diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8592 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 122,
+  "global_step": 1218,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0024630541871921183,
+      "grad_norm": 6.755204349027433,
+      "learning_rate": 1.639344262295082e-07,
+      "loss": 1.7545,
+      "step": 1
+    },
+    {
+      "epoch": 0.0049261083743842365,
+      "grad_norm": 6.1608262456668,
+      "learning_rate": 3.278688524590164e-07,
+      "loss": 1.6567,
+      "step": 2
+    },
+    {
+      "epoch": 0.007389162561576354,
+      "grad_norm": 7.276454148357188,
+      "learning_rate": 4.918032786885246e-07,
+      "loss": 1.6064,
+      "step": 3
+    },
+    {
+      "epoch": 0.009852216748768473,
+      "grad_norm": 7.441939367386136,
+      "learning_rate": 6.557377049180328e-07,
+      "loss": 1.694,
+      "step": 4
+    },
+    {
+      "epoch": 0.012315270935960592,
+      "grad_norm": 6.801120317958336,
+      "learning_rate": 8.196721311475409e-07,
+      "loss": 1.674,
+      "step": 5
+    },
+    {
+      "epoch": 0.014778325123152709,
+      "grad_norm": 6.668480467421702,
+      "learning_rate": 9.836065573770493e-07,
+      "loss": 1.6892,
+      "step": 6
+    },
+    {
+      "epoch": 0.017241379310344827,
+      "grad_norm": 5.778850655509419,
+      "learning_rate": 1.1475409836065575e-06,
+      "loss": 1.65,
+      "step": 7
+    },
+    {
+      "epoch": 0.019704433497536946,
+      "grad_norm": 5.881726898089619,
+      "learning_rate": 1.3114754098360657e-06,
+      "loss": 1.4504,
+      "step": 8
+    },
+    {
+      "epoch": 0.022167487684729065,
+      "grad_norm": 6.267210495801273,
+      "learning_rate": 1.4754098360655739e-06,
+      "loss": 1.5861,
+      "step": 9
+    },
+    {
+      "epoch": 0.024630541871921183,
+      "grad_norm": 6.71823927246709,
+      "learning_rate": 1.6393442622950819e-06,
+      "loss": 1.6732,
+      "step": 10
+    },
+    {
+      "epoch": 0.027093596059113302,
+      "grad_norm": 5.63826797566246,
+      "learning_rate": 1.8032786885245903e-06,
+      "loss": 1.6,
+      "step": 11
+    },
+    {
+      "epoch": 0.029556650246305417,
+      "grad_norm": 5.210021434600727,
+      "learning_rate": 1.9672131147540985e-06,
+      "loss": 1.5305,
+      "step": 12
+    },
+    {
+      "epoch": 0.03201970443349754,
+      "grad_norm": 4.812748295707652,
+      "learning_rate": 2.1311475409836067e-06,
+      "loss": 1.4732,
+      "step": 13
+    },
+    {
+      "epoch": 0.034482758620689655,
+      "grad_norm": 5.049180015239499,
+      "learning_rate": 2.295081967213115e-06,
+      "loss": 1.5632,
+      "step": 14
+    },
+    {
+      "epoch": 0.03694581280788178,
+      "grad_norm": 5.427126932487668,
+      "learning_rate": 2.459016393442623e-06,
+      "loss": 1.5503,
+      "step": 15
+    },
+    {
+      "epoch": 0.03940886699507389,
+      "grad_norm": 7.015669043048,
+      "learning_rate": 2.6229508196721314e-06,
+      "loss": 1.5676,
+      "step": 16
+    },
+    {
+      "epoch": 0.04187192118226601,
+      "grad_norm": 6.548605897195016,
+      "learning_rate": 2.786885245901639e-06,
+      "loss": 1.4395,
+      "step": 17
+    },
+    {
+      "epoch": 0.04433497536945813,
+      "grad_norm": 7.1754245326754775,
+      "learning_rate": 2.9508196721311478e-06,
+      "loss": 1.4365,
+      "step": 18
+    },
+    {
+      "epoch": 0.046798029556650245,
+      "grad_norm": 3.578361669984036,
+      "learning_rate": 3.114754098360656e-06,
+      "loss": 1.3229,
+      "step": 19
+    },
+    {
+      "epoch": 0.04926108374384237,
+      "grad_norm": 3.569098802613537,
+      "learning_rate": 3.2786885245901638e-06,
+      "loss": 1.4392,
+      "step": 20
+    },
+    {
+      "epoch": 0.05172413793103448,
+      "grad_norm": 4.251402903866085,
+      "learning_rate": 3.4426229508196724e-06,
+      "loss": 1.4364,
+      "step": 21
+    },
+    {
+      "epoch": 0.054187192118226604,
+      "grad_norm": 2.923147322312773,
+      "learning_rate": 3.6065573770491806e-06,
+      "loss": 1.3143,
+      "step": 22
+    },
+    {
+      "epoch": 0.05665024630541872,
+      "grad_norm": 2.4265557126296033,
+      "learning_rate": 3.7704918032786884e-06,
+      "loss": 1.213,
+      "step": 23
+    },
+    {
+      "epoch": 0.059113300492610835,
+      "grad_norm": 2.732829589092375,
+      "learning_rate": 3.934426229508197e-06,
+      "loss": 1.3925,
+      "step": 24
+    },
+    {
+      "epoch": 0.06157635467980296,
+      "grad_norm": 1.9706056115152149,
+      "learning_rate": 4.098360655737705e-06,
+      "loss": 1.3419,
+      "step": 25
+    },
+    {
+      "epoch": 0.06403940886699508,
+      "grad_norm": 2.440362765282649,
+      "learning_rate": 4.2622950819672135e-06,
+      "loss": 1.3138,
+      "step": 26
+    },
+    {
+      "epoch": 0.0665024630541872,
+      "grad_norm": 2.774947236392866,
+      "learning_rate": 4.426229508196722e-06,
+      "loss": 1.2859,
+      "step": 27
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 1.7546870950917484,
+      "learning_rate": 4.59016393442623e-06,
+      "loss": 1.2653,
+      "step": 28
+    },
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 1.884315429049546,
+      "learning_rate": 4.754098360655738e-06,
+      "loss": 1.2559,
+      "step": 29
+    },
+    {
+      "epoch": 0.07389162561576355,
+      "grad_norm": 3.5506868423316535,
+      "learning_rate": 4.918032786885246e-06,
+      "loss": 1.2376,
+      "step": 30
+    },
+    {
+      "epoch": 0.07635467980295567,
+      "grad_norm": 2.7390642339131883,
+      "learning_rate": 5.0819672131147545e-06,
+      "loss": 1.2772,
+      "step": 31
+    },
+    {
+      "epoch": 0.07881773399014778,
+      "grad_norm": 1.9058135736884207,
+      "learning_rate": 5.245901639344263e-06,
+      "loss": 1.2745,
+      "step": 32
+    },
+    {
+      "epoch": 0.0812807881773399,
+      "grad_norm": 1.701642992762258,
+      "learning_rate": 5.409836065573772e-06,
+      "loss": 1.2432,
+      "step": 33
+    },
+    {
+      "epoch": 0.08374384236453201,
+      "grad_norm": 1.7090970944969652,
+      "learning_rate": 5.573770491803278e-06,
+      "loss": 1.1373,
+      "step": 34
+    },
+    {
+      "epoch": 0.08620689655172414,
+      "grad_norm": 1.6361001969458995,
+      "learning_rate": 5.737704918032787e-06,
+      "loss": 1.2228,
+      "step": 35
+    },
+    {
+      "epoch": 0.08866995073891626,
+      "grad_norm": 1.5739967057073647,
+      "learning_rate": 5.9016393442622956e-06,
+      "loss": 1.2146,
+      "step": 36
+    },
+    {
+      "epoch": 0.09113300492610837,
+      "grad_norm": 1.9219671708001531,
+      "learning_rate": 6.065573770491804e-06,
+      "loss": 1.2395,
+      "step": 37
+    },
+    {
+      "epoch": 0.09359605911330049,
+      "grad_norm": 1.5805901639635929,
+      "learning_rate": 6.229508196721312e-06,
+      "loss": 1.2511,
+      "step": 38
+    },
+    {
+      "epoch": 0.0960591133004926,
+      "grad_norm": 1.6114052863863946,
+      "learning_rate": 6.393442622950821e-06,
+      "loss": 1.1572,
+      "step": 39
+    },
+    {
+      "epoch": 0.09852216748768473,
+      "grad_norm": 1.3413556524297487,
+      "learning_rate": 6.5573770491803276e-06,
+      "loss": 1.1696,
+      "step": 40
+    },
+    {
+      "epoch": 0.10098522167487685,
+      "grad_norm": 1.573160047417312,
+      "learning_rate": 6.721311475409837e-06,
+      "loss": 1.2352,
+      "step": 41
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 1.7131097668621476,
+      "learning_rate": 6.885245901639345e-06,
+      "loss": 1.2543,
+      "step": 42
+    },
+    {
+      "epoch": 0.10591133004926108,
+      "grad_norm": 1.649219562452614,
+      "learning_rate": 7.049180327868853e-06,
+      "loss": 1.1574,
+      "step": 43
+    },
+    {
+      "epoch": 0.10837438423645321,
+      "grad_norm": 1.4440157459964449,
+      "learning_rate": 7.213114754098361e-06,
+      "loss": 1.1804,
+      "step": 44
+    },
+    {
+      "epoch": 0.11083743842364532,
+      "grad_norm": 1.637172226460499,
+      "learning_rate": 7.3770491803278695e-06,
+      "loss": 1.1763,
+      "step": 45
+    },
+    {
+      "epoch": 0.11330049261083744,
+      "grad_norm": 1.4639788299395275,
+      "learning_rate": 7.540983606557377e-06,
+      "loss": 1.1729,
+      "step": 46
+    },
+    {
+      "epoch": 0.11576354679802955,
+      "grad_norm": 1.426069674647583,
+      "learning_rate": 7.704918032786886e-06,
+      "loss": 1.2131,
+      "step": 47
+    },
+    {
+      "epoch": 0.11822660098522167,
+      "grad_norm": 1.4255315692661334,
+      "learning_rate": 7.868852459016394e-06,
+      "loss": 1.1882,
+      "step": 48
+    },
+    {
+      "epoch": 0.1206896551724138,
+      "grad_norm": 1.4132839188094533,
+      "learning_rate": 8.032786885245902e-06,
+      "loss": 1.3373,
+      "step": 49
+    },
+    {
+      "epoch": 0.12315270935960591,
+      "grad_norm": 1.4367298882920272,
+      "learning_rate": 8.19672131147541e-06,
+      "loss": 1.2286,
+      "step": 50
+    },
+    {
+      "epoch": 0.12561576354679804,
+      "grad_norm": 1.5382762004097146,
+      "learning_rate": 8.360655737704919e-06,
+      "loss": 1.2261,
+      "step": 51
+    },
+    {
+      "epoch": 0.12807881773399016,
+      "grad_norm": 1.453100963106598,
+      "learning_rate": 8.524590163934427e-06,
+      "loss": 1.0797,
+      "step": 52
+    },
+    {
+      "epoch": 0.13054187192118227,
+      "grad_norm": 1.325292686308896,
+      "learning_rate": 8.688524590163935e-06,
+      "loss": 1.1637,
+      "step": 53
+    },
+    {
+      "epoch": 0.1330049261083744,
+      "grad_norm": 3.0471451077188823,
+      "learning_rate": 8.852459016393443e-06,
+      "loss": 1.1376,
+      "step": 54
+    },
+    {
+      "epoch": 0.1354679802955665,
+      "grad_norm": 1.3350496520763624,
+      "learning_rate": 9.016393442622952e-06,
+      "loss": 1.2465,
+      "step": 55
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 1.2528620379790851,
+      "learning_rate": 9.18032786885246e-06,
+      "loss": 1.2961,
+      "step": 56
+    },
+    {
+      "epoch": 0.14039408866995073,
+      "grad_norm": 1.4539284894534732,
+      "learning_rate": 9.344262295081968e-06,
+      "loss": 1.2051,
+      "step": 57
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.3530535108609614,
+      "learning_rate": 9.508196721311476e-06,
+      "loss": 1.1876,
+      "step": 58
+    },
+    {
+      "epoch": 0.14532019704433496,
+      "grad_norm": 1.8022207600960927,
+      "learning_rate": 9.672131147540984e-06,
+      "loss": 1.2107,
+      "step": 59
+    },
+    {
+      "epoch": 0.1477832512315271,
+      "grad_norm": 1.5427089810504193,
+      "learning_rate": 9.836065573770493e-06,
+      "loss": 1.184,
+      "step": 60
+    },
+    {
+      "epoch": 0.15024630541871922,
+      "grad_norm": 1.3509881006277293,
+      "learning_rate": 1e-05,
+      "loss": 1.2428,
+      "step": 61
+    },
+    {
+      "epoch": 0.15270935960591134,
+      "grad_norm": 1.4535922047987058,
+      "learning_rate": 1.0163934426229509e-05,
+      "loss": 1.2097,
+      "step": 62
+    },
+    {
+      "epoch": 0.15517241379310345,
+      "grad_norm": 1.3286695373805288,
+      "learning_rate": 1.0327868852459017e-05,
+      "loss": 1.2187,
+      "step": 63
+    },
+    {
+      "epoch": 0.15763546798029557,
+      "grad_norm": 1.956202653305833,
+      "learning_rate": 1.0491803278688525e-05,
+      "loss": 1.3457,
+      "step": 64
+    },
+    {
+      "epoch": 0.16009852216748768,
+      "grad_norm": 1.7097271683751625,
+      "learning_rate": 1.0655737704918034e-05,
+      "loss": 1.1926,
+      "step": 65
+    },
+    {
+      "epoch": 0.1625615763546798,
+      "grad_norm": 1.918082364013381,
+      "learning_rate": 1.0819672131147544e-05,
+      "loss": 1.24,
+      "step": 66
+    },
+    {
+      "epoch": 0.16502463054187191,
+      "grad_norm": 1.4310867049835891,
+      "learning_rate": 1.0983606557377052e-05,
+      "loss": 1.166,
+      "step": 67
+    },
+    {
+      "epoch": 0.16748768472906403,
+      "grad_norm": 1.4449776751367742,
+      "learning_rate": 1.1147540983606557e-05,
+      "loss": 1.2541,
+      "step": 68
+    },
+    {
+      "epoch": 0.16995073891625614,
+      "grad_norm": 1.9928171397001568,
+      "learning_rate": 1.1311475409836066e-05,
+      "loss": 1.2344,
+      "step": 69
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 1.3422535392670518,
+      "learning_rate": 1.1475409836065575e-05,
+      "loss": 1.1888,
+      "step": 70
+    },
+    {
+      "epoch": 0.1748768472906404,
+      "grad_norm": 1.4523806562654191,
+      "learning_rate": 1.1639344262295083e-05,
+      "loss": 1.2256,
+      "step": 71
+    },
+    {
+      "epoch": 0.17733990147783252,
+      "grad_norm": 4.130022632453735,
+      "learning_rate": 1.1803278688524591e-05,
+      "loss": 1.1419,
+      "step": 72
+    },
+    {
+      "epoch": 0.17980295566502463,
+      "grad_norm": 1.3419677092565894,
+      "learning_rate": 1.19672131147541e-05,
+      "loss": 1.2626,
+      "step": 73
+    },
+    {
+      "epoch": 0.18226600985221675,
+      "grad_norm": 1.2920391970939145,
+      "learning_rate": 1.2131147540983608e-05,
+      "loss": 1.1537,
+      "step": 74
+    },
+    {
+      "epoch": 0.18472906403940886,
+      "grad_norm": 1.3690697927831479,
+      "learning_rate": 1.2295081967213116e-05,
+      "loss": 1.1883,
+      "step": 75
+    },
+    {
+      "epoch": 0.18719211822660098,
+      "grad_norm": 1.3455032620994691,
+      "learning_rate": 1.2459016393442624e-05,
+      "loss": 1.161,
+      "step": 76
+    },
+    {
+      "epoch": 0.1896551724137931,
+      "grad_norm": 1.3785307807986422,
+      "learning_rate": 1.2622950819672132e-05,
+      "loss": 1.2395,
+      "step": 77
+    },
+    {
+      "epoch": 0.1921182266009852,
+      "grad_norm": 1.6416828196678115,
+      "learning_rate": 1.2786885245901642e-05,
+      "loss": 1.2244,
+      "step": 78
+    },
+    {
+      "epoch": 0.19458128078817735,
+      "grad_norm": 1.3829015067383406,
+      "learning_rate": 1.295081967213115e-05,
+      "loss": 1.2362,
+      "step": 79
+    },
+    {
+      "epoch": 0.19704433497536947,
+      "grad_norm": 1.4435032423119754,
+      "learning_rate": 1.3114754098360655e-05,
+      "loss": 1.2797,
+      "step": 80
+    },
+    {
+      "epoch": 0.19950738916256158,
+      "grad_norm": 1.6051281198721252,
+      "learning_rate": 1.3278688524590165e-05,
+      "loss": 1.2826,
+      "step": 81
+    },
+    {
+      "epoch": 0.2019704433497537,
+      "grad_norm": 1.2928645158224772,
+      "learning_rate": 1.3442622950819673e-05,
+      "loss": 1.0685,
+      "step": 82
+    },
+    {
+      "epoch": 0.2044334975369458,
+      "grad_norm": 1.2607032302995402,
+      "learning_rate": 1.3606557377049181e-05,
+      "loss": 1.184,
+      "step": 83
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 1.3202314182643196,
+      "learning_rate": 1.377049180327869e-05,
+      "loss": 1.219,
+      "step": 84
+    },
+    {
+      "epoch": 0.20935960591133004,
+      "grad_norm": 1.4805215710398998,
+      "learning_rate": 1.3934426229508198e-05,
+      "loss": 1.2491,
+      "step": 85
+    },
+    {
+      "epoch": 0.21182266009852216,
+      "grad_norm": 1.448126450670855,
+      "learning_rate": 1.4098360655737706e-05,
+      "loss": 1.1326,
+      "step": 86
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 1.3592880484296994,
+      "learning_rate": 1.4262295081967214e-05,
+      "loss": 1.2039,
+      "step": 87
+    },
+    {
+      "epoch": 0.21674876847290642,
+      "grad_norm": 1.490428746897636,
+      "learning_rate": 1.4426229508196722e-05,
+      "loss": 1.1613,
+      "step": 88
+    },
+    {
+      "epoch": 0.21921182266009853,
+      "grad_norm": 1.3752578580264638,
+      "learning_rate": 1.459016393442623e-05,
+      "loss": 1.1717,
+      "step": 89
+    },
+    {
+      "epoch": 0.22167487684729065,
+      "grad_norm": 1.3787876326681114,
+      "learning_rate": 1.4754098360655739e-05,
+      "loss": 1.1852,
+      "step": 90
+    },
+    {
+      "epoch": 0.22413793103448276,
+      "grad_norm": 1.3033991865967125,
+      "learning_rate": 1.4918032786885249e-05,
+      "loss": 1.2958,
+      "step": 91
+    },
+    {
+      "epoch": 0.22660098522167488,
+      "grad_norm": 1.4410533111103268,
+      "learning_rate": 1.5081967213114754e-05,
+      "loss": 1.1426,
+      "step": 92
+    },
+    {
+      "epoch": 0.229064039408867,
+      "grad_norm": 1.267690597740285,
+      "learning_rate": 1.5245901639344264e-05,
+      "loss": 1.2456,
+      "step": 93
+    },
+    {
+      "epoch": 0.2315270935960591,
+      "grad_norm": 1.2732820474488957,
+      "learning_rate": 1.5409836065573772e-05,
+      "loss": 1.0927,
+      "step": 94
+    },
+    {
+      "epoch": 0.23399014778325122,
+      "grad_norm": 1.3825572149981824,
+      "learning_rate": 1.5573770491803278e-05,
+      "loss": 1.1705,
+      "step": 95
+    },
+    {
+      "epoch": 0.23645320197044334,
+      "grad_norm": 1.3911275919817174,
+      "learning_rate": 1.5737704918032788e-05,
+      "loss": 1.3077,
+      "step": 96
+    },
+    {
+      "epoch": 0.23891625615763548,
+      "grad_norm": 1.3800954472663725,
+      "learning_rate": 1.5901639344262295e-05,
+      "loss": 1.2748,
+      "step": 97
+    },
+    {
+      "epoch": 0.2413793103448276,
+      "grad_norm": 1.24170315494731,
+      "learning_rate": 1.6065573770491805e-05,
+      "loss": 1.2311,
+      "step": 98
+    },
+    {
+      "epoch": 0.2438423645320197,
+      "grad_norm": 1.224022878473549,
+      "learning_rate": 1.6229508196721314e-05,
+      "loss": 1.2036,
+      "step": 99
+    },
+    {
+      "epoch": 0.24630541871921183,
+      "grad_norm": 1.1439233408832539,
+      "learning_rate": 1.639344262295082e-05,
+      "loss": 1.193,
+      "step": 100
+    },
+    {
+      "epoch": 0.24876847290640394,
+      "grad_norm": 1.2947682653487784,
+      "learning_rate": 1.655737704918033e-05,
+      "loss": 1.2291,
+      "step": 101
+    },
+    {
+      "epoch": 0.2512315270935961,
+      "grad_norm": 1.3058225981987805,
+      "learning_rate": 1.6721311475409837e-05,
+      "loss": 1.1754,
+      "step": 102
+    },
+    {
+      "epoch": 0.2536945812807882,
+      "grad_norm": 1.302849394197415,
+      "learning_rate": 1.6885245901639347e-05,
+      "loss": 1.1601,
+      "step": 103
+    },
+    {
+      "epoch": 0.2561576354679803,
+      "grad_norm": 1.3741821110684322,
+      "learning_rate": 1.7049180327868854e-05,
+      "loss": 1.1909,
+      "step": 104
+    },
+    {
+      "epoch": 0.25862068965517243,
+      "grad_norm": 1.2930766633975732,
+      "learning_rate": 1.721311475409836e-05,
+      "loss": 1.1943,
+      "step": 105
+    },
+    {
+      "epoch": 0.26108374384236455,
+      "grad_norm": 1.1903446658458594,
+      "learning_rate": 1.737704918032787e-05,
+      "loss": 1.0855,
+      "step": 106
+    },
+    {
+      "epoch": 0.26354679802955666,
+      "grad_norm": 1.2250949803334914,
+      "learning_rate": 1.7540983606557377e-05,
+      "loss": 1.2148,
+      "step": 107
+    },
+    {
+      "epoch": 0.2660098522167488,
+      "grad_norm": 1.2682867905243251,
+      "learning_rate": 1.7704918032786887e-05,
+      "loss": 1.1985,
+      "step": 108
+    },
+    {
+      "epoch": 0.2684729064039409,
+      "grad_norm": 1.332130987734679,
+      "learning_rate": 1.7868852459016393e-05,
+      "loss": 1.2871,
+      "step": 109
+    },
+    {
+      "epoch": 0.270935960591133,
+      "grad_norm": 1.353173414954392,
+      "learning_rate": 1.8032786885245903e-05,
+      "loss": 1.2565,
+      "step": 110
+    },
+    {
+      "epoch": 0.2733990147783251,
+      "grad_norm": 1.2412860406443336,
+      "learning_rate": 1.8196721311475413e-05,
+      "loss": 1.1267,
+      "step": 111
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 1.306470828577769,
+      "learning_rate": 1.836065573770492e-05,
+      "loss": 1.2686,
+      "step": 112
+    },
+    {
+      "epoch": 0.27832512315270935,
+      "grad_norm": 1.314493617982923,
+      "learning_rate": 1.852459016393443e-05,
+      "loss": 1.2142,
+      "step": 113
+    },
+    {
+      "epoch": 0.28078817733990147,
+      "grad_norm": 1.2545002988873488,
+      "learning_rate": 1.8688524590163936e-05,
+      "loss": 1.1789,
+      "step": 114
+    },
+    {
+      "epoch": 0.2832512315270936,
+      "grad_norm": 1.4957826455258296,
+      "learning_rate": 1.8852459016393446e-05,
+      "loss": 1.1381,
+      "step": 115
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 1.2616977276541765,
+      "learning_rate": 1.9016393442622952e-05,
+      "loss": 1.1655,
+      "step": 116
+    },
+    {
+      "epoch": 0.2881773399014778,
+      "grad_norm": 1.4200150617955847,
+      "learning_rate": 1.918032786885246e-05,
+      "loss": 1.1579,
+      "step": 117
+    },
+    {
+      "epoch": 0.29064039408866993,
+      "grad_norm": 1.280841296999978,
+      "learning_rate": 1.934426229508197e-05,
+      "loss": 1.1745,
+      "step": 118
+    },
+    {
+      "epoch": 0.29310344827586204,
+      "grad_norm": 1.431370479393712,
+      "learning_rate": 1.9508196721311475e-05,
+      "loss": 1.0881,
+      "step": 119
+    },
+    {
+      "epoch": 0.2955665024630542,
+      "grad_norm": 1.3945095370442866,
+      "learning_rate": 1.9672131147540985e-05,
+      "loss": 1.2475,
+      "step": 120
+    },
+    {
+      "epoch": 0.29802955665024633,
+      "grad_norm": 1.2768548618993538,
+      "learning_rate": 1.9836065573770492e-05,
+      "loss": 1.208,
+      "step": 121
+    },
+    {
+      "epoch": 0.30049261083743845,
+      "grad_norm": 1.2835127226225345,
+      "learning_rate": 2e-05,
+      "loss": 1.1334,
+      "step": 122
+    },
+    {
+      "epoch": 0.30295566502463056,
+      "grad_norm": 1.346417817235094,
+      "learning_rate": 1.999995891830594e-05,
+      "loss": 1.1599,
+      "step": 123
+    },
+    {
+      "epoch": 0.3054187192118227,
+      "grad_norm": 1.269667726663668,
+      "learning_rate": 1.9999835673561284e-05,
+      "loss": 1.1884,
+      "step": 124
+    },
+    {
+      "epoch": 0.3078817733990148,
+      "grad_norm": 1.2540186656341483,
+      "learning_rate": 1.9999630266778667e-05,
+      "loss": 1.2117,
+      "step": 125
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 1.4143099910749737,
+      "learning_rate": 1.9999342699645774e-05,
+      "loss": 1.1938,
+      "step": 126
+    },
+    {
+      "epoch": 0.312807881773399,
+      "grad_norm": 1.1536311312987986,
+      "learning_rate": 1.9998972974525354e-05,
+      "loss": 1.1861,
+      "step": 127
+    },
+    {
+      "epoch": 0.31527093596059114,
+      "grad_norm": 1.302613671700403,
+      "learning_rate": 1.9998521094455198e-05,
+      "loss": 1.2304,
+      "step": 128
+    },
+    {
+      "epoch": 0.31773399014778325,
+      "grad_norm": 1.1964334455281025,
+      "learning_rate": 1.9997987063148097e-05,
+      "loss": 1.2261,
+      "step": 129
+    },
+    {
+      "epoch": 0.32019704433497537,
+      "grad_norm": 1.1822704307676777,
+      "learning_rate": 1.9997370884991842e-05,
+      "loss": 1.184,
+      "step": 130
+    },
+    {
+      "epoch": 0.3226600985221675,
+      "grad_norm": 1.22219459666233,
+      "learning_rate": 1.9996672565049158e-05,
+      "loss": 1.2066,
+      "step": 131
+    },
+    {
+      "epoch": 0.3251231527093596,
+      "grad_norm": 1.2286717932679192,
+      "learning_rate": 1.9995892109057675e-05,
+      "loss": 1.195,
+      "step": 132
+    },
+    {
+      "epoch": 0.3275862068965517,
+      "grad_norm": 1.3696833472218313,
+      "learning_rate": 1.9995029523429892e-05,
+      "loss": 1.2068,
+      "step": 133
+    },
+    {
+      "epoch": 0.33004926108374383,
+      "grad_norm": 1.2838616154927274,
+      "learning_rate": 1.99940848152531e-05,
+      "loss": 1.2182,
+      "step": 134
+    },
+    {
+      "epoch": 0.33251231527093594,
+      "grad_norm": 1.2042219437492114,
+      "learning_rate": 1.9993057992289336e-05,
+      "loss": 1.2719,
+      "step": 135
+    },
+    {
+      "epoch": 0.33497536945812806,
+      "grad_norm": 1.2738233578477811,
+      "learning_rate": 1.9991949062975336e-05,
+      "loss": 1.2291,
+      "step": 136
+    },
+    {
+      "epoch": 0.3374384236453202,
+      "grad_norm": 1.2525695616237198,
+      "learning_rate": 1.999075803642243e-05,
+      "loss": 1.1082,
+      "step": 137
+    },
+    {
+      "epoch": 0.3399014778325123,
+      "grad_norm": 1.2886546096500846,
+      "learning_rate": 1.9989484922416503e-05,
+      "loss": 1.2668,
+      "step": 138
+    },
+    {
+      "epoch": 0.34236453201970446,
+      "grad_norm": 1.2475813831853189,
+      "learning_rate": 1.998812973141789e-05,
+      "loss": 1.1992,
+      "step": 139
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 1.2435298839915614,
+      "learning_rate": 1.9986692474561292e-05,
+      "loss": 1.1272,
+      "step": 140
+    },
+    {
+      "epoch": 0.3472906403940887,
+      "grad_norm": 1.2161159416407954,
+      "learning_rate": 1.9985173163655706e-05,
+      "loss": 1.1604,
+      "step": 141
+    },
+    {
+      "epoch": 0.3497536945812808,
+      "grad_norm": 1.3209301107834823,
+      "learning_rate": 1.9983571811184297e-05,
+      "loss": 1.3596,
+      "step": 142
+    },
+    {
+      "epoch": 0.3522167487684729,
+      "grad_norm": 1.2025600444530125,
+      "learning_rate": 1.998188843030433e-05,
+      "loss": 1.1783,
+      "step": 143
+    },
+    {
+      "epoch": 0.35467980295566504,
+      "grad_norm": 1.264846939356213,
+      "learning_rate": 1.9980123034847025e-05,
+      "loss": 1.1358,
+      "step": 144
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 1.1130177503920067,
+      "learning_rate": 1.9978275639317476e-05,
+      "loss": 1.2869,
+      "step": 145
+    },
+    {
+      "epoch": 0.35960591133004927,
+      "grad_norm": 1.3231040741748257,
+      "learning_rate": 1.9976346258894502e-05,
+      "loss": 1.2137,
+      "step": 146
+    },
+    {
+      "epoch": 0.3620689655172414,
+      "grad_norm": 1.3228546138037165,
+      "learning_rate": 1.9974334909430553e-05,
+      "loss": 1.2011,
+      "step": 147
+    },
+    {
+      "epoch": 0.3645320197044335,
+      "grad_norm": 1.2112479027414027,
+      "learning_rate": 1.9972241607451552e-05,
+      "loss": 1.2163,
+      "step": 148
+    },
+    {
+      "epoch": 0.3669950738916256,
+      "grad_norm": 3.0314255496444424,
+      "learning_rate": 1.9970066370156783e-05,
+      "loss": 1.2251,
+      "step": 149
+    },
+    {
+      "epoch": 0.3694581280788177,
+      "grad_norm": 1.3138861829332404,
+      "learning_rate": 1.996780921541873e-05,
+      "loss": 1.1836,
+      "step": 150
+    },
+    {
+      "epoch": 0.37192118226600984,
+      "grad_norm": 1.263655885008596,
+      "learning_rate": 1.9965470161782942e-05,
+      "loss": 1.2689,
+      "step": 151
+    },
+    {
+      "epoch": 0.37438423645320196,
+      "grad_norm": 1.2698621113330835,
+      "learning_rate": 1.9963049228467875e-05,
+      "loss": 1.1665,
+      "step": 152
+    },
+    {
+      "epoch": 0.3768472906403941,
+      "grad_norm": 1.2201512424338394,
+      "learning_rate": 1.996054643536474e-05,
+      "loss": 1.2165,
+      "step": 153
+    },
+    {
+      "epoch": 0.3793103448275862,
+      "grad_norm": 1.1977096655194543,
+      "learning_rate": 1.9957961803037325e-05,
+      "loss": 1.2366,
+      "step": 154
+    },
+    {
+      "epoch": 0.3817733990147783,
+      "grad_norm": 1.2423714555528018,
+      "learning_rate": 1.9955295352721854e-05,
+      "loss": 1.1961,
+      "step": 155
+    },
+    {
+      "epoch": 0.3842364532019704,
+      "grad_norm": 1.3232608807922,
+      "learning_rate": 1.9952547106326787e-05,
+      "loss": 1.2157,
+      "step": 156
+    },
+    {
+      "epoch": 0.3866995073891626,
+      "grad_norm": 1.1545598332466134,
+      "learning_rate": 1.9949717086432637e-05,
+      "loss": 1.2075,
+      "step": 157
+    },
+    {
+      "epoch": 0.3891625615763547,
+      "grad_norm": 1.1485059030104625,
+      "learning_rate": 1.9946805316291817e-05,
+      "loss": 1.2725,
+      "step": 158
+    },
+    {
+      "epoch": 0.3916256157635468,
+      "grad_norm": 1.2516166246690759,
+      "learning_rate": 1.994381181982841e-05,
+      "loss": 1.1786,
+      "step": 159
+    },
+    {
+      "epoch": 0.39408866995073893,
+      "grad_norm": 1.2184886163016218,
+      "learning_rate": 1.9940736621638e-05,
+      "loss": 1.2017,
+      "step": 160
+    },
+    {
+      "epoch": 0.39655172413793105,
+      "grad_norm": 1.2419283614575665,
+      "learning_rate": 1.993757974698746e-05,
+      "loss": 1.1513,
+      "step": 161
+    },
+    {
+      "epoch": 0.39901477832512317,
+      "grad_norm": 1.1858119511145377,
+      "learning_rate": 1.993434122181474e-05,
+      "loss": 1.1983,
+      "step": 162
+    },
+    {
+      "epoch": 0.4014778325123153,
+      "grad_norm": 1.3195984393597104,
+      "learning_rate": 1.9931021072728658e-05,
+      "loss": 1.1303,
+      "step": 163
+    },
+    {
+      "epoch": 0.4039408866995074,
+      "grad_norm": 1.219853317479262,
+      "learning_rate": 1.992761932700868e-05,
+      "loss": 1.1482,
+      "step": 164
+    },
+    {
+      "epoch": 0.4064039408866995,
+      "grad_norm": 1.267830422424957,
+      "learning_rate": 1.9924136012604714e-05,
+      "loss": 1.141,
+      "step": 165
+    },
+    {
+      "epoch": 0.4088669950738916,
+      "grad_norm": 1.216480881515268,
+      "learning_rate": 1.9920571158136837e-05,
+      "loss": 1.1279,
+      "step": 166
+    },
+    {
+      "epoch": 0.41133004926108374,
+      "grad_norm": 1.2245362182376216,
+      "learning_rate": 1.9916924792895112e-05,
+      "loss": 1.2717,
+      "step": 167
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 1.2595827428269755,
+      "learning_rate": 1.9913196946839304e-05,
+      "loss": 1.2269,
+      "step": 168
+    },
+    {
+      "epoch": 0.41625615763546797,
+      "grad_norm": 1.3641245509975088,
+      "learning_rate": 1.9909387650598665e-05,
+      "loss": 1.1724,
+      "step": 169
+    },
+    {
+      "epoch": 0.4187192118226601,
+      "grad_norm": 1.260501522434016,
+      "learning_rate": 1.990549693547166e-05,
+      "loss": 1.2654,
+      "step": 170
+    },
+    {
+      "epoch": 0.4211822660098522,
+      "grad_norm": 1.2619119502151588,
+      "learning_rate": 1.9901524833425724e-05,
+      "loss": 1.1777,
+      "step": 171
+    },
+    {
+      "epoch": 0.4236453201970443,
+      "grad_norm": 1.3434378239164741,
+      "learning_rate": 1.9897471377096992e-05,
+      "loss": 1.2009,
+      "step": 172
+    },
+    {
+      "epoch": 0.42610837438423643,
+      "grad_norm": 1.2380055984160956,
+      "learning_rate": 1.9893336599790034e-05,
+      "loss": 1.2525,
+      "step": 173
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 1.2072881530148862,
+      "learning_rate": 1.9889120535477584e-05,
+      "loss": 1.2491,
+      "step": 174
+    },
+    {
+      "epoch": 0.43103448275862066,
+      "grad_norm": 1.0766964722475039,
+      "learning_rate": 1.9884823218800255e-05,
+      "loss": 1.0366,
+      "step": 175
+    },
+    {
+      "epoch": 0.43349753694581283,
+      "grad_norm": 1.3160107751509542,
+      "learning_rate": 1.9880444685066252e-05,
+      "loss": 1.209,
+      "step": 176
+    },
+    {
+      "epoch": 0.43596059113300495,
+      "grad_norm": 1.3430756495849807,
+      "learning_rate": 1.9875984970251095e-05,
+      "loss": 1.1439,
+      "step": 177
+    },
+    {
+      "epoch": 0.43842364532019706,
+      "grad_norm": 1.231198194546705,
+      "learning_rate": 1.987144411099731e-05,
+      "loss": 1.1447,
+      "step": 178
+    },
+    {
+      "epoch": 0.4408866995073892,
+      "grad_norm": 1.207002262890835,
+      "learning_rate": 1.9866822144614143e-05,
+      "loss": 1.2291,
+      "step": 179
+    },
+    {
+      "epoch": 0.4433497536945813,
+      "grad_norm": 1.3095521975408642,
+      "learning_rate": 1.9862119109077226e-05,
+      "loss": 1.2764,
+      "step": 180
+    },
+    {
+      "epoch": 0.4458128078817734,
+      "grad_norm": 1.201524388139198,
+      "learning_rate": 1.9857335043028297e-05,
+      "loss": 1.1659,
+      "step": 181
+    },
+    {
+      "epoch": 0.4482758620689655,
+      "grad_norm": 1.1608566584620457,
+      "learning_rate": 1.985246998577486e-05,
+      "loss": 1.2003,
+      "step": 182
+    },
+    {
+      "epoch": 0.45073891625615764,
+      "grad_norm": 1.3091957188051122,
+      "learning_rate": 1.984752397728988e-05,
+      "loss": 1.1489,
+      "step": 183
+    },
+    {
+      "epoch": 0.45320197044334976,
+      "grad_norm": 1.1032447685084816,
+      "learning_rate": 1.984249705821143e-05,
+      "loss": 1.2502,
+      "step": 184
+    },
+    {
+      "epoch": 0.45566502463054187,
+      "grad_norm": 1.2821867240068512,
+      "learning_rate": 1.983738926984239e-05,
+      "loss": 1.2042,
+      "step": 185
+    },
+    {
+      "epoch": 0.458128078817734,
+      "grad_norm": 1.2338430972855412,
+      "learning_rate": 1.9832200654150077e-05,
+      "loss": 1.2205,
+      "step": 186
+    },
+    {
+      "epoch": 0.4605911330049261,
+      "grad_norm": 1.3306732862750001,
+      "learning_rate": 1.9826931253765907e-05,
+      "loss": 1.2431,
+      "step": 187
+    },
+    {
+      "epoch": 0.4630541871921182,
+      "grad_norm": 1.2859704589328107,
+      "learning_rate": 1.9821581111985072e-05,
+      "loss": 1.1479,
+      "step": 188
+    },
+    {
+      "epoch": 0.46551724137931033,
+      "grad_norm": 1.1233030342200305,
+      "learning_rate": 1.9816150272766136e-05,
+      "loss": 1.2477,
+      "step": 189
+    },
+    {
+      "epoch": 0.46798029556650245,
+      "grad_norm": 1.2485067030879888,
+      "learning_rate": 1.981063878073073e-05,
+      "loss": 1.1717,
+      "step": 190
+    },
+    {
+      "epoch": 0.47044334975369456,
+      "grad_norm": 1.2188124029371652,
+      "learning_rate": 1.9805046681163124e-05,
+      "loss": 1.2328,
+      "step": 191
+    },
+    {
+      "epoch": 0.4729064039408867,
+      "grad_norm": 1.104067774008823,
+      "learning_rate": 1.979937402000991e-05,
+      "loss": 1.1088,
+      "step": 192
+    },
+    {
+      "epoch": 0.4753694581280788,
+      "grad_norm": 1.2062591255910127,
+      "learning_rate": 1.9793620843879594e-05,
+      "loss": 1.1787,
+      "step": 193
+    },
+    {
+      "epoch": 0.47783251231527096,
+      "grad_norm": 1.1520948755515033,
+      "learning_rate": 1.9787787200042224e-05,
+      "loss": 1.211,
+      "step": 194
+    },
+    {
+      "epoch": 0.4802955665024631,
+      "grad_norm": 1.1382559643242922,
+      "learning_rate": 1.9781873136428985e-05,
+      "loss": 1.2036,
+      "step": 195
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 1.1719359827068931,
+      "learning_rate": 1.977587870163184e-05,
+      "loss": 1.2282,
+      "step": 196
+    },
+    {
+      "epoch": 0.4852216748768473,
+      "grad_norm": 1.173908706920403,
+      "learning_rate": 1.9769803944903084e-05,
+      "loss": 1.1701,
+      "step": 197
+    },
+    {
+      "epoch": 0.4876847290640394,
+      "grad_norm": 1.2974583795887777,
+      "learning_rate": 1.9763648916154982e-05,
+      "loss": 1.1916,
+      "step": 198
+    },
+    {
+      "epoch": 0.49014778325123154,
+      "grad_norm": 1.1486984786807202,
+      "learning_rate": 1.9757413665959337e-05,
+      "loss": 1.175,
+      "step": 199
+    },
+    {
+      "epoch": 0.49261083743842365,
+      "grad_norm": 1.221271596149515,
+      "learning_rate": 1.975109824554707e-05,
+      "loss": 1.1911,
+      "step": 200
+    },
+    {
+      "epoch": 0.49507389162561577,
+      "grad_norm": 1.1965417960869649,
+      "learning_rate": 1.9744702706807825e-05,
+      "loss": 1.1619,
+      "step": 201
+    },
+    {
+      "epoch": 0.4975369458128079,
+      "grad_norm": 1.2243203243336844,
+      "learning_rate": 1.973822710228951e-05,
+      "loss": 1.2238,
+      "step": 202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2957754414537586,
+      "learning_rate": 1.9731671485197884e-05,
+      "loss": 1.2679,
+      "step": 203
+    },
+    {
+      "epoch": 0.5024630541871922,
+      "grad_norm": 1.1325395452571845,
+      "learning_rate": 1.972503590939612e-05,
+      "loss": 1.2317,
+      "step": 204
+    },
+    {
+      "epoch": 0.5049261083743842,
+      "grad_norm": 1.2147615165721934,
+      "learning_rate": 1.971832042940436e-05,
+      "loss": 1.1789,
+      "step": 205
+    },
+    {
+      "epoch": 0.5073891625615764,
+      "grad_norm": 1.225349750025557,
+      "learning_rate": 1.971152510039926e-05,
+      "loss": 1.2282,
+      "step": 206
+    },
+    {
+      "epoch": 0.5098522167487685,
+      "grad_norm": 1.174075438585791,
+      "learning_rate": 1.9704649978213545e-05,
+      "loss": 1.1857,
+      "step": 207
+    },
+    {
+      "epoch": 0.5123152709359606,
+      "grad_norm": 1.2208947603484308,
+      "learning_rate": 1.9697695119335547e-05,
+      "loss": 1.0999,
+      "step": 208
+    },
+    {
+      "epoch": 0.5147783251231527,
+      "grad_norm": 1.2316689586919194,
+      "learning_rate": 1.9690660580908746e-05,
+      "loss": 1.2547,
+      "step": 209
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 1.099172990032186,
+      "learning_rate": 1.9683546420731292e-05,
+      "loss": 1.2453,
+      "step": 210
+    },
+    {
+      "epoch": 0.5197044334975369,
+      "grad_norm": 1.2088679194404248,
+      "learning_rate": 1.9676352697255535e-05,
+      "loss": 1.2397,
+      "step": 211
+    },
+    {
+      "epoch": 0.5221674876847291,
+      "grad_norm": 1.160985220214582,
+      "learning_rate": 1.9669079469587548e-05,
+      "loss": 1.1987,
+      "step": 212
+    },
+    {
+      "epoch": 0.5246305418719212,
+      "grad_norm": 1.120082172887501,
+      "learning_rate": 1.9661726797486625e-05,
+      "loss": 1.2121,
+      "step": 213
+    },
+    {
+      "epoch": 0.5270935960591133,
+      "grad_norm": 1.115565048872266,
+      "learning_rate": 1.965429474136482e-05,
+      "loss": 1.2217,
+      "step": 214
+    },
+    {
+      "epoch": 0.5295566502463054,
+      "grad_norm": 1.250204164997804,
+      "learning_rate": 1.964678336228642e-05,
+      "loss": 1.2421,
+      "step": 215
+    },
+    {
+      "epoch": 0.5320197044334976,
+      "grad_norm": 1.150896566690045,
+      "learning_rate": 1.963919272196746e-05,
+      "loss": 1.2221,
+      "step": 216
+    },
+    {
+      "epoch": 0.5344827586206896,
+      "grad_norm": 1.0954406750353816,
+      "learning_rate": 1.9631522882775217e-05,
+      "loss": 1.2473,
+      "step": 217
+    },
+    {
+      "epoch": 0.5369458128078818,
+      "grad_norm": 1.1351476869831179,
+      "learning_rate": 1.9623773907727682e-05,
+      "loss": 1.2446,
+      "step": 218
+    },
+    {
+      "epoch": 0.5394088669950738,
+      "grad_norm": 1.21936352277132,
+      "learning_rate": 1.9615945860493063e-05,
+      "loss": 1.224,
+      "step": 219
+    },
+    {
+      "epoch": 0.541871921182266,
+      "grad_norm": 1.3092839029528676,
+      "learning_rate": 1.9608038805389253e-05,
+      "loss": 1.2218,
+      "step": 220
+    },
+    {
+      "epoch": 0.5443349753694581,
+      "grad_norm": 1.225745056279308,
+      "learning_rate": 1.9600052807383285e-05,
+      "loss": 1.2438,
+      "step": 221
+    },
+    {
+      "epoch": 0.5467980295566502,
+      "grad_norm": 1.2504511495884272,
+      "learning_rate": 1.9591987932090836e-05,
+      "loss": 1.2647,
+      "step": 222
+    },
+    {
+      "epoch": 0.5492610837438424,
+      "grad_norm": 1.1593114686406394,
+      "learning_rate": 1.9583844245775647e-05,
+      "loss": 1.1673,
+      "step": 223
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 1.1284204513542304,
+      "learning_rate": 1.9575621815349e-05,
+      "loss": 1.1771,
+      "step": 224
+    },
+    {
+      "epoch": 0.5541871921182266,
+      "grad_norm": 1.1779061820003602,
+      "learning_rate": 1.9567320708369178e-05,
+      "loss": 1.2122,
+      "step": 225
+    },
+    {
+      "epoch": 0.5566502463054187,
+      "grad_norm": 1.1254709635496583,
+      "learning_rate": 1.9558940993040885e-05,
+      "loss": 1.0954,
+      "step": 226
+    },
+    {
+      "epoch": 0.5591133004926109,
+      "grad_norm": 1.1637900628918603,
+      "learning_rate": 1.95504827382147e-05,
+      "loss": 1.2657,
+      "step": 227
+    },
+    {
+      "epoch": 0.5615763546798029,
+      "grad_norm": 1.1915952438789714,
+      "learning_rate": 1.954194601338651e-05,
+      "loss": 1.2246,
+      "step": 228
+    },
+    {
+      "epoch": 0.5640394088669951,
+      "grad_norm": 1.149186856509955,
+      "learning_rate": 1.9533330888696943e-05,
+      "loss": 1.1612,
+      "step": 229
+    },
+    {
+      "epoch": 0.5665024630541872,
+      "grad_norm": 1.2535552487243269,
+      "learning_rate": 1.952463743493078e-05,
+      "loss": 1.1355,
+      "step": 230
+    },
+    {
+      "epoch": 0.5689655172413793,
+      "grad_norm": 1.1630334081566465,
+      "learning_rate": 1.9515865723516375e-05,
+      "loss": 1.1364,
+      "step": 231
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 1.243998377167235,
+      "learning_rate": 1.9507015826525096e-05,
+      "loss": 1.2279,
+      "step": 232
+    },
+    {
+      "epoch": 0.5738916256157636,
+      "grad_norm": 1.1689830445690776,
+      "learning_rate": 1.9498087816670685e-05,
+      "loss": 1.22,
+      "step": 233
+    },
+    {
+      "epoch": 0.5763546798029556,
+      "grad_norm": 1.286419187045414,
+      "learning_rate": 1.9489081767308696e-05,
+      "loss": 1.2364,
+      "step": 234
+    },
+    {
+      "epoch": 0.5788177339901478,
+      "grad_norm": 1.141561567847589,
+      "learning_rate": 1.9479997752435886e-05,
+      "loss": 1.2269,
+      "step": 235
+    },
+    {
+      "epoch": 0.5812807881773399,
+      "grad_norm": 1.110732255693052,
+      "learning_rate": 1.9470835846689596e-05,
+      "loss": 1.2456,
+      "step": 236
+    },
+    {
+      "epoch": 0.583743842364532,
+      "grad_norm": 1.1763550128335267,
+      "learning_rate": 1.946159612534715e-05,
+      "loss": 1.3204,
+      "step": 237
+    },
+    {
+      "epoch": 0.5862068965517241,
+      "grad_norm": 1.0953171538119415,
+      "learning_rate": 1.9452278664325227e-05,
+      "loss": 1.1944,
+      "step": 238
+    },
+    {
+      "epoch": 0.5886699507389163,
+      "grad_norm": 1.1128666155800393,
+      "learning_rate": 1.9442883540179243e-05,
+      "loss": 1.1549,
+      "step": 239
+    },
+    {
+      "epoch": 0.5911330049261084,
+      "grad_norm": 1.1402680347959282,
+      "learning_rate": 1.9433410830102724e-05,
+      "loss": 1.1204,
+      "step": 240
+    },
+    {
+      "epoch": 0.5935960591133005,
+      "grad_norm": 1.1584732744133148,
+      "learning_rate": 1.9423860611926667e-05,
+      "loss": 1.2345,
+      "step": 241
+    },
+    {
+      "epoch": 0.5960591133004927,
+      "grad_norm": 1.1967676816789183,
+      "learning_rate": 1.9414232964118893e-05,
+      "loss": 1.1648,
+      "step": 242
+    },
+    {
+      "epoch": 0.5985221674876847,
+      "grad_norm": 1.148420632167128,
+      "learning_rate": 1.9404527965783423e-05,
+      "loss": 1.2316,
+      "step": 243
+    },
+    {
+      "epoch": 0.6009852216748769,
+      "grad_norm": 1.0993871108489186,
+      "learning_rate": 1.939474569665981e-05,
+      "loss": 1.2144,
+      "step": 244
+    },
+    {
+      "epoch": 0.603448275862069,
+      "grad_norm": 1.1538934147765967,
+      "learning_rate": 1.9384886237122496e-05,
+      "loss": 1.2159,
+      "step": 245
+    },
+    {
+      "epoch": 0.6059113300492611,
+      "grad_norm": 1.1881182717459382,
+      "learning_rate": 1.937494966818014e-05,
+      "loss": 1.1254,
+      "step": 246
+    },
+    {
+      "epoch": 0.6083743842364532,
+      "grad_norm": 1.1614146488111312,
+      "learning_rate": 1.936493607147495e-05,
+      "loss": 1.1451,
+      "step": 247
+    },
+    {
+      "epoch": 0.6108374384236454,
+      "grad_norm": 1.225877996046424,
+      "learning_rate": 1.9354845529282042e-05,
+      "loss": 1.1092,
+      "step": 248
+    },
+    {
+      "epoch": 0.6133004926108374,
+      "grad_norm": 1.1601458205451087,
+      "learning_rate": 1.9344678124508718e-05,
+      "loss": 1.1736,
+      "step": 249
+    },
+    {
+      "epoch": 0.6157635467980296,
+      "grad_norm": 1.2271749826001248,
+      "learning_rate": 1.933443394069383e-05,
+      "loss": 1.1853,
+      "step": 250
+    },
+    {
+      "epoch": 0.6182266009852216,
+      "grad_norm": 1.1936727009457124,
+      "learning_rate": 1.9324113062007056e-05,
+      "loss": 1.208,
+      "step": 251
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 1.1054034533946946,
+      "learning_rate": 1.9313715573248238e-05,
+      "loss": 1.1716,
+      "step": 252
+    },
+    {
+      "epoch": 0.6231527093596059,
+      "grad_norm": 1.123817935455876,
+      "learning_rate": 1.9303241559846664e-05,
+      "loss": 1.1773,
+      "step": 253
+    },
+    {
+      "epoch": 0.625615763546798,
+      "grad_norm": 1.1409313887780133,
+      "learning_rate": 1.9292691107860374e-05,
+      "loss": 1.2331,
+      "step": 254
+    },
+    {
+      "epoch": 0.6280788177339901,
+      "grad_norm": 1.2830663692347828,
+      "learning_rate": 1.928206430397546e-05,
+      "loss": 1.122,
+      "step": 255
+    },
+    {
+      "epoch": 0.6305418719211823,
+      "grad_norm": 1.2456312128305413,
+      "learning_rate": 1.927136123550534e-05,
+      "loss": 1.1737,
+      "step": 256
+    },
+    {
+      "epoch": 0.6330049261083743,
+      "grad_norm": 1.1375738643935585,
+      "learning_rate": 1.9260581990390056e-05,
+      "loss": 1.2649,
+      "step": 257
+    },
+    {
+      "epoch": 0.6354679802955665,
+      "grad_norm": 1.2435773355834052,
+      "learning_rate": 1.9249726657195534e-05,
+      "loss": 1.1599,
+      "step": 258
+    },
+    {
+      "epoch": 0.6379310344827587,
+      "grad_norm": 1.1393125638945556,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 1.093,
+      "step": 259
+    },
+    {
+      "epoch": 0.6403940886699507,
+      "grad_norm": 1.1721175896368747,
+      "learning_rate": 1.922778808395759e-05,
+      "loss": 1.1583,
+      "step": 260
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 1.110374859755865,
+      "learning_rate": 1.921670502416892e-05,
+      "loss": 1.1272,
+      "step": 261
+    },
+    {
+      "epoch": 0.645320197044335,
+      "grad_norm": 1.1938044690392797,
+      "learning_rate": 1.9205546236809037e-05,
+      "loss": 1.2681,
+      "step": 262
+    },
+    {
+      "epoch": 0.6477832512315271,
+      "grad_norm": 1.172075534192954,
+      "learning_rate": 1.919431181356231e-05,
+      "loss": 1.1932,
+      "step": 263
+    },
+    {
+      "epoch": 0.6502463054187192,
+      "grad_norm": 1.1848898863031847,
+      "learning_rate": 1.9183001846734573e-05,
+      "loss": 1.1863,
+      "step": 264
+    },
+    {
+      "epoch": 0.6527093596059114,
+      "grad_norm": 1.2191290632712344,
+      "learning_rate": 1.9171616429252345e-05,
+      "loss": 1.2369,
+      "step": 265
+    },
+    {
+      "epoch": 0.6551724137931034,
+      "grad_norm": 1.1380638213371261,
+      "learning_rate": 1.9160155654662075e-05,
+      "loss": 1.1865,
+      "step": 266
+    },
+    {
+      "epoch": 0.6576354679802956,
+      "grad_norm": 1.1272497464414688,
+      "learning_rate": 1.9148619617129364e-05,
+      "loss": 1.1075,
+      "step": 267
+    },
+    {
+      "epoch": 0.6600985221674877,
+      "grad_norm": 1.0977343874233538,
+      "learning_rate": 1.9137008411438213e-05,
+      "loss": 1.1057,
+      "step": 268
+    },
+    {
+      "epoch": 0.6625615763546798,
+      "grad_norm": 1.2422871579673707,
+      "learning_rate": 1.9125322132990215e-05,
+      "loss": 1.1109,
+      "step": 269
+    },
+    {
+      "epoch": 0.6650246305418719,
+      "grad_norm": 1.1461084844660117,
+      "learning_rate": 1.9113560877803798e-05,
+      "loss": 1.1637,
+      "step": 270
+    },
+    {
+      "epoch": 0.6674876847290641,
+      "grad_norm": 1.1539868553279071,
+      "learning_rate": 1.910172474251341e-05,
+      "loss": 1.1443,
+      "step": 271
+    },
+    {
+      "epoch": 0.6699507389162561,
+      "grad_norm": 1.3547337982819572,
+      "learning_rate": 1.9089813824368765e-05,
+      "loss": 1.1825,
+      "step": 272
+    },
+    {
+      "epoch": 0.6724137931034483,
+      "grad_norm": 1.0894241345563804,
+      "learning_rate": 1.907782822123399e-05,
+      "loss": 1.1977,
+      "step": 273
+    },
+    {
+      "epoch": 0.6748768472906403,
+      "grad_norm": 2.522327757778015,
+      "learning_rate": 1.9065768031586864e-05,
+      "loss": 1.1979,
+      "step": 274
+    },
+    {
+      "epoch": 0.6773399014778325,
+      "grad_norm": 1.2145690116670471,
+      "learning_rate": 1.905363335451799e-05,
+      "loss": 1.1391,
+      "step": 275
+    },
+    {
+      "epoch": 0.6798029556650246,
+      "grad_norm": 1.1303357112611645,
+      "learning_rate": 1.9041424289729994e-05,
+      "loss": 1.2104,
+      "step": 276
+    },
+    {
+      "epoch": 0.6822660098522167,
+      "grad_norm": 1.0164044250424984,
+      "learning_rate": 1.9029140937536676e-05,
+      "loss": 1.1788,
+      "step": 277
+    },
+    {
+      "epoch": 0.6847290640394089,
+      "grad_norm": 1.1631362609482145,
+      "learning_rate": 1.901678339886223e-05,
+      "loss": 1.2575,
+      "step": 278
+    },
+    {
+      "epoch": 0.687192118226601,
+      "grad_norm": 1.2778336171835305,
+      "learning_rate": 1.9004351775240376e-05,
+      "loss": 1.1017,
+      "step": 279
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 1.1076836924059956,
+      "learning_rate": 1.8991846168813547e-05,
+      "loss": 1.2713,
+      "step": 280
+    },
+    {
+      "epoch": 0.6921182266009852,
+      "grad_norm": 1.1314655583323705,
+      "learning_rate": 1.897926668233204e-05,
+      "loss": 1.1888,
+      "step": 281
+    },
+    {
+      "epoch": 0.6945812807881774,
+      "grad_norm": 1.1855719624964784,
+      "learning_rate": 1.896661341915318e-05,
+      "loss": 1.2381,
+      "step": 282
+    },
+    {
+      "epoch": 0.6970443349753694,
+      "grad_norm": 1.2117978393007827,
+      "learning_rate": 1.8953886483240465e-05,
+      "loss": 1.1747,
+      "step": 283
+    },
+    {
+      "epoch": 0.6995073891625616,
+      "grad_norm": 1.22190887716409,
+      "learning_rate": 1.8941085979162714e-05,
+      "loss": 1.2231,
+      "step": 284
+    },
+    {
+      "epoch": 0.7019704433497537,
+      "grad_norm": 1.1468822822027778,
+      "learning_rate": 1.8928212012093204e-05,
+      "loss": 1.1037,
+      "step": 285
+    },
+    {
+      "epoch": 0.7044334975369458,
+      "grad_norm": 1.0379019544598291,
+      "learning_rate": 1.891526468780881e-05,
+      "loss": 1.2211,
+      "step": 286
+    },
+    {
+      "epoch": 0.7068965517241379,
+      "grad_norm": 1.1470946681059266,
+      "learning_rate": 1.8902244112689128e-05,
+      "loss": 1.2476,
+      "step": 287
+    },
+    {
+      "epoch": 0.7093596059113301,
+      "grad_norm": 1.0986922726965034,
+      "learning_rate": 1.8889150393715627e-05,
+      "loss": 1.2508,
+      "step": 288
+    },
+    {
+      "epoch": 0.7118226600985221,
+      "grad_norm": 1.1723376060384838,
+      "learning_rate": 1.8875983638470732e-05,
+      "loss": 1.108,
+      "step": 289
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 1.2597061971205685,
+      "learning_rate": 1.8862743955136966e-05,
+      "loss": 1.2004,
+      "step": 290
+    },
+    {
+      "epoch": 0.7167487684729064,
+      "grad_norm": 1.0920190465684365,
+      "learning_rate": 1.8849431452496053e-05,
+      "loss": 1.0691,
+      "step": 291
+    },
+    {
+      "epoch": 0.7192118226600985,
+      "grad_norm": 1.234494843278685,
+      "learning_rate": 1.8836046239928025e-05,
+      "loss": 1.1709,
+      "step": 292
+    },
+    {
+      "epoch": 0.7216748768472906,
+      "grad_norm": 1.146006391923531,
+      "learning_rate": 1.8822588427410324e-05,
+      "loss": 1.2534,
+      "step": 293
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 1.190538234002207,
+      "learning_rate": 1.8809058125516894e-05,
+      "loss": 1.1848,
+      "step": 294
+    },
+    {
+      "epoch": 0.7266009852216748,
+      "grad_norm": 1.208970619825956,
+      "learning_rate": 1.8795455445417286e-05,
+      "loss": 1.2027,
+      "step": 295
+    },
+    {
+      "epoch": 0.729064039408867,
+      "grad_norm": 1.1374832152081942,
+      "learning_rate": 1.8781780498875727e-05,
+      "loss": 1.1937,
+      "step": 296
+    },
+    {
+      "epoch": 0.7315270935960592,
+      "grad_norm": 1.3364141405885752,
+      "learning_rate": 1.8768033398250203e-05,
+      "loss": 1.2581,
+      "step": 297
+    },
+    {
+      "epoch": 0.7339901477832512,
+      "grad_norm": 1.16606193718879,
+      "learning_rate": 1.8754214256491564e-05,
+      "loss": 1.2148,
+      "step": 298
+    },
+    {
+      "epoch": 0.7364532019704434,
+      "grad_norm": 1.1111001047278033,
+      "learning_rate": 1.874032318714255e-05,
+      "loss": 1.2825,
+      "step": 299
+    },
+    {
+      "epoch": 0.7389162561576355,
+      "grad_norm": 1.1775459910682609,
+      "learning_rate": 1.8726360304336896e-05,
+      "loss": 1.2107,
+      "step": 300
+    },
+    {
+      "epoch": 0.7413793103448276,
+      "grad_norm": 1.1126730344985154,
+      "learning_rate": 1.8712325722798376e-05,
+      "loss": 1.2195,
+      "step": 301
+    },
+    {
+      "epoch": 0.7438423645320197,
+      "grad_norm": 1.071239376393792,
+      "learning_rate": 1.8698219557839875e-05,
+      "loss": 1.1619,
+      "step": 302
+    },
+    {
+      "epoch": 0.7463054187192119,
+      "grad_norm": 1.213727432025452,
+      "learning_rate": 1.8684041925362412e-05,
+      "loss": 1.13,
+      "step": 303
+    },
+    {
+      "epoch": 0.7487684729064039,
+      "grad_norm": 1.0809178873153813,
+      "learning_rate": 1.866979294185423e-05,
+      "loss": 1.1999,
+      "step": 304
+    },
+    {
+      "epoch": 0.7512315270935961,
+      "grad_norm": 1.1723696872547285,
+      "learning_rate": 1.8655472724389798e-05,
+      "loss": 1.1627,
+      "step": 305
+    },
+    {
+      "epoch": 0.7536945812807881,
+      "grad_norm": 1.1130417951086524,
+      "learning_rate": 1.864108139062888e-05,
+      "loss": 1.2837,
+      "step": 306
+    },
+    {
+      "epoch": 0.7561576354679803,
+      "grad_norm": 0.989072189163068,
+      "learning_rate": 1.8626619058815546e-05,
+      "loss": 1.1429,
+      "step": 307
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 1.1073822606450792,
+      "learning_rate": 1.8612085847777215e-05,
+      "loss": 1.1501,
+      "step": 308
+    },
+    {
+      "epoch": 0.7610837438423645,
+      "grad_norm": 1.171468689734966,
+      "learning_rate": 1.859748187692367e-05,
+      "loss": 1.1497,
+      "step": 309
+    },
+    {
+      "epoch": 0.7635467980295566,
+      "grad_norm": 1.134123155785567,
+      "learning_rate": 1.858280726624609e-05,
+      "loss": 1.2505,
+      "step": 310
+    },
+    {
+      "epoch": 0.7660098522167488,
+      "grad_norm": 1.1568070693554542,
+      "learning_rate": 1.8568062136316047e-05,
+      "loss": 1.173,
+      "step": 311
+    },
+    {
+      "epoch": 0.7684729064039408,
+      "grad_norm": 1.1786164733403517,
+      "learning_rate": 1.855324660828452e-05,
+      "loss": 1.3101,
+      "step": 312
+    },
+    {
+      "epoch": 0.770935960591133,
+      "grad_norm": 1.1516980964817813,
+      "learning_rate": 1.853836080388091e-05,
+      "loss": 1.1491,
+      "step": 313
+    },
+    {
+      "epoch": 0.7733990147783252,
+      "grad_norm": 1.1607105202666188,
+      "learning_rate": 1.8523404845412028e-05,
+      "loss": 1.1573,
+      "step": 314
+    },
+    {
+      "epoch": 0.7758620689655172,
+      "grad_norm": 1.2215633687630816,
+      "learning_rate": 1.8508378855761097e-05,
+      "loss": 1.1812,
+      "step": 315
+    },
+    {
+      "epoch": 0.7783251231527094,
+      "grad_norm": 1.1163715503978044,
+      "learning_rate": 1.849328295838674e-05,
+      "loss": 1.1803,
+      "step": 316
+    },
+    {
+      "epoch": 0.7807881773399015,
+      "grad_norm": 1.119203309438992,
+      "learning_rate": 1.8478117277321967e-05,
+      "loss": 1.1608,
+      "step": 317
+    },
+    {
+      "epoch": 0.7832512315270936,
+      "grad_norm": 1.1277883201622099,
+      "learning_rate": 1.8462881937173144e-05,
+      "loss": 1.2806,
+      "step": 318
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 1.2201333631258782,
+      "learning_rate": 1.844757706311899e-05,
+      "loss": 1.2105,
+      "step": 319
+    },
+    {
+      "epoch": 0.7881773399014779,
+      "grad_norm": 1.1377305185003392,
+      "learning_rate": 1.8432202780909542e-05,
+      "loss": 1.2321,
+      "step": 320
+    },
+    {
+      "epoch": 0.7906403940886699,
+      "grad_norm": 1.1811650603904118,
+      "learning_rate": 1.8416759216865104e-05,
+      "loss": 1.2735,
+      "step": 321
+    },
+    {
+      "epoch": 0.7931034482758621,
+      "grad_norm": 1.3545418684139807,
+      "learning_rate": 1.8401246497875238e-05,
+      "loss": 1.0759,
+      "step": 322
+    },
+    {
+      "epoch": 0.7955665024630542,
+      "grad_norm": 1.0451387839811885,
+      "learning_rate": 1.838566475139769e-05,
+      "loss": 1.0708,
+      "step": 323
+    },
+    {
+      "epoch": 0.7980295566502463,
+      "grad_norm": 1.0753675941248633,
+      "learning_rate": 1.8370014105457378e-05,
+      "loss": 1.1107,
+      "step": 324
+    },
+    {
+      "epoch": 0.8004926108374384,
+      "grad_norm": 1.1922876818399974,
+      "learning_rate": 1.8354294688645303e-05,
+      "loss": 1.3381,
+      "step": 325
+    },
+    {
+      "epoch": 0.8029556650246306,
+      "grad_norm": 1.1939594364688118,
+      "learning_rate": 1.8338506630117527e-05,
+      "loss": 1.1409,
+      "step": 326
+    },
+    {
+      "epoch": 0.8054187192118226,
+      "grad_norm": 1.18442161488455,
+      "learning_rate": 1.8322650059594087e-05,
+      "loss": 1.123,
+      "step": 327
+    },
+    {
+      "epoch": 0.8078817733990148,
+      "grad_norm": 1.1238133211722114,
+      "learning_rate": 1.8306725107357933e-05,
+      "loss": 1.2287,
+      "step": 328
+    },
+    {
+      "epoch": 0.8103448275862069,
+      "grad_norm": 1.1396376634626217,
+      "learning_rate": 1.8290731904253874e-05,
+      "loss": 1.1696,
+      "step": 329
+    },
+    {
+      "epoch": 0.812807881773399,
+      "grad_norm": 1.104245428958938,
+      "learning_rate": 1.827467058168748e-05,
+      "loss": 1.1477,
+      "step": 330
+    },
+    {
+      "epoch": 0.8152709359605911,
+      "grad_norm": 1.0173077068409289,
+      "learning_rate": 1.8258541271624025e-05,
+      "loss": 1.1542,
+      "step": 331
+    },
+    {
+      "epoch": 0.8177339901477833,
+      "grad_norm": 1.1244744556794186,
+      "learning_rate": 1.824234410658738e-05,
+      "loss": 1.0486,
+      "step": 332
+    },
+    {
+      "epoch": 0.8201970443349754,
+      "grad_norm": 1.1579675032712282,
+      "learning_rate": 1.8226079219658944e-05,
+      "loss": 1.1863,
+      "step": 333
+    },
+    {
+      "epoch": 0.8226600985221675,
+      "grad_norm": 1.0784521850364304,
+      "learning_rate": 1.8209746744476538e-05,
+      "loss": 1.0729,
+      "step": 334
+    },
+    {
+      "epoch": 0.8251231527093597,
+      "grad_norm": 1.2563302920668962,
+      "learning_rate": 1.819334681523331e-05,
+      "loss": 1.1819,
+      "step": 335
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 1.0556548027317063,
+      "learning_rate": 1.817687956667664e-05,
+      "loss": 1.1346,
+      "step": 336
+    },
+    {
+      "epoch": 0.8300492610837439,
+      "grad_norm": 1.162947818681008,
+      "learning_rate": 1.816034513410702e-05,
+      "loss": 1.2799,
+      "step": 337
+    },
+    {
+      "epoch": 0.8325123152709359,
+      "grad_norm": 1.164865306549888,
+      "learning_rate": 1.8143743653376944e-05,
+      "loss": 1.1781,
+      "step": 338
+    },
+    {
+      "epoch": 0.8349753694581281,
+      "grad_norm": 1.2084232932617742,
+      "learning_rate": 1.8127075260889807e-05,
+      "loss": 1.3102,
+      "step": 339
+    },
+    {
+      "epoch": 0.8374384236453202,
+      "grad_norm": 1.2149928942814014,
+      "learning_rate": 1.811034009359877e-05,
+      "loss": 1.1121,
+      "step": 340
+    },
+    {
+      "epoch": 0.8399014778325123,
+      "grad_norm": 1.2306877304680588,
+      "learning_rate": 1.8093538289005635e-05,
+      "loss": 1.145,
+      "step": 341
+    },
+    {
+      "epoch": 0.8423645320197044,
+      "grad_norm": 1.091540803035678,
+      "learning_rate": 1.8076669985159726e-05,
+      "loss": 1.1959,
+      "step": 342
+    },
+    {
+      "epoch": 0.8448275862068966,
+      "grad_norm": 1.3613289131151323,
+      "learning_rate": 1.8059735320656738e-05,
+      "loss": 1.2083,
+      "step": 343
+    },
+    {
+      "epoch": 0.8472906403940886,
+      "grad_norm": 1.2232906657806824,
+      "learning_rate": 1.8042734434637615e-05,
+      "loss": 1.1454,
+      "step": 344
+    },
+    {
+      "epoch": 0.8497536945812808,
+      "grad_norm": 1.087381733569275,
+      "learning_rate": 1.8025667466787395e-05,
+      "loss": 1.2648,
+      "step": 345
+    },
+    {
+      "epoch": 0.8522167487684729,
+      "grad_norm": 1.1778176248205627,
+      "learning_rate": 1.8008534557334064e-05,
+      "loss": 1.2206,
+      "step": 346
+    },
+    {
+      "epoch": 0.854679802955665,
+      "grad_norm": 1.0706060522109717,
+      "learning_rate": 1.799133584704742e-05,
+      "loss": 1.1613,
+      "step": 347
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 1.1963795904044179,
+      "learning_rate": 1.7974071477237887e-05,
+      "loss": 1.2371,
+      "step": 348
+    },
+    {
+      "epoch": 0.8596059113300493,
+      "grad_norm": 1.123023416025367,
+      "learning_rate": 1.7956741589755383e-05,
+      "loss": 1.2587,
+      "step": 349
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 1.182261809701674,
+      "learning_rate": 1.7939346326988127e-05,
+      "loss": 1.1528,
+      "step": 350
+    },
+    {
+      "epoch": 0.8645320197044335,
+      "grad_norm": 1.1800429964314092,
+      "learning_rate": 1.7921885831861497e-05,
+      "loss": 1.1786,
+      "step": 351
+    },
+    {
+      "epoch": 0.8669950738916257,
+      "grad_norm": 1.0474477880571182,
+      "learning_rate": 1.7904360247836838e-05,
+      "loss": 1.1392,
+      "step": 352
+    },
+    {
+      "epoch": 0.8694581280788177,
+      "grad_norm": 1.1750124829217656,
+      "learning_rate": 1.788676971891028e-05,
+      "loss": 1.1595,
+      "step": 353
+    },
+    {
+      "epoch": 0.8719211822660099,
+      "grad_norm": 1.1514700470746069,
+      "learning_rate": 1.7869114389611574e-05,
+      "loss": 1.2797,
+      "step": 354
+    },
+    {
+      "epoch": 0.874384236453202,
+      "grad_norm": 1.154059319490481,
+      "learning_rate": 1.7851394405002885e-05,
+      "loss": 1.1745,
+      "step": 355
+    },
+    {
+      "epoch": 0.8768472906403941,
+      "grad_norm": 1.1453159493554437,
+      "learning_rate": 1.7833609910677613e-05,
+      "loss": 1.2188,
+      "step": 356
+    },
+    {
+      "epoch": 0.8793103448275862,
+      "grad_norm": 1.002442417060651,
+      "learning_rate": 1.781576105275919e-05,
+      "loss": 1.1681,
+      "step": 357
+    },
+    {
+      "epoch": 0.8817733990147784,
+      "grad_norm": 1.1572955275962467,
+      "learning_rate": 1.7797847977899873e-05,
+      "loss": 1.2138,
+      "step": 358
+    },
+    {
+      "epoch": 0.8842364532019704,
+      "grad_norm": 1.1397441441376521,
+      "learning_rate": 1.777987083327956e-05,
+      "loss": 1.2463,
+      "step": 359
+    },
+    {
+      "epoch": 0.8866995073891626,
+      "grad_norm": 1.182544658777311,
+      "learning_rate": 1.7761829766604556e-05,
+      "loss": 1.2148,
+      "step": 360
+    },
+    {
+      "epoch": 0.8891625615763546,
+      "grad_norm": 1.0226000788614955,
+      "learning_rate": 1.7743724926106387e-05,
+      "loss": 1.1402,
+      "step": 361
+    },
+    {
+      "epoch": 0.8916256157635468,
+      "grad_norm": 1.0975942899654156,
+      "learning_rate": 1.7725556460540553e-05,
+      "loss": 1.1199,
+      "step": 362
+    },
+    {
+      "epoch": 0.8940886699507389,
+      "grad_norm": 1.2060301743460258,
+      "learning_rate": 1.770732451918532e-05,
+      "loss": 1.1245,
+      "step": 363
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 1.079957648347576,
+      "learning_rate": 1.7689029251840492e-05,
+      "loss": 1.2386,
+      "step": 364
+    },
+    {
+      "epoch": 0.8990147783251231,
+      "grad_norm": 1.1712605709384634,
+      "learning_rate": 1.7670670808826193e-05,
+      "loss": 1.1207,
+      "step": 365
+    },
+    {
+      "epoch": 0.9014778325123153,
+      "grad_norm": 1.0974275615482618,
+      "learning_rate": 1.7652249340981608e-05,
+      "loss": 1.1442,
+      "step": 366
+    },
+    {
+      "epoch": 0.9039408866995073,
+      "grad_norm": 1.2576416089085674,
+      "learning_rate": 1.7633764999663753e-05,
+      "loss": 1.2465,
+      "step": 367
+    },
+    {
+      "epoch": 0.9064039408866995,
+      "grad_norm": 1.170403586740402,
+      "learning_rate": 1.7615217936746246e-05,
+      "loss": 1.2322,
+      "step": 368
+    },
+    {
+      "epoch": 0.9088669950738916,
+      "grad_norm": 1.3529765939710923,
+      "learning_rate": 1.7596608304618037e-05,
+      "loss": 1.1621,
+      "step": 369
+    },
+    {
+      "epoch": 0.9113300492610837,
+      "grad_norm": 1.19970464647962,
+      "learning_rate": 1.757793625618217e-05,
+      "loss": 1.1752,
+      "step": 370
+    },
+    {
+      "epoch": 0.9137931034482759,
+      "grad_norm": 1.045866124271059,
+      "learning_rate": 1.7559201944854515e-05,
+      "loss": 1.1167,
+      "step": 371
+    },
+    {
+      "epoch": 0.916256157635468,
+      "grad_norm": 1.2380282266877778,
+      "learning_rate": 1.7540405524562533e-05,
+      "loss": 1.1254,
+      "step": 372
+    },
+    {
+      "epoch": 0.9187192118226601,
+      "grad_norm": 1.2107504915610179,
+      "learning_rate": 1.752154714974397e-05,
+      "loss": 1.1664,
+      "step": 373
+    },
+    {
+      "epoch": 0.9211822660098522,
+      "grad_norm": 1.1048664298823,
+      "learning_rate": 1.750262697534563e-05,
+      "loss": 1.1247,
+      "step": 374
+    },
+    {
+      "epoch": 0.9236453201970444,
+      "grad_norm": 1.2029399481805099,
+      "learning_rate": 1.748364515682207e-05,
+      "loss": 1.2152,
+      "step": 375
+    },
+    {
+      "epoch": 0.9261083743842364,
+      "grad_norm": 1.2155006815831146,
+      "learning_rate": 1.7464601850134353e-05,
+      "loss": 1.199,
+      "step": 376
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 1.121059722915666,
+      "learning_rate": 1.744549721174873e-05,
+      "loss": 1.1952,
+      "step": 377
+    },
+    {
+      "epoch": 0.9310344827586207,
+      "grad_norm": 1.3679700954916463,
+      "learning_rate": 1.742633139863538e-05,
+      "loss": 1.1492,
+      "step": 378
+    },
+    {
+      "epoch": 0.9334975369458128,
+      "grad_norm": 1.1654891929567284,
+      "learning_rate": 1.740710456826713e-05,
+      "loss": 1.2148,
+      "step": 379
+    },
+    {
+      "epoch": 0.9359605911330049,
+      "grad_norm": 1.1914479232753978,
+      "learning_rate": 1.738781687861812e-05,
+      "loss": 1.2196,
+      "step": 380
+    },
+    {
+      "epoch": 0.9384236453201971,
+      "grad_norm": 1.308899573274587,
+      "learning_rate": 1.7368468488162547e-05,
+      "loss": 1.1854,
+      "step": 381
+    },
+    {
+      "epoch": 0.9408866995073891,
+      "grad_norm": 1.1135936633329866,
+      "learning_rate": 1.7349059555873348e-05,
+      "loss": 1.2045,
+      "step": 382
+    },
+    {
+      "epoch": 0.9433497536945813,
+      "grad_norm": 1.040732626175033,
+      "learning_rate": 1.732959024122088e-05,
+      "loss": 1.1462,
+      "step": 383
+    },
+    {
+      "epoch": 0.9458128078817734,
+      "grad_norm": 1.117833464279004,
+      "learning_rate": 1.731006070417163e-05,
+      "loss": 1.2479,
+      "step": 384
+    },
+    {
+      "epoch": 0.9482758620689655,
+      "grad_norm": 1.0208867441853824,
+      "learning_rate": 1.7290471105186893e-05,
+      "loss": 1.1074,
+      "step": 385
+    },
+    {
+      "epoch": 0.9507389162561576,
+      "grad_norm": 1.1191058461375822,
+      "learning_rate": 1.7270821605221448e-05,
+      "loss": 1.1831,
+      "step": 386
+    },
+    {
+      "epoch": 0.9532019704433498,
+      "grad_norm": 1.0649430293753392,
+      "learning_rate": 1.725111236572225e-05,
+      "loss": 1.0715,
+      "step": 387
+    },
+    {
+      "epoch": 0.9556650246305419,
+      "grad_norm": 1.034587430317254,
+      "learning_rate": 1.7231343548627085e-05,
+      "loss": 1.1789,
+      "step": 388
+    },
+    {
+      "epoch": 0.958128078817734,
+      "grad_norm": 1.1630666684749342,
+      "learning_rate": 1.7211515316363252e-05,
+      "loss": 1.1998,
+      "step": 389
+    },
+    {
+      "epoch": 0.9605911330049262,
+      "grad_norm": 1.1040157838812632,
+      "learning_rate": 1.7191627831846226e-05,
+      "loss": 1.1184,
+      "step": 390
+    },
+    {
+      "epoch": 0.9630541871921182,
+      "grad_norm": 1.0866918345230143,
+      "learning_rate": 1.7171681258478316e-05,
+      "loss": 1.2642,
+      "step": 391
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 1.1160664302169045,
+      "learning_rate": 1.7151675760147325e-05,
+      "loss": 1.1768,
+      "step": 392
+    },
+    {
+      "epoch": 0.9679802955665024,
+      "grad_norm": 1.0898737202560378,
+      "learning_rate": 1.7131611501225215e-05,
+      "loss": 1.077,
+      "step": 393
+    },
+    {
+      "epoch": 0.9704433497536946,
+      "grad_norm": 1.171609009119484,
+      "learning_rate": 1.7111488646566728e-05,
+      "loss": 1.0983,
+      "step": 394
+    },
+    {
+      "epoch": 0.9729064039408867,
+      "grad_norm": 1.143712763362563,
+      "learning_rate": 1.7091307361508057e-05,
+      "loss": 1.161,
+      "step": 395
+    },
+    {
+      "epoch": 0.9753694581280788,
+      "grad_norm": 1.1019522335964358,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.2167,
+      "step": 396
+    },
+    {
+      "epoch": 0.9778325123152709,
+      "grad_norm": 1.0502571018254019,
+      "learning_rate": 1.7050770163933985e-05,
+      "loss": 1.1345,
+      "step": 397
+    },
+    {
+      "epoch": 0.9802955665024631,
+      "grad_norm": 1.1531281985520245,
+      "learning_rate": 1.7030414584485938e-05,
+      "loss": 1.1951,
+      "step": 398
+    },
+    {
+      "epoch": 0.9827586206896551,
+      "grad_norm": 1.1454655606151676,
+      "learning_rate": 1.701000124076967e-05,
+      "loss": 1.1806,
+      "step": 399
+    },
+    {
+      "epoch": 0.9852216748768473,
+      "grad_norm": 1.027735356294518,
+      "learning_rate": 1.6989530300508126e-05,
+      "loss": 1.2236,
+      "step": 400
+    },
+    {
+      "epoch": 0.9876847290640394,
+      "grad_norm": 1.2813557837011695,
+      "learning_rate": 1.6969001931897492e-05,
+      "loss": 1.2562,
+      "step": 401
+    },
+    {
+      "epoch": 0.9901477832512315,
+      "grad_norm": 1.1138267385331413,
+      "learning_rate": 1.6948416303605796e-05,
+      "loss": 1.1348,
+      "step": 402
+    },
+    {
+      "epoch": 0.9926108374384236,
+      "grad_norm": 1.1495412077219034,
+      "learning_rate": 1.692777358477154e-05,
+      "loss": 1.1724,
+      "step": 403
+    },
+    {
+      "epoch": 0.9950738916256158,
+      "grad_norm": 1.2200742520734178,
+      "learning_rate": 1.690707394500229e-05,
+      "loss": 1.172,
+      "step": 404
+    },
+    {
+      "epoch": 0.9975369458128078,
+      "grad_norm": 1.1441115823994843,
+      "learning_rate": 1.6886317554373304e-05,
+      "loss": 1.26,
+      "step": 405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.147905758611473,
+      "learning_rate": 1.6865504583426117e-05,
+      "loss": 1.2232,
+      "step": 406
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.1330950260162354,
+      "eval_runtime": 1.197,
+      "eval_samples_per_second": 44.277,
+      "eval_steps_per_second": 1.671,
+      "step": 406
+    },
+    {
+      "epoch": 1.0024630541871922,
+      "grad_norm": 1.170172561000041,
+      "learning_rate": 1.684463520316715e-05,
+      "loss": 0.8731,
+      "step": 407
+    },
+    {
+      "epoch": 1.0049261083743843,
+      "grad_norm": 1.2237706574828835,
+      "learning_rate": 1.6823709585066308e-05,
+      "loss": 0.8808,
+      "step": 408
+    },
+    {
+      "epoch": 1.0073891625615763,
+      "grad_norm": 1.2298905240162896,
+      "learning_rate": 1.6802727901055555e-05,
+      "loss": 0.7646,
+      "step": 409
+    },
+    {
+      "epoch": 1.0098522167487685,
+      "grad_norm": 1.0593730690187393,
+      "learning_rate": 1.6781690323527512e-05,
+      "loss": 0.8189,
+      "step": 410
+    },
+    {
+      "epoch": 1.0123152709359606,
+      "grad_norm": 1.1332111051976046,
+      "learning_rate": 1.6760597025334046e-05,
+      "loss": 0.8166,
+      "step": 411
+    },
+    {
+      "epoch": 1.0147783251231528,
+      "grad_norm": 1.1310612789825578,
+      "learning_rate": 1.6739448179784846e-05,
+      "loss": 0.8595,
+      "step": 412
+    },
+    {
+      "epoch": 1.0172413793103448,
+      "grad_norm": 1.3093854189073562,
+      "learning_rate": 1.6718243960645984e-05,
+      "loss": 0.7923,
+      "step": 413
+    },
+    {
+      "epoch": 1.019704433497537,
+      "grad_norm": 1.2336577734057612,
+      "learning_rate": 1.669698454213852e-05,
+      "loss": 0.8729,
+      "step": 414
+    },
+    {
+      "epoch": 1.022167487684729,
+      "grad_norm": 1.2237802524743282,
+      "learning_rate": 1.6675670098937034e-05,
+      "loss": 0.7861,
+      "step": 415
+    },
+    {
+      "epoch": 1.0246305418719213,
+      "grad_norm": 1.2118327123494963,
+      "learning_rate": 1.665430080616821e-05,
+      "loss": 0.7735,
+      "step": 416
+    },
+    {
+      "epoch": 1.0270935960591132,
+      "grad_norm": 1.3244396464334183,
+      "learning_rate": 1.66328768394094e-05,
+      "loss": 0.8035,
+      "step": 417
+    },
+    {
+      "epoch": 1.0295566502463054,
+      "grad_norm": 1.3411534528299813,
+      "learning_rate": 1.6611398374687172e-05,
+      "loss": 0.7913,
+      "step": 418
+    },
+    {
+      "epoch": 1.0320197044334976,
+      "grad_norm": 1.3938146293948532,
+      "learning_rate": 1.6589865588475872e-05,
+      "loss": 0.8104,
+      "step": 419
+    },
+    {
+      "epoch": 1.0344827586206897,
+      "grad_norm": 1.2846576685911195,
+      "learning_rate": 1.6568278657696166e-05,
+      "loss": 0.8342,
+      "step": 420
+    },
+    {
+      "epoch": 1.0369458128078817,
+      "grad_norm": 1.2818325858164603,
+      "learning_rate": 1.6546637759713588e-05,
+      "loss": 0.8619,
+      "step": 421
+    },
+    {
+      "epoch": 1.0394088669950738,
+      "grad_norm": 1.2219684847999306,
+      "learning_rate": 1.6524943072337094e-05,
+      "loss": 0.9003,
+      "step": 422
+    },
+    {
+      "epoch": 1.041871921182266,
+      "grad_norm": 1.1858559822945252,
+      "learning_rate": 1.6503194773817583e-05,
+      "loss": 0.7633,
+      "step": 423
+    },
+    {
+      "epoch": 1.0443349753694582,
+      "grad_norm": 1.3685348411308935,
+      "learning_rate": 1.6481393042846442e-05,
+      "loss": 0.814,
+      "step": 424
+    },
+    {
+      "epoch": 1.0467980295566504,
+      "grad_norm": 1.2588090443708368,
+      "learning_rate": 1.6459538058554088e-05,
+      "loss": 0.7668,
+      "step": 425
+    },
+    {
+      "epoch": 1.0492610837438423,
+      "grad_norm": 1.1730454766579244,
+      "learning_rate": 1.6437630000508466e-05,
+      "loss": 0.8318,
+      "step": 426
+    },
+    {
+      "epoch": 1.0517241379310345,
+      "grad_norm": 1.1178207203491073,
+      "learning_rate": 1.6415669048713608e-05,
+      "loss": 0.8044,
+      "step": 427
+    },
+    {
+      "epoch": 1.0541871921182266,
+      "grad_norm": 1.1675768038300958,
+      "learning_rate": 1.6393655383608132e-05,
+      "loss": 0.904,
+      "step": 428
+    },
+    {
+      "epoch": 1.0566502463054188,
+      "grad_norm": 1.2060282468793229,
+      "learning_rate": 1.6371589186063778e-05,
+      "loss": 0.7863,
+      "step": 429
+    },
+    {
+      "epoch": 1.0591133004926108,
+      "grad_norm": 1.1889344636454922,
+      "learning_rate": 1.634947063738389e-05,
+      "loss": 0.8245,
+      "step": 430
+    },
+    {
+      "epoch": 1.061576354679803,
+      "grad_norm": 1.1473348605104958,
+      "learning_rate": 1.6327299919301967e-05,
+      "loss": 0.8468,
+      "step": 431
+    },
+    {
+      "epoch": 1.064039408866995,
+      "grad_norm": 1.460006662118422,
+      "learning_rate": 1.630507721398013e-05,
+      "loss": 0.7999,
+      "step": 432
+    },
+    {
+      "epoch": 1.0665024630541873,
+      "grad_norm": 1.1264533086549082,
+      "learning_rate": 1.6282802704007668e-05,
+      "loss": 0.6685,
+      "step": 433
+    },
+    {
+      "epoch": 1.0689655172413792,
+      "grad_norm": 1.3107475661495822,
+      "learning_rate": 1.6260476572399494e-05,
+      "loss": 0.8646,
+      "step": 434
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 1.1024793231954173,
+      "learning_rate": 1.6238099002594672e-05,
+      "loss": 0.8254,
+      "step": 435
+    },
+    {
+      "epoch": 1.0738916256157636,
+      "grad_norm": 1.1437338176478864,
+      "learning_rate": 1.6215670178454893e-05,
+      "loss": 0.7006,
+      "step": 436
+    },
+    {
+      "epoch": 1.0763546798029557,
+      "grad_norm": 1.2641208800236743,
+      "learning_rate": 1.6193190284262982e-05,
+      "loss": 0.7703,
+      "step": 437
+    },
+    {
+      "epoch": 1.0788177339901477,
+      "grad_norm": 1.241387163403215,
+      "learning_rate": 1.6170659504721365e-05,
+      "loss": 0.8022,
+      "step": 438
+    },
+    {
+      "epoch": 1.0812807881773399,
+      "grad_norm": 1.302912389668731,
+      "learning_rate": 1.6148078024950553e-05,
+      "loss": 0.8705,
+      "step": 439
+    },
+    {
+      "epoch": 1.083743842364532,
+      "grad_norm": 1.1043132768619937,
+      "learning_rate": 1.6125446030487642e-05,
+      "loss": 0.7861,
+      "step": 440
+    },
+    {
+      "epoch": 1.0862068965517242,
+      "grad_norm": 1.2300049891797116,
+      "learning_rate": 1.610276370728477e-05,
+      "loss": 0.7969,
+      "step": 441
+    },
+    {
+      "epoch": 1.0886699507389164,
+      "grad_norm": 1.2137604817685268,
+      "learning_rate": 1.608003124170758e-05,
+      "loss": 0.7967,
+      "step": 442
+    },
+    {
+      "epoch": 1.0911330049261083,
+      "grad_norm": 1.1329733898744914,
+      "learning_rate": 1.6057248820533712e-05,
+      "loss": 0.8096,
+      "step": 443
+    },
+    {
+      "epoch": 1.0935960591133005,
+      "grad_norm": 1.183662689517777,
+      "learning_rate": 1.6034416630951265e-05,
+      "loss": 0.7719,
+      "step": 444
+    },
+    {
+      "epoch": 1.0960591133004927,
+      "grad_norm": 1.1511925588354042,
+      "learning_rate": 1.6011534860557238e-05,
+      "loss": 0.8161,
+      "step": 445
+    },
+    {
+      "epoch": 1.0985221674876848,
+      "grad_norm": 1.145735279858507,
+      "learning_rate": 1.598860369735601e-05,
+      "loss": 0.7658,
+      "step": 446
+    },
+    {
+      "epoch": 1.1009852216748768,
+      "grad_norm": 1.2799792617369568,
+      "learning_rate": 1.5965623329757795e-05,
+      "loss": 0.7895,
+      "step": 447
+    },
+    {
+      "epoch": 1.103448275862069,
+      "grad_norm": 1.4027373630169981,
+      "learning_rate": 1.594259394657707e-05,
+      "loss": 0.8342,
+      "step": 448
+    },
+    {
+      "epoch": 1.1059113300492611,
+      "grad_norm": 1.2011206142993758,
+      "learning_rate": 1.5919515737031052e-05,
+      "loss": 0.8278,
+      "step": 449
+    },
+    {
+      "epoch": 1.1083743842364533,
+      "grad_norm": 1.25576321484377,
+      "learning_rate": 1.589638889073813e-05,
+      "loss": 0.7734,
+      "step": 450
+    },
+    {
+      "epoch": 1.1108374384236452,
+      "grad_norm": 1.1615466895373159,
+      "learning_rate": 1.587321359771631e-05,
+      "loss": 0.8156,
+      "step": 451
+    },
+    {
+      "epoch": 1.1133004926108374,
+      "grad_norm": 1.0816311974629496,
+      "learning_rate": 1.584999004838165e-05,
+      "loss": 0.8087,
+      "step": 452
+    },
+    {
+      "epoch": 1.1157635467980296,
+      "grad_norm": 1.226745834741543,
+      "learning_rate": 1.58267184335467e-05,
+      "loss": 0.7927,
+      "step": 453
+    },
+    {
+      "epoch": 1.1182266009852218,
+      "grad_norm": 1.2317871780870304,
+      "learning_rate": 1.5803398944418934e-05,
+      "loss": 0.8278,
+      "step": 454
+    },
+    {
+      "epoch": 1.1206896551724137,
+      "grad_norm": 1.1296849188768545,
+      "learning_rate": 1.5780031772599174e-05,
+      "loss": 0.867,
+      "step": 455
+    },
+    {
+      "epoch": 1.1231527093596059,
+      "grad_norm": 1.2021358932472797,
+      "learning_rate": 1.5756617110080023e-05,
+      "loss": 0.8528,
+      "step": 456
+    },
+    {
+      "epoch": 1.125615763546798,
+      "grad_norm": 1.2362728257188624,
+      "learning_rate": 1.573315514924428e-05,
+      "loss": 0.7861,
+      "step": 457
+    },
+    {
+      "epoch": 1.1280788177339902,
+      "grad_norm": 1.1978579078305969,
+      "learning_rate": 1.570964608286336e-05,
+      "loss": 0.8233,
+      "step": 458
+    },
+    {
+      "epoch": 1.1305418719211824,
+      "grad_norm": 1.112676355763924,
+      "learning_rate": 1.5686090104095726e-05,
+      "loss": 0.8448,
+      "step": 459
+    },
+    {
+      "epoch": 1.1330049261083743,
+      "grad_norm": 1.211505830357936,
+      "learning_rate": 1.5662487406485273e-05,
+      "loss": 0.8085,
+      "step": 460
+    },
+    {
+      "epoch": 1.1354679802955665,
+      "grad_norm": 1.1577469697804295,
+      "learning_rate": 1.5638838183959768e-05,
+      "loss": 0.8649,
+      "step": 461
+    },
+    {
+      "epoch": 1.1379310344827587,
+      "grad_norm": 1.0983751868077105,
+      "learning_rate": 1.561514263082923e-05,
+      "loss": 0.8016,
+      "step": 462
+    },
+    {
+      "epoch": 1.1403940886699506,
+      "grad_norm": 1.1740547761112214,
+      "learning_rate": 1.5591400941784354e-05,
+      "loss": 0.7641,
+      "step": 463
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 1.2318617428976384,
+      "learning_rate": 1.5567613311894908e-05,
+      "loss": 0.8657,
+      "step": 464
+    },
+    {
+      "epoch": 1.145320197044335,
+      "grad_norm": 1.1436631487613726,
+      "learning_rate": 1.554377993660811e-05,
+      "loss": 0.7574,
+      "step": 465
+    },
+    {
+      "epoch": 1.1477832512315271,
+      "grad_norm": 1.1703552054952053,
+      "learning_rate": 1.5519901011747046e-05,
+      "loss": 0.8259,
+      "step": 466
+    },
+    {
+      "epoch": 1.1502463054187193,
+      "grad_norm": 1.3066168128610796,
+      "learning_rate": 1.5495976733509058e-05,
+      "loss": 0.8162,
+      "step": 467
+    },
+    {
+      "epoch": 1.1527093596059113,
+      "grad_norm": 1.4001859677641861,
+      "learning_rate": 1.5472007298464117e-05,
+      "loss": 0.8298,
+      "step": 468
+    },
+    {
+      "epoch": 1.1551724137931034,
+      "grad_norm": 1.201916818749201,
+      "learning_rate": 1.544799290355323e-05,
+      "loss": 0.8599,
+      "step": 469
+    },
+    {
+      "epoch": 1.1576354679802956,
+      "grad_norm": 1.3128386015318791,
+      "learning_rate": 1.5423933746086793e-05,
+      "loss": 0.8812,
+      "step": 470
+    },
+    {
+      "epoch": 1.1600985221674878,
+      "grad_norm": 1.1665217900013063,
+      "learning_rate": 1.5399830023743004e-05,
+      "loss": 0.8279,
+      "step": 471
+    },
+    {
+      "epoch": 1.1625615763546797,
+      "grad_norm": 1.2208438395459311,
+      "learning_rate": 1.5375681934566203e-05,
+      "loss": 0.8202,
+      "step": 472
+    },
+    {
+      "epoch": 1.1650246305418719,
+      "grad_norm": 1.1704164201666705,
+      "learning_rate": 1.5351489676965283e-05,
+      "loss": 0.8314,
+      "step": 473
+    },
+    {
+      "epoch": 1.167487684729064,
+      "grad_norm": 1.2284782658704982,
+      "learning_rate": 1.532725344971202e-05,
+      "loss": 0.815,
+      "step": 474
+    },
+    {
+      "epoch": 1.1699507389162562,
+      "grad_norm": 1.1594969035013099,
+      "learning_rate": 1.5302973451939472e-05,
+      "loss": 0.7647,
+      "step": 475
+    },
+    {
+      "epoch": 1.1724137931034484,
+      "grad_norm": 1.1280002852635664,
+      "learning_rate": 1.527864988314033e-05,
+      "loss": 0.8097,
+      "step": 476
+    },
+    {
+      "epoch": 1.1748768472906403,
+      "grad_norm": 1.2872121766964812,
+      "learning_rate": 1.525428294316527e-05,
+      "loss": 0.8599,
+      "step": 477
+    },
+    {
+      "epoch": 1.1773399014778325,
+      "grad_norm": 1.185034098959484,
+      "learning_rate": 1.5229872832221336e-05,
+      "loss": 0.7931,
+      "step": 478
+    },
+    {
+      "epoch": 1.1798029556650247,
+      "grad_norm": 1.1623622074294984,
+      "learning_rate": 1.5205419750870261e-05,
+      "loss": 0.8038,
+      "step": 479
+    },
+    {
+      "epoch": 1.1822660098522166,
+      "grad_norm": 1.1952324634758382,
+      "learning_rate": 1.5180923900026847e-05,
+      "loss": 0.7815,
+      "step": 480
+    },
+    {
+      "epoch": 1.1847290640394088,
+      "grad_norm": 1.2525927357406512,
+      "learning_rate": 1.5156385480957312e-05,
+      "loss": 0.8219,
+      "step": 481
+    },
+    {
+      "epoch": 1.187192118226601,
+      "grad_norm": 1.2038365526465693,
+      "learning_rate": 1.5131804695277612e-05,
+      "loss": 0.8542,
+      "step": 482
+    },
+    {
+      "epoch": 1.1896551724137931,
+      "grad_norm": 1.2858558767933885,
+      "learning_rate": 1.5107181744951818e-05,
+      "loss": 0.8084,
+      "step": 483
+    },
+    {
+      "epoch": 1.1921182266009853,
+      "grad_norm": 1.2085049715583003,
+      "learning_rate": 1.5082516832290424e-05,
+      "loss": 0.7713,
+      "step": 484
+    },
+    {
+      "epoch": 1.1945812807881773,
+      "grad_norm": 1.3373843829371959,
+      "learning_rate": 1.5057810159948715e-05,
+      "loss": 0.8367,
+      "step": 485
+    },
+    {
+      "epoch": 1.1970443349753694,
+      "grad_norm": 1.2811437423188181,
+      "learning_rate": 1.5033061930925081e-05,
+      "loss": 0.8871,
+      "step": 486
+    },
+    {
+      "epoch": 1.1995073891625616,
+      "grad_norm": 1.1471284944785145,
+      "learning_rate": 1.5008272348559359e-05,
+      "loss": 0.8107,
+      "step": 487
+    },
+    {
+      "epoch": 1.2019704433497538,
+      "grad_norm": 1.1844999665131086,
+      "learning_rate": 1.4983441616531152e-05,
+      "loss": 0.7747,
+      "step": 488
+    },
+    {
+      "epoch": 1.2044334975369457,
+      "grad_norm": 1.3337471389382305,
+      "learning_rate": 1.4958569938858169e-05,
+      "loss": 0.8162,
+      "step": 489
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 1.270924528424841,
+      "learning_rate": 1.4933657519894542e-05,
+      "loss": 0.7758,
+      "step": 490
+    },
+    {
+      "epoch": 1.20935960591133,
+      "grad_norm": 1.1929882615048306,
+      "learning_rate": 1.4908704564329144e-05,
+      "loss": 0.8396,
+      "step": 491
+    },
+    {
+      "epoch": 1.2118226600985222,
+      "grad_norm": 1.1581136787960526,
+      "learning_rate": 1.4883711277183917e-05,
+      "loss": 0.758,
+      "step": 492
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 1.152599758870068,
+      "learning_rate": 1.485867786381217e-05,
+      "loss": 0.7743,
+      "step": 493
+    },
+    {
+      "epoch": 1.2167487684729064,
+      "grad_norm": 1.0976649382440848,
+      "learning_rate": 1.483360452989691e-05,
+      "loss": 0.7754,
+      "step": 494
+    },
+    {
+      "epoch": 1.2192118226600985,
+      "grad_norm": 1.3639201768384124,
+      "learning_rate": 1.4808491481449146e-05,
+      "loss": 0.7712,
+      "step": 495
+    },
+    {
+      "epoch": 1.2216748768472907,
+      "grad_norm": 1.1221870115996668,
+      "learning_rate": 1.4783338924806191e-05,
+      "loss": 0.7303,
+      "step": 496
+    },
+    {
+      "epoch": 1.2241379310344827,
+      "grad_norm": 1.204352211214488,
+      "learning_rate": 1.4758147066629975e-05,
+      "loss": 0.8276,
+      "step": 497
+    },
+    {
+      "epoch": 1.2266009852216748,
+      "grad_norm": 1.1667595156552748,
+      "learning_rate": 1.4732916113905336e-05,
+      "loss": 0.823,
+      "step": 498
+    },
+    {
+      "epoch": 1.229064039408867,
+      "grad_norm": 1.2298487963431766,
+      "learning_rate": 1.4707646273938335e-05,
+      "loss": 0.8354,
+      "step": 499
+    },
+    {
+      "epoch": 1.2315270935960592,
+      "grad_norm": 1.1513021643269783,
+      "learning_rate": 1.4682337754354534e-05,
+      "loss": 0.789,
+      "step": 500
+    },
+    {
+      "epoch": 1.2339901477832513,
+      "grad_norm": 1.1589932247552988,
+      "learning_rate": 1.465699076309731e-05,
+      "loss": 0.7952,
+      "step": 501
+    },
+    {
+      "epoch": 1.2364532019704433,
+      "grad_norm": 1.3026280395484064,
+      "learning_rate": 1.4631605508426124e-05,
+      "loss": 0.7599,
+      "step": 502
+    },
+    {
+      "epoch": 1.2389162561576355,
+      "grad_norm": 1.2515180429856536,
+      "learning_rate": 1.4606182198914835e-05,
+      "loss": 0.7587,
+      "step": 503
+    },
+    {
+      "epoch": 1.2413793103448276,
+      "grad_norm": 1.2677344180296528,
+      "learning_rate": 1.4580721043449968e-05,
+      "loss": 0.8503,
+      "step": 504
+    },
+    {
+      "epoch": 1.2438423645320198,
+      "grad_norm": 1.1694068857858309,
+      "learning_rate": 1.4555222251228997e-05,
+      "loss": 0.8142,
+      "step": 505
+    },
+    {
+      "epoch": 1.2463054187192117,
+      "grad_norm": 1.1193301587944917,
+      "learning_rate": 1.4529686031758642e-05,
+      "loss": 0.7871,
+      "step": 506
+    },
+    {
+      "epoch": 1.248768472906404,
+      "grad_norm": 1.1140491179830212,
+      "learning_rate": 1.450411259485314e-05,
+      "loss": 0.7957,
+      "step": 507
+    },
+    {
+      "epoch": 1.251231527093596,
+      "grad_norm": 1.3933130922322663,
+      "learning_rate": 1.4478502150632503e-05,
+      "loss": 0.8492,
+      "step": 508
+    },
+    {
+      "epoch": 1.2536945812807883,
+      "grad_norm": 1.2235960349776593,
+      "learning_rate": 1.4452854909520824e-05,
+      "loss": 0.7768,
+      "step": 509
+    },
+    {
+      "epoch": 1.2561576354679804,
+      "grad_norm": 1.1958254542517641,
+      "learning_rate": 1.4427171082244523e-05,
+      "loss": 0.8251,
+      "step": 510
+    },
+    {
+      "epoch": 1.2586206896551724,
+      "grad_norm": 1.2124657124656038,
+      "learning_rate": 1.4401450879830628e-05,
+      "loss": 0.8213,
+      "step": 511
+    },
+    {
+      "epoch": 1.2610837438423645,
+      "grad_norm": 1.1042709061463936,
+      "learning_rate": 1.4375694513605037e-05,
+      "loss": 0.7067,
+      "step": 512
+    },
+    {
+      "epoch": 1.2635467980295567,
+      "grad_norm": 1.1598541174611159,
+      "learning_rate": 1.4349902195190777e-05,
+      "loss": 0.8232,
+      "step": 513
+    },
+    {
+      "epoch": 1.2660098522167487,
+      "grad_norm": 1.1972923586973774,
+      "learning_rate": 1.4324074136506283e-05,
+      "loss": 0.7975,
+      "step": 514
+    },
+    {
+      "epoch": 1.2684729064039408,
+      "grad_norm": 1.2731865009477907,
+      "learning_rate": 1.429821054976363e-05,
+      "loss": 0.9027,
+      "step": 515
+    },
+    {
+      "epoch": 1.270935960591133,
+      "grad_norm": 1.155103011471133,
+      "learning_rate": 1.427231164746681e-05,
+      "loss": 0.7706,
+      "step": 516
+    },
+    {
+      "epoch": 1.2733990147783252,
+      "grad_norm": 1.2198465256227577,
+      "learning_rate": 1.424637764240998e-05,
+      "loss": 0.7563,
+      "step": 517
+    },
+    {
+      "epoch": 1.2758620689655173,
+      "grad_norm": 1.366993263615919,
+      "learning_rate": 1.4220408747675714e-05,
+      "loss": 0.7796,
+      "step": 518
+    },
+    {
+      "epoch": 1.2783251231527093,
+      "grad_norm": 1.2735003122389987,
+      "learning_rate": 1.419440517663325e-05,
+      "loss": 0.8419,
+      "step": 519
+    },
+    {
+      "epoch": 1.2807881773399015,
+      "grad_norm": 1.2122112834944652,
+      "learning_rate": 1.4168367142936736e-05,
+      "loss": 0.8188,
+      "step": 520
+    },
+    {
+      "epoch": 1.2832512315270936,
+      "grad_norm": 1.269392551768455,
+      "learning_rate": 1.4142294860523475e-05,
+      "loss": 0.7833,
+      "step": 521
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 1.3049386062650268,
+      "learning_rate": 1.4116188543612182e-05,
+      "loss": 0.7456,
+      "step": 522
+    },
+    {
+      "epoch": 1.2881773399014778,
+      "grad_norm": 1.2858511023159993,
+      "learning_rate": 1.4090048406701196e-05,
+      "loss": 0.8879,
+      "step": 523
+    },
+    {
+      "epoch": 1.29064039408867,
+      "grad_norm": 1.2398498419628219,
+      "learning_rate": 1.4063874664566734e-05,
+      "loss": 0.7817,
+      "step": 524
+    },
+    {
+      "epoch": 1.293103448275862,
+      "grad_norm": 1.4051514467224941,
+      "learning_rate": 1.4037667532261143e-05,
+      "loss": 0.8403,
+      "step": 525
+    },
+    {
+      "epoch": 1.2955665024630543,
+      "grad_norm": 1.1862173180052564,
+      "learning_rate": 1.4011427225111091e-05,
+      "loss": 0.7812,
+      "step": 526
+    },
+    {
+      "epoch": 1.2980295566502464,
+      "grad_norm": 1.200169338755385,
+      "learning_rate": 1.3985153958715833e-05,
+      "loss": 0.8574,
+      "step": 527
+    },
+    {
+      "epoch": 1.3004926108374384,
+      "grad_norm": 1.2982257564785258,
+      "learning_rate": 1.3958847948945428e-05,
+      "loss": 0.7913,
+      "step": 528
+    },
+    {
+      "epoch": 1.3029556650246306,
+      "grad_norm": 1.270166983252296,
+      "learning_rate": 1.3932509411938969e-05,
+      "loss": 0.8953,
+      "step": 529
+    },
+    {
+      "epoch": 1.3054187192118227,
+      "grad_norm": 1.4017248461544733,
+      "learning_rate": 1.3906138564102794e-05,
+      "loss": 0.7397,
+      "step": 530
+    },
+    {
+      "epoch": 1.3078817733990147,
+      "grad_norm": 1.303169921173925,
+      "learning_rate": 1.387973562210873e-05,
+      "loss": 0.8169,
+      "step": 531
+    },
+    {
+      "epoch": 1.3103448275862069,
+      "grad_norm": 1.2084993489539493,
+      "learning_rate": 1.3853300802892285e-05,
+      "loss": 0.7616,
+      "step": 532
+    },
+    {
+      "epoch": 1.312807881773399,
+      "grad_norm": 1.2014613849373372,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.8159,
+      "step": 533
+    },
+    {
+      "epoch": 1.3152709359605912,
+      "grad_norm": 1.2453837988149308,
+      "learning_rate": 1.380033640184213e-05,
+      "loss": 0.7579,
+      "step": 534
+    },
+    {
+      "epoch": 1.3177339901477834,
+      "grad_norm": 1.2381290859750014,
+      "learning_rate": 1.3773807255181877e-05,
+      "loss": 0.7243,
+      "step": 535
+    },
+    {
+      "epoch": 1.3201970443349753,
+      "grad_norm": 1.27147120221371,
+      "learning_rate": 1.3747247101642605e-05,
+      "loss": 0.8649,
+      "step": 536
+    },
+    {
+      "epoch": 1.3226600985221675,
+      "grad_norm": 1.158154491413447,
+      "learning_rate": 1.3720656159451528e-05,
+      "loss": 0.8011,
+      "step": 537
+    },
+    {
+      "epoch": 1.3251231527093597,
+      "grad_norm": 1.3017650011669712,
+      "learning_rate": 1.369403464708884e-05,
+      "loss": 0.9275,
+      "step": 538
+    },
+    {
+      "epoch": 1.3275862068965516,
+      "grad_norm": 1.3104691915896813,
+      "learning_rate": 1.3667382783285903e-05,
+      "loss": 0.7761,
+      "step": 539
+    },
+    {
+      "epoch": 1.3300492610837438,
+      "grad_norm": 1.243683495971976,
+      "learning_rate": 1.3640700787023465e-05,
+      "loss": 0.7895,
+      "step": 540
+    },
+    {
+      "epoch": 1.332512315270936,
+      "grad_norm": 1.159975802070526,
+      "learning_rate": 1.3613988877529844e-05,
+      "loss": 0.8894,
+      "step": 541
+    },
+    {
+      "epoch": 1.3349753694581281,
+      "grad_norm": 1.2019839635245768,
+      "learning_rate": 1.358724727427914e-05,
+      "loss": 0.748,
+      "step": 542
+    },
+    {
+      "epoch": 1.3374384236453203,
+      "grad_norm": 1.2207648910154842,
+      "learning_rate": 1.3560476196989422e-05,
+      "loss": 0.8096,
+      "step": 543
+    },
+    {
+      "epoch": 1.3399014778325122,
+      "grad_norm": 1.1793273477069977,
+      "learning_rate": 1.3533675865620937e-05,
+      "loss": 0.8036,
+      "step": 544
+    },
+    {
+      "epoch": 1.3423645320197044,
+      "grad_norm": 1.238454138783358,
+      "learning_rate": 1.3506846500374285e-05,
+      "loss": 0.8674,
+      "step": 545
+    },
+    {
+      "epoch": 1.3448275862068966,
+      "grad_norm": 1.1648860296895283,
+      "learning_rate": 1.3479988321688619e-05,
+      "loss": 0.7865,
+      "step": 546
+    },
+    {
+      "epoch": 1.3472906403940887,
+      "grad_norm": 1.3201781887381099,
+      "learning_rate": 1.345310155023984e-05,
+      "loss": 0.7725,
+      "step": 547
+    },
+    {
+      "epoch": 1.3497536945812807,
+      "grad_norm": 1.233172978252316,
+      "learning_rate": 1.3426186406938769e-05,
+      "loss": 0.7894,
+      "step": 548
+    },
+    {
+      "epoch": 1.3522167487684729,
+      "grad_norm": 1.3015241821070775,
+      "learning_rate": 1.3399243112929341e-05,
+      "loss": 0.8477,
+      "step": 549
+    },
+    {
+      "epoch": 1.354679802955665,
+      "grad_norm": 1.1360609110835636,
+      "learning_rate": 1.337227188958679e-05,
+      "loss": 0.7782,
+      "step": 550
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 1.239717775421825,
+      "learning_rate": 1.3345272958515825e-05,
+      "loss": 0.8363,
+      "step": 551
+    },
+    {
+      "epoch": 1.3596059113300494,
+      "grad_norm": 1.1884708952944945,
+      "learning_rate": 1.3318246541548812e-05,
+      "loss": 0.7989,
+      "step": 552
+    },
+    {
+      "epoch": 1.3620689655172413,
+      "grad_norm": 1.1971749649447105,
+      "learning_rate": 1.3291192860743951e-05,
+      "loss": 0.8045,
+      "step": 553
+    },
+    {
+      "epoch": 1.3645320197044335,
+      "grad_norm": 1.2046416000547713,
+      "learning_rate": 1.3264112138383445e-05,
+      "loss": 0.812,
+      "step": 554
+    },
+    {
+      "epoch": 1.3669950738916257,
+      "grad_norm": 1.483500253022278,
+      "learning_rate": 1.3237004596971687e-05,
+      "loss": 0.7654,
+      "step": 555
+    },
+    {
+      "epoch": 1.3694581280788176,
+      "grad_norm": 1.26823767855671,
+      "learning_rate": 1.3209870459233422e-05,
+      "loss": 0.8295,
+      "step": 556
+    },
+    {
+      "epoch": 1.3719211822660098,
+      "grad_norm": 1.2613408139731488,
+      "learning_rate": 1.3182709948111921e-05,
+      "loss": 0.832,
+      "step": 557
+    },
+    {
+      "epoch": 1.374384236453202,
+      "grad_norm": 1.4258160312538462,
+      "learning_rate": 1.315552328676714e-05,
+      "loss": 0.8343,
+      "step": 558
+    },
+    {
+      "epoch": 1.3768472906403941,
+      "grad_norm": 1.2572159389054705,
+      "learning_rate": 1.3128310698573904e-05,
+      "loss": 0.8219,
+      "step": 559
+    },
+    {
+      "epoch": 1.3793103448275863,
+      "grad_norm": 1.3652002769654354,
+      "learning_rate": 1.3101072407120056e-05,
+      "loss": 0.8296,
+      "step": 560
+    },
+    {
+      "epoch": 1.3817733990147782,
+      "grad_norm": 1.2118828805605262,
+      "learning_rate": 1.3073808636204628e-05,
+      "loss": 0.8171,
+      "step": 561
+    },
+    {
+      "epoch": 1.3842364532019704,
+      "grad_norm": 1.0587624278234267,
+      "learning_rate": 1.3046519609836002e-05,
+      "loss": 0.7874,
+      "step": 562
+    },
+    {
+      "epoch": 1.3866995073891626,
+      "grad_norm": 1.3329375643263082,
+      "learning_rate": 1.3019205552230058e-05,
+      "loss": 0.7417,
+      "step": 563
+    },
+    {
+      "epoch": 1.3891625615763548,
+      "grad_norm": 1.1031096935561104,
+      "learning_rate": 1.2991866687808355e-05,
+      "loss": 0.8319,
+      "step": 564
+    },
+    {
+      "epoch": 1.3916256157635467,
+      "grad_norm": 1.2095226759533384,
+      "learning_rate": 1.2964503241196258e-05,
+      "loss": 0.7734,
+      "step": 565
+    },
+    {
+      "epoch": 1.3940886699507389,
+      "grad_norm": 1.1731433362549062,
+      "learning_rate": 1.2937115437221119e-05,
+      "loss": 0.7528,
+      "step": 566
+    },
+    {
+      "epoch": 1.396551724137931,
+      "grad_norm": 1.2621437514360003,
+      "learning_rate": 1.290970350091042e-05,
+      "loss": 0.8612,
+      "step": 567
+    },
+    {
+      "epoch": 1.3990147783251232,
+      "grad_norm": 1.2047222483728302,
+      "learning_rate": 1.2882267657489908e-05,
+      "loss": 0.8118,
+      "step": 568
+    },
+    {
+      "epoch": 1.4014778325123154,
+      "grad_norm": 1.1955457472270847,
+      "learning_rate": 1.2854808132381778e-05,
+      "loss": 0.7544,
+      "step": 569
+    },
+    {
+      "epoch": 1.4039408866995073,
+      "grad_norm": 1.1779805142874884,
+      "learning_rate": 1.2827325151202783e-05,
+      "loss": 0.9057,
+      "step": 570
+    },
+    {
+      "epoch": 1.4064039408866995,
+      "grad_norm": 1.2035470197594993,
+      "learning_rate": 1.2799818939762411e-05,
+      "loss": 0.8084,
+      "step": 571
+    },
+    {
+      "epoch": 1.4088669950738917,
+      "grad_norm": 1.2170557325385891,
+      "learning_rate": 1.2772289724061015e-05,
+      "loss": 0.8076,
+      "step": 572
+    },
+    {
+      "epoch": 1.4113300492610836,
+      "grad_norm": 1.24883879131097,
+      "learning_rate": 1.2744737730287961e-05,
+      "loss": 0.8122,
+      "step": 573
+    },
+    {
+      "epoch": 1.4137931034482758,
+      "grad_norm": 1.2582272148017177,
+      "learning_rate": 1.2717163184819761e-05,
+      "loss": 0.7615,
+      "step": 574
+    },
+    {
+      "epoch": 1.416256157635468,
+      "grad_norm": 1.2360932667952746,
+      "learning_rate": 1.2689566314218229e-05,
+      "loss": 0.8374,
+      "step": 575
+    },
+    {
+      "epoch": 1.4187192118226601,
+      "grad_norm": 1.2678887173349591,
+      "learning_rate": 1.2661947345228593e-05,
+      "loss": 0.7963,
+      "step": 576
+    },
+    {
+      "epoch": 1.4211822660098523,
+      "grad_norm": 1.1823971178974648,
+      "learning_rate": 1.2634306504777669e-05,
+      "loss": 0.8145,
+      "step": 577
+    },
+    {
+      "epoch": 1.4236453201970443,
+      "grad_norm": 1.229572078421072,
+      "learning_rate": 1.2606644019971967e-05,
+      "loss": 0.803,
+      "step": 578
+    },
+    {
+      "epoch": 1.4261083743842364,
+      "grad_norm": 1.2782034591059763,
+      "learning_rate": 1.257896011809583e-05,
+      "loss": 0.8445,
+      "step": 579
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 1.2226128616190663,
+      "learning_rate": 1.255125502660958e-05,
+      "loss": 0.7574,
+      "step": 580
+    },
+    {
+      "epoch": 1.4310344827586206,
+      "grad_norm": 1.420285940394074,
+      "learning_rate": 1.2523528973147631e-05,
+      "loss": 0.76,
+      "step": 581
+    },
+    {
+      "epoch": 1.4334975369458127,
+      "grad_norm": 1.2613007411069364,
+      "learning_rate": 1.2495782185516638e-05,
+      "loss": 0.8747,
+      "step": 582
+    },
+    {
+      "epoch": 1.435960591133005,
+      "grad_norm": 1.1822960919312775,
+      "learning_rate": 1.2468014891693603e-05,
+      "loss": 0.8049,
+      "step": 583
+    },
+    {
+      "epoch": 1.438423645320197,
+      "grad_norm": 1.2101681634206956,
+      "learning_rate": 1.2440227319824024e-05,
+      "loss": 0.8206,
+      "step": 584
+    },
+    {
+      "epoch": 1.4408866995073892,
+      "grad_norm": 1.246552003408417,
+      "learning_rate": 1.2412419698220002e-05,
+      "loss": 0.8058,
+      "step": 585
+    },
+    {
+      "epoch": 1.4433497536945814,
+      "grad_norm": 1.252003589868731,
+      "learning_rate": 1.2384592255358385e-05,
+      "loss": 0.7911,
+      "step": 586
+    },
+    {
+      "epoch": 1.4458128078817734,
+      "grad_norm": 1.3125874626445564,
+      "learning_rate": 1.2356745219878865e-05,
+      "loss": 0.8202,
+      "step": 587
+    },
+    {
+      "epoch": 1.4482758620689655,
+      "grad_norm": 1.0989266100217372,
+      "learning_rate": 1.2328878820582122e-05,
+      "loss": 0.7612,
+      "step": 588
+    },
+    {
+      "epoch": 1.4507389162561577,
+      "grad_norm": 1.2101322573111324,
+      "learning_rate": 1.2300993286427937e-05,
+      "loss": 0.7667,
+      "step": 589
+    },
+    {
+      "epoch": 1.4532019704433496,
+      "grad_norm": 1.2465878646002861,
+      "learning_rate": 1.2273088846533303e-05,
+      "loss": 0.8296,
+      "step": 590
+    },
+    {
+      "epoch": 1.4556650246305418,
+      "grad_norm": 1.0874134160794056,
+      "learning_rate": 1.2245165730170556e-05,
+      "loss": 0.7107,
+      "step": 591
+    },
+    {
+      "epoch": 1.458128078817734,
+      "grad_norm": 1.284251767300427,
+      "learning_rate": 1.2217224166765478e-05,
+      "loss": 0.8011,
+      "step": 592
+    },
+    {
+      "epoch": 1.4605911330049262,
+      "grad_norm": 1.4418013529990912,
+      "learning_rate": 1.2189264385895422e-05,
+      "loss": 0.8479,
+      "step": 593
+    },
+    {
+      "epoch": 1.4630541871921183,
+      "grad_norm": 1.128848117268779,
+      "learning_rate": 1.216128661728742e-05,
+      "loss": 0.8192,
+      "step": 594
+    },
+    {
+      "epoch": 1.4655172413793103,
+      "grad_norm": 1.2677008006682624,
+      "learning_rate": 1.2133291090816298e-05,
+      "loss": 0.8309,
+      "step": 595
+    },
+    {
+      "epoch": 1.4679802955665024,
+      "grad_norm": 1.1558220045437095,
+      "learning_rate": 1.2105278036502787e-05,
+      "loss": 0.7888,
+      "step": 596
+    },
+    {
+      "epoch": 1.4704433497536946,
+      "grad_norm": 1.2794639420073919,
+      "learning_rate": 1.207724768451164e-05,
+      "loss": 0.8222,
+      "step": 597
+    },
+    {
+      "epoch": 1.4729064039408866,
+      "grad_norm": 2.586562988605663,
+      "learning_rate": 1.204920026514971e-05,
+      "loss": 0.8208,
+      "step": 598
+    },
+    {
+      "epoch": 1.4753694581280787,
+      "grad_norm": 1.395439984519839,
+      "learning_rate": 1.202113600886411e-05,
+      "loss": 0.759,
+      "step": 599
+    },
+    {
+      "epoch": 1.477832512315271,
+      "grad_norm": 1.275618575143965,
+      "learning_rate": 1.1993055146240273e-05,
+      "loss": 0.8169,
+      "step": 600
+    },
+    {
+      "epoch": 1.480295566502463,
+      "grad_norm": 1.2638902431289813,
+      "learning_rate": 1.1964957908000084e-05,
+      "loss": 0.7543,
+      "step": 601
+    },
+    {
+      "epoch": 1.4827586206896552,
+      "grad_norm": 1.407480083344989,
+      "learning_rate": 1.1936844524999966e-05,
+      "loss": 0.864,
+      "step": 602
+    },
+    {
+      "epoch": 1.4852216748768474,
+      "grad_norm": 1.278431047578547,
+      "learning_rate": 1.1908715228229007e-05,
+      "loss": 0.8488,
+      "step": 603
+    },
+    {
+      "epoch": 1.4876847290640394,
+      "grad_norm": 1.3204794332396885,
+      "learning_rate": 1.1880570248807033e-05,
+      "loss": 0.8228,
+      "step": 604
+    },
+    {
+      "epoch": 1.4901477832512315,
+      "grad_norm": 1.190588198115564,
+      "learning_rate": 1.1852409817982732e-05,
+      "loss": 0.7722,
+      "step": 605
+    },
+    {
+      "epoch": 1.4926108374384237,
+      "grad_norm": 1.2204957343358656,
+      "learning_rate": 1.1824234167131748e-05,
+      "loss": 0.8303,
+      "step": 606
+    },
+    {
+      "epoch": 1.4950738916256157,
+      "grad_norm": 1.2067469783703244,
+      "learning_rate": 1.1796043527754775e-05,
+      "loss": 0.8247,
+      "step": 607
+    },
+    {
+      "epoch": 1.4975369458128078,
+      "grad_norm": 1.088386408975514,
+      "learning_rate": 1.1767838131475654e-05,
+      "loss": 0.8317,
+      "step": 608
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.246678038985525,
+      "learning_rate": 1.1739618210039476e-05,
+      "loss": 0.8387,
+      "step": 609
+    },
+    {
+      "epoch": 1.5024630541871922,
+      "grad_norm": 1.3081553621591417,
+      "learning_rate": 1.171138399531068e-05,
+      "loss": 0.8095,
+      "step": 610
+    },
+    {
+      "epoch": 1.5049261083743843,
+      "grad_norm": 1.3480579606791177,
+      "learning_rate": 1.1683135719271143e-05,
+      "loss": 0.7528,
+      "step": 611
+    },
+    {
+      "epoch": 1.5073891625615765,
+      "grad_norm": 1.215326489274603,
+      "learning_rate": 1.1654873614018266e-05,
+      "loss": 0.708,
+      "step": 612
+    },
+    {
+      "epoch": 1.5098522167487685,
+      "grad_norm": 1.3029442751043114,
+      "learning_rate": 1.1626597911763085e-05,
+      "loss": 0.8625,
+      "step": 613
+    },
+    {
+      "epoch": 1.5123152709359606,
+      "grad_norm": 1.104253255699839,
+      "learning_rate": 1.1598308844828348e-05,
+      "loss": 0.777,
+      "step": 614
+    },
+    {
+      "epoch": 1.5147783251231526,
+      "grad_norm": 1.1975148176578576,
+      "learning_rate": 1.1570006645646614e-05,
+      "loss": 0.7645,
+      "step": 615
+    },
+    {
+      "epoch": 1.5172413793103448,
+      "grad_norm": 1.1057283346083475,
+      "learning_rate": 1.1541691546758343e-05,
+      "loss": 0.7522,
+      "step": 616
+    },
+    {
+      "epoch": 1.519704433497537,
+      "grad_norm": 1.247641914108553,
+      "learning_rate": 1.1513363780809974e-05,
+      "loss": 0.7629,
+      "step": 617
+    },
+    {
+      "epoch": 1.522167487684729,
+      "grad_norm": 1.1767999538616116,
+      "learning_rate": 1.1485023580552039e-05,
+      "loss": 0.8078,
+      "step": 618
+    },
+    {
+      "epoch": 1.5246305418719213,
+      "grad_norm": 1.154080030079345,
+      "learning_rate": 1.145667117883722e-05,
+      "loss": 0.8732,
+      "step": 619
+    },
+    {
+      "epoch": 1.5270935960591134,
+      "grad_norm": 1.3640651689745824,
+      "learning_rate": 1.1428306808618456e-05,
+      "loss": 0.8185,
+      "step": 620
+    },
+    {
+      "epoch": 1.5295566502463054,
+      "grad_norm": 1.1533101313155347,
+      "learning_rate": 1.1399930702947025e-05,
+      "loss": 0.756,
+      "step": 621
+    },
+    {
+      "epoch": 1.5320197044334976,
+      "grad_norm": 1.2145801515793824,
+      "learning_rate": 1.1371543094970624e-05,
+      "loss": 0.7755,
+      "step": 622
+    },
+    {
+      "epoch": 1.5344827586206895,
+      "grad_norm": 1.337484122565318,
+      "learning_rate": 1.1343144217931457e-05,
+      "loss": 0.8264,
+      "step": 623
+    },
+    {
+      "epoch": 1.5369458128078817,
+      "grad_norm": 1.247835813056646,
+      "learning_rate": 1.131473430516432e-05,
+      "loss": 0.832,
+      "step": 624
+    },
+    {
+      "epoch": 1.5394088669950738,
+      "grad_norm": 1.1929733226316261,
+      "learning_rate": 1.1286313590094686e-05,
+      "loss": 0.7755,
+      "step": 625
+    },
+    {
+      "epoch": 1.541871921182266,
+      "grad_norm": 1.1688725473547292,
+      "learning_rate": 1.1257882306236776e-05,
+      "loss": 0.8177,
+      "step": 626
+    },
+    {
+      "epoch": 1.5443349753694582,
+      "grad_norm": 1.2425031440894645,
+      "learning_rate": 1.1229440687191649e-05,
+      "loss": 0.8898,
+      "step": 627
+    },
+    {
+      "epoch": 1.5467980295566504,
+      "grad_norm": 1.1812642152122668,
+      "learning_rate": 1.1200988966645286e-05,
+      "loss": 0.7424,
+      "step": 628
+    },
+    {
+      "epoch": 1.5492610837438425,
+      "grad_norm": 1.1886090069259614,
+      "learning_rate": 1.1172527378366664e-05,
+      "loss": 0.8885,
+      "step": 629
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 1.1149791418784638,
+      "learning_rate": 1.1144056156205834e-05,
+      "loss": 0.8056,
+      "step": 630
+    },
+    {
+      "epoch": 1.5541871921182266,
+      "grad_norm": 1.3036055961654982,
+      "learning_rate": 1.1115575534092003e-05,
+      "loss": 0.812,
+      "step": 631
+    },
+    {
+      "epoch": 1.5566502463054186,
+      "grad_norm": 1.252057147056369,
+      "learning_rate": 1.1087085746031612e-05,
+      "loss": 0.7681,
+      "step": 632
+    },
+    {
+      "epoch": 1.5591133004926108,
+      "grad_norm": 1.3566281419965618,
+      "learning_rate": 1.1058587026106413e-05,
+      "loss": 0.7321,
+      "step": 633
+    },
+    {
+      "epoch": 1.561576354679803,
+      "grad_norm": 1.1725228108543888,
+      "learning_rate": 1.1030079608471544e-05,
+      "loss": 0.7562,
+      "step": 634
+    },
+    {
+      "epoch": 1.564039408866995,
+      "grad_norm": 1.133276219658579,
+      "learning_rate": 1.1001563727353611e-05,
+      "loss": 0.6928,
+      "step": 635
+    },
+    {
+      "epoch": 1.5665024630541873,
+      "grad_norm": 1.2399506492177108,
+      "learning_rate": 1.0973039617048748e-05,
+      "loss": 0.8037,
+      "step": 636
+    },
+    {
+      "epoch": 1.5689655172413794,
+      "grad_norm": 1.1452906045733329,
+      "learning_rate": 1.0944507511920715e-05,
+      "loss": 0.7353,
+      "step": 637
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 1.2357644096150233,
+      "learning_rate": 1.091596764639895e-05,
+      "loss": 0.7519,
+      "step": 638
+    },
+    {
+      "epoch": 1.5738916256157636,
+      "grad_norm": 1.1687791749301346,
+      "learning_rate": 1.0887420254976661e-05,
+      "loss": 0.7385,
+      "step": 639
+    },
+    {
+      "epoch": 1.5763546798029555,
+      "grad_norm": 1.2588415734204763,
+      "learning_rate": 1.0858865572208892e-05,
+      "loss": 0.8075,
+      "step": 640
+    },
+    {
+      "epoch": 1.5788177339901477,
+      "grad_norm": 1.3002483204113997,
+      "learning_rate": 1.0830303832710584e-05,
+      "loss": 0.8216,
+      "step": 641
+    },
+    {
+      "epoch": 1.5812807881773399,
+      "grad_norm": 1.4354003831505446,
+      "learning_rate": 1.080173527115467e-05,
+      "loss": 0.8193,
+      "step": 642
+    },
+    {
+      "epoch": 1.583743842364532,
+      "grad_norm": 1.2282360832315742,
+      "learning_rate": 1.0773160122270127e-05,
+      "loss": 0.7992,
+      "step": 643
+    },
+    {
+      "epoch": 1.5862068965517242,
+      "grad_norm": 1.2132237663333358,
+      "learning_rate": 1.0744578620840065e-05,
+      "loss": 0.7955,
+      "step": 644
+    },
+    {
+      "epoch": 1.5886699507389164,
+      "grad_norm": 1.3025822357605237,
+      "learning_rate": 1.071599100169978e-05,
+      "loss": 0.8444,
+      "step": 645
+    },
+    {
+      "epoch": 1.5911330049261085,
+      "grad_norm": 1.1876807828120155,
+      "learning_rate": 1.0687397499734842e-05,
+      "loss": 0.8145,
+      "step": 646
+    },
+    {
+      "epoch": 1.5935960591133005,
+      "grad_norm": 1.2750470339757678,
+      "learning_rate": 1.0658798349879144e-05,
+      "loss": 0.7826,
+      "step": 647
+    },
+    {
+      "epoch": 1.5960591133004927,
+      "grad_norm": 1.2533975680525946,
+      "learning_rate": 1.0630193787112994e-05,
+      "loss": 0.7337,
+      "step": 648
+    },
+    {
+      "epoch": 1.5985221674876846,
+      "grad_norm": 1.2909643766125924,
+      "learning_rate": 1.0601584046461173e-05,
+      "loss": 0.8309,
+      "step": 649
+    },
+    {
+      "epoch": 1.6009852216748768,
+      "grad_norm": 1.249588707970312,
+      "learning_rate": 1.0572969362991e-05,
+      "loss": 0.7868,
+      "step": 650
+    },
+    {
+      "epoch": 1.603448275862069,
+      "grad_norm": 1.304369904929036,
+      "learning_rate": 1.0544349971810413e-05,
+      "loss": 0.7627,
+      "step": 651
+    },
+    {
+      "epoch": 1.6059113300492611,
+      "grad_norm": 1.1585386153804442,
+      "learning_rate": 1.0515726108066025e-05,
+      "loss": 0.8094,
+      "step": 652
+    },
+    {
+      "epoch": 1.6083743842364533,
+      "grad_norm": 1.0747780193643772,
+      "learning_rate": 1.0487098006941197e-05,
+      "loss": 0.724,
+      "step": 653
+    },
+    {
+      "epoch": 1.6108374384236455,
+      "grad_norm": 1.2348917591677222,
+      "learning_rate": 1.0458465903654107e-05,
+      "loss": 0.8228,
+      "step": 654
+    },
+    {
+      "epoch": 1.6133004926108374,
+      "grad_norm": 1.3812549245755192,
+      "learning_rate": 1.0429830033455821e-05,
+      "loss": 0.776,
+      "step": 655
+    },
+    {
+      "epoch": 1.6157635467980296,
+      "grad_norm": 1.2320624793236556,
+      "learning_rate": 1.0401190631628348e-05,
+      "loss": 0.7745,
+      "step": 656
+    },
+    {
+      "epoch": 1.6182266009852215,
+      "grad_norm": 1.3097067583361632,
+      "learning_rate": 1.037254793348272e-05,
+      "loss": 0.8509,
+      "step": 657
+    },
+    {
+      "epoch": 1.6206896551724137,
+      "grad_norm": 1.1886890882736885,
+      "learning_rate": 1.034390217435704e-05,
+      "loss": 0.7648,
+      "step": 658
+    },
+    {
+      "epoch": 1.6231527093596059,
+      "grad_norm": 1.2406532841684434,
+      "learning_rate": 1.031525358961458e-05,
+      "loss": 0.7229,
+      "step": 659
+    },
+    {
+      "epoch": 1.625615763546798,
+      "grad_norm": 1.1866637347128004,
+      "learning_rate": 1.0286602414641818e-05,
+      "loss": 0.7521,
+      "step": 660
+    },
+    {
+      "epoch": 1.6280788177339902,
+      "grad_norm": 1.3094453688037277,
+      "learning_rate": 1.0257948884846507e-05,
+      "loss": 0.787,
+      "step": 661
+    },
+    {
+      "epoch": 1.6305418719211824,
+      "grad_norm": 1.1461382835800795,
+      "learning_rate": 1.0229293235655768e-05,
+      "loss": 0.7752,
+      "step": 662
+    },
+    {
+      "epoch": 1.6330049261083743,
+      "grad_norm": 1.1057948517781475,
+      "learning_rate": 1.0200635702514115e-05,
+      "loss": 0.7768,
+      "step": 663
+    },
+    {
+      "epoch": 1.6354679802955665,
+      "grad_norm": 1.3130929606549313,
+      "learning_rate": 1.0171976520881552e-05,
+      "loss": 0.7752,
+      "step": 664
+    },
+    {
+      "epoch": 1.6379310344827587,
+      "grad_norm": 1.253600846885877,
+      "learning_rate": 1.0143315926231625e-05,
+      "loss": 0.861,
+      "step": 665
+    },
+    {
+      "epoch": 1.6403940886699506,
+      "grad_norm": 1.3659343098585952,
+      "learning_rate": 1.011465415404949e-05,
+      "loss": 0.8735,
+      "step": 666
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 1.2506223559792482,
+      "learning_rate": 1.0085991439829981e-05,
+      "loss": 0.8135,
+      "step": 667
+    },
+    {
+      "epoch": 1.645320197044335,
+      "grad_norm": 1.2158152139702334,
+      "learning_rate": 1.005732801907567e-05,
+      "loss": 0.8085,
+      "step": 668
+    },
+    {
+      "epoch": 1.6477832512315271,
+      "grad_norm": 1.1629809276922662,
+      "learning_rate": 1.0028664127294924e-05,
+      "loss": 0.8169,
+      "step": 669
+    },
+    {
+      "epoch": 1.6502463054187193,
+      "grad_norm": 1.2721403571520649,
+      "learning_rate": 1e-05,
+      "loss": 0.7487,
+      "step": 670
+    },
+    {
+      "epoch": 1.6527093596059115,
+      "grad_norm": 1.1451388884519247,
+      "learning_rate": 9.971335872705076e-06,
+      "loss": 0.816,
+      "step": 671
+    },
+    {
+      "epoch": 1.6551724137931034,
+      "grad_norm": 1.2378201272908131,
+      "learning_rate": 9.942671980924336e-06,
+      "loss": 0.8639,
+      "step": 672
+    },
+    {
+      "epoch": 1.6576354679802956,
+      "grad_norm": 1.1150340954310294,
+      "learning_rate": 9.91400856017002e-06,
+      "loss": 0.7564,
+      "step": 673
+    },
+    {
+      "epoch": 1.6600985221674875,
+      "grad_norm": 1.197759728586669,
+      "learning_rate": 9.88534584595051e-06,
+      "loss": 0.7742,
+      "step": 674
+    },
+    {
+      "epoch": 1.6625615763546797,
+      "grad_norm": 1.15635948693778,
+      "learning_rate": 9.856684073768378e-06,
+      "loss": 0.7488,
+      "step": 675
+    },
+    {
+      "epoch": 1.6650246305418719,
+      "grad_norm": 1.2703505000100057,
+      "learning_rate": 9.82802347911845e-06,
+      "loss": 0.8409,
+      "step": 676
+    },
+    {
+      "epoch": 1.667487684729064,
+      "grad_norm": 1.283316271945116,
+      "learning_rate": 9.799364297485889e-06,
+      "loss": 0.802,
+      "step": 677
+    },
+    {
+      "epoch": 1.6699507389162562,
+      "grad_norm": 1.2653379703519836,
+      "learning_rate": 9.770706764344235e-06,
+      "loss": 0.787,
+      "step": 678
+    },
+    {
+      "epoch": 1.6724137931034484,
+      "grad_norm": 1.1968302845383292,
+      "learning_rate": 9.742051115153494e-06,
+      "loss": 0.8629,
+      "step": 679
+    },
+    {
+      "epoch": 1.6748768472906403,
+      "grad_norm": 1.2912384164148434,
+      "learning_rate": 9.713397585358189e-06,
+      "loss": 0.8705,
+      "step": 680
+    },
+    {
+      "epoch": 1.6773399014778325,
+      "grad_norm": 1.2397918632625542,
+      "learning_rate": 9.684746410385423e-06,
+      "loss": 0.7341,
+      "step": 681
+    },
+    {
+      "epoch": 1.6798029556650245,
+      "grad_norm": 1.3277884112924991,
+      "learning_rate": 9.65609782564296e-06,
+      "loss": 0.8532,
+      "step": 682
+    },
+    {
+      "epoch": 1.6822660098522166,
+      "grad_norm": 1.28238548578091,
+      "learning_rate": 9.627452066517287e-06,
+      "loss": 0.8299,
+      "step": 683
+    },
+    {
+      "epoch": 1.6847290640394088,
+      "grad_norm": 1.3869509650283263,
+      "learning_rate": 9.598809368371656e-06,
+      "loss": 0.8729,
+      "step": 684
+    },
+    {
+      "epoch": 1.687192118226601,
+      "grad_norm": 1.2710993545680642,
+      "learning_rate": 9.57016996654418e-06,
+      "loss": 0.8646,
+      "step": 685
+    },
+    {
+      "epoch": 1.6896551724137931,
+      "grad_norm": 1.156095752221962,
+      "learning_rate": 9.541534096345896e-06,
+      "loss": 0.8558,
+      "step": 686
+    },
+    {
+      "epoch": 1.6921182266009853,
+      "grad_norm": 1.201653063155239,
+      "learning_rate": 9.512901993058806e-06,
+      "loss": 0.8423,
+      "step": 687
+    },
+    {
+      "epoch": 1.6945812807881775,
+      "grad_norm": 1.2880254562095468,
+      "learning_rate": 9.484273891933982e-06,
+      "loss": 0.8183,
+      "step": 688
+    },
+    {
+      "epoch": 1.6970443349753694,
+      "grad_norm": 1.2926407596817835,
+      "learning_rate": 9.45565002818959e-06,
+      "loss": 0.7902,
+      "step": 689
+    },
+    {
+      "epoch": 1.6995073891625616,
+      "grad_norm": 1.216102218118117,
+      "learning_rate": 9.427030637009002e-06,
+      "loss": 0.8244,
+      "step": 690
+    },
+    {
+      "epoch": 1.7019704433497536,
+      "grad_norm": 1.2445243112067728,
+      "learning_rate": 9.398415953538832e-06,
+      "loss": 0.7736,
+      "step": 691
+    },
+    {
+      "epoch": 1.7044334975369457,
+      "grad_norm": 1.3549158029332782,
+      "learning_rate": 9.369806212887008e-06,
+      "loss": 0.8151,
+      "step": 692
+    },
+    {
+      "epoch": 1.706896551724138,
+      "grad_norm": 1.1910343277545914,
+      "learning_rate": 9.341201650120857e-06,
+      "loss": 0.7932,
+      "step": 693
+    },
+    {
+      "epoch": 1.70935960591133,
+      "grad_norm": 1.3045491584940332,
+      "learning_rate": 9.312602500265162e-06,
+      "loss": 0.7487,
+      "step": 694
+    },
+    {
+      "epoch": 1.7118226600985222,
+      "grad_norm": 1.1737401234269127,
+      "learning_rate": 9.284008998300221e-06,
+      "loss": 0.8448,
+      "step": 695
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 1.2426206685194015,
+      "learning_rate": 9.255421379159935e-06,
+      "loss": 0.7532,
+      "step": 696
+    },
+    {
+      "epoch": 1.7167487684729064,
+      "grad_norm": 1.26162246950879,
+      "learning_rate": 9.226839877729875e-06,
+      "loss": 0.7648,
+      "step": 697
+    },
+    {
+      "epoch": 1.7192118226600985,
+      "grad_norm": 1.1970069692977665,
+      "learning_rate": 9.198264728845332e-06,
+      "loss": 0.7629,
+      "step": 698
+    },
+    {
+      "epoch": 1.7216748768472905,
+      "grad_norm": 1.4179536172358826,
+      "learning_rate": 9.16969616728942e-06,
+      "loss": 0.7358,
+      "step": 699
+    },
+    {
+      "epoch": 1.7241379310344827,
+      "grad_norm": 1.3060032953802012,
+      "learning_rate": 9.14113442779111e-06,
+      "loss": 0.8003,
+      "step": 700
+    },
+    {
+      "epoch": 1.7266009852216748,
+      "grad_norm": 1.3638840356512967,
+      "learning_rate": 9.112579745023339e-06,
+      "loss": 0.8707,
+      "step": 701
+    },
+    {
+      "epoch": 1.729064039408867,
+      "grad_norm": 1.274103423512333,
+      "learning_rate": 9.084032353601053e-06,
+      "loss": 0.8248,
+      "step": 702
+    },
+    {
+      "epoch": 1.7315270935960592,
+      "grad_norm": 1.3121386439291902,
+      "learning_rate": 9.055492488079288e-06,
+      "loss": 0.7968,
+      "step": 703
+    },
+    {
+      "epoch": 1.7339901477832513,
+      "grad_norm": 1.2623503907948912,
+      "learning_rate": 9.026960382951253e-06,
+      "loss": 0.8412,
+      "step": 704
+    },
+    {
+      "epoch": 1.7364532019704435,
+      "grad_norm": 1.1912986831884471,
+      "learning_rate": 8.998436272646394e-06,
+      "loss": 0.8316,
+      "step": 705
+    },
+    {
+      "epoch": 1.7389162561576355,
+      "grad_norm": 1.2132466602982988,
+      "learning_rate": 8.969920391528459e-06,
+      "loss": 0.8217,
+      "step": 706
+    },
+    {
+      "epoch": 1.7413793103448276,
+      "grad_norm": 1.2357849567091015,
+      "learning_rate": 8.941412973893594e-06,
+      "loss": 0.8701,
+      "step": 707
+    },
+    {
+      "epoch": 1.7438423645320196,
+      "grad_norm": 1.275012627707386,
+      "learning_rate": 8.912914253968391e-06,
+      "loss": 0.8012,
+      "step": 708
+    },
+    {
+      "epoch": 1.7463054187192117,
+      "grad_norm": 1.174149099633243,
+      "learning_rate": 8.884424465907999e-06,
+      "loss": 0.8607,
+      "step": 709
+    },
+    {
+      "epoch": 1.748768472906404,
+      "grad_norm": 1.6184748441516579,
+      "learning_rate": 8.855943843794171e-06,
+      "loss": 0.9043,
+      "step": 710
+    },
+    {
+      "epoch": 1.751231527093596,
+      "grad_norm": 1.2547177929861402,
+      "learning_rate": 8.827472621633338e-06,
+      "loss": 0.8844,
+      "step": 711
+    },
+    {
+      "epoch": 1.7536945812807883,
+      "grad_norm": 1.2999933279306415,
+      "learning_rate": 8.799011033354716e-06,
+      "loss": 0.7509,
+      "step": 712
+    },
+    {
+      "epoch": 1.7561576354679804,
+      "grad_norm": 1.3512966798664634,
+      "learning_rate": 8.770559312808356e-06,
+      "loss": 0.7192,
+      "step": 713
+    },
+    {
+      "epoch": 1.7586206896551724,
+      "grad_norm": 1.1979086115469524,
+      "learning_rate": 8.742117693763229e-06,
+      "loss": 0.7903,
+      "step": 714
+    },
+    {
+      "epoch": 1.7610837438423645,
+      "grad_norm": 1.2591818706386428,
+      "learning_rate": 8.713686409905314e-06,
+      "loss": 0.8037,
+      "step": 715
+    },
+    {
+      "epoch": 1.7635467980295565,
+      "grad_norm": 1.1869529668782393,
+      "learning_rate": 8.685265694835681e-06,
+      "loss": 0.7899,
+      "step": 716
+    },
+    {
+      "epoch": 1.7660098522167487,
+      "grad_norm": 1.2087344895859873,
+      "learning_rate": 8.656855782068546e-06,
+      "loss": 0.8248,
+      "step": 717
+    },
+    {
+      "epoch": 1.7684729064039408,
+      "grad_norm": 1.4091522785146284,
+      "learning_rate": 8.628456905029383e-06,
+      "loss": 0.7503,
+      "step": 718
+    },
+    {
+      "epoch": 1.770935960591133,
+      "grad_norm": 1.1773230497063034,
+      "learning_rate": 8.600069297052978e-06,
+      "loss": 0.8082,
+      "step": 719
+    },
+    {
+      "epoch": 1.7733990147783252,
+      "grad_norm": 1.1988553170569236,
+      "learning_rate": 8.571693191381545e-06,
+      "loss": 0.8413,
+      "step": 720
+    },
+    {
+      "epoch": 1.7758620689655173,
+      "grad_norm": 1.4854104696878407,
+      "learning_rate": 8.543328821162784e-06,
+      "loss": 0.805,
+      "step": 721
+    },
+    {
+      "epoch": 1.7783251231527095,
+      "grad_norm": 1.3600020004005307,
+      "learning_rate": 8.514976419447963e-06,
+      "loss": 0.8134,
+      "step": 722
+    },
+    {
+      "epoch": 1.7807881773399015,
+      "grad_norm": 1.3016359654240084,
+      "learning_rate": 8.486636219190027e-06,
+      "loss": 0.7808,
+      "step": 723
+    },
+    {
+      "epoch": 1.7832512315270936,
+      "grad_norm": 1.4123020489150648,
+      "learning_rate": 8.458308453241664e-06,
+      "loss": 0.8469,
+      "step": 724
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 1.2340962541899692,
+      "learning_rate": 8.429993354353389e-06,
+      "loss": 0.7705,
+      "step": 725
+    },
+    {
+      "epoch": 1.7881773399014778,
+      "grad_norm": 1.345461310154665,
+      "learning_rate": 8.401691155171654e-06,
+      "loss": 0.7969,
+      "step": 726
+    },
+    {
+      "epoch": 1.79064039408867,
+      "grad_norm": 1.2824781625684334,
+      "learning_rate": 8.373402088236919e-06,
+      "loss": 0.8231,
+      "step": 727
+    },
+    {
+      "epoch": 1.793103448275862,
+      "grad_norm": 1.2880786724912396,
+      "learning_rate": 8.345126385981737e-06,
+      "loss": 0.7968,
+      "step": 728
+    },
+    {
+      "epoch": 1.7955665024630543,
+      "grad_norm": 1.3102693221194743,
+      "learning_rate": 8.316864280728862e-06,
+      "loss": 0.8681,
+      "step": 729
+    },
+    {
+      "epoch": 1.7980295566502464,
+      "grad_norm": 1.2738329969474977,
+      "learning_rate": 8.288616004689321e-06,
+      "loss": 0.7903,
+      "step": 730
+    },
+    {
+      "epoch": 1.8004926108374384,
+      "grad_norm": 1.2385329222417236,
+      "learning_rate": 8.260381789960524e-06,
+      "loss": 0.7438,
+      "step": 731
+    },
+    {
+      "epoch": 1.8029556650246306,
+      "grad_norm": 1.218177612092328,
+      "learning_rate": 8.23216186852435e-06,
+      "loss": 0.7472,
+      "step": 732
+    },
+    {
+      "epoch": 1.8054187192118225,
+      "grad_norm": 1.2120129147650411,
+      "learning_rate": 8.203956472245226e-06,
+      "loss": 0.8197,
+      "step": 733
+    },
+    {
+      "epoch": 1.8078817733990147,
+      "grad_norm": 1.2165040571597452,
+      "learning_rate": 8.175765832868252e-06,
+      "loss": 0.8366,
+      "step": 734
+    },
+    {
+      "epoch": 1.8103448275862069,
+      "grad_norm": 1.2998375662780255,
+      "learning_rate": 8.147590182017271e-06,
+      "loss": 0.7067,
+      "step": 735
+    },
+    {
+      "epoch": 1.812807881773399,
+      "grad_norm": 1.3898387357761792,
+      "learning_rate": 8.119429751192972e-06,
+      "loss": 0.7683,
+      "step": 736
+    },
+    {
+      "epoch": 1.8152709359605912,
+      "grad_norm": 1.1222228103474585,
+      "learning_rate": 8.091284771770995e-06,
+      "loss": 0.787,
+      "step": 737
+    },
+    {
+      "epoch": 1.8177339901477834,
+      "grad_norm": 1.2359381809546215,
+      "learning_rate": 8.063155475000037e-06,
+      "loss": 0.8227,
+      "step": 738
+    },
+    {
+      "epoch": 1.8201970443349755,
+      "grad_norm": 1.234570113587752,
+      "learning_rate": 8.035042091999921e-06,
+      "loss": 0.8029,
+      "step": 739
+    },
+    {
+      "epoch": 1.8226600985221675,
+      "grad_norm": 1.1813222913469996,
+      "learning_rate": 8.006944853759732e-06,
+      "loss": 0.8681,
+      "step": 740
+    },
+    {
+      "epoch": 1.8251231527093597,
+      "grad_norm": 1.2472125922928459,
+      "learning_rate": 7.978863991135894e-06,
+      "loss": 0.679,
+      "step": 741
+    },
+    {
+      "epoch": 1.8275862068965516,
+      "grad_norm": 1.2471563896518065,
+      "learning_rate": 7.950799734850292e-06,
+      "loss": 0.7532,
+      "step": 742
+    },
+    {
+      "epoch": 1.8300492610837438,
+      "grad_norm": 1.507755416817376,
+      "learning_rate": 7.922752315488367e-06,
+      "loss": 0.7974,
+      "step": 743
+    },
+    {
+      "epoch": 1.832512315270936,
+      "grad_norm": 1.320569346347928,
+      "learning_rate": 7.894721963497214e-06,
+      "loss": 0.8542,
+      "step": 744
+    },
+    {
+      "epoch": 1.8349753694581281,
+      "grad_norm": 1.2435703377857028,
+      "learning_rate": 7.866708909183702e-06,
+      "loss": 0.8112,
+      "step": 745
+    },
+    {
+      "epoch": 1.8374384236453203,
+      "grad_norm": 1.1787441094130966,
+      "learning_rate": 7.838713382712583e-06,
+      "loss": 0.7813,
+      "step": 746
+    },
+    {
+      "epoch": 1.8399014778325125,
+      "grad_norm": 1.2673174056554444,
+      "learning_rate": 7.810735614104581e-06,
+      "loss": 0.8036,
+      "step": 747
+    },
+    {
+      "epoch": 1.8423645320197044,
+      "grad_norm": 1.2382779286818382,
+      "learning_rate": 7.782775833234522e-06,
+      "loss": 0.7841,
+      "step": 748
+    },
+    {
+      "epoch": 1.8448275862068966,
+      "grad_norm": 1.3171967077329367,
+      "learning_rate": 7.754834269829447e-06,
+      "loss": 0.7876,
+      "step": 749
+    },
+    {
+      "epoch": 1.8472906403940885,
+      "grad_norm": 1.2095899405689357,
+      "learning_rate": 7.726911153466699e-06,
+      "loss": 0.7662,
+      "step": 750
+    },
+    {
+      "epoch": 1.8497536945812807,
+      "grad_norm": 1.2643242733519557,
+      "learning_rate": 7.699006713572068e-06,
+      "loss": 0.7129,
+      "step": 751
+    },
+    {
+      "epoch": 1.8522167487684729,
+      "grad_norm": 1.1614731016488853,
+      "learning_rate": 7.67112117941788e-06,
+      "loss": 0.8317,
+      "step": 752
+    },
+    {
+      "epoch": 1.854679802955665,
+      "grad_norm": 1.3044211749689816,
+      "learning_rate": 7.643254780121137e-06,
+      "loss": 0.8999,
+      "step": 753
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 1.2531731384620277,
+      "learning_rate": 7.615407744641618e-06,
+      "loss": 0.8214,
+      "step": 754
+    },
+    {
+      "epoch": 1.8596059113300494,
+      "grad_norm": 1.290177573912193,
+      "learning_rate": 7.58758030178e-06,
+      "loss": 0.7333,
+      "step": 755
+    },
+    {
+      "epoch": 1.8620689655172413,
+      "grad_norm": 1.3536272539617957,
+      "learning_rate": 7.559772680175979e-06,
+      "loss": 0.8637,
+      "step": 756
+    },
+    {
+      "epoch": 1.8645320197044335,
+      "grad_norm": 1.2127897309680038,
+      "learning_rate": 7.531985108306401e-06,
+      "loss": 0.8286,
+      "step": 757
+    },
+    {
+      "epoch": 1.8669950738916257,
+      "grad_norm": 1.2227694425501479,
+      "learning_rate": 7.504217814483364e-06,
+      "loss": 0.7792,
+      "step": 758
+    },
+    {
+      "epoch": 1.8694581280788176,
+      "grad_norm": 1.2686793829146474,
+      "learning_rate": 7.476471026852368e-06,
+      "loss": 0.7377,
+      "step": 759
+    },
+    {
+      "epoch": 1.8719211822660098,
+      "grad_norm": 1.2636415457291812,
+      "learning_rate": 7.448744973390423e-06,
+      "loss": 0.8641,
+      "step": 760
+    },
+    {
+      "epoch": 1.874384236453202,
+      "grad_norm": 1.2832085131759972,
+      "learning_rate": 7.42103988190417e-06,
+      "loss": 0.8076,
+      "step": 761
+    },
+    {
+      "epoch": 1.8768472906403941,
+      "grad_norm": 1.2736423070302447,
+      "learning_rate": 7.393355980028039e-06,
+      "loss": 0.8,
+      "step": 762
+    },
+    {
+      "epoch": 1.8793103448275863,
+      "grad_norm": 1.2361714774594441,
+      "learning_rate": 7.365693495222333e-06,
+      "loss": 0.8062,
+      "step": 763
+    },
+    {
+      "epoch": 1.8817733990147785,
+      "grad_norm": 1.3355989082868924,
+      "learning_rate": 7.338052654771407e-06,
+      "loss": 0.8897,
+      "step": 764
+    },
+    {
+      "epoch": 1.8842364532019704,
+      "grad_norm": 1.299030856035472,
+      "learning_rate": 7.310433685781777e-06,
+      "loss": 0.8515,
+      "step": 765
+    },
+    {
+      "epoch": 1.8866995073891626,
+      "grad_norm": 1.2367119216007498,
+      "learning_rate": 7.282836815180241e-06,
+      "loss": 0.7439,
+      "step": 766
+    },
+    {
+      "epoch": 1.8891625615763545,
+      "grad_norm": 1.1881454119265882,
+      "learning_rate": 7.25526226971204e-06,
+      "loss": 0.7721,
+      "step": 767
+    },
+    {
+      "epoch": 1.8916256157635467,
+      "grad_norm": 1.3040948309731069,
+      "learning_rate": 7.227710275938987e-06,
+      "loss": 0.7644,
+      "step": 768
+    },
+    {
+      "epoch": 1.8940886699507389,
+      "grad_norm": 1.2197987861110628,
+      "learning_rate": 7.200181060237591e-06,
+      "loss": 0.8303,
+      "step": 769
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 1.2828570731035869,
+      "learning_rate": 7.172674848797218e-06,
+      "loss": 0.8328,
+      "step": 770
+    },
+    {
+      "epoch": 1.8990147783251232,
+      "grad_norm": 1.4559968851181917,
+      "learning_rate": 7.1451918676182265e-06,
+      "loss": 0.9782,
+      "step": 771
+    },
+    {
+      "epoch": 1.9014778325123154,
+      "grad_norm": 1.2492933183068,
+      "learning_rate": 7.117732342510093e-06,
+      "loss": 0.8212,
+      "step": 772
+    },
+    {
+      "epoch": 1.9039408866995073,
+      "grad_norm": 1.4018960307121615,
+      "learning_rate": 7.090296499089586e-06,
+      "loss": 0.8024,
+      "step": 773
+    },
+    {
+      "epoch": 1.9064039408866995,
+      "grad_norm": 1.2182986083781042,
+      "learning_rate": 7.062884562778883e-06,
+      "loss": 0.8012,
+      "step": 774
+    },
+    {
+      "epoch": 1.9088669950738915,
+      "grad_norm": 1.2231046689175582,
+      "learning_rate": 7.035496758803744e-06,
+      "loss": 0.7983,
+      "step": 775
+    },
+    {
+      "epoch": 1.9113300492610836,
+      "grad_norm": 1.1444633906430899,
+      "learning_rate": 7.008133312191649e-06,
+      "loss": 0.7521,
+      "step": 776
+    },
+    {
+      "epoch": 1.9137931034482758,
+      "grad_norm": 1.372421100342843,
+      "learning_rate": 6.980794447769945e-06,
+      "loss": 0.8037,
+      "step": 777
+    },
+    {
+      "epoch": 1.916256157635468,
+      "grad_norm": 1.1305865286401904,
+      "learning_rate": 6.953480390164001e-06,
+      "loss": 0.7388,
+      "step": 778
+    },
+    {
+      "epoch": 1.9187192118226601,
+      "grad_norm": 1.2569974070865955,
+      "learning_rate": 6.926191363795375e-06,
+      "loss": 0.7452,
+      "step": 779
+    },
+    {
+      "epoch": 1.9211822660098523,
+      "grad_norm": 1.3412993062423042,
+      "learning_rate": 6.898927592879945e-06,
+      "loss": 0.7688,
+      "step": 780
+    },
+    {
+      "epoch": 1.9236453201970445,
+      "grad_norm": 1.2090819385708156,
+      "learning_rate": 6.871689301426101e-06,
+      "loss": 0.8282,
+      "step": 781
+    },
+    {
+      "epoch": 1.9261083743842364,
+      "grad_norm": 1.1780932434820832,
+      "learning_rate": 6.844476713232863e-06,
+      "loss": 0.7317,
+      "step": 782
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 1.1991428015699919,
+      "learning_rate": 6.817290051888084e-06,
+      "loss": 0.7313,
+      "step": 783
+    },
+    {
+      "epoch": 1.9310344827586206,
+      "grad_norm": 1.1845907412384689,
+      "learning_rate": 6.790129540766581e-06,
+      "loss": 0.6905,
+      "step": 784
+    },
+    {
+      "epoch": 1.9334975369458127,
+      "grad_norm": 1.1823628385970768,
+      "learning_rate": 6.762995403028315e-06,
+      "loss": 0.8612,
+      "step": 785
+    },
+    {
+      "epoch": 1.935960591133005,
+      "grad_norm": 1.2900978323337295,
+      "learning_rate": 6.735887861616555e-06,
+      "loss": 0.8631,
+      "step": 786
+    },
+    {
+      "epoch": 1.938423645320197,
+      "grad_norm": 1.3719330308323607,
+      "learning_rate": 6.708807139256053e-06,
+      "loss": 0.7995,
+      "step": 787
+    },
+    {
+      "epoch": 1.9408866995073892,
+      "grad_norm": 1.1448296688437882,
+      "learning_rate": 6.68175345845119e-06,
+      "loss": 0.776,
+      "step": 788
+    },
+    {
+      "epoch": 1.9433497536945814,
+      "grad_norm": 1.2807302118426145,
+      "learning_rate": 6.654727041484174e-06,
+      "loss": 0.8262,
+      "step": 789
+    },
+    {
+      "epoch": 1.9458128078817734,
+      "grad_norm": 1.1881387398139545,
+      "learning_rate": 6.627728110413214e-06,
+      "loss": 0.7318,
+      "step": 790
+    },
+    {
+      "epoch": 1.9482758620689655,
+      "grad_norm": 1.1956389735552402,
+      "learning_rate": 6.600756887070661e-06,
+      "loss": 0.7791,
+      "step": 791
+    },
+    {
+      "epoch": 1.9507389162561575,
+      "grad_norm": 1.266273567831439,
+      "learning_rate": 6.5738135930612355e-06,
+      "loss": 0.7846,
+      "step": 792
+    },
+    {
+      "epoch": 1.9532019704433496,
+      "grad_norm": 1.2813147551757353,
+      "learning_rate": 6.546898449760162e-06,
+      "loss": 0.7578,
+      "step": 793
+    },
+    {
+      "epoch": 1.9556650246305418,
+      "grad_norm": 1.2695127045670673,
+      "learning_rate": 6.520011678311382e-06,
+      "loss": 0.8526,
+      "step": 794
+    },
+    {
+      "epoch": 1.958128078817734,
+      "grad_norm": 1.2579803354821724,
+      "learning_rate": 6.493153499625719e-06,
+      "loss": 0.7402,
+      "step": 795
+    },
+    {
+      "epoch": 1.9605911330049262,
+      "grad_norm": 1.2582817860754603,
+      "learning_rate": 6.466324134379066e-06,
+      "loss": 0.8348,
+      "step": 796
+    },
+    {
+      "epoch": 1.9630541871921183,
+      "grad_norm": 1.2385063087201058,
+      "learning_rate": 6.4395238030105786e-06,
+      "loss": 0.8478,
+      "step": 797
+    },
+    {
+      "epoch": 1.9655172413793105,
+      "grad_norm": 1.195808457349546,
+      "learning_rate": 6.412752725720864e-06,
+      "loss": 0.7572,
+      "step": 798
+    },
+    {
+      "epoch": 1.9679802955665024,
+      "grad_norm": 1.3577088416317467,
+      "learning_rate": 6.386011122470159e-06,
+      "loss": 0.8267,
+      "step": 799
+    },
+    {
+      "epoch": 1.9704433497536946,
+      "grad_norm": 1.2137305749817913,
+      "learning_rate": 6.359299212976535e-06,
+      "loss": 0.7496,
+      "step": 800
+    },
+    {
+      "epoch": 1.9729064039408866,
+      "grad_norm": 1.2600435645277315,
+      "learning_rate": 6.332617216714099e-06,
+      "loss": 0.8067,
+      "step": 801
+    },
+    {
+      "epoch": 1.9753694581280787,
+      "grad_norm": 1.2359007086022071,
+      "learning_rate": 6.305965352911162e-06,
+      "loss": 0.7763,
+      "step": 802
+    },
+    {
+      "epoch": 1.977832512315271,
+      "grad_norm": 1.181088102437968,
+      "learning_rate": 6.279343840548475e-06,
+      "loss": 0.7879,
+      "step": 803
+    },
+    {
+      "epoch": 1.980295566502463,
+      "grad_norm": 1.219894312126708,
+      "learning_rate": 6.252752898357397e-06,
+      "loss": 0.7964,
+      "step": 804
+    },
+    {
+      "epoch": 1.9827586206896552,
+      "grad_norm": 1.2905088691096278,
+      "learning_rate": 6.226192744818125e-06,
+      "loss": 0.7987,
+      "step": 805
+    },
+    {
+      "epoch": 1.9852216748768474,
+      "grad_norm": 1.2774010508339309,
+      "learning_rate": 6.1996635981578755e-06,
+      "loss": 0.7586,
+      "step": 806
+    },
+    {
+      "epoch": 1.9876847290640394,
+      "grad_norm": 1.2990616896725224,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.8053,
+      "step": 807
+    },
+    {
+      "epoch": 1.9901477832512315,
+      "grad_norm": 1.2875472902668321,
+      "learning_rate": 6.146699197107715e-06,
+      "loss": 0.8085,
+      "step": 808
+    },
+    {
+      "epoch": 1.9926108374384235,
+      "grad_norm": 1.297497657303358,
+      "learning_rate": 6.1202643778912755e-06,
+      "loss": 0.7358,
+      "step": 809
+    },
+    {
+      "epoch": 1.9950738916256157,
+      "grad_norm": 1.3375207542876968,
+      "learning_rate": 6.093861435897208e-06,
+      "loss": 0.7534,
+      "step": 810
+    },
+    {
+      "epoch": 1.9975369458128078,
+      "grad_norm": 1.1428449791874038,
+      "learning_rate": 6.067490588061034e-06,
+      "loss": 0.835,
+      "step": 811
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1221671996355929,
+      "learning_rate": 6.041152051054575e-06,
+      "loss": 0.7618,
+      "step": 812
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1783102750778198,
+      "eval_runtime": 1.0243,
+      "eval_samples_per_second": 51.741,
+      "eval_steps_per_second": 1.952,
+      "step": 812
+    },
+    {
+      "epoch": 2.002463054187192,
+      "grad_norm": 1.588914754445093,
+      "learning_rate": 6.014846041284168e-06,
+      "loss": 0.4776,
+      "step": 813
+    },
+    {
+      "epoch": 2.0049261083743843,
+      "grad_norm": 1.5350096343791197,
+      "learning_rate": 5.988572774888913e-06,
+      "loss": 0.5285,
+      "step": 814
+    },
+    {
+      "epoch": 2.0073891625615765,
+      "grad_norm": 1.432797078056913,
+      "learning_rate": 5.962332467738858e-06,
+      "loss": 0.5086,
+      "step": 815
+    },
+    {
+      "epoch": 2.0098522167487687,
+      "grad_norm": 1.5156168986625058,
+      "learning_rate": 5.936125335433265e-06,
+      "loss": 0.5012,
+      "step": 816
+    },
+    {
+      "epoch": 2.0123152709359604,
+      "grad_norm": 1.3629511567458108,
+      "learning_rate": 5.909951593298811e-06,
+      "loss": 0.5097,
+      "step": 817
+    },
+    {
+      "epoch": 2.0147783251231526,
+      "grad_norm": 1.455888069592663,
+      "learning_rate": 5.883811456387821e-06,
+      "loss": 0.47,
+      "step": 818
+    },
+    {
+      "epoch": 2.0172413793103448,
+      "grad_norm": 1.3828572842517384,
+      "learning_rate": 5.857705139476525e-06,
+      "loss": 0.4102,
+      "step": 819
+    },
+    {
+      "epoch": 2.019704433497537,
+      "grad_norm": 1.3017575377730861,
+      "learning_rate": 5.831632857063271e-06,
+      "loss": 0.5274,
+      "step": 820
+    },
+    {
+      "epoch": 2.022167487684729,
+      "grad_norm": 1.108056587908784,
+      "learning_rate": 5.805594823366751e-06,
+      "loss": 0.4378,
+      "step": 821
+    },
+    {
+      "epoch": 2.0246305418719213,
+      "grad_norm": 1.5460641450442574,
+      "learning_rate": 5.779591252324286e-06,
+      "loss": 0.4807,
+      "step": 822
+    },
+    {
+      "epoch": 2.0270935960591134,
+      "grad_norm": 1.8603857162424282,
+      "learning_rate": 5.753622357590021e-06,
+      "loss": 0.4289,
+      "step": 823
+    },
+    {
+      "epoch": 2.0295566502463056,
+      "grad_norm": 2.16456724815034,
+      "learning_rate": 5.7276883525331915e-06,
+      "loss": 0.4917,
+      "step": 824
+    },
+    {
+      "epoch": 2.0320197044334973,
+      "grad_norm": 1.9616754366803704,
+      "learning_rate": 5.701789450236377e-06,
+      "loss": 0.4341,
+      "step": 825
+    },
+    {
+      "epoch": 2.0344827586206895,
+      "grad_norm": 1.5161982759740398,
+      "learning_rate": 5.675925863493721e-06,
+      "loss": 0.4152,
+      "step": 826
+    },
+    {
+      "epoch": 2.0369458128078817,
+      "grad_norm": 1.375299941206896,
+      "learning_rate": 5.650097804809224e-06,
+      "loss": 0.4969,
+      "step": 827
+    },
+    {
+      "epoch": 2.039408866995074,
+      "grad_norm": 1.6895547296036824,
+      "learning_rate": 5.6243054863949675e-06,
+      "loss": 0.4731,
+      "step": 828
+    },
+    {
+      "epoch": 2.041871921182266,
+      "grad_norm": 1.3043534085251032,
+      "learning_rate": 5.598549120169376e-06,
+      "loss": 0.4922,
+      "step": 829
+    },
+    {
+      "epoch": 2.044334975369458,
+      "grad_norm": 1.2409519312491848,
+      "learning_rate": 5.5728289177554805e-06,
+      "loss": 0.478,
+      "step": 830
+    },
+    {
+      "epoch": 2.0467980295566504,
+      "grad_norm": 1.1417491546587637,
+      "learning_rate": 5.54714509047918e-06,
+      "loss": 0.4711,
+      "step": 831
+    },
+    {
+      "epoch": 2.0492610837438425,
+      "grad_norm": 1.1867496227796592,
+      "learning_rate": 5.521497849367501e-06,
+      "loss": 0.4489,
+      "step": 832
+    },
+    {
+      "epoch": 2.0517241379310347,
+      "grad_norm": 1.1942192014348725,
+      "learning_rate": 5.49588740514686e-06,
+      "loss": 0.4616,
+      "step": 833
+    },
+    {
+      "epoch": 2.0541871921182264,
+      "grad_norm": 1.2845719235227278,
+      "learning_rate": 5.4703139682413585e-06,
+      "loss": 0.4909,
+      "step": 834
+    },
+    {
+      "epoch": 2.0566502463054186,
+      "grad_norm": 1.193612579105303,
+      "learning_rate": 5.444777748771006e-06,
+      "loss": 0.4631,
+      "step": 835
+    },
+    {
+      "epoch": 2.0591133004926108,
+      "grad_norm": 1.3477404139677929,
+      "learning_rate": 5.419278956550037e-06,
+      "loss": 0.4235,
+      "step": 836
+    },
+    {
+      "epoch": 2.061576354679803,
+      "grad_norm": 1.3198425088476715,
+      "learning_rate": 5.393817801085168e-06,
+      "loss": 0.449,
+      "step": 837
+    },
+    {
+      "epoch": 2.064039408866995,
+      "grad_norm": 1.214429975002499,
+      "learning_rate": 5.368394491573876e-06,
+      "loss": 0.4041,
+      "step": 838
+    },
+    {
+      "epoch": 2.0665024630541873,
+      "grad_norm": 1.2538441675331544,
+      "learning_rate": 5.343009236902696e-06,
+      "loss": 0.4219,
+      "step": 839
+    },
+    {
+      "epoch": 2.0689655172413794,
+      "grad_norm": 1.3246926778992256,
+      "learning_rate": 5.31766224564547e-06,
+      "loss": 0.4486,
+      "step": 840
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 1.4256185360385263,
+      "learning_rate": 5.292353726061665e-06,
+      "loss": 0.4448,
+      "step": 841
+    },
+    {
+      "epoch": 2.0738916256157633,
+      "grad_norm": 1.3944025060384952,
+      "learning_rate": 5.267083886094668e-06,
+      "loss": 0.4167,
+      "step": 842
+    },
+    {
+      "epoch": 2.0763546798029555,
+      "grad_norm": 1.4379516389643867,
+      "learning_rate": 5.241852933370026e-06,
+      "loss": 0.4192,
+      "step": 843
+    },
+    {
+      "epoch": 2.0788177339901477,
+      "grad_norm": 1.457537115812056,
+      "learning_rate": 5.216661075193814e-06,
+      "loss": 0.4773,
+      "step": 844
+    },
+    {
+      "epoch": 2.08128078817734,
+      "grad_norm": 1.370868110022366,
+      "learning_rate": 5.191508518550856e-06,
+      "loss": 0.4553,
+      "step": 845
+    },
+    {
+      "epoch": 2.083743842364532,
+      "grad_norm": 1.3292778397149203,
+      "learning_rate": 5.166395470103092e-06,
+      "loss": 0.4155,
+      "step": 846
+    },
+    {
+      "epoch": 2.086206896551724,
+      "grad_norm": 1.3856581809648412,
+      "learning_rate": 5.141322136187837e-06,
+      "loss": 0.424,
+      "step": 847
+    },
+    {
+      "epoch": 2.0886699507389164,
+      "grad_norm": 1.4148998837049225,
+      "learning_rate": 5.116288722816087e-06,
+      "loss": 0.4565,
+      "step": 848
+    },
+    {
+      "epoch": 2.0911330049261085,
+      "grad_norm": 1.3557392037433267,
+      "learning_rate": 5.091295435670856e-06,
+      "loss": 0.4058,
+      "step": 849
+    },
+    {
+      "epoch": 2.0935960591133007,
+      "grad_norm": 1.4072507899870548,
+      "learning_rate": 5.06634248010546e-06,
+      "loss": 0.4387,
+      "step": 850
+    },
+    {
+      "epoch": 2.0960591133004924,
+      "grad_norm": 1.4501048050197551,
+      "learning_rate": 5.041430061141833e-06,
+      "loss": 0.3873,
+      "step": 851
+    },
+    {
+      "epoch": 2.0985221674876846,
+      "grad_norm": 1.3717223462653934,
+      "learning_rate": 5.016558383468851e-06,
+      "loss": 0.4427,
+      "step": 852
+    },
+    {
+      "epoch": 2.100985221674877,
+      "grad_norm": 1.4196314455655044,
+      "learning_rate": 4.991727651440644e-06,
+      "loss": 0.4925,
+      "step": 853
+    },
+    {
+      "epoch": 2.103448275862069,
+      "grad_norm": 1.3996902906624666,
+      "learning_rate": 4.9669380690749215e-06,
+      "loss": 0.4299,
+      "step": 854
+    },
+    {
+      "epoch": 2.105911330049261,
+      "grad_norm": 1.413379187957181,
+      "learning_rate": 4.942189840051288e-06,
+      "loss": 0.4626,
+      "step": 855
+    },
+    {
+      "epoch": 2.1083743842364533,
+      "grad_norm": 1.4796501050209723,
+      "learning_rate": 4.91748316770958e-06,
+      "loss": 0.3961,
+      "step": 856
+    },
+    {
+      "epoch": 2.1108374384236455,
+      "grad_norm": 1.2866241512949912,
+      "learning_rate": 4.892818255048188e-06,
+      "loss": 0.3805,
+      "step": 857
+    },
+    {
+      "epoch": 2.1133004926108376,
+      "grad_norm": 1.4281302554244624,
+      "learning_rate": 4.868195304722391e-06,
+      "loss": 0.4653,
+      "step": 858
+    },
+    {
+      "epoch": 2.1157635467980294,
+      "grad_norm": 1.3873106217233924,
+      "learning_rate": 4.843614519042693e-06,
+      "loss": 0.453,
+      "step": 859
+    },
+    {
+      "epoch": 2.1182266009852215,
+      "grad_norm": 1.4992551543729367,
+      "learning_rate": 4.819076099973152e-06,
+      "loss": 0.4783,
+      "step": 860
+    },
+    {
+      "epoch": 2.1206896551724137,
+      "grad_norm": 1.4555068645308868,
+      "learning_rate": 4.794580249129745e-06,
+      "loss": 0.4386,
+      "step": 861
+    },
+    {
+      "epoch": 2.123152709359606,
+      "grad_norm": 1.3677150471453048,
+      "learning_rate": 4.77012716777867e-06,
+      "loss": 0.4535,
+      "step": 862
+    },
+    {
+      "epoch": 2.125615763546798,
+      "grad_norm": 1.4592593749677938,
+      "learning_rate": 4.745717056834729e-06,
+      "loss": 0.4784,
+      "step": 863
+    },
+    {
+      "epoch": 2.12807881773399,
+      "grad_norm": 1.3058905622596233,
+      "learning_rate": 4.721350116859675e-06,
+      "loss": 0.4758,
+      "step": 864
+    },
+    {
+      "epoch": 2.1305418719211824,
+      "grad_norm": 1.3796160548753895,
+      "learning_rate": 4.697026548060528e-06,
+      "loss": 0.3942,
+      "step": 865
+    },
+    {
+      "epoch": 2.1330049261083746,
+      "grad_norm": 1.5689985557097283,
+      "learning_rate": 4.672746550287985e-06,
+      "loss": 0.3992,
+      "step": 866
+    },
+    {
+      "epoch": 2.1354679802955667,
+      "grad_norm": 1.4119990346597284,
+      "learning_rate": 4.64851032303472e-06,
+      "loss": 0.4385,
+      "step": 867
+    },
+    {
+      "epoch": 2.1379310344827585,
+      "grad_norm": 1.316583638450487,
+      "learning_rate": 4.6243180654337975e-06,
+      "loss": 0.4356,
+      "step": 868
+    },
+    {
+      "epoch": 2.1403940886699506,
+      "grad_norm": 1.311193542574191,
+      "learning_rate": 4.600169976257003e-06,
+      "loss": 0.4192,
+      "step": 869
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 1.423052457452422,
+      "learning_rate": 4.576066253913209e-06,
+      "loss": 0.4245,
+      "step": 870
+    },
+    {
+      "epoch": 2.145320197044335,
+      "grad_norm": 1.4693513004626029,
+      "learning_rate": 4.5520070964467736e-06,
+      "loss": 0.4587,
+      "step": 871
+    },
+    {
+      "epoch": 2.147783251231527,
+      "grad_norm": 1.384490207674686,
+      "learning_rate": 4.527992701535884e-06,
+      "loss": 0.4481,
+      "step": 872
+    },
+    {
+      "epoch": 2.1502463054187193,
+      "grad_norm": 1.4328977470040005,
+      "learning_rate": 4.504023266490945e-06,
+      "loss": 0.4682,
+      "step": 873
+    },
+    {
+      "epoch": 2.1527093596059115,
+      "grad_norm": 1.37449823240775,
+      "learning_rate": 4.480098988252958e-06,
+      "loss": 0.4532,
+      "step": 874
+    },
+    {
+      "epoch": 2.1551724137931036,
+      "grad_norm": 1.335432347499691,
+      "learning_rate": 4.456220063391895e-06,
+      "loss": 0.4634,
+      "step": 875
+    },
+    {
+      "epoch": 2.1576354679802954,
+      "grad_norm": 1.4102490254577829,
+      "learning_rate": 4.432386688105095e-06,
+      "loss": 0.4549,
+      "step": 876
+    },
+    {
+      "epoch": 2.1600985221674875,
+      "grad_norm": 1.28216640915284,
+      "learning_rate": 4.4085990582156455e-06,
+      "loss": 0.4421,
+      "step": 877
+    },
+    {
+      "epoch": 2.1625615763546797,
+      "grad_norm": 1.366684303324961,
+      "learning_rate": 4.384857369170772e-06,
+      "loss": 0.4011,
+      "step": 878
+    },
+    {
+      "epoch": 2.165024630541872,
+      "grad_norm": 1.1996920269907791,
+      "learning_rate": 4.361161816040234e-06,
+      "loss": 0.4414,
+      "step": 879
+    },
+    {
+      "epoch": 2.167487684729064,
+      "grad_norm": 1.3944536716456477,
+      "learning_rate": 4.337512593514729e-06,
+      "loss": 0.4815,
+      "step": 880
+    },
+    {
+      "epoch": 2.1699507389162562,
+      "grad_norm": 1.3901493244785412,
+      "learning_rate": 4.313909895904278e-06,
+      "loss": 0.4426,
+      "step": 881
+    },
+    {
+      "epoch": 2.1724137931034484,
+      "grad_norm": 1.32233283408806,
+      "learning_rate": 4.290353917136639e-06,
+      "loss": 0.4536,
+      "step": 882
+    },
+    {
+      "epoch": 2.1748768472906406,
+      "grad_norm": 1.3272136478529442,
+      "learning_rate": 4.2668448507557256e-06,
+      "loss": 0.4312,
+      "step": 883
+    },
+    {
+      "epoch": 2.1773399014778327,
+      "grad_norm": 1.4155708824743911,
+      "learning_rate": 4.243382889919981e-06,
+      "loss": 0.4473,
+      "step": 884
+    },
+    {
+      "epoch": 2.1798029556650245,
+      "grad_norm": 1.2008275476687367,
+      "learning_rate": 4.219968227400826e-06,
+      "loss": 0.4038,
+      "step": 885
+    },
+    {
+      "epoch": 2.1822660098522166,
+      "grad_norm": 1.3564194305345516,
+      "learning_rate": 4.1966010555810696e-06,
+      "loss": 0.3835,
+      "step": 886
+    },
+    {
+      "epoch": 2.184729064039409,
+      "grad_norm": 1.3534838740160056,
+      "learning_rate": 4.173281566453301e-06,
+      "loss": 0.4761,
+      "step": 887
+    },
+    {
+      "epoch": 2.187192118226601,
+      "grad_norm": 1.3017077196675206,
+      "learning_rate": 4.1500099516183555e-06,
+      "loss": 0.4116,
+      "step": 888
+    },
+    {
+      "epoch": 2.189655172413793,
+      "grad_norm": 1.3703991469997996,
+      "learning_rate": 4.126786402283692e-06,
+      "loss": 0.4547,
+      "step": 889
+    },
+    {
+      "epoch": 2.1921182266009853,
+      "grad_norm": 1.3984987043061732,
+      "learning_rate": 4.1036111092618725e-06,
+      "loss": 0.4199,
+      "step": 890
+    },
+    {
+      "epoch": 2.1945812807881775,
+      "grad_norm": 1.2738074952870384,
+      "learning_rate": 4.080484262968955e-06,
+      "loss": 0.4075,
+      "step": 891
+    },
+    {
+      "epoch": 2.1970443349753697,
+      "grad_norm": 1.3801369942516273,
+      "learning_rate": 4.057406053422933e-06,
+      "loss": 0.4569,
+      "step": 892
+    },
+    {
+      "epoch": 2.1995073891625614,
+      "grad_norm": 1.311085074555953,
+      "learning_rate": 4.0343766702422075e-06,
+      "loss": 0.4554,
+      "step": 893
+    },
+    {
+      "epoch": 2.2019704433497536,
+      "grad_norm": 1.452184967833434,
+      "learning_rate": 4.011396302643989e-06,
+      "loss": 0.4341,
+      "step": 894
+    },
+    {
+      "epoch": 2.2044334975369457,
+      "grad_norm": 1.445857836967331,
+      "learning_rate": 3.9884651394427635e-06,
+      "loss": 0.4505,
+      "step": 895
+    },
+    {
+      "epoch": 2.206896551724138,
+      "grad_norm": 1.3382342043578157,
+      "learning_rate": 3.965583369048737e-06,
+      "loss": 0.4403,
+      "step": 896
+    },
+    {
+      "epoch": 2.20935960591133,
+      "grad_norm": 1.3057248679862405,
+      "learning_rate": 3.94275117946629e-06,
+      "loss": 0.4696,
+      "step": 897
+    },
+    {
+      "epoch": 2.2118226600985222,
+      "grad_norm": 1.2809589334496911,
+      "learning_rate": 3.919968758292425e-06,
+      "loss": 0.4799,
+      "step": 898
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 1.3557106703635813,
+      "learning_rate": 3.897236292715235e-06,
+      "loss": 0.4113,
+      "step": 899
+    },
+    {
+      "epoch": 2.2167487684729066,
+      "grad_norm": 1.2588564408575982,
+      "learning_rate": 3.874553969512358e-06,
+      "loss": 0.3738,
+      "step": 900
+    },
+    {
+      "epoch": 2.2192118226600988,
+      "grad_norm": 1.3209859270505637,
+      "learning_rate": 3.851921975049448e-06,
+      "loss": 0.4744,
+      "step": 901
+    },
+    {
+      "epoch": 2.2216748768472905,
+      "grad_norm": 1.4353782918125415,
+      "learning_rate": 3.82934049527864e-06,
+      "loss": 0.4334,
+      "step": 902
+    },
+    {
+      "epoch": 2.2241379310344827,
+      "grad_norm": 1.370166127859558,
+      "learning_rate": 3.8068097157370198e-06,
+      "loss": 0.4198,
+      "step": 903
+    },
+    {
+      "epoch": 2.226600985221675,
+      "grad_norm": 1.3785248139686699,
+      "learning_rate": 3.784329821545105e-06,
+      "loss": 0.493,
+      "step": 904
+    },
+    {
+      "epoch": 2.229064039408867,
+      "grad_norm": 1.4012086147216676,
+      "learning_rate": 3.761900997405332e-06,
+      "loss": 0.4261,
+      "step": 905
+    },
+    {
+      "epoch": 2.231527093596059,
+      "grad_norm": 1.4598475179344954,
+      "learning_rate": 3.739523427600509e-06,
+      "loss": 0.4882,
+      "step": 906
+    },
+    {
+      "epoch": 2.2339901477832513,
+      "grad_norm": 1.3382645355980185,
+      "learning_rate": 3.7171972959923306e-06,
+      "loss": 0.389,
+      "step": 907
+    },
+    {
+      "epoch": 2.2364532019704435,
+      "grad_norm": 1.6313561328255533,
+      "learning_rate": 3.6949227860198712e-06,
+      "loss": 0.4413,
+      "step": 908
+    },
+    {
+      "epoch": 2.2389162561576357,
+      "grad_norm": 1.3363884950954974,
+      "learning_rate": 3.672700080698035e-06,
+      "loss": 0.4373,
+      "step": 909
+    },
+    {
+      "epoch": 2.2413793103448274,
+      "grad_norm": 1.2852328199400724,
+      "learning_rate": 3.650529362616113e-06,
+      "loss": 0.4235,
+      "step": 910
+    },
+    {
+      "epoch": 2.2438423645320196,
+      "grad_norm": 1.3035202286924483,
+      "learning_rate": 3.628410813936224e-06,
+      "loss": 0.4315,
+      "step": 911
+    },
+    {
+      "epoch": 2.2463054187192117,
+      "grad_norm": 1.413447124980615,
+      "learning_rate": 3.606344616391867e-06,
+      "loss": 0.4366,
+      "step": 912
+    },
+    {
+      "epoch": 2.248768472906404,
+      "grad_norm": 1.4223611021696319,
+      "learning_rate": 3.584330951286398e-06,
+      "loss": 0.3895,
+      "step": 913
+    },
+    {
+      "epoch": 2.251231527093596,
+      "grad_norm": 1.255207942893995,
+      "learning_rate": 3.5623699994915363e-06,
+      "loss": 0.3745,
+      "step": 914
+    },
+    {
+      "epoch": 2.2536945812807883,
+      "grad_norm": 1.4374623501035038,
+      "learning_rate": 3.540461941445915e-06,
+      "loss": 0.4634,
+      "step": 915
+    },
+    {
+      "epoch": 2.2561576354679804,
+      "grad_norm": 1.5027013931244602,
+      "learning_rate": 3.5186069571535575e-06,
+      "loss": 0.4446,
+      "step": 916
+    },
+    {
+      "epoch": 2.2586206896551726,
+      "grad_norm": 1.375540410395641,
+      "learning_rate": 3.4968052261824194e-06,
+      "loss": 0.424,
+      "step": 917
+    },
+    {
+      "epoch": 2.2610837438423648,
+      "grad_norm": 1.3719507565794786,
+      "learning_rate": 3.475056927662912e-06,
+      "loss": 0.4747,
+      "step": 918
+    },
+    {
+      "epoch": 2.2635467980295565,
+      "grad_norm": 1.4772716615579669,
+      "learning_rate": 3.453362240286414e-06,
+      "loss": 0.428,
+      "step": 919
+    },
+    {
+      "epoch": 2.2660098522167487,
+      "grad_norm": 1.4999802906013076,
+      "learning_rate": 3.4317213423038386e-06,
+      "loss": 0.5345,
+      "step": 920
+    },
+    {
+      "epoch": 2.268472906403941,
+      "grad_norm": 1.2596303941165061,
+      "learning_rate": 3.410134411524133e-06,
+      "loss": 0.4026,
+      "step": 921
+    },
+    {
+      "epoch": 2.270935960591133,
+      "grad_norm": 1.5112210817307412,
+      "learning_rate": 3.388601625312833e-06,
+      "loss": 0.5103,
+      "step": 922
+    },
+    {
+      "epoch": 2.273399014778325,
+      "grad_norm": 1.35817191719229,
+      "learning_rate": 3.3671231605906052e-06,
+      "loss": 0.436,
+      "step": 923
+    },
+    {
+      "epoch": 2.2758620689655173,
+      "grad_norm": 1.325168955125457,
+      "learning_rate": 3.345699193831795e-06,
+      "loss": 0.4255,
+      "step": 924
+    },
+    {
+      "epoch": 2.2783251231527095,
+      "grad_norm": 1.3448844379285703,
+      "learning_rate": 3.3243299010629706e-06,
+      "loss": 0.4303,
+      "step": 925
+    },
+    {
+      "epoch": 2.2807881773399012,
+      "grad_norm": 1.3721766094695604,
+      "learning_rate": 3.3030154578614783e-06,
+      "loss": 0.47,
+      "step": 926
+    },
+    {
+      "epoch": 2.2832512315270934,
+      "grad_norm": 1.4236530463185595,
+      "learning_rate": 3.281756039354016e-06,
+      "loss": 0.4362,
+      "step": 927
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 1.467129239194111,
+      "learning_rate": 3.2605518202151577e-06,
+      "loss": 0.4473,
+      "step": 928
+    },
+    {
+      "epoch": 2.2881773399014778,
+      "grad_norm": 1.4686674541246747,
+      "learning_rate": 3.239402974665956e-06,
+      "loss": 0.4715,
+      "step": 929
+    },
+    {
+      "epoch": 2.29064039408867,
+      "grad_norm": 1.4915559240963885,
+      "learning_rate": 3.218309676472492e-06,
+      "loss": 0.4558,
+      "step": 930
+    },
+    {
+      "epoch": 2.293103448275862,
+      "grad_norm": 1.1865656841276544,
+      "learning_rate": 3.1972720989444473e-06,
+      "loss": 0.3997,
+      "step": 931
+    },
+    {
+      "epoch": 2.2955665024630543,
+      "grad_norm": 1.3572809159894121,
+      "learning_rate": 3.1762904149336947e-06,
+      "loss": 0.4496,
+      "step": 932
+    },
+    {
+      "epoch": 2.2980295566502464,
+      "grad_norm": 1.3456093986659265,
+      "learning_rate": 3.155364796832848e-06,
+      "loss": 0.3912,
+      "step": 933
+    },
+    {
+      "epoch": 2.3004926108374386,
+      "grad_norm": 1.3017956784715656,
+      "learning_rate": 3.134495416573884e-06,
+      "loss": 0.4628,
+      "step": 934
+    },
+    {
+      "epoch": 2.302955665024631,
+      "grad_norm": 1.3513307758646476,
+      "learning_rate": 3.113682445626701e-06,
+      "loss": 0.4348,
+      "step": 935
+    },
+    {
+      "epoch": 2.3054187192118225,
+      "grad_norm": 1.2330294168030709,
+      "learning_rate": 3.0929260549977116e-06,
+      "loss": 0.4316,
+      "step": 936
+    },
+    {
+      "epoch": 2.3078817733990147,
+      "grad_norm": 1.3349058841719377,
+      "learning_rate": 3.072226415228463e-06,
+      "loss": 0.4577,
+      "step": 937
+    },
+    {
+      "epoch": 2.310344827586207,
+      "grad_norm": 1.4275976029414144,
+      "learning_rate": 3.0515836963942056e-06,
+      "loss": 0.4152,
+      "step": 938
+    },
+    {
+      "epoch": 2.312807881773399,
+      "grad_norm": 1.5018768489602432,
+      "learning_rate": 3.030998068102512e-06,
+      "loss": 0.5112,
+      "step": 939
+    },
+    {
+      "epoch": 2.315270935960591,
+      "grad_norm": 1.422756215437372,
+      "learning_rate": 3.01046969949188e-06,
+      "loss": 0.4346,
+      "step": 940
+    },
+    {
+      "epoch": 2.3177339901477834,
+      "grad_norm": 1.3363316718497398,
+      "learning_rate": 2.989998759230335e-06,
+      "loss": 0.4312,
+      "step": 941
+    },
+    {
+      "epoch": 2.3201970443349755,
+      "grad_norm": 1.448983885454986,
+      "learning_rate": 2.9695854155140648e-06,
+      "loss": 0.4377,
+      "step": 942
+    },
+    {
+      "epoch": 2.3226600985221673,
+      "grad_norm": 1.4581804013672377,
+      "learning_rate": 2.9492298360660164e-06,
+      "loss": 0.418,
+      "step": 943
+    },
+    {
+      "epoch": 2.3251231527093594,
+      "grad_norm": 1.319855334345226,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.4423,
+      "step": 944
+    },
+    {
+      "epoch": 2.3275862068965516,
+      "grad_norm": 1.580740545777575,
+      "learning_rate": 2.9086926384919457e-06,
+      "loss": 0.433,
+      "step": 945
+    },
+    {
+      "epoch": 2.3300492610837438,
+      "grad_norm": 1.3795581605796603,
+      "learning_rate": 2.8885113534332742e-06,
+      "loss": 0.3646,
+      "step": 946
+    },
+    {
+      "epoch": 2.332512315270936,
+      "grad_norm": 1.5096790836694218,
+      "learning_rate": 2.868388498774787e-06,
+      "loss": 0.4518,
+      "step": 947
+    },
+    {
+      "epoch": 2.334975369458128,
+      "grad_norm": 1.4423367771248332,
+      "learning_rate": 2.8483242398526723e-06,
+      "loss": 0.444,
+      "step": 948
+    },
+    {
+      "epoch": 2.3374384236453203,
+      "grad_norm": 1.336413605383313,
+      "learning_rate": 2.8283187415216882e-06,
+      "loss": 0.4333,
+      "step": 949
+    },
+    {
+      "epoch": 2.3399014778325125,
+      "grad_norm": 1.5454395454107182,
+      "learning_rate": 2.80837216815378e-06,
+      "loss": 0.5254,
+      "step": 950
+    },
+    {
+      "epoch": 2.3423645320197046,
+      "grad_norm": 1.3720388605714537,
+      "learning_rate": 2.788484683636753e-06,
+      "loss": 0.4129,
+      "step": 951
+    },
+    {
+      "epoch": 2.344827586206897,
+      "grad_norm": 1.4461430817880851,
+      "learning_rate": 2.7686564513729198e-06,
+      "loss": 0.4638,
+      "step": 952
+    },
+    {
+      "epoch": 2.3472906403940885,
+      "grad_norm": 1.347253269261669,
+      "learning_rate": 2.7488876342777505e-06,
+      "loss": 0.449,
+      "step": 953
+    },
+    {
+      "epoch": 2.3497536945812807,
+      "grad_norm": 1.2862954699428446,
+      "learning_rate": 2.7291783947785544e-06,
+      "loss": 0.4624,
+      "step": 954
+    },
+    {
+      "epoch": 2.352216748768473,
+      "grad_norm": 1.327859784741237,
+      "learning_rate": 2.7095288948131114e-06,
+      "loss": 0.4315,
+      "step": 955
+    },
+    {
+      "epoch": 2.354679802955665,
+      "grad_norm": 1.3669530176992353,
+      "learning_rate": 2.689939295828371e-06,
+      "loss": 0.417,
+      "step": 956
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 1.3073976430707428,
+      "learning_rate": 2.6704097587791245e-06,
+      "loss": 0.4108,
+      "step": 957
+    },
+    {
+      "epoch": 2.3596059113300494,
+      "grad_norm": 1.5054897938372562,
+      "learning_rate": 2.650940444126654e-06,
+      "loss": 0.4266,
+      "step": 958
+    },
+    {
+      "epoch": 2.3620689655172415,
+      "grad_norm": 1.3119458436488454,
+      "learning_rate": 2.6315315118374528e-06,
+      "loss": 0.4102,
+      "step": 959
+    },
+    {
+      "epoch": 2.3645320197044333,
+      "grad_norm": 1.3132832324610446,
+      "learning_rate": 2.6121831213818825e-06,
+      "loss": 0.4094,
+      "step": 960
+    },
+    {
+      "epoch": 2.3669950738916254,
+      "grad_norm": 1.4849618655102403,
+      "learning_rate": 2.5928954317328735e-06,
+      "loss": 0.4476,
+      "step": 961
+    },
+    {
+      "epoch": 2.3694581280788176,
+      "grad_norm": 1.427387910440588,
+      "learning_rate": 2.5736686013646226e-06,
+      "loss": 0.4543,
+      "step": 962
+    },
+    {
+      "epoch": 2.37192118226601,
+      "grad_norm": 1.2309953803898444,
+      "learning_rate": 2.5545027882512742e-06,
+      "loss": 0.4183,
+      "step": 963
+    },
+    {
+      "epoch": 2.374384236453202,
+      "grad_norm": 1.3941449396948984,
+      "learning_rate": 2.535398149865651e-06,
+      "loss": 0.4184,
+      "step": 964
+    },
+    {
+      "epoch": 2.376847290640394,
+      "grad_norm": 1.431910469884643,
+      "learning_rate": 2.5163548431779306e-06,
+      "loss": 0.4735,
+      "step": 965
+    },
+    {
+      "epoch": 2.3793103448275863,
+      "grad_norm": 1.4183629016207275,
+      "learning_rate": 2.4973730246543736e-06,
+      "loss": 0.5079,
+      "step": 966
+    },
+    {
+      "epoch": 2.3817733990147785,
+      "grad_norm": 1.4273244371031524,
+      "learning_rate": 2.4784528502560334e-06,
+      "loss": 0.524,
+      "step": 967
+    },
+    {
+      "epoch": 2.3842364532019706,
+      "grad_norm": 1.3868244399506982,
+      "learning_rate": 2.4595944754374723e-06,
+      "loss": 0.4884,
+      "step": 968
+    },
+    {
+      "epoch": 2.386699507389163,
+      "grad_norm": 1.3187397273960737,
+      "learning_rate": 2.440798055145487e-06,
+      "loss": 0.427,
+      "step": 969
+    },
+    {
+      "epoch": 2.3891625615763545,
+      "grad_norm": 1.3870998235065886,
+      "learning_rate": 2.422063743817832e-06,
+      "loss": 0.4765,
+      "step": 970
+    },
+    {
+      "epoch": 2.3916256157635467,
+      "grad_norm": 1.2139987766819027,
+      "learning_rate": 2.403391695381966e-06,
+      "loss": 0.4608,
+      "step": 971
+    },
+    {
+      "epoch": 2.394088669950739,
+      "grad_norm": 1.3780286984521148,
+      "learning_rate": 2.3847820632537565e-06,
+      "loss": 0.426,
+      "step": 972
+    },
+    {
+      "epoch": 2.396551724137931,
+      "grad_norm": 1.4032638634017782,
+      "learning_rate": 2.36623500033625e-06,
+      "loss": 0.4371,
+      "step": 973
+    },
+    {
+      "epoch": 2.399014778325123,
+      "grad_norm": 1.3587077465348367,
+      "learning_rate": 2.347750659018397e-06,
+      "loss": 0.437,
+      "step": 974
+    },
+    {
+      "epoch": 2.4014778325123154,
+      "grad_norm": 1.43116767013501,
+      "learning_rate": 2.329329191173808e-06,
+      "loss": 0.3933,
+      "step": 975
+    },
+    {
+      "epoch": 2.4039408866995076,
+      "grad_norm": 1.3497230934560185,
+      "learning_rate": 2.3109707481595113e-06,
+      "loss": 0.4169,
+      "step": 976
+    },
+    {
+      "epoch": 2.4064039408866993,
+      "grad_norm": 1.3400589572690111,
+      "learning_rate": 2.2926754808146865e-06,
+      "loss": 0.3997,
+      "step": 977
+    },
+    {
+      "epoch": 2.4088669950738915,
+      "grad_norm": 1.3958444262770742,
+      "learning_rate": 2.27444353945945e-06,
+      "loss": 0.476,
+      "step": 978
+    },
+    {
+      "epoch": 2.4113300492610836,
+      "grad_norm": 1.4303832315562959,
+      "learning_rate": 2.2562750738936155e-06,
+      "loss": 0.4577,
+      "step": 979
+    },
+    {
+      "epoch": 2.413793103448276,
+      "grad_norm": 1.3703616543417236,
+      "learning_rate": 2.2381702333954436e-06,
+      "loss": 0.4663,
+      "step": 980
+    },
+    {
+      "epoch": 2.416256157635468,
+      "grad_norm": 1.6574867237646813,
+      "learning_rate": 2.2201291667204463e-06,
+      "loss": 0.4068,
+      "step": 981
+    },
+    {
+      "epoch": 2.41871921182266,
+      "grad_norm": 1.3490464268889546,
+      "learning_rate": 2.2021520221001304e-06,
+      "loss": 0.4121,
+      "step": 982
+    },
+    {
+      "epoch": 2.4211822660098523,
+      "grad_norm": 1.3141040536885171,
+      "learning_rate": 2.1842389472408133e-06,
+      "loss": 0.4159,
+      "step": 983
+    },
+    {
+      "epoch": 2.4236453201970445,
+      "grad_norm": 1.4427159259839646,
+      "learning_rate": 2.16639008932239e-06,
+      "loss": 0.4554,
+      "step": 984
+    },
+    {
+      "epoch": 2.4261083743842367,
+      "grad_norm": 1.545319208298366,
+      "learning_rate": 2.148605594997115e-06,
+      "loss": 0.4607,
+      "step": 985
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 1.4101543056654273,
+      "learning_rate": 2.130885610388428e-06,
+      "loss": 0.4611,
+      "step": 986
+    },
+    {
+      "epoch": 2.4310344827586206,
+      "grad_norm": 1.408589768999158,
+      "learning_rate": 2.113230281089722e-06,
+      "loss": 0.5082,
+      "step": 987
+    },
+    {
+      "epoch": 2.4334975369458127,
+      "grad_norm": 1.4867698223664976,
+      "learning_rate": 2.0956397521631666e-06,
+      "loss": 0.3851,
+      "step": 988
+    },
+    {
+      "epoch": 2.435960591133005,
+      "grad_norm": 1.4481977377492585,
+      "learning_rate": 2.0781141681385053e-06,
+      "loss": 0.4593,
+      "step": 989
+    },
+    {
+      "epoch": 2.438423645320197,
+      "grad_norm": 1.3773299330669084,
+      "learning_rate": 2.0606536730118767e-06,
+      "loss": 0.4188,
+      "step": 990
+    },
+    {
+      "epoch": 2.4408866995073892,
+      "grad_norm": 1.3379429270969756,
+      "learning_rate": 2.043258410244622e-06,
+      "loss": 0.4046,
+      "step": 991
+    },
+    {
+      "epoch": 2.4433497536945814,
+      "grad_norm": 1.3034613781216384,
+      "learning_rate": 2.0259285227621152e-06,
+      "loss": 0.3752,
+      "step": 992
+    },
+    {
+      "epoch": 2.4458128078817736,
+      "grad_norm": 1.4005214997111823,
+      "learning_rate": 2.008664152952583e-06,
+      "loss": 0.4352,
+      "step": 993
+    },
+    {
+      "epoch": 2.4482758620689653,
+      "grad_norm": 1.5390251658602347,
+      "learning_rate": 1.9914654426659374e-06,
+      "loss": 0.5227,
+      "step": 994
+    },
+    {
+      "epoch": 2.4507389162561575,
+      "grad_norm": 1.4120408248618947,
+      "learning_rate": 1.974332533212611e-06,
+      "loss": 0.4376,
+      "step": 995
+    },
+    {
+      "epoch": 2.4532019704433496,
+      "grad_norm": 1.3534741415957616,
+      "learning_rate": 1.9572655653623884e-06,
+      "loss": 0.3917,
+      "step": 996
+    },
+    {
+      "epoch": 2.455665024630542,
+      "grad_norm": 1.550754037613496,
+      "learning_rate": 1.9402646793432624e-06,
+      "loss": 0.4496,
+      "step": 997
+    },
+    {
+      "epoch": 2.458128078817734,
+      "grad_norm": 1.4863648579252822,
+      "learning_rate": 1.9233300148402767e-06,
+      "loss": 0.4432,
+      "step": 998
+    },
+    {
+      "epoch": 2.460591133004926,
+      "grad_norm": 1.3251851474313068,
+      "learning_rate": 1.9064617109943662e-06,
+      "loss": 0.4588,
+      "step": 999
+    },
+    {
+      "epoch": 2.4630541871921183,
+      "grad_norm": 1.4516448819832148,
+      "learning_rate": 1.88965990640123e-06,
+      "loss": 0.4167,
+      "step": 1000
+    },
+    {
+      "epoch": 2.4655172413793105,
+      "grad_norm": 1.339371445046325,
+      "learning_rate": 1.8729247391101958e-06,
+      "loss": 0.4513,
+      "step": 1001
+    },
+    {
+      "epoch": 2.4679802955665027,
+      "grad_norm": 1.3197724180115045,
+      "learning_rate": 1.8562563466230577e-06,
+      "loss": 0.4873,
+      "step": 1002
+    },
+    {
+      "epoch": 2.4704433497536944,
+      "grad_norm": 1.4044603616044546,
+      "learning_rate": 1.8396548658929858e-06,
+      "loss": 0.4649,
+      "step": 1003
+    },
+    {
+      "epoch": 2.4729064039408866,
+      "grad_norm": 1.4038106742511758,
+      "learning_rate": 1.823120433323361e-06,
+      "loss": 0.464,
+      "step": 1004
+    },
+    {
+      "epoch": 2.4753694581280787,
+      "grad_norm": 1.349029195487974,
+      "learning_rate": 1.8066531847666891e-06,
+      "loss": 0.4926,
+      "step": 1005
+    },
+    {
+      "epoch": 2.477832512315271,
+      "grad_norm": 1.2537116259162797,
+      "learning_rate": 1.7902532555234653e-06,
+      "loss": 0.3811,
+      "step": 1006
+    },
+    {
+      "epoch": 2.480295566502463,
+      "grad_norm": 1.2774057635731395,
+      "learning_rate": 1.7739207803410575e-06,
+      "loss": 0.461,
+      "step": 1007
+    },
+    {
+      "epoch": 2.4827586206896552,
+      "grad_norm": 1.3408082654485425,
+      "learning_rate": 1.757655893412622e-06,
+      "loss": 0.4722,
+      "step": 1008
+    },
+    {
+      "epoch": 2.4852216748768474,
+      "grad_norm": 1.3890623678372538,
+      "learning_rate": 1.7414587283759776e-06,
+      "loss": 0.4164,
+      "step": 1009
+    },
+    {
+      "epoch": 2.4876847290640396,
+      "grad_norm": 1.4576147307163487,
+      "learning_rate": 1.7253294183125223e-06,
+      "loss": 0.3919,
+      "step": 1010
+    },
+    {
+      "epoch": 2.4901477832512313,
+      "grad_norm": 1.3055433104777268,
+      "learning_rate": 1.7092680957461294e-06,
+      "loss": 0.4028,
+      "step": 1011
+    },
+    {
+      "epoch": 2.4926108374384235,
+      "grad_norm": 1.3368502170428882,
+      "learning_rate": 1.6932748926420695e-06,
+      "loss": 0.3992,
+      "step": 1012
+    },
+    {
+      "epoch": 2.4950738916256157,
+      "grad_norm": 1.4901190196895604,
+      "learning_rate": 1.6773499404059157e-06,
+      "loss": 0.4728,
+      "step": 1013
+    },
+    {
+      "epoch": 2.497536945812808,
+      "grad_norm": 1.5122569306577867,
+      "learning_rate": 1.661493369882473e-06,
+      "loss": 0.4247,
+      "step": 1014
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.4714052264601012,
+      "learning_rate": 1.6457053113546972e-06,
+      "loss": 0.4547,
+      "step": 1015
+    },
+    {
+      "epoch": 2.502463054187192,
+      "grad_norm": 1.4859293780019633,
+      "learning_rate": 1.6299858945426251e-06,
+      "loss": 0.4518,
+      "step": 1016
+    },
+    {
+      "epoch": 2.5049261083743843,
+      "grad_norm": 1.486632547771064,
+      "learning_rate": 1.6143352486023113e-06,
+      "loss": 0.402,
+      "step": 1017
+    },
+    {
+      "epoch": 2.5073891625615765,
+      "grad_norm": 1.427110611406604,
+      "learning_rate": 1.5987535021247668e-06,
+      "loss": 0.4676,
+      "step": 1018
+    },
+    {
+      "epoch": 2.5098522167487687,
+      "grad_norm": 1.3565824039794343,
+      "learning_rate": 1.583240783134896e-06,
+      "loss": 0.4373,
+      "step": 1019
+    },
+    {
+      "epoch": 2.512315270935961,
+      "grad_norm": 1.388950797396656,
+      "learning_rate": 1.5677972190904623e-06,
+      "loss": 0.4437,
+      "step": 1020
+    },
+    {
+      "epoch": 2.5147783251231526,
+      "grad_norm": 1.4626865626616046,
+      "learning_rate": 1.5524229368810129e-06,
+      "loss": 0.4564,
+      "step": 1021
+    },
+    {
+      "epoch": 2.5172413793103448,
+      "grad_norm": 1.3078143175629777,
+      "learning_rate": 1.537118062826859e-06,
+      "loss": 0.3984,
+      "step": 1022
+    },
+    {
+      "epoch": 2.519704433497537,
+      "grad_norm": 1.4062068932602083,
+      "learning_rate": 1.5218827226780397e-06,
+      "loss": 0.4528,
+      "step": 1023
+    },
+    {
+      "epoch": 2.522167487684729,
+      "grad_norm": 1.2604439262195677,
+      "learning_rate": 1.5067170416132603e-06,
+      "loss": 0.4476,
+      "step": 1024
+    },
+    {
+      "epoch": 2.5246305418719213,
+      "grad_norm": 1.3101537714680547,
+      "learning_rate": 1.491621144238905e-06,
+      "loss": 0.4145,
+      "step": 1025
+    },
+    {
+      "epoch": 2.5270935960591134,
+      "grad_norm": 1.5517058798932588,
+      "learning_rate": 1.4765951545879732e-06,
+      "loss": 0.502,
+      "step": 1026
+    },
+    {
+      "epoch": 2.529556650246305,
+      "grad_norm": 1.3144508577188376,
+      "learning_rate": 1.4616391961190924e-06,
+      "loss": 0.4445,
+      "step": 1027
+    },
+    {
+      "epoch": 2.5320197044334973,
+      "grad_norm": 1.31218852025624,
+      "learning_rate": 1.4467533917154842e-06,
+      "loss": 0.3828,
+      "step": 1028
+    },
+    {
+      "epoch": 2.5344827586206895,
+      "grad_norm": 1.3551984182117738,
+      "learning_rate": 1.4319378636839554e-06,
+      "loss": 0.4262,
+      "step": 1029
+    },
+    {
+      "epoch": 2.5369458128078817,
+      "grad_norm": 1.3617229573222573,
+      "learning_rate": 1.4171927337539103e-06,
+      "loss": 0.4686,
+      "step": 1030
+    },
+    {
+      "epoch": 2.539408866995074,
+      "grad_norm": 1.4355849903262687,
+      "learning_rate": 1.4025181230763307e-06,
+      "loss": 0.4173,
+      "step": 1031
+    },
+    {
+      "epoch": 2.541871921182266,
+      "grad_norm": 1.3098348624851555,
+      "learning_rate": 1.3879141522227878e-06,
+      "loss": 0.4374,
+      "step": 1032
+    },
+    {
+      "epoch": 2.544334975369458,
+      "grad_norm": 1.4008843098400272,
+      "learning_rate": 1.3733809411844567e-06,
+      "loss": 0.414,
+      "step": 1033
+    },
+    {
+      "epoch": 2.5467980295566504,
+      "grad_norm": 1.25062011595787,
+      "learning_rate": 1.3589186093711227e-06,
+      "loss": 0.3916,
+      "step": 1034
+    },
+    {
+      "epoch": 2.5492610837438425,
+      "grad_norm": 1.368087123801389,
+      "learning_rate": 1.3445272756102023e-06,
+      "loss": 0.4363,
+      "step": 1035
+    },
+    {
+      "epoch": 2.5517241379310347,
+      "grad_norm": 1.4968318064914437,
+      "learning_rate": 1.3302070581457716e-06,
+      "loss": 0.5321,
+      "step": 1036
+    },
+    {
+      "epoch": 2.554187192118227,
+      "grad_norm": 1.3476313491953373,
+      "learning_rate": 1.3159580746375887e-06,
+      "loss": 0.4289,
+      "step": 1037
+    },
+    {
+      "epoch": 2.5566502463054186,
+      "grad_norm": 1.30154671356043,
+      "learning_rate": 1.3017804421601298e-06,
+      "loss": 0.4328,
+      "step": 1038
+    },
+    {
+      "epoch": 2.5591133004926108,
+      "grad_norm": 1.4316207239035015,
+      "learning_rate": 1.287674277201626e-06,
+      "loss": 0.4165,
+      "step": 1039
+    },
+    {
+      "epoch": 2.561576354679803,
+      "grad_norm": 1.3989984676524476,
+      "learning_rate": 1.273639695663108e-06,
+      "loss": 0.3953,
+      "step": 1040
+    },
+    {
+      "epoch": 2.564039408866995,
+      "grad_norm": 1.2361979965982464,
+      "learning_rate": 1.2596768128574522e-06,
+      "loss": 0.4622,
+      "step": 1041
+    },
+    {
+      "epoch": 2.5665024630541873,
+      "grad_norm": 1.2907013618840952,
+      "learning_rate": 1.245785743508441e-06,
+      "loss": 0.3547,
+      "step": 1042
+    },
+    {
+      "epoch": 2.5689655172413794,
+      "grad_norm": 1.4727445621471031,
+      "learning_rate": 1.2319666017497977e-06,
+      "loss": 0.4452,
+      "step": 1043
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 1.519408076376249,
+      "learning_rate": 1.2182195011242747e-06,
+      "loss": 0.4428,
+      "step": 1044
+    },
+    {
+      "epoch": 2.5738916256157633,
+      "grad_norm": 1.3424765740735816,
+      "learning_rate": 1.2045445545827162e-06,
+      "loss": 0.4107,
+      "step": 1045
+    },
+    {
+      "epoch": 2.5763546798029555,
+      "grad_norm": 1.340550761623763,
+      "learning_rate": 1.1909418744831048e-06,
+      "loss": 0.3926,
+      "step": 1046
+    },
+    {
+      "epoch": 2.5788177339901477,
+      "grad_norm": 1.4558936374768694,
+      "learning_rate": 1.1774115725896806e-06,
+      "loss": 0.4602,
+      "step": 1047
+    },
+    {
+      "epoch": 2.58128078817734,
+      "grad_norm": 1.4657668803500534,
+      "learning_rate": 1.1639537600719764e-06,
+      "loss": 0.4226,
+      "step": 1048
+    },
+    {
+      "epoch": 2.583743842364532,
+      "grad_norm": 1.3345263428816123,
+      "learning_rate": 1.1505685475039486e-06,
+      "loss": 0.3708,
+      "step": 1049
+    },
+    {
+      "epoch": 2.586206896551724,
+      "grad_norm": 1.4749605076562247,
+      "learning_rate": 1.1372560448630377e-06,
+      "loss": 0.4363,
+      "step": 1050
+    },
+    {
+      "epoch": 2.5886699507389164,
+      "grad_norm": 1.431981232264409,
+      "learning_rate": 1.12401636152927e-06,
+      "loss": 0.4795,
+      "step": 1051
+    },
+    {
+      "epoch": 2.5911330049261085,
+      "grad_norm": 1.3571379220485589,
+      "learning_rate": 1.1108496062843743e-06,
+      "loss": 0.4076,
+      "step": 1052
+    },
+    {
+      "epoch": 2.5935960591133007,
+      "grad_norm": 1.3520041830037657,
+      "learning_rate": 1.0977558873108717e-06,
+      "loss": 0.4183,
+      "step": 1053
+    },
+    {
+      "epoch": 2.596059113300493,
+      "grad_norm": 1.3058571055783685,
+      "learning_rate": 1.0847353121911952e-06,
+      "loss": 0.4331,
+      "step": 1054
+    },
+    {
+      "epoch": 2.5985221674876846,
+      "grad_norm": 1.4374198061923347,
+      "learning_rate": 1.0717879879068004e-06,
+      "loss": 0.4091,
+      "step": 1055
+    },
+    {
+      "epoch": 2.600985221674877,
+      "grad_norm": 1.3061693650322,
+      "learning_rate": 1.0589140208372872e-06,
+      "loss": 0.4089,
+      "step": 1056
+    },
+    {
+      "epoch": 2.603448275862069,
+      "grad_norm": 1.36698375818668,
+      "learning_rate": 1.0461135167595348e-06,
+      "loss": 0.4146,
+      "step": 1057
+    },
+    {
+      "epoch": 2.605911330049261,
+      "grad_norm": 1.537130547151082,
+      "learning_rate": 1.0333865808468203e-06,
+      "loss": 0.4389,
+      "step": 1058
+    },
+    {
+      "epoch": 2.6083743842364533,
+      "grad_norm": 1.6015974087516027,
+      "learning_rate": 1.0207333176679624e-06,
+      "loss": 0.4566,
+      "step": 1059
+    },
+    {
+      "epoch": 2.6108374384236455,
+      "grad_norm": 1.3859922413299193,
+      "learning_rate": 1.008153831186457e-06,
+      "loss": 0.432,
+      "step": 1060
+    },
+    {
+      "epoch": 2.613300492610837,
+      "grad_norm": 1.5522488584730805,
+      "learning_rate": 9.956482247596266e-07,
+      "loss": 0.4596,
+      "step": 1061
+    },
+    {
+      "epoch": 2.6157635467980294,
+      "grad_norm": 1.3239821501159612,
+      "learning_rate": 9.83216601137773e-07,
+      "loss": 0.4649,
+      "step": 1062
+    },
+    {
+      "epoch": 2.6182266009852215,
+      "grad_norm": 1.3842707127541798,
+      "learning_rate": 9.70859062463324e-07,
+      "loss": 0.4364,
+      "step": 1063
+    },
+    {
+      "epoch": 2.6206896551724137,
+      "grad_norm": 1.558492097909847,
+      "learning_rate": 9.58575710270011e-07,
+      "loss": 0.4847,
+      "step": 1064
+    },
+    {
+      "epoch": 2.623152709359606,
+      "grad_norm": 1.3737478189910985,
+      "learning_rate": 9.463666454820119e-07,
+      "loss": 0.437,
+      "step": 1065
+    },
+    {
+      "epoch": 2.625615763546798,
+      "grad_norm": 1.4758278770209496,
+      "learning_rate": 9.342319684131396e-07,
+      "loss": 0.4438,
+      "step": 1066
+    },
+    {
+      "epoch": 2.62807881773399,
+      "grad_norm": 1.3742717201385994,
+      "learning_rate": 9.221717787660145e-07,
+      "loss": 0.4192,
+      "step": 1067
+    },
+    {
+      "epoch": 2.6305418719211824,
+      "grad_norm": 1.391248413249548,
+      "learning_rate": 9.101861756312369e-07,
+      "loss": 0.4463,
+      "step": 1068
+    },
+    {
+      "epoch": 2.6330049261083746,
+      "grad_norm": 1.3232578628568767,
+      "learning_rate": 8.982752574865905e-07,
+      "loss": 0.3847,
+      "step": 1069
+    },
+    {
+      "epoch": 2.6354679802955667,
+      "grad_norm": 1.414081425829948,
+      "learning_rate": 8.864391221962065e-07,
+      "loss": 0.421,
+      "step": 1070
+    },
+    {
+      "epoch": 2.637931034482759,
+      "grad_norm": 1.3239040844855803,
+      "learning_rate": 8.746778670097877e-07,
+      "loss": 0.446,
+      "step": 1071
+    },
+    {
+      "epoch": 2.6403940886699506,
+      "grad_norm": 1.458269181429966,
+      "learning_rate": 8.629915885617912e-07,
+      "loss": 0.4243,
+      "step": 1072
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 1.3909862349162998,
+      "learning_rate": 8.513803828706369e-07,
+      "loss": 0.4255,
+      "step": 1073
+    },
+    {
+      "epoch": 2.645320197044335,
+      "grad_norm": 1.3319983604151855,
+      "learning_rate": 8.398443453379268e-07,
+      "loss": 0.4361,
+      "step": 1074
+    },
+    {
+      "epoch": 2.647783251231527,
+      "grad_norm": 1.3060574220765917,
+      "learning_rate": 8.283835707476551e-07,
+      "loss": 0.4402,
+      "step": 1075
+    },
+    {
+      "epoch": 2.6502463054187193,
+      "grad_norm": 1.7912238778913565,
+      "learning_rate": 8.169981532654269e-07,
+      "loss": 0.4523,
+      "step": 1076
+    },
+    {
+      "epoch": 2.6527093596059115,
+      "grad_norm": 1.4354069440463493,
+      "learning_rate": 8.05688186437693e-07,
+      "loss": 0.417,
+      "step": 1077
+    },
+    {
+      "epoch": 2.655172413793103,
+      "grad_norm": 1.6364704389830589,
+      "learning_rate": 7.944537631909666e-07,
+      "loss": 0.4122,
+      "step": 1078
+    },
+    {
+      "epoch": 2.6576354679802954,
+      "grad_norm": 1.3269293395822788,
+      "learning_rate": 7.832949758310804e-07,
+      "loss": 0.4222,
+      "step": 1079
+    },
+    {
+      "epoch": 2.6600985221674875,
+      "grad_norm": 1.394316541881774,
+      "learning_rate": 7.722119160424113e-07,
+      "loss": 0.4552,
+      "step": 1080
+    },
+    {
+      "epoch": 2.6625615763546797,
+      "grad_norm": 1.3563979863588622,
+      "learning_rate": 7.612046748871327e-07,
+      "loss": 0.4848,
+      "step": 1081
+    },
+    {
+      "epoch": 2.665024630541872,
+      "grad_norm": 1.3529428918942261,
+      "learning_rate": 7.502733428044684e-07,
+      "loss": 0.4341,
+      "step": 1082
+    },
+    {
+      "epoch": 2.667487684729064,
+      "grad_norm": 1.3470724416934157,
+      "learning_rate": 7.394180096099457e-07,
+      "loss": 0.4334,
+      "step": 1083
+    },
+    {
+      "epoch": 2.6699507389162562,
+      "grad_norm": 1.5532911154258444,
+      "learning_rate": 7.286387644946602e-07,
+      "loss": 0.4485,
+      "step": 1084
+    },
+    {
+      "epoch": 2.6724137931034484,
+      "grad_norm": 1.372281812338366,
+      "learning_rate": 7.179356960245409e-07,
+      "loss": 0.4638,
+      "step": 1085
+    },
+    {
+      "epoch": 2.6748768472906406,
+      "grad_norm": 1.3559481517815584,
+      "learning_rate": 7.073088921396287e-07,
+      "loss": 0.4693,
+      "step": 1086
+    },
+    {
+      "epoch": 2.6773399014778327,
+      "grad_norm": 1.4449092818033946,
+      "learning_rate": 6.9675844015334e-07,
+      "loss": 0.4967,
+      "step": 1087
+    },
+    {
+      "epoch": 2.6798029556650245,
+      "grad_norm": 1.3778805903089744,
+      "learning_rate": 6.862844267517643e-07,
+      "loss": 0.3764,
+      "step": 1088
+    },
+    {
+      "epoch": 2.6822660098522166,
+      "grad_norm": 1.3559112266356532,
+      "learning_rate": 6.758869379929456e-07,
+      "loss": 0.3792,
+      "step": 1089
+    },
+    {
+      "epoch": 2.684729064039409,
+      "grad_norm": 1.4387713905156232,
+      "learning_rate": 6.655660593061719e-07,
+      "loss": 0.4379,
+      "step": 1090
+    },
+    {
+      "epoch": 2.687192118226601,
+      "grad_norm": 1.5256195873397334,
+      "learning_rate": 6.553218754912838e-07,
+      "loss": 0.457,
+      "step": 1091
+    },
+    {
+      "epoch": 2.689655172413793,
+      "grad_norm": 1.5205581450905843,
+      "learning_rate": 6.451544707179635e-07,
+      "loss": 0.4619,
+      "step": 1092
+    },
+    {
+      "epoch": 2.6921182266009853,
+      "grad_norm": 1.474272309236532,
+      "learning_rate": 6.350639285250515e-07,
+      "loss": 0.3903,
+      "step": 1093
+    },
+    {
+      "epoch": 2.6945812807881775,
+      "grad_norm": 1.3300532536134841,
+      "learning_rate": 6.250503318198664e-07,
+      "loss": 0.4552,
+      "step": 1094
+    },
+    {
+      "epoch": 2.697044334975369,
+      "grad_norm": 1.4139848129260824,
+      "learning_rate": 6.151137628775051e-07,
+      "loss": 0.4698,
+      "step": 1095
+    },
+    {
+      "epoch": 2.6995073891625614,
+      "grad_norm": 1.226025554947605,
+      "learning_rate": 6.052543033401892e-07,
+      "loss": 0.3841,
+      "step": 1096
+    },
+    {
+      "epoch": 2.7019704433497536,
+      "grad_norm": 1.4122996854950705,
+      "learning_rate": 5.954720342165787e-07,
+      "loss": 0.4281,
+      "step": 1097
+    },
+    {
+      "epoch": 2.7044334975369457,
+      "grad_norm": 1.446119176145452,
+      "learning_rate": 5.857670358811096e-07,
+      "loss": 0.3884,
+      "step": 1098
+    },
+    {
+      "epoch": 2.706896551724138,
+      "grad_norm": 1.3038488193707176,
+      "learning_rate": 5.761393880733379e-07,
+      "loss": 0.4048,
+      "step": 1099
+    },
+    {
+      "epoch": 2.70935960591133,
+      "grad_norm": 1.420186224130169,
+      "learning_rate": 5.665891698972769e-07,
+      "loss": 0.4686,
+      "step": 1100
+    },
+    {
+      "epoch": 2.7118226600985222,
+      "grad_norm": 1.3963315653391744,
+      "learning_rate": 5.571164598207579e-07,
+      "loss": 0.4772,
+      "step": 1101
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 1.306261358257357,
+      "learning_rate": 5.477213356747746e-07,
+      "loss": 0.4153,
+      "step": 1102
+    },
+    {
+      "epoch": 2.7167487684729066,
+      "grad_norm": 1.4008098065369066,
+      "learning_rate": 5.384038746528519e-07,
+      "loss": 0.4318,
+      "step": 1103
+    },
+    {
+      "epoch": 2.7192118226600988,
+      "grad_norm": 1.3047796433855312,
+      "learning_rate": 5.291641533104053e-07,
+      "loss": 0.4073,
+      "step": 1104
+    },
+    {
+      "epoch": 2.7216748768472905,
+      "grad_norm": 1.416061505856914,
+      "learning_rate": 5.200022475641154e-07,
+      "loss": 0.4531,
+      "step": 1105
+    },
+    {
+      "epoch": 2.7241379310344827,
+      "grad_norm": 1.3635390947792194,
+      "learning_rate": 5.109182326913053e-07,
+      "loss": 0.4688,
+      "step": 1106
+    },
+    {
+      "epoch": 2.726600985221675,
+      "grad_norm": 1.942327146657886,
+      "learning_rate": 5.019121833293161e-07,
+      "loss": 0.5167,
+      "step": 1107
+    },
+    {
+      "epoch": 2.729064039408867,
+      "grad_norm": 1.4336132910953188,
+      "learning_rate": 4.929841734749063e-07,
+      "loss": 0.3942,
+      "step": 1108
+    },
+    {
+      "epoch": 2.731527093596059,
+      "grad_norm": 1.4685779227239883,
+      "learning_rate": 4.841342764836243e-07,
+      "loss": 0.4152,
+      "step": 1109
+    },
+    {
+      "epoch": 2.7339901477832513,
+      "grad_norm": 1.52996421859389,
+      "learning_rate": 4.7536256506922507e-07,
+      "loss": 0.4611,
+      "step": 1110
+    },
+    {
+      "epoch": 2.7364532019704435,
+      "grad_norm": 1.3271448894618918,
+      "learning_rate": 4.666691113030608e-07,
+      "loss": 0.4948,
+      "step": 1111
+    },
+    {
+      "epoch": 2.7389162561576352,
+      "grad_norm": 1.369850094592236,
+      "learning_rate": 4.580539866134914e-07,
+      "loss": 0.4391,
+      "step": 1112
+    },
+    {
+      "epoch": 2.7413793103448274,
+      "grad_norm": 1.4211350076903349,
+      "learning_rate": 4.4951726178530387e-07,
+      "loss": 0.4468,
+      "step": 1113
+    },
+    {
+      "epoch": 2.7438423645320196,
+      "grad_norm": 1.4215463950267333,
+      "learning_rate": 4.410590069591192e-07,
+      "loss": 0.4379,
+      "step": 1114
+    },
+    {
+      "epoch": 2.7463054187192117,
+      "grad_norm": 1.3031888566506113,
+      "learning_rate": 4.3267929163082424e-07,
+      "loss": 0.4403,
+      "step": 1115
+    },
+    {
+      "epoch": 2.748768472906404,
+      "grad_norm": 1.3468251194756362,
+      "learning_rate": 4.2437818465100313e-07,
+      "loss": 0.4513,
+      "step": 1116
+    },
+    {
+      "epoch": 2.751231527093596,
+      "grad_norm": 1.5145994039502364,
+      "learning_rate": 4.1615575422435774e-07,
+      "loss": 0.4522,
+      "step": 1117
+    },
+    {
+      "epoch": 2.7536945812807883,
+      "grad_norm": 1.3776203676171312,
+      "learning_rate": 4.0801206790916815e-07,
+      "loss": 0.4349,
+      "step": 1118
+    },
+    {
+      "epoch": 2.7561576354679804,
+      "grad_norm": 1.43650020163057,
+      "learning_rate": 3.999471926167142e-07,
+      "loss": 0.4672,
+      "step": 1119
+    },
+    {
+      "epoch": 2.7586206896551726,
+      "grad_norm": 1.3270521430757325,
+      "learning_rate": 3.919611946107493e-07,
+      "loss": 0.4231,
+      "step": 1120
+    },
+    {
+      "epoch": 2.7610837438423648,
+      "grad_norm": 1.407470004370219,
+      "learning_rate": 3.840541395069375e-07,
+      "loss": 0.4416,
+      "step": 1121
+    },
+    {
+      "epoch": 2.7635467980295565,
+      "grad_norm": 1.3017281874471514,
+      "learning_rate": 3.762260922723182e-07,
+      "loss": 0.4185,
+      "step": 1122
+    },
+    {
+      "epoch": 2.7660098522167487,
+      "grad_norm": 1.474360524583935,
+      "learning_rate": 3.6847711722478476e-07,
+      "loss": 0.4903,
+      "step": 1123
+    },
+    {
+      "epoch": 2.768472906403941,
+      "grad_norm": 1.4457912167764524,
+      "learning_rate": 3.6080727803254003e-07,
+      "loss": 0.4746,
+      "step": 1124
+    },
+    {
+      "epoch": 2.770935960591133,
+      "grad_norm": 1.4057732939699066,
+      "learning_rate": 3.532166377135815e-07,
+      "loss": 0.4343,
+      "step": 1125
+    },
+    {
+      "epoch": 2.773399014778325,
+      "grad_norm": 1.3699173186753695,
+      "learning_rate": 3.457052586351817e-07,
+      "loss": 0.4489,
+      "step": 1126
+    },
+    {
+      "epoch": 2.7758620689655173,
+      "grad_norm": 1.4768301100654644,
+      "learning_rate": 3.3827320251337613e-07,
+      "loss": 0.4164,
+      "step": 1127
+    },
+    {
+      "epoch": 2.7783251231527095,
+      "grad_norm": 1.3425587097486684,
+      "learning_rate": 3.309205304124552e-07,
+      "loss": 0.4031,
+      "step": 1128
+    },
+    {
+      "epoch": 2.7807881773399012,
+      "grad_norm": 1.4025448625702555,
+      "learning_rate": 3.2364730274446533e-07,
+      "loss": 0.4656,
+      "step": 1129
+    },
+    {
+      "epoch": 2.7832512315270934,
+      "grad_norm": 1.312010310420369,
+      "learning_rate": 3.1645357926870957e-07,
+      "loss": 0.411,
+      "step": 1130
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 1.3175412500572485,
+      "learning_rate": 3.0933941909125573e-07,
+      "loss": 0.4566,
+      "step": 1131
+    },
+    {
+      "epoch": 2.7881773399014778,
+      "grad_norm": 1.2052023157460234,
+      "learning_rate": 3.0230488066445465e-07,
+      "loss": 0.4399,
+      "step": 1132
+    },
+    {
+      "epoch": 2.79064039408867,
+      "grad_norm": 1.3424640534914472,
+      "learning_rate": 2.95350021786458e-07,
+      "loss": 0.4757,
+      "step": 1133
+    },
+    {
+      "epoch": 2.793103448275862,
+      "grad_norm": 1.3679133640930732,
+      "learning_rate": 2.8847489960074136e-07,
+      "loss": 0.3832,
+      "step": 1134
+    },
+    {
+      "epoch": 2.7955665024630543,
+      "grad_norm": 1.3863631101371345,
+      "learning_rate": 2.81679570595641e-07,
+      "loss": 0.4106,
+      "step": 1135
+    },
+    {
+      "epoch": 2.7980295566502464,
+      "grad_norm": 1.386315730456819,
+      "learning_rate": 2.7496409060387973e-07,
+      "loss": 0.4553,
+      "step": 1136
+    },
+    {
+      "epoch": 2.8004926108374386,
+      "grad_norm": 1.4383553157966882,
+      "learning_rate": 2.683285148021164e-07,
+      "loss": 0.4475,
+      "step": 1137
+    },
+    {
+      "epoch": 2.802955665024631,
+      "grad_norm": 1.298956430164076,
+      "learning_rate": 2.6177289771049274e-07,
+      "loss": 0.3686,
+      "step": 1138
+    },
+    {
+      "epoch": 2.8054187192118225,
+      "grad_norm": 1.3945210346948373,
+      "learning_rate": 2.55297293192176e-07,
+      "loss": 0.4656,
+      "step": 1139
+    },
+    {
+      "epoch": 2.8078817733990147,
+      "grad_norm": 1.4422687544824127,
+      "learning_rate": 2.489017544529315e-07,
+      "loss": 0.4349,
+      "step": 1140
+    },
+    {
+      "epoch": 2.810344827586207,
+      "grad_norm": 1.4051201414009515,
+      "learning_rate": 2.425863340406676e-07,
+      "loss": 0.449,
+      "step": 1141
+    },
+    {
+      "epoch": 2.812807881773399,
+      "grad_norm": 1.511620491245075,
+      "learning_rate": 2.3635108384502003e-07,
+      "loss": 0.424,
+      "step": 1142
+    },
+    {
+      "epoch": 2.815270935960591,
+      "grad_norm": 1.3615308306926566,
+      "learning_rate": 2.3019605509691957e-07,
+      "loss": 0.4707,
+      "step": 1143
+    },
+    {
+      "epoch": 2.8177339901477834,
+      "grad_norm": 1.3122510673967738,
+      "learning_rate": 2.2412129836816287e-07,
+      "loss": 0.4608,
+      "step": 1144
+    },
+    {
+      "epoch": 2.8201970443349755,
+      "grad_norm": 1.5609500827224883,
+      "learning_rate": 2.181268635710143e-07,
+      "loss": 0.4626,
+      "step": 1145
+    },
+    {
+      "epoch": 2.8226600985221673,
+      "grad_norm": 1.3276734874890772,
+      "learning_rate": 2.1221279995777833e-07,
+      "loss": 0.4268,
+      "step": 1146
+    },
+    {
+      "epoch": 2.8251231527093594,
+      "grad_norm": 1.5550713760087707,
+      "learning_rate": 2.0637915612040537e-07,
+      "loss": 0.4504,
+      "step": 1147
+    },
+    {
+      "epoch": 2.8275862068965516,
+      "grad_norm": 1.4114060292517756,
+      "learning_rate": 2.0062597999009114e-07,
+      "loss": 0.4328,
+      "step": 1148
+    },
+    {
+      "epoch": 2.8300492610837438,
+      "grad_norm": 1.3296504964733822,
+      "learning_rate": 1.9495331883687906e-07,
+      "loss": 0.4492,
+      "step": 1149
+    },
+    {
+      "epoch": 2.832512315270936,
+      "grad_norm": 1.3967315241716747,
+      "learning_rate": 1.8936121926927508e-07,
+      "loss": 0.384,
+      "step": 1150
+    },
+    {
+      "epoch": 2.834975369458128,
+      "grad_norm": 1.3061573634541395,
+      "learning_rate": 1.838497272338635e-07,
+      "loss": 0.3879,
+      "step": 1151
+    },
+    {
+      "epoch": 2.8374384236453203,
+      "grad_norm": 1.4896924794719835,
+      "learning_rate": 1.7841888801493178e-07,
+      "loss": 0.4503,
+      "step": 1152
+    },
+    {
+      "epoch": 2.8399014778325125,
+      "grad_norm": 1.4610870611163946,
+      "learning_rate": 1.7306874623409296e-07,
+      "loss": 0.4363,
+      "step": 1153
+    },
+    {
+      "epoch": 2.8423645320197046,
+      "grad_norm": 1.3263638207206525,
+      "learning_rate": 1.677993458499272e-07,
+      "loss": 0.4566,
+      "step": 1154
+    },
+    {
+      "epoch": 2.844827586206897,
+      "grad_norm": 1.4580139446193467,
+      "learning_rate": 1.6261073015761076e-07,
+      "loss": 0.4284,
+      "step": 1155
+    },
+    {
+      "epoch": 2.8472906403940885,
+      "grad_norm": 1.3351801013546014,
+      "learning_rate": 1.5750294178856872e-07,
+      "loss": 0.4884,
+      "step": 1156
+    },
+    {
+      "epoch": 2.8497536945812807,
+      "grad_norm": 1.384344254748622,
+      "learning_rate": 1.5247602271012296e-07,
+      "loss": 0.456,
+      "step": 1157
+    },
+    {
+      "epoch": 2.852216748768473,
+      "grad_norm": 1.4290403975748904,
+      "learning_rate": 1.4753001422514125e-07,
+      "loss": 0.4098,
+      "step": 1158
+    },
+    {
+      "epoch": 2.854679802955665,
+      "grad_norm": 1.3394080696527035,
+      "learning_rate": 1.4266495697170536e-07,
+      "loss": 0.4353,
+      "step": 1159
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 1.2953260892700496,
+      "learning_rate": 1.378808909227769e-07,
+      "loss": 0.4156,
+      "step": 1160
+    },
+    {
+      "epoch": 2.8596059113300494,
+      "grad_norm": 1.592838699996501,
+      "learning_rate": 1.331778553858598e-07,
+      "loss": 0.401,
+      "step": 1161
+    },
+    {
+      "epoch": 2.862068965517241,
+      "grad_norm": 1.292887059883994,
+      "learning_rate": 1.2855588900269057e-07,
+      "loss": 0.4166,
+      "step": 1162
+    },
+    {
+      "epoch": 2.8645320197044333,
+      "grad_norm": 1.334020298576052,
+      "learning_rate": 1.2401502974890735e-07,
+      "loss": 0.4077,
+      "step": 1163
+    },
+    {
+      "epoch": 2.8669950738916254,
+      "grad_norm": 1.4003479763543634,
+      "learning_rate": 1.1955531493375137e-07,
+      "loss": 0.4428,
+      "step": 1164
+    },
+    {
+      "epoch": 2.8694581280788176,
+      "grad_norm": 1.3371298304471708,
+      "learning_rate": 1.1517678119975062e-07,
+      "loss": 0.3542,
+      "step": 1165
+    },
+    {
+      "epoch": 2.87192118226601,
+      "grad_norm": 1.5132598513073474,
+      "learning_rate": 1.1087946452241871e-07,
+      "loss": 0.481,
+      "step": 1166
+    },
+    {
+      "epoch": 2.874384236453202,
+      "grad_norm": 1.4306721449483724,
+      "learning_rate": 1.0666340020996868e-07,
+      "loss": 0.4654,
+      "step": 1167
+    },
+    {
+      "epoch": 2.876847290640394,
+      "grad_norm": 1.300078468522143,
+      "learning_rate": 1.0252862290301092e-07,
+      "loss": 0.4041,
+      "step": 1168
+    },
+    {
+      "epoch": 2.8793103448275863,
+      "grad_norm": 1.359042203659106,
+      "learning_rate": 9.847516657427891e-08,
+      "loss": 0.3943,
+      "step": 1169
+    },
+    {
+      "epoch": 2.8817733990147785,
+      "grad_norm": 1.355797631548456,
+      "learning_rate": 9.45030645283418e-08,
+      "loss": 0.4422,
+      "step": 1170
+    },
+    {
+      "epoch": 2.8842364532019706,
+      "grad_norm": 1.3787583225860216,
+      "learning_rate": 9.061234940133668e-08,
+      "loss": 0.4256,
+      "step": 1171
+    },
+    {
+      "epoch": 2.886699507389163,
+      "grad_norm": 1.3829540460417096,
+      "learning_rate": 8.68030531606967e-08,
+      "loss": 0.4438,
+      "step": 1172
+    },
+    {
+      "epoch": 2.8891625615763545,
+      "grad_norm": 1.3190642104581738,
+      "learning_rate": 8.307520710489015e-08,
+      "loss": 0.3978,
+      "step": 1173
+    },
+    {
+      "epoch": 2.8916256157635467,
+      "grad_norm": 1.3711227022684378,
+      "learning_rate": 7.94288418631639e-08,
+      "loss": 0.4115,
+      "step": 1174
+    },
+    {
+      "epoch": 2.894088669950739,
+      "grad_norm": 1.4510508675522023,
+      "learning_rate": 7.586398739528933e-08,
+      "loss": 0.4698,
+      "step": 1175
+    },
+    {
+      "epoch": 2.896551724137931,
+      "grad_norm": 1.3184974428570875,
+      "learning_rate": 7.238067299131901e-08,
+      "loss": 0.389,
+      "step": 1176
+    },
+    {
+      "epoch": 2.899014778325123,
+      "grad_norm": 1.3653014334389277,
+      "learning_rate": 6.897892727134592e-08,
+      "loss": 0.4308,
+      "step": 1177
+    },
+    {
+      "epoch": 2.9014778325123154,
+      "grad_norm": 1.321809991631339,
+      "learning_rate": 6.565877818526245e-08,
+      "loss": 0.4117,
+      "step": 1178
+    },
+    {
+      "epoch": 2.903940886699507,
+      "grad_norm": 1.3555580571460746,
+      "learning_rate": 6.242025301254062e-08,
+      "loss": 0.3806,
+      "step": 1179
+    },
+    {
+      "epoch": 2.9064039408866993,
+      "grad_norm": 1.3860459105744272,
+      "learning_rate": 5.926337836199891e-08,
+      "loss": 0.4864,
+      "step": 1180
+    },
+    {
+      "epoch": 2.9088669950738915,
+      "grad_norm": 1.3401810024199852,
+      "learning_rate": 5.6188180171590225e-08,
+      "loss": 0.4078,
+      "step": 1181
+    },
+    {
+      "epoch": 2.9113300492610836,
+      "grad_norm": 1.3808336053194565,
+      "learning_rate": 5.319468370818537e-08,
+      "loss": 0.4417,
+      "step": 1182
+    },
+    {
+      "epoch": 2.913793103448276,
+      "grad_norm": 1.398845932630647,
+      "learning_rate": 5.028291356736326e-08,
+      "loss": 0.4235,
+      "step": 1183
+    },
+    {
+      "epoch": 2.916256157635468,
+      "grad_norm": 1.304228879076456,
+      "learning_rate": 4.7452893673216596e-08,
+      "loss": 0.4148,
+      "step": 1184
+    },
+    {
+      "epoch": 2.91871921182266,
+      "grad_norm": 1.373532769519456,
+      "learning_rate": 4.470464727814538e-08,
+      "loss": 0.497,
+      "step": 1185
+    },
+    {
+      "epoch": 2.9211822660098523,
+      "grad_norm": 1.3849509132720759,
+      "learning_rate": 4.203819696267486e-08,
+      "loss": 0.4583,
+      "step": 1186
+    },
+    {
+      "epoch": 2.9236453201970445,
+      "grad_norm": 1.4790204293759686,
+      "learning_rate": 3.945356463526451e-08,
+      "loss": 0.4058,
+      "step": 1187
+    },
+    {
+      "epoch": 2.9261083743842367,
+      "grad_norm": 1.3153126869192464,
+      "learning_rate": 3.6950771532126004e-08,
+      "loss": 0.4136,
+      "step": 1188
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 1.393226889730239,
+      "learning_rate": 3.452983821705891e-08,
+      "loss": 0.4095,
+      "step": 1189
+    },
+    {
+      "epoch": 2.9310344827586206,
+      "grad_norm": 1.352911127468965,
+      "learning_rate": 3.2190784581270786e-08,
+      "loss": 0.4081,
+      "step": 1190
+    },
+    {
+      "epoch": 2.9334975369458127,
+      "grad_norm": 1.3965701203598222,
+      "learning_rate": 2.993362984321735e-08,
+      "loss": 0.4549,
+      "step": 1191
+    },
+    {
+      "epoch": 2.935960591133005,
+      "grad_norm": 1.3673527644952992,
+      "learning_rate": 2.7758392548449253e-08,
+      "loss": 0.3992,
+      "step": 1192
+    },
+    {
+      "epoch": 2.938423645320197,
+      "grad_norm": 1.485424112708791,
+      "learning_rate": 2.566509056944999e-08,
+      "loss": 0.4147,
+      "step": 1193
+    },
+    {
+      "epoch": 2.9408866995073892,
+      "grad_norm": 1.1690074167987112,
+      "learning_rate": 2.3653741105499338e-08,
+      "loss": 0.4106,
+      "step": 1194
+    },
+    {
+      "epoch": 2.9433497536945814,
+      "grad_norm": 1.252783061343205,
+      "learning_rate": 2.1724360682527925e-08,
+      "loss": 0.4156,
+      "step": 1195
+    },
+    {
+      "epoch": 2.945812807881773,
+      "grad_norm": 1.389408951454765,
+      "learning_rate": 1.9876965152975102e-08,
+      "loss": 0.464,
+      "step": 1196
+    },
+    {
+      "epoch": 2.9482758620689653,
+      "grad_norm": 1.3968147793959094,
+      "learning_rate": 1.8111569695672403e-08,
+      "loss": 0.432,
+      "step": 1197
+    },
+    {
+      "epoch": 2.9507389162561575,
+      "grad_norm": 1.5930391109955138,
+      "learning_rate": 1.6428188815703627e-08,
+      "loss": 0.4468,
+      "step": 1198
+    },
+    {
+      "epoch": 2.9532019704433496,
+      "grad_norm": 1.3360863318274727,
+      "learning_rate": 1.4826836344299378e-08,
+      "loss": 0.4274,
+      "step": 1199
+    },
+    {
+      "epoch": 2.955665024630542,
+      "grad_norm": 1.3365993533383909,
+      "learning_rate": 1.3307525438711611e-08,
+      "loss": 0.4588,
+      "step": 1200
+    },
+    {
+      "epoch": 2.958128078817734,
+      "grad_norm": 1.3043736063095712,
+      "learning_rate": 1.187026858211482e-08,
+      "loss": 0.4493,
+      "step": 1201
+    },
+    {
+      "epoch": 2.960591133004926,
+      "grad_norm": 1.3765447349138107,
+      "learning_rate": 1.0515077583498346e-08,
+      "loss": 0.4329,
+      "step": 1202
+    },
+    {
+      "epoch": 2.9630541871921183,
+      "grad_norm": 1.5023486505786556,
+      "learning_rate": 9.241963577569791e-09,
+      "loss": 0.4085,
+      "step": 1203
+    },
+    {
+      "epoch": 2.9655172413793105,
+      "grad_norm": 1.4333440994627213,
+      "learning_rate": 8.050937024666195e-09,
+      "loss": 0.4496,
+      "step": 1204
+    },
+    {
+      "epoch": 2.9679802955665027,
+      "grad_norm": 1.3538625913735858,
+      "learning_rate": 6.942007710665222e-09,
+      "loss": 0.4102,
+      "step": 1205
+    },
+    {
+      "epoch": 2.970443349753695,
+      "grad_norm": 1.5283975307469613,
+      "learning_rate": 5.9151847469041125e-09,
+      "loss": 0.4659,
+      "step": 1206
+    },
+    {
+      "epoch": 2.9729064039408866,
+      "grad_norm": 1.3822731216133375,
+      "learning_rate": 4.970476570110849e-09,
+      "loss": 0.454,
+      "step": 1207
+    },
+    {
+      "epoch": 2.9753694581280787,
+      "grad_norm": 1.6025137704623584,
+      "learning_rate": 4.1078909423253325e-09,
+      "loss": 0.434,
+      "step": 1208
+    },
+    {
+      "epoch": 2.977832512315271,
+      "grad_norm": 1.2741444971281997,
+      "learning_rate": 3.327434950846087e-09,
+      "loss": 0.4034,
+      "step": 1209
+    },
+    {
+      "epoch": 2.980295566502463,
+      "grad_norm": 1.3053641477690985,
+      "learning_rate": 2.629115008160321e-09,
+      "loss": 0.43,
+      "step": 1210
+    },
+    {
+      "epoch": 2.9827586206896552,
+      "grad_norm": 1.3805692332839812,
+      "learning_rate": 2.0129368519050674e-09,
+      "loss": 0.4448,
+      "step": 1211
+    },
+    {
+      "epoch": 2.9852216748768474,
+      "grad_norm": 1.3762241029858107,
+      "learning_rate": 1.4789055448061195e-09,
+      "loss": 0.3958,
+      "step": 1212
+    },
+    {
+      "epoch": 2.987684729064039,
+      "grad_norm": 1.4598142007937973,
+      "learning_rate": 1.027025474648058e-09,
+      "loss": 0.3828,
+      "step": 1213
+    },
+    {
+      "epoch": 2.9901477832512313,
+      "grad_norm": 1.2336752634495343,
+      "learning_rate": 6.573003542276191e-10,
+      "loss": 0.4129,
+      "step": 1214
+    },
+    {
+      "epoch": 2.9926108374384235,
+      "grad_norm": 1.3455401184837332,
+      "learning_rate": 3.697332213348226e-10,
+      "loss": 0.4113,
+      "step": 1215
+    },
+    {
+      "epoch": 2.9950738916256157,
+      "grad_norm": 1.5424282237808498,
+      "learning_rate": 1.6432643871633346e-10,
+      "loss": 0.4441,
+      "step": 1216
+    },
+    {
+      "epoch": 2.997536945812808,
+      "grad_norm": 1.3758637143082293,
+      "learning_rate": 4.108169406435991e-11,
+      "loss": 0.4782,
+      "step": 1217
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.4103477538199638,
+      "learning_rate": 0.0,
+      "loss": 0.4558,
+      "step": 1218
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.3741660118103027,
+      "eval_runtime": 1.0549,
+      "eval_samples_per_second": 50.243,
+      "eval_steps_per_second": 1.896,
+      "step": 1218
+    },
+    {
+      "epoch": 3.0,
+      "step": 1218,
+      "total_flos": 74579642941440.0,
+      "train_loss": 0.8193829917212817,
+      "train_runtime": 13183.5138,
+      "train_samples_per_second": 11.821,
+      "train_steps_per_second": 0.092
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1218,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 609,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 74579642941440.0,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}