{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 3959, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.002525890376357666, "grad_norm": 4348443.0, "learning_rate": 4.987370548118212e-05, "loss": 11.2101, "step": 10 }, { "epoch": 0.005051780752715332, "grad_norm": 6754872.5, "learning_rate": 4.974741096236424e-05, "loss": 9.3824, "step": 20 }, { "epoch": 0.007577671129072998, "grad_norm": 5965014.0, "learning_rate": 4.962111644354635e-05, "loss": 7.8958, "step": 30 }, { "epoch": 0.010103561505430665, "grad_norm": 2165972.25, "learning_rate": 4.949482192472847e-05, "loss": 7.421, "step": 40 }, { "epoch": 0.01262945188178833, "grad_norm": 3444951.25, "learning_rate": 4.936852740591059e-05, "loss": 7.2076, "step": 50 }, { "epoch": 0.015155342258145996, "grad_norm": 1671924.125, "learning_rate": 4.92422328870927e-05, "loss": 7.121, "step": 60 }, { "epoch": 0.017681232634503663, "grad_norm": 1477403.75, "learning_rate": 4.911593836827482e-05, "loss": 6.8903, "step": 70 }, { "epoch": 0.02020712301086133, "grad_norm": 1541815.25, "learning_rate": 4.898964384945693e-05, "loss": 6.7837, "step": 80 }, { "epoch": 0.022733013387218996, "grad_norm": 9298988.0, "learning_rate": 4.886334933063905e-05, "loss": 6.9096, "step": 90 }, { "epoch": 0.02525890376357666, "grad_norm": 1347703.0, "learning_rate": 4.873705481182117e-05, "loss": 6.4829, "step": 100 }, { "epoch": 0.027784794139934326, "grad_norm": 744675.0625, "learning_rate": 4.861076029300329e-05, "loss": 6.292, "step": 110 }, { "epoch": 0.030310684516291993, "grad_norm": 1124820.25, "learning_rate": 4.84844657741854e-05, "loss": 5.9801, "step": 120 }, { "epoch": 0.032836574892649656, "grad_norm": 926771.5, "learning_rate": 4.835817125536752e-05, "loss": 6.5963, "step": 130 }, { "epoch": 0.035362465269007326, "grad_norm": 929847.5, "learning_rate": 4.823187673654964e-05, "loss": 6.235, "step": 140 }, { "epoch": 0.03788835564536499, "grad_norm": 2526786.75, "learning_rate": 4.810558221773176e-05, "loss": 5.9566, "step": 150 }, { "epoch": 0.04041424602172266, "grad_norm": 914504.5, "learning_rate": 4.797928769891387e-05, "loss": 5.9669, "step": 160 }, { "epoch": 0.04294013639808032, "grad_norm": 795588.8125, "learning_rate": 4.785299318009598e-05, "loss": 5.7371, "step": 170 }, { "epoch": 0.04546602677443799, "grad_norm": 954969.25, "learning_rate": 4.77266986612781e-05, "loss": 5.5412, "step": 180 }, { "epoch": 0.047991917150795656, "grad_norm": 964213.375, "learning_rate": 4.760040414246022e-05, "loss": 5.5661, "step": 190 }, { "epoch": 0.05051780752715332, "grad_norm": 1348911.125, "learning_rate": 4.747410962364234e-05, "loss": 5.7872, "step": 200 }, { "epoch": 0.05304369790351099, "grad_norm": 793627.5, "learning_rate": 4.7347815104824456e-05, "loss": 5.6837, "step": 210 }, { "epoch": 0.05556958827986865, "grad_norm": 762008.625, "learning_rate": 4.722152058600657e-05, "loss": 5.3987, "step": 220 }, { "epoch": 0.05809547865622632, "grad_norm": 724261.0625, "learning_rate": 4.709522606718869e-05, "loss": 5.4719, "step": 230 }, { "epoch": 0.060621369032583985, "grad_norm": 1137774.75, "learning_rate": 4.6968931548370805e-05, "loss": 5.4004, "step": 240 }, { "epoch": 0.06314725940894166, "grad_norm": 752281.3125, "learning_rate": 4.6842637029552924e-05, "loss": 5.3922, "step": 250 }, { "epoch": 0.06567314978529931, "grad_norm": 558181.375, "learning_rate": 4.6716342510735036e-05, "loss": 5.2856, "step": 260 }, { "epoch": 0.06819904016165698, "grad_norm": 927075.5, "learning_rate": 4.6590047991917155e-05, "loss": 5.1218, "step": 270 }, { "epoch": 0.07072493053801465, "grad_norm": 1678542.0, "learning_rate": 4.646375347309927e-05, "loss": 5.0974, "step": 280 }, { "epoch": 0.07325082091437232, "grad_norm": 1064729.5, "learning_rate": 4.6337458954281386e-05, "loss": 5.2749, "step": 290 }, { "epoch": 0.07577671129072998, "grad_norm": 1193670.375, "learning_rate": 4.6211164435463505e-05, "loss": 5.1134, "step": 300 }, { "epoch": 0.07830260166708765, "grad_norm": 979860.8125, "learning_rate": 4.6084869916645617e-05, "loss": 4.9763, "step": 310 }, { "epoch": 0.08082849204344532, "grad_norm": 868298.8125, "learning_rate": 4.5958575397827735e-05, "loss": 5.0413, "step": 320 }, { "epoch": 0.08335438241980297, "grad_norm": 1145414.625, "learning_rate": 4.5832280879009854e-05, "loss": 4.8234, "step": 330 }, { "epoch": 0.08588027279616064, "grad_norm": 1516276.375, "learning_rate": 4.570598636019197e-05, "loss": 4.7487, "step": 340 }, { "epoch": 0.08840616317251832, "grad_norm": 1310860.25, "learning_rate": 4.557969184137409e-05, "loss": 4.869, "step": 350 }, { "epoch": 0.09093205354887599, "grad_norm": 961273.9375, "learning_rate": 4.5453397322556204e-05, "loss": 5.2202, "step": 360 }, { "epoch": 0.09345794392523364, "grad_norm": 888523.3125, "learning_rate": 4.532710280373832e-05, "loss": 5.1888, "step": 370 }, { "epoch": 0.09598383430159131, "grad_norm": 1027426.9375, "learning_rate": 4.5200808284920434e-05, "loss": 4.8891, "step": 380 }, { "epoch": 0.09850972467794898, "grad_norm": 1760436.75, "learning_rate": 4.507451376610255e-05, "loss": 4.7838, "step": 390 }, { "epoch": 0.10103561505430664, "grad_norm": 835284.875, "learning_rate": 4.4948219247284665e-05, "loss": 4.9398, "step": 400 }, { "epoch": 0.10356150543066431, "grad_norm": 1165127.125, "learning_rate": 4.4821924728466784e-05, "loss": 5.0785, "step": 410 }, { "epoch": 0.10608739580702198, "grad_norm": 1061038.75, "learning_rate": 4.46956302096489e-05, "loss": 4.9144, "step": 420 }, { "epoch": 0.10861328618337965, "grad_norm": 1058956.875, "learning_rate": 4.456933569083102e-05, "loss": 4.7658, "step": 430 }, { "epoch": 0.1111391765597373, "grad_norm": 838523.1875, "learning_rate": 4.444304117201314e-05, "loss": 4.9675, "step": 440 }, { "epoch": 0.11366506693609497, "grad_norm": 1297820.5, "learning_rate": 4.431674665319525e-05, "loss": 4.8902, "step": 450 }, { "epoch": 0.11619095731245264, "grad_norm": 841045.25, "learning_rate": 4.419045213437737e-05, "loss": 4.7037, "step": 460 }, { "epoch": 0.1187168476888103, "grad_norm": 1158120.75, "learning_rate": 4.406415761555949e-05, "loss": 4.7918, "step": 470 }, { "epoch": 0.12124273806516797, "grad_norm": 646241.4375, "learning_rate": 4.39378630967416e-05, "loss": 4.7433, "step": 480 }, { "epoch": 0.12376862844152564, "grad_norm": 641319.3125, "learning_rate": 4.381156857792372e-05, "loss": 4.608, "step": 490 }, { "epoch": 0.1262945188178833, "grad_norm": 1063831.625, "learning_rate": 4.368527405910583e-05, "loss": 4.8131, "step": 500 }, { "epoch": 0.12882040919424098, "grad_norm": 1552814.75, "learning_rate": 4.355897954028795e-05, "loss": 4.6981, "step": 510 }, { "epoch": 0.13134629957059862, "grad_norm": 915990.25, "learning_rate": 4.343268502147007e-05, "loss": 4.8095, "step": 520 }, { "epoch": 0.1338721899469563, "grad_norm": 838948.625, "learning_rate": 4.330639050265219e-05, "loss": 4.8535, "step": 530 }, { "epoch": 0.13639808032331396, "grad_norm": 697365.9375, "learning_rate": 4.31800959838343e-05, "loss": 5.0165, "step": 540 }, { "epoch": 0.13892397069967163, "grad_norm": 853774.25, "learning_rate": 4.305380146501642e-05, "loss": 4.6141, "step": 550 }, { "epoch": 0.1414498610760293, "grad_norm": 714720.875, "learning_rate": 4.292750694619854e-05, "loss": 4.6416, "step": 560 }, { "epoch": 0.14397575145238697, "grad_norm": 953777.125, "learning_rate": 4.280121242738066e-05, "loss": 4.8611, "step": 570 }, { "epoch": 0.14650164182874464, "grad_norm": 1033054.5, "learning_rate": 4.2674917908562776e-05, "loss": 4.2724, "step": 580 }, { "epoch": 0.1490275322051023, "grad_norm": 877196.1875, "learning_rate": 4.254862338974489e-05, "loss": 4.7521, "step": 590 }, { "epoch": 0.15155342258145996, "grad_norm": 1127457.0, "learning_rate": 4.2422328870927e-05, "loss": 4.3, "step": 600 }, { "epoch": 0.15407931295781763, "grad_norm": 627419.4375, "learning_rate": 4.229603435210912e-05, "loss": 4.5331, "step": 610 }, { "epoch": 0.1566052033341753, "grad_norm": 691871.3125, "learning_rate": 4.216973983329124e-05, "loss": 4.5695, "step": 620 }, { "epoch": 0.15913109371053297, "grad_norm": 637530.0625, "learning_rate": 4.2043445314473356e-05, "loss": 4.4994, "step": 630 }, { "epoch": 0.16165698408689064, "grad_norm": 692771.5625, "learning_rate": 4.191715079565547e-05, "loss": 4.403, "step": 640 }, { "epoch": 0.1641828744632483, "grad_norm": 738835.1875, "learning_rate": 4.179085627683759e-05, "loss": 4.5743, "step": 650 }, { "epoch": 0.16670876483960595, "grad_norm": 1637430.25, "learning_rate": 4.1664561758019706e-05, "loss": 4.5377, "step": 660 }, { "epoch": 0.16923465521596362, "grad_norm": 760962.875, "learning_rate": 4.1538267239201825e-05, "loss": 4.5168, "step": 670 }, { "epoch": 0.1717605455923213, "grad_norm": 1142129.25, "learning_rate": 4.1411972720383937e-05, "loss": 4.2554, "step": 680 }, { "epoch": 0.17428643596867896, "grad_norm": 614473.6875, "learning_rate": 4.1285678201566055e-05, "loss": 4.8585, "step": 690 }, { "epoch": 0.17681232634503663, "grad_norm": 1348435.875, "learning_rate": 4.115938368274817e-05, "loss": 4.4252, "step": 700 }, { "epoch": 0.1793382167213943, "grad_norm": 1183288.25, "learning_rate": 4.1033089163930286e-05, "loss": 4.4951, "step": 710 }, { "epoch": 0.18186410709775197, "grad_norm": 579328.4375, "learning_rate": 4.0906794645112405e-05, "loss": 4.322, "step": 720 }, { "epoch": 0.1843899974741096, "grad_norm": 1145040.25, "learning_rate": 4.078050012629452e-05, "loss": 4.1513, "step": 730 }, { "epoch": 0.18691588785046728, "grad_norm": 1142186.875, "learning_rate": 4.0654205607476636e-05, "loss": 4.1724, "step": 740 }, { "epoch": 0.18944177822682495, "grad_norm": 1073306.125, "learning_rate": 4.0527911088658754e-05, "loss": 4.2603, "step": 750 }, { "epoch": 0.19196766860318262, "grad_norm": 856185.875, "learning_rate": 4.040161656984087e-05, "loss": 4.1743, "step": 760 }, { "epoch": 0.1944935589795403, "grad_norm": 1259884.75, "learning_rate": 4.027532205102299e-05, "loss": 4.4978, "step": 770 }, { "epoch": 0.19701944935589796, "grad_norm": 988211.375, "learning_rate": 4.0149027532205104e-05, "loss": 4.4756, "step": 780 }, { "epoch": 0.19954533973225563, "grad_norm": 1129890.375, "learning_rate": 4.002273301338722e-05, "loss": 4.4766, "step": 790 }, { "epoch": 0.20207123010861328, "grad_norm": 1127913.0, "learning_rate": 3.989643849456934e-05, "loss": 4.1696, "step": 800 }, { "epoch": 0.20459712048497095, "grad_norm": 643195.0625, "learning_rate": 3.9770143975751453e-05, "loss": 4.2825, "step": 810 }, { "epoch": 0.20712301086132862, "grad_norm": 585030.1875, "learning_rate": 3.9643849456933565e-05, "loss": 4.5449, "step": 820 }, { "epoch": 0.20964890123768629, "grad_norm": 964589.9375, "learning_rate": 3.9517554938115684e-05, "loss": 4.2085, "step": 830 }, { "epoch": 0.21217479161404396, "grad_norm": 1307139.375, "learning_rate": 3.93912604192978e-05, "loss": 4.2636, "step": 840 }, { "epoch": 0.21470068199040163, "grad_norm": 1642451.125, "learning_rate": 3.926496590047992e-05, "loss": 4.1194, "step": 850 }, { "epoch": 0.2172265723667593, "grad_norm": 767172.5625, "learning_rate": 3.913867138166204e-05, "loss": 3.9216, "step": 860 }, { "epoch": 0.21975246274311694, "grad_norm": 1053997.0, "learning_rate": 3.901237686284415e-05, "loss": 4.4344, "step": 870 }, { "epoch": 0.2222783531194746, "grad_norm": 1236466.5, "learning_rate": 3.888608234402627e-05, "loss": 4.5718, "step": 880 }, { "epoch": 0.22480424349583228, "grad_norm": 1578047.5, "learning_rate": 3.875978782520839e-05, "loss": 4.2614, "step": 890 }, { "epoch": 0.22733013387218995, "grad_norm": 1341923.75, "learning_rate": 3.863349330639051e-05, "loss": 4.2356, "step": 900 }, { "epoch": 0.22985602424854762, "grad_norm": 1259278.75, "learning_rate": 3.850719878757262e-05, "loss": 4.1374, "step": 910 }, { "epoch": 0.2323819146249053, "grad_norm": 1208085.125, "learning_rate": 3.838090426875473e-05, "loss": 4.3741, "step": 920 }, { "epoch": 0.23490780500126293, "grad_norm": 713612.375, "learning_rate": 3.825460974993685e-05, "loss": 4.2538, "step": 930 }, { "epoch": 0.2374336953776206, "grad_norm": 773853.25, "learning_rate": 3.812831523111897e-05, "loss": 4.0877, "step": 940 }, { "epoch": 0.23995958575397827, "grad_norm": 1925775.125, "learning_rate": 3.800202071230109e-05, "loss": 3.8277, "step": 950 }, { "epoch": 0.24248547613033594, "grad_norm": 814602.125, "learning_rate": 3.78757261934832e-05, "loss": 4.016, "step": 960 }, { "epoch": 0.2450113665066936, "grad_norm": 1109952.0, "learning_rate": 3.774943167466532e-05, "loss": 3.9932, "step": 970 }, { "epoch": 0.24753725688305128, "grad_norm": 744110.125, "learning_rate": 3.762313715584744e-05, "loss": 4.0229, "step": 980 }, { "epoch": 0.2500631472594089, "grad_norm": 904255.75, "learning_rate": 3.749684263702956e-05, "loss": 4.4272, "step": 990 }, { "epoch": 0.2525890376357666, "grad_norm": 829616.3125, "learning_rate": 3.7370548118211676e-05, "loss": 4.3034, "step": 1000 }, { "epoch": 0.25511492801212426, "grad_norm": 1236616.375, "learning_rate": 3.724425359939379e-05, "loss": 3.9168, "step": 1010 }, { "epoch": 0.25764081838848196, "grad_norm": 696752.75, "learning_rate": 3.711795908057591e-05, "loss": 4.2012, "step": 1020 }, { "epoch": 0.2601667087648396, "grad_norm": 680934.875, "learning_rate": 3.699166456175802e-05, "loss": 4.2596, "step": 1030 }, { "epoch": 0.26269259914119725, "grad_norm": 1334373.125, "learning_rate": 3.686537004294014e-05, "loss": 4.461, "step": 1040 }, { "epoch": 0.26521848951755495, "grad_norm": 1002765.875, "learning_rate": 3.6739075524122257e-05, "loss": 4.2859, "step": 1050 }, { "epoch": 0.2677443798939126, "grad_norm": 507567.5, "learning_rate": 3.661278100530437e-05, "loss": 3.8771, "step": 1060 }, { "epoch": 0.2702702702702703, "grad_norm": 968382.5, "learning_rate": 3.648648648648649e-05, "loss": 4.1681, "step": 1070 }, { "epoch": 0.2727961606466279, "grad_norm": 1306089.125, "learning_rate": 3.6360191967668606e-05, "loss": 4.2425, "step": 1080 }, { "epoch": 0.2753220510229856, "grad_norm": 1039757.5, "learning_rate": 3.6233897448850725e-05, "loss": 4.3264, "step": 1090 }, { "epoch": 0.27784794139934327, "grad_norm": 1464928.5, "learning_rate": 3.610760293003284e-05, "loss": 4.2827, "step": 1100 }, { "epoch": 0.2803738317757009, "grad_norm": 824237.6875, "learning_rate": 3.5981308411214956e-05, "loss": 3.8312, "step": 1110 }, { "epoch": 0.2828997221520586, "grad_norm": 942513.125, "learning_rate": 3.5855013892397074e-05, "loss": 4.1363, "step": 1120 }, { "epoch": 0.28542561252841625, "grad_norm": 697307.1875, "learning_rate": 3.5728719373579186e-05, "loss": 3.9892, "step": 1130 }, { "epoch": 0.28795150290477395, "grad_norm": 1249596.5, "learning_rate": 3.5602424854761305e-05, "loss": 4.0763, "step": 1140 }, { "epoch": 0.2904773932811316, "grad_norm": 963218.0625, "learning_rate": 3.547613033594342e-05, "loss": 4.1946, "step": 1150 }, { "epoch": 0.2930032836574893, "grad_norm": 1025589.625, "learning_rate": 3.5349835817125536e-05, "loss": 4.1447, "step": 1160 }, { "epoch": 0.29552917403384693, "grad_norm": 696336.3125, "learning_rate": 3.5223541298307655e-05, "loss": 4.1982, "step": 1170 }, { "epoch": 0.2980550644102046, "grad_norm": 826398.875, "learning_rate": 3.5097246779489773e-05, "loss": 4.1935, "step": 1180 }, { "epoch": 0.30058095478656227, "grad_norm": 527454.375, "learning_rate": 3.497095226067189e-05, "loss": 3.8515, "step": 1190 }, { "epoch": 0.3031068451629199, "grad_norm": 1154428.125, "learning_rate": 3.4844657741854004e-05, "loss": 4.123, "step": 1200 }, { "epoch": 0.3056327355392776, "grad_norm": 843105.25, "learning_rate": 3.471836322303612e-05, "loss": 4.0649, "step": 1210 }, { "epoch": 0.30815862591563525, "grad_norm": 677728.0, "learning_rate": 3.459206870421824e-05, "loss": 4.1858, "step": 1220 }, { "epoch": 0.31068451629199295, "grad_norm": 843633.8125, "learning_rate": 3.4465774185400354e-05, "loss": 3.9904, "step": 1230 }, { "epoch": 0.3132104066683506, "grad_norm": 1079652.625, "learning_rate": 3.433947966658247e-05, "loss": 4.1296, "step": 1240 }, { "epoch": 0.31573629704470824, "grad_norm": 1375458.875, "learning_rate": 3.4213185147764585e-05, "loss": 3.845, "step": 1250 }, { "epoch": 0.31826218742106593, "grad_norm": 1572799.75, "learning_rate": 3.40868906289467e-05, "loss": 3.8965, "step": 1260 }, { "epoch": 0.3207880777974236, "grad_norm": 626954.0625, "learning_rate": 3.396059611012882e-05, "loss": 4.2417, "step": 1270 }, { "epoch": 0.3233139681737813, "grad_norm": 778080.8125, "learning_rate": 3.383430159131094e-05, "loss": 4.0585, "step": 1280 }, { "epoch": 0.3258398585501389, "grad_norm": 1268475.375, "learning_rate": 3.370800707249305e-05, "loss": 3.5957, "step": 1290 }, { "epoch": 0.3283657489264966, "grad_norm": 983964.5, "learning_rate": 3.358171255367517e-05, "loss": 4.4113, "step": 1300 }, { "epoch": 0.33089163930285426, "grad_norm": 1136581.5, "learning_rate": 3.345541803485729e-05, "loss": 4.2545, "step": 1310 }, { "epoch": 0.3334175296792119, "grad_norm": 1313611.375, "learning_rate": 3.332912351603941e-05, "loss": 4.0466, "step": 1320 }, { "epoch": 0.3359434200555696, "grad_norm": 595656.0625, "learning_rate": 3.320282899722152e-05, "loss": 4.1121, "step": 1330 }, { "epoch": 0.33846931043192724, "grad_norm": 1134474.625, "learning_rate": 3.307653447840364e-05, "loss": 3.8602, "step": 1340 }, { "epoch": 0.34099520080828494, "grad_norm": 932185.5, "learning_rate": 3.295023995958575e-05, "loss": 4.0099, "step": 1350 }, { "epoch": 0.3435210911846426, "grad_norm": 992654.875, "learning_rate": 3.282394544076787e-05, "loss": 3.8907, "step": 1360 }, { "epoch": 0.3460469815610003, "grad_norm": 798724.1875, "learning_rate": 3.269765092194999e-05, "loss": 3.4421, "step": 1370 }, { "epoch": 0.3485728719373579, "grad_norm": 1000527.125, "learning_rate": 3.257135640313211e-05, "loss": 4.0111, "step": 1380 }, { "epoch": 0.35109876231371556, "grad_norm": 1207395.875, "learning_rate": 3.244506188431422e-05, "loss": 4.332, "step": 1390 }, { "epoch": 0.35362465269007326, "grad_norm": 736049.8125, "learning_rate": 3.231876736549634e-05, "loss": 4.0475, "step": 1400 }, { "epoch": 0.3561505430664309, "grad_norm": 902288.75, "learning_rate": 3.219247284667846e-05, "loss": 3.7045, "step": 1410 }, { "epoch": 0.3586764334427886, "grad_norm": 1179507.375, "learning_rate": 3.2066178327860577e-05, "loss": 3.7423, "step": 1420 }, { "epoch": 0.36120232381914624, "grad_norm": 1319296.25, "learning_rate": 3.193988380904269e-05, "loss": 3.7602, "step": 1430 }, { "epoch": 0.36372821419550394, "grad_norm": 1040202.3125, "learning_rate": 3.181358929022481e-05, "loss": 3.871, "step": 1440 }, { "epoch": 0.3662541045718616, "grad_norm": 892262.9375, "learning_rate": 3.168729477140692e-05, "loss": 4.1026, "step": 1450 }, { "epoch": 0.3687799949482192, "grad_norm": 1401292.5, "learning_rate": 3.156100025258904e-05, "loss": 3.9609, "step": 1460 }, { "epoch": 0.3713058853245769, "grad_norm": 1242199.5, "learning_rate": 3.143470573377116e-05, "loss": 3.9826, "step": 1470 }, { "epoch": 0.37383177570093457, "grad_norm": 953926.4375, "learning_rate": 3.130841121495327e-05, "loss": 3.7857, "step": 1480 }, { "epoch": 0.37635766607729226, "grad_norm": 1005284.125, "learning_rate": 3.118211669613539e-05, "loss": 3.8462, "step": 1490 }, { "epoch": 0.3788835564536499, "grad_norm": 617377.0, "learning_rate": 3.1055822177317506e-05, "loss": 4.0298, "step": 1500 }, { "epoch": 0.3814094468300076, "grad_norm": 1349709.0, "learning_rate": 3.0929527658499625e-05, "loss": 4.108, "step": 1510 }, { "epoch": 0.38393533720636525, "grad_norm": 863097.375, "learning_rate": 3.0803233139681744e-05, "loss": 3.7545, "step": 1520 }, { "epoch": 0.3864612275827229, "grad_norm": 1074415.75, "learning_rate": 3.0676938620863856e-05, "loss": 3.7492, "step": 1530 }, { "epoch": 0.3889871179590806, "grad_norm": 1269458.125, "learning_rate": 3.0550644102045975e-05, "loss": 4.0702, "step": 1540 }, { "epoch": 0.39151300833543823, "grad_norm": 623983.6875, "learning_rate": 3.042434958322809e-05, "loss": 3.7522, "step": 1550 }, { "epoch": 0.3940388987117959, "grad_norm": 847152.8125, "learning_rate": 3.029805506441021e-05, "loss": 3.9036, "step": 1560 }, { "epoch": 0.39656478908815357, "grad_norm": 1389634.5, "learning_rate": 3.017176054559232e-05, "loss": 3.7832, "step": 1570 }, { "epoch": 0.39909067946451127, "grad_norm": 983121.3125, "learning_rate": 3.004546602677444e-05, "loss": 3.6402, "step": 1580 }, { "epoch": 0.4016165698408689, "grad_norm": 1020898.8125, "learning_rate": 2.9919171507956555e-05, "loss": 3.6429, "step": 1590 }, { "epoch": 0.40414246021722655, "grad_norm": 847203.9375, "learning_rate": 2.9792876989138674e-05, "loss": 3.8984, "step": 1600 }, { "epoch": 0.40666835059358425, "grad_norm": 1133185.5, "learning_rate": 2.9666582470320793e-05, "loss": 3.7906, "step": 1610 }, { "epoch": 0.4091942409699419, "grad_norm": 1127204.625, "learning_rate": 2.9540287951502905e-05, "loss": 4.143, "step": 1620 }, { "epoch": 0.4117201313462996, "grad_norm": 1315488.625, "learning_rate": 2.9413993432685023e-05, "loss": 3.8023, "step": 1630 }, { "epoch": 0.41424602172265723, "grad_norm": 776120.5625, "learning_rate": 2.928769891386714e-05, "loss": 3.4144, "step": 1640 }, { "epoch": 0.41677191209901493, "grad_norm": 832920.9375, "learning_rate": 2.9161404395049257e-05, "loss": 3.6427, "step": 1650 }, { "epoch": 0.41929780247537257, "grad_norm": 1192992.25, "learning_rate": 2.9035109876231376e-05, "loss": 3.9277, "step": 1660 }, { "epoch": 0.4218236928517302, "grad_norm": 918210.4375, "learning_rate": 2.8908815357413488e-05, "loss": 3.6729, "step": 1670 }, { "epoch": 0.4243495832280879, "grad_norm": 924215.25, "learning_rate": 2.8782520838595607e-05, "loss": 3.7461, "step": 1680 }, { "epoch": 0.42687547360444555, "grad_norm": 715260.8125, "learning_rate": 2.8656226319777722e-05, "loss": 3.7783, "step": 1690 }, { "epoch": 0.42940136398080325, "grad_norm": 1466543.5, "learning_rate": 2.852993180095984e-05, "loss": 3.8766, "step": 1700 }, { "epoch": 0.4319272543571609, "grad_norm": 669040.375, "learning_rate": 2.8403637282141953e-05, "loss": 3.9012, "step": 1710 }, { "epoch": 0.4344531447335186, "grad_norm": 2193973.75, "learning_rate": 2.8277342763324072e-05, "loss": 3.3848, "step": 1720 }, { "epoch": 0.43697903510987623, "grad_norm": 831080.625, "learning_rate": 2.815104824450619e-05, "loss": 3.6074, "step": 1730 }, { "epoch": 0.4395049254862339, "grad_norm": 960923.75, "learning_rate": 2.8024753725688306e-05, "loss": 3.6234, "step": 1740 }, { "epoch": 0.4420308158625916, "grad_norm": 1489447.5, "learning_rate": 2.7898459206870425e-05, "loss": 3.6699, "step": 1750 }, { "epoch": 0.4445567062389492, "grad_norm": 611681.375, "learning_rate": 2.7772164688052537e-05, "loss": 3.389, "step": 1760 }, { "epoch": 0.4470825966153069, "grad_norm": 2048307.875, "learning_rate": 2.7645870169234656e-05, "loss": 3.8727, "step": 1770 }, { "epoch": 0.44960848699166456, "grad_norm": 855010.75, "learning_rate": 2.7519575650416774e-05, "loss": 3.839, "step": 1780 }, { "epoch": 0.45213437736802226, "grad_norm": 954567.125, "learning_rate": 2.739328113159889e-05, "loss": 3.507, "step": 1790 }, { "epoch": 0.4546602677443799, "grad_norm": 1057372.25, "learning_rate": 2.726698661278101e-05, "loss": 4.0177, "step": 1800 }, { "epoch": 0.45718615812073754, "grad_norm": 1196149.375, "learning_rate": 2.714069209396312e-05, "loss": 3.6044, "step": 1810 }, { "epoch": 0.45971204849709524, "grad_norm": 664877.5, "learning_rate": 2.701439757514524e-05, "loss": 3.7145, "step": 1820 }, { "epoch": 0.4622379388734529, "grad_norm": 1227852.0, "learning_rate": 2.6888103056327358e-05, "loss": 3.7588, "step": 1830 }, { "epoch": 0.4647638292498106, "grad_norm": 1403871.0, "learning_rate": 2.6761808537509473e-05, "loss": 3.672, "step": 1840 }, { "epoch": 0.4672897196261682, "grad_norm": 1831221.875, "learning_rate": 2.663551401869159e-05, "loss": 3.8373, "step": 1850 }, { "epoch": 0.46981561000252586, "grad_norm": 1008128.75, "learning_rate": 2.6509219499873704e-05, "loss": 3.6977, "step": 1860 }, { "epoch": 0.47234150037888356, "grad_norm": 704558.8125, "learning_rate": 2.6382924981055823e-05, "loss": 3.9759, "step": 1870 }, { "epoch": 0.4748673907552412, "grad_norm": 1501172.875, "learning_rate": 2.6256630462237942e-05, "loss": 3.3989, "step": 1880 }, { "epoch": 0.4773932811315989, "grad_norm": 1022078.375, "learning_rate": 2.613033594342006e-05, "loss": 3.7731, "step": 1890 }, { "epoch": 0.47991917150795654, "grad_norm": 892236.75, "learning_rate": 2.6004041424602173e-05, "loss": 3.6186, "step": 1900 }, { "epoch": 0.48244506188431424, "grad_norm": 592257.875, "learning_rate": 2.5877746905784288e-05, "loss": 3.3699, "step": 1910 }, { "epoch": 0.4849709522606719, "grad_norm": 761213.9375, "learning_rate": 2.5751452386966407e-05, "loss": 3.6678, "step": 1920 }, { "epoch": 0.4874968426370295, "grad_norm": 1076466.0, "learning_rate": 2.5625157868148525e-05, "loss": 3.4251, "step": 1930 }, { "epoch": 0.4900227330133872, "grad_norm": 1370577.375, "learning_rate": 2.5498863349330644e-05, "loss": 3.6471, "step": 1940 }, { "epoch": 0.49254862338974487, "grad_norm": 682627.5625, "learning_rate": 2.5372568830512756e-05, "loss": 3.4545, "step": 1950 }, { "epoch": 0.49507451376610256, "grad_norm": 812808.125, "learning_rate": 2.524627431169487e-05, "loss": 3.64, "step": 1960 }, { "epoch": 0.4976004041424602, "grad_norm": 1190569.0, "learning_rate": 2.511997979287699e-05, "loss": 3.5669, "step": 1970 }, { "epoch": 0.5001262945188178, "grad_norm": 1053372.125, "learning_rate": 2.4993685274059106e-05, "loss": 3.561, "step": 1980 }, { "epoch": 0.5026521848951756, "grad_norm": 1154567.75, "learning_rate": 2.4867390755241225e-05, "loss": 3.7747, "step": 1990 }, { "epoch": 0.5051780752715332, "grad_norm": 1105378.875, "learning_rate": 2.4741096236423343e-05, "loss": 3.5388, "step": 2000 }, { "epoch": 0.5077039656478909, "grad_norm": 975523.9375, "learning_rate": 2.4614801717605455e-05, "loss": 3.7567, "step": 2010 }, { "epoch": 0.5102298560242485, "grad_norm": 1316863.5, "learning_rate": 2.4488507198787574e-05, "loss": 3.4911, "step": 2020 }, { "epoch": 0.5127557464006062, "grad_norm": 1235055.375, "learning_rate": 2.436221267996969e-05, "loss": 3.3298, "step": 2030 }, { "epoch": 0.5152816367769639, "grad_norm": 1190251.5, "learning_rate": 2.4235918161151808e-05, "loss": 3.7792, "step": 2040 }, { "epoch": 0.5178075271533216, "grad_norm": 1828954.875, "learning_rate": 2.4109623642333924e-05, "loss": 3.3407, "step": 2050 }, { "epoch": 0.5203334175296792, "grad_norm": 734125.1875, "learning_rate": 2.398332912351604e-05, "loss": 3.6435, "step": 2060 }, { "epoch": 0.5228593079060369, "grad_norm": 1114272.0, "learning_rate": 2.3857034604698158e-05, "loss": 3.603, "step": 2070 }, { "epoch": 0.5253851982823945, "grad_norm": 1464147.0, "learning_rate": 2.3730740085880273e-05, "loss": 3.6587, "step": 2080 }, { "epoch": 0.5279110886587522, "grad_norm": 2074921.0, "learning_rate": 2.3604445567062392e-05, "loss": 3.7505, "step": 2090 }, { "epoch": 0.5304369790351099, "grad_norm": 1063793.625, "learning_rate": 2.3478151048244507e-05, "loss": 3.8397, "step": 2100 }, { "epoch": 0.5329628694114675, "grad_norm": 702127.6875, "learning_rate": 2.3351856529426626e-05, "loss": 3.4248, "step": 2110 }, { "epoch": 0.5354887597878252, "grad_norm": 1237508.5, "learning_rate": 2.3225562010608738e-05, "loss": 3.6119, "step": 2120 }, { "epoch": 0.5380146501641829, "grad_norm": 678885.625, "learning_rate": 2.3099267491790857e-05, "loss": 3.4523, "step": 2130 }, { "epoch": 0.5405405405405406, "grad_norm": 1506679.875, "learning_rate": 2.2972972972972976e-05, "loss": 3.4496, "step": 2140 }, { "epoch": 0.5430664309168982, "grad_norm": 1016625.6875, "learning_rate": 2.284667845415509e-05, "loss": 3.6607, "step": 2150 }, { "epoch": 0.5455923212932559, "grad_norm": 916826.3125, "learning_rate": 2.272038393533721e-05, "loss": 3.6147, "step": 2160 }, { "epoch": 0.5481182116696135, "grad_norm": 1150088.625, "learning_rate": 2.2594089416519322e-05, "loss": 3.3045, "step": 2170 }, { "epoch": 0.5506441020459713, "grad_norm": 1215418.875, "learning_rate": 2.246779489770144e-05, "loss": 3.6112, "step": 2180 }, { "epoch": 0.5531699924223289, "grad_norm": 610014.5, "learning_rate": 2.2341500378883556e-05, "loss": 3.5236, "step": 2190 }, { "epoch": 0.5556958827986865, "grad_norm": 834213.625, "learning_rate": 2.2215205860065675e-05, "loss": 3.4063, "step": 2200 }, { "epoch": 0.5582217731750442, "grad_norm": 976218.8125, "learning_rate": 2.2088911341247793e-05, "loss": 3.311, "step": 2210 }, { "epoch": 0.5607476635514018, "grad_norm": 940372.0, "learning_rate": 2.196261682242991e-05, "loss": 3.6945, "step": 2220 }, { "epoch": 0.5632735539277596, "grad_norm": 903718.9375, "learning_rate": 2.1836322303612024e-05, "loss": 3.3955, "step": 2230 }, { "epoch": 0.5657994443041172, "grad_norm": 791363.5, "learning_rate": 2.171002778479414e-05, "loss": 3.5137, "step": 2240 }, { "epoch": 0.5683253346804749, "grad_norm": 958748.625, "learning_rate": 2.158373326597626e-05, "loss": 3.3765, "step": 2250 }, { "epoch": 0.5708512250568325, "grad_norm": 1431520.25, "learning_rate": 2.1457438747158374e-05, "loss": 3.1066, "step": 2260 }, { "epoch": 0.5733771154331903, "grad_norm": 793628.1875, "learning_rate": 2.1331144228340493e-05, "loss": 3.5486, "step": 2270 }, { "epoch": 0.5759030058095479, "grad_norm": 935731.3125, "learning_rate": 2.1204849709522608e-05, "loss": 3.427, "step": 2280 }, { "epoch": 0.5784288961859055, "grad_norm": 1024607.625, "learning_rate": 2.1078555190704723e-05, "loss": 3.2279, "step": 2290 }, { "epoch": 0.5809547865622632, "grad_norm": 1027219.3125, "learning_rate": 2.0952260671886842e-05, "loss": 3.7803, "step": 2300 }, { "epoch": 0.5834806769386208, "grad_norm": 762017.8125, "learning_rate": 2.0825966153068957e-05, "loss": 3.7272, "step": 2310 }, { "epoch": 0.5860065673149786, "grad_norm": 1251829.125, "learning_rate": 2.0699671634251076e-05, "loss": 3.6896, "step": 2320 }, { "epoch": 0.5885324576913362, "grad_norm": 841272.5, "learning_rate": 2.057337711543319e-05, "loss": 3.3452, "step": 2330 }, { "epoch": 0.5910583480676939, "grad_norm": 1126517.625, "learning_rate": 2.0447082596615307e-05, "loss": 3.386, "step": 2340 }, { "epoch": 0.5935842384440515, "grad_norm": 982217.3125, "learning_rate": 2.0320788077797426e-05, "loss": 3.7347, "step": 2350 }, { "epoch": 0.5961101288204091, "grad_norm": 644312.75, "learning_rate": 2.019449355897954e-05, "loss": 3.1686, "step": 2360 }, { "epoch": 0.5986360191967669, "grad_norm": 1135732.625, "learning_rate": 2.006819904016166e-05, "loss": 3.6648, "step": 2370 }, { "epoch": 0.6011619095731245, "grad_norm": 886122.0, "learning_rate": 1.9941904521343775e-05, "loss": 3.3114, "step": 2380 }, { "epoch": 0.6036877999494822, "grad_norm": 1676023.875, "learning_rate": 1.981561000252589e-05, "loss": 3.8232, "step": 2390 }, { "epoch": 0.6062136903258398, "grad_norm": 1056139.0, "learning_rate": 1.9689315483708006e-05, "loss": 3.3829, "step": 2400 }, { "epoch": 0.6087395807021975, "grad_norm": 1303276.625, "learning_rate": 1.9563020964890125e-05, "loss": 3.7356, "step": 2410 }, { "epoch": 0.6112654710785552, "grad_norm": 1168222.5, "learning_rate": 1.9436726446072244e-05, "loss": 3.4672, "step": 2420 }, { "epoch": 0.6137913614549129, "grad_norm": 974863.5, "learning_rate": 1.931043192725436e-05, "loss": 3.7479, "step": 2430 }, { "epoch": 0.6163172518312705, "grad_norm": 727352.375, "learning_rate": 1.9184137408436474e-05, "loss": 3.4183, "step": 2440 }, { "epoch": 0.6188431422076281, "grad_norm": 1103530.0, "learning_rate": 1.905784288961859e-05, "loss": 3.5818, "step": 2450 }, { "epoch": 0.6213690325839859, "grad_norm": 866723.3125, "learning_rate": 1.893154837080071e-05, "loss": 3.4123, "step": 2460 }, { "epoch": 0.6238949229603435, "grad_norm": 1022545.8125, "learning_rate": 1.8805253851982824e-05, "loss": 3.1739, "step": 2470 }, { "epoch": 0.6264208133367012, "grad_norm": 1114618.75, "learning_rate": 1.8678959333164943e-05, "loss": 3.6795, "step": 2480 }, { "epoch": 0.6289467037130588, "grad_norm": 1178612.875, "learning_rate": 1.8552664814347058e-05, "loss": 3.3061, "step": 2490 }, { "epoch": 0.6314725940894165, "grad_norm": 757466.0, "learning_rate": 1.8426370295529173e-05, "loss": 3.4692, "step": 2500 }, { "epoch": 0.6339984844657742, "grad_norm": 932833.75, "learning_rate": 1.8300075776711292e-05, "loss": 3.3951, "step": 2510 }, { "epoch": 0.6365243748421319, "grad_norm": 811280.75, "learning_rate": 1.8173781257893408e-05, "loss": 3.42, "step": 2520 }, { "epoch": 0.6390502652184895, "grad_norm": 828370.875, "learning_rate": 1.8047486739075526e-05, "loss": 3.5068, "step": 2530 }, { "epoch": 0.6415761555948472, "grad_norm": 796540.6875, "learning_rate": 1.7921192220257642e-05, "loss": 3.4725, "step": 2540 }, { "epoch": 0.6441020459712048, "grad_norm": 1101170.0, "learning_rate": 1.7794897701439757e-05, "loss": 3.5315, "step": 2550 }, { "epoch": 0.6466279363475625, "grad_norm": 1084591.375, "learning_rate": 1.7668603182621876e-05, "loss": 3.3357, "step": 2560 }, { "epoch": 0.6491538267239202, "grad_norm": 846566.375, "learning_rate": 1.754230866380399e-05, "loss": 3.4352, "step": 2570 }, { "epoch": 0.6516797171002778, "grad_norm": 817152.9375, "learning_rate": 1.741601414498611e-05, "loss": 3.3913, "step": 2580 }, { "epoch": 0.6542056074766355, "grad_norm": 896931.875, "learning_rate": 1.7289719626168225e-05, "loss": 3.3725, "step": 2590 }, { "epoch": 0.6567314978529932, "grad_norm": 580791.0, "learning_rate": 1.716342510735034e-05, "loss": 3.3803, "step": 2600 }, { "epoch": 0.6592573882293509, "grad_norm": 817709.375, "learning_rate": 1.7037130588532456e-05, "loss": 3.4564, "step": 2610 }, { "epoch": 0.6617832786057085, "grad_norm": 1013458.75, "learning_rate": 1.6910836069714575e-05, "loss": 3.4659, "step": 2620 }, { "epoch": 0.6643091689820662, "grad_norm": 855889.0625, "learning_rate": 1.6784541550896694e-05, "loss": 3.1951, "step": 2630 }, { "epoch": 0.6668350593584238, "grad_norm": 1114807.125, "learning_rate": 1.665824703207881e-05, "loss": 3.2604, "step": 2640 }, { "epoch": 0.6693609497347816, "grad_norm": 1472619.75, "learning_rate": 1.6531952513260925e-05, "loss": 3.2474, "step": 2650 }, { "epoch": 0.6718868401111392, "grad_norm": 933670.625, "learning_rate": 1.640565799444304e-05, "loss": 3.2033, "step": 2660 }, { "epoch": 0.6744127304874968, "grad_norm": 752061.375, "learning_rate": 1.627936347562516e-05, "loss": 3.5689, "step": 2670 }, { "epoch": 0.6769386208638545, "grad_norm": 1359021.375, "learning_rate": 1.6153068956807274e-05, "loss": 3.289, "step": 2680 }, { "epoch": 0.6794645112402121, "grad_norm": 755337.0625, "learning_rate": 1.6026774437989393e-05, "loss": 3.4877, "step": 2690 }, { "epoch": 0.6819904016165699, "grad_norm": 797617.3125, "learning_rate": 1.5900479919171508e-05, "loss": 3.4128, "step": 2700 }, { "epoch": 0.6845162919929275, "grad_norm": 754639.75, "learning_rate": 1.5774185400353624e-05, "loss": 3.4954, "step": 2710 }, { "epoch": 0.6870421823692852, "grad_norm": 853287.1875, "learning_rate": 1.5647890881535742e-05, "loss": 3.2564, "step": 2720 }, { "epoch": 0.6895680727456428, "grad_norm": 554583.1875, "learning_rate": 1.5521596362717858e-05, "loss": 3.2221, "step": 2730 }, { "epoch": 0.6920939631220006, "grad_norm": 722306.8125, "learning_rate": 1.5395301843899977e-05, "loss": 3.1547, "step": 2740 }, { "epoch": 0.6946198534983582, "grad_norm": 833465.875, "learning_rate": 1.5269007325082092e-05, "loss": 3.1672, "step": 2750 }, { "epoch": 0.6971457438747158, "grad_norm": 801353.3125, "learning_rate": 1.5142712806264209e-05, "loss": 3.4491, "step": 2760 }, { "epoch": 0.6996716342510735, "grad_norm": 621898.875, "learning_rate": 1.5016418287446326e-05, "loss": 3.4968, "step": 2770 }, { "epoch": 0.7021975246274311, "grad_norm": 926238.0625, "learning_rate": 1.4890123768628441e-05, "loss": 3.5613, "step": 2780 }, { "epoch": 0.7047234150037889, "grad_norm": 991762.4375, "learning_rate": 1.476382924981056e-05, "loss": 3.1534, "step": 2790 }, { "epoch": 0.7072493053801465, "grad_norm": 876887.3125, "learning_rate": 1.4637534730992674e-05, "loss": 3.1666, "step": 2800 }, { "epoch": 0.7097751957565042, "grad_norm": 748469.75, "learning_rate": 1.4511240212174793e-05, "loss": 3.3201, "step": 2810 }, { "epoch": 0.7123010861328618, "grad_norm": 913840.125, "learning_rate": 1.4384945693356908e-05, "loss": 3.3263, "step": 2820 }, { "epoch": 0.7148269765092194, "grad_norm": 776508.25, "learning_rate": 1.4258651174539025e-05, "loss": 3.3736, "step": 2830 }, { "epoch": 0.7173528668855772, "grad_norm": 1003093.6875, "learning_rate": 1.4132356655721144e-05, "loss": 3.3369, "step": 2840 }, { "epoch": 0.7198787572619348, "grad_norm": 480733.9375, "learning_rate": 1.400606213690326e-05, "loss": 3.4599, "step": 2850 }, { "epoch": 0.7224046476382925, "grad_norm": 921517.8125, "learning_rate": 1.3879767618085376e-05, "loss": 3.3327, "step": 2860 }, { "epoch": 0.7249305380146501, "grad_norm": 1106659.0, "learning_rate": 1.3753473099267492e-05, "loss": 3.4347, "step": 2870 }, { "epoch": 0.7274564283910079, "grad_norm": 506619.90625, "learning_rate": 1.3627178580449609e-05, "loss": 3.2675, "step": 2880 }, { "epoch": 0.7299823187673655, "grad_norm": 862991.0625, "learning_rate": 1.3500884061631724e-05, "loss": 3.2461, "step": 2890 }, { "epoch": 0.7325082091437232, "grad_norm": 835895.8125, "learning_rate": 1.3374589542813843e-05, "loss": 3.3976, "step": 2900 }, { "epoch": 0.7350340995200808, "grad_norm": 699905.9375, "learning_rate": 1.324829502399596e-05, "loss": 3.4889, "step": 2910 }, { "epoch": 0.7375599898964385, "grad_norm": 824201.75, "learning_rate": 1.3122000505178075e-05, "loss": 3.548, "step": 2920 }, { "epoch": 0.7400858802727962, "grad_norm": 871363.75, "learning_rate": 1.2995705986360193e-05, "loss": 3.5843, "step": 2930 }, { "epoch": 0.7426117706491538, "grad_norm": 627847.1875, "learning_rate": 1.2869411467542308e-05, "loss": 3.1553, "step": 2940 }, { "epoch": 0.7451376610255115, "grad_norm": 758722.4375, "learning_rate": 1.2743116948724427e-05, "loss": 3.4285, "step": 2950 }, { "epoch": 0.7476635514018691, "grad_norm": 851608.8125, "learning_rate": 1.2616822429906542e-05, "loss": 3.3232, "step": 2960 }, { "epoch": 0.7501894417782268, "grad_norm": 917633.8125, "learning_rate": 1.249052791108866e-05, "loss": 3.1254, "step": 2970 }, { "epoch": 0.7527153321545845, "grad_norm": 1110027.375, "learning_rate": 1.2364233392270776e-05, "loss": 3.4867, "step": 2980 }, { "epoch": 0.7552412225309422, "grad_norm": 817171.625, "learning_rate": 1.2237938873452892e-05, "loss": 3.4251, "step": 2990 }, { "epoch": 0.7577671129072998, "grad_norm": 1074504.25, "learning_rate": 1.211164435463501e-05, "loss": 3.4581, "step": 3000 }, { "epoch": 0.7602930032836575, "grad_norm": 777395.4375, "learning_rate": 1.1985349835817126e-05, "loss": 3.3139, "step": 3010 }, { "epoch": 0.7628188936600152, "grad_norm": 657749.75, "learning_rate": 1.1859055316999243e-05, "loss": 3.2542, "step": 3020 }, { "epoch": 0.7653447840363728, "grad_norm": 536018.1875, "learning_rate": 1.173276079818136e-05, "loss": 3.4512, "step": 3030 }, { "epoch": 0.7678706744127305, "grad_norm": 1034483.5625, "learning_rate": 1.1606466279363475e-05, "loss": 3.3698, "step": 3040 }, { "epoch": 0.7703965647890881, "grad_norm": 1303588.5, "learning_rate": 1.1480171760545592e-05, "loss": 3.1435, "step": 3050 }, { "epoch": 0.7729224551654458, "grad_norm": 961475.75, "learning_rate": 1.135387724172771e-05, "loss": 3.1625, "step": 3060 }, { "epoch": 0.7754483455418035, "grad_norm": 811610.3125, "learning_rate": 1.1227582722909827e-05, "loss": 3.2445, "step": 3070 }, { "epoch": 0.7779742359181612, "grad_norm": 573133.25, "learning_rate": 1.1101288204091944e-05, "loss": 3.2385, "step": 3080 }, { "epoch": 0.7805001262945188, "grad_norm": 719160.0625, "learning_rate": 1.0974993685274059e-05, "loss": 3.4963, "step": 3090 }, { "epoch": 0.7830260166708765, "grad_norm": 625430.0625, "learning_rate": 1.0848699166456176e-05, "loss": 3.219, "step": 3100 }, { "epoch": 0.7855519070472341, "grad_norm": 847258.625, "learning_rate": 1.0722404647638293e-05, "loss": 3.3977, "step": 3110 }, { "epoch": 0.7880777974235919, "grad_norm": 909602.0625, "learning_rate": 1.0596110128820409e-05, "loss": 3.1271, "step": 3120 }, { "epoch": 0.7906036877999495, "grad_norm": 2303105.75, "learning_rate": 1.0469815610002526e-05, "loss": 3.1443, "step": 3130 }, { "epoch": 0.7931295781763071, "grad_norm": 627608.9375, "learning_rate": 1.0343521091184644e-05, "loss": 3.1124, "step": 3140 }, { "epoch": 0.7956554685526648, "grad_norm": 683938.75, "learning_rate": 1.021722657236676e-05, "loss": 3.5188, "step": 3150 }, { "epoch": 0.7981813589290225, "grad_norm": 742955.25, "learning_rate": 1.0090932053548877e-05, "loss": 3.3953, "step": 3160 }, { "epoch": 0.8007072493053802, "grad_norm": 871770.5, "learning_rate": 9.964637534730994e-06, "loss": 3.2334, "step": 3170 }, { "epoch": 0.8032331396817378, "grad_norm": 751897.5625, "learning_rate": 9.83834301591311e-06, "loss": 3.2724, "step": 3180 }, { "epoch": 0.8057590300580955, "grad_norm": 740156.9375, "learning_rate": 9.712048497095226e-06, "loss": 3.344, "step": 3190 }, { "epoch": 0.8082849204344531, "grad_norm": 634736.5, "learning_rate": 9.585753978277342e-06, "loss": 3.1028, "step": 3200 }, { "epoch": 0.8108108108108109, "grad_norm": 1092338.25, "learning_rate": 9.45945945945946e-06, "loss": 3.5916, "step": 3210 }, { "epoch": 0.8133367011871685, "grad_norm": 915638.875, "learning_rate": 9.333164940641578e-06, "loss": 3.126, "step": 3220 }, { "epoch": 0.8158625915635261, "grad_norm": 919722.3125, "learning_rate": 9.206870421823693e-06, "loss": 3.1032, "step": 3230 }, { "epoch": 0.8183884819398838, "grad_norm": 791748.625, "learning_rate": 9.08057590300581e-06, "loss": 3.4718, "step": 3240 }, { "epoch": 0.8209143723162414, "grad_norm": 607153.8125, "learning_rate": 8.954281384187927e-06, "loss": 2.989, "step": 3250 }, { "epoch": 0.8234402626925992, "grad_norm": 800489.625, "learning_rate": 8.827986865370043e-06, "loss": 3.1845, "step": 3260 }, { "epoch": 0.8259661530689568, "grad_norm": 748984.9375, "learning_rate": 8.70169234655216e-06, "loss": 3.2721, "step": 3270 }, { "epoch": 0.8284920434453145, "grad_norm": 686388.0625, "learning_rate": 8.575397827734277e-06, "loss": 3.178, "step": 3280 }, { "epoch": 0.8310179338216721, "grad_norm": 634646.5, "learning_rate": 8.449103308916394e-06, "loss": 3.1601, "step": 3290 }, { "epoch": 0.8335438241980299, "grad_norm": 973350.375, "learning_rate": 8.322808790098511e-06, "loss": 3.1181, "step": 3300 }, { "epoch": 0.8360697145743875, "grad_norm": 802978.1875, "learning_rate": 8.196514271280626e-06, "loss": 3.1694, "step": 3310 }, { "epoch": 0.8385956049507451, "grad_norm": 543352.25, "learning_rate": 8.070219752462743e-06, "loss": 3.0763, "step": 3320 }, { "epoch": 0.8411214953271028, "grad_norm": 965499.0625, "learning_rate": 7.94392523364486e-06, "loss": 3.3403, "step": 3330 }, { "epoch": 0.8436473857034604, "grad_norm": 1031962.0, "learning_rate": 7.817630714826977e-06, "loss": 3.1295, "step": 3340 }, { "epoch": 0.8461732760798182, "grad_norm": 900595.875, "learning_rate": 7.691336196009095e-06, "loss": 3.0841, "step": 3350 }, { "epoch": 0.8486991664561758, "grad_norm": 878680.5625, "learning_rate": 7.565041677191211e-06, "loss": 3.2706, "step": 3360 }, { "epoch": 0.8512250568325335, "grad_norm": 541207.75, "learning_rate": 7.438747158373327e-06, "loss": 3.2398, "step": 3370 }, { "epoch": 0.8537509472088911, "grad_norm": 689078.875, "learning_rate": 7.312452639555443e-06, "loss": 3.0991, "step": 3380 }, { "epoch": 0.8562768375852488, "grad_norm": 831574.1875, "learning_rate": 7.18615812073756e-06, "loss": 3.2631, "step": 3390 }, { "epoch": 0.8588027279616065, "grad_norm": 777584.0, "learning_rate": 7.0598636019196766e-06, "loss": 3.4074, "step": 3400 }, { "epoch": 0.8613286183379641, "grad_norm": 742726.625, "learning_rate": 6.9335690831017945e-06, "loss": 3.166, "step": 3410 }, { "epoch": 0.8638545087143218, "grad_norm": 580725.0625, "learning_rate": 6.807274564283911e-06, "loss": 3.1598, "step": 3420 }, { "epoch": 0.8663803990906794, "grad_norm": 588559.3125, "learning_rate": 6.680980045466027e-06, "loss": 3.0765, "step": 3430 }, { "epoch": 0.8689062894670372, "grad_norm": 512794.3125, "learning_rate": 6.554685526648144e-06, "loss": 3.3832, "step": 3440 }, { "epoch": 0.8714321798433948, "grad_norm": 876623.1875, "learning_rate": 6.42839100783026e-06, "loss": 3.4546, "step": 3450 }, { "epoch": 0.8739580702197525, "grad_norm": 716818.125, "learning_rate": 6.3020964890123765e-06, "loss": 3.2477, "step": 3460 }, { "epoch": 0.8764839605961101, "grad_norm": 1114704.75, "learning_rate": 6.175801970194494e-06, "loss": 3.1437, "step": 3470 }, { "epoch": 0.8790098509724678, "grad_norm": 639491.625, "learning_rate": 6.049507451376611e-06, "loss": 3.1617, "step": 3480 }, { "epoch": 0.8815357413488255, "grad_norm": 717235.0625, "learning_rate": 5.923212932558727e-06, "loss": 3.1024, "step": 3490 }, { "epoch": 0.8840616317251832, "grad_norm": 736368.0, "learning_rate": 5.796918413740844e-06, "loss": 3.3481, "step": 3500 }, { "epoch": 0.8865875221015408, "grad_norm": 527303.1875, "learning_rate": 5.670623894922961e-06, "loss": 3.0436, "step": 3510 }, { "epoch": 0.8891134124778984, "grad_norm": 1408582.375, "learning_rate": 5.544329376105077e-06, "loss": 3.3846, "step": 3520 }, { "epoch": 0.8916393028542561, "grad_norm": 557347.875, "learning_rate": 5.418034857287194e-06, "loss": 3.3017, "step": 3530 }, { "epoch": 0.8941651932306138, "grad_norm": 533418.25, "learning_rate": 5.2917403384693106e-06, "loss": 2.9057, "step": 3540 }, { "epoch": 0.8966910836069715, "grad_norm": 940764.0, "learning_rate": 5.165445819651428e-06, "loss": 2.9862, "step": 3550 }, { "epoch": 0.8992169739833291, "grad_norm": 753501.4375, "learning_rate": 5.039151300833544e-06, "loss": 3.0074, "step": 3560 }, { "epoch": 0.9017428643596868, "grad_norm": 641765.1875, "learning_rate": 4.912856782015661e-06, "loss": 3.2161, "step": 3570 }, { "epoch": 0.9042687547360445, "grad_norm": 846279.5625, "learning_rate": 4.786562263197777e-06, "loss": 3.1604, "step": 3580 }, { "epoch": 0.9067946451124022, "grad_norm": 623566.875, "learning_rate": 4.660267744379894e-06, "loss": 3.1331, "step": 3590 }, { "epoch": 0.9093205354887598, "grad_norm": 711430.625, "learning_rate": 4.533973225562011e-06, "loss": 3.2304, "step": 3600 }, { "epoch": 0.9118464258651174, "grad_norm": 656724.5, "learning_rate": 4.4076787067441276e-06, "loss": 3.4538, "step": 3610 }, { "epoch": 0.9143723162414751, "grad_norm": 595240.0, "learning_rate": 4.281384187926244e-06, "loss": 3.2896, "step": 3620 }, { "epoch": 0.9168982066178328, "grad_norm": 662666.375, "learning_rate": 4.155089669108361e-06, "loss": 3.0247, "step": 3630 }, { "epoch": 0.9194240969941905, "grad_norm": 671792.5625, "learning_rate": 4.028795150290478e-06, "loss": 3.2846, "step": 3640 }, { "epoch": 0.9219499873705481, "grad_norm": 946814.3125, "learning_rate": 3.902500631472594e-06, "loss": 3.2046, "step": 3650 }, { "epoch": 0.9244758777469058, "grad_norm": 795244.75, "learning_rate": 3.776206112654711e-06, "loss": 3.1295, "step": 3660 }, { "epoch": 0.9270017681232634, "grad_norm": 670945.125, "learning_rate": 3.649911593836828e-06, "loss": 3.1636, "step": 3670 }, { "epoch": 0.9295276584996212, "grad_norm": 752728.5625, "learning_rate": 3.5236170750189446e-06, "loss": 3.3796, "step": 3680 }, { "epoch": 0.9320535488759788, "grad_norm": 798886.1875, "learning_rate": 3.3973225562010608e-06, "loss": 3.101, "step": 3690 }, { "epoch": 0.9345794392523364, "grad_norm": 470130.46875, "learning_rate": 3.2710280373831774e-06, "loss": 3.3339, "step": 3700 }, { "epoch": 0.9371053296286941, "grad_norm": 437737.59375, "learning_rate": 3.1447335185652945e-06, "loss": 3.0232, "step": 3710 }, { "epoch": 0.9396312200050517, "grad_norm": 497686.5625, "learning_rate": 3.018438999747411e-06, "loss": 3.2697, "step": 3720 }, { "epoch": 0.9421571103814095, "grad_norm": 851300.5, "learning_rate": 2.892144480929528e-06, "loss": 3.4282, "step": 3730 }, { "epoch": 0.9446830007577671, "grad_norm": 715128.5, "learning_rate": 2.7658499621116445e-06, "loss": 3.1052, "step": 3740 }, { "epoch": 0.9472088911341248, "grad_norm": 904129.3125, "learning_rate": 2.639555443293761e-06, "loss": 3.1504, "step": 3750 }, { "epoch": 0.9497347815104824, "grad_norm": 897233.625, "learning_rate": 2.5132609244758778e-06, "loss": 3.2904, "step": 3760 }, { "epoch": 0.9522606718868402, "grad_norm": 722250.1875, "learning_rate": 2.3869664056579944e-06, "loss": 3.1392, "step": 3770 }, { "epoch": 0.9547865622631978, "grad_norm": 721083.9375, "learning_rate": 2.2606718868401115e-06, "loss": 3.137, "step": 3780 }, { "epoch": 0.9573124526395554, "grad_norm": 637189.25, "learning_rate": 2.1343773680222277e-06, "loss": 3.1711, "step": 3790 }, { "epoch": 0.9598383430159131, "grad_norm": 858492.0, "learning_rate": 2.008082849204345e-06, "loss": 3.2843, "step": 3800 }, { "epoch": 0.9623642333922707, "grad_norm": 714733.625, "learning_rate": 1.8817883303864613e-06, "loss": 3.305, "step": 3810 }, { "epoch": 0.9648901237686285, "grad_norm": 655039.75, "learning_rate": 1.7554938115685781e-06, "loss": 3.4754, "step": 3820 }, { "epoch": 0.9674160141449861, "grad_norm": 537881.875, "learning_rate": 1.6291992927506948e-06, "loss": 3.1539, "step": 3830 }, { "epoch": 0.9699419045213438, "grad_norm": 654013.3125, "learning_rate": 1.5029047739328114e-06, "loss": 3.3037, "step": 3840 }, { "epoch": 0.9724677948977014, "grad_norm": 654474.0, "learning_rate": 1.376610255114928e-06, "loss": 3.1471, "step": 3850 }, { "epoch": 0.974993685274059, "grad_norm": 925252.9375, "learning_rate": 1.2503157362970447e-06, "loss": 3.0344, "step": 3860 }, { "epoch": 0.9775195756504168, "grad_norm": 690294.6875, "learning_rate": 1.1240212174791616e-06, "loss": 3.0319, "step": 3870 }, { "epoch": 0.9800454660267744, "grad_norm": 654045.1875, "learning_rate": 9.977266986612783e-07, "loss": 3.2279, "step": 3880 }, { "epoch": 0.9825713564031321, "grad_norm": 593023.5, "learning_rate": 8.714321798433948e-07, "loss": 3.3599, "step": 3890 }, { "epoch": 0.9850972467794897, "grad_norm": 451262.59375, "learning_rate": 7.451376610255115e-07, "loss": 2.9683, "step": 3900 }, { "epoch": 0.9876231371558475, "grad_norm": 814919.1875, "learning_rate": 6.188431422076282e-07, "loss": 3.4731, "step": 3910 }, { "epoch": 0.9901490275322051, "grad_norm": 596075.0, "learning_rate": 4.925486233897449e-07, "loss": 3.0758, "step": 3920 }, { "epoch": 0.9926749179085628, "grad_norm": 436784.15625, "learning_rate": 3.6625410457186164e-07, "loss": 3.0476, "step": 3930 }, { "epoch": 0.9952008082849204, "grad_norm": 646142.9375, "learning_rate": 2.399595857539783e-07, "loss": 3.0169, "step": 3940 }, { "epoch": 0.997726698661278, "grad_norm": 558096.9375, "learning_rate": 1.1366506693609499e-07, "loss": 3.1317, "step": 3950 }, { "epoch": 1.0, "step": 3959, "total_flos": 2.3966994296890982e+17, "train_loss": 3.909515940200563, "train_runtime": 39767.0605, "train_samples_per_second": 0.398, "train_steps_per_second": 0.1 } ], "logging_steps": 10, "max_steps": 3959, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 3959, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 2.3966994296890982e+17, "train_batch_size": 4, "trial_name": null, "trial_params": null }