{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 15165, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0019784350578692256, "grad_norm": 19.637782613049417, "learning_rate": 5.932762030323006e-07, "loss": 8.2273, "step": 10 }, { "epoch": 0.003956870115738451, "grad_norm": 19.83226249327557, "learning_rate": 1.2524719841793013e-06, "loss": 8.167, "step": 20 }, { "epoch": 0.005935305173607676, "grad_norm": 17.787035299527364, "learning_rate": 1.9116677653263018e-06, "loss": 7.8304, "step": 30 }, { "epoch": 0.007913740231476903, "grad_norm": 6.953509182433938, "learning_rate": 2.570863546473303e-06, "loss": 7.0706, "step": 40 }, { "epoch": 0.009892175289346126, "grad_norm": 4.3091682221623735, "learning_rate": 3.2300593276203036e-06, "loss": 6.5597, "step": 50 }, { "epoch": 0.011870610347215352, "grad_norm": 3.0516932716653735, "learning_rate": 3.889255108767304e-06, "loss": 6.2279, "step": 60 }, { "epoch": 0.013849045405084578, "grad_norm": 2.7480308952739625, "learning_rate": 4.548450889914305e-06, "loss": 5.9488, "step": 70 }, { "epoch": 0.015827480462953805, "grad_norm": 3.533312550024619, "learning_rate": 5.207646671061305e-06, "loss": 5.6328, "step": 80 }, { "epoch": 0.01780591552082303, "grad_norm": 3.654444836350219, "learning_rate": 5.866842452208306e-06, "loss": 5.1732, "step": 90 }, { "epoch": 0.019784350578692253, "grad_norm": 4.086440552060253, "learning_rate": 6.5260382333553066e-06, "loss": 4.6653, "step": 100 }, { "epoch": 0.02176278563656148, "grad_norm": 4.138046221068996, "learning_rate": 7.1852340145023085e-06, "loss": 4.2061, "step": 110 }, { "epoch": 0.023741220694430704, "grad_norm": 5.7178401280922735, "learning_rate": 7.844429795649308e-06, "loss": 3.8123, "step": 120 }, { "epoch": 0.02571965575229993, "grad_norm": 5.106742130319203, "learning_rate": 8.503625576796309e-06, "loss": 3.512, "step": 130 }, { "epoch": 0.027698090810169156, "grad_norm": 4.573421695623524, "learning_rate": 9.162821357943308e-06, "loss": 3.2543, "step": 140 }, { "epoch": 0.029676525868038383, "grad_norm": 5.128687988140343, "learning_rate": 9.82201713909031e-06, "loss": 3.0517, "step": 150 }, { "epoch": 0.03165496092590761, "grad_norm": 4.880606343802238, "learning_rate": 1.048121292023731e-05, "loss": 2.8845, "step": 160 }, { "epoch": 0.033633395983776834, "grad_norm": 8.127044152874786, "learning_rate": 1.1140408701384312e-05, "loss": 2.748, "step": 170 }, { "epoch": 0.03561183104164606, "grad_norm": 8.485194090796208, "learning_rate": 1.1799604482531313e-05, "loss": 2.6414, "step": 180 }, { "epoch": 0.03759026609951528, "grad_norm": 6.612294272350823, "learning_rate": 1.2458800263678314e-05, "loss": 2.5598, "step": 190 }, { "epoch": 0.039568701157384506, "grad_norm": 12.16444744207085, "learning_rate": 1.3117996044825315e-05, "loss": 2.4886, "step": 200 }, { "epoch": 0.04154713621525374, "grad_norm": 18.63069790020457, "learning_rate": 1.3777191825972313e-05, "loss": 2.4395, "step": 210 }, { "epoch": 0.04352557127312296, "grad_norm": 11.412955192769877, "learning_rate": 1.4436387607119314e-05, "loss": 2.383, "step": 220 }, { "epoch": 0.045504006330992185, "grad_norm": 11.202485180848445, "learning_rate": 1.5095583388266315e-05, "loss": 2.3341, "step": 230 }, { "epoch": 0.04748244138886141, "grad_norm": 31.060419144311027, "learning_rate": 1.5754779169413318e-05, "loss": 2.3041, "step": 240 }, { "epoch": 0.04946087644673064, "grad_norm": 22.03166390729224, "learning_rate": 1.6413974950560317e-05, "loss": 2.2843, "step": 250 }, { "epoch": 0.05143931150459986, "grad_norm": 32.84287275099665, "learning_rate": 1.707317073170732e-05, "loss": 2.2874, "step": 260 }, { "epoch": 0.05341774656246909, "grad_norm": 13.792898956806326, "learning_rate": 1.7732366512854316e-05, "loss": 2.2134, "step": 270 }, { "epoch": 0.05539618162033831, "grad_norm": 9.703481654155626, "learning_rate": 1.839156229400132e-05, "loss": 2.1511, "step": 280 }, { "epoch": 0.057374616678207535, "grad_norm": 18.160904957620193, "learning_rate": 1.905075807514832e-05, "loss": 2.1061, "step": 290 }, { "epoch": 0.059353051736076766, "grad_norm": 31.18164400199025, "learning_rate": 1.970995385629532e-05, "loss": 2.0923, "step": 300 }, { "epoch": 0.06133148679394599, "grad_norm": 18.352406637861762, "learning_rate": 2.0369149637442324e-05, "loss": 2.0124, "step": 310 }, { "epoch": 0.06330992185181522, "grad_norm": 13.168827033502875, "learning_rate": 2.102834541858932e-05, "loss": 1.9455, "step": 320 }, { "epoch": 0.06528835690968444, "grad_norm": 12.98291760357499, "learning_rate": 2.1687541199736323e-05, "loss": 1.8702, "step": 330 }, { "epoch": 0.06726679196755367, "grad_norm": 11.475153571817428, "learning_rate": 2.2346736980883322e-05, "loss": 1.7829, "step": 340 }, { "epoch": 0.06924522702542289, "grad_norm": 14.551237589997115, "learning_rate": 2.3005932762030325e-05, "loss": 1.6901, "step": 350 }, { "epoch": 0.07122366208329212, "grad_norm": 9.667096995303945, "learning_rate": 2.3665128543177324e-05, "loss": 1.5745, "step": 360 }, { "epoch": 0.07320209714116134, "grad_norm": 13.163411291543234, "learning_rate": 2.4324324324324327e-05, "loss": 1.4658, "step": 370 }, { "epoch": 0.07518053219903056, "grad_norm": 9.940351335922, "learning_rate": 2.4983520105471327e-05, "loss": 1.3668, "step": 380 }, { "epoch": 0.07715896725689979, "grad_norm": 13.992341034680026, "learning_rate": 2.564271588661833e-05, "loss": 1.2807, "step": 390 }, { "epoch": 0.07913740231476901, "grad_norm": 7.4831699972814105, "learning_rate": 2.630191166776533e-05, "loss": 1.211, "step": 400 }, { "epoch": 0.08111583737263825, "grad_norm": 9.56614415520462, "learning_rate": 2.6961107448912325e-05, "loss": 1.1591, "step": 410 }, { "epoch": 0.08309427243050747, "grad_norm": 9.870662837309114, "learning_rate": 2.762030323005933e-05, "loss": 1.1007, "step": 420 }, { "epoch": 0.0850727074883767, "grad_norm": 7.296161468394743, "learning_rate": 2.8279499011206327e-05, "loss": 1.0605, "step": 430 }, { "epoch": 0.08705114254624592, "grad_norm": 9.934379350610987, "learning_rate": 2.8938694792353333e-05, "loss": 1.0221, "step": 440 }, { "epoch": 0.08902957760411515, "grad_norm": 8.322773668200096, "learning_rate": 2.959789057350033e-05, "loss": 0.9699, "step": 450 }, { "epoch": 0.09100801266198437, "grad_norm": 11.476438775344631, "learning_rate": 3.025708635464733e-05, "loss": 0.9331, "step": 460 }, { "epoch": 0.0929864477198536, "grad_norm": 7.972860756568456, "learning_rate": 3.0916282135794335e-05, "loss": 0.9203, "step": 470 }, { "epoch": 0.09496488277772282, "grad_norm": 11.520251589320083, "learning_rate": 3.1575477916941334e-05, "loss": 0.8832, "step": 480 }, { "epoch": 0.09694331783559204, "grad_norm": 9.465237203788785, "learning_rate": 3.2234673698088334e-05, "loss": 0.8484, "step": 490 }, { "epoch": 0.09892175289346128, "grad_norm": 13.258709856888437, "learning_rate": 3.289386947923533e-05, "loss": 0.8216, "step": 500 }, { "epoch": 0.1009001879513305, "grad_norm": 9.067024221441216, "learning_rate": 3.355306526038233e-05, "loss": 0.806, "step": 510 }, { "epoch": 0.10287862300919973, "grad_norm": 9.967066995989914, "learning_rate": 3.421226104152934e-05, "loss": 0.7768, "step": 520 }, { "epoch": 0.10485705806706895, "grad_norm": 6.990448616699275, "learning_rate": 3.487145682267634e-05, "loss": 0.7528, "step": 530 }, { "epoch": 0.10683549312493817, "grad_norm": 8.863709916752113, "learning_rate": 3.553065260382334e-05, "loss": 0.7225, "step": 540 }, { "epoch": 0.1088139281828074, "grad_norm": 8.219483179811172, "learning_rate": 3.618984838497034e-05, "loss": 0.6902, "step": 550 }, { "epoch": 0.11079236324067662, "grad_norm": 8.18954618815931, "learning_rate": 3.684904416611734e-05, "loss": 0.6664, "step": 560 }, { "epoch": 0.11277079829854585, "grad_norm": 6.99844977800662, "learning_rate": 3.750823994726434e-05, "loss": 0.6444, "step": 570 }, { "epoch": 0.11474923335641507, "grad_norm": 7.503046833377609, "learning_rate": 3.8167435728411335e-05, "loss": 0.6149, "step": 580 }, { "epoch": 0.11672766841428431, "grad_norm": 6.722656169187286, "learning_rate": 3.882663150955834e-05, "loss": 0.6168, "step": 590 }, { "epoch": 0.11870610347215353, "grad_norm": 7.030946593625458, "learning_rate": 3.948582729070534e-05, "loss": 0.5881, "step": 600 }, { "epoch": 0.12068453853002276, "grad_norm": 9.774246488301465, "learning_rate": 4.014502307185235e-05, "loss": 0.5607, "step": 610 }, { "epoch": 0.12266297358789198, "grad_norm": 7.078903913735779, "learning_rate": 4.0804218852999346e-05, "loss": 0.5405, "step": 620 }, { "epoch": 0.1246414086457612, "grad_norm": 7.402770055900705, "learning_rate": 4.146341463414634e-05, "loss": 0.5212, "step": 630 }, { "epoch": 0.12661984370363044, "grad_norm": 7.182019494715211, "learning_rate": 4.2122610415293345e-05, "loss": 0.4983, "step": 640 }, { "epoch": 0.12859827876149965, "grad_norm": 5.276303108063209, "learning_rate": 4.2781806196440345e-05, "loss": 0.4771, "step": 650 }, { "epoch": 0.1305767138193689, "grad_norm": 4.7737357936486235, "learning_rate": 4.344100197758735e-05, "loss": 0.4604, "step": 660 }, { "epoch": 0.1325551488772381, "grad_norm": 5.4871112446079655, "learning_rate": 4.4100197758734343e-05, "loss": 0.4532, "step": 670 }, { "epoch": 0.13453358393510734, "grad_norm": 6.595412809013484, "learning_rate": 4.475939353988134e-05, "loss": 0.4311, "step": 680 }, { "epoch": 0.13651201899297655, "grad_norm": 4.841829206426501, "learning_rate": 4.541858932102835e-05, "loss": 0.4181, "step": 690 }, { "epoch": 0.13849045405084578, "grad_norm": 4.7217934333031435, "learning_rate": 4.607778510217535e-05, "loss": 0.3928, "step": 700 }, { "epoch": 0.140468889108715, "grad_norm": 6.026086612922775, "learning_rate": 4.673698088332235e-05, "loss": 0.3897, "step": 710 }, { "epoch": 0.14244732416658423, "grad_norm": 4.396456746375336, "learning_rate": 4.739617666446935e-05, "loss": 0.3701, "step": 720 }, { "epoch": 0.14442575922445347, "grad_norm": 6.14790077100577, "learning_rate": 4.805537244561635e-05, "loss": 0.3488, "step": 730 }, { "epoch": 0.14640419428232268, "grad_norm": 5.390368007008413, "learning_rate": 4.871456822676335e-05, "loss": 0.3357, "step": 740 }, { "epoch": 0.14838262934019192, "grad_norm": 4.441570962528891, "learning_rate": 4.937376400791035e-05, "loss": 0.3203, "step": 750 }, { "epoch": 0.15036106439806113, "grad_norm": 5.270754499685052, "learning_rate": 5.003295978905735e-05, "loss": 0.3071, "step": 760 }, { "epoch": 0.15233949945593037, "grad_norm": 4.071551636960703, "learning_rate": 5.069215557020436e-05, "loss": 0.2991, "step": 770 }, { "epoch": 0.15431793451379958, "grad_norm": 3.0561647726741783, "learning_rate": 5.135135135135135e-05, "loss": 0.2832, "step": 780 }, { "epoch": 0.15629636957166881, "grad_norm": 2.9516027116400427, "learning_rate": 5.201054713249836e-05, "loss": 0.2681, "step": 790 }, { "epoch": 0.15827480462953802, "grad_norm": 4.621200447731901, "learning_rate": 5.2669742913645356e-05, "loss": 0.26, "step": 800 }, { "epoch": 0.16025323968740726, "grad_norm": 3.041429947319794, "learning_rate": 5.332893869479235e-05, "loss": 0.2471, "step": 810 }, { "epoch": 0.1622316747452765, "grad_norm": 2.995343942067045, "learning_rate": 5.3988134475939355e-05, "loss": 0.2369, "step": 820 }, { "epoch": 0.1642101098031457, "grad_norm": 3.1918861691745795, "learning_rate": 5.464733025708636e-05, "loss": 0.2279, "step": 830 }, { "epoch": 0.16618854486101495, "grad_norm": 2.1160708161237194, "learning_rate": 5.5306526038233354e-05, "loss": 0.2168, "step": 840 }, { "epoch": 0.16816697991888416, "grad_norm": 2.849132153355752, "learning_rate": 5.596572181938036e-05, "loss": 0.2107, "step": 850 }, { "epoch": 0.1701454149767534, "grad_norm": 2.5306963483355673, "learning_rate": 5.662491760052736e-05, "loss": 0.1974, "step": 860 }, { "epoch": 0.1721238500346226, "grad_norm": 3.134461921573245, "learning_rate": 5.7284113381674366e-05, "loss": 0.1898, "step": 870 }, { "epoch": 0.17410228509249184, "grad_norm": 4.467213018542406, "learning_rate": 5.794330916282136e-05, "loss": 0.1918, "step": 880 }, { "epoch": 0.17608072015036105, "grad_norm": 1.9384617338789845, "learning_rate": 5.8602504943968364e-05, "loss": 0.1846, "step": 890 }, { "epoch": 0.1780591552082303, "grad_norm": 2.22914110082071, "learning_rate": 5.9261700725115364e-05, "loss": 0.176, "step": 900 }, { "epoch": 0.18003759026609953, "grad_norm": 2.6195670505855104, "learning_rate": 5.9920896506262356e-05, "loss": 0.1645, "step": 910 }, { "epoch": 0.18201602532396874, "grad_norm": 5.623481408126527, "learning_rate": 6.058009228740936e-05, "loss": 0.1614, "step": 920 }, { "epoch": 0.18399446038183798, "grad_norm": 2.541850203721254, "learning_rate": 6.123928806855637e-05, "loss": 0.1562, "step": 930 }, { "epoch": 0.1859728954397072, "grad_norm": 2.530270886035028, "learning_rate": 6.189848384970337e-05, "loss": 0.1498, "step": 940 }, { "epoch": 0.18795133049757642, "grad_norm": 1.8747639348929863, "learning_rate": 6.255767963085037e-05, "loss": 0.1418, "step": 950 }, { "epoch": 0.18992976555544563, "grad_norm": 1.6302710016776405, "learning_rate": 6.321687541199737e-05, "loss": 0.1353, "step": 960 }, { "epoch": 0.19190820061331487, "grad_norm": 1.374949112823051, "learning_rate": 6.387607119314437e-05, "loss": 0.1303, "step": 970 }, { "epoch": 0.19388663567118408, "grad_norm": 2.300554535314371, "learning_rate": 6.453526697429137e-05, "loss": 0.1262, "step": 980 }, { "epoch": 0.19586507072905332, "grad_norm": 1.6596926456190018, "learning_rate": 6.519446275543837e-05, "loss": 0.1338, "step": 990 }, { "epoch": 0.19784350578692256, "grad_norm": 1.6448857589126298, "learning_rate": 6.585365853658538e-05, "loss": 0.1282, "step": 1000 }, { "epoch": 0.19982194084479177, "grad_norm": 1.4480849991974878, "learning_rate": 6.651285431773236e-05, "loss": 0.116, "step": 1010 }, { "epoch": 0.201800375902661, "grad_norm": 1.8869991728949147, "learning_rate": 6.717205009887936e-05, "loss": 0.1104, "step": 1020 }, { "epoch": 0.20377881096053022, "grad_norm": 1.6293318832625847, "learning_rate": 6.783124588002638e-05, "loss": 0.1061, "step": 1030 }, { "epoch": 0.20575724601839945, "grad_norm": 1.2427464212813581, "learning_rate": 6.849044166117338e-05, "loss": 0.1022, "step": 1040 }, { "epoch": 0.20773568107626866, "grad_norm": 1.2404060214766448, "learning_rate": 6.914963744232038e-05, "loss": 0.0974, "step": 1050 }, { "epoch": 0.2097141161341379, "grad_norm": 6.9065034819409625, "learning_rate": 6.980883322346737e-05, "loss": 0.0977, "step": 1060 }, { "epoch": 0.2116925511920071, "grad_norm": 2.190513083820012, "learning_rate": 7.046802900461437e-05, "loss": 0.1328, "step": 1070 }, { "epoch": 0.21367098624987635, "grad_norm": 1.263188586172193, "learning_rate": 7.112722478576137e-05, "loss": 0.0917, "step": 1080 }, { "epoch": 0.2156494213077456, "grad_norm": 1.9784069526825092, "learning_rate": 7.178642056690837e-05, "loss": 0.0871, "step": 1090 }, { "epoch": 0.2176278563656148, "grad_norm": 0.8523617558822155, "learning_rate": 7.244561634805537e-05, "loss": 0.085, "step": 1100 }, { "epoch": 0.21960629142348403, "grad_norm": 0.8144860765656456, "learning_rate": 7.310481212920237e-05, "loss": 0.0775, "step": 1110 }, { "epoch": 0.22158472648135324, "grad_norm": 0.8322446467787168, "learning_rate": 7.376400791034937e-05, "loss": 0.072, "step": 1120 }, { "epoch": 0.22356316153922248, "grad_norm": 0.8840235823881915, "learning_rate": 7.442320369149638e-05, "loss": 0.0781, "step": 1130 }, { "epoch": 0.2255415965970917, "grad_norm": 0.7406476619212958, "learning_rate": 7.508239947264338e-05, "loss": 0.067, "step": 1140 }, { "epoch": 0.22752003165496093, "grad_norm": 0.6135846645545082, "learning_rate": 7.574159525379037e-05, "loss": 0.0625, "step": 1150 }, { "epoch": 0.22949846671283014, "grad_norm": 0.8497087947713994, "learning_rate": 7.640079103493738e-05, "loss": 0.0588, "step": 1160 }, { "epoch": 0.23147690177069938, "grad_norm": 0.5684521607051449, "learning_rate": 7.705998681608438e-05, "loss": 0.0559, "step": 1170 }, { "epoch": 0.23345533682856862, "grad_norm": 0.44642133269859746, "learning_rate": 7.771918259723138e-05, "loss": 0.0524, "step": 1180 }, { "epoch": 0.23543377188643783, "grad_norm": 0.7417253246437843, "learning_rate": 7.837837837837838e-05, "loss": 0.049, "step": 1190 }, { "epoch": 0.23741220694430706, "grad_norm": 0.6699824985744847, "learning_rate": 7.903757415952538e-05, "loss": 0.0489, "step": 1200 }, { "epoch": 0.23939064200217627, "grad_norm": 0.4584409365864605, "learning_rate": 7.969676994067239e-05, "loss": 0.0459, "step": 1210 }, { "epoch": 0.2413690770600455, "grad_norm": 0.44232745755322544, "learning_rate": 8.035596572181938e-05, "loss": 0.0436, "step": 1220 }, { "epoch": 0.24334751211791472, "grad_norm": 0.3867957221399885, "learning_rate": 8.101516150296639e-05, "loss": 0.0407, "step": 1230 }, { "epoch": 0.24532594717578396, "grad_norm": 0.41432418911839947, "learning_rate": 8.167435728411339e-05, "loss": 0.0401, "step": 1240 }, { "epoch": 0.24730438223365317, "grad_norm": 0.43958767606215066, "learning_rate": 8.233355306526038e-05, "loss": 0.0385, "step": 1250 }, { "epoch": 0.2492828172915224, "grad_norm": 2.3478935687105587, "learning_rate": 8.299274884640739e-05, "loss": 0.0391, "step": 1260 }, { "epoch": 0.2512612523493916, "grad_norm": 0.7921257388969846, "learning_rate": 8.365194462755439e-05, "loss": 0.0391, "step": 1270 }, { "epoch": 0.2532396874072609, "grad_norm": 0.4159095258664668, "learning_rate": 8.431114040870139e-05, "loss": 0.0364, "step": 1280 }, { "epoch": 0.2552181224651301, "grad_norm": 0.3574794027980809, "learning_rate": 8.497033618984839e-05, "loss": 0.0347, "step": 1290 }, { "epoch": 0.2571965575229993, "grad_norm": 0.3793357762723568, "learning_rate": 8.562953197099539e-05, "loss": 0.0338, "step": 1300 }, { "epoch": 0.2591749925808685, "grad_norm": 0.39061129654899224, "learning_rate": 8.62887277521424e-05, "loss": 0.0322, "step": 1310 }, { "epoch": 0.2611534276387378, "grad_norm": 0.3507422200583872, "learning_rate": 8.694792353328939e-05, "loss": 0.0318, "step": 1320 }, { "epoch": 0.263131862696607, "grad_norm": 0.3316497467819614, "learning_rate": 8.760711931443639e-05, "loss": 0.0309, "step": 1330 }, { "epoch": 0.2651102977544762, "grad_norm": 0.4422978758582334, "learning_rate": 8.82663150955834e-05, "loss": 0.0302, "step": 1340 }, { "epoch": 0.2670887328123454, "grad_norm": 0.42532269482645996, "learning_rate": 8.892551087673038e-05, "loss": 0.0292, "step": 1350 }, { "epoch": 0.2690671678702147, "grad_norm": 1.2751835724526872, "learning_rate": 8.95847066578774e-05, "loss": 0.0322, "step": 1360 }, { "epoch": 0.2710456029280839, "grad_norm": 0.41974258023847216, "learning_rate": 9.02439024390244e-05, "loss": 0.0299, "step": 1370 }, { "epoch": 0.2730240379859531, "grad_norm": 0.32580553825660563, "learning_rate": 9.09030982201714e-05, "loss": 0.0285, "step": 1380 }, { "epoch": 0.27500247304382236, "grad_norm": 0.362899277572937, "learning_rate": 9.15622940013184e-05, "loss": 0.0272, "step": 1390 }, { "epoch": 0.27698090810169157, "grad_norm": 0.31286682503756197, "learning_rate": 9.22214897824654e-05, "loss": 0.0263, "step": 1400 }, { "epoch": 0.2789593431595608, "grad_norm": 0.26773424358272674, "learning_rate": 9.288068556361241e-05, "loss": 0.0255, "step": 1410 }, { "epoch": 0.28093777821743, "grad_norm": 0.40448389034455734, "learning_rate": 9.35398813447594e-05, "loss": 0.0252, "step": 1420 }, { "epoch": 0.28291621327529926, "grad_norm": 0.2783819383288086, "learning_rate": 9.41990771259064e-05, "loss": 0.0241, "step": 1430 }, { "epoch": 0.28489464833316847, "grad_norm": 0.29961272536024675, "learning_rate": 9.48582729070534e-05, "loss": 0.0236, "step": 1440 }, { "epoch": 0.2868730833910377, "grad_norm": 0.24347322009574, "learning_rate": 9.551746868820039e-05, "loss": 0.0228, "step": 1450 }, { "epoch": 0.28885151844890694, "grad_norm": 0.23839323438206578, "learning_rate": 9.61766644693474e-05, "loss": 0.0226, "step": 1460 }, { "epoch": 0.29082995350677615, "grad_norm": 0.26473311552442963, "learning_rate": 9.68358602504944e-05, "loss": 0.0217, "step": 1470 }, { "epoch": 0.29280838856464536, "grad_norm": 0.25862281572683987, "learning_rate": 9.74950560316414e-05, "loss": 0.0211, "step": 1480 }, { "epoch": 0.29478682362251457, "grad_norm": 0.2277270114727975, "learning_rate": 9.81542518127884e-05, "loss": 0.0205, "step": 1490 }, { "epoch": 0.29676525868038384, "grad_norm": 0.24087395405359055, "learning_rate": 9.88134475939354e-05, "loss": 0.02, "step": 1500 }, { "epoch": 0.29874369373825305, "grad_norm": 0.22063387099417836, "learning_rate": 9.94726433750824e-05, "loss": 0.0193, "step": 1510 }, { "epoch": 0.30072212879612226, "grad_norm": 0.20714882477980437, "learning_rate": 9.999999470139298e-05, "loss": 0.0187, "step": 1520 }, { "epoch": 0.30270056385399147, "grad_norm": 0.2302853547965489, "learning_rate": 9.99998092502654e-05, "loss": 0.0184, "step": 1530 }, { "epoch": 0.30467899891186073, "grad_norm": 0.24697567028746728, "learning_rate": 9.999935886991012e-05, "loss": 0.0184, "step": 1540 }, { "epoch": 0.30665743396972994, "grad_norm": 0.2162360843440974, "learning_rate": 9.99986435627135e-05, "loss": 0.0179, "step": 1550 }, { "epoch": 0.30863586902759915, "grad_norm": 0.22785434831280413, "learning_rate": 9.99976633324657e-05, "loss": 0.0174, "step": 1560 }, { "epoch": 0.3106143040854684, "grad_norm": 0.21011992583277403, "learning_rate": 9.999641818436056e-05, "loss": 0.0169, "step": 1570 }, { "epoch": 0.31259273914333763, "grad_norm": 0.23882667469933117, "learning_rate": 9.999490812499562e-05, "loss": 0.0169, "step": 1580 }, { "epoch": 0.31457117420120684, "grad_norm": 0.22869025276994928, "learning_rate": 9.999313316237211e-05, "loss": 0.0162, "step": 1590 }, { "epoch": 0.31654960925907605, "grad_norm": 0.19936558362244672, "learning_rate": 9.999109330589484e-05, "loss": 0.0158, "step": 1600 }, { "epoch": 0.3185280443169453, "grad_norm": 0.18776511028399503, "learning_rate": 9.998878856637221e-05, "loss": 0.0157, "step": 1610 }, { "epoch": 0.3205064793748145, "grad_norm": 0.22047831859103678, "learning_rate": 9.998621895601612e-05, "loss": 0.0151, "step": 1620 }, { "epoch": 0.32248491443268373, "grad_norm": 0.21909712933711048, "learning_rate": 9.998338448844192e-05, "loss": 0.0153, "step": 1630 }, { "epoch": 0.324463349490553, "grad_norm": 0.18342955578512224, "learning_rate": 9.998028517866833e-05, "loss": 0.0147, "step": 1640 }, { "epoch": 0.3264417845484222, "grad_norm": 0.19614528373745874, "learning_rate": 9.997692104311738e-05, "loss": 0.0144, "step": 1650 }, { "epoch": 0.3284202196062914, "grad_norm": 0.1910028678576799, "learning_rate": 9.997329209961429e-05, "loss": 0.0141, "step": 1660 }, { "epoch": 0.33039865466416063, "grad_norm": 0.1875581912159569, "learning_rate": 9.996939836738737e-05, "loss": 0.0138, "step": 1670 }, { "epoch": 0.3323770897220299, "grad_norm": 0.21318059322872368, "learning_rate": 9.9965239867068e-05, "loss": 0.0138, "step": 1680 }, { "epoch": 0.3343555247798991, "grad_norm": 0.15100583828371544, "learning_rate": 9.996081662069043e-05, "loss": 0.0138, "step": 1690 }, { "epoch": 0.3363339598377683, "grad_norm": 0.16665345852934302, "learning_rate": 9.995612865169168e-05, "loss": 0.0136, "step": 1700 }, { "epoch": 0.3383123948956375, "grad_norm": 0.9719045950170202, "learning_rate": 9.995117598491145e-05, "loss": 0.0277, "step": 1710 }, { "epoch": 0.3402908299535068, "grad_norm": 3.161728848159066, "learning_rate": 9.994595864659197e-05, "loss": 0.0406, "step": 1720 }, { "epoch": 0.342269265011376, "grad_norm": 3.6147429552746417, "learning_rate": 9.994047666437785e-05, "loss": 0.0542, "step": 1730 }, { "epoch": 0.3442477000692452, "grad_norm": 0.2735534000567824, "learning_rate": 9.993473006731592e-05, "loss": 0.022, "step": 1740 }, { "epoch": 0.3462261351271145, "grad_norm": 0.1810803504663682, "learning_rate": 9.992871888585518e-05, "loss": 0.016, "step": 1750 }, { "epoch": 0.3482045701849837, "grad_norm": 0.14853009791786845, "learning_rate": 9.992244315184647e-05, "loss": 0.0143, "step": 1760 }, { "epoch": 0.3501830052428529, "grad_norm": 0.1508942913302995, "learning_rate": 9.991590289854243e-05, "loss": 0.0136, "step": 1770 }, { "epoch": 0.3521614403007221, "grad_norm": 0.14224740250692022, "learning_rate": 9.990909816059725e-05, "loss": 0.0126, "step": 1780 }, { "epoch": 0.35413987535859137, "grad_norm": 0.13888816460678802, "learning_rate": 9.99020289740666e-05, "loss": 0.0124, "step": 1790 }, { "epoch": 0.3561183104164606, "grad_norm": 0.14941569438589294, "learning_rate": 9.989469537640728e-05, "loss": 0.0122, "step": 1800 }, { "epoch": 0.3580967454743298, "grad_norm": 0.14008372469353464, "learning_rate": 9.988709740647713e-05, "loss": 0.0118, "step": 1810 }, { "epoch": 0.36007518053219906, "grad_norm": 0.1471251068367602, "learning_rate": 9.987923510453478e-05, "loss": 0.0119, "step": 1820 }, { "epoch": 0.36205361559006827, "grad_norm": 0.11583487666072863, "learning_rate": 9.987110851223947e-05, "loss": 0.0118, "step": 1830 }, { "epoch": 0.3640320506479375, "grad_norm": 0.14009606580200387, "learning_rate": 9.986271767265079e-05, "loss": 0.0117, "step": 1840 }, { "epoch": 0.3660104857058067, "grad_norm": 0.15359018530615084, "learning_rate": 9.985406263022849e-05, "loss": 0.0115, "step": 1850 }, { "epoch": 0.36798892076367595, "grad_norm": 0.13794481098483324, "learning_rate": 9.984514343083221e-05, "loss": 0.0115, "step": 1860 }, { "epoch": 0.36996735582154516, "grad_norm": 0.1338257809163921, "learning_rate": 9.983596012172127e-05, "loss": 0.0113, "step": 1870 }, { "epoch": 0.3719457908794144, "grad_norm": 0.11194861485256917, "learning_rate": 9.982651275155441e-05, "loss": 0.011, "step": 1880 }, { "epoch": 0.3739242259372836, "grad_norm": 0.13570239156331954, "learning_rate": 9.981680137038949e-05, "loss": 0.0108, "step": 1890 }, { "epoch": 0.37590266099515285, "grad_norm": 0.12947791555980262, "learning_rate": 9.980682602968329e-05, "loss": 0.0107, "step": 1900 }, { "epoch": 0.37788109605302206, "grad_norm": 0.11846270267188591, "learning_rate": 9.97965867822912e-05, "loss": 0.0105, "step": 1910 }, { "epoch": 0.37985953111089127, "grad_norm": 0.12775236471738088, "learning_rate": 9.978608368246693e-05, "loss": 0.0108, "step": 1920 }, { "epoch": 0.38183796616876053, "grad_norm": 0.11510206230282105, "learning_rate": 9.977531678586227e-05, "loss": 0.0102, "step": 1930 }, { "epoch": 0.38381640122662974, "grad_norm": 0.1167495841995747, "learning_rate": 9.976428614952675e-05, "loss": 0.0101, "step": 1940 }, { "epoch": 0.38579483628449895, "grad_norm": 0.1371366840889005, "learning_rate": 9.975299183190734e-05, "loss": 0.0102, "step": 1950 }, { "epoch": 0.38777327134236816, "grad_norm": 0.12495466419939394, "learning_rate": 9.974143389284818e-05, "loss": 0.0099, "step": 1960 }, { "epoch": 0.38975170640023743, "grad_norm": 0.1222265321586572, "learning_rate": 9.97296123935902e-05, "loss": 0.0099, "step": 1970 }, { "epoch": 0.39173014145810664, "grad_norm": 0.10581669983906558, "learning_rate": 9.971752739677088e-05, "loss": 0.0098, "step": 1980 }, { "epoch": 0.39370857651597585, "grad_norm": 0.1031261631363871, "learning_rate": 9.970517896642382e-05, "loss": 0.0096, "step": 1990 }, { "epoch": 0.3956870115738451, "grad_norm": 0.1098854853995119, "learning_rate": 9.969256716797848e-05, "loss": 0.0098, "step": 2000 }, { "epoch": 0.3976654466317143, "grad_norm": 0.11173021888171392, "learning_rate": 9.967969206825977e-05, "loss": 0.0093, "step": 2010 }, { "epoch": 0.39964388168958354, "grad_norm": 0.11939922842820136, "learning_rate": 9.96665537354878e-05, "loss": 0.0096, "step": 2020 }, { "epoch": 0.40162231674745275, "grad_norm": 0.12434579253503734, "learning_rate": 9.965315223927741e-05, "loss": 0.0095, "step": 2030 }, { "epoch": 0.403600751805322, "grad_norm": 0.10397011769498186, "learning_rate": 9.963948765063778e-05, "loss": 0.0096, "step": 2040 }, { "epoch": 0.4055791868631912, "grad_norm": 0.10874306421551215, "learning_rate": 9.962556004197217e-05, "loss": 0.0093, "step": 2050 }, { "epoch": 0.40755762192106043, "grad_norm": 0.13220931485486098, "learning_rate": 9.961136948707752e-05, "loss": 0.0093, "step": 2060 }, { "epoch": 0.40953605697892964, "grad_norm": 0.1296809440842965, "learning_rate": 9.959691606114393e-05, "loss": 0.0093, "step": 2070 }, { "epoch": 0.4115144920367989, "grad_norm": 0.1179090438361135, "learning_rate": 9.958219984075441e-05, "loss": 0.0093, "step": 2080 }, { "epoch": 0.4134929270946681, "grad_norm": 0.11884656631643735, "learning_rate": 9.956722090388438e-05, "loss": 0.009, "step": 2090 }, { "epoch": 0.4154713621525373, "grad_norm": 0.11716687338707497, "learning_rate": 9.955197932990132e-05, "loss": 0.009, "step": 2100 }, { "epoch": 0.4174497972104066, "grad_norm": 0.10664497146695569, "learning_rate": 9.953647519956433e-05, "loss": 0.0091, "step": 2110 }, { "epoch": 0.4194282322682758, "grad_norm": 0.11903304887750939, "learning_rate": 9.952070859502361e-05, "loss": 0.0088, "step": 2120 }, { "epoch": 0.421406667326145, "grad_norm": 0.09738834474502249, "learning_rate": 9.950467959982021e-05, "loss": 0.0088, "step": 2130 }, { "epoch": 0.4233851023840142, "grad_norm": 0.09920220236467324, "learning_rate": 9.948838829888543e-05, "loss": 0.0088, "step": 2140 }, { "epoch": 0.4253635374418835, "grad_norm": 0.09619743321048915, "learning_rate": 9.947183477854042e-05, "loss": 0.0088, "step": 2150 }, { "epoch": 0.4273419724997527, "grad_norm": 0.0974254141180528, "learning_rate": 9.945501912649579e-05, "loss": 0.0087, "step": 2160 }, { "epoch": 0.4293204075576219, "grad_norm": 0.09443148747455482, "learning_rate": 9.943794143185096e-05, "loss": 0.0087, "step": 2170 }, { "epoch": 0.4312988426154912, "grad_norm": 0.09350468571118675, "learning_rate": 9.942060178509396e-05, "loss": 0.0086, "step": 2180 }, { "epoch": 0.4332772776733604, "grad_norm": 0.09070704335772732, "learning_rate": 9.940300027810067e-05, "loss": 0.0085, "step": 2190 }, { "epoch": 0.4352557127312296, "grad_norm": 0.12110216964097328, "learning_rate": 9.938513700413454e-05, "loss": 0.0087, "step": 2200 }, { "epoch": 0.4372341477890988, "grad_norm": 0.09890163894885512, "learning_rate": 9.9367012057846e-05, "loss": 0.0085, "step": 2210 }, { "epoch": 0.43921258284696807, "grad_norm": 0.11128753159509124, "learning_rate": 9.934862553527198e-05, "loss": 0.0084, "step": 2220 }, { "epoch": 0.4411910179048373, "grad_norm": 0.10195595147527749, "learning_rate": 9.932997753383537e-05, "loss": 0.0084, "step": 2230 }, { "epoch": 0.4431694529627065, "grad_norm": 0.1035635647276319, "learning_rate": 9.931106815234459e-05, "loss": 0.0084, "step": 2240 }, { "epoch": 0.4451478880205757, "grad_norm": 0.12526375331576606, "learning_rate": 9.929189749099296e-05, "loss": 0.0082, "step": 2250 }, { "epoch": 0.44712632307844496, "grad_norm": 0.10124687119197032, "learning_rate": 9.927246565135826e-05, "loss": 0.0085, "step": 2260 }, { "epoch": 0.4491047581363142, "grad_norm": 0.12655923047390247, "learning_rate": 9.92527727364021e-05, "loss": 0.0084, "step": 2270 }, { "epoch": 0.4510831931941834, "grad_norm": 0.12234433416748676, "learning_rate": 9.923281885046949e-05, "loss": 0.0082, "step": 2280 }, { "epoch": 0.45306162825205265, "grad_norm": 0.10038433023346938, "learning_rate": 9.921260409928814e-05, "loss": 0.0081, "step": 2290 }, { "epoch": 0.45504006330992186, "grad_norm": 0.10592374911299951, "learning_rate": 9.919212858996804e-05, "loss": 0.0082, "step": 2300 }, { "epoch": 0.45701849836779107, "grad_norm": 0.11256516289439368, "learning_rate": 9.917139243100087e-05, "loss": 0.0083, "step": 2310 }, { "epoch": 0.4589969334256603, "grad_norm": 0.10941118440635397, "learning_rate": 9.915039573225926e-05, "loss": 0.0083, "step": 2320 }, { "epoch": 0.46097536848352955, "grad_norm": 0.09076166231590097, "learning_rate": 9.912913860499649e-05, "loss": 0.0081, "step": 2330 }, { "epoch": 0.46295380354139876, "grad_norm": 0.09539136003050565, "learning_rate": 9.910762116184562e-05, "loss": 0.0079, "step": 2340 }, { "epoch": 0.46493223859926797, "grad_norm": 0.09920800558172575, "learning_rate": 9.908584351681911e-05, "loss": 0.0079, "step": 2350 }, { "epoch": 0.46691067365713723, "grad_norm": 0.07883437880669862, "learning_rate": 9.906380578530809e-05, "loss": 0.0078, "step": 2360 }, { "epoch": 0.46888910871500644, "grad_norm": 0.08844392359513846, "learning_rate": 9.904150808408178e-05, "loss": 0.0076, "step": 2370 }, { "epoch": 0.47086754377287565, "grad_norm": 0.08782650023677235, "learning_rate": 9.901895053128689e-05, "loss": 0.0079, "step": 2380 }, { "epoch": 0.47284597883074486, "grad_norm": 0.10263452740495278, "learning_rate": 9.899613324644698e-05, "loss": 0.0078, "step": 2390 }, { "epoch": 0.4748244138886141, "grad_norm": 0.07539712980685903, "learning_rate": 9.897305635046183e-05, "loss": 0.0077, "step": 2400 }, { "epoch": 0.47680284894648334, "grad_norm": 0.09154103611980138, "learning_rate": 9.894971996560678e-05, "loss": 0.0078, "step": 2410 }, { "epoch": 0.47878128400435255, "grad_norm": 0.0949233826823644, "learning_rate": 9.89261242155321e-05, "loss": 0.0075, "step": 2420 }, { "epoch": 0.48075971906222176, "grad_norm": 0.08791399663054178, "learning_rate": 9.890226922526237e-05, "loss": 0.0076, "step": 2430 }, { "epoch": 0.482738154120091, "grad_norm": 0.08812362893437024, "learning_rate": 9.887815512119574e-05, "loss": 0.0077, "step": 2440 }, { "epoch": 0.48471658917796023, "grad_norm": 0.08865882789841384, "learning_rate": 9.885378203110331e-05, "loss": 0.0077, "step": 2450 }, { "epoch": 0.48669502423582944, "grad_norm": 0.08302609862666924, "learning_rate": 9.882915008412847e-05, "loss": 0.0075, "step": 2460 }, { "epoch": 0.4886734592936987, "grad_norm": 0.0894044049025858, "learning_rate": 9.880425941078617e-05, "loss": 0.0076, "step": 2470 }, { "epoch": 0.4906518943515679, "grad_norm": 0.0875976993567813, "learning_rate": 9.877911014296223e-05, "loss": 0.0075, "step": 2480 }, { "epoch": 0.49263032940943713, "grad_norm": 0.07532701277591465, "learning_rate": 9.875370241391269e-05, "loss": 0.0075, "step": 2490 }, { "epoch": 0.49460876446730634, "grad_norm": 0.07966759962069439, "learning_rate": 9.872803635826307e-05, "loss": 0.0075, "step": 2500 }, { "epoch": 0.4965871995251756, "grad_norm": 0.09858397509750241, "learning_rate": 9.870211211200766e-05, "loss": 0.0073, "step": 2510 }, { "epoch": 0.4985656345830448, "grad_norm": 0.09319554063457361, "learning_rate": 9.867592981250878e-05, "loss": 0.0074, "step": 2520 }, { "epoch": 0.500544069640914, "grad_norm": 0.12781117254542, "learning_rate": 9.864948959849609e-05, "loss": 0.0074, "step": 2530 }, { "epoch": 0.5025225046987832, "grad_norm": 0.26482548077665524, "learning_rate": 9.862279161006585e-05, "loss": 0.0098, "step": 2540 }, { "epoch": 0.5045009397566524, "grad_norm": 1.3380179010337696, "learning_rate": 9.859583598868012e-05, "loss": 0.0138, "step": 2550 }, { "epoch": 0.5064793748145218, "grad_norm": 0.5324160366432114, "learning_rate": 9.856862287716613e-05, "loss": 0.0651, "step": 2560 }, { "epoch": 0.508457809872391, "grad_norm": 0.20795257366814654, "learning_rate": 9.854115241971535e-05, "loss": 0.0156, "step": 2570 }, { "epoch": 0.5104362449302602, "grad_norm": 0.1358973049313345, "learning_rate": 9.85134247618829e-05, "loss": 0.0104, "step": 2580 }, { "epoch": 0.5124146799881294, "grad_norm": 0.11303985103375866, "learning_rate": 9.848544005058668e-05, "loss": 0.0088, "step": 2590 }, { "epoch": 0.5143931150459986, "grad_norm": 0.07561229737159056, "learning_rate": 9.845719843410663e-05, "loss": 0.0081, "step": 2600 }, { "epoch": 0.5163715501038678, "grad_norm": 0.08491679400058451, "learning_rate": 9.842870006208389e-05, "loss": 0.0078, "step": 2610 }, { "epoch": 0.518349985161737, "grad_norm": 0.08114020479286879, "learning_rate": 9.839994508552006e-05, "loss": 0.0077, "step": 2620 }, { "epoch": 0.5203284202196063, "grad_norm": 0.07439429710508837, "learning_rate": 9.837093365677644e-05, "loss": 0.0073, "step": 2630 }, { "epoch": 0.5223068552774756, "grad_norm": 0.07588934608576213, "learning_rate": 9.83416659295731e-05, "loss": 0.0075, "step": 2640 }, { "epoch": 0.5242852903353448, "grad_norm": 0.07727254521424115, "learning_rate": 9.831214205898814e-05, "loss": 0.0074, "step": 2650 }, { "epoch": 0.526263725393214, "grad_norm": 0.0770544746274172, "learning_rate": 9.82823622014569e-05, "loss": 0.0073, "step": 2660 }, { "epoch": 0.5282421604510832, "grad_norm": 0.07507052238833033, "learning_rate": 9.825232651477109e-05, "loss": 0.0072, "step": 2670 }, { "epoch": 0.5302205955089524, "grad_norm": 0.08038501538455785, "learning_rate": 9.82220351580779e-05, "loss": 0.0071, "step": 2680 }, { "epoch": 0.5321990305668216, "grad_norm": 0.07064608286766139, "learning_rate": 9.819148829187932e-05, "loss": 0.007, "step": 2690 }, { "epoch": 0.5341774656246908, "grad_norm": 0.06202757033774752, "learning_rate": 9.816068607803107e-05, "loss": 0.0071, "step": 2700 }, { "epoch": 0.5361559006825601, "grad_norm": 0.06498278906559196, "learning_rate": 9.812962867974191e-05, "loss": 0.007, "step": 2710 }, { "epoch": 0.5381343357404293, "grad_norm": 0.06889168223262838, "learning_rate": 9.809831626157274e-05, "loss": 0.0069, "step": 2720 }, { "epoch": 0.5401127707982986, "grad_norm": 0.07658902219040475, "learning_rate": 9.806674898943569e-05, "loss": 0.007, "step": 2730 }, { "epoch": 0.5420912058561678, "grad_norm": 0.065227283289949, "learning_rate": 9.803492703059323e-05, "loss": 0.007, "step": 2740 }, { "epoch": 0.544069640914037, "grad_norm": 0.07285925886476977, "learning_rate": 9.800285055365736e-05, "loss": 0.0069, "step": 2750 }, { "epoch": 0.5460480759719062, "grad_norm": 0.07094206575560595, "learning_rate": 9.797051972858867e-05, "loss": 0.0069, "step": 2760 }, { "epoch": 0.5480265110297754, "grad_norm": 0.07226595292182501, "learning_rate": 9.79379347266954e-05, "loss": 0.0068, "step": 2770 }, { "epoch": 0.5500049460876447, "grad_norm": 0.07467132239176189, "learning_rate": 9.79050957206326e-05, "loss": 0.007, "step": 2780 }, { "epoch": 0.5519833811455139, "grad_norm": 0.10178890401636641, "learning_rate": 9.787200288440119e-05, "loss": 0.0069, "step": 2790 }, { "epoch": 0.5539618162033831, "grad_norm": 0.08092189483997364, "learning_rate": 9.783865639334705e-05, "loss": 0.0069, "step": 2800 }, { "epoch": 0.5559402512612523, "grad_norm": 0.05878477071038792, "learning_rate": 9.780505642416001e-05, "loss": 0.0067, "step": 2810 }, { "epoch": 0.5579186863191216, "grad_norm": 0.05530654925471683, "learning_rate": 9.777120315487306e-05, "loss": 0.0066, "step": 2820 }, { "epoch": 0.5598971213769908, "grad_norm": 0.0686536401780331, "learning_rate": 9.773709676486129e-05, "loss": 0.0067, "step": 2830 }, { "epoch": 0.56187555643486, "grad_norm": 0.06035416233845031, "learning_rate": 9.770273743484097e-05, "loss": 0.0068, "step": 2840 }, { "epoch": 0.5638539914927293, "grad_norm": 0.05803243981437283, "learning_rate": 9.766812534686863e-05, "loss": 0.0066, "step": 2850 }, { "epoch": 0.5658324265505985, "grad_norm": 0.06995933649531104, "learning_rate": 9.763326068434e-05, "loss": 0.0066, "step": 2860 }, { "epoch": 0.5678108616084677, "grad_norm": 0.0605332765649623, "learning_rate": 9.75981436319892e-05, "loss": 0.0066, "step": 2870 }, { "epoch": 0.5697892966663369, "grad_norm": 0.06737579928910571, "learning_rate": 9.756277437588759e-05, "loss": 0.0067, "step": 2880 }, { "epoch": 0.5717677317242061, "grad_norm": 0.06189558705755954, "learning_rate": 9.752715310344288e-05, "loss": 0.0066, "step": 2890 }, { "epoch": 0.5737461667820754, "grad_norm": 0.06854647227258914, "learning_rate": 9.74912800033981e-05, "loss": 0.0067, "step": 2900 }, { "epoch": 0.5757246018399446, "grad_norm": 0.06334049765668723, "learning_rate": 9.745515526583065e-05, "loss": 0.0067, "step": 2910 }, { "epoch": 0.5777030368978139, "grad_norm": 0.07265687833634417, "learning_rate": 9.741877908215123e-05, "loss": 0.0067, "step": 2920 }, { "epoch": 0.5796814719556831, "grad_norm": 0.06912150801140689, "learning_rate": 9.738215164510286e-05, "loss": 0.0067, "step": 2930 }, { "epoch": 0.5816599070135523, "grad_norm": 0.06380076622359117, "learning_rate": 9.734527314875985e-05, "loss": 0.0066, "step": 2940 }, { "epoch": 0.5836383420714215, "grad_norm": 0.0748767974202622, "learning_rate": 9.730814378852678e-05, "loss": 0.0067, "step": 2950 }, { "epoch": 0.5856167771292907, "grad_norm": 0.06800504603033558, "learning_rate": 9.727076376113745e-05, "loss": 0.0065, "step": 2960 }, { "epoch": 0.5875952121871599, "grad_norm": 0.059588811804270385, "learning_rate": 9.723313326465384e-05, "loss": 0.0065, "step": 2970 }, { "epoch": 0.5895736472450291, "grad_norm": 0.05354398140471879, "learning_rate": 9.71952524984651e-05, "loss": 0.0064, "step": 2980 }, { "epoch": 0.5915520823028984, "grad_norm": 0.06992506261058357, "learning_rate": 9.715712166328644e-05, "loss": 0.0065, "step": 2990 }, { "epoch": 0.5935305173607677, "grad_norm": 0.07328603458194191, "learning_rate": 9.711874096115805e-05, "loss": 0.0064, "step": 3000 }, { "epoch": 0.5955089524186369, "grad_norm": 0.05746748382423511, "learning_rate": 9.708011059544415e-05, "loss": 0.0064, "step": 3010 }, { "epoch": 0.5974873874765061, "grad_norm": 0.05877461678841951, "learning_rate": 9.704123077083176e-05, "loss": 0.0065, "step": 3020 }, { "epoch": 0.5994658225343753, "grad_norm": 0.051196419888608125, "learning_rate": 9.700210169332968e-05, "loss": 0.0063, "step": 3030 }, { "epoch": 0.6014442575922445, "grad_norm": 0.06331233148415444, "learning_rate": 9.696272357026746e-05, "loss": 0.0063, "step": 3040 }, { "epoch": 0.6034226926501137, "grad_norm": 0.061860321869425514, "learning_rate": 9.692309661029421e-05, "loss": 0.0064, "step": 3050 }, { "epoch": 0.6054011277079829, "grad_norm": 0.067519120947699, "learning_rate": 9.68832210233775e-05, "loss": 0.0066, "step": 3060 }, { "epoch": 0.6073795627658523, "grad_norm": 0.07524005887985293, "learning_rate": 9.684309702080233e-05, "loss": 0.0065, "step": 3070 }, { "epoch": 0.6093579978237215, "grad_norm": 0.06070609348728366, "learning_rate": 9.680272481516992e-05, "loss": 0.0065, "step": 3080 }, { "epoch": 0.6113364328815907, "grad_norm": 0.12817129560256088, "learning_rate": 9.676210462039664e-05, "loss": 0.0064, "step": 3090 }, { "epoch": 0.6133148679394599, "grad_norm": 0.06537155344505334, "learning_rate": 9.672123665171285e-05, "loss": 0.0065, "step": 3100 }, { "epoch": 0.6152933029973291, "grad_norm": 0.08187746090238779, "learning_rate": 9.668012112566174e-05, "loss": 0.0065, "step": 3110 }, { "epoch": 0.6172717380551983, "grad_norm": 0.06284585792273872, "learning_rate": 9.663875826009824e-05, "loss": 0.0063, "step": 3120 }, { "epoch": 0.6192501731130675, "grad_norm": 0.058579617677076326, "learning_rate": 9.659714827418786e-05, "loss": 0.0064, "step": 3130 }, { "epoch": 0.6212286081709368, "grad_norm": 0.05671718811117095, "learning_rate": 9.655529138840543e-05, "loss": 0.0063, "step": 3140 }, { "epoch": 0.623207043228806, "grad_norm": 0.058903454628710904, "learning_rate": 9.651318782453407e-05, "loss": 0.0062, "step": 3150 }, { "epoch": 0.6251854782866753, "grad_norm": 0.05670298829502456, "learning_rate": 9.64708378056639e-05, "loss": 0.0062, "step": 3160 }, { "epoch": 0.6271639133445445, "grad_norm": 0.05428876166111194, "learning_rate": 9.642824155619095e-05, "loss": 0.0062, "step": 3170 }, { "epoch": 0.6291423484024137, "grad_norm": 0.06254574460432358, "learning_rate": 9.638539930181592e-05, "loss": 0.0063, "step": 3180 }, { "epoch": 0.6311207834602829, "grad_norm": 0.06672441168101174, "learning_rate": 9.634231126954296e-05, "loss": 0.0064, "step": 3190 }, { "epoch": 0.6330992185181521, "grad_norm": 0.05090452979591465, "learning_rate": 9.629897768767853e-05, "loss": 0.0062, "step": 3200 }, { "epoch": 0.6350776535760214, "grad_norm": 0.05765847904822037, "learning_rate": 9.625539878583016e-05, "loss": 0.0063, "step": 3210 }, { "epoch": 0.6370560886338906, "grad_norm": 0.05041445650810007, "learning_rate": 9.621157479490521e-05, "loss": 0.0063, "step": 3220 }, { "epoch": 0.6390345236917598, "grad_norm": 0.04947260615273494, "learning_rate": 9.616750594710973e-05, "loss": 0.0064, "step": 3230 }, { "epoch": 0.641012958749629, "grad_norm": 0.052930154093284505, "learning_rate": 9.612319247594707e-05, "loss": 0.0062, "step": 3240 }, { "epoch": 0.6429913938074983, "grad_norm": 0.054125747626247174, "learning_rate": 9.607863461621685e-05, "loss": 0.0062, "step": 3250 }, { "epoch": 0.6449698288653675, "grad_norm": 0.05710421262162926, "learning_rate": 9.603383260401352e-05, "loss": 0.0063, "step": 3260 }, { "epoch": 0.6469482639232367, "grad_norm": 0.05398629342840845, "learning_rate": 9.598878667672526e-05, "loss": 0.0063, "step": 3270 }, { "epoch": 0.648926698981106, "grad_norm": 0.051946230834141204, "learning_rate": 9.594349707303262e-05, "loss": 0.0062, "step": 3280 }, { "epoch": 0.6509051340389752, "grad_norm": 0.05510427993338031, "learning_rate": 9.589796403290732e-05, "loss": 0.0062, "step": 3290 }, { "epoch": 0.6528835690968444, "grad_norm": 0.05585402519049233, "learning_rate": 9.585218779761093e-05, "loss": 0.0062, "step": 3300 }, { "epoch": 0.6548620041547136, "grad_norm": 0.05801028715682649, "learning_rate": 9.580616860969364e-05, "loss": 0.0062, "step": 3310 }, { "epoch": 0.6568404392125828, "grad_norm": 0.053932109311021714, "learning_rate": 9.575990671299295e-05, "loss": 0.0062, "step": 3320 }, { "epoch": 0.658818874270452, "grad_norm": 0.05924132673349397, "learning_rate": 9.571340235263236e-05, "loss": 0.0061, "step": 3330 }, { "epoch": 0.6607973093283213, "grad_norm": 0.06107219838029144, "learning_rate": 9.566665577502008e-05, "loss": 0.0062, "step": 3340 }, { "epoch": 0.6627757443861905, "grad_norm": 0.055948574313692906, "learning_rate": 9.561966722784774e-05, "loss": 0.0063, "step": 3350 }, { "epoch": 0.6647541794440598, "grad_norm": 0.056699774353642714, "learning_rate": 9.557243696008912e-05, "loss": 0.0062, "step": 3360 }, { "epoch": 0.666732614501929, "grad_norm": 0.06647034643900732, "learning_rate": 9.552496522199872e-05, "loss": 0.0062, "step": 3370 }, { "epoch": 0.6687110495597982, "grad_norm": 0.062109360688086826, "learning_rate": 9.54772522651105e-05, "loss": 0.0062, "step": 3380 }, { "epoch": 0.6706894846176674, "grad_norm": 0.0443074645474432, "learning_rate": 9.54292983422366e-05, "loss": 0.0062, "step": 3390 }, { "epoch": 0.6726679196755366, "grad_norm": 0.7126545771408139, "learning_rate": 9.538110370746587e-05, "loss": 0.0135, "step": 3400 }, { "epoch": 0.6746463547334058, "grad_norm": 1.838740056177238, "learning_rate": 9.533266861616264e-05, "loss": 0.0157, "step": 3410 }, { "epoch": 0.676624789791275, "grad_norm": 2.313205467450454, "learning_rate": 9.528399332496533e-05, "loss": 0.0903, "step": 3420 }, { "epoch": 0.6786032248491444, "grad_norm": 0.19913027405735842, "learning_rate": 9.523507809178506e-05, "loss": 0.0211, "step": 3430 }, { "epoch": 0.6805816599070136, "grad_norm": 0.11775151842160198, "learning_rate": 9.518592317580432e-05, "loss": 0.0103, "step": 3440 }, { "epoch": 0.6825600949648828, "grad_norm": 0.07434810113025182, "learning_rate": 9.513652883747558e-05, "loss": 0.008, "step": 3450 }, { "epoch": 0.684538530022752, "grad_norm": 0.07557817167064859, "learning_rate": 9.508689533851991e-05, "loss": 0.0073, "step": 3460 }, { "epoch": 0.6865169650806212, "grad_norm": 0.06071379140394132, "learning_rate": 9.503702294192563e-05, "loss": 0.0068, "step": 3470 }, { "epoch": 0.6884954001384904, "grad_norm": 0.063060210146848, "learning_rate": 9.498691191194682e-05, "loss": 0.0067, "step": 3480 }, { "epoch": 0.6904738351963596, "grad_norm": 0.05653952131526833, "learning_rate": 9.493656251410204e-05, "loss": 0.0066, "step": 3490 }, { "epoch": 0.692452270254229, "grad_norm": 0.07485922990062185, "learning_rate": 9.488597501517287e-05, "loss": 0.0066, "step": 3500 }, { "epoch": 0.6944307053120982, "grad_norm": 0.05282320165364128, "learning_rate": 9.483514968320243e-05, "loss": 0.0064, "step": 3510 }, { "epoch": 0.6964091403699674, "grad_norm": 0.05066162976905024, "learning_rate": 9.478408678749412e-05, "loss": 0.0062, "step": 3520 }, { "epoch": 0.6983875754278366, "grad_norm": 0.06086114944598497, "learning_rate": 9.473278659861e-05, "loss": 0.0063, "step": 3530 }, { "epoch": 0.7003660104857058, "grad_norm": 0.0585517952529397, "learning_rate": 9.46812493883695e-05, "loss": 0.0062, "step": 3540 }, { "epoch": 0.702344445543575, "grad_norm": 0.05465672190075681, "learning_rate": 9.462947542984795e-05, "loss": 0.0061, "step": 3550 }, { "epoch": 0.7043228806014442, "grad_norm": 0.05460020875531365, "learning_rate": 9.457746499737508e-05, "loss": 0.0063, "step": 3560 }, { "epoch": 0.7063013156593135, "grad_norm": 0.05449513153115839, "learning_rate": 9.45252183665336e-05, "loss": 0.0061, "step": 3570 }, { "epoch": 0.7082797507171827, "grad_norm": 0.05912074293349835, "learning_rate": 9.44727358141578e-05, "loss": 0.0063, "step": 3580 }, { "epoch": 0.710258185775052, "grad_norm": 0.05139678969575524, "learning_rate": 9.442001761833194e-05, "loss": 0.0062, "step": 3590 }, { "epoch": 0.7122366208329212, "grad_norm": 0.042382018800827065, "learning_rate": 9.436706405838892e-05, "loss": 0.006, "step": 3600 }, { "epoch": 0.7142150558907904, "grad_norm": 0.05503643239385127, "learning_rate": 9.431387541490872e-05, "loss": 0.006, "step": 3610 }, { "epoch": 0.7161934909486596, "grad_norm": 0.03781684994462094, "learning_rate": 9.426045196971695e-05, "loss": 0.006, "step": 3620 }, { "epoch": 0.7181719260065288, "grad_norm": 0.0527864556724502, "learning_rate": 9.420679400588334e-05, "loss": 0.0059, "step": 3630 }, { "epoch": 0.7201503610643981, "grad_norm": 0.05459511268525548, "learning_rate": 9.41529018077202e-05, "loss": 0.006, "step": 3640 }, { "epoch": 0.7221287961222673, "grad_norm": 0.04960495815440276, "learning_rate": 9.409877566078101e-05, "loss": 0.006, "step": 3650 }, { "epoch": 0.7241072311801365, "grad_norm": 0.055543567730973464, "learning_rate": 9.404441585185883e-05, "loss": 0.0061, "step": 3660 }, { "epoch": 0.7260856662380057, "grad_norm": 0.04180945934142346, "learning_rate": 9.398982266898481e-05, "loss": 0.0059, "step": 3670 }, { "epoch": 0.728064101295875, "grad_norm": 0.05250715491470673, "learning_rate": 9.393499640142662e-05, "loss": 0.006, "step": 3680 }, { "epoch": 0.7300425363537442, "grad_norm": 0.04952053334810146, "learning_rate": 9.387993733968702e-05, "loss": 0.006, "step": 3690 }, { "epoch": 0.7320209714116134, "grad_norm": 0.04601226439857965, "learning_rate": 9.382464577550218e-05, "loss": 0.006, "step": 3700 }, { "epoch": 0.7339994064694826, "grad_norm": 0.04661713153431959, "learning_rate": 9.376912200184029e-05, "loss": 0.0059, "step": 3710 }, { "epoch": 0.7359778415273519, "grad_norm": 0.04372508938479629, "learning_rate": 9.371336631289984e-05, "loss": 0.006, "step": 3720 }, { "epoch": 0.7379562765852211, "grad_norm": 0.0439169479340367, "learning_rate": 9.365737900410821e-05, "loss": 0.0059, "step": 3730 }, { "epoch": 0.7399347116430903, "grad_norm": 0.042307373252861456, "learning_rate": 9.360116037212002e-05, "loss": 0.0058, "step": 3740 }, { "epoch": 0.7419131467009595, "grad_norm": 0.04272057933986619, "learning_rate": 9.354471071481557e-05, "loss": 0.0058, "step": 3750 }, { "epoch": 0.7438915817588287, "grad_norm": 0.05490715602102348, "learning_rate": 9.348803033129929e-05, "loss": 0.0059, "step": 3760 }, { "epoch": 0.745870016816698, "grad_norm": 0.05448304267785827, "learning_rate": 9.343111952189814e-05, "loss": 0.0059, "step": 3770 }, { "epoch": 0.7478484518745672, "grad_norm": 0.04831361026502636, "learning_rate": 9.337397858816e-05, "loss": 0.0059, "step": 3780 }, { "epoch": 0.7498268869324365, "grad_norm": 0.0437452756514658, "learning_rate": 9.331660783285208e-05, "loss": 0.0059, "step": 3790 }, { "epoch": 0.7518053219903057, "grad_norm": 0.04723328131719851, "learning_rate": 9.325900755995938e-05, "loss": 0.0059, "step": 3800 }, { "epoch": 0.7537837570481749, "grad_norm": 0.053497571434574016, "learning_rate": 9.320117807468292e-05, "loss": 0.006, "step": 3810 }, { "epoch": 0.7557621921060441, "grad_norm": 0.043023889738370805, "learning_rate": 9.314311968343833e-05, "loss": 0.0059, "step": 3820 }, { "epoch": 0.7577406271639133, "grad_norm": 0.04948447693500051, "learning_rate": 9.308483269385406e-05, "loss": 0.006, "step": 3830 }, { "epoch": 0.7597190622217825, "grad_norm": 0.04733269561829835, "learning_rate": 9.302631741476984e-05, "loss": 0.0058, "step": 3840 }, { "epoch": 0.7616974972796517, "grad_norm": 0.044990806731528936, "learning_rate": 9.2967574156235e-05, "loss": 0.0058, "step": 3850 }, { "epoch": 0.7636759323375211, "grad_norm": 0.048432209335944686, "learning_rate": 9.290860322950687e-05, "loss": 0.0058, "step": 3860 }, { "epoch": 0.7656543673953903, "grad_norm": 0.038321655151852915, "learning_rate": 9.284940494704906e-05, "loss": 0.0059, "step": 3870 }, { "epoch": 0.7676328024532595, "grad_norm": 0.05363631230889327, "learning_rate": 9.278997962252986e-05, "loss": 0.0058, "step": 3880 }, { "epoch": 0.7696112375111287, "grad_norm": 0.04286373379738482, "learning_rate": 9.273032757082062e-05, "loss": 0.0058, "step": 3890 }, { "epoch": 0.7715896725689979, "grad_norm": 0.046838338429783105, "learning_rate": 9.267044910799395e-05, "loss": 0.0058, "step": 3900 }, { "epoch": 0.7735681076268671, "grad_norm": 0.046811141134852845, "learning_rate": 9.261034455132218e-05, "loss": 0.0058, "step": 3910 }, { "epoch": 0.7755465426847363, "grad_norm": 0.03964249052501702, "learning_rate": 9.255001421927558e-05, "loss": 0.0058, "step": 3920 }, { "epoch": 0.7775249777426056, "grad_norm": 0.04405760678095594, "learning_rate": 9.248945843152075e-05, "loss": 0.0057, "step": 3930 }, { "epoch": 0.7795034128004749, "grad_norm": 0.04096050990973934, "learning_rate": 9.242867750891887e-05, "loss": 0.0058, "step": 3940 }, { "epoch": 0.7814818478583441, "grad_norm": 0.036328319424726495, "learning_rate": 9.236767177352403e-05, "loss": 0.0057, "step": 3950 }, { "epoch": 0.7834602829162133, "grad_norm": 0.048579201981942856, "learning_rate": 9.230644154858152e-05, "loss": 0.0057, "step": 3960 }, { "epoch": 0.7854387179740825, "grad_norm": 0.04006922209678304, "learning_rate": 9.224498715852607e-05, "loss": 0.0058, "step": 3970 }, { "epoch": 0.7874171530319517, "grad_norm": 0.04008246199222222, "learning_rate": 9.218330892898024e-05, "loss": 0.0057, "step": 3980 }, { "epoch": 0.7893955880898209, "grad_norm": 0.037160627169548884, "learning_rate": 9.212140718675257e-05, "loss": 0.0056, "step": 3990 }, { "epoch": 0.7913740231476902, "grad_norm": 0.04850173678881574, "learning_rate": 9.205928225983593e-05, "loss": 0.0057, "step": 4000 }, { "epoch": 0.7933524582055594, "grad_norm": 0.04085935917242858, "learning_rate": 9.199693447740577e-05, "loss": 0.0057, "step": 4010 }, { "epoch": 0.7953308932634287, "grad_norm": 0.0487202999012432, "learning_rate": 9.193436416981833e-05, "loss": 0.0057, "step": 4020 }, { "epoch": 0.7973093283212979, "grad_norm": 0.042996874004967385, "learning_rate": 9.187157166860894e-05, "loss": 0.0057, "step": 4030 }, { "epoch": 0.7992877633791671, "grad_norm": 0.06090370759091184, "learning_rate": 9.180855730649025e-05, "loss": 0.0057, "step": 4040 }, { "epoch": 0.8012661984370363, "grad_norm": 0.05005113866030296, "learning_rate": 9.174532141735045e-05, "loss": 0.0058, "step": 4050 }, { "epoch": 0.8032446334949055, "grad_norm": 0.04025804669524248, "learning_rate": 9.168186433625156e-05, "loss": 0.0058, "step": 4060 }, { "epoch": 0.8052230685527747, "grad_norm": 0.04363718260162186, "learning_rate": 9.161818639942753e-05, "loss": 0.0058, "step": 4070 }, { "epoch": 0.807201503610644, "grad_norm": 0.050484493072529946, "learning_rate": 9.155428794428258e-05, "loss": 0.0058, "step": 4080 }, { "epoch": 0.8091799386685132, "grad_norm": 0.05089878391740369, "learning_rate": 9.14901693093894e-05, "loss": 0.0058, "step": 4090 }, { "epoch": 0.8111583737263824, "grad_norm": 0.05010210032660769, "learning_rate": 9.14258308344873e-05, "loss": 0.0057, "step": 4100 }, { "epoch": 0.8131368087842517, "grad_norm": 0.04069468974688542, "learning_rate": 9.136127286048038e-05, "loss": 0.0058, "step": 4110 }, { "epoch": 0.8151152438421209, "grad_norm": 0.04116330872660485, "learning_rate": 9.129649572943587e-05, "loss": 0.0057, "step": 4120 }, { "epoch": 0.8170936788999901, "grad_norm": 0.04534311185009215, "learning_rate": 9.123149978458216e-05, "loss": 0.0057, "step": 4130 }, { "epoch": 0.8190721139578593, "grad_norm": 0.04234503902077635, "learning_rate": 9.116628537030711e-05, "loss": 0.0058, "step": 4140 }, { "epoch": 0.8210505490157286, "grad_norm": 0.03932929250995006, "learning_rate": 9.11008528321561e-05, "loss": 0.0057, "step": 4150 }, { "epoch": 0.8230289840735978, "grad_norm": 0.05483577641000977, "learning_rate": 9.103520251683029e-05, "loss": 0.0057, "step": 4160 }, { "epoch": 0.825007419131467, "grad_norm": 0.039911135968876764, "learning_rate": 9.096933477218475e-05, "loss": 0.0057, "step": 4170 }, { "epoch": 0.8269858541893362, "grad_norm": 0.06434040856491034, "learning_rate": 9.090324994722663e-05, "loss": 0.0056, "step": 4180 }, { "epoch": 0.8289642892472054, "grad_norm": 0.044709409529248005, "learning_rate": 9.083694839211329e-05, "loss": 0.0057, "step": 4190 }, { "epoch": 0.8309427243050747, "grad_norm": 0.06623829057108145, "learning_rate": 9.07704304581505e-05, "loss": 0.0057, "step": 4200 }, { "epoch": 0.8329211593629439, "grad_norm": 0.04900286747901559, "learning_rate": 9.070369649779047e-05, "loss": 0.0056, "step": 4210 }, { "epoch": 0.8348995944208132, "grad_norm": 0.047587755870750065, "learning_rate": 9.063674686463008e-05, "loss": 0.0057, "step": 4220 }, { "epoch": 0.8368780294786824, "grad_norm": 0.03403272538519831, "learning_rate": 9.056958191340901e-05, "loss": 0.0058, "step": 4230 }, { "epoch": 0.8388564645365516, "grad_norm": 0.044790324654289844, "learning_rate": 9.050220200000774e-05, "loss": 0.0057, "step": 4240 }, { "epoch": 0.8408348995944208, "grad_norm": 0.04203584615111322, "learning_rate": 9.043460748144584e-05, "loss": 0.0057, "step": 4250 }, { "epoch": 0.84281333465229, "grad_norm": 0.047785666109283975, "learning_rate": 9.036679871587992e-05, "loss": 0.0057, "step": 4260 }, { "epoch": 0.8447917697101592, "grad_norm": 0.04531771474817144, "learning_rate": 9.029877606260186e-05, "loss": 0.0057, "step": 4270 }, { "epoch": 0.8467702047680284, "grad_norm": 0.04283329056423652, "learning_rate": 9.023053988203679e-05, "loss": 0.0056, "step": 4280 }, { "epoch": 0.8487486398258978, "grad_norm": 0.044946740709006755, "learning_rate": 9.016209053574127e-05, "loss": 0.0056, "step": 4290 }, { "epoch": 0.850727074883767, "grad_norm": 0.04430364627039971, "learning_rate": 9.009342838640134e-05, "loss": 0.0056, "step": 4300 }, { "epoch": 0.8527055099416362, "grad_norm": 0.04548141341995741, "learning_rate": 9.002455379783057e-05, "loss": 0.0056, "step": 4310 }, { "epoch": 0.8546839449995054, "grad_norm": 0.06068254235119247, "learning_rate": 8.99554671349682e-05, "loss": 0.0058, "step": 4320 }, { "epoch": 0.8566623800573746, "grad_norm": 0.050410634600550126, "learning_rate": 8.988616876387717e-05, "loss": 0.0057, "step": 4330 }, { "epoch": 0.8586408151152438, "grad_norm": 0.03642438536583747, "learning_rate": 8.98166590517421e-05, "loss": 0.0057, "step": 4340 }, { "epoch": 0.860619250173113, "grad_norm": 0.042453351674432066, "learning_rate": 8.974693836686754e-05, "loss": 0.0057, "step": 4350 }, { "epoch": 0.8625976852309823, "grad_norm": 0.059892261452290846, "learning_rate": 8.967700707867584e-05, "loss": 0.0057, "step": 4360 }, { "epoch": 0.8645761202888516, "grad_norm": 0.04510576050539382, "learning_rate": 8.960686555770523e-05, "loss": 0.0056, "step": 4370 }, { "epoch": 0.8665545553467208, "grad_norm": 0.05420822964894672, "learning_rate": 8.953651417560791e-05, "loss": 0.0056, "step": 4380 }, { "epoch": 0.86853299040459, "grad_norm": 0.041286990742185684, "learning_rate": 8.946595330514806e-05, "loss": 0.0056, "step": 4390 }, { "epoch": 0.8705114254624592, "grad_norm": 0.044651522507349696, "learning_rate": 8.939518332019985e-05, "loss": 0.0056, "step": 4400 }, { "epoch": 0.8724898605203284, "grad_norm": 0.024344357858870437, "learning_rate": 8.932420459574544e-05, "loss": 0.0056, "step": 4410 }, { "epoch": 0.8744682955781976, "grad_norm": 0.04834073635907548, "learning_rate": 8.925301750787306e-05, "loss": 0.0056, "step": 4420 }, { "epoch": 0.8764467306360668, "grad_norm": 0.0681315254129069, "learning_rate": 8.918162243377493e-05, "loss": 0.0058, "step": 4430 }, { "epoch": 0.8784251656939361, "grad_norm": 0.042436322460547654, "learning_rate": 8.911001975174536e-05, "loss": 0.0055, "step": 4440 }, { "epoch": 0.8804036007518053, "grad_norm": 0.03577857281883983, "learning_rate": 8.903820984117864e-05, "loss": 0.0056, "step": 4450 }, { "epoch": 0.8823820358096746, "grad_norm": 0.05210607835647707, "learning_rate": 8.896619308256711e-05, "loss": 0.0056, "step": 4460 }, { "epoch": 0.8843604708675438, "grad_norm": 0.03398692335903851, "learning_rate": 8.88939698574991e-05, "loss": 0.0054, "step": 4470 }, { "epoch": 0.886338905925413, "grad_norm": 0.035988957121509055, "learning_rate": 8.882154054865695e-05, "loss": 0.0055, "step": 4480 }, { "epoch": 0.8883173409832822, "grad_norm": 0.040709661403919795, "learning_rate": 8.874890553981492e-05, "loss": 0.0055, "step": 4490 }, { "epoch": 0.8902957760411514, "grad_norm": 0.032549783994869345, "learning_rate": 8.867606521583725e-05, "loss": 0.0056, "step": 4500 }, { "epoch": 0.8922742110990207, "grad_norm": 0.03959821311155834, "learning_rate": 8.860301996267602e-05, "loss": 0.0055, "step": 4510 }, { "epoch": 0.8942526461568899, "grad_norm": 0.03399382401331832, "learning_rate": 8.852977016736912e-05, "loss": 0.0056, "step": 4520 }, { "epoch": 0.8962310812147591, "grad_norm": 0.03527564691513799, "learning_rate": 8.84563162180383e-05, "loss": 0.0056, "step": 4530 }, { "epoch": 0.8982095162726283, "grad_norm": 0.046909675502707464, "learning_rate": 8.838265850388698e-05, "loss": 0.0055, "step": 4540 }, { "epoch": 0.9001879513304976, "grad_norm": 0.04112314221218313, "learning_rate": 8.830879741519832e-05, "loss": 0.0055, "step": 4550 }, { "epoch": 0.9021663863883668, "grad_norm": 0.029301685068256162, "learning_rate": 8.823473334333298e-05, "loss": 0.0054, "step": 4560 }, { "epoch": 0.904144821446236, "grad_norm": 0.04035117861654918, "learning_rate": 8.816046668072726e-05, "loss": 0.0055, "step": 4570 }, { "epoch": 0.9061232565041053, "grad_norm": 0.03183133712799434, "learning_rate": 8.808599782089081e-05, "loss": 0.0054, "step": 4580 }, { "epoch": 0.9081016915619745, "grad_norm": 0.03963598813293878, "learning_rate": 8.80113271584047e-05, "loss": 0.0055, "step": 4590 }, { "epoch": 0.9100801266198437, "grad_norm": 0.04460141663017717, "learning_rate": 8.793645508891926e-05, "loss": 0.0056, "step": 4600 }, { "epoch": 0.9120585616777129, "grad_norm": 0.04253749261648797, "learning_rate": 8.786138200915199e-05, "loss": 0.0056, "step": 4610 }, { "epoch": 0.9140369967355821, "grad_norm": 0.03745074598611024, "learning_rate": 8.778610831688548e-05, "loss": 0.0055, "step": 4620 }, { "epoch": 0.9160154317934514, "grad_norm": 0.04054460416648101, "learning_rate": 8.771063441096526e-05, "loss": 0.0055, "step": 4630 }, { "epoch": 0.9179938668513206, "grad_norm": 0.04706894664328689, "learning_rate": 8.763496069129773e-05, "loss": 0.0056, "step": 4640 }, { "epoch": 0.9199723019091899, "grad_norm": 0.045942457363925146, "learning_rate": 8.755908755884804e-05, "loss": 0.0055, "step": 4650 }, { "epoch": 0.9219507369670591, "grad_norm": 0.06270144515647959, "learning_rate": 8.74830154156379e-05, "loss": 0.0056, "step": 4660 }, { "epoch": 0.9239291720249283, "grad_norm": 0.04673214308487745, "learning_rate": 8.740674466474356e-05, "loss": 0.0056, "step": 4670 }, { "epoch": 0.9259076070827975, "grad_norm": 0.038901485608572116, "learning_rate": 8.733027571029358e-05, "loss": 0.0057, "step": 4680 }, { "epoch": 0.9278860421406667, "grad_norm": 0.040496171014254186, "learning_rate": 8.725360895746671e-05, "loss": 0.0055, "step": 4690 }, { "epoch": 0.9298644771985359, "grad_norm": 0.0374153756385262, "learning_rate": 8.717674481248978e-05, "loss": 0.0056, "step": 4700 }, { "epoch": 0.9318429122564051, "grad_norm": 0.025535819289588646, "learning_rate": 8.709968368263553e-05, "loss": 0.0055, "step": 4710 }, { "epoch": 0.9338213473142745, "grad_norm": 0.03879829254204007, "learning_rate": 8.70224259762204e-05, "loss": 0.0054, "step": 4720 }, { "epoch": 0.9357997823721437, "grad_norm": 0.05800476178411971, "learning_rate": 8.694497210260246e-05, "loss": 0.0055, "step": 4730 }, { "epoch": 0.9377782174300129, "grad_norm": 0.03803197143653136, "learning_rate": 8.686732247217917e-05, "loss": 0.0056, "step": 4740 }, { "epoch": 0.9397566524878821, "grad_norm": 0.03883433226580444, "learning_rate": 8.678947749638526e-05, "loss": 0.0055, "step": 4750 }, { "epoch": 0.9417350875457513, "grad_norm": 0.049511735420544906, "learning_rate": 8.671143758769044e-05, "loss": 0.0056, "step": 4760 }, { "epoch": 0.9437135226036205, "grad_norm": 0.04330654439138215, "learning_rate": 8.663320315959736e-05, "loss": 0.0055, "step": 4770 }, { "epoch": 0.9456919576614897, "grad_norm": 0.03938926247394834, "learning_rate": 8.655477462663937e-05, "loss": 0.0056, "step": 4780 }, { "epoch": 0.9476703927193589, "grad_norm": 0.043100978246848436, "learning_rate": 8.647615240437821e-05, "loss": 0.0055, "step": 4790 }, { "epoch": 0.9496488277772283, "grad_norm": 0.046229415784614536, "learning_rate": 8.6397336909402e-05, "loss": 0.0055, "step": 4800 }, { "epoch": 0.9516272628350975, "grad_norm": 0.04348275355338679, "learning_rate": 8.631832855932289e-05, "loss": 0.0055, "step": 4810 }, { "epoch": 0.9536056978929667, "grad_norm": 0.04714054671540608, "learning_rate": 8.62391277727749e-05, "loss": 0.0055, "step": 4820 }, { "epoch": 0.9555841329508359, "grad_norm": 0.040059125010807065, "learning_rate": 8.615973496941169e-05, "loss": 0.0057, "step": 4830 }, { "epoch": 0.9575625680087051, "grad_norm": 0.08003097755568261, "learning_rate": 8.608015056990437e-05, "loss": 0.0056, "step": 4840 }, { "epoch": 0.9595410030665743, "grad_norm": 0.05750105466944878, "learning_rate": 8.600037499593918e-05, "loss": 0.0055, "step": 4850 }, { "epoch": 0.9615194381244435, "grad_norm": 0.0351261794361651, "learning_rate": 8.592040867021538e-05, "loss": 0.0055, "step": 4860 }, { "epoch": 0.9634978731823128, "grad_norm": 0.028601432945424794, "learning_rate": 8.584025201644292e-05, "loss": 0.0054, "step": 4870 }, { "epoch": 0.965476308240182, "grad_norm": 0.046094943573683336, "learning_rate": 8.575990545934023e-05, "loss": 0.0055, "step": 4880 }, { "epoch": 0.9674547432980513, "grad_norm": 0.04651640942126606, "learning_rate": 8.567936942463195e-05, "loss": 0.0055, "step": 4890 }, { "epoch": 0.9694331783559205, "grad_norm": 0.035963030931479324, "learning_rate": 8.559864433904671e-05, "loss": 0.0055, "step": 4900 }, { "epoch": 0.9714116134137897, "grad_norm": 0.04812055206677903, "learning_rate": 8.551773063031484e-05, "loss": 0.0055, "step": 4910 }, { "epoch": 0.9733900484716589, "grad_norm": 0.038444786548038186, "learning_rate": 8.543662872716607e-05, "loss": 0.0055, "step": 4920 }, { "epoch": 0.9753684835295281, "grad_norm": 0.04120869481487097, "learning_rate": 8.535533905932738e-05, "loss": 0.0054, "step": 4930 }, { "epoch": 0.9773469185873974, "grad_norm": 0.04349911950600627, "learning_rate": 8.527386205752056e-05, "loss": 0.0055, "step": 4940 }, { "epoch": 0.9793253536452666, "grad_norm": 0.0353721620579985, "learning_rate": 8.519219815346004e-05, "loss": 0.0055, "step": 4950 }, { "epoch": 0.9813037887031358, "grad_norm": 0.04813339576005731, "learning_rate": 8.511034777985058e-05, "loss": 0.0055, "step": 4960 }, { "epoch": 0.983282223761005, "grad_norm": 0.03570129472474123, "learning_rate": 8.502831137038496e-05, "loss": 0.0054, "step": 4970 }, { "epoch": 0.9852606588188743, "grad_norm": 0.033113097357052625, "learning_rate": 8.494608935974168e-05, "loss": 0.0054, "step": 4980 }, { "epoch": 0.9872390938767435, "grad_norm": 0.05624145428937703, "learning_rate": 8.486368218358268e-05, "loss": 0.0055, "step": 4990 }, { "epoch": 0.9892175289346127, "grad_norm": 0.04104291146605793, "learning_rate": 8.478109027855103e-05, "loss": 0.0054, "step": 5000 }, { "epoch": 0.991195963992482, "grad_norm": 0.04482533714278806, "learning_rate": 8.469831408226858e-05, "loss": 0.0055, "step": 5010 }, { "epoch": 0.9931743990503512, "grad_norm": 0.027511907821805445, "learning_rate": 8.46153540333337e-05, "loss": 0.0055, "step": 5020 }, { "epoch": 0.9951528341082204, "grad_norm": 0.041338440003761766, "learning_rate": 8.453221057131886e-05, "loss": 0.0054, "step": 5030 }, { "epoch": 0.9971312691660896, "grad_norm": 0.04529153009675916, "learning_rate": 8.444888413676844e-05, "loss": 0.0054, "step": 5040 }, { "epoch": 0.9991097042239588, "grad_norm": 0.04355117524619956, "learning_rate": 8.436537517119624e-05, "loss": 0.0054, "step": 5050 }, { "epoch": 1.0009892175289345, "grad_norm": 0.03055515304537575, "learning_rate": 8.428168411708332e-05, "loss": 0.0054, "step": 5060 }, { "epoch": 1.0029676525868039, "grad_norm": 0.047560373488857215, "learning_rate": 8.419781141787548e-05, "loss": 0.0054, "step": 5070 }, { "epoch": 1.004946087644673, "grad_norm": 0.02683507205249473, "learning_rate": 8.411375751798099e-05, "loss": 0.0054, "step": 5080 }, { "epoch": 1.0069245227025423, "grad_norm": 0.029124427414850166, "learning_rate": 8.402952286276824e-05, "loss": 0.0054, "step": 5090 }, { "epoch": 1.0089029577604116, "grad_norm": 0.033621264791519626, "learning_rate": 8.394510789856339e-05, "loss": 0.0053, "step": 5100 }, { "epoch": 1.0108813928182807, "grad_norm": 0.03186973120057544, "learning_rate": 8.386051307264797e-05, "loss": 0.0054, "step": 5110 }, { "epoch": 1.01285982787615, "grad_norm": 0.07109531783852259, "learning_rate": 8.377573883325653e-05, "loss": 0.0054, "step": 5120 }, { "epoch": 1.0148382629340191, "grad_norm": 0.09703735539944616, "learning_rate": 8.369078562957425e-05, "loss": 0.0057, "step": 5130 }, { "epoch": 1.0168166979918885, "grad_norm": 0.04542084962318844, "learning_rate": 8.360565391173459e-05, "loss": 0.0061, "step": 5140 }, { "epoch": 1.0187951330497576, "grad_norm": 0.058115327887701636, "learning_rate": 8.352034413081686e-05, "loss": 0.0056, "step": 5150 }, { "epoch": 1.0207735681076269, "grad_norm": 0.07643286599383742, "learning_rate": 8.343485673884388e-05, "loss": 0.0057, "step": 5160 }, { "epoch": 1.0227520031654962, "grad_norm": 0.0730225610278571, "learning_rate": 8.334919218877957e-05, "loss": 0.0056, "step": 5170 }, { "epoch": 1.0247304382233653, "grad_norm": 0.04576258232677307, "learning_rate": 8.326335093452651e-05, "loss": 0.0055, "step": 5180 }, { "epoch": 1.0267088732812346, "grad_norm": 0.0325850395708169, "learning_rate": 8.317733343092357e-05, "loss": 0.0054, "step": 5190 }, { "epoch": 1.0286873083391037, "grad_norm": 0.03183559014995542, "learning_rate": 8.309114013374353e-05, "loss": 0.0053, "step": 5200 }, { "epoch": 1.030665743396973, "grad_norm": 0.04957936795082493, "learning_rate": 8.300477149969057e-05, "loss": 0.0053, "step": 5210 }, { "epoch": 1.0326441784548421, "grad_norm": 0.044147221696157916, "learning_rate": 8.291822798639797e-05, "loss": 0.0054, "step": 5220 }, { "epoch": 1.0346226135127115, "grad_norm": 0.022582884820835147, "learning_rate": 8.28315100524256e-05, "loss": 0.0054, "step": 5230 }, { "epoch": 1.0366010485705808, "grad_norm": 0.023883108532179793, "learning_rate": 8.274461815725751e-05, "loss": 0.0054, "step": 5240 }, { "epoch": 1.0385794836284499, "grad_norm": 0.03679863419557376, "learning_rate": 8.265755276129953e-05, "loss": 0.0054, "step": 5250 }, { "epoch": 1.0405579186863192, "grad_norm": 0.0454671679634965, "learning_rate": 8.257031432587676e-05, "loss": 0.0055, "step": 5260 }, { "epoch": 1.0425363537441883, "grad_norm": 0.04142698492573628, "learning_rate": 8.24829033132312e-05, "loss": 0.0054, "step": 5270 }, { "epoch": 1.0445147888020576, "grad_norm": 0.04033757143532992, "learning_rate": 8.239532018651924e-05, "loss": 0.0053, "step": 5280 }, { "epoch": 1.0464932238599267, "grad_norm": 0.04027827090668591, "learning_rate": 8.230756540980929e-05, "loss": 0.0053, "step": 5290 }, { "epoch": 1.048471658917796, "grad_norm": 0.02637364908905128, "learning_rate": 8.22196394480792e-05, "loss": 0.0054, "step": 5300 }, { "epoch": 1.0504500939756654, "grad_norm": 0.056833182371949155, "learning_rate": 8.213154276721388e-05, "loss": 0.0054, "step": 5310 }, { "epoch": 1.0524285290335345, "grad_norm": 0.031217740637365373, "learning_rate": 8.204327583400287e-05, "loss": 0.0053, "step": 5320 }, { "epoch": 1.0544069640914038, "grad_norm": 0.02844798702968925, "learning_rate": 8.195483911613772e-05, "loss": 0.0054, "step": 5330 }, { "epoch": 1.0563853991492729, "grad_norm": 0.038643639613203266, "learning_rate": 8.186623308220967e-05, "loss": 0.0054, "step": 5340 }, { "epoch": 1.0583638342071422, "grad_norm": 0.042432315795242086, "learning_rate": 8.177745820170705e-05, "loss": 0.0053, "step": 5350 }, { "epoch": 1.0603422692650113, "grad_norm": 0.03535263178736307, "learning_rate": 8.168851494501287e-05, "loss": 0.0054, "step": 5360 }, { "epoch": 1.0623207043228806, "grad_norm": 0.04412203322412542, "learning_rate": 8.159940378340232e-05, "loss": 0.0054, "step": 5370 }, { "epoch": 1.0642991393807497, "grad_norm": 0.04245798106108046, "learning_rate": 8.15101251890402e-05, "loss": 0.0054, "step": 5380 }, { "epoch": 1.066277574438619, "grad_norm": 0.03199350913269745, "learning_rate": 8.142067963497851e-05, "loss": 0.0053, "step": 5390 }, { "epoch": 1.0682560094964884, "grad_norm": 0.04794158747470023, "learning_rate": 8.133106759515388e-05, "loss": 0.0053, "step": 5400 }, { "epoch": 1.0702344445543575, "grad_norm": 0.030961777586157032, "learning_rate": 8.124128954438508e-05, "loss": 0.0054, "step": 5410 }, { "epoch": 1.0722128796122268, "grad_norm": 0.03152795250709625, "learning_rate": 8.115134595837055e-05, "loss": 0.0053, "step": 5420 }, { "epoch": 1.0741913146700959, "grad_norm": 0.025506156241943186, "learning_rate": 8.106123731368578e-05, "loss": 0.0053, "step": 5430 }, { "epoch": 1.0761697497279652, "grad_norm": 0.03219012376436614, "learning_rate": 8.097096408778086e-05, "loss": 0.0053, "step": 5440 }, { "epoch": 1.0781481847858343, "grad_norm": 0.028477732978055153, "learning_rate": 8.088052675897795e-05, "loss": 0.0053, "step": 5450 }, { "epoch": 1.0801266198437036, "grad_norm": 0.03399240671941344, "learning_rate": 8.078992580646871e-05, "loss": 0.0053, "step": 5460 }, { "epoch": 1.082105054901573, "grad_norm": 0.03868025099458124, "learning_rate": 8.06991617103118e-05, "loss": 0.0053, "step": 5470 }, { "epoch": 1.084083489959442, "grad_norm": 0.04088910382579558, "learning_rate": 8.060823495143024e-05, "loss": 0.0053, "step": 5480 }, { "epoch": 1.0860619250173114, "grad_norm": 0.02844467650618694, "learning_rate": 8.051714601160903e-05, "loss": 0.0052, "step": 5490 }, { "epoch": 1.0880403600751805, "grad_norm": 0.02926836345161859, "learning_rate": 8.042589537349243e-05, "loss": 0.0052, "step": 5500 }, { "epoch": 1.0900187951330498, "grad_norm": 0.030229836986547978, "learning_rate": 8.033448352058154e-05, "loss": 0.0052, "step": 5510 }, { "epoch": 1.0919972301909189, "grad_norm": 0.032951214929387296, "learning_rate": 8.024291093723162e-05, "loss": 0.0053, "step": 5520 }, { "epoch": 1.0939756652487882, "grad_norm": 0.03508718490481177, "learning_rate": 8.01511781086496e-05, "loss": 0.0053, "step": 5530 }, { "epoch": 1.0959541003066575, "grad_norm": 0.022637612720568006, "learning_rate": 8.005928552089148e-05, "loss": 0.0052, "step": 5540 }, { "epoch": 1.0979325353645266, "grad_norm": 0.0395739148284348, "learning_rate": 7.996723366085977e-05, "loss": 0.0053, "step": 5550 }, { "epoch": 1.099910970422396, "grad_norm": 0.043102899116423427, "learning_rate": 7.98750230163009e-05, "loss": 0.0053, "step": 5560 }, { "epoch": 1.101889405480265, "grad_norm": 0.028811703589080282, "learning_rate": 7.978265407580261e-05, "loss": 0.0054, "step": 5570 }, { "epoch": 1.1038678405381344, "grad_norm": 0.04603280202575731, "learning_rate": 7.969012732879144e-05, "loss": 0.0053, "step": 5580 }, { "epoch": 1.1058462755960035, "grad_norm": 0.029764986729849122, "learning_rate": 7.959744326553001e-05, "loss": 0.0053, "step": 5590 }, { "epoch": 1.1078247106538728, "grad_norm": 0.025666958289447685, "learning_rate": 7.950460237711458e-05, "loss": 0.0053, "step": 5600 }, { "epoch": 1.109803145711742, "grad_norm": 0.030968526076831825, "learning_rate": 7.94116051554723e-05, "loss": 0.0053, "step": 5610 }, { "epoch": 1.1117815807696112, "grad_norm": 0.030380259307121488, "learning_rate": 7.931845209335871e-05, "loss": 0.0053, "step": 5620 }, { "epoch": 1.1137600158274805, "grad_norm": 0.030934524480240706, "learning_rate": 7.922514368435505e-05, "loss": 0.0052, "step": 5630 }, { "epoch": 1.1157384508853496, "grad_norm": 0.03684454858170427, "learning_rate": 7.91316804228657e-05, "loss": 0.0053, "step": 5640 }, { "epoch": 1.117716885943219, "grad_norm": 0.0306288322163976, "learning_rate": 7.903806280411557e-05, "loss": 0.0052, "step": 5650 }, { "epoch": 1.119695321001088, "grad_norm": 0.0444930109980131, "learning_rate": 7.89442913241474e-05, "loss": 0.0053, "step": 5660 }, { "epoch": 1.1216737560589574, "grad_norm": 0.049678037944609614, "learning_rate": 7.885036647981921e-05, "loss": 0.0054, "step": 5670 }, { "epoch": 1.1236521911168267, "grad_norm": 0.02831510844151286, "learning_rate": 7.875628876880161e-05, "loss": 0.0053, "step": 5680 }, { "epoch": 1.1256306261746958, "grad_norm": 0.0289977730202135, "learning_rate": 7.866205868957524e-05, "loss": 0.0053, "step": 5690 }, { "epoch": 1.127609061232565, "grad_norm": 0.0317147439916211, "learning_rate": 7.8567676741428e-05, "loss": 0.0054, "step": 5700 }, { "epoch": 1.1295874962904342, "grad_norm": 0.029466927048270367, "learning_rate": 7.847314342445258e-05, "loss": 0.0053, "step": 5710 }, { "epoch": 1.1315659313483035, "grad_norm": 0.03250558063084724, "learning_rate": 7.837845923954364e-05, "loss": 0.0053, "step": 5720 }, { "epoch": 1.1335443664061726, "grad_norm": 0.04132101024688395, "learning_rate": 7.828362468839525e-05, "loss": 0.0053, "step": 5730 }, { "epoch": 1.135522801464042, "grad_norm": 0.038393775413381924, "learning_rate": 7.818864027349822e-05, "loss": 0.0053, "step": 5740 }, { "epoch": 1.1375012365219113, "grad_norm": 0.03899980893950007, "learning_rate": 7.809350649813743e-05, "loss": 0.0053, "step": 5750 }, { "epoch": 1.1394796715797804, "grad_norm": 0.032472191555044536, "learning_rate": 7.799822386638915e-05, "loss": 0.0053, "step": 5760 }, { "epoch": 1.1414581066376497, "grad_norm": 0.029404384356372573, "learning_rate": 7.790279288311837e-05, "loss": 0.0053, "step": 5770 }, { "epoch": 1.1434365416955188, "grad_norm": 0.026865065518501798, "learning_rate": 7.78072140539762e-05, "loss": 0.0052, "step": 5780 }, { "epoch": 1.145414976753388, "grad_norm": 0.03203832588668241, "learning_rate": 7.771148788539704e-05, "loss": 0.0053, "step": 5790 }, { "epoch": 1.1473934118112572, "grad_norm": 0.02716708690762755, "learning_rate": 7.761561488459603e-05, "loss": 0.0053, "step": 5800 }, { "epoch": 1.1493718468691265, "grad_norm": 0.032831740791798764, "learning_rate": 7.751959555956632e-05, "loss": 0.0052, "step": 5810 }, { "epoch": 1.1513502819269958, "grad_norm": 0.05107091688212574, "learning_rate": 7.742343041907634e-05, "loss": 0.0053, "step": 5820 }, { "epoch": 1.153328716984865, "grad_norm": 0.037252820336677664, "learning_rate": 7.732711997266719e-05, "loss": 0.0053, "step": 5830 }, { "epoch": 1.1553071520427343, "grad_norm": 0.027851822398685163, "learning_rate": 7.723066473064986e-05, "loss": 0.0053, "step": 5840 }, { "epoch": 1.1572855871006034, "grad_norm": 0.04326527014549261, "learning_rate": 7.713406520410252e-05, "loss": 0.0053, "step": 5850 }, { "epoch": 1.1592640221584727, "grad_norm": 0.07278979067831416, "learning_rate": 7.703732190486792e-05, "loss": 0.0054, "step": 5860 }, { "epoch": 1.1612424572163418, "grad_norm": 0.03395572859323836, "learning_rate": 7.694043534555055e-05, "loss": 0.0055, "step": 5870 }, { "epoch": 1.163220892274211, "grad_norm": 0.050318472905328156, "learning_rate": 7.684340603951399e-05, "loss": 0.0054, "step": 5880 }, { "epoch": 1.1651993273320804, "grad_norm": 0.06002742001305542, "learning_rate": 7.674623450087822e-05, "loss": 0.0056, "step": 5890 }, { "epoch": 1.1671777623899495, "grad_norm": 0.06795650064920666, "learning_rate": 7.664892124451676e-05, "loss": 0.0055, "step": 5900 }, { "epoch": 1.1691561974478188, "grad_norm": 0.021558586037374296, "learning_rate": 7.655146678605414e-05, "loss": 0.0054, "step": 5910 }, { "epoch": 1.171134632505688, "grad_norm": 0.0338823721497897, "learning_rate": 7.645387164186298e-05, "loss": 0.0054, "step": 5920 }, { "epoch": 1.1731130675635573, "grad_norm": 0.02570157372632155, "learning_rate": 7.635613632906141e-05, "loss": 0.0053, "step": 5930 }, { "epoch": 1.1750915026214264, "grad_norm": 0.01772587183015968, "learning_rate": 7.625826136551022e-05, "loss": 0.0052, "step": 5940 }, { "epoch": 1.1770699376792957, "grad_norm": 0.030184690153560394, "learning_rate": 7.616024726981015e-05, "loss": 0.0053, "step": 5950 }, { "epoch": 1.179048372737165, "grad_norm": 0.040273409138632796, "learning_rate": 7.606209456129914e-05, "loss": 0.0052, "step": 5960 }, { "epoch": 1.181026807795034, "grad_norm": 0.042593741522192294, "learning_rate": 7.596380376004965e-05, "loss": 0.0052, "step": 5970 }, { "epoch": 1.1830052428529034, "grad_norm": 0.01992982832550707, "learning_rate": 7.586537538686574e-05, "loss": 0.0053, "step": 5980 }, { "epoch": 1.1849836779107725, "grad_norm": 0.03746835123236025, "learning_rate": 7.576680996328045e-05, "loss": 0.0054, "step": 5990 }, { "epoch": 1.1869621129686418, "grad_norm": 0.04360095909075448, "learning_rate": 7.566810801155304e-05, "loss": 0.0053, "step": 6000 }, { "epoch": 1.188940548026511, "grad_norm": 0.038421977235073596, "learning_rate": 7.556927005466612e-05, "loss": 0.0054, "step": 6010 }, { "epoch": 1.1909189830843803, "grad_norm": 0.02343201491962226, "learning_rate": 7.547029661632298e-05, "loss": 0.0053, "step": 6020 }, { "epoch": 1.1928974181422496, "grad_norm": 0.029392605969832238, "learning_rate": 7.537118822094474e-05, "loss": 0.0053, "step": 6030 }, { "epoch": 1.1948758532001187, "grad_norm": 0.036651429554557045, "learning_rate": 7.52719453936676e-05, "loss": 0.0053, "step": 6040 }, { "epoch": 1.196854288257988, "grad_norm": 0.03350816279577848, "learning_rate": 7.517256866034011e-05, "loss": 0.0053, "step": 6050 }, { "epoch": 1.198832723315857, "grad_norm": 0.030987790995458092, "learning_rate": 7.507305854752027e-05, "loss": 0.0053, "step": 6060 }, { "epoch": 1.2008111583737264, "grad_norm": 0.0346706313861039, "learning_rate": 7.497341558247286e-05, "loss": 0.0054, "step": 6070 }, { "epoch": 1.2027895934315955, "grad_norm": 0.04150858873559331, "learning_rate": 7.487364029316655e-05, "loss": 0.0054, "step": 6080 }, { "epoch": 1.2047680284894648, "grad_norm": 0.039861874099186935, "learning_rate": 7.47737332082712e-05, "loss": 0.0054, "step": 6090 }, { "epoch": 1.2067464635473342, "grad_norm": 0.029847263959674315, "learning_rate": 7.467369485715492e-05, "loss": 0.0053, "step": 6100 }, { "epoch": 1.2087248986052033, "grad_norm": 0.03251532698303107, "learning_rate": 7.457352576988144e-05, "loss": 0.0053, "step": 6110 }, { "epoch": 1.2107033336630726, "grad_norm": 0.03324515956888663, "learning_rate": 7.447322647720714e-05, "loss": 0.0053, "step": 6120 }, { "epoch": 1.2126817687209417, "grad_norm": 0.02582021604427929, "learning_rate": 7.437279751057833e-05, "loss": 0.0053, "step": 6130 }, { "epoch": 1.214660203778811, "grad_norm": 0.026584943642491917, "learning_rate": 7.42722394021284e-05, "loss": 0.0052, "step": 6140 }, { "epoch": 1.21663863883668, "grad_norm": 0.041612106128670515, "learning_rate": 7.417155268467505e-05, "loss": 0.0052, "step": 6150 }, { "epoch": 1.2186170738945494, "grad_norm": 0.0252663917734944, "learning_rate": 7.407073789171737e-05, "loss": 0.0052, "step": 6160 }, { "epoch": 1.2205955089524188, "grad_norm": 0.03864471301279432, "learning_rate": 7.396979555743315e-05, "loss": 0.0054, "step": 6170 }, { "epoch": 1.2225739440102878, "grad_norm": 0.04324779207434955, "learning_rate": 7.386872621667588e-05, "loss": 0.0054, "step": 6180 }, { "epoch": 1.2245523790681572, "grad_norm": 0.04161587344763828, "learning_rate": 7.376753040497206e-05, "loss": 0.0053, "step": 6190 }, { "epoch": 1.2265308141260263, "grad_norm": 0.034050862201234336, "learning_rate": 7.366620865851832e-05, "loss": 0.0053, "step": 6200 }, { "epoch": 1.2285092491838956, "grad_norm": 0.03846658141676986, "learning_rate": 7.356476151417851e-05, "loss": 0.0053, "step": 6210 }, { "epoch": 1.2304876842417647, "grad_norm": 0.03714661098610616, "learning_rate": 7.3463189509481e-05, "loss": 0.0053, "step": 6220 }, { "epoch": 1.232466119299634, "grad_norm": 0.03364793846119566, "learning_rate": 7.336149318261565e-05, "loss": 0.0053, "step": 6230 }, { "epoch": 1.2344445543575033, "grad_norm": 0.02741367842510466, "learning_rate": 7.325967307243116e-05, "loss": 0.0053, "step": 6240 }, { "epoch": 1.2364229894153724, "grad_norm": 0.02386153598149066, "learning_rate": 7.3157729718432e-05, "loss": 0.0052, "step": 6250 }, { "epoch": 1.2384014244732418, "grad_norm": 0.03125022763748143, "learning_rate": 7.305566366077573e-05, "loss": 0.0052, "step": 6260 }, { "epoch": 1.2403798595311109, "grad_norm": 0.022068034492936674, "learning_rate": 7.295347544027006e-05, "loss": 0.0053, "step": 6270 }, { "epoch": 1.2423582945889802, "grad_norm": 10.603476932719744, "learning_rate": 7.285116559836996e-05, "loss": 0.0707, "step": 6280 }, { "epoch": 1.2443367296468493, "grad_norm": 2.9530759844921963, "learning_rate": 7.274873467717487e-05, "loss": 0.1911, "step": 6290 }, { "epoch": 1.2463151647047186, "grad_norm": 0.3065754040033182, "learning_rate": 7.264618321942575e-05, "loss": 0.0249, "step": 6300 }, { "epoch": 1.248293599762588, "grad_norm": 0.0874582737799162, "learning_rate": 7.254351176850222e-05, "loss": 0.0097, "step": 6310 }, { "epoch": 1.250272034820457, "grad_norm": 0.051066157901512234, "learning_rate": 7.244072086841977e-05, "loss": 0.0068, "step": 6320 }, { "epoch": 1.2522504698783261, "grad_norm": 0.05372403619062444, "learning_rate": 7.23378110638267e-05, "loss": 0.0061, "step": 6330 }, { "epoch": 1.2542289049361954, "grad_norm": 0.03587803967553707, "learning_rate": 7.223478290000142e-05, "loss": 0.0057, "step": 6340 }, { "epoch": 1.2562073399940648, "grad_norm": 0.038595620222516454, "learning_rate": 7.213163692284943e-05, "loss": 0.0055, "step": 6350 }, { "epoch": 1.2581857750519339, "grad_norm": 0.028562026034737585, "learning_rate": 7.202837367890054e-05, "loss": 0.0055, "step": 6360 }, { "epoch": 1.2601642101098032, "grad_norm": 0.03297829637949374, "learning_rate": 7.192499371530581e-05, "loss": 0.0054, "step": 6370 }, { "epoch": 1.2621426451676725, "grad_norm": 0.03326825451031355, "learning_rate": 7.182149757983484e-05, "loss": 0.0054, "step": 6380 }, { "epoch": 1.2641210802255416, "grad_norm": 0.019147193320279744, "learning_rate": 7.171788582087277e-05, "loss": 0.0054, "step": 6390 }, { "epoch": 1.2660995152834107, "grad_norm": 0.021929696194293285, "learning_rate": 7.16141589874173e-05, "loss": 0.0053, "step": 6400 }, { "epoch": 1.26807795034128, "grad_norm": 0.028002974111680408, "learning_rate": 7.151031762907597e-05, "loss": 0.0053, "step": 6410 }, { "epoch": 1.2700563853991493, "grad_norm": 0.16605004309563082, "learning_rate": 7.140636229606308e-05, "loss": 0.0054, "step": 6420 }, { "epoch": 1.2720348204570184, "grad_norm": 0.07475610341082747, "learning_rate": 7.130229353919684e-05, "loss": 0.0056, "step": 6430 }, { "epoch": 1.2740132555148878, "grad_norm": 0.04380860812935727, "learning_rate": 7.119811190989649e-05, "loss": 0.0054, "step": 6440 }, { "epoch": 1.275991690572757, "grad_norm": 0.0325190351684636, "learning_rate": 7.10938179601793e-05, "loss": 0.0054, "step": 6450 }, { "epoch": 1.2779701256306262, "grad_norm": 0.021781537026848753, "learning_rate": 7.098941224265769e-05, "loss": 0.0053, "step": 6460 }, { "epoch": 1.2799485606884953, "grad_norm": 0.0270850066007043, "learning_rate": 7.088489531053627e-05, "loss": 0.0053, "step": 6470 }, { "epoch": 1.2819269957463646, "grad_norm": 0.023427530657391633, "learning_rate": 7.0780267717609e-05, "loss": 0.0052, "step": 6480 }, { "epoch": 1.283905430804234, "grad_norm": 0.026373573304602972, "learning_rate": 7.067553001825609e-05, "loss": 0.0053, "step": 6490 }, { "epoch": 1.285883865862103, "grad_norm": 0.02989133589034671, "learning_rate": 7.057068276744123e-05, "loss": 0.0053, "step": 6500 }, { "epoch": 1.2878623009199723, "grad_norm": 0.022264441604619878, "learning_rate": 7.04657265207086e-05, "loss": 0.0052, "step": 6510 }, { "epoch": 1.2898407359778417, "grad_norm": 0.029856404765650494, "learning_rate": 7.036066183417981e-05, "loss": 0.0052, "step": 6520 }, { "epoch": 1.2918191710357108, "grad_norm": 0.01839794068995146, "learning_rate": 7.025548926455117e-05, "loss": 0.0051, "step": 6530 }, { "epoch": 1.2937976060935799, "grad_norm": 0.028680713297060564, "learning_rate": 7.015020936909052e-05, "loss": 0.0052, "step": 6540 }, { "epoch": 1.2957760411514492, "grad_norm": 0.024895542675878957, "learning_rate": 7.004482270563441e-05, "loss": 0.0052, "step": 6550 }, { "epoch": 1.2977544762093185, "grad_norm": 0.03195886177028069, "learning_rate": 6.993932983258514e-05, "loss": 0.0052, "step": 6560 }, { "epoch": 1.2997329112671876, "grad_norm": 0.018707038907315402, "learning_rate": 6.983373130890774e-05, "loss": 0.0051, "step": 6570 }, { "epoch": 1.301711346325057, "grad_norm": 0.025053441620057796, "learning_rate": 6.972802769412706e-05, "loss": 0.0051, "step": 6580 }, { "epoch": 1.3036897813829262, "grad_norm": 0.02784525014955868, "learning_rate": 6.962221954832476e-05, "loss": 0.0051, "step": 6590 }, { "epoch": 1.3056682164407953, "grad_norm": 0.026357933874572047, "learning_rate": 6.95163074321364e-05, "loss": 0.0052, "step": 6600 }, { "epoch": 1.3076466514986644, "grad_norm": 0.017494405172256713, "learning_rate": 6.941029190674844e-05, "loss": 0.0051, "step": 6610 }, { "epoch": 1.3096250865565338, "grad_norm": 0.022737548942178238, "learning_rate": 6.93041735338952e-05, "loss": 0.0052, "step": 6620 }, { "epoch": 1.311603521614403, "grad_norm": 0.025209369509202235, "learning_rate": 6.919795287585604e-05, "loss": 0.0052, "step": 6630 }, { "epoch": 1.3135819566722722, "grad_norm": 0.02674995083847258, "learning_rate": 6.909163049545223e-05, "loss": 0.0051, "step": 6640 }, { "epoch": 1.3155603917301415, "grad_norm": 0.030351836741292392, "learning_rate": 6.898520695604404e-05, "loss": 0.0051, "step": 6650 }, { "epoch": 1.3175388267880108, "grad_norm": 0.016653010851606594, "learning_rate": 6.887868282152775e-05, "loss": 0.0051, "step": 6660 }, { "epoch": 1.31951726184588, "grad_norm": 0.02010973632606591, "learning_rate": 6.877205865633262e-05, "loss": 0.0051, "step": 6670 }, { "epoch": 1.321495696903749, "grad_norm": 0.01901241478903529, "learning_rate": 6.866533502541802e-05, "loss": 0.0052, "step": 6680 }, { "epoch": 1.3234741319616183, "grad_norm": 0.019888260124643883, "learning_rate": 6.855851249427024e-05, "loss": 0.0051, "step": 6690 }, { "epoch": 1.3254525670194877, "grad_norm": 0.01761253293523921, "learning_rate": 6.845159162889965e-05, "loss": 0.0051, "step": 6700 }, { "epoch": 1.3274310020773568, "grad_norm": 0.021259454587723152, "learning_rate": 6.834457299583768e-05, "loss": 0.0051, "step": 6710 }, { "epoch": 1.329409437135226, "grad_norm": 0.01584901495067816, "learning_rate": 6.823745716213377e-05, "loss": 0.005, "step": 6720 }, { "epoch": 1.3313878721930952, "grad_norm": 0.021675416020175562, "learning_rate": 6.813024469535236e-05, "loss": 0.0051, "step": 6730 }, { "epoch": 1.3333663072509645, "grad_norm": 0.017256488165444287, "learning_rate": 6.802293616356997e-05, "loss": 0.0051, "step": 6740 }, { "epoch": 1.3353447423088336, "grad_norm": 0.01988095994815182, "learning_rate": 6.791553213537209e-05, "loss": 0.0051, "step": 6750 }, { "epoch": 1.337323177366703, "grad_norm": 0.029010702803392147, "learning_rate": 6.780803317985021e-05, "loss": 0.0051, "step": 6760 }, { "epoch": 1.3393016124245722, "grad_norm": 0.02229326528449447, "learning_rate": 6.77004398665988e-05, "loss": 0.0051, "step": 6770 }, { "epoch": 1.3412800474824413, "grad_norm": 0.024231994266251184, "learning_rate": 6.759275276571233e-05, "loss": 0.0051, "step": 6780 }, { "epoch": 1.3432584825403107, "grad_norm": 0.019760509190009986, "learning_rate": 6.748497244778214e-05, "loss": 0.0051, "step": 6790 }, { "epoch": 1.3452369175981798, "grad_norm": 0.016710462119061714, "learning_rate": 6.737709948389356e-05, "loss": 0.0051, "step": 6800 }, { "epoch": 1.347215352656049, "grad_norm": 0.01887818681230359, "learning_rate": 6.726913444562283e-05, "loss": 0.0051, "step": 6810 }, { "epoch": 1.3491937877139182, "grad_norm": 0.015721172581988692, "learning_rate": 6.716107790503398e-05, "loss": 0.0051, "step": 6820 }, { "epoch": 1.3511722227717875, "grad_norm": 0.020616954688510183, "learning_rate": 6.705293043467589e-05, "loss": 0.005, "step": 6830 }, { "epoch": 1.3531506578296568, "grad_norm": 0.02169976718548275, "learning_rate": 6.69446926075793e-05, "loss": 0.0051, "step": 6840 }, { "epoch": 1.355129092887526, "grad_norm": 0.017649101893321903, "learning_rate": 6.683636499725365e-05, "loss": 0.0051, "step": 6850 }, { "epoch": 1.3571075279453952, "grad_norm": 0.023464058138752606, "learning_rate": 6.672794817768414e-05, "loss": 0.0051, "step": 6860 }, { "epoch": 1.3590859630032643, "grad_norm": 0.017944826253035673, "learning_rate": 6.661944272332867e-05, "loss": 0.005, "step": 6870 }, { "epoch": 1.3610643980611337, "grad_norm": 0.013136162236096264, "learning_rate": 6.651084920911473e-05, "loss": 0.0051, "step": 6880 }, { "epoch": 1.3630428331190028, "grad_norm": 0.02017698183742181, "learning_rate": 6.640216821043646e-05, "loss": 0.0051, "step": 6890 }, { "epoch": 1.365021268176872, "grad_norm": 0.015804745118237505, "learning_rate": 6.629340030315148e-05, "loss": 0.0051, "step": 6900 }, { "epoch": 1.3669997032347414, "grad_norm": 0.02231726477759663, "learning_rate": 6.618454606357796e-05, "loss": 0.005, "step": 6910 }, { "epoch": 1.3689781382926105, "grad_norm": 0.014535587590450497, "learning_rate": 6.607560606849149e-05, "loss": 0.0051, "step": 6920 }, { "epoch": 1.3709565733504798, "grad_norm": 0.014638216120479886, "learning_rate": 6.596658089512208e-05, "loss": 0.005, "step": 6930 }, { "epoch": 1.372935008408349, "grad_norm": 0.02323694463583421, "learning_rate": 6.585747112115096e-05, "loss": 0.0051, "step": 6940 }, { "epoch": 1.3749134434662182, "grad_norm": 0.01574608265542015, "learning_rate": 6.574827732470779e-05, "loss": 0.0051, "step": 6950 }, { "epoch": 1.3768918785240873, "grad_norm": 0.029953066042503078, "learning_rate": 6.563900008436727e-05, "loss": 0.0051, "step": 6960 }, { "epoch": 1.3788703135819567, "grad_norm": 0.02122171639692881, "learning_rate": 6.552963997914635e-05, "loss": 0.005, "step": 6970 }, { "epoch": 1.380848748639826, "grad_norm": 0.0211923854555959, "learning_rate": 6.542019758850097e-05, "loss": 0.0051, "step": 6980 }, { "epoch": 1.382827183697695, "grad_norm": 0.017206584323745094, "learning_rate": 6.531067349232313e-05, "loss": 0.005, "step": 6990 }, { "epoch": 1.3848056187555644, "grad_norm": 0.02368424130441156, "learning_rate": 6.52010682709377e-05, "loss": 0.0051, "step": 7000 }, { "epoch": 1.3867840538134335, "grad_norm": 0.018986517916588494, "learning_rate": 6.509138250509946e-05, "loss": 0.0051, "step": 7010 }, { "epoch": 1.3887624888713028, "grad_norm": 0.021783652193033926, "learning_rate": 6.498161677598993e-05, "loss": 0.0051, "step": 7020 }, { "epoch": 1.390740923929172, "grad_norm": 0.01989221560555763, "learning_rate": 6.48717716652143e-05, "loss": 0.005, "step": 7030 }, { "epoch": 1.3927193589870412, "grad_norm": 0.028824501466034488, "learning_rate": 6.476184775479841e-05, "loss": 0.005, "step": 7040 }, { "epoch": 1.3946977940449106, "grad_norm": 0.013118894131611164, "learning_rate": 6.465184562718563e-05, "loss": 0.005, "step": 7050 }, { "epoch": 1.3966762291027797, "grad_norm": 0.030419341589724876, "learning_rate": 6.454176586523373e-05, "loss": 0.0051, "step": 7060 }, { "epoch": 1.398654664160649, "grad_norm": 0.018289203329966308, "learning_rate": 6.443160905221187e-05, "loss": 0.005, "step": 7070 }, { "epoch": 1.400633099218518, "grad_norm": 0.011335676134755196, "learning_rate": 6.432137577179749e-05, "loss": 0.005, "step": 7080 }, { "epoch": 1.4026115342763874, "grad_norm": 0.019403411549867747, "learning_rate": 6.421106660807314e-05, "loss": 0.005, "step": 7090 }, { "epoch": 1.4045899693342565, "grad_norm": 0.016219224353396837, "learning_rate": 6.410068214552351e-05, "loss": 0.005, "step": 7100 }, { "epoch": 1.4065684043921258, "grad_norm": 0.01349969499784592, "learning_rate": 6.399022296903224e-05, "loss": 0.005, "step": 7110 }, { "epoch": 1.4085468394499951, "grad_norm": 0.0237897046108694, "learning_rate": 6.387968966387881e-05, "loss": 0.0051, "step": 7120 }, { "epoch": 1.4105252745078642, "grad_norm": 0.024510245392803796, "learning_rate": 6.376908281573556e-05, "loss": 0.0051, "step": 7130 }, { "epoch": 1.4125037095657336, "grad_norm": 0.019887673658650275, "learning_rate": 6.365840301066445e-05, "loss": 0.005, "step": 7140 }, { "epoch": 1.4144821446236027, "grad_norm": 0.023947486968730117, "learning_rate": 6.354765083511401e-05, "loss": 0.0051, "step": 7150 }, { "epoch": 1.416460579681472, "grad_norm": 0.020353519957312532, "learning_rate": 6.343682687591626e-05, "loss": 0.005, "step": 7160 }, { "epoch": 1.418439014739341, "grad_norm": 0.016826958038086558, "learning_rate": 6.332593172028356e-05, "loss": 0.005, "step": 7170 }, { "epoch": 1.4204174497972104, "grad_norm": 0.020008407320600014, "learning_rate": 6.32149659558055e-05, "loss": 0.0051, "step": 7180 }, { "epoch": 1.4223958848550797, "grad_norm": 0.025437068408178262, "learning_rate": 6.310393017044582e-05, "loss": 0.0051, "step": 7190 }, { "epoch": 1.4243743199129488, "grad_norm": 0.018363317692837968, "learning_rate": 6.299282495253925e-05, "loss": 0.005, "step": 7200 }, { "epoch": 1.4263527549708181, "grad_norm": 0.022959479829482075, "learning_rate": 6.288165089078842e-05, "loss": 0.0051, "step": 7210 }, { "epoch": 1.4283311900286872, "grad_norm": 0.01912696487761665, "learning_rate": 6.277040857426077e-05, "loss": 0.0051, "step": 7220 }, { "epoch": 1.4303096250865566, "grad_norm": 0.019023963098169925, "learning_rate": 6.265909859238535e-05, "loss": 0.0051, "step": 7230 }, { "epoch": 1.4322880601444257, "grad_norm": 0.025756968836663428, "learning_rate": 6.254772153494977e-05, "loss": 0.0051, "step": 7240 }, { "epoch": 1.434266495202295, "grad_norm": 0.02490517252114092, "learning_rate": 6.243627799209703e-05, "loss": 0.005, "step": 7250 }, { "epoch": 1.4362449302601643, "grad_norm": 0.024378475943524002, "learning_rate": 6.232476855432243e-05, "loss": 0.0051, "step": 7260 }, { "epoch": 1.4382233653180334, "grad_norm": 0.023754439546808573, "learning_rate": 6.22131938124704e-05, "loss": 0.0051, "step": 7270 }, { "epoch": 1.4402018003759027, "grad_norm": 0.014664055217834237, "learning_rate": 6.210155435773138e-05, "loss": 0.0051, "step": 7280 }, { "epoch": 1.4421802354337718, "grad_norm": 0.015002712603649751, "learning_rate": 6.198985078163876e-05, "loss": 0.005, "step": 7290 }, { "epoch": 1.4441586704916411, "grad_norm": 0.015046014945227256, "learning_rate": 6.187808367606561e-05, "loss": 0.005, "step": 7300 }, { "epoch": 1.4461371055495102, "grad_norm": 0.019315429968942022, "learning_rate": 6.176625363322163e-05, "loss": 0.0051, "step": 7310 }, { "epoch": 1.4481155406073796, "grad_norm": 0.01744248504368862, "learning_rate": 6.165436124565007e-05, "loss": 0.005, "step": 7320 }, { "epoch": 1.450093975665249, "grad_norm": 0.023906053351486815, "learning_rate": 6.154240710622445e-05, "loss": 0.005, "step": 7330 }, { "epoch": 1.452072410723118, "grad_norm": 0.017899151235512036, "learning_rate": 6.143039180814548e-05, "loss": 0.005, "step": 7340 }, { "epoch": 1.4540508457809873, "grad_norm": 0.03603011604556944, "learning_rate": 6.131831594493799e-05, "loss": 0.005, "step": 7350 }, { "epoch": 1.4560292808388564, "grad_norm": 0.015281731059167402, "learning_rate": 6.120618011044766e-05, "loss": 0.005, "step": 7360 }, { "epoch": 1.4580077158967257, "grad_norm": 0.013057675147702603, "learning_rate": 6.109398489883795e-05, "loss": 0.005, "step": 7370 }, { "epoch": 1.4599861509545948, "grad_norm": 0.026849784386253833, "learning_rate": 6.0981730904586954e-05, "loss": 0.005, "step": 7380 }, { "epoch": 1.4619645860124642, "grad_norm": 0.024668194697485976, "learning_rate": 6.086941872248424e-05, "loss": 0.005, "step": 7390 }, { "epoch": 1.4639430210703335, "grad_norm": 0.01485927096641802, "learning_rate": 6.0757048947627636e-05, "loss": 0.005, "step": 7400 }, { "epoch": 1.4659214561282026, "grad_norm": 0.015471832387451756, "learning_rate": 6.0644622175420204e-05, "loss": 0.0051, "step": 7410 }, { "epoch": 1.467899891186072, "grad_norm": 0.008500689302769012, "learning_rate": 6.0532139001566946e-05, "loss": 0.005, "step": 7420 }, { "epoch": 1.469878326243941, "grad_norm": 0.01978624221199307, "learning_rate": 6.041960002207174e-05, "loss": 0.005, "step": 7430 }, { "epoch": 1.4718567613018103, "grad_norm": 0.0170530811077618, "learning_rate": 6.0307005833234195e-05, "loss": 0.005, "step": 7440 }, { "epoch": 1.4738351963596794, "grad_norm": 0.03423738675798128, "learning_rate": 6.019435703164639e-05, "loss": 0.005, "step": 7450 }, { "epoch": 1.4758136314175487, "grad_norm": 0.012343973469495919, "learning_rate": 6.008165421418981e-05, "loss": 0.005, "step": 7460 }, { "epoch": 1.477792066475418, "grad_norm": 0.01709058608629632, "learning_rate": 5.9968897978032135e-05, "loss": 0.005, "step": 7470 }, { "epoch": 1.4797705015332872, "grad_norm": 0.012741495867049412, "learning_rate": 5.9856088920624096e-05, "loss": 0.005, "step": 7480 }, { "epoch": 1.4817489365911565, "grad_norm": 0.01740107134613611, "learning_rate": 5.974322763969631e-05, "loss": 0.005, "step": 7490 }, { "epoch": 1.4837273716490256, "grad_norm": 0.02479099330141599, "learning_rate": 5.963031473325609e-05, "loss": 0.0051, "step": 7500 }, { "epoch": 1.485705806706895, "grad_norm": 0.022565232069577567, "learning_rate": 5.9517350799584307e-05, "loss": 0.005, "step": 7510 }, { "epoch": 1.487684241764764, "grad_norm": 0.02216359796942625, "learning_rate": 5.9404336437232175e-05, "loss": 0.0051, "step": 7520 }, { "epoch": 1.4896626768226333, "grad_norm": 0.022622625844978875, "learning_rate": 5.929127224501818e-05, "loss": 0.005, "step": 7530 }, { "epoch": 1.4916411118805026, "grad_norm": 0.023137418065357185, "learning_rate": 5.917815882202477e-05, "loss": 0.0051, "step": 7540 }, { "epoch": 1.4936195469383717, "grad_norm": 0.019787188447914057, "learning_rate": 5.9064996767595236e-05, "loss": 0.005, "step": 7550 }, { "epoch": 1.495597981996241, "grad_norm": 0.009377442994281712, "learning_rate": 5.8951786681330625e-05, "loss": 0.0051, "step": 7560 }, { "epoch": 1.4975764170541102, "grad_norm": 0.013951202038574532, "learning_rate": 5.883852916308641e-05, "loss": 0.005, "step": 7570 }, { "epoch": 1.4995548521119795, "grad_norm": 0.021086360976989682, "learning_rate": 5.872522481296941e-05, "loss": 0.0051, "step": 7580 }, { "epoch": 1.5015332871698486, "grad_norm": 0.019856003192296866, "learning_rate": 5.861187423133464e-05, "loss": 0.0051, "step": 7590 }, { "epoch": 1.503511722227718, "grad_norm": 0.019496054137121212, "learning_rate": 5.8498478018782e-05, "loss": 0.005, "step": 7600 }, { "epoch": 1.5054901572855872, "grad_norm": 0.015043724778703605, "learning_rate": 5.83850367761532e-05, "loss": 0.005, "step": 7610 }, { "epoch": 1.5074685923434563, "grad_norm": 0.004238497682294309, "learning_rate": 5.8271551104528554e-05, "loss": 0.005, "step": 7620 }, { "epoch": 1.5094470274013254, "grad_norm": 0.011578901720175248, "learning_rate": 5.815802160522379e-05, "loss": 0.005, "step": 7630 }, { "epoch": 1.5114254624591947, "grad_norm": 0.021180582251365355, "learning_rate": 5.804444887978684e-05, "loss": 0.005, "step": 7640 }, { "epoch": 1.513403897517064, "grad_norm": 0.019812035361194903, "learning_rate": 5.7930833529994696e-05, "loss": 0.005, "step": 7650 }, { "epoch": 1.5153823325749332, "grad_norm": 0.01996149624229686, "learning_rate": 5.7817176157850194e-05, "loss": 0.005, "step": 7660 }, { "epoch": 1.5173607676328025, "grad_norm": 0.018232046733898024, "learning_rate": 5.770347736557884e-05, "loss": 0.005, "step": 7670 }, { "epoch": 1.5193392026906718, "grad_norm": 0.023129797478894258, "learning_rate": 5.758973775562557e-05, "loss": 0.005, "step": 7680 }, { "epoch": 1.521317637748541, "grad_norm": 0.015540240065626319, "learning_rate": 5.7475957930651646e-05, "loss": 0.005, "step": 7690 }, { "epoch": 1.52329607280641, "grad_norm": 0.026294421490536977, "learning_rate": 5.736213849353138e-05, "loss": 0.005, "step": 7700 }, { "epoch": 1.5252745078642793, "grad_norm": 0.01737954111989701, "learning_rate": 5.724828004734899e-05, "loss": 0.005, "step": 7710 }, { "epoch": 1.5272529429221486, "grad_norm": 0.020002105346910028, "learning_rate": 5.713438319539538e-05, "loss": 0.005, "step": 7720 }, { "epoch": 1.5292313779800177, "grad_norm": 0.024972699481939076, "learning_rate": 5.702044854116494e-05, "loss": 0.005, "step": 7730 }, { "epoch": 1.531209813037887, "grad_norm": 0.021915353952822143, "learning_rate": 5.6906476688352406e-05, "loss": 0.005, "step": 7740 }, { "epoch": 1.5331882480957564, "grad_norm": 0.013065163342528125, "learning_rate": 5.679246824084955e-05, "loss": 0.0049, "step": 7750 }, { "epoch": 1.5351666831536255, "grad_norm": 0.016434610001804666, "learning_rate": 5.6678423802742074e-05, "loss": 0.005, "step": 7760 }, { "epoch": 1.5371451182114946, "grad_norm": 0.011347921448019065, "learning_rate": 5.65643439783064e-05, "loss": 0.005, "step": 7770 }, { "epoch": 1.539123553269364, "grad_norm": 0.0210193810791627, "learning_rate": 5.645022937200641e-05, "loss": 0.0049, "step": 7780 }, { "epoch": 1.5411019883272332, "grad_norm": 0.018957713774373476, "learning_rate": 5.6336080588490326e-05, "loss": 0.005, "step": 7790 }, { "epoch": 1.5430804233851023, "grad_norm": 0.013521677459780612, "learning_rate": 5.622189823258743e-05, "loss": 0.005, "step": 7800 }, { "epoch": 1.5450588584429716, "grad_norm": 0.016913820806355274, "learning_rate": 5.610768290930488e-05, "loss": 0.005, "step": 7810 }, { "epoch": 1.547037293500841, "grad_norm": 0.018638305189176706, "learning_rate": 5.599343522382454e-05, "loss": 0.0051, "step": 7820 }, { "epoch": 1.54901572855871, "grad_norm": 0.019451747976412023, "learning_rate": 5.587915578149976e-05, "loss": 0.005, "step": 7830 }, { "epoch": 1.5509941636165792, "grad_norm": 0.021775789031113036, "learning_rate": 5.576484518785211e-05, "loss": 0.005, "step": 7840 }, { "epoch": 1.5529725986744485, "grad_norm": 0.0132383882484977, "learning_rate": 5.5650504048568255e-05, "loss": 0.005, "step": 7850 }, { "epoch": 1.5549510337323178, "grad_norm": 0.014276323337133742, "learning_rate": 5.553613296949671e-05, "loss": 0.0049, "step": 7860 }, { "epoch": 1.556929468790187, "grad_norm": 0.019701037540286685, "learning_rate": 5.542173255664462e-05, "loss": 0.005, "step": 7870 }, { "epoch": 1.5589079038480562, "grad_norm": 0.03305321556911838, "learning_rate": 5.530730341617455e-05, "loss": 0.005, "step": 7880 }, { "epoch": 1.5608863389059255, "grad_norm": 0.025402148702672267, "learning_rate": 5.519284615440129e-05, "loss": 0.0051, "step": 7890 }, { "epoch": 1.5628647739637946, "grad_norm": 0.020254793952455018, "learning_rate": 5.5078361377788645e-05, "loss": 0.005, "step": 7900 }, { "epoch": 1.5648432090216637, "grad_norm": 0.009165490638382819, "learning_rate": 5.4963849692946166e-05, "loss": 0.005, "step": 7910 }, { "epoch": 1.566821644079533, "grad_norm": 0.03615677983885278, "learning_rate": 5.484931170662606e-05, "loss": 0.0051, "step": 7920 }, { "epoch": 1.5688000791374024, "grad_norm": 0.016977517976904895, "learning_rate": 5.4734748025719804e-05, "loss": 0.005, "step": 7930 }, { "epoch": 1.5707785141952715, "grad_norm": 0.02058438180331575, "learning_rate": 5.462015925725509e-05, "loss": 0.0049, "step": 7940 }, { "epoch": 1.5727569492531408, "grad_norm": 0.009976596641837683, "learning_rate": 5.4505546008392505e-05, "loss": 0.0049, "step": 7950 }, { "epoch": 1.5747353843110101, "grad_norm": 0.019310230224817013, "learning_rate": 5.439090888642237e-05, "loss": 0.005, "step": 7960 }, { "epoch": 1.5767138193688792, "grad_norm": 0.018253577641422897, "learning_rate": 5.427624849876145e-05, "loss": 0.005, "step": 7970 }, { "epoch": 1.5786922544267483, "grad_norm": 0.028761144748151373, "learning_rate": 5.4161565452949856e-05, "loss": 0.005, "step": 7980 }, { "epoch": 1.5806706894846176, "grad_norm": 0.01829809542554448, "learning_rate": 5.40468603566477e-05, "loss": 0.005, "step": 7990 }, { "epoch": 1.582649124542487, "grad_norm": 0.015390464151235236, "learning_rate": 5.393213381763197e-05, "loss": 0.005, "step": 8000 }, { "epoch": 1.584627559600356, "grad_norm": 0.02021095969993427, "learning_rate": 5.3817386443793235e-05, "loss": 0.005, "step": 8010 }, { "epoch": 1.5866059946582254, "grad_norm": 0.025682965467523678, "learning_rate": 5.370261884313249e-05, "loss": 0.005, "step": 8020 }, { "epoch": 1.5885844297160947, "grad_norm": 0.02755057285676576, "learning_rate": 5.35878316237579e-05, "loss": 0.005, "step": 8030 }, { "epoch": 1.5905628647739638, "grad_norm": 0.023945481285186383, "learning_rate": 5.347302539388155e-05, "loss": 0.005, "step": 8040 }, { "epoch": 1.592541299831833, "grad_norm": 0.017878902181083733, "learning_rate": 5.335820076181629e-05, "loss": 0.005, "step": 8050 }, { "epoch": 1.5945197348897022, "grad_norm": 0.01979889576145264, "learning_rate": 5.324335833597245e-05, "loss": 0.005, "step": 8060 }, { "epoch": 1.5964981699475715, "grad_norm": 0.026622064999786324, "learning_rate": 5.312849872485468e-05, "loss": 0.005, "step": 8070 }, { "epoch": 1.5984766050054406, "grad_norm": 0.01965849907616097, "learning_rate": 5.3013622537058635e-05, "loss": 0.005, "step": 8080 }, { "epoch": 1.60045504006331, "grad_norm": 0.010519352152671338, "learning_rate": 5.289873038126786e-05, "loss": 0.005, "step": 8090 }, { "epoch": 1.6024334751211793, "grad_norm": 0.026551189500221485, "learning_rate": 5.2783822866250466e-05, "loss": 0.0051, "step": 8100 }, { "epoch": 1.6044119101790484, "grad_norm": 0.020851330326579053, "learning_rate": 5.266890060085595e-05, "loss": 0.0051, "step": 8110 }, { "epoch": 1.6063903452369175, "grad_norm": 0.01851436671553028, "learning_rate": 5.255396419401198e-05, "loss": 0.005, "step": 8120 }, { "epoch": 1.6083687802947868, "grad_norm": 0.017262362377280727, "learning_rate": 5.2439014254721166e-05, "loss": 0.005, "step": 8130 }, { "epoch": 1.6103472153526561, "grad_norm": 0.014697898495626496, "learning_rate": 5.232405139205778e-05, "loss": 0.005, "step": 8140 }, { "epoch": 1.6123256504105252, "grad_norm": 0.022247270684886817, "learning_rate": 5.220907621516461e-05, "loss": 0.005, "step": 8150 }, { "epoch": 1.6143040854683945, "grad_norm": 0.019402545756230762, "learning_rate": 5.209408933324967e-05, "loss": 0.005, "step": 8160 }, { "epoch": 1.6162825205262639, "grad_norm": 0.022450552351294117, "learning_rate": 5.197909135558302e-05, "loss": 0.005, "step": 8170 }, { "epoch": 1.618260955584133, "grad_norm": 0.03358907700028808, "learning_rate": 5.186408289149346e-05, "loss": 0.005, "step": 8180 }, { "epoch": 1.620239390642002, "grad_norm": 0.018087538434867097, "learning_rate": 5.1749064550365414e-05, "loss": 0.005, "step": 8190 }, { "epoch": 1.6222178256998714, "grad_norm": 0.016573774611214762, "learning_rate": 5.163403694163559e-05, "loss": 0.005, "step": 8200 }, { "epoch": 1.6241962607577407, "grad_norm": 0.00963485992227461, "learning_rate": 5.1519000674789833e-05, "loss": 0.005, "step": 8210 }, { "epoch": 1.6261746958156098, "grad_norm": 0.016804767027719464, "learning_rate": 5.140395635935987e-05, "loss": 0.005, "step": 8220 }, { "epoch": 1.6281531308734791, "grad_norm": 0.027999442593928323, "learning_rate": 5.128890460492004e-05, "loss": 0.005, "step": 8230 }, { "epoch": 1.6301315659313484, "grad_norm": 0.014270350937864097, "learning_rate": 5.117384602108413e-05, "loss": 0.005, "step": 8240 }, { "epoch": 1.6321100009892175, "grad_norm": 0.04057786802895742, "learning_rate": 5.1058781217502084e-05, "loss": 0.0051, "step": 8250 }, { "epoch": 1.6340884360470866, "grad_norm": 0.01970321239695706, "learning_rate": 5.094371080385685e-05, "loss": 0.005, "step": 8260 }, { "epoch": 1.636066871104956, "grad_norm": 0.019021304303569034, "learning_rate": 5.082863538986102e-05, "loss": 0.0052, "step": 8270 }, { "epoch": 1.6380453061628253, "grad_norm": 0.02445925994960959, "learning_rate": 5.071355558525379e-05, "loss": 0.005, "step": 8280 }, { "epoch": 1.6400237412206944, "grad_norm": 0.0221777296600126, "learning_rate": 5.059847199979752e-05, "loss": 0.005, "step": 8290 }, { "epoch": 1.6420021762785637, "grad_norm": 0.013318345734434276, "learning_rate": 5.048338524327466e-05, "loss": 0.005, "step": 8300 }, { "epoch": 1.643980611336433, "grad_norm": 0.01614997772464178, "learning_rate": 5.036829592548446e-05, "loss": 0.005, "step": 8310 }, { "epoch": 1.6459590463943021, "grad_norm": 0.014391610740361764, "learning_rate": 5.02532046562397e-05, "loss": 0.005, "step": 8320 }, { "epoch": 1.6479374814521712, "grad_norm": 0.023949265101614464, "learning_rate": 5.013811204536353e-05, "loss": 0.005, "step": 8330 }, { "epoch": 1.6499159165100405, "grad_norm": 0.020986958733913052, "learning_rate": 5.002301870268623e-05, "loss": 0.005, "step": 8340 }, { "epoch": 1.6518943515679099, "grad_norm": 0.012079374530200023, "learning_rate": 4.9907925238041925e-05, "loss": 0.005, "step": 8350 }, { "epoch": 1.653872786625779, "grad_norm": 0.024296664486975334, "learning_rate": 4.9792832261265384e-05, "loss": 0.005, "step": 8360 }, { "epoch": 1.6558512216836483, "grad_norm": 0.016275701860330602, "learning_rate": 4.967774038218882e-05, "loss": 0.005, "step": 8370 }, { "epoch": 1.6578296567415176, "grad_norm": 0.020560168693868723, "learning_rate": 4.95626502106386e-05, "loss": 0.005, "step": 8380 }, { "epoch": 1.6598080917993867, "grad_norm": 0.017225604563944782, "learning_rate": 4.944756235643205e-05, "loss": 0.005, "step": 8390 }, { "epoch": 1.6617865268572558, "grad_norm": 0.02533008991214159, "learning_rate": 4.933247742937425e-05, "loss": 0.005, "step": 8400 }, { "epoch": 1.6637649619151251, "grad_norm": 0.016554780374138798, "learning_rate": 4.92173960392547e-05, "loss": 0.0051, "step": 8410 }, { "epoch": 1.6657433969729945, "grad_norm": 0.017829554256082163, "learning_rate": 4.910231879584427e-05, "loss": 0.0051, "step": 8420 }, { "epoch": 1.6677218320308635, "grad_norm": 0.010298815907596645, "learning_rate": 4.8987246308891716e-05, "loss": 0.005, "step": 8430 }, { "epoch": 1.6697002670887326, "grad_norm": 0.01999975074366446, "learning_rate": 4.887217918812072e-05, "loss": 0.005, "step": 8440 }, { "epoch": 1.6716787021466022, "grad_norm": 0.02559617560384222, "learning_rate": 4.875711804322643e-05, "loss": 0.005, "step": 8450 }, { "epoch": 1.6736571372044713, "grad_norm": 0.02917082861660753, "learning_rate": 4.8642063483872415e-05, "loss": 0.005, "step": 8460 }, { "epoch": 1.6756355722623404, "grad_norm": 0.012992554400041802, "learning_rate": 4.85270161196873e-05, "loss": 0.0051, "step": 8470 }, { "epoch": 1.6776140073202097, "grad_norm": 0.015062801590814197, "learning_rate": 4.841197656026159e-05, "loss": 0.005, "step": 8480 }, { "epoch": 1.679592442378079, "grad_norm": 0.028408253950640647, "learning_rate": 4.8296945415144466e-05, "loss": 0.005, "step": 8490 }, { "epoch": 1.6815708774359481, "grad_norm": 0.016849293030548024, "learning_rate": 4.818192329384047e-05, "loss": 0.005, "step": 8500 }, { "epoch": 1.6835493124938172, "grad_norm": 0.010601199016409856, "learning_rate": 4.806691080580638e-05, "loss": 0.005, "step": 8510 }, { "epoch": 1.6855277475516868, "grad_norm": 0.018245122418775395, "learning_rate": 4.795190856044792e-05, "loss": 0.005, "step": 8520 }, { "epoch": 1.6875061826095559, "grad_norm": 0.026069194411481738, "learning_rate": 4.783691716711652e-05, "loss": 0.005, "step": 8530 }, { "epoch": 1.689484617667425, "grad_norm": 0.013212093225055526, "learning_rate": 4.772193723510613e-05, "loss": 0.005, "step": 8540 }, { "epoch": 1.6914630527252943, "grad_norm": 0.007776032709454633, "learning_rate": 4.7606969373649986e-05, "loss": 0.005, "step": 8550 }, { "epoch": 1.6934414877831636, "grad_norm": 0.02935697050945925, "learning_rate": 4.7492014191917334e-05, "loss": 0.005, "step": 8560 }, { "epoch": 1.6954199228410327, "grad_norm": 0.014226775843143202, "learning_rate": 4.737707229901022e-05, "loss": 0.005, "step": 8570 }, { "epoch": 1.6973983578989018, "grad_norm": 0.018861684894261748, "learning_rate": 4.726214430396033e-05, "loss": 0.005, "step": 8580 }, { "epoch": 1.6993767929567714, "grad_norm": 0.043791815622162805, "learning_rate": 4.71472308157257e-05, "loss": 0.0051, "step": 8590 }, { "epoch": 1.7013552280146405, "grad_norm": 0.039601624350728705, "learning_rate": 4.7032332443187467e-05, "loss": 0.0052, "step": 8600 }, { "epoch": 1.7033336630725096, "grad_norm": 0.021886175299307906, "learning_rate": 4.69174497951467e-05, "loss": 0.0051, "step": 8610 }, { "epoch": 1.7053120981303789, "grad_norm": 0.026242786592011687, "learning_rate": 4.680258348032115e-05, "loss": 0.005, "step": 8620 }, { "epoch": 1.7072905331882482, "grad_norm": 0.021579641517114754, "learning_rate": 4.668773410734201e-05, "loss": 0.005, "step": 8630 }, { "epoch": 1.7092689682461173, "grad_norm": 0.015146096369545345, "learning_rate": 4.657290228475073e-05, "loss": 0.0051, "step": 8640 }, { "epoch": 1.7112474033039864, "grad_norm": 0.02382543824628399, "learning_rate": 4.6458088620995736e-05, "loss": 0.005, "step": 8650 }, { "epoch": 1.713225838361856, "grad_norm": 0.016554690628563396, "learning_rate": 4.6343293724429266e-05, "loss": 0.005, "step": 8660 }, { "epoch": 1.715204273419725, "grad_norm": 0.018982688582958013, "learning_rate": 4.622851820330412e-05, "loss": 0.005, "step": 8670 }, { "epoch": 1.7171827084775941, "grad_norm": 0.018140088086666048, "learning_rate": 4.6113762665770385e-05, "loss": 0.005, "step": 8680 }, { "epoch": 1.7191611435354635, "grad_norm": 0.0187869564288913, "learning_rate": 4.599902771987232e-05, "loss": 0.0049, "step": 8690 }, { "epoch": 1.7211395785933328, "grad_norm": 0.02317509663792515, "learning_rate": 4.588431397354508e-05, "loss": 0.005, "step": 8700 }, { "epoch": 1.7231180136512019, "grad_norm": 0.015881235895266826, "learning_rate": 4.576962203461144e-05, "loss": 0.0051, "step": 8710 }, { "epoch": 1.725096448709071, "grad_norm": 0.022320761968210312, "learning_rate": 4.5654952510778646e-05, "loss": 0.005, "step": 8720 }, { "epoch": 1.7270748837669405, "grad_norm": 0.016373366022838926, "learning_rate": 4.5540306009635194e-05, "loss": 0.005, "step": 8730 }, { "epoch": 1.7290533188248096, "grad_norm": 0.006682719367650814, "learning_rate": 4.54256831386476e-05, "loss": 0.005, "step": 8740 }, { "epoch": 1.7310317538826787, "grad_norm": 0.02702070484211182, "learning_rate": 4.531108450515712e-05, "loss": 0.005, "step": 8750 }, { "epoch": 1.733010188940548, "grad_norm": 0.017441167597767725, "learning_rate": 4.5196510716376645e-05, "loss": 0.005, "step": 8760 }, { "epoch": 1.7349886239984174, "grad_norm": 0.011914925084002794, "learning_rate": 4.508196237938741e-05, "loss": 0.0051, "step": 8770 }, { "epoch": 1.7369670590562865, "grad_norm": 0.05696947996398024, "learning_rate": 4.4967440101135736e-05, "loss": 0.0051, "step": 8780 }, { "epoch": 1.7389454941141556, "grad_norm": 0.012894653392142639, "learning_rate": 4.485294448842996e-05, "loss": 0.0052, "step": 8790 }, { "epoch": 1.7409239291720249, "grad_norm": 0.0230843164804097, "learning_rate": 4.473847614793705e-05, "loss": 0.0052, "step": 8800 }, { "epoch": 1.7429023642298942, "grad_norm": 0.02485372397393329, "learning_rate": 4.462403568617951e-05, "loss": 0.005, "step": 8810 }, { "epoch": 1.7448807992877633, "grad_norm": 0.061795447964485016, "learning_rate": 4.4509623709532125e-05, "loss": 0.005, "step": 8820 }, { "epoch": 1.7468592343456326, "grad_norm": 0.028492791976632723, "learning_rate": 4.439524082421872e-05, "loss": 0.0051, "step": 8830 }, { "epoch": 1.748837669403502, "grad_norm": 0.02236933055851863, "learning_rate": 4.4280887636309035e-05, "loss": 0.0051, "step": 8840 }, { "epoch": 1.750816104461371, "grad_norm": 0.02510067858677648, "learning_rate": 4.416656475171537e-05, "loss": 0.005, "step": 8850 }, { "epoch": 1.7527945395192401, "grad_norm": 0.011442433071206909, "learning_rate": 4.405227277618953e-05, "loss": 0.0051, "step": 8860 }, { "epoch": 1.7547729745771095, "grad_norm": 0.020268678143223168, "learning_rate": 4.393801231531952e-05, "loss": 0.0051, "step": 8870 }, { "epoch": 1.7567514096349788, "grad_norm": 0.03781910062743573, "learning_rate": 4.382378397452635e-05, "loss": 0.005, "step": 8880 }, { "epoch": 1.7587298446928479, "grad_norm": 0.05689305141298917, "learning_rate": 4.3709588359060884e-05, "loss": 0.0051, "step": 8890 }, { "epoch": 1.7607082797507172, "grad_norm": 0.026010881723224247, "learning_rate": 4.359542607400051e-05, "loss": 0.0051, "step": 8900 }, { "epoch": 1.7626867148085865, "grad_norm": 0.02145317863570537, "learning_rate": 4.34812977242461e-05, "loss": 0.0051, "step": 8910 }, { "epoch": 1.7646651498664556, "grad_norm": 0.02809194558249643, "learning_rate": 4.336720391451864e-05, "loss": 0.0051, "step": 8920 }, { "epoch": 1.7666435849243247, "grad_norm": 0.019370471881264263, "learning_rate": 4.325314524935614e-05, "loss": 0.005, "step": 8930 }, { "epoch": 1.768622019982194, "grad_norm": 0.02780342142236549, "learning_rate": 4.3139122333110406e-05, "loss": 0.005, "step": 8940 }, { "epoch": 1.7706004550400634, "grad_norm": 0.024920458831282365, "learning_rate": 4.302513576994378e-05, "loss": 0.0051, "step": 8950 }, { "epoch": 1.7725788900979325, "grad_norm": 0.01705026947719004, "learning_rate": 4.291118616382603e-05, "loss": 0.0051, "step": 8960 }, { "epoch": 1.7745573251558018, "grad_norm": 0.014632692046212694, "learning_rate": 4.279727411853108e-05, "loss": 0.005, "step": 8970 }, { "epoch": 1.776535760213671, "grad_norm": 0.024436583499265307, "learning_rate": 4.268340023763381e-05, "loss": 0.005, "step": 8980 }, { "epoch": 1.7785141952715402, "grad_norm": 0.011906148319669288, "learning_rate": 4.256956512450696e-05, "loss": 0.005, "step": 8990 }, { "epoch": 1.7804926303294093, "grad_norm": 0.0237424528148114, "learning_rate": 4.2455769382317756e-05, "loss": 0.005, "step": 9000 }, { "epoch": 1.7824710653872786, "grad_norm": 0.011843464895126344, "learning_rate": 4.234201361402488e-05, "loss": 0.0049, "step": 9010 }, { "epoch": 1.784449500445148, "grad_norm": 0.010988029964509503, "learning_rate": 4.2228298422375166e-05, "loss": 0.005, "step": 9020 }, { "epoch": 1.786427935503017, "grad_norm": 0.01941413104339946, "learning_rate": 4.21146244099005e-05, "loss": 0.005, "step": 9030 }, { "epoch": 1.7884063705608864, "grad_norm": 0.01954790471491711, "learning_rate": 4.2000992178914546e-05, "loss": 0.0049, "step": 9040 }, { "epoch": 1.7903848056187557, "grad_norm": 0.018723549260691984, "learning_rate": 4.188740233150956e-05, "loss": 0.005, "step": 9050 }, { "epoch": 1.7923632406766248, "grad_norm": 0.027182370822105752, "learning_rate": 4.177385546955328e-05, "loss": 0.0049, "step": 9060 }, { "epoch": 1.7943416757344939, "grad_norm": 0.009635418273728323, "learning_rate": 4.166035219468562e-05, "loss": 0.005, "step": 9070 }, { "epoch": 1.7963201107923632, "grad_norm": 0.017751590284168492, "learning_rate": 4.154689310831558e-05, "loss": 0.0049, "step": 9080 }, { "epoch": 1.7982985458502325, "grad_norm": 0.014414229973402207, "learning_rate": 4.1433478811618034e-05, "loss": 0.0049, "step": 9090 }, { "epoch": 1.8002769809081016, "grad_norm": 0.010865515740287669, "learning_rate": 4.13201099055305e-05, "loss": 0.005, "step": 9100 }, { "epoch": 1.802255415965971, "grad_norm": 0.011044959419167263, "learning_rate": 4.1206786990750005e-05, "loss": 0.0049, "step": 9110 }, { "epoch": 1.8042338510238403, "grad_norm": 0.018354180646890848, "learning_rate": 4.1093510667729906e-05, "loss": 0.005, "step": 9120 }, { "epoch": 1.8062122860817094, "grad_norm": 0.016577865875752655, "learning_rate": 4.098028153667666e-05, "loss": 0.0049, "step": 9130 }, { "epoch": 1.8081907211395785, "grad_norm": 0.031766982336052546, "learning_rate": 4.086710019754665e-05, "loss": 0.0049, "step": 9140 }, { "epoch": 1.8101691561974478, "grad_norm": 0.03367755758353847, "learning_rate": 4.075396725004308e-05, "loss": 0.005, "step": 9150 }, { "epoch": 1.812147591255317, "grad_norm": 0.016761724534782127, "learning_rate": 4.064088329361274e-05, "loss": 0.005, "step": 9160 }, { "epoch": 1.8141260263131862, "grad_norm": 0.011106442361591966, "learning_rate": 4.05278489274428e-05, "loss": 0.005, "step": 9170 }, { "epoch": 1.8161044613710555, "grad_norm": 0.0250782950939159, "learning_rate": 4.041486475045769e-05, "loss": 0.0049, "step": 9180 }, { "epoch": 1.8180828964289248, "grad_norm": 0.006910355474365653, "learning_rate": 4.0301931361315937e-05, "loss": 0.005, "step": 9190 }, { "epoch": 1.820061331486794, "grad_norm": 0.016513473982056808, "learning_rate": 4.0189049358406906e-05, "loss": 0.0049, "step": 9200 }, { "epoch": 1.822039766544663, "grad_norm": 0.012204809660953057, "learning_rate": 4.007621933984773e-05, "loss": 0.005, "step": 9210 }, { "epoch": 1.8240182016025324, "grad_norm": 0.011967354725089347, "learning_rate": 3.996344190348007e-05, "loss": 0.0049, "step": 9220 }, { "epoch": 1.8259966366604017, "grad_norm": 0.006167104031920807, "learning_rate": 3.985071764686701e-05, "loss": 0.0049, "step": 9230 }, { "epoch": 1.8279750717182708, "grad_norm": 0.00707501949224515, "learning_rate": 3.973804716728981e-05, "loss": 0.0049, "step": 9240 }, { "epoch": 1.82995350677614, "grad_norm": 0.018138963709493968, "learning_rate": 3.962543106174483e-05, "loss": 0.0049, "step": 9250 }, { "epoch": 1.8319319418340094, "grad_norm": 0.01951639828060566, "learning_rate": 3.951286992694029e-05, "loss": 0.0049, "step": 9260 }, { "epoch": 1.8339103768918785, "grad_norm": 0.010540475407839318, "learning_rate": 3.940036435929318e-05, "loss": 0.005, "step": 9270 }, { "epoch": 1.8358888119497476, "grad_norm": 0.012982147903557634, "learning_rate": 3.928791495492601e-05, "loss": 0.005, "step": 9280 }, { "epoch": 1.837867247007617, "grad_norm": 0.008643530326982863, "learning_rate": 3.9175522309663724e-05, "loss": 0.0049, "step": 9290 }, { "epoch": 1.8398456820654863, "grad_norm": 0.009359217032841021, "learning_rate": 3.906318701903055e-05, "loss": 0.0049, "step": 9300 }, { "epoch": 1.8418241171233554, "grad_norm": 0.02060071123665169, "learning_rate": 3.8950909678246784e-05, "loss": 0.005, "step": 9310 }, { "epoch": 1.8438025521812247, "grad_norm": 0.0140322239307105, "learning_rate": 3.883869088222568e-05, "loss": 0.005, "step": 9320 }, { "epoch": 1.845780987239094, "grad_norm": 0.014222205704974178, "learning_rate": 3.872653122557028e-05, "loss": 0.0049, "step": 9330 }, { "epoch": 1.847759422296963, "grad_norm": 0.017102931290158593, "learning_rate": 3.861443130257031e-05, "loss": 0.0049, "step": 9340 }, { "epoch": 1.8497378573548322, "grad_norm": 0.01992360298803752, "learning_rate": 3.850239170719891e-05, "loss": 0.005, "step": 9350 }, { "epoch": 1.8517162924127015, "grad_norm": 0.011049410230869512, "learning_rate": 3.8390413033109644e-05, "loss": 0.0049, "step": 9360 }, { "epoch": 1.8536947274705708, "grad_norm": 0.012688780873361064, "learning_rate": 3.827849587363323e-05, "loss": 0.005, "step": 9370 }, { "epoch": 1.85567316252844, "grad_norm": 0.005326913034240854, "learning_rate": 3.816664082177447e-05, "loss": 0.005, "step": 9380 }, { "epoch": 1.8576515975863093, "grad_norm": 0.02892923733504959, "learning_rate": 3.8054848470209095e-05, "loss": 0.005, "step": 9390 }, { "epoch": 1.8596300326441786, "grad_norm": 0.013321593100630264, "learning_rate": 3.794311941128055e-05, "loss": 0.005, "step": 9400 }, { "epoch": 1.8616084677020477, "grad_norm": 0.012239520740623264, "learning_rate": 3.783145423699701e-05, "loss": 0.0049, "step": 9410 }, { "epoch": 1.8635869027599168, "grad_norm": 0.01624110084692981, "learning_rate": 3.771985353902805e-05, "loss": 0.005, "step": 9420 }, { "epoch": 1.865565337817786, "grad_norm": 0.022475732988488623, "learning_rate": 3.7608317908701706e-05, "loss": 0.0049, "step": 9430 }, { "epoch": 1.8675437728756554, "grad_norm": 0.018995515985658698, "learning_rate": 3.749684793700115e-05, "loss": 0.0049, "step": 9440 }, { "epoch": 1.8695222079335245, "grad_norm": 0.008687968883649247, "learning_rate": 3.738544421456175e-05, "loss": 0.0049, "step": 9450 }, { "epoch": 1.8715006429913938, "grad_norm": 0.008380597543778895, "learning_rate": 3.727410733166779e-05, "loss": 0.0049, "step": 9460 }, { "epoch": 1.8734790780492632, "grad_norm": 0.017898913806399033, "learning_rate": 3.716283787824939e-05, "loss": 0.0049, "step": 9470 }, { "epoch": 1.8754575131071323, "grad_norm": 0.012327701093323443, "learning_rate": 3.7051636443879425e-05, "loss": 0.0049, "step": 9480 }, { "epoch": 1.8774359481650014, "grad_norm": 0.01449448586891039, "learning_rate": 3.694050361777034e-05, "loss": 0.0049, "step": 9490 }, { "epoch": 1.8794143832228707, "grad_norm": 0.01816085515329901, "learning_rate": 3.682943998877105e-05, "loss": 0.0049, "step": 9500 }, { "epoch": 1.88139281828074, "grad_norm": 0.011030098817933364, "learning_rate": 3.6718446145363854e-05, "loss": 0.0051, "step": 9510 }, { "epoch": 1.883371253338609, "grad_norm": 0.021027755236505564, "learning_rate": 3.660752267566123e-05, "loss": 0.005, "step": 9520 }, { "epoch": 1.8853496883964784, "grad_norm": 0.02395104323402294, "learning_rate": 3.649667016740281e-05, "loss": 0.0052, "step": 9530 }, { "epoch": 1.8873281234543478, "grad_norm": 0.039044672469468904, "learning_rate": 3.638588920795225e-05, "loss": 0.005, "step": 9540 }, { "epoch": 1.8893065585122168, "grad_norm": 0.01635197064260792, "learning_rate": 3.627518038429403e-05, "loss": 0.0049, "step": 9550 }, { "epoch": 1.891284993570086, "grad_norm": 0.011813375780556735, "learning_rate": 3.61645442830305e-05, "loss": 0.0049, "step": 9560 }, { "epoch": 1.8932634286279553, "grad_norm": 0.012569298315875155, "learning_rate": 3.6053981490378577e-05, "loss": 0.005, "step": 9570 }, { "epoch": 1.8952418636858246, "grad_norm": 0.018653679675960082, "learning_rate": 3.5943492592166836e-05, "loss": 0.0049, "step": 9580 }, { "epoch": 1.8972202987436937, "grad_norm": 0.01863963283717833, "learning_rate": 3.5833078173832256e-05, "loss": 0.0049, "step": 9590 }, { "epoch": 1.899198733801563, "grad_norm": 0.01606051040828435, "learning_rate": 3.572273882041723e-05, "loss": 0.0049, "step": 9600 }, { "epoch": 1.9011771688594323, "grad_norm": 0.009014670413311715, "learning_rate": 3.561247511656637e-05, "loss": 0.0049, "step": 9610 }, { "epoch": 1.9031556039173014, "grad_norm": 0.005669079199557325, "learning_rate": 3.5502287646523455e-05, "loss": 0.005, "step": 9620 }, { "epoch": 1.9051340389751705, "grad_norm": 0.015934695599857506, "learning_rate": 3.5392176994128354e-05, "loss": 0.0049, "step": 9630 }, { "epoch": 1.9071124740330399, "grad_norm": 0.019284645621864032, "learning_rate": 3.528214374281388e-05, "loss": 0.0049, "step": 9640 }, { "epoch": 1.9090909090909092, "grad_norm": 0.01661569686241377, "learning_rate": 3.517218847560274e-05, "loss": 0.0049, "step": 9650 }, { "epoch": 1.9110693441487783, "grad_norm": 0.012127525310201333, "learning_rate": 3.506231177510446e-05, "loss": 0.005, "step": 9660 }, { "epoch": 1.9130477792066476, "grad_norm": 0.014575058388823363, "learning_rate": 3.4952514223512235e-05, "loss": 0.005, "step": 9670 }, { "epoch": 1.915026214264517, "grad_norm": 0.003064086759791757, "learning_rate": 3.484279640259989e-05, "loss": 0.005, "step": 9680 }, { "epoch": 1.917004649322386, "grad_norm": 0.014318392162680205, "learning_rate": 3.473315889371881e-05, "loss": 0.005, "step": 9690 }, { "epoch": 1.918983084380255, "grad_norm": 0.020467164881671208, "learning_rate": 3.462360227779482e-05, "loss": 0.005, "step": 9700 }, { "epoch": 1.9209615194381244, "grad_norm": 0.016404942786163595, "learning_rate": 3.4514127135325103e-05, "loss": 0.0049, "step": 9710 }, { "epoch": 1.9229399544959938, "grad_norm": 0.016665347448256065, "learning_rate": 3.440473404637519e-05, "loss": 0.0049, "step": 9720 }, { "epoch": 1.9249183895538629, "grad_norm": 0.009492326400500326, "learning_rate": 3.4295423590575826e-05, "loss": 0.0049, "step": 9730 }, { "epoch": 1.9268968246117322, "grad_norm": 0.012063419084035692, "learning_rate": 3.4186196347119906e-05, "loss": 0.005, "step": 9740 }, { "epoch": 1.9288752596696015, "grad_norm": 0.01653986654076417, "learning_rate": 3.407705289475942e-05, "loss": 0.0049, "step": 9750 }, { "epoch": 1.9308536947274706, "grad_norm": 0.014158432284028726, "learning_rate": 3.3967993811802406e-05, "loss": 0.0049, "step": 9760 }, { "epoch": 1.9328321297853397, "grad_norm": 0.014801155851601893, "learning_rate": 3.385901967610982e-05, "loss": 0.0049, "step": 9770 }, { "epoch": 1.934810564843209, "grad_norm": 0.020138881320358355, "learning_rate": 3.375013106509254e-05, "loss": 0.0049, "step": 9780 }, { "epoch": 1.9367889999010783, "grad_norm": 0.005775955316403096, "learning_rate": 3.364132855570829e-05, "loss": 0.0049, "step": 9790 }, { "epoch": 1.9387674349589474, "grad_norm": 0.015870019015717367, "learning_rate": 3.353261272445854e-05, "loss": 0.005, "step": 9800 }, { "epoch": 1.9407458700168165, "grad_norm": 0.019882625928423242, "learning_rate": 3.342398414738555e-05, "loss": 0.005, "step": 9810 }, { "epoch": 1.942724305074686, "grad_norm": 0.013650719055648913, "learning_rate": 3.3315443400069186e-05, "loss": 0.005, "step": 9820 }, { "epoch": 1.9447027401325552, "grad_norm": 0.012836048441123401, "learning_rate": 3.3206991057623974e-05, "loss": 0.0049, "step": 9830 }, { "epoch": 1.9466811751904243, "grad_norm": 0.013360820008600078, "learning_rate": 3.3098627694696046e-05, "loss": 0.0049, "step": 9840 }, { "epoch": 1.9486596102482936, "grad_norm": 0.014935975291671814, "learning_rate": 3.299035388545999e-05, "loss": 0.0049, "step": 9850 }, { "epoch": 1.950638045306163, "grad_norm": 0.01789013838796938, "learning_rate": 3.288217020361594e-05, "loss": 0.0049, "step": 9860 }, { "epoch": 1.952616480364032, "grad_norm": 0.013736476553476436, "learning_rate": 3.277407722238646e-05, "loss": 0.0049, "step": 9870 }, { "epoch": 1.9545949154219011, "grad_norm": 0.00448479749733958, "learning_rate": 3.266607551451357e-05, "loss": 0.0049, "step": 9880 }, { "epoch": 1.9565733504797707, "grad_norm": 0.013783039713030023, "learning_rate": 3.255816565225559e-05, "loss": 0.0049, "step": 9890 }, { "epoch": 1.9585517855376398, "grad_norm": 0.011344086111911986, "learning_rate": 3.245034820738426e-05, "loss": 0.0049, "step": 9900 }, { "epoch": 1.9605302205955089, "grad_norm": 0.013015564395883795, "learning_rate": 3.234262375118161e-05, "loss": 0.0049, "step": 9910 }, { "epoch": 1.9625086556533782, "grad_norm": 0.016826697511562384, "learning_rate": 3.223499285443694e-05, "loss": 0.005, "step": 9920 }, { "epoch": 1.9644870907112475, "grad_norm": 0.006584899924699131, "learning_rate": 3.212745608744385e-05, "loss": 0.0049, "step": 9930 }, { "epoch": 1.9664655257691166, "grad_norm": 0.027455782735791412, "learning_rate": 3.202001401999715e-05, "loss": 0.0049, "step": 9940 }, { "epoch": 1.9684439608269857, "grad_norm": 0.017991916130041037, "learning_rate": 3.191266722138989e-05, "loss": 0.0049, "step": 9950 }, { "epoch": 1.9704223958848552, "grad_norm": 0.015476323051794756, "learning_rate": 3.180541626041035e-05, "loss": 0.0049, "step": 9960 }, { "epoch": 1.9724008309427243, "grad_norm": 0.012359998344034548, "learning_rate": 3.169826170533896e-05, "loss": 0.0049, "step": 9970 }, { "epoch": 1.9743792660005934, "grad_norm": 0.027990942326033677, "learning_rate": 3.1591204123945383e-05, "loss": 0.0049, "step": 9980 }, { "epoch": 1.9763577010584628, "grad_norm": 0.008097200609733055, "learning_rate": 3.1484244083485396e-05, "loss": 0.005, "step": 9990 }, { "epoch": 1.978336136116332, "grad_norm": 0.010714357827724104, "learning_rate": 3.137738215069801e-05, "loss": 0.0049, "step": 10000 }, { "epoch": 1.9803145711742012, "grad_norm": 0.013861333521204026, "learning_rate": 3.127061889180235e-05, "loss": 0.005, "step": 10010 }, { "epoch": 1.9822930062320703, "grad_norm": 0.021013556963888597, "learning_rate": 3.1163954872494726e-05, "loss": 0.0049, "step": 10020 }, { "epoch": 1.9842714412899398, "grad_norm": 0.023032431321792685, "learning_rate": 3.1057390657945653e-05, "loss": 0.0049, "step": 10030 }, { "epoch": 1.986249876347809, "grad_norm": 0.02311298677513279, "learning_rate": 3.095092681279674e-05, "loss": 0.0049, "step": 10040 }, { "epoch": 1.988228311405678, "grad_norm": 0.014153470523932454, "learning_rate": 3.084456390115786e-05, "loss": 0.005, "step": 10050 }, { "epoch": 1.9902067464635473, "grad_norm": 0.017992715560194567, "learning_rate": 3.0738302486604045e-05, "loss": 0.0049, "step": 10060 }, { "epoch": 1.9921851815214167, "grad_norm": 0.017207085901280433, "learning_rate": 3.0632143132172504e-05, "loss": 0.0049, "step": 10070 }, { "epoch": 1.9941636165792858, "grad_norm": 0.019727043631419708, "learning_rate": 3.052608640035972e-05, "loss": 0.0049, "step": 10080 }, { "epoch": 1.9961420516371549, "grad_norm": 0.009506835377934682, "learning_rate": 3.042013285311839e-05, "loss": 0.0049, "step": 10090 }, { "epoch": 1.9981204866950244, "grad_norm": 0.009360806161705221, "learning_rate": 3.0314283051854485e-05, "loss": 0.0049, "step": 10100 }, { "epoch": 2.0, "grad_norm": 0.012575335946312303, "learning_rate": 3.0208537557424277e-05, "loss": 0.0049, "step": 10110 }, { "epoch": 2.001978435057869, "grad_norm": 0.013941044371290949, "learning_rate": 3.0102896930131354e-05, "loss": 0.0049, "step": 10120 }, { "epoch": 2.0039568701157386, "grad_norm": 0.021306355907437013, "learning_rate": 2.9997361729723606e-05, "loss": 0.005, "step": 10130 }, { "epoch": 2.0059353051736077, "grad_norm": 0.010711789163300268, "learning_rate": 2.9891932515390382e-05, "loss": 0.0049, "step": 10140 }, { "epoch": 2.007913740231477, "grad_norm": 0.011287639768519818, "learning_rate": 2.9786609845759418e-05, "loss": 0.0049, "step": 10150 }, { "epoch": 2.009892175289346, "grad_norm": 0.024588158963295925, "learning_rate": 2.9681394278893903e-05, "loss": 0.005, "step": 10160 }, { "epoch": 2.0118706103472155, "grad_norm": 0.020806940868060404, "learning_rate": 2.9576286372289542e-05, "loss": 0.0049, "step": 10170 }, { "epoch": 2.0138490454050846, "grad_norm": 0.00583345103161499, "learning_rate": 2.947128668287161e-05, "loss": 0.0049, "step": 10180 }, { "epoch": 2.0158274804629537, "grad_norm": 0.0197551121452667, "learning_rate": 2.936639576699194e-05, "loss": 0.0049, "step": 10190 }, { "epoch": 2.0178059155208232, "grad_norm": 0.006520152222178015, "learning_rate": 2.9261614180426034e-05, "loss": 0.0049, "step": 10200 }, { "epoch": 2.0197843505786923, "grad_norm": 0.014101510383200082, "learning_rate": 2.915694247837014e-05, "loss": 0.0049, "step": 10210 }, { "epoch": 2.0217627856365614, "grad_norm": 0.008292058057169932, "learning_rate": 2.90523812154382e-05, "loss": 0.0049, "step": 10220 }, { "epoch": 2.0237412206944305, "grad_norm": 0.007226852882654017, "learning_rate": 2.894793094565904e-05, "loss": 0.0049, "step": 10230 }, { "epoch": 2.0257196557523, "grad_norm": 0.011509519347663555, "learning_rate": 2.8843592222473354e-05, "loss": 0.0049, "step": 10240 }, { "epoch": 2.027698090810169, "grad_norm": 0.014355986920592891, "learning_rate": 2.8739365598730818e-05, "loss": 0.0049, "step": 10250 }, { "epoch": 2.0296765258680383, "grad_norm": 0.00788986238196745, "learning_rate": 2.8635251626687076e-05, "loss": 0.0049, "step": 10260 }, { "epoch": 2.031654960925908, "grad_norm": 0.012012777088431013, "learning_rate": 2.853125085800096e-05, "loss": 0.0049, "step": 10270 }, { "epoch": 2.033633395983777, "grad_norm": 0.02383439861059609, "learning_rate": 2.8427363843731393e-05, "loss": 0.0049, "step": 10280 }, { "epoch": 2.035611831041646, "grad_norm": 0.020173838199515726, "learning_rate": 2.832359113433463e-05, "loss": 0.0049, "step": 10290 }, { "epoch": 2.037590266099515, "grad_norm": 0.006241970776659051, "learning_rate": 2.8219933279661247e-05, "loss": 0.005, "step": 10300 }, { "epoch": 2.0395687011573846, "grad_norm": 0.007210399338500074, "learning_rate": 2.8116390828953253e-05, "loss": 0.0049, "step": 10310 }, { "epoch": 2.0415471362152537, "grad_norm": 0.01027563592050959, "learning_rate": 2.801296433084115e-05, "loss": 0.0049, "step": 10320 }, { "epoch": 2.043525571273123, "grad_norm": 0.0152798170546597, "learning_rate": 2.7909654333341084e-05, "loss": 0.0049, "step": 10330 }, { "epoch": 2.0455040063309924, "grad_norm": 0.018626466674510334, "learning_rate": 2.7806461383851895e-05, "loss": 0.0049, "step": 10340 }, { "epoch": 2.0474824413888615, "grad_norm": 0.01302839073947395, "learning_rate": 2.7703386029152246e-05, "loss": 0.0049, "step": 10350 }, { "epoch": 2.0494608764467306, "grad_norm": 0.007291290463857292, "learning_rate": 2.76004288153977e-05, "loss": 0.0049, "step": 10360 }, { "epoch": 2.0514393115045997, "grad_norm": 0.0070181339392829745, "learning_rate": 2.749759028811786e-05, "loss": 0.0049, "step": 10370 }, { "epoch": 2.0534177465624692, "grad_norm": 0.0028466445302643456, "learning_rate": 2.7394870992213394e-05, "loss": 0.0049, "step": 10380 }, { "epoch": 2.0553961816203383, "grad_norm": 0.01900130612247899, "learning_rate": 2.7292271471953284e-05, "loss": 0.0049, "step": 10390 }, { "epoch": 2.0573746166782074, "grad_norm": 0.011150432897110945, "learning_rate": 2.7189792270971827e-05, "loss": 0.0049, "step": 10400 }, { "epoch": 2.059353051736077, "grad_norm": 0.006067838560177696, "learning_rate": 2.7087433932265825e-05, "loss": 0.0049, "step": 10410 }, { "epoch": 2.061331486793946, "grad_norm": 0.010875343636820188, "learning_rate": 2.698519699819163e-05, "loss": 0.0049, "step": 10420 }, { "epoch": 2.063309921851815, "grad_norm": 0.007969782655879313, "learning_rate": 2.688308201046236e-05, "loss": 0.0049, "step": 10430 }, { "epoch": 2.0652883569096843, "grad_norm": 0.010318739676634077, "learning_rate": 2.6781089510144996e-05, "loss": 0.0049, "step": 10440 }, { "epoch": 2.067266791967554, "grad_norm": 0.014941766358376614, "learning_rate": 2.667922003765745e-05, "loss": 0.0049, "step": 10450 }, { "epoch": 2.069245227025423, "grad_norm": 0.01014366606053396, "learning_rate": 2.657747413276581e-05, "loss": 0.0049, "step": 10460 }, { "epoch": 2.071223662083292, "grad_norm": 0.01570714720191582, "learning_rate": 2.647585233458142e-05, "loss": 0.0048, "step": 10470 }, { "epoch": 2.0732020971411615, "grad_norm": 0.0076638855412125, "learning_rate": 2.637435518155802e-05, "loss": 0.0049, "step": 10480 }, { "epoch": 2.0751805321990306, "grad_norm": 0.008069808196170728, "learning_rate": 2.627298321148893e-05, "loss": 0.0049, "step": 10490 }, { "epoch": 2.0771589672568997, "grad_norm": 0.006758835042487589, "learning_rate": 2.617173696150411e-05, "loss": 0.0048, "step": 10500 }, { "epoch": 2.079137402314769, "grad_norm": 0.0042150156912977, "learning_rate": 2.6070616968067445e-05, "loss": 0.0049, "step": 10510 }, { "epoch": 2.0811158373726384, "grad_norm": 0.009664860614324406, "learning_rate": 2.5969623766973815e-05, "loss": 0.0049, "step": 10520 }, { "epoch": 2.0830942724305075, "grad_norm": 0.006663867350119711, "learning_rate": 2.5868757893346284e-05, "loss": 0.0048, "step": 10530 }, { "epoch": 2.0850727074883766, "grad_norm": 0.005683282754922332, "learning_rate": 2.5768019881633237e-05, "loss": 0.0049, "step": 10540 }, { "epoch": 2.087051142546246, "grad_norm": 0.026149506234740973, "learning_rate": 2.5667410265605617e-05, "loss": 0.0049, "step": 10550 }, { "epoch": 2.0890295776041152, "grad_norm": 0.017561508949794108, "learning_rate": 2.556692957835398e-05, "loss": 0.0049, "step": 10560 }, { "epoch": 2.0910080126619843, "grad_norm": 0.015149893340226575, "learning_rate": 2.5466578352285774e-05, "loss": 0.0049, "step": 10570 }, { "epoch": 2.0929864477198534, "grad_norm": 0.008455563318015342, "learning_rate": 2.5366357119122486e-05, "loss": 0.0049, "step": 10580 }, { "epoch": 2.094964882777723, "grad_norm": 0.005503673626861871, "learning_rate": 2.5266266409896834e-05, "loss": 0.0049, "step": 10590 }, { "epoch": 2.096943317835592, "grad_norm": 0.012354362571497529, "learning_rate": 2.5166306754949908e-05, "loss": 0.0049, "step": 10600 }, { "epoch": 2.098921752893461, "grad_norm": 0.011856551646159496, "learning_rate": 2.5066478683928457e-05, "loss": 0.005, "step": 10610 }, { "epoch": 2.1009001879513307, "grad_norm": 0.01273108380835395, "learning_rate": 2.4966782725781916e-05, "loss": 0.0049, "step": 10620 }, { "epoch": 2.1028786230092, "grad_norm": 0.01145778259498361, "learning_rate": 2.4867219408759794e-05, "loss": 0.0049, "step": 10630 }, { "epoch": 2.104857058067069, "grad_norm": 0.016182509446048304, "learning_rate": 2.4767789260408753e-05, "loss": 0.0049, "step": 10640 }, { "epoch": 2.106835493124938, "grad_norm": 0.011174888771408948, "learning_rate": 2.466849280756985e-05, "loss": 0.0049, "step": 10650 }, { "epoch": 2.1088139281828076, "grad_norm": 0.0130954542738303, "learning_rate": 2.4569330576375743e-05, "loss": 0.0049, "step": 10660 }, { "epoch": 2.1107923632406767, "grad_norm": 0.012141587563539374, "learning_rate": 2.4470303092247926e-05, "loss": 0.0049, "step": 10670 }, { "epoch": 2.1127707982985457, "grad_norm": 0.008166994941807925, "learning_rate": 2.4371410879893857e-05, "loss": 0.0049, "step": 10680 }, { "epoch": 2.114749233356415, "grad_norm": 0.04892731219443696, "learning_rate": 2.4272654463304328e-05, "loss": 0.0049, "step": 10690 }, { "epoch": 2.1167276684142844, "grad_norm": 0.02210234781975422, "learning_rate": 2.4174034365750513e-05, "loss": 0.0049, "step": 10700 }, { "epoch": 2.1187061034721535, "grad_norm": 0.014818306143350475, "learning_rate": 2.4075551109781358e-05, "loss": 0.0049, "step": 10710 }, { "epoch": 2.1206845385300226, "grad_norm": 0.014542060408597218, "learning_rate": 2.3977205217220707e-05, "loss": 0.0049, "step": 10720 }, { "epoch": 2.122662973587892, "grad_norm": 0.01406469356339733, "learning_rate": 2.387899720916459e-05, "loss": 0.0049, "step": 10730 }, { "epoch": 2.1246414086457612, "grad_norm": 0.009606607156167207, "learning_rate": 2.3780927605978397e-05, "loss": 0.0049, "step": 10740 }, { "epoch": 2.1266198437036303, "grad_norm": 0.003715932480320162, "learning_rate": 2.3682996927294214e-05, "loss": 0.0049, "step": 10750 }, { "epoch": 2.1285982787614994, "grad_norm": 0.011448115872551134, "learning_rate": 2.3585205692008e-05, "loss": 0.0049, "step": 10760 }, { "epoch": 2.130576713819369, "grad_norm": 0.010931701386815025, "learning_rate": 2.348755441827686e-05, "loss": 0.0049, "step": 10770 }, { "epoch": 2.132555148877238, "grad_norm": 0.012760652235348856, "learning_rate": 2.3390043623516293e-05, "loss": 0.0049, "step": 10780 }, { "epoch": 2.134533583935107, "grad_norm": 0.0045646344507789816, "learning_rate": 2.3292673824397487e-05, "loss": 0.0049, "step": 10790 }, { "epoch": 2.1365120189929767, "grad_norm": 0.008782734561312273, "learning_rate": 2.319544553684449e-05, "loss": 0.0049, "step": 10800 }, { "epoch": 2.138490454050846, "grad_norm": 0.012163885468131179, "learning_rate": 2.3098359276031578e-05, "loss": 0.0049, "step": 10810 }, { "epoch": 2.140468889108715, "grad_norm": 0.013599964829061955, "learning_rate": 2.3001415556380484e-05, "loss": 0.0049, "step": 10820 }, { "epoch": 2.142447324166584, "grad_norm": 0.004114916172487493, "learning_rate": 2.290461489155768e-05, "loss": 0.0049, "step": 10830 }, { "epoch": 2.1444257592244536, "grad_norm": 0.00880311413802491, "learning_rate": 2.280795779447159e-05, "loss": 0.0049, "step": 10840 }, { "epoch": 2.1464041942823227, "grad_norm": 0.009700421990048027, "learning_rate": 2.271144477726999e-05, "loss": 0.0049, "step": 10850 }, { "epoch": 2.1483826293401918, "grad_norm": 0.006270591813073498, "learning_rate": 2.261507635133724e-05, "loss": 0.0049, "step": 10860 }, { "epoch": 2.1503610643980613, "grad_norm": 0.012081342163184194, "learning_rate": 2.2518853027291487e-05, "loss": 0.0049, "step": 10870 }, { "epoch": 2.1523394994559304, "grad_norm": 0.00904387193997516, "learning_rate": 2.242277531498212e-05, "loss": 0.0049, "step": 10880 }, { "epoch": 2.1543179345137995, "grad_norm": 0.0075131288365753056, "learning_rate": 2.2326843723486973e-05, "loss": 0.0049, "step": 10890 }, { "epoch": 2.1562963695716686, "grad_norm": 0.008542159300405496, "learning_rate": 2.223105876110962e-05, "loss": 0.0049, "step": 10900 }, { "epoch": 2.158274804629538, "grad_norm": 0.0335764787739486, "learning_rate": 2.213542093537675e-05, "loss": 0.0049, "step": 10910 }, { "epoch": 2.1602532396874072, "grad_norm": 0.001990531513768214, "learning_rate": 2.2039930753035354e-05, "loss": 0.0048, "step": 10920 }, { "epoch": 2.1622316747452763, "grad_norm": 0.016309185283180646, "learning_rate": 2.1944588720050197e-05, "loss": 0.0049, "step": 10930 }, { "epoch": 2.164210109803146, "grad_norm": 0.013924872107033366, "learning_rate": 2.1849395341601026e-05, "loss": 0.0049, "step": 10940 }, { "epoch": 2.166188544861015, "grad_norm": 0.00572948993312922, "learning_rate": 2.1754351122079923e-05, "loss": 0.0049, "step": 10950 }, { "epoch": 2.168166979918884, "grad_norm": 0.011942250361249152, "learning_rate": 2.1659456565088654e-05, "loss": 0.0049, "step": 10960 }, { "epoch": 2.170145414976753, "grad_norm": 0.012924764236534225, "learning_rate": 2.156471217343598e-05, "loss": 0.0049, "step": 10970 }, { "epoch": 2.1721238500346227, "grad_norm": 0.007958646616792086, "learning_rate": 2.1470118449134974e-05, "loss": 0.0049, "step": 10980 }, { "epoch": 2.174102285092492, "grad_norm": 0.0060741944946902686, "learning_rate": 2.1375675893400372e-05, "loss": 0.0048, "step": 10990 }, { "epoch": 2.176080720150361, "grad_norm": 0.006323049206733025, "learning_rate": 2.128138500664597e-05, "loss": 0.0049, "step": 11000 }, { "epoch": 2.1780591552082305, "grad_norm": 0.00717949129900845, "learning_rate": 2.1187246288481917e-05, "loss": 0.0049, "step": 11010 }, { "epoch": 2.1800375902660996, "grad_norm": 0.007920351394777083, "learning_rate": 2.109326023771206e-05, "loss": 0.0049, "step": 11020 }, { "epoch": 2.1820160253239687, "grad_norm": 0.01440783561812086, "learning_rate": 2.0999427352331362e-05, "loss": 0.0049, "step": 11030 }, { "epoch": 2.1839944603818378, "grad_norm": 0.012739025752214196, "learning_rate": 2.0905748129523166e-05, "loss": 0.0049, "step": 11040 }, { "epoch": 2.1859728954397073, "grad_norm": 0.006014400765555164, "learning_rate": 2.081222306565666e-05, "loss": 0.0049, "step": 11050 }, { "epoch": 2.1879513304975764, "grad_norm": 0.01638106765505794, "learning_rate": 2.0718852656284195e-05, "loss": 0.0049, "step": 11060 }, { "epoch": 2.1899297655554455, "grad_norm": 0.016582172712300853, "learning_rate": 2.0625637396138664e-05, "loss": 0.0049, "step": 11070 }, { "epoch": 2.191908200613315, "grad_norm": 0.010863640478265692, "learning_rate": 2.053257777913089e-05, "loss": 0.0049, "step": 11080 }, { "epoch": 2.193886635671184, "grad_norm": 0.010740731361364763, "learning_rate": 2.043967429834703e-05, "loss": 0.0049, "step": 11090 }, { "epoch": 2.1958650707290532, "grad_norm": 0.011235614778445133, "learning_rate": 2.034692744604586e-05, "loss": 0.0049, "step": 11100 }, { "epoch": 2.1978435057869223, "grad_norm": 0.015395962469702274, "learning_rate": 2.025433771365632e-05, "loss": 0.0049, "step": 11110 }, { "epoch": 2.199821940844792, "grad_norm": 0.006395775696335573, "learning_rate": 2.0161905591774826e-05, "loss": 0.0049, "step": 11120 }, { "epoch": 2.201800375902661, "grad_norm": 0.006831622650523668, "learning_rate": 2.0069631570162624e-05, "loss": 0.0049, "step": 11130 }, { "epoch": 2.20377881096053, "grad_norm": 0.0042288977652911435, "learning_rate": 1.9977516137743302e-05, "loss": 0.0049, "step": 11140 }, { "epoch": 2.2057572460183996, "grad_norm": 0.02581949364447971, "learning_rate": 1.9885559782600126e-05, "loss": 0.0049, "step": 11150 }, { "epoch": 2.2077356810762687, "grad_norm": 0.013927238045077924, "learning_rate": 1.97937629919735e-05, "loss": 0.0049, "step": 11160 }, { "epoch": 2.209714116134138, "grad_norm": 0.0031532293243946847, "learning_rate": 1.97021262522583e-05, "loss": 0.0048, "step": 11170 }, { "epoch": 2.211692551192007, "grad_norm": 0.008346776999060588, "learning_rate": 1.9610650049001407e-05, "loss": 0.0048, "step": 11180 }, { "epoch": 2.2136709862498765, "grad_norm": 0.009163181443904768, "learning_rate": 1.951933486689907e-05, "loss": 0.0048, "step": 11190 }, { "epoch": 2.2156494213077456, "grad_norm": 0.005485348687788032, "learning_rate": 1.9428181189794353e-05, "loss": 0.0049, "step": 11200 }, { "epoch": 2.2176278563656147, "grad_norm": 0.010169597425668853, "learning_rate": 1.9337189500674573e-05, "loss": 0.0049, "step": 11210 }, { "epoch": 2.219606291423484, "grad_norm": 0.0043042636892453015, "learning_rate": 1.9246360281668697e-05, "loss": 0.0048, "step": 11220 }, { "epoch": 2.2215847264813533, "grad_norm": 0.003848310049274622, "learning_rate": 1.9155694014044877e-05, "loss": 0.0049, "step": 11230 }, { "epoch": 2.2235631615392224, "grad_norm": 0.006793887630929106, "learning_rate": 1.9065191178207825e-05, "loss": 0.0048, "step": 11240 }, { "epoch": 2.2255415965970915, "grad_norm": 0.010092825280993978, "learning_rate": 1.8974852253696307e-05, "loss": 0.0048, "step": 11250 }, { "epoch": 2.227520031654961, "grad_norm": 0.006180027293137317, "learning_rate": 1.8884677719180594e-05, "loss": 0.0049, "step": 11260 }, { "epoch": 2.22949846671283, "grad_norm": 0.0026729274803594326, "learning_rate": 1.879466805245986e-05, "loss": 0.0049, "step": 11270 }, { "epoch": 2.2314769017706992, "grad_norm": 0.013385712655096128, "learning_rate": 1.87048237304598e-05, "loss": 0.0049, "step": 11280 }, { "epoch": 2.233455336828569, "grad_norm": 0.006329624208782768, "learning_rate": 1.8615145229229935e-05, "loss": 0.0048, "step": 11290 }, { "epoch": 2.235433771886438, "grad_norm": 0.012087040985852523, "learning_rate": 1.8525633023941208e-05, "loss": 0.0049, "step": 11300 }, { "epoch": 2.237412206944307, "grad_norm": 0.009733786165679728, "learning_rate": 1.8436287588883417e-05, "loss": 0.0049, "step": 11310 }, { "epoch": 2.239390642002176, "grad_norm": 0.00572724887320622, "learning_rate": 1.8347109397462713e-05, "loss": 0.0048, "step": 11320 }, { "epoch": 2.2413690770600456, "grad_norm": 0.016255483917779867, "learning_rate": 1.8258098922199107e-05, "loss": 0.0049, "step": 11330 }, { "epoch": 2.2433475121179147, "grad_norm": 0.013917639324411119, "learning_rate": 1.8169256634723887e-05, "loss": 0.0048, "step": 11340 }, { "epoch": 2.245325947175784, "grad_norm": 0.004589778135368418, "learning_rate": 1.808058300577724e-05, "loss": 0.0049, "step": 11350 }, { "epoch": 2.2473043822336534, "grad_norm": 0.009002097077938054, "learning_rate": 1.7992078505205683e-05, "loss": 0.0049, "step": 11360 }, { "epoch": 2.2492828172915225, "grad_norm": 0.0024830640386171218, "learning_rate": 1.7903743601959577e-05, "loss": 0.0049, "step": 11370 }, { "epoch": 2.2512612523493916, "grad_norm": 0.009735206509308217, "learning_rate": 1.7815578764090674e-05, "loss": 0.0049, "step": 11380 }, { "epoch": 2.2532396874072607, "grad_norm": 0.004728173416438885, "learning_rate": 1.7727584458749608e-05, "loss": 0.0049, "step": 11390 }, { "epoch": 2.25521812246513, "grad_norm": 0.008760238143583239, "learning_rate": 1.7639761152183415e-05, "loss": 0.0048, "step": 11400 }, { "epoch": 2.2571965575229993, "grad_norm": 0.009417011870408052, "learning_rate": 1.7552109309733066e-05, "loss": 0.0049, "step": 11410 }, { "epoch": 2.2591749925808684, "grad_norm": 0.006802278390508995, "learning_rate": 1.746462939583105e-05, "loss": 0.0049, "step": 11420 }, { "epoch": 2.261153427638738, "grad_norm": 0.013570524195058645, "learning_rate": 1.737732187399886e-05, "loss": 0.0049, "step": 11430 }, { "epoch": 2.263131862696607, "grad_norm": 0.004200316753900831, "learning_rate": 1.7290187206844538e-05, "loss": 0.0049, "step": 11440 }, { "epoch": 2.265110297754476, "grad_norm": 0.005723131865755404, "learning_rate": 1.720322585606027e-05, "loss": 0.0049, "step": 11450 }, { "epoch": 2.2670887328123452, "grad_norm": 0.004616765917297674, "learning_rate": 1.711643828241985e-05, "loss": 0.0049, "step": 11460 }, { "epoch": 2.269067167870215, "grad_norm": 0.004569882812925609, "learning_rate": 1.7029824945776345e-05, "loss": 0.0049, "step": 11470 }, { "epoch": 2.271045602928084, "grad_norm": 0.0055773439212271955, "learning_rate": 1.6943386305059593e-05, "loss": 0.0049, "step": 11480 }, { "epoch": 2.273024037985953, "grad_norm": 0.009499254877603265, "learning_rate": 1.6857122818273792e-05, "loss": 0.0048, "step": 11490 }, { "epoch": 2.2750024730438225, "grad_norm": 0.005591547095852957, "learning_rate": 1.6771034942495064e-05, "loss": 0.0049, "step": 11500 }, { "epoch": 2.2769809081016916, "grad_norm": 0.015353527818775333, "learning_rate": 1.6685123133869044e-05, "loss": 0.0049, "step": 11510 }, { "epoch": 2.2789593431595607, "grad_norm": 0.0011053990548616935, "learning_rate": 1.659938784760843e-05, "loss": 0.0048, "step": 11520 }, { "epoch": 2.28093777821743, "grad_norm": 0.011947427639315774, "learning_rate": 1.651382953799063e-05, "loss": 0.0049, "step": 11530 }, { "epoch": 2.2829162132752994, "grad_norm": 0.010797154753672608, "learning_rate": 1.642844865835531e-05, "loss": 0.0049, "step": 11540 }, { "epoch": 2.2848946483331685, "grad_norm": 0.0051180982917368, "learning_rate": 1.634324566110203e-05, "loss": 0.0048, "step": 11550 }, { "epoch": 2.2868730833910376, "grad_norm": 0.006800675035697962, "learning_rate": 1.6258220997687756e-05, "loss": 0.0049, "step": 11560 }, { "epoch": 2.288851518448907, "grad_norm": 0.005325612374829352, "learning_rate": 1.6173375118624594e-05, "loss": 0.0049, "step": 11570 }, { "epoch": 2.290829953506776, "grad_norm": 0.017622792607976806, "learning_rate": 1.6088708473477342e-05, "loss": 0.0049, "step": 11580 }, { "epoch": 2.2928083885646453, "grad_norm": 0.007093754966971193, "learning_rate": 1.6004221510861057e-05, "loss": 0.0049, "step": 11590 }, { "epoch": 2.2947868236225144, "grad_norm": 0.0059812707732046335, "learning_rate": 1.5919914678438776e-05, "loss": 0.0049, "step": 11600 }, { "epoch": 2.296765258680384, "grad_norm": 0.006251732260953764, "learning_rate": 1.583578842291908e-05, "loss": 0.0049, "step": 11610 }, { "epoch": 2.298743693738253, "grad_norm": 0.0023477684198637463, "learning_rate": 1.5751843190053754e-05, "loss": 0.0049, "step": 11620 }, { "epoch": 2.300722128796122, "grad_norm": 0.005772482877827245, "learning_rate": 1.5668079424635424e-05, "loss": 0.0049, "step": 11630 }, { "epoch": 2.3027005638539917, "grad_norm": 0.008687925963376168, "learning_rate": 1.5584497570495143e-05, "loss": 0.0049, "step": 11640 }, { "epoch": 2.304678998911861, "grad_norm": 0.012291788740960144, "learning_rate": 1.5501098070500135e-05, "loss": 0.0049, "step": 11650 }, { "epoch": 2.30665743396973, "grad_norm": 0.001397751722120564, "learning_rate": 1.5417881366551395e-05, "loss": 0.0049, "step": 11660 }, { "epoch": 2.308635869027599, "grad_norm": 0.005997343292672973, "learning_rate": 1.5334847899581346e-05, "loss": 0.0049, "step": 11670 }, { "epoch": 2.3106143040854685, "grad_norm": 0.021624963011345838, "learning_rate": 1.5251998109551508e-05, "loss": 0.0049, "step": 11680 }, { "epoch": 2.3125927391433376, "grad_norm": 0.0148563375587159, "learning_rate": 1.516933243545019e-05, "loss": 0.0049, "step": 11690 }, { "epoch": 2.3145711742012067, "grad_norm": 0.005291109907345976, "learning_rate": 1.5086851315290117e-05, "loss": 0.0049, "step": 11700 }, { "epoch": 2.3165496092590763, "grad_norm": 0.010907174277179854, "learning_rate": 1.5004555186106122e-05, "loss": 0.0048, "step": 11710 }, { "epoch": 2.3185280443169454, "grad_norm": 0.0029041725171997513, "learning_rate": 1.4922444483952886e-05, "loss": 0.0049, "step": 11720 }, { "epoch": 2.3205064793748145, "grad_norm": 0.0023293587077477446, "learning_rate": 1.4840519643902573e-05, "loss": 0.0048, "step": 11730 }, { "epoch": 2.3224849144326836, "grad_norm": 0.00612357980628851, "learning_rate": 1.4758781100042517e-05, "loss": 0.0049, "step": 11740 }, { "epoch": 2.324463349490553, "grad_norm": 0.002160244818194235, "learning_rate": 1.4677229285472988e-05, "loss": 0.0049, "step": 11750 }, { "epoch": 2.326441784548422, "grad_norm": 0.008436757421284993, "learning_rate": 1.459586463230478e-05, "loss": 0.0049, "step": 11760 }, { "epoch": 2.3284202196062913, "grad_norm": 0.011196970905323966, "learning_rate": 1.4514687571657049e-05, "loss": 0.0048, "step": 11770 }, { "epoch": 2.330398654664161, "grad_norm": 0.0072654972776874435, "learning_rate": 1.4433698533654955e-05, "loss": 0.0049, "step": 11780 }, { "epoch": 2.33237708972203, "grad_norm": 0.009557180746201517, "learning_rate": 1.4352897947427396e-05, "loss": 0.0048, "step": 11790 }, { "epoch": 2.334355524779899, "grad_norm": 0.00658138544973912, "learning_rate": 1.4272286241104754e-05, "loss": 0.0048, "step": 11800 }, { "epoch": 2.336333959837768, "grad_norm": 0.008412967785435075, "learning_rate": 1.4191863841816611e-05, "loss": 0.0049, "step": 11810 }, { "epoch": 2.3383123948956377, "grad_norm": 0.008721658831858246, "learning_rate": 1.4111631175689443e-05, "loss": 0.0049, "step": 11820 }, { "epoch": 2.340290829953507, "grad_norm": 0.010238385103802183, "learning_rate": 1.4031588667844476e-05, "loss": 0.0049, "step": 11830 }, { "epoch": 2.342269265011376, "grad_norm": 0.007909078551715566, "learning_rate": 1.3951736742395277e-05, "loss": 0.0048, "step": 11840 }, { "epoch": 2.3442477000692454, "grad_norm": 0.006849652143996244, "learning_rate": 1.3872075822445668e-05, "loss": 0.0049, "step": 11850 }, { "epoch": 2.3462261351271145, "grad_norm": 0.012974772429818504, "learning_rate": 1.3792606330087376e-05, "loss": 0.0049, "step": 11860 }, { "epoch": 2.3482045701849836, "grad_norm": 0.009897585915273335, "learning_rate": 1.371332868639783e-05, "loss": 0.0049, "step": 11870 }, { "epoch": 2.3501830052428527, "grad_norm": 0.006783570958005859, "learning_rate": 1.3634243311437945e-05, "loss": 0.0049, "step": 11880 }, { "epoch": 2.3521614403007223, "grad_norm": 0.0025321485098646597, "learning_rate": 1.355535062424984e-05, "loss": 0.0049, "step": 11890 }, { "epoch": 2.3541398753585914, "grad_norm": 0.005552045014758698, "learning_rate": 1.3476651042854699e-05, "loss": 0.0049, "step": 11900 }, { "epoch": 2.3561183104164605, "grad_norm": 0.006902769224809915, "learning_rate": 1.3398144984250494e-05, "loss": 0.0049, "step": 11910 }, { "epoch": 2.35809674547433, "grad_norm": 0.0033323197124538843, "learning_rate": 1.3319832864409804e-05, "loss": 0.0049, "step": 11920 }, { "epoch": 2.360075180532199, "grad_norm": 0.0033458551784442347, "learning_rate": 1.3241715098277613e-05, "loss": 0.0049, "step": 11930 }, { "epoch": 2.362053615590068, "grad_norm": 0.005517692232340451, "learning_rate": 1.3163792099769067e-05, "loss": 0.0049, "step": 11940 }, { "epoch": 2.3640320506479373, "grad_norm": 0.00456913352621798, "learning_rate": 1.3086064281767346e-05, "loss": 0.0049, "step": 11950 }, { "epoch": 2.366010485705807, "grad_norm": 0.013581702453518283, "learning_rate": 1.3008532056121437e-05, "loss": 0.0049, "step": 11960 }, { "epoch": 2.367988920763676, "grad_norm": 0.006987769083666098, "learning_rate": 1.2931195833643978e-05, "loss": 0.0049, "step": 11970 }, { "epoch": 2.369967355821545, "grad_norm": 0.013864106586640359, "learning_rate": 1.2854056024109013e-05, "loss": 0.0049, "step": 11980 }, { "epoch": 2.3719457908794146, "grad_norm": 0.006378464902308953, "learning_rate": 1.2777113036249928e-05, "loss": 0.0049, "step": 11990 }, { "epoch": 2.3739242259372837, "grad_norm": 0.006052402168879356, "learning_rate": 1.2700367277757209e-05, "loss": 0.0048, "step": 12000 }, { "epoch": 2.375902660995153, "grad_norm": 0.005611012662614181, "learning_rate": 1.2623819155276267e-05, "loss": 0.0048, "step": 12010 }, { "epoch": 2.377881096053022, "grad_norm": 0.011091616052583533, "learning_rate": 1.2547469074405376e-05, "loss": 0.0048, "step": 12020 }, { "epoch": 2.3798595311108914, "grad_norm": 0.0059991229307814935, "learning_rate": 1.2471317439693436e-05, "loss": 0.0049, "step": 12030 }, { "epoch": 2.3818379661687605, "grad_norm": 0.0036471025193809963, "learning_rate": 1.2395364654637853e-05, "loss": 0.0049, "step": 12040 }, { "epoch": 2.3838164012266296, "grad_norm": 0.009916774656531334, "learning_rate": 1.2319611121682439e-05, "loss": 0.0049, "step": 12050 }, { "epoch": 2.385794836284499, "grad_norm": 0.006853651954051589, "learning_rate": 1.224405724221519e-05, "loss": 0.0049, "step": 12060 }, { "epoch": 2.3877732713423683, "grad_norm": 0.00786152654285284, "learning_rate": 1.2168703416566274e-05, "loss": 0.0049, "step": 12070 }, { "epoch": 2.3897517064002374, "grad_norm": 0.005661371944320141, "learning_rate": 1.2093550044005818e-05, "loss": 0.0048, "step": 12080 }, { "epoch": 2.3917301414581065, "grad_norm": 0.006514403726336657, "learning_rate": 1.2018597522741854e-05, "loss": 0.0049, "step": 12090 }, { "epoch": 2.393708576515976, "grad_norm": 0.005352919231840983, "learning_rate": 1.1943846249918162e-05, "loss": 0.0049, "step": 12100 }, { "epoch": 2.395687011573845, "grad_norm": 0.010720840837532983, "learning_rate": 1.1869296621612209e-05, "loss": 0.0049, "step": 12110 }, { "epoch": 2.397665446631714, "grad_norm": 0.004419647264721168, "learning_rate": 1.1794949032832991e-05, "loss": 0.0048, "step": 12120 }, { "epoch": 2.3996438816895838, "grad_norm": 0.007786006688762261, "learning_rate": 1.1720803877518992e-05, "loss": 0.0048, "step": 12130 }, { "epoch": 2.401622316747453, "grad_norm": 0.0031282874209113977, "learning_rate": 1.1646861548536092e-05, "loss": 0.0049, "step": 12140 }, { "epoch": 2.403600751805322, "grad_norm": 0.007667307393383208, "learning_rate": 1.1573122437675466e-05, "loss": 0.0049, "step": 12150 }, { "epoch": 2.405579186863191, "grad_norm": 0.0022204175167783746, "learning_rate": 1.1499586935651529e-05, "loss": 0.0048, "step": 12160 }, { "epoch": 2.4075576219210606, "grad_norm": 0.015539721679331816, "learning_rate": 1.1426255432099848e-05, "loss": 0.0049, "step": 12170 }, { "epoch": 2.4095360569789297, "grad_norm": 0.007904182756752157, "learning_rate": 1.1353128315575052e-05, "loss": 0.0049, "step": 12180 }, { "epoch": 2.411514492036799, "grad_norm": 0.00659239626250241, "learning_rate": 1.128020597354884e-05, "loss": 0.0049, "step": 12190 }, { "epoch": 2.4134929270946683, "grad_norm": 0.014293711219792777, "learning_rate": 1.1207488792407889e-05, "loss": 0.0049, "step": 12200 }, { "epoch": 2.4154713621525374, "grad_norm": 0.007638323156803352, "learning_rate": 1.1134977157451793e-05, "loss": 0.0049, "step": 12210 }, { "epoch": 2.4174497972104065, "grad_norm": 0.004615411554642655, "learning_rate": 1.1062671452891055e-05, "loss": 0.0049, "step": 12220 }, { "epoch": 2.4194282322682756, "grad_norm": 0.006864940517880195, "learning_rate": 1.0990572061845034e-05, "loss": 0.0049, "step": 12230 }, { "epoch": 2.421406667326145, "grad_norm": 0.002295918536425843, "learning_rate": 1.091867936633988e-05, "loss": 0.0048, "step": 12240 }, { "epoch": 2.4233851023840143, "grad_norm": 0.006212537108813713, "learning_rate": 1.0846993747306589e-05, "loss": 0.0049, "step": 12250 }, { "epoch": 2.4253635374418834, "grad_norm": 0.009631024408064383, "learning_rate": 1.0775515584578933e-05, "loss": 0.0048, "step": 12260 }, { "epoch": 2.427341972499753, "grad_norm": 0.0024554663735303077, "learning_rate": 1.0704245256891421e-05, "loss": 0.0049, "step": 12270 }, { "epoch": 2.429320407557622, "grad_norm": 0.005035953573646754, "learning_rate": 1.0633183141877362e-05, "loss": 0.0048, "step": 12280 }, { "epoch": 2.431298842615491, "grad_norm": 0.012735347489310435, "learning_rate": 1.0562329616066824e-05, "loss": 0.0048, "step": 12290 }, { "epoch": 2.43327727767336, "grad_norm": 0.004761102975364903, "learning_rate": 1.0491685054884647e-05, "loss": 0.0048, "step": 12300 }, { "epoch": 2.4352557127312298, "grad_norm": 0.007262890509045137, "learning_rate": 1.0421249832648417e-05, "loss": 0.0049, "step": 12310 }, { "epoch": 2.437234147789099, "grad_norm": 0.002847541800771913, "learning_rate": 1.0351024322566538e-05, "loss": 0.0049, "step": 12320 }, { "epoch": 2.439212582846968, "grad_norm": 0.0029123055612013653, "learning_rate": 1.028100889673625e-05, "loss": 0.0049, "step": 12330 }, { "epoch": 2.4411910179048375, "grad_norm": 0.005597742332676149, "learning_rate": 1.0211203926141604e-05, "loss": 0.0049, "step": 12340 }, { "epoch": 2.4431694529627066, "grad_norm": 0.009055713718602731, "learning_rate": 1.0141609780651584e-05, "loss": 0.0049, "step": 12350 }, { "epoch": 2.4451478880205757, "grad_norm": 0.003452189235684107, "learning_rate": 1.0072226829018016e-05, "loss": 0.0049, "step": 12360 }, { "epoch": 2.447126323078445, "grad_norm": 0.007355438737602597, "learning_rate": 1.0003055438873765e-05, "loss": 0.0049, "step": 12370 }, { "epoch": 2.4491047581363143, "grad_norm": 0.003082475771675204, "learning_rate": 9.934095976730678e-06, "loss": 0.0048, "step": 12380 }, { "epoch": 2.4510831931941834, "grad_norm": 0.00555349813602752, "learning_rate": 9.865348807977698e-06, "loss": 0.0048, "step": 12390 }, { "epoch": 2.4530616282520525, "grad_norm": 0.009849566930876161, "learning_rate": 9.79681429687891e-06, "loss": 0.0049, "step": 12400 }, { "epoch": 2.455040063309922, "grad_norm": 0.007966215298846962, "learning_rate": 9.728492806571576e-06, "loss": 0.0049, "step": 12410 }, { "epoch": 2.457018498367791, "grad_norm": 0.006040416982212873, "learning_rate": 9.6603846990643e-06, "loss": 0.0049, "step": 12420 }, { "epoch": 2.4589969334256603, "grad_norm": 0.01875501659275948, "learning_rate": 9.592490335234994e-06, "loss": 0.0049, "step": 12430 }, { "epoch": 2.4609753684835294, "grad_norm": 0.01205155068967829, "learning_rate": 9.524810074829072e-06, "loss": 0.0049, "step": 12440 }, { "epoch": 2.462953803541399, "grad_norm": 0.0044818784924866096, "learning_rate": 9.457344276457485e-06, "loss": 0.0049, "step": 12450 }, { "epoch": 2.464932238599268, "grad_norm": 0.002270084120141928, "learning_rate": 9.390093297594837e-06, "loss": 0.0049, "step": 12460 }, { "epoch": 2.466910673657137, "grad_norm": 0.005104783068399605, "learning_rate": 9.323057494577498e-06, "loss": 0.0048, "step": 12470 }, { "epoch": 2.4688891087150067, "grad_norm": 0.002472573685685777, "learning_rate": 9.25623722260166e-06, "loss": 0.0049, "step": 12480 }, { "epoch": 2.4708675437728758, "grad_norm": 0.00519307094685369, "learning_rate": 9.18963283572155e-06, "loss": 0.0048, "step": 12490 }, { "epoch": 2.472845978830745, "grad_norm": 0.010318476895607595, "learning_rate": 9.123244686847499e-06, "loss": 0.0049, "step": 12500 }, { "epoch": 2.474824413888614, "grad_norm": 0.0011839441273652051, "learning_rate": 9.057073127744064e-06, "loss": 0.0048, "step": 12510 }, { "epoch": 2.4768028489464835, "grad_norm": 0.01631501901611383, "learning_rate": 8.991118509028179e-06, "loss": 0.0049, "step": 12520 }, { "epoch": 2.4787812840043526, "grad_norm": 0.006300710853853944, "learning_rate": 8.925381180167319e-06, "loss": 0.0049, "step": 12530 }, { "epoch": 2.4807597190622217, "grad_norm": 0.028895757905357707, "learning_rate": 8.859861489477594e-06, "loss": 0.0048, "step": 12540 }, { "epoch": 2.4827381541200912, "grad_norm": 0.007594548089638376, "learning_rate": 8.794559784121937e-06, "loss": 0.0048, "step": 12550 }, { "epoch": 2.4847165891779603, "grad_norm": 0.014690523272605963, "learning_rate": 8.729476410108284e-06, "loss": 0.0049, "step": 12560 }, { "epoch": 2.4866950242358294, "grad_norm": 0.007568342130201554, "learning_rate": 8.664611712287719e-06, "loss": 0.0049, "step": 12570 }, { "epoch": 2.4886734592936985, "grad_norm": 0.004367164196902389, "learning_rate": 8.599966034352635e-06, "loss": 0.0048, "step": 12580 }, { "epoch": 2.490651894351568, "grad_norm": 0.004524892566788348, "learning_rate": 8.535539718834929e-06, "loss": 0.0048, "step": 12590 }, { "epoch": 2.492630329409437, "grad_norm": 0.009470848217927695, "learning_rate": 8.471333107104196e-06, "loss": 0.0049, "step": 12600 }, { "epoch": 2.4946087644673063, "grad_norm": 0.00925020418991596, "learning_rate": 8.407346539365869e-06, "loss": 0.0048, "step": 12610 }, { "epoch": 2.496587199525176, "grad_norm": 0.015456727196564764, "learning_rate": 8.343580354659491e-06, "loss": 0.0049, "step": 12620 }, { "epoch": 2.498565634583045, "grad_norm": 0.009795671371077898, "learning_rate": 8.280034890856885e-06, "loss": 0.0049, "step": 12630 }, { "epoch": 2.500544069640914, "grad_norm": 0.015413743597848445, "learning_rate": 8.216710484660334e-06, "loss": 0.0048, "step": 12640 }, { "epoch": 2.502522504698783, "grad_norm": 0.0063919962796238055, "learning_rate": 8.15360747160086e-06, "loss": 0.0048, "step": 12650 }, { "epoch": 2.5045009397566522, "grad_norm": 0.0033460011314181853, "learning_rate": 8.090726186036357e-06, "loss": 0.0048, "step": 12660 }, { "epoch": 2.5064793748145218, "grad_norm": 0.01185456464446356, "learning_rate": 8.02806696114992e-06, "loss": 0.0049, "step": 12670 }, { "epoch": 2.508457809872391, "grad_norm": 0.003481058095849719, "learning_rate": 7.965630128948038e-06, "loss": 0.0048, "step": 12680 }, { "epoch": 2.5104362449302604, "grad_norm": 0.013521543253812633, "learning_rate": 7.90341602025877e-06, "loss": 0.0049, "step": 12690 }, { "epoch": 2.5124146799881295, "grad_norm": 0.0027327457923147333, "learning_rate": 7.841424964730105e-06, "loss": 0.0048, "step": 12700 }, { "epoch": 2.5143931150459986, "grad_norm": 0.007769744751264851, "learning_rate": 7.779657290828146e-06, "loss": 0.0048, "step": 12710 }, { "epoch": 2.5163715501038677, "grad_norm": 0.002254573029200986, "learning_rate": 7.718113325835396e-06, "loss": 0.0049, "step": 12720 }, { "epoch": 2.518349985161737, "grad_norm": 0.004204850769055069, "learning_rate": 7.656793395848983e-06, "loss": 0.0049, "step": 12730 }, { "epoch": 2.5203284202196063, "grad_norm": 0.0008526760639661153, "learning_rate": 7.595697825778986e-06, "loss": 0.0048, "step": 12740 }, { "epoch": 2.5223068552774754, "grad_norm": 0.009944031752570833, "learning_rate": 7.5348269393466896e-06, "loss": 0.0049, "step": 12750 }, { "epoch": 2.524285290335345, "grad_norm": 0.005767282860186193, "learning_rate": 7.474181059082852e-06, "loss": 0.0049, "step": 12760 }, { "epoch": 2.526263725393214, "grad_norm": 0.0067308121768407895, "learning_rate": 7.413760506326051e-06, "loss": 0.0049, "step": 12770 }, { "epoch": 2.528242160451083, "grad_norm": 0.004008950818976804, "learning_rate": 7.353565601220874e-06, "loss": 0.0049, "step": 12780 }, { "epoch": 2.5302205955089523, "grad_norm": 0.00142063025617739, "learning_rate": 7.293596662716362e-06, "loss": 0.0048, "step": 12790 }, { "epoch": 2.5321990305668214, "grad_norm": 0.002960916421932482, "learning_rate": 7.233854008564206e-06, "loss": 0.0048, "step": 12800 }, { "epoch": 2.534177465624691, "grad_norm": 0.008054306863253048, "learning_rate": 7.174337955317123e-06, "loss": 0.0048, "step": 12810 }, { "epoch": 2.53615590068256, "grad_norm": 0.0028438536334512853, "learning_rate": 7.115048818327152e-06, "loss": 0.0049, "step": 12820 }, { "epoch": 2.5381343357404296, "grad_norm": 0.0038741092904798586, "learning_rate": 7.055986911744017e-06, "loss": 0.0049, "step": 12830 }, { "epoch": 2.5401127707982987, "grad_norm": 0.014112701526527393, "learning_rate": 6.997152548513397e-06, "loss": 0.0048, "step": 12840 }, { "epoch": 2.5420912058561678, "grad_norm": 0.004121095914865367, "learning_rate": 6.938546040375327e-06, "loss": 0.0049, "step": 12850 }, { "epoch": 2.544069640914037, "grad_norm": 0.0032779712605284996, "learning_rate": 6.8801676978625386e-06, "loss": 0.0048, "step": 12860 }, { "epoch": 2.546048075971906, "grad_norm": 0.0072806694719493035, "learning_rate": 6.822017830298788e-06, "loss": 0.0049, "step": 12870 }, { "epoch": 2.5480265110297755, "grad_norm": 0.004364603977723651, "learning_rate": 6.764096745797255e-06, "loss": 0.0049, "step": 12880 }, { "epoch": 2.5500049460876446, "grad_norm": 0.004562991607910926, "learning_rate": 6.706404751258871e-06, "loss": 0.0048, "step": 12890 }, { "epoch": 2.551983381145514, "grad_norm": 0.0054345864633303514, "learning_rate": 6.648942152370702e-06, "loss": 0.0049, "step": 12900 }, { "epoch": 2.5539618162033833, "grad_norm": 0.014948963593315757, "learning_rate": 6.591709253604356e-06, "loss": 0.0049, "step": 12910 }, { "epoch": 2.5559402512612523, "grad_norm": 0.009247579370882641, "learning_rate": 6.534706358214337e-06, "loss": 0.0048, "step": 12920 }, { "epoch": 2.5579186863191214, "grad_norm": 0.0012427365548921213, "learning_rate": 6.477933768236471e-06, "loss": 0.0049, "step": 12930 }, { "epoch": 2.5598971213769905, "grad_norm": 0.006536648495584513, "learning_rate": 6.421391784486264e-06, "loss": 0.0048, "step": 12940 }, { "epoch": 2.56187555643486, "grad_norm": 0.0015705205777996674, "learning_rate": 6.365080706557352e-06, "loss": 0.0048, "step": 12950 }, { "epoch": 2.563853991492729, "grad_norm": 0.00241187495611733, "learning_rate": 6.309000832819867e-06, "loss": 0.0048, "step": 12960 }, { "epoch": 2.5658324265505987, "grad_norm": 0.002494242934921268, "learning_rate": 6.25315246041891e-06, "loss": 0.0048, "step": 12970 }, { "epoch": 2.567810861608468, "grad_norm": 0.007796848150321629, "learning_rate": 6.197535885272909e-06, "loss": 0.0048, "step": 12980 }, { "epoch": 2.569789296666337, "grad_norm": 0.006326477757674995, "learning_rate": 6.142151402072132e-06, "loss": 0.0049, "step": 12990 }, { "epoch": 2.571767731724206, "grad_norm": 0.0017967395331649962, "learning_rate": 6.086999304277069e-06, "loss": 0.0048, "step": 13000 }, { "epoch": 2.573746166782075, "grad_norm": 0.0008539932450443773, "learning_rate": 6.032079884116876e-06, "loss": 0.0048, "step": 13010 }, { "epoch": 2.5757246018399447, "grad_norm": 0.005021577941545847, "learning_rate": 5.977393432587874e-06, "loss": 0.0048, "step": 13020 }, { "epoch": 2.5777030368978138, "grad_norm": 0.001697911372506787, "learning_rate": 5.922940239451935e-06, "loss": 0.0048, "step": 13030 }, { "epoch": 2.5796814719556833, "grad_norm": 0.01597336590967697, "learning_rate": 5.868720593235017e-06, "loss": 0.0049, "step": 13040 }, { "epoch": 2.5816599070135524, "grad_norm": 0.013353917166044805, "learning_rate": 5.814734781225589e-06, "loss": 0.0048, "step": 13050 }, { "epoch": 2.5836383420714215, "grad_norm": 0.007269597960881115, "learning_rate": 5.760983089473143e-06, "loss": 0.0049, "step": 13060 }, { "epoch": 2.5856167771292906, "grad_norm": 0.008903545850398004, "learning_rate": 5.707465802786654e-06, "loss": 0.0049, "step": 13070 }, { "epoch": 2.5875952121871597, "grad_norm": 0.004086266569955447, "learning_rate": 5.654183204733049e-06, "loss": 0.0048, "step": 13080 }, { "epoch": 2.5895736472450293, "grad_norm": 0.006768263337846134, "learning_rate": 5.601135577635775e-06, "loss": 0.0048, "step": 13090 }, { "epoch": 2.5915520823028984, "grad_norm": 0.0018643702743263491, "learning_rate": 5.54832320257323e-06, "loss": 0.0049, "step": 13100 }, { "epoch": 2.593530517360768, "grad_norm": 0.003243581435947785, "learning_rate": 5.495746359377335e-06, "loss": 0.0048, "step": 13110 }, { "epoch": 2.595508952418637, "grad_norm": 0.004728917566029483, "learning_rate": 5.443405326631984e-06, "loss": 0.0048, "step": 13120 }, { "epoch": 2.597487387476506, "grad_norm": 0.007299261063508918, "learning_rate": 5.391300381671627e-06, "loss": 0.0049, "step": 13130 }, { "epoch": 2.599465822534375, "grad_norm": 0.00925059665146975, "learning_rate": 5.339431800579781e-06, "loss": 0.0049, "step": 13140 }, { "epoch": 2.6014442575922443, "grad_norm": 0.0029994743407914994, "learning_rate": 5.287799858187548e-06, "loss": 0.0048, "step": 13150 }, { "epoch": 2.603422692650114, "grad_norm": 0.009967454996753144, "learning_rate": 5.236404828072183e-06, "loss": 0.0049, "step": 13160 }, { "epoch": 2.605401127707983, "grad_norm": 0.004916751211667762, "learning_rate": 5.185246982555641e-06, "loss": 0.0049, "step": 13170 }, { "epoch": 2.6073795627658525, "grad_norm": 0.0026891268589918035, "learning_rate": 5.134326592703131e-06, "loss": 0.0048, "step": 13180 }, { "epoch": 2.6093579978237216, "grad_norm": 0.0055139944046099995, "learning_rate": 5.083643928321679e-06, "loss": 0.0048, "step": 13190 }, { "epoch": 2.6113364328815907, "grad_norm": 0.004852980769599085, "learning_rate": 5.033199257958671e-06, "loss": 0.0049, "step": 13200 }, { "epoch": 2.6133148679394598, "grad_norm": 0.004180897063901075, "learning_rate": 4.9829928489004975e-06, "loss": 0.0048, "step": 13210 }, { "epoch": 2.615293302997329, "grad_norm": 0.01232822850597102, "learning_rate": 4.933024967171063e-06, "loss": 0.0048, "step": 13220 }, { "epoch": 2.6172717380551984, "grad_norm": 0.005438613885841797, "learning_rate": 4.88329587753043e-06, "loss": 0.0048, "step": 13230 }, { "epoch": 2.6192501731130675, "grad_norm": 0.0033909609945998323, "learning_rate": 4.8338058434733935e-06, "loss": 0.0049, "step": 13240 }, { "epoch": 2.621228608170937, "grad_norm": 0.004593092615156945, "learning_rate": 4.784555127228091e-06, "loss": 0.0048, "step": 13250 }, { "epoch": 2.623207043228806, "grad_norm": 0.004028559108808125, "learning_rate": 4.735543989754587e-06, "loss": 0.0049, "step": 13260 }, { "epoch": 2.6251854782866753, "grad_norm": 0.0212067612491117, "learning_rate": 4.686772690743529e-06, "loss": 0.0049, "step": 13270 }, { "epoch": 2.6271639133445444, "grad_norm": 0.008831743163722167, "learning_rate": 4.638241488614758e-06, "loss": 0.0049, "step": 13280 }, { "epoch": 2.6291423484024135, "grad_norm": 0.009059456374134666, "learning_rate": 4.5899506405159335e-06, "loss": 0.0049, "step": 13290 }, { "epoch": 2.631120783460283, "grad_norm": 0.004640326652481303, "learning_rate": 4.5419004023211695e-06, "loss": 0.0049, "step": 13300 }, { "epoch": 2.633099218518152, "grad_norm": 0.0026435523214844315, "learning_rate": 4.494091028629699e-06, "loss": 0.0049, "step": 13310 }, { "epoch": 2.6350776535760216, "grad_norm": 0.005140898880418435, "learning_rate": 4.446522772764494e-06, "loss": 0.0048, "step": 13320 }, { "epoch": 2.6370560886338907, "grad_norm": 0.006468965626314602, "learning_rate": 4.3991958867709345e-06, "loss": 0.0048, "step": 13330 }, { "epoch": 2.63903452369176, "grad_norm": 0.0015928423639179384, "learning_rate": 4.352110621415484e-06, "loss": 0.0048, "step": 13340 }, { "epoch": 2.641012958749629, "grad_norm": 0.0005951192881108467, "learning_rate": 4.305267226184356e-06, "loss": 0.0048, "step": 13350 }, { "epoch": 2.642991393807498, "grad_norm": 0.002605225661400826, "learning_rate": 4.258665949282192e-06, "loss": 0.0048, "step": 13360 }, { "epoch": 2.6449698288653676, "grad_norm": 0.010031024380662348, "learning_rate": 4.212307037630747e-06, "loss": 0.0048, "step": 13370 }, { "epoch": 2.6469482639232367, "grad_norm": 0.005629433330281618, "learning_rate": 4.1661907368675535e-06, "loss": 0.0049, "step": 13380 }, { "epoch": 2.648926698981106, "grad_norm": 0.0032700055091920734, "learning_rate": 4.1203172913446775e-06, "loss": 0.0048, "step": 13390 }, { "epoch": 2.6509051340389753, "grad_norm": 0.005462413521384711, "learning_rate": 4.074686944127365e-06, "loss": 0.0049, "step": 13400 }, { "epoch": 2.6528835690968444, "grad_norm": 0.0025017049596064937, "learning_rate": 4.029299936992792e-06, "loss": 0.0048, "step": 13410 }, { "epoch": 2.6548620041547135, "grad_norm": 0.0010113777824508547, "learning_rate": 3.984156510428777e-06, "loss": 0.0048, "step": 13420 }, { "epoch": 2.6568404392125826, "grad_norm": 0.006312205218192591, "learning_rate": 3.939256903632494e-06, "loss": 0.0048, "step": 13430 }, { "epoch": 2.658818874270452, "grad_norm": 0.005231179727696471, "learning_rate": 3.89460135450922e-06, "loss": 0.0048, "step": 13440 }, { "epoch": 2.6607973093283213, "grad_norm": 0.004445603794580928, "learning_rate": 3.850190099671047e-06, "loss": 0.0049, "step": 13450 }, { "epoch": 2.6627757443861904, "grad_norm": 0.007171322814320959, "learning_rate": 3.8060233744356633e-06, "loss": 0.0048, "step": 13460 }, { "epoch": 2.66475417944406, "grad_norm": 0.004664685113014176, "learning_rate": 3.7621014128250977e-06, "loss": 0.0048, "step": 13470 }, { "epoch": 2.666732614501929, "grad_norm": 0.002360786867478785, "learning_rate": 3.7184244475644616e-06, "loss": 0.0049, "step": 13480 }, { "epoch": 2.668711049559798, "grad_norm": 0.00503157369246434, "learning_rate": 3.6749927100807434e-06, "loss": 0.0048, "step": 13490 }, { "epoch": 2.670689484617667, "grad_norm": 0.006881519238343002, "learning_rate": 3.6318064305015332e-06, "loss": 0.0049, "step": 13500 }, { "epoch": 2.6726679196755367, "grad_norm": 0.0007378899971421895, "learning_rate": 3.5888658376538655e-06, "loss": 0.0048, "step": 13510 }, { "epoch": 2.674646354733406, "grad_norm": 0.008433483282637192, "learning_rate": 3.5461711590629774e-06, "loss": 0.0049, "step": 13520 }, { "epoch": 2.676624789791275, "grad_norm": 0.003901899511946218, "learning_rate": 3.5037226209510887e-06, "loss": 0.0048, "step": 13530 }, { "epoch": 2.6786032248491445, "grad_norm": 0.0017661622853674477, "learning_rate": 3.4615204482362386e-06, "loss": 0.0049, "step": 13540 }, { "epoch": 2.6805816599070136, "grad_norm": 0.003310248630767662, "learning_rate": 3.419564864531044e-06, "loss": 0.0049, "step": 13550 }, { "epoch": 2.6825600949648827, "grad_norm": 0.005126772470457341, "learning_rate": 3.3778560921415738e-06, "loss": 0.0048, "step": 13560 }, { "epoch": 2.684538530022752, "grad_norm": 0.004965879628756141, "learning_rate": 3.3363943520661132e-06, "loss": 0.0049, "step": 13570 }, { "epoch": 2.6865169650806213, "grad_norm": 0.005421902804820821, "learning_rate": 3.2951798639940467e-06, "loss": 0.0049, "step": 13580 }, { "epoch": 2.6884954001384904, "grad_norm": 0.0030647045961330714, "learning_rate": 3.2542128463046497e-06, "loss": 0.0049, "step": 13590 }, { "epoch": 2.6904738351963595, "grad_norm": 0.0015048859352265213, "learning_rate": 3.2134935160659686e-06, "loss": 0.0048, "step": 13600 }, { "epoch": 2.692452270254229, "grad_norm": 0.0026285450591928693, "learning_rate": 3.173022089033645e-06, "loss": 0.0049, "step": 13610 }, { "epoch": 2.694430705312098, "grad_norm": 0.003483204049444364, "learning_rate": 3.1327987796497537e-06, "loss": 0.0048, "step": 13620 }, { "epoch": 2.6964091403699673, "grad_norm": 0.003138742982197174, "learning_rate": 3.0928238010417275e-06, "loss": 0.0048, "step": 13630 }, { "epoch": 2.6983875754278364, "grad_norm": 0.004981790064871517, "learning_rate": 3.0530973650211793e-06, "loss": 0.0048, "step": 13640 }, { "epoch": 2.700366010485706, "grad_norm": 0.00582828499856072, "learning_rate": 3.0136196820827855e-06, "loss": 0.0049, "step": 13650 }, { "epoch": 2.702344445543575, "grad_norm": 0.006862689769200456, "learning_rate": 2.974390961403195e-06, "loss": 0.0049, "step": 13660 }, { "epoch": 2.704322880601444, "grad_norm": 0.006858303373264722, "learning_rate": 2.935411410839889e-06, "loss": 0.0048, "step": 13670 }, { "epoch": 2.7063013156593136, "grad_norm": 0.0061425644080476734, "learning_rate": 2.8966812369301043e-06, "loss": 0.0048, "step": 13680 }, { "epoch": 2.7082797507171827, "grad_norm": 0.003213330807126191, "learning_rate": 2.8582006448897135e-06, "loss": 0.0048, "step": 13690 }, { "epoch": 2.710258185775052, "grad_norm": 0.0052257090393458714, "learning_rate": 2.8199698386121687e-06, "loss": 0.0048, "step": 13700 }, { "epoch": 2.712236620832921, "grad_norm": 0.009628881876217424, "learning_rate": 2.7819890206674084e-06, "loss": 0.0049, "step": 13710 }, { "epoch": 2.7142150558907905, "grad_norm": 0.005027999552882633, "learning_rate": 2.7442583923007746e-06, "loss": 0.0048, "step": 13720 }, { "epoch": 2.7161934909486596, "grad_norm": 0.009323820727992444, "learning_rate": 2.706778153431949e-06, "loss": 0.0048, "step": 13730 }, { "epoch": 2.7181719260065287, "grad_norm": 0.004600323934939696, "learning_rate": 2.669548502653918e-06, "loss": 0.0049, "step": 13740 }, { "epoch": 2.7201503610643982, "grad_norm": 0.008440113987286935, "learning_rate": 2.6325696372318688e-06, "loss": 0.0049, "step": 13750 }, { "epoch": 2.7221287961222673, "grad_norm": 0.0029263441486813793, "learning_rate": 2.5958417531022017e-06, "loss": 0.0049, "step": 13760 }, { "epoch": 2.7241072311801364, "grad_norm": 0.005052793182728446, "learning_rate": 2.559365044871459e-06, "loss": 0.0049, "step": 13770 }, { "epoch": 2.7260856662380055, "grad_norm": 0.004494716424711221, "learning_rate": 2.523139705815297e-06, "loss": 0.0048, "step": 13780 }, { "epoch": 2.728064101295875, "grad_norm": 0.0027326453267776684, "learning_rate": 2.4871659278774883e-06, "loss": 0.0049, "step": 13790 }, { "epoch": 2.730042536353744, "grad_norm": 0.004964712767683836, "learning_rate": 2.4514439016688374e-06, "loss": 0.0049, "step": 13800 }, { "epoch": 2.7320209714116133, "grad_norm": 0.001588210859820669, "learning_rate": 2.415973816466255e-06, "loss": 0.0048, "step": 13810 }, { "epoch": 2.733999406469483, "grad_norm": 0.0036082088060689272, "learning_rate": 2.3807558602117143e-06, "loss": 0.0048, "step": 13820 }, { "epoch": 2.735977841527352, "grad_norm": 0.004177964514163221, "learning_rate": 2.3457902195112236e-06, "loss": 0.0048, "step": 13830 }, { "epoch": 2.737956276585221, "grad_norm": 0.005193936600566022, "learning_rate": 2.311077079633911e-06, "loss": 0.0048, "step": 13840 }, { "epoch": 2.73993471164309, "grad_norm": 0.0034441914412583034, "learning_rate": 2.2766166245109744e-06, "loss": 0.0049, "step": 13850 }, { "epoch": 2.7419131467009596, "grad_norm": 0.004016964016843684, "learning_rate": 2.2424090367347605e-06, "loss": 0.0048, "step": 13860 }, { "epoch": 2.7438915817588287, "grad_norm": 0.0036180168191544512, "learning_rate": 2.2084544975577383e-06, "loss": 0.0049, "step": 13870 }, { "epoch": 2.745870016816698, "grad_norm": 0.005503024809430983, "learning_rate": 2.174753186891604e-06, "loss": 0.0048, "step": 13880 }, { "epoch": 2.7478484518745674, "grad_norm": 0.004243844436966538, "learning_rate": 2.1413052833062773e-06, "loss": 0.0048, "step": 13890 }, { "epoch": 2.7498268869324365, "grad_norm": 0.004285618924864715, "learning_rate": 2.1081109640289755e-06, "loss": 0.0048, "step": 13900 }, { "epoch": 2.7518053219903056, "grad_norm": 0.010242620619672321, "learning_rate": 2.0751704049432943e-06, "loss": 0.0049, "step": 13910 }, { "epoch": 2.7537837570481747, "grad_norm": 0.003893931419828782, "learning_rate": 2.0424837805882224e-06, "loss": 0.0049, "step": 13920 }, { "epoch": 2.7557621921060442, "grad_norm": 0.0019250160860279186, "learning_rate": 2.0100512641572587e-06, "loss": 0.0049, "step": 13930 }, { "epoch": 2.7577406271639133, "grad_norm": 0.004792600357030465, "learning_rate": 1.9778730274975012e-06, "loss": 0.0048, "step": 13940 }, { "epoch": 2.7597190622217824, "grad_norm": 0.0025149201915899227, "learning_rate": 1.945949241108708e-06, "loss": 0.0048, "step": 13950 }, { "epoch": 2.761697497279652, "grad_norm": 0.0011817467921534325, "learning_rate": 1.914280074142405e-06, "loss": 0.0049, "step": 13960 }, { "epoch": 2.763675932337521, "grad_norm": 0.005226392595291648, "learning_rate": 1.8828656944009814e-06, "loss": 0.0048, "step": 13970 }, { "epoch": 2.76565436739539, "grad_norm": 0.003617197591234684, "learning_rate": 1.8517062683368335e-06, "loss": 0.0048, "step": 13980 }, { "epoch": 2.7676328024532593, "grad_norm": 0.010585311050759013, "learning_rate": 1.820801961051427e-06, "loss": 0.0049, "step": 13990 }, { "epoch": 2.769611237511129, "grad_norm": 0.004021684260786729, "learning_rate": 1.7901529362944869e-06, "loss": 0.0048, "step": 14000 }, { "epoch": 2.771589672568998, "grad_norm": 0.0037546978079979615, "learning_rate": 1.759759356463081e-06, "loss": 0.0049, "step": 14010 }, { "epoch": 2.773568107626867, "grad_norm": 0.004731520815419526, "learning_rate": 1.7296213826007712e-06, "loss": 0.0048, "step": 14020 }, { "epoch": 2.7755465426847366, "grad_norm": 0.0011531736547068102, "learning_rate": 1.6997391743967695e-06, "loss": 0.0049, "step": 14030 }, { "epoch": 2.7775249777426056, "grad_norm": 0.006879359703001387, "learning_rate": 1.6701128901850937e-06, "loss": 0.0049, "step": 14040 }, { "epoch": 2.7795034128004747, "grad_norm": 0.004791362264315265, "learning_rate": 1.6407426869437026e-06, "loss": 0.0048, "step": 14050 }, { "epoch": 2.781481847858344, "grad_norm": 0.01216875043745859, "learning_rate": 1.6116287202937063e-06, "loss": 0.0048, "step": 14060 }, { "epoch": 2.7834602829162134, "grad_norm": 0.014237425378251076, "learning_rate": 1.5827711444985016e-06, "loss": 0.0048, "step": 14070 }, { "epoch": 2.7854387179740825, "grad_norm": 0.004184868700467683, "learning_rate": 1.554170112462977e-06, "loss": 0.0049, "step": 14080 }, { "epoch": 2.7874171530319516, "grad_norm": 0.004558770615845977, "learning_rate": 1.5258257757327033e-06, "loss": 0.0048, "step": 14090 }, { "epoch": 2.789395588089821, "grad_norm": 0.002582327508931167, "learning_rate": 1.4977382844931166e-06, "loss": 0.0048, "step": 14100 }, { "epoch": 2.7913740231476902, "grad_norm": 0.004949911493626684, "learning_rate": 1.4699077875687251e-06, "loss": 0.0048, "step": 14110 }, { "epoch": 2.7933524582055593, "grad_norm": 0.0016939808739032589, "learning_rate": 1.442334432422332e-06, "loss": 0.0049, "step": 14120 }, { "epoch": 2.7953308932634284, "grad_norm": 0.0051998298490565364, "learning_rate": 1.415018365154258e-06, "loss": 0.0049, "step": 14130 }, { "epoch": 2.797309328321298, "grad_norm": 0.005949982547839485, "learning_rate": 1.3879597305015368e-06, "loss": 0.0049, "step": 14140 }, { "epoch": 2.799287763379167, "grad_norm": 0.005152079793393017, "learning_rate": 1.3611586718371871e-06, "loss": 0.0049, "step": 14150 }, { "epoch": 2.801266198437036, "grad_norm": 0.0021400510880142942, "learning_rate": 1.3346153311694309e-06, "loss": 0.0048, "step": 14160 }, { "epoch": 2.8032446334949057, "grad_norm": 0.0019312786188725396, "learning_rate": 1.3083298491409213e-06, "loss": 0.0048, "step": 14170 }, { "epoch": 2.805223068552775, "grad_norm": 0.0013928867682459114, "learning_rate": 1.2823023650280541e-06, "loss": 0.0048, "step": 14180 }, { "epoch": 2.807201503610644, "grad_norm": 0.00483001712892883, "learning_rate": 1.2565330167401745e-06, "loss": 0.0048, "step": 14190 }, { "epoch": 2.809179938668513, "grad_norm": 0.007505111889805549, "learning_rate": 1.2310219408188773e-06, "loss": 0.0049, "step": 14200 }, { "epoch": 2.8111583737263826, "grad_norm": 0.0052904666567675455, "learning_rate": 1.2057692724372794e-06, "loss": 0.0048, "step": 14210 }, { "epoch": 2.8131368087842517, "grad_norm": 0.004338996948583365, "learning_rate": 1.1807751453992654e-06, "loss": 0.0049, "step": 14220 }, { "epoch": 2.8151152438421208, "grad_norm": 0.003530984332917518, "learning_rate": 1.156039692138855e-06, "loss": 0.0049, "step": 14230 }, { "epoch": 2.8170936788999903, "grad_norm": 0.0037448103425007785, "learning_rate": 1.1315630437194359e-06, "loss": 0.0048, "step": 14240 }, { "epoch": 2.8190721139578594, "grad_norm": 0.0031079497556411413, "learning_rate": 1.1073453298330982e-06, "loss": 0.0049, "step": 14250 }, { "epoch": 2.8210505490157285, "grad_norm": 0.0033532018115630525, "learning_rate": 1.0833866787999248e-06, "loss": 0.0048, "step": 14260 }, { "epoch": 2.8230289840735976, "grad_norm": 0.0014324613793292667, "learning_rate": 1.0596872175673455e-06, "loss": 0.0049, "step": 14270 }, { "epoch": 2.825007419131467, "grad_norm": 0.004369184236434416, "learning_rate": 1.0362470717094397e-06, "loss": 0.0048, "step": 14280 }, { "epoch": 2.8269858541893362, "grad_norm": 0.005731694170409952, "learning_rate": 1.0130663654262684e-06, "loss": 0.0048, "step": 14290 }, { "epoch": 2.8289642892472053, "grad_norm": 0.007502654651256006, "learning_rate": 9.901452215432428e-07, "loss": 0.0048, "step": 14300 }, { "epoch": 2.830942724305075, "grad_norm": 0.0016282043774877885, "learning_rate": 9.674837615104349e-07, "loss": 0.0048, "step": 14310 }, { "epoch": 2.832921159362944, "grad_norm": 0.004379866917176618, "learning_rate": 9.450821054019732e-07, "loss": 0.0048, "step": 14320 }, { "epoch": 2.834899594420813, "grad_norm": 0.001786943115187066, "learning_rate": 9.229403719153817e-07, "loss": 0.0049, "step": 14330 }, { "epoch": 2.836878029478682, "grad_norm": 0.0013321729174920794, "learning_rate": 9.010586783709418e-07, "loss": 0.0048, "step": 14340 }, { "epoch": 2.8388564645365517, "grad_norm": 0.006966572397858866, "learning_rate": 8.794371407111091e-07, "loss": 0.0049, "step": 14350 }, { "epoch": 2.840834899594421, "grad_norm": 0.000598975408119532, "learning_rate": 8.580758734998639e-07, "loss": 0.0048, "step": 14360 }, { "epoch": 2.84281333465229, "grad_norm": 0.004184093628574164, "learning_rate": 8.36974989922118e-07, "loss": 0.0048, "step": 14370 }, { "epoch": 2.8447917697101595, "grad_norm": 0.00597862564574173, "learning_rate": 8.161346017831195e-07, "loss": 0.0048, "step": 14380 }, { "epoch": 2.8467702047680286, "grad_norm": 0.002478651945953765, "learning_rate": 7.955548195078433e-07, "loss": 0.0048, "step": 14390 }, { "epoch": 2.8487486398258977, "grad_norm": 0.0021844725282744453, "learning_rate": 7.752357521404186e-07, "loss": 0.0049, "step": 14400 }, { "epoch": 2.8507270748837668, "grad_norm": 0.005677004597278862, "learning_rate": 7.551775073435574e-07, "loss": 0.0048, "step": 14410 }, { "epoch": 2.8527055099416363, "grad_norm": 0.005829436202596139, "learning_rate": 7.353801913979719e-07, "loss": 0.0048, "step": 14420 }, { "epoch": 2.8546839449995054, "grad_norm": 0.006649880031649109, "learning_rate": 7.158439092018076e-07, "loss": 0.0049, "step": 14430 }, { "epoch": 2.8566623800573745, "grad_norm": 0.0020718784854737386, "learning_rate": 6.965687642701113e-07, "loss": 0.0048, "step": 14440 }, { "epoch": 2.858640815115244, "grad_norm": 0.0021127137397771887, "learning_rate": 6.775548587342529e-07, "loss": 0.0048, "step": 14450 }, { "epoch": 2.860619250173113, "grad_norm": 0.0015409484242964512, "learning_rate": 6.588022933414095e-07, "loss": 0.0048, "step": 14460 }, { "epoch": 2.8625976852309822, "grad_norm": 0.004308262757931374, "learning_rate": 6.403111674539997e-07, "loss": 0.0048, "step": 14470 }, { "epoch": 2.8645761202888513, "grad_norm": 0.005783903244574432, "learning_rate": 6.220815790492051e-07, "loss": 0.0048, "step": 14480 }, { "epoch": 2.866554555346721, "grad_norm": 0.003079496865396833, "learning_rate": 6.041136247184054e-07, "loss": 0.0049, "step": 14490 }, { "epoch": 2.86853299040459, "grad_norm": 0.0028643836154889695, "learning_rate": 5.86407399666683e-07, "loss": 0.0048, "step": 14500 }, { "epoch": 2.870511425462459, "grad_norm": 0.005466105161282796, "learning_rate": 5.689629977123412e-07, "loss": 0.0048, "step": 14510 }, { "epoch": 2.8724898605203286, "grad_norm": 0.005519867678213713, "learning_rate": 5.517805112863539e-07, "loss": 0.0048, "step": 14520 }, { "epoch": 2.8744682955781977, "grad_norm": 0.005182784903990588, "learning_rate": 5.348600314319385e-07, "loss": 0.0048, "step": 14530 }, { "epoch": 2.876446730636067, "grad_norm": 0.009325660933351328, "learning_rate": 5.182016478040231e-07, "loss": 0.0048, "step": 14540 }, { "epoch": 2.878425165693936, "grad_norm": 0.008607084025388996, "learning_rate": 5.018054486687962e-07, "loss": 0.0048, "step": 14550 }, { "epoch": 2.8804036007518055, "grad_norm": 0.0009766197499738547, "learning_rate": 4.85671520903247e-07, "loss": 0.0049, "step": 14560 }, { "epoch": 2.8823820358096746, "grad_norm": 0.00734680562283183, "learning_rate": 4.697999499946704e-07, "loss": 0.0048, "step": 14570 }, { "epoch": 2.8843604708675437, "grad_norm": 0.007936605056646107, "learning_rate": 4.5419082004026226e-07, "loss": 0.0049, "step": 14580 }, { "epoch": 2.886338905925413, "grad_norm": 0.00842045357896325, "learning_rate": 4.3884421374661975e-07, "loss": 0.0048, "step": 14590 }, { "epoch": 2.8883173409832823, "grad_norm": 0.0019392364228458473, "learning_rate": 4.2376021242934714e-07, "loss": 0.0048, "step": 14600 }, { "epoch": 2.8902957760411514, "grad_norm": 0.00418628922923065, "learning_rate": 4.089388960126117e-07, "loss": 0.0049, "step": 14610 }, { "epoch": 2.8922742110990205, "grad_norm": 0.0061032788788575165, "learning_rate": 3.943803430287052e-07, "loss": 0.0048, "step": 14620 }, { "epoch": 2.89425264615689, "grad_norm": 0.0057638804814412305, "learning_rate": 3.8008463061764977e-07, "loss": 0.0049, "step": 14630 }, { "epoch": 2.896231081214759, "grad_norm": 0.013137752053138455, "learning_rate": 3.660518345267705e-07, "loss": 0.0049, "step": 14640 }, { "epoch": 2.8982095162726282, "grad_norm": 0.0033241561557892335, "learning_rate": 3.5228202911030685e-07, "loss": 0.0049, "step": 14650 }, { "epoch": 2.900187951330498, "grad_norm": 0.004562301398920238, "learning_rate": 3.38775287329024e-07, "loss": 0.0048, "step": 14660 }, { "epoch": 2.902166386388367, "grad_norm": 0.004116869480337898, "learning_rate": 3.255316807498077e-07, "loss": 0.0049, "step": 14670 }, { "epoch": 2.904144821446236, "grad_norm": 0.002837615843071466, "learning_rate": 3.125512795452812e-07, "loss": 0.0049, "step": 14680 }, { "epoch": 2.906123256504105, "grad_norm": 0.005530720000228801, "learning_rate": 2.998341524934778e-07, "loss": 0.0049, "step": 14690 }, { "epoch": 2.9081016915619746, "grad_norm": 0.004973268891462566, "learning_rate": 2.873803669774133e-07, "loss": 0.0048, "step": 14700 }, { "epoch": 2.9100801266198437, "grad_norm": 0.00431853321643085, "learning_rate": 2.751899889847864e-07, "loss": 0.0048, "step": 14710 }, { "epoch": 2.912058561677713, "grad_norm": 0.003434411255473935, "learning_rate": 2.632630831075844e-07, "loss": 0.0048, "step": 14720 }, { "epoch": 2.9140369967355824, "grad_norm": 0.006690729853517542, "learning_rate": 2.5159971254176684e-07, "loss": 0.0048, "step": 14730 }, { "epoch": 2.9160154317934515, "grad_norm": 0.008193662778541994, "learning_rate": 2.401999390869325e-07, "loss": 0.0048, "step": 14740 }, { "epoch": 2.9179938668513206, "grad_norm": 0.000768126958930955, "learning_rate": 2.290638231459641e-07, "loss": 0.0048, "step": 14750 }, { "epoch": 2.9199723019091897, "grad_norm": 0.005307160961582646, "learning_rate": 2.1819142372475066e-07, "loss": 0.0049, "step": 14760 }, { "epoch": 2.921950736967059, "grad_norm": 0.004756913266892915, "learning_rate": 2.0758279843183793e-07, "loss": 0.0049, "step": 14770 }, { "epoch": 2.9239291720249283, "grad_norm": 0.005730632138802588, "learning_rate": 1.972380034781285e-07, "loss": 0.0049, "step": 14780 }, { "epoch": 2.9259076070827974, "grad_norm": 0.005801143033581654, "learning_rate": 1.8715709367660982e-07, "loss": 0.0048, "step": 14790 }, { "epoch": 2.927886042140667, "grad_norm": 0.006487623330286614, "learning_rate": 1.773401224420379e-07, "loss": 0.0048, "step": 14800 }, { "epoch": 2.929864477198536, "grad_norm": 0.0038017145967274776, "learning_rate": 1.6778714179065957e-07, "loss": 0.0049, "step": 14810 }, { "epoch": 2.931842912256405, "grad_norm": 0.0030548272702137767, "learning_rate": 1.584982023399517e-07, "loss": 0.0049, "step": 14820 }, { "epoch": 2.9338213473142742, "grad_norm": 0.004147087216017951, "learning_rate": 1.4947335330832702e-07, "loss": 0.0049, "step": 14830 }, { "epoch": 2.935799782372144, "grad_norm": 0.00174197477409132, "learning_rate": 1.4071264251489523e-07, "loss": 0.0048, "step": 14840 }, { "epoch": 2.937778217430013, "grad_norm": 0.004563306692685047, "learning_rate": 1.3221611637919685e-07, "loss": 0.0049, "step": 14850 }, { "epoch": 2.939756652487882, "grad_norm": 0.0031793290723380074, "learning_rate": 1.2398381992096975e-07, "loss": 0.0048, "step": 14860 }, { "epoch": 2.9417350875457515, "grad_norm": 0.003738732829065381, "learning_rate": 1.160157967598996e-07, "loss": 0.0048, "step": 14870 }, { "epoch": 2.9437135226036206, "grad_norm": 0.004333297335736699, "learning_rate": 1.0831208911539215e-07, "loss": 0.0049, "step": 14880 }, { "epoch": 2.9456919576614897, "grad_norm": 0.00512236297143814, "learning_rate": 1.0087273780634566e-07, "loss": 0.0048, "step": 14890 }, { "epoch": 2.947670392719359, "grad_norm": 0.007813787606281491, "learning_rate": 9.36977822509455e-08, "loss": 0.0049, "step": 14900 }, { "epoch": 2.9496488277772284, "grad_norm": 0.004358594184095444, "learning_rate": 8.678726046644214e-08, "loss": 0.0049, "step": 14910 }, { "epoch": 2.9516272628350975, "grad_norm": 0.0014288837553849292, "learning_rate": 8.014120906896239e-08, "loss": 0.0048, "step": 14920 }, { "epoch": 2.9536056978929666, "grad_norm": 0.0024985450686210893, "learning_rate": 7.375966327330952e-08, "loss": 0.0048, "step": 14930 }, { "epoch": 2.955584132950836, "grad_norm": 0.0037448442660448816, "learning_rate": 6.76426568927635e-08, "loss": 0.0048, "step": 14940 }, { "epoch": 2.957562568008705, "grad_norm": 0.006883183434147885, "learning_rate": 6.179022233893106e-08, "loss": 0.0048, "step": 14950 }, { "epoch": 2.9595410030665743, "grad_norm": 0.00808865230971712, "learning_rate": 5.6202390621551415e-08, "loss": 0.0049, "step": 14960 }, { "epoch": 2.9615194381244434, "grad_norm": 0.00811738358324938, "learning_rate": 5.0879191348324194e-08, "loss": 0.0048, "step": 14970 }, { "epoch": 2.963497873182313, "grad_norm": 0.005500668190329635, "learning_rate": 4.5820652724787305e-08, "loss": 0.0048, "step": 14980 }, { "epoch": 2.965476308240182, "grad_norm": 0.001645277197555448, "learning_rate": 4.102680155413929e-08, "loss": 0.0048, "step": 14990 }, { "epoch": 2.967454743298051, "grad_norm": 0.003610182640832566, "learning_rate": 3.649766323710058e-08, "loss": 0.0049, "step": 15000 }, { "epoch": 2.9694331783559207, "grad_norm": 0.0010555004969539858, "learning_rate": 3.223326177178021e-08, "loss": 0.0048, "step": 15010 }, { "epoch": 2.97141161341379, "grad_norm": 0.007560583603421006, "learning_rate": 2.8233619753564867e-08, "loss": 0.0048, "step": 15020 }, { "epoch": 2.973390048471659, "grad_norm": 0.0015282841198763321, "learning_rate": 2.4498758374968954e-08, "loss": 0.0048, "step": 15030 }, { "epoch": 2.975368483529528, "grad_norm": 0.0019451237851731734, "learning_rate": 2.1028697425551356e-08, "loss": 0.0048, "step": 15040 }, { "epoch": 2.9773469185873975, "grad_norm": 0.005313362335658087, "learning_rate": 1.78234552917933e-08, "loss": 0.0048, "step": 15050 }, { "epoch": 2.9793253536452666, "grad_norm": 0.0018777829285824509, "learning_rate": 1.4883048957003986e-08, "loss": 0.0048, "step": 15060 }, { "epoch": 2.9813037887031357, "grad_norm": 0.004736391195712116, "learning_rate": 1.220749400123733e-08, "loss": 0.0049, "step": 15070 }, { "epoch": 2.9832822237610053, "grad_norm": 0.003374872513195771, "learning_rate": 9.796804601197584e-09, "loss": 0.0048, "step": 15080 }, { "epoch": 2.9852606588188744, "grad_norm": 0.004455443521052089, "learning_rate": 7.650993530178285e-09, "loss": 0.0048, "step": 15090 }, { "epoch": 2.9872390938767435, "grad_norm": 0.0016684203100349737, "learning_rate": 5.77007215798453e-09, "loss": 0.0048, "step": 15100 }, { "epoch": 2.9892175289346126, "grad_norm": 0.011799365111640984, "learning_rate": 4.154050450871916e-09, "loss": 0.0049, "step": 15110 }, { "epoch": 2.991195963992482, "grad_norm": 0.0017881864431620192, "learning_rate": 2.8029369715021345e-09, "loss": 0.0049, "step": 15120 }, { "epoch": 2.993174399050351, "grad_norm": 0.004520916857726118, "learning_rate": 1.7167388788985606e-09, "loss": 0.0049, "step": 15130 }, { "epoch": 2.9951528341082203, "grad_norm": 0.010133547019265035, "learning_rate": 8.954619283851884e-10, "loss": 0.0049, "step": 15140 }, { "epoch": 2.99713126916609, "grad_norm": 0.006355893772964033, "learning_rate": 3.3911047158663447e-10, "loss": 0.0048, "step": 15150 }, { "epoch": 2.999109704223959, "grad_norm": 0.006182268850304125, "learning_rate": 4.7687456389278186e-11, "loss": 0.0049, "step": 15160 }, { "epoch": 3.0, "step": 15165, "total_flos": 1373930750214144.0, "train_loss": 0.11494380943673674, "train_runtime": 41155.9364, "train_samples_per_second": 188.625, "train_steps_per_second": 0.368 } ], "logging_steps": 10, "max_steps": 15165, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 1800, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1373930750214144.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }