss": 1.2936, "step": 25850 }, { "epoch": 0.6811918222737345, "grad_norm": 17.020627975463867, "learning_rate": 4.8732150419254996e-06, "loss": 1.1448, "step": 25860 }, { "epoch": 0.6814552375182332, "grad_norm": 3.2352802753448486, "learning_rate": 4.865891838156497e-06, "loss": 1.238, "step": 25870 }, { "epoch": 0.681718652762732, "grad_norm": 8.99996280670166, "learning_rate": 4.858572371415878e-06, "loss": 1.4112, "step": 25880 }, { "epoch": 0.6819820680072307, "grad_norm": 6.6104350090026855, "learning_rate": 4.851256647031365e-06, "loss": 1.2233, "step": 25890 }, { "epoch": 0.6822454832517295, "grad_norm": 10.195953369140625, "learning_rate": 4.843944670327942e-06, "loss": 1.1534, "step": 25900 }, { "epoch": 0.6825088984962282, "grad_norm": 6.369474411010742, "learning_rate": 4.836636446627868e-06, "loss": 1.2197, "step": 25910 }, { "epoch": 0.682772313740727, "grad_norm": 10.754475593566895, "learning_rate": 4.8293319812506755e-06, "loss": 1.2241, "step": 25920 }, { "epoch": 0.6830357289852257, "grad_norm": 8.090713500976562, "learning_rate": 4.822031279513162e-06, "loss": 1.2638, "step": 25930 }, { "epoch": 0.6832991442297245, "grad_norm": 4.408010959625244, "learning_rate": 4.814734346729374e-06, "loss": 1.1602, "step": 25940 }, { "epoch": 0.6835625594742232, "grad_norm": 10.687789916992188, "learning_rate": 4.80744118821063e-06, "loss": 1.2971, "step": 25950 }, { "epoch": 0.6838259747187219, "grad_norm": 5.895262718200684, "learning_rate": 4.800151809265491e-06, "loss": 1.2278, "step": 25960 }, { "epoch": 0.6840893899632207, "grad_norm": 9.807201385498047, "learning_rate": 4.792866215199764e-06, "loss": 1.1605, "step": 25970 }, { "epoch": 0.6843528052077194, "grad_norm": 7.876060485839844, "learning_rate": 4.785584411316515e-06, "loss": 1.3415, "step": 25980 }, { "epoch": 0.6846162204522181, "grad_norm": 11.125879287719727, "learning_rate": 4.778306402916036e-06, "loss": 1.2953, "step": 25990 }, { "epoch": 0.6848796356967168, "grad_norm": 6.682347774505615, "learning_rate": 4.771032195295868e-06, "loss": 1.2004, "step": 26000 }, { "epoch": 0.6851430509412156, "grad_norm": 6.505674839019775, "learning_rate": 4.763761793750785e-06, "loss": 1.2084, "step": 26010 }, { "epoch": 0.6854064661857143, "grad_norm": 10.68661117553711, "learning_rate": 4.756495203572775e-06, "loss": 1.2385, "step": 26020 }, { "epoch": 0.6856698814302131, "grad_norm": 7.965817928314209, "learning_rate": 4.749232430051067e-06, "loss": 1.2114, "step": 26030 }, { "epoch": 0.6859332966747118, "grad_norm": 23.90220069885254, "learning_rate": 4.741973478472115e-06, "loss": 1.3824, "step": 26040 }, { "epoch": 0.6861967119192105, "grad_norm": 10.84085464477539, "learning_rate": 4.734718354119575e-06, "loss": 1.2672, "step": 26050 }, { "epoch": 0.6864601271637093, "grad_norm": 7.181231498718262, "learning_rate": 4.727467062274334e-06, "loss": 1.1759, "step": 26060 }, { "epoch": 0.686723542408208, "grad_norm": 11.738880157470703, "learning_rate": 4.7202196082144765e-06, "loss": 1.3251, "step": 26070 }, { "epoch": 0.6869869576527068, "grad_norm": 6.119255542755127, "learning_rate": 4.7129759972152985e-06, "loss": 1.2819, "step": 26080 }, { "epoch": 0.6872503728972055, "grad_norm": 8.480938911437988, "learning_rate": 4.705736234549304e-06, "loss": 1.2304, "step": 26090 }, { "epoch": 0.6875137881417043, "grad_norm": 7.583914279937744, "learning_rate": 4.698500325486184e-06, "loss": 1.1382, "step": 26100 }, { "epoch": 0.687777203386203, "grad_norm": 10.518040657043457, "learning_rate": 4.691268275292838e-06, "loss": 1.184, "step": 26110 }, { "epoch": 0.6880406186307017, "grad_norm": 10.468603134155273, "learning_rate": 4.684040089233347e-06, "loss": 1.4255, "step": 26120 }, { "epoch": 0.6883040338752004, "grad_norm": 8.233214378356934, "learning_rate": 4.676815772568978e-06, "loss": 1.2506, "step": 26130 }, { "epoch": 0.6885674491196991, "grad_norm": 9.4558744430542, "learning_rate": 4.669595330558191e-06, "loss": 1.2317, "step": 26140 }, { "epoch": 0.6888308643641979, "grad_norm": 15.419086456298828, "learning_rate": 4.662378768456623e-06, "loss": 1.281, "step": 26150 }, { "epoch": 0.6890942796086966, "grad_norm": 12.238937377929688, "learning_rate": 4.6551660915170796e-06, "loss": 1.1414, "step": 26160 }, { "epoch": 0.6893576948531954, "grad_norm": 8.086023330688477, "learning_rate": 4.6479573049895474e-06, "loss": 1.3172, "step": 26170 }, { "epoch": 0.6896211100976941, "grad_norm": 12.26124382019043, "learning_rate": 4.640752414121176e-06, "loss": 1.1635, "step": 26180 }, { "epoch": 0.6898845253421929, "grad_norm": 12.110819816589355, "learning_rate": 4.633551424156275e-06, "loss": 1.1649, "step": 26190 }, { "epoch": 0.6901479405866916, "grad_norm": 11.336727142333984, "learning_rate": 4.626354340336326e-06, "loss": 1.175, "step": 26200 }, { "epoch": 0.6904113558311903, "grad_norm": 6.492225646972656, "learning_rate": 4.619161167899964e-06, "loss": 1.2481, "step": 26210 }, { "epoch": 0.6906747710756891, "grad_norm": 5.356593608856201, "learning_rate": 4.611971912082971e-06, "loss": 1.0758, "step": 26220 }, { "epoch": 0.6909381863201878, "grad_norm": 7.507218360900879, "learning_rate": 4.604786578118278e-06, "loss": 1.2235, "step": 26230 }, { "epoch": 0.6912016015646866, "grad_norm": 11.786884307861328, "learning_rate": 4.597605171235974e-06, "loss": 1.1808, "step": 26240 }, { "epoch": 0.6914650168091853, "grad_norm": 8.464447021484375, "learning_rate": 4.590427696663271e-06, "loss": 1.1944, "step": 26250 }, { "epoch": 0.6917284320536841, "grad_norm": 6.287354946136475, "learning_rate": 4.583254159624537e-06, "loss": 1.2996, "step": 26260 }, { "epoch": 0.6919918472981827, "grad_norm": 6.0503458976745605, "learning_rate": 4.576084565341262e-06, "loss": 1.1879, "step": 26270 }, { "epoch": 0.6922552625426815, "grad_norm": 38.46751403808594, "learning_rate": 4.5689189190320646e-06, "loss": 1.2748, "step": 26280 }, { "epoch": 0.6925186777871802, "grad_norm": 9.215950965881348, "learning_rate": 4.561757225912703e-06, "loss": 1.2878, "step": 26290 }, { "epoch": 0.6927820930316789, "grad_norm": 18.78495216369629, "learning_rate": 4.554599491196042e-06, "loss": 1.2687, "step": 26300 }, { "epoch": 0.6930455082761777, "grad_norm": 111.27274322509766, "learning_rate": 4.547445720092076e-06, "loss": 1.2576, "step": 26310 }, { "epoch": 0.6933089235206764, "grad_norm": 19.893360137939453, "learning_rate": 4.540295917807919e-06, "loss": 1.2667, "step": 26320 }, { "epoch": 0.6935723387651752, "grad_norm": 10.132737159729004, "learning_rate": 4.533150089547773e-06, "loss": 1.2948, "step": 26330 }, { "epoch": 0.6938357540096739, "grad_norm": 5.417535781860352, "learning_rate": 4.5260082405129705e-06, "loss": 1.3479, "step": 26340 }, { "epoch": 0.6940991692541727, "grad_norm": 14.898469924926758, "learning_rate": 4.5188703759019425e-06, "loss": 1.262, "step": 26350 }, { "epoch": 0.6943625844986714, "grad_norm": 6.699496269226074, "learning_rate": 4.511736500910211e-06, "loss": 1.2263, "step": 26360 }, { "epoch": 0.6946259997431702, "grad_norm": 9.356054306030273, "learning_rate": 4.504606620730405e-06, "loss": 1.2499, "step": 26370 }, { "epoch": 0.6948894149876689, "grad_norm": 4.419393062591553, "learning_rate": 4.497480740552241e-06, "loss": 1.2819, "step": 26380 }, { "epoch": 0.6951528302321676, "grad_norm": 11.327170372009277, "learning_rate": 4.490358865562515e-06, "loss": 1.2072, "step": 26390 }, { "epoch": 0.6954162454766664, "grad_norm": 13.916216850280762, "learning_rate": 4.483241000945128e-06, "loss": 1.2838, "step": 26400 }, { "epoch": 0.695679660721165, "grad_norm": 8.555780410766602, "learning_rate": 4.4761271518810415e-06, "loss": 1.2076, "step": 26410 }, { "epoch": 0.6959430759656638, "grad_norm": 3.542222738265991, "learning_rate": 4.469017323548305e-06, "loss": 1.1717, "step": 26420 }, { "epoch": 0.6962064912101625, "grad_norm": 10.457159996032715, "learning_rate": 4.461911521122047e-06, "loss": 1.1913, "step": 26430 }, { "epoch": 0.6964699064546613, "grad_norm": 6.690451622009277, "learning_rate": 4.454809749774449e-06, "loss": 1.2599, "step": 26440 }, { "epoch": 0.69673332169916, "grad_norm": 14.857367515563965, "learning_rate": 4.447712014674769e-06, "loss": 1.3214, "step": 26450 }, { "epoch": 0.6969967369436588, "grad_norm": 5.9561662673950195, "learning_rate": 4.440618320989329e-06, "loss": 1.2782, "step": 26460 }, { "epoch": 0.6972601521881575, "grad_norm": 7.599686145782471, "learning_rate": 4.4335286738815005e-06, "loss": 1.2418, "step": 26470 }, { "epoch": 0.6975235674326562, "grad_norm": 13.278491020202637, "learning_rate": 4.426443078511722e-06, "loss": 1.1854, "step": 26480 }, { "epoch": 0.697786982677155, "grad_norm": 15.866960525512695, "learning_rate": 4.419361540037471e-06, "loss": 1.2792, "step": 26490 }, { "epoch": 0.6980503979216537, "grad_norm": 12.472566604614258, "learning_rate": 4.4122840636132734e-06, "loss": 1.2913, "step": 26500 }, { "epoch": 0.6983138131661525, "grad_norm": 12.251481056213379, "learning_rate": 4.405210654390705e-06, "loss": 1.1481, "step": 26510 }, { "epoch": 0.6985772284106512, "grad_norm": 3.742027521133423, "learning_rate": 4.398141317518382e-06, "loss": 1.186, "step": 26520 }, { "epoch": 0.69884064365515, "grad_norm": 10.487710952758789, "learning_rate": 4.391076058141949e-06, "loss": 1.1703, "step": 26530 }, { "epoch": 0.6991040588996487, "grad_norm": 12.822783470153809, "learning_rate": 4.38401488140408e-06, "loss": 1.1961, "step": 26540 }, { "epoch": 0.6993674741441475, "grad_norm": 14.203514099121094, "learning_rate": 4.376957792444494e-06, "loss": 1.1991, "step": 26550 }, { "epoch": 0.6996308893886461, "grad_norm": 6.6220173835754395, "learning_rate": 4.369904796399914e-06, "loss": 1.2545, "step": 26560 }, { "epoch": 0.6998943046331448, "grad_norm": 7.648800849914551, "learning_rate": 4.362855898404102e-06, "loss": 1.0459, "step": 26570 }, { "epoch": 0.7001577198776436, "grad_norm": 10.704607009887695, "learning_rate": 4.3558111035878206e-06, "loss": 1.2009, "step": 26580 }, { "epoch": 0.7004211351221423, "grad_norm": 15.9629487991333, "learning_rate": 4.348770417078862e-06, "loss": 1.2185, "step": 26590 }, { "epoch": 0.7006845503666411, "grad_norm": 9.63800048828125, "learning_rate": 4.341733844002015e-06, "loss": 1.2104, "step": 26600 }, { "epoch": 0.7009479656111398, "grad_norm": 7.120676040649414, "learning_rate": 4.334701389479078e-06, "loss": 1.1035, "step": 26610 }, { "epoch": 0.7012113808556386, "grad_norm": 8.560417175292969, "learning_rate": 4.327673058628854e-06, "loss": 1.1586, "step": 26620 }, { "epoch": 0.7014747961001373, "grad_norm": 16.64198875427246, "learning_rate": 4.320648856567148e-06, "loss": 1.3308, "step": 26630 }, { "epoch": 0.701738211344636, "grad_norm": 11.895329475402832, "learning_rate": 4.3136287884067515e-06, "loss": 1.209, "step": 26640 }, { "epoch": 0.7020016265891348, "grad_norm": 10.462653160095215, "learning_rate": 4.306612859257446e-06, "loss": 1.1209, "step": 26650 }, { "epoch": 0.7022650418336335, "grad_norm": 6.716972827911377, "learning_rate": 4.299601074226014e-06, "loss": 1.2627, "step": 26660 }, { "epoch": 0.7025284570781323, "grad_norm": 13.060144424438477, "learning_rate": 4.2925934384162035e-06, "loss": 1.3103, "step": 26670 }, { "epoch": 0.702791872322631, "grad_norm": 13.93526840209961, "learning_rate": 4.285589956928759e-06, "loss": 1.2486, "step": 26680 }, { "epoch": 0.7030552875671298, "grad_norm": 8.40973949432373, "learning_rate": 4.278590634861388e-06, "loss": 1.2954, "step": 26690 }, { "epoch": 0.7033187028116284, "grad_norm": 7.413169860839844, "learning_rate": 4.271595477308775e-06, "loss": 1.0858, "step": 26700 }, { "epoch": 0.7035821180561272, "grad_norm": 3.585737705230713, "learning_rate": 4.264604489362576e-06, "loss": 1.092, "step": 26710 }, { "epoch": 0.7038455333006259, "grad_norm": 10.374383926391602, "learning_rate": 4.257617676111413e-06, "loss": 1.3224, "step": 26720 }, { "epoch": 0.7041089485451246, "grad_norm": 14.501031875610352, "learning_rate": 4.25063504264086e-06, "loss": 1.3247, "step": 26730 }, { "epoch": 0.7043723637896234, "grad_norm": 17.889558792114258, "learning_rate": 4.243656594033464e-06, "loss": 1.3065, "step": 26740 }, { "epoch": 0.7046357790341221, "grad_norm": 10.550854682922363, "learning_rate": 4.2366823353687105e-06, "loss": 1.2566, "step": 26750 }, { "epoch": 0.7048991942786209, "grad_norm": 7.20853328704834, "learning_rate": 4.229712271723038e-06, "loss": 1.2376, "step": 26760 }, { "epoch": 0.7051626095231196, "grad_norm": 5.194488525390625, "learning_rate": 4.222746408169844e-06, "loss": 1.172, "step": 26770 }, { "epoch": 0.7054260247676184, "grad_norm": 8.624528884887695, "learning_rate": 4.2157847497794516e-06, "loss": 1.2937, "step": 26780 }, { "epoch": 0.7056894400121171, "grad_norm": 5.13702917098999, "learning_rate": 4.208827301619135e-06, "loss": 1.2639, "step": 26790 }, { "epoch": 0.7059528552566159, "grad_norm": 16.630889892578125, "learning_rate": 4.201874068753108e-06, "loss": 1.2405, "step": 26800 }, { "epoch": 0.7062162705011146, "grad_norm": 6.863016605377197, "learning_rate": 4.1949250562424925e-06, "loss": 1.3355, "step": 26810 }, { "epoch": 0.7064796857456133, "grad_norm": 4.846906661987305, "learning_rate": 4.187980269145363e-06, "loss": 1.3134, "step": 26820 }, { "epoch": 0.7067431009901121, "grad_norm": 12.716536521911621, "learning_rate": 4.181039712516713e-06, "loss": 1.2702, "step": 26830 }, { "epoch": 0.7070065162346107, "grad_norm": 7.544820308685303, "learning_rate": 4.174103391408446e-06, "loss": 1.1471, "step": 26840 }, { "epoch": 0.7072699314791095, "grad_norm": 10.287220001220703, "learning_rate": 4.167171310869398e-06, "loss": 1.142, "step": 26850 }, { "epoch": 0.7075333467236082, "grad_norm": 9.600506782531738, "learning_rate": 4.160243475945305e-06, "loss": 1.1947, "step": 26860 }, { "epoch": 0.707796761968107, "grad_norm": 19.258453369140625, "learning_rate": 4.1533198916788155e-06, "loss": 1.2426, "step": 26870 }, { "epoch": 0.7080601772126057, "grad_norm": 9.592848777770996, "learning_rate": 4.146400563109494e-06, "loss": 1.2521, "step": 26880 }, { "epoch": 0.7083235924571045, "grad_norm": 7.419201850891113, "learning_rate": 4.139485495273792e-06, "loss": 1.043, "step": 26890 }, { "epoch": 0.7085870077016032, "grad_norm": 11.28185749053955, "learning_rate": 4.132574693205076e-06, "loss": 1.249, "step": 26900 }, { "epoch": 0.7088504229461019, "grad_norm": 3.7899718284606934, "learning_rate": 4.125668161933594e-06, "loss": 1.1904, "step": 26910 }, { "epoch": 0.7091138381906007, "grad_norm": 9.760143280029297, "learning_rate": 4.1187659064864875e-06, "loss": 1.2099, "step": 26920 }, { "epoch": 0.7093772534350994, "grad_norm": 11.602234840393066, "learning_rate": 4.1118679318877915e-06, "loss": 1.1547, "step": 26930 }, { "epoch": 0.7096406686795982, "grad_norm": 8.613829612731934, "learning_rate": 4.104974243158427e-06, "loss": 1.2753, "step": 26940 }, { "epoch": 0.7099040839240969, "grad_norm": 6.963531970977783, "learning_rate": 4.098084845316187e-06, "loss": 1.1203, "step": 26950 }, { "epoch": 0.7101674991685957, "grad_norm": 12.72823715209961, "learning_rate": 4.091199743375741e-06, "loss": 1.1829, "step": 26960 }, { "epoch": 0.7104309144130944, "grad_norm": 11.918359756469727, "learning_rate": 4.084318942348642e-06, "loss": 1.2105, "step": 26970 }, { "epoch": 0.7106943296575932, "grad_norm": 8.371241569519043, "learning_rate": 4.0774424472433e-06, "loss": 1.2699, "step": 26980 }, { "epoch": 0.7109577449020918, "grad_norm": 16.880403518676758, "learning_rate": 4.070570263064998e-06, "loss": 1.2558, "step": 26990 }, { "epoch": 0.7112211601465905, "grad_norm": 21.592655181884766, "learning_rate": 4.0637023948158864e-06, "loss": 1.1944, "step": 27000 }, { "epoch": 0.7114845753910893, "grad_norm": 6.380605697631836, "learning_rate": 4.05683884749496e-06, "loss": 1.2168, "step": 27010 }, { "epoch": 0.711747990635588, "grad_norm": 4.519338607788086, "learning_rate": 4.049979626098076e-06, "loss": 1.2473, "step": 27020 }, { "epoch": 0.7120114058800868, "grad_norm": 5.7306108474731445, "learning_rate": 4.0431247356179494e-06, "loss": 1.1218, "step": 27030 }, { "epoch": 0.7122748211245855, "grad_norm": 8.584068298339844, "learning_rate": 4.036274181044127e-06, "loss": 1.2691, "step": 27040 }, { "epoch": 0.7125382363690843, "grad_norm": 16.393827438354492, "learning_rate": 4.029427967363019e-06, "loss": 1.382, "step": 27050 }, { "epoch": 0.712801651613583, "grad_norm": 17.976877212524414, "learning_rate": 4.022586099557859e-06, "loss": 1.2275, "step": 27060 }, { "epoch": 0.7130650668580818, "grad_norm": 8.9276704788208, "learning_rate": 4.0157485826087225e-06, "loss": 1.0862, "step": 27070 }, { "epoch": 0.7133284821025805, "grad_norm": 7.40914249420166, "learning_rate": 4.0089154214925266e-06, "loss": 1.2812, "step": 27080 }, { "epoch": 0.7135918973470792, "grad_norm": 5.1152567863464355, "learning_rate": 4.002086621183002e-06, "loss": 1.3035, "step": 27090 }, { "epoch": 0.713855312591578, "grad_norm": 8.336875915527344, "learning_rate": 3.99526218665072e-06, "loss": 1.3142, "step": 27100 }, { "epoch": 0.7141187278360767, "grad_norm": 12.616101264953613, "learning_rate": 3.988442122863076e-06, "loss": 1.2744, "step": 27110 }, { "epoch": 0.7143821430805755, "grad_norm": 5.930954456329346, "learning_rate": 3.981626434784258e-06, "loss": 1.1754, "step": 27120 }, { "epoch": 0.7146455583250741, "grad_norm": 7.624250888824463, "learning_rate": 3.974815127375297e-06, "loss": 1.2693, "step": 27130 }, { "epoch": 0.714908973569573, "grad_norm": 5.490057468414307, "learning_rate": 3.968008205594032e-06, "loss": 1.3082, "step": 27140 }, { "epoch": 0.7151723888140716, "grad_norm": 19.924509048461914, "learning_rate": 3.961205674395091e-06, "loss": 1.2517, "step": 27150 }, { "epoch": 0.7154358040585704, "grad_norm": 9.019606590270996, "learning_rate": 3.95440753872993e-06, "loss": 1.2262, "step": 27160 }, { "epoch": 0.7156992193030691, "grad_norm": 8.202618598937988, "learning_rate": 3.947613803546787e-06, "loss": 1.1618, "step": 27170 }, { "epoch": 0.7159626345475678, "grad_norm": 12.091175079345703, "learning_rate": 3.940824473790703e-06, "loss": 1.262, "step": 27180 }, { "epoch": 0.7162260497920666, "grad_norm": 10.056657791137695, "learning_rate": 3.934039554403521e-06, "loss": 1.2811, "step": 27190 }, { "epoch": 0.7164894650365653, "grad_norm": 6.162553310394287, "learning_rate": 3.927259050323857e-06, "loss": 1.143, "step": 27200 }, { "epoch": 0.7167528802810641, "grad_norm": 8.208976745605469, "learning_rate": 3.9204829664871314e-06, "loss": 1.2152, "step": 27210 }, { "epoch": 0.7170162955255628, "grad_norm": 6.672464847564697, "learning_rate": 3.913711307825531e-06, "loss": 1.2598, "step": 27220 }, { "epoch": 0.7172797107700616, "grad_norm": 5.915503978729248, "learning_rate": 3.9069440792680336e-06, "loss": 1.3071, "step": 27230 }, { "epoch": 0.7175431260145603, "grad_norm": 16.773019790649414, "learning_rate": 3.900181285740384e-06, "loss": 1.1622, "step": 27240 }, { "epoch": 0.717806541259059, "grad_norm": 7.119536399841309, "learning_rate": 3.8934229321651075e-06, "loss": 1.26, "step": 27250 }, { "epoch": 0.7180699565035578, "grad_norm": 4.531097412109375, "learning_rate": 3.886669023461486e-06, "loss": 1.3047, "step": 27260 }, { "epoch": 0.7183333717480564, "grad_norm": 8.247146606445312, "learning_rate": 3.87991956454558e-06, "loss": 1.1749, "step": 27270 }, { "epoch": 0.7185967869925552, "grad_norm": 7.823760509490967, "learning_rate": 3.873174560330199e-06, "loss": 1.1969, "step": 27280 }, { "epoch": 0.7188602022370539, "grad_norm": 8.736512184143066, "learning_rate": 3.866434015724912e-06, "loss": 1.2317, "step": 27290 }, { "epoch": 0.7191236174815527, "grad_norm": 32.64698791503906, "learning_rate": 3.85969793563605e-06, "loss": 1.1847, "step": 27300 }, { "epoch": 0.7193870327260514, "grad_norm": 9.443113327026367, "learning_rate": 3.8529663249666895e-06, "loss": 1.2246, "step": 27310 }, { "epoch": 0.7196504479705502, "grad_norm": 16.1470890045166, "learning_rate": 3.846239188616651e-06, "loss": 1.2331, "step": 27320 }, { "epoch": 0.7199138632150489, "grad_norm": 5.221819877624512, "learning_rate": 3.839516531482496e-06, "loss": 1.1908, "step": 27330 }, { "epoch": 0.7201772784595476, "grad_norm": 8.22486686706543, "learning_rate": 3.832798358457539e-06, "loss": 1.2274, "step": 27340 }, { "epoch": 0.7204406937040464, "grad_norm": 9.44852352142334, "learning_rate": 3.826084674431815e-06, "loss": 1.3021, "step": 27350 }, { "epoch": 0.7207041089485451, "grad_norm": 10.664459228515625, "learning_rate": 3.8193754842921025e-06, "loss": 1.243, "step": 27360 }, { "epoch": 0.7209675241930439, "grad_norm": 8.17872428894043, "learning_rate": 3.812670792921904e-06, "loss": 1.1209, "step": 27370 }, { "epoch": 0.7212309394375426, "grad_norm": 12.051916122436523, "learning_rate": 3.8059706052014436e-06, "loss": 1.3766, "step": 27380 }, { "epoch": 0.7214943546820414, "grad_norm": 11.754029273986816, "learning_rate": 3.799274926007679e-06, "loss": 1.1954, "step": 27390 }, { "epoch": 0.7217577699265401, "grad_norm": 7.017317771911621, "learning_rate": 3.7925837602142712e-06, "loss": 1.1508, "step": 27400 }, { "epoch": 0.7220211851710389, "grad_norm": 16.88252830505371, "learning_rate": 3.785897112691609e-06, "loss": 1.2186, "step": 27410 }, { "epoch": 0.7222846004155375, "grad_norm": 4.564225673675537, "learning_rate": 3.779214988306794e-06, "loss": 1.3978, "step": 27420 }, { "epoch": 0.7225480156600362, "grad_norm": 7.190403461456299, "learning_rate": 3.772537391923613e-06, "loss": 1.3138, "step": 27430 }, { "epoch": 0.722811430904535, "grad_norm": 4.266348361968994, "learning_rate": 3.765864328402582e-06, "loss": 1.2387, "step": 27440 }, { "epoch": 0.7230748461490337, "grad_norm": 13.613656997680664, "learning_rate": 3.75919580260091e-06, "loss": 1.1954, "step": 27450 }, { "epoch": 0.7233382613935325, "grad_norm": 7.081454277038574, "learning_rate": 3.7525318193724967e-06, "loss": 1.1263, "step": 27460 }, { "epoch": 0.7236016766380312, "grad_norm": 17.199438095092773, "learning_rate": 3.7458723835679444e-06, "loss": 1.2437, "step": 27470 }, { "epoch": 0.72386509188253, "grad_norm": 90.51868438720703, "learning_rate": 3.7392175000345387e-06, "loss": 1.3021, "step": 27480 }, { "epoch": 0.7241285071270287, "grad_norm": 11.227408409118652, "learning_rate": 3.7325671736162515e-06, "loss": 1.2069, "step": 27490 }, { "epoch": 0.7243919223715275, "grad_norm": 9.376782417297363, "learning_rate": 3.7259214091537422e-06, "loss": 1.1013, "step": 27500 }, { "epoch": 0.7246553376160262, "grad_norm": 8.863065719604492, "learning_rate": 3.7192802114843517e-06, "loss": 1.1985, "step": 27510 }, { "epoch": 0.7249187528605249, "grad_norm": 5.11561393737793, "learning_rate": 3.7126435854420863e-06, "loss": 1.2011, "step": 27520 }, { "epoch": 0.7251821681050237, "grad_norm": 8.133282661437988, "learning_rate": 3.7060115358576363e-06, "loss": 1.2471, "step": 27530 }, { "epoch": 0.7254455833495224, "grad_norm": 4.464186668395996, "learning_rate": 3.699384067558355e-06, "loss": 1.1695, "step": 27540 }, { "epoch": 0.7257089985940212, "grad_norm": 12.826943397521973, "learning_rate": 3.6927611853682554e-06, "loss": 1.156, "step": 27550 }, { "epoch": 0.7259724138385198, "grad_norm": 11.135642051696777, "learning_rate": 3.6861428941080277e-06, "loss": 1.1632, "step": 27560 }, { "epoch": 0.7262358290830186, "grad_norm": 6.3439764976501465, "learning_rate": 3.679529198595004e-06, "loss": 1.2144, "step": 27570 }, { "epoch": 0.7264992443275173, "grad_norm": 8.332908630371094, "learning_rate": 3.672920103643185e-06, "loss": 1.2348, "step": 27580 }, { "epoch": 0.7267626595720161, "grad_norm": 4.791558742523193, "learning_rate": 3.6663156140632137e-06, "loss": 1.2726, "step": 27590 }, { "epoch": 0.7270260748165148, "grad_norm": 10.161201477050781, "learning_rate": 3.65971573466238e-06, "loss": 1.2408, "step": 27600 }, { "epoch": 0.7272894900610135, "grad_norm": 6.7998857498168945, "learning_rate": 3.6531204702446245e-06, "loss": 1.0737, "step": 27610 }, { "epoch": 0.7275529053055123, "grad_norm": 7.088876724243164, "learning_rate": 3.646529825610531e-06, "loss": 1.1997, "step": 27620 }, { "epoch": 0.727816320550011, "grad_norm": 8.473071098327637, "learning_rate": 3.63994380555731e-06, "loss": 1.2839, "step": 27630 }, { "epoch": 0.7280797357945098, "grad_norm": 6.355103969573975, "learning_rate": 3.6333624148788094e-06, "loss": 1.2391, "step": 27640 }, { "epoch": 0.7283431510390085, "grad_norm": 17.870521545410156, "learning_rate": 3.626785658365517e-06, "loss": 1.2885, "step": 27650 }, { "epoch": 0.7286065662835073, "grad_norm": 8.336323738098145, "learning_rate": 3.6202135408045314e-06, "loss": 1.3269, "step": 27660 }, { "epoch": 0.728869981528006, "grad_norm": 8.01080322265625, "learning_rate": 3.6136460669795903e-06, "loss": 1.2479, "step": 27670 }, { "epoch": 0.7291333967725048, "grad_norm": 15.984819412231445, "learning_rate": 3.6070832416710375e-06, "loss": 1.3053, "step": 27680 }, { "epoch": 0.7293968120170035, "grad_norm": 4.790171146392822, "learning_rate": 3.600525069655847e-06, "loss": 1.1353, "step": 27690 }, { "epoch": 0.7296602272615021, "grad_norm": 19.164060592651367, "learning_rate": 3.593971555707595e-06, "loss": 1.3046, "step": 27700 }, { "epoch": 0.729923642506001, "grad_norm": 4.733276844024658, "learning_rate": 3.5874227045964683e-06, "loss": 1.1326, "step": 27710 }, { "epoch": 0.7301870577504996, "grad_norm": 14.153071403503418, "learning_rate": 3.580878521089264e-06, "loss": 1.245, "step": 27720 }, { "epoch": 0.7304504729949984, "grad_norm": 69.43119812011719, "learning_rate": 3.5743390099493835e-06, "loss": 1.1434, "step": 27730 }, { "epoch": 0.7307138882394971, "grad_norm": 12.831033706665039, "learning_rate": 3.567804175936821e-06, "loss": 1.3792, "step": 27740 }, { "epoch": 0.7309773034839959, "grad_norm": 11.46628475189209, "learning_rate": 3.5612740238081643e-06, "loss": 1.2688, "step": 27750 }, { "epoch": 0.7312407187284946, "grad_norm": 6.316324234008789, "learning_rate": 3.5547485583166076e-06, "loss": 1.1629, "step": 27760 }, { "epoch": 0.7315041339729933, "grad_norm": 7.9033284187316895, "learning_rate": 3.548227784211914e-06, "loss": 1.1249, "step": 27770 }, { "epoch": 0.7317675492174921, "grad_norm": 17.419740676879883, "learning_rate": 3.5417117062404483e-06, "loss": 1.2182, "step": 27780 }, { "epoch": 0.7320309644619908, "grad_norm": 12.671055793762207, "learning_rate": 3.5352003291451555e-06, "loss": 1.3151, "step": 27790 }, { "epoch": 0.7322943797064896, "grad_norm": 11.4597749710083, "learning_rate": 3.5286936576655404e-06, "loss": 1.3401, "step": 27800 }, { "epoch": 0.7325577949509883, "grad_norm": 8.529892921447754, "learning_rate": 3.5221916965377057e-06, "loss": 1.3419, "step": 27810 }, { "epoch": 0.7328212101954871, "grad_norm": 6.525213718414307, "learning_rate": 3.515694450494318e-06, "loss": 1.2133, "step": 27820 }, { "epoch": 0.7330846254399858, "grad_norm": 6.203141689300537, "learning_rate": 3.5092019242646025e-06, "loss": 1.281, "step": 27830 }, { "epoch": 0.7333480406844846, "grad_norm": 4.599928855895996, "learning_rate": 3.5027141225743656e-06, "loss": 1.1847, "step": 27840 }, { "epoch": 0.7336114559289832, "grad_norm": 6.271599769592285, "learning_rate": 3.4962310501459606e-06, "loss": 1.1693, "step": 27850 }, { "epoch": 0.7338748711734819, "grad_norm": 7.340092182159424, "learning_rate": 3.489752711698301e-06, "loss": 1.2112, "step": 27860 }, { "epoch": 0.7341382864179807, "grad_norm": 7.582787036895752, "learning_rate": 3.4832791119468624e-06, "loss": 1.2962, "step": 27870 }, { "epoch": 0.7344017016624794, "grad_norm": 8.689005851745605, "learning_rate": 3.476810255603662e-06, "loss": 1.2808, "step": 27880 }, { "epoch": 0.7346651169069782, "grad_norm": 14.772737503051758, "learning_rate": 3.470346147377269e-06, "loss": 1.3084, "step": 27890 }, { "epoch": 0.7349285321514769, "grad_norm": 8.314057350158691, "learning_rate": 3.463886791972804e-06, "loss": 1.1936, "step": 27900 }, { "epoch": 0.7351919473959757, "grad_norm": 6.526559352874756, "learning_rate": 3.4574321940919054e-06, "loss": 1.2585, "step": 27910 }, { "epoch": 0.7354553626404744, "grad_norm": 9.90458869934082, "learning_rate": 3.4509823584327706e-06, "loss": 1.1911, "step": 27920 }, { "epoch": 0.7357187778849732, "grad_norm": 7.201440811157227, "learning_rate": 3.4445372896901276e-06, "loss": 1.1673, "step": 27930 }, { "epoch": 0.7359821931294719, "grad_norm": 13.639850616455078, "learning_rate": 3.438096992555221e-06, "loss": 1.3175, "step": 27940 }, { "epoch": 0.7362456083739706, "grad_norm": 7.421821117401123, "learning_rate": 3.4316614717158405e-06, "loss": 1.3027, "step": 27950 }, { "epoch": 0.7365090236184694, "grad_norm": 9.729570388793945, "learning_rate": 3.425230731856286e-06, "loss": 1.2021, "step": 27960 }, { "epoch": 0.7367724388629681, "grad_norm": 6.487699031829834, "learning_rate": 3.4188047776573785e-06, "loss": 1.1681, "step": 27970 }, { "epoch": 0.7370358541074669, "grad_norm": 7.422901153564453, "learning_rate": 3.412383613796465e-06, "loss": 1.2198, "step": 27980 }, { "epoch": 0.7372992693519655, "grad_norm": 5.198636054992676, "learning_rate": 3.4059672449473922e-06, "loss": 1.2278, "step": 27990 }, { "epoch": 0.7375626845964643, "grad_norm": 7.601644992828369, "learning_rate": 3.3995556757805327e-06, "loss": 1.1354, "step": 28000 }, { "epoch": 0.737826099840963, "grad_norm": 9.192791938781738, "learning_rate": 3.393148910962751e-06, "loss": 1.2167, "step": 28010 }, { "epoch": 0.7380895150854618, "grad_norm": 15.241044998168945, "learning_rate": 3.3867469551574197e-06, "loss": 1.1059, "step": 28020 }, { "epoch": 0.7383529303299605, "grad_norm": 239.47024536132812, "learning_rate": 3.3803498130244105e-06, "loss": 1.2194, "step": 28030 }, { "epoch": 0.7386163455744592, "grad_norm": 5.6883463859558105, "learning_rate": 3.3739574892201e-06, "loss": 1.1658, "step": 28040 }, { "epoch": 0.738879760818958, "grad_norm": 9.623893737792969, "learning_rate": 3.3675699883973433e-06, "loss": 1.2525, "step": 28050 }, { "epoch": 0.7391431760634567, "grad_norm": 19.118925094604492, "learning_rate": 3.361187315205491e-06, "loss": 1.2901, "step": 28060 }, { "epoch": 0.7394065913079555, "grad_norm": 5.679615020751953, "learning_rate": 3.354809474290385e-06, "loss": 1.2783, "step": 28070 }, { "epoch": 0.7396700065524542, "grad_norm": 11.751688957214355, "learning_rate": 3.3484364702943397e-06, "loss": 1.1817, "step": 28080 }, { "epoch": 0.739933421796953, "grad_norm": 8.829705238342285, "learning_rate": 3.342068307856157e-06, "loss": 1.1387, "step": 28090 }, { "epoch": 0.7401968370414517, "grad_norm": 10.12314510345459, "learning_rate": 3.335704991611115e-06, "loss": 1.1294, "step": 28100 }, { "epoch": 0.7404602522859505, "grad_norm": 17.57094955444336, "learning_rate": 3.32934652619096e-06, "loss": 1.2984, "step": 28110 }, { "epoch": 0.7407236675304492, "grad_norm": 4.487314224243164, "learning_rate": 3.322992916223903e-06, "loss": 1.2661, "step": 28120 }, { "epoch": 0.7409870827749478, "grad_norm": 6.947227954864502, "learning_rate": 3.316644166334634e-06, "loss": 1.2303, "step": 28130 }, { "epoch": 0.7412504980194466, "grad_norm": 7.996845245361328, "learning_rate": 3.310300281144293e-06, "loss": 1.2491, "step": 28140 }, { "epoch": 0.7415139132639453, "grad_norm": 12.407132148742676, "learning_rate": 3.303961265270489e-06, "loss": 1.1199, "step": 28150 }, { "epoch": 0.7417773285084441, "grad_norm": 11.668746948242188, "learning_rate": 3.2976271233272794e-06, "loss": 1.0999, "step": 28160 }, { "epoch": 0.7420407437529428, "grad_norm": 16.312633514404297, "learning_rate": 3.2912978599251723e-06, "loss": 1.213, "step": 28170 }, { "epoch": 0.7423041589974416, "grad_norm": 9.091997146606445, "learning_rate": 3.284973479671135e-06, "loss": 1.1734, "step": 28180 }, { "epoch": 0.7425675742419403, "grad_norm": 10.747648239135742, "learning_rate": 3.2786539871685707e-06, "loss": 1.1941, "step": 28190 }, { "epoch": 0.7428309894864391, "grad_norm": 13.304875373840332, "learning_rate": 3.2723393870173293e-06, "loss": 1.1475, "step": 28200 }, { "epoch": 0.7430944047309378, "grad_norm": 23.057933807373047, "learning_rate": 3.266029683813705e-06, "loss": 1.2475, "step": 28210 }, { "epoch": 0.7433578199754365, "grad_norm": 17.446489334106445, "learning_rate": 3.2597248821504103e-06, "loss": 1.209, "step": 28220 }, { "epoch": 0.7436212352199353, "grad_norm": 12.732297897338867, "learning_rate": 3.2534249866166055e-06, "loss": 1.2108, "step": 28230 }, { "epoch": 0.743884650464434, "grad_norm": 10.92783260345459, "learning_rate": 3.24713000179788e-06, "loss": 1.0834, "step": 28240 }, { "epoch": 0.7441480657089328, "grad_norm": 6.13063383102417, "learning_rate": 3.2408399322762386e-06, "loss": 1.1568, "step": 28250 }, { "epoch": 0.7444114809534315, "grad_norm": 5.784814834594727, "learning_rate": 3.2345547826301184e-06, "loss": 1.2044, "step": 28260 }, { "epoch": 0.7446748961979303, "grad_norm": 9.758633613586426, "learning_rate": 3.2282745574343687e-06, "loss": 1.2105, "step": 28270 }, { "epoch": 0.744938311442429, "grad_norm": 5.850472927093506, "learning_rate": 3.2219992612602534e-06, "loss": 1.1666, "step": 28280 }, { "epoch": 0.7452017266869276, "grad_norm": 12.295234680175781, "learning_rate": 3.2157288986754575e-06, "loss": 1.1458, "step": 28290 }, { "epoch": 0.7454651419314264, "grad_norm": 14.055192947387695, "learning_rate": 3.2094634742440643e-06, "loss": 1.0723, "step": 28300 }, { "epoch": 0.7457285571759251, "grad_norm": 7.126120090484619, "learning_rate": 3.2032029925265707e-06, "loss": 1.2262, "step": 28310 }, { "epoch": 0.7459919724204239, "grad_norm": 10.581390380859375, "learning_rate": 3.1969474580798755e-06, "loss": 1.2049, "step": 28320 }, { "epoch": 0.7462553876649226, "grad_norm": 91.3353042602539, "learning_rate": 3.19069687545727e-06, "loss": 1.3099, "step": 28330 }, { "epoch": 0.7465188029094214, "grad_norm": 8.80888843536377, "learning_rate": 3.184451249208443e-06, "loss": 1.2608, "step": 28340 }, { "epoch": 0.7467822181539201, "grad_norm": 8.281060218811035, "learning_rate": 3.178210583879483e-06, "loss": 1.1929, "step": 28350 }, { "epoch": 0.7470456333984189, "grad_norm": 29.88004493713379, "learning_rate": 3.1719748840128574e-06, "loss": 1.2491, "step": 28360 }, { "epoch": 0.7473090486429176, "grad_norm": 4.409859657287598, "learning_rate": 3.165744154147429e-06, "loss": 1.1833, "step": 28370 }, { "epoch": 0.7475724638874163, "grad_norm": 5.630708694458008, "learning_rate": 3.1595183988184354e-06, "loss": 1.2773, "step": 28380 }, { "epoch": 0.7478358791319151, "grad_norm": 8.95604133605957, "learning_rate": 3.153297622557493e-06, "loss": 1.0849, "step": 28390 }, { "epoch": 0.7480992943764138, "grad_norm": 11.138555526733398, "learning_rate": 3.1470818298926e-06, "loss": 1.1724, "step": 28400 }, { "epoch": 0.7483627096209126, "grad_norm": 9.953782081604004, "learning_rate": 3.1408710253481277e-06, "loss": 1.2353, "step": 28410 }, { "epoch": 0.7486261248654112, "grad_norm": 7.278004169464111, "learning_rate": 3.1346652134448098e-06, "loss": 1.2747, "step": 28420 }, { "epoch": 0.74888954010991, "grad_norm": 7.015289306640625, "learning_rate": 3.1284643986997454e-06, "loss": 1.3521, "step": 28430 }, { "epoch": 0.7491529553544087, "grad_norm": 31.50489044189453, "learning_rate": 3.122268585626408e-06, "loss": 1.1604, "step": 28440 }, { "epoch": 0.7494163705989075, "grad_norm": 9.79479694366455, "learning_rate": 3.1160777787346153e-06, "loss": 1.1961, "step": 28450 }, { "epoch": 0.7496797858434062, "grad_norm": 10.374861717224121, "learning_rate": 3.109891982530554e-06, "loss": 1.1244, "step": 28460 }, { "epoch": 0.7499432010879049, "grad_norm": 22.22563362121582, "learning_rate": 3.103711201516757e-06, "loss": 1.2418, "step": 28470 }, { "epoch": 0.7502066163324037, "grad_norm": 181.2485809326172, "learning_rate": 3.0975354401921022e-06, "loss": 1.1937, "step": 28480 }, { "epoch": 0.7504700315769024, "grad_norm": 19.17630386352539, "learning_rate": 3.0913647030518267e-06, "loss": 1.307, "step": 28490 }, { "epoch": 0.7507334468214012, "grad_norm": 6.6467108726501465, "learning_rate": 3.085198994587497e-06, "loss": 1.1995, "step": 28500 }, { "epoch": 0.7509968620658999, "grad_norm": 8.172994613647461, "learning_rate": 3.079038319287029e-06, "loss": 1.2474, "step": 28510 }, { "epoch": 0.7512602773103987, "grad_norm": 18.543067932128906, "learning_rate": 3.0728826816346733e-06, "loss": 1.1796, "step": 28520 }, { "epoch": 0.7515236925548974, "grad_norm": 8.461370468139648, "learning_rate": 3.066732086111011e-06, "loss": 1.2926, "step": 28530 }, { "epoch": 0.7517871077993962, "grad_norm": 32.905635833740234, "learning_rate": 3.060586537192948e-06, "loss": 1.2061, "step": 28540 }, { "epoch": 0.7520505230438949, "grad_norm": 7.27686882019043, "learning_rate": 3.05444603935373e-06, "loss": 1.3205, "step": 28550 }, { "epoch": 0.7523139382883935, "grad_norm": 16.090869903564453, "learning_rate": 3.0483105970629145e-06, "loss": 1.2253, "step": 28560 }, { "epoch": 0.7525773535328923, "grad_norm": 15.524304389953613, "learning_rate": 3.0421802147863857e-06, "loss": 1.2639, "step": 28570 }, { "epoch": 0.752840768777391, "grad_norm": 7.1260762214660645, "learning_rate": 3.0360548969863477e-06, "loss": 1.1626, "step": 28580 }, { "epoch": 0.7531041840218898, "grad_norm": 11.451452255249023, "learning_rate": 3.029934648121303e-06, "loss": 1.2278, "step": 28590 }, { "epoch": 0.7533675992663885, "grad_norm": 8.683025360107422, "learning_rate": 3.023819472646079e-06, "loss": 1.2497, "step": 28600 }, { "epoch": 0.7536310145108873, "grad_norm": 5.360408306121826, "learning_rate": 3.0177093750118103e-06, "loss": 1.1293, "step": 28610 }, { "epoch": 0.753894429755386, "grad_norm": 14.305228233337402, "learning_rate": 3.011604359665925e-06, "loss": 1.1803, "step": 28620 }, { "epoch": 0.7541578449998848, "grad_norm": 12.235884666442871, "learning_rate": 3.005504431052163e-06, "loss": 1.163, "step": 28630 }, { "epoch": 0.7544212602443835, "grad_norm": 13.817008018493652, "learning_rate": 2.999409593610555e-06, "loss": 1.117, "step": 28640 }, { "epoch": 0.7546846754888822, "grad_norm": 5.548008918762207, "learning_rate": 2.9933198517774233e-06, "loss": 1.2165, "step": 28650 }, { "epoch": 0.754948090733381, "grad_norm": 7.306845664978027, "learning_rate": 2.987235209985393e-06, "loss": 1.2513, "step": 28660 }, { "epoch": 0.7552115059778797, "grad_norm": 10.334417343139648, "learning_rate": 2.9811556726633615e-06, "loss": 1.1674, "step": 28670 }, { "epoch": 0.7554749212223785, "grad_norm": 7.412516117095947, "learning_rate": 2.975081244236526e-06, "loss": 1.2057, "step": 28680 }, { "epoch": 0.7557383364668772, "grad_norm": 9.014336585998535, "learning_rate": 2.9690119291263543e-06, "loss": 1.1513, "step": 28690 }, { "epoch": 0.756001751711376, "grad_norm": 14.351969718933105, "learning_rate": 2.962947731750593e-06, "loss": 1.2096, "step": 28700 }, { "epoch": 0.7562651669558746, "grad_norm": 6.201923370361328, "learning_rate": 2.956888656523269e-06, "loss": 1.1669, "step": 28710 }, { "epoch": 0.7565285822003734, "grad_norm": 6.256604194641113, "learning_rate": 2.9508347078546816e-06, "loss": 1.1903, "step": 28720 }, { "epoch": 0.7567919974448721, "grad_norm": 5.771930694580078, "learning_rate": 2.9447858901513894e-06, "loss": 1.1866, "step": 28730 }, { "epoch": 0.7570554126893708, "grad_norm": 16.40662956237793, "learning_rate": 2.9387422078162266e-06, "loss": 1.218, "step": 28740 }, { "epoch": 0.7573188279338696, "grad_norm": 15.207331657409668, "learning_rate": 2.932703665248283e-06, "loss": 1.2242, "step": 28750 }, { "epoch": 0.7575822431783683, "grad_norm": 8.450349807739258, "learning_rate": 2.926670266842906e-06, "loss": 1.2411, "step": 28760 }, { "epoch": 0.7578456584228671, "grad_norm": 5.260526180267334, "learning_rate": 2.9206420169917083e-06, "loss": 1.1494, "step": 28770 }, { "epoch": 0.7581090736673658, "grad_norm": 15.014945983886719, "learning_rate": 2.9146189200825413e-06, "loss": 1.1618, "step": 28780 }, { "epoch": 0.7583724889118646, "grad_norm": 7.597702503204346, "learning_rate": 2.9086009804995196e-06, "loss": 1.1926, "step": 28790 }, { "epoch": 0.7586359041563633, "grad_norm": 10.684842109680176, "learning_rate": 2.9025882026229946e-06, "loss": 1.163, "step": 28800 }, { "epoch": 0.758899319400862, "grad_norm": 6.940189838409424, "learning_rate": 2.8965805908295595e-06, "loss": 1.2202, "step": 28810 }, { "epoch": 0.7591627346453608, "grad_norm": 15.791461944580078, "learning_rate": 2.890578149492054e-06, "loss": 1.2313, "step": 28820 }, { "epoch": 0.7594261498898595, "grad_norm": 5.58223819732666, "learning_rate": 2.884580882979554e-06, "loss": 1.2363, "step": 28830 }, { "epoch": 0.7596895651343583, "grad_norm": 12.057272911071777, "learning_rate": 2.878588795657363e-06, "loss": 1.2575, "step": 28840 }, { "epoch": 0.759952980378857, "grad_norm": 7.4193196296691895, "learning_rate": 2.8726018918870146e-06, "loss": 1.1674, "step": 28850 }, { "epoch": 0.7602163956233557, "grad_norm": 7.747781753540039, "learning_rate": 2.866620176026279e-06, "loss": 1.1734, "step": 28860 }, { "epoch": 0.7604798108678544, "grad_norm": 9.720925331115723, "learning_rate": 2.860643652429137e-06, "loss": 1.204, "step": 28870 }, { "epoch": 0.7607432261123532, "grad_norm": 5.958893775939941, "learning_rate": 2.8546723254457988e-06, "loss": 1.1502, "step": 28880 }, { "epoch": 0.7610066413568519, "grad_norm": 6.58300256729126, "learning_rate": 2.848706199422696e-06, "loss": 1.1615, "step": 28890 }, { "epoch": 0.7612700566013506, "grad_norm": 6.317964553833008, "learning_rate": 2.842745278702458e-06, "loss": 1.0715, "step": 28900 }, { "epoch": 0.7615334718458494, "grad_norm": 9.768638610839844, "learning_rate": 2.836789567623941e-06, "loss": 1.1557, "step": 28910 }, { "epoch": 0.7617968870903481, "grad_norm": 6.416312217712402, "learning_rate": 2.8308390705222057e-06, "loss": 1.1307, "step": 28920 }, { "epoch": 0.7620603023348469, "grad_norm": 6.566009998321533, "learning_rate": 2.8248937917285113e-06, "loss": 1.1914, "step": 28930 }, { "epoch": 0.7623237175793456, "grad_norm": 8.535893440246582, "learning_rate": 2.818953735570329e-06, "loss": 1.2963, "step": 28940 }, { "epoch": 0.7625871328238444, "grad_norm": 8.804706573486328, "learning_rate": 2.8130189063713187e-06, "loss": 1.258, "step": 28950 }, { "epoch": 0.7628505480683431, "grad_norm": 10.209192276000977, "learning_rate": 2.8070893084513372e-06, "loss": 1.2119, "step": 28960 }, { "epoch": 0.7631139633128419, "grad_norm": 8.886204719543457, "learning_rate": 2.8011649461264435e-06, "loss": 1.2349, "step": 28970 }, { "epoch": 0.7633773785573406, "grad_norm": 10.782876014709473, "learning_rate": 2.7952458237088707e-06, "loss": 1.2566, "step": 28980 }, { "epoch": 0.7636407938018392, "grad_norm": 8.938876152038574, "learning_rate": 2.789331945507049e-06, "loss": 1.2707, "step": 28990 }, { "epoch": 0.763904209046338, "grad_norm": 20.352174758911133, "learning_rate": 2.783423315825594e-06, "loss": 1.1739, "step": 29000 }, { "epoch": 0.7641676242908367, "grad_norm": 4.359538555145264, "learning_rate": 2.7775199389652828e-06, "loss": 1.2797, "step": 29010 }, { "epoch": 0.7644310395353355, "grad_norm": 10.400784492492676, "learning_rate": 2.771621819223088e-06, "loss": 1.2643, "step": 29020 }, { "epoch": 0.7646944547798342, "grad_norm": 10.794914245605469, "learning_rate": 2.7657289608921502e-06, "loss": 1.2051, "step": 29030 }, { "epoch": 0.764957870024333, "grad_norm": 6.815999507904053, "learning_rate": 2.759841368261773e-06, "loss": 1.207, "step": 29040 }, { "epoch": 0.7652212852688317, "grad_norm": 10.733039855957031, "learning_rate": 2.7539590456174402e-06, "loss": 1.1434, "step": 29050 }, { "epoch": 0.7654847005133305, "grad_norm": 10.242233276367188, "learning_rate": 2.7480819972407893e-06, "loss": 1.1787, "step": 29060 }, { "epoch": 0.7657481157578292, "grad_norm": 5.773466110229492, "learning_rate": 2.7422102274096185e-06, "loss": 1.0793, "step": 29070 }, { "epoch": 0.7660115310023279, "grad_norm": 7.101656913757324, "learning_rate": 2.736343740397893e-06, "loss": 1.169, "step": 29080 }, { "epoch": 0.7662749462468267, "grad_norm": 8.725702285766602, "learning_rate": 2.730482540475722e-06, "loss": 1.1255, "step": 29090 }, { "epoch": 0.7665383614913254, "grad_norm": 9.821623802185059, "learning_rate": 2.7246266319093774e-06, "loss": 1.2055, "step": 29100 }, { "epoch": 0.7668017767358242, "grad_norm": 10.057585716247559, "learning_rate": 2.7187760189612665e-06, "loss": 1.2679, "step": 29110 }, { "epoch": 0.7670651919803229, "grad_norm": 12.664830207824707, "learning_rate": 2.712930705889957e-06, "loss": 1.111, "step": 29120 }, { "epoch": 0.7673286072248217, "grad_norm": 6.862995624542236, "learning_rate": 2.707090696950143e-06, "loss": 1.256, "step": 29130 }, { "epoch": 0.7675920224693203, "grad_norm": 6.412499904632568, "learning_rate": 2.701255996392672e-06, "loss": 1.2801, "step": 29140 }, { "epoch": 0.7678554377138191, "grad_norm": 7.135316848754883, "learning_rate": 2.695426608464521e-06, "loss": 1.2615, "step": 29150 }, { "epoch": 0.7681188529583178, "grad_norm": 10.006275177001953, "learning_rate": 2.689602537408794e-06, "loss": 1.1688, "step": 29160 }, { "epoch": 0.7683822682028165, "grad_norm": 8.497063636779785, "learning_rate": 2.68378378746474e-06, "loss": 1.1008, "step": 29170 }, { "epoch": 0.7686456834473153, "grad_norm": 149.290771484375, "learning_rate": 2.677970362867719e-06, "loss": 1.3264, "step": 29180 }, { "epoch": 0.768909098691814, "grad_norm": 13.960959434509277, "learning_rate": 2.6721622678492255e-06, "loss": 1.1417, "step": 29190 }, { "epoch": 0.7691725139363128, "grad_norm": 19.05010223388672, "learning_rate": 2.666359506636874e-06, "loss": 1.1165, "step": 29200 }, { "epoch": 0.7694359291808115, "grad_norm": 13.392934799194336, "learning_rate": 2.660562083454391e-06, "loss": 1.2419, "step": 29210 }, { "epoch": 0.7696993444253103, "grad_norm": 9.286114692687988, "learning_rate": 2.654770002521617e-06, "loss": 1.2393, "step": 29220 }, { "epoch": 0.769962759669809, "grad_norm": 11.260431289672852, "learning_rate": 2.648983268054515e-06, "loss": 1.1537, "step": 29230 }, { "epoch": 0.7702261749143078, "grad_norm": 6.862528324127197, "learning_rate": 2.6432018842651395e-06, "loss": 1.168, "step": 29240 }, { "epoch": 0.7704895901588065, "grad_norm": 7.785955905914307, "learning_rate": 2.6374258553616693e-06, "loss": 1.1334, "step": 29250 }, { "epoch": 0.7707530054033052, "grad_norm": 4.850693225860596, "learning_rate": 2.631655185548371e-06, "loss": 1.1102, "step": 29260 }, { "epoch": 0.771016420647804, "grad_norm": 5.961513042449951, "learning_rate": 2.6258898790256117e-06, "loss": 1.108, "step": 29270 }, { "epoch": 0.7712798358923026, "grad_norm": 7.133718013763428, "learning_rate": 2.620129939989865e-06, "loss": 1.2256, "step": 29280 }, { "epoch": 0.7715432511368014, "grad_norm": 7.862963676452637, "learning_rate": 2.614375372633685e-06, "loss": 1.2005, "step": 29290 }, { "epoch": 0.7718066663813001, "grad_norm": 11.230714797973633, "learning_rate": 2.6086261811457234e-06, "loss": 1.145, "step": 29300 }, { "epoch": 0.7720700816257989, "grad_norm": 10.648667335510254, "learning_rate": 2.602882369710724e-06, "loss": 1.2441, "step": 29310 }, { "epoch": 0.7723334968702976, "grad_norm": 7.801405906677246, "learning_rate": 2.597143942509496e-06, "loss": 1.1544, "step": 29320 }, { "epoch": 0.7725969121147963, "grad_norm": 17.442501068115234, "learning_rate": 2.591410903718946e-06, "loss": 1.1998, "step": 29330 }, { "epoch": 0.7728603273592951, "grad_norm": 8.34951114654541, "learning_rate": 2.585683257512057e-06, "loss": 1.1808, "step": 29340 }, { "epoch": 0.7731237426037938, "grad_norm": 7.824574947357178, "learning_rate": 2.579961008057877e-06, "loss": 1.1223, "step": 29350 }, { "epoch": 0.7733871578482926, "grad_norm": 12.186628341674805, "learning_rate": 2.5742441595215394e-06, "loss": 1.1924, "step": 29360 }, { "epoch": 0.7736505730927913, "grad_norm": 7.276239395141602, "learning_rate": 2.5685327160642357e-06, "loss": 1.1141, "step": 29370 }, { "epoch": 0.7739139883372901, "grad_norm": 10.465269088745117, "learning_rate": 2.5628266818432246e-06, "loss": 1.2721, "step": 29380 }, { "epoch": 0.7741774035817888, "grad_norm": 14.45191478729248, "learning_rate": 2.557126061011831e-06, "loss": 1.2609, "step": 29390 }, { "epoch": 0.7744408188262876, "grad_norm": 19.268863677978516, "learning_rate": 2.5514308577194413e-06, "loss": 1.1946, "step": 29400 }, { "epoch": 0.7747042340707863, "grad_norm": 6.567240238189697, "learning_rate": 2.54574107611149e-06, "loss": 1.1466, "step": 29410 }, { "epoch": 0.774967649315285, "grad_norm": 17.037254333496094, "learning_rate": 2.5400567203294757e-06, "loss": 1.2487, "step": 29420 }, { "epoch": 0.7752310645597837, "grad_norm": 6.074851989746094, "learning_rate": 2.5343777945109404e-06, "loss": 1.1965, "step": 29430 }, { "epoch": 0.7754944798042824, "grad_norm": 15.437520980834961, "learning_rate": 2.528704302789472e-06, "loss": 1.3025, "step": 29440 }, { "epoch": 0.7757578950487812, "grad_norm": 7.9437713623046875, "learning_rate": 2.5230362492947136e-06, "loss": 1.2582, "step": 29450 }, { "epoch": 0.7760213102932799, "grad_norm": 8.013522148132324, "learning_rate": 2.5173736381523373e-06, "loss": 1.1951, "step": 29460 }, { "epoch": 0.7762847255377787, "grad_norm": 12.880861282348633, "learning_rate": 2.5117164734840636e-06, "loss": 1.1582, "step": 29470 }, { "epoch": 0.7765481407822774, "grad_norm": 12.082415580749512, "learning_rate": 2.506064759407645e-06, "loss": 1.1545, "step": 29480 }, { "epoch": 0.7768115560267762, "grad_norm": 5.783811092376709, "learning_rate": 2.5004185000368608e-06, "loss": 1.271, "step": 29490 }, { "epoch": 0.7770749712712749, "grad_norm": 8.522090911865234, "learning_rate": 2.494777699481529e-06, "loss": 1.2646, "step": 29500 }, { "epoch": 0.7773383865157736, "grad_norm": 8.932174682617188, "learning_rate": 2.489142361847493e-06, "loss": 1.1694, "step": 29510 }, { "epoch": 0.7776018017602724, "grad_norm": 7.147249698638916, "learning_rate": 2.483512491236616e-06, "loss": 1.2839, "step": 29520 }, { "epoch": 0.7778652170047711, "grad_norm": 9.566771507263184, "learning_rate": 2.4778880917467795e-06, "loss": 1.1721, "step": 29530 }, { "epoch": 0.7781286322492699, "grad_norm": 9.908282279968262, "learning_rate": 2.4722691674718923e-06, "loss": 1.1978, "step": 29540 }, { "epoch": 0.7783920474937686, "grad_norm": 7.336192607879639, "learning_rate": 2.4666557225018663e-06, "loss": 1.1948, "step": 29550 }, { "epoch": 0.7786554627382674, "grad_norm": 5.521805286407471, "learning_rate": 2.4610477609226356e-06, "loss": 1.1967, "step": 29560 }, { "epoch": 0.778918877982766, "grad_norm": 8.708727836608887, "learning_rate": 2.455445286816137e-06, "loss": 1.346, "step": 29570 }, { "epoch": 0.7791822932272648, "grad_norm": 11.33962345123291, "learning_rate": 2.4498483042603103e-06, "loss": 1.1213, "step": 29580 }, { "epoch": 0.7794457084717635, "grad_norm": 10.437017440795898, "learning_rate": 2.444256817329109e-06, "loss": 1.0929, "step": 29590 }, { "epoch": 0.7797091237162622, "grad_norm": 5.898515701293945, "learning_rate": 2.4386708300924735e-06, "loss": 1.169, "step": 29600 }, { "epoch": 0.779972538960761, "grad_norm": 7.564632415771484, "learning_rate": 2.433090346616348e-06, "loss": 1.2477, "step": 29610 }, { "epoch": 0.7802359542052597, "grad_norm": 6.833156108856201, "learning_rate": 2.4275153709626764e-06, "loss": 1.161, "step": 29620 }, { "epoch": 0.7804993694497585, "grad_norm": 7.021008014678955, "learning_rate": 2.4219459071893803e-06, "loss": 1.1873, "step": 29630 }, { "epoch": 0.7807627846942572, "grad_norm": 9.238037109375, "learning_rate": 2.4163819593503758e-06, "loss": 1.142, "step": 29640 }, { "epoch": 0.781026199938756, "grad_norm": 12.667337417602539, "learning_rate": 2.4108235314955687e-06, "loss": 1.1803, "step": 29650 }, { "epoch": 0.7812896151832547, "grad_norm": 8.852887153625488, "learning_rate": 2.405270627670836e-06, "loss": 1.2027, "step": 29660 }, { "epoch": 0.7815530304277535, "grad_norm": 9.440607070922852, "learning_rate": 2.3997232519180446e-06, "loss": 1.1204, "step": 29670 }, { "epoch": 0.7818164456722522, "grad_norm": 6.9737091064453125, "learning_rate": 2.394181408275037e-06, "loss": 1.249, "step": 29680 }, { "epoch": 0.7820798609167509, "grad_norm": 11.870588302612305, "learning_rate": 2.3886451007756162e-06, "loss": 1.2429, "step": 29690 }, { "epoch": 0.7823432761612497, "grad_norm": 92.65271759033203, "learning_rate": 2.3831143334495677e-06, "loss": 1.2367, "step": 29700 }, { "epoch": 0.7826066914057483, "grad_norm": 7.394147872924805, "learning_rate": 2.3775891103226454e-06, "loss": 1.323, "step": 29710 }, { "epoch": 0.7828701066502471, "grad_norm": 13.865079879760742, "learning_rate": 2.3720694354165584e-06, "loss": 1.3226, "step": 29720 }, { "epoch": 0.7831335218947458, "grad_norm": 8.259990692138672, "learning_rate": 2.366555312748986e-06, "loss": 1.1864, "step": 29730 }, { "epoch": 0.7833969371392446, "grad_norm": 11.749423027038574, "learning_rate": 2.3610467463335627e-06, "loss": 1.1787, "step": 29740 }, { "epoch": 0.7836603523837433, "grad_norm": 6.162956714630127, "learning_rate": 2.355543740179873e-06, "loss": 1.2101, "step": 29750 }, { "epoch": 0.7839237676282421, "grad_norm": 5.713232040405273, "learning_rate": 2.350046298293467e-06, "loss": 1.1573, "step": 29760 }, { "epoch": 0.7841871828727408, "grad_norm": 15.664694786071777, "learning_rate": 2.3445544246758313e-06, "loss": 1.2584, "step": 29770 }, { "epoch": 0.7844505981172395, "grad_norm": 9.85895824432373, "learning_rate": 2.3390681233244105e-06, "loss": 1.2127, "step": 29780 }, { "epoch": 0.7847140133617383, "grad_norm": 12.996437072753906, "learning_rate": 2.333587398232586e-06, "loss": 1.1112, "step": 29790 }, { "epoch": 0.784977428606237, "grad_norm": 4.515507698059082, "learning_rate": 2.3281122533896814e-06, "loss": 1.154, "step": 29800 }, { "epoch": 0.7852408438507358, "grad_norm": 14.356064796447754, "learning_rate": 2.322642692780961e-06, "loss": 1.1639, "step": 29810 }, { "epoch": 0.7855042590952345, "grad_norm": 12.914161682128906, "learning_rate": 2.317178720387625e-06, "loss": 1.203, "step": 29820 }, { "epoch": 0.7857676743397333, "grad_norm": 8.196971893310547, "learning_rate": 2.3117203401868003e-06, "loss": 1.1648, "step": 29830 }, { "epoch": 0.786031089584232, "grad_norm": 14.145028114318848, "learning_rate": 2.3062675561515536e-06, "loss": 1.2294, "step": 29840 }, { "epoch": 0.7862945048287308, "grad_norm": 8.980778694152832, "learning_rate": 2.300820372250867e-06, "loss": 1.2127, "step": 29850 }, { "epoch": 0.7865579200732294, "grad_norm": 10.700285911560059, "learning_rate": 2.2953787924496494e-06, "loss": 1.1653, "step": 29860 }, { "epoch": 0.7868213353177281, "grad_norm": 7.736145496368408, "learning_rate": 2.289942820708738e-06, "loss": 1.2301, "step": 29870 }, { "epoch": 0.7870847505622269, "grad_norm": 7.432765483856201, "learning_rate": 2.284512460984878e-06, "loss": 1.1063, "step": 29880 }, { "epoch": 0.7873481658067256, "grad_norm": 5.816494941711426, "learning_rate": 2.279087717230738e-06, "loss": 1.2144, "step": 29890 }, { "epoch": 0.7876115810512244, "grad_norm": 11.162300109863281, "learning_rate": 2.27366859339489e-06, "loss": 1.1636, "step": 29900 }, { "epoch": 0.7878749962957231, "grad_norm": 8.670113563537598, "learning_rate": 2.2682550934218262e-06, "loss": 1.136, "step": 29910 }, { "epoch": 0.7881384115402219, "grad_norm": 8.336100578308105, "learning_rate": 2.262847221251935e-06, "loss": 1.2706, "step": 29920 }, { "epoch": 0.7884018267847206, "grad_norm": 15.423498153686523, "learning_rate": 2.257444980821516e-06, "loss": 1.1808, "step": 29930 }, { "epoch": 0.7886652420292193, "grad_norm": 7.66867733001709, "learning_rate": 2.2520483760627642e-06, "loss": 1.365, "step": 29940 }, { "epoch": 0.7889286572737181, "grad_norm": 5.870673656463623, "learning_rate": 2.2466574109037743e-06, "loss": 1.1725, "step": 29950 }, { "epoch": 0.7891920725182168, "grad_norm": 8.252575874328613, "learning_rate": 2.241272089268539e-06, "loss": 1.1429, "step": 29960 }, { "epoch": 0.7894554877627156, "grad_norm": 14.552165031433105, "learning_rate": 2.235892415076938e-06, "loss": 1.1883, "step": 29970 }, { "epoch": 0.7897189030072143, "grad_norm": 7.39256477355957, "learning_rate": 2.230518392244743e-06, "loss": 1.1468, "step": 29980 }, { "epoch": 0.7899823182517131, "grad_norm": 7.381988525390625, "learning_rate": 2.225150024683619e-06, "loss": 1.1685, "step": 29990 }, { "epoch": 0.7902457334962117, "grad_norm": 10.897844314575195, "learning_rate": 2.219787316301096e-06, "loss": 1.27, "step": 30000 }, { "epoch": 0.7905091487407105, "grad_norm": 7.39321756362915, "learning_rate": 2.214430271000603e-06, "loss": 1.2735, "step": 30010 }, { "epoch": 0.7907725639852092, "grad_norm": 7.9408416748046875, "learning_rate": 2.209078892681442e-06, "loss": 1.142, "step": 30020 }, { "epoch": 0.7910359792297079, "grad_norm": 12.226337432861328, "learning_rate": 2.2037331852387843e-06, "loss": 1.1792, "step": 30030 }, { "epoch": 0.7912993944742067, "grad_norm": 6.207977294921875, "learning_rate": 2.1983931525636836e-06, "loss": 1.1798, "step": 30040 }, { "epoch": 0.7915628097187054, "grad_norm": 27.126054763793945, "learning_rate": 2.1930587985430528e-06, "loss": 1.2528, "step": 30050 }, { "epoch": 0.7918262249632042, "grad_norm": 5.866193771362305, "learning_rate": 2.187730127059674e-06, "loss": 1.1347, "step": 30060 }, { "epoch": 0.7920896402077029, "grad_norm": 7.08782958984375, "learning_rate": 2.1824071419922e-06, "loss": 1.1463, "step": 30070 }, { "epoch": 0.7923530554522017, "grad_norm": 7.024013519287109, "learning_rate": 2.1770898472151337e-06, "loss": 1.0605, "step": 30080 }, { "epoch": 0.7926164706967004, "grad_norm": 7.243834018707275, "learning_rate": 2.1717782465988457e-06, "loss": 1.1988, "step": 30090 }, { "epoch": 0.7928798859411992, "grad_norm": 6.959171295166016, "learning_rate": 2.1664723440095626e-06, "loss": 1.1962, "step": 30100 }, { "epoch": 0.7931433011856979, "grad_norm": 10.731844902038574, "learning_rate": 2.161172143309348e-06, "loss": 1.1998, "step": 30110 }, { "epoch": 0.7934067164301966, "grad_norm": 8.12711238861084, "learning_rate": 2.155877648356132e-06, "loss": 1.1448, "step": 30120 }, { "epoch": 0.7936701316746954, "grad_norm": 9.20328140258789, "learning_rate": 2.1505888630036877e-06, "loss": 1.2241, "step": 30130 }, { "epoch": 0.793933546919194, "grad_norm": 12.966363906860352, "learning_rate": 2.145305791101626e-06, "loss": 1.1203, "step": 30140 }, { "epoch": 0.7941969621636928, "grad_norm": 11.48041820526123, "learning_rate": 2.1400284364954093e-06, "loss": 1.1471, "step": 30150 }, { "epoch": 0.7944603774081915, "grad_norm": 5.0922675132751465, "learning_rate": 2.134756803026328e-06, "loss": 1.2408, "step": 30160 }, { "epoch": 0.7947237926526903, "grad_norm": 12.124212265014648, "learning_rate": 2.129490894531512e-06, "loss": 1.0427, "step": 30170 }, { "epoch": 0.794987207897189, "grad_norm": 7.154329299926758, "learning_rate": 2.1242307148439258e-06, "loss": 1.2521, "step": 30180 }, { "epoch": 0.7952506231416878, "grad_norm": 11.58328914642334, "learning_rate": 2.1189762677923662e-06, "loss": 1.1745, "step": 30190 }, { "epoch": 0.7955140383861865, "grad_norm": 7.410404205322266, "learning_rate": 2.1137275572014525e-06, "loss": 1.21, "step": 30200 }, { "epoch": 0.7957774536306852, "grad_norm": 9.289090156555176, "learning_rate": 2.108484586891627e-06, "loss": 1.2256, "step": 30210 }, { "epoch": 0.796040868875184, "grad_norm": 10.899478912353516, "learning_rate": 2.1032473606791628e-06, "loss": 1.1164, "step": 30220 }, { "epoch": 0.7963042841196827, "grad_norm": 6.359890937805176, "learning_rate": 2.0980158823761397e-06, "loss": 1.2263, "step": 30230 }, { "epoch": 0.7965676993641815, "grad_norm": 9.507186889648438, "learning_rate": 2.092790155790465e-06, "loss": 1.1782, "step": 30240 }, { "epoch": 0.7968311146086802, "grad_norm": 11.528604507446289, "learning_rate": 2.0875701847258513e-06, "loss": 1.1979, "step": 30250 }, { "epoch": 0.797094529853179, "grad_norm": 11.746885299682617, "learning_rate": 2.08235597298183e-06, "loss": 1.2236, "step": 30260 }, { "epoch": 0.7973579450976777, "grad_norm": 11.046670913696289, "learning_rate": 2.0771475243537307e-06, "loss": 1.1462, "step": 30270 }, { "epoch": 0.7976213603421765, "grad_norm": 12.598626136779785, "learning_rate": 2.071944842632693e-06, "loss": 1.238, "step": 30280 }, { "epoch": 0.7978847755866751, "grad_norm": 7.527684211730957, "learning_rate": 2.0667479316056594e-06, "loss": 1.0968, "step": 30290 }, { "epoch": 0.7981481908311738, "grad_norm": 5.641901969909668, "learning_rate": 2.061556795055375e-06, "loss": 1.2431, "step": 30300 }, { "epoch": 0.7984116060756726, "grad_norm": 6.267887115478516, "learning_rate": 2.0563714367603748e-06, "loss": 1.1648, "step": 30310 }, { "epoch": 0.7986750213201713, "grad_norm": 6.753127098083496, "learning_rate": 2.05119186049499e-06, "loss": 1.1743, "step": 30320 }, { "epoch": 0.7989384365646701, "grad_norm": 9.947144508361816, "learning_rate": 2.0460180700293485e-06, "loss": 1.1304, "step": 30330 }, { "epoch": 0.7992018518091688, "grad_norm": 9.089081764221191, "learning_rate": 2.0408500691293587e-06, "loss": 1.2436, "step": 30340 }, { "epoch": 0.7994652670536676, "grad_norm": 14.342866897583008, "learning_rate": 2.0356878615567223e-06, "loss": 1.1603, "step": 30350 }, { "epoch": 0.7997286822981663, "grad_norm": 6.698002815246582, "learning_rate": 2.03053145106892e-06, "loss": 1.154, "step": 30360 }, { "epoch": 0.7999920975426651, "grad_norm": 13.20721435546875, "learning_rate": 2.02538084141921e-06, "loss": 1.2414, "step": 30370 }, { "epoch": 0.8002555127871638, "grad_norm": 8.23073959350586, "learning_rate": 2.0202360363566377e-06, "loss": 1.1799, "step": 30380 }, { "epoch": 0.8005189280316625, "grad_norm": 8.671072959899902, "learning_rate": 2.0150970396260106e-06, "loss": 1.2781, "step": 30390 }, { "epoch": 0.8007823432761613, "grad_norm": 10.397076606750488, "learning_rate": 2.0099638549679213e-06, "loss": 1.2991, "step": 30400 }, { "epoch": 0.80104575852066, "grad_norm": 9.48668098449707, "learning_rate": 2.0048364861187276e-06, "loss": 1.2025, "step": 30410 }, { "epoch": 0.8013091737651588, "grad_norm": 5.827513217926025, "learning_rate": 1.999714936810545e-06, "loss": 1.1281, "step": 30420 }, { "epoch": 0.8015725890096574, "grad_norm": 8.886529922485352, "learning_rate": 1.994599210771263e-06, "loss": 1.1126, "step": 30430 }, { "epoch": 0.8018360042541562, "grad_norm": 7.082366943359375, "learning_rate": 1.989489311724535e-06, "loss": 1.0692, "step": 30440 }, { "epoch": 0.8020994194986549, "grad_norm": 11.16085433959961, "learning_rate": 1.9843852433897626e-06, "loss": 1.16, "step": 30450 }, { "epoch": 0.8023628347431536, "grad_norm": 10.1416654586792, "learning_rate": 1.979287009482109e-06, "loss": 1.1018, "step": 30460 }, { "epoch": 0.8026262499876524, "grad_norm": 11.169856071472168, "learning_rate": 1.9741946137124978e-06, "loss": 1.2392, "step": 30470 }, { "epoch": 0.8028896652321511, "grad_norm": 8.751354217529297, "learning_rate": 1.9691080597875844e-06, "loss": 1.1052, "step": 30480 }, { "epoch": 0.8031530804766499, "grad_norm": 21.854312896728516, "learning_rate": 1.9640273514097886e-06, "loss": 1.178, "step": 30490 }, { "epoch": 0.8034164957211486, "grad_norm": 7.638960838317871, "learning_rate": 1.958952492277273e-06, "loss": 1.0936, "step": 30500 }, { "epoch": 0.8036799109656474, "grad_norm": 6.640834808349609, "learning_rate": 1.953883486083934e-06, "loss": 1.2324, "step": 30510 }, { "epoch": 0.8039433262101461, "grad_norm": 9.722416877746582, "learning_rate": 1.948820336519418e-06, "loss": 1.1237, "step": 30520 }, { "epoch": 0.8042067414546449, "grad_norm": 8.796442985534668, "learning_rate": 1.9437630472691026e-06, "loss": 1.1802, "step": 30530 }, { "epoch": 0.8044701566991436, "grad_norm": 13.944660186767578, "learning_rate": 1.9387116220140965e-06, "loss": 1.1418, "step": 30540 }, { "epoch": 0.8047335719436423, "grad_norm": 7.697389125823975, "learning_rate": 1.9336660644312524e-06, "loss": 1.1231, "step": 30550 }, { "epoch": 0.804996987188141, "grad_norm": 11.009883880615234, "learning_rate": 1.928626378193138e-06, "loss": 1.1298, "step": 30560 }, { "epoch": 0.8052604024326397, "grad_norm": 5.224260330200195, "learning_rate": 1.923592566968058e-06, "loss": 1.1393, "step": 30570 }, { "epoch": 0.8055238176771385, "grad_norm": 10.568594932556152, "learning_rate": 1.918564634420034e-06, "loss": 1.1475, "step": 30580 }, { "epoch": 0.8057872329216372, "grad_norm": 13.124283790588379, "learning_rate": 1.9135425842088083e-06, "loss": 1.1887, "step": 30590 }, { "epoch": 0.806050648166136, "grad_norm": 5.619673728942871, "learning_rate": 1.908526419989848e-06, "loss": 1.1704, "step": 30600 }, { "epoch": 0.8063140634106347, "grad_norm": 5.663410663604736, "learning_rate": 1.9035161454143347e-06, "loss": 1.1284, "step": 30610 }, { "epoch": 0.8065774786551335, "grad_norm": 11.685273170471191, "learning_rate": 1.8985117641291562e-06, "loss": 1.221, "step": 30620 }, { "epoch": 0.8068408938996322, "grad_norm": 10.362303733825684, "learning_rate": 1.893513279776913e-06, "loss": 1.286, "step": 30630 }, { "epoch": 0.8071043091441309, "grad_norm": 7.370004653930664, "learning_rate": 1.8885206959959212e-06, "loss": 1.2271, "step": 30640 }, { "epoch": 0.8073677243886297, "grad_norm": 10.271496772766113, "learning_rate": 1.883534016420191e-06, "loss": 1.1456, "step": 30650 }, { "epoch": 0.8076311396331284, "grad_norm": 6.52615213394165, "learning_rate": 1.8785532446794442e-06, "loss": 1.1787, "step": 30660 }, { "epoch": 0.8078945548776272, "grad_norm": 12.689484596252441, "learning_rate": 1.8735783843990962e-06, "loss": 1.2581, "step": 30670 }, { "epoch": 0.8081579701221259, "grad_norm": 14.378952026367188, "learning_rate": 1.8686094392002585e-06, "loss": 1.1537, "step": 30680 }, { "epoch": 0.8084213853666247, "grad_norm": 12.024587631225586, "learning_rate": 1.8636464126997443e-06, "loss": 1.0988, "step": 30690 }, { "epoch": 0.8086848006111234, "grad_norm": 16.845104217529297, "learning_rate": 1.8586893085100556e-06, "loss": 1.2898, "step": 30700 }, { "epoch": 0.8089482158556222, "grad_norm": 6.919190406799316, "learning_rate": 1.8537381302393787e-06, "loss": 1.2093, "step": 30710 }, { "epoch": 0.8092116311001208, "grad_norm": 5.981680870056152, "learning_rate": 1.8487928814915945e-06, "loss": 1.1422, "step": 30720 }, { "epoch": 0.8094750463446195, "grad_norm": 13.231403350830078, "learning_rate": 1.8438535658662615e-06, "loss": 1.2616, "step": 30730 }, { "epoch": 0.8097384615891183, "grad_norm": 11.066880226135254, "learning_rate": 1.8389201869586205e-06, "loss": 1.1545, "step": 30740 }, { "epoch": 0.810001876833617, "grad_norm": 6.695717811584473, "learning_rate": 1.8339927483595965e-06, "loss": 1.2404, "step": 30750 }, { "epoch": 0.8102652920781158, "grad_norm": 8.397920608520508, "learning_rate": 1.829071253655782e-06, "loss": 1.1426, "step": 30760 }, { "epoch": 0.8105287073226145, "grad_norm": 9.284100532531738, "learning_rate": 1.8241557064294513e-06, "loss": 1.2197, "step": 30770 }, { "epoch": 0.8107921225671133, "grad_norm": 10.304032325744629, "learning_rate": 1.819246110258549e-06, "loss": 1.2595, "step": 30780 }, { "epoch": 0.811055537811612, "grad_norm": 8.334758758544922, "learning_rate": 1.8143424687166767e-06, "loss": 1.2529, "step": 30790 }, { "epoch": 0.8113189530561108, "grad_norm": 7.986595153808594, "learning_rate": 1.8094447853731145e-06, "loss": 1.1339, "step": 30800 }, { "epoch": 0.8115823683006095, "grad_norm": 5.3527655601501465, "learning_rate": 1.8045530637928033e-06, "loss": 1.2074, "step": 30810 }, { "epoch": 0.8118457835451082, "grad_norm": 7.946804046630859, "learning_rate": 1.7996673075363391e-06, "loss": 1.2215, "step": 30820 }, { "epoch": 0.812109198789607, "grad_norm": 7.503734111785889, "learning_rate": 1.794787520159983e-06, "loss": 1.0723, "step": 30830 }, { "epoch": 0.8123726140341057, "grad_norm": 6.672111511230469, "learning_rate": 1.7899137052156445e-06, "loss": 1.1157, "step": 30840 }, { "epoch": 0.8126360292786045, "grad_norm": 8.291959762573242, "learning_rate": 1.7850458662508884e-06, "loss": 1.2212, "step": 30850 }, { "epoch": 0.8128994445231031, "grad_norm": 12.518600463867188, "learning_rate": 1.7801840068089337e-06, "loss": 1.1289, "step": 30860 }, { "epoch": 0.8131628597676019, "grad_norm": 6.536609172821045, "learning_rate": 1.7753281304286397e-06, "loss": 1.1697, "step": 30870 }, { "epoch": 0.8134262750121006, "grad_norm": 8.193613052368164, "learning_rate": 1.7704782406445197e-06, "loss": 1.1448, "step": 30880 }, { "epoch": 0.8136896902565994, "grad_norm": 13.966492652893066, "learning_rate": 1.7656343409867215e-06, "loss": 1.201, "step": 30890 }, { "epoch": 0.8139531055010981, "grad_norm": 20.347383499145508, "learning_rate": 1.7607964349810337e-06, "loss": 1.2047, "step": 30900 }, { "epoch": 0.8142165207455968, "grad_norm": 7.036605358123779, "learning_rate": 1.7559645261488856e-06, "loss": 1.1768, "step": 30910 }, { "epoch": 0.8144799359900956, "grad_norm": 7.031517028808594, "learning_rate": 1.7511386180073442e-06, "loss": 1.1316, "step": 30920 }, { "epoch": 0.8147433512345943, "grad_norm": 16.96132469177246, "learning_rate": 1.7463187140690974e-06, "loss": 1.1893, "step": 30930 }, { "epoch": 0.8150067664790931, "grad_norm": 8.5442476272583, "learning_rate": 1.7415048178424765e-06, "loss": 1.2012, "step": 30940 }, { "epoch": 0.8152701817235918, "grad_norm": 11.770572662353516, "learning_rate": 1.7366969328314299e-06, "loss": 1.1133, "step": 30950 }, { "epoch": 0.8155335969680906, "grad_norm": 7.239218235015869, "learning_rate": 1.7318950625355323e-06, "loss": 1.229, "step": 30960 }, { "epoch": 0.8157970122125893, "grad_norm": 8.630463600158691, "learning_rate": 1.727099210449983e-06, "loss": 1.1614, "step": 30970 }, { "epoch": 0.816060427457088, "grad_norm": 7.736532688140869, "learning_rate": 1.722309380065602e-06, "loss": 1.1189, "step": 30980 }, { "epoch": 0.8163238427015868, "grad_norm": 7.92650032043457, "learning_rate": 1.7175255748688236e-06, "loss": 1.153, "step": 30990 }, { "epoch": 0.8165872579460854, "grad_norm": 8.156909942626953, "learning_rate": 1.7127477983416917e-06, "loss": 1.1043, "step": 31000 }, { "epoch": 0.8168506731905842, "grad_norm": 7.331878185272217, "learning_rate": 1.7079760539618729e-06, "loss": 1.151, "step": 31010 }, { "epoch": 0.8171140884350829, "grad_norm": 7.292756080627441, "learning_rate": 1.7032103452026327e-06, "loss": 1.1558, "step": 31020 }, { "epoch": 0.8173775036795817, "grad_norm": 12.1723051071167, "learning_rate": 1.6984506755328522e-06, "loss": 1.1534, "step": 31030 }, { "epoch": 0.8176409189240804, "grad_norm": 7.21678352355957, "learning_rate": 1.6936970484170111e-06, "loss": 1.0971, "step": 31040 }, { "epoch": 0.8179043341685792, "grad_norm": 9.061594009399414, "learning_rate": 1.6889494673151897e-06, "loss": 1.1066, "step": 31050 }, { "epoch": 0.8181677494130779, "grad_norm": 6.573923587799072, "learning_rate": 1.6842079356830732e-06, "loss": 1.2331, "step": 31060 }, { "epoch": 0.8184311646575766, "grad_norm": 9.935912132263184, "learning_rate": 1.679472456971938e-06, "loss": 1.1589, "step": 31070 }, { "epoch": 0.8186945799020754, "grad_norm": 10.924025535583496, "learning_rate": 1.6747430346286576e-06, "loss": 1.1589, "step": 31080 }, { "epoch": 0.8189579951465741, "grad_norm": 7.645543575286865, "learning_rate": 1.6700196720957006e-06, "loss": 1.1754, "step": 31090 }, { "epoch": 0.8192214103910729, "grad_norm": 6.837041854858398, "learning_rate": 1.665302372811114e-06, "loss": 1.1601, "step": 31100 }, { "epoch": 0.8194848256355716, "grad_norm": 9.495421409606934, "learning_rate": 1.6605911402085407e-06, "loss": 1.0987, "step": 31110 }, { "epoch": 0.8197482408800704, "grad_norm": 6.673137187957764, "learning_rate": 1.6558859777172088e-06, "loss": 1.102, "step": 31120 }, { "epoch": 0.820011656124569, "grad_norm": 11.590087890625, "learning_rate": 1.6511868887619197e-06, "loss": 1.118, "step": 31130 }, { "epoch": 0.8202750713690679, "grad_norm": 9.012484550476074, "learning_rate": 1.6464938767630634e-06, "loss": 1.1177, "step": 31140 }, { "epoch": 0.8205384866135665, "grad_norm": 5.9406418800354, "learning_rate": 1.6418069451365993e-06, "loss": 1.2369, "step": 31150 }, { "epoch": 0.8208019018580652, "grad_norm": 6.872724533081055, "learning_rate": 1.6371260972940618e-06, "loss": 1.1546, "step": 31160 }, { "epoch": 0.821065317102564, "grad_norm": 12.503777503967285, "learning_rate": 1.6324513366425631e-06, "loss": 1.1608, "step": 31170 }, { "epoch": 0.8213287323470627, "grad_norm": 9.994277954101562, "learning_rate": 1.6277826665847762e-06, "loss": 1.1954, "step": 31180 }, { "epoch": 0.8215921475915615, "grad_norm": 15.936970710754395, "learning_rate": 1.623120090518948e-06, "loss": 1.1555, "step": 31190 }, { "epoch": 0.8218555628360602, "grad_norm": 13.967419624328613, "learning_rate": 1.618463611838892e-06, "loss": 1.1344, "step": 31200 }, { "epoch": 0.822118978080559, "grad_norm": 6.840645790100098, "learning_rate": 1.6138132339339662e-06, "loss": 1.2307, "step": 31210 }, { "epoch": 0.8223823933250577, "grad_norm": 11.96902084350586, "learning_rate": 1.609168960189108e-06, "loss": 1.1939, "step": 31220 }, { "epoch": 0.8226458085695565, "grad_norm": 5.833017349243164, "learning_rate": 1.6045307939848042e-06, "loss": 1.2028, "step": 31230 }, { "epoch": 0.8229092238140552, "grad_norm": 8.82912826538086, "learning_rate": 1.599898738697092e-06, "loss": 1.2014, "step": 31240 }, { "epoch": 0.8231726390585539, "grad_norm": 7.119807720184326, "learning_rate": 1.595272797697569e-06, "loss": 1.1353, "step": 31250 }, { "epoch": 0.8234360543030527, "grad_norm": 10.981313705444336, "learning_rate": 1.5906529743533738e-06, "loss": 1.2312, "step": 31260 }, { "epoch": 0.8236994695475514, "grad_norm": 7.623471736907959, "learning_rate": 1.5860392720271945e-06, "loss": 1.0894, "step": 31270 }, { "epoch": 0.8239628847920502, "grad_norm": 13.504652976989746, "learning_rate": 1.5814316940772668e-06, "loss": 1.2434, "step": 31280 }, { "epoch": 0.8242263000365488, "grad_norm": 7.754531383514404, "learning_rate": 1.5768302438573701e-06, "loss": 1.0537, "step": 31290 }, { "epoch": 0.8244897152810476, "grad_norm": 9.678184509277344, "learning_rate": 1.5722349247168167e-06, "loss": 1.1731, "step": 31300 }, { "epoch": 0.8247531305255463, "grad_norm": 9.057154655456543, "learning_rate": 1.5676457400004597e-06, "loss": 1.1228, "step": 31310 }, { "epoch": 0.8250165457700451, "grad_norm": 5.831321716308594, "learning_rate": 1.5630626930486891e-06, "loss": 1.1123, "step": 31320 }, { "epoch": 0.8252799610145438, "grad_norm": 10.18830394744873, "learning_rate": 1.5584857871974224e-06, "loss": 1.2498, "step": 31330 }, { "epoch": 0.8255433762590425, "grad_norm": 14.060593605041504, "learning_rate": 1.5539150257781144e-06, "loss": 1.1883, "step": 31340 }, { "epoch": 0.8258067915035413, "grad_norm": 13.48275375366211, "learning_rate": 1.5493504121177383e-06, "loss": 1.1684, "step": 31350 }, { "epoch": 0.82607020674804, "grad_norm": 12.0016450881958, "learning_rate": 1.544791949538802e-06, "loss": 1.0894, "step": 31360 }, { "epoch": 0.8263336219925388, "grad_norm": 10.265583038330078, "learning_rate": 1.540239641359329e-06, "loss": 1.073, "step": 31370 }, { "epoch": 0.8265970372370375, "grad_norm": 8.88713550567627, "learning_rate": 1.5356934908928634e-06, "loss": 1.1769, "step": 31380 }, { "epoch": 0.8268604524815363, "grad_norm": 8.522123336791992, "learning_rate": 1.5311535014484724e-06, "loss": 1.092, "step": 31390 }, { "epoch": 0.827123867726035, "grad_norm": 8.581297874450684, "learning_rate": 1.5266196763307362e-06, "loss": 1.256, "step": 31400 }, { "epoch": 0.8273872829705338, "grad_norm": 9.425699234008789, "learning_rate": 1.522092018839747e-06, "loss": 1.2262, "step": 31410 }, { "epoch": 0.8276506982150325, "grad_norm": 6.643393516540527, "learning_rate": 1.5175705322711053e-06, "loss": 1.149, "step": 31420 }, { "epoch": 0.8279141134595311, "grad_norm": 9.888055801391602, "learning_rate": 1.5130552199159277e-06, "loss": 1.1955, "step": 31430 }, { "epoch": 0.8281775287040299, "grad_norm": 9.682290077209473, "learning_rate": 1.508546085060827e-06, "loss": 1.1433, "step": 31440 }, { "epoch": 0.8284409439485286, "grad_norm": 8.159740447998047, "learning_rate": 1.504043130987929e-06, "loss": 1.2235, "step": 31450 }, { "epoch": 0.8287043591930274, "grad_norm": 7.289303302764893, "learning_rate": 1.4995463609748528e-06, "loss": 1.1531, "step": 31460 }, { "epoch": 0.8289677744375261, "grad_norm": 8.978445053100586, "learning_rate": 1.4950557782947183e-06, "loss": 1.1545, "step": 31470 }, { "epoch": 0.8292311896820249, "grad_norm": 6.11592960357666, "learning_rate": 1.4905713862161464e-06, "loss": 1.0461, "step": 31480 }, { "epoch": 0.8294946049265236, "grad_norm": 18.631071090698242, "learning_rate": 1.4860931880032447e-06, "loss": 1.0722, "step": 31490 }, { "epoch": 0.8297580201710223, "grad_norm": 7.065129280090332, "learning_rate": 1.4816211869156194e-06, "loss": 1.0993, "step": 31500 }, { "epoch": 0.8300214354155211, "grad_norm": 6.307096481323242, "learning_rate": 1.477155386208362e-06, "loss": 1.1809, "step": 31510 }, { "epoch": 0.8302848506600198, "grad_norm": 9.172913551330566, "learning_rate": 1.4726957891320516e-06, "loss": 1.1323, "step": 31520 }, { "epoch": 0.8305482659045186, "grad_norm": 9.846457481384277, "learning_rate": 1.4682423989327487e-06, "loss": 1.1859, "step": 31530 }, { "epoch": 0.8308116811490173, "grad_norm": 13.42812442779541, "learning_rate": 1.4637952188520032e-06, "loss": 1.2412, "step": 31540 }, { "epoch": 0.8310750963935161, "grad_norm": 9.484488487243652, "learning_rate": 1.4593542521268356e-06, "loss": 1.2043, "step": 31550 }, { "epoch": 0.8313385116380148, "grad_norm": 6.812854766845703, "learning_rate": 1.4549195019897512e-06, "loss": 1.0403, "step": 31560 }, { "epoch": 0.8316019268825136, "grad_norm": 9.425481796264648, "learning_rate": 1.4504909716687332e-06, "loss": 1.1561, "step": 31570 }, { "epoch": 0.8318653421270122, "grad_norm": 6.462778568267822, "learning_rate": 1.4460686643872212e-06, "loss": 1.0601, "step": 31580 }, { "epoch": 0.8321287573715109, "grad_norm": 10.865950584411621, "learning_rate": 1.4416525833641426e-06, "loss": 1.1366, "step": 31590 }, { "epoch": 0.8323921726160097, "grad_norm": 6.0995869636535645, "learning_rate": 1.4372427318138882e-06, "loss": 1.3141, "step": 31600 }, { "epoch": 0.8326555878605084, "grad_norm": 13.180307388305664, "learning_rate": 1.432839112946306e-06, "loss": 1.1443, "step": 31610 }, { "epoch": 0.8329190031050072, "grad_norm": 9.40456485748291, "learning_rate": 1.4284417299667208e-06, "loss": 1.1565, "step": 31620 }, { "epoch": 0.8331824183495059, "grad_norm": 9.118408203125, "learning_rate": 1.4240505860759079e-06, "loss": 1.1625, "step": 31630 }, { "epoch": 0.8334458335940047, "grad_norm": 12.277379989624023, "learning_rate": 1.4196656844701028e-06, "loss": 1.1086, "step": 31640 }, { "epoch": 0.8337092488385034, "grad_norm": 11.576123237609863, "learning_rate": 1.4152870283410046e-06, "loss": 1.1474, "step": 31650 }, { "epoch": 0.8339726640830022, "grad_norm": 8.516620635986328, "learning_rate": 1.4109146208757563e-06, "loss": 1.1369, "step": 31660 }, { "epoch": 0.8342360793275009, "grad_norm": 7.281489372253418, "learning_rate": 1.4065484652569616e-06, "loss": 1.0954, "step": 31670 }, { "epoch": 0.8344994945719996, "grad_norm": 8.071185111999512, "learning_rate": 1.4021885646626677e-06, "loss": 1.2363, "step": 31680 }, { "epoch": 0.8347629098164984, "grad_norm": 13.013749122619629, "learning_rate": 1.397834922266369e-06, "loss": 1.2071, "step": 31690 }, { "epoch": 0.835026325060997, "grad_norm": 9.298771858215332, "learning_rate": 1.3934875412370075e-06, "loss": 1.1243, "step": 31700 }, { "epoch": 0.8352897403054959, "grad_norm": 11.528572082519531, "learning_rate": 1.3891464247389696e-06, "loss": 1.157, "step": 31710 }, { "epoch": 0.8355531555499945, "grad_norm": 9.760581970214844, "learning_rate": 1.3848115759320757e-06, "loss": 1.1696, "step": 31720 }, { "epoch": 0.8358165707944933, "grad_norm": 8.217482566833496, "learning_rate": 1.3804829979715862e-06, "loss": 1.1982, "step": 31730 }, { "epoch": 0.836079986038992, "grad_norm": 6.121707439422607, "learning_rate": 1.3761606940082018e-06, "loss": 1.2329, "step": 31740 }, { "epoch": 0.8363434012834908, "grad_norm": 9.802854537963867, "learning_rate": 1.3718446671880482e-06, "loss": 1.1399, "step": 31750 }, { "epoch": 0.8366068165279895, "grad_norm": 9.835196495056152, "learning_rate": 1.367534920652689e-06, "loss": 1.1493, "step": 31760 }, { "epoch": 0.8368702317724882, "grad_norm": 5.81845235824585, "learning_rate": 1.363231457539116e-06, "loss": 1.1631, "step": 31770 }, { "epoch": 0.837133647016987, "grad_norm": 10.56919002532959, "learning_rate": 1.3589342809797423e-06, "loss": 1.1535, "step": 31780 }, { "epoch": 0.8373970622614857, "grad_norm": 6.401373863220215, "learning_rate": 1.3546433941024084e-06, "loss": 1.0689, "step": 31790 }, { "epoch": 0.8376604775059845, "grad_norm": 9.20201587677002, "learning_rate": 1.3503588000303791e-06, "loss": 1.217, "step": 31800 }, { "epoch": 0.8379238927504832, "grad_norm": 15.267066955566406, "learning_rate": 1.346080501882332e-06, "loss": 1.2464, "step": 31810 }, { "epoch": 0.838187307994982, "grad_norm": 13.017342567443848, "learning_rate": 1.3418085027723703e-06, "loss": 1.1573, "step": 31820 }, { "epoch": 0.8384507232394807, "grad_norm": 7.63369083404541, "learning_rate": 1.3375428058100081e-06, "loss": 1.0973, "step": 31830 }, { "epoch": 0.8387141384839795, "grad_norm": 9.179638862609863, "learning_rate": 1.3332834141001671e-06, "loss": 1.1093, "step": 31840 }, { "epoch": 0.8389775537284782, "grad_norm": 14.927624702453613, "learning_rate": 1.3290303307431918e-06, "loss": 1.1711, "step": 31850 }, { "epoch": 0.8392409689729768, "grad_norm": 8.259521484375, "learning_rate": 1.3247835588348211e-06, "loss": 1.291, "step": 31860 }, { "epoch": 0.8395043842174756, "grad_norm": 12.4782075881958, "learning_rate": 1.3205431014662106e-06, "loss": 1.1318, "step": 31870 }, { "epoch": 0.8397677994619743, "grad_norm": 8.319181442260742, "learning_rate": 1.3163089617239199e-06, "loss": 1.1766, "step": 31880 }, { "epoch": 0.8400312147064731, "grad_norm": 9.566555976867676, "learning_rate": 1.3120811426898972e-06, "loss": 1.2385, "step": 31890 }, { "epoch": 0.8402946299509718, "grad_norm": 7.779786109924316, "learning_rate": 1.3078596474415029e-06, "loss": 1.2096, "step": 31900 }, { "epoch": 0.8405580451954706, "grad_norm": 13.955842018127441, "learning_rate": 1.3036444790514913e-06, "loss": 1.1596, "step": 31910 }, { "epoch": 0.8408214604399693, "grad_norm": 5.74229097366333, "learning_rate": 1.2994356405880083e-06, "loss": 1.2116, "step": 31920 }, { "epoch": 0.8410848756844681, "grad_norm": 6.232998847961426, "learning_rate": 1.2952331351145976e-06, "loss": 1.1322, "step": 31930 }, { "epoch": 0.8413482909289668, "grad_norm": 16.023740768432617, "learning_rate": 1.2910369656901878e-06, "loss": 1.0641, "step": 31940 }, { "epoch": 0.8416117061734655, "grad_norm": 7.124833583831787, "learning_rate": 1.286847135369096e-06, "loss": 1.2735, "step": 31950 }, { "epoch": 0.8418751214179643, "grad_norm": 5.208352565765381, "learning_rate": 1.2826636472010312e-06, "loss": 1.2098, "step": 31960 }, { "epoch": 0.842138536662463, "grad_norm": 7.167943477630615, "learning_rate": 1.2784865042310779e-06, "loss": 1.1647, "step": 31970 }, { "epoch": 0.8424019519069618, "grad_norm": 7.650773048400879, "learning_rate": 1.2743157094997083e-06, "loss": 1.167, "step": 31980 }, { "epoch": 0.8426653671514605, "grad_norm": 5.722565650939941, "learning_rate": 1.2701512660427772e-06, "loss": 1.1989, "step": 31990 }, { "epoch": 0.8429287823959593, "grad_norm": 6.9415283203125, "learning_rate": 1.2659931768915001e-06, "loss": 1.216, "step": 32000 }, { "epoch": 0.8431921976404579, "grad_norm": 7.182684898376465, "learning_rate": 1.2618414450724825e-06, "loss": 1.1671, "step": 32010 }, { "epoch": 0.8434556128849566, "grad_norm": 9.801552772521973, "learning_rate": 1.2576960736077016e-06, "loss": 1.1655, "step": 32020 }, { "epoch": 0.8437190281294554, "grad_norm": 9.985540390014648, "learning_rate": 1.2535570655144969e-06, "loss": 1.2447, "step": 32030 }, { "epoch": 0.8439824433739541, "grad_norm": 15.624730110168457, "learning_rate": 1.249424423805584e-06, "loss": 1.137, "step": 32040 }, { "epoch": 0.8442458586184529, "grad_norm": 49.671630859375, "learning_rate": 1.2452981514890405e-06, "loss": 1.2742, "step": 32050 }, { "epoch": 0.8445092738629516, "grad_norm": 11.738286018371582, "learning_rate": 1.2411782515683045e-06, "loss": 1.1929, "step": 32060 }, { "epoch": 0.8447726891074504, "grad_norm": 9.703991889953613, "learning_rate": 1.2370647270421831e-06, "loss": 1.1553, "step": 32070 }, { "epoch": 0.8450361043519491, "grad_norm": 8.218047142028809, "learning_rate": 1.2329575809048423e-06, "loss": 1.1089, "step": 32080 }, { "epoch": 0.8452995195964479, "grad_norm": 10.105463981628418, "learning_rate": 1.2288568161457992e-06, "loss": 1.0829, "step": 32090 }, { "epoch": 0.8455629348409466, "grad_norm": 14.89763355255127, "learning_rate": 1.2247624357499288e-06, "loss": 1.0326, "step": 32100 }, { "epoch": 0.8458263500854453, "grad_norm": 11.415549278259277, "learning_rate": 1.2206744426974616e-06, "loss": 1.1758, "step": 32110 }, { "epoch": 0.8460897653299441, "grad_norm": 9.858724594116211, "learning_rate": 1.2165928399639748e-06, "loss": 1.1349, "step": 32120 }, { "epoch": 0.8463531805744428, "grad_norm": 9.781495094299316, "learning_rate": 1.2125176305204001e-06, "loss": 1.0875, "step": 32130 }, { "epoch": 0.8466165958189416, "grad_norm": 12.387809753417969, "learning_rate": 1.208448817333009e-06, "loss": 1.1417, "step": 32140 }, { "epoch": 0.8468800110634402, "grad_norm": 14.604401588439941, "learning_rate": 1.2043864033634178e-06, "loss": 1.1496, "step": 32150 }, { "epoch": 0.847143426307939, "grad_norm": 18.11266326904297, "learning_rate": 1.2003303915685937e-06, "loss": 1.0867, "step": 32160 }, { "epoch": 0.8474068415524377, "grad_norm": 6.647791385650635, "learning_rate": 1.1962807849008306e-06, "loss": 1.1521, "step": 32170 }, { "epoch": 0.8476702567969365, "grad_norm": 8.048619270324707, "learning_rate": 1.1922375863077707e-06, "loss": 1.1745, "step": 32180 }, { "epoch": 0.8479336720414352, "grad_norm": 6.168464660644531, "learning_rate": 1.1882007987323886e-06, "loss": 1.1516, "step": 32190 }, { "epoch": 0.8481970872859339, "grad_norm": 6.283260822296143, "learning_rate": 1.1841704251129927e-06, "loss": 1.0721, "step": 32200 }, { "epoch": 0.8484605025304327, "grad_norm": 16.147626876831055, "learning_rate": 1.1801464683832175e-06, "loss": 1.0989, "step": 32210 }, { "epoch": 0.8487239177749314, "grad_norm": 9.637327194213867, "learning_rate": 1.176128931472037e-06, "loss": 1.1807, "step": 32220 }, { "epoch": 0.8489873330194302, "grad_norm": 7.452266216278076, "learning_rate": 1.1721178173037417e-06, "loss": 1.2392, "step": 32230 }, { "epoch": 0.8492507482639289, "grad_norm": 13.176429748535156, "learning_rate": 1.1681131287979563e-06, "loss": 1.1803, "step": 32240 }, { "epoch": 0.8495141635084277, "grad_norm": 7.590648651123047, "learning_rate": 1.1641148688696235e-06, "loss": 1.1163, "step": 32250 }, { "epoch": 0.8497775787529264, "grad_norm": 17.394153594970703, "learning_rate": 1.1601230404290042e-06, "loss": 1.1477, "step": 32260 }, { "epoch": 0.8500409939974252, "grad_norm": 8.838793754577637, "learning_rate": 1.1561376463816843e-06, "loss": 1.1167, "step": 32270 }, { "epoch": 0.8503044092419239, "grad_norm": 8.388134002685547, "learning_rate": 1.1521586896285608e-06, "loss": 1.1269, "step": 32280 }, { "epoch": 0.8505678244864225, "grad_norm": 6.998394966125488, "learning_rate": 1.1481861730658494e-06, "loss": 1.1771, "step": 32290 }, { "epoch": 0.8508312397309213, "grad_norm": 9.368738174438477, "learning_rate": 1.1442200995850772e-06, "loss": 1.0813, "step": 32300 }, { "epoch": 0.85109465497542, "grad_norm": 14.990349769592285, "learning_rate": 1.140260472073078e-06, "loss": 1.1932, "step": 32310 }, { "epoch": 0.8513580702199188, "grad_norm": 15.991227149963379, "learning_rate": 1.1363072934119957e-06, "loss": 1.1258, "step": 32320 }, { "epoch": 0.8516214854644175, "grad_norm": 8.31185245513916, "learning_rate": 1.1323605664792824e-06, "loss": 1.1251, "step": 32330 }, { "epoch": 0.8518849007089163, "grad_norm": 17.082847595214844, "learning_rate": 1.1284202941476908e-06, "loss": 1.27, "step": 32340 }, { "epoch": 0.852148315953415, "grad_norm": 8.107457160949707, "learning_rate": 1.1244864792852783e-06, "loss": 1.2204, "step": 32350 }, { "epoch": 0.8524117311979138, "grad_norm": 9.99099063873291, "learning_rate": 1.1205591247554004e-06, "loss": 1.1258, "step": 32360 }, { "epoch": 0.8526751464424125, "grad_norm": 5.88415002822876, "learning_rate": 1.1166382334167081e-06, "loss": 1.1274, "step": 32370 }, { "epoch": 0.8529385616869112, "grad_norm": 7.295480728149414, "learning_rate": 1.1127238081231517e-06, "loss": 1.1829, "step": 32380 }, { "epoch": 0.85320197693141, "grad_norm": 7.197556495666504, "learning_rate": 1.1088158517239755e-06, "loss": 1.0533, "step": 32390 }, { "epoch": 0.8534653921759087, "grad_norm": 7.740823268890381, "learning_rate": 1.1049143670637118e-06, "loss": 1.0898, "step": 32400 }, { "epoch": 0.8537288074204075, "grad_norm": 8.950605392456055, "learning_rate": 1.1010193569821813e-06, "loss": 1.1399, "step": 32410 }, { "epoch": 0.8539922226649062, "grad_norm": 8.77425479888916, "learning_rate": 1.0971308243144985e-06, "loss": 1.2119, "step": 32420 }, { "epoch": 0.854255637909405, "grad_norm": 8.838737487792969, "learning_rate": 1.0932487718910555e-06, "loss": 1.1894, "step": 32430 }, { "epoch": 0.8545190531539036, "grad_norm": 5.256621360778809, "learning_rate": 1.0893732025375348e-06, "loss": 1.0847, "step": 32440 }, { "epoch": 0.8547824683984024, "grad_norm": 12.977181434631348, "learning_rate": 1.0855041190748915e-06, "loss": 1.1446, "step": 32450 }, { "epoch": 0.8550458836429011, "grad_norm": 5.654670238494873, "learning_rate": 1.081641524319369e-06, "loss": 1.1205, "step": 32460 }, { "epoch": 0.8553092988873998, "grad_norm": 20.45656967163086, "learning_rate": 1.07778542108248e-06, "loss": 1.1704, "step": 32470 }, { "epoch": 0.8555727141318986, "grad_norm": 20.9130916595459, "learning_rate": 1.0739358121710152e-06, "loss": 1.1356, "step": 32480 }, { "epoch": 0.8558361293763973, "grad_norm": 6.402110576629639, "learning_rate": 1.0700927003870398e-06, "loss": 1.325, "step": 32490 }, { "epoch": 0.8560995446208961, "grad_norm": 8.539005279541016, "learning_rate": 1.0662560885278883e-06, "loss": 1.0658, "step": 32500 }, { "epoch": 0.8563629598653948, "grad_norm": 9.427186012268066, "learning_rate": 1.0624259793861647e-06, "loss": 1.0978, "step": 32510 }, { "epoch": 0.8566263751098936, "grad_norm": 18.484560012817383, "learning_rate": 1.0586023757497355e-06, "loss": 1.1965, "step": 32520 }, { "epoch": 0.8568897903543923, "grad_norm": 13.262128829956055, "learning_rate": 1.0547852804017388e-06, "loss": 1.2014, "step": 32530 }, { "epoch": 0.8571532055988911, "grad_norm": 12.884847640991211, "learning_rate": 1.0509746961205701e-06, "loss": 1.1143, "step": 32540 }, { "epoch": 0.8574166208433898, "grad_norm": 6.777890682220459, "learning_rate": 1.047170625679892e-06, "loss": 1.1649, "step": 32550 }, { "epoch": 0.8576800360878885, "grad_norm": 8.76011848449707, "learning_rate": 1.0433730718486168e-06, "loss": 1.1533, "step": 32560 }, { "epoch": 0.8579434513323873, "grad_norm": 8.164602279663086, "learning_rate": 1.0395820373909183e-06, "loss": 1.2923, "step": 32570 }, { "epoch": 0.8582068665768859, "grad_norm": 21.255773544311523, "learning_rate": 1.0357975250662256e-06, "loss": 1.2673, "step": 32580 }, { "epoch": 0.8584702818213847, "grad_norm": 13.915183067321777, "learning_rate": 1.032019537629223e-06, "loss": 1.1797, "step": 32590 }, { "epoch": 0.8587336970658834, "grad_norm": 6.027990818023682, "learning_rate": 1.0282480778298365e-06, "loss": 1.1785, "step": 32600 }, { "epoch": 0.8589971123103822, "grad_norm": 14.588454246520996, "learning_rate": 1.0244831484132523e-06, "loss": 1.2101, "step": 32610 }, { "epoch": 0.8592605275548809, "grad_norm": 6.58778715133667, "learning_rate": 1.0207247521198937e-06, "loss": 1.0428, "step": 32620 }, { "epoch": 0.8595239427993796, "grad_norm": 9.921648979187012, "learning_rate": 1.0169728916854305e-06, "loss": 1.1358, "step": 32630 }, { "epoch": 0.8597873580438784, "grad_norm": 8.758599281311035, "learning_rate": 1.0132275698407813e-06, "loss": 1.1278, "step": 32640 }, { "epoch": 0.8600507732883771, "grad_norm": 7.8956804275512695, "learning_rate": 1.0094887893120964e-06, "loss": 1.1941, "step": 32650 }, { "epoch": 0.8603141885328759, "grad_norm": 9.253400802612305, "learning_rate": 1.0057565528207725e-06, "loss": 1.1498, "step": 32660 }, { "epoch": 0.8605776037773746, "grad_norm": 8.5781888961792, "learning_rate": 1.002030863083443e-06, "loss": 1.1951, "step": 32670 }, { "epoch": 0.8608410190218734, "grad_norm": 11.353352546691895, "learning_rate": 9.983117228119643e-07, "loss": 1.2482, "step": 32680 }, { "epoch": 0.8611044342663721, "grad_norm": 10.097792625427246, "learning_rate": 9.945991347134398e-07, "loss": 1.1426, "step": 32690 }, { "epoch": 0.8613678495108709, "grad_norm": 17.720996856689453, "learning_rate": 9.908931014901991e-07, "loss": 1.0248, "step": 32700 }, { "epoch": 0.8616312647553696, "grad_norm": 22.136411666870117, "learning_rate": 9.871936258397962e-07, "loss": 1.2, "step": 32710 }, { "epoch": 0.8618946799998682, "grad_norm": 9.966521263122559, "learning_rate": 9.835007104550187e-07, "loss": 1.137, "step": 32720 }, { "epoch": 0.862158095244367, "grad_norm": 7.82261848449707, "learning_rate": 9.79814358023874e-07, "loss": 1.0849, "step": 32730 }, { "epoch": 0.8624215104888657, "grad_norm": 7.936743259429932, "learning_rate": 9.76134571229591e-07, "loss": 1.2001, "step": 32740 }, { "epoch": 0.8626849257333645, "grad_norm": 10.880586624145508, "learning_rate": 9.724613527506287e-07, "loss": 1.1676, "step": 32750 }, { "epoch": 0.8629483409778632, "grad_norm": 9.771432876586914, "learning_rate": 9.68794705260655e-07, "loss": 1.119, "step": 32760 }, { "epoch": 0.863211756222362, "grad_norm": 15.077018737792969, "learning_rate": 9.651346314285614e-07, "loss": 1.2368, "step": 32770 }, { "epoch": 0.8634751714668607, "grad_norm": 15.520211219787598, "learning_rate": 9.614811339184515e-07, "loss": 1.2019, "step": 32780 }, { "epoch": 0.8637385867113595, "grad_norm": 7.021361351013184, "learning_rate": 9.578342153896403e-07, "loss": 1.2731, "step": 32790 }, { "epoch": 0.8640020019558582, "grad_norm": 10.737369537353516, "learning_rate": 9.541938784966598e-07, "loss": 1.1596, "step": 32800 }, { "epoch": 0.8642654172003569, "grad_norm": 8.75694465637207, "learning_rate": 9.505601258892483e-07, "loss": 1.1587, "step": 32810 }, { "epoch": 0.8645288324448557, "grad_norm": 7.977107524871826, "learning_rate": 9.469329602123512e-07, "loss": 1.1878, "step": 32820 }, { "epoch": 0.8647922476893544, "grad_norm": 19.376012802124023, "learning_rate": 9.433123841061165e-07, "loss": 1.2563, "step": 32830 }, { "epoch": 0.8650556629338532, "grad_norm": 13.567084312438965, "learning_rate": 9.396984002059029e-07, "loss": 1.0984, "step": 32840 }, { "epoch": 0.8653190781783519, "grad_norm": 12.21002197265625, "learning_rate": 9.360910111422627e-07, "loss": 1.2243, "step": 32850 }, { "epoch": 0.8655824934228507, "grad_norm": 14.12449836730957, "learning_rate": 9.324902195409535e-07, "loss": 1.1744, "step": 32860 }, { "epoch": 0.8658459086673493, "grad_norm": 7.414440631866455, "learning_rate": 9.288960280229331e-07, "loss": 1.1798, "step": 32870 }, { "epoch": 0.8661093239118481, "grad_norm": 7.436563491821289, "learning_rate": 9.253084392043466e-07, "loss": 1.0936, "step": 32880 }, { "epoch": 0.8663727391563468, "grad_norm": 11.315138816833496, "learning_rate": 9.217274556965372e-07, "loss": 1.2198, "step": 32890 }, { "epoch": 0.8666361544008455, "grad_norm": 7.812894344329834, "learning_rate": 9.181530801060445e-07, "loss": 1.1023, "step": 32900 }, { "epoch": 0.8668995696453443, "grad_norm": 6.790420055389404, "learning_rate": 9.145853150345907e-07, "loss": 1.1526, "step": 32910 }, { "epoch": 0.867162984889843, "grad_norm": 11.079233169555664, "learning_rate": 9.11024163079094e-07, "loss": 1.1071, "step": 32920 }, { "epoch": 0.8674264001343418, "grad_norm": 7.48032808303833, "learning_rate": 9.07469626831653e-07, "loss": 1.1058, "step": 32930 }, { "epoch": 0.8676898153788405, "grad_norm": 9.705724716186523, "learning_rate": 9.039217088795515e-07, "loss": 1.1675, "step": 32940 }, { "epoch": 0.8679532306233393, "grad_norm": 5.903825759887695, "learning_rate": 9.003804118052617e-07, "loss": 1.1198, "step": 32950 }, { "epoch": 0.868216645867838, "grad_norm": 11.331876754760742, "learning_rate": 8.968457381864282e-07, "loss": 1.2655, "step": 32960 }, { "epoch": 0.8684800611123368, "grad_norm": 13.345621109008789, "learning_rate": 8.933176905958796e-07, "loss": 1.2225, "step": 32970 }, { "epoch": 0.8687434763568355, "grad_norm": 9.482664108276367, "learning_rate": 8.897962716016262e-07, "loss": 1.111, "step": 32980 }, { "epoch": 0.8690068916013342, "grad_norm": 10.475421905517578, "learning_rate": 8.862814837668388e-07, "loss": 1.1319, "step": 32990 }, { "epoch": 0.869270306845833, "grad_norm": 12.020146369934082, "learning_rate": 8.827733296498753e-07, "loss": 1.156, "step": 33000 }, { "epoch": 0.8695337220903316, "grad_norm": 8.262234687805176, "learning_rate": 8.79271811804262e-07, "loss": 1.054, "step": 33010 }, { "epoch": 0.8697971373348304, "grad_norm": 7.831429481506348, "learning_rate": 8.757769327786892e-07, "loss": 1.2027, "step": 33020 }, { "epoch": 0.8700605525793291, "grad_norm": 13.1566162109375, "learning_rate": 8.722886951170229e-07, "loss": 1.1941, "step": 33030 }, { "epoch": 0.8703239678238279, "grad_norm": 9.970398902893066, "learning_rate": 8.688071013582889e-07, "loss": 1.2146, "step": 33040 }, { "epoch": 0.8705873830683266, "grad_norm": 11.162369728088379, "learning_rate": 8.653321540366754e-07, "loss": 1.1333, "step": 33050 }, { "epoch": 0.8708507983128254, "grad_norm": 11.680798530578613, "learning_rate": 8.618638556815429e-07, "loss": 1.1461, "step": 33060 }, { "epoch": 0.8711142135573241, "grad_norm": 13.578390121459961, "learning_rate": 8.584022088174004e-07, "loss": 1.157, "step": 33070 }, { "epoch": 0.8713776288018228, "grad_norm": 10.638934135437012, "learning_rate": 8.549472159639216e-07, "loss": 1.1288, "step": 33080 }, { "epoch": 0.8716410440463216, "grad_norm": 8.945267677307129, "learning_rate": 8.514988796359402e-07, "loss": 1.1521, "step": 33090 }, { "epoch": 0.8719044592908203, "grad_norm": 7.636659145355225, "learning_rate": 8.480572023434375e-07, "loss": 1.1185, "step": 33100 }, { "epoch": 0.8721678745353191, "grad_norm": 10.493993759155273, "learning_rate": 8.446221865915483e-07, "loss": 1.166, "step": 33110 }, { "epoch": 0.8724312897798178, "grad_norm": 11.306140899658203, "learning_rate": 8.411938348805648e-07, "loss": 1.1564, "step": 33120 }, { "epoch": 0.8726947050243166, "grad_norm": 9.309571266174316, "learning_rate": 8.377721497059222e-07, "loss": 1.1599, "step": 33130 }, { "epoch": 0.8729581202688153, "grad_norm": 8.761427879333496, "learning_rate": 8.34357133558209e-07, "loss": 1.1623, "step": 33140 }, { "epoch": 0.8732215355133139, "grad_norm": 8.290459632873535, "learning_rate": 8.309487889231549e-07, "loss": 1.1039, "step": 33150 }, { "epoch": 0.8734849507578127, "grad_norm": 6.5263590812683105, "learning_rate": 8.275471182816319e-07, "loss": 1.1608, "step": 33160 }, { "epoch": 0.8737483660023114, "grad_norm": 6.421176910400391, "learning_rate": 8.241521241096595e-07, "loss": 1.1158, "step": 33170 }, { "epoch": 0.8740117812468102, "grad_norm": 10.207563400268555, "learning_rate": 8.207638088783976e-07, "loss": 1.1064, "step": 33180 }, { "epoch": 0.8742751964913089, "grad_norm": 7.8762736320495605, "learning_rate": 8.173821750541389e-07, "loss": 1.1368, "step": 33190 }, { "epoch": 0.8745386117358077, "grad_norm": 10.871671676635742, "learning_rate": 8.140072250983166e-07, "loss": 1.2476, "step": 33200 }, { "epoch": 0.8748020269803064, "grad_norm": 10.234232902526855, "learning_rate": 8.106389614674992e-07, "loss": 1.1057, "step": 33210 }, { "epoch": 0.8750654422248052, "grad_norm": 10.188984870910645, "learning_rate": 8.072773866133854e-07, "loss": 1.175, "step": 33220 }, { "epoch": 0.8753288574693039, "grad_norm": 6.831078052520752, "learning_rate": 8.039225029828113e-07, "loss": 1.1081, "step": 33230 }, { "epoch": 0.8755922727138026, "grad_norm": 14.557104110717773, "learning_rate": 8.005743130177357e-07, "loss": 1.1407, "step": 33240 }, { "epoch": 0.8758556879583014, "grad_norm": 12.310924530029297, "learning_rate": 7.97232819155247e-07, "loss": 1.1935, "step": 33250 }, { "epoch": 0.8761191032028001, "grad_norm": 12.728763580322266, "learning_rate": 7.93898023827564e-07, "loss": 1.1896, "step": 33260 }, { "epoch": 0.8763825184472989, "grad_norm": 10.874765396118164, "learning_rate": 7.905699294620228e-07, "loss": 1.1816, "step": 33270 }, { "epoch": 0.8766459336917976, "grad_norm": 15.096443176269531, "learning_rate": 7.87248538481088e-07, "loss": 1.1601, "step": 33280 }, { "epoch": 0.8769093489362964, "grad_norm": 13.569689750671387, "learning_rate": 7.839338533023433e-07, "loss": 1.0653, "step": 33290 }, { "epoch": 0.877172764180795, "grad_norm": 7.381311893463135, "learning_rate": 7.806258763384888e-07, "loss": 0.9433, "step": 33300 }, { "epoch": 0.8774361794252938, "grad_norm": 14.047696113586426, "learning_rate": 7.77324609997343e-07, "loss": 1.0483, "step": 33310 }, { "epoch": 0.8776995946697925, "grad_norm": 20.547494888305664, "learning_rate": 7.740300566818437e-07, "loss": 1.1856, "step": 33320 }, { "epoch": 0.8779630099142912, "grad_norm": 7.950221538543701, "learning_rate": 7.707422187900349e-07, "loss": 1.245, "step": 33330 }, { "epoch": 0.87822642515879, "grad_norm": 90.43871307373047, "learning_rate": 7.67461098715081e-07, "loss": 1.1519, "step": 33340 }, { "epoch": 0.8784898404032887, "grad_norm": 8.950170516967773, "learning_rate": 7.641866988452506e-07, "loss": 1.1112, "step": 33350 }, { "epoch": 0.8787532556477875, "grad_norm": 9.244402885437012, "learning_rate": 7.609190215639217e-07, "loss": 0.997, "step": 33360 }, { "epoch": 0.8790166708922862, "grad_norm": 9.939420700073242, "learning_rate": 7.576580692495794e-07, "loss": 1.1845, "step": 33370 }, { "epoch": 0.879280086136785, "grad_norm": 9.907596588134766, "learning_rate": 7.544038442758184e-07, "loss": 1.135, "step": 33380 }, { "epoch": 0.8795435013812837, "grad_norm": 9.714869499206543, "learning_rate": 7.511563490113283e-07, "loss": 1.177, "step": 33390 }, { "epoch": 0.8798069166257825, "grad_norm": 25.097639083862305, "learning_rate": 7.479155858199083e-07, "loss": 1.2004, "step": 33400 }, { "epoch": 0.8800703318702812, "grad_norm": 12.760353088378906, "learning_rate": 7.446815570604527e-07, "loss": 1.1779, "step": 33410 }, { "epoch": 0.8803337471147799, "grad_norm": 6.628438949584961, "learning_rate": 7.414542650869517e-07, "loss": 1.3113, "step": 33420 }, { "epoch": 0.8805971623592787, "grad_norm": 8.37238883972168, "learning_rate": 7.382337122484995e-07, "loss": 1.2333, "step": 33430 }, { "epoch": 0.8808605776037773, "grad_norm": 9.207893371582031, "learning_rate": 7.350199008892778e-07, "loss": 1.1481, "step": 33440 }, { "epoch": 0.8811239928482761, "grad_norm": 10.915332794189453, "learning_rate": 7.318128333485663e-07, "loss": 1.1524, "step": 33450 }, { "epoch": 0.8813874080927748, "grad_norm": 14.058781623840332, "learning_rate": 7.286125119607346e-07, "loss": 1.0823, "step": 33460 }, { "epoch": 0.8816508233372736, "grad_norm": 8.432967185974121, "learning_rate": 7.254189390552369e-07, "loss": 1.1943, "step": 33470 }, { "epoch": 0.8819142385817723, "grad_norm": 8.47348690032959, "learning_rate": 7.222321169566227e-07, "loss": 1.145, "step": 33480 }, { "epoch": 0.8821776538262711, "grad_norm": 10.444677352905273, "learning_rate": 7.190520479845253e-07, "loss": 1.2015, "step": 33490 }, { "epoch": 0.8824410690707698, "grad_norm": 16.779993057250977, "learning_rate": 7.1587873445366e-07, "loss": 1.0062, "step": 33500 }, { "epoch": 0.8827044843152685, "grad_norm": 14.761578559875488, "learning_rate": 7.127121786738278e-07, "loss": 1.2153, "step": 33510 }, { "epoch": 0.8829678995597673, "grad_norm": 11.269522666931152, "learning_rate": 7.095523829499107e-07, "loss": 1.1257, "step": 33520 }, { "epoch": 0.883231314804266, "grad_norm": 9.841747283935547, "learning_rate": 7.063993495818666e-07, "loss": 1.1598, "step": 33530 }, { "epoch": 0.8834947300487648, "grad_norm": 13.81451416015625, "learning_rate": 7.032530808647364e-07, "loss": 1.1321, "step": 33540 }, { "epoch": 0.8837581452932635, "grad_norm": 10.095303535461426, "learning_rate": 7.001135790886315e-07, "loss": 1.154, "step": 33550 }, { "epoch": 0.8840215605377623, "grad_norm": 8.073668479919434, "learning_rate": 6.969808465387451e-07, "loss": 1.2709, "step": 33560 }, { "epoch": 0.884284975782261, "grad_norm": 9.129645347595215, "learning_rate": 6.938548854953364e-07, "loss": 1.1518, "step": 33570 }, { "epoch": 0.8845483910267598, "grad_norm": 8.365315437316895, "learning_rate": 6.907356982337377e-07, "loss": 1.1327, "step": 33580 }, { "epoch": 0.8848118062712584, "grad_norm": 12.932005882263184, "learning_rate": 6.876232870243515e-07, "loss": 1.28, "step": 33590 }, { "epoch": 0.8850752215157571, "grad_norm": 10.146544456481934, "learning_rate": 6.845176541326504e-07, "loss": 1.1187, "step": 33600 }, { "epoch": 0.8853386367602559, "grad_norm": 7.70832633972168, "learning_rate": 6.814188018191703e-07, "loss": 1.1366, "step": 33610 }, { "epoch": 0.8856020520047546, "grad_norm": 8.484920501708984, "learning_rate": 6.783267323395082e-07, "loss": 1.0771, "step": 33620 }, { "epoch": 0.8858654672492534, "grad_norm": 7.143806457519531, "learning_rate": 6.752414479443326e-07, "loss": 1.1287, "step": 33630 }, { "epoch": 0.8861288824937521, "grad_norm": 14.096282958984375, "learning_rate": 6.721629508793648e-07, "loss": 1.1925, "step": 33640 }, { "epoch": 0.8863922977382509, "grad_norm": 9.240187644958496, "learning_rate": 6.690912433853924e-07, "loss": 1.1855, "step": 33650 }, { "epoch": 0.8866557129827496, "grad_norm": 11.503029823303223, "learning_rate": 6.660263276982593e-07, "loss": 1.2127, "step": 33660 }, { "epoch": 0.8869191282272483, "grad_norm": 7.248349189758301, "learning_rate": 6.629682060488596e-07, "loss": 1.1358, "step": 33670 }, { "epoch": 0.8871825434717471, "grad_norm": 13.219300270080566, "learning_rate": 6.599168806631496e-07, "loss": 1.1697, "step": 33680 }, { "epoch": 0.8874459587162458, "grad_norm": 6.271262168884277, "learning_rate": 6.568723537621368e-07, "loss": 1.1618, "step": 33690 }, { "epoch": 0.8877093739607446, "grad_norm": 8.143566131591797, "learning_rate": 6.53834627561879e-07, "loss": 1.1378, "step": 33700 }, { "epoch": 0.8879727892052433, "grad_norm": 9.58728313446045, "learning_rate": 6.50803704273485e-07, "loss": 1.1235, "step": 33710 }, { "epoch": 0.888236204449742, "grad_norm": 17.685007095336914, "learning_rate": 6.477795861031122e-07, "loss": 1.1556, "step": 33720 }, { "epoch": 0.8884996196942407, "grad_norm": 15.417120933532715, "learning_rate": 6.4476227525196e-07, "loss": 1.0611, "step": 33730 }, { "epoch": 0.8887630349387395, "grad_norm": 12.311787605285645, "learning_rate": 6.417517739162804e-07, "loss": 1.2015, "step": 33740 }, { "epoch": 0.8890264501832382, "grad_norm": 11.058486938476562, "learning_rate": 6.387480842873639e-07, "loss": 1.0907, "step": 33750 }, { "epoch": 0.8892898654277369, "grad_norm": 11.436271667480469, "learning_rate": 6.357512085515439e-07, "loss": 1.14, "step": 33760 }, { "epoch": 0.8895532806722357, "grad_norm": 14.99983024597168, "learning_rate": 6.327611488901964e-07, "loss": 1.1333, "step": 33770 }, { "epoch": 0.8898166959167344, "grad_norm": 10.018074035644531, "learning_rate": 6.297779074797306e-07, "loss": 1.1686, "step": 33780 }, { "epoch": 0.8900801111612332, "grad_norm": 12.41059684753418, "learning_rate": 6.268014864915972e-07, "loss": 1.1251, "step": 33790 }, { "epoch": 0.8903435264057319, "grad_norm": 10.920182228088379, "learning_rate": 6.238318880922845e-07, "loss": 1.1535, "step": 33800 }, { "epoch": 0.8906069416502307, "grad_norm": 8.647796630859375, "learning_rate": 6.208691144433077e-07, "loss": 1.0321, "step": 33810 }, { "epoch": 0.8908703568947294, "grad_norm": 17.016326904296875, "learning_rate": 6.179131677012218e-07, "loss": 1.1692, "step": 33820 }, { "epoch": 0.8911337721392282, "grad_norm": 6.892329692840576, "learning_rate": 6.149640500176068e-07, "loss": 1.1825, "step": 33830 }, { "epoch": 0.8913971873837269, "grad_norm": 9.525321960449219, "learning_rate": 6.120217635390746e-07, "loss": 1.121, "step": 33840 }, { "epoch": 0.8916606026282256, "grad_norm": 13.722214698791504, "learning_rate": 6.090863104072664e-07, "loss": 1.1909, "step": 33850 }, { "epoch": 0.8919240178727244, "grad_norm": 11.378910064697266, "learning_rate": 6.061576927588452e-07, "loss": 1.1035, "step": 33860 }, { "epoch": 0.892187433117223, "grad_norm": 16.482786178588867, "learning_rate": 6.032359127255027e-07, "loss": 1.1253, "step": 33870 }, { "epoch": 0.8924508483617218, "grad_norm": 14.005863189697266, "learning_rate": 6.003209724339531e-07, "loss": 1.1326, "step": 33880 }, { "epoch": 0.8927142636062205, "grad_norm": 10.064102172851562, "learning_rate": 5.974128740059271e-07, "loss": 1.1292, "step": 33890 }, { "epoch": 0.8929776788507193, "grad_norm": 9.63267993927002, "learning_rate": 5.945116195581813e-07, "loss": 1.1384, "step": 33900 }, { "epoch": 0.893241094095218, "grad_norm": 15.928169250488281, "learning_rate": 5.916172112024898e-07, "loss": 1.1708, "step": 33910 }, { "epoch": 0.8935045093397168, "grad_norm": 12.025712013244629, "learning_rate": 5.887296510456408e-07, "loss": 1.2091, "step": 33920 }, { "epoch": 0.8937679245842155, "grad_norm": 7.9270920753479, "learning_rate": 5.858489411894397e-07, "loss": 1.1342, "step": 33930 }, { "epoch": 0.8940313398287142, "grad_norm": 16.032176971435547, "learning_rate": 5.829750837307068e-07, "loss": 1.1861, "step": 33940 }, { "epoch": 0.894294755073213, "grad_norm": 6.138912677764893, "learning_rate": 5.801080807612691e-07, "loss": 1.0836, "step": 33950 }, { "epoch": 0.8945581703177117, "grad_norm": 8.317326545715332, "learning_rate": 5.772479343679715e-07, "loss": 1.0801, "step": 33960 }, { "epoch": 0.8948215855622105, "grad_norm": 12.230846405029297, "learning_rate": 5.743946466326666e-07, "loss": 1.1327, "step": 33970 }, { "epoch": 0.8950850008067092, "grad_norm": 13.969659805297852, "learning_rate": 5.715482196322098e-07, "loss": 1.1457, "step": 33980 }, { "epoch": 0.895348416051208, "grad_norm": 15.057814598083496, "learning_rate": 5.687086554384658e-07, "loss": 1.1134, "step": 33990 }, { "epoch": 0.8956118312957067, "grad_norm": 8.711353302001953, "learning_rate": 5.65875956118308e-07, "loss": 1.0638, "step": 34000 }, { "epoch": 0.8958752465402054, "grad_norm": 7.880745887756348, "learning_rate": 5.63050123733605e-07, "loss": 1.1945, "step": 34010 }, { "epoch": 0.8961386617847041, "grad_norm": 10.515129089355469, "learning_rate": 5.602311603412336e-07, "loss": 1.0468, "step": 34020 }, { "epoch": 0.8964020770292028, "grad_norm": 9.075164794921875, "learning_rate": 5.57419067993068e-07, "loss": 1.0661, "step": 34030 }, { "epoch": 0.8966654922737016, "grad_norm": 16.623886108398438, "learning_rate": 5.546138487359809e-07, "loss": 1.0438, "step": 34040 }, { "epoch": 0.8969289075182003, "grad_norm": 14.254822731018066, "learning_rate": 5.518155046118434e-07, "loss": 1.2021, "step": 34050 }, { "epoch": 0.8971923227626991, "grad_norm": 5.599329948425293, "learning_rate": 5.490240376575207e-07, "loss": 1.1711, "step": 34060 }, { "epoch": 0.8974557380071978, "grad_norm": 22.06282615661621, "learning_rate": 5.46239449904874e-07, "loss": 1.1248, "step": 34070 }, { "epoch": 0.8977191532516966, "grad_norm": 11.838096618652344, "learning_rate": 5.434617433807598e-07, "loss": 1.2427, "step": 34080 }, { "epoch": 0.8979825684961953, "grad_norm": 12.729964256286621, "learning_rate": 5.406909201070154e-07, "loss": 1.069, "step": 34090 }, { "epoch": 0.8982459837406941, "grad_norm": 12.28861141204834, "learning_rate": 5.379269821004795e-07, "loss": 1.1839, "step": 34100 }, { "epoch": 0.8985093989851928, "grad_norm": 9.865238189697266, "learning_rate": 5.351699313729763e-07, "loss": 1.2031, "step": 34110 }, { "epoch": 0.8987728142296915, "grad_norm": 15.585233688354492, "learning_rate": 5.324197699313105e-07, "loss": 1.2736, "step": 34120 }, { "epoch": 0.8990362294741903, "grad_norm": 9.083556175231934, "learning_rate": 5.296764997772819e-07, "loss": 1.1695, "step": 34130 }, { "epoch": 0.899299644718689, "grad_norm": 8.976278305053711, "learning_rate": 5.269401229076665e-07, "loss": 1.181, "step": 34140 }, { "epoch": 0.8995630599631877, "grad_norm": 9.39061164855957, "learning_rate": 5.242106413142245e-07, "loss": 1.1615, "step": 34150 }, { "epoch": 0.8998264752076864, "grad_norm": 9.694668769836426, "learning_rate": 5.214880569837e-07, "loss": 1.1567, "step": 34160 }, { "epoch": 0.9000898904521852, "grad_norm": 8.248828887939453, "learning_rate": 5.187723718978165e-07, "loss": 1.1277, "step": 34170 }, { "epoch": 0.9003533056966839, "grad_norm": 9.504439353942871, "learning_rate": 5.160635880332709e-07, "loss": 1.193, "step": 34180 }, { "epoch": 0.9006167209411826, "grad_norm": 10.71901798248291, "learning_rate": 5.133617073617425e-07, "loss": 1.1369, "step": 34190 }, { "epoch": 0.9008801361856814, "grad_norm": 7.78312873840332, "learning_rate": 5.106667318498837e-07, "loss": 1.1649, "step": 34200 }, { "epoch": 0.9011435514301801, "grad_norm": 11.043210983276367, "learning_rate": 5.079786634593175e-07, "loss": 1.0707, "step": 34210 }, { "epoch": 0.9014069666746789, "grad_norm": 8.47862434387207, "learning_rate": 5.052975041466457e-07, "loss": 1.0984, "step": 34220 }, { "epoch": 0.9016703819191776, "grad_norm": 10.428630828857422, "learning_rate": 5.026232558634337e-07, "loss": 1.125, "step": 34230 }, { "epoch": 0.9019337971636764, "grad_norm": 11.397324562072754, "learning_rate": 4.999559205562254e-07, "loss": 1.241, "step": 34240 }, { "epoch": 0.9021972124081751, "grad_norm": 10.000570297241211, "learning_rate": 4.972955001665258e-07, "loss": 1.1933, "step": 34250 }, { "epoch": 0.9024606276526739, "grad_norm": 7.003478050231934, "learning_rate": 4.946419966308058e-07, "loss": 1.1676, "step": 34260 }, { "epoch": 0.9027240428971726, "grad_norm": 12.588276863098145, "learning_rate": 4.919954118805093e-07, "loss": 1.048, "step": 34270 }, { "epoch": 0.9029874581416713, "grad_norm": 9.712860107421875, "learning_rate": 4.893557478420375e-07, "loss": 1.1383, "step": 34280 }, { "epoch": 0.90325087338617, "grad_norm": 13.447404861450195, "learning_rate": 4.867230064367567e-07, "loss": 1.1227, "step": 34290 }, { "epoch": 0.9035142886306687, "grad_norm": 14.642350196838379, "learning_rate": 4.840971895809921e-07, "loss": 1.1254, "step": 34300 }, { "epoch": 0.9037777038751675, "grad_norm": 11.322317123413086, "learning_rate": 4.814782991860323e-07, "loss": 1.2321, "step": 34310 }, { "epoch": 0.9040411191196662, "grad_norm": 13.353717803955078, "learning_rate": 4.788663371581204e-07, "loss": 1.1621, "step": 34320 }, { "epoch": 0.904304534364165, "grad_norm": 13.853574752807617, "learning_rate": 4.7626130539846105e-07, "loss": 1.221, "step": 34330 }, { "epoch": 0.9045679496086637, "grad_norm": 17.469995498657227, "learning_rate": 4.736632058032098e-07, "loss": 1.1177, "step": 34340 }, { "epoch": 0.9048313648531625, "grad_norm": 8.299126625061035, "learning_rate": 4.710720402634772e-07, "loss": 1.1688, "step": 34350 }, { "epoch": 0.9050947800976612, "grad_norm": 8.922969818115234, "learning_rate": 4.684878106653301e-07, "loss": 1.1105, "step": 34360 }, { "epoch": 0.9053581953421599, "grad_norm": 7.330361843109131, "learning_rate": 4.659105188897817e-07, "loss": 1.1551, "step": 34370 }, { "epoch": 0.9056216105866587, "grad_norm": 11.607035636901855, "learning_rate": 4.6334016681280126e-07, "loss": 1.2517, "step": 34380 }, { "epoch": 0.9058850258311574, "grad_norm": 8.082463264465332, "learning_rate": 4.607767563053023e-07, "loss": 1.1116, "step": 34390 }, { "epoch": 0.9061484410756562, "grad_norm": 8.443191528320312, "learning_rate": 4.5822028923314664e-07, "loss": 1.2351, "step": 34400 }, { "epoch": 0.9064118563201549, "grad_norm": 8.197968482971191, "learning_rate": 4.5567076745714144e-07, "loss": 1.3143, "step": 34410 }, { "epoch": 0.9066752715646537, "grad_norm": 8.638983726501465, "learning_rate": 4.5312819283304223e-07, "loss": 1.2197, "step": 34420 }, { "epoch": 0.9069386868091524, "grad_norm": 10.592315673828125, "learning_rate": 4.505925672115419e-07, "loss": 1.1241, "step": 34430 }, { "epoch": 0.9072021020536511, "grad_norm": 7.927863121032715, "learning_rate": 4.480638924382796e-07, "loss": 1.1288, "step": 34440 }, { "epoch": 0.9074655172981498, "grad_norm": 8.67530345916748, "learning_rate": 4.455421703538354e-07, "loss": 1.1285, "step": 34450 }, { "epoch": 0.9077289325426485, "grad_norm": 11.051576614379883, "learning_rate": 4.4302740279372426e-07, "loss": 1.2415, "step": 34460 }, { "epoch": 0.9079923477871473, "grad_norm": 7.310313701629639, "learning_rate": 4.4051959158840417e-07, "loss": 1.2447, "step": 34470 }, { "epoch": 0.908255763031646, "grad_norm": 13.749628067016602, "learning_rate": 4.3801873856326725e-07, "loss": 1.1648, "step": 34480 }, { "epoch": 0.9085191782761448, "grad_norm": 11.915290832519531, "learning_rate": 4.355248455386385e-07, "loss": 1.1078, "step": 34490 }, { "epoch": 0.9087825935206435, "grad_norm": 9.331755638122559, "learning_rate": 4.330379143297836e-07, "loss": 1.2672, "step": 34500 }, { "epoch": 0.9090460087651423, "grad_norm": 11.98525333404541, "learning_rate": 4.3055794674689457e-07, "loss": 1.1114, "step": 34510 }, { "epoch": 0.909309424009641, "grad_norm": 11.972107887268066, "learning_rate": 4.280849445950952e-07, "loss": 1.1328, "step": 34520 }, { "epoch": 0.9095728392541398, "grad_norm": 9.174764633178711, "learning_rate": 4.256189096744445e-07, "loss": 1.102, "step": 34530 }, { "epoch": 0.9098362544986385, "grad_norm": 12.22992992401123, "learning_rate": 4.2315984377992226e-07, "loss": 1.0238, "step": 34540 }, { "epoch": 0.9100996697431372, "grad_norm": 8.637992858886719, "learning_rate": 4.207077487014444e-07, "loss": 1.1902, "step": 34550 }, { "epoch": 0.910363084987636, "grad_norm": 10.551188468933105, "learning_rate": 4.182626262238465e-07, "loss": 1.1797, "step": 34560 }, { "epoch": 0.9106265002321347, "grad_norm": 10.139881134033203, "learning_rate": 4.158244781268894e-07, "loss": 1.1283, "step": 34570 }, { "epoch": 0.9108899154766334, "grad_norm": 9.849143028259277, "learning_rate": 4.133933061852602e-07, "loss": 1.3071, "step": 34580 }, { "epoch": 0.9111533307211321, "grad_norm": 10.614712715148926, "learning_rate": 4.109691121685677e-07, "loss": 1.099, "step": 34590 }, { "epoch": 0.9114167459656309, "grad_norm": 9.403322219848633, "learning_rate": 4.085518978413394e-07, "loss": 1.2214, "step": 34600 }, { "epoch": 0.9116801612101296, "grad_norm": 12.368501663208008, "learning_rate": 4.0614166496302454e-07, "loss": 1.2258, "step": 34610 }, { "epoch": 0.9119435764546284, "grad_norm": 8.237711906433105, "learning_rate": 4.0373841528799084e-07, "loss": 1.2282, "step": 34620 }, { "epoch": 0.9122069916991271, "grad_norm": 22.661710739135742, "learning_rate": 4.0134215056551793e-07, "loss": 1.1052, "step": 34630 }, { "epoch": 0.9124704069436258, "grad_norm": 9.472006797790527, "learning_rate": 3.9895287253981065e-07, "loss": 1.22, "step": 34640 }, { "epoch": 0.9127338221881246, "grad_norm": 8.794376373291016, "learning_rate": 3.9657058294997885e-07, "loss": 1.3375, "step": 34650 }, { "epoch": 0.9129972374326233, "grad_norm": 11.453802108764648, "learning_rate": 3.941952835300533e-07, "loss": 1.0883, "step": 34660 }, { "epoch": 0.9132606526771221, "grad_norm": 8.617049217224121, "learning_rate": 3.918269760089721e-07, "loss": 1.0843, "step": 34670 }, { "epoch": 0.9135240679216208, "grad_norm": 14.568305969238281, "learning_rate": 3.89465662110583e-07, "loss": 1.2083, "step": 34680 }, { "epoch": 0.9137874831661196, "grad_norm": 6.929413795471191, "learning_rate": 3.8711134355364665e-07, "loss": 1.1827, "step": 34690 }, { "epoch": 0.9140508984106183, "grad_norm": 13.545747756958008, "learning_rate": 3.847640220518345e-07, "loss": 1.1858, "step": 34700 }, { "epoch": 0.914314313655117, "grad_norm": 10.864808082580566, "learning_rate": 3.8242369931371647e-07, "loss": 1.1733, "step": 34710 }, { "epoch": 0.9145777288996157, "grad_norm": 11.020844459533691, "learning_rate": 3.800903770427744e-07, "loss": 1.1124, "step": 34720 }, { "epoch": 0.9148411441441144, "grad_norm": 12.68433666229248, "learning_rate": 3.7776405693739417e-07, "loss": 1.1573, "step": 34730 }, { "epoch": 0.9151045593886132, "grad_norm": 8.306726455688477, "learning_rate": 3.754447406908612e-07, "loss": 1.1885, "step": 34740 }, { "epoch": 0.9153679746331119, "grad_norm": 15.842392921447754, "learning_rate": 3.7313242999136856e-07, "loss": 1.1713, "step": 34750 }, { "epoch": 0.9156313898776107, "grad_norm": 6.022691249847412, "learning_rate": 3.708271265220087e-07, "loss": 1.0568, "step": 34760 }, { "epoch": 0.9158948051221094, "grad_norm": 6.986219882965088, "learning_rate": 3.6852883196076605e-07, "loss": 1.1169, "step": 34770 }, { "epoch": 0.9161582203666082, "grad_norm": 7.124515533447266, "learning_rate": 3.6623754798053357e-07, "loss": 1.1288, "step": 34780 }, { "epoch": 0.9164216356111069, "grad_norm": 8.407670974731445, "learning_rate": 3.6395327624909714e-07, "loss": 1.151, "step": 34790 }, { "epoch": 0.9166850508556056, "grad_norm": 9.635004997253418, "learning_rate": 3.616760184291368e-07, "loss": 1.0792, "step": 34800 }, { "epoch": 0.9169484661001044, "grad_norm": 14.190855979919434, "learning_rate": 3.59405776178231e-07, "loss": 1.038, "step": 34810 }, { "epoch": 0.9172118813446031, "grad_norm": 17.446435928344727, "learning_rate": 3.571425511488491e-07, "loss": 1.13, "step": 34820 }, { "epoch": 0.9174752965891019, "grad_norm": 15.851126670837402, "learning_rate": 3.548863449883522e-07, "loss": 1.1885, "step": 34830 }, { "epoch": 0.9177387118336006, "grad_norm": 12.545223236083984, "learning_rate": 3.5263715933899543e-07, "loss": 1.0151, "step": 34840 }, { "epoch": 0.9180021270780994, "grad_norm": 10.077058792114258, "learning_rate": 3.503949958379205e-07, "loss": 1.0801, "step": 34850 }, { "epoch": 0.918265542322598, "grad_norm": 4.87405252456665, "learning_rate": 3.481598561171595e-07, "loss": 1.1137, "step": 34860 }, { "epoch": 0.9185289575670968, "grad_norm": 15.157437324523926, "learning_rate": 3.4593174180363563e-07, "loss": 1.1275, "step": 34870 }, { "epoch": 0.9187923728115955, "grad_norm": 8.567394256591797, "learning_rate": 3.437106545191493e-07, "loss": 1.1791, "step": 34880 }, { "epoch": 0.9190557880560942, "grad_norm": 13.740571022033691, "learning_rate": 3.414965958803929e-07, "loss": 1.0996, "step": 34890 }, { "epoch": 0.919319203300593, "grad_norm": 12.134849548339844, "learning_rate": 3.3928956749894384e-07, "loss": 1.1445, "step": 34900 }, { "epoch": 0.9195826185450917, "grad_norm": 6.4669036865234375, "learning_rate": 3.3708957098125716e-07, "loss": 1.1691, "step": 34910 }, { "epoch": 0.9198460337895905, "grad_norm": 16.00926399230957, "learning_rate": 3.3489660792867526e-07, "loss": 1.0953, "step": 34920 }, { "epoch": 0.9201094490340892, "grad_norm": 10.941378593444824, "learning_rate": 3.327106799374147e-07, "loss": 1.1562, "step": 34930 }, { "epoch": 0.920372864278588, "grad_norm": 18.4073429107666, "learning_rate": 3.305317885985759e-07, "loss": 1.2504, "step": 34940 }, { "epoch": 0.9206362795230867, "grad_norm": 7.866973400115967, "learning_rate": 3.2835993549813593e-07, "loss": 1.0611, "step": 34950 }, { "epoch": 0.9208996947675855, "grad_norm": 9.33503532409668, "learning_rate": 3.2619512221694905e-07, "loss": 1.1868, "step": 34960 }, { "epoch": 0.9211631100120842, "grad_norm": 8.085517883300781, "learning_rate": 3.24037350330747e-07, "loss": 1.0021, "step": 34970 }, { "epoch": 0.9214265252565829, "grad_norm": 10.626887321472168, "learning_rate": 3.218866214101313e-07, "loss": 1.1625, "step": 34980 }, { "epoch": 0.9216899405010817, "grad_norm": 9.1748046875, "learning_rate": 3.197429370205829e-07, "loss": 1.1755, "step": 34990 }, { "epoch": 0.9219533557455803, "grad_norm": 12.799250602722168, "learning_rate": 3.176062987224504e-07, "loss": 1.0913, "step": 35000 }, { "epoch": 0.9222167709900791, "grad_norm": 8.19770336151123, "learning_rate": 3.1547670807095864e-07, "loss": 1.2066, "step": 35010 }, { "epoch": 0.9224801862345778, "grad_norm": 13.099564552307129, "learning_rate": 3.133541666161977e-07, "loss": 1.1312, "step": 35020 }, { "epoch": 0.9227436014790766, "grad_norm": 8.530671119689941, "learning_rate": 3.112386759031294e-07, "loss": 1.1215, "step": 35030 }, { "epoch": 0.9230070167235753, "grad_norm": 15.649774551391602, "learning_rate": 3.0913023747158324e-07, "loss": 1.0338, "step": 35040 }, { "epoch": 0.9232704319680741, "grad_norm": 12.086459159851074, "learning_rate": 3.070288528562526e-07, "loss": 1.2182, "step": 35050 }, { "epoch": 0.9235338472125728, "grad_norm": 25.359580993652344, "learning_rate": 3.0493452358670183e-07, "loss": 1.137, "step": 35060 }, { "epoch": 0.9237972624570715, "grad_norm": 30.076513290405273, "learning_rate": 3.0284725118735593e-07, "loss": 1.1399, "step": 35070 }, { "epoch": 0.9240606777015703, "grad_norm": 11.856143951416016, "learning_rate": 3.007670371775051e-07, "loss": 1.2501, "step": 35080 }, { "epoch": 0.924324092946069, "grad_norm": 7.266757011413574, "learning_rate": 2.986938830712982e-07, "loss": 1.0819, "step": 35090 }, { "epoch": 0.9245875081905678, "grad_norm": 14.82813835144043, "learning_rate": 2.9662779037775145e-07, "loss": 1.1701, "step": 35100 }, { "epoch": 0.9248509234350665, "grad_norm": 9.683534622192383, "learning_rate": 2.945687606007364e-07, "loss": 1.079, "step": 35110 }, { "epoch": 0.9251143386795653, "grad_norm": 13.479316711425781, "learning_rate": 2.9251679523898627e-07, "loss": 1.2272, "step": 35120 }, { "epoch": 0.925377753924064, "grad_norm": 11.112903594970703, "learning_rate": 2.904718957860897e-07, "loss": 1.0942, "step": 35130 }, { "epoch": 0.9256411691685628, "grad_norm": 8.884119033813477, "learning_rate": 2.8843406373049277e-07, "loss": 1.1549, "step": 35140 }, { "epoch": 0.9259045844130614, "grad_norm": 7.286893367767334, "learning_rate": 2.864033005555011e-07, "loss": 1.0823, "step": 35150 }, { "epoch": 0.9261679996575601, "grad_norm": 15.727523803710938, "learning_rate": 2.843796077392691e-07, "loss": 1.1133, "step": 35160 }, { "epoch": 0.9264314149020589, "grad_norm": 10.252971649169922, "learning_rate": 2.8236298675480946e-07, "loss": 1.1411, "step": 35170 }, { "epoch": 0.9266948301465576, "grad_norm": 7.825354099273682, "learning_rate": 2.803534390699869e-07, "loss": 1.188, "step": 35180 }, { "epoch": 0.9269582453910564, "grad_norm": 10.369479179382324, "learning_rate": 2.7835096614751366e-07, "loss": 1.2233, "step": 35190 }, { "epoch": 0.9272216606355551, "grad_norm": 13.462965965270996, "learning_rate": 2.763555694449549e-07, "loss": 1.0858, "step": 35200 }, { "epoch": 0.9274850758800539, "grad_norm": 14.940919876098633, "learning_rate": 2.743672504147277e-07, "loss": 1.0888, "step": 35210 }, { "epoch": 0.9277484911245526, "grad_norm": 9.79853630065918, "learning_rate": 2.723860105040932e-07, "loss": 1.2033, "step": 35220 }, { "epoch": 0.9280119063690514, "grad_norm": 8.61441421508789, "learning_rate": 2.7041185115516234e-07, "loss": 1.2043, "step": 35230 }, { "epoch": 0.9282753216135501, "grad_norm": 7.251466751098633, "learning_rate": 2.684447738048934e-07, "loss": 1.0968, "step": 35240 }, { "epoch": 0.9285387368580488, "grad_norm": 12.215846061706543, "learning_rate": 2.664847798850834e-07, "loss": 1.0821, "step": 35250 }, { "epoch": 0.9288021521025476, "grad_norm": 9.009248733520508, "learning_rate": 2.6453187082238096e-07, "loss": 1.118, "step": 35260 }, { "epoch": 0.9290655673470463, "grad_norm": 6.85825252532959, "learning_rate": 2.625860480382758e-07, "loss": 1.1162, "step": 35270 }, { "epoch": 0.9293289825915451, "grad_norm": 11.131779670715332, "learning_rate": 2.6064731294909606e-07, "loss": 1.1468, "step": 35280 }, { "epoch": 0.9295923978360437, "grad_norm": 9.182358741760254, "learning_rate": 2.587156669660151e-07, "loss": 1.0536, "step": 35290 }, { "epoch": 0.9298558130805425, "grad_norm": 9.992034912109375, "learning_rate": 2.5679111149504387e-07, "loss": 1.1114, "step": 35300 }, { "epoch": 0.9301192283250412, "grad_norm": 17.44693374633789, "learning_rate": 2.5487364793703173e-07, "loss": 1.0908, "step": 35310 }, { "epoch": 0.9303826435695399, "grad_norm": 7.336899280548096, "learning_rate": 2.5296327768767004e-07, "loss": 1.0439, "step": 35320 }, { "epoch": 0.9306460588140387, "grad_norm": 16.147218704223633, "learning_rate": 2.5106000213748206e-07, "loss": 1.116, "step": 35330 }, { "epoch": 0.9309094740585374, "grad_norm": 6.154407024383545, "learning_rate": 2.4916382267182957e-07, "loss": 1.1687, "step": 35340 }, { "epoch": 0.9311728893030362, "grad_norm": 9.446101188659668, "learning_rate": 2.472747406709097e-07, "loss": 1.073, "step": 35350 }, { "epoch": 0.9314363045475349, "grad_norm": 10.34638500213623, "learning_rate": 2.4539275750975035e-07, "loss": 1.2464, "step": 35360 }, { "epoch": 0.9316997197920337, "grad_norm": 9.2090482711792, "learning_rate": 2.4351787455821675e-07, "loss": 1.1657, "step": 35370 }, { "epoch": 0.9319631350365324, "grad_norm": 13.727375984191895, "learning_rate": 2.4165009318100286e-07, "loss": 1.1746, "step": 35380 }, { "epoch": 0.9322265502810312, "grad_norm": 15.226557731628418, "learning_rate": 2.397894147376345e-07, "loss": 1.2238, "step": 35390 }, { "epoch": 0.9324899655255299, "grad_norm": 6.827721118927002, "learning_rate": 2.379358405824661e-07, "loss": 1.1269, "step": 35400 }, { "epoch": 0.9327533807700286, "grad_norm": 17.900949478149414, "learning_rate": 2.3608937206468395e-07, "loss": 1.1731, "step": 35410 }, { "epoch": 0.9330167960145274, "grad_norm": 11.064220428466797, "learning_rate": 2.3425001052829853e-07, "loss": 1.1744, "step": 35420 }, { "epoch": 0.933280211259026, "grad_norm": 5.221346378326416, "learning_rate": 2.324177573121522e-07, "loss": 1.0594, "step": 35430 }, { "epoch": 0.9335436265035248, "grad_norm": 9.296138763427734, "learning_rate": 2.3059261374990705e-07, "loss": 1.0955, "step": 35440 }, { "epoch": 0.9338070417480235, "grad_norm": 18.757396697998047, "learning_rate": 2.2877458117005592e-07, "loss": 1.1081, "step": 35450 }, { "epoch": 0.9340704569925223, "grad_norm": 13.723665237426758, "learning_rate": 2.269636608959125e-07, "loss": 1.096, "step": 35460 }, { "epoch": 0.934333872237021, "grad_norm": 13.280776977539062, "learning_rate": 2.251598542456135e-07, "loss": 1.2109, "step": 35470 }, { "epoch": 0.9345972874815198, "grad_norm": 10.779294967651367, "learning_rate": 2.2336316253211866e-07, "loss": 1.0302, "step": 35480 }, { "epoch": 0.9348607027260185, "grad_norm": 10.112801551818848, "learning_rate": 2.2157358706320963e-07, "loss": 1.0797, "step": 35490 }, { "epoch": 0.9351241179705172, "grad_norm": 9.152740478515625, "learning_rate": 2.197911291414867e-07, "loss": 1.174, "step": 35500 }, { "epoch": 0.935387533215016, "grad_norm": 9.863764762878418, "learning_rate": 2.180157900643709e-07, "loss": 1.1716, "step": 35510 }, { "epoch": 0.9356509484595147, "grad_norm": 6.8824543952941895, "learning_rate": 2.1624757112409967e-07, "loss": 1.1851, "step": 35520 }, { "epoch": 0.9359143637040135, "grad_norm": 9.15234661102295, "learning_rate": 2.1448647360773013e-07, "loss": 1.0704, "step": 35530 }, { "epoch": 0.9361777789485122, "grad_norm": 11.470961570739746, "learning_rate": 2.1273249879713242e-07, "loss": 1.1245, "step": 35540 }, { "epoch": 0.936441194193011, "grad_norm": 5.655580043792725, "learning_rate": 2.1098564796899867e-07, "loss": 1.075, "step": 35550 }, { "epoch": 0.9367046094375097, "grad_norm": 10.25605583190918, "learning_rate": 2.0924592239482734e-07, "loss": 1.1982, "step": 35560 }, { "epoch": 0.9369680246820085, "grad_norm": 20.1669921875, "learning_rate": 2.0751332334093544e-07, "loss": 1.1473, "step": 35570 }, { "epoch": 0.9372314399265071, "grad_norm": 12.648122787475586, "learning_rate": 2.0578785206845308e-07, "loss": 1.0911, "step": 35580 }, { "epoch": 0.9374948551710058, "grad_norm": 10.0946683883667, "learning_rate": 2.0406950983331897e-07, "loss": 1.1288, "step": 35590 }, { "epoch": 0.9377582704155046, "grad_norm": 7.147735595703125, "learning_rate": 2.0235829788628703e-07, "loss": 1.2308, "step": 35600 }, { "epoch": 0.9380216856600033, "grad_norm": 11.543400764465332, "learning_rate": 2.006542174729176e-07, "loss": 1.1465, "step": 35610 }, { "epoch": 0.9382851009045021, "grad_norm": 9.378731727600098, "learning_rate": 1.9895726983357955e-07, "loss": 1.0524, "step": 35620 }, { "epoch": 0.9385485161490008, "grad_norm": 9.152973175048828, "learning_rate": 1.9726745620345488e-07, "loss": 1.1366, "step": 35630 }, { "epoch": 0.9388119313934996, "grad_norm": 9.166072845458984, "learning_rate": 1.955847778125275e-07, "loss": 1.1543, "step": 35640 }, { "epoch": 0.9390753466379983, "grad_norm": 7.036546230316162, "learning_rate": 1.9390923588558986e-07, "loss": 1.1212, "step": 35650 }, { "epoch": 0.9393387618824971, "grad_norm": 10.675925254821777, "learning_rate": 1.9224083164224304e-07, "loss": 1.1766, "step": 35660 }, { "epoch": 0.9396021771269958, "grad_norm": 10.443366050720215, "learning_rate": 1.9057956629688456e-07, "loss": 1.0874, "step": 35670 }, { "epoch": 0.9398655923714945, "grad_norm": 11.058358192443848, "learning_rate": 1.8892544105872378e-07, "loss": 1.2272, "step": 35680 }, { "epoch": 0.9401290076159933, "grad_norm": 11.761453628540039, "learning_rate": 1.8727845713176985e-07, "loss": 1.1955, "step": 35690 }, { "epoch": 0.940392422860492, "grad_norm": 7.073545932769775, "learning_rate": 1.8563861571483267e-07, "loss": 1.1088, "step": 35700 }, { "epoch": 0.9406558381049908, "grad_norm": 11.718839645385742, "learning_rate": 1.8400591800152522e-07, "loss": 1.2169, "step": 35710 }, { "epoch": 0.9409192533494894, "grad_norm": 10.538962364196777, "learning_rate": 1.8238036518026025e-07, "loss": 1.1526, "step": 35720 }, { "epoch": 0.9411826685939882, "grad_norm": 9.23486042022705, "learning_rate": 1.8076195843424793e-07, "loss": 1.1415, "step": 35730 }, { "epoch": 0.9414460838384869, "grad_norm": 8.318564414978027, "learning_rate": 1.7915069894149928e-07, "loss": 1.1478, "step": 35740 }, { "epoch": 0.9417094990829857, "grad_norm": 7.397567272186279, "learning_rate": 1.775465878748228e-07, "loss": 1.1524, "step": 35750 }, { "epoch": 0.9419729143274844, "grad_norm": 7.76917839050293, "learning_rate": 1.7594962640182235e-07, "loss": 1.172, "step": 35760 }, { "epoch": 0.9422363295719831, "grad_norm": 9.835089683532715, "learning_rate": 1.743598156848991e-07, "loss": 1.1116, "step": 35770 }, { "epoch": 0.9424997448164819, "grad_norm": 11.257411003112793, "learning_rate": 1.7277715688124863e-07, "loss": 1.12, "step": 35780 }, { "epoch": 0.9427631600609806, "grad_norm": 5.076475620269775, "learning_rate": 1.7120165114285936e-07, "loss": 1.0766, "step": 35790 }, { "epoch": 0.9430265753054794, "grad_norm": 6.8095245361328125, "learning_rate": 1.6963329961651732e-07, "loss": 1.138, "step": 35800 }, { "epoch": 0.9432899905499781, "grad_norm": 15.803105354309082, "learning_rate": 1.680721034437971e-07, "loss": 1.1102, "step": 35810 }, { "epoch": 0.9435534057944769, "grad_norm": 11.27157211303711, "learning_rate": 1.6651806376106638e-07, "loss": 1.2321, "step": 35820 }, { "epoch": 0.9438168210389756, "grad_norm": 10.862615585327148, "learning_rate": 1.6497118169948367e-07, "loss": 1.1638, "step": 35830 }, { "epoch": 0.9440802362834743, "grad_norm": 7.417665004730225, "learning_rate": 1.6343145838499609e-07, "loss": 1.1762, "step": 35840 }, { "epoch": 0.9443436515279731, "grad_norm": 9.022090911865234, "learning_rate": 1.6189889493834376e-07, "loss": 1.1525, "step": 35850 }, { "epoch": 0.9446070667724717, "grad_norm": 8.349259376525879, "learning_rate": 1.6037349247505328e-07, "loss": 1.2028, "step": 35860 }, { "epoch": 0.9448704820169705, "grad_norm": 16.508569717407227, "learning_rate": 1.5885525210543762e-07, "loss": 1.2098, "step": 35870 }, { "epoch": 0.9451338972614692, "grad_norm": 11.990571022033691, "learning_rate": 1.573441749345972e-07, "loss": 1.0736, "step": 35880 }, { "epoch": 0.945397312505968, "grad_norm": 12.153646469116211, "learning_rate": 1.5584026206241997e-07, "loss": 1.1627, "step": 35890 }, { "epoch": 0.9456607277504667, "grad_norm": 19.36845588684082, "learning_rate": 1.5434351458357588e-07, "loss": 1.1352, "step": 35900 }, { "epoch": 0.9459241429949655, "grad_norm": 11.167912483215332, "learning_rate": 1.5285393358752566e-07, "loss": 1.1546, "step": 35910 }, { "epoch": 0.9461875582394642, "grad_norm": 5.993119239807129, "learning_rate": 1.5137152015850643e-07, "loss": 1.1473, "step": 35920 }, { "epoch": 0.9464509734839629, "grad_norm": 9.531357765197754, "learning_rate": 1.4989627537554063e-07, "loss": 1.0646, "step": 35930 }, { "epoch": 0.9467143887284617, "grad_norm": 21.36967658996582, "learning_rate": 1.4842820031243599e-07, "loss": 1.1948, "step": 35940 }, { "epoch": 0.9469778039729604, "grad_norm": 15.26762580871582, "learning_rate": 1.469672960377766e-07, "loss": 1.0856, "step": 35950 }, { "epoch": 0.9472412192174592, "grad_norm": 11.183869361877441, "learning_rate": 1.4551356361493074e-07, "loss": 1.1623, "step": 35960 }, { "epoch": 0.9475046344619579, "grad_norm": 9.264765739440918, "learning_rate": 1.4406700410204532e-07, "loss": 1.0568, "step": 35970 }, { "epoch": 0.9477680497064567, "grad_norm": 10.81629467010498, "learning_rate": 1.4262761855204477e-07, "loss": 1.0574, "step": 35980 }, { "epoch": 0.9480314649509554, "grad_norm": 10.056268692016602, "learning_rate": 1.411954080126321e-07, "loss": 1.159, "step": 35990 }, { "epoch": 0.9482948801954542, "grad_norm": 12.725666046142578, "learning_rate": 1.3977037352629118e-07, "loss": 1.2299, "step": 36000 }, { "epoch": 0.9485582954399528, "grad_norm": 10.038346290588379, "learning_rate": 1.38352516130279e-07, "loss": 1.1731, "step": 36010 }, { "epoch": 0.9488217106844515, "grad_norm": 9.279366493225098, "learning_rate": 1.3694183685662776e-07, "loss": 1.1327, "step": 36020 }, { "epoch": 0.9490851259289503, "grad_norm": 8.229722023010254, "learning_rate": 1.3553833673214945e-07, "loss": 1.0968, "step": 36030 }, { "epoch": 0.949348541173449, "grad_norm": 9.047844886779785, "learning_rate": 1.3414201677842353e-07, "loss": 1.164, "step": 36040 }, { "epoch": 0.9496119564179478, "grad_norm": 20.195104598999023, "learning_rate": 1.3275287801181037e-07, "loss": 1.1583, "step": 36050 }, { "epoch": 0.9498753716624465, "grad_norm": 7.857487201690674, "learning_rate": 1.3137092144343887e-07, "loss": 1.011, "step": 36060 }, { "epoch": 0.9501387869069453, "grad_norm": 17.092037200927734, "learning_rate": 1.2999614807921225e-07, "loss": 1.2103, "step": 36070 }, { "epoch": 0.950402202151444, "grad_norm": 9.902064323425293, "learning_rate": 1.286285589198022e-07, "loss": 1.0598, "step": 36080 }, { "epoch": 0.9506656173959428, "grad_norm": 16.6181583404541, "learning_rate": 1.272681549606547e-07, "loss": 1.132, "step": 36090 }, { "epoch": 0.9509290326404415, "grad_norm": 9.175812721252441, "learning_rate": 1.2591493719198323e-07, "loss": 1.1785, "step": 36100 }, { "epoch": 0.9511924478849402, "grad_norm": 8.035114288330078, "learning_rate": 1.245689065987732e-07, "loss": 1.148, "step": 36110 }, { "epoch": 0.951455863129439, "grad_norm": 13.891192436218262, "learning_rate": 1.2323006416077532e-07, "loss": 1.0255, "step": 36120 }, { "epoch": 0.9517192783739377, "grad_norm": 8.833988189697266, "learning_rate": 1.218984108525112e-07, "loss": 1.181, "step": 36130 }, { "epoch": 0.9519826936184365, "grad_norm": 8.799610137939453, "learning_rate": 1.2057394764326657e-07, "loss": 1.238, "step": 36140 }, { "epoch": 0.9522461088629351, "grad_norm": 7.695852756500244, "learning_rate": 1.1925667549709585e-07, "loss": 1.1547, "step": 36150 }, { "epoch": 0.952509524107434, "grad_norm": 9.022981643676758, "learning_rate": 1.1794659537281982e-07, "loss": 1.111, "step": 36160 }, { "epoch": 0.9527729393519326, "grad_norm": 5.551951885223389, "learning_rate": 1.1664370822402127e-07, "loss": 1.0575, "step": 36170 }, { "epoch": 0.9530363545964314, "grad_norm": 9.394152641296387, "learning_rate": 1.1534801499905157e-07, "loss": 1.1397, "step": 36180 }, { "epoch": 0.9532997698409301, "grad_norm": 8.241857528686523, "learning_rate": 1.1405951664101967e-07, "loss": 1.2463, "step": 36190 }, { "epoch": 0.9535631850854288, "grad_norm": 10.305855751037598, "learning_rate": 1.1277821408780532e-07, "loss": 1.0424, "step": 36200 }, { "epoch": 0.9538266003299276, "grad_norm": 7.444577693939209, "learning_rate": 1.1150410827204361e-07, "loss": 1.2198, "step": 36210 }, { "epoch": 0.9540900155744263, "grad_norm": 10.578062057495117, "learning_rate": 1.1023720012113493e-07, "loss": 1.1199, "step": 36220 }, { "epoch": 0.9543534308189251, "grad_norm": 8.359997749328613, "learning_rate": 1.0897749055724049e-07, "loss": 1.2524, "step": 36230 }, { "epoch": 0.9546168460634238, "grad_norm": 7.529768466949463, "learning_rate": 1.0772498049728019e-07, "loss": 1.1977, "step": 36240 }, { "epoch": 0.9548802613079226, "grad_norm": 8.476798057556152, "learning_rate": 1.0647967085293476e-07, "loss": 1.1383, "step": 36250 }, { "epoch": 0.9551436765524213, "grad_norm": 7.85594367980957, "learning_rate": 1.0524156253064355e-07, "loss": 1.1233, "step": 36260 }, { "epoch": 0.9554070917969201, "grad_norm": 12.828485488891602, "learning_rate": 1.0401065643160236e-07, "loss": 1.1588, "step": 36270 }, { "epoch": 0.9556705070414188, "grad_norm": 19.179515838623047, "learning_rate": 1.0278695345177003e-07, "loss": 1.2898, "step": 36280 }, { "epoch": 0.9559339222859174, "grad_norm": 9.386930465698242, "learning_rate": 1.0157045448185631e-07, "loss": 1.1095, "step": 36290 }, { "epoch": 0.9561973375304162, "grad_norm": 6.440732479095459, "learning_rate": 1.0036116040733069e-07, "loss": 1.0846, "step": 36300 }, { "epoch": 0.9564607527749149, "grad_norm": 8.739423751831055, "learning_rate": 9.91590721084168e-08, "loss": 1.0687, "step": 36310 }, { "epoch": 0.9567241680194137, "grad_norm": 9.1159029006958, "learning_rate": 9.796419046009254e-08, "loss": 1.1312, "step": 36320 }, { "epoch": 0.9569875832639124, "grad_norm": 6.196286201477051, "learning_rate": 9.677651633209439e-08, "loss": 1.121, "step": 36330 }, { "epoch": 0.9572509985084112, "grad_norm": 8.255705833435059, "learning_rate": 9.559605058890864e-08, "loss": 1.0968, "step": 36340 }, { "epoch": 0.9575144137529099, "grad_norm": 9.942719459533691, "learning_rate": 9.442279408977462e-08, "loss": 1.1311, "step": 36350 }, { "epoch": 0.9577778289974086, "grad_norm": 8.991029739379883, "learning_rate": 9.325674768868697e-08, "loss": 1.1126, "step": 36360 }, { "epoch": 0.9580412442419074, "grad_norm": 10.643099784851074, "learning_rate": 9.209791223439013e-08, "loss": 1.1126, "step": 36370 }, { "epoch": 0.9583046594864061, "grad_norm": 12.688947677612305, "learning_rate": 9.094628857037935e-08, "loss": 1.1884, "step": 36380 }, { "epoch": 0.9585680747309049, "grad_norm": 9.372642517089844, "learning_rate": 8.98018775349041e-08, "loss": 1.2314, "step": 36390 }, { "epoch": 0.9588314899754036, "grad_norm": 7.384982109069824, "learning_rate": 8.866467996095918e-08, "loss": 1.1349, "step": 36400 }, { "epoch": 0.9590949052199024, "grad_norm": 13.505167961120605, "learning_rate": 8.753469667629133e-08, "loss": 1.1281, "step": 36410 }, { "epoch": 0.9593583204644011, "grad_norm": 9.017455101013184, "learning_rate": 8.641192850339597e-08, "loss": 1.011, "step": 36420 }, { "epoch": 0.9596217357088999, "grad_norm": 11.889460563659668, "learning_rate": 8.529637625951714e-08, "loss": 1.1207, "step": 36430 }, { "epoch": 0.9598851509533985, "grad_norm": 9.88829517364502, "learning_rate": 8.418804075664533e-08, "loss": 1.1362, "step": 36440 }, { "epoch": 0.9601485661978972, "grad_norm": 8.951654434204102, "learning_rate": 8.308692280151853e-08, "loss": 1.2125, "step": 36450 }, { "epoch": 0.960411981442396, "grad_norm": 9.401697158813477, "learning_rate": 8.19930231956223e-08, "loss": 1.1503, "step": 36460 }, { "epoch": 0.9606753966868947, "grad_norm": 12.372333526611328, "learning_rate": 8.090634273518638e-08, "loss": 1.2169, "step": 36470 }, { "epoch": 0.9609388119313935, "grad_norm": 9.986246109008789, "learning_rate": 7.982688221118806e-08, "loss": 1.1022, "step": 36480 }, { "epoch": 0.9612022271758922, "grad_norm": 9.671172142028809, "learning_rate": 7.875464240934661e-08, "loss": 1.1152, "step": 36490 }, { "epoch": 0.961465642420391, "grad_norm": 7.804324626922607, "learning_rate": 7.76896241101266e-08, "loss": 1.0919, "step": 36500 }, { "epoch": 0.9617290576648897, "grad_norm": 7.977505683898926, "learning_rate": 7.663182808873682e-08, "loss": 1.0773, "step": 36510 }, { "epoch": 0.9619924729093885, "grad_norm": 7.611614227294922, "learning_rate": 7.558125511512915e-08, "loss": 1.1061, "step": 36520 }, { "epoch": 0.9622558881538872, "grad_norm": 11.49341869354248, "learning_rate": 7.453790595399746e-08, "loss": 1.0123, "step": 36530 }, { "epoch": 0.9625193033983859, "grad_norm": 11.881976127624512, "learning_rate": 7.350178136477648e-08, "loss": 1.034, "step": 36540 }, { "epoch": 0.9627827186428847, "grad_norm": 7.419308662414551, "learning_rate": 7.247288210164516e-08, "loss": 1.0123, "step": 36550 }, { "epoch": 0.9630461338873834, "grad_norm": 11.622288703918457, "learning_rate": 7.145120891351886e-08, "loss": 1.2354, "step": 36560 }, { "epoch": 0.9633095491318822, "grad_norm": 9.981803894042969, "learning_rate": 7.04367625440583e-08, "loss": 1.1524, "step": 36570 }, { "epoch": 0.9635729643763808, "grad_norm": 10.852524757385254, "learning_rate": 6.942954373165944e-08, "loss": 1.1494, "step": 36580 }, { "epoch": 0.9638363796208796, "grad_norm": 8.66505241394043, "learning_rate": 6.842955320946143e-08, "loss": 1.0491, "step": 36590 }, { "epoch": 0.9640997948653783, "grad_norm": 12.68745231628418, "learning_rate": 6.743679170533868e-08, "loss": 1.047, "step": 36600 }, { "epoch": 0.9643632101098771, "grad_norm": 13.881913185119629, "learning_rate": 6.645125994190538e-08, "loss": 1.2583, "step": 36610 }, { "epoch": 0.9646266253543758, "grad_norm": 9.526371955871582, "learning_rate": 6.547295863651326e-08, "loss": 1.3079, "step": 36620 }, { "epoch": 0.9648900405988745, "grad_norm": 6.9679975509643555, "learning_rate": 6.450188850124938e-08, "loss": 1.1213, "step": 36630 }, { "epoch": 0.9651534558433733, "grad_norm": 13.912981986999512, "learning_rate": 6.353805024294168e-08, "loss": 1.0787, "step": 36640 }, { "epoch": 0.965416871087872, "grad_norm": 22.561283111572266, "learning_rate": 6.258144456314896e-08, "loss": 1.0768, "step": 36650 }, { "epoch": 0.9656802863323708, "grad_norm": 8.544466972351074, "learning_rate": 6.163207215816869e-08, "loss": 1.0532, "step": 36660 }, { "epoch": 0.9659437015768695, "grad_norm": 8.309013366699219, "learning_rate": 6.068993371903254e-08, "loss": 0.9796, "step": 36670 }, { "epoch": 0.9662071168213683, "grad_norm": 13.463048934936523, "learning_rate": 5.97550299315075e-08, "loss": 1.1832, "step": 36680 }, { "epoch": 0.966470532065867, "grad_norm": 10.513821601867676, "learning_rate": 5.882736147609369e-08, "loss": 1.1185, "step": 36690 }, { "epoch": 0.9667339473103658, "grad_norm": 12.073942184448242, "learning_rate": 5.7906929028023196e-08, "loss": 1.121, "step": 36700 }, { "epoch": 0.9669973625548645, "grad_norm": 12.310080528259277, "learning_rate": 5.699373325726454e-08, "loss": 1.0284, "step": 36710 }, { "epoch": 0.9672607777993631, "grad_norm": 12.44055461883545, "learning_rate": 5.6087774828517124e-08, "loss": 1.1958, "step": 36720 }, { "epoch": 0.967524193043862, "grad_norm": 10.837935447692871, "learning_rate": 5.5189054401211253e-08, "loss": 1.0873, "step": 36730 }, { "epoch": 0.9677876082883606, "grad_norm": 12.407529830932617, "learning_rate": 5.429757262951141e-08, "loss": 1.1674, "step": 36740 }, { "epoch": 0.9680510235328594, "grad_norm": 11.307591438293457, "learning_rate": 5.3413330162309654e-08, "loss": 1.0961, "step": 36750 }, { "epoch": 0.9683144387773581, "grad_norm": 11.309042930603027, "learning_rate": 5.2536327643233356e-08, "loss": 1.0781, "step": 36760 }, { "epoch": 0.9685778540218569, "grad_norm": 10.898759841918945, "learning_rate": 5.166656571063633e-08, "loss": 1.1103, "step": 36770 }, { "epoch": 0.9688412692663556, "grad_norm": 10.572040557861328, "learning_rate": 5.080404499760216e-08, "loss": 1.2051, "step": 36780 }, { "epoch": 0.9691046845108544, "grad_norm": 15.641359329223633, "learning_rate": 4.9948766131945324e-08, "loss": 1.1375, "step": 36790 }, { "epoch": 0.9693680997553531, "grad_norm": 10.344244956970215, "learning_rate": 4.9100729736208944e-08, "loss": 1.0547, "step": 36800 }, { "epoch": 0.9696315149998518, "grad_norm": 7.862698078155518, "learning_rate": 4.82599364276648e-08, "loss": 1.11, "step": 36810 }, { "epoch": 0.9698949302443506, "grad_norm": 9.584527969360352, "learning_rate": 4.742638681831113e-08, "loss": 1.0532, "step": 36820 }, { "epoch": 0.9701583454888493, "grad_norm": 11.668069839477539, "learning_rate": 4.6600081514872606e-08, "loss": 1.1725, "step": 36830 }, { "epoch": 0.9704217607333481, "grad_norm": 18.925735473632812, "learning_rate": 4.578102111880367e-08, "loss": 1.2774, "step": 36840 }, { "epoch": 0.9706851759778468, "grad_norm": 8.283143997192383, "learning_rate": 4.49692062262852e-08, "loss": 1.3305, "step": 36850 }, { "epoch": 0.9709485912223456, "grad_norm": 11.803452491760254, "learning_rate": 4.416463742822341e-08, "loss": 1.1376, "step": 36860 }, { "epoch": 0.9712120064668442, "grad_norm": 11.225531578063965, "learning_rate": 4.3367315310247626e-08, "loss": 1.0768, "step": 36870 }, { "epoch": 0.9714754217113429, "grad_norm": 18.636228561401367, "learning_rate": 4.257724045271694e-08, "loss": 0.9991, "step": 36880 }, { "epoch": 0.9717388369558417, "grad_norm": 15.59659481048584, "learning_rate": 4.1794413430711336e-08, "loss": 1.2337, "step": 36890 }, { "epoch": 0.9720022522003404, "grad_norm": 7.88625955581665, "learning_rate": 4.101883481403946e-08, "loss": 1.1413, "step": 36900 }, { "epoch": 0.9722656674448392, "grad_norm": 8.369081497192383, "learning_rate": 4.0250505167229727e-08, "loss": 1.1504, "step": 36910 }, { "epoch": 0.9725290826893379, "grad_norm": 25.276561737060547, "learning_rate": 3.94894250495359e-08, "loss": 1.1788, "step": 36920 }, { "epoch": 0.9727924979338367, "grad_norm": 9.80546760559082, "learning_rate": 3.8735595014935954e-08, "loss": 1.2007, "step": 36930 }, { "epoch": 0.9730559131783354, "grad_norm": 9.60522174835205, "learning_rate": 3.798901561212875e-08, "loss": 1.2547, "step": 36940 }, { "epoch": 0.9733193284228342, "grad_norm": 7.3806586265563965, "learning_rate": 3.724968738453738e-08, "loss": 1.1535, "step": 36950 }, { "epoch": 0.9735827436673329, "grad_norm": 13.008478164672852, "learning_rate": 3.651761087030581e-08, "loss": 1.0804, "step": 36960 }, { "epoch": 0.9738461589118316, "grad_norm": 6.863667964935303, "learning_rate": 3.579278660230001e-08, "loss": 0.885, "step": 36970 }, { "epoch": 0.9741095741563304, "grad_norm": 11.556983947753906, "learning_rate": 3.5075215108106855e-08, "loss": 1.1473, "step": 36980 }, { "epoch": 0.9743729894008291, "grad_norm": 13.247139930725098, "learning_rate": 3.4364896910034084e-08, "loss": 1.1134, "step": 36990 }, { "epoch": 0.9746364046453279, "grad_norm": 9.239654541015625, "learning_rate": 3.366183252510924e-08, "loss": 1.1546, "step": 37000 }, { "epoch": 0.9748998198898265, "grad_norm": 9.277846336364746, "learning_rate": 3.296602246508296e-08, "loss": 1.025, "step": 37010 }, { "epoch": 0.9751632351343253, "grad_norm": 10.646523475646973, "learning_rate": 3.227746723642122e-08, "loss": 1.2091, "step": 37020 }, { "epoch": 0.975426650378824, "grad_norm": 10.912693977355957, "learning_rate": 3.15961673403109e-08, "loss": 0.9881, "step": 37030 }, { "epoch": 0.9756900656233228, "grad_norm": 10.229850769042969, "learning_rate": 3.0922123272660865e-08, "loss": 1.041, "step": 37040 }, { "epoch": 0.9759534808678215, "grad_norm": 16.380971908569336, "learning_rate": 3.025533552409421e-08, "loss": 1.2437, "step": 37050 }, { "epoch": 0.9762168961123202, "grad_norm": 10.567758560180664, "learning_rate": 2.9595804579956034e-08, "loss": 1.2084, "step": 37060 }, { "epoch": 0.976480311356819, "grad_norm": 10.075678825378418, "learning_rate": 2.894353092030566e-08, "loss": 1.1905, "step": 37070 }, { "epoch": 0.9767437266013177, "grad_norm": 10.458196640014648, "learning_rate": 2.82985150199222e-08, "loss": 1.1378, "step": 37080 }, { "epoch": 0.9770071418458165, "grad_norm": 8.47217082977295, "learning_rate": 2.7660757348301203e-08, "loss": 1.0725, "step": 37090 }, { "epoch": 0.9772705570903152, "grad_norm": 13.67198657989502, "learning_rate": 2.7030258369656892e-08, "loss": 1.0842, "step": 37100 }, { "epoch": 0.977533972334814, "grad_norm": 12.729284286499023, "learning_rate": 2.6407018542917717e-08, "loss": 1.1352, "step": 37110 }, { "epoch": 0.9777973875793127, "grad_norm": 6.472448825836182, "learning_rate": 2.579103832172858e-08, "loss": 1.2346, "step": 37120 }, { "epoch": 0.9780608028238115, "grad_norm": 10.462120056152344, "learning_rate": 2.5182318154451934e-08, "loss": 1.1004, "step": 37130 }, { "epoch": 0.9783242180683102, "grad_norm": 9.162874221801758, "learning_rate": 2.4580858484163362e-08, "loss": 1.0859, "step": 37140 }, { "epoch": 0.9785876333128088, "grad_norm": 27.292343139648438, "learning_rate": 2.3986659748654883e-08, "loss": 1.2228, "step": 37150 }, { "epoch": 0.9788510485573076, "grad_norm": 5.657218933105469, "learning_rate": 2.3399722380434974e-08, "loss": 1.1921, "step": 37160 }, { "epoch": 0.9791144638018063, "grad_norm": 9.713508605957031, "learning_rate": 2.2820046806725227e-08, "loss": 1.1352, "step": 37170 }, { "epoch": 0.9793778790463051, "grad_norm": 25.21521759033203, "learning_rate": 2.2247633449459237e-08, "loss": 1.1455, "step": 37180 }, { "epoch": 0.9796412942908038, "grad_norm": 9.405596733093262, "learning_rate": 2.168248272528928e-08, "loss": 1.1601, "step": 37190 }, { "epoch": 0.9799047095353026, "grad_norm": 10.610394477844238, "learning_rate": 2.11245950455774e-08, "loss": 1.1553, "step": 37200 }, { "epoch": 0.9801681247798013, "grad_norm": 8.835819244384766, "learning_rate": 2.057397081639989e-08, "loss": 1.145, "step": 37210 }, { "epoch": 0.9804315400243001, "grad_norm": 29.819286346435547, "learning_rate": 2.003061043854726e-08, "loss": 1.0828, "step": 37220 }, { "epoch": 0.9806949552687988, "grad_norm": 8.96506404876709, "learning_rate": 1.9494514307523137e-08, "loss": 1.0038, "step": 37230 }, { "epoch": 0.9809583705132975, "grad_norm": 9.260284423828125, "learning_rate": 1.896568281354094e-08, "loss": 1.0476, "step": 37240 }, { "epoch": 0.9812217857577963, "grad_norm": 9.093425750732422, "learning_rate": 1.844411634152943e-08, "loss": 1.0549, "step": 37250 }, { "epoch": 0.981485201002295, "grad_norm": 16.045421600341797, "learning_rate": 1.7929815271126028e-08, "loss": 1.1138, "step": 37260 }, { "epoch": 0.9817486162467938, "grad_norm": 7.838864326477051, "learning_rate": 1.7422779976682403e-08, "loss": 1.2042, "step": 37270 }, { "epoch": 0.9820120314912925, "grad_norm": 11.226157188415527, "learning_rate": 1.692301082726222e-08, "loss": 1.1571, "step": 37280 }, { "epoch": 0.9822754467357913, "grad_norm": 12.788179397583008, "learning_rate": 1.6430508186636717e-08, "loss": 1.1458, "step": 37290 }, { "epoch": 0.98253886198029, "grad_norm": 11.170735359191895, "learning_rate": 1.594527241329136e-08, "loss": 1.1564, "step": 37300 }, { "epoch": 0.9828022772247887, "grad_norm": 14.343019485473633, "learning_rate": 1.546730386041917e-08, "loss": 1.0764, "step": 37310 }, { "epoch": 0.9830656924692874, "grad_norm": 9.570619583129883, "learning_rate": 1.4996602875927412e-08, "loss": 1.0768, "step": 37320 }, { "epoch": 0.9833291077137861, "grad_norm": 11.917278289794922, "learning_rate": 1.4533169802429803e-08, "loss": 1.0995, "step": 37330 }, { "epoch": 0.9835925229582849, "grad_norm": 12.944307327270508, "learning_rate": 1.4077004977250952e-08, "loss": 1.0946, "step": 37340 }, { "epoch": 0.9838559382027836, "grad_norm": 8.70278263092041, "learning_rate": 1.3628108732426371e-08, "loss": 1.1488, "step": 37350 }, { "epoch": 0.9841193534472824, "grad_norm": 11.318720817565918, "learning_rate": 1.3186481394699135e-08, "loss": 1.1978, "step": 37360 }, { "epoch": 0.9843827686917811, "grad_norm": 8.020891189575195, "learning_rate": 1.2752123285522111e-08, "loss": 1.0356, "step": 37370 }, { "epoch": 0.9846461839362799, "grad_norm": 6.4645538330078125, "learning_rate": 1.232503472105795e-08, "loss": 1.1697, "step": 37380 }, { "epoch": 0.9849095991807786, "grad_norm": 15.600767135620117, "learning_rate": 1.1905216012176869e-08, "loss": 1.2071, "step": 37390 }, { "epoch": 0.9851730144252773, "grad_norm": 14.369098663330078, "learning_rate": 1.1492667464456652e-08, "loss": 1.1402, "step": 37400 }, { "epoch": 0.9854364296697761, "grad_norm": 12.011136054992676, "learning_rate": 1.1087389378184877e-08, "loss": 1.0541, "step": 37410 }, { "epoch": 0.9856998449142748, "grad_norm": 11.413780212402344, "learning_rate": 1.0689382048355567e-08, "loss": 1.1509, "step": 37420 }, { "epoch": 0.9859632601587736, "grad_norm": 13.161310195922852, "learning_rate": 1.0298645764672543e-08, "loss": 1.0617, "step": 37430 }, { "epoch": 0.9862266754032722, "grad_norm": 14.529175758361816, "learning_rate": 9.915180811546076e-09, "loss": 1.1029, "step": 37440 }, { "epoch": 0.986490090647771, "grad_norm": 8.117080688476562, "learning_rate": 9.538987468092897e-09, "loss": 1.0878, "step": 37450 }, { "epoch": 0.9867535058922697, "grad_norm": 10.85071086883545, "learning_rate": 9.170066008137302e-09, "loss": 1.1666, "step": 37460 }, { "epoch": 0.9870169211367685, "grad_norm": 7.760594367980957, "learning_rate": 8.808416700211153e-09, "loss": 1.2474, "step": 37470 }, { "epoch": 0.9872803363812672, "grad_norm": 10.442521095275879, "learning_rate": 8.454039807553881e-09, "loss": 1.0193, "step": 37480 }, { "epoch": 0.9875437516257659, "grad_norm": 13.61961841583252, "learning_rate": 8.106935588109155e-09, "loss": 1.0304, "step": 37490 }, { "epoch": 0.9878071668702647, "grad_norm": 9.076234817504883, "learning_rate": 7.767104294528205e-09, "loss": 1.1811, "step": 37500 }, { "epoch": 0.9880705821147634, "grad_norm": 11.785021781921387, "learning_rate": 7.434546174168722e-09, "loss": 1.1379, "step": 37510 }, { "epoch": 0.9883339973592622, "grad_norm": 9.68836498260498, "learning_rate": 7.109261469093742e-09, "loss": 1.2178, "step": 37520 }, { "epoch": 0.9885974126037609, "grad_norm": 12.241327285766602, "learning_rate": 6.7912504160716484e-09, "loss": 1.2114, "step": 37530 }, { "epoch": 0.9888608278482597, "grad_norm": 7.527163505554199, "learning_rate": 6.48051324657839e-09, "loss": 1.1169, "step": 37540 }, { "epoch": 0.9891242430927584, "grad_norm": 8.736496925354004, "learning_rate": 6.177050186793043e-09, "loss": 1.0317, "step": 37550 }, { "epoch": 0.9893876583372572, "grad_norm": 13.77798080444336, "learning_rate": 5.880861457602249e-09, "loss": 1.1827, "step": 37560 }, { "epoch": 0.9896510735817559, "grad_norm": 11.816718101501465, "learning_rate": 5.5919472745968875e-09, "loss": 1.0985, "step": 37570 }, { "epoch": 0.9899144888262545, "grad_norm": 15.230541229248047, "learning_rate": 5.310307848070962e-09, "loss": 1.0616, "step": 37580 }, { "epoch": 0.9901779040707533, "grad_norm": 8.248568534851074, "learning_rate": 5.035943383026043e-09, "loss": 1.0327, "step": 37590 }, { "epoch": 0.990441319315252, "grad_norm": 8.132597923278809, "learning_rate": 4.7688540791668295e-09, "loss": 1.1544, "step": 37600 }, { "epoch": 0.9907047345597508, "grad_norm": 14.392091751098633, "learning_rate": 4.509040130903364e-09, "loss": 1.114, "step": 37610 }, { "epoch": 0.9909681498042495, "grad_norm": 8.423932075500488, "learning_rate": 4.256501727349927e-09, "loss": 1.1063, "step": 37620 }, { "epoch": 0.9912315650487483, "grad_norm": 8.729024887084961, "learning_rate": 4.011239052323923e-09, "loss": 1.0748, "step": 37630 }, { "epoch": 0.991494980293247, "grad_norm": 17.652915954589844, "learning_rate": 3.773252284348105e-09, "loss": 1.1656, "step": 37640 }, { "epoch": 0.9917583955377458, "grad_norm": 20.125713348388672, "learning_rate": 3.5425415966505726e-09, "loss": 1.1486, "step": 37650 }, { "epoch": 0.9920218107822445, "grad_norm": 12.343775749206543, "learning_rate": 3.319107157159218e-09, "loss": 1.0857, "step": 37660 }, { "epoch": 0.9922852260267432, "grad_norm": 10.12013053894043, "learning_rate": 3.1029491285083923e-09, "loss": 1.1936, "step": 37670 }, { "epoch": 0.992548641271242, "grad_norm": 9.91645336151123, "learning_rate": 2.8940676680377922e-09, "loss": 1.1639, "step": 37680 }, { "epoch": 0.9928120565157407, "grad_norm": 9.207098007202148, "learning_rate": 2.6924629277869098e-09, "loss": 1.011, "step": 37690 }, { "epoch": 0.9930754717602395, "grad_norm": 13.363540649414062, "learning_rate": 2.4981350545005833e-09, "loss": 1.2361, "step": 37700 }, { "epoch": 0.9933388870047382, "grad_norm": 12.32223892211914, "learning_rate": 2.311084189626778e-09, "loss": 1.1561, "step": 37710 }, { "epoch": 0.993602302249237, "grad_norm": 14.073246955871582, "learning_rate": 2.1313104693176934e-09, "loss": 1.247, "step": 37720 }, { "epoch": 0.9938657174937356, "grad_norm": 11.04294204711914, "learning_rate": 1.958814024425326e-09, "loss": 0.9954, "step": 37730 }, { "epoch": 0.9941291327382344, "grad_norm": 8.011744499206543, "learning_rate": 1.793594980509239e-09, "loss": 1.1036, "step": 37740 }, { "epoch": 0.9943925479827331, "grad_norm": 274.66278076171875, "learning_rate": 1.6356534578276795e-09, "loss": 1.1784, "step": 37750 }, { "epoch": 0.9946559632272318, "grad_norm": 14.736515045166016, "learning_rate": 1.4849895713431316e-09, "loss": 1.1025, "step": 37760 }, { "epoch": 0.9949193784717306, "grad_norm": 10.031840324401855, "learning_rate": 1.341603430723426e-09, "loss": 1.2254, "step": 37770 }, { "epoch": 0.9951827937162293, "grad_norm": 18.174930572509766, "learning_rate": 1.205495140336188e-09, "loss": 1.1696, "step": 37780 }, { "epoch": 0.9954462089607281, "grad_norm": 14.761902809143066, "learning_rate": 1.076664799249949e-09, "loss": 1.1105, "step": 37790 }, { "epoch": 0.9957096242052268, "grad_norm": 12.710417747497559, "learning_rate": 9.551125012419172e-10, "loss": 1.124, "step": 37800 }, { "epoch": 0.9959730394497256, "grad_norm": 10.125197410583496, "learning_rate": 8.408383347846549e-10, "loss": 1.2128, "step": 37810 }, { "epoch": 0.9962364546942243, "grad_norm": 10.500262260437012, "learning_rate": 7.338423830582919e-10, "loss": 1.1741, "step": 37820 }, { "epoch": 0.9964998699387231, "grad_norm": 14.972089767456055, "learning_rate": 6.341247239416426e-10, "loss": 1.1737, "step": 37830 }, { "epoch": 0.9967632851832218, "grad_norm": 10.579835891723633, "learning_rate": 5.416854300199781e-10, "loss": 1.2571, "step": 37840 }, { "epoch": 0.9970267004277205, "grad_norm": 11.301043510437012, "learning_rate": 4.5652456857503414e-10, "loss": 1.0952, "step": 37850 }, { "epoch": 0.9972901156722193, "grad_norm": 8.719541549682617, "learning_rate": 3.7864220159722354e-10, "loss": 1.0867, "step": 37860 }, { "epoch": 0.997553530916718, "grad_norm": 11.111146926879883, "learning_rate": 3.080383857734237e-10, "loss": 1.1613, "step": 37870 }, { "epoch": 0.9978169461612167, "grad_norm": 8.674304962158203, "learning_rate": 2.4471317249585846e-10, "loss": 1.2788, "step": 37880 }, { "epoch": 0.9980803614057154, "grad_norm": 9.656017303466797, "learning_rate": 1.8866660785654689e-10, "loss": 1.1277, "step": 37890 }, { "epoch": 0.9983437766502142, "grad_norm": 7.973507404327393, "learning_rate": 1.3989873265285446e-10, "loss": 1.1206, "step": 37900 }, { "epoch": 0.9986071918947129, "grad_norm": 8.449443817138672, "learning_rate": 9.840958238194198e-11, "loss": 1.0563, "step": 37910 }, { "epoch": 0.9988706071392117, "grad_norm": 8.268136024475098, "learning_rate": 6.419918724187569e-11, "loss": 1.0823, "step": 37920 }, { "epoch": 0.9991340223837104, "grad_norm": 8.095474243164062, "learning_rate": 3.726757213384779e-11, "loss": 1.1064, "step": 37930 }, { "epoch": 0.9993974376282091, "grad_norm": 18.85165786743164, "learning_rate": 1.7614756661066267e-11, "loss": 1.101, "step": 37940 }, { "epoch": 0.9996608528727079, "grad_norm": 9.734877586364746, "learning_rate": 5.240755128754771e-12, "loss": 1.188, "step": 37950 }, { "epoch": 0.9999242681172066, "grad_norm": 9.190556526184082, "learning_rate": 1.455765441527035e-13, "loss": 1.1125, "step": 37960 }, { "epoch": 0.9999769511661064, "step": 37962, "total_flos": 2.3899935777450728e+20, "train_loss": 1.3406300857599565, "train_runtime": 316428.2007, "train_samples_per_second": 7.678, "train_steps_per_second": 0.12 } ], "logging_steps": 10, "max_steps": 37962, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.3899935777450728e+20, "train_batch_size": 1, "trial_name": null, "trial_params": null }