diff --git "a/sft/smoe_sigmoidgating/trainer_state.json" "b/sft/smoe_sigmoidgating/trainer_state.json" new file mode 100644--- /dev/null +++ "b/sft/smoe_sigmoidgating/trainer_state.json" @@ -0,0 +1,58254 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999398785546805, + "eval_steps": 500, + "global_step": 8316, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00012024289063909097, + "grad_norm": 42.9054718724962, + "learning_rate": 0.0, + "loss": 1.8842, + "step": 1 + }, + { + "epoch": 0.00024048578127818193, + "grad_norm": 56.01716786578133, + "learning_rate": 5.021476677069823e-07, + "loss": 1.9782, + "step": 2 + }, + { + "epoch": 0.0003607286719172729, + "grad_norm": 40.99744090955937, + "learning_rate": 7.958852231401551e-07, + "loss": 1.6601, + "step": 3 + }, + { + "epoch": 0.00048097156255636386, + "grad_norm": 36.11456718404119, + "learning_rate": 1.0042953354139647e-06, + "loss": 1.6417, + "step": 4 + }, + { + "epoch": 0.0006012144531954548, + "grad_norm": 53.44524569229124, + "learning_rate": 1.1659507774310057e-06, + "loss": 1.9841, + "step": 5 + }, + { + "epoch": 0.0007214573438345458, + "grad_norm": 45.45778701598468, + "learning_rate": 1.2980328908471373e-06, + "loss": 1.6823, + "step": 6 + }, + { + "epoch": 0.0008417002344736367, + "grad_norm": 4.569475667911875, + "learning_rate": 1.4097067265369432e-06, + "loss": 0.8152, + "step": 7 + }, + { + "epoch": 0.0009619431251127277, + "grad_norm": 32.33896779259698, + "learning_rate": 1.506443003120947e-06, + "loss": 1.531, + "step": 8 + }, + { + "epoch": 0.0010821860157518186, + "grad_norm": 14.578404463760236, + "learning_rate": 1.5917704462803102e-06, + "loss": 1.5061, + "step": 9 + }, + { + "epoch": 0.0012024289063909096, + "grad_norm": 11.296802061486527, + "learning_rate": 1.6680984451379884e-06, + "loss": 1.5111, + "step": 10 + }, + { + "epoch": 0.0013226717970300007, + "grad_norm": 11.353964334254165, + "learning_rate": 1.7371455188905097e-06, + "loss": 1.3273, + "step": 11 + }, + { + "epoch": 0.0014429146876690916, + "grad_norm": 8.39854287724588, + "learning_rate": 1.8001805585541196e-06, + "loss": 1.3087, + "step": 12 + }, + { + "epoch": 0.0015631575783081825, + "grad_norm": 5.0919298535809645, + "learning_rate": 1.8581671739548328e-06, + "loss": 1.3616, + "step": 13 + }, + { + "epoch": 0.0016834004689472734, + "grad_norm": 5.102695029763435, + "learning_rate": 1.9118543942439254e-06, + "loss": 1.2501, + "step": 14 + }, + { + "epoch": 0.0018036433595863645, + "grad_norm": 4.444236263061862, + "learning_rate": 1.961836000571161e-06, + "loss": 1.2257, + "step": 15 + }, + { + "epoch": 0.0019238862502254555, + "grad_norm": 3.856628799999913, + "learning_rate": 2.0085906708279293e-06, + "loss": 0.6485, + "step": 16 + }, + { + "epoch": 0.0020441291408645466, + "grad_norm": 3.5706274114914787, + "learning_rate": 2.0525099325728135e-06, + "loss": 1.2446, + "step": 17 + }, + { + "epoch": 0.0021643720315036373, + "grad_norm": 3.647900690943871, + "learning_rate": 2.0939181139872922e-06, + "loss": 0.7238, + "step": 18 + }, + { + "epoch": 0.0022846149221427284, + "grad_norm": 3.514211684232565, + "learning_rate": 2.1330868934640175e-06, + "loss": 1.1197, + "step": 19 + }, + { + "epoch": 0.002404857812781819, + "grad_norm": 3.616665038241768, + "learning_rate": 2.170246112844971e-06, + "loss": 0.7659, + "step": 20 + }, + { + "epoch": 0.0025251007034209102, + "grad_norm": 2.8978392608272454, + "learning_rate": 2.2055919496770983e-06, + "loss": 1.1266, + "step": 21 + }, + { + "epoch": 0.0026453435940600014, + "grad_norm": 4.655755548018, + "learning_rate": 2.2392931865974923e-06, + "loss": 1.0177, + "step": 22 + }, + { + "epoch": 0.002765586484699092, + "grad_norm": 4.8103012412546695, + "learning_rate": 2.271496085962064e-06, + "loss": 1.1461, + "step": 23 + }, + { + "epoch": 0.002885829375338183, + "grad_norm": 2.65883116131481, + "learning_rate": 2.3023282262611022e-06, + "loss": 1.1386, + "step": 24 + }, + { + "epoch": 0.003006072265977274, + "grad_norm": 2.385848815508872, + "learning_rate": 2.3319015548620114e-06, + "loss": 1.0834, + "step": 25 + }, + { + "epoch": 0.003126315156616365, + "grad_norm": 2.506290342239313, + "learning_rate": 2.3603148416618152e-06, + "loss": 0.9941, + "step": 26 + }, + { + "epoch": 0.003246558047255456, + "grad_norm": 2.068751945379443, + "learning_rate": 2.3876556694204647e-06, + "loss": 1.0867, + "step": 27 + }, + { + "epoch": 0.003366800937894547, + "grad_norm": 2.1316820439471775, + "learning_rate": 2.414002061950908e-06, + "loss": 1.0285, + "step": 28 + }, + { + "epoch": 0.003487043828533638, + "grad_norm": 2.20978151976203, + "learning_rate": 2.4394238264681557e-06, + "loss": 1.0964, + "step": 29 + }, + { + "epoch": 0.003607286719172729, + "grad_norm": 1.8821247158380945, + "learning_rate": 2.4639836682781433e-06, + "loss": 1.0753, + "step": 30 + }, + { + "epoch": 0.00372752960981182, + "grad_norm": 2.080291421905818, + "learning_rate": 2.487738122623307e-06, + "loss": 1.1211, + "step": 31 + }, + { + "epoch": 0.003847772500450911, + "grad_norm": 2.5253122646509643, + "learning_rate": 2.510738338534912e-06, + "loss": 1.0968, + "step": 32 + }, + { + "epoch": 0.003968015391090002, + "grad_norm": 2.1037677978560616, + "learning_rate": 2.5330307420306648e-06, + "loss": 1.1194, + "step": 33 + }, + { + "epoch": 0.004088258281729093, + "grad_norm": 1.803801744423376, + "learning_rate": 2.554657600279796e-06, + "loss": 0.9519, + "step": 34 + }, + { + "epoch": 0.004208501172368184, + "grad_norm": 1.813600009050609, + "learning_rate": 2.5756575039679493e-06, + "loss": 1.1068, + "step": 35 + }, + { + "epoch": 0.0043287440630072746, + "grad_norm": 1.776026042373559, + "learning_rate": 2.5960657816942747e-06, + "loss": 1.0005, + "step": 36 + }, + { + "epoch": 0.004448986953646365, + "grad_norm": 1.5328115947382408, + "learning_rate": 2.6159148575788668e-06, + "loss": 0.6133, + "step": 37 + }, + { + "epoch": 0.004569229844285457, + "grad_norm": 2.0354539064132737, + "learning_rate": 2.635234561171e-06, + "loss": 1.0885, + "step": 38 + }, + { + "epoch": 0.0046894727349245475, + "grad_norm": 2.066615651126547, + "learning_rate": 2.6540523970949877e-06, + "loss": 1.0363, + "step": 39 + }, + { + "epoch": 0.004809715625563638, + "grad_norm": 2.4133467616274937, + "learning_rate": 2.6723937805519533e-06, + "loss": 1.0535, + "step": 40 + }, + { + "epoch": 0.00492995851620273, + "grad_norm": 2.034451470644476, + "learning_rate": 2.690282243737839e-06, + "loss": 1.0643, + "step": 41 + }, + { + "epoch": 0.0050502014068418205, + "grad_norm": 2.036108925219593, + "learning_rate": 2.7077396173840807e-06, + "loss": 1.1169, + "step": 42 + }, + { + "epoch": 0.005170444297480911, + "grad_norm": 1.8209750561878049, + "learning_rate": 2.7247861909342594e-06, + "loss": 1.0429, + "step": 43 + }, + { + "epoch": 0.005290687188120003, + "grad_norm": 2.1484756743767597, + "learning_rate": 2.7414408543044743e-06, + "loss": 0.9383, + "step": 44 + }, + { + "epoch": 0.005410930078759093, + "grad_norm": 2.5284580798756053, + "learning_rate": 2.7577212237113157e-06, + "loss": 0.999, + "step": 45 + }, + { + "epoch": 0.005531172969398184, + "grad_norm": 1.7699356795871108, + "learning_rate": 2.7736437536690466e-06, + "loss": 1.1637, + "step": 46 + }, + { + "epoch": 0.005651415860037276, + "grad_norm": 1.7284985212783734, + "learning_rate": 2.789223836941131e-06, + "loss": 1.1723, + "step": 47 + }, + { + "epoch": 0.005771658750676366, + "grad_norm": 2.1074049050087136, + "learning_rate": 2.8044758939680847e-06, + "loss": 1.2127, + "step": 48 + }, + { + "epoch": 0.005891901641315457, + "grad_norm": 2.0028752878019014, + "learning_rate": 2.8194134530738863e-06, + "loss": 1.1116, + "step": 49 + }, + { + "epoch": 0.006012144531954548, + "grad_norm": 2.2516528522435695, + "learning_rate": 2.834049222568994e-06, + "loss": 1.0608, + "step": 50 + }, + { + "epoch": 0.006132387422593639, + "grad_norm": 1.6517769492958847, + "learning_rate": 2.848395155712969e-06, + "loss": 1.0113, + "step": 51 + }, + { + "epoch": 0.00625263031323273, + "grad_norm": 2.0943767691262973, + "learning_rate": 2.8624625093687977e-06, + "loss": 1.106, + "step": 52 + }, + { + "epoch": 0.006372873203871821, + "grad_norm": 1.8590404305660617, + "learning_rate": 2.876261897070029e-06, + "loss": 0.9734, + "step": 53 + }, + { + "epoch": 0.006493116094510912, + "grad_norm": 1.93010049331016, + "learning_rate": 2.889803337127447e-06, + "loss": 1.0949, + "step": 54 + }, + { + "epoch": 0.006613358985150003, + "grad_norm": 2.0311769912510327, + "learning_rate": 2.903096296321516e-06, + "loss": 1.0257, + "step": 55 + }, + { + "epoch": 0.006733601875789094, + "grad_norm": 1.8097742565862314, + "learning_rate": 2.9161497296578907e-06, + "loss": 1.0102, + "step": 56 + }, + { + "epoch": 0.006853844766428185, + "grad_norm": 2.0800133636852265, + "learning_rate": 2.928972116604173e-06, + "loss": 0.9766, + "step": 57 + }, + { + "epoch": 0.006974087657067276, + "grad_norm": 1.6952952722473953, + "learning_rate": 2.9415714941751377e-06, + "loss": 1.1239, + "step": 58 + }, + { + "epoch": 0.007094330547706367, + "grad_norm": 1.8939899103768032, + "learning_rate": 2.9539554871897396e-06, + "loss": 1.0465, + "step": 59 + }, + { + "epoch": 0.007214573438345458, + "grad_norm": 1.958422138013253, + "learning_rate": 2.9661313359851253e-06, + "loss": 1.062, + "step": 60 + }, + { + "epoch": 0.007334816328984549, + "grad_norm": 1.7582466776076209, + "learning_rate": 2.978105921839922e-06, + "loss": 1.0789, + "step": 61 + }, + { + "epoch": 0.00745505921962364, + "grad_norm": 1.873789905431085, + "learning_rate": 2.9898857903302893e-06, + "loss": 0.8582, + "step": 62 + }, + { + "epoch": 0.007575302110262731, + "grad_norm": 2.5945258761071455, + "learning_rate": 3.001477172817253e-06, + "loss": 1.0536, + "step": 63 + }, + { + "epoch": 0.007695545000901822, + "grad_norm": 2.1059034249169417, + "learning_rate": 3.012886006241894e-06, + "loss": 1.0821, + "step": 64 + }, + { + "epoch": 0.007815787891540913, + "grad_norm": 1.7097717215324966, + "learning_rate": 3.0241179513858383e-06, + "loss": 0.9837, + "step": 65 + }, + { + "epoch": 0.007936030782180003, + "grad_norm": 1.9633349470877801, + "learning_rate": 3.035178409737647e-06, + "loss": 1.0059, + "step": 66 + }, + { + "epoch": 0.008056273672819095, + "grad_norm": 1.913274049881658, + "learning_rate": 3.046072539090907e-06, + "loss": 1.0003, + "step": 67 + }, + { + "epoch": 0.008176516563458186, + "grad_norm": 2.0148328067146593, + "learning_rate": 3.056805267986779e-06, + "loss": 1.1684, + "step": 68 + }, + { + "epoch": 0.008296759454097276, + "grad_norm": 1.7243738586045927, + "learning_rate": 3.0673813091022194e-06, + "loss": 1.0336, + "step": 69 + }, + { + "epoch": 0.008417002344736368, + "grad_norm": 1.2650530900482255, + "learning_rate": 3.0778051716749317e-06, + "loss": 0.6242, + "step": 70 + }, + { + "epoch": 0.008537245235375458, + "grad_norm": 1.8263490315557696, + "learning_rate": 3.0880811730470094e-06, + "loss": 1.0048, + "step": 71 + }, + { + "epoch": 0.008657488126014549, + "grad_norm": 1.1283731224692688, + "learning_rate": 3.098213449401257e-06, + "loss": 0.5885, + "step": 72 + }, + { + "epoch": 0.00877773101665364, + "grad_norm": 1.8608873131532275, + "learning_rate": 3.1082059657570015e-06, + "loss": 1.1054, + "step": 73 + }, + { + "epoch": 0.00889797390729273, + "grad_norm": 1.7330239334655, + "learning_rate": 3.1180625252858496e-06, + "loss": 1.049, + "step": 74 + }, + { + "epoch": 0.009018216797931822, + "grad_norm": 2.3806687013476218, + "learning_rate": 3.1277867780021663e-06, + "loss": 1.0246, + "step": 75 + }, + { + "epoch": 0.009138459688570914, + "grad_norm": 1.7187223952417439, + "learning_rate": 3.1373822288779824e-06, + "loss": 1.0051, + "step": 76 + }, + { + "epoch": 0.009258702579210003, + "grad_norm": 1.6599135995693506, + "learning_rate": 3.1468522454274533e-06, + "loss": 0.9046, + "step": 77 + }, + { + "epoch": 0.009378945469849095, + "grad_norm": 1.7240265739587846, + "learning_rate": 3.15620006480197e-06, + "loss": 1.0074, + "step": 78 + }, + { + "epoch": 0.009499188360488187, + "grad_norm": 2.5035391551849533, + "learning_rate": 3.1654288004333087e-06, + "loss": 1.0059, + "step": 79 + }, + { + "epoch": 0.009619431251127276, + "grad_norm": 2.22300827376267, + "learning_rate": 3.1745414482589353e-06, + "loss": 0.8833, + "step": 80 + }, + { + "epoch": 0.009739674141766368, + "grad_norm": 1.9796715209605102, + "learning_rate": 3.1835408925606204e-06, + "loss": 1.0088, + "step": 81 + }, + { + "epoch": 0.00985991703240546, + "grad_norm": 2.1569794236547146, + "learning_rate": 3.1924299114448214e-06, + "loss": 1.0125, + "step": 82 + }, + { + "epoch": 0.00998015992304455, + "grad_norm": 1.6921851498912437, + "learning_rate": 3.2012111819909055e-06, + "loss": 0.9264, + "step": 83 + }, + { + "epoch": 0.010100402813683641, + "grad_norm": 1.8791578604188046, + "learning_rate": 3.2098872850910627e-06, + "loss": 1.0543, + "step": 84 + }, + { + "epoch": 0.010220645704322733, + "grad_norm": 1.6818223028812016, + "learning_rate": 3.2184607100038194e-06, + "loss": 0.967, + "step": 85 + }, + { + "epoch": 0.010340888594961822, + "grad_norm": 1.7142580461539945, + "learning_rate": 3.2269338586412414e-06, + "loss": 1.0434, + "step": 86 + }, + { + "epoch": 0.010461131485600914, + "grad_norm": 1.954563801096647, + "learning_rate": 3.2353090496083106e-06, + "loss": 1.0715, + "step": 87 + }, + { + "epoch": 0.010581374376240005, + "grad_norm": 1.5811513455340482, + "learning_rate": 3.2435885220114572e-06, + "loss": 0.8852, + "step": 88 + }, + { + "epoch": 0.010701617266879095, + "grad_norm": 1.8876004577581194, + "learning_rate": 3.2517744390519113e-06, + "loss": 1.0133, + "step": 89 + }, + { + "epoch": 0.010821860157518187, + "grad_norm": 1.7752047568410065, + "learning_rate": 3.259868891418298e-06, + "loss": 0.8808, + "step": 90 + }, + { + "epoch": 0.010942103048157278, + "grad_norm": 1.7622319254354826, + "learning_rate": 3.2678739004917757e-06, + "loss": 0.9708, + "step": 91 + }, + { + "epoch": 0.011062345938796368, + "grad_norm": 1.5680599473130195, + "learning_rate": 3.275791421376029e-06, + "loss": 0.9848, + "step": 92 + }, + { + "epoch": 0.01118258882943546, + "grad_norm": 1.7714651289696666, + "learning_rate": 3.2836233457634622e-06, + "loss": 1.0332, + "step": 93 + }, + { + "epoch": 0.011302831720074551, + "grad_norm": 1.7044995561835823, + "learning_rate": 3.2913715046481135e-06, + "loss": 0.9545, + "step": 94 + }, + { + "epoch": 0.011423074610713641, + "grad_norm": 1.8203509412643988, + "learning_rate": 3.299037670895023e-06, + "loss": 0.9729, + "step": 95 + }, + { + "epoch": 0.011543317501352733, + "grad_norm": 1.6871351624743998, + "learning_rate": 3.3066235616750667e-06, + "loss": 0.939, + "step": 96 + }, + { + "epoch": 0.011663560391991824, + "grad_norm": 1.945024338793043, + "learning_rate": 3.3141308407736276e-06, + "loss": 1.0312, + "step": 97 + }, + { + "epoch": 0.011783803282630914, + "grad_norm": 1.6536875641931401, + "learning_rate": 3.321561120780869e-06, + "loss": 0.9596, + "step": 98 + }, + { + "epoch": 0.011904046173270006, + "grad_norm": 1.8295162156588731, + "learning_rate": 3.3289159651708192e-06, + "loss": 1.1218, + "step": 99 + }, + { + "epoch": 0.012024289063909096, + "grad_norm": 1.8085799741941273, + "learning_rate": 3.3361968902759768e-06, + "loss": 1.0618, + "step": 100 + }, + { + "epoch": 0.012144531954548187, + "grad_norm": 2.1584189742881, + "learning_rate": 3.343405367163663e-06, + "loss": 1.0439, + "step": 101 + }, + { + "epoch": 0.012264774845187279, + "grad_norm": 1.9154915593577924, + "learning_rate": 3.350542823419951e-06, + "loss": 0.9273, + "step": 102 + }, + { + "epoch": 0.012385017735826368, + "grad_norm": 3.002351475074631, + "learning_rate": 3.3576106448465615e-06, + "loss": 1.1064, + "step": 103 + }, + { + "epoch": 0.01250526062646546, + "grad_norm": 1.7094613592171706, + "learning_rate": 3.3646101770757797e-06, + "loss": 0.9803, + "step": 104 + }, + { + "epoch": 0.012625503517104552, + "grad_norm": 1.4358162348270307, + "learning_rate": 3.371542727108104e-06, + "loss": 0.9471, + "step": 105 + }, + { + "epoch": 0.012745746407743641, + "grad_norm": 1.89941801644431, + "learning_rate": 3.3784095647770114e-06, + "loss": 1.0734, + "step": 106 + }, + { + "epoch": 0.012865989298382733, + "grad_norm": 1.6835688259466017, + "learning_rate": 3.3852119241449547e-06, + "loss": 1.04, + "step": 107 + }, + { + "epoch": 0.012986232189021825, + "grad_norm": 1.7975891684325944, + "learning_rate": 3.3919510048344295e-06, + "loss": 1.0528, + "step": 108 + }, + { + "epoch": 0.013106475079660914, + "grad_norm": 1.8627317886005, + "learning_rate": 3.3986279732976907e-06, + "loss": 0.973, + "step": 109 + }, + { + "epoch": 0.013226717970300006, + "grad_norm": 1.8766932788924675, + "learning_rate": 3.4052439640284983e-06, + "loss": 1.031, + "step": 110 + }, + { + "epoch": 0.013346960860939098, + "grad_norm": 1.6729077865852626, + "learning_rate": 3.4118000807190217e-06, + "loss": 0.9279, + "step": 111 + }, + { + "epoch": 0.013467203751578187, + "grad_norm": 1.671494057295879, + "learning_rate": 3.4182973973648723e-06, + "loss": 0.8596, + "step": 112 + }, + { + "epoch": 0.013587446642217279, + "grad_norm": 2.2398077800564855, + "learning_rate": 3.424736959321014e-06, + "loss": 1.1074, + "step": 113 + }, + { + "epoch": 0.01370768953285637, + "grad_norm": 1.7495487176217848, + "learning_rate": 3.431119784311155e-06, + "loss": 1.0071, + "step": 114 + }, + { + "epoch": 0.01382793242349546, + "grad_norm": 1.4511018019915218, + "learning_rate": 3.43744686339307e-06, + "loss": 0.8319, + "step": 115 + }, + { + "epoch": 0.013948175314134552, + "grad_norm": 1.7668559416858678, + "learning_rate": 3.44371916188212e-06, + "loss": 0.9993, + "step": 116 + }, + { + "epoch": 0.014068418204773643, + "grad_norm": 1.855228834201376, + "learning_rate": 3.449937620235143e-06, + "loss": 0.9942, + "step": 117 + }, + { + "epoch": 0.014188661095412733, + "grad_norm": 1.8719644133494637, + "learning_rate": 3.456103154896722e-06, + "loss": 0.954, + "step": 118 + }, + { + "epoch": 0.014308903986051825, + "grad_norm": 1.7537807768053684, + "learning_rate": 3.462216659109757e-06, + "loss": 1.0254, + "step": 119 + }, + { + "epoch": 0.014429146876690916, + "grad_norm": 2.113366201107061, + "learning_rate": 3.4682790036921077e-06, + "loss": 0.9923, + "step": 120 + }, + { + "epoch": 0.014549389767330006, + "grad_norm": 1.9938940779427516, + "learning_rate": 3.4742910377810193e-06, + "loss": 0.8896, + "step": 121 + }, + { + "epoch": 0.014669632657969098, + "grad_norm": 1.7979440577230255, + "learning_rate": 3.4802535895469042e-06, + "loss": 0.989, + "step": 122 + }, + { + "epoch": 0.01478987554860819, + "grad_norm": 1.8568796034132815, + "learning_rate": 3.4861674668779934e-06, + "loss": 0.9803, + "step": 123 + }, + { + "epoch": 0.01491011843924728, + "grad_norm": 1.6121854223898633, + "learning_rate": 3.492033458037272e-06, + "loss": 0.9236, + "step": 124 + }, + { + "epoch": 0.01503036132988637, + "grad_norm": 2.0168150707948302, + "learning_rate": 3.497852332293018e-06, + "loss": 1.019, + "step": 125 + }, + { + "epoch": 0.015150604220525462, + "grad_norm": 1.6971646225071664, + "learning_rate": 3.5036248405242356e-06, + "loss": 1.0329, + "step": 126 + }, + { + "epoch": 0.015270847111164552, + "grad_norm": 1.7166710504766, + "learning_rate": 3.509351715802146e-06, + "loss": 0.9456, + "step": 127 + }, + { + "epoch": 0.015391090001803644, + "grad_norm": 1.8545781296938648, + "learning_rate": 3.5150336739488763e-06, + "loss": 0.9296, + "step": 128 + }, + { + "epoch": 0.015511332892442733, + "grad_norm": 1.5635883355678308, + "learning_rate": 3.5206714140744143e-06, + "loss": 0.9166, + "step": 129 + }, + { + "epoch": 0.015631575783081827, + "grad_norm": 2.039798296886234, + "learning_rate": 3.5262656190928208e-06, + "loss": 1.0507, + "step": 130 + }, + { + "epoch": 0.015751818673720917, + "grad_norm": 1.0511630474330889, + "learning_rate": 3.5318169562186737e-06, + "loss": 0.7173, + "step": 131 + }, + { + "epoch": 0.015872061564360006, + "grad_norm": 2.075285498959399, + "learning_rate": 3.5373260774446292e-06, + "loss": 0.9097, + "step": 132 + }, + { + "epoch": 0.0159923044549991, + "grad_norm": 1.5016991210772352, + "learning_rate": 3.542793620000961e-06, + "loss": 0.9755, + "step": 133 + }, + { + "epoch": 0.01611254734563819, + "grad_norm": 1.8951122533519131, + "learning_rate": 3.5482202067978894e-06, + "loss": 0.979, + "step": 134 + }, + { + "epoch": 0.01623279023627728, + "grad_norm": 1.8305983806431652, + "learning_rate": 3.553606446851471e-06, + "loss": 0.8983, + "step": 135 + }, + { + "epoch": 0.016353033126916373, + "grad_norm": 1.678779463065641, + "learning_rate": 3.5589529356937613e-06, + "loss": 0.9503, + "step": 136 + }, + { + "epoch": 0.016473276017555463, + "grad_norm": 1.542233526128579, + "learning_rate": 3.5642602557679627e-06, + "loss": 0.868, + "step": 137 + }, + { + "epoch": 0.016593518908194552, + "grad_norm": 2.126668947850422, + "learning_rate": 3.569528976809202e-06, + "loss": 0.9323, + "step": 138 + }, + { + "epoch": 0.016713761798833646, + "grad_norm": 1.6496141040254557, + "learning_rate": 3.5747596562115522e-06, + "loss": 0.9975, + "step": 139 + }, + { + "epoch": 0.016834004689472735, + "grad_norm": 1.912072778493789, + "learning_rate": 3.5799528393819138e-06, + "loss": 1.0592, + "step": 140 + }, + { + "epoch": 0.016954247580111825, + "grad_norm": 1.686518489651471, + "learning_rate": 3.585109060081286e-06, + "loss": 0.9608, + "step": 141 + }, + { + "epoch": 0.017074490470750915, + "grad_norm": 1.7011533050749876, + "learning_rate": 3.590228840753992e-06, + "loss": 0.8822, + "step": 142 + }, + { + "epoch": 0.01719473336139001, + "grad_norm": 1.770390518772268, + "learning_rate": 3.5953126928453423e-06, + "loss": 0.9829, + "step": 143 + }, + { + "epoch": 0.017314976252029098, + "grad_norm": 1.8024633003339912, + "learning_rate": 3.600361117108239e-06, + "loss": 0.9097, + "step": 144 + }, + { + "epoch": 0.017435219142668188, + "grad_norm": 1.694880557957819, + "learning_rate": 3.6053746038991616e-06, + "loss": 1.0737, + "step": 145 + }, + { + "epoch": 0.01755546203330728, + "grad_norm": 1.0982551434190448, + "learning_rate": 3.6103536334639843e-06, + "loss": 0.5885, + "step": 146 + }, + { + "epoch": 0.01767570492394637, + "grad_norm": 1.8265641551014342, + "learning_rate": 3.615298676214041e-06, + "loss": 0.9304, + "step": 147 + }, + { + "epoch": 0.01779594781458546, + "grad_norm": 1.7865940038770227, + "learning_rate": 3.6202101929928317e-06, + "loss": 0.9897, + "step": 148 + }, + { + "epoch": 0.017916190705224554, + "grad_norm": 1.6605827232552812, + "learning_rate": 3.6250886353337413e-06, + "loss": 0.9753, + "step": 149 + }, + { + "epoch": 0.018036433595863644, + "grad_norm": 1.9653523509236837, + "learning_rate": 3.6299344457091488e-06, + "loss": 0.9623, + "step": 150 + }, + { + "epoch": 0.018156676486502734, + "grad_norm": 2.0414725490334256, + "learning_rate": 3.634748057771256e-06, + "loss": 1.0743, + "step": 151 + }, + { + "epoch": 0.018276919377141827, + "grad_norm": 1.4883407620924165, + "learning_rate": 3.639529896584965e-06, + "loss": 0.9481, + "step": 152 + }, + { + "epoch": 0.018397162267780917, + "grad_norm": 2.2163467286285976, + "learning_rate": 3.6442803788531233e-06, + "loss": 1.0596, + "step": 153 + }, + { + "epoch": 0.018517405158420007, + "grad_norm": 1.666847441724027, + "learning_rate": 3.6489999131344357e-06, + "loss": 1.0649, + "step": 154 + }, + { + "epoch": 0.0186376480490591, + "grad_norm": 1.635068439252819, + "learning_rate": 3.653688900054313e-06, + "loss": 0.9871, + "step": 155 + }, + { + "epoch": 0.01875789093969819, + "grad_norm": 1.6799155856677688, + "learning_rate": 3.6583477325089526e-06, + "loss": 0.9116, + "step": 156 + }, + { + "epoch": 0.01887813383033728, + "grad_norm": 1.9430395735412942, + "learning_rate": 3.6629767958628916e-06, + "loss": 1.1773, + "step": 157 + }, + { + "epoch": 0.018998376720976373, + "grad_norm": 1.7760400706831247, + "learning_rate": 3.667576468140291e-06, + "loss": 0.9871, + "step": 158 + }, + { + "epoch": 0.019118619611615463, + "grad_norm": 1.929773738030621, + "learning_rate": 3.672147120210184e-06, + "loss": 1.0025, + "step": 159 + }, + { + "epoch": 0.019238862502254553, + "grad_norm": 1.6970523748548505, + "learning_rate": 3.6766891159659177e-06, + "loss": 0.9542, + "step": 160 + }, + { + "epoch": 0.019359105392893646, + "grad_norm": 2.0309769667010062, + "learning_rate": 3.6812028124990075e-06, + "loss": 1.0382, + "step": 161 + }, + { + "epoch": 0.019479348283532736, + "grad_norm": 2.2662412660307365, + "learning_rate": 3.6856885602676016e-06, + "loss": 0.9561, + "step": 162 + }, + { + "epoch": 0.019599591174171826, + "grad_norm": 1.903059619099958, + "learning_rate": 3.6901467032597733e-06, + "loss": 1.063, + "step": 163 + }, + { + "epoch": 0.01971983406481092, + "grad_norm": 1.9938489979049399, + "learning_rate": 3.694577579151804e-06, + "loss": 0.9939, + "step": 164 + }, + { + "epoch": 0.01984007695545001, + "grad_norm": 2.0301508350427815, + "learning_rate": 3.6989815194616703e-06, + "loss": 0.8515, + "step": 165 + }, + { + "epoch": 0.0199603198460891, + "grad_norm": 2.3702013828331814, + "learning_rate": 3.703358849697888e-06, + "loss": 0.9257, + "step": 166 + }, + { + "epoch": 0.020080562736728192, + "grad_norm": 1.588006159671973, + "learning_rate": 3.7077098895038803e-06, + "loss": 0.9069, + "step": 167 + }, + { + "epoch": 0.020200805627367282, + "grad_norm": 1.8662114485216463, + "learning_rate": 3.712034952798045e-06, + "loss": 1.0805, + "step": 168 + }, + { + "epoch": 0.02032104851800637, + "grad_norm": 1.8860217084312874, + "learning_rate": 3.7163343479096656e-06, + "loss": 1.0089, + "step": 169 + }, + { + "epoch": 0.020441291408645465, + "grad_norm": 1.7522766426968355, + "learning_rate": 3.720608377710802e-06, + "loss": 0.9346, + "step": 170 + }, + { + "epoch": 0.020561534299284555, + "grad_norm": 1.8840481197828973, + "learning_rate": 3.7248573397443277e-06, + "loss": 1.005, + "step": 171 + }, + { + "epoch": 0.020681777189923645, + "grad_norm": 2.001747559949108, + "learning_rate": 3.729081526348224e-06, + "loss": 1.0693, + "step": 172 + }, + { + "epoch": 0.020802020080562738, + "grad_norm": 1.629635508102032, + "learning_rate": 3.7332812247762777e-06, + "loss": 0.9336, + "step": 173 + }, + { + "epoch": 0.020922262971201828, + "grad_norm": 1.8232722634939007, + "learning_rate": 3.737456717315293e-06, + "loss": 1.0936, + "step": 174 + }, + { + "epoch": 0.021042505861840918, + "grad_norm": 1.6623711385477706, + "learning_rate": 3.7416082813989552e-06, + "loss": 0.9828, + "step": 175 + }, + { + "epoch": 0.02116274875248001, + "grad_norm": 1.6634109009345528, + "learning_rate": 3.745736189718439e-06, + "loss": 0.9912, + "step": 176 + }, + { + "epoch": 0.0212829916431191, + "grad_norm": 1.8327738023658096, + "learning_rate": 3.749840710329894e-06, + "loss": 0.849, + "step": 177 + }, + { + "epoch": 0.02140323453375819, + "grad_norm": 2.383201892491691, + "learning_rate": 3.7539221067588938e-06, + "loss": 1.1698, + "step": 178 + }, + { + "epoch": 0.021523477424397284, + "grad_norm": 2.718398411258505, + "learning_rate": 3.757980638101964e-06, + "loss": 1.1898, + "step": 179 + }, + { + "epoch": 0.021643720315036374, + "grad_norm": 1.9019056751837464, + "learning_rate": 3.7620165591252806e-06, + "loss": 1.0365, + "step": 180 + }, + { + "epoch": 0.021763963205675464, + "grad_norm": 1.7084616248487798, + "learning_rate": 3.766030120360636e-06, + "loss": 1.0091, + "step": 181 + }, + { + "epoch": 0.021884206096314557, + "grad_norm": 4.5357712912651404, + "learning_rate": 3.7700215681987578e-06, + "loss": 1.004, + "step": 182 + }, + { + "epoch": 0.022004448986953647, + "grad_norm": 1.464771970897619, + "learning_rate": 3.7739911449800767e-06, + "loss": 0.9392, + "step": 183 + }, + { + "epoch": 0.022124691877592736, + "grad_norm": 1.6389962967556726, + "learning_rate": 3.7779390890830114e-06, + "loss": 0.8808, + "step": 184 + }, + { + "epoch": 0.02224493476823183, + "grad_norm": 1.7653281713469793, + "learning_rate": 3.7818656350098723e-06, + "loss": 0.9675, + "step": 185 + }, + { + "epoch": 0.02236517765887092, + "grad_norm": 2.142563213898091, + "learning_rate": 3.7857710134704447e-06, + "loss": 0.9573, + "step": 186 + }, + { + "epoch": 0.02248542054951001, + "grad_norm": 1.862554252397602, + "learning_rate": 3.7896554514633234e-06, + "loss": 0.93, + "step": 187 + }, + { + "epoch": 0.022605663440149103, + "grad_norm": 1.8387357124899006, + "learning_rate": 3.7935191723550955e-06, + "loss": 0.9516, + "step": 188 + }, + { + "epoch": 0.022725906330788193, + "grad_norm": 1.6538261815845865, + "learning_rate": 3.797362395957408e-06, + "loss": 0.9694, + "step": 189 + }, + { + "epoch": 0.022846149221427282, + "grad_norm": 2.239537390895657, + "learning_rate": 3.8011853386020055e-06, + "loss": 0.9908, + "step": 190 + }, + { + "epoch": 0.022966392112066376, + "grad_norm": 2.0906872339493954, + "learning_rate": 3.804988213213804e-06, + "loss": 1.0419, + "step": 191 + }, + { + "epoch": 0.023086635002705466, + "grad_norm": 1.0321295484211859, + "learning_rate": 3.808771229382049e-06, + "loss": 0.6334, + "step": 192 + }, + { + "epoch": 0.023206877893344555, + "grad_norm": 1.806655999195564, + "learning_rate": 3.8125345934296324e-06, + "loss": 0.9288, + "step": 193 + }, + { + "epoch": 0.02332712078398365, + "grad_norm": 1.9110420529076715, + "learning_rate": 3.81627850848061e-06, + "loss": 0.9964, + "step": 194 + }, + { + "epoch": 0.02344736367462274, + "grad_norm": 2.000862412480368, + "learning_rate": 3.820003174525994e-06, + "loss": 1.0268, + "step": 195 + }, + { + "epoch": 0.02356760656526183, + "grad_norm": 2.011950433222909, + "learning_rate": 3.823708788487851e-06, + "loss": 0.9302, + "step": 196 + }, + { + "epoch": 0.02368784945590092, + "grad_norm": 1.6859431679447414, + "learning_rate": 3.827395544281781e-06, + "loss": 0.9408, + "step": 197 + }, + { + "epoch": 0.02380809234654001, + "grad_norm": 2.0968629839534394, + "learning_rate": 3.831063632877802e-06, + "loss": 0.9021, + "step": 198 + }, + { + "epoch": 0.0239283352371791, + "grad_norm": 2.112830940823987, + "learning_rate": 3.834713242359712e-06, + "loss": 0.8745, + "step": 199 + }, + { + "epoch": 0.02404857812781819, + "grad_norm": 1.6118482085623567, + "learning_rate": 3.838344557982959e-06, + "loss": 0.9797, + "step": 200 + }, + { + "epoch": 0.024168821018457284, + "grad_norm": 2.418213667043824, + "learning_rate": 3.841957762231063e-06, + "loss": 1.0588, + "step": 201 + }, + { + "epoch": 0.024289063909096374, + "grad_norm": 1.8148665927167018, + "learning_rate": 3.8455530348706454e-06, + "loss": 0.9809, + "step": 202 + }, + { + "epoch": 0.024409306799735464, + "grad_norm": 1.6966976009813963, + "learning_rate": 3.849130553005099e-06, + "loss": 0.8785, + "step": 203 + }, + { + "epoch": 0.024529549690374557, + "grad_norm": 1.7268541875048264, + "learning_rate": 3.852690491126933e-06, + "loss": 0.9681, + "step": 204 + }, + { + "epoch": 0.024649792581013647, + "grad_norm": 1.9311248738762126, + "learning_rate": 3.856233021168845e-06, + "loss": 1.0305, + "step": 205 + }, + { + "epoch": 0.024770035471652737, + "grad_norm": 1.7656166279011962, + "learning_rate": 3.859758312553544e-06, + "loss": 0.9874, + "step": 206 + }, + { + "epoch": 0.02489027836229183, + "grad_norm": 1.8695485019804445, + "learning_rate": 3.8632665322423735e-06, + "loss": 0.9985, + "step": 207 + }, + { + "epoch": 0.02501052125293092, + "grad_norm": 1.6722986272858151, + "learning_rate": 3.866757844782762e-06, + "loss": 0.9204, + "step": 208 + }, + { + "epoch": 0.02513076414357001, + "grad_norm": 2.0179376863399248, + "learning_rate": 3.870232412354527e-06, + "loss": 1.0222, + "step": 209 + }, + { + "epoch": 0.025251007034209103, + "grad_norm": 1.6363864461195288, + "learning_rate": 3.873690394815086e-06, + "loss": 1.0001, + "step": 210 + }, + { + "epoch": 0.025371249924848193, + "grad_norm": 1.9521428805143832, + "learning_rate": 3.877131949743587e-06, + "loss": 1.0206, + "step": 211 + }, + { + "epoch": 0.025491492815487283, + "grad_norm": 1.7325801325510024, + "learning_rate": 3.880557232483993e-06, + "loss": 0.8909, + "step": 212 + }, + { + "epoch": 0.025611735706126376, + "grad_norm": 1.6210832342702983, + "learning_rate": 3.883966396187164e-06, + "loss": 0.956, + "step": 213 + }, + { + "epoch": 0.025731978596765466, + "grad_norm": 1.9933828320731213, + "learning_rate": 3.887359591851937e-06, + "loss": 1.0142, + "step": 214 + }, + { + "epoch": 0.025852221487404556, + "grad_norm": 1.4791943543041177, + "learning_rate": 3.890736968365265e-06, + "loss": 0.9887, + "step": 215 + }, + { + "epoch": 0.02597246437804365, + "grad_norm": 1.611881933119715, + "learning_rate": 3.894098672541412e-06, + "loss": 0.9411, + "step": 216 + }, + { + "epoch": 0.02609270726868274, + "grad_norm": 1.640830762766537, + "learning_rate": 3.89744484916025e-06, + "loss": 0.8531, + "step": 217 + }, + { + "epoch": 0.02621295015932183, + "grad_norm": 1.7542777230935904, + "learning_rate": 3.900775641004673e-06, + "loss": 1.0033, + "step": 218 + }, + { + "epoch": 0.026333193049960922, + "grad_norm": 2.294944730604273, + "learning_rate": 3.904091188897156e-06, + "loss": 0.9857, + "step": 219 + }, + { + "epoch": 0.026453435940600012, + "grad_norm": 2.1422171136435506, + "learning_rate": 3.90739163173548e-06, + "loss": 0.9456, + "step": 220 + }, + { + "epoch": 0.026573678831239102, + "grad_norm": 2.196690788522508, + "learning_rate": 3.910677106527646e-06, + "loss": 1.0074, + "step": 221 + }, + { + "epoch": 0.026693921721878195, + "grad_norm": 2.3691997976689123, + "learning_rate": 3.913947748426004e-06, + "loss": 0.9499, + "step": 222 + }, + { + "epoch": 0.026814164612517285, + "grad_norm": 2.721346262261746, + "learning_rate": 3.9172036907606136e-06, + "loss": 0.9274, + "step": 223 + }, + { + "epoch": 0.026934407503156375, + "grad_norm": 1.638061045854048, + "learning_rate": 3.920445065071855e-06, + "loss": 1.026, + "step": 224 + }, + { + "epoch": 0.027054650393795468, + "grad_norm": 2.068274304420871, + "learning_rate": 3.923672001142322e-06, + "loss": 0.9141, + "step": 225 + }, + { + "epoch": 0.027174893284434558, + "grad_norm": 1.58955976869541, + "learning_rate": 3.926884627027996e-06, + "loss": 0.9255, + "step": 226 + }, + { + "epoch": 0.027295136175073648, + "grad_norm": 1.7792921851354904, + "learning_rate": 3.930083069088744e-06, + "loss": 0.8716, + "step": 227 + }, + { + "epoch": 0.02741537906571274, + "grad_norm": 1.0085995580149931, + "learning_rate": 3.933267452018137e-06, + "loss": 0.5965, + "step": 228 + }, + { + "epoch": 0.02753562195635183, + "grad_norm": 1.7576179227160855, + "learning_rate": 3.936437898872622e-06, + "loss": 0.904, + "step": 229 + }, + { + "epoch": 0.02765586484699092, + "grad_norm": 1.776020605534795, + "learning_rate": 3.9395945311000525e-06, + "loss": 0.9026, + "step": 230 + }, + { + "epoch": 0.027776107737630014, + "grad_norm": 2.2969030261762717, + "learning_rate": 3.942737468567608e-06, + "loss": 1.0288, + "step": 231 + }, + { + "epoch": 0.027896350628269104, + "grad_norm": 1.838009124081518, + "learning_rate": 3.9458668295891026e-06, + "loss": 0.9578, + "step": 232 + }, + { + "epoch": 0.028016593518908194, + "grad_norm": 2.06334426246341, + "learning_rate": 3.948982730951712e-06, + "loss": 0.9787, + "step": 233 + }, + { + "epoch": 0.028136836409547287, + "grad_norm": 1.9780369864073162, + "learning_rate": 3.9520852879421254e-06, + "loss": 0.9178, + "step": 234 + }, + { + "epoch": 0.028257079300186377, + "grad_norm": 1.888616077817853, + "learning_rate": 3.955174614372137e-06, + "loss": 0.9117, + "step": 235 + }, + { + "epoch": 0.028377322190825467, + "grad_norm": 1.9333968245286357, + "learning_rate": 3.9582508226037045e-06, + "loss": 0.9351, + "step": 236 + }, + { + "epoch": 0.02849756508146456, + "grad_norm": 2.2014765422818026, + "learning_rate": 3.9613140235734636e-06, + "loss": 1.0706, + "step": 237 + }, + { + "epoch": 0.02861780797210365, + "grad_norm": 2.0121544819750485, + "learning_rate": 3.96436432681674e-06, + "loss": 0.9411, + "step": 238 + }, + { + "epoch": 0.02873805086274274, + "grad_norm": 1.913003309255989, + "learning_rate": 3.967401840491044e-06, + "loss": 0.997, + "step": 239 + }, + { + "epoch": 0.028858293753381833, + "grad_norm": 2.1344533344599745, + "learning_rate": 3.97042667139909e-06, + "loss": 0.972, + "step": 240 + }, + { + "epoch": 0.028978536644020923, + "grad_norm": 2.1192908167902753, + "learning_rate": 3.973438925011327e-06, + "loss": 0.9716, + "step": 241 + }, + { + "epoch": 0.029098779534660012, + "grad_norm": 2.100909844346555, + "learning_rate": 3.976438705488002e-06, + "loss": 0.9894, + "step": 242 + }, + { + "epoch": 0.029219022425299106, + "grad_norm": 2.4542991598101334, + "learning_rate": 3.9794261157007744e-06, + "loss": 1.0796, + "step": 243 + }, + { + "epoch": 0.029339265315938196, + "grad_norm": 2.1317605578891534, + "learning_rate": 3.982401257253887e-06, + "loss": 1.0394, + "step": 244 + }, + { + "epoch": 0.029459508206577285, + "grad_norm": 1.8452815790517036, + "learning_rate": 3.985364230504893e-06, + "loss": 0.9426, + "step": 245 + }, + { + "epoch": 0.02957975109721638, + "grad_norm": 2.0296872509727883, + "learning_rate": 3.988315134584976e-06, + "loss": 0.9661, + "step": 246 + }, + { + "epoch": 0.02969999398785547, + "grad_norm": 1.772112649794248, + "learning_rate": 3.991254067418851e-06, + "loss": 0.8747, + "step": 247 + }, + { + "epoch": 0.02982023687849456, + "grad_norm": 1.825359763650484, + "learning_rate": 3.994181125744254e-06, + "loss": 0.8892, + "step": 248 + }, + { + "epoch": 0.02994047976913365, + "grad_norm": 1.8006078011811224, + "learning_rate": 3.99709640513106e-06, + "loss": 0.7952, + "step": 249 + }, + { + "epoch": 0.03006072265977274, + "grad_norm": 2.179500229970487, + "learning_rate": 4e-06, + "loss": 0.9695, + "step": 250 + }, + { + "epoch": 0.03018096555041183, + "grad_norm": 2.1696435456272045, + "learning_rate": 3.999999848300794e-06, + "loss": 1.0014, + "step": 251 + }, + { + "epoch": 0.030301208441050925, + "grad_norm": 1.6704911538797294, + "learning_rate": 3.999999393203203e-06, + "loss": 0.9547, + "step": 252 + }, + { + "epoch": 0.030421451331690014, + "grad_norm": 1.7177113949586669, + "learning_rate": 3.999998634707293e-06, + "loss": 0.9174, + "step": 253 + }, + { + "epoch": 0.030541694222329104, + "grad_norm": 2.2290166881650224, + "learning_rate": 3.999997572813182e-06, + "loss": 1.0433, + "step": 254 + }, + { + "epoch": 0.030661937112968194, + "grad_norm": 1.977998114157801, + "learning_rate": 3.999996207521028e-06, + "loss": 0.9509, + "step": 255 + }, + { + "epoch": 0.030782180003607287, + "grad_norm": 2.687123738053769, + "learning_rate": 3.999994538831039e-06, + "loss": 0.9465, + "step": 256 + }, + { + "epoch": 0.030902422894246377, + "grad_norm": 2.175756813884604, + "learning_rate": 3.99999256674347e-06, + "loss": 0.923, + "step": 257 + }, + { + "epoch": 0.031022665784885467, + "grad_norm": 0.9927704367703919, + "learning_rate": 3.999990291258618e-06, + "loss": 0.5371, + "step": 258 + }, + { + "epoch": 0.03114290867552456, + "grad_norm": 2.3988978040520794, + "learning_rate": 3.999987712376829e-06, + "loss": 0.9258, + "step": 259 + }, + { + "epoch": 0.031263151566163654, + "grad_norm": 2.453879465116673, + "learning_rate": 3.999984830098494e-06, + "loss": 0.8985, + "step": 260 + }, + { + "epoch": 0.03138339445680274, + "grad_norm": 2.513452385461466, + "learning_rate": 3.999981644424051e-06, + "loss": 1.0662, + "step": 261 + }, + { + "epoch": 0.03150363734744183, + "grad_norm": 2.483561256582522, + "learning_rate": 3.999978155353982e-06, + "loss": 0.9613, + "step": 262 + }, + { + "epoch": 0.03162388023808092, + "grad_norm": 2.090151755194156, + "learning_rate": 3.9999743628888186e-06, + "loss": 0.8662, + "step": 263 + }, + { + "epoch": 0.03174412312872001, + "grad_norm": 2.493891883580711, + "learning_rate": 3.999970267029133e-06, + "loss": 0.9536, + "step": 264 + }, + { + "epoch": 0.0318643660193591, + "grad_norm": 2.054245585794895, + "learning_rate": 3.999965867775548e-06, + "loss": 0.859, + "step": 265 + }, + { + "epoch": 0.0319846089099982, + "grad_norm": 3.0822580541746363, + "learning_rate": 3.9999611651287315e-06, + "loss": 0.9491, + "step": 266 + }, + { + "epoch": 0.03210485180063729, + "grad_norm": 2.900837277529639, + "learning_rate": 3.999956159089396e-06, + "loss": 0.8852, + "step": 267 + }, + { + "epoch": 0.03222509469127638, + "grad_norm": 2.1255051629491257, + "learning_rate": 3.999950849658302e-06, + "loss": 0.8508, + "step": 268 + }, + { + "epoch": 0.03234533758191547, + "grad_norm": 2.6061245894961664, + "learning_rate": 3.999945236836254e-06, + "loss": 0.9244, + "step": 269 + }, + { + "epoch": 0.03246558047255456, + "grad_norm": 3.289052061666928, + "learning_rate": 3.999939320624103e-06, + "loss": 1.045, + "step": 270 + }, + { + "epoch": 0.03258582336319365, + "grad_norm": 2.129161910358449, + "learning_rate": 3.999933101022749e-06, + "loss": 0.954, + "step": 271 + }, + { + "epoch": 0.032706066253832745, + "grad_norm": 2.2280020360201123, + "learning_rate": 3.999926578033132e-06, + "loss": 0.9124, + "step": 272 + }, + { + "epoch": 0.032826309144471835, + "grad_norm": 2.255703584650297, + "learning_rate": 3.999919751656244e-06, + "loss": 0.7146, + "step": 273 + }, + { + "epoch": 0.032946552035110925, + "grad_norm": 2.5050021672953293, + "learning_rate": 3.9999126218931195e-06, + "loss": 0.8226, + "step": 274 + }, + { + "epoch": 0.033066794925750015, + "grad_norm": 2.4832142931507155, + "learning_rate": 3.99990518874484e-06, + "loss": 0.9527, + "step": 275 + }, + { + "epoch": 0.033187037816389105, + "grad_norm": 2.255282452945954, + "learning_rate": 3.999897452212534e-06, + "loss": 0.9653, + "step": 276 + }, + { + "epoch": 0.033307280707028195, + "grad_norm": 2.0705409661675676, + "learning_rate": 3.999889412297374e-06, + "loss": 1.047, + "step": 277 + }, + { + "epoch": 0.03342752359766729, + "grad_norm": 2.007341381566588, + "learning_rate": 3.999881069000581e-06, + "loss": 0.8308, + "step": 278 + }, + { + "epoch": 0.03354776648830638, + "grad_norm": 2.990450282558808, + "learning_rate": 3.99987242232342e-06, + "loss": 0.9346, + "step": 279 + }, + { + "epoch": 0.03366800937894547, + "grad_norm": 1.925949924966798, + "learning_rate": 3.9998634722672026e-06, + "loss": 0.8386, + "step": 280 + }, + { + "epoch": 0.03378825226958456, + "grad_norm": 2.1131418009385987, + "learning_rate": 3.999854218833286e-06, + "loss": 0.8308, + "step": 281 + }, + { + "epoch": 0.03390849516022365, + "grad_norm": 2.1461522380800746, + "learning_rate": 3.999844662023075e-06, + "loss": 0.8712, + "step": 282 + }, + { + "epoch": 0.03402873805086274, + "grad_norm": 1.9260293363592886, + "learning_rate": 3.999834801838018e-06, + "loss": 0.9457, + "step": 283 + }, + { + "epoch": 0.03414898094150183, + "grad_norm": 2.4203489356256727, + "learning_rate": 3.9998246382796115e-06, + "loss": 0.789, + "step": 284 + }, + { + "epoch": 0.03426922383214093, + "grad_norm": 2.2716649017364445, + "learning_rate": 3.999814171349399e-06, + "loss": 0.9841, + "step": 285 + }, + { + "epoch": 0.03438946672278002, + "grad_norm": 2.1464863743308475, + "learning_rate": 3.9998034010489655e-06, + "loss": 0.7778, + "step": 286 + }, + { + "epoch": 0.03450970961341911, + "grad_norm": 2.397075850564979, + "learning_rate": 3.999792327379946e-06, + "loss": 0.8073, + "step": 287 + }, + { + "epoch": 0.034629952504058197, + "grad_norm": 2.328516268014139, + "learning_rate": 3.999780950344021e-06, + "loss": 1.0165, + "step": 288 + }, + { + "epoch": 0.034750195394697286, + "grad_norm": 2.4717219366820755, + "learning_rate": 3.999769269942916e-06, + "loss": 0.8692, + "step": 289 + }, + { + "epoch": 0.034870438285336376, + "grad_norm": 1.7393722025348128, + "learning_rate": 3.999757286178402e-06, + "loss": 0.8428, + "step": 290 + }, + { + "epoch": 0.03499068117597547, + "grad_norm": 1.841576539793148, + "learning_rate": 3.999744999052299e-06, + "loss": 0.9333, + "step": 291 + }, + { + "epoch": 0.03511092406661456, + "grad_norm": 0.9750729301270854, + "learning_rate": 3.9997324085664675e-06, + "loss": 0.6128, + "step": 292 + }, + { + "epoch": 0.03523116695725365, + "grad_norm": 2.014165778874728, + "learning_rate": 3.999719514722821e-06, + "loss": 0.9487, + "step": 293 + }, + { + "epoch": 0.03535140984789274, + "grad_norm": 2.53166233048142, + "learning_rate": 3.999706317523314e-06, + "loss": 0.7869, + "step": 294 + }, + { + "epoch": 0.03547165273853183, + "grad_norm": 2.7793218379665574, + "learning_rate": 3.999692816969948e-06, + "loss": 0.8994, + "step": 295 + }, + { + "epoch": 0.03559189562917092, + "grad_norm": 1.0413996290610283, + "learning_rate": 3.999679013064772e-06, + "loss": 0.6973, + "step": 296 + }, + { + "epoch": 0.03571213851981002, + "grad_norm": 2.5908744222581728, + "learning_rate": 3.99966490580988e-06, + "loss": 0.9158, + "step": 297 + }, + { + "epoch": 0.03583238141044911, + "grad_norm": 2.6189682098009044, + "learning_rate": 3.999650495207411e-06, + "loss": 0.7117, + "step": 298 + }, + { + "epoch": 0.0359526243010882, + "grad_norm": 2.712934302794765, + "learning_rate": 3.999635781259553e-06, + "loss": 0.9377, + "step": 299 + }, + { + "epoch": 0.03607286719172729, + "grad_norm": 0.9308465096324603, + "learning_rate": 3.999620763968535e-06, + "loss": 0.5255, + "step": 300 + }, + { + "epoch": 0.03619311008236638, + "grad_norm": 2.7471150090513725, + "learning_rate": 3.999605443336638e-06, + "loss": 0.8844, + "step": 301 + }, + { + "epoch": 0.03631335297300547, + "grad_norm": 2.6400586485850557, + "learning_rate": 3.999589819366185e-06, + "loss": 0.9458, + "step": 302 + }, + { + "epoch": 0.036433595863644565, + "grad_norm": 2.2290521990421595, + "learning_rate": 3.999573892059547e-06, + "loss": 0.8761, + "step": 303 + }, + { + "epoch": 0.036553838754283655, + "grad_norm": 2.1241875385602396, + "learning_rate": 3.999557661419138e-06, + "loss": 0.8565, + "step": 304 + }, + { + "epoch": 0.036674081644922744, + "grad_norm": 2.0682408746448653, + "learning_rate": 3.9995411274474225e-06, + "loss": 0.8453, + "step": 305 + }, + { + "epoch": 0.036794324535561834, + "grad_norm": 2.388531335297334, + "learning_rate": 3.999524290146908e-06, + "loss": 0.8555, + "step": 306 + }, + { + "epoch": 0.036914567426200924, + "grad_norm": 2.1689334407036647, + "learning_rate": 3.9995071495201485e-06, + "loss": 0.9549, + "step": 307 + }, + { + "epoch": 0.037034810316840014, + "grad_norm": 2.247301178519492, + "learning_rate": 3.999489705569744e-06, + "loss": 1.0165, + "step": 308 + }, + { + "epoch": 0.03715505320747911, + "grad_norm": 2.147869687039901, + "learning_rate": 3.999471958298341e-06, + "loss": 0.9023, + "step": 309 + }, + { + "epoch": 0.0372752960981182, + "grad_norm": 1.9246001113101912, + "learning_rate": 3.999453907708631e-06, + "loss": 0.7933, + "step": 310 + }, + { + "epoch": 0.03739553898875729, + "grad_norm": 1.9056021418065383, + "learning_rate": 3.999435553803353e-06, + "loss": 0.839, + "step": 311 + }, + { + "epoch": 0.03751578187939638, + "grad_norm": 2.445743608101598, + "learning_rate": 3.999416896585292e-06, + "loss": 0.8854, + "step": 312 + }, + { + "epoch": 0.03763602477003547, + "grad_norm": 4.0911055315421585, + "learning_rate": 3.9993979360572775e-06, + "loss": 0.9254, + "step": 313 + }, + { + "epoch": 0.03775626766067456, + "grad_norm": 2.446031556120067, + "learning_rate": 3.999378672222185e-06, + "loss": 0.8771, + "step": 314 + }, + { + "epoch": 0.03787651055131366, + "grad_norm": 2.6185463181995923, + "learning_rate": 3.9993591050829385e-06, + "loss": 0.8599, + "step": 315 + }, + { + "epoch": 0.037996753441952746, + "grad_norm": 1.8840099658054552, + "learning_rate": 3.999339234642506e-06, + "loss": 0.8175, + "step": 316 + }, + { + "epoch": 0.038116996332591836, + "grad_norm": 2.128725600671181, + "learning_rate": 3.9993190609038994e-06, + "loss": 0.8806, + "step": 317 + }, + { + "epoch": 0.038237239223230926, + "grad_norm": 2.0591662658583982, + "learning_rate": 3.999298583870182e-06, + "loss": 0.8638, + "step": 318 + }, + { + "epoch": 0.038357482113870016, + "grad_norm": 1.9314014066831868, + "learning_rate": 3.999277803544458e-06, + "loss": 0.8054, + "step": 319 + }, + { + "epoch": 0.038477725004509106, + "grad_norm": 0.9476618301969493, + "learning_rate": 3.999256719929882e-06, + "loss": 0.6251, + "step": 320 + }, + { + "epoch": 0.0385979678951482, + "grad_norm": 1.2147759912367353, + "learning_rate": 3.999235333029651e-06, + "loss": 0.6736, + "step": 321 + }, + { + "epoch": 0.03871821078578729, + "grad_norm": 1.9012571517538959, + "learning_rate": 3.999213642847009e-06, + "loss": 0.8424, + "step": 322 + }, + { + "epoch": 0.03883845367642638, + "grad_norm": 1.8622286400473225, + "learning_rate": 3.999191649385247e-06, + "loss": 0.939, + "step": 323 + }, + { + "epoch": 0.03895869656706547, + "grad_norm": 0.9277167388845902, + "learning_rate": 3.999169352647702e-06, + "loss": 0.5999, + "step": 324 + }, + { + "epoch": 0.03907893945770456, + "grad_norm": 2.2808609722201765, + "learning_rate": 3.999146752637755e-06, + "loss": 0.8696, + "step": 325 + }, + { + "epoch": 0.03919918234834365, + "grad_norm": 2.7379449700453407, + "learning_rate": 3.999123849358836e-06, + "loss": 0.9454, + "step": 326 + }, + { + "epoch": 0.03931942523898275, + "grad_norm": 2.142503234611571, + "learning_rate": 3.999100642814418e-06, + "loss": 0.7836, + "step": 327 + }, + { + "epoch": 0.03943966812962184, + "grad_norm": 2.2853886934398138, + "learning_rate": 3.999077133008022e-06, + "loss": 0.9161, + "step": 328 + }, + { + "epoch": 0.03955991102026093, + "grad_norm": 1.9391698716873067, + "learning_rate": 3.9990533199432145e-06, + "loss": 0.933, + "step": 329 + }, + { + "epoch": 0.03968015391090002, + "grad_norm": 2.3960618172760593, + "learning_rate": 3.999029203623608e-06, + "loss": 0.7927, + "step": 330 + }, + { + "epoch": 0.03980039680153911, + "grad_norm": 2.138114738183261, + "learning_rate": 3.99900478405286e-06, + "loss": 0.9006, + "step": 331 + }, + { + "epoch": 0.0399206396921782, + "grad_norm": 2.470879421050894, + "learning_rate": 3.998980061234676e-06, + "loss": 0.8594, + "step": 332 + }, + { + "epoch": 0.040040882582817294, + "grad_norm": 3.5375484234087065, + "learning_rate": 3.9989550351728055e-06, + "loss": 0.8142, + "step": 333 + }, + { + "epoch": 0.040161125473456384, + "grad_norm": 2.2031986954720484, + "learning_rate": 3.998929705871046e-06, + "loss": 0.881, + "step": 334 + }, + { + "epoch": 0.040281368364095474, + "grad_norm": 2.4638143526959944, + "learning_rate": 3.99890407333324e-06, + "loss": 0.9204, + "step": 335 + }, + { + "epoch": 0.040401611254734564, + "grad_norm": 1.6881437133520154, + "learning_rate": 3.998878137563275e-06, + "loss": 0.8959, + "step": 336 + }, + { + "epoch": 0.040521854145373654, + "grad_norm": 2.3500196902485015, + "learning_rate": 3.998851898565085e-06, + "loss": 0.887, + "step": 337 + }, + { + "epoch": 0.04064209703601274, + "grad_norm": 1.8703994259117949, + "learning_rate": 3.998825356342653e-06, + "loss": 0.8639, + "step": 338 + }, + { + "epoch": 0.04076233992665183, + "grad_norm": 2.2862461718677904, + "learning_rate": 3.998798510900003e-06, + "loss": 0.7612, + "step": 339 + }, + { + "epoch": 0.04088258281729093, + "grad_norm": 2.0858022810295553, + "learning_rate": 3.998771362241207e-06, + "loss": 0.8612, + "step": 340 + }, + { + "epoch": 0.04100282570793002, + "grad_norm": 1.814621834359269, + "learning_rate": 3.998743910370385e-06, + "loss": 0.8958, + "step": 341 + }, + { + "epoch": 0.04112306859856911, + "grad_norm": 2.2707747238948404, + "learning_rate": 3.998716155291702e-06, + "loss": 0.7532, + "step": 342 + }, + { + "epoch": 0.0412433114892082, + "grad_norm": 1.7116802815406884, + "learning_rate": 3.998688097009366e-06, + "loss": 0.9268, + "step": 343 + }, + { + "epoch": 0.04136355437984729, + "grad_norm": 2.4292451020321253, + "learning_rate": 3.998659735527636e-06, + "loss": 0.835, + "step": 344 + }, + { + "epoch": 0.04148379727048638, + "grad_norm": 1.6920807442844512, + "learning_rate": 3.998631070850813e-06, + "loss": 0.7974, + "step": 345 + }, + { + "epoch": 0.041604040161125476, + "grad_norm": 2.1187810214762104, + "learning_rate": 3.9986021029832455e-06, + "loss": 0.859, + "step": 346 + }, + { + "epoch": 0.041724283051764566, + "grad_norm": 2.9273358074372666, + "learning_rate": 3.9985728319293285e-06, + "loss": 0.9634, + "step": 347 + }, + { + "epoch": 0.041844525942403656, + "grad_norm": 2.298477236821112, + "learning_rate": 3.998543257693501e-06, + "loss": 0.8778, + "step": 348 + }, + { + "epoch": 0.041964768833042745, + "grad_norm": 1.8314938926376252, + "learning_rate": 3.998513380280251e-06, + "loss": 0.907, + "step": 349 + }, + { + "epoch": 0.042085011723681835, + "grad_norm": 2.0986887062124047, + "learning_rate": 3.99848319969411e-06, + "loss": 0.9816, + "step": 350 + }, + { + "epoch": 0.042205254614320925, + "grad_norm": 2.460159033310092, + "learning_rate": 3.9984527159396564e-06, + "loss": 0.8306, + "step": 351 + }, + { + "epoch": 0.04232549750496002, + "grad_norm": 2.3736342468483387, + "learning_rate": 3.9984219290215154e-06, + "loss": 0.8748, + "step": 352 + }, + { + "epoch": 0.04244574039559911, + "grad_norm": 1.6667262146166528, + "learning_rate": 3.998390838944356e-06, + "loss": 0.9108, + "step": 353 + }, + { + "epoch": 0.0425659832862382, + "grad_norm": 2.1125673226909507, + "learning_rate": 3.998359445712895e-06, + "loss": 0.9323, + "step": 354 + }, + { + "epoch": 0.04268622617687729, + "grad_norm": 2.3703470326915013, + "learning_rate": 3.9983277493318955e-06, + "loss": 0.8392, + "step": 355 + }, + { + "epoch": 0.04280646906751638, + "grad_norm": 2.0180632754822376, + "learning_rate": 3.998295749806165e-06, + "loss": 0.8323, + "step": 356 + }, + { + "epoch": 0.04292671195815547, + "grad_norm": 2.163418450211737, + "learning_rate": 3.998263447140558e-06, + "loss": 0.8681, + "step": 357 + }, + { + "epoch": 0.04304695484879457, + "grad_norm": 1.9109114263819056, + "learning_rate": 3.998230841339976e-06, + "loss": 0.8489, + "step": 358 + }, + { + "epoch": 0.04316719773943366, + "grad_norm": 2.2779533918132384, + "learning_rate": 3.998197932409363e-06, + "loss": 0.8687, + "step": 359 + }, + { + "epoch": 0.04328744063007275, + "grad_norm": 3.0137603212778696, + "learning_rate": 3.9981647203537125e-06, + "loss": 0.8913, + "step": 360 + }, + { + "epoch": 0.04340768352071184, + "grad_norm": 1.9173163957488786, + "learning_rate": 3.998131205178063e-06, + "loss": 0.9871, + "step": 361 + }, + { + "epoch": 0.04352792641135093, + "grad_norm": 2.4241429692335443, + "learning_rate": 3.998097386887498e-06, + "loss": 0.791, + "step": 362 + }, + { + "epoch": 0.04364816930199002, + "grad_norm": 2.0387413970559933, + "learning_rate": 3.998063265487148e-06, + "loss": 0.8661, + "step": 363 + }, + { + "epoch": 0.043768412192629114, + "grad_norm": 2.009690412146443, + "learning_rate": 3.99802884098219e-06, + "loss": 0.8385, + "step": 364 + }, + { + "epoch": 0.043888655083268203, + "grad_norm": 2.402638387832586, + "learning_rate": 3.997994113377845e-06, + "loss": 0.8505, + "step": 365 + }, + { + "epoch": 0.04400889797390729, + "grad_norm": 2.076572153544035, + "learning_rate": 3.9979590826793815e-06, + "loss": 0.8578, + "step": 366 + }, + { + "epoch": 0.04412914086454638, + "grad_norm": 1.9146027717253873, + "learning_rate": 3.997923748892113e-06, + "loss": 0.8338, + "step": 367 + }, + { + "epoch": 0.04424938375518547, + "grad_norm": 1.828801152315052, + "learning_rate": 3.9978881120214015e-06, + "loss": 0.9011, + "step": 368 + }, + { + "epoch": 0.04436962664582456, + "grad_norm": 2.244905319936076, + "learning_rate": 3.997852172072652e-06, + "loss": 0.8174, + "step": 369 + }, + { + "epoch": 0.04448986953646366, + "grad_norm": 2.615602747960389, + "learning_rate": 3.9978159290513155e-06, + "loss": 0.924, + "step": 370 + }, + { + "epoch": 0.04461011242710275, + "grad_norm": 1.8748007986427984, + "learning_rate": 3.997779382962892e-06, + "loss": 0.8221, + "step": 371 + }, + { + "epoch": 0.04473035531774184, + "grad_norm": 1.7116197593127893, + "learning_rate": 3.997742533812924e-06, + "loss": 0.7591, + "step": 372 + }, + { + "epoch": 0.04485059820838093, + "grad_norm": 2.660281095782068, + "learning_rate": 3.997705381607001e-06, + "loss": 0.9548, + "step": 373 + }, + { + "epoch": 0.04497084109902002, + "grad_norm": 1.1895006793558587, + "learning_rate": 3.997667926350761e-06, + "loss": 0.6044, + "step": 374 + }, + { + "epoch": 0.04509108398965911, + "grad_norm": 0.910110014039863, + "learning_rate": 3.997630168049886e-06, + "loss": 0.5793, + "step": 375 + }, + { + "epoch": 0.045211326880298205, + "grad_norm": 1.8769037754267808, + "learning_rate": 3.997592106710101e-06, + "loss": 0.7997, + "step": 376 + }, + { + "epoch": 0.045331569770937295, + "grad_norm": 2.3607131497965947, + "learning_rate": 3.997553742337182e-06, + "loss": 0.6892, + "step": 377 + }, + { + "epoch": 0.045451812661576385, + "grad_norm": 1.8603492556580186, + "learning_rate": 3.997515074936949e-06, + "loss": 0.9302, + "step": 378 + }, + { + "epoch": 0.045572055552215475, + "grad_norm": 2.1986931126941283, + "learning_rate": 3.997476104515268e-06, + "loss": 0.9067, + "step": 379 + }, + { + "epoch": 0.045692298442854565, + "grad_norm": 1.8661453978140263, + "learning_rate": 3.9974368310780485e-06, + "loss": 0.7853, + "step": 380 + }, + { + "epoch": 0.045812541333493655, + "grad_norm": 2.2489362957363364, + "learning_rate": 3.997397254631251e-06, + "loss": 0.7761, + "step": 381 + }, + { + "epoch": 0.04593278422413275, + "grad_norm": 0.8321058571189303, + "learning_rate": 3.997357375180878e-06, + "loss": 0.6025, + "step": 382 + }, + { + "epoch": 0.04605302711477184, + "grad_norm": 1.9747667561192153, + "learning_rate": 3.997317192732979e-06, + "loss": 0.779, + "step": 383 + }, + { + "epoch": 0.04617327000541093, + "grad_norm": 1.8420285739566988, + "learning_rate": 3.99727670729365e-06, + "loss": 0.8485, + "step": 384 + }, + { + "epoch": 0.04629351289605002, + "grad_norm": 1.6965338882223127, + "learning_rate": 3.997235918869033e-06, + "loss": 0.7955, + "step": 385 + }, + { + "epoch": 0.04641375578668911, + "grad_norm": 2.0250967873884487, + "learning_rate": 3.997194827465315e-06, + "loss": 0.8501, + "step": 386 + }, + { + "epoch": 0.0465339986773282, + "grad_norm": 2.5666274348476725, + "learning_rate": 3.997153433088728e-06, + "loss": 0.9555, + "step": 387 + }, + { + "epoch": 0.0466542415679673, + "grad_norm": 2.060362777802386, + "learning_rate": 3.997111735745554e-06, + "loss": 0.8455, + "step": 388 + }, + { + "epoch": 0.04677448445860639, + "grad_norm": 2.043601410390084, + "learning_rate": 3.997069735442118e-06, + "loss": 0.8519, + "step": 389 + }, + { + "epoch": 0.04689472734924548, + "grad_norm": 1.5927445475135156, + "learning_rate": 3.997027432184792e-06, + "loss": 0.8153, + "step": 390 + }, + { + "epoch": 0.04701497023988457, + "grad_norm": 1.9202491892597011, + "learning_rate": 3.99698482597999e-06, + "loss": 0.9079, + "step": 391 + }, + { + "epoch": 0.04713521313052366, + "grad_norm": 0.8840494183367718, + "learning_rate": 3.99694191683418e-06, + "loss": 0.641, + "step": 392 + }, + { + "epoch": 0.047255456021162746, + "grad_norm": 2.1685779162467296, + "learning_rate": 3.996898704753867e-06, + "loss": 0.8511, + "step": 393 + }, + { + "epoch": 0.04737569891180184, + "grad_norm": 2.3809644192436696, + "learning_rate": 3.996855189745609e-06, + "loss": 0.9005, + "step": 394 + }, + { + "epoch": 0.04749594180244093, + "grad_norm": 2.140357676054475, + "learning_rate": 3.996811371816007e-06, + "loss": 0.9524, + "step": 395 + }, + { + "epoch": 0.04761618469308002, + "grad_norm": 3.261494776179908, + "learning_rate": 3.996767250971707e-06, + "loss": 0.8028, + "step": 396 + }, + { + "epoch": 0.04773642758371911, + "grad_norm": 1.7991434788001095, + "learning_rate": 3.996722827219403e-06, + "loss": 0.8878, + "step": 397 + }, + { + "epoch": 0.0478566704743582, + "grad_norm": 2.5431940356897123, + "learning_rate": 3.996678100565833e-06, + "loss": 0.8575, + "step": 398 + }, + { + "epoch": 0.04797691336499729, + "grad_norm": 3.662940530680075, + "learning_rate": 3.996633071017783e-06, + "loss": 0.9273, + "step": 399 + }, + { + "epoch": 0.04809715625563638, + "grad_norm": 2.2355162366986865, + "learning_rate": 3.996587738582084e-06, + "loss": 0.8385, + "step": 400 + }, + { + "epoch": 0.04821739914627548, + "grad_norm": 2.3230728530265146, + "learning_rate": 3.9965421032656115e-06, + "loss": 0.8816, + "step": 401 + }, + { + "epoch": 0.04833764203691457, + "grad_norm": 2.4031098919655234, + "learning_rate": 3.99649616507529e-06, + "loss": 0.9725, + "step": 402 + }, + { + "epoch": 0.04845788492755366, + "grad_norm": 0.8949096079418112, + "learning_rate": 3.996449924018088e-06, + "loss": 0.6324, + "step": 403 + }, + { + "epoch": 0.04857812781819275, + "grad_norm": 1.7396888777033765, + "learning_rate": 3.99640338010102e-06, + "loss": 0.8151, + "step": 404 + }, + { + "epoch": 0.04869837070883184, + "grad_norm": 1.89027704236853, + "learning_rate": 3.996356533331146e-06, + "loss": 0.8109, + "step": 405 + }, + { + "epoch": 0.04881861359947093, + "grad_norm": 2.719221052381337, + "learning_rate": 3.996309383715573e-06, + "loss": 0.6623, + "step": 406 + }, + { + "epoch": 0.048938856490110025, + "grad_norm": 8.370235625201165, + "learning_rate": 3.996261931261454e-06, + "loss": 0.77, + "step": 407 + }, + { + "epoch": 0.049059099380749115, + "grad_norm": 1.7217795288425177, + "learning_rate": 3.996214175975987e-06, + "loss": 0.881, + "step": 408 + }, + { + "epoch": 0.049179342271388204, + "grad_norm": 3.2535105598621876, + "learning_rate": 3.996166117866417e-06, + "loss": 0.8136, + "step": 409 + }, + { + "epoch": 0.049299585162027294, + "grad_norm": 1.9616233568534078, + "learning_rate": 3.996117756940035e-06, + "loss": 0.8854, + "step": 410 + }, + { + "epoch": 0.049419828052666384, + "grad_norm": 2.0447444164732986, + "learning_rate": 3.996069093204175e-06, + "loss": 0.9998, + "step": 411 + }, + { + "epoch": 0.049540070943305474, + "grad_norm": 2.3525204149063765, + "learning_rate": 3.996020126666221e-06, + "loss": 0.8947, + "step": 412 + }, + { + "epoch": 0.04966031383394457, + "grad_norm": 2.0653601957133554, + "learning_rate": 3.995970857333601e-06, + "loss": 0.84, + "step": 413 + }, + { + "epoch": 0.04978055672458366, + "grad_norm": 1.892862052602647, + "learning_rate": 3.995921285213789e-06, + "loss": 0.8132, + "step": 414 + }, + { + "epoch": 0.04990079961522275, + "grad_norm": 2.2833392883371633, + "learning_rate": 3.995871410314305e-06, + "loss": 0.8378, + "step": 415 + }, + { + "epoch": 0.05002104250586184, + "grad_norm": 0.9315792359413485, + "learning_rate": 3.995821232642714e-06, + "loss": 0.598, + "step": 416 + }, + { + "epoch": 0.05014128539650093, + "grad_norm": 1.9940762411366568, + "learning_rate": 3.995770752206629e-06, + "loss": 0.8361, + "step": 417 + }, + { + "epoch": 0.05026152828714002, + "grad_norm": 2.0972441938090034, + "learning_rate": 3.995719969013709e-06, + "loss": 0.9915, + "step": 418 + }, + { + "epoch": 0.05038177117777912, + "grad_norm": 3.3913467974966984, + "learning_rate": 3.995668883071655e-06, + "loss": 0.9043, + "step": 419 + }, + { + "epoch": 0.050502014068418206, + "grad_norm": 2.4435796966138863, + "learning_rate": 3.995617494388219e-06, + "loss": 0.937, + "step": 420 + }, + { + "epoch": 0.050622256959057296, + "grad_norm": 2.435701146627721, + "learning_rate": 3.995565802971196e-06, + "loss": 0.8304, + "step": 421 + }, + { + "epoch": 0.050742499849696386, + "grad_norm": 1.8561177935545483, + "learning_rate": 3.995513808828427e-06, + "loss": 0.6972, + "step": 422 + }, + { + "epoch": 0.050862742740335476, + "grad_norm": 1.9626284296083427, + "learning_rate": 3.9954615119678e-06, + "loss": 0.7875, + "step": 423 + }, + { + "epoch": 0.050982985630974566, + "grad_norm": 2.281622242277963, + "learning_rate": 3.995408912397248e-06, + "loss": 0.8301, + "step": 424 + }, + { + "epoch": 0.05110322852161366, + "grad_norm": 2.5717433668636778, + "learning_rate": 3.99535601012475e-06, + "loss": 0.9488, + "step": 425 + }, + { + "epoch": 0.05122347141225275, + "grad_norm": 1.63275365404016, + "learning_rate": 3.995302805158333e-06, + "loss": 0.7677, + "step": 426 + }, + { + "epoch": 0.05134371430289184, + "grad_norm": 2.2904310074037406, + "learning_rate": 3.9952492975060665e-06, + "loss": 0.8594, + "step": 427 + }, + { + "epoch": 0.05146395719353093, + "grad_norm": 2.2851858805504186, + "learning_rate": 3.995195487176067e-06, + "loss": 0.8867, + "step": 428 + }, + { + "epoch": 0.05158420008417002, + "grad_norm": 1.7564988038112999, + "learning_rate": 3.995141374176499e-06, + "loss": 0.8737, + "step": 429 + }, + { + "epoch": 0.05170444297480911, + "grad_norm": 0.8870758480319428, + "learning_rate": 3.995086958515572e-06, + "loss": 0.6325, + "step": 430 + }, + { + "epoch": 0.05182468586544821, + "grad_norm": 0.8655859101761695, + "learning_rate": 3.995032240201538e-06, + "loss": 0.6004, + "step": 431 + }, + { + "epoch": 0.0519449287560873, + "grad_norm": 0.9477832127343596, + "learning_rate": 3.9949772192427e-06, + "loss": 0.6322, + "step": 432 + }, + { + "epoch": 0.05206517164672639, + "grad_norm": 1.8556528742936453, + "learning_rate": 3.994921895647405e-06, + "loss": 0.8126, + "step": 433 + }, + { + "epoch": 0.05218541453736548, + "grad_norm": 0.8380492089214953, + "learning_rate": 3.994866269424043e-06, + "loss": 0.5543, + "step": 434 + }, + { + "epoch": 0.05230565742800457, + "grad_norm": 2.290562265325483, + "learning_rate": 3.9948103405810545e-06, + "loss": 0.8077, + "step": 435 + }, + { + "epoch": 0.05242590031864366, + "grad_norm": 1.970242494675081, + "learning_rate": 3.994754109126923e-06, + "loss": 0.8748, + "step": 436 + }, + { + "epoch": 0.052546143209282754, + "grad_norm": 2.29790344050023, + "learning_rate": 3.994697575070181e-06, + "loss": 0.9486, + "step": 437 + }, + { + "epoch": 0.052666386099921844, + "grad_norm": 2.109294635513379, + "learning_rate": 3.994640738419402e-06, + "loss": 0.9294, + "step": 438 + }, + { + "epoch": 0.052786628990560934, + "grad_norm": 2.1168705889574513, + "learning_rate": 3.9945835991832075e-06, + "loss": 0.8289, + "step": 439 + }, + { + "epoch": 0.052906871881200024, + "grad_norm": 2.102167886092731, + "learning_rate": 3.994526157370268e-06, + "loss": 0.9456, + "step": 440 + }, + { + "epoch": 0.053027114771839114, + "grad_norm": 0.8896682686799043, + "learning_rate": 3.994468412989296e-06, + "loss": 0.5935, + "step": 441 + }, + { + "epoch": 0.053147357662478203, + "grad_norm": 2.3379503846398446, + "learning_rate": 3.994410366049052e-06, + "loss": 0.9465, + "step": 442 + }, + { + "epoch": 0.0532676005531173, + "grad_norm": 2.6495325755895336, + "learning_rate": 3.994352016558341e-06, + "loss": 0.8633, + "step": 443 + }, + { + "epoch": 0.05338784344375639, + "grad_norm": 2.1671419233770433, + "learning_rate": 3.994293364526014e-06, + "loss": 0.7571, + "step": 444 + }, + { + "epoch": 0.05350808633439548, + "grad_norm": 1.990065038662938, + "learning_rate": 3.99423440996097e-06, + "loss": 0.8604, + "step": 445 + }, + { + "epoch": 0.05362832922503457, + "grad_norm": 2.497723496587497, + "learning_rate": 3.994175152872152e-06, + "loss": 0.8432, + "step": 446 + }, + { + "epoch": 0.05374857211567366, + "grad_norm": 1.9304809174719622, + "learning_rate": 3.994115593268548e-06, + "loss": 0.7961, + "step": 447 + }, + { + "epoch": 0.05386881500631275, + "grad_norm": 2.9081024500917247, + "learning_rate": 3.994055731159195e-06, + "loss": 0.842, + "step": 448 + }, + { + "epoch": 0.053989057896951846, + "grad_norm": 1.876901615956365, + "learning_rate": 3.993995566553172e-06, + "loss": 0.8888, + "step": 449 + }, + { + "epoch": 0.054109300787590936, + "grad_norm": 1.9428925073234051, + "learning_rate": 3.993935099459607e-06, + "loss": 0.7914, + "step": 450 + }, + { + "epoch": 0.054229543678230026, + "grad_norm": 2.0544347571353256, + "learning_rate": 3.993874329887673e-06, + "loss": 0.754, + "step": 451 + }, + { + "epoch": 0.054349786568869116, + "grad_norm": 2.338735276486299, + "learning_rate": 3.993813257846589e-06, + "loss": 0.8934, + "step": 452 + }, + { + "epoch": 0.054470029459508205, + "grad_norm": 2.335725296231204, + "learning_rate": 3.993751883345619e-06, + "loss": 0.9612, + "step": 453 + }, + { + "epoch": 0.054590272350147295, + "grad_norm": 2.304172264308465, + "learning_rate": 3.993690206394073e-06, + "loss": 0.8925, + "step": 454 + }, + { + "epoch": 0.054710515240786385, + "grad_norm": 2.305527976712451, + "learning_rate": 3.993628227001307e-06, + "loss": 0.8967, + "step": 455 + }, + { + "epoch": 0.05483075813142548, + "grad_norm": 2.2370717240807547, + "learning_rate": 3.993565945176726e-06, + "loss": 0.7373, + "step": 456 + }, + { + "epoch": 0.05495100102206457, + "grad_norm": 3.0566802488446116, + "learning_rate": 3.993503360929776e-06, + "loss": 0.857, + "step": 457 + }, + { + "epoch": 0.05507124391270366, + "grad_norm": 1.637454866558984, + "learning_rate": 3.99344047426995e-06, + "loss": 0.828, + "step": 458 + }, + { + "epoch": 0.05519148680334275, + "grad_norm": 2.2434599452497346, + "learning_rate": 3.993377285206789e-06, + "loss": 0.9619, + "step": 459 + }, + { + "epoch": 0.05531172969398184, + "grad_norm": 1.750732064396709, + "learning_rate": 3.99331379374988e-06, + "loss": 0.8784, + "step": 460 + }, + { + "epoch": 0.05543197258462093, + "grad_norm": 2.000966315478277, + "learning_rate": 3.993249999908852e-06, + "loss": 0.8262, + "step": 461 + }, + { + "epoch": 0.05555221547526003, + "grad_norm": 1.9858982517496937, + "learning_rate": 3.993185903693384e-06, + "loss": 0.8863, + "step": 462 + }, + { + "epoch": 0.05567245836589912, + "grad_norm": 1.9805412203980508, + "learning_rate": 3.9931215051131995e-06, + "loss": 0.8324, + "step": 463 + }, + { + "epoch": 0.05579270125653821, + "grad_norm": 6.690203539114142, + "learning_rate": 3.993056804178068e-06, + "loss": 0.8135, + "step": 464 + }, + { + "epoch": 0.0559129441471773, + "grad_norm": 1.862048302991942, + "learning_rate": 3.992991800897803e-06, + "loss": 0.8645, + "step": 465 + }, + { + "epoch": 0.05603318703781639, + "grad_norm": 1.942715792262372, + "learning_rate": 3.9929264952822665e-06, + "loss": 0.9166, + "step": 466 + }, + { + "epoch": 0.05615342992845548, + "grad_norm": 2.516908117000179, + "learning_rate": 3.992860887341366e-06, + "loss": 0.9006, + "step": 467 + }, + { + "epoch": 0.056273672819094574, + "grad_norm": 2.102583059514121, + "learning_rate": 3.992794977085052e-06, + "loss": 0.8404, + "step": 468 + }, + { + "epoch": 0.056393915709733664, + "grad_norm": 2.9260061506101116, + "learning_rate": 3.992728764523326e-06, + "loss": 0.8631, + "step": 469 + }, + { + "epoch": 0.05651415860037275, + "grad_norm": 1.6170814958558835, + "learning_rate": 3.99266224966623e-06, + "loss": 0.8256, + "step": 470 + }, + { + "epoch": 0.05663440149101184, + "grad_norm": 2.252225834528322, + "learning_rate": 3.992595432523855e-06, + "loss": 0.8934, + "step": 471 + }, + { + "epoch": 0.05675464438165093, + "grad_norm": 1.9163620195288202, + "learning_rate": 3.992528313106338e-06, + "loss": 0.8764, + "step": 472 + }, + { + "epoch": 0.05687488727229002, + "grad_norm": 2.168724574398983, + "learning_rate": 3.9924608914238595e-06, + "loss": 0.8454, + "step": 473 + }, + { + "epoch": 0.05699513016292912, + "grad_norm": 2.64389052863018, + "learning_rate": 3.992393167486648e-06, + "loss": 0.8656, + "step": 474 + }, + { + "epoch": 0.05711537305356821, + "grad_norm": 2.4610043471203804, + "learning_rate": 3.992325141304977e-06, + "loss": 0.8272, + "step": 475 + }, + { + "epoch": 0.0572356159442073, + "grad_norm": 2.470995497461321, + "learning_rate": 3.992256812889166e-06, + "loss": 0.8807, + "step": 476 + }, + { + "epoch": 0.05735585883484639, + "grad_norm": 2.1063226160951056, + "learning_rate": 3.992188182249582e-06, + "loss": 0.7904, + "step": 477 + }, + { + "epoch": 0.05747610172548548, + "grad_norm": 2.3047261897573468, + "learning_rate": 3.992119249396633e-06, + "loss": 0.9347, + "step": 478 + }, + { + "epoch": 0.05759634461612457, + "grad_norm": 2.058466464021748, + "learning_rate": 3.992050014340778e-06, + "loss": 0.8404, + "step": 479 + }, + { + "epoch": 0.057716587506763666, + "grad_norm": 0.894891225022712, + "learning_rate": 3.99198047709252e-06, + "loss": 0.5523, + "step": 480 + }, + { + "epoch": 0.057836830397402755, + "grad_norm": 1.7179403999213374, + "learning_rate": 3.991910637662408e-06, + "loss": 0.8065, + "step": 481 + }, + { + "epoch": 0.057957073288041845, + "grad_norm": 1.7909407438808944, + "learning_rate": 3.9918404960610355e-06, + "loss": 0.8168, + "step": 482 + }, + { + "epoch": 0.058077316178680935, + "grad_norm": 2.48017134197601, + "learning_rate": 3.991770052299043e-06, + "loss": 0.794, + "step": 483 + }, + { + "epoch": 0.058197559069320025, + "grad_norm": 2.4559712026069582, + "learning_rate": 3.991699306387118e-06, + "loss": 0.9063, + "step": 484 + }, + { + "epoch": 0.058317801959959115, + "grad_norm": 2.908878081566914, + "learning_rate": 3.991628258335991e-06, + "loss": 0.8035, + "step": 485 + }, + { + "epoch": 0.05843804485059821, + "grad_norm": 2.838130218830363, + "learning_rate": 3.991556908156442e-06, + "loss": 0.912, + "step": 486 + }, + { + "epoch": 0.0585582877412373, + "grad_norm": 1.860867390244812, + "learning_rate": 3.9914852558592914e-06, + "loss": 0.8938, + "step": 487 + }, + { + "epoch": 0.05867853063187639, + "grad_norm": 3.8529895668947893, + "learning_rate": 3.991413301455413e-06, + "loss": 0.8457, + "step": 488 + }, + { + "epoch": 0.05879877352251548, + "grad_norm": 3.6703927152726004, + "learning_rate": 3.991341044955719e-06, + "loss": 0.7984, + "step": 489 + }, + { + "epoch": 0.05891901641315457, + "grad_norm": 1.9547345662720699, + "learning_rate": 3.991268486371172e-06, + "loss": 0.8309, + "step": 490 + }, + { + "epoch": 0.05903925930379366, + "grad_norm": 2.358985818255116, + "learning_rate": 3.991195625712779e-06, + "loss": 0.9009, + "step": 491 + }, + { + "epoch": 0.05915950219443276, + "grad_norm": 2.141519401424999, + "learning_rate": 3.991122462991592e-06, + "loss": 0.8262, + "step": 492 + }, + { + "epoch": 0.05927974508507185, + "grad_norm": 3.285928497084448, + "learning_rate": 3.991048998218712e-06, + "loss": 0.8536, + "step": 493 + }, + { + "epoch": 0.05939998797571094, + "grad_norm": 2.5146326986764747, + "learning_rate": 3.990975231405281e-06, + "loss": 0.7954, + "step": 494 + }, + { + "epoch": 0.05952023086635003, + "grad_norm": 2.153834773258912, + "learning_rate": 3.990901162562491e-06, + "loss": 0.8062, + "step": 495 + }, + { + "epoch": 0.05964047375698912, + "grad_norm": 3.3316528534773826, + "learning_rate": 3.9908267917015765e-06, + "loss": 0.925, + "step": 496 + }, + { + "epoch": 0.059760716647628206, + "grad_norm": 1.9940154422097016, + "learning_rate": 3.990752118833821e-06, + "loss": 0.9476, + "step": 497 + }, + { + "epoch": 0.0598809595382673, + "grad_norm": 1.7500820821858092, + "learning_rate": 3.990677143970553e-06, + "loss": 0.7944, + "step": 498 + }, + { + "epoch": 0.06000120242890639, + "grad_norm": 2.0449144131907797, + "learning_rate": 3.990601867123144e-06, + "loss": 0.8301, + "step": 499 + }, + { + "epoch": 0.06012144531954548, + "grad_norm": 2.3467560881497818, + "learning_rate": 3.990526288303014e-06, + "loss": 0.8754, + "step": 500 + }, + { + "epoch": 0.06024168821018457, + "grad_norm": 1.7600837784030852, + "learning_rate": 3.9904504075216295e-06, + "loss": 0.9239, + "step": 501 + }, + { + "epoch": 0.06036193110082366, + "grad_norm": 2.1486472287547005, + "learning_rate": 3.990374224790501e-06, + "loss": 0.9577, + "step": 502 + }, + { + "epoch": 0.06048217399146275, + "grad_norm": 1.9995457610235496, + "learning_rate": 3.990297740121185e-06, + "loss": 0.735, + "step": 503 + }, + { + "epoch": 0.06060241688210185, + "grad_norm": 1.837372693692639, + "learning_rate": 3.990220953525284e-06, + "loss": 0.7954, + "step": 504 + }, + { + "epoch": 0.06072265977274094, + "grad_norm": 2.497477307679545, + "learning_rate": 3.9901438650144465e-06, + "loss": 0.7672, + "step": 505 + }, + { + "epoch": 0.06084290266338003, + "grad_norm": 2.3013491488289373, + "learning_rate": 3.990066474600367e-06, + "loss": 0.9318, + "step": 506 + }, + { + "epoch": 0.06096314555401912, + "grad_norm": 1.8434413716258915, + "learning_rate": 3.989988782294786e-06, + "loss": 0.6957, + "step": 507 + }, + { + "epoch": 0.06108338844465821, + "grad_norm": 1.7221010312471536, + "learning_rate": 3.989910788109489e-06, + "loss": 0.9589, + "step": 508 + }, + { + "epoch": 0.0612036313352973, + "grad_norm": 2.0445756883350663, + "learning_rate": 3.989832492056307e-06, + "loss": 0.7661, + "step": 509 + }, + { + "epoch": 0.06132387422593639, + "grad_norm": 2.5148324606512586, + "learning_rate": 3.989753894147119e-06, + "loss": 0.8333, + "step": 510 + }, + { + "epoch": 0.061444117116575485, + "grad_norm": 1.747502861402622, + "learning_rate": 3.989674994393846e-06, + "loss": 0.8124, + "step": 511 + }, + { + "epoch": 0.061564360007214575, + "grad_norm": 2.3286556176856226, + "learning_rate": 3.98959579280846e-06, + "loss": 0.9522, + "step": 512 + }, + { + "epoch": 0.061684602897853665, + "grad_norm": 2.0226771277348057, + "learning_rate": 3.989516289402973e-06, + "loss": 0.8546, + "step": 513 + }, + { + "epoch": 0.061804845788492754, + "grad_norm": 2.412656306604069, + "learning_rate": 3.989436484189447e-06, + "loss": 0.8291, + "step": 514 + }, + { + "epoch": 0.061925088679131844, + "grad_norm": 2.6009073434860026, + "learning_rate": 3.9893563771799885e-06, + "loss": 0.8348, + "step": 515 + }, + { + "epoch": 0.062045331569770934, + "grad_norm": 2.486540995772486, + "learning_rate": 3.989275968386749e-06, + "loss": 0.8861, + "step": 516 + }, + { + "epoch": 0.06216557446041003, + "grad_norm": 1.9767277463811217, + "learning_rate": 3.989195257821926e-06, + "loss": 0.7796, + "step": 517 + }, + { + "epoch": 0.06228581735104912, + "grad_norm": 2.015823472771479, + "learning_rate": 3.989114245497765e-06, + "loss": 0.859, + "step": 518 + }, + { + "epoch": 0.06240606024168821, + "grad_norm": 2.158045808746071, + "learning_rate": 3.989032931426554e-06, + "loss": 0.9645, + "step": 519 + }, + { + "epoch": 0.06252630313232731, + "grad_norm": 2.1208051903237024, + "learning_rate": 3.9889513156206295e-06, + "loss": 0.8812, + "step": 520 + }, + { + "epoch": 0.06264654602296639, + "grad_norm": 2.4736054392978137, + "learning_rate": 3.988869398092371e-06, + "loss": 0.7664, + "step": 521 + }, + { + "epoch": 0.06276678891360549, + "grad_norm": 2.2739247851853897, + "learning_rate": 3.988787178854206e-06, + "loss": 0.8137, + "step": 522 + }, + { + "epoch": 0.06288703180424457, + "grad_norm": 2.4183829450436862, + "learning_rate": 3.988704657918608e-06, + "loss": 0.8957, + "step": 523 + }, + { + "epoch": 0.06300727469488367, + "grad_norm": 2.5537394126340263, + "learning_rate": 3.988621835298094e-06, + "loss": 0.8298, + "step": 524 + }, + { + "epoch": 0.06312751758552275, + "grad_norm": 1.8859549609294866, + "learning_rate": 3.988538711005229e-06, + "loss": 0.9367, + "step": 525 + }, + { + "epoch": 0.06324776047616185, + "grad_norm": 2.558894714825145, + "learning_rate": 3.988455285052622e-06, + "loss": 0.9038, + "step": 526 + }, + { + "epoch": 0.06336800336680094, + "grad_norm": 2.5484504741476437, + "learning_rate": 3.98837155745293e-06, + "loss": 0.8534, + "step": 527 + }, + { + "epoch": 0.06348824625744003, + "grad_norm": 2.3231325833225327, + "learning_rate": 3.988287528218854e-06, + "loss": 0.7865, + "step": 528 + }, + { + "epoch": 0.06360848914807912, + "grad_norm": 2.0120523416904206, + "learning_rate": 3.98820319736314e-06, + "loss": 0.911, + "step": 529 + }, + { + "epoch": 0.0637287320387182, + "grad_norm": 1.9312037021062616, + "learning_rate": 3.988118564898582e-06, + "loss": 0.8641, + "step": 530 + }, + { + "epoch": 0.0638489749293573, + "grad_norm": 2.592760931584717, + "learning_rate": 3.988033630838019e-06, + "loss": 0.9153, + "step": 531 + }, + { + "epoch": 0.0639692178199964, + "grad_norm": 1.7734134353147692, + "learning_rate": 3.987948395194334e-06, + "loss": 0.8959, + "step": 532 + }, + { + "epoch": 0.06408946071063548, + "grad_norm": 2.2940896720820367, + "learning_rate": 3.987862857980458e-06, + "loss": 0.7926, + "step": 533 + }, + { + "epoch": 0.06420970360127458, + "grad_norm": 2.05388251371079, + "learning_rate": 3.987777019209368e-06, + "loss": 0.7892, + "step": 534 + }, + { + "epoch": 0.06432994649191366, + "grad_norm": 3.853342631459705, + "learning_rate": 3.987690878894084e-06, + "loss": 0.8237, + "step": 535 + }, + { + "epoch": 0.06445018938255276, + "grad_norm": 2.3868046571845, + "learning_rate": 3.987604437047673e-06, + "loss": 0.8765, + "step": 536 + }, + { + "epoch": 0.06457043227319184, + "grad_norm": 1.9317661472187266, + "learning_rate": 3.987517693683251e-06, + "loss": 0.7914, + "step": 537 + }, + { + "epoch": 0.06469067516383094, + "grad_norm": 2.72506506605949, + "learning_rate": 3.9874306488139745e-06, + "loss": 0.9755, + "step": 538 + }, + { + "epoch": 0.06481091805447003, + "grad_norm": 1.9252800279531173, + "learning_rate": 3.987343302453049e-06, + "loss": 0.8863, + "step": 539 + }, + { + "epoch": 0.06493116094510912, + "grad_norm": 1.83957503314394, + "learning_rate": 3.987255654613724e-06, + "loss": 0.8381, + "step": 540 + }, + { + "epoch": 0.06505140383574821, + "grad_norm": 2.5697495791964386, + "learning_rate": 3.987167705309296e-06, + "loss": 0.7287, + "step": 541 + }, + { + "epoch": 0.0651716467263873, + "grad_norm": 1.985774830564598, + "learning_rate": 3.987079454553108e-06, + "loss": 0.9674, + "step": 542 + }, + { + "epoch": 0.0652918896170264, + "grad_norm": 1.7989398536402992, + "learning_rate": 3.986990902358546e-06, + "loss": 0.93, + "step": 543 + }, + { + "epoch": 0.06541213250766549, + "grad_norm": 2.0546032558329586, + "learning_rate": 3.986902048739045e-06, + "loss": 0.9506, + "step": 544 + }, + { + "epoch": 0.06553237539830457, + "grad_norm": 2.4649635926627957, + "learning_rate": 3.986812893708082e-06, + "loss": 0.8312, + "step": 545 + }, + { + "epoch": 0.06565261828894367, + "grad_norm": 1.9902636045755955, + "learning_rate": 3.9867234372791826e-06, + "loss": 0.8341, + "step": 546 + }, + { + "epoch": 0.06577286117958275, + "grad_norm": 1.5766283246486703, + "learning_rate": 3.986633679465918e-06, + "loss": 0.8864, + "step": 547 + }, + { + "epoch": 0.06589310407022185, + "grad_norm": 2.6758105645349057, + "learning_rate": 3.986543620281904e-06, + "loss": 0.8327, + "step": 548 + }, + { + "epoch": 0.06601334696086093, + "grad_norm": 1.6833836541664733, + "learning_rate": 3.986453259740802e-06, + "loss": 0.9262, + "step": 549 + }, + { + "epoch": 0.06613358985150003, + "grad_norm": 2.7403399990223294, + "learning_rate": 3.986362597856319e-06, + "loss": 0.808, + "step": 550 + }, + { + "epoch": 0.06625383274213913, + "grad_norm": 2.152499710774153, + "learning_rate": 3.986271634642211e-06, + "loss": 0.8407, + "step": 551 + }, + { + "epoch": 0.06637407563277821, + "grad_norm": 1.9537509983130972, + "learning_rate": 3.986180370112274e-06, + "loss": 0.835, + "step": 552 + }, + { + "epoch": 0.0664943185234173, + "grad_norm": 2.064548746455084, + "learning_rate": 3.986088804280354e-06, + "loss": 0.7633, + "step": 553 + }, + { + "epoch": 0.06661456141405639, + "grad_norm": 2.0678611205670068, + "learning_rate": 3.985996937160342e-06, + "loss": 0.9629, + "step": 554 + }, + { + "epoch": 0.06673480430469549, + "grad_norm": 2.027297649704662, + "learning_rate": 3.985904768766173e-06, + "loss": 0.7007, + "step": 555 + }, + { + "epoch": 0.06685504719533458, + "grad_norm": 2.1930274882126755, + "learning_rate": 3.98581229911183e-06, + "loss": 0.786, + "step": 556 + }, + { + "epoch": 0.06697529008597367, + "grad_norm": 1.7300466879570002, + "learning_rate": 3.985719528211341e-06, + "loss": 0.9332, + "step": 557 + }, + { + "epoch": 0.06709553297661276, + "grad_norm": 0.8460197648857632, + "learning_rate": 3.985626456078777e-06, + "loss": 0.6323, + "step": 558 + }, + { + "epoch": 0.06721577586725185, + "grad_norm": 2.327071700469755, + "learning_rate": 3.985533082728259e-06, + "loss": 0.8859, + "step": 559 + }, + { + "epoch": 0.06733601875789094, + "grad_norm": 1.7309592350567775, + "learning_rate": 3.985439408173951e-06, + "loss": 0.7631, + "step": 560 + }, + { + "epoch": 0.06745626164853002, + "grad_norm": 2.03276756524306, + "learning_rate": 3.9853454324300634e-06, + "loss": 0.7274, + "step": 561 + }, + { + "epoch": 0.06757650453916912, + "grad_norm": 2.0209771593436843, + "learning_rate": 3.985251155510852e-06, + "loss": 0.804, + "step": 562 + }, + { + "epoch": 0.06769674742980822, + "grad_norm": 1.859880715206094, + "learning_rate": 3.98515657743062e-06, + "loss": 0.824, + "step": 563 + }, + { + "epoch": 0.0678169903204473, + "grad_norm": 1.9485447392935176, + "learning_rate": 3.985061698203711e-06, + "loss": 0.7981, + "step": 564 + }, + { + "epoch": 0.0679372332110864, + "grad_norm": 0.8947495439863469, + "learning_rate": 3.984966517844523e-06, + "loss": 0.6404, + "step": 565 + }, + { + "epoch": 0.06805747610172548, + "grad_norm": 6.321775665483335, + "learning_rate": 3.984871036367492e-06, + "loss": 0.8314, + "step": 566 + }, + { + "epoch": 0.06817771899236458, + "grad_norm": 1.8695569187670553, + "learning_rate": 3.984775253787102e-06, + "loss": 0.8466, + "step": 567 + }, + { + "epoch": 0.06829796188300366, + "grad_norm": 3.1780668810366146, + "learning_rate": 3.984679170117885e-06, + "loss": 0.9087, + "step": 568 + }, + { + "epoch": 0.06841820477364276, + "grad_norm": 2.2350430856942487, + "learning_rate": 3.984582785374415e-06, + "loss": 0.796, + "step": 569 + }, + { + "epoch": 0.06853844766428185, + "grad_norm": 2.0201027981879, + "learning_rate": 3.9844860995713155e-06, + "loss": 0.8248, + "step": 570 + }, + { + "epoch": 0.06865869055492094, + "grad_norm": 2.588321474305656, + "learning_rate": 3.9843891127232524e-06, + "loss": 0.8476, + "step": 571 + }, + { + "epoch": 0.06877893344556003, + "grad_norm": 2.903682600167775, + "learning_rate": 3.984291824844938e-06, + "loss": 0.6861, + "step": 572 + }, + { + "epoch": 0.06889917633619912, + "grad_norm": 2.545669363915113, + "learning_rate": 3.984194235951132e-06, + "loss": 0.8783, + "step": 573 + }, + { + "epoch": 0.06901941922683821, + "grad_norm": 2.7779832149407735, + "learning_rate": 3.9840963460566375e-06, + "loss": 0.8707, + "step": 574 + }, + { + "epoch": 0.06913966211747731, + "grad_norm": 1.6028416433379076, + "learning_rate": 3.983998155176305e-06, + "loss": 0.9032, + "step": 575 + }, + { + "epoch": 0.06925990500811639, + "grad_norm": 0.8343230370617646, + "learning_rate": 3.9838996633250305e-06, + "loss": 0.5713, + "step": 576 + }, + { + "epoch": 0.06938014789875549, + "grad_norm": 2.5409371835269527, + "learning_rate": 3.983800870517753e-06, + "loss": 0.8929, + "step": 577 + }, + { + "epoch": 0.06950039078939457, + "grad_norm": 4.418962038124424, + "learning_rate": 3.983701776769463e-06, + "loss": 0.8126, + "step": 578 + }, + { + "epoch": 0.06962063368003367, + "grad_norm": 2.0103357987289394, + "learning_rate": 3.9836023820951885e-06, + "loss": 0.8647, + "step": 579 + }, + { + "epoch": 0.06974087657067275, + "grad_norm": 2.5974983798362516, + "learning_rate": 3.983502686510011e-06, + "loss": 0.6983, + "step": 580 + }, + { + "epoch": 0.06986111946131185, + "grad_norm": 3.023630535262126, + "learning_rate": 3.9834026900290525e-06, + "loss": 0.7524, + "step": 581 + }, + { + "epoch": 0.06998136235195095, + "grad_norm": 1.9246055000626965, + "learning_rate": 3.983302392667482e-06, + "loss": 1.0182, + "step": 582 + }, + { + "epoch": 0.07010160524259003, + "grad_norm": 1.848398716730111, + "learning_rate": 3.983201794440517e-06, + "loss": 0.9474, + "step": 583 + }, + { + "epoch": 0.07022184813322913, + "grad_norm": 1.8517352317364326, + "learning_rate": 3.9831008953634165e-06, + "loss": 0.6991, + "step": 584 + }, + { + "epoch": 0.07034209102386821, + "grad_norm": 1.7174546649174625, + "learning_rate": 3.9829996954514864e-06, + "loss": 0.8239, + "step": 585 + }, + { + "epoch": 0.0704623339145073, + "grad_norm": 2.1129215519609836, + "learning_rate": 3.982898194720079e-06, + "loss": 0.8539, + "step": 586 + }, + { + "epoch": 0.0705825768051464, + "grad_norm": 2.2468659508195294, + "learning_rate": 3.982796393184592e-06, + "loss": 0.8392, + "step": 587 + }, + { + "epoch": 0.07070281969578548, + "grad_norm": 0.8031552735793954, + "learning_rate": 3.98269429086047e-06, + "loss": 0.6273, + "step": 588 + }, + { + "epoch": 0.07082306258642458, + "grad_norm": 2.3846573488741116, + "learning_rate": 3.982591887763199e-06, + "loss": 0.8797, + "step": 589 + }, + { + "epoch": 0.07094330547706366, + "grad_norm": 5.923787877480321, + "learning_rate": 3.982489183908316e-06, + "loss": 0.8396, + "step": 590 + }, + { + "epoch": 0.07106354836770276, + "grad_norm": 1.8950840723517064, + "learning_rate": 3.982386179311399e-06, + "loss": 0.8615, + "step": 591 + }, + { + "epoch": 0.07118379125834184, + "grad_norm": 2.765594309835895, + "learning_rate": 3.982282873988075e-06, + "loss": 0.9081, + "step": 592 + }, + { + "epoch": 0.07130403414898094, + "grad_norm": 1.8390276935560312, + "learning_rate": 3.982179267954016e-06, + "loss": 0.8795, + "step": 593 + }, + { + "epoch": 0.07142427703962004, + "grad_norm": 2.07914020722804, + "learning_rate": 3.982075361224937e-06, + "loss": 0.9744, + "step": 594 + }, + { + "epoch": 0.07154451993025912, + "grad_norm": 1.9612205270725032, + "learning_rate": 3.981971153816602e-06, + "loss": 0.8992, + "step": 595 + }, + { + "epoch": 0.07166476282089822, + "grad_norm": 1.7524820165214823, + "learning_rate": 3.981866645744819e-06, + "loss": 0.9699, + "step": 596 + }, + { + "epoch": 0.0717850057115373, + "grad_norm": 3.9838201398273667, + "learning_rate": 3.9817618370254416e-06, + "loss": 0.8347, + "step": 597 + }, + { + "epoch": 0.0719052486021764, + "grad_norm": 2.115099338155566, + "learning_rate": 3.9816567276743684e-06, + "loss": 0.8908, + "step": 598 + }, + { + "epoch": 0.0720254914928155, + "grad_norm": 1.9931230928657717, + "learning_rate": 3.9815513177075466e-06, + "loss": 0.7962, + "step": 599 + }, + { + "epoch": 0.07214573438345458, + "grad_norm": 1.488048144730666, + "learning_rate": 3.9814456071409646e-06, + "loss": 0.7113, + "step": 600 + }, + { + "epoch": 0.07226597727409367, + "grad_norm": 2.4004051263928323, + "learning_rate": 3.981339595990659e-06, + "loss": 0.8881, + "step": 601 + }, + { + "epoch": 0.07238622016473276, + "grad_norm": 2.096735860696633, + "learning_rate": 3.981233284272713e-06, + "loss": 0.8351, + "step": 602 + }, + { + "epoch": 0.07250646305537185, + "grad_norm": 1.6824439563457996, + "learning_rate": 3.981126672003253e-06, + "loss": 0.9044, + "step": 603 + }, + { + "epoch": 0.07262670594601094, + "grad_norm": 2.413161717718188, + "learning_rate": 3.981019759198451e-06, + "loss": 0.794, + "step": 604 + }, + { + "epoch": 0.07274694883665003, + "grad_norm": 2.120418096736646, + "learning_rate": 3.980912545874528e-06, + "loss": 0.8526, + "step": 605 + }, + { + "epoch": 0.07286719172728913, + "grad_norm": 2.5550006826142297, + "learning_rate": 3.980805032047746e-06, + "loss": 0.8707, + "step": 606 + }, + { + "epoch": 0.07298743461792821, + "grad_norm": 2.0996943115087556, + "learning_rate": 3.980697217734415e-06, + "loss": 0.8213, + "step": 607 + }, + { + "epoch": 0.07310767750856731, + "grad_norm": 1.8004319520404604, + "learning_rate": 3.980589102950891e-06, + "loss": 0.9315, + "step": 608 + }, + { + "epoch": 0.07322792039920639, + "grad_norm": 2.4189770495874208, + "learning_rate": 3.9804806877135755e-06, + "loss": 0.7818, + "step": 609 + }, + { + "epoch": 0.07334816328984549, + "grad_norm": 3.124580202516102, + "learning_rate": 3.980371972038915e-06, + "loss": 0.883, + "step": 610 + }, + { + "epoch": 0.07346840618048459, + "grad_norm": 1.6697601061539564, + "learning_rate": 3.980262955943399e-06, + "loss": 0.8468, + "step": 611 + }, + { + "epoch": 0.07358864907112367, + "grad_norm": 3.1724087872029605, + "learning_rate": 3.980153639443569e-06, + "loss": 0.8824, + "step": 612 + }, + { + "epoch": 0.07370889196176277, + "grad_norm": 2.4629166652382795, + "learning_rate": 3.980044022556005e-06, + "loss": 0.8146, + "step": 613 + }, + { + "epoch": 0.07382913485240185, + "grad_norm": 1.9594575076702312, + "learning_rate": 3.9799341052973375e-06, + "loss": 0.744, + "step": 614 + }, + { + "epoch": 0.07394937774304094, + "grad_norm": 2.923379602475686, + "learning_rate": 3.979823887684241e-06, + "loss": 0.7849, + "step": 615 + }, + { + "epoch": 0.07406962063368003, + "grad_norm": 2.2105431086994662, + "learning_rate": 3.979713369733434e-06, + "loss": 0.8703, + "step": 616 + }, + { + "epoch": 0.07418986352431912, + "grad_norm": 2.0041627478082082, + "learning_rate": 3.979602551461683e-06, + "loss": 0.8632, + "step": 617 + }, + { + "epoch": 0.07431010641495822, + "grad_norm": 2.2604465629135855, + "learning_rate": 3.979491432885799e-06, + "loss": 0.9213, + "step": 618 + }, + { + "epoch": 0.0744303493055973, + "grad_norm": 1.9581003523954714, + "learning_rate": 3.97938001402264e-06, + "loss": 0.8407, + "step": 619 + }, + { + "epoch": 0.0745505921962364, + "grad_norm": 2.728491071018501, + "learning_rate": 3.979268294889105e-06, + "loss": 0.8217, + "step": 620 + }, + { + "epoch": 0.07467083508687548, + "grad_norm": 1.8364483061841466, + "learning_rate": 3.979156275502143e-06, + "loss": 0.7497, + "step": 621 + }, + { + "epoch": 0.07479107797751458, + "grad_norm": 2.3211671430500953, + "learning_rate": 3.979043955878749e-06, + "loss": 0.94, + "step": 622 + }, + { + "epoch": 0.07491132086815366, + "grad_norm": 1.9635767103014232, + "learning_rate": 3.978931336035959e-06, + "loss": 0.8426, + "step": 623 + }, + { + "epoch": 0.07503156375879276, + "grad_norm": 2.3398386552401753, + "learning_rate": 3.9788184159908595e-06, + "loss": 0.8364, + "step": 624 + }, + { + "epoch": 0.07515180664943186, + "grad_norm": 2.97758534396427, + "learning_rate": 3.97870519576058e-06, + "loss": 0.8478, + "step": 625 + }, + { + "epoch": 0.07527204954007094, + "grad_norm": 2.4737219660762197, + "learning_rate": 3.978591675362295e-06, + "loss": 0.8286, + "step": 626 + }, + { + "epoch": 0.07539229243071004, + "grad_norm": 1.7811893653588138, + "learning_rate": 3.978477854813226e-06, + "loss": 0.8922, + "step": 627 + }, + { + "epoch": 0.07551253532134912, + "grad_norm": 2.018946195817924, + "learning_rate": 3.97836373413064e-06, + "loss": 0.8387, + "step": 628 + }, + { + "epoch": 0.07563277821198822, + "grad_norm": 2.9971057918334156, + "learning_rate": 3.978249313331848e-06, + "loss": 0.7612, + "step": 629 + }, + { + "epoch": 0.07575302110262731, + "grad_norm": 2.6856647701394825, + "learning_rate": 3.978134592434208e-06, + "loss": 0.6397, + "step": 630 + }, + { + "epoch": 0.0758732639932664, + "grad_norm": 1.0029005661165855, + "learning_rate": 3.978019571455123e-06, + "loss": 0.5951, + "step": 631 + }, + { + "epoch": 0.07599350688390549, + "grad_norm": 2.162794271744229, + "learning_rate": 3.977904250412042e-06, + "loss": 0.8566, + "step": 632 + }, + { + "epoch": 0.07611374977454458, + "grad_norm": 2.190554312138252, + "learning_rate": 3.97778862932246e-06, + "loss": 0.8706, + "step": 633 + }, + { + "epoch": 0.07623399266518367, + "grad_norm": 2.181241155130986, + "learning_rate": 3.9776727082039144e-06, + "loss": 0.9512, + "step": 634 + }, + { + "epoch": 0.07635423555582276, + "grad_norm": 0.8184730978816739, + "learning_rate": 3.977556487073991e-06, + "loss": 0.5561, + "step": 635 + }, + { + "epoch": 0.07647447844646185, + "grad_norm": 1.7521048557745231, + "learning_rate": 3.97743996595032e-06, + "loss": 0.8266, + "step": 636 + }, + { + "epoch": 0.07659472133710095, + "grad_norm": 1.5408587357755164, + "learning_rate": 3.9773231448505804e-06, + "loss": 0.8289, + "step": 637 + }, + { + "epoch": 0.07671496422774003, + "grad_norm": 2.0130518375853277, + "learning_rate": 3.977206023792491e-06, + "loss": 0.793, + "step": 638 + }, + { + "epoch": 0.07683520711837913, + "grad_norm": 2.14986624634342, + "learning_rate": 3.97708860279382e-06, + "loss": 0.823, + "step": 639 + }, + { + "epoch": 0.07695545000901821, + "grad_norm": 1.665907575674502, + "learning_rate": 3.97697088187238e-06, + "loss": 0.794, + "step": 640 + }, + { + "epoch": 0.07707569289965731, + "grad_norm": 2.4790639669052625, + "learning_rate": 3.976852861046029e-06, + "loss": 0.9312, + "step": 641 + }, + { + "epoch": 0.0771959357902964, + "grad_norm": 1.6477350121199754, + "learning_rate": 3.97673454033267e-06, + "loss": 0.8046, + "step": 642 + }, + { + "epoch": 0.07731617868093549, + "grad_norm": 2.087119692577553, + "learning_rate": 3.976615919750254e-06, + "loss": 0.8354, + "step": 643 + }, + { + "epoch": 0.07743642157157458, + "grad_norm": 1.9676971042111382, + "learning_rate": 3.976496999316775e-06, + "loss": 0.8828, + "step": 644 + }, + { + "epoch": 0.07755666446221367, + "grad_norm": 2.1465738987672083, + "learning_rate": 3.976377779050271e-06, + "loss": 0.848, + "step": 645 + }, + { + "epoch": 0.07767690735285276, + "grad_norm": 2.0644157067032327, + "learning_rate": 3.976258258968831e-06, + "loss": 0.8678, + "step": 646 + }, + { + "epoch": 0.07779715024349185, + "grad_norm": 2.3408070716276534, + "learning_rate": 3.976138439090583e-06, + "loss": 0.7528, + "step": 647 + }, + { + "epoch": 0.07791739313413094, + "grad_norm": 2.4916662550025355, + "learning_rate": 3.976018319433706e-06, + "loss": 0.8717, + "step": 648 + }, + { + "epoch": 0.07803763602477004, + "grad_norm": 2.366363482688606, + "learning_rate": 3.9758979000164205e-06, + "loss": 0.9344, + "step": 649 + }, + { + "epoch": 0.07815787891540912, + "grad_norm": 2.197203124111086, + "learning_rate": 3.975777180856995e-06, + "loss": 0.7401, + "step": 650 + }, + { + "epoch": 0.07827812180604822, + "grad_norm": 4.33749532122181, + "learning_rate": 3.975656161973742e-06, + "loss": 0.885, + "step": 651 + }, + { + "epoch": 0.0783983646966873, + "grad_norm": 2.4669955816148175, + "learning_rate": 3.9755348433850194e-06, + "loss": 0.9064, + "step": 652 + }, + { + "epoch": 0.0785186075873264, + "grad_norm": 0.9744609956798889, + "learning_rate": 3.975413225109232e-06, + "loss": 0.6373, + "step": 653 + }, + { + "epoch": 0.0786388504779655, + "grad_norm": 2.762749766800515, + "learning_rate": 3.975291307164829e-06, + "loss": 0.9619, + "step": 654 + }, + { + "epoch": 0.07875909336860458, + "grad_norm": 1.929255757701785, + "learning_rate": 3.975169089570306e-06, + "loss": 0.8683, + "step": 655 + }, + { + "epoch": 0.07887933625924368, + "grad_norm": 2.2332509020875384, + "learning_rate": 3.975046572344202e-06, + "loss": 0.9321, + "step": 656 + }, + { + "epoch": 0.07899957914988276, + "grad_norm": 1.6246606032193542, + "learning_rate": 3.974923755505103e-06, + "loss": 0.7243, + "step": 657 + }, + { + "epoch": 0.07911982204052186, + "grad_norm": 1.65459748366386, + "learning_rate": 3.974800639071641e-06, + "loss": 0.9214, + "step": 658 + }, + { + "epoch": 0.07924006493116094, + "grad_norm": 2.061642208319206, + "learning_rate": 3.974677223062492e-06, + "loss": 1.0214, + "step": 659 + }, + { + "epoch": 0.07936030782180004, + "grad_norm": 2.2353003659766295, + "learning_rate": 3.974553507496378e-06, + "loss": 0.7625, + "step": 660 + }, + { + "epoch": 0.07948055071243913, + "grad_norm": 2.7527877375883074, + "learning_rate": 3.974429492392068e-06, + "loss": 0.9124, + "step": 661 + }, + { + "epoch": 0.07960079360307822, + "grad_norm": 1.890864081280554, + "learning_rate": 3.974305177768373e-06, + "loss": 0.9222, + "step": 662 + }, + { + "epoch": 0.07972103649371731, + "grad_norm": 2.0991995194082755, + "learning_rate": 3.974180563644152e-06, + "loss": 0.8853, + "step": 663 + }, + { + "epoch": 0.0798412793843564, + "grad_norm": 2.165736353269663, + "learning_rate": 3.97405565003831e-06, + "loss": 0.9047, + "step": 664 + }, + { + "epoch": 0.07996152227499549, + "grad_norm": 2.0233944921451936, + "learning_rate": 3.973930436969794e-06, + "loss": 0.8099, + "step": 665 + }, + { + "epoch": 0.08008176516563459, + "grad_norm": 1.8638206996594726, + "learning_rate": 3.973804924457602e-06, + "loss": 0.8753, + "step": 666 + }, + { + "epoch": 0.08020200805627367, + "grad_norm": 1.7173276391087977, + "learning_rate": 3.973679112520771e-06, + "loss": 0.866, + "step": 667 + }, + { + "epoch": 0.08032225094691277, + "grad_norm": 1.8686028387501141, + "learning_rate": 3.973553001178389e-06, + "loss": 1.0068, + "step": 668 + }, + { + "epoch": 0.08044249383755185, + "grad_norm": 2.0399962877725972, + "learning_rate": 3.973426590449585e-06, + "loss": 0.7665, + "step": 669 + }, + { + "epoch": 0.08056273672819095, + "grad_norm": 1.8614316193026925, + "learning_rate": 3.9732998803535364e-06, + "loss": 0.7687, + "step": 670 + }, + { + "epoch": 0.08068297961883003, + "grad_norm": 2.2352230356712854, + "learning_rate": 3.973172870909465e-06, + "loss": 0.8647, + "step": 671 + }, + { + "epoch": 0.08080322250946913, + "grad_norm": 5.380531481048567, + "learning_rate": 3.973045562136638e-06, + "loss": 0.8253, + "step": 672 + }, + { + "epoch": 0.08092346540010822, + "grad_norm": 2.1311466852020433, + "learning_rate": 3.972917954054368e-06, + "loss": 0.9225, + "step": 673 + }, + { + "epoch": 0.08104370829074731, + "grad_norm": 2.173672817059872, + "learning_rate": 3.972790046682013e-06, + "loss": 0.8357, + "step": 674 + }, + { + "epoch": 0.0811639511813864, + "grad_norm": 1.9184838125874704, + "learning_rate": 3.972661840038977e-06, + "loss": 0.8031, + "step": 675 + }, + { + "epoch": 0.08128419407202549, + "grad_norm": 2.1331576672390584, + "learning_rate": 3.972533334144707e-06, + "loss": 0.8452, + "step": 676 + }, + { + "epoch": 0.08140443696266458, + "grad_norm": 2.0249691619408736, + "learning_rate": 3.972404529018699e-06, + "loss": 0.805, + "step": 677 + }, + { + "epoch": 0.08152467985330367, + "grad_norm": 1.9059130886833167, + "learning_rate": 3.972275424680493e-06, + "loss": 0.8693, + "step": 678 + }, + { + "epoch": 0.08164492274394276, + "grad_norm": 1.9952337573592065, + "learning_rate": 3.972146021149673e-06, + "loss": 0.926, + "step": 679 + }, + { + "epoch": 0.08176516563458186, + "grad_norm": 2.1758652949922945, + "learning_rate": 3.972016318445868e-06, + "loss": 0.7925, + "step": 680 + }, + { + "epoch": 0.08188540852522094, + "grad_norm": 1.8662603781961544, + "learning_rate": 3.971886316588757e-06, + "loss": 0.9402, + "step": 681 + }, + { + "epoch": 0.08200565141586004, + "grad_norm": 2.3075208930395332, + "learning_rate": 3.9717560155980595e-06, + "loss": 0.7568, + "step": 682 + }, + { + "epoch": 0.08212589430649912, + "grad_norm": 1.922395775180989, + "learning_rate": 3.971625415493542e-06, + "loss": 0.9314, + "step": 683 + }, + { + "epoch": 0.08224613719713822, + "grad_norm": 1.9541193147736553, + "learning_rate": 3.971494516295017e-06, + "loss": 0.8798, + "step": 684 + }, + { + "epoch": 0.08236638008777732, + "grad_norm": 1.7919913226964874, + "learning_rate": 3.971363318022341e-06, + "loss": 0.8604, + "step": 685 + }, + { + "epoch": 0.0824866229784164, + "grad_norm": 1.9219214581092399, + "learning_rate": 3.971231820695417e-06, + "loss": 0.7095, + "step": 686 + }, + { + "epoch": 0.0826068658690555, + "grad_norm": 2.0156171931625706, + "learning_rate": 3.971100024334193e-06, + "loss": 0.8232, + "step": 687 + }, + { + "epoch": 0.08272710875969458, + "grad_norm": 1.956432989659954, + "learning_rate": 3.970967928958663e-06, + "loss": 0.8795, + "step": 688 + }, + { + "epoch": 0.08284735165033368, + "grad_norm": 1.6712532669982705, + "learning_rate": 3.970835534588865e-06, + "loss": 0.8391, + "step": 689 + }, + { + "epoch": 0.08296759454097276, + "grad_norm": 1.7780170213466018, + "learning_rate": 3.970702841244883e-06, + "loss": 0.8697, + "step": 690 + }, + { + "epoch": 0.08308783743161186, + "grad_norm": 1.6964873712431088, + "learning_rate": 3.970569848946847e-06, + "loss": 0.8428, + "step": 691 + }, + { + "epoch": 0.08320808032225095, + "grad_norm": 2.110076674564288, + "learning_rate": 3.970436557714932e-06, + "loss": 0.8443, + "step": 692 + }, + { + "epoch": 0.08332832321289003, + "grad_norm": 2.0764197904001915, + "learning_rate": 3.970302967569358e-06, + "loss": 0.8875, + "step": 693 + }, + { + "epoch": 0.08344856610352913, + "grad_norm": 2.0192524996136676, + "learning_rate": 3.9701690785303896e-06, + "loss": 0.7033, + "step": 694 + }, + { + "epoch": 0.08356880899416821, + "grad_norm": 2.3313581303532307, + "learning_rate": 3.970034890618339e-06, + "loss": 0.909, + "step": 695 + }, + { + "epoch": 0.08368905188480731, + "grad_norm": 2.0080270615460147, + "learning_rate": 3.969900403853562e-06, + "loss": 0.8957, + "step": 696 + }, + { + "epoch": 0.08380929477544641, + "grad_norm": 1.7562476080067049, + "learning_rate": 3.96976561825646e-06, + "loss": 0.7929, + "step": 697 + }, + { + "epoch": 0.08392953766608549, + "grad_norm": 1.8990464980151316, + "learning_rate": 3.969630533847479e-06, + "loss": 0.8924, + "step": 698 + }, + { + "epoch": 0.08404978055672459, + "grad_norm": 1.9408614010035012, + "learning_rate": 3.969495150647113e-06, + "loss": 0.8581, + "step": 699 + }, + { + "epoch": 0.08417002344736367, + "grad_norm": 1.8640220337126217, + "learning_rate": 3.969359468675899e-06, + "loss": 0.7779, + "step": 700 + }, + { + "epoch": 0.08429026633800277, + "grad_norm": 1.8997041848440352, + "learning_rate": 3.969223487954418e-06, + "loss": 0.9049, + "step": 701 + }, + { + "epoch": 0.08441050922864185, + "grad_norm": 2.2726734883273263, + "learning_rate": 3.969087208503301e-06, + "loss": 0.841, + "step": 702 + }, + { + "epoch": 0.08453075211928095, + "grad_norm": 2.040857219480975, + "learning_rate": 3.968950630343219e-06, + "loss": 0.8641, + "step": 703 + }, + { + "epoch": 0.08465099500992004, + "grad_norm": 2.01852740309481, + "learning_rate": 3.968813753494892e-06, + "loss": 0.9486, + "step": 704 + }, + { + "epoch": 0.08477123790055913, + "grad_norm": 2.1387903746779826, + "learning_rate": 3.968676577979084e-06, + "loss": 0.7656, + "step": 705 + }, + { + "epoch": 0.08489148079119822, + "grad_norm": 2.1335494895543063, + "learning_rate": 3.968539103816605e-06, + "loss": 0.795, + "step": 706 + }, + { + "epoch": 0.0850117236818373, + "grad_norm": 1.894508471794446, + "learning_rate": 3.9684013310283085e-06, + "loss": 0.9115, + "step": 707 + }, + { + "epoch": 0.0851319665724764, + "grad_norm": 2.0178698673160564, + "learning_rate": 3.9682632596350956e-06, + "loss": 0.6599, + "step": 708 + }, + { + "epoch": 0.0852522094631155, + "grad_norm": 2.068591514578703, + "learning_rate": 3.968124889657911e-06, + "loss": 0.7962, + "step": 709 + }, + { + "epoch": 0.08537245235375458, + "grad_norm": 2.473291396876297, + "learning_rate": 3.967986221117746e-06, + "loss": 0.9202, + "step": 710 + }, + { + "epoch": 0.08549269524439368, + "grad_norm": 1.8736755779773586, + "learning_rate": 3.967847254035635e-06, + "loss": 0.8754, + "step": 711 + }, + { + "epoch": 0.08561293813503276, + "grad_norm": 2.1400560054402544, + "learning_rate": 3.967707988432661e-06, + "loss": 0.884, + "step": 712 + }, + { + "epoch": 0.08573318102567186, + "grad_norm": 2.0983103721439016, + "learning_rate": 3.967568424329949e-06, + "loss": 0.8858, + "step": 713 + }, + { + "epoch": 0.08585342391631094, + "grad_norm": 0.8246277070277874, + "learning_rate": 3.967428561748671e-06, + "loss": 0.5567, + "step": 714 + }, + { + "epoch": 0.08597366680695004, + "grad_norm": 1.8102915232974246, + "learning_rate": 3.967288400710045e-06, + "loss": 0.8869, + "step": 715 + }, + { + "epoch": 0.08609390969758914, + "grad_norm": 2.301112411836758, + "learning_rate": 3.9671479412353335e-06, + "loss": 0.8951, + "step": 716 + }, + { + "epoch": 0.08621415258822822, + "grad_norm": 1.893297182843913, + "learning_rate": 3.967007183345843e-06, + "loss": 0.7509, + "step": 717 + }, + { + "epoch": 0.08633439547886732, + "grad_norm": 2.3317901640065126, + "learning_rate": 3.966866127062927e-06, + "loss": 0.9179, + "step": 718 + }, + { + "epoch": 0.0864546383695064, + "grad_norm": 0.8768113665286893, + "learning_rate": 3.966724772407982e-06, + "loss": 0.6281, + "step": 719 + }, + { + "epoch": 0.0865748812601455, + "grad_norm": 2.0787399644087587, + "learning_rate": 3.966583119402454e-06, + "loss": 0.8957, + "step": 720 + }, + { + "epoch": 0.08669512415078459, + "grad_norm": 2.026901990797249, + "learning_rate": 3.9664411680678305e-06, + "loss": 0.8344, + "step": 721 + }, + { + "epoch": 0.08681536704142367, + "grad_norm": 0.8554796888792988, + "learning_rate": 3.966298918425644e-06, + "loss": 0.6159, + "step": 722 + }, + { + "epoch": 0.08693560993206277, + "grad_norm": 2.120649484436276, + "learning_rate": 3.966156370497476e-06, + "loss": 0.8325, + "step": 723 + }, + { + "epoch": 0.08705585282270185, + "grad_norm": 1.8616275766629302, + "learning_rate": 3.96601352430495e-06, + "loss": 0.8992, + "step": 724 + }, + { + "epoch": 0.08717609571334095, + "grad_norm": 1.586891977009777, + "learning_rate": 3.965870379869735e-06, + "loss": 0.8402, + "step": 725 + }, + { + "epoch": 0.08729633860398003, + "grad_norm": 2.0358864321858183, + "learning_rate": 3.965726937213547e-06, + "loss": 0.8781, + "step": 726 + }, + { + "epoch": 0.08741658149461913, + "grad_norm": 3.7525642528789382, + "learning_rate": 3.965583196358144e-06, + "loss": 0.8256, + "step": 727 + }, + { + "epoch": 0.08753682438525823, + "grad_norm": 2.540112463195989, + "learning_rate": 3.965439157325335e-06, + "loss": 0.763, + "step": 728 + }, + { + "epoch": 0.08765706727589731, + "grad_norm": 3.010348432530975, + "learning_rate": 3.965294820136968e-06, + "loss": 0.7806, + "step": 729 + }, + { + "epoch": 0.08777731016653641, + "grad_norm": 1.8822352367302977, + "learning_rate": 3.965150184814938e-06, + "loss": 0.8816, + "step": 730 + }, + { + "epoch": 0.08789755305717549, + "grad_norm": 2.31447826001647, + "learning_rate": 3.965005251381189e-06, + "loss": 0.7784, + "step": 731 + }, + { + "epoch": 0.08801779594781459, + "grad_norm": 0.9764400558003605, + "learning_rate": 3.964860019857705e-06, + "loss": 0.6477, + "step": 732 + }, + { + "epoch": 0.08813803883845367, + "grad_norm": 2.1052438477835222, + "learning_rate": 3.964714490266518e-06, + "loss": 0.8507, + "step": 733 + }, + { + "epoch": 0.08825828172909277, + "grad_norm": 0.9036339127838094, + "learning_rate": 3.964568662629706e-06, + "loss": 0.6475, + "step": 734 + }, + { + "epoch": 0.08837852461973186, + "grad_norm": 2.7922112991693493, + "learning_rate": 3.9644225369693895e-06, + "loss": 0.8681, + "step": 735 + }, + { + "epoch": 0.08849876751037095, + "grad_norm": 1.9902619873766887, + "learning_rate": 3.964276113307735e-06, + "loss": 0.8789, + "step": 736 + }, + { + "epoch": 0.08861901040101004, + "grad_norm": 1.8634070268001235, + "learning_rate": 3.9641293916669574e-06, + "loss": 0.8253, + "step": 737 + }, + { + "epoch": 0.08873925329164913, + "grad_norm": 4.024581242925555, + "learning_rate": 3.9639823720693115e-06, + "loss": 0.8366, + "step": 738 + }, + { + "epoch": 0.08885949618228822, + "grad_norm": 0.8616049531447161, + "learning_rate": 3.963835054537102e-06, + "loss": 0.6021, + "step": 739 + }, + { + "epoch": 0.08897973907292732, + "grad_norm": 2.23188376094623, + "learning_rate": 3.963687439092676e-06, + "loss": 0.6273, + "step": 740 + }, + { + "epoch": 0.0890999819635664, + "grad_norm": 1.9215663258850553, + "learning_rate": 3.963539525758427e-06, + "loss": 0.8202, + "step": 741 + }, + { + "epoch": 0.0892202248542055, + "grad_norm": 1.9164435077067106, + "learning_rate": 3.9633913145567925e-06, + "loss": 0.6946, + "step": 742 + }, + { + "epoch": 0.08934046774484458, + "grad_norm": 1.8264293142404915, + "learning_rate": 3.9632428055102575e-06, + "loss": 0.8218, + "step": 743 + }, + { + "epoch": 0.08946071063548368, + "grad_norm": 2.1181705909160575, + "learning_rate": 3.9630939986413495e-06, + "loss": 0.684, + "step": 744 + }, + { + "epoch": 0.08958095352612276, + "grad_norm": 1.9347211537293916, + "learning_rate": 3.962944893972643e-06, + "loss": 0.7871, + "step": 745 + }, + { + "epoch": 0.08970119641676186, + "grad_norm": 2.502623992957826, + "learning_rate": 3.962795491526756e-06, + "loss": 0.9334, + "step": 746 + }, + { + "epoch": 0.08982143930740095, + "grad_norm": 1.9468854802861413, + "learning_rate": 3.962645791326354e-06, + "loss": 0.9089, + "step": 747 + }, + { + "epoch": 0.08994168219804004, + "grad_norm": 1.8621775559238107, + "learning_rate": 3.962495793394146e-06, + "loss": 0.8415, + "step": 748 + }, + { + "epoch": 0.09006192508867913, + "grad_norm": 0.7576179254601532, + "learning_rate": 3.9623454977528864e-06, + "loss": 0.6129, + "step": 749 + }, + { + "epoch": 0.09018216797931822, + "grad_norm": 1.8881116191916434, + "learning_rate": 3.962194904425375e-06, + "loss": 0.8654, + "step": 750 + }, + { + "epoch": 0.09030241086995731, + "grad_norm": 2.272632784179731, + "learning_rate": 3.9620440134344566e-06, + "loss": 0.6931, + "step": 751 + }, + { + "epoch": 0.09042265376059641, + "grad_norm": 2.5751934608449925, + "learning_rate": 3.9618928248030215e-06, + "loss": 0.8418, + "step": 752 + }, + { + "epoch": 0.0905428966512355, + "grad_norm": 2.0527557459044536, + "learning_rate": 3.961741338554005e-06, + "loss": 0.8441, + "step": 753 + }, + { + "epoch": 0.09066313954187459, + "grad_norm": 1.7773834559438464, + "learning_rate": 3.9615895547103865e-06, + "loss": 0.7665, + "step": 754 + }, + { + "epoch": 0.09078338243251367, + "grad_norm": 2.230277619465275, + "learning_rate": 3.961437473295193e-06, + "loss": 0.7895, + "step": 755 + }, + { + "epoch": 0.09090362532315277, + "grad_norm": 2.5146244499766097, + "learning_rate": 3.961285094331495e-06, + "loss": 0.741, + "step": 756 + }, + { + "epoch": 0.09102386821379185, + "grad_norm": 1.7328228577836569, + "learning_rate": 3.961132417842406e-06, + "loss": 0.8668, + "step": 757 + }, + { + "epoch": 0.09114411110443095, + "grad_norm": 3.2843624793484585, + "learning_rate": 3.960979443851089e-06, + "loss": 0.7643, + "step": 758 + }, + { + "epoch": 0.09126435399507005, + "grad_norm": 1.6893479154605262, + "learning_rate": 3.96082617238075e-06, + "loss": 0.7918, + "step": 759 + }, + { + "epoch": 0.09138459688570913, + "grad_norm": 2.062563692901184, + "learning_rate": 3.960672603454639e-06, + "loss": 0.8079, + "step": 760 + }, + { + "epoch": 0.09150483977634823, + "grad_norm": 2.8042025771163286, + "learning_rate": 3.960518737096054e-06, + "loss": 0.7864, + "step": 761 + }, + { + "epoch": 0.09162508266698731, + "grad_norm": 2.1268962926076243, + "learning_rate": 3.960364573328334e-06, + "loss": 0.7512, + "step": 762 + }, + { + "epoch": 0.0917453255576264, + "grad_norm": 1.954829517453253, + "learning_rate": 3.9602101121748675e-06, + "loss": 0.9018, + "step": 763 + }, + { + "epoch": 0.0918655684482655, + "grad_norm": 1.9556603417474967, + "learning_rate": 3.960055353659085e-06, + "loss": 0.7453, + "step": 764 + }, + { + "epoch": 0.09198581133890459, + "grad_norm": 2.0425380741525463, + "learning_rate": 3.959900297804465e-06, + "loss": 0.8428, + "step": 765 + }, + { + "epoch": 0.09210605422954368, + "grad_norm": 4.88722534968998, + "learning_rate": 3.9597449446345276e-06, + "loss": 0.7854, + "step": 766 + }, + { + "epoch": 0.09222629712018277, + "grad_norm": 2.2411014704680015, + "learning_rate": 3.95958929417284e-06, + "loss": 0.8487, + "step": 767 + }, + { + "epoch": 0.09234654001082186, + "grad_norm": 0.7480713763096423, + "learning_rate": 3.9594333464430145e-06, + "loss": 0.5892, + "step": 768 + }, + { + "epoch": 0.09246678290146094, + "grad_norm": 2.004363426125122, + "learning_rate": 3.959277101468709e-06, + "loss": 0.9004, + "step": 769 + }, + { + "epoch": 0.09258702579210004, + "grad_norm": 3.250192343081992, + "learning_rate": 3.959120559273624e-06, + "loss": 0.7963, + "step": 770 + }, + { + "epoch": 0.09270726868273914, + "grad_norm": 2.1275770605096733, + "learning_rate": 3.958963719881509e-06, + "loss": 0.8471, + "step": 771 + }, + { + "epoch": 0.09282751157337822, + "grad_norm": 2.015780657295132, + "learning_rate": 3.958806583316154e-06, + "loss": 0.952, + "step": 772 + }, + { + "epoch": 0.09294775446401732, + "grad_norm": 2.2247417300055914, + "learning_rate": 3.9586491496013985e-06, + "loss": 0.8031, + "step": 773 + }, + { + "epoch": 0.0930679973546564, + "grad_norm": 2.275620922827794, + "learning_rate": 3.958491418761124e-06, + "loss": 0.8463, + "step": 774 + }, + { + "epoch": 0.0931882402452955, + "grad_norm": 2.1477364800913166, + "learning_rate": 3.958333390819258e-06, + "loss": 0.7445, + "step": 775 + }, + { + "epoch": 0.0933084831359346, + "grad_norm": 2.5370905804249975, + "learning_rate": 3.9581750657997754e-06, + "loss": 0.8149, + "step": 776 + }, + { + "epoch": 0.09342872602657368, + "grad_norm": 1.753980877144593, + "learning_rate": 3.95801644372669e-06, + "loss": 0.9017, + "step": 777 + }, + { + "epoch": 0.09354896891721277, + "grad_norm": 2.1067752217238045, + "learning_rate": 3.957857524624068e-06, + "loss": 0.8556, + "step": 778 + }, + { + "epoch": 0.09366921180785186, + "grad_norm": 1.5759864140229949, + "learning_rate": 3.957698308516016e-06, + "loss": 0.9055, + "step": 779 + }, + { + "epoch": 0.09378945469849095, + "grad_norm": 2.2936043841215983, + "learning_rate": 3.957538795426688e-06, + "loss": 0.8366, + "step": 780 + }, + { + "epoch": 0.09390969758913004, + "grad_norm": 2.3673674763665145, + "learning_rate": 3.9573789853802804e-06, + "loss": 0.7922, + "step": 781 + }, + { + "epoch": 0.09402994047976913, + "grad_norm": 2.1925204839151893, + "learning_rate": 3.957218878401037e-06, + "loss": 0.763, + "step": 782 + }, + { + "epoch": 0.09415018337040823, + "grad_norm": 2.383537607397648, + "learning_rate": 3.957058474513246e-06, + "loss": 0.9062, + "step": 783 + }, + { + "epoch": 0.09427042626104731, + "grad_norm": 1.8354895511954317, + "learning_rate": 3.956897773741241e-06, + "loss": 0.7994, + "step": 784 + }, + { + "epoch": 0.09439066915168641, + "grad_norm": 1.6266648493949454, + "learning_rate": 3.956736776109398e-06, + "loss": 0.7263, + "step": 785 + }, + { + "epoch": 0.09451091204232549, + "grad_norm": 1.7938391872656527, + "learning_rate": 3.956575481642143e-06, + "loss": 0.8458, + "step": 786 + }, + { + "epoch": 0.09463115493296459, + "grad_norm": 2.5388800452774176, + "learning_rate": 3.956413890363943e-06, + "loss": 0.763, + "step": 787 + }, + { + "epoch": 0.09475139782360369, + "grad_norm": 2.1698391380983497, + "learning_rate": 3.956252002299312e-06, + "loss": 0.8298, + "step": 788 + }, + { + "epoch": 0.09487164071424277, + "grad_norm": 1.9842965418406326, + "learning_rate": 3.956089817472807e-06, + "loss": 0.924, + "step": 789 + }, + { + "epoch": 0.09499188360488187, + "grad_norm": 2.2790446660818953, + "learning_rate": 3.955927335909032e-06, + "loss": 0.8646, + "step": 790 + }, + { + "epoch": 0.09511212649552095, + "grad_norm": 1.968953924365528, + "learning_rate": 3.955764557632634e-06, + "loss": 0.7657, + "step": 791 + }, + { + "epoch": 0.09523236938616005, + "grad_norm": 2.423883405204121, + "learning_rate": 3.955601482668309e-06, + "loss": 0.9586, + "step": 792 + }, + { + "epoch": 0.09535261227679913, + "grad_norm": 1.7548537381947404, + "learning_rate": 3.955438111040794e-06, + "loss": 0.897, + "step": 793 + }, + { + "epoch": 0.09547285516743823, + "grad_norm": 1.8747150840453186, + "learning_rate": 3.955274442774873e-06, + "loss": 0.8134, + "step": 794 + }, + { + "epoch": 0.09559309805807732, + "grad_norm": 2.3260913133483534, + "learning_rate": 3.9551104778953725e-06, + "loss": 0.7324, + "step": 795 + }, + { + "epoch": 0.0957133409487164, + "grad_norm": 2.3077913752101034, + "learning_rate": 3.954946216427167e-06, + "loss": 0.8687, + "step": 796 + }, + { + "epoch": 0.0958335838393555, + "grad_norm": 0.8924507367405203, + "learning_rate": 3.954781658395176e-06, + "loss": 0.6181, + "step": 797 + }, + { + "epoch": 0.09595382672999458, + "grad_norm": 2.156493131779206, + "learning_rate": 3.95461680382436e-06, + "loss": 0.939, + "step": 798 + }, + { + "epoch": 0.09607406962063368, + "grad_norm": 2.6777758661195605, + "learning_rate": 3.9544516527397295e-06, + "loss": 0.8713, + "step": 799 + }, + { + "epoch": 0.09619431251127276, + "grad_norm": 1.8096581904508566, + "learning_rate": 3.954286205166338e-06, + "loss": 0.8118, + "step": 800 + }, + { + "epoch": 0.09631455540191186, + "grad_norm": 1.9708029321929383, + "learning_rate": 3.954120461129282e-06, + "loss": 0.8465, + "step": 801 + }, + { + "epoch": 0.09643479829255096, + "grad_norm": 1.9018517705957305, + "learning_rate": 3.953954420653706e-06, + "loss": 0.8439, + "step": 802 + }, + { + "epoch": 0.09655504118319004, + "grad_norm": 2.2052365187866547, + "learning_rate": 3.953788083764798e-06, + "loss": 0.8904, + "step": 803 + }, + { + "epoch": 0.09667528407382914, + "grad_norm": 2.037804928294475, + "learning_rate": 3.953621450487792e-06, + "loss": 0.9325, + "step": 804 + }, + { + "epoch": 0.09679552696446822, + "grad_norm": 0.8590840280554511, + "learning_rate": 3.953454520847964e-06, + "loss": 0.6126, + "step": 805 + }, + { + "epoch": 0.09691576985510732, + "grad_norm": 2.1848043999316493, + "learning_rate": 3.9532872948706395e-06, + "loss": 0.7492, + "step": 806 + }, + { + "epoch": 0.09703601274574641, + "grad_norm": 2.2551489301673864, + "learning_rate": 3.9531197725811845e-06, + "loss": 0.8405, + "step": 807 + }, + { + "epoch": 0.0971562556363855, + "grad_norm": 1.8335726827379084, + "learning_rate": 3.952951954005013e-06, + "loss": 0.8857, + "step": 808 + }, + { + "epoch": 0.0972764985270246, + "grad_norm": 1.6689354554181557, + "learning_rate": 3.952783839167584e-06, + "loss": 0.8531, + "step": 809 + }, + { + "epoch": 0.09739674141766368, + "grad_norm": 2.9770326418306037, + "learning_rate": 3.952615428094398e-06, + "loss": 0.766, + "step": 810 + }, + { + "epoch": 0.09751698430830277, + "grad_norm": 1.7396725690724124, + "learning_rate": 3.952446720811004e-06, + "loss": 0.7454, + "step": 811 + }, + { + "epoch": 0.09763722719894186, + "grad_norm": 0.8312814405418193, + "learning_rate": 3.952277717342995e-06, + "loss": 0.6368, + "step": 812 + }, + { + "epoch": 0.09775747008958095, + "grad_norm": 1.7810395099684695, + "learning_rate": 3.952108417716009e-06, + "loss": 0.8624, + "step": 813 + }, + { + "epoch": 0.09787771298022005, + "grad_norm": 1.8307695090423748, + "learning_rate": 3.951938821955727e-06, + "loss": 0.8606, + "step": 814 + }, + { + "epoch": 0.09799795587085913, + "grad_norm": 1.5597195109050068, + "learning_rate": 3.9517689300878786e-06, + "loss": 0.7729, + "step": 815 + }, + { + "epoch": 0.09811819876149823, + "grad_norm": 1.832318899513611, + "learning_rate": 3.951598742138236e-06, + "loss": 0.8024, + "step": 816 + }, + { + "epoch": 0.09823844165213731, + "grad_norm": 1.9491759393000396, + "learning_rate": 3.951428258132615e-06, + "loss": 0.8112, + "step": 817 + }, + { + "epoch": 0.09835868454277641, + "grad_norm": 1.8105561586407417, + "learning_rate": 3.951257478096879e-06, + "loss": 0.8577, + "step": 818 + }, + { + "epoch": 0.0984789274334155, + "grad_norm": 2.4540411337939183, + "learning_rate": 3.951086402056936e-06, + "loss": 0.6901, + "step": 819 + }, + { + "epoch": 0.09859917032405459, + "grad_norm": 1.655428988819492, + "learning_rate": 3.950915030038735e-06, + "loss": 0.8421, + "step": 820 + }, + { + "epoch": 0.09871941321469369, + "grad_norm": 2.440932571236202, + "learning_rate": 3.9507433620682765e-06, + "loss": 0.854, + "step": 821 + }, + { + "epoch": 0.09883965610533277, + "grad_norm": 1.5737404676516824, + "learning_rate": 3.9505713981716e-06, + "loss": 0.8845, + "step": 822 + }, + { + "epoch": 0.09895989899597187, + "grad_norm": 1.8323790783297105, + "learning_rate": 3.950399138374795e-06, + "loss": 0.8179, + "step": 823 + }, + { + "epoch": 0.09908014188661095, + "grad_norm": 1.5661087324363383, + "learning_rate": 3.95022658270399e-06, + "loss": 0.753, + "step": 824 + }, + { + "epoch": 0.09920038477725004, + "grad_norm": 1.8600802326330645, + "learning_rate": 3.9500537311853635e-06, + "loss": 0.7944, + "step": 825 + }, + { + "epoch": 0.09932062766788914, + "grad_norm": 2.140142121809715, + "learning_rate": 3.949880583845136e-06, + "loss": 0.8506, + "step": 826 + }, + { + "epoch": 0.09944087055852822, + "grad_norm": 1.7670051784212821, + "learning_rate": 3.949707140709575e-06, + "loss": 0.8257, + "step": 827 + }, + { + "epoch": 0.09956111344916732, + "grad_norm": 2.16710088731428, + "learning_rate": 3.949533401804991e-06, + "loss": 0.8524, + "step": 828 + }, + { + "epoch": 0.0996813563398064, + "grad_norm": 2.5693143164414787, + "learning_rate": 3.949359367157739e-06, + "loss": 0.9234, + "step": 829 + }, + { + "epoch": 0.0998015992304455, + "grad_norm": 1.9636972772136378, + "learning_rate": 3.949185036794222e-06, + "loss": 0.7619, + "step": 830 + }, + { + "epoch": 0.0999218421210846, + "grad_norm": 1.6105143393654202, + "learning_rate": 3.949010410740884e-06, + "loss": 0.797, + "step": 831 + }, + { + "epoch": 0.10004208501172368, + "grad_norm": 1.8868160499140818, + "learning_rate": 3.948835489024216e-06, + "loss": 0.8742, + "step": 832 + }, + { + "epoch": 0.10016232790236278, + "grad_norm": 1.8604148713801294, + "learning_rate": 3.948660271670755e-06, + "loss": 0.9099, + "step": 833 + }, + { + "epoch": 0.10028257079300186, + "grad_norm": 2.4171058211924206, + "learning_rate": 3.948484758707079e-06, + "loss": 0.8595, + "step": 834 + }, + { + "epoch": 0.10040281368364096, + "grad_norm": 1.8775118215623041, + "learning_rate": 3.948308950159815e-06, + "loss": 0.848, + "step": 835 + }, + { + "epoch": 0.10052305657428004, + "grad_norm": 2.5609760870362654, + "learning_rate": 3.9481328460556326e-06, + "loss": 0.7711, + "step": 836 + }, + { + "epoch": 0.10064329946491914, + "grad_norm": 1.9620855246052549, + "learning_rate": 3.9479564464212455e-06, + "loss": 0.9057, + "step": 837 + }, + { + "epoch": 0.10076354235555823, + "grad_norm": 2.3776960630230337, + "learning_rate": 3.947779751283414e-06, + "loss": 0.7865, + "step": 838 + }, + { + "epoch": 0.10088378524619732, + "grad_norm": 1.8288307457891708, + "learning_rate": 3.947602760668944e-06, + "loss": 0.7704, + "step": 839 + }, + { + "epoch": 0.10100402813683641, + "grad_norm": 1.7461224239990405, + "learning_rate": 3.947425474604684e-06, + "loss": 0.7239, + "step": 840 + }, + { + "epoch": 0.1011242710274755, + "grad_norm": 1.9956531374292052, + "learning_rate": 3.947247893117528e-06, + "loss": 0.9347, + "step": 841 + }, + { + "epoch": 0.10124451391811459, + "grad_norm": 2.7947127012703117, + "learning_rate": 3.947070016234413e-06, + "loss": 0.7017, + "step": 842 + }, + { + "epoch": 0.10136475680875369, + "grad_norm": 2.671634774813937, + "learning_rate": 3.946891843982326e-06, + "loss": 0.7691, + "step": 843 + }, + { + "epoch": 0.10148499969939277, + "grad_norm": 2.62467118883248, + "learning_rate": 3.9467133763882935e-06, + "loss": 0.7675, + "step": 844 + }, + { + "epoch": 0.10160524259003187, + "grad_norm": 1.8500406312346869, + "learning_rate": 3.9465346134793905e-06, + "loss": 0.8705, + "step": 845 + }, + { + "epoch": 0.10172548548067095, + "grad_norm": 2.289196960874597, + "learning_rate": 3.9463555552827335e-06, + "loss": 0.8104, + "step": 846 + }, + { + "epoch": 0.10184572837131005, + "grad_norm": 2.5289208327784194, + "learning_rate": 3.946176201825487e-06, + "loss": 0.8837, + "step": 847 + }, + { + "epoch": 0.10196597126194913, + "grad_norm": 2.1257188366549715, + "learning_rate": 3.9459965531348575e-06, + "loss": 0.8496, + "step": 848 + }, + { + "epoch": 0.10208621415258823, + "grad_norm": 2.960756321499676, + "learning_rate": 3.945816609238098e-06, + "loss": 0.8798, + "step": 849 + }, + { + "epoch": 0.10220645704322733, + "grad_norm": 2.0473578662468634, + "learning_rate": 3.945636370162507e-06, + "loss": 0.868, + "step": 850 + }, + { + "epoch": 0.10232669993386641, + "grad_norm": 1.697175088274094, + "learning_rate": 3.945455835935425e-06, + "loss": 0.8005, + "step": 851 + }, + { + "epoch": 0.1024469428245055, + "grad_norm": 2.048810087839157, + "learning_rate": 3.94527500658424e-06, + "loss": 0.7647, + "step": 852 + }, + { + "epoch": 0.10256718571514459, + "grad_norm": 1.692394223535174, + "learning_rate": 3.945093882136382e-06, + "loss": 0.8238, + "step": 853 + }, + { + "epoch": 0.10268742860578368, + "grad_norm": 1.7905887939627405, + "learning_rate": 3.944912462619329e-06, + "loss": 0.8536, + "step": 854 + }, + { + "epoch": 0.10280767149642277, + "grad_norm": 2.6580051681950527, + "learning_rate": 3.9447307480606025e-06, + "loss": 0.8266, + "step": 855 + }, + { + "epoch": 0.10292791438706186, + "grad_norm": 1.94853672487705, + "learning_rate": 3.944548738487767e-06, + "loss": 0.9126, + "step": 856 + }, + { + "epoch": 0.10304815727770096, + "grad_norm": 2.1623030027063033, + "learning_rate": 3.944366433928434e-06, + "loss": 0.9193, + "step": 857 + }, + { + "epoch": 0.10316840016834004, + "grad_norm": 1.5534722260496143, + "learning_rate": 3.9441838344102594e-06, + "loss": 0.8477, + "step": 858 + }, + { + "epoch": 0.10328864305897914, + "grad_norm": 2.236910361910848, + "learning_rate": 3.944000939960943e-06, + "loss": 0.6936, + "step": 859 + }, + { + "epoch": 0.10340888594961822, + "grad_norm": 1.5913415550149526, + "learning_rate": 3.943817750608229e-06, + "loss": 0.8044, + "step": 860 + }, + { + "epoch": 0.10352912884025732, + "grad_norm": 2.233518422935719, + "learning_rate": 3.943634266379908e-06, + "loss": 0.8401, + "step": 861 + }, + { + "epoch": 0.10364937173089642, + "grad_norm": 1.7635882622890375, + "learning_rate": 3.943450487303815e-06, + "loss": 0.8596, + "step": 862 + }, + { + "epoch": 0.1037696146215355, + "grad_norm": 2.1787142204283874, + "learning_rate": 3.943266413407827e-06, + "loss": 0.8648, + "step": 863 + }, + { + "epoch": 0.1038898575121746, + "grad_norm": 2.2276960733175573, + "learning_rate": 3.94308204471987e-06, + "loss": 0.8631, + "step": 864 + }, + { + "epoch": 0.10401010040281368, + "grad_norm": 2.4069159440822694, + "learning_rate": 3.942897381267912e-06, + "loss": 0.752, + "step": 865 + }, + { + "epoch": 0.10413034329345278, + "grad_norm": 2.5123661328233324, + "learning_rate": 3.942712423079965e-06, + "loss": 0.6733, + "step": 866 + }, + { + "epoch": 0.10425058618409186, + "grad_norm": 2.284480381320965, + "learning_rate": 3.942527170184088e-06, + "loss": 0.9131, + "step": 867 + }, + { + "epoch": 0.10437082907473096, + "grad_norm": 2.6817460674940907, + "learning_rate": 3.942341622608385e-06, + "loss": 0.7964, + "step": 868 + }, + { + "epoch": 0.10449107196537005, + "grad_norm": 1.593521550578913, + "learning_rate": 3.942155780381001e-06, + "loss": 0.7845, + "step": 869 + }, + { + "epoch": 0.10461131485600914, + "grad_norm": 1.8659564952174172, + "learning_rate": 3.94196964353013e-06, + "loss": 0.7724, + "step": 870 + }, + { + "epoch": 0.10473155774664823, + "grad_norm": 1.7694277701531256, + "learning_rate": 3.941783212084008e-06, + "loss": 0.8126, + "step": 871 + }, + { + "epoch": 0.10485180063728732, + "grad_norm": 2.5304173171353854, + "learning_rate": 3.941596486070916e-06, + "loss": 0.8008, + "step": 872 + }, + { + "epoch": 0.10497204352792641, + "grad_norm": 2.123275618175494, + "learning_rate": 3.941409465519182e-06, + "loss": 0.6069, + "step": 873 + }, + { + "epoch": 0.10509228641856551, + "grad_norm": 1.606991085576113, + "learning_rate": 3.941222150457176e-06, + "loss": 0.86, + "step": 874 + }, + { + "epoch": 0.10521252930920459, + "grad_norm": 2.4099396000807984, + "learning_rate": 3.941034540913311e-06, + "loss": 0.725, + "step": 875 + }, + { + "epoch": 0.10533277219984369, + "grad_norm": 1.7808154730584729, + "learning_rate": 3.940846636916051e-06, + "loss": 0.8328, + "step": 876 + }, + { + "epoch": 0.10545301509048277, + "grad_norm": 2.0347205224606904, + "learning_rate": 3.940658438493899e-06, + "loss": 0.8756, + "step": 877 + }, + { + "epoch": 0.10557325798112187, + "grad_norm": 2.0001337483507204, + "learning_rate": 3.940469945675405e-06, + "loss": 0.7777, + "step": 878 + }, + { + "epoch": 0.10569350087176095, + "grad_norm": 1.9227363083039082, + "learning_rate": 3.940281158489163e-06, + "loss": 0.9219, + "step": 879 + }, + { + "epoch": 0.10581374376240005, + "grad_norm": 1.665101063041034, + "learning_rate": 3.940092076963812e-06, + "loss": 0.8428, + "step": 880 + }, + { + "epoch": 0.10593398665303914, + "grad_norm": 2.453193614996316, + "learning_rate": 3.9399027011280355e-06, + "loss": 0.8035, + "step": 881 + }, + { + "epoch": 0.10605422954367823, + "grad_norm": 2.0825066515757933, + "learning_rate": 3.939713031010561e-06, + "loss": 0.786, + "step": 882 + }, + { + "epoch": 0.10617447243431732, + "grad_norm": 2.0746077982674413, + "learning_rate": 3.939523066640163e-06, + "loss": 0.788, + "step": 883 + }, + { + "epoch": 0.10629471532495641, + "grad_norm": 1.7539259605744486, + "learning_rate": 3.939332808045657e-06, + "loss": 0.8203, + "step": 884 + }, + { + "epoch": 0.1064149582155955, + "grad_norm": 1.7187309264078385, + "learning_rate": 3.939142255255906e-06, + "loss": 0.8516, + "step": 885 + }, + { + "epoch": 0.1065352011062346, + "grad_norm": 2.9181617356384204, + "learning_rate": 3.938951408299817e-06, + "loss": 0.8797, + "step": 886 + }, + { + "epoch": 0.10665544399687368, + "grad_norm": 0.8074745135226923, + "learning_rate": 3.938760267206342e-06, + "loss": 0.5451, + "step": 887 + }, + { + "epoch": 0.10677568688751278, + "grad_norm": 2.311792007528311, + "learning_rate": 3.938568832004475e-06, + "loss": 0.7995, + "step": 888 + }, + { + "epoch": 0.10689592977815186, + "grad_norm": 2.0206135371384812, + "learning_rate": 3.938377102723257e-06, + "loss": 0.7677, + "step": 889 + }, + { + "epoch": 0.10701617266879096, + "grad_norm": 3.1288879622592902, + "learning_rate": 3.938185079391774e-06, + "loss": 0.846, + "step": 890 + }, + { + "epoch": 0.10713641555943004, + "grad_norm": 2.4422260342399382, + "learning_rate": 3.937992762039157e-06, + "loss": 1.0709, + "step": 891 + }, + { + "epoch": 0.10725665845006914, + "grad_norm": 1.5581883259316969, + "learning_rate": 3.937800150694577e-06, + "loss": 0.8094, + "step": 892 + }, + { + "epoch": 0.10737690134070824, + "grad_norm": 2.253826913478586, + "learning_rate": 3.937607245387255e-06, + "loss": 0.7816, + "step": 893 + }, + { + "epoch": 0.10749714423134732, + "grad_norm": 2.084203906951221, + "learning_rate": 3.937414046146455e-06, + "loss": 0.729, + "step": 894 + }, + { + "epoch": 0.10761738712198642, + "grad_norm": 2.130669156123598, + "learning_rate": 3.9372205530014845e-06, + "loss": 0.7708, + "step": 895 + }, + { + "epoch": 0.1077376300126255, + "grad_norm": 1.903653612307942, + "learning_rate": 3.937026765981696e-06, + "loss": 0.726, + "step": 896 + }, + { + "epoch": 0.1078578729032646, + "grad_norm": 1.9908560099320882, + "learning_rate": 3.936832685116488e-06, + "loss": 0.8058, + "step": 897 + }, + { + "epoch": 0.10797811579390369, + "grad_norm": 2.818681070707191, + "learning_rate": 3.936638310435301e-06, + "loss": 0.9162, + "step": 898 + }, + { + "epoch": 0.10809835868454278, + "grad_norm": 2.015702120466253, + "learning_rate": 3.936443641967623e-06, + "loss": 0.8242, + "step": 899 + }, + { + "epoch": 0.10821860157518187, + "grad_norm": 1.8546931201621488, + "learning_rate": 3.936248679742983e-06, + "loss": 0.8395, + "step": 900 + }, + { + "epoch": 0.10833884446582095, + "grad_norm": 1.060701474887171, + "learning_rate": 3.936053423790959e-06, + "loss": 0.7032, + "step": 901 + }, + { + "epoch": 0.10845908735646005, + "grad_norm": 1.6081145452344128, + "learning_rate": 3.935857874141168e-06, + "loss": 0.7744, + "step": 902 + }, + { + "epoch": 0.10857933024709913, + "grad_norm": 2.1521425793577817, + "learning_rate": 3.935662030823279e-06, + "loss": 0.8519, + "step": 903 + }, + { + "epoch": 0.10869957313773823, + "grad_norm": 2.5732872164584877, + "learning_rate": 3.935465893866998e-06, + "loss": 0.7391, + "step": 904 + }, + { + "epoch": 0.10881981602837733, + "grad_norm": 1.8866120934203827, + "learning_rate": 3.935269463302079e-06, + "loss": 0.8114, + "step": 905 + }, + { + "epoch": 0.10894005891901641, + "grad_norm": 1.7806857369783624, + "learning_rate": 3.935072739158322e-06, + "loss": 0.7779, + "step": 906 + }, + { + "epoch": 0.10906030180965551, + "grad_norm": 1.7084106877499299, + "learning_rate": 3.934875721465569e-06, + "loss": 0.8099, + "step": 907 + }, + { + "epoch": 0.10918054470029459, + "grad_norm": 2.427558383102071, + "learning_rate": 3.9346784102537076e-06, + "loss": 0.7294, + "step": 908 + }, + { + "epoch": 0.10930078759093369, + "grad_norm": 1.721549441928058, + "learning_rate": 3.934480805552669e-06, + "loss": 0.7885, + "step": 909 + }, + { + "epoch": 0.10942103048157277, + "grad_norm": 2.013786009283639, + "learning_rate": 3.93428290739243e-06, + "loss": 0.8892, + "step": 910 + }, + { + "epoch": 0.10954127337221187, + "grad_norm": 2.9330836435455057, + "learning_rate": 3.9340847158030125e-06, + "loss": 0.8133, + "step": 911 + }, + { + "epoch": 0.10966151626285096, + "grad_norm": 1.8270248624115308, + "learning_rate": 3.9338862308144814e-06, + "loss": 0.7661, + "step": 912 + }, + { + "epoch": 0.10978175915349005, + "grad_norm": 1.6884608182509526, + "learning_rate": 3.933687452456946e-06, + "loss": 0.8569, + "step": 913 + }, + { + "epoch": 0.10990200204412914, + "grad_norm": 2.1006951226127377, + "learning_rate": 3.933488380760562e-06, + "loss": 0.8858, + "step": 914 + }, + { + "epoch": 0.11002224493476823, + "grad_norm": 2.335125836365436, + "learning_rate": 3.9332890157555286e-06, + "loss": 0.8774, + "step": 915 + }, + { + "epoch": 0.11014248782540732, + "grad_norm": 1.8015016782076787, + "learning_rate": 3.933089357472088e-06, + "loss": 0.7746, + "step": 916 + }, + { + "epoch": 0.11026273071604642, + "grad_norm": 1.8128905697645636, + "learning_rate": 3.932889405940529e-06, + "loss": 0.8744, + "step": 917 + }, + { + "epoch": 0.1103829736066855, + "grad_norm": 2.322275686282347, + "learning_rate": 3.932689161191184e-06, + "loss": 0.8127, + "step": 918 + }, + { + "epoch": 0.1105032164973246, + "grad_norm": 2.0161283589023573, + "learning_rate": 3.93248862325443e-06, + "loss": 0.8907, + "step": 919 + }, + { + "epoch": 0.11062345938796368, + "grad_norm": 0.9477627737582651, + "learning_rate": 3.932287792160688e-06, + "loss": 0.6456, + "step": 920 + }, + { + "epoch": 0.11074370227860278, + "grad_norm": 2.3995134257763504, + "learning_rate": 3.932086667940424e-06, + "loss": 0.8214, + "step": 921 + }, + { + "epoch": 0.11086394516924186, + "grad_norm": 2.0774087460303754, + "learning_rate": 3.93188525062415e-06, + "loss": 0.827, + "step": 922 + }, + { + "epoch": 0.11098418805988096, + "grad_norm": 1.775416542929975, + "learning_rate": 3.931683540242418e-06, + "loss": 0.865, + "step": 923 + }, + { + "epoch": 0.11110443095052006, + "grad_norm": 2.792690364595342, + "learning_rate": 3.9314815368258295e-06, + "loss": 0.9254, + "step": 924 + }, + { + "epoch": 0.11122467384115914, + "grad_norm": 4.12683573247613, + "learning_rate": 3.9312792404050275e-06, + "loss": 0.7983, + "step": 925 + }, + { + "epoch": 0.11134491673179824, + "grad_norm": 1.8593948747870963, + "learning_rate": 3.9310766510107e-06, + "loss": 0.7829, + "step": 926 + }, + { + "epoch": 0.11146515962243732, + "grad_norm": 1.8823315651202297, + "learning_rate": 3.9308737686735806e-06, + "loss": 0.9307, + "step": 927 + }, + { + "epoch": 0.11158540251307641, + "grad_norm": 2.0659359142486067, + "learning_rate": 3.9306705934244455e-06, + "loss": 0.8374, + "step": 928 + }, + { + "epoch": 0.11170564540371551, + "grad_norm": 1.7493526078699448, + "learning_rate": 3.930467125294116e-06, + "loss": 0.8859, + "step": 929 + }, + { + "epoch": 0.1118258882943546, + "grad_norm": 0.9397392550983408, + "learning_rate": 3.930263364313458e-06, + "loss": 0.6064, + "step": 930 + }, + { + "epoch": 0.11194613118499369, + "grad_norm": 2.2158884212001153, + "learning_rate": 3.930059310513384e-06, + "loss": 0.8473, + "step": 931 + }, + { + "epoch": 0.11206637407563277, + "grad_norm": 1.75858761122457, + "learning_rate": 3.929854963924846e-06, + "loss": 0.848, + "step": 932 + }, + { + "epoch": 0.11218661696627187, + "grad_norm": 2.086835689260112, + "learning_rate": 3.929650324578845e-06, + "loss": 0.779, + "step": 933 + }, + { + "epoch": 0.11230685985691095, + "grad_norm": 2.351406561158009, + "learning_rate": 3.929445392506423e-06, + "loss": 0.8263, + "step": 934 + }, + { + "epoch": 0.11242710274755005, + "grad_norm": 1.8034427858631152, + "learning_rate": 3.92924016773867e-06, + "loss": 0.7685, + "step": 935 + }, + { + "epoch": 0.11254734563818915, + "grad_norm": 2.1966864471046637, + "learning_rate": 3.9290346503067175e-06, + "loss": 0.7488, + "step": 936 + }, + { + "epoch": 0.11266758852882823, + "grad_norm": 1.7136197598952039, + "learning_rate": 3.9288288402417415e-06, + "loss": 0.8035, + "step": 937 + }, + { + "epoch": 0.11278783141946733, + "grad_norm": 3.2140027935915607, + "learning_rate": 3.928622737574964e-06, + "loss": 0.7168, + "step": 938 + }, + { + "epoch": 0.11290807431010641, + "grad_norm": 1.7634800397902226, + "learning_rate": 3.928416342337652e-06, + "loss": 0.9179, + "step": 939 + }, + { + "epoch": 0.1130283172007455, + "grad_norm": 1.7860573262009807, + "learning_rate": 3.928209654561113e-06, + "loss": 0.8359, + "step": 940 + }, + { + "epoch": 0.1131485600913846, + "grad_norm": 1.9606270715363892, + "learning_rate": 3.928002674276703e-06, + "loss": 0.8237, + "step": 941 + }, + { + "epoch": 0.11326880298202369, + "grad_norm": 2.1098483084438926, + "learning_rate": 3.92779540151582e-06, + "loss": 0.7728, + "step": 942 + }, + { + "epoch": 0.11338904587266278, + "grad_norm": 3.3199764306024613, + "learning_rate": 3.927587836309907e-06, + "loss": 0.8631, + "step": 943 + }, + { + "epoch": 0.11350928876330187, + "grad_norm": 1.8593588435649173, + "learning_rate": 3.927379978690452e-06, + "loss": 0.7924, + "step": 944 + }, + { + "epoch": 0.11362953165394096, + "grad_norm": 2.710260576058457, + "learning_rate": 3.927171828688987e-06, + "loss": 0.8868, + "step": 945 + }, + { + "epoch": 0.11374977454458005, + "grad_norm": 2.0397738681074222, + "learning_rate": 3.926963386337088e-06, + "loss": 0.8271, + "step": 946 + }, + { + "epoch": 0.11387001743521914, + "grad_norm": 2.3226752276885096, + "learning_rate": 3.926754651666375e-06, + "loss": 0.7209, + "step": 947 + }, + { + "epoch": 0.11399026032585824, + "grad_norm": 3.33380888089853, + "learning_rate": 3.926545624708513e-06, + "loss": 0.7915, + "step": 948 + }, + { + "epoch": 0.11411050321649732, + "grad_norm": 1.7485974686083445, + "learning_rate": 3.926336305495213e-06, + "loss": 0.8649, + "step": 949 + }, + { + "epoch": 0.11423074610713642, + "grad_norm": 2.6257508430207204, + "learning_rate": 3.926126694058226e-06, + "loss": 0.8991, + "step": 950 + }, + { + "epoch": 0.1143509889977755, + "grad_norm": 1.4624306295521992, + "learning_rate": 3.92591679042935e-06, + "loss": 0.8263, + "step": 951 + }, + { + "epoch": 0.1144712318884146, + "grad_norm": 2.780809860549582, + "learning_rate": 3.92570659464043e-06, + "loss": 0.8316, + "step": 952 + }, + { + "epoch": 0.1145914747790537, + "grad_norm": 2.0078584539594604, + "learning_rate": 3.925496106723349e-06, + "loss": 0.8048, + "step": 953 + }, + { + "epoch": 0.11471171766969278, + "grad_norm": 1.8526106817048527, + "learning_rate": 3.9252853267100405e-06, + "loss": 0.848, + "step": 954 + }, + { + "epoch": 0.11483196056033187, + "grad_norm": 1.8893779077990205, + "learning_rate": 3.9250742546324786e-06, + "loss": 0.8439, + "step": 955 + }, + { + "epoch": 0.11495220345097096, + "grad_norm": 1.7385535551944251, + "learning_rate": 3.924862890522683e-06, + "loss": 0.8723, + "step": 956 + }, + { + "epoch": 0.11507244634161005, + "grad_norm": 2.155986051022448, + "learning_rate": 3.9246512344127174e-06, + "loss": 0.8823, + "step": 957 + }, + { + "epoch": 0.11519268923224914, + "grad_norm": 1.8645113996841107, + "learning_rate": 3.9244392863346895e-06, + "loss": 0.8271, + "step": 958 + }, + { + "epoch": 0.11531293212288823, + "grad_norm": 2.856722909855173, + "learning_rate": 3.9242270463207524e-06, + "loss": 0.9326, + "step": 959 + }, + { + "epoch": 0.11543317501352733, + "grad_norm": 2.9559290253993318, + "learning_rate": 3.924014514403102e-06, + "loss": 0.864, + "step": 960 + }, + { + "epoch": 0.11555341790416641, + "grad_norm": 4.322544507778782, + "learning_rate": 3.92380169061398e-06, + "loss": 0.9204, + "step": 961 + }, + { + "epoch": 0.11567366079480551, + "grad_norm": 2.1894524716351613, + "learning_rate": 3.9235885749856705e-06, + "loss": 0.8543, + "step": 962 + }, + { + "epoch": 0.1157939036854446, + "grad_norm": 1.8902655646030588, + "learning_rate": 3.9233751675505035e-06, + "loss": 0.8334, + "step": 963 + }, + { + "epoch": 0.11591414657608369, + "grad_norm": 1.9272940924238875, + "learning_rate": 3.923161468340853e-06, + "loss": 0.8595, + "step": 964 + }, + { + "epoch": 0.11603438946672277, + "grad_norm": 1.7733538771756323, + "learning_rate": 3.9229474773891374e-06, + "loss": 0.8259, + "step": 965 + }, + { + "epoch": 0.11615463235736187, + "grad_norm": 2.056500349486711, + "learning_rate": 3.922733194727818e-06, + "loss": 0.8541, + "step": 966 + }, + { + "epoch": 0.11627487524800097, + "grad_norm": 2.9530400662596112, + "learning_rate": 3.922518620389402e-06, + "loss": 0.8791, + "step": 967 + }, + { + "epoch": 0.11639511813864005, + "grad_norm": 1.7410438786936477, + "learning_rate": 3.922303754406439e-06, + "loss": 0.9021, + "step": 968 + }, + { + "epoch": 0.11651536102927915, + "grad_norm": 2.025408753007712, + "learning_rate": 3.922088596811526e-06, + "loss": 0.7999, + "step": 969 + }, + { + "epoch": 0.11663560391991823, + "grad_norm": 2.254858599202771, + "learning_rate": 3.9218731476373e-06, + "loss": 0.8795, + "step": 970 + }, + { + "epoch": 0.11675584681055733, + "grad_norm": 1.9586368758418369, + "learning_rate": 3.9216574069164455e-06, + "loss": 0.8549, + "step": 971 + }, + { + "epoch": 0.11687608970119642, + "grad_norm": 1.569813055826368, + "learning_rate": 3.921441374681691e-06, + "loss": 0.8119, + "step": 972 + }, + { + "epoch": 0.1169963325918355, + "grad_norm": 2.2748680202222573, + "learning_rate": 3.921225050965808e-06, + "loss": 0.6614, + "step": 973 + }, + { + "epoch": 0.1171165754824746, + "grad_norm": 2.1223790329968373, + "learning_rate": 3.921008435801612e-06, + "loss": 0.7531, + "step": 974 + }, + { + "epoch": 0.11723681837311369, + "grad_norm": 2.5501674597113815, + "learning_rate": 3.920791529221963e-06, + "loss": 0.7831, + "step": 975 + }, + { + "epoch": 0.11735706126375278, + "grad_norm": 1.733510807317247, + "learning_rate": 3.920574331259768e-06, + "loss": 0.7717, + "step": 976 + }, + { + "epoch": 0.11747730415439187, + "grad_norm": 3.907331515320613, + "learning_rate": 3.9203568419479716e-06, + "loss": 0.812, + "step": 977 + }, + { + "epoch": 0.11759754704503096, + "grad_norm": 1.8465581296132134, + "learning_rate": 3.92013906131957e-06, + "loss": 0.7606, + "step": 978 + }, + { + "epoch": 0.11771778993567006, + "grad_norm": 1.5157887478493464, + "learning_rate": 3.9199209894076e-06, + "loss": 0.8296, + "step": 979 + }, + { + "epoch": 0.11783803282630914, + "grad_norm": 2.816532380855009, + "learning_rate": 3.919702626245142e-06, + "loss": 0.9139, + "step": 980 + }, + { + "epoch": 0.11795827571694824, + "grad_norm": 2.1035227655911184, + "learning_rate": 3.919483971865322e-06, + "loss": 0.6813, + "step": 981 + }, + { + "epoch": 0.11807851860758732, + "grad_norm": 1.9231264335421183, + "learning_rate": 3.91926502630131e-06, + "loss": 0.8893, + "step": 982 + }, + { + "epoch": 0.11819876149822642, + "grad_norm": 1.796670454241854, + "learning_rate": 3.91904578958632e-06, + "loss": 0.7273, + "step": 983 + }, + { + "epoch": 0.11831900438886551, + "grad_norm": 2.078030117995205, + "learning_rate": 3.918826261753608e-06, + "loss": 0.8511, + "step": 984 + }, + { + "epoch": 0.1184392472795046, + "grad_norm": 2.4711312507343095, + "learning_rate": 3.918606442836478e-06, + "loss": 0.7271, + "step": 985 + }, + { + "epoch": 0.1185594901701437, + "grad_norm": 1.7936620099221627, + "learning_rate": 3.918386332868277e-06, + "loss": 0.7855, + "step": 986 + }, + { + "epoch": 0.11867973306078278, + "grad_norm": 1.719124146399629, + "learning_rate": 3.918165931882394e-06, + "loss": 0.9521, + "step": 987 + }, + { + "epoch": 0.11879997595142187, + "grad_norm": 2.439849699112684, + "learning_rate": 3.917945239912264e-06, + "loss": 0.7775, + "step": 988 + }, + { + "epoch": 0.11892021884206096, + "grad_norm": 2.0209917878102535, + "learning_rate": 3.917724256991367e-06, + "loss": 0.758, + "step": 989 + }, + { + "epoch": 0.11904046173270005, + "grad_norm": 2.0099140257027295, + "learning_rate": 3.9175029831532245e-06, + "loss": 0.8294, + "step": 990 + }, + { + "epoch": 0.11916070462333915, + "grad_norm": 2.0172781855190416, + "learning_rate": 3.917281418431404e-06, + "loss": 0.887, + "step": 991 + }, + { + "epoch": 0.11928094751397823, + "grad_norm": 4.281846944583136, + "learning_rate": 3.917059562859516e-06, + "loss": 0.7805, + "step": 992 + }, + { + "epoch": 0.11940119040461733, + "grad_norm": 2.2201149194201926, + "learning_rate": 3.916837416471218e-06, + "loss": 0.8966, + "step": 993 + }, + { + "epoch": 0.11952143329525641, + "grad_norm": 2.705805337156206, + "learning_rate": 3.916614979300207e-06, + "loss": 0.7402, + "step": 994 + }, + { + "epoch": 0.11964167618589551, + "grad_norm": 1.9171387533318962, + "learning_rate": 3.9163922513802274e-06, + "loss": 0.796, + "step": 995 + }, + { + "epoch": 0.1197619190765346, + "grad_norm": 3.322302367889118, + "learning_rate": 3.916169232745067e-06, + "loss": 0.8436, + "step": 996 + }, + { + "epoch": 0.11988216196717369, + "grad_norm": 2.73156113936466, + "learning_rate": 3.915945923428559e-06, + "loss": 0.9293, + "step": 997 + }, + { + "epoch": 0.12000240485781279, + "grad_norm": 2.3848212108335503, + "learning_rate": 3.915722323464577e-06, + "loss": 0.8419, + "step": 998 + }, + { + "epoch": 0.12012264774845187, + "grad_norm": 2.093715952126672, + "learning_rate": 3.91549843288704e-06, + "loss": 0.7249, + "step": 999 + }, + { + "epoch": 0.12024289063909097, + "grad_norm": 1.9594943750937817, + "learning_rate": 3.915274251729916e-06, + "loss": 0.8007, + "step": 1000 + }, + { + "epoch": 0.12036313352973005, + "grad_norm": 2.1129113964851847, + "learning_rate": 3.91504978002721e-06, + "loss": 0.9114, + "step": 1001 + }, + { + "epoch": 0.12048337642036915, + "grad_norm": 1.9940633880273695, + "learning_rate": 3.914825017812974e-06, + "loss": 0.7815, + "step": 1002 + }, + { + "epoch": 0.12060361931100824, + "grad_norm": 2.188921657172827, + "learning_rate": 3.9145999651213065e-06, + "loss": 0.7405, + "step": 1003 + }, + { + "epoch": 0.12072386220164733, + "grad_norm": 2.4166623038767754, + "learning_rate": 3.9143746219863465e-06, + "loss": 0.9024, + "step": 1004 + }, + { + "epoch": 0.12084410509228642, + "grad_norm": 0.9244570375758007, + "learning_rate": 3.914148988442278e-06, + "loss": 0.648, + "step": 1005 + }, + { + "epoch": 0.1209643479829255, + "grad_norm": 2.6963165624446495, + "learning_rate": 3.91392306452333e-06, + "loss": 0.9678, + "step": 1006 + }, + { + "epoch": 0.1210845908735646, + "grad_norm": 3.1436921131424254, + "learning_rate": 3.913696850263774e-06, + "loss": 0.6836, + "step": 1007 + }, + { + "epoch": 0.1212048337642037, + "grad_norm": 1.973107375125752, + "learning_rate": 3.913470345697929e-06, + "loss": 0.8013, + "step": 1008 + }, + { + "epoch": 0.12132507665484278, + "grad_norm": 2.0853556417961414, + "learning_rate": 3.913243550860153e-06, + "loss": 0.8701, + "step": 1009 + }, + { + "epoch": 0.12144531954548188, + "grad_norm": 1.756909625982056, + "learning_rate": 3.913016465784852e-06, + "loss": 0.7737, + "step": 1010 + }, + { + "epoch": 0.12156556243612096, + "grad_norm": 2.6438874626044493, + "learning_rate": 3.912789090506474e-06, + "loss": 0.7381, + "step": 1011 + }, + { + "epoch": 0.12168580532676006, + "grad_norm": 2.4500403965324966, + "learning_rate": 3.9125614250595114e-06, + "loss": 0.7277, + "step": 1012 + }, + { + "epoch": 0.12180604821739914, + "grad_norm": 3.0178980981751584, + "learning_rate": 3.912333469478502e-06, + "loss": 0.9021, + "step": 1013 + }, + { + "epoch": 0.12192629110803824, + "grad_norm": 3.056629392950609, + "learning_rate": 3.912105223798025e-06, + "loss": 0.7914, + "step": 1014 + }, + { + "epoch": 0.12204653399867733, + "grad_norm": 1.01131534266286, + "learning_rate": 3.9118766880527065e-06, + "loss": 0.6769, + "step": 1015 + }, + { + "epoch": 0.12216677688931642, + "grad_norm": 1.7702592779443556, + "learning_rate": 3.9116478622772145e-06, + "loss": 0.7487, + "step": 1016 + }, + { + "epoch": 0.12228701977995551, + "grad_norm": 3.0998547366163076, + "learning_rate": 3.911418746506261e-06, + "loss": 0.888, + "step": 1017 + }, + { + "epoch": 0.1224072626705946, + "grad_norm": 1.682239244272194, + "learning_rate": 3.911189340774604e-06, + "loss": 0.797, + "step": 1018 + }, + { + "epoch": 0.1225275055612337, + "grad_norm": 1.7769010222975257, + "learning_rate": 3.910959645117043e-06, + "loss": 0.8033, + "step": 1019 + }, + { + "epoch": 0.12264774845187278, + "grad_norm": 0.8168578290703427, + "learning_rate": 3.910729659568423e-06, + "loss": 0.5678, + "step": 1020 + }, + { + "epoch": 0.12276799134251187, + "grad_norm": 1.7116684419176404, + "learning_rate": 3.9104993841636344e-06, + "loss": 0.8292, + "step": 1021 + }, + { + "epoch": 0.12288823423315097, + "grad_norm": 1.8349713319623102, + "learning_rate": 3.910268818937608e-06, + "loss": 0.8168, + "step": 1022 + }, + { + "epoch": 0.12300847712379005, + "grad_norm": 2.5008755104536133, + "learning_rate": 3.9100379639253196e-06, + "loss": 0.8841, + "step": 1023 + }, + { + "epoch": 0.12312872001442915, + "grad_norm": 2.8622776994633514, + "learning_rate": 3.909806819161791e-06, + "loss": 0.874, + "step": 1024 + }, + { + "epoch": 0.12324896290506823, + "grad_norm": 2.1027828488050107, + "learning_rate": 3.909575384682086e-06, + "loss": 0.8734, + "step": 1025 + }, + { + "epoch": 0.12336920579570733, + "grad_norm": 1.726618632098895, + "learning_rate": 3.9093436605213144e-06, + "loss": 0.6986, + "step": 1026 + }, + { + "epoch": 0.12348944868634643, + "grad_norm": 1.7724551534426976, + "learning_rate": 3.909111646714627e-06, + "loss": 0.8112, + "step": 1027 + }, + { + "epoch": 0.12360969157698551, + "grad_norm": 2.1895198760801975, + "learning_rate": 3.9088793432972206e-06, + "loss": 0.7363, + "step": 1028 + }, + { + "epoch": 0.1237299344676246, + "grad_norm": 2.0276452577158794, + "learning_rate": 3.908646750304336e-06, + "loss": 0.834, + "step": 1029 + }, + { + "epoch": 0.12385017735826369, + "grad_norm": 1.6093318896399098, + "learning_rate": 3.908413867771257e-06, + "loss": 0.8758, + "step": 1030 + }, + { + "epoch": 0.12397042024890279, + "grad_norm": 1.6360472255223577, + "learning_rate": 3.908180695733311e-06, + "loss": 0.8223, + "step": 1031 + }, + { + "epoch": 0.12409066313954187, + "grad_norm": 1.8260864890633057, + "learning_rate": 3.907947234225871e-06, + "loss": 0.8402, + "step": 1032 + }, + { + "epoch": 0.12421090603018096, + "grad_norm": 2.092593656606538, + "learning_rate": 3.907713483284352e-06, + "loss": 0.88, + "step": 1033 + }, + { + "epoch": 0.12433114892082006, + "grad_norm": 2.59496444398486, + "learning_rate": 3.907479442944216e-06, + "loss": 0.9798, + "step": 1034 + }, + { + "epoch": 0.12445139181145914, + "grad_norm": 2.140791503667864, + "learning_rate": 3.907245113240963e-06, + "loss": 0.9341, + "step": 1035 + }, + { + "epoch": 0.12457163470209824, + "grad_norm": 1.8139118089277733, + "learning_rate": 3.907010494210144e-06, + "loss": 0.753, + "step": 1036 + }, + { + "epoch": 0.12469187759273732, + "grad_norm": 2.1015534241207368, + "learning_rate": 3.9067755858873495e-06, + "loss": 0.9332, + "step": 1037 + }, + { + "epoch": 0.12481212048337642, + "grad_norm": 0.8664105689557696, + "learning_rate": 3.906540388308214e-06, + "loss": 0.629, + "step": 1038 + }, + { + "epoch": 0.12493236337401552, + "grad_norm": 1.765420277297495, + "learning_rate": 3.906304901508417e-06, + "loss": 0.8226, + "step": 1039 + }, + { + "epoch": 0.12505260626465461, + "grad_norm": 2.4819456494549077, + "learning_rate": 3.9060691255236835e-06, + "loss": 0.7682, + "step": 1040 + }, + { + "epoch": 0.1251728491552937, + "grad_norm": 1.7756044701769393, + "learning_rate": 3.905833060389778e-06, + "loss": 0.8139, + "step": 1041 + }, + { + "epoch": 0.12529309204593278, + "grad_norm": 2.4356755006256776, + "learning_rate": 3.905596706142513e-06, + "loss": 0.7972, + "step": 1042 + }, + { + "epoch": 0.12541333493657186, + "grad_norm": 2.06447437459499, + "learning_rate": 3.9053600628177435e-06, + "loss": 0.8707, + "step": 1043 + }, + { + "epoch": 0.12553357782721097, + "grad_norm": 2.0551547255358495, + "learning_rate": 3.905123130451367e-06, + "loss": 0.8597, + "step": 1044 + }, + { + "epoch": 0.12565382071785006, + "grad_norm": 2.605481469439296, + "learning_rate": 3.904885909079326e-06, + "loss": 0.8032, + "step": 1045 + }, + { + "epoch": 0.12577406360848914, + "grad_norm": 2.355297278493856, + "learning_rate": 3.904648398737607e-06, + "loss": 0.7953, + "step": 1046 + }, + { + "epoch": 0.12589430649912825, + "grad_norm": 2.1831884127367047, + "learning_rate": 3.9044105994622406e-06, + "loss": 0.7853, + "step": 1047 + }, + { + "epoch": 0.12601454938976733, + "grad_norm": 1.897279299741131, + "learning_rate": 3.9041725112893005e-06, + "loss": 0.8233, + "step": 1048 + }, + { + "epoch": 0.12613479228040642, + "grad_norm": 2.0596483341522105, + "learning_rate": 3.903934134254904e-06, + "loss": 0.7636, + "step": 1049 + }, + { + "epoch": 0.1262550351710455, + "grad_norm": 2.1494671937648953, + "learning_rate": 3.903695468395213e-06, + "loss": 0.8619, + "step": 1050 + }, + { + "epoch": 0.1263752780616846, + "grad_norm": 2.155365251003731, + "learning_rate": 3.903456513746434e-06, + "loss": 0.5722, + "step": 1051 + }, + { + "epoch": 0.1264955209523237, + "grad_norm": 2.278672421254852, + "learning_rate": 3.903217270344815e-06, + "loss": 0.8832, + "step": 1052 + }, + { + "epoch": 0.12661576384296278, + "grad_norm": 1.732650311707799, + "learning_rate": 3.902977738226648e-06, + "loss": 0.8297, + "step": 1053 + }, + { + "epoch": 0.12673600673360189, + "grad_norm": 2.1931846868394733, + "learning_rate": 3.902737917428273e-06, + "loss": 0.9198, + "step": 1054 + }, + { + "epoch": 0.12685624962424097, + "grad_norm": 1.8401829757103607, + "learning_rate": 3.902497807986068e-06, + "loss": 0.8545, + "step": 1055 + }, + { + "epoch": 0.12697649251488005, + "grad_norm": 2.1689170619150158, + "learning_rate": 3.902257409936458e-06, + "loss": 0.8466, + "step": 1056 + }, + { + "epoch": 0.12709673540551916, + "grad_norm": 2.08394501925116, + "learning_rate": 3.902016723315912e-06, + "loss": 0.8508, + "step": 1057 + }, + { + "epoch": 0.12721697829615825, + "grad_norm": 2.13036924716643, + "learning_rate": 3.901775748160941e-06, + "loss": 0.7066, + "step": 1058 + }, + { + "epoch": 0.12733722118679733, + "grad_norm": 0.8185123127714373, + "learning_rate": 3.901534484508101e-06, + "loss": 0.6095, + "step": 1059 + }, + { + "epoch": 0.1274574640774364, + "grad_norm": 2.1812211523908362, + "learning_rate": 3.901292932393991e-06, + "loss": 0.7531, + "step": 1060 + }, + { + "epoch": 0.12757770696807552, + "grad_norm": 2.1898085265741023, + "learning_rate": 3.9010510918552555e-06, + "loss": 0.8681, + "step": 1061 + }, + { + "epoch": 0.1276979498587146, + "grad_norm": 2.44494557903237, + "learning_rate": 3.900808962928581e-06, + "loss": 0.7664, + "step": 1062 + }, + { + "epoch": 0.1278181927493537, + "grad_norm": 2.373461375515096, + "learning_rate": 3.900566545650698e-06, + "loss": 0.9027, + "step": 1063 + }, + { + "epoch": 0.1279384356399928, + "grad_norm": 2.368792603444279, + "learning_rate": 3.900323840058381e-06, + "loss": 0.8326, + "step": 1064 + }, + { + "epoch": 0.12805867853063188, + "grad_norm": 1.9158605732529232, + "learning_rate": 3.900080846188449e-06, + "loss": 0.822, + "step": 1065 + }, + { + "epoch": 0.12817892142127096, + "grad_norm": 1.7936952403146873, + "learning_rate": 3.8998375640777625e-06, + "loss": 0.8243, + "step": 1066 + }, + { + "epoch": 0.12829916431191005, + "grad_norm": 0.7127191206639696, + "learning_rate": 3.899593993763229e-06, + "loss": 0.527, + "step": 1067 + }, + { + "epoch": 0.12841940720254916, + "grad_norm": 2.516343103118067, + "learning_rate": 3.899350135281796e-06, + "loss": 0.8301, + "step": 1068 + }, + { + "epoch": 0.12853965009318824, + "grad_norm": 2.0959858130418967, + "learning_rate": 3.8991059886704585e-06, + "loss": 0.8028, + "step": 1069 + }, + { + "epoch": 0.12865989298382732, + "grad_norm": 1.9261934098552511, + "learning_rate": 3.898861553966252e-06, + "loss": 0.8371, + "step": 1070 + }, + { + "epoch": 0.12878013587446643, + "grad_norm": 1.6882499158812063, + "learning_rate": 3.898616831206257e-06, + "loss": 0.8869, + "step": 1071 + }, + { + "epoch": 0.12890037876510552, + "grad_norm": 1.9495574106557179, + "learning_rate": 3.8983718204276e-06, + "loss": 0.7772, + "step": 1072 + }, + { + "epoch": 0.1290206216557446, + "grad_norm": 3.2106946867304162, + "learning_rate": 3.898126521667446e-06, + "loss": 0.8296, + "step": 1073 + }, + { + "epoch": 0.12914086454638368, + "grad_norm": 1.9335187801122788, + "learning_rate": 3.897880934963007e-06, + "loss": 0.8399, + "step": 1074 + }, + { + "epoch": 0.1292611074370228, + "grad_norm": 2.3183838506369137, + "learning_rate": 3.89763506035154e-06, + "loss": 0.8006, + "step": 1075 + }, + { + "epoch": 0.12938135032766188, + "grad_norm": 1.658919935454736, + "learning_rate": 3.897388897870343e-06, + "loss": 0.8231, + "step": 1076 + }, + { + "epoch": 0.12950159321830096, + "grad_norm": 1.9038479789422809, + "learning_rate": 3.89714244755676e-06, + "loss": 0.7657, + "step": 1077 + }, + { + "epoch": 0.12962183610894007, + "grad_norm": 2.7569946286630316, + "learning_rate": 3.896895709448175e-06, + "loss": 0.8728, + "step": 1078 + }, + { + "epoch": 0.12974207899957915, + "grad_norm": 3.091559356993247, + "learning_rate": 3.896648683582019e-06, + "loss": 0.7848, + "step": 1079 + }, + { + "epoch": 0.12986232189021824, + "grad_norm": 2.17949563552689, + "learning_rate": 3.896401369995766e-06, + "loss": 0.8176, + "step": 1080 + }, + { + "epoch": 0.12998256478085732, + "grad_norm": 1.8535697776352376, + "learning_rate": 3.896153768726932e-06, + "loss": 0.8037, + "step": 1081 + }, + { + "epoch": 0.13010280767149643, + "grad_norm": 2.0948725865114186, + "learning_rate": 3.8959058798130806e-06, + "loss": 0.8915, + "step": 1082 + }, + { + "epoch": 0.1302230505621355, + "grad_norm": 1.7756193745949298, + "learning_rate": 3.895657703291814e-06, + "loss": 0.7671, + "step": 1083 + }, + { + "epoch": 0.1303432934527746, + "grad_norm": 2.51174333480265, + "learning_rate": 3.895409239200781e-06, + "loss": 0.8087, + "step": 1084 + }, + { + "epoch": 0.1304635363434137, + "grad_norm": 2.1371687113703954, + "learning_rate": 3.895160487577673e-06, + "loss": 0.9305, + "step": 1085 + }, + { + "epoch": 0.1305837792340528, + "grad_norm": 0.7918746616450116, + "learning_rate": 3.894911448460226e-06, + "loss": 0.611, + "step": 1086 + }, + { + "epoch": 0.13070402212469187, + "grad_norm": 2.1577765548109804, + "learning_rate": 3.8946621218862195e-06, + "loss": 0.7391, + "step": 1087 + }, + { + "epoch": 0.13082426501533098, + "grad_norm": 1.9499347113414496, + "learning_rate": 3.894412507893475e-06, + "loss": 0.9009, + "step": 1088 + }, + { + "epoch": 0.13094450790597006, + "grad_norm": 2.0556486369299503, + "learning_rate": 3.894162606519859e-06, + "loss": 0.7304, + "step": 1089 + }, + { + "epoch": 0.13106475079660915, + "grad_norm": 2.0284318594824016, + "learning_rate": 3.893912417803282e-06, + "loss": 0.7814, + "step": 1090 + }, + { + "epoch": 0.13118499368724823, + "grad_norm": 1.9705630839620805, + "learning_rate": 3.8936619417816975e-06, + "loss": 0.7832, + "step": 1091 + }, + { + "epoch": 0.13130523657788734, + "grad_norm": 1.8566520692170436, + "learning_rate": 3.8934111784931015e-06, + "loss": 0.7212, + "step": 1092 + }, + { + "epoch": 0.13142547946852642, + "grad_norm": 0.9340681841751669, + "learning_rate": 3.893160127975535e-06, + "loss": 0.5916, + "step": 1093 + }, + { + "epoch": 0.1315457223591655, + "grad_norm": 2.6418850392317146, + "learning_rate": 3.8929087902670826e-06, + "loss": 0.8241, + "step": 1094 + }, + { + "epoch": 0.13166596524980462, + "grad_norm": 1.085043669670001, + "learning_rate": 3.8926571654058715e-06, + "loss": 0.6086, + "step": 1095 + }, + { + "epoch": 0.1317862081404437, + "grad_norm": 2.3245484190435843, + "learning_rate": 3.892405253430074e-06, + "loss": 0.7801, + "step": 1096 + }, + { + "epoch": 0.13190645103108278, + "grad_norm": 1.9162464994335113, + "learning_rate": 3.892153054377904e-06, + "loss": 0.831, + "step": 1097 + }, + { + "epoch": 0.13202669392172187, + "grad_norm": 1.0169486590191619, + "learning_rate": 3.891900568287619e-06, + "loss": 0.5951, + "step": 1098 + }, + { + "epoch": 0.13214693681236098, + "grad_norm": 3.5554576668244913, + "learning_rate": 3.891647795197523e-06, + "loss": 0.7289, + "step": 1099 + }, + { + "epoch": 0.13226717970300006, + "grad_norm": 1.990473050676487, + "learning_rate": 3.8913947351459605e-06, + "loss": 0.6954, + "step": 1100 + }, + { + "epoch": 0.13238742259363914, + "grad_norm": 1.937863100544994, + "learning_rate": 3.89114138817132e-06, + "loss": 0.694, + "step": 1101 + }, + { + "epoch": 0.13250766548427825, + "grad_norm": 1.8618848514456323, + "learning_rate": 3.890887754312035e-06, + "loss": 0.858, + "step": 1102 + }, + { + "epoch": 0.13262790837491734, + "grad_norm": 1.74648312999704, + "learning_rate": 3.890633833606581e-06, + "loss": 0.8872, + "step": 1103 + }, + { + "epoch": 0.13274815126555642, + "grad_norm": 2.037028149222562, + "learning_rate": 3.890379626093477e-06, + "loss": 0.7097, + "step": 1104 + }, + { + "epoch": 0.1328683941561955, + "grad_norm": 1.96824318553666, + "learning_rate": 3.890125131811287e-06, + "loss": 0.9384, + "step": 1105 + }, + { + "epoch": 0.1329886370468346, + "grad_norm": 1.834356723154963, + "learning_rate": 3.889870350798618e-06, + "loss": 0.7618, + "step": 1106 + }, + { + "epoch": 0.1331088799374737, + "grad_norm": 2.1386727668377135, + "learning_rate": 3.889615283094119e-06, + "loss": 0.7899, + "step": 1107 + }, + { + "epoch": 0.13322912282811278, + "grad_norm": 2.2860566802386315, + "learning_rate": 3.889359928736485e-06, + "loss": 0.8622, + "step": 1108 + }, + { + "epoch": 0.1333493657187519, + "grad_norm": 2.192906483254431, + "learning_rate": 3.889104287764451e-06, + "loss": 0.9208, + "step": 1109 + }, + { + "epoch": 0.13346960860939097, + "grad_norm": 3.4746268037698775, + "learning_rate": 3.888848360216798e-06, + "loss": 0.9199, + "step": 1110 + }, + { + "epoch": 0.13358985150003005, + "grad_norm": 0.9109166779273113, + "learning_rate": 3.888592146132351e-06, + "loss": 0.5671, + "step": 1111 + }, + { + "epoch": 0.13371009439066917, + "grad_norm": 1.8227215980427613, + "learning_rate": 3.888335645549978e-06, + "loss": 0.796, + "step": 1112 + }, + { + "epoch": 0.13383033728130825, + "grad_norm": 2.3614560479914632, + "learning_rate": 3.888078858508588e-06, + "loss": 0.8331, + "step": 1113 + }, + { + "epoch": 0.13395058017194733, + "grad_norm": 5.08787338048901, + "learning_rate": 3.8878217850471365e-06, + "loss": 0.8517, + "step": 1114 + }, + { + "epoch": 0.13407082306258641, + "grad_norm": 1.81566278152419, + "learning_rate": 3.887564425204621e-06, + "loss": 0.7475, + "step": 1115 + }, + { + "epoch": 0.13419106595322552, + "grad_norm": 0.8534098893584259, + "learning_rate": 3.887306779020083e-06, + "loss": 0.5483, + "step": 1116 + }, + { + "epoch": 0.1343113088438646, + "grad_norm": 2.545985809688174, + "learning_rate": 3.887048846532608e-06, + "loss": 0.7193, + "step": 1117 + }, + { + "epoch": 0.1344315517345037, + "grad_norm": 0.7700880331025293, + "learning_rate": 3.8867906277813224e-06, + "loss": 0.5831, + "step": 1118 + }, + { + "epoch": 0.1345517946251428, + "grad_norm": 2.154786207387717, + "learning_rate": 3.886532122805399e-06, + "loss": 0.7464, + "step": 1119 + }, + { + "epoch": 0.13467203751578188, + "grad_norm": 2.0191204900285604, + "learning_rate": 3.886273331644053e-06, + "loss": 0.9058, + "step": 1120 + }, + { + "epoch": 0.13479228040642097, + "grad_norm": 2.3528099426669074, + "learning_rate": 3.886014254336542e-06, + "loss": 0.8321, + "step": 1121 + }, + { + "epoch": 0.13491252329706005, + "grad_norm": 1.7137674416754243, + "learning_rate": 3.885754890922168e-06, + "loss": 0.9278, + "step": 1122 + }, + { + "epoch": 0.13503276618769916, + "grad_norm": 1.83505571982373, + "learning_rate": 3.885495241440277e-06, + "loss": 0.7948, + "step": 1123 + }, + { + "epoch": 0.13515300907833824, + "grad_norm": 1.8478438622870954, + "learning_rate": 3.885235305930257e-06, + "loss": 0.7521, + "step": 1124 + }, + { + "epoch": 0.13527325196897733, + "grad_norm": 1.9497913444587525, + "learning_rate": 3.884975084431539e-06, + "loss": 0.8601, + "step": 1125 + }, + { + "epoch": 0.13539349485961644, + "grad_norm": 2.3915269932739847, + "learning_rate": 3.8847145769836e-06, + "loss": 0.9277, + "step": 1126 + }, + { + "epoch": 0.13551373775025552, + "grad_norm": 3.071645199939922, + "learning_rate": 3.884453783625959e-06, + "loss": 0.6804, + "step": 1127 + }, + { + "epoch": 0.1356339806408946, + "grad_norm": 2.068363524694846, + "learning_rate": 3.884192704398176e-06, + "loss": 0.8574, + "step": 1128 + }, + { + "epoch": 0.13575422353153369, + "grad_norm": 1.6534769127142073, + "learning_rate": 3.883931339339858e-06, + "loss": 0.7579, + "step": 1129 + }, + { + "epoch": 0.1358744664221728, + "grad_norm": 2.6443541790100378, + "learning_rate": 3.883669688490654e-06, + "loss": 0.8004, + "step": 1130 + }, + { + "epoch": 0.13599470931281188, + "grad_norm": 1.9977133314968112, + "learning_rate": 3.883407751890256e-06, + "loss": 0.8622, + "step": 1131 + }, + { + "epoch": 0.13611495220345096, + "grad_norm": 1.677273883617122, + "learning_rate": 3.8831455295783994e-06, + "loss": 0.8598, + "step": 1132 + }, + { + "epoch": 0.13623519509409007, + "grad_norm": 1.7521494199405407, + "learning_rate": 3.882883021594864e-06, + "loss": 0.7498, + "step": 1133 + }, + { + "epoch": 0.13635543798472916, + "grad_norm": 2.4582815218708136, + "learning_rate": 3.8826202279794705e-06, + "loss": 0.8775, + "step": 1134 + }, + { + "epoch": 0.13647568087536824, + "grad_norm": 2.0144041637235217, + "learning_rate": 3.882357148772085e-06, + "loss": 0.7123, + "step": 1135 + }, + { + "epoch": 0.13659592376600732, + "grad_norm": 2.5699297495343143, + "learning_rate": 3.882093784012617e-06, + "loss": 0.849, + "step": 1136 + }, + { + "epoch": 0.13671616665664643, + "grad_norm": 1.7623212845469016, + "learning_rate": 3.881830133741019e-06, + "loss": 0.8452, + "step": 1137 + }, + { + "epoch": 0.13683640954728551, + "grad_norm": 1.9389956454326018, + "learning_rate": 3.881566197997285e-06, + "loss": 0.7713, + "step": 1138 + }, + { + "epoch": 0.1369566524379246, + "grad_norm": 1.6888201459410366, + "learning_rate": 3.881301976821456e-06, + "loss": 0.7542, + "step": 1139 + }, + { + "epoch": 0.1370768953285637, + "grad_norm": 1.7268108344677706, + "learning_rate": 3.881037470253612e-06, + "loss": 0.9122, + "step": 1140 + }, + { + "epoch": 0.1371971382192028, + "grad_norm": 2.836335629767306, + "learning_rate": 3.88077267833388e-06, + "loss": 0.8074, + "step": 1141 + }, + { + "epoch": 0.13731738110984187, + "grad_norm": 2.073487118642592, + "learning_rate": 3.880507601102427e-06, + "loss": 0.8441, + "step": 1142 + }, + { + "epoch": 0.13743762400048098, + "grad_norm": 1.8219345135398823, + "learning_rate": 3.880242238599467e-06, + "loss": 0.8295, + "step": 1143 + }, + { + "epoch": 0.13755786689112007, + "grad_norm": 1.7101043344758502, + "learning_rate": 3.879976590865254e-06, + "loss": 0.8399, + "step": 1144 + }, + { + "epoch": 0.13767810978175915, + "grad_norm": 1.85754021910385, + "learning_rate": 3.879710657940087e-06, + "loss": 0.8774, + "step": 1145 + }, + { + "epoch": 0.13779835267239823, + "grad_norm": 1.9694211512187838, + "learning_rate": 3.879444439864308e-06, + "loss": 0.7057, + "step": 1146 + }, + { + "epoch": 0.13791859556303734, + "grad_norm": 1.8191717281553645, + "learning_rate": 3.879177936678301e-06, + "loss": 0.8693, + "step": 1147 + }, + { + "epoch": 0.13803883845367643, + "grad_norm": 3.55402155077169, + "learning_rate": 3.878911148422496e-06, + "loss": 0.7767, + "step": 1148 + }, + { + "epoch": 0.1381590813443155, + "grad_norm": 2.374112816794964, + "learning_rate": 3.878644075137364e-06, + "loss": 0.7129, + "step": 1149 + }, + { + "epoch": 0.13827932423495462, + "grad_norm": 1.9934782745884398, + "learning_rate": 3.878376716863418e-06, + "loss": 0.8006, + "step": 1150 + }, + { + "epoch": 0.1383995671255937, + "grad_norm": 2.5030769486404987, + "learning_rate": 3.878109073641219e-06, + "loss": 0.7267, + "step": 1151 + }, + { + "epoch": 0.13851981001623279, + "grad_norm": 1.6361604525874986, + "learning_rate": 3.877841145511366e-06, + "loss": 0.8194, + "step": 1152 + }, + { + "epoch": 0.13864005290687187, + "grad_norm": 1.7208526927360883, + "learning_rate": 3.8775729325145035e-06, + "loss": 0.8388, + "step": 1153 + }, + { + "epoch": 0.13876029579751098, + "grad_norm": 0.8064276758225439, + "learning_rate": 3.877304434691321e-06, + "loss": 0.6496, + "step": 1154 + }, + { + "epoch": 0.13888053868815006, + "grad_norm": 1.763733044278556, + "learning_rate": 3.877035652082548e-06, + "loss": 0.8041, + "step": 1155 + }, + { + "epoch": 0.13900078157878915, + "grad_norm": 2.0590137726377318, + "learning_rate": 3.87676658472896e-06, + "loss": 0.8605, + "step": 1156 + }, + { + "epoch": 0.13912102446942826, + "grad_norm": 2.6258155827319296, + "learning_rate": 3.876497232671372e-06, + "loss": 0.8659, + "step": 1157 + }, + { + "epoch": 0.13924126736006734, + "grad_norm": 3.307150209460491, + "learning_rate": 3.876227595950647e-06, + "loss": 0.8478, + "step": 1158 + }, + { + "epoch": 0.13936151025070642, + "grad_norm": 1.6181695140316288, + "learning_rate": 3.875957674607686e-06, + "loss": 0.7998, + "step": 1159 + }, + { + "epoch": 0.1394817531413455, + "grad_norm": 2.1437086184764422, + "learning_rate": 3.8756874686834386e-06, + "loss": 0.8914, + "step": 1160 + }, + { + "epoch": 0.13960199603198462, + "grad_norm": 1.9411364239286455, + "learning_rate": 3.875416978218893e-06, + "loss": 0.8087, + "step": 1161 + }, + { + "epoch": 0.1397222389226237, + "grad_norm": 2.960799000100166, + "learning_rate": 3.8751462032550835e-06, + "loss": 0.8349, + "step": 1162 + }, + { + "epoch": 0.13984248181326278, + "grad_norm": 2.200411593342197, + "learning_rate": 3.874875143833085e-06, + "loss": 0.8367, + "step": 1163 + }, + { + "epoch": 0.1399627247039019, + "grad_norm": 1.8196152428141819, + "learning_rate": 3.874603799994019e-06, + "loss": 0.6999, + "step": 1164 + }, + { + "epoch": 0.14008296759454097, + "grad_norm": 2.0232154989521494, + "learning_rate": 3.874332171779046e-06, + "loss": 0.8761, + "step": 1165 + }, + { + "epoch": 0.14020321048518006, + "grad_norm": 1.699854449043788, + "learning_rate": 3.874060259229373e-06, + "loss": 0.7601, + "step": 1166 + }, + { + "epoch": 0.14032345337581917, + "grad_norm": 2.184987287157031, + "learning_rate": 3.873788062386249e-06, + "loss": 0.9486, + "step": 1167 + }, + { + "epoch": 0.14044369626645825, + "grad_norm": 1.9373603893555835, + "learning_rate": 3.873515581290965e-06, + "loss": 0.8349, + "step": 1168 + }, + { + "epoch": 0.14056393915709733, + "grad_norm": 2.022075533466347, + "learning_rate": 3.8732428159848575e-06, + "loss": 0.7644, + "step": 1169 + }, + { + "epoch": 0.14068418204773642, + "grad_norm": 2.2257994297901265, + "learning_rate": 3.872969766509304e-06, + "loss": 0.793, + "step": 1170 + }, + { + "epoch": 0.14080442493837553, + "grad_norm": 0.7663223438431829, + "learning_rate": 3.872696432905726e-06, + "loss": 0.5569, + "step": 1171 + }, + { + "epoch": 0.1409246678290146, + "grad_norm": 2.6709782394956676, + "learning_rate": 3.872422815215589e-06, + "loss": 0.7265, + "step": 1172 + }, + { + "epoch": 0.1410449107196537, + "grad_norm": 4.6223468903292115, + "learning_rate": 3.8721489134803994e-06, + "loss": 0.757, + "step": 1173 + }, + { + "epoch": 0.1411651536102928, + "grad_norm": 2.346171844432628, + "learning_rate": 3.871874727741707e-06, + "loss": 0.7408, + "step": 1174 + }, + { + "epoch": 0.1412853965009319, + "grad_norm": 1.7663036156337075, + "learning_rate": 3.871600258041108e-06, + "loss": 0.9691, + "step": 1175 + }, + { + "epoch": 0.14140563939157097, + "grad_norm": 2.2657796872704896, + "learning_rate": 3.871325504420238e-06, + "loss": 0.8701, + "step": 1176 + }, + { + "epoch": 0.14152588228221005, + "grad_norm": 2.0137704997141057, + "learning_rate": 3.871050466920776e-06, + "loss": 0.8302, + "step": 1177 + }, + { + "epoch": 0.14164612517284916, + "grad_norm": 1.9364820698115592, + "learning_rate": 3.870775145584447e-06, + "loss": 0.8122, + "step": 1178 + }, + { + "epoch": 0.14176636806348825, + "grad_norm": 4.031986139788344, + "learning_rate": 3.8704995404530145e-06, + "loss": 0.6614, + "step": 1179 + }, + { + "epoch": 0.14188661095412733, + "grad_norm": 1.667869975897475, + "learning_rate": 3.87022365156829e-06, + "loss": 0.8587, + "step": 1180 + }, + { + "epoch": 0.14200685384476644, + "grad_norm": 2.0499097152925616, + "learning_rate": 3.869947478972123e-06, + "loss": 0.8184, + "step": 1181 + }, + { + "epoch": 0.14212709673540552, + "grad_norm": 1.8855944456584308, + "learning_rate": 3.869671022706412e-06, + "loss": 0.8318, + "step": 1182 + }, + { + "epoch": 0.1422473396260446, + "grad_norm": 1.8825471303240287, + "learning_rate": 3.869394282813092e-06, + "loss": 0.6567, + "step": 1183 + }, + { + "epoch": 0.1423675825166837, + "grad_norm": 5.963378086650695, + "learning_rate": 3.869117259334147e-06, + "loss": 0.9085, + "step": 1184 + }, + { + "epoch": 0.1424878254073228, + "grad_norm": 1.7965067610535137, + "learning_rate": 3.868839952311599e-06, + "loss": 0.8264, + "step": 1185 + }, + { + "epoch": 0.14260806829796188, + "grad_norm": 2.202980848425667, + "learning_rate": 3.868562361787516e-06, + "loss": 0.8158, + "step": 1186 + }, + { + "epoch": 0.14272831118860096, + "grad_norm": 2.0901670401956896, + "learning_rate": 3.868284487804009e-06, + "loss": 0.6992, + "step": 1187 + }, + { + "epoch": 0.14284855407924008, + "grad_norm": 1.5689778895719653, + "learning_rate": 3.86800633040323e-06, + "loss": 0.7859, + "step": 1188 + }, + { + "epoch": 0.14296879696987916, + "grad_norm": 2.007530341008165, + "learning_rate": 3.867727889627376e-06, + "loss": 0.7894, + "step": 1189 + }, + { + "epoch": 0.14308903986051824, + "grad_norm": 2.4938143499575993, + "learning_rate": 3.867449165518687e-06, + "loss": 0.8006, + "step": 1190 + }, + { + "epoch": 0.14320928275115732, + "grad_norm": 3.878724185170488, + "learning_rate": 3.867170158119444e-06, + "loss": 0.7237, + "step": 1191 + }, + { + "epoch": 0.14332952564179643, + "grad_norm": 2.2085748456450185, + "learning_rate": 3.866890867471972e-06, + "loss": 0.7698, + "step": 1192 + }, + { + "epoch": 0.14344976853243552, + "grad_norm": 2.4081094666372844, + "learning_rate": 3.86661129361864e-06, + "loss": 0.9053, + "step": 1193 + }, + { + "epoch": 0.1435700114230746, + "grad_norm": 1.881992905646968, + "learning_rate": 3.866331436601859e-06, + "loss": 0.8669, + "step": 1194 + }, + { + "epoch": 0.1436902543137137, + "grad_norm": 2.154473388302139, + "learning_rate": 3.866051296464083e-06, + "loss": 0.7495, + "step": 1195 + }, + { + "epoch": 0.1438104972043528, + "grad_norm": 2.0171281990536043, + "learning_rate": 3.86577087324781e-06, + "loss": 0.8584, + "step": 1196 + }, + { + "epoch": 0.14393074009499188, + "grad_norm": 2.0678076219782224, + "learning_rate": 3.865490166995578e-06, + "loss": 0.7804, + "step": 1197 + }, + { + "epoch": 0.144050982985631, + "grad_norm": 2.353611962337163, + "learning_rate": 3.86520917774997e-06, + "loss": 0.8534, + "step": 1198 + }, + { + "epoch": 0.14417122587627007, + "grad_norm": 5.753787576211688, + "learning_rate": 3.864927905553614e-06, + "loss": 0.7539, + "step": 1199 + }, + { + "epoch": 0.14429146876690915, + "grad_norm": 2.0155521350080594, + "learning_rate": 3.8646463504491765e-06, + "loss": 0.8961, + "step": 1200 + }, + { + "epoch": 0.14441171165754824, + "grad_norm": 1.8733461079325948, + "learning_rate": 3.8643645124793705e-06, + "loss": 0.844, + "step": 1201 + }, + { + "epoch": 0.14453195454818735, + "grad_norm": 1.9953892130731186, + "learning_rate": 3.8640823916869515e-06, + "loss": 0.7542, + "step": 1202 + }, + { + "epoch": 0.14465219743882643, + "grad_norm": 3.097821725512009, + "learning_rate": 3.863799988114714e-06, + "loss": 0.7909, + "step": 1203 + }, + { + "epoch": 0.1447724403294655, + "grad_norm": 2.840618715184879, + "learning_rate": 3.863517301805502e-06, + "loss": 0.7251, + "step": 1204 + }, + { + "epoch": 0.14489268322010462, + "grad_norm": 2.6109309325943175, + "learning_rate": 3.863234332802196e-06, + "loss": 0.9795, + "step": 1205 + }, + { + "epoch": 0.1450129261107437, + "grad_norm": 2.073747426882775, + "learning_rate": 3.862951081147723e-06, + "loss": 0.7538, + "step": 1206 + }, + { + "epoch": 0.1451331690013828, + "grad_norm": 2.090203541239713, + "learning_rate": 3.862667546885053e-06, + "loss": 0.7964, + "step": 1207 + }, + { + "epoch": 0.14525341189202187, + "grad_norm": 2.101350114151874, + "learning_rate": 3.8623837300571965e-06, + "loss": 0.7465, + "step": 1208 + }, + { + "epoch": 0.14537365478266098, + "grad_norm": 1.7697478913328173, + "learning_rate": 3.8620996307072085e-06, + "loss": 0.8473, + "step": 1209 + }, + { + "epoch": 0.14549389767330007, + "grad_norm": 2.262739777899844, + "learning_rate": 3.861815248878188e-06, + "loss": 0.6497, + "step": 1210 + }, + { + "epoch": 0.14561414056393915, + "grad_norm": 2.356455227846421, + "learning_rate": 3.861530584613274e-06, + "loss": 0.8052, + "step": 1211 + }, + { + "epoch": 0.14573438345457826, + "grad_norm": 5.140091610400116, + "learning_rate": 3.86124563795565e-06, + "loss": 0.8254, + "step": 1212 + }, + { + "epoch": 0.14585462634521734, + "grad_norm": 1.8669305728371222, + "learning_rate": 3.860960408948543e-06, + "loss": 0.7086, + "step": 1213 + }, + { + "epoch": 0.14597486923585642, + "grad_norm": 2.4020414332771565, + "learning_rate": 3.860674897635222e-06, + "loss": 0.9099, + "step": 1214 + }, + { + "epoch": 0.1460951121264955, + "grad_norm": 2.2693520243749465, + "learning_rate": 3.860389104058998e-06, + "loss": 0.8429, + "step": 1215 + }, + { + "epoch": 0.14621535501713462, + "grad_norm": 1.8358089320840634, + "learning_rate": 3.860103028263227e-06, + "loss": 0.7324, + "step": 1216 + }, + { + "epoch": 0.1463355979077737, + "grad_norm": 2.20323776064769, + "learning_rate": 3.859816670291304e-06, + "loss": 0.7102, + "step": 1217 + }, + { + "epoch": 0.14645584079841278, + "grad_norm": 4.073498670486715, + "learning_rate": 3.859530030186672e-06, + "loss": 0.9061, + "step": 1218 + }, + { + "epoch": 0.1465760836890519, + "grad_norm": 2.377406757985225, + "learning_rate": 3.859243107992813e-06, + "loss": 0.8437, + "step": 1219 + }, + { + "epoch": 0.14669632657969098, + "grad_norm": 3.9037361295748645, + "learning_rate": 3.858955903753252e-06, + "loss": 0.8033, + "step": 1220 + }, + { + "epoch": 0.14681656947033006, + "grad_norm": 1.7330262034306763, + "learning_rate": 3.858668417511559e-06, + "loss": 0.8385, + "step": 1221 + }, + { + "epoch": 0.14693681236096917, + "grad_norm": 2.2726831397193172, + "learning_rate": 3.8583806493113445e-06, + "loss": 0.7754, + "step": 1222 + }, + { + "epoch": 0.14705705525160825, + "grad_norm": 1.9543493476943297, + "learning_rate": 3.858092599196263e-06, + "loss": 0.8252, + "step": 1223 + }, + { + "epoch": 0.14717729814224734, + "grad_norm": 2.138286343318628, + "learning_rate": 3.857804267210012e-06, + "loss": 0.8316, + "step": 1224 + }, + { + "epoch": 0.14729754103288642, + "grad_norm": 2.066147786293296, + "learning_rate": 3.857515653396331e-06, + "loss": 0.8848, + "step": 1225 + }, + { + "epoch": 0.14741778392352553, + "grad_norm": 2.176189172060084, + "learning_rate": 3.857226757799002e-06, + "loss": 0.8781, + "step": 1226 + }, + { + "epoch": 0.1475380268141646, + "grad_norm": 2.293273281689938, + "learning_rate": 3.85693758046185e-06, + "loss": 0.7543, + "step": 1227 + }, + { + "epoch": 0.1476582697048037, + "grad_norm": 2.317292376850265, + "learning_rate": 3.8566481214287435e-06, + "loss": 0.8349, + "step": 1228 + }, + { + "epoch": 0.1477785125954428, + "grad_norm": 2.1444529812675275, + "learning_rate": 3.8563583807435935e-06, + "loss": 0.9122, + "step": 1229 + }, + { + "epoch": 0.1478987554860819, + "grad_norm": 1.9313012298406251, + "learning_rate": 3.856068358450353e-06, + "loss": 0.7878, + "step": 1230 + }, + { + "epoch": 0.14801899837672097, + "grad_norm": 1.679770179310522, + "learning_rate": 3.8557780545930186e-06, + "loss": 0.8684, + "step": 1231 + }, + { + "epoch": 0.14813924126736006, + "grad_norm": 1.8420551074521037, + "learning_rate": 3.855487469215628e-06, + "loss": 0.8016, + "step": 1232 + }, + { + "epoch": 0.14825948415799917, + "grad_norm": 2.3454916014576868, + "learning_rate": 3.855196602362264e-06, + "loss": 0.7294, + "step": 1233 + }, + { + "epoch": 0.14837972704863825, + "grad_norm": 2.113084935573175, + "learning_rate": 3.854905454077051e-06, + "loss": 0.951, + "step": 1234 + }, + { + "epoch": 0.14849996993927733, + "grad_norm": 2.7536156034292394, + "learning_rate": 3.854614024404155e-06, + "loss": 0.8888, + "step": 1235 + }, + { + "epoch": 0.14862021282991644, + "grad_norm": 1.8164596713834444, + "learning_rate": 3.8543223133877865e-06, + "loss": 0.9052, + "step": 1236 + }, + { + "epoch": 0.14874045572055553, + "grad_norm": 2.00905122655001, + "learning_rate": 3.854030321072198e-06, + "loss": 0.892, + "step": 1237 + }, + { + "epoch": 0.1488606986111946, + "grad_norm": 1.9581335317533581, + "learning_rate": 3.853738047501682e-06, + "loss": 0.7381, + "step": 1238 + }, + { + "epoch": 0.1489809415018337, + "grad_norm": 2.0068562527484515, + "learning_rate": 3.85344549272058e-06, + "loss": 0.786, + "step": 1239 + }, + { + "epoch": 0.1491011843924728, + "grad_norm": 1.9077465249041057, + "learning_rate": 3.853152656773269e-06, + "loss": 0.8288, + "step": 1240 + }, + { + "epoch": 0.14922142728311188, + "grad_norm": 1.7309119214635105, + "learning_rate": 3.852859539704174e-06, + "loss": 0.8563, + "step": 1241 + }, + { + "epoch": 0.14934167017375097, + "grad_norm": 1.953968262784635, + "learning_rate": 3.85256614155776e-06, + "loss": 0.7688, + "step": 1242 + }, + { + "epoch": 0.14946191306439008, + "grad_norm": 2.0796898707857094, + "learning_rate": 3.852272462378535e-06, + "loss": 0.7543, + "step": 1243 + }, + { + "epoch": 0.14958215595502916, + "grad_norm": 2.214874751767157, + "learning_rate": 3.85197850221105e-06, + "loss": 0.7929, + "step": 1244 + }, + { + "epoch": 0.14970239884566824, + "grad_norm": 1.6040082167940557, + "learning_rate": 3.851684261099899e-06, + "loss": 0.7668, + "step": 1245 + }, + { + "epoch": 0.14982264173630733, + "grad_norm": 1.9221774847904338, + "learning_rate": 3.851389739089718e-06, + "loss": 0.8678, + "step": 1246 + }, + { + "epoch": 0.14994288462694644, + "grad_norm": 1.9088690853880332, + "learning_rate": 3.851094936225186e-06, + "loss": 0.8137, + "step": 1247 + }, + { + "epoch": 0.15006312751758552, + "grad_norm": 1.4941451727106037, + "learning_rate": 3.850799852551024e-06, + "loss": 0.7734, + "step": 1248 + }, + { + "epoch": 0.1501833704082246, + "grad_norm": 2.3121084964400573, + "learning_rate": 3.850504488111995e-06, + "loss": 0.8713, + "step": 1249 + }, + { + "epoch": 0.15030361329886371, + "grad_norm": 1.7625363772270302, + "learning_rate": 3.850208842952907e-06, + "loss": 0.8365, + "step": 1250 + }, + { + "epoch": 0.1504238561895028, + "grad_norm": 1.6443980835846146, + "learning_rate": 3.849912917118608e-06, + "loss": 0.801, + "step": 1251 + }, + { + "epoch": 0.15054409908014188, + "grad_norm": 0.8954645477127093, + "learning_rate": 3.849616710653992e-06, + "loss": 0.5942, + "step": 1252 + }, + { + "epoch": 0.150664341970781, + "grad_norm": 1.8429113517383875, + "learning_rate": 3.84932022360399e-06, + "loss": 0.7567, + "step": 1253 + }, + { + "epoch": 0.15078458486142007, + "grad_norm": 2.5783392614860747, + "learning_rate": 3.849023456013581e-06, + "loss": 0.8431, + "step": 1254 + }, + { + "epoch": 0.15090482775205916, + "grad_norm": 2.023421710921144, + "learning_rate": 3.848726407927784e-06, + "loss": 0.6344, + "step": 1255 + }, + { + "epoch": 0.15102507064269824, + "grad_norm": 2.3293804905373543, + "learning_rate": 3.84842907939166e-06, + "loss": 0.8786, + "step": 1256 + }, + { + "epoch": 0.15114531353333735, + "grad_norm": 3.0582962768416273, + "learning_rate": 3.8481314704503146e-06, + "loss": 0.7188, + "step": 1257 + }, + { + "epoch": 0.15126555642397643, + "grad_norm": 2.1480015537817536, + "learning_rate": 3.847833581148895e-06, + "loss": 0.8932, + "step": 1258 + }, + { + "epoch": 0.15138579931461552, + "grad_norm": 2.2028656935441946, + "learning_rate": 3.84753541153259e-06, + "loss": 0.8237, + "step": 1259 + }, + { + "epoch": 0.15150604220525463, + "grad_norm": 2.331431453894054, + "learning_rate": 3.847236961646633e-06, + "loss": 0.8425, + "step": 1260 + }, + { + "epoch": 0.1516262850958937, + "grad_norm": 2.434419690800465, + "learning_rate": 3.846938231536296e-06, + "loss": 0.7939, + "step": 1261 + }, + { + "epoch": 0.1517465279865328, + "grad_norm": 2.0267933004536185, + "learning_rate": 3.8466392212468995e-06, + "loss": 0.8129, + "step": 1262 + }, + { + "epoch": 0.15186677087717187, + "grad_norm": 0.8317356211248115, + "learning_rate": 3.8463399308238e-06, + "loss": 0.62, + "step": 1263 + }, + { + "epoch": 0.15198701376781099, + "grad_norm": 1.793202303964878, + "learning_rate": 3.846040360312402e-06, + "loss": 0.6496, + "step": 1264 + }, + { + "epoch": 0.15210725665845007, + "grad_norm": 3.2950661564901806, + "learning_rate": 3.8457405097581485e-06, + "loss": 0.8338, + "step": 1265 + }, + { + "epoch": 0.15222749954908915, + "grad_norm": 2.1770201191706247, + "learning_rate": 3.8454403792065275e-06, + "loss": 0.7878, + "step": 1266 + }, + { + "epoch": 0.15234774243972826, + "grad_norm": 2.091828218682947, + "learning_rate": 3.845139968703068e-06, + "loss": 0.8592, + "step": 1267 + }, + { + "epoch": 0.15246798533036734, + "grad_norm": 2.0029989646265225, + "learning_rate": 3.844839278293342e-06, + "loss": 0.8418, + "step": 1268 + }, + { + "epoch": 0.15258822822100643, + "grad_norm": 2.3441698842637373, + "learning_rate": 3.8445383080229654e-06, + "loss": 0.7682, + "step": 1269 + }, + { + "epoch": 0.1527084711116455, + "grad_norm": 2.458057183310543, + "learning_rate": 3.844237057937593e-06, + "loss": 0.7533, + "step": 1270 + }, + { + "epoch": 0.15282871400228462, + "grad_norm": 2.422134287090562, + "learning_rate": 3.843935528082926e-06, + "loss": 0.7901, + "step": 1271 + }, + { + "epoch": 0.1529489568929237, + "grad_norm": 1.7034669585297186, + "learning_rate": 3.843633718504704e-06, + "loss": 0.8575, + "step": 1272 + }, + { + "epoch": 0.1530691997835628, + "grad_norm": 4.761971696201565, + "learning_rate": 3.843331629248715e-06, + "loss": 0.9119, + "step": 1273 + }, + { + "epoch": 0.1531894426742019, + "grad_norm": 3.293065767680408, + "learning_rate": 3.843029260360782e-06, + "loss": 0.7719, + "step": 1274 + }, + { + "epoch": 0.15330968556484098, + "grad_norm": 1.76140414048773, + "learning_rate": 3.8427266118867755e-06, + "loss": 0.7928, + "step": 1275 + }, + { + "epoch": 0.15342992845548006, + "grad_norm": 1.969733461851142, + "learning_rate": 3.842423683872608e-06, + "loss": 0.8437, + "step": 1276 + }, + { + "epoch": 0.15355017134611917, + "grad_norm": 2.3216483127428074, + "learning_rate": 3.842120476364232e-06, + "loss": 0.7909, + "step": 1277 + }, + { + "epoch": 0.15367041423675826, + "grad_norm": 3.030926559338822, + "learning_rate": 3.841816989407644e-06, + "loss": 0.8492, + "step": 1278 + }, + { + "epoch": 0.15379065712739734, + "grad_norm": 2.058045518755425, + "learning_rate": 3.841513223048884e-06, + "loss": 0.7771, + "step": 1279 + }, + { + "epoch": 0.15391090001803642, + "grad_norm": 2.627528562236276, + "learning_rate": 3.841209177334031e-06, + "loss": 0.8071, + "step": 1280 + }, + { + "epoch": 0.15403114290867553, + "grad_norm": 2.661945743807498, + "learning_rate": 3.84090485230921e-06, + "loss": 0.7572, + "step": 1281 + }, + { + "epoch": 0.15415138579931462, + "grad_norm": 2.6744433991523873, + "learning_rate": 3.840600248020588e-06, + "loss": 0.7683, + "step": 1282 + }, + { + "epoch": 0.1542716286899537, + "grad_norm": 2.1506000524225177, + "learning_rate": 3.840295364514371e-06, + "loss": 0.8065, + "step": 1283 + }, + { + "epoch": 0.1543918715805928, + "grad_norm": 2.2032233682951454, + "learning_rate": 3.83999020183681e-06, + "loss": 0.7977, + "step": 1284 + }, + { + "epoch": 0.1545121144712319, + "grad_norm": 1.888418817776501, + "learning_rate": 3.839684760034199e-06, + "loss": 0.7912, + "step": 1285 + }, + { + "epoch": 0.15463235736187098, + "grad_norm": 3.274384791706704, + "learning_rate": 3.8393790391528716e-06, + "loss": 0.6588, + "step": 1286 + }, + { + "epoch": 0.15475260025251006, + "grad_norm": 2.266686131823865, + "learning_rate": 3.8390730392392075e-06, + "loss": 0.8977, + "step": 1287 + }, + { + "epoch": 0.15487284314314917, + "grad_norm": 2.139258512400883, + "learning_rate": 3.838766760339626e-06, + "loss": 0.8059, + "step": 1288 + }, + { + "epoch": 0.15499308603378825, + "grad_norm": 2.791872633742369, + "learning_rate": 3.838460202500587e-06, + "loss": 0.8154, + "step": 1289 + }, + { + "epoch": 0.15511332892442733, + "grad_norm": 2.0132609853734937, + "learning_rate": 3.838153365768599e-06, + "loss": 0.7465, + "step": 1290 + }, + { + "epoch": 0.15523357181506645, + "grad_norm": 4.7584614929648525, + "learning_rate": 3.837846250190206e-06, + "loss": 0.768, + "step": 1291 + }, + { + "epoch": 0.15535381470570553, + "grad_norm": 2.0082347143924237, + "learning_rate": 3.837538855811998e-06, + "loss": 0.7802, + "step": 1292 + }, + { + "epoch": 0.1554740575963446, + "grad_norm": 2.270167985321701, + "learning_rate": 3.837231182680606e-06, + "loss": 0.7203, + "step": 1293 + }, + { + "epoch": 0.1555943004869837, + "grad_norm": 1.6341113167738972, + "learning_rate": 3.836923230842706e-06, + "loss": 0.7692, + "step": 1294 + }, + { + "epoch": 0.1557145433776228, + "grad_norm": 1.9989599861693461, + "learning_rate": 3.836615000345011e-06, + "loss": 0.8196, + "step": 1295 + }, + { + "epoch": 0.1558347862682619, + "grad_norm": 2.685394171794926, + "learning_rate": 3.836306491234282e-06, + "loss": 0.7856, + "step": 1296 + }, + { + "epoch": 0.15595502915890097, + "grad_norm": 2.0452193924567728, + "learning_rate": 3.835997703557317e-06, + "loss": 0.7676, + "step": 1297 + }, + { + "epoch": 0.15607527204954008, + "grad_norm": 1.7629306456479636, + "learning_rate": 3.83568863736096e-06, + "loss": 0.8043, + "step": 1298 + }, + { + "epoch": 0.15619551494017916, + "grad_norm": 2.2978195013102765, + "learning_rate": 3.8353792926920975e-06, + "loss": 0.9036, + "step": 1299 + }, + { + "epoch": 0.15631575783081825, + "grad_norm": 2.078106758209704, + "learning_rate": 3.835069669597655e-06, + "loss": 0.8344, + "step": 1300 + }, + { + "epoch": 0.15643600072145733, + "grad_norm": 2.5540101439499354, + "learning_rate": 3.834759768124603e-06, + "loss": 0.8084, + "step": 1301 + }, + { + "epoch": 0.15655624361209644, + "grad_norm": 2.4705536446567415, + "learning_rate": 3.834449588319953e-06, + "loss": 0.7732, + "step": 1302 + }, + { + "epoch": 0.15667648650273552, + "grad_norm": 1.7262023273931872, + "learning_rate": 3.834139130230758e-06, + "loss": 0.8565, + "step": 1303 + }, + { + "epoch": 0.1567967293933746, + "grad_norm": 1.7576763185095867, + "learning_rate": 3.833828393904117e-06, + "loss": 0.8146, + "step": 1304 + }, + { + "epoch": 0.15691697228401372, + "grad_norm": 2.354933644714496, + "learning_rate": 3.833517379387165e-06, + "loss": 0.7863, + "step": 1305 + }, + { + "epoch": 0.1570372151746528, + "grad_norm": 2.7612307709576536, + "learning_rate": 3.833206086727085e-06, + "loss": 0.8963, + "step": 1306 + }, + { + "epoch": 0.15715745806529188, + "grad_norm": 2.268368027323885, + "learning_rate": 3.8328945159710994e-06, + "loss": 0.7142, + "step": 1307 + }, + { + "epoch": 0.157277700955931, + "grad_norm": 2.2978394228788446, + "learning_rate": 3.832582667166473e-06, + "loss": 0.8906, + "step": 1308 + }, + { + "epoch": 0.15739794384657008, + "grad_norm": 2.2482477358545037, + "learning_rate": 3.8322705403605125e-06, + "loss": 0.8253, + "step": 1309 + }, + { + "epoch": 0.15751818673720916, + "grad_norm": 2.1298234835988317, + "learning_rate": 3.831958135600568e-06, + "loss": 0.8152, + "step": 1310 + }, + { + "epoch": 0.15763842962784824, + "grad_norm": 2.1530288741270303, + "learning_rate": 3.831645452934032e-06, + "loss": 0.8077, + "step": 1311 + }, + { + "epoch": 0.15775867251848735, + "grad_norm": 1.943701755133471, + "learning_rate": 3.831332492408336e-06, + "loss": 0.8116, + "step": 1312 + }, + { + "epoch": 0.15787891540912644, + "grad_norm": 1.9099014830022438, + "learning_rate": 3.831019254070957e-06, + "loss": 0.696, + "step": 1313 + }, + { + "epoch": 0.15799915829976552, + "grad_norm": 3.577220819164299, + "learning_rate": 3.8307057379694135e-06, + "loss": 0.9666, + "step": 1314 + }, + { + "epoch": 0.15811940119040463, + "grad_norm": 2.2482676724893422, + "learning_rate": 3.830391944151264e-06, + "loss": 0.838, + "step": 1315 + }, + { + "epoch": 0.1582396440810437, + "grad_norm": 1.8022619022987583, + "learning_rate": 3.830077872664114e-06, + "loss": 0.6793, + "step": 1316 + }, + { + "epoch": 0.1583598869716828, + "grad_norm": 3.7631999077696365, + "learning_rate": 3.829763523555604e-06, + "loss": 0.7391, + "step": 1317 + }, + { + "epoch": 0.15848012986232188, + "grad_norm": 2.4753996709387738, + "learning_rate": 3.829448896873423e-06, + "loss": 0.7978, + "step": 1318 + }, + { + "epoch": 0.158600372752961, + "grad_norm": 2.139720530755722, + "learning_rate": 3.829133992665299e-06, + "loss": 0.7986, + "step": 1319 + }, + { + "epoch": 0.15872061564360007, + "grad_norm": 2.347831529113433, + "learning_rate": 3.828818810979002e-06, + "loss": 0.902, + "step": 1320 + }, + { + "epoch": 0.15884085853423915, + "grad_norm": 2.7358429374452657, + "learning_rate": 3.8285033518623454e-06, + "loss": 0.8115, + "step": 1321 + }, + { + "epoch": 0.15896110142487826, + "grad_norm": 2.929813799360939, + "learning_rate": 3.8281876153631845e-06, + "loss": 0.8229, + "step": 1322 + }, + { + "epoch": 0.15908134431551735, + "grad_norm": 2.044618017434855, + "learning_rate": 3.827871601529416e-06, + "loss": 0.6584, + "step": 1323 + }, + { + "epoch": 0.15920158720615643, + "grad_norm": 1.7281013638913247, + "learning_rate": 3.827555310408979e-06, + "loss": 0.8043, + "step": 1324 + }, + { + "epoch": 0.1593218300967955, + "grad_norm": 1.6037842761306813, + "learning_rate": 3.827238742049854e-06, + "loss": 0.8357, + "step": 1325 + }, + { + "epoch": 0.15944207298743462, + "grad_norm": 1.725515125594144, + "learning_rate": 3.826921896500066e-06, + "loss": 0.524, + "step": 1326 + }, + { + "epoch": 0.1595623158780737, + "grad_norm": 1.9486198651978224, + "learning_rate": 3.826604773807678e-06, + "loss": 0.792, + "step": 1327 + }, + { + "epoch": 0.1596825587687128, + "grad_norm": 2.905695380336787, + "learning_rate": 3.826287374020798e-06, + "loss": 0.754, + "step": 1328 + }, + { + "epoch": 0.1598028016593519, + "grad_norm": 2.699458816827469, + "learning_rate": 3.825969697187575e-06, + "loss": 0.8229, + "step": 1329 + }, + { + "epoch": 0.15992304454999098, + "grad_norm": 1.82243112546449, + "learning_rate": 3.8256517433562015e-06, + "loss": 0.7014, + "step": 1330 + }, + { + "epoch": 0.16004328744063007, + "grad_norm": 2.284646732225916, + "learning_rate": 3.82533351257491e-06, + "loss": 0.9303, + "step": 1331 + }, + { + "epoch": 0.16016353033126918, + "grad_norm": 1.873127192743219, + "learning_rate": 3.825015004891975e-06, + "loss": 0.8936, + "step": 1332 + }, + { + "epoch": 0.16028377322190826, + "grad_norm": 1.7198316465090369, + "learning_rate": 3.824696220355716e-06, + "loss": 0.7656, + "step": 1333 + }, + { + "epoch": 0.16040401611254734, + "grad_norm": 1.6132060257412992, + "learning_rate": 3.824377159014491e-06, + "loss": 0.7926, + "step": 1334 + }, + { + "epoch": 0.16052425900318643, + "grad_norm": 1.98184643741983, + "learning_rate": 3.824057820916702e-06, + "loss": 0.8568, + "step": 1335 + }, + { + "epoch": 0.16064450189382554, + "grad_norm": 2.6583931747764473, + "learning_rate": 3.8237382061107904e-06, + "loss": 0.7256, + "step": 1336 + }, + { + "epoch": 0.16076474478446462, + "grad_norm": 2.1889443125326737, + "learning_rate": 3.823418314645243e-06, + "loss": 0.7963, + "step": 1337 + }, + { + "epoch": 0.1608849876751037, + "grad_norm": 1.8454970794795182, + "learning_rate": 3.823098146568588e-06, + "loss": 0.757, + "step": 1338 + }, + { + "epoch": 0.1610052305657428, + "grad_norm": 1.9044792831307638, + "learning_rate": 3.822777701929394e-06, + "loss": 0.7317, + "step": 1339 + }, + { + "epoch": 0.1611254734563819, + "grad_norm": 1.9151603197364362, + "learning_rate": 3.8224569807762714e-06, + "loss": 0.741, + "step": 1340 + }, + { + "epoch": 0.16124571634702098, + "grad_norm": 2.573989140135672, + "learning_rate": 3.822135983157873e-06, + "loss": 0.7707, + "step": 1341 + }, + { + "epoch": 0.16136595923766006, + "grad_norm": 1.9702464879932426, + "learning_rate": 3.821814709122896e-06, + "loss": 0.8476, + "step": 1342 + }, + { + "epoch": 0.16148620212829917, + "grad_norm": 2.088521892298617, + "learning_rate": 3.821493158720076e-06, + "loss": 0.8531, + "step": 1343 + }, + { + "epoch": 0.16160644501893826, + "grad_norm": 5.2664205702895535, + "learning_rate": 3.821171331998191e-06, + "loss": 0.7505, + "step": 1344 + }, + { + "epoch": 0.16172668790957734, + "grad_norm": 0.7176116968731532, + "learning_rate": 3.820849229006064e-06, + "loss": 0.5453, + "step": 1345 + }, + { + "epoch": 0.16184693080021645, + "grad_norm": 1.9637282980163462, + "learning_rate": 3.8205268497925564e-06, + "loss": 0.7297, + "step": 1346 + }, + { + "epoch": 0.16196717369085553, + "grad_norm": 2.226171091042424, + "learning_rate": 3.8202041944065725e-06, + "loss": 0.7905, + "step": 1347 + }, + { + "epoch": 0.16208741658149461, + "grad_norm": 2.025406987299807, + "learning_rate": 3.819881262897061e-06, + "loss": 0.7426, + "step": 1348 + }, + { + "epoch": 0.1622076594721337, + "grad_norm": 1.8770273057011335, + "learning_rate": 3.819558055313008e-06, + "loss": 0.7441, + "step": 1349 + }, + { + "epoch": 0.1623279023627728, + "grad_norm": 2.654551602590432, + "learning_rate": 3.819234571703444e-06, + "loss": 0.7807, + "step": 1350 + }, + { + "epoch": 0.1624481452534119, + "grad_norm": 1.9601598770758777, + "learning_rate": 3.8189108121174435e-06, + "loss": 0.8645, + "step": 1351 + }, + { + "epoch": 0.16256838814405097, + "grad_norm": 1.6652200460656386, + "learning_rate": 3.818586776604118e-06, + "loss": 0.844, + "step": 1352 + }, + { + "epoch": 0.16268863103469008, + "grad_norm": 2.334371494765503, + "learning_rate": 3.818262465212625e-06, + "loss": 0.6238, + "step": 1353 + }, + { + "epoch": 0.16280887392532917, + "grad_norm": 1.9609930022998474, + "learning_rate": 3.817937877992161e-06, + "loss": 0.7706, + "step": 1354 + }, + { + "epoch": 0.16292911681596825, + "grad_norm": 2.308868650021954, + "learning_rate": 3.817613014991967e-06, + "loss": 0.8765, + "step": 1355 + }, + { + "epoch": 0.16304935970660733, + "grad_norm": 2.04408376692285, + "learning_rate": 3.817287876261323e-06, + "loss": 0.7773, + "step": 1356 + }, + { + "epoch": 0.16316960259724644, + "grad_norm": 2.0720279136603814, + "learning_rate": 3.816962461849553e-06, + "loss": 0.8087, + "step": 1357 + }, + { + "epoch": 0.16328984548788553, + "grad_norm": 1.7724110325705167, + "learning_rate": 3.8166367718060235e-06, + "loss": 0.8525, + "step": 1358 + }, + { + "epoch": 0.1634100883785246, + "grad_norm": 2.6487784827277503, + "learning_rate": 3.816310806180139e-06, + "loss": 0.7731, + "step": 1359 + }, + { + "epoch": 0.16353033126916372, + "grad_norm": 1.964857340286002, + "learning_rate": 3.81598456502135e-06, + "loss": 0.8159, + "step": 1360 + }, + { + "epoch": 0.1636505741598028, + "grad_norm": 2.0324501600172393, + "learning_rate": 3.8156580483791455e-06, + "loss": 0.8774, + "step": 1361 + }, + { + "epoch": 0.16377081705044189, + "grad_norm": 2.3009047677477095, + "learning_rate": 3.815331256303059e-06, + "loss": 0.7783, + "step": 1362 + }, + { + "epoch": 0.163891059941081, + "grad_norm": 2.208372967442338, + "learning_rate": 3.815004188842665e-06, + "loss": 0.7816, + "step": 1363 + }, + { + "epoch": 0.16401130283172008, + "grad_norm": 1.579790728145889, + "learning_rate": 3.814676846047578e-06, + "loss": 0.8026, + "step": 1364 + }, + { + "epoch": 0.16413154572235916, + "grad_norm": 1.7752418245880197, + "learning_rate": 3.8143492279674565e-06, + "loss": 0.7037, + "step": 1365 + }, + { + "epoch": 0.16425178861299825, + "grad_norm": 0.8943122362443968, + "learning_rate": 3.8140213346519997e-06, + "loss": 0.5848, + "step": 1366 + }, + { + "epoch": 0.16437203150363736, + "grad_norm": 1.797369847367285, + "learning_rate": 3.813693166150948e-06, + "loss": 0.7797, + "step": 1367 + }, + { + "epoch": 0.16449227439427644, + "grad_norm": 3.212952216117476, + "learning_rate": 3.813364722514086e-06, + "loss": 0.8688, + "step": 1368 + }, + { + "epoch": 0.16461251728491552, + "grad_norm": 2.0703682370150536, + "learning_rate": 3.8130360037912368e-06, + "loss": 0.818, + "step": 1369 + }, + { + "epoch": 0.16473276017555463, + "grad_norm": 2.6513400306505166, + "learning_rate": 3.812707010032268e-06, + "loss": 0.8285, + "step": 1370 + }, + { + "epoch": 0.16485300306619372, + "grad_norm": 1.947887325235876, + "learning_rate": 3.8123777412870863e-06, + "loss": 0.7999, + "step": 1371 + }, + { + "epoch": 0.1649732459568328, + "grad_norm": 2.05356512127968, + "learning_rate": 3.812048197605643e-06, + "loss": 0.7902, + "step": 1372 + }, + { + "epoch": 0.16509348884747188, + "grad_norm": 2.128926856264952, + "learning_rate": 3.8117183790379277e-06, + "loss": 0.824, + "step": 1373 + }, + { + "epoch": 0.165213731738111, + "grad_norm": 2.7551047023809354, + "learning_rate": 3.811388285633976e-06, + "loss": 0.9604, + "step": 1374 + }, + { + "epoch": 0.16533397462875007, + "grad_norm": 1.9545762430633167, + "learning_rate": 3.811057917443861e-06, + "loss": 0.6269, + "step": 1375 + }, + { + "epoch": 0.16545421751938916, + "grad_norm": 0.8780024650035135, + "learning_rate": 3.8107272745177e-06, + "loss": 0.6844, + "step": 1376 + }, + { + "epoch": 0.16557446041002827, + "grad_norm": 1.7245419792580483, + "learning_rate": 3.8103963569056513e-06, + "loss": 0.7995, + "step": 1377 + }, + { + "epoch": 0.16569470330066735, + "grad_norm": 1.9880736716683094, + "learning_rate": 3.8100651646579146e-06, + "loss": 0.885, + "step": 1378 + }, + { + "epoch": 0.16581494619130643, + "grad_norm": 1.9394045667149429, + "learning_rate": 3.8097336978247317e-06, + "loss": 0.9254, + "step": 1379 + }, + { + "epoch": 0.16593518908194552, + "grad_norm": 2.3240809808498706, + "learning_rate": 3.8094019564563854e-06, + "loss": 0.9052, + "step": 1380 + }, + { + "epoch": 0.16605543197258463, + "grad_norm": 2.2773866731187584, + "learning_rate": 3.809069940603201e-06, + "loss": 0.7539, + "step": 1381 + }, + { + "epoch": 0.1661756748632237, + "grad_norm": 3.6156764654059184, + "learning_rate": 3.8087376503155452e-06, + "loss": 0.7907, + "step": 1382 + }, + { + "epoch": 0.1662959177538628, + "grad_norm": 0.9128451269496627, + "learning_rate": 3.808405085643826e-06, + "loss": 0.564, + "step": 1383 + }, + { + "epoch": 0.1664161606445019, + "grad_norm": 2.082100070868027, + "learning_rate": 3.8080722466384925e-06, + "loss": 0.894, + "step": 1384 + }, + { + "epoch": 0.166536403535141, + "grad_norm": 2.124483382312894, + "learning_rate": 3.8077391333500376e-06, + "loss": 0.7178, + "step": 1385 + }, + { + "epoch": 0.16665664642578007, + "grad_norm": 1.5889218068362414, + "learning_rate": 3.8074057458289934e-06, + "loss": 0.7778, + "step": 1386 + }, + { + "epoch": 0.16677688931641918, + "grad_norm": 2.306290098425808, + "learning_rate": 3.807072084125934e-06, + "loss": 0.8383, + "step": 1387 + }, + { + "epoch": 0.16689713220705826, + "grad_norm": 2.291882129134842, + "learning_rate": 3.806738148291477e-06, + "loss": 0.8209, + "step": 1388 + }, + { + "epoch": 0.16701737509769735, + "grad_norm": 2.0675484109397755, + "learning_rate": 3.8064039383762793e-06, + "loss": 0.7182, + "step": 1389 + }, + { + "epoch": 0.16713761798833643, + "grad_norm": 2.323029747735122, + "learning_rate": 3.8060694544310396e-06, + "loss": 0.7792, + "step": 1390 + }, + { + "epoch": 0.16725786087897554, + "grad_norm": 1.9355637504388585, + "learning_rate": 3.8057346965065006e-06, + "loss": 0.7955, + "step": 1391 + }, + { + "epoch": 0.16737810376961462, + "grad_norm": 1.626068350918974, + "learning_rate": 3.805399664653443e-06, + "loss": 0.8499, + "step": 1392 + }, + { + "epoch": 0.1674983466602537, + "grad_norm": 2.3443429832540126, + "learning_rate": 3.805064358922692e-06, + "loss": 0.7551, + "step": 1393 + }, + { + "epoch": 0.16761858955089282, + "grad_norm": 1.7217381943981607, + "learning_rate": 3.8047287793651136e-06, + "loss": 0.8232, + "step": 1394 + }, + { + "epoch": 0.1677388324415319, + "grad_norm": 2.024468472269905, + "learning_rate": 3.8043929260316137e-06, + "loss": 0.8909, + "step": 1395 + }, + { + "epoch": 0.16785907533217098, + "grad_norm": 2.067944248171043, + "learning_rate": 3.8040567989731417e-06, + "loss": 0.8415, + "step": 1396 + }, + { + "epoch": 0.16797931822281006, + "grad_norm": 2.011946285675965, + "learning_rate": 3.8037203982406876e-06, + "loss": 0.807, + "step": 1397 + }, + { + "epoch": 0.16809956111344918, + "grad_norm": 1.8953013948899853, + "learning_rate": 3.8033837238852835e-06, + "loss": 0.7377, + "step": 1398 + }, + { + "epoch": 0.16821980400408826, + "grad_norm": 2.396506460969438, + "learning_rate": 3.8030467759580017e-06, + "loss": 0.7029, + "step": 1399 + }, + { + "epoch": 0.16834004689472734, + "grad_norm": 2.1384341869916703, + "learning_rate": 3.802709554509958e-06, + "loss": 0.8733, + "step": 1400 + }, + { + "epoch": 0.16846028978536645, + "grad_norm": 1.670415865462693, + "learning_rate": 3.8023720595923083e-06, + "loss": 0.7991, + "step": 1401 + }, + { + "epoch": 0.16858053267600553, + "grad_norm": 2.2650821430056323, + "learning_rate": 3.80203429125625e-06, + "loss": 0.8878, + "step": 1402 + }, + { + "epoch": 0.16870077556664462, + "grad_norm": 1.691214958104197, + "learning_rate": 3.8016962495530225e-06, + "loss": 0.7105, + "step": 1403 + }, + { + "epoch": 0.1688210184572837, + "grad_norm": 2.481884553804936, + "learning_rate": 3.8013579345339063e-06, + "loss": 0.7835, + "step": 1404 + }, + { + "epoch": 0.1689412613479228, + "grad_norm": 2.471019908514528, + "learning_rate": 3.801019346250224e-06, + "loss": 0.7054, + "step": 1405 + }, + { + "epoch": 0.1690615042385619, + "grad_norm": 2.760909980502927, + "learning_rate": 3.8006804847533395e-06, + "loss": 0.8432, + "step": 1406 + }, + { + "epoch": 0.16918174712920098, + "grad_norm": 2.2563974770826527, + "learning_rate": 3.8003413500946556e-06, + "loss": 0.8643, + "step": 1407 + }, + { + "epoch": 0.1693019900198401, + "grad_norm": 3.721813318620657, + "learning_rate": 3.8000019423256216e-06, + "loss": 0.8435, + "step": 1408 + }, + { + "epoch": 0.16942223291047917, + "grad_norm": 1.6547242100203337, + "learning_rate": 3.7996622614977234e-06, + "loss": 0.8825, + "step": 1409 + }, + { + "epoch": 0.16954247580111825, + "grad_norm": 1.7598855912346205, + "learning_rate": 3.799322307662492e-06, + "loss": 0.7993, + "step": 1410 + }, + { + "epoch": 0.16966271869175734, + "grad_norm": 2.693570135001885, + "learning_rate": 3.798982080871496e-06, + "loss": 0.8462, + "step": 1411 + }, + { + "epoch": 0.16978296158239645, + "grad_norm": 2.4448471845367563, + "learning_rate": 3.798641581176349e-06, + "loss": 0.683, + "step": 1412 + }, + { + "epoch": 0.16990320447303553, + "grad_norm": 1.825916619234495, + "learning_rate": 3.7983008086287044e-06, + "loss": 0.7552, + "step": 1413 + }, + { + "epoch": 0.1700234473636746, + "grad_norm": 1.9553098131636586, + "learning_rate": 3.797959763280257e-06, + "loss": 0.8032, + "step": 1414 + }, + { + "epoch": 0.17014369025431372, + "grad_norm": 6.384665548217671, + "learning_rate": 3.797618445182743e-06, + "loss": 0.7953, + "step": 1415 + }, + { + "epoch": 0.1702639331449528, + "grad_norm": 2.044312598577308, + "learning_rate": 3.79727685438794e-06, + "loss": 0.8517, + "step": 1416 + }, + { + "epoch": 0.1703841760355919, + "grad_norm": 0.8457083404591635, + "learning_rate": 3.796934990947667e-06, + "loss": 0.6174, + "step": 1417 + }, + { + "epoch": 0.170504418926231, + "grad_norm": 0.8845601295833487, + "learning_rate": 3.7965928549137854e-06, + "loss": 0.6261, + "step": 1418 + }, + { + "epoch": 0.17062466181687008, + "grad_norm": 1.944284312905497, + "learning_rate": 3.7962504463381953e-06, + "loss": 0.7865, + "step": 1419 + }, + { + "epoch": 0.17074490470750917, + "grad_norm": 2.56961430397708, + "learning_rate": 3.7959077652728412e-06, + "loss": 0.7882, + "step": 1420 + }, + { + "epoch": 0.17086514759814825, + "grad_norm": 1.9422763893121, + "learning_rate": 3.795564811769707e-06, + "loss": 0.7842, + "step": 1421 + }, + { + "epoch": 0.17098539048878736, + "grad_norm": 2.151343693130905, + "learning_rate": 3.795221585880818e-06, + "loss": 0.7895, + "step": 1422 + }, + { + "epoch": 0.17110563337942644, + "grad_norm": 1.9023990192256803, + "learning_rate": 3.794878087658242e-06, + "loss": 0.9199, + "step": 1423 + }, + { + "epoch": 0.17122587627006552, + "grad_norm": 1.853309547536973, + "learning_rate": 3.7945343171540873e-06, + "loss": 0.7937, + "step": 1424 + }, + { + "epoch": 0.17134611916070464, + "grad_norm": 1.8600484445887535, + "learning_rate": 3.7941902744205033e-06, + "loss": 0.7943, + "step": 1425 + }, + { + "epoch": 0.17146636205134372, + "grad_norm": 2.040684781798039, + "learning_rate": 3.7938459595096817e-06, + "loss": 0.8428, + "step": 1426 + }, + { + "epoch": 0.1715866049419828, + "grad_norm": 1.925595164255141, + "learning_rate": 3.7935013724738545e-06, + "loss": 0.8709, + "step": 1427 + }, + { + "epoch": 0.17170684783262188, + "grad_norm": 1.757416337193813, + "learning_rate": 3.7931565133652945e-06, + "loss": 0.7841, + "step": 1428 + }, + { + "epoch": 0.171827090723261, + "grad_norm": 2.612936638660614, + "learning_rate": 3.792811382236317e-06, + "loss": 0.6906, + "step": 1429 + }, + { + "epoch": 0.17194733361390008, + "grad_norm": 2.021373977963767, + "learning_rate": 3.792465979139279e-06, + "loss": 0.7884, + "step": 1430 + }, + { + "epoch": 0.17206757650453916, + "grad_norm": 0.9327167238175187, + "learning_rate": 3.792120304126576e-06, + "loss": 0.6572, + "step": 1431 + }, + { + "epoch": 0.17218781939517827, + "grad_norm": 1.8141564659470863, + "learning_rate": 3.791774357250649e-06, + "loss": 0.8413, + "step": 1432 + }, + { + "epoch": 0.17230806228581735, + "grad_norm": 2.8820521894993587, + "learning_rate": 3.7914281385639757e-06, + "loss": 0.7982, + "step": 1433 + }, + { + "epoch": 0.17242830517645644, + "grad_norm": 1.9566691476696776, + "learning_rate": 3.7910816481190784e-06, + "loss": 0.8015, + "step": 1434 + }, + { + "epoch": 0.17254854806709552, + "grad_norm": 1.843518517397321, + "learning_rate": 3.7907348859685193e-06, + "loss": 0.7509, + "step": 1435 + }, + { + "epoch": 0.17266879095773463, + "grad_norm": 1.928349451884057, + "learning_rate": 3.790387852164902e-06, + "loss": 0.8152, + "step": 1436 + }, + { + "epoch": 0.1727890338483737, + "grad_norm": 1.7019575124652966, + "learning_rate": 3.7900405467608707e-06, + "loss": 0.7789, + "step": 1437 + }, + { + "epoch": 0.1729092767390128, + "grad_norm": 2.9815711786095656, + "learning_rate": 3.7896929698091114e-06, + "loss": 0.7973, + "step": 1438 + }, + { + "epoch": 0.1730295196296519, + "grad_norm": 2.4747188215733944, + "learning_rate": 3.7893451213623518e-06, + "loss": 0.6953, + "step": 1439 + }, + { + "epoch": 0.173149762520291, + "grad_norm": 1.9617205268490248, + "learning_rate": 3.7889970014733606e-06, + "loss": 0.8357, + "step": 1440 + }, + { + "epoch": 0.17327000541093007, + "grad_norm": 1.759274565572501, + "learning_rate": 3.7886486101949463e-06, + "loss": 0.7926, + "step": 1441 + }, + { + "epoch": 0.17339024830156918, + "grad_norm": 2.3780917071344208, + "learning_rate": 3.7882999475799594e-06, + "loss": 0.8866, + "step": 1442 + }, + { + "epoch": 0.17351049119220827, + "grad_norm": 2.107318695342552, + "learning_rate": 3.787951013681293e-06, + "loss": 0.82, + "step": 1443 + }, + { + "epoch": 0.17363073408284735, + "grad_norm": 1.9924175653193021, + "learning_rate": 3.787601808551879e-06, + "loss": 0.7858, + "step": 1444 + }, + { + "epoch": 0.17375097697348643, + "grad_norm": 2.248834626077595, + "learning_rate": 3.7872523322446926e-06, + "loss": 0.8534, + "step": 1445 + }, + { + "epoch": 0.17387121986412554, + "grad_norm": 2.545062397190061, + "learning_rate": 3.7869025848127478e-06, + "loss": 0.612, + "step": 1446 + }, + { + "epoch": 0.17399146275476463, + "grad_norm": 2.699125914902506, + "learning_rate": 3.786552566309102e-06, + "loss": 0.8124, + "step": 1447 + }, + { + "epoch": 0.1741117056454037, + "grad_norm": 2.2800277152802964, + "learning_rate": 3.7862022767868517e-06, + "loss": 0.8718, + "step": 1448 + }, + { + "epoch": 0.17423194853604282, + "grad_norm": 1.9607286076018928, + "learning_rate": 3.7858517162991367e-06, + "loss": 0.8518, + "step": 1449 + }, + { + "epoch": 0.1743521914266819, + "grad_norm": 2.2241146083364556, + "learning_rate": 3.7855008848991363e-06, + "loss": 0.6172, + "step": 1450 + }, + { + "epoch": 0.17447243431732098, + "grad_norm": 1.8239004114678956, + "learning_rate": 3.7851497826400714e-06, + "loss": 0.7858, + "step": 1451 + }, + { + "epoch": 0.17459267720796007, + "grad_norm": 1.7478728990982089, + "learning_rate": 3.7847984095752034e-06, + "loss": 0.7636, + "step": 1452 + }, + { + "epoch": 0.17471292009859918, + "grad_norm": 1.7922157584692182, + "learning_rate": 3.784446765757836e-06, + "loss": 0.8096, + "step": 1453 + }, + { + "epoch": 0.17483316298923826, + "grad_norm": 2.1483995468948844, + "learning_rate": 3.7840948512413133e-06, + "loss": 0.7808, + "step": 1454 + }, + { + "epoch": 0.17495340587987734, + "grad_norm": 1.8558001734228142, + "learning_rate": 3.7837426660790196e-06, + "loss": 0.7986, + "step": 1455 + }, + { + "epoch": 0.17507364877051645, + "grad_norm": 1.9410375479083697, + "learning_rate": 3.783390210324382e-06, + "loss": 0.8293, + "step": 1456 + }, + { + "epoch": 0.17519389166115554, + "grad_norm": 1.9139099683125198, + "learning_rate": 3.7830374840308676e-06, + "loss": 0.7303, + "step": 1457 + }, + { + "epoch": 0.17531413455179462, + "grad_norm": 2.357099838194555, + "learning_rate": 3.7826844872519842e-06, + "loss": 0.8518, + "step": 1458 + }, + { + "epoch": 0.1754343774424337, + "grad_norm": 2.458950827250802, + "learning_rate": 3.782331220041282e-06, + "loss": 0.7382, + "step": 1459 + }, + { + "epoch": 0.17555462033307281, + "grad_norm": 2.3226364263207597, + "learning_rate": 3.7819776824523504e-06, + "loss": 0.8354, + "step": 1460 + }, + { + "epoch": 0.1756748632237119, + "grad_norm": 2.232109361669722, + "learning_rate": 3.7816238745388213e-06, + "loss": 0.8457, + "step": 1461 + }, + { + "epoch": 0.17579510611435098, + "grad_norm": 3.1101657494536696, + "learning_rate": 3.781269796354367e-06, + "loss": 0.8748, + "step": 1462 + }, + { + "epoch": 0.1759153490049901, + "grad_norm": 1.8425528308845938, + "learning_rate": 3.7809154479527006e-06, + "loss": 0.8691, + "step": 1463 + }, + { + "epoch": 0.17603559189562917, + "grad_norm": 2.0603113989043798, + "learning_rate": 3.780560829387577e-06, + "loss": 0.8429, + "step": 1464 + }, + { + "epoch": 0.17615583478626826, + "grad_norm": 0.8594839200440224, + "learning_rate": 3.7802059407127915e-06, + "loss": 0.5798, + "step": 1465 + }, + { + "epoch": 0.17627607767690734, + "grad_norm": 2.009255451069218, + "learning_rate": 3.7798507819821797e-06, + "loss": 0.8745, + "step": 1466 + }, + { + "epoch": 0.17639632056754645, + "grad_norm": 2.3602649955358004, + "learning_rate": 3.7794953532496197e-06, + "loss": 0.8036, + "step": 1467 + }, + { + "epoch": 0.17651656345818553, + "grad_norm": 0.8870583778682818, + "learning_rate": 3.7791396545690295e-06, + "loss": 0.5802, + "step": 1468 + }, + { + "epoch": 0.17663680634882462, + "grad_norm": 2.2133086197787057, + "learning_rate": 3.7787836859943685e-06, + "loss": 0.8257, + "step": 1469 + }, + { + "epoch": 0.17675704923946373, + "grad_norm": 2.586171580435244, + "learning_rate": 3.7784274475796363e-06, + "loss": 0.797, + "step": 1470 + }, + { + "epoch": 0.1768772921301028, + "grad_norm": 2.2919723191312364, + "learning_rate": 3.7780709393788745e-06, + "loss": 0.7617, + "step": 1471 + }, + { + "epoch": 0.1769975350207419, + "grad_norm": 2.291399599176868, + "learning_rate": 3.777714161446165e-06, + "loss": 0.7618, + "step": 1472 + }, + { + "epoch": 0.177117777911381, + "grad_norm": 2.013607938030784, + "learning_rate": 3.7773571138356304e-06, + "loss": 0.7091, + "step": 1473 + }, + { + "epoch": 0.17723802080202009, + "grad_norm": 2.461175719341931, + "learning_rate": 3.776999796601435e-06, + "loss": 0.8976, + "step": 1474 + }, + { + "epoch": 0.17735826369265917, + "grad_norm": 1.9622004723822535, + "learning_rate": 3.776642209797783e-06, + "loss": 0.7314, + "step": 1475 + }, + { + "epoch": 0.17747850658329825, + "grad_norm": 2.9459583980733184, + "learning_rate": 3.7762843534789205e-06, + "loss": 0.7844, + "step": 1476 + }, + { + "epoch": 0.17759874947393736, + "grad_norm": 2.3544917824553777, + "learning_rate": 3.7759262276991343e-06, + "loss": 0.8919, + "step": 1477 + }, + { + "epoch": 0.17771899236457644, + "grad_norm": 2.3108959305166614, + "learning_rate": 3.7755678325127506e-06, + "loss": 0.8209, + "step": 1478 + }, + { + "epoch": 0.17783923525521553, + "grad_norm": 2.14993622263861, + "learning_rate": 3.7752091679741393e-06, + "loss": 0.7602, + "step": 1479 + }, + { + "epoch": 0.17795947814585464, + "grad_norm": 3.058962121365469, + "learning_rate": 3.774850234137708e-06, + "loss": 0.7922, + "step": 1480 + }, + { + "epoch": 0.17807972103649372, + "grad_norm": 2.110437648276002, + "learning_rate": 3.7744910310579076e-06, + "loss": 0.831, + "step": 1481 + }, + { + "epoch": 0.1781999639271328, + "grad_norm": 2.0372356547553823, + "learning_rate": 3.774131558789229e-06, + "loss": 0.8656, + "step": 1482 + }, + { + "epoch": 0.1783202068177719, + "grad_norm": 2.356892068740919, + "learning_rate": 3.773771817386203e-06, + "loss": 0.7154, + "step": 1483 + }, + { + "epoch": 0.178440449708411, + "grad_norm": 1.6390652462888098, + "learning_rate": 3.773411806903403e-06, + "loss": 0.7981, + "step": 1484 + }, + { + "epoch": 0.17856069259905008, + "grad_norm": 1.717440088662525, + "learning_rate": 3.7730515273954415e-06, + "loss": 0.9561, + "step": 1485 + }, + { + "epoch": 0.17868093548968916, + "grad_norm": 2.7609089442418733, + "learning_rate": 3.772690978916973e-06, + "loss": 0.8516, + "step": 1486 + }, + { + "epoch": 0.17880117838032827, + "grad_norm": 2.1165147676806564, + "learning_rate": 3.772330161522693e-06, + "loss": 0.8777, + "step": 1487 + }, + { + "epoch": 0.17892142127096736, + "grad_norm": 2.488056549993504, + "learning_rate": 3.7719690752673365e-06, + "loss": 0.8091, + "step": 1488 + }, + { + "epoch": 0.17904166416160644, + "grad_norm": 1.962974250492429, + "learning_rate": 3.7716077202056796e-06, + "loss": 0.7946, + "step": 1489 + }, + { + "epoch": 0.17916190705224552, + "grad_norm": 3.0400497284385652, + "learning_rate": 3.7712460963925404e-06, + "loss": 0.9418, + "step": 1490 + }, + { + "epoch": 0.17928214994288463, + "grad_norm": 2.0139344688899166, + "learning_rate": 3.7708842038827775e-06, + "loss": 0.7545, + "step": 1491 + }, + { + "epoch": 0.17940239283352372, + "grad_norm": 2.0199471198610404, + "learning_rate": 3.770522042731288e-06, + "loss": 0.8637, + "step": 1492 + }, + { + "epoch": 0.1795226357241628, + "grad_norm": 1.8382923364203654, + "learning_rate": 3.7701596129930122e-06, + "loss": 0.8904, + "step": 1493 + }, + { + "epoch": 0.1796428786148019, + "grad_norm": 2.1484408291877615, + "learning_rate": 3.7697969147229315e-06, + "loss": 0.7456, + "step": 1494 + }, + { + "epoch": 0.179763121505441, + "grad_norm": 1.878357614847117, + "learning_rate": 3.7694339479760647e-06, + "loss": 0.8592, + "step": 1495 + }, + { + "epoch": 0.17988336439608008, + "grad_norm": 0.792594783903059, + "learning_rate": 3.769070712807476e-06, + "loss": 0.5734, + "step": 1496 + }, + { + "epoch": 0.18000360728671919, + "grad_norm": 1.7401824498163063, + "learning_rate": 3.768707209272266e-06, + "loss": 0.798, + "step": 1497 + }, + { + "epoch": 0.18012385017735827, + "grad_norm": 2.0905239246151734, + "learning_rate": 3.768343437425579e-06, + "loss": 0.7764, + "step": 1498 + }, + { + "epoch": 0.18024409306799735, + "grad_norm": 2.360687699095052, + "learning_rate": 3.7679793973225987e-06, + "loss": 0.8638, + "step": 1499 + }, + { + "epoch": 0.18036433595863643, + "grad_norm": 0.8553382390769361, + "learning_rate": 3.767615089018549e-06, + "loss": 0.6171, + "step": 1500 + }, + { + "epoch": 0.18048457884927555, + "grad_norm": 2.185452947566946, + "learning_rate": 3.7672505125686966e-06, + "loss": 0.8711, + "step": 1501 + }, + { + "epoch": 0.18060482173991463, + "grad_norm": 2.787703495552167, + "learning_rate": 3.7668856680283455e-06, + "loss": 0.8585, + "step": 1502 + }, + { + "epoch": 0.1807250646305537, + "grad_norm": 1.85905511492636, + "learning_rate": 3.7665205554528437e-06, + "loss": 0.8373, + "step": 1503 + }, + { + "epoch": 0.18084530752119282, + "grad_norm": 2.074933577244242, + "learning_rate": 3.7661551748975782e-06, + "loss": 0.7464, + "step": 1504 + }, + { + "epoch": 0.1809655504118319, + "grad_norm": 0.8224332371315782, + "learning_rate": 3.7657895264179772e-06, + "loss": 0.6058, + "step": 1505 + }, + { + "epoch": 0.181085793302471, + "grad_norm": 1.6287630965086175, + "learning_rate": 3.765423610069509e-06, + "loss": 0.7461, + "step": 1506 + }, + { + "epoch": 0.18120603619311007, + "grad_norm": 1.8613757740818664, + "learning_rate": 3.765057425907683e-06, + "loss": 0.7298, + "step": 1507 + }, + { + "epoch": 0.18132627908374918, + "grad_norm": 1.9462109065309945, + "learning_rate": 3.764690973988048e-06, + "loss": 0.7952, + "step": 1508 + }, + { + "epoch": 0.18144652197438826, + "grad_norm": 1.7019835680606574, + "learning_rate": 3.7643242543661967e-06, + "loss": 0.742, + "step": 1509 + }, + { + "epoch": 0.18156676486502735, + "grad_norm": 0.8267122489977845, + "learning_rate": 3.7639572670977573e-06, + "loss": 0.6054, + "step": 1510 + }, + { + "epoch": 0.18168700775566646, + "grad_norm": 1.6292600544784184, + "learning_rate": 3.7635900122384042e-06, + "loss": 0.771, + "step": 1511 + }, + { + "epoch": 0.18180725064630554, + "grad_norm": 2.2573157163241615, + "learning_rate": 3.7632224898438477e-06, + "loss": 0.8781, + "step": 1512 + }, + { + "epoch": 0.18192749353694462, + "grad_norm": 1.5306547136643092, + "learning_rate": 3.762854699969842e-06, + "loss": 0.7927, + "step": 1513 + }, + { + "epoch": 0.1820477364275837, + "grad_norm": 1.9354436483100435, + "learning_rate": 3.762486642672179e-06, + "loss": 0.7395, + "step": 1514 + }, + { + "epoch": 0.18216797931822282, + "grad_norm": 1.8694013933864864, + "learning_rate": 3.7621183180066946e-06, + "loss": 0.8793, + "step": 1515 + }, + { + "epoch": 0.1822882222088619, + "grad_norm": 1.6954674715855182, + "learning_rate": 3.7617497260292625e-06, + "loss": 0.7455, + "step": 1516 + }, + { + "epoch": 0.18240846509950098, + "grad_norm": 2.5172326299977503, + "learning_rate": 3.7613808667957967e-06, + "loss": 0.7963, + "step": 1517 + }, + { + "epoch": 0.1825287079901401, + "grad_norm": 1.8376802005768809, + "learning_rate": 3.7610117403622547e-06, + "loss": 0.9157, + "step": 1518 + }, + { + "epoch": 0.18264895088077918, + "grad_norm": 1.683573782463886, + "learning_rate": 3.7606423467846313e-06, + "loss": 0.9094, + "step": 1519 + }, + { + "epoch": 0.18276919377141826, + "grad_norm": 1.5540511012469522, + "learning_rate": 3.760272686118964e-06, + "loss": 0.8049, + "step": 1520 + }, + { + "epoch": 0.18288943666205737, + "grad_norm": 1.9911988441807238, + "learning_rate": 3.7599027584213297e-06, + "loss": 0.9335, + "step": 1521 + }, + { + "epoch": 0.18300967955269645, + "grad_norm": 1.9218683010077608, + "learning_rate": 3.7595325637478465e-06, + "loss": 0.7877, + "step": 1522 + }, + { + "epoch": 0.18312992244333554, + "grad_norm": 1.7138662137194527, + "learning_rate": 3.7591621021546723e-06, + "loss": 0.8201, + "step": 1523 + }, + { + "epoch": 0.18325016533397462, + "grad_norm": 1.79334296682434, + "learning_rate": 3.7587913736980062e-06, + "loss": 0.8204, + "step": 1524 + }, + { + "epoch": 0.18337040822461373, + "grad_norm": 1.6913029945342966, + "learning_rate": 3.7584203784340865e-06, + "loss": 0.8517, + "step": 1525 + }, + { + "epoch": 0.1834906511152528, + "grad_norm": 2.44876469201222, + "learning_rate": 3.7580491164191938e-06, + "loss": 0.8631, + "step": 1526 + }, + { + "epoch": 0.1836108940058919, + "grad_norm": 0.7539512739424622, + "learning_rate": 3.757677587709648e-06, + "loss": 0.614, + "step": 1527 + }, + { + "epoch": 0.183731136896531, + "grad_norm": 2.1224723817715643, + "learning_rate": 3.7573057923618095e-06, + "loss": 0.76, + "step": 1528 + }, + { + "epoch": 0.1838513797871701, + "grad_norm": 8.061231883071981, + "learning_rate": 3.7569337304320793e-06, + "loss": 0.7517, + "step": 1529 + }, + { + "epoch": 0.18397162267780917, + "grad_norm": 0.8570383035096557, + "learning_rate": 3.756561401976899e-06, + "loss": 0.646, + "step": 1530 + }, + { + "epoch": 0.18409186556844825, + "grad_norm": 1.8561515756323383, + "learning_rate": 3.7561888070527514e-06, + "loss": 0.8396, + "step": 1531 + }, + { + "epoch": 0.18421210845908736, + "grad_norm": 2.0299980413493213, + "learning_rate": 3.7558159457161577e-06, + "loss": 0.8102, + "step": 1532 + }, + { + "epoch": 0.18433235134972645, + "grad_norm": 2.995027235413585, + "learning_rate": 3.755442818023681e-06, + "loss": 0.7942, + "step": 1533 + }, + { + "epoch": 0.18445259424036553, + "grad_norm": 2.189364047794897, + "learning_rate": 3.7550694240319246e-06, + "loss": 0.7695, + "step": 1534 + }, + { + "epoch": 0.18457283713100464, + "grad_norm": 2.361709753202968, + "learning_rate": 3.7546957637975326e-06, + "loss": 0.7829, + "step": 1535 + }, + { + "epoch": 0.18469308002164372, + "grad_norm": 2.397300226920008, + "learning_rate": 3.7543218373771873e-06, + "loss": 0.7477, + "step": 1536 + }, + { + "epoch": 0.1848133229122828, + "grad_norm": 1.3846965821006425, + "learning_rate": 3.753947644827615e-06, + "loss": 0.7816, + "step": 1537 + }, + { + "epoch": 0.1849335658029219, + "grad_norm": 0.9655860485627926, + "learning_rate": 3.753573186205579e-06, + "loss": 0.5728, + "step": 1538 + }, + { + "epoch": 0.185053808693561, + "grad_norm": 2.085490889725762, + "learning_rate": 3.753198461567885e-06, + "loss": 0.784, + "step": 1539 + }, + { + "epoch": 0.18517405158420008, + "grad_norm": 2.1597037919359656, + "learning_rate": 3.7528234709713783e-06, + "loss": 0.923, + "step": 1540 + }, + { + "epoch": 0.18529429447483917, + "grad_norm": 1.9266802254478017, + "learning_rate": 3.7524482144729447e-06, + "loss": 0.8502, + "step": 1541 + }, + { + "epoch": 0.18541453736547828, + "grad_norm": 2.039662707625078, + "learning_rate": 3.7520726921295106e-06, + "loss": 0.8443, + "step": 1542 + }, + { + "epoch": 0.18553478025611736, + "grad_norm": 1.9282262420041363, + "learning_rate": 3.751696903998042e-06, + "loss": 0.7309, + "step": 1543 + }, + { + "epoch": 0.18565502314675644, + "grad_norm": 3.3302794454358304, + "learning_rate": 3.7513208501355456e-06, + "loss": 0.705, + "step": 1544 + }, + { + "epoch": 0.18577526603739553, + "grad_norm": 1.7619344575016798, + "learning_rate": 3.750944530599069e-06, + "loss": 0.8425, + "step": 1545 + }, + { + "epoch": 0.18589550892803464, + "grad_norm": 2.234813407089741, + "learning_rate": 3.7505679454456992e-06, + "loss": 0.8138, + "step": 1546 + }, + { + "epoch": 0.18601575181867372, + "grad_norm": 1.7646948465768901, + "learning_rate": 3.750191094732564e-06, + "loss": 0.7077, + "step": 1547 + }, + { + "epoch": 0.1861359947093128, + "grad_norm": 2.451321078315058, + "learning_rate": 3.7498139785168313e-06, + "loss": 0.7627, + "step": 1548 + }, + { + "epoch": 0.1862562375999519, + "grad_norm": 1.7101018249578717, + "learning_rate": 3.749436596855709e-06, + "loss": 0.7805, + "step": 1549 + }, + { + "epoch": 0.186376480490591, + "grad_norm": 2.1188836429724156, + "learning_rate": 3.749058949806446e-06, + "loss": 0.915, + "step": 1550 + }, + { + "epoch": 0.18649672338123008, + "grad_norm": 1.8581987394963162, + "learning_rate": 3.748681037426331e-06, + "loss": 0.8516, + "step": 1551 + }, + { + "epoch": 0.1866169662718692, + "grad_norm": 2.0386349752360915, + "learning_rate": 3.7483028597726936e-06, + "loss": 0.9236, + "step": 1552 + }, + { + "epoch": 0.18673720916250827, + "grad_norm": 1.7455598497098481, + "learning_rate": 3.7479244169029017e-06, + "loss": 0.6344, + "step": 1553 + }, + { + "epoch": 0.18685745205314735, + "grad_norm": 2.3121078777356723, + "learning_rate": 3.7475457088743658e-06, + "loss": 0.7486, + "step": 1554 + }, + { + "epoch": 0.18697769494378644, + "grad_norm": 1.804450081405124, + "learning_rate": 3.7471667357445348e-06, + "loss": 0.7527, + "step": 1555 + }, + { + "epoch": 0.18709793783442555, + "grad_norm": 1.9582289268852946, + "learning_rate": 3.7467874975709e-06, + "loss": 0.7383, + "step": 1556 + }, + { + "epoch": 0.18721818072506463, + "grad_norm": 2.2248683976247374, + "learning_rate": 3.7464079944109904e-06, + "loss": 0.7968, + "step": 1557 + }, + { + "epoch": 0.18733842361570371, + "grad_norm": 1.8759763083192176, + "learning_rate": 3.746028226322376e-06, + "loss": 0.7853, + "step": 1558 + }, + { + "epoch": 0.18745866650634282, + "grad_norm": 1.7459968693629773, + "learning_rate": 3.745648193362669e-06, + "loss": 0.7606, + "step": 1559 + }, + { + "epoch": 0.1875789093969819, + "grad_norm": 2.4007506507937006, + "learning_rate": 3.745267895589518e-06, + "loss": 0.7329, + "step": 1560 + }, + { + "epoch": 0.187699152287621, + "grad_norm": 2.3464566687162245, + "learning_rate": 3.7448873330606154e-06, + "loss": 0.8287, + "step": 1561 + }, + { + "epoch": 0.18781939517826007, + "grad_norm": 1.99845576274243, + "learning_rate": 3.7445065058336914e-06, + "loss": 0.8817, + "step": 1562 + }, + { + "epoch": 0.18793963806889918, + "grad_norm": 1.9453085780582955, + "learning_rate": 3.7441254139665176e-06, + "loss": 0.8674, + "step": 1563 + }, + { + "epoch": 0.18805988095953827, + "grad_norm": 1.6970607339784283, + "learning_rate": 3.743744057516905e-06, + "loss": 0.8282, + "step": 1564 + }, + { + "epoch": 0.18818012385017735, + "grad_norm": 3.232193510626602, + "learning_rate": 3.743362436542706e-06, + "loss": 0.8864, + "step": 1565 + }, + { + "epoch": 0.18830036674081646, + "grad_norm": 1.7098629856430723, + "learning_rate": 3.7429805511018115e-06, + "loss": 0.7709, + "step": 1566 + }, + { + "epoch": 0.18842060963145554, + "grad_norm": 1.7107665094083222, + "learning_rate": 3.7425984012521524e-06, + "loss": 0.7916, + "step": 1567 + }, + { + "epoch": 0.18854085252209463, + "grad_norm": 0.7519833645375603, + "learning_rate": 3.7422159870517025e-06, + "loss": 0.6059, + "step": 1568 + }, + { + "epoch": 0.1886610954127337, + "grad_norm": 1.5850723988938509, + "learning_rate": 3.7418333085584717e-06, + "loss": 0.7952, + "step": 1569 + }, + { + "epoch": 0.18878133830337282, + "grad_norm": 1.9538769304530645, + "learning_rate": 3.7414503658305128e-06, + "loss": 0.9187, + "step": 1570 + }, + { + "epoch": 0.1889015811940119, + "grad_norm": 2.3189785908923546, + "learning_rate": 3.7410671589259185e-06, + "loss": 0.779, + "step": 1571 + }, + { + "epoch": 0.18902182408465099, + "grad_norm": 2.2796000954786777, + "learning_rate": 3.7406836879028205e-06, + "loss": 0.8048, + "step": 1572 + }, + { + "epoch": 0.1891420669752901, + "grad_norm": 2.0001809515156754, + "learning_rate": 3.7402999528193907e-06, + "loss": 0.7731, + "step": 1573 + }, + { + "epoch": 0.18926230986592918, + "grad_norm": 2.373210634460232, + "learning_rate": 3.739915953733842e-06, + "loss": 0.8651, + "step": 1574 + }, + { + "epoch": 0.18938255275656826, + "grad_norm": 1.6312651534856504, + "learning_rate": 3.7395316907044264e-06, + "loss": 0.8244, + "step": 1575 + }, + { + "epoch": 0.18950279564720737, + "grad_norm": 2.1074329619871706, + "learning_rate": 3.7391471637894364e-06, + "loss": 0.7979, + "step": 1576 + }, + { + "epoch": 0.18962303853784646, + "grad_norm": 1.9383570713525984, + "learning_rate": 3.738762373047205e-06, + "loss": 0.8574, + "step": 1577 + }, + { + "epoch": 0.18974328142848554, + "grad_norm": 1.616983520617287, + "learning_rate": 3.738377318536103e-06, + "loss": 0.8396, + "step": 1578 + }, + { + "epoch": 0.18986352431912462, + "grad_norm": 2.125328823902479, + "learning_rate": 3.7379920003145447e-06, + "loss": 0.7193, + "step": 1579 + }, + { + "epoch": 0.18998376720976373, + "grad_norm": 2.43038331889894, + "learning_rate": 3.7376064184409817e-06, + "loss": 0.8436, + "step": 1580 + }, + { + "epoch": 0.19010401010040281, + "grad_norm": 1.929292337211838, + "learning_rate": 3.7372205729739063e-06, + "loss": 0.878, + "step": 1581 + }, + { + "epoch": 0.1902242529910419, + "grad_norm": 1.9352075489130345, + "learning_rate": 3.7368344639718514e-06, + "loss": 0.7208, + "step": 1582 + }, + { + "epoch": 0.190344495881681, + "grad_norm": 1.5143915249957667, + "learning_rate": 3.7364480914933895e-06, + "loss": 0.8103, + "step": 1583 + }, + { + "epoch": 0.1904647387723201, + "grad_norm": 2.0590876469977664, + "learning_rate": 3.7360614555971325e-06, + "loss": 0.8226, + "step": 1584 + }, + { + "epoch": 0.19058498166295917, + "grad_norm": 2.0913333895805595, + "learning_rate": 3.735674556341733e-06, + "loss": 0.8564, + "step": 1585 + }, + { + "epoch": 0.19070522455359826, + "grad_norm": 2.0828310806121, + "learning_rate": 3.7352873937858835e-06, + "loss": 0.839, + "step": 1586 + }, + { + "epoch": 0.19082546744423737, + "grad_norm": 1.9443887869120182, + "learning_rate": 3.734899967988316e-06, + "loss": 0.7306, + "step": 1587 + }, + { + "epoch": 0.19094571033487645, + "grad_norm": 1.7512545176195229, + "learning_rate": 3.7345122790078026e-06, + "loss": 0.8452, + "step": 1588 + }, + { + "epoch": 0.19106595322551553, + "grad_norm": 2.928649230083079, + "learning_rate": 3.7341243269031556e-06, + "loss": 0.9437, + "step": 1589 + }, + { + "epoch": 0.19118619611615464, + "grad_norm": 1.5927906514915133, + "learning_rate": 3.7337361117332275e-06, + "loss": 0.7814, + "step": 1590 + }, + { + "epoch": 0.19130643900679373, + "grad_norm": 1.9626479269269872, + "learning_rate": 3.7333476335569087e-06, + "loss": 0.7789, + "step": 1591 + }, + { + "epoch": 0.1914266818974328, + "grad_norm": 2.7233135464376286, + "learning_rate": 3.7329588924331325e-06, + "loss": 0.6822, + "step": 1592 + }, + { + "epoch": 0.1915469247880719, + "grad_norm": 1.856031537285418, + "learning_rate": 3.732569888420871e-06, + "loss": 0.8345, + "step": 1593 + }, + { + "epoch": 0.191667167678711, + "grad_norm": 2.039323110478544, + "learning_rate": 3.732180621579134e-06, + "loss": 0.834, + "step": 1594 + }, + { + "epoch": 0.1917874105693501, + "grad_norm": 1.7782803593294711, + "learning_rate": 3.7317910919669745e-06, + "loss": 0.8152, + "step": 1595 + }, + { + "epoch": 0.19190765345998917, + "grad_norm": 3.0403541624665276, + "learning_rate": 3.7314012996434826e-06, + "loss": 0.7744, + "step": 1596 + }, + { + "epoch": 0.19202789635062828, + "grad_norm": 1.9297893049435757, + "learning_rate": 3.7310112446677907e-06, + "loss": 0.8204, + "step": 1597 + }, + { + "epoch": 0.19214813924126736, + "grad_norm": 2.875683910460407, + "learning_rate": 3.7306209270990695e-06, + "loss": 0.6929, + "step": 1598 + }, + { + "epoch": 0.19226838213190645, + "grad_norm": 1.755381552037691, + "learning_rate": 3.7302303469965292e-06, + "loss": 0.8718, + "step": 1599 + }, + { + "epoch": 0.19238862502254553, + "grad_norm": 2.0784854432918394, + "learning_rate": 3.7298395044194206e-06, + "loss": 0.7155, + "step": 1600 + }, + { + "epoch": 0.19250886791318464, + "grad_norm": 1.8692487248937655, + "learning_rate": 3.7294483994270356e-06, + "loss": 0.9495, + "step": 1601 + }, + { + "epoch": 0.19262911080382372, + "grad_norm": 2.0970373963462845, + "learning_rate": 3.7290570320787033e-06, + "loss": 0.789, + "step": 1602 + }, + { + "epoch": 0.1927493536944628, + "grad_norm": 2.0383575507051264, + "learning_rate": 3.728665402433793e-06, + "loss": 0.7253, + "step": 1603 + }, + { + "epoch": 0.19286959658510192, + "grad_norm": 2.4261808450189397, + "learning_rate": 3.7282735105517164e-06, + "loss": 0.8719, + "step": 1604 + }, + { + "epoch": 0.192989839475741, + "grad_norm": 2.444748298467326, + "learning_rate": 3.727881356491922e-06, + "loss": 0.6822, + "step": 1605 + }, + { + "epoch": 0.19311008236638008, + "grad_norm": 1.7789804154165578, + "learning_rate": 3.7274889403139002e-06, + "loss": 0.765, + "step": 1606 + }, + { + "epoch": 0.1932303252570192, + "grad_norm": 2.992503176936835, + "learning_rate": 3.727096262077179e-06, + "loss": 0.7919, + "step": 1607 + }, + { + "epoch": 0.19335056814765827, + "grad_norm": 1.8092888635349043, + "learning_rate": 3.7267033218413285e-06, + "loss": 0.8602, + "step": 1608 + }, + { + "epoch": 0.19347081103829736, + "grad_norm": 2.122308709788365, + "learning_rate": 3.726310119665957e-06, + "loss": 0.8214, + "step": 1609 + }, + { + "epoch": 0.19359105392893644, + "grad_norm": 1.826907999926297, + "learning_rate": 3.725916655610713e-06, + "loss": 0.8649, + "step": 1610 + }, + { + "epoch": 0.19371129681957555, + "grad_norm": 2.400966541055562, + "learning_rate": 3.725522929735284e-06, + "loss": 0.7675, + "step": 1611 + }, + { + "epoch": 0.19383153971021463, + "grad_norm": 2.2158922946177766, + "learning_rate": 3.725128942099399e-06, + "loss": 0.7522, + "step": 1612 + }, + { + "epoch": 0.19395178260085372, + "grad_norm": 1.6690233099642742, + "learning_rate": 3.7247346927628245e-06, + "loss": 0.8046, + "step": 1613 + }, + { + "epoch": 0.19407202549149283, + "grad_norm": 1.9334213025808202, + "learning_rate": 3.7243401817853694e-06, + "loss": 0.8018, + "step": 1614 + }, + { + "epoch": 0.1941922683821319, + "grad_norm": 2.0882413906713206, + "learning_rate": 3.723945409226879e-06, + "loss": 0.7318, + "step": 1615 + }, + { + "epoch": 0.194312511272771, + "grad_norm": 2.224390945227361, + "learning_rate": 3.723550375147241e-06, + "loss": 0.813, + "step": 1616 + }, + { + "epoch": 0.19443275416341008, + "grad_norm": 2.0745833196855177, + "learning_rate": 3.7231550796063816e-06, + "loss": 0.8107, + "step": 1617 + }, + { + "epoch": 0.1945529970540492, + "grad_norm": 2.009828735191087, + "learning_rate": 3.722759522664266e-06, + "loss": 0.6633, + "step": 1618 + }, + { + "epoch": 0.19467323994468827, + "grad_norm": 1.9543619130193228, + "learning_rate": 3.7223637043809016e-06, + "loss": 0.8256, + "step": 1619 + }, + { + "epoch": 0.19479348283532735, + "grad_norm": 3.7322799360194017, + "learning_rate": 3.7219676248163322e-06, + "loss": 0.8679, + "step": 1620 + }, + { + "epoch": 0.19491372572596646, + "grad_norm": 1.7599196625927325, + "learning_rate": 3.721571284030643e-06, + "loss": 0.9408, + "step": 1621 + }, + { + "epoch": 0.19503396861660555, + "grad_norm": 2.067850743793758, + "learning_rate": 3.7211746820839587e-06, + "loss": 0.795, + "step": 1622 + }, + { + "epoch": 0.19515421150724463, + "grad_norm": 1.7281815200585766, + "learning_rate": 3.7207778190364437e-06, + "loss": 0.815, + "step": 1623 + }, + { + "epoch": 0.1952744543978837, + "grad_norm": 1.9796294892564839, + "learning_rate": 3.720380694948302e-06, + "loss": 0.7504, + "step": 1624 + }, + { + "epoch": 0.19539469728852282, + "grad_norm": 1.0439166501341801, + "learning_rate": 3.719983309879777e-06, + "loss": 0.7132, + "step": 1625 + }, + { + "epoch": 0.1955149401791619, + "grad_norm": 1.6724230144441787, + "learning_rate": 3.719585663891151e-06, + "loss": 0.7816, + "step": 1626 + }, + { + "epoch": 0.195635183069801, + "grad_norm": 2.294552621556817, + "learning_rate": 3.719187757042747e-06, + "loss": 0.8028, + "step": 1627 + }, + { + "epoch": 0.1957554259604401, + "grad_norm": 0.7327511060166902, + "learning_rate": 3.7187895893949275e-06, + "loss": 0.5513, + "step": 1628 + }, + { + "epoch": 0.19587566885107918, + "grad_norm": 2.5826906409712427, + "learning_rate": 3.7183911610080937e-06, + "loss": 0.7661, + "step": 1629 + }, + { + "epoch": 0.19599591174171827, + "grad_norm": 2.3197671268594324, + "learning_rate": 3.7179924719426872e-06, + "loss": 0.764, + "step": 1630 + }, + { + "epoch": 0.19611615463235738, + "grad_norm": 2.2874702409986867, + "learning_rate": 3.7175935222591885e-06, + "loss": 0.7658, + "step": 1631 + }, + { + "epoch": 0.19623639752299646, + "grad_norm": 3.5121042211878963, + "learning_rate": 3.717194312018118e-06, + "loss": 0.7458, + "step": 1632 + }, + { + "epoch": 0.19635664041363554, + "grad_norm": 1.9054267856783278, + "learning_rate": 3.716794841280036e-06, + "loss": 0.7668, + "step": 1633 + }, + { + "epoch": 0.19647688330427462, + "grad_norm": 3.8509139623389297, + "learning_rate": 3.7163951101055407e-06, + "loss": 0.787, + "step": 1634 + }, + { + "epoch": 0.19659712619491373, + "grad_norm": 1.8690640265415976, + "learning_rate": 3.715995118555273e-06, + "loss": 0.7982, + "step": 1635 + }, + { + "epoch": 0.19671736908555282, + "grad_norm": 2.054213088351234, + "learning_rate": 3.71559486668991e-06, + "loss": 0.869, + "step": 1636 + }, + { + "epoch": 0.1968376119761919, + "grad_norm": 1.6227628273416543, + "learning_rate": 3.715194354570169e-06, + "loss": 0.7801, + "step": 1637 + }, + { + "epoch": 0.196957854866831, + "grad_norm": 1.91796589091723, + "learning_rate": 3.714793582256809e-06, + "loss": 0.8416, + "step": 1638 + }, + { + "epoch": 0.1970780977574701, + "grad_norm": 2.394629521554787, + "learning_rate": 3.7143925498106253e-06, + "loss": 0.8518, + "step": 1639 + }, + { + "epoch": 0.19719834064810918, + "grad_norm": 1.7795213710555249, + "learning_rate": 3.7139912572924558e-06, + "loss": 0.8027, + "step": 1640 + }, + { + "epoch": 0.19731858353874826, + "grad_norm": 2.3580497404939953, + "learning_rate": 3.7135897047631744e-06, + "loss": 0.815, + "step": 1641 + }, + { + "epoch": 0.19743882642938737, + "grad_norm": 2.1561321281612953, + "learning_rate": 3.713187892283698e-06, + "loss": 0.7672, + "step": 1642 + }, + { + "epoch": 0.19755906932002645, + "grad_norm": 2.320694914364894, + "learning_rate": 3.71278581991498e-06, + "loss": 0.8845, + "step": 1643 + }, + { + "epoch": 0.19767931221066554, + "grad_norm": 1.8869894082358898, + "learning_rate": 3.712383487718015e-06, + "loss": 0.7956, + "step": 1644 + }, + { + "epoch": 0.19779955510130465, + "grad_norm": 7.908168394126007, + "learning_rate": 3.7119808957538365e-06, + "loss": 0.8691, + "step": 1645 + }, + { + "epoch": 0.19791979799194373, + "grad_norm": 2.0872699039673948, + "learning_rate": 3.711578044083517e-06, + "loss": 0.811, + "step": 1646 + }, + { + "epoch": 0.1980400408825828, + "grad_norm": 1.7065485570813366, + "learning_rate": 3.7111749327681698e-06, + "loss": 0.7505, + "step": 1647 + }, + { + "epoch": 0.1981602837732219, + "grad_norm": 2.097874464906757, + "learning_rate": 3.7107715618689455e-06, + "loss": 0.8744, + "step": 1648 + }, + { + "epoch": 0.198280526663861, + "grad_norm": 1.486938511637296, + "learning_rate": 3.710367931447035e-06, + "loss": 0.8401, + "step": 1649 + }, + { + "epoch": 0.1984007695545001, + "grad_norm": 2.3770875289394073, + "learning_rate": 3.70996404156367e-06, + "loss": 0.8747, + "step": 1650 + }, + { + "epoch": 0.19852101244513917, + "grad_norm": 1.7275356812387417, + "learning_rate": 3.7095598922801187e-06, + "loss": 0.7384, + "step": 1651 + }, + { + "epoch": 0.19864125533577828, + "grad_norm": 2.6862459139928188, + "learning_rate": 3.7091554836576914e-06, + "loss": 0.7715, + "step": 1652 + }, + { + "epoch": 0.19876149822641737, + "grad_norm": 1.7728675228828088, + "learning_rate": 3.708750815757736e-06, + "loss": 0.8381, + "step": 1653 + }, + { + "epoch": 0.19888174111705645, + "grad_norm": 2.247258741692157, + "learning_rate": 3.7083458886416407e-06, + "loss": 0.7448, + "step": 1654 + }, + { + "epoch": 0.19900198400769553, + "grad_norm": 1.8803742001848853, + "learning_rate": 3.707940702370832e-06, + "loss": 0.887, + "step": 1655 + }, + { + "epoch": 0.19912222689833464, + "grad_norm": 0.7693167450431561, + "learning_rate": 3.707535257006777e-06, + "loss": 0.5832, + "step": 1656 + }, + { + "epoch": 0.19924246978897373, + "grad_norm": 2.469806107235227, + "learning_rate": 3.707129552610981e-06, + "loss": 0.8967, + "step": 1657 + }, + { + "epoch": 0.1993627126796128, + "grad_norm": 2.297831433520527, + "learning_rate": 3.70672358924499e-06, + "loss": 0.7437, + "step": 1658 + }, + { + "epoch": 0.19948295557025192, + "grad_norm": 1.7817454401961208, + "learning_rate": 3.706317366970386e-06, + "loss": 0.7899, + "step": 1659 + }, + { + "epoch": 0.199603198460891, + "grad_norm": 1.9177374612922713, + "learning_rate": 3.705910885848795e-06, + "loss": 0.8443, + "step": 1660 + }, + { + "epoch": 0.19972344135153008, + "grad_norm": 2.3941590360561262, + "learning_rate": 3.705504145941879e-06, + "loss": 0.8581, + "step": 1661 + }, + { + "epoch": 0.1998436842421692, + "grad_norm": 1.9068093651722615, + "learning_rate": 3.7050971473113403e-06, + "loss": 0.7958, + "step": 1662 + }, + { + "epoch": 0.19996392713280828, + "grad_norm": 1.6829446838091846, + "learning_rate": 3.7046898900189196e-06, + "loss": 0.8045, + "step": 1663 + }, + { + "epoch": 0.20008417002344736, + "grad_norm": 1.6672759214083044, + "learning_rate": 3.704282374126398e-06, + "loss": 0.8397, + "step": 1664 + }, + { + "epoch": 0.20020441291408644, + "grad_norm": 1.7183378245569114, + "learning_rate": 3.7038745996955954e-06, + "loss": 0.8843, + "step": 1665 + }, + { + "epoch": 0.20032465580472555, + "grad_norm": 3.4024447963701325, + "learning_rate": 3.703466566788371e-06, + "loss": 0.7326, + "step": 1666 + }, + { + "epoch": 0.20044489869536464, + "grad_norm": 1.995297267440449, + "learning_rate": 3.703058275466622e-06, + "loss": 0.7537, + "step": 1667 + }, + { + "epoch": 0.20056514158600372, + "grad_norm": 1.6726789591431188, + "learning_rate": 3.7026497257922877e-06, + "loss": 0.7844, + "step": 1668 + }, + { + "epoch": 0.20068538447664283, + "grad_norm": 1.6561538380471545, + "learning_rate": 3.7022409178273436e-06, + "loss": 0.8639, + "step": 1669 + }, + { + "epoch": 0.2008056273672819, + "grad_norm": 1.8135805526441209, + "learning_rate": 3.7018318516338054e-06, + "loss": 0.785, + "step": 1670 + }, + { + "epoch": 0.200925870257921, + "grad_norm": 2.4696709568918167, + "learning_rate": 3.7014225272737284e-06, + "loss": 0.828, + "step": 1671 + }, + { + "epoch": 0.20104611314856008, + "grad_norm": 2.3012539389357456, + "learning_rate": 3.701012944809207e-06, + "loss": 0.7517, + "step": 1672 + }, + { + "epoch": 0.2011663560391992, + "grad_norm": 1.994817336857683, + "learning_rate": 3.700603104302374e-06, + "loss": 0.7992, + "step": 1673 + }, + { + "epoch": 0.20128659892983827, + "grad_norm": 0.9044294157180273, + "learning_rate": 3.7001930058154027e-06, + "loss": 0.5604, + "step": 1674 + }, + { + "epoch": 0.20140684182047736, + "grad_norm": 2.8242358026477747, + "learning_rate": 3.6997826494105037e-06, + "loss": 0.8105, + "step": 1675 + }, + { + "epoch": 0.20152708471111647, + "grad_norm": 2.1893562070849653, + "learning_rate": 3.6993720351499286e-06, + "loss": 0.6947, + "step": 1676 + }, + { + "epoch": 0.20164732760175555, + "grad_norm": 2.079293746988888, + "learning_rate": 3.6989611630959666e-06, + "loss": 0.7774, + "step": 1677 + }, + { + "epoch": 0.20176757049239463, + "grad_norm": 0.6872042756854103, + "learning_rate": 3.6985500333109474e-06, + "loss": 0.5842, + "step": 1678 + }, + { + "epoch": 0.20188781338303372, + "grad_norm": 2.404177974980654, + "learning_rate": 3.6981386458572385e-06, + "loss": 0.7731, + "step": 1679 + }, + { + "epoch": 0.20200805627367283, + "grad_norm": 2.4100141931646135, + "learning_rate": 3.6977270007972468e-06, + "loss": 0.7675, + "step": 1680 + }, + { + "epoch": 0.2021282991643119, + "grad_norm": 2.163634550561575, + "learning_rate": 3.6973150981934196e-06, + "loss": 0.7317, + "step": 1681 + }, + { + "epoch": 0.202248542054951, + "grad_norm": 2.520078898208991, + "learning_rate": 3.6969029381082415e-06, + "loss": 0.853, + "step": 1682 + }, + { + "epoch": 0.2023687849455901, + "grad_norm": 2.1038405540229776, + "learning_rate": 3.696490520604237e-06, + "loss": 0.8046, + "step": 1683 + }, + { + "epoch": 0.20248902783622919, + "grad_norm": 1.684273991714427, + "learning_rate": 3.696077845743968e-06, + "loss": 0.8104, + "step": 1684 + }, + { + "epoch": 0.20260927072686827, + "grad_norm": 2.5160420705859137, + "learning_rate": 3.69566491359004e-06, + "loss": 0.7419, + "step": 1685 + }, + { + "epoch": 0.20272951361750738, + "grad_norm": 2.145930821941801, + "learning_rate": 3.695251724205092e-06, + "loss": 0.7056, + "step": 1686 + }, + { + "epoch": 0.20284975650814646, + "grad_norm": 1.724440154849176, + "learning_rate": 3.6948382776518054e-06, + "loss": 0.8633, + "step": 1687 + }, + { + "epoch": 0.20296999939878554, + "grad_norm": 2.0146732892765553, + "learning_rate": 3.6944245739929e-06, + "loss": 0.8039, + "step": 1688 + }, + { + "epoch": 0.20309024228942463, + "grad_norm": 5.353241817092752, + "learning_rate": 3.6940106132911332e-06, + "loss": 0.7245, + "step": 1689 + }, + { + "epoch": 0.20321048518006374, + "grad_norm": 2.1136381796253865, + "learning_rate": 3.6935963956093037e-06, + "loss": 0.8938, + "step": 1690 + }, + { + "epoch": 0.20333072807070282, + "grad_norm": 3.2822846734107523, + "learning_rate": 3.6931819210102474e-06, + "loss": 0.7007, + "step": 1691 + }, + { + "epoch": 0.2034509709613419, + "grad_norm": 1.9054501215488295, + "learning_rate": 3.6927671895568402e-06, + "loss": 0.853, + "step": 1692 + }, + { + "epoch": 0.20357121385198101, + "grad_norm": 1.8266699453426554, + "learning_rate": 3.692352201311996e-06, + "loss": 0.8733, + "step": 1693 + }, + { + "epoch": 0.2036914567426201, + "grad_norm": 2.2170192791793513, + "learning_rate": 3.6919369563386687e-06, + "loss": 0.7734, + "step": 1694 + }, + { + "epoch": 0.20381169963325918, + "grad_norm": 4.549573169851435, + "learning_rate": 3.69152145469985e-06, + "loss": 0.7938, + "step": 1695 + }, + { + "epoch": 0.20393194252389826, + "grad_norm": 1.7255949183145611, + "learning_rate": 3.691105696458572e-06, + "loss": 0.8275, + "step": 1696 + }, + { + "epoch": 0.20405218541453737, + "grad_norm": 2.580420102507734, + "learning_rate": 3.690689681677904e-06, + "loss": 0.696, + "step": 1697 + }, + { + "epoch": 0.20417242830517646, + "grad_norm": 1.7373977301708248, + "learning_rate": 3.690273410420956e-06, + "loss": 0.8857, + "step": 1698 + }, + { + "epoch": 0.20429267119581554, + "grad_norm": 3.410558093538272, + "learning_rate": 3.689856882750875e-06, + "loss": 0.7744, + "step": 1699 + }, + { + "epoch": 0.20441291408645465, + "grad_norm": 1.7543159804899646, + "learning_rate": 3.6894400987308486e-06, + "loss": 0.7923, + "step": 1700 + }, + { + "epoch": 0.20453315697709373, + "grad_norm": 2.065817274877813, + "learning_rate": 3.6890230584241024e-06, + "loss": 0.8604, + "step": 1701 + }, + { + "epoch": 0.20465339986773282, + "grad_norm": 1.0707246863815154, + "learning_rate": 3.6886057618939016e-06, + "loss": 0.6642, + "step": 1702 + }, + { + "epoch": 0.2047736427583719, + "grad_norm": 2.1393029801563315, + "learning_rate": 3.6881882092035492e-06, + "loss": 0.7042, + "step": 1703 + }, + { + "epoch": 0.204893885649011, + "grad_norm": 0.9356373721255861, + "learning_rate": 3.6877704004163873e-06, + "loss": 0.6126, + "step": 1704 + }, + { + "epoch": 0.2050141285396501, + "grad_norm": 1.9020680691906362, + "learning_rate": 3.6873523355957984e-06, + "loss": 0.7844, + "step": 1705 + }, + { + "epoch": 0.20513437143028918, + "grad_norm": 0.993333260106291, + "learning_rate": 3.686934014805201e-06, + "loss": 0.6417, + "step": 1706 + }, + { + "epoch": 0.20525461432092829, + "grad_norm": 1.765977469852458, + "learning_rate": 3.6865154381080552e-06, + "loss": 0.8152, + "step": 1707 + }, + { + "epoch": 0.20537485721156737, + "grad_norm": 2.080112854573629, + "learning_rate": 3.6860966055678585e-06, + "loss": 0.8365, + "step": 1708 + }, + { + "epoch": 0.20549510010220645, + "grad_norm": 1.801333129785857, + "learning_rate": 3.685677517248147e-06, + "loss": 0.8692, + "step": 1709 + }, + { + "epoch": 0.20561534299284553, + "grad_norm": 2.084298622299052, + "learning_rate": 3.6852581732124967e-06, + "loss": 0.8111, + "step": 1710 + }, + { + "epoch": 0.20573558588348465, + "grad_norm": 1.8976445681014906, + "learning_rate": 3.6848385735245213e-06, + "loss": 0.7658, + "step": 1711 + }, + { + "epoch": 0.20585582877412373, + "grad_norm": 1.7886392017653523, + "learning_rate": 3.6844187182478734e-06, + "loss": 0.8702, + "step": 1712 + }, + { + "epoch": 0.2059760716647628, + "grad_norm": 1.9786750450565256, + "learning_rate": 3.683998607446246e-06, + "loss": 0.7594, + "step": 1713 + }, + { + "epoch": 0.20609631455540192, + "grad_norm": 2.149305244682699, + "learning_rate": 3.6835782411833686e-06, + "loss": 0.7591, + "step": 1714 + }, + { + "epoch": 0.206216557446041, + "grad_norm": 1.7741841496361273, + "learning_rate": 3.68315761952301e-06, + "loss": 0.748, + "step": 1715 + }, + { + "epoch": 0.2063368003366801, + "grad_norm": 1.8657979223728216, + "learning_rate": 3.6827367425289797e-06, + "loss": 0.8428, + "step": 1716 + }, + { + "epoch": 0.2064570432273192, + "grad_norm": 2.1431420000658585, + "learning_rate": 3.6823156102651225e-06, + "loss": 0.7337, + "step": 1717 + }, + { + "epoch": 0.20657728611795828, + "grad_norm": 1.8683704156567345, + "learning_rate": 3.6818942227953257e-06, + "loss": 0.7163, + "step": 1718 + }, + { + "epoch": 0.20669752900859736, + "grad_norm": 2.0837870167428916, + "learning_rate": 3.681472580183512e-06, + "loss": 0.6958, + "step": 1719 + }, + { + "epoch": 0.20681777189923645, + "grad_norm": 1.8818442993295024, + "learning_rate": 3.6810506824936455e-06, + "loss": 0.8709, + "step": 1720 + }, + { + "epoch": 0.20693801478987556, + "grad_norm": 1.0605859907323842, + "learning_rate": 3.680628529789726e-06, + "loss": 0.6254, + "step": 1721 + }, + { + "epoch": 0.20705825768051464, + "grad_norm": 1.8497158450909605, + "learning_rate": 3.680206122135796e-06, + "loss": 0.8715, + "step": 1722 + }, + { + "epoch": 0.20717850057115372, + "grad_norm": 1.930969183298098, + "learning_rate": 3.6797834595959323e-06, + "loss": 0.7892, + "step": 1723 + }, + { + "epoch": 0.20729874346179283, + "grad_norm": 2.7862376378205718, + "learning_rate": 3.679360542234254e-06, + "loss": 0.784, + "step": 1724 + }, + { + "epoch": 0.20741898635243192, + "grad_norm": 1.7160811076792217, + "learning_rate": 3.678937370114916e-06, + "loss": 0.7267, + "step": 1725 + }, + { + "epoch": 0.207539229243071, + "grad_norm": 1.9020362934536739, + "learning_rate": 3.678513943302114e-06, + "loss": 0.7883, + "step": 1726 + }, + { + "epoch": 0.20765947213371008, + "grad_norm": 1.9796634519060303, + "learning_rate": 3.678090261860082e-06, + "loss": 0.8605, + "step": 1727 + }, + { + "epoch": 0.2077797150243492, + "grad_norm": 2.0793069332208467, + "learning_rate": 3.6776663258530906e-06, + "loss": 0.7882, + "step": 1728 + }, + { + "epoch": 0.20789995791498828, + "grad_norm": 2.085547249967774, + "learning_rate": 3.6772421353454516e-06, + "loss": 0.7146, + "step": 1729 + }, + { + "epoch": 0.20802020080562736, + "grad_norm": 2.0721782985915, + "learning_rate": 3.6768176904015153e-06, + "loss": 0.8885, + "step": 1730 + }, + { + "epoch": 0.20814044369626647, + "grad_norm": 2.1326903144015854, + "learning_rate": 3.6763929910856674e-06, + "loss": 0.6076, + "step": 1731 + }, + { + "epoch": 0.20826068658690555, + "grad_norm": 2.1680689333200838, + "learning_rate": 3.6759680374623365e-06, + "loss": 0.7838, + "step": 1732 + }, + { + "epoch": 0.20838092947754464, + "grad_norm": 3.096745899639031, + "learning_rate": 3.675542829595986e-06, + "loss": 0.7682, + "step": 1733 + }, + { + "epoch": 0.20850117236818372, + "grad_norm": 1.7214380225581847, + "learning_rate": 3.6751173675511213e-06, + "loss": 0.7969, + "step": 1734 + }, + { + "epoch": 0.20862141525882283, + "grad_norm": 2.474346047521349, + "learning_rate": 3.674691651392283e-06, + "loss": 0.88, + "step": 1735 + }, + { + "epoch": 0.2087416581494619, + "grad_norm": 5.267573598846313, + "learning_rate": 3.674265681184053e-06, + "loss": 0.7636, + "step": 1736 + }, + { + "epoch": 0.208861901040101, + "grad_norm": 1.708030565305499, + "learning_rate": 3.6738394569910504e-06, + "loss": 0.8709, + "step": 1737 + }, + { + "epoch": 0.2089821439307401, + "grad_norm": 2.535836009628918, + "learning_rate": 3.6734129788779333e-06, + "loss": 0.8373, + "step": 1738 + }, + { + "epoch": 0.2091023868213792, + "grad_norm": 1.7175478112064033, + "learning_rate": 3.6729862469093976e-06, + "loss": 0.912, + "step": 1739 + }, + { + "epoch": 0.20922262971201827, + "grad_norm": 2.950645414967404, + "learning_rate": 3.6725592611501782e-06, + "loss": 0.8382, + "step": 1740 + }, + { + "epoch": 0.20934287260265738, + "grad_norm": 1.7241260028294951, + "learning_rate": 3.6721320216650496e-06, + "loss": 0.7701, + "step": 1741 + }, + { + "epoch": 0.20946311549329646, + "grad_norm": 1.982164007243429, + "learning_rate": 3.6717045285188215e-06, + "loss": 0.8386, + "step": 1742 + }, + { + "epoch": 0.20958335838393555, + "grad_norm": 1.8970901040500214, + "learning_rate": 3.671276781776346e-06, + "loss": 0.8737, + "step": 1743 + }, + { + "epoch": 0.20970360127457463, + "grad_norm": 1.9493072428520657, + "learning_rate": 3.6708487815025128e-06, + "loss": 0.6768, + "step": 1744 + }, + { + "epoch": 0.20982384416521374, + "grad_norm": 2.1415754004401206, + "learning_rate": 3.6704205277622463e-06, + "loss": 0.7526, + "step": 1745 + }, + { + "epoch": 0.20994408705585282, + "grad_norm": 2.029197477489297, + "learning_rate": 3.6699920206205146e-06, + "loss": 0.8078, + "step": 1746 + }, + { + "epoch": 0.2100643299464919, + "grad_norm": 1.7310594364321918, + "learning_rate": 3.669563260142321e-06, + "loss": 0.8242, + "step": 1747 + }, + { + "epoch": 0.21018457283713102, + "grad_norm": 1.976663579354065, + "learning_rate": 3.6691342463927083e-06, + "loss": 0.8465, + "step": 1748 + }, + { + "epoch": 0.2103048157277701, + "grad_norm": 2.1345496177455505, + "learning_rate": 3.668704979436758e-06, + "loss": 0.8241, + "step": 1749 + }, + { + "epoch": 0.21042505861840918, + "grad_norm": 4.399305566260661, + "learning_rate": 3.668275459339588e-06, + "loss": 0.7949, + "step": 1750 + }, + { + "epoch": 0.21054530150904827, + "grad_norm": 2.024419715531181, + "learning_rate": 3.667845686166358e-06, + "loss": 0.812, + "step": 1751 + }, + { + "epoch": 0.21066554439968738, + "grad_norm": 2.2750007000597203, + "learning_rate": 3.6674156599822634e-06, + "loss": 0.8659, + "step": 1752 + }, + { + "epoch": 0.21078578729032646, + "grad_norm": 2.027096889511943, + "learning_rate": 3.666985380852539e-06, + "loss": 0.8234, + "step": 1753 + }, + { + "epoch": 0.21090603018096554, + "grad_norm": 2.399898518959948, + "learning_rate": 3.6665548488424576e-06, + "loss": 0.756, + "step": 1754 + }, + { + "epoch": 0.21102627307160465, + "grad_norm": 1.6727180138150768, + "learning_rate": 3.6661240640173307e-06, + "loss": 0.8837, + "step": 1755 + }, + { + "epoch": 0.21114651596224374, + "grad_norm": 0.8748323416166183, + "learning_rate": 3.6656930264425085e-06, + "loss": 0.58, + "step": 1756 + }, + { + "epoch": 0.21126675885288282, + "grad_norm": 2.119458209029609, + "learning_rate": 3.665261736183378e-06, + "loss": 0.7618, + "step": 1757 + }, + { + "epoch": 0.2113870017435219, + "grad_norm": 2.8402391496895465, + "learning_rate": 3.664830193305366e-06, + "loss": 0.8937, + "step": 1758 + }, + { + "epoch": 0.211507244634161, + "grad_norm": 2.7679866901088728, + "learning_rate": 3.6643983978739373e-06, + "loss": 0.7794, + "step": 1759 + }, + { + "epoch": 0.2116274875248001, + "grad_norm": 2.0159482474950514, + "learning_rate": 3.663966349954596e-06, + "loss": 0.8286, + "step": 1760 + }, + { + "epoch": 0.21174773041543918, + "grad_norm": 0.7889601386121536, + "learning_rate": 3.6635340496128816e-06, + "loss": 0.5975, + "step": 1761 + }, + { + "epoch": 0.2118679733060783, + "grad_norm": 1.6461276101140845, + "learning_rate": 3.6631014969143747e-06, + "loss": 0.9309, + "step": 1762 + }, + { + "epoch": 0.21198821619671737, + "grad_norm": 1.847356340452052, + "learning_rate": 3.662668691924693e-06, + "loss": 0.8948, + "step": 1763 + }, + { + "epoch": 0.21210845908735645, + "grad_norm": 2.024368131036461, + "learning_rate": 3.6622356347094927e-06, + "loss": 0.7236, + "step": 1764 + }, + { + "epoch": 0.21222870197799554, + "grad_norm": 2.4764155876796674, + "learning_rate": 3.6618023253344684e-06, + "loss": 0.7889, + "step": 1765 + }, + { + "epoch": 0.21234894486863465, + "grad_norm": 1.7272021459459241, + "learning_rate": 3.6613687638653527e-06, + "loss": 0.8386, + "step": 1766 + }, + { + "epoch": 0.21246918775927373, + "grad_norm": 1.7143917661174541, + "learning_rate": 3.660934950367916e-06, + "loss": 0.7814, + "step": 1767 + }, + { + "epoch": 0.21258943064991281, + "grad_norm": 1.8068982739633634, + "learning_rate": 3.660500884907968e-06, + "loss": 0.8398, + "step": 1768 + }, + { + "epoch": 0.21270967354055192, + "grad_norm": 0.8316505169940883, + "learning_rate": 3.660066567551356e-06, + "loss": 0.6013, + "step": 1769 + }, + { + "epoch": 0.212829916431191, + "grad_norm": 2.4770681544919775, + "learning_rate": 3.6596319983639657e-06, + "loss": 0.8472, + "step": 1770 + }, + { + "epoch": 0.2129501593218301, + "grad_norm": 1.7150309467025902, + "learning_rate": 3.6591971774117214e-06, + "loss": 0.8742, + "step": 1771 + }, + { + "epoch": 0.2130704022124692, + "grad_norm": 2.2560797249322206, + "learning_rate": 3.6587621047605833e-06, + "loss": 0.8083, + "step": 1772 + }, + { + "epoch": 0.21319064510310828, + "grad_norm": 2.1745295048595876, + "learning_rate": 3.6583267804765542e-06, + "loss": 0.8756, + "step": 1773 + }, + { + "epoch": 0.21331088799374737, + "grad_norm": 1.8885766905791066, + "learning_rate": 3.6578912046256702e-06, + "loss": 0.8632, + "step": 1774 + }, + { + "epoch": 0.21343113088438645, + "grad_norm": 2.3088416984617868, + "learning_rate": 3.6574553772740083e-06, + "loss": 0.777, + "step": 1775 + }, + { + "epoch": 0.21355137377502556, + "grad_norm": 0.8724422878973241, + "learning_rate": 3.657019298487684e-06, + "loss": 0.6196, + "step": 1776 + }, + { + "epoch": 0.21367161666566464, + "grad_norm": 1.7741291865087963, + "learning_rate": 3.6565829683328495e-06, + "loss": 0.8462, + "step": 1777 + }, + { + "epoch": 0.21379185955630373, + "grad_norm": 1.748598627569905, + "learning_rate": 3.6561463868756965e-06, + "loss": 0.8603, + "step": 1778 + }, + { + "epoch": 0.21391210244694284, + "grad_norm": 1.6323147978012984, + "learning_rate": 3.655709554182452e-06, + "loss": 0.7851, + "step": 1779 + }, + { + "epoch": 0.21403234533758192, + "grad_norm": 1.7863655918912753, + "learning_rate": 3.6552724703193855e-06, + "loss": 0.8493, + "step": 1780 + }, + { + "epoch": 0.214152588228221, + "grad_norm": 0.7957962813082216, + "learning_rate": 3.654835135352801e-06, + "loss": 0.5596, + "step": 1781 + }, + { + "epoch": 0.21427283111886009, + "grad_norm": 1.8239139408422298, + "learning_rate": 3.654397549349043e-06, + "loss": 0.8821, + "step": 1782 + }, + { + "epoch": 0.2143930740094992, + "grad_norm": 1.9328039540724067, + "learning_rate": 3.653959712374491e-06, + "loss": 0.7605, + "step": 1783 + }, + { + "epoch": 0.21451331690013828, + "grad_norm": 1.7521647279802979, + "learning_rate": 3.6535216244955663e-06, + "loss": 0.8325, + "step": 1784 + }, + { + "epoch": 0.21463355979077736, + "grad_norm": 1.6114638022897343, + "learning_rate": 3.653083285778726e-06, + "loss": 0.7193, + "step": 1785 + }, + { + "epoch": 0.21475380268141647, + "grad_norm": 2.3359238231858184, + "learning_rate": 3.6526446962904653e-06, + "loss": 0.8215, + "step": 1786 + }, + { + "epoch": 0.21487404557205556, + "grad_norm": 1.6271921923272843, + "learning_rate": 3.652205856097318e-06, + "loss": 0.7462, + "step": 1787 + }, + { + "epoch": 0.21499428846269464, + "grad_norm": 2.17147241607299, + "learning_rate": 3.651766765265856e-06, + "loss": 0.7969, + "step": 1788 + }, + { + "epoch": 0.21511453135333372, + "grad_norm": 2.7545873354369808, + "learning_rate": 3.65132742386269e-06, + "loss": 0.8191, + "step": 1789 + }, + { + "epoch": 0.21523477424397283, + "grad_norm": 1.9189253711392298, + "learning_rate": 3.6508878319544656e-06, + "loss": 0.8561, + "step": 1790 + }, + { + "epoch": 0.21535501713461191, + "grad_norm": 2.6450824923712632, + "learning_rate": 3.65044798960787e-06, + "loss": 0.8232, + "step": 1791 + }, + { + "epoch": 0.215475260025251, + "grad_norm": 1.8248995505084111, + "learning_rate": 3.650007896889627e-06, + "loss": 0.7943, + "step": 1792 + }, + { + "epoch": 0.2155955029158901, + "grad_norm": 29.985274706407978, + "learning_rate": 3.6495675538664974e-06, + "loss": 0.8121, + "step": 1793 + }, + { + "epoch": 0.2157157458065292, + "grad_norm": 1.662865436516418, + "learning_rate": 3.649126960605282e-06, + "loss": 0.8274, + "step": 1794 + }, + { + "epoch": 0.21583598869716827, + "grad_norm": 2.194062226511019, + "learning_rate": 3.6486861171728174e-06, + "loss": 0.8377, + "step": 1795 + }, + { + "epoch": 0.21595623158780738, + "grad_norm": 1.610283456337243, + "learning_rate": 3.6482450236359803e-06, + "loss": 0.791, + "step": 1796 + }, + { + "epoch": 0.21607647447844647, + "grad_norm": 2.1879598951552146, + "learning_rate": 3.647803680061683e-06, + "loss": 0.7776, + "step": 1797 + }, + { + "epoch": 0.21619671736908555, + "grad_norm": 2.549189259168263, + "learning_rate": 3.6473620865168776e-06, + "loss": 0.7645, + "step": 1798 + }, + { + "epoch": 0.21631696025972463, + "grad_norm": 1.9478245337689986, + "learning_rate": 3.646920243068554e-06, + "loss": 0.8205, + "step": 1799 + }, + { + "epoch": 0.21643720315036374, + "grad_norm": 1.6018311065958943, + "learning_rate": 3.6464781497837384e-06, + "loss": 0.7545, + "step": 1800 + }, + { + "epoch": 0.21655744604100283, + "grad_norm": 3.5451634516515287, + "learning_rate": 3.6460358067294965e-06, + "loss": 0.7327, + "step": 1801 + }, + { + "epoch": 0.2166776889316419, + "grad_norm": 1.902314548150805, + "learning_rate": 3.645593213972932e-06, + "loss": 0.7819, + "step": 1802 + }, + { + "epoch": 0.21679793182228102, + "grad_norm": 2.2426239545877222, + "learning_rate": 3.6451503715811852e-06, + "loss": 0.8099, + "step": 1803 + }, + { + "epoch": 0.2169181747129201, + "grad_norm": 1.8367948437646606, + "learning_rate": 3.6447072796214345e-06, + "loss": 0.8076, + "step": 1804 + }, + { + "epoch": 0.21703841760355919, + "grad_norm": 0.9028740517723388, + "learning_rate": 3.644263938160898e-06, + "loss": 0.6319, + "step": 1805 + }, + { + "epoch": 0.21715866049419827, + "grad_norm": 3.4745591467174775, + "learning_rate": 3.6438203472668293e-06, + "loss": 0.7338, + "step": 1806 + }, + { + "epoch": 0.21727890338483738, + "grad_norm": 2.4702190422509536, + "learning_rate": 3.6433765070065206e-06, + "loss": 0.829, + "step": 1807 + }, + { + "epoch": 0.21739914627547646, + "grad_norm": 2.6557589806389865, + "learning_rate": 3.6429324174473025e-06, + "loss": 0.8847, + "step": 1808 + }, + { + "epoch": 0.21751938916611555, + "grad_norm": 2.1232541710406005, + "learning_rate": 3.6424880786565425e-06, + "loss": 0.8558, + "step": 1809 + }, + { + "epoch": 0.21763963205675466, + "grad_norm": 1.8841368263757734, + "learning_rate": 3.6420434907016482e-06, + "loss": 0.8094, + "step": 1810 + }, + { + "epoch": 0.21775987494739374, + "grad_norm": 3.1308401848483935, + "learning_rate": 3.6415986536500606e-06, + "loss": 0.8175, + "step": 1811 + }, + { + "epoch": 0.21788011783803282, + "grad_norm": 1.7021471068420286, + "learning_rate": 3.641153567569263e-06, + "loss": 0.8119, + "step": 1812 + }, + { + "epoch": 0.2180003607286719, + "grad_norm": 2.1809546502954866, + "learning_rate": 3.640708232526774e-06, + "loss": 0.963, + "step": 1813 + }, + { + "epoch": 0.21812060361931102, + "grad_norm": 2.093875368834349, + "learning_rate": 3.6402626485901504e-06, + "loss": 0.7944, + "step": 1814 + }, + { + "epoch": 0.2182408465099501, + "grad_norm": 2.9411376182009534, + "learning_rate": 3.639816815826988e-06, + "loss": 0.7869, + "step": 1815 + }, + { + "epoch": 0.21836108940058918, + "grad_norm": 2.044021247805588, + "learning_rate": 3.6393707343049176e-06, + "loss": 0.7864, + "step": 1816 + }, + { + "epoch": 0.2184813322912283, + "grad_norm": 2.2411133855063103, + "learning_rate": 3.6389244040916104e-06, + "loss": 0.7388, + "step": 1817 + }, + { + "epoch": 0.21860157518186737, + "grad_norm": 2.140825025582692, + "learning_rate": 3.6384778252547747e-06, + "loss": 0.7981, + "step": 1818 + }, + { + "epoch": 0.21872181807250646, + "grad_norm": 3.0135134879696976, + "learning_rate": 3.638030997862155e-06, + "loss": 0.785, + "step": 1819 + }, + { + "epoch": 0.21884206096314554, + "grad_norm": 0.7708663203080336, + "learning_rate": 3.6375839219815356e-06, + "loss": 0.5959, + "step": 1820 + }, + { + "epoch": 0.21896230385378465, + "grad_norm": 2.1783614962021307, + "learning_rate": 3.6371365976807375e-06, + "loss": 0.8369, + "step": 1821 + }, + { + "epoch": 0.21908254674442373, + "grad_norm": 1.817305934782502, + "learning_rate": 3.6366890250276185e-06, + "loss": 0.8427, + "step": 1822 + }, + { + "epoch": 0.21920278963506282, + "grad_norm": 1.9842994255032405, + "learning_rate": 3.6362412040900764e-06, + "loss": 0.9022, + "step": 1823 + }, + { + "epoch": 0.21932303252570193, + "grad_norm": 2.049431172979952, + "learning_rate": 3.635793134936044e-06, + "loss": 0.8088, + "step": 1824 + }, + { + "epoch": 0.219443275416341, + "grad_norm": 2.2587930734535293, + "learning_rate": 3.635344817633494e-06, + "loss": 0.7355, + "step": 1825 + }, + { + "epoch": 0.2195635183069801, + "grad_norm": 3.813503446007994, + "learning_rate": 3.634896252250436e-06, + "loss": 0.7662, + "step": 1826 + }, + { + "epoch": 0.2196837611976192, + "grad_norm": 1.8784997702611885, + "learning_rate": 3.6344474388549157e-06, + "loss": 0.8216, + "step": 1827 + }, + { + "epoch": 0.2198040040882583, + "grad_norm": 2.2558906586034304, + "learning_rate": 3.6339983775150183e-06, + "loss": 0.8107, + "step": 1828 + }, + { + "epoch": 0.21992424697889737, + "grad_norm": 2.395742486530549, + "learning_rate": 3.6335490682988664e-06, + "loss": 0.8574, + "step": 1829 + }, + { + "epoch": 0.22004448986953645, + "grad_norm": 2.039801803385497, + "learning_rate": 3.63309951127462e-06, + "loss": 0.8359, + "step": 1830 + }, + { + "epoch": 0.22016473276017556, + "grad_norm": 2.3075609817795386, + "learning_rate": 3.6326497065104757e-06, + "loss": 0.7589, + "step": 1831 + }, + { + "epoch": 0.22028497565081465, + "grad_norm": 3.1078401544183443, + "learning_rate": 3.6321996540746697e-06, + "loss": 0.7879, + "step": 1832 + }, + { + "epoch": 0.22040521854145373, + "grad_norm": 1.7539036672186579, + "learning_rate": 3.6317493540354733e-06, + "loss": 0.8111, + "step": 1833 + }, + { + "epoch": 0.22052546143209284, + "grad_norm": 2.4838385812584125, + "learning_rate": 3.6312988064611976e-06, + "loss": 0.7738, + "step": 1834 + }, + { + "epoch": 0.22064570432273192, + "grad_norm": 1.8413450429177136, + "learning_rate": 3.6308480114201896e-06, + "loss": 0.8206, + "step": 1835 + }, + { + "epoch": 0.220765947213371, + "grad_norm": 1.7603715367899706, + "learning_rate": 3.630396968980835e-06, + "loss": 0.7691, + "step": 1836 + }, + { + "epoch": 0.2208861901040101, + "grad_norm": 2.2859179669977796, + "learning_rate": 3.6299456792115575e-06, + "loss": 0.8359, + "step": 1837 + }, + { + "epoch": 0.2210064329946492, + "grad_norm": 3.293101243581204, + "learning_rate": 3.629494142180815e-06, + "loss": 0.8124, + "step": 1838 + }, + { + "epoch": 0.22112667588528828, + "grad_norm": 2.4359609195349305, + "learning_rate": 3.6290423579571075e-06, + "loss": 0.8531, + "step": 1839 + }, + { + "epoch": 0.22124691877592736, + "grad_norm": 1.691001015413003, + "learning_rate": 3.6285903266089694e-06, + "loss": 0.8074, + "step": 1840 + }, + { + "epoch": 0.22136716166656648, + "grad_norm": 2.760477739534204, + "learning_rate": 3.628138048204974e-06, + "loss": 0.7769, + "step": 1841 + }, + { + "epoch": 0.22148740455720556, + "grad_norm": 2.333761777597794, + "learning_rate": 3.6276855228137304e-06, + "loss": 0.7681, + "step": 1842 + }, + { + "epoch": 0.22160764744784464, + "grad_norm": 3.4338109421850707, + "learning_rate": 3.6272327505038874e-06, + "loss": 0.8263, + "step": 1843 + }, + { + "epoch": 0.22172789033848372, + "grad_norm": 1.9204992316814813, + "learning_rate": 3.626779731344131e-06, + "loss": 0.7856, + "step": 1844 + }, + { + "epoch": 0.22184813322912283, + "grad_norm": 2.207109784104673, + "learning_rate": 3.6263264654031814e-06, + "loss": 0.86, + "step": 1845 + }, + { + "epoch": 0.22196837611976192, + "grad_norm": 0.7067299300867006, + "learning_rate": 3.6258729527498008e-06, + "loss": 0.593, + "step": 1846 + }, + { + "epoch": 0.222088619010401, + "grad_norm": 2.098476597584336, + "learning_rate": 3.6254191934527854e-06, + "loss": 0.6462, + "step": 1847 + }, + { + "epoch": 0.2222088619010401, + "grad_norm": 1.9721611383616864, + "learning_rate": 3.6249651875809715e-06, + "loss": 0.6552, + "step": 1848 + }, + { + "epoch": 0.2223291047916792, + "grad_norm": 2.0157800659147824, + "learning_rate": 3.62451093520323e-06, + "loss": 0.8979, + "step": 1849 + }, + { + "epoch": 0.22244934768231828, + "grad_norm": 2.698880253060763, + "learning_rate": 3.6240564363884714e-06, + "loss": 0.9205, + "step": 1850 + }, + { + "epoch": 0.2225695905729574, + "grad_norm": 2.05331792984589, + "learning_rate": 3.623601691205643e-06, + "loss": 0.7116, + "step": 1851 + }, + { + "epoch": 0.22268983346359647, + "grad_norm": 1.8932058830526066, + "learning_rate": 3.623146699723729e-06, + "loss": 0.8213, + "step": 1852 + }, + { + "epoch": 0.22281007635423555, + "grad_norm": 2.271134196191272, + "learning_rate": 3.6226914620117507e-06, + "loss": 0.7785, + "step": 1853 + }, + { + "epoch": 0.22293031924487464, + "grad_norm": 2.6411770576482394, + "learning_rate": 3.622235978138768e-06, + "loss": 0.8151, + "step": 1854 + }, + { + "epoch": 0.22305056213551375, + "grad_norm": 2.524896630919011, + "learning_rate": 3.621780248173877e-06, + "loss": 0.8179, + "step": 1855 + }, + { + "epoch": 0.22317080502615283, + "grad_norm": 0.8333022389968608, + "learning_rate": 3.6213242721862125e-06, + "loss": 0.6107, + "step": 1856 + }, + { + "epoch": 0.2232910479167919, + "grad_norm": 1.716974763474823, + "learning_rate": 3.620868050244945e-06, + "loss": 0.7587, + "step": 1857 + }, + { + "epoch": 0.22341129080743102, + "grad_norm": 2.2497137735355786, + "learning_rate": 3.6204115824192817e-06, + "loss": 0.7868, + "step": 1858 + }, + { + "epoch": 0.2235315336980701, + "grad_norm": 2.510345985610004, + "learning_rate": 3.619954868778471e-06, + "loss": 0.7684, + "step": 1859 + }, + { + "epoch": 0.2236517765887092, + "grad_norm": 2.0193837921791804, + "learning_rate": 3.6194979093917944e-06, + "loss": 0.8348, + "step": 1860 + }, + { + "epoch": 0.22377201947934827, + "grad_norm": 1.9516124514317632, + "learning_rate": 3.6190407043285724e-06, + "loss": 0.8772, + "step": 1861 + }, + { + "epoch": 0.22389226236998738, + "grad_norm": 2.035356098108131, + "learning_rate": 3.618583253658163e-06, + "loss": 0.7646, + "step": 1862 + }, + { + "epoch": 0.22401250526062647, + "grad_norm": 1.977337108707918, + "learning_rate": 3.618125557449961e-06, + "loss": 0.8678, + "step": 1863 + }, + { + "epoch": 0.22413274815126555, + "grad_norm": 1.9603899917112004, + "learning_rate": 3.6176676157733983e-06, + "loss": 0.8383, + "step": 1864 + }, + { + "epoch": 0.22425299104190466, + "grad_norm": 2.441050462356673, + "learning_rate": 3.6172094286979443e-06, + "loss": 0.7649, + "step": 1865 + }, + { + "epoch": 0.22437323393254374, + "grad_norm": 1.5248192414265562, + "learning_rate": 3.6167509962931064e-06, + "loss": 0.8202, + "step": 1866 + }, + { + "epoch": 0.22449347682318282, + "grad_norm": 2.6657030530192634, + "learning_rate": 3.6162923186284276e-06, + "loss": 0.7848, + "step": 1867 + }, + { + "epoch": 0.2246137197138219, + "grad_norm": 2.1985100642675137, + "learning_rate": 3.6158333957734888e-06, + "loss": 0.8695, + "step": 1868 + }, + { + "epoch": 0.22473396260446102, + "grad_norm": 1.9617311611717234, + "learning_rate": 3.6153742277979088e-06, + "loss": 0.8313, + "step": 1869 + }, + { + "epoch": 0.2248542054951001, + "grad_norm": 2.3314872886839684, + "learning_rate": 3.6149148147713434e-06, + "loss": 0.7928, + "step": 1870 + }, + { + "epoch": 0.22497444838573918, + "grad_norm": 1.7595527681078802, + "learning_rate": 3.614455156763484e-06, + "loss": 0.8695, + "step": 1871 + }, + { + "epoch": 0.2250946912763783, + "grad_norm": 2.3283661583034077, + "learning_rate": 3.613995253844061e-06, + "loss": 0.7219, + "step": 1872 + }, + { + "epoch": 0.22521493416701738, + "grad_norm": 1.850611897389653, + "learning_rate": 3.6135351060828414e-06, + "loss": 0.8116, + "step": 1873 + }, + { + "epoch": 0.22533517705765646, + "grad_norm": 1.9837538580565504, + "learning_rate": 3.6130747135496285e-06, + "loss": 0.697, + "step": 1874 + }, + { + "epoch": 0.22545541994829554, + "grad_norm": 1.9176079125110217, + "learning_rate": 3.6126140763142646e-06, + "loss": 0.6624, + "step": 1875 + }, + { + "epoch": 0.22557566283893465, + "grad_norm": 2.8044193745335733, + "learning_rate": 3.6121531944466275e-06, + "loss": 0.8725, + "step": 1876 + }, + { + "epoch": 0.22569590572957374, + "grad_norm": 2.379772009948839, + "learning_rate": 3.611692068016633e-06, + "loss": 0.7881, + "step": 1877 + }, + { + "epoch": 0.22581614862021282, + "grad_norm": 2.231730249689498, + "learning_rate": 3.611230697094233e-06, + "loss": 0.7551, + "step": 1878 + }, + { + "epoch": 0.22593639151085193, + "grad_norm": 2.0091181565700227, + "learning_rate": 3.6107690817494173e-06, + "loss": 0.8815, + "step": 1879 + }, + { + "epoch": 0.226056634401491, + "grad_norm": 2.203737322618844, + "learning_rate": 3.6103072220522117e-06, + "loss": 0.7147, + "step": 1880 + }, + { + "epoch": 0.2261768772921301, + "grad_norm": 1.8695630400426082, + "learning_rate": 3.609845118072682e-06, + "loss": 0.9235, + "step": 1881 + }, + { + "epoch": 0.2262971201827692, + "grad_norm": 1.8666866380894724, + "learning_rate": 3.6093827698809276e-06, + "loss": 0.8008, + "step": 1882 + }, + { + "epoch": 0.2264173630734083, + "grad_norm": 2.63811359951626, + "learning_rate": 3.6089201775470864e-06, + "loss": 0.8502, + "step": 1883 + }, + { + "epoch": 0.22653760596404737, + "grad_norm": 1.4251938537447995, + "learning_rate": 3.6084573411413334e-06, + "loss": 0.7804, + "step": 1884 + }, + { + "epoch": 0.22665784885468646, + "grad_norm": 1.9977713326852438, + "learning_rate": 3.607994260733881e-06, + "loss": 0.8139, + "step": 1885 + }, + { + "epoch": 0.22677809174532557, + "grad_norm": 1.645881008310767, + "learning_rate": 3.6075309363949776e-06, + "loss": 0.7521, + "step": 1886 + }, + { + "epoch": 0.22689833463596465, + "grad_norm": 2.061762822508762, + "learning_rate": 3.6070673681949094e-06, + "loss": 0.8137, + "step": 1887 + }, + { + "epoch": 0.22701857752660373, + "grad_norm": 1.9888299510908616, + "learning_rate": 3.606603556203999e-06, + "loss": 0.8233, + "step": 1888 + }, + { + "epoch": 0.22713882041724284, + "grad_norm": 1.9953833583118328, + "learning_rate": 3.6061395004926066e-06, + "loss": 0.8476, + "step": 1889 + }, + { + "epoch": 0.22725906330788193, + "grad_norm": 2.169232304396141, + "learning_rate": 3.605675201131129e-06, + "loss": 0.8484, + "step": 1890 + }, + { + "epoch": 0.227379306198521, + "grad_norm": 2.5592537763902534, + "learning_rate": 3.60521065819e-06, + "loss": 0.8054, + "step": 1891 + }, + { + "epoch": 0.2274995490891601, + "grad_norm": 3.4702326724623798, + "learning_rate": 3.60474587173969e-06, + "loss": 0.8834, + "step": 1892 + }, + { + "epoch": 0.2276197919797992, + "grad_norm": 2.213040271894218, + "learning_rate": 3.6042808418507084e-06, + "loss": 0.8502, + "step": 1893 + }, + { + "epoch": 0.22774003487043828, + "grad_norm": 2.009199205364759, + "learning_rate": 3.6038155685935976e-06, + "loss": 0.7746, + "step": 1894 + }, + { + "epoch": 0.22786027776107737, + "grad_norm": 2.0639371777591906, + "learning_rate": 3.6033500520389404e-06, + "loss": 0.7093, + "step": 1895 + }, + { + "epoch": 0.22798052065171648, + "grad_norm": 0.7963203294915684, + "learning_rate": 3.6028842922573553e-06, + "loss": 0.6482, + "step": 1896 + }, + { + "epoch": 0.22810076354235556, + "grad_norm": 0.8550780636581322, + "learning_rate": 3.602418289319497e-06, + "loss": 0.6292, + "step": 1897 + }, + { + "epoch": 0.22822100643299464, + "grad_norm": 1.998270121134844, + "learning_rate": 3.601952043296059e-06, + "loss": 0.7383, + "step": 1898 + }, + { + "epoch": 0.22834124932363373, + "grad_norm": 2.2084866476521126, + "learning_rate": 3.6014855542577696e-06, + "loss": 0.8171, + "step": 1899 + }, + { + "epoch": 0.22846149221427284, + "grad_norm": 1.8658906910394277, + "learning_rate": 3.6010188222753943e-06, + "loss": 0.8508, + "step": 1900 + }, + { + "epoch": 0.22858173510491192, + "grad_norm": 0.8909281164643224, + "learning_rate": 3.6005518474197372e-06, + "loss": 0.6421, + "step": 1901 + }, + { + "epoch": 0.228701977995551, + "grad_norm": 2.2821453897831843, + "learning_rate": 3.6000846297616373e-06, + "loss": 0.792, + "step": 1902 + }, + { + "epoch": 0.22882222088619011, + "grad_norm": 2.6406407101857927, + "learning_rate": 3.5996171693719717e-06, + "loss": 0.7352, + "step": 1903 + }, + { + "epoch": 0.2289424637768292, + "grad_norm": 0.8979374364647239, + "learning_rate": 3.5991494663216528e-06, + "loss": 0.6484, + "step": 1904 + }, + { + "epoch": 0.22906270666746828, + "grad_norm": 2.119583857017389, + "learning_rate": 3.5986815206816314e-06, + "loss": 0.8828, + "step": 1905 + }, + { + "epoch": 0.2291829495581074, + "grad_norm": 1.7683208443682723, + "learning_rate": 3.598213332522895e-06, + "loss": 0.7583, + "step": 1906 + }, + { + "epoch": 0.22930319244874647, + "grad_norm": 2.2934694539562694, + "learning_rate": 3.597744901916466e-06, + "loss": 0.7805, + "step": 1907 + }, + { + "epoch": 0.22942343533938556, + "grad_norm": 2.2846916052387867, + "learning_rate": 3.5972762289334058e-06, + "loss": 0.7703, + "step": 1908 + }, + { + "epoch": 0.22954367823002464, + "grad_norm": 1.9256773760097587, + "learning_rate": 3.5968073136448116e-06, + "loss": 0.8648, + "step": 1909 + }, + { + "epoch": 0.22966392112066375, + "grad_norm": 1.6678066487573313, + "learning_rate": 3.596338156121818e-06, + "loss": 0.9181, + "step": 1910 + }, + { + "epoch": 0.22978416401130283, + "grad_norm": 0.7543733562966733, + "learning_rate": 3.595868756435595e-06, + "loss": 0.594, + "step": 1911 + }, + { + "epoch": 0.22990440690194192, + "grad_norm": 2.3694716911741907, + "learning_rate": 3.5953991146573504e-06, + "loss": 0.8101, + "step": 1912 + }, + { + "epoch": 0.23002464979258103, + "grad_norm": 2.26078669131412, + "learning_rate": 3.5949292308583294e-06, + "loss": 0.8351, + "step": 1913 + }, + { + "epoch": 0.2301448926832201, + "grad_norm": 1.9188026122792117, + "learning_rate": 3.594459105109811e-06, + "loss": 0.8154, + "step": 1914 + }, + { + "epoch": 0.2302651355738592, + "grad_norm": 1.897493229629038, + "learning_rate": 3.593988737483115e-06, + "loss": 0.8191, + "step": 1915 + }, + { + "epoch": 0.23038537846449827, + "grad_norm": 1.9660839012109712, + "learning_rate": 3.5935181280495947e-06, + "loss": 0.7934, + "step": 1916 + }, + { + "epoch": 0.23050562135513739, + "grad_norm": 0.8174364808396549, + "learning_rate": 3.5930472768806412e-06, + "loss": 0.5436, + "step": 1917 + }, + { + "epoch": 0.23062586424577647, + "grad_norm": 1.9932110718748974, + "learning_rate": 3.5925761840476826e-06, + "loss": 0.7767, + "step": 1918 + }, + { + "epoch": 0.23074610713641555, + "grad_norm": 1.8282745821800015, + "learning_rate": 3.592104849622183e-06, + "loss": 0.8254, + "step": 1919 + }, + { + "epoch": 0.23086635002705466, + "grad_norm": 1.4492768227182853, + "learning_rate": 3.591633273675644e-06, + "loss": 0.74, + "step": 1920 + }, + { + "epoch": 0.23098659291769374, + "grad_norm": 0.9151438324346186, + "learning_rate": 3.591161456279602e-06, + "loss": 0.5829, + "step": 1921 + }, + { + "epoch": 0.23110683580833283, + "grad_norm": 1.5326040081431234, + "learning_rate": 3.590689397505633e-06, + "loss": 0.8036, + "step": 1922 + }, + { + "epoch": 0.2312270786989719, + "grad_norm": 1.637073050596366, + "learning_rate": 3.590217097425347e-06, + "loss": 0.8708, + "step": 1923 + }, + { + "epoch": 0.23134732158961102, + "grad_norm": 2.0813503619215687, + "learning_rate": 3.589744556110391e-06, + "loss": 0.7191, + "step": 1924 + }, + { + "epoch": 0.2314675644802501, + "grad_norm": 1.5604428291249666, + "learning_rate": 3.58927177363245e-06, + "loss": 0.8475, + "step": 1925 + }, + { + "epoch": 0.2315878073708892, + "grad_norm": 1.9919821240645643, + "learning_rate": 3.5887987500632447e-06, + "loss": 0.7424, + "step": 1926 + }, + { + "epoch": 0.2317080502615283, + "grad_norm": 1.8830342775528126, + "learning_rate": 3.5883254854745325e-06, + "loss": 0.8496, + "step": 1927 + }, + { + "epoch": 0.23182829315216738, + "grad_norm": 2.4042173445289605, + "learning_rate": 3.587851979938107e-06, + "loss": 0.761, + "step": 1928 + }, + { + "epoch": 0.23194853604280646, + "grad_norm": 1.92817325871694, + "learning_rate": 3.5873782335257985e-06, + "loss": 0.7837, + "step": 1929 + }, + { + "epoch": 0.23206877893344555, + "grad_norm": 2.9490615903904103, + "learning_rate": 3.5869042463094744e-06, + "loss": 0.7958, + "step": 1930 + }, + { + "epoch": 0.23218902182408466, + "grad_norm": 1.979527898917048, + "learning_rate": 3.586430018361038e-06, + "loss": 0.7794, + "step": 1931 + }, + { + "epoch": 0.23230926471472374, + "grad_norm": 1.8190460383632567, + "learning_rate": 3.5859555497524283e-06, + "loss": 0.7652, + "step": 1932 + }, + { + "epoch": 0.23242950760536282, + "grad_norm": 2.2354724029769266, + "learning_rate": 3.5854808405556237e-06, + "loss": 0.931, + "step": 1933 + }, + { + "epoch": 0.23254975049600193, + "grad_norm": 2.1641495339419663, + "learning_rate": 3.5850058908426355e-06, + "loss": 0.7648, + "step": 1934 + }, + { + "epoch": 0.23266999338664102, + "grad_norm": 1.8372031536496314, + "learning_rate": 3.584530700685514e-06, + "loss": 0.8601, + "step": 1935 + }, + { + "epoch": 0.2327902362772801, + "grad_norm": 2.067213740798586, + "learning_rate": 3.5840552701563448e-06, + "loss": 0.8885, + "step": 1936 + }, + { + "epoch": 0.2329104791679192, + "grad_norm": 2.2296699122247823, + "learning_rate": 3.5835795993272513e-06, + "loss": 0.8296, + "step": 1937 + }, + { + "epoch": 0.2330307220585583, + "grad_norm": 1.9300868435824698, + "learning_rate": 3.583103688270391e-06, + "loss": 0.7257, + "step": 1938 + }, + { + "epoch": 0.23315096494919738, + "grad_norm": 2.1370755582883576, + "learning_rate": 3.58262753705796e-06, + "loss": 0.8976, + "step": 1939 + }, + { + "epoch": 0.23327120783983646, + "grad_norm": 0.7637046943500293, + "learning_rate": 3.5821511457621902e-06, + "loss": 0.5555, + "step": 1940 + }, + { + "epoch": 0.23339145073047557, + "grad_norm": 3.3847284331482355, + "learning_rate": 3.5816745144553497e-06, + "loss": 0.8166, + "step": 1941 + }, + { + "epoch": 0.23351169362111465, + "grad_norm": 2.016838477306123, + "learning_rate": 3.5811976432097424e-06, + "loss": 0.768, + "step": 1942 + }, + { + "epoch": 0.23363193651175373, + "grad_norm": 2.3265928821104294, + "learning_rate": 3.58072053209771e-06, + "loss": 0.8543, + "step": 1943 + }, + { + "epoch": 0.23375217940239285, + "grad_norm": 2.1822358141711273, + "learning_rate": 3.5802431811916296e-06, + "loss": 0.7924, + "step": 1944 + }, + { + "epoch": 0.23387242229303193, + "grad_norm": 1.7961974483652385, + "learning_rate": 3.579765590563916e-06, + "loss": 0.8199, + "step": 1945 + }, + { + "epoch": 0.233992665183671, + "grad_norm": 2.1377908921428803, + "learning_rate": 3.579287760287017e-06, + "loss": 0.8265, + "step": 1946 + }, + { + "epoch": 0.2341129080743101, + "grad_norm": 1.6454849196293289, + "learning_rate": 3.578809690433421e-06, + "loss": 0.7358, + "step": 1947 + }, + { + "epoch": 0.2342331509649492, + "grad_norm": 2.145260212727132, + "learning_rate": 3.578331381075651e-06, + "loss": 0.8237, + "step": 1948 + }, + { + "epoch": 0.2343533938555883, + "grad_norm": 2.2156536364083492, + "learning_rate": 3.5778528322862646e-06, + "loss": 0.7082, + "step": 1949 + }, + { + "epoch": 0.23447363674622737, + "grad_norm": 1.636097865364991, + "learning_rate": 3.5773740441378585e-06, + "loss": 0.8721, + "step": 1950 + }, + { + "epoch": 0.23459387963686648, + "grad_norm": 1.63237292929957, + "learning_rate": 3.5768950167030633e-06, + "loss": 0.7441, + "step": 1951 + }, + { + "epoch": 0.23471412252750556, + "grad_norm": 1.791439226781116, + "learning_rate": 3.576415750054548e-06, + "loss": 0.7951, + "step": 1952 + }, + { + "epoch": 0.23483436541814465, + "grad_norm": 1.9831634788014159, + "learning_rate": 3.5759362442650172e-06, + "loss": 0.8629, + "step": 1953 + }, + { + "epoch": 0.23495460830878373, + "grad_norm": 2.0189989816286147, + "learning_rate": 3.5754564994072113e-06, + "loss": 0.8602, + "step": 1954 + }, + { + "epoch": 0.23507485119942284, + "grad_norm": 2.184089617989981, + "learning_rate": 3.5749765155539067e-06, + "loss": 0.6107, + "step": 1955 + }, + { + "epoch": 0.23519509409006192, + "grad_norm": 2.238933261887027, + "learning_rate": 3.574496292777917e-06, + "loss": 0.9344, + "step": 1956 + }, + { + "epoch": 0.235315336980701, + "grad_norm": 1.984261336788413, + "learning_rate": 3.574015831152092e-06, + "loss": 0.7263, + "step": 1957 + }, + { + "epoch": 0.23543557987134012, + "grad_norm": 2.2550306242832, + "learning_rate": 3.573535130749316e-06, + "loss": 0.842, + "step": 1958 + }, + { + "epoch": 0.2355558227619792, + "grad_norm": 1.67447524319946, + "learning_rate": 3.5730541916425127e-06, + "loss": 0.7411, + "step": 1959 + }, + { + "epoch": 0.23567606565261828, + "grad_norm": 2.0472035617515645, + "learning_rate": 3.572573013904639e-06, + "loss": 0.8651, + "step": 1960 + }, + { + "epoch": 0.2357963085432574, + "grad_norm": 1.864345734302935, + "learning_rate": 3.572091597608689e-06, + "loss": 0.9345, + "step": 1961 + }, + { + "epoch": 0.23591655143389648, + "grad_norm": 2.187409753300238, + "learning_rate": 3.571609942827694e-06, + "loss": 0.7417, + "step": 1962 + }, + { + "epoch": 0.23603679432453556, + "grad_norm": 1.7475270100308473, + "learning_rate": 3.57112804963472e-06, + "loss": 0.8897, + "step": 1963 + }, + { + "epoch": 0.23615703721517464, + "grad_norm": 1.7216274796661022, + "learning_rate": 3.57064591810287e-06, + "loss": 0.7705, + "step": 1964 + }, + { + "epoch": 0.23627728010581375, + "grad_norm": 2.207276997345706, + "learning_rate": 3.570163548305284e-06, + "loss": 0.808, + "step": 1965 + }, + { + "epoch": 0.23639752299645284, + "grad_norm": 2.014827600113397, + "learning_rate": 3.569680940315135e-06, + "loss": 0.7016, + "step": 1966 + }, + { + "epoch": 0.23651776588709192, + "grad_norm": 2.0861036827382953, + "learning_rate": 3.5691980942056356e-06, + "loss": 0.8285, + "step": 1967 + }, + { + "epoch": 0.23663800877773103, + "grad_norm": 1.6849392428891266, + "learning_rate": 3.5687150100500332e-06, + "loss": 0.7985, + "step": 1968 + }, + { + "epoch": 0.2367582516683701, + "grad_norm": 1.5816467571445865, + "learning_rate": 3.568231687921611e-06, + "loss": 0.7491, + "step": 1969 + }, + { + "epoch": 0.2368784945590092, + "grad_norm": 1.832868372593199, + "learning_rate": 3.5677481278936883e-06, + "loss": 0.8094, + "step": 1970 + }, + { + "epoch": 0.23699873744964828, + "grad_norm": 0.8428406937197143, + "learning_rate": 3.5672643300396214e-06, + "loss": 0.5785, + "step": 1971 + }, + { + "epoch": 0.2371189803402874, + "grad_norm": 2.9564400287224863, + "learning_rate": 3.566780294432802e-06, + "loss": 0.678, + "step": 1972 + }, + { + "epoch": 0.23723922323092647, + "grad_norm": 2.79607034541799, + "learning_rate": 3.566296021146657e-06, + "loss": 0.7633, + "step": 1973 + }, + { + "epoch": 0.23735946612156555, + "grad_norm": 1.9471424488530593, + "learning_rate": 3.565811510254652e-06, + "loss": 0.7351, + "step": 1974 + }, + { + "epoch": 0.23747970901220466, + "grad_norm": 0.8350102281800588, + "learning_rate": 3.5653267618302845e-06, + "loss": 0.5841, + "step": 1975 + }, + { + "epoch": 0.23759995190284375, + "grad_norm": 1.76612387841979, + "learning_rate": 3.564841775947093e-06, + "loss": 0.866, + "step": 1976 + }, + { + "epoch": 0.23772019479348283, + "grad_norm": 2.0782463784013623, + "learning_rate": 3.5643565526786475e-06, + "loss": 0.7662, + "step": 1977 + }, + { + "epoch": 0.2378404376841219, + "grad_norm": 1.6017313054141298, + "learning_rate": 3.5638710920985574e-06, + "loss": 0.7752, + "step": 1978 + }, + { + "epoch": 0.23796068057476102, + "grad_norm": 2.059928446888102, + "learning_rate": 3.5633853942804655e-06, + "loss": 0.8304, + "step": 1979 + }, + { + "epoch": 0.2380809234654001, + "grad_norm": 2.362353535305214, + "learning_rate": 3.5628994592980527e-06, + "loss": 0.7745, + "step": 1980 + }, + { + "epoch": 0.2382011663560392, + "grad_norm": 1.8555207897718728, + "learning_rate": 3.562413287225034e-06, + "loss": 0.7098, + "step": 1981 + }, + { + "epoch": 0.2383214092466783, + "grad_norm": 2.2062090126963576, + "learning_rate": 3.5619268781351623e-06, + "loss": 0.9028, + "step": 1982 + }, + { + "epoch": 0.23844165213731738, + "grad_norm": 2.0025795751945967, + "learning_rate": 3.5614402321022256e-06, + "loss": 0.7794, + "step": 1983 + }, + { + "epoch": 0.23856189502795647, + "grad_norm": 2.0127269403673993, + "learning_rate": 3.5609533492000463e-06, + "loss": 0.8781, + "step": 1984 + }, + { + "epoch": 0.23868213791859555, + "grad_norm": 2.130115634483718, + "learning_rate": 3.560466229502485e-06, + "loss": 0.7911, + "step": 1985 + }, + { + "epoch": 0.23880238080923466, + "grad_norm": 2.5989125361768184, + "learning_rate": 3.5599788730834384e-06, + "loss": 0.9053, + "step": 1986 + }, + { + "epoch": 0.23892262369987374, + "grad_norm": 2.2403045084583693, + "learning_rate": 3.559491280016836e-06, + "loss": 0.7945, + "step": 1987 + }, + { + "epoch": 0.23904286659051283, + "grad_norm": 2.037756022650681, + "learning_rate": 3.5590034503766465e-06, + "loss": 0.7135, + "step": 1988 + }, + { + "epoch": 0.23916310948115194, + "grad_norm": 2.3312765206881543, + "learning_rate": 3.558515384236874e-06, + "loss": 0.8226, + "step": 1989 + }, + { + "epoch": 0.23928335237179102, + "grad_norm": 1.8877661662611247, + "learning_rate": 3.558027081671556e-06, + "loss": 0.8354, + "step": 1990 + }, + { + "epoch": 0.2394035952624301, + "grad_norm": 2.529272178028556, + "learning_rate": 3.557538542754769e-06, + "loss": 0.6879, + "step": 1991 + }, + { + "epoch": 0.2395238381530692, + "grad_norm": 1.9457634705684135, + "learning_rate": 3.557049767560623e-06, + "loss": 0.6738, + "step": 1992 + }, + { + "epoch": 0.2396440810437083, + "grad_norm": 2.0152988613552667, + "learning_rate": 3.5565607561632655e-06, + "loss": 0.869, + "step": 1993 + }, + { + "epoch": 0.23976432393434738, + "grad_norm": 2.3851383052367052, + "learning_rate": 3.5560715086368787e-06, + "loss": 0.7987, + "step": 1994 + }, + { + "epoch": 0.23988456682498646, + "grad_norm": 2.0376317642761435, + "learning_rate": 3.5555820250556816e-06, + "loss": 0.8317, + "step": 1995 + }, + { + "epoch": 0.24000480971562557, + "grad_norm": 2.3593133577290017, + "learning_rate": 3.5550923054939278e-06, + "loss": 0.7003, + "step": 1996 + }, + { + "epoch": 0.24012505260626466, + "grad_norm": 1.983726455558371, + "learning_rate": 3.5546023500259083e-06, + "loss": 0.7498, + "step": 1997 + }, + { + "epoch": 0.24024529549690374, + "grad_norm": 2.1555094898628195, + "learning_rate": 3.5541121587259477e-06, + "loss": 0.8178, + "step": 1998 + }, + { + "epoch": 0.24036553838754285, + "grad_norm": 0.8453039412144512, + "learning_rate": 3.553621731668408e-06, + "loss": 0.5802, + "step": 1999 + }, + { + "epoch": 0.24048578127818193, + "grad_norm": 1.729630794818972, + "learning_rate": 3.553131068927688e-06, + "loss": 0.84, + "step": 2000 + }, + { + "epoch": 0.24060602416882101, + "grad_norm": 2.5595091136033115, + "learning_rate": 3.552640170578219e-06, + "loss": 0.8076, + "step": 2001 + }, + { + "epoch": 0.2407262670594601, + "grad_norm": 1.942601777973459, + "learning_rate": 3.5521490366944703e-06, + "loss": 0.7892, + "step": 2002 + }, + { + "epoch": 0.2408465099500992, + "grad_norm": 3.286390945702907, + "learning_rate": 3.5516576673509474e-06, + "loss": 0.8081, + "step": 2003 + }, + { + "epoch": 0.2409667528407383, + "grad_norm": 1.6791292904211383, + "learning_rate": 3.5511660626221896e-06, + "loss": 0.8648, + "step": 2004 + }, + { + "epoch": 0.24108699573137737, + "grad_norm": 2.2796593960447464, + "learning_rate": 3.5506742225827744e-06, + "loss": 0.9032, + "step": 2005 + }, + { + "epoch": 0.24120723862201648, + "grad_norm": 2.5928202421917126, + "learning_rate": 3.5501821473073116e-06, + "loss": 0.9069, + "step": 2006 + }, + { + "epoch": 0.24132748151265557, + "grad_norm": 2.0061872858721026, + "learning_rate": 3.54968983687045e-06, + "loss": 0.873, + "step": 2007 + }, + { + "epoch": 0.24144772440329465, + "grad_norm": 2.8062416807191553, + "learning_rate": 3.549197291346872e-06, + "loss": 0.9017, + "step": 2008 + }, + { + "epoch": 0.24156796729393373, + "grad_norm": 4.207253622909528, + "learning_rate": 3.548704510811297e-06, + "loss": 0.8035, + "step": 2009 + }, + { + "epoch": 0.24168821018457284, + "grad_norm": 3.5132134198373426, + "learning_rate": 3.5482114953384787e-06, + "loss": 0.7592, + "step": 2010 + }, + { + "epoch": 0.24180845307521193, + "grad_norm": 2.133317265847416, + "learning_rate": 3.5477182450032077e-06, + "loss": 0.8577, + "step": 2011 + }, + { + "epoch": 0.241928695965851, + "grad_norm": 7.984313367392995, + "learning_rate": 3.5472247598803097e-06, + "loss": 0.8459, + "step": 2012 + }, + { + "epoch": 0.24204893885649012, + "grad_norm": 2.2486889516316357, + "learning_rate": 3.546731040044645e-06, + "loss": 0.8566, + "step": 2013 + }, + { + "epoch": 0.2421691817471292, + "grad_norm": 1.9438921787478456, + "learning_rate": 3.546237085571112e-06, + "loss": 0.7562, + "step": 2014 + }, + { + "epoch": 0.24228942463776829, + "grad_norm": 2.016715058095545, + "learning_rate": 3.5457428965346425e-06, + "loss": 0.737, + "step": 2015 + }, + { + "epoch": 0.2424096675284074, + "grad_norm": 1.8281835562954376, + "learning_rate": 3.545248473010205e-06, + "loss": 0.7509, + "step": 2016 + }, + { + "epoch": 0.24252991041904648, + "grad_norm": 1.657152829527798, + "learning_rate": 3.544753815072802e-06, + "loss": 0.8845, + "step": 2017 + }, + { + "epoch": 0.24265015330968556, + "grad_norm": 1.992332243803502, + "learning_rate": 3.544258922797474e-06, + "loss": 0.8857, + "step": 2018 + }, + { + "epoch": 0.24277039620032465, + "grad_norm": 1.7096531197143898, + "learning_rate": 3.543763796259295e-06, + "loss": 0.7902, + "step": 2019 + }, + { + "epoch": 0.24289063909096376, + "grad_norm": 1.8171545968721858, + "learning_rate": 3.5432684355333754e-06, + "loss": 0.9159, + "step": 2020 + }, + { + "epoch": 0.24301088198160284, + "grad_norm": 1.956455351560141, + "learning_rate": 3.5427728406948613e-06, + "loss": 0.7744, + "step": 2021 + }, + { + "epoch": 0.24313112487224192, + "grad_norm": 0.7704687072345425, + "learning_rate": 3.542277011818934e-06, + "loss": 0.5799, + "step": 2022 + }, + { + "epoch": 0.24325136776288103, + "grad_norm": 2.1036812124719075, + "learning_rate": 3.5417809489808104e-06, + "loss": 0.7536, + "step": 2023 + }, + { + "epoch": 0.24337161065352012, + "grad_norm": 2.025077714723903, + "learning_rate": 3.5412846522557422e-06, + "loss": 0.7322, + "step": 2024 + }, + { + "epoch": 0.2434918535441592, + "grad_norm": 1.972867988326952, + "learning_rate": 3.540788121719018e-06, + "loss": 0.7465, + "step": 2025 + }, + { + "epoch": 0.24361209643479828, + "grad_norm": 1.8830305424349005, + "learning_rate": 3.5402913574459604e-06, + "loss": 0.8327, + "step": 2026 + }, + { + "epoch": 0.2437323393254374, + "grad_norm": 1.7466973065947857, + "learning_rate": 3.5397943595119297e-06, + "loss": 0.868, + "step": 2027 + }, + { + "epoch": 0.24385258221607647, + "grad_norm": 2.6894408026962178, + "learning_rate": 3.5392971279923177e-06, + "loss": 0.7826, + "step": 2028 + }, + { + "epoch": 0.24397282510671556, + "grad_norm": 2.24092188614659, + "learning_rate": 3.5387996629625557e-06, + "loss": 0.8435, + "step": 2029 + }, + { + "epoch": 0.24409306799735467, + "grad_norm": 0.813983882688254, + "learning_rate": 3.5383019644981083e-06, + "loss": 0.5516, + "step": 2030 + }, + { + "epoch": 0.24421331088799375, + "grad_norm": 2.812690728396202, + "learning_rate": 3.5378040326744763e-06, + "loss": 0.741, + "step": 2031 + }, + { + "epoch": 0.24433355377863283, + "grad_norm": 4.023851095625994, + "learning_rate": 3.5373058675671946e-06, + "loss": 0.8663, + "step": 2032 + }, + { + "epoch": 0.24445379666927192, + "grad_norm": 2.5157364530337922, + "learning_rate": 3.536807469251836e-06, + "loss": 0.7348, + "step": 2033 + }, + { + "epoch": 0.24457403955991103, + "grad_norm": 1.9612272132147892, + "learning_rate": 3.5363088378040055e-06, + "loss": 0.8312, + "step": 2034 + }, + { + "epoch": 0.2446942824505501, + "grad_norm": 0.7603527242522792, + "learning_rate": 3.5358099732993463e-06, + "loss": 0.6437, + "step": 2035 + }, + { + "epoch": 0.2448145253411892, + "grad_norm": 2.1116919331994772, + "learning_rate": 3.535310875813535e-06, + "loss": 0.9014, + "step": 2036 + }, + { + "epoch": 0.2449347682318283, + "grad_norm": 1.7686645541091612, + "learning_rate": 3.5348115454222843e-06, + "loss": 0.8233, + "step": 2037 + }, + { + "epoch": 0.2450550111224674, + "grad_norm": 2.0269062253679984, + "learning_rate": 3.5343119822013425e-06, + "loss": 0.8681, + "step": 2038 + }, + { + "epoch": 0.24517525401310647, + "grad_norm": 1.891368877257527, + "learning_rate": 3.533812186226493e-06, + "loss": 0.7827, + "step": 2039 + }, + { + "epoch": 0.24529549690374555, + "grad_norm": 1.828615041408532, + "learning_rate": 3.5333121575735545e-06, + "loss": 0.7623, + "step": 2040 + }, + { + "epoch": 0.24541573979438466, + "grad_norm": 2.2612502382024213, + "learning_rate": 3.532811896318381e-06, + "loss": 0.7567, + "step": 2041 + }, + { + "epoch": 0.24553598268502375, + "grad_norm": 2.0650060052299466, + "learning_rate": 3.5323114025368615e-06, + "loss": 0.8277, + "step": 2042 + }, + { + "epoch": 0.24565622557566283, + "grad_norm": 2.030217539213658, + "learning_rate": 3.53181067630492e-06, + "loss": 0.82, + "step": 2043 + }, + { + "epoch": 0.24577646846630194, + "grad_norm": 1.6793137245371101, + "learning_rate": 3.5313097176985175e-06, + "loss": 0.7657, + "step": 2044 + }, + { + "epoch": 0.24589671135694102, + "grad_norm": 1.8188544725123956, + "learning_rate": 3.5308085267936482e-06, + "loss": 0.8169, + "step": 2045 + }, + { + "epoch": 0.2460169542475801, + "grad_norm": 2.04608208826681, + "learning_rate": 3.530307103666342e-06, + "loss": 0.9062, + "step": 2046 + }, + { + "epoch": 0.24613719713821922, + "grad_norm": 2.4009759746399673, + "learning_rate": 3.5298054483926658e-06, + "loss": 0.8103, + "step": 2047 + }, + { + "epoch": 0.2462574400288583, + "grad_norm": 2.2064387513595243, + "learning_rate": 3.5293035610487187e-06, + "loss": 0.8343, + "step": 2048 + }, + { + "epoch": 0.24637768291949738, + "grad_norm": 0.7336630515290132, + "learning_rate": 3.5288014417106374e-06, + "loss": 0.6198, + "step": 2049 + }, + { + "epoch": 0.24649792581013646, + "grad_norm": 1.8125911558182763, + "learning_rate": 3.528299090454593e-06, + "loss": 0.7619, + "step": 2050 + }, + { + "epoch": 0.24661816870077558, + "grad_norm": 2.2258508448433805, + "learning_rate": 3.527796507356792e-06, + "loss": 0.8366, + "step": 2051 + }, + { + "epoch": 0.24673841159141466, + "grad_norm": 2.503910725272668, + "learning_rate": 3.527293692493475e-06, + "loss": 0.9093, + "step": 2052 + }, + { + "epoch": 0.24685865448205374, + "grad_norm": 2.5561292587156, + "learning_rate": 3.52679064594092e-06, + "loss": 0.7429, + "step": 2053 + }, + { + "epoch": 0.24697889737269285, + "grad_norm": 2.305003309433686, + "learning_rate": 3.5262873677754375e-06, + "loss": 0.7509, + "step": 2054 + }, + { + "epoch": 0.24709914026333193, + "grad_norm": 1.722622479650565, + "learning_rate": 3.5257838580733745e-06, + "loss": 0.8094, + "step": 2055 + }, + { + "epoch": 0.24721938315397102, + "grad_norm": 2.050246512315284, + "learning_rate": 3.5252801169111138e-06, + "loss": 0.8818, + "step": 2056 + }, + { + "epoch": 0.2473396260446101, + "grad_norm": 1.9519861181500342, + "learning_rate": 3.524776144365072e-06, + "loss": 0.802, + "step": 2057 + }, + { + "epoch": 0.2474598689352492, + "grad_norm": 1.850533565696106, + "learning_rate": 3.5242719405117016e-06, + "loss": 0.8017, + "step": 2058 + }, + { + "epoch": 0.2475801118258883, + "grad_norm": 3.1005624076370957, + "learning_rate": 3.5237675054274893e-06, + "loss": 0.7622, + "step": 2059 + }, + { + "epoch": 0.24770035471652738, + "grad_norm": 2.2148657096298945, + "learning_rate": 3.5232628391889584e-06, + "loss": 0.8135, + "step": 2060 + }, + { + "epoch": 0.2478205976071665, + "grad_norm": 2.4785092761812892, + "learning_rate": 3.522757941872666e-06, + "loss": 0.644, + "step": 2061 + }, + { + "epoch": 0.24794084049780557, + "grad_norm": 1.8895327352740199, + "learning_rate": 3.5222528135552042e-06, + "loss": 0.827, + "step": 2062 + }, + { + "epoch": 0.24806108338844465, + "grad_norm": 1.8827549462603588, + "learning_rate": 3.521747454313201e-06, + "loss": 0.8065, + "step": 2063 + }, + { + "epoch": 0.24818132627908374, + "grad_norm": 2.4946001864691727, + "learning_rate": 3.521241864223319e-06, + "loss": 0.6737, + "step": 2064 + }, + { + "epoch": 0.24830156916972285, + "grad_norm": 0.8041245055799774, + "learning_rate": 3.5207360433622552e-06, + "loss": 0.6209, + "step": 2065 + }, + { + "epoch": 0.24842181206036193, + "grad_norm": 2.043255466742783, + "learning_rate": 3.5202299918067437e-06, + "loss": 0.7512, + "step": 2066 + }, + { + "epoch": 0.248542054951001, + "grad_norm": 3.074716917824446, + "learning_rate": 3.519723709633551e-06, + "loss": 0.7029, + "step": 2067 + }, + { + "epoch": 0.24866229784164012, + "grad_norm": 1.8119418501586422, + "learning_rate": 3.519217196919479e-06, + "loss": 0.8384, + "step": 2068 + }, + { + "epoch": 0.2487825407322792, + "grad_norm": 1.9056367356691855, + "learning_rate": 3.518710453741367e-06, + "loss": 0.7421, + "step": 2069 + }, + { + "epoch": 0.2489027836229183, + "grad_norm": 2.363570468324169, + "learning_rate": 3.518203480176086e-06, + "loss": 0.6846, + "step": 2070 + }, + { + "epoch": 0.2490230265135574, + "grad_norm": 1.7105192605154969, + "learning_rate": 3.517696276300545e-06, + "loss": 0.8144, + "step": 2071 + }, + { + "epoch": 0.24914326940419648, + "grad_norm": 2.809897426620093, + "learning_rate": 3.517188842191685e-06, + "loss": 0.7032, + "step": 2072 + }, + { + "epoch": 0.24926351229483557, + "grad_norm": 1.6843136976194564, + "learning_rate": 3.5166811779264837e-06, + "loss": 0.7516, + "step": 2073 + }, + { + "epoch": 0.24938375518547465, + "grad_norm": 2.1066723499140956, + "learning_rate": 3.5161732835819545e-06, + "loss": 0.7855, + "step": 2074 + }, + { + "epoch": 0.24950399807611376, + "grad_norm": 2.0591334987157697, + "learning_rate": 3.515665159235143e-06, + "loss": 0.8468, + "step": 2075 + }, + { + "epoch": 0.24962424096675284, + "grad_norm": 1.9198354428026914, + "learning_rate": 3.5151568049631318e-06, + "loss": 0.7593, + "step": 2076 + }, + { + "epoch": 0.24974448385739192, + "grad_norm": 1.7721175604674384, + "learning_rate": 3.5146482208430385e-06, + "loss": 0.8103, + "step": 2077 + }, + { + "epoch": 0.24986472674803104, + "grad_norm": 2.1558290153531936, + "learning_rate": 3.514139406952014e-06, + "loss": 0.6843, + "step": 2078 + }, + { + "epoch": 0.24998496963867012, + "grad_norm": 1.7644957585617547, + "learning_rate": 3.5136303633672454e-06, + "loss": 0.8397, + "step": 2079 + }, + { + "epoch": 0.25010521252930923, + "grad_norm": 1.7117620423490747, + "learning_rate": 3.5131210901659544e-06, + "loss": 0.7551, + "step": 2080 + }, + { + "epoch": 0.2502254554199483, + "grad_norm": 2.6389463608043715, + "learning_rate": 3.5126115874253967e-06, + "loss": 0.8266, + "step": 2081 + }, + { + "epoch": 0.2503456983105874, + "grad_norm": 3.586271781549426, + "learning_rate": 3.5121018552228644e-06, + "loss": 0.8175, + "step": 2082 + }, + { + "epoch": 0.2504659412012265, + "grad_norm": 2.0906777687426734, + "learning_rate": 3.5115918936356827e-06, + "loss": 0.7662, + "step": 2083 + }, + { + "epoch": 0.25058618409186556, + "grad_norm": 1.8249568385415784, + "learning_rate": 3.5110817027412123e-06, + "loss": 0.7946, + "step": 2084 + }, + { + "epoch": 0.25070642698250467, + "grad_norm": 2.3234711286090337, + "learning_rate": 3.5105712826168493e-06, + "loss": 0.6937, + "step": 2085 + }, + { + "epoch": 0.2508266698731437, + "grad_norm": 1.8528389695506653, + "learning_rate": 3.5100606333400235e-06, + "loss": 0.7068, + "step": 2086 + }, + { + "epoch": 0.25094691276378284, + "grad_norm": 2.0165475316233294, + "learning_rate": 3.5095497549882006e-06, + "loss": 0.7791, + "step": 2087 + }, + { + "epoch": 0.25106715565442195, + "grad_norm": 2.3640181976624874, + "learning_rate": 3.50903864763888e-06, + "loss": 0.7373, + "step": 2088 + }, + { + "epoch": 0.251187398545061, + "grad_norm": 3.7773464009572524, + "learning_rate": 3.5085273113695965e-06, + "loss": 0.7692, + "step": 2089 + }, + { + "epoch": 0.2513076414357001, + "grad_norm": 1.788962948623887, + "learning_rate": 3.508015746257919e-06, + "loss": 0.7883, + "step": 2090 + }, + { + "epoch": 0.2514278843263392, + "grad_norm": 2.0950341912440686, + "learning_rate": 3.5075039523814518e-06, + "loss": 0.8351, + "step": 2091 + }, + { + "epoch": 0.2515481272169783, + "grad_norm": 1.9256196795733334, + "learning_rate": 3.506991929817834e-06, + "loss": 0.8173, + "step": 2092 + }, + { + "epoch": 0.2516683701076174, + "grad_norm": 1.7940796711984857, + "learning_rate": 3.506479678644738e-06, + "loss": 0.833, + "step": 2093 + }, + { + "epoch": 0.2517886129982565, + "grad_norm": 4.709303539318681, + "learning_rate": 3.505967198939873e-06, + "loss": 0.7457, + "step": 2094 + }, + { + "epoch": 0.25190885588889556, + "grad_norm": 2.004788388622175, + "learning_rate": 3.5054544907809813e-06, + "loss": 0.7931, + "step": 2095 + }, + { + "epoch": 0.25202909877953467, + "grad_norm": 1.8920346514127606, + "learning_rate": 3.50494155424584e-06, + "loss": 0.813, + "step": 2096 + }, + { + "epoch": 0.2521493416701738, + "grad_norm": 1.9380405656369954, + "learning_rate": 3.504428389412262e-06, + "loss": 0.8337, + "step": 2097 + }, + { + "epoch": 0.25226958456081283, + "grad_norm": 2.1396016368683557, + "learning_rate": 3.5039149963580927e-06, + "loss": 0.7414, + "step": 2098 + }, + { + "epoch": 0.25238982745145194, + "grad_norm": 2.0176510349922077, + "learning_rate": 3.503401375161215e-06, + "loss": 0.7124, + "step": 2099 + }, + { + "epoch": 0.252510070342091, + "grad_norm": 1.617067046908692, + "learning_rate": 3.502887525899544e-06, + "loss": 0.8425, + "step": 2100 + }, + { + "epoch": 0.2526303132327301, + "grad_norm": 1.8070097675027836, + "learning_rate": 3.50237344865103e-06, + "loss": 0.8288, + "step": 2101 + }, + { + "epoch": 0.2527505561233692, + "grad_norm": 1.975652553097098, + "learning_rate": 3.501859143493658e-06, + "loss": 0.7736, + "step": 2102 + }, + { + "epoch": 0.2528707990140083, + "grad_norm": 0.9214453908477236, + "learning_rate": 3.5013446105054488e-06, + "loss": 0.6058, + "step": 2103 + }, + { + "epoch": 0.2529910419046474, + "grad_norm": 1.747104702009217, + "learning_rate": 3.5008298497644555e-06, + "loss": 0.7552, + "step": 2104 + }, + { + "epoch": 0.2531112847952865, + "grad_norm": 1.7488328583878874, + "learning_rate": 3.500314861348767e-06, + "loss": 0.8815, + "step": 2105 + }, + { + "epoch": 0.25323152768592555, + "grad_norm": 1.9048892307867622, + "learning_rate": 3.499799645336507e-06, + "loss": 0.7754, + "step": 2106 + }, + { + "epoch": 0.25335177057656466, + "grad_norm": 1.4595419189925682, + "learning_rate": 3.4992842018058336e-06, + "loss": 0.8733, + "step": 2107 + }, + { + "epoch": 0.25347201346720377, + "grad_norm": 2.1575788742977435, + "learning_rate": 3.4987685308349384e-06, + "loss": 0.8869, + "step": 2108 + }, + { + "epoch": 0.2535922563578428, + "grad_norm": 2.1384840233923756, + "learning_rate": 3.4982526325020497e-06, + "loss": 0.6276, + "step": 2109 + }, + { + "epoch": 0.25371249924848194, + "grad_norm": 2.4852673349553944, + "learning_rate": 3.4977365068854273e-06, + "loss": 0.8312, + "step": 2110 + }, + { + "epoch": 0.25383274213912105, + "grad_norm": 1.8665145874516165, + "learning_rate": 3.4972201540633676e-06, + "loss": 0.7406, + "step": 2111 + }, + { + "epoch": 0.2539529850297601, + "grad_norm": 1.764694508815092, + "learning_rate": 3.4967035741142008e-06, + "loss": 0.8527, + "step": 2112 + }, + { + "epoch": 0.2540732279203992, + "grad_norm": 1.8490175551122168, + "learning_rate": 3.4961867671162917e-06, + "loss": 0.8216, + "step": 2113 + }, + { + "epoch": 0.2541934708110383, + "grad_norm": 2.194749318712952, + "learning_rate": 3.4956697331480402e-06, + "loss": 0.7808, + "step": 2114 + }, + { + "epoch": 0.2543137137016774, + "grad_norm": 2.508939780834235, + "learning_rate": 3.495152472287879e-06, + "loss": 0.8057, + "step": 2115 + }, + { + "epoch": 0.2544339565923165, + "grad_norm": 1.6504705827293487, + "learning_rate": 3.4946349846142766e-06, + "loss": 0.7419, + "step": 2116 + }, + { + "epoch": 0.25455419948295555, + "grad_norm": 1.9365644349622086, + "learning_rate": 3.4941172702057353e-06, + "loss": 0.7633, + "step": 2117 + }, + { + "epoch": 0.25467444237359466, + "grad_norm": 1.7693058232428018, + "learning_rate": 3.4935993291407924e-06, + "loss": 0.8105, + "step": 2118 + }, + { + "epoch": 0.25479468526423377, + "grad_norm": 2.16749130352876, + "learning_rate": 3.4930811614980183e-06, + "loss": 0.7217, + "step": 2119 + }, + { + "epoch": 0.2549149281548728, + "grad_norm": 1.6941134523864938, + "learning_rate": 3.4925627673560198e-06, + "loss": 0.7967, + "step": 2120 + }, + { + "epoch": 0.25503517104551193, + "grad_norm": 1.7513404208575825, + "learning_rate": 3.4920441467934357e-06, + "loss": 0.8852, + "step": 2121 + }, + { + "epoch": 0.25515541393615104, + "grad_norm": 2.709159245859772, + "learning_rate": 3.491525299888941e-06, + "loss": 0.8362, + "step": 2122 + }, + { + "epoch": 0.2552756568267901, + "grad_norm": 0.8813002808748207, + "learning_rate": 3.491006226721244e-06, + "loss": 0.6271, + "step": 2123 + }, + { + "epoch": 0.2553958997174292, + "grad_norm": 2.1241192956099293, + "learning_rate": 3.4904869273690882e-06, + "loss": 0.7845, + "step": 2124 + }, + { + "epoch": 0.2555161426080683, + "grad_norm": 1.8916566414364357, + "learning_rate": 3.489967401911251e-06, + "loss": 0.8981, + "step": 2125 + }, + { + "epoch": 0.2556363854987074, + "grad_norm": 2.24401270781538, + "learning_rate": 3.4894476504265428e-06, + "loss": 0.7032, + "step": 2126 + }, + { + "epoch": 0.2557566283893465, + "grad_norm": 0.7999373664775373, + "learning_rate": 3.4889276729938104e-06, + "loss": 0.5455, + "step": 2127 + }, + { + "epoch": 0.2558768712799856, + "grad_norm": 1.806781429268132, + "learning_rate": 3.488407469691934e-06, + "loss": 0.8112, + "step": 2128 + }, + { + "epoch": 0.25599711417062465, + "grad_norm": 2.206987390673162, + "learning_rate": 3.487887040599828e-06, + "loss": 0.8206, + "step": 2129 + }, + { + "epoch": 0.25611735706126376, + "grad_norm": 2.215516864413689, + "learning_rate": 3.4873663857964407e-06, + "loss": 0.7689, + "step": 2130 + }, + { + "epoch": 0.2562375999519028, + "grad_norm": 1.9144210790953793, + "learning_rate": 3.4868455053607556e-06, + "loss": 0.6691, + "step": 2131 + }, + { + "epoch": 0.2563578428425419, + "grad_norm": 2.30138457957895, + "learning_rate": 3.486324399371789e-06, + "loss": 0.7249, + "step": 2132 + }, + { + "epoch": 0.25647808573318104, + "grad_norm": 1.9176518631024513, + "learning_rate": 3.485803067908593e-06, + "loss": 0.7915, + "step": 2133 + }, + { + "epoch": 0.2565983286238201, + "grad_norm": 1.7806122290600839, + "learning_rate": 3.485281511050253e-06, + "loss": 0.8005, + "step": 2134 + }, + { + "epoch": 0.2567185715144592, + "grad_norm": 2.1892785607609797, + "learning_rate": 3.484759728875889e-06, + "loss": 0.9039, + "step": 2135 + }, + { + "epoch": 0.2568388144050983, + "grad_norm": 1.7430400045578454, + "learning_rate": 3.4842377214646543e-06, + "loss": 0.8081, + "step": 2136 + }, + { + "epoch": 0.25695905729573737, + "grad_norm": 1.5834725302026083, + "learning_rate": 3.483715488895737e-06, + "loss": 0.6679, + "step": 2137 + }, + { + "epoch": 0.2570793001863765, + "grad_norm": 1.7752021378948237, + "learning_rate": 3.48319303124836e-06, + "loss": 0.7865, + "step": 2138 + }, + { + "epoch": 0.2571995430770156, + "grad_norm": 2.8255304901848706, + "learning_rate": 3.4826703486017798e-06, + "loss": 0.6825, + "step": 2139 + }, + { + "epoch": 0.25731978596765465, + "grad_norm": 1.5277547063915762, + "learning_rate": 3.4821474410352867e-06, + "loss": 0.7774, + "step": 2140 + }, + { + "epoch": 0.25744002885829376, + "grad_norm": 0.9096508997136336, + "learning_rate": 3.481624308628205e-06, + "loss": 0.627, + "step": 2141 + }, + { + "epoch": 0.25756027174893287, + "grad_norm": 2.746365139005665, + "learning_rate": 3.481100951459893e-06, + "loss": 1.0183, + "step": 2142 + }, + { + "epoch": 0.2576805146395719, + "grad_norm": 1.6494845278714736, + "learning_rate": 3.4805773696097453e-06, + "loss": 0.795, + "step": 2143 + }, + { + "epoch": 0.25780075753021103, + "grad_norm": 2.02853129450819, + "learning_rate": 3.4800535631571874e-06, + "loss": 0.8839, + "step": 2144 + }, + { + "epoch": 0.25792100042085014, + "grad_norm": 2.173925625515891, + "learning_rate": 3.4795295321816804e-06, + "loss": 0.7703, + "step": 2145 + }, + { + "epoch": 0.2580412433114892, + "grad_norm": 2.114781775973931, + "learning_rate": 3.47900527676272e-06, + "loss": 0.92, + "step": 2146 + }, + { + "epoch": 0.2581614862021283, + "grad_norm": 1.9511302069108736, + "learning_rate": 3.478480796979835e-06, + "loss": 0.8925, + "step": 2147 + }, + { + "epoch": 0.25828172909276736, + "grad_norm": 1.622978713971893, + "learning_rate": 3.4779560929125894e-06, + "loss": 0.7834, + "step": 2148 + }, + { + "epoch": 0.2584019719834065, + "grad_norm": 0.6736734185018092, + "learning_rate": 3.4774311646405783e-06, + "loss": 0.5694, + "step": 2149 + }, + { + "epoch": 0.2585222148740456, + "grad_norm": 1.9618303310514016, + "learning_rate": 3.476906012243435e-06, + "loss": 0.8453, + "step": 2150 + }, + { + "epoch": 0.25864245776468464, + "grad_norm": 1.6166874946484813, + "learning_rate": 3.476380635800824e-06, + "loss": 0.8196, + "step": 2151 + }, + { + "epoch": 0.25876270065532375, + "grad_norm": 2.088423144496524, + "learning_rate": 3.475855035392444e-06, + "loss": 0.8679, + "step": 2152 + }, + { + "epoch": 0.25888294354596286, + "grad_norm": 1.7762254487501292, + "learning_rate": 3.475329211098029e-06, + "loss": 0.7176, + "step": 2153 + }, + { + "epoch": 0.2590031864366019, + "grad_norm": 1.6293630714111553, + "learning_rate": 3.4748031629973453e-06, + "loss": 0.8281, + "step": 2154 + }, + { + "epoch": 0.25912342932724103, + "grad_norm": 0.9117840793350125, + "learning_rate": 3.4742768911701944e-06, + "loss": 0.5669, + "step": 2155 + }, + { + "epoch": 0.25924367221788014, + "grad_norm": 2.43828384316162, + "learning_rate": 3.4737503956964113e-06, + "loss": 0.718, + "step": 2156 + }, + { + "epoch": 0.2593639151085192, + "grad_norm": 2.0776421695911163, + "learning_rate": 3.473223676655865e-06, + "loss": 0.6763, + "step": 2157 + }, + { + "epoch": 0.2594841579991583, + "grad_norm": 1.8785448262052467, + "learning_rate": 3.472696734128459e-06, + "loss": 0.8084, + "step": 2158 + }, + { + "epoch": 0.2596044008897974, + "grad_norm": 3.6811631922695462, + "learning_rate": 3.4721695681941286e-06, + "loss": 0.7623, + "step": 2159 + }, + { + "epoch": 0.25972464378043647, + "grad_norm": 1.9325891832941366, + "learning_rate": 3.471642178932845e-06, + "loss": 0.8288, + "step": 2160 + }, + { + "epoch": 0.2598448866710756, + "grad_norm": 3.9779845207684184, + "learning_rate": 3.471114566424613e-06, + "loss": 0.909, + "step": 2161 + }, + { + "epoch": 0.25996512956171464, + "grad_norm": 1.951271459350275, + "learning_rate": 3.4705867307494715e-06, + "loss": 0.7642, + "step": 2162 + }, + { + "epoch": 0.26008537245235375, + "grad_norm": 2.5475487315240404, + "learning_rate": 3.470058671987492e-06, + "loss": 0.8482, + "step": 2163 + }, + { + "epoch": 0.26020561534299286, + "grad_norm": 1.9718143895139664, + "learning_rate": 3.4695303902187805e-06, + "loss": 0.8512, + "step": 2164 + }, + { + "epoch": 0.2603258582336319, + "grad_norm": 2.392751408744049, + "learning_rate": 3.469001885523478e-06, + "loss": 0.7935, + "step": 2165 + }, + { + "epoch": 0.260446101124271, + "grad_norm": 1.7551792342342447, + "learning_rate": 3.4684731579817568e-06, + "loss": 0.8186, + "step": 2166 + }, + { + "epoch": 0.26056634401491013, + "grad_norm": 1.6754707813544119, + "learning_rate": 3.4679442076738247e-06, + "loss": 0.7703, + "step": 2167 + }, + { + "epoch": 0.2606865869055492, + "grad_norm": 1.9307758181252368, + "learning_rate": 3.4674150346799245e-06, + "loss": 0.8472, + "step": 2168 + }, + { + "epoch": 0.2608068297961883, + "grad_norm": 2.396199825838129, + "learning_rate": 3.4668856390803295e-06, + "loss": 0.8032, + "step": 2169 + }, + { + "epoch": 0.2609270726868274, + "grad_norm": 2.1945848446878164, + "learning_rate": 3.4663560209553495e-06, + "loss": 0.9039, + "step": 2170 + }, + { + "epoch": 0.26104731557746647, + "grad_norm": 1.7770981839071165, + "learning_rate": 3.4658261803853267e-06, + "loss": 0.7998, + "step": 2171 + }, + { + "epoch": 0.2611675584681056, + "grad_norm": 1.949614459829668, + "learning_rate": 3.4652961174506383e-06, + "loss": 0.8073, + "step": 2172 + }, + { + "epoch": 0.2612878013587447, + "grad_norm": 0.9846718355734907, + "learning_rate": 3.464765832231694e-06, + "loss": 0.5815, + "step": 2173 + }, + { + "epoch": 0.26140804424938374, + "grad_norm": 2.1561089068704606, + "learning_rate": 3.4642353248089373e-06, + "loss": 0.7127, + "step": 2174 + }, + { + "epoch": 0.26152828714002285, + "grad_norm": 3.552899592219028, + "learning_rate": 3.463704595262846e-06, + "loss": 0.8082, + "step": 2175 + }, + { + "epoch": 0.26164853003066196, + "grad_norm": 1.9588747155483424, + "learning_rate": 3.463173643673931e-06, + "loss": 0.7154, + "step": 2176 + }, + { + "epoch": 0.261768772921301, + "grad_norm": 0.913313599626222, + "learning_rate": 3.4626424701227387e-06, + "loss": 0.6361, + "step": 2177 + }, + { + "epoch": 0.26188901581194013, + "grad_norm": 0.8351419174634542, + "learning_rate": 3.4621110746898452e-06, + "loss": 0.5583, + "step": 2178 + }, + { + "epoch": 0.2620092587025792, + "grad_norm": 1.5572190113618969, + "learning_rate": 3.4615794574558654e-06, + "loss": 0.7494, + "step": 2179 + }, + { + "epoch": 0.2621295015932183, + "grad_norm": 2.3539115857590227, + "learning_rate": 3.4610476185014436e-06, + "loss": 0.8543, + "step": 2180 + }, + { + "epoch": 0.2622497444838574, + "grad_norm": 1.5998561794492263, + "learning_rate": 3.4605155579072597e-06, + "loss": 0.803, + "step": 2181 + }, + { + "epoch": 0.26236998737449646, + "grad_norm": 1.768341540294494, + "learning_rate": 3.459983275754027e-06, + "loss": 0.7198, + "step": 2182 + }, + { + "epoch": 0.26249023026513557, + "grad_norm": 2.2714563500864156, + "learning_rate": 3.4594507721224918e-06, + "loss": 0.8047, + "step": 2183 + }, + { + "epoch": 0.2626104731557747, + "grad_norm": 2.9800922760981847, + "learning_rate": 3.4589180470934353e-06, + "loss": 0.8264, + "step": 2184 + }, + { + "epoch": 0.26273071604641374, + "grad_norm": 1.7837969991243632, + "learning_rate": 3.4583851007476713e-06, + "loss": 0.7763, + "step": 2185 + }, + { + "epoch": 0.26285095893705285, + "grad_norm": 2.186323557990864, + "learning_rate": 3.4578519331660464e-06, + "loss": 0.6882, + "step": 2186 + }, + { + "epoch": 0.26297120182769196, + "grad_norm": 1.7921078678389013, + "learning_rate": 3.4573185444294426e-06, + "loss": 0.8302, + "step": 2187 + }, + { + "epoch": 0.263091444718331, + "grad_norm": 1.6972995486891365, + "learning_rate": 3.456784934618774e-06, + "loss": 0.7941, + "step": 2188 + }, + { + "epoch": 0.2632116876089701, + "grad_norm": 2.0487588847109084, + "learning_rate": 3.4562511038149897e-06, + "loss": 0.8065, + "step": 2189 + }, + { + "epoch": 0.26333193049960923, + "grad_norm": 0.8721693779358752, + "learning_rate": 3.4557170520990705e-06, + "loss": 0.5782, + "step": 2190 + }, + { + "epoch": 0.2634521733902483, + "grad_norm": 1.576640724683796, + "learning_rate": 3.4551827795520324e-06, + "loss": 0.87, + "step": 2191 + }, + { + "epoch": 0.2635724162808874, + "grad_norm": 1.8717605581310601, + "learning_rate": 3.4546482862549226e-06, + "loss": 0.8508, + "step": 2192 + }, + { + "epoch": 0.2636926591715265, + "grad_norm": 2.3687303003785374, + "learning_rate": 3.4541135722888253e-06, + "loss": 0.7945, + "step": 2193 + }, + { + "epoch": 0.26381290206216557, + "grad_norm": 1.6200487560423438, + "learning_rate": 3.453578637734854e-06, + "loss": 0.8098, + "step": 2194 + }, + { + "epoch": 0.2639331449528047, + "grad_norm": 1.6248661132823035, + "learning_rate": 3.4530434826741605e-06, + "loss": 0.7879, + "step": 2195 + }, + { + "epoch": 0.26405338784344373, + "grad_norm": 1.7447723235770922, + "learning_rate": 3.452508107187926e-06, + "loss": 0.697, + "step": 2196 + }, + { + "epoch": 0.26417363073408284, + "grad_norm": 1.9438747623549675, + "learning_rate": 3.451972511357366e-06, + "loss": 0.7782, + "step": 2197 + }, + { + "epoch": 0.26429387362472195, + "grad_norm": 1.7612289119890194, + "learning_rate": 3.45143669526373e-06, + "loss": 0.8557, + "step": 2198 + }, + { + "epoch": 0.264414116515361, + "grad_norm": 0.7863718747410531, + "learning_rate": 3.450900658988302e-06, + "loss": 0.6329, + "step": 2199 + }, + { + "epoch": 0.2645343594060001, + "grad_norm": 2.2099135760846913, + "learning_rate": 3.450364402612397e-06, + "loss": 0.7882, + "step": 2200 + }, + { + "epoch": 0.26465460229663923, + "grad_norm": 1.9030422048614388, + "learning_rate": 3.449827926217366e-06, + "loss": 0.8427, + "step": 2201 + }, + { + "epoch": 0.2647748451872783, + "grad_norm": 1.8543055693071202, + "learning_rate": 3.449291229884591e-06, + "loss": 0.8084, + "step": 2202 + }, + { + "epoch": 0.2648950880779174, + "grad_norm": 1.8357821784364938, + "learning_rate": 3.4487543136954887e-06, + "loss": 0.8735, + "step": 2203 + }, + { + "epoch": 0.2650153309685565, + "grad_norm": 1.6995379970171463, + "learning_rate": 3.448217177731509e-06, + "loss": 0.9182, + "step": 2204 + }, + { + "epoch": 0.26513557385919556, + "grad_norm": 1.8359291238796627, + "learning_rate": 3.4476798220741348e-06, + "loss": 0.7927, + "step": 2205 + }, + { + "epoch": 0.26525581674983467, + "grad_norm": 1.7755156453076322, + "learning_rate": 3.4471422468048826e-06, + "loss": 0.7852, + "step": 2206 + }, + { + "epoch": 0.2653760596404738, + "grad_norm": 4.027241214368225, + "learning_rate": 3.4466044520053022e-06, + "loss": 0.7386, + "step": 2207 + }, + { + "epoch": 0.26549630253111284, + "grad_norm": 1.8462282110024861, + "learning_rate": 3.446066437756977e-06, + "loss": 0.6053, + "step": 2208 + }, + { + "epoch": 0.26561654542175195, + "grad_norm": 2.017907321658076, + "learning_rate": 3.4455282041415224e-06, + "loss": 0.7583, + "step": 2209 + }, + { + "epoch": 0.265736788312391, + "grad_norm": 2.089825268244796, + "learning_rate": 3.4449897512405894e-06, + "loss": 0.8717, + "step": 2210 + }, + { + "epoch": 0.2658570312030301, + "grad_norm": 2.0418972611768242, + "learning_rate": 3.444451079135859e-06, + "loss": 0.7593, + "step": 2211 + }, + { + "epoch": 0.2659772740936692, + "grad_norm": 1.8005134565448537, + "learning_rate": 3.4439121879090493e-06, + "loss": 0.7448, + "step": 2212 + }, + { + "epoch": 0.2660975169843083, + "grad_norm": 1.929589025787917, + "learning_rate": 3.4433730776419082e-06, + "loss": 0.8348, + "step": 2213 + }, + { + "epoch": 0.2662177598749474, + "grad_norm": 2.052769102606755, + "learning_rate": 3.4428337484162183e-06, + "loss": 0.8122, + "step": 2214 + }, + { + "epoch": 0.2663380027655865, + "grad_norm": 2.0096901748774667, + "learning_rate": 3.442294200313797e-06, + "loss": 0.8469, + "step": 2215 + }, + { + "epoch": 0.26645824565622556, + "grad_norm": 0.7730401685056608, + "learning_rate": 3.4417544334164916e-06, + "loss": 0.5281, + "step": 2216 + }, + { + "epoch": 0.26657848854686467, + "grad_norm": 2.1426948508612864, + "learning_rate": 3.4412144478061854e-06, + "loss": 0.7821, + "step": 2217 + }, + { + "epoch": 0.2666987314375038, + "grad_norm": 1.9111578875862034, + "learning_rate": 3.4406742435647925e-06, + "loss": 0.762, + "step": 2218 + }, + { + "epoch": 0.26681897432814283, + "grad_norm": 1.897607378528621, + "learning_rate": 3.440133820774263e-06, + "loss": 0.7923, + "step": 2219 + }, + { + "epoch": 0.26693921721878194, + "grad_norm": 1.9509681694605607, + "learning_rate": 3.439593179516578e-06, + "loss": 0.8292, + "step": 2220 + }, + { + "epoch": 0.26705946010942105, + "grad_norm": 2.097058202071358, + "learning_rate": 3.4390523198737524e-06, + "loss": 0.8124, + "step": 2221 + }, + { + "epoch": 0.2671797030000601, + "grad_norm": 1.7570242915984655, + "learning_rate": 3.4385112419278333e-06, + "loss": 0.7326, + "step": 2222 + }, + { + "epoch": 0.2672999458906992, + "grad_norm": 0.8047173503783418, + "learning_rate": 3.4379699457609033e-06, + "loss": 0.6484, + "step": 2223 + }, + { + "epoch": 0.26742018878133833, + "grad_norm": 1.8867169808642068, + "learning_rate": 3.4374284314550755e-06, + "loss": 0.9079, + "step": 2224 + }, + { + "epoch": 0.2675404316719774, + "grad_norm": 1.9809616649217128, + "learning_rate": 3.436886699092498e-06, + "loss": 0.818, + "step": 2225 + }, + { + "epoch": 0.2676606745626165, + "grad_norm": 3.058521878894967, + "learning_rate": 3.4363447487553502e-06, + "loss": 0.724, + "step": 2226 + }, + { + "epoch": 0.26778091745325555, + "grad_norm": 1.933722536055989, + "learning_rate": 3.4358025805258455e-06, + "loss": 0.7843, + "step": 2227 + }, + { + "epoch": 0.26790116034389466, + "grad_norm": 2.542564910840268, + "learning_rate": 3.435260194486232e-06, + "loss": 0.8391, + "step": 2228 + }, + { + "epoch": 0.2680214032345338, + "grad_norm": 2.4101667357576804, + "learning_rate": 3.4347175907187875e-06, + "loss": 0.8227, + "step": 2229 + }, + { + "epoch": 0.26814164612517283, + "grad_norm": 1.7992202712403333, + "learning_rate": 3.4341747693058254e-06, + "loss": 0.883, + "step": 2230 + }, + { + "epoch": 0.26826188901581194, + "grad_norm": 1.8393425066323867, + "learning_rate": 3.4336317303296916e-06, + "loss": 0.7781, + "step": 2231 + }, + { + "epoch": 0.26838213190645105, + "grad_norm": 2.1397952956931894, + "learning_rate": 3.4330884738727635e-06, + "loss": 0.7561, + "step": 2232 + }, + { + "epoch": 0.2685023747970901, + "grad_norm": 1.9210573023084037, + "learning_rate": 3.4325450000174535e-06, + "loss": 0.7126, + "step": 2233 + }, + { + "epoch": 0.2686226176877292, + "grad_norm": 2.8500169720060713, + "learning_rate": 3.4320013088462067e-06, + "loss": 0.7521, + "step": 2234 + }, + { + "epoch": 0.2687428605783683, + "grad_norm": 1.5222365120327634, + "learning_rate": 3.431457400441499e-06, + "loss": 0.8209, + "step": 2235 + }, + { + "epoch": 0.2688631034690074, + "grad_norm": 0.9251313966783287, + "learning_rate": 3.4309132748858424e-06, + "loss": 0.6098, + "step": 2236 + }, + { + "epoch": 0.2689833463596465, + "grad_norm": 1.7461560129419145, + "learning_rate": 3.430368932261779e-06, + "loss": 0.8462, + "step": 2237 + }, + { + "epoch": 0.2691035892502856, + "grad_norm": 1.9258931651684572, + "learning_rate": 3.429824372651886e-06, + "loss": 0.7585, + "step": 2238 + }, + { + "epoch": 0.26922383214092466, + "grad_norm": 2.0414851620499515, + "learning_rate": 3.4292795961387732e-06, + "loss": 0.8425, + "step": 2239 + }, + { + "epoch": 0.26934407503156377, + "grad_norm": 2.1099992746650207, + "learning_rate": 3.4287346028050818e-06, + "loss": 0.8768, + "step": 2240 + }, + { + "epoch": 0.2694643179222028, + "grad_norm": 1.6714168883533669, + "learning_rate": 3.4281893927334866e-06, + "loss": 0.7984, + "step": 2241 + }, + { + "epoch": 0.26958456081284193, + "grad_norm": 2.762957132068368, + "learning_rate": 3.4276439660066963e-06, + "loss": 0.7571, + "step": 2242 + }, + { + "epoch": 0.26970480370348104, + "grad_norm": 2.075673887830901, + "learning_rate": 3.427098322707452e-06, + "loss": 0.855, + "step": 2243 + }, + { + "epoch": 0.2698250465941201, + "grad_norm": 3.805203443945789, + "learning_rate": 3.426552462918526e-06, + "loss": 0.9069, + "step": 2244 + }, + { + "epoch": 0.2699452894847592, + "grad_norm": 2.038981819749036, + "learning_rate": 3.426006386722726e-06, + "loss": 0.7299, + "step": 2245 + }, + { + "epoch": 0.2700655323753983, + "grad_norm": 1.9984614892016361, + "learning_rate": 3.4254600942028914e-06, + "loss": 0.9232, + "step": 2246 + }, + { + "epoch": 0.2701857752660374, + "grad_norm": 2.412738918669122, + "learning_rate": 3.424913585441893e-06, + "loss": 0.8317, + "step": 2247 + }, + { + "epoch": 0.2703060181566765, + "grad_norm": 6.059320304194863, + "learning_rate": 3.4243668605226374e-06, + "loss": 0.8723, + "step": 2248 + }, + { + "epoch": 0.2704262610473156, + "grad_norm": 2.1350426943438725, + "learning_rate": 3.423819919528061e-06, + "loss": 0.8307, + "step": 2249 + }, + { + "epoch": 0.27054650393795465, + "grad_norm": 2.059609207936424, + "learning_rate": 3.4232727625411355e-06, + "loss": 0.7878, + "step": 2250 + }, + { + "epoch": 0.27066674682859376, + "grad_norm": 1.739806596815787, + "learning_rate": 3.4227253896448626e-06, + "loss": 0.8708, + "step": 2251 + }, + { + "epoch": 0.2707869897192329, + "grad_norm": 2.631564515224836, + "learning_rate": 3.42217780092228e-06, + "loss": 0.8295, + "step": 2252 + }, + { + "epoch": 0.27090723260987193, + "grad_norm": 0.8022065666616628, + "learning_rate": 3.421629996456456e-06, + "loss": 0.6043, + "step": 2253 + }, + { + "epoch": 0.27102747550051104, + "grad_norm": 2.31885089500524, + "learning_rate": 3.421081976330491e-06, + "loss": 0.8285, + "step": 2254 + }, + { + "epoch": 0.27114771839115015, + "grad_norm": 2.051314221108488, + "learning_rate": 3.4205337406275207e-06, + "loss": 0.8798, + "step": 2255 + }, + { + "epoch": 0.2712679612817892, + "grad_norm": 2.268501192431844, + "learning_rate": 3.4199852894307114e-06, + "loss": 0.7705, + "step": 2256 + }, + { + "epoch": 0.2713882041724283, + "grad_norm": 2.1252313375772096, + "learning_rate": 3.419436622823262e-06, + "loss": 0.7914, + "step": 2257 + }, + { + "epoch": 0.27150844706306737, + "grad_norm": 1.9140647421064674, + "learning_rate": 3.4188877408884063e-06, + "loss": 0.7475, + "step": 2258 + }, + { + "epoch": 0.2716286899537065, + "grad_norm": 2.3158857973499445, + "learning_rate": 3.4183386437094088e-06, + "loss": 0.6551, + "step": 2259 + }, + { + "epoch": 0.2717489328443456, + "grad_norm": 2.342335795669633, + "learning_rate": 3.417789331369565e-06, + "loss": 0.8244, + "step": 2260 + }, + { + "epoch": 0.27186917573498465, + "grad_norm": 2.0234928786310578, + "learning_rate": 3.4172398039522088e-06, + "loss": 0.919, + "step": 2261 + }, + { + "epoch": 0.27198941862562376, + "grad_norm": 1.8279710362474952, + "learning_rate": 3.4166900615407e-06, + "loss": 0.8025, + "step": 2262 + }, + { + "epoch": 0.27210966151626287, + "grad_norm": 2.1377161686325565, + "learning_rate": 3.416140104218436e-06, + "loss": 0.7542, + "step": 2263 + }, + { + "epoch": 0.2722299044069019, + "grad_norm": 0.8565903022203799, + "learning_rate": 3.4155899320688437e-06, + "loss": 0.6982, + "step": 2264 + }, + { + "epoch": 0.27235014729754103, + "grad_norm": 2.237916287531229, + "learning_rate": 3.415039545175384e-06, + "loss": 0.7508, + "step": 2265 + }, + { + "epoch": 0.27247039018818014, + "grad_norm": 2.085420808841988, + "learning_rate": 3.414488943621551e-06, + "loss": 0.6549, + "step": 2266 + }, + { + "epoch": 0.2725906330788192, + "grad_norm": 1.809205711759887, + "learning_rate": 3.41393812749087e-06, + "loss": 0.7417, + "step": 2267 + }, + { + "epoch": 0.2727108759694583, + "grad_norm": 2.555228826350193, + "learning_rate": 3.4133870968668984e-06, + "loss": 0.724, + "step": 2268 + }, + { + "epoch": 0.2728311188600974, + "grad_norm": 2.5180103058304892, + "learning_rate": 3.412835851833229e-06, + "loss": 0.7892, + "step": 2269 + }, + { + "epoch": 0.2729513617507365, + "grad_norm": 1.8221434197534723, + "learning_rate": 3.4122843924734834e-06, + "loss": 0.7901, + "step": 2270 + }, + { + "epoch": 0.2730716046413756, + "grad_norm": 2.0701848980429167, + "learning_rate": 3.411732718871319e-06, + "loss": 0.8867, + "step": 2271 + }, + { + "epoch": 0.27319184753201464, + "grad_norm": 1.928368739421541, + "learning_rate": 3.4111808311104227e-06, + "loss": 0.7912, + "step": 2272 + }, + { + "epoch": 0.27331209042265375, + "grad_norm": 1.9394290304079853, + "learning_rate": 3.410628729274517e-06, + "loss": 0.6941, + "step": 2273 + }, + { + "epoch": 0.27343233331329286, + "grad_norm": 2.1019669040865616, + "learning_rate": 3.4100764134473546e-06, + "loss": 0.8278, + "step": 2274 + }, + { + "epoch": 0.2735525762039319, + "grad_norm": 2.2553907031412357, + "learning_rate": 3.4095238837127215e-06, + "loss": 0.8564, + "step": 2275 + }, + { + "epoch": 0.27367281909457103, + "grad_norm": 1.9769684144826853, + "learning_rate": 3.4089711401544355e-06, + "loss": 0.8032, + "step": 2276 + }, + { + "epoch": 0.27379306198521014, + "grad_norm": 3.016790925587734, + "learning_rate": 3.4084181828563486e-06, + "loss": 0.6746, + "step": 2277 + }, + { + "epoch": 0.2739133048758492, + "grad_norm": 1.6356959823231623, + "learning_rate": 3.4078650119023428e-06, + "loss": 0.7101, + "step": 2278 + }, + { + "epoch": 0.2740335477664883, + "grad_norm": 2.1893276753757114, + "learning_rate": 3.4073116273763337e-06, + "loss": 0.7458, + "step": 2279 + }, + { + "epoch": 0.2741537906571274, + "grad_norm": 2.453042970488588, + "learning_rate": 3.40675802936227e-06, + "loss": 0.8206, + "step": 2280 + }, + { + "epoch": 0.27427403354776647, + "grad_norm": 2.202915046373347, + "learning_rate": 3.4062042179441318e-06, + "loss": 0.7214, + "step": 2281 + }, + { + "epoch": 0.2743942764384056, + "grad_norm": 1.8406339272766716, + "learning_rate": 3.4056501932059314e-06, + "loss": 0.8137, + "step": 2282 + }, + { + "epoch": 0.2745145193290447, + "grad_norm": 0.782302567007736, + "learning_rate": 3.405095955231715e-06, + "loss": 0.5812, + "step": 2283 + }, + { + "epoch": 0.27463476221968375, + "grad_norm": 3.620757152134941, + "learning_rate": 3.4045415041055585e-06, + "loss": 0.9436, + "step": 2284 + }, + { + "epoch": 0.27475500511032286, + "grad_norm": 2.4004488286907093, + "learning_rate": 3.4039868399115728e-06, + "loss": 0.7903, + "step": 2285 + }, + { + "epoch": 0.27487524800096197, + "grad_norm": 1.8093040834539666, + "learning_rate": 3.4034319627339003e-06, + "loss": 0.8059, + "step": 2286 + }, + { + "epoch": 0.274995490891601, + "grad_norm": 2.3531334784406934, + "learning_rate": 3.402876872656715e-06, + "loss": 0.7002, + "step": 2287 + }, + { + "epoch": 0.27511573378224013, + "grad_norm": 1.8812478556573249, + "learning_rate": 3.402321569764223e-06, + "loss": 0.9025, + "step": 2288 + }, + { + "epoch": 0.2752359766728792, + "grad_norm": 1.849369988262248, + "learning_rate": 3.4017660541406635e-06, + "loss": 0.8392, + "step": 2289 + }, + { + "epoch": 0.2753562195635183, + "grad_norm": 1.786023619970842, + "learning_rate": 3.4012103258703092e-06, + "loss": 0.7444, + "step": 2290 + }, + { + "epoch": 0.2754764624541574, + "grad_norm": 2.220357393244464, + "learning_rate": 3.4006543850374616e-06, + "loss": 0.8323, + "step": 2291 + }, + { + "epoch": 0.27559670534479647, + "grad_norm": 2.066857791594833, + "learning_rate": 3.400098231726458e-06, + "loss": 0.7543, + "step": 2292 + }, + { + "epoch": 0.2757169482354356, + "grad_norm": 2.169304183819608, + "learning_rate": 3.3995418660216657e-06, + "loss": 0.8724, + "step": 2293 + }, + { + "epoch": 0.2758371911260747, + "grad_norm": 3.7171699308643227, + "learning_rate": 3.3989852880074848e-06, + "loss": 0.8108, + "step": 2294 + }, + { + "epoch": 0.27595743401671374, + "grad_norm": 0.7610671557622143, + "learning_rate": 3.398428497768348e-06, + "loss": 0.6075, + "step": 2295 + }, + { + "epoch": 0.27607767690735285, + "grad_norm": 1.7726166964103036, + "learning_rate": 3.3978714953887205e-06, + "loss": 0.7264, + "step": 2296 + }, + { + "epoch": 0.27619791979799196, + "grad_norm": 2.1699981264027186, + "learning_rate": 3.397314280953098e-06, + "loss": 0.8631, + "step": 2297 + }, + { + "epoch": 0.276318162688631, + "grad_norm": 2.0010475753016483, + "learning_rate": 3.3967568545460108e-06, + "loss": 0.806, + "step": 2298 + }, + { + "epoch": 0.27643840557927013, + "grad_norm": 1.8475300461490036, + "learning_rate": 3.3961992162520185e-06, + "loss": 0.8128, + "step": 2299 + }, + { + "epoch": 0.27655864846990924, + "grad_norm": 2.375484517516931, + "learning_rate": 3.3956413661557156e-06, + "loss": 0.7208, + "step": 2300 + }, + { + "epoch": 0.2766788913605483, + "grad_norm": 2.0580602312829375, + "learning_rate": 3.3950833043417273e-06, + "loss": 0.6714, + "step": 2301 + }, + { + "epoch": 0.2767991342511874, + "grad_norm": 2.5780500921290987, + "learning_rate": 3.3945250308947105e-06, + "loss": 0.741, + "step": 2302 + }, + { + "epoch": 0.2769193771418265, + "grad_norm": 1.2888315167254965, + "learning_rate": 3.3939665458993556e-06, + "loss": 0.6842, + "step": 2303 + }, + { + "epoch": 0.27703962003246557, + "grad_norm": 2.1935572004231667, + "learning_rate": 3.3934078494403843e-06, + "loss": 0.7715, + "step": 2304 + }, + { + "epoch": 0.2771598629231047, + "grad_norm": 2.4403628629806544, + "learning_rate": 3.3928489416025495e-06, + "loss": 0.8187, + "step": 2305 + }, + { + "epoch": 0.27728010581374374, + "grad_norm": 2.5695061157890144, + "learning_rate": 3.392289822470638e-06, + "loss": 0.7958, + "step": 2306 + }, + { + "epoch": 0.27740034870438285, + "grad_norm": 2.1321465986282444, + "learning_rate": 3.3917304921294674e-06, + "loss": 0.764, + "step": 2307 + }, + { + "epoch": 0.27752059159502196, + "grad_norm": 1.7656225018778995, + "learning_rate": 3.3911709506638876e-06, + "loss": 0.8099, + "step": 2308 + }, + { + "epoch": 0.277640834485661, + "grad_norm": 2.1272234784274198, + "learning_rate": 3.390611198158781e-06, + "loss": 0.8222, + "step": 2309 + }, + { + "epoch": 0.2777610773763001, + "grad_norm": 2.07192481255813, + "learning_rate": 3.3900512346990612e-06, + "loss": 0.9078, + "step": 2310 + }, + { + "epoch": 0.27788132026693924, + "grad_norm": 1.9506127030688736, + "learning_rate": 3.389491060369674e-06, + "loss": 0.6629, + "step": 2311 + }, + { + "epoch": 0.2780015631575783, + "grad_norm": 2.5593116277317653, + "learning_rate": 3.388930675255598e-06, + "loss": 0.903, + "step": 2312 + }, + { + "epoch": 0.2781218060482174, + "grad_norm": 2.6783090499255016, + "learning_rate": 3.388370079441843e-06, + "loss": 0.8068, + "step": 2313 + }, + { + "epoch": 0.2782420489388565, + "grad_norm": 2.4269719787287753, + "learning_rate": 3.3878092730134505e-06, + "loss": 0.9358, + "step": 2314 + }, + { + "epoch": 0.27836229182949557, + "grad_norm": 1.581687155143009, + "learning_rate": 3.3872482560554947e-06, + "loss": 0.811, + "step": 2315 + }, + { + "epoch": 0.2784825347201347, + "grad_norm": 0.823251201539563, + "learning_rate": 3.386687028653082e-06, + "loss": 0.5699, + "step": 2316 + }, + { + "epoch": 0.2786027776107738, + "grad_norm": 1.8276238691982505, + "learning_rate": 3.386125590891349e-06, + "loss": 0.8543, + "step": 2317 + }, + { + "epoch": 0.27872302050141284, + "grad_norm": 2.2099618005687667, + "learning_rate": 3.3855639428554657e-06, + "loss": 0.8356, + "step": 2318 + }, + { + "epoch": 0.27884326339205195, + "grad_norm": 1.9137556803274538, + "learning_rate": 3.385002084630635e-06, + "loss": 0.8058, + "step": 2319 + }, + { + "epoch": 0.278963506282691, + "grad_norm": 2.299332633845237, + "learning_rate": 3.384440016302088e-06, + "loss": 0.8517, + "step": 2320 + }, + { + "epoch": 0.2790837491733301, + "grad_norm": 2.138189447764737, + "learning_rate": 3.3838777379550923e-06, + "loss": 0.6228, + "step": 2321 + }, + { + "epoch": 0.27920399206396923, + "grad_norm": 2.266320167965958, + "learning_rate": 3.383315249674944e-06, + "loss": 0.7896, + "step": 2322 + }, + { + "epoch": 0.2793242349546083, + "grad_norm": 2.1137659782649454, + "learning_rate": 3.3827525515469715e-06, + "loss": 0.8654, + "step": 2323 + }, + { + "epoch": 0.2794444778452474, + "grad_norm": 2.531182370636381, + "learning_rate": 3.3821896436565367e-06, + "loss": 0.7165, + "step": 2324 + }, + { + "epoch": 0.2795647207358865, + "grad_norm": 1.7696243497428714, + "learning_rate": 3.381626526089032e-06, + "loss": 0.7103, + "step": 2325 + }, + { + "epoch": 0.27968496362652556, + "grad_norm": 2.073122816202545, + "learning_rate": 3.3810631989298815e-06, + "loss": 0.7951, + "step": 2326 + }, + { + "epoch": 0.2798052065171647, + "grad_norm": 3.283846890196678, + "learning_rate": 3.3804996622645423e-06, + "loss": 0.8556, + "step": 2327 + }, + { + "epoch": 0.2799254494078038, + "grad_norm": 2.721821157490072, + "learning_rate": 3.3799359161785015e-06, + "loss": 0.8972, + "step": 2328 + }, + { + "epoch": 0.28004569229844284, + "grad_norm": 2.3194882762470392, + "learning_rate": 3.3793719607572798e-06, + "loss": 0.8603, + "step": 2329 + }, + { + "epoch": 0.28016593518908195, + "grad_norm": 1.9719389867300317, + "learning_rate": 3.378807796086428e-06, + "loss": 0.7722, + "step": 2330 + }, + { + "epoch": 0.28028617807972106, + "grad_norm": 2.021084039087489, + "learning_rate": 3.37824342225153e-06, + "loss": 0.7706, + "step": 2331 + }, + { + "epoch": 0.2804064209703601, + "grad_norm": 2.042697983371387, + "learning_rate": 3.3776788393382006e-06, + "loss": 0.7898, + "step": 2332 + }, + { + "epoch": 0.2805266638609992, + "grad_norm": 2.136127285454471, + "learning_rate": 3.3771140474320872e-06, + "loss": 0.775, + "step": 2333 + }, + { + "epoch": 0.28064690675163834, + "grad_norm": 2.856174366156187, + "learning_rate": 3.3765490466188664e-06, + "loss": 0.7969, + "step": 2334 + }, + { + "epoch": 0.2807671496422774, + "grad_norm": 2.862232552240553, + "learning_rate": 3.3759838369842508e-06, + "loss": 0.7473, + "step": 2335 + }, + { + "epoch": 0.2808873925329165, + "grad_norm": 1.958822196111948, + "learning_rate": 3.375418418613981e-06, + "loss": 0.7357, + "step": 2336 + }, + { + "epoch": 0.28100763542355556, + "grad_norm": 2.7145757090026477, + "learning_rate": 3.374852791593831e-06, + "loss": 0.8434, + "step": 2337 + }, + { + "epoch": 0.28112787831419467, + "grad_norm": 2.5410569090132062, + "learning_rate": 3.374286956009605e-06, + "loss": 0.5442, + "step": 2338 + }, + { + "epoch": 0.2812481212048338, + "grad_norm": 2.132376378222686, + "learning_rate": 3.3737209119471405e-06, + "loss": 0.7578, + "step": 2339 + }, + { + "epoch": 0.28136836409547283, + "grad_norm": 3.762032722529577, + "learning_rate": 3.373154659492306e-06, + "loss": 0.645, + "step": 2340 + }, + { + "epoch": 0.28148860698611194, + "grad_norm": 1.9665868470863876, + "learning_rate": 3.3725881987310016e-06, + "loss": 0.8517, + "step": 2341 + }, + { + "epoch": 0.28160884987675106, + "grad_norm": 6.721955993817016, + "learning_rate": 3.372021529749159e-06, + "loss": 0.8847, + "step": 2342 + }, + { + "epoch": 0.2817290927673901, + "grad_norm": 2.002332281168384, + "learning_rate": 3.3714546526327405e-06, + "loss": 0.931, + "step": 2343 + }, + { + "epoch": 0.2818493356580292, + "grad_norm": 2.6430586034806542, + "learning_rate": 3.3708875674677423e-06, + "loss": 0.8899, + "step": 2344 + }, + { + "epoch": 0.28196957854866833, + "grad_norm": 2.065108759050366, + "learning_rate": 3.37032027434019e-06, + "loss": 0.847, + "step": 2345 + }, + { + "epoch": 0.2820898214393074, + "grad_norm": 2.062334111796141, + "learning_rate": 3.369752773336141e-06, + "loss": 0.8371, + "step": 2346 + }, + { + "epoch": 0.2822100643299465, + "grad_norm": 1.6579115109085074, + "learning_rate": 3.3691850645416864e-06, + "loss": 0.7929, + "step": 2347 + }, + { + "epoch": 0.2823303072205856, + "grad_norm": 1.8494927793528408, + "learning_rate": 3.368617148042945e-06, + "loss": 0.8393, + "step": 2348 + }, + { + "epoch": 0.28245055011122466, + "grad_norm": 1.714357763433917, + "learning_rate": 3.368049023926071e-06, + "loss": 0.8504, + "step": 2349 + }, + { + "epoch": 0.2825707930018638, + "grad_norm": 1.998501096761907, + "learning_rate": 3.3674806922772476e-06, + "loss": 0.8366, + "step": 2350 + }, + { + "epoch": 0.28269103589250283, + "grad_norm": 2.1278315359993285, + "learning_rate": 3.3669121531826904e-06, + "loss": 0.7558, + "step": 2351 + }, + { + "epoch": 0.28281127878314194, + "grad_norm": 2.022502904260387, + "learning_rate": 3.366343406728647e-06, + "loss": 0.8359, + "step": 2352 + }, + { + "epoch": 0.28293152167378105, + "grad_norm": 1.780575206163468, + "learning_rate": 3.3657744530013946e-06, + "loss": 0.691, + "step": 2353 + }, + { + "epoch": 0.2830517645644201, + "grad_norm": 2.213644035814558, + "learning_rate": 3.3652052920872437e-06, + "loss": 0.7222, + "step": 2354 + }, + { + "epoch": 0.2831720074550592, + "grad_norm": 1.9848785362267407, + "learning_rate": 3.3646359240725355e-06, + "loss": 0.8615, + "step": 2355 + }, + { + "epoch": 0.2832922503456983, + "grad_norm": 2.0314361486970025, + "learning_rate": 3.364066349043643e-06, + "loss": 0.6791, + "step": 2356 + }, + { + "epoch": 0.2834124932363374, + "grad_norm": 1.8677069926926382, + "learning_rate": 3.363496567086969e-06, + "loss": 0.8244, + "step": 2357 + }, + { + "epoch": 0.2835327361269765, + "grad_norm": 6.189711666756225, + "learning_rate": 3.3629265782889506e-06, + "loss": 0.7634, + "step": 2358 + }, + { + "epoch": 0.2836529790176156, + "grad_norm": 1.878045088961033, + "learning_rate": 3.362356382736054e-06, + "loss": 0.7224, + "step": 2359 + }, + { + "epoch": 0.28377322190825466, + "grad_norm": 2.0858191981683145, + "learning_rate": 3.361785980514777e-06, + "loss": 0.9147, + "step": 2360 + }, + { + "epoch": 0.28389346479889377, + "grad_norm": 5.95175868096854, + "learning_rate": 3.361215371711649e-06, + "loss": 0.7748, + "step": 2361 + }, + { + "epoch": 0.2840137076895329, + "grad_norm": 1.8075542717221376, + "learning_rate": 3.3606445564132326e-06, + "loss": 0.8405, + "step": 2362 + }, + { + "epoch": 0.28413395058017193, + "grad_norm": 2.0612178632036557, + "learning_rate": 3.360073534706118e-06, + "loss": 0.8291, + "step": 2363 + }, + { + "epoch": 0.28425419347081105, + "grad_norm": 2.1612724745173724, + "learning_rate": 3.35950230667693e-06, + "loss": 0.7724, + "step": 2364 + }, + { + "epoch": 0.28437443636145016, + "grad_norm": 2.163521158423924, + "learning_rate": 3.358930872412323e-06, + "loss": 0.8678, + "step": 2365 + }, + { + "epoch": 0.2844946792520892, + "grad_norm": 1.74618536860272, + "learning_rate": 3.3583592319989825e-06, + "loss": 0.8169, + "step": 2366 + }, + { + "epoch": 0.2846149221427283, + "grad_norm": 2.1421688109625356, + "learning_rate": 3.357787385523627e-06, + "loss": 0.6929, + "step": 2367 + }, + { + "epoch": 0.2847351650333674, + "grad_norm": 1.9451901194366026, + "learning_rate": 3.3572153330730048e-06, + "loss": 0.8307, + "step": 2368 + }, + { + "epoch": 0.2848554079240065, + "grad_norm": 0.8808744644936419, + "learning_rate": 3.3566430747338956e-06, + "loss": 0.6478, + "step": 2369 + }, + { + "epoch": 0.2849756508146456, + "grad_norm": 2.221711023130662, + "learning_rate": 3.35607061059311e-06, + "loss": 0.8683, + "step": 2370 + }, + { + "epoch": 0.28509589370528465, + "grad_norm": 1.7776057794270956, + "learning_rate": 3.3554979407374917e-06, + "loss": 0.757, + "step": 2371 + }, + { + "epoch": 0.28521613659592376, + "grad_norm": 1.572429691071056, + "learning_rate": 3.3549250652539134e-06, + "loss": 0.7397, + "step": 2372 + }, + { + "epoch": 0.2853363794865629, + "grad_norm": 1.9510158289311725, + "learning_rate": 3.3543519842292794e-06, + "loss": 0.8253, + "step": 2373 + }, + { + "epoch": 0.28545662237720193, + "grad_norm": 1.9223217328663453, + "learning_rate": 3.353778697750527e-06, + "loss": 0.8416, + "step": 2374 + }, + { + "epoch": 0.28557686526784104, + "grad_norm": 1.8483762729336237, + "learning_rate": 3.353205205904622e-06, + "loss": 0.8972, + "step": 2375 + }, + { + "epoch": 0.28569710815848015, + "grad_norm": 2.0535377215349833, + "learning_rate": 3.3526315087785637e-06, + "loss": 0.7256, + "step": 2376 + }, + { + "epoch": 0.2858173510491192, + "grad_norm": 1.8982065947338012, + "learning_rate": 3.3520576064593805e-06, + "loss": 0.812, + "step": 2377 + }, + { + "epoch": 0.2859375939397583, + "grad_norm": 1.850175941967507, + "learning_rate": 3.3514834990341337e-06, + "loss": 0.8219, + "step": 2378 + }, + { + "epoch": 0.2860578368303974, + "grad_norm": 2.510236857663732, + "learning_rate": 3.3509091865899144e-06, + "loss": 0.9361, + "step": 2379 + }, + { + "epoch": 0.2861780797210365, + "grad_norm": 2.2824303695570256, + "learning_rate": 3.350334669213846e-06, + "loss": 0.7119, + "step": 2380 + }, + { + "epoch": 0.2862983226116756, + "grad_norm": 2.5358382661010754, + "learning_rate": 3.3497599469930816e-06, + "loss": 0.7599, + "step": 2381 + }, + { + "epoch": 0.28641856550231465, + "grad_norm": 2.4526549453863113, + "learning_rate": 3.349185020014807e-06, + "loss": 0.8444, + "step": 2382 + }, + { + "epoch": 0.28653880839295376, + "grad_norm": 5.093896407381177, + "learning_rate": 3.348609888366237e-06, + "loss": 0.7441, + "step": 2383 + }, + { + "epoch": 0.28665905128359287, + "grad_norm": 1.9620572119098916, + "learning_rate": 3.348034552134619e-06, + "loss": 0.6407, + "step": 2384 + }, + { + "epoch": 0.2867792941742319, + "grad_norm": 1.8364739466230842, + "learning_rate": 3.3474590114072316e-06, + "loss": 0.8454, + "step": 2385 + }, + { + "epoch": 0.28689953706487104, + "grad_norm": 1.9101245464042074, + "learning_rate": 3.3468832662713836e-06, + "loss": 0.8414, + "step": 2386 + }, + { + "epoch": 0.28701977995551015, + "grad_norm": 2.0783801388256298, + "learning_rate": 3.346307316814415e-06, + "loss": 0.8472, + "step": 2387 + }, + { + "epoch": 0.2871400228461492, + "grad_norm": 2.0703153790148026, + "learning_rate": 3.3457311631236965e-06, + "loss": 0.7679, + "step": 2388 + }, + { + "epoch": 0.2872602657367883, + "grad_norm": 1.7058561632088138, + "learning_rate": 3.345154805286631e-06, + "loss": 0.8505, + "step": 2389 + }, + { + "epoch": 0.2873805086274274, + "grad_norm": 2.9248421846511214, + "learning_rate": 3.344578243390651e-06, + "loss": 0.7757, + "step": 2390 + }, + { + "epoch": 0.2875007515180665, + "grad_norm": 2.1109233683487933, + "learning_rate": 3.3440014775232206e-06, + "loss": 0.783, + "step": 2391 + }, + { + "epoch": 0.2876209944087056, + "grad_norm": 2.008442345769606, + "learning_rate": 3.343424507771834e-06, + "loss": 0.7219, + "step": 2392 + }, + { + "epoch": 0.2877412372993447, + "grad_norm": 1.9609330013877058, + "learning_rate": 3.342847334224018e-06, + "loss": 0.8714, + "step": 2393 + }, + { + "epoch": 0.28786148018998375, + "grad_norm": 0.9551147266699624, + "learning_rate": 3.342269956967329e-06, + "loss": 0.6257, + "step": 2394 + }, + { + "epoch": 0.28798172308062286, + "grad_norm": 4.4246233160176995, + "learning_rate": 3.341692376089355e-06, + "loss": 0.7306, + "step": 2395 + }, + { + "epoch": 0.288101965971262, + "grad_norm": 3.2565429555718093, + "learning_rate": 3.3411145916777146e-06, + "loss": 0.8456, + "step": 2396 + }, + { + "epoch": 0.28822220886190103, + "grad_norm": 13.691693225416142, + "learning_rate": 3.3405366038200566e-06, + "loss": 0.9248, + "step": 2397 + }, + { + "epoch": 0.28834245175254014, + "grad_norm": 2.736694834150292, + "learning_rate": 3.3399584126040617e-06, + "loss": 0.8555, + "step": 2398 + }, + { + "epoch": 0.2884626946431792, + "grad_norm": 1.9519921030760323, + "learning_rate": 3.339380018117441e-06, + "loss": 0.9131, + "step": 2399 + }, + { + "epoch": 0.2885829375338183, + "grad_norm": 2.9093085161555647, + "learning_rate": 3.3388014204479366e-06, + "loss": 0.7871, + "step": 2400 + }, + { + "epoch": 0.2887031804244574, + "grad_norm": 2.241764467320534, + "learning_rate": 3.338222619683321e-06, + "loss": 0.9263, + "step": 2401 + }, + { + "epoch": 0.2888234233150965, + "grad_norm": 2.157683407445953, + "learning_rate": 3.337643615911398e-06, + "loss": 0.7446, + "step": 2402 + }, + { + "epoch": 0.2889436662057356, + "grad_norm": 3.3214643567291913, + "learning_rate": 3.3370644092200026e-06, + "loss": 0.7953, + "step": 2403 + }, + { + "epoch": 0.2890639090963747, + "grad_norm": 7.58281205729702, + "learning_rate": 3.3364849996969985e-06, + "loss": 0.789, + "step": 2404 + }, + { + "epoch": 0.28918415198701375, + "grad_norm": 2.3553546732391, + "learning_rate": 3.335905387430283e-06, + "loss": 0.8605, + "step": 2405 + }, + { + "epoch": 0.28930439487765286, + "grad_norm": 1.90707617194601, + "learning_rate": 3.335325572507782e-06, + "loss": 0.8369, + "step": 2406 + }, + { + "epoch": 0.28942463776829197, + "grad_norm": 1.7772277123436424, + "learning_rate": 3.3347455550174537e-06, + "loss": 0.7429, + "step": 2407 + }, + { + "epoch": 0.289544880658931, + "grad_norm": 1.94559338851232, + "learning_rate": 3.3341653350472864e-06, + "loss": 0.6908, + "step": 2408 + }, + { + "epoch": 0.28966512354957014, + "grad_norm": 2.1895530875234552, + "learning_rate": 3.333584912685298e-06, + "loss": 0.703, + "step": 2409 + }, + { + "epoch": 0.28978536644020925, + "grad_norm": 0.8787949894312976, + "learning_rate": 3.3330042880195385e-06, + "loss": 0.5561, + "step": 2410 + }, + { + "epoch": 0.2899056093308483, + "grad_norm": 1.837733976557641, + "learning_rate": 3.3324234611380888e-06, + "loss": 0.7912, + "step": 2411 + }, + { + "epoch": 0.2900258522214874, + "grad_norm": 10.720339287470177, + "learning_rate": 3.3318424321290596e-06, + "loss": 0.8238, + "step": 2412 + }, + { + "epoch": 0.2901460951121265, + "grad_norm": 0.8337420931302487, + "learning_rate": 3.3312612010805917e-06, + "loss": 0.6001, + "step": 2413 + }, + { + "epoch": 0.2902663380027656, + "grad_norm": 1.7236837838824899, + "learning_rate": 3.330679768080858e-06, + "loss": 0.71, + "step": 2414 + }, + { + "epoch": 0.2903865808934047, + "grad_norm": 2.2360190365344352, + "learning_rate": 3.3300981332180627e-06, + "loss": 0.8402, + "step": 2415 + }, + { + "epoch": 0.29050682378404374, + "grad_norm": 1.8017117626229038, + "learning_rate": 3.3295162965804373e-06, + "loss": 0.805, + "step": 2416 + }, + { + "epoch": 0.29062706667468285, + "grad_norm": 2.2294025712250716, + "learning_rate": 3.328934258256247e-06, + "loss": 0.7892, + "step": 2417 + }, + { + "epoch": 0.29074730956532197, + "grad_norm": 2.007663350187363, + "learning_rate": 3.3283520183337856e-06, + "loss": 0.6742, + "step": 2418 + }, + { + "epoch": 0.290867552455961, + "grad_norm": 1.8173160366476946, + "learning_rate": 3.3277695769013797e-06, + "loss": 0.6973, + "step": 2419 + }, + { + "epoch": 0.29098779534660013, + "grad_norm": 1.8441287152843688, + "learning_rate": 3.327186934047385e-06, + "loss": 0.7763, + "step": 2420 + }, + { + "epoch": 0.29110803823723924, + "grad_norm": 2.067043501840485, + "learning_rate": 3.3266040898601877e-06, + "loss": 0.6651, + "step": 2421 + }, + { + "epoch": 0.2912282811278783, + "grad_norm": 1.8527363464486533, + "learning_rate": 3.3260210444282045e-06, + "loss": 0.7871, + "step": 2422 + }, + { + "epoch": 0.2913485240185174, + "grad_norm": 2.2243083050154038, + "learning_rate": 3.325437797839883e-06, + "loss": 0.7403, + "step": 2423 + }, + { + "epoch": 0.2914687669091565, + "grad_norm": 2.413497513021575, + "learning_rate": 3.3248543501837015e-06, + "loss": 0.7589, + "step": 2424 + }, + { + "epoch": 0.2915890097997956, + "grad_norm": 1.884965986759505, + "learning_rate": 3.3242707015481684e-06, + "loss": 0.7813, + "step": 2425 + }, + { + "epoch": 0.2917092526904347, + "grad_norm": 1.8370028387638557, + "learning_rate": 3.323686852021823e-06, + "loss": 0.8132, + "step": 2426 + }, + { + "epoch": 0.2918294955810738, + "grad_norm": 2.237808150910614, + "learning_rate": 3.323102801693235e-06, + "loss": 0.8037, + "step": 2427 + }, + { + "epoch": 0.29194973847171285, + "grad_norm": 3.2583689408824204, + "learning_rate": 3.322518550651003e-06, + "loss": 0.8136, + "step": 2428 + }, + { + "epoch": 0.29206998136235196, + "grad_norm": 1.6848635387912412, + "learning_rate": 3.3219340989837586e-06, + "loss": 0.8171, + "step": 2429 + }, + { + "epoch": 0.292190224252991, + "grad_norm": 1.9076894109635332, + "learning_rate": 3.3213494467801625e-06, + "loss": 0.8117, + "step": 2430 + }, + { + "epoch": 0.2923104671436301, + "grad_norm": 5.639979782609348, + "learning_rate": 3.3207645941289063e-06, + "loss": 0.7215, + "step": 2431 + }, + { + "epoch": 0.29243071003426924, + "grad_norm": 1.7980509351659228, + "learning_rate": 3.320179541118711e-06, + "loss": 0.8076, + "step": 2432 + }, + { + "epoch": 0.2925509529249083, + "grad_norm": 1.003853703767541, + "learning_rate": 3.3195942878383293e-06, + "loss": 0.6036, + "step": 2433 + }, + { + "epoch": 0.2926711958155474, + "grad_norm": 2.7361193515332918, + "learning_rate": 3.319008834376543e-06, + "loss": 0.7847, + "step": 2434 + }, + { + "epoch": 0.2927914387061865, + "grad_norm": 2.3896157122554875, + "learning_rate": 3.3184231808221654e-06, + "loss": 0.8924, + "step": 2435 + }, + { + "epoch": 0.29291168159682557, + "grad_norm": 2.253252433739142, + "learning_rate": 3.3178373272640394e-06, + "loss": 0.6387, + "step": 2436 + }, + { + "epoch": 0.2930319244874647, + "grad_norm": 2.21805593890964, + "learning_rate": 3.3172512737910387e-06, + "loss": 0.8654, + "step": 2437 + }, + { + "epoch": 0.2931521673781038, + "grad_norm": 2.0892414116266957, + "learning_rate": 3.3166650204920674e-06, + "loss": 0.8914, + "step": 2438 + }, + { + "epoch": 0.29327241026874284, + "grad_norm": 2.3744494151327826, + "learning_rate": 3.316078567456059e-06, + "loss": 0.8167, + "step": 2439 + }, + { + "epoch": 0.29339265315938196, + "grad_norm": 1.5711603864729955, + "learning_rate": 3.3154919147719786e-06, + "loss": 0.7654, + "step": 2440 + }, + { + "epoch": 0.29351289605002107, + "grad_norm": 1.9020035143758593, + "learning_rate": 3.31490506252882e-06, + "loss": 0.8689, + "step": 2441 + }, + { + "epoch": 0.2936331389406601, + "grad_norm": 1.7640561586034575, + "learning_rate": 3.31431801081561e-06, + "loss": 0.8423, + "step": 2442 + }, + { + "epoch": 0.29375338183129923, + "grad_norm": 0.9070182043643892, + "learning_rate": 3.313730759721402e-06, + "loss": 0.6787, + "step": 2443 + }, + { + "epoch": 0.29387362472193834, + "grad_norm": 1.8971147992111885, + "learning_rate": 3.313143309335282e-06, + "loss": 0.8667, + "step": 2444 + }, + { + "epoch": 0.2939938676125774, + "grad_norm": 1.9784206657555103, + "learning_rate": 3.3125556597463665e-06, + "loss": 0.8546, + "step": 2445 + }, + { + "epoch": 0.2941141105032165, + "grad_norm": 1.5673498037314708, + "learning_rate": 3.311967811043801e-06, + "loss": 0.6689, + "step": 2446 + }, + { + "epoch": 0.29423435339385556, + "grad_norm": 2.0126591826800238, + "learning_rate": 3.3113797633167617e-06, + "loss": 0.8208, + "step": 2447 + }, + { + "epoch": 0.2943545962844947, + "grad_norm": 3.040908684194636, + "learning_rate": 3.310791516654455e-06, + "loss": 0.7025, + "step": 2448 + }, + { + "epoch": 0.2944748391751338, + "grad_norm": 4.901965755160934, + "learning_rate": 3.3102030711461177e-06, + "loss": 0.8036, + "step": 2449 + }, + { + "epoch": 0.29459508206577284, + "grad_norm": 1.9237928048391615, + "learning_rate": 3.3096144268810156e-06, + "loss": 0.6916, + "step": 2450 + }, + { + "epoch": 0.29471532495641195, + "grad_norm": 2.0502388373086062, + "learning_rate": 3.3090255839484462e-06, + "loss": 0.7354, + "step": 2451 + }, + { + "epoch": 0.29483556784705106, + "grad_norm": 1.8643236743129556, + "learning_rate": 3.3084365424377366e-06, + "loss": 0.8576, + "step": 2452 + }, + { + "epoch": 0.2949558107376901, + "grad_norm": 0.7411814245246326, + "learning_rate": 3.307847302438245e-06, + "loss": 0.5608, + "step": 2453 + }, + { + "epoch": 0.2950760536283292, + "grad_norm": 2.569784834942488, + "learning_rate": 3.3072578640393562e-06, + "loss": 0.787, + "step": 2454 + }, + { + "epoch": 0.29519629651896834, + "grad_norm": 1.9498814620155414, + "learning_rate": 3.3066682273304886e-06, + "loss": 0.804, + "step": 2455 + }, + { + "epoch": 0.2953165394096074, + "grad_norm": 2.020149062671514, + "learning_rate": 3.3060783924010904e-06, + "loss": 0.7951, + "step": 2456 + }, + { + "epoch": 0.2954367823002465, + "grad_norm": 2.918740452622149, + "learning_rate": 3.3054883593406387e-06, + "loss": 0.8535, + "step": 2457 + }, + { + "epoch": 0.2955570251908856, + "grad_norm": 3.6472456167619502, + "learning_rate": 3.3048981282386404e-06, + "loss": 0.6576, + "step": 2458 + }, + { + "epoch": 0.29567726808152467, + "grad_norm": 2.419882465377241, + "learning_rate": 3.304307699184634e-06, + "loss": 0.827, + "step": 2459 + }, + { + "epoch": 0.2957975109721638, + "grad_norm": 1.8517751450899815, + "learning_rate": 3.3037170722681866e-06, + "loss": 0.7928, + "step": 2460 + }, + { + "epoch": 0.29591775386280283, + "grad_norm": 1.919438313837684, + "learning_rate": 3.3031262475788956e-06, + "loss": 0.6884, + "step": 2461 + }, + { + "epoch": 0.29603799675344195, + "grad_norm": 1.7726519353014645, + "learning_rate": 3.3025352252063897e-06, + "loss": 0.7362, + "step": 2462 + }, + { + "epoch": 0.29615823964408106, + "grad_norm": 1.795874275513781, + "learning_rate": 3.3019440052403252e-06, + "loss": 0.7617, + "step": 2463 + }, + { + "epoch": 0.2962784825347201, + "grad_norm": 2.069407715211907, + "learning_rate": 3.30135258777039e-06, + "loss": 0.7092, + "step": 2464 + }, + { + "epoch": 0.2963987254253592, + "grad_norm": 2.1308166331786875, + "learning_rate": 3.3007609728863024e-06, + "loss": 0.706, + "step": 2465 + }, + { + "epoch": 0.29651896831599833, + "grad_norm": 1.8148838020729745, + "learning_rate": 3.300169160677809e-06, + "loss": 0.7408, + "step": 2466 + }, + { + "epoch": 0.2966392112066374, + "grad_norm": 2.995384480502487, + "learning_rate": 3.2995771512346878e-06, + "loss": 0.7809, + "step": 2467 + }, + { + "epoch": 0.2967594540972765, + "grad_norm": 3.074869671052339, + "learning_rate": 3.298984944646746e-06, + "loss": 0.7427, + "step": 2468 + }, + { + "epoch": 0.2968796969879156, + "grad_norm": 1.854070582382663, + "learning_rate": 3.298392541003822e-06, + "loss": 0.8191, + "step": 2469 + }, + { + "epoch": 0.29699993987855466, + "grad_norm": 1.7703235191547715, + "learning_rate": 3.2977999403957806e-06, + "loss": 0.899, + "step": 2470 + }, + { + "epoch": 0.2971201827691938, + "grad_norm": 2.2092803028321084, + "learning_rate": 3.2972071429125207e-06, + "loss": 0.6725, + "step": 2471 + }, + { + "epoch": 0.2972404256598329, + "grad_norm": 1.977945077918547, + "learning_rate": 3.2966141486439682e-06, + "loss": 0.8899, + "step": 2472 + }, + { + "epoch": 0.29736066855047194, + "grad_norm": 3.285180344646331, + "learning_rate": 3.29602095768008e-06, + "loss": 0.6526, + "step": 2473 + }, + { + "epoch": 0.29748091144111105, + "grad_norm": 2.683511746637611, + "learning_rate": 3.2954275701108437e-06, + "loss": 0.65, + "step": 2474 + }, + { + "epoch": 0.29760115433175016, + "grad_norm": 2.4811664711371377, + "learning_rate": 3.294833986026275e-06, + "loss": 0.6956, + "step": 2475 + }, + { + "epoch": 0.2977213972223892, + "grad_norm": 1.9982355202138673, + "learning_rate": 3.29424020551642e-06, + "loss": 0.8558, + "step": 2476 + }, + { + "epoch": 0.2978416401130283, + "grad_norm": 1.9134013376988284, + "learning_rate": 3.2936462286713546e-06, + "loss": 0.7254, + "step": 2477 + }, + { + "epoch": 0.2979618830036674, + "grad_norm": 2.0523385084511485, + "learning_rate": 3.2930520555811846e-06, + "loss": 0.7809, + "step": 2478 + }, + { + "epoch": 0.2980821258943065, + "grad_norm": 1.9262936666738912, + "learning_rate": 3.292457686336046e-06, + "loss": 0.8085, + "step": 2479 + }, + { + "epoch": 0.2982023687849456, + "grad_norm": 0.8805411497743435, + "learning_rate": 3.291863121026105e-06, + "loss": 0.6131, + "step": 2480 + }, + { + "epoch": 0.29832261167558466, + "grad_norm": 2.1053120230976714, + "learning_rate": 3.2912683597415547e-06, + "loss": 0.7672, + "step": 2481 + }, + { + "epoch": 0.29844285456622377, + "grad_norm": 2.0653087102572707, + "learning_rate": 3.2906734025726213e-06, + "loss": 0.7903, + "step": 2482 + }, + { + "epoch": 0.2985630974568629, + "grad_norm": 1.8511991506492163, + "learning_rate": 3.290078249609559e-06, + "loss": 0.8821, + "step": 2483 + }, + { + "epoch": 0.29868334034750194, + "grad_norm": 2.954553329350574, + "learning_rate": 3.2894829009426514e-06, + "loss": 0.895, + "step": 2484 + }, + { + "epoch": 0.29880358323814105, + "grad_norm": 2.8189432724965338, + "learning_rate": 3.288887356662213e-06, + "loss": 0.7804, + "step": 2485 + }, + { + "epoch": 0.29892382612878016, + "grad_norm": 1.0625107147574109, + "learning_rate": 3.288291616858588e-06, + "loss": 0.5978, + "step": 2486 + }, + { + "epoch": 0.2990440690194192, + "grad_norm": 1.7914970524662779, + "learning_rate": 3.287695681622149e-06, + "loss": 0.7709, + "step": 2487 + }, + { + "epoch": 0.2991643119100583, + "grad_norm": 1.7861694532403938, + "learning_rate": 3.2870995510432982e-06, + "loss": 0.8134, + "step": 2488 + }, + { + "epoch": 0.29928455480069743, + "grad_norm": 1.9758551617331936, + "learning_rate": 3.2865032252124697e-06, + "loss": 0.774, + "step": 2489 + }, + { + "epoch": 0.2994047976913365, + "grad_norm": 1.4977855230584571, + "learning_rate": 3.2859067042201243e-06, + "loss": 0.784, + "step": 2490 + }, + { + "epoch": 0.2995250405819756, + "grad_norm": 2.3456181519528085, + "learning_rate": 3.2853099881567544e-06, + "loss": 0.7867, + "step": 2491 + }, + { + "epoch": 0.29964528347261465, + "grad_norm": 1.790211737379545, + "learning_rate": 3.284713077112881e-06, + "loss": 0.7993, + "step": 2492 + }, + { + "epoch": 0.29976552636325376, + "grad_norm": 2.6667602840521036, + "learning_rate": 3.284115971179056e-06, + "loss": 0.873, + "step": 2493 + }, + { + "epoch": 0.2998857692538929, + "grad_norm": 2.285609012972399, + "learning_rate": 3.283518670445859e-06, + "loss": 0.7922, + "step": 2494 + }, + { + "epoch": 0.30000601214453193, + "grad_norm": 0.6922249984056221, + "learning_rate": 3.2829211750038995e-06, + "loss": 0.5441, + "step": 2495 + }, + { + "epoch": 0.30012625503517104, + "grad_norm": 1.7704511865901615, + "learning_rate": 3.2823234849438183e-06, + "loss": 0.896, + "step": 2496 + }, + { + "epoch": 0.30024649792581015, + "grad_norm": 2.407519315362852, + "learning_rate": 3.2817256003562836e-06, + "loss": 0.7655, + "step": 2497 + }, + { + "epoch": 0.3003667408164492, + "grad_norm": 2.0043141803131914, + "learning_rate": 3.281127521331995e-06, + "loss": 0.6654, + "step": 2498 + }, + { + "epoch": 0.3004869837070883, + "grad_norm": 0.8887846822718313, + "learning_rate": 3.2805292479616798e-06, + "loss": 0.6071, + "step": 2499 + }, + { + "epoch": 0.30060722659772743, + "grad_norm": 3.452695655886179, + "learning_rate": 3.2799307803360955e-06, + "loss": 0.9219, + "step": 2500 + }, + { + "epoch": 0.3007274694883665, + "grad_norm": 1.5733622831305132, + "learning_rate": 3.27933211854603e-06, + "loss": 0.8162, + "step": 2501 + }, + { + "epoch": 0.3008477123790056, + "grad_norm": 1.5510684883594485, + "learning_rate": 3.278733262682299e-06, + "loss": 0.8768, + "step": 2502 + }, + { + "epoch": 0.3009679552696447, + "grad_norm": 2.3501194172812108, + "learning_rate": 3.2781342128357484e-06, + "loss": 0.829, + "step": 2503 + }, + { + "epoch": 0.30108819816028376, + "grad_norm": 3.4590786663108086, + "learning_rate": 3.2775349690972547e-06, + "loss": 0.8081, + "step": 2504 + }, + { + "epoch": 0.30120844105092287, + "grad_norm": 0.7622711526119627, + "learning_rate": 3.276935531557722e-06, + "loss": 0.5187, + "step": 2505 + }, + { + "epoch": 0.301328683941562, + "grad_norm": 2.0350770550510164, + "learning_rate": 3.2763359003080837e-06, + "loss": 0.7951, + "step": 2506 + }, + { + "epoch": 0.30144892683220104, + "grad_norm": 0.8073615300291753, + "learning_rate": 3.2757360754393047e-06, + "loss": 0.6246, + "step": 2507 + }, + { + "epoch": 0.30156916972284015, + "grad_norm": 3.136969873392517, + "learning_rate": 3.2751360570423767e-06, + "loss": 0.646, + "step": 2508 + }, + { + "epoch": 0.3016894126134792, + "grad_norm": 2.204270950859873, + "learning_rate": 3.2745358452083236e-06, + "loss": 0.7682, + "step": 2509 + }, + { + "epoch": 0.3018096555041183, + "grad_norm": 1.4561544034957516, + "learning_rate": 3.2739354400281955e-06, + "loss": 0.825, + "step": 2510 + }, + { + "epoch": 0.3019298983947574, + "grad_norm": 0.8771364831715037, + "learning_rate": 3.2733348415930744e-06, + "loss": 0.637, + "step": 2511 + }, + { + "epoch": 0.3020501412853965, + "grad_norm": 2.6040263837541273, + "learning_rate": 3.27273404999407e-06, + "loss": 0.8115, + "step": 2512 + }, + { + "epoch": 0.3021703841760356, + "grad_norm": 0.8279761414522818, + "learning_rate": 3.272133065322322e-06, + "loss": 0.6056, + "step": 2513 + }, + { + "epoch": 0.3022906270666747, + "grad_norm": 2.3065036665190393, + "learning_rate": 3.271531887669e-06, + "loss": 0.7984, + "step": 2514 + }, + { + "epoch": 0.30241086995731375, + "grad_norm": 3.691867246553472, + "learning_rate": 3.2709305171253015e-06, + "loss": 0.6451, + "step": 2515 + }, + { + "epoch": 0.30253111284795287, + "grad_norm": 1.908833550883533, + "learning_rate": 3.2703289537824536e-06, + "loss": 0.7781, + "step": 2516 + }, + { + "epoch": 0.302651355738592, + "grad_norm": 3.224038381409602, + "learning_rate": 3.269727197731714e-06, + "loss": 0.7932, + "step": 2517 + }, + { + "epoch": 0.30277159862923103, + "grad_norm": 1.7649757069112704, + "learning_rate": 3.269125249064367e-06, + "loss": 0.7824, + "step": 2518 + }, + { + "epoch": 0.30289184151987014, + "grad_norm": 1.7015939532719089, + "learning_rate": 3.2685231078717297e-06, + "loss": 0.8405, + "step": 2519 + }, + { + "epoch": 0.30301208441050925, + "grad_norm": 1.95928472040196, + "learning_rate": 3.267920774245145e-06, + "loss": 0.7549, + "step": 2520 + }, + { + "epoch": 0.3031323273011483, + "grad_norm": 2.162621634706335, + "learning_rate": 3.2673182482759876e-06, + "loss": 0.8508, + "step": 2521 + }, + { + "epoch": 0.3032525701917874, + "grad_norm": 2.998577575433244, + "learning_rate": 3.266715530055659e-06, + "loss": 0.6717, + "step": 2522 + }, + { + "epoch": 0.30337281308242653, + "grad_norm": 1.7932590756529017, + "learning_rate": 3.2661126196755927e-06, + "loss": 0.8052, + "step": 2523 + }, + { + "epoch": 0.3034930559730656, + "grad_norm": 0.7881589569714444, + "learning_rate": 3.265509517227248e-06, + "loss": 0.5597, + "step": 2524 + }, + { + "epoch": 0.3036132988637047, + "grad_norm": 1.8341156056408623, + "learning_rate": 3.264906222802115e-06, + "loss": 0.8185, + "step": 2525 + }, + { + "epoch": 0.30373354175434375, + "grad_norm": 2.7518041934560524, + "learning_rate": 3.264302736491715e-06, + "loss": 0.7926, + "step": 2526 + }, + { + "epoch": 0.30385378464498286, + "grad_norm": 2.1745896971853647, + "learning_rate": 3.263699058387594e-06, + "loss": 0.8802, + "step": 2527 + }, + { + "epoch": 0.30397402753562197, + "grad_norm": 2.514027722238479, + "learning_rate": 3.2630951885813315e-06, + "loss": 0.9147, + "step": 2528 + }, + { + "epoch": 0.304094270426261, + "grad_norm": 2.1468734319200857, + "learning_rate": 3.262491127164533e-06, + "loss": 0.7867, + "step": 2529 + }, + { + "epoch": 0.30421451331690014, + "grad_norm": 2.3049508671685466, + "learning_rate": 3.2618868742288337e-06, + "loss": 0.8134, + "step": 2530 + }, + { + "epoch": 0.30433475620753925, + "grad_norm": 2.2842271790902453, + "learning_rate": 3.261282429865899e-06, + "loss": 0.7326, + "step": 2531 + }, + { + "epoch": 0.3044549990981783, + "grad_norm": 2.4238429779900974, + "learning_rate": 3.2606777941674225e-06, + "loss": 0.7294, + "step": 2532 + }, + { + "epoch": 0.3045752419888174, + "grad_norm": 2.3949272985712726, + "learning_rate": 3.2600729672251276e-06, + "loss": 0.847, + "step": 2533 + }, + { + "epoch": 0.3046954848794565, + "grad_norm": 2.2416873600038922, + "learning_rate": 3.259467949130765e-06, + "loss": 0.6607, + "step": 2534 + }, + { + "epoch": 0.3048157277700956, + "grad_norm": 2.542105860885179, + "learning_rate": 3.2588627399761164e-06, + "loss": 0.8321, + "step": 2535 + }, + { + "epoch": 0.3049359706607347, + "grad_norm": 2.0438545596939233, + "learning_rate": 3.2582573398529903e-06, + "loss": 0.7143, + "step": 2536 + }, + { + "epoch": 0.3050562135513738, + "grad_norm": 2.027419483923824, + "learning_rate": 3.2576517488532265e-06, + "loss": 0.7432, + "step": 2537 + }, + { + "epoch": 0.30517645644201286, + "grad_norm": 1.7489006813033594, + "learning_rate": 3.257045967068692e-06, + "loss": 0.88, + "step": 2538 + }, + { + "epoch": 0.30529669933265197, + "grad_norm": 2.179613228356313, + "learning_rate": 3.2564399945912848e-06, + "loss": 0.8282, + "step": 2539 + }, + { + "epoch": 0.305416942223291, + "grad_norm": 2.3264522815879674, + "learning_rate": 3.2558338315129287e-06, + "loss": 0.8286, + "step": 2540 + }, + { + "epoch": 0.30553718511393013, + "grad_norm": 1.9560232402770439, + "learning_rate": 3.2552274779255785e-06, + "loss": 0.7663, + "step": 2541 + }, + { + "epoch": 0.30565742800456924, + "grad_norm": 1.9958675129405323, + "learning_rate": 3.2546209339212184e-06, + "loss": 0.7698, + "step": 2542 + }, + { + "epoch": 0.3057776708952083, + "grad_norm": 1.7157931469507532, + "learning_rate": 3.25401419959186e-06, + "loss": 0.7804, + "step": 2543 + }, + { + "epoch": 0.3058979137858474, + "grad_norm": 1.9272499905941411, + "learning_rate": 3.253407275029545e-06, + "loss": 0.7679, + "step": 2544 + }, + { + "epoch": 0.3060181566764865, + "grad_norm": 1.961751578909709, + "learning_rate": 3.2528001603263425e-06, + "loss": 0.8059, + "step": 2545 + }, + { + "epoch": 0.3061383995671256, + "grad_norm": 1.805512871196952, + "learning_rate": 3.2521928555743514e-06, + "loss": 0.8177, + "step": 2546 + }, + { + "epoch": 0.3062586424577647, + "grad_norm": 2.319850066371604, + "learning_rate": 3.2515853608657e-06, + "loss": 0.6736, + "step": 2547 + }, + { + "epoch": 0.3063788853484038, + "grad_norm": 2.866262510183377, + "learning_rate": 3.250977676292545e-06, + "loss": 0.7597, + "step": 2548 + }, + { + "epoch": 0.30649912823904285, + "grad_norm": 2.3905463112241057, + "learning_rate": 3.2503698019470712e-06, + "loss": 0.7942, + "step": 2549 + }, + { + "epoch": 0.30661937112968196, + "grad_norm": 2.0099220919337424, + "learning_rate": 3.249761737921492e-06, + "loss": 0.791, + "step": 2550 + }, + { + "epoch": 0.30673961402032107, + "grad_norm": 2.075424275052759, + "learning_rate": 3.249153484308051e-06, + "loss": 0.7484, + "step": 2551 + }, + { + "epoch": 0.3068598569109601, + "grad_norm": 1.906972904398899, + "learning_rate": 3.2485450411990194e-06, + "loss": 0.7814, + "step": 2552 + }, + { + "epoch": 0.30698009980159924, + "grad_norm": 1.8162572656054268, + "learning_rate": 3.2479364086866983e-06, + "loss": 0.8294, + "step": 2553 + }, + { + "epoch": 0.30710034269223835, + "grad_norm": 1.8064738098661774, + "learning_rate": 3.247327586863416e-06, + "loss": 0.8208, + "step": 2554 + }, + { + "epoch": 0.3072205855828774, + "grad_norm": 1.9496535922741645, + "learning_rate": 3.2467185758215304e-06, + "loss": 0.7823, + "step": 2555 + }, + { + "epoch": 0.3073408284735165, + "grad_norm": 2.283804521416186, + "learning_rate": 3.246109375653428e-06, + "loss": 0.8566, + "step": 2556 + }, + { + "epoch": 0.30746107136415557, + "grad_norm": 1.9539967536938347, + "learning_rate": 3.2454999864515243e-06, + "loss": 0.786, + "step": 2557 + }, + { + "epoch": 0.3075813142547947, + "grad_norm": 1.7540970254637, + "learning_rate": 3.244890408308263e-06, + "loss": 0.6949, + "step": 2558 + }, + { + "epoch": 0.3077015571454338, + "grad_norm": 3.2774108225127843, + "learning_rate": 3.2442806413161165e-06, + "loss": 0.6213, + "step": 2559 + }, + { + "epoch": 0.30782180003607285, + "grad_norm": 2.2286984555054796, + "learning_rate": 3.243670685567586e-06, + "loss": 0.7637, + "step": 2560 + }, + { + "epoch": 0.30794204292671196, + "grad_norm": 2.2112196887123785, + "learning_rate": 3.2430605411552012e-06, + "loss": 0.808, + "step": 2561 + }, + { + "epoch": 0.30806228581735107, + "grad_norm": 0.9196147365476968, + "learning_rate": 3.2424502081715205e-06, + "loss": 0.7061, + "step": 2562 + }, + { + "epoch": 0.3081825287079901, + "grad_norm": 1.7464193770144034, + "learning_rate": 3.241839686709132e-06, + "loss": 0.787, + "step": 2563 + }, + { + "epoch": 0.30830277159862923, + "grad_norm": 2.322936966013346, + "learning_rate": 3.2412289768606495e-06, + "loss": 0.8231, + "step": 2564 + }, + { + "epoch": 0.30842301448926834, + "grad_norm": 1.7514741683043884, + "learning_rate": 3.240618078718718e-06, + "loss": 0.8323, + "step": 2565 + }, + { + "epoch": 0.3085432573799074, + "grad_norm": 2.1383224689607054, + "learning_rate": 3.240006992376011e-06, + "loss": 0.7463, + "step": 2566 + }, + { + "epoch": 0.3086635002705465, + "grad_norm": 2.4460857497224247, + "learning_rate": 3.2393957179252284e-06, + "loss": 0.7694, + "step": 2567 + }, + { + "epoch": 0.3087837431611856, + "grad_norm": 1.818052075373569, + "learning_rate": 3.2387842554591016e-06, + "loss": 0.8134, + "step": 2568 + }, + { + "epoch": 0.3089039860518247, + "grad_norm": 2.3546033696141357, + "learning_rate": 3.238172605070388e-06, + "loss": 0.8822, + "step": 2569 + }, + { + "epoch": 0.3090242289424638, + "grad_norm": 2.3304243573964967, + "learning_rate": 3.2375607668518745e-06, + "loss": 0.799, + "step": 2570 + }, + { + "epoch": 0.30914447183310284, + "grad_norm": 2.2498493144121974, + "learning_rate": 3.236948740896377e-06, + "loss": 0.908, + "step": 2571 + }, + { + "epoch": 0.30926471472374195, + "grad_norm": 1.7512800439821015, + "learning_rate": 3.2363365272967384e-06, + "loss": 0.8444, + "step": 2572 + }, + { + "epoch": 0.30938495761438106, + "grad_norm": 1.9100809843060755, + "learning_rate": 3.235724126145832e-06, + "loss": 0.8218, + "step": 2573 + }, + { + "epoch": 0.3095052005050201, + "grad_norm": 1.7861350817115849, + "learning_rate": 3.235111537536558e-06, + "loss": 0.7791, + "step": 2574 + }, + { + "epoch": 0.30962544339565923, + "grad_norm": 1.8895466148681068, + "learning_rate": 3.2344987615618456e-06, + "loss": 0.8332, + "step": 2575 + }, + { + "epoch": 0.30974568628629834, + "grad_norm": 1.568067473614143, + "learning_rate": 3.2338857983146533e-06, + "loss": 0.7892, + "step": 2576 + }, + { + "epoch": 0.3098659291769374, + "grad_norm": 2.326614811746622, + "learning_rate": 3.233272647887966e-06, + "loss": 0.7688, + "step": 2577 + }, + { + "epoch": 0.3099861720675765, + "grad_norm": 1.666221543317333, + "learning_rate": 3.2326593103747985e-06, + "loss": 0.9031, + "step": 2578 + }, + { + "epoch": 0.3101064149582156, + "grad_norm": 11.668621952966426, + "learning_rate": 3.2320457858681936e-06, + "loss": 0.8597, + "step": 2579 + }, + { + "epoch": 0.31022665784885467, + "grad_norm": 2.1734085706352797, + "learning_rate": 3.2314320744612228e-06, + "loss": 0.8568, + "step": 2580 + }, + { + "epoch": 0.3103469007394938, + "grad_norm": 1.9684309928024637, + "learning_rate": 3.2308181762469854e-06, + "loss": 0.7688, + "step": 2581 + }, + { + "epoch": 0.3104671436301329, + "grad_norm": 2.10494031626704, + "learning_rate": 3.230204091318609e-06, + "loss": 0.7967, + "step": 2582 + }, + { + "epoch": 0.31058738652077195, + "grad_norm": 1.9310414124467845, + "learning_rate": 3.2295898197692503e-06, + "loss": 0.8525, + "step": 2583 + }, + { + "epoch": 0.31070762941141106, + "grad_norm": 2.248308393577134, + "learning_rate": 3.228975361692094e-06, + "loss": 0.7985, + "step": 2584 + }, + { + "epoch": 0.31082787230205017, + "grad_norm": 2.7078264397880334, + "learning_rate": 3.228360717180352e-06, + "loss": 0.8097, + "step": 2585 + }, + { + "epoch": 0.3109481151926892, + "grad_norm": 0.8217236325078483, + "learning_rate": 3.227745886327266e-06, + "loss": 0.5944, + "step": 2586 + }, + { + "epoch": 0.31106835808332833, + "grad_norm": 0.8196441470805945, + "learning_rate": 3.227130869226105e-06, + "loss": 0.5589, + "step": 2587 + }, + { + "epoch": 0.3111886009739674, + "grad_norm": 2.338111503702841, + "learning_rate": 3.226515665970167e-06, + "loss": 0.8345, + "step": 2588 + }, + { + "epoch": 0.3113088438646065, + "grad_norm": 2.1292663829567786, + "learning_rate": 3.225900276652777e-06, + "loss": 0.8705, + "step": 2589 + }, + { + "epoch": 0.3114290867552456, + "grad_norm": 1.522267758174867, + "learning_rate": 3.2252847013672906e-06, + "loss": 0.7673, + "step": 2590 + }, + { + "epoch": 0.31154932964588467, + "grad_norm": 2.107198312690077, + "learning_rate": 3.224668940207089e-06, + "loss": 0.7682, + "step": 2591 + }, + { + "epoch": 0.3116695725365238, + "grad_norm": 1.824149871767446, + "learning_rate": 3.2240529932655828e-06, + "loss": 0.8685, + "step": 2592 + }, + { + "epoch": 0.3117898154271629, + "grad_norm": 2.676691454594071, + "learning_rate": 3.223436860636211e-06, + "loss": 0.8945, + "step": 2593 + }, + { + "epoch": 0.31191005831780194, + "grad_norm": 1.7255547115348153, + "learning_rate": 3.2228205424124403e-06, + "loss": 0.7486, + "step": 2594 + }, + { + "epoch": 0.31203030120844105, + "grad_norm": 3.0865938285434225, + "learning_rate": 3.222204038687765e-06, + "loss": 0.7583, + "step": 2595 + }, + { + "epoch": 0.31215054409908016, + "grad_norm": 1.637758420980504, + "learning_rate": 3.221587349555709e-06, + "loss": 0.8881, + "step": 2596 + }, + { + "epoch": 0.3122707869897192, + "grad_norm": 1.6529412237749896, + "learning_rate": 3.2209704751098236e-06, + "loss": 0.6994, + "step": 2597 + }, + { + "epoch": 0.31239102988035833, + "grad_norm": 2.291858514789415, + "learning_rate": 3.2203534154436875e-06, + "loss": 0.8362, + "step": 2598 + }, + { + "epoch": 0.31251127277099744, + "grad_norm": 2.254343079757924, + "learning_rate": 3.2197361706509084e-06, + "loss": 0.7629, + "step": 2599 + }, + { + "epoch": 0.3126315156616365, + "grad_norm": 2.8544945261707833, + "learning_rate": 3.2191187408251228e-06, + "loss": 0.846, + "step": 2600 + }, + { + "epoch": 0.3127517585522756, + "grad_norm": 2.4212045049061093, + "learning_rate": 3.218501126059993e-06, + "loss": 0.7965, + "step": 2601 + }, + { + "epoch": 0.31287200144291466, + "grad_norm": 2.3343753758523373, + "learning_rate": 3.2178833264492116e-06, + "loss": 0.8171, + "step": 2602 + }, + { + "epoch": 0.31299224433355377, + "grad_norm": 1.784189642850749, + "learning_rate": 3.217265342086498e-06, + "loss": 0.7651, + "step": 2603 + }, + { + "epoch": 0.3131124872241929, + "grad_norm": 2.5526576012388733, + "learning_rate": 3.216647173065599e-06, + "loss": 0.7434, + "step": 2604 + }, + { + "epoch": 0.31323273011483194, + "grad_norm": 1.7186592771688445, + "learning_rate": 3.216028819480292e-06, + "loss": 0.7425, + "step": 2605 + }, + { + "epoch": 0.31335297300547105, + "grad_norm": 3.683722290514752, + "learning_rate": 3.2154102814243793e-06, + "loss": 0.7572, + "step": 2606 + }, + { + "epoch": 0.31347321589611016, + "grad_norm": 2.0547752216599875, + "learning_rate": 3.2147915589916937e-06, + "loss": 0.6742, + "step": 2607 + }, + { + "epoch": 0.3135934587867492, + "grad_norm": 1.8549598397485803, + "learning_rate": 3.2141726522760938e-06, + "loss": 0.8316, + "step": 2608 + }, + { + "epoch": 0.3137137016773883, + "grad_norm": 0.7098237420633047, + "learning_rate": 3.213553561371469e-06, + "loss": 0.5269, + "step": 2609 + }, + { + "epoch": 0.31383394456802743, + "grad_norm": 2.5761348556854964, + "learning_rate": 3.212934286371733e-06, + "loss": 0.9697, + "step": 2610 + }, + { + "epoch": 0.3139541874586665, + "grad_norm": 2.0995587294037126, + "learning_rate": 3.2123148273708304e-06, + "loss": 0.8304, + "step": 2611 + }, + { + "epoch": 0.3140744303493056, + "grad_norm": 1.9256207319916225, + "learning_rate": 3.211695184462733e-06, + "loss": 0.7703, + "step": 2612 + }, + { + "epoch": 0.3141946732399447, + "grad_norm": 0.8970789049543927, + "learning_rate": 3.2110753577414383e-06, + "loss": 0.6052, + "step": 2613 + }, + { + "epoch": 0.31431491613058377, + "grad_norm": 2.3950656711438616, + "learning_rate": 3.2104553473009757e-06, + "loss": 0.7995, + "step": 2614 + }, + { + "epoch": 0.3144351590212229, + "grad_norm": 2.0352552393527312, + "learning_rate": 3.209835153235399e-06, + "loss": 0.6863, + "step": 2615 + }, + { + "epoch": 0.314555401911862, + "grad_norm": 1.8807637951928216, + "learning_rate": 3.2092147756387916e-06, + "loss": 0.6841, + "step": 2616 + }, + { + "epoch": 0.31467564480250104, + "grad_norm": 2.2464799897189747, + "learning_rate": 3.208594214605264e-06, + "loss": 0.8401, + "step": 2617 + }, + { + "epoch": 0.31479588769314015, + "grad_norm": 2.0179569446367567, + "learning_rate": 3.2079734702289553e-06, + "loss": 0.7742, + "step": 2618 + }, + { + "epoch": 0.3149161305837792, + "grad_norm": 0.8167094712490216, + "learning_rate": 3.207352542604031e-06, + "loss": 0.6042, + "step": 2619 + }, + { + "epoch": 0.3150363734744183, + "grad_norm": 1.6464048753327012, + "learning_rate": 3.2067314318246864e-06, + "loss": 0.7884, + "step": 2620 + }, + { + "epoch": 0.31515661636505743, + "grad_norm": 1.8914064416014722, + "learning_rate": 3.206110137985143e-06, + "loss": 0.7736, + "step": 2621 + }, + { + "epoch": 0.3152768592556965, + "grad_norm": 1.9324700683605045, + "learning_rate": 3.2054886611796505e-06, + "loss": 0.9223, + "step": 2622 + }, + { + "epoch": 0.3153971021463356, + "grad_norm": 0.9257041158613699, + "learning_rate": 3.204867001502487e-06, + "loss": 0.6363, + "step": 2623 + }, + { + "epoch": 0.3155173450369747, + "grad_norm": 4.332991415374936, + "learning_rate": 3.2042451590479567e-06, + "loss": 0.8109, + "step": 2624 + }, + { + "epoch": 0.31563758792761376, + "grad_norm": 1.9565992049762195, + "learning_rate": 3.203623133910394e-06, + "loss": 0.8722, + "step": 2625 + }, + { + "epoch": 0.31575783081825287, + "grad_norm": 4.826584060833196, + "learning_rate": 3.203000926184158e-06, + "loss": 0.7768, + "step": 2626 + }, + { + "epoch": 0.315878073708892, + "grad_norm": 1.7357537228108806, + "learning_rate": 3.202378535963639e-06, + "loss": 0.7813, + "step": 2627 + }, + { + "epoch": 0.31599831659953104, + "grad_norm": 2.5183592916884776, + "learning_rate": 3.2017559633432516e-06, + "loss": 0.8437, + "step": 2628 + }, + { + "epoch": 0.31611855949017015, + "grad_norm": 1.9832859204606288, + "learning_rate": 3.2011332084174398e-06, + "loss": 0.6715, + "step": 2629 + }, + { + "epoch": 0.31623880238080926, + "grad_norm": 1.7602507232764217, + "learning_rate": 3.2005102712806756e-06, + "loss": 0.8982, + "step": 2630 + }, + { + "epoch": 0.3163590452714483, + "grad_norm": 2.233170084419466, + "learning_rate": 3.1998871520274575e-06, + "loss": 0.7345, + "step": 2631 + }, + { + "epoch": 0.3164792881620874, + "grad_norm": 3.67340990816347, + "learning_rate": 3.199263850752312e-06, + "loss": 0.8546, + "step": 2632 + }, + { + "epoch": 0.31659953105272653, + "grad_norm": 2.1905383126905664, + "learning_rate": 3.198640367549795e-06, + "loss": 0.8596, + "step": 2633 + }, + { + "epoch": 0.3167197739433656, + "grad_norm": 1.7609425122410574, + "learning_rate": 3.198016702514487e-06, + "loss": 0.8602, + "step": 2634 + }, + { + "epoch": 0.3168400168340047, + "grad_norm": 1.924686639224735, + "learning_rate": 3.1973928557409972e-06, + "loss": 0.8472, + "step": 2635 + }, + { + "epoch": 0.31696025972464376, + "grad_norm": 2.213080779320339, + "learning_rate": 3.1967688273239636e-06, + "loss": 0.7163, + "step": 2636 + }, + { + "epoch": 0.31708050261528287, + "grad_norm": 1.9368661028105878, + "learning_rate": 3.1961446173580503e-06, + "loss": 0.8279, + "step": 2637 + }, + { + "epoch": 0.317200745505922, + "grad_norm": 1.8849656073066055, + "learning_rate": 3.1955202259379502e-06, + "loss": 0.7751, + "step": 2638 + }, + { + "epoch": 0.31732098839656103, + "grad_norm": 1.92389975477374, + "learning_rate": 3.194895653158381e-06, + "loss": 0.8314, + "step": 2639 + }, + { + "epoch": 0.31744123128720014, + "grad_norm": 0.7776100047753195, + "learning_rate": 3.194270899114093e-06, + "loss": 0.556, + "step": 2640 + }, + { + "epoch": 0.31756147417783925, + "grad_norm": 2.0061320237502587, + "learning_rate": 3.193645963899858e-06, + "loss": 0.8248, + "step": 2641 + }, + { + "epoch": 0.3176817170684783, + "grad_norm": 2.0606985605824715, + "learning_rate": 3.193020847610479e-06, + "loss": 0.8446, + "step": 2642 + }, + { + "epoch": 0.3178019599591174, + "grad_norm": 2.265467144958782, + "learning_rate": 3.192395550340787e-06, + "loss": 0.715, + "step": 2643 + }, + { + "epoch": 0.31792220284975653, + "grad_norm": 2.200831958212297, + "learning_rate": 3.191770072185638e-06, + "loss": 0.7744, + "step": 2644 + }, + { + "epoch": 0.3180424457403956, + "grad_norm": 2.597649458109072, + "learning_rate": 3.191144413239916e-06, + "loss": 0.7323, + "step": 2645 + }, + { + "epoch": 0.3181626886310347, + "grad_norm": 2.4321459141802677, + "learning_rate": 3.190518573598534e-06, + "loss": 0.8852, + "step": 2646 + }, + { + "epoch": 0.3182829315216738, + "grad_norm": 1.6020438610649457, + "learning_rate": 3.1898925533564308e-06, + "loss": 0.7782, + "step": 2647 + }, + { + "epoch": 0.31840317441231286, + "grad_norm": 4.508166638087769, + "learning_rate": 3.1892663526085733e-06, + "loss": 0.6468, + "step": 2648 + }, + { + "epoch": 0.31852341730295197, + "grad_norm": 0.7583618532483958, + "learning_rate": 3.188639971449956e-06, + "loss": 0.5703, + "step": 2649 + }, + { + "epoch": 0.318643660193591, + "grad_norm": 2.025785402175291, + "learning_rate": 3.1880134099756e-06, + "loss": 0.7274, + "step": 2650 + }, + { + "epoch": 0.31876390308423014, + "grad_norm": 1.8306832305513556, + "learning_rate": 3.1873866682805535e-06, + "loss": 0.698, + "step": 2651 + }, + { + "epoch": 0.31888414597486925, + "grad_norm": 1.7501164806292933, + "learning_rate": 3.186759746459894e-06, + "loss": 0.8927, + "step": 2652 + }, + { + "epoch": 0.3190043888655083, + "grad_norm": 1.8735394193209578, + "learning_rate": 3.1861326446087246e-06, + "loss": 0.7979, + "step": 2653 + }, + { + "epoch": 0.3191246317561474, + "grad_norm": 2.3782387785192194, + "learning_rate": 3.1855053628221763e-06, + "loss": 0.7236, + "step": 2654 + }, + { + "epoch": 0.3192448746467865, + "grad_norm": 2.7195856848893056, + "learning_rate": 3.184877901195407e-06, + "loss": 0.9082, + "step": 2655 + }, + { + "epoch": 0.3193651175374256, + "grad_norm": 0.8061654305276732, + "learning_rate": 3.184250259823602e-06, + "loss": 0.6285, + "step": 2656 + }, + { + "epoch": 0.3194853604280647, + "grad_norm": 2.2468453650715388, + "learning_rate": 3.183622438801974e-06, + "loss": 0.8198, + "step": 2657 + }, + { + "epoch": 0.3196056033187038, + "grad_norm": 2.358569986805546, + "learning_rate": 3.1829944382257637e-06, + "loss": 0.758, + "step": 2658 + }, + { + "epoch": 0.31972584620934286, + "grad_norm": 2.4663906170624794, + "learning_rate": 3.1823662581902373e-06, + "loss": 0.8252, + "step": 2659 + }, + { + "epoch": 0.31984608909998197, + "grad_norm": 2.7657188787529154, + "learning_rate": 3.1817378987906896e-06, + "loss": 0.7505, + "step": 2660 + }, + { + "epoch": 0.3199663319906211, + "grad_norm": 2.2926257005214894, + "learning_rate": 3.181109360122442e-06, + "loss": 0.8097, + "step": 2661 + }, + { + "epoch": 0.32008657488126013, + "grad_norm": 2.7303293872025898, + "learning_rate": 3.1804806422808445e-06, + "loss": 0.7936, + "step": 2662 + }, + { + "epoch": 0.32020681777189924, + "grad_norm": 1.650436563247494, + "learning_rate": 3.1798517453612714e-06, + "loss": 0.7397, + "step": 2663 + }, + { + "epoch": 0.32032706066253835, + "grad_norm": 1.861988186758819, + "learning_rate": 3.1792226694591265e-06, + "loss": 0.762, + "step": 2664 + }, + { + "epoch": 0.3204473035531774, + "grad_norm": 2.011138654605723, + "learning_rate": 3.178593414669841e-06, + "loss": 0.8113, + "step": 2665 + }, + { + "epoch": 0.3205675464438165, + "grad_norm": 5.482261837484263, + "learning_rate": 3.1779639810888707e-06, + "loss": 0.7068, + "step": 2666 + }, + { + "epoch": 0.3206877893344556, + "grad_norm": 2.9899017129008802, + "learning_rate": 3.1773343688117013e-06, + "loss": 0.7676, + "step": 2667 + }, + { + "epoch": 0.3208080322250947, + "grad_norm": 2.1480468394821783, + "learning_rate": 3.1767045779338445e-06, + "loss": 0.8464, + "step": 2668 + }, + { + "epoch": 0.3209282751157338, + "grad_norm": 2.43385100497473, + "learning_rate": 3.176074608550839e-06, + "loss": 0.9182, + "step": 2669 + }, + { + "epoch": 0.32104851800637285, + "grad_norm": 2.411123517537268, + "learning_rate": 3.17544446075825e-06, + "loss": 0.8331, + "step": 2670 + }, + { + "epoch": 0.32116876089701196, + "grad_norm": 1.671947370727556, + "learning_rate": 3.174814134651671e-06, + "loss": 0.7107, + "step": 2671 + }, + { + "epoch": 0.3212890037876511, + "grad_norm": 1.6419903115045933, + "learning_rate": 3.1741836303267215e-06, + "loss": 0.811, + "step": 2672 + }, + { + "epoch": 0.32140924667829013, + "grad_norm": 1.8091365914090407, + "learning_rate": 3.1735529478790496e-06, + "loss": 0.7574, + "step": 2673 + }, + { + "epoch": 0.32152948956892924, + "grad_norm": 1.9719776157091413, + "learning_rate": 3.172922087404328e-06, + "loss": 0.8048, + "step": 2674 + }, + { + "epoch": 0.32164973245956835, + "grad_norm": 0.7767309716075999, + "learning_rate": 3.1722910489982586e-06, + "loss": 0.5526, + "step": 2675 + }, + { + "epoch": 0.3217699753502074, + "grad_norm": 6.0932159972353555, + "learning_rate": 3.1716598327565694e-06, + "loss": 0.8051, + "step": 2676 + }, + { + "epoch": 0.3218902182408465, + "grad_norm": 1.5917215341161874, + "learning_rate": 3.171028438775015e-06, + "loss": 0.8447, + "step": 2677 + }, + { + "epoch": 0.3220104611314856, + "grad_norm": 2.163886879084966, + "learning_rate": 3.170396867149377e-06, + "loss": 0.8442, + "step": 2678 + }, + { + "epoch": 0.3221307040221247, + "grad_norm": 3.4279512429169583, + "learning_rate": 3.1697651179754653e-06, + "loss": 0.865, + "step": 2679 + }, + { + "epoch": 0.3222509469127638, + "grad_norm": 6.600356132840278, + "learning_rate": 3.1691331913491153e-06, + "loss": 0.736, + "step": 2680 + }, + { + "epoch": 0.32237118980340285, + "grad_norm": 2.5407980284005784, + "learning_rate": 3.1685010873661898e-06, + "loss": 0.8485, + "step": 2681 + }, + { + "epoch": 0.32249143269404196, + "grad_norm": 2.6107541397151035, + "learning_rate": 3.167868806122578e-06, + "loss": 0.8063, + "step": 2682 + }, + { + "epoch": 0.32261167558468107, + "grad_norm": 1.840975338507001, + "learning_rate": 3.1672363477141968e-06, + "loss": 0.6596, + "step": 2683 + }, + { + "epoch": 0.3227319184753201, + "grad_norm": 1.9183689795307828, + "learning_rate": 3.1666037122369903e-06, + "loss": 0.8582, + "step": 2684 + }, + { + "epoch": 0.32285216136595923, + "grad_norm": 2.830378861965002, + "learning_rate": 3.165970899786928e-06, + "loss": 0.8711, + "step": 2685 + }, + { + "epoch": 0.32297240425659834, + "grad_norm": 1.7985504963339671, + "learning_rate": 3.1653379104600067e-06, + "loss": 0.7544, + "step": 2686 + }, + { + "epoch": 0.3230926471472374, + "grad_norm": 1.521757637527424, + "learning_rate": 3.164704744352251e-06, + "loss": 0.7004, + "step": 2687 + }, + { + "epoch": 0.3232128900378765, + "grad_norm": 1.7586087606256313, + "learning_rate": 3.164071401559713e-06, + "loss": 0.8148, + "step": 2688 + }, + { + "epoch": 0.3233331329285156, + "grad_norm": 1.830736495790602, + "learning_rate": 3.1634378821784674e-06, + "loss": 0.7165, + "step": 2689 + }, + { + "epoch": 0.3234533758191547, + "grad_norm": 2.0322275010509605, + "learning_rate": 3.1628041863046208e-06, + "loss": 0.7445, + "step": 2690 + }, + { + "epoch": 0.3235736187097938, + "grad_norm": 2.400219978741552, + "learning_rate": 3.162170314034304e-06, + "loss": 0.9144, + "step": 2691 + }, + { + "epoch": 0.3236938616004329, + "grad_norm": 1.5537694012511651, + "learning_rate": 3.1615362654636738e-06, + "loss": 0.8135, + "step": 2692 + }, + { + "epoch": 0.32381410449107195, + "grad_norm": 1.9131228261821251, + "learning_rate": 3.1609020406889163e-06, + "loss": 0.875, + "step": 2693 + }, + { + "epoch": 0.32393434738171106, + "grad_norm": 1.7230673924316147, + "learning_rate": 3.1602676398062416e-06, + "loss": 0.8577, + "step": 2694 + }, + { + "epoch": 0.3240545902723502, + "grad_norm": 2.2261588818314344, + "learning_rate": 3.1596330629118886e-06, + "loss": 0.6227, + "step": 2695 + }, + { + "epoch": 0.32417483316298923, + "grad_norm": 1.8701952726232385, + "learning_rate": 3.1589983101021223e-06, + "loss": 0.7342, + "step": 2696 + }, + { + "epoch": 0.32429507605362834, + "grad_norm": 2.0074052805900386, + "learning_rate": 3.1583633814732337e-06, + "loss": 0.8535, + "step": 2697 + }, + { + "epoch": 0.3244153189442674, + "grad_norm": 3.357560488339909, + "learning_rate": 3.157728277121541e-06, + "loss": 0.7215, + "step": 2698 + }, + { + "epoch": 0.3245355618349065, + "grad_norm": 2.759330958255773, + "learning_rate": 3.1570929971433897e-06, + "loss": 0.7915, + "step": 2699 + }, + { + "epoch": 0.3246558047255456, + "grad_norm": 2.0578572586278203, + "learning_rate": 3.1564575416351504e-06, + "loss": 0.8452, + "step": 2700 + }, + { + "epoch": 0.32477604761618467, + "grad_norm": 1.9742043039135282, + "learning_rate": 3.155821910693221e-06, + "loss": 0.7482, + "step": 2701 + }, + { + "epoch": 0.3248962905068238, + "grad_norm": 1.9745323400063512, + "learning_rate": 3.1551861044140275e-06, + "loss": 0.8673, + "step": 2702 + }, + { + "epoch": 0.3250165333974629, + "grad_norm": 1.8861569660114053, + "learning_rate": 3.15455012289402e-06, + "loss": 0.7817, + "step": 2703 + }, + { + "epoch": 0.32513677628810195, + "grad_norm": 1.7935661035569248, + "learning_rate": 3.153913966229677e-06, + "loss": 0.8452, + "step": 2704 + }, + { + "epoch": 0.32525701917874106, + "grad_norm": 0.6503599234084855, + "learning_rate": 3.1532776345175027e-06, + "loss": 0.5032, + "step": 2705 + }, + { + "epoch": 0.32537726206938017, + "grad_norm": 1.794869207329684, + "learning_rate": 3.1526411278540285e-06, + "loss": 0.795, + "step": 2706 + }, + { + "epoch": 0.3254975049600192, + "grad_norm": 2.4807619897717395, + "learning_rate": 3.1520044463358116e-06, + "loss": 0.8165, + "step": 2707 + }, + { + "epoch": 0.32561774785065833, + "grad_norm": 1.62523326666714, + "learning_rate": 3.151367590059436e-06, + "loss": 0.8072, + "step": 2708 + }, + { + "epoch": 0.32573799074129745, + "grad_norm": 2.057783571820562, + "learning_rate": 3.1507305591215117e-06, + "loss": 0.8693, + "step": 2709 + }, + { + "epoch": 0.3258582336319365, + "grad_norm": 0.6892187784696965, + "learning_rate": 3.150093353618677e-06, + "loss": 0.5581, + "step": 2710 + }, + { + "epoch": 0.3259784765225756, + "grad_norm": 2.450715746747509, + "learning_rate": 3.149455973647596e-06, + "loss": 0.8941, + "step": 2711 + }, + { + "epoch": 0.32609871941321467, + "grad_norm": 1.8957086109873549, + "learning_rate": 3.1488184193049563e-06, + "loss": 0.7786, + "step": 2712 + }, + { + "epoch": 0.3262189623038538, + "grad_norm": 1.7795119924974545, + "learning_rate": 3.1481806906874767e-06, + "loss": 0.727, + "step": 2713 + }, + { + "epoch": 0.3263392051944929, + "grad_norm": 1.8946891797495724, + "learning_rate": 3.147542787891899e-06, + "loss": 0.8783, + "step": 2714 + }, + { + "epoch": 0.32645944808513194, + "grad_norm": 1.8322719911954224, + "learning_rate": 3.1469047110149926e-06, + "loss": 0.7559, + "step": 2715 + }, + { + "epoch": 0.32657969097577105, + "grad_norm": 2.050436866545509, + "learning_rate": 3.146266460153554e-06, + "loss": 0.856, + "step": 2716 + }, + { + "epoch": 0.32669993386641016, + "grad_norm": 1.7057398050767565, + "learning_rate": 3.145628035404404e-06, + "loss": 0.806, + "step": 2717 + }, + { + "epoch": 0.3268201767570492, + "grad_norm": 0.9080156358368306, + "learning_rate": 3.1449894368643922e-06, + "loss": 0.5755, + "step": 2718 + }, + { + "epoch": 0.32694041964768833, + "grad_norm": 2.1883691423463794, + "learning_rate": 3.1443506646303934e-06, + "loss": 0.7181, + "step": 2719 + }, + { + "epoch": 0.32706066253832744, + "grad_norm": 2.395115841670196, + "learning_rate": 3.1437117187993086e-06, + "loss": 0.6709, + "step": 2720 + }, + { + "epoch": 0.3271809054289665, + "grad_norm": 1.6285391926861534, + "learning_rate": 3.143072599468065e-06, + "loss": 0.8047, + "step": 2721 + }, + { + "epoch": 0.3273011483196056, + "grad_norm": 1.6405944094757845, + "learning_rate": 3.1424333067336174e-06, + "loss": 0.7644, + "step": 2722 + }, + { + "epoch": 0.3274213912102447, + "grad_norm": 1.7906901189603783, + "learning_rate": 3.141793840692945e-06, + "loss": 0.7889, + "step": 2723 + }, + { + "epoch": 0.32754163410088377, + "grad_norm": 2.3521002981443027, + "learning_rate": 3.1411542014430553e-06, + "loss": 0.6191, + "step": 2724 + }, + { + "epoch": 0.3276618769915229, + "grad_norm": 1.9060507498246164, + "learning_rate": 3.1405143890809804e-06, + "loss": 0.8217, + "step": 2725 + }, + { + "epoch": 0.327782119882162, + "grad_norm": 1.7744061349661244, + "learning_rate": 3.1398744037037796e-06, + "loss": 0.7042, + "step": 2726 + }, + { + "epoch": 0.32790236277280105, + "grad_norm": 2.1686106788232307, + "learning_rate": 3.139234245408538e-06, + "loss": 0.8518, + "step": 2727 + }, + { + "epoch": 0.32802260566344016, + "grad_norm": 1.4238694842994184, + "learning_rate": 3.1385939142923666e-06, + "loss": 0.7646, + "step": 2728 + }, + { + "epoch": 0.3281428485540792, + "grad_norm": 4.44259286829579, + "learning_rate": 3.137953410452405e-06, + "loss": 0.7887, + "step": 2729 + }, + { + "epoch": 0.3282630914447183, + "grad_norm": 1.7866239628504224, + "learning_rate": 3.1373127339858146e-06, + "loss": 0.746, + "step": 2730 + }, + { + "epoch": 0.32838333433535744, + "grad_norm": 1.8803657050242508, + "learning_rate": 3.136671884989787e-06, + "loss": 0.7514, + "step": 2731 + }, + { + "epoch": 0.3285035772259965, + "grad_norm": 2.7485015427944384, + "learning_rate": 3.1360308635615383e-06, + "loss": 0.8718, + "step": 2732 + }, + { + "epoch": 0.3286238201166356, + "grad_norm": 2.069108144869967, + "learning_rate": 3.135389669798311e-06, + "loss": 0.7928, + "step": 2733 + }, + { + "epoch": 0.3287440630072747, + "grad_norm": 1.829446429131104, + "learning_rate": 3.134748303797373e-06, + "loss": 0.803, + "step": 2734 + }, + { + "epoch": 0.32886430589791377, + "grad_norm": 2.2039367046233065, + "learning_rate": 3.1341067656560203e-06, + "loss": 0.8206, + "step": 2735 + }, + { + "epoch": 0.3289845487885529, + "grad_norm": 6.6625519770156805, + "learning_rate": 3.133465055471572e-06, + "loss": 0.8767, + "step": 2736 + }, + { + "epoch": 0.329104791679192, + "grad_norm": 2.8301520907238213, + "learning_rate": 3.1328231733413767e-06, + "loss": 0.6651, + "step": 2737 + }, + { + "epoch": 0.32922503456983104, + "grad_norm": 2.5352172844215417, + "learning_rate": 3.1321811193628067e-06, + "loss": 0.9164, + "step": 2738 + }, + { + "epoch": 0.32934527746047015, + "grad_norm": 2.093001616228102, + "learning_rate": 3.131538893633261e-06, + "loss": 0.7099, + "step": 2739 + }, + { + "epoch": 0.32946552035110926, + "grad_norm": 2.428834231766926, + "learning_rate": 3.130896496250165e-06, + "loss": 0.7807, + "step": 2740 + }, + { + "epoch": 0.3295857632417483, + "grad_norm": 2.307081731081845, + "learning_rate": 3.1302539273109693e-06, + "loss": 0.8803, + "step": 2741 + }, + { + "epoch": 0.32970600613238743, + "grad_norm": 1.9727846009450891, + "learning_rate": 3.1296111869131513e-06, + "loss": 0.8117, + "step": 2742 + }, + { + "epoch": 0.32982624902302654, + "grad_norm": 1.8857646923016258, + "learning_rate": 3.1289682751542153e-06, + "loss": 0.8627, + "step": 2743 + }, + { + "epoch": 0.3299464919136656, + "grad_norm": 1.945129998773266, + "learning_rate": 3.1283251921316883e-06, + "loss": 0.7187, + "step": 2744 + }, + { + "epoch": 0.3300667348043047, + "grad_norm": 2.0310997540587463, + "learning_rate": 3.1276819379431277e-06, + "loss": 0.8185, + "step": 2745 + }, + { + "epoch": 0.33018697769494376, + "grad_norm": 1.944383943464212, + "learning_rate": 3.1270385126861134e-06, + "loss": 0.7617, + "step": 2746 + }, + { + "epoch": 0.3303072205855829, + "grad_norm": 1.7379800942820742, + "learning_rate": 3.1263949164582533e-06, + "loss": 0.8291, + "step": 2747 + }, + { + "epoch": 0.330427463476222, + "grad_norm": 2.0568786077661905, + "learning_rate": 3.1257511493571797e-06, + "loss": 0.7903, + "step": 2748 + }, + { + "epoch": 0.33054770636686104, + "grad_norm": 2.076101824062228, + "learning_rate": 3.125107211480552e-06, + "loss": 0.7882, + "step": 2749 + }, + { + "epoch": 0.33066794925750015, + "grad_norm": 1.7419270547425887, + "learning_rate": 3.124463102926054e-06, + "loss": 0.8017, + "step": 2750 + }, + { + "epoch": 0.33078819214813926, + "grad_norm": 0.7669251749028984, + "learning_rate": 3.1238188237913984e-06, + "loss": 0.6162, + "step": 2751 + }, + { + "epoch": 0.3309084350387783, + "grad_norm": 1.9028422429038605, + "learning_rate": 3.1231743741743202e-06, + "loss": 0.7679, + "step": 2752 + }, + { + "epoch": 0.3310286779294174, + "grad_norm": 2.9029606077627315, + "learning_rate": 3.122529754172582e-06, + "loss": 0.8434, + "step": 2753 + }, + { + "epoch": 0.33114892082005654, + "grad_norm": 1.7950343780625801, + "learning_rate": 3.1218849638839736e-06, + "loss": 0.7275, + "step": 2754 + }, + { + "epoch": 0.3312691637106956, + "grad_norm": 2.266222434362805, + "learning_rate": 3.121240003406307e-06, + "loss": 0.7906, + "step": 2755 + }, + { + "epoch": 0.3313894066013347, + "grad_norm": 2.053402575072638, + "learning_rate": 3.120594872837425e-06, + "loss": 0.7267, + "step": 2756 + }, + { + "epoch": 0.3315096494919738, + "grad_norm": 0.8411762401332563, + "learning_rate": 3.1199495722751906e-06, + "loss": 0.6242, + "step": 2757 + }, + { + "epoch": 0.33162989238261287, + "grad_norm": 1.699604771054835, + "learning_rate": 3.1193041018174972e-06, + "loss": 0.841, + "step": 2758 + }, + { + "epoch": 0.331750135273252, + "grad_norm": 2.01219156875353, + "learning_rate": 3.118658461562261e-06, + "loss": 0.9517, + "step": 2759 + }, + { + "epoch": 0.33187037816389103, + "grad_norm": 1.453083981170892, + "learning_rate": 3.118012651607426e-06, + "loss": 0.8521, + "step": 2760 + }, + { + "epoch": 0.33199062105453014, + "grad_norm": 2.0088855532929815, + "learning_rate": 3.1173666720509603e-06, + "loss": 0.8357, + "step": 2761 + }, + { + "epoch": 0.33211086394516925, + "grad_norm": 1.7136767333192044, + "learning_rate": 3.116720522990859e-06, + "loss": 0.6819, + "step": 2762 + }, + { + "epoch": 0.3322311068358083, + "grad_norm": 7.5321553426110786, + "learning_rate": 3.116074204525142e-06, + "loss": 0.6226, + "step": 2763 + }, + { + "epoch": 0.3323513497264474, + "grad_norm": 1.588200425147545, + "learning_rate": 3.1154277167518553e-06, + "loss": 0.8416, + "step": 2764 + }, + { + "epoch": 0.33247159261708653, + "grad_norm": 0.7887234856551665, + "learning_rate": 3.114781059769072e-06, + "loss": 0.595, + "step": 2765 + }, + { + "epoch": 0.3325918355077256, + "grad_norm": 3.0911789817071806, + "learning_rate": 3.1141342336748874e-06, + "loss": 0.681, + "step": 2766 + }, + { + "epoch": 0.3327120783983647, + "grad_norm": 1.6156073327638805, + "learning_rate": 3.1134872385674253e-06, + "loss": 0.8225, + "step": 2767 + }, + { + "epoch": 0.3328323212890038, + "grad_norm": 1.7539876736302413, + "learning_rate": 3.1128400745448353e-06, + "loss": 0.8636, + "step": 2768 + }, + { + "epoch": 0.33295256417964286, + "grad_norm": 2.4062982529595214, + "learning_rate": 3.11219274170529e-06, + "loss": 0.6354, + "step": 2769 + }, + { + "epoch": 0.333072807070282, + "grad_norm": 2.0023397825918434, + "learning_rate": 3.1115452401469903e-06, + "loss": 0.8177, + "step": 2770 + }, + { + "epoch": 0.3331930499609211, + "grad_norm": 2.2290372246448777, + "learning_rate": 3.1108975699681613e-06, + "loss": 0.8669, + "step": 2771 + }, + { + "epoch": 0.33331329285156014, + "grad_norm": 1.8719868290633472, + "learning_rate": 3.1102497312670542e-06, + "loss": 0.7223, + "step": 2772 + }, + { + "epoch": 0.33343353574219925, + "grad_norm": 2.2721741826959505, + "learning_rate": 3.109601724141946e-06, + "loss": 0.8164, + "step": 2773 + }, + { + "epoch": 0.33355377863283836, + "grad_norm": 1.9058692426791586, + "learning_rate": 3.108953548691138e-06, + "loss": 0.684, + "step": 2774 + }, + { + "epoch": 0.3336740215234774, + "grad_norm": 2.814512931193594, + "learning_rate": 3.108305205012959e-06, + "loss": 0.731, + "step": 2775 + }, + { + "epoch": 0.3337942644141165, + "grad_norm": 2.3106222003386345, + "learning_rate": 3.107656693205761e-06, + "loss": 0.8791, + "step": 2776 + }, + { + "epoch": 0.3339145073047556, + "grad_norm": 2.854167260682724, + "learning_rate": 3.107008013367924e-06, + "loss": 0.7096, + "step": 2777 + }, + { + "epoch": 0.3340347501953947, + "grad_norm": 2.4856656159019357, + "learning_rate": 3.1063591655978507e-06, + "loss": 0.8766, + "step": 2778 + }, + { + "epoch": 0.3341549930860338, + "grad_norm": 1.922188834021063, + "learning_rate": 3.105710149993972e-06, + "loss": 0.8051, + "step": 2779 + }, + { + "epoch": 0.33427523597667286, + "grad_norm": 2.4889349206100153, + "learning_rate": 3.1050609666547427e-06, + "loss": 0.859, + "step": 2780 + }, + { + "epoch": 0.33439547886731197, + "grad_norm": 1.8447456257906947, + "learning_rate": 3.104411615678644e-06, + "loss": 0.7751, + "step": 2781 + }, + { + "epoch": 0.3345157217579511, + "grad_norm": 2.2517507946901985, + "learning_rate": 3.1037620971641803e-06, + "loss": 0.7453, + "step": 2782 + }, + { + "epoch": 0.33463596464859013, + "grad_norm": 2.82203051369849, + "learning_rate": 3.1031124112098844e-06, + "loss": 0.6551, + "step": 2783 + }, + { + "epoch": 0.33475620753922924, + "grad_norm": 1.8391736018896023, + "learning_rate": 3.1024625579143127e-06, + "loss": 0.7237, + "step": 2784 + }, + { + "epoch": 0.33487645042986836, + "grad_norm": 2.420217623738089, + "learning_rate": 3.101812537376048e-06, + "loss": 0.7355, + "step": 2785 + }, + { + "epoch": 0.3349966933205074, + "grad_norm": 2.0502089510239894, + "learning_rate": 3.1011623496936973e-06, + "loss": 0.8437, + "step": 2786 + }, + { + "epoch": 0.3351169362111465, + "grad_norm": 1.9689002206802535, + "learning_rate": 3.100511994965893e-06, + "loss": 0.7045, + "step": 2787 + }, + { + "epoch": 0.33523717910178563, + "grad_norm": 1.674400062729442, + "learning_rate": 3.0998614732912947e-06, + "loss": 0.8463, + "step": 2788 + }, + { + "epoch": 0.3353574219924247, + "grad_norm": 2.035293193756615, + "learning_rate": 3.0992107847685855e-06, + "loss": 0.6811, + "step": 2789 + }, + { + "epoch": 0.3354776648830638, + "grad_norm": 3.296624919934316, + "learning_rate": 3.0985599294964736e-06, + "loss": 0.7988, + "step": 2790 + }, + { + "epoch": 0.33559790777370285, + "grad_norm": 1.9482311124176124, + "learning_rate": 3.097908907573695e-06, + "loss": 0.7104, + "step": 2791 + }, + { + "epoch": 0.33571815066434196, + "grad_norm": 2.3071130435135894, + "learning_rate": 3.0972577190990067e-06, + "loss": 0.9039, + "step": 2792 + }, + { + "epoch": 0.3358383935549811, + "grad_norm": 1.844981800392762, + "learning_rate": 3.096606364171196e-06, + "loss": 0.8147, + "step": 2793 + }, + { + "epoch": 0.33595863644562013, + "grad_norm": 1.8904396785988757, + "learning_rate": 3.0959548428890703e-06, + "loss": 0.8541, + "step": 2794 + }, + { + "epoch": 0.33607887933625924, + "grad_norm": 1.746460263830889, + "learning_rate": 3.095303155351468e-06, + "loss": 0.8414, + "step": 2795 + }, + { + "epoch": 0.33619912222689835, + "grad_norm": 2.711215222119383, + "learning_rate": 3.0946513016572464e-06, + "loss": 0.7946, + "step": 2796 + }, + { + "epoch": 0.3363193651175374, + "grad_norm": 1.9730348129372004, + "learning_rate": 3.0939992819052938e-06, + "loss": 0.7816, + "step": 2797 + }, + { + "epoch": 0.3364396080081765, + "grad_norm": 2.3754104460366197, + "learning_rate": 3.0933470961945193e-06, + "loss": 0.8233, + "step": 2798 + }, + { + "epoch": 0.3365598508988156, + "grad_norm": 2.1323815214241977, + "learning_rate": 3.0926947446238597e-06, + "loss": 0.693, + "step": 2799 + }, + { + "epoch": 0.3366800937894547, + "grad_norm": 2.2601962590030595, + "learning_rate": 3.092042227292276e-06, + "loss": 0.8366, + "step": 2800 + }, + { + "epoch": 0.3368003366800938, + "grad_norm": 1.61490126730217, + "learning_rate": 3.0913895442987557e-06, + "loss": 0.887, + "step": 2801 + }, + { + "epoch": 0.3369205795707329, + "grad_norm": 2.121049574924143, + "learning_rate": 3.090736695742308e-06, + "loss": 0.8642, + "step": 2802 + }, + { + "epoch": 0.33704082246137196, + "grad_norm": 2.5402432087981204, + "learning_rate": 3.0900836817219713e-06, + "loss": 0.5237, + "step": 2803 + }, + { + "epoch": 0.33716106535201107, + "grad_norm": 1.6713623259065757, + "learning_rate": 3.089430502336807e-06, + "loss": 0.8396, + "step": 2804 + }, + { + "epoch": 0.3372813082426502, + "grad_norm": 2.328285774977672, + "learning_rate": 3.088777157685902e-06, + "loss": 0.9025, + "step": 2805 + }, + { + "epoch": 0.33740155113328923, + "grad_norm": 2.472366774067945, + "learning_rate": 3.088123647868367e-06, + "loss": 0.8627, + "step": 2806 + }, + { + "epoch": 0.33752179402392835, + "grad_norm": 5.125534774047553, + "learning_rate": 3.0874699729833405e-06, + "loss": 0.8215, + "step": 2807 + }, + { + "epoch": 0.3376420369145674, + "grad_norm": 1.770416173230853, + "learning_rate": 3.086816133129983e-06, + "loss": 0.8006, + "step": 2808 + }, + { + "epoch": 0.3377622798052065, + "grad_norm": 2.087403543036936, + "learning_rate": 3.0861621284074826e-06, + "loss": 0.7714, + "step": 2809 + }, + { + "epoch": 0.3378825226958456, + "grad_norm": 1.5681174959905553, + "learning_rate": 3.085507958915051e-06, + "loss": 0.7352, + "step": 2810 + }, + { + "epoch": 0.3380027655864847, + "grad_norm": 4.511960347630381, + "learning_rate": 3.084853624751925e-06, + "loss": 0.7217, + "step": 2811 + }, + { + "epoch": 0.3381230084771238, + "grad_norm": 2.073684652355452, + "learning_rate": 3.0841991260173668e-06, + "loss": 0.8656, + "step": 2812 + }, + { + "epoch": 0.3382432513677629, + "grad_norm": 2.020624746457717, + "learning_rate": 3.0835444628106634e-06, + "loss": 0.8126, + "step": 2813 + }, + { + "epoch": 0.33836349425840195, + "grad_norm": 2.4194694139155897, + "learning_rate": 3.082889635231126e-06, + "loss": 0.8334, + "step": 2814 + }, + { + "epoch": 0.33848373714904106, + "grad_norm": 2.546196501988702, + "learning_rate": 3.0822346433780925e-06, + "loss": 0.7732, + "step": 2815 + }, + { + "epoch": 0.3386039800396802, + "grad_norm": 2.0868894997750216, + "learning_rate": 3.0815794873509237e-06, + "loss": 0.8829, + "step": 2816 + }, + { + "epoch": 0.33872422293031923, + "grad_norm": 3.270288377464187, + "learning_rate": 3.0809241672490066e-06, + "loss": 0.7333, + "step": 2817 + }, + { + "epoch": 0.33884446582095834, + "grad_norm": 1.922421385468567, + "learning_rate": 3.080268683171753e-06, + "loss": 0.8542, + "step": 2818 + }, + { + "epoch": 0.33896470871159745, + "grad_norm": 2.1113141224385346, + "learning_rate": 3.0796130352185985e-06, + "loss": 0.8988, + "step": 2819 + }, + { + "epoch": 0.3390849516022365, + "grad_norm": 2.035462808492722, + "learning_rate": 3.0789572234890057e-06, + "loss": 0.6707, + "step": 2820 + }, + { + "epoch": 0.3392051944928756, + "grad_norm": 1.6845713032152885, + "learning_rate": 3.0783012480824596e-06, + "loss": 0.7741, + "step": 2821 + }, + { + "epoch": 0.33932543738351467, + "grad_norm": 3.5222699246015954, + "learning_rate": 3.077645109098471e-06, + "loss": 0.7484, + "step": 2822 + }, + { + "epoch": 0.3394456802741538, + "grad_norm": 1.7191156865282808, + "learning_rate": 3.076988806636577e-06, + "loss": 0.7252, + "step": 2823 + }, + { + "epoch": 0.3395659231647929, + "grad_norm": 1.9734554673537525, + "learning_rate": 3.0763323407963377e-06, + "loss": 0.8895, + "step": 2824 + }, + { + "epoch": 0.33968616605543195, + "grad_norm": 1.9205053147309725, + "learning_rate": 3.075675711677337e-06, + "loss": 0.8035, + "step": 2825 + }, + { + "epoch": 0.33980640894607106, + "grad_norm": 1.9785618979041042, + "learning_rate": 3.0750189193791865e-06, + "loss": 0.7835, + "step": 2826 + }, + { + "epoch": 0.33992665183671017, + "grad_norm": 1.9008464207826528, + "learning_rate": 3.0743619640015203e-06, + "loss": 0.7109, + "step": 2827 + }, + { + "epoch": 0.3400468947273492, + "grad_norm": 1.9188751751455635, + "learning_rate": 3.073704845643999e-06, + "loss": 0.929, + "step": 2828 + }, + { + "epoch": 0.34016713761798834, + "grad_norm": 3.034414154020126, + "learning_rate": 3.0730475644063063e-06, + "loss": 0.7934, + "step": 2829 + }, + { + "epoch": 0.34028738050862745, + "grad_norm": 1.9745772573272002, + "learning_rate": 3.072390120388151e-06, + "loss": 0.6552, + "step": 2830 + }, + { + "epoch": 0.3404076233992665, + "grad_norm": 2.245809679726604, + "learning_rate": 3.071732513689267e-06, + "loss": 0.7296, + "step": 2831 + }, + { + "epoch": 0.3405278662899056, + "grad_norm": 2.3214951914356066, + "learning_rate": 3.0710747444094134e-06, + "loss": 0.6796, + "step": 2832 + }, + { + "epoch": 0.3406481091805447, + "grad_norm": 1.9465428640024667, + "learning_rate": 3.070416812648372e-06, + "loss": 0.648, + "step": 2833 + }, + { + "epoch": 0.3407683520711838, + "grad_norm": 2.43636628219378, + "learning_rate": 3.069758718505951e-06, + "loss": 0.6574, + "step": 2834 + }, + { + "epoch": 0.3408885949618229, + "grad_norm": 1.7127219174685506, + "learning_rate": 3.0691004620819836e-06, + "loss": 0.8011, + "step": 2835 + }, + { + "epoch": 0.341008837852462, + "grad_norm": 0.80595372397048, + "learning_rate": 3.0684420434763254e-06, + "loss": 0.6033, + "step": 2836 + }, + { + "epoch": 0.34112908074310105, + "grad_norm": 1.9428624423539789, + "learning_rate": 3.06778346278886e-06, + "loss": 0.7707, + "step": 2837 + }, + { + "epoch": 0.34124932363374016, + "grad_norm": 1.7592245576272905, + "learning_rate": 3.0671247201194906e-06, + "loss": 0.7957, + "step": 2838 + }, + { + "epoch": 0.3413695665243792, + "grad_norm": 1.7895421880122808, + "learning_rate": 3.066465815568151e-06, + "loss": 0.7576, + "step": 2839 + }, + { + "epoch": 0.34148980941501833, + "grad_norm": 1.7359794549825, + "learning_rate": 3.0658067492347947e-06, + "loss": 0.6915, + "step": 2840 + }, + { + "epoch": 0.34161005230565744, + "grad_norm": 2.8818352756480095, + "learning_rate": 3.065147521219402e-06, + "loss": 0.6828, + "step": 2841 + }, + { + "epoch": 0.3417302951962965, + "grad_norm": 1.5259881992108784, + "learning_rate": 3.064488131621977e-06, + "loss": 0.743, + "step": 2842 + }, + { + "epoch": 0.3418505380869356, + "grad_norm": 1.84872953763397, + "learning_rate": 3.063828580542549e-06, + "loss": 0.7463, + "step": 2843 + }, + { + "epoch": 0.3419707809775747, + "grad_norm": 1.9293836593804323, + "learning_rate": 3.0631688680811706e-06, + "loss": 0.7399, + "step": 2844 + }, + { + "epoch": 0.3420910238682138, + "grad_norm": 2.142657352397988, + "learning_rate": 3.062508994337921e-06, + "loss": 0.7604, + "step": 2845 + }, + { + "epoch": 0.3422112667588529, + "grad_norm": 2.373455951109156, + "learning_rate": 3.0618489594129013e-06, + "loss": 0.8042, + "step": 2846 + }, + { + "epoch": 0.342331509649492, + "grad_norm": 2.1044027995444963, + "learning_rate": 3.061188763406239e-06, + "loss": 0.7182, + "step": 2847 + }, + { + "epoch": 0.34245175254013105, + "grad_norm": 2.2309054667006802, + "learning_rate": 3.060528406418085e-06, + "loss": 0.828, + "step": 2848 + }, + { + "epoch": 0.34257199543077016, + "grad_norm": 1.9395618542698494, + "learning_rate": 3.0598678885486145e-06, + "loss": 0.6208, + "step": 2849 + }, + { + "epoch": 0.34269223832140927, + "grad_norm": 1.6854303878898367, + "learning_rate": 3.0592072098980282e-06, + "loss": 0.7483, + "step": 2850 + }, + { + "epoch": 0.3428124812120483, + "grad_norm": 2.4084235938910505, + "learning_rate": 3.0585463705665514e-06, + "loss": 0.7379, + "step": 2851 + }, + { + "epoch": 0.34293272410268744, + "grad_norm": 2.292382966426946, + "learning_rate": 3.0578853706544304e-06, + "loss": 0.7076, + "step": 2852 + }, + { + "epoch": 0.34305296699332655, + "grad_norm": 2.332400907678743, + "learning_rate": 3.0572242102619404e-06, + "loss": 0.6621, + "step": 2853 + }, + { + "epoch": 0.3431732098839656, + "grad_norm": 1.8769117146348924, + "learning_rate": 3.0565628894893784e-06, + "loss": 0.8094, + "step": 2854 + }, + { + "epoch": 0.3432934527746047, + "grad_norm": 1.804774955182773, + "learning_rate": 3.0559014084370655e-06, + "loss": 0.752, + "step": 2855 + }, + { + "epoch": 0.34341369566524377, + "grad_norm": 1.8239376006662804, + "learning_rate": 3.055239767205349e-06, + "loss": 0.794, + "step": 2856 + }, + { + "epoch": 0.3435339385558829, + "grad_norm": 2.7231553439834864, + "learning_rate": 3.054577965894599e-06, + "loss": 0.7812, + "step": 2857 + }, + { + "epoch": 0.343654181446522, + "grad_norm": 1.5495864531829329, + "learning_rate": 3.0539160046052094e-06, + "loss": 0.7054, + "step": 2858 + }, + { + "epoch": 0.34377442433716104, + "grad_norm": 2.1291716523930155, + "learning_rate": 3.0532538834376003e-06, + "loss": 0.7047, + "step": 2859 + }, + { + "epoch": 0.34389466722780015, + "grad_norm": 1.9271920319564677, + "learning_rate": 3.0525916024922143e-06, + "loss": 0.7871, + "step": 2860 + }, + { + "epoch": 0.34401491011843927, + "grad_norm": 2.962291397927103, + "learning_rate": 3.0519291618695193e-06, + "loss": 0.8497, + "step": 2861 + }, + { + "epoch": 0.3441351530090783, + "grad_norm": 1.6506231200652737, + "learning_rate": 3.0512665616700065e-06, + "loss": 0.7598, + "step": 2862 + }, + { + "epoch": 0.34425539589971743, + "grad_norm": 1.9772823666344583, + "learning_rate": 3.0506038019941933e-06, + "loss": 0.8972, + "step": 2863 + }, + { + "epoch": 0.34437563879035654, + "grad_norm": 3.149031590147916, + "learning_rate": 3.049940882942617e-06, + "loss": 0.6912, + "step": 2864 + }, + { + "epoch": 0.3444958816809956, + "grad_norm": 1.883941691668673, + "learning_rate": 3.0492778046158448e-06, + "loss": 0.807, + "step": 2865 + }, + { + "epoch": 0.3446161245716347, + "grad_norm": 2.8008512224849795, + "learning_rate": 3.0486145671144633e-06, + "loss": 0.7774, + "step": 2866 + }, + { + "epoch": 0.3447363674622738, + "grad_norm": 2.660276704321317, + "learning_rate": 3.047951170539086e-06, + "loss": 0.7656, + "step": 2867 + }, + { + "epoch": 0.3448566103529129, + "grad_norm": 1.8370158173903361, + "learning_rate": 3.047287614990349e-06, + "loss": 0.8446, + "step": 2868 + }, + { + "epoch": 0.344976853243552, + "grad_norm": 3.307571343238539, + "learning_rate": 3.046623900568914e-06, + "loss": 0.63, + "step": 2869 + }, + { + "epoch": 0.34509709613419104, + "grad_norm": 3.134663860059634, + "learning_rate": 3.045960027375465e-06, + "loss": 0.7042, + "step": 2870 + }, + { + "epoch": 0.34521733902483015, + "grad_norm": 2.669549387221868, + "learning_rate": 3.045295995510711e-06, + "loss": 0.8249, + "step": 2871 + }, + { + "epoch": 0.34533758191546926, + "grad_norm": 1.8186238851442222, + "learning_rate": 3.0446318050753865e-06, + "loss": 0.7369, + "step": 2872 + }, + { + "epoch": 0.3454578248061083, + "grad_norm": 1.9921730788961802, + "learning_rate": 3.0439674561702474e-06, + "loss": 0.7787, + "step": 2873 + }, + { + "epoch": 0.3455780676967474, + "grad_norm": 3.060369879614036, + "learning_rate": 3.043302948896076e-06, + "loss": 0.8865, + "step": 2874 + }, + { + "epoch": 0.34569831058738654, + "grad_norm": 3.4552815465307454, + "learning_rate": 3.0426382833536756e-06, + "loss": 0.6087, + "step": 2875 + }, + { + "epoch": 0.3458185534780256, + "grad_norm": 2.9854066673336734, + "learning_rate": 3.041973459643877e-06, + "loss": 0.7793, + "step": 2876 + }, + { + "epoch": 0.3459387963686647, + "grad_norm": 2.0155414889283, + "learning_rate": 3.0413084778675334e-06, + "loss": 0.6774, + "step": 2877 + }, + { + "epoch": 0.3460590392593038, + "grad_norm": 2.194606898393551, + "learning_rate": 3.0406433381255214e-06, + "loss": 0.8478, + "step": 2878 + }, + { + "epoch": 0.34617928214994287, + "grad_norm": 3.8306413126539187, + "learning_rate": 3.0399780405187425e-06, + "loss": 0.8355, + "step": 2879 + }, + { + "epoch": 0.346299525040582, + "grad_norm": 2.31706910194216, + "learning_rate": 3.0393125851481216e-06, + "loss": 0.7901, + "step": 2880 + }, + { + "epoch": 0.3464197679312211, + "grad_norm": 2.1079571515239897, + "learning_rate": 3.038646972114608e-06, + "loss": 0.8673, + "step": 2881 + }, + { + "epoch": 0.34654001082186014, + "grad_norm": 1.8956616991202826, + "learning_rate": 3.037981201519174e-06, + "loss": 0.6809, + "step": 2882 + }, + { + "epoch": 0.34666025371249926, + "grad_norm": 3.0045868458504925, + "learning_rate": 3.0373152734628175e-06, + "loss": 0.7237, + "step": 2883 + }, + { + "epoch": 0.34678049660313837, + "grad_norm": 2.013097198209082, + "learning_rate": 3.0366491880465584e-06, + "loss": 0.7666, + "step": 2884 + }, + { + "epoch": 0.3469007394937774, + "grad_norm": 1.8551869389198015, + "learning_rate": 3.035982945371443e-06, + "loss": 0.825, + "step": 2885 + }, + { + "epoch": 0.34702098238441653, + "grad_norm": 2.0192624176954648, + "learning_rate": 3.035316545538537e-06, + "loss": 0.8632, + "step": 2886 + }, + { + "epoch": 0.3471412252750556, + "grad_norm": 2.744513587449615, + "learning_rate": 3.034649988648935e-06, + "loss": 0.8016, + "step": 2887 + }, + { + "epoch": 0.3472614681656947, + "grad_norm": 1.67301502754134, + "learning_rate": 3.033983274803752e-06, + "loss": 0.8104, + "step": 2888 + }, + { + "epoch": 0.3473817110563338, + "grad_norm": 2.182698238497891, + "learning_rate": 3.0333164041041283e-06, + "loss": 0.7313, + "step": 2889 + }, + { + "epoch": 0.34750195394697286, + "grad_norm": 1.9884064940331758, + "learning_rate": 3.032649376651228e-06, + "loss": 0.7213, + "step": 2890 + }, + { + "epoch": 0.347622196837612, + "grad_norm": 1.5938364641259282, + "learning_rate": 3.031982192546238e-06, + "loss": 0.7619, + "step": 2891 + }, + { + "epoch": 0.3477424397282511, + "grad_norm": 2.1906578159459182, + "learning_rate": 3.0313148518903696e-06, + "loss": 0.9541, + "step": 2892 + }, + { + "epoch": 0.34786268261889014, + "grad_norm": 2.1116634811863086, + "learning_rate": 3.030647354784859e-06, + "loss": 0.8227, + "step": 2893 + }, + { + "epoch": 0.34798292550952925, + "grad_norm": 2.110753379037428, + "learning_rate": 3.029979701330964e-06, + "loss": 0.7754, + "step": 2894 + }, + { + "epoch": 0.34810316840016836, + "grad_norm": 2.212789992816112, + "learning_rate": 3.029311891629966e-06, + "loss": 0.8066, + "step": 2895 + }, + { + "epoch": 0.3482234112908074, + "grad_norm": 2.0684848995860894, + "learning_rate": 3.0286439257831744e-06, + "loss": 0.7555, + "step": 2896 + }, + { + "epoch": 0.3483436541814465, + "grad_norm": 2.0409331528642602, + "learning_rate": 3.0279758038919156e-06, + "loss": 0.7147, + "step": 2897 + }, + { + "epoch": 0.34846389707208564, + "grad_norm": 2.106301335917534, + "learning_rate": 3.0273075260575455e-06, + "loss": 0.791, + "step": 2898 + }, + { + "epoch": 0.3485841399627247, + "grad_norm": 1.964525261428918, + "learning_rate": 3.0266390923814396e-06, + "loss": 0.8127, + "step": 2899 + }, + { + "epoch": 0.3487043828533638, + "grad_norm": 1.956813168129083, + "learning_rate": 3.0259705029650008e-06, + "loss": 0.8249, + "step": 2900 + }, + { + "epoch": 0.34882462574400286, + "grad_norm": 1.8158606581479677, + "learning_rate": 3.025301757909652e-06, + "loss": 0.7376, + "step": 2901 + }, + { + "epoch": 0.34894486863464197, + "grad_norm": 1.6804911975079588, + "learning_rate": 3.024632857316842e-06, + "loss": 0.8115, + "step": 2902 + }, + { + "epoch": 0.3490651115252811, + "grad_norm": 1.989540585629523, + "learning_rate": 3.0239638012880412e-06, + "loss": 0.7822, + "step": 2903 + }, + { + "epoch": 0.34918535441592014, + "grad_norm": 2.3716597910513526, + "learning_rate": 3.0232945899247466e-06, + "loss": 0.8202, + "step": 2904 + }, + { + "epoch": 0.34930559730655925, + "grad_norm": 2.4115508510124855, + "learning_rate": 3.022625223328476e-06, + "loss": 0.7795, + "step": 2905 + }, + { + "epoch": 0.34942584019719836, + "grad_norm": 1.4790100319182322, + "learning_rate": 3.0219557016007723e-06, + "loss": 0.6969, + "step": 2906 + }, + { + "epoch": 0.3495460830878374, + "grad_norm": 1.8996123771792799, + "learning_rate": 3.021286024843202e-06, + "loss": 0.7085, + "step": 2907 + }, + { + "epoch": 0.3496663259784765, + "grad_norm": 1.0874025630051418, + "learning_rate": 3.0206161931573526e-06, + "loss": 0.6491, + "step": 2908 + }, + { + "epoch": 0.34978656886911563, + "grad_norm": 1.5958889837828854, + "learning_rate": 3.0199462066448388e-06, + "loss": 0.9333, + "step": 2909 + }, + { + "epoch": 0.3499068117597547, + "grad_norm": 1.9779156663001598, + "learning_rate": 3.019276065407296e-06, + "loss": 0.6926, + "step": 2910 + }, + { + "epoch": 0.3500270546503938, + "grad_norm": 1.792519706394295, + "learning_rate": 3.018605769546385e-06, + "loss": 0.8093, + "step": 2911 + }, + { + "epoch": 0.3501472975410329, + "grad_norm": 1.7909424079397775, + "learning_rate": 3.017935319163788e-06, + "loss": 0.803, + "step": 2912 + }, + { + "epoch": 0.35026754043167196, + "grad_norm": 2.139339980190686, + "learning_rate": 3.017264714361213e-06, + "loss": 0.7112, + "step": 2913 + }, + { + "epoch": 0.3503877833223111, + "grad_norm": 1.9926741660087486, + "learning_rate": 3.016593955240389e-06, + "loss": 0.8279, + "step": 2914 + }, + { + "epoch": 0.3505080262129502, + "grad_norm": 0.8296081289484452, + "learning_rate": 3.015923041903071e-06, + "loss": 0.6381, + "step": 2915 + }, + { + "epoch": 0.35062826910358924, + "grad_norm": 2.111337158861339, + "learning_rate": 3.0152519744510347e-06, + "loss": 0.8413, + "step": 2916 + }, + { + "epoch": 0.35074851199422835, + "grad_norm": 2.0693753406613093, + "learning_rate": 3.014580752986081e-06, + "loss": 0.8414, + "step": 2917 + }, + { + "epoch": 0.3508687548848674, + "grad_norm": 2.1068694881795995, + "learning_rate": 3.0139093776100345e-06, + "loss": 0.7915, + "step": 2918 + }, + { + "epoch": 0.3509889977755065, + "grad_norm": 1.6467334201476722, + "learning_rate": 3.013237848424741e-06, + "loss": 0.7578, + "step": 2919 + }, + { + "epoch": 0.35110924066614563, + "grad_norm": 2.0725724463230635, + "learning_rate": 3.012566165532072e-06, + "loss": 0.7551, + "step": 2920 + }, + { + "epoch": 0.3512294835567847, + "grad_norm": 2.104264461649378, + "learning_rate": 3.0118943290339207e-06, + "loss": 0.7698, + "step": 2921 + }, + { + "epoch": 0.3513497264474238, + "grad_norm": 1.9904703556278602, + "learning_rate": 3.011222339032204e-06, + "loss": 0.6928, + "step": 2922 + }, + { + "epoch": 0.3514699693380629, + "grad_norm": 1.8781117367395133, + "learning_rate": 3.0105501956288626e-06, + "loss": 0.7029, + "step": 2923 + }, + { + "epoch": 0.35159021222870196, + "grad_norm": 2.202831492250815, + "learning_rate": 3.0098778989258602e-06, + "loss": 0.7378, + "step": 2924 + }, + { + "epoch": 0.35171045511934107, + "grad_norm": 1.8808307269203246, + "learning_rate": 3.009205449025183e-06, + "loss": 0.8891, + "step": 2925 + }, + { + "epoch": 0.3518306980099802, + "grad_norm": 1.9304873395740298, + "learning_rate": 3.008532846028842e-06, + "loss": 0.6316, + "step": 2926 + }, + { + "epoch": 0.35195094090061924, + "grad_norm": 2.325686002862625, + "learning_rate": 3.0078600900388694e-06, + "loss": 0.7197, + "step": 2927 + }, + { + "epoch": 0.35207118379125835, + "grad_norm": 1.8030646487603412, + "learning_rate": 3.007187181157323e-06, + "loss": 0.7432, + "step": 2928 + }, + { + "epoch": 0.35219142668189746, + "grad_norm": 2.185489384472021, + "learning_rate": 3.006514119486282e-06, + "loss": 0.6876, + "step": 2929 + }, + { + "epoch": 0.3523116695725365, + "grad_norm": 1.8191362763873633, + "learning_rate": 3.005840905127849e-06, + "loss": 0.7025, + "step": 2930 + }, + { + "epoch": 0.3524319124631756, + "grad_norm": 2.4485049142910897, + "learning_rate": 3.0051675381841516e-06, + "loss": 0.8746, + "step": 2931 + }, + { + "epoch": 0.3525521553538147, + "grad_norm": 1.5691383489051438, + "learning_rate": 3.0044940187573363e-06, + "loss": 0.7766, + "step": 2932 + }, + { + "epoch": 0.3526723982444538, + "grad_norm": 2.0335355371503274, + "learning_rate": 3.003820346949578e-06, + "loss": 0.6595, + "step": 2933 + }, + { + "epoch": 0.3527926411350929, + "grad_norm": 1.8586411510578802, + "learning_rate": 3.003146522863071e-06, + "loss": 0.7957, + "step": 2934 + }, + { + "epoch": 0.35291288402573195, + "grad_norm": 2.399192268824165, + "learning_rate": 3.0024725466000345e-06, + "loss": 0.8711, + "step": 2935 + }, + { + "epoch": 0.35303312691637107, + "grad_norm": 2.854579218171635, + "learning_rate": 3.0017984182627087e-06, + "loss": 0.795, + "step": 2936 + }, + { + "epoch": 0.3531533698070102, + "grad_norm": 2.0679305283113054, + "learning_rate": 3.00112413795336e-06, + "loss": 0.8331, + "step": 2937 + }, + { + "epoch": 0.35327361269764923, + "grad_norm": 1.9815320653522837, + "learning_rate": 3.000449705774275e-06, + "loss": 0.8052, + "step": 2938 + }, + { + "epoch": 0.35339385558828834, + "grad_norm": 2.1128816068079175, + "learning_rate": 2.9997751218277654e-06, + "loss": 0.7167, + "step": 2939 + }, + { + "epoch": 0.35351409847892745, + "grad_norm": 1.8745449149671956, + "learning_rate": 2.999100386216166e-06, + "loss": 0.783, + "step": 2940 + }, + { + "epoch": 0.3536343413695665, + "grad_norm": 8.024615564377816, + "learning_rate": 2.998425499041831e-06, + "loss": 0.7517, + "step": 2941 + }, + { + "epoch": 0.3537545842602056, + "grad_norm": 1.3278298442710625, + "learning_rate": 2.997750460407142e-06, + "loss": 0.6462, + "step": 2942 + }, + { + "epoch": 0.35387482715084473, + "grad_norm": 2.2592770799282587, + "learning_rate": 2.997075270414501e-06, + "loss": 0.7086, + "step": 2943 + }, + { + "epoch": 0.3539950700414838, + "grad_norm": 0.7090272916725112, + "learning_rate": 2.9963999291663347e-06, + "loss": 0.5779, + "step": 2944 + }, + { + "epoch": 0.3541153129321229, + "grad_norm": 2.3457990717263297, + "learning_rate": 2.9957244367650915e-06, + "loss": 0.7431, + "step": 2945 + }, + { + "epoch": 0.354235555822762, + "grad_norm": 2.058931558243144, + "learning_rate": 2.9950487933132425e-06, + "loss": 0.8471, + "step": 2946 + }, + { + "epoch": 0.35435579871340106, + "grad_norm": 2.7419096107637526, + "learning_rate": 2.994372998913283e-06, + "loss": 0.7198, + "step": 2947 + }, + { + "epoch": 0.35447604160404017, + "grad_norm": 2.4489501611164077, + "learning_rate": 2.99369705366773e-06, + "loss": 0.6283, + "step": 2948 + }, + { + "epoch": 0.3545962844946792, + "grad_norm": 2.567020708494689, + "learning_rate": 2.9930209576791244e-06, + "loss": 0.8291, + "step": 2949 + }, + { + "epoch": 0.35471652738531834, + "grad_norm": 2.3836122265164956, + "learning_rate": 2.9923447110500285e-06, + "loss": 0.6378, + "step": 2950 + }, + { + "epoch": 0.35483677027595745, + "grad_norm": 1.5465747272087502, + "learning_rate": 2.9916683138830295e-06, + "loss": 0.7593, + "step": 2951 + }, + { + "epoch": 0.3549570131665965, + "grad_norm": 2.036266529257363, + "learning_rate": 2.9909917662807353e-06, + "loss": 0.8178, + "step": 2952 + }, + { + "epoch": 0.3550772560572356, + "grad_norm": 2.3436890657769083, + "learning_rate": 2.9903150683457783e-06, + "loss": 0.7084, + "step": 2953 + }, + { + "epoch": 0.3551974989478747, + "grad_norm": 2.2917668638577458, + "learning_rate": 2.9896382201808126e-06, + "loss": 0.6631, + "step": 2954 + }, + { + "epoch": 0.3553177418385138, + "grad_norm": 2.1865335781991844, + "learning_rate": 2.988961221888516e-06, + "loss": 0.816, + "step": 2955 + }, + { + "epoch": 0.3554379847291529, + "grad_norm": 3.3372401389921422, + "learning_rate": 2.988284073571589e-06, + "loss": 0.7962, + "step": 2956 + }, + { + "epoch": 0.355558227619792, + "grad_norm": 2.3747002261855297, + "learning_rate": 2.9876067753327528e-06, + "loss": 0.7363, + "step": 2957 + }, + { + "epoch": 0.35567847051043106, + "grad_norm": 2.999823882973369, + "learning_rate": 2.986929327274754e-06, + "loss": 0.8101, + "step": 2958 + }, + { + "epoch": 0.35579871340107017, + "grad_norm": 1.842855915595118, + "learning_rate": 2.9862517295003617e-06, + "loss": 0.7895, + "step": 2959 + }, + { + "epoch": 0.3559189562917093, + "grad_norm": 1.5560676927527, + "learning_rate": 2.9855739821123654e-06, + "loss": 0.727, + "step": 2960 + }, + { + "epoch": 0.35603919918234833, + "grad_norm": 1.7403271240452882, + "learning_rate": 2.98489608521358e-06, + "loss": 0.821, + "step": 2961 + }, + { + "epoch": 0.35615944207298744, + "grad_norm": 2.6320292499157585, + "learning_rate": 2.9842180389068425e-06, + "loss": 0.8016, + "step": 2962 + }, + { + "epoch": 0.35627968496362655, + "grad_norm": 0.7709289482888705, + "learning_rate": 2.98353984329501e-06, + "loss": 0.5931, + "step": 2963 + }, + { + "epoch": 0.3563999278542656, + "grad_norm": 1.7731360628833757, + "learning_rate": 2.982861498480965e-06, + "loss": 0.7109, + "step": 2964 + }, + { + "epoch": 0.3565201707449047, + "grad_norm": 1.654072247492142, + "learning_rate": 2.9821830045676122e-06, + "loss": 0.8311, + "step": 2965 + }, + { + "epoch": 0.3566404136355438, + "grad_norm": 2.227484903509186, + "learning_rate": 2.9815043616578793e-06, + "loss": 0.7405, + "step": 2966 + }, + { + "epoch": 0.3567606565261829, + "grad_norm": 2.703555034322105, + "learning_rate": 2.9808255698547145e-06, + "loss": 0.7757, + "step": 2967 + }, + { + "epoch": 0.356880899416822, + "grad_norm": 2.170983989366615, + "learning_rate": 2.9801466292610913e-06, + "loss": 0.8045, + "step": 2968 + }, + { + "epoch": 0.35700114230746105, + "grad_norm": 3.4036036019111964, + "learning_rate": 2.979467539980003e-06, + "loss": 0.8152, + "step": 2969 + }, + { + "epoch": 0.35712138519810016, + "grad_norm": 2.1299477365735218, + "learning_rate": 2.978788302114468e-06, + "loss": 0.7751, + "step": 2970 + }, + { + "epoch": 0.35724162808873927, + "grad_norm": 2.446993996173585, + "learning_rate": 2.9781089157675255e-06, + "loss": 0.8399, + "step": 2971 + }, + { + "epoch": 0.3573618709793783, + "grad_norm": 1.5060831539455286, + "learning_rate": 2.977429381042238e-06, + "loss": 0.8891, + "step": 2972 + }, + { + "epoch": 0.35748211387001744, + "grad_norm": 2.3031497286293288, + "learning_rate": 2.9767496980416913e-06, + "loss": 0.8897, + "step": 2973 + }, + { + "epoch": 0.35760235676065655, + "grad_norm": 2.2470985354458604, + "learning_rate": 2.9760698668689914e-06, + "loss": 0.8242, + "step": 2974 + }, + { + "epoch": 0.3577225996512956, + "grad_norm": 1.83830310113049, + "learning_rate": 2.975389887627269e-06, + "loss": 0.7169, + "step": 2975 + }, + { + "epoch": 0.3578428425419347, + "grad_norm": 2.530730394320057, + "learning_rate": 2.9747097604196764e-06, + "loss": 0.9004, + "step": 2976 + }, + { + "epoch": 0.3579630854325738, + "grad_norm": 0.6847610752022021, + "learning_rate": 2.9740294853493875e-06, + "loss": 0.5674, + "step": 2977 + }, + { + "epoch": 0.3580833283232129, + "grad_norm": 2.061555819706591, + "learning_rate": 2.9733490625196008e-06, + "loss": 0.6784, + "step": 2978 + }, + { + "epoch": 0.358203571213852, + "grad_norm": 2.6049129688273265, + "learning_rate": 2.9726684920335353e-06, + "loss": 0.763, + "step": 2979 + }, + { + "epoch": 0.35832381410449105, + "grad_norm": 2.9440996420003125, + "learning_rate": 2.971987773994432e-06, + "loss": 0.8244, + "step": 2980 + }, + { + "epoch": 0.35844405699513016, + "grad_norm": 2.789914553984831, + "learning_rate": 2.9713069085055566e-06, + "loss": 0.8303, + "step": 2981 + }, + { + "epoch": 0.35856429988576927, + "grad_norm": 1.6750722954933601, + "learning_rate": 2.9706258956701958e-06, + "loss": 0.7903, + "step": 2982 + }, + { + "epoch": 0.3586845427764083, + "grad_norm": 2.5876024308869456, + "learning_rate": 2.9699447355916575e-06, + "loss": 0.7871, + "step": 2983 + }, + { + "epoch": 0.35880478566704743, + "grad_norm": 1.981167701913255, + "learning_rate": 2.969263428373275e-06, + "loss": 0.7398, + "step": 2984 + }, + { + "epoch": 0.35892502855768654, + "grad_norm": 2.5013696733219475, + "learning_rate": 2.9685819741184007e-06, + "loss": 0.8009, + "step": 2985 + }, + { + "epoch": 0.3590452714483256, + "grad_norm": 2.479749103590977, + "learning_rate": 2.967900372930411e-06, + "loss": 0.6947, + "step": 2986 + }, + { + "epoch": 0.3591655143389647, + "grad_norm": 2.6161482384470847, + "learning_rate": 2.9672186249127046e-06, + "loss": 0.7988, + "step": 2987 + }, + { + "epoch": 0.3592857572296038, + "grad_norm": 2.033744336889972, + "learning_rate": 2.9665367301687014e-06, + "loss": 0.7964, + "step": 2988 + }, + { + "epoch": 0.3594060001202429, + "grad_norm": 1.8954910835734613, + "learning_rate": 2.965854688801845e-06, + "loss": 0.7659, + "step": 2989 + }, + { + "epoch": 0.359526243010882, + "grad_norm": 2.217650665015998, + "learning_rate": 2.9651725009156005e-06, + "loss": 0.7691, + "step": 2990 + }, + { + "epoch": 0.3596464859015211, + "grad_norm": 1.6921792323367542, + "learning_rate": 2.964490166613454e-06, + "loss": 0.7454, + "step": 2991 + }, + { + "epoch": 0.35976672879216015, + "grad_norm": 0.7728292211812997, + "learning_rate": 2.963807685998917e-06, + "loss": 0.5778, + "step": 2992 + }, + { + "epoch": 0.35988697168279926, + "grad_norm": 1.6749162909244162, + "learning_rate": 2.9631250591755196e-06, + "loss": 0.7834, + "step": 2993 + }, + { + "epoch": 0.36000721457343837, + "grad_norm": 1.695030716361739, + "learning_rate": 2.962442286246817e-06, + "loss": 0.5779, + "step": 2994 + }, + { + "epoch": 0.3601274574640774, + "grad_norm": 1.6885279597440495, + "learning_rate": 2.9617593673163853e-06, + "loss": 0.7019, + "step": 2995 + }, + { + "epoch": 0.36024770035471654, + "grad_norm": 2.2670789591429155, + "learning_rate": 2.9610763024878216e-06, + "loss": 0.7814, + "step": 2996 + }, + { + "epoch": 0.3603679432453556, + "grad_norm": 2.131273881437234, + "learning_rate": 2.960393091864747e-06, + "loss": 0.9225, + "step": 2997 + }, + { + "epoch": 0.3604881861359947, + "grad_norm": 1.8449415462300527, + "learning_rate": 2.959709735550804e-06, + "loss": 0.7486, + "step": 2998 + }, + { + "epoch": 0.3606084290266338, + "grad_norm": 2.2694497282583277, + "learning_rate": 2.9590262336496575e-06, + "loss": 0.7563, + "step": 2999 + }, + { + "epoch": 0.36072867191727287, + "grad_norm": 2.4221248374789686, + "learning_rate": 2.9583425862649936e-06, + "loss": 0.8599, + "step": 3000 + }, + { + "epoch": 0.360848914807912, + "grad_norm": 2.153861450570671, + "learning_rate": 2.9576587935005215e-06, + "loss": 0.7421, + "step": 3001 + }, + { + "epoch": 0.3609691576985511, + "grad_norm": 49.109909418823, + "learning_rate": 2.9569748554599713e-06, + "loss": 0.7284, + "step": 3002 + }, + { + "epoch": 0.36108940058919015, + "grad_norm": 2.432269034324631, + "learning_rate": 2.956290772247097e-06, + "loss": 0.7346, + "step": 3003 + }, + { + "epoch": 0.36120964347982926, + "grad_norm": 1.7425562720605632, + "learning_rate": 2.9556065439656724e-06, + "loss": 0.7405, + "step": 3004 + }, + { + "epoch": 0.36132988637046837, + "grad_norm": 1.7375108018717735, + "learning_rate": 2.9549221707194952e-06, + "loss": 0.8221, + "step": 3005 + }, + { + "epoch": 0.3614501292611074, + "grad_norm": 2.172989593579249, + "learning_rate": 2.954237652612384e-06, + "loss": 0.7365, + "step": 3006 + }, + { + "epoch": 0.36157037215174653, + "grad_norm": 1.9720588610455376, + "learning_rate": 2.9535529897481796e-06, + "loss": 0.8489, + "step": 3007 + }, + { + "epoch": 0.36169061504238564, + "grad_norm": 2.9690561237112156, + "learning_rate": 2.9528681822307446e-06, + "loss": 0.777, + "step": 3008 + }, + { + "epoch": 0.3618108579330247, + "grad_norm": 1.9302222866069367, + "learning_rate": 2.952183230163964e-06, + "loss": 0.8282, + "step": 3009 + }, + { + "epoch": 0.3619311008236638, + "grad_norm": 2.176062081257802, + "learning_rate": 2.9514981336517448e-06, + "loss": 0.7367, + "step": 3010 + }, + { + "epoch": 0.36205134371430286, + "grad_norm": 1.9543035994693674, + "learning_rate": 2.950812892798015e-06, + "loss": 0.8189, + "step": 3011 + }, + { + "epoch": 0.362171586604942, + "grad_norm": 1.7662624457501936, + "learning_rate": 2.9501275077067256e-06, + "loss": 0.8768, + "step": 3012 + }, + { + "epoch": 0.3622918294955811, + "grad_norm": 1.6041347781227913, + "learning_rate": 2.949441978481848e-06, + "loss": 0.8894, + "step": 3013 + }, + { + "epoch": 0.36241207238622014, + "grad_norm": 1.9929440725990202, + "learning_rate": 2.9487563052273778e-06, + "loss": 0.806, + "step": 3014 + }, + { + "epoch": 0.36253231527685925, + "grad_norm": 1.9281067957088294, + "learning_rate": 2.94807048804733e-06, + "loss": 0.8592, + "step": 3015 + }, + { + "epoch": 0.36265255816749836, + "grad_norm": 1.8521772563220482, + "learning_rate": 2.9473845270457434e-06, + "loss": 0.9041, + "step": 3016 + }, + { + "epoch": 0.3627728010581374, + "grad_norm": 2.1076429249366804, + "learning_rate": 2.946698422326677e-06, + "loss": 0.6985, + "step": 3017 + }, + { + "epoch": 0.36289304394877653, + "grad_norm": 2.191188536248193, + "learning_rate": 2.946012173994213e-06, + "loss": 0.8041, + "step": 3018 + }, + { + "epoch": 0.36301328683941564, + "grad_norm": 1.4023078765106047, + "learning_rate": 2.945325782152454e-06, + "loss": 0.6861, + "step": 3019 + }, + { + "epoch": 0.3631335297300547, + "grad_norm": 2.249861077152453, + "learning_rate": 2.9446392469055257e-06, + "loss": 0.7929, + "step": 3020 + }, + { + "epoch": 0.3632537726206938, + "grad_norm": 1.7776659327663527, + "learning_rate": 2.9439525683575745e-06, + "loss": 0.8008, + "step": 3021 + }, + { + "epoch": 0.3633740155113329, + "grad_norm": 2.22453253046465, + "learning_rate": 2.9432657466127694e-06, + "loss": 0.7582, + "step": 3022 + }, + { + "epoch": 0.36349425840197197, + "grad_norm": 1.7284018958657161, + "learning_rate": 2.9425787817753007e-06, + "loss": 0.7697, + "step": 3023 + }, + { + "epoch": 0.3636145012926111, + "grad_norm": 1.6581008933536834, + "learning_rate": 2.94189167394938e-06, + "loss": 0.7209, + "step": 3024 + }, + { + "epoch": 0.3637347441832502, + "grad_norm": 1.7282485669402012, + "learning_rate": 2.941204423239241e-06, + "loss": 0.8119, + "step": 3025 + }, + { + "epoch": 0.36385498707388925, + "grad_norm": 1.9278402405707955, + "learning_rate": 2.9405170297491395e-06, + "loss": 0.7681, + "step": 3026 + }, + { + "epoch": 0.36397522996452836, + "grad_norm": 2.0787301007188312, + "learning_rate": 2.939829493583353e-06, + "loss": 0.8071, + "step": 3027 + }, + { + "epoch": 0.3640954728551674, + "grad_norm": 2.5636562437028223, + "learning_rate": 2.939141814846179e-06, + "loss": 0.8417, + "step": 3028 + }, + { + "epoch": 0.3642157157458065, + "grad_norm": 1.6248465979135072, + "learning_rate": 2.938453993641938e-06, + "loss": 0.8299, + "step": 3029 + }, + { + "epoch": 0.36433595863644563, + "grad_norm": 4.954186166393938, + "learning_rate": 2.937766030074973e-06, + "loss": 0.7077, + "step": 3030 + }, + { + "epoch": 0.3644562015270847, + "grad_norm": 2.251158945634613, + "learning_rate": 2.937077924249646e-06, + "loss": 0.8325, + "step": 3031 + }, + { + "epoch": 0.3645764444177238, + "grad_norm": 2.1750272201062506, + "learning_rate": 2.9363896762703443e-06, + "loss": 0.7627, + "step": 3032 + }, + { + "epoch": 0.3646966873083629, + "grad_norm": 1.779856220181918, + "learning_rate": 2.9357012862414725e-06, + "loss": 0.847, + "step": 3033 + }, + { + "epoch": 0.36481693019900197, + "grad_norm": 1.9437949761096642, + "learning_rate": 2.9350127542674593e-06, + "loss": 0.7184, + "step": 3034 + }, + { + "epoch": 0.3649371730896411, + "grad_norm": 1.9277286239584617, + "learning_rate": 2.934324080452755e-06, + "loss": 0.7691, + "step": 3035 + }, + { + "epoch": 0.3650574159802802, + "grad_norm": 2.859914593254996, + "learning_rate": 2.9336352649018307e-06, + "loss": 0.7827, + "step": 3036 + }, + { + "epoch": 0.36517765887091924, + "grad_norm": 2.303031507481103, + "learning_rate": 2.9329463077191783e-06, + "loss": 0.7042, + "step": 3037 + }, + { + "epoch": 0.36529790176155835, + "grad_norm": 2.30844730228973, + "learning_rate": 2.9322572090093135e-06, + "loss": 0.6486, + "step": 3038 + }, + { + "epoch": 0.36541814465219746, + "grad_norm": 5.140907349412567, + "learning_rate": 2.9315679688767713e-06, + "loss": 0.7874, + "step": 3039 + }, + { + "epoch": 0.3655383875428365, + "grad_norm": 2.69725024462713, + "learning_rate": 2.9308785874261085e-06, + "loss": 0.6655, + "step": 3040 + }, + { + "epoch": 0.36565863043347563, + "grad_norm": 2.0529985920213276, + "learning_rate": 2.9301890647619045e-06, + "loss": 0.823, + "step": 3041 + }, + { + "epoch": 0.36577887332411474, + "grad_norm": 1.9495930589418509, + "learning_rate": 2.929499400988759e-06, + "loss": 0.8059, + "step": 3042 + }, + { + "epoch": 0.3658991162147538, + "grad_norm": 1.747281142297447, + "learning_rate": 2.9288095962112927e-06, + "loss": 0.6588, + "step": 3043 + }, + { + "epoch": 0.3660193591053929, + "grad_norm": 1.9497619972725648, + "learning_rate": 2.9281196505341503e-06, + "loss": 0.8565, + "step": 3044 + }, + { + "epoch": 0.36613960199603196, + "grad_norm": 2.2105468836284015, + "learning_rate": 2.9274295640619946e-06, + "loss": 0.7908, + "step": 3045 + }, + { + "epoch": 0.36625984488667107, + "grad_norm": 1.8713742485918863, + "learning_rate": 2.9267393368995103e-06, + "loss": 0.7919, + "step": 3046 + }, + { + "epoch": 0.3663800877773102, + "grad_norm": 2.6100544301930864, + "learning_rate": 2.926048969151407e-06, + "loss": 0.7517, + "step": 3047 + }, + { + "epoch": 0.36650033066794924, + "grad_norm": 1.9133022744895525, + "learning_rate": 2.92535846092241e-06, + "loss": 0.6904, + "step": 3048 + }, + { + "epoch": 0.36662057355858835, + "grad_norm": 2.5737850767453088, + "learning_rate": 2.9246678123172704e-06, + "loss": 0.8295, + "step": 3049 + }, + { + "epoch": 0.36674081644922746, + "grad_norm": 4.500725581694439, + "learning_rate": 2.9239770234407596e-06, + "loss": 0.7531, + "step": 3050 + }, + { + "epoch": 0.3668610593398665, + "grad_norm": 1.8306417503124581, + "learning_rate": 2.9232860943976686e-06, + "loss": 0.6881, + "step": 3051 + }, + { + "epoch": 0.3669813022305056, + "grad_norm": 1.6840949142374733, + "learning_rate": 2.9225950252928115e-06, + "loss": 0.8447, + "step": 3052 + }, + { + "epoch": 0.36710154512114473, + "grad_norm": 2.351986463481849, + "learning_rate": 2.9219038162310217e-06, + "loss": 0.8263, + "step": 3053 + }, + { + "epoch": 0.3672217880117838, + "grad_norm": 2.5843164255130433, + "learning_rate": 2.921212467317157e-06, + "loss": 0.8337, + "step": 3054 + }, + { + "epoch": 0.3673420309024229, + "grad_norm": 1.9453284328068658, + "learning_rate": 2.920520978656093e-06, + "loss": 0.8087, + "step": 3055 + }, + { + "epoch": 0.367462273793062, + "grad_norm": 1.9925651641107096, + "learning_rate": 2.919829350352729e-06, + "loss": 0.7735, + "step": 3056 + }, + { + "epoch": 0.36758251668370107, + "grad_norm": 0.7690514942483787, + "learning_rate": 2.919137582511983e-06, + "loss": 0.6012, + "step": 3057 + }, + { + "epoch": 0.3677027595743402, + "grad_norm": 2.054750158710708, + "learning_rate": 2.918445675238797e-06, + "loss": 0.6461, + "step": 3058 + }, + { + "epoch": 0.36782300246497923, + "grad_norm": 2.1530331928407134, + "learning_rate": 2.917753628638132e-06, + "loss": 0.704, + "step": 3059 + }, + { + "epoch": 0.36794324535561834, + "grad_norm": 2.1324793516995197, + "learning_rate": 2.9170614428149716e-06, + "loss": 0.713, + "step": 3060 + }, + { + "epoch": 0.36806348824625745, + "grad_norm": 2.698212851586817, + "learning_rate": 2.9163691178743195e-06, + "loss": 0.8754, + "step": 3061 + }, + { + "epoch": 0.3681837311368965, + "grad_norm": 1.880120469961269, + "learning_rate": 2.9156766539212006e-06, + "loss": 0.7797, + "step": 3062 + }, + { + "epoch": 0.3683039740275356, + "grad_norm": 2.123066864340694, + "learning_rate": 2.9149840510606614e-06, + "loss": 0.7238, + "step": 3063 + }, + { + "epoch": 0.36842421691817473, + "grad_norm": 1.0265289092120393, + "learning_rate": 2.914291309397769e-06, + "loss": 0.6416, + "step": 3064 + }, + { + "epoch": 0.3685444598088138, + "grad_norm": 1.9073585663803827, + "learning_rate": 2.9135984290376117e-06, + "loss": 0.7938, + "step": 3065 + }, + { + "epoch": 0.3686647026994529, + "grad_norm": 1.745366706791833, + "learning_rate": 2.9129054100853e-06, + "loss": 0.8286, + "step": 3066 + }, + { + "epoch": 0.368784945590092, + "grad_norm": 1.7580163456023465, + "learning_rate": 2.912212252645963e-06, + "loss": 0.7649, + "step": 3067 + }, + { + "epoch": 0.36890518848073106, + "grad_norm": 2.2689493979691204, + "learning_rate": 2.9115189568247523e-06, + "loss": 0.7653, + "step": 3068 + }, + { + "epoch": 0.36902543137137017, + "grad_norm": 2.18953414072902, + "learning_rate": 2.910825522726841e-06, + "loss": 0.9266, + "step": 3069 + }, + { + "epoch": 0.3691456742620093, + "grad_norm": 2.6388903456120625, + "learning_rate": 2.9101319504574215e-06, + "loss": 0.7758, + "step": 3070 + }, + { + "epoch": 0.36926591715264834, + "grad_norm": 1.7045821509601549, + "learning_rate": 2.909438240121709e-06, + "loss": 0.7687, + "step": 3071 + }, + { + "epoch": 0.36938616004328745, + "grad_norm": 1.8115860838224391, + "learning_rate": 2.908744391824939e-06, + "loss": 0.7128, + "step": 3072 + }, + { + "epoch": 0.36950640293392656, + "grad_norm": 1.9668770615597502, + "learning_rate": 2.908050405672367e-06, + "loss": 0.7958, + "step": 3073 + }, + { + "epoch": 0.3696266458245656, + "grad_norm": 1.8838118437580504, + "learning_rate": 2.9073562817692703e-06, + "loss": 0.8002, + "step": 3074 + }, + { + "epoch": 0.3697468887152047, + "grad_norm": 0.7228105534957359, + "learning_rate": 2.9066620202209468e-06, + "loss": 0.5658, + "step": 3075 + }, + { + "epoch": 0.3698671316058438, + "grad_norm": 1.806674870346888, + "learning_rate": 2.905967621132716e-06, + "loss": 0.7804, + "step": 3076 + }, + { + "epoch": 0.3699873744964829, + "grad_norm": 2.756003069899445, + "learning_rate": 2.9052730846099172e-06, + "loss": 0.7523, + "step": 3077 + }, + { + "epoch": 0.370107617387122, + "grad_norm": 0.885332544855931, + "learning_rate": 2.9045784107579123e-06, + "loss": 0.6098, + "step": 3078 + }, + { + "epoch": 0.37022786027776106, + "grad_norm": 1.9091528380864597, + "learning_rate": 2.9038835996820807e-06, + "loss": 0.6783, + "step": 3079 + }, + { + "epoch": 0.37034810316840017, + "grad_norm": 1.7660452418553458, + "learning_rate": 2.903188651487826e-06, + "loss": 0.8022, + "step": 3080 + }, + { + "epoch": 0.3704683460590393, + "grad_norm": 2.1512626799188133, + "learning_rate": 2.902493566280571e-06, + "loss": 0.8656, + "step": 3081 + }, + { + "epoch": 0.37058858894967833, + "grad_norm": 1.8590970512817762, + "learning_rate": 2.9017983441657595e-06, + "loss": 0.8206, + "step": 3082 + }, + { + "epoch": 0.37070883184031744, + "grad_norm": 2.2502280225711466, + "learning_rate": 2.9011029852488564e-06, + "loss": 0.7632, + "step": 3083 + }, + { + "epoch": 0.37082907473095655, + "grad_norm": 0.992983761365559, + "learning_rate": 2.9004074896353465e-06, + "loss": 0.6254, + "step": 3084 + }, + { + "epoch": 0.3709493176215956, + "grad_norm": 1.8161742359591875, + "learning_rate": 2.8997118574307362e-06, + "loss": 0.8195, + "step": 3085 + }, + { + "epoch": 0.3710695605122347, + "grad_norm": 2.001436671188798, + "learning_rate": 2.899016088740553e-06, + "loss": 0.7482, + "step": 3086 + }, + { + "epoch": 0.37118980340287383, + "grad_norm": 1.8189317435470485, + "learning_rate": 2.898320183670344e-06, + "loss": 0.7964, + "step": 3087 + }, + { + "epoch": 0.3713100462935129, + "grad_norm": 1.7942003993885114, + "learning_rate": 2.8976241423256767e-06, + "loss": 0.8941, + "step": 3088 + }, + { + "epoch": 0.371430289184152, + "grad_norm": 2.725316590983082, + "learning_rate": 2.896927964812142e-06, + "loss": 0.6893, + "step": 3089 + }, + { + "epoch": 0.37155053207479105, + "grad_norm": 2.585856548782501, + "learning_rate": 2.8962316512353465e-06, + "loss": 0.7559, + "step": 3090 + }, + { + "epoch": 0.37167077496543016, + "grad_norm": 1.6274834509270384, + "learning_rate": 2.8955352017009233e-06, + "loss": 0.7511, + "step": 3091 + }, + { + "epoch": 0.3717910178560693, + "grad_norm": 2.0184071583349152, + "learning_rate": 2.8948386163145212e-06, + "loss": 0.7843, + "step": 3092 + }, + { + "epoch": 0.3719112607467083, + "grad_norm": 1.7892239426978986, + "learning_rate": 2.8941418951818135e-06, + "loss": 0.7973, + "step": 3093 + }, + { + "epoch": 0.37203150363734744, + "grad_norm": 2.387061461272281, + "learning_rate": 2.8934450384084903e-06, + "loss": 0.7211, + "step": 3094 + }, + { + "epoch": 0.37215174652798655, + "grad_norm": 3.3021822331937867, + "learning_rate": 2.8927480461002653e-06, + "loss": 0.7036, + "step": 3095 + }, + { + "epoch": 0.3722719894186256, + "grad_norm": 2.830612016421927, + "learning_rate": 2.892050918362872e-06, + "loss": 0.8621, + "step": 3096 + }, + { + "epoch": 0.3723922323092647, + "grad_norm": 0.8601407887301811, + "learning_rate": 2.8913536553020626e-06, + "loss": 0.5602, + "step": 3097 + }, + { + "epoch": 0.3725124751999038, + "grad_norm": 2.01424626946598, + "learning_rate": 2.8906562570236137e-06, + "loss": 0.8568, + "step": 3098 + }, + { + "epoch": 0.3726327180905429, + "grad_norm": 1.5416216027486775, + "learning_rate": 2.889958723633318e-06, + "loss": 0.766, + "step": 3099 + }, + { + "epoch": 0.372752960981182, + "grad_norm": 2.158071439704253, + "learning_rate": 2.889261055236992e-06, + "loss": 0.7401, + "step": 3100 + }, + { + "epoch": 0.3728732038718211, + "grad_norm": 1.7665112052149639, + "learning_rate": 2.8885632519404704e-06, + "loss": 0.8349, + "step": 3101 + }, + { + "epoch": 0.37299344676246016, + "grad_norm": 3.975943826389894, + "learning_rate": 2.8878653138496107e-06, + "loss": 0.7641, + "step": 3102 + }, + { + "epoch": 0.37311368965309927, + "grad_norm": 2.710160641571824, + "learning_rate": 2.8871672410702878e-06, + "loss": 0.7657, + "step": 3103 + }, + { + "epoch": 0.3732339325437384, + "grad_norm": 2.218242748464036, + "learning_rate": 2.8864690337084008e-06, + "loss": 0.823, + "step": 3104 + }, + { + "epoch": 0.37335417543437743, + "grad_norm": 1.859092049747786, + "learning_rate": 2.885770691869866e-06, + "loss": 0.7869, + "step": 3105 + }, + { + "epoch": 0.37347441832501654, + "grad_norm": 2.120160819737572, + "learning_rate": 2.8850722156606207e-06, + "loss": 0.7539, + "step": 3106 + }, + { + "epoch": 0.3735946612156556, + "grad_norm": 1.612875825126228, + "learning_rate": 2.8843736051866252e-06, + "loss": 0.6763, + "step": 3107 + }, + { + "epoch": 0.3737149041062947, + "grad_norm": 1.6081249975092533, + "learning_rate": 2.8836748605538557e-06, + "loss": 0.6918, + "step": 3108 + }, + { + "epoch": 0.3738351469969338, + "grad_norm": 2.0031232289322807, + "learning_rate": 2.882975981868313e-06, + "loss": 0.6371, + "step": 3109 + }, + { + "epoch": 0.3739553898875729, + "grad_norm": 2.388917799584956, + "learning_rate": 2.882276969236016e-06, + "loss": 0.6995, + "step": 3110 + }, + { + "epoch": 0.374075632778212, + "grad_norm": 1.9963342928542012, + "learning_rate": 2.881577822763005e-06, + "loss": 0.7754, + "step": 3111 + }, + { + "epoch": 0.3741958756688511, + "grad_norm": 1.9038022695017105, + "learning_rate": 2.880878542555338e-06, + "loss": 0.8837, + "step": 3112 + }, + { + "epoch": 0.37431611855949015, + "grad_norm": 2.0686674672054637, + "learning_rate": 2.8801791287190976e-06, + "loss": 0.8081, + "step": 3113 + }, + { + "epoch": 0.37443636145012926, + "grad_norm": 2.853938870957575, + "learning_rate": 2.8794795813603817e-06, + "loss": 0.8655, + "step": 3114 + }, + { + "epoch": 0.3745566043407684, + "grad_norm": 1.8167676340562033, + "learning_rate": 2.878779900585314e-06, + "loss": 0.8227, + "step": 3115 + }, + { + "epoch": 0.37467684723140743, + "grad_norm": 1.6507828647011669, + "learning_rate": 2.8780800865000336e-06, + "loss": 0.7577, + "step": 3116 + }, + { + "epoch": 0.37479709012204654, + "grad_norm": 0.9899988439822913, + "learning_rate": 2.877380139210702e-06, + "loss": 0.5913, + "step": 3117 + }, + { + "epoch": 0.37491733301268565, + "grad_norm": 1.9981211539414339, + "learning_rate": 2.876680058823501e-06, + "loss": 0.7671, + "step": 3118 + }, + { + "epoch": 0.3750375759033247, + "grad_norm": 2.259078360641849, + "learning_rate": 2.8759798454446314e-06, + "loss": 0.6628, + "step": 3119 + }, + { + "epoch": 0.3751578187939638, + "grad_norm": 1.9291216062991619, + "learning_rate": 2.8752794991803173e-06, + "loss": 0.8162, + "step": 3120 + }, + { + "epoch": 0.37527806168460287, + "grad_norm": 2.0553634169995196, + "learning_rate": 2.8745790201367976e-06, + "loss": 0.755, + "step": 3121 + }, + { + "epoch": 0.375398304575242, + "grad_norm": 1.9857617436272694, + "learning_rate": 2.8738784084203373e-06, + "loss": 0.8425, + "step": 3122 + }, + { + "epoch": 0.3755185474658811, + "grad_norm": 1.6647436332941796, + "learning_rate": 2.873177664137216e-06, + "loss": 0.7882, + "step": 3123 + }, + { + "epoch": 0.37563879035652015, + "grad_norm": 2.2671959002054605, + "learning_rate": 2.8724767873937384e-06, + "loss": 0.6924, + "step": 3124 + }, + { + "epoch": 0.37575903324715926, + "grad_norm": 2.435236638063726, + "learning_rate": 2.871775778296225e-06, + "loss": 0.881, + "step": 3125 + }, + { + "epoch": 0.37587927613779837, + "grad_norm": 2.1521755921294874, + "learning_rate": 2.8710746369510196e-06, + "loss": 0.7924, + "step": 3126 + }, + { + "epoch": 0.3759995190284374, + "grad_norm": 2.4498081988559104, + "learning_rate": 2.8703733634644846e-06, + "loss": 0.8319, + "step": 3127 + }, + { + "epoch": 0.37611976191907653, + "grad_norm": 2.5107536775614663, + "learning_rate": 2.869671957943002e-06, + "loss": 0.798, + "step": 3128 + }, + { + "epoch": 0.37624000480971564, + "grad_norm": 1.8962013669903377, + "learning_rate": 2.8689704204929747e-06, + "loss": 0.7507, + "step": 3129 + }, + { + "epoch": 0.3763602477003547, + "grad_norm": 2.1637105827774077, + "learning_rate": 2.8682687512208253e-06, + "loss": 0.819, + "step": 3130 + }, + { + "epoch": 0.3764804905909938, + "grad_norm": 1.9619578104567403, + "learning_rate": 2.8675669502329972e-06, + "loss": 0.8155, + "step": 3131 + }, + { + "epoch": 0.3766007334816329, + "grad_norm": 2.614082886319378, + "learning_rate": 2.866865017635952e-06, + "loss": 0.865, + "step": 3132 + }, + { + "epoch": 0.376720976372272, + "grad_norm": 1.5689130781906704, + "learning_rate": 2.866162953536174e-06, + "loss": 0.7993, + "step": 3133 + }, + { + "epoch": 0.3768412192629111, + "grad_norm": 1.6424912510236793, + "learning_rate": 2.8654607580401634e-06, + "loss": 0.7531, + "step": 3134 + }, + { + "epoch": 0.3769614621535502, + "grad_norm": 0.8960670380739182, + "learning_rate": 2.8647584312544446e-06, + "loss": 0.652, + "step": 3135 + }, + { + "epoch": 0.37708170504418925, + "grad_norm": 1.3889883890538657, + "learning_rate": 2.864055973285559e-06, + "loss": 0.8544, + "step": 3136 + }, + { + "epoch": 0.37720194793482836, + "grad_norm": 1.8210344970175614, + "learning_rate": 2.8633533842400698e-06, + "loss": 0.8643, + "step": 3137 + }, + { + "epoch": 0.3773221908254674, + "grad_norm": 1.8741490022339373, + "learning_rate": 2.862650664224558e-06, + "loss": 0.7843, + "step": 3138 + }, + { + "epoch": 0.37744243371610653, + "grad_norm": 1.9750927070395345, + "learning_rate": 2.861947813345627e-06, + "loss": 0.7004, + "step": 3139 + }, + { + "epoch": 0.37756267660674564, + "grad_norm": 1.7977335330111888, + "learning_rate": 2.8612448317098974e-06, + "loss": 0.729, + "step": 3140 + }, + { + "epoch": 0.3776829194973847, + "grad_norm": 2.088803269387897, + "learning_rate": 2.8605417194240114e-06, + "loss": 0.8342, + "step": 3141 + }, + { + "epoch": 0.3778031623880238, + "grad_norm": 1.8276041707252408, + "learning_rate": 2.8598384765946315e-06, + "loss": 0.7959, + "step": 3142 + }, + { + "epoch": 0.3779234052786629, + "grad_norm": 2.0554037839974404, + "learning_rate": 2.8591351033284377e-06, + "loss": 0.7231, + "step": 3143 + }, + { + "epoch": 0.37804364816930197, + "grad_norm": 2.065126884755603, + "learning_rate": 2.8584315997321325e-06, + "loss": 0.8423, + "step": 3144 + }, + { + "epoch": 0.3781638910599411, + "grad_norm": 2.4199937959071307, + "learning_rate": 2.8577279659124356e-06, + "loss": 0.7922, + "step": 3145 + }, + { + "epoch": 0.3782841339505802, + "grad_norm": 1.7447607768346747, + "learning_rate": 2.857024201976089e-06, + "loss": 0.8359, + "step": 3146 + }, + { + "epoch": 0.37840437684121925, + "grad_norm": 2.3171348519495236, + "learning_rate": 2.8563203080298516e-06, + "loss": 0.7404, + "step": 3147 + }, + { + "epoch": 0.37852461973185836, + "grad_norm": 2.0614369282535616, + "learning_rate": 2.855616284180505e-06, + "loss": 0.9075, + "step": 3148 + }, + { + "epoch": 0.37864486262249747, + "grad_norm": 0.8805764498840652, + "learning_rate": 2.8549121305348477e-06, + "loss": 0.6612, + "step": 3149 + }, + { + "epoch": 0.3787651055131365, + "grad_norm": 2.7046328181530064, + "learning_rate": 2.8542078471997006e-06, + "loss": 0.83, + "step": 3150 + }, + { + "epoch": 0.37888534840377563, + "grad_norm": 2.021802912138927, + "learning_rate": 2.8535034342819013e-06, + "loss": 0.765, + "step": 3151 + }, + { + "epoch": 0.37900559129441475, + "grad_norm": 1.6526879928236713, + "learning_rate": 2.85279889188831e-06, + "loss": 0.7252, + "step": 3152 + }, + { + "epoch": 0.3791258341850538, + "grad_norm": 1.780734941403075, + "learning_rate": 2.852094220125805e-06, + "loss": 0.8142, + "step": 3153 + }, + { + "epoch": 0.3792460770756929, + "grad_norm": 2.19003965908006, + "learning_rate": 2.8513894191012846e-06, + "loss": 0.7155, + "step": 3154 + }, + { + "epoch": 0.37936631996633197, + "grad_norm": 1.5734785815182175, + "learning_rate": 2.8506844889216664e-06, + "loss": 0.7926, + "step": 3155 + }, + { + "epoch": 0.3794865628569711, + "grad_norm": 0.8777934171259882, + "learning_rate": 2.849979429693887e-06, + "loss": 0.6289, + "step": 3156 + }, + { + "epoch": 0.3796068057476102, + "grad_norm": 2.0543339066070065, + "learning_rate": 2.8492742415249042e-06, + "loss": 0.7503, + "step": 3157 + }, + { + "epoch": 0.37972704863824924, + "grad_norm": 1.7311967476597414, + "learning_rate": 2.848568924521694e-06, + "loss": 0.765, + "step": 3158 + }, + { + "epoch": 0.37984729152888835, + "grad_norm": 1.9310856128838798, + "learning_rate": 2.8478634787912526e-06, + "loss": 0.7391, + "step": 3159 + }, + { + "epoch": 0.37996753441952746, + "grad_norm": 1.9417849503317717, + "learning_rate": 2.847157904440596e-06, + "loss": 0.7686, + "step": 3160 + }, + { + "epoch": 0.3800877773101665, + "grad_norm": 1.5509505864086461, + "learning_rate": 2.846452201576759e-06, + "loss": 0.7439, + "step": 3161 + }, + { + "epoch": 0.38020802020080563, + "grad_norm": 0.845664712857265, + "learning_rate": 2.845746370306795e-06, + "loss": 0.6278, + "step": 3162 + }, + { + "epoch": 0.38032826309144474, + "grad_norm": 2.286586633325241, + "learning_rate": 2.84504041073778e-06, + "loss": 0.7879, + "step": 3163 + }, + { + "epoch": 0.3804485059820838, + "grad_norm": 1.9287962303134276, + "learning_rate": 2.844334322976806e-06, + "loss": 0.7928, + "step": 3164 + }, + { + "epoch": 0.3805687488727229, + "grad_norm": 1.778033447590377, + "learning_rate": 2.8436281071309866e-06, + "loss": 0.8384, + "step": 3165 + }, + { + "epoch": 0.380688991763362, + "grad_norm": 0.7359692738795055, + "learning_rate": 2.842921763307455e-06, + "loss": 0.53, + "step": 3166 + }, + { + "epoch": 0.38080923465400107, + "grad_norm": 1.8652168632604003, + "learning_rate": 2.842215291613361e-06, + "loss": 0.8273, + "step": 3167 + }, + { + "epoch": 0.3809294775446402, + "grad_norm": 0.7949507049656274, + "learning_rate": 2.8415086921558774e-06, + "loss": 0.5937, + "step": 3168 + }, + { + "epoch": 0.38104972043527924, + "grad_norm": 1.6935685852416726, + "learning_rate": 2.840801965042194e-06, + "loss": 0.7919, + "step": 3169 + }, + { + "epoch": 0.38116996332591835, + "grad_norm": 1.8142524665772648, + "learning_rate": 2.840095110379521e-06, + "loss": 0.8396, + "step": 3170 + }, + { + "epoch": 0.38129020621655746, + "grad_norm": 0.7468212236014151, + "learning_rate": 2.8393881282750884e-06, + "loss": 0.5398, + "step": 3171 + }, + { + "epoch": 0.3814104491071965, + "grad_norm": 1.7687944979512646, + "learning_rate": 2.838681018836144e-06, + "loss": 0.7937, + "step": 3172 + }, + { + "epoch": 0.3815306919978356, + "grad_norm": 2.153081805456153, + "learning_rate": 2.837973782169955e-06, + "loss": 0.7945, + "step": 3173 + }, + { + "epoch": 0.38165093488847474, + "grad_norm": 0.8262619799093169, + "learning_rate": 2.8372664183838096e-06, + "loss": 0.5921, + "step": 3174 + }, + { + "epoch": 0.3817711777791138, + "grad_norm": 2.5090099102385572, + "learning_rate": 2.836558927585015e-06, + "loss": 0.6962, + "step": 3175 + }, + { + "epoch": 0.3818914206697529, + "grad_norm": 1.7975474550564532, + "learning_rate": 2.8358513098808957e-06, + "loss": 0.8246, + "step": 3176 + }, + { + "epoch": 0.382011663560392, + "grad_norm": 1.7664823222999344, + "learning_rate": 2.835143565378798e-06, + "loss": 0.7752, + "step": 3177 + }, + { + "epoch": 0.38213190645103107, + "grad_norm": 1.908000517507199, + "learning_rate": 2.8344356941860847e-06, + "loss": 0.795, + "step": 3178 + }, + { + "epoch": 0.3822521493416702, + "grad_norm": 2.302423121259227, + "learning_rate": 2.8337276964101403e-06, + "loss": 0.6679, + "step": 3179 + }, + { + "epoch": 0.3823723922323093, + "grad_norm": 2.079154311801487, + "learning_rate": 2.833019572158367e-06, + "loss": 0.7654, + "step": 3180 + }, + { + "epoch": 0.38249263512294834, + "grad_norm": 1.8243676723164481, + "learning_rate": 2.8323113215381872e-06, + "loss": 0.807, + "step": 3181 + }, + { + "epoch": 0.38261287801358745, + "grad_norm": 1.9032526058017336, + "learning_rate": 2.831602944657042e-06, + "loss": 0.7629, + "step": 3182 + }, + { + "epoch": 0.38273312090422656, + "grad_norm": 2.1318151356778285, + "learning_rate": 2.830894441622391e-06, + "loss": 0.7505, + "step": 3183 + }, + { + "epoch": 0.3828533637948656, + "grad_norm": 2.033828507252178, + "learning_rate": 2.8301858125417134e-06, + "loss": 0.8024, + "step": 3184 + }, + { + "epoch": 0.38297360668550473, + "grad_norm": 1.838458469082591, + "learning_rate": 2.8294770575225082e-06, + "loss": 0.7507, + "step": 3185 + }, + { + "epoch": 0.3830938495761438, + "grad_norm": 1.8784470978578778, + "learning_rate": 2.828768176672293e-06, + "loss": 0.8431, + "step": 3186 + }, + { + "epoch": 0.3832140924667829, + "grad_norm": 2.646527977615901, + "learning_rate": 2.8280591700986044e-06, + "loss": 0.7233, + "step": 3187 + }, + { + "epoch": 0.383334335357422, + "grad_norm": 1.745870076644859, + "learning_rate": 2.827350037908999e-06, + "loss": 0.7518, + "step": 3188 + }, + { + "epoch": 0.38345457824806106, + "grad_norm": 3.6419848084151263, + "learning_rate": 2.8266407802110496e-06, + "loss": 0.791, + "step": 3189 + }, + { + "epoch": 0.3835748211387002, + "grad_norm": 2.640384008482432, + "learning_rate": 2.8259313971123515e-06, + "loss": 0.7631, + "step": 3190 + }, + { + "epoch": 0.3836950640293393, + "grad_norm": 1.6019602211955728, + "learning_rate": 2.8252218887205166e-06, + "loss": 0.7868, + "step": 3191 + }, + { + "epoch": 0.38381530691997834, + "grad_norm": 1.9414559862608616, + "learning_rate": 2.824512255143178e-06, + "loss": 0.8099, + "step": 3192 + }, + { + "epoch": 0.38393554981061745, + "grad_norm": 1.7994412217139637, + "learning_rate": 2.8238024964879855e-06, + "loss": 0.7959, + "step": 3193 + }, + { + "epoch": 0.38405579270125656, + "grad_norm": 1.9222112842143262, + "learning_rate": 2.8230926128626095e-06, + "loss": 0.7768, + "step": 3194 + }, + { + "epoch": 0.3841760355918956, + "grad_norm": 1.9628745918188477, + "learning_rate": 2.822382604374738e-06, + "loss": 0.8017, + "step": 3195 + }, + { + "epoch": 0.3842962784825347, + "grad_norm": 2.036282569994769, + "learning_rate": 2.8216724711320793e-06, + "loss": 0.6615, + "step": 3196 + }, + { + "epoch": 0.38441652137317384, + "grad_norm": 1.5062905966958486, + "learning_rate": 2.820962213242361e-06, + "loss": 0.7967, + "step": 3197 + }, + { + "epoch": 0.3845367642638129, + "grad_norm": 2.2397574305325416, + "learning_rate": 2.8202518308133264e-06, + "loss": 0.8444, + "step": 3198 + }, + { + "epoch": 0.384657007154452, + "grad_norm": 2.017613103218093, + "learning_rate": 2.8195413239527426e-06, + "loss": 0.7366, + "step": 3199 + }, + { + "epoch": 0.38477725004509106, + "grad_norm": 1.9162103552054843, + "learning_rate": 2.8188306927683906e-06, + "loss": 0.8079, + "step": 3200 + }, + { + "epoch": 0.38489749293573017, + "grad_norm": 2.0527621339033537, + "learning_rate": 2.818119937368074e-06, + "loss": 0.7522, + "step": 3201 + }, + { + "epoch": 0.3850177358263693, + "grad_norm": 2.0011863766051157, + "learning_rate": 2.817409057859613e-06, + "loss": 0.6644, + "step": 3202 + }, + { + "epoch": 0.38513797871700833, + "grad_norm": 2.025140347738781, + "learning_rate": 2.8166980543508482e-06, + "loss": 0.7988, + "step": 3203 + }, + { + "epoch": 0.38525822160764744, + "grad_norm": 1.8281253912509596, + "learning_rate": 2.815986926949638e-06, + "loss": 0.8047, + "step": 3204 + }, + { + "epoch": 0.38537846449828655, + "grad_norm": 1.775722157963203, + "learning_rate": 2.8152756757638597e-06, + "loss": 0.8112, + "step": 3205 + }, + { + "epoch": 0.3854987073889256, + "grad_norm": 1.921929831614288, + "learning_rate": 2.8145643009014093e-06, + "loss": 0.8538, + "step": 3206 + }, + { + "epoch": 0.3856189502795647, + "grad_norm": 1.9544718256705795, + "learning_rate": 2.813852802470202e-06, + "loss": 0.794, + "step": 3207 + }, + { + "epoch": 0.38573919317020383, + "grad_norm": 1.8283808602502365, + "learning_rate": 2.8131411805781717e-06, + "loss": 0.7293, + "step": 3208 + }, + { + "epoch": 0.3858594360608429, + "grad_norm": 2.521015244743136, + "learning_rate": 2.8124294353332707e-06, + "loss": 0.6549, + "step": 3209 + }, + { + "epoch": 0.385979678951482, + "grad_norm": 2.2949242644601924, + "learning_rate": 2.8117175668434713e-06, + "loss": 0.7789, + "step": 3210 + }, + { + "epoch": 0.3860999218421211, + "grad_norm": 2.5149044165829326, + "learning_rate": 2.811005575216762e-06, + "loss": 0.7109, + "step": 3211 + }, + { + "epoch": 0.38622016473276016, + "grad_norm": 1.4875780581982205, + "learning_rate": 2.8102934605611513e-06, + "loss": 0.7926, + "step": 3212 + }, + { + "epoch": 0.3863404076233993, + "grad_norm": 2.097146311460496, + "learning_rate": 2.8095812229846665e-06, + "loss": 0.6792, + "step": 3213 + }, + { + "epoch": 0.3864606505140384, + "grad_norm": 2.3493349798213345, + "learning_rate": 2.808868862595355e-06, + "loss": 0.6943, + "step": 3214 + }, + { + "epoch": 0.38658089340467744, + "grad_norm": 3.1312066050433014, + "learning_rate": 2.8081563795012795e-06, + "loss": 0.8004, + "step": 3215 + }, + { + "epoch": 0.38670113629531655, + "grad_norm": 1.8042672729045952, + "learning_rate": 2.807443773810524e-06, + "loss": 0.7467, + "step": 3216 + }, + { + "epoch": 0.3868213791859556, + "grad_norm": 1.7251024243450592, + "learning_rate": 2.80673104563119e-06, + "loss": 0.8981, + "step": 3217 + }, + { + "epoch": 0.3869416220765947, + "grad_norm": 2.4300866489110673, + "learning_rate": 2.8060181950713976e-06, + "loss": 0.7881, + "step": 3218 + }, + { + "epoch": 0.3870618649672338, + "grad_norm": 2.1236985653132314, + "learning_rate": 2.805305222239286e-06, + "loss": 0.8134, + "step": 3219 + }, + { + "epoch": 0.3871821078578729, + "grad_norm": 1.8474253859054868, + "learning_rate": 2.8045921272430118e-06, + "loss": 0.7406, + "step": 3220 + }, + { + "epoch": 0.387302350748512, + "grad_norm": 2.21917838675708, + "learning_rate": 2.803878910190753e-06, + "loss": 0.7823, + "step": 3221 + }, + { + "epoch": 0.3874225936391511, + "grad_norm": 2.4334090700077224, + "learning_rate": 2.8031655711907017e-06, + "loss": 0.8265, + "step": 3222 + }, + { + "epoch": 0.38754283652979016, + "grad_norm": 2.4683790290341032, + "learning_rate": 2.8024521103510723e-06, + "loss": 0.8106, + "step": 3223 + }, + { + "epoch": 0.38766307942042927, + "grad_norm": 1.9197108171162525, + "learning_rate": 2.8017385277800952e-06, + "loss": 0.7567, + "step": 3224 + }, + { + "epoch": 0.3877833223110684, + "grad_norm": 2.1123072835383376, + "learning_rate": 2.8010248235860213e-06, + "loss": 0.7573, + "step": 3225 + }, + { + "epoch": 0.38790356520170743, + "grad_norm": 0.8357952389034028, + "learning_rate": 2.8003109978771192e-06, + "loss": 0.628, + "step": 3226 + }, + { + "epoch": 0.38802380809234654, + "grad_norm": 1.9906738514797164, + "learning_rate": 2.799597050761674e-06, + "loss": 0.7939, + "step": 3227 + }, + { + "epoch": 0.38814405098298566, + "grad_norm": 1.8603707860719383, + "learning_rate": 2.7988829823479924e-06, + "loss": 0.7987, + "step": 3228 + }, + { + "epoch": 0.3882642938736247, + "grad_norm": 1.7134168245082855, + "learning_rate": 2.7981687927443976e-06, + "loss": 0.6423, + "step": 3229 + }, + { + "epoch": 0.3883845367642638, + "grad_norm": 1.991162683520579, + "learning_rate": 2.797454482059231e-06, + "loss": 0.8614, + "step": 3230 + }, + { + "epoch": 0.3885047796549029, + "grad_norm": 1.715982741004564, + "learning_rate": 2.7967400504008537e-06, + "loss": 0.8519, + "step": 3231 + }, + { + "epoch": 0.388625022545542, + "grad_norm": 0.792921055454036, + "learning_rate": 2.7960254978776456e-06, + "loss": 0.5754, + "step": 3232 + }, + { + "epoch": 0.3887452654361811, + "grad_norm": 3.0060795535425475, + "learning_rate": 2.7953108245980006e-06, + "loss": 0.8215, + "step": 3233 + }, + { + "epoch": 0.38886550832682015, + "grad_norm": 1.7199188419534492, + "learning_rate": 2.7945960306703365e-06, + "loss": 0.7377, + "step": 3234 + }, + { + "epoch": 0.38898575121745926, + "grad_norm": 1.7318238061437206, + "learning_rate": 2.7938811162030865e-06, + "loss": 0.6624, + "step": 3235 + }, + { + "epoch": 0.3891059941080984, + "grad_norm": 2.1812589716248607, + "learning_rate": 2.793166081304702e-06, + "loss": 0.8281, + "step": 3236 + }, + { + "epoch": 0.38922623699873743, + "grad_norm": 1.754355120408644, + "learning_rate": 2.7924509260836543e-06, + "loss": 0.8321, + "step": 3237 + }, + { + "epoch": 0.38934647988937654, + "grad_norm": 1.435048625401326, + "learning_rate": 2.791735650648431e-06, + "loss": 0.6876, + "step": 3238 + }, + { + "epoch": 0.38946672278001565, + "grad_norm": 2.0837141324751016, + "learning_rate": 2.791020255107538e-06, + "loss": 0.7496, + "step": 3239 + }, + { + "epoch": 0.3895869656706547, + "grad_norm": 1.607188693407162, + "learning_rate": 2.7903047395695023e-06, + "loss": 0.8121, + "step": 3240 + }, + { + "epoch": 0.3897072085612938, + "grad_norm": 2.041953773812225, + "learning_rate": 2.789589104142865e-06, + "loss": 0.9043, + "step": 3241 + }, + { + "epoch": 0.3898274514519329, + "grad_norm": 1.6172864211046243, + "learning_rate": 2.7888733489361895e-06, + "loss": 0.7744, + "step": 3242 + }, + { + "epoch": 0.389947694342572, + "grad_norm": 0.7330329075539354, + "learning_rate": 2.788157474058054e-06, + "loss": 0.5875, + "step": 3243 + }, + { + "epoch": 0.3900679372332111, + "grad_norm": 1.5710253260056541, + "learning_rate": 2.7874414796170555e-06, + "loss": 0.7042, + "step": 3244 + }, + { + "epoch": 0.3901881801238502, + "grad_norm": 2.4836872036448785, + "learning_rate": 2.7867253657218113e-06, + "loss": 0.8531, + "step": 3245 + }, + { + "epoch": 0.39030842301448926, + "grad_norm": 2.041896677360445, + "learning_rate": 2.7860091324809544e-06, + "loss": 0.7317, + "step": 3246 + }, + { + "epoch": 0.39042866590512837, + "grad_norm": 1.6541797727102474, + "learning_rate": 2.7852927800031377e-06, + "loss": 0.8184, + "step": 3247 + }, + { + "epoch": 0.3905489087957674, + "grad_norm": 2.2455335366244755, + "learning_rate": 2.7845763083970298e-06, + "loss": 0.83, + "step": 3248 + }, + { + "epoch": 0.39066915168640653, + "grad_norm": 2.0499470494572813, + "learning_rate": 2.7838597177713205e-06, + "loss": 0.8245, + "step": 3249 + }, + { + "epoch": 0.39078939457704565, + "grad_norm": 1.7747926617804037, + "learning_rate": 2.7831430082347143e-06, + "loss": 0.7389, + "step": 3250 + }, + { + "epoch": 0.3909096374676847, + "grad_norm": 2.0069255131316632, + "learning_rate": 2.7824261798959373e-06, + "loss": 0.8247, + "step": 3251 + }, + { + "epoch": 0.3910298803583238, + "grad_norm": 2.026400013853089, + "learning_rate": 2.78170923286373e-06, + "loss": 0.801, + "step": 3252 + }, + { + "epoch": 0.3911501232489629, + "grad_norm": 2.2855205538936874, + "learning_rate": 2.780992167246854e-06, + "loss": 0.842, + "step": 3253 + }, + { + "epoch": 0.391270366139602, + "grad_norm": 0.9885443838475424, + "learning_rate": 2.7802749831540883e-06, + "loss": 0.7222, + "step": 3254 + }, + { + "epoch": 0.3913906090302411, + "grad_norm": 2.4759064295748847, + "learning_rate": 2.7795576806942268e-06, + "loss": 0.8219, + "step": 3255 + }, + { + "epoch": 0.3915108519208802, + "grad_norm": 0.7798194758268385, + "learning_rate": 2.778840259976085e-06, + "loss": 0.5497, + "step": 3256 + }, + { + "epoch": 0.39163109481151925, + "grad_norm": 2.3108445505428707, + "learning_rate": 2.778122721108495e-06, + "loss": 0.7741, + "step": 3257 + }, + { + "epoch": 0.39175133770215836, + "grad_norm": 2.3277014020571176, + "learning_rate": 2.7774050642003076e-06, + "loss": 0.8884, + "step": 3258 + }, + { + "epoch": 0.3918715805927975, + "grad_norm": 11.95352902197094, + "learning_rate": 2.7766872893603896e-06, + "loss": 0.9382, + "step": 3259 + }, + { + "epoch": 0.39199182348343653, + "grad_norm": 1.926725082776845, + "learning_rate": 2.7759693966976275e-06, + "loss": 0.7363, + "step": 3260 + }, + { + "epoch": 0.39211206637407564, + "grad_norm": 1.8560105345936766, + "learning_rate": 2.7752513863209242e-06, + "loss": 0.8557, + "step": 3261 + }, + { + "epoch": 0.39223230926471475, + "grad_norm": 1.7356850427634414, + "learning_rate": 2.774533258339203e-06, + "loss": 0.845, + "step": 3262 + }, + { + "epoch": 0.3923525521553538, + "grad_norm": 2.8212736821902507, + "learning_rate": 2.7738150128614014e-06, + "loss": 0.8109, + "step": 3263 + }, + { + "epoch": 0.3924727950459929, + "grad_norm": 2.104512353079139, + "learning_rate": 2.7730966499964777e-06, + "loss": 0.9018, + "step": 3264 + }, + { + "epoch": 0.39259303793663197, + "grad_norm": 2.497373154543523, + "learning_rate": 2.772378169853408e-06, + "loss": 0.8144, + "step": 3265 + }, + { + "epoch": 0.3927132808272711, + "grad_norm": 1.717066338996621, + "learning_rate": 2.771659572541183e-06, + "loss": 0.7473, + "step": 3266 + }, + { + "epoch": 0.3928335237179102, + "grad_norm": 4.482095065700695, + "learning_rate": 2.7709408581688143e-06, + "loss": 0.8698, + "step": 3267 + }, + { + "epoch": 0.39295376660854925, + "grad_norm": 1.6531605246175607, + "learning_rate": 2.7702220268453307e-06, + "loss": 0.8806, + "step": 3268 + }, + { + "epoch": 0.39307400949918836, + "grad_norm": 2.120310152146706, + "learning_rate": 2.7695030786797785e-06, + "loss": 0.8464, + "step": 3269 + }, + { + "epoch": 0.39319425238982747, + "grad_norm": 2.60555650371687, + "learning_rate": 2.7687840137812206e-06, + "loss": 0.7494, + "step": 3270 + }, + { + "epoch": 0.3933144952804665, + "grad_norm": 0.8101347461520908, + "learning_rate": 2.7680648322587395e-06, + "loss": 0.6213, + "step": 3271 + }, + { + "epoch": 0.39343473817110564, + "grad_norm": 1.8550726859357778, + "learning_rate": 2.7673455342214334e-06, + "loss": 0.8133, + "step": 3272 + }, + { + "epoch": 0.39355498106174475, + "grad_norm": 1.9507756255668005, + "learning_rate": 2.7666261197784198e-06, + "loss": 0.7652, + "step": 3273 + }, + { + "epoch": 0.3936752239523838, + "grad_norm": 2.253339514394706, + "learning_rate": 2.7659065890388336e-06, + "loss": 0.7705, + "step": 3274 + }, + { + "epoch": 0.3937954668430229, + "grad_norm": 2.516289854155708, + "learning_rate": 2.7651869421118266e-06, + "loss": 0.8561, + "step": 3275 + }, + { + "epoch": 0.393915709733662, + "grad_norm": 1.849791633625147, + "learning_rate": 2.76446717910657e-06, + "loss": 0.8337, + "step": 3276 + }, + { + "epoch": 0.3940359526243011, + "grad_norm": 2.5700319826978504, + "learning_rate": 2.763747300132249e-06, + "loss": 0.783, + "step": 3277 + }, + { + "epoch": 0.3941561955149402, + "grad_norm": 1.7160690900192102, + "learning_rate": 2.7630273052980704e-06, + "loss": 0.8704, + "step": 3278 + }, + { + "epoch": 0.39427643840557924, + "grad_norm": 2.8933755521140507, + "learning_rate": 2.7623071947132554e-06, + "loss": 0.6792, + "step": 3279 + }, + { + "epoch": 0.39439668129621835, + "grad_norm": 1.9861204025596597, + "learning_rate": 2.7615869684870458e-06, + "loss": 0.7896, + "step": 3280 + }, + { + "epoch": 0.39451692418685746, + "grad_norm": 1.6896991528546117, + "learning_rate": 2.7608666267286986e-06, + "loss": 0.8507, + "step": 3281 + }, + { + "epoch": 0.3946371670774965, + "grad_norm": 5.069858054557285, + "learning_rate": 2.760146169547489e-06, + "loss": 0.8723, + "step": 3282 + }, + { + "epoch": 0.39475740996813563, + "grad_norm": 1.4627109310270336, + "learning_rate": 2.75942559705271e-06, + "loss": 0.7684, + "step": 3283 + }, + { + "epoch": 0.39487765285877474, + "grad_norm": 2.853027134832805, + "learning_rate": 2.7587049093536713e-06, + "loss": 0.8951, + "step": 3284 + }, + { + "epoch": 0.3949978957494138, + "grad_norm": 1.9471904468468162, + "learning_rate": 2.757984106559701e-06, + "loss": 0.8119, + "step": 3285 + }, + { + "epoch": 0.3951181386400529, + "grad_norm": 2.5071225747331, + "learning_rate": 2.7572631887801446e-06, + "loss": 0.7208, + "step": 3286 + }, + { + "epoch": 0.395238381530692, + "grad_norm": 1.6573231657124992, + "learning_rate": 2.7565421561243654e-06, + "loss": 0.7685, + "step": 3287 + }, + { + "epoch": 0.3953586244213311, + "grad_norm": 2.172963480263775, + "learning_rate": 2.7558210087017413e-06, + "loss": 0.8311, + "step": 3288 + }, + { + "epoch": 0.3954788673119702, + "grad_norm": 1.8655172882424123, + "learning_rate": 2.7550997466216724e-06, + "loss": 0.7392, + "step": 3289 + }, + { + "epoch": 0.3955991102026093, + "grad_norm": 1.8295231224907078, + "learning_rate": 2.7543783699935714e-06, + "loss": 0.8159, + "step": 3290 + }, + { + "epoch": 0.39571935309324835, + "grad_norm": 2.5087961899896323, + "learning_rate": 2.753656878926872e-06, + "loss": 0.8659, + "step": 3291 + }, + { + "epoch": 0.39583959598388746, + "grad_norm": 2.7235655832729933, + "learning_rate": 2.752935273531023e-06, + "loss": 0.7468, + "step": 3292 + }, + { + "epoch": 0.39595983887452657, + "grad_norm": 2.0896233273067217, + "learning_rate": 2.752213553915492e-06, + "loss": 0.7896, + "step": 3293 + }, + { + "epoch": 0.3960800817651656, + "grad_norm": 0.8254139627460197, + "learning_rate": 2.751491720189762e-06, + "loss": 0.6613, + "step": 3294 + }, + { + "epoch": 0.39620032465580474, + "grad_norm": 1.9468931341168643, + "learning_rate": 2.7507697724633364e-06, + "loss": 0.9154, + "step": 3295 + }, + { + "epoch": 0.3963205675464438, + "grad_norm": 0.7882259958219482, + "learning_rate": 2.7500477108457327e-06, + "loss": 0.5473, + "step": 3296 + }, + { + "epoch": 0.3964408104370829, + "grad_norm": 2.004272903302004, + "learning_rate": 2.7493255354464877e-06, + "loss": 0.8095, + "step": 3297 + }, + { + "epoch": 0.396561053327722, + "grad_norm": 2.035415369330793, + "learning_rate": 2.748603246375156e-06, + "loss": 0.7667, + "step": 3298 + }, + { + "epoch": 0.39668129621836107, + "grad_norm": 3.3022140011771364, + "learning_rate": 2.7478808437413055e-06, + "loss": 0.7119, + "step": 3299 + }, + { + "epoch": 0.3968015391090002, + "grad_norm": 1.8793550141784354, + "learning_rate": 2.7471583276545263e-06, + "loss": 0.6693, + "step": 3300 + }, + { + "epoch": 0.3969217819996393, + "grad_norm": 2.3263496952329628, + "learning_rate": 2.7464356982244224e-06, + "loss": 0.708, + "step": 3301 + }, + { + "epoch": 0.39704202489027834, + "grad_norm": 0.7928482866160823, + "learning_rate": 2.745712955560617e-06, + "loss": 0.6183, + "step": 3302 + }, + { + "epoch": 0.39716226778091746, + "grad_norm": 4.553533024986593, + "learning_rate": 2.7449900997727496e-06, + "loss": 0.7772, + "step": 3303 + }, + { + "epoch": 0.39728251067155657, + "grad_norm": 1.8797027105285191, + "learning_rate": 2.744267130970476e-06, + "loss": 0.8456, + "step": 3304 + }, + { + "epoch": 0.3974027535621956, + "grad_norm": 2.1045482654578977, + "learning_rate": 2.7435440492634697e-06, + "loss": 0.7753, + "step": 3305 + }, + { + "epoch": 0.39752299645283473, + "grad_norm": 2.3577672267685332, + "learning_rate": 2.7428208547614228e-06, + "loss": 0.6832, + "step": 3306 + }, + { + "epoch": 0.39764323934347384, + "grad_norm": 1.8127558297445094, + "learning_rate": 2.742097547574043e-06, + "loss": 0.7768, + "step": 3307 + }, + { + "epoch": 0.3977634822341129, + "grad_norm": 2.065534267077571, + "learning_rate": 2.7413741278110544e-06, + "loss": 0.7775, + "step": 3308 + }, + { + "epoch": 0.397883725124752, + "grad_norm": 2.11017757265178, + "learning_rate": 2.7406505955822016e-06, + "loss": 0.6914, + "step": 3309 + }, + { + "epoch": 0.39800396801539106, + "grad_norm": 2.1964697994567413, + "learning_rate": 2.7399269509972415e-06, + "loss": 0.666, + "step": 3310 + }, + { + "epoch": 0.3981242109060302, + "grad_norm": 2.8497875803706605, + "learning_rate": 2.7392031941659514e-06, + "loss": 0.8548, + "step": 3311 + }, + { + "epoch": 0.3982444537966693, + "grad_norm": 2.131854502395688, + "learning_rate": 2.7384793251981244e-06, + "loss": 0.8627, + "step": 3312 + }, + { + "epoch": 0.39836469668730834, + "grad_norm": 1.963288983209501, + "learning_rate": 2.737755344203571e-06, + "loss": 0.8132, + "step": 3313 + }, + { + "epoch": 0.39848493957794745, + "grad_norm": 1.7322792932095612, + "learning_rate": 2.7370312512921186e-06, + "loss": 0.7986, + "step": 3314 + }, + { + "epoch": 0.39860518246858656, + "grad_norm": 2.7533580044925725, + "learning_rate": 2.736307046573611e-06, + "loss": 0.7688, + "step": 3315 + }, + { + "epoch": 0.3987254253592256, + "grad_norm": 1.8462291149696313, + "learning_rate": 2.73558273015791e-06, + "loss": 0.8222, + "step": 3316 + }, + { + "epoch": 0.3988456682498647, + "grad_norm": 2.3362500035922245, + "learning_rate": 2.734858302154894e-06, + "loss": 0.7022, + "step": 3317 + }, + { + "epoch": 0.39896591114050384, + "grad_norm": 2.0631990115189596, + "learning_rate": 2.734133762674457e-06, + "loss": 0.7673, + "step": 3318 + }, + { + "epoch": 0.3990861540311429, + "grad_norm": 1.7701672890913498, + "learning_rate": 2.7334091118265124e-06, + "loss": 0.71, + "step": 3319 + }, + { + "epoch": 0.399206396921782, + "grad_norm": 0.6834470973062658, + "learning_rate": 2.732684349720989e-06, + "loss": 0.5794, + "step": 3320 + }, + { + "epoch": 0.3993266398124211, + "grad_norm": 1.6106408335359315, + "learning_rate": 2.7319594764678318e-06, + "loss": 0.7602, + "step": 3321 + }, + { + "epoch": 0.39944688270306017, + "grad_norm": 1.6971813396074442, + "learning_rate": 2.7312344921770044e-06, + "loss": 0.8404, + "step": 3322 + }, + { + "epoch": 0.3995671255936993, + "grad_norm": 2.000817370284194, + "learning_rate": 2.7305093969584857e-06, + "loss": 0.7879, + "step": 3323 + }, + { + "epoch": 0.3996873684843384, + "grad_norm": 1.7144242383983503, + "learning_rate": 2.729784190922272e-06, + "loss": 0.8044, + "step": 3324 + }, + { + "epoch": 0.39980761137497745, + "grad_norm": 0.9610900757777832, + "learning_rate": 2.729058874178378e-06, + "loss": 0.5721, + "step": 3325 + }, + { + "epoch": 0.39992785426561656, + "grad_norm": 2.0079028632465374, + "learning_rate": 2.7283334468368315e-06, + "loss": 0.6967, + "step": 3326 + }, + { + "epoch": 0.4000480971562556, + "grad_norm": 1.9935352628338399, + "learning_rate": 2.72760790900768e-06, + "loss": 0.7375, + "step": 3327 + }, + { + "epoch": 0.4001683400468947, + "grad_norm": 1.9209476497612237, + "learning_rate": 2.7268822608009875e-06, + "loss": 0.7881, + "step": 3328 + }, + { + "epoch": 0.40028858293753383, + "grad_norm": 1.9550154425359916, + "learning_rate": 2.726156502326834e-06, + "loss": 0.7846, + "step": 3329 + }, + { + "epoch": 0.4004088258281729, + "grad_norm": 0.7033913849501272, + "learning_rate": 2.725430633695316e-06, + "loss": 0.6029, + "step": 3330 + }, + { + "epoch": 0.400529068718812, + "grad_norm": 0.912936536020504, + "learning_rate": 2.7247046550165485e-06, + "loss": 0.5794, + "step": 3331 + }, + { + "epoch": 0.4006493116094511, + "grad_norm": 1.5211345774888003, + "learning_rate": 2.7239785664006606e-06, + "loss": 0.766, + "step": 3332 + }, + { + "epoch": 0.40076955450009016, + "grad_norm": 0.7835212879392762, + "learning_rate": 2.7232523679578002e-06, + "loss": 0.6185, + "step": 3333 + }, + { + "epoch": 0.4008897973907293, + "grad_norm": 2.584917679883381, + "learning_rate": 2.7225260597981295e-06, + "loss": 0.7965, + "step": 3334 + }, + { + "epoch": 0.4010100402813684, + "grad_norm": 2.5048328416981946, + "learning_rate": 2.721799642031831e-06, + "loss": 0.7863, + "step": 3335 + }, + { + "epoch": 0.40113028317200744, + "grad_norm": 1.971929737425566, + "learning_rate": 2.721073114769101e-06, + "loss": 0.7799, + "step": 3336 + }, + { + "epoch": 0.40125052606264655, + "grad_norm": 2.024864135705506, + "learning_rate": 2.7203464781201523e-06, + "loss": 0.7528, + "step": 3337 + }, + { + "epoch": 0.40137076895328566, + "grad_norm": 2.131870383081774, + "learning_rate": 2.719619732195215e-06, + "loss": 0.7874, + "step": 3338 + }, + { + "epoch": 0.4014910118439247, + "grad_norm": 1.4460107714619839, + "learning_rate": 2.7188928771045377e-06, + "loss": 0.7296, + "step": 3339 + }, + { + "epoch": 0.4016112547345638, + "grad_norm": 1.8023747408831179, + "learning_rate": 2.7181659129583815e-06, + "loss": 0.8025, + "step": 3340 + }, + { + "epoch": 0.4017314976252029, + "grad_norm": 2.4227926177143035, + "learning_rate": 2.7174388398670276e-06, + "loss": 0.76, + "step": 3341 + }, + { + "epoch": 0.401851740515842, + "grad_norm": 1.9394646189983868, + "learning_rate": 2.716711657940773e-06, + "loss": 0.9284, + "step": 3342 + }, + { + "epoch": 0.4019719834064811, + "grad_norm": 0.8608627273019323, + "learning_rate": 2.7159843672899284e-06, + "loss": 0.5661, + "step": 3343 + }, + { + "epoch": 0.40209222629712016, + "grad_norm": 1.9421892333387403, + "learning_rate": 2.715256968024825e-06, + "loss": 0.8188, + "step": 3344 + }, + { + "epoch": 0.40221246918775927, + "grad_norm": 1.5454000511863546, + "learning_rate": 2.7145294602558083e-06, + "loss": 0.8269, + "step": 3345 + }, + { + "epoch": 0.4023327120783984, + "grad_norm": 2.260497189917249, + "learning_rate": 2.713801844093241e-06, + "loss": 0.7084, + "step": 3346 + }, + { + "epoch": 0.40245295496903744, + "grad_norm": 1.8670972219692794, + "learning_rate": 2.7130741196475014e-06, + "loss": 0.8921, + "step": 3347 + }, + { + "epoch": 0.40257319785967655, + "grad_norm": 1.9569823359850604, + "learning_rate": 2.7123462870289848e-06, + "loss": 0.7995, + "step": 3348 + }, + { + "epoch": 0.40269344075031566, + "grad_norm": 1.565583200715253, + "learning_rate": 2.711618346348102e-06, + "loss": 0.8145, + "step": 3349 + }, + { + "epoch": 0.4028136836409547, + "grad_norm": 2.3547963563377996, + "learning_rate": 2.7108902977152825e-06, + "loss": 0.6377, + "step": 3350 + }, + { + "epoch": 0.4029339265315938, + "grad_norm": 2.4655128211742303, + "learning_rate": 2.7101621412409704e-06, + "loss": 0.7522, + "step": 3351 + }, + { + "epoch": 0.40305416942223293, + "grad_norm": 1.7716854364089287, + "learning_rate": 2.7094338770356256e-06, + "loss": 0.8614, + "step": 3352 + }, + { + "epoch": 0.403174412312872, + "grad_norm": 2.1995883990103775, + "learning_rate": 2.708705505209726e-06, + "loss": 0.6531, + "step": 3353 + }, + { + "epoch": 0.4032946552035111, + "grad_norm": 2.4060310165042287, + "learning_rate": 2.7079770258737646e-06, + "loss": 0.9213, + "step": 3354 + }, + { + "epoch": 0.4034148980941502, + "grad_norm": 2.1034458046095903, + "learning_rate": 2.707248439138251e-06, + "loss": 0.7613, + "step": 3355 + }, + { + "epoch": 0.40353514098478926, + "grad_norm": 1.8346346513806586, + "learning_rate": 2.7065197451137114e-06, + "loss": 0.6587, + "step": 3356 + }, + { + "epoch": 0.4036553838754284, + "grad_norm": 2.5723980729769864, + "learning_rate": 2.7057909439106894e-06, + "loss": 0.6796, + "step": 3357 + }, + { + "epoch": 0.40377562676606743, + "grad_norm": 1.8466816240983368, + "learning_rate": 2.7050620356397417e-06, + "loss": 0.7846, + "step": 3358 + }, + { + "epoch": 0.40389586965670654, + "grad_norm": 2.1302167026445527, + "learning_rate": 2.7043330204114437e-06, + "loss": 0.7266, + "step": 3359 + }, + { + "epoch": 0.40401611254734565, + "grad_norm": 1.9399403616006814, + "learning_rate": 2.7036038983363862e-06, + "loss": 0.8627, + "step": 3360 + }, + { + "epoch": 0.4041363554379847, + "grad_norm": 1.7437919447581451, + "learning_rate": 2.702874669525177e-06, + "loss": 0.8444, + "step": 3361 + }, + { + "epoch": 0.4042565983286238, + "grad_norm": 2.31012676221484, + "learning_rate": 2.7021453340884394e-06, + "loss": 0.702, + "step": 3362 + }, + { + "epoch": 0.40437684121926293, + "grad_norm": 2.075820677831115, + "learning_rate": 2.7014158921368125e-06, + "loss": 0.7395, + "step": 3363 + }, + { + "epoch": 0.404497084109902, + "grad_norm": 1.9717410542355418, + "learning_rate": 2.700686343780953e-06, + "loss": 0.8581, + "step": 3364 + }, + { + "epoch": 0.4046173270005411, + "grad_norm": 1.9909646264022887, + "learning_rate": 2.699956689131532e-06, + "loss": 0.8846, + "step": 3365 + }, + { + "epoch": 0.4047375698911802, + "grad_norm": 2.285346755213515, + "learning_rate": 2.699226928299238e-06, + "loss": 0.858, + "step": 3366 + }, + { + "epoch": 0.40485781278181926, + "grad_norm": 2.4044722888957626, + "learning_rate": 2.698497061394774e-06, + "loss": 0.7969, + "step": 3367 + }, + { + "epoch": 0.40497805567245837, + "grad_norm": 1.6374413982126403, + "learning_rate": 2.6977670885288627e-06, + "loss": 0.812, + "step": 3368 + }, + { + "epoch": 0.4050982985630975, + "grad_norm": 2.0295078729635168, + "learning_rate": 2.6970370098122378e-06, + "loss": 0.757, + "step": 3369 + }, + { + "epoch": 0.40521854145373654, + "grad_norm": 1.5231952181139985, + "learning_rate": 2.6963068253556535e-06, + "loss": 0.8702, + "step": 3370 + }, + { + "epoch": 0.40533878434437565, + "grad_norm": 1.8529605014088923, + "learning_rate": 2.6955765352698763e-06, + "loss": 0.8601, + "step": 3371 + }, + { + "epoch": 0.40545902723501476, + "grad_norm": 1.872782858657445, + "learning_rate": 2.6948461396656923e-06, + "loss": 0.734, + "step": 3372 + }, + { + "epoch": 0.4055792701256538, + "grad_norm": 1.983149593187554, + "learning_rate": 2.6941156386539013e-06, + "loss": 0.7491, + "step": 3373 + }, + { + "epoch": 0.4056995130162929, + "grad_norm": 2.0030256273493756, + "learning_rate": 2.6933850323453203e-06, + "loss": 0.8151, + "step": 3374 + }, + { + "epoch": 0.405819755906932, + "grad_norm": 3.0675595586912676, + "learning_rate": 2.6926543208507806e-06, + "loss": 0.7528, + "step": 3375 + }, + { + "epoch": 0.4059399987975711, + "grad_norm": 2.279074480701854, + "learning_rate": 2.6919235042811316e-06, + "loss": 0.8099, + "step": 3376 + }, + { + "epoch": 0.4060602416882102, + "grad_norm": 2.3786883575422673, + "learning_rate": 2.691192582747237e-06, + "loss": 0.7685, + "step": 3377 + }, + { + "epoch": 0.40618048457884925, + "grad_norm": 1.9618973529535129, + "learning_rate": 2.6904615563599765e-06, + "loss": 0.751, + "step": 3378 + }, + { + "epoch": 0.40630072746948837, + "grad_norm": 1.7531867303706845, + "learning_rate": 2.6897304252302477e-06, + "loss": 0.8319, + "step": 3379 + }, + { + "epoch": 0.4064209703601275, + "grad_norm": 0.8845968527591063, + "learning_rate": 2.688999189468962e-06, + "loss": 0.5485, + "step": 3380 + }, + { + "epoch": 0.40654121325076653, + "grad_norm": 3.637876332534605, + "learning_rate": 2.6882678491870464e-06, + "loss": 0.7642, + "step": 3381 + }, + { + "epoch": 0.40666145614140564, + "grad_norm": 1.762973110467775, + "learning_rate": 2.6875364044954453e-06, + "loss": 0.7169, + "step": 3382 + }, + { + "epoch": 0.40678169903204475, + "grad_norm": 1.6457120955354256, + "learning_rate": 2.6868048555051185e-06, + "loss": 0.8267, + "step": 3383 + }, + { + "epoch": 0.4069019419226838, + "grad_norm": 3.0090494362043114, + "learning_rate": 2.686073202327041e-06, + "loss": 0.869, + "step": 3384 + }, + { + "epoch": 0.4070221848133229, + "grad_norm": 1.7826356016052052, + "learning_rate": 2.6853414450722043e-06, + "loss": 0.7359, + "step": 3385 + }, + { + "epoch": 0.40714242770396203, + "grad_norm": 1.5698482909069982, + "learning_rate": 2.684609583851616e-06, + "loss": 0.8525, + "step": 3386 + }, + { + "epoch": 0.4072626705946011, + "grad_norm": 1.5543534576555242, + "learning_rate": 2.683877618776297e-06, + "loss": 0.8093, + "step": 3387 + }, + { + "epoch": 0.4073829134852402, + "grad_norm": 2.198208456572894, + "learning_rate": 2.6831455499572876e-06, + "loss": 0.7499, + "step": 3388 + }, + { + "epoch": 0.40750315637587925, + "grad_norm": 2.4203422372172017, + "learning_rate": 2.682413377505641e-06, + "loss": 0.781, + "step": 3389 + }, + { + "epoch": 0.40762339926651836, + "grad_norm": 1.898501359783868, + "learning_rate": 2.6816811015324284e-06, + "loss": 0.7742, + "step": 3390 + }, + { + "epoch": 0.40774364215715747, + "grad_norm": 0.7352902005596168, + "learning_rate": 2.6809487221487343e-06, + "loss": 0.5666, + "step": 3391 + }, + { + "epoch": 0.4078638850477965, + "grad_norm": 2.3193510484705113, + "learning_rate": 2.6802162394656605e-06, + "loss": 0.8317, + "step": 3392 + }, + { + "epoch": 0.40798412793843564, + "grad_norm": 1.70259442945582, + "learning_rate": 2.679483653594324e-06, + "loss": 0.7245, + "step": 3393 + }, + { + "epoch": 0.40810437082907475, + "grad_norm": 2.723922632006546, + "learning_rate": 2.678750964645857e-06, + "loss": 0.769, + "step": 3394 + }, + { + "epoch": 0.4082246137197138, + "grad_norm": 7.856262682871873, + "learning_rate": 2.6780181727314094e-06, + "loss": 0.8371, + "step": 3395 + }, + { + "epoch": 0.4083448566103529, + "grad_norm": 2.0324986068138986, + "learning_rate": 2.6772852779621435e-06, + "loss": 0.7851, + "step": 3396 + }, + { + "epoch": 0.408465099500992, + "grad_norm": 2.0749865474035047, + "learning_rate": 2.676552280449239e-06, + "loss": 0.8665, + "step": 3397 + }, + { + "epoch": 0.4085853423916311, + "grad_norm": 2.895192759261678, + "learning_rate": 2.6758191803038917e-06, + "loss": 0.7654, + "step": 3398 + }, + { + "epoch": 0.4087055852822702, + "grad_norm": 1.639958564188327, + "learning_rate": 2.6750859776373125e-06, + "loss": 0.8313, + "step": 3399 + }, + { + "epoch": 0.4088258281729093, + "grad_norm": 0.8000384351318963, + "learning_rate": 2.674352672560727e-06, + "loss": 0.6054, + "step": 3400 + }, + { + "epoch": 0.40894607106354836, + "grad_norm": 2.3899376572973243, + "learning_rate": 2.673619265185377e-06, + "loss": 0.7713, + "step": 3401 + }, + { + "epoch": 0.40906631395418747, + "grad_norm": 1.6500325373129114, + "learning_rate": 2.672885755622521e-06, + "loss": 0.785, + "step": 3402 + }, + { + "epoch": 0.4091865568448266, + "grad_norm": 2.302479970142832, + "learning_rate": 2.67215214398343e-06, + "loss": 0.7091, + "step": 3403 + }, + { + "epoch": 0.40930679973546563, + "grad_norm": 3.199048979822781, + "learning_rate": 2.671418430379393e-06, + "loss": 0.7827, + "step": 3404 + }, + { + "epoch": 0.40942704262610474, + "grad_norm": 1.8357216566106411, + "learning_rate": 2.670684614921715e-06, + "loss": 0.8383, + "step": 3405 + }, + { + "epoch": 0.4095472855167438, + "grad_norm": 1.976451885762003, + "learning_rate": 2.6699506977217128e-06, + "loss": 0.6974, + "step": 3406 + }, + { + "epoch": 0.4096675284073829, + "grad_norm": 2.114418724931758, + "learning_rate": 2.6692166788907233e-06, + "loss": 0.7071, + "step": 3407 + }, + { + "epoch": 0.409787771298022, + "grad_norm": 2.3611806061434355, + "learning_rate": 2.6684825585400957e-06, + "loss": 0.7718, + "step": 3408 + }, + { + "epoch": 0.4099080141886611, + "grad_norm": 0.8295692230230388, + "learning_rate": 2.6677483367811947e-06, + "loss": 0.652, + "step": 3409 + }, + { + "epoch": 0.4100282570793002, + "grad_norm": 1.622684442884068, + "learning_rate": 2.6670140137254028e-06, + "loss": 0.7584, + "step": 3410 + }, + { + "epoch": 0.4101484999699393, + "grad_norm": 2.5311354623005906, + "learning_rate": 2.666279589484115e-06, + "loss": 0.9033, + "step": 3411 + }, + { + "epoch": 0.41026874286057835, + "grad_norm": 2.241685551938436, + "learning_rate": 2.6655450641687435e-06, + "loss": 0.8178, + "step": 3412 + }, + { + "epoch": 0.41038898575121746, + "grad_norm": 1.7510357244127133, + "learning_rate": 2.664810437890715e-06, + "loss": 0.6987, + "step": 3413 + }, + { + "epoch": 0.41050922864185657, + "grad_norm": 1.9641915484559744, + "learning_rate": 2.6640757107614714e-06, + "loss": 0.7995, + "step": 3414 + }, + { + "epoch": 0.4106294715324956, + "grad_norm": 2.0099315815758425, + "learning_rate": 2.6633408828924697e-06, + "loss": 0.6986, + "step": 3415 + }, + { + "epoch": 0.41074971442313474, + "grad_norm": 2.0662827914590456, + "learning_rate": 2.662605954395185e-06, + "loss": 0.7035, + "step": 3416 + }, + { + "epoch": 0.41086995731377385, + "grad_norm": 1.700205549267359, + "learning_rate": 2.6618709253811027e-06, + "loss": 0.8397, + "step": 3417 + }, + { + "epoch": 0.4109902002044129, + "grad_norm": 1.748068629294628, + "learning_rate": 2.6611357959617277e-06, + "loss": 0.8792, + "step": 3418 + }, + { + "epoch": 0.411110443095052, + "grad_norm": 2.3640952481802127, + "learning_rate": 2.660400566248578e-06, + "loss": 0.9149, + "step": 3419 + }, + { + "epoch": 0.41123068598569107, + "grad_norm": 3.2316406781189557, + "learning_rate": 2.6596652363531876e-06, + "loss": 0.6815, + "step": 3420 + }, + { + "epoch": 0.4113509288763302, + "grad_norm": 1.6256067903724158, + "learning_rate": 2.6589298063871055e-06, + "loss": 0.7869, + "step": 3421 + }, + { + "epoch": 0.4114711717669693, + "grad_norm": 2.1491409537276187, + "learning_rate": 2.658194276461895e-06, + "loss": 0.7074, + "step": 3422 + }, + { + "epoch": 0.41159141465760835, + "grad_norm": 2.3332647131866024, + "learning_rate": 2.6574586466891368e-06, + "loss": 0.6756, + "step": 3423 + }, + { + "epoch": 0.41171165754824746, + "grad_norm": 1.9379462681851638, + "learning_rate": 2.6567229171804247e-06, + "loss": 0.6518, + "step": 3424 + }, + { + "epoch": 0.41183190043888657, + "grad_norm": 2.299973661852423, + "learning_rate": 2.655987088047368e-06, + "loss": 0.8784, + "step": 3425 + }, + { + "epoch": 0.4119521433295256, + "grad_norm": 1.9613505988399484, + "learning_rate": 2.6552511594015912e-06, + "loss": 0.788, + "step": 3426 + }, + { + "epoch": 0.41207238622016473, + "grad_norm": 2.0449873978429367, + "learning_rate": 2.654515131354735e-06, + "loss": 0.8563, + "step": 3427 + }, + { + "epoch": 0.41219262911080384, + "grad_norm": 2.044325310695964, + "learning_rate": 2.653779004018453e-06, + "loss": 0.8555, + "step": 3428 + }, + { + "epoch": 0.4123128720014429, + "grad_norm": 2.69938771758249, + "learning_rate": 2.653042777504417e-06, + "loss": 0.8223, + "step": 3429 + }, + { + "epoch": 0.412433114892082, + "grad_norm": 1.7684233778386138, + "learning_rate": 2.6523064519243105e-06, + "loss": 0.8037, + "step": 3430 + }, + { + "epoch": 0.4125533577827211, + "grad_norm": 2.176223901617364, + "learning_rate": 2.6515700273898333e-06, + "loss": 0.7953, + "step": 3431 + }, + { + "epoch": 0.4126736006733602, + "grad_norm": 3.6470868961911425, + "learning_rate": 2.6508335040127018e-06, + "loss": 0.6919, + "step": 3432 + }, + { + "epoch": 0.4127938435639993, + "grad_norm": 1.6660286476905304, + "learning_rate": 2.6500968819046446e-06, + "loss": 0.7805, + "step": 3433 + }, + { + "epoch": 0.4129140864546384, + "grad_norm": 4.714795811621972, + "learning_rate": 2.649360161177408e-06, + "loss": 0.5963, + "step": 3434 + }, + { + "epoch": 0.41303432934527745, + "grad_norm": 3.4187132280732793, + "learning_rate": 2.6486233419427504e-06, + "loss": 0.735, + "step": 3435 + }, + { + "epoch": 0.41315457223591656, + "grad_norm": 3.0169298580564674, + "learning_rate": 2.6478864243124484e-06, + "loss": 0.7613, + "step": 3436 + }, + { + "epoch": 0.4132748151265556, + "grad_norm": 2.219633544369219, + "learning_rate": 2.6471494083982903e-06, + "loss": 0.8617, + "step": 3437 + }, + { + "epoch": 0.4133950580171947, + "grad_norm": 1.8184836045061799, + "learning_rate": 2.6464122943120818e-06, + "loss": 0.7547, + "step": 3438 + }, + { + "epoch": 0.41351530090783384, + "grad_norm": 3.2982605829322833, + "learning_rate": 2.645675082165642e-06, + "loss": 0.8272, + "step": 3439 + }, + { + "epoch": 0.4136355437984729, + "grad_norm": 2.9030754224734387, + "learning_rate": 2.644937772070806e-06, + "loss": 0.7555, + "step": 3440 + }, + { + "epoch": 0.413755786689112, + "grad_norm": 2.7047440635897373, + "learning_rate": 2.6442003641394225e-06, + "loss": 0.8387, + "step": 3441 + }, + { + "epoch": 0.4138760295797511, + "grad_norm": 1.4783616181039931, + "learning_rate": 2.643462858483356e-06, + "loss": 0.8404, + "step": 3442 + }, + { + "epoch": 0.41399627247039017, + "grad_norm": 2.44060729114394, + "learning_rate": 2.6427252552144856e-06, + "loss": 0.7299, + "step": 3443 + }, + { + "epoch": 0.4141165153610293, + "grad_norm": 1.897770454422139, + "learning_rate": 2.6419875544447044e-06, + "loss": 0.7546, + "step": 3444 + }, + { + "epoch": 0.4142367582516684, + "grad_norm": 1.6867982323539361, + "learning_rate": 2.6412497562859218e-06, + "loss": 0.724, + "step": 3445 + }, + { + "epoch": 0.41435700114230745, + "grad_norm": 2.48038685580854, + "learning_rate": 2.6405118608500617e-06, + "loss": 0.7592, + "step": 3446 + }, + { + "epoch": 0.41447724403294656, + "grad_norm": 1.80482211581099, + "learning_rate": 2.6397738682490613e-06, + "loss": 0.8189, + "step": 3447 + }, + { + "epoch": 0.41459748692358567, + "grad_norm": 1.9405035917725086, + "learning_rate": 2.6390357785948734e-06, + "loss": 0.7533, + "step": 3448 + }, + { + "epoch": 0.4147177298142247, + "grad_norm": 1.7251365607566485, + "learning_rate": 2.6382975919994667e-06, + "loss": 0.8061, + "step": 3449 + }, + { + "epoch": 0.41483797270486383, + "grad_norm": 3.247865122414038, + "learning_rate": 2.637559308574822e-06, + "loss": 0.7312, + "step": 3450 + }, + { + "epoch": 0.4149582155955029, + "grad_norm": 1.8276026705017083, + "learning_rate": 2.6368209284329376e-06, + "loss": 0.711, + "step": 3451 + }, + { + "epoch": 0.415078458486142, + "grad_norm": 4.03272900323106, + "learning_rate": 2.636082451685825e-06, + "loss": 0.7658, + "step": 3452 + }, + { + "epoch": 0.4151987013767811, + "grad_norm": 1.9814981948917474, + "learning_rate": 2.6353438784455094e-06, + "loss": 0.8641, + "step": 3453 + }, + { + "epoch": 0.41531894426742016, + "grad_norm": 2.436307957969282, + "learning_rate": 2.6346052088240326e-06, + "loss": 0.7234, + "step": 3454 + }, + { + "epoch": 0.4154391871580593, + "grad_norm": 2.0369312358442797, + "learning_rate": 2.63386644293345e-06, + "loss": 0.7772, + "step": 3455 + }, + { + "epoch": 0.4155594300486984, + "grad_norm": 3.186616962263361, + "learning_rate": 2.633127580885833e-06, + "loss": 0.838, + "step": 3456 + }, + { + "epoch": 0.41567967293933744, + "grad_norm": 2.716207170144391, + "learning_rate": 2.632388622793265e-06, + "loss": 0.6559, + "step": 3457 + }, + { + "epoch": 0.41579991582997655, + "grad_norm": 1.930296149386077, + "learning_rate": 2.6316495687678457e-06, + "loss": 0.6886, + "step": 3458 + }, + { + "epoch": 0.41592015872061566, + "grad_norm": 2.5418028143600404, + "learning_rate": 2.6309104189216887e-06, + "loss": 0.7712, + "step": 3459 + }, + { + "epoch": 0.4160404016112547, + "grad_norm": 2.2142256693455655, + "learning_rate": 2.630171173366923e-06, + "loss": 0.7557, + "step": 3460 + }, + { + "epoch": 0.41616064450189383, + "grad_norm": 2.38220357149606, + "learning_rate": 2.629431832215691e-06, + "loss": 0.7507, + "step": 3461 + }, + { + "epoch": 0.41628088739253294, + "grad_norm": 2.734873611765773, + "learning_rate": 2.628692395580151e-06, + "loss": 0.8784, + "step": 3462 + }, + { + "epoch": 0.416401130283172, + "grad_norm": 1.8169039136792482, + "learning_rate": 2.6279528635724747e-06, + "loss": 0.795, + "step": 3463 + }, + { + "epoch": 0.4165213731738111, + "grad_norm": 2.3793447807523314, + "learning_rate": 2.627213236304848e-06, + "loss": 0.7896, + "step": 3464 + }, + { + "epoch": 0.4166416160644502, + "grad_norm": 2.151171688802312, + "learning_rate": 2.626473513889472e-06, + "loss": 0.7178, + "step": 3465 + }, + { + "epoch": 0.41676185895508927, + "grad_norm": 1.9021383875734033, + "learning_rate": 2.625733696438562e-06, + "loss": 0.8332, + "step": 3466 + }, + { + "epoch": 0.4168821018457284, + "grad_norm": 2.0849547625054927, + "learning_rate": 2.6249937840643476e-06, + "loss": 0.7547, + "step": 3467 + }, + { + "epoch": 0.41700234473636744, + "grad_norm": 1.9212020067860918, + "learning_rate": 2.6242537768790733e-06, + "loss": 0.6743, + "step": 3468 + }, + { + "epoch": 0.41712258762700655, + "grad_norm": 1.790652971565176, + "learning_rate": 2.6235136749949975e-06, + "loss": 0.6913, + "step": 3469 + }, + { + "epoch": 0.41724283051764566, + "grad_norm": 2.5710667949829533, + "learning_rate": 2.6227734785243924e-06, + "loss": 0.6197, + "step": 3470 + }, + { + "epoch": 0.4173630734082847, + "grad_norm": 1.807401127940412, + "learning_rate": 2.6220331875795466e-06, + "loss": 0.7956, + "step": 3471 + }, + { + "epoch": 0.4174833162989238, + "grad_norm": 1.6543048736847286, + "learning_rate": 2.62129280227276e-06, + "loss": 0.7559, + "step": 3472 + }, + { + "epoch": 0.41760355918956293, + "grad_norm": 1.9310053599444714, + "learning_rate": 2.62055232271635e-06, + "loss": 0.6893, + "step": 3473 + }, + { + "epoch": 0.417723802080202, + "grad_norm": 2.060459167638696, + "learning_rate": 2.619811749022646e-06, + "loss": 0.8828, + "step": 3474 + }, + { + "epoch": 0.4178440449708411, + "grad_norm": 2.4055039871871022, + "learning_rate": 2.6190710813039917e-06, + "loss": 0.722, + "step": 3475 + }, + { + "epoch": 0.4179642878614802, + "grad_norm": 2.1377411511117277, + "learning_rate": 2.618330319672747e-06, + "loss": 0.8423, + "step": 3476 + }, + { + "epoch": 0.41808453075211927, + "grad_norm": 2.096552553129119, + "learning_rate": 2.617589464241284e-06, + "loss": 0.9246, + "step": 3477 + }, + { + "epoch": 0.4182047736427584, + "grad_norm": 1.9852476829301038, + "learning_rate": 2.6168485151219914e-06, + "loss": 0.7473, + "step": 3478 + }, + { + "epoch": 0.4183250165333975, + "grad_norm": 2.1044872027203145, + "learning_rate": 2.616107472427269e-06, + "loss": 0.7174, + "step": 3479 + }, + { + "epoch": 0.41844525942403654, + "grad_norm": 2.4285838082743316, + "learning_rate": 2.615366336269533e-06, + "loss": 0.7693, + "step": 3480 + }, + { + "epoch": 0.41856550231467565, + "grad_norm": 2.199641293846394, + "learning_rate": 2.6146251067612126e-06, + "loss": 0.8105, + "step": 3481 + }, + { + "epoch": 0.41868574520531476, + "grad_norm": 1.8652540595329108, + "learning_rate": 2.6138837840147525e-06, + "loss": 0.8333, + "step": 3482 + }, + { + "epoch": 0.4188059880959538, + "grad_norm": 2.2964496867529762, + "learning_rate": 2.6131423681426103e-06, + "loss": 0.7749, + "step": 3483 + }, + { + "epoch": 0.41892623098659293, + "grad_norm": 1.6108697412635453, + "learning_rate": 2.6124008592572587e-06, + "loss": 0.7307, + "step": 3484 + }, + { + "epoch": 0.419046473877232, + "grad_norm": 2.0275779985695257, + "learning_rate": 2.6116592574711835e-06, + "loss": 0.8217, + "step": 3485 + }, + { + "epoch": 0.4191667167678711, + "grad_norm": 1.92397200148071, + "learning_rate": 2.6109175628968853e-06, + "loss": 0.8476, + "step": 3486 + }, + { + "epoch": 0.4192869596585102, + "grad_norm": 1.9657803496815125, + "learning_rate": 2.610175775646878e-06, + "loss": 0.8316, + "step": 3487 + }, + { + "epoch": 0.41940720254914926, + "grad_norm": 2.3175976216023217, + "learning_rate": 2.6094338958336907e-06, + "loss": 0.74, + "step": 3488 + }, + { + "epoch": 0.41952744543978837, + "grad_norm": 2.6880991068348847, + "learning_rate": 2.608691923569867e-06, + "loss": 0.8258, + "step": 3489 + }, + { + "epoch": 0.4196476883304275, + "grad_norm": 1.9772054357847166, + "learning_rate": 2.6079498589679616e-06, + "loss": 0.7645, + "step": 3490 + }, + { + "epoch": 0.41976793122106654, + "grad_norm": 1.9135148491085165, + "learning_rate": 2.6072077021405465e-06, + "loss": 0.7654, + "step": 3491 + }, + { + "epoch": 0.41988817411170565, + "grad_norm": 1.8435382398131703, + "learning_rate": 2.6064654532002054e-06, + "loss": 0.6933, + "step": 3492 + }, + { + "epoch": 0.42000841700234476, + "grad_norm": 1.4986938511859316, + "learning_rate": 2.6057231122595375e-06, + "loss": 0.7644, + "step": 3493 + }, + { + "epoch": 0.4201286598929838, + "grad_norm": 1.6998637761906041, + "learning_rate": 2.604980679431154e-06, + "loss": 0.7323, + "step": 3494 + }, + { + "epoch": 0.4202489027836229, + "grad_norm": 2.373346053993053, + "learning_rate": 2.604238154827684e-06, + "loss": 0.7487, + "step": 3495 + }, + { + "epoch": 0.42036914567426203, + "grad_norm": 2.1732868024412433, + "learning_rate": 2.6034955385617656e-06, + "loss": 0.7326, + "step": 3496 + }, + { + "epoch": 0.4204893885649011, + "grad_norm": 0.7355034052097684, + "learning_rate": 2.6027528307460544e-06, + "loss": 0.6174, + "step": 3497 + }, + { + "epoch": 0.4206096314555402, + "grad_norm": 2.0079581953689587, + "learning_rate": 2.602010031493217e-06, + "loss": 0.8686, + "step": 3498 + }, + { + "epoch": 0.42072987434617926, + "grad_norm": 1.8093814988093702, + "learning_rate": 2.6012671409159367e-06, + "loss": 0.8691, + "step": 3499 + }, + { + "epoch": 0.42085011723681837, + "grad_norm": 1.865568025316453, + "learning_rate": 2.6005241591269097e-06, + "loss": 0.8228, + "step": 3500 + }, + { + "epoch": 0.4209703601274575, + "grad_norm": 1.6823497876342128, + "learning_rate": 2.5997810862388454e-06, + "loss": 0.8005, + "step": 3501 + }, + { + "epoch": 0.42109060301809653, + "grad_norm": 2.106875146296068, + "learning_rate": 2.599037922364467e-06, + "loss": 0.7613, + "step": 3502 + }, + { + "epoch": 0.42121084590873564, + "grad_norm": 2.0961485831033757, + "learning_rate": 2.5982946676165112e-06, + "loss": 0.7573, + "step": 3503 + }, + { + "epoch": 0.42133108879937475, + "grad_norm": 0.7396172439609504, + "learning_rate": 2.5975513221077313e-06, + "loss": 0.5766, + "step": 3504 + }, + { + "epoch": 0.4214513316900138, + "grad_norm": 2.400864655982089, + "learning_rate": 2.5968078859508897e-06, + "loss": 0.8949, + "step": 3505 + }, + { + "epoch": 0.4215715745806529, + "grad_norm": 2.0604929550853948, + "learning_rate": 2.5960643592587673e-06, + "loss": 0.8021, + "step": 3506 + }, + { + "epoch": 0.42169181747129203, + "grad_norm": 1.8030286819410892, + "learning_rate": 2.5953207421441553e-06, + "loss": 0.8191, + "step": 3507 + }, + { + "epoch": 0.4218120603619311, + "grad_norm": 2.1736421051361474, + "learning_rate": 2.5945770347198603e-06, + "loss": 0.7528, + "step": 3508 + }, + { + "epoch": 0.4219323032525702, + "grad_norm": 1.8173652988350444, + "learning_rate": 2.593833237098701e-06, + "loss": 0.8222, + "step": 3509 + }, + { + "epoch": 0.4220525461432093, + "grad_norm": 2.1101787025022647, + "learning_rate": 2.593089349393512e-06, + "loss": 0.6351, + "step": 3510 + }, + { + "epoch": 0.42217278903384836, + "grad_norm": 2.5446438055586422, + "learning_rate": 2.592345371717141e-06, + "loss": 0.8417, + "step": 3511 + }, + { + "epoch": 0.42229303192448747, + "grad_norm": 2.1878710840732096, + "learning_rate": 2.591601304182448e-06, + "loss": 0.7213, + "step": 3512 + }, + { + "epoch": 0.4224132748151266, + "grad_norm": 1.9822750243869667, + "learning_rate": 2.5908571469023067e-06, + "loss": 0.7969, + "step": 3513 + }, + { + "epoch": 0.42253351770576564, + "grad_norm": 2.850517687710718, + "learning_rate": 2.5901128999896067e-06, + "loss": 0.7587, + "step": 3514 + }, + { + "epoch": 0.42265376059640475, + "grad_norm": 1.6643014655662995, + "learning_rate": 2.5893685635572487e-06, + "loss": 0.6904, + "step": 3515 + }, + { + "epoch": 0.4227740034870438, + "grad_norm": 2.293953938993731, + "learning_rate": 2.5886241377181483e-06, + "loss": 0.6999, + "step": 3516 + }, + { + "epoch": 0.4228942463776829, + "grad_norm": 1.981323358760095, + "learning_rate": 2.587879622585234e-06, + "loss": 0.8163, + "step": 3517 + }, + { + "epoch": 0.423014489268322, + "grad_norm": 1.8768894847675517, + "learning_rate": 2.5871350182714486e-06, + "loss": 0.7661, + "step": 3518 + }, + { + "epoch": 0.4231347321589611, + "grad_norm": 1.9421551064132194, + "learning_rate": 2.586390324889748e-06, + "loss": 0.8044, + "step": 3519 + }, + { + "epoch": 0.4232549750496002, + "grad_norm": 1.8987010882709574, + "learning_rate": 2.5856455425531003e-06, + "loss": 0.6756, + "step": 3520 + }, + { + "epoch": 0.4233752179402393, + "grad_norm": 1.7891343281415517, + "learning_rate": 2.5849006713744902e-06, + "loss": 0.8059, + "step": 3521 + }, + { + "epoch": 0.42349546083087836, + "grad_norm": 2.2588830013837438, + "learning_rate": 2.5841557114669135e-06, + "loss": 0.7367, + "step": 3522 + }, + { + "epoch": 0.42361570372151747, + "grad_norm": 2.6807345021483013, + "learning_rate": 2.58341066294338e-06, + "loss": 0.6802, + "step": 3523 + }, + { + "epoch": 0.4237359466121566, + "grad_norm": 2.178410966399934, + "learning_rate": 2.5826655259169124e-06, + "loss": 0.8665, + "step": 3524 + }, + { + "epoch": 0.42385618950279563, + "grad_norm": 1.911979756623207, + "learning_rate": 2.5819203005005475e-06, + "loss": 0.904, + "step": 3525 + }, + { + "epoch": 0.42397643239343474, + "grad_norm": 1.630146898893485, + "learning_rate": 2.581174986807336e-06, + "loss": 0.7899, + "step": 3526 + }, + { + "epoch": 0.42409667528407385, + "grad_norm": 2.0999125417027, + "learning_rate": 2.580429584950341e-06, + "loss": 0.9141, + "step": 3527 + }, + { + "epoch": 0.4242169181747129, + "grad_norm": 2.4969474275895838, + "learning_rate": 2.5796840950426397e-06, + "loss": 0.6723, + "step": 3528 + }, + { + "epoch": 0.424337161065352, + "grad_norm": 2.393921318596868, + "learning_rate": 2.578938517197322e-06, + "loss": 0.6625, + "step": 3529 + }, + { + "epoch": 0.4244574039559911, + "grad_norm": 2.2208389334640843, + "learning_rate": 2.5781928515274916e-06, + "loss": 0.6297, + "step": 3530 + }, + { + "epoch": 0.4245776468466302, + "grad_norm": 1.9553309685833367, + "learning_rate": 2.577447098146265e-06, + "loss": 0.6854, + "step": 3531 + }, + { + "epoch": 0.4246978897372693, + "grad_norm": 1.6646909288271163, + "learning_rate": 2.5767012571667724e-06, + "loss": 0.7913, + "step": 3532 + }, + { + "epoch": 0.42481813262790835, + "grad_norm": 1.7432250623955767, + "learning_rate": 2.5759553287021587e-06, + "loss": 0.6916, + "step": 3533 + }, + { + "epoch": 0.42493837551854746, + "grad_norm": 2.1554012010089423, + "learning_rate": 2.5752093128655786e-06, + "loss": 0.7801, + "step": 3534 + }, + { + "epoch": 0.4250586184091866, + "grad_norm": 2.189014019861543, + "learning_rate": 2.574463209770204e-06, + "loss": 0.7388, + "step": 3535 + }, + { + "epoch": 0.42517886129982563, + "grad_norm": 2.0322673890661194, + "learning_rate": 2.5737170195292165e-06, + "loss": 0.7967, + "step": 3536 + }, + { + "epoch": 0.42529910419046474, + "grad_norm": 3.289994567588333, + "learning_rate": 2.572970742255814e-06, + "loss": 0.7866, + "step": 3537 + }, + { + "epoch": 0.42541934708110385, + "grad_norm": 1.7170200311362118, + "learning_rate": 2.5722243780632046e-06, + "loss": 0.8168, + "step": 3538 + }, + { + "epoch": 0.4255395899717429, + "grad_norm": 0.7593525447586551, + "learning_rate": 2.5714779270646125e-06, + "loss": 0.6049, + "step": 3539 + }, + { + "epoch": 0.425659832862382, + "grad_norm": 2.463632131666959, + "learning_rate": 2.5707313893732735e-06, + "loss": 0.7855, + "step": 3540 + }, + { + "epoch": 0.4257800757530211, + "grad_norm": 2.748411153089101, + "learning_rate": 2.5699847651024364e-06, + "loss": 0.7753, + "step": 3541 + }, + { + "epoch": 0.4259003186436602, + "grad_norm": 2.3565255446673343, + "learning_rate": 2.5692380543653627e-06, + "loss": 0.7763, + "step": 3542 + }, + { + "epoch": 0.4260205615342993, + "grad_norm": 2.0031841679968134, + "learning_rate": 2.5684912572753293e-06, + "loss": 0.7085, + "step": 3543 + }, + { + "epoch": 0.4261408044249384, + "grad_norm": 1.745821596650919, + "learning_rate": 2.5677443739456245e-06, + "loss": 0.8407, + "step": 3544 + }, + { + "epoch": 0.42626104731557746, + "grad_norm": 3.384737449392251, + "learning_rate": 2.5669974044895495e-06, + "loss": 0.799, + "step": 3545 + }, + { + "epoch": 0.42638129020621657, + "grad_norm": 1.9552509181960993, + "learning_rate": 2.5662503490204187e-06, + "loss": 0.7939, + "step": 3546 + }, + { + "epoch": 0.4265015330968556, + "grad_norm": 1.9296377075195992, + "learning_rate": 2.5655032076515603e-06, + "loss": 0.7663, + "step": 3547 + }, + { + "epoch": 0.42662177598749473, + "grad_norm": 2.2364427295620795, + "learning_rate": 2.5647559804963155e-06, + "loss": 0.8314, + "step": 3548 + }, + { + "epoch": 0.42674201887813384, + "grad_norm": 2.0236414508324096, + "learning_rate": 2.5640086676680364e-06, + "loss": 0.7895, + "step": 3549 + }, + { + "epoch": 0.4268622617687729, + "grad_norm": 2.2029725150507646, + "learning_rate": 2.5632612692800923e-06, + "loss": 0.8153, + "step": 3550 + }, + { + "epoch": 0.426982504659412, + "grad_norm": 2.004855579873194, + "learning_rate": 2.5625137854458603e-06, + "loss": 0.7616, + "step": 3551 + }, + { + "epoch": 0.4271027475500511, + "grad_norm": 1.8504681149621511, + "learning_rate": 2.561766216278735e-06, + "loss": 0.8073, + "step": 3552 + }, + { + "epoch": 0.4272229904406902, + "grad_norm": 2.404959691873723, + "learning_rate": 2.561018561892121e-06, + "loss": 0.817, + "step": 3553 + }, + { + "epoch": 0.4273432333313293, + "grad_norm": 1.7270185463797993, + "learning_rate": 2.5602708223994363e-06, + "loss": 0.774, + "step": 3554 + }, + { + "epoch": 0.4274634762219684, + "grad_norm": 2.705336707801024, + "learning_rate": 2.559522997914115e-06, + "loss": 0.6762, + "step": 3555 + }, + { + "epoch": 0.42758371911260745, + "grad_norm": 2.0736956897726673, + "learning_rate": 2.558775088549599e-06, + "loss": 0.8532, + "step": 3556 + }, + { + "epoch": 0.42770396200324656, + "grad_norm": 2.7937038921331956, + "learning_rate": 2.5580270944193467e-06, + "loss": 0.6745, + "step": 3557 + }, + { + "epoch": 0.4278242048938857, + "grad_norm": 0.748543278301092, + "learning_rate": 2.557279015636827e-06, + "loss": 0.556, + "step": 3558 + }, + { + "epoch": 0.42794444778452473, + "grad_norm": 0.7728975511296671, + "learning_rate": 2.5565308523155245e-06, + "loss": 0.6124, + "step": 3559 + }, + { + "epoch": 0.42806469067516384, + "grad_norm": 2.5478398654328562, + "learning_rate": 2.5557826045689336e-06, + "loss": 0.838, + "step": 3560 + }, + { + "epoch": 0.4281849335658029, + "grad_norm": 0.865575477753466, + "learning_rate": 2.5550342725105643e-06, + "loss": 0.5895, + "step": 3561 + }, + { + "epoch": 0.428305176456442, + "grad_norm": 1.9204197778371035, + "learning_rate": 2.554285856253937e-06, + "loss": 0.8165, + "step": 3562 + }, + { + "epoch": 0.4284254193470811, + "grad_norm": 1.947878512520528, + "learning_rate": 2.5535373559125855e-06, + "loss": 0.7787, + "step": 3563 + }, + { + "epoch": 0.42854566223772017, + "grad_norm": 2.080810876291516, + "learning_rate": 2.552788771600057e-06, + "loss": 0.8191, + "step": 3564 + }, + { + "epoch": 0.4286659051283593, + "grad_norm": 1.8059976289254103, + "learning_rate": 2.5520401034299118e-06, + "loss": 0.8242, + "step": 3565 + }, + { + "epoch": 0.4287861480189984, + "grad_norm": 7.569637240910847, + "learning_rate": 2.551291351515722e-06, + "loss": 0.8907, + "step": 3566 + }, + { + "epoch": 0.42890639090963745, + "grad_norm": 1.779863788557906, + "learning_rate": 2.5505425159710726e-06, + "loss": 0.8574, + "step": 3567 + }, + { + "epoch": 0.42902663380027656, + "grad_norm": 1.8393754950421284, + "learning_rate": 2.549793596909561e-06, + "loss": 0.8277, + "step": 3568 + }, + { + "epoch": 0.42914687669091567, + "grad_norm": 2.125504714267827, + "learning_rate": 2.5490445944447976e-06, + "loss": 0.6672, + "step": 3569 + }, + { + "epoch": 0.4292671195815547, + "grad_norm": 2.1670622708994336, + "learning_rate": 2.548295508690406e-06, + "loss": 0.6569, + "step": 3570 + }, + { + "epoch": 0.42938736247219383, + "grad_norm": 1.7864052307718572, + "learning_rate": 2.5475463397600217e-06, + "loss": 0.7685, + "step": 3571 + }, + { + "epoch": 0.42950760536283294, + "grad_norm": 2.425110065282677, + "learning_rate": 2.546797087767293e-06, + "loss": 0.7834, + "step": 3572 + }, + { + "epoch": 0.429627848253472, + "grad_norm": 1.9615371820609637, + "learning_rate": 2.546047752825881e-06, + "loss": 0.8723, + "step": 3573 + }, + { + "epoch": 0.4297480911441111, + "grad_norm": 4.504771775573098, + "learning_rate": 2.5452983350494595e-06, + "loss": 0.9318, + "step": 3574 + }, + { + "epoch": 0.4298683340347502, + "grad_norm": 2.4613778825339763, + "learning_rate": 2.544548834551713e-06, + "loss": 0.6595, + "step": 3575 + }, + { + "epoch": 0.4299885769253893, + "grad_norm": 2.13666585964876, + "learning_rate": 2.5437992514463424e-06, + "loss": 0.9485, + "step": 3576 + }, + { + "epoch": 0.4301088198160284, + "grad_norm": 1.975000239022156, + "learning_rate": 2.5430495858470565e-06, + "loss": 0.8854, + "step": 3577 + }, + { + "epoch": 0.43022906270666744, + "grad_norm": 2.7306792651490563, + "learning_rate": 2.54229983786758e-06, + "loss": 0.7859, + "step": 3578 + }, + { + "epoch": 0.43034930559730655, + "grad_norm": 1.8452596141656203, + "learning_rate": 2.541550007621651e-06, + "loss": 0.8535, + "step": 3579 + }, + { + "epoch": 0.43046954848794566, + "grad_norm": 1.9128493220626148, + "learning_rate": 2.5408000952230156e-06, + "loss": 0.8039, + "step": 3580 + }, + { + "epoch": 0.4305897913785847, + "grad_norm": 1.957571598463349, + "learning_rate": 2.5400501007854357e-06, + "loss": 0.9118, + "step": 3581 + }, + { + "epoch": 0.43071003426922383, + "grad_norm": 1.9735609944681487, + "learning_rate": 2.539300024422685e-06, + "loss": 0.7581, + "step": 3582 + }, + { + "epoch": 0.43083027715986294, + "grad_norm": 0.8001516332569508, + "learning_rate": 2.538549866248549e-06, + "loss": 0.6101, + "step": 3583 + }, + { + "epoch": 0.430950520050502, + "grad_norm": 1.8516726561666057, + "learning_rate": 2.5377996263768274e-06, + "loss": 0.8152, + "step": 3584 + }, + { + "epoch": 0.4310707629411411, + "grad_norm": 1.9148922114502367, + "learning_rate": 2.5370493049213293e-06, + "loss": 0.6888, + "step": 3585 + }, + { + "epoch": 0.4311910058317802, + "grad_norm": 1.940946405550331, + "learning_rate": 2.536298901995878e-06, + "loss": 0.8069, + "step": 3586 + }, + { + "epoch": 0.43131124872241927, + "grad_norm": 2.095119199434454, + "learning_rate": 2.535548417714311e-06, + "loss": 0.8039, + "step": 3587 + }, + { + "epoch": 0.4314314916130584, + "grad_norm": 1.5684322658526728, + "learning_rate": 2.534797852190474e-06, + "loss": 0.8765, + "step": 3588 + }, + { + "epoch": 0.4315517345036975, + "grad_norm": 2.7521063800533088, + "learning_rate": 2.5340472055382283e-06, + "loss": 0.8235, + "step": 3589 + }, + { + "epoch": 0.43167197739433655, + "grad_norm": 1.8754317740161217, + "learning_rate": 2.5332964778714468e-06, + "loss": 0.8129, + "step": 3590 + }, + { + "epoch": 0.43179222028497566, + "grad_norm": 1.8584856102083331, + "learning_rate": 2.5325456693040123e-06, + "loss": 0.6674, + "step": 3591 + }, + { + "epoch": 0.43191246317561477, + "grad_norm": 2.2736037150968067, + "learning_rate": 2.531794779949824e-06, + "loss": 0.759, + "step": 3592 + }, + { + "epoch": 0.4320327060662538, + "grad_norm": 2.1872415662422613, + "learning_rate": 2.5310438099227903e-06, + "loss": 0.8849, + "step": 3593 + }, + { + "epoch": 0.43215294895689293, + "grad_norm": 0.8093796011893376, + "learning_rate": 2.530292759336833e-06, + "loss": 0.5347, + "step": 3594 + }, + { + "epoch": 0.432273191847532, + "grad_norm": 2.300982895516919, + "learning_rate": 2.5295416283058855e-06, + "loss": 0.7094, + "step": 3595 + }, + { + "epoch": 0.4323934347381711, + "grad_norm": 1.628973431196678, + "learning_rate": 2.5287904169438943e-06, + "loss": 0.661, + "step": 3596 + }, + { + "epoch": 0.4325136776288102, + "grad_norm": 2.412704530258438, + "learning_rate": 2.528039125364817e-06, + "loss": 0.6529, + "step": 3597 + }, + { + "epoch": 0.43263392051944927, + "grad_norm": 2.1397312275388085, + "learning_rate": 2.5272877536826246e-06, + "loss": 0.7602, + "step": 3598 + }, + { + "epoch": 0.4327541634100884, + "grad_norm": 2.4009908696977735, + "learning_rate": 2.5265363020112986e-06, + "loss": 0.7124, + "step": 3599 + }, + { + "epoch": 0.4328744063007275, + "grad_norm": 2.0134156412811697, + "learning_rate": 2.5257847704648344e-06, + "loss": 0.8453, + "step": 3600 + }, + { + "epoch": 0.43299464919136654, + "grad_norm": 2.6129414746649826, + "learning_rate": 2.525033159157239e-06, + "loss": 0.759, + "step": 3601 + }, + { + "epoch": 0.43311489208200565, + "grad_norm": 1.9805725792748703, + "learning_rate": 2.52428146820253e-06, + "loss": 0.776, + "step": 3602 + }, + { + "epoch": 0.43323513497264476, + "grad_norm": 2.21559068575222, + "learning_rate": 2.52352969771474e-06, + "loss": 0.8236, + "step": 3603 + }, + { + "epoch": 0.4333553778632838, + "grad_norm": 1.92389702842979, + "learning_rate": 2.5227778478079106e-06, + "loss": 0.8896, + "step": 3604 + }, + { + "epoch": 0.43347562075392293, + "grad_norm": 1.8157804990484085, + "learning_rate": 2.522025918596098e-06, + "loss": 0.7731, + "step": 3605 + }, + { + "epoch": 0.43359586364456204, + "grad_norm": 1.566101501195992, + "learning_rate": 2.521273910193368e-06, + "loss": 0.6617, + "step": 3606 + }, + { + "epoch": 0.4337161065352011, + "grad_norm": 2.299494592449999, + "learning_rate": 2.5205218227138006e-06, + "loss": 0.8771, + "step": 3607 + }, + { + "epoch": 0.4338363494258402, + "grad_norm": 2.1988757121727023, + "learning_rate": 2.519769656271486e-06, + "loss": 0.8005, + "step": 3608 + }, + { + "epoch": 0.43395659231647926, + "grad_norm": 2.1116562552171585, + "learning_rate": 2.5190174109805285e-06, + "loss": 0.6831, + "step": 3609 + }, + { + "epoch": 0.43407683520711837, + "grad_norm": 1.842872410771183, + "learning_rate": 2.518265086955042e-06, + "loss": 0.6414, + "step": 3610 + }, + { + "epoch": 0.4341970780977575, + "grad_norm": 2.290545507199221, + "learning_rate": 2.5175126843091534e-06, + "loss": 0.8406, + "step": 3611 + }, + { + "epoch": 0.43431732098839654, + "grad_norm": 2.0673717413791897, + "learning_rate": 2.5167602031570034e-06, + "loss": 0.7603, + "step": 3612 + }, + { + "epoch": 0.43443756387903565, + "grad_norm": 1.6772085660375897, + "learning_rate": 2.51600764361274e-06, + "loss": 0.7408, + "step": 3613 + }, + { + "epoch": 0.43455780676967476, + "grad_norm": 2.5724419118478736, + "learning_rate": 2.5152550057905283e-06, + "loss": 0.8018, + "step": 3614 + }, + { + "epoch": 0.4346780496603138, + "grad_norm": 2.2284272533283342, + "learning_rate": 2.5145022898045415e-06, + "loss": 0.7722, + "step": 3615 + }, + { + "epoch": 0.4347982925509529, + "grad_norm": 2.3528876642144643, + "learning_rate": 2.5137494957689664e-06, + "loss": 0.9026, + "step": 3616 + }, + { + "epoch": 0.43491853544159204, + "grad_norm": 0.7731447147332066, + "learning_rate": 2.5129966237980016e-06, + "loss": 0.5741, + "step": 3617 + }, + { + "epoch": 0.4350387783322311, + "grad_norm": 1.9921148548718823, + "learning_rate": 2.512243674005857e-06, + "loss": 0.7873, + "step": 3618 + }, + { + "epoch": 0.4351590212228702, + "grad_norm": 1.9325283005037532, + "learning_rate": 2.5114906465067537e-06, + "loss": 0.8684, + "step": 3619 + }, + { + "epoch": 0.4352792641135093, + "grad_norm": 2.123846975428461, + "learning_rate": 2.5107375414149264e-06, + "loss": 0.7549, + "step": 3620 + }, + { + "epoch": 0.43539950700414837, + "grad_norm": 2.36573248232286, + "learning_rate": 2.5099843588446197e-06, + "loss": 0.7301, + "step": 3621 + }, + { + "epoch": 0.4355197498947875, + "grad_norm": 1.7462337702308932, + "learning_rate": 2.509231098910091e-06, + "loss": 0.618, + "step": 3622 + }, + { + "epoch": 0.4356399927854266, + "grad_norm": 2.065783228274914, + "learning_rate": 2.508477761725611e-06, + "loss": 0.7543, + "step": 3623 + }, + { + "epoch": 0.43576023567606564, + "grad_norm": 2.0174079520440835, + "learning_rate": 2.507724347405458e-06, + "loss": 0.8114, + "step": 3624 + }, + { + "epoch": 0.43588047856670475, + "grad_norm": 2.076437358084944, + "learning_rate": 2.5069708560639243e-06, + "loss": 0.8256, + "step": 3625 + }, + { + "epoch": 0.4360007214573438, + "grad_norm": 2.460316413131077, + "learning_rate": 2.5062172878153158e-06, + "loss": 0.6174, + "step": 3626 + }, + { + "epoch": 0.4361209643479829, + "grad_norm": 2.970625756614969, + "learning_rate": 2.505463642773947e-06, + "loss": 0.874, + "step": 3627 + }, + { + "epoch": 0.43624120723862203, + "grad_norm": 2.7202041946227338, + "learning_rate": 2.504709921054146e-06, + "loss": 0.7552, + "step": 3628 + }, + { + "epoch": 0.4363614501292611, + "grad_norm": 4.16274493988886, + "learning_rate": 2.50395612277025e-06, + "loss": 0.8374, + "step": 3629 + }, + { + "epoch": 0.4364816930199002, + "grad_norm": 2.005732189629541, + "learning_rate": 2.503202248036612e-06, + "loss": 0.7366, + "step": 3630 + }, + { + "epoch": 0.4366019359105393, + "grad_norm": 1.832701212828799, + "learning_rate": 2.5024482969675927e-06, + "loss": 0.7389, + "step": 3631 + }, + { + "epoch": 0.43672217880117836, + "grad_norm": 1.959107780182744, + "learning_rate": 2.501694269677566e-06, + "loss": 0.8479, + "step": 3632 + }, + { + "epoch": 0.4368424216918175, + "grad_norm": 1.8928974460927779, + "learning_rate": 2.500940166280918e-06, + "loss": 0.8134, + "step": 3633 + }, + { + "epoch": 0.4369626645824566, + "grad_norm": 1.9435785984626657, + "learning_rate": 2.500185986892045e-06, + "loss": 0.7994, + "step": 3634 + }, + { + "epoch": 0.43708290747309564, + "grad_norm": 2.566015576684759, + "learning_rate": 2.499431731625355e-06, + "loss": 0.778, + "step": 3635 + }, + { + "epoch": 0.43720315036373475, + "grad_norm": 2.092318942532781, + "learning_rate": 2.4986774005952686e-06, + "loss": 0.7977, + "step": 3636 + }, + { + "epoch": 0.43732339325437386, + "grad_norm": 2.860871803950274, + "learning_rate": 2.4979229939162166e-06, + "loss": 0.8513, + "step": 3637 + }, + { + "epoch": 0.4374436361450129, + "grad_norm": 1.889520495911006, + "learning_rate": 2.4971685117026433e-06, + "loss": 0.8069, + "step": 3638 + }, + { + "epoch": 0.437563879035652, + "grad_norm": 1.740083801400029, + "learning_rate": 2.4964139540690018e-06, + "loss": 0.7715, + "step": 3639 + }, + { + "epoch": 0.4376841219262911, + "grad_norm": 1.7924169551494367, + "learning_rate": 2.495659321129758e-06, + "loss": 0.7335, + "step": 3640 + }, + { + "epoch": 0.4378043648169302, + "grad_norm": 2.2477954024544062, + "learning_rate": 2.494904612999389e-06, + "loss": 0.758, + "step": 3641 + }, + { + "epoch": 0.4379246077075693, + "grad_norm": 0.7605445223058235, + "learning_rate": 2.4941498297923843e-06, + "loss": 0.5658, + "step": 3642 + }, + { + "epoch": 0.43804485059820836, + "grad_norm": 1.8731929972656003, + "learning_rate": 2.4933949716232424e-06, + "loss": 0.7027, + "step": 3643 + }, + { + "epoch": 0.43816509348884747, + "grad_norm": 2.279890099269281, + "learning_rate": 2.492640038606476e-06, + "loss": 0.7385, + "step": 3644 + }, + { + "epoch": 0.4382853363794866, + "grad_norm": 1.856896295004875, + "learning_rate": 2.491885030856608e-06, + "loss": 0.7905, + "step": 3645 + }, + { + "epoch": 0.43840557927012563, + "grad_norm": 2.0941672976829753, + "learning_rate": 2.4911299484881713e-06, + "loss": 0.8307, + "step": 3646 + }, + { + "epoch": 0.43852582216076474, + "grad_norm": 1.5905879947577195, + "learning_rate": 2.490374791615712e-06, + "loss": 0.8116, + "step": 3647 + }, + { + "epoch": 0.43864606505140386, + "grad_norm": 2.9909042116448434, + "learning_rate": 2.4896195603537867e-06, + "loss": 0.7853, + "step": 3648 + }, + { + "epoch": 0.4387663079420429, + "grad_norm": 1.9690688949745074, + "learning_rate": 2.488864254816964e-06, + "loss": 0.7404, + "step": 3649 + }, + { + "epoch": 0.438886550832682, + "grad_norm": 2.1410438517641346, + "learning_rate": 2.4881088751198218e-06, + "loss": 0.6861, + "step": 3650 + }, + { + "epoch": 0.43900679372332113, + "grad_norm": 2.4294207754648545, + "learning_rate": 2.4873534213769517e-06, + "loss": 0.6457, + "step": 3651 + }, + { + "epoch": 0.4391270366139602, + "grad_norm": 1.7856307119158648, + "learning_rate": 2.4865978937029547e-06, + "loss": 0.7215, + "step": 3652 + }, + { + "epoch": 0.4392472795045993, + "grad_norm": 1.654830900467799, + "learning_rate": 2.485842292212445e-06, + "loss": 0.6666, + "step": 3653 + }, + { + "epoch": 0.4393675223952384, + "grad_norm": 1.917130980711696, + "learning_rate": 2.485086617020045e-06, + "loss": 0.8105, + "step": 3654 + }, + { + "epoch": 0.43948776528587746, + "grad_norm": 2.4861047828731113, + "learning_rate": 2.4843308682403903e-06, + "loss": 0.8223, + "step": 3655 + }, + { + "epoch": 0.4396080081765166, + "grad_norm": 1.7634468484239003, + "learning_rate": 2.4835750459881294e-06, + "loss": 0.8292, + "step": 3656 + }, + { + "epoch": 0.43972825106715563, + "grad_norm": 2.322526793556343, + "learning_rate": 2.4828191503779177e-06, + "loss": 0.8187, + "step": 3657 + }, + { + "epoch": 0.43984849395779474, + "grad_norm": 2.111023322449571, + "learning_rate": 2.482063181524425e-06, + "loss": 0.8988, + "step": 3658 + }, + { + "epoch": 0.43996873684843385, + "grad_norm": 2.2794328532971595, + "learning_rate": 2.4813071395423307e-06, + "loss": 0.8238, + "step": 3659 + }, + { + "epoch": 0.4400889797390729, + "grad_norm": 1.9893315207487614, + "learning_rate": 2.4805510245463263e-06, + "loss": 0.6504, + "step": 3660 + }, + { + "epoch": 0.440209222629712, + "grad_norm": 1.9614660151151073, + "learning_rate": 2.4797948366511137e-06, + "loss": 0.6083, + "step": 3661 + }, + { + "epoch": 0.4403294655203511, + "grad_norm": 2.082787582122491, + "learning_rate": 2.4790385759714055e-06, + "loss": 0.7664, + "step": 3662 + }, + { + "epoch": 0.4404497084109902, + "grad_norm": 1.7435162869593321, + "learning_rate": 2.478282242621926e-06, + "loss": 0.7129, + "step": 3663 + }, + { + "epoch": 0.4405699513016293, + "grad_norm": 0.8597157583026221, + "learning_rate": 2.477525836717411e-06, + "loss": 0.5964, + "step": 3664 + }, + { + "epoch": 0.4406901941922684, + "grad_norm": 2.081413019800871, + "learning_rate": 2.476769358372606e-06, + "loss": 0.7978, + "step": 3665 + }, + { + "epoch": 0.44081043708290746, + "grad_norm": 2.0748602670706893, + "learning_rate": 2.4760128077022683e-06, + "loss": 0.7524, + "step": 3666 + }, + { + "epoch": 0.44093067997354657, + "grad_norm": 1.5541169926509115, + "learning_rate": 2.4752561848211672e-06, + "loss": 0.6864, + "step": 3667 + }, + { + "epoch": 0.4410509228641857, + "grad_norm": 1.8324872000586816, + "learning_rate": 2.4744994898440797e-06, + "loss": 0.7164, + "step": 3668 + }, + { + "epoch": 0.44117116575482473, + "grad_norm": 1.9100278097395, + "learning_rate": 2.473742722885797e-06, + "loss": 0.8423, + "step": 3669 + }, + { + "epoch": 0.44129140864546385, + "grad_norm": 3.4974196322671207, + "learning_rate": 2.4729858840611197e-06, + "loss": 0.6592, + "step": 3670 + }, + { + "epoch": 0.4414116515361029, + "grad_norm": 1.914358123993848, + "learning_rate": 2.4722289734848605e-06, + "loss": 0.7296, + "step": 3671 + }, + { + "epoch": 0.441531894426742, + "grad_norm": 2.351232866775414, + "learning_rate": 2.471471991271841e-06, + "loss": 0.7841, + "step": 3672 + }, + { + "epoch": 0.4416521373173811, + "grad_norm": 1.9614068795571387, + "learning_rate": 2.470714937536896e-06, + "loss": 0.7964, + "step": 3673 + }, + { + "epoch": 0.4417723802080202, + "grad_norm": 1.915265448775462, + "learning_rate": 2.469957812394868e-06, + "loss": 0.7113, + "step": 3674 + }, + { + "epoch": 0.4418926230986593, + "grad_norm": 2.689200018869105, + "learning_rate": 2.4692006159606148e-06, + "loss": 0.7685, + "step": 3675 + }, + { + "epoch": 0.4420128659892984, + "grad_norm": 1.8425988952100103, + "learning_rate": 2.468443348349e-06, + "loss": 0.7892, + "step": 3676 + }, + { + "epoch": 0.44213310887993745, + "grad_norm": 3.1218804620001954, + "learning_rate": 2.467686009674902e-06, + "loss": 0.8288, + "step": 3677 + }, + { + "epoch": 0.44225335177057656, + "grad_norm": 2.7403822822921233, + "learning_rate": 2.466928600053209e-06, + "loss": 0.8573, + "step": 3678 + }, + { + "epoch": 0.4423735946612157, + "grad_norm": 1.897192904182369, + "learning_rate": 2.466171119598818e-06, + "loss": 0.7225, + "step": 3679 + }, + { + "epoch": 0.44249383755185473, + "grad_norm": 2.7803449551194195, + "learning_rate": 2.465413568426639e-06, + "loss": 0.7809, + "step": 3680 + }, + { + "epoch": 0.44261408044249384, + "grad_norm": 1.5969821841703862, + "learning_rate": 2.464655946651591e-06, + "loss": 0.8106, + "step": 3681 + }, + { + "epoch": 0.44273432333313295, + "grad_norm": 2.0070997821159446, + "learning_rate": 2.4638982543886065e-06, + "loss": 0.8164, + "step": 3682 + }, + { + "epoch": 0.442854566223772, + "grad_norm": 2.436854448175561, + "learning_rate": 2.4631404917526254e-06, + "loss": 0.8747, + "step": 3683 + }, + { + "epoch": 0.4429748091144111, + "grad_norm": 1.6810095973242944, + "learning_rate": 2.4623826588586e-06, + "loss": 0.7942, + "step": 3684 + }, + { + "epoch": 0.4430950520050502, + "grad_norm": 1.6826521858210883, + "learning_rate": 2.461624755821492e-06, + "loss": 0.8302, + "step": 3685 + }, + { + "epoch": 0.4432152948956893, + "grad_norm": 3.205205176774137, + "learning_rate": 2.4608667827562763e-06, + "loss": 0.7693, + "step": 3686 + }, + { + "epoch": 0.4433355377863284, + "grad_norm": 1.9953509178436146, + "learning_rate": 2.460108739777936e-06, + "loss": 0.9064, + "step": 3687 + }, + { + "epoch": 0.44345578067696745, + "grad_norm": 1.5570979003115974, + "learning_rate": 2.4593506270014656e-06, + "loss": 0.7666, + "step": 3688 + }, + { + "epoch": 0.44357602356760656, + "grad_norm": 1.9620472476881499, + "learning_rate": 2.45859244454187e-06, + "loss": 0.8254, + "step": 3689 + }, + { + "epoch": 0.44369626645824567, + "grad_norm": 1.6985265966381877, + "learning_rate": 2.4578341925141655e-06, + "loss": 0.6701, + "step": 3690 + }, + { + "epoch": 0.4438165093488847, + "grad_norm": 1.9078909268142126, + "learning_rate": 2.457075871033378e-06, + "loss": 0.7321, + "step": 3691 + }, + { + "epoch": 0.44393675223952384, + "grad_norm": 3.580128604347815, + "learning_rate": 2.4563174802145445e-06, + "loss": 0.8908, + "step": 3692 + }, + { + "epoch": 0.44405699513016295, + "grad_norm": 0.6476931838673601, + "learning_rate": 2.455559020172712e-06, + "loss": 0.487, + "step": 3693 + }, + { + "epoch": 0.444177238020802, + "grad_norm": 1.9380148541970728, + "learning_rate": 2.4548004910229385e-06, + "loss": 0.8976, + "step": 3694 + }, + { + "epoch": 0.4442974809114411, + "grad_norm": 1.7469578595376845, + "learning_rate": 2.4540418928802913e-06, + "loss": 0.8723, + "step": 3695 + }, + { + "epoch": 0.4444177238020802, + "grad_norm": 2.1296497411592195, + "learning_rate": 2.4532832258598506e-06, + "loss": 0.6652, + "step": 3696 + }, + { + "epoch": 0.4445379666927193, + "grad_norm": 1.788763695963656, + "learning_rate": 2.4525244900767047e-06, + "loss": 0.8107, + "step": 3697 + }, + { + "epoch": 0.4446582095833584, + "grad_norm": 0.7757863300998462, + "learning_rate": 2.4517656856459536e-06, + "loss": 0.6047, + "step": 3698 + }, + { + "epoch": 0.4447784524739975, + "grad_norm": 1.822964549571088, + "learning_rate": 2.4510068126827073e-06, + "loss": 0.6833, + "step": 3699 + }, + { + "epoch": 0.44489869536463655, + "grad_norm": 2.178923004512717, + "learning_rate": 2.450247871302086e-06, + "loss": 0.8281, + "step": 3700 + }, + { + "epoch": 0.44501893825527566, + "grad_norm": 2.1089734860920046, + "learning_rate": 2.44948886161922e-06, + "loss": 0.8438, + "step": 3701 + }, + { + "epoch": 0.4451391811459148, + "grad_norm": 1.602450984972933, + "learning_rate": 2.4487297837492524e-06, + "loss": 0.8527, + "step": 3702 + }, + { + "epoch": 0.44525942403655383, + "grad_norm": 2.3654423188357425, + "learning_rate": 2.4479706378073323e-06, + "loss": 0.6205, + "step": 3703 + }, + { + "epoch": 0.44537966692719294, + "grad_norm": 2.302393712542699, + "learning_rate": 2.447211423908623e-06, + "loss": 0.8438, + "step": 3704 + }, + { + "epoch": 0.445499909817832, + "grad_norm": 2.3900247456055626, + "learning_rate": 2.4464521421682966e-06, + "loss": 0.7557, + "step": 3705 + }, + { + "epoch": 0.4456201527084711, + "grad_norm": 1.3666196679353475, + "learning_rate": 2.4456927927015345e-06, + "loss": 0.8787, + "step": 3706 + }, + { + "epoch": 0.4457403955991102, + "grad_norm": 2.852520507417827, + "learning_rate": 2.4449333756235307e-06, + "loss": 0.7689, + "step": 3707 + }, + { + "epoch": 0.4458606384897493, + "grad_norm": 2.262783610234979, + "learning_rate": 2.4441738910494876e-06, + "loss": 0.7938, + "step": 3708 + }, + { + "epoch": 0.4459808813803884, + "grad_norm": 1.8741589249241386, + "learning_rate": 2.4434143390946176e-06, + "loss": 0.8224, + "step": 3709 + }, + { + "epoch": 0.4461011242710275, + "grad_norm": 2.3992149261076454, + "learning_rate": 2.4426547198741457e-06, + "loss": 0.8599, + "step": 3710 + }, + { + "epoch": 0.44622136716166655, + "grad_norm": 2.3493454325916834, + "learning_rate": 2.441895033503305e-06, + "loss": 0.7446, + "step": 3711 + }, + { + "epoch": 0.44634161005230566, + "grad_norm": 3.0429869272964787, + "learning_rate": 2.4411352800973375e-06, + "loss": 0.824, + "step": 3712 + }, + { + "epoch": 0.44646185294294477, + "grad_norm": 3.2076463005019016, + "learning_rate": 2.4403754597715005e-06, + "loss": 0.7595, + "step": 3713 + }, + { + "epoch": 0.4465820958335838, + "grad_norm": 2.0049557084972944, + "learning_rate": 2.4396155726410553e-06, + "loss": 0.9347, + "step": 3714 + }, + { + "epoch": 0.44670233872422294, + "grad_norm": 3.3892058465733204, + "learning_rate": 2.438855618821278e-06, + "loss": 0.9193, + "step": 3715 + }, + { + "epoch": 0.44682258161486205, + "grad_norm": 1.9710056764548984, + "learning_rate": 2.4380955984274517e-06, + "loss": 0.6796, + "step": 3716 + }, + { + "epoch": 0.4469428245055011, + "grad_norm": 1.7460815975440502, + "learning_rate": 2.4373355115748716e-06, + "loss": 0.774, + "step": 3717 + }, + { + "epoch": 0.4470630673961402, + "grad_norm": 1.9139404257947035, + "learning_rate": 2.436575358378842e-06, + "loss": 0.7268, + "step": 3718 + }, + { + "epoch": 0.44718331028677927, + "grad_norm": 2.487989758915078, + "learning_rate": 2.4358151389546782e-06, + "loss": 0.8395, + "step": 3719 + }, + { + "epoch": 0.4473035531774184, + "grad_norm": 2.943739574762176, + "learning_rate": 2.4350548534177035e-06, + "loss": 0.763, + "step": 3720 + }, + { + "epoch": 0.4474237960680575, + "grad_norm": 1.6526044635389727, + "learning_rate": 2.434294501883254e-06, + "loss": 0.6737, + "step": 3721 + }, + { + "epoch": 0.44754403895869654, + "grad_norm": 1.7359130500561792, + "learning_rate": 2.433534084466674e-06, + "loss": 0.6626, + "step": 3722 + }, + { + "epoch": 0.44766428184933565, + "grad_norm": 1.7122071315073688, + "learning_rate": 2.4327736012833178e-06, + "loss": 0.7148, + "step": 3723 + }, + { + "epoch": 0.44778452473997477, + "grad_norm": 2.229512291849784, + "learning_rate": 2.4320130524485506e-06, + "loss": 0.772, + "step": 3724 + }, + { + "epoch": 0.4479047676306138, + "grad_norm": 1.612128189749859, + "learning_rate": 2.431252438077746e-06, + "loss": 0.8006, + "step": 3725 + }, + { + "epoch": 0.44802501052125293, + "grad_norm": 2.497955630785708, + "learning_rate": 2.4304917582862906e-06, + "loss": 0.7782, + "step": 3726 + }, + { + "epoch": 0.44814525341189204, + "grad_norm": 1.9408003477528661, + "learning_rate": 2.4297310131895774e-06, + "loss": 0.8826, + "step": 3727 + }, + { + "epoch": 0.4482654963025311, + "grad_norm": 3.278445107656002, + "learning_rate": 2.4289702029030113e-06, + "loss": 0.7606, + "step": 3728 + }, + { + "epoch": 0.4483857391931702, + "grad_norm": 1.8304888448617571, + "learning_rate": 2.4282093275420057e-06, + "loss": 0.8355, + "step": 3729 + }, + { + "epoch": 0.4485059820838093, + "grad_norm": 2.163873216634574, + "learning_rate": 2.4274483872219863e-06, + "loss": 0.7099, + "step": 3730 + }, + { + "epoch": 0.4486262249744484, + "grad_norm": 2.290182523158336, + "learning_rate": 2.426687382058386e-06, + "loss": 0.9388, + "step": 3731 + }, + { + "epoch": 0.4487464678650875, + "grad_norm": 3.978873010687909, + "learning_rate": 2.425926312166649e-06, + "loss": 0.5984, + "step": 3732 + }, + { + "epoch": 0.4488667107557266, + "grad_norm": 2.084982321135952, + "learning_rate": 2.42516517766223e-06, + "loss": 0.7354, + "step": 3733 + }, + { + "epoch": 0.44898695364636565, + "grad_norm": 2.0959361036622193, + "learning_rate": 2.4244039786605907e-06, + "loss": 0.6839, + "step": 3734 + }, + { + "epoch": 0.44910719653700476, + "grad_norm": 2.174374027717989, + "learning_rate": 2.4236427152772055e-06, + "loss": 0.8333, + "step": 3735 + }, + { + "epoch": 0.4492274394276438, + "grad_norm": 0.837815892725972, + "learning_rate": 2.422881387627557e-06, + "loss": 0.5736, + "step": 3736 + }, + { + "epoch": 0.4493476823182829, + "grad_norm": 1.6992507405393804, + "learning_rate": 2.422119995827139e-06, + "loss": 0.7787, + "step": 3737 + }, + { + "epoch": 0.44946792520892204, + "grad_norm": 2.5816235472901727, + "learning_rate": 2.4213585399914528e-06, + "loss": 0.7465, + "step": 3738 + }, + { + "epoch": 0.4495881680995611, + "grad_norm": 1.7234788517906392, + "learning_rate": 2.4205970202360113e-06, + "loss": 0.8573, + "step": 3739 + }, + { + "epoch": 0.4497084109902002, + "grad_norm": 1.9325605618362378, + "learning_rate": 2.4198354366763354e-06, + "loss": 0.7852, + "step": 3740 + }, + { + "epoch": 0.4498286538808393, + "grad_norm": 2.0250381560321844, + "learning_rate": 2.4190737894279587e-06, + "loss": 0.7864, + "step": 3741 + }, + { + "epoch": 0.44994889677147837, + "grad_norm": 2.464850907836774, + "learning_rate": 2.4183120786064203e-06, + "loss": 0.8152, + "step": 3742 + }, + { + "epoch": 0.4500691396621175, + "grad_norm": 2.9332832751193205, + "learning_rate": 2.417550304327273e-06, + "loss": 0.8676, + "step": 3743 + }, + { + "epoch": 0.4501893825527566, + "grad_norm": 1.6074694861664711, + "learning_rate": 2.4167884667060763e-06, + "loss": 0.7598, + "step": 3744 + }, + { + "epoch": 0.45030962544339564, + "grad_norm": 2.2280960958566904, + "learning_rate": 2.4160265658584e-06, + "loss": 0.879, + "step": 3745 + }, + { + "epoch": 0.45042986833403476, + "grad_norm": 1.9902489898414573, + "learning_rate": 2.4152646018998253e-06, + "loss": 0.6907, + "step": 3746 + }, + { + "epoch": 0.45055011122467387, + "grad_norm": 1.9503401777457807, + "learning_rate": 2.4145025749459403e-06, + "loss": 0.7256, + "step": 3747 + }, + { + "epoch": 0.4506703541153129, + "grad_norm": 2.909746425369182, + "learning_rate": 2.413740485112344e-06, + "loss": 0.7098, + "step": 3748 + }, + { + "epoch": 0.45079059700595203, + "grad_norm": 1.7967636073870954, + "learning_rate": 2.412978332514646e-06, + "loss": 0.8264, + "step": 3749 + }, + { + "epoch": 0.4509108398965911, + "grad_norm": 1.8747878908346185, + "learning_rate": 2.4122161172684623e-06, + "loss": 0.7257, + "step": 3750 + }, + { + "epoch": 0.4510310827872302, + "grad_norm": 2.140058679221833, + "learning_rate": 2.4114538394894216e-06, + "loss": 0.8493, + "step": 3751 + }, + { + "epoch": 0.4511513256778693, + "grad_norm": 1.7925723100918298, + "learning_rate": 2.410691499293161e-06, + "loss": 0.8348, + "step": 3752 + }, + { + "epoch": 0.45127156856850836, + "grad_norm": 1.7696034667244973, + "learning_rate": 2.409929096795326e-06, + "loss": 0.7432, + "step": 3753 + }, + { + "epoch": 0.4513918114591475, + "grad_norm": 1.8516384060909437, + "learning_rate": 2.409166632111573e-06, + "loss": 0.7947, + "step": 3754 + }, + { + "epoch": 0.4515120543497866, + "grad_norm": 2.0321702486491207, + "learning_rate": 2.4084041053575674e-06, + "loss": 0.8092, + "step": 3755 + }, + { + "epoch": 0.45163229724042564, + "grad_norm": 1.9460153803594933, + "learning_rate": 2.4076415166489834e-06, + "loss": 0.7277, + "step": 3756 + }, + { + "epoch": 0.45175254013106475, + "grad_norm": 1.6187675975888762, + "learning_rate": 2.406878866101506e-06, + "loss": 0.7919, + "step": 3757 + }, + { + "epoch": 0.45187278302170386, + "grad_norm": 2.6733406909958704, + "learning_rate": 2.4061161538308273e-06, + "loss": 0.7917, + "step": 3758 + }, + { + "epoch": 0.4519930259123429, + "grad_norm": 1.928386604949221, + "learning_rate": 2.4053533799526523e-06, + "loss": 0.8977, + "step": 3759 + }, + { + "epoch": 0.452113268802982, + "grad_norm": 1.7789066359930412, + "learning_rate": 2.404590544582691e-06, + "loss": 0.8658, + "step": 3760 + }, + { + "epoch": 0.45223351169362114, + "grad_norm": 1.7369242311070836, + "learning_rate": 2.403827647836666e-06, + "loss": 0.8115, + "step": 3761 + }, + { + "epoch": 0.4523537545842602, + "grad_norm": 2.054340869978665, + "learning_rate": 2.4030646898303075e-06, + "loss": 0.7023, + "step": 3762 + }, + { + "epoch": 0.4524739974748993, + "grad_norm": 2.2056982285192523, + "learning_rate": 2.4023016706793566e-06, + "loss": 0.8201, + "step": 3763 + }, + { + "epoch": 0.4525942403655384, + "grad_norm": 0.7933647031619226, + "learning_rate": 2.401538590499561e-06, + "loss": 0.5694, + "step": 3764 + }, + { + "epoch": 0.45271448325617747, + "grad_norm": 2.108418678797871, + "learning_rate": 2.400775449406682e-06, + "loss": 0.7249, + "step": 3765 + }, + { + "epoch": 0.4528347261468166, + "grad_norm": 2.8454521031830406, + "learning_rate": 2.400012247516485e-06, + "loss": 0.732, + "step": 3766 + }, + { + "epoch": 0.45295496903745563, + "grad_norm": 1.8242477643403276, + "learning_rate": 2.3992489849447484e-06, + "loss": 0.9046, + "step": 3767 + }, + { + "epoch": 0.45307521192809475, + "grad_norm": 1.6632550926893126, + "learning_rate": 2.3984856618072584e-06, + "loss": 0.7906, + "step": 3768 + }, + { + "epoch": 0.45319545481873386, + "grad_norm": 1.9920246730908016, + "learning_rate": 2.3977222782198098e-06, + "loss": 0.7447, + "step": 3769 + }, + { + "epoch": 0.4533156977093729, + "grad_norm": 1.868295316975803, + "learning_rate": 2.3969588342982077e-06, + "loss": 0.7583, + "step": 3770 + }, + { + "epoch": 0.453435940600012, + "grad_norm": 1.557933925843555, + "learning_rate": 2.396195330158267e-06, + "loss": 0.732, + "step": 3771 + }, + { + "epoch": 0.45355618349065113, + "grad_norm": 2.060749119962157, + "learning_rate": 2.3954317659158094e-06, + "loss": 0.801, + "step": 3772 + }, + { + "epoch": 0.4536764263812902, + "grad_norm": 0.9070121585699732, + "learning_rate": 2.394668141686667e-06, + "loss": 0.5695, + "step": 3773 + }, + { + "epoch": 0.4537966692719293, + "grad_norm": 1.9974515891385574, + "learning_rate": 2.3939044575866813e-06, + "loss": 0.7011, + "step": 3774 + }, + { + "epoch": 0.4539169121625684, + "grad_norm": 2.1521137758547138, + "learning_rate": 2.3931407137317024e-06, + "loss": 0.7644, + "step": 3775 + }, + { + "epoch": 0.45403715505320746, + "grad_norm": 2.138479675296278, + "learning_rate": 2.3923769102375907e-06, + "loss": 0.8539, + "step": 3776 + }, + { + "epoch": 0.4541573979438466, + "grad_norm": 1.9611126946841226, + "learning_rate": 2.391613047220213e-06, + "loss": 0.7913, + "step": 3777 + }, + { + "epoch": 0.4542776408344857, + "grad_norm": 3.020548383931774, + "learning_rate": 2.390849124795447e-06, + "loss": 0.796, + "step": 3778 + }, + { + "epoch": 0.45439788372512474, + "grad_norm": 2.0866048938191812, + "learning_rate": 2.3900851430791804e-06, + "loss": 0.8447, + "step": 3779 + }, + { + "epoch": 0.45451812661576385, + "grad_norm": 2.2022009460549734, + "learning_rate": 2.389321102187307e-06, + "loss": 0.8513, + "step": 3780 + }, + { + "epoch": 0.4546383695064029, + "grad_norm": 1.7760663147350517, + "learning_rate": 2.3885570022357326e-06, + "loss": 0.8214, + "step": 3781 + }, + { + "epoch": 0.454758612397042, + "grad_norm": 0.832648178407929, + "learning_rate": 2.38779284334037e-06, + "loss": 0.6103, + "step": 3782 + }, + { + "epoch": 0.4548788552876811, + "grad_norm": 1.97105454486408, + "learning_rate": 2.387028625617141e-06, + "loss": 0.7882, + "step": 3783 + }, + { + "epoch": 0.4549990981783202, + "grad_norm": 1.8602194271069432, + "learning_rate": 2.3862643491819766e-06, + "loss": 0.8571, + "step": 3784 + }, + { + "epoch": 0.4551193410689593, + "grad_norm": 1.9466393782823745, + "learning_rate": 2.3855000141508186e-06, + "loss": 0.845, + "step": 3785 + }, + { + "epoch": 0.4552395839595984, + "grad_norm": 2.5286026285384895, + "learning_rate": 2.3847356206396143e-06, + "loss": 0.8447, + "step": 3786 + }, + { + "epoch": 0.45535982685023746, + "grad_norm": 1.5122091596943772, + "learning_rate": 2.3839711687643227e-06, + "loss": 0.789, + "step": 3787 + }, + { + "epoch": 0.45548006974087657, + "grad_norm": 2.0269365728661475, + "learning_rate": 2.38320665864091e-06, + "loss": 0.7452, + "step": 3788 + }, + { + "epoch": 0.4556003126315157, + "grad_norm": 1.7074057301404035, + "learning_rate": 2.3824420903853516e-06, + "loss": 0.8229, + "step": 3789 + }, + { + "epoch": 0.45572055552215474, + "grad_norm": 2.7152320638994443, + "learning_rate": 2.3816774641136324e-06, + "loss": 0.8258, + "step": 3790 + }, + { + "epoch": 0.45584079841279385, + "grad_norm": 2.120221431072369, + "learning_rate": 2.380912779941745e-06, + "loss": 0.7188, + "step": 3791 + }, + { + "epoch": 0.45596104130343296, + "grad_norm": 1.9404952984158008, + "learning_rate": 2.3801480379856918e-06, + "loss": 0.8417, + "step": 3792 + }, + { + "epoch": 0.456081284194072, + "grad_norm": 1.6725539495550998, + "learning_rate": 2.379383238361484e-06, + "loss": 0.8411, + "step": 3793 + }, + { + "epoch": 0.4562015270847111, + "grad_norm": 1.997225744639005, + "learning_rate": 2.3786183811851407e-06, + "loss": 0.8033, + "step": 3794 + }, + { + "epoch": 0.45632176997535023, + "grad_norm": 1.8192348974307935, + "learning_rate": 2.3778534665726892e-06, + "loss": 0.8009, + "step": 3795 + }, + { + "epoch": 0.4564420128659893, + "grad_norm": 2.348338099525151, + "learning_rate": 2.377088494640168e-06, + "loss": 0.7278, + "step": 3796 + }, + { + "epoch": 0.4565622557566284, + "grad_norm": 2.288519055951229, + "learning_rate": 2.3763234655036216e-06, + "loss": 0.7821, + "step": 3797 + }, + { + "epoch": 0.45668249864726745, + "grad_norm": 2.1968547792982775, + "learning_rate": 2.3755583792791046e-06, + "loss": 0.8727, + "step": 3798 + }, + { + "epoch": 0.45680274153790656, + "grad_norm": 2.2430659577004155, + "learning_rate": 2.3747932360826803e-06, + "loss": 0.7566, + "step": 3799 + }, + { + "epoch": 0.4569229844285457, + "grad_norm": 1.9239863141855926, + "learning_rate": 2.3740280360304205e-06, + "loss": 0.8297, + "step": 3800 + }, + { + "epoch": 0.45704322731918473, + "grad_norm": 2.053853725198491, + "learning_rate": 2.3732627792384038e-06, + "loss": 0.6826, + "step": 3801 + }, + { + "epoch": 0.45716347020982384, + "grad_norm": 2.707103117281951, + "learning_rate": 2.3724974658227207e-06, + "loss": 0.7534, + "step": 3802 + }, + { + "epoch": 0.45728371310046295, + "grad_norm": 1.899801509929426, + "learning_rate": 2.3717320958994687e-06, + "loss": 0.7156, + "step": 3803 + }, + { + "epoch": 0.457403955991102, + "grad_norm": 2.0283921311235065, + "learning_rate": 2.3709666695847534e-06, + "loss": 0.7109, + "step": 3804 + }, + { + "epoch": 0.4575241988817411, + "grad_norm": 2.081156419821011, + "learning_rate": 2.370201186994689e-06, + "loss": 0.7077, + "step": 3805 + }, + { + "epoch": 0.45764444177238023, + "grad_norm": 2.3497088150455854, + "learning_rate": 2.369435648245399e-06, + "loss": 0.7025, + "step": 3806 + }, + { + "epoch": 0.4577646846630193, + "grad_norm": 1.8077767753533167, + "learning_rate": 2.368670053453015e-06, + "loss": 0.8534, + "step": 3807 + }, + { + "epoch": 0.4578849275536584, + "grad_norm": 2.457653458112457, + "learning_rate": 2.3679044027336757e-06, + "loss": 0.7448, + "step": 3808 + }, + { + "epoch": 0.4580051704442975, + "grad_norm": 2.650533741156032, + "learning_rate": 2.3671386962035326e-06, + "loss": 0.6978, + "step": 3809 + }, + { + "epoch": 0.45812541333493656, + "grad_norm": 2.153109045758539, + "learning_rate": 2.3663729339787405e-06, + "loss": 0.6917, + "step": 3810 + }, + { + "epoch": 0.45824565622557567, + "grad_norm": 2.538943713417279, + "learning_rate": 2.365607116175466e-06, + "loss": 0.7402, + "step": 3811 + }, + { + "epoch": 0.4583658991162148, + "grad_norm": 3.9331348988753825, + "learning_rate": 2.3648412429098825e-06, + "loss": 0.6766, + "step": 3812 + }, + { + "epoch": 0.45848614200685384, + "grad_norm": 1.8128386214067136, + "learning_rate": 2.364075314298172e-06, + "loss": 0.821, + "step": 3813 + }, + { + "epoch": 0.45860638489749295, + "grad_norm": 1.9535899715086518, + "learning_rate": 2.3633093304565267e-06, + "loss": 0.7102, + "step": 3814 + }, + { + "epoch": 0.458726627788132, + "grad_norm": 2.170650329002707, + "learning_rate": 2.3625432915011443e-06, + "loss": 0.6321, + "step": 3815 + }, + { + "epoch": 0.4588468706787711, + "grad_norm": 1.9881368108720554, + "learning_rate": 2.3617771975482334e-06, + "loss": 0.6575, + "step": 3816 + }, + { + "epoch": 0.4589671135694102, + "grad_norm": 1.7074930714233485, + "learning_rate": 2.3610110487140083e-06, + "loss": 0.7519, + "step": 3817 + }, + { + "epoch": 0.4590873564600493, + "grad_norm": 1.7705478064032956, + "learning_rate": 2.360244845114695e-06, + "loss": 0.8092, + "step": 3818 + }, + { + "epoch": 0.4592075993506884, + "grad_norm": 2.233884210637253, + "learning_rate": 2.3594785868665245e-06, + "loss": 0.6897, + "step": 3819 + }, + { + "epoch": 0.4593278422413275, + "grad_norm": 2.0621202292629066, + "learning_rate": 2.3587122740857386e-06, + "loss": 0.8109, + "step": 3820 + }, + { + "epoch": 0.45944808513196655, + "grad_norm": 1.5649992981199208, + "learning_rate": 2.357945906888586e-06, + "loss": 0.7882, + "step": 3821 + }, + { + "epoch": 0.45956832802260567, + "grad_norm": 3.8676660106811473, + "learning_rate": 2.357179485391324e-06, + "loss": 0.805, + "step": 3822 + }, + { + "epoch": 0.4596885709132448, + "grad_norm": 1.9322727815951055, + "learning_rate": 2.3564130097102173e-06, + "loss": 0.8662, + "step": 3823 + }, + { + "epoch": 0.45980881380388383, + "grad_norm": 1.6947806090885347, + "learning_rate": 2.355646479961541e-06, + "loss": 0.7526, + "step": 3824 + }, + { + "epoch": 0.45992905669452294, + "grad_norm": 1.9652865993390134, + "learning_rate": 2.354879896261576e-06, + "loss": 0.7259, + "step": 3825 + }, + { + "epoch": 0.46004929958516205, + "grad_norm": 2.247876118555593, + "learning_rate": 2.3541132587266133e-06, + "loss": 0.5737, + "step": 3826 + }, + { + "epoch": 0.4601695424758011, + "grad_norm": 2.086714239197729, + "learning_rate": 2.3533465674729515e-06, + "loss": 0.6942, + "step": 3827 + }, + { + "epoch": 0.4602897853664402, + "grad_norm": 2.425870982797107, + "learning_rate": 2.352579822616895e-06, + "loss": 0.7355, + "step": 3828 + }, + { + "epoch": 0.4604100282570793, + "grad_norm": 1.9710968198990475, + "learning_rate": 2.351813024274761e-06, + "loss": 0.782, + "step": 3829 + }, + { + "epoch": 0.4605302711477184, + "grad_norm": 2.8677901380468738, + "learning_rate": 2.3510461725628693e-06, + "loss": 0.7414, + "step": 3830 + }, + { + "epoch": 0.4606505140383575, + "grad_norm": 1.7484476152251875, + "learning_rate": 2.350279267597554e-06, + "loss": 0.7184, + "step": 3831 + }, + { + "epoch": 0.46077075692899655, + "grad_norm": 2.845382892339116, + "learning_rate": 2.3495123094951515e-06, + "loss": 0.839, + "step": 3832 + }, + { + "epoch": 0.46089099981963566, + "grad_norm": 2.040692142240734, + "learning_rate": 2.34874529837201e-06, + "loss": 0.7698, + "step": 3833 + }, + { + "epoch": 0.46101124271027477, + "grad_norm": 1.9844016726646148, + "learning_rate": 2.347978234344483e-06, + "loss": 0.8001, + "step": 3834 + }, + { + "epoch": 0.4611314856009138, + "grad_norm": 1.678862542124432, + "learning_rate": 2.347211117528935e-06, + "loss": 0.6927, + "step": 3835 + }, + { + "epoch": 0.46125172849155294, + "grad_norm": 1.6983283856250893, + "learning_rate": 2.3464439480417374e-06, + "loss": 0.7211, + "step": 3836 + }, + { + "epoch": 0.46137197138219205, + "grad_norm": 3.619556285788861, + "learning_rate": 2.3456767259992676e-06, + "loss": 0.774, + "step": 3837 + }, + { + "epoch": 0.4614922142728311, + "grad_norm": 2.2854098709084227, + "learning_rate": 2.3449094515179135e-06, + "loss": 0.8926, + "step": 3838 + }, + { + "epoch": 0.4616124571634702, + "grad_norm": 1.6801197294304824, + "learning_rate": 2.34414212471407e-06, + "loss": 0.8196, + "step": 3839 + }, + { + "epoch": 0.4617327000541093, + "grad_norm": 1.8300078387691854, + "learning_rate": 2.3433747457041394e-06, + "loss": 0.7333, + "step": 3840 + }, + { + "epoch": 0.4618529429447484, + "grad_norm": 2.0901234662397283, + "learning_rate": 2.342607314604533e-06, + "loss": 0.8562, + "step": 3841 + }, + { + "epoch": 0.4619731858353875, + "grad_norm": 2.1853128671011666, + "learning_rate": 2.3418398315316694e-06, + "loss": 0.8437, + "step": 3842 + }, + { + "epoch": 0.4620934287260266, + "grad_norm": 4.910892889564821, + "learning_rate": 2.3410722966019755e-06, + "loss": 0.786, + "step": 3843 + }, + { + "epoch": 0.46221367161666566, + "grad_norm": 1.8903073130640653, + "learning_rate": 2.3403047099318848e-06, + "loss": 0.6596, + "step": 3844 + }, + { + "epoch": 0.46233391450730477, + "grad_norm": 2.10272003401377, + "learning_rate": 2.3395370716378405e-06, + "loss": 0.7557, + "step": 3845 + }, + { + "epoch": 0.4624541573979438, + "grad_norm": 2.145533593239509, + "learning_rate": 2.338769381836292e-06, + "loss": 0.7329, + "step": 3846 + }, + { + "epoch": 0.46257440028858293, + "grad_norm": 2.337178898752286, + "learning_rate": 2.3380016406436984e-06, + "loss": 0.7321, + "step": 3847 + }, + { + "epoch": 0.46269464317922204, + "grad_norm": 2.045303556861297, + "learning_rate": 2.337233848176524e-06, + "loss": 0.8214, + "step": 3848 + }, + { + "epoch": 0.4628148860698611, + "grad_norm": 2.194101237031301, + "learning_rate": 2.3364660045512435e-06, + "loss": 0.8338, + "step": 3849 + }, + { + "epoch": 0.4629351289605002, + "grad_norm": 0.7645989763771407, + "learning_rate": 2.335698109884337e-06, + "loss": 0.5833, + "step": 3850 + }, + { + "epoch": 0.4630553718511393, + "grad_norm": 0.7939863393771093, + "learning_rate": 2.334930164292294e-06, + "loss": 0.5992, + "step": 3851 + }, + { + "epoch": 0.4631756147417784, + "grad_norm": 2.1819669228484617, + "learning_rate": 2.334162167891612e-06, + "loss": 0.8039, + "step": 3852 + }, + { + "epoch": 0.4632958576324175, + "grad_norm": 2.399809603291386, + "learning_rate": 2.333394120798795e-06, + "loss": 0.7528, + "step": 3853 + }, + { + "epoch": 0.4634161005230566, + "grad_norm": 1.9821100600486024, + "learning_rate": 2.3326260231303545e-06, + "loss": 0.7272, + "step": 3854 + }, + { + "epoch": 0.46353634341369565, + "grad_norm": 1.6413160775938294, + "learning_rate": 2.331857875002811e-06, + "loss": 0.8682, + "step": 3855 + }, + { + "epoch": 0.46365658630433476, + "grad_norm": 1.8226417406847881, + "learning_rate": 2.3310896765326916e-06, + "loss": 0.7677, + "step": 3856 + }, + { + "epoch": 0.46377682919497387, + "grad_norm": 1.551530402428726, + "learning_rate": 2.330321427836531e-06, + "loss": 0.8468, + "step": 3857 + }, + { + "epoch": 0.4638970720856129, + "grad_norm": 1.6676640466754244, + "learning_rate": 2.3295531290308733e-06, + "loss": 0.8328, + "step": 3858 + }, + { + "epoch": 0.46401731497625204, + "grad_norm": 3.2755762364950853, + "learning_rate": 2.3287847802322678e-06, + "loss": 0.7637, + "step": 3859 + }, + { + "epoch": 0.4641375578668911, + "grad_norm": 3.2642109560954475, + "learning_rate": 2.3280163815572723e-06, + "loss": 0.8455, + "step": 3860 + }, + { + "epoch": 0.4642578007575302, + "grad_norm": 2.2085396532329558, + "learning_rate": 2.3272479331224522e-06, + "loss": 0.7719, + "step": 3861 + }, + { + "epoch": 0.4643780436481693, + "grad_norm": 1.563136695839414, + "learning_rate": 2.3264794350443817e-06, + "loss": 0.7846, + "step": 3862 + }, + { + "epoch": 0.46449828653880837, + "grad_norm": 2.4434726606494013, + "learning_rate": 2.3257108874396396e-06, + "loss": 0.7909, + "step": 3863 + }, + { + "epoch": 0.4646185294294475, + "grad_norm": 2.4538490538928373, + "learning_rate": 2.3249422904248152e-06, + "loss": 0.7416, + "step": 3864 + }, + { + "epoch": 0.4647387723200866, + "grad_norm": 1.3936304768174046, + "learning_rate": 2.324173644116504e-06, + "loss": 0.8758, + "step": 3865 + }, + { + "epoch": 0.46485901521072565, + "grad_norm": 1.8462614637882895, + "learning_rate": 2.3234049486313087e-06, + "loss": 0.8162, + "step": 3866 + }, + { + "epoch": 0.46497925810136476, + "grad_norm": 1.7637340581161292, + "learning_rate": 2.322636204085839e-06, + "loss": 0.7617, + "step": 3867 + }, + { + "epoch": 0.46509950099200387, + "grad_norm": 2.2385069515281373, + "learning_rate": 2.3218674105967143e-06, + "loss": 0.7968, + "step": 3868 + }, + { + "epoch": 0.4652197438826429, + "grad_norm": 1.9396608055149873, + "learning_rate": 2.3210985682805593e-06, + "loss": 0.8387, + "step": 3869 + }, + { + "epoch": 0.46533998677328203, + "grad_norm": 2.287510055770034, + "learning_rate": 2.320329677254007e-06, + "loss": 0.686, + "step": 3870 + }, + { + "epoch": 0.46546022966392114, + "grad_norm": 2.104694287714672, + "learning_rate": 2.319560737633697e-06, + "loss": 0.7295, + "step": 3871 + }, + { + "epoch": 0.4655804725545602, + "grad_norm": 1.6519056961492495, + "learning_rate": 2.3187917495362775e-06, + "loss": 0.6874, + "step": 3872 + }, + { + "epoch": 0.4657007154451993, + "grad_norm": 2.4439439935588427, + "learning_rate": 2.318022713078403e-06, + "loss": 0.7708, + "step": 3873 + }, + { + "epoch": 0.4658209583358384, + "grad_norm": 2.540140433747067, + "learning_rate": 2.3172536283767354e-06, + "loss": 0.8548, + "step": 3874 + }, + { + "epoch": 0.4659412012264775, + "grad_norm": 1.9608951883196601, + "learning_rate": 2.3164844955479447e-06, + "loss": 0.807, + "step": 3875 + }, + { + "epoch": 0.4660614441171166, + "grad_norm": 1.6916253562480934, + "learning_rate": 2.3157153147087082e-06, + "loss": 0.7094, + "step": 3876 + }, + { + "epoch": 0.46618168700775564, + "grad_norm": 2.1380243026451247, + "learning_rate": 2.314946085975709e-06, + "loss": 0.833, + "step": 3877 + }, + { + "epoch": 0.46630192989839475, + "grad_norm": 1.7403236306331675, + "learning_rate": 2.3141768094656393e-06, + "loss": 0.829, + "step": 3878 + }, + { + "epoch": 0.46642217278903386, + "grad_norm": 2.284192421605697, + "learning_rate": 2.3134074852951966e-06, + "loss": 0.8323, + "step": 3879 + }, + { + "epoch": 0.4665424156796729, + "grad_norm": 1.8059166359739993, + "learning_rate": 2.312638113581088e-06, + "loss": 0.7808, + "step": 3880 + }, + { + "epoch": 0.46666265857031203, + "grad_norm": 2.5971804075450624, + "learning_rate": 2.311868694440027e-06, + "loss": 0.7908, + "step": 3881 + }, + { + "epoch": 0.46678290146095114, + "grad_norm": 0.7406122456027796, + "learning_rate": 2.3110992279887323e-06, + "loss": 0.6254, + "step": 3882 + }, + { + "epoch": 0.4669031443515902, + "grad_norm": 3.114951495777249, + "learning_rate": 2.310329714343932e-06, + "loss": 0.8604, + "step": 3883 + }, + { + "epoch": 0.4670233872422293, + "grad_norm": 2.1412053122146872, + "learning_rate": 2.309560153622361e-06, + "loss": 0.8229, + "step": 3884 + }, + { + "epoch": 0.4671436301328684, + "grad_norm": 2.3009992667822434, + "learning_rate": 2.3087905459407602e-06, + "loss": 0.748, + "step": 3885 + }, + { + "epoch": 0.46726387302350747, + "grad_norm": 0.8068346062794445, + "learning_rate": 2.3080208914158795e-06, + "loss": 0.6295, + "step": 3886 + }, + { + "epoch": 0.4673841159141466, + "grad_norm": 5.046301652683185, + "learning_rate": 2.3072511901644753e-06, + "loss": 0.7227, + "step": 3887 + }, + { + "epoch": 0.4675043588047857, + "grad_norm": 1.9237805975450644, + "learning_rate": 2.306481442303309e-06, + "loss": 0.8084, + "step": 3888 + }, + { + "epoch": 0.46762460169542475, + "grad_norm": 1.9243950178016453, + "learning_rate": 2.3057116479491515e-06, + "loss": 0.7327, + "step": 3889 + }, + { + "epoch": 0.46774484458606386, + "grad_norm": 2.766350430106142, + "learning_rate": 2.30494180721878e-06, + "loss": 0.7633, + "step": 3890 + }, + { + "epoch": 0.4678650874767029, + "grad_norm": 2.0648029070772016, + "learning_rate": 2.3041719202289794e-06, + "loss": 0.9038, + "step": 3891 + }, + { + "epoch": 0.467985330367342, + "grad_norm": 2.651584252452025, + "learning_rate": 2.30340198709654e-06, + "loss": 0.8099, + "step": 3892 + }, + { + "epoch": 0.46810557325798113, + "grad_norm": 1.9339347769815436, + "learning_rate": 2.3026320079382605e-06, + "loss": 0.7478, + "step": 3893 + }, + { + "epoch": 0.4682258161486202, + "grad_norm": 3.3351309379712646, + "learning_rate": 2.3018619828709454e-06, + "loss": 0.7668, + "step": 3894 + }, + { + "epoch": 0.4683460590392593, + "grad_norm": 1.9554428385390612, + "learning_rate": 2.3010919120114084e-06, + "loss": 0.8238, + "step": 3895 + }, + { + "epoch": 0.4684663019298984, + "grad_norm": 2.6449832501809523, + "learning_rate": 2.3003217954764672e-06, + "loss": 0.6614, + "step": 3896 + }, + { + "epoch": 0.46858654482053747, + "grad_norm": 1.964481509965626, + "learning_rate": 2.299551633382949e-06, + "loss": 0.8006, + "step": 3897 + }, + { + "epoch": 0.4687067877111766, + "grad_norm": 2.8090573751878347, + "learning_rate": 2.2987814258476854e-06, + "loss": 0.8585, + "step": 3898 + }, + { + "epoch": 0.4688270306018157, + "grad_norm": 2.367480370058494, + "learning_rate": 2.2980111729875177e-06, + "loss": 0.6834, + "step": 3899 + }, + { + "epoch": 0.46894727349245474, + "grad_norm": 2.080709585416767, + "learning_rate": 2.2972408749192917e-06, + "loss": 0.8257, + "step": 3900 + }, + { + "epoch": 0.46906751638309385, + "grad_norm": 2.739685569180313, + "learning_rate": 2.296470531759861e-06, + "loss": 0.6711, + "step": 3901 + }, + { + "epoch": 0.46918775927373296, + "grad_norm": 1.8824413139446226, + "learning_rate": 2.2957001436260866e-06, + "loss": 0.7975, + "step": 3902 + }, + { + "epoch": 0.469308002164372, + "grad_norm": 1.644364735923988, + "learning_rate": 2.294929710634836e-06, + "loss": 0.7326, + "step": 3903 + }, + { + "epoch": 0.46942824505501113, + "grad_norm": 1.7592396685790357, + "learning_rate": 2.2941592329029823e-06, + "loss": 0.6171, + "step": 3904 + }, + { + "epoch": 0.46954848794565024, + "grad_norm": 5.28524322325426, + "learning_rate": 2.2933887105474067e-06, + "loss": 0.7924, + "step": 3905 + }, + { + "epoch": 0.4696687308362893, + "grad_norm": 1.6016805465737245, + "learning_rate": 2.2926181436849974e-06, + "loss": 0.8173, + "step": 3906 + }, + { + "epoch": 0.4697889737269284, + "grad_norm": 1.6731962170725243, + "learning_rate": 2.2918475324326478e-06, + "loss": 0.7296, + "step": 3907 + }, + { + "epoch": 0.46990921661756746, + "grad_norm": 2.3210960632127797, + "learning_rate": 2.2910768769072603e-06, + "loss": 0.9155, + "step": 3908 + }, + { + "epoch": 0.47002945950820657, + "grad_norm": 3.184724234914793, + "learning_rate": 2.2903061772257417e-06, + "loss": 0.7644, + "step": 3909 + }, + { + "epoch": 0.4701497023988457, + "grad_norm": 1.5392654207717056, + "learning_rate": 2.289535433505007e-06, + "loss": 0.786, + "step": 3910 + }, + { + "epoch": 0.47026994528948474, + "grad_norm": 2.196533949358861, + "learning_rate": 2.2887646458619767e-06, + "loss": 0.6378, + "step": 3911 + }, + { + "epoch": 0.47039018818012385, + "grad_norm": 2.115008143957164, + "learning_rate": 2.2879938144135797e-06, + "loss": 0.769, + "step": 3912 + }, + { + "epoch": 0.47051043107076296, + "grad_norm": 2.098468758129416, + "learning_rate": 2.2872229392767496e-06, + "loss": 0.7513, + "step": 3913 + }, + { + "epoch": 0.470630673961402, + "grad_norm": 1.5932415917984397, + "learning_rate": 2.286452020568428e-06, + "loss": 0.7492, + "step": 3914 + }, + { + "epoch": 0.4707509168520411, + "grad_norm": 2.8273488871350385, + "learning_rate": 2.2856810584055637e-06, + "loss": 0.7374, + "step": 3915 + }, + { + "epoch": 0.47087115974268023, + "grad_norm": 1.6774144605950925, + "learning_rate": 2.2849100529051085e-06, + "loss": 0.6795, + "step": 3916 + }, + { + "epoch": 0.4709914026333193, + "grad_norm": 2.1924262027652226, + "learning_rate": 2.284139004184026e-06, + "loss": 0.8062, + "step": 3917 + }, + { + "epoch": 0.4711116455239584, + "grad_norm": 2.155183500721794, + "learning_rate": 2.2833679123592814e-06, + "loss": 0.7436, + "step": 3918 + }, + { + "epoch": 0.4712318884145975, + "grad_norm": 1.7313689566276735, + "learning_rate": 2.2825967775478508e-06, + "loss": 0.64, + "step": 3919 + }, + { + "epoch": 0.47135213130523657, + "grad_norm": 2.209713007041034, + "learning_rate": 2.2818255998667135e-06, + "loss": 0.8436, + "step": 3920 + }, + { + "epoch": 0.4714723741958757, + "grad_norm": 1.5899922069622492, + "learning_rate": 2.2810543794328566e-06, + "loss": 0.7944, + "step": 3921 + }, + { + "epoch": 0.4715926170865148, + "grad_norm": 1.788747701476811, + "learning_rate": 2.2802831163632735e-06, + "loss": 0.8276, + "step": 3922 + }, + { + "epoch": 0.47171285997715384, + "grad_norm": 1.694572814741178, + "learning_rate": 2.279511810774965e-06, + "loss": 0.7481, + "step": 3923 + }, + { + "epoch": 0.47183310286779295, + "grad_norm": 1.9467317237608461, + "learning_rate": 2.2787404627849364e-06, + "loss": 0.7192, + "step": 3924 + }, + { + "epoch": 0.471953345758432, + "grad_norm": 1.9325333587215023, + "learning_rate": 2.277969072510202e-06, + "loss": 0.7946, + "step": 3925 + }, + { + "epoch": 0.4720735886490711, + "grad_norm": 1.603337044476968, + "learning_rate": 2.2771976400677803e-06, + "loss": 0.8165, + "step": 3926 + }, + { + "epoch": 0.47219383153971023, + "grad_norm": 1.7501743774728562, + "learning_rate": 2.2764261655746965e-06, + "loss": 0.7935, + "step": 3927 + }, + { + "epoch": 0.4723140744303493, + "grad_norm": 1.6350242565117235, + "learning_rate": 2.2756546491479832e-06, + "loss": 0.7607, + "step": 3928 + }, + { + "epoch": 0.4724343173209884, + "grad_norm": 2.1919795355711678, + "learning_rate": 2.274883090904679e-06, + "loss": 0.806, + "step": 3929 + }, + { + "epoch": 0.4725545602116275, + "grad_norm": 2.2306887222555605, + "learning_rate": 2.2741114909618283e-06, + "loss": 0.684, + "step": 3930 + }, + { + "epoch": 0.47267480310226656, + "grad_norm": 1.797969584240649, + "learning_rate": 2.2733398494364828e-06, + "loss": 0.7226, + "step": 3931 + }, + { + "epoch": 0.47279504599290567, + "grad_norm": 2.07162330444981, + "learning_rate": 2.272568166445699e-06, + "loss": 0.8506, + "step": 3932 + }, + { + "epoch": 0.4729152888835448, + "grad_norm": 2.1013401296734835, + "learning_rate": 2.271796442106541e-06, + "loss": 0.6514, + "step": 3933 + }, + { + "epoch": 0.47303553177418384, + "grad_norm": 0.8221303869380961, + "learning_rate": 2.271024676536079e-06, + "loss": 0.5657, + "step": 3934 + }, + { + "epoch": 0.47315577466482295, + "grad_norm": 1.9889616336104539, + "learning_rate": 2.2702528698513894e-06, + "loss": 0.7407, + "step": 3935 + }, + { + "epoch": 0.47327601755546206, + "grad_norm": 2.0769096778915612, + "learning_rate": 2.269481022169554e-06, + "loss": 0.7877, + "step": 3936 + }, + { + "epoch": 0.4733962604461011, + "grad_norm": 3.0765099330613412, + "learning_rate": 2.2687091336076614e-06, + "loss": 0.8057, + "step": 3937 + }, + { + "epoch": 0.4735165033367402, + "grad_norm": 2.223596571742992, + "learning_rate": 2.267937204282807e-06, + "loss": 0.8044, + "step": 3938 + }, + { + "epoch": 0.4736367462273793, + "grad_norm": 4.436420148100127, + "learning_rate": 2.2671652343120926e-06, + "loss": 0.7963, + "step": 3939 + }, + { + "epoch": 0.4737569891180184, + "grad_norm": 1.6402016229668699, + "learning_rate": 2.2663932238126236e-06, + "loss": 0.8115, + "step": 3940 + }, + { + "epoch": 0.4738772320086575, + "grad_norm": 1.4201860562517674, + "learning_rate": 2.265621172901515e-06, + "loss": 0.8085, + "step": 3941 + }, + { + "epoch": 0.47399747489929656, + "grad_norm": 2.1026651545736663, + "learning_rate": 2.2648490816958854e-06, + "loss": 0.7194, + "step": 3942 + }, + { + "epoch": 0.47411771778993567, + "grad_norm": 2.306086745854704, + "learning_rate": 2.264076950312861e-06, + "loss": 0.737, + "step": 3943 + }, + { + "epoch": 0.4742379606805748, + "grad_norm": 2.1268308269835816, + "learning_rate": 2.2633047788695727e-06, + "loss": 0.8307, + "step": 3944 + }, + { + "epoch": 0.47435820357121383, + "grad_norm": 1.7516439754106448, + "learning_rate": 2.262532567483159e-06, + "loss": 0.6439, + "step": 3945 + }, + { + "epoch": 0.47447844646185294, + "grad_norm": 2.0561798825816995, + "learning_rate": 2.2617603162707635e-06, + "loss": 0.8072, + "step": 3946 + }, + { + "epoch": 0.47459868935249205, + "grad_norm": 1.7214938370159183, + "learning_rate": 2.2609880253495363e-06, + "loss": 0.8322, + "step": 3947 + }, + { + "epoch": 0.4747189322431311, + "grad_norm": 1.9617427069425255, + "learning_rate": 2.260215694836633e-06, + "loss": 0.8704, + "step": 3948 + }, + { + "epoch": 0.4748391751337702, + "grad_norm": 2.03298453586024, + "learning_rate": 2.2594433248492157e-06, + "loss": 0.6486, + "step": 3949 + }, + { + "epoch": 0.47495941802440933, + "grad_norm": 1.922647398517916, + "learning_rate": 2.2586709155044527e-06, + "loss": 0.8093, + "step": 3950 + }, + { + "epoch": 0.4750796609150484, + "grad_norm": 1.8116905772793361, + "learning_rate": 2.2578984669195167e-06, + "loss": 0.7633, + "step": 3951 + }, + { + "epoch": 0.4751999038056875, + "grad_norm": 1.9709582584244183, + "learning_rate": 2.2571259792115887e-06, + "loss": 0.6844, + "step": 3952 + }, + { + "epoch": 0.4753201466963266, + "grad_norm": 1.8129997879760416, + "learning_rate": 2.2563534524978544e-06, + "loss": 0.7942, + "step": 3953 + }, + { + "epoch": 0.47544038958696566, + "grad_norm": 1.6805637646659393, + "learning_rate": 2.2555808868955052e-06, + "loss": 0.7126, + "step": 3954 + }, + { + "epoch": 0.47556063247760477, + "grad_norm": 2.215634482380923, + "learning_rate": 2.254808282521738e-06, + "loss": 0.7426, + "step": 3955 + }, + { + "epoch": 0.4756808753682438, + "grad_norm": 1.9042822891068, + "learning_rate": 2.2540356394937573e-06, + "loss": 0.8133, + "step": 3956 + }, + { + "epoch": 0.47580111825888294, + "grad_norm": 1.9724662462641451, + "learning_rate": 2.253262957928772e-06, + "loss": 0.8435, + "step": 3957 + }, + { + "epoch": 0.47592136114952205, + "grad_norm": 1.7611868112354874, + "learning_rate": 2.2524902379439976e-06, + "loss": 0.7255, + "step": 3958 + }, + { + "epoch": 0.4760416040401611, + "grad_norm": 0.7500817333826381, + "learning_rate": 2.251717479656655e-06, + "loss": 0.6374, + "step": 3959 + }, + { + "epoch": 0.4761618469308002, + "grad_norm": 1.7022428415265316, + "learning_rate": 2.2509446831839704e-06, + "loss": 0.7661, + "step": 3960 + }, + { + "epoch": 0.4762820898214393, + "grad_norm": 2.7869523130456377, + "learning_rate": 2.250171848643177e-06, + "loss": 0.8315, + "step": 3961 + }, + { + "epoch": 0.4764023327120784, + "grad_norm": 1.7561746384277215, + "learning_rate": 2.249398976151513e-06, + "loss": 0.867, + "step": 3962 + }, + { + "epoch": 0.4765225756027175, + "grad_norm": 2.348314240673343, + "learning_rate": 2.248626065826223e-06, + "loss": 0.7923, + "step": 3963 + }, + { + "epoch": 0.4766428184933566, + "grad_norm": 0.7753152882757639, + "learning_rate": 2.2478531177845564e-06, + "loss": 0.6266, + "step": 3964 + }, + { + "epoch": 0.47676306138399566, + "grad_norm": 1.8218224283946574, + "learning_rate": 2.247080132143769e-06, + "loss": 0.8513, + "step": 3965 + }, + { + "epoch": 0.47688330427463477, + "grad_norm": 2.3738854453288063, + "learning_rate": 2.246307109021121e-06, + "loss": 0.6937, + "step": 3966 + }, + { + "epoch": 0.4770035471652739, + "grad_norm": 1.842376391277206, + "learning_rate": 2.2455340485338817e-06, + "loss": 0.8285, + "step": 3967 + }, + { + "epoch": 0.47712379005591293, + "grad_norm": 2.2449824229276607, + "learning_rate": 2.244760950799322e-06, + "loss": 0.6802, + "step": 3968 + }, + { + "epoch": 0.47724403294655204, + "grad_norm": 1.887697964827741, + "learning_rate": 2.2439878159347203e-06, + "loss": 0.7309, + "step": 3969 + }, + { + "epoch": 0.4773642758371911, + "grad_norm": 0.7374585932083084, + "learning_rate": 2.2432146440573616e-06, + "loss": 0.5543, + "step": 3970 + }, + { + "epoch": 0.4774845187278302, + "grad_norm": 1.9387343381818485, + "learning_rate": 2.242441435284534e-06, + "loss": 0.668, + "step": 3971 + }, + { + "epoch": 0.4776047616184693, + "grad_norm": 2.1387668536272963, + "learning_rate": 2.2416681897335337e-06, + "loss": 0.8557, + "step": 3972 + }, + { + "epoch": 0.4777250045091084, + "grad_norm": 1.8390419543211671, + "learning_rate": 2.240894907521661e-06, + "loss": 0.6709, + "step": 3973 + }, + { + "epoch": 0.4778452473997475, + "grad_norm": 1.7774707120046456, + "learning_rate": 2.240121588766223e-06, + "loss": 0.6367, + "step": 3974 + }, + { + "epoch": 0.4779654902903866, + "grad_norm": 1.9261759572091262, + "learning_rate": 2.239348233584531e-06, + "loss": 0.7189, + "step": 3975 + }, + { + "epoch": 0.47808573318102565, + "grad_norm": 1.8374819488676755, + "learning_rate": 2.2385748420939013e-06, + "loss": 0.812, + "step": 3976 + }, + { + "epoch": 0.47820597607166476, + "grad_norm": 1.8668209977156907, + "learning_rate": 2.2378014144116583e-06, + "loss": 0.7305, + "step": 3977 + }, + { + "epoch": 0.4783262189623039, + "grad_norm": 1.9500268396340008, + "learning_rate": 2.23702795065513e-06, + "loss": 0.7986, + "step": 3978 + }, + { + "epoch": 0.47844646185294293, + "grad_norm": 0.9881443034342896, + "learning_rate": 2.2362544509416493e-06, + "loss": 0.6745, + "step": 3979 + }, + { + "epoch": 0.47856670474358204, + "grad_norm": 2.1891778233474612, + "learning_rate": 2.2354809153885572e-06, + "loss": 0.8309, + "step": 3980 + }, + { + "epoch": 0.47868694763422115, + "grad_norm": 1.9353859350463083, + "learning_rate": 2.234707344113197e-06, + "loss": 0.8365, + "step": 3981 + }, + { + "epoch": 0.4788071905248602, + "grad_norm": 1.7035654180991366, + "learning_rate": 2.233933737232919e-06, + "loss": 0.7763, + "step": 3982 + }, + { + "epoch": 0.4789274334154993, + "grad_norm": 2.6231508553943765, + "learning_rate": 2.2331600948650793e-06, + "loss": 0.7839, + "step": 3983 + }, + { + "epoch": 0.4790476763061384, + "grad_norm": 1.71745903776982, + "learning_rate": 2.2323864171270386e-06, + "loss": 0.8069, + "step": 3984 + }, + { + "epoch": 0.4791679191967775, + "grad_norm": 1.8960446949048708, + "learning_rate": 2.231612704136164e-06, + "loss": 0.7291, + "step": 3985 + }, + { + "epoch": 0.4792881620874166, + "grad_norm": 2.304277907359901, + "learning_rate": 2.2308389560098253e-06, + "loss": 0.7571, + "step": 3986 + }, + { + "epoch": 0.47940840497805565, + "grad_norm": 4.87389263389524, + "learning_rate": 2.2300651728654008e-06, + "loss": 0.7775, + "step": 3987 + }, + { + "epoch": 0.47952864786869476, + "grad_norm": 0.7443610237917044, + "learning_rate": 2.229291354820272e-06, + "loss": 0.6021, + "step": 3988 + }, + { + "epoch": 0.47964889075933387, + "grad_norm": 2.1294032085533185, + "learning_rate": 2.228517501991828e-06, + "loss": 0.7613, + "step": 3989 + }, + { + "epoch": 0.4797691336499729, + "grad_norm": 0.9991683481969569, + "learning_rate": 2.22774361449746e-06, + "loss": 0.6117, + "step": 3990 + }, + { + "epoch": 0.47988937654061203, + "grad_norm": 2.983188573070767, + "learning_rate": 2.2269696924545668e-06, + "loss": 0.71, + "step": 3991 + }, + { + "epoch": 0.48000961943125114, + "grad_norm": 2.2788238165236407, + "learning_rate": 2.2261957359805523e-06, + "loss": 0.7847, + "step": 3992 + }, + { + "epoch": 0.4801298623218902, + "grad_norm": 2.0786682795497335, + "learning_rate": 2.225421745192823e-06, + "loss": 0.7437, + "step": 3993 + }, + { + "epoch": 0.4802501052125293, + "grad_norm": 2.0316578382259047, + "learning_rate": 2.2246477202087955e-06, + "loss": 0.7851, + "step": 3994 + }, + { + "epoch": 0.4803703481031684, + "grad_norm": 1.6856920659376555, + "learning_rate": 2.223873661145887e-06, + "loss": 0.8328, + "step": 3995 + }, + { + "epoch": 0.4804905909938075, + "grad_norm": 1.7202191402869988, + "learning_rate": 2.2230995681215226e-06, + "loss": 0.7159, + "step": 3996 + }, + { + "epoch": 0.4806108338844466, + "grad_norm": 1.8456576548631867, + "learning_rate": 2.2223254412531305e-06, + "loss": 0.7836, + "step": 3997 + }, + { + "epoch": 0.4807310767750857, + "grad_norm": 1.8814971254576682, + "learning_rate": 2.221551280658146e-06, + "loss": 0.8275, + "step": 3998 + }, + { + "epoch": 0.48085131966572475, + "grad_norm": 1.7451747037335585, + "learning_rate": 2.2207770864540085e-06, + "loss": 0.741, + "step": 3999 + }, + { + "epoch": 0.48097156255636386, + "grad_norm": 1.8624088956325489, + "learning_rate": 2.220002858758162e-06, + "loss": 0.7304, + "step": 4000 + }, + { + "epoch": 0.481091805447003, + "grad_norm": 0.9036913630350664, + "learning_rate": 2.2192285976880573e-06, + "loss": 0.6091, + "step": 4001 + }, + { + "epoch": 0.48121204833764203, + "grad_norm": 2.3049551081061255, + "learning_rate": 2.2184543033611485e-06, + "loss": 0.8124, + "step": 4002 + }, + { + "epoch": 0.48133229122828114, + "grad_norm": 7.925259019728324, + "learning_rate": 2.2176799758948957e-06, + "loss": 0.8244, + "step": 4003 + }, + { + "epoch": 0.4814525341189202, + "grad_norm": 1.9092834600553854, + "learning_rate": 2.2169056154067635e-06, + "loss": 0.7399, + "step": 4004 + }, + { + "epoch": 0.4815727770095593, + "grad_norm": 1.8157285020646228, + "learning_rate": 2.216131222014222e-06, + "loss": 0.8271, + "step": 4005 + }, + { + "epoch": 0.4816930199001984, + "grad_norm": 2.8697951928940357, + "learning_rate": 2.2153567958347455e-06, + "loss": 0.8084, + "step": 4006 + }, + { + "epoch": 0.48181326279083747, + "grad_norm": 4.339876712155426, + "learning_rate": 2.214582336985815e-06, + "loss": 0.8113, + "step": 4007 + }, + { + "epoch": 0.4819335056814766, + "grad_norm": 3.8952110867969103, + "learning_rate": 2.2138078455849142e-06, + "loss": 0.6656, + "step": 4008 + }, + { + "epoch": 0.4820537485721157, + "grad_norm": 2.4542452447182055, + "learning_rate": 2.2130333217495334e-06, + "loss": 0.7964, + "step": 4009 + }, + { + "epoch": 0.48217399146275475, + "grad_norm": 3.4703843673966004, + "learning_rate": 2.2122587655971665e-06, + "loss": 0.6785, + "step": 4010 + }, + { + "epoch": 0.48229423435339386, + "grad_norm": 1.7324379078360779, + "learning_rate": 2.211484177245314e-06, + "loss": 0.6407, + "step": 4011 + }, + { + "epoch": 0.48241447724403297, + "grad_norm": 4.754234233691448, + "learning_rate": 2.21070955681148e-06, + "loss": 0.7284, + "step": 4012 + }, + { + "epoch": 0.482534720134672, + "grad_norm": 1.7382999719190175, + "learning_rate": 2.209934904413174e-06, + "loss": 0.7869, + "step": 4013 + }, + { + "epoch": 0.48265496302531113, + "grad_norm": 2.1880528432612327, + "learning_rate": 2.2091602201679095e-06, + "loss": 0.7188, + "step": 4014 + }, + { + "epoch": 0.48277520591595025, + "grad_norm": 2.498176863616093, + "learning_rate": 2.208385504193206e-06, + "loss": 0.8365, + "step": 4015 + }, + { + "epoch": 0.4828954488065893, + "grad_norm": 2.111168219256817, + "learning_rate": 2.2076107566065873e-06, + "loss": 0.8149, + "step": 4016 + }, + { + "epoch": 0.4830156916972284, + "grad_norm": 2.2545041459807487, + "learning_rate": 2.2068359775255816e-06, + "loss": 0.7638, + "step": 4017 + }, + { + "epoch": 0.48313593458786747, + "grad_norm": 2.388194325730461, + "learning_rate": 2.206061167067723e-06, + "loss": 0.7943, + "step": 4018 + }, + { + "epoch": 0.4832561774785066, + "grad_norm": 2.0365950464536993, + "learning_rate": 2.205286325350549e-06, + "loss": 0.7953, + "step": 4019 + }, + { + "epoch": 0.4833764203691457, + "grad_norm": 2.5246117283891145, + "learning_rate": 2.204511452491603e-06, + "loss": 0.736, + "step": 4020 + }, + { + "epoch": 0.48349666325978474, + "grad_norm": 1.6888747796436918, + "learning_rate": 2.2037365486084316e-06, + "loss": 0.7561, + "step": 4021 + }, + { + "epoch": 0.48361690615042385, + "grad_norm": 1.8669654362998087, + "learning_rate": 2.2029616138185886e-06, + "loss": 0.7773, + "step": 4022 + }, + { + "epoch": 0.48373714904106296, + "grad_norm": 1.923606402792679, + "learning_rate": 2.202186648239629e-06, + "loss": 0.8315, + "step": 4023 + }, + { + "epoch": 0.483857391931702, + "grad_norm": 2.5006985642058095, + "learning_rate": 2.201411651989117e-06, + "loss": 0.7172, + "step": 4024 + }, + { + "epoch": 0.48397763482234113, + "grad_norm": 2.0117865393614838, + "learning_rate": 2.2006366251846167e-06, + "loss": 0.7875, + "step": 4025 + }, + { + "epoch": 0.48409787771298024, + "grad_norm": 1.9015762065308834, + "learning_rate": 2.1998615679436997e-06, + "loss": 0.7582, + "step": 4026 + }, + { + "epoch": 0.4842181206036193, + "grad_norm": 2.606832762041971, + "learning_rate": 2.199086480383942e-06, + "loss": 0.7812, + "step": 4027 + }, + { + "epoch": 0.4843383634942584, + "grad_norm": 3.093409972337648, + "learning_rate": 2.1983113626229234e-06, + "loss": 0.6789, + "step": 4028 + }, + { + "epoch": 0.4844586063848975, + "grad_norm": 1.9158083680152234, + "learning_rate": 2.1975362147782293e-06, + "loss": 0.7866, + "step": 4029 + }, + { + "epoch": 0.48457884927553657, + "grad_norm": 0.7501630605822698, + "learning_rate": 2.196761036967448e-06, + "loss": 0.542, + "step": 4030 + }, + { + "epoch": 0.4846990921661757, + "grad_norm": 2.9194531527009353, + "learning_rate": 2.1959858293081743e-06, + "loss": 0.7824, + "step": 4031 + }, + { + "epoch": 0.4848193350568148, + "grad_norm": 1.765709393940259, + "learning_rate": 2.1952105919180056e-06, + "loss": 0.7646, + "step": 4032 + }, + { + "epoch": 0.48493957794745385, + "grad_norm": 1.9646375180077613, + "learning_rate": 2.1944353249145456e-06, + "loss": 0.6803, + "step": 4033 + }, + { + "epoch": 0.48505982083809296, + "grad_norm": 1.6682446003784621, + "learning_rate": 2.193660028415401e-06, + "loss": 0.7471, + "step": 4034 + }, + { + "epoch": 0.485180063728732, + "grad_norm": 1.8492985658620575, + "learning_rate": 2.1928847025381852e-06, + "loss": 0.8193, + "step": 4035 + }, + { + "epoch": 0.4853003066193711, + "grad_norm": 1.7403202742109953, + "learning_rate": 2.192109347400512e-06, + "loss": 0.8426, + "step": 4036 + }, + { + "epoch": 0.48542054951001024, + "grad_norm": 2.010942446355064, + "learning_rate": 2.191333963120004e-06, + "loss": 0.7958, + "step": 4037 + }, + { + "epoch": 0.4855407924006493, + "grad_norm": 2.51521421123159, + "learning_rate": 2.190558549814286e-06, + "loss": 0.7075, + "step": 4038 + }, + { + "epoch": 0.4856610352912884, + "grad_norm": 1.9476394524062235, + "learning_rate": 2.1897831076009872e-06, + "loss": 0.8011, + "step": 4039 + }, + { + "epoch": 0.4857812781819275, + "grad_norm": 3.1458440279410427, + "learning_rate": 2.1890076365977426e-06, + "loss": 0.8047, + "step": 4040 + }, + { + "epoch": 0.48590152107256657, + "grad_norm": 0.8624641037810337, + "learning_rate": 2.188232136922189e-06, + "loss": 0.5283, + "step": 4041 + }, + { + "epoch": 0.4860217639632057, + "grad_norm": 2.112372860665945, + "learning_rate": 2.187456608691971e-06, + "loss": 0.7747, + "step": 4042 + }, + { + "epoch": 0.4861420068538448, + "grad_norm": 1.986805966195844, + "learning_rate": 2.1866810520247334e-06, + "loss": 0.8758, + "step": 4043 + }, + { + "epoch": 0.48626224974448384, + "grad_norm": 1.7482980217411723, + "learning_rate": 2.185905467038129e-06, + "loss": 0.6634, + "step": 4044 + }, + { + "epoch": 0.48638249263512295, + "grad_norm": 1.691205301307338, + "learning_rate": 2.1851298538498127e-06, + "loss": 0.7754, + "step": 4045 + }, + { + "epoch": 0.48650273552576206, + "grad_norm": 4.207475984066236, + "learning_rate": 2.184354212577446e-06, + "loss": 0.8022, + "step": 4046 + }, + { + "epoch": 0.4866229784164011, + "grad_norm": 2.8146261867495785, + "learning_rate": 2.1835785433386907e-06, + "loss": 0.6326, + "step": 4047 + }, + { + "epoch": 0.48674322130704023, + "grad_norm": 1.973440365702405, + "learning_rate": 2.182802846251216e-06, + "loss": 0.6572, + "step": 4048 + }, + { + "epoch": 0.4868634641976793, + "grad_norm": 1.8689568089392596, + "learning_rate": 2.182027121432696e-06, + "loss": 0.7215, + "step": 4049 + }, + { + "epoch": 0.4869837070883184, + "grad_norm": 1.7444629264602998, + "learning_rate": 2.1812513690008054e-06, + "loss": 0.8274, + "step": 4050 + }, + { + "epoch": 0.4871039499789575, + "grad_norm": 2.8060187365558242, + "learning_rate": 2.180475589073227e-06, + "loss": 0.8026, + "step": 4051 + }, + { + "epoch": 0.48722419286959656, + "grad_norm": 1.7521612581696746, + "learning_rate": 2.1796997817676456e-06, + "loss": 0.7351, + "step": 4052 + }, + { + "epoch": 0.4873444357602357, + "grad_norm": 1.6445505317771794, + "learning_rate": 2.1789239472017494e-06, + "loss": 0.677, + "step": 4053 + }, + { + "epoch": 0.4874646786508748, + "grad_norm": 2.813123930494626, + "learning_rate": 2.1781480854932326e-06, + "loss": 0.7313, + "step": 4054 + }, + { + "epoch": 0.48758492154151384, + "grad_norm": 1.9946241847978086, + "learning_rate": 2.1773721967597933e-06, + "loss": 0.7966, + "step": 4055 + }, + { + "epoch": 0.48770516443215295, + "grad_norm": 0.9001991581436354, + "learning_rate": 2.1765962811191322e-06, + "loss": 0.5743, + "step": 4056 + }, + { + "epoch": 0.48782540732279206, + "grad_norm": 0.8478170848796985, + "learning_rate": 2.1758203386889566e-06, + "loss": 0.6211, + "step": 4057 + }, + { + "epoch": 0.4879456502134311, + "grad_norm": 2.296567208563888, + "learning_rate": 2.1750443695869746e-06, + "loss": 0.844, + "step": 4058 + }, + { + "epoch": 0.4880658931040702, + "grad_norm": 1.8622095636839393, + "learning_rate": 2.174268373930901e-06, + "loss": 0.8616, + "step": 4059 + }, + { + "epoch": 0.48818613599470934, + "grad_norm": 2.0230478036675437, + "learning_rate": 2.1734923518384537e-06, + "loss": 0.8055, + "step": 4060 + }, + { + "epoch": 0.4883063788853484, + "grad_norm": 1.8542020630047442, + "learning_rate": 2.1727163034273547e-06, + "loss": 0.8282, + "step": 4061 + }, + { + "epoch": 0.4884266217759875, + "grad_norm": 2.098153792323784, + "learning_rate": 2.17194022881533e-06, + "loss": 0.7929, + "step": 4062 + }, + { + "epoch": 0.4885468646666266, + "grad_norm": 1.764335836853542, + "learning_rate": 2.1711641281201092e-06, + "loss": 0.68, + "step": 4063 + }, + { + "epoch": 0.48866710755726567, + "grad_norm": 2.283704508478156, + "learning_rate": 2.1703880014594264e-06, + "loss": 0.8031, + "step": 4064 + }, + { + "epoch": 0.4887873504479048, + "grad_norm": 1.8413915906281344, + "learning_rate": 2.1696118489510182e-06, + "loss": 0.7423, + "step": 4065 + }, + { + "epoch": 0.48890759333854383, + "grad_norm": 1.8538165958555166, + "learning_rate": 2.1688356707126286e-06, + "loss": 0.7232, + "step": 4066 + }, + { + "epoch": 0.48902783622918294, + "grad_norm": 2.109221728372453, + "learning_rate": 2.168059466862001e-06, + "loss": 0.702, + "step": 4067 + }, + { + "epoch": 0.48914807911982205, + "grad_norm": 2.0147859468189226, + "learning_rate": 2.167283237516887e-06, + "loss": 0.8245, + "step": 4068 + }, + { + "epoch": 0.4892683220104611, + "grad_norm": 2.330042516623347, + "learning_rate": 2.1665069827950383e-06, + "loss": 0.756, + "step": 4069 + }, + { + "epoch": 0.4893885649011002, + "grad_norm": 1.866085732979487, + "learning_rate": 2.1657307028142126e-06, + "loss": 0.8709, + "step": 4070 + }, + { + "epoch": 0.48950880779173933, + "grad_norm": 2.2201413372515377, + "learning_rate": 2.164954397692171e-06, + "loss": 0.6867, + "step": 4071 + }, + { + "epoch": 0.4896290506823784, + "grad_norm": 1.09062604275629, + "learning_rate": 2.164178067546678e-06, + "loss": 0.7728, + "step": 4072 + }, + { + "epoch": 0.4897492935730175, + "grad_norm": 1.7399773370428857, + "learning_rate": 2.163401712495504e-06, + "loss": 0.9112, + "step": 4073 + }, + { + "epoch": 0.4898695364636566, + "grad_norm": 1.6510515503532373, + "learning_rate": 2.1626253326564194e-06, + "loss": 0.7966, + "step": 4074 + }, + { + "epoch": 0.48998977935429566, + "grad_norm": 1.8225933405831207, + "learning_rate": 2.161848928147201e-06, + "loss": 0.7731, + "step": 4075 + }, + { + "epoch": 0.4901100222449348, + "grad_norm": 2.1837713889268398, + "learning_rate": 2.161072499085629e-06, + "loss": 0.8122, + "step": 4076 + }, + { + "epoch": 0.4902302651355739, + "grad_norm": 1.61990445184916, + "learning_rate": 2.160296045589487e-06, + "loss": 0.8369, + "step": 4077 + }, + { + "epoch": 0.49035050802621294, + "grad_norm": 1.7382159617303765, + "learning_rate": 2.159519567776562e-06, + "loss": 0.7025, + "step": 4078 + }, + { + "epoch": 0.49047075091685205, + "grad_norm": 3.1199654269136894, + "learning_rate": 2.1587430657646463e-06, + "loss": 0.7076, + "step": 4079 + }, + { + "epoch": 0.4905909938074911, + "grad_norm": 1.9187717349533042, + "learning_rate": 2.157966539671533e-06, + "loss": 0.7806, + "step": 4080 + }, + { + "epoch": 0.4907112366981302, + "grad_norm": 1.9999861120695486, + "learning_rate": 2.157189989615021e-06, + "loss": 0.6733, + "step": 4081 + }, + { + "epoch": 0.4908314795887693, + "grad_norm": 1.7937517305691442, + "learning_rate": 2.156413415712913e-06, + "loss": 0.7584, + "step": 4082 + }, + { + "epoch": 0.4909517224794084, + "grad_norm": 1.6989527000772002, + "learning_rate": 2.155636818083014e-06, + "loss": 0.7877, + "step": 4083 + }, + { + "epoch": 0.4910719653700475, + "grad_norm": 1.812994396262075, + "learning_rate": 2.154860196843134e-06, + "loss": 0.8488, + "step": 4084 + }, + { + "epoch": 0.4911922082606866, + "grad_norm": 1.7585791378354718, + "learning_rate": 2.154083552111085e-06, + "loss": 0.7713, + "step": 4085 + }, + { + "epoch": 0.49131245115132566, + "grad_norm": 1.9668538484020923, + "learning_rate": 2.1533068840046834e-06, + "loss": 0.8276, + "step": 4086 + }, + { + "epoch": 0.49143269404196477, + "grad_norm": 2.7075217767264133, + "learning_rate": 2.152530192641749e-06, + "loss": 0.6208, + "step": 4087 + }, + { + "epoch": 0.4915529369326039, + "grad_norm": 8.045800709538932, + "learning_rate": 2.1517534781401068e-06, + "loss": 0.7298, + "step": 4088 + }, + { + "epoch": 0.49167317982324293, + "grad_norm": 2.106647791214575, + "learning_rate": 2.150976740617581e-06, + "loss": 0.7035, + "step": 4089 + }, + { + "epoch": 0.49179342271388204, + "grad_norm": 2.167342691839957, + "learning_rate": 2.150199980192006e-06, + "loss": 0.7175, + "step": 4090 + }, + { + "epoch": 0.49191366560452116, + "grad_norm": 1.947182610520487, + "learning_rate": 2.1494231969812114e-06, + "loss": 0.8095, + "step": 4091 + }, + { + "epoch": 0.4920339084951602, + "grad_norm": 2.768538153059092, + "learning_rate": 2.1486463911030372e-06, + "loss": 0.8172, + "step": 4092 + }, + { + "epoch": 0.4921541513857993, + "grad_norm": 2.044191304662402, + "learning_rate": 2.147869562675324e-06, + "loss": 0.7465, + "step": 4093 + }, + { + "epoch": 0.49227439427643843, + "grad_norm": 1.8220698833107276, + "learning_rate": 2.147092711815915e-06, + "loss": 0.7275, + "step": 4094 + }, + { + "epoch": 0.4923946371670775, + "grad_norm": 2.4961736006642887, + "learning_rate": 2.1463158386426593e-06, + "loss": 0.8688, + "step": 4095 + }, + { + "epoch": 0.4925148800577166, + "grad_norm": 2.168918930642119, + "learning_rate": 2.145538943273407e-06, + "loss": 0.776, + "step": 4096 + }, + { + "epoch": 0.49263512294835565, + "grad_norm": 2.303589743692412, + "learning_rate": 2.144762025826013e-06, + "loss": 0.7232, + "step": 4097 + }, + { + "epoch": 0.49275536583899476, + "grad_norm": 2.61287185358806, + "learning_rate": 2.143985086418334e-06, + "loss": 0.8742, + "step": 4098 + }, + { + "epoch": 0.4928756087296339, + "grad_norm": 1.4468133505422782, + "learning_rate": 2.1432081251682324e-06, + "loss": 0.7657, + "step": 4099 + }, + { + "epoch": 0.49299585162027293, + "grad_norm": 1.7882260695277108, + "learning_rate": 2.142431142193572e-06, + "loss": 0.8725, + "step": 4100 + }, + { + "epoch": 0.49311609451091204, + "grad_norm": 2.064467329916539, + "learning_rate": 2.1416541376122207e-06, + "loss": 0.7173, + "step": 4101 + }, + { + "epoch": 0.49323633740155115, + "grad_norm": 1.8391000982053347, + "learning_rate": 2.1408771115420496e-06, + "loss": 0.7387, + "step": 4102 + }, + { + "epoch": 0.4933565802921902, + "grad_norm": 1.9050620708215418, + "learning_rate": 2.140100064100932e-06, + "loss": 0.655, + "step": 4103 + }, + { + "epoch": 0.4934768231828293, + "grad_norm": 3.4355145355836503, + "learning_rate": 2.139322995406746e-06, + "loss": 0.7645, + "step": 4104 + }, + { + "epoch": 0.4935970660734684, + "grad_norm": 2.0008381041194925, + "learning_rate": 2.1385459055773727e-06, + "loss": 0.8037, + "step": 4105 + }, + { + "epoch": 0.4937173089641075, + "grad_norm": 2.0864173822928014, + "learning_rate": 2.137768794730696e-06, + "loss": 0.7442, + "step": 4106 + }, + { + "epoch": 0.4938375518547466, + "grad_norm": 2.262299300085854, + "learning_rate": 2.1369916629846026e-06, + "loss": 0.8105, + "step": 4107 + }, + { + "epoch": 0.4939577947453857, + "grad_norm": 4.624805961868727, + "learning_rate": 2.136214510456983e-06, + "loss": 0.754, + "step": 4108 + }, + { + "epoch": 0.49407803763602476, + "grad_norm": 0.9052983911196386, + "learning_rate": 2.1354373372657296e-06, + "loss": 0.6323, + "step": 4109 + }, + { + "epoch": 0.49419828052666387, + "grad_norm": 1.7397075159515607, + "learning_rate": 2.1346601435287404e-06, + "loss": 0.7105, + "step": 4110 + }, + { + "epoch": 0.494318523417303, + "grad_norm": 2.0972245267584504, + "learning_rate": 2.1338829293639144e-06, + "loss": 0.8057, + "step": 4111 + }, + { + "epoch": 0.49443876630794203, + "grad_norm": 2.508586253176084, + "learning_rate": 2.1331056948891547e-06, + "loss": 0.8323, + "step": 4112 + }, + { + "epoch": 0.49455900919858115, + "grad_norm": 2.1812847670268276, + "learning_rate": 2.1323284402223666e-06, + "loss": 0.767, + "step": 4113 + }, + { + "epoch": 0.4946792520892202, + "grad_norm": 1.8227854290905336, + "learning_rate": 2.1315511654814597e-06, + "loss": 0.8811, + "step": 4114 + }, + { + "epoch": 0.4947994949798593, + "grad_norm": 1.934235622333341, + "learning_rate": 2.1307738707843456e-06, + "loss": 0.7904, + "step": 4115 + }, + { + "epoch": 0.4949197378704984, + "grad_norm": 2.001255833214021, + "learning_rate": 2.1299965562489385e-06, + "loss": 0.7011, + "step": 4116 + }, + { + "epoch": 0.4950399807611375, + "grad_norm": 1.4160522456202136, + "learning_rate": 2.129219221993158e-06, + "loss": 0.7903, + "step": 4117 + }, + { + "epoch": 0.4951602236517766, + "grad_norm": 0.795544055138417, + "learning_rate": 2.128441868134924e-06, + "loss": 0.5998, + "step": 4118 + }, + { + "epoch": 0.4952804665424157, + "grad_norm": 2.5584347741045046, + "learning_rate": 2.1276644947921606e-06, + "loss": 0.8377, + "step": 4119 + }, + { + "epoch": 0.49540070943305475, + "grad_norm": 1.8864490640104643, + "learning_rate": 2.126887102082795e-06, + "loss": 0.8274, + "step": 4120 + }, + { + "epoch": 0.49552095232369386, + "grad_norm": 1.6853239369512727, + "learning_rate": 2.126109690124757e-06, + "loss": 0.7048, + "step": 4121 + }, + { + "epoch": 0.495641195214333, + "grad_norm": 1.6317500257905562, + "learning_rate": 2.1253322590359786e-06, + "loss": 0.7146, + "step": 4122 + }, + { + "epoch": 0.49576143810497203, + "grad_norm": 2.153619128749345, + "learning_rate": 2.124554808934397e-06, + "loss": 0.7502, + "step": 4123 + }, + { + "epoch": 0.49588168099561114, + "grad_norm": 3.150839067151384, + "learning_rate": 2.1237773399379496e-06, + "loss": 0.7342, + "step": 4124 + }, + { + "epoch": 0.49600192388625025, + "grad_norm": 1.8134459296265095, + "learning_rate": 2.122999852164578e-06, + "loss": 0.8736, + "step": 4125 + }, + { + "epoch": 0.4961221667768893, + "grad_norm": 2.2895426067679034, + "learning_rate": 2.122222345732227e-06, + "loss": 0.586, + "step": 4126 + }, + { + "epoch": 0.4962424096675284, + "grad_norm": 1.7390230460545548, + "learning_rate": 2.121444820758843e-06, + "loss": 0.8357, + "step": 4127 + }, + { + "epoch": 0.49636265255816747, + "grad_norm": 3.384271813507691, + "learning_rate": 2.120667277362376e-06, + "loss": 0.7878, + "step": 4128 + }, + { + "epoch": 0.4964828954488066, + "grad_norm": 2.144345245910949, + "learning_rate": 2.1198897156607796e-06, + "loss": 0.85, + "step": 4129 + }, + { + "epoch": 0.4966031383394457, + "grad_norm": 2.1370014791666496, + "learning_rate": 2.1191121357720085e-06, + "loss": 0.7429, + "step": 4130 + }, + { + "epoch": 0.49672338123008475, + "grad_norm": 1.861090245612295, + "learning_rate": 2.1183345378140206e-06, + "loss": 0.7499, + "step": 4131 + }, + { + "epoch": 0.49684362412072386, + "grad_norm": 0.8790920485037, + "learning_rate": 2.1175569219047783e-06, + "loss": 0.6202, + "step": 4132 + }, + { + "epoch": 0.49696386701136297, + "grad_norm": 1.5886474157269672, + "learning_rate": 2.1167792881622437e-06, + "loss": 0.7364, + "step": 4133 + }, + { + "epoch": 0.497084109902002, + "grad_norm": 1.580584130292383, + "learning_rate": 2.116001636704384e-06, + "loss": 0.8119, + "step": 4134 + }, + { + "epoch": 0.49720435279264114, + "grad_norm": 2.130409983547309, + "learning_rate": 2.1152239676491685e-06, + "loss": 0.8039, + "step": 4135 + }, + { + "epoch": 0.49732459568328025, + "grad_norm": 2.2933158026993676, + "learning_rate": 2.114446281114569e-06, + "loss": 0.742, + "step": 4136 + }, + { + "epoch": 0.4974448385739193, + "grad_norm": 1.7001619374057637, + "learning_rate": 2.1136685772185587e-06, + "loss": 0.7637, + "step": 4137 + }, + { + "epoch": 0.4975650814645584, + "grad_norm": 1.8535147245865173, + "learning_rate": 2.1128908560791163e-06, + "loss": 0.7821, + "step": 4138 + }, + { + "epoch": 0.4976853243551975, + "grad_norm": 2.8671781867185255, + "learning_rate": 2.1121131178142203e-06, + "loss": 0.7825, + "step": 4139 + }, + { + "epoch": 0.4978055672458366, + "grad_norm": 1.6983445297386262, + "learning_rate": 2.1113353625418544e-06, + "loss": 0.8307, + "step": 4140 + }, + { + "epoch": 0.4979258101364757, + "grad_norm": 1.7907998257652062, + "learning_rate": 2.1105575903800017e-06, + "loss": 0.7986, + "step": 4141 + }, + { + "epoch": 0.4980460530271148, + "grad_norm": 1.9982651100040802, + "learning_rate": 2.1097798014466502e-06, + "loss": 0.8583, + "step": 4142 + }, + { + "epoch": 0.49816629591775385, + "grad_norm": 2.2391062975138456, + "learning_rate": 2.109001995859791e-06, + "loss": 0.5913, + "step": 4143 + }, + { + "epoch": 0.49828653880839296, + "grad_norm": 0.8085839091268806, + "learning_rate": 2.108224173737415e-06, + "loss": 0.6022, + "step": 4144 + }, + { + "epoch": 0.498406781699032, + "grad_norm": 1.9763643670573252, + "learning_rate": 2.1074463351975183e-06, + "loss": 0.7639, + "step": 4145 + }, + { + "epoch": 0.49852702458967113, + "grad_norm": 1.8633807333693417, + "learning_rate": 2.106668480358098e-06, + "loss": 0.7182, + "step": 4146 + }, + { + "epoch": 0.49864726748031024, + "grad_norm": 1.7534156571970283, + "learning_rate": 2.105890609337154e-06, + "loss": 0.7172, + "step": 4147 + }, + { + "epoch": 0.4987675103709493, + "grad_norm": 0.7030045724098763, + "learning_rate": 2.1051127222526883e-06, + "loss": 0.6385, + "step": 4148 + }, + { + "epoch": 0.4988877532615884, + "grad_norm": 1.7319661501160193, + "learning_rate": 2.1043348192227067e-06, + "loss": 0.811, + "step": 4149 + }, + { + "epoch": 0.4990079961522275, + "grad_norm": 1.8340999416547026, + "learning_rate": 2.1035569003652156e-06, + "loss": 0.6197, + "step": 4150 + }, + { + "epoch": 0.4991282390428666, + "grad_norm": 2.1652537898845425, + "learning_rate": 2.1027789657982255e-06, + "loss": 0.8244, + "step": 4151 + }, + { + "epoch": 0.4992484819335057, + "grad_norm": 2.050101961531021, + "learning_rate": 2.1020010156397482e-06, + "loss": 0.7761, + "step": 4152 + }, + { + "epoch": 0.4993687248241448, + "grad_norm": 1.514321228533717, + "learning_rate": 2.101223050007797e-06, + "loss": 0.7767, + "step": 4153 + }, + { + "epoch": 0.49948896771478385, + "grad_norm": 0.824854998417021, + "learning_rate": 2.1004450690203904e-06, + "loss": 0.538, + "step": 4154 + }, + { + "epoch": 0.49960921060542296, + "grad_norm": 0.8954778934585053, + "learning_rate": 2.099667072795546e-06, + "loss": 0.6342, + "step": 4155 + }, + { + "epoch": 0.49972945349606207, + "grad_norm": 1.907699721593507, + "learning_rate": 2.0988890614512864e-06, + "loss": 0.7971, + "step": 4156 + }, + { + "epoch": 0.4998496963867011, + "grad_norm": 1.8729016324275432, + "learning_rate": 2.098111035105635e-06, + "loss": 0.8457, + "step": 4157 + }, + { + "epoch": 0.49996993927734024, + "grad_norm": 1.660427008596239, + "learning_rate": 2.0973329938766176e-06, + "loss": 0.7391, + "step": 4158 + }, + { + "epoch": 0.5000901821679793, + "grad_norm": 2.5089553653157095, + "learning_rate": 2.0965549378822618e-06, + "loss": 0.7962, + "step": 4159 + }, + { + "epoch": 0.5002104250586185, + "grad_norm": 2.040654171418239, + "learning_rate": 2.095776867240599e-06, + "loss": 0.8451, + "step": 4160 + }, + { + "epoch": 0.5003306679492575, + "grad_norm": 2.2383680609733285, + "learning_rate": 2.094998782069661e-06, + "loss": 0.8305, + "step": 4161 + }, + { + "epoch": 0.5004509108398966, + "grad_norm": 1.9001335448464491, + "learning_rate": 2.0942206824874845e-06, + "loss": 0.7569, + "step": 4162 + }, + { + "epoch": 0.5005711537305357, + "grad_norm": 2.005293040942822, + "learning_rate": 2.093442568612105e-06, + "loss": 0.7902, + "step": 4163 + }, + { + "epoch": 0.5006913966211748, + "grad_norm": 1.5725113386464051, + "learning_rate": 2.0926644405615613e-06, + "loss": 0.8516, + "step": 4164 + }, + { + "epoch": 0.5008116395118138, + "grad_norm": 1.9939575111239463, + "learning_rate": 2.091886298453897e-06, + "loss": 0.8163, + "step": 4165 + }, + { + "epoch": 0.500931882402453, + "grad_norm": 2.1085531258388683, + "learning_rate": 2.091108142407153e-06, + "loss": 0.7305, + "step": 4166 + }, + { + "epoch": 0.5010521252930921, + "grad_norm": 0.882290035698174, + "learning_rate": 2.090329972539377e-06, + "loss": 0.6258, + "step": 4167 + }, + { + "epoch": 0.5011723681837311, + "grad_norm": 1.6705580185671636, + "learning_rate": 2.089551788968616e-06, + "loss": 0.6898, + "step": 4168 + }, + { + "epoch": 0.5012926110743702, + "grad_norm": 0.8495038197158159, + "learning_rate": 2.08877359181292e-06, + "loss": 0.6091, + "step": 4169 + }, + { + "epoch": 0.5014128539650093, + "grad_norm": 2.506972888375404, + "learning_rate": 2.0879953811903396e-06, + "loss": 0.8587, + "step": 4170 + }, + { + "epoch": 0.5015330968556484, + "grad_norm": 3.1878555697914073, + "learning_rate": 2.08721715721893e-06, + "loss": 0.7886, + "step": 4171 + }, + { + "epoch": 0.5016533397462875, + "grad_norm": 1.9583478142493782, + "learning_rate": 2.0864389200167477e-06, + "loss": 0.7767, + "step": 4172 + }, + { + "epoch": 0.5017735826369266, + "grad_norm": 1.8204931079510882, + "learning_rate": 2.0856606697018504e-06, + "loss": 0.7958, + "step": 4173 + }, + { + "epoch": 0.5018938255275657, + "grad_norm": 2.1462880982898347, + "learning_rate": 2.084882406392297e-06, + "loss": 0.736, + "step": 4174 + }, + { + "epoch": 0.5020140684182047, + "grad_norm": 2.254149425485314, + "learning_rate": 2.0841041302061496e-06, + "loss": 0.7132, + "step": 4175 + }, + { + "epoch": 0.5021343113088439, + "grad_norm": 2.2407203685942507, + "learning_rate": 2.083325841261473e-06, + "loss": 0.7584, + "step": 4176 + }, + { + "epoch": 0.502254554199483, + "grad_norm": 2.105541447391176, + "learning_rate": 2.0825475396763322e-06, + "loss": 0.6725, + "step": 4177 + }, + { + "epoch": 0.502374797090122, + "grad_norm": 1.3680437295126076, + "learning_rate": 2.081769225568796e-06, + "loss": 0.6601, + "step": 4178 + }, + { + "epoch": 0.5024950399807612, + "grad_norm": 1.5514683966704927, + "learning_rate": 2.0809908990569327e-06, + "loss": 0.7624, + "step": 4179 + }, + { + "epoch": 0.5026152828714002, + "grad_norm": 1.9538852280222212, + "learning_rate": 2.0802125602588146e-06, + "loss": 0.7992, + "step": 4180 + }, + { + "epoch": 0.5027355257620393, + "grad_norm": 1.8854428986749554, + "learning_rate": 2.0794342092925146e-06, + "loss": 0.6696, + "step": 4181 + }, + { + "epoch": 0.5028557686526784, + "grad_norm": 2.1643104824283794, + "learning_rate": 2.078655846276108e-06, + "loss": 0.6809, + "step": 4182 + }, + { + "epoch": 0.5029760115433175, + "grad_norm": 2.349509118938589, + "learning_rate": 2.0778774713276727e-06, + "loss": 0.6904, + "step": 4183 + }, + { + "epoch": 0.5030962544339566, + "grad_norm": 2.22020684352597, + "learning_rate": 2.077099084565287e-06, + "loss": 0.6849, + "step": 4184 + }, + { + "epoch": 0.5032164973245957, + "grad_norm": 3.666899081291369, + "learning_rate": 2.0763206861070313e-06, + "loss": 0.6585, + "step": 4185 + }, + { + "epoch": 0.5033367402152348, + "grad_norm": 1.9783009358972163, + "learning_rate": 2.0755422760709876e-06, + "loss": 0.7543, + "step": 4186 + }, + { + "epoch": 0.5034569831058738, + "grad_norm": 2.08658489796305, + "learning_rate": 2.0747638545752417e-06, + "loss": 0.7823, + "step": 4187 + }, + { + "epoch": 0.503577225996513, + "grad_norm": 1.904742473406678, + "learning_rate": 2.073985421737878e-06, + "loss": 0.8343, + "step": 4188 + }, + { + "epoch": 0.5036974688871521, + "grad_norm": 2.3526970543863097, + "learning_rate": 2.0732069776769844e-06, + "loss": 0.7439, + "step": 4189 + }, + { + "epoch": 0.5038177117777911, + "grad_norm": 1.93839563157569, + "learning_rate": 2.072428522510651e-06, + "loss": 0.7391, + "step": 4190 + }, + { + "epoch": 0.5039379546684303, + "grad_norm": 2.7202285604656633, + "learning_rate": 2.071650056356968e-06, + "loss": 0.768, + "step": 4191 + }, + { + "epoch": 0.5040581975590693, + "grad_norm": 2.2106614226925956, + "learning_rate": 2.070871579334028e-06, + "loss": 0.806, + "step": 4192 + }, + { + "epoch": 0.5041784404497084, + "grad_norm": 1.7739923088268261, + "learning_rate": 2.0700930915599264e-06, + "loss": 0.7234, + "step": 4193 + }, + { + "epoch": 0.5042986833403476, + "grad_norm": 2.088760808417928, + "learning_rate": 2.0693145931527583e-06, + "loss": 0.784, + "step": 4194 + }, + { + "epoch": 0.5044189262309866, + "grad_norm": 1.5777874576170183, + "learning_rate": 2.068536084230622e-06, + "loss": 0.7825, + "step": 4195 + }, + { + "epoch": 0.5045391691216257, + "grad_norm": 3.6102083903685127, + "learning_rate": 2.067757564911616e-06, + "loss": 0.8884, + "step": 4196 + }, + { + "epoch": 0.5046594120122648, + "grad_norm": 2.0637078649793286, + "learning_rate": 2.0669790353138407e-06, + "loss": 0.9289, + "step": 4197 + }, + { + "epoch": 0.5047796549029039, + "grad_norm": 2.4400253907430356, + "learning_rate": 2.0662004955553995e-06, + "loss": 0.7326, + "step": 4198 + }, + { + "epoch": 0.5048998977935429, + "grad_norm": 1.8735117410007343, + "learning_rate": 2.065421945754395e-06, + "loss": 0.776, + "step": 4199 + }, + { + "epoch": 0.505020140684182, + "grad_norm": 1.7136426468828967, + "learning_rate": 2.0646433860289344e-06, + "loss": 0.785, + "step": 4200 + }, + { + "epoch": 0.5051403835748212, + "grad_norm": 2.5793991177214504, + "learning_rate": 2.0638648164971233e-06, + "loss": 0.8312, + "step": 4201 + }, + { + "epoch": 0.5052606264654602, + "grad_norm": 1.7423636172105, + "learning_rate": 2.06308623727707e-06, + "loss": 0.896, + "step": 4202 + }, + { + "epoch": 0.5053808693560993, + "grad_norm": 2.391130082373458, + "learning_rate": 2.0623076484868846e-06, + "loss": 0.7663, + "step": 4203 + }, + { + "epoch": 0.5055011122467384, + "grad_norm": 0.8487586662288257, + "learning_rate": 2.061529050244679e-06, + "loss": 0.6073, + "step": 4204 + }, + { + "epoch": 0.5056213551373775, + "grad_norm": 2.0041941534471936, + "learning_rate": 2.060750442668565e-06, + "loss": 0.7479, + "step": 4205 + }, + { + "epoch": 0.5057415980280165, + "grad_norm": 2.838498475099053, + "learning_rate": 2.059971825876657e-06, + "loss": 0.652, + "step": 4206 + }, + { + "epoch": 0.5058618409186557, + "grad_norm": 2.0273059995807934, + "learning_rate": 2.0591931999870713e-06, + "loss": 0.7679, + "step": 4207 + }, + { + "epoch": 0.5059820838092948, + "grad_norm": 0.8225611429764865, + "learning_rate": 2.0584145651179234e-06, + "loss": 0.5758, + "step": 4208 + }, + { + "epoch": 0.5061023266999338, + "grad_norm": 2.7202309269207947, + "learning_rate": 2.0576359213873327e-06, + "loss": 0.8017, + "step": 4209 + }, + { + "epoch": 0.506222569590573, + "grad_norm": 2.3346604138277343, + "learning_rate": 2.056857268913419e-06, + "loss": 0.708, + "step": 4210 + }, + { + "epoch": 0.506342812481212, + "grad_norm": 2.2398608959807094, + "learning_rate": 2.056078607814303e-06, + "loss": 0.848, + "step": 4211 + }, + { + "epoch": 0.5064630553718511, + "grad_norm": 1.839837141085861, + "learning_rate": 2.055299938208106e-06, + "loss": 0.7913, + "step": 4212 + }, + { + "epoch": 0.5065832982624903, + "grad_norm": 1.6759901650124274, + "learning_rate": 2.0545212602129526e-06, + "loss": 0.8609, + "step": 4213 + }, + { + "epoch": 0.5067035411531293, + "grad_norm": 2.132515352111151, + "learning_rate": 2.0537425739469673e-06, + "loss": 0.6635, + "step": 4214 + }, + { + "epoch": 0.5068237840437684, + "grad_norm": 0.8523488744761741, + "learning_rate": 2.052963879528276e-06, + "loss": 0.5955, + "step": 4215 + }, + { + "epoch": 0.5069440269344075, + "grad_norm": 1.9828721009683978, + "learning_rate": 2.052185177075007e-06, + "loss": 0.7711, + "step": 4216 + }, + { + "epoch": 0.5070642698250466, + "grad_norm": 2.050033113165659, + "learning_rate": 2.051406466705288e-06, + "loss": 0.8304, + "step": 4217 + }, + { + "epoch": 0.5071845127156857, + "grad_norm": 1.8315819553343622, + "learning_rate": 2.0506277485372486e-06, + "loss": 0.8101, + "step": 4218 + }, + { + "epoch": 0.5073047556063248, + "grad_norm": 4.23585692652355, + "learning_rate": 2.04984902268902e-06, + "loss": 0.6799, + "step": 4219 + }, + { + "epoch": 0.5074249984969639, + "grad_norm": 2.5797790769869366, + "learning_rate": 2.0490702892787345e-06, + "loss": 0.7673, + "step": 4220 + }, + { + "epoch": 0.5075452413876029, + "grad_norm": 1.6359989158374952, + "learning_rate": 2.0482915484245246e-06, + "loss": 0.6201, + "step": 4221 + }, + { + "epoch": 0.5076654842782421, + "grad_norm": 2.369381834929441, + "learning_rate": 2.047512800244526e-06, + "loss": 0.8478, + "step": 4222 + }, + { + "epoch": 0.5077857271688812, + "grad_norm": 2.2826042858420794, + "learning_rate": 2.046734044856873e-06, + "loss": 0.7916, + "step": 4223 + }, + { + "epoch": 0.5079059700595202, + "grad_norm": 2.1907642532471776, + "learning_rate": 2.045955282379702e-06, + "loss": 0.8138, + "step": 4224 + }, + { + "epoch": 0.5080262129501594, + "grad_norm": 3.0849522816632287, + "learning_rate": 2.045176512931152e-06, + "loss": 0.7669, + "step": 4225 + }, + { + "epoch": 0.5081464558407984, + "grad_norm": 2.132108914647579, + "learning_rate": 2.0443977366293604e-06, + "loss": 0.7653, + "step": 4226 + }, + { + "epoch": 0.5082666987314375, + "grad_norm": 1.67772648801715, + "learning_rate": 2.043618953592468e-06, + "loss": 0.774, + "step": 4227 + }, + { + "epoch": 0.5083869416220766, + "grad_norm": 1.8329298991582497, + "learning_rate": 2.0428401639386144e-06, + "loss": 0.8205, + "step": 4228 + }, + { + "epoch": 0.5085071845127157, + "grad_norm": 0.8341534195329855, + "learning_rate": 2.042061367785943e-06, + "loss": 0.5823, + "step": 4229 + }, + { + "epoch": 0.5086274274033548, + "grad_norm": 2.1284367314820223, + "learning_rate": 2.041282565252594e-06, + "loss": 0.7516, + "step": 4230 + }, + { + "epoch": 0.5087476702939938, + "grad_norm": 4.557898094882658, + "learning_rate": 2.040503756456714e-06, + "loss": 0.7755, + "step": 4231 + }, + { + "epoch": 0.508867913184633, + "grad_norm": 2.006910900934152, + "learning_rate": 2.0397249415164456e-06, + "loss": 0.7886, + "step": 4232 + }, + { + "epoch": 0.508988156075272, + "grad_norm": 1.6739605881257016, + "learning_rate": 2.0389461205499354e-06, + "loss": 0.8039, + "step": 4233 + }, + { + "epoch": 0.5091083989659111, + "grad_norm": 2.9668997369429704, + "learning_rate": 2.03816729367533e-06, + "loss": 0.7399, + "step": 4234 + }, + { + "epoch": 0.5092286418565503, + "grad_norm": 2.072952602464762, + "learning_rate": 2.0373884610107765e-06, + "loss": 0.7119, + "step": 4235 + }, + { + "epoch": 0.5093488847471893, + "grad_norm": 3.01485025998787, + "learning_rate": 2.0366096226744225e-06, + "loss": 0.7016, + "step": 4236 + }, + { + "epoch": 0.5094691276378284, + "grad_norm": 1.794212691016042, + "learning_rate": 2.035830778784418e-06, + "loss": 0.77, + "step": 4237 + }, + { + "epoch": 0.5095893705284675, + "grad_norm": 1.936973807961817, + "learning_rate": 2.0350519294589134e-06, + "loss": 0.8022, + "step": 4238 + }, + { + "epoch": 0.5097096134191066, + "grad_norm": 1.6506794513126157, + "learning_rate": 2.0342730748160588e-06, + "loss": 0.8299, + "step": 4239 + }, + { + "epoch": 0.5098298563097456, + "grad_norm": 2.095345643511755, + "learning_rate": 2.033494214974006e-06, + "loss": 0.7148, + "step": 4240 + }, + { + "epoch": 0.5099500992003848, + "grad_norm": 1.9381840636532974, + "learning_rate": 2.0327153500509067e-06, + "loss": 0.8449, + "step": 4241 + }, + { + "epoch": 0.5100703420910239, + "grad_norm": 2.241613228124324, + "learning_rate": 2.031936480164916e-06, + "loss": 0.8482, + "step": 4242 + }, + { + "epoch": 0.5101905849816629, + "grad_norm": 1.787108040897726, + "learning_rate": 2.0311576054341857e-06, + "loss": 0.806, + "step": 4243 + }, + { + "epoch": 0.5103108278723021, + "grad_norm": 1.7593348713663675, + "learning_rate": 2.0303787259768715e-06, + "loss": 0.6338, + "step": 4244 + }, + { + "epoch": 0.5104310707629411, + "grad_norm": 2.694884158389045, + "learning_rate": 2.0295998419111294e-06, + "loss": 0.6977, + "step": 4245 + }, + { + "epoch": 0.5105513136535802, + "grad_norm": 2.480852424805254, + "learning_rate": 2.028820953355115e-06, + "loss": 0.7451, + "step": 4246 + }, + { + "epoch": 0.5106715565442194, + "grad_norm": 2.3674284053881487, + "learning_rate": 2.0280420604269834e-06, + "loss": 0.7895, + "step": 4247 + }, + { + "epoch": 0.5107917994348584, + "grad_norm": 0.7346147997934853, + "learning_rate": 2.027263163244895e-06, + "loss": 0.5897, + "step": 4248 + }, + { + "epoch": 0.5109120423254975, + "grad_norm": 1.5942813511123175, + "learning_rate": 2.026484261927005e-06, + "loss": 0.7441, + "step": 4249 + }, + { + "epoch": 0.5110322852161366, + "grad_norm": 2.075903831253695, + "learning_rate": 2.025705356591475e-06, + "loss": 0.7425, + "step": 4250 + }, + { + "epoch": 0.5111525281067757, + "grad_norm": 0.7732695965631674, + "learning_rate": 2.024926447356462e-06, + "loss": 0.5798, + "step": 4251 + }, + { + "epoch": 0.5112727709974147, + "grad_norm": 3.731980326119233, + "learning_rate": 2.024147534340127e-06, + "loss": 0.7917, + "step": 4252 + }, + { + "epoch": 0.5113930138880539, + "grad_norm": 1.8701311474470803, + "learning_rate": 2.02336861766063e-06, + "loss": 0.8038, + "step": 4253 + }, + { + "epoch": 0.511513256778693, + "grad_norm": 1.6908759503785253, + "learning_rate": 2.0225896974361327e-06, + "loss": 0.79, + "step": 4254 + }, + { + "epoch": 0.511633499669332, + "grad_norm": 0.8715468797620969, + "learning_rate": 2.0218107737847962e-06, + "loss": 0.5995, + "step": 4255 + }, + { + "epoch": 0.5117537425599712, + "grad_norm": 2.1012400553488972, + "learning_rate": 2.0210318468247826e-06, + "loss": 0.7495, + "step": 4256 + }, + { + "epoch": 0.5118739854506102, + "grad_norm": 1.7997250055484348, + "learning_rate": 2.020252916674255e-06, + "loss": 0.8234, + "step": 4257 + }, + { + "epoch": 0.5119942283412493, + "grad_norm": 1.7427108047142525, + "learning_rate": 2.019473983451375e-06, + "loss": 0.8112, + "step": 4258 + }, + { + "epoch": 0.5121144712318885, + "grad_norm": 1.980311460851391, + "learning_rate": 2.0186950472743076e-06, + "loss": 0.7199, + "step": 4259 + }, + { + "epoch": 0.5122347141225275, + "grad_norm": 1.8409676328442208, + "learning_rate": 2.0179161082612162e-06, + "loss": 0.7433, + "step": 4260 + }, + { + "epoch": 0.5123549570131666, + "grad_norm": 2.4266967028900384, + "learning_rate": 2.017137166530266e-06, + "loss": 0.7309, + "step": 4261 + }, + { + "epoch": 0.5124751999038056, + "grad_norm": 2.086304021630742, + "learning_rate": 2.0163582221996213e-06, + "loss": 0.8085, + "step": 4262 + }, + { + "epoch": 0.5125954427944448, + "grad_norm": 2.586326270253309, + "learning_rate": 2.015579275387446e-06, + "loss": 0.6827, + "step": 4263 + }, + { + "epoch": 0.5127156856850839, + "grad_norm": 1.9230126216115164, + "learning_rate": 2.0148003262119085e-06, + "loss": 0.6883, + "step": 4264 + }, + { + "epoch": 0.5128359285757229, + "grad_norm": 1.9257849641032896, + "learning_rate": 2.0140213747911728e-06, + "loss": 0.7672, + "step": 4265 + }, + { + "epoch": 0.5129561714663621, + "grad_norm": 1.9821400709881343, + "learning_rate": 2.013242421243406e-06, + "loss": 0.8058, + "step": 4266 + }, + { + "epoch": 0.5130764143570011, + "grad_norm": 2.466190991631051, + "learning_rate": 2.012463465686774e-06, + "loss": 0.7968, + "step": 4267 + }, + { + "epoch": 0.5131966572476402, + "grad_norm": 0.8177401218593219, + "learning_rate": 2.0116845082394446e-06, + "loss": 0.5484, + "step": 4268 + }, + { + "epoch": 0.5133169001382794, + "grad_norm": 1.8282273744789486, + "learning_rate": 2.0109055490195836e-06, + "loss": 0.7878, + "step": 4269 + }, + { + "epoch": 0.5134371430289184, + "grad_norm": 2.431906125691287, + "learning_rate": 2.0101265881453605e-06, + "loss": 0.6494, + "step": 4270 + }, + { + "epoch": 0.5135573859195575, + "grad_norm": 1.9289762599348421, + "learning_rate": 2.009347625734941e-06, + "loss": 0.7893, + "step": 4271 + }, + { + "epoch": 0.5136776288101966, + "grad_norm": 2.5186843749932146, + "learning_rate": 2.0085686619064954e-06, + "loss": 0.7574, + "step": 4272 + }, + { + "epoch": 0.5137978717008357, + "grad_norm": 2.445017251419959, + "learning_rate": 2.00778969677819e-06, + "loss": 0.8331, + "step": 4273 + }, + { + "epoch": 0.5139181145914747, + "grad_norm": 1.925760079474468, + "learning_rate": 2.0070107304681934e-06, + "loss": 0.6438, + "step": 4274 + }, + { + "epoch": 0.5140383574821139, + "grad_norm": 1.6681204018899283, + "learning_rate": 2.006231763094675e-06, + "loss": 0.7876, + "step": 4275 + }, + { + "epoch": 0.514158600372753, + "grad_norm": 2.5190720723780338, + "learning_rate": 2.0054527947758027e-06, + "loss": 0.8798, + "step": 4276 + }, + { + "epoch": 0.514278843263392, + "grad_norm": 0.7428193464593557, + "learning_rate": 2.004673825629746e-06, + "loss": 0.5595, + "step": 4277 + }, + { + "epoch": 0.5143990861540312, + "grad_norm": 1.6422713965155797, + "learning_rate": 2.0038948557746744e-06, + "loss": 0.7273, + "step": 4278 + }, + { + "epoch": 0.5145193290446702, + "grad_norm": 1.6965926298964191, + "learning_rate": 2.0031158853287558e-06, + "loss": 0.7571, + "step": 4279 + }, + { + "epoch": 0.5146395719353093, + "grad_norm": 2.7092408249310624, + "learning_rate": 2.0023369144101593e-06, + "loss": 0.7143, + "step": 4280 + }, + { + "epoch": 0.5147598148259485, + "grad_norm": 1.7878609832980206, + "learning_rate": 2.0015579431370555e-06, + "loss": 0.7675, + "step": 4281 + }, + { + "epoch": 0.5148800577165875, + "grad_norm": 2.3718193238650818, + "learning_rate": 2.000778971627612e-06, + "loss": 0.7079, + "step": 4282 + }, + { + "epoch": 0.5150003006072266, + "grad_norm": 2.5964563796656295, + "learning_rate": 2e-06, + "loss": 0.9043, + "step": 4283 + }, + { + "epoch": 0.5151205434978657, + "grad_norm": 2.257771633530731, + "learning_rate": 1.9992210283723878e-06, + "loss": 0.8627, + "step": 4284 + }, + { + "epoch": 0.5152407863885048, + "grad_norm": 1.5232278532881536, + "learning_rate": 1.9984420568629448e-06, + "loss": 0.7925, + "step": 4285 + }, + { + "epoch": 0.5153610292791438, + "grad_norm": 2.1911733030920555, + "learning_rate": 1.9976630855898405e-06, + "loss": 0.7886, + "step": 4286 + }, + { + "epoch": 0.515481272169783, + "grad_norm": 2.4326073865719864, + "learning_rate": 1.9968841146712445e-06, + "loss": 0.7507, + "step": 4287 + }, + { + "epoch": 0.5156015150604221, + "grad_norm": 1.6227222030673139, + "learning_rate": 1.996105144225326e-06, + "loss": 0.7195, + "step": 4288 + }, + { + "epoch": 0.5157217579510611, + "grad_norm": 1.8885419735817959, + "learning_rate": 1.995326174370254e-06, + "loss": 0.7907, + "step": 4289 + }, + { + "epoch": 0.5158420008417003, + "grad_norm": 2.634240824222649, + "learning_rate": 1.994547205224197e-06, + "loss": 0.7357, + "step": 4290 + }, + { + "epoch": 0.5159622437323393, + "grad_norm": 1.9743456565364574, + "learning_rate": 1.993768236905325e-06, + "loss": 0.678, + "step": 4291 + }, + { + "epoch": 0.5160824866229784, + "grad_norm": 3.1299672131755947, + "learning_rate": 1.992989269531807e-06, + "loss": 0.6617, + "step": 4292 + }, + { + "epoch": 0.5162027295136175, + "grad_norm": 2.5997525354140625, + "learning_rate": 1.99221030322181e-06, + "loss": 0.6891, + "step": 4293 + }, + { + "epoch": 0.5163229724042566, + "grad_norm": 1.5001446336316924, + "learning_rate": 1.991431338093505e-06, + "loss": 0.8097, + "step": 4294 + }, + { + "epoch": 0.5164432152948957, + "grad_norm": 1.7830906108611355, + "learning_rate": 1.9906523742650587e-06, + "loss": 0.7986, + "step": 4295 + }, + { + "epoch": 0.5165634581855347, + "grad_norm": 1.9292403100652633, + "learning_rate": 1.9898734118546397e-06, + "loss": 0.7733, + "step": 4296 + }, + { + "epoch": 0.5166837010761739, + "grad_norm": 1.502878288594785, + "learning_rate": 1.989094450980416e-06, + "loss": 0.8079, + "step": 4297 + }, + { + "epoch": 0.516803943966813, + "grad_norm": 3.090475488022099, + "learning_rate": 1.9883154917605556e-06, + "loss": 0.7713, + "step": 4298 + }, + { + "epoch": 0.516924186857452, + "grad_norm": 1.834721068924875, + "learning_rate": 1.9875365343132262e-06, + "loss": 0.8363, + "step": 4299 + }, + { + "epoch": 0.5170444297480912, + "grad_norm": 2.1830953667325264, + "learning_rate": 1.9867575787565946e-06, + "loss": 0.8514, + "step": 4300 + }, + { + "epoch": 0.5171646726387302, + "grad_norm": 2.6127712054229333, + "learning_rate": 1.9859786252088275e-06, + "loss": 0.865, + "step": 4301 + }, + { + "epoch": 0.5172849155293693, + "grad_norm": 2.5387176749022977, + "learning_rate": 1.9851996737880914e-06, + "loss": 0.6746, + "step": 4302 + }, + { + "epoch": 0.5174051584200084, + "grad_norm": 2.1867475032789963, + "learning_rate": 1.9844207246125537e-06, + "loss": 0.7525, + "step": 4303 + }, + { + "epoch": 0.5175254013106475, + "grad_norm": 2.158298348790115, + "learning_rate": 1.983641777800379e-06, + "loss": 0.6888, + "step": 4304 + }, + { + "epoch": 0.5176456442012866, + "grad_norm": 0.7541166812356339, + "learning_rate": 1.9828628334697343e-06, + "loss": 0.5884, + "step": 4305 + }, + { + "epoch": 0.5177658870919257, + "grad_norm": 0.7760370284246987, + "learning_rate": 1.982083891738784e-06, + "loss": 0.5479, + "step": 4306 + }, + { + "epoch": 0.5178861299825648, + "grad_norm": 1.6193915933033454, + "learning_rate": 1.9813049527256923e-06, + "loss": 0.8353, + "step": 4307 + }, + { + "epoch": 0.5180063728732038, + "grad_norm": 2.336945281276535, + "learning_rate": 1.9805260165486252e-06, + "loss": 0.8212, + "step": 4308 + }, + { + "epoch": 0.518126615763843, + "grad_norm": 2.0207069855613855, + "learning_rate": 1.9797470833257457e-06, + "loss": 0.8667, + "step": 4309 + }, + { + "epoch": 0.5182468586544821, + "grad_norm": 1.835431812525539, + "learning_rate": 1.9789681531752177e-06, + "loss": 0.7794, + "step": 4310 + }, + { + "epoch": 0.5183671015451211, + "grad_norm": 1.5848448799593369, + "learning_rate": 1.978189226215204e-06, + "loss": 0.7284, + "step": 4311 + }, + { + "epoch": 0.5184873444357603, + "grad_norm": 1.99207889036433, + "learning_rate": 1.9774103025638675e-06, + "loss": 0.7769, + "step": 4312 + }, + { + "epoch": 0.5186075873263993, + "grad_norm": 1.720626638836397, + "learning_rate": 1.9766313823393696e-06, + "loss": 0.7694, + "step": 4313 + }, + { + "epoch": 0.5187278302170384, + "grad_norm": 2.3153941627586367, + "learning_rate": 1.975852465659873e-06, + "loss": 0.6921, + "step": 4314 + }, + { + "epoch": 0.5188480731076776, + "grad_norm": 2.3860902348903963, + "learning_rate": 1.9750735526435377e-06, + "loss": 0.7097, + "step": 4315 + }, + { + "epoch": 0.5189683159983166, + "grad_norm": 2.2828980854761465, + "learning_rate": 1.974294643408525e-06, + "loss": 0.792, + "step": 4316 + }, + { + "epoch": 0.5190885588889557, + "grad_norm": 1.8213912976081459, + "learning_rate": 1.9735157380729947e-06, + "loss": 0.6744, + "step": 4317 + }, + { + "epoch": 0.5192088017795948, + "grad_norm": 1.8241739206299104, + "learning_rate": 1.9727368367551053e-06, + "loss": 0.8479, + "step": 4318 + }, + { + "epoch": 0.5193290446702339, + "grad_norm": 2.0040394997120816, + "learning_rate": 1.9719579395730164e-06, + "loss": 0.6885, + "step": 4319 + }, + { + "epoch": 0.5194492875608729, + "grad_norm": 2.1415505740196714, + "learning_rate": 1.9711790466448854e-06, + "loss": 0.9408, + "step": 4320 + }, + { + "epoch": 0.5195695304515121, + "grad_norm": 2.0714618299661214, + "learning_rate": 1.9704001580888704e-06, + "loss": 0.7227, + "step": 4321 + }, + { + "epoch": 0.5196897733421512, + "grad_norm": 1.8552637247951165, + "learning_rate": 1.9696212740231283e-06, + "loss": 0.8804, + "step": 4322 + }, + { + "epoch": 0.5198100162327902, + "grad_norm": 2.3097372305234556, + "learning_rate": 1.9688423945658146e-06, + "loss": 0.8167, + "step": 4323 + }, + { + "epoch": 0.5199302591234293, + "grad_norm": 2.0838865562870446, + "learning_rate": 1.9680635198350845e-06, + "loss": 0.7301, + "step": 4324 + }, + { + "epoch": 0.5200505020140684, + "grad_norm": 1.97896902871052, + "learning_rate": 1.967284649949093e-06, + "loss": 0.7306, + "step": 4325 + }, + { + "epoch": 0.5201707449047075, + "grad_norm": 2.0408210041031465, + "learning_rate": 1.966505785025994e-06, + "loss": 0.7253, + "step": 4326 + }, + { + "epoch": 0.5202909877953465, + "grad_norm": 1.8008234842440913, + "learning_rate": 1.965726925183941e-06, + "loss": 0.7656, + "step": 4327 + }, + { + "epoch": 0.5204112306859857, + "grad_norm": 1.870660337451461, + "learning_rate": 1.964948070541087e-06, + "loss": 0.849, + "step": 4328 + }, + { + "epoch": 0.5205314735766248, + "grad_norm": 2.4756083771480526, + "learning_rate": 1.9641692212155816e-06, + "loss": 0.6933, + "step": 4329 + }, + { + "epoch": 0.5206517164672638, + "grad_norm": 2.3456926559989633, + "learning_rate": 1.9633903773255777e-06, + "loss": 0.7285, + "step": 4330 + }, + { + "epoch": 0.520771959357903, + "grad_norm": 2.3146638798425894, + "learning_rate": 1.9626115389892237e-06, + "loss": 0.7494, + "step": 4331 + }, + { + "epoch": 0.520892202248542, + "grad_norm": 1.870748213238381, + "learning_rate": 1.96183270632467e-06, + "loss": 0.8589, + "step": 4332 + }, + { + "epoch": 0.5210124451391811, + "grad_norm": 1.8961547816591637, + "learning_rate": 1.9610538794500644e-06, + "loss": 0.7953, + "step": 4333 + }, + { + "epoch": 0.5211326880298203, + "grad_norm": 0.7935742100974317, + "learning_rate": 1.9602750584835542e-06, + "loss": 0.5945, + "step": 4334 + }, + { + "epoch": 0.5212529309204593, + "grad_norm": 2.1287954879085027, + "learning_rate": 1.959496243543286e-06, + "loss": 0.831, + "step": 4335 + }, + { + "epoch": 0.5213731738110984, + "grad_norm": 2.0835786547863155, + "learning_rate": 1.9587174347474057e-06, + "loss": 0.8002, + "step": 4336 + }, + { + "epoch": 0.5214934167017375, + "grad_norm": 2.464507888666778, + "learning_rate": 1.9579386322140574e-06, + "loss": 0.8282, + "step": 4337 + }, + { + "epoch": 0.5216136595923766, + "grad_norm": 1.700328666106769, + "learning_rate": 1.9571598360613854e-06, + "loss": 0.8123, + "step": 4338 + }, + { + "epoch": 0.5217339024830157, + "grad_norm": 2.0653649286168765, + "learning_rate": 1.956381046407532e-06, + "loss": 0.694, + "step": 4339 + }, + { + "epoch": 0.5218541453736548, + "grad_norm": 1.884754366668124, + "learning_rate": 1.9556022633706394e-06, + "loss": 0.8629, + "step": 4340 + }, + { + "epoch": 0.5219743882642939, + "grad_norm": 1.8491331492419598, + "learning_rate": 1.954823487068848e-06, + "loss": 0.801, + "step": 4341 + }, + { + "epoch": 0.5220946311549329, + "grad_norm": 1.8535747297568919, + "learning_rate": 1.9540447176202976e-06, + "loss": 0.8169, + "step": 4342 + }, + { + "epoch": 0.5222148740455721, + "grad_norm": 0.8791694077604209, + "learning_rate": 1.9532659551431272e-06, + "loss": 0.608, + "step": 4343 + }, + { + "epoch": 0.5223351169362112, + "grad_norm": 1.8410854158825547, + "learning_rate": 1.9524871997554744e-06, + "loss": 0.6799, + "step": 4344 + }, + { + "epoch": 0.5224553598268502, + "grad_norm": 2.1915610627621547, + "learning_rate": 1.951708451575475e-06, + "loss": 0.8074, + "step": 4345 + }, + { + "epoch": 0.5225756027174894, + "grad_norm": 2.6488233977189695, + "learning_rate": 1.9509297107212657e-06, + "loss": 0.829, + "step": 4346 + }, + { + "epoch": 0.5226958456081284, + "grad_norm": 1.6587495476406604, + "learning_rate": 1.95015097731098e-06, + "loss": 0.7936, + "step": 4347 + }, + { + "epoch": 0.5228160884987675, + "grad_norm": 2.2948387617928527, + "learning_rate": 1.949372251462751e-06, + "loss": 0.8234, + "step": 4348 + }, + { + "epoch": 0.5229363313894067, + "grad_norm": 14.22821628567246, + "learning_rate": 1.9485935332947124e-06, + "loss": 0.8315, + "step": 4349 + }, + { + "epoch": 0.5230565742800457, + "grad_norm": 2.4854031720451926, + "learning_rate": 1.947814822924993e-06, + "loss": 0.8359, + "step": 4350 + }, + { + "epoch": 0.5231768171706848, + "grad_norm": 1.921735991126053, + "learning_rate": 1.9470361204717236e-06, + "loss": 0.8347, + "step": 4351 + }, + { + "epoch": 0.5232970600613239, + "grad_norm": 1.6388813016827206, + "learning_rate": 1.9462574260530326e-06, + "loss": 0.8123, + "step": 4352 + }, + { + "epoch": 0.523417302951963, + "grad_norm": 1.8018943672846226, + "learning_rate": 1.9454787397870472e-06, + "loss": 0.8127, + "step": 4353 + }, + { + "epoch": 0.523537545842602, + "grad_norm": 2.167035205993408, + "learning_rate": 1.944700061791894e-06, + "loss": 0.726, + "step": 4354 + }, + { + "epoch": 0.5236577887332411, + "grad_norm": 2.0448203221038077, + "learning_rate": 1.943921392185698e-06, + "loss": 0.6501, + "step": 4355 + }, + { + "epoch": 0.5237780316238803, + "grad_norm": 2.0671011724344983, + "learning_rate": 1.9431427310865814e-06, + "loss": 0.7721, + "step": 4356 + }, + { + "epoch": 0.5238982745145193, + "grad_norm": 1.8787426153610185, + "learning_rate": 1.942364078612667e-06, + "loss": 0.7891, + "step": 4357 + }, + { + "epoch": 0.5240185174051584, + "grad_norm": 2.2351559427975145, + "learning_rate": 1.9415854348820765e-06, + "loss": 0.7566, + "step": 4358 + }, + { + "epoch": 0.5241387602957975, + "grad_norm": 9.31084698848963, + "learning_rate": 1.940806800012929e-06, + "loss": 0.688, + "step": 4359 + }, + { + "epoch": 0.5242590031864366, + "grad_norm": 1.9720069949894583, + "learning_rate": 1.9400281741233432e-06, + "loss": 0.6329, + "step": 4360 + }, + { + "epoch": 0.5243792460770756, + "grad_norm": 0.6689460475945437, + "learning_rate": 1.939249557331435e-06, + "loss": 0.5258, + "step": 4361 + }, + { + "epoch": 0.5244994889677148, + "grad_norm": 1.9804523775827196, + "learning_rate": 1.938470949755321e-06, + "loss": 0.7352, + "step": 4362 + }, + { + "epoch": 0.5246197318583539, + "grad_norm": 0.8245127366043996, + "learning_rate": 1.937692351513115e-06, + "loss": 0.5571, + "step": 4363 + }, + { + "epoch": 0.5247399747489929, + "grad_norm": 1.6820545639640658, + "learning_rate": 1.9369137627229297e-06, + "loss": 0.8126, + "step": 4364 + }, + { + "epoch": 0.5248602176396321, + "grad_norm": 2.0305313012651345, + "learning_rate": 1.936135183502877e-06, + "loss": 0.8877, + "step": 4365 + }, + { + "epoch": 0.5249804605302711, + "grad_norm": 2.250170595271547, + "learning_rate": 1.935356613971066e-06, + "loss": 0.8072, + "step": 4366 + }, + { + "epoch": 0.5251007034209102, + "grad_norm": 2.18174400528615, + "learning_rate": 1.9345780542456047e-06, + "loss": 0.7716, + "step": 4367 + }, + { + "epoch": 0.5252209463115494, + "grad_norm": 2.2077456238074435, + "learning_rate": 1.9337995044446007e-06, + "loss": 0.7212, + "step": 4368 + }, + { + "epoch": 0.5253411892021884, + "grad_norm": 1.896566968665246, + "learning_rate": 1.9330209646861596e-06, + "loss": 0.8044, + "step": 4369 + }, + { + "epoch": 0.5254614320928275, + "grad_norm": 1.6246363159499402, + "learning_rate": 1.9322424350883843e-06, + "loss": 0.7791, + "step": 4370 + }, + { + "epoch": 0.5255816749834666, + "grad_norm": 1.6180044128819893, + "learning_rate": 1.931463915769379e-06, + "loss": 0.7906, + "step": 4371 + }, + { + "epoch": 0.5257019178741057, + "grad_norm": 2.22684891849547, + "learning_rate": 1.930685406847242e-06, + "loss": 0.7401, + "step": 4372 + }, + { + "epoch": 0.5258221607647448, + "grad_norm": 2.00515904221496, + "learning_rate": 1.9299069084400734e-06, + "loss": 0.8212, + "step": 4373 + }, + { + "epoch": 0.5259424036553839, + "grad_norm": 7.525270993241938, + "learning_rate": 1.9291284206659717e-06, + "loss": 0.7036, + "step": 4374 + }, + { + "epoch": 0.526062646546023, + "grad_norm": 2.030473766195949, + "learning_rate": 1.928349943643032e-06, + "loss": 0.717, + "step": 4375 + }, + { + "epoch": 0.526182889436662, + "grad_norm": 1.7285290733719423, + "learning_rate": 1.9275714774893493e-06, + "loss": 0.821, + "step": 4376 + }, + { + "epoch": 0.5263031323273012, + "grad_norm": 1.99426448485934, + "learning_rate": 1.9267930223230154e-06, + "loss": 0.7364, + "step": 4377 + }, + { + "epoch": 0.5264233752179402, + "grad_norm": 2.0052077916792763, + "learning_rate": 1.9260145782621224e-06, + "loss": 0.7889, + "step": 4378 + }, + { + "epoch": 0.5265436181085793, + "grad_norm": 2.5391403421060446, + "learning_rate": 1.925236145424758e-06, + "loss": 0.887, + "step": 4379 + }, + { + "epoch": 0.5266638609992185, + "grad_norm": 0.711541600379048, + "learning_rate": 1.924457723929012e-06, + "loss": 0.5762, + "step": 4380 + }, + { + "epoch": 0.5267841038898575, + "grad_norm": 1.6468142069357816, + "learning_rate": 1.9236793138929685e-06, + "loss": 0.8311, + "step": 4381 + }, + { + "epoch": 0.5269043467804966, + "grad_norm": 2.05406730777811, + "learning_rate": 1.9229009154347133e-06, + "loss": 0.8113, + "step": 4382 + }, + { + "epoch": 0.5270245896711357, + "grad_norm": 2.1773533509162206, + "learning_rate": 1.922122528672327e-06, + "loss": 0.8046, + "step": 4383 + }, + { + "epoch": 0.5271448325617748, + "grad_norm": 2.334554511627696, + "learning_rate": 1.9213441537238914e-06, + "loss": 0.7787, + "step": 4384 + }, + { + "epoch": 0.5272650754524139, + "grad_norm": 0.8451957325838013, + "learning_rate": 1.920565790707485e-06, + "loss": 0.574, + "step": 4385 + }, + { + "epoch": 0.527385318343053, + "grad_norm": 2.126735539511952, + "learning_rate": 1.9197874397411853e-06, + "loss": 0.6665, + "step": 4386 + }, + { + "epoch": 0.5275055612336921, + "grad_norm": 3.6672505578502177, + "learning_rate": 1.919009100943067e-06, + "loss": 0.6716, + "step": 4387 + }, + { + "epoch": 0.5276258041243311, + "grad_norm": 2.19206306827616, + "learning_rate": 1.9182307744312043e-06, + "loss": 0.6661, + "step": 4388 + }, + { + "epoch": 0.5277460470149702, + "grad_norm": 2.483874864439351, + "learning_rate": 1.9174524603236676e-06, + "loss": 0.7677, + "step": 4389 + }, + { + "epoch": 0.5278662899056094, + "grad_norm": 2.0493495048665866, + "learning_rate": 1.916674158738527e-06, + "loss": 0.7655, + "step": 4390 + }, + { + "epoch": 0.5279865327962484, + "grad_norm": 2.33280954839848, + "learning_rate": 1.9158958697938506e-06, + "loss": 0.6128, + "step": 4391 + }, + { + "epoch": 0.5281067756868875, + "grad_norm": 2.581313779472477, + "learning_rate": 1.9151175936077032e-06, + "loss": 0.8595, + "step": 4392 + }, + { + "epoch": 0.5282270185775266, + "grad_norm": 1.5849792140957562, + "learning_rate": 1.9143393302981507e-06, + "loss": 0.7933, + "step": 4393 + }, + { + "epoch": 0.5283472614681657, + "grad_norm": 2.060192898542591, + "learning_rate": 1.913561079983252e-06, + "loss": 0.8381, + "step": 4394 + }, + { + "epoch": 0.5284675043588047, + "grad_norm": 2.904401950436142, + "learning_rate": 1.9127828427810693e-06, + "loss": 0.7543, + "step": 4395 + }, + { + "epoch": 0.5285877472494439, + "grad_norm": 1.9574135047247947, + "learning_rate": 1.9120046188096607e-06, + "loss": 0.8113, + "step": 4396 + }, + { + "epoch": 0.528707990140083, + "grad_norm": 5.666398359940391, + "learning_rate": 1.9112264081870804e-06, + "loss": 0.743, + "step": 4397 + }, + { + "epoch": 0.528828233030722, + "grad_norm": 2.086371787363753, + "learning_rate": 1.9104482110313843e-06, + "loss": 0.7642, + "step": 4398 + }, + { + "epoch": 0.5289484759213612, + "grad_norm": 1.8367214395948779, + "learning_rate": 1.909670027460623e-06, + "loss": 0.748, + "step": 4399 + }, + { + "epoch": 0.5290687188120002, + "grad_norm": 1.8714684606937242, + "learning_rate": 1.908891857592847e-06, + "loss": 0.7281, + "step": 4400 + }, + { + "epoch": 0.5291889617026393, + "grad_norm": 2.814771455529346, + "learning_rate": 1.9081137015461034e-06, + "loss": 0.9087, + "step": 4401 + }, + { + "epoch": 0.5293092045932785, + "grad_norm": 2.016353859824958, + "learning_rate": 1.9073355594384383e-06, + "loss": 0.9094, + "step": 4402 + }, + { + "epoch": 0.5294294474839175, + "grad_norm": 1.9751947669073087, + "learning_rate": 1.906557431387895e-06, + "loss": 0.8085, + "step": 4403 + }, + { + "epoch": 0.5295496903745566, + "grad_norm": 1.9880096309973532, + "learning_rate": 1.905779317512516e-06, + "loss": 0.7922, + "step": 4404 + }, + { + "epoch": 0.5296699332651957, + "grad_norm": 4.473267945684855, + "learning_rate": 1.9050012179303385e-06, + "loss": 0.8136, + "step": 4405 + }, + { + "epoch": 0.5297901761558348, + "grad_norm": 2.280202442183072, + "learning_rate": 1.904223132759401e-06, + "loss": 0.6946, + "step": 4406 + }, + { + "epoch": 0.5299104190464738, + "grad_norm": 2.1737731762133685, + "learning_rate": 1.9034450621177383e-06, + "loss": 0.6938, + "step": 4407 + }, + { + "epoch": 0.530030661937113, + "grad_norm": 1.871126369757415, + "learning_rate": 1.9026670061233824e-06, + "loss": 0.7051, + "step": 4408 + }, + { + "epoch": 0.5301509048277521, + "grad_norm": 1.7692821740015423, + "learning_rate": 1.901888964894365e-06, + "loss": 0.8125, + "step": 4409 + }, + { + "epoch": 0.5302711477183911, + "grad_norm": 1.8473141375677462, + "learning_rate": 1.9011109385487134e-06, + "loss": 0.6776, + "step": 4410 + }, + { + "epoch": 0.5303913906090303, + "grad_norm": 3.138428788424196, + "learning_rate": 1.900332927204454e-06, + "loss": 0.6677, + "step": 4411 + }, + { + "epoch": 0.5305116334996693, + "grad_norm": 1.6912900959683714, + "learning_rate": 1.8995549309796097e-06, + "loss": 0.7724, + "step": 4412 + }, + { + "epoch": 0.5306318763903084, + "grad_norm": 1.8799118554980547, + "learning_rate": 1.8987769499922028e-06, + "loss": 0.7695, + "step": 4413 + }, + { + "epoch": 0.5307521192809476, + "grad_norm": 2.421005782147656, + "learning_rate": 1.897998984360252e-06, + "loss": 0.719, + "step": 4414 + }, + { + "epoch": 0.5308723621715866, + "grad_norm": 1.5349722016641187, + "learning_rate": 1.897221034201775e-06, + "loss": 0.7907, + "step": 4415 + }, + { + "epoch": 0.5309926050622257, + "grad_norm": 1.5133229499550305, + "learning_rate": 1.8964430996347842e-06, + "loss": 0.6698, + "step": 4416 + }, + { + "epoch": 0.5311128479528648, + "grad_norm": 1.6674996996283082, + "learning_rate": 1.8956651807772931e-06, + "loss": 0.8284, + "step": 4417 + }, + { + "epoch": 0.5312330908435039, + "grad_norm": 1.6606612570598083, + "learning_rate": 1.8948872777473115e-06, + "loss": 0.8416, + "step": 4418 + }, + { + "epoch": 0.531353333734143, + "grad_norm": 1.9764465780529503, + "learning_rate": 1.8941093906628458e-06, + "loss": 0.6361, + "step": 4419 + }, + { + "epoch": 0.531473576624782, + "grad_norm": 2.1937162086949105, + "learning_rate": 1.893331519641902e-06, + "loss": 0.7134, + "step": 4420 + }, + { + "epoch": 0.5315938195154212, + "grad_norm": 2.680982449216214, + "learning_rate": 1.8925536648024815e-06, + "loss": 0.7368, + "step": 4421 + }, + { + "epoch": 0.5317140624060602, + "grad_norm": 2.1216788424581767, + "learning_rate": 1.8917758262625849e-06, + "loss": 0.7616, + "step": 4422 + }, + { + "epoch": 0.5318343052966993, + "grad_norm": 1.717549754592315, + "learning_rate": 1.8909980041402089e-06, + "loss": 0.8104, + "step": 4423 + }, + { + "epoch": 0.5319545481873384, + "grad_norm": 2.477226288704611, + "learning_rate": 1.8902201985533494e-06, + "loss": 0.6577, + "step": 4424 + }, + { + "epoch": 0.5320747910779775, + "grad_norm": 1.8490507577352755, + "learning_rate": 1.8894424096199983e-06, + "loss": 0.7555, + "step": 4425 + }, + { + "epoch": 0.5321950339686166, + "grad_norm": 2.6559432806312753, + "learning_rate": 1.8886646374581463e-06, + "loss": 0.8647, + "step": 4426 + }, + { + "epoch": 0.5323152768592557, + "grad_norm": 1.7086821021743317, + "learning_rate": 1.8878868821857795e-06, + "loss": 0.7173, + "step": 4427 + }, + { + "epoch": 0.5324355197498948, + "grad_norm": 2.1584476936511634, + "learning_rate": 1.8871091439208838e-06, + "loss": 0.7615, + "step": 4428 + }, + { + "epoch": 0.5325557626405338, + "grad_norm": 3.3542854967254176, + "learning_rate": 1.8863314227814414e-06, + "loss": 0.7754, + "step": 4429 + }, + { + "epoch": 0.532676005531173, + "grad_norm": 2.706552614555809, + "learning_rate": 1.8855537188854313e-06, + "loss": 0.4921, + "step": 4430 + }, + { + "epoch": 0.5327962484218121, + "grad_norm": 1.9778602169776387, + "learning_rate": 1.8847760323508315e-06, + "loss": 0.7902, + "step": 4431 + }, + { + "epoch": 0.5329164913124511, + "grad_norm": 1.8515312095605645, + "learning_rate": 1.883998363295616e-06, + "loss": 0.7568, + "step": 4432 + }, + { + "epoch": 0.5330367342030903, + "grad_norm": 0.8860273985017003, + "learning_rate": 1.8832207118377565e-06, + "loss": 0.6265, + "step": 4433 + }, + { + "epoch": 0.5331569770937293, + "grad_norm": 2.4709381836060063, + "learning_rate": 1.882443078095222e-06, + "loss": 0.7065, + "step": 4434 + }, + { + "epoch": 0.5332772199843684, + "grad_norm": 0.882868098305854, + "learning_rate": 1.8816654621859794e-06, + "loss": 0.6691, + "step": 4435 + }, + { + "epoch": 0.5333974628750076, + "grad_norm": 2.3241934125785413, + "learning_rate": 1.8808878642279915e-06, + "loss": 0.7317, + "step": 4436 + }, + { + "epoch": 0.5335177057656466, + "grad_norm": 2.6740724327987997, + "learning_rate": 1.8801102843392209e-06, + "loss": 0.6606, + "step": 4437 + }, + { + "epoch": 0.5336379486562857, + "grad_norm": 1.6116141140616826, + "learning_rate": 1.8793327226376238e-06, + "loss": 0.8559, + "step": 4438 + }, + { + "epoch": 0.5337581915469248, + "grad_norm": 1.7470162022390119, + "learning_rate": 1.8785551792411569e-06, + "loss": 0.8087, + "step": 4439 + }, + { + "epoch": 0.5338784344375639, + "grad_norm": 2.055312609133863, + "learning_rate": 1.8777776542677733e-06, + "loss": 0.8307, + "step": 4440 + }, + { + "epoch": 0.5339986773282029, + "grad_norm": 1.8920995463460903, + "learning_rate": 1.8770001478354216e-06, + "loss": 0.7363, + "step": 4441 + }, + { + "epoch": 0.5341189202188421, + "grad_norm": 2.4378265748964263, + "learning_rate": 1.8762226600620504e-06, + "loss": 0.846, + "step": 4442 + }, + { + "epoch": 0.5342391631094812, + "grad_norm": 2.6597442980171913, + "learning_rate": 1.8754451910656031e-06, + "loss": 0.5928, + "step": 4443 + }, + { + "epoch": 0.5343594060001202, + "grad_norm": 4.192604938756495, + "learning_rate": 1.8746677409640212e-06, + "loss": 0.8262, + "step": 4444 + }, + { + "epoch": 0.5344796488907594, + "grad_norm": 1.7707266233119145, + "learning_rate": 1.8738903098752432e-06, + "loss": 0.8495, + "step": 4445 + }, + { + "epoch": 0.5345998917813984, + "grad_norm": 2.10280030962307, + "learning_rate": 1.8731128979172052e-06, + "loss": 0.7359, + "step": 4446 + }, + { + "epoch": 0.5347201346720375, + "grad_norm": 25.880634671054, + "learning_rate": 1.8723355052078394e-06, + "loss": 0.6801, + "step": 4447 + }, + { + "epoch": 0.5348403775626767, + "grad_norm": 3.318767592101517, + "learning_rate": 1.8715581318650765e-06, + "loss": 0.7857, + "step": 4448 + }, + { + "epoch": 0.5349606204533157, + "grad_norm": 2.2936378555532166, + "learning_rate": 1.8707807780068422e-06, + "loss": 0.8211, + "step": 4449 + }, + { + "epoch": 0.5350808633439548, + "grad_norm": 2.034882686335912, + "learning_rate": 1.8700034437510611e-06, + "loss": 0.6609, + "step": 4450 + }, + { + "epoch": 0.5352011062345938, + "grad_norm": 2.428518621287308, + "learning_rate": 1.8692261292156549e-06, + "loss": 0.8242, + "step": 4451 + }, + { + "epoch": 0.535321349125233, + "grad_norm": 1.9676110742569464, + "learning_rate": 1.8684488345185401e-06, + "loss": 0.8074, + "step": 4452 + }, + { + "epoch": 0.535441592015872, + "grad_norm": 2.195941658412372, + "learning_rate": 1.8676715597776332e-06, + "loss": 0.7904, + "step": 4453 + }, + { + "epoch": 0.5355618349065111, + "grad_norm": 1.7407051930882451, + "learning_rate": 1.8668943051108455e-06, + "loss": 0.764, + "step": 4454 + }, + { + "epoch": 0.5356820777971503, + "grad_norm": 1.766265820366915, + "learning_rate": 1.8661170706360856e-06, + "loss": 0.7664, + "step": 4455 + }, + { + "epoch": 0.5358023206877893, + "grad_norm": 2.259881678496833, + "learning_rate": 1.8653398564712594e-06, + "loss": 0.8191, + "step": 4456 + }, + { + "epoch": 0.5359225635784284, + "grad_norm": 1.7171671427942343, + "learning_rate": 1.8645626627342704e-06, + "loss": 0.8247, + "step": 4457 + }, + { + "epoch": 0.5360428064690675, + "grad_norm": 2.859500236061648, + "learning_rate": 1.8637854895430172e-06, + "loss": 0.8138, + "step": 4458 + }, + { + "epoch": 0.5361630493597066, + "grad_norm": 2.863941652502793, + "learning_rate": 1.8630083370153978e-06, + "loss": 0.7014, + "step": 4459 + }, + { + "epoch": 0.5362832922503457, + "grad_norm": 0.758552190268826, + "learning_rate": 1.8622312052693041e-06, + "loss": 0.5547, + "step": 4460 + }, + { + "epoch": 0.5364035351409848, + "grad_norm": 4.286640421390033, + "learning_rate": 1.8614540944226267e-06, + "loss": 0.724, + "step": 4461 + }, + { + "epoch": 0.5365237780316239, + "grad_norm": 1.9402451909844554, + "learning_rate": 1.8606770045932537e-06, + "loss": 0.6827, + "step": 4462 + }, + { + "epoch": 0.5366440209222629, + "grad_norm": 1.8543992986009603, + "learning_rate": 1.859899935899068e-06, + "loss": 0.8227, + "step": 4463 + }, + { + "epoch": 0.5367642638129021, + "grad_norm": 1.874739183723572, + "learning_rate": 1.8591228884579506e-06, + "loss": 0.7925, + "step": 4464 + }, + { + "epoch": 0.5368845067035412, + "grad_norm": 2.275277458290973, + "learning_rate": 1.8583458623877795e-06, + "loss": 0.8215, + "step": 4465 + }, + { + "epoch": 0.5370047495941802, + "grad_norm": 1.7669997138264397, + "learning_rate": 1.8575688578064281e-06, + "loss": 0.7454, + "step": 4466 + }, + { + "epoch": 0.5371249924848194, + "grad_norm": 1.8934586152861947, + "learning_rate": 1.8567918748317674e-06, + "loss": 0.7691, + "step": 4467 + }, + { + "epoch": 0.5372452353754584, + "grad_norm": 7.732408308194085, + "learning_rate": 1.8560149135816659e-06, + "loss": 0.832, + "step": 4468 + }, + { + "epoch": 0.5373654782660975, + "grad_norm": 2.44516185747423, + "learning_rate": 1.8552379741739873e-06, + "loss": 0.8503, + "step": 4469 + }, + { + "epoch": 0.5374857211567367, + "grad_norm": 0.9380525232384724, + "learning_rate": 1.8544610567265935e-06, + "loss": 0.5581, + "step": 4470 + }, + { + "epoch": 0.5376059640473757, + "grad_norm": 2.332210872512073, + "learning_rate": 1.853684161357341e-06, + "loss": 0.839, + "step": 4471 + }, + { + "epoch": 0.5377262069380148, + "grad_norm": 1.847431709565685, + "learning_rate": 1.852907288184085e-06, + "loss": 0.7748, + "step": 4472 + }, + { + "epoch": 0.5378464498286539, + "grad_norm": 7.057892503299747, + "learning_rate": 1.8521304373246762e-06, + "loss": 0.7068, + "step": 4473 + }, + { + "epoch": 0.537966692719293, + "grad_norm": 2.457213766478436, + "learning_rate": 1.8513536088969626e-06, + "loss": 0.8953, + "step": 4474 + }, + { + "epoch": 0.538086935609932, + "grad_norm": 1.7985714409314286, + "learning_rate": 1.8505768030187884e-06, + "loss": 0.81, + "step": 4475 + }, + { + "epoch": 0.5382071785005712, + "grad_norm": 1.5833570411647235, + "learning_rate": 1.849800019807995e-06, + "loss": 0.8064, + "step": 4476 + }, + { + "epoch": 0.5383274213912103, + "grad_norm": 2.8511557929142963, + "learning_rate": 1.8490232593824186e-06, + "loss": 0.7122, + "step": 4477 + }, + { + "epoch": 0.5384476642818493, + "grad_norm": 1.6124799091359072, + "learning_rate": 1.8482465218598935e-06, + "loss": 0.8511, + "step": 4478 + }, + { + "epoch": 0.5385679071724885, + "grad_norm": 2.019873702567195, + "learning_rate": 1.8474698073582508e-06, + "loss": 0.8387, + "step": 4479 + }, + { + "epoch": 0.5386881500631275, + "grad_norm": 2.2192152368337372, + "learning_rate": 1.8466931159953166e-06, + "loss": 0.8753, + "step": 4480 + }, + { + "epoch": 0.5388083929537666, + "grad_norm": 1.8959715097677048, + "learning_rate": 1.8459164478889158e-06, + "loss": 0.8424, + "step": 4481 + }, + { + "epoch": 0.5389286358444056, + "grad_norm": 1.6986534844630086, + "learning_rate": 1.8451398031568663e-06, + "loss": 0.7616, + "step": 4482 + }, + { + "epoch": 0.5390488787350448, + "grad_norm": 1.5660564383985374, + "learning_rate": 1.844363181916986e-06, + "loss": 0.7513, + "step": 4483 + }, + { + "epoch": 0.5391691216256839, + "grad_norm": 2.031229107089065, + "learning_rate": 1.8435865842870868e-06, + "loss": 0.8328, + "step": 4484 + }, + { + "epoch": 0.5392893645163229, + "grad_norm": 1.8895261108863548, + "learning_rate": 1.8428100103849787e-06, + "loss": 0.7258, + "step": 4485 + }, + { + "epoch": 0.5394096074069621, + "grad_norm": 2.0899115150374272, + "learning_rate": 1.842033460328467e-06, + "loss": 0.7356, + "step": 4486 + }, + { + "epoch": 0.5395298502976011, + "grad_norm": 1.6974262473218227, + "learning_rate": 1.8412569342353541e-06, + "loss": 0.7526, + "step": 4487 + }, + { + "epoch": 0.5396500931882402, + "grad_norm": 2.015026743027488, + "learning_rate": 1.840480432223438e-06, + "loss": 0.8557, + "step": 4488 + }, + { + "epoch": 0.5397703360788794, + "grad_norm": 1.853725730958349, + "learning_rate": 1.8397039544105131e-06, + "loss": 0.7791, + "step": 4489 + }, + { + "epoch": 0.5398905789695184, + "grad_norm": 1.6557083683761582, + "learning_rate": 1.8389275009143711e-06, + "loss": 0.7016, + "step": 4490 + }, + { + "epoch": 0.5400108218601575, + "grad_norm": 1.8765436176299761, + "learning_rate": 1.8381510718527988e-06, + "loss": 0.7379, + "step": 4491 + }, + { + "epoch": 0.5401310647507966, + "grad_norm": 3.814823130736054, + "learning_rate": 1.8373746673435812e-06, + "loss": 0.6385, + "step": 4492 + }, + { + "epoch": 0.5402513076414357, + "grad_norm": 1.7584561674172208, + "learning_rate": 1.8365982875044964e-06, + "loss": 0.7931, + "step": 4493 + }, + { + "epoch": 0.5403715505320748, + "grad_norm": 2.4095080228863144, + "learning_rate": 1.8358219324533217e-06, + "loss": 0.7616, + "step": 4494 + }, + { + "epoch": 0.5404917934227139, + "grad_norm": 1.8010935481556096, + "learning_rate": 1.8350456023078292e-06, + "loss": 0.7064, + "step": 4495 + }, + { + "epoch": 0.540612036313353, + "grad_norm": 2.097367648653756, + "learning_rate": 1.8342692971857874e-06, + "loss": 0.7845, + "step": 4496 + }, + { + "epoch": 0.540732279203992, + "grad_norm": 2.7366349826268768, + "learning_rate": 1.833493017204962e-06, + "loss": 0.7153, + "step": 4497 + }, + { + "epoch": 0.5408525220946312, + "grad_norm": 1.823377787078926, + "learning_rate": 1.8327167624831134e-06, + "loss": 0.7818, + "step": 4498 + }, + { + "epoch": 0.5409727649852702, + "grad_norm": 1.9661827302885375, + "learning_rate": 1.831940533137999e-06, + "loss": 0.7124, + "step": 4499 + }, + { + "epoch": 0.5410930078759093, + "grad_norm": 1.6992616143794321, + "learning_rate": 1.8311643292873718e-06, + "loss": 0.7235, + "step": 4500 + }, + { + "epoch": 0.5412132507665485, + "grad_norm": 1.863847370635968, + "learning_rate": 1.8303881510489818e-06, + "loss": 0.8789, + "step": 4501 + }, + { + "epoch": 0.5413334936571875, + "grad_norm": 2.5542187741945197, + "learning_rate": 1.829611998540574e-06, + "loss": 0.6947, + "step": 4502 + }, + { + "epoch": 0.5414537365478266, + "grad_norm": 1.8208165596451977, + "learning_rate": 1.8288358718798914e-06, + "loss": 0.7991, + "step": 4503 + }, + { + "epoch": 0.5415739794384657, + "grad_norm": 1.7342343659990591, + "learning_rate": 1.8280597711846703e-06, + "loss": 0.7273, + "step": 4504 + }, + { + "epoch": 0.5416942223291048, + "grad_norm": 3.4215247963543787, + "learning_rate": 1.8272836965726455e-06, + "loss": 0.8379, + "step": 4505 + }, + { + "epoch": 0.5418144652197439, + "grad_norm": 2.041144583246566, + "learning_rate": 1.8265076481615461e-06, + "loss": 0.7872, + "step": 4506 + }, + { + "epoch": 0.541934708110383, + "grad_norm": 7.813948840264536, + "learning_rate": 1.8257316260690987e-06, + "loss": 0.8816, + "step": 4507 + }, + { + "epoch": 0.5420549510010221, + "grad_norm": 1.93129839049518, + "learning_rate": 1.8249556304130254e-06, + "loss": 0.7635, + "step": 4508 + }, + { + "epoch": 0.5421751938916611, + "grad_norm": 2.1983824201786053, + "learning_rate": 1.824179661311044e-06, + "loss": 0.6969, + "step": 4509 + }, + { + "epoch": 0.5422954367823003, + "grad_norm": 1.9874716798696004, + "learning_rate": 1.823403718880868e-06, + "loss": 0.8014, + "step": 4510 + }, + { + "epoch": 0.5424156796729394, + "grad_norm": 2.260045409294954, + "learning_rate": 1.822627803240207e-06, + "loss": 0.6631, + "step": 4511 + }, + { + "epoch": 0.5425359225635784, + "grad_norm": 2.0055838361291896, + "learning_rate": 1.8218519145067675e-06, + "loss": 0.8562, + "step": 4512 + }, + { + "epoch": 0.5426561654542175, + "grad_norm": 1.789773991131577, + "learning_rate": 1.8210760527982508e-06, + "loss": 0.8957, + "step": 4513 + }, + { + "epoch": 0.5427764083448566, + "grad_norm": 2.0386272565143972, + "learning_rate": 1.8203002182323552e-06, + "loss": 0.7525, + "step": 4514 + }, + { + "epoch": 0.5428966512354957, + "grad_norm": 1.9256091553588075, + "learning_rate": 1.819524410926773e-06, + "loss": 0.7619, + "step": 4515 + }, + { + "epoch": 0.5430168941261347, + "grad_norm": 1.540571760453188, + "learning_rate": 1.8187486309991944e-06, + "loss": 0.7725, + "step": 4516 + }, + { + "epoch": 0.5431371370167739, + "grad_norm": 1.8785016740553389, + "learning_rate": 1.817972878567304e-06, + "loss": 0.7762, + "step": 4517 + }, + { + "epoch": 0.543257379907413, + "grad_norm": 1.7639029553203782, + "learning_rate": 1.8171971537487834e-06, + "loss": 0.7666, + "step": 4518 + }, + { + "epoch": 0.543377622798052, + "grad_norm": 1.9328748811663667, + "learning_rate": 1.8164214566613093e-06, + "loss": 0.8115, + "step": 4519 + }, + { + "epoch": 0.5434978656886912, + "grad_norm": 2.7104034337496032, + "learning_rate": 1.8156457874225547e-06, + "loss": 0.6638, + "step": 4520 + }, + { + "epoch": 0.5436181085793302, + "grad_norm": 1.7989482031888377, + "learning_rate": 1.814870146150187e-06, + "loss": 0.8086, + "step": 4521 + }, + { + "epoch": 0.5437383514699693, + "grad_norm": 2.1044373542740504, + "learning_rate": 1.814094532961871e-06, + "loss": 0.7907, + "step": 4522 + }, + { + "epoch": 0.5438585943606085, + "grad_norm": 2.0083781236056946, + "learning_rate": 1.8133189479752666e-06, + "loss": 0.8389, + "step": 4523 + }, + { + "epoch": 0.5439788372512475, + "grad_norm": 1.8985919654077439, + "learning_rate": 1.8125433913080292e-06, + "loss": 0.8249, + "step": 4524 + }, + { + "epoch": 0.5440990801418866, + "grad_norm": 2.697179175558405, + "learning_rate": 1.811767863077811e-06, + "loss": 0.8454, + "step": 4525 + }, + { + "epoch": 0.5442193230325257, + "grad_norm": 1.8713618267641141, + "learning_rate": 1.8109923634022577e-06, + "loss": 0.786, + "step": 4526 + }, + { + "epoch": 0.5443395659231648, + "grad_norm": 2.2652991751164353, + "learning_rate": 1.8102168923990128e-06, + "loss": 0.8688, + "step": 4527 + }, + { + "epoch": 0.5444598088138038, + "grad_norm": 1.7650381817888834, + "learning_rate": 1.809441450185714e-06, + "loss": 0.8003, + "step": 4528 + }, + { + "epoch": 0.544580051704443, + "grad_norm": 2.155796169122381, + "learning_rate": 1.8086660368799958e-06, + "loss": 0.7325, + "step": 4529 + }, + { + "epoch": 0.5447002945950821, + "grad_norm": 1.6257416426578595, + "learning_rate": 1.807890652599488e-06, + "loss": 0.7766, + "step": 4530 + }, + { + "epoch": 0.5448205374857211, + "grad_norm": 2.399504681973143, + "learning_rate": 1.8071152974618156e-06, + "loss": 0.8271, + "step": 4531 + }, + { + "epoch": 0.5449407803763603, + "grad_norm": 2.3741271774241035, + "learning_rate": 1.806339971584599e-06, + "loss": 0.7867, + "step": 4532 + }, + { + "epoch": 0.5450610232669993, + "grad_norm": 1.9800549437866157, + "learning_rate": 1.8055646750854546e-06, + "loss": 0.8571, + "step": 4533 + }, + { + "epoch": 0.5451812661576384, + "grad_norm": 2.3113459207488933, + "learning_rate": 1.8047894080819945e-06, + "loss": 0.8205, + "step": 4534 + }, + { + "epoch": 0.5453015090482776, + "grad_norm": 0.7444545454159237, + "learning_rate": 1.8040141706918258e-06, + "loss": 0.6322, + "step": 4535 + }, + { + "epoch": 0.5454217519389166, + "grad_norm": 2.9251180184837167, + "learning_rate": 1.8032389630325525e-06, + "loss": 0.7711, + "step": 4536 + }, + { + "epoch": 0.5455419948295557, + "grad_norm": 2.298516549748505, + "learning_rate": 1.8024637852217707e-06, + "loss": 0.7573, + "step": 4537 + }, + { + "epoch": 0.5456622377201948, + "grad_norm": 2.0226035737817742, + "learning_rate": 1.8016886373770766e-06, + "loss": 0.851, + "step": 4538 + }, + { + "epoch": 0.5457824806108339, + "grad_norm": 1.7296496103714774, + "learning_rate": 1.8009135196160579e-06, + "loss": 0.7954, + "step": 4539 + }, + { + "epoch": 0.545902723501473, + "grad_norm": 1.7533216967879477, + "learning_rate": 1.8001384320563e-06, + "loss": 0.8442, + "step": 4540 + }, + { + "epoch": 0.5460229663921121, + "grad_norm": 0.7855290528964132, + "learning_rate": 1.7993633748153833e-06, + "loss": 0.578, + "step": 4541 + }, + { + "epoch": 0.5461432092827512, + "grad_norm": 1.9437663969961279, + "learning_rate": 1.7985883480108834e-06, + "loss": 0.7285, + "step": 4542 + }, + { + "epoch": 0.5462634521733902, + "grad_norm": 1.6134065667127175, + "learning_rate": 1.797813351760371e-06, + "loss": 0.7238, + "step": 4543 + }, + { + "epoch": 0.5463836950640293, + "grad_norm": 1.752049268042864, + "learning_rate": 1.7970383861814116e-06, + "loss": 0.787, + "step": 4544 + }, + { + "epoch": 0.5465039379546685, + "grad_norm": 2.058426967864357, + "learning_rate": 1.7962634513915684e-06, + "loss": 0.7438, + "step": 4545 + }, + { + "epoch": 0.5466241808453075, + "grad_norm": 1.807938986584289, + "learning_rate": 1.7954885475083969e-06, + "loss": 0.7959, + "step": 4546 + }, + { + "epoch": 0.5467444237359466, + "grad_norm": 2.4013733431321063, + "learning_rate": 1.7947136746494513e-06, + "loss": 0.7336, + "step": 4547 + }, + { + "epoch": 0.5468646666265857, + "grad_norm": 2.654521065311338, + "learning_rate": 1.793938832932277e-06, + "loss": 0.884, + "step": 4548 + }, + { + "epoch": 0.5469849095172248, + "grad_norm": 2.070751215677297, + "learning_rate": 1.7931640224744185e-06, + "loss": 0.7056, + "step": 4549 + }, + { + "epoch": 0.5471051524078638, + "grad_norm": 1.5071576529688957, + "learning_rate": 1.7923892433934127e-06, + "loss": 0.7389, + "step": 4550 + }, + { + "epoch": 0.547225395298503, + "grad_norm": 1.7235691826073842, + "learning_rate": 1.7916144958067939e-06, + "loss": 0.7882, + "step": 4551 + }, + { + "epoch": 0.5473456381891421, + "grad_norm": 1.722940918806703, + "learning_rate": 1.7908397798320905e-06, + "loss": 0.7896, + "step": 4552 + }, + { + "epoch": 0.5474658810797811, + "grad_norm": 1.9672564942105724, + "learning_rate": 1.7900650955868265e-06, + "loss": 0.756, + "step": 4553 + }, + { + "epoch": 0.5475861239704203, + "grad_norm": 1.706557987348971, + "learning_rate": 1.7892904431885202e-06, + "loss": 0.769, + "step": 4554 + }, + { + "epoch": 0.5477063668610593, + "grad_norm": 2.0516715887254953, + "learning_rate": 1.788515822754686e-06, + "loss": 0.7541, + "step": 4555 + }, + { + "epoch": 0.5478266097516984, + "grad_norm": 2.139910279291627, + "learning_rate": 1.7877412344028335e-06, + "loss": 0.7845, + "step": 4556 + }, + { + "epoch": 0.5479468526423376, + "grad_norm": 2.2589770419037882, + "learning_rate": 1.7869666782504668e-06, + "loss": 0.7813, + "step": 4557 + }, + { + "epoch": 0.5480670955329766, + "grad_norm": 2.295460583381025, + "learning_rate": 1.7861921544150867e-06, + "loss": 0.6921, + "step": 4558 + }, + { + "epoch": 0.5481873384236157, + "grad_norm": 2.108382606181879, + "learning_rate": 1.7854176630141856e-06, + "loss": 0.7696, + "step": 4559 + }, + { + "epoch": 0.5483075813142548, + "grad_norm": 2.602765878953021, + "learning_rate": 1.784643204165255e-06, + "loss": 0.8464, + "step": 4560 + }, + { + "epoch": 0.5484278242048939, + "grad_norm": 2.4380391820586422, + "learning_rate": 1.7838687779857783e-06, + "loss": 0.7761, + "step": 4561 + }, + { + "epoch": 0.5485480670955329, + "grad_norm": 1.9980574949696381, + "learning_rate": 1.7830943845932366e-06, + "loss": 0.6471, + "step": 4562 + }, + { + "epoch": 0.5486683099861721, + "grad_norm": 1.617141206972531, + "learning_rate": 1.7823200241051044e-06, + "loss": 0.7553, + "step": 4563 + }, + { + "epoch": 0.5487885528768112, + "grad_norm": 2.4579383619012423, + "learning_rate": 1.7815456966388513e-06, + "loss": 0.8132, + "step": 4564 + }, + { + "epoch": 0.5489087957674502, + "grad_norm": 2.2562214350279084, + "learning_rate": 1.780771402311943e-06, + "loss": 0.8187, + "step": 4565 + }, + { + "epoch": 0.5490290386580894, + "grad_norm": 2.2125240604794976, + "learning_rate": 1.7799971412418374e-06, + "loss": 0.7929, + "step": 4566 + }, + { + "epoch": 0.5491492815487284, + "grad_norm": 1.8663671758496803, + "learning_rate": 1.7792229135459918e-06, + "loss": 0.7388, + "step": 4567 + }, + { + "epoch": 0.5492695244393675, + "grad_norm": 0.7468154771780424, + "learning_rate": 1.7784487193418538e-06, + "loss": 0.6179, + "step": 4568 + }, + { + "epoch": 0.5493897673300067, + "grad_norm": 3.134345309876595, + "learning_rate": 1.7776745587468698e-06, + "loss": 0.615, + "step": 4569 + }, + { + "epoch": 0.5495100102206457, + "grad_norm": 2.237654087152708, + "learning_rate": 1.7769004318784776e-06, + "loss": 0.8215, + "step": 4570 + }, + { + "epoch": 0.5496302531112848, + "grad_norm": 1.7034496033035063, + "learning_rate": 1.776126338854113e-06, + "loss": 0.8064, + "step": 4571 + }, + { + "epoch": 0.5497504960019239, + "grad_norm": 1.837435626504796, + "learning_rate": 1.7753522797912044e-06, + "loss": 0.8517, + "step": 4572 + }, + { + "epoch": 0.549870738892563, + "grad_norm": 2.1310656873369993, + "learning_rate": 1.7745782548071765e-06, + "loss": 0.7051, + "step": 4573 + }, + { + "epoch": 0.549990981783202, + "grad_norm": 1.9386108044473485, + "learning_rate": 1.7738042640194482e-06, + "loss": 0.7459, + "step": 4574 + }, + { + "epoch": 0.5501112246738411, + "grad_norm": 1.6340216581106888, + "learning_rate": 1.7730303075454335e-06, + "loss": 0.7056, + "step": 4575 + }, + { + "epoch": 0.5502314675644803, + "grad_norm": 1.9331226121682368, + "learning_rate": 1.7722563855025402e-06, + "loss": 0.8494, + "step": 4576 + }, + { + "epoch": 0.5503517104551193, + "grad_norm": 2.39432165028188, + "learning_rate": 1.7714824980081721e-06, + "loss": 0.716, + "step": 4577 + }, + { + "epoch": 0.5504719533457584, + "grad_norm": 1.8663654512949697, + "learning_rate": 1.7707086451797276e-06, + "loss": 0.7465, + "step": 4578 + }, + { + "epoch": 0.5505921962363975, + "grad_norm": 0.7121479503166872, + "learning_rate": 1.7699348271345993e-06, + "loss": 0.5231, + "step": 4579 + }, + { + "epoch": 0.5507124391270366, + "grad_norm": 0.7173612280303213, + "learning_rate": 1.7691610439901753e-06, + "loss": 0.5454, + "step": 4580 + }, + { + "epoch": 0.5508326820176757, + "grad_norm": 1.7357138200341014, + "learning_rate": 1.7683872958638367e-06, + "loss": 0.7583, + "step": 4581 + }, + { + "epoch": 0.5509529249083148, + "grad_norm": 2.0631869934808043, + "learning_rate": 1.7676135828729614e-06, + "loss": 0.8427, + "step": 4582 + }, + { + "epoch": 0.5510731677989539, + "grad_norm": 1.9450395384939947, + "learning_rate": 1.7668399051349205e-06, + "loss": 0.8361, + "step": 4583 + }, + { + "epoch": 0.5511934106895929, + "grad_norm": 2.1110456844140892, + "learning_rate": 1.766066262767081e-06, + "loss": 0.8368, + "step": 4584 + }, + { + "epoch": 0.5513136535802321, + "grad_norm": 2.08668636066946, + "learning_rate": 1.765292655886803e-06, + "loss": 0.7777, + "step": 4585 + }, + { + "epoch": 0.5514338964708712, + "grad_norm": 2.142480867137965, + "learning_rate": 1.764519084611443e-06, + "loss": 0.7094, + "step": 4586 + }, + { + "epoch": 0.5515541393615102, + "grad_norm": 1.9849537883923702, + "learning_rate": 1.7637455490583505e-06, + "loss": 0.7812, + "step": 4587 + }, + { + "epoch": 0.5516743822521494, + "grad_norm": 2.019908995136368, + "learning_rate": 1.7629720493448701e-06, + "loss": 0.774, + "step": 4588 + }, + { + "epoch": 0.5517946251427884, + "grad_norm": 2.0048470175734834, + "learning_rate": 1.7621985855883418e-06, + "loss": 0.8532, + "step": 4589 + }, + { + "epoch": 0.5519148680334275, + "grad_norm": 1.8152368862790624, + "learning_rate": 1.7614251579060983e-06, + "loss": 0.7309, + "step": 4590 + }, + { + "epoch": 0.5520351109240667, + "grad_norm": 1.7931087663619711, + "learning_rate": 1.76065176641547e-06, + "loss": 0.849, + "step": 4591 + }, + { + "epoch": 0.5521553538147057, + "grad_norm": 2.201884360032296, + "learning_rate": 1.759878411233777e-06, + "loss": 0.7777, + "step": 4592 + }, + { + "epoch": 0.5522755967053448, + "grad_norm": 2.1874669753715286, + "learning_rate": 1.7591050924783388e-06, + "loss": 0.7648, + "step": 4593 + }, + { + "epoch": 0.5523958395959839, + "grad_norm": 0.8428383599975585, + "learning_rate": 1.7583318102664661e-06, + "loss": 0.5794, + "step": 4594 + }, + { + "epoch": 0.552516082486623, + "grad_norm": 1.9332826306282576, + "learning_rate": 1.757558564715466e-06, + "loss": 0.7932, + "step": 4595 + }, + { + "epoch": 0.552636325377262, + "grad_norm": 2.66454554678396, + "learning_rate": 1.7567853559426386e-06, + "loss": 0.7463, + "step": 4596 + }, + { + "epoch": 0.5527565682679012, + "grad_norm": 3.158474280064986, + "learning_rate": 1.7560121840652797e-06, + "loss": 0.7593, + "step": 4597 + }, + { + "epoch": 0.5528768111585403, + "grad_norm": 3.78873755251288, + "learning_rate": 1.7552390492006782e-06, + "loss": 0.7029, + "step": 4598 + }, + { + "epoch": 0.5529970540491793, + "grad_norm": 5.381684870322401, + "learning_rate": 1.7544659514661184e-06, + "loss": 0.6535, + "step": 4599 + }, + { + "epoch": 0.5531172969398185, + "grad_norm": 3.5967028884077243, + "learning_rate": 1.7536928909788786e-06, + "loss": 0.8005, + "step": 4600 + }, + { + "epoch": 0.5532375398304575, + "grad_norm": 0.8939070876993923, + "learning_rate": 1.752919867856231e-06, + "loss": 0.6192, + "step": 4601 + }, + { + "epoch": 0.5533577827210966, + "grad_norm": 1.7345024955557813, + "learning_rate": 1.7521468822154436e-06, + "loss": 0.791, + "step": 4602 + }, + { + "epoch": 0.5534780256117358, + "grad_norm": 1.7081196193840629, + "learning_rate": 1.751373934173777e-06, + "loss": 0.7501, + "step": 4603 + }, + { + "epoch": 0.5535982685023748, + "grad_norm": 1.6318769192985665, + "learning_rate": 1.750601023848487e-06, + "loss": 0.7385, + "step": 4604 + }, + { + "epoch": 0.5537185113930139, + "grad_norm": 2.3479923762988912, + "learning_rate": 1.749828151356823e-06, + "loss": 0.7434, + "step": 4605 + }, + { + "epoch": 0.553838754283653, + "grad_norm": 2.7918507719301635, + "learning_rate": 1.7490553168160297e-06, + "loss": 0.7586, + "step": 4606 + }, + { + "epoch": 0.5539589971742921, + "grad_norm": 2.0275923442389754, + "learning_rate": 1.748282520343345e-06, + "loss": 0.7653, + "step": 4607 + }, + { + "epoch": 0.5540792400649311, + "grad_norm": 2.2992009433441774, + "learning_rate": 1.7475097620560023e-06, + "loss": 0.7894, + "step": 4608 + }, + { + "epoch": 0.5541994829555702, + "grad_norm": 1.7623584879886234, + "learning_rate": 1.746737042071228e-06, + "loss": 0.7135, + "step": 4609 + }, + { + "epoch": 0.5543197258462094, + "grad_norm": 1.803756738069867, + "learning_rate": 1.7459643605062424e-06, + "loss": 0.7955, + "step": 4610 + }, + { + "epoch": 0.5544399687368484, + "grad_norm": 1.6058462028733371, + "learning_rate": 1.745191717478262e-06, + "loss": 0.8124, + "step": 4611 + }, + { + "epoch": 0.5545602116274875, + "grad_norm": 1.9701434456936118, + "learning_rate": 1.7444191131044948e-06, + "loss": 0.8011, + "step": 4612 + }, + { + "epoch": 0.5546804545181266, + "grad_norm": 2.406648875282584, + "learning_rate": 1.7436465475021456e-06, + "loss": 0.7338, + "step": 4613 + }, + { + "epoch": 0.5548006974087657, + "grad_norm": 2.643682481701688, + "learning_rate": 1.7428740207884111e-06, + "loss": 0.7184, + "step": 4614 + }, + { + "epoch": 0.5549209402994048, + "grad_norm": 1.6964276421336446, + "learning_rate": 1.7421015330804833e-06, + "loss": 0.6161, + "step": 4615 + }, + { + "epoch": 0.5550411831900439, + "grad_norm": 5.056032735827255, + "learning_rate": 1.7413290844955475e-06, + "loss": 0.7265, + "step": 4616 + }, + { + "epoch": 0.555161426080683, + "grad_norm": 1.9361527588808807, + "learning_rate": 1.7405566751507843e-06, + "loss": 0.7845, + "step": 4617 + }, + { + "epoch": 0.555281668971322, + "grad_norm": 1.777969820193401, + "learning_rate": 1.7397843051633668e-06, + "loss": 0.6807, + "step": 4618 + }, + { + "epoch": 0.5554019118619612, + "grad_norm": 1.9863266846418082, + "learning_rate": 1.739011974650464e-06, + "loss": 0.7194, + "step": 4619 + }, + { + "epoch": 0.5555221547526003, + "grad_norm": 2.371246584345359, + "learning_rate": 1.7382396837292365e-06, + "loss": 0.7717, + "step": 4620 + }, + { + "epoch": 0.5556423976432393, + "grad_norm": 1.7638471311430666, + "learning_rate": 1.737467432516841e-06, + "loss": 0.7356, + "step": 4621 + }, + { + "epoch": 0.5557626405338785, + "grad_norm": 2.610086127135358, + "learning_rate": 1.7366952211304274e-06, + "loss": 0.7539, + "step": 4622 + }, + { + "epoch": 0.5558828834245175, + "grad_norm": 2.2018101874370246, + "learning_rate": 1.735923049687139e-06, + "loss": 0.8417, + "step": 4623 + }, + { + "epoch": 0.5560031263151566, + "grad_norm": 1.582637868362655, + "learning_rate": 1.7351509183041144e-06, + "loss": 0.7406, + "step": 4624 + }, + { + "epoch": 0.5561233692057957, + "grad_norm": 1.992942995916628, + "learning_rate": 1.7343788270984852e-06, + "loss": 0.7209, + "step": 4625 + }, + { + "epoch": 0.5562436120964348, + "grad_norm": 2.031118300643391, + "learning_rate": 1.7336067761873764e-06, + "loss": 0.751, + "step": 4626 + }, + { + "epoch": 0.5563638549870739, + "grad_norm": 1.9463009448724713, + "learning_rate": 1.7328347656879076e-06, + "loss": 0.759, + "step": 4627 + }, + { + "epoch": 0.556484097877713, + "grad_norm": 2.3526006795843117, + "learning_rate": 1.7320627957171927e-06, + "loss": 0.6872, + "step": 4628 + }, + { + "epoch": 0.5566043407683521, + "grad_norm": 2.6516191394446045, + "learning_rate": 1.7312908663923382e-06, + "loss": 0.8179, + "step": 4629 + }, + { + "epoch": 0.5567245836589911, + "grad_norm": 1.9889298675738774, + "learning_rate": 1.7305189778304463e-06, + "loss": 0.6811, + "step": 4630 + }, + { + "epoch": 0.5568448265496303, + "grad_norm": 2.0503569827133226, + "learning_rate": 1.729747130148611e-06, + "loss": 0.8032, + "step": 4631 + }, + { + "epoch": 0.5569650694402694, + "grad_norm": 2.242859317736834, + "learning_rate": 1.7289753234639208e-06, + "loss": 0.7727, + "step": 4632 + }, + { + "epoch": 0.5570853123309084, + "grad_norm": 1.9068326841227174, + "learning_rate": 1.7282035578934592e-06, + "loss": 0.7687, + "step": 4633 + }, + { + "epoch": 0.5572055552215476, + "grad_norm": 1.7335083316837119, + "learning_rate": 1.727431833554301e-06, + "loss": 0.7896, + "step": 4634 + }, + { + "epoch": 0.5573257981121866, + "grad_norm": 1.9406881865751147, + "learning_rate": 1.7266601505635175e-06, + "loss": 0.7741, + "step": 4635 + }, + { + "epoch": 0.5574460410028257, + "grad_norm": 2.193797393026063, + "learning_rate": 1.7258885090381717e-06, + "loss": 0.7614, + "step": 4636 + }, + { + "epoch": 0.5575662838934649, + "grad_norm": 2.75839633543165, + "learning_rate": 1.7251169090953213e-06, + "loss": 0.7942, + "step": 4637 + }, + { + "epoch": 0.5576865267841039, + "grad_norm": 2.6186700521934103, + "learning_rate": 1.7243453508520168e-06, + "loss": 0.7775, + "step": 4638 + }, + { + "epoch": 0.557806769674743, + "grad_norm": 2.2826638217058606, + "learning_rate": 1.7235738344253038e-06, + "loss": 0.8522, + "step": 4639 + }, + { + "epoch": 0.557927012565382, + "grad_norm": 1.772964817862043, + "learning_rate": 1.72280235993222e-06, + "loss": 0.8255, + "step": 4640 + }, + { + "epoch": 0.5580472554560212, + "grad_norm": 2.4076484480711695, + "learning_rate": 1.722030927489798e-06, + "loss": 0.6974, + "step": 4641 + }, + { + "epoch": 0.5581674983466602, + "grad_norm": 1.716166687656825, + "learning_rate": 1.7212595372150634e-06, + "loss": 0.7462, + "step": 4642 + }, + { + "epoch": 0.5582877412372993, + "grad_norm": 2.397913275969555, + "learning_rate": 1.720488189225035e-06, + "loss": 0.7385, + "step": 4643 + }, + { + "epoch": 0.5584079841279385, + "grad_norm": 2.3578574245210198, + "learning_rate": 1.7197168836367265e-06, + "loss": 0.8004, + "step": 4644 + }, + { + "epoch": 0.5585282270185775, + "grad_norm": 1.7402684885913533, + "learning_rate": 1.7189456205671433e-06, + "loss": 0.8233, + "step": 4645 + }, + { + "epoch": 0.5586484699092166, + "grad_norm": 6.557905641575563, + "learning_rate": 1.7181744001332866e-06, + "loss": 0.8273, + "step": 4646 + }, + { + "epoch": 0.5587687127998557, + "grad_norm": 1.9924914558256168, + "learning_rate": 1.7174032224521493e-06, + "loss": 0.6343, + "step": 4647 + }, + { + "epoch": 0.5588889556904948, + "grad_norm": 1.654721760597753, + "learning_rate": 1.7166320876407184e-06, + "loss": 0.7027, + "step": 4648 + }, + { + "epoch": 0.5590091985811338, + "grad_norm": 1.9664342672892934, + "learning_rate": 1.7158609958159742e-06, + "loss": 0.6783, + "step": 4649 + }, + { + "epoch": 0.559129441471773, + "grad_norm": 1.9951951006568402, + "learning_rate": 1.7150899470948911e-06, + "loss": 0.7849, + "step": 4650 + }, + { + "epoch": 0.5592496843624121, + "grad_norm": 0.8166215238615424, + "learning_rate": 1.7143189415944365e-06, + "loss": 0.5668, + "step": 4651 + }, + { + "epoch": 0.5593699272530511, + "grad_norm": 1.9217758773425155, + "learning_rate": 1.7135479794315714e-06, + "loss": 0.7642, + "step": 4652 + }, + { + "epoch": 0.5594901701436903, + "grad_norm": 1.8089968271306747, + "learning_rate": 1.7127770607232502e-06, + "loss": 0.7925, + "step": 4653 + }, + { + "epoch": 0.5596104130343293, + "grad_norm": 2.0500200875391354, + "learning_rate": 1.7120061855864204e-06, + "loss": 0.8029, + "step": 4654 + }, + { + "epoch": 0.5597306559249684, + "grad_norm": 1.91454225094137, + "learning_rate": 1.7112353541380233e-06, + "loss": 0.7123, + "step": 4655 + }, + { + "epoch": 0.5598508988156076, + "grad_norm": 1.569552716971864, + "learning_rate": 1.7104645664949931e-06, + "loss": 0.7249, + "step": 4656 + }, + { + "epoch": 0.5599711417062466, + "grad_norm": 1.780336580815032, + "learning_rate": 1.7096938227742584e-06, + "loss": 0.7228, + "step": 4657 + }, + { + "epoch": 0.5600913845968857, + "grad_norm": 2.318561959675016, + "learning_rate": 1.70892312309274e-06, + "loss": 0.8475, + "step": 4658 + }, + { + "epoch": 0.5602116274875248, + "grad_norm": 1.935478879040259, + "learning_rate": 1.7081524675673523e-06, + "loss": 0.6847, + "step": 4659 + }, + { + "epoch": 0.5603318703781639, + "grad_norm": 0.7813243830556628, + "learning_rate": 1.7073818563150026e-06, + "loss": 0.597, + "step": 4660 + }, + { + "epoch": 0.560452113268803, + "grad_norm": 2.0571844789290727, + "learning_rate": 1.7066112894525935e-06, + "loss": 0.8688, + "step": 4661 + }, + { + "epoch": 0.5605723561594421, + "grad_norm": 1.8268478366478065, + "learning_rate": 1.7058407670970177e-06, + "loss": 0.7315, + "step": 4662 + }, + { + "epoch": 0.5606925990500812, + "grad_norm": 1.7173026493191352, + "learning_rate": 1.7050702893651643e-06, + "loss": 0.6145, + "step": 4663 + }, + { + "epoch": 0.5608128419407202, + "grad_norm": 2.264108189555537, + "learning_rate": 1.7042998563739134e-06, + "loss": 0.7579, + "step": 4664 + }, + { + "epoch": 0.5609330848313594, + "grad_norm": 2.0337715343795955, + "learning_rate": 1.703529468240139e-06, + "loss": 0.7217, + "step": 4665 + }, + { + "epoch": 0.5610533277219985, + "grad_norm": 2.3479158127881052, + "learning_rate": 1.7027591250807088e-06, + "loss": 0.7481, + "step": 4666 + }, + { + "epoch": 0.5611735706126375, + "grad_norm": 2.1801287023367317, + "learning_rate": 1.7019888270124825e-06, + "loss": 0.8482, + "step": 4667 + }, + { + "epoch": 0.5612938135032767, + "grad_norm": 4.010818157493614, + "learning_rate": 1.7012185741523147e-06, + "loss": 0.8221, + "step": 4668 + }, + { + "epoch": 0.5614140563939157, + "grad_norm": 2.011432634765992, + "learning_rate": 1.7004483666170514e-06, + "loss": 0.6265, + "step": 4669 + }, + { + "epoch": 0.5615342992845548, + "grad_norm": 1.9865082815685613, + "learning_rate": 1.699678204523533e-06, + "loss": 0.8144, + "step": 4670 + }, + { + "epoch": 0.5616545421751938, + "grad_norm": 2.8235198405281228, + "learning_rate": 1.6989080879885918e-06, + "loss": 0.6857, + "step": 4671 + }, + { + "epoch": 0.561774785065833, + "grad_norm": 0.9354183291878543, + "learning_rate": 1.6981380171290544e-06, + "loss": 0.6104, + "step": 4672 + }, + { + "epoch": 0.5618950279564721, + "grad_norm": 1.8535669478454948, + "learning_rate": 1.6973679920617396e-06, + "loss": 0.7471, + "step": 4673 + }, + { + "epoch": 0.5620152708471111, + "grad_norm": 2.7326539645222803, + "learning_rate": 1.6965980129034603e-06, + "loss": 0.8525, + "step": 4674 + }, + { + "epoch": 0.5621355137377503, + "grad_norm": 1.714782713757522, + "learning_rate": 1.6958280797710209e-06, + "loss": 0.769, + "step": 4675 + }, + { + "epoch": 0.5622557566283893, + "grad_norm": 0.729328346492844, + "learning_rate": 1.6950581927812198e-06, + "loss": 0.5478, + "step": 4676 + }, + { + "epoch": 0.5623759995190284, + "grad_norm": 1.9917454247750708, + "learning_rate": 1.6942883520508486e-06, + "loss": 0.7982, + "step": 4677 + }, + { + "epoch": 0.5624962424096676, + "grad_norm": 1.8949058555628913, + "learning_rate": 1.693518557696691e-06, + "loss": 0.7749, + "step": 4678 + }, + { + "epoch": 0.5626164853003066, + "grad_norm": 2.1949684172076513, + "learning_rate": 1.6927488098355252e-06, + "loss": 0.8936, + "step": 4679 + }, + { + "epoch": 0.5627367281909457, + "grad_norm": 0.9522833625855899, + "learning_rate": 1.6919791085841201e-06, + "loss": 0.6321, + "step": 4680 + }, + { + "epoch": 0.5628569710815848, + "grad_norm": 2.2295821209754942, + "learning_rate": 1.6912094540592396e-06, + "loss": 0.7925, + "step": 4681 + }, + { + "epoch": 0.5629772139722239, + "grad_norm": 2.7334272213449635, + "learning_rate": 1.6904398463776393e-06, + "loss": 0.8211, + "step": 4682 + }, + { + "epoch": 0.5630974568628629, + "grad_norm": 2.275064941150075, + "learning_rate": 1.6896702856560683e-06, + "loss": 0.7284, + "step": 4683 + }, + { + "epoch": 0.5632176997535021, + "grad_norm": 2.6654282813698478, + "learning_rate": 1.6889007720112677e-06, + "loss": 0.6984, + "step": 4684 + }, + { + "epoch": 0.5633379426441412, + "grad_norm": 2.4134752861798137, + "learning_rate": 1.6881313055599734e-06, + "loss": 0.7738, + "step": 4685 + }, + { + "epoch": 0.5634581855347802, + "grad_norm": 2.8659758541264, + "learning_rate": 1.6873618864189117e-06, + "loss": 0.8264, + "step": 4686 + }, + { + "epoch": 0.5635784284254194, + "grad_norm": 5.412587591083616, + "learning_rate": 1.686592514704803e-06, + "loss": 0.7881, + "step": 4687 + }, + { + "epoch": 0.5636986713160584, + "grad_norm": 2.411912077583907, + "learning_rate": 1.685823190534361e-06, + "loss": 0.7073, + "step": 4688 + }, + { + "epoch": 0.5638189142066975, + "grad_norm": 2.0728842830205108, + "learning_rate": 1.6850539140242907e-06, + "loss": 0.8435, + "step": 4689 + }, + { + "epoch": 0.5639391570973367, + "grad_norm": 1.9724505931047753, + "learning_rate": 1.684284685291292e-06, + "loss": 0.8271, + "step": 4690 + }, + { + "epoch": 0.5640593999879757, + "grad_norm": 1.9306446744015757, + "learning_rate": 1.683515504452055e-06, + "loss": 0.8171, + "step": 4691 + }, + { + "epoch": 0.5641796428786148, + "grad_norm": 1.5919411061478108, + "learning_rate": 1.6827463716232648e-06, + "loss": 0.6678, + "step": 4692 + }, + { + "epoch": 0.5642998857692539, + "grad_norm": 1.9628778713438284, + "learning_rate": 1.6819772869215972e-06, + "loss": 0.7621, + "step": 4693 + }, + { + "epoch": 0.564420128659893, + "grad_norm": 1.7673251313463252, + "learning_rate": 1.6812082504637228e-06, + "loss": 0.8223, + "step": 4694 + }, + { + "epoch": 0.564540371550532, + "grad_norm": 1.506080541931739, + "learning_rate": 1.6804392623663025e-06, + "loss": 0.7439, + "step": 4695 + }, + { + "epoch": 0.5646606144411712, + "grad_norm": 2.0904603993119095, + "learning_rate": 1.6796703227459935e-06, + "loss": 0.7837, + "step": 4696 + }, + { + "epoch": 0.5647808573318103, + "grad_norm": 1.8666863187822644, + "learning_rate": 1.6789014317194407e-06, + "loss": 0.7633, + "step": 4697 + }, + { + "epoch": 0.5649011002224493, + "grad_norm": 2.1383164480627426, + "learning_rate": 1.6781325894032853e-06, + "loss": 0.7289, + "step": 4698 + }, + { + "epoch": 0.5650213431130885, + "grad_norm": 2.420873127105236, + "learning_rate": 1.6773637959141608e-06, + "loss": 0.9195, + "step": 4699 + }, + { + "epoch": 0.5651415860037275, + "grad_norm": 2.153325959282562, + "learning_rate": 1.6765950513686915e-06, + "loss": 0.6699, + "step": 4700 + }, + { + "epoch": 0.5652618288943666, + "grad_norm": 1.751192367879688, + "learning_rate": 1.675826355883496e-06, + "loss": 0.769, + "step": 4701 + }, + { + "epoch": 0.5653820717850057, + "grad_norm": 1.8786381710676376, + "learning_rate": 1.6750577095751848e-06, + "loss": 0.8003, + "step": 4702 + }, + { + "epoch": 0.5655023146756448, + "grad_norm": 2.1270285628911374, + "learning_rate": 1.6742891125603605e-06, + "loss": 0.7309, + "step": 4703 + }, + { + "epoch": 0.5656225575662839, + "grad_norm": 2.363540206467127, + "learning_rate": 1.6735205649556185e-06, + "loss": 0.7206, + "step": 4704 + }, + { + "epoch": 0.5657428004569229, + "grad_norm": 1.6582581923184814, + "learning_rate": 1.6727520668775476e-06, + "loss": 0.8501, + "step": 4705 + }, + { + "epoch": 0.5658630433475621, + "grad_norm": 1.6781073656345484, + "learning_rate": 1.6719836184427275e-06, + "loss": 0.7586, + "step": 4706 + }, + { + "epoch": 0.5659832862382012, + "grad_norm": 1.8031315557139207, + "learning_rate": 1.671215219767733e-06, + "loss": 0.6442, + "step": 4707 + }, + { + "epoch": 0.5661035291288402, + "grad_norm": 1.9374352721200223, + "learning_rate": 1.670446870969127e-06, + "loss": 0.7668, + "step": 4708 + }, + { + "epoch": 0.5662237720194794, + "grad_norm": 4.621482955265609, + "learning_rate": 1.6696785721634685e-06, + "loss": 0.8019, + "step": 4709 + }, + { + "epoch": 0.5663440149101184, + "grad_norm": 1.9053980690272143, + "learning_rate": 1.6689103234673086e-06, + "loss": 0.7415, + "step": 4710 + }, + { + "epoch": 0.5664642578007575, + "grad_norm": 2.653528951529726, + "learning_rate": 1.668142124997189e-06, + "loss": 0.7799, + "step": 4711 + }, + { + "epoch": 0.5665845006913967, + "grad_norm": 0.7368280145679964, + "learning_rate": 1.6673739768696453e-06, + "loss": 0.5991, + "step": 4712 + }, + { + "epoch": 0.5667047435820357, + "grad_norm": 2.314387504838481, + "learning_rate": 1.6666058792012052e-06, + "loss": 0.7777, + "step": 4713 + }, + { + "epoch": 0.5668249864726748, + "grad_norm": 0.919417062571282, + "learning_rate": 1.6658378321083878e-06, + "loss": 0.6887, + "step": 4714 + }, + { + "epoch": 0.5669452293633139, + "grad_norm": 2.3384719419103166, + "learning_rate": 1.6650698357077055e-06, + "loss": 0.833, + "step": 4715 + }, + { + "epoch": 0.567065472253953, + "grad_norm": 2.434992160353681, + "learning_rate": 1.6643018901156632e-06, + "loss": 0.8228, + "step": 4716 + }, + { + "epoch": 0.567185715144592, + "grad_norm": 2.4656344685535876, + "learning_rate": 1.6635339954487566e-06, + "loss": 0.7977, + "step": 4717 + }, + { + "epoch": 0.5673059580352312, + "grad_norm": 2.5094312155615595, + "learning_rate": 1.6627661518234765e-06, + "loss": 0.8215, + "step": 4718 + }, + { + "epoch": 0.5674262009258703, + "grad_norm": 1.6817820419768128, + "learning_rate": 1.661998359356302e-06, + "loss": 0.8576, + "step": 4719 + }, + { + "epoch": 0.5675464438165093, + "grad_norm": 0.7644313148346243, + "learning_rate": 1.6612306181637077e-06, + "loss": 0.5589, + "step": 4720 + }, + { + "epoch": 0.5676666867071485, + "grad_norm": 2.879123715762576, + "learning_rate": 1.6604629283621598e-06, + "loss": 0.6602, + "step": 4721 + }, + { + "epoch": 0.5677869295977875, + "grad_norm": 2.149690131102509, + "learning_rate": 1.6596952900681152e-06, + "loss": 0.7453, + "step": 4722 + }, + { + "epoch": 0.5679071724884266, + "grad_norm": 2.213757660743214, + "learning_rate": 1.658927703398025e-06, + "loss": 0.8198, + "step": 4723 + }, + { + "epoch": 0.5680274153790658, + "grad_norm": 2.2769408613926045, + "learning_rate": 1.6581601684683309e-06, + "loss": 0.7867, + "step": 4724 + }, + { + "epoch": 0.5681476582697048, + "grad_norm": 2.5766181935154906, + "learning_rate": 1.6573926853954674e-06, + "loss": 0.6905, + "step": 4725 + }, + { + "epoch": 0.5682679011603439, + "grad_norm": 1.8059579580151262, + "learning_rate": 1.6566252542958608e-06, + "loss": 0.8337, + "step": 4726 + }, + { + "epoch": 0.568388144050983, + "grad_norm": 1.869695471482496, + "learning_rate": 1.6558578752859305e-06, + "loss": 0.7857, + "step": 4727 + }, + { + "epoch": 0.5685083869416221, + "grad_norm": 1.8497289923041977, + "learning_rate": 1.6550905484820865e-06, + "loss": 0.7928, + "step": 4728 + }, + { + "epoch": 0.5686286298322611, + "grad_norm": 4.9766027427974535, + "learning_rate": 1.6543232740007328e-06, + "loss": 0.7927, + "step": 4729 + }, + { + "epoch": 0.5687488727229003, + "grad_norm": 2.1699345105041092, + "learning_rate": 1.653556051958263e-06, + "loss": 0.6757, + "step": 4730 + }, + { + "epoch": 0.5688691156135394, + "grad_norm": 1.7347749815913796, + "learning_rate": 1.6527888824710642e-06, + "loss": 0.743, + "step": 4731 + }, + { + "epoch": 0.5689893585041784, + "grad_norm": 2.839074787623468, + "learning_rate": 1.6520217656555166e-06, + "loss": 0.7703, + "step": 4732 + }, + { + "epoch": 0.5691096013948175, + "grad_norm": 1.7451486099052702, + "learning_rate": 1.65125470162799e-06, + "loss": 0.7128, + "step": 4733 + }, + { + "epoch": 0.5692298442854566, + "grad_norm": 3.098061226039687, + "learning_rate": 1.6504876905048485e-06, + "loss": 0.6984, + "step": 4734 + }, + { + "epoch": 0.5693500871760957, + "grad_norm": 2.3065425325446, + "learning_rate": 1.6497207324024464e-06, + "loss": 0.723, + "step": 4735 + }, + { + "epoch": 0.5694703300667348, + "grad_norm": 2.7708691377166894, + "learning_rate": 1.6489538274371305e-06, + "loss": 0.8286, + "step": 4736 + }, + { + "epoch": 0.5695905729573739, + "grad_norm": 2.111098426089184, + "learning_rate": 1.6481869757252396e-06, + "loss": 0.8353, + "step": 4737 + }, + { + "epoch": 0.569710815848013, + "grad_norm": 1.6401538717378576, + "learning_rate": 1.647420177383105e-06, + "loss": 0.7224, + "step": 4738 + }, + { + "epoch": 0.569831058738652, + "grad_norm": 1.8867874922777141, + "learning_rate": 1.646653432527049e-06, + "loss": 0.7279, + "step": 4739 + }, + { + "epoch": 0.5699513016292912, + "grad_norm": 1.5011434964811285, + "learning_rate": 1.645886741273387e-06, + "loss": 0.747, + "step": 4740 + }, + { + "epoch": 0.5700715445199303, + "grad_norm": 1.9653920191737786, + "learning_rate": 1.645120103738424e-06, + "loss": 0.7404, + "step": 4741 + }, + { + "epoch": 0.5701917874105693, + "grad_norm": 2.3480684296223586, + "learning_rate": 1.6443535200384591e-06, + "loss": 0.8479, + "step": 4742 + }, + { + "epoch": 0.5703120303012085, + "grad_norm": 1.8533595253559578, + "learning_rate": 1.6435869902897827e-06, + "loss": 0.7087, + "step": 4743 + }, + { + "epoch": 0.5704322731918475, + "grad_norm": 0.7992712904141102, + "learning_rate": 1.6428205146086764e-06, + "loss": 0.6215, + "step": 4744 + }, + { + "epoch": 0.5705525160824866, + "grad_norm": 1.5589652706190533, + "learning_rate": 1.6420540931114142e-06, + "loss": 0.71, + "step": 4745 + }, + { + "epoch": 0.5706727589731257, + "grad_norm": 2.2169705629575382, + "learning_rate": 1.6412877259142616e-06, + "loss": 0.798, + "step": 4746 + }, + { + "epoch": 0.5707930018637648, + "grad_norm": 2.061679676953369, + "learning_rate": 1.6405214131334757e-06, + "loss": 0.7458, + "step": 4747 + }, + { + "epoch": 0.5709132447544039, + "grad_norm": 1.9749201408368657, + "learning_rate": 1.6397551548853052e-06, + "loss": 0.8032, + "step": 4748 + }, + { + "epoch": 0.571033487645043, + "grad_norm": 1.6966824951008124, + "learning_rate": 1.6389889512859917e-06, + "loss": 0.7125, + "step": 4749 + }, + { + "epoch": 0.5711537305356821, + "grad_norm": 0.8188149360939047, + "learning_rate": 1.638222802451767e-06, + "loss": 0.6037, + "step": 4750 + }, + { + "epoch": 0.5712739734263211, + "grad_norm": 1.7818403102424285, + "learning_rate": 1.6374567084988561e-06, + "loss": 0.7588, + "step": 4751 + }, + { + "epoch": 0.5713942163169603, + "grad_norm": 1.7022058649054628, + "learning_rate": 1.6366906695434738e-06, + "loss": 0.7699, + "step": 4752 + }, + { + "epoch": 0.5715144592075994, + "grad_norm": 1.9725436641177798, + "learning_rate": 1.6359246857018275e-06, + "loss": 0.8599, + "step": 4753 + }, + { + "epoch": 0.5716347020982384, + "grad_norm": 2.5517163278774824, + "learning_rate": 1.6351587570901178e-06, + "loss": 0.7842, + "step": 4754 + }, + { + "epoch": 0.5717549449888776, + "grad_norm": 2.530187313096694, + "learning_rate": 1.634392883824534e-06, + "loss": 0.7618, + "step": 4755 + }, + { + "epoch": 0.5718751878795166, + "grad_norm": 3.97250777035548, + "learning_rate": 1.6336270660212595e-06, + "loss": 0.6844, + "step": 4756 + }, + { + "epoch": 0.5719954307701557, + "grad_norm": 2.047537077776465, + "learning_rate": 1.6328613037964676e-06, + "loss": 0.6729, + "step": 4757 + }, + { + "epoch": 0.5721156736607949, + "grad_norm": 2.0101945454596435, + "learning_rate": 1.6320955972663241e-06, + "loss": 0.6863, + "step": 4758 + }, + { + "epoch": 0.5722359165514339, + "grad_norm": 3.9123575300413296, + "learning_rate": 1.6313299465469857e-06, + "loss": 0.6558, + "step": 4759 + }, + { + "epoch": 0.572356159442073, + "grad_norm": 2.6045824456496907, + "learning_rate": 1.6305643517546014e-06, + "loss": 0.8058, + "step": 4760 + }, + { + "epoch": 0.5724764023327121, + "grad_norm": 1.971417330725632, + "learning_rate": 1.629798813005311e-06, + "loss": 0.8497, + "step": 4761 + }, + { + "epoch": 0.5725966452233512, + "grad_norm": 2.2174235932648894, + "learning_rate": 1.6290333304152473e-06, + "loss": 0.7144, + "step": 4762 + }, + { + "epoch": 0.5727168881139902, + "grad_norm": 1.7543948709947834, + "learning_rate": 1.6282679041005314e-06, + "loss": 0.5643, + "step": 4763 + }, + { + "epoch": 0.5728371310046293, + "grad_norm": 2.218037343081481, + "learning_rate": 1.6275025341772789e-06, + "loss": 0.8734, + "step": 4764 + }, + { + "epoch": 0.5729573738952685, + "grad_norm": 2.5150083176947406, + "learning_rate": 1.626737220761596e-06, + "loss": 0.8222, + "step": 4765 + }, + { + "epoch": 0.5730776167859075, + "grad_norm": 2.03649881490162, + "learning_rate": 1.62597196396958e-06, + "loss": 0.7944, + "step": 4766 + }, + { + "epoch": 0.5731978596765466, + "grad_norm": 1.7902659404051897, + "learning_rate": 1.6252067639173197e-06, + "loss": 0.8558, + "step": 4767 + }, + { + "epoch": 0.5733181025671857, + "grad_norm": 2.1407585380835656, + "learning_rate": 1.6244416207208956e-06, + "loss": 0.706, + "step": 4768 + }, + { + "epoch": 0.5734383454578248, + "grad_norm": 1.918455912763736, + "learning_rate": 1.6236765344963787e-06, + "loss": 0.7358, + "step": 4769 + }, + { + "epoch": 0.5735585883484638, + "grad_norm": 2.751586629749837, + "learning_rate": 1.6229115053598322e-06, + "loss": 0.7005, + "step": 4770 + }, + { + "epoch": 0.573678831239103, + "grad_norm": 1.8922799803659152, + "learning_rate": 1.6221465334273108e-06, + "loss": 0.7329, + "step": 4771 + }, + { + "epoch": 0.5737990741297421, + "grad_norm": 1.9730389214421273, + "learning_rate": 1.6213816188148593e-06, + "loss": 0.6147, + "step": 4772 + }, + { + "epoch": 0.5739193170203811, + "grad_norm": 1.7802948649629304, + "learning_rate": 1.6206167616385162e-06, + "loss": 0.7761, + "step": 4773 + }, + { + "epoch": 0.5740395599110203, + "grad_norm": 2.2119832615417367, + "learning_rate": 1.6198519620143078e-06, + "loss": 0.7431, + "step": 4774 + }, + { + "epoch": 0.5741598028016593, + "grad_norm": 1.5277496344931554, + "learning_rate": 1.6190872200582546e-06, + "loss": 0.7858, + "step": 4775 + }, + { + "epoch": 0.5742800456922984, + "grad_norm": 2.088854746449335, + "learning_rate": 1.6183225358863676e-06, + "loss": 0.7816, + "step": 4776 + }, + { + "epoch": 0.5744002885829376, + "grad_norm": 2.0603151019308634, + "learning_rate": 1.617557909614648e-06, + "loss": 0.7188, + "step": 4777 + }, + { + "epoch": 0.5745205314735766, + "grad_norm": 3.1594024626824, + "learning_rate": 1.6167933413590899e-06, + "loss": 0.8634, + "step": 4778 + }, + { + "epoch": 0.5746407743642157, + "grad_norm": 2.3870765430338103, + "learning_rate": 1.6160288312356773e-06, + "loss": 0.9114, + "step": 4779 + }, + { + "epoch": 0.5747610172548548, + "grad_norm": 1.8493941603812505, + "learning_rate": 1.6152643793603857e-06, + "loss": 0.8203, + "step": 4780 + }, + { + "epoch": 0.5748812601454939, + "grad_norm": 1.7760420843069828, + "learning_rate": 1.6144999858491815e-06, + "loss": 0.8799, + "step": 4781 + }, + { + "epoch": 0.575001503036133, + "grad_norm": 1.8108496551667546, + "learning_rate": 1.6137356508180232e-06, + "loss": 0.8603, + "step": 4782 + }, + { + "epoch": 0.5751217459267721, + "grad_norm": 1.847979204853519, + "learning_rate": 1.6129713743828593e-06, + "loss": 0.8152, + "step": 4783 + }, + { + "epoch": 0.5752419888174112, + "grad_norm": 1.667279424877032, + "learning_rate": 1.6122071566596306e-06, + "loss": 0.7489, + "step": 4784 + }, + { + "epoch": 0.5753622317080502, + "grad_norm": 2.3454420975186108, + "learning_rate": 1.6114429977642674e-06, + "loss": 0.8406, + "step": 4785 + }, + { + "epoch": 0.5754824745986894, + "grad_norm": 1.7992437919964375, + "learning_rate": 1.6106788978126926e-06, + "loss": 0.7401, + "step": 4786 + }, + { + "epoch": 0.5756027174893285, + "grad_norm": 2.1833658662464, + "learning_rate": 1.6099148569208196e-06, + "loss": 0.7948, + "step": 4787 + }, + { + "epoch": 0.5757229603799675, + "grad_norm": 1.663986705791113, + "learning_rate": 1.6091508752045523e-06, + "loss": 0.642, + "step": 4788 + }, + { + "epoch": 0.5758432032706067, + "grad_norm": 1.6791801218609135, + "learning_rate": 1.608386952779787e-06, + "loss": 0.8691, + "step": 4789 + }, + { + "epoch": 0.5759634461612457, + "grad_norm": 1.5659154565432374, + "learning_rate": 1.6076230897624098e-06, + "loss": 0.7488, + "step": 4790 + }, + { + "epoch": 0.5760836890518848, + "grad_norm": 1.978120272903459, + "learning_rate": 1.6068592862682974e-06, + "loss": 0.7763, + "step": 4791 + }, + { + "epoch": 0.576203931942524, + "grad_norm": 2.0434063357100136, + "learning_rate": 1.6060955424133187e-06, + "loss": 0.7398, + "step": 4792 + }, + { + "epoch": 0.576324174833163, + "grad_norm": 1.8543535916976546, + "learning_rate": 1.6053318583133332e-06, + "loss": 0.9008, + "step": 4793 + }, + { + "epoch": 0.5764444177238021, + "grad_norm": 2.2252465357979028, + "learning_rate": 1.6045682340841907e-06, + "loss": 0.7608, + "step": 4794 + }, + { + "epoch": 0.5765646606144411, + "grad_norm": 0.7728337715574921, + "learning_rate": 1.6038046698417336e-06, + "loss": 0.581, + "step": 4795 + }, + { + "epoch": 0.5766849035050803, + "grad_norm": 2.4009649522376995, + "learning_rate": 1.6030411657017919e-06, + "loss": 0.6934, + "step": 4796 + }, + { + "epoch": 0.5768051463957193, + "grad_norm": 1.8341517427915628, + "learning_rate": 1.6022777217801903e-06, + "loss": 0.8456, + "step": 4797 + }, + { + "epoch": 0.5769253892863584, + "grad_norm": 2.175600218989282, + "learning_rate": 1.601514338192742e-06, + "loss": 0.7419, + "step": 4798 + }, + { + "epoch": 0.5770456321769976, + "grad_norm": 2.1189945856216914, + "learning_rate": 1.6007510150552514e-06, + "loss": 0.7284, + "step": 4799 + }, + { + "epoch": 0.5771658750676366, + "grad_norm": 1.6110035332392965, + "learning_rate": 1.599987752483515e-06, + "loss": 0.6258, + "step": 4800 + }, + { + "epoch": 0.5772861179582757, + "grad_norm": 1.814590038311991, + "learning_rate": 1.5992245505933184e-06, + "loss": 0.6848, + "step": 4801 + }, + { + "epoch": 0.5774063608489148, + "grad_norm": 1.9924386857194172, + "learning_rate": 1.5984614095004388e-06, + "loss": 0.7135, + "step": 4802 + }, + { + "epoch": 0.5775266037395539, + "grad_norm": 2.106654921189091, + "learning_rate": 1.5976983293206438e-06, + "loss": 0.8096, + "step": 4803 + }, + { + "epoch": 0.577646846630193, + "grad_norm": 1.742641783190534, + "learning_rate": 1.5969353101696928e-06, + "loss": 0.7153, + "step": 4804 + }, + { + "epoch": 0.5777670895208321, + "grad_norm": 1.6821258589186838, + "learning_rate": 1.5961723521633341e-06, + "loss": 0.8014, + "step": 4805 + }, + { + "epoch": 0.5778873324114712, + "grad_norm": 2.2438342065337924, + "learning_rate": 1.5954094554173097e-06, + "loss": 0.9219, + "step": 4806 + }, + { + "epoch": 0.5780075753021102, + "grad_norm": 2.386650521509697, + "learning_rate": 1.5946466200473482e-06, + "loss": 0.7943, + "step": 4807 + }, + { + "epoch": 0.5781278181927494, + "grad_norm": 2.509861759467191, + "learning_rate": 1.5938838461691723e-06, + "loss": 0.8346, + "step": 4808 + }, + { + "epoch": 0.5782480610833884, + "grad_norm": 2.288652298663396, + "learning_rate": 1.593121133898494e-06, + "loss": 0.8376, + "step": 4809 + }, + { + "epoch": 0.5783683039740275, + "grad_norm": 2.2141150835101917, + "learning_rate": 1.592358483351016e-06, + "loss": 0.7976, + "step": 4810 + }, + { + "epoch": 0.5784885468646667, + "grad_norm": 2.13526657941657, + "learning_rate": 1.5915958946424326e-06, + "loss": 0.7252, + "step": 4811 + }, + { + "epoch": 0.5786087897553057, + "grad_norm": 1.5807308173969663, + "learning_rate": 1.5908333678884271e-06, + "loss": 0.7464, + "step": 4812 + }, + { + "epoch": 0.5787290326459448, + "grad_norm": 2.260025998531452, + "learning_rate": 1.5900709032046743e-06, + "loss": 0.744, + "step": 4813 + }, + { + "epoch": 0.5788492755365839, + "grad_norm": 2.0441746261862637, + "learning_rate": 1.5893085007068391e-06, + "loss": 0.783, + "step": 4814 + }, + { + "epoch": 0.578969518427223, + "grad_norm": 1.902236966811778, + "learning_rate": 1.5885461605105786e-06, + "loss": 0.7155, + "step": 4815 + }, + { + "epoch": 0.579089761317862, + "grad_norm": 1.874609270709878, + "learning_rate": 1.5877838827315375e-06, + "loss": 0.7698, + "step": 4816 + }, + { + "epoch": 0.5792100042085012, + "grad_norm": 1.8372557756740902, + "learning_rate": 1.587021667485355e-06, + "loss": 0.708, + "step": 4817 + }, + { + "epoch": 0.5793302470991403, + "grad_norm": 1.921675757027945, + "learning_rate": 1.5862595148876559e-06, + "loss": 0.7886, + "step": 4818 + }, + { + "epoch": 0.5794504899897793, + "grad_norm": 1.9974640623638715, + "learning_rate": 1.58549742505406e-06, + "loss": 0.7643, + "step": 4819 + }, + { + "epoch": 0.5795707328804185, + "grad_norm": 2.0889215163575354, + "learning_rate": 1.5847353981001747e-06, + "loss": 0.7583, + "step": 4820 + }, + { + "epoch": 0.5796909757710575, + "grad_norm": 2.544921125527518, + "learning_rate": 1.5839734341415993e-06, + "loss": 0.7014, + "step": 4821 + }, + { + "epoch": 0.5798112186616966, + "grad_norm": 2.292482704754634, + "learning_rate": 1.5832115332939238e-06, + "loss": 0.771, + "step": 4822 + }, + { + "epoch": 0.5799314615523358, + "grad_norm": 2.43051457204023, + "learning_rate": 1.5824496956727272e-06, + "loss": 0.7515, + "step": 4823 + }, + { + "epoch": 0.5800517044429748, + "grad_norm": 1.879282194027501, + "learning_rate": 1.5816879213935797e-06, + "loss": 0.7344, + "step": 4824 + }, + { + "epoch": 0.5801719473336139, + "grad_norm": 1.5513026516422215, + "learning_rate": 1.5809262105720416e-06, + "loss": 0.7965, + "step": 4825 + }, + { + "epoch": 0.580292190224253, + "grad_norm": 1.6496588354263346, + "learning_rate": 1.5801645633236644e-06, + "loss": 0.7995, + "step": 4826 + }, + { + "epoch": 0.5804124331148921, + "grad_norm": 1.7766323803902904, + "learning_rate": 1.579402979763989e-06, + "loss": 0.7753, + "step": 4827 + }, + { + "epoch": 0.5805326760055312, + "grad_norm": 2.2995700724609285, + "learning_rate": 1.578641460008548e-06, + "loss": 0.8145, + "step": 4828 + }, + { + "epoch": 0.5806529188961702, + "grad_norm": 2.127656286216792, + "learning_rate": 1.5778800041728613e-06, + "loss": 0.6865, + "step": 4829 + }, + { + "epoch": 0.5807731617868094, + "grad_norm": 1.5103037912570358, + "learning_rate": 1.577118612372443e-06, + "loss": 0.6618, + "step": 4830 + }, + { + "epoch": 0.5808934046774484, + "grad_norm": 1.9726578816322151, + "learning_rate": 1.5763572847227943e-06, + "loss": 0.7092, + "step": 4831 + }, + { + "epoch": 0.5810136475680875, + "grad_norm": 1.825434865482775, + "learning_rate": 1.5755960213394091e-06, + "loss": 0.815, + "step": 4832 + }, + { + "epoch": 0.5811338904587267, + "grad_norm": 1.836567222870847, + "learning_rate": 1.5748348223377703e-06, + "loss": 0.7837, + "step": 4833 + }, + { + "epoch": 0.5812541333493657, + "grad_norm": 1.6527379785370857, + "learning_rate": 1.5740736878333507e-06, + "loss": 0.7816, + "step": 4834 + }, + { + "epoch": 0.5813743762400048, + "grad_norm": 3.551407929735224, + "learning_rate": 1.5733126179416143e-06, + "loss": 0.7788, + "step": 4835 + }, + { + "epoch": 0.5814946191306439, + "grad_norm": 1.8385430210818163, + "learning_rate": 1.5725516127780137e-06, + "loss": 0.732, + "step": 4836 + }, + { + "epoch": 0.581614862021283, + "grad_norm": 2.690421800549919, + "learning_rate": 1.5717906724579943e-06, + "loss": 0.8913, + "step": 4837 + }, + { + "epoch": 0.581735104911922, + "grad_norm": 2.7351278086199344, + "learning_rate": 1.571029797096989e-06, + "loss": 0.6887, + "step": 4838 + }, + { + "epoch": 0.5818553478025612, + "grad_norm": 2.274190771362592, + "learning_rate": 1.570268986810423e-06, + "loss": 0.7892, + "step": 4839 + }, + { + "epoch": 0.5819755906932003, + "grad_norm": 1.9626510853528993, + "learning_rate": 1.5695082417137096e-06, + "loss": 0.7488, + "step": 4840 + }, + { + "epoch": 0.5820958335838393, + "grad_norm": 1.8657673179234724, + "learning_rate": 1.5687475619222539e-06, + "loss": 0.7579, + "step": 4841 + }, + { + "epoch": 0.5822160764744785, + "grad_norm": 2.9554936028044896, + "learning_rate": 1.5679869475514496e-06, + "loss": 0.7412, + "step": 4842 + }, + { + "epoch": 0.5823363193651175, + "grad_norm": 2.2621719881303974, + "learning_rate": 1.567226398716682e-06, + "loss": 0.8176, + "step": 4843 + }, + { + "epoch": 0.5824565622557566, + "grad_norm": 1.9568079289919018, + "learning_rate": 1.566465915533326e-06, + "loss": 0.6194, + "step": 4844 + }, + { + "epoch": 0.5825768051463958, + "grad_norm": 1.9120463437844333, + "learning_rate": 1.5657054981167458e-06, + "loss": 0.8818, + "step": 4845 + }, + { + "epoch": 0.5826970480370348, + "grad_norm": 2.1602158495112223, + "learning_rate": 1.5649451465822965e-06, + "loss": 0.6857, + "step": 4846 + }, + { + "epoch": 0.5828172909276739, + "grad_norm": 2.6247126104073653, + "learning_rate": 1.5641848610453218e-06, + "loss": 0.837, + "step": 4847 + }, + { + "epoch": 0.582937533818313, + "grad_norm": 2.2359919532740276, + "learning_rate": 1.563424641621158e-06, + "loss": 0.8652, + "step": 4848 + }, + { + "epoch": 0.5830577767089521, + "grad_norm": 3.209611976439931, + "learning_rate": 1.5626644884251282e-06, + "loss": 0.7078, + "step": 4849 + }, + { + "epoch": 0.5831780195995911, + "grad_norm": 1.7126800860087994, + "learning_rate": 1.5619044015725488e-06, + "loss": 0.8858, + "step": 4850 + }, + { + "epoch": 0.5832982624902303, + "grad_norm": 3.090854561421621, + "learning_rate": 1.5611443811787224e-06, + "loss": 0.8768, + "step": 4851 + }, + { + "epoch": 0.5834185053808694, + "grad_norm": 2.1174604514436, + "learning_rate": 1.560384427358945e-06, + "loss": 0.6906, + "step": 4852 + }, + { + "epoch": 0.5835387482715084, + "grad_norm": 1.6908286432107495, + "learning_rate": 1.5596245402284998e-06, + "loss": 0.7364, + "step": 4853 + }, + { + "epoch": 0.5836589911621476, + "grad_norm": 1.7188935653375226, + "learning_rate": 1.5588647199026619e-06, + "loss": 0.8249, + "step": 4854 + }, + { + "epoch": 0.5837792340527866, + "grad_norm": 3.2003930327328303, + "learning_rate": 1.5581049664966956e-06, + "loss": 0.8798, + "step": 4855 + }, + { + "epoch": 0.5838994769434257, + "grad_norm": 0.9994855988205913, + "learning_rate": 1.5573452801258545e-06, + "loss": 0.6511, + "step": 4856 + }, + { + "epoch": 0.5840197198340649, + "grad_norm": 1.9063238536098526, + "learning_rate": 1.5565856609053824e-06, + "loss": 0.6394, + "step": 4857 + }, + { + "epoch": 0.5841399627247039, + "grad_norm": 2.4680931086543434, + "learning_rate": 1.5558261089505127e-06, + "loss": 0.8043, + "step": 4858 + }, + { + "epoch": 0.584260205615343, + "grad_norm": 2.084839263867733, + "learning_rate": 1.5550666243764697e-06, + "loss": 0.8039, + "step": 4859 + }, + { + "epoch": 0.584380448505982, + "grad_norm": 3.0934969095301295, + "learning_rate": 1.554307207298465e-06, + "loss": 0.7793, + "step": 4860 + }, + { + "epoch": 0.5845006913966212, + "grad_norm": 1.9583797111159946, + "learning_rate": 1.553547857831704e-06, + "loss": 0.7909, + "step": 4861 + }, + { + "epoch": 0.5846209342872603, + "grad_norm": 0.8982451523252671, + "learning_rate": 1.5527885760913771e-06, + "loss": 0.6418, + "step": 4862 + }, + { + "epoch": 0.5847411771778993, + "grad_norm": 2.2555875275866577, + "learning_rate": 1.552029362192668e-06, + "loss": 0.7674, + "step": 4863 + }, + { + "epoch": 0.5848614200685385, + "grad_norm": 1.8229363649303085, + "learning_rate": 1.5512702162507478e-06, + "loss": 0.7298, + "step": 4864 + }, + { + "epoch": 0.5849816629591775, + "grad_norm": 1.133529699134202, + "learning_rate": 1.5505111383807792e-06, + "loss": 0.5581, + "step": 4865 + }, + { + "epoch": 0.5851019058498166, + "grad_norm": 2.0390960573672476, + "learning_rate": 1.5497521286979138e-06, + "loss": 0.8072, + "step": 4866 + }, + { + "epoch": 0.5852221487404557, + "grad_norm": 2.655827028594021, + "learning_rate": 1.5489931873172927e-06, + "loss": 0.7522, + "step": 4867 + }, + { + "epoch": 0.5853423916310948, + "grad_norm": 1.7050442162426906, + "learning_rate": 1.5482343143540467e-06, + "loss": 0.7907, + "step": 4868 + }, + { + "epoch": 0.5854626345217339, + "grad_norm": 2.5043441222962226, + "learning_rate": 1.547475509923295e-06, + "loss": 0.8298, + "step": 4869 + }, + { + "epoch": 0.585582877412373, + "grad_norm": 0.7751302978747971, + "learning_rate": 1.5467167741401495e-06, + "loss": 0.5607, + "step": 4870 + }, + { + "epoch": 0.5857031203030121, + "grad_norm": 3.10625051579001, + "learning_rate": 1.5459581071197083e-06, + "loss": 0.7199, + "step": 4871 + }, + { + "epoch": 0.5858233631936511, + "grad_norm": 2.01110237826892, + "learning_rate": 1.5451995089770624e-06, + "loss": 0.8378, + "step": 4872 + }, + { + "epoch": 0.5859436060842903, + "grad_norm": 1.3392685525793695, + "learning_rate": 1.5444409798272885e-06, + "loss": 0.7199, + "step": 4873 + }, + { + "epoch": 0.5860638489749294, + "grad_norm": 1.9049196447611185, + "learning_rate": 1.543682519785456e-06, + "loss": 0.8144, + "step": 4874 + }, + { + "epoch": 0.5861840918655684, + "grad_norm": 2.2516993357304727, + "learning_rate": 1.5429241289666219e-06, + "loss": 0.8082, + "step": 4875 + }, + { + "epoch": 0.5863043347562076, + "grad_norm": 2.8703180420001275, + "learning_rate": 1.5421658074858342e-06, + "loss": 0.7005, + "step": 4876 + }, + { + "epoch": 0.5864245776468466, + "grad_norm": 2.3628246023823976, + "learning_rate": 1.5414075554581298e-06, + "loss": 0.6741, + "step": 4877 + }, + { + "epoch": 0.5865448205374857, + "grad_norm": 2.1594407067466777, + "learning_rate": 1.5406493729985348e-06, + "loss": 0.7868, + "step": 4878 + }, + { + "epoch": 0.5866650634281249, + "grad_norm": 1.9407362213451536, + "learning_rate": 1.5398912602220644e-06, + "loss": 0.7203, + "step": 4879 + }, + { + "epoch": 0.5867853063187639, + "grad_norm": 2.323915195677424, + "learning_rate": 1.539133217243724e-06, + "loss": 0.7886, + "step": 4880 + }, + { + "epoch": 0.586905549209403, + "grad_norm": 2.0969268837889965, + "learning_rate": 1.5383752441785081e-06, + "loss": 0.7708, + "step": 4881 + }, + { + "epoch": 0.5870257921000421, + "grad_norm": 2.1580068093334734, + "learning_rate": 1.5376173411414003e-06, + "loss": 0.8594, + "step": 4882 + }, + { + "epoch": 0.5871460349906812, + "grad_norm": 10.762578303354632, + "learning_rate": 1.5368595082473753e-06, + "loss": 0.7938, + "step": 4883 + }, + { + "epoch": 0.5872662778813202, + "grad_norm": 2.0062085584121587, + "learning_rate": 1.5361017456113935e-06, + "loss": 0.7858, + "step": 4884 + }, + { + "epoch": 0.5873865207719594, + "grad_norm": 2.1795527372094816, + "learning_rate": 1.5353440533484085e-06, + "loss": 0.8606, + "step": 4885 + }, + { + "epoch": 0.5875067636625985, + "grad_norm": 3.6523848750609145, + "learning_rate": 1.534586431573361e-06, + "loss": 0.6639, + "step": 4886 + }, + { + "epoch": 0.5876270065532375, + "grad_norm": 1.9243567345629837, + "learning_rate": 1.5338288804011817e-06, + "loss": 0.7958, + "step": 4887 + }, + { + "epoch": 0.5877472494438767, + "grad_norm": 1.8531408865699592, + "learning_rate": 1.533071399946791e-06, + "loss": 0.7142, + "step": 4888 + }, + { + "epoch": 0.5878674923345157, + "grad_norm": 1.8202702595144298, + "learning_rate": 1.5323139903250977e-06, + "loss": 0.5779, + "step": 4889 + }, + { + "epoch": 0.5879877352251548, + "grad_norm": 5.601959682625947, + "learning_rate": 1.5315566516510002e-06, + "loss": 0.7736, + "step": 4890 + }, + { + "epoch": 0.5881079781157939, + "grad_norm": 1.9271868978629014, + "learning_rate": 1.5307993840393857e-06, + "loss": 0.6832, + "step": 4891 + }, + { + "epoch": 0.588228221006433, + "grad_norm": 1.9231270534143652, + "learning_rate": 1.530042187605132e-06, + "loss": 0.8094, + "step": 4892 + }, + { + "epoch": 0.5883484638970721, + "grad_norm": 1.7991237335359644, + "learning_rate": 1.5292850624631044e-06, + "loss": 0.8438, + "step": 4893 + }, + { + "epoch": 0.5884687067877111, + "grad_norm": 1.8696841861645117, + "learning_rate": 1.5285280087281593e-06, + "loss": 0.8085, + "step": 4894 + }, + { + "epoch": 0.5885889496783503, + "grad_norm": 0.657390851690157, + "learning_rate": 1.5277710265151398e-06, + "loss": 0.5673, + "step": 4895 + }, + { + "epoch": 0.5887091925689893, + "grad_norm": 3.185187604280616, + "learning_rate": 1.5270141159388803e-06, + "loss": 0.779, + "step": 4896 + }, + { + "epoch": 0.5888294354596284, + "grad_norm": 1.6456103918323264, + "learning_rate": 1.526257277114203e-06, + "loss": 0.809, + "step": 4897 + }, + { + "epoch": 0.5889496783502676, + "grad_norm": 2.4618589583467685, + "learning_rate": 1.5255005101559201e-06, + "loss": 0.7969, + "step": 4898 + }, + { + "epoch": 0.5890699212409066, + "grad_norm": 2.3728847118589487, + "learning_rate": 1.524743815178833e-06, + "loss": 0.7688, + "step": 4899 + }, + { + "epoch": 0.5891901641315457, + "grad_norm": 1.8755653164909907, + "learning_rate": 1.5239871922977315e-06, + "loss": 0.811, + "step": 4900 + }, + { + "epoch": 0.5893104070221848, + "grad_norm": 2.066824800327552, + "learning_rate": 1.523230641627394e-06, + "loss": 0.9045, + "step": 4901 + }, + { + "epoch": 0.5894306499128239, + "grad_norm": 2.0745504512761643, + "learning_rate": 1.5224741632825888e-06, + "loss": 0.7299, + "step": 4902 + }, + { + "epoch": 0.589550892803463, + "grad_norm": 1.6527909199096882, + "learning_rate": 1.521717757378074e-06, + "loss": 0.6991, + "step": 4903 + }, + { + "epoch": 0.5896711356941021, + "grad_norm": 1.9378004764011005, + "learning_rate": 1.5209614240285943e-06, + "loss": 0.6951, + "step": 4904 + }, + { + "epoch": 0.5897913785847412, + "grad_norm": 2.3469453519189147, + "learning_rate": 1.520205163348887e-06, + "loss": 0.849, + "step": 4905 + }, + { + "epoch": 0.5899116214753802, + "grad_norm": 0.7353217332927381, + "learning_rate": 1.519448975453674e-06, + "loss": 0.5697, + "step": 4906 + }, + { + "epoch": 0.5900318643660194, + "grad_norm": 2.1664141605644094, + "learning_rate": 1.5186928604576696e-06, + "loss": 0.7674, + "step": 4907 + }, + { + "epoch": 0.5901521072566585, + "grad_norm": 2.16315195917162, + "learning_rate": 1.5179368184755752e-06, + "loss": 0.7761, + "step": 4908 + }, + { + "epoch": 0.5902723501472975, + "grad_norm": 1.5964659941576298, + "learning_rate": 1.5171808496220821e-06, + "loss": 0.8302, + "step": 4909 + }, + { + "epoch": 0.5903925930379367, + "grad_norm": 1.7239041814529759, + "learning_rate": 1.5164249540118708e-06, + "loss": 0.8218, + "step": 4910 + }, + { + "epoch": 0.5905128359285757, + "grad_norm": 1.950551221650711, + "learning_rate": 1.5156691317596093e-06, + "loss": 0.8344, + "step": 4911 + }, + { + "epoch": 0.5906330788192148, + "grad_norm": 2.080510885146456, + "learning_rate": 1.5149133829799556e-06, + "loss": 0.6726, + "step": 4912 + }, + { + "epoch": 0.590753321709854, + "grad_norm": 1.8826461655204756, + "learning_rate": 1.5141577077875556e-06, + "loss": 0.8134, + "step": 4913 + }, + { + "epoch": 0.590873564600493, + "grad_norm": 1.9306447978932646, + "learning_rate": 1.5134021062970451e-06, + "loss": 0.7248, + "step": 4914 + }, + { + "epoch": 0.5909938074911321, + "grad_norm": 2.218281118789672, + "learning_rate": 1.5126465786230483e-06, + "loss": 0.8125, + "step": 4915 + }, + { + "epoch": 0.5911140503817712, + "grad_norm": 1.7885160319543174, + "learning_rate": 1.5118911248801787e-06, + "loss": 0.8275, + "step": 4916 + }, + { + "epoch": 0.5912342932724103, + "grad_norm": 2.0034821714098747, + "learning_rate": 1.5111357451830364e-06, + "loss": 0.7994, + "step": 4917 + }, + { + "epoch": 0.5913545361630493, + "grad_norm": 2.0164380466772704, + "learning_rate": 1.5103804396462131e-06, + "loss": 0.7126, + "step": 4918 + }, + { + "epoch": 0.5914747790536885, + "grad_norm": 2.1390289147960906, + "learning_rate": 1.5096252083842877e-06, + "loss": 0.8083, + "step": 4919 + }, + { + "epoch": 0.5915950219443276, + "grad_norm": 1.8561281337615245, + "learning_rate": 1.5088700515118285e-06, + "loss": 0.8551, + "step": 4920 + }, + { + "epoch": 0.5917152648349666, + "grad_norm": 1.8872726606675372, + "learning_rate": 1.508114969143392e-06, + "loss": 0.6699, + "step": 4921 + }, + { + "epoch": 0.5918355077256057, + "grad_norm": 6.368911528480922, + "learning_rate": 1.5073599613935238e-06, + "loss": 0.7769, + "step": 4922 + }, + { + "epoch": 0.5919557506162448, + "grad_norm": 2.164729377095902, + "learning_rate": 1.5066050283767574e-06, + "loss": 0.5805, + "step": 4923 + }, + { + "epoch": 0.5920759935068839, + "grad_norm": 1.9012556469474162, + "learning_rate": 1.505850170207616e-06, + "loss": 0.8366, + "step": 4924 + }, + { + "epoch": 0.592196236397523, + "grad_norm": 2.0465135327974493, + "learning_rate": 1.505095387000611e-06, + "loss": 0.7885, + "step": 4925 + }, + { + "epoch": 0.5923164792881621, + "grad_norm": 2.1095684951761684, + "learning_rate": 1.504340678870242e-06, + "loss": 0.7444, + "step": 4926 + }, + { + "epoch": 0.5924367221788012, + "grad_norm": 2.060682594204852, + "learning_rate": 1.5035860459309989e-06, + "loss": 0.9015, + "step": 4927 + }, + { + "epoch": 0.5925569650694402, + "grad_norm": 2.110863958358236, + "learning_rate": 1.5028314882973568e-06, + "loss": 0.64, + "step": 4928 + }, + { + "epoch": 0.5926772079600794, + "grad_norm": 2.7597139627706047, + "learning_rate": 1.502077006083783e-06, + "loss": 0.8497, + "step": 4929 + }, + { + "epoch": 0.5927974508507184, + "grad_norm": 1.8269091092089247, + "learning_rate": 1.5013225994047315e-06, + "loss": 0.7674, + "step": 4930 + }, + { + "epoch": 0.5929176937413575, + "grad_norm": 1.715287205528197, + "learning_rate": 1.5005682683746452e-06, + "loss": 0.8093, + "step": 4931 + }, + { + "epoch": 0.5930379366319967, + "grad_norm": 2.029833255553057, + "learning_rate": 1.4998140131079553e-06, + "loss": 0.7297, + "step": 4932 + }, + { + "epoch": 0.5931581795226357, + "grad_norm": 1.843581499061775, + "learning_rate": 1.4990598337190821e-06, + "loss": 0.7388, + "step": 4933 + }, + { + "epoch": 0.5932784224132748, + "grad_norm": 1.730741113450631, + "learning_rate": 1.4983057303224338e-06, + "loss": 0.6832, + "step": 4934 + }, + { + "epoch": 0.5933986653039139, + "grad_norm": 1.7254859018824857, + "learning_rate": 1.4975517030324072e-06, + "loss": 0.8833, + "step": 4935 + }, + { + "epoch": 0.593518908194553, + "grad_norm": 0.7950946891823201, + "learning_rate": 1.4967977519633882e-06, + "loss": 0.6186, + "step": 4936 + }, + { + "epoch": 0.593639151085192, + "grad_norm": 2.0410532386635785, + "learning_rate": 1.4960438772297494e-06, + "loss": 0.789, + "step": 4937 + }, + { + "epoch": 0.5937593939758312, + "grad_norm": 2.415298886018306, + "learning_rate": 1.495290078945855e-06, + "loss": 0.7359, + "step": 4938 + }, + { + "epoch": 0.5938796368664703, + "grad_norm": 1.7357613461285832, + "learning_rate": 1.4945363572260529e-06, + "loss": 0.746, + "step": 4939 + }, + { + "epoch": 0.5939998797571093, + "grad_norm": 1.966595879051277, + "learning_rate": 1.4937827121846845e-06, + "loss": 0.6843, + "step": 4940 + }, + { + "epoch": 0.5941201226477485, + "grad_norm": 1.4955504703403784, + "learning_rate": 1.4930291439360755e-06, + "loss": 0.737, + "step": 4941 + }, + { + "epoch": 0.5942403655383875, + "grad_norm": 2.053451572498291, + "learning_rate": 1.4922756525945427e-06, + "loss": 0.7966, + "step": 4942 + }, + { + "epoch": 0.5943606084290266, + "grad_norm": 0.7869152274540931, + "learning_rate": 1.4915222382743894e-06, + "loss": 0.5966, + "step": 4943 + }, + { + "epoch": 0.5944808513196658, + "grad_norm": 2.000353781880982, + "learning_rate": 1.4907689010899085e-06, + "loss": 0.7267, + "step": 4944 + }, + { + "epoch": 0.5946010942103048, + "grad_norm": 1.982854666235246, + "learning_rate": 1.4900156411553804e-06, + "loss": 0.6266, + "step": 4945 + }, + { + "epoch": 0.5947213371009439, + "grad_norm": 1.8688180100941598, + "learning_rate": 1.4892624585850739e-06, + "loss": 0.8553, + "step": 4946 + }, + { + "epoch": 0.594841579991583, + "grad_norm": 2.0885984899792853, + "learning_rate": 1.4885093534932465e-06, + "loss": 0.7996, + "step": 4947 + }, + { + "epoch": 0.5949618228822221, + "grad_norm": 2.557795509277286, + "learning_rate": 1.4877563259941433e-06, + "loss": 0.7134, + "step": 4948 + }, + { + "epoch": 0.5950820657728612, + "grad_norm": 1.7438834792155153, + "learning_rate": 1.4870033762019988e-06, + "loss": 0.6795, + "step": 4949 + }, + { + "epoch": 0.5952023086635003, + "grad_norm": 1.5877969298925387, + "learning_rate": 1.4862505042310334e-06, + "loss": 0.7348, + "step": 4950 + }, + { + "epoch": 0.5953225515541394, + "grad_norm": 1.6586191039768674, + "learning_rate": 1.4854977101954587e-06, + "loss": 0.6982, + "step": 4951 + }, + { + "epoch": 0.5954427944447784, + "grad_norm": 1.8593640367200752, + "learning_rate": 1.4847449942094716e-06, + "loss": 0.8617, + "step": 4952 + }, + { + "epoch": 0.5955630373354175, + "grad_norm": 2.2215418913311007, + "learning_rate": 1.4839923563872598e-06, + "loss": 0.866, + "step": 4953 + }, + { + "epoch": 0.5956832802260567, + "grad_norm": 1.7615227770018997, + "learning_rate": 1.483239796842997e-06, + "loss": 0.7582, + "step": 4954 + }, + { + "epoch": 0.5958035231166957, + "grad_norm": 1.7698369383190433, + "learning_rate": 1.4824873156908462e-06, + "loss": 0.8384, + "step": 4955 + }, + { + "epoch": 0.5959237660073348, + "grad_norm": 1.5922946643048306, + "learning_rate": 1.4817349130449584e-06, + "loss": 0.7548, + "step": 4956 + }, + { + "epoch": 0.5960440088979739, + "grad_norm": 1.8855439313506388, + "learning_rate": 1.4809825890194717e-06, + "loss": 0.8294, + "step": 4957 + }, + { + "epoch": 0.596164251788613, + "grad_norm": 1.817117859284152, + "learning_rate": 1.4802303437285139e-06, + "loss": 0.7798, + "step": 4958 + }, + { + "epoch": 0.596284494679252, + "grad_norm": 2.1428814432492005, + "learning_rate": 1.4794781772861994e-06, + "loss": 0.8122, + "step": 4959 + }, + { + "epoch": 0.5964047375698912, + "grad_norm": 2.286519274989271, + "learning_rate": 1.4787260898066324e-06, + "loss": 0.6693, + "step": 4960 + }, + { + "epoch": 0.5965249804605303, + "grad_norm": 3.038422583355374, + "learning_rate": 1.4779740814039023e-06, + "loss": 0.8672, + "step": 4961 + }, + { + "epoch": 0.5966452233511693, + "grad_norm": 1.99791179360706, + "learning_rate": 1.4772221521920894e-06, + "loss": 0.6891, + "step": 4962 + }, + { + "epoch": 0.5967654662418085, + "grad_norm": 1.9039386431881533, + "learning_rate": 1.4764703022852598e-06, + "loss": 0.7441, + "step": 4963 + }, + { + "epoch": 0.5968857091324475, + "grad_norm": 2.4110087119047527, + "learning_rate": 1.4757185317974696e-06, + "loss": 0.7765, + "step": 4964 + }, + { + "epoch": 0.5970059520230866, + "grad_norm": 2.2922715746978954, + "learning_rate": 1.474966840842761e-06, + "loss": 0.7113, + "step": 4965 + }, + { + "epoch": 0.5971261949137258, + "grad_norm": 1.8347054101079203, + "learning_rate": 1.4742152295351655e-06, + "loss": 0.8748, + "step": 4966 + }, + { + "epoch": 0.5972464378043648, + "grad_norm": 2.326210514903941, + "learning_rate": 1.4734636979887016e-06, + "loss": 0.649, + "step": 4967 + }, + { + "epoch": 0.5973666806950039, + "grad_norm": 2.0700852755244443, + "learning_rate": 1.4727122463173755e-06, + "loss": 0.9144, + "step": 4968 + }, + { + "epoch": 0.597486923585643, + "grad_norm": 2.0970137480583904, + "learning_rate": 1.471960874635183e-06, + "loss": 0.6499, + "step": 4969 + }, + { + "epoch": 0.5976071664762821, + "grad_norm": 2.414702987887818, + "learning_rate": 1.4712095830561055e-06, + "loss": 0.7165, + "step": 4970 + }, + { + "epoch": 0.5977274093669211, + "grad_norm": 1.936058492628113, + "learning_rate": 1.4704583716941147e-06, + "loss": 0.8152, + "step": 4971 + }, + { + "epoch": 0.5978476522575603, + "grad_norm": 1.7711270107516601, + "learning_rate": 1.4697072406631672e-06, + "loss": 0.725, + "step": 4972 + }, + { + "epoch": 0.5979678951481994, + "grad_norm": 1.684339955377587, + "learning_rate": 1.4689561900772097e-06, + "loss": 0.7357, + "step": 4973 + }, + { + "epoch": 0.5980881380388384, + "grad_norm": 2.4185464936038823, + "learning_rate": 1.4682052200501758e-06, + "loss": 0.733, + "step": 4974 + }, + { + "epoch": 0.5982083809294776, + "grad_norm": 1.8284248041009805, + "learning_rate": 1.4674543306959876e-06, + "loss": 0.8064, + "step": 4975 + }, + { + "epoch": 0.5983286238201166, + "grad_norm": 2.1989284073024966, + "learning_rate": 1.4667035221285535e-06, + "loss": 0.8546, + "step": 4976 + }, + { + "epoch": 0.5984488667107557, + "grad_norm": 1.87185035651134, + "learning_rate": 1.4659527944617715e-06, + "loss": 0.7431, + "step": 4977 + }, + { + "epoch": 0.5985691096013949, + "grad_norm": 1.9721947460618698, + "learning_rate": 1.465202147809526e-06, + "loss": 0.7604, + "step": 4978 + }, + { + "epoch": 0.5986893524920339, + "grad_norm": 2.1605335754525914, + "learning_rate": 1.4644515822856888e-06, + "loss": 0.7684, + "step": 4979 + }, + { + "epoch": 0.598809595382673, + "grad_norm": 0.7732469343216536, + "learning_rate": 1.4637010980041215e-06, + "loss": 0.5651, + "step": 4980 + }, + { + "epoch": 0.5989298382733121, + "grad_norm": 2.274000380006891, + "learning_rate": 1.4629506950786707e-06, + "loss": 0.9048, + "step": 4981 + }, + { + "epoch": 0.5990500811639512, + "grad_norm": 0.827219918325185, + "learning_rate": 1.4622003736231733e-06, + "loss": 0.5612, + "step": 4982 + }, + { + "epoch": 0.5991703240545903, + "grad_norm": 1.9836403996382268, + "learning_rate": 1.461450133751451e-06, + "loss": 0.8089, + "step": 4983 + }, + { + "epoch": 0.5992905669452293, + "grad_norm": 1.93407512799355, + "learning_rate": 1.4606999755773153e-06, + "loss": 0.7588, + "step": 4984 + }, + { + "epoch": 0.5994108098358685, + "grad_norm": 1.7403880178869784, + "learning_rate": 1.4599498992145643e-06, + "loss": 0.824, + "step": 4985 + }, + { + "epoch": 0.5995310527265075, + "grad_norm": 2.3862185289596676, + "learning_rate": 1.4591999047769846e-06, + "loss": 0.7135, + "step": 4986 + }, + { + "epoch": 0.5996512956171466, + "grad_norm": 1.7771913567909179, + "learning_rate": 1.4584499923783486e-06, + "loss": 0.7553, + "step": 4987 + }, + { + "epoch": 0.5997715385077858, + "grad_norm": 3.3541303992038354, + "learning_rate": 1.457700162132419e-06, + "loss": 0.7651, + "step": 4988 + }, + { + "epoch": 0.5998917813984248, + "grad_norm": 1.8689761991337575, + "learning_rate": 1.4569504141529433e-06, + "loss": 0.7261, + "step": 4989 + }, + { + "epoch": 0.6000120242890639, + "grad_norm": 2.4913415698857064, + "learning_rate": 1.456200748553658e-06, + "loss": 0.7228, + "step": 4990 + }, + { + "epoch": 0.600132267179703, + "grad_norm": 1.8534861683891553, + "learning_rate": 1.455451165448287e-06, + "loss": 0.7881, + "step": 4991 + }, + { + "epoch": 0.6002525100703421, + "grad_norm": 2.503708187369287, + "learning_rate": 1.4547016649505407e-06, + "loss": 0.741, + "step": 4992 + }, + { + "epoch": 0.6003727529609811, + "grad_norm": 2.0291585865943453, + "learning_rate": 1.4539522471741193e-06, + "loss": 0.8524, + "step": 4993 + }, + { + "epoch": 0.6004929958516203, + "grad_norm": 4.411418723957301, + "learning_rate": 1.4532029122327067e-06, + "loss": 0.7174, + "step": 4994 + }, + { + "epoch": 0.6006132387422594, + "grad_norm": 1.8621438831663275, + "learning_rate": 1.4524536602399783e-06, + "loss": 0.7554, + "step": 4995 + }, + { + "epoch": 0.6007334816328984, + "grad_norm": 1.6980901368191417, + "learning_rate": 1.4517044913095938e-06, + "loss": 0.7782, + "step": 4996 + }, + { + "epoch": 0.6008537245235376, + "grad_norm": 1.7767408081864664, + "learning_rate": 1.4509554055552022e-06, + "loss": 0.8205, + "step": 4997 + }, + { + "epoch": 0.6009739674141766, + "grad_norm": 2.917584374784108, + "learning_rate": 1.450206403090439e-06, + "loss": 0.8454, + "step": 4998 + }, + { + "epoch": 0.6010942103048157, + "grad_norm": 1.8944892799998372, + "learning_rate": 1.4494574840289274e-06, + "loss": 0.8638, + "step": 4999 + }, + { + "epoch": 0.6012144531954549, + "grad_norm": 1.6873206997030055, + "learning_rate": 1.4487086484842782e-06, + "loss": 0.7433, + "step": 5000 + }, + { + "epoch": 0.6013346960860939, + "grad_norm": 2.571295367681217, + "learning_rate": 1.4479598965700878e-06, + "loss": 0.6047, + "step": 5001 + }, + { + "epoch": 0.601454938976733, + "grad_norm": 2.376878096212115, + "learning_rate": 1.4472112283999427e-06, + "loss": 0.6942, + "step": 5002 + }, + { + "epoch": 0.6015751818673721, + "grad_norm": 2.5218744781380154, + "learning_rate": 1.4464626440874143e-06, + "loss": 0.6934, + "step": 5003 + }, + { + "epoch": 0.6016954247580112, + "grad_norm": 2.6374366065628374, + "learning_rate": 1.4457141437460636e-06, + "loss": 0.7522, + "step": 5004 + }, + { + "epoch": 0.6018156676486502, + "grad_norm": 2.0110222360680057, + "learning_rate": 1.444965727489436e-06, + "loss": 0.7324, + "step": 5005 + }, + { + "epoch": 0.6019359105392894, + "grad_norm": 1.8653896087440496, + "learning_rate": 1.444217395431066e-06, + "loss": 0.6326, + "step": 5006 + }, + { + "epoch": 0.6020561534299285, + "grad_norm": 0.8020558703456286, + "learning_rate": 1.4434691476844755e-06, + "loss": 0.5588, + "step": 5007 + }, + { + "epoch": 0.6021763963205675, + "grad_norm": 2.608180520778285, + "learning_rate": 1.4427209843631729e-06, + "loss": 0.674, + "step": 5008 + }, + { + "epoch": 0.6022966392112067, + "grad_norm": 3.589080696242511, + "learning_rate": 1.4419729055806534e-06, + "loss": 0.8176, + "step": 5009 + }, + { + "epoch": 0.6024168821018457, + "grad_norm": 1.7738015897639858, + "learning_rate": 1.441224911450401e-06, + "loss": 0.8265, + "step": 5010 + }, + { + "epoch": 0.6025371249924848, + "grad_norm": 1.6632086484015396, + "learning_rate": 1.4404770020858851e-06, + "loss": 0.8241, + "step": 5011 + }, + { + "epoch": 0.602657367883124, + "grad_norm": 2.098998707524193, + "learning_rate": 1.439729177600563e-06, + "loss": 0.8625, + "step": 5012 + }, + { + "epoch": 0.602777610773763, + "grad_norm": 2.058865900164661, + "learning_rate": 1.4389814381078793e-06, + "loss": 0.7305, + "step": 5013 + }, + { + "epoch": 0.6028978536644021, + "grad_norm": 1.979616843440234, + "learning_rate": 1.438233783721265e-06, + "loss": 0.8001, + "step": 5014 + }, + { + "epoch": 0.6030180965550412, + "grad_norm": 2.093655712225152, + "learning_rate": 1.43748621455414e-06, + "loss": 0.779, + "step": 5015 + }, + { + "epoch": 0.6031383394456803, + "grad_norm": 2.4212889916033102, + "learning_rate": 1.4367387307199082e-06, + "loss": 0.8092, + "step": 5016 + }, + { + "epoch": 0.6032585823363193, + "grad_norm": 1.7749833415149674, + "learning_rate": 1.4359913323319632e-06, + "loss": 0.8252, + "step": 5017 + }, + { + "epoch": 0.6033788252269584, + "grad_norm": 1.7387242127843503, + "learning_rate": 1.4352440195036847e-06, + "loss": 0.7783, + "step": 5018 + }, + { + "epoch": 0.6034990681175976, + "grad_norm": 1.6470185450729846, + "learning_rate": 1.4344967923484395e-06, + "loss": 0.8046, + "step": 5019 + }, + { + "epoch": 0.6036193110082366, + "grad_norm": 2.005232522661695, + "learning_rate": 1.433749650979581e-06, + "loss": 0.7222, + "step": 5020 + }, + { + "epoch": 0.6037395538988757, + "grad_norm": 2.0339420965499286, + "learning_rate": 1.433002595510451e-06, + "loss": 0.6843, + "step": 5021 + }, + { + "epoch": 0.6038597967895148, + "grad_norm": 2.1178035051082476, + "learning_rate": 1.4322556260543757e-06, + "loss": 0.7223, + "step": 5022 + }, + { + "epoch": 0.6039800396801539, + "grad_norm": 0.9204345699887146, + "learning_rate": 1.4315087427246703e-06, + "loss": 0.6276, + "step": 5023 + }, + { + "epoch": 0.604100282570793, + "grad_norm": 0.8742573856546852, + "learning_rate": 1.4307619456346372e-06, + "loss": 0.5856, + "step": 5024 + }, + { + "epoch": 0.6042205254614321, + "grad_norm": 1.908227926227805, + "learning_rate": 1.430015234897564e-06, + "loss": 0.7445, + "step": 5025 + }, + { + "epoch": 0.6043407683520712, + "grad_norm": 1.6407728582693801, + "learning_rate": 1.4292686106267274e-06, + "loss": 0.6668, + "step": 5026 + }, + { + "epoch": 0.6044610112427102, + "grad_norm": 1.6357458313858666, + "learning_rate": 1.4285220729353876e-06, + "loss": 0.7766, + "step": 5027 + }, + { + "epoch": 0.6045812541333494, + "grad_norm": 2.829071392389613, + "learning_rate": 1.4277756219367957e-06, + "loss": 0.7862, + "step": 5028 + }, + { + "epoch": 0.6047014970239885, + "grad_norm": 2.0390151445874984, + "learning_rate": 1.4270292577441864e-06, + "loss": 0.802, + "step": 5029 + }, + { + "epoch": 0.6048217399146275, + "grad_norm": 1.6956375443317586, + "learning_rate": 1.4262829804707836e-06, + "loss": 0.7225, + "step": 5030 + }, + { + "epoch": 0.6049419828052667, + "grad_norm": 1.598949999792726, + "learning_rate": 1.4255367902297958e-06, + "loss": 0.7005, + "step": 5031 + }, + { + "epoch": 0.6050622256959057, + "grad_norm": 2.1860112164864884, + "learning_rate": 1.4247906871344215e-06, + "loss": 0.7958, + "step": 5032 + }, + { + "epoch": 0.6051824685865448, + "grad_norm": 2.6510538776668575, + "learning_rate": 1.4240446712978415e-06, + "loss": 0.7567, + "step": 5033 + }, + { + "epoch": 0.605302711477184, + "grad_norm": 2.0858329146665833, + "learning_rate": 1.423298742833227e-06, + "loss": 0.7486, + "step": 5034 + }, + { + "epoch": 0.605422954367823, + "grad_norm": 2.0821092315441327, + "learning_rate": 1.4225529018537352e-06, + "loss": 0.7186, + "step": 5035 + }, + { + "epoch": 0.6055431972584621, + "grad_norm": 1.6705834221652052, + "learning_rate": 1.4218071484725082e-06, + "loss": 0.7775, + "step": 5036 + }, + { + "epoch": 0.6056634401491012, + "grad_norm": 1.9481530955506585, + "learning_rate": 1.4210614828026786e-06, + "loss": 0.766, + "step": 5037 + }, + { + "epoch": 0.6057836830397403, + "grad_norm": 1.597989819314011, + "learning_rate": 1.4203159049573605e-06, + "loss": 0.7464, + "step": 5038 + }, + { + "epoch": 0.6059039259303793, + "grad_norm": 2.0270864216806137, + "learning_rate": 1.4195704150496593e-06, + "loss": 0.8725, + "step": 5039 + }, + { + "epoch": 0.6060241688210185, + "grad_norm": 1.8683257524039931, + "learning_rate": 1.4188250131926639e-06, + "loss": 0.7423, + "step": 5040 + }, + { + "epoch": 0.6061444117116576, + "grad_norm": 2.3466126316886395, + "learning_rate": 1.4180796994994525e-06, + "loss": 0.8087, + "step": 5041 + }, + { + "epoch": 0.6062646546022966, + "grad_norm": 2.0791498396300216, + "learning_rate": 1.4173344740830877e-06, + "loss": 0.725, + "step": 5042 + }, + { + "epoch": 0.6063848974929358, + "grad_norm": 1.7840560761692048, + "learning_rate": 1.4165893370566206e-06, + "loss": 0.7126, + "step": 5043 + }, + { + "epoch": 0.6065051403835748, + "grad_norm": 1.7608028492392782, + "learning_rate": 1.4158442885330865e-06, + "loss": 0.7778, + "step": 5044 + }, + { + "epoch": 0.6066253832742139, + "grad_norm": 2.2800597129497944, + "learning_rate": 1.4150993286255094e-06, + "loss": 0.7913, + "step": 5045 + }, + { + "epoch": 0.6067456261648531, + "grad_norm": 2.086753428466291, + "learning_rate": 1.4143544574468993e-06, + "loss": 0.7988, + "step": 5046 + }, + { + "epoch": 0.6068658690554921, + "grad_norm": 1.7598498934977502, + "learning_rate": 1.4136096751102523e-06, + "loss": 0.824, + "step": 5047 + }, + { + "epoch": 0.6069861119461312, + "grad_norm": 2.1553275305074697, + "learning_rate": 1.4128649817285516e-06, + "loss": 0.8366, + "step": 5048 + }, + { + "epoch": 0.6071063548367702, + "grad_norm": 2.700562690452169, + "learning_rate": 1.412120377414766e-06, + "loss": 0.6336, + "step": 5049 + }, + { + "epoch": 0.6072265977274094, + "grad_norm": 1.5864590693885596, + "learning_rate": 1.4113758622818522e-06, + "loss": 0.7165, + "step": 5050 + }, + { + "epoch": 0.6073468406180484, + "grad_norm": 1.866708734015606, + "learning_rate": 1.410631436442751e-06, + "loss": 0.8334, + "step": 5051 + }, + { + "epoch": 0.6074670835086875, + "grad_norm": 2.0966144155390825, + "learning_rate": 1.4098871000103936e-06, + "loss": 0.8731, + "step": 5052 + }, + { + "epoch": 0.6075873263993267, + "grad_norm": 1.672472909271754, + "learning_rate": 1.409142853097693e-06, + "loss": 0.8287, + "step": 5053 + }, + { + "epoch": 0.6077075692899657, + "grad_norm": 2.0153539904184465, + "learning_rate": 1.408398695817553e-06, + "loss": 0.8005, + "step": 5054 + }, + { + "epoch": 0.6078278121806048, + "grad_norm": 1.839305240035413, + "learning_rate": 1.4076546282828593e-06, + "loss": 0.7034, + "step": 5055 + }, + { + "epoch": 0.6079480550712439, + "grad_norm": 2.02847934442901, + "learning_rate": 1.4069106506064874e-06, + "loss": 0.6621, + "step": 5056 + }, + { + "epoch": 0.608068297961883, + "grad_norm": 1.7604852320694413, + "learning_rate": 1.4061667629012989e-06, + "loss": 0.7866, + "step": 5057 + }, + { + "epoch": 0.608188540852522, + "grad_norm": 1.5698902074728942, + "learning_rate": 1.40542296528014e-06, + "loss": 0.8361, + "step": 5058 + }, + { + "epoch": 0.6083087837431612, + "grad_norm": 2.0654083322994707, + "learning_rate": 1.4046792578558452e-06, + "loss": 0.7691, + "step": 5059 + }, + { + "epoch": 0.6084290266338003, + "grad_norm": 5.960748072022923, + "learning_rate": 1.4039356407412325e-06, + "loss": 0.7664, + "step": 5060 + }, + { + "epoch": 0.6085492695244393, + "grad_norm": 0.7968452672460341, + "learning_rate": 1.40319211404911e-06, + "loss": 0.572, + "step": 5061 + }, + { + "epoch": 0.6086695124150785, + "grad_norm": 1.9272595161799821, + "learning_rate": 1.4024486778922691e-06, + "loss": 0.9119, + "step": 5062 + }, + { + "epoch": 0.6087897553057176, + "grad_norm": 1.8056942992414946, + "learning_rate": 1.4017053323834884e-06, + "loss": 0.78, + "step": 5063 + }, + { + "epoch": 0.6089099981963566, + "grad_norm": 3.616296777986371, + "learning_rate": 1.4009620776355333e-06, + "loss": 0.7622, + "step": 5064 + }, + { + "epoch": 0.6090302410869958, + "grad_norm": 1.7295248587487921, + "learning_rate": 1.4002189137611553e-06, + "loss": 0.794, + "step": 5065 + }, + { + "epoch": 0.6091504839776348, + "grad_norm": 1.7726664610527028, + "learning_rate": 1.3994758408730901e-06, + "loss": 0.7039, + "step": 5066 + }, + { + "epoch": 0.6092707268682739, + "grad_norm": 4.519239200265765, + "learning_rate": 1.3987328590840629e-06, + "loss": 0.7696, + "step": 5067 + }, + { + "epoch": 0.609390969758913, + "grad_norm": 1.8388501377703015, + "learning_rate": 1.397989968506783e-06, + "loss": 0.8678, + "step": 5068 + }, + { + "epoch": 0.6095112126495521, + "grad_norm": 2.305501916815499, + "learning_rate": 1.3972471692539458e-06, + "loss": 0.7365, + "step": 5069 + }, + { + "epoch": 0.6096314555401912, + "grad_norm": 2.0303659716166274, + "learning_rate": 1.3965044614382348e-06, + "loss": 0.7603, + "step": 5070 + }, + { + "epoch": 0.6097516984308303, + "grad_norm": 2.841770531181589, + "learning_rate": 1.3957618451723162e-06, + "loss": 0.7562, + "step": 5071 + }, + { + "epoch": 0.6098719413214694, + "grad_norm": 1.8980284827158316, + "learning_rate": 1.3950193205688457e-06, + "loss": 0.7212, + "step": 5072 + }, + { + "epoch": 0.6099921842121084, + "grad_norm": 1.8414773674018696, + "learning_rate": 1.3942768877404627e-06, + "loss": 0.8424, + "step": 5073 + }, + { + "epoch": 0.6101124271027476, + "grad_norm": 1.6088787165487641, + "learning_rate": 1.393534546799795e-06, + "loss": 0.741, + "step": 5074 + }, + { + "epoch": 0.6102326699933867, + "grad_norm": 1.6164664167413552, + "learning_rate": 1.3927922978594536e-06, + "loss": 0.6838, + "step": 5075 + }, + { + "epoch": 0.6103529128840257, + "grad_norm": 0.7854306704607564, + "learning_rate": 1.3920501410320387e-06, + "loss": 0.5746, + "step": 5076 + }, + { + "epoch": 0.6104731557746649, + "grad_norm": 2.1281306822668724, + "learning_rate": 1.3913080764301333e-06, + "loss": 0.7655, + "step": 5077 + }, + { + "epoch": 0.6105933986653039, + "grad_norm": 1.9442953491572823, + "learning_rate": 1.3905661041663085e-06, + "loss": 0.7172, + "step": 5078 + }, + { + "epoch": 0.610713641555943, + "grad_norm": 2.126261953491842, + "learning_rate": 1.389824224353122e-06, + "loss": 0.663, + "step": 5079 + }, + { + "epoch": 0.610833884446582, + "grad_norm": 2.601724430577688, + "learning_rate": 1.389082437103115e-06, + "loss": 0.7746, + "step": 5080 + }, + { + "epoch": 0.6109541273372212, + "grad_norm": 1.8976476462598288, + "learning_rate": 1.3883407425288172e-06, + "loss": 0.7823, + "step": 5081 + }, + { + "epoch": 0.6110743702278603, + "grad_norm": 3.319664602458348, + "learning_rate": 1.3875991407427417e-06, + "loss": 0.8024, + "step": 5082 + }, + { + "epoch": 0.6111946131184993, + "grad_norm": 0.798803501635705, + "learning_rate": 1.38685763185739e-06, + "loss": 0.5824, + "step": 5083 + }, + { + "epoch": 0.6113148560091385, + "grad_norm": 2.8099568206890586, + "learning_rate": 1.3861162159852476e-06, + "loss": 0.6871, + "step": 5084 + }, + { + "epoch": 0.6114350988997775, + "grad_norm": 1.6323689127425234, + "learning_rate": 1.3853748932387875e-06, + "loss": 0.8044, + "step": 5085 + }, + { + "epoch": 0.6115553417904166, + "grad_norm": 2.2086732950584156, + "learning_rate": 1.3846336637304671e-06, + "loss": 0.7516, + "step": 5086 + }, + { + "epoch": 0.6116755846810558, + "grad_norm": 1.952913013398134, + "learning_rate": 1.3838925275727316e-06, + "loss": 0.8315, + "step": 5087 + }, + { + "epoch": 0.6117958275716948, + "grad_norm": 1.695973069840877, + "learning_rate": 1.3831514848780089e-06, + "loss": 0.7939, + "step": 5088 + }, + { + "epoch": 0.6119160704623339, + "grad_norm": 2.373922505149509, + "learning_rate": 1.3824105357587152e-06, + "loss": 0.9263, + "step": 5089 + }, + { + "epoch": 0.612036313352973, + "grad_norm": 1.5747193707123126, + "learning_rate": 1.381669680327253e-06, + "loss": 0.829, + "step": 5090 + }, + { + "epoch": 0.6121565562436121, + "grad_norm": 1.9422587129238131, + "learning_rate": 1.380928918696008e-06, + "loss": 0.7163, + "step": 5091 + }, + { + "epoch": 0.6122767991342511, + "grad_norm": 2.496066623116485, + "learning_rate": 1.3801882509773548e-06, + "loss": 0.7256, + "step": 5092 + }, + { + "epoch": 0.6123970420248903, + "grad_norm": 1.843853381863098, + "learning_rate": 1.3794476772836503e-06, + "loss": 0.8258, + "step": 5093 + }, + { + "epoch": 0.6125172849155294, + "grad_norm": 1.6800390540987786, + "learning_rate": 1.3787071977272402e-06, + "loss": 0.8487, + "step": 5094 + }, + { + "epoch": 0.6126375278061684, + "grad_norm": 2.587141695876168, + "learning_rate": 1.3779668124204535e-06, + "loss": 0.7257, + "step": 5095 + }, + { + "epoch": 0.6127577706968076, + "grad_norm": 1.6056116787664858, + "learning_rate": 1.3772265214756074e-06, + "loss": 0.8091, + "step": 5096 + }, + { + "epoch": 0.6128780135874466, + "grad_norm": 5.136872472747355, + "learning_rate": 1.3764863250050025e-06, + "loss": 0.7595, + "step": 5097 + }, + { + "epoch": 0.6129982564780857, + "grad_norm": 2.491323291334017, + "learning_rate": 1.3757462231209272e-06, + "loss": 0.8109, + "step": 5098 + }, + { + "epoch": 0.6131184993687249, + "grad_norm": 1.9215324530778035, + "learning_rate": 1.3750062159356525e-06, + "loss": 0.8899, + "step": 5099 + }, + { + "epoch": 0.6132387422593639, + "grad_norm": 1.7640646056846032, + "learning_rate": 1.3742663035614382e-06, + "loss": 0.8328, + "step": 5100 + }, + { + "epoch": 0.613358985150003, + "grad_norm": 2.793567513259946, + "learning_rate": 1.3735264861105283e-06, + "loss": 0.8021, + "step": 5101 + }, + { + "epoch": 0.6134792280406421, + "grad_norm": 1.9907339739951457, + "learning_rate": 1.372786763695152e-06, + "loss": 0.7905, + "step": 5102 + }, + { + "epoch": 0.6135994709312812, + "grad_norm": 5.800940062535628, + "learning_rate": 1.3720471364275257e-06, + "loss": 0.7765, + "step": 5103 + }, + { + "epoch": 0.6137197138219203, + "grad_norm": 3.366600605030104, + "learning_rate": 1.3713076044198486e-06, + "loss": 0.7922, + "step": 5104 + }, + { + "epoch": 0.6138399567125594, + "grad_norm": 2.049159282326111, + "learning_rate": 1.3705681677843086e-06, + "loss": 0.811, + "step": 5105 + }, + { + "epoch": 0.6139601996031985, + "grad_norm": 0.7887779704845983, + "learning_rate": 1.3698288266330768e-06, + "loss": 0.6065, + "step": 5106 + }, + { + "epoch": 0.6140804424938375, + "grad_norm": 3.0031127675389855, + "learning_rate": 1.3690895810783113e-06, + "loss": 0.7252, + "step": 5107 + }, + { + "epoch": 0.6142006853844767, + "grad_norm": 1.8653335624805572, + "learning_rate": 1.3683504312321543e-06, + "loss": 0.7128, + "step": 5108 + }, + { + "epoch": 0.6143209282751158, + "grad_norm": 1.7705290888303788, + "learning_rate": 1.3676113772067355e-06, + "loss": 0.8019, + "step": 5109 + }, + { + "epoch": 0.6144411711657548, + "grad_norm": 2.104833390263715, + "learning_rate": 1.3668724191141671e-06, + "loss": 0.7276, + "step": 5110 + }, + { + "epoch": 0.6145614140563939, + "grad_norm": 2.068283872153861, + "learning_rate": 1.3661335570665493e-06, + "loss": 0.6689, + "step": 5111 + }, + { + "epoch": 0.614681656947033, + "grad_norm": 2.0480542462361986, + "learning_rate": 1.3653947911759676e-06, + "loss": 0.6996, + "step": 5112 + }, + { + "epoch": 0.6148018998376721, + "grad_norm": 1.632351458861664, + "learning_rate": 1.3646561215544904e-06, + "loss": 0.7507, + "step": 5113 + }, + { + "epoch": 0.6149221427283111, + "grad_norm": 1.9951664213210611, + "learning_rate": 1.363917548314176e-06, + "loss": 0.8016, + "step": 5114 + }, + { + "epoch": 0.6150423856189503, + "grad_norm": 1.655241749327953, + "learning_rate": 1.3631790715670626e-06, + "loss": 0.7388, + "step": 5115 + }, + { + "epoch": 0.6151626285095894, + "grad_norm": 1.939292140399068, + "learning_rate": 1.3624406914251783e-06, + "loss": 0.8627, + "step": 5116 + }, + { + "epoch": 0.6152828714002284, + "grad_norm": 1.8775753137185476, + "learning_rate": 1.3617024080005335e-06, + "loss": 0.888, + "step": 5117 + }, + { + "epoch": 0.6154031142908676, + "grad_norm": 1.6899046542567173, + "learning_rate": 1.3609642214051266e-06, + "loss": 0.7481, + "step": 5118 + }, + { + "epoch": 0.6155233571815066, + "grad_norm": 2.0025475489463456, + "learning_rate": 1.3602261317509385e-06, + "loss": 0.6617, + "step": 5119 + }, + { + "epoch": 0.6156436000721457, + "grad_norm": 2.453217717259095, + "learning_rate": 1.3594881391499387e-06, + "loss": 0.8286, + "step": 5120 + }, + { + "epoch": 0.6157638429627849, + "grad_norm": 1.713465108195073, + "learning_rate": 1.3587502437140778e-06, + "loss": 0.794, + "step": 5121 + }, + { + "epoch": 0.6158840858534239, + "grad_norm": 2.8640020902082606, + "learning_rate": 1.3580124455552952e-06, + "loss": 0.8567, + "step": 5122 + }, + { + "epoch": 0.616004328744063, + "grad_norm": 1.802491540558285, + "learning_rate": 1.3572747447855148e-06, + "loss": 0.8765, + "step": 5123 + }, + { + "epoch": 0.6161245716347021, + "grad_norm": 1.9831079958749656, + "learning_rate": 1.356537141516644e-06, + "loss": 0.6929, + "step": 5124 + }, + { + "epoch": 0.6162448145253412, + "grad_norm": 2.001884168972487, + "learning_rate": 1.3557996358605775e-06, + "loss": 0.6237, + "step": 5125 + }, + { + "epoch": 0.6163650574159802, + "grad_norm": 2.274209641884243, + "learning_rate": 1.3550622279291941e-06, + "loss": 0.7094, + "step": 5126 + }, + { + "epoch": 0.6164853003066194, + "grad_norm": 1.4122061162077373, + "learning_rate": 1.354324917834358e-06, + "loss": 0.83, + "step": 5127 + }, + { + "epoch": 0.6166055431972585, + "grad_norm": 2.1841920091113822, + "learning_rate": 1.353587705687918e-06, + "loss": 0.777, + "step": 5128 + }, + { + "epoch": 0.6167257860878975, + "grad_norm": 3.991004843322522, + "learning_rate": 1.3528505916017096e-06, + "loss": 0.7231, + "step": 5129 + }, + { + "epoch": 0.6168460289785367, + "grad_norm": 2.2223525552146737, + "learning_rate": 1.3521135756875514e-06, + "loss": 0.8878, + "step": 5130 + }, + { + "epoch": 0.6169662718691757, + "grad_norm": 1.4303327261369978, + "learning_rate": 1.3513766580572496e-06, + "loss": 0.8641, + "step": 5131 + }, + { + "epoch": 0.6170865147598148, + "grad_norm": 2.7828848673574407, + "learning_rate": 1.3506398388225924e-06, + "loss": 0.775, + "step": 5132 + }, + { + "epoch": 0.617206757650454, + "grad_norm": 1.791059953147248, + "learning_rate": 1.349903118095355e-06, + "loss": 0.7179, + "step": 5133 + }, + { + "epoch": 0.617327000541093, + "grad_norm": 1.7335509671007336, + "learning_rate": 1.349166495987298e-06, + "loss": 0.737, + "step": 5134 + }, + { + "epoch": 0.6174472434317321, + "grad_norm": 0.8293442297993426, + "learning_rate": 1.348429972610166e-06, + "loss": 0.6089, + "step": 5135 + }, + { + "epoch": 0.6175674863223712, + "grad_norm": 0.8579939061528659, + "learning_rate": 1.3476935480756897e-06, + "loss": 0.5789, + "step": 5136 + }, + { + "epoch": 0.6176877292130103, + "grad_norm": 2.1724616706021758, + "learning_rate": 1.346957222495583e-06, + "loss": 0.762, + "step": 5137 + }, + { + "epoch": 0.6178079721036493, + "grad_norm": 3.547026912444074, + "learning_rate": 1.3462209959815466e-06, + "loss": 0.7159, + "step": 5138 + }, + { + "epoch": 0.6179282149942885, + "grad_norm": 1.8781494392924907, + "learning_rate": 1.345484868645265e-06, + "loss": 0.7493, + "step": 5139 + }, + { + "epoch": 0.6180484578849276, + "grad_norm": 2.4283264862082703, + "learning_rate": 1.3447488405984088e-06, + "loss": 0.7871, + "step": 5140 + }, + { + "epoch": 0.6181687007755666, + "grad_norm": 2.185484475343573, + "learning_rate": 1.3440129119526322e-06, + "loss": 0.7071, + "step": 5141 + }, + { + "epoch": 0.6182889436662057, + "grad_norm": 0.8299177665648259, + "learning_rate": 1.3432770828195762e-06, + "loss": 0.5119, + "step": 5142 + }, + { + "epoch": 0.6184091865568448, + "grad_norm": 2.4157350542233247, + "learning_rate": 1.3425413533108635e-06, + "loss": 0.7111, + "step": 5143 + }, + { + "epoch": 0.6185294294474839, + "grad_norm": 2.2924876967571093, + "learning_rate": 1.341805723538105e-06, + "loss": 0.7129, + "step": 5144 + }, + { + "epoch": 0.618649672338123, + "grad_norm": 1.576532221069533, + "learning_rate": 1.3410701936128948e-06, + "loss": 0.7777, + "step": 5145 + }, + { + "epoch": 0.6187699152287621, + "grad_norm": 2.307482353465342, + "learning_rate": 1.340334763646812e-06, + "loss": 0.8514, + "step": 5146 + }, + { + "epoch": 0.6188901581194012, + "grad_norm": 1.8152787841986069, + "learning_rate": 1.3395994337514218e-06, + "loss": 0.7464, + "step": 5147 + }, + { + "epoch": 0.6190104010100402, + "grad_norm": 3.1255727624523137, + "learning_rate": 1.3388642040382725e-06, + "loss": 0.7882, + "step": 5148 + }, + { + "epoch": 0.6191306439006794, + "grad_norm": 3.3757296585840875, + "learning_rate": 1.3381290746188975e-06, + "loss": 0.8455, + "step": 5149 + }, + { + "epoch": 0.6192508867913185, + "grad_norm": 1.7082756622008133, + "learning_rate": 1.3373940456048152e-06, + "loss": 0.6731, + "step": 5150 + }, + { + "epoch": 0.6193711296819575, + "grad_norm": 1.6092332203458266, + "learning_rate": 1.3366591171075299e-06, + "loss": 0.5934, + "step": 5151 + }, + { + "epoch": 0.6194913725725967, + "grad_norm": 1.9081751998060628, + "learning_rate": 1.335924289238529e-06, + "loss": 0.9097, + "step": 5152 + }, + { + "epoch": 0.6196116154632357, + "grad_norm": 1.698957962544407, + "learning_rate": 1.3351895621092859e-06, + "loss": 0.7683, + "step": 5153 + }, + { + "epoch": 0.6197318583538748, + "grad_norm": 2.0657048612499085, + "learning_rate": 1.3344549358312567e-06, + "loss": 0.768, + "step": 5154 + }, + { + "epoch": 0.619852101244514, + "grad_norm": 1.9537145106435385, + "learning_rate": 1.3337204105158852e-06, + "loss": 0.7876, + "step": 5155 + }, + { + "epoch": 0.619972344135153, + "grad_norm": 1.9272497431756999, + "learning_rate": 1.332985986274597e-06, + "loss": 0.7265, + "step": 5156 + }, + { + "epoch": 0.6200925870257921, + "grad_norm": 2.1063343880158687, + "learning_rate": 1.3322516632188047e-06, + "loss": 0.7532, + "step": 5157 + }, + { + "epoch": 0.6202128299164312, + "grad_norm": 1.7244891668052986, + "learning_rate": 1.3315174414599045e-06, + "loss": 0.6725, + "step": 5158 + }, + { + "epoch": 0.6203330728070703, + "grad_norm": 2.028582185490293, + "learning_rate": 1.3307833211092768e-06, + "loss": 0.7578, + "step": 5159 + }, + { + "epoch": 0.6204533156977093, + "grad_norm": 1.73476604830164, + "learning_rate": 1.3300493022782873e-06, + "loss": 0.7596, + "step": 5160 + }, + { + "epoch": 0.6205735585883485, + "grad_norm": 5.60921569427991, + "learning_rate": 1.3293153850782855e-06, + "loss": 0.7316, + "step": 5161 + }, + { + "epoch": 0.6206938014789876, + "grad_norm": 1.7835532572724313, + "learning_rate": 1.3285815696206069e-06, + "loss": 0.7168, + "step": 5162 + }, + { + "epoch": 0.6208140443696266, + "grad_norm": 1.9643352601379407, + "learning_rate": 1.32784785601657e-06, + "loss": 0.7715, + "step": 5163 + }, + { + "epoch": 0.6209342872602658, + "grad_norm": 3.017258117892434, + "learning_rate": 1.3271142443774798e-06, + "loss": 0.7368, + "step": 5164 + }, + { + "epoch": 0.6210545301509048, + "grad_norm": 1.9820644112753592, + "learning_rate": 1.3263807348146228e-06, + "loss": 0.8226, + "step": 5165 + }, + { + "epoch": 0.6211747730415439, + "grad_norm": 2.0165073326577203, + "learning_rate": 1.3256473274392733e-06, + "loss": 0.7395, + "step": 5166 + }, + { + "epoch": 0.6212950159321831, + "grad_norm": 1.869646376664915, + "learning_rate": 1.3249140223626873e-06, + "loss": 0.7038, + "step": 5167 + }, + { + "epoch": 0.6214152588228221, + "grad_norm": 1.9182082762777404, + "learning_rate": 1.3241808196961077e-06, + "loss": 0.7569, + "step": 5168 + }, + { + "epoch": 0.6215355017134612, + "grad_norm": 1.8315856651998164, + "learning_rate": 1.3234477195507608e-06, + "loss": 0.7075, + "step": 5169 + }, + { + "epoch": 0.6216557446041003, + "grad_norm": 1.9882615843514195, + "learning_rate": 1.322714722037857e-06, + "loss": 0.6256, + "step": 5170 + }, + { + "epoch": 0.6217759874947394, + "grad_norm": 2.25740908282303, + "learning_rate": 1.321981827268591e-06, + "loss": 0.7779, + "step": 5171 + }, + { + "epoch": 0.6218962303853784, + "grad_norm": 1.7567914243596516, + "learning_rate": 1.3212490353541426e-06, + "loss": 0.8159, + "step": 5172 + }, + { + "epoch": 0.6220164732760175, + "grad_norm": 2.0570168871003265, + "learning_rate": 1.3205163464056762e-06, + "loss": 0.8081, + "step": 5173 + }, + { + "epoch": 0.6221367161666567, + "grad_norm": 3.2405252998634615, + "learning_rate": 1.319783760534339e-06, + "loss": 0.7262, + "step": 5174 + }, + { + "epoch": 0.6222569590572957, + "grad_norm": 2.143153792238219, + "learning_rate": 1.319051277851266e-06, + "loss": 0.7597, + "step": 5175 + }, + { + "epoch": 0.6223772019479348, + "grad_norm": 1.9736950848932349, + "learning_rate": 1.3183188984675716e-06, + "loss": 0.8467, + "step": 5176 + }, + { + "epoch": 0.6224974448385739, + "grad_norm": 2.504081065343437, + "learning_rate": 1.3175866224943586e-06, + "loss": 0.7158, + "step": 5177 + }, + { + "epoch": 0.622617687729213, + "grad_norm": 2.2486089008842205, + "learning_rate": 1.316854450042712e-06, + "loss": 0.7422, + "step": 5178 + }, + { + "epoch": 0.622737930619852, + "grad_norm": 1.891384508148995, + "learning_rate": 1.3161223812237024e-06, + "loss": 0.7485, + "step": 5179 + }, + { + "epoch": 0.6228581735104912, + "grad_norm": 4.785185347488063, + "learning_rate": 1.3153904161483842e-06, + "loss": 0.8545, + "step": 5180 + }, + { + "epoch": 0.6229784164011303, + "grad_norm": 2.0172719215320507, + "learning_rate": 1.3146585549277953e-06, + "loss": 0.8693, + "step": 5181 + }, + { + "epoch": 0.6230986592917693, + "grad_norm": 2.3054490722312866, + "learning_rate": 1.3139267976729591e-06, + "loss": 0.79, + "step": 5182 + }, + { + "epoch": 0.6232189021824085, + "grad_norm": 1.8223269862140103, + "learning_rate": 1.3131951444948815e-06, + "loss": 0.7257, + "step": 5183 + }, + { + "epoch": 0.6233391450730476, + "grad_norm": 1.8437688794220108, + "learning_rate": 1.3124635955045546e-06, + "loss": 0.7643, + "step": 5184 + }, + { + "epoch": 0.6234593879636866, + "grad_norm": 2.0645092511735452, + "learning_rate": 1.3117321508129537e-06, + "loss": 0.844, + "step": 5185 + }, + { + "epoch": 0.6235796308543258, + "grad_norm": 1.5021426474420332, + "learning_rate": 1.3110008105310388e-06, + "loss": 0.7652, + "step": 5186 + }, + { + "epoch": 0.6236998737449648, + "grad_norm": 2.1219914400177937, + "learning_rate": 1.3102695747697526e-06, + "loss": 0.784, + "step": 5187 + }, + { + "epoch": 0.6238201166356039, + "grad_norm": 3.7240565289264405, + "learning_rate": 1.3095384436400237e-06, + "loss": 0.9058, + "step": 5188 + }, + { + "epoch": 0.623940359526243, + "grad_norm": 2.066326291244543, + "learning_rate": 1.3088074172527633e-06, + "loss": 0.832, + "step": 5189 + }, + { + "epoch": 0.6240606024168821, + "grad_norm": 9.979582351171246, + "learning_rate": 1.3080764957188684e-06, + "loss": 0.7226, + "step": 5190 + }, + { + "epoch": 0.6241808453075212, + "grad_norm": 1.766354840427666, + "learning_rate": 1.3073456791492192e-06, + "loss": 0.7108, + "step": 5191 + }, + { + "epoch": 0.6243010881981603, + "grad_norm": 1.9807011996459871, + "learning_rate": 1.3066149676546801e-06, + "loss": 0.7873, + "step": 5192 + }, + { + "epoch": 0.6244213310887994, + "grad_norm": 1.7679757192679142, + "learning_rate": 1.3058843613460985e-06, + "loss": 0.6684, + "step": 5193 + }, + { + "epoch": 0.6245415739794384, + "grad_norm": 1.949054718760459, + "learning_rate": 1.3051538603343075e-06, + "loss": 0.7527, + "step": 5194 + }, + { + "epoch": 0.6246618168700776, + "grad_norm": 1.7677789563480724, + "learning_rate": 1.3044234647301235e-06, + "loss": 0.6808, + "step": 5195 + }, + { + "epoch": 0.6247820597607167, + "grad_norm": 1.866981207652389, + "learning_rate": 1.303693174644347e-06, + "loss": 0.7244, + "step": 5196 + }, + { + "epoch": 0.6249023026513557, + "grad_norm": 1.969790108241516, + "learning_rate": 1.3029629901877625e-06, + "loss": 0.81, + "step": 5197 + }, + { + "epoch": 0.6250225455419949, + "grad_norm": 2.9454637078850205, + "learning_rate": 1.3022329114711376e-06, + "loss": 0.7768, + "step": 5198 + }, + { + "epoch": 0.6251427884326339, + "grad_norm": 2.29001886484963, + "learning_rate": 1.3015029386052256e-06, + "loss": 0.7017, + "step": 5199 + }, + { + "epoch": 0.625263031323273, + "grad_norm": 2.313996526690805, + "learning_rate": 1.3007730717007622e-06, + "loss": 0.7302, + "step": 5200 + }, + { + "epoch": 0.6253832742139122, + "grad_norm": 1.7374948158461259, + "learning_rate": 1.3000433108684676e-06, + "loss": 0.7547, + "step": 5201 + }, + { + "epoch": 0.6255035171045512, + "grad_norm": 2.537198365422323, + "learning_rate": 1.2993136562190467e-06, + "loss": 0.8103, + "step": 5202 + }, + { + "epoch": 0.6256237599951903, + "grad_norm": 1.8192520654647486, + "learning_rate": 1.2985841078631871e-06, + "loss": 0.708, + "step": 5203 + }, + { + "epoch": 0.6257440028858293, + "grad_norm": 1.8083488677569104, + "learning_rate": 1.2978546659115608e-06, + "loss": 0.7846, + "step": 5204 + }, + { + "epoch": 0.6258642457764685, + "grad_norm": 1.9414020745522251, + "learning_rate": 1.2971253304748228e-06, + "loss": 0.856, + "step": 5205 + }, + { + "epoch": 0.6259844886671075, + "grad_norm": 1.8139811416242946, + "learning_rate": 1.296396101663614e-06, + "loss": 0.7514, + "step": 5206 + }, + { + "epoch": 0.6261047315577466, + "grad_norm": 2.1517092674799336, + "learning_rate": 1.2956669795885565e-06, + "loss": 0.8452, + "step": 5207 + }, + { + "epoch": 0.6262249744483858, + "grad_norm": 1.8275544670118644, + "learning_rate": 1.294937964360259e-06, + "loss": 0.6816, + "step": 5208 + }, + { + "epoch": 0.6263452173390248, + "grad_norm": 2.1943900454360405, + "learning_rate": 1.2942090560893108e-06, + "loss": 0.7147, + "step": 5209 + }, + { + "epoch": 0.6264654602296639, + "grad_norm": 1.8737283845129324, + "learning_rate": 1.2934802548862882e-06, + "loss": 0.6057, + "step": 5210 + }, + { + "epoch": 0.626585703120303, + "grad_norm": 2.067255837023862, + "learning_rate": 1.292751560861749e-06, + "loss": 0.8304, + "step": 5211 + }, + { + "epoch": 0.6267059460109421, + "grad_norm": 1.8863118050715126, + "learning_rate": 1.2920229741262354e-06, + "loss": 0.8001, + "step": 5212 + }, + { + "epoch": 0.6268261889015811, + "grad_norm": 2.062428097483061, + "learning_rate": 1.2912944947902739e-06, + "loss": 0.756, + "step": 5213 + }, + { + "epoch": 0.6269464317922203, + "grad_norm": 2.2968647236496835, + "learning_rate": 1.2905661229643742e-06, + "loss": 0.7233, + "step": 5214 + }, + { + "epoch": 0.6270666746828594, + "grad_norm": 2.264864353219121, + "learning_rate": 1.2898378587590299e-06, + "loss": 0.8509, + "step": 5215 + }, + { + "epoch": 0.6271869175734984, + "grad_norm": 2.2353877193381515, + "learning_rate": 1.2891097022847173e-06, + "loss": 0.8834, + "step": 5216 + }, + { + "epoch": 0.6273071604641376, + "grad_norm": 1.669867049538605, + "learning_rate": 1.2883816536518978e-06, + "loss": 0.6703, + "step": 5217 + }, + { + "epoch": 0.6274274033547766, + "grad_norm": 1.788865124338599, + "learning_rate": 1.2876537129710155e-06, + "loss": 0.8197, + "step": 5218 + }, + { + "epoch": 0.6275476462454157, + "grad_norm": 2.4911524617365286, + "learning_rate": 1.286925880352499e-06, + "loss": 0.7566, + "step": 5219 + }, + { + "epoch": 0.6276678891360549, + "grad_norm": 1.9619638257245056, + "learning_rate": 1.2861981559067592e-06, + "loss": 0.7136, + "step": 5220 + }, + { + "epoch": 0.6277881320266939, + "grad_norm": 1.7735486772953495, + "learning_rate": 1.2854705397441917e-06, + "loss": 0.802, + "step": 5221 + }, + { + "epoch": 0.627908374917333, + "grad_norm": 2.2228808566077807, + "learning_rate": 1.2847430319751747e-06, + "loss": 0.7812, + "step": 5222 + }, + { + "epoch": 0.6280286178079721, + "grad_norm": 2.910124953952945, + "learning_rate": 1.2840156327100712e-06, + "loss": 0.6723, + "step": 5223 + }, + { + "epoch": 0.6281488606986112, + "grad_norm": 1.7586282151385553, + "learning_rate": 1.2832883420592272e-06, + "loss": 0.7267, + "step": 5224 + }, + { + "epoch": 0.6282691035892503, + "grad_norm": 2.0314611471912545, + "learning_rate": 1.282561160132972e-06, + "loss": 0.6439, + "step": 5225 + }, + { + "epoch": 0.6283893464798894, + "grad_norm": 1.7124015779822568, + "learning_rate": 1.2818340870416186e-06, + "loss": 0.8116, + "step": 5226 + }, + { + "epoch": 0.6285095893705285, + "grad_norm": 1.8100404823370595, + "learning_rate": 1.2811071228954626e-06, + "loss": 0.7611, + "step": 5227 + }, + { + "epoch": 0.6286298322611675, + "grad_norm": 1.9182740879383449, + "learning_rate": 1.2803802678047846e-06, + "loss": 0.8129, + "step": 5228 + }, + { + "epoch": 0.6287500751518067, + "grad_norm": 5.315790212172973, + "learning_rate": 1.279653521879848e-06, + "loss": 0.7461, + "step": 5229 + }, + { + "epoch": 0.6288703180424458, + "grad_norm": 2.154786649971865, + "learning_rate": 1.2789268852308997e-06, + "loss": 0.8445, + "step": 5230 + }, + { + "epoch": 0.6289905609330848, + "grad_norm": 1.9986052418118132, + "learning_rate": 1.2782003579681688e-06, + "loss": 0.713, + "step": 5231 + }, + { + "epoch": 0.629110803823724, + "grad_norm": 1.6913369673355472, + "learning_rate": 1.2774739402018701e-06, + "loss": 0.7449, + "step": 5232 + }, + { + "epoch": 0.629231046714363, + "grad_norm": 1.7341947032245302, + "learning_rate": 1.2767476320422002e-06, + "loss": 0.735, + "step": 5233 + }, + { + "epoch": 0.6293512896050021, + "grad_norm": 0.7879925640621893, + "learning_rate": 1.2760214335993392e-06, + "loss": 0.5722, + "step": 5234 + }, + { + "epoch": 0.6294715324956413, + "grad_norm": 2.4379756170028326, + "learning_rate": 1.2752953449834514e-06, + "loss": 0.5898, + "step": 5235 + }, + { + "epoch": 0.6295917753862803, + "grad_norm": 2.0952505172220897, + "learning_rate": 1.2745693663046836e-06, + "loss": 0.805, + "step": 5236 + }, + { + "epoch": 0.6297120182769194, + "grad_norm": 2.7148978468081477, + "learning_rate": 1.2738434976731662e-06, + "loss": 0.8079, + "step": 5237 + }, + { + "epoch": 0.6298322611675584, + "grad_norm": 1.8185344083235246, + "learning_rate": 1.2731177391990125e-06, + "loss": 0.7549, + "step": 5238 + }, + { + "epoch": 0.6299525040581976, + "grad_norm": 1.968314470830081, + "learning_rate": 1.2723920909923203e-06, + "loss": 0.8241, + "step": 5239 + }, + { + "epoch": 0.6300727469488366, + "grad_norm": 0.8667917465463956, + "learning_rate": 1.2716665531631688e-06, + "loss": 0.6052, + "step": 5240 + }, + { + "epoch": 0.6301929898394757, + "grad_norm": 1.7668223414419066, + "learning_rate": 1.270941125821623e-06, + "loss": 0.7738, + "step": 5241 + }, + { + "epoch": 0.6303132327301149, + "grad_norm": 1.654146910516613, + "learning_rate": 1.2702158090777278e-06, + "loss": 0.755, + "step": 5242 + }, + { + "epoch": 0.6304334756207539, + "grad_norm": 2.190208284410769, + "learning_rate": 1.2694906030415148e-06, + "loss": 0.7536, + "step": 5243 + }, + { + "epoch": 0.630553718511393, + "grad_norm": 3.1278679466764103, + "learning_rate": 1.2687655078229958e-06, + "loss": 0.8319, + "step": 5244 + }, + { + "epoch": 0.6306739614020321, + "grad_norm": 2.1356354640116004, + "learning_rate": 1.2680405235321678e-06, + "loss": 0.6946, + "step": 5245 + }, + { + "epoch": 0.6307942042926712, + "grad_norm": 3.300856369337523, + "learning_rate": 1.267315650279011e-06, + "loss": 0.7963, + "step": 5246 + }, + { + "epoch": 0.6309144471833102, + "grad_norm": 2.049371725430538, + "learning_rate": 1.2665908881734874e-06, + "loss": 0.7502, + "step": 5247 + }, + { + "epoch": 0.6310346900739494, + "grad_norm": 2.0412144322283736, + "learning_rate": 1.2658662373255432e-06, + "loss": 0.8567, + "step": 5248 + }, + { + "epoch": 0.6311549329645885, + "grad_norm": 0.7257234794693546, + "learning_rate": 1.2651416978451063e-06, + "loss": 0.5233, + "step": 5249 + }, + { + "epoch": 0.6312751758552275, + "grad_norm": 1.8709091063562726, + "learning_rate": 1.2644172698420903e-06, + "loss": 0.6555, + "step": 5250 + }, + { + "epoch": 0.6313954187458667, + "grad_norm": 1.860672313843352, + "learning_rate": 1.2636929534263892e-06, + "loss": 0.8525, + "step": 5251 + }, + { + "epoch": 0.6315156616365057, + "grad_norm": 1.7181612566983242, + "learning_rate": 1.2629687487078821e-06, + "loss": 0.778, + "step": 5252 + }, + { + "epoch": 0.6316359045271448, + "grad_norm": 2.497155096698107, + "learning_rate": 1.2622446557964293e-06, + "loss": 0.7708, + "step": 5253 + }, + { + "epoch": 0.631756147417784, + "grad_norm": 1.6655735086637429, + "learning_rate": 1.261520674801876e-06, + "loss": 0.7182, + "step": 5254 + }, + { + "epoch": 0.631876390308423, + "grad_norm": 2.508764923918031, + "learning_rate": 1.2607968058340488e-06, + "loss": 0.728, + "step": 5255 + }, + { + "epoch": 0.6319966331990621, + "grad_norm": 1.8205504691638474, + "learning_rate": 1.2600730490027583e-06, + "loss": 0.7315, + "step": 5256 + }, + { + "epoch": 0.6321168760897012, + "grad_norm": 1.7417125653268064, + "learning_rate": 1.2593494044177984e-06, + "loss": 0.8044, + "step": 5257 + }, + { + "epoch": 0.6322371189803403, + "grad_norm": 2.2840533863782984, + "learning_rate": 1.2586258721889448e-06, + "loss": 0.81, + "step": 5258 + }, + { + "epoch": 0.6323573618709794, + "grad_norm": 2.037966612762311, + "learning_rate": 1.2579024524259573e-06, + "loss": 0.8178, + "step": 5259 + }, + { + "epoch": 0.6324776047616185, + "grad_norm": 1.983879207640993, + "learning_rate": 1.2571791452385768e-06, + "loss": 0.9161, + "step": 5260 + }, + { + "epoch": 0.6325978476522576, + "grad_norm": 1.8945058290014853, + "learning_rate": 1.2564559507365301e-06, + "loss": 0.7751, + "step": 5261 + }, + { + "epoch": 0.6327180905428966, + "grad_norm": 2.870004044905414, + "learning_rate": 1.2557328690295244e-06, + "loss": 0.7926, + "step": 5262 + }, + { + "epoch": 0.6328383334335358, + "grad_norm": 1.690426866693198, + "learning_rate": 1.255009900227251e-06, + "loss": 0.7608, + "step": 5263 + }, + { + "epoch": 0.6329585763241748, + "grad_norm": 1.943720645008253, + "learning_rate": 1.254287044439383e-06, + "loss": 0.7952, + "step": 5264 + }, + { + "epoch": 0.6330788192148139, + "grad_norm": 0.7848793414500542, + "learning_rate": 1.2535643017755776e-06, + "loss": 0.5443, + "step": 5265 + }, + { + "epoch": 0.6331990621054531, + "grad_norm": 2.079861711159862, + "learning_rate": 1.2528416723454737e-06, + "loss": 0.7225, + "step": 5266 + }, + { + "epoch": 0.6333193049960921, + "grad_norm": 1.501943838738157, + "learning_rate": 1.2521191562586945e-06, + "loss": 0.7114, + "step": 5267 + }, + { + "epoch": 0.6334395478867312, + "grad_norm": 1.950653465525314, + "learning_rate": 1.2513967536248445e-06, + "loss": 0.7716, + "step": 5268 + }, + { + "epoch": 0.6335597907773702, + "grad_norm": 1.8247725567116653, + "learning_rate": 1.2506744645535117e-06, + "loss": 0.8117, + "step": 5269 + }, + { + "epoch": 0.6336800336680094, + "grad_norm": 2.1377605569431815, + "learning_rate": 1.249952289154267e-06, + "loss": 0.6125, + "step": 5270 + }, + { + "epoch": 0.6338002765586485, + "grad_norm": 1.757328356721823, + "learning_rate": 1.2492302275366635e-06, + "loss": 0.7655, + "step": 5271 + }, + { + "epoch": 0.6339205194492875, + "grad_norm": 3.3592753550810452, + "learning_rate": 1.2485082798102377e-06, + "loss": 0.6682, + "step": 5272 + }, + { + "epoch": 0.6340407623399267, + "grad_norm": 2.2057053625870817, + "learning_rate": 1.2477864460845084e-06, + "loss": 0.688, + "step": 5273 + }, + { + "epoch": 0.6341610052305657, + "grad_norm": 4.45703039411692, + "learning_rate": 1.2470647264689776e-06, + "loss": 0.7453, + "step": 5274 + }, + { + "epoch": 0.6342812481212048, + "grad_norm": 2.0286811430522493, + "learning_rate": 1.2463431210731282e-06, + "loss": 0.725, + "step": 5275 + }, + { + "epoch": 0.634401491011844, + "grad_norm": 3.13347194513837, + "learning_rate": 1.2456216300064289e-06, + "loss": 0.7679, + "step": 5276 + }, + { + "epoch": 0.634521733902483, + "grad_norm": 1.7916560801111954, + "learning_rate": 1.244900253378328e-06, + "loss": 0.7853, + "step": 5277 + }, + { + "epoch": 0.6346419767931221, + "grad_norm": 1.9982762060624482, + "learning_rate": 1.2441789912982583e-06, + "loss": 0.7052, + "step": 5278 + }, + { + "epoch": 0.6347622196837612, + "grad_norm": 1.8080682176362883, + "learning_rate": 1.2434578438756346e-06, + "loss": 0.6459, + "step": 5279 + }, + { + "epoch": 0.6348824625744003, + "grad_norm": 1.8915884542331052, + "learning_rate": 1.242736811219855e-06, + "loss": 0.7841, + "step": 5280 + }, + { + "epoch": 0.6350027054650393, + "grad_norm": 2.4311282679033575, + "learning_rate": 1.2420158934402988e-06, + "loss": 0.824, + "step": 5281 + }, + { + "epoch": 0.6351229483556785, + "grad_norm": 2.061521471421078, + "learning_rate": 1.2412950906463286e-06, + "loss": 0.8496, + "step": 5282 + }, + { + "epoch": 0.6352431912463176, + "grad_norm": 1.8526058556998806, + "learning_rate": 1.2405744029472902e-06, + "loss": 0.9005, + "step": 5283 + }, + { + "epoch": 0.6353634341369566, + "grad_norm": 2.0830454691055316, + "learning_rate": 1.2398538304525108e-06, + "loss": 0.7688, + "step": 5284 + }, + { + "epoch": 0.6354836770275958, + "grad_norm": 2.396262409653665, + "learning_rate": 1.2391333732713016e-06, + "loss": 0.7682, + "step": 5285 + }, + { + "epoch": 0.6356039199182348, + "grad_norm": 2.0930062866262333, + "learning_rate": 1.2384130315129543e-06, + "loss": 0.7897, + "step": 5286 + }, + { + "epoch": 0.6357241628088739, + "grad_norm": 2.7251011120996944, + "learning_rate": 1.2376928052867447e-06, + "loss": 0.7401, + "step": 5287 + }, + { + "epoch": 0.6358444056995131, + "grad_norm": 2.0925492216858896, + "learning_rate": 1.2369726947019299e-06, + "loss": 0.7768, + "step": 5288 + }, + { + "epoch": 0.6359646485901521, + "grad_norm": 2.111000847307159, + "learning_rate": 1.2362526998677511e-06, + "loss": 0.6729, + "step": 5289 + }, + { + "epoch": 0.6360848914807912, + "grad_norm": 2.3251888526592857, + "learning_rate": 1.2355328208934301e-06, + "loss": 0.8497, + "step": 5290 + }, + { + "epoch": 0.6362051343714303, + "grad_norm": 1.6626444043996802, + "learning_rate": 1.2348130578881728e-06, + "loss": 0.7263, + "step": 5291 + }, + { + "epoch": 0.6363253772620694, + "grad_norm": 2.303279537156039, + "learning_rate": 1.2340934109611664e-06, + "loss": 0.7656, + "step": 5292 + }, + { + "epoch": 0.6364456201527084, + "grad_norm": 2.773352565270767, + "learning_rate": 1.2333738802215798e-06, + "loss": 0.6997, + "step": 5293 + }, + { + "epoch": 0.6365658630433476, + "grad_norm": 2.2217037258314924, + "learning_rate": 1.2326544657785668e-06, + "loss": 0.816, + "step": 5294 + }, + { + "epoch": 0.6366861059339867, + "grad_norm": 2.2060036756773664, + "learning_rate": 1.2319351677412608e-06, + "loss": 0.7515, + "step": 5295 + }, + { + "epoch": 0.6368063488246257, + "grad_norm": 1.910026436667705, + "learning_rate": 1.2312159862187796e-06, + "loss": 0.7435, + "step": 5296 + }, + { + "epoch": 0.6369265917152649, + "grad_norm": 1.5419781945761157, + "learning_rate": 1.2304969213202217e-06, + "loss": 0.7616, + "step": 5297 + }, + { + "epoch": 0.6370468346059039, + "grad_norm": 4.301650400368699, + "learning_rate": 1.2297779731546692e-06, + "loss": 0.7973, + "step": 5298 + }, + { + "epoch": 0.637167077496543, + "grad_norm": 2.3450914740056215, + "learning_rate": 1.2290591418311853e-06, + "loss": 0.784, + "step": 5299 + }, + { + "epoch": 0.637287320387182, + "grad_norm": 1.677373951754181, + "learning_rate": 1.2283404274588172e-06, + "loss": 0.7177, + "step": 5300 + }, + { + "epoch": 0.6374075632778212, + "grad_norm": 0.7610856775534312, + "learning_rate": 1.227621830146592e-06, + "loss": 0.5278, + "step": 5301 + }, + { + "epoch": 0.6375278061684603, + "grad_norm": 1.792191081222242, + "learning_rate": 1.2269033500035217e-06, + "loss": 0.7967, + "step": 5302 + }, + { + "epoch": 0.6376480490590993, + "grad_norm": 1.8166071298741764, + "learning_rate": 1.2261849871385988e-06, + "loss": 0.7415, + "step": 5303 + }, + { + "epoch": 0.6377682919497385, + "grad_norm": 2.1436157495929806, + "learning_rate": 1.2254667416607972e-06, + "loss": 0.6373, + "step": 5304 + }, + { + "epoch": 0.6378885348403776, + "grad_norm": 2.0739614898284646, + "learning_rate": 1.2247486136790756e-06, + "loss": 0.8332, + "step": 5305 + }, + { + "epoch": 0.6380087777310166, + "grad_norm": 2.494915560285378, + "learning_rate": 1.2240306033023726e-06, + "loss": 0.8111, + "step": 5306 + }, + { + "epoch": 0.6381290206216558, + "grad_norm": 1.7918022533242055, + "learning_rate": 1.223312710639611e-06, + "loss": 0.7276, + "step": 5307 + }, + { + "epoch": 0.6382492635122948, + "grad_norm": 2.0297814562564005, + "learning_rate": 1.2225949357996928e-06, + "loss": 0.8721, + "step": 5308 + }, + { + "epoch": 0.6383695064029339, + "grad_norm": 1.6269000755719436, + "learning_rate": 1.221877278891505e-06, + "loss": 0.8022, + "step": 5309 + }, + { + "epoch": 0.638489749293573, + "grad_norm": 2.0700827417100105, + "learning_rate": 1.221159740023915e-06, + "loss": 0.7182, + "step": 5310 + }, + { + "epoch": 0.6386099921842121, + "grad_norm": 2.0769748803549573, + "learning_rate": 1.2204423193057735e-06, + "loss": 0.7336, + "step": 5311 + }, + { + "epoch": 0.6387302350748512, + "grad_norm": 0.8678148423336014, + "learning_rate": 1.2197250168459122e-06, + "loss": 0.6335, + "step": 5312 + }, + { + "epoch": 0.6388504779654903, + "grad_norm": 2.0940084724874257, + "learning_rate": 1.2190078327531454e-06, + "loss": 0.7515, + "step": 5313 + }, + { + "epoch": 0.6389707208561294, + "grad_norm": 1.4174831599289817, + "learning_rate": 1.2182907671362697e-06, + "loss": 0.7282, + "step": 5314 + }, + { + "epoch": 0.6390909637467684, + "grad_norm": 2.223403563679437, + "learning_rate": 1.2175738201040626e-06, + "loss": 0.7917, + "step": 5315 + }, + { + "epoch": 0.6392112066374076, + "grad_norm": 2.0511545186715097, + "learning_rate": 1.2168569917652855e-06, + "loss": 0.7874, + "step": 5316 + }, + { + "epoch": 0.6393314495280467, + "grad_norm": 1.5746848502556032, + "learning_rate": 1.2161402822286797e-06, + "loss": 0.6449, + "step": 5317 + }, + { + "epoch": 0.6394516924186857, + "grad_norm": 1.8630376052705817, + "learning_rate": 1.2154236916029703e-06, + "loss": 0.7901, + "step": 5318 + }, + { + "epoch": 0.6395719353093249, + "grad_norm": 3.765495456350906, + "learning_rate": 1.2147072199968627e-06, + "loss": 0.745, + "step": 5319 + }, + { + "epoch": 0.6396921781999639, + "grad_norm": 1.7307977987849796, + "learning_rate": 1.2139908675190454e-06, + "loss": 0.7164, + "step": 5320 + }, + { + "epoch": 0.639812421090603, + "grad_norm": 1.8162006220764095, + "learning_rate": 1.2132746342781883e-06, + "loss": 0.7534, + "step": 5321 + }, + { + "epoch": 0.6399326639812422, + "grad_norm": 3.0694428199851993, + "learning_rate": 1.2125585203829442e-06, + "loss": 0.8006, + "step": 5322 + }, + { + "epoch": 0.6400529068718812, + "grad_norm": 1.9468968691753326, + "learning_rate": 1.211842525941946e-06, + "loss": 0.7439, + "step": 5323 + }, + { + "epoch": 0.6401731497625203, + "grad_norm": 2.057433407564507, + "learning_rate": 1.2111266510638105e-06, + "loss": 0.791, + "step": 5324 + }, + { + "epoch": 0.6402933926531594, + "grad_norm": 1.731657149711258, + "learning_rate": 1.2104108958571346e-06, + "loss": 0.8023, + "step": 5325 + }, + { + "epoch": 0.6404136355437985, + "grad_norm": 1.5950182188422553, + "learning_rate": 1.2096952604304975e-06, + "loss": 0.76, + "step": 5326 + }, + { + "epoch": 0.6405338784344375, + "grad_norm": 2.110172488317114, + "learning_rate": 1.2089797448924616e-06, + "loss": 0.7045, + "step": 5327 + }, + { + "epoch": 0.6406541213250767, + "grad_norm": 6.346867138450634, + "learning_rate": 1.2082643493515692e-06, + "loss": 0.6666, + "step": 5328 + }, + { + "epoch": 0.6407743642157158, + "grad_norm": 1.7457673111163405, + "learning_rate": 1.207549073916346e-06, + "loss": 0.8232, + "step": 5329 + }, + { + "epoch": 0.6408946071063548, + "grad_norm": 2.1541397161906923, + "learning_rate": 1.2068339186952976e-06, + "loss": 0.7764, + "step": 5330 + }, + { + "epoch": 0.6410148499969939, + "grad_norm": 2.128745312688374, + "learning_rate": 1.2061188837969136e-06, + "loss": 0.7363, + "step": 5331 + }, + { + "epoch": 0.641135092887633, + "grad_norm": 4.419631658542642, + "learning_rate": 1.2054039693296631e-06, + "loss": 0.8493, + "step": 5332 + }, + { + "epoch": 0.6412553357782721, + "grad_norm": 1.6668833432812995, + "learning_rate": 1.2046891754019992e-06, + "loss": 0.8099, + "step": 5333 + }, + { + "epoch": 0.6413755786689112, + "grad_norm": 2.3517816891332166, + "learning_rate": 1.2039745021223548e-06, + "loss": 0.8257, + "step": 5334 + }, + { + "epoch": 0.6414958215595503, + "grad_norm": 0.8056064284546418, + "learning_rate": 1.2032599495991456e-06, + "loss": 0.5708, + "step": 5335 + }, + { + "epoch": 0.6416160644501894, + "grad_norm": 2.0073405501532497, + "learning_rate": 1.2025455179407685e-06, + "loss": 0.7028, + "step": 5336 + }, + { + "epoch": 0.6417363073408284, + "grad_norm": 2.5787213184945466, + "learning_rate": 1.2018312072556022e-06, + "loss": 0.7435, + "step": 5337 + }, + { + "epoch": 0.6418565502314676, + "grad_norm": 2.0058734005564904, + "learning_rate": 1.2011170176520077e-06, + "loss": 0.7496, + "step": 5338 + }, + { + "epoch": 0.6419767931221066, + "grad_norm": 1.6451158367986114, + "learning_rate": 1.2004029492383256e-06, + "loss": 0.816, + "step": 5339 + }, + { + "epoch": 0.6420970360127457, + "grad_norm": 1.9615849493740616, + "learning_rate": 1.1996890021228814e-06, + "loss": 0.7424, + "step": 5340 + }, + { + "epoch": 0.6422172789033849, + "grad_norm": 1.560636248076346, + "learning_rate": 1.1989751764139785e-06, + "loss": 0.7009, + "step": 5341 + }, + { + "epoch": 0.6423375217940239, + "grad_norm": 1.7545101086294599, + "learning_rate": 1.1982614722199044e-06, + "loss": 0.8263, + "step": 5342 + }, + { + "epoch": 0.642457764684663, + "grad_norm": 2.4500373798555013, + "learning_rate": 1.1975478896489276e-06, + "loss": 0.7816, + "step": 5343 + }, + { + "epoch": 0.6425780075753021, + "grad_norm": 1.9831496532986461, + "learning_rate": 1.1968344288092981e-06, + "loss": 0.7666, + "step": 5344 + }, + { + "epoch": 0.6426982504659412, + "grad_norm": 2.076335279721736, + "learning_rate": 1.1961210898092468e-06, + "loss": 0.651, + "step": 5345 + }, + { + "epoch": 0.6428184933565803, + "grad_norm": 2.3146702660534086, + "learning_rate": 1.1954078727569874e-06, + "loss": 0.8055, + "step": 5346 + }, + { + "epoch": 0.6429387362472194, + "grad_norm": 1.696487933389444, + "learning_rate": 1.1946947777607141e-06, + "loss": 0.786, + "step": 5347 + }, + { + "epoch": 0.6430589791378585, + "grad_norm": 2.1210424933378684, + "learning_rate": 1.1939818049286024e-06, + "loss": 0.8095, + "step": 5348 + }, + { + "epoch": 0.6431792220284975, + "grad_norm": 1.8007789992444583, + "learning_rate": 1.1932689543688101e-06, + "loss": 0.759, + "step": 5349 + }, + { + "epoch": 0.6432994649191367, + "grad_norm": 1.6833413769117493, + "learning_rate": 1.1925562261894756e-06, + "loss": 0.7241, + "step": 5350 + }, + { + "epoch": 0.6434197078097758, + "grad_norm": 2.3802707811395387, + "learning_rate": 1.1918436204987207e-06, + "loss": 0.7818, + "step": 5351 + }, + { + "epoch": 0.6435399507004148, + "grad_norm": 4.580959236941274, + "learning_rate": 1.191131137404645e-06, + "loss": 0.8196, + "step": 5352 + }, + { + "epoch": 0.643660193591054, + "grad_norm": 2.555240116370109, + "learning_rate": 1.190418777015333e-06, + "loss": 0.7769, + "step": 5353 + }, + { + "epoch": 0.643780436481693, + "grad_norm": 1.5365797436692914, + "learning_rate": 1.1897065394388487e-06, + "loss": 0.7418, + "step": 5354 + }, + { + "epoch": 0.6439006793723321, + "grad_norm": 1.5513522924836534, + "learning_rate": 1.1889944247832385e-06, + "loss": 0.7719, + "step": 5355 + }, + { + "epoch": 0.6440209222629713, + "grad_norm": 4.808228814340546, + "learning_rate": 1.1882824331565283e-06, + "loss": 0.7096, + "step": 5356 + }, + { + "epoch": 0.6441411651536103, + "grad_norm": 2.225284571076636, + "learning_rate": 1.1875705646667287e-06, + "loss": 0.8898, + "step": 5357 + }, + { + "epoch": 0.6442614080442494, + "grad_norm": 2.2548384036337206, + "learning_rate": 1.1868588194218282e-06, + "loss": 0.7621, + "step": 5358 + }, + { + "epoch": 0.6443816509348885, + "grad_norm": 1.930981221677168, + "learning_rate": 1.1861471975297979e-06, + "loss": 0.7511, + "step": 5359 + }, + { + "epoch": 0.6445018938255276, + "grad_norm": 1.7234201964740061, + "learning_rate": 1.185435699098591e-06, + "loss": 0.7127, + "step": 5360 + }, + { + "epoch": 0.6446221367161666, + "grad_norm": 2.2552395319964766, + "learning_rate": 1.1847243242361403e-06, + "loss": 0.7813, + "step": 5361 + }, + { + "epoch": 0.6447423796068057, + "grad_norm": 1.7347769056785332, + "learning_rate": 1.1840130730503624e-06, + "loss": 0.7852, + "step": 5362 + }, + { + "epoch": 0.6448626224974449, + "grad_norm": 1.8884475401201615, + "learning_rate": 1.1833019456491518e-06, + "loss": 0.7567, + "step": 5363 + }, + { + "epoch": 0.6449828653880839, + "grad_norm": 2.386194649199773, + "learning_rate": 1.1825909421403871e-06, + "loss": 0.7922, + "step": 5364 + }, + { + "epoch": 0.645103108278723, + "grad_norm": 1.849962597546217, + "learning_rate": 1.181880062631926e-06, + "loss": 0.7665, + "step": 5365 + }, + { + "epoch": 0.6452233511693621, + "grad_norm": 2.537393907856628, + "learning_rate": 1.1811693072316093e-06, + "loss": 0.8526, + "step": 5366 + }, + { + "epoch": 0.6453435940600012, + "grad_norm": 3.0902634828691693, + "learning_rate": 1.1804586760472574e-06, + "loss": 0.8467, + "step": 5367 + }, + { + "epoch": 0.6454638369506402, + "grad_norm": 2.415185660892514, + "learning_rate": 1.1797481691866736e-06, + "loss": 0.8059, + "step": 5368 + }, + { + "epoch": 0.6455840798412794, + "grad_norm": 2.0417339028426524, + "learning_rate": 1.1790377867576393e-06, + "loss": 0.835, + "step": 5369 + }, + { + "epoch": 0.6457043227319185, + "grad_norm": 2.1575077092993284, + "learning_rate": 1.1783275288679203e-06, + "loss": 0.766, + "step": 5370 + }, + { + "epoch": 0.6458245656225575, + "grad_norm": 0.8521229710643597, + "learning_rate": 1.177617395625262e-06, + "loss": 0.5715, + "step": 5371 + }, + { + "epoch": 0.6459448085131967, + "grad_norm": 1.7500569470540464, + "learning_rate": 1.1769073871373908e-06, + "loss": 0.7555, + "step": 5372 + }, + { + "epoch": 0.6460650514038357, + "grad_norm": 2.053766892886674, + "learning_rate": 1.176197503512015e-06, + "loss": 0.8467, + "step": 5373 + }, + { + "epoch": 0.6461852942944748, + "grad_norm": 2.4745615869497817, + "learning_rate": 1.1754877448568223e-06, + "loss": 0.825, + "step": 5374 + }, + { + "epoch": 0.646305537185114, + "grad_norm": 2.076259607594841, + "learning_rate": 1.1747781112794837e-06, + "loss": 0.9007, + "step": 5375 + }, + { + "epoch": 0.646425780075753, + "grad_norm": 1.9783737869861537, + "learning_rate": 1.1740686028876487e-06, + "loss": 0.8311, + "step": 5376 + }, + { + "epoch": 0.6465460229663921, + "grad_norm": 2.2098676162010324, + "learning_rate": 1.1733592197889507e-06, + "loss": 0.7493, + "step": 5377 + }, + { + "epoch": 0.6466662658570312, + "grad_norm": 1.9056193449812895, + "learning_rate": 1.1726499620910014e-06, + "loss": 0.7271, + "step": 5378 + }, + { + "epoch": 0.6467865087476703, + "grad_norm": 2.0670921759277543, + "learning_rate": 1.1719408299013955e-06, + "loss": 0.7829, + "step": 5379 + }, + { + "epoch": 0.6469067516383094, + "grad_norm": 2.285267153926042, + "learning_rate": 1.1712318233277067e-06, + "loss": 0.7656, + "step": 5380 + }, + { + "epoch": 0.6470269945289485, + "grad_norm": 0.7736783278861706, + "learning_rate": 1.1705229424774916e-06, + "loss": 0.5793, + "step": 5381 + }, + { + "epoch": 0.6471472374195876, + "grad_norm": 1.9174776711090171, + "learning_rate": 1.1698141874582867e-06, + "loss": 0.6396, + "step": 5382 + }, + { + "epoch": 0.6472674803102266, + "grad_norm": 2.0744237995987698, + "learning_rate": 1.169105558377609e-06, + "loss": 0.7201, + "step": 5383 + }, + { + "epoch": 0.6473877232008658, + "grad_norm": 1.6562107729315099, + "learning_rate": 1.1683970553429587e-06, + "loss": 0.7871, + "step": 5384 + }, + { + "epoch": 0.6475079660915048, + "grad_norm": 1.9442792852589141, + "learning_rate": 1.1676886784618128e-06, + "loss": 0.8244, + "step": 5385 + }, + { + "epoch": 0.6476282089821439, + "grad_norm": 2.2665401320574605, + "learning_rate": 1.1669804278416332e-06, + "loss": 0.8435, + "step": 5386 + }, + { + "epoch": 0.6477484518727831, + "grad_norm": 1.9749738619076724, + "learning_rate": 1.1662723035898602e-06, + "loss": 0.7163, + "step": 5387 + }, + { + "epoch": 0.6478686947634221, + "grad_norm": 2.1820292045673817, + "learning_rate": 1.165564305813915e-06, + "loss": 0.8237, + "step": 5388 + }, + { + "epoch": 0.6479889376540612, + "grad_norm": 1.9560519450297487, + "learning_rate": 1.1648564346212019e-06, + "loss": 0.8189, + "step": 5389 + }, + { + "epoch": 0.6481091805447003, + "grad_norm": 1.8944615931339166, + "learning_rate": 1.164148690119104e-06, + "loss": 0.7629, + "step": 5390 + }, + { + "epoch": 0.6482294234353394, + "grad_norm": 2.090322393347467, + "learning_rate": 1.163441072414985e-06, + "loss": 0.7403, + "step": 5391 + }, + { + "epoch": 0.6483496663259785, + "grad_norm": 1.8958958067049971, + "learning_rate": 1.16273358161619e-06, + "loss": 0.7035, + "step": 5392 + }, + { + "epoch": 0.6484699092166175, + "grad_norm": 1.8592997944472844, + "learning_rate": 1.1620262178300446e-06, + "loss": 0.8363, + "step": 5393 + }, + { + "epoch": 0.6485901521072567, + "grad_norm": 1.8106237763468855, + "learning_rate": 1.1613189811638563e-06, + "loss": 0.7606, + "step": 5394 + }, + { + "epoch": 0.6487103949978957, + "grad_norm": 1.760976766234828, + "learning_rate": 1.1606118717249117e-06, + "loss": 0.7862, + "step": 5395 + }, + { + "epoch": 0.6488306378885348, + "grad_norm": 1.7300169206354066, + "learning_rate": 1.1599048896204787e-06, + "loss": 0.6833, + "step": 5396 + }, + { + "epoch": 0.648950880779174, + "grad_norm": 1.817805057200212, + "learning_rate": 1.1591980349578061e-06, + "loss": 0.8101, + "step": 5397 + }, + { + "epoch": 0.649071123669813, + "grad_norm": 0.7467548177900218, + "learning_rate": 1.158491307844123e-06, + "loss": 0.544, + "step": 5398 + }, + { + "epoch": 0.6491913665604521, + "grad_norm": 1.6673833101041047, + "learning_rate": 1.1577847083866387e-06, + "loss": 0.8431, + "step": 5399 + }, + { + "epoch": 0.6493116094510912, + "grad_norm": 1.9628699761818142, + "learning_rate": 1.1570782366925453e-06, + "loss": 0.7284, + "step": 5400 + }, + { + "epoch": 0.6494318523417303, + "grad_norm": 1.6675726176725203, + "learning_rate": 1.1563718928690132e-06, + "loss": 0.7593, + "step": 5401 + }, + { + "epoch": 0.6495520952323693, + "grad_norm": 2.4990398470543695, + "learning_rate": 1.1556656770231942e-06, + "loss": 0.7161, + "step": 5402 + }, + { + "epoch": 0.6496723381230085, + "grad_norm": 1.8281306079224215, + "learning_rate": 1.1549595892622207e-06, + "loss": 0.7622, + "step": 5403 + }, + { + "epoch": 0.6497925810136476, + "grad_norm": 0.8270499971399341, + "learning_rate": 1.1542536296932047e-06, + "loss": 0.5912, + "step": 5404 + }, + { + "epoch": 0.6499128239042866, + "grad_norm": 1.7349473850726242, + "learning_rate": 1.1535477984232414e-06, + "loss": 0.7093, + "step": 5405 + }, + { + "epoch": 0.6500330667949258, + "grad_norm": 1.864351361157148, + "learning_rate": 1.152842095559404e-06, + "loss": 0.7712, + "step": 5406 + }, + { + "epoch": 0.6501533096855648, + "grad_norm": 1.6711647450193914, + "learning_rate": 1.1521365212087474e-06, + "loss": 0.7717, + "step": 5407 + }, + { + "epoch": 0.6502735525762039, + "grad_norm": 1.7877993726456767, + "learning_rate": 1.1514310754783062e-06, + "loss": 0.7112, + "step": 5408 + }, + { + "epoch": 0.6503937954668431, + "grad_norm": 3.696033883199568, + "learning_rate": 1.1507257584750964e-06, + "loss": 0.7347, + "step": 5409 + }, + { + "epoch": 0.6505140383574821, + "grad_norm": 1.8778291339108972, + "learning_rate": 1.150020570306113e-06, + "loss": 0.7784, + "step": 5410 + }, + { + "epoch": 0.6506342812481212, + "grad_norm": 2.875511621479926, + "learning_rate": 1.1493155110783338e-06, + "loss": 0.7532, + "step": 5411 + }, + { + "epoch": 0.6507545241387603, + "grad_norm": 2.1023821171974006, + "learning_rate": 1.1486105808987155e-06, + "loss": 0.7146, + "step": 5412 + }, + { + "epoch": 0.6508747670293994, + "grad_norm": 1.7813216747118465, + "learning_rate": 1.1479057798741947e-06, + "loss": 0.8127, + "step": 5413 + }, + { + "epoch": 0.6509950099200384, + "grad_norm": 0.8025613491577994, + "learning_rate": 1.14720110811169e-06, + "loss": 0.5335, + "step": 5414 + }, + { + "epoch": 0.6511152528106776, + "grad_norm": 2.0320037617010307, + "learning_rate": 1.146496565718098e-06, + "loss": 0.7723, + "step": 5415 + }, + { + "epoch": 0.6512354957013167, + "grad_norm": 2.744128287616839, + "learning_rate": 1.1457921528002996e-06, + "loss": 0.7607, + "step": 5416 + }, + { + "epoch": 0.6513557385919557, + "grad_norm": 2.736799636503595, + "learning_rate": 1.1450878694651522e-06, + "loss": 0.7257, + "step": 5417 + }, + { + "epoch": 0.6514759814825949, + "grad_norm": 2.2799806590682072, + "learning_rate": 1.1443837158194954e-06, + "loss": 0.6404, + "step": 5418 + }, + { + "epoch": 0.651596224373234, + "grad_norm": 1.892767582751367, + "learning_rate": 1.1436796919701484e-06, + "loss": 0.7478, + "step": 5419 + }, + { + "epoch": 0.651716467263873, + "grad_norm": 1.9581461336740815, + "learning_rate": 1.1429757980239115e-06, + "loss": 0.6204, + "step": 5420 + }, + { + "epoch": 0.6518367101545122, + "grad_norm": 2.785410480211835, + "learning_rate": 1.1422720340875636e-06, + "loss": 0.823, + "step": 5421 + }, + { + "epoch": 0.6519569530451512, + "grad_norm": 1.897220299854611, + "learning_rate": 1.1415684002678671e-06, + "loss": 0.7941, + "step": 5422 + }, + { + "epoch": 0.6520771959357903, + "grad_norm": 2.206596939429715, + "learning_rate": 1.1408648966715617e-06, + "loss": 0.7772, + "step": 5423 + }, + { + "epoch": 0.6521974388264293, + "grad_norm": 1.8570756559475012, + "learning_rate": 1.1401615234053683e-06, + "loss": 0.7314, + "step": 5424 + }, + { + "epoch": 0.6523176817170685, + "grad_norm": 11.062881376944777, + "learning_rate": 1.1394582805759885e-06, + "loss": 0.7619, + "step": 5425 + }, + { + "epoch": 0.6524379246077076, + "grad_norm": 3.234889040286592, + "learning_rate": 1.1387551682901022e-06, + "loss": 0.76, + "step": 5426 + }, + { + "epoch": 0.6525581674983466, + "grad_norm": 2.06022634318272, + "learning_rate": 1.138052186654373e-06, + "loss": 0.7099, + "step": 5427 + }, + { + "epoch": 0.6526784103889858, + "grad_norm": 2.385318726329456, + "learning_rate": 1.1373493357754417e-06, + "loss": 0.8861, + "step": 5428 + }, + { + "epoch": 0.6527986532796248, + "grad_norm": 1.912458471707519, + "learning_rate": 1.1366466157599303e-06, + "loss": 0.775, + "step": 5429 + }, + { + "epoch": 0.6529188961702639, + "grad_norm": 2.115320262063452, + "learning_rate": 1.1359440267144412e-06, + "loss": 0.7663, + "step": 5430 + }, + { + "epoch": 0.653039139060903, + "grad_norm": 2.015808806799109, + "learning_rate": 1.1352415687455556e-06, + "loss": 0.7435, + "step": 5431 + }, + { + "epoch": 0.6531593819515421, + "grad_norm": 2.5743100695700103, + "learning_rate": 1.1345392419598362e-06, + "loss": 0.6444, + "step": 5432 + }, + { + "epoch": 0.6532796248421812, + "grad_norm": 2.6637592758382387, + "learning_rate": 1.1338370464638263e-06, + "loss": 0.7203, + "step": 5433 + }, + { + "epoch": 0.6533998677328203, + "grad_norm": 5.394004227005403, + "learning_rate": 1.1331349823640474e-06, + "loss": 0.6379, + "step": 5434 + }, + { + "epoch": 0.6535201106234594, + "grad_norm": 4.29487746537977, + "learning_rate": 1.132433049767003e-06, + "loss": 0.7834, + "step": 5435 + }, + { + "epoch": 0.6536403535140984, + "grad_norm": 1.7190144422145248, + "learning_rate": 1.1317312487791748e-06, + "loss": 0.8128, + "step": 5436 + }, + { + "epoch": 0.6537605964047376, + "grad_norm": 2.1874925885756142, + "learning_rate": 1.1310295795070253e-06, + "loss": 0.7368, + "step": 5437 + }, + { + "epoch": 0.6538808392953767, + "grad_norm": 1.8716252792461119, + "learning_rate": 1.1303280420569982e-06, + "loss": 0.8129, + "step": 5438 + }, + { + "epoch": 0.6540010821860157, + "grad_norm": 1.9223594365119032, + "learning_rate": 1.1296266365355158e-06, + "loss": 0.7734, + "step": 5439 + }, + { + "epoch": 0.6541213250766549, + "grad_norm": 1.897545498992277, + "learning_rate": 1.1289253630489806e-06, + "loss": 0.7384, + "step": 5440 + }, + { + "epoch": 0.6542415679672939, + "grad_norm": 2.595682733942198, + "learning_rate": 1.1282242217037753e-06, + "loss": 0.7287, + "step": 5441 + }, + { + "epoch": 0.654361810857933, + "grad_norm": 2.614472942884184, + "learning_rate": 1.127523212606262e-06, + "loss": 0.6215, + "step": 5442 + }, + { + "epoch": 0.6544820537485722, + "grad_norm": 1.6501143907139912, + "learning_rate": 1.1268223358627835e-06, + "loss": 0.7364, + "step": 5443 + }, + { + "epoch": 0.6546022966392112, + "grad_norm": 1.7811618582350464, + "learning_rate": 1.126121591579663e-06, + "loss": 0.7173, + "step": 5444 + }, + { + "epoch": 0.6547225395298503, + "grad_norm": 1.8340809627038512, + "learning_rate": 1.1254209798632018e-06, + "loss": 0.6929, + "step": 5445 + }, + { + "epoch": 0.6548427824204894, + "grad_norm": 1.9621568513442915, + "learning_rate": 1.124720500819683e-06, + "loss": 0.8438, + "step": 5446 + }, + { + "epoch": 0.6549630253111285, + "grad_norm": 1.9245978195373115, + "learning_rate": 1.1240201545553682e-06, + "loss": 0.8258, + "step": 5447 + }, + { + "epoch": 0.6550832682017675, + "grad_norm": 1.7651027480954837, + "learning_rate": 1.1233199411764987e-06, + "loss": 0.7335, + "step": 5448 + }, + { + "epoch": 0.6552035110924067, + "grad_norm": 3.131723419157043, + "learning_rate": 1.1226198607892978e-06, + "loss": 0.694, + "step": 5449 + }, + { + "epoch": 0.6553237539830458, + "grad_norm": 1.8725233251241906, + "learning_rate": 1.1219199134999664e-06, + "loss": 0.8034, + "step": 5450 + }, + { + "epoch": 0.6554439968736848, + "grad_norm": 2.0971643876388226, + "learning_rate": 1.1212200994146863e-06, + "loss": 0.7911, + "step": 5451 + }, + { + "epoch": 0.655564239764324, + "grad_norm": 2.295375204560665, + "learning_rate": 1.120520418639618e-06, + "loss": 0.7619, + "step": 5452 + }, + { + "epoch": 0.655684482654963, + "grad_norm": 1.9154577658310545, + "learning_rate": 1.119820871280903e-06, + "loss": 0.8345, + "step": 5453 + }, + { + "epoch": 0.6558047255456021, + "grad_norm": 2.080391357907662, + "learning_rate": 1.1191214574446614e-06, + "loss": 0.7384, + "step": 5454 + }, + { + "epoch": 0.6559249684362413, + "grad_norm": 1.7206063389485, + "learning_rate": 1.118422177236995e-06, + "loss": 0.8023, + "step": 5455 + }, + { + "epoch": 0.6560452113268803, + "grad_norm": 2.330498733353204, + "learning_rate": 1.1177230307639835e-06, + "loss": 0.8607, + "step": 5456 + }, + { + "epoch": 0.6561654542175194, + "grad_norm": 2.7606155803676033, + "learning_rate": 1.1170240181316865e-06, + "loss": 0.7846, + "step": 5457 + }, + { + "epoch": 0.6562856971081584, + "grad_norm": 2.266025376294794, + "learning_rate": 1.1163251394461442e-06, + "loss": 0.8027, + "step": 5458 + }, + { + "epoch": 0.6564059399987976, + "grad_norm": 1.899503683216674, + "learning_rate": 1.1156263948133746e-06, + "loss": 0.827, + "step": 5459 + }, + { + "epoch": 0.6565261828894366, + "grad_norm": 1.7258969228030014, + "learning_rate": 1.1149277843393787e-06, + "loss": 0.7784, + "step": 5460 + }, + { + "epoch": 0.6566464257800757, + "grad_norm": 2.2783663565589505, + "learning_rate": 1.1142293081301342e-06, + "loss": 0.6363, + "step": 5461 + }, + { + "epoch": 0.6567666686707149, + "grad_norm": 1.8402305892876245, + "learning_rate": 1.1135309662915995e-06, + "loss": 0.6794, + "step": 5462 + }, + { + "epoch": 0.6568869115613539, + "grad_norm": 2.2796099267197594, + "learning_rate": 1.112832758929712e-06, + "loss": 0.6095, + "step": 5463 + }, + { + "epoch": 0.657007154451993, + "grad_norm": 3.1061146574258465, + "learning_rate": 1.11213468615039e-06, + "loss": 0.751, + "step": 5464 + }, + { + "epoch": 0.6571273973426321, + "grad_norm": 1.5723162719909378, + "learning_rate": 1.1114367480595292e-06, + "loss": 0.7542, + "step": 5465 + }, + { + "epoch": 0.6572476402332712, + "grad_norm": 2.6424389990746366, + "learning_rate": 1.1107389447630086e-06, + "loss": 0.8165, + "step": 5466 + }, + { + "epoch": 0.6573678831239103, + "grad_norm": 5.942909426773555, + "learning_rate": 1.1100412763666818e-06, + "loss": 0.7862, + "step": 5467 + }, + { + "epoch": 0.6574881260145494, + "grad_norm": 2.1300589984011995, + "learning_rate": 1.1093437429763865e-06, + "loss": 0.8015, + "step": 5468 + }, + { + "epoch": 0.6576083689051885, + "grad_norm": 2.217908565380554, + "learning_rate": 1.1086463446979361e-06, + "loss": 0.7427, + "step": 5469 + }, + { + "epoch": 0.6577286117958275, + "grad_norm": 1.9616434110928356, + "learning_rate": 1.1079490816371277e-06, + "loss": 0.7758, + "step": 5470 + }, + { + "epoch": 0.6578488546864667, + "grad_norm": 2.213464916619026, + "learning_rate": 1.1072519538997352e-06, + "loss": 0.7493, + "step": 5471 + }, + { + "epoch": 0.6579690975771058, + "grad_norm": 1.6939337261256295, + "learning_rate": 1.1065549615915095e-06, + "loss": 0.8293, + "step": 5472 + }, + { + "epoch": 0.6580893404677448, + "grad_norm": 2.1989693915304267, + "learning_rate": 1.105858104818187e-06, + "loss": 0.7848, + "step": 5473 + }, + { + "epoch": 0.658209583358384, + "grad_norm": 2.431818478372794, + "learning_rate": 1.105161383685478e-06, + "loss": 0.7526, + "step": 5474 + }, + { + "epoch": 0.658329826249023, + "grad_norm": 0.7335002234236904, + "learning_rate": 1.1044647982990771e-06, + "loss": 0.5635, + "step": 5475 + }, + { + "epoch": 0.6584500691396621, + "grad_norm": 2.2305058411087963, + "learning_rate": 1.1037683487646536e-06, + "loss": 0.6517, + "step": 5476 + }, + { + "epoch": 0.6585703120303013, + "grad_norm": 1.8057636173711356, + "learning_rate": 1.1030720351878583e-06, + "loss": 0.7763, + "step": 5477 + }, + { + "epoch": 0.6586905549209403, + "grad_norm": 0.8157348792650169, + "learning_rate": 1.102375857674323e-06, + "loss": 0.5769, + "step": 5478 + }, + { + "epoch": 0.6588107978115794, + "grad_norm": 1.7076144760653085, + "learning_rate": 1.1016798163296561e-06, + "loss": 0.9039, + "step": 5479 + }, + { + "epoch": 0.6589310407022185, + "grad_norm": 1.885888211065282, + "learning_rate": 1.1009839112594471e-06, + "loss": 0.66, + "step": 5480 + }, + { + "epoch": 0.6590512835928576, + "grad_norm": 2.0840597666409177, + "learning_rate": 1.1002881425692638e-06, + "loss": 0.7198, + "step": 5481 + }, + { + "epoch": 0.6591715264834966, + "grad_norm": 1.8967922286902117, + "learning_rate": 1.0995925103646532e-06, + "loss": 0.7573, + "step": 5482 + }, + { + "epoch": 0.6592917693741358, + "grad_norm": 2.958819515054688, + "learning_rate": 1.0988970147511437e-06, + "loss": 0.6725, + "step": 5483 + }, + { + "epoch": 0.6594120122647749, + "grad_norm": 2.7000103243877676, + "learning_rate": 1.0982016558342405e-06, + "loss": 0.8127, + "step": 5484 + }, + { + "epoch": 0.6595322551554139, + "grad_norm": 1.8610384896883025, + "learning_rate": 1.0975064337194291e-06, + "loss": 0.7139, + "step": 5485 + }, + { + "epoch": 0.6596524980460531, + "grad_norm": 2.024966100715123, + "learning_rate": 1.0968113485121743e-06, + "loss": 0.7089, + "step": 5486 + }, + { + "epoch": 0.6597727409366921, + "grad_norm": 1.9135757761659584, + "learning_rate": 1.0961164003179185e-06, + "loss": 0.8037, + "step": 5487 + }, + { + "epoch": 0.6598929838273312, + "grad_norm": 1.9345319841372866, + "learning_rate": 1.0954215892420884e-06, + "loss": 0.8476, + "step": 5488 + }, + { + "epoch": 0.6600132267179702, + "grad_norm": 1.72749827990743, + "learning_rate": 1.094726915390082e-06, + "loss": 0.7061, + "step": 5489 + }, + { + "epoch": 0.6601334696086094, + "grad_norm": 1.6933021403741215, + "learning_rate": 1.0940323788672836e-06, + "loss": 0.6989, + "step": 5490 + }, + { + "epoch": 0.6602537124992485, + "grad_norm": 1.690494282687701, + "learning_rate": 1.0933379797790522e-06, + "loss": 0.7417, + "step": 5491 + }, + { + "epoch": 0.6603739553898875, + "grad_norm": 2.229874780671683, + "learning_rate": 1.0926437182307293e-06, + "loss": 0.7172, + "step": 5492 + }, + { + "epoch": 0.6604941982805267, + "grad_norm": 1.6358112741940232, + "learning_rate": 1.0919495943276338e-06, + "loss": 0.7836, + "step": 5493 + }, + { + "epoch": 0.6606144411711657, + "grad_norm": 2.404048259161856, + "learning_rate": 1.0912556081750611e-06, + "loss": 0.7682, + "step": 5494 + }, + { + "epoch": 0.6607346840618048, + "grad_norm": 2.6158997057607856, + "learning_rate": 1.0905617598782909e-06, + "loss": 0.766, + "step": 5495 + }, + { + "epoch": 0.660854926952444, + "grad_norm": 2.5234426020048075, + "learning_rate": 1.0898680495425775e-06, + "loss": 0.8134, + "step": 5496 + }, + { + "epoch": 0.660975169843083, + "grad_norm": 1.7447490933666396, + "learning_rate": 1.0891744772731594e-06, + "loss": 0.8062, + "step": 5497 + }, + { + "epoch": 0.6610954127337221, + "grad_norm": 1.6123258333997181, + "learning_rate": 1.088481043175248e-06, + "loss": 0.6611, + "step": 5498 + }, + { + "epoch": 0.6612156556243612, + "grad_norm": 1.692066090350424, + "learning_rate": 1.0877877473540368e-06, + "loss": 0.7564, + "step": 5499 + }, + { + "epoch": 0.6613358985150003, + "grad_norm": 2.472649018335841, + "learning_rate": 1.0870945899147002e-06, + "loss": 0.7281, + "step": 5500 + }, + { + "epoch": 0.6614561414056394, + "grad_norm": 1.7503250365173528, + "learning_rate": 1.0864015709623879e-06, + "loss": 0.7612, + "step": 5501 + }, + { + "epoch": 0.6615763842962785, + "grad_norm": 2.3417204143650356, + "learning_rate": 1.0857086906022313e-06, + "loss": 0.8096, + "step": 5502 + }, + { + "epoch": 0.6616966271869176, + "grad_norm": 2.2535339788495192, + "learning_rate": 1.0850159489393388e-06, + "loss": 0.7364, + "step": 5503 + }, + { + "epoch": 0.6618168700775566, + "grad_norm": 1.7900491184549443, + "learning_rate": 1.0843233460787992e-06, + "loss": 0.826, + "step": 5504 + }, + { + "epoch": 0.6619371129681958, + "grad_norm": 1.8036629547032066, + "learning_rate": 1.0836308821256805e-06, + "loss": 0.7817, + "step": 5505 + }, + { + "epoch": 0.6620573558588349, + "grad_norm": 2.092980770177666, + "learning_rate": 1.0829385571850282e-06, + "loss": 0.7837, + "step": 5506 + }, + { + "epoch": 0.6621775987494739, + "grad_norm": 2.5635213909770815, + "learning_rate": 1.0822463713618679e-06, + "loss": 0.8386, + "step": 5507 + }, + { + "epoch": 0.6622978416401131, + "grad_norm": 2.0452884028612948, + "learning_rate": 1.0815543247612034e-06, + "loss": 0.8548, + "step": 5508 + }, + { + "epoch": 0.6624180845307521, + "grad_norm": 1.6873954457989182, + "learning_rate": 1.0808624174880168e-06, + "loss": 0.8354, + "step": 5509 + }, + { + "epoch": 0.6625383274213912, + "grad_norm": 1.8122763002402464, + "learning_rate": 1.080170649647272e-06, + "loss": 0.8039, + "step": 5510 + }, + { + "epoch": 0.6626585703120303, + "grad_norm": 4.349429582049695, + "learning_rate": 1.0794790213439068e-06, + "loss": 0.6774, + "step": 5511 + }, + { + "epoch": 0.6627788132026694, + "grad_norm": 3.010086111820346, + "learning_rate": 1.078787532682843e-06, + "loss": 0.7921, + "step": 5512 + }, + { + "epoch": 0.6628990560933085, + "grad_norm": 2.291196694011541, + "learning_rate": 1.0780961837689773e-06, + "loss": 0.7568, + "step": 5513 + }, + { + "epoch": 0.6630192989839476, + "grad_norm": 1.5673562686371927, + "learning_rate": 1.0774049747071883e-06, + "loss": 0.7001, + "step": 5514 + }, + { + "epoch": 0.6631395418745867, + "grad_norm": 2.351237734043184, + "learning_rate": 1.076713905602332e-06, + "loss": 0.6846, + "step": 5515 + }, + { + "epoch": 0.6632597847652257, + "grad_norm": 2.2966800529115705, + "learning_rate": 1.07602297655924e-06, + "loss": 0.8196, + "step": 5516 + }, + { + "epoch": 0.6633800276558649, + "grad_norm": 1.8165499722626115, + "learning_rate": 1.0753321876827292e-06, + "loss": 0.8193, + "step": 5517 + }, + { + "epoch": 0.663500270546504, + "grad_norm": 2.1926130212167685, + "learning_rate": 1.0746415390775893e-06, + "loss": 0.7452, + "step": 5518 + }, + { + "epoch": 0.663620513437143, + "grad_norm": 2.1491477208549226, + "learning_rate": 1.0739510308485939e-06, + "loss": 0.768, + "step": 5519 + }, + { + "epoch": 0.6637407563277821, + "grad_norm": 0.8215748111793472, + "learning_rate": 1.07326066310049e-06, + "loss": 0.6262, + "step": 5520 + }, + { + "epoch": 0.6638609992184212, + "grad_norm": 1.9815830203541274, + "learning_rate": 1.0725704359380059e-06, + "loss": 0.7986, + "step": 5521 + }, + { + "epoch": 0.6639812421090603, + "grad_norm": 2.08281951926008, + "learning_rate": 1.0718803494658497e-06, + "loss": 0.7261, + "step": 5522 + }, + { + "epoch": 0.6641014849996993, + "grad_norm": 2.200879515057793, + "learning_rate": 1.071190403788707e-06, + "loss": 0.8414, + "step": 5523 + }, + { + "epoch": 0.6642217278903385, + "grad_norm": 1.8108406363543472, + "learning_rate": 1.0705005990112415e-06, + "loss": 0.7586, + "step": 5524 + }, + { + "epoch": 0.6643419707809776, + "grad_norm": 2.3481524001902097, + "learning_rate": 1.0698109352380957e-06, + "loss": 0.7514, + "step": 5525 + }, + { + "epoch": 0.6644622136716166, + "grad_norm": 2.4532250062032204, + "learning_rate": 1.0691214125738909e-06, + "loss": 0.7776, + "step": 5526 + }, + { + "epoch": 0.6645824565622558, + "grad_norm": 0.8082889590586151, + "learning_rate": 1.0684320311232287e-06, + "loss": 0.5752, + "step": 5527 + }, + { + "epoch": 0.6647026994528948, + "grad_norm": 2.172805806224689, + "learning_rate": 1.0677427909906865e-06, + "loss": 0.8201, + "step": 5528 + }, + { + "epoch": 0.6648229423435339, + "grad_norm": 1.729673043374122, + "learning_rate": 1.0670536922808216e-06, + "loss": 0.7241, + "step": 5529 + }, + { + "epoch": 0.6649431852341731, + "grad_norm": 1.9832355501876413, + "learning_rate": 1.06636473509817e-06, + "loss": 0.7208, + "step": 5530 + }, + { + "epoch": 0.6650634281248121, + "grad_norm": 1.9822282969104208, + "learning_rate": 1.0656759195472447e-06, + "loss": 0.8128, + "step": 5531 + }, + { + "epoch": 0.6651836710154512, + "grad_norm": 0.7828224379466407, + "learning_rate": 1.0649872457325414e-06, + "loss": 0.5981, + "step": 5532 + }, + { + "epoch": 0.6653039139060903, + "grad_norm": 0.8679281628696307, + "learning_rate": 1.0642987137585278e-06, + "loss": 0.5522, + "step": 5533 + }, + { + "epoch": 0.6654241567967294, + "grad_norm": 4.118675006652601, + "learning_rate": 1.0636103237296561e-06, + "loss": 0.8287, + "step": 5534 + }, + { + "epoch": 0.6655443996873684, + "grad_norm": 4.962910801870142, + "learning_rate": 1.062922075750353e-06, + "loss": 0.8416, + "step": 5535 + }, + { + "epoch": 0.6656646425780076, + "grad_norm": 2.155170004350039, + "learning_rate": 1.0622339699250267e-06, + "loss": 0.7199, + "step": 5536 + }, + { + "epoch": 0.6657848854686467, + "grad_norm": 1.7412218916158517, + "learning_rate": 1.0615460063580624e-06, + "loss": 0.7999, + "step": 5537 + }, + { + "epoch": 0.6659051283592857, + "grad_norm": 2.0683066962056404, + "learning_rate": 1.060858185153821e-06, + "loss": 0.7315, + "step": 5538 + }, + { + "epoch": 0.6660253712499249, + "grad_norm": 2.160539865491025, + "learning_rate": 1.0601705064166474e-06, + "loss": 0.7667, + "step": 5539 + }, + { + "epoch": 0.666145614140564, + "grad_norm": 1.9930322389497777, + "learning_rate": 1.0594829702508596e-06, + "loss": 0.7364, + "step": 5540 + }, + { + "epoch": 0.666265857031203, + "grad_norm": 1.8409385582463194, + "learning_rate": 1.0587955767607592e-06, + "loss": 0.553, + "step": 5541 + }, + { + "epoch": 0.6663860999218422, + "grad_norm": 2.39488449104834, + "learning_rate": 1.0581083260506206e-06, + "loss": 0.7787, + "step": 5542 + }, + { + "epoch": 0.6665063428124812, + "grad_norm": 2.214723829356569, + "learning_rate": 1.0574212182246993e-06, + "loss": 0.7685, + "step": 5543 + }, + { + "epoch": 0.6666265857031203, + "grad_norm": 2.844127085071551, + "learning_rate": 1.0567342533872303e-06, + "loss": 0.7591, + "step": 5544 + }, + { + "epoch": 0.6667468285937594, + "grad_norm": 1.7617311329205538, + "learning_rate": 1.0560474316424255e-06, + "loss": 0.8131, + "step": 5545 + }, + { + "epoch": 0.6668670714843985, + "grad_norm": 2.432527997599708, + "learning_rate": 1.0553607530944746e-06, + "loss": 0.738, + "step": 5546 + }, + { + "epoch": 0.6669873143750376, + "grad_norm": 2.119179889649202, + "learning_rate": 1.0546742178475463e-06, + "loss": 0.8993, + "step": 5547 + }, + { + "epoch": 0.6671075572656767, + "grad_norm": 1.7852960778029792, + "learning_rate": 1.0539878260057868e-06, + "loss": 0.8693, + "step": 5548 + }, + { + "epoch": 0.6672278001563158, + "grad_norm": 14.029683388844202, + "learning_rate": 1.0533015776733226e-06, + "loss": 0.6901, + "step": 5549 + }, + { + "epoch": 0.6673480430469548, + "grad_norm": 2.2095579554494584, + "learning_rate": 1.0526154729542566e-06, + "loss": 0.7886, + "step": 5550 + }, + { + "epoch": 0.6674682859375939, + "grad_norm": 2.324133812129708, + "learning_rate": 1.0519295119526699e-06, + "loss": 0.8083, + "step": 5551 + }, + { + "epoch": 0.667588528828233, + "grad_norm": 1.7801127233066514, + "learning_rate": 1.0512436947726227e-06, + "loss": 0.8328, + "step": 5552 + }, + { + "epoch": 0.6677087717188721, + "grad_norm": 12.136254190296295, + "learning_rate": 1.0505580215181517e-06, + "loss": 0.662, + "step": 5553 + }, + { + "epoch": 0.6678290146095112, + "grad_norm": 0.8539616447273471, + "learning_rate": 1.0498724922932753e-06, + "loss": 0.5656, + "step": 5554 + }, + { + "epoch": 0.6679492575001503, + "grad_norm": 2.1730091235008357, + "learning_rate": 1.0491871072019851e-06, + "loss": 0.8695, + "step": 5555 + }, + { + "epoch": 0.6680695003907894, + "grad_norm": 1.8410958405163689, + "learning_rate": 1.0485018663482555e-06, + "loss": 0.6425, + "step": 5556 + }, + { + "epoch": 0.6681897432814284, + "grad_norm": 2.4730024773338117, + "learning_rate": 1.0478167698360354e-06, + "loss": 0.7147, + "step": 5557 + }, + { + "epoch": 0.6683099861720676, + "grad_norm": 2.0241925219092116, + "learning_rate": 1.0471318177692556e-06, + "loss": 0.7101, + "step": 5558 + }, + { + "epoch": 0.6684302290627067, + "grad_norm": 2.322048064773511, + "learning_rate": 1.046447010251821e-06, + "loss": 0.763, + "step": 5559 + }, + { + "epoch": 0.6685504719533457, + "grad_norm": 1.8767109376260243, + "learning_rate": 1.0457623473876157e-06, + "loss": 0.7639, + "step": 5560 + }, + { + "epoch": 0.6686707148439849, + "grad_norm": 1.9438386412399282, + "learning_rate": 1.0450778292805046e-06, + "loss": 0.7117, + "step": 5561 + }, + { + "epoch": 0.6687909577346239, + "grad_norm": 1.5485290777415857, + "learning_rate": 1.0443934560343267e-06, + "loss": 0.7901, + "step": 5562 + }, + { + "epoch": 0.668911200625263, + "grad_norm": 1.9872232256415112, + "learning_rate": 1.0437092277529034e-06, + "loss": 0.7855, + "step": 5563 + }, + { + "epoch": 0.6690314435159022, + "grad_norm": 2.5696324895374865, + "learning_rate": 1.0430251445400292e-06, + "loss": 0.7442, + "step": 5564 + }, + { + "epoch": 0.6691516864065412, + "grad_norm": 2.0263668345255716, + "learning_rate": 1.0423412064994787e-06, + "loss": 0.6292, + "step": 5565 + }, + { + "epoch": 0.6692719292971803, + "grad_norm": 1.958302098559467, + "learning_rate": 1.0416574137350064e-06, + "loss": 0.7447, + "step": 5566 + }, + { + "epoch": 0.6693921721878194, + "grad_norm": 2.3157319861445176, + "learning_rate": 1.0409737663503428e-06, + "loss": 0.8119, + "step": 5567 + }, + { + "epoch": 0.6695124150784585, + "grad_norm": 5.907264904844361, + "learning_rate": 1.040290264449196e-06, + "loss": 0.8318, + "step": 5568 + }, + { + "epoch": 0.6696326579690975, + "grad_norm": 2.3398387571354804, + "learning_rate": 1.0396069081352532e-06, + "loss": 0.6407, + "step": 5569 + }, + { + "epoch": 0.6697529008597367, + "grad_norm": 0.7886139378421052, + "learning_rate": 1.0389236975121782e-06, + "loss": 0.5614, + "step": 5570 + }, + { + "epoch": 0.6698731437503758, + "grad_norm": 2.1183596811324232, + "learning_rate": 1.0382406326836147e-06, + "loss": 0.722, + "step": 5571 + }, + { + "epoch": 0.6699933866410148, + "grad_norm": 1.8715726044554535, + "learning_rate": 1.0375577137531828e-06, + "loss": 0.7636, + "step": 5572 + }, + { + "epoch": 0.670113629531654, + "grad_norm": 1.8735201399824994, + "learning_rate": 1.0368749408244802e-06, + "loss": 0.7262, + "step": 5573 + }, + { + "epoch": 0.670233872422293, + "grad_norm": 1.9366025384319214, + "learning_rate": 1.0361923140010836e-06, + "loss": 0.7879, + "step": 5574 + }, + { + "epoch": 0.6703541153129321, + "grad_norm": 2.25495767118053, + "learning_rate": 1.0355098333865455e-06, + "loss": 0.6364, + "step": 5575 + }, + { + "epoch": 0.6704743582035713, + "grad_norm": 1.9343382355827667, + "learning_rate": 1.0348274990844006e-06, + "loss": 0.6973, + "step": 5576 + }, + { + "epoch": 0.6705946010942103, + "grad_norm": 1.854367027458433, + "learning_rate": 1.034145311198155e-06, + "loss": 0.7243, + "step": 5577 + }, + { + "epoch": 0.6707148439848494, + "grad_norm": 1.8739090606787971, + "learning_rate": 1.0334632698312989e-06, + "loss": 0.6436, + "step": 5578 + }, + { + "epoch": 0.6708350868754885, + "grad_norm": 2.067985061333303, + "learning_rate": 1.032781375087295e-06, + "loss": 0.7537, + "step": 5579 + }, + { + "epoch": 0.6709553297661276, + "grad_norm": 1.5853429722879258, + "learning_rate": 1.0320996270695891e-06, + "loss": 0.6766, + "step": 5580 + }, + { + "epoch": 0.6710755726567667, + "grad_norm": 1.7686958277733515, + "learning_rate": 1.0314180258815998e-06, + "loss": 0.7347, + "step": 5581 + }, + { + "epoch": 0.6711958155474057, + "grad_norm": 1.9427476389763465, + "learning_rate": 1.0307365716267247e-06, + "loss": 0.7422, + "step": 5582 + }, + { + "epoch": 0.6713160584380449, + "grad_norm": 2.1801813037968967, + "learning_rate": 1.0300552644083423e-06, + "loss": 0.7881, + "step": 5583 + }, + { + "epoch": 0.6714363013286839, + "grad_norm": 2.4616225485011025, + "learning_rate": 1.0293741043298036e-06, + "loss": 0.7297, + "step": 5584 + }, + { + "epoch": 0.671556544219323, + "grad_norm": 2.713287829768152, + "learning_rate": 1.0286930914944436e-06, + "loss": 0.7183, + "step": 5585 + }, + { + "epoch": 0.6716767871099621, + "grad_norm": 2.4829311373714065, + "learning_rate": 1.0280122260055684e-06, + "loss": 0.7788, + "step": 5586 + }, + { + "epoch": 0.6717970300006012, + "grad_norm": 2.202305201598718, + "learning_rate": 1.0273315079664652e-06, + "loss": 0.824, + "step": 5587 + }, + { + "epoch": 0.6719172728912403, + "grad_norm": 2.2131485414155168, + "learning_rate": 1.0266509374803992e-06, + "loss": 0.7468, + "step": 5588 + }, + { + "epoch": 0.6720375157818794, + "grad_norm": 2.6433062061426735, + "learning_rate": 1.0259705146506123e-06, + "loss": 0.8467, + "step": 5589 + }, + { + "epoch": 0.6721577586725185, + "grad_norm": 1.9981228841007355, + "learning_rate": 1.025290239580324e-06, + "loss": 0.781, + "step": 5590 + }, + { + "epoch": 0.6722780015631575, + "grad_norm": 2.352952008012075, + "learning_rate": 1.0246101123727313e-06, + "loss": 0.758, + "step": 5591 + }, + { + "epoch": 0.6723982444537967, + "grad_norm": 1.9382961852279728, + "learning_rate": 1.0239301331310085e-06, + "loss": 0.788, + "step": 5592 + }, + { + "epoch": 0.6725184873444358, + "grad_norm": 1.6470912841996124, + "learning_rate": 1.0232503019583088e-06, + "loss": 0.8897, + "step": 5593 + }, + { + "epoch": 0.6726387302350748, + "grad_norm": 2.197780864069698, + "learning_rate": 1.0225706189577619e-06, + "loss": 0.7019, + "step": 5594 + }, + { + "epoch": 0.672758973125714, + "grad_norm": 1.9233364348273612, + "learning_rate": 1.021891084232475e-06, + "loss": 0.7506, + "step": 5595 + }, + { + "epoch": 0.672879216016353, + "grad_norm": 2.362036713078237, + "learning_rate": 1.0212116978855325e-06, + "loss": 0.8083, + "step": 5596 + }, + { + "epoch": 0.6729994589069921, + "grad_norm": 2.4677934362750693, + "learning_rate": 1.020532460019997e-06, + "loss": 0.7896, + "step": 5597 + }, + { + "epoch": 0.6731197017976313, + "grad_norm": 2.411043223388753, + "learning_rate": 1.0198533707389096e-06, + "loss": 0.7132, + "step": 5598 + }, + { + "epoch": 0.6732399446882703, + "grad_norm": 1.956593965266113, + "learning_rate": 1.0191744301452853e-06, + "loss": 0.7377, + "step": 5599 + }, + { + "epoch": 0.6733601875789094, + "grad_norm": 1.7659754447525924, + "learning_rate": 1.0184956383421208e-06, + "loss": 0.7029, + "step": 5600 + }, + { + "epoch": 0.6734804304695485, + "grad_norm": 2.21382109430402, + "learning_rate": 1.017816995432387e-06, + "loss": 0.6578, + "step": 5601 + }, + { + "epoch": 0.6736006733601876, + "grad_norm": 2.0135489721374547, + "learning_rate": 1.0171385015190353e-06, + "loss": 0.7468, + "step": 5602 + }, + { + "epoch": 0.6737209162508266, + "grad_norm": 2.0792611823708342, + "learning_rate": 1.0164601567049908e-06, + "loss": 0.7383, + "step": 5603 + }, + { + "epoch": 0.6738411591414658, + "grad_norm": 1.7181235819145004, + "learning_rate": 1.015781961093158e-06, + "loss": 0.8077, + "step": 5604 + }, + { + "epoch": 0.6739614020321049, + "grad_norm": 1.677331025528849, + "learning_rate": 1.0151039147864197e-06, + "loss": 0.7726, + "step": 5605 + }, + { + "epoch": 0.6740816449227439, + "grad_norm": 2.0916327404092607, + "learning_rate": 1.0144260178876336e-06, + "loss": 0.6633, + "step": 5606 + }, + { + "epoch": 0.6742018878133831, + "grad_norm": 2.4808786609196143, + "learning_rate": 1.0137482704996388e-06, + "loss": 0.6718, + "step": 5607 + }, + { + "epoch": 0.6743221307040221, + "grad_norm": 2.008636661695812, + "learning_rate": 1.0130706727252461e-06, + "loss": 0.7949, + "step": 5608 + }, + { + "epoch": 0.6744423735946612, + "grad_norm": 2.5665086235170538, + "learning_rate": 1.0123932246672468e-06, + "loss": 0.6849, + "step": 5609 + }, + { + "epoch": 0.6745626164853004, + "grad_norm": 0.7607779929043754, + "learning_rate": 1.0117159264284114e-06, + "loss": 0.5586, + "step": 5610 + }, + { + "epoch": 0.6746828593759394, + "grad_norm": 1.638641830005217, + "learning_rate": 1.0110387781114837e-06, + "loss": 0.7732, + "step": 5611 + }, + { + "epoch": 0.6748031022665785, + "grad_norm": 2.373979047880455, + "learning_rate": 1.0103617798191872e-06, + "loss": 0.7762, + "step": 5612 + }, + { + "epoch": 0.6749233451572175, + "grad_norm": 2.504954292809296, + "learning_rate": 1.0096849316542217e-06, + "loss": 0.835, + "step": 5613 + }, + { + "epoch": 0.6750435880478567, + "grad_norm": 1.954978002829349, + "learning_rate": 1.0090082337192643e-06, + "loss": 0.7558, + "step": 5614 + }, + { + "epoch": 0.6751638309384957, + "grad_norm": 2.2319921845945303, + "learning_rate": 1.0083316861169705e-06, + "loss": 0.789, + "step": 5615 + }, + { + "epoch": 0.6752840738291348, + "grad_norm": 1.9797852670423686, + "learning_rate": 1.0076552889499713e-06, + "loss": 0.7219, + "step": 5616 + }, + { + "epoch": 0.675404316719774, + "grad_norm": 1.9419218487792853, + "learning_rate": 1.006979042320876e-06, + "loss": 0.7392, + "step": 5617 + }, + { + "epoch": 0.675524559610413, + "grad_norm": 4.878302018291307, + "learning_rate": 1.0063029463322702e-06, + "loss": 0.6337, + "step": 5618 + }, + { + "epoch": 0.6756448025010521, + "grad_norm": 2.156431992079139, + "learning_rate": 1.0056270010867164e-06, + "loss": 0.7603, + "step": 5619 + }, + { + "epoch": 0.6757650453916912, + "grad_norm": 3.657010382462561, + "learning_rate": 1.004951206686758e-06, + "loss": 0.7958, + "step": 5620 + }, + { + "epoch": 0.6758852882823303, + "grad_norm": 2.0680808653377585, + "learning_rate": 1.0042755632349087e-06, + "loss": 0.719, + "step": 5621 + }, + { + "epoch": 0.6760055311729694, + "grad_norm": 2.034122958882532, + "learning_rate": 1.0036000708336653e-06, + "loss": 0.6325, + "step": 5622 + }, + { + "epoch": 0.6761257740636085, + "grad_norm": 2.67013778873901, + "learning_rate": 1.0029247295854984e-06, + "loss": 0.8044, + "step": 5623 + }, + { + "epoch": 0.6762460169542476, + "grad_norm": 2.630528486551855, + "learning_rate": 1.0022495395928588e-06, + "loss": 0.7203, + "step": 5624 + }, + { + "epoch": 0.6763662598448866, + "grad_norm": 0.8019146597398861, + "learning_rate": 1.0015745009581697e-06, + "loss": 0.6245, + "step": 5625 + }, + { + "epoch": 0.6764865027355258, + "grad_norm": 1.8986395583137414, + "learning_rate": 1.0008996137838343e-06, + "loss": 0.6731, + "step": 5626 + }, + { + "epoch": 0.6766067456261649, + "grad_norm": 2.2060366389603105, + "learning_rate": 1.000224878172234e-06, + "loss": 0.8071, + "step": 5627 + }, + { + "epoch": 0.6767269885168039, + "grad_norm": 2.08774479784093, + "learning_rate": 9.99550294225724e-07, + "loss": 0.7354, + "step": 5628 + }, + { + "epoch": 0.6768472314074431, + "grad_norm": 2.0947763005822844, + "learning_rate": 9.988758620466402e-07, + "loss": 0.7259, + "step": 5629 + }, + { + "epoch": 0.6769674742980821, + "grad_norm": 2.0237994367417564, + "learning_rate": 9.982015817372917e-07, + "loss": 0.7633, + "step": 5630 + }, + { + "epoch": 0.6770877171887212, + "grad_norm": 1.9054317281519675, + "learning_rate": 9.975274533999657e-07, + "loss": 0.8283, + "step": 5631 + }, + { + "epoch": 0.6772079600793603, + "grad_norm": 3.0840786900753923, + "learning_rate": 9.96853477136929e-07, + "loss": 0.8466, + "step": 5632 + }, + { + "epoch": 0.6773282029699994, + "grad_norm": 2.0744277072978887, + "learning_rate": 9.96179653050422e-07, + "loss": 0.751, + "step": 5633 + }, + { + "epoch": 0.6774484458606385, + "grad_norm": 2.1479552681454654, + "learning_rate": 9.955059812426635e-07, + "loss": 0.7466, + "step": 5634 + }, + { + "epoch": 0.6775686887512776, + "grad_norm": 2.1993696523494592, + "learning_rate": 9.948324618158493e-07, + "loss": 0.8311, + "step": 5635 + }, + { + "epoch": 0.6776889316419167, + "grad_norm": 2.3767928333509545, + "learning_rate": 9.941590948721502e-07, + "loss": 0.7845, + "step": 5636 + }, + { + "epoch": 0.6778091745325557, + "grad_norm": 1.7418876355674113, + "learning_rate": 9.934858805137188e-07, + "loss": 0.7664, + "step": 5637 + }, + { + "epoch": 0.6779294174231949, + "grad_norm": 1.9115878554142587, + "learning_rate": 9.92812818842677e-07, + "loss": 0.8129, + "step": 5638 + }, + { + "epoch": 0.678049660313834, + "grad_norm": 2.131194342728286, + "learning_rate": 9.921399099611306e-07, + "loss": 0.6442, + "step": 5639 + }, + { + "epoch": 0.678169903204473, + "grad_norm": 1.6960904494314528, + "learning_rate": 9.914671539711588e-07, + "loss": 0.6943, + "step": 5640 + }, + { + "epoch": 0.6782901460951122, + "grad_norm": 25.51917338883097, + "learning_rate": 9.90794550974817e-07, + "loss": 0.7851, + "step": 5641 + }, + { + "epoch": 0.6784103889857512, + "grad_norm": 2.124119071122922, + "learning_rate": 9.901221010741407e-07, + "loss": 0.8172, + "step": 5642 + }, + { + "epoch": 0.6785306318763903, + "grad_norm": 2.4972311899742925, + "learning_rate": 9.894498043711375e-07, + "loss": 0.7494, + "step": 5643 + }, + { + "epoch": 0.6786508747670293, + "grad_norm": 41.25209955594306, + "learning_rate": 9.887776609677962e-07, + "loss": 0.6986, + "step": 5644 + }, + { + "epoch": 0.6787711176576685, + "grad_norm": 2.6315387848248784, + "learning_rate": 9.88105670966079e-07, + "loss": 0.7259, + "step": 5645 + }, + { + "epoch": 0.6788913605483076, + "grad_norm": 2.233825829515114, + "learning_rate": 9.874338344679283e-07, + "loss": 0.7938, + "step": 5646 + }, + { + "epoch": 0.6790116034389466, + "grad_norm": 2.1154903352102368, + "learning_rate": 9.86762151575259e-07, + "loss": 0.7435, + "step": 5647 + }, + { + "epoch": 0.6791318463295858, + "grad_norm": 1.487494553628135, + "learning_rate": 9.860906223899651e-07, + "loss": 0.8063, + "step": 5648 + }, + { + "epoch": 0.6792520892202248, + "grad_norm": 3.1275799591796227, + "learning_rate": 9.854192470139184e-07, + "loss": 0.762, + "step": 5649 + }, + { + "epoch": 0.6793723321108639, + "grad_norm": 2.061852902712882, + "learning_rate": 9.847480255489645e-07, + "loss": 0.7227, + "step": 5650 + }, + { + "epoch": 0.6794925750015031, + "grad_norm": 2.197830873475227, + "learning_rate": 9.840769580969295e-07, + "loss": 0.696, + "step": 5651 + }, + { + "epoch": 0.6796128178921421, + "grad_norm": 1.9470094686394517, + "learning_rate": 9.834060447596114e-07, + "loss": 0.8031, + "step": 5652 + }, + { + "epoch": 0.6797330607827812, + "grad_norm": 2.0275508355894947, + "learning_rate": 9.827352856387868e-07, + "loss": 0.7905, + "step": 5653 + }, + { + "epoch": 0.6798533036734203, + "grad_norm": 0.7935936631159276, + "learning_rate": 9.820646808362118e-07, + "loss": 0.6433, + "step": 5654 + }, + { + "epoch": 0.6799735465640594, + "grad_norm": 2.3103809442215097, + "learning_rate": 9.813942304536154e-07, + "loss": 0.7305, + "step": 5655 + }, + { + "epoch": 0.6800937894546984, + "grad_norm": 1.8937726891294666, + "learning_rate": 9.807239345927043e-07, + "loss": 0.6418, + "step": 5656 + }, + { + "epoch": 0.6802140323453376, + "grad_norm": 2.128457006248308, + "learning_rate": 9.80053793355162e-07, + "loss": 0.7282, + "step": 5657 + }, + { + "epoch": 0.6803342752359767, + "grad_norm": 2.1272738016195483, + "learning_rate": 9.793838068426472e-07, + "loss": 0.7542, + "step": 5658 + }, + { + "epoch": 0.6804545181266157, + "grad_norm": 2.6512258249607794, + "learning_rate": 9.78713975156799e-07, + "loss": 0.6174, + "step": 5659 + }, + { + "epoch": 0.6805747610172549, + "grad_norm": 1.9099236407170186, + "learning_rate": 9.780442983992273e-07, + "loss": 0.7257, + "step": 5660 + }, + { + "epoch": 0.680695003907894, + "grad_norm": 1.7861490973663035, + "learning_rate": 9.773747766715238e-07, + "loss": 0.7244, + "step": 5661 + }, + { + "epoch": 0.680815246798533, + "grad_norm": 1.7073859712608623, + "learning_rate": 9.767054100752536e-07, + "loss": 0.8065, + "step": 5662 + }, + { + "epoch": 0.6809354896891722, + "grad_norm": 1.9456694248714288, + "learning_rate": 9.760361987119584e-07, + "loss": 0.8232, + "step": 5663 + }, + { + "epoch": 0.6810557325798112, + "grad_norm": 2.205251762912704, + "learning_rate": 9.753671426831592e-07, + "loss": 0.6803, + "step": 5664 + }, + { + "epoch": 0.6811759754704503, + "grad_norm": 1.9201573705751431, + "learning_rate": 9.746982420903483e-07, + "loss": 0.7989, + "step": 5665 + }, + { + "epoch": 0.6812962183610894, + "grad_norm": 1.574465294550857, + "learning_rate": 9.740294970349993e-07, + "loss": 0.7498, + "step": 5666 + }, + { + "epoch": 0.6814164612517285, + "grad_norm": 0.9001671106113394, + "learning_rate": 9.733609076185594e-07, + "loss": 0.6105, + "step": 5667 + }, + { + "epoch": 0.6815367041423676, + "grad_norm": 1.9836730917124261, + "learning_rate": 9.72692473942455e-07, + "loss": 0.8433, + "step": 5668 + }, + { + "epoch": 0.6816569470330067, + "grad_norm": 1.70129408107269, + "learning_rate": 9.720241961080849e-07, + "loss": 0.7783, + "step": 5669 + }, + { + "epoch": 0.6817771899236458, + "grad_norm": 2.212425351365277, + "learning_rate": 9.713560742168259e-07, + "loss": 0.7304, + "step": 5670 + }, + { + "epoch": 0.6818974328142848, + "grad_norm": 2.134320407241856, + "learning_rate": 9.706881083700333e-07, + "loss": 0.7182, + "step": 5671 + }, + { + "epoch": 0.682017675704924, + "grad_norm": 2.070754669762633, + "learning_rate": 9.700202986690357e-07, + "loss": 0.8325, + "step": 5672 + }, + { + "epoch": 0.682137918595563, + "grad_norm": 3.4198384952892593, + "learning_rate": 9.693526452151413e-07, + "loss": 0.6662, + "step": 5673 + }, + { + "epoch": 0.6822581614862021, + "grad_norm": 1.6950138373921335, + "learning_rate": 9.686851481096305e-07, + "loss": 0.7588, + "step": 5674 + }, + { + "epoch": 0.6823784043768413, + "grad_norm": 1.9445461004417366, + "learning_rate": 9.68017807453762e-07, + "loss": 0.7291, + "step": 5675 + }, + { + "epoch": 0.6824986472674803, + "grad_norm": 2.0134014314330564, + "learning_rate": 9.673506233487721e-07, + "loss": 0.7415, + "step": 5676 + }, + { + "epoch": 0.6826188901581194, + "grad_norm": 1.7513004648042785, + "learning_rate": 9.666835958958717e-07, + "loss": 0.8619, + "step": 5677 + }, + { + "epoch": 0.6827391330487584, + "grad_norm": 2.943774077039607, + "learning_rate": 9.660167251962484e-07, + "loss": 0.8088, + "step": 5678 + }, + { + "epoch": 0.6828593759393976, + "grad_norm": 1.7462208678123594, + "learning_rate": 9.653500113510654e-07, + "loss": 0.7855, + "step": 5679 + }, + { + "epoch": 0.6829796188300367, + "grad_norm": 5.496895694314194, + "learning_rate": 9.646834544614627e-07, + "loss": 0.674, + "step": 5680 + }, + { + "epoch": 0.6830998617206757, + "grad_norm": 1.8983426168458706, + "learning_rate": 9.64017054628558e-07, + "loss": 0.7655, + "step": 5681 + }, + { + "epoch": 0.6832201046113149, + "grad_norm": 1.7369085828479571, + "learning_rate": 9.63350811953441e-07, + "loss": 0.7939, + "step": 5682 + }, + { + "epoch": 0.6833403475019539, + "grad_norm": 2.853445941301359, + "learning_rate": 9.626847265371826e-07, + "loss": 0.7077, + "step": 5683 + }, + { + "epoch": 0.683460590392593, + "grad_norm": 2.0460826708889504, + "learning_rate": 9.620187984808262e-07, + "loss": 0.7915, + "step": 5684 + }, + { + "epoch": 0.6835808332832322, + "grad_norm": 2.0008943465448534, + "learning_rate": 9.613530278853919e-07, + "loss": 0.8624, + "step": 5685 + }, + { + "epoch": 0.6837010761738712, + "grad_norm": 2.003277120304039, + "learning_rate": 9.60687414851879e-07, + "loss": 0.7492, + "step": 5686 + }, + { + "epoch": 0.6838213190645103, + "grad_norm": 3.260466375164992, + "learning_rate": 9.600219594812575e-07, + "loss": 0.7776, + "step": 5687 + }, + { + "epoch": 0.6839415619551494, + "grad_norm": 1.8043944546133353, + "learning_rate": 9.593566618744786e-07, + "loss": 0.7339, + "step": 5688 + }, + { + "epoch": 0.6840618048457885, + "grad_norm": 2.491485209752416, + "learning_rate": 9.58691522132466e-07, + "loss": 0.7387, + "step": 5689 + }, + { + "epoch": 0.6841820477364275, + "grad_norm": 2.061490129556834, + "learning_rate": 9.58026540356123e-07, + "loss": 0.8538, + "step": 5690 + }, + { + "epoch": 0.6843022906270667, + "grad_norm": 1.6773020994561372, + "learning_rate": 9.573617166463246e-07, + "loss": 0.8707, + "step": 5691 + }, + { + "epoch": 0.6844225335177058, + "grad_norm": 2.5776490668065906, + "learning_rate": 9.56697051103924e-07, + "loss": 0.6028, + "step": 5692 + }, + { + "epoch": 0.6845427764083448, + "grad_norm": 1.9339326811931596, + "learning_rate": 9.560325438297522e-07, + "loss": 0.8135, + "step": 5693 + }, + { + "epoch": 0.684663019298984, + "grad_norm": 2.0658240841310587, + "learning_rate": 9.553681949246127e-07, + "loss": 0.8719, + "step": 5694 + }, + { + "epoch": 0.684783262189623, + "grad_norm": 1.834573181988334, + "learning_rate": 9.547040044892886e-07, + "loss": 0.7589, + "step": 5695 + }, + { + "epoch": 0.6849035050802621, + "grad_norm": 0.8805318422229883, + "learning_rate": 9.540399726245354e-07, + "loss": 0.6012, + "step": 5696 + }, + { + "epoch": 0.6850237479709013, + "grad_norm": 1.9523005462054845, + "learning_rate": 9.533760994310859e-07, + "loss": 0.6917, + "step": 5697 + }, + { + "epoch": 0.6851439908615403, + "grad_norm": 2.3311332366583475, + "learning_rate": 9.527123850096508e-07, + "loss": 0.7567, + "step": 5698 + }, + { + "epoch": 0.6852642337521794, + "grad_norm": 2.1166132186789754, + "learning_rate": 9.520488294609142e-07, + "loss": 0.7248, + "step": 5699 + }, + { + "epoch": 0.6853844766428185, + "grad_norm": 0.7558687588998815, + "learning_rate": 9.513854328855368e-07, + "loss": 0.5386, + "step": 5700 + }, + { + "epoch": 0.6855047195334576, + "grad_norm": 1.99445277058532, + "learning_rate": 9.507221953841558e-07, + "loss": 0.816, + "step": 5701 + }, + { + "epoch": 0.6856249624240967, + "grad_norm": 1.6581258407573387, + "learning_rate": 9.500591170573824e-07, + "loss": 0.7835, + "step": 5702 + }, + { + "epoch": 0.6857452053147358, + "grad_norm": 2.172751051046064, + "learning_rate": 9.493961980058078e-07, + "loss": 0.7524, + "step": 5703 + }, + { + "epoch": 0.6858654482053749, + "grad_norm": 1.9125742829301637, + "learning_rate": 9.48733438329993e-07, + "loss": 0.6822, + "step": 5704 + }, + { + "epoch": 0.6859856910960139, + "grad_norm": 1.6903587427794053, + "learning_rate": 9.480708381304807e-07, + "loss": 0.7454, + "step": 5705 + }, + { + "epoch": 0.6861059339866531, + "grad_norm": 2.2486349839573028, + "learning_rate": 9.474083975077858e-07, + "loss": 0.8424, + "step": 5706 + }, + { + "epoch": 0.6862261768772921, + "grad_norm": 3.3987231605875836, + "learning_rate": 9.467461165623994e-07, + "loss": 0.8039, + "step": 5707 + }, + { + "epoch": 0.6863464197679312, + "grad_norm": 1.8965767740670083, + "learning_rate": 9.46083995394791e-07, + "loss": 0.8029, + "step": 5708 + }, + { + "epoch": 0.6864666626585703, + "grad_norm": 1.8694936644127191, + "learning_rate": 9.454220341054012e-07, + "loss": 0.6341, + "step": 5709 + }, + { + "epoch": 0.6865869055492094, + "grad_norm": 2.035548898962724, + "learning_rate": 9.447602327946512e-07, + "loss": 0.8088, + "step": 5710 + }, + { + "epoch": 0.6867071484398485, + "grad_norm": 1.9688728536391062, + "learning_rate": 9.440985915629338e-07, + "loss": 0.7686, + "step": 5711 + }, + { + "epoch": 0.6868273913304875, + "grad_norm": 2.6559216464676236, + "learning_rate": 9.434371105106223e-07, + "loss": 0.7389, + "step": 5712 + }, + { + "epoch": 0.6869476342211267, + "grad_norm": 1.643453575646101, + "learning_rate": 9.427757897380602e-07, + "loss": 0.7104, + "step": 5713 + }, + { + "epoch": 0.6870678771117658, + "grad_norm": 2.0692570137070057, + "learning_rate": 9.421146293455695e-07, + "loss": 0.8476, + "step": 5714 + }, + { + "epoch": 0.6871881200024048, + "grad_norm": 1.8796768353065343, + "learning_rate": 9.414536294334489e-07, + "loss": 0.6901, + "step": 5715 + }, + { + "epoch": 0.687308362893044, + "grad_norm": 4.511323880177023, + "learning_rate": 9.407927901019708e-07, + "loss": 0.6958, + "step": 5716 + }, + { + "epoch": 0.687428605783683, + "grad_norm": 1.9985339632829935, + "learning_rate": 9.401321114513854e-07, + "loss": 0.769, + "step": 5717 + }, + { + "epoch": 0.6875488486743221, + "grad_norm": 1.6198903224457946, + "learning_rate": 9.394715935819155e-07, + "loss": 0.7541, + "step": 5718 + }, + { + "epoch": 0.6876690915649613, + "grad_norm": 1.9980487842121866, + "learning_rate": 9.388112365937608e-07, + "loss": 0.6246, + "step": 5719 + }, + { + "epoch": 0.6877893344556003, + "grad_norm": 2.148709921471402, + "learning_rate": 9.381510405870985e-07, + "loss": 0.8289, + "step": 5720 + }, + { + "epoch": 0.6879095773462394, + "grad_norm": 7.458952410984798, + "learning_rate": 9.374910056620791e-07, + "loss": 0.7808, + "step": 5721 + }, + { + "epoch": 0.6880298202368785, + "grad_norm": 1.7355555734839272, + "learning_rate": 9.368311319188293e-07, + "loss": 0.816, + "step": 5722 + }, + { + "epoch": 0.6881500631275176, + "grad_norm": 1.857327014574602, + "learning_rate": 9.361714194574515e-07, + "loss": 0.7986, + "step": 5723 + }, + { + "epoch": 0.6882703060181566, + "grad_norm": 0.7406849559224901, + "learning_rate": 9.355118683780228e-07, + "loss": 0.5833, + "step": 5724 + }, + { + "epoch": 0.6883905489087958, + "grad_norm": 2.225334819504455, + "learning_rate": 9.348524787805987e-07, + "loss": 0.7951, + "step": 5725 + }, + { + "epoch": 0.6885107917994349, + "grad_norm": 3.454050207686512, + "learning_rate": 9.341932507652053e-07, + "loss": 0.8556, + "step": 5726 + }, + { + "epoch": 0.6886310346900739, + "grad_norm": 1.9119776367691252, + "learning_rate": 9.335341844318489e-07, + "loss": 0.7881, + "step": 5727 + }, + { + "epoch": 0.6887512775807131, + "grad_norm": 1.9191335338272844, + "learning_rate": 9.328752798805091e-07, + "loss": 0.7354, + "step": 5728 + }, + { + "epoch": 0.6888715204713521, + "grad_norm": 2.6495972741205134, + "learning_rate": 9.322165372111399e-07, + "loss": 0.7656, + "step": 5729 + }, + { + "epoch": 0.6889917633619912, + "grad_norm": 1.971751392679846, + "learning_rate": 9.315579565236747e-07, + "loss": 0.7605, + "step": 5730 + }, + { + "epoch": 0.6891120062526304, + "grad_norm": 1.7837856381658184, + "learning_rate": 9.308995379180162e-07, + "loss": 0.7388, + "step": 5731 + }, + { + "epoch": 0.6892322491432694, + "grad_norm": 0.7488753708335635, + "learning_rate": 9.302412814940488e-07, + "loss": 0.5959, + "step": 5732 + }, + { + "epoch": 0.6893524920339085, + "grad_norm": 2.0008982786891933, + "learning_rate": 9.295831873516276e-07, + "loss": 0.7186, + "step": 5733 + }, + { + "epoch": 0.6894727349245476, + "grad_norm": 1.6715897646577624, + "learning_rate": 9.289252555905873e-07, + "loss": 0.7612, + "step": 5734 + }, + { + "epoch": 0.6895929778151867, + "grad_norm": 1.8901657933980243, + "learning_rate": 9.282674863107334e-07, + "loss": 0.7677, + "step": 5735 + }, + { + "epoch": 0.6897132207058257, + "grad_norm": 3.1358355541223073, + "learning_rate": 9.276098796118488e-07, + "loss": 0.7678, + "step": 5736 + }, + { + "epoch": 0.6898334635964649, + "grad_norm": 1.740055987016412, + "learning_rate": 9.269524355936938e-07, + "loss": 0.6635, + "step": 5737 + }, + { + "epoch": 0.689953706487104, + "grad_norm": 1.855939817361305, + "learning_rate": 9.262951543560002e-07, + "loss": 0.8553, + "step": 5738 + }, + { + "epoch": 0.690073949377743, + "grad_norm": 2.562445384118096, + "learning_rate": 9.256380359984795e-07, + "loss": 0.8688, + "step": 5739 + }, + { + "epoch": 0.6901941922683821, + "grad_norm": 1.8442946453939866, + "learning_rate": 9.249810806208139e-07, + "loss": 0.7511, + "step": 5740 + }, + { + "epoch": 0.6903144351590212, + "grad_norm": 1.9748000780153094, + "learning_rate": 9.243242883226627e-07, + "loss": 0.8055, + "step": 5741 + }, + { + "epoch": 0.6904346780496603, + "grad_norm": 2.4769554428140275, + "learning_rate": 9.236676592036628e-07, + "loss": 0.701, + "step": 5742 + }, + { + "epoch": 0.6905549209402994, + "grad_norm": 1.7751411623633002, + "learning_rate": 9.230111933634228e-07, + "loss": 0.7389, + "step": 5743 + }, + { + "epoch": 0.6906751638309385, + "grad_norm": 1.6029906813361965, + "learning_rate": 9.223548909015288e-07, + "loss": 0.8112, + "step": 5744 + }, + { + "epoch": 0.6907954067215776, + "grad_norm": 1.958766145426626, + "learning_rate": 9.216987519175407e-07, + "loss": 0.7251, + "step": 5745 + }, + { + "epoch": 0.6909156496122166, + "grad_norm": 1.6646955118904927, + "learning_rate": 9.210427765109942e-07, + "loss": 0.6891, + "step": 5746 + }, + { + "epoch": 0.6910358925028558, + "grad_norm": 2.2853073200598515, + "learning_rate": 9.20386964781402e-07, + "loss": 0.8161, + "step": 5747 + }, + { + "epoch": 0.6911561353934949, + "grad_norm": 2.1221444634421713, + "learning_rate": 9.197313168282472e-07, + "loss": 0.8488, + "step": 5748 + }, + { + "epoch": 0.6912763782841339, + "grad_norm": 2.2485497357188846, + "learning_rate": 9.190758327509935e-07, + "loss": 0.7258, + "step": 5749 + }, + { + "epoch": 0.6913966211747731, + "grad_norm": 0.9607071949448451, + "learning_rate": 9.184205126490767e-07, + "loss": 0.645, + "step": 5750 + }, + { + "epoch": 0.6915168640654121, + "grad_norm": 1.117552424231663, + "learning_rate": 9.177653566219075e-07, + "loss": 0.5971, + "step": 5751 + }, + { + "epoch": 0.6916371069560512, + "grad_norm": 2.2682341186142367, + "learning_rate": 9.171103647688744e-07, + "loss": 0.765, + "step": 5752 + }, + { + "epoch": 0.6917573498466904, + "grad_norm": 1.7649174175448477, + "learning_rate": 9.164555371893367e-07, + "loss": 0.695, + "step": 5753 + }, + { + "epoch": 0.6918775927373294, + "grad_norm": 1.8909955883166303, + "learning_rate": 9.158008739826333e-07, + "loss": 0.7577, + "step": 5754 + }, + { + "epoch": 0.6919978356279685, + "grad_norm": 1.7358734943116634, + "learning_rate": 9.151463752480744e-07, + "loss": 0.8685, + "step": 5755 + }, + { + "epoch": 0.6921180785186076, + "grad_norm": 1.3918485937538503, + "learning_rate": 9.144920410849493e-07, + "loss": 0.8064, + "step": 5756 + }, + { + "epoch": 0.6922383214092467, + "grad_norm": 1.7769917233799761, + "learning_rate": 9.138378715925176e-07, + "loss": 0.8096, + "step": 5757 + }, + { + "epoch": 0.6923585642998857, + "grad_norm": 1.902653850996559, + "learning_rate": 9.131838668700167e-07, + "loss": 0.8135, + "step": 5758 + }, + { + "epoch": 0.6924788071905249, + "grad_norm": 1.811032194174748, + "learning_rate": 9.125300270166598e-07, + "loss": 0.8696, + "step": 5759 + }, + { + "epoch": 0.692599050081164, + "grad_norm": 1.8035911764010704, + "learning_rate": 9.118763521316324e-07, + "loss": 0.8625, + "step": 5760 + }, + { + "epoch": 0.692719292971803, + "grad_norm": 1.8235665243831134, + "learning_rate": 9.112228423140987e-07, + "loss": 0.7641, + "step": 5761 + }, + { + "epoch": 0.6928395358624422, + "grad_norm": 2.5137715590361998, + "learning_rate": 9.105694976631932e-07, + "loss": 0.8692, + "step": 5762 + }, + { + "epoch": 0.6929597787530812, + "grad_norm": 2.1661939349791446, + "learning_rate": 9.099163182780283e-07, + "loss": 0.7308, + "step": 5763 + }, + { + "epoch": 0.6930800216437203, + "grad_norm": 3.4825786527472427, + "learning_rate": 9.092633042576916e-07, + "loss": 0.5002, + "step": 5764 + }, + { + "epoch": 0.6932002645343595, + "grad_norm": 2.062124622748905, + "learning_rate": 9.086104557012446e-07, + "loss": 0.5684, + "step": 5765 + }, + { + "epoch": 0.6933205074249985, + "grad_norm": 2.0369455156413503, + "learning_rate": 9.079577727077239e-07, + "loss": 0.6578, + "step": 5766 + }, + { + "epoch": 0.6934407503156376, + "grad_norm": 2.272342588640394, + "learning_rate": 9.073052553761404e-07, + "loss": 0.7253, + "step": 5767 + }, + { + "epoch": 0.6935609932062767, + "grad_norm": 1.6589836018207744, + "learning_rate": 9.066529038054805e-07, + "loss": 0.7807, + "step": 5768 + }, + { + "epoch": 0.6936812360969158, + "grad_norm": 1.7407721684972868, + "learning_rate": 9.060007180947071e-07, + "loss": 0.748, + "step": 5769 + }, + { + "epoch": 0.6938014789875548, + "grad_norm": 1.9486634222132615, + "learning_rate": 9.053486983427534e-07, + "loss": 0.7354, + "step": 5770 + }, + { + "epoch": 0.6939217218781939, + "grad_norm": 2.3662966824061304, + "learning_rate": 9.046968446485326e-07, + "loss": 0.7141, + "step": 5771 + }, + { + "epoch": 0.6940419647688331, + "grad_norm": 3.456294335741434, + "learning_rate": 9.040451571109295e-07, + "loss": 0.7082, + "step": 5772 + }, + { + "epoch": 0.6941622076594721, + "grad_norm": 0.8485719156309732, + "learning_rate": 9.033936358288042e-07, + "loss": 0.6051, + "step": 5773 + }, + { + "epoch": 0.6942824505501112, + "grad_norm": 1.604072056081738, + "learning_rate": 9.027422809009937e-07, + "loss": 0.8249, + "step": 5774 + }, + { + "epoch": 0.6944026934407503, + "grad_norm": 1.8154612061882007, + "learning_rate": 9.020910924263054e-07, + "loss": 0.837, + "step": 5775 + }, + { + "epoch": 0.6945229363313894, + "grad_norm": 0.817921705815058, + "learning_rate": 9.014400705035261e-07, + "loss": 0.5823, + "step": 5776 + }, + { + "epoch": 0.6946431792220285, + "grad_norm": 2.057289593530568, + "learning_rate": 9.00789215231414e-07, + "loss": 0.7732, + "step": 5777 + }, + { + "epoch": 0.6947634221126676, + "grad_norm": 1.8200367076022541, + "learning_rate": 9.001385267087056e-07, + "loss": 0.8243, + "step": 5778 + }, + { + "epoch": 0.6948836650033067, + "grad_norm": 1.5406233719822946, + "learning_rate": 8.994880050341072e-07, + "loss": 0.7081, + "step": 5779 + }, + { + "epoch": 0.6950039078939457, + "grad_norm": 1.8071032452455644, + "learning_rate": 8.988376503063026e-07, + "loss": 0.7783, + "step": 5780 + }, + { + "epoch": 0.6951241507845849, + "grad_norm": 1.911573387515356, + "learning_rate": 8.981874626239521e-07, + "loss": 0.8194, + "step": 5781 + }, + { + "epoch": 0.695244393675224, + "grad_norm": 2.0970262543977127, + "learning_rate": 8.975374420856872e-07, + "loss": 0.8897, + "step": 5782 + }, + { + "epoch": 0.695364636565863, + "grad_norm": 2.144882423539651, + "learning_rate": 8.968875887901157e-07, + "loss": 0.7259, + "step": 5783 + }, + { + "epoch": 0.6954848794565022, + "grad_norm": 3.290441518451975, + "learning_rate": 8.9623790283582e-07, + "loss": 0.6364, + "step": 5784 + }, + { + "epoch": 0.6956051223471412, + "grad_norm": 2.5119585127472117, + "learning_rate": 8.955883843213561e-07, + "loss": 0.7733, + "step": 5785 + }, + { + "epoch": 0.6957253652377803, + "grad_norm": 1.945382480817997, + "learning_rate": 8.949390333452569e-07, + "loss": 0.8755, + "step": 5786 + }, + { + "epoch": 0.6958456081284194, + "grad_norm": 1.6613994655723534, + "learning_rate": 8.942898500060279e-07, + "loss": 0.6809, + "step": 5787 + }, + { + "epoch": 0.6959658510190585, + "grad_norm": 4.880526417965083, + "learning_rate": 8.936408344021493e-07, + "loss": 0.7261, + "step": 5788 + }, + { + "epoch": 0.6960860939096976, + "grad_norm": 2.004506398655838, + "learning_rate": 8.929919866320765e-07, + "loss": 0.7131, + "step": 5789 + }, + { + "epoch": 0.6962063368003367, + "grad_norm": 2.014036040209968, + "learning_rate": 8.923433067942385e-07, + "loss": 0.818, + "step": 5790 + }, + { + "epoch": 0.6963265796909758, + "grad_norm": 2.0113302207486896, + "learning_rate": 8.916947949870417e-07, + "loss": 0.684, + "step": 5791 + }, + { + "epoch": 0.6964468225816148, + "grad_norm": 0.7529824877877161, + "learning_rate": 8.910464513088615e-07, + "loss": 0.5818, + "step": 5792 + }, + { + "epoch": 0.696567065472254, + "grad_norm": 1.9406204936148739, + "learning_rate": 8.903982758580542e-07, + "loss": 0.7918, + "step": 5793 + }, + { + "epoch": 0.696687308362893, + "grad_norm": 2.3203576080356023, + "learning_rate": 8.897502687329457e-07, + "loss": 0.8099, + "step": 5794 + }, + { + "epoch": 0.6968075512535321, + "grad_norm": 2.1639850478651725, + "learning_rate": 8.891024300318382e-07, + "loss": 0.804, + "step": 5795 + }, + { + "epoch": 0.6969277941441713, + "grad_norm": 2.0846160246030294, + "learning_rate": 8.884547598530103e-07, + "loss": 0.7605, + "step": 5796 + }, + { + "epoch": 0.6970480370348103, + "grad_norm": 1.7926983930750693, + "learning_rate": 8.8780725829471e-07, + "loss": 0.7535, + "step": 5797 + }, + { + "epoch": 0.6971682799254494, + "grad_norm": 1.982817872433189, + "learning_rate": 8.87159925455165e-07, + "loss": 0.7823, + "step": 5798 + }, + { + "epoch": 0.6972885228160886, + "grad_norm": 2.190511755526236, + "learning_rate": 8.865127614325738e-07, + "loss": 0.7368, + "step": 5799 + }, + { + "epoch": 0.6974087657067276, + "grad_norm": 1.8140081510854877, + "learning_rate": 8.85865766325113e-07, + "loss": 0.6672, + "step": 5800 + }, + { + "epoch": 0.6975290085973667, + "grad_norm": 2.1812095660202937, + "learning_rate": 8.852189402309287e-07, + "loss": 0.7257, + "step": 5801 + }, + { + "epoch": 0.6976492514880057, + "grad_norm": 2.96963375389343, + "learning_rate": 8.845722832481441e-07, + "loss": 0.7442, + "step": 5802 + }, + { + "epoch": 0.6977694943786449, + "grad_norm": 1.9572035053152397, + "learning_rate": 8.83925795474858e-07, + "loss": 0.7755, + "step": 5803 + }, + { + "epoch": 0.6978897372692839, + "grad_norm": 3.506416977766999, + "learning_rate": 8.832794770091414e-07, + "loss": 0.595, + "step": 5804 + }, + { + "epoch": 0.698009980159923, + "grad_norm": 1.9260360208789307, + "learning_rate": 8.826333279490401e-07, + "loss": 0.8302, + "step": 5805 + }, + { + "epoch": 0.6981302230505622, + "grad_norm": 2.256772233962725, + "learning_rate": 8.819873483925748e-07, + "loss": 0.687, + "step": 5806 + }, + { + "epoch": 0.6982504659412012, + "grad_norm": 2.3261162201111345, + "learning_rate": 8.81341538437739e-07, + "loss": 0.7526, + "step": 5807 + }, + { + "epoch": 0.6983707088318403, + "grad_norm": 1.872117306114984, + "learning_rate": 8.80695898182503e-07, + "loss": 0.6845, + "step": 5808 + }, + { + "epoch": 0.6984909517224794, + "grad_norm": 0.8263980564887614, + "learning_rate": 8.800504277248093e-07, + "loss": 0.6507, + "step": 5809 + }, + { + "epoch": 0.6986111946131185, + "grad_norm": 1.9124024958469978, + "learning_rate": 8.794051271625753e-07, + "loss": 0.7552, + "step": 5810 + }, + { + "epoch": 0.6987314375037575, + "grad_norm": 1.6312197715326222, + "learning_rate": 8.787599965936925e-07, + "loss": 0.832, + "step": 5811 + }, + { + "epoch": 0.6988516803943967, + "grad_norm": 1.6607366287908265, + "learning_rate": 8.781150361160261e-07, + "loss": 0.7226, + "step": 5812 + }, + { + "epoch": 0.6989719232850358, + "grad_norm": 1.7304305853843835, + "learning_rate": 8.774702458274181e-07, + "loss": 0.7368, + "step": 5813 + }, + { + "epoch": 0.6990921661756748, + "grad_norm": 2.7811352995462704, + "learning_rate": 8.768256258256799e-07, + "loss": 0.7088, + "step": 5814 + }, + { + "epoch": 0.699212409066314, + "grad_norm": 2.120485446879093, + "learning_rate": 8.76181176208602e-07, + "loss": 0.7429, + "step": 5815 + }, + { + "epoch": 0.699332651956953, + "grad_norm": 1.9365020150573788, + "learning_rate": 8.755368970739461e-07, + "loss": 0.7361, + "step": 5816 + }, + { + "epoch": 0.6994528948475921, + "grad_norm": 2.3437082922721286, + "learning_rate": 8.748927885194479e-07, + "loss": 0.6182, + "step": 5817 + }, + { + "epoch": 0.6995731377382313, + "grad_norm": 0.8291905942815861, + "learning_rate": 8.742488506428209e-07, + "loss": 0.5743, + "step": 5818 + }, + { + "epoch": 0.6996933806288703, + "grad_norm": 1.856656306846841, + "learning_rate": 8.736050835417466e-07, + "loss": 0.7821, + "step": 5819 + }, + { + "epoch": 0.6998136235195094, + "grad_norm": 3.867993943100204, + "learning_rate": 8.729614873138862e-07, + "loss": 0.6184, + "step": 5820 + }, + { + "epoch": 0.6999338664101485, + "grad_norm": 1.997109111977927, + "learning_rate": 8.723180620568716e-07, + "loss": 0.7788, + "step": 5821 + }, + { + "epoch": 0.7000541093007876, + "grad_norm": 2.8242195941929062, + "learning_rate": 8.716748078683116e-07, + "loss": 0.8524, + "step": 5822 + }, + { + "epoch": 0.7001743521914267, + "grad_norm": 1.9845170097900386, + "learning_rate": 8.710317248457855e-07, + "loss": 0.6883, + "step": 5823 + }, + { + "epoch": 0.7002945950820658, + "grad_norm": 1.8460782108952447, + "learning_rate": 8.703888130868482e-07, + "loss": 0.7189, + "step": 5824 + }, + { + "epoch": 0.7004148379727049, + "grad_norm": 2.0671102842411617, + "learning_rate": 8.697460726890307e-07, + "loss": 0.8236, + "step": 5825 + }, + { + "epoch": 0.7005350808633439, + "grad_norm": 2.5390440015852587, + "learning_rate": 8.691035037498354e-07, + "loss": 0.9104, + "step": 5826 + }, + { + "epoch": 0.7006553237539831, + "grad_norm": 1.9223423211328752, + "learning_rate": 8.684611063667391e-07, + "loss": 0.7228, + "step": 5827 + }, + { + "epoch": 0.7007755666446221, + "grad_norm": 1.842866200839876, + "learning_rate": 8.678188806371935e-07, + "loss": 0.7755, + "step": 5828 + }, + { + "epoch": 0.7008958095352612, + "grad_norm": 1.74970058877079, + "learning_rate": 8.671768266586228e-07, + "loss": 0.8568, + "step": 5829 + }, + { + "epoch": 0.7010160524259004, + "grad_norm": 3.670077918989939, + "learning_rate": 8.665349445284275e-07, + "loss": 0.785, + "step": 5830 + }, + { + "epoch": 0.7011362953165394, + "grad_norm": 1.486755149343607, + "learning_rate": 8.658932343439799e-07, + "loss": 0.8082, + "step": 5831 + }, + { + "epoch": 0.7012565382071785, + "grad_norm": 2.1237108022430267, + "learning_rate": 8.65251696202627e-07, + "loss": 0.7851, + "step": 5832 + }, + { + "epoch": 0.7013767810978175, + "grad_norm": 2.3346368236860187, + "learning_rate": 8.646103302016896e-07, + "loss": 0.8771, + "step": 5833 + }, + { + "epoch": 0.7014970239884567, + "grad_norm": 2.7776665050684013, + "learning_rate": 8.639691364384614e-07, + "loss": 0.8894, + "step": 5834 + }, + { + "epoch": 0.7016172668790958, + "grad_norm": 1.9676142247159851, + "learning_rate": 8.633281150102136e-07, + "loss": 0.73, + "step": 5835 + }, + { + "epoch": 0.7017375097697348, + "grad_norm": 4.027815664412817, + "learning_rate": 8.626872660141855e-07, + "loss": 0.6851, + "step": 5836 + }, + { + "epoch": 0.701857752660374, + "grad_norm": 2.1122321100063193, + "learning_rate": 8.620465895475957e-07, + "loss": 0.7502, + "step": 5837 + }, + { + "epoch": 0.701977995551013, + "grad_norm": 1.5026961573358644, + "learning_rate": 8.614060857076333e-07, + "loss": 0.7543, + "step": 5838 + }, + { + "epoch": 0.7020982384416521, + "grad_norm": 2.0307065970481006, + "learning_rate": 8.60765754591462e-07, + "loss": 0.7491, + "step": 5839 + }, + { + "epoch": 0.7022184813322913, + "grad_norm": 1.8014036745464448, + "learning_rate": 8.601255962962211e-07, + "loss": 0.7341, + "step": 5840 + }, + { + "epoch": 0.7023387242229303, + "grad_norm": 4.8064674119922985, + "learning_rate": 8.594856109190194e-07, + "loss": 0.7297, + "step": 5841 + }, + { + "epoch": 0.7024589671135694, + "grad_norm": 1.6876004507049298, + "learning_rate": 8.588457985569446e-07, + "loss": 0.6942, + "step": 5842 + }, + { + "epoch": 0.7025792100042085, + "grad_norm": 2.184179783558189, + "learning_rate": 8.582061593070542e-07, + "loss": 0.7248, + "step": 5843 + }, + { + "epoch": 0.7026994528948476, + "grad_norm": 2.1723018748169536, + "learning_rate": 8.57566693266383e-07, + "loss": 0.7629, + "step": 5844 + }, + { + "epoch": 0.7028196957854866, + "grad_norm": 3.7570976481496903, + "learning_rate": 8.569274005319354e-07, + "loss": 0.698, + "step": 5845 + }, + { + "epoch": 0.7029399386761258, + "grad_norm": 2.007537229199881, + "learning_rate": 8.562882812006913e-07, + "loss": 0.8, + "step": 5846 + }, + { + "epoch": 0.7030601815667649, + "grad_norm": 1.7144324432884772, + "learning_rate": 8.556493353696066e-07, + "loss": 0.781, + "step": 5847 + }, + { + "epoch": 0.7031804244574039, + "grad_norm": 2.3920098575784428, + "learning_rate": 8.550105631356077e-07, + "loss": 0.6803, + "step": 5848 + }, + { + "epoch": 0.7033006673480431, + "grad_norm": 2.0911953272174095, + "learning_rate": 8.543719645955961e-07, + "loss": 0.7783, + "step": 5849 + }, + { + "epoch": 0.7034209102386821, + "grad_norm": 1.736530648296875, + "learning_rate": 8.537335398464467e-07, + "loss": 0.7496, + "step": 5850 + }, + { + "epoch": 0.7035411531293212, + "grad_norm": 3.1388616213744127, + "learning_rate": 8.53095288985007e-07, + "loss": 0.8536, + "step": 5851 + }, + { + "epoch": 0.7036613960199604, + "grad_norm": 1.9731725640253783, + "learning_rate": 8.524572121081009e-07, + "loss": 0.824, + "step": 5852 + }, + { + "epoch": 0.7037816389105994, + "grad_norm": 1.9297091216447455, + "learning_rate": 8.518193093125232e-07, + "loss": 0.6256, + "step": 5853 + }, + { + "epoch": 0.7039018818012385, + "grad_norm": 1.7681578976215984, + "learning_rate": 8.511815806950436e-07, + "loss": 0.813, + "step": 5854 + }, + { + "epoch": 0.7040221246918776, + "grad_norm": 1.7595524293590141, + "learning_rate": 8.505440263524044e-07, + "loss": 0.7827, + "step": 5855 + }, + { + "epoch": 0.7041423675825167, + "grad_norm": 2.8968720437109883, + "learning_rate": 8.49906646381322e-07, + "loss": 0.8849, + "step": 5856 + }, + { + "epoch": 0.7042626104731557, + "grad_norm": 2.0119740148273118, + "learning_rate": 8.492694408784884e-07, + "loss": 0.727, + "step": 5857 + }, + { + "epoch": 0.7043828533637949, + "grad_norm": 2.505343829407901, + "learning_rate": 8.486324099405642e-07, + "loss": 0.6299, + "step": 5858 + }, + { + "epoch": 0.704503096254434, + "grad_norm": 3.5475115099902332, + "learning_rate": 8.479955536641887e-07, + "loss": 0.7482, + "step": 5859 + }, + { + "epoch": 0.704623339145073, + "grad_norm": 1.9332760944881957, + "learning_rate": 8.473588721459716e-07, + "loss": 0.6674, + "step": 5860 + }, + { + "epoch": 0.7047435820357122, + "grad_norm": 2.0560502439363733, + "learning_rate": 8.467223654824967e-07, + "loss": 0.7097, + "step": 5861 + }, + { + "epoch": 0.7048638249263512, + "grad_norm": 2.371108229465674, + "learning_rate": 8.460860337703233e-07, + "loss": 0.627, + "step": 5862 + }, + { + "epoch": 0.7049840678169903, + "grad_norm": 1.8981571068935683, + "learning_rate": 8.454498771059797e-07, + "loss": 0.7094, + "step": 5863 + }, + { + "epoch": 0.7051043107076294, + "grad_norm": 2.2562663450203813, + "learning_rate": 8.448138955859725e-07, + "loss": 0.8413, + "step": 5864 + }, + { + "epoch": 0.7052245535982685, + "grad_norm": 2.092311649755688, + "learning_rate": 8.44178089306778e-07, + "loss": 0.9025, + "step": 5865 + }, + { + "epoch": 0.7053447964889076, + "grad_norm": 1.8213664920496386, + "learning_rate": 8.4354245836485e-07, + "loss": 0.768, + "step": 5866 + }, + { + "epoch": 0.7054650393795466, + "grad_norm": 1.5516758411571934, + "learning_rate": 8.429070028566108e-07, + "loss": 0.7282, + "step": 5867 + }, + { + "epoch": 0.7055852822701858, + "grad_norm": 11.009361098597703, + "learning_rate": 8.422717228784586e-07, + "loss": 0.7509, + "step": 5868 + }, + { + "epoch": 0.7057055251608249, + "grad_norm": 1.980705472804619, + "learning_rate": 8.416366185267663e-07, + "loss": 0.6961, + "step": 5869 + }, + { + "epoch": 0.7058257680514639, + "grad_norm": 1.7241599511516184, + "learning_rate": 8.410016898978778e-07, + "loss": 0.7801, + "step": 5870 + }, + { + "epoch": 0.7059460109421031, + "grad_norm": 1.7400593439483711, + "learning_rate": 8.403669370881115e-07, + "loss": 0.791, + "step": 5871 + }, + { + "epoch": 0.7060662538327421, + "grad_norm": 1.69302621839831, + "learning_rate": 8.397323601937587e-07, + "loss": 0.7912, + "step": 5872 + }, + { + "epoch": 0.7061864967233812, + "grad_norm": 5.050294266993513, + "learning_rate": 8.390979593110838e-07, + "loss": 0.773, + "step": 5873 + }, + { + "epoch": 0.7063067396140204, + "grad_norm": 1.7342481137161367, + "learning_rate": 8.384637345363262e-07, + "loss": 0.8186, + "step": 5874 + }, + { + "epoch": 0.7064269825046594, + "grad_norm": 1.787267125572578, + "learning_rate": 8.378296859656964e-07, + "loss": 0.7711, + "step": 5875 + }, + { + "epoch": 0.7065472253952985, + "grad_norm": 2.0508830889892646, + "learning_rate": 8.371958136953792e-07, + "loss": 0.683, + "step": 5876 + }, + { + "epoch": 0.7066674682859376, + "grad_norm": 2.6162524930210327, + "learning_rate": 8.365621178215326e-07, + "loss": 0.6706, + "step": 5877 + }, + { + "epoch": 0.7067877111765767, + "grad_norm": 2.0711299793782842, + "learning_rate": 8.359285984402871e-07, + "loss": 0.7552, + "step": 5878 + }, + { + "epoch": 0.7069079540672157, + "grad_norm": 2.0566733159349844, + "learning_rate": 8.352952556477489e-07, + "loss": 0.7416, + "step": 5879 + }, + { + "epoch": 0.7070281969578549, + "grad_norm": 2.2338352218278787, + "learning_rate": 8.34662089539993e-07, + "loss": 0.7723, + "step": 5880 + }, + { + "epoch": 0.707148439848494, + "grad_norm": 2.1981221203951917, + "learning_rate": 8.340291002130722e-07, + "loss": 0.7999, + "step": 5881 + }, + { + "epoch": 0.707268682739133, + "grad_norm": 2.2528338070868688, + "learning_rate": 8.3339628776301e-07, + "loss": 0.7985, + "step": 5882 + }, + { + "epoch": 0.7073889256297722, + "grad_norm": 2.04249336736529, + "learning_rate": 8.327636522858033e-07, + "loss": 0.5766, + "step": 5883 + }, + { + "epoch": 0.7075091685204112, + "grad_norm": 2.02497728594062, + "learning_rate": 8.321311938774225e-07, + "loss": 0.7808, + "step": 5884 + }, + { + "epoch": 0.7076294114110503, + "grad_norm": 4.332319630138749, + "learning_rate": 8.314989126338104e-07, + "loss": 0.7987, + "step": 5885 + }, + { + "epoch": 0.7077496543016895, + "grad_norm": 6.730070858450572, + "learning_rate": 8.308668086508847e-07, + "loss": 0.8429, + "step": 5886 + }, + { + "epoch": 0.7078698971923285, + "grad_norm": 2.0513810506797343, + "learning_rate": 8.302348820245342e-07, + "loss": 0.7397, + "step": 5887 + }, + { + "epoch": 0.7079901400829676, + "grad_norm": 2.566323382057286, + "learning_rate": 8.296031328506232e-07, + "loss": 0.7053, + "step": 5888 + }, + { + "epoch": 0.7081103829736067, + "grad_norm": 2.388518458732077, + "learning_rate": 8.289715612249857e-07, + "loss": 0.7591, + "step": 5889 + }, + { + "epoch": 0.7082306258642458, + "grad_norm": 2.2034159799600728, + "learning_rate": 8.283401672434305e-07, + "loss": 0.7781, + "step": 5890 + }, + { + "epoch": 0.7083508687548848, + "grad_norm": 2.0525338957911465, + "learning_rate": 8.277089510017412e-07, + "loss": 0.7073, + "step": 5891 + }, + { + "epoch": 0.708471111645524, + "grad_norm": 1.7311609403653183, + "learning_rate": 8.270779125956719e-07, + "loss": 0.8249, + "step": 5892 + }, + { + "epoch": 0.7085913545361631, + "grad_norm": 2.1085983542031292, + "learning_rate": 8.264470521209505e-07, + "loss": 0.7993, + "step": 5893 + }, + { + "epoch": 0.7087115974268021, + "grad_norm": 2.1822357049168555, + "learning_rate": 8.258163696732785e-07, + "loss": 0.773, + "step": 5894 + }, + { + "epoch": 0.7088318403174413, + "grad_norm": 2.2969309482118248, + "learning_rate": 8.251858653483288e-07, + "loss": 0.7722, + "step": 5895 + }, + { + "epoch": 0.7089520832080803, + "grad_norm": 2.1756939142803327, + "learning_rate": 8.245555392417501e-07, + "loss": 0.8615, + "step": 5896 + }, + { + "epoch": 0.7090723260987194, + "grad_norm": 1.8306770443901168, + "learning_rate": 8.239253914491613e-07, + "loss": 0.7866, + "step": 5897 + }, + { + "epoch": 0.7091925689893585, + "grad_norm": 1.8367270212705216, + "learning_rate": 8.232954220661556e-07, + "loss": 0.7586, + "step": 5898 + }, + { + "epoch": 0.7093128118799976, + "grad_norm": 2.5303539997969318, + "learning_rate": 8.226656311882989e-07, + "loss": 0.7052, + "step": 5899 + }, + { + "epoch": 0.7094330547706367, + "grad_norm": 2.1182036829224407, + "learning_rate": 8.22036018911129e-07, + "loss": 0.7626, + "step": 5900 + }, + { + "epoch": 0.7095532976612757, + "grad_norm": 2.3631387967138493, + "learning_rate": 8.214065853301599e-07, + "loss": 0.8155, + "step": 5901 + }, + { + "epoch": 0.7096735405519149, + "grad_norm": 0.7993022006292462, + "learning_rate": 8.207773305408734e-07, + "loss": 0.5823, + "step": 5902 + }, + { + "epoch": 0.709793783442554, + "grad_norm": 2.2971164291030135, + "learning_rate": 8.201482546387288e-07, + "loss": 0.8058, + "step": 5903 + }, + { + "epoch": 0.709914026333193, + "grad_norm": 1.7573131614987991, + "learning_rate": 8.195193577191553e-07, + "loss": 0.92, + "step": 5904 + }, + { + "epoch": 0.7100342692238322, + "grad_norm": 1.6470645049942825, + "learning_rate": 8.188906398775579e-07, + "loss": 0.848, + "step": 5905 + }, + { + "epoch": 0.7101545121144712, + "grad_norm": 2.1266007845289234, + "learning_rate": 8.18262101209311e-07, + "loss": 0.6906, + "step": 5906 + }, + { + "epoch": 0.7102747550051103, + "grad_norm": 1.697203745882754, + "learning_rate": 8.176337418097626e-07, + "loss": 0.7083, + "step": 5907 + }, + { + "epoch": 0.7103949978957494, + "grad_norm": 1.9716811990188918, + "learning_rate": 8.170055617742364e-07, + "loss": 0.8035, + "step": 5908 + }, + { + "epoch": 0.7105152407863885, + "grad_norm": 2.067300584948065, + "learning_rate": 8.163775611980252e-07, + "loss": 0.7111, + "step": 5909 + }, + { + "epoch": 0.7106354836770276, + "grad_norm": 1.7589495562877695, + "learning_rate": 8.157497401763982e-07, + "loss": 0.7976, + "step": 5910 + }, + { + "epoch": 0.7107557265676667, + "grad_norm": 1.9104877332519699, + "learning_rate": 8.151220988045935e-07, + "loss": 0.78, + "step": 5911 + }, + { + "epoch": 0.7108759694583058, + "grad_norm": 2.070825592374501, + "learning_rate": 8.144946371778234e-07, + "loss": 0.8316, + "step": 5912 + }, + { + "epoch": 0.7109962123489448, + "grad_norm": 1.7734871743929264, + "learning_rate": 8.138673553912751e-07, + "loss": 0.7813, + "step": 5913 + }, + { + "epoch": 0.711116455239584, + "grad_norm": 2.320330378916282, + "learning_rate": 8.132402535401059e-07, + "loss": 0.5711, + "step": 5914 + }, + { + "epoch": 0.711236698130223, + "grad_norm": 1.7806068731732987, + "learning_rate": 8.126133317194465e-07, + "loss": 0.7456, + "step": 5915 + }, + { + "epoch": 0.7113569410208621, + "grad_norm": 1.8616709278414234, + "learning_rate": 8.11986590024401e-07, + "loss": 0.7474, + "step": 5916 + }, + { + "epoch": 0.7114771839115013, + "grad_norm": 1.79844411257167, + "learning_rate": 8.113600285500442e-07, + "loss": 0.6921, + "step": 5917 + }, + { + "epoch": 0.7115974268021403, + "grad_norm": 1.7622425459456388, + "learning_rate": 8.107336473914268e-07, + "loss": 0.7471, + "step": 5918 + }, + { + "epoch": 0.7117176696927794, + "grad_norm": 0.7773804775743711, + "learning_rate": 8.101074466435694e-07, + "loss": 0.5583, + "step": 5919 + }, + { + "epoch": 0.7118379125834186, + "grad_norm": 1.665361855280722, + "learning_rate": 8.094814264014662e-07, + "loss": 0.68, + "step": 5920 + }, + { + "epoch": 0.7119581554740576, + "grad_norm": 2.4668073132944235, + "learning_rate": 8.088555867600844e-07, + "loss": 0.8224, + "step": 5921 + }, + { + "epoch": 0.7120783983646967, + "grad_norm": 1.851246802863658, + "learning_rate": 8.08229927814362e-07, + "loss": 0.6096, + "step": 5922 + }, + { + "epoch": 0.7121986412553358, + "grad_norm": 1.7566871261973585, + "learning_rate": 8.076044496592134e-07, + "loss": 0.6547, + "step": 5923 + }, + { + "epoch": 0.7123188841459749, + "grad_norm": 2.412395012503155, + "learning_rate": 8.069791523895204e-07, + "loss": 0.7858, + "step": 5924 + }, + { + "epoch": 0.7124391270366139, + "grad_norm": 1.9614626724538216, + "learning_rate": 8.063540361001422e-07, + "loss": 0.7758, + "step": 5925 + }, + { + "epoch": 0.7125593699272531, + "grad_norm": 2.1392063535548878, + "learning_rate": 8.057291008859069e-07, + "loss": 0.8036, + "step": 5926 + }, + { + "epoch": 0.7126796128178922, + "grad_norm": 1.860722414233792, + "learning_rate": 8.051043468416187e-07, + "loss": 0.686, + "step": 5927 + }, + { + "epoch": 0.7127998557085312, + "grad_norm": 2.222249454755284, + "learning_rate": 8.044797740620506e-07, + "loss": 0.8321, + "step": 5928 + }, + { + "epoch": 0.7129200985991703, + "grad_norm": 2.125429110116385, + "learning_rate": 8.038553826419494e-07, + "loss": 0.7942, + "step": 5929 + }, + { + "epoch": 0.7130403414898094, + "grad_norm": 1.7155765019355826, + "learning_rate": 8.032311726760364e-07, + "loss": 0.8155, + "step": 5930 + }, + { + "epoch": 0.7131605843804485, + "grad_norm": 1.812890898546898, + "learning_rate": 8.026071442590022e-07, + "loss": 0.6987, + "step": 5931 + }, + { + "epoch": 0.7132808272710875, + "grad_norm": 1.942750522949714, + "learning_rate": 8.019832974855134e-07, + "loss": 0.811, + "step": 5932 + }, + { + "epoch": 0.7134010701617267, + "grad_norm": 2.473496714574183, + "learning_rate": 8.013596324502052e-07, + "loss": 0.8275, + "step": 5933 + }, + { + "epoch": 0.7135213130523658, + "grad_norm": 1.859442316968466, + "learning_rate": 8.007361492476872e-07, + "loss": 0.7986, + "step": 5934 + }, + { + "epoch": 0.7136415559430048, + "grad_norm": 1.7160948618920246, + "learning_rate": 8.001128479725426e-07, + "loss": 0.7945, + "step": 5935 + }, + { + "epoch": 0.713761798833644, + "grad_norm": 5.524891447493951, + "learning_rate": 7.994897287193248e-07, + "loss": 0.8144, + "step": 5936 + }, + { + "epoch": 0.713882041724283, + "grad_norm": 2.2879736071311467, + "learning_rate": 7.988667915825605e-07, + "loss": 0.8424, + "step": 5937 + }, + { + "epoch": 0.7140022846149221, + "grad_norm": 2.177886109133788, + "learning_rate": 7.982440366567491e-07, + "loss": 0.761, + "step": 5938 + }, + { + "epoch": 0.7141225275055613, + "grad_norm": 1.8097132602149983, + "learning_rate": 7.97621464036361e-07, + "loss": 0.7538, + "step": 5939 + }, + { + "epoch": 0.7142427703962003, + "grad_norm": 1.7351671076833792, + "learning_rate": 7.969990738158417e-07, + "loss": 0.6813, + "step": 5940 + }, + { + "epoch": 0.7143630132868394, + "grad_norm": 2.170220602574003, + "learning_rate": 7.963768660896062e-07, + "loss": 0.8523, + "step": 5941 + }, + { + "epoch": 0.7144832561774785, + "grad_norm": 2.0530976501222487, + "learning_rate": 7.957548409520432e-07, + "loss": 0.8293, + "step": 5942 + }, + { + "epoch": 0.7146034990681176, + "grad_norm": 2.233936186696135, + "learning_rate": 7.951329984975135e-07, + "loss": 0.8497, + "step": 5943 + }, + { + "epoch": 0.7147237419587567, + "grad_norm": 0.7258594762325575, + "learning_rate": 7.94511338820349e-07, + "loss": 0.5431, + "step": 5944 + }, + { + "epoch": 0.7148439848493958, + "grad_norm": 1.9629526916104556, + "learning_rate": 7.938898620148575e-07, + "loss": 0.7885, + "step": 5945 + }, + { + "epoch": 0.7149642277400349, + "grad_norm": 1.9841392557279354, + "learning_rate": 7.932685681753135e-07, + "loss": 0.7147, + "step": 5946 + }, + { + "epoch": 0.7150844706306739, + "grad_norm": 2.2435803490047297, + "learning_rate": 7.92647457395969e-07, + "loss": 0.6306, + "step": 5947 + }, + { + "epoch": 0.7152047135213131, + "grad_norm": 2.251205121598589, + "learning_rate": 7.920265297710444e-07, + "loss": 0.7443, + "step": 5948 + }, + { + "epoch": 0.7153249564119522, + "grad_norm": 1.9581540478823511, + "learning_rate": 7.914057853947363e-07, + "loss": 0.7347, + "step": 5949 + }, + { + "epoch": 0.7154451993025912, + "grad_norm": 1.9278968812108492, + "learning_rate": 7.907852243612089e-07, + "loss": 0.6318, + "step": 5950 + }, + { + "epoch": 0.7155654421932304, + "grad_norm": 1.8655008654933651, + "learning_rate": 7.901648467646009e-07, + "loss": 0.7277, + "step": 5951 + }, + { + "epoch": 0.7156856850838694, + "grad_norm": 1.6492833864384067, + "learning_rate": 7.895446526990244e-07, + "loss": 0.7309, + "step": 5952 + }, + { + "epoch": 0.7158059279745085, + "grad_norm": 1.595448056849772, + "learning_rate": 7.889246422585609e-07, + "loss": 0.7559, + "step": 5953 + }, + { + "epoch": 0.7159261708651476, + "grad_norm": 2.6952714557563295, + "learning_rate": 7.883048155372675e-07, + "loss": 0.7357, + "step": 5954 + }, + { + "epoch": 0.7160464137557867, + "grad_norm": 2.5554005039097256, + "learning_rate": 7.876851726291698e-07, + "loss": 0.7163, + "step": 5955 + }, + { + "epoch": 0.7161666566464258, + "grad_norm": 2.2576063649412927, + "learning_rate": 7.870657136282666e-07, + "loss": 0.7865, + "step": 5956 + }, + { + "epoch": 0.7162868995370649, + "grad_norm": 1.5449039733081031, + "learning_rate": 7.86446438628531e-07, + "loss": 0.8205, + "step": 5957 + }, + { + "epoch": 0.716407142427704, + "grad_norm": 0.7734517568179782, + "learning_rate": 7.858273477239059e-07, + "loss": 0.5689, + "step": 5958 + }, + { + "epoch": 0.716527385318343, + "grad_norm": 1.7166611896662083, + "learning_rate": 7.852084410083067e-07, + "loss": 0.7166, + "step": 5959 + }, + { + "epoch": 0.7166476282089821, + "grad_norm": 3.188957629051803, + "learning_rate": 7.84589718575621e-07, + "loss": 0.6437, + "step": 5960 + }, + { + "epoch": 0.7167678710996213, + "grad_norm": 1.9231380871069292, + "learning_rate": 7.83971180519708e-07, + "loss": 0.689, + "step": 5961 + }, + { + "epoch": 0.7168881139902603, + "grad_norm": 2.2068369011843525, + "learning_rate": 7.833528269344008e-07, + "loss": 0.7611, + "step": 5962 + }, + { + "epoch": 0.7170083568808994, + "grad_norm": 7.698245594256667, + "learning_rate": 7.827346579135023e-07, + "loss": 0.7796, + "step": 5963 + }, + { + "epoch": 0.7171285997715385, + "grad_norm": 2.1290519454517507, + "learning_rate": 7.821166735507885e-07, + "loss": 0.8371, + "step": 5964 + }, + { + "epoch": 0.7172488426621776, + "grad_norm": 1.7677811816851627, + "learning_rate": 7.81498873940007e-07, + "loss": 0.6908, + "step": 5965 + }, + { + "epoch": 0.7173690855528166, + "grad_norm": 38.69291567758641, + "learning_rate": 7.808812591748768e-07, + "loss": 0.7847, + "step": 5966 + }, + { + "epoch": 0.7174893284434558, + "grad_norm": 2.0613934409510466, + "learning_rate": 7.802638293490915e-07, + "loss": 0.6551, + "step": 5967 + }, + { + "epoch": 0.7176095713340949, + "grad_norm": 1.6081616170000488, + "learning_rate": 7.796465845563123e-07, + "loss": 0.7699, + "step": 5968 + }, + { + "epoch": 0.7177298142247339, + "grad_norm": 1.79118634224626, + "learning_rate": 7.790295248901766e-07, + "loss": 0.7976, + "step": 5969 + }, + { + "epoch": 0.7178500571153731, + "grad_norm": 1.7741514272947634, + "learning_rate": 7.784126504442902e-07, + "loss": 0.6301, + "step": 5970 + }, + { + "epoch": 0.7179703000060121, + "grad_norm": 1.5922737015888582, + "learning_rate": 7.777959613122351e-07, + "loss": 0.6771, + "step": 5971 + }, + { + "epoch": 0.7180905428966512, + "grad_norm": 1.7288161539148301, + "learning_rate": 7.771794575875604e-07, + "loss": 0.7845, + "step": 5972 + }, + { + "epoch": 0.7182107857872904, + "grad_norm": 2.6604656202033845, + "learning_rate": 7.765631393637888e-07, + "loss": 0.7762, + "step": 5973 + }, + { + "epoch": 0.7183310286779294, + "grad_norm": 2.547260091489255, + "learning_rate": 7.75947006734417e-07, + "loss": 0.4899, + "step": 5974 + }, + { + "epoch": 0.7184512715685685, + "grad_norm": 2.0924641090879064, + "learning_rate": 7.753310597929101e-07, + "loss": 0.8308, + "step": 5975 + }, + { + "epoch": 0.7185715144592076, + "grad_norm": 0.7626066461620153, + "learning_rate": 7.747152986327095e-07, + "loss": 0.5533, + "step": 5976 + }, + { + "epoch": 0.7186917573498467, + "grad_norm": 1.6748631948144284, + "learning_rate": 7.740997233472228e-07, + "loss": 0.6832, + "step": 5977 + }, + { + "epoch": 0.7188120002404857, + "grad_norm": 2.2949368350576225, + "learning_rate": 7.734843340298329e-07, + "loss": 0.7098, + "step": 5978 + }, + { + "epoch": 0.7189322431311249, + "grad_norm": 6.9256314316847, + "learning_rate": 7.72869130773895e-07, + "loss": 0.7635, + "step": 5979 + }, + { + "epoch": 0.719052486021764, + "grad_norm": 0.7999143047926458, + "learning_rate": 7.722541136727343e-07, + "loss": 0.5938, + "step": 5980 + }, + { + "epoch": 0.719172728912403, + "grad_norm": 1.93778583510111, + "learning_rate": 7.716392828196483e-07, + "loss": 0.8073, + "step": 5981 + }, + { + "epoch": 0.7192929718030422, + "grad_norm": 2.513413968078649, + "learning_rate": 7.710246383079064e-07, + "loss": 0.78, + "step": 5982 + }, + { + "epoch": 0.7194132146936812, + "grad_norm": 2.6942663843981087, + "learning_rate": 7.704101802307492e-07, + "loss": 0.9267, + "step": 5983 + }, + { + "epoch": 0.7195334575843203, + "grad_norm": 2.0874717299323127, + "learning_rate": 7.697959086813912e-07, + "loss": 0.8757, + "step": 5984 + }, + { + "epoch": 0.7196537004749595, + "grad_norm": 1.6692514244851648, + "learning_rate": 7.691818237530145e-07, + "loss": 0.8059, + "step": 5985 + }, + { + "epoch": 0.7197739433655985, + "grad_norm": 1.9270597782929257, + "learning_rate": 7.685679255387774e-07, + "loss": 0.7739, + "step": 5986 + }, + { + "epoch": 0.7198941862562376, + "grad_norm": 2.047122852956719, + "learning_rate": 7.679542141318065e-07, + "loss": 0.7668, + "step": 5987 + }, + { + "epoch": 0.7200144291468767, + "grad_norm": 1.8690658761666705, + "learning_rate": 7.673406896252013e-07, + "loss": 0.7542, + "step": 5988 + }, + { + "epoch": 0.7201346720375158, + "grad_norm": 1.6691431561072987, + "learning_rate": 7.667273521120347e-07, + "loss": 0.7881, + "step": 5989 + }, + { + "epoch": 0.7202549149281549, + "grad_norm": 2.081555396232401, + "learning_rate": 7.661142016853468e-07, + "loss": 0.7989, + "step": 5990 + }, + { + "epoch": 0.7203751578187939, + "grad_norm": 1.7104853899303083, + "learning_rate": 7.655012384381543e-07, + "loss": 0.7493, + "step": 5991 + }, + { + "epoch": 0.7204954007094331, + "grad_norm": 1.9598142286740052, + "learning_rate": 7.648884624634415e-07, + "loss": 0.8211, + "step": 5992 + }, + { + "epoch": 0.7206156436000721, + "grad_norm": 1.819042892730477, + "learning_rate": 7.642758738541683e-07, + "loss": 0.8935, + "step": 5993 + }, + { + "epoch": 0.7207358864907112, + "grad_norm": 0.7748348829281676, + "learning_rate": 7.636634727032621e-07, + "loss": 0.6084, + "step": 5994 + }, + { + "epoch": 0.7208561293813504, + "grad_norm": 1.9843182893594284, + "learning_rate": 7.630512591036231e-07, + "loss": 0.794, + "step": 5995 + }, + { + "epoch": 0.7209763722719894, + "grad_norm": 2.2283205821267584, + "learning_rate": 7.624392331481255e-07, + "loss": 0.6485, + "step": 5996 + }, + { + "epoch": 0.7210966151626285, + "grad_norm": 0.7823442425353928, + "learning_rate": 7.618273949296115e-07, + "loss": 0.5186, + "step": 5997 + }, + { + "epoch": 0.7212168580532676, + "grad_norm": 2.0232154989521494, + "learning_rate": 7.612157445408987e-07, + "loss": 0.689, + "step": 5998 + }, + { + "epoch": 0.7213371009439067, + "grad_norm": 2.141786580371484, + "learning_rate": 7.606042820747716e-07, + "loss": 0.7469, + "step": 5999 + }, + { + "epoch": 0.7214573438345457, + "grad_norm": 3.2262453380552514, + "learning_rate": 7.599930076239889e-07, + "loss": 0.8565, + "step": 6000 + }, + { + "epoch": 0.7215775867251849, + "grad_norm": 2.02651131358684, + "learning_rate": 7.593819212812818e-07, + "loss": 0.712, + "step": 6001 + }, + { + "epoch": 0.721697829615824, + "grad_norm": 2.094734871232065, + "learning_rate": 7.587710231393508e-07, + "loss": 0.7195, + "step": 6002 + }, + { + "epoch": 0.721818072506463, + "grad_norm": 2.1903050557708177, + "learning_rate": 7.581603132908685e-07, + "loss": 0.8416, + "step": 6003 + }, + { + "epoch": 0.7219383153971022, + "grad_norm": 2.340194344723412, + "learning_rate": 7.575497918284795e-07, + "loss": 0.792, + "step": 6004 + }, + { + "epoch": 0.7220585582877412, + "grad_norm": 2.01823776407359, + "learning_rate": 7.569394588447984e-07, + "loss": 0.7504, + "step": 6005 + }, + { + "epoch": 0.7221788011783803, + "grad_norm": 2.338025031054501, + "learning_rate": 7.563293144324146e-07, + "loss": 0.7788, + "step": 6006 + }, + { + "epoch": 0.7222990440690195, + "grad_norm": 1.888037558295174, + "learning_rate": 7.557193586838834e-07, + "loss": 0.8002, + "step": 6007 + }, + { + "epoch": 0.7224192869596585, + "grad_norm": 2.3322266156584974, + "learning_rate": 7.551095916917371e-07, + "loss": 0.7142, + "step": 6008 + }, + { + "epoch": 0.7225395298502976, + "grad_norm": 7.470094184741331, + "learning_rate": 7.545000135484758e-07, + "loss": 0.6744, + "step": 6009 + }, + { + "epoch": 0.7226597727409367, + "grad_norm": 2.031469127866423, + "learning_rate": 7.538906243465714e-07, + "loss": 0.6305, + "step": 6010 + }, + { + "epoch": 0.7227800156315758, + "grad_norm": 2.1164524730930805, + "learning_rate": 7.5328142417847e-07, + "loss": 0.7896, + "step": 6011 + }, + { + "epoch": 0.7229002585222148, + "grad_norm": 1.6042909780449206, + "learning_rate": 7.526724131365838e-07, + "loss": 0.6955, + "step": 6012 + }, + { + "epoch": 0.723020501412854, + "grad_norm": 1.720867586449217, + "learning_rate": 7.520635913133017e-07, + "loss": 0.7092, + "step": 6013 + }, + { + "epoch": 0.7231407443034931, + "grad_norm": 1.7753477185796152, + "learning_rate": 7.514549588009798e-07, + "loss": 0.8229, + "step": 6014 + }, + { + "epoch": 0.7232609871941321, + "grad_norm": 2.6434428512315242, + "learning_rate": 7.508465156919492e-07, + "loss": 0.7083, + "step": 6015 + }, + { + "epoch": 0.7233812300847713, + "grad_norm": 2.763641335685948, + "learning_rate": 7.502382620785083e-07, + "loss": 0.6142, + "step": 6016 + }, + { + "epoch": 0.7235014729754103, + "grad_norm": 0.8167801872165064, + "learning_rate": 7.496301980529289e-07, + "loss": 0.6262, + "step": 6017 + }, + { + "epoch": 0.7236217158660494, + "grad_norm": 1.9200496534046474, + "learning_rate": 7.490223237074547e-07, + "loss": 0.7478, + "step": 6018 + }, + { + "epoch": 0.7237419587566886, + "grad_norm": 1.7851423601073415, + "learning_rate": 7.484146391342989e-07, + "loss": 0.6618, + "step": 6019 + }, + { + "epoch": 0.7238622016473276, + "grad_norm": 2.491957029011965, + "learning_rate": 7.478071444256484e-07, + "loss": 0.5755, + "step": 6020 + }, + { + "epoch": 0.7239824445379667, + "grad_norm": 1.842523393107219, + "learning_rate": 7.471998396736579e-07, + "loss": 0.8004, + "step": 6021 + }, + { + "epoch": 0.7241026874286057, + "grad_norm": 1.822544350522695, + "learning_rate": 7.465927249704549e-07, + "loss": 0.7638, + "step": 6022 + }, + { + "epoch": 0.7242229303192449, + "grad_norm": 1.7784600074696493, + "learning_rate": 7.459858004081398e-07, + "loss": 0.7798, + "step": 6023 + }, + { + "epoch": 0.724343173209884, + "grad_norm": 0.6611555312520095, + "learning_rate": 7.453790660787815e-07, + "loss": 0.5804, + "step": 6024 + }, + { + "epoch": 0.724463416100523, + "grad_norm": 2.215197230528605, + "learning_rate": 7.447725220744214e-07, + "loss": 0.6411, + "step": 6025 + }, + { + "epoch": 0.7245836589911622, + "grad_norm": 2.4638757996591956, + "learning_rate": 7.441661684870717e-07, + "loss": 0.7756, + "step": 6026 + }, + { + "epoch": 0.7247039018818012, + "grad_norm": 1.6985991652824088, + "learning_rate": 7.435600054087152e-07, + "loss": 0.8207, + "step": 6027 + }, + { + "epoch": 0.7248241447724403, + "grad_norm": 2.2719562332343513, + "learning_rate": 7.42954032931308e-07, + "loss": 0.7488, + "step": 6028 + }, + { + "epoch": 0.7249443876630794, + "grad_norm": 1.7917738446256055, + "learning_rate": 7.423482511467733e-07, + "loss": 0.7522, + "step": 6029 + }, + { + "epoch": 0.7250646305537185, + "grad_norm": 2.06293968369165, + "learning_rate": 7.417426601470099e-07, + "loss": 0.6485, + "step": 6030 + }, + { + "epoch": 0.7251848734443576, + "grad_norm": 2.610779618352857, + "learning_rate": 7.411372600238841e-07, + "loss": 0.79, + "step": 6031 + }, + { + "epoch": 0.7253051163349967, + "grad_norm": 2.0416117550478616, + "learning_rate": 7.405320508692346e-07, + "loss": 0.7406, + "step": 6032 + }, + { + "epoch": 0.7254253592256358, + "grad_norm": 2.004288129503065, + "learning_rate": 7.399270327748727e-07, + "loss": 0.7618, + "step": 6033 + }, + { + "epoch": 0.7255456021162748, + "grad_norm": 2.20481656113787, + "learning_rate": 7.39322205832577e-07, + "loss": 0.7442, + "step": 6034 + }, + { + "epoch": 0.725665845006914, + "grad_norm": 2.3525050102610243, + "learning_rate": 7.387175701341009e-07, + "loss": 0.812, + "step": 6035 + }, + { + "epoch": 0.7257860878975531, + "grad_norm": 2.062967189703573, + "learning_rate": 7.381131257711659e-07, + "loss": 0.7269, + "step": 6036 + }, + { + "epoch": 0.7259063307881921, + "grad_norm": 1.8348378236082559, + "learning_rate": 7.375088728354677e-07, + "loss": 0.8378, + "step": 6037 + }, + { + "epoch": 0.7260265736788313, + "grad_norm": 1.5127212394132625, + "learning_rate": 7.369048114186691e-07, + "loss": 0.6744, + "step": 6038 + }, + { + "epoch": 0.7261468165694703, + "grad_norm": 1.8826370474081189, + "learning_rate": 7.363009416124055e-07, + "loss": 0.8321, + "step": 6039 + }, + { + "epoch": 0.7262670594601094, + "grad_norm": 2.201658802620037, + "learning_rate": 7.356972635082852e-07, + "loss": 0.6338, + "step": 6040 + }, + { + "epoch": 0.7263873023507486, + "grad_norm": 1.8113830018425185, + "learning_rate": 7.35093777197884e-07, + "loss": 0.7567, + "step": 6041 + }, + { + "epoch": 0.7265075452413876, + "grad_norm": 2.1660685691864043, + "learning_rate": 7.344904827727525e-07, + "loss": 0.8567, + "step": 6042 + }, + { + "epoch": 0.7266277881320267, + "grad_norm": 4.79628250488847, + "learning_rate": 7.338873803244076e-07, + "loss": 0.7459, + "step": 6043 + }, + { + "epoch": 0.7267480310226658, + "grad_norm": 1.7835158275504042, + "learning_rate": 7.332844699443401e-07, + "loss": 0.8106, + "step": 6044 + }, + { + "epoch": 0.7268682739133049, + "grad_norm": 1.841859009423822, + "learning_rate": 7.326817517240121e-07, + "loss": 0.7561, + "step": 6045 + }, + { + "epoch": 0.7269885168039439, + "grad_norm": 1.9419804113561034, + "learning_rate": 7.320792257548545e-07, + "loss": 0.8364, + "step": 6046 + }, + { + "epoch": 0.7271087596945831, + "grad_norm": 2.393019530392378, + "learning_rate": 7.314768921282704e-07, + "loss": 0.7667, + "step": 6047 + }, + { + "epoch": 0.7272290025852222, + "grad_norm": 2.5978555932675857, + "learning_rate": 7.30874750935633e-07, + "loss": 0.7257, + "step": 6048 + }, + { + "epoch": 0.7273492454758612, + "grad_norm": 1.9189757521838264, + "learning_rate": 7.30272802268286e-07, + "loss": 0.7964, + "step": 6049 + }, + { + "epoch": 0.7274694883665004, + "grad_norm": 1.7926538394475207, + "learning_rate": 7.29671046217547e-07, + "loss": 0.761, + "step": 6050 + }, + { + "epoch": 0.7275897312571394, + "grad_norm": 1.939632472786039, + "learning_rate": 7.290694828746988e-07, + "loss": 0.8276, + "step": 6051 + }, + { + "epoch": 0.7277099741477785, + "grad_norm": 1.939432165180518, + "learning_rate": 7.284681123310004e-07, + "loss": 0.8634, + "step": 6052 + }, + { + "epoch": 0.7278302170384175, + "grad_norm": 1.6507839479076836, + "learning_rate": 7.27866934677678e-07, + "loss": 0.8005, + "step": 6053 + }, + { + "epoch": 0.7279504599290567, + "grad_norm": 1.8841885172361328, + "learning_rate": 7.272659500059297e-07, + "loss": 0.7849, + "step": 6054 + }, + { + "epoch": 0.7280707028196958, + "grad_norm": 2.5468766850190936, + "learning_rate": 7.266651584069264e-07, + "loss": 0.8036, + "step": 6055 + }, + { + "epoch": 0.7281909457103348, + "grad_norm": 1.7116926085906052, + "learning_rate": 7.260645599718045e-07, + "loss": 0.5752, + "step": 6056 + }, + { + "epoch": 0.728311188600974, + "grad_norm": 2.2437751279821185, + "learning_rate": 7.254641547916767e-07, + "loss": 0.6786, + "step": 6057 + }, + { + "epoch": 0.728431431491613, + "grad_norm": 1.8774486447357208, + "learning_rate": 7.248639429576226e-07, + "loss": 0.6951, + "step": 6058 + }, + { + "epoch": 0.7285516743822521, + "grad_norm": 1.7125234003626202, + "learning_rate": 7.242639245606959e-07, + "loss": 0.7199, + "step": 6059 + }, + { + "epoch": 0.7286719172728913, + "grad_norm": 1.921765145984652, + "learning_rate": 7.236640996919168e-07, + "loss": 0.828, + "step": 6060 + }, + { + "epoch": 0.7287921601635303, + "grad_norm": 1.5873973122372966, + "learning_rate": 7.230644684422782e-07, + "loss": 0.7054, + "step": 6061 + }, + { + "epoch": 0.7289124030541694, + "grad_norm": 1.7923479189660891, + "learning_rate": 7.224650309027451e-07, + "loss": 0.8183, + "step": 6062 + }, + { + "epoch": 0.7290326459448085, + "grad_norm": 2.066289714563105, + "learning_rate": 7.218657871642506e-07, + "loss": 0.688, + "step": 6063 + }, + { + "epoch": 0.7291528888354476, + "grad_norm": 2.0609088887042732, + "learning_rate": 7.212667373177012e-07, + "loss": 0.6286, + "step": 6064 + }, + { + "epoch": 0.7292731317260867, + "grad_norm": 1.7429968489258953, + "learning_rate": 7.206678814539704e-07, + "loss": 0.7576, + "step": 6065 + }, + { + "epoch": 0.7293933746167258, + "grad_norm": 1.667752968903445, + "learning_rate": 7.20069219663904e-07, + "loss": 0.7268, + "step": 6066 + }, + { + "epoch": 0.7295136175073649, + "grad_norm": 1.655716288243412, + "learning_rate": 7.1947075203832e-07, + "loss": 0.7998, + "step": 6067 + }, + { + "epoch": 0.7296338603980039, + "grad_norm": 0.8773383503014838, + "learning_rate": 7.188724786680049e-07, + "loss": 0.6014, + "step": 6068 + }, + { + "epoch": 0.7297541032886431, + "grad_norm": 1.7317247504123732, + "learning_rate": 7.182743996437162e-07, + "loss": 0.7597, + "step": 6069 + }, + { + "epoch": 0.7298743461792822, + "grad_norm": 1.9760042993058604, + "learning_rate": 7.176765150561819e-07, + "loss": 0.6895, + "step": 6070 + }, + { + "epoch": 0.7299945890699212, + "grad_norm": 1.8904350752841914, + "learning_rate": 7.170788249961002e-07, + "loss": 0.8067, + "step": 6071 + }, + { + "epoch": 0.7301148319605604, + "grad_norm": 1.9263990544817433, + "learning_rate": 7.164813295541418e-07, + "loss": 0.8859, + "step": 6072 + }, + { + "epoch": 0.7302350748511994, + "grad_norm": 2.121051598230165, + "learning_rate": 7.15884028820944e-07, + "loss": 0.7072, + "step": 6073 + }, + { + "epoch": 0.7303553177418385, + "grad_norm": 2.3261376417637405, + "learning_rate": 7.152869228871185e-07, + "loss": 0.6065, + "step": 6074 + }, + { + "epoch": 0.7304755606324776, + "grad_norm": 1.7092582788315733, + "learning_rate": 7.146900118432457e-07, + "loss": 0.7273, + "step": 6075 + }, + { + "epoch": 0.7305958035231167, + "grad_norm": 1.688650233712241, + "learning_rate": 7.140932957798753e-07, + "loss": 0.8602, + "step": 6076 + }, + { + "epoch": 0.7307160464137558, + "grad_norm": 1.8769531885121282, + "learning_rate": 7.134967747875309e-07, + "loss": 0.7147, + "step": 6077 + }, + { + "epoch": 0.7308362893043949, + "grad_norm": 1.8820206492605664, + "learning_rate": 7.129004489567014e-07, + "loss": 0.8236, + "step": 6078 + }, + { + "epoch": 0.730956532195034, + "grad_norm": 2.255980596546843, + "learning_rate": 7.123043183778512e-07, + "loss": 0.7852, + "step": 6079 + }, + { + "epoch": 0.731076775085673, + "grad_norm": 1.5783061310714361, + "learning_rate": 7.117083831414114e-07, + "loss": 0.6568, + "step": 6080 + }, + { + "epoch": 0.7311970179763122, + "grad_norm": 1.758412576610307, + "learning_rate": 7.11112643337787e-07, + "loss": 0.704, + "step": 6081 + }, + { + "epoch": 0.7313172608669513, + "grad_norm": 2.564740480824185, + "learning_rate": 7.10517099057349e-07, + "loss": 0.7727, + "step": 6082 + }, + { + "epoch": 0.7314375037575903, + "grad_norm": 2.3805589368992277, + "learning_rate": 7.099217503904411e-07, + "loss": 0.6146, + "step": 6083 + }, + { + "epoch": 0.7315577466482295, + "grad_norm": 1.9273576764553415, + "learning_rate": 7.093265974273788e-07, + "loss": 0.9001, + "step": 6084 + }, + { + "epoch": 0.7316779895388685, + "grad_norm": 1.9135154720943806, + "learning_rate": 7.087316402584447e-07, + "loss": 0.7235, + "step": 6085 + }, + { + "epoch": 0.7317982324295076, + "grad_norm": 1.922791177354589, + "learning_rate": 7.081368789738953e-07, + "loss": 0.8619, + "step": 6086 + }, + { + "epoch": 0.7319184753201466, + "grad_norm": 2.3197239601407795, + "learning_rate": 7.075423136639537e-07, + "loss": 0.7811, + "step": 6087 + }, + { + "epoch": 0.7320387182107858, + "grad_norm": 1.7333020852402272, + "learning_rate": 7.069479444188149e-07, + "loss": 0.7506, + "step": 6088 + }, + { + "epoch": 0.7321589611014249, + "grad_norm": 1.6943029394976583, + "learning_rate": 7.063537713286453e-07, + "loss": 0.8238, + "step": 6089 + }, + { + "epoch": 0.7322792039920639, + "grad_norm": 2.0088099514288236, + "learning_rate": 7.057597944835803e-07, + "loss": 0.8131, + "step": 6090 + }, + { + "epoch": 0.7323994468827031, + "grad_norm": 1.6487573973147764, + "learning_rate": 7.051660139737253e-07, + "loss": 0.7463, + "step": 6091 + }, + { + "epoch": 0.7325196897733421, + "grad_norm": 2.001578185163708, + "learning_rate": 7.045724298891565e-07, + "loss": 0.7682, + "step": 6092 + }, + { + "epoch": 0.7326399326639812, + "grad_norm": 2.0580211907232795, + "learning_rate": 7.039790423199192e-07, + "loss": 0.6971, + "step": 6093 + }, + { + "epoch": 0.7327601755546204, + "grad_norm": 2.225125354278864, + "learning_rate": 7.033858513560322e-07, + "loss": 0.7854, + "step": 6094 + }, + { + "epoch": 0.7328804184452594, + "grad_norm": 2.2390618951853902, + "learning_rate": 7.027928570874794e-07, + "loss": 0.7726, + "step": 6095 + }, + { + "epoch": 0.7330006613358985, + "grad_norm": 1.9007245639763346, + "learning_rate": 7.022000596042194e-07, + "loss": 0.8557, + "step": 6096 + }, + { + "epoch": 0.7331209042265376, + "grad_norm": 2.3187537303802013, + "learning_rate": 7.016074589961784e-07, + "loss": 0.8271, + "step": 6097 + }, + { + "epoch": 0.7332411471171767, + "grad_norm": 1.6419196697151432, + "learning_rate": 7.01015055353253e-07, + "loss": 0.6722, + "step": 6098 + }, + { + "epoch": 0.7333613900078157, + "grad_norm": 1.8059677273051895, + "learning_rate": 7.004228487653123e-07, + "loss": 0.7829, + "step": 6099 + }, + { + "epoch": 0.7334816328984549, + "grad_norm": 1.742659705791469, + "learning_rate": 6.998308393221906e-07, + "loss": 0.7854, + "step": 6100 + }, + { + "epoch": 0.733601875789094, + "grad_norm": 2.3382589485081247, + "learning_rate": 6.992390271136977e-07, + "loss": 0.7195, + "step": 6101 + }, + { + "epoch": 0.733722118679733, + "grad_norm": 1.7129016906311254, + "learning_rate": 6.986474122296094e-07, + "loss": 0.8594, + "step": 6102 + }, + { + "epoch": 0.7338423615703722, + "grad_norm": 2.080359956516068, + "learning_rate": 6.980559947596751e-07, + "loss": 0.722, + "step": 6103 + }, + { + "epoch": 0.7339626044610112, + "grad_norm": 1.9287559330194464, + "learning_rate": 6.974647747936109e-07, + "loss": 0.7625, + "step": 6104 + }, + { + "epoch": 0.7340828473516503, + "grad_norm": 2.013798914101596, + "learning_rate": 6.968737524211039e-07, + "loss": 0.8264, + "step": 6105 + }, + { + "epoch": 0.7342030902422895, + "grad_norm": 1.992631271900912, + "learning_rate": 6.962829277318132e-07, + "loss": 0.8017, + "step": 6106 + }, + { + "epoch": 0.7343233331329285, + "grad_norm": 1.7774550853693523, + "learning_rate": 6.956923008153652e-07, + "loss": 0.8423, + "step": 6107 + }, + { + "epoch": 0.7344435760235676, + "grad_norm": 2.225568368732762, + "learning_rate": 6.951018717613593e-07, + "loss": 0.8487, + "step": 6108 + }, + { + "epoch": 0.7345638189142067, + "grad_norm": 2.0478490011756754, + "learning_rate": 6.945116406593614e-07, + "loss": 0.7826, + "step": 6109 + }, + { + "epoch": 0.7346840618048458, + "grad_norm": 2.154518980610586, + "learning_rate": 6.939216075989089e-07, + "loss": 0.7479, + "step": 6110 + }, + { + "epoch": 0.7348043046954849, + "grad_norm": 1.657530020072275, + "learning_rate": 6.933317726695109e-07, + "loss": 0.6643, + "step": 6111 + }, + { + "epoch": 0.734924547586124, + "grad_norm": 2.594877250869975, + "learning_rate": 6.92742135960644e-07, + "loss": 0.8045, + "step": 6112 + }, + { + "epoch": 0.7350447904767631, + "grad_norm": 0.8271276838453062, + "learning_rate": 6.921526975617556e-07, + "loss": 0.5573, + "step": 6113 + }, + { + "epoch": 0.7351650333674021, + "grad_norm": 1.7981190894238253, + "learning_rate": 6.915634575622631e-07, + "loss": 0.7583, + "step": 6114 + }, + { + "epoch": 0.7352852762580413, + "grad_norm": 1.9611213263562897, + "learning_rate": 6.909744160515532e-07, + "loss": 0.7116, + "step": 6115 + }, + { + "epoch": 0.7354055191486804, + "grad_norm": 1.9504939554057743, + "learning_rate": 6.903855731189849e-07, + "loss": 0.6946, + "step": 6116 + }, + { + "epoch": 0.7355257620393194, + "grad_norm": 2.229874566831343, + "learning_rate": 6.897969288538825e-07, + "loss": 0.8268, + "step": 6117 + }, + { + "epoch": 0.7356460049299585, + "grad_norm": 1.7800560680945707, + "learning_rate": 6.892084833455452e-07, + "loss": 0.8174, + "step": 6118 + }, + { + "epoch": 0.7357662478205976, + "grad_norm": 1.5882784098962401, + "learning_rate": 6.886202366832384e-07, + "loss": 0.8408, + "step": 6119 + }, + { + "epoch": 0.7358864907112367, + "grad_norm": 5.910515430778496, + "learning_rate": 6.880321889561987e-07, + "loss": 0.7397, + "step": 6120 + }, + { + "epoch": 0.7360067336018757, + "grad_norm": 2.0651630202409925, + "learning_rate": 6.874443402536338e-07, + "loss": 0.6563, + "step": 6121 + }, + { + "epoch": 0.7361269764925149, + "grad_norm": 1.7171983824394634, + "learning_rate": 6.868566906647177e-07, + "loss": 0.81, + "step": 6122 + }, + { + "epoch": 0.736247219383154, + "grad_norm": 1.7333224427663578, + "learning_rate": 6.862692402785984e-07, + "loss": 0.8373, + "step": 6123 + }, + { + "epoch": 0.736367462273793, + "grad_norm": 0.6956786145492838, + "learning_rate": 6.856819891843899e-07, + "loss": 0.4967, + "step": 6124 + }, + { + "epoch": 0.7364877051644322, + "grad_norm": 2.157498758256009, + "learning_rate": 6.8509493747118e-07, + "loss": 0.7293, + "step": 6125 + }, + { + "epoch": 0.7366079480550712, + "grad_norm": 2.0854974060993623, + "learning_rate": 6.845080852280221e-07, + "loss": 0.8845, + "step": 6126 + }, + { + "epoch": 0.7367281909457103, + "grad_norm": 1.670895583720532, + "learning_rate": 6.839214325439409e-07, + "loss": 0.7485, + "step": 6127 + }, + { + "epoch": 0.7368484338363495, + "grad_norm": 1.7277150859614172, + "learning_rate": 6.833349795079327e-07, + "loss": 0.7188, + "step": 6128 + }, + { + "epoch": 0.7369686767269885, + "grad_norm": 1.676276856163904, + "learning_rate": 6.827487262089613e-07, + "loss": 0.6894, + "step": 6129 + }, + { + "epoch": 0.7370889196176276, + "grad_norm": 0.8976758299070803, + "learning_rate": 6.821626727359606e-07, + "loss": 0.568, + "step": 6130 + }, + { + "epoch": 0.7372091625082667, + "grad_norm": 2.091477256598351, + "learning_rate": 6.815768191778348e-07, + "loss": 0.7797, + "step": 6131 + }, + { + "epoch": 0.7373294053989058, + "grad_norm": 2.002504687738349, + "learning_rate": 6.809911656234569e-07, + "loss": 0.7278, + "step": 6132 + }, + { + "epoch": 0.7374496482895448, + "grad_norm": 2.391467575488285, + "learning_rate": 6.804057121616707e-07, + "loss": 0.7873, + "step": 6133 + }, + { + "epoch": 0.737569891180184, + "grad_norm": 3.135297669510892, + "learning_rate": 6.798204588812888e-07, + "loss": 0.7244, + "step": 6134 + }, + { + "epoch": 0.7376901340708231, + "grad_norm": 1.8154849105077608, + "learning_rate": 6.792354058710937e-07, + "loss": 0.7593, + "step": 6135 + }, + { + "epoch": 0.7378103769614621, + "grad_norm": 2.049208497557333, + "learning_rate": 6.786505532198374e-07, + "loss": 0.6538, + "step": 6136 + }, + { + "epoch": 0.7379306198521013, + "grad_norm": 2.567165593374797, + "learning_rate": 6.780659010162411e-07, + "loss": 0.8539, + "step": 6137 + }, + { + "epoch": 0.7380508627427403, + "grad_norm": 1.6816811728379542, + "learning_rate": 6.774814493489975e-07, + "loss": 0.8365, + "step": 6138 + }, + { + "epoch": 0.7381711056333794, + "grad_norm": 1.742385921383562, + "learning_rate": 6.768971983067655e-07, + "loss": 0.669, + "step": 6139 + }, + { + "epoch": 0.7382913485240186, + "grad_norm": 1.0535819016183143, + "learning_rate": 6.763131479781772e-07, + "loss": 0.6782, + "step": 6140 + }, + { + "epoch": 0.7384115914146576, + "grad_norm": 2.213861372056213, + "learning_rate": 6.757292984518316e-07, + "loss": 0.7679, + "step": 6141 + }, + { + "epoch": 0.7385318343052967, + "grad_norm": 0.7520130084084459, + "learning_rate": 6.751456498162981e-07, + "loss": 0.5653, + "step": 6142 + }, + { + "epoch": 0.7386520771959358, + "grad_norm": 2.0029975362563537, + "learning_rate": 6.745622021601174e-07, + "loss": 0.859, + "step": 6143 + }, + { + "epoch": 0.7387723200865749, + "grad_norm": 2.021601370406658, + "learning_rate": 6.739789555717954e-07, + "loss": 0.6992, + "step": 6144 + }, + { + "epoch": 0.738892562977214, + "grad_norm": 3.3712977129961867, + "learning_rate": 6.733959101398124e-07, + "loss": 0.7804, + "step": 6145 + }, + { + "epoch": 0.7390128058678531, + "grad_norm": 1.735524664270297, + "learning_rate": 6.728130659526143e-07, + "loss": 0.8146, + "step": 6146 + }, + { + "epoch": 0.7391330487584922, + "grad_norm": 2.997835332011491, + "learning_rate": 6.7223042309862e-07, + "loss": 0.7139, + "step": 6147 + }, + { + "epoch": 0.7392532916491312, + "grad_norm": 2.066073471993386, + "learning_rate": 6.716479816662144e-07, + "loss": 0.7381, + "step": 6148 + }, + { + "epoch": 0.7393735345397703, + "grad_norm": 1.9350447325747873, + "learning_rate": 6.710657417437531e-07, + "loss": 0.7288, + "step": 6149 + }, + { + "epoch": 0.7394937774304094, + "grad_norm": 2.919500191645576, + "learning_rate": 6.704837034195628e-07, + "loss": 0.8037, + "step": 6150 + }, + { + "epoch": 0.7396140203210485, + "grad_norm": 2.10525619198249, + "learning_rate": 6.699018667819376e-07, + "loss": 0.8532, + "step": 6151 + }, + { + "epoch": 0.7397342632116876, + "grad_norm": 1.613562238482881, + "learning_rate": 6.693202319191415e-07, + "loss": 0.7324, + "step": 6152 + }, + { + "epoch": 0.7398545061023267, + "grad_norm": 2.2304258860195563, + "learning_rate": 6.687387989194084e-07, + "loss": 0.7542, + "step": 6153 + }, + { + "epoch": 0.7399747489929658, + "grad_norm": 1.9207053618372487, + "learning_rate": 6.681575678709404e-07, + "loss": 0.8017, + "step": 6154 + }, + { + "epoch": 0.7400949918836048, + "grad_norm": 1.8690682360304853, + "learning_rate": 6.67576538861911e-07, + "loss": 0.7106, + "step": 6155 + }, + { + "epoch": 0.740215234774244, + "grad_norm": 1.6054552360764114, + "learning_rate": 6.669957119804612e-07, + "loss": 0.8223, + "step": 6156 + }, + { + "epoch": 0.7403354776648831, + "grad_norm": 2.7958027579790814, + "learning_rate": 6.66415087314702e-07, + "loss": 0.7296, + "step": 6157 + }, + { + "epoch": 0.7404557205555221, + "grad_norm": 2.0796073273324405, + "learning_rate": 6.65834664952714e-07, + "loss": 0.7334, + "step": 6158 + }, + { + "epoch": 0.7405759634461613, + "grad_norm": 1.7837671263313002, + "learning_rate": 6.652544449825457e-07, + "loss": 0.7627, + "step": 6159 + }, + { + "epoch": 0.7406962063368003, + "grad_norm": 1.6281221913260397, + "learning_rate": 6.646744274922182e-07, + "loss": 0.7663, + "step": 6160 + }, + { + "epoch": 0.7408164492274394, + "grad_norm": 2.5316136416854413, + "learning_rate": 6.640946125697171e-07, + "loss": 0.7624, + "step": 6161 + }, + { + "epoch": 0.7409366921180786, + "grad_norm": 2.037747481592499, + "learning_rate": 6.635150003030017e-07, + "loss": 0.7628, + "step": 6162 + }, + { + "epoch": 0.7410569350087176, + "grad_norm": 2.142956987961307, + "learning_rate": 6.629355907799981e-07, + "loss": 0.8656, + "step": 6163 + }, + { + "epoch": 0.7411771778993567, + "grad_norm": 2.47372456834856, + "learning_rate": 6.623563840886015e-07, + "loss": 0.6934, + "step": 6164 + }, + { + "epoch": 0.7412974207899958, + "grad_norm": 4.328997913017307, + "learning_rate": 6.617773803166795e-07, + "loss": 0.6998, + "step": 6165 + }, + { + "epoch": 0.7414176636806349, + "grad_norm": 2.437305931777569, + "learning_rate": 6.611985795520634e-07, + "loss": 0.8255, + "step": 6166 + }, + { + "epoch": 0.7415379065712739, + "grad_norm": 1.9137683875996416, + "learning_rate": 6.606199818825588e-07, + "loss": 0.7817, + "step": 6167 + }, + { + "epoch": 0.7416581494619131, + "grad_norm": 3.840625640645606, + "learning_rate": 6.600415873959377e-07, + "loss": 0.8191, + "step": 6168 + }, + { + "epoch": 0.7417783923525522, + "grad_norm": 2.1805385445851186, + "learning_rate": 6.594633961799437e-07, + "loss": 0.6459, + "step": 6169 + }, + { + "epoch": 0.7418986352431912, + "grad_norm": 1.6669186004325884, + "learning_rate": 6.588854083222857e-07, + "loss": 0.8205, + "step": 6170 + }, + { + "epoch": 0.7420188781338304, + "grad_norm": 2.249997668794908, + "learning_rate": 6.583076239106444e-07, + "loss": 0.8135, + "step": 6171 + }, + { + "epoch": 0.7421391210244694, + "grad_norm": 2.4374122603837645, + "learning_rate": 6.577300430326707e-07, + "loss": 0.7631, + "step": 6172 + }, + { + "epoch": 0.7422593639151085, + "grad_norm": 2.2004328301944946, + "learning_rate": 6.571526657759821e-07, + "loss": 0.7289, + "step": 6173 + }, + { + "epoch": 0.7423796068057477, + "grad_norm": 1.6699208755936337, + "learning_rate": 6.565754922281663e-07, + "loss": 0.7066, + "step": 6174 + }, + { + "epoch": 0.7424998496963867, + "grad_norm": 1.8577816330997998, + "learning_rate": 6.559985224767801e-07, + "loss": 0.7906, + "step": 6175 + }, + { + "epoch": 0.7426200925870258, + "grad_norm": 2.413697545386888, + "learning_rate": 6.55421756609349e-07, + "loss": 0.7606, + "step": 6176 + }, + { + "epoch": 0.7427403354776649, + "grad_norm": 1.8608804586718284, + "learning_rate": 6.54845194713369e-07, + "loss": 0.7881, + "step": 6177 + }, + { + "epoch": 0.742860578368304, + "grad_norm": 2.0243371561534524, + "learning_rate": 6.542688368763034e-07, + "loss": 0.7968, + "step": 6178 + }, + { + "epoch": 0.742980821258943, + "grad_norm": 1.6863576059810814, + "learning_rate": 6.536926831855854e-07, + "loss": 0.774, + "step": 6179 + }, + { + "epoch": 0.7431010641495821, + "grad_norm": 2.3093944006738583, + "learning_rate": 6.531167337286165e-07, + "loss": 0.7399, + "step": 6180 + }, + { + "epoch": 0.7432213070402213, + "grad_norm": 1.477051669033824, + "learning_rate": 6.52540988592768e-07, + "loss": 0.7992, + "step": 6181 + }, + { + "epoch": 0.7433415499308603, + "grad_norm": 2.2401682836466543, + "learning_rate": 6.519654478653814e-07, + "loss": 0.8367, + "step": 6182 + }, + { + "epoch": 0.7434617928214994, + "grad_norm": 0.8824834970114505, + "learning_rate": 6.51390111633763e-07, + "loss": 0.562, + "step": 6183 + }, + { + "epoch": 0.7435820357121385, + "grad_norm": 1.661979266766105, + "learning_rate": 6.508149799851932e-07, + "loss": 0.764, + "step": 6184 + }, + { + "epoch": 0.7437022786027776, + "grad_norm": 1.8750345226924363, + "learning_rate": 6.502400530069183e-07, + "loss": 0.6113, + "step": 6185 + }, + { + "epoch": 0.7438225214934167, + "grad_norm": 2.0808227925408542, + "learning_rate": 6.496653307861535e-07, + "loss": 0.6899, + "step": 6186 + }, + { + "epoch": 0.7439427643840558, + "grad_norm": 1.895479951837527, + "learning_rate": 6.490908134100857e-07, + "loss": 0.6624, + "step": 6187 + }, + { + "epoch": 0.7440630072746949, + "grad_norm": 2.027550482821127, + "learning_rate": 6.48516500965866e-07, + "loss": 0.7008, + "step": 6188 + }, + { + "epoch": 0.7441832501653339, + "grad_norm": 1.5365176000856369, + "learning_rate": 6.479423935406192e-07, + "loss": 0.8202, + "step": 6189 + }, + { + "epoch": 0.7443034930559731, + "grad_norm": 0.8179178070848093, + "learning_rate": 6.473684912214357e-07, + "loss": 0.6207, + "step": 6190 + }, + { + "epoch": 0.7444237359466122, + "grad_norm": 1.9189174815093757, + "learning_rate": 6.467947940953778e-07, + "loss": 0.6976, + "step": 6191 + }, + { + "epoch": 0.7445439788372512, + "grad_norm": 2.3578083824490843, + "learning_rate": 6.462213022494732e-07, + "loss": 0.725, + "step": 6192 + }, + { + "epoch": 0.7446642217278904, + "grad_norm": 0.7881247736592463, + "learning_rate": 6.456480157707201e-07, + "loss": 0.6115, + "step": 6193 + }, + { + "epoch": 0.7447844646185294, + "grad_norm": 2.1997325561261865, + "learning_rate": 6.450749347460866e-07, + "loss": 0.8555, + "step": 6194 + }, + { + "epoch": 0.7449047075091685, + "grad_norm": 1.7914760917028574, + "learning_rate": 6.445020592625083e-07, + "loss": 0.7929, + "step": 6195 + }, + { + "epoch": 0.7450249503998077, + "grad_norm": 2.1929297497756677, + "learning_rate": 6.4392938940689e-07, + "loss": 0.8058, + "step": 6196 + }, + { + "epoch": 0.7451451932904467, + "grad_norm": 2.826211608733948, + "learning_rate": 6.433569252661049e-07, + "loss": 0.7139, + "step": 6197 + }, + { + "epoch": 0.7452654361810858, + "grad_norm": 2.259805294865716, + "learning_rate": 6.427846669269952e-07, + "loss": 0.7164, + "step": 6198 + }, + { + "epoch": 0.7453856790717249, + "grad_norm": 2.0971562022188825, + "learning_rate": 6.422126144763729e-07, + "loss": 0.8292, + "step": 6199 + }, + { + "epoch": 0.745505921962364, + "grad_norm": 2.1909455547802255, + "learning_rate": 6.416407680010174e-07, + "loss": 0.77, + "step": 6200 + }, + { + "epoch": 0.745626164853003, + "grad_norm": 2.7005125442323714, + "learning_rate": 6.410691275876774e-07, + "loss": 0.8118, + "step": 6201 + }, + { + "epoch": 0.7457464077436422, + "grad_norm": 2.202604516414874, + "learning_rate": 6.404976933230704e-07, + "loss": 0.7732, + "step": 6202 + }, + { + "epoch": 0.7458666506342813, + "grad_norm": 1.8953275599750112, + "learning_rate": 6.399264652938813e-07, + "loss": 0.731, + "step": 6203 + }, + { + "epoch": 0.7459868935249203, + "grad_norm": 1.9441659758242695, + "learning_rate": 6.393554435867679e-07, + "loss": 0.7462, + "step": 6204 + }, + { + "epoch": 0.7461071364155595, + "grad_norm": 2.0697240605005756, + "learning_rate": 6.387846282883502e-07, + "loss": 0.8388, + "step": 6205 + }, + { + "epoch": 0.7462273793061985, + "grad_norm": 1.9826698486381262, + "learning_rate": 6.38214019485223e-07, + "loss": 0.7751, + "step": 6206 + }, + { + "epoch": 0.7463476221968376, + "grad_norm": 1.9941842039840643, + "learning_rate": 6.376436172639461e-07, + "loss": 0.7182, + "step": 6207 + }, + { + "epoch": 0.7464678650874768, + "grad_norm": 2.853295371911243, + "learning_rate": 6.370734217110487e-07, + "loss": 0.656, + "step": 6208 + }, + { + "epoch": 0.7465881079781158, + "grad_norm": 2.1539245449724023, + "learning_rate": 6.36503432913031e-07, + "loss": 0.6432, + "step": 6209 + }, + { + "epoch": 0.7467083508687549, + "grad_norm": 2.0145213336207886, + "learning_rate": 6.359336509563569e-07, + "loss": 0.7008, + "step": 6210 + }, + { + "epoch": 0.7468285937593939, + "grad_norm": 4.211145852336207, + "learning_rate": 6.353640759274641e-07, + "loss": 0.8096, + "step": 6211 + }, + { + "epoch": 0.7469488366500331, + "grad_norm": 5.1258255828582175, + "learning_rate": 6.347947079127556e-07, + "loss": 0.7487, + "step": 6212 + }, + { + "epoch": 0.7470690795406721, + "grad_norm": 2.0268189443717364, + "learning_rate": 6.342255469986053e-07, + "loss": 0.7715, + "step": 6213 + }, + { + "epoch": 0.7471893224313112, + "grad_norm": 1.9405671729964353, + "learning_rate": 6.336565932713533e-07, + "loss": 0.7707, + "step": 6214 + }, + { + "epoch": 0.7473095653219504, + "grad_norm": 1.8547419680637314, + "learning_rate": 6.330878468173088e-07, + "loss": 0.7779, + "step": 6215 + }, + { + "epoch": 0.7474298082125894, + "grad_norm": 1.7823624065309283, + "learning_rate": 6.32519307722752e-07, + "loss": 0.7358, + "step": 6216 + }, + { + "epoch": 0.7475500511032285, + "grad_norm": 0.8551410064649018, + "learning_rate": 6.31950976073929e-07, + "loss": 0.5499, + "step": 6217 + }, + { + "epoch": 0.7476702939938676, + "grad_norm": 2.512128212214108, + "learning_rate": 6.31382851957055e-07, + "loss": 0.8136, + "step": 6218 + }, + { + "epoch": 0.7477905368845067, + "grad_norm": 1.8959149842617462, + "learning_rate": 6.308149354583143e-07, + "loss": 0.7223, + "step": 6219 + }, + { + "epoch": 0.7479107797751458, + "grad_norm": 1.9006886890482042, + "learning_rate": 6.302472266638586e-07, + "loss": 0.816, + "step": 6220 + }, + { + "epoch": 0.7480310226657849, + "grad_norm": 3.060044374580515, + "learning_rate": 6.296797256598101e-07, + "loss": 0.7051, + "step": 6221 + }, + { + "epoch": 0.748151265556424, + "grad_norm": 1.7776653962907174, + "learning_rate": 6.291124325322576e-07, + "loss": 0.8137, + "step": 6222 + }, + { + "epoch": 0.748271508447063, + "grad_norm": 1.6007888041625686, + "learning_rate": 6.285453473672595e-07, + "loss": 0.63, + "step": 6223 + }, + { + "epoch": 0.7483917513377022, + "grad_norm": 2.072430718233703, + "learning_rate": 6.279784702508415e-07, + "loss": 0.7583, + "step": 6224 + }, + { + "epoch": 0.7485119942283412, + "grad_norm": 0.8024109051763954, + "learning_rate": 6.274118012689979e-07, + "loss": 0.5865, + "step": 6225 + }, + { + "epoch": 0.7486322371189803, + "grad_norm": 1.6289165688606386, + "learning_rate": 6.268453405076943e-07, + "loss": 0.6865, + "step": 6226 + }, + { + "epoch": 0.7487524800096195, + "grad_norm": 2.0792013263552853, + "learning_rate": 6.262790880528592e-07, + "loss": 0.8276, + "step": 6227 + }, + { + "epoch": 0.7488727229002585, + "grad_norm": 2.4690539981689392, + "learning_rate": 6.257130439903951e-07, + "loss": 0.8058, + "step": 6228 + }, + { + "epoch": 0.7489929657908976, + "grad_norm": 1.7815263767993135, + "learning_rate": 6.251472084061695e-07, + "loss": 0.8114, + "step": 6229 + }, + { + "epoch": 0.7491132086815367, + "grad_norm": 2.102644517719414, + "learning_rate": 6.245815813860191e-07, + "loss": 0.8915, + "step": 6230 + }, + { + "epoch": 0.7492334515721758, + "grad_norm": 3.521531225856533, + "learning_rate": 6.240161630157495e-07, + "loss": 0.7124, + "step": 6231 + }, + { + "epoch": 0.7493536944628149, + "grad_norm": 2.687815625486407, + "learning_rate": 6.23450953381133e-07, + "loss": 0.7113, + "step": 6232 + }, + { + "epoch": 0.749473937353454, + "grad_norm": 2.9300737050130805, + "learning_rate": 6.228859525679131e-07, + "loss": 0.6861, + "step": 6233 + }, + { + "epoch": 0.7495941802440931, + "grad_norm": 6.727969918454809, + "learning_rate": 6.223211606617986e-07, + "loss": 0.8042, + "step": 6234 + }, + { + "epoch": 0.7497144231347321, + "grad_norm": 1.8826742794222455, + "learning_rate": 6.217565777484701e-07, + "loss": 0.8415, + "step": 6235 + }, + { + "epoch": 0.7498346660253713, + "grad_norm": 1.9238064992241057, + "learning_rate": 6.211922039135722e-07, + "loss": 0.8061, + "step": 6236 + }, + { + "epoch": 0.7499549089160104, + "grad_norm": 1.854894159593004, + "learning_rate": 6.206280392427201e-07, + "loss": 0.8184, + "step": 6237 + }, + { + "epoch": 0.7500751518066494, + "grad_norm": 1.5497289205201525, + "learning_rate": 6.200640838214983e-07, + "loss": 0.7412, + "step": 6238 + }, + { + "epoch": 0.7501953946972886, + "grad_norm": 1.770220152962228, + "learning_rate": 6.195003377354578e-07, + "loss": 0.6725, + "step": 6239 + }, + { + "epoch": 0.7503156375879276, + "grad_norm": 2.483760445836318, + "learning_rate": 6.189368010701183e-07, + "loss": 0.733, + "step": 6240 + }, + { + "epoch": 0.7504358804785667, + "grad_norm": 2.1194873438279767, + "learning_rate": 6.183734739109683e-07, + "loss": 0.7698, + "step": 6241 + }, + { + "epoch": 0.7505561233692057, + "grad_norm": 2.1813012715519537, + "learning_rate": 6.178103563434629e-07, + "loss": 0.6868, + "step": 6242 + }, + { + "epoch": 0.7506763662598449, + "grad_norm": 1.7887397041797564, + "learning_rate": 6.172474484530283e-07, + "loss": 0.8427, + "step": 6243 + }, + { + "epoch": 0.750796609150484, + "grad_norm": 2.231503649059665, + "learning_rate": 6.166847503250563e-07, + "loss": 0.7607, + "step": 6244 + }, + { + "epoch": 0.750916852041123, + "grad_norm": 2.498771556400597, + "learning_rate": 6.161222620449078e-07, + "loss": 0.7934, + "step": 6245 + }, + { + "epoch": 0.7510370949317622, + "grad_norm": 2.103667385680059, + "learning_rate": 6.155599836979117e-07, + "loss": 0.8065, + "step": 6246 + }, + { + "epoch": 0.7511573378224012, + "grad_norm": 2.7605007134844124, + "learning_rate": 6.149979153693649e-07, + "loss": 0.8237, + "step": 6247 + }, + { + "epoch": 0.7512775807130403, + "grad_norm": 2.0133294333279124, + "learning_rate": 6.144360571445343e-07, + "loss": 0.7692, + "step": 6248 + }, + { + "epoch": 0.7513978236036795, + "grad_norm": 1.815429950157996, + "learning_rate": 6.138744091086509e-07, + "loss": 0.8035, + "step": 6249 + }, + { + "epoch": 0.7515180664943185, + "grad_norm": 2.426487621716774, + "learning_rate": 6.133129713469183e-07, + "loss": 0.735, + "step": 6250 + }, + { + "epoch": 0.7516383093849576, + "grad_norm": 2.0490528198032147, + "learning_rate": 6.127517439445053e-07, + "loss": 0.6429, + "step": 6251 + }, + { + "epoch": 0.7517585522755967, + "grad_norm": 1.9462439823241355, + "learning_rate": 6.121907269865498e-07, + "loss": 0.8219, + "step": 6252 + }, + { + "epoch": 0.7518787951662358, + "grad_norm": 0.938184297683879, + "learning_rate": 6.116299205581577e-07, + "loss": 0.673, + "step": 6253 + }, + { + "epoch": 0.7519990380568748, + "grad_norm": 2.411356770715916, + "learning_rate": 6.110693247444018e-07, + "loss": 0.687, + "step": 6254 + }, + { + "epoch": 0.752119280947514, + "grad_norm": 1.7257990465330204, + "learning_rate": 6.105089396303258e-07, + "loss": 0.8243, + "step": 6255 + }, + { + "epoch": 0.7522395238381531, + "grad_norm": 1.803634534538438, + "learning_rate": 6.099487653009383e-07, + "loss": 0.7587, + "step": 6256 + }, + { + "epoch": 0.7523597667287921, + "grad_norm": 2.0832306518681643, + "learning_rate": 6.093888018412192e-07, + "loss": 0.8333, + "step": 6257 + }, + { + "epoch": 0.7524800096194313, + "grad_norm": 0.7189895811082084, + "learning_rate": 6.088290493361125e-07, + "loss": 0.547, + "step": 6258 + }, + { + "epoch": 0.7526002525100703, + "grad_norm": 2.1221068265453233, + "learning_rate": 6.082695078705322e-07, + "loss": 0.7218, + "step": 6259 + }, + { + "epoch": 0.7527204954007094, + "grad_norm": 2.2147300731409687, + "learning_rate": 6.077101775293618e-07, + "loss": 0.6922, + "step": 6260 + }, + { + "epoch": 0.7528407382913486, + "grad_norm": 2.893066035403457, + "learning_rate": 6.071510583974504e-07, + "loss": 0.8339, + "step": 6261 + }, + { + "epoch": 0.7529609811819876, + "grad_norm": 1.951862018406451, + "learning_rate": 6.065921505596161e-07, + "loss": 0.7295, + "step": 6262 + }, + { + "epoch": 0.7530812240726267, + "grad_norm": 1.8846463338625248, + "learning_rate": 6.060334541006445e-07, + "loss": 0.7734, + "step": 6263 + }, + { + "epoch": 0.7532014669632658, + "grad_norm": 1.6471994098601135, + "learning_rate": 6.05474969105289e-07, + "loss": 0.692, + "step": 6264 + }, + { + "epoch": 0.7533217098539049, + "grad_norm": 2.7559654383049415, + "learning_rate": 6.049166956582725e-07, + "loss": 0.7407, + "step": 6265 + }, + { + "epoch": 0.753441952744544, + "grad_norm": 1.9724270224759979, + "learning_rate": 6.043586338442841e-07, + "loss": 0.8786, + "step": 6266 + }, + { + "epoch": 0.7535621956351831, + "grad_norm": 2.710760457087049, + "learning_rate": 6.038007837479815e-07, + "loss": 0.7349, + "step": 6267 + }, + { + "epoch": 0.7536824385258222, + "grad_norm": 1.860758419052158, + "learning_rate": 6.032431454539897e-07, + "loss": 0.641, + "step": 6268 + }, + { + "epoch": 0.7538026814164612, + "grad_norm": 1.6181289217810404, + "learning_rate": 6.026857190469014e-07, + "loss": 0.8178, + "step": 6269 + }, + { + "epoch": 0.7539229243071004, + "grad_norm": 1.779929089580348, + "learning_rate": 6.0212850461128e-07, + "loss": 0.7393, + "step": 6270 + }, + { + "epoch": 0.7540431671977395, + "grad_norm": 2.2197871403414546, + "learning_rate": 6.015715022316516e-07, + "loss": 0.7482, + "step": 6271 + }, + { + "epoch": 0.7541634100883785, + "grad_norm": 2.9246520854388502, + "learning_rate": 6.010147119925154e-07, + "loss": 0.7824, + "step": 6272 + }, + { + "epoch": 0.7542836529790176, + "grad_norm": 1.9391147898494536, + "learning_rate": 6.004581339783348e-07, + "loss": 0.6708, + "step": 6273 + }, + { + "epoch": 0.7544038958696567, + "grad_norm": 3.0528751079704723, + "learning_rate": 5.999017682735425e-07, + "loss": 0.688, + "step": 6274 + }, + { + "epoch": 0.7545241387602958, + "grad_norm": 2.0170735912827142, + "learning_rate": 5.993456149625387e-07, + "loss": 0.6634, + "step": 6275 + }, + { + "epoch": 0.7546443816509348, + "grad_norm": 1.9323525500581058, + "learning_rate": 5.987896741296909e-07, + "loss": 0.8295, + "step": 6276 + }, + { + "epoch": 0.754764624541574, + "grad_norm": 1.8442331747802947, + "learning_rate": 5.982339458593361e-07, + "loss": 0.7867, + "step": 6277 + }, + { + "epoch": 0.7548848674322131, + "grad_norm": 1.6073376991600847, + "learning_rate": 5.976784302357767e-07, + "loss": 0.8409, + "step": 6278 + }, + { + "epoch": 0.7550051103228521, + "grad_norm": 2.0333702172649732, + "learning_rate": 5.971231273432855e-07, + "loss": 0.7361, + "step": 6279 + }, + { + "epoch": 0.7551253532134913, + "grad_norm": 2.090312812429899, + "learning_rate": 5.965680372661e-07, + "loss": 0.5458, + "step": 6280 + }, + { + "epoch": 0.7552455961041303, + "grad_norm": 1.7873507644138178, + "learning_rate": 5.960131600884266e-07, + "loss": 0.5599, + "step": 6281 + }, + { + "epoch": 0.7553658389947694, + "grad_norm": 1.742613188735647, + "learning_rate": 5.954584958944413e-07, + "loss": 0.7635, + "step": 6282 + }, + { + "epoch": 0.7554860818854086, + "grad_norm": 1.9962863418624757, + "learning_rate": 5.949040447682854e-07, + "loss": 0.8168, + "step": 6283 + }, + { + "epoch": 0.7556063247760476, + "grad_norm": 2.9056361380741267, + "learning_rate": 5.943498067940686e-07, + "loss": 0.6874, + "step": 6284 + }, + { + "epoch": 0.7557265676666867, + "grad_norm": 1.786783824514559, + "learning_rate": 5.937957820558686e-07, + "loss": 0.8216, + "step": 6285 + }, + { + "epoch": 0.7558468105573258, + "grad_norm": 0.860961524642019, + "learning_rate": 5.932419706377296e-07, + "loss": 0.6541, + "step": 6286 + }, + { + "epoch": 0.7559670534479649, + "grad_norm": 1.9023753952746836, + "learning_rate": 5.92688372623666e-07, + "loss": 0.7473, + "step": 6287 + }, + { + "epoch": 0.7560872963386039, + "grad_norm": 2.3492144144092433, + "learning_rate": 5.921349880976574e-07, + "loss": 0.7415, + "step": 6288 + }, + { + "epoch": 0.7562075392292431, + "grad_norm": 1.7595610335600051, + "learning_rate": 5.915818171436515e-07, + "loss": 0.822, + "step": 6289 + }, + { + "epoch": 0.7563277821198822, + "grad_norm": 4.011312700056768, + "learning_rate": 5.910288598455642e-07, + "loss": 0.7462, + "step": 6290 + }, + { + "epoch": 0.7564480250105212, + "grad_norm": 2.475758804214726, + "learning_rate": 5.90476116287278e-07, + "loss": 0.7472, + "step": 6291 + }, + { + "epoch": 0.7565682679011604, + "grad_norm": 1.9002980349842855, + "learning_rate": 5.899235865526456e-07, + "loss": 0.6768, + "step": 6292 + }, + { + "epoch": 0.7566885107917994, + "grad_norm": 1.817853781520103, + "learning_rate": 5.893712707254825e-07, + "loss": 0.8248, + "step": 6293 + }, + { + "epoch": 0.7568087536824385, + "grad_norm": 2.7125987478819087, + "learning_rate": 5.888191688895769e-07, + "loss": 0.6633, + "step": 6294 + }, + { + "epoch": 0.7569289965730777, + "grad_norm": 2.1025338462981047, + "learning_rate": 5.882672811286813e-07, + "loss": 0.6289, + "step": 6295 + }, + { + "epoch": 0.7570492394637167, + "grad_norm": 2.2894624224026368, + "learning_rate": 5.877156075265166e-07, + "loss": 0.6982, + "step": 6296 + }, + { + "epoch": 0.7571694823543558, + "grad_norm": 2.589226328351995, + "learning_rate": 5.871641481667715e-07, + "loss": 0.6997, + "step": 6297 + }, + { + "epoch": 0.7572897252449949, + "grad_norm": 1.666984750276321, + "learning_rate": 5.866129031331011e-07, + "loss": 0.8454, + "step": 6298 + }, + { + "epoch": 0.757409968135634, + "grad_norm": 2.2331906894659888, + "learning_rate": 5.8606187250913e-07, + "loss": 0.8357, + "step": 6299 + }, + { + "epoch": 0.757530211026273, + "grad_norm": 1.8702965553260535, + "learning_rate": 5.855110563784482e-07, + "loss": 0.8434, + "step": 6300 + }, + { + "epoch": 0.7576504539169122, + "grad_norm": 1.6373056893412181, + "learning_rate": 5.849604548246156e-07, + "loss": 0.6436, + "step": 6301 + }, + { + "epoch": 0.7577706968075513, + "grad_norm": 1.9808162106020124, + "learning_rate": 5.844100679311565e-07, + "loss": 0.8057, + "step": 6302 + }, + { + "epoch": 0.7578909396981903, + "grad_norm": 2.691538947257136, + "learning_rate": 5.838598957815637e-07, + "loss": 0.7654, + "step": 6303 + }, + { + "epoch": 0.7580111825888295, + "grad_norm": 1.5338644168772488, + "learning_rate": 5.833099384592996e-07, + "loss": 0.8567, + "step": 6304 + }, + { + "epoch": 0.7581314254794685, + "grad_norm": 1.8677775196963062, + "learning_rate": 5.827601960477913e-07, + "loss": 0.7187, + "step": 6305 + }, + { + "epoch": 0.7582516683701076, + "grad_norm": 2.0323198288830766, + "learning_rate": 5.822106686304344e-07, + "loss": 0.7111, + "step": 6306 + }, + { + "epoch": 0.7583719112607467, + "grad_norm": 1.9198623290971042, + "learning_rate": 5.816613562905919e-07, + "loss": 0.5797, + "step": 6307 + }, + { + "epoch": 0.7584921541513858, + "grad_norm": 1.5027345526952716, + "learning_rate": 5.811122591115933e-07, + "loss": 0.6996, + "step": 6308 + }, + { + "epoch": 0.7586123970420249, + "grad_norm": 2.6971514192195887, + "learning_rate": 5.805633771767376e-07, + "loss": 0.717, + "step": 6309 + }, + { + "epoch": 0.7587326399326639, + "grad_norm": 1.776898607116665, + "learning_rate": 5.800147105692888e-07, + "loss": 0.7838, + "step": 6310 + }, + { + "epoch": 0.7588528828233031, + "grad_norm": 1.6869230343685087, + "learning_rate": 5.794662593724795e-07, + "loss": 0.7908, + "step": 6311 + }, + { + "epoch": 0.7589731257139422, + "grad_norm": 2.679801115513477, + "learning_rate": 5.789180236695091e-07, + "loss": 0.7542, + "step": 6312 + }, + { + "epoch": 0.7590933686045812, + "grad_norm": 1.9962655010350097, + "learning_rate": 5.78370003543544e-07, + "loss": 0.8485, + "step": 6313 + }, + { + "epoch": 0.7592136114952204, + "grad_norm": 1.8661233590983959, + "learning_rate": 5.778221990777203e-07, + "loss": 0.8404, + "step": 6314 + }, + { + "epoch": 0.7593338543858594, + "grad_norm": 2.011063255995867, + "learning_rate": 5.772746103551372e-07, + "loss": 0.8298, + "step": 6315 + }, + { + "epoch": 0.7594540972764985, + "grad_norm": 1.6522188612903672, + "learning_rate": 5.767272374588648e-07, + "loss": 0.7194, + "step": 6316 + }, + { + "epoch": 0.7595743401671377, + "grad_norm": 1.6538675555071083, + "learning_rate": 5.76180080471939e-07, + "loss": 0.7815, + "step": 6317 + }, + { + "epoch": 0.7596945830577767, + "grad_norm": 2.281267035433771, + "learning_rate": 5.756331394773631e-07, + "loss": 0.7248, + "step": 6318 + }, + { + "epoch": 0.7598148259484158, + "grad_norm": 2.063491814011313, + "learning_rate": 5.750864145581071e-07, + "loss": 0.7654, + "step": 6319 + }, + { + "epoch": 0.7599350688390549, + "grad_norm": 4.044095649863408, + "learning_rate": 5.745399057971085e-07, + "loss": 0.8641, + "step": 6320 + }, + { + "epoch": 0.760055311729694, + "grad_norm": 4.371988404379891, + "learning_rate": 5.739936132772738e-07, + "loss": 0.7561, + "step": 6321 + }, + { + "epoch": 0.760175554620333, + "grad_norm": 2.71658335325899, + "learning_rate": 5.734475370814733e-07, + "loss": 0.7521, + "step": 6322 + }, + { + "epoch": 0.7602957975109722, + "grad_norm": 1.7648136671659345, + "learning_rate": 5.729016772925483e-07, + "loss": 0.7877, + "step": 6323 + }, + { + "epoch": 0.7604160404016113, + "grad_norm": 3.3160646080653016, + "learning_rate": 5.723560339933038e-07, + "loss": 0.7127, + "step": 6324 + }, + { + "epoch": 0.7605362832922503, + "grad_norm": 2.325744640919468, + "learning_rate": 5.71810607266513e-07, + "loss": 0.6543, + "step": 6325 + }, + { + "epoch": 0.7606565261828895, + "grad_norm": 1.7977950684531028, + "learning_rate": 5.712653971949184e-07, + "loss": 0.6048, + "step": 6326 + }, + { + "epoch": 0.7607767690735285, + "grad_norm": 2.8590024278062702, + "learning_rate": 5.707204038612268e-07, + "loss": 0.7617, + "step": 6327 + }, + { + "epoch": 0.7608970119641676, + "grad_norm": 2.2679623878697255, + "learning_rate": 5.701756273481138e-07, + "loss": 0.7449, + "step": 6328 + }, + { + "epoch": 0.7610172548548068, + "grad_norm": 1.7122784937702953, + "learning_rate": 5.696310677382212e-07, + "loss": 0.7411, + "step": 6329 + }, + { + "epoch": 0.7611374977454458, + "grad_norm": 0.8894334820239862, + "learning_rate": 5.690867251141576e-07, + "loss": 0.6179, + "step": 6330 + }, + { + "epoch": 0.7612577406360849, + "grad_norm": 2.736355396818281, + "learning_rate": 5.685425995585013e-07, + "loss": 0.9176, + "step": 6331 + }, + { + "epoch": 0.761377983526724, + "grad_norm": 0.7648382621155223, + "learning_rate": 5.679986911537935e-07, + "loss": 0.5905, + "step": 6332 + }, + { + "epoch": 0.7614982264173631, + "grad_norm": 1.8998613708517234, + "learning_rate": 5.674549999825462e-07, + "loss": 0.6797, + "step": 6333 + }, + { + "epoch": 0.7616184693080021, + "grad_norm": 0.9817465248774957, + "learning_rate": 5.669115261272363e-07, + "loss": 0.7157, + "step": 6334 + }, + { + "epoch": 0.7617387121986413, + "grad_norm": 4.583438224748209, + "learning_rate": 5.663682696703081e-07, + "loss": 0.7291, + "step": 6335 + }, + { + "epoch": 0.7618589550892804, + "grad_norm": 1.9675361887840348, + "learning_rate": 5.658252306941746e-07, + "loss": 0.8213, + "step": 6336 + }, + { + "epoch": 0.7619791979799194, + "grad_norm": 2.2676546673666356, + "learning_rate": 5.65282409281212e-07, + "loss": 0.755, + "step": 6337 + }, + { + "epoch": 0.7620994408705585, + "grad_norm": 2.403889476679173, + "learning_rate": 5.64739805513768e-07, + "loss": 0.7048, + "step": 6338 + }, + { + "epoch": 0.7622196837611976, + "grad_norm": 0.8028822342702453, + "learning_rate": 5.641974194741541e-07, + "loss": 0.5585, + "step": 6339 + }, + { + "epoch": 0.7623399266518367, + "grad_norm": 0.8638615103700453, + "learning_rate": 5.636552512446502e-07, + "loss": 0.6374, + "step": 6340 + }, + { + "epoch": 0.7624601695424758, + "grad_norm": 1.8252508931603864, + "learning_rate": 5.631133009075027e-07, + "loss": 0.7816, + "step": 6341 + }, + { + "epoch": 0.7625804124331149, + "grad_norm": 1.899341132717827, + "learning_rate": 5.625715685449242e-07, + "loss": 0.6898, + "step": 6342 + }, + { + "epoch": 0.762700655323754, + "grad_norm": 1.717657747867486, + "learning_rate": 5.620300542390966e-07, + "loss": 0.7186, + "step": 6343 + }, + { + "epoch": 0.762820898214393, + "grad_norm": 1.8446240050774663, + "learning_rate": 5.614887580721659e-07, + "loss": 0.8567, + "step": 6344 + }, + { + "epoch": 0.7629411411050322, + "grad_norm": 1.885636234319523, + "learning_rate": 5.609476801262481e-07, + "loss": 0.7421, + "step": 6345 + }, + { + "epoch": 0.7630613839956712, + "grad_norm": 2.1397847106521426, + "learning_rate": 5.604068204834223e-07, + "loss": 0.6429, + "step": 6346 + }, + { + "epoch": 0.7631816268863103, + "grad_norm": 2.036917307093314, + "learning_rate": 5.598661792257367e-07, + "loss": 0.7669, + "step": 6347 + }, + { + "epoch": 0.7633018697769495, + "grad_norm": 2.148638795612522, + "learning_rate": 5.593257564352071e-07, + "loss": 0.7606, + "step": 6348 + }, + { + "epoch": 0.7634221126675885, + "grad_norm": 1.6466006490212428, + "learning_rate": 5.58785552193815e-07, + "loss": 0.7589, + "step": 6349 + }, + { + "epoch": 0.7635423555582276, + "grad_norm": 1.8845773869730673, + "learning_rate": 5.582455665835086e-07, + "loss": 0.7618, + "step": 6350 + }, + { + "epoch": 0.7636625984488667, + "grad_norm": 2.5080408485529606, + "learning_rate": 5.577057996862036e-07, + "loss": 0.7295, + "step": 6351 + }, + { + "epoch": 0.7637828413395058, + "grad_norm": 1.6179169562286555, + "learning_rate": 5.571662515837814e-07, + "loss": 0.763, + "step": 6352 + }, + { + "epoch": 0.7639030842301449, + "grad_norm": 1.7498508117298937, + "learning_rate": 5.566269223580926e-07, + "loss": 0.8353, + "step": 6353 + }, + { + "epoch": 0.764023327120784, + "grad_norm": 1.7386901374565698, + "learning_rate": 5.560878120909511e-07, + "loss": 0.758, + "step": 6354 + }, + { + "epoch": 0.7641435700114231, + "grad_norm": 0.8585654173351097, + "learning_rate": 5.55548920864141e-07, + "loss": 0.586, + "step": 6355 + }, + { + "epoch": 0.7642638129020621, + "grad_norm": 2.222918932364494, + "learning_rate": 5.550102487594113e-07, + "loss": 0.7766, + "step": 6356 + }, + { + "epoch": 0.7643840557927013, + "grad_norm": 1.5585636910429665, + "learning_rate": 5.54471795858477e-07, + "loss": 0.714, + "step": 6357 + }, + { + "epoch": 0.7645042986833404, + "grad_norm": 2.04353794315992, + "learning_rate": 5.539335622430235e-07, + "loss": 0.8343, + "step": 6358 + }, + { + "epoch": 0.7646245415739794, + "grad_norm": 2.162167266649587, + "learning_rate": 5.533955479946975e-07, + "loss": 0.7463, + "step": 6359 + }, + { + "epoch": 0.7647447844646186, + "grad_norm": 0.8722414060340394, + "learning_rate": 5.528577531951173e-07, + "loss": 0.658, + "step": 6360 + }, + { + "epoch": 0.7648650273552576, + "grad_norm": 2.216925287098476, + "learning_rate": 5.523201779258653e-07, + "loss": 0.7366, + "step": 6361 + }, + { + "epoch": 0.7649852702458967, + "grad_norm": 2.3202162186397657, + "learning_rate": 5.517828222684912e-07, + "loss": 0.8484, + "step": 6362 + }, + { + "epoch": 0.7651055131365359, + "grad_norm": 0.7780352636474998, + "learning_rate": 5.512456863045117e-07, + "loss": 0.5911, + "step": 6363 + }, + { + "epoch": 0.7652257560271749, + "grad_norm": 1.950919408951115, + "learning_rate": 5.507087701154089e-07, + "loss": 0.7456, + "step": 6364 + }, + { + "epoch": 0.765345998917814, + "grad_norm": 1.8547121452873006, + "learning_rate": 5.50172073782634e-07, + "loss": 0.7541, + "step": 6365 + }, + { + "epoch": 0.7654662418084531, + "grad_norm": 1.7298004030839158, + "learning_rate": 5.496355973876023e-07, + "loss": 0.8797, + "step": 6366 + }, + { + "epoch": 0.7655864846990922, + "grad_norm": 1.64793496810968, + "learning_rate": 5.490993410116984e-07, + "loss": 0.7077, + "step": 6367 + }, + { + "epoch": 0.7657067275897312, + "grad_norm": 1.6346220365880884, + "learning_rate": 5.485633047362704e-07, + "loss": 0.7007, + "step": 6368 + }, + { + "epoch": 0.7658269704803703, + "grad_norm": 1.8390450657461375, + "learning_rate": 5.480274886426341e-07, + "loss": 0.7893, + "step": 6369 + }, + { + "epoch": 0.7659472133710095, + "grad_norm": 2.482245246131933, + "learning_rate": 5.474918928120744e-07, + "loss": 0.7806, + "step": 6370 + }, + { + "epoch": 0.7660674562616485, + "grad_norm": 1.8134995696827452, + "learning_rate": 5.469565173258392e-07, + "loss": 0.8767, + "step": 6371 + }, + { + "epoch": 0.7661876991522876, + "grad_norm": 1.8940053937567312, + "learning_rate": 5.464213622651454e-07, + "loss": 0.6407, + "step": 6372 + }, + { + "epoch": 0.7663079420429267, + "grad_norm": 1.6663235629098552, + "learning_rate": 5.458864277111753e-07, + "loss": 0.8436, + "step": 6373 + }, + { + "epoch": 0.7664281849335658, + "grad_norm": 2.3481052876899784, + "learning_rate": 5.453517137450769e-07, + "loss": 0.6987, + "step": 6374 + }, + { + "epoch": 0.7665484278242048, + "grad_norm": 1.77669786727639, + "learning_rate": 5.448172204479684e-07, + "loss": 0.765, + "step": 6375 + }, + { + "epoch": 0.766668670714844, + "grad_norm": 1.7260881005232098, + "learning_rate": 5.442829479009294e-07, + "loss": 0.7507, + "step": 6376 + }, + { + "epoch": 0.7667889136054831, + "grad_norm": 2.139890892955162, + "learning_rate": 5.437488961850103e-07, + "loss": 0.7175, + "step": 6377 + }, + { + "epoch": 0.7669091564961221, + "grad_norm": 1.78131625403613, + "learning_rate": 5.432150653812258e-07, + "loss": 0.7602, + "step": 6378 + }, + { + "epoch": 0.7670293993867613, + "grad_norm": 2.1894465913428385, + "learning_rate": 5.42681455570557e-07, + "loss": 0.8322, + "step": 6379 + }, + { + "epoch": 0.7671496422774003, + "grad_norm": 1.6854298927847553, + "learning_rate": 5.42148066833954e-07, + "loss": 0.647, + "step": 6380 + }, + { + "epoch": 0.7672698851680394, + "grad_norm": 2.420963435745025, + "learning_rate": 5.416148992523289e-07, + "loss": 0.756, + "step": 6381 + }, + { + "epoch": 0.7673901280586786, + "grad_norm": 1.7599019835011755, + "learning_rate": 5.410819529065644e-07, + "loss": 0.7884, + "step": 6382 + }, + { + "epoch": 0.7675103709493176, + "grad_norm": 2.038137174802164, + "learning_rate": 5.405492278775079e-07, + "loss": 0.6549, + "step": 6383 + }, + { + "epoch": 0.7676306138399567, + "grad_norm": 3.076562785182027, + "learning_rate": 5.400167242459732e-07, + "loss": 0.7995, + "step": 6384 + }, + { + "epoch": 0.7677508567305958, + "grad_norm": 1.8853704401183784, + "learning_rate": 5.394844420927405e-07, + "loss": 0.8095, + "step": 6385 + }, + { + "epoch": 0.7678710996212349, + "grad_norm": 2.160000842058936, + "learning_rate": 5.389523814985562e-07, + "loss": 0.7355, + "step": 6386 + }, + { + "epoch": 0.767991342511874, + "grad_norm": 1.7755132953830306, + "learning_rate": 5.384205425441344e-07, + "loss": 0.7645, + "step": 6387 + }, + { + "epoch": 0.7681115854025131, + "grad_norm": 1.7658146444102496, + "learning_rate": 5.378889253101537e-07, + "loss": 0.8461, + "step": 6388 + }, + { + "epoch": 0.7682318282931522, + "grad_norm": 1.6977183064241044, + "learning_rate": 5.373575298772617e-07, + "loss": 0.8119, + "step": 6389 + }, + { + "epoch": 0.7683520711837912, + "grad_norm": 0.7599588653449928, + "learning_rate": 5.368263563260689e-07, + "loss": 0.6133, + "step": 6390 + }, + { + "epoch": 0.7684723140744304, + "grad_norm": 1.6254333505146656, + "learning_rate": 5.362954047371537e-07, + "loss": 0.6479, + "step": 6391 + }, + { + "epoch": 0.7685925569650695, + "grad_norm": 1.8938508060428931, + "learning_rate": 5.357646751910627e-07, + "loss": 0.7279, + "step": 6392 + }, + { + "epoch": 0.7687127998557085, + "grad_norm": 2.0823511733126527, + "learning_rate": 5.352341677683061e-07, + "loss": 0.803, + "step": 6393 + }, + { + "epoch": 0.7688330427463477, + "grad_norm": 1.9105042060699926, + "learning_rate": 5.347038825493617e-07, + "loss": 0.7938, + "step": 6394 + }, + { + "epoch": 0.7689532856369867, + "grad_norm": 2.0448213714706585, + "learning_rate": 5.341738196146732e-07, + "loss": 0.6902, + "step": 6395 + }, + { + "epoch": 0.7690735285276258, + "grad_norm": 2.0414423008336673, + "learning_rate": 5.336439790446503e-07, + "loss": 0.7355, + "step": 6396 + }, + { + "epoch": 0.769193771418265, + "grad_norm": 1.8247917631308739, + "learning_rate": 5.331143609196711e-07, + "loss": 0.6261, + "step": 6397 + }, + { + "epoch": 0.769314014308904, + "grad_norm": 1.8713930404310266, + "learning_rate": 5.325849653200758e-07, + "loss": 0.7751, + "step": 6398 + }, + { + "epoch": 0.7694342571995431, + "grad_norm": 2.2082224644114876, + "learning_rate": 5.32055792326175e-07, + "loss": 0.7692, + "step": 6399 + }, + { + "epoch": 0.7695545000901821, + "grad_norm": 1.9459809529782783, + "learning_rate": 5.315268420182437e-07, + "loss": 0.7325, + "step": 6400 + }, + { + "epoch": 0.7696747429808213, + "grad_norm": 1.7960342430554554, + "learning_rate": 5.309981144765221e-07, + "loss": 0.7656, + "step": 6401 + }, + { + "epoch": 0.7697949858714603, + "grad_norm": 2.3108972717456577, + "learning_rate": 5.304696097812196e-07, + "loss": 0.765, + "step": 6402 + }, + { + "epoch": 0.7699152287620994, + "grad_norm": 3.9545329003050065, + "learning_rate": 5.299413280125078e-07, + "loss": 0.6072, + "step": 6403 + }, + { + "epoch": 0.7700354716527386, + "grad_norm": 2.0430784469533405, + "learning_rate": 5.294132692505284e-07, + "loss": 0.7312, + "step": 6404 + }, + { + "epoch": 0.7701557145433776, + "grad_norm": 2.222185296175715, + "learning_rate": 5.288854335753861e-07, + "loss": 0.793, + "step": 6405 + }, + { + "epoch": 0.7702759574340167, + "grad_norm": 2.993371156681515, + "learning_rate": 5.283578210671551e-07, + "loss": 0.7628, + "step": 6406 + }, + { + "epoch": 0.7703962003246558, + "grad_norm": 2.580414466390889, + "learning_rate": 5.278304318058719e-07, + "loss": 0.7677, + "step": 6407 + }, + { + "epoch": 0.7705164432152949, + "grad_norm": 2.64448391802985, + "learning_rate": 5.273032658715411e-07, + "loss": 0.7899, + "step": 6408 + }, + { + "epoch": 0.7706366861059339, + "grad_norm": 1.7266703438198059, + "learning_rate": 5.267763233441347e-07, + "loss": 0.7623, + "step": 6409 + }, + { + "epoch": 0.7707569289965731, + "grad_norm": 2.671664938003003, + "learning_rate": 5.26249604303588e-07, + "loss": 0.6979, + "step": 6410 + }, + { + "epoch": 0.7708771718872122, + "grad_norm": 2.0616923976311186, + "learning_rate": 5.257231088298057e-07, + "loss": 0.7866, + "step": 6411 + }, + { + "epoch": 0.7709974147778512, + "grad_norm": 0.8168525753431065, + "learning_rate": 5.25196837002655e-07, + "loss": 0.5403, + "step": 6412 + }, + { + "epoch": 0.7711176576684904, + "grad_norm": 2.049591591323292, + "learning_rate": 5.24670788901971e-07, + "loss": 0.6822, + "step": 6413 + }, + { + "epoch": 0.7712379005591294, + "grad_norm": 2.043017764260264, + "learning_rate": 5.241449646075557e-07, + "loss": 0.6925, + "step": 6414 + }, + { + "epoch": 0.7713581434497685, + "grad_norm": 2.569949417210199, + "learning_rate": 5.236193641991762e-07, + "loss": 0.7308, + "step": 6415 + }, + { + "epoch": 0.7714783863404077, + "grad_norm": 2.261792328317962, + "learning_rate": 5.23093987756565e-07, + "loss": 0.7004, + "step": 6416 + }, + { + "epoch": 0.7715986292310467, + "grad_norm": 1.8465769157679752, + "learning_rate": 5.225688353594217e-07, + "loss": 0.7549, + "step": 6417 + }, + { + "epoch": 0.7717188721216858, + "grad_norm": 2.1110268235808682, + "learning_rate": 5.220439070874108e-07, + "loss": 0.7818, + "step": 6418 + }, + { + "epoch": 0.7718391150123249, + "grad_norm": 1.6702046829697, + "learning_rate": 5.215192030201652e-07, + "loss": 0.7161, + "step": 6419 + }, + { + "epoch": 0.771959357902964, + "grad_norm": 1.814300760661619, + "learning_rate": 5.209947232372798e-07, + "loss": 0.866, + "step": 6420 + }, + { + "epoch": 0.772079600793603, + "grad_norm": 1.6807087477263387, + "learning_rate": 5.204704678183196e-07, + "loss": 0.8127, + "step": 6421 + }, + { + "epoch": 0.7721998436842422, + "grad_norm": 2.1299075510441465, + "learning_rate": 5.19946436842813e-07, + "loss": 0.8602, + "step": 6422 + }, + { + "epoch": 0.7723200865748813, + "grad_norm": 2.5947826122451128, + "learning_rate": 5.194226303902546e-07, + "loss": 0.6887, + "step": 6423 + }, + { + "epoch": 0.7724403294655203, + "grad_norm": 2.2567556475062314, + "learning_rate": 5.188990485401072e-07, + "loss": 0.71, + "step": 6424 + }, + { + "epoch": 0.7725605723561595, + "grad_norm": 2.0583534172818307, + "learning_rate": 5.183756913717954e-07, + "loss": 0.8664, + "step": 6425 + }, + { + "epoch": 0.7726808152467985, + "grad_norm": 2.5819029898824275, + "learning_rate": 5.178525589647136e-07, + "loss": 0.7348, + "step": 6426 + }, + { + "epoch": 0.7728010581374376, + "grad_norm": 2.221299976136581, + "learning_rate": 5.173296513982197e-07, + "loss": 0.7899, + "step": 6427 + }, + { + "epoch": 0.7729213010280768, + "grad_norm": 2.2563888126690372, + "learning_rate": 5.168069687516398e-07, + "loss": 0.6492, + "step": 6428 + }, + { + "epoch": 0.7730415439187158, + "grad_norm": 2.0835529084524245, + "learning_rate": 5.16284511104263e-07, + "loss": 0.7251, + "step": 6429 + }, + { + "epoch": 0.7731617868093549, + "grad_norm": 2.898132308151116, + "learning_rate": 5.157622785353457e-07, + "loss": 0.8138, + "step": 6430 + }, + { + "epoch": 0.7732820296999939, + "grad_norm": 0.6578392132117052, + "learning_rate": 5.152402711241113e-07, + "loss": 0.6044, + "step": 6431 + }, + { + "epoch": 0.7734022725906331, + "grad_norm": 1.826640643168799, + "learning_rate": 5.147184889497465e-07, + "loss": 0.8314, + "step": 6432 + }, + { + "epoch": 0.7735225154812722, + "grad_norm": 2.1649534959622008, + "learning_rate": 5.141969320914072e-07, + "loss": 0.8002, + "step": 6433 + }, + { + "epoch": 0.7736427583719112, + "grad_norm": 2.8005457823024376, + "learning_rate": 5.136756006282113e-07, + "loss": 0.6274, + "step": 6434 + }, + { + "epoch": 0.7737630012625504, + "grad_norm": 3.131153609263635, + "learning_rate": 5.131544946392446e-07, + "loss": 0.8521, + "step": 6435 + }, + { + "epoch": 0.7738832441531894, + "grad_norm": 2.589956059617049, + "learning_rate": 5.126336142035592e-07, + "loss": 0.6473, + "step": 6436 + }, + { + "epoch": 0.7740034870438285, + "grad_norm": 2.381079471920062, + "learning_rate": 5.121129594001721e-07, + "loss": 0.7341, + "step": 6437 + }, + { + "epoch": 0.7741237299344677, + "grad_norm": 1.5823116854551693, + "learning_rate": 5.115925303080661e-07, + "loss": 0.8162, + "step": 6438 + }, + { + "epoch": 0.7742439728251067, + "grad_norm": 2.241173172509984, + "learning_rate": 5.110723270061899e-07, + "loss": 0.7973, + "step": 6439 + }, + { + "epoch": 0.7743642157157458, + "grad_norm": 1.805237393182916, + "learning_rate": 5.105523495734572e-07, + "loss": 0.8016, + "step": 6440 + }, + { + "epoch": 0.7744844586063849, + "grad_norm": 1.6119438335960894, + "learning_rate": 5.100325980887499e-07, + "loss": 0.7512, + "step": 6441 + }, + { + "epoch": 0.774604701497024, + "grad_norm": 2.4998190814358723, + "learning_rate": 5.095130726309116e-07, + "loss": 0.8345, + "step": 6442 + }, + { + "epoch": 0.774724944387663, + "grad_norm": 0.8013811423090549, + "learning_rate": 5.089937732787559e-07, + "loss": 0.5901, + "step": 6443 + }, + { + "epoch": 0.7748451872783022, + "grad_norm": 2.403040541416013, + "learning_rate": 5.084747001110592e-07, + "loss": 0.6676, + "step": 6444 + }, + { + "epoch": 0.7749654301689413, + "grad_norm": 1.7115989347142102, + "learning_rate": 5.07955853206564e-07, + "loss": 0.704, + "step": 6445 + }, + { + "epoch": 0.7750856730595803, + "grad_norm": 4.296573142806296, + "learning_rate": 5.074372326439807e-07, + "loss": 0.7136, + "step": 6446 + }, + { + "epoch": 0.7752059159502195, + "grad_norm": 2.3325653742453123, + "learning_rate": 5.069188385019814e-07, + "loss": 0.7448, + "step": 6447 + }, + { + "epoch": 0.7753261588408585, + "grad_norm": 2.3553752216350547, + "learning_rate": 5.064006708592077e-07, + "loss": 0.6194, + "step": 6448 + }, + { + "epoch": 0.7754464017314976, + "grad_norm": 2.4891093506507644, + "learning_rate": 5.058827297942641e-07, + "loss": 0.7644, + "step": 6449 + }, + { + "epoch": 0.7755666446221368, + "grad_norm": 2.032116397069571, + "learning_rate": 5.053650153857237e-07, + "loss": 0.7535, + "step": 6450 + }, + { + "epoch": 0.7756868875127758, + "grad_norm": 1.6976490004797682, + "learning_rate": 5.048475277121214e-07, + "loss": 0.7, + "step": 6451 + }, + { + "epoch": 0.7758071304034149, + "grad_norm": 1.7258017404495996, + "learning_rate": 5.043302668519598e-07, + "loss": 0.7748, + "step": 6452 + }, + { + "epoch": 0.775927373294054, + "grad_norm": 1.7579019481364953, + "learning_rate": 5.038132328837079e-07, + "loss": 0.7262, + "step": 6453 + }, + { + "epoch": 0.7760476161846931, + "grad_norm": 3.1732275573614857, + "learning_rate": 5.032964258857993e-07, + "loss": 0.7422, + "step": 6454 + }, + { + "epoch": 0.7761678590753321, + "grad_norm": 1.9985896144356228, + "learning_rate": 5.027798459366329e-07, + "loss": 0.6836, + "step": 6455 + }, + { + "epoch": 0.7762881019659713, + "grad_norm": 2.286577666230923, + "learning_rate": 5.02263493114573e-07, + "loss": 0.6373, + "step": 6456 + }, + { + "epoch": 0.7764083448566104, + "grad_norm": 2.2946060287972263, + "learning_rate": 5.017473674979502e-07, + "loss": 0.772, + "step": 6457 + }, + { + "epoch": 0.7765285877472494, + "grad_norm": 0.7539851093061384, + "learning_rate": 5.01231469165061e-07, + "loss": 0.5836, + "step": 6458 + }, + { + "epoch": 0.7766488306378886, + "grad_norm": 0.837282934099801, + "learning_rate": 5.007157981941663e-07, + "loss": 0.5687, + "step": 6459 + }, + { + "epoch": 0.7767690735285276, + "grad_norm": 0.8920914472994619, + "learning_rate": 5.002003546634928e-07, + "loss": 0.6751, + "step": 6460 + }, + { + "epoch": 0.7768893164191667, + "grad_norm": 1.6420939091255802, + "learning_rate": 4.996851386512331e-07, + "loss": 0.7663, + "step": 6461 + }, + { + "epoch": 0.7770095593098058, + "grad_norm": 1.649985287340633, + "learning_rate": 4.991701502355444e-07, + "loss": 0.832, + "step": 6462 + }, + { + "epoch": 0.7771298022004449, + "grad_norm": 1.587401968264528, + "learning_rate": 4.986553894945518e-07, + "loss": 0.7587, + "step": 6463 + }, + { + "epoch": 0.777250045091084, + "grad_norm": 2.0577542585677144, + "learning_rate": 4.981408565063416e-07, + "loss": 0.8687, + "step": 6464 + }, + { + "epoch": 0.777370287981723, + "grad_norm": 1.9525174836415118, + "learning_rate": 4.976265513489701e-07, + "loss": 0.7676, + "step": 6465 + }, + { + "epoch": 0.7774905308723622, + "grad_norm": 3.015177797354279, + "learning_rate": 4.971124741004562e-07, + "loss": 0.8076, + "step": 6466 + }, + { + "epoch": 0.7776107737630013, + "grad_norm": 1.8834479434943965, + "learning_rate": 4.965986248387846e-07, + "loss": 0.7685, + "step": 6467 + }, + { + "epoch": 0.7777310166536403, + "grad_norm": 1.5846230457567991, + "learning_rate": 4.960850036419073e-07, + "loss": 0.7762, + "step": 6468 + }, + { + "epoch": 0.7778512595442795, + "grad_norm": 1.8333812692906557, + "learning_rate": 4.955716105877378e-07, + "loss": 0.7896, + "step": 6469 + }, + { + "epoch": 0.7779715024349185, + "grad_norm": 1.8352073353769625, + "learning_rate": 4.950584457541598e-07, + "loss": 0.8293, + "step": 6470 + }, + { + "epoch": 0.7780917453255576, + "grad_norm": 1.4189001129995606, + "learning_rate": 4.945455092190183e-07, + "loss": 0.8228, + "step": 6471 + }, + { + "epoch": 0.7782119882161967, + "grad_norm": 0.8308535555486432, + "learning_rate": 4.940328010601271e-07, + "loss": 0.5606, + "step": 6472 + }, + { + "epoch": 0.7783322311068358, + "grad_norm": 2.0652671515539094, + "learning_rate": 4.935203213552621e-07, + "loss": 0.7691, + "step": 6473 + }, + { + "epoch": 0.7784524739974749, + "grad_norm": 2.02743653542642, + "learning_rate": 4.930080701821662e-07, + "loss": 0.6704, + "step": 6474 + }, + { + "epoch": 0.778572716888114, + "grad_norm": 1.879030283047936, + "learning_rate": 4.92496047618548e-07, + "loss": 0.7768, + "step": 6475 + }, + { + "epoch": 0.7786929597787531, + "grad_norm": 1.7504532090933884, + "learning_rate": 4.919842537420811e-07, + "loss": 0.7808, + "step": 6476 + }, + { + "epoch": 0.7788132026693921, + "grad_norm": 1.733926938577983, + "learning_rate": 4.91472688630404e-07, + "loss": 0.796, + "step": 6477 + }, + { + "epoch": 0.7789334455600313, + "grad_norm": 1.7576193470268153, + "learning_rate": 4.909613523611202e-07, + "loss": 0.745, + "step": 6478 + }, + { + "epoch": 0.7790536884506704, + "grad_norm": 2.098245832675671, + "learning_rate": 4.904502450117991e-07, + "loss": 0.7506, + "step": 6479 + }, + { + "epoch": 0.7791739313413094, + "grad_norm": 3.475276134689344, + "learning_rate": 4.899393666599762e-07, + "loss": 0.7331, + "step": 6480 + }, + { + "epoch": 0.7792941742319486, + "grad_norm": 2.3934874507653965, + "learning_rate": 4.894287173831506e-07, + "loss": 0.7283, + "step": 6481 + }, + { + "epoch": 0.7794144171225876, + "grad_norm": 1.9435250523166774, + "learning_rate": 4.889182972587877e-07, + "loss": 0.8476, + "step": 6482 + }, + { + "epoch": 0.7795346600132267, + "grad_norm": 1.7465070197015937, + "learning_rate": 4.884081063643177e-07, + "loss": 0.6638, + "step": 6483 + }, + { + "epoch": 0.7796549029038659, + "grad_norm": 0.866978216262792, + "learning_rate": 4.878981447771353e-07, + "loss": 0.5255, + "step": 6484 + }, + { + "epoch": 0.7797751457945049, + "grad_norm": 1.8310406249459996, + "learning_rate": 4.873884125746035e-07, + "loss": 0.7353, + "step": 6485 + }, + { + "epoch": 0.779895388685144, + "grad_norm": 2.8310242013385283, + "learning_rate": 4.868789098340456e-07, + "loss": 0.7281, + "step": 6486 + }, + { + "epoch": 0.7800156315757831, + "grad_norm": 2.38968595125336, + "learning_rate": 4.863696366327543e-07, + "loss": 0.7379, + "step": 6487 + }, + { + "epoch": 0.7801358744664222, + "grad_norm": 1.9581443073137845, + "learning_rate": 4.85860593047986e-07, + "loss": 0.7838, + "step": 6488 + }, + { + "epoch": 0.7802561173570612, + "grad_norm": 1.6125708394421743, + "learning_rate": 4.853517791569613e-07, + "loss": 0.7464, + "step": 6489 + }, + { + "epoch": 0.7803763602477004, + "grad_norm": 1.9759840288404829, + "learning_rate": 4.848431950368684e-07, + "loss": 0.6649, + "step": 6490 + }, + { + "epoch": 0.7804966031383395, + "grad_norm": 0.7143137066328855, + "learning_rate": 4.843348407648569e-07, + "loss": 0.5587, + "step": 6491 + }, + { + "epoch": 0.7806168460289785, + "grad_norm": 3.2280342495506416, + "learning_rate": 4.838267164180457e-07, + "loss": 0.8313, + "step": 6492 + }, + { + "epoch": 0.7807370889196176, + "grad_norm": 28.763516905711587, + "learning_rate": 4.833188220735156e-07, + "loss": 0.8392, + "step": 6493 + }, + { + "epoch": 0.7808573318102567, + "grad_norm": 2.095722692563588, + "learning_rate": 4.828111578083152e-07, + "loss": 0.7501, + "step": 6494 + }, + { + "epoch": 0.7809775747008958, + "grad_norm": 2.0613529599247737, + "learning_rate": 4.823037236994556e-07, + "loss": 0.8179, + "step": 6495 + }, + { + "epoch": 0.7810978175915348, + "grad_norm": 0.7295534243535928, + "learning_rate": 4.817965198239136e-07, + "loss": 0.5638, + "step": 6496 + }, + { + "epoch": 0.781218060482174, + "grad_norm": 2.194122752286118, + "learning_rate": 4.812895462586331e-07, + "loss": 0.749, + "step": 6497 + }, + { + "epoch": 0.7813383033728131, + "grad_norm": 1.655969847971314, + "learning_rate": 4.807828030805207e-07, + "loss": 0.8259, + "step": 6498 + }, + { + "epoch": 0.7814585462634521, + "grad_norm": 1.8072622848306399, + "learning_rate": 4.802762903664495e-07, + "loss": 0.6854, + "step": 6499 + }, + { + "epoch": 0.7815787891540913, + "grad_norm": 2.257663075079493, + "learning_rate": 4.797700081932565e-07, + "loss": 0.742, + "step": 6500 + }, + { + "epoch": 0.7816990320447303, + "grad_norm": 9.05947299258187, + "learning_rate": 4.792639566377442e-07, + "loss": 0.8248, + "step": 6501 + }, + { + "epoch": 0.7818192749353694, + "grad_norm": 2.1109820990153567, + "learning_rate": 4.78758135776681e-07, + "loss": 0.7763, + "step": 6502 + }, + { + "epoch": 0.7819395178260086, + "grad_norm": 2.250704549047265, + "learning_rate": 4.782525456867989e-07, + "loss": 0.7905, + "step": 6503 + }, + { + "epoch": 0.7820597607166476, + "grad_norm": 1.9130793323568338, + "learning_rate": 4.777471864447959e-07, + "loss": 0.8311, + "step": 6504 + }, + { + "epoch": 0.7821800036072867, + "grad_norm": 1.9354649592829265, + "learning_rate": 4.772420581273344e-07, + "loss": 0.811, + "step": 6505 + }, + { + "epoch": 0.7823002464979258, + "grad_norm": 1.8547436391533954, + "learning_rate": 4.7673716081104134e-07, + "loss": 0.7667, + "step": 6506 + }, + { + "epoch": 0.7824204893885649, + "grad_norm": 1.8454782177371625, + "learning_rate": 4.762324945725109e-07, + "loss": 0.847, + "step": 6507 + }, + { + "epoch": 0.782540732279204, + "grad_norm": 1.7897638670246505, + "learning_rate": 4.7572805948829844e-07, + "loss": 0.7594, + "step": 6508 + }, + { + "epoch": 0.7826609751698431, + "grad_norm": 1.680059489377653, + "learning_rate": 4.7522385563492795e-07, + "loss": 0.7119, + "step": 6509 + }, + { + "epoch": 0.7827812180604822, + "grad_norm": 1.8914935426291737, + "learning_rate": 4.747198830888863e-07, + "loss": 0.7097, + "step": 6510 + }, + { + "epoch": 0.7829014609511212, + "grad_norm": 2.004774355474857, + "learning_rate": 4.742161419266251e-07, + "loss": 0.6857, + "step": 6511 + }, + { + "epoch": 0.7830217038417604, + "grad_norm": 5.423877533140551, + "learning_rate": 4.7371263222456304e-07, + "loss": 0.6542, + "step": 6512 + }, + { + "epoch": 0.7831419467323995, + "grad_norm": 0.8089946913786663, + "learning_rate": 4.7320935405908004e-07, + "loss": 0.6151, + "step": 6513 + }, + { + "epoch": 0.7832621896230385, + "grad_norm": 2.6863188698639555, + "learning_rate": 4.7270630750652475e-07, + "loss": 0.8444, + "step": 6514 + }, + { + "epoch": 0.7833824325136777, + "grad_norm": 1.7269650503530727, + "learning_rate": 4.7220349264320746e-07, + "loss": 0.8068, + "step": 6515 + }, + { + "epoch": 0.7835026754043167, + "grad_norm": 0.7410603044761609, + "learning_rate": 4.71700909545407e-07, + "loss": 0.5494, + "step": 6516 + }, + { + "epoch": 0.7836229182949558, + "grad_norm": 2.1352900273678426, + "learning_rate": 4.711985582893627e-07, + "loss": 0.7745, + "step": 6517 + }, + { + "epoch": 0.783743161185595, + "grad_norm": 2.3095139478455136, + "learning_rate": 4.706964389512811e-07, + "loss": 0.7187, + "step": 6518 + }, + { + "epoch": 0.783863404076234, + "grad_norm": 2.05088122896174, + "learning_rate": 4.701945516073345e-07, + "loss": 0.881, + "step": 6519 + }, + { + "epoch": 0.7839836469668731, + "grad_norm": 2.0231473854671966, + "learning_rate": 4.696928963336577e-07, + "loss": 0.7543, + "step": 6520 + }, + { + "epoch": 0.7841038898575122, + "grad_norm": 0.8642598470028137, + "learning_rate": 4.6919147320635224e-07, + "loss": 0.6104, + "step": 6521 + }, + { + "epoch": 0.7842241327481513, + "grad_norm": 2.1998515599197033, + "learning_rate": 4.6869028230148286e-07, + "loss": 0.7341, + "step": 6522 + }, + { + "epoch": 0.7843443756387903, + "grad_norm": 4.693685494092681, + "learning_rate": 4.6818932369507957e-07, + "loss": 0.6062, + "step": 6523 + }, + { + "epoch": 0.7844646185294295, + "grad_norm": 2.1415166181355523, + "learning_rate": 4.676885974631386e-07, + "loss": 0.8943, + "step": 6524 + }, + { + "epoch": 0.7845848614200686, + "grad_norm": 2.329118145577757, + "learning_rate": 4.67188103681619e-07, + "loss": 0.8179, + "step": 6525 + }, + { + "epoch": 0.7847051043107076, + "grad_norm": 2.265569225808638, + "learning_rate": 4.666878424264453e-07, + "loss": 0.7005, + "step": 6526 + }, + { + "epoch": 0.7848253472013467, + "grad_norm": 1.9259083920615323, + "learning_rate": 4.661878137735069e-07, + "loss": 0.7393, + "step": 6527 + }, + { + "epoch": 0.7849455900919858, + "grad_norm": 1.761877894350119, + "learning_rate": 4.656880177986571e-07, + "loss": 0.7481, + "step": 6528 + }, + { + "epoch": 0.7850658329826249, + "grad_norm": 2.1981074776166647, + "learning_rate": 4.6518845457771607e-07, + "loss": 0.8194, + "step": 6529 + }, + { + "epoch": 0.7851860758732639, + "grad_norm": 1.7904887948152959, + "learning_rate": 4.646891241864652e-07, + "loss": 0.799, + "step": 6530 + }, + { + "epoch": 0.7853063187639031, + "grad_norm": 2.1167728257278213, + "learning_rate": 4.6419002670065397e-07, + "loss": 0.7391, + "step": 6531 + }, + { + "epoch": 0.7854265616545422, + "grad_norm": 1.9489004605330458, + "learning_rate": 4.6369116219599445e-07, + "loss": 0.8691, + "step": 6532 + }, + { + "epoch": 0.7855468045451812, + "grad_norm": 1.8482742405151502, + "learning_rate": 4.631925307481637e-07, + "loss": 0.7939, + "step": 6533 + }, + { + "epoch": 0.7856670474358204, + "grad_norm": 2.286256913080291, + "learning_rate": 4.6269413243280533e-07, + "loss": 0.7604, + "step": 6534 + }, + { + "epoch": 0.7857872903264594, + "grad_norm": 2.8328058275332513, + "learning_rate": 4.621959673255236e-07, + "loss": 0.7494, + "step": 6535 + }, + { + "epoch": 0.7859075332170985, + "grad_norm": 2.2234502088300157, + "learning_rate": 4.6169803550189135e-07, + "loss": 0.9035, + "step": 6536 + }, + { + "epoch": 0.7860277761077377, + "grad_norm": 3.121467119211471, + "learning_rate": 4.6120033703744355e-07, + "loss": 0.7782, + "step": 6537 + }, + { + "epoch": 0.7861480189983767, + "grad_norm": 1.8155583853656116, + "learning_rate": 4.607028720076822e-07, + "loss": 0.7812, + "step": 6538 + }, + { + "epoch": 0.7862682618890158, + "grad_norm": 1.8212058040621086, + "learning_rate": 4.6020564048807074e-07, + "loss": 0.7368, + "step": 6539 + }, + { + "epoch": 0.7863885047796549, + "grad_norm": 2.511162537589244, + "learning_rate": 4.5970864255403883e-07, + "loss": 0.7222, + "step": 6540 + }, + { + "epoch": 0.786508747670294, + "grad_norm": 1.787422194488162, + "learning_rate": 4.59211878280982e-07, + "loss": 0.8254, + "step": 6541 + }, + { + "epoch": 0.786628990560933, + "grad_norm": 2.4148981814511608, + "learning_rate": 4.587153477442578e-07, + "loss": 0.7066, + "step": 6542 + }, + { + "epoch": 0.7867492334515722, + "grad_norm": 2.6687354665892054, + "learning_rate": 4.582190510191899e-07, + "loss": 0.8201, + "step": 6543 + }, + { + "epoch": 0.7868694763422113, + "grad_norm": 2.114268412574467, + "learning_rate": 4.5772298818106625e-07, + "loss": 0.8758, + "step": 6544 + }, + { + "epoch": 0.7869897192328503, + "grad_norm": 3.364659752386356, + "learning_rate": 4.572271593051384e-07, + "loss": 0.723, + "step": 6545 + }, + { + "epoch": 0.7871099621234895, + "grad_norm": 1.7409341178098388, + "learning_rate": 4.567315644666245e-07, + "loss": 0.783, + "step": 6546 + }, + { + "epoch": 0.7872302050141285, + "grad_norm": 2.1714201526371086, + "learning_rate": 4.5623620374070507e-07, + "loss": 0.8496, + "step": 6547 + }, + { + "epoch": 0.7873504479047676, + "grad_norm": 0.7791882967539029, + "learning_rate": 4.557410772025263e-07, + "loss": 0.5847, + "step": 6548 + }, + { + "epoch": 0.7874706907954068, + "grad_norm": 1.9263819749936206, + "learning_rate": 4.5524618492719803e-07, + "loss": 0.667, + "step": 6549 + }, + { + "epoch": 0.7875909336860458, + "grad_norm": 2.139419773267406, + "learning_rate": 4.54751526989795e-07, + "loss": 0.7896, + "step": 6550 + }, + { + "epoch": 0.7877111765766849, + "grad_norm": 1.98847963433871, + "learning_rate": 4.5425710346535775e-07, + "loss": 0.7949, + "step": 6551 + }, + { + "epoch": 0.787831419467324, + "grad_norm": 2.0585372309010737, + "learning_rate": 4.537629144288877e-07, + "loss": 0.8194, + "step": 6552 + }, + { + "epoch": 0.7879516623579631, + "grad_norm": 2.7578595295210944, + "learning_rate": 4.5326895995535477e-07, + "loss": 0.7547, + "step": 6553 + }, + { + "epoch": 0.7880719052486022, + "grad_norm": 2.2068985890716157, + "learning_rate": 4.527752401196907e-07, + "loss": 0.8478, + "step": 6554 + }, + { + "epoch": 0.7881921481392413, + "grad_norm": 1.8819773234380532, + "learning_rate": 4.5228175499679254e-07, + "loss": 0.6739, + "step": 6555 + }, + { + "epoch": 0.7883123910298804, + "grad_norm": 0.8566693803420175, + "learning_rate": 4.5178850466152174e-07, + "loss": 0.545, + "step": 6556 + }, + { + "epoch": 0.7884326339205194, + "grad_norm": 1.8299843225693517, + "learning_rate": 4.512954891887031e-07, + "loss": 0.825, + "step": 6557 + }, + { + "epoch": 0.7885528768111585, + "grad_norm": 2.418767596828906, + "learning_rate": 4.5080270865312806e-07, + "loss": 0.8374, + "step": 6558 + }, + { + "epoch": 0.7886731197017977, + "grad_norm": 1.9718977574699244, + "learning_rate": 4.5031016312954985e-07, + "loss": 0.7115, + "step": 6559 + }, + { + "epoch": 0.7887933625924367, + "grad_norm": 1.8158072191414434, + "learning_rate": 4.498178526926886e-07, + "loss": 0.7463, + "step": 6560 + }, + { + "epoch": 0.7889136054830758, + "grad_norm": 2.0027758884759046, + "learning_rate": 4.4932577741722635e-07, + "loss": 0.7319, + "step": 6561 + }, + { + "epoch": 0.7890338483737149, + "grad_norm": 1.6692740628295064, + "learning_rate": 4.4883393737780985e-07, + "loss": 0.7437, + "step": 6562 + }, + { + "epoch": 0.789154091264354, + "grad_norm": 2.9644416466044743, + "learning_rate": 4.4834233264905254e-07, + "loss": 0.7856, + "step": 6563 + }, + { + "epoch": 0.789274334154993, + "grad_norm": 2.3343106221382453, + "learning_rate": 4.478509633055294e-07, + "loss": 0.7163, + "step": 6564 + }, + { + "epoch": 0.7893945770456322, + "grad_norm": 2.3290804752059024, + "learning_rate": 4.473598294217813e-07, + "loss": 0.8031, + "step": 6565 + }, + { + "epoch": 0.7895148199362713, + "grad_norm": 3.3727888705520574, + "learning_rate": 4.468689310723124e-07, + "loss": 0.7114, + "step": 6566 + }, + { + "epoch": 0.7896350628269103, + "grad_norm": 1.8284242825184411, + "learning_rate": 4.463782683315913e-07, + "loss": 0.7884, + "step": 6567 + }, + { + "epoch": 0.7897553057175495, + "grad_norm": 1.9303506182871246, + "learning_rate": 4.458878412740523e-07, + "loss": 0.7334, + "step": 6568 + }, + { + "epoch": 0.7898755486081885, + "grad_norm": 2.3983955503889445, + "learning_rate": 4.453976499740919e-07, + "loss": 0.7756, + "step": 6569 + }, + { + "epoch": 0.7899957914988276, + "grad_norm": 2.3493104207021736, + "learning_rate": 4.4490769450607215e-07, + "loss": 0.7839, + "step": 6570 + }, + { + "epoch": 0.7901160343894668, + "grad_norm": 1.9787411951868028, + "learning_rate": 4.4441797494431845e-07, + "loss": 0.7334, + "step": 6571 + }, + { + "epoch": 0.7902362772801058, + "grad_norm": 2.090172287259997, + "learning_rate": 4.439284913631207e-07, + "loss": 0.7875, + "step": 6572 + }, + { + "epoch": 0.7903565201707449, + "grad_norm": 1.8422682028606374, + "learning_rate": 4.434392438367347e-07, + "loss": 0.8391, + "step": 6573 + }, + { + "epoch": 0.790476763061384, + "grad_norm": 1.777864082380684, + "learning_rate": 4.4295023243937677e-07, + "loss": 0.7495, + "step": 6574 + }, + { + "epoch": 0.7905970059520231, + "grad_norm": 1.765402720815451, + "learning_rate": 4.4246145724523123e-07, + "loss": 0.8058, + "step": 6575 + }, + { + "epoch": 0.7907172488426621, + "grad_norm": 3.5876824947287904, + "learning_rate": 4.41972918328444e-07, + "loss": 0.7694, + "step": 6576 + }, + { + "epoch": 0.7908374917333013, + "grad_norm": 2.187560598351497, + "learning_rate": 4.4148461576312646e-07, + "loss": 0.773, + "step": 6577 + }, + { + "epoch": 0.7909577346239404, + "grad_norm": 1.4595921487602816, + "learning_rate": 4.4099654962335343e-07, + "loss": 0.748, + "step": 6578 + }, + { + "epoch": 0.7910779775145794, + "grad_norm": 1.8069773097384334, + "learning_rate": 4.405087199831636e-07, + "loss": 0.7535, + "step": 6579 + }, + { + "epoch": 0.7911982204052186, + "grad_norm": 2.4969731126589645, + "learning_rate": 4.400211269165619e-07, + "loss": 0.6753, + "step": 6580 + }, + { + "epoch": 0.7913184632958576, + "grad_norm": 1.4825712789604708, + "learning_rate": 4.3953377049751416e-07, + "loss": 0.7679, + "step": 6581 + }, + { + "epoch": 0.7914387061864967, + "grad_norm": 4.6727310468066765, + "learning_rate": 4.390466507999537e-07, + "loss": 0.7901, + "step": 6582 + }, + { + "epoch": 0.7915589490771359, + "grad_norm": 2.066541699763833, + "learning_rate": 4.385597678977748e-07, + "loss": 0.7647, + "step": 6583 + }, + { + "epoch": 0.7916791919677749, + "grad_norm": 1.9421470042492925, + "learning_rate": 4.3807312186483726e-07, + "loss": 0.7592, + "step": 6584 + }, + { + "epoch": 0.791799434858414, + "grad_norm": 2.596802994410351, + "learning_rate": 4.375867127749655e-07, + "loss": 0.7856, + "step": 6585 + }, + { + "epoch": 0.7919196777490531, + "grad_norm": 2.197951498855423, + "learning_rate": 4.3710054070194744e-07, + "loss": 0.6758, + "step": 6586 + }, + { + "epoch": 0.7920399206396922, + "grad_norm": 2.464108316579732, + "learning_rate": 4.3661460571953455e-07, + "loss": 0.6639, + "step": 6587 + }, + { + "epoch": 0.7921601635303313, + "grad_norm": 1.6132406825398484, + "learning_rate": 4.36128907901443e-07, + "loss": 0.6843, + "step": 6588 + }, + { + "epoch": 0.7922804064209703, + "grad_norm": 4.559028872763948, + "learning_rate": 4.356434473213519e-07, + "loss": 0.7351, + "step": 6589 + }, + { + "epoch": 0.7924006493116095, + "grad_norm": 1.6651995718547445, + "learning_rate": 4.351582240529068e-07, + "loss": 0.8025, + "step": 6590 + }, + { + "epoch": 0.7925208922022485, + "grad_norm": 0.6922890582185228, + "learning_rate": 4.346732381697149e-07, + "loss": 0.5824, + "step": 6591 + }, + { + "epoch": 0.7926411350928876, + "grad_norm": 1.9473465545886313, + "learning_rate": 4.3418848974534825e-07, + "loss": 0.8142, + "step": 6592 + }, + { + "epoch": 0.7927613779835267, + "grad_norm": 1.779550913190381, + "learning_rate": 4.3370397885334276e-07, + "loss": 0.6924, + "step": 6593 + }, + { + "epoch": 0.7928816208741658, + "grad_norm": 2.075047559078231, + "learning_rate": 4.3321970556719777e-07, + "loss": 0.7609, + "step": 6594 + }, + { + "epoch": 0.7930018637648049, + "grad_norm": 2.430690251811251, + "learning_rate": 4.3273566996037856e-07, + "loss": 0.7208, + "step": 6595 + }, + { + "epoch": 0.793122106655444, + "grad_norm": 2.0216253111724716, + "learning_rate": 4.322518721063113e-07, + "loss": 0.809, + "step": 6596 + }, + { + "epoch": 0.7932423495460831, + "grad_norm": 1.878036075460858, + "learning_rate": 4.3176831207838906e-07, + "loss": 0.7014, + "step": 6597 + }, + { + "epoch": 0.7933625924367221, + "grad_norm": 2.1410980815630674, + "learning_rate": 4.3128498994996685e-07, + "loss": 0.7487, + "step": 6598 + }, + { + "epoch": 0.7934828353273613, + "grad_norm": 2.0134454816357783, + "learning_rate": 4.308019057943646e-07, + "loss": 0.7215, + "step": 6599 + }, + { + "epoch": 0.7936030782180004, + "grad_norm": 1.6602442819031304, + "learning_rate": 4.3031905968486535e-07, + "loss": 0.7474, + "step": 6600 + }, + { + "epoch": 0.7937233211086394, + "grad_norm": 2.175766566447872, + "learning_rate": 4.298364516947162e-07, + "loss": 0.6856, + "step": 6601 + }, + { + "epoch": 0.7938435639992786, + "grad_norm": 2.2147602153008115, + "learning_rate": 4.293540818971295e-07, + "loss": 0.6593, + "step": 6602 + }, + { + "epoch": 0.7939638068899176, + "grad_norm": 2.128704656452909, + "learning_rate": 4.2887195036527934e-07, + "loss": 0.7734, + "step": 6603 + }, + { + "epoch": 0.7940840497805567, + "grad_norm": 2.422478286913554, + "learning_rate": 4.28390057172306e-07, + "loss": 0.737, + "step": 6604 + }, + { + "epoch": 0.7942042926711959, + "grad_norm": 2.2068063266886795, + "learning_rate": 4.279084023913111e-07, + "loss": 0.7277, + "step": 6605 + }, + { + "epoch": 0.7943245355618349, + "grad_norm": 1.873443720586915, + "learning_rate": 4.2742698609536096e-07, + "loss": 0.7012, + "step": 6606 + }, + { + "epoch": 0.794444778452474, + "grad_norm": 1.912444509062718, + "learning_rate": 4.2694580835748706e-07, + "loss": 0.79, + "step": 6607 + }, + { + "epoch": 0.7945650213431131, + "grad_norm": 2.255040562639162, + "learning_rate": 4.264648692506836e-07, + "loss": 0.7449, + "step": 6608 + }, + { + "epoch": 0.7946852642337522, + "grad_norm": 1.8930590385665564, + "learning_rate": 4.2598416884790824e-07, + "loss": 0.7268, + "step": 6609 + }, + { + "epoch": 0.7948055071243912, + "grad_norm": 1.998591045956083, + "learning_rate": 4.255037072220828e-07, + "loss": 0.8128, + "step": 6610 + }, + { + "epoch": 0.7949257500150304, + "grad_norm": 1.7554595118552467, + "learning_rate": 4.2502348444609293e-07, + "loss": 0.7214, + "step": 6611 + }, + { + "epoch": 0.7950459929056695, + "grad_norm": 1.825276364342846, + "learning_rate": 4.2454350059278844e-07, + "loss": 0.6975, + "step": 6612 + }, + { + "epoch": 0.7951662357963085, + "grad_norm": 1.8069442576353638, + "learning_rate": 4.240637557349824e-07, + "loss": 0.8464, + "step": 6613 + }, + { + "epoch": 0.7952864786869477, + "grad_norm": 1.8873847749415142, + "learning_rate": 4.235842499454516e-07, + "loss": 0.6646, + "step": 6614 + }, + { + "epoch": 0.7954067215775867, + "grad_norm": 1.9297170907041372, + "learning_rate": 4.2310498329693687e-07, + "loss": 0.8293, + "step": 6615 + }, + { + "epoch": 0.7955269644682258, + "grad_norm": 1.57060345283905, + "learning_rate": 4.2262595586214164e-07, + "loss": 0.8103, + "step": 6616 + }, + { + "epoch": 0.795647207358865, + "grad_norm": 1.663535021070162, + "learning_rate": 4.221471677137358e-07, + "loss": 0.7712, + "step": 6617 + }, + { + "epoch": 0.795767450249504, + "grad_norm": 1.6341782110423442, + "learning_rate": 4.216686189243492e-07, + "loss": 0.7059, + "step": 6618 + }, + { + "epoch": 0.7958876931401431, + "grad_norm": 1.7034360269145794, + "learning_rate": 4.211903095665785e-07, + "loss": 0.7264, + "step": 6619 + }, + { + "epoch": 0.7960079360307821, + "grad_norm": 1.9664623957224006, + "learning_rate": 4.2071223971298277e-07, + "loss": 0.758, + "step": 6620 + }, + { + "epoch": 0.7961281789214213, + "grad_norm": 2.2857848778109884, + "learning_rate": 4.2023440943608433e-07, + "loss": 0.6152, + "step": 6621 + }, + { + "epoch": 0.7962484218120603, + "grad_norm": 1.6235766412697936, + "learning_rate": 4.1975681880837023e-07, + "loss": 0.7833, + "step": 6622 + }, + { + "epoch": 0.7963686647026994, + "grad_norm": 2.040106261421656, + "learning_rate": 4.192794679022895e-07, + "loss": 0.8305, + "step": 6623 + }, + { + "epoch": 0.7964889075933386, + "grad_norm": 1.8190950571045557, + "learning_rate": 4.1880235679025743e-07, + "loss": 0.7232, + "step": 6624 + }, + { + "epoch": 0.7966091504839776, + "grad_norm": 1.784154565025647, + "learning_rate": 4.1832548554464986e-07, + "loss": 0.6408, + "step": 6625 + }, + { + "epoch": 0.7967293933746167, + "grad_norm": 0.7758000443593519, + "learning_rate": 4.178488542378098e-07, + "loss": 0.5881, + "step": 6626 + }, + { + "epoch": 0.7968496362652558, + "grad_norm": 1.7646430331613425, + "learning_rate": 4.173724629420401e-07, + "loss": 0.89, + "step": 6627 + }, + { + "epoch": 0.7969698791558949, + "grad_norm": 4.1582382150498045, + "learning_rate": 4.168963117296087e-07, + "loss": 0.6855, + "step": 6628 + }, + { + "epoch": 0.797090122046534, + "grad_norm": 2.188705548107913, + "learning_rate": 4.1642040067274876e-07, + "loss": 0.7565, + "step": 6629 + }, + { + "epoch": 0.7972103649371731, + "grad_norm": 2.1805882934034955, + "learning_rate": 4.1594472984365493e-07, + "loss": 0.735, + "step": 6630 + }, + { + "epoch": 0.7973306078278122, + "grad_norm": 2.169925061351452, + "learning_rate": 4.154692993144862e-07, + "loss": 0.7814, + "step": 6631 + }, + { + "epoch": 0.7974508507184512, + "grad_norm": 2.0695711931958405, + "learning_rate": 4.1499410915736476e-07, + "loss": 0.7161, + "step": 6632 + }, + { + "epoch": 0.7975710936090904, + "grad_norm": 0.7763144421204199, + "learning_rate": 4.145191594443762e-07, + "loss": 0.6424, + "step": 6633 + }, + { + "epoch": 0.7976913364997295, + "grad_norm": 2.0280735708807582, + "learning_rate": 4.140444502475713e-07, + "loss": 0.7092, + "step": 6634 + }, + { + "epoch": 0.7978115793903685, + "grad_norm": 2.0759562024180354, + "learning_rate": 4.1356998163896216e-07, + "loss": 0.701, + "step": 6635 + }, + { + "epoch": 0.7979318222810077, + "grad_norm": 2.10094116555603, + "learning_rate": 4.130957536905255e-07, + "loss": 0.7527, + "step": 6636 + }, + { + "epoch": 0.7980520651716467, + "grad_norm": 2.471845014657929, + "learning_rate": 4.1262176647420134e-07, + "loss": 0.7165, + "step": 6637 + }, + { + "epoch": 0.7981723080622858, + "grad_norm": 1.6537762287623243, + "learning_rate": 4.121480200618923e-07, + "loss": 0.8018, + "step": 6638 + }, + { + "epoch": 0.798292550952925, + "grad_norm": 8.646246793734496, + "learning_rate": 4.116745145254674e-07, + "loss": 0.8038, + "step": 6639 + }, + { + "epoch": 0.798412793843564, + "grad_norm": 0.7831467681928166, + "learning_rate": 4.1120124993675476e-07, + "loss": 0.5807, + "step": 6640 + }, + { + "epoch": 0.7985330367342031, + "grad_norm": 2.0319269079444062, + "learning_rate": 4.107282263675498e-07, + "loss": 0.6265, + "step": 6641 + }, + { + "epoch": 0.7986532796248422, + "grad_norm": 0.7867120161024027, + "learning_rate": 4.1025544388960907e-07, + "loss": 0.5253, + "step": 6642 + }, + { + "epoch": 0.7987735225154813, + "grad_norm": 2.0435436599581203, + "learning_rate": 4.097829025746538e-07, + "loss": 0.7185, + "step": 6643 + }, + { + "epoch": 0.7988937654061203, + "grad_norm": 0.674477880371124, + "learning_rate": 4.0931060249436757e-07, + "loss": 0.6113, + "step": 6644 + }, + { + "epoch": 0.7990140082967595, + "grad_norm": 1.9798739591612995, + "learning_rate": 4.088385437203978e-07, + "loss": 0.6972, + "step": 6645 + }, + { + "epoch": 0.7991342511873986, + "grad_norm": 2.1091805827649965, + "learning_rate": 4.083667263243564e-07, + "loss": 0.7846, + "step": 6646 + }, + { + "epoch": 0.7992544940780376, + "grad_norm": 1.9151855910684177, + "learning_rate": 4.0789515037781653e-07, + "loss": 0.7215, + "step": 6647 + }, + { + "epoch": 0.7993747369686768, + "grad_norm": 2.135445558926272, + "learning_rate": 4.0742381595231755e-07, + "loss": 0.8301, + "step": 6648 + }, + { + "epoch": 0.7994949798593158, + "grad_norm": 2.7588327049921895, + "learning_rate": 4.06952723119359e-07, + "loss": 0.786, + "step": 6649 + }, + { + "epoch": 0.7996152227499549, + "grad_norm": 1.8778931866618145, + "learning_rate": 4.0648187195040504e-07, + "loss": 0.6791, + "step": 6650 + }, + { + "epoch": 0.799735465640594, + "grad_norm": 0.8460220193792994, + "learning_rate": 4.060112625168848e-07, + "loss": 0.676, + "step": 6651 + }, + { + "epoch": 0.7998557085312331, + "grad_norm": 3.3359978199135223, + "learning_rate": 4.055408948901886e-07, + "loss": 0.7439, + "step": 6652 + }, + { + "epoch": 0.7999759514218722, + "grad_norm": 1.8099702742016655, + "learning_rate": 4.050707691416708e-07, + "loss": 0.7163, + "step": 6653 + }, + { + "epoch": 0.8000961943125112, + "grad_norm": 0.6858718188910584, + "learning_rate": 4.046008853426495e-07, + "loss": 0.5979, + "step": 6654 + }, + { + "epoch": 0.8002164372031504, + "grad_norm": 2.3931604042731975, + "learning_rate": 4.0413124356440464e-07, + "loss": 0.6319, + "step": 6655 + }, + { + "epoch": 0.8003366800937894, + "grad_norm": 1.7718840529868212, + "learning_rate": 4.0366184387818223e-07, + "loss": 0.8264, + "step": 6656 + }, + { + "epoch": 0.8004569229844285, + "grad_norm": 1.8231951764200702, + "learning_rate": 4.0319268635518797e-07, + "loss": 0.8556, + "step": 6657 + }, + { + "epoch": 0.8005771658750677, + "grad_norm": 2.7348304587194163, + "learning_rate": 4.027237710665943e-07, + "loss": 0.7557, + "step": 6658 + }, + { + "epoch": 0.8006974087657067, + "grad_norm": 2.043621476824794, + "learning_rate": 4.022550980835344e-07, + "loss": 0.6976, + "step": 6659 + }, + { + "epoch": 0.8008176516563458, + "grad_norm": 2.059099341598597, + "learning_rate": 4.017866674771051e-07, + "loss": 0.7971, + "step": 6660 + }, + { + "epoch": 0.8009378945469849, + "grad_norm": 1.6841405701852927, + "learning_rate": 4.013184793183688e-07, + "loss": 0.7468, + "step": 6661 + }, + { + "epoch": 0.801058137437624, + "grad_norm": 2.2400876091105184, + "learning_rate": 4.008505336783472e-07, + "loss": 0.731, + "step": 6662 + }, + { + "epoch": 0.801178380328263, + "grad_norm": 1.8972145191493615, + "learning_rate": 4.003828306280284e-07, + "loss": 0.8101, + "step": 6663 + }, + { + "epoch": 0.8012986232189022, + "grad_norm": 2.0553912563932712, + "learning_rate": 3.999153702383626e-07, + "loss": 0.7801, + "step": 6664 + }, + { + "epoch": 0.8014188661095413, + "grad_norm": 2.8001517595308667, + "learning_rate": 3.9944815258026263e-07, + "loss": 0.7371, + "step": 6665 + }, + { + "epoch": 0.8015391090001803, + "grad_norm": 1.7396534507767456, + "learning_rate": 3.989811777246057e-07, + "loss": 0.8333, + "step": 6666 + }, + { + "epoch": 0.8016593518908195, + "grad_norm": 0.8624502817281188, + "learning_rate": 3.985144457422305e-07, + "loss": 0.6623, + "step": 6667 + }, + { + "epoch": 0.8017795947814585, + "grad_norm": 2.154087696312702, + "learning_rate": 3.9804795670394096e-07, + "loss": 0.7667, + "step": 6668 + }, + { + "epoch": 0.8018998376720976, + "grad_norm": 1.750942793699749, + "learning_rate": 3.975817106805022e-07, + "loss": 0.7062, + "step": 6669 + }, + { + "epoch": 0.8020200805627368, + "grad_norm": 1.8363219812254303, + "learning_rate": 3.97115707742645e-07, + "loss": 0.6456, + "step": 6670 + }, + { + "epoch": 0.8021403234533758, + "grad_norm": 2.067285246206863, + "learning_rate": 3.966499479610599e-07, + "loss": 0.6495, + "step": 6671 + }, + { + "epoch": 0.8022605663440149, + "grad_norm": 1.9212955128448128, + "learning_rate": 3.9618443140640225e-07, + "loss": 0.6465, + "step": 6672 + }, + { + "epoch": 0.802380809234654, + "grad_norm": 0.690825611856684, + "learning_rate": 3.957191581492918e-07, + "loss": 0.5145, + "step": 6673 + }, + { + "epoch": 0.8025010521252931, + "grad_norm": 3.280612702015061, + "learning_rate": 3.952541282603097e-07, + "loss": 0.7179, + "step": 6674 + }, + { + "epoch": 0.8026212950159322, + "grad_norm": 2.278082333598227, + "learning_rate": 3.9478934181000013e-07, + "loss": 0.8401, + "step": 6675 + }, + { + "epoch": 0.8027415379065713, + "grad_norm": 2.2289637416801678, + "learning_rate": 3.943247988688714e-07, + "loss": 0.841, + "step": 6676 + }, + { + "epoch": 0.8028617807972104, + "grad_norm": 1.9219382671084115, + "learning_rate": 3.938604995073933e-07, + "loss": 0.7244, + "step": 6677 + }, + { + "epoch": 0.8029820236878494, + "grad_norm": 1.9424842589340756, + "learning_rate": 3.9339644379600157e-07, + "loss": 0.6582, + "step": 6678 + }, + { + "epoch": 0.8031022665784886, + "grad_norm": 2.1120601941113812, + "learning_rate": 3.929326318050907e-07, + "loss": 0.7165, + "step": 6679 + }, + { + "epoch": 0.8032225094691277, + "grad_norm": 2.840051782431456, + "learning_rate": 3.924690636050225e-07, + "loss": 0.7959, + "step": 6680 + }, + { + "epoch": 0.8033427523597667, + "grad_norm": 2.6574142597251424, + "learning_rate": 3.9200573926611915e-07, + "loss": 0.7262, + "step": 6681 + }, + { + "epoch": 0.8034629952504058, + "grad_norm": 2.0140930978420295, + "learning_rate": 3.9154265885866613e-07, + "loss": 0.7295, + "step": 6682 + }, + { + "epoch": 0.8035832381410449, + "grad_norm": 5.537968579997723, + "learning_rate": 3.9107982245291394e-07, + "loss": 0.7566, + "step": 6683 + }, + { + "epoch": 0.803703481031684, + "grad_norm": 4.315297200918029, + "learning_rate": 3.9061723011907245e-07, + "loss": 0.7761, + "step": 6684 + }, + { + "epoch": 0.803823723922323, + "grad_norm": 2.04877599135292, + "learning_rate": 3.901548819273179e-07, + "loss": 0.7918, + "step": 6685 + }, + { + "epoch": 0.8039439668129622, + "grad_norm": 3.7075823905777687, + "learning_rate": 3.896927779477881e-07, + "loss": 0.6937, + "step": 6686 + }, + { + "epoch": 0.8040642097036013, + "grad_norm": 2.4314770746956484, + "learning_rate": 3.892309182505833e-07, + "loss": 0.6745, + "step": 6687 + }, + { + "epoch": 0.8041844525942403, + "grad_norm": 2.3531350995539824, + "learning_rate": 3.887693029057675e-07, + "loss": 0.8672, + "step": 6688 + }, + { + "epoch": 0.8043046954848795, + "grad_norm": 2.392970013345702, + "learning_rate": 3.8830793198336684e-07, + "loss": 0.8135, + "step": 6689 + }, + { + "epoch": 0.8044249383755185, + "grad_norm": 1.8126866474731615, + "learning_rate": 3.878468055533721e-07, + "loss": 0.7074, + "step": 6690 + }, + { + "epoch": 0.8045451812661576, + "grad_norm": 2.5799051323533893, + "learning_rate": 3.8738592368573464e-07, + "loss": 0.8518, + "step": 6691 + }, + { + "epoch": 0.8046654241567968, + "grad_norm": 2.180044056158497, + "learning_rate": 3.8692528645037137e-07, + "loss": 0.8845, + "step": 6692 + }, + { + "epoch": 0.8047856670474358, + "grad_norm": 2.387938641902051, + "learning_rate": 3.8646489391715907e-07, + "loss": 0.7846, + "step": 6693 + }, + { + "epoch": 0.8049059099380749, + "grad_norm": 2.502366662374614, + "learning_rate": 3.8600474615593903e-07, + "loss": 0.8794, + "step": 6694 + }, + { + "epoch": 0.805026152828714, + "grad_norm": 0.8001265291647236, + "learning_rate": 3.8554484323651605e-07, + "loss": 0.5965, + "step": 6695 + }, + { + "epoch": 0.8051463957193531, + "grad_norm": 2.153050910549769, + "learning_rate": 3.85085185228657e-07, + "loss": 0.7944, + "step": 6696 + }, + { + "epoch": 0.8052666386099921, + "grad_norm": 2.416011282963964, + "learning_rate": 3.8462577220209114e-07, + "loss": 0.7377, + "step": 6697 + }, + { + "epoch": 0.8053868815006313, + "grad_norm": 0.7195389191159774, + "learning_rate": 3.8416660422651127e-07, + "loss": 0.5909, + "step": 6698 + }, + { + "epoch": 0.8055071243912704, + "grad_norm": 1.7812302571591256, + "learning_rate": 3.837076813715723e-07, + "loss": 0.681, + "step": 6699 + }, + { + "epoch": 0.8056273672819094, + "grad_norm": 1.7016357069845085, + "learning_rate": 3.832490037068941e-07, + "loss": 0.7514, + "step": 6700 + }, + { + "epoch": 0.8057476101725486, + "grad_norm": 2.0689241180443485, + "learning_rate": 3.827905713020554e-07, + "loss": 0.7636, + "step": 6701 + }, + { + "epoch": 0.8058678530631876, + "grad_norm": 2.3118034550965483, + "learning_rate": 3.823323842266017e-07, + "loss": 0.7001, + "step": 6702 + }, + { + "epoch": 0.8059880959538267, + "grad_norm": 7.030851225671009, + "learning_rate": 3.818744425500393e-07, + "loss": 0.7321, + "step": 6703 + }, + { + "epoch": 0.8061083388444659, + "grad_norm": 1.7689875763305383, + "learning_rate": 3.8141674634183675e-07, + "loss": 0.8098, + "step": 6704 + }, + { + "epoch": 0.8062285817351049, + "grad_norm": 1.7394595149702625, + "learning_rate": 3.809592956714278e-07, + "loss": 0.6642, + "step": 6705 + }, + { + "epoch": 0.806348824625744, + "grad_norm": 2.2041784703227694, + "learning_rate": 3.805020906082057e-07, + "loss": 0.7506, + "step": 6706 + }, + { + "epoch": 0.8064690675163831, + "grad_norm": 2.220338601932747, + "learning_rate": 3.8004513122152917e-07, + "loss": 0.8144, + "step": 6707 + }, + { + "epoch": 0.8065893104070222, + "grad_norm": 2.0437727853980587, + "learning_rate": 3.79588417580718e-07, + "loss": 0.6763, + "step": 6708 + }, + { + "epoch": 0.8067095532976613, + "grad_norm": 1.866640274111043, + "learning_rate": 3.791319497550558e-07, + "loss": 0.7737, + "step": 6709 + }, + { + "epoch": 0.8068297961883004, + "grad_norm": 1.874978955468652, + "learning_rate": 3.78675727813788e-07, + "loss": 0.7093, + "step": 6710 + }, + { + "epoch": 0.8069500390789395, + "grad_norm": 1.7246243053926198, + "learning_rate": 3.782197518261225e-07, + "loss": 0.737, + "step": 6711 + }, + { + "epoch": 0.8070702819695785, + "grad_norm": 2.876598079861266, + "learning_rate": 3.777640218612319e-07, + "loss": 0.9592, + "step": 6712 + }, + { + "epoch": 0.8071905248602176, + "grad_norm": 2.2314408250091637, + "learning_rate": 3.773085379882488e-07, + "loss": 0.7328, + "step": 6713 + }, + { + "epoch": 0.8073107677508568, + "grad_norm": 1.908959885844687, + "learning_rate": 3.768533002762715e-07, + "loss": 0.764, + "step": 6714 + }, + { + "epoch": 0.8074310106414958, + "grad_norm": 1.8302612864418975, + "learning_rate": 3.763983087943572e-07, + "loss": 0.7712, + "step": 6715 + }, + { + "epoch": 0.8075512535321349, + "grad_norm": 1.9828334437497432, + "learning_rate": 3.759435636115282e-07, + "loss": 0.8136, + "step": 6716 + }, + { + "epoch": 0.807671496422774, + "grad_norm": 1.9417090688285903, + "learning_rate": 3.7548906479676967e-07, + "loss": 0.7321, + "step": 6717 + }, + { + "epoch": 0.8077917393134131, + "grad_norm": 1.986994538720607, + "learning_rate": 3.7503481241902855e-07, + "loss": 0.7162, + "step": 6718 + }, + { + "epoch": 0.8079119822040521, + "grad_norm": 1.7651382720644964, + "learning_rate": 3.745808065472145e-07, + "loss": 0.8002, + "step": 6719 + }, + { + "epoch": 0.8080322250946913, + "grad_norm": 1.6155236572120257, + "learning_rate": 3.741270472501994e-07, + "loss": 0.7657, + "step": 6720 + }, + { + "epoch": 0.8081524679853304, + "grad_norm": 1.6714117531401878, + "learning_rate": 3.736735345968183e-07, + "loss": 0.7288, + "step": 6721 + }, + { + "epoch": 0.8082727108759694, + "grad_norm": 1.5099712986776272, + "learning_rate": 3.7322026865586986e-07, + "loss": 0.7905, + "step": 6722 + }, + { + "epoch": 0.8083929537666086, + "grad_norm": 2.33777222297687, + "learning_rate": 3.7276724949611206e-07, + "loss": 0.7366, + "step": 6723 + }, + { + "epoch": 0.8085131966572476, + "grad_norm": 1.9252896252031582, + "learning_rate": 3.723144771862694e-07, + "loss": 0.7555, + "step": 6724 + }, + { + "epoch": 0.8086334395478867, + "grad_norm": 2.1457679392902413, + "learning_rate": 3.718619517950263e-07, + "loss": 0.773, + "step": 6725 + }, + { + "epoch": 0.8087536824385259, + "grad_norm": 2.035718023922961, + "learning_rate": 3.714096733910301e-07, + "loss": 0.773, + "step": 6726 + }, + { + "epoch": 0.8088739253291649, + "grad_norm": 1.835850879979561, + "learning_rate": 3.709576420428926e-07, + "loss": 0.7074, + "step": 6727 + }, + { + "epoch": 0.808994168219804, + "grad_norm": 5.004954744147723, + "learning_rate": 3.7050585781918463e-07, + "loss": 0.7397, + "step": 6728 + }, + { + "epoch": 0.8091144111104431, + "grad_norm": 2.6286853170304885, + "learning_rate": 3.700543207884428e-07, + "loss": 0.6881, + "step": 6729 + }, + { + "epoch": 0.8092346540010822, + "grad_norm": 2.0721820954683943, + "learning_rate": 3.6960303101916466e-07, + "loss": 0.7085, + "step": 6730 + }, + { + "epoch": 0.8093548968917212, + "grad_norm": 0.9042215336502494, + "learning_rate": 3.6915198857981047e-07, + "loss": 0.5554, + "step": 6731 + }, + { + "epoch": 0.8094751397823604, + "grad_norm": 1.7637376403375535, + "learning_rate": 3.687011935388027e-07, + "loss": 0.6809, + "step": 6732 + }, + { + "epoch": 0.8095953826729995, + "grad_norm": 2.360465397770848, + "learning_rate": 3.6825064596452646e-07, + "loss": 0.7339, + "step": 6733 + }, + { + "epoch": 0.8097156255636385, + "grad_norm": 1.8087119935108438, + "learning_rate": 3.678003459253305e-07, + "loss": 0.7083, + "step": 6734 + }, + { + "epoch": 0.8098358684542777, + "grad_norm": 2.591676181722594, + "learning_rate": 3.673502934895236e-07, + "loss": 0.7442, + "step": 6735 + }, + { + "epoch": 0.8099561113449167, + "grad_norm": 0.7551734433938944, + "learning_rate": 3.669004887253802e-07, + "loss": 0.5805, + "step": 6736 + }, + { + "epoch": 0.8100763542355558, + "grad_norm": 1.679738456485427, + "learning_rate": 3.664509317011335e-07, + "loss": 0.7927, + "step": 6737 + }, + { + "epoch": 0.810196597126195, + "grad_norm": 2.239618936891477, + "learning_rate": 3.6600162248498134e-07, + "loss": 0.7394, + "step": 6738 + }, + { + "epoch": 0.810316840016834, + "grad_norm": 1.9640818155654471, + "learning_rate": 3.6555256114508426e-07, + "loss": 0.7649, + "step": 6739 + }, + { + "epoch": 0.8104370829074731, + "grad_norm": 1.9283916122127502, + "learning_rate": 3.651037477495642e-07, + "loss": 0.7322, + "step": 6740 + }, + { + "epoch": 0.8105573257981122, + "grad_norm": 7.84424385856575, + "learning_rate": 3.6465518236650584e-07, + "loss": 0.6816, + "step": 6741 + }, + { + "epoch": 0.8106775686887513, + "grad_norm": 1.836392220533042, + "learning_rate": 3.642068650639558e-07, + "loss": 0.7859, + "step": 6742 + }, + { + "epoch": 0.8107978115793903, + "grad_norm": 1.9324991229648747, + "learning_rate": 3.6375879590992334e-07, + "loss": 0.649, + "step": 6743 + }, + { + "epoch": 0.8109180544700295, + "grad_norm": 2.745522321608573, + "learning_rate": 3.6331097497238173e-07, + "loss": 0.8148, + "step": 6744 + }, + { + "epoch": 0.8110382973606686, + "grad_norm": 2.1021669787679365, + "learning_rate": 3.628634023192627e-07, + "loss": 0.802, + "step": 6745 + }, + { + "epoch": 0.8111585402513076, + "grad_norm": 2.0260741975684735, + "learning_rate": 3.624160780184644e-07, + "loss": 0.753, + "step": 6746 + }, + { + "epoch": 0.8112787831419467, + "grad_norm": 2.4222772079705117, + "learning_rate": 3.6196900213784496e-07, + "loss": 0.7507, + "step": 6747 + }, + { + "epoch": 0.8113990260325858, + "grad_norm": 1.9956195066372238, + "learning_rate": 3.6152217474522527e-07, + "loss": 0.8695, + "step": 6748 + }, + { + "epoch": 0.8115192689232249, + "grad_norm": 1.7717033343439688, + "learning_rate": 3.6107559590838975e-07, + "loss": 0.7312, + "step": 6749 + }, + { + "epoch": 0.811639511813864, + "grad_norm": 2.3805973950921504, + "learning_rate": 3.606292656950822e-07, + "loss": 0.6652, + "step": 6750 + }, + { + "epoch": 0.8117597547045031, + "grad_norm": 2.0239438634784825, + "learning_rate": 3.601831841730121e-07, + "loss": 0.8686, + "step": 6751 + }, + { + "epoch": 0.8118799975951422, + "grad_norm": 1.730835197691387, + "learning_rate": 3.5973735140984916e-07, + "loss": 0.7307, + "step": 6752 + }, + { + "epoch": 0.8120002404857812, + "grad_norm": 2.0237706915284766, + "learning_rate": 3.5929176747322607e-07, + "loss": 0.7974, + "step": 6753 + }, + { + "epoch": 0.8121204833764204, + "grad_norm": 0.834473477670657, + "learning_rate": 3.588464324307372e-07, + "loss": 0.5621, + "step": 6754 + }, + { + "epoch": 0.8122407262670595, + "grad_norm": 2.127894393817281, + "learning_rate": 3.584013463499391e-07, + "loss": 0.7583, + "step": 6755 + }, + { + "epoch": 0.8123609691576985, + "grad_norm": 0.744140865295852, + "learning_rate": 3.579565092983521e-07, + "loss": 0.6447, + "step": 6756 + }, + { + "epoch": 0.8124812120483377, + "grad_norm": 2.0331910469216785, + "learning_rate": 3.575119213434565e-07, + "loss": 0.8445, + "step": 6757 + }, + { + "epoch": 0.8126014549389767, + "grad_norm": 1.8286179750023204, + "learning_rate": 3.5706758255269765e-07, + "loss": 0.824, + "step": 6758 + }, + { + "epoch": 0.8127216978296158, + "grad_norm": 2.116337116248689, + "learning_rate": 3.566234929934795e-07, + "loss": 0.6986, + "step": 6759 + }, + { + "epoch": 0.812841940720255, + "grad_norm": 1.5276333663582726, + "learning_rate": 3.561796527331706e-07, + "loss": 0.7161, + "step": 6760 + }, + { + "epoch": 0.812962183610894, + "grad_norm": 1.9081981272324522, + "learning_rate": 3.5573606183910163e-07, + "loss": 0.7815, + "step": 6761 + }, + { + "epoch": 0.8130824265015331, + "grad_norm": 1.7730360206865103, + "learning_rate": 3.5529272037856493e-07, + "loss": 0.7853, + "step": 6762 + }, + { + "epoch": 0.8132026693921722, + "grad_norm": 0.7267144977360693, + "learning_rate": 3.548496284188149e-07, + "loss": 0.539, + "step": 6763 + }, + { + "epoch": 0.8133229122828113, + "grad_norm": 1.783864896050553, + "learning_rate": 3.544067860270681e-07, + "loss": 0.7924, + "step": 6764 + }, + { + "epoch": 0.8134431551734503, + "grad_norm": 2.087646127485973, + "learning_rate": 3.539641932705029e-07, + "loss": 0.7125, + "step": 6765 + }, + { + "epoch": 0.8135633980640895, + "grad_norm": 2.337813832639965, + "learning_rate": 3.53521850216262e-07, + "loss": 0.7727, + "step": 6766 + }, + { + "epoch": 0.8136836409547286, + "grad_norm": 1.8424276362045453, + "learning_rate": 3.530797569314461e-07, + "loss": 0.7693, + "step": 6767 + }, + { + "epoch": 0.8138038838453676, + "grad_norm": 1.9822397232934308, + "learning_rate": 3.5263791348312235e-07, + "loss": 0.7792, + "step": 6768 + }, + { + "epoch": 0.8139241267360068, + "grad_norm": 2.137279485738515, + "learning_rate": 3.521963199383171e-07, + "loss": 0.7076, + "step": 6769 + }, + { + "epoch": 0.8140443696266458, + "grad_norm": 2.047402233915013, + "learning_rate": 3.517549763640197e-07, + "loss": 0.7662, + "step": 6770 + }, + { + "epoch": 0.8141646125172849, + "grad_norm": 1.8288935162648068, + "learning_rate": 3.513138828271829e-07, + "loss": 0.7171, + "step": 6771 + }, + { + "epoch": 0.8142848554079241, + "grad_norm": 1.9428918324006248, + "learning_rate": 3.508730393947179e-07, + "loss": 0.7011, + "step": 6772 + }, + { + "epoch": 0.8144050982985631, + "grad_norm": 2.029943779791694, + "learning_rate": 3.504324461335024e-07, + "loss": 0.7211, + "step": 6773 + }, + { + "epoch": 0.8145253411892022, + "grad_norm": 1.7570089580500758, + "learning_rate": 3.499921031103732e-07, + "loss": 0.8846, + "step": 6774 + }, + { + "epoch": 0.8146455840798413, + "grad_norm": 1.7880933377057067, + "learning_rate": 3.4955201039212987e-07, + "loss": 0.7843, + "step": 6775 + }, + { + "epoch": 0.8147658269704804, + "grad_norm": 1.9456340723166252, + "learning_rate": 3.4911216804553465e-07, + "loss": 0.6553, + "step": 6776 + }, + { + "epoch": 0.8148860698611194, + "grad_norm": 2.15185762928542, + "learning_rate": 3.4867257613731017e-07, + "loss": 0.7142, + "step": 6777 + }, + { + "epoch": 0.8150063127517585, + "grad_norm": 2.7939626352519173, + "learning_rate": 3.4823323473414343e-07, + "loss": 0.8594, + "step": 6778 + }, + { + "epoch": 0.8151265556423977, + "grad_norm": 2.03407513683553, + "learning_rate": 3.477941439026812e-07, + "loss": 0.7604, + "step": 6779 + }, + { + "epoch": 0.8152467985330367, + "grad_norm": 2.278265476997621, + "learning_rate": 3.473553037095349e-07, + "loss": 0.7335, + "step": 6780 + }, + { + "epoch": 0.8153670414236758, + "grad_norm": 1.8262641108266038, + "learning_rate": 3.469167142212743e-07, + "loss": 0.8363, + "step": 6781 + }, + { + "epoch": 0.8154872843143149, + "grad_norm": 2.3986467400649616, + "learning_rate": 3.4647837550443337e-07, + "loss": 0.6331, + "step": 6782 + }, + { + "epoch": 0.815607527204954, + "grad_norm": 3.9999406333332534, + "learning_rate": 3.460402876255086e-07, + "loss": 0.7471, + "step": 6783 + }, + { + "epoch": 0.815727770095593, + "grad_norm": 2.158239138203489, + "learning_rate": 3.456024506509574e-07, + "loss": 0.7182, + "step": 6784 + }, + { + "epoch": 0.8158480129862322, + "grad_norm": 1.6488908623366243, + "learning_rate": 3.4516486464719873e-07, + "loss": 0.7483, + "step": 6785 + }, + { + "epoch": 0.8159682558768713, + "grad_norm": 1.6981029135314054, + "learning_rate": 3.4472752968061445e-07, + "loss": 0.6263, + "step": 6786 + }, + { + "epoch": 0.8160884987675103, + "grad_norm": 1.98384603824354, + "learning_rate": 3.442904458175475e-07, + "loss": 0.7372, + "step": 6787 + }, + { + "epoch": 0.8162087416581495, + "grad_norm": 1.5317604615414735, + "learning_rate": 3.438536131243044e-07, + "loss": 0.7605, + "step": 6788 + }, + { + "epoch": 0.8163289845487885, + "grad_norm": 2.163780992838009, + "learning_rate": 3.434170316671503e-07, + "loss": 0.6245, + "step": 6789 + }, + { + "epoch": 0.8164492274394276, + "grad_norm": 2.348364800846326, + "learning_rate": 3.4298070151231583e-07, + "loss": 0.9029, + "step": 6790 + }, + { + "epoch": 0.8165694703300668, + "grad_norm": 2.3990164688581825, + "learning_rate": 3.425446227259916e-07, + "loss": 0.6065, + "step": 6791 + }, + { + "epoch": 0.8166897132207058, + "grad_norm": 1.9701180322016938, + "learning_rate": 3.421087953743296e-07, + "loss": 0.8254, + "step": 6792 + }, + { + "epoch": 0.8168099561113449, + "grad_norm": 2.5702791066884734, + "learning_rate": 3.416732195234464e-07, + "loss": 0.8043, + "step": 6793 + }, + { + "epoch": 0.816930199001984, + "grad_norm": 1.5710375426266576, + "learning_rate": 3.4123789523941613e-07, + "loss": 0.7974, + "step": 6794 + }, + { + "epoch": 0.8170504418926231, + "grad_norm": 1.6081522027571475, + "learning_rate": 3.4080282258827884e-07, + "loss": 0.6405, + "step": 6795 + }, + { + "epoch": 0.8171706847832622, + "grad_norm": 2.1650696762184953, + "learning_rate": 3.403680016360342e-07, + "loss": 0.7311, + "step": 6796 + }, + { + "epoch": 0.8172909276739013, + "grad_norm": 1.878881125173101, + "learning_rate": 3.3993343244864403e-07, + "loss": 0.6776, + "step": 6797 + }, + { + "epoch": 0.8174111705645404, + "grad_norm": 1.8613117933807717, + "learning_rate": 3.394991150920323e-07, + "loss": 0.7317, + "step": 6798 + }, + { + "epoch": 0.8175314134551794, + "grad_norm": 2.1244931458012277, + "learning_rate": 3.3906504963208396e-07, + "loss": 0.7473, + "step": 6799 + }, + { + "epoch": 0.8176516563458186, + "grad_norm": 1.8413061983087375, + "learning_rate": 3.3863123613464774e-07, + "loss": 0.6665, + "step": 6800 + }, + { + "epoch": 0.8177718992364577, + "grad_norm": 1.800623888931011, + "learning_rate": 3.381976746655317e-07, + "loss": 0.7528, + "step": 6801 + }, + { + "epoch": 0.8178921421270967, + "grad_norm": 2.0844941020835472, + "learning_rate": 3.3776436529050756e-07, + "loss": 0.6785, + "step": 6802 + }, + { + "epoch": 0.8180123850177359, + "grad_norm": 1.6869549047506371, + "learning_rate": 3.373313080753073e-07, + "loss": 0.7293, + "step": 6803 + }, + { + "epoch": 0.8181326279083749, + "grad_norm": 1.5893339422265322, + "learning_rate": 3.3689850308562527e-07, + "loss": 0.7818, + "step": 6804 + }, + { + "epoch": 0.818252870799014, + "grad_norm": 1.878033282536532, + "learning_rate": 3.364659503871183e-07, + "loss": 0.7827, + "step": 6805 + }, + { + "epoch": 0.8183731136896532, + "grad_norm": 1.915464673948537, + "learning_rate": 3.3603365004540417e-07, + "loss": 0.8421, + "step": 6806 + }, + { + "epoch": 0.8184933565802922, + "grad_norm": 2.695092764480524, + "learning_rate": 3.356016021260624e-07, + "loss": 0.7733, + "step": 6807 + }, + { + "epoch": 0.8186135994709313, + "grad_norm": 2.3868831696511594, + "learning_rate": 3.35169806694634e-07, + "loss": 0.6539, + "step": 6808 + }, + { + "epoch": 0.8187338423615703, + "grad_norm": 0.7327086434044144, + "learning_rate": 3.3473826381662186e-07, + "loss": 0.607, + "step": 6809 + }, + { + "epoch": 0.8188540852522095, + "grad_norm": 1.950787541719617, + "learning_rate": 3.3430697355749216e-07, + "loss": 0.8171, + "step": 6810 + }, + { + "epoch": 0.8189743281428485, + "grad_norm": 1.960063359540473, + "learning_rate": 3.3387593598266907e-07, + "loss": 0.7578, + "step": 6811 + }, + { + "epoch": 0.8190945710334876, + "grad_norm": 1.8395963530837554, + "learning_rate": 3.3344515115754225e-07, + "loss": 0.7853, + "step": 6812 + }, + { + "epoch": 0.8192148139241268, + "grad_norm": 1.922993156369351, + "learning_rate": 3.33014619147461e-07, + "loss": 0.8001, + "step": 6813 + }, + { + "epoch": 0.8193350568147658, + "grad_norm": 1.8438963104644759, + "learning_rate": 3.325843400177362e-07, + "loss": 0.7213, + "step": 6814 + }, + { + "epoch": 0.8194552997054049, + "grad_norm": 2.586454126664693, + "learning_rate": 3.32154313833642e-07, + "loss": 0.7383, + "step": 6815 + }, + { + "epoch": 0.819575542596044, + "grad_norm": 2.1603978385828313, + "learning_rate": 3.3172454066041164e-07, + "loss": 0.6034, + "step": 6816 + }, + { + "epoch": 0.8196957854866831, + "grad_norm": 2.035521022447252, + "learning_rate": 3.3129502056324234e-07, + "loss": 0.7663, + "step": 6817 + }, + { + "epoch": 0.8198160283773221, + "grad_norm": 0.8209386706334811, + "learning_rate": 3.3086575360729165e-07, + "loss": 0.5994, + "step": 6818 + }, + { + "epoch": 0.8199362712679613, + "grad_norm": 2.7811352138191996, + "learning_rate": 3.3043673985767906e-07, + "loss": 0.7165, + "step": 6819 + }, + { + "epoch": 0.8200565141586004, + "grad_norm": 1.6649148396888467, + "learning_rate": 3.3000797937948564e-07, + "loss": 0.7729, + "step": 6820 + }, + { + "epoch": 0.8201767570492394, + "grad_norm": 0.9463849565322084, + "learning_rate": 3.295794722377534e-07, + "loss": 0.6513, + "step": 6821 + }, + { + "epoch": 0.8202969999398786, + "grad_norm": 2.1955950680719556, + "learning_rate": 3.291512184974876e-07, + "loss": 0.802, + "step": 6822 + }, + { + "epoch": 0.8204172428305176, + "grad_norm": 1.6972627453181484, + "learning_rate": 3.2872321822365346e-07, + "loss": 0.6676, + "step": 6823 + }, + { + "epoch": 0.8205374857211567, + "grad_norm": 1.9807350837350983, + "learning_rate": 3.282954714811783e-07, + "loss": 0.7332, + "step": 6824 + }, + { + "epoch": 0.8206577286117959, + "grad_norm": 2.22903176966087, + "learning_rate": 3.2786797833495093e-07, + "loss": 0.71, + "step": 6825 + }, + { + "epoch": 0.8207779715024349, + "grad_norm": 2.180099393696086, + "learning_rate": 3.274407388498213e-07, + "loss": 0.7332, + "step": 6826 + }, + { + "epoch": 0.820898214393074, + "grad_norm": 1.8559960188329683, + "learning_rate": 3.270137530906021e-07, + "loss": 0.7427, + "step": 6827 + }, + { + "epoch": 0.8210184572837131, + "grad_norm": 1.801158211862452, + "learning_rate": 3.265870211220665e-07, + "loss": 0.8378, + "step": 6828 + }, + { + "epoch": 0.8211387001743522, + "grad_norm": 2.081852488331632, + "learning_rate": 3.2616054300894934e-07, + "loss": 0.8235, + "step": 6829 + }, + { + "epoch": 0.8212589430649913, + "grad_norm": 2.5143175222193124, + "learning_rate": 3.2573431881594693e-07, + "loss": 0.846, + "step": 6830 + }, + { + "epoch": 0.8213791859556304, + "grad_norm": 2.146506256464049, + "learning_rate": 3.2530834860771663e-07, + "loss": 0.6609, + "step": 6831 + }, + { + "epoch": 0.8214994288462695, + "grad_norm": 2.1248873512478643, + "learning_rate": 3.248826324488794e-07, + "loss": 0.7451, + "step": 6832 + }, + { + "epoch": 0.8216196717369085, + "grad_norm": 1.843929993200452, + "learning_rate": 3.244571704040138e-07, + "loss": 0.8783, + "step": 6833 + }, + { + "epoch": 0.8217399146275477, + "grad_norm": 1.8609914725393715, + "learning_rate": 3.2403196253766374e-07, + "loss": 0.7407, + "step": 6834 + }, + { + "epoch": 0.8218601575181868, + "grad_norm": 2.358904254763658, + "learning_rate": 3.2360700891433254e-07, + "loss": 0.7952, + "step": 6835 + }, + { + "epoch": 0.8219804004088258, + "grad_norm": 0.801498612800678, + "learning_rate": 3.231823095984847e-07, + "loss": 0.5736, + "step": 6836 + }, + { + "epoch": 0.822100643299465, + "grad_norm": 1.9942789506153424, + "learning_rate": 3.2275786465454814e-07, + "loss": 0.7626, + "step": 6837 + }, + { + "epoch": 0.822220886190104, + "grad_norm": 1.8971377346327494, + "learning_rate": 3.2233367414690917e-07, + "loss": 0.7613, + "step": 6838 + }, + { + "epoch": 0.8223411290807431, + "grad_norm": 2.2805517512837112, + "learning_rate": 3.219097381399183e-07, + "loss": 0.8534, + "step": 6839 + }, + { + "epoch": 0.8224613719713821, + "grad_norm": 2.1454866987325922, + "learning_rate": 3.2148605669788584e-07, + "loss": 0.8128, + "step": 6840 + }, + { + "epoch": 0.8225816148620213, + "grad_norm": 2.2425362451998754, + "learning_rate": 3.2106262988508405e-07, + "loss": 0.7773, + "step": 6841 + }, + { + "epoch": 0.8227018577526604, + "grad_norm": 2.006412478133464, + "learning_rate": 3.206394577657465e-07, + "loss": 0.7489, + "step": 6842 + }, + { + "epoch": 0.8228221006432994, + "grad_norm": 2.221064367915104, + "learning_rate": 3.202165404040675e-07, + "loss": 0.7228, + "step": 6843 + }, + { + "epoch": 0.8229423435339386, + "grad_norm": 1.8982629892022145, + "learning_rate": 3.1979387786420396e-07, + "loss": 0.7474, + "step": 6844 + }, + { + "epoch": 0.8230625864245776, + "grad_norm": 2.338491211299437, + "learning_rate": 3.1937147021027346e-07, + "loss": 0.8255, + "step": 6845 + }, + { + "epoch": 0.8231828293152167, + "grad_norm": 2.1254137421780857, + "learning_rate": 3.189493175063547e-07, + "loss": 0.7682, + "step": 6846 + }, + { + "epoch": 0.8233030722058559, + "grad_norm": 1.6948332922320861, + "learning_rate": 3.1852741981648776e-07, + "loss": 0.6756, + "step": 6847 + }, + { + "epoch": 0.8234233150964949, + "grad_norm": 1.9705299928981992, + "learning_rate": 3.1810577720467404e-07, + "loss": 0.703, + "step": 6848 + }, + { + "epoch": 0.823543557987134, + "grad_norm": 1.5740179118357533, + "learning_rate": 3.176843897348769e-07, + "loss": 0.564, + "step": 6849 + }, + { + "epoch": 0.8236638008777731, + "grad_norm": 4.635053687582009, + "learning_rate": 3.1726325747102034e-07, + "loss": 0.7614, + "step": 6850 + }, + { + "epoch": 0.8237840437684122, + "grad_norm": 1.8844752270817957, + "learning_rate": 3.1684238047698974e-07, + "loss": 0.6444, + "step": 6851 + }, + { + "epoch": 0.8239042866590512, + "grad_norm": 1.9513473966387365, + "learning_rate": 3.1642175881663155e-07, + "loss": 0.537, + "step": 6852 + }, + { + "epoch": 0.8240245295496904, + "grad_norm": 2.8074205094155404, + "learning_rate": 3.160013925537537e-07, + "loss": 0.8422, + "step": 6853 + }, + { + "epoch": 0.8241447724403295, + "grad_norm": 1.9051287120479825, + "learning_rate": 3.155812817521266e-07, + "loss": 0.7582, + "step": 6854 + }, + { + "epoch": 0.8242650153309685, + "grad_norm": 2.186795802528898, + "learning_rate": 3.151614264754787e-07, + "loss": 0.7848, + "step": 6855 + }, + { + "epoch": 0.8243852582216077, + "grad_norm": 2.1544500384696135, + "learning_rate": 3.147418267875035e-07, + "loss": 0.7998, + "step": 6856 + }, + { + "epoch": 0.8245055011122467, + "grad_norm": 2.312229140631482, + "learning_rate": 3.1432248275185315e-07, + "loss": 0.6563, + "step": 6857 + }, + { + "epoch": 0.8246257440028858, + "grad_norm": 2.968981924534552, + "learning_rate": 3.139033944321412e-07, + "loss": 0.7784, + "step": 6858 + }, + { + "epoch": 0.824745986893525, + "grad_norm": 1.6249945713832852, + "learning_rate": 3.1348456189194507e-07, + "loss": 0.7922, + "step": 6859 + }, + { + "epoch": 0.824866229784164, + "grad_norm": 1.6562625236757498, + "learning_rate": 3.1306598519479876e-07, + "loss": 0.8334, + "step": 6860 + }, + { + "epoch": 0.8249864726748031, + "grad_norm": 1.9316035345236742, + "learning_rate": 3.1264766440420177e-07, + "loss": 0.7877, + "step": 6861 + }, + { + "epoch": 0.8251067155654422, + "grad_norm": 3.333624938289701, + "learning_rate": 3.122295995836124e-07, + "loss": 0.6903, + "step": 6862 + }, + { + "epoch": 0.8252269584560813, + "grad_norm": 1.8260619431316696, + "learning_rate": 3.118117907964508e-07, + "loss": 0.7812, + "step": 6863 + }, + { + "epoch": 0.8253472013467203, + "grad_norm": 1.9197827867981128, + "learning_rate": 3.1139423810609856e-07, + "loss": 0.8046, + "step": 6864 + }, + { + "epoch": 0.8254674442373595, + "grad_norm": 1.99483556342236, + "learning_rate": 3.1097694157589714e-07, + "loss": 0.7577, + "step": 6865 + }, + { + "epoch": 0.8255876871279986, + "grad_norm": 3.4549221695829817, + "learning_rate": 3.105599012691511e-07, + "loss": 0.7667, + "step": 6866 + }, + { + "epoch": 0.8257079300186376, + "grad_norm": 1.5824923373213104, + "learning_rate": 3.101431172491249e-07, + "loss": 0.8228, + "step": 6867 + }, + { + "epoch": 0.8258281729092768, + "grad_norm": 2.862791761163505, + "learning_rate": 3.097265895790444e-07, + "loss": 0.7179, + "step": 6868 + }, + { + "epoch": 0.8259484157999158, + "grad_norm": 3.4279644576447956, + "learning_rate": 3.093103183220962e-07, + "loss": 0.8352, + "step": 6869 + }, + { + "epoch": 0.8260686586905549, + "grad_norm": 0.8302802344712302, + "learning_rate": 3.0889430354142796e-07, + "loss": 0.5939, + "step": 6870 + }, + { + "epoch": 0.826188901581194, + "grad_norm": 1.9072204370981385, + "learning_rate": 3.084785453001497e-07, + "loss": 0.7031, + "step": 6871 + }, + { + "epoch": 0.8263091444718331, + "grad_norm": 2.45342850478062, + "learning_rate": 3.080630436613314e-07, + "loss": 0.8231, + "step": 6872 + }, + { + "epoch": 0.8264293873624722, + "grad_norm": 1.9714920083100937, + "learning_rate": 3.076477986880039e-07, + "loss": 0.8608, + "step": 6873 + }, + { + "epoch": 0.8265496302531112, + "grad_norm": 2.091534595448696, + "learning_rate": 3.0723281044315986e-07, + "loss": 0.6958, + "step": 6874 + }, + { + "epoch": 0.8266698731437504, + "grad_norm": 2.1359413308790858, + "learning_rate": 3.068180789897521e-07, + "loss": 0.7692, + "step": 6875 + }, + { + "epoch": 0.8267901160343895, + "grad_norm": 1.5012848437583017, + "learning_rate": 3.064036043906966e-07, + "loss": 0.8161, + "step": 6876 + }, + { + "epoch": 0.8269103589250285, + "grad_norm": 1.9257434895432752, + "learning_rate": 3.059893867088668e-07, + "loss": 0.685, + "step": 6877 + }, + { + "epoch": 0.8270306018156677, + "grad_norm": 2.516236319810317, + "learning_rate": 3.055754260071004e-07, + "loss": 0.6695, + "step": 6878 + }, + { + "epoch": 0.8271508447063067, + "grad_norm": 2.2562622239053822, + "learning_rate": 3.051617223481948e-07, + "loss": 0.7357, + "step": 6879 + }, + { + "epoch": 0.8272710875969458, + "grad_norm": 2.390652226311721, + "learning_rate": 3.047482757949078e-07, + "loss": 0.7538, + "step": 6880 + }, + { + "epoch": 0.827391330487585, + "grad_norm": 2.1117487231747822, + "learning_rate": 3.043350864099605e-07, + "loss": 0.8583, + "step": 6881 + }, + { + "epoch": 0.827511573378224, + "grad_norm": 2.1234249560930887, + "learning_rate": 3.039221542560315e-07, + "loss": 0.8145, + "step": 6882 + }, + { + "epoch": 0.8276318162688631, + "grad_norm": 5.732927269565261, + "learning_rate": 3.0350947939576356e-07, + "loss": 0.7426, + "step": 6883 + }, + { + "epoch": 0.8277520591595022, + "grad_norm": 1.8268858793850224, + "learning_rate": 3.0309706189175876e-07, + "loss": 0.7282, + "step": 6884 + }, + { + "epoch": 0.8278723020501413, + "grad_norm": 0.7738173351662105, + "learning_rate": 3.0268490180658045e-07, + "loss": 0.5746, + "step": 6885 + }, + { + "epoch": 0.8279925449407803, + "grad_norm": 2.315437925043339, + "learning_rate": 3.0227299920275305e-07, + "loss": 0.7964, + "step": 6886 + }, + { + "epoch": 0.8281127878314195, + "grad_norm": 2.670619736658896, + "learning_rate": 3.018613541427613e-07, + "loss": 0.8599, + "step": 6887 + }, + { + "epoch": 0.8282330307220586, + "grad_norm": 1.6842114937145818, + "learning_rate": 3.0144996668905243e-07, + "loss": 0.74, + "step": 6888 + }, + { + "epoch": 0.8283532736126976, + "grad_norm": 2.1238051028229052, + "learning_rate": 3.010388369040331e-07, + "loss": 0.8219, + "step": 6889 + }, + { + "epoch": 0.8284735165033368, + "grad_norm": 1.5512554683173945, + "learning_rate": 3.0062796485007156e-07, + "loss": 0.8314, + "step": 6890 + }, + { + "epoch": 0.8285937593939758, + "grad_norm": 2.5428312999887814, + "learning_rate": 3.002173505894965e-07, + "loss": 0.6553, + "step": 6891 + }, + { + "epoch": 0.8287140022846149, + "grad_norm": 3.100612542946876, + "learning_rate": 2.998069941845973e-07, + "loss": 0.6275, + "step": 6892 + }, + { + "epoch": 0.8288342451752541, + "grad_norm": 0.7265997026004493, + "learning_rate": 2.993968956976258e-07, + "loss": 0.5748, + "step": 6893 + }, + { + "epoch": 0.8289544880658931, + "grad_norm": 2.032859282134874, + "learning_rate": 2.9898705519079313e-07, + "loss": 0.7087, + "step": 6894 + }, + { + "epoch": 0.8290747309565322, + "grad_norm": 1.8004365206277562, + "learning_rate": 2.985774727262715e-07, + "loss": 0.7464, + "step": 6895 + }, + { + "epoch": 0.8291949738471713, + "grad_norm": 1.8218287100562995, + "learning_rate": 2.981681483661949e-07, + "loss": 0.8217, + "step": 6896 + }, + { + "epoch": 0.8293152167378104, + "grad_norm": 2.6192325629953475, + "learning_rate": 2.9775908217265633e-07, + "loss": 0.7108, + "step": 6897 + }, + { + "epoch": 0.8294354596284494, + "grad_norm": 1.2578001791783013, + "learning_rate": 2.9735027420771253e-07, + "loss": 0.5035, + "step": 6898 + }, + { + "epoch": 0.8295557025190886, + "grad_norm": 1.937397308089252, + "learning_rate": 2.969417245333774e-07, + "loss": 0.7208, + "step": 6899 + }, + { + "epoch": 0.8296759454097277, + "grad_norm": 2.1397392500910306, + "learning_rate": 2.9653343321162915e-07, + "loss": 0.7865, + "step": 6900 + }, + { + "epoch": 0.8297961883003667, + "grad_norm": 2.807217192974258, + "learning_rate": 2.9612540030440446e-07, + "loss": 0.6478, + "step": 6901 + }, + { + "epoch": 0.8299164311910058, + "grad_norm": 0.8706275821600945, + "learning_rate": 2.9571762587360206e-07, + "loss": 0.6422, + "step": 6902 + }, + { + "epoch": 0.8300366740816449, + "grad_norm": 1.6858541445061201, + "learning_rate": 2.953101099810806e-07, + "loss": 0.7415, + "step": 6903 + }, + { + "epoch": 0.830156916972284, + "grad_norm": 2.3077461248627085, + "learning_rate": 2.9490285268865965e-07, + "loss": 0.8274, + "step": 6904 + }, + { + "epoch": 0.830277159862923, + "grad_norm": 2.378177825701939, + "learning_rate": 2.9449585405812085e-07, + "loss": 0.7914, + "step": 6905 + }, + { + "epoch": 0.8303974027535622, + "grad_norm": 1.6977301731156802, + "learning_rate": 2.940891141512043e-07, + "loss": 0.7427, + "step": 6906 + }, + { + "epoch": 0.8305176456442013, + "grad_norm": 2.0112715437015614, + "learning_rate": 2.9368263302961385e-07, + "loss": 0.7228, + "step": 6907 + }, + { + "epoch": 0.8306378885348403, + "grad_norm": 1.83856441782706, + "learning_rate": 2.9327641075501075e-07, + "loss": 0.8017, + "step": 6908 + }, + { + "epoch": 0.8307581314254795, + "grad_norm": 5.604285691520608, + "learning_rate": 2.9287044738901866e-07, + "loss": 0.6741, + "step": 6909 + }, + { + "epoch": 0.8308783743161186, + "grad_norm": 2.3955265235714416, + "learning_rate": 2.9246474299322274e-07, + "loss": 0.9137, + "step": 6910 + }, + { + "epoch": 0.8309986172067576, + "grad_norm": 1.0478442246037047, + "learning_rate": 2.920592976291678e-07, + "loss": 0.6319, + "step": 6911 + }, + { + "epoch": 0.8311188600973968, + "grad_norm": 2.07347424035483, + "learning_rate": 2.916541113583595e-07, + "loss": 0.8086, + "step": 6912 + }, + { + "epoch": 0.8312391029880358, + "grad_norm": 2.2004251372767767, + "learning_rate": 2.912491842422642e-07, + "loss": 0.6637, + "step": 6913 + }, + { + "epoch": 0.8313593458786749, + "grad_norm": 1.8029409436938733, + "learning_rate": 2.9084451634230857e-07, + "loss": 0.7145, + "step": 6914 + }, + { + "epoch": 0.831479588769314, + "grad_norm": 2.5133132263748683, + "learning_rate": 2.9044010771988125e-07, + "loss": 0.7179, + "step": 6915 + }, + { + "epoch": 0.8315998316599531, + "grad_norm": 1.9866644200476589, + "learning_rate": 2.900359584363303e-07, + "loss": 0.7253, + "step": 6916 + }, + { + "epoch": 0.8317200745505922, + "grad_norm": 2.048706982052174, + "learning_rate": 2.8963206855296494e-07, + "loss": 0.8469, + "step": 6917 + }, + { + "epoch": 0.8318403174412313, + "grad_norm": 1.6890235663486979, + "learning_rate": 2.892284381310548e-07, + "loss": 0.7716, + "step": 6918 + }, + { + "epoch": 0.8319605603318704, + "grad_norm": 2.2450433812697157, + "learning_rate": 2.888250672318302e-07, + "loss": 0.7277, + "step": 6919 + }, + { + "epoch": 0.8320808032225094, + "grad_norm": 1.6256054703926421, + "learning_rate": 2.884219559164831e-07, + "loss": 0.6901, + "step": 6920 + }, + { + "epoch": 0.8322010461131486, + "grad_norm": 1.8922774604561732, + "learning_rate": 2.880191042461635e-07, + "loss": 0.8176, + "step": 6921 + }, + { + "epoch": 0.8323212890037877, + "grad_norm": 1.8737259032753053, + "learning_rate": 2.876165122819849e-07, + "loss": 0.8049, + "step": 6922 + }, + { + "epoch": 0.8324415318944267, + "grad_norm": 1.774171181716776, + "learning_rate": 2.872141800850201e-07, + "loss": 0.7969, + "step": 6923 + }, + { + "epoch": 0.8325617747850659, + "grad_norm": 1.7719883985075044, + "learning_rate": 2.868121077163024e-07, + "loss": 0.7347, + "step": 6924 + }, + { + "epoch": 0.8326820176757049, + "grad_norm": 1.9611440602969576, + "learning_rate": 2.864102952368257e-07, + "loss": 0.7242, + "step": 6925 + }, + { + "epoch": 0.832802260566344, + "grad_norm": 1.3087956059333143, + "learning_rate": 2.860087427075444e-07, + "loss": 0.5958, + "step": 6926 + }, + { + "epoch": 0.8329225034569832, + "grad_norm": 2.5508048040403697, + "learning_rate": 2.856074501893744e-07, + "loss": 0.8659, + "step": 6927 + }, + { + "epoch": 0.8330427463476222, + "grad_norm": 1.9898795128773608, + "learning_rate": 2.8520641774319054e-07, + "loss": 0.8199, + "step": 6928 + }, + { + "epoch": 0.8331629892382613, + "grad_norm": 2.135582881754424, + "learning_rate": 2.848056454298309e-07, + "loss": 0.7622, + "step": 6929 + }, + { + "epoch": 0.8332832321289004, + "grad_norm": 1.9429067420037893, + "learning_rate": 2.844051333100905e-07, + "loss": 0.6519, + "step": 6930 + }, + { + "epoch": 0.8334034750195395, + "grad_norm": 1.9514864953346176, + "learning_rate": 2.840048814447269e-07, + "loss": 0.838, + "step": 6931 + }, + { + "epoch": 0.8335237179101785, + "grad_norm": 3.5043863331227256, + "learning_rate": 2.836048898944587e-07, + "loss": 0.7505, + "step": 6932 + }, + { + "epoch": 0.8336439608008177, + "grad_norm": 3.2020872342332662, + "learning_rate": 2.832051587199642e-07, + "loss": 0.727, + "step": 6933 + }, + { + "epoch": 0.8337642036914568, + "grad_norm": 0.8160523007826214, + "learning_rate": 2.828056879818821e-07, + "loss": 0.578, + "step": 6934 + }, + { + "epoch": 0.8338844465820958, + "grad_norm": 2.0800828244151726, + "learning_rate": 2.824064777408117e-07, + "loss": 0.8399, + "step": 6935 + }, + { + "epoch": 0.8340046894727349, + "grad_norm": 2.3879275593209384, + "learning_rate": 2.8200752805731263e-07, + "loss": 0.759, + "step": 6936 + }, + { + "epoch": 0.834124932363374, + "grad_norm": 1.5050088382854303, + "learning_rate": 2.8160883899190625e-07, + "loss": 0.8128, + "step": 6937 + }, + { + "epoch": 0.8342451752540131, + "grad_norm": 2.3158492500972567, + "learning_rate": 2.8121041060507234e-07, + "loss": 0.7404, + "step": 6938 + }, + { + "epoch": 0.8343654181446521, + "grad_norm": 1.682216213490785, + "learning_rate": 2.808122429572528e-07, + "loss": 0.7147, + "step": 6939 + }, + { + "epoch": 0.8344856610352913, + "grad_norm": 3.7048063042752, + "learning_rate": 2.804143361088489e-07, + "loss": 0.7655, + "step": 6940 + }, + { + "epoch": 0.8346059039259304, + "grad_norm": 2.3251228177025665, + "learning_rate": 2.8001669012022277e-07, + "loss": 0.7867, + "step": 6941 + }, + { + "epoch": 0.8347261468165694, + "grad_norm": 1.6719890805120596, + "learning_rate": 2.7961930505169795e-07, + "loss": 0.6965, + "step": 6942 + }, + { + "epoch": 0.8348463897072086, + "grad_norm": 2.3703902833431103, + "learning_rate": 2.792221809635558e-07, + "loss": 0.7635, + "step": 6943 + }, + { + "epoch": 0.8349666325978476, + "grad_norm": 1.8360861860088138, + "learning_rate": 2.788253179160411e-07, + "loss": 0.7509, + "step": 6944 + }, + { + "epoch": 0.8350868754884867, + "grad_norm": 3.403723803645993, + "learning_rate": 2.7842871596935725e-07, + "loss": 0.6535, + "step": 6945 + }, + { + "epoch": 0.8352071183791259, + "grad_norm": 1.8627299326859015, + "learning_rate": 2.780323751836682e-07, + "loss": 0.6976, + "step": 6946 + }, + { + "epoch": 0.8353273612697649, + "grad_norm": 1.5995602629298662, + "learning_rate": 2.7763629561909876e-07, + "loss": 0.7878, + "step": 6947 + }, + { + "epoch": 0.835447604160404, + "grad_norm": 2.044217080221505, + "learning_rate": 2.772404773357335e-07, + "loss": 0.775, + "step": 6948 + }, + { + "epoch": 0.8355678470510431, + "grad_norm": 1.9523633158337999, + "learning_rate": 2.7684492039361853e-07, + "loss": 0.7852, + "step": 6949 + }, + { + "epoch": 0.8356880899416822, + "grad_norm": 2.5951776826625657, + "learning_rate": 2.764496248527586e-07, + "loss": 0.8371, + "step": 6950 + }, + { + "epoch": 0.8358083328323213, + "grad_norm": 2.5602676157844932, + "learning_rate": 2.760545907731211e-07, + "loss": 0.7873, + "step": 6951 + }, + { + "epoch": 0.8359285757229604, + "grad_norm": 1.698509471670383, + "learning_rate": 2.75659818214631e-07, + "loss": 0.6817, + "step": 6952 + }, + { + "epoch": 0.8360488186135995, + "grad_norm": 1.8220397874694125, + "learning_rate": 2.752653072371749e-07, + "loss": 0.7842, + "step": 6953 + }, + { + "epoch": 0.8361690615042385, + "grad_norm": 1.9242092936163802, + "learning_rate": 2.7487105790060105e-07, + "loss": 0.7496, + "step": 6954 + }, + { + "epoch": 0.8362893043948777, + "grad_norm": 1.9845417583413523, + "learning_rate": 2.7447707026471587e-07, + "loss": 0.6968, + "step": 6955 + }, + { + "epoch": 0.8364095472855168, + "grad_norm": 2.3184656056182535, + "learning_rate": 2.740833443892874e-07, + "loss": 0.801, + "step": 6956 + }, + { + "epoch": 0.8365297901761558, + "grad_norm": 2.2983466383908753, + "learning_rate": 2.7368988033404327e-07, + "loss": 0.7993, + "step": 6957 + }, + { + "epoch": 0.836650033066795, + "grad_norm": 1.5798191758995834, + "learning_rate": 2.732966781586712e-07, + "loss": 0.8492, + "step": 6958 + }, + { + "epoch": 0.836770275957434, + "grad_norm": 1.7407688129399521, + "learning_rate": 2.729037379228205e-07, + "loss": 0.6677, + "step": 6959 + }, + { + "epoch": 0.8368905188480731, + "grad_norm": 2.6542685804058963, + "learning_rate": 2.725110596860998e-07, + "loss": 0.8059, + "step": 6960 + }, + { + "epoch": 0.8370107617387123, + "grad_norm": 1.9029220548012917, + "learning_rate": 2.7211864350807776e-07, + "loss": 0.7078, + "step": 6961 + }, + { + "epoch": 0.8371310046293513, + "grad_norm": 2.0653167910191463, + "learning_rate": 2.717264894482836e-07, + "loss": 0.7387, + "step": 6962 + }, + { + "epoch": 0.8372512475199904, + "grad_norm": 2.0054409404114883, + "learning_rate": 2.7133459756620646e-07, + "loss": 0.8153, + "step": 6963 + }, + { + "epoch": 0.8373714904106295, + "grad_norm": 1.6855091784738534, + "learning_rate": 2.7094296792129733e-07, + "loss": 0.7427, + "step": 6964 + }, + { + "epoch": 0.8374917333012686, + "grad_norm": 1.7823655500171505, + "learning_rate": 2.7055160057296424e-07, + "loss": 0.7578, + "step": 6965 + }, + { + "epoch": 0.8376119761919076, + "grad_norm": 1.694189728246701, + "learning_rate": 2.7016049558057896e-07, + "loss": 0.7217, + "step": 6966 + }, + { + "epoch": 0.8377322190825467, + "grad_norm": 1.7028343582640593, + "learning_rate": 2.6976965300347074e-07, + "loss": 0.7121, + "step": 6967 + }, + { + "epoch": 0.8378524619731859, + "grad_norm": 3.0725176708436495, + "learning_rate": 2.693790729009309e-07, + "loss": 0.7025, + "step": 6968 + }, + { + "epoch": 0.8379727048638249, + "grad_norm": 1.7928492022117581, + "learning_rate": 2.6898875533220946e-07, + "loss": 0.8871, + "step": 6969 + }, + { + "epoch": 0.838092947754464, + "grad_norm": 1.771182268956968, + "learning_rate": 2.685987003565171e-07, + "loss": 0.8213, + "step": 6970 + }, + { + "epoch": 0.8382131906451031, + "grad_norm": 2.4957805311660146, + "learning_rate": 2.6820890803302566e-07, + "loss": 0.7618, + "step": 6971 + }, + { + "epoch": 0.8383334335357422, + "grad_norm": 2.149311567581251, + "learning_rate": 2.6781937842086557e-07, + "loss": 0.819, + "step": 6972 + }, + { + "epoch": 0.8384536764263812, + "grad_norm": 1.8707918628381082, + "learning_rate": 2.6743011157912933e-07, + "loss": 0.678, + "step": 6973 + }, + { + "epoch": 0.8385739193170204, + "grad_norm": 2.35169440118164, + "learning_rate": 2.6704110756686725e-07, + "loss": 0.656, + "step": 6974 + }, + { + "epoch": 0.8386941622076595, + "grad_norm": 1.8781133870354494, + "learning_rate": 2.6665236644309085e-07, + "loss": 0.8434, + "step": 6975 + }, + { + "epoch": 0.8388144050982985, + "grad_norm": 2.221741392515354, + "learning_rate": 2.662638882667727e-07, + "loss": 0.7988, + "step": 6976 + }, + { + "epoch": 0.8389346479889377, + "grad_norm": 2.0456192008127725, + "learning_rate": 2.658756730968443e-07, + "loss": 0.734, + "step": 6977 + }, + { + "epoch": 0.8390548908795767, + "grad_norm": 2.1486376859860874, + "learning_rate": 2.654877209921975e-07, + "loss": 0.8825, + "step": 6978 + }, + { + "epoch": 0.8391751337702158, + "grad_norm": 1.9969324190245206, + "learning_rate": 2.651000320116843e-07, + "loss": 0.63, + "step": 6979 + }, + { + "epoch": 0.839295376660855, + "grad_norm": 2.053042373271968, + "learning_rate": 2.647126062141163e-07, + "loss": 0.7624, + "step": 6980 + }, + { + "epoch": 0.839415619551494, + "grad_norm": 1.9475104848557419, + "learning_rate": 2.643254436582669e-07, + "loss": 0.8402, + "step": 6981 + }, + { + "epoch": 0.8395358624421331, + "grad_norm": 1.9416782487952227, + "learning_rate": 2.6393854440286743e-07, + "loss": 0.8291, + "step": 6982 + }, + { + "epoch": 0.8396561053327722, + "grad_norm": 2.2843240379984167, + "learning_rate": 2.6355190850661045e-07, + "loss": 0.7123, + "step": 6983 + }, + { + "epoch": 0.8397763482234113, + "grad_norm": 1.6721618976604222, + "learning_rate": 2.631655360281486e-07, + "loss": 0.8673, + "step": 6984 + }, + { + "epoch": 0.8398965911140504, + "grad_norm": 2.4575535351033904, + "learning_rate": 2.6277942702609323e-07, + "loss": 0.6605, + "step": 6985 + }, + { + "epoch": 0.8400168340046895, + "grad_norm": 2.225119353961441, + "learning_rate": 2.623935815590186e-07, + "loss": 0.8769, + "step": 6986 + }, + { + "epoch": 0.8401370768953286, + "grad_norm": 1.718062107104871, + "learning_rate": 2.6200799968545516e-07, + "loss": 0.8096, + "step": 6987 + }, + { + "epoch": 0.8402573197859676, + "grad_norm": 0.7977956801306381, + "learning_rate": 2.616226814638969e-07, + "loss": 0.5655, + "step": 6988 + }, + { + "epoch": 0.8403775626766068, + "grad_norm": 2.014594117549631, + "learning_rate": 2.612376269527954e-07, + "loss": 0.7817, + "step": 6989 + }, + { + "epoch": 0.8404978055672458, + "grad_norm": 1.7630885311710665, + "learning_rate": 2.608528362105635e-07, + "loss": 0.6818, + "step": 6990 + }, + { + "epoch": 0.8406180484578849, + "grad_norm": 1.8317156359183921, + "learning_rate": 2.6046830929557374e-07, + "loss": 0.7362, + "step": 6991 + }, + { + "epoch": 0.8407382913485241, + "grad_norm": 2.4825148426238375, + "learning_rate": 2.6008404626615776e-07, + "loss": 0.8486, + "step": 6992 + }, + { + "epoch": 0.8408585342391631, + "grad_norm": 2.6010675732767794, + "learning_rate": 2.597000471806092e-07, + "loss": 0.7398, + "step": 6993 + }, + { + "epoch": 0.8409787771298022, + "grad_norm": 1.903336972613399, + "learning_rate": 2.593163120971793e-07, + "loss": 0.7324, + "step": 6994 + }, + { + "epoch": 0.8410990200204413, + "grad_norm": 2.109867292813966, + "learning_rate": 2.5893284107408165e-07, + "loss": 0.6906, + "step": 6995 + }, + { + "epoch": 0.8412192629110804, + "grad_norm": 1.947096837418226, + "learning_rate": 2.5854963416948726e-07, + "loss": 0.7835, + "step": 6996 + }, + { + "epoch": 0.8413395058017195, + "grad_norm": 2.058410983866783, + "learning_rate": 2.5816669144152816e-07, + "loss": 0.6972, + "step": 6997 + }, + { + "epoch": 0.8414597486923585, + "grad_norm": 0.9115925862575354, + "learning_rate": 2.5778401294829777e-07, + "loss": 0.6636, + "step": 6998 + }, + { + "epoch": 0.8415799915829977, + "grad_norm": 1.7217592425474257, + "learning_rate": 2.574015987478473e-07, + "loss": 0.6497, + "step": 6999 + }, + { + "epoch": 0.8417002344736367, + "grad_norm": 2.0104860547466044, + "learning_rate": 2.570194488981887e-07, + "loss": 0.8742, + "step": 7000 + }, + { + "epoch": 0.8418204773642758, + "grad_norm": 0.8501194266961253, + "learning_rate": 2.566375634572939e-07, + "loss": 0.6035, + "step": 7001 + }, + { + "epoch": 0.841940720254915, + "grad_norm": 1.6951554045427626, + "learning_rate": 2.562559424830943e-07, + "loss": 0.7655, + "step": 7002 + }, + { + "epoch": 0.842060963145554, + "grad_norm": 2.3834364777589445, + "learning_rate": 2.5587458603348256e-07, + "loss": 0.7092, + "step": 7003 + }, + { + "epoch": 0.8421812060361931, + "grad_norm": 2.0290019580122824, + "learning_rate": 2.554934941663085e-07, + "loss": 0.8449, + "step": 7004 + }, + { + "epoch": 0.8423014489268322, + "grad_norm": 2.5305346196249925, + "learning_rate": 2.5511266693938484e-07, + "loss": 0.7391, + "step": 7005 + }, + { + "epoch": 0.8424216918174713, + "grad_norm": 1.549261390918704, + "learning_rate": 2.547321044104822e-07, + "loss": 0.7834, + "step": 7006 + }, + { + "epoch": 0.8425419347081103, + "grad_norm": 1.7989046658313714, + "learning_rate": 2.5435180663733113e-07, + "loss": 0.7733, + "step": 7007 + }, + { + "epoch": 0.8426621775987495, + "grad_norm": 2.7045573590557943, + "learning_rate": 2.539717736776241e-07, + "loss": 0.7257, + "step": 7008 + }, + { + "epoch": 0.8427824204893886, + "grad_norm": 1.61689210119593, + "learning_rate": 2.535920055890097e-07, + "loss": 0.766, + "step": 7009 + }, + { + "epoch": 0.8429026633800276, + "grad_norm": 2.519963852759374, + "learning_rate": 2.5321250242910006e-07, + "loss": 0.6491, + "step": 7010 + }, + { + "epoch": 0.8430229062706668, + "grad_norm": 2.0160603607541976, + "learning_rate": 2.5283326425546493e-07, + "loss": 0.8659, + "step": 7011 + }, + { + "epoch": 0.8431431491613058, + "grad_norm": 1.9572945606101342, + "learning_rate": 2.5245429112563443e-07, + "loss": 0.6992, + "step": 7012 + }, + { + "epoch": 0.8432633920519449, + "grad_norm": 1.7696519014771614, + "learning_rate": 2.5207558309709865e-07, + "loss": 0.8223, + "step": 7013 + }, + { + "epoch": 0.8433836349425841, + "grad_norm": 0.672055175930072, + "learning_rate": 2.516971402273065e-07, + "loss": 0.5634, + "step": 7014 + }, + { + "epoch": 0.8435038778332231, + "grad_norm": 4.274243712870922, + "learning_rate": 2.513189625736687e-07, + "loss": 0.6761, + "step": 7015 + }, + { + "epoch": 0.8436241207238622, + "grad_norm": 2.112521389283011, + "learning_rate": 2.509410501935534e-07, + "loss": 0.7176, + "step": 7016 + }, + { + "epoch": 0.8437443636145013, + "grad_norm": 2.2062350567103595, + "learning_rate": 2.5056340314429116e-07, + "loss": 0.7496, + "step": 7017 + }, + { + "epoch": 0.8438646065051404, + "grad_norm": 2.472238128649037, + "learning_rate": 2.5018602148316904e-07, + "loss": 0.8078, + "step": 7018 + }, + { + "epoch": 0.8439848493957794, + "grad_norm": 2.1014654995112307, + "learning_rate": 2.498089052674359e-07, + "loss": 0.8043, + "step": 7019 + }, + { + "epoch": 0.8441050922864186, + "grad_norm": 2.2648584581827196, + "learning_rate": 2.494320545543007e-07, + "loss": 0.7566, + "step": 7020 + }, + { + "epoch": 0.8442253351770577, + "grad_norm": 1.7182386157733922, + "learning_rate": 2.490554694009308e-07, + "loss": 0.6706, + "step": 7021 + }, + { + "epoch": 0.8443455780676967, + "grad_norm": 1.6800269205343386, + "learning_rate": 2.4867914986445426e-07, + "loss": 0.7852, + "step": 7022 + }, + { + "epoch": 0.8444658209583359, + "grad_norm": 1.8816878889569182, + "learning_rate": 2.483030960019581e-07, + "loss": 0.7133, + "step": 7023 + }, + { + "epoch": 0.8445860638489749, + "grad_norm": 0.7510561659035467, + "learning_rate": 2.479273078704891e-07, + "loss": 0.5549, + "step": 7024 + }, + { + "epoch": 0.844706306739614, + "grad_norm": 0.7986792166782087, + "learning_rate": 2.475517855270552e-07, + "loss": 0.6477, + "step": 7025 + }, + { + "epoch": 0.8448265496302532, + "grad_norm": 2.0243736664219107, + "learning_rate": 2.4717652902862143e-07, + "loss": 0.7328, + "step": 7026 + }, + { + "epoch": 0.8449467925208922, + "grad_norm": 2.85098051638119, + "learning_rate": 2.4680153843211495e-07, + "loss": 0.8181, + "step": 7027 + }, + { + "epoch": 0.8450670354115313, + "grad_norm": 1.6489011284293091, + "learning_rate": 2.464268137944212e-07, + "loss": 0.7252, + "step": 7028 + }, + { + "epoch": 0.8451872783021703, + "grad_norm": 2.0688677659181662, + "learning_rate": 2.46052355172385e-07, + "loss": 0.7836, + "step": 7029 + }, + { + "epoch": 0.8453075211928095, + "grad_norm": 2.3608882674228533, + "learning_rate": 2.456781626228128e-07, + "loss": 0.7476, + "step": 7030 + }, + { + "epoch": 0.8454277640834486, + "grad_norm": 0.9291225608199368, + "learning_rate": 2.453042362024675e-07, + "loss": 0.6642, + "step": 7031 + }, + { + "epoch": 0.8455480069740876, + "grad_norm": 1.4909932415436145, + "learning_rate": 2.449305759680751e-07, + "loss": 0.7368, + "step": 7032 + }, + { + "epoch": 0.8456682498647268, + "grad_norm": 1.555546049058718, + "learning_rate": 2.445571819763188e-07, + "loss": 0.7527, + "step": 7033 + }, + { + "epoch": 0.8457884927553658, + "grad_norm": 1.6706439325506457, + "learning_rate": 2.4418405428384227e-07, + "loss": 0.5863, + "step": 7034 + }, + { + "epoch": 0.8459087356460049, + "grad_norm": 2.147445838254102, + "learning_rate": 2.4381119294724864e-07, + "loss": 0.7222, + "step": 7035 + }, + { + "epoch": 0.846028978536644, + "grad_norm": 6.673633209767065, + "learning_rate": 2.434385980231004e-07, + "loss": 0.5478, + "step": 7036 + }, + { + "epoch": 0.8461492214272831, + "grad_norm": 1.6406103224324846, + "learning_rate": 2.4306626956792043e-07, + "loss": 0.6544, + "step": 7037 + }, + { + "epoch": 0.8462694643179222, + "grad_norm": 1.8294458426784148, + "learning_rate": 2.4269420763819017e-07, + "loss": 0.7608, + "step": 7038 + }, + { + "epoch": 0.8463897072085613, + "grad_norm": 3.033418803705132, + "learning_rate": 2.4232241229035223e-07, + "loss": 0.8423, + "step": 7039 + }, + { + "epoch": 0.8465099500992004, + "grad_norm": 0.7567305165363957, + "learning_rate": 2.419508835808064e-07, + "loss": 0.5678, + "step": 7040 + }, + { + "epoch": 0.8466301929898394, + "grad_norm": 1.9780235469407834, + "learning_rate": 2.415796215659134e-07, + "loss": 0.6323, + "step": 7041 + }, + { + "epoch": 0.8467504358804786, + "grad_norm": 2.126700226011734, + "learning_rate": 2.412086263019939e-07, + "loss": 0.7773, + "step": 7042 + }, + { + "epoch": 0.8468706787711177, + "grad_norm": 1.7715954729969359, + "learning_rate": 2.408378978453276e-07, + "loss": 0.8019, + "step": 7043 + }, + { + "epoch": 0.8469909216617567, + "grad_norm": 0.8259217805842346, + "learning_rate": 2.404674362521533e-07, + "loss": 0.6407, + "step": 7044 + }, + { + "epoch": 0.8471111645523959, + "grad_norm": 2.462164872500418, + "learning_rate": 2.4009724157866997e-07, + "loss": 0.7419, + "step": 7045 + }, + { + "epoch": 0.8472314074430349, + "grad_norm": 2.085616640709211, + "learning_rate": 2.3972731388103564e-07, + "loss": 0.7632, + "step": 7046 + }, + { + "epoch": 0.847351650333674, + "grad_norm": 0.8172831229577632, + "learning_rate": 2.393576532153687e-07, + "loss": 0.6242, + "step": 7047 + }, + { + "epoch": 0.8474718932243132, + "grad_norm": 0.9390487276359003, + "learning_rate": 2.389882596377453e-07, + "loss": 0.5785, + "step": 7048 + }, + { + "epoch": 0.8475921361149522, + "grad_norm": 1.8058702300626526, + "learning_rate": 2.386191332042031e-07, + "loss": 0.7706, + "step": 7049 + }, + { + "epoch": 0.8477123790055913, + "grad_norm": 1.8729030327164555, + "learning_rate": 2.3825027397073794e-07, + "loss": 0.7362, + "step": 7050 + }, + { + "epoch": 0.8478326218962304, + "grad_norm": 2.1289296560138213, + "learning_rate": 2.3788168199330515e-07, + "loss": 0.6699, + "step": 7051 + }, + { + "epoch": 0.8479528647868695, + "grad_norm": 1.5918987218184906, + "learning_rate": 2.3751335732782074e-07, + "loss": 0.7317, + "step": 7052 + }, + { + "epoch": 0.8480731076775085, + "grad_norm": 1.9660246013797973, + "learning_rate": 2.371453000301582e-07, + "loss": 0.7945, + "step": 7053 + }, + { + "epoch": 0.8481933505681477, + "grad_norm": 2.5035603919679006, + "learning_rate": 2.3677751015615222e-07, + "loss": 0.7449, + "step": 7054 + }, + { + "epoch": 0.8483135934587868, + "grad_norm": 1.7632787188104895, + "learning_rate": 2.3640998776159593e-07, + "loss": 0.8553, + "step": 7055 + }, + { + "epoch": 0.8484338363494258, + "grad_norm": 2.618188376280008, + "learning_rate": 2.3604273290224253e-07, + "loss": 0.8162, + "step": 7056 + }, + { + "epoch": 0.848554079240065, + "grad_norm": 1.8531814129172914, + "learning_rate": 2.356757456338039e-07, + "loss": 0.7513, + "step": 7057 + }, + { + "epoch": 0.848674322130704, + "grad_norm": 0.774310534413234, + "learning_rate": 2.3530902601195147e-07, + "loss": 0.5906, + "step": 7058 + }, + { + "epoch": 0.8487945650213431, + "grad_norm": 2.299370687884564, + "learning_rate": 2.34942574092317e-07, + "loss": 0.7875, + "step": 7059 + }, + { + "epoch": 0.8489148079119821, + "grad_norm": 2.0917706599226835, + "learning_rate": 2.3457638993049045e-07, + "loss": 0.7742, + "step": 7060 + }, + { + "epoch": 0.8490350508026213, + "grad_norm": 1.918675931314806, + "learning_rate": 2.3421047358202252e-07, + "loss": 0.6428, + "step": 7061 + }, + { + "epoch": 0.8491552936932604, + "grad_norm": 2.8142815775215, + "learning_rate": 2.3384482510242144e-07, + "loss": 0.8387, + "step": 7062 + }, + { + "epoch": 0.8492755365838994, + "grad_norm": 1.9713351518796536, + "learning_rate": 2.3347944454715575e-07, + "loss": 0.7753, + "step": 7063 + }, + { + "epoch": 0.8493957794745386, + "grad_norm": 1.973933948644701, + "learning_rate": 2.331143319716542e-07, + "loss": 0.678, + "step": 7064 + }, + { + "epoch": 0.8495160223651776, + "grad_norm": 2.2738419530225578, + "learning_rate": 2.3274948743130363e-07, + "loss": 0.6571, + "step": 7065 + }, + { + "epoch": 0.8496362652558167, + "grad_norm": 1.6938601834539644, + "learning_rate": 2.3238491098145085e-07, + "loss": 0.7981, + "step": 7066 + }, + { + "epoch": 0.8497565081464559, + "grad_norm": 2.1699873590760395, + "learning_rate": 2.3202060267740141e-07, + "loss": 0.7349, + "step": 7067 + }, + { + "epoch": 0.8498767510370949, + "grad_norm": 2.105350752940639, + "learning_rate": 2.3165656257442044e-07, + "loss": 0.7749, + "step": 7068 + }, + { + "epoch": 0.849996993927734, + "grad_norm": 1.9508011077126821, + "learning_rate": 2.31292790727734e-07, + "loss": 0.9048, + "step": 7069 + }, + { + "epoch": 0.8501172368183731, + "grad_norm": 2.1488246499543, + "learning_rate": 2.3092928719252392e-07, + "loss": 0.8061, + "step": 7070 + }, + { + "epoch": 0.8502374797090122, + "grad_norm": 2.072984461099082, + "learning_rate": 2.3056605202393475e-07, + "loss": 0.7853, + "step": 7071 + }, + { + "epoch": 0.8503577225996513, + "grad_norm": 2.1228587638388183, + "learning_rate": 2.3020308527706888e-07, + "loss": 0.67, + "step": 7072 + }, + { + "epoch": 0.8504779654902904, + "grad_norm": 1.8005394762961138, + "learning_rate": 2.2984038700698715e-07, + "loss": 0.8925, + "step": 7073 + }, + { + "epoch": 0.8505982083809295, + "grad_norm": 1.629923916544685, + "learning_rate": 2.2947795726871222e-07, + "loss": 0.7923, + "step": 7074 + }, + { + "epoch": 0.8507184512715685, + "grad_norm": 1.869369572536251, + "learning_rate": 2.2911579611722253e-07, + "loss": 0.864, + "step": 7075 + }, + { + "epoch": 0.8508386941622077, + "grad_norm": 1.773479041062548, + "learning_rate": 2.2875390360745905e-07, + "loss": 0.8768, + "step": 7076 + }, + { + "epoch": 0.8509589370528468, + "grad_norm": 1.7272243008144879, + "learning_rate": 2.2839227979432008e-07, + "loss": 0.7771, + "step": 7077 + }, + { + "epoch": 0.8510791799434858, + "grad_norm": 1.8515863215849586, + "learning_rate": 2.2803092473266373e-07, + "loss": 0.8532, + "step": 7078 + }, + { + "epoch": 0.851199422834125, + "grad_norm": 2.3899503266986355, + "learning_rate": 2.2766983847730724e-07, + "loss": 0.8641, + "step": 7079 + }, + { + "epoch": 0.851319665724764, + "grad_norm": 1.9397562487344626, + "learning_rate": 2.2730902108302663e-07, + "loss": 0.6711, + "step": 7080 + }, + { + "epoch": 0.8514399086154031, + "grad_norm": 1.647104818389109, + "learning_rate": 2.269484726045583e-07, + "loss": 0.6901, + "step": 7081 + }, + { + "epoch": 0.8515601515060423, + "grad_norm": 1.707322573736016, + "learning_rate": 2.2658819309659672e-07, + "loss": 0.7999, + "step": 7082 + }, + { + "epoch": 0.8516803943966813, + "grad_norm": 1.9019383932699907, + "learning_rate": 2.2622818261379706e-07, + "loss": 0.8563, + "step": 7083 + }, + { + "epoch": 0.8518006372873204, + "grad_norm": 1.995649613092154, + "learning_rate": 2.2586844121077142e-07, + "loss": 0.7517, + "step": 7084 + }, + { + "epoch": 0.8519208801779595, + "grad_norm": 1.9166245732315355, + "learning_rate": 2.2550896894209215e-07, + "loss": 0.7261, + "step": 7085 + }, + { + "epoch": 0.8520411230685986, + "grad_norm": 0.6939603108729374, + "learning_rate": 2.2514976586229184e-07, + "loss": 0.5661, + "step": 7086 + }, + { + "epoch": 0.8521613659592376, + "grad_norm": 0.7629539451748616, + "learning_rate": 2.247908320258609e-07, + "loss": 0.5465, + "step": 7087 + }, + { + "epoch": 0.8522816088498768, + "grad_norm": 2.14386987823645, + "learning_rate": 2.2443216748724914e-07, + "loss": 0.8031, + "step": 7088 + }, + { + "epoch": 0.8524018517405159, + "grad_norm": 1.9277848350158757, + "learning_rate": 2.2407377230086588e-07, + "loss": 0.7451, + "step": 7089 + }, + { + "epoch": 0.8525220946311549, + "grad_norm": 3.6980501965694916, + "learning_rate": 2.23715646521079e-07, + "loss": 0.8351, + "step": 7090 + }, + { + "epoch": 0.852642337521794, + "grad_norm": 1.9020386124110145, + "learning_rate": 2.2335779020221724e-07, + "loss": 0.8419, + "step": 7091 + }, + { + "epoch": 0.8527625804124331, + "grad_norm": 0.8088390328456506, + "learning_rate": 2.2300020339856497e-07, + "loss": 0.5642, + "step": 7092 + }, + { + "epoch": 0.8528828233030722, + "grad_norm": 2.280374750994749, + "learning_rate": 2.2264288616436966e-07, + "loss": 0.7798, + "step": 7093 + }, + { + "epoch": 0.8530030661937112, + "grad_norm": 1.7896248545805475, + "learning_rate": 2.222858385538351e-07, + "loss": 0.7282, + "step": 7094 + }, + { + "epoch": 0.8531233090843504, + "grad_norm": 1.637638025072305, + "learning_rate": 2.2192906062112527e-07, + "loss": 0.6815, + "step": 7095 + }, + { + "epoch": 0.8532435519749895, + "grad_norm": 1.7193143438291698, + "learning_rate": 2.2157255242036377e-07, + "loss": 0.7115, + "step": 7096 + }, + { + "epoch": 0.8533637948656285, + "grad_norm": 1.6386164404639483, + "learning_rate": 2.2121631400563135e-07, + "loss": 0.7483, + "step": 7097 + }, + { + "epoch": 0.8534840377562677, + "grad_norm": 0.8035839995657347, + "learning_rate": 2.208603454309701e-07, + "loss": 0.53, + "step": 7098 + }, + { + "epoch": 0.8536042806469067, + "grad_norm": 2.141223127794418, + "learning_rate": 2.2050464675037994e-07, + "loss": 0.7077, + "step": 7099 + }, + { + "epoch": 0.8537245235375458, + "grad_norm": 2.3022630256038297, + "learning_rate": 2.2014921801782016e-07, + "loss": 0.7294, + "step": 7100 + }, + { + "epoch": 0.853844766428185, + "grad_norm": 1.9304617745409691, + "learning_rate": 2.1979405928720872e-07, + "loss": 0.7413, + "step": 7101 + }, + { + "epoch": 0.853965009318824, + "grad_norm": 1.6717623378548816, + "learning_rate": 2.1943917061242257e-07, + "loss": 0.7958, + "step": 7102 + }, + { + "epoch": 0.8540852522094631, + "grad_norm": 1.6719524329961974, + "learning_rate": 2.1908455204729903e-07, + "loss": 0.6641, + "step": 7103 + }, + { + "epoch": 0.8542054951001022, + "grad_norm": 2.236866235750503, + "learning_rate": 2.1873020364563265e-07, + "loss": 0.7828, + "step": 7104 + }, + { + "epoch": 0.8543257379907413, + "grad_norm": 2.156409769082306, + "learning_rate": 2.183761254611789e-07, + "loss": 0.7627, + "step": 7105 + }, + { + "epoch": 0.8544459808813804, + "grad_norm": 1.9648964118346601, + "learning_rate": 2.1802231754764987e-07, + "loss": 0.7096, + "step": 7106 + }, + { + "epoch": 0.8545662237720195, + "grad_norm": 2.0514554323380123, + "learning_rate": 2.17668779958718e-07, + "loss": 0.7636, + "step": 7107 + }, + { + "epoch": 0.8546864666626586, + "grad_norm": 2.6243878513647463, + "learning_rate": 2.1731551274801553e-07, + "loss": 0.8092, + "step": 7108 + }, + { + "epoch": 0.8548067095532976, + "grad_norm": 3.0330529897345344, + "learning_rate": 2.169625159691324e-07, + "loss": 0.6178, + "step": 7109 + }, + { + "epoch": 0.8549269524439368, + "grad_norm": 5.418882924069812, + "learning_rate": 2.1660978967561784e-07, + "loss": 0.7437, + "step": 7110 + }, + { + "epoch": 0.8550471953345758, + "grad_norm": 2.3372807037788244, + "learning_rate": 2.1625733392098035e-07, + "loss": 0.7922, + "step": 7111 + }, + { + "epoch": 0.8551674382252149, + "grad_norm": 1.7519701358305286, + "learning_rate": 2.159051487586867e-07, + "loss": 0.7935, + "step": 7112 + }, + { + "epoch": 0.8552876811158541, + "grad_norm": 20.644522750116508, + "learning_rate": 2.155532342421642e-07, + "loss": 0.7284, + "step": 7113 + }, + { + "epoch": 0.8554079240064931, + "grad_norm": 1.7758688371827551, + "learning_rate": 2.1520159042479636e-07, + "loss": 0.7817, + "step": 7114 + }, + { + "epoch": 0.8555281668971322, + "grad_norm": 2.2507430015420415, + "learning_rate": 2.148502173599287e-07, + "loss": 0.7126, + "step": 7115 + }, + { + "epoch": 0.8556484097877713, + "grad_norm": 1.679683809498126, + "learning_rate": 2.1449911510086372e-07, + "loss": 0.6578, + "step": 7116 + }, + { + "epoch": 0.8557686526784104, + "grad_norm": 1.905257232299901, + "learning_rate": 2.141482837008628e-07, + "loss": 0.7689, + "step": 7117 + }, + { + "epoch": 0.8558888955690495, + "grad_norm": 1.9017861428314689, + "learning_rate": 2.1379772321314826e-07, + "loss": 0.7274, + "step": 7118 + }, + { + "epoch": 0.8560091384596886, + "grad_norm": 2.0860992254648645, + "learning_rate": 2.1344743369089802e-07, + "loss": 0.8187, + "step": 7119 + }, + { + "epoch": 0.8561293813503277, + "grad_norm": 1.7413300598098185, + "learning_rate": 2.130974151872522e-07, + "loss": 0.8229, + "step": 7120 + }, + { + "epoch": 0.8562496242409667, + "grad_norm": 1.8201006983052985, + "learning_rate": 2.1274766775530773e-07, + "loss": 0.7856, + "step": 7121 + }, + { + "epoch": 0.8563698671316058, + "grad_norm": 1.9970645820516806, + "learning_rate": 2.1239819144812077e-07, + "loss": 0.805, + "step": 7122 + }, + { + "epoch": 0.856490110022245, + "grad_norm": 2.0444420485666766, + "learning_rate": 2.1204898631870716e-07, + "loss": 0.7021, + "step": 7123 + }, + { + "epoch": 0.856610352912884, + "grad_norm": 1.8582581443572221, + "learning_rate": 2.1170005242004006e-07, + "loss": 0.7638, + "step": 7124 + }, + { + "epoch": 0.8567305958035231, + "grad_norm": 2.272568579248533, + "learning_rate": 2.1135138980505384e-07, + "loss": 0.7838, + "step": 7125 + }, + { + "epoch": 0.8568508386941622, + "grad_norm": 1.8061568318662133, + "learning_rate": 2.110029985266395e-07, + "loss": 0.7272, + "step": 7126 + }, + { + "epoch": 0.8569710815848013, + "grad_norm": 1.706532001596555, + "learning_rate": 2.1065487863764787e-07, + "loss": 0.7428, + "step": 7127 + }, + { + "epoch": 0.8570913244754403, + "grad_norm": 1.878440370914472, + "learning_rate": 2.1030703019088846e-07, + "loss": 0.859, + "step": 7128 + }, + { + "epoch": 0.8572115673660795, + "grad_norm": 1.9856272673481359, + "learning_rate": 2.099594532391291e-07, + "loss": 0.7145, + "step": 7129 + }, + { + "epoch": 0.8573318102567186, + "grad_norm": 1.5500614984679133, + "learning_rate": 2.0961214783509806e-07, + "loss": 0.7934, + "step": 7130 + }, + { + "epoch": 0.8574520531473576, + "grad_norm": 1.8095245931977413, + "learning_rate": 2.0926511403148051e-07, + "loss": 0.751, + "step": 7131 + }, + { + "epoch": 0.8575722960379968, + "grad_norm": 1.9501693945448517, + "learning_rate": 2.0891835188092143e-07, + "loss": 0.7599, + "step": 7132 + }, + { + "epoch": 0.8576925389286358, + "grad_norm": 1.8883709672756142, + "learning_rate": 2.0857186143602434e-07, + "loss": 0.8209, + "step": 7133 + }, + { + "epoch": 0.8578127818192749, + "grad_norm": 1.8184779668337985, + "learning_rate": 2.0822564274935094e-07, + "loss": 0.6731, + "step": 7134 + }, + { + "epoch": 0.8579330247099141, + "grad_norm": 1.709151010291254, + "learning_rate": 2.078796958734239e-07, + "loss": 0.6738, + "step": 7135 + }, + { + "epoch": 0.8580532676005531, + "grad_norm": 1.873811853977134, + "learning_rate": 2.0753402086072124e-07, + "loss": 0.7541, + "step": 7136 + }, + { + "epoch": 0.8581735104911922, + "grad_norm": 2.3324432264739308, + "learning_rate": 2.071886177636828e-07, + "loss": 0.7547, + "step": 7137 + }, + { + "epoch": 0.8582937533818313, + "grad_norm": 1.7757136992039362, + "learning_rate": 2.0684348663470575e-07, + "loss": 0.8352, + "step": 7138 + }, + { + "epoch": 0.8584139962724704, + "grad_norm": 1.9641246655321318, + "learning_rate": 2.0649862752614555e-07, + "loss": 0.618, + "step": 7139 + }, + { + "epoch": 0.8585342391631094, + "grad_norm": 0.7664695285879173, + "learning_rate": 2.0615404049031838e-07, + "loss": 0.5708, + "step": 7140 + }, + { + "epoch": 0.8586544820537486, + "grad_norm": 5.090091819309737, + "learning_rate": 2.0580972557949616e-07, + "loss": 0.7792, + "step": 7141 + }, + { + "epoch": 0.8587747249443877, + "grad_norm": 0.8089536520126377, + "learning_rate": 2.054656828459125e-07, + "loss": 0.5422, + "step": 7142 + }, + { + "epoch": 0.8588949678350267, + "grad_norm": 1.8446821022633262, + "learning_rate": 2.051219123417578e-07, + "loss": 0.7757, + "step": 7143 + }, + { + "epoch": 0.8590152107256659, + "grad_norm": 1.98059635390815, + "learning_rate": 2.0477841411918196e-07, + "loss": 0.6077, + "step": 7144 + }, + { + "epoch": 0.859135453616305, + "grad_norm": 1.9493087717188367, + "learning_rate": 2.0443518823029326e-07, + "loss": 0.7531, + "step": 7145 + }, + { + "epoch": 0.859255696506944, + "grad_norm": 2.4023798746013605, + "learning_rate": 2.0409223472715854e-07, + "loss": 0.7666, + "step": 7146 + }, + { + "epoch": 0.8593759393975832, + "grad_norm": 1.8313516686244573, + "learning_rate": 2.0374955366180434e-07, + "loss": 0.7483, + "step": 7147 + }, + { + "epoch": 0.8594961822882222, + "grad_norm": 3.181033437736018, + "learning_rate": 2.034071450862147e-07, + "loss": 0.7285, + "step": 7148 + }, + { + "epoch": 0.8596164251788613, + "grad_norm": 1.8088100625316286, + "learning_rate": 2.030650090523327e-07, + "loss": 0.7697, + "step": 7149 + }, + { + "epoch": 0.8597366680695004, + "grad_norm": 2.211024198416075, + "learning_rate": 2.0272314561205995e-07, + "loss": 0.5986, + "step": 7150 + }, + { + "epoch": 0.8598569109601395, + "grad_norm": 1.8540890352377875, + "learning_rate": 2.023815548172567e-07, + "loss": 0.7324, + "step": 7151 + }, + { + "epoch": 0.8599771538507786, + "grad_norm": 1.6624590531662997, + "learning_rate": 2.0204023671974267e-07, + "loss": 0.6618, + "step": 7152 + }, + { + "epoch": 0.8600973967414177, + "grad_norm": 2.283433887838532, + "learning_rate": 2.0169919137129532e-07, + "loss": 0.8115, + "step": 7153 + }, + { + "epoch": 0.8602176396320568, + "grad_norm": 2.3609143218765265, + "learning_rate": 2.013584188236508e-07, + "loss": 0.7104, + "step": 7154 + }, + { + "epoch": 0.8603378825226958, + "grad_norm": 1.7811593819037914, + "learning_rate": 2.0101791912850396e-07, + "loss": 0.7995, + "step": 7155 + }, + { + "epoch": 0.8604581254133349, + "grad_norm": 2.509314256742869, + "learning_rate": 2.006776923375082e-07, + "loss": 0.6366, + "step": 7156 + }, + { + "epoch": 0.860578368303974, + "grad_norm": 1.713237383724904, + "learning_rate": 2.003377385022764e-07, + "loss": 0.7183, + "step": 7157 + }, + { + "epoch": 0.8606986111946131, + "grad_norm": 1.946194245918603, + "learning_rate": 1.9999805767437826e-07, + "loss": 0.7831, + "step": 7158 + }, + { + "epoch": 0.8608188540852522, + "grad_norm": 1.7817350697508054, + "learning_rate": 1.9965864990534386e-07, + "loss": 0.7184, + "step": 7159 + }, + { + "epoch": 0.8609390969758913, + "grad_norm": 1.8394974627673026, + "learning_rate": 1.9931951524666092e-07, + "loss": 0.7814, + "step": 7160 + }, + { + "epoch": 0.8610593398665304, + "grad_norm": 1.7768688196161313, + "learning_rate": 1.9898065374977534e-07, + "loss": 0.8124, + "step": 7161 + }, + { + "epoch": 0.8611795827571694, + "grad_norm": 1.9724105228682582, + "learning_rate": 1.9864206546609342e-07, + "loss": 0.7271, + "step": 7162 + }, + { + "epoch": 0.8612998256478086, + "grad_norm": 1.8807735562181023, + "learning_rate": 1.983037504469771e-07, + "loss": 0.8458, + "step": 7163 + }, + { + "epoch": 0.8614200685384477, + "grad_norm": 1.747420999162889, + "learning_rate": 1.9796570874374984e-07, + "loss": 0.6704, + "step": 7164 + }, + { + "epoch": 0.8615403114290867, + "grad_norm": 1.7620896585452732, + "learning_rate": 1.976279404076917e-07, + "loss": 0.7757, + "step": 7165 + }, + { + "epoch": 0.8616605543197259, + "grad_norm": 1.7743869878102623, + "learning_rate": 1.9729044549004193e-07, + "loss": 0.7619, + "step": 7166 + }, + { + "epoch": 0.8617807972103649, + "grad_norm": 1.917031115283879, + "learning_rate": 1.9695322404199822e-07, + "loss": 0.706, + "step": 7167 + }, + { + "epoch": 0.861901040101004, + "grad_norm": 2.6488314985381205, + "learning_rate": 1.9661627611471654e-07, + "loss": 0.8246, + "step": 7168 + }, + { + "epoch": 0.8620212829916432, + "grad_norm": 2.0465293767196724, + "learning_rate": 1.9627960175931246e-07, + "loss": 0.7011, + "step": 7169 + }, + { + "epoch": 0.8621415258822822, + "grad_norm": 2.4289323915171117, + "learning_rate": 1.9594320102685847e-07, + "loss": 0.7453, + "step": 7170 + }, + { + "epoch": 0.8622617687729213, + "grad_norm": 2.132349208495798, + "learning_rate": 1.956070739683864e-07, + "loss": 0.6505, + "step": 7171 + }, + { + "epoch": 0.8623820116635604, + "grad_norm": 1.5695249945675882, + "learning_rate": 1.9527122063488678e-07, + "loss": 0.746, + "step": 7172 + }, + { + "epoch": 0.8625022545541995, + "grad_norm": 2.000245913645963, + "learning_rate": 1.9493564107730755e-07, + "loss": 0.8074, + "step": 7173 + }, + { + "epoch": 0.8626224974448385, + "grad_norm": 2.1138617375830604, + "learning_rate": 1.9460033534655684e-07, + "loss": 0.6143, + "step": 7174 + }, + { + "epoch": 0.8627427403354777, + "grad_norm": 1.6976037078324362, + "learning_rate": 1.9426530349349978e-07, + "loss": 0.8448, + "step": 7175 + }, + { + "epoch": 0.8628629832261168, + "grad_norm": 1.9100577050125112, + "learning_rate": 1.9393054556896038e-07, + "loss": 0.6477, + "step": 7176 + }, + { + "epoch": 0.8629832261167558, + "grad_norm": 2.3973032105066454, + "learning_rate": 1.9359606162372133e-07, + "loss": 0.6962, + "step": 7177 + }, + { + "epoch": 0.863103469007395, + "grad_norm": 1.9035650010870782, + "learning_rate": 1.9326185170852293e-07, + "loss": 0.7082, + "step": 7178 + }, + { + "epoch": 0.863223711898034, + "grad_norm": 1.8190588174404871, + "learning_rate": 1.9292791587406598e-07, + "loss": 0.7227, + "step": 7179 + }, + { + "epoch": 0.8633439547886731, + "grad_norm": 1.9608256395510293, + "learning_rate": 1.9259425417100661e-07, + "loss": 0.8707, + "step": 7180 + }, + { + "epoch": 0.8634641976793123, + "grad_norm": 2.2071469436974507, + "learning_rate": 1.9226086664996234e-07, + "loss": 0.7504, + "step": 7181 + }, + { + "epoch": 0.8635844405699513, + "grad_norm": 2.7010046078961474, + "learning_rate": 1.9192775336150712e-07, + "loss": 0.7431, + "step": 7182 + }, + { + "epoch": 0.8637046834605904, + "grad_norm": 0.783785101973002, + "learning_rate": 1.915949143561739e-07, + "loss": 0.5631, + "step": 7183 + }, + { + "epoch": 0.8638249263512295, + "grad_norm": 17.202440239550683, + "learning_rate": 1.9126234968445498e-07, + "loss": 0.7842, + "step": 7184 + }, + { + "epoch": 0.8639451692418686, + "grad_norm": 1.4882677933692825, + "learning_rate": 1.9093005939679884e-07, + "loss": 0.6802, + "step": 7185 + }, + { + "epoch": 0.8640654121325076, + "grad_norm": 1.9834030053646317, + "learning_rate": 1.9059804354361452e-07, + "loss": 0.7725, + "step": 7186 + }, + { + "epoch": 0.8641856550231467, + "grad_norm": 1.7439173163534634, + "learning_rate": 1.902663021752684e-07, + "loss": 0.7063, + "step": 7187 + }, + { + "epoch": 0.8643058979137859, + "grad_norm": 2.2937428404153204, + "learning_rate": 1.8993483534208556e-07, + "loss": 0.8223, + "step": 7188 + }, + { + "epoch": 0.8644261408044249, + "grad_norm": 2.5716414798665967, + "learning_rate": 1.8960364309434884e-07, + "loss": 0.7564, + "step": 7189 + }, + { + "epoch": 0.864546383695064, + "grad_norm": 1.7632536365812816, + "learning_rate": 1.8927272548229967e-07, + "loss": 0.7861, + "step": 7190 + }, + { + "epoch": 0.8646666265857031, + "grad_norm": 1.6380090111658117, + "learning_rate": 1.8894208255613876e-07, + "loss": 0.835, + "step": 7191 + }, + { + "epoch": 0.8647868694763422, + "grad_norm": 2.438355515805675, + "learning_rate": 1.8861171436602397e-07, + "loss": 0.781, + "step": 7192 + }, + { + "epoch": 0.8649071123669813, + "grad_norm": 2.2801591538210992, + "learning_rate": 1.882816209620719e-07, + "loss": 0.8073, + "step": 7193 + }, + { + "epoch": 0.8650273552576204, + "grad_norm": 2.1324619102877604, + "learning_rate": 1.8795180239435738e-07, + "loss": 0.7765, + "step": 7194 + }, + { + "epoch": 0.8651475981482595, + "grad_norm": 4.168570935294028, + "learning_rate": 1.8762225871291348e-07, + "loss": 0.7624, + "step": 7195 + }, + { + "epoch": 0.8652678410388985, + "grad_norm": 1.9639430626581407, + "learning_rate": 1.8729298996773201e-07, + "loss": 0.8137, + "step": 7196 + }, + { + "epoch": 0.8653880839295377, + "grad_norm": 0.8476569180112395, + "learning_rate": 1.8696399620876301e-07, + "loss": 0.6094, + "step": 7197 + }, + { + "epoch": 0.8655083268201768, + "grad_norm": 2.3460170716805226, + "learning_rate": 1.866352774859141e-07, + "loss": 0.7994, + "step": 7198 + }, + { + "epoch": 0.8656285697108158, + "grad_norm": 3.4065474634037263, + "learning_rate": 1.8630683384905188e-07, + "loss": 0.6945, + "step": 7199 + }, + { + "epoch": 0.865748812601455, + "grad_norm": 1.8950162601024976, + "learning_rate": 1.8597866534800045e-07, + "loss": 0.8915, + "step": 7200 + }, + { + "epoch": 0.865869055492094, + "grad_norm": 2.5171669919524122, + "learning_rate": 1.8565077203254398e-07, + "loss": 0.7462, + "step": 7201 + }, + { + "epoch": 0.8659892983827331, + "grad_norm": 3.4012013388632774, + "learning_rate": 1.8532315395242203e-07, + "loss": 0.7353, + "step": 7202 + }, + { + "epoch": 0.8661095412733723, + "grad_norm": 2.376546055362546, + "learning_rate": 1.849958111573353e-07, + "loss": 0.7286, + "step": 7203 + }, + { + "epoch": 0.8662297841640113, + "grad_norm": 1.7563270226147583, + "learning_rate": 1.8466874369694074e-07, + "loss": 0.6438, + "step": 7204 + }, + { + "epoch": 0.8663500270546504, + "grad_norm": 2.260026104025182, + "learning_rate": 1.843419516208542e-07, + "loss": 0.7066, + "step": 7205 + }, + { + "epoch": 0.8664702699452895, + "grad_norm": 2.0874147363484474, + "learning_rate": 1.8401543497865047e-07, + "loss": 0.8017, + "step": 7206 + }, + { + "epoch": 0.8665905128359286, + "grad_norm": 3.7931170920721553, + "learning_rate": 1.836891938198608e-07, + "loss": 0.6468, + "step": 7207 + }, + { + "epoch": 0.8667107557265676, + "grad_norm": 2.843387454890369, + "learning_rate": 1.8336322819397677e-07, + "loss": 0.7151, + "step": 7208 + }, + { + "epoch": 0.8668309986172068, + "grad_norm": 2.132068658027392, + "learning_rate": 1.8303753815044654e-07, + "loss": 0.6291, + "step": 7209 + }, + { + "epoch": 0.8669512415078459, + "grad_norm": 2.350839801641051, + "learning_rate": 1.827121237386773e-07, + "loss": 0.7114, + "step": 7210 + }, + { + "epoch": 0.8670714843984849, + "grad_norm": 2.383643633865306, + "learning_rate": 1.8238698500803374e-07, + "loss": 0.7585, + "step": 7211 + }, + { + "epoch": 0.8671917272891241, + "grad_norm": 0.7263175797658631, + "learning_rate": 1.820621220078391e-07, + "loss": 0.5629, + "step": 7212 + }, + { + "epoch": 0.8673119701797631, + "grad_norm": 2.0469566503648533, + "learning_rate": 1.8173753478737553e-07, + "loss": 0.6839, + "step": 7213 + }, + { + "epoch": 0.8674322130704022, + "grad_norm": 2.0219577400930437, + "learning_rate": 1.8141322339588205e-07, + "loss": 0.8015, + "step": 7214 + }, + { + "epoch": 0.8675524559610414, + "grad_norm": 3.890204349436041, + "learning_rate": 1.810891878825569e-07, + "loss": 0.7027, + "step": 7215 + }, + { + "epoch": 0.8676726988516804, + "grad_norm": 1.9345452327729773, + "learning_rate": 1.8076542829655561e-07, + "loss": 0.7216, + "step": 7216 + }, + { + "epoch": 0.8677929417423195, + "grad_norm": 2.474571992490925, + "learning_rate": 1.8044194468699203e-07, + "loss": 0.7971, + "step": 7217 + }, + { + "epoch": 0.8679131846329585, + "grad_norm": 2.2242280199496194, + "learning_rate": 1.8011873710293912e-07, + "loss": 0.7592, + "step": 7218 + }, + { + "epoch": 0.8680334275235977, + "grad_norm": 1.9072124365436298, + "learning_rate": 1.7979580559342677e-07, + "loss": 0.6946, + "step": 7219 + }, + { + "epoch": 0.8681536704142367, + "grad_norm": 1.8473438216542293, + "learning_rate": 1.7947315020744358e-07, + "loss": 0.667, + "step": 7220 + }, + { + "epoch": 0.8682739133048758, + "grad_norm": 2.0211131764036065, + "learning_rate": 1.7915077099393594e-07, + "loss": 0.8039, + "step": 7221 + }, + { + "epoch": 0.868394156195515, + "grad_norm": 6.25286097848495, + "learning_rate": 1.788286680018083e-07, + "loss": 0.7421, + "step": 7222 + }, + { + "epoch": 0.868514399086154, + "grad_norm": 1.7476028645682962, + "learning_rate": 1.7850684127992443e-07, + "loss": 0.7264, + "step": 7223 + }, + { + "epoch": 0.8686346419767931, + "grad_norm": 1.7524029718766985, + "learning_rate": 1.7818529087710378e-07, + "loss": 0.7082, + "step": 7224 + }, + { + "epoch": 0.8687548848674322, + "grad_norm": 1.7390125579380558, + "learning_rate": 1.7786401684212637e-07, + "loss": 0.8424, + "step": 7225 + }, + { + "epoch": 0.8688751277580713, + "grad_norm": 0.7466854924428583, + "learning_rate": 1.7754301922372883e-07, + "loss": 0.5598, + "step": 7226 + }, + { + "epoch": 0.8689953706487104, + "grad_norm": 2.111527877236973, + "learning_rate": 1.7722229807060617e-07, + "loss": 0.815, + "step": 7227 + }, + { + "epoch": 0.8691156135393495, + "grad_norm": 2.084975002696996, + "learning_rate": 1.7690185343141172e-07, + "loss": 0.8262, + "step": 7228 + }, + { + "epoch": 0.8692358564299886, + "grad_norm": 2.9125619251396935, + "learning_rate": 1.7658168535475615e-07, + "loss": 0.7016, + "step": 7229 + }, + { + "epoch": 0.8693560993206276, + "grad_norm": 1.5508159950649583, + "learning_rate": 1.7626179388920948e-07, + "loss": 0.6468, + "step": 7230 + }, + { + "epoch": 0.8694763422112668, + "grad_norm": 1.559504804387937, + "learning_rate": 1.7594217908329866e-07, + "loss": 0.8063, + "step": 7231 + }, + { + "epoch": 0.8695965851019059, + "grad_norm": 2.093378745799302, + "learning_rate": 1.7562284098550895e-07, + "loss": 0.7397, + "step": 7232 + }, + { + "epoch": 0.8697168279925449, + "grad_norm": 0.8537353651766787, + "learning_rate": 1.753037796442838e-07, + "loss": 0.6232, + "step": 7233 + }, + { + "epoch": 0.8698370708831841, + "grad_norm": 2.527971004138717, + "learning_rate": 1.74984995108024e-07, + "loss": 0.7568, + "step": 7234 + }, + { + "epoch": 0.8699573137738231, + "grad_norm": 2.0416220316257196, + "learning_rate": 1.7466648742508981e-07, + "loss": 0.8328, + "step": 7235 + }, + { + "epoch": 0.8700775566644622, + "grad_norm": 2.122497768333529, + "learning_rate": 1.7434825664379837e-07, + "loss": 0.8471, + "step": 7236 + }, + { + "epoch": 0.8701977995551013, + "grad_norm": 2.6191092196977603, + "learning_rate": 1.740303028124246e-07, + "loss": 0.8637, + "step": 7237 + }, + { + "epoch": 0.8703180424457404, + "grad_norm": 1.9460196684248183, + "learning_rate": 1.7371262597920212e-07, + "loss": 0.7593, + "step": 7238 + }, + { + "epoch": 0.8704382853363795, + "grad_norm": 1.5789908120804248, + "learning_rate": 1.7339522619232195e-07, + "loss": 0.7671, + "step": 7239 + }, + { + "epoch": 0.8705585282270186, + "grad_norm": 1.981408853510372, + "learning_rate": 1.730781034999338e-07, + "loss": 0.7551, + "step": 7240 + }, + { + "epoch": 0.8706787711176577, + "grad_norm": 1.931594832661523, + "learning_rate": 1.7276125795014497e-07, + "loss": 0.7403, + "step": 7241 + }, + { + "epoch": 0.8707990140082967, + "grad_norm": 2.1204961282577375, + "learning_rate": 1.7244468959102054e-07, + "loss": 0.6787, + "step": 7242 + }, + { + "epoch": 0.8709192568989359, + "grad_norm": 2.1167478210430795, + "learning_rate": 1.7212839847058348e-07, + "loss": 0.8585, + "step": 7243 + }, + { + "epoch": 0.871039499789575, + "grad_norm": 1.9860907518755706, + "learning_rate": 1.718123846368147e-07, + "loss": 0.7401, + "step": 7244 + }, + { + "epoch": 0.871159742680214, + "grad_norm": 1.6854755832982367, + "learning_rate": 1.714966481376543e-07, + "loss": 0.72, + "step": 7245 + }, + { + "epoch": 0.8712799855708532, + "grad_norm": 1.9121573795764821, + "learning_rate": 1.7118118902099797e-07, + "loss": 0.8313, + "step": 7246 + }, + { + "epoch": 0.8714002284614922, + "grad_norm": 1.7598193432465925, + "learning_rate": 1.7086600733470146e-07, + "loss": 0.8089, + "step": 7247 + }, + { + "epoch": 0.8715204713521313, + "grad_norm": 2.084317038672427, + "learning_rate": 1.7055110312657738e-07, + "loss": 0.7752, + "step": 7248 + }, + { + "epoch": 0.8716407142427703, + "grad_norm": 2.209408828066312, + "learning_rate": 1.702364764443962e-07, + "loss": 0.7434, + "step": 7249 + }, + { + "epoch": 0.8717609571334095, + "grad_norm": 1.8924910738897018, + "learning_rate": 1.6992212733588685e-07, + "loss": 0.7298, + "step": 7250 + }, + { + "epoch": 0.8718812000240486, + "grad_norm": 1.887663610579958, + "learning_rate": 1.6960805584873538e-07, + "loss": 0.7496, + "step": 7251 + }, + { + "epoch": 0.8720014429146876, + "grad_norm": 1.5554369936143753, + "learning_rate": 1.6929426203058684e-07, + "loss": 0.7847, + "step": 7252 + }, + { + "epoch": 0.8721216858053268, + "grad_norm": 2.031918577057231, + "learning_rate": 1.689807459290431e-07, + "loss": 0.8027, + "step": 7253 + }, + { + "epoch": 0.8722419286959658, + "grad_norm": 4.730624989355739, + "learning_rate": 1.6866750759166437e-07, + "loss": 0.7141, + "step": 7254 + }, + { + "epoch": 0.8723621715866049, + "grad_norm": 2.2654282056907387, + "learning_rate": 1.6835454706596865e-07, + "loss": 0.7713, + "step": 7255 + }, + { + "epoch": 0.8724824144772441, + "grad_norm": 1.8625698050916404, + "learning_rate": 1.680418643994317e-07, + "loss": 0.7412, + "step": 7256 + }, + { + "epoch": 0.8726026573678831, + "grad_norm": 0.9049901097110256, + "learning_rate": 1.6772945963948738e-07, + "loss": 0.6454, + "step": 7257 + }, + { + "epoch": 0.8727229002585222, + "grad_norm": 2.2196154988300303, + "learning_rate": 1.6741733283352733e-07, + "loss": 0.7774, + "step": 7258 + }, + { + "epoch": 0.8728431431491613, + "grad_norm": 1.5672655293444808, + "learning_rate": 1.6710548402890102e-07, + "loss": 0.8406, + "step": 7259 + }, + { + "epoch": 0.8729633860398004, + "grad_norm": 1.8061605279535446, + "learning_rate": 1.6679391327291527e-07, + "loss": 0.6684, + "step": 7260 + }, + { + "epoch": 0.8730836289304394, + "grad_norm": 2.684819148559849, + "learning_rate": 1.6648262061283492e-07, + "loss": 0.6823, + "step": 7261 + }, + { + "epoch": 0.8732038718210786, + "grad_norm": 1.8629542929611433, + "learning_rate": 1.6617160609588353e-07, + "loss": 0.7425, + "step": 7262 + }, + { + "epoch": 0.8733241147117177, + "grad_norm": 2.37419837674296, + "learning_rate": 1.6586086976924163e-07, + "loss": 0.7233, + "step": 7263 + }, + { + "epoch": 0.8734443576023567, + "grad_norm": 2.2513732957728148, + "learning_rate": 1.6555041168004747e-07, + "loss": 0.7873, + "step": 7264 + }, + { + "epoch": 0.8735646004929959, + "grad_norm": 1.9884082326454242, + "learning_rate": 1.6524023187539715e-07, + "loss": 0.6923, + "step": 7265 + }, + { + "epoch": 0.873684843383635, + "grad_norm": 1.8333326975503454, + "learning_rate": 1.649303304023446e-07, + "loss": 0.7513, + "step": 7266 + }, + { + "epoch": 0.873805086274274, + "grad_norm": 1.667097306247393, + "learning_rate": 1.6462070730790246e-07, + "loss": 0.793, + "step": 7267 + }, + { + "epoch": 0.8739253291649132, + "grad_norm": 2.411759843096216, + "learning_rate": 1.6431136263903912e-07, + "loss": 0.7892, + "step": 7268 + }, + { + "epoch": 0.8740455720555522, + "grad_norm": 1.9806120630662136, + "learning_rate": 1.6400229644268282e-07, + "loss": 0.7376, + "step": 7269 + }, + { + "epoch": 0.8741658149461913, + "grad_norm": 1.9628850376941624, + "learning_rate": 1.6369350876571852e-07, + "loss": 0.8148, + "step": 7270 + }, + { + "epoch": 0.8742860578368304, + "grad_norm": 2.086620433323625, + "learning_rate": 1.6338499965498874e-07, + "loss": 0.8217, + "step": 7271 + }, + { + "epoch": 0.8744063007274695, + "grad_norm": 1.5391865791612045, + "learning_rate": 1.630767691572943e-07, + "loss": 0.7748, + "step": 7272 + }, + { + "epoch": 0.8745265436181086, + "grad_norm": 0.756232630317397, + "learning_rate": 1.6276881731939306e-07, + "loss": 0.5353, + "step": 7273 + }, + { + "epoch": 0.8746467865087477, + "grad_norm": 1.7333826198662106, + "learning_rate": 1.6246114418800193e-07, + "loss": 0.7539, + "step": 7274 + }, + { + "epoch": 0.8747670293993868, + "grad_norm": 1.7669033048000062, + "learning_rate": 1.6215374980979423e-07, + "loss": 0.7696, + "step": 7275 + }, + { + "epoch": 0.8748872722900258, + "grad_norm": 1.927438823456432, + "learning_rate": 1.6184663423140133e-07, + "loss": 0.6905, + "step": 7276 + }, + { + "epoch": 0.875007515180665, + "grad_norm": 1.9785994335925277, + "learning_rate": 1.615397974994126e-07, + "loss": 0.6405, + "step": 7277 + }, + { + "epoch": 0.875127758071304, + "grad_norm": 1.486959917129832, + "learning_rate": 1.6123323966037438e-07, + "loss": 0.809, + "step": 7278 + }, + { + "epoch": 0.8752480009619431, + "grad_norm": 1.880470243547684, + "learning_rate": 1.6092696076079216e-07, + "loss": 0.789, + "step": 7279 + }, + { + "epoch": 0.8753682438525822, + "grad_norm": 1.77551833093194, + "learning_rate": 1.6062096084712785e-07, + "loss": 0.7406, + "step": 7280 + }, + { + "epoch": 0.8754884867432213, + "grad_norm": 1.9412614549641531, + "learning_rate": 1.6031523996580098e-07, + "loss": 0.7055, + "step": 7281 + }, + { + "epoch": 0.8756087296338604, + "grad_norm": 2.323616937251393, + "learning_rate": 1.6000979816318981e-07, + "loss": 0.6626, + "step": 7282 + }, + { + "epoch": 0.8757289725244994, + "grad_norm": 2.3504118781629058, + "learning_rate": 1.5970463548562886e-07, + "loss": 0.7502, + "step": 7283 + }, + { + "epoch": 0.8758492154151386, + "grad_norm": 1.96172307909173, + "learning_rate": 1.5939975197941192e-07, + "loss": 0.7148, + "step": 7284 + }, + { + "epoch": 0.8759694583057777, + "grad_norm": 0.8174178139907367, + "learning_rate": 1.5909514769078892e-07, + "loss": 0.5331, + "step": 7285 + }, + { + "epoch": 0.8760897011964167, + "grad_norm": 1.6148989051143803, + "learning_rate": 1.5879082266596867e-07, + "loss": 0.7834, + "step": 7286 + }, + { + "epoch": 0.8762099440870559, + "grad_norm": 1.789039245187803, + "learning_rate": 1.5848677695111645e-07, + "loss": 0.7238, + "step": 7287 + }, + { + "epoch": 0.8763301869776949, + "grad_norm": 2.5473511607646047, + "learning_rate": 1.5818301059235562e-07, + "loss": 0.704, + "step": 7288 + }, + { + "epoch": 0.876450429868334, + "grad_norm": 2.916378297628401, + "learning_rate": 1.578795236357684e-07, + "loss": 0.8179, + "step": 7289 + }, + { + "epoch": 0.8765706727589732, + "grad_norm": 2.130832521599814, + "learning_rate": 1.5757631612739218e-07, + "loss": 0.8567, + "step": 7290 + }, + { + "epoch": 0.8766909156496122, + "grad_norm": 0.7904444763594811, + "learning_rate": 1.572733881132242e-07, + "loss": 0.6145, + "step": 7291 + }, + { + "epoch": 0.8768111585402513, + "grad_norm": 0.8026431140860693, + "learning_rate": 1.5697073963921814e-07, + "loss": 0.5856, + "step": 7292 + }, + { + "epoch": 0.8769314014308904, + "grad_norm": 2.0383085413517046, + "learning_rate": 1.566683707512857e-07, + "loss": 0.8587, + "step": 7293 + }, + { + "epoch": 0.8770516443215295, + "grad_norm": 2.07143201968653, + "learning_rate": 1.5636628149529553e-07, + "loss": 0.8015, + "step": 7294 + }, + { + "epoch": 0.8771718872121685, + "grad_norm": 3.3057308488914368, + "learning_rate": 1.560644719170743e-07, + "loss": 0.8004, + "step": 7295 + }, + { + "epoch": 0.8772921301028077, + "grad_norm": 1.7471619481234648, + "learning_rate": 1.5576294206240692e-07, + "loss": 0.7218, + "step": 7296 + }, + { + "epoch": 0.8774123729934468, + "grad_norm": 2.5492591230426176, + "learning_rate": 1.5546169197703507e-07, + "loss": 0.6813, + "step": 7297 + }, + { + "epoch": 0.8775326158840858, + "grad_norm": 2.5595920158668273, + "learning_rate": 1.5516072170665774e-07, + "loss": 0.7804, + "step": 7298 + }, + { + "epoch": 0.877652858774725, + "grad_norm": 2.0066233872766155, + "learning_rate": 1.5486003129693214e-07, + "loss": 0.8674, + "step": 7299 + }, + { + "epoch": 0.877773101665364, + "grad_norm": 1.8932596560143442, + "learning_rate": 1.545596207934725e-07, + "loss": 0.784, + "step": 7300 + }, + { + "epoch": 0.8778933445560031, + "grad_norm": 1.7933299380250198, + "learning_rate": 1.5425949024185147e-07, + "loss": 0.7816, + "step": 7301 + }, + { + "epoch": 0.8780135874466423, + "grad_norm": 2.0012128252035146, + "learning_rate": 1.5395963968759818e-07, + "loss": 0.68, + "step": 7302 + }, + { + "epoch": 0.8781338303372813, + "grad_norm": 1.655757614936119, + "learning_rate": 1.536600691761998e-07, + "loss": 0.6471, + "step": 7303 + }, + { + "epoch": 0.8782540732279204, + "grad_norm": 1.803804255758041, + "learning_rate": 1.5336077875310084e-07, + "loss": 0.7193, + "step": 7304 + }, + { + "epoch": 0.8783743161185595, + "grad_norm": 1.945528194618286, + "learning_rate": 1.5306176846370321e-07, + "loss": 0.7431, + "step": 7305 + }, + { + "epoch": 0.8784945590091986, + "grad_norm": 1.9851095687259714, + "learning_rate": 1.5276303835336712e-07, + "loss": 0.7434, + "step": 7306 + }, + { + "epoch": 0.8786148018998376, + "grad_norm": 0.78961128564507, + "learning_rate": 1.524645884674094e-07, + "loss": 0.536, + "step": 7307 + }, + { + "epoch": 0.8787350447904768, + "grad_norm": 2.1911969144389283, + "learning_rate": 1.521664188511047e-07, + "loss": 0.7958, + "step": 7308 + }, + { + "epoch": 0.8788552876811159, + "grad_norm": 6.478125812624316, + "learning_rate": 1.518685295496851e-07, + "loss": 0.8085, + "step": 7309 + }, + { + "epoch": 0.8789755305717549, + "grad_norm": 1.742235670578449, + "learning_rate": 1.5157092060833975e-07, + "loss": 0.8593, + "step": 7310 + }, + { + "epoch": 0.879095773462394, + "grad_norm": 1.6662535234830422, + "learning_rate": 1.5127359207221658e-07, + "loss": 0.6681, + "step": 7311 + }, + { + "epoch": 0.8792160163530331, + "grad_norm": 2.2116071677619082, + "learning_rate": 1.5097654398641923e-07, + "loss": 0.7336, + "step": 7312 + }, + { + "epoch": 0.8793362592436722, + "grad_norm": 1.3086827034394377, + "learning_rate": 1.5067977639601014e-07, + "loss": 0.7321, + "step": 7313 + }, + { + "epoch": 0.8794565021343113, + "grad_norm": 7.727407935688462, + "learning_rate": 1.5038328934600864e-07, + "loss": 0.7115, + "step": 7314 + }, + { + "epoch": 0.8795767450249504, + "grad_norm": 2.219833109656896, + "learning_rate": 1.5008708288139161e-07, + "loss": 0.7025, + "step": 7315 + }, + { + "epoch": 0.8796969879155895, + "grad_norm": 2.039935279642321, + "learning_rate": 1.497911570470931e-07, + "loss": 0.737, + "step": 7316 + }, + { + "epoch": 0.8798172308062285, + "grad_norm": 1.6413483977619419, + "learning_rate": 1.494955118880048e-07, + "loss": 0.857, + "step": 7317 + }, + { + "epoch": 0.8799374736968677, + "grad_norm": 1.7061779412225027, + "learning_rate": 1.4920014744897634e-07, + "loss": 0.7313, + "step": 7318 + }, + { + "epoch": 0.8800577165875068, + "grad_norm": 1.8451764682434444, + "learning_rate": 1.4890506377481392e-07, + "loss": 0.8701, + "step": 7319 + }, + { + "epoch": 0.8801779594781458, + "grad_norm": 1.6117762456892746, + "learning_rate": 1.486102609102815e-07, + "loss": 0.6399, + "step": 7320 + }, + { + "epoch": 0.880298202368785, + "grad_norm": 2.5459586525231837, + "learning_rate": 1.483157389001004e-07, + "loss": 0.8616, + "step": 7321 + }, + { + "epoch": 0.880418445259424, + "grad_norm": 2.052235115234286, + "learning_rate": 1.4802149778894933e-07, + "loss": 0.7861, + "step": 7322 + }, + { + "epoch": 0.8805386881500631, + "grad_norm": 1.8017691924606745, + "learning_rate": 1.4772753762146484e-07, + "loss": 0.879, + "step": 7323 + }, + { + "epoch": 0.8806589310407023, + "grad_norm": 1.5601508219265865, + "learning_rate": 1.474338584422401e-07, + "loss": 0.7093, + "step": 7324 + }, + { + "epoch": 0.8807791739313413, + "grad_norm": 1.7973035881744006, + "learning_rate": 1.4714046029582595e-07, + "loss": 0.7616, + "step": 7325 + }, + { + "epoch": 0.8808994168219804, + "grad_norm": 1.882059033585849, + "learning_rate": 1.46847343226731e-07, + "loss": 0.7618, + "step": 7326 + }, + { + "epoch": 0.8810196597126195, + "grad_norm": 1.9143764316255831, + "learning_rate": 1.465545072794203e-07, + "loss": 0.702, + "step": 7327 + }, + { + "epoch": 0.8811399026032586, + "grad_norm": 1.5567002039747189, + "learning_rate": 1.4626195249831774e-07, + "loss": 0.7551, + "step": 7328 + }, + { + "epoch": 0.8812601454938976, + "grad_norm": 1.7715438614081973, + "learning_rate": 1.4596967892780244e-07, + "loss": 0.7215, + "step": 7329 + }, + { + "epoch": 0.8813803883845368, + "grad_norm": 1.6731089378956787, + "learning_rate": 1.4567768661221314e-07, + "loss": 0.7513, + "step": 7330 + }, + { + "epoch": 0.8815006312751759, + "grad_norm": 1.9691878543359094, + "learning_rate": 1.4538597559584442e-07, + "loss": 0.7475, + "step": 7331 + }, + { + "epoch": 0.8816208741658149, + "grad_norm": 2.6258995694597993, + "learning_rate": 1.4509454592294823e-07, + "loss": 0.792, + "step": 7332 + }, + { + "epoch": 0.8817411170564541, + "grad_norm": 1.9423205795790255, + "learning_rate": 1.448033976377354e-07, + "loss": 0.7961, + "step": 7333 + }, + { + "epoch": 0.8818613599470931, + "grad_norm": 2.04178586598901, + "learning_rate": 1.445125307843713e-07, + "loss": 0.7455, + "step": 7334 + }, + { + "epoch": 0.8819816028377322, + "grad_norm": 1.6304499860414063, + "learning_rate": 1.442219454069813e-07, + "loss": 0.7623, + "step": 7335 + }, + { + "epoch": 0.8821018457283714, + "grad_norm": 1.8348057931519577, + "learning_rate": 1.4393164154964676e-07, + "loss": 0.6661, + "step": 7336 + }, + { + "epoch": 0.8822220886190104, + "grad_norm": 2.1482573832524086, + "learning_rate": 1.4364161925640649e-07, + "loss": 0.9417, + "step": 7337 + }, + { + "epoch": 0.8823423315096495, + "grad_norm": 1.8534316917203648, + "learning_rate": 1.4335187857125663e-07, + "loss": 0.8566, + "step": 7338 + }, + { + "epoch": 0.8824625744002886, + "grad_norm": 1.96092795563461, + "learning_rate": 1.4306241953815023e-07, + "loss": 0.758, + "step": 7339 + }, + { + "epoch": 0.8825828172909277, + "grad_norm": 1.7959179527130957, + "learning_rate": 1.4277324220099862e-07, + "loss": 0.7142, + "step": 7340 + }, + { + "epoch": 0.8827030601815667, + "grad_norm": 2.1279227408531636, + "learning_rate": 1.4248434660366938e-07, + "loss": 0.7436, + "step": 7341 + }, + { + "epoch": 0.8828233030722058, + "grad_norm": 1.9609716647998015, + "learning_rate": 1.4219573278998808e-07, + "loss": 0.7105, + "step": 7342 + }, + { + "epoch": 0.882943545962845, + "grad_norm": 1.9876255235680305, + "learning_rate": 1.4190740080373685e-07, + "loss": 0.6537, + "step": 7343 + }, + { + "epoch": 0.883063788853484, + "grad_norm": 1.9358855874011058, + "learning_rate": 1.4161935068865538e-07, + "loss": 0.8414, + "step": 7344 + }, + { + "epoch": 0.8831840317441231, + "grad_norm": 2.104043849974027, + "learning_rate": 1.4133158248844113e-07, + "loss": 0.7603, + "step": 7345 + }, + { + "epoch": 0.8833042746347622, + "grad_norm": 1.7487055214276035, + "learning_rate": 1.4104409624674785e-07, + "loss": 0.7382, + "step": 7346 + }, + { + "epoch": 0.8834245175254013, + "grad_norm": 1.8013154600745231, + "learning_rate": 1.407568920071873e-07, + "loss": 0.7886, + "step": 7347 + }, + { + "epoch": 0.8835447604160404, + "grad_norm": 2.1427094272381395, + "learning_rate": 1.4046996981332782e-07, + "loss": 0.6888, + "step": 7348 + }, + { + "epoch": 0.8836650033066795, + "grad_norm": 2.1144339472677784, + "learning_rate": 1.4018332970869516e-07, + "loss": 0.7882, + "step": 7349 + }, + { + "epoch": 0.8837852461973186, + "grad_norm": 1.8431826219577274, + "learning_rate": 1.398969717367733e-07, + "loss": 0.855, + "step": 7350 + }, + { + "epoch": 0.8839054890879576, + "grad_norm": 1.9008060953858128, + "learning_rate": 1.396108959410014e-07, + "loss": 0.7639, + "step": 7351 + }, + { + "epoch": 0.8840257319785968, + "grad_norm": 1.5489863619120887, + "learning_rate": 1.3932510236477745e-07, + "loss": 0.8156, + "step": 7352 + }, + { + "epoch": 0.8841459748692359, + "grad_norm": 1.9155165151522884, + "learning_rate": 1.3903959105145636e-07, + "loss": 0.5645, + "step": 7353 + }, + { + "epoch": 0.8842662177598749, + "grad_norm": 1.9383980915319956, + "learning_rate": 1.387543620443492e-07, + "loss": 0.8386, + "step": 7354 + }, + { + "epoch": 0.8843864606505141, + "grad_norm": 1.8149350350683504, + "learning_rate": 1.3846941538672606e-07, + "loss": 0.8424, + "step": 7355 + }, + { + "epoch": 0.8845067035411531, + "grad_norm": 2.1140322664048625, + "learning_rate": 1.3818475112181193e-07, + "loss": 0.8119, + "step": 7356 + }, + { + "epoch": 0.8846269464317922, + "grad_norm": 2.0953140110860975, + "learning_rate": 1.3790036929279091e-07, + "loss": 0.7996, + "step": 7357 + }, + { + "epoch": 0.8847471893224313, + "grad_norm": 2.470695598123416, + "learning_rate": 1.3761626994280363e-07, + "loss": 0.5896, + "step": 7358 + }, + { + "epoch": 0.8848674322130704, + "grad_norm": 1.9226952019891757, + "learning_rate": 1.3733245311494735e-07, + "loss": 0.7382, + "step": 7359 + }, + { + "epoch": 0.8849876751037095, + "grad_norm": 2.0608120571364257, + "learning_rate": 1.3704891885227676e-07, + "loss": 0.7105, + "step": 7360 + }, + { + "epoch": 0.8851079179943486, + "grad_norm": 1.9204852033510393, + "learning_rate": 1.367656671978037e-07, + "loss": 0.7773, + "step": 7361 + }, + { + "epoch": 0.8852281608849877, + "grad_norm": 1.822652663232681, + "learning_rate": 1.36482698194498e-07, + "loss": 0.7384, + "step": 7362 + }, + { + "epoch": 0.8853484037756267, + "grad_norm": 2.0328023993837143, + "learning_rate": 1.3620001188528506e-07, + "loss": 0.7237, + "step": 7363 + }, + { + "epoch": 0.8854686466662659, + "grad_norm": 2.243948852738144, + "learning_rate": 1.3591760831304865e-07, + "loss": 0.7368, + "step": 7364 + }, + { + "epoch": 0.885588889556905, + "grad_norm": 1.7530429814102193, + "learning_rate": 1.356354875206287e-07, + "loss": 0.7922, + "step": 7365 + }, + { + "epoch": 0.885709132447544, + "grad_norm": 2.1378758730206027, + "learning_rate": 1.3535364955082296e-07, + "loss": 0.7003, + "step": 7366 + }, + { + "epoch": 0.8858293753381832, + "grad_norm": 3.576678520910617, + "learning_rate": 1.3507209444638613e-07, + "loss": 0.645, + "step": 7367 + }, + { + "epoch": 0.8859496182288222, + "grad_norm": 1.9394909873255386, + "learning_rate": 1.347908222500298e-07, + "loss": 0.7432, + "step": 7368 + }, + { + "epoch": 0.8860698611194613, + "grad_norm": 2.071258328869407, + "learning_rate": 1.3450983300442276e-07, + "loss": 0.7013, + "step": 7369 + }, + { + "epoch": 0.8861901040101005, + "grad_norm": 2.225338033650994, + "learning_rate": 1.3422912675219068e-07, + "loss": 0.7371, + "step": 7370 + }, + { + "epoch": 0.8863103469007395, + "grad_norm": 2.151040238079, + "learning_rate": 1.339487035359166e-07, + "loss": 0.7944, + "step": 7371 + }, + { + "epoch": 0.8864305897913786, + "grad_norm": 1.5636357566517787, + "learning_rate": 1.336685633981409e-07, + "loss": 0.8465, + "step": 7372 + }, + { + "epoch": 0.8865508326820177, + "grad_norm": 2.008528407371978, + "learning_rate": 1.333887063813597e-07, + "loss": 0.7501, + "step": 7373 + }, + { + "epoch": 0.8866710755726568, + "grad_norm": 1.7452098819682837, + "learning_rate": 1.331091325280278e-07, + "loss": 0.6628, + "step": 7374 + }, + { + "epoch": 0.8867913184632958, + "grad_norm": 1.7597483508143912, + "learning_rate": 1.3282984188055625e-07, + "loss": 0.7861, + "step": 7375 + }, + { + "epoch": 0.8869115613539349, + "grad_norm": 1.8290202239322622, + "learning_rate": 1.3255083448131288e-07, + "loss": 0.8027, + "step": 7376 + }, + { + "epoch": 0.8870318042445741, + "grad_norm": 3.8987733623640013, + "learning_rate": 1.3227211037262365e-07, + "loss": 0.7963, + "step": 7377 + }, + { + "epoch": 0.8871520471352131, + "grad_norm": 2.3367396989837967, + "learning_rate": 1.319936695967696e-07, + "loss": 0.8563, + "step": 7378 + }, + { + "epoch": 0.8872722900258522, + "grad_norm": 2.913933881524005, + "learning_rate": 1.3171551219599097e-07, + "loss": 0.8209, + "step": 7379 + }, + { + "epoch": 0.8873925329164913, + "grad_norm": 2.105974182724181, + "learning_rate": 1.3143763821248377e-07, + "loss": 0.787, + "step": 7380 + }, + { + "epoch": 0.8875127758071304, + "grad_norm": 5.548355635627498, + "learning_rate": 1.3116004768840118e-07, + "loss": 0.726, + "step": 7381 + }, + { + "epoch": 0.8876330186977694, + "grad_norm": 1.765132869225574, + "learning_rate": 1.3088274066585348e-07, + "loss": 0.7437, + "step": 7382 + }, + { + "epoch": 0.8877532615884086, + "grad_norm": 2.3772640479420235, + "learning_rate": 1.3060571718690749e-07, + "loss": 0.9057, + "step": 7383 + }, + { + "epoch": 0.8878735044790477, + "grad_norm": 0.7665678952350313, + "learning_rate": 1.3032897729358805e-07, + "loss": 0.5697, + "step": 7384 + }, + { + "epoch": 0.8879937473696867, + "grad_norm": 2.2163714702299053, + "learning_rate": 1.3005252102787645e-07, + "loss": 0.805, + "step": 7385 + }, + { + "epoch": 0.8881139902603259, + "grad_norm": 1.6411902044462763, + "learning_rate": 1.297763484317105e-07, + "loss": 0.734, + "step": 7386 + }, + { + "epoch": 0.888234233150965, + "grad_norm": 2.171129057419703, + "learning_rate": 1.2950045954698551e-07, + "loss": 0.7121, + "step": 7387 + }, + { + "epoch": 0.888354476041604, + "grad_norm": 1.7557167275334304, + "learning_rate": 1.2922485441555343e-07, + "loss": 0.7594, + "step": 7388 + }, + { + "epoch": 0.8884747189322432, + "grad_norm": 1.7571641192440215, + "learning_rate": 1.2894953307922363e-07, + "loss": 0.8187, + "step": 7389 + }, + { + "epoch": 0.8885949618228822, + "grad_norm": 1.9294963541226269, + "learning_rate": 1.2867449557976208e-07, + "loss": 0.8465, + "step": 7390 + }, + { + "epoch": 0.8887152047135213, + "grad_norm": 1.831678930050053, + "learning_rate": 1.283997419588916e-07, + "loss": 0.7614, + "step": 7391 + }, + { + "epoch": 0.8888354476041604, + "grad_norm": 1.9200011010961555, + "learning_rate": 1.2812527225829216e-07, + "loss": 0.6277, + "step": 7392 + }, + { + "epoch": 0.8889556904947995, + "grad_norm": 2.1146038662018496, + "learning_rate": 1.2785108651960052e-07, + "loss": 0.7693, + "step": 7393 + }, + { + "epoch": 0.8890759333854386, + "grad_norm": 1.8918679305348731, + "learning_rate": 1.2757718478441094e-07, + "loss": 0.8132, + "step": 7394 + }, + { + "epoch": 0.8891961762760777, + "grad_norm": 1.6917453628550898, + "learning_rate": 1.2730356709427302e-07, + "loss": 0.7783, + "step": 7395 + }, + { + "epoch": 0.8893164191667168, + "grad_norm": 1.569297956233759, + "learning_rate": 1.2703023349069542e-07, + "loss": 0.6054, + "step": 7396 + }, + { + "epoch": 0.8894366620573558, + "grad_norm": 1.832797607985775, + "learning_rate": 1.2675718401514223e-07, + "loss": 0.6234, + "step": 7397 + }, + { + "epoch": 0.889556904947995, + "grad_norm": 4.9391770231008465, + "learning_rate": 1.264844187090346e-07, + "loss": 0.7468, + "step": 7398 + }, + { + "epoch": 0.889677147838634, + "grad_norm": 2.0969863475451556, + "learning_rate": 1.262119376137516e-07, + "loss": 0.7544, + "step": 7399 + }, + { + "epoch": 0.8897973907292731, + "grad_norm": 2.3840470914734033, + "learning_rate": 1.2593974077062707e-07, + "loss": 0.8516, + "step": 7400 + }, + { + "epoch": 0.8899176336199123, + "grad_norm": 1.494803806304373, + "learning_rate": 1.2566782822095423e-07, + "loss": 0.6339, + "step": 7401 + }, + { + "epoch": 0.8900378765105513, + "grad_norm": 1.7583300019484265, + "learning_rate": 1.2539620000598162e-07, + "loss": 0.7201, + "step": 7402 + }, + { + "epoch": 0.8901581194011904, + "grad_norm": 1.7079340382821946, + "learning_rate": 1.2512485616691492e-07, + "loss": 0.7974, + "step": 7403 + }, + { + "epoch": 0.8902783622918296, + "grad_norm": 1.7424851920156799, + "learning_rate": 1.2485379674491681e-07, + "loss": 0.8109, + "step": 7404 + }, + { + "epoch": 0.8903986051824686, + "grad_norm": 2.1607111240620545, + "learning_rate": 1.2458302178110657e-07, + "loss": 0.795, + "step": 7405 + }, + { + "epoch": 0.8905188480731077, + "grad_norm": 2.0706784482736618, + "learning_rate": 1.2431253131656118e-07, + "loss": 0.8277, + "step": 7406 + }, + { + "epoch": 0.8906390909637467, + "grad_norm": 1.8172327273801299, + "learning_rate": 1.240423253923133e-07, + "loss": 0.7687, + "step": 7407 + }, + { + "epoch": 0.8907593338543859, + "grad_norm": 2.286363696758987, + "learning_rate": 1.237724040493533e-07, + "loss": 0.6957, + "step": 7408 + }, + { + "epoch": 0.8908795767450249, + "grad_norm": 5.1272129770458434, + "learning_rate": 1.2350276732862773e-07, + "loss": 0.7374, + "step": 7409 + }, + { + "epoch": 0.890999819635664, + "grad_norm": 0.848129281108399, + "learning_rate": 1.2323341527103993e-07, + "loss": 0.5672, + "step": 7410 + }, + { + "epoch": 0.8911200625263032, + "grad_norm": 2.2266779284087748, + "learning_rate": 1.2296434791745135e-07, + "loss": 0.8508, + "step": 7411 + }, + { + "epoch": 0.8912403054169422, + "grad_norm": 1.992684036907059, + "learning_rate": 1.2269556530867875e-07, + "loss": 0.7712, + "step": 7412 + }, + { + "epoch": 0.8913605483075813, + "grad_norm": 2.56991657581818, + "learning_rate": 1.2242706748549614e-07, + "loss": 0.821, + "step": 7413 + }, + { + "epoch": 0.8914807911982204, + "grad_norm": 1.9448924396261835, + "learning_rate": 1.2215885448863473e-07, + "loss": 0.8192, + "step": 7414 + }, + { + "epoch": 0.8916010340888595, + "grad_norm": 1.9166508826006696, + "learning_rate": 1.2189092635878152e-07, + "loss": 0.8043, + "step": 7415 + }, + { + "epoch": 0.8917212769794985, + "grad_norm": 1.6490613290313791, + "learning_rate": 1.216232831365822e-07, + "loss": 0.7758, + "step": 7416 + }, + { + "epoch": 0.8918415198701377, + "grad_norm": 1.7487500358837853, + "learning_rate": 1.2135592486263678e-07, + "loss": 0.813, + "step": 7417 + }, + { + "epoch": 0.8919617627607768, + "grad_norm": 1.8910590059867343, + "learning_rate": 1.2108885157750415e-07, + "loss": 0.6142, + "step": 7418 + }, + { + "epoch": 0.8920820056514158, + "grad_norm": 1.7544836100793617, + "learning_rate": 1.2082206332169897e-07, + "loss": 0.8057, + "step": 7419 + }, + { + "epoch": 0.892202248542055, + "grad_norm": 2.265941834493628, + "learning_rate": 1.2055556013569225e-07, + "loss": 0.7377, + "step": 7420 + }, + { + "epoch": 0.892322491432694, + "grad_norm": 1.6607574451422638, + "learning_rate": 1.2028934205991315e-07, + "loss": 0.8252, + "step": 7421 + }, + { + "epoch": 0.8924427343233331, + "grad_norm": 1.412132801162338, + "learning_rate": 1.2002340913474607e-07, + "loss": 0.767, + "step": 7422 + }, + { + "epoch": 0.8925629772139723, + "grad_norm": 1.871144654411776, + "learning_rate": 1.1975776140053317e-07, + "loss": 0.7448, + "step": 7423 + }, + { + "epoch": 0.8926832201046113, + "grad_norm": 1.9082196800115094, + "learning_rate": 1.194923988975729e-07, + "loss": 0.7425, + "step": 7424 + }, + { + "epoch": 0.8928034629952504, + "grad_norm": 2.2487986854427247, + "learning_rate": 1.192273216661206e-07, + "loss": 0.7366, + "step": 7425 + }, + { + "epoch": 0.8929237058858895, + "grad_norm": 0.7824175312228283, + "learning_rate": 1.189625297463881e-07, + "loss": 0.576, + "step": 7426 + }, + { + "epoch": 0.8930439487765286, + "grad_norm": 1.6547587176766305, + "learning_rate": 1.1869802317854394e-07, + "loss": 0.8008, + "step": 7427 + }, + { + "epoch": 0.8931641916671677, + "grad_norm": 3.2134426767362636, + "learning_rate": 1.1843380200271425e-07, + "loss": 0.7251, + "step": 7428 + }, + { + "epoch": 0.8932844345578068, + "grad_norm": 1.679850228877006, + "learning_rate": 1.181698662589805e-07, + "loss": 0.8102, + "step": 7429 + }, + { + "epoch": 0.8934046774484459, + "grad_norm": 1.8772522432946601, + "learning_rate": 1.1790621598738249e-07, + "loss": 0.7605, + "step": 7430 + }, + { + "epoch": 0.8935249203390849, + "grad_norm": 1.8692995520045768, + "learning_rate": 1.1764285122791461e-07, + "loss": 0.7465, + "step": 7431 + }, + { + "epoch": 0.8936451632297241, + "grad_norm": 2.0321488959059657, + "learning_rate": 1.173797720205294e-07, + "loss": 0.7811, + "step": 7432 + }, + { + "epoch": 0.8937654061203631, + "grad_norm": 2.796778416164705, + "learning_rate": 1.1711697840513602e-07, + "loss": 0.7234, + "step": 7433 + }, + { + "epoch": 0.8938856490110022, + "grad_norm": 2.5820504561502027, + "learning_rate": 1.1685447042160012e-07, + "loss": 0.7151, + "step": 7434 + }, + { + "epoch": 0.8940058919016414, + "grad_norm": 1.7363321756974335, + "learning_rate": 1.1659224810974367e-07, + "loss": 0.7172, + "step": 7435 + }, + { + "epoch": 0.8941261347922804, + "grad_norm": 1.6899596055976118, + "learning_rate": 1.1633031150934591e-07, + "loss": 0.6857, + "step": 7436 + }, + { + "epoch": 0.8942463776829195, + "grad_norm": 1.8294870893925177, + "learning_rate": 1.1606866066014176e-07, + "loss": 0.7963, + "step": 7437 + }, + { + "epoch": 0.8943666205735585, + "grad_norm": 2.122596503446397, + "learning_rate": 1.1580729560182434e-07, + "loss": 0.7558, + "step": 7438 + }, + { + "epoch": 0.8944868634641977, + "grad_norm": 1.6622020368376071, + "learning_rate": 1.1554621637404171e-07, + "loss": 0.7106, + "step": 7439 + }, + { + "epoch": 0.8946071063548368, + "grad_norm": 2.3748399780974347, + "learning_rate": 1.1528542301639999e-07, + "loss": 0.6116, + "step": 7440 + }, + { + "epoch": 0.8947273492454758, + "grad_norm": 2.432498103549763, + "learning_rate": 1.1502491556846105e-07, + "loss": 0.8249, + "step": 7441 + }, + { + "epoch": 0.894847592136115, + "grad_norm": 2.598465642816817, + "learning_rate": 1.1476469406974331e-07, + "loss": 0.8132, + "step": 7442 + }, + { + "epoch": 0.894967835026754, + "grad_norm": 1.7326023562582094, + "learning_rate": 1.1450475855972341e-07, + "loss": 0.7738, + "step": 7443 + }, + { + "epoch": 0.8950880779173931, + "grad_norm": 2.28257785977688, + "learning_rate": 1.1424510907783158e-07, + "loss": 0.708, + "step": 7444 + }, + { + "epoch": 0.8952083208080323, + "grad_norm": 1.6870303736758008, + "learning_rate": 1.1398574566345787e-07, + "loss": 0.8297, + "step": 7445 + }, + { + "epoch": 0.8953285636986713, + "grad_norm": 2.061629949914514, + "learning_rate": 1.1372666835594702e-07, + "loss": 0.8299, + "step": 7446 + }, + { + "epoch": 0.8954488065893104, + "grad_norm": 2.0971766657088295, + "learning_rate": 1.1346787719460071e-07, + "loss": 0.721, + "step": 7447 + }, + { + "epoch": 0.8955690494799495, + "grad_norm": 1.818801842438952, + "learning_rate": 1.1320937221867732e-07, + "loss": 0.7241, + "step": 7448 + }, + { + "epoch": 0.8956892923705886, + "grad_norm": 2.1977690395536795, + "learning_rate": 1.1295115346739192e-07, + "loss": 0.8024, + "step": 7449 + }, + { + "epoch": 0.8958095352612276, + "grad_norm": 2.190103670740846, + "learning_rate": 1.1269322097991629e-07, + "loss": 0.7311, + "step": 7450 + }, + { + "epoch": 0.8959297781518668, + "grad_norm": 2.136310322168233, + "learning_rate": 1.1243557479537846e-07, + "loss": 0.6789, + "step": 7451 + }, + { + "epoch": 0.8960500210425059, + "grad_norm": 2.3016501892332406, + "learning_rate": 1.121782149528634e-07, + "loss": 0.6905, + "step": 7452 + }, + { + "epoch": 0.8961702639331449, + "grad_norm": 2.166618517805287, + "learning_rate": 1.1192114149141208e-07, + "loss": 0.793, + "step": 7453 + }, + { + "epoch": 0.8962905068237841, + "grad_norm": 3.2283936262302015, + "learning_rate": 1.1166435445002197e-07, + "loss": 0.6536, + "step": 7454 + }, + { + "epoch": 0.8964107497144231, + "grad_norm": 2.8932293681028187, + "learning_rate": 1.1140785386764818e-07, + "loss": 0.6824, + "step": 7455 + }, + { + "epoch": 0.8965309926050622, + "grad_norm": 1.904887353365025, + "learning_rate": 1.1115163978320153e-07, + "loss": 0.7007, + "step": 7456 + }, + { + "epoch": 0.8966512354957014, + "grad_norm": 2.7377788175268525, + "learning_rate": 1.1089571223554917e-07, + "loss": 0.8305, + "step": 7457 + }, + { + "epoch": 0.8967714783863404, + "grad_norm": 1.8373102777849604, + "learning_rate": 1.1064007126351537e-07, + "loss": 0.8589, + "step": 7458 + }, + { + "epoch": 0.8968917212769795, + "grad_norm": 2.2777471953822657, + "learning_rate": 1.1038471690588003e-07, + "loss": 0.7712, + "step": 7459 + }, + { + "epoch": 0.8970119641676186, + "grad_norm": 1.8907835712612484, + "learning_rate": 1.1012964920138145e-07, + "loss": 0.8047, + "step": 7460 + }, + { + "epoch": 0.8971322070582577, + "grad_norm": 2.104617709423691, + "learning_rate": 1.0987486818871205e-07, + "loss": 0.7632, + "step": 7461 + }, + { + "epoch": 0.8972524499488967, + "grad_norm": 3.2169025637711446, + "learning_rate": 1.0962037390652245e-07, + "loss": 0.7365, + "step": 7462 + }, + { + "epoch": 0.8973726928395359, + "grad_norm": 2.031693395491478, + "learning_rate": 1.0936616639341911e-07, + "loss": 0.7214, + "step": 7463 + }, + { + "epoch": 0.897492935730175, + "grad_norm": 0.8852663623167045, + "learning_rate": 1.0911224568796473e-07, + "loss": 0.5484, + "step": 7464 + }, + { + "epoch": 0.897613178620814, + "grad_norm": 1.8747214428295624, + "learning_rate": 1.0885861182867984e-07, + "loss": 0.71, + "step": 7465 + }, + { + "epoch": 0.8977334215114532, + "grad_norm": 1.82269543713148, + "learning_rate": 1.0860526485403942e-07, + "loss": 0.7103, + "step": 7466 + }, + { + "epoch": 0.8978536644020922, + "grad_norm": 1.6298817885425934, + "learning_rate": 1.0835220480247675e-07, + "loss": 0.7722, + "step": 7467 + }, + { + "epoch": 0.8979739072927313, + "grad_norm": 1.985429438589618, + "learning_rate": 1.0809943171238067e-07, + "loss": 0.8402, + "step": 7468 + }, + { + "epoch": 0.8980941501833704, + "grad_norm": 2.211287831294979, + "learning_rate": 1.078469456220965e-07, + "loss": 0.6348, + "step": 7469 + }, + { + "epoch": 0.8982143930740095, + "grad_norm": 1.7728398866385582, + "learning_rate": 1.0759474656992606e-07, + "loss": 0.7012, + "step": 7470 + }, + { + "epoch": 0.8983346359646486, + "grad_norm": 2.916595421783782, + "learning_rate": 1.0734283459412785e-07, + "loss": 0.7813, + "step": 7471 + }, + { + "epoch": 0.8984548788552876, + "grad_norm": 2.5436348928890626, + "learning_rate": 1.0709120973291707e-07, + "loss": 0.8089, + "step": 7472 + }, + { + "epoch": 0.8985751217459268, + "grad_norm": 2.2772446049722554, + "learning_rate": 1.0683987202446475e-07, + "loss": 0.7796, + "step": 7473 + }, + { + "epoch": 0.8986953646365659, + "grad_norm": 2.0425433268439894, + "learning_rate": 1.0658882150689862e-07, + "loss": 0.7018, + "step": 7474 + }, + { + "epoch": 0.8988156075272049, + "grad_norm": 4.586306399531202, + "learning_rate": 1.0633805821830288e-07, + "loss": 0.7868, + "step": 7475 + }, + { + "epoch": 0.8989358504178441, + "grad_norm": 2.2652195008551477, + "learning_rate": 1.0608758219671753e-07, + "loss": 0.8319, + "step": 7476 + }, + { + "epoch": 0.8990560933084831, + "grad_norm": 1.6815534296590446, + "learning_rate": 1.0583739348014065e-07, + "loss": 0.7096, + "step": 7477 + }, + { + "epoch": 0.8991763361991222, + "grad_norm": 1.8361170904100983, + "learning_rate": 1.0558749210652518e-07, + "loss": 0.8507, + "step": 7478 + }, + { + "epoch": 0.8992965790897613, + "grad_norm": 1.6293378299384702, + "learning_rate": 1.053378781137808e-07, + "loss": 0.8558, + "step": 7479 + }, + { + "epoch": 0.8994168219804004, + "grad_norm": 1.93429879329612, + "learning_rate": 1.0508855153977392e-07, + "loss": 0.7783, + "step": 7480 + }, + { + "epoch": 0.8995370648710395, + "grad_norm": 2.4262181869850314, + "learning_rate": 1.0483951242232669e-07, + "loss": 0.6703, + "step": 7481 + }, + { + "epoch": 0.8996573077616786, + "grad_norm": 0.9822175740642183, + "learning_rate": 1.0459076079921936e-07, + "loss": 0.5772, + "step": 7482 + }, + { + "epoch": 0.8997775506523177, + "grad_norm": 2.2291094217050733, + "learning_rate": 1.0434229670818618e-07, + "loss": 0.8527, + "step": 7483 + }, + { + "epoch": 0.8998977935429567, + "grad_norm": 1.5963277726431349, + "learning_rate": 1.0409412018691944e-07, + "loss": 0.8031, + "step": 7484 + }, + { + "epoch": 0.9000180364335959, + "grad_norm": 2.1338022471703644, + "learning_rate": 1.0384623127306724e-07, + "loss": 0.7555, + "step": 7485 + }, + { + "epoch": 0.900138279324235, + "grad_norm": 1.8449580873727152, + "learning_rate": 1.0359863000423397e-07, + "loss": 0.7946, + "step": 7486 + }, + { + "epoch": 0.900258522214874, + "grad_norm": 1.5605367151724578, + "learning_rate": 1.0335131641798112e-07, + "loss": 0.7198, + "step": 7487 + }, + { + "epoch": 0.9003787651055132, + "grad_norm": 0.8236718555484087, + "learning_rate": 1.0310429055182512e-07, + "loss": 0.5563, + "step": 7488 + }, + { + "epoch": 0.9004990079961522, + "grad_norm": 1.7447362482939477, + "learning_rate": 1.0285755244324024e-07, + "loss": 0.7444, + "step": 7489 + }, + { + "epoch": 0.9006192508867913, + "grad_norm": 1.5139493035832443, + "learning_rate": 1.0261110212965629e-07, + "loss": 0.6883, + "step": 7490 + }, + { + "epoch": 0.9007394937774305, + "grad_norm": 2.025071239395712, + "learning_rate": 1.023649396484596e-07, + "loss": 0.7947, + "step": 7491 + }, + { + "epoch": 0.9008597366680695, + "grad_norm": 2.0284158742007805, + "learning_rate": 1.0211906503699275e-07, + "loss": 0.6816, + "step": 7492 + }, + { + "epoch": 0.9009799795587086, + "grad_norm": 2.340565157890568, + "learning_rate": 1.0187347833255455e-07, + "loss": 0.8262, + "step": 7493 + }, + { + "epoch": 0.9011002224493477, + "grad_norm": 1.818900219458604, + "learning_rate": 1.0162817957240056e-07, + "loss": 0.7956, + "step": 7494 + }, + { + "epoch": 0.9012204653399868, + "grad_norm": 0.8899258328727294, + "learning_rate": 1.0138316879374253e-07, + "loss": 0.6305, + "step": 7495 + }, + { + "epoch": 0.9013407082306258, + "grad_norm": 3.346504431689269, + "learning_rate": 1.0113844603374833e-07, + "loss": 0.7429, + "step": 7496 + }, + { + "epoch": 0.901460951121265, + "grad_norm": 2.2487789655583095, + "learning_rate": 1.0089401132954178e-07, + "loss": 0.7201, + "step": 7497 + }, + { + "epoch": 0.9015811940119041, + "grad_norm": 1.7037583319725451, + "learning_rate": 1.006498647182037e-07, + "loss": 0.7226, + "step": 7498 + }, + { + "epoch": 0.9017014369025431, + "grad_norm": 2.1642093539762124, + "learning_rate": 1.004060062367713e-07, + "loss": 0.7223, + "step": 7499 + }, + { + "epoch": 0.9018216797931822, + "grad_norm": 1.8693607722856627, + "learning_rate": 1.0016243592223728e-07, + "loss": 0.7006, + "step": 7500 + }, + { + "epoch": 0.9019419226838213, + "grad_norm": 2.80227591456678, + "learning_rate": 9.991915381155114e-08, + "loss": 0.6634, + "step": 7501 + }, + { + "epoch": 0.9020621655744604, + "grad_norm": 2.16378738362211, + "learning_rate": 9.967615994161871e-08, + "loss": 0.7566, + "step": 7502 + }, + { + "epoch": 0.9021824084650995, + "grad_norm": 1.7270752848708757, + "learning_rate": 9.943345434930161e-08, + "loss": 0.7837, + "step": 7503 + }, + { + "epoch": 0.9023026513557386, + "grad_norm": 2.3694330524003266, + "learning_rate": 9.919103707141885e-08, + "loss": 0.6951, + "step": 7504 + }, + { + "epoch": 0.9024228942463777, + "grad_norm": 1.8389081580672135, + "learning_rate": 9.89489081447441e-08, + "loss": 0.7717, + "step": 7505 + }, + { + "epoch": 0.9025431371370167, + "grad_norm": 3.0943229992372157, + "learning_rate": 9.870706760600844e-08, + "loss": 0.8351, + "step": 7506 + }, + { + "epoch": 0.9026633800276559, + "grad_norm": 2.03240394093147, + "learning_rate": 9.846551549189918e-08, + "loss": 0.7289, + "step": 7507 + }, + { + "epoch": 0.902783622918295, + "grad_norm": 2.0334588585343463, + "learning_rate": 9.822425183905902e-08, + "loss": 0.6929, + "step": 7508 + }, + { + "epoch": 0.902903865808934, + "grad_norm": 0.9229355546785226, + "learning_rate": 9.798327668408823e-08, + "loss": 0.7523, + "step": 7509 + }, + { + "epoch": 0.9030241086995732, + "grad_norm": 2.1070008657242743, + "learning_rate": 9.774259006354158e-08, + "loss": 0.6943, + "step": 7510 + }, + { + "epoch": 0.9031443515902122, + "grad_norm": 2.493684038327192, + "learning_rate": 9.750219201393184e-08, + "loss": 0.7654, + "step": 7511 + }, + { + "epoch": 0.9032645944808513, + "grad_norm": 2.0268127098710367, + "learning_rate": 9.726208257172697e-08, + "loss": 0.7781, + "step": 7512 + }, + { + "epoch": 0.9033848373714904, + "grad_norm": 1.9215268075518757, + "learning_rate": 9.702226177335115e-08, + "loss": 0.7505, + "step": 7513 + }, + { + "epoch": 0.9035050802621295, + "grad_norm": 1.5675577321592795, + "learning_rate": 9.67827296551853e-08, + "loss": 0.7281, + "step": 7514 + }, + { + "epoch": 0.9036253231527686, + "grad_norm": 2.129808930974442, + "learning_rate": 9.65434862535659e-08, + "loss": 0.6864, + "step": 7515 + }, + { + "epoch": 0.9037455660434077, + "grad_norm": 3.581501396369582, + "learning_rate": 9.630453160478635e-08, + "loss": 0.66, + "step": 7516 + }, + { + "epoch": 0.9038658089340468, + "grad_norm": 1.5818023139233461, + "learning_rate": 9.60658657450959e-08, + "loss": 0.8226, + "step": 7517 + }, + { + "epoch": 0.9039860518246858, + "grad_norm": 2.248418145961319, + "learning_rate": 9.582748871069979e-08, + "loss": 0.7972, + "step": 7518 + }, + { + "epoch": 0.904106294715325, + "grad_norm": 2.7731026460762407, + "learning_rate": 9.558940053775954e-08, + "loss": 0.8361, + "step": 7519 + }, + { + "epoch": 0.904226537605964, + "grad_norm": 2.0588003557968646, + "learning_rate": 9.535160126239294e-08, + "loss": 0.6802, + "step": 7520 + }, + { + "epoch": 0.9043467804966031, + "grad_norm": 1.5969788250698747, + "learning_rate": 9.511409092067424e-08, + "loss": 0.7073, + "step": 7521 + }, + { + "epoch": 0.9044670233872423, + "grad_norm": 2.0924868972714465, + "learning_rate": 9.487686954863327e-08, + "loss": 0.683, + "step": 7522 + }, + { + "epoch": 0.9045872662778813, + "grad_norm": 1.9770493944294654, + "learning_rate": 9.46399371822566e-08, + "loss": 0.7771, + "step": 7523 + }, + { + "epoch": 0.9047075091685204, + "grad_norm": 2.1153221781392735, + "learning_rate": 9.440329385748657e-08, + "loss": 0.732, + "step": 7524 + }, + { + "epoch": 0.9048277520591596, + "grad_norm": 1.8054925353308715, + "learning_rate": 9.416693961022137e-08, + "loss": 0.7114, + "step": 7525 + }, + { + "epoch": 0.9049479949497986, + "grad_norm": 1.6785101024242988, + "learning_rate": 9.393087447631654e-08, + "loss": 0.772, + "step": 7526 + }, + { + "epoch": 0.9050682378404377, + "grad_norm": 1.7420737738332963, + "learning_rate": 9.36950984915823e-08, + "loss": 0.7285, + "step": 7527 + }, + { + "epoch": 0.9051884807310768, + "grad_norm": 1.8019275120160025, + "learning_rate": 9.345961169178607e-08, + "loss": 0.6954, + "step": 7528 + }, + { + "epoch": 0.9053087236217159, + "grad_norm": 1.555171258701234, + "learning_rate": 9.322441411265081e-08, + "loss": 0.7268, + "step": 7529 + }, + { + "epoch": 0.9054289665123549, + "grad_norm": 1.871292263004778, + "learning_rate": 9.298950578985554e-08, + "loss": 0.7351, + "step": 7530 + }, + { + "epoch": 0.905549209402994, + "grad_norm": 1.900601444920068, + "learning_rate": 9.275488675903665e-08, + "loss": 0.7097, + "step": 7531 + }, + { + "epoch": 0.9056694522936332, + "grad_norm": 2.083207762430674, + "learning_rate": 9.252055705578454e-08, + "loss": 0.7439, + "step": 7532 + }, + { + "epoch": 0.9057896951842722, + "grad_norm": 1.788063136673647, + "learning_rate": 9.228651671564747e-08, + "loss": 0.721, + "step": 7533 + }, + { + "epoch": 0.9059099380749113, + "grad_norm": 1.6094600321703576, + "learning_rate": 9.205276577412901e-08, + "loss": 0.7795, + "step": 7534 + }, + { + "epoch": 0.9060301809655504, + "grad_norm": 2.514250101187147, + "learning_rate": 9.181930426668905e-08, + "loss": 0.7716, + "step": 7535 + }, + { + "epoch": 0.9061504238561895, + "grad_norm": 1.5770132095422278, + "learning_rate": 9.158613222874346e-08, + "loss": 0.6777, + "step": 7536 + }, + { + "epoch": 0.9062706667468285, + "grad_norm": 1.6500520437876018, + "learning_rate": 9.135324969566394e-08, + "loss": 0.825, + "step": 7537 + }, + { + "epoch": 0.9063909096374677, + "grad_norm": 2.0303052612570207, + "learning_rate": 9.112065670277913e-08, + "loss": 0.7612, + "step": 7538 + }, + { + "epoch": 0.9065111525281068, + "grad_norm": 1.9950034192660675, + "learning_rate": 9.088835328537303e-08, + "loss": 0.7296, + "step": 7539 + }, + { + "epoch": 0.9066313954187458, + "grad_norm": 2.491937606863186, + "learning_rate": 9.065633947868568e-08, + "loss": 0.7223, + "step": 7540 + }, + { + "epoch": 0.906751638309385, + "grad_norm": 2.2545533201774024, + "learning_rate": 9.042461531791379e-08, + "loss": 0.8036, + "step": 7541 + }, + { + "epoch": 0.906871881200024, + "grad_norm": 1.6093743333537378, + "learning_rate": 9.019318083820903e-08, + "loss": 0.7791, + "step": 7542 + }, + { + "epoch": 0.9069921240906631, + "grad_norm": 2.002796840121407, + "learning_rate": 8.996203607468045e-08, + "loss": 0.8561, + "step": 7543 + }, + { + "epoch": 0.9071123669813023, + "grad_norm": 1.5234799110182329, + "learning_rate": 8.973118106239241e-08, + "loss": 0.756, + "step": 7544 + }, + { + "epoch": 0.9072326098719413, + "grad_norm": 2.0138456785747656, + "learning_rate": 8.95006158363656e-08, + "loss": 0.9482, + "step": 7545 + }, + { + "epoch": 0.9073528527625804, + "grad_norm": 1.9548883031561606, + "learning_rate": 8.9270340431576e-08, + "loss": 0.7788, + "step": 7546 + }, + { + "epoch": 0.9074730956532195, + "grad_norm": 2.071610529664056, + "learning_rate": 8.904035488295658e-08, + "loss": 0.7386, + "step": 7547 + }, + { + "epoch": 0.9075933385438586, + "grad_norm": 0.6790161544285451, + "learning_rate": 8.881065922539632e-08, + "loss": 0.5329, + "step": 7548 + }, + { + "epoch": 0.9077135814344977, + "grad_norm": 1.9010843420246597, + "learning_rate": 8.85812534937389e-08, + "loss": 0.7422, + "step": 7549 + }, + { + "epoch": 0.9078338243251368, + "grad_norm": 2.7825119070396416, + "learning_rate": 8.835213772278583e-08, + "loss": 0.6827, + "step": 7550 + }, + { + "epoch": 0.9079540672157759, + "grad_norm": 2.1170926794236107, + "learning_rate": 8.812331194729373e-08, + "loss": 0.7965, + "step": 7551 + }, + { + "epoch": 0.9080743101064149, + "grad_norm": 1.8741089929052173, + "learning_rate": 8.789477620197461e-08, + "loss": 0.7264, + "step": 7552 + }, + { + "epoch": 0.9081945529970541, + "grad_norm": 2.436582147965001, + "learning_rate": 8.766653052149831e-08, + "loss": 0.7948, + "step": 7553 + }, + { + "epoch": 0.9083147958876931, + "grad_norm": 2.0823344569908757, + "learning_rate": 8.743857494048823e-08, + "loss": 0.7508, + "step": 7554 + }, + { + "epoch": 0.9084350387783322, + "grad_norm": 1.9996859184648759, + "learning_rate": 8.721090949352605e-08, + "loss": 0.628, + "step": 7555 + }, + { + "epoch": 0.9085552816689714, + "grad_norm": 1.87862313213404, + "learning_rate": 8.698353421514793e-08, + "loss": 0.7373, + "step": 7556 + }, + { + "epoch": 0.9086755245596104, + "grad_norm": 2.015817085990175, + "learning_rate": 8.67564491398467e-08, + "loss": 0.8008, + "step": 7557 + }, + { + "epoch": 0.9087957674502495, + "grad_norm": 1.8011513948155216, + "learning_rate": 8.652965430207104e-08, + "loss": 0.7402, + "step": 7558 + }, + { + "epoch": 0.9089160103408886, + "grad_norm": 2.1232326113986977, + "learning_rate": 8.630314973622521e-08, + "loss": 0.6595, + "step": 7559 + }, + { + "epoch": 0.9090362532315277, + "grad_norm": 1.8162460420569333, + "learning_rate": 8.607693547666995e-08, + "loss": 0.7128, + "step": 7560 + }, + { + "epoch": 0.9091564961221668, + "grad_norm": 0.9008443686214899, + "learning_rate": 8.585101155772201e-08, + "loss": 0.5805, + "step": 7561 + }, + { + "epoch": 0.9092767390128058, + "grad_norm": 1.8162825347443763, + "learning_rate": 8.562537801365377e-08, + "loss": 0.6907, + "step": 7562 + }, + { + "epoch": 0.909396981903445, + "grad_norm": 2.0345453381458944, + "learning_rate": 8.540003487869362e-08, + "loss": 0.7048, + "step": 7563 + }, + { + "epoch": 0.909517224794084, + "grad_norm": 1.9735986856702827, + "learning_rate": 8.517498218702557e-08, + "loss": 0.7975, + "step": 7564 + }, + { + "epoch": 0.9096374676847231, + "grad_norm": 1.7309082024362905, + "learning_rate": 8.49502199727905e-08, + "loss": 0.6954, + "step": 7565 + }, + { + "epoch": 0.9097577105753623, + "grad_norm": 2.2859434157716727, + "learning_rate": 8.472574827008428e-08, + "loss": 0.668, + "step": 7566 + }, + { + "epoch": 0.9098779534660013, + "grad_norm": 2.1548212818996126, + "learning_rate": 8.450156711295942e-08, + "loss": 0.8434, + "step": 7567 + }, + { + "epoch": 0.9099981963566404, + "grad_norm": 2.131775991913716, + "learning_rate": 8.427767653542383e-08, + "loss": 0.8676, + "step": 7568 + }, + { + "epoch": 0.9101184392472795, + "grad_norm": 2.108655227494225, + "learning_rate": 8.405407657144125e-08, + "loss": 0.7026, + "step": 7569 + }, + { + "epoch": 0.9102386821379186, + "grad_norm": 1.7984431845859334, + "learning_rate": 8.383076725493232e-08, + "loss": 0.7255, + "step": 7570 + }, + { + "epoch": 0.9103589250285576, + "grad_norm": 2.011437138966483, + "learning_rate": 8.360774861977216e-08, + "loss": 0.6832, + "step": 7571 + }, + { + "epoch": 0.9104791679191968, + "grad_norm": 2.0302245853009104, + "learning_rate": 8.338502069979281e-08, + "loss": 0.7457, + "step": 7572 + }, + { + "epoch": 0.9105994108098359, + "grad_norm": 29.474077635125518, + "learning_rate": 8.316258352878214e-08, + "loss": 0.8064, + "step": 7573 + }, + { + "epoch": 0.9107196537004749, + "grad_norm": 1.8387879016737954, + "learning_rate": 8.294043714048338e-08, + "loss": 0.7135, + "step": 7574 + }, + { + "epoch": 0.9108398965911141, + "grad_norm": 0.7664648237797781, + "learning_rate": 8.271858156859624e-08, + "loss": 0.6049, + "step": 7575 + }, + { + "epoch": 0.9109601394817531, + "grad_norm": 1.86689883758069, + "learning_rate": 8.249701684677557e-08, + "loss": 0.7427, + "step": 7576 + }, + { + "epoch": 0.9110803823723922, + "grad_norm": 1.8793103582532449, + "learning_rate": 8.227574300863294e-08, + "loss": 0.8129, + "step": 7577 + }, + { + "epoch": 0.9112006252630314, + "grad_norm": 1.6485935516979167, + "learning_rate": 8.205476008773548e-08, + "loss": 0.6991, + "step": 7578 + }, + { + "epoch": 0.9113208681536704, + "grad_norm": 1.9495169310224678, + "learning_rate": 8.183406811760596e-08, + "loss": 0.829, + "step": 7579 + }, + { + "epoch": 0.9114411110443095, + "grad_norm": 1.5104531996244093, + "learning_rate": 8.161366713172313e-08, + "loss": 0.741, + "step": 7580 + }, + { + "epoch": 0.9115613539349486, + "grad_norm": 2.607680815794093, + "learning_rate": 8.139355716352137e-08, + "loss": 0.8421, + "step": 7581 + }, + { + "epoch": 0.9116815968255877, + "grad_norm": 1.6293672417037857, + "learning_rate": 8.117373824639196e-08, + "loss": 0.6995, + "step": 7582 + }, + { + "epoch": 0.9118018397162267, + "grad_norm": 0.7334162358182788, + "learning_rate": 8.095421041368067e-08, + "loss": 0.5927, + "step": 7583 + }, + { + "epoch": 0.9119220826068659, + "grad_norm": 1.865236164511576, + "learning_rate": 8.073497369868999e-08, + "loss": 0.7052, + "step": 7584 + }, + { + "epoch": 0.912042325497505, + "grad_norm": 1.8016571099483365, + "learning_rate": 8.051602813467772e-08, + "loss": 0.7594, + "step": 7585 + }, + { + "epoch": 0.912162568388144, + "grad_norm": 1.6069620060966239, + "learning_rate": 8.029737375485756e-08, + "loss": 0.7136, + "step": 7586 + }, + { + "epoch": 0.9122828112787832, + "grad_norm": 1.8124005520602775, + "learning_rate": 8.007901059239986e-08, + "loss": 0.728, + "step": 7587 + }, + { + "epoch": 0.9124030541694222, + "grad_norm": 1.6684190914941188, + "learning_rate": 7.986093868042964e-08, + "loss": 0.808, + "step": 7588 + }, + { + "epoch": 0.9125232970600613, + "grad_norm": 1.9399285631693233, + "learning_rate": 7.964315805202826e-08, + "loss": 0.683, + "step": 7589 + }, + { + "epoch": 0.9126435399507005, + "grad_norm": 2.214298349335655, + "learning_rate": 7.942566874023304e-08, + "loss": 0.737, + "step": 7590 + }, + { + "epoch": 0.9127637828413395, + "grad_norm": 2.293740761553794, + "learning_rate": 7.920847077803649e-08, + "loss": 0.6998, + "step": 7591 + }, + { + "epoch": 0.9128840257319786, + "grad_norm": 1.8102518147813889, + "learning_rate": 7.899156419838826e-08, + "loss": 0.8252, + "step": 7592 + }, + { + "epoch": 0.9130042686226177, + "grad_norm": 1.950794263619864, + "learning_rate": 7.87749490341918e-08, + "loss": 0.6588, + "step": 7593 + }, + { + "epoch": 0.9131245115132568, + "grad_norm": 1.8797937149137596, + "learning_rate": 7.855862531830836e-08, + "loss": 0.8352, + "step": 7594 + }, + { + "epoch": 0.9132447544038959, + "grad_norm": 1.7077616999016938, + "learning_rate": 7.834259308355373e-08, + "loss": 0.7306, + "step": 7595 + }, + { + "epoch": 0.9133649972945349, + "grad_norm": 1.9833794446868822, + "learning_rate": 7.812685236269989e-08, + "loss": 0.7497, + "step": 7596 + }, + { + "epoch": 0.9134852401851741, + "grad_norm": 0.8063866906028981, + "learning_rate": 7.791140318847445e-08, + "loss": 0.5872, + "step": 7597 + }, + { + "epoch": 0.9136054830758131, + "grad_norm": 1.504901665551519, + "learning_rate": 7.769624559356081e-08, + "loss": 0.8025, + "step": 7598 + }, + { + "epoch": 0.9137257259664522, + "grad_norm": 3.3568929329923853, + "learning_rate": 7.748137961059842e-08, + "loss": 0.759, + "step": 7599 + }, + { + "epoch": 0.9138459688570914, + "grad_norm": 2.368367017720742, + "learning_rate": 7.726680527218211e-08, + "loss": 0.6656, + "step": 7600 + }, + { + "epoch": 0.9139662117477304, + "grad_norm": 1.8187376749466608, + "learning_rate": 7.70525226108627e-08, + "loss": 0.7591, + "step": 7601 + }, + { + "epoch": 0.9140864546383695, + "grad_norm": 1.886322927714677, + "learning_rate": 7.683853165914666e-08, + "loss": 0.8047, + "step": 7602 + }, + { + "epoch": 0.9142066975290086, + "grad_norm": 1.7549014931662583, + "learning_rate": 7.662483244949602e-08, + "loss": 0.7746, + "step": 7603 + }, + { + "epoch": 0.9143269404196477, + "grad_norm": 2.7772895818436427, + "learning_rate": 7.641142501432951e-08, + "loss": 0.8141, + "step": 7604 + }, + { + "epoch": 0.9144471833102867, + "grad_norm": 1.6324112686648813, + "learning_rate": 7.619830938602013e-08, + "loss": 0.7374, + "step": 7605 + }, + { + "epoch": 0.9145674262009259, + "grad_norm": 1.9792587526548764, + "learning_rate": 7.598548559689777e-08, + "loss": 0.8278, + "step": 7606 + }, + { + "epoch": 0.914687669091565, + "grad_norm": 2.7381321843341895, + "learning_rate": 7.577295367924751e-08, + "loss": 0.8136, + "step": 7607 + }, + { + "epoch": 0.914807911982204, + "grad_norm": 2.8814920849165997, + "learning_rate": 7.556071366531002e-08, + "loss": 0.8271, + "step": 7608 + }, + { + "epoch": 0.9149281548728432, + "grad_norm": 2.592840931619155, + "learning_rate": 7.53487655872822e-08, + "loss": 0.7953, + "step": 7609 + }, + { + "epoch": 0.9150483977634822, + "grad_norm": 1.7144542764489041, + "learning_rate": 7.513710947731656e-08, + "loss": 0.743, + "step": 7610 + }, + { + "epoch": 0.9151686406541213, + "grad_norm": 1.8428686589401873, + "learning_rate": 7.492574536752095e-08, + "loss": 0.8526, + "step": 7611 + }, + { + "epoch": 0.9152888835447605, + "grad_norm": 2.173365897693333, + "learning_rate": 7.471467328995907e-08, + "loss": 0.7812, + "step": 7612 + }, + { + "epoch": 0.9154091264353995, + "grad_norm": 2.4108330813408974, + "learning_rate": 7.450389327665018e-08, + "loss": 0.6065, + "step": 7613 + }, + { + "epoch": 0.9155293693260386, + "grad_norm": 2.41808805648967, + "learning_rate": 7.429340535957029e-08, + "loss": 0.6818, + "step": 7614 + }, + { + "epoch": 0.9156496122166777, + "grad_norm": 2.6668067736377004, + "learning_rate": 7.40832095706494e-08, + "loss": 0.7098, + "step": 7615 + }, + { + "epoch": 0.9157698551073168, + "grad_norm": 1.8854769140351135, + "learning_rate": 7.387330594177443e-08, + "loss": 0.8041, + "step": 7616 + }, + { + "epoch": 0.9158900979979558, + "grad_norm": 1.6029137842191927, + "learning_rate": 7.366369450478749e-08, + "loss": 0.7962, + "step": 7617 + }, + { + "epoch": 0.916010340888595, + "grad_norm": 1.6843687605874986, + "learning_rate": 7.345437529148646e-08, + "loss": 0.6659, + "step": 7618 + }, + { + "epoch": 0.9161305837792341, + "grad_norm": 2.3524841327517243, + "learning_rate": 7.324534833362483e-08, + "loss": 0.7252, + "step": 7619 + }, + { + "epoch": 0.9162508266698731, + "grad_norm": 1.9579170339188519, + "learning_rate": 7.303661366291192e-08, + "loss": 0.6839, + "step": 7620 + }, + { + "epoch": 0.9163710695605123, + "grad_norm": 3.650116430019425, + "learning_rate": 7.28281713110126e-08, + "loss": 0.8193, + "step": 7621 + }, + { + "epoch": 0.9164913124511513, + "grad_norm": 2.067124701574714, + "learning_rate": 7.262002130954759e-08, + "loss": 0.7757, + "step": 7622 + }, + { + "epoch": 0.9166115553417904, + "grad_norm": 1.7901400190604635, + "learning_rate": 7.241216369009296e-08, + "loss": 0.7926, + "step": 7623 + }, + { + "epoch": 0.9167317982324296, + "grad_norm": 1.849001952897026, + "learning_rate": 7.220459848418037e-08, + "loss": 0.6693, + "step": 7624 + }, + { + "epoch": 0.9168520411230686, + "grad_norm": 1.9754553547581322, + "learning_rate": 7.199732572329708e-08, + "loss": 0.8019, + "step": 7625 + }, + { + "epoch": 0.9169722840137077, + "grad_norm": 2.2679545035129967, + "learning_rate": 7.179034543888684e-08, + "loss": 0.765, + "step": 7626 + }, + { + "epoch": 0.9170925269043467, + "grad_norm": 2.1037154390575385, + "learning_rate": 7.158365766234808e-08, + "loss": 0.7762, + "step": 7627 + }, + { + "epoch": 0.9172127697949859, + "grad_norm": 1.865976747104219, + "learning_rate": 7.137726242503527e-08, + "loss": 0.7273, + "step": 7628 + }, + { + "epoch": 0.917333012685625, + "grad_norm": 2.6414846792821964, + "learning_rate": 7.11711597582585e-08, + "loss": 0.7888, + "step": 7629 + }, + { + "epoch": 0.917453255576264, + "grad_norm": 1.7790027965804576, + "learning_rate": 7.096534969328271e-08, + "loss": 0.8019, + "step": 7630 + }, + { + "epoch": 0.9175734984669032, + "grad_norm": 2.1029732092485482, + "learning_rate": 7.075983226132987e-08, + "loss": 0.8449, + "step": 7631 + }, + { + "epoch": 0.9176937413575422, + "grad_norm": 2.695767659994526, + "learning_rate": 7.055460749357656e-08, + "loss": 0.8021, + "step": 7632 + }, + { + "epoch": 0.9178139842481813, + "grad_norm": 1.8546425997948048, + "learning_rate": 7.034967542115521e-08, + "loss": 0.7052, + "step": 7633 + }, + { + "epoch": 0.9179342271388204, + "grad_norm": 2.3461675763780714, + "learning_rate": 7.014503607515388e-08, + "loss": 0.7564, + "step": 7634 + }, + { + "epoch": 0.9180544700294595, + "grad_norm": 2.0390646046594743, + "learning_rate": 6.994068948661592e-08, + "loss": 0.6848, + "step": 7635 + }, + { + "epoch": 0.9181747129200986, + "grad_norm": 2.013922156866659, + "learning_rate": 6.973663568654142e-08, + "loss": 0.771, + "step": 7636 + }, + { + "epoch": 0.9182949558107377, + "grad_norm": 2.104145377274887, + "learning_rate": 6.953287470588386e-08, + "loss": 0.6605, + "step": 7637 + }, + { + "epoch": 0.9184151987013768, + "grad_norm": 2.464114992771506, + "learning_rate": 6.932940657555452e-08, + "loss": 0.8635, + "step": 7638 + }, + { + "epoch": 0.9185354415920158, + "grad_norm": 1.456093248436198, + "learning_rate": 6.912623132641938e-08, + "loss": 0.7687, + "step": 7639 + }, + { + "epoch": 0.918655684482655, + "grad_norm": 1.8452194306679228, + "learning_rate": 6.892334898929952e-08, + "loss": 0.7707, + "step": 7640 + }, + { + "epoch": 0.918775927373294, + "grad_norm": 1.9230108858654396, + "learning_rate": 6.872075959497236e-08, + "loss": 0.8497, + "step": 7641 + }, + { + "epoch": 0.9188961702639331, + "grad_norm": 1.8037536318630736, + "learning_rate": 6.85184631741702e-08, + "loss": 0.8345, + "step": 7642 + }, + { + "epoch": 0.9190164131545723, + "grad_norm": 2.0037158778968482, + "learning_rate": 6.831645975758161e-08, + "loss": 0.7821, + "step": 7643 + }, + { + "epoch": 0.9191366560452113, + "grad_norm": 1.8155675777212097, + "learning_rate": 6.811474937585026e-08, + "loss": 0.6754, + "step": 7644 + }, + { + "epoch": 0.9192568989358504, + "grad_norm": 1.6166560824857235, + "learning_rate": 6.79133320595755e-08, + "loss": 0.8016, + "step": 7645 + }, + { + "epoch": 0.9193771418264896, + "grad_norm": 2.2591595178948567, + "learning_rate": 6.771220783931198e-08, + "loss": 0.7612, + "step": 7646 + }, + { + "epoch": 0.9194973847171286, + "grad_norm": 0.8532195810841308, + "learning_rate": 6.751137674556994e-08, + "loss": 0.6466, + "step": 7647 + }, + { + "epoch": 0.9196176276077677, + "grad_norm": 2.391068261633953, + "learning_rate": 6.731083880881572e-08, + "loss": 0.7769, + "step": 7648 + }, + { + "epoch": 0.9197378704984068, + "grad_norm": 2.1638969055057755, + "learning_rate": 6.711059405947072e-08, + "loss": 0.8129, + "step": 7649 + }, + { + "epoch": 0.9198581133890459, + "grad_norm": 1.6764697814502487, + "learning_rate": 6.691064252791156e-08, + "loss": 0.7771, + "step": 7650 + }, + { + "epoch": 0.9199783562796849, + "grad_norm": 1.5660490546746417, + "learning_rate": 6.67109842444713e-08, + "loss": 0.7825, + "step": 7651 + }, + { + "epoch": 0.9200985991703241, + "grad_norm": 2.2595977588971317, + "learning_rate": 6.651161923943704e-08, + "loss": 0.7724, + "step": 7652 + }, + { + "epoch": 0.9202188420609632, + "grad_norm": 1.8233848479186683, + "learning_rate": 6.631254754305326e-08, + "loss": 0.7759, + "step": 7653 + }, + { + "epoch": 0.9203390849516022, + "grad_norm": 4.464383212523508, + "learning_rate": 6.611376918551848e-08, + "loss": 0.7813, + "step": 7654 + }, + { + "epoch": 0.9204593278422414, + "grad_norm": 1.9552039716614686, + "learning_rate": 6.591528419698744e-08, + "loss": 0.8004, + "step": 7655 + }, + { + "epoch": 0.9205795707328804, + "grad_norm": 2.2944056932791033, + "learning_rate": 6.571709260756986e-08, + "loss": 0.8358, + "step": 7656 + }, + { + "epoch": 0.9206998136235195, + "grad_norm": 2.3722271040966216, + "learning_rate": 6.551919444733122e-08, + "loss": 0.7691, + "step": 7657 + }, + { + "epoch": 0.9208200565141585, + "grad_norm": 2.489757920106375, + "learning_rate": 6.53215897462931e-08, + "loss": 0.657, + "step": 7658 + }, + { + "epoch": 0.9209402994047977, + "grad_norm": 2.3401441174330366, + "learning_rate": 6.512427853443103e-08, + "loss": 0.7513, + "step": 7659 + }, + { + "epoch": 0.9210605422954368, + "grad_norm": 1.759731347430799, + "learning_rate": 6.492726084167799e-08, + "loss": 0.7553, + "step": 7660 + }, + { + "epoch": 0.9211807851860758, + "grad_norm": 0.7872382879588307, + "learning_rate": 6.473053669792072e-08, + "loss": 0.5752, + "step": 7661 + }, + { + "epoch": 0.921301028076715, + "grad_norm": 2.0003692762877843, + "learning_rate": 6.453410613300248e-08, + "loss": 0.7266, + "step": 7662 + }, + { + "epoch": 0.921421270967354, + "grad_norm": 1.6827281310910813, + "learning_rate": 6.43379691767214e-08, + "loss": 0.5872, + "step": 7663 + }, + { + "epoch": 0.9215415138579931, + "grad_norm": 0.7275709311654872, + "learning_rate": 6.414212585883105e-08, + "loss": 0.5532, + "step": 7664 + }, + { + "epoch": 0.9216617567486323, + "grad_norm": 2.4950786789163892, + "learning_rate": 6.394657620904143e-08, + "loss": 0.7028, + "step": 7665 + }, + { + "epoch": 0.9217819996392713, + "grad_norm": 1.975683989568877, + "learning_rate": 6.375132025701657e-08, + "loss": 0.7224, + "step": 7666 + }, + { + "epoch": 0.9219022425299104, + "grad_norm": 2.3595410219099664, + "learning_rate": 6.355635803237724e-08, + "loss": 0.6983, + "step": 7667 + }, + { + "epoch": 0.9220224854205495, + "grad_norm": 2.868398800298046, + "learning_rate": 6.336168956469867e-08, + "loss": 0.8018, + "step": 7668 + }, + { + "epoch": 0.9221427283111886, + "grad_norm": 1.8987238789633243, + "learning_rate": 6.316731488351168e-08, + "loss": 0.7203, + "step": 7669 + }, + { + "epoch": 0.9222629712018277, + "grad_norm": 1.7365770536989822, + "learning_rate": 6.297323401830334e-08, + "loss": 0.6345, + "step": 7670 + }, + { + "epoch": 0.9223832140924668, + "grad_norm": 2.1214735277741816, + "learning_rate": 6.277944699851523e-08, + "loss": 0.6879, + "step": 7671 + }, + { + "epoch": 0.9225034569831059, + "grad_norm": 1.8105119127768154, + "learning_rate": 6.25859538535447e-08, + "loss": 0.735, + "step": 7672 + }, + { + "epoch": 0.9226236998737449, + "grad_norm": 2.6307865124236507, + "learning_rate": 6.239275461274474e-08, + "loss": 0.7865, + "step": 7673 + }, + { + "epoch": 0.9227439427643841, + "grad_norm": 1.918108342640015, + "learning_rate": 6.219984930542299e-08, + "loss": 0.8626, + "step": 7674 + }, + { + "epoch": 0.9228641856550232, + "grad_norm": 2.615469297380172, + "learning_rate": 6.200723796084383e-08, + "loss": 0.7584, + "step": 7675 + }, + { + "epoch": 0.9229844285456622, + "grad_norm": 0.7718952099088718, + "learning_rate": 6.181492060822546e-08, + "loss": 0.6305, + "step": 7676 + }, + { + "epoch": 0.9231046714363014, + "grad_norm": 2.0462484784082755, + "learning_rate": 6.162289727674274e-08, + "loss": 0.8199, + "step": 7677 + }, + { + "epoch": 0.9232249143269404, + "grad_norm": 2.0071468214568964, + "learning_rate": 6.143116799552527e-08, + "loss": 0.8803, + "step": 7678 + }, + { + "epoch": 0.9233451572175795, + "grad_norm": 2.610099646175193, + "learning_rate": 6.123973279365802e-08, + "loss": 0.5633, + "step": 7679 + }, + { + "epoch": 0.9234654001082186, + "grad_norm": 1.7619692334661796, + "learning_rate": 6.10485917001824e-08, + "loss": 0.7773, + "step": 7680 + }, + { + "epoch": 0.9235856429988577, + "grad_norm": 3.2547601905063726, + "learning_rate": 6.085774474409322e-08, + "loss": 0.8118, + "step": 7681 + }, + { + "epoch": 0.9237058858894968, + "grad_norm": 2.478166030986671, + "learning_rate": 6.066719195434267e-08, + "loss": 0.702, + "step": 7682 + }, + { + "epoch": 0.9238261287801359, + "grad_norm": 4.5714523067879975, + "learning_rate": 6.047693335983717e-08, + "loss": 0.6722, + "step": 7683 + }, + { + "epoch": 0.923946371670775, + "grad_norm": 2.501820854841865, + "learning_rate": 6.028696898943853e-08, + "loss": 0.8245, + "step": 7684 + }, + { + "epoch": 0.924066614561414, + "grad_norm": 2.18344033782281, + "learning_rate": 6.00972988719648e-08, + "loss": 0.7083, + "step": 7685 + }, + { + "epoch": 0.9241868574520532, + "grad_norm": 2.681835681464375, + "learning_rate": 5.990792303618807e-08, + "loss": 0.7074, + "step": 7686 + }, + { + "epoch": 0.9243071003426923, + "grad_norm": 2.3030522121485486, + "learning_rate": 5.971884151083695e-08, + "loss": 0.6959, + "step": 7687 + }, + { + "epoch": 0.9244273432333313, + "grad_norm": 1.8869609795737265, + "learning_rate": 5.9530054324595124e-08, + "loss": 0.7452, + "step": 7688 + }, + { + "epoch": 0.9245475861239704, + "grad_norm": 0.7310256467121865, + "learning_rate": 5.934156150610103e-08, + "loss": 0.5755, + "step": 7689 + }, + { + "epoch": 0.9246678290146095, + "grad_norm": 4.512628002940185, + "learning_rate": 5.915336308394914e-08, + "loss": 0.7908, + "step": 7690 + }, + { + "epoch": 0.9247880719052486, + "grad_norm": 1.6593661234415287, + "learning_rate": 5.89654590866886e-08, + "loss": 0.7742, + "step": 7691 + }, + { + "epoch": 0.9249083147958876, + "grad_norm": 1.8995007963407313, + "learning_rate": 5.877784954282483e-08, + "loss": 0.8877, + "step": 7692 + }, + { + "epoch": 0.9250285576865268, + "grad_norm": 5.25299931182363, + "learning_rate": 5.8590534480817963e-08, + "loss": 0.7228, + "step": 7693 + }, + { + "epoch": 0.9251488005771659, + "grad_norm": 2.935018262140234, + "learning_rate": 5.840351392908349e-08, + "loss": 0.7293, + "step": 7694 + }, + { + "epoch": 0.9252690434678049, + "grad_norm": 4.559930992608181, + "learning_rate": 5.821678791599205e-08, + "loss": 0.7162, + "step": 7695 + }, + { + "epoch": 0.9253892863584441, + "grad_norm": 1.801179324688766, + "learning_rate": 5.803035646986965e-08, + "loss": 0.8089, + "step": 7696 + }, + { + "epoch": 0.9255095292490831, + "grad_norm": 2.5765075233567276, + "learning_rate": 5.7844219618998766e-08, + "loss": 0.673, + "step": 7697 + }, + { + "epoch": 0.9256297721397222, + "grad_norm": 1.770048827882547, + "learning_rate": 5.765837739161505e-08, + "loss": 0.7154, + "step": 7698 + }, + { + "epoch": 0.9257500150303614, + "grad_norm": 1.757705210484429, + "learning_rate": 5.7472829815911504e-08, + "loss": 0.7466, + "step": 7699 + }, + { + "epoch": 0.9258702579210004, + "grad_norm": 2.2983054552976263, + "learning_rate": 5.7287576920035164e-08, + "loss": 0.8144, + "step": 7700 + }, + { + "epoch": 0.9259905008116395, + "grad_norm": 1.9945816314585878, + "learning_rate": 5.7102618732088435e-08, + "loss": 0.775, + "step": 7701 + }, + { + "epoch": 0.9261107437022786, + "grad_norm": 1.6660607031104764, + "learning_rate": 5.6917955280130216e-08, + "loss": 0.7503, + "step": 7702 + }, + { + "epoch": 0.9262309865929177, + "grad_norm": 1.9931932491126156, + "learning_rate": 5.6733586592172755e-08, + "loss": 0.7217, + "step": 7703 + }, + { + "epoch": 0.9263512294835567, + "grad_norm": 1.9522734960255161, + "learning_rate": 5.6549512696185244e-08, + "loss": 0.8008, + "step": 7704 + }, + { + "epoch": 0.9264714723741959, + "grad_norm": 1.8127355751323306, + "learning_rate": 5.636573362009156e-08, + "loss": 0.6895, + "step": 7705 + }, + { + "epoch": 0.926591715264835, + "grad_norm": 2.0992072425586508, + "learning_rate": 5.618224939177074e-08, + "loss": 0.7713, + "step": 7706 + }, + { + "epoch": 0.926711958155474, + "grad_norm": 1.683580013493483, + "learning_rate": 5.599906003905719e-08, + "loss": 0.7009, + "step": 7707 + }, + { + "epoch": 0.9268322010461132, + "grad_norm": 2.245113152114142, + "learning_rate": 5.581616558974023e-08, + "loss": 0.8172, + "step": 7708 + }, + { + "epoch": 0.9269524439367522, + "grad_norm": 1.9509964597739318, + "learning_rate": 5.5633566071565444e-08, + "loss": 0.7895, + "step": 7709 + }, + { + "epoch": 0.9270726868273913, + "grad_norm": 2.038747361191399, + "learning_rate": 5.5451261512232896e-08, + "loss": 0.7104, + "step": 7710 + }, + { + "epoch": 0.9271929297180305, + "grad_norm": 1.8809018392710952, + "learning_rate": 5.5269251939397576e-08, + "loss": 0.6295, + "step": 7711 + }, + { + "epoch": 0.9273131726086695, + "grad_norm": 2.1907497789486645, + "learning_rate": 5.508753738067073e-08, + "loss": 0.7673, + "step": 7712 + }, + { + "epoch": 0.9274334154993086, + "grad_norm": 2.947515741230198, + "learning_rate": 5.4906117863617875e-08, + "loss": 0.7964, + "step": 7713 + }, + { + "epoch": 0.9275536583899477, + "grad_norm": 1.866020125011746, + "learning_rate": 5.4724993415760533e-08, + "loss": 0.7816, + "step": 7714 + }, + { + "epoch": 0.9276739012805868, + "grad_norm": 2.5483532693439455, + "learning_rate": 5.454416406457496e-08, + "loss": 0.749, + "step": 7715 + }, + { + "epoch": 0.9277941441712259, + "grad_norm": 3.109424073105219, + "learning_rate": 5.436362983749299e-08, + "loss": 0.7404, + "step": 7716 + }, + { + "epoch": 0.927914387061865, + "grad_norm": 1.986524003984901, + "learning_rate": 5.418339076190137e-08, + "loss": 0.6521, + "step": 7717 + }, + { + "epoch": 0.9280346299525041, + "grad_norm": 1.7790994880137345, + "learning_rate": 5.400344686514202e-08, + "loss": 0.8908, + "step": 7718 + }, + { + "epoch": 0.9281548728431431, + "grad_norm": 2.459496648511253, + "learning_rate": 5.38237981745131e-08, + "loss": 0.6668, + "step": 7719 + }, + { + "epoch": 0.9282751157337822, + "grad_norm": 2.177987916154504, + "learning_rate": 5.364444471726592e-08, + "loss": 0.816, + "step": 7720 + }, + { + "epoch": 0.9283953586244214, + "grad_norm": 2.5186694186740675, + "learning_rate": 5.346538652060939e-08, + "loss": 0.8042, + "step": 7721 + }, + { + "epoch": 0.9285156015150604, + "grad_norm": 2.426785321146904, + "learning_rate": 5.3286623611705994e-08, + "loss": 0.7073, + "step": 7722 + }, + { + "epoch": 0.9286358444056995, + "grad_norm": 1.6406299409337801, + "learning_rate": 5.3108156017673824e-08, + "loss": 0.6058, + "step": 7723 + }, + { + "epoch": 0.9287560872963386, + "grad_norm": 1.8547501306797285, + "learning_rate": 5.2929983765586775e-08, + "loss": 0.7231, + "step": 7724 + }, + { + "epoch": 0.9288763301869777, + "grad_norm": 1.9795838435011859, + "learning_rate": 5.275210688247278e-08, + "loss": 0.6303, + "step": 7725 + }, + { + "epoch": 0.9289965730776167, + "grad_norm": 1.9700797298195074, + "learning_rate": 5.257452539531604e-08, + "loss": 0.8524, + "step": 7726 + }, + { + "epoch": 0.9291168159682559, + "grad_norm": 1.7692984118336814, + "learning_rate": 5.2397239331055445e-08, + "loss": 0.6865, + "step": 7727 + }, + { + "epoch": 0.929237058858895, + "grad_norm": 2.0134406266875726, + "learning_rate": 5.2220248716585036e-08, + "loss": 0.8135, + "step": 7728 + }, + { + "epoch": 0.929357301749534, + "grad_norm": 2.2852998085248575, + "learning_rate": 5.204355357875445e-08, + "loss": 0.7575, + "step": 7729 + }, + { + "epoch": 0.9294775446401732, + "grad_norm": 2.170465794333967, + "learning_rate": 5.1867153944367584e-08, + "loss": 0.706, + "step": 7730 + }, + { + "epoch": 0.9295977875308122, + "grad_norm": 1.6315717591741017, + "learning_rate": 5.16910498401848e-08, + "loss": 0.7397, + "step": 7731 + }, + { + "epoch": 0.9297180304214513, + "grad_norm": 2.8465233997084742, + "learning_rate": 5.151524129292073e-08, + "loss": 0.834, + "step": 7732 + }, + { + "epoch": 0.9298382733120905, + "grad_norm": 2.12868964816976, + "learning_rate": 5.1339728329245155e-08, + "loss": 0.6727, + "step": 7733 + }, + { + "epoch": 0.9299585162027295, + "grad_norm": 2.2703486167308533, + "learning_rate": 5.116451097578367e-08, + "loss": 0.803, + "step": 7734 + }, + { + "epoch": 0.9300787590933686, + "grad_norm": 1.8499725210881182, + "learning_rate": 5.0989589259115895e-08, + "loss": 0.7472, + "step": 7735 + }, + { + "epoch": 0.9301990019840077, + "grad_norm": 2.3158261889987743, + "learning_rate": 5.081496320577816e-08, + "loss": 0.7204, + "step": 7736 + }, + { + "epoch": 0.9303192448746468, + "grad_norm": 0.9391489151024768, + "learning_rate": 5.0640632842260835e-08, + "loss": 0.6123, + "step": 7737 + }, + { + "epoch": 0.9304394877652858, + "grad_norm": 1.5282930121862797, + "learning_rate": 5.0466598195009426e-08, + "loss": 0.7279, + "step": 7738 + }, + { + "epoch": 0.930559730655925, + "grad_norm": 1.9553765098772253, + "learning_rate": 5.0292859290425036e-08, + "loss": 0.7082, + "step": 7739 + }, + { + "epoch": 0.9306799735465641, + "grad_norm": 2.259400228567923, + "learning_rate": 5.011941615486348e-08, + "loss": 0.7808, + "step": 7740 + }, + { + "epoch": 0.9308002164372031, + "grad_norm": 2.141185158161424, + "learning_rate": 4.994626881463659e-08, + "loss": 0.8489, + "step": 7741 + }, + { + "epoch": 0.9309204593278423, + "grad_norm": 1.9436580869280007, + "learning_rate": 4.9773417296009814e-08, + "loss": 0.7114, + "step": 7742 + }, + { + "epoch": 0.9310407022184813, + "grad_norm": 3.747279388265666, + "learning_rate": 4.960086162520527e-08, + "loss": 0.6528, + "step": 7743 + }, + { + "epoch": 0.9311609451091204, + "grad_norm": 2.026578255197286, + "learning_rate": 4.942860182839936e-08, + "loss": 0.8303, + "step": 7744 + }, + { + "epoch": 0.9312811879997596, + "grad_norm": 1.91373886174923, + "learning_rate": 4.925663793172341e-08, + "loss": 0.7992, + "step": 7745 + }, + { + "epoch": 0.9314014308903986, + "grad_norm": 0.7989694335113358, + "learning_rate": 4.908496996126477e-08, + "loss": 0.5653, + "step": 7746 + }, + { + "epoch": 0.9315216737810377, + "grad_norm": 2.463410990329342, + "learning_rate": 4.89135979430646e-08, + "loss": 0.7678, + "step": 7747 + }, + { + "epoch": 0.9316419166716768, + "grad_norm": 2.177311412296023, + "learning_rate": 4.874252190312078e-08, + "loss": 0.8551, + "step": 7748 + }, + { + "epoch": 0.9317621595623159, + "grad_norm": 1.9914726021788203, + "learning_rate": 4.857174186738477e-08, + "loss": 0.6498, + "step": 7749 + }, + { + "epoch": 0.931882402452955, + "grad_norm": 2.4629043711823964, + "learning_rate": 4.840125786176408e-08, + "loss": 0.7362, + "step": 7750 + }, + { + "epoch": 0.932002645343594, + "grad_norm": 2.101721208103153, + "learning_rate": 4.823106991212067e-08, + "loss": 0.7748, + "step": 7751 + }, + { + "epoch": 0.9321228882342332, + "grad_norm": 1.9623200303673227, + "learning_rate": 4.806117804427212e-08, + "loss": 0.8407, + "step": 7752 + }, + { + "epoch": 0.9322431311248722, + "grad_norm": 2.0298659083986563, + "learning_rate": 4.7891582283990926e-08, + "loss": 0.6476, + "step": 7753 + }, + { + "epoch": 0.9323633740155113, + "grad_norm": 1.8104736576687066, + "learning_rate": 4.772228265700473e-08, + "loss": 0.7311, + "step": 7754 + }, + { + "epoch": 0.9324836169061504, + "grad_norm": 2.1268143481614494, + "learning_rate": 4.75532791889961e-08, + "loss": 0.7667, + "step": 7755 + }, + { + "epoch": 0.9326038597967895, + "grad_norm": 1.7979485000261652, + "learning_rate": 4.738457190560252e-08, + "loss": 0.6581, + "step": 7756 + }, + { + "epoch": 0.9327241026874286, + "grad_norm": 2.0190131520810772, + "learning_rate": 4.721616083241664e-08, + "loss": 0.7921, + "step": 7757 + }, + { + "epoch": 0.9328443455780677, + "grad_norm": 1.8254783578267366, + "learning_rate": 4.7048045994986684e-08, + "loss": 0.7809, + "step": 7758 + }, + { + "epoch": 0.9329645884687068, + "grad_norm": 3.6761121611061753, + "learning_rate": 4.688022741881559e-08, + "loss": 0.9122, + "step": 7759 + }, + { + "epoch": 0.9330848313593458, + "grad_norm": 1.612175735871409, + "learning_rate": 4.671270512936076e-08, + "loss": 0.757, + "step": 7760 + }, + { + "epoch": 0.933205074249985, + "grad_norm": 1.8169344882523504, + "learning_rate": 4.6545479152035884e-08, + "loss": 0.829, + "step": 7761 + }, + { + "epoch": 0.9333253171406241, + "grad_norm": 1.9833548018418221, + "learning_rate": 4.637854951220821e-08, + "loss": 0.7653, + "step": 7762 + }, + { + "epoch": 0.9334455600312631, + "grad_norm": 2.3473685242191675, + "learning_rate": 4.621191623520171e-08, + "loss": 0.7504, + "step": 7763 + }, + { + "epoch": 0.9335658029219023, + "grad_norm": 2.4801498087889056, + "learning_rate": 4.604557934629372e-08, + "loss": 0.8512, + "step": 7764 + }, + { + "epoch": 0.9336860458125413, + "grad_norm": 1.8115576563479698, + "learning_rate": 4.587953887071805e-08, + "loss": 0.8066, + "step": 7765 + }, + { + "epoch": 0.9338062887031804, + "grad_norm": 2.8230405172696327, + "learning_rate": 4.5713794833662554e-08, + "loss": 0.8618, + "step": 7766 + }, + { + "epoch": 0.9339265315938196, + "grad_norm": 1.900966513903178, + "learning_rate": 4.5548347260270236e-08, + "loss": 0.6349, + "step": 7767 + }, + { + "epoch": 0.9340467744844586, + "grad_norm": 1.6563080831455648, + "learning_rate": 4.538319617564012e-08, + "loss": 0.6986, + "step": 7768 + }, + { + "epoch": 0.9341670173750977, + "grad_norm": 1.904763063929977, + "learning_rate": 4.521834160482485e-08, + "loss": 0.7494, + "step": 7769 + }, + { + "epoch": 0.9342872602657368, + "grad_norm": 1.577657347024493, + "learning_rate": 4.5053783572832846e-08, + "loss": 0.8256, + "step": 7770 + }, + { + "epoch": 0.9344075031563759, + "grad_norm": 1.7068546299817142, + "learning_rate": 4.488952210462771e-08, + "loss": 0.7614, + "step": 7771 + }, + { + "epoch": 0.9345277460470149, + "grad_norm": 1.9124391483773642, + "learning_rate": 4.4725557225127495e-08, + "loss": 0.8626, + "step": 7772 + }, + { + "epoch": 0.9346479889376541, + "grad_norm": 1.57368107930372, + "learning_rate": 4.456188895920565e-08, + "loss": 0.7961, + "step": 7773 + }, + { + "epoch": 0.9347682318282932, + "grad_norm": 2.077648595827538, + "learning_rate": 4.439851733169031e-08, + "loss": 0.8551, + "step": 7774 + }, + { + "epoch": 0.9348884747189322, + "grad_norm": 2.515621019443182, + "learning_rate": 4.4235442367365204e-08, + "loss": 0.694, + "step": 7775 + }, + { + "epoch": 0.9350087176095714, + "grad_norm": 2.0823404107647026, + "learning_rate": 4.4072664090968545e-08, + "loss": 0.7961, + "step": 7776 + }, + { + "epoch": 0.9351289605002104, + "grad_norm": 1.7926197917498952, + "learning_rate": 4.391018252719347e-08, + "loss": 0.8551, + "step": 7777 + }, + { + "epoch": 0.9352492033908495, + "grad_norm": 1.7480476251525974, + "learning_rate": 4.374799770068849e-08, + "loss": 0.6968, + "step": 7778 + }, + { + "epoch": 0.9353694462814887, + "grad_norm": 1.7692686311313002, + "learning_rate": 4.358610963605658e-08, + "loss": 0.7499, + "step": 7779 + }, + { + "epoch": 0.9354896891721277, + "grad_norm": 2.38292256085469, + "learning_rate": 4.342451835785677e-08, + "loss": 0.6879, + "step": 7780 + }, + { + "epoch": 0.9356099320627668, + "grad_norm": 1.6887352978499097, + "learning_rate": 4.3263223890601665e-08, + "loss": 0.7507, + "step": 7781 + }, + { + "epoch": 0.9357301749534058, + "grad_norm": 3.130660461846376, + "learning_rate": 4.31022262587597e-08, + "loss": 0.7977, + "step": 7782 + }, + { + "epoch": 0.935850417844045, + "grad_norm": 1.6005960665481818, + "learning_rate": 4.2941525486754225e-08, + "loss": 0.6647, + "step": 7783 + }, + { + "epoch": 0.935970660734684, + "grad_norm": 1.934026989422068, + "learning_rate": 4.278112159896286e-08, + "loss": 0.8018, + "step": 7784 + }, + { + "epoch": 0.9360909036253231, + "grad_norm": 1.7757033606656676, + "learning_rate": 4.2621014619719896e-08, + "loss": 0.6772, + "step": 7785 + }, + { + "epoch": 0.9362111465159623, + "grad_norm": 0.7283044258363464, + "learning_rate": 4.246120457331215e-08, + "loss": 0.5868, + "step": 7786 + }, + { + "epoch": 0.9363313894066013, + "grad_norm": 2.081645078076256, + "learning_rate": 4.2301691483983325e-08, + "loss": 0.7258, + "step": 7787 + }, + { + "epoch": 0.9364516322972404, + "grad_norm": 1.9316409952840268, + "learning_rate": 4.214247537593163e-08, + "loss": 0.757, + "step": 7788 + }, + { + "epoch": 0.9365718751878795, + "grad_norm": 1.8683479565545265, + "learning_rate": 4.1983556273309293e-08, + "loss": 0.8123, + "step": 7789 + }, + { + "epoch": 0.9366921180785186, + "grad_norm": 2.1468423366144123, + "learning_rate": 4.182493420022526e-08, + "loss": 0.6919, + "step": 7790 + }, + { + "epoch": 0.9368123609691577, + "grad_norm": 1.797100285214943, + "learning_rate": 4.166660918074139e-08, + "loss": 0.7842, + "step": 7791 + }, + { + "epoch": 0.9369326038597968, + "grad_norm": 2.452884443600727, + "learning_rate": 4.15085812388758e-08, + "loss": 0.7363, + "step": 7792 + }, + { + "epoch": 0.9370528467504359, + "grad_norm": 1.7360607580724225, + "learning_rate": 4.135085039860153e-08, + "loss": 0.7899, + "step": 7793 + }, + { + "epoch": 0.9371730896410749, + "grad_norm": 2.4542704052418305, + "learning_rate": 4.1193416683845906e-08, + "loss": 0.7912, + "step": 7794 + }, + { + "epoch": 0.9372933325317141, + "grad_norm": 2.1773629868297144, + "learning_rate": 4.103628011849136e-08, + "loss": 0.8334, + "step": 7795 + }, + { + "epoch": 0.9374135754223532, + "grad_norm": 2.0196338384016648, + "learning_rate": 4.0879440726375506e-08, + "loss": 0.7617, + "step": 7796 + }, + { + "epoch": 0.9375338183129922, + "grad_norm": 2.460714760040168, + "learning_rate": 4.0722898531291074e-08, + "loss": 0.5647, + "step": 7797 + }, + { + "epoch": 0.9376540612036314, + "grad_norm": 2.2207659307370315, + "learning_rate": 4.0566653556985295e-08, + "loss": 0.7743, + "step": 7798 + }, + { + "epoch": 0.9377743040942704, + "grad_norm": 2.337352413475352, + "learning_rate": 4.0410705827159886e-08, + "loss": 0.8225, + "step": 7799 + }, + { + "epoch": 0.9378945469849095, + "grad_norm": 2.21869369220474, + "learning_rate": 4.0255055365472356e-08, + "loss": 0.7202, + "step": 7800 + }, + { + "epoch": 0.9380147898755486, + "grad_norm": 2.2110775744814464, + "learning_rate": 4.009970219553471e-08, + "loss": 0.7535, + "step": 7801 + }, + { + "epoch": 0.9381350327661877, + "grad_norm": 2.3799593494053166, + "learning_rate": 3.99446463409141e-08, + "loss": 0.7696, + "step": 7802 + }, + { + "epoch": 0.9382552756568268, + "grad_norm": 2.253267353021091, + "learning_rate": 3.978988782513215e-08, + "loss": 0.699, + "step": 7803 + }, + { + "epoch": 0.9383755185474659, + "grad_norm": 1.7438875123633915, + "learning_rate": 3.963542667166586e-08, + "loss": 0.7694, + "step": 7804 + }, + { + "epoch": 0.938495761438105, + "grad_norm": 1.7632280131269564, + "learning_rate": 3.9481262903946486e-08, + "loss": 0.6837, + "step": 7805 + }, + { + "epoch": 0.938616004328744, + "grad_norm": 0.7946701963568177, + "learning_rate": 3.932739654536066e-08, + "loss": 0.5453, + "step": 7806 + }, + { + "epoch": 0.9387362472193832, + "grad_norm": 2.015225393027934, + "learning_rate": 3.917382761925014e-08, + "loss": 0.74, + "step": 7807 + }, + { + "epoch": 0.9388564901100223, + "grad_norm": 1.8096864501444176, + "learning_rate": 3.9020556148910754e-08, + "loss": 0.7955, + "step": 7808 + }, + { + "epoch": 0.9389767330006613, + "grad_norm": 0.7261099687919332, + "learning_rate": 3.8867582157593895e-08, + "loss": 0.5681, + "step": 7809 + }, + { + "epoch": 0.9390969758913005, + "grad_norm": 1.7136493946598275, + "learning_rate": 3.871490566850544e-08, + "loss": 0.7733, + "step": 7810 + }, + { + "epoch": 0.9392172187819395, + "grad_norm": 1.5715715665284553, + "learning_rate": 3.856252670480642e-08, + "loss": 0.7089, + "step": 7811 + }, + { + "epoch": 0.9393374616725786, + "grad_norm": 1.9332664752230075, + "learning_rate": 3.841044528961279e-08, + "loss": 0.8132, + "step": 7812 + }, + { + "epoch": 0.9394577045632178, + "grad_norm": 6.455626217748457, + "learning_rate": 3.825866144599477e-08, + "loss": 0.7915, + "step": 7813 + }, + { + "epoch": 0.9395779474538568, + "grad_norm": 2.9693994715153345, + "learning_rate": 3.8107175196978145e-08, + "loss": 0.7561, + "step": 7814 + }, + { + "epoch": 0.9396981903444959, + "grad_norm": 1.8859005372123776, + "learning_rate": 3.7955986565542996e-08, + "loss": 0.7758, + "step": 7815 + }, + { + "epoch": 0.9398184332351349, + "grad_norm": 1.9734636826035528, + "learning_rate": 3.780509557462497e-08, + "loss": 0.6846, + "step": 7816 + }, + { + "epoch": 0.9399386761257741, + "grad_norm": 1.6249952316214218, + "learning_rate": 3.765450224711375e-08, + "loss": 0.7571, + "step": 7817 + }, + { + "epoch": 0.9400589190164131, + "grad_norm": 1.6896099039882044, + "learning_rate": 3.750420660585396e-08, + "loss": 0.7993, + "step": 7818 + }, + { + "epoch": 0.9401791619070522, + "grad_norm": 3.920502076595901, + "learning_rate": 3.735420867364603e-08, + "loss": 0.7994, + "step": 7819 + }, + { + "epoch": 0.9402994047976914, + "grad_norm": 2.7493388508273586, + "learning_rate": 3.7204508473244186e-08, + "loss": 0.6229, + "step": 7820 + }, + { + "epoch": 0.9404196476883304, + "grad_norm": 2.3363324600125934, + "learning_rate": 3.7055106027357395e-08, + "loss": 0.6978, + "step": 7821 + }, + { + "epoch": 0.9405398905789695, + "grad_norm": 2.4792786153526927, + "learning_rate": 3.690600135865063e-08, + "loss": 0.7248, + "step": 7822 + }, + { + "epoch": 0.9406601334696086, + "grad_norm": 0.7930117816242268, + "learning_rate": 3.675719448974246e-08, + "loss": 0.5809, + "step": 7823 + }, + { + "epoch": 0.9407803763602477, + "grad_norm": 1.8841043685991208, + "learning_rate": 3.6608685443207054e-08, + "loss": 0.6014, + "step": 7824 + }, + { + "epoch": 0.9409006192508867, + "grad_norm": 2.0055381866905266, + "learning_rate": 3.646047424157306e-08, + "loss": 0.6745, + "step": 7825 + }, + { + "epoch": 0.9410208621415259, + "grad_norm": 2.123956985311063, + "learning_rate": 3.631256090732382e-08, + "loss": 0.6919, + "step": 7826 + }, + { + "epoch": 0.941141105032165, + "grad_norm": 1.6658004496567955, + "learning_rate": 3.6164945462897833e-08, + "loss": 0.8293, + "step": 7827 + }, + { + "epoch": 0.941261347922804, + "grad_norm": 1.7574965807690195, + "learning_rate": 3.6017627930687856e-08, + "loss": 0.7617, + "step": 7828 + }, + { + "epoch": 0.9413815908134432, + "grad_norm": 3.038876407682121, + "learning_rate": 3.587060833304267e-08, + "loss": 0.7765, + "step": 7829 + }, + { + "epoch": 0.9415018337040822, + "grad_norm": 2.06083056770424, + "learning_rate": 3.5723886692264225e-08, + "loss": 0.6431, + "step": 7830 + }, + { + "epoch": 0.9416220765947213, + "grad_norm": 2.2486552352403595, + "learning_rate": 3.557746303061071e-08, + "loss": 0.6165, + "step": 7831 + }, + { + "epoch": 0.9417423194853605, + "grad_norm": 1.650053199719605, + "learning_rate": 3.543133737029391e-08, + "loss": 0.7261, + "step": 7832 + }, + { + "epoch": 0.9418625623759995, + "grad_norm": 1.810830630028163, + "learning_rate": 3.5285509733481214e-08, + "loss": 0.6959, + "step": 7833 + }, + { + "epoch": 0.9419828052666386, + "grad_norm": 1.7515347426679693, + "learning_rate": 3.513998014229469e-08, + "loss": 0.7668, + "step": 7834 + }, + { + "epoch": 0.9421030481572777, + "grad_norm": 2.3486045915014473, + "learning_rate": 3.499474861881069e-08, + "loss": 0.869, + "step": 7835 + }, + { + "epoch": 0.9422232910479168, + "grad_norm": 2.8384072554842867, + "learning_rate": 3.4849815185061136e-08, + "loss": 0.6841, + "step": 7836 + }, + { + "epoch": 0.9423435339385559, + "grad_norm": 2.012114550717717, + "learning_rate": 3.470517986303223e-08, + "loss": 0.764, + "step": 7837 + }, + { + "epoch": 0.942463776829195, + "grad_norm": 1.7858914587138628, + "learning_rate": 3.4560842674664856e-08, + "loss": 0.7998, + "step": 7838 + }, + { + "epoch": 0.9425840197198341, + "grad_norm": 2.1515300895116987, + "learning_rate": 3.441680364185506e-08, + "loss": 0.7565, + "step": 7839 + }, + { + "epoch": 0.9427042626104731, + "grad_norm": 2.2729692009560694, + "learning_rate": 3.427306278645314e-08, + "loss": 0.7531, + "step": 7840 + }, + { + "epoch": 0.9428245055011123, + "grad_norm": 1.811260457242339, + "learning_rate": 3.4129620130264767e-08, + "loss": 0.7298, + "step": 7841 + }, + { + "epoch": 0.9429447483917514, + "grad_norm": 2.2797651292056473, + "learning_rate": 3.398647569505009e-08, + "loss": 0.7846, + "step": 7842 + }, + { + "epoch": 0.9430649912823904, + "grad_norm": 2.471434280710217, + "learning_rate": 3.384362950252373e-08, + "loss": 0.7497, + "step": 7843 + }, + { + "epoch": 0.9431852341730296, + "grad_norm": 1.808361458742778, + "learning_rate": 3.3701081574355473e-08, + "loss": 0.5689, + "step": 7844 + }, + { + "epoch": 0.9433054770636686, + "grad_norm": 0.6504890894315699, + "learning_rate": 3.3558831932169796e-08, + "loss": 0.5171, + "step": 7845 + }, + { + "epoch": 0.9434257199543077, + "grad_norm": 1.8757887135094646, + "learning_rate": 3.341688059754588e-08, + "loss": 0.8863, + "step": 7846 + }, + { + "epoch": 0.9435459628449467, + "grad_norm": 2.020613183158966, + "learning_rate": 3.327522759201762e-08, + "loss": 0.7805, + "step": 7847 + }, + { + "epoch": 0.9436662057355859, + "grad_norm": 2.5340240724740424, + "learning_rate": 3.313387293707359e-08, + "loss": 0.6699, + "step": 7848 + }, + { + "epoch": 0.943786448626225, + "grad_norm": 2.2797473504810015, + "learning_rate": 3.29928166541571e-08, + "loss": 0.6936, + "step": 7849 + }, + { + "epoch": 0.943906691516864, + "grad_norm": 2.0861355691089245, + "learning_rate": 3.2852058764666346e-08, + "loss": 0.8071, + "step": 7850 + }, + { + "epoch": 0.9440269344075032, + "grad_norm": 1.706233416466242, + "learning_rate": 3.2711599289954264e-08, + "loss": 0.6887, + "step": 7851 + }, + { + "epoch": 0.9441471772981422, + "grad_norm": 1.8587455004475357, + "learning_rate": 3.257143825132847e-08, + "loss": 0.7797, + "step": 7852 + }, + { + "epoch": 0.9442674201887813, + "grad_norm": 1.8199514268325028, + "learning_rate": 3.243157567005106e-08, + "loss": 0.7646, + "step": 7853 + }, + { + "epoch": 0.9443876630794205, + "grad_norm": 2.0365450581249593, + "learning_rate": 3.2292011567339296e-08, + "loss": 0.645, + "step": 7854 + }, + { + "epoch": 0.9445079059700595, + "grad_norm": 2.17608322705258, + "learning_rate": 3.21527459643649e-08, + "loss": 0.5674, + "step": 7855 + }, + { + "epoch": 0.9446281488606986, + "grad_norm": 5.270346316227117, + "learning_rate": 3.2013778882254536e-08, + "loss": 0.7413, + "step": 7856 + }, + { + "epoch": 0.9447483917513377, + "grad_norm": 1.8959517669015191, + "learning_rate": 3.1875110342088676e-08, + "loss": 0.766, + "step": 7857 + }, + { + "epoch": 0.9448686346419768, + "grad_norm": 1.7739962735228336, + "learning_rate": 3.1736740364904035e-08, + "loss": 0.6574, + "step": 7858 + }, + { + "epoch": 0.9449888775326158, + "grad_norm": 1.9670468789910829, + "learning_rate": 3.159866897169094e-08, + "loss": 0.7712, + "step": 7859 + }, + { + "epoch": 0.945109120423255, + "grad_norm": 2.040098665132783, + "learning_rate": 3.146089618339487e-08, + "loss": 0.76, + "step": 7860 + }, + { + "epoch": 0.9452293633138941, + "grad_norm": 1.7613754445460457, + "learning_rate": 3.132342202091554e-08, + "loss": 0.6827, + "step": 7861 + }, + { + "epoch": 0.9453496062045331, + "grad_norm": 2.128286961065485, + "learning_rate": 3.1186246505107595e-08, + "loss": 0.6865, + "step": 7862 + }, + { + "epoch": 0.9454698490951723, + "grad_norm": 1.8293067177559243, + "learning_rate": 3.104936965678084e-08, + "loss": 0.842, + "step": 7863 + }, + { + "epoch": 0.9455900919858113, + "grad_norm": 2.214686043320203, + "learning_rate": 3.091279149669956e-08, + "loss": 0.8258, + "step": 7864 + }, + { + "epoch": 0.9457103348764504, + "grad_norm": 2.1669569799134014, + "learning_rate": 3.0776512045581624e-08, + "loss": 0.7388, + "step": 7865 + }, + { + "epoch": 0.9458305777670896, + "grad_norm": 1.9384542391858526, + "learning_rate": 3.0640531324101384e-08, + "loss": 0.7795, + "step": 7866 + }, + { + "epoch": 0.9459508206577286, + "grad_norm": 1.7906186195246543, + "learning_rate": 3.0504849352886554e-08, + "loss": 0.7596, + "step": 7867 + }, + { + "epoch": 0.9460710635483677, + "grad_norm": 2.386661010643137, + "learning_rate": 3.036946615252023e-08, + "loss": 0.7217, + "step": 7868 + }, + { + "epoch": 0.9461913064390068, + "grad_norm": 2.204575189856475, + "learning_rate": 3.0234381743539984e-08, + "loss": 0.6716, + "step": 7869 + }, + { + "epoch": 0.9463115493296459, + "grad_norm": 2.2165446537062543, + "learning_rate": 3.0099596146437863e-08, + "loss": 0.8012, + "step": 7870 + }, + { + "epoch": 0.946431792220285, + "grad_norm": 0.7928388893267431, + "learning_rate": 2.996510938166086e-08, + "loss": 0.6007, + "step": 7871 + }, + { + "epoch": 0.9465520351109241, + "grad_norm": 1.8945773089600924, + "learning_rate": 2.983092146960997e-08, + "loss": 0.7366, + "step": 7872 + }, + { + "epoch": 0.9466722780015632, + "grad_norm": 2.27966786742826, + "learning_rate": 2.9697032430642256e-08, + "loss": 0.8049, + "step": 7873 + }, + { + "epoch": 0.9467925208922022, + "grad_norm": 2.544438513101534, + "learning_rate": 2.9563442285067906e-08, + "loss": 0.7421, + "step": 7874 + }, + { + "epoch": 0.9469127637828414, + "grad_norm": 3.488428332740407, + "learning_rate": 2.943015105315294e-08, + "loss": 0.8021, + "step": 7875 + }, + { + "epoch": 0.9470330066734804, + "grad_norm": 2.6133479407771576, + "learning_rate": 2.929715875511718e-08, + "loss": 0.6703, + "step": 7876 + }, + { + "epoch": 0.9471532495641195, + "grad_norm": 2.192679023697804, + "learning_rate": 2.9164465411135375e-08, + "loss": 0.6997, + "step": 7877 + }, + { + "epoch": 0.9472734924547586, + "grad_norm": 1.8317382838531948, + "learning_rate": 2.9032071041337426e-08, + "loss": 0.8079, + "step": 7878 + }, + { + "epoch": 0.9473937353453977, + "grad_norm": 1.9336309872519213, + "learning_rate": 2.889997566580704e-08, + "loss": 0.7297, + "step": 7879 + }, + { + "epoch": 0.9475139782360368, + "grad_norm": 1.878079619152949, + "learning_rate": 2.8768179304583086e-08, + "loss": 0.7078, + "step": 7880 + }, + { + "epoch": 0.9476342211266758, + "grad_norm": 1.7046525037269398, + "learning_rate": 2.8636681977659117e-08, + "loss": 0.74, + "step": 7881 + }, + { + "epoch": 0.947754464017315, + "grad_norm": 2.2219622287715977, + "learning_rate": 2.850548370498318e-08, + "loss": 0.789, + "step": 7882 + }, + { + "epoch": 0.9478747069079541, + "grad_norm": 1.4993595504441817, + "learning_rate": 2.8374584506457798e-08, + "loss": 0.7138, + "step": 7883 + }, + { + "epoch": 0.9479949497985931, + "grad_norm": 2.3710456855841984, + "learning_rate": 2.824398440193998e-08, + "loss": 0.677, + "step": 7884 + }, + { + "epoch": 0.9481151926892323, + "grad_norm": 43.23833894911002, + "learning_rate": 2.811368341124232e-08, + "loss": 0.7182, + "step": 7885 + }, + { + "epoch": 0.9482354355798713, + "grad_norm": 2.318967487181311, + "learning_rate": 2.7983681554131222e-08, + "loss": 0.6868, + "step": 7886 + }, + { + "epoch": 0.9483556784705104, + "grad_norm": 2.0040883953753816, + "learning_rate": 2.7853978850327365e-08, + "loss": 0.7033, + "step": 7887 + }, + { + "epoch": 0.9484759213611496, + "grad_norm": 1.9603131276705403, + "learning_rate": 2.7724575319507225e-08, + "loss": 0.8759, + "step": 7888 + }, + { + "epoch": 0.9485961642517886, + "grad_norm": 1.9996569458477798, + "learning_rate": 2.759547098130044e-08, + "loss": 0.7759, + "step": 7889 + }, + { + "epoch": 0.9487164071424277, + "grad_norm": 1.8619770467963968, + "learning_rate": 2.746666585529267e-08, + "loss": 0.7704, + "step": 7890 + }, + { + "epoch": 0.9488366500330668, + "grad_norm": 2.1913610987508214, + "learning_rate": 2.73381599610234e-08, + "loss": 0.7434, + "step": 7891 + }, + { + "epoch": 0.9489568929237059, + "grad_norm": 1.8535230855520295, + "learning_rate": 2.7209953317987033e-08, + "loss": 0.7178, + "step": 7892 + }, + { + "epoch": 0.9490771358143449, + "grad_norm": 5.032646980394359, + "learning_rate": 2.7082045945631793e-08, + "loss": 0.7844, + "step": 7893 + }, + { + "epoch": 0.9491973787049841, + "grad_norm": 1.9512873435005715, + "learning_rate": 2.6954437863361712e-08, + "loss": 0.6955, + "step": 7894 + }, + { + "epoch": 0.9493176215956232, + "grad_norm": 2.1339314079379825, + "learning_rate": 2.6827129090534862e-08, + "loss": 0.7112, + "step": 7895 + }, + { + "epoch": 0.9494378644862622, + "grad_norm": 1.8881494377446324, + "learning_rate": 2.670011964646335e-08, + "loss": 0.7748, + "step": 7896 + }, + { + "epoch": 0.9495581073769014, + "grad_norm": 2.203384708761656, + "learning_rate": 2.657340955041487e-08, + "loss": 0.6838, + "step": 7897 + }, + { + "epoch": 0.9496783502675404, + "grad_norm": 1.9178918018948918, + "learning_rate": 2.6446998821611167e-08, + "loss": 0.7158, + "step": 7898 + }, + { + "epoch": 0.9497985931581795, + "grad_norm": 3.276254247773992, + "learning_rate": 2.6320887479228228e-08, + "loss": 0.7181, + "step": 7899 + }, + { + "epoch": 0.9499188360488187, + "grad_norm": 2.470333798513942, + "learning_rate": 2.619507554239786e-08, + "loss": 0.7335, + "step": 7900 + }, + { + "epoch": 0.9500390789394577, + "grad_norm": 3.1491869467527875, + "learning_rate": 2.606956303020502e-08, + "loss": 0.6984, + "step": 7901 + }, + { + "epoch": 0.9501593218300968, + "grad_norm": 1.7589490818767006, + "learning_rate": 2.5944349961690036e-08, + "loss": 0.8436, + "step": 7902 + }, + { + "epoch": 0.9502795647207359, + "grad_norm": 1.7073173370510655, + "learning_rate": 2.581943635584749e-08, + "loss": 0.7364, + "step": 7903 + }, + { + "epoch": 0.950399807611375, + "grad_norm": 1.6066969283144086, + "learning_rate": 2.569482223162689e-08, + "loss": 0.6584, + "step": 7904 + }, + { + "epoch": 0.950520050502014, + "grad_norm": 1.719357400552413, + "learning_rate": 2.5570507607932e-08, + "loss": 0.7285, + "step": 7905 + }, + { + "epoch": 0.9506402933926532, + "grad_norm": 2.821881563130432, + "learning_rate": 2.54464925036213e-08, + "loss": 0.6443, + "step": 7906 + }, + { + "epoch": 0.9507605362832923, + "grad_norm": 1.9307199411212106, + "learning_rate": 2.532277693750773e-08, + "loss": 0.6099, + "step": 7907 + }, + { + "epoch": 0.9508807791739313, + "grad_norm": 2.093420315877106, + "learning_rate": 2.5199360928358948e-08, + "loss": 0.7622, + "step": 7908 + }, + { + "epoch": 0.9510010220645704, + "grad_norm": 1.7118714452945478, + "learning_rate": 2.507624449489665e-08, + "loss": 0.8711, + "step": 7909 + }, + { + "epoch": 0.9511212649552095, + "grad_norm": 2.2257551904708017, + "learning_rate": 2.495342765579811e-08, + "loss": 0.6535, + "step": 7910 + }, + { + "epoch": 0.9512415078458486, + "grad_norm": 1.747832318143652, + "learning_rate": 2.4830910429693984e-08, + "loss": 0.716, + "step": 7911 + }, + { + "epoch": 0.9513617507364877, + "grad_norm": 1.8531738223355614, + "learning_rate": 2.470869283517052e-08, + "loss": 0.7972, + "step": 7912 + }, + { + "epoch": 0.9514819936271268, + "grad_norm": 1.7256504601600624, + "learning_rate": 2.458677489076777e-08, + "loss": 0.7755, + "step": 7913 + }, + { + "epoch": 0.9516022365177659, + "grad_norm": 2.7538097175290934, + "learning_rate": 2.446515661498072e-08, + "loss": 0.8339, + "step": 7914 + }, + { + "epoch": 0.9517224794084049, + "grad_norm": 1.992542788669532, + "learning_rate": 2.434383802625861e-08, + "loss": 0.7504, + "step": 7915 + }, + { + "epoch": 0.9518427222990441, + "grad_norm": 1.8512227837251005, + "learning_rate": 2.4222819143005168e-08, + "loss": 0.7406, + "step": 7916 + }, + { + "epoch": 0.9519629651896832, + "grad_norm": 1.8154194438085123, + "learning_rate": 2.4102099983579706e-08, + "loss": 0.8129, + "step": 7917 + }, + { + "epoch": 0.9520832080803222, + "grad_norm": 1.8762877810319385, + "learning_rate": 2.3981680566294236e-08, + "loss": 0.779, + "step": 7918 + }, + { + "epoch": 0.9522034509709614, + "grad_norm": 2.4767624446409546, + "learning_rate": 2.3861560909416822e-08, + "loss": 0.7322, + "step": 7919 + }, + { + "epoch": 0.9523236938616004, + "grad_norm": 2.035905637806489, + "learning_rate": 2.3741741031169325e-08, + "loss": 0.8256, + "step": 7920 + }, + { + "epoch": 0.9524439367522395, + "grad_norm": 1.9258417889904467, + "learning_rate": 2.3622220949728544e-08, + "loss": 0.7133, + "step": 7921 + }, + { + "epoch": 0.9525641796428787, + "grad_norm": 2.4961832475213632, + "learning_rate": 2.3503000683225526e-08, + "loss": 0.6131, + "step": 7922 + }, + { + "epoch": 0.9526844225335177, + "grad_norm": 2.0355758380897395, + "learning_rate": 2.3384080249745585e-08, + "loss": 0.8496, + "step": 7923 + }, + { + "epoch": 0.9528046654241568, + "grad_norm": 2.191275036805346, + "learning_rate": 2.3265459667329178e-08, + "loss": 0.8355, + "step": 7924 + }, + { + "epoch": 0.9529249083147959, + "grad_norm": 2.2391866879786075, + "learning_rate": 2.31471389539708e-08, + "loss": 0.8663, + "step": 7925 + }, + { + "epoch": 0.953045151205435, + "grad_norm": 2.2563769782942455, + "learning_rate": 2.3029118127619872e-08, + "loss": 0.7314, + "step": 7926 + }, + { + "epoch": 0.953165394096074, + "grad_norm": 2.192980630840412, + "learning_rate": 2.2911397206179628e-08, + "loss": 0.8713, + "step": 7927 + }, + { + "epoch": 0.9532856369867132, + "grad_norm": 4.662264836903179, + "learning_rate": 2.279397620750845e-08, + "loss": 0.6261, + "step": 7928 + }, + { + "epoch": 0.9534058798773523, + "grad_norm": 2.000495849177929, + "learning_rate": 2.2676855149419195e-08, + "loss": 0.791, + "step": 7929 + }, + { + "epoch": 0.9535261227679913, + "grad_norm": 2.8686499750655114, + "learning_rate": 2.2560034049678988e-08, + "loss": 0.7623, + "step": 7930 + }, + { + "epoch": 0.9536463656586305, + "grad_norm": 1.7492657892360444, + "learning_rate": 2.2443512926008988e-08, + "loss": 0.758, + "step": 7931 + }, + { + "epoch": 0.9537666085492695, + "grad_norm": 2.735353654386026, + "learning_rate": 2.2327291796085946e-08, + "loss": 0.6975, + "step": 7932 + }, + { + "epoch": 0.9538868514399086, + "grad_norm": 2.328920164581776, + "learning_rate": 2.2211370677540197e-08, + "loss": 0.7791, + "step": 7933 + }, + { + "epoch": 0.9540070943305478, + "grad_norm": 2.370476379965608, + "learning_rate": 2.2095749587957012e-08, + "loss": 0.7866, + "step": 7934 + }, + { + "epoch": 0.9541273372211868, + "grad_norm": 7.571023724572332, + "learning_rate": 2.1980428544876138e-08, + "loss": 0.6964, + "step": 7935 + }, + { + "epoch": 0.9542475801118259, + "grad_norm": 1.9610495364338796, + "learning_rate": 2.1865407565791584e-08, + "loss": 0.7445, + "step": 7936 + }, + { + "epoch": 0.954367823002465, + "grad_norm": 2.0320667678797104, + "learning_rate": 2.175068666815183e-08, + "loss": 0.7796, + "step": 7937 + }, + { + "epoch": 0.9544880658931041, + "grad_norm": 2.2857702750773634, + "learning_rate": 2.163626586935985e-08, + "loss": 0.7933, + "step": 7938 + }, + { + "epoch": 0.9546083087837431, + "grad_norm": 2.122310170198176, + "learning_rate": 2.1522145186773755e-08, + "loss": 0.6362, + "step": 7939 + }, + { + "epoch": 0.9547285516743822, + "grad_norm": 2.1957559918192984, + "learning_rate": 2.140832463770481e-08, + "loss": 0.8565, + "step": 7940 + }, + { + "epoch": 0.9548487945650214, + "grad_norm": 2.2732000587453975, + "learning_rate": 2.129480423941987e-08, + "loss": 0.7589, + "step": 7941 + }, + { + "epoch": 0.9549690374556604, + "grad_norm": 1.6268244551648354, + "learning_rate": 2.1181584009140052e-08, + "loss": 0.8051, + "step": 7942 + }, + { + "epoch": 0.9550892803462995, + "grad_norm": 2.5505913142340333, + "learning_rate": 2.10686639640405e-08, + "loss": 0.8427, + "step": 7943 + }, + { + "epoch": 0.9552095232369386, + "grad_norm": 1.7408600270130088, + "learning_rate": 2.0956044121251294e-08, + "loss": 0.814, + "step": 7944 + }, + { + "epoch": 0.9553297661275777, + "grad_norm": 1.8321109583581345, + "learning_rate": 2.084372449785654e-08, + "loss": 0.8114, + "step": 7945 + }, + { + "epoch": 0.9554500090182168, + "grad_norm": 2.0366128405977886, + "learning_rate": 2.0731705110895282e-08, + "loss": 0.6928, + "step": 7946 + }, + { + "epoch": 0.9555702519088559, + "grad_norm": 1.753666578879563, + "learning_rate": 2.0619985977360587e-08, + "loss": 0.8682, + "step": 7947 + }, + { + "epoch": 0.955690494799495, + "grad_norm": 1.7330458763376597, + "learning_rate": 2.0508567114200237e-08, + "loss": 0.7682, + "step": 7948 + }, + { + "epoch": 0.955810737690134, + "grad_norm": 1.9181714232474147, + "learning_rate": 2.0397448538316485e-08, + "loss": 0.7886, + "step": 7949 + }, + { + "epoch": 0.9559309805807732, + "grad_norm": 2.111503262097501, + "learning_rate": 2.028663026656563e-08, + "loss": 0.6689, + "step": 7950 + }, + { + "epoch": 0.9560512234714122, + "grad_norm": 1.9943756652467504, + "learning_rate": 2.0176112315758885e-08, + "loss": 0.7186, + "step": 7951 + }, + { + "epoch": 0.9561714663620513, + "grad_norm": 3.1918635700201197, + "learning_rate": 2.0065894702661957e-08, + "loss": 0.696, + "step": 7952 + }, + { + "epoch": 0.9562917092526905, + "grad_norm": 1.6601423195647163, + "learning_rate": 1.9955977443994577e-08, + "loss": 0.7817, + "step": 7953 + }, + { + "epoch": 0.9564119521433295, + "grad_norm": 2.2614123939078037, + "learning_rate": 1.9846360556430965e-08, + "loss": 0.62, + "step": 7954 + }, + { + "epoch": 0.9565321950339686, + "grad_norm": 2.2574716066452685, + "learning_rate": 1.973704405660004e-08, + "loss": 0.6187, + "step": 7955 + }, + { + "epoch": 0.9566524379246077, + "grad_norm": 1.5082207952144506, + "learning_rate": 1.9628027961085203e-08, + "loss": 0.783, + "step": 7956 + }, + { + "epoch": 0.9567726808152468, + "grad_norm": 1.7845395815652894, + "learning_rate": 1.9519312286423894e-08, + "loss": 0.8395, + "step": 7957 + }, + { + "epoch": 0.9568929237058859, + "grad_norm": 1.818123479470738, + "learning_rate": 1.9410897049108255e-08, + "loss": 0.7791, + "step": 7958 + }, + { + "epoch": 0.957013166596525, + "grad_norm": 2.7130209532398575, + "learning_rate": 1.9302782265584905e-08, + "loss": 0.915, + "step": 7959 + }, + { + "epoch": 0.9571334094871641, + "grad_norm": 2.3202518750629015, + "learning_rate": 1.9194967952254282e-08, + "loss": 0.871, + "step": 7960 + }, + { + "epoch": 0.9572536523778031, + "grad_norm": 2.525250613173329, + "learning_rate": 1.9087454125472635e-08, + "loss": 0.8151, + "step": 7961 + }, + { + "epoch": 0.9573738952684423, + "grad_norm": 2.1489317169488857, + "learning_rate": 1.8980240801548696e-08, + "loss": 0.79, + "step": 7962 + }, + { + "epoch": 0.9574941381590814, + "grad_norm": 1.7076207589918544, + "learning_rate": 1.8873327996747458e-08, + "loss": 0.7407, + "step": 7963 + }, + { + "epoch": 0.9576143810497204, + "grad_norm": 1.9096605402685678, + "learning_rate": 1.8766715727287053e-08, + "loss": 0.6599, + "step": 7964 + }, + { + "epoch": 0.9577346239403596, + "grad_norm": 1.6656345589025296, + "learning_rate": 1.8660404009340546e-08, + "loss": 0.7938, + "step": 7965 + }, + { + "epoch": 0.9578548668309986, + "grad_norm": 0.8822916232803015, + "learning_rate": 1.8554392859035485e-08, + "loss": 0.5958, + "step": 7966 + }, + { + "epoch": 0.9579751097216377, + "grad_norm": 1.7705126602939698, + "learning_rate": 1.8448682292453444e-08, + "loss": 0.7913, + "step": 7967 + }, + { + "epoch": 0.9580953526122769, + "grad_norm": 1.8716512657793276, + "learning_rate": 1.8343272325631154e-08, + "loss": 0.6604, + "step": 7968 + }, + { + "epoch": 0.9582155955029159, + "grad_norm": 2.432817803819935, + "learning_rate": 1.8238162974558492e-08, + "loss": 0.7848, + "step": 7969 + }, + { + "epoch": 0.958335838393555, + "grad_norm": 1.8298191143752456, + "learning_rate": 1.8133354255181144e-08, + "loss": 0.7477, + "step": 7970 + }, + { + "epoch": 0.958456081284194, + "grad_norm": 1.891875806946294, + "learning_rate": 1.802884618339795e-08, + "loss": 0.7519, + "step": 7971 + }, + { + "epoch": 0.9585763241748332, + "grad_norm": 3.3153960325979623, + "learning_rate": 1.7924638775062894e-08, + "loss": 0.8072, + "step": 7972 + }, + { + "epoch": 0.9586965670654722, + "grad_norm": 2.31267526323281, + "learning_rate": 1.7820732045984444e-08, + "loss": 0.815, + "step": 7973 + }, + { + "epoch": 0.9588168099561113, + "grad_norm": 1.8222300371658924, + "learning_rate": 1.7717126011924655e-08, + "loss": 0.7428, + "step": 7974 + }, + { + "epoch": 0.9589370528467505, + "grad_norm": 3.218652520046208, + "learning_rate": 1.7613820688600957e-08, + "loss": 0.769, + "step": 7975 + }, + { + "epoch": 0.9590572957373895, + "grad_norm": 1.807653326809861, + "learning_rate": 1.7510816091684588e-08, + "loss": 0.7916, + "step": 7976 + }, + { + "epoch": 0.9591775386280286, + "grad_norm": 3.515214277223356, + "learning_rate": 1.740811223680083e-08, + "loss": 0.7923, + "step": 7977 + }, + { + "epoch": 0.9592977815186677, + "grad_norm": 2.16483070152092, + "learning_rate": 1.7305709139530334e-08, + "loss": 0.7469, + "step": 7978 + }, + { + "epoch": 0.9594180244093068, + "grad_norm": 2.26777231467573, + "learning_rate": 1.7203606815407334e-08, + "loss": 0.753, + "step": 7979 + }, + { + "epoch": 0.9595382672999458, + "grad_norm": 1.8075021040443848, + "learning_rate": 1.7101805279920557e-08, + "loss": 0.7974, + "step": 7980 + }, + { + "epoch": 0.959658510190585, + "grad_norm": 1.9528387241369969, + "learning_rate": 1.7000304548513643e-08, + "loss": 0.8134, + "step": 7981 + }, + { + "epoch": 0.9597787530812241, + "grad_norm": 2.4120648955286628, + "learning_rate": 1.6899104636583394e-08, + "loss": 0.8339, + "step": 7982 + }, + { + "epoch": 0.9598989959718631, + "grad_norm": 0.7473735518019586, + "learning_rate": 1.6798205559482638e-08, + "loss": 0.6201, + "step": 7983 + }, + { + "epoch": 0.9600192388625023, + "grad_norm": 1.7328273982378453, + "learning_rate": 1.669760733251713e-08, + "loss": 0.7675, + "step": 7984 + }, + { + "epoch": 0.9601394817531413, + "grad_norm": 1.7909816127197453, + "learning_rate": 1.659730997094755e-08, + "loss": 0.825, + "step": 7985 + }, + { + "epoch": 0.9602597246437804, + "grad_norm": 1.760418329482638, + "learning_rate": 1.6497313489989283e-08, + "loss": 0.6213, + "step": 7986 + }, + { + "epoch": 0.9603799675344196, + "grad_norm": 2.1681068720077583, + "learning_rate": 1.639761790481131e-08, + "loss": 0.7056, + "step": 7987 + }, + { + "epoch": 0.9605002104250586, + "grad_norm": 10.106914427177133, + "learning_rate": 1.6298223230537754e-08, + "loss": 0.7941, + "step": 7988 + }, + { + "epoch": 0.9606204533156977, + "grad_norm": 2.794576798201936, + "learning_rate": 1.619912948224611e-08, + "loss": 0.7008, + "step": 7989 + }, + { + "epoch": 0.9607406962063368, + "grad_norm": 3.1024430031655092, + "learning_rate": 1.6100336674969682e-08, + "loss": 0.6168, + "step": 7990 + }, + { + "epoch": 0.9608609390969759, + "grad_norm": 1.8978161206988993, + "learning_rate": 1.600184482369449e-08, + "loss": 0.7776, + "step": 7991 + }, + { + "epoch": 0.960981181987615, + "grad_norm": 2.575467861596359, + "learning_rate": 1.5903653943362126e-08, + "loss": 0.8957, + "step": 7992 + }, + { + "epoch": 0.9611014248782541, + "grad_norm": 2.0192443525814623, + "learning_rate": 1.580576404886802e-08, + "loss": 0.7728, + "step": 7993 + }, + { + "epoch": 0.9612216677688932, + "grad_norm": 2.0189167910429773, + "learning_rate": 1.570817515506162e-08, + "loss": 0.8037, + "step": 7994 + }, + { + "epoch": 0.9613419106595322, + "grad_norm": 2.092714765458755, + "learning_rate": 1.561088727674753e-08, + "loss": 0.8137, + "step": 7995 + }, + { + "epoch": 0.9614621535501714, + "grad_norm": 2.38931427970524, + "learning_rate": 1.551390042868417e-08, + "loss": 0.7178, + "step": 7996 + }, + { + "epoch": 0.9615823964408104, + "grad_norm": 1.9532485312496977, + "learning_rate": 1.5417214625584207e-08, + "loss": 0.7103, + "step": 7997 + }, + { + "epoch": 0.9617026393314495, + "grad_norm": 1.956497393677767, + "learning_rate": 1.5320829882114806e-08, + "loss": 0.8567, + "step": 7998 + }, + { + "epoch": 0.9618228822220887, + "grad_norm": 2.5888321916250274, + "learning_rate": 1.5224746212897378e-08, + "loss": 0.7915, + "step": 7999 + }, + { + "epoch": 0.9619431251127277, + "grad_norm": 2.0386750887282465, + "learning_rate": 1.512896363250804e-08, + "loss": 0.7772, + "step": 8000 + }, + { + "epoch": 0.9620633680033668, + "grad_norm": 1.7768720399105333, + "learning_rate": 1.503348215547673e-08, + "loss": 0.7591, + "step": 8001 + }, + { + "epoch": 0.962183610894006, + "grad_norm": 1.974749370558914, + "learning_rate": 1.4938301796288078e-08, + "loss": 0.8111, + "step": 8002 + }, + { + "epoch": 0.962303853784645, + "grad_norm": 2.6543384630146574, + "learning_rate": 1.4843422569380537e-08, + "loss": 0.8197, + "step": 8003 + }, + { + "epoch": 0.9624240966752841, + "grad_norm": 1.9196885239755048, + "learning_rate": 1.4748844489147483e-08, + "loss": 0.8279, + "step": 8004 + }, + { + "epoch": 0.9625443395659231, + "grad_norm": 2.113688262272457, + "learning_rate": 1.4654567569936326e-08, + "loss": 0.7145, + "step": 8005 + }, + { + "epoch": 0.9626645824565623, + "grad_norm": 2.68275040836276, + "learning_rate": 1.456059182604874e-08, + "loss": 0.8303, + "step": 8006 + }, + { + "epoch": 0.9627848253472013, + "grad_norm": 1.918831873833735, + "learning_rate": 1.4466917271740653e-08, + "loss": 0.7739, + "step": 8007 + }, + { + "epoch": 0.9629050682378404, + "grad_norm": 1.9748582088901845, + "learning_rate": 1.4373543921222697e-08, + "loss": 0.6788, + "step": 8008 + }, + { + "epoch": 0.9630253111284796, + "grad_norm": 1.7188921782901796, + "learning_rate": 1.428047178865932e-08, + "loss": 0.7774, + "step": 8009 + }, + { + "epoch": 0.9631455540191186, + "grad_norm": 1.628312329664746, + "learning_rate": 1.4187700888169451e-08, + "loss": 0.7456, + "step": 8010 + }, + { + "epoch": 0.9632657969097577, + "grad_norm": 0.7809873902503234, + "learning_rate": 1.40952312338265e-08, + "loss": 0.5698, + "step": 8011 + }, + { + "epoch": 0.9633860398003968, + "grad_norm": 1.9682281650265758, + "learning_rate": 1.4003062839657909e-08, + "loss": 0.6838, + "step": 8012 + }, + { + "epoch": 0.9635062826910359, + "grad_norm": 1.7129139392966857, + "learning_rate": 1.391119571964583e-08, + "loss": 0.8025, + "step": 8013 + }, + { + "epoch": 0.9636265255816749, + "grad_norm": 1.928538547972276, + "learning_rate": 1.3819629887726225e-08, + "loss": 0.7383, + "step": 8014 + }, + { + "epoch": 0.9637467684723141, + "grad_norm": 1.7054189929003414, + "learning_rate": 1.3728365357789317e-08, + "loss": 0.7615, + "step": 8015 + }, + { + "epoch": 0.9638670113629532, + "grad_norm": 2.393318603526969, + "learning_rate": 1.3637402143680254e-08, + "loss": 0.768, + "step": 8016 + }, + { + "epoch": 0.9639872542535922, + "grad_norm": 0.7460732582034373, + "learning_rate": 1.3546740259197998e-08, + "loss": 0.5507, + "step": 8017 + }, + { + "epoch": 0.9641074971442314, + "grad_norm": 2.4374225310706192, + "learning_rate": 1.3456379718095989e-08, + "loss": 0.7114, + "step": 8018 + }, + { + "epoch": 0.9642277400348704, + "grad_norm": 0.9717152651307054, + "learning_rate": 1.3366320534081487e-08, + "loss": 0.6206, + "step": 8019 + }, + { + "epoch": 0.9643479829255095, + "grad_norm": 5.8844787750574445, + "learning_rate": 1.3276562720816675e-08, + "loss": 0.7651, + "step": 8020 + }, + { + "epoch": 0.9644682258161487, + "grad_norm": 2.1330565148801397, + "learning_rate": 1.3187106291917549e-08, + "loss": 0.827, + "step": 8021 + }, + { + "epoch": 0.9645884687067877, + "grad_norm": 1.7733173749798414, + "learning_rate": 1.309795126095503e-08, + "loss": 0.7132, + "step": 8022 + }, + { + "epoch": 0.9647087115974268, + "grad_norm": 2.514028955097107, + "learning_rate": 1.3009097641453192e-08, + "loss": 0.8031, + "step": 8023 + }, + { + "epoch": 0.9648289544880659, + "grad_norm": 2.0978918529307027, + "learning_rate": 1.2920545446891474e-08, + "loss": 0.7569, + "step": 8024 + }, + { + "epoch": 0.964949197378705, + "grad_norm": 1.9613464658510604, + "learning_rate": 1.2832294690703127e-08, + "loss": 0.7071, + "step": 8025 + }, + { + "epoch": 0.965069440269344, + "grad_norm": 1.9056999163291166, + "learning_rate": 1.2744345386275668e-08, + "loss": 0.7783, + "step": 8026 + }, + { + "epoch": 0.9651896831599832, + "grad_norm": 1.8884010160167493, + "learning_rate": 1.265669754695109e-08, + "loss": 0.7911, + "step": 8027 + }, + { + "epoch": 0.9653099260506223, + "grad_norm": 1.9787819806404516, + "learning_rate": 1.2569351186025201e-08, + "loss": 0.8237, + "step": 8028 + }, + { + "epoch": 0.9654301689412613, + "grad_norm": 1.5469184638188314, + "learning_rate": 1.2482306316748737e-08, + "loss": 0.7547, + "step": 8029 + }, + { + "epoch": 0.9655504118319005, + "grad_norm": 3.7107272761177974, + "learning_rate": 1.2395562952326021e-08, + "loss": 0.7878, + "step": 8030 + }, + { + "epoch": 0.9656706547225395, + "grad_norm": 2.353401858797277, + "learning_rate": 1.2309121105916309e-08, + "loss": 0.8154, + "step": 8031 + }, + { + "epoch": 0.9657908976131786, + "grad_norm": 2.0934975315295365, + "learning_rate": 1.222298079063222e-08, + "loss": 0.6958, + "step": 8032 + }, + { + "epoch": 0.9659111405038178, + "grad_norm": 2.014889368470461, + "learning_rate": 1.2137142019541524e-08, + "loss": 0.729, + "step": 8033 + }, + { + "epoch": 0.9660313833944568, + "grad_norm": 2.564972498785347, + "learning_rate": 1.2051604805666027e-08, + "loss": 0.7332, + "step": 8034 + }, + { + "epoch": 0.9661516262850959, + "grad_norm": 2.269586611678201, + "learning_rate": 1.196636916198135e-08, + "loss": 0.7859, + "step": 8035 + }, + { + "epoch": 0.9662718691757349, + "grad_norm": 2.0501727845483377, + "learning_rate": 1.1881435101418036e-08, + "loss": 0.7699, + "step": 8036 + }, + { + "epoch": 0.9663921120663741, + "grad_norm": 0.7444070254081978, + "learning_rate": 1.1796802636860003e-08, + "loss": 0.6556, + "step": 8037 + }, + { + "epoch": 0.9665123549570132, + "grad_norm": 2.0283193721410915, + "learning_rate": 1.1712471781146316e-08, + "loss": 0.7391, + "step": 8038 + }, + { + "epoch": 0.9666325978476522, + "grad_norm": 1.8756110149448164, + "learning_rate": 1.1628442547069628e-08, + "loss": 0.6738, + "step": 8039 + }, + { + "epoch": 0.9667528407382914, + "grad_norm": 2.2165374469618553, + "learning_rate": 1.1544714947377521e-08, + "loss": 0.7786, + "step": 8040 + }, + { + "epoch": 0.9668730836289304, + "grad_norm": 2.1059807489281464, + "learning_rate": 1.1461288994770945e-08, + "loss": 0.7022, + "step": 8041 + }, + { + "epoch": 0.9669933265195695, + "grad_norm": 1.847726510064924, + "learning_rate": 1.1378164701906002e-08, + "loss": 0.7719, + "step": 8042 + }, + { + "epoch": 0.9671135694102087, + "grad_norm": 1.8580894834817532, + "learning_rate": 1.1295342081392156e-08, + "loss": 0.6673, + "step": 8043 + }, + { + "epoch": 0.9672338123008477, + "grad_norm": 1.6030003489912181, + "learning_rate": 1.1212821145793804e-08, + "loss": 0.6974, + "step": 8044 + }, + { + "epoch": 0.9673540551914868, + "grad_norm": 2.0072479047955567, + "learning_rate": 1.1130601907629156e-08, + "loss": 0.7891, + "step": 8045 + }, + { + "epoch": 0.9674742980821259, + "grad_norm": 0.8334835433685495, + "learning_rate": 1.1048684379370899e-08, + "loss": 0.6497, + "step": 8046 + }, + { + "epoch": 0.967594540972765, + "grad_norm": 2.1673655727636203, + "learning_rate": 1.0967068573445759e-08, + "loss": 0.7471, + "step": 8047 + }, + { + "epoch": 0.967714783863404, + "grad_norm": 2.0792025877067193, + "learning_rate": 1.0885754502234945e-08, + "loss": 0.6527, + "step": 8048 + }, + { + "epoch": 0.9678350267540432, + "grad_norm": 2.3320086556367237, + "learning_rate": 1.08047421780737e-08, + "loss": 0.7849, + "step": 8049 + }, + { + "epoch": 0.9679552696446823, + "grad_norm": 2.180744639444197, + "learning_rate": 1.0724031613251305e-08, + "loss": 0.7432, + "step": 8050 + }, + { + "epoch": 0.9680755125353213, + "grad_norm": 2.0972778433637265, + "learning_rate": 1.0643622820011744e-08, + "loss": 0.6694, + "step": 8051 + }, + { + "epoch": 0.9681957554259605, + "grad_norm": 2.792609003547023, + "learning_rate": 1.0563515810552814e-08, + "loss": 0.6866, + "step": 8052 + }, + { + "epoch": 0.9683159983165995, + "grad_norm": 1.7170437059255281, + "learning_rate": 1.0483710597026795e-08, + "loss": 0.7387, + "step": 8053 + }, + { + "epoch": 0.9684362412072386, + "grad_norm": 2.0374913502140974, + "learning_rate": 1.0404207191540227e-08, + "loss": 0.7414, + "step": 8054 + }, + { + "epoch": 0.9685564840978778, + "grad_norm": 2.117030401841352, + "learning_rate": 1.0325005606153236e-08, + "loss": 0.753, + "step": 8055 + }, + { + "epoch": 0.9686767269885168, + "grad_norm": 2.6893837115736, + "learning_rate": 1.0246105852881104e-08, + "loss": 0.7961, + "step": 8056 + }, + { + "epoch": 0.9687969698791559, + "grad_norm": 1.9757128913731832, + "learning_rate": 1.0167507943692476e-08, + "loss": 0.7869, + "step": 8057 + }, + { + "epoch": 0.968917212769795, + "grad_norm": 2.0377498216119996, + "learning_rate": 1.008921189051093e-08, + "loss": 0.7166, + "step": 8058 + }, + { + "epoch": 0.9690374556604341, + "grad_norm": 1.909886003657431, + "learning_rate": 1.0011217705213848e-08, + "loss": 0.7756, + "step": 8059 + }, + { + "epoch": 0.9691576985510731, + "grad_norm": 1.7169626826395092, + "learning_rate": 9.933525399632658e-09, + "loss": 0.748, + "step": 8060 + }, + { + "epoch": 0.9692779414417123, + "grad_norm": 1.7857783796524502, + "learning_rate": 9.856134985553488e-09, + "loss": 0.6581, + "step": 8061 + }, + { + "epoch": 0.9693981843323514, + "grad_norm": 2.3848103137128867, + "learning_rate": 9.77904647471628e-09, + "loss": 0.7356, + "step": 8062 + }, + { + "epoch": 0.9695184272229904, + "grad_norm": 1.5344102905938721, + "learning_rate": 9.702259878815454e-09, + "loss": 0.7388, + "step": 8063 + }, + { + "epoch": 0.9696386701136296, + "grad_norm": 2.252923337929934, + "learning_rate": 9.625775209499254e-09, + "loss": 0.7495, + "step": 8064 + }, + { + "epoch": 0.9697589130042686, + "grad_norm": 2.4483913745929495, + "learning_rate": 9.549592478370172e-09, + "loss": 0.745, + "step": 8065 + }, + { + "epoch": 0.9698791558949077, + "grad_norm": 1.7205996877385998, + "learning_rate": 9.473711696985632e-09, + "loss": 0.7942, + "step": 8066 + }, + { + "epoch": 0.9699993987855468, + "grad_norm": 2.157170527685829, + "learning_rate": 9.398132876856201e-09, + "loss": 0.7635, + "step": 8067 + }, + { + "epoch": 0.9701196416761859, + "grad_norm": 0.7777238232956833, + "learning_rate": 9.322856029447379e-09, + "loss": 0.6084, + "step": 8068 + }, + { + "epoch": 0.970239884566825, + "grad_norm": 2.1326718685750112, + "learning_rate": 9.247881166178695e-09, + "loss": 0.804, + "step": 8069 + }, + { + "epoch": 0.970360127457464, + "grad_norm": 2.10108834629154, + "learning_rate": 9.173208298423274e-09, + "loss": 0.7729, + "step": 8070 + }, + { + "epoch": 0.9704803703481032, + "grad_norm": 2.04004432162588, + "learning_rate": 9.09883743750961e-09, + "loss": 0.7621, + "step": 8071 + }, + { + "epoch": 0.9706006132387422, + "grad_norm": 1.7164515993276572, + "learning_rate": 9.024768594719124e-09, + "loss": 0.8408, + "step": 8072 + }, + { + "epoch": 0.9707208561293813, + "grad_norm": 1.9492640060033166, + "learning_rate": 8.95100178128816e-09, + "loss": 0.7227, + "step": 8073 + }, + { + "epoch": 0.9708410990200205, + "grad_norm": 1.9129363815010707, + "learning_rate": 8.877537008407321e-09, + "loss": 0.6985, + "step": 8074 + }, + { + "epoch": 0.9709613419106595, + "grad_norm": 1.6298069647710989, + "learning_rate": 8.804374287221028e-09, + "loss": 0.6905, + "step": 8075 + }, + { + "epoch": 0.9710815848012986, + "grad_norm": 1.6644619028092293, + "learning_rate": 8.731513628827958e-09, + "loss": 0.8502, + "step": 8076 + }, + { + "epoch": 0.9712018276919377, + "grad_norm": 2.027971526982494, + "learning_rate": 8.658955044280825e-09, + "loss": 0.8294, + "step": 8077 + }, + { + "epoch": 0.9713220705825768, + "grad_norm": 1.4685683442455004, + "learning_rate": 8.586698544587268e-09, + "loss": 0.7771, + "step": 8078 + }, + { + "epoch": 0.9714423134732159, + "grad_norm": 1.9185626009902457, + "learning_rate": 8.514744140707853e-09, + "loss": 0.7487, + "step": 8079 + }, + { + "epoch": 0.971562556363855, + "grad_norm": 1.7051142003808226, + "learning_rate": 8.443091843558515e-09, + "loss": 0.7673, + "step": 8080 + }, + { + "epoch": 0.9716827992544941, + "grad_norm": 2.4874701263320795, + "learning_rate": 8.37174166400878e-09, + "loss": 0.6475, + "step": 8081 + }, + { + "epoch": 0.9718030421451331, + "grad_norm": 3.560480716695424, + "learning_rate": 8.300693612881992e-09, + "loss": 0.8512, + "step": 8082 + }, + { + "epoch": 0.9719232850357723, + "grad_norm": 2.0362219189510147, + "learning_rate": 8.22994770095664e-09, + "loss": 0.8158, + "step": 8083 + }, + { + "epoch": 0.9720435279264114, + "grad_norm": 2.237564691215762, + "learning_rate": 8.159503938964585e-09, + "loss": 0.763, + "step": 8084 + }, + { + "epoch": 0.9721637708170504, + "grad_norm": 1.9382876364169526, + "learning_rate": 8.089362337592164e-09, + "loss": 0.7033, + "step": 8085 + }, + { + "epoch": 0.9722840137076896, + "grad_norm": 1.6936119445617426, + "learning_rate": 8.019522907479536e-09, + "loss": 0.7262, + "step": 8086 + }, + { + "epoch": 0.9724042565983286, + "grad_norm": 1.9938030797894217, + "learning_rate": 7.949985659221558e-09, + "loss": 0.778, + "step": 8087 + }, + { + "epoch": 0.9725244994889677, + "grad_norm": 2.1870897180812046, + "learning_rate": 7.880750603366904e-09, + "loss": 0.7908, + "step": 8088 + }, + { + "epoch": 0.9726447423796069, + "grad_norm": 1.7741156806386567, + "learning_rate": 7.811817750418282e-09, + "loss": 0.8014, + "step": 8089 + }, + { + "epoch": 0.9727649852702459, + "grad_norm": 1.7705648405692558, + "learning_rate": 7.743187110833105e-09, + "loss": 0.8007, + "step": 8090 + }, + { + "epoch": 0.972885228160885, + "grad_norm": 1.6740011795659622, + "learning_rate": 7.674858695022602e-09, + "loss": 0.8095, + "step": 8091 + }, + { + "epoch": 0.9730054710515241, + "grad_norm": 4.223142018991694, + "learning_rate": 7.606832513351591e-09, + "loss": 0.7648, + "step": 8092 + }, + { + "epoch": 0.9731257139421632, + "grad_norm": 0.8627406019673347, + "learning_rate": 7.539108576140264e-09, + "loss": 0.6399, + "step": 8093 + }, + { + "epoch": 0.9732459568328022, + "grad_norm": 1.9238143688039278, + "learning_rate": 7.471686893661732e-09, + "loss": 0.704, + "step": 8094 + }, + { + "epoch": 0.9733661997234414, + "grad_norm": 1.673146343814681, + "learning_rate": 7.4045674761442636e-09, + "loss": 0.6476, + "step": 8095 + }, + { + "epoch": 0.9734864426140805, + "grad_norm": 1.7804325302611195, + "learning_rate": 7.337750333769488e-09, + "loss": 0.747, + "step": 8096 + }, + { + "epoch": 0.9736066855047195, + "grad_norm": 2.571367319830765, + "learning_rate": 7.2712354766737425e-09, + "loss": 0.7286, + "step": 8097 + }, + { + "epoch": 0.9737269283953586, + "grad_norm": 1.589683730861664, + "learning_rate": 7.2050229149469565e-09, + "loss": 0.8145, + "step": 8098 + }, + { + "epoch": 0.9738471712859977, + "grad_norm": 1.7989512514270602, + "learning_rate": 7.139112658633984e-09, + "loss": 0.6365, + "step": 8099 + }, + { + "epoch": 0.9739674141766368, + "grad_norm": 2.1135779435184396, + "learning_rate": 7.073504717733048e-09, + "loss": 0.7017, + "step": 8100 + }, + { + "epoch": 0.9740876570672758, + "grad_norm": 0.7476878686674568, + "learning_rate": 7.008199102196855e-09, + "loss": 0.573, + "step": 8101 + }, + { + "epoch": 0.974207899957915, + "grad_norm": 1.0756173756322003, + "learning_rate": 6.9431958219321464e-09, + "loss": 0.5903, + "step": 8102 + }, + { + "epoch": 0.9743281428485541, + "grad_norm": 1.7437446375391825, + "learning_rate": 6.878494886800146e-09, + "loss": 0.7778, + "step": 8103 + }, + { + "epoch": 0.9744483857391931, + "grad_norm": 1.9153365896648518, + "learning_rate": 6.814096306615669e-09, + "loss": 0.7631, + "step": 8104 + }, + { + "epoch": 0.9745686286298323, + "grad_norm": 2.194214135436571, + "learning_rate": 6.750000091148011e-09, + "loss": 0.654, + "step": 8105 + }, + { + "epoch": 0.9746888715204713, + "grad_norm": 1.8566650389019848, + "learning_rate": 6.686206250120729e-09, + "loss": 0.7306, + "step": 8106 + }, + { + "epoch": 0.9748091144111104, + "grad_norm": 2.053537024926539, + "learning_rate": 6.622714793210749e-09, + "loss": 0.7471, + "step": 8107 + }, + { + "epoch": 0.9749293573017496, + "grad_norm": 1.69838095871794, + "learning_rate": 6.559525730050364e-09, + "loss": 0.7854, + "step": 8108 + }, + { + "epoch": 0.9750496001923886, + "grad_norm": 1.8741816960273945, + "learning_rate": 6.496639070224574e-09, + "loss": 0.7642, + "step": 8109 + }, + { + "epoch": 0.9751698430830277, + "grad_norm": 2.5487285024532387, + "learning_rate": 6.4340548232739714e-09, + "loss": 0.8416, + "step": 8110 + }, + { + "epoch": 0.9752900859736668, + "grad_norm": 1.7383596338322733, + "learning_rate": 6.371772998692071e-09, + "loss": 0.7946, + "step": 8111 + }, + { + "epoch": 0.9754103288643059, + "grad_norm": 4.962410775638869, + "learning_rate": 6.309793605927094e-09, + "loss": 0.6532, + "step": 8112 + }, + { + "epoch": 0.975530571754945, + "grad_norm": 2.772346389408916, + "learning_rate": 6.248116654381297e-09, + "loss": 0.8027, + "step": 8113 + }, + { + "epoch": 0.9756508146455841, + "grad_norm": 2.0905282584075344, + "learning_rate": 6.186742153410751e-09, + "loss": 0.7289, + "step": 8114 + }, + { + "epoch": 0.9757710575362232, + "grad_norm": 1.9298822713927803, + "learning_rate": 6.125670112326453e-09, + "loss": 0.8764, + "step": 8115 + }, + { + "epoch": 0.9758913004268622, + "grad_norm": 2.429262473658224, + "learning_rate": 6.064900540392548e-09, + "loss": 0.7066, + "step": 8116 + }, + { + "epoch": 0.9760115433175014, + "grad_norm": 1.9251668387110137, + "learning_rate": 6.0044334468278835e-09, + "loss": 0.7887, + "step": 8117 + }, + { + "epoch": 0.9761317862081405, + "grad_norm": 5.653234362875528, + "learning_rate": 5.944268840805345e-09, + "loss": 0.7173, + "step": 8118 + }, + { + "epoch": 0.9762520290987795, + "grad_norm": 3.162237097627072, + "learning_rate": 5.88440673145163e-09, + "loss": 0.6451, + "step": 8119 + }, + { + "epoch": 0.9763722719894187, + "grad_norm": 2.54188113532394, + "learning_rate": 5.824847127848142e-09, + "loss": 0.8306, + "step": 8120 + }, + { + "epoch": 0.9764925148800577, + "grad_norm": 1.9090060962481747, + "learning_rate": 5.765590039029433e-09, + "loss": 0.7914, + "step": 8121 + }, + { + "epoch": 0.9766127577706968, + "grad_norm": 1.6594738803289488, + "learning_rate": 5.706635473985422e-09, + "loss": 0.7109, + "step": 8122 + }, + { + "epoch": 0.976733000661336, + "grad_norm": 2.5787115181267333, + "learning_rate": 5.6479834416591764e-09, + "loss": 0.8523, + "step": 8123 + }, + { + "epoch": 0.976853243551975, + "grad_norm": 1.9037707105625918, + "learning_rate": 5.589633950947803e-09, + "loss": 0.6903, + "step": 8124 + }, + { + "epoch": 0.9769734864426141, + "grad_norm": 2.115577676939688, + "learning_rate": 5.5315870107035535e-09, + "loss": 0.7123, + "step": 8125 + }, + { + "epoch": 0.9770937293332532, + "grad_norm": 1.7564348714054312, + "learning_rate": 5.473842629731607e-09, + "loss": 0.7905, + "step": 8126 + }, + { + "epoch": 0.9772139722238923, + "grad_norm": 2.179403854017201, + "learning_rate": 5.416400816792066e-09, + "loss": 0.7853, + "step": 8127 + }, + { + "epoch": 0.9773342151145313, + "grad_norm": 2.3544725427555626, + "learning_rate": 5.359261580598407e-09, + "loss": 0.7843, + "step": 8128 + }, + { + "epoch": 0.9774544580051704, + "grad_norm": 3.6975916473399972, + "learning_rate": 5.302424929819027e-09, + "loss": 0.782, + "step": 8129 + }, + { + "epoch": 0.9775747008958096, + "grad_norm": 2.1216680546606437, + "learning_rate": 5.24589087307592e-09, + "loss": 0.7316, + "step": 8130 + }, + { + "epoch": 0.9776949437864486, + "grad_norm": 1.5482533796764157, + "learning_rate": 5.189659418944891e-09, + "loss": 0.6503, + "step": 8131 + }, + { + "epoch": 0.9778151866770877, + "grad_norm": 2.39218477713681, + "learning_rate": 5.133730575956674e-09, + "loss": 0.7864, + "step": 8132 + }, + { + "epoch": 0.9779354295677268, + "grad_norm": 1.9674415478875726, + "learning_rate": 5.0781043525953696e-09, + "loss": 0.724, + "step": 8133 + }, + { + "epoch": 0.9780556724583659, + "grad_norm": 1.9587374195895035, + "learning_rate": 5.0227807572995605e-09, + "loss": 0.7412, + "step": 8134 + }, + { + "epoch": 0.9781759153490049, + "grad_norm": 2.693166126956158, + "learning_rate": 4.967759798461646e-09, + "loss": 0.6831, + "step": 8135 + }, + { + "epoch": 0.9782961582396441, + "grad_norm": 24.76682888438246, + "learning_rate": 4.913041484428282e-09, + "loss": 0.7498, + "step": 8136 + }, + { + "epoch": 0.9784164011302832, + "grad_norm": 1.8359025667803544, + "learning_rate": 4.858625823500384e-09, + "loss": 0.7443, + "step": 8137 + }, + { + "epoch": 0.9785366440209222, + "grad_norm": 2.0461142487607904, + "learning_rate": 4.80451282393246e-09, + "loss": 0.7356, + "step": 8138 + }, + { + "epoch": 0.9786568869115614, + "grad_norm": 1.989783478368122, + "learning_rate": 4.750702493933722e-09, + "loss": 0.6812, + "step": 8139 + }, + { + "epoch": 0.9787771298022004, + "grad_norm": 1.946009499568825, + "learning_rate": 4.697194841666974e-09, + "loss": 0.8562, + "step": 8140 + }, + { + "epoch": 0.9788973726928395, + "grad_norm": 1.8294208856841936, + "learning_rate": 4.6439898752492764e-09, + "loss": 0.8229, + "step": 8141 + }, + { + "epoch": 0.9790176155834787, + "grad_norm": 0.7634032574600866, + "learning_rate": 4.591087602751731e-09, + "loss": 0.637, + "step": 8142 + }, + { + "epoch": 0.9791378584741177, + "grad_norm": 1.7569431443978158, + "learning_rate": 4.538488032199916e-09, + "loss": 0.7226, + "step": 8143 + }, + { + "epoch": 0.9792581013647568, + "grad_norm": 2.036004942027991, + "learning_rate": 4.486191171572784e-09, + "loss": 0.6931, + "step": 8144 + }, + { + "epoch": 0.9793783442553959, + "grad_norm": 1.7173293744973543, + "learning_rate": 4.434197028803766e-09, + "loss": 0.7794, + "step": 8145 + }, + { + "epoch": 0.979498587146035, + "grad_norm": 2.045427815426425, + "learning_rate": 4.3825056117805514e-09, + "loss": 0.819, + "step": 8146 + }, + { + "epoch": 0.979618830036674, + "grad_norm": 2.1863893141506674, + "learning_rate": 4.331116928344425e-09, + "loss": 0.7945, + "step": 8147 + }, + { + "epoch": 0.9797390729273132, + "grad_norm": 2.160724254796834, + "learning_rate": 4.28003098629115e-09, + "loss": 0.633, + "step": 8148 + }, + { + "epoch": 0.9798593158179523, + "grad_norm": 1.8544791383369557, + "learning_rate": 4.229247793370305e-09, + "loss": 0.785, + "step": 8149 + }, + { + "epoch": 0.9799795587085913, + "grad_norm": 1.820225725848017, + "learning_rate": 4.178767357285951e-09, + "loss": 0.7036, + "step": 8150 + }, + { + "epoch": 0.9800998015992305, + "grad_norm": 1.8435032566291139, + "learning_rate": 4.128589685695516e-09, + "loss": 0.7186, + "step": 8151 + }, + { + "epoch": 0.9802200444898695, + "grad_norm": 1.8515944981123873, + "learning_rate": 4.078714786211135e-09, + "loss": 0.8464, + "step": 8152 + }, + { + "epoch": 0.9803402873805086, + "grad_norm": 7.868164971550307, + "learning_rate": 4.029142666398977e-09, + "loss": 0.7693, + "step": 8153 + }, + { + "epoch": 0.9804605302711478, + "grad_norm": 2.022084730257749, + "learning_rate": 3.979873333778805e-09, + "loss": 0.8011, + "step": 8154 + }, + { + "epoch": 0.9805807731617868, + "grad_norm": 1.9357304028939537, + "learning_rate": 3.930906795824862e-09, + "loss": 0.7411, + "step": 8155 + }, + { + "epoch": 0.9807010160524259, + "grad_norm": 2.0621469368885545, + "learning_rate": 3.882243059965207e-09, + "loss": 0.7703, + "step": 8156 + }, + { + "epoch": 0.980821258943065, + "grad_norm": 2.242878239266032, + "learning_rate": 3.833882133582156e-09, + "loss": 0.6563, + "step": 8157 + }, + { + "epoch": 0.9809415018337041, + "grad_norm": 6.783201437877706, + "learning_rate": 3.785824024012285e-09, + "loss": 0.7832, + "step": 8158 + }, + { + "epoch": 0.9810617447243432, + "grad_norm": 1.629623584463165, + "learning_rate": 3.738068738545541e-09, + "loss": 0.7868, + "step": 8159 + }, + { + "epoch": 0.9811819876149822, + "grad_norm": 2.511196242308198, + "learning_rate": 3.6906162844265733e-09, + "loss": 0.7941, + "step": 8160 + }, + { + "epoch": 0.9813022305056214, + "grad_norm": 1.735747816433605, + "learning_rate": 3.643466668853845e-09, + "loss": 0.7171, + "step": 8161 + }, + { + "epoch": 0.9814224733962604, + "grad_norm": 2.561178843755771, + "learning_rate": 3.59661989898008e-09, + "loss": 0.7583, + "step": 8162 + }, + { + "epoch": 0.9815427162868995, + "grad_norm": 1.7881612714964046, + "learning_rate": 3.5500759819115934e-09, + "loss": 0.7722, + "step": 8163 + }, + { + "epoch": 0.9816629591775387, + "grad_norm": 1.8411292507462618, + "learning_rate": 3.5038349247094034e-09, + "loss": 0.8146, + "step": 8164 + }, + { + "epoch": 0.9817832020681777, + "grad_norm": 2.012459805484066, + "learning_rate": 3.4578967343878994e-09, + "loss": 0.7792, + "step": 8165 + }, + { + "epoch": 0.9819034449588168, + "grad_norm": 1.7970742820030063, + "learning_rate": 3.4122614179161733e-09, + "loss": 0.8144, + "step": 8166 + }, + { + "epoch": 0.9820236878494559, + "grad_norm": 1.7956959712458793, + "learning_rate": 3.36692898221691e-09, + "loss": 0.7859, + "step": 8167 + }, + { + "epoch": 0.982143930740095, + "grad_norm": 1.7082152287045462, + "learning_rate": 3.3218994341668305e-09, + "loss": 0.7375, + "step": 8168 + }, + { + "epoch": 0.982264173630734, + "grad_norm": 1.5446438352296663, + "learning_rate": 3.2771727805971373e-09, + "loss": 0.753, + "step": 8169 + }, + { + "epoch": 0.9823844165213732, + "grad_norm": 1.7358474666894856, + "learning_rate": 3.232749028292847e-09, + "loss": 0.7771, + "step": 8170 + }, + { + "epoch": 0.9825046594120123, + "grad_norm": 1.7028262375032264, + "learning_rate": 3.188628183992792e-09, + "loss": 0.8885, + "step": 8171 + }, + { + "epoch": 0.9826249023026513, + "grad_norm": 0.7498144476559142, + "learning_rate": 3.1448102543902844e-09, + "loss": 0.6255, + "step": 8172 + }, + { + "epoch": 0.9827451451932905, + "grad_norm": 1.9339930883005974, + "learning_rate": 3.1012952461324515e-09, + "loss": 0.6822, + "step": 8173 + }, + { + "epoch": 0.9828653880839295, + "grad_norm": 2.205059095356329, + "learning_rate": 3.0580831658204575e-09, + "loss": 0.7403, + "step": 8174 + }, + { + "epoch": 0.9829856309745686, + "grad_norm": 1.776415437836835, + "learning_rate": 3.015174020009281e-09, + "loss": 0.7796, + "step": 8175 + }, + { + "epoch": 0.9831058738652078, + "grad_norm": 1.8684292419118391, + "learning_rate": 2.9725678152086043e-09, + "loss": 0.7511, + "step": 8176 + }, + { + "epoch": 0.9832261167558468, + "grad_norm": 2.9019645350765053, + "learning_rate": 2.930264557881257e-09, + "loss": 0.8283, + "step": 8177 + }, + { + "epoch": 0.9833463596464859, + "grad_norm": 0.946323043795744, + "learning_rate": 2.8882642544452163e-09, + "loss": 0.5818, + "step": 8178 + }, + { + "epoch": 0.983466602537125, + "grad_norm": 2.281927974757841, + "learning_rate": 2.8465669112716083e-09, + "loss": 0.751, + "step": 8179 + }, + { + "epoch": 0.9835868454277641, + "grad_norm": 1.8512938743198601, + "learning_rate": 2.8051725346858177e-09, + "loss": 0.764, + "step": 8180 + }, + { + "epoch": 0.9837070883184031, + "grad_norm": 2.227604779349002, + "learning_rate": 2.7640811309674883e-09, + "loss": 0.7107, + "step": 8181 + }, + { + "epoch": 0.9838273312090423, + "grad_norm": 1.5840892827563124, + "learning_rate": 2.7232927063498557e-09, + "loss": 0.8086, + "step": 8182 + }, + { + "epoch": 0.9839475740996814, + "grad_norm": 1.8324992348746125, + "learning_rate": 2.682807267020859e-09, + "loss": 0.6924, + "step": 8183 + }, + { + "epoch": 0.9840678169903204, + "grad_norm": 1.7129116426785649, + "learning_rate": 2.642624819121808e-09, + "loss": 0.6251, + "step": 8184 + }, + { + "epoch": 0.9841880598809596, + "grad_norm": 2.1103935890655663, + "learning_rate": 2.6027453687487154e-09, + "loss": 0.6242, + "step": 8185 + }, + { + "epoch": 0.9843083027715986, + "grad_norm": 2.277736204693196, + "learning_rate": 2.5631689219509643e-09, + "loss": 0.5403, + "step": 8186 + }, + { + "epoch": 0.9844285456622377, + "grad_norm": 1.7347385609677464, + "learning_rate": 2.523895484732197e-09, + "loss": 0.8373, + "step": 8187 + }, + { + "epoch": 0.9845487885528769, + "grad_norm": 1.938761700238869, + "learning_rate": 2.4849250630505357e-09, + "loss": 0.7495, + "step": 8188 + }, + { + "epoch": 0.9846690314435159, + "grad_norm": 1.903765763776691, + "learning_rate": 2.4462576628172528e-09, + "loss": 0.7321, + "step": 8189 + }, + { + "epoch": 0.984789274334155, + "grad_norm": 1.9258576352830896, + "learning_rate": 2.407893289898766e-09, + "loss": 0.7461, + "step": 8190 + }, + { + "epoch": 0.984909517224794, + "grad_norm": 2.1483433928252937, + "learning_rate": 2.3698319501144202e-09, + "loss": 0.844, + "step": 8191 + }, + { + "epoch": 0.9850297601154332, + "grad_norm": 1.9642120619998278, + "learning_rate": 2.3320736492382644e-09, + "loss": 0.7351, + "step": 8192 + }, + { + "epoch": 0.9851500030060723, + "grad_norm": 2.006804454398729, + "learning_rate": 2.29461839299816e-09, + "loss": 0.6817, + "step": 8193 + }, + { + "epoch": 0.9852702458967113, + "grad_norm": 1.5861157096370049, + "learning_rate": 2.257466187076229e-09, + "loss": 0.8043, + "step": 8194 + }, + { + "epoch": 0.9853904887873505, + "grad_norm": 1.8894748184129113, + "learning_rate": 2.2206170371081854e-09, + "loss": 0.7152, + "step": 8195 + }, + { + "epoch": 0.9855107316779895, + "grad_norm": 1.710884267597261, + "learning_rate": 2.1840709486842247e-09, + "loss": 0.8518, + "step": 8196 + }, + { + "epoch": 0.9856309745686286, + "grad_norm": 1.9269054296356949, + "learning_rate": 2.1478279273481335e-09, + "loss": 0.7952, + "step": 8197 + }, + { + "epoch": 0.9857512174592677, + "grad_norm": 2.110870283455774, + "learning_rate": 2.1118879785981815e-09, + "loss": 0.7988, + "step": 8198 + }, + { + "epoch": 0.9858714603499068, + "grad_norm": 1.829173512833397, + "learning_rate": 2.0762511078862288e-09, + "loss": 0.7971, + "step": 8199 + }, + { + "epoch": 0.9859917032405459, + "grad_norm": 2.2982277552103785, + "learning_rate": 2.0409173206186183e-09, + "loss": 0.6534, + "step": 8200 + }, + { + "epoch": 0.986111946131185, + "grad_norm": 2.005851292431379, + "learning_rate": 2.0058866221550617e-09, + "loss": 0.877, + "step": 8201 + }, + { + "epoch": 0.9862321890218241, + "grad_norm": 2.1268596365903707, + "learning_rate": 1.971159017809976e-09, + "loss": 0.755, + "step": 8202 + }, + { + "epoch": 0.9863524319124631, + "grad_norm": 2.4645182399156043, + "learning_rate": 1.93673451285159e-09, + "loss": 0.7846, + "step": 8203 + }, + { + "epoch": 0.9864726748031023, + "grad_norm": 0.7546056556945698, + "learning_rate": 1.9026131125019495e-09, + "loss": 0.5655, + "step": 8204 + }, + { + "epoch": 0.9865929176937414, + "grad_norm": 1.8153009148410435, + "learning_rate": 1.8687948219371363e-09, + "loss": 0.8734, + "step": 8205 + }, + { + "epoch": 0.9867131605843804, + "grad_norm": 2.3018959855245043, + "learning_rate": 1.835279646287491e-09, + "loss": 0.8923, + "step": 8206 + }, + { + "epoch": 0.9868334034750196, + "grad_norm": 1.7429669608017626, + "learning_rate": 1.8020675906371685e-09, + "loss": 0.7677, + "step": 8207 + }, + { + "epoch": 0.9869536463656586, + "grad_norm": 2.0948797565631323, + "learning_rate": 1.7691586600243612e-09, + "loss": 0.7582, + "step": 8208 + }, + { + "epoch": 0.9870738892562977, + "grad_norm": 2.4594134743382106, + "learning_rate": 1.7365528594415202e-09, + "loss": 0.8769, + "step": 8209 + }, + { + "epoch": 0.9871941321469369, + "grad_norm": 1.5748863603073335, + "learning_rate": 1.7042501938346888e-09, + "loss": 0.6747, + "step": 8210 + }, + { + "epoch": 0.9873143750375759, + "grad_norm": 2.0437762850783074, + "learning_rate": 1.6722506681043913e-09, + "loss": 0.77, + "step": 8211 + }, + { + "epoch": 0.987434617928215, + "grad_norm": 2.109049567090952, + "learning_rate": 1.640554287104745e-09, + "loss": 0.6917, + "step": 8212 + }, + { + "epoch": 0.9875548608188541, + "grad_norm": 2.0052457203207656, + "learning_rate": 1.609161055644348e-09, + "loss": 0.8016, + "step": 8213 + }, + { + "epoch": 0.9876751037094932, + "grad_norm": 2.7504266928047136, + "learning_rate": 1.5780709784849467e-09, + "loss": 0.6879, + "step": 8214 + }, + { + "epoch": 0.9877953466001322, + "grad_norm": 37.61451265801539, + "learning_rate": 1.5472840603436565e-09, + "loss": 0.8262, + "step": 8215 + }, + { + "epoch": 0.9879155894907714, + "grad_norm": 1.9876044119993403, + "learning_rate": 1.5168003058900757e-09, + "loss": 0.7811, + "step": 8216 + }, + { + "epoch": 0.9880358323814105, + "grad_norm": 2.1590519267309336, + "learning_rate": 1.4866197197491715e-09, + "loss": 0.9221, + "step": 8217 + }, + { + "epoch": 0.9881560752720495, + "grad_norm": 3.0989891926914583, + "learning_rate": 1.4567423064988371e-09, + "loss": 0.7875, + "step": 8218 + }, + { + "epoch": 0.9882763181626887, + "grad_norm": 2.0328514242000852, + "learning_rate": 1.4271680706718913e-09, + "loss": 0.7838, + "step": 8219 + }, + { + "epoch": 0.9883965610533277, + "grad_norm": 2.0751509645996618, + "learning_rate": 1.3978970167543013e-09, + "loss": 0.8327, + "step": 8220 + }, + { + "epoch": 0.9885168039439668, + "grad_norm": 1.948715909570394, + "learning_rate": 1.3689291491867372e-09, + "loss": 0.7745, + "step": 8221 + }, + { + "epoch": 0.988637046834606, + "grad_norm": 2.6947063261973914, + "learning_rate": 1.3402644723636836e-09, + "loss": 0.7424, + "step": 8222 + }, + { + "epoch": 0.988757289725245, + "grad_norm": 2.049653009961119, + "learning_rate": 1.311902990633218e-09, + "loss": 0.8386, + "step": 8223 + }, + { + "epoch": 0.9888775326158841, + "grad_norm": 1.6940740465376913, + "learning_rate": 1.2838447082978987e-09, + "loss": 0.7157, + "step": 8224 + }, + { + "epoch": 0.9889977755065231, + "grad_norm": 2.8107876121164415, + "learning_rate": 1.2560896296143208e-09, + "loss": 0.8328, + "step": 8225 + }, + { + "epoch": 0.9891180183971623, + "grad_norm": 2.243162468198443, + "learning_rate": 1.2286377587926722e-09, + "loss": 0.8231, + "step": 8226 + }, + { + "epoch": 0.9892382612878013, + "grad_norm": 2.423071879310622, + "learning_rate": 1.2014890999973992e-09, + "loss": 0.7529, + "step": 8227 + }, + { + "epoch": 0.9893585041784404, + "grad_norm": 1.8433018155820948, + "learning_rate": 1.1746436573472073e-09, + "loss": 0.7885, + "step": 8228 + }, + { + "epoch": 0.9894787470690796, + "grad_norm": 2.0418008124468803, + "learning_rate": 1.1481014349141726e-09, + "loss": 0.6995, + "step": 8229 + }, + { + "epoch": 0.9895989899597186, + "grad_norm": 1.8287903968588939, + "learning_rate": 1.121862436724852e-09, + "loss": 0.847, + "step": 8230 + }, + { + "epoch": 0.9897192328503577, + "grad_norm": 1.712727624843085, + "learning_rate": 1.0959266667598388e-09, + "loss": 0.7056, + "step": 8231 + }, + { + "epoch": 0.9898394757409968, + "grad_norm": 2.283853273525665, + "learning_rate": 1.0702941289533196e-09, + "loss": 0.7537, + "step": 8232 + }, + { + "epoch": 0.9899597186316359, + "grad_norm": 1.9802077382859553, + "learning_rate": 1.0449648271939615e-09, + "loss": 0.8916, + "step": 8233 + }, + { + "epoch": 0.990079961522275, + "grad_norm": 1.8185702651311992, + "learning_rate": 1.0199387653240243e-09, + "loss": 0.7292, + "step": 8234 + }, + { + "epoch": 0.9902002044129141, + "grad_norm": 1.570478743678017, + "learning_rate": 9.952159471400267e-10, + "loss": 0.7101, + "step": 8235 + }, + { + "epoch": 0.9903204473035532, + "grad_norm": 2.220131135280756, + "learning_rate": 9.707963763923022e-10, + "loss": 0.8454, + "step": 8236 + }, + { + "epoch": 0.9904406901941922, + "grad_norm": 1.7166279958177866, + "learning_rate": 9.466800567854427e-10, + "loss": 0.7933, + "step": 8237 + }, + { + "epoch": 0.9905609330848314, + "grad_norm": 2.042563053497896, + "learning_rate": 9.228669919778553e-10, + "loss": 0.6794, + "step": 8238 + }, + { + "epoch": 0.9906811759754705, + "grad_norm": 3.5290747500313158, + "learning_rate": 8.993571855817617e-10, + "loss": 0.7958, + "step": 8239 + }, + { + "epoch": 0.9908014188661095, + "grad_norm": 2.575051620114137, + "learning_rate": 8.761506411638642e-10, + "loss": 0.7514, + "step": 8240 + }, + { + "epoch": 0.9909216617567487, + "grad_norm": 1.7542175152703692, + "learning_rate": 8.53247362244236e-10, + "loss": 0.747, + "step": 8241 + }, + { + "epoch": 0.9910419046473877, + "grad_norm": 1.7641187336205677, + "learning_rate": 8.306473522976532e-10, + "loss": 0.686, + "step": 8242 + }, + { + "epoch": 0.9911621475380268, + "grad_norm": 1.7738345873760557, + "learning_rate": 8.083506147522623e-10, + "loss": 0.7164, + "step": 8243 + }, + { + "epoch": 0.991282390428666, + "grad_norm": 2.084940239761141, + "learning_rate": 7.863571529906909e-10, + "loss": 0.8599, + "step": 8244 + }, + { + "epoch": 0.991402633319305, + "grad_norm": 0.7370344664693734, + "learning_rate": 7.646669703489372e-10, + "loss": 0.5974, + "step": 8245 + }, + { + "epoch": 0.9915228762099441, + "grad_norm": 2.0626787050451125, + "learning_rate": 7.432800701177023e-10, + "loss": 0.5751, + "step": 8246 + }, + { + "epoch": 0.9916431191005832, + "grad_norm": 0.8184932699352083, + "learning_rate": 7.221964555415017e-10, + "loss": 0.5783, + "step": 8247 + }, + { + "epoch": 0.9917633619912223, + "grad_norm": 1.766433125780312, + "learning_rate": 7.01416129818222e-10, + "loss": 0.7506, + "step": 8248 + }, + { + "epoch": 0.9918836048818613, + "grad_norm": 2.3691268379583894, + "learning_rate": 6.809390961006745e-10, + "loss": 0.5923, + "step": 8249 + }, + { + "epoch": 0.9920038477725005, + "grad_norm": 1.7684299839990514, + "learning_rate": 6.607653574948191e-10, + "loss": 0.6868, + "step": 8250 + }, + { + "epoch": 0.9921240906631396, + "grad_norm": 1.7818500779575657, + "learning_rate": 6.408949170613187e-10, + "loss": 0.8201, + "step": 8251 + }, + { + "epoch": 0.9922443335537786, + "grad_norm": 1.9493800767329168, + "learning_rate": 6.213277778144288e-10, + "loss": 0.8226, + "step": 8252 + }, + { + "epoch": 0.9923645764444178, + "grad_norm": 2.036316992756534, + "learning_rate": 6.020639427224416e-10, + "loss": 0.6748, + "step": 8253 + }, + { + "epoch": 0.9924848193350568, + "grad_norm": 2.0609898674062, + "learning_rate": 5.831034147076864e-10, + "loss": 0.7383, + "step": 8254 + }, + { + "epoch": 0.9926050622256959, + "grad_norm": 0.7021808855520223, + "learning_rate": 5.644461966463065e-10, + "loss": 0.5575, + "step": 8255 + }, + { + "epoch": 0.9927253051163349, + "grad_norm": 1.8089011406906659, + "learning_rate": 5.460922913687049e-10, + "loss": 0.7614, + "step": 8256 + }, + { + "epoch": 0.9928455480069741, + "grad_norm": 2.0264577822368954, + "learning_rate": 5.280417016593208e-10, + "loss": 0.7562, + "step": 8257 + }, + { + "epoch": 0.9929657908976132, + "grad_norm": 1.7490322979692512, + "learning_rate": 5.102944302559642e-10, + "loss": 0.7524, + "step": 8258 + }, + { + "epoch": 0.9930860337882522, + "grad_norm": 2.0496997706445383, + "learning_rate": 4.9285047985137e-10, + "loss": 0.7992, + "step": 8259 + }, + { + "epoch": 0.9932062766788914, + "grad_norm": 1.6344973983300508, + "learning_rate": 4.757098530916436e-10, + "loss": 0.7443, + "step": 8260 + }, + { + "epoch": 0.9933265195695304, + "grad_norm": 2.4462636249613077, + "learning_rate": 4.5887255257670563e-10, + "loss": 0.7822, + "step": 8261 + }, + { + "epoch": 0.9934467624601695, + "grad_norm": 3.6179087766298954, + "learning_rate": 4.4233858086117906e-10, + "loss": 0.7694, + "step": 8262 + }, + { + "epoch": 0.9935670053508087, + "grad_norm": 2.184257283674847, + "learning_rate": 4.261079404528356e-10, + "loss": 0.6783, + "step": 8263 + }, + { + "epoch": 0.9936872482414477, + "grad_norm": 1.6927602515450366, + "learning_rate": 4.1018063381437205e-10, + "loss": 0.6885, + "step": 8264 + }, + { + "epoch": 0.9938074911320868, + "grad_norm": 0.8739620251098305, + "learning_rate": 3.9455666336141167e-10, + "loss": 0.6115, + "step": 8265 + }, + { + "epoch": 0.9939277340227259, + "grad_norm": 2.6292293627760475, + "learning_rate": 3.7923603146450267e-10, + "loss": 0.8363, + "step": 8266 + }, + { + "epoch": 0.994047976913365, + "grad_norm": 2.013686201338768, + "learning_rate": 3.642187404473418e-10, + "loss": 0.8066, + "step": 8267 + }, + { + "epoch": 0.994168219804004, + "grad_norm": 2.5785921540538825, + "learning_rate": 3.495047925885508e-10, + "loss": 0.8662, + "step": 8268 + }, + { + "epoch": 0.9942884626946432, + "grad_norm": 2.2958811990925576, + "learning_rate": 3.350941901199e-10, + "loss": 0.8342, + "step": 8269 + }, + { + "epoch": 0.9944087055852823, + "grad_norm": 2.4201390105116265, + "learning_rate": 3.2098693522764066e-10, + "loss": 0.8311, + "step": 8270 + }, + { + "epoch": 0.9945289484759213, + "grad_norm": 1.840483989398257, + "learning_rate": 3.071830300516165e-10, + "loss": 0.8107, + "step": 8271 + }, + { + "epoch": 0.9946491913665605, + "grad_norm": 2.4030061134327285, + "learning_rate": 2.9368247668615234e-10, + "loss": 0.7147, + "step": 8272 + }, + { + "epoch": 0.9947694342571995, + "grad_norm": 2.862691154927219, + "learning_rate": 2.804852771789434e-10, + "loss": 0.6171, + "step": 8273 + }, + { + "epoch": 0.9948896771478386, + "grad_norm": 1.825175979699645, + "learning_rate": 2.675914335321661e-10, + "loss": 0.5584, + "step": 8274 + }, + { + "epoch": 0.9950099200384778, + "grad_norm": 2.750482516872732, + "learning_rate": 2.550009477018111e-10, + "loss": 0.7972, + "step": 8275 + }, + { + "epoch": 0.9951301629291168, + "grad_norm": 1.9010351172208098, + "learning_rate": 2.4271382159790634e-10, + "loss": 0.631, + "step": 8276 + }, + { + "epoch": 0.9952504058197559, + "grad_norm": 1.9110474798910542, + "learning_rate": 2.3073005708429406e-10, + "loss": 0.8608, + "step": 8277 + }, + { + "epoch": 0.995370648710395, + "grad_norm": 1.8944742410478683, + "learning_rate": 2.190496559788535e-10, + "loss": 0.7278, + "step": 8278 + }, + { + "epoch": 0.9954908916010341, + "grad_norm": 2.546309144986292, + "learning_rate": 2.0767262005372265e-10, + "loss": 0.7667, + "step": 8279 + }, + { + "epoch": 0.9956111344916732, + "grad_norm": 2.079669808442195, + "learning_rate": 1.965989510346322e-10, + "loss": 0.7544, + "step": 8280 + }, + { + "epoch": 0.9957313773823123, + "grad_norm": 2.1260578663900507, + "learning_rate": 1.8582865060134955e-10, + "loss": 0.7162, + "step": 8281 + }, + { + "epoch": 0.9958516202729514, + "grad_norm": 0.7919688568253643, + "learning_rate": 1.7536172038790098e-10, + "loss": 0.5581, + "step": 8282 + }, + { + "epoch": 0.9959718631635904, + "grad_norm": 2.948157436571695, + "learning_rate": 1.651981619819054e-10, + "loss": 0.6928, + "step": 8283 + }, + { + "epoch": 0.9960921060542296, + "grad_norm": 2.7809573244553563, + "learning_rate": 1.5533797692546257e-10, + "loss": 0.7113, + "step": 8284 + }, + { + "epoch": 0.9962123489448687, + "grad_norm": 1.936307232320587, + "learning_rate": 1.4578116671404296e-10, + "loss": 0.8421, + "step": 8285 + }, + { + "epoch": 0.9963325918355077, + "grad_norm": 2.0337130358806257, + "learning_rate": 1.3652773279759777e-10, + "loss": 0.7157, + "step": 8286 + }, + { + "epoch": 0.9964528347261468, + "grad_norm": 1.6169787285985657, + "learning_rate": 1.2757767657989305e-10, + "loss": 0.6319, + "step": 8287 + }, + { + "epoch": 0.9965730776167859, + "grad_norm": 5.226283318286337, + "learning_rate": 1.1893099941850948e-10, + "loss": 0.871, + "step": 8288 + }, + { + "epoch": 0.996693320507425, + "grad_norm": 2.1796507473379982, + "learning_rate": 1.105877026252866e-10, + "loss": 0.7793, + "step": 8289 + }, + { + "epoch": 0.996813563398064, + "grad_norm": 1.8889210486323424, + "learning_rate": 1.0254778746565663e-10, + "loss": 0.7244, + "step": 8290 + }, + { + "epoch": 0.9969338062887032, + "grad_norm": 2.022963774028164, + "learning_rate": 9.481125515953259e-11, + "loss": 0.7363, + "step": 8291 + }, + { + "epoch": 0.9970540491793423, + "grad_norm": 1.7213150304672729, + "learning_rate": 8.737810688064228e-11, + "loss": 0.7996, + "step": 8292 + }, + { + "epoch": 0.9971742920699813, + "grad_norm": 3.683070350913465, + "learning_rate": 8.024834375608414e-11, + "loss": 0.7974, + "step": 8293 + }, + { + "epoch": 0.9972945349606205, + "grad_norm": 0.8203717891520426, + "learning_rate": 7.342196686788149e-11, + "loss": 0.6284, + "step": 8294 + }, + { + "epoch": 0.9974147778512595, + "grad_norm": 1.9391778633418464, + "learning_rate": 6.689897725142834e-11, + "loss": 0.6881, + "step": 8295 + }, + { + "epoch": 0.9975350207418986, + "grad_norm": 2.075545351911882, + "learning_rate": 6.067937589615545e-11, + "loss": 0.8883, + "step": 8296 + }, + { + "epoch": 0.9976552636325378, + "grad_norm": 0.7667505981970345, + "learning_rate": 5.476316374575241e-11, + "loss": 0.5773, + "step": 8297 + }, + { + "epoch": 0.9977755065231768, + "grad_norm": 3.4164807028675583, + "learning_rate": 4.9150341697723476e-11, + "loss": 0.7313, + "step": 8298 + }, + { + "epoch": 0.9978957494138159, + "grad_norm": 1.6757754879298883, + "learning_rate": 4.384091060338768e-11, + "loss": 0.6673, + "step": 8299 + }, + { + "epoch": 0.998015992304455, + "grad_norm": 3.151373239942282, + "learning_rate": 3.883487126810081e-11, + "loss": 0.7372, + "step": 8300 + }, + { + "epoch": 0.9981362351950941, + "grad_norm": 1.6811575910279712, + "learning_rate": 3.41322244516995e-11, + "loss": 0.7922, + "step": 8301 + }, + { + "epoch": 0.9982564780857331, + "grad_norm": 1.6653384479284328, + "learning_rate": 2.9732970866946925e-11, + "loss": 0.6331, + "step": 8302 + }, + { + "epoch": 0.9983767209763723, + "grad_norm": 2.4895014623058076, + "learning_rate": 2.563711118175327e-11, + "loss": 0.788, + "step": 8303 + }, + { + "epoch": 0.9984969638670114, + "grad_norm": 2.083245873204864, + "learning_rate": 2.184464601717728e-11, + "loss": 0.8421, + "step": 8304 + }, + { + "epoch": 0.9986172067576504, + "grad_norm": 2.9791592862409035, + "learning_rate": 1.8355575948758585e-11, + "loss": 0.7775, + "step": 8305 + }, + { + "epoch": 0.9987374496482896, + "grad_norm": 2.011422203947182, + "learning_rate": 1.5169901505407424e-11, + "loss": 0.7334, + "step": 8306 + }, + { + "epoch": 0.9988576925389286, + "grad_norm": 2.230753491293024, + "learning_rate": 1.228762317073695e-11, + "loss": 0.744, + "step": 8307 + }, + { + "epoch": 0.9989779354295677, + "grad_norm": 2.0141082498046505, + "learning_rate": 9.70874138195299e-12, + "loss": 0.7998, + "step": 8308 + }, + { + "epoch": 0.9990981783202069, + "grad_norm": 1.5173114949874482, + "learning_rate": 7.433256530076093e-12, + "loss": 0.7499, + "step": 8309 + }, + { + "epoch": 0.9992184212108459, + "grad_norm": 2.182631169558123, + "learning_rate": 5.46116896038562e-12, + "loss": 0.7603, + "step": 8310 + }, + { + "epoch": 0.999338664101485, + "grad_norm": 2.006686477478852, + "learning_rate": 3.792478972197699e-12, + "loss": 0.6223, + "step": 8311 + }, + { + "epoch": 0.9994589069921241, + "grad_norm": 2.1219844739311786, + "learning_rate": 2.4271868181990895e-12, + "loss": 0.7012, + "step": 8312 + }, + { + "epoch": 0.9995791498827632, + "grad_norm": 2.16334626502045, + "learning_rate": 1.3652927060014973e-12, + "loss": 0.8153, + "step": 8313 + }, + { + "epoch": 0.9996993927734023, + "grad_norm": 2.449754671542567, + "learning_rate": 6.067967965872612e-13, + "loss": 0.6444, + "step": 8314 + }, + { + "epoch": 0.9998196356640414, + "grad_norm": 1.6151239618020978, + "learning_rate": 1.5169920497548615e-13, + "loss": 0.7736, + "step": 8315 + }, + { + "epoch": 0.9999398785546805, + "grad_norm": 1.1435982087516565, + "learning_rate": 0.0, + "loss": 0.5503, + "step": 8316 + }, + { + "epoch": 0.9999398785546805, + "step": 8316, + "total_flos": 6.738231971282944e+17, + "train_loss": 0.7759840532836287, + "train_runtime": 117446.8965, + "train_samples_per_second": 2.832, + "train_steps_per_second": 0.071 + } + ], + "logging_steps": 1.0, + "max_steps": 8316, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 6.738231971282944e+17, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +}