diff --git "a/limo/full/checkpoint-1545/trainer_state.json" "b/limo/full/checkpoint-1545/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/limo/full/checkpoint-1545/trainer_state.json"
@@ -0,0 +1,10848 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 15.0,
+  "eval_steps": 500,
+  "global_step": 1545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.009708737864077669,
+      "grad_norm": 32.0991353240983,
+      "learning_rate": 4.999994831641374e-06,
+      "loss": 4.1052,
+      "step": 1
+    },
+    {
+      "epoch": 0.019417475728155338,
+      "grad_norm": 29.524023707455505,
+      "learning_rate": 4.9999793265868636e-06,
+      "loss": 3.2535,
+      "step": 2
+    },
+    {
+      "epoch": 0.02912621359223301,
+      "grad_norm": 25.915328826410054,
+      "learning_rate": 4.999953484900578e-06,
+      "loss": 3.0603,
+      "step": 3
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 26.925318996290773,
+      "learning_rate": 4.9999173066893655e-06,
+      "loss": 3.8736,
+      "step": 4
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 17.49179072124481,
+      "learning_rate": 4.9998707921028104e-06,
+      "loss": 3.2503,
+      "step": 5
+    },
+    {
+      "epoch": 0.05825242718446602,
+      "grad_norm": 11.174584909912863,
+      "learning_rate": 4.999813941333237e-06,
+      "loss": 2.0295,
+      "step": 6
+    },
+    {
+      "epoch": 0.06796116504854369,
+      "grad_norm": 9.90835381349434,
+      "learning_rate": 4.999746754615704e-06,
+      "loss": 1.9601,
+      "step": 7
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 2.252755813363907,
+      "learning_rate": 4.9996692322280085e-06,
+      "loss": 1.1375,
+      "step": 8
+    },
+    {
+      "epoch": 0.08737864077669903,
+      "grad_norm": 9.417040995705188,
+      "learning_rate": 4.999581374490681e-06,
+      "loss": 2.4889,
+      "step": 9
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 5.128297805941186,
+      "learning_rate": 4.999483181766986e-06,
+      "loss": 1.9038,
+      "step": 10
+    },
+    {
+      "epoch": 0.10679611650485436,
+      "grad_norm": 2.6424619563549627,
+      "learning_rate": 4.999374654462919e-06,
+      "loss": 1.3898,
+      "step": 11
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 5.089573735718794,
+      "learning_rate": 4.999255793027207e-06,
+      "loss": 2.1286,
+      "step": 12
+    },
+    {
+      "epoch": 0.1262135922330097,
+      "grad_norm": 2.251991573691883,
+      "learning_rate": 4.999126597951305e-06,
+      "loss": 1.0427,
+      "step": 13
+    },
+    {
+      "epoch": 0.13592233009708737,
+      "grad_norm": 2.2920823891389333,
+      "learning_rate": 4.998987069769394e-06,
+      "loss": 1.1071,
+      "step": 14
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 2.8377769456674247,
+      "learning_rate": 4.998837209058379e-06,
+      "loss": 1.6229,
+      "step": 15
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 7.727434690224115,
+      "learning_rate": 4.998677016437888e-06,
+      "loss": 1.3171,
+      "step": 16
+    },
+    {
+      "epoch": 0.1650485436893204,
+      "grad_norm": 3.74374589959768,
+      "learning_rate": 4.998506492570266e-06,
+      "loss": 1.7463,
+      "step": 17
+    },
+    {
+      "epoch": 0.17475728155339806,
+      "grad_norm": 3.837005912107839,
+      "learning_rate": 4.998325638160576e-06,
+      "loss": 1.3993,
+      "step": 18
+    },
+    {
+      "epoch": 0.18446601941747573,
+      "grad_norm": 5.797983985086783,
+      "learning_rate": 4.998134453956596e-06,
+      "loss": 1.1204,
+      "step": 19
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 2.7548345494912807,
+      "learning_rate": 4.997932940748811e-06,
+      "loss": 1.2081,
+      "step": 20
+    },
+    {
+      "epoch": 0.20388349514563106,
+      "grad_norm": 5.045371621372198,
+      "learning_rate": 4.997721099370416e-06,
+      "loss": 1.2124,
+      "step": 21
+    },
+    {
+      "epoch": 0.21359223300970873,
+      "grad_norm": 2.051110624952864,
+      "learning_rate": 4.997498930697308e-06,
+      "loss": 1.3137,
+      "step": 22
+    },
+    {
+      "epoch": 0.22330097087378642,
+      "grad_norm": 2.691454094294257,
+      "learning_rate": 4.997266435648086e-06,
+      "loss": 1.3098,
+      "step": 23
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 2.4865134695003017,
+      "learning_rate": 4.997023615184044e-06,
+      "loss": 1.6151,
+      "step": 24
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 2.8075788095017757,
+      "learning_rate": 4.996770470309167e-06,
+      "loss": 1.6531,
+      "step": 25
+    },
+    {
+      "epoch": 0.2524271844660194,
+      "grad_norm": 1.768040918463024,
+      "learning_rate": 4.996507002070131e-06,
+      "loss": 1.1902,
+      "step": 26
+    },
+    {
+      "epoch": 0.2621359223300971,
+      "grad_norm": 1.955363000769332,
+      "learning_rate": 4.996233211556295e-06,
+      "loss": 1.2454,
+      "step": 27
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 4.241739298469607,
+      "learning_rate": 4.9959490998996974e-06,
+      "loss": 0.8855,
+      "step": 28
+    },
+    {
+      "epoch": 0.2815533980582524,
+      "grad_norm": 7.222330025323884,
+      "learning_rate": 4.995654668275049e-06,
+      "loss": 1.2271,
+      "step": 29
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 2.3651169987619527,
+      "learning_rate": 4.995349917899735e-06,
+      "loss": 1.3329,
+      "step": 30
+    },
+    {
+      "epoch": 0.30097087378640774,
+      "grad_norm": 1.6948963634507868,
+      "learning_rate": 4.9950348500338005e-06,
+      "loss": 0.9562,
+      "step": 31
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.3136462562718183,
+      "learning_rate": 4.994709465979954e-06,
+      "loss": 0.9246,
+      "step": 32
+    },
+    {
+      "epoch": 0.32038834951456313,
+      "grad_norm": 3.6835006171200306,
+      "learning_rate": 4.994373767083556e-06,
+      "loss": 1.0724,
+      "step": 33
+    },
+    {
+      "epoch": 0.3300970873786408,
+      "grad_norm": 3.163430858725886,
+      "learning_rate": 4.994027754732616e-06,
+      "loss": 1.1139,
+      "step": 34
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 1.4746343819625023,
+      "learning_rate": 4.993671430357788e-06,
+      "loss": 0.9994,
+      "step": 35
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.9654224512814882,
+      "learning_rate": 4.99330479543236e-06,
+      "loss": 1.3667,
+      "step": 36
+    },
+    {
+      "epoch": 0.3592233009708738,
+      "grad_norm": 1.2495201306162087,
+      "learning_rate": 4.992927851472254e-06,
+      "loss": 1.1252,
+      "step": 37
+    },
+    {
+      "epoch": 0.36893203883495146,
+      "grad_norm": 1.5824371832462278,
+      "learning_rate": 4.992540600036014e-06,
+      "loss": 1.3831,
+      "step": 38
+    },
+    {
+      "epoch": 0.3786407766990291,
+      "grad_norm": 1.4702657605395504,
+      "learning_rate": 4.992143042724805e-06,
+      "loss": 1.1461,
+      "step": 39
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.4506087672413808,
+      "learning_rate": 4.991735181182401e-06,
+      "loss": 1.4195,
+      "step": 40
+    },
+    {
+      "epoch": 0.39805825242718446,
+      "grad_norm": 1.075865370590295,
+      "learning_rate": 4.991317017095182e-06,
+      "loss": 0.9813,
+      "step": 41
+    },
+    {
+      "epoch": 0.4077669902912621,
+      "grad_norm": 1.2265855895293323,
+      "learning_rate": 4.990888552192126e-06,
+      "loss": 1.0049,
+      "step": 42
+    },
+    {
+      "epoch": 0.4174757281553398,
+      "grad_norm": 2.4151552794703437,
+      "learning_rate": 4.9904497882448004e-06,
+      "loss": 1.1099,
+      "step": 43
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.6677746145821908,
+      "learning_rate": 4.990000727067357e-06,
+      "loss": 1.1481,
+      "step": 44
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 1.5396824114800838,
+      "learning_rate": 4.989541370516523e-06,
+      "loss": 0.9952,
+      "step": 45
+    },
+    {
+      "epoch": 0.44660194174757284,
+      "grad_norm": 1.3928969634273873,
+      "learning_rate": 4.989071720491595e-06,
+      "loss": 0.9147,
+      "step": 46
+    },
+    {
+      "epoch": 0.4563106796116505,
+      "grad_norm": 0.9492551299059091,
+      "learning_rate": 4.988591778934428e-06,
+      "loss": 0.6869,
+      "step": 47
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.2714456311245477,
+      "learning_rate": 4.9881015478294294e-06,
+      "loss": 0.8276,
+      "step": 48
+    },
+    {
+      "epoch": 0.47572815533980584,
+      "grad_norm": 2.6314825001613658,
+      "learning_rate": 4.987601029203553e-06,
+      "loss": 1.0247,
+      "step": 49
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 1.416971495316372,
+      "learning_rate": 4.987090225126285e-06,
+      "loss": 1.012,
+      "step": 50
+    },
+    {
+      "epoch": 0.49514563106796117,
+      "grad_norm": 1.361560829125586,
+      "learning_rate": 4.98656913770964e-06,
+      "loss": 0.8907,
+      "step": 51
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.3035984152044737,
+      "learning_rate": 4.986037769108154e-06,
+      "loss": 1.056,
+      "step": 52
+    },
+    {
+      "epoch": 0.5145631067961165,
+      "grad_norm": 1.0150205143363307,
+      "learning_rate": 4.9854961215188676e-06,
+      "loss": 1.0017,
+      "step": 53
+    },
+    {
+      "epoch": 0.5242718446601942,
+      "grad_norm": 0.9609277033036215,
+      "learning_rate": 4.984944197181324e-06,
+      "loss": 0.7601,
+      "step": 54
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 1.1898667428428673,
+      "learning_rate": 4.9843819983775575e-06,
+      "loss": 0.8858,
+      "step": 55
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.0139938825198491,
+      "learning_rate": 4.983809527432086e-06,
+      "loss": 0.8071,
+      "step": 56
+    },
+    {
+      "epoch": 0.5533980582524272,
+      "grad_norm": 1.247418732762796,
+      "learning_rate": 4.983226786711895e-06,
+      "loss": 0.9675,
+      "step": 57
+    },
+    {
+      "epoch": 0.5631067961165048,
+      "grad_norm": 0.8942637413037233,
+      "learning_rate": 4.982633778626437e-06,
+      "loss": 0.8187,
+      "step": 58
+    },
+    {
+      "epoch": 0.5728155339805825,
+      "grad_norm": 1.3943617986028647,
+      "learning_rate": 4.982030505627613e-06,
+      "loss": 1.0678,
+      "step": 59
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.8610131821728051,
+      "learning_rate": 4.98141697020977e-06,
+      "loss": 0.8306,
+      "step": 60
+    },
+    {
+      "epoch": 0.5922330097087378,
+      "grad_norm": 1.0768921632881472,
+      "learning_rate": 4.9807931749096836e-06,
+      "loss": 0.81,
+      "step": 61
+    },
+    {
+      "epoch": 0.6019417475728155,
+      "grad_norm": 0.9625381291373968,
+      "learning_rate": 4.980159122306551e-06,
+      "loss": 0.892,
+      "step": 62
+    },
+    {
+      "epoch": 0.6116504854368932,
+      "grad_norm": 0.9617828022691955,
+      "learning_rate": 4.979514815021984e-06,
+      "loss": 1.0243,
+      "step": 63
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.2606711145146943,
+      "learning_rate": 4.978860255719989e-06,
+      "loss": 0.7773,
+      "step": 64
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.8531956488376088,
+      "learning_rate": 4.978195447106965e-06,
+      "loss": 0.9458,
+      "step": 65
+    },
+    {
+      "epoch": 0.6407766990291263,
+      "grad_norm": 0.8576435230419439,
+      "learning_rate": 4.9775203919316864e-06,
+      "loss": 0.7812,
+      "step": 66
+    },
+    {
+      "epoch": 0.6504854368932039,
+      "grad_norm": 0.9030199864232913,
+      "learning_rate": 4.976835092985297e-06,
+      "loss": 0.9382,
+      "step": 67
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 0.8483972987027173,
+      "learning_rate": 4.976139553101291e-06,
+      "loss": 0.8671,
+      "step": 68
+    },
+    {
+      "epoch": 0.6699029126213593,
+      "grad_norm": 0.8960545150162272,
+      "learning_rate": 4.975433775155509e-06,
+      "loss": 0.8646,
+      "step": 69
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 1.0017538898162217,
+      "learning_rate": 4.974717762066123e-06,
+      "loss": 0.8805,
+      "step": 70
+    },
+    {
+      "epoch": 0.6893203883495146,
+      "grad_norm": 0.9212309895628247,
+      "learning_rate": 4.973991516793621e-06,
+      "loss": 1.0576,
+      "step": 71
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 0.9305642899936102,
+      "learning_rate": 4.973255042340801e-06,
+      "loss": 0.8486,
+      "step": 72
+    },
+    {
+      "epoch": 0.7087378640776699,
+      "grad_norm": 1.0773851273416222,
+      "learning_rate": 4.972508341752754e-06,
+      "loss": 1.0583,
+      "step": 73
+    },
+    {
+      "epoch": 0.7184466019417476,
+      "grad_norm": 0.892091799769842,
+      "learning_rate": 4.9717514181168534e-06,
+      "loss": 0.7527,
+      "step": 74
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.8481741026230679,
+      "learning_rate": 4.970984274562741e-06,
+      "loss": 0.7125,
+      "step": 75
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 0.9312066959231862,
+      "learning_rate": 4.970206914262315e-06,
+      "loss": 0.8687,
+      "step": 76
+    },
+    {
+      "epoch": 0.7475728155339806,
+      "grad_norm": 0.9215979104056321,
+      "learning_rate": 4.969419340429717e-06,
+      "loss": 0.7691,
+      "step": 77
+    },
+    {
+      "epoch": 0.7572815533980582,
+      "grad_norm": 1.227175476240732,
+      "learning_rate": 4.968621556321319e-06,
+      "loss": 0.92,
+      "step": 78
+    },
+    {
+      "epoch": 0.7669902912621359,
+      "grad_norm": 0.9330601553607981,
+      "learning_rate": 4.967813565235708e-06,
+      "loss": 0.8216,
+      "step": 79
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.8252798793274818,
+      "learning_rate": 4.966995370513675e-06,
+      "loss": 0.7061,
+      "step": 80
+    },
+    {
+      "epoch": 0.7864077669902912,
+      "grad_norm": 0.9922301648204198,
+      "learning_rate": 4.966166975538197e-06,
+      "loss": 1.0408,
+      "step": 81
+    },
+    {
+      "epoch": 0.7961165048543689,
+      "grad_norm": 0.7891923797199208,
+      "learning_rate": 4.965328383734429e-06,
+      "loss": 0.7595,
+      "step": 82
+    },
+    {
+      "epoch": 0.8058252427184466,
+      "grad_norm": 1.0034345889698408,
+      "learning_rate": 4.964479598569686e-06,
+      "loss": 1.0233,
+      "step": 83
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.9634692778615994,
+      "learning_rate": 4.963620623553428e-06,
+      "loss": 0.929,
+      "step": 84
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.8794808183269647,
+      "learning_rate": 4.962751462237248e-06,
+      "loss": 0.7247,
+      "step": 85
+    },
+    {
+      "epoch": 0.8349514563106796,
+      "grad_norm": 0.8991558039524955,
+      "learning_rate": 4.9618721182148564e-06,
+      "loss": 0.6248,
+      "step": 86
+    },
+    {
+      "epoch": 0.8446601941747572,
+      "grad_norm": 0.9291777621093169,
+      "learning_rate": 4.960982595122064e-06,
+      "loss": 0.7035,
+      "step": 87
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 0.85791569259972,
+      "learning_rate": 4.960082896636773e-06,
+      "loss": 0.9148,
+      "step": 88
+    },
+    {
+      "epoch": 0.8640776699029126,
+      "grad_norm": 0.9063445307146435,
+      "learning_rate": 4.959173026478952e-06,
+      "loss": 0.7805,
+      "step": 89
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.7587205124423355,
+      "learning_rate": 4.958252988410631e-06,
+      "loss": 0.6329,
+      "step": 90
+    },
+    {
+      "epoch": 0.883495145631068,
+      "grad_norm": 1.0713074090013364,
+      "learning_rate": 4.9573227862358794e-06,
+      "loss": 0.7955,
+      "step": 91
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 0.8477622625159322,
+      "learning_rate": 4.956382423800791e-06,
+      "loss": 0.8325,
+      "step": 92
+    },
+    {
+      "epoch": 0.9029126213592233,
+      "grad_norm": 0.9269059925838263,
+      "learning_rate": 4.955431904993471e-06,
+      "loss": 0.8194,
+      "step": 93
+    },
+    {
+      "epoch": 0.912621359223301,
+      "grad_norm": 1.0475028792580197,
+      "learning_rate": 4.954471233744015e-06,
+      "loss": 0.5835,
+      "step": 94
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.8985260034505762,
+      "learning_rate": 4.9535004140245005e-06,
+      "loss": 0.8063,
+      "step": 95
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 1.0048572531965805,
+      "learning_rate": 4.952519449848962e-06,
+      "loss": 0.8127,
+      "step": 96
+    },
+    {
+      "epoch": 0.941747572815534,
+      "grad_norm": 0.9062750156316196,
+      "learning_rate": 4.951528345273379e-06,
+      "loss": 0.7181,
+      "step": 97
+    },
+    {
+      "epoch": 0.9514563106796117,
+      "grad_norm": 0.8778949835317967,
+      "learning_rate": 4.950527104395659e-06,
+      "loss": 0.8103,
+      "step": 98
+    },
+    {
+      "epoch": 0.9611650485436893,
+      "grad_norm": 0.9728187484090823,
+      "learning_rate": 4.9495157313556185e-06,
+      "loss": 0.6329,
+      "step": 99
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.7352370132544686,
+      "learning_rate": 4.94849423033497e-06,
+      "loss": 0.5231,
+      "step": 100
+    },
+    {
+      "epoch": 0.9805825242718447,
+      "grad_norm": 0.8479762742075609,
+      "learning_rate": 4.9474626055573e-06,
+      "loss": 0.9551,
+      "step": 101
+    },
+    {
+      "epoch": 0.9902912621359223,
+      "grad_norm": 0.7351142523348447,
+      "learning_rate": 4.946420861288051e-06,
+      "loss": 0.738,
+      "step": 102
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9009004817884025,
+      "learning_rate": 4.9453690018345144e-06,
+      "loss": 0.7789,
+      "step": 103
+    },
+    {
+      "epoch": 1.0097087378640777,
+      "grad_norm": 0.8057060436062744,
+      "learning_rate": 4.944307031545797e-06,
+      "loss": 0.6629,
+      "step": 104
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.8977736266341005,
+      "learning_rate": 4.943234954812812e-06,
+      "loss": 0.9053,
+      "step": 105
+    },
+    {
+      "epoch": 1.029126213592233,
+      "grad_norm": 0.7717440949849311,
+      "learning_rate": 4.942152776068264e-06,
+      "loss": 0.6404,
+      "step": 106
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 0.890503065931168,
+      "learning_rate": 4.941060499786622e-06,
+      "loss": 0.9117,
+      "step": 107
+    },
+    {
+      "epoch": 1.0485436893203883,
+      "grad_norm": 0.8814259279520221,
+      "learning_rate": 4.939958130484106e-06,
+      "loss": 0.7055,
+      "step": 108
+    },
+    {
+      "epoch": 1.058252427184466,
+      "grad_norm": 0.671081647050562,
+      "learning_rate": 4.938845672718668e-06,
+      "loss": 0.6671,
+      "step": 109
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 0.8278909841687843,
+      "learning_rate": 4.937723131089974e-06,
+      "loss": 0.7318,
+      "step": 110
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 0.9269878913315724,
+      "learning_rate": 4.93659051023938e-06,
+      "loss": 0.7646,
+      "step": 111
+    },
+    {
+      "epoch": 1.087378640776699,
+      "grad_norm": 0.665235455755867,
+      "learning_rate": 4.93544781484992e-06,
+      "loss": 0.4998,
+      "step": 112
+    },
+    {
+      "epoch": 1.0970873786407767,
+      "grad_norm": 0.8487063874467249,
+      "learning_rate": 4.9342950496462815e-06,
+      "loss": 0.7435,
+      "step": 113
+    },
+    {
+      "epoch": 1.1067961165048543,
+      "grad_norm": 0.8430924624096062,
+      "learning_rate": 4.933132219394786e-06,
+      "loss": 0.6992,
+      "step": 114
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.8993084000328078,
+      "learning_rate": 4.931959328903376e-06,
+      "loss": 0.8961,
+      "step": 115
+    },
+    {
+      "epoch": 1.1262135922330097,
+      "grad_norm": 0.7373642852662377,
+      "learning_rate": 4.930776383021584e-06,
+      "loss": 0.6722,
+      "step": 116
+    },
+    {
+      "epoch": 1.1359223300970873,
+      "grad_norm": 0.8356211654383496,
+      "learning_rate": 4.92958338664052e-06,
+      "loss": 0.528,
+      "step": 117
+    },
+    {
+      "epoch": 1.145631067961165,
+      "grad_norm": 0.8788366905889159,
+      "learning_rate": 4.928380344692853e-06,
+      "loss": 0.6369,
+      "step": 118
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 0.70400203786099,
+      "learning_rate": 4.927167262152784e-06,
+      "loss": 0.6961,
+      "step": 119
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.7933675823948615,
+      "learning_rate": 4.925944144036027e-06,
+      "loss": 0.7316,
+      "step": 120
+    },
+    {
+      "epoch": 1.174757281553398,
+      "grad_norm": 0.7413576987999055,
+      "learning_rate": 4.924710995399796e-06,
+      "loss": 0.6764,
+      "step": 121
+    },
+    {
+      "epoch": 1.1844660194174756,
+      "grad_norm": 0.8385780569550869,
+      "learning_rate": 4.923467821342773e-06,
+      "loss": 0.7602,
+      "step": 122
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 0.8531356062842188,
+      "learning_rate": 4.922214627005092e-06,
+      "loss": 0.749,
+      "step": 123
+    },
+    {
+      "epoch": 1.203883495145631,
+      "grad_norm": 0.8216291410900652,
+      "learning_rate": 4.920951417568323e-06,
+      "loss": 0.8079,
+      "step": 124
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.8320005263963047,
+      "learning_rate": 4.919678198255438e-06,
+      "loss": 0.6805,
+      "step": 125
+    },
+    {
+      "epoch": 1.2233009708737863,
+      "grad_norm": 0.8268979479586152,
+      "learning_rate": 4.918394974330801e-06,
+      "loss": 0.7583,
+      "step": 126
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 0.7645562011859167,
+      "learning_rate": 4.917101751100142e-06,
+      "loss": 0.7109,
+      "step": 127
+    },
+    {
+      "epoch": 1.2427184466019416,
+      "grad_norm": 0.8421630378850258,
+      "learning_rate": 4.915798533910534e-06,
+      "loss": 0.714,
+      "step": 128
+    },
+    {
+      "epoch": 1.2524271844660193,
+      "grad_norm": 0.7552330522382976,
+      "learning_rate": 4.9144853281503715e-06,
+      "loss": 0.5679,
+      "step": 129
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.852273134838057,
+      "learning_rate": 4.91316213924935e-06,
+      "loss": 0.7952,
+      "step": 130
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 0.8114602189230878,
+      "learning_rate": 4.911828972678441e-06,
+      "loss": 0.6176,
+      "step": 131
+    },
+    {
+      "epoch": 1.2815533980582523,
+      "grad_norm": 0.8164393262307038,
+      "learning_rate": 4.91048583394987e-06,
+      "loss": 0.6477,
+      "step": 132
+    },
+    {
+      "epoch": 1.29126213592233,
+      "grad_norm": 0.8581250231339308,
+      "learning_rate": 4.909132728617095e-06,
+      "loss": 0.9476,
+      "step": 133
+    },
+    {
+      "epoch": 1.3009708737864076,
+      "grad_norm": 0.766359434418812,
+      "learning_rate": 4.907769662274785e-06,
+      "loss": 0.6528,
+      "step": 134
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.6976692264719603,
+      "learning_rate": 4.90639664055879e-06,
+      "loss": 0.5637,
+      "step": 135
+    },
+    {
+      "epoch": 1.3203883495145632,
+      "grad_norm": 0.7314627468283406,
+      "learning_rate": 4.905013669146127e-06,
+      "loss": 0.6096,
+      "step": 136
+    },
+    {
+      "epoch": 1.3300970873786409,
+      "grad_norm": 0.7773436570942428,
+      "learning_rate": 4.903620753754949e-06,
+      "loss": 0.7461,
+      "step": 137
+    },
+    {
+      "epoch": 1.3398058252427185,
+      "grad_norm": 0.8921198866699622,
+      "learning_rate": 4.902217900144524e-06,
+      "loss": 0.9358,
+      "step": 138
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 0.7375485723614339,
+      "learning_rate": 4.900805114115214e-06,
+      "loss": 0.8942,
+      "step": 139
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.7942794591041569,
+      "learning_rate": 4.899382401508446e-06,
+      "loss": 0.5492,
+      "step": 140
+    },
+    {
+      "epoch": 1.3689320388349515,
+      "grad_norm": 0.8078627057423822,
+      "learning_rate": 4.8979497682066916e-06,
+      "loss": 0.7808,
+      "step": 141
+    },
+    {
+      "epoch": 1.3786407766990292,
+      "grad_norm": 0.6837805173092772,
+      "learning_rate": 4.89650722013344e-06,
+      "loss": 0.5391,
+      "step": 142
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 0.9832481462127313,
+      "learning_rate": 4.895054763253177e-06,
+      "loss": 0.7406,
+      "step": 143
+    },
+    {
+      "epoch": 1.3980582524271845,
+      "grad_norm": 0.769076719995811,
+      "learning_rate": 4.8935924035713564e-06,
+      "loss": 0.5929,
+      "step": 144
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.8430007967016218,
+      "learning_rate": 4.892120147134378e-06,
+      "loss": 0.699,
+      "step": 145
+    },
+    {
+      "epoch": 1.4174757281553398,
+      "grad_norm": 0.9122194953590461,
+      "learning_rate": 4.8906380000295615e-06,
+      "loss": 0.6895,
+      "step": 146
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 0.7898786206966036,
+      "learning_rate": 4.889145968385121e-06,
+      "loss": 0.6528,
+      "step": 147
+    },
+    {
+      "epoch": 1.4368932038834952,
+      "grad_norm": 0.7666616277046007,
+      "learning_rate": 4.887644058370139e-06,
+      "loss": 0.531,
+      "step": 148
+    },
+    {
+      "epoch": 1.4466019417475728,
+      "grad_norm": 0.7653718997690864,
+      "learning_rate": 4.886132276194544e-06,
+      "loss": 0.5768,
+      "step": 149
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.90831243558374,
+      "learning_rate": 4.884610628109082e-06,
+      "loss": 0.5652,
+      "step": 150
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 0.7442744457963555,
+      "learning_rate": 4.883079120405292e-06,
+      "loss": 0.6688,
+      "step": 151
+    },
+    {
+      "epoch": 1.4757281553398058,
+      "grad_norm": 0.7344457110605118,
+      "learning_rate": 4.881537759415478e-06,
+      "loss": 0.5314,
+      "step": 152
+    },
+    {
+      "epoch": 1.4854368932038835,
+      "grad_norm": 1.1695813268075903,
+      "learning_rate": 4.879986551512684e-06,
+      "loss": 0.7273,
+      "step": 153
+    },
+    {
+      "epoch": 1.4951456310679612,
+      "grad_norm": 0.6738881315688398,
+      "learning_rate": 4.878425503110672e-06,
+      "loss": 0.5607,
+      "step": 154
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.8581543167197856,
+      "learning_rate": 4.876854620663887e-06,
+      "loss": 0.6943,
+      "step": 155
+    },
+    {
+      "epoch": 1.5145631067961165,
+      "grad_norm": 0.8060496260215426,
+      "learning_rate": 4.875273910667434e-06,
+      "loss": 0.6212,
+      "step": 156
+    },
+    {
+      "epoch": 1.5242718446601942,
+      "grad_norm": 0.8453811238639173,
+      "learning_rate": 4.873683379657057e-06,
+      "loss": 0.5456,
+      "step": 157
+    },
+    {
+      "epoch": 1.5339805825242718,
+      "grad_norm": 0.8844708129131296,
+      "learning_rate": 4.8720830342091015e-06,
+      "loss": 0.9448,
+      "step": 158
+    },
+    {
+      "epoch": 1.5436893203883495,
+      "grad_norm": 0.861907269409231,
+      "learning_rate": 4.870472880940496e-06,
+      "loss": 0.7925,
+      "step": 159
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 1.0829307275890252,
+      "learning_rate": 4.868852926508721e-06,
+      "loss": 0.8343,
+      "step": 160
+    },
+    {
+      "epoch": 1.5631067961165048,
+      "grad_norm": 0.875424595240487,
+      "learning_rate": 4.867223177611779e-06,
+      "loss": 0.7158,
+      "step": 161
+    },
+    {
+      "epoch": 1.5728155339805825,
+      "grad_norm": 0.6659133214470974,
+      "learning_rate": 4.865583640988173e-06,
+      "loss": 0.5684,
+      "step": 162
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 0.772588673794732,
+      "learning_rate": 4.863934323416871e-06,
+      "loss": 0.6486,
+      "step": 163
+    },
+    {
+      "epoch": 1.5922330097087378,
+      "grad_norm": 0.741499702229279,
+      "learning_rate": 4.862275231717288e-06,
+      "loss": 0.5758,
+      "step": 164
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.920336013887185,
+      "learning_rate": 4.860606372749247e-06,
+      "loss": 0.5468,
+      "step": 165
+    },
+    {
+      "epoch": 1.6116504854368932,
+      "grad_norm": 0.7093065195416415,
+      "learning_rate": 4.858927753412958e-06,
+      "loss": 0.6127,
+      "step": 166
+    },
+    {
+      "epoch": 1.6213592233009708,
+      "grad_norm": 0.8110174812695655,
+      "learning_rate": 4.857239380648985e-06,
+      "loss": 0.6017,
+      "step": 167
+    },
+    {
+      "epoch": 1.6310679611650487,
+      "grad_norm": 0.7375215564893128,
+      "learning_rate": 4.855541261438223e-06,
+      "loss": 0.6753,
+      "step": 168
+    },
+    {
+      "epoch": 1.6407766990291264,
+      "grad_norm": 0.7500973220432647,
+      "learning_rate": 4.8538334028018605e-06,
+      "loss": 0.8343,
+      "step": 169
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.6948713105727569,
+      "learning_rate": 4.8521158118013605e-06,
+      "loss": 0.5493,
+      "step": 170
+    },
+    {
+      "epoch": 1.6601941747572817,
+      "grad_norm": 0.773755912005905,
+      "learning_rate": 4.850388495538423e-06,
+      "loss": 0.6271,
+      "step": 171
+    },
+    {
+      "epoch": 1.6699029126213594,
+      "grad_norm": 0.8759327452228973,
+      "learning_rate": 4.84865146115496e-06,
+      "loss": 0.5641,
+      "step": 172
+    },
+    {
+      "epoch": 1.679611650485437,
+      "grad_norm": 0.8673872477886911,
+      "learning_rate": 4.846904715833066e-06,
+      "loss": 0.5295,
+      "step": 173
+    },
+    {
+      "epoch": 1.6893203883495147,
+      "grad_norm": 0.7435589108212891,
+      "learning_rate": 4.8451482667949836e-06,
+      "loss": 0.7122,
+      "step": 174
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.7510040594244859,
+      "learning_rate": 4.843382121303082e-06,
+      "loss": 0.6579,
+      "step": 175
+    },
+    {
+      "epoch": 1.70873786407767,
+      "grad_norm": 0.7254228155893363,
+      "learning_rate": 4.841606286659819e-06,
+      "loss": 0.5591,
+      "step": 176
+    },
+    {
+      "epoch": 1.7184466019417477,
+      "grad_norm": 0.7786787819553973,
+      "learning_rate": 4.839820770207714e-06,
+      "loss": 0.5417,
+      "step": 177
+    },
+    {
+      "epoch": 1.7281553398058254,
+      "grad_norm": 0.7424090664501856,
+      "learning_rate": 4.8380255793293195e-06,
+      "loss": 0.5679,
+      "step": 178
+    },
+    {
+      "epoch": 1.737864077669903,
+      "grad_norm": 0.880425961769788,
+      "learning_rate": 4.8362207214471864e-06,
+      "loss": 0.5504,
+      "step": 179
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.7692052821337514,
+      "learning_rate": 4.83440620402384e-06,
+      "loss": 0.6217,
+      "step": 180
+    },
+    {
+      "epoch": 1.7572815533980584,
+      "grad_norm": 0.7571562044181578,
+      "learning_rate": 4.832582034561738e-06,
+      "loss": 0.641,
+      "step": 181
+    },
+    {
+      "epoch": 1.766990291262136,
+      "grad_norm": 0.8286514373124609,
+      "learning_rate": 4.830748220603251e-06,
+      "loss": 0.7462,
+      "step": 182
+    },
+    {
+      "epoch": 1.7766990291262137,
+      "grad_norm": 0.8663863630421635,
+      "learning_rate": 4.828904769730628e-06,
+      "loss": 0.7598,
+      "step": 183
+    },
+    {
+      "epoch": 1.7864077669902914,
+      "grad_norm": 0.8392430546644409,
+      "learning_rate": 4.827051689565958e-06,
+      "loss": 0.6278,
+      "step": 184
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.7431256674232911,
+      "learning_rate": 4.825188987771149e-06,
+      "loss": 0.7471,
+      "step": 185
+    },
+    {
+      "epoch": 1.8058252427184467,
+      "grad_norm": 0.775602420628486,
+      "learning_rate": 4.82331667204789e-06,
+      "loss": 0.4658,
+      "step": 186
+    },
+    {
+      "epoch": 1.8155339805825244,
+      "grad_norm": 0.8220707289832607,
+      "learning_rate": 4.821434750137619e-06,
+      "loss": 0.6212,
+      "step": 187
+    },
+    {
+      "epoch": 1.825242718446602,
+      "grad_norm": 0.8721039703571108,
+      "learning_rate": 4.819543229821494e-06,
+      "loss": 0.7135,
+      "step": 188
+    },
+    {
+      "epoch": 1.8349514563106797,
+      "grad_norm": 0.7196046947557964,
+      "learning_rate": 4.8176421189203605e-06,
+      "loss": 0.5464,
+      "step": 189
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.7844213665577071,
+      "learning_rate": 4.815731425294716e-06,
+      "loss": 0.6536,
+      "step": 190
+    },
+    {
+      "epoch": 1.854368932038835,
+      "grad_norm": 0.7921297235454372,
+      "learning_rate": 4.813811156844681e-06,
+      "loss": 0.7183,
+      "step": 191
+    },
+    {
+      "epoch": 1.8640776699029127,
+      "grad_norm": 0.863196612505686,
+      "learning_rate": 4.811881321509964e-06,
+      "loss": 0.6976,
+      "step": 192
+    },
+    {
+      "epoch": 1.8737864077669903,
+      "grad_norm": 0.7755881818234687,
+      "learning_rate": 4.809941927269829e-06,
+      "loss": 0.8491,
+      "step": 193
+    },
+    {
+      "epoch": 1.883495145631068,
+      "grad_norm": 0.8186186296199729,
+      "learning_rate": 4.807992982143064e-06,
+      "loss": 0.6343,
+      "step": 194
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.913496737962807,
+      "learning_rate": 4.806034494187949e-06,
+      "loss": 0.8436,
+      "step": 195
+    },
+    {
+      "epoch": 1.9029126213592233,
+      "grad_norm": 0.7590143443155779,
+      "learning_rate": 4.804066471502216e-06,
+      "loss": 0.603,
+      "step": 196
+    },
+    {
+      "epoch": 1.912621359223301,
+      "grad_norm": 0.766119646590294,
+      "learning_rate": 4.802088922223024e-06,
+      "loss": 0.5167,
+      "step": 197
+    },
+    {
+      "epoch": 1.9223300970873787,
+      "grad_norm": 0.7274017571399753,
+      "learning_rate": 4.80010185452692e-06,
+      "loss": 0.6725,
+      "step": 198
+    },
+    {
+      "epoch": 1.9320388349514563,
+      "grad_norm": 0.73780689049117,
+      "learning_rate": 4.798105276629806e-06,
+      "loss": 0.521,
+      "step": 199
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.7791155049733209,
+      "learning_rate": 4.796099196786908e-06,
+      "loss": 0.583,
+      "step": 200
+    },
+    {
+      "epoch": 1.9514563106796117,
+      "grad_norm": 0.8004953963543238,
+      "learning_rate": 4.794083623292737e-06,
+      "loss": 0.7842,
+      "step": 201
+    },
+    {
+      "epoch": 1.9611650485436893,
+      "grad_norm": 0.7807030577892377,
+      "learning_rate": 4.792058564481058e-06,
+      "loss": 0.5534,
+      "step": 202
+    },
+    {
+      "epoch": 1.970873786407767,
+      "grad_norm": 0.7092633632036252,
+      "learning_rate": 4.7900240287248554e-06,
+      "loss": 0.7083,
+      "step": 203
+    },
+    {
+      "epoch": 1.9805825242718447,
+      "grad_norm": 0.8875925228430702,
+      "learning_rate": 4.7879800244362975e-06,
+      "loss": 0.7782,
+      "step": 204
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.8909791320483021,
+      "learning_rate": 4.785926560066703e-06,
+      "loss": 0.7681,
+      "step": 205
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7275875515555469,
+      "learning_rate": 4.783863644106502e-06,
+      "loss": 0.649,
+      "step": 206
+    },
+    {
+      "epoch": 2.0097087378640777,
+      "grad_norm": 0.802116606621558,
+      "learning_rate": 4.781791285085209e-06,
+      "loss": 0.5568,
+      "step": 207
+    },
+    {
+      "epoch": 2.0194174757281553,
+      "grad_norm": 0.7406594271373391,
+      "learning_rate": 4.779709491571378e-06,
+      "loss": 0.5144,
+      "step": 208
+    },
+    {
+      "epoch": 2.029126213592233,
+      "grad_norm": 0.6857989185878482,
+      "learning_rate": 4.777618272172573e-06,
+      "loss": 0.6449,
+      "step": 209
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.734748210095666,
+      "learning_rate": 4.775517635535332e-06,
+      "loss": 0.7377,
+      "step": 210
+    },
+    {
+      "epoch": 2.0485436893203883,
+      "grad_norm": 0.7708218856857236,
+      "learning_rate": 4.77340759034513e-06,
+      "loss": 0.6855,
+      "step": 211
+    },
+    {
+      "epoch": 2.058252427184466,
+      "grad_norm": 0.7279897754942517,
+      "learning_rate": 4.771288145326343e-06,
+      "loss": 0.6684,
+      "step": 212
+    },
+    {
+      "epoch": 2.0679611650485437,
+      "grad_norm": 0.674474791686522,
+      "learning_rate": 4.769159309242213e-06,
+      "loss": 0.438,
+      "step": 213
+    },
+    {
+      "epoch": 2.0776699029126213,
+      "grad_norm": 0.792505397680811,
+      "learning_rate": 4.767021090894809e-06,
+      "loss": 0.6831,
+      "step": 214
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 0.744545194492276,
+      "learning_rate": 4.764873499124997e-06,
+      "loss": 0.6976,
+      "step": 215
+    },
+    {
+      "epoch": 2.0970873786407767,
+      "grad_norm": 0.7354909274219694,
+      "learning_rate": 4.762716542812395e-06,
+      "loss": 0.5495,
+      "step": 216
+    },
+    {
+      "epoch": 2.1067961165048543,
+      "grad_norm": 0.7133200950602842,
+      "learning_rate": 4.7605502308753415e-06,
+      "loss": 0.5687,
+      "step": 217
+    },
+    {
+      "epoch": 2.116504854368932,
+      "grad_norm": 0.7982098464573162,
+      "learning_rate": 4.758374572270859e-06,
+      "loss": 0.6886,
+      "step": 218
+    },
+    {
+      "epoch": 2.1262135922330097,
+      "grad_norm": 0.6913449166288441,
+      "learning_rate": 4.756189575994614e-06,
+      "loss": 0.3812,
+      "step": 219
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.7192629747712952,
+      "learning_rate": 4.753995251080884e-06,
+      "loss": 0.575,
+      "step": 220
+    },
+    {
+      "epoch": 2.145631067961165,
+      "grad_norm": 0.6248283732386594,
+      "learning_rate": 4.7517916066025126e-06,
+      "loss": 0.463,
+      "step": 221
+    },
+    {
+      "epoch": 2.1553398058252426,
+      "grad_norm": 0.8287445715621566,
+      "learning_rate": 4.7495786516708806e-06,
+      "loss": 0.7071,
+      "step": 222
+    },
+    {
+      "epoch": 2.1650485436893203,
+      "grad_norm": 0.8318311223138934,
+      "learning_rate": 4.747356395435865e-06,
+      "loss": 0.4322,
+      "step": 223
+    },
+    {
+      "epoch": 2.174757281553398,
+      "grad_norm": 0.8533840933951986,
+      "learning_rate": 4.745124847085799e-06,
+      "loss": 0.568,
+      "step": 224
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 0.7901361363585824,
+      "learning_rate": 4.742884015847436e-06,
+      "loss": 0.5393,
+      "step": 225
+    },
+    {
+      "epoch": 2.1941747572815533,
+      "grad_norm": 0.8111627610336898,
+      "learning_rate": 4.740633910985911e-06,
+      "loss": 0.5969,
+      "step": 226
+    },
+    {
+      "epoch": 2.203883495145631,
+      "grad_norm": 0.7671540334640228,
+      "learning_rate": 4.738374541804704e-06,
+      "loss": 0.4576,
+      "step": 227
+    },
+    {
+      "epoch": 2.2135922330097086,
+      "grad_norm": 0.7242562019069873,
+      "learning_rate": 4.7361059176456e-06,
+      "loss": 0.5186,
+      "step": 228
+    },
+    {
+      "epoch": 2.2233009708737863,
+      "grad_norm": 0.7763865691421442,
+      "learning_rate": 4.733828047888647e-06,
+      "loss": 0.6682,
+      "step": 229
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.6906133600693856,
+      "learning_rate": 4.731540941952126e-06,
+      "loss": 0.4051,
+      "step": 230
+    },
+    {
+      "epoch": 2.2427184466019416,
+      "grad_norm": 0.7481490824065317,
+      "learning_rate": 4.7292446092925016e-06,
+      "loss": 0.4902,
+      "step": 231
+    },
+    {
+      "epoch": 2.2524271844660193,
+      "grad_norm": 0.7044106920537051,
+      "learning_rate": 4.726939059404392e-06,
+      "loss": 0.4499,
+      "step": 232
+    },
+    {
+      "epoch": 2.262135922330097,
+      "grad_norm": 0.6853728707124397,
+      "learning_rate": 4.724624301820524e-06,
+      "loss": 0.4757,
+      "step": 233
+    },
+    {
+      "epoch": 2.2718446601941746,
+      "grad_norm": 0.9364489830239873,
+      "learning_rate": 4.722300346111695e-06,
+      "loss": 0.6159,
+      "step": 234
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 0.8138300584150231,
+      "learning_rate": 4.719967201886734e-06,
+      "loss": 0.5365,
+      "step": 235
+    },
+    {
+      "epoch": 2.29126213592233,
+      "grad_norm": 0.7475372553230331,
+      "learning_rate": 4.717624878792461e-06,
+      "loss": 0.7347,
+      "step": 236
+    },
+    {
+      "epoch": 2.3009708737864076,
+      "grad_norm": 0.9795714505586022,
+      "learning_rate": 4.715273386513651e-06,
+      "loss": 0.7175,
+      "step": 237
+    },
+    {
+      "epoch": 2.3106796116504853,
+      "grad_norm": 0.7332975150286841,
+      "learning_rate": 4.712912734772988e-06,
+      "loss": 0.5204,
+      "step": 238
+    },
+    {
+      "epoch": 2.320388349514563,
+      "grad_norm": 0.6488936041709285,
+      "learning_rate": 4.710542933331025e-06,
+      "loss": 0.5825,
+      "step": 239
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.7371958257036496,
+      "learning_rate": 4.708163991986152e-06,
+      "loss": 0.6679,
+      "step": 240
+    },
+    {
+      "epoch": 2.3398058252427183,
+      "grad_norm": 0.8214266450639852,
+      "learning_rate": 4.705775920574546e-06,
+      "loss": 0.6434,
+      "step": 241
+    },
+    {
+      "epoch": 2.349514563106796,
+      "grad_norm": 0.7702926046174587,
+      "learning_rate": 4.703378728970134e-06,
+      "loss": 0.4755,
+      "step": 242
+    },
+    {
+      "epoch": 2.3592233009708736,
+      "grad_norm": 0.6996698003201337,
+      "learning_rate": 4.700972427084551e-06,
+      "loss": 0.592,
+      "step": 243
+    },
+    {
+      "epoch": 2.3689320388349513,
+      "grad_norm": 0.7560377791738715,
+      "learning_rate": 4.698557024867105e-06,
+      "loss": 0.4565,
+      "step": 244
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 0.7633818347510686,
+      "learning_rate": 4.696132532304727e-06,
+      "loss": 0.6659,
+      "step": 245
+    },
+    {
+      "epoch": 2.3883495145631066,
+      "grad_norm": 0.7234976470337168,
+      "learning_rate": 4.693698959421935e-06,
+      "loss": 0.5678,
+      "step": 246
+    },
+    {
+      "epoch": 2.3980582524271843,
+      "grad_norm": 0.750666487544434,
+      "learning_rate": 4.691256316280789e-06,
+      "loss": 0.8332,
+      "step": 247
+    },
+    {
+      "epoch": 2.407766990291262,
+      "grad_norm": 0.7493376372555444,
+      "learning_rate": 4.688804612980855e-06,
+      "loss": 0.4837,
+      "step": 248
+    },
+    {
+      "epoch": 2.4174757281553396,
+      "grad_norm": 0.6751745421938132,
+      "learning_rate": 4.686343859659158e-06,
+      "loss": 0.6951,
+      "step": 249
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.734057362640619,
+      "learning_rate": 4.683874066490143e-06,
+      "loss": 0.5507,
+      "step": 250
+    },
+    {
+      "epoch": 2.436893203883495,
+      "grad_norm": 0.7227243519681933,
+      "learning_rate": 4.681395243685631e-06,
+      "loss": 0.56,
+      "step": 251
+    },
+    {
+      "epoch": 2.4466019417475726,
+      "grad_norm": 0.7751770457742055,
+      "learning_rate": 4.67890740149478e-06,
+      "loss": 0.518,
+      "step": 252
+    },
+    {
+      "epoch": 2.4563106796116507,
+      "grad_norm": 0.8010398356492873,
+      "learning_rate": 4.676410550204036e-06,
+      "loss": 0.6707,
+      "step": 253
+    },
+    {
+      "epoch": 2.466019417475728,
+      "grad_norm": 0.6322959904355947,
+      "learning_rate": 4.673904700137098e-06,
+      "loss": 0.3992,
+      "step": 254
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.6951997535390679,
+      "learning_rate": 4.671389861654873e-06,
+      "loss": 0.4745,
+      "step": 255
+    },
+    {
+      "epoch": 2.4854368932038833,
+      "grad_norm": 0.7106909584689424,
+      "learning_rate": 4.668866045155428e-06,
+      "loss": 0.596,
+      "step": 256
+    },
+    {
+      "epoch": 2.4951456310679614,
+      "grad_norm": 0.8847918134161639,
+      "learning_rate": 4.666333261073956e-06,
+      "loss": 0.6131,
+      "step": 257
+    },
+    {
+      "epoch": 2.5048543689320386,
+      "grad_norm": 0.7875255376725511,
+      "learning_rate": 4.6637915198827265e-06,
+      "loss": 0.6808,
+      "step": 258
+    },
+    {
+      "epoch": 2.5145631067961167,
+      "grad_norm": 0.74457845575171,
+      "learning_rate": 4.661240832091042e-06,
+      "loss": 0.534,
+      "step": 259
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.850799682933559,
+      "learning_rate": 4.658681208245198e-06,
+      "loss": 0.6302,
+      "step": 260
+    },
+    {
+      "epoch": 2.533980582524272,
+      "grad_norm": 0.7621063847057238,
+      "learning_rate": 4.65611265892844e-06,
+      "loss": 0.5924,
+      "step": 261
+    },
+    {
+      "epoch": 2.5436893203883493,
+      "grad_norm": 0.8162383044494845,
+      "learning_rate": 4.653535194760912e-06,
+      "loss": 0.5497,
+      "step": 262
+    },
+    {
+      "epoch": 2.5533980582524274,
+      "grad_norm": 0.7920894927810362,
+      "learning_rate": 4.650948826399624e-06,
+      "loss": 0.392,
+      "step": 263
+    },
+    {
+      "epoch": 2.5631067961165046,
+      "grad_norm": 0.7383821620603919,
+      "learning_rate": 4.648353564538397e-06,
+      "loss": 0.5049,
+      "step": 264
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 0.7650056257507512,
+      "learning_rate": 4.645749419907829e-06,
+      "loss": 0.4973,
+      "step": 265
+    },
+    {
+      "epoch": 2.58252427184466,
+      "grad_norm": 0.7120510594433922,
+      "learning_rate": 4.64313640327524e-06,
+      "loss": 0.406,
+      "step": 266
+    },
+    {
+      "epoch": 2.592233009708738,
+      "grad_norm": 0.7835448578972172,
+      "learning_rate": 4.640514525444637e-06,
+      "loss": 0.6122,
+      "step": 267
+    },
+    {
+      "epoch": 2.6019417475728153,
+      "grad_norm": 0.9134409217464837,
+      "learning_rate": 4.637883797256663e-06,
+      "loss": 0.845,
+      "step": 268
+    },
+    {
+      "epoch": 2.6116504854368934,
+      "grad_norm": 0.7387942613013898,
+      "learning_rate": 4.635244229588558e-06,
+      "loss": 0.5296,
+      "step": 269
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.6722442278200302,
+      "learning_rate": 4.632595833354105e-06,
+      "loss": 0.6111,
+      "step": 270
+    },
+    {
+      "epoch": 2.6310679611650487,
+      "grad_norm": 0.790100355706828,
+      "learning_rate": 4.629938619503593e-06,
+      "loss": 0.4931,
+      "step": 271
+    },
+    {
+      "epoch": 2.6407766990291264,
+      "grad_norm": 0.7457720631344418,
+      "learning_rate": 4.627272599023772e-06,
+      "loss": 0.6932,
+      "step": 272
+    },
+    {
+      "epoch": 2.650485436893204,
+      "grad_norm": 0.7742836603714502,
+      "learning_rate": 4.6245977829378e-06,
+      "loss": 0.8069,
+      "step": 273
+    },
+    {
+      "epoch": 2.6601941747572817,
+      "grad_norm": 0.7522977861605561,
+      "learning_rate": 4.6219141823052035e-06,
+      "loss": 0.4594,
+      "step": 274
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.7897785740037921,
+      "learning_rate": 4.619221808221833e-06,
+      "loss": 0.7502,
+      "step": 275
+    },
+    {
+      "epoch": 2.679611650485437,
+      "grad_norm": 0.6542147168655412,
+      "learning_rate": 4.616520671819812e-06,
+      "loss": 0.455,
+      "step": 276
+    },
+    {
+      "epoch": 2.6893203883495147,
+      "grad_norm": 1.3324292561191493,
+      "learning_rate": 4.613810784267492e-06,
+      "loss": 0.615,
+      "step": 277
+    },
+    {
+      "epoch": 2.6990291262135924,
+      "grad_norm": 0.6366824455259196,
+      "learning_rate": 4.61109215676941e-06,
+      "loss": 0.4398,
+      "step": 278
+    },
+    {
+      "epoch": 2.70873786407767,
+      "grad_norm": 0.7464337608282651,
+      "learning_rate": 4.608364800566241e-06,
+      "loss": 0.587,
+      "step": 279
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.6895928222340304,
+      "learning_rate": 4.605628726934747e-06,
+      "loss": 0.6703,
+      "step": 280
+    },
+    {
+      "epoch": 2.7281553398058254,
+      "grad_norm": 0.6878653541621508,
+      "learning_rate": 4.602883947187738e-06,
+      "loss": 0.5948,
+      "step": 281
+    },
+    {
+      "epoch": 2.737864077669903,
+      "grad_norm": 0.6926203162726917,
+      "learning_rate": 4.600130472674017e-06,
+      "loss": 0.4607,
+      "step": 282
+    },
+    {
+      "epoch": 2.7475728155339807,
+      "grad_norm": 0.7781936564011823,
+      "learning_rate": 4.5973683147783405e-06,
+      "loss": 0.7486,
+      "step": 283
+    },
+    {
+      "epoch": 2.7572815533980584,
+      "grad_norm": 0.7501081998810354,
+      "learning_rate": 4.594597484921365e-06,
+      "loss": 0.5782,
+      "step": 284
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 0.7647372977005877,
+      "learning_rate": 4.5918179945596055e-06,
+      "loss": 0.5159,
+      "step": 285
+    },
+    {
+      "epoch": 2.7766990291262137,
+      "grad_norm": 0.6977238852528742,
+      "learning_rate": 4.589029855185384e-06,
+      "loss": 0.5334,
+      "step": 286
+    },
+    {
+      "epoch": 2.7864077669902914,
+      "grad_norm": 0.9130101610006348,
+      "learning_rate": 4.586233078326785e-06,
+      "loss": 0.8354,
+      "step": 287
+    },
+    {
+      "epoch": 2.796116504854369,
+      "grad_norm": 0.8318292434477851,
+      "learning_rate": 4.583427675547602e-06,
+      "loss": 0.6258,
+      "step": 288
+    },
+    {
+      "epoch": 2.8058252427184467,
+      "grad_norm": 0.7324981921403528,
+      "learning_rate": 4.580613658447301e-06,
+      "loss": 0.709,
+      "step": 289
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.7523751933345019,
+      "learning_rate": 4.577791038660959e-06,
+      "loss": 0.4851,
+      "step": 290
+    },
+    {
+      "epoch": 2.825242718446602,
+      "grad_norm": 0.7542503570491638,
+      "learning_rate": 4.574959827859226e-06,
+      "loss": 0.4493,
+      "step": 291
+    },
+    {
+      "epoch": 2.8349514563106797,
+      "grad_norm": 0.741187265691656,
+      "learning_rate": 4.572120037748273e-06,
+      "loss": 0.4682,
+      "step": 292
+    },
+    {
+      "epoch": 2.8446601941747574,
+      "grad_norm": 0.7141919069039983,
+      "learning_rate": 4.5692716800697415e-06,
+      "loss": 0.6235,
+      "step": 293
+    },
+    {
+      "epoch": 2.854368932038835,
+      "grad_norm": 0.7414161387983556,
+      "learning_rate": 4.566414766600698e-06,
+      "loss": 0.5613,
+      "step": 294
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.700849509216115,
+      "learning_rate": 4.563549309153589e-06,
+      "loss": 0.443,
+      "step": 295
+    },
+    {
+      "epoch": 2.8737864077669903,
+      "grad_norm": 0.7854807058221909,
+      "learning_rate": 4.56067531957618e-06,
+      "loss": 0.4657,
+      "step": 296
+    },
+    {
+      "epoch": 2.883495145631068,
+      "grad_norm": 0.7744368603433763,
+      "learning_rate": 4.557792809751519e-06,
+      "loss": 0.6192,
+      "step": 297
+    },
+    {
+      "epoch": 2.8932038834951457,
+      "grad_norm": 0.811186979862698,
+      "learning_rate": 4.554901791597883e-06,
+      "loss": 0.5432,
+      "step": 298
+    },
+    {
+      "epoch": 2.9029126213592233,
+      "grad_norm": 0.7610230989372687,
+      "learning_rate": 4.552002277068725e-06,
+      "loss": 0.5689,
+      "step": 299
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.738268676455508,
+      "learning_rate": 4.549094278152631e-06,
+      "loss": 0.6102,
+      "step": 300
+    },
+    {
+      "epoch": 2.9223300970873787,
+      "grad_norm": 0.8265276954113333,
+      "learning_rate": 4.546177806873266e-06,
+      "loss": 0.4803,
+      "step": 301
+    },
+    {
+      "epoch": 2.9320388349514563,
+      "grad_norm": 0.7994330788168778,
+      "learning_rate": 4.543252875289326e-06,
+      "loss": 0.5232,
+      "step": 302
+    },
+    {
+      "epoch": 2.941747572815534,
+      "grad_norm": 0.7803228028067196,
+      "learning_rate": 4.540319495494486e-06,
+      "loss": 0.5785,
+      "step": 303
+    },
+    {
+      "epoch": 2.9514563106796117,
+      "grad_norm": 0.812661521852206,
+      "learning_rate": 4.537377679617353e-06,
+      "loss": 0.5857,
+      "step": 304
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 0.8017682940418599,
+      "learning_rate": 4.534427439821416e-06,
+      "loss": 0.4679,
+      "step": 305
+    },
+    {
+      "epoch": 2.970873786407767,
+      "grad_norm": 0.7557388320666087,
+      "learning_rate": 4.531468788304992e-06,
+      "loss": 0.4511,
+      "step": 306
+    },
+    {
+      "epoch": 2.9805825242718447,
+      "grad_norm": 0.7990082356636408,
+      "learning_rate": 4.5285017373011784e-06,
+      "loss": 0.4999,
+      "step": 307
+    },
+    {
+      "epoch": 2.9902912621359223,
+      "grad_norm": 0.8410650831835784,
+      "learning_rate": 4.5255262990778024e-06,
+      "loss": 0.4279,
+      "step": 308
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.67842206115732,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.4778,
+      "step": 309
+    },
+    {
+      "epoch": 3.0097087378640777,
+      "grad_norm": 0.7208972107613384,
+      "learning_rate": 4.519550310217013e-06,
+      "loss": 0.5166,
+      "step": 310
+    },
+    {
+      "epoch": 3.0194174757281553,
+      "grad_norm": 0.73349262960244,
+      "learning_rate": 4.516549784288442e-06,
+      "loss": 0.4455,
+      "step": 311
+    },
+    {
+      "epoch": 3.029126213592233,
+      "grad_norm": 0.6956083097152967,
+      "learning_rate": 4.513540920557892e-06,
+      "loss": 0.4371,
+      "step": 312
+    },
+    {
+      "epoch": 3.0388349514563107,
+      "grad_norm": 0.7915417702341847,
+      "learning_rate": 4.510523731466072e-06,
+      "loss": 0.6448,
+      "step": 313
+    },
+    {
+      "epoch": 3.0485436893203883,
+      "grad_norm": 0.5990172469379881,
+      "learning_rate": 4.507498229488116e-06,
+      "loss": 0.3384,
+      "step": 314
+    },
+    {
+      "epoch": 3.058252427184466,
+      "grad_norm": 0.8318855981134111,
+      "learning_rate": 4.504464427133527e-06,
+      "loss": 0.4508,
+      "step": 315
+    },
+    {
+      "epoch": 3.0679611650485437,
+      "grad_norm": 0.8758812451211815,
+      "learning_rate": 4.501422336946126e-06,
+      "loss": 0.4328,
+      "step": 316
+    },
+    {
+      "epoch": 3.0776699029126213,
+      "grad_norm": 0.7844299628970943,
+      "learning_rate": 4.498371971504005e-06,
+      "loss": 0.4222,
+      "step": 317
+    },
+    {
+      "epoch": 3.087378640776699,
+      "grad_norm": 0.7195150581975005,
+      "learning_rate": 4.49531334341947e-06,
+      "loss": 0.4417,
+      "step": 318
+    },
+    {
+      "epoch": 3.0970873786407767,
+      "grad_norm": 0.7671602770728359,
+      "learning_rate": 4.49224646533899e-06,
+      "loss": 0.42,
+      "step": 319
+    },
+    {
+      "epoch": 3.1067961165048543,
+      "grad_norm": 0.6929615407642132,
+      "learning_rate": 4.489171349943144e-06,
+      "loss": 0.4332,
+      "step": 320
+    },
+    {
+      "epoch": 3.116504854368932,
+      "grad_norm": 0.7086258123510658,
+      "learning_rate": 4.486088009946575e-06,
+      "loss": 0.3632,
+      "step": 321
+    },
+    {
+      "epoch": 3.1262135922330097,
+      "grad_norm": 0.7708151250786913,
+      "learning_rate": 4.482996458097926e-06,
+      "loss": 0.4975,
+      "step": 322
+    },
+    {
+      "epoch": 3.1359223300970873,
+      "grad_norm": 0.846155524161367,
+      "learning_rate": 4.479896707179796e-06,
+      "loss": 0.6871,
+      "step": 323
+    },
+    {
+      "epoch": 3.145631067961165,
+      "grad_norm": 0.8231242232551517,
+      "learning_rate": 4.476788770008685e-06,
+      "loss": 0.6574,
+      "step": 324
+    },
+    {
+      "epoch": 3.1553398058252426,
+      "grad_norm": 0.8549632502448101,
+      "learning_rate": 4.473672659434941e-06,
+      "loss": 0.5856,
+      "step": 325
+    },
+    {
+      "epoch": 3.1650485436893203,
+      "grad_norm": 0.7228128426745009,
+      "learning_rate": 4.470548388342704e-06,
+      "loss": 0.4776,
+      "step": 326
+    },
+    {
+      "epoch": 3.174757281553398,
+      "grad_norm": 0.8108949680823984,
+      "learning_rate": 4.467415969649858e-06,
+      "loss": 0.4874,
+      "step": 327
+    },
+    {
+      "epoch": 3.1844660194174756,
+      "grad_norm": 0.6851857911673899,
+      "learning_rate": 4.464275416307973e-06,
+      "loss": 0.4994,
+      "step": 328
+    },
+    {
+      "epoch": 3.1941747572815533,
+      "grad_norm": 0.761950009953729,
+      "learning_rate": 4.461126741302253e-06,
+      "loss": 0.4929,
+      "step": 329
+    },
+    {
+      "epoch": 3.203883495145631,
+      "grad_norm": 0.8432338612654883,
+      "learning_rate": 4.457969957651485e-06,
+      "loss": 0.4137,
+      "step": 330
+    },
+    {
+      "epoch": 3.2135922330097086,
+      "grad_norm": 0.7741138798177974,
+      "learning_rate": 4.454805078407979e-06,
+      "loss": 0.6696,
+      "step": 331
+    },
+    {
+      "epoch": 3.2233009708737863,
+      "grad_norm": 0.71765697779996,
+      "learning_rate": 4.451632116657521e-06,
+      "loss": 0.4506,
+      "step": 332
+    },
+    {
+      "epoch": 3.233009708737864,
+      "grad_norm": 0.6253829454649015,
+      "learning_rate": 4.448451085519314e-06,
+      "loss": 0.3586,
+      "step": 333
+    },
+    {
+      "epoch": 3.2427184466019416,
+      "grad_norm": 0.9686569714580476,
+      "learning_rate": 4.445261998145927e-06,
+      "loss": 0.4832,
+      "step": 334
+    },
+    {
+      "epoch": 3.2524271844660193,
+      "grad_norm": 0.682456331535341,
+      "learning_rate": 4.442064867723236e-06,
+      "loss": 0.4737,
+      "step": 335
+    },
+    {
+      "epoch": 3.262135922330097,
+      "grad_norm": 0.801934212380896,
+      "learning_rate": 4.438859707470376e-06,
+      "loss": 0.4988,
+      "step": 336
+    },
+    {
+      "epoch": 3.2718446601941746,
+      "grad_norm": 0.6210847995583545,
+      "learning_rate": 4.435646530639679e-06,
+      "loss": 0.3549,
+      "step": 337
+    },
+    {
+      "epoch": 3.2815533980582523,
+      "grad_norm": 0.7693142736109597,
+      "learning_rate": 4.432425350516627e-06,
+      "loss": 0.4612,
+      "step": 338
+    },
+    {
+      "epoch": 3.29126213592233,
+      "grad_norm": 0.7469011823723339,
+      "learning_rate": 4.42919618041979e-06,
+      "loss": 0.3343,
+      "step": 339
+    },
+    {
+      "epoch": 3.3009708737864076,
+      "grad_norm": 0.6080757362122666,
+      "learning_rate": 4.425959033700776e-06,
+      "loss": 0.2422,
+      "step": 340
+    },
+    {
+      "epoch": 3.3106796116504853,
+      "grad_norm": 0.7696858723921102,
+      "learning_rate": 4.422713923744174e-06,
+      "loss": 0.5103,
+      "step": 341
+    },
+    {
+      "epoch": 3.320388349514563,
+      "grad_norm": 0.8778656215024975,
+      "learning_rate": 4.419460863967496e-06,
+      "loss": 0.5335,
+      "step": 342
+    },
+    {
+      "epoch": 3.3300970873786406,
+      "grad_norm": 0.6994532036876416,
+      "learning_rate": 4.416199867821126e-06,
+      "loss": 0.5022,
+      "step": 343
+    },
+    {
+      "epoch": 3.3398058252427183,
+      "grad_norm": 0.7969656894134531,
+      "learning_rate": 4.412930948788263e-06,
+      "loss": 0.4246,
+      "step": 344
+    },
+    {
+      "epoch": 3.349514563106796,
+      "grad_norm": 0.8154555970057127,
+      "learning_rate": 4.409654120384863e-06,
+      "loss": 0.6664,
+      "step": 345
+    },
+    {
+      "epoch": 3.3592233009708736,
+      "grad_norm": 0.7113779335682774,
+      "learning_rate": 4.406369396159585e-06,
+      "loss": 0.6024,
+      "step": 346
+    },
+    {
+      "epoch": 3.3689320388349513,
+      "grad_norm": 0.7946176349594999,
+      "learning_rate": 4.403076789693735e-06,
+      "loss": 0.6273,
+      "step": 347
+    },
+    {
+      "epoch": 3.378640776699029,
+      "grad_norm": 0.6725243345900541,
+      "learning_rate": 4.399776314601212e-06,
+      "loss": 0.422,
+      "step": 348
+    },
+    {
+      "epoch": 3.3883495145631066,
+      "grad_norm": 0.8474457763132652,
+      "learning_rate": 4.396467984528445e-06,
+      "loss": 0.5515,
+      "step": 349
+    },
+    {
+      "epoch": 3.3980582524271843,
+      "grad_norm": 0.7635758657443303,
+      "learning_rate": 4.393151813154345e-06,
+      "loss": 0.417,
+      "step": 350
+    },
+    {
+      "epoch": 3.407766990291262,
+      "grad_norm": 0.7301687492460268,
+      "learning_rate": 4.3898278141902396e-06,
+      "loss": 0.4335,
+      "step": 351
+    },
+    {
+      "epoch": 3.4174757281553396,
+      "grad_norm": 0.8217888155242994,
+      "learning_rate": 4.386496001379826e-06,
+      "loss": 0.5301,
+      "step": 352
+    },
+    {
+      "epoch": 3.4271844660194173,
+      "grad_norm": 0.700440895441813,
+      "learning_rate": 4.383156388499106e-06,
+      "loss": 0.5289,
+      "step": 353
+    },
+    {
+      "epoch": 3.436893203883495,
+      "grad_norm": 0.676777993038823,
+      "learning_rate": 4.3798089893563335e-06,
+      "loss": 0.4079,
+      "step": 354
+    },
+    {
+      "epoch": 3.4466019417475726,
+      "grad_norm": 0.7359945465771945,
+      "learning_rate": 4.3764538177919555e-06,
+      "loss": 0.3024,
+      "step": 355
+    },
+    {
+      "epoch": 3.4563106796116507,
+      "grad_norm": 0.7431315486816701,
+      "learning_rate": 4.3730908876785574e-06,
+      "loss": 0.4715,
+      "step": 356
+    },
+    {
+      "epoch": 3.466019417475728,
+      "grad_norm": 0.858728130994808,
+      "learning_rate": 4.3697202129208e-06,
+      "loss": 0.5126,
+      "step": 357
+    },
+    {
+      "epoch": 3.475728155339806,
+      "grad_norm": 0.7006328621067651,
+      "learning_rate": 4.36634180745537e-06,
+      "loss": 0.3219,
+      "step": 358
+    },
+    {
+      "epoch": 3.4854368932038833,
+      "grad_norm": 0.8347983820993932,
+      "learning_rate": 4.3629556852509145e-06,
+      "loss": 0.7038,
+      "step": 359
+    },
+    {
+      "epoch": 3.4951456310679614,
+      "grad_norm": 0.6736438089942041,
+      "learning_rate": 4.35956186030799e-06,
+      "loss": 0.388,
+      "step": 360
+    },
+    {
+      "epoch": 3.5048543689320386,
+      "grad_norm": 0.7313432635941632,
+      "learning_rate": 4.356160346659001e-06,
+      "loss": 0.4803,
+      "step": 361
+    },
+    {
+      "epoch": 3.5145631067961167,
+      "grad_norm": 0.914378556446873,
+      "learning_rate": 4.3527511583681384e-06,
+      "loss": 0.9472,
+      "step": 362
+    },
+    {
+      "epoch": 3.524271844660194,
+      "grad_norm": 0.7928555788899265,
+      "learning_rate": 4.34933430953133e-06,
+      "loss": 0.5195,
+      "step": 363
+    },
+    {
+      "epoch": 3.533980582524272,
+      "grad_norm": 0.8552957566999786,
+      "learning_rate": 4.345909814276177e-06,
+      "loss": 0.5432,
+      "step": 364
+    },
+    {
+      "epoch": 3.5436893203883493,
+      "grad_norm": 0.6593786360114855,
+      "learning_rate": 4.3424776867618935e-06,
+      "loss": 0.5773,
+      "step": 365
+    },
+    {
+      "epoch": 3.5533980582524274,
+      "grad_norm": 0.7453073470450546,
+      "learning_rate": 4.339037941179253e-06,
+      "loss": 0.5414,
+      "step": 366
+    },
+    {
+      "epoch": 3.5631067961165046,
+      "grad_norm": 0.7285347424646024,
+      "learning_rate": 4.335590591750526e-06,
+      "loss": 0.4336,
+      "step": 367
+    },
+    {
+      "epoch": 3.5728155339805827,
+      "grad_norm": 0.7229430966497608,
+      "learning_rate": 4.332135652729423e-06,
+      "loss": 0.4226,
+      "step": 368
+    },
+    {
+      "epoch": 3.58252427184466,
+      "grad_norm": 0.7264295125746805,
+      "learning_rate": 4.328673138401036e-06,
+      "loss": 0.5502,
+      "step": 369
+    },
+    {
+      "epoch": 3.592233009708738,
+      "grad_norm": 0.7134753381323626,
+      "learning_rate": 4.325203063081776e-06,
+      "loss": 0.2876,
+      "step": 370
+    },
+    {
+      "epoch": 3.6019417475728153,
+      "grad_norm": 0.7636004423408037,
+      "learning_rate": 4.32172544111932e-06,
+      "loss": 0.3651,
+      "step": 371
+    },
+    {
+      "epoch": 3.6116504854368934,
+      "grad_norm": 0.731718208729643,
+      "learning_rate": 4.318240286892544e-06,
+      "loss": 0.4391,
+      "step": 372
+    },
+    {
+      "epoch": 3.6213592233009706,
+      "grad_norm": 0.6392786785401195,
+      "learning_rate": 4.314747614811471e-06,
+      "loss": 0.4575,
+      "step": 373
+    },
+    {
+      "epoch": 3.6310679611650487,
+      "grad_norm": 0.7082359191028809,
+      "learning_rate": 4.3112474393172055e-06,
+      "loss": 0.3473,
+      "step": 374
+    },
+    {
+      "epoch": 3.6407766990291264,
+      "grad_norm": 0.704495139070046,
+      "learning_rate": 4.307739774881878e-06,
+      "loss": 0.4346,
+      "step": 375
+    },
+    {
+      "epoch": 3.650485436893204,
+      "grad_norm": 0.7500124236230435,
+      "learning_rate": 4.304224636008582e-06,
+      "loss": 0.2937,
+      "step": 376
+    },
+    {
+      "epoch": 3.6601941747572817,
+      "grad_norm": 0.7473030765489458,
+      "learning_rate": 4.300702037231318e-06,
+      "loss": 0.5837,
+      "step": 377
+    },
+    {
+      "epoch": 3.6699029126213594,
+      "grad_norm": 0.6924606297437474,
+      "learning_rate": 4.297171993114927e-06,
+      "loss": 0.2863,
+      "step": 378
+    },
+    {
+      "epoch": 3.679611650485437,
+      "grad_norm": 0.7683874913737846,
+      "learning_rate": 4.2936345182550365e-06,
+      "loss": 0.5933,
+      "step": 379
+    },
+    {
+      "epoch": 3.6893203883495147,
+      "grad_norm": 0.738704129251105,
+      "learning_rate": 4.290089627277998e-06,
+      "loss": 0.3695,
+      "step": 380
+    },
+    {
+      "epoch": 3.6990291262135924,
+      "grad_norm": 0.730122979591364,
+      "learning_rate": 4.286537334840825e-06,
+      "loss": 0.5314,
+      "step": 381
+    },
+    {
+      "epoch": 3.70873786407767,
+      "grad_norm": 0.8561211215862192,
+      "learning_rate": 4.2829776556311355e-06,
+      "loss": 0.4077,
+      "step": 382
+    },
+    {
+      "epoch": 3.7184466019417477,
+      "grad_norm": 0.8670647072409673,
+      "learning_rate": 4.279410604367088e-06,
+      "loss": 0.6157,
+      "step": 383
+    },
+    {
+      "epoch": 3.7281553398058254,
+      "grad_norm": 0.8439911656034701,
+      "learning_rate": 4.275836195797323e-06,
+      "loss": 0.5611,
+      "step": 384
+    },
+    {
+      "epoch": 3.737864077669903,
+      "grad_norm": 0.8111642418779228,
+      "learning_rate": 4.2722544447008995e-06,
+      "loss": 0.6242,
+      "step": 385
+    },
+    {
+      "epoch": 3.7475728155339807,
+      "grad_norm": 0.8234444876560004,
+      "learning_rate": 4.268665365887238e-06,
+      "loss": 0.626,
+      "step": 386
+    },
+    {
+      "epoch": 3.7572815533980584,
+      "grad_norm": 0.7499486492298939,
+      "learning_rate": 4.265068974196056e-06,
+      "loss": 0.3372,
+      "step": 387
+    },
+    {
+      "epoch": 3.766990291262136,
+      "grad_norm": 0.8486583520088408,
+      "learning_rate": 4.261465284497307e-06,
+      "loss": 0.3682,
+      "step": 388
+    },
+    {
+      "epoch": 3.7766990291262137,
+      "grad_norm": 0.8004979648644168,
+      "learning_rate": 4.257854311691118e-06,
+      "loss": 0.5395,
+      "step": 389
+    },
+    {
+      "epoch": 3.7864077669902914,
+      "grad_norm": 0.7151044131463972,
+      "learning_rate": 4.254236070707734e-06,
+      "loss": 0.3921,
+      "step": 390
+    },
+    {
+      "epoch": 3.796116504854369,
+      "grad_norm": 0.6664683762506176,
+      "learning_rate": 4.250610576507445e-06,
+      "loss": 0.357,
+      "step": 391
+    },
+    {
+      "epoch": 3.8058252427184467,
+      "grad_norm": 0.8605859568498817,
+      "learning_rate": 4.246977844080537e-06,
+      "loss": 0.5304,
+      "step": 392
+    },
+    {
+      "epoch": 3.8155339805825244,
+      "grad_norm": 0.8012964647317763,
+      "learning_rate": 4.24333788844722e-06,
+      "loss": 0.4708,
+      "step": 393
+    },
+    {
+      "epoch": 3.825242718446602,
+      "grad_norm": 0.8135374110450408,
+      "learning_rate": 4.239690724657571e-06,
+      "loss": 0.4547,
+      "step": 394
+    },
+    {
+      "epoch": 3.8349514563106797,
+      "grad_norm": 1.1239630061318624,
+      "learning_rate": 4.236036367791471e-06,
+      "loss": 0.5222,
+      "step": 395
+    },
+    {
+      "epoch": 3.8446601941747574,
+      "grad_norm": 0.8761219541832255,
+      "learning_rate": 4.23237483295854e-06,
+      "loss": 0.584,
+      "step": 396
+    },
+    {
+      "epoch": 3.854368932038835,
+      "grad_norm": 0.7205334687878558,
+      "learning_rate": 4.228706135298081e-06,
+      "loss": 0.383,
+      "step": 397
+    },
+    {
+      "epoch": 3.8640776699029127,
+      "grad_norm": 0.778638883459709,
+      "learning_rate": 4.225030289979006e-06,
+      "loss": 0.3611,
+      "step": 398
+    },
+    {
+      "epoch": 3.8737864077669903,
+      "grad_norm": 0.775480657637448,
+      "learning_rate": 4.221347312199788e-06,
+      "loss": 0.3474,
+      "step": 399
+    },
+    {
+      "epoch": 3.883495145631068,
+      "grad_norm": 0.9124181162931588,
+      "learning_rate": 4.2176572171883865e-06,
+      "loss": 0.5554,
+      "step": 400
+    },
+    {
+      "epoch": 3.8932038834951457,
+      "grad_norm": 0.7674226802327085,
+      "learning_rate": 4.213960020202187e-06,
+      "loss": 0.5775,
+      "step": 401
+    },
+    {
+      "epoch": 3.9029126213592233,
+      "grad_norm": 0.8374874478064829,
+      "learning_rate": 4.2102557365279435e-06,
+      "loss": 0.6738,
+      "step": 402
+    },
+    {
+      "epoch": 3.912621359223301,
+      "grad_norm": 0.789430401614793,
+      "learning_rate": 4.206544381481708e-06,
+      "loss": 0.5645,
+      "step": 403
+    },
+    {
+      "epoch": 3.9223300970873787,
+      "grad_norm": 0.8301680910378548,
+      "learning_rate": 4.202825970408772e-06,
+      "loss": 0.3362,
+      "step": 404
+    },
+    {
+      "epoch": 3.9320388349514563,
+      "grad_norm": 0.7434161207431128,
+      "learning_rate": 4.199100518683601e-06,
+      "loss": 0.3865,
+      "step": 405
+    },
+    {
+      "epoch": 3.941747572815534,
+      "grad_norm": 0.8023276174001709,
+      "learning_rate": 4.195368041709772e-06,
+      "loss": 0.557,
+      "step": 406
+    },
+    {
+      "epoch": 3.9514563106796117,
+      "grad_norm": 0.6897944866344006,
+      "learning_rate": 4.191628554919907e-06,
+      "loss": 0.4306,
+      "step": 407
+    },
+    {
+      "epoch": 3.9611650485436893,
+      "grad_norm": 0.7587292046885276,
+      "learning_rate": 4.187882073775615e-06,
+      "loss": 0.4791,
+      "step": 408
+    },
+    {
+      "epoch": 3.970873786407767,
+      "grad_norm": 0.7402776540544128,
+      "learning_rate": 4.184128613767422e-06,
+      "loss": 0.4587,
+      "step": 409
+    },
+    {
+      "epoch": 3.9805825242718447,
+      "grad_norm": 0.918829002078532,
+      "learning_rate": 4.18036819041471e-06,
+      "loss": 0.4834,
+      "step": 410
+    },
+    {
+      "epoch": 3.9902912621359223,
+      "grad_norm": 0.6951873113715458,
+      "learning_rate": 4.17660081926565e-06,
+      "loss": 0.3421,
+      "step": 411
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.6863501093124053,
+      "learning_rate": 4.172826515897146e-06,
+      "loss": 0.3388,
+      "step": 412
+    },
+    {
+      "epoch": 4.009708737864078,
+      "grad_norm": 0.7114971026737045,
+      "learning_rate": 4.169045295914757e-06,
+      "loss": 0.5117,
+      "step": 413
+    },
+    {
+      "epoch": 4.019417475728155,
+      "grad_norm": 0.669639982894413,
+      "learning_rate": 4.165257174952647e-06,
+      "loss": 0.2925,
+      "step": 414
+    },
+    {
+      "epoch": 4.029126213592233,
+      "grad_norm": 0.7192571477195556,
+      "learning_rate": 4.161462168673508e-06,
+      "loss": 0.4189,
+      "step": 415
+    },
+    {
+      "epoch": 4.038834951456311,
+      "grad_norm": 0.7186614597357849,
+      "learning_rate": 4.157660292768502e-06,
+      "loss": 0.5147,
+      "step": 416
+    },
+    {
+      "epoch": 4.048543689320389,
+      "grad_norm": 0.7615450499237213,
+      "learning_rate": 4.1538515629571985e-06,
+      "loss": 0.3839,
+      "step": 417
+    },
+    {
+      "epoch": 4.058252427184466,
+      "grad_norm": 0.7825364795958354,
+      "learning_rate": 4.1500359949875e-06,
+      "loss": 0.5142,
+      "step": 418
+    },
+    {
+      "epoch": 4.067961165048544,
+      "grad_norm": 0.7567301120413946,
+      "learning_rate": 4.1462136046355864e-06,
+      "loss": 0.3761,
+      "step": 419
+    },
+    {
+      "epoch": 4.077669902912621,
+      "grad_norm": 0.7496382866591231,
+      "learning_rate": 4.142384407705846e-06,
+      "loss": 0.6578,
+      "step": 420
+    },
+    {
+      "epoch": 4.087378640776699,
+      "grad_norm": 0.6832493076636806,
+      "learning_rate": 4.138548420030808e-06,
+      "loss": 0.2888,
+      "step": 421
+    },
+    {
+      "epoch": 4.097087378640777,
+      "grad_norm": 0.8082055763247596,
+      "learning_rate": 4.13470565747108e-06,
+      "loss": 0.4464,
+      "step": 422
+    },
+    {
+      "epoch": 4.106796116504855,
+      "grad_norm": 0.6922107618500224,
+      "learning_rate": 4.130856135915282e-06,
+      "loss": 0.414,
+      "step": 423
+    },
+    {
+      "epoch": 4.116504854368932,
+      "grad_norm": 0.8568390451993608,
+      "learning_rate": 4.126999871279982e-06,
+      "loss": 0.3524,
+      "step": 424
+    },
+    {
+      "epoch": 4.12621359223301,
+      "grad_norm": 0.7555701643934601,
+      "learning_rate": 4.123136879509626e-06,
+      "loss": 0.4294,
+      "step": 425
+    },
+    {
+      "epoch": 4.135922330097087,
+      "grad_norm": 0.7172542927002353,
+      "learning_rate": 4.119267176576475e-06,
+      "loss": 0.464,
+      "step": 426
+    },
+    {
+      "epoch": 4.145631067961165,
+      "grad_norm": 0.6682558575648311,
+      "learning_rate": 4.11539077848054e-06,
+      "loss": 0.4873,
+      "step": 427
+    },
+    {
+      "epoch": 4.155339805825243,
+      "grad_norm": 0.8598473815632454,
+      "learning_rate": 4.111507701249513e-06,
+      "loss": 0.4819,
+      "step": 428
+    },
+    {
+      "epoch": 4.165048543689321,
+      "grad_norm": 0.9167731886550053,
+      "learning_rate": 4.107617960938702e-06,
+      "loss": 0.1954,
+      "step": 429
+    },
+    {
+      "epoch": 4.174757281553398,
+      "grad_norm": 0.8392679998978078,
+      "learning_rate": 4.103721573630965e-06,
+      "loss": 0.3716,
+      "step": 430
+    },
+    {
+      "epoch": 4.184466019417476,
+      "grad_norm": 0.7039689710893157,
+      "learning_rate": 4.099818555436645e-06,
+      "loss": 0.399,
+      "step": 431
+    },
+    {
+      "epoch": 4.194174757281553,
+      "grad_norm": 0.8962803885867973,
+      "learning_rate": 4.095908922493499e-06,
+      "loss": 0.6665,
+      "step": 432
+    },
+    {
+      "epoch": 4.203883495145631,
+      "grad_norm": 1.919580266091257,
+      "learning_rate": 4.091992690966636e-06,
+      "loss": 0.2248,
+      "step": 433
+    },
+    {
+      "epoch": 4.213592233009709,
+      "grad_norm": 0.7602103224056724,
+      "learning_rate": 4.088069877048447e-06,
+      "loss": 0.3179,
+      "step": 434
+    },
+    {
+      "epoch": 4.223300970873787,
+      "grad_norm": 0.6467738319616958,
+      "learning_rate": 4.084140496958539e-06,
+      "loss": 0.2963,
+      "step": 435
+    },
+    {
+      "epoch": 4.233009708737864,
+      "grad_norm": 0.7220900194096829,
+      "learning_rate": 4.080204566943668e-06,
+      "loss": 0.2475,
+      "step": 436
+    },
+    {
+      "epoch": 4.242718446601942,
+      "grad_norm": 1.189054603182352,
+      "learning_rate": 4.076262103277673e-06,
+      "loss": 0.2943,
+      "step": 437
+    },
+    {
+      "epoch": 4.252427184466019,
+      "grad_norm": 0.7169713980642173,
+      "learning_rate": 4.072313122261406e-06,
+      "loss": 0.6236,
+      "step": 438
+    },
+    {
+      "epoch": 4.262135922330097,
+      "grad_norm": 0.631689118975417,
+      "learning_rate": 4.068357640222668e-06,
+      "loss": 0.2852,
+      "step": 439
+    },
+    {
+      "epoch": 4.271844660194175,
+      "grad_norm": 0.756930350000182,
+      "learning_rate": 4.06439567351614e-06,
+      "loss": 0.3246,
+      "step": 440
+    },
+    {
+      "epoch": 4.281553398058253,
+      "grad_norm": 0.6706899017753607,
+      "learning_rate": 4.0604272385233105e-06,
+      "loss": 0.3887,
+      "step": 441
+    },
+    {
+      "epoch": 4.29126213592233,
+      "grad_norm": 0.6775487389857546,
+      "learning_rate": 4.056452351652418e-06,
+      "loss": 0.2223,
+      "step": 442
+    },
+    {
+      "epoch": 4.300970873786408,
+      "grad_norm": 0.8187523228643651,
+      "learning_rate": 4.052471029338375e-06,
+      "loss": 0.4288,
+      "step": 443
+    },
+    {
+      "epoch": 4.310679611650485,
+      "grad_norm": 0.6200886929947741,
+      "learning_rate": 4.048483288042703e-06,
+      "loss": 0.24,
+      "step": 444
+    },
+    {
+      "epoch": 4.320388349514563,
+      "grad_norm": 0.7372457408545906,
+      "learning_rate": 4.0444891442534615e-06,
+      "loss": 0.3116,
+      "step": 445
+    },
+    {
+      "epoch": 4.330097087378641,
+      "grad_norm": 0.7260528356835414,
+      "learning_rate": 4.040488614485187e-06,
+      "loss": 0.4287,
+      "step": 446
+    },
+    {
+      "epoch": 4.339805825242719,
+      "grad_norm": 0.8887283733391886,
+      "learning_rate": 4.036481715278818e-06,
+      "loss": 0.4366,
+      "step": 447
+    },
+    {
+      "epoch": 4.349514563106796,
+      "grad_norm": 0.7417768044969674,
+      "learning_rate": 4.032468463201626e-06,
+      "loss": 0.4489,
+      "step": 448
+    },
+    {
+      "epoch": 4.359223300970874,
+      "grad_norm": 0.767167243913028,
+      "learning_rate": 4.028448874847152e-06,
+      "loss": 0.1889,
+      "step": 449
+    },
+    {
+      "epoch": 4.368932038834951,
+      "grad_norm": 0.7548738530081098,
+      "learning_rate": 4.024422966835137e-06,
+      "loss": 0.3836,
+      "step": 450
+    },
+    {
+      "epoch": 4.378640776699029,
+      "grad_norm": 0.7267980574762086,
+      "learning_rate": 4.0203907558114475e-06,
+      "loss": 0.3113,
+      "step": 451
+    },
+    {
+      "epoch": 4.388349514563107,
+      "grad_norm": 0.7384964389802123,
+      "learning_rate": 4.016352258448016e-06,
+      "loss": 0.2558,
+      "step": 452
+    },
+    {
+      "epoch": 4.398058252427185,
+      "grad_norm": 0.7354848571749719,
+      "learning_rate": 4.0123074914427635e-06,
+      "loss": 0.2631,
+      "step": 453
+    },
+    {
+      "epoch": 4.407766990291262,
+      "grad_norm": 0.6121502281910712,
+      "learning_rate": 4.008256471519536e-06,
+      "loss": 0.2496,
+      "step": 454
+    },
+    {
+      "epoch": 4.41747572815534,
+      "grad_norm": 0.8179123108880375,
+      "learning_rate": 4.004199215428032e-06,
+      "loss": 0.3859,
+      "step": 455
+    },
+    {
+      "epoch": 4.427184466019417,
+      "grad_norm": 0.9831209055765248,
+      "learning_rate": 4.000135739943735e-06,
+      "loss": 0.4455,
+      "step": 456
+    },
+    {
+      "epoch": 4.436893203883495,
+      "grad_norm": 0.8368630657095354,
+      "learning_rate": 3.996066061867844e-06,
+      "loss": 0.4183,
+      "step": 457
+    },
+    {
+      "epoch": 4.446601941747573,
+      "grad_norm": 0.7510366020026782,
+      "learning_rate": 3.991990198027203e-06,
+      "loss": 0.3243,
+      "step": 458
+    },
+    {
+      "epoch": 4.456310679611651,
+      "grad_norm": 0.7739088378814717,
+      "learning_rate": 3.987908165274233e-06,
+      "loss": 0.3937,
+      "step": 459
+    },
+    {
+      "epoch": 4.466019417475728,
+      "grad_norm": 0.7451064795923457,
+      "learning_rate": 3.9838199804868635e-06,
+      "loss": 0.4762,
+      "step": 460
+    },
+    {
+      "epoch": 4.475728155339806,
+      "grad_norm": 0.8005653610588523,
+      "learning_rate": 3.979725660568456e-06,
+      "loss": 0.4203,
+      "step": 461
+    },
+    {
+      "epoch": 4.485436893203883,
+      "grad_norm": 0.7571415862327683,
+      "learning_rate": 3.975625222447742e-06,
+      "loss": 0.2966,
+      "step": 462
+    },
+    {
+      "epoch": 4.495145631067961,
+      "grad_norm": 0.9407682824067278,
+      "learning_rate": 3.97151868307875e-06,
+      "loss": 0.4631,
+      "step": 463
+    },
+    {
+      "epoch": 4.504854368932039,
+      "grad_norm": 0.8280322987446443,
+      "learning_rate": 3.9674060594407345e-06,
+      "loss": 0.3439,
+      "step": 464
+    },
+    {
+      "epoch": 4.514563106796117,
+      "grad_norm": 0.7056881797171722,
+      "learning_rate": 3.963287368538105e-06,
+      "loss": 0.2049,
+      "step": 465
+    },
+    {
+      "epoch": 4.524271844660194,
+      "grad_norm": 0.9332990417689767,
+      "learning_rate": 3.959162627400361e-06,
+      "loss": 0.4755,
+      "step": 466
+    },
+    {
+      "epoch": 4.533980582524272,
+      "grad_norm": 1.1403012923588804,
+      "learning_rate": 3.9550318530820145e-06,
+      "loss": 0.4185,
+      "step": 467
+    },
+    {
+      "epoch": 4.543689320388349,
+      "grad_norm": 0.689075963184153,
+      "learning_rate": 3.9508950626625244e-06,
+      "loss": 0.2702,
+      "step": 468
+    },
+    {
+      "epoch": 4.553398058252427,
+      "grad_norm": 0.8020665017742961,
+      "learning_rate": 3.946752273246224e-06,
+      "loss": 0.4359,
+      "step": 469
+    },
+    {
+      "epoch": 4.563106796116505,
+      "grad_norm": 0.8212257599019372,
+      "learning_rate": 3.942603501962249e-06,
+      "loss": 0.352,
+      "step": 470
+    },
+    {
+      "epoch": 4.572815533980583,
+      "grad_norm": 0.7742300518258042,
+      "learning_rate": 3.9384487659644716e-06,
+      "loss": 0.2489,
+      "step": 471
+    },
+    {
+      "epoch": 4.58252427184466,
+      "grad_norm": 0.7529500735779854,
+      "learning_rate": 3.934288082431423e-06,
+      "loss": 0.4064,
+      "step": 472
+    },
+    {
+      "epoch": 4.592233009708738,
+      "grad_norm": 0.7285209333659659,
+      "learning_rate": 3.930121468566227e-06,
+      "loss": 0.4903,
+      "step": 473
+    },
+    {
+      "epoch": 4.601941747572815,
+      "grad_norm": 0.702088268251986,
+      "learning_rate": 3.925948941596528e-06,
+      "loss": 0.3524,
+      "step": 474
+    },
+    {
+      "epoch": 4.611650485436893,
+      "grad_norm": 0.6123252794922119,
+      "learning_rate": 3.92177051877442e-06,
+      "loss": 0.291,
+      "step": 475
+    },
+    {
+      "epoch": 4.621359223300971,
+      "grad_norm": 0.743156646498327,
+      "learning_rate": 3.917586217376369e-06,
+      "loss": 0.435,
+      "step": 476
+    },
+    {
+      "epoch": 4.631067961165049,
+      "grad_norm": 0.704106227231604,
+      "learning_rate": 3.913396054703155e-06,
+      "loss": 0.4408,
+      "step": 477
+    },
+    {
+      "epoch": 4.640776699029126,
+      "grad_norm": 0.7127662571747935,
+      "learning_rate": 3.909200048079786e-06,
+      "loss": 0.4177,
+      "step": 478
+    },
+    {
+      "epoch": 4.650485436893204,
+      "grad_norm": 0.748418825692663,
+      "learning_rate": 3.9049982148554384e-06,
+      "loss": 0.5271,
+      "step": 479
+    },
+    {
+      "epoch": 4.660194174757281,
+      "grad_norm": 1.0250685486452722,
+      "learning_rate": 3.900790572403376e-06,
+      "loss": 0.3461,
+      "step": 480
+    },
+    {
+      "epoch": 4.669902912621359,
+      "grad_norm": 0.7829872249544201,
+      "learning_rate": 3.896577138120881e-06,
+      "loss": 0.4164,
+      "step": 481
+    },
+    {
+      "epoch": 4.679611650485437,
+      "grad_norm": 0.9232004605724133,
+      "learning_rate": 3.892357929429187e-06,
+      "loss": 0.3579,
+      "step": 482
+    },
+    {
+      "epoch": 4.689320388349515,
+      "grad_norm": 0.784529648802751,
+      "learning_rate": 3.8881329637734e-06,
+      "loss": 0.2236,
+      "step": 483
+    },
+    {
+      "epoch": 4.699029126213592,
+      "grad_norm": 0.7262690017605888,
+      "learning_rate": 3.883902258622431e-06,
+      "loss": 0.3756,
+      "step": 484
+    },
+    {
+      "epoch": 4.70873786407767,
+      "grad_norm": 0.8102856355194789,
+      "learning_rate": 3.8796658314689205e-06,
+      "loss": 0.4178,
+      "step": 485
+    },
+    {
+      "epoch": 4.718446601941747,
+      "grad_norm": 0.7486005625149885,
+      "learning_rate": 3.875423699829168e-06,
+      "loss": 0.2835,
+      "step": 486
+    },
+    {
+      "epoch": 4.728155339805825,
+      "grad_norm": 0.7379335654656654,
+      "learning_rate": 3.871175881243061e-06,
+      "loss": 0.3489,
+      "step": 487
+    },
+    {
+      "epoch": 4.737864077669903,
+      "grad_norm": 0.8104226180405947,
+      "learning_rate": 3.866922393273999e-06,
+      "loss": 0.4974,
+      "step": 488
+    },
+    {
+      "epoch": 4.747572815533981,
+      "grad_norm": 0.7845427197820135,
+      "learning_rate": 3.862663253508822e-06,
+      "loss": 0.3453,
+      "step": 489
+    },
+    {
+      "epoch": 4.757281553398058,
+      "grad_norm": 0.7436234031119154,
+      "learning_rate": 3.858398479557739e-06,
+      "loss": 0.3273,
+      "step": 490
+    },
+    {
+      "epoch": 4.766990291262136,
+      "grad_norm": 2.217227869082823,
+      "learning_rate": 3.8541280890542565e-06,
+      "loss": 0.3229,
+      "step": 491
+    },
+    {
+      "epoch": 4.776699029126213,
+      "grad_norm": 1.0428680689060519,
+      "learning_rate": 3.849852099655102e-06,
+      "loss": 0.4476,
+      "step": 492
+    },
+    {
+      "epoch": 4.786407766990291,
+      "grad_norm": 0.7093952303779135,
+      "learning_rate": 3.845570529040151e-06,
+      "loss": 0.3531,
+      "step": 493
+    },
+    {
+      "epoch": 4.796116504854369,
+      "grad_norm": 0.7919634533614799,
+      "learning_rate": 3.841283394912361e-06,
+      "loss": 0.3435,
+      "step": 494
+    },
+    {
+      "epoch": 4.805825242718447,
+      "grad_norm": 0.7136080403274071,
+      "learning_rate": 3.836990714997686e-06,
+      "loss": 0.4444,
+      "step": 495
+    },
+    {
+      "epoch": 4.815533980582524,
+      "grad_norm": 0.8376006583714606,
+      "learning_rate": 3.832692507045015e-06,
+      "loss": 0.4478,
+      "step": 496
+    },
+    {
+      "epoch": 4.825242718446602,
+      "grad_norm": 0.6616123052279387,
+      "learning_rate": 3.828388788826091e-06,
+      "loss": 0.5166,
+      "step": 497
+    },
+    {
+      "epoch": 4.834951456310679,
+      "grad_norm": 0.7483368464194841,
+      "learning_rate": 3.824079578135442e-06,
+      "loss": 0.4151,
+      "step": 498
+    },
+    {
+      "epoch": 4.844660194174757,
+      "grad_norm": 0.646917604155803,
+      "learning_rate": 3.819764892790307e-06,
+      "loss": 0.4058,
+      "step": 499
+    },
+    {
+      "epoch": 4.854368932038835,
+      "grad_norm": 0.8267691480427513,
+      "learning_rate": 3.815444750630555e-06,
+      "loss": 0.3406,
+      "step": 500
+    },
+    {
+      "epoch": 4.864077669902913,
+      "grad_norm": 0.8928434754303194,
+      "learning_rate": 3.811119169518624e-06,
+      "loss": 0.541,
+      "step": 501
+    },
+    {
+      "epoch": 4.87378640776699,
+      "grad_norm": 0.9334290516881728,
+      "learning_rate": 3.8067881673394363e-06,
+      "loss": 0.6994,
+      "step": 502
+    },
+    {
+      "epoch": 4.883495145631068,
+      "grad_norm": 0.7272706791055442,
+      "learning_rate": 3.802451762000331e-06,
+      "loss": 0.3112,
+      "step": 503
+    },
+    {
+      "epoch": 4.893203883495145,
+      "grad_norm": 0.6753535996733946,
+      "learning_rate": 3.7981099714309856e-06,
+      "loss": 0.2749,
+      "step": 504
+    },
+    {
+      "epoch": 4.902912621359223,
+      "grad_norm": 0.8446207256750283,
+      "learning_rate": 3.7937628135833453e-06,
+      "loss": 0.354,
+      "step": 505
+    },
+    {
+      "epoch": 4.9126213592233015,
+      "grad_norm": 0.8427902134243664,
+      "learning_rate": 3.7894103064315463e-06,
+      "loss": 0.3765,
+      "step": 506
+    },
+    {
+      "epoch": 4.922330097087379,
+      "grad_norm": 0.7801637629685254,
+      "learning_rate": 3.7850524679718424e-06,
+      "loss": 0.3014,
+      "step": 507
+    },
+    {
+      "epoch": 4.932038834951456,
+      "grad_norm": 0.7448345484733,
+      "learning_rate": 3.7806893162225328e-06,
+      "loss": 0.2862,
+      "step": 508
+    },
+    {
+      "epoch": 4.941747572815534,
+      "grad_norm": 0.7506722271057215,
+      "learning_rate": 3.7763208692238818e-06,
+      "loss": 0.4005,
+      "step": 509
+    },
+    {
+      "epoch": 4.951456310679612,
+      "grad_norm": 0.7507341810173735,
+      "learning_rate": 3.7719471450380518e-06,
+      "loss": 0.3154,
+      "step": 510
+    },
+    {
+      "epoch": 4.961165048543689,
+      "grad_norm": 0.7509927539126146,
+      "learning_rate": 3.7675681617490212e-06,
+      "loss": 0.5283,
+      "step": 511
+    },
+    {
+      "epoch": 4.970873786407767,
+      "grad_norm": 0.7780851180094496,
+      "learning_rate": 3.7631839374625167e-06,
+      "loss": 0.5371,
+      "step": 512
+    },
+    {
+      "epoch": 4.980582524271845,
+      "grad_norm": 0.7269269443801724,
+      "learning_rate": 3.758794490305932e-06,
+      "loss": 0.3471,
+      "step": 513
+    },
+    {
+      "epoch": 4.990291262135923,
+      "grad_norm": 0.7023346099879195,
+      "learning_rate": 3.7543998384282565e-06,
+      "loss": 0.3844,
+      "step": 514
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.6668852497704912,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.4468,
+      "step": 515
+    },
+    {
+      "epoch": 5.009708737864078,
+      "grad_norm": 0.7476449787043878,
+      "learning_rate": 3.745594993213118e-06,
+      "loss": 0.3995,
+      "step": 516
+    },
+    {
+      "epoch": 5.019417475728155,
+      "grad_norm": 0.6983005205382378,
+      "learning_rate": 3.7411848362809324e-06,
+      "loss": 0.5712,
+      "step": 517
+    },
+    {
+      "epoch": 5.029126213592233,
+      "grad_norm": 0.7248404467665612,
+      "learning_rate": 3.7367695474380623e-06,
+      "loss": 0.3509,
+      "step": 518
+    },
+    {
+      "epoch": 5.038834951456311,
+      "grad_norm": 0.8246975738832559,
+      "learning_rate": 3.7323491449403444e-06,
+      "loss": 0.344,
+      "step": 519
+    },
+    {
+      "epoch": 5.048543689320389,
+      "grad_norm": 0.6636977636917745,
+      "learning_rate": 3.7279236470647593e-06,
+      "loss": 0.1996,
+      "step": 520
+    },
+    {
+      "epoch": 5.058252427184466,
+      "grad_norm": 0.8867501754473929,
+      "learning_rate": 3.723493072109355e-06,
+      "loss": 0.5021,
+      "step": 521
+    },
+    {
+      "epoch": 5.067961165048544,
+      "grad_norm": 0.8626723097108108,
+      "learning_rate": 3.719057438393172e-06,
+      "loss": 0.2895,
+      "step": 522
+    },
+    {
+      "epoch": 5.077669902912621,
+      "grad_norm": 1.0193146429465378,
+      "learning_rate": 3.714616764256166e-06,
+      "loss": 0.4664,
+      "step": 523
+    },
+    {
+      "epoch": 5.087378640776699,
+      "grad_norm": 0.8843905485635875,
+      "learning_rate": 3.7101710680591353e-06,
+      "loss": 0.3435,
+      "step": 524
+    },
+    {
+      "epoch": 5.097087378640777,
+      "grad_norm": 0.9512264521315206,
+      "learning_rate": 3.7057203681836407e-06,
+      "loss": 0.3871,
+      "step": 525
+    },
+    {
+      "epoch": 5.106796116504855,
+      "grad_norm": 0.8490880861462835,
+      "learning_rate": 3.701264683031934e-06,
+      "loss": 0.4215,
+      "step": 526
+    },
+    {
+      "epoch": 5.116504854368932,
+      "grad_norm": 0.7923161136499023,
+      "learning_rate": 3.6968040310268766e-06,
+      "loss": 0.3106,
+      "step": 527
+    },
+    {
+      "epoch": 5.12621359223301,
+      "grad_norm": 0.7158609870667556,
+      "learning_rate": 3.692338430611869e-06,
+      "loss": 0.3569,
+      "step": 528
+    },
+    {
+      "epoch": 5.135922330097087,
+      "grad_norm": 0.793510070177626,
+      "learning_rate": 3.687867900250771e-06,
+      "loss": 0.2498,
+      "step": 529
+    },
+    {
+      "epoch": 5.145631067961165,
+      "grad_norm": 0.789606227818232,
+      "learning_rate": 3.683392458427825e-06,
+      "loss": 0.2286,
+      "step": 530
+    },
+    {
+      "epoch": 5.155339805825243,
+      "grad_norm": 0.8813406233305229,
+      "learning_rate": 3.6789121236475818e-06,
+      "loss": 0.3946,
+      "step": 531
+    },
+    {
+      "epoch": 5.165048543689321,
+      "grad_norm": 0.7309365109221931,
+      "learning_rate": 3.674426914434824e-06,
+      "loss": 0.36,
+      "step": 532
+    },
+    {
+      "epoch": 5.174757281553398,
+      "grad_norm": 0.7734597901253387,
+      "learning_rate": 3.6699368493344856e-06,
+      "loss": 0.3541,
+      "step": 533
+    },
+    {
+      "epoch": 5.184466019417476,
+      "grad_norm": 0.684738131548144,
+      "learning_rate": 3.665441946911582e-06,
+      "loss": 0.2127,
+      "step": 534
+    },
+    {
+      "epoch": 5.194174757281553,
+      "grad_norm": 0.7131381186080148,
+      "learning_rate": 3.660942225751126e-06,
+      "loss": 0.2997,
+      "step": 535
+    },
+    {
+      "epoch": 5.203883495145631,
+      "grad_norm": 0.8304343144105206,
+      "learning_rate": 3.6564377044580558e-06,
+      "loss": 0.2777,
+      "step": 536
+    },
+    {
+      "epoch": 5.213592233009709,
+      "grad_norm": 0.7512779127156971,
+      "learning_rate": 3.6519284016571567e-06,
+      "loss": 0.2193,
+      "step": 537
+    },
+    {
+      "epoch": 5.223300970873787,
+      "grad_norm": 0.8246283116758232,
+      "learning_rate": 3.647414335992985e-06,
+      "loss": 0.3837,
+      "step": 538
+    },
+    {
+      "epoch": 5.233009708737864,
+      "grad_norm": 0.8396621277026458,
+      "learning_rate": 3.642895526129787e-06,
+      "loss": 0.3746,
+      "step": 539
+    },
+    {
+      "epoch": 5.242718446601942,
+      "grad_norm": 0.7935778168524156,
+      "learning_rate": 3.638371990751428e-06,
+      "loss": 0.3837,
+      "step": 540
+    },
+    {
+      "epoch": 5.252427184466019,
+      "grad_norm": 0.8991486533621065,
+      "learning_rate": 3.63384374856131e-06,
+      "loss": 0.3581,
+      "step": 541
+    },
+    {
+      "epoch": 5.262135922330097,
+      "grad_norm": 0.9432724876796261,
+      "learning_rate": 3.629310818282297e-06,
+      "loss": 0.4725,
+      "step": 542
+    },
+    {
+      "epoch": 5.271844660194175,
+      "grad_norm": 0.7506118004771117,
+      "learning_rate": 3.6247732186566365e-06,
+      "loss": 0.3419,
+      "step": 543
+    },
+    {
+      "epoch": 5.281553398058253,
+      "grad_norm": 0.7598462241955665,
+      "learning_rate": 3.6202309684458813e-06,
+      "loss": 0.2133,
+      "step": 544
+    },
+    {
+      "epoch": 5.29126213592233,
+      "grad_norm": 0.8062442517121977,
+      "learning_rate": 3.615684086430815e-06,
+      "loss": 0.2294,
+      "step": 545
+    },
+    {
+      "epoch": 5.300970873786408,
+      "grad_norm": 0.9107115654245546,
+      "learning_rate": 3.61113259141137e-06,
+      "loss": 0.3629,
+      "step": 546
+    },
+    {
+      "epoch": 5.310679611650485,
+      "grad_norm": 0.6755535448056335,
+      "learning_rate": 3.606576502206554e-06,
+      "loss": 0.1603,
+      "step": 547
+    },
+    {
+      "epoch": 5.320388349514563,
+      "grad_norm": 0.7340467593460501,
+      "learning_rate": 3.602015837654369e-06,
+      "loss": 0.1743,
+      "step": 548
+    },
+    {
+      "epoch": 5.330097087378641,
+      "grad_norm": 0.7628965161854933,
+      "learning_rate": 3.5974506166117355e-06,
+      "loss": 0.1826,
+      "step": 549
+    },
+    {
+      "epoch": 5.339805825242719,
+      "grad_norm": 0.8396283760259973,
+      "learning_rate": 3.592880857954413e-06,
+      "loss": 0.4369,
+      "step": 550
+    },
+    {
+      "epoch": 5.349514563106796,
+      "grad_norm": 0.7344763407362891,
+      "learning_rate": 3.588306580576922e-06,
+      "loss": 0.1936,
+      "step": 551
+    },
+    {
+      "epoch": 5.359223300970874,
+      "grad_norm": 0.6947637843679245,
+      "learning_rate": 3.583727803392468e-06,
+      "loss": 0.1258,
+      "step": 552
+    },
+    {
+      "epoch": 5.368932038834951,
+      "grad_norm": 0.8331318012858376,
+      "learning_rate": 3.57914454533286e-06,
+      "loss": 0.3562,
+      "step": 553
+    },
+    {
+      "epoch": 5.378640776699029,
+      "grad_norm": 1.0821814432994845,
+      "learning_rate": 3.5745568253484363e-06,
+      "loss": 0.2078,
+      "step": 554
+    },
+    {
+      "epoch": 5.388349514563107,
+      "grad_norm": 0.7957795333648919,
+      "learning_rate": 3.5699646624079824e-06,
+      "loss": 0.2882,
+      "step": 555
+    },
+    {
+      "epoch": 5.398058252427185,
+      "grad_norm": 0.6525938882878057,
+      "learning_rate": 3.5653680754986543e-06,
+      "loss": 0.1807,
+      "step": 556
+    },
+    {
+      "epoch": 5.407766990291262,
+      "grad_norm": 0.76383658803048,
+      "learning_rate": 3.560767083625899e-06,
+      "loss": 0.4813,
+      "step": 557
+    },
+    {
+      "epoch": 5.41747572815534,
+      "grad_norm": 0.787353335592656,
+      "learning_rate": 3.556161705813378e-06,
+      "loss": 0.228,
+      "step": 558
+    },
+    {
+      "epoch": 5.427184466019417,
+      "grad_norm": 0.94797988039084,
+      "learning_rate": 3.5515519611028863e-06,
+      "loss": 0.3629,
+      "step": 559
+    },
+    {
+      "epoch": 5.436893203883495,
+      "grad_norm": 0.8381630759676166,
+      "learning_rate": 3.5469378685542742e-06,
+      "loss": 0.3928,
+      "step": 560
+    },
+    {
+      "epoch": 5.446601941747573,
+      "grad_norm": 0.8908509257555599,
+      "learning_rate": 3.542319447245372e-06,
+      "loss": 0.5126,
+      "step": 561
+    },
+    {
+      "epoch": 5.456310679611651,
+      "grad_norm": 0.9904764252647209,
+      "learning_rate": 3.537696716271904e-06,
+      "loss": 0.4614,
+      "step": 562
+    },
+    {
+      "epoch": 5.466019417475728,
+      "grad_norm": 0.8971864824848211,
+      "learning_rate": 3.533069694747415e-06,
+      "loss": 0.2269,
+      "step": 563
+    },
+    {
+      "epoch": 5.475728155339806,
+      "grad_norm": 0.7345775221122297,
+      "learning_rate": 3.528438401803192e-06,
+      "loss": 0.2796,
+      "step": 564
+    },
+    {
+      "epoch": 5.485436893203883,
+      "grad_norm": 0.7565695930856697,
+      "learning_rate": 3.52380285658818e-06,
+      "loss": 0.2396,
+      "step": 565
+    },
+    {
+      "epoch": 5.495145631067961,
+      "grad_norm": 0.9700827286254838,
+      "learning_rate": 3.5191630782689074e-06,
+      "loss": 0.272,
+      "step": 566
+    },
+    {
+      "epoch": 5.504854368932039,
+      "grad_norm": 0.8209558296520447,
+      "learning_rate": 3.5145190860294043e-06,
+      "loss": 0.2344,
+      "step": 567
+    },
+    {
+      "epoch": 5.514563106796117,
+      "grad_norm": 0.6284068735301757,
+      "learning_rate": 3.5098708990711254e-06,
+      "loss": 0.1831,
+      "step": 568
+    },
+    {
+      "epoch": 5.524271844660194,
+      "grad_norm": 0.7757264445876239,
+      "learning_rate": 3.505218536612869e-06,
+      "loss": 0.3549,
+      "step": 569
+    },
+    {
+      "epoch": 5.533980582524272,
+      "grad_norm": 0.7719191434722412,
+      "learning_rate": 3.500562017890695e-06,
+      "loss": 0.17,
+      "step": 570
+    },
+    {
+      "epoch": 5.543689320388349,
+      "grad_norm": 0.8121655179279875,
+      "learning_rate": 3.495901362157853e-06,
+      "loss": 0.1626,
+      "step": 571
+    },
+    {
+      "epoch": 5.553398058252427,
+      "grad_norm": 0.9163623739345025,
+      "learning_rate": 3.4912365886846934e-06,
+      "loss": 0.5342,
+      "step": 572
+    },
+    {
+      "epoch": 5.563106796116505,
+      "grad_norm": 0.8394373837788445,
+      "learning_rate": 3.4865677167585942e-06,
+      "loss": 0.3415,
+      "step": 573
+    },
+    {
+      "epoch": 5.572815533980583,
+      "grad_norm": 0.7873980583937358,
+      "learning_rate": 3.4818947656838796e-06,
+      "loss": 0.224,
+      "step": 574
+    },
+    {
+      "epoch": 5.58252427184466,
+      "grad_norm": 0.7467839551701586,
+      "learning_rate": 3.4772177547817387e-06,
+      "loss": 0.16,
+      "step": 575
+    },
+    {
+      "epoch": 5.592233009708738,
+      "grad_norm": 0.8007193223538516,
+      "learning_rate": 3.472536703390148e-06,
+      "loss": 0.2182,
+      "step": 576
+    },
+    {
+      "epoch": 5.601941747572815,
+      "grad_norm": 0.6853648803562354,
+      "learning_rate": 3.467851630863789e-06,
+      "loss": 0.2209,
+      "step": 577
+    },
+    {
+      "epoch": 5.611650485436893,
+      "grad_norm": 0.7720375849124096,
+      "learning_rate": 3.463162556573969e-06,
+      "loss": 0.3613,
+      "step": 578
+    },
+    {
+      "epoch": 5.621359223300971,
+      "grad_norm": 0.7587365273730499,
+      "learning_rate": 3.4584694999085424e-06,
+      "loss": 0.4704,
+      "step": 579
+    },
+    {
+      "epoch": 5.631067961165049,
+      "grad_norm": 0.7050203026843053,
+      "learning_rate": 3.4537724802718294e-06,
+      "loss": 0.246,
+      "step": 580
+    },
+    {
+      "epoch": 5.640776699029126,
+      "grad_norm": 0.7213230006573722,
+      "learning_rate": 3.4490715170845356e-06,
+      "loss": 0.2496,
+      "step": 581
+    },
+    {
+      "epoch": 5.650485436893204,
+      "grad_norm": 1.5911224887444453,
+      "learning_rate": 3.4443666297836715e-06,
+      "loss": 0.2489,
+      "step": 582
+    },
+    {
+      "epoch": 5.660194174757281,
+      "grad_norm": 0.8993220255740552,
+      "learning_rate": 3.4396578378224734e-06,
+      "loss": 0.2876,
+      "step": 583
+    },
+    {
+      "epoch": 5.669902912621359,
+      "grad_norm": 0.8319084775768927,
+      "learning_rate": 3.4349451606703214e-06,
+      "loss": 0.3687,
+      "step": 584
+    },
+    {
+      "epoch": 5.679611650485437,
+      "grad_norm": 0.7249535982488255,
+      "learning_rate": 3.430228617812661e-06,
+      "loss": 0.199,
+      "step": 585
+    },
+    {
+      "epoch": 5.689320388349515,
+      "grad_norm": 0.5875254159252349,
+      "learning_rate": 3.4255082287509183e-06,
+      "loss": 0.1397,
+      "step": 586
+    },
+    {
+      "epoch": 5.699029126213592,
+      "grad_norm": 1.7482143348929224,
+      "learning_rate": 3.420784013002426e-06,
+      "loss": 0.2072,
+      "step": 587
+    },
+    {
+      "epoch": 5.70873786407767,
+      "grad_norm": 0.6823640493082458,
+      "learning_rate": 3.416055990100336e-06,
+      "loss": 0.2213,
+      "step": 588
+    },
+    {
+      "epoch": 5.718446601941747,
+      "grad_norm": 0.7955612652181657,
+      "learning_rate": 3.4113241795935427e-06,
+      "loss": 0.2398,
+      "step": 589
+    },
+    {
+      "epoch": 5.728155339805825,
+      "grad_norm": 0.7758046902122557,
+      "learning_rate": 3.4065886010466014e-06,
+      "loss": 0.3072,
+      "step": 590
+    },
+    {
+      "epoch": 5.737864077669903,
+      "grad_norm": 0.7533825958833359,
+      "learning_rate": 3.401849274039647e-06,
+      "loss": 0.3601,
+      "step": 591
+    },
+    {
+      "epoch": 5.747572815533981,
+      "grad_norm": 0.6544575762213413,
+      "learning_rate": 3.3971062181683117e-06,
+      "loss": 0.2311,
+      "step": 592
+    },
+    {
+      "epoch": 5.757281553398058,
+      "grad_norm": 0.8250888056688914,
+      "learning_rate": 3.3923594530436477e-06,
+      "loss": 0.3673,
+      "step": 593
+    },
+    {
+      "epoch": 5.766990291262136,
+      "grad_norm": 0.8332226268708626,
+      "learning_rate": 3.387608998292041e-06,
+      "loss": 0.29,
+      "step": 594
+    },
+    {
+      "epoch": 5.776699029126213,
+      "grad_norm": 0.9912332225583187,
+      "learning_rate": 3.382854873555137e-06,
+      "loss": 0.191,
+      "step": 595
+    },
+    {
+      "epoch": 5.786407766990291,
+      "grad_norm": 0.7564530094127416,
+      "learning_rate": 3.3780970984897504e-06,
+      "loss": 0.1793,
+      "step": 596
+    },
+    {
+      "epoch": 5.796116504854369,
+      "grad_norm": 0.8710313622557714,
+      "learning_rate": 3.373335692767793e-06,
+      "loss": 0.395,
+      "step": 597
+    },
+    {
+      "epoch": 5.805825242718447,
+      "grad_norm": 0.7534705198649296,
+      "learning_rate": 3.3685706760761865e-06,
+      "loss": 0.3471,
+      "step": 598
+    },
+    {
+      "epoch": 5.815533980582524,
+      "grad_norm": 0.7266924212415772,
+      "learning_rate": 3.3638020681167827e-06,
+      "loss": 0.2073,
+      "step": 599
+    },
+    {
+      "epoch": 5.825242718446602,
+      "grad_norm": 0.6145017524574724,
+      "learning_rate": 3.3590298886062833e-06,
+      "loss": 0.2466,
+      "step": 600
+    },
+    {
+      "epoch": 5.834951456310679,
+      "grad_norm": 0.8159007931540191,
+      "learning_rate": 3.354254157276155e-06,
+      "loss": 0.4331,
+      "step": 601
+    },
+    {
+      "epoch": 5.844660194174757,
+      "grad_norm": 0.9049844421997624,
+      "learning_rate": 3.3494748938725525e-06,
+      "loss": 0.2736,
+      "step": 602
+    },
+    {
+      "epoch": 5.854368932038835,
+      "grad_norm": 0.8353175945697331,
+      "learning_rate": 3.3446921181562326e-06,
+      "loss": 0.2368,
+      "step": 603
+    },
+    {
+      "epoch": 5.864077669902913,
+      "grad_norm": 0.8795663444794419,
+      "learning_rate": 3.3399058499024767e-06,
+      "loss": 0.138,
+      "step": 604
+    },
+    {
+      "epoch": 5.87378640776699,
+      "grad_norm": 0.9133299820724685,
+      "learning_rate": 3.3351161089010055e-06,
+      "loss": 0.375,
+      "step": 605
+    },
+    {
+      "epoch": 5.883495145631068,
+      "grad_norm": 0.8571212547288527,
+      "learning_rate": 3.330322914955897e-06,
+      "loss": 0.3399,
+      "step": 606
+    },
+    {
+      "epoch": 5.893203883495145,
+      "grad_norm": 0.8491851116183573,
+      "learning_rate": 3.325526287885509e-06,
+      "loss": 0.3133,
+      "step": 607
+    },
+    {
+      "epoch": 5.902912621359223,
+      "grad_norm": 0.7068426870475476,
+      "learning_rate": 3.3207262475223913e-06,
+      "loss": 0.3358,
+      "step": 608
+    },
+    {
+      "epoch": 5.9126213592233015,
+      "grad_norm": 0.8823539443693329,
+      "learning_rate": 3.315922813713209e-06,
+      "loss": 0.463,
+      "step": 609
+    },
+    {
+      "epoch": 5.922330097087379,
+      "grad_norm": 0.8082936904142045,
+      "learning_rate": 3.3111160063186553e-06,
+      "loss": 0.2456,
+      "step": 610
+    },
+    {
+      "epoch": 5.932038834951456,
+      "grad_norm": 0.8575417238390962,
+      "learning_rate": 3.3063058452133756e-06,
+      "loss": 0.356,
+      "step": 611
+    },
+    {
+      "epoch": 5.941747572815534,
+      "grad_norm": 0.9195204085165103,
+      "learning_rate": 3.301492350285879e-06,
+      "loss": 0.3163,
+      "step": 612
+    },
+    {
+      "epoch": 5.951456310679612,
+      "grad_norm": 0.7507019258704335,
+      "learning_rate": 3.296675541438461e-06,
+      "loss": 0.1877,
+      "step": 613
+    },
+    {
+      "epoch": 5.961165048543689,
+      "grad_norm": 0.8316963047488544,
+      "learning_rate": 3.2918554385871163e-06,
+      "loss": 0.408,
+      "step": 614
+    },
+    {
+      "epoch": 5.970873786407767,
+      "grad_norm": 0.7669332051739475,
+      "learning_rate": 3.2870320616614626e-06,
+      "loss": 0.1992,
+      "step": 615
+    },
+    {
+      "epoch": 5.980582524271845,
+      "grad_norm": 0.8022721375687455,
+      "learning_rate": 3.282205430604653e-06,
+      "loss": 0.2359,
+      "step": 616
+    },
+    {
+      "epoch": 5.990291262135923,
+      "grad_norm": 0.8261290006148897,
+      "learning_rate": 3.2773755653732954e-06,
+      "loss": 0.2569,
+      "step": 617
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.6791133651060106,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.2842,
+      "step": 618
+    },
+    {
+      "epoch": 6.009708737864078,
+      "grad_norm": 0.8707791160221153,
+      "learning_rate": 3.267706212280146e-06,
+      "loss": 0.1839,
+      "step": 619
+    },
+    {
+      "epoch": 6.019417475728155,
+      "grad_norm": 0.7963631564153827,
+      "learning_rate": 3.2628667643981036e-06,
+      "loss": 0.1891,
+      "step": 620
+    },
+    {
+      "epoch": 6.029126213592233,
+      "grad_norm": 1.0279730453071867,
+      "learning_rate": 3.2580241623008426e-06,
+      "loss": 0.2358,
+      "step": 621
+    },
+    {
+      "epoch": 6.038834951456311,
+      "grad_norm": 1.397048986070453,
+      "learning_rate": 3.2531784260110067e-06,
+      "loss": 0.2829,
+      "step": 622
+    },
+    {
+      "epoch": 6.048543689320389,
+      "grad_norm": 0.7957544850491932,
+      "learning_rate": 3.2483295755641986e-06,
+      "loss": 0.2457,
+      "step": 623
+    },
+    {
+      "epoch": 6.058252427184466,
+      "grad_norm": 1.465656842900919,
+      "learning_rate": 3.243477631008897e-06,
+      "loss": 0.3518,
+      "step": 624
+    },
+    {
+      "epoch": 6.067961165048544,
+      "grad_norm": 0.9566010742642381,
+      "learning_rate": 3.238622612406373e-06,
+      "loss": 0.1091,
+      "step": 625
+    },
+    {
+      "epoch": 6.077669902912621,
+      "grad_norm": 1.094395509090259,
+      "learning_rate": 3.233764539830608e-06,
+      "loss": 0.2517,
+      "step": 626
+    },
+    {
+      "epoch": 6.087378640776699,
+      "grad_norm": 0.786465657330707,
+      "learning_rate": 3.228903433368212e-06,
+      "loss": 0.1099,
+      "step": 627
+    },
+    {
+      "epoch": 6.097087378640777,
+      "grad_norm": 0.7448290029077708,
+      "learning_rate": 3.224039313118338e-06,
+      "loss": 0.0839,
+      "step": 628
+    },
+    {
+      "epoch": 6.106796116504855,
+      "grad_norm": 0.8159852737100751,
+      "learning_rate": 3.2191721991925993e-06,
+      "loss": 0.1697,
+      "step": 629
+    },
+    {
+      "epoch": 6.116504854368932,
+      "grad_norm": 0.783092292843629,
+      "learning_rate": 3.21430211171499e-06,
+      "loss": 0.2721,
+      "step": 630
+    },
+    {
+      "epoch": 6.12621359223301,
+      "grad_norm": 0.909457312270043,
+      "learning_rate": 3.209429070821795e-06,
+      "loss": 0.4595,
+      "step": 631
+    },
+    {
+      "epoch": 6.135922330097087,
+      "grad_norm": 0.7768261343108038,
+      "learning_rate": 3.2045530966615136e-06,
+      "loss": 0.2819,
+      "step": 632
+    },
+    {
+      "epoch": 6.145631067961165,
+      "grad_norm": 0.6646429241432731,
+      "learning_rate": 3.1996742093947724e-06,
+      "loss": 0.2084,
+      "step": 633
+    },
+    {
+      "epoch": 6.155339805825243,
+      "grad_norm": 1.8192215128102225,
+      "learning_rate": 3.1947924291942423e-06,
+      "loss": 0.2926,
+      "step": 634
+    },
+    {
+      "epoch": 6.165048543689321,
+      "grad_norm": 0.9265469467709787,
+      "learning_rate": 3.189907776244556e-06,
+      "loss": 0.1355,
+      "step": 635
+    },
+    {
+      "epoch": 6.174757281553398,
+      "grad_norm": 0.8092435744627986,
+      "learning_rate": 3.185020270742225e-06,
+      "loss": 0.2611,
+      "step": 636
+    },
+    {
+      "epoch": 6.184466019417476,
+      "grad_norm": 0.9280566002154651,
+      "learning_rate": 3.180129932895553e-06,
+      "loss": 0.1779,
+      "step": 637
+    },
+    {
+      "epoch": 6.194174757281553,
+      "grad_norm": 0.790228426680879,
+      "learning_rate": 3.1752367829245563e-06,
+      "loss": 0.3931,
+      "step": 638
+    },
+    {
+      "epoch": 6.203883495145631,
+      "grad_norm": 1.0755642442690179,
+      "learning_rate": 3.1703408410608777e-06,
+      "loss": 0.4567,
+      "step": 639
+    },
+    {
+      "epoch": 6.213592233009709,
+      "grad_norm": 0.755742369789303,
+      "learning_rate": 3.1654421275477045e-06,
+      "loss": 0.2269,
+      "step": 640
+    },
+    {
+      "epoch": 6.223300970873787,
+      "grad_norm": 0.9315816211452175,
+      "learning_rate": 3.1605406626396826e-06,
+      "loss": 0.2207,
+      "step": 641
+    },
+    {
+      "epoch": 6.233009708737864,
+      "grad_norm": 0.9545007552402858,
+      "learning_rate": 3.155636466602836e-06,
+      "loss": 0.2818,
+      "step": 642
+    },
+    {
+      "epoch": 6.242718446601942,
+      "grad_norm": 0.8741584596148888,
+      "learning_rate": 3.150729559714478e-06,
+      "loss": 0.376,
+      "step": 643
+    },
+    {
+      "epoch": 6.252427184466019,
+      "grad_norm": 0.8956284681200907,
+      "learning_rate": 3.145819962263134e-06,
+      "loss": 0.2309,
+      "step": 644
+    },
+    {
+      "epoch": 6.262135922330097,
+      "grad_norm": 0.9302001043992213,
+      "learning_rate": 3.1409076945484513e-06,
+      "loss": 0.1773,
+      "step": 645
+    },
+    {
+      "epoch": 6.271844660194175,
+      "grad_norm": 0.7020688549614059,
+      "learning_rate": 3.135992776881119e-06,
+      "loss": 0.1055,
+      "step": 646
+    },
+    {
+      "epoch": 6.281553398058253,
+      "grad_norm": 0.8984373280427396,
+      "learning_rate": 3.1310752295827818e-06,
+      "loss": 0.1185,
+      "step": 647
+    },
+    {
+      "epoch": 6.29126213592233,
+      "grad_norm": 0.7618424845029574,
+      "learning_rate": 3.1261550729859602e-06,
+      "loss": 0.2409,
+      "step": 648
+    },
+    {
+      "epoch": 6.300970873786408,
+      "grad_norm": 0.8351950915081744,
+      "learning_rate": 3.12123232743396e-06,
+      "loss": 0.1645,
+      "step": 649
+    },
+    {
+      "epoch": 6.310679611650485,
+      "grad_norm": 0.8458831464212704,
+      "learning_rate": 3.116307013280793e-06,
+      "loss": 0.4059,
+      "step": 650
+    },
+    {
+      "epoch": 6.320388349514563,
+      "grad_norm": 0.6691280560723778,
+      "learning_rate": 3.1113791508910913e-06,
+      "loss": 0.2279,
+      "step": 651
+    },
+    {
+      "epoch": 6.330097087378641,
+      "grad_norm": 0.7854776986942449,
+      "learning_rate": 3.106448760640022e-06,
+      "loss": 0.3258,
+      "step": 652
+    },
+    {
+      "epoch": 6.339805825242719,
+      "grad_norm": 0.6901204342705916,
+      "learning_rate": 3.1015158629132066e-06,
+      "loss": 0.1222,
+      "step": 653
+    },
+    {
+      "epoch": 6.349514563106796,
+      "grad_norm": 0.7644991091087948,
+      "learning_rate": 3.096580478106631e-06,
+      "loss": 0.2448,
+      "step": 654
+    },
+    {
+      "epoch": 6.359223300970874,
+      "grad_norm": 0.8458191662291118,
+      "learning_rate": 3.0916426266265676e-06,
+      "loss": 0.1768,
+      "step": 655
+    },
+    {
+      "epoch": 6.368932038834951,
+      "grad_norm": 0.8114001055076686,
+      "learning_rate": 3.086702328889486e-06,
+      "loss": 0.1425,
+      "step": 656
+    },
+    {
+      "epoch": 6.378640776699029,
+      "grad_norm": 1.056470365811026,
+      "learning_rate": 3.0817596053219697e-06,
+      "loss": 0.192,
+      "step": 657
+    },
+    {
+      "epoch": 6.388349514563107,
+      "grad_norm": 0.8235670285844128,
+      "learning_rate": 3.076814476360634e-06,
+      "loss": 0.1868,
+      "step": 658
+    },
+    {
+      "epoch": 6.398058252427185,
+      "grad_norm": 0.5428827520625176,
+      "learning_rate": 3.071866962452038e-06,
+      "loss": 0.0962,
+      "step": 659
+    },
+    {
+      "epoch": 6.407766990291262,
+      "grad_norm": 1.008018313016207,
+      "learning_rate": 3.066917084052603e-06,
+      "loss": 0.3406,
+      "step": 660
+    },
+    {
+      "epoch": 6.41747572815534,
+      "grad_norm": 0.8178626481888325,
+      "learning_rate": 3.061964861628527e-06,
+      "loss": 0.3335,
+      "step": 661
+    },
+    {
+      "epoch": 6.427184466019417,
+      "grad_norm": 0.8894315711207563,
+      "learning_rate": 3.057010315655698e-06,
+      "loss": 0.1577,
+      "step": 662
+    },
+    {
+      "epoch": 6.436893203883495,
+      "grad_norm": 0.8286795084689285,
+      "learning_rate": 3.0520534666196134e-06,
+      "loss": 0.2201,
+      "step": 663
+    },
+    {
+      "epoch": 6.446601941747573,
+      "grad_norm": 0.6036483911933364,
+      "learning_rate": 3.0470943350152914e-06,
+      "loss": 0.2444,
+      "step": 664
+    },
+    {
+      "epoch": 6.456310679611651,
+      "grad_norm": 0.6709778515906393,
+      "learning_rate": 3.042132941347189e-06,
+      "loss": 0.3274,
+      "step": 665
+    },
+    {
+      "epoch": 6.466019417475728,
+      "grad_norm": 0.7976957124768677,
+      "learning_rate": 3.037169306129115e-06,
+      "loss": 0.1896,
+      "step": 666
+    },
+    {
+      "epoch": 6.475728155339806,
+      "grad_norm": 0.7689359311211758,
+      "learning_rate": 3.0322034498841475e-06,
+      "loss": 0.2606,
+      "step": 667
+    },
+    {
+      "epoch": 6.485436893203883,
+      "grad_norm": 0.6978195308947686,
+      "learning_rate": 3.027235393144547e-06,
+      "loss": 0.2605,
+      "step": 668
+    },
+    {
+      "epoch": 6.495145631067961,
+      "grad_norm": 0.7202725995379047,
+      "learning_rate": 3.0222651564516715e-06,
+      "loss": 0.3485,
+      "step": 669
+    },
+    {
+      "epoch": 6.504854368932039,
+      "grad_norm": 0.8910034600894114,
+      "learning_rate": 3.017292760355896e-06,
+      "loss": 0.1461,
+      "step": 670
+    },
+    {
+      "epoch": 6.514563106796117,
+      "grad_norm": 0.9239014497460577,
+      "learning_rate": 3.0123182254165194e-06,
+      "loss": 0.0908,
+      "step": 671
+    },
+    {
+      "epoch": 6.524271844660194,
+      "grad_norm": 1.237740412456114,
+      "learning_rate": 3.0073415722016875e-06,
+      "loss": 0.4193,
+      "step": 672
+    },
+    {
+      "epoch": 6.533980582524272,
+      "grad_norm": 0.7735186784296664,
+      "learning_rate": 3.002362821288302e-06,
+      "loss": 0.1549,
+      "step": 673
+    },
+    {
+      "epoch": 6.543689320388349,
+      "grad_norm": 0.7673008373922907,
+      "learning_rate": 2.9973819932619404e-06,
+      "loss": 0.258,
+      "step": 674
+    },
+    {
+      "epoch": 6.553398058252427,
+      "grad_norm": 0.9318778367739176,
+      "learning_rate": 2.9923991087167657e-06,
+      "loss": 0.23,
+      "step": 675
+    },
+    {
+      "epoch": 6.563106796116505,
+      "grad_norm": 0.7309274196915427,
+      "learning_rate": 2.987414188255446e-06,
+      "loss": 0.3513,
+      "step": 676
+    },
+    {
+      "epoch": 6.572815533980583,
+      "grad_norm": 1.3271998778860932,
+      "learning_rate": 2.9824272524890664e-06,
+      "loss": 0.2651,
+      "step": 677
+    },
+    {
+      "epoch": 6.58252427184466,
+      "grad_norm": 0.8381939012078518,
+      "learning_rate": 2.977438322037046e-06,
+      "loss": 0.1581,
+      "step": 678
+    },
+    {
+      "epoch": 6.592233009708738,
+      "grad_norm": 0.6919842843107336,
+      "learning_rate": 2.9724474175270485e-06,
+      "loss": 0.1908,
+      "step": 679
+    },
+    {
+      "epoch": 6.601941747572815,
+      "grad_norm": 0.7191226845250073,
+      "learning_rate": 2.967454559594903e-06,
+      "loss": 0.1725,
+      "step": 680
+    },
+    {
+      "epoch": 6.611650485436893,
+      "grad_norm": 0.7901314352404131,
+      "learning_rate": 2.9624597688845126e-06,
+      "loss": 0.2598,
+      "step": 681
+    },
+    {
+      "epoch": 6.621359223300971,
+      "grad_norm": 0.8001240477856122,
+      "learning_rate": 2.957463066047773e-06,
+      "loss": 0.2253,
+      "step": 682
+    },
+    {
+      "epoch": 6.631067961165049,
+      "grad_norm": 0.9336908746395779,
+      "learning_rate": 2.9524644717444866e-06,
+      "loss": 0.1203,
+      "step": 683
+    },
+    {
+      "epoch": 6.640776699029126,
+      "grad_norm": 0.902680866331855,
+      "learning_rate": 2.9474640066422757e-06,
+      "loss": 0.3975,
+      "step": 684
+    },
+    {
+      "epoch": 6.650485436893204,
+      "grad_norm": 0.9361390212501749,
+      "learning_rate": 2.9424616914164982e-06,
+      "loss": 0.1326,
+      "step": 685
+    },
+    {
+      "epoch": 6.660194174757281,
+      "grad_norm": 1.1810674711179745,
+      "learning_rate": 2.9374575467501605e-06,
+      "loss": 0.1679,
+      "step": 686
+    },
+    {
+      "epoch": 6.669902912621359,
+      "grad_norm": 0.8359120548897934,
+      "learning_rate": 2.9324515933338343e-06,
+      "loss": 0.1959,
+      "step": 687
+    },
+    {
+      "epoch": 6.679611650485437,
+      "grad_norm": 0.8188298340548763,
+      "learning_rate": 2.9274438518655703e-06,
+      "loss": 0.333,
+      "step": 688
+    },
+    {
+      "epoch": 6.689320388349515,
+      "grad_norm": 0.630837189067903,
+      "learning_rate": 2.9224343430508105e-06,
+      "loss": 0.1625,
+      "step": 689
+    },
+    {
+      "epoch": 6.699029126213592,
+      "grad_norm": 0.6939282578828986,
+      "learning_rate": 2.917423087602306e-06,
+      "loss": 0.2996,
+      "step": 690
+    },
+    {
+      "epoch": 6.70873786407767,
+      "grad_norm": 0.8998041887412566,
+      "learning_rate": 2.9124101062400283e-06,
+      "loss": 0.2907,
+      "step": 691
+    },
+    {
+      "epoch": 6.718446601941747,
+      "grad_norm": 0.8389800760003714,
+      "learning_rate": 2.907395419691087e-06,
+      "loss": 0.3268,
+      "step": 692
+    },
+    {
+      "epoch": 6.728155339805825,
+      "grad_norm": 0.9441856769971552,
+      "learning_rate": 2.9023790486896404e-06,
+      "loss": 0.1834,
+      "step": 693
+    },
+    {
+      "epoch": 6.737864077669903,
+      "grad_norm": 0.7207079417413059,
+      "learning_rate": 2.8973610139768114e-06,
+      "loss": 0.1713,
+      "step": 694
+    },
+    {
+      "epoch": 6.747572815533981,
+      "grad_norm": 0.9416472237882838,
+      "learning_rate": 2.8923413363006038e-06,
+      "loss": 0.3246,
+      "step": 695
+    },
+    {
+      "epoch": 6.757281553398058,
+      "grad_norm": 0.7806422775042208,
+      "learning_rate": 2.887320036415811e-06,
+      "loss": 0.1888,
+      "step": 696
+    },
+    {
+      "epoch": 6.766990291262136,
+      "grad_norm": 0.9389934047803904,
+      "learning_rate": 2.882297135083937e-06,
+      "loss": 0.1117,
+      "step": 697
+    },
+    {
+      "epoch": 6.776699029126213,
+      "grad_norm": 1.0816268250392724,
+      "learning_rate": 2.877272653073107e-06,
+      "loss": 0.1527,
+      "step": 698
+    },
+    {
+      "epoch": 6.786407766990291,
+      "grad_norm": 0.803771712642974,
+      "learning_rate": 2.87224661115798e-06,
+      "loss": 0.1918,
+      "step": 699
+    },
+    {
+      "epoch": 6.796116504854369,
+      "grad_norm": 0.76795552970201,
+      "learning_rate": 2.8672190301196655e-06,
+      "loss": 0.2738,
+      "step": 700
+    },
+    {
+      "epoch": 6.805825242718447,
+      "grad_norm": 0.891590775155618,
+      "learning_rate": 2.8621899307456376e-06,
+      "loss": 0.2631,
+      "step": 701
+    },
+    {
+      "epoch": 6.815533980582524,
+      "grad_norm": 0.9199990875863763,
+      "learning_rate": 2.8571593338296473e-06,
+      "loss": 0.213,
+      "step": 702
+    },
+    {
+      "epoch": 6.825242718446602,
+      "grad_norm": 0.5739691108494726,
+      "learning_rate": 2.8521272601716376e-06,
+      "loss": 0.1101,
+      "step": 703
+    },
+    {
+      "epoch": 6.834951456310679,
+      "grad_norm": 1.0392563744404688,
+      "learning_rate": 2.8470937305776567e-06,
+      "loss": 0.4097,
+      "step": 704
+    },
+    {
+      "epoch": 6.844660194174757,
+      "grad_norm": 0.8306609969259794,
+      "learning_rate": 2.842058765859776e-06,
+      "loss": 0.1619,
+      "step": 705
+    },
+    {
+      "epoch": 6.854368932038835,
+      "grad_norm": 0.8615685770113111,
+      "learning_rate": 2.837022386835996e-06,
+      "loss": 0.1667,
+      "step": 706
+    },
+    {
+      "epoch": 6.864077669902913,
+      "grad_norm": 0.6463762969432196,
+      "learning_rate": 2.8319846143301676e-06,
+      "loss": 0.1625,
+      "step": 707
+    },
+    {
+      "epoch": 6.87378640776699,
+      "grad_norm": 0.7927082014544528,
+      "learning_rate": 2.826945469171903e-06,
+      "loss": 0.1654,
+      "step": 708
+    },
+    {
+      "epoch": 6.883495145631068,
+      "grad_norm": 0.6577362367813564,
+      "learning_rate": 2.82190497219649e-06,
+      "loss": 0.1615,
+      "step": 709
+    },
+    {
+      "epoch": 6.893203883495145,
+      "grad_norm": 1.02468338437526,
+      "learning_rate": 2.8168631442448046e-06,
+      "loss": 0.2807,
+      "step": 710
+    },
+    {
+      "epoch": 6.902912621359223,
+      "grad_norm": 0.7697195438915208,
+      "learning_rate": 2.8118200061632273e-06,
+      "loss": 0.1068,
+      "step": 711
+    },
+    {
+      "epoch": 6.9126213592233015,
+      "grad_norm": 0.8385493264196137,
+      "learning_rate": 2.8067755788035544e-06,
+      "loss": 0.3729,
+      "step": 712
+    },
+    {
+      "epoch": 6.922330097087379,
+      "grad_norm": 1.0076993590471206,
+      "learning_rate": 2.801729883022915e-06,
+      "loss": 0.172,
+      "step": 713
+    },
+    {
+      "epoch": 6.932038834951456,
+      "grad_norm": 0.7627948189362331,
+      "learning_rate": 2.7966829396836804e-06,
+      "loss": 0.2992,
+      "step": 714
+    },
+    {
+      "epoch": 6.941747572815534,
+      "grad_norm": 0.8903733137540231,
+      "learning_rate": 2.791634769653381e-06,
+      "loss": 0.1731,
+      "step": 715
+    },
+    {
+      "epoch": 6.951456310679612,
+      "grad_norm": 0.8121494855617393,
+      "learning_rate": 2.78658539380462e-06,
+      "loss": 0.3445,
+      "step": 716
+    },
+    {
+      "epoch": 6.961165048543689,
+      "grad_norm": 0.6831177053115758,
+      "learning_rate": 2.781534833014985e-06,
+      "loss": 0.2196,
+      "step": 717
+    },
+    {
+      "epoch": 6.970873786407767,
+      "grad_norm": 0.8060126813685334,
+      "learning_rate": 2.7764831081669635e-06,
+      "loss": 0.3198,
+      "step": 718
+    },
+    {
+      "epoch": 6.980582524271845,
+      "grad_norm": 0.559970759452349,
+      "learning_rate": 2.771430240147856e-06,
+      "loss": 0.1326,
+      "step": 719
+    },
+    {
+      "epoch": 6.990291262135923,
+      "grad_norm": 1.018187535764793,
+      "learning_rate": 2.7663762498496905e-06,
+      "loss": 0.2745,
+      "step": 720
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.7403875608736046,
+      "learning_rate": 2.761321158169134e-06,
+      "loss": 0.1285,
+      "step": 721
+    },
+    {
+      "epoch": 7.009708737864078,
+      "grad_norm": 0.836665956792195,
+      "learning_rate": 2.7562649860074077e-06,
+      "loss": 0.1438,
+      "step": 722
+    },
+    {
+      "epoch": 7.019417475728155,
+      "grad_norm": 0.9057591504597606,
+      "learning_rate": 2.7512077542702005e-06,
+      "loss": 0.4176,
+      "step": 723
+    },
+    {
+      "epoch": 7.029126213592233,
+      "grad_norm": 0.7310746968198095,
+      "learning_rate": 2.746149483867582e-06,
+      "loss": 0.125,
+      "step": 724
+    },
+    {
+      "epoch": 7.038834951456311,
+      "grad_norm": 0.8094466402699106,
+      "learning_rate": 2.741090195713917e-06,
+      "loss": 0.1358,
+      "step": 725
+    },
+    {
+      "epoch": 7.048543689320389,
+      "grad_norm": 0.8397086363403945,
+      "learning_rate": 2.736029910727777e-06,
+      "loss": 0.1356,
+      "step": 726
+    },
+    {
+      "epoch": 7.058252427184466,
+      "grad_norm": 1.3541284896158823,
+      "learning_rate": 2.730968649831858e-06,
+      "loss": 0.1502,
+      "step": 727
+    },
+    {
+      "epoch": 7.067961165048544,
+      "grad_norm": 1.1229775978036463,
+      "learning_rate": 2.7259064339528875e-06,
+      "loss": 0.3025,
+      "step": 728
+    },
+    {
+      "epoch": 7.077669902912621,
+      "grad_norm": 0.7963179988105072,
+      "learning_rate": 2.720843284021543e-06,
+      "loss": 0.2234,
+      "step": 729
+    },
+    {
+      "epoch": 7.087378640776699,
+      "grad_norm": 0.9155996246197351,
+      "learning_rate": 2.7157792209723654e-06,
+      "loss": 0.1298,
+      "step": 730
+    },
+    {
+      "epoch": 7.097087378640777,
+      "grad_norm": 0.8846875801793995,
+      "learning_rate": 2.7107142657436696e-06,
+      "loss": 0.234,
+      "step": 731
+    },
+    {
+      "epoch": 7.106796116504855,
+      "grad_norm": 0.7661475643829027,
+      "learning_rate": 2.705648439277459e-06,
+      "loss": 0.1238,
+      "step": 732
+    },
+    {
+      "epoch": 7.116504854368932,
+      "grad_norm": 0.6344540911037735,
+      "learning_rate": 2.7005817625193398e-06,
+      "loss": 0.1464,
+      "step": 733
+    },
+    {
+      "epoch": 7.12621359223301,
+      "grad_norm": 0.9875032525499696,
+      "learning_rate": 2.695514256418435e-06,
+      "loss": 0.1332,
+      "step": 734
+    },
+    {
+      "epoch": 7.135922330097087,
+      "grad_norm": 0.8841032967053356,
+      "learning_rate": 2.6904459419272955e-06,
+      "loss": 0.1499,
+      "step": 735
+    },
+    {
+      "epoch": 7.145631067961165,
+      "grad_norm": 1.0727410414660654,
+      "learning_rate": 2.685376840001814e-06,
+      "loss": 0.1845,
+      "step": 736
+    },
+    {
+      "epoch": 7.155339805825243,
+      "grad_norm": 0.7603950939388796,
+      "learning_rate": 2.6803069716011405e-06,
+      "loss": 0.0635,
+      "step": 737
+    },
+    {
+      "epoch": 7.165048543689321,
+      "grad_norm": 0.8222924068539089,
+      "learning_rate": 2.6752363576875933e-06,
+      "loss": 0.1862,
+      "step": 738
+    },
+    {
+      "epoch": 7.174757281553398,
+      "grad_norm": 0.8437247717724511,
+      "learning_rate": 2.6701650192265734e-06,
+      "loss": 0.1873,
+      "step": 739
+    },
+    {
+      "epoch": 7.184466019417476,
+      "grad_norm": 0.9780684483534852,
+      "learning_rate": 2.6650929771864776e-06,
+      "loss": 0.1166,
+      "step": 740
+    },
+    {
+      "epoch": 7.194174757281553,
+      "grad_norm": 0.8398834065613339,
+      "learning_rate": 2.660020252538611e-06,
+      "loss": 0.2882,
+      "step": 741
+    },
+    {
+      "epoch": 7.203883495145631,
+      "grad_norm": 0.6876195874715586,
+      "learning_rate": 2.6549468662571026e-06,
+      "loss": 0.1951,
+      "step": 742
+    },
+    {
+      "epoch": 7.213592233009709,
+      "grad_norm": 1.0489870539764072,
+      "learning_rate": 2.6498728393188157e-06,
+      "loss": 0.2443,
+      "step": 743
+    },
+    {
+      "epoch": 7.223300970873787,
+      "grad_norm": 0.9997066200403729,
+      "learning_rate": 2.6447981927032634e-06,
+      "loss": 0.1229,
+      "step": 744
+    },
+    {
+      "epoch": 7.233009708737864,
+      "grad_norm": 0.7857265262109651,
+      "learning_rate": 2.639722947392521e-06,
+      "loss": 0.3217,
+      "step": 745
+    },
+    {
+      "epoch": 7.242718446601942,
+      "grad_norm": 0.8361751064417774,
+      "learning_rate": 2.6346471243711376e-06,
+      "loss": 0.1699,
+      "step": 746
+    },
+    {
+      "epoch": 7.252427184466019,
+      "grad_norm": 0.6354672587304362,
+      "learning_rate": 2.629570744626052e-06,
+      "loss": 0.2213,
+      "step": 747
+    },
+    {
+      "epoch": 7.262135922330097,
+      "grad_norm": 0.691741853784474,
+      "learning_rate": 2.624493829146507e-06,
+      "loss": 0.2447,
+      "step": 748
+    },
+    {
+      "epoch": 7.271844660194175,
+      "grad_norm": 0.9444925946366645,
+      "learning_rate": 2.619416398923957e-06,
+      "loss": 0.3385,
+      "step": 749
+    },
+    {
+      "epoch": 7.281553398058253,
+      "grad_norm": 0.8079469525410382,
+      "learning_rate": 2.614338474951987e-06,
+      "loss": 0.0788,
+      "step": 750
+    },
+    {
+      "epoch": 7.29126213592233,
+      "grad_norm": 0.6978663980764732,
+      "learning_rate": 2.6092600782262213e-06,
+      "loss": 0.1165,
+      "step": 751
+    },
+    {
+      "epoch": 7.300970873786408,
+      "grad_norm": 0.7118818478635398,
+      "learning_rate": 2.6041812297442417e-06,
+      "loss": 0.1376,
+      "step": 752
+    },
+    {
+      "epoch": 7.310679611650485,
+      "grad_norm": 0.776609030527654,
+      "learning_rate": 2.5991019505054965e-06,
+      "loss": 0.1688,
+      "step": 753
+    },
+    {
+      "epoch": 7.320388349514563,
+      "grad_norm": 0.6957555306995469,
+      "learning_rate": 2.5940222615112143e-06,
+      "loss": 0.112,
+      "step": 754
+    },
+    {
+      "epoch": 7.330097087378641,
+      "grad_norm": 0.6104826494631463,
+      "learning_rate": 2.5889421837643186e-06,
+      "loss": 0.1296,
+      "step": 755
+    },
+    {
+      "epoch": 7.339805825242719,
+      "grad_norm": 0.7019412036534555,
+      "learning_rate": 2.5838617382693415e-06,
+      "loss": 0.2872,
+      "step": 756
+    },
+    {
+      "epoch": 7.349514563106796,
+      "grad_norm": 0.91073647902368,
+      "learning_rate": 2.5787809460323337e-06,
+      "loss": 0.1448,
+      "step": 757
+    },
+    {
+      "epoch": 7.359223300970874,
+      "grad_norm": 0.7283515289799285,
+      "learning_rate": 2.57369982806078e-06,
+      "loss": 0.2016,
+      "step": 758
+    },
+    {
+      "epoch": 7.368932038834951,
+      "grad_norm": 0.8223274511720547,
+      "learning_rate": 2.5686184053635127e-06,
+      "loss": 0.1389,
+      "step": 759
+    },
+    {
+      "epoch": 7.378640776699029,
+      "grad_norm": 0.7050953772126068,
+      "learning_rate": 2.563536698950624e-06,
+      "loss": 0.1134,
+      "step": 760
+    },
+    {
+      "epoch": 7.388349514563107,
+      "grad_norm": 0.5692667770218556,
+      "learning_rate": 2.5584547298333772e-06,
+      "loss": 0.0578,
+      "step": 761
+    },
+    {
+      "epoch": 7.398058252427185,
+      "grad_norm": 0.6216050157913948,
+      "learning_rate": 2.5533725190241255e-06,
+      "loss": 0.0397,
+      "step": 762
+    },
+    {
+      "epoch": 7.407766990291262,
+      "grad_norm": 0.6761069814061764,
+      "learning_rate": 2.5482900875362184e-06,
+      "loss": 0.1089,
+      "step": 763
+    },
+    {
+      "epoch": 7.41747572815534,
+      "grad_norm": 1.0199372568203338,
+      "learning_rate": 2.543207456383919e-06,
+      "loss": 0.1276,
+      "step": 764
+    },
+    {
+      "epoch": 7.427184466019417,
+      "grad_norm": 0.8577131476418797,
+      "learning_rate": 2.538124646582315e-06,
+      "loss": 0.2917,
+      "step": 765
+    },
+    {
+      "epoch": 7.436893203883495,
+      "grad_norm": 0.7214685312852196,
+      "learning_rate": 2.533041679147235e-06,
+      "loss": 0.1534,
+      "step": 766
+    },
+    {
+      "epoch": 7.446601941747573,
+      "grad_norm": 0.7949192422368898,
+      "learning_rate": 2.527958575095157e-06,
+      "loss": 0.1386,
+      "step": 767
+    },
+    {
+      "epoch": 7.456310679611651,
+      "grad_norm": 0.9275307940564848,
+      "learning_rate": 2.522875355443124e-06,
+      "loss": 0.2946,
+      "step": 768
+    },
+    {
+      "epoch": 7.466019417475728,
+      "grad_norm": 0.9634584569650436,
+      "learning_rate": 2.5177920412086586e-06,
+      "loss": 0.258,
+      "step": 769
+    },
+    {
+      "epoch": 7.475728155339806,
+      "grad_norm": 0.8797775717303168,
+      "learning_rate": 2.512708653409674e-06,
+      "loss": 0.1895,
+      "step": 770
+    },
+    {
+      "epoch": 7.485436893203883,
+      "grad_norm": 0.7020792385043,
+      "learning_rate": 2.507625213064386e-06,
+      "loss": 0.1305,
+      "step": 771
+    },
+    {
+      "epoch": 7.495145631067961,
+      "grad_norm": 0.6721656885922416,
+      "learning_rate": 2.5025417411912307e-06,
+      "loss": 0.0923,
+      "step": 772
+    },
+    {
+      "epoch": 7.504854368932039,
+      "grad_norm": 0.9046829126351137,
+      "learning_rate": 2.4974582588087697e-06,
+      "loss": 0.2464,
+      "step": 773
+    },
+    {
+      "epoch": 7.514563106796117,
+      "grad_norm": 0.8690288633052726,
+      "learning_rate": 2.492374786935614e-06,
+      "loss": 0.2224,
+      "step": 774
+    },
+    {
+      "epoch": 7.524271844660194,
+      "grad_norm": 0.8553378382148052,
+      "learning_rate": 2.487291346590326e-06,
+      "loss": 0.2824,
+      "step": 775
+    },
+    {
+      "epoch": 7.533980582524272,
+      "grad_norm": 0.6651089727018585,
+      "learning_rate": 2.4822079587913414e-06,
+      "loss": 0.197,
+      "step": 776
+    },
+    {
+      "epoch": 7.543689320388349,
+      "grad_norm": 0.7135997471244006,
+      "learning_rate": 2.4771246445568763e-06,
+      "loss": 0.098,
+      "step": 777
+    },
+    {
+      "epoch": 7.553398058252427,
+      "grad_norm": 1.0277044305758947,
+      "learning_rate": 2.472041424904844e-06,
+      "loss": 0.123,
+      "step": 778
+    },
+    {
+      "epoch": 7.563106796116505,
+      "grad_norm": 0.7865144489394965,
+      "learning_rate": 2.466958320852766e-06,
+      "loss": 0.1981,
+      "step": 779
+    },
+    {
+      "epoch": 7.572815533980583,
+      "grad_norm": 0.987488048755019,
+      "learning_rate": 2.4618753534176854e-06,
+      "loss": 0.169,
+      "step": 780
+    },
+    {
+      "epoch": 7.58252427184466,
+      "grad_norm": 0.8166300006729655,
+      "learning_rate": 2.4567925436160823e-06,
+      "loss": 0.1999,
+      "step": 781
+    },
+    {
+      "epoch": 7.592233009708738,
+      "grad_norm": 0.81581841653359,
+      "learning_rate": 2.4517099124637824e-06,
+      "loss": 0.2586,
+      "step": 782
+    },
+    {
+      "epoch": 7.601941747572815,
+      "grad_norm": 1.108761709333979,
+      "learning_rate": 2.4466274809758757e-06,
+      "loss": 0.1429,
+      "step": 783
+    },
+    {
+      "epoch": 7.611650485436893,
+      "grad_norm": 0.7640567025062378,
+      "learning_rate": 2.4415452701666236e-06,
+      "loss": 0.1963,
+      "step": 784
+    },
+    {
+      "epoch": 7.621359223300971,
+      "grad_norm": 0.7929183080114451,
+      "learning_rate": 2.436463301049378e-06,
+      "loss": 0.2358,
+      "step": 785
+    },
+    {
+      "epoch": 7.631067961165049,
+      "grad_norm": 0.9165311722428774,
+      "learning_rate": 2.431381594636488e-06,
+      "loss": 0.2054,
+      "step": 786
+    },
+    {
+      "epoch": 7.640776699029126,
+      "grad_norm": 0.911092735481885,
+      "learning_rate": 2.42630017193922e-06,
+      "loss": 0.3815,
+      "step": 787
+    },
+    {
+      "epoch": 7.650485436893204,
+      "grad_norm": 0.7386605111856521,
+      "learning_rate": 2.4212190539676667e-06,
+      "loss": 0.055,
+      "step": 788
+    },
+    {
+      "epoch": 7.660194174757281,
+      "grad_norm": 0.7852123838015137,
+      "learning_rate": 2.4161382617306585e-06,
+      "loss": 0.2142,
+      "step": 789
+    },
+    {
+      "epoch": 7.669902912621359,
+      "grad_norm": 0.8671129620068961,
+      "learning_rate": 2.4110578162356814e-06,
+      "loss": 0.1295,
+      "step": 790
+    },
+    {
+      "epoch": 7.679611650485437,
+      "grad_norm": 0.8379544581414955,
+      "learning_rate": 2.405977738488786e-06,
+      "loss": 0.1811,
+      "step": 791
+    },
+    {
+      "epoch": 7.689320388349515,
+      "grad_norm": 0.8279537297736859,
+      "learning_rate": 2.4008980494945044e-06,
+      "loss": 0.1764,
+      "step": 792
+    },
+    {
+      "epoch": 7.699029126213592,
+      "grad_norm": 0.7533139895873169,
+      "learning_rate": 2.3958187702557587e-06,
+      "loss": 0.0803,
+      "step": 793
+    },
+    {
+      "epoch": 7.70873786407767,
+      "grad_norm": 0.7827211446349897,
+      "learning_rate": 2.39073992177378e-06,
+      "loss": 0.0769,
+      "step": 794
+    },
+    {
+      "epoch": 7.718446601941747,
+      "grad_norm": 0.870755479127296,
+      "learning_rate": 2.385661525048014e-06,
+      "loss": 0.1638,
+      "step": 795
+    },
+    {
+      "epoch": 7.728155339805825,
+      "grad_norm": 0.9060633189944168,
+      "learning_rate": 2.3805836010760435e-06,
+      "loss": 0.1948,
+      "step": 796
+    },
+    {
+      "epoch": 7.737864077669903,
+      "grad_norm": 0.785817720430874,
+      "learning_rate": 2.375506170853494e-06,
+      "loss": 0.1063,
+      "step": 797
+    },
+    {
+      "epoch": 7.747572815533981,
+      "grad_norm": 0.6084463077588024,
+      "learning_rate": 2.3704292553739487e-06,
+      "loss": 0.0805,
+      "step": 798
+    },
+    {
+      "epoch": 7.757281553398058,
+      "grad_norm": 0.6284328864307693,
+      "learning_rate": 2.3653528756288636e-06,
+      "loss": 0.1343,
+      "step": 799
+    },
+    {
+      "epoch": 7.766990291262136,
+      "grad_norm": 1.1072324614685045,
+      "learning_rate": 2.3602770526074804e-06,
+      "loss": 0.238,
+      "step": 800
+    },
+    {
+      "epoch": 7.776699029126213,
+      "grad_norm": 0.8771044918281113,
+      "learning_rate": 2.3552018072967375e-06,
+      "loss": 0.1482,
+      "step": 801
+    },
+    {
+      "epoch": 7.786407766990291,
+      "grad_norm": 0.6690650637514127,
+      "learning_rate": 2.3501271606811848e-06,
+      "loss": 0.0995,
+      "step": 802
+    },
+    {
+      "epoch": 7.796116504854369,
+      "grad_norm": 0.7709052345206647,
+      "learning_rate": 2.345053133742898e-06,
+      "loss": 0.2694,
+      "step": 803
+    },
+    {
+      "epoch": 7.805825242718447,
+      "grad_norm": 0.6349998040433816,
+      "learning_rate": 2.3399797474613894e-06,
+      "loss": 0.0663,
+      "step": 804
+    },
+    {
+      "epoch": 7.815533980582524,
+      "grad_norm": 1.0261288250181526,
+      "learning_rate": 2.334907022813523e-06,
+      "loss": 0.1044,
+      "step": 805
+    },
+    {
+      "epoch": 7.825242718446602,
+      "grad_norm": 0.6267009972409442,
+      "learning_rate": 2.329834980773427e-06,
+      "loss": 0.2633,
+      "step": 806
+    },
+    {
+      "epoch": 7.834951456310679,
+      "grad_norm": 0.918811037432451,
+      "learning_rate": 2.324763642312407e-06,
+      "loss": 0.1622,
+      "step": 807
+    },
+    {
+      "epoch": 7.844660194174757,
+      "grad_norm": 0.739956492123394,
+      "learning_rate": 2.3196930283988603e-06,
+      "loss": 0.2448,
+      "step": 808
+    },
+    {
+      "epoch": 7.854368932038835,
+      "grad_norm": 0.7010116121906006,
+      "learning_rate": 2.3146231599981865e-06,
+      "loss": 0.2406,
+      "step": 809
+    },
+    {
+      "epoch": 7.864077669902913,
+      "grad_norm": 0.7415696256419213,
+      "learning_rate": 2.3095540580727054e-06,
+      "loss": 0.279,
+      "step": 810
+    },
+    {
+      "epoch": 7.87378640776699,
+      "grad_norm": 0.7935876878721858,
+      "learning_rate": 2.304485743581566e-06,
+      "loss": 0.2569,
+      "step": 811
+    },
+    {
+      "epoch": 7.883495145631068,
+      "grad_norm": 1.2683791645513949,
+      "learning_rate": 2.299418237480661e-06,
+      "loss": 0.0492,
+      "step": 812
+    },
+    {
+      "epoch": 7.893203883495145,
+      "grad_norm": 0.8300472882833771,
+      "learning_rate": 2.294351560722542e-06,
+      "loss": 0.2097,
+      "step": 813
+    },
+    {
+      "epoch": 7.902912621359223,
+      "grad_norm": 0.770677712112457,
+      "learning_rate": 2.2892857342563316e-06,
+      "loss": 0.0987,
+      "step": 814
+    },
+    {
+      "epoch": 7.9126213592233015,
+      "grad_norm": 0.810806375980374,
+      "learning_rate": 2.2842207790276355e-06,
+      "loss": 0.2034,
+      "step": 815
+    },
+    {
+      "epoch": 7.922330097087379,
+      "grad_norm": 0.865025284529186,
+      "learning_rate": 2.279156715978457e-06,
+      "loss": 0.2431,
+      "step": 816
+    },
+    {
+      "epoch": 7.932038834951456,
+      "grad_norm": 0.7151367709362322,
+      "learning_rate": 2.274093566047113e-06,
+      "loss": 0.1682,
+      "step": 817
+    },
+    {
+      "epoch": 7.941747572815534,
+      "grad_norm": 0.7316758258717895,
+      "learning_rate": 2.2690313501681426e-06,
+      "loss": 0.1433,
+      "step": 818
+    },
+    {
+      "epoch": 7.951456310679612,
+      "grad_norm": 0.7394666643705587,
+      "learning_rate": 2.263970089272223e-06,
+      "loss": 0.178,
+      "step": 819
+    },
+    {
+      "epoch": 7.961165048543689,
+      "grad_norm": 0.654876224823624,
+      "learning_rate": 2.2589098042860838e-06,
+      "loss": 0.1848,
+      "step": 820
+    },
+    {
+      "epoch": 7.970873786407767,
+      "grad_norm": 0.6696537239231014,
+      "learning_rate": 2.2538505161324186e-06,
+      "loss": 0.0858,
+      "step": 821
+    },
+    {
+      "epoch": 7.980582524271845,
+      "grad_norm": 0.9132395771066292,
+      "learning_rate": 2.2487922457298007e-06,
+      "loss": 0.086,
+      "step": 822
+    },
+    {
+      "epoch": 7.990291262135923,
+      "grad_norm": 0.8888174000325908,
+      "learning_rate": 2.243735013992593e-06,
+      "loss": 0.2508,
+      "step": 823
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.7061956513505121,
+      "learning_rate": 2.238678841830867e-06,
+      "loss": 0.1326,
+      "step": 824
+    },
+    {
+      "epoch": 8.009708737864077,
+      "grad_norm": 0.6096642773859534,
+      "learning_rate": 2.2336237501503103e-06,
+      "loss": 0.0859,
+      "step": 825
+    },
+    {
+      "epoch": 8.019417475728156,
+      "grad_norm": 0.874150385218761,
+      "learning_rate": 2.2285697598521446e-06,
+      "loss": 0.0705,
+      "step": 826
+    },
+    {
+      "epoch": 8.029126213592233,
+      "grad_norm": 0.6120565251058548,
+      "learning_rate": 2.2235168918330374e-06,
+      "loss": 0.147,
+      "step": 827
+    },
+    {
+      "epoch": 8.03883495145631,
+      "grad_norm": 0.7460242405023572,
+      "learning_rate": 2.2184651669850164e-06,
+      "loss": 0.1725,
+      "step": 828
+    },
+    {
+      "epoch": 8.048543689320388,
+      "grad_norm": 0.47300009134877163,
+      "learning_rate": 2.2134146061953814e-06,
+      "loss": 0.0186,
+      "step": 829
+    },
+    {
+      "epoch": 8.058252427184467,
+      "grad_norm": 0.978149051086085,
+      "learning_rate": 2.2083652303466196e-06,
+      "loss": 0.2288,
+      "step": 830
+    },
+    {
+      "epoch": 8.067961165048544,
+      "grad_norm": 0.9177514365909158,
+      "learning_rate": 2.20331706031632e-06,
+      "loss": 0.1992,
+      "step": 831
+    },
+    {
+      "epoch": 8.077669902912621,
+      "grad_norm": 0.6550516508961114,
+      "learning_rate": 2.1982701169770853e-06,
+      "loss": 0.1029,
+      "step": 832
+    },
+    {
+      "epoch": 8.087378640776699,
+      "grad_norm": 0.7730840644031025,
+      "learning_rate": 2.1932244211964456e-06,
+      "loss": 0.144,
+      "step": 833
+    },
+    {
+      "epoch": 8.097087378640778,
+      "grad_norm": 0.5752559068978315,
+      "learning_rate": 2.1881799938367735e-06,
+      "loss": 0.0994,
+      "step": 834
+    },
+    {
+      "epoch": 8.106796116504855,
+      "grad_norm": 0.697423153476562,
+      "learning_rate": 2.1831368557551962e-06,
+      "loss": 0.0829,
+      "step": 835
+    },
+    {
+      "epoch": 8.116504854368932,
+      "grad_norm": 0.9600930670561553,
+      "learning_rate": 2.1780950278035114e-06,
+      "loss": 0.1914,
+      "step": 836
+    },
+    {
+      "epoch": 8.12621359223301,
+      "grad_norm": 0.8777974546702457,
+      "learning_rate": 2.173054530828098e-06,
+      "loss": 0.0827,
+      "step": 837
+    },
+    {
+      "epoch": 8.135922330097088,
+      "grad_norm": 0.6781280331911594,
+      "learning_rate": 2.168015385669833e-06,
+      "loss": 0.0882,
+      "step": 838
+    },
+    {
+      "epoch": 8.145631067961165,
+      "grad_norm": 1.002740237182081,
+      "learning_rate": 2.162977613164005e-06,
+      "loss": 0.2869,
+      "step": 839
+    },
+    {
+      "epoch": 8.155339805825243,
+      "grad_norm": 1.1050895301986552,
+      "learning_rate": 2.157941234140225e-06,
+      "loss": 0.2794,
+      "step": 840
+    },
+    {
+      "epoch": 8.16504854368932,
+      "grad_norm": 0.6988453726767133,
+      "learning_rate": 2.1529062694223437e-06,
+      "loss": 0.153,
+      "step": 841
+    },
+    {
+      "epoch": 8.174757281553399,
+      "grad_norm": 0.4102856314598248,
+      "learning_rate": 2.147872739828364e-06,
+      "loss": 0.0422,
+      "step": 842
+    },
+    {
+      "epoch": 8.184466019417476,
+      "grad_norm": 0.6732421099515787,
+      "learning_rate": 2.142840666170354e-06,
+      "loss": 0.0614,
+      "step": 843
+    },
+    {
+      "epoch": 8.194174757281553,
+      "grad_norm": 0.5179106274827621,
+      "learning_rate": 2.1378100692543637e-06,
+      "loss": 0.075,
+      "step": 844
+    },
+    {
+      "epoch": 8.20388349514563,
+      "grad_norm": 0.6248672879860213,
+      "learning_rate": 2.1327809698803354e-06,
+      "loss": 0.1265,
+      "step": 845
+    },
+    {
+      "epoch": 8.21359223300971,
+      "grad_norm": 0.6770539867953491,
+      "learning_rate": 2.1277533888420203e-06,
+      "loss": 0.1378,
+      "step": 846
+    },
+    {
+      "epoch": 8.223300970873787,
+      "grad_norm": 0.5868543420862828,
+      "learning_rate": 2.1227273469268932e-06,
+      "loss": 0.0833,
+      "step": 847
+    },
+    {
+      "epoch": 8.233009708737864,
+      "grad_norm": 0.6864008955143607,
+      "learning_rate": 2.117702864916063e-06,
+      "loss": 0.0868,
+      "step": 848
+    },
+    {
+      "epoch": 8.242718446601941,
+      "grad_norm": 1.103909018912092,
+      "learning_rate": 2.1126799635841897e-06,
+      "loss": 0.1473,
+      "step": 849
+    },
+    {
+      "epoch": 8.25242718446602,
+      "grad_norm": 0.8183458395730451,
+      "learning_rate": 2.1076586636993975e-06,
+      "loss": 0.1373,
+      "step": 850
+    },
+    {
+      "epoch": 8.262135922330097,
+      "grad_norm": 0.715445629716324,
+      "learning_rate": 2.102638986023189e-06,
+      "loss": 0.105,
+      "step": 851
+    },
+    {
+      "epoch": 8.271844660194175,
+      "grad_norm": 0.7455938407063636,
+      "learning_rate": 2.0976209513103604e-06,
+      "loss": 0.1504,
+      "step": 852
+    },
+    {
+      "epoch": 8.281553398058252,
+      "grad_norm": 0.7023437585403879,
+      "learning_rate": 2.0926045803089135e-06,
+      "loss": 0.0817,
+      "step": 853
+    },
+    {
+      "epoch": 8.29126213592233,
+      "grad_norm": 1.2518758042906253,
+      "learning_rate": 2.087589893759972e-06,
+      "loss": 0.0936,
+      "step": 854
+    },
+    {
+      "epoch": 8.300970873786408,
+      "grad_norm": 0.5636684222747409,
+      "learning_rate": 2.0825769123976954e-06,
+      "loss": 0.1198,
+      "step": 855
+    },
+    {
+      "epoch": 8.310679611650485,
+      "grad_norm": 0.7849653948615573,
+      "learning_rate": 2.077565656949191e-06,
+      "loss": 0.1976,
+      "step": 856
+    },
+    {
+      "epoch": 8.320388349514563,
+      "grad_norm": 0.785519727376682,
+      "learning_rate": 2.072556148134431e-06,
+      "loss": 0.1825,
+      "step": 857
+    },
+    {
+      "epoch": 8.330097087378642,
+      "grad_norm": 0.7359895734584899,
+      "learning_rate": 2.0675484066661666e-06,
+      "loss": 0.1522,
+      "step": 858
+    },
+    {
+      "epoch": 8.339805825242719,
+      "grad_norm": 0.5593109629965097,
+      "learning_rate": 2.0625424532498407e-06,
+      "loss": 0.0722,
+      "step": 859
+    },
+    {
+      "epoch": 8.349514563106796,
+      "grad_norm": 0.6111935300174707,
+      "learning_rate": 2.057538308583502e-06,
+      "loss": 0.0648,
+      "step": 860
+    },
+    {
+      "epoch": 8.359223300970873,
+      "grad_norm": 1.1576569133730004,
+      "learning_rate": 2.0525359933577243e-06,
+      "loss": 0.2763,
+      "step": 861
+    },
+    {
+      "epoch": 8.368932038834952,
+      "grad_norm": 0.8296811567162217,
+      "learning_rate": 2.047535528255514e-06,
+      "loss": 0.1847,
+      "step": 862
+    },
+    {
+      "epoch": 8.37864077669903,
+      "grad_norm": 0.8205871431038276,
+      "learning_rate": 2.0425369339522276e-06,
+      "loss": 0.0166,
+      "step": 863
+    },
+    {
+      "epoch": 8.388349514563107,
+      "grad_norm": 0.6447713146205751,
+      "learning_rate": 2.0375402311154886e-06,
+      "loss": 0.1415,
+      "step": 864
+    },
+    {
+      "epoch": 8.398058252427184,
+      "grad_norm": 0.4030106740717843,
+      "learning_rate": 2.0325454404050983e-06,
+      "loss": 0.0142,
+      "step": 865
+    },
+    {
+      "epoch": 8.407766990291263,
+      "grad_norm": 1.0171802020707985,
+      "learning_rate": 2.0275525824729523e-06,
+      "loss": 0.2016,
+      "step": 866
+    },
+    {
+      "epoch": 8.41747572815534,
+      "grad_norm": 0.6994124188590608,
+      "learning_rate": 2.022561677962955e-06,
+      "loss": 0.1537,
+      "step": 867
+    },
+    {
+      "epoch": 8.427184466019417,
+      "grad_norm": 0.7402451386749993,
+      "learning_rate": 2.017572747510934e-06,
+      "loss": 0.1046,
+      "step": 868
+    },
+    {
+      "epoch": 8.436893203883495,
+      "grad_norm": 0.7846549321219498,
+      "learning_rate": 2.012585811744555e-06,
+      "loss": 0.1425,
+      "step": 869
+    },
+    {
+      "epoch": 8.446601941747574,
+      "grad_norm": 1.440584552164248,
+      "learning_rate": 2.0076008912832355e-06,
+      "loss": 0.2161,
+      "step": 870
+    },
+    {
+      "epoch": 8.45631067961165,
+      "grad_norm": 0.9407034033302534,
+      "learning_rate": 2.002618006738061e-06,
+      "loss": 0.2049,
+      "step": 871
+    },
+    {
+      "epoch": 8.466019417475728,
+      "grad_norm": 0.742221931800229,
+      "learning_rate": 1.9976371787116992e-06,
+      "loss": 0.1399,
+      "step": 872
+    },
+    {
+      "epoch": 8.475728155339805,
+      "grad_norm": 0.5517495585998253,
+      "learning_rate": 1.9926584277983134e-06,
+      "loss": 0.0586,
+      "step": 873
+    },
+    {
+      "epoch": 8.485436893203884,
+      "grad_norm": 0.9431771754400559,
+      "learning_rate": 1.9876817745834805e-06,
+      "loss": 0.1939,
+      "step": 874
+    },
+    {
+      "epoch": 8.495145631067961,
+      "grad_norm": 0.6915916778894676,
+      "learning_rate": 1.9827072396441044e-06,
+      "loss": 0.1943,
+      "step": 875
+    },
+    {
+      "epoch": 8.504854368932039,
+      "grad_norm": 0.7708115028681229,
+      "learning_rate": 1.9777348435483285e-06,
+      "loss": 0.1584,
+      "step": 876
+    },
+    {
+      "epoch": 8.514563106796116,
+      "grad_norm": 1.105190714246728,
+      "learning_rate": 1.972764606855454e-06,
+      "loss": 0.2334,
+      "step": 877
+    },
+    {
+      "epoch": 8.524271844660195,
+      "grad_norm": 0.6628366560401108,
+      "learning_rate": 1.9677965501158534e-06,
+      "loss": 0.1279,
+      "step": 878
+    },
+    {
+      "epoch": 8.533980582524272,
+      "grad_norm": 0.6913345579869764,
+      "learning_rate": 1.9628306938708857e-06,
+      "loss": 0.1757,
+      "step": 879
+    },
+    {
+      "epoch": 8.54368932038835,
+      "grad_norm": 0.6844557174447111,
+      "learning_rate": 1.957867058652812e-06,
+      "loss": 0.125,
+      "step": 880
+    },
+    {
+      "epoch": 8.553398058252426,
+      "grad_norm": 0.6999128351223273,
+      "learning_rate": 1.952905664984709e-06,
+      "loss": 0.2437,
+      "step": 881
+    },
+    {
+      "epoch": 8.563106796116505,
+      "grad_norm": 0.575971990243284,
+      "learning_rate": 1.947946533380387e-06,
+      "loss": 0.0885,
+      "step": 882
+    },
+    {
+      "epoch": 8.572815533980583,
+      "grad_norm": 0.5928962221916307,
+      "learning_rate": 1.9429896843443025e-06,
+      "loss": 0.0689,
+      "step": 883
+    },
+    {
+      "epoch": 8.58252427184466,
+      "grad_norm": 0.7868027361400175,
+      "learning_rate": 1.938035138371474e-06,
+      "loss": 0.3051,
+      "step": 884
+    },
+    {
+      "epoch": 8.592233009708737,
+      "grad_norm": 0.5972483182228643,
+      "learning_rate": 1.933082915947398e-06,
+      "loss": 0.1386,
+      "step": 885
+    },
+    {
+      "epoch": 8.601941747572816,
+      "grad_norm": 0.6175357677063205,
+      "learning_rate": 1.928133037547963e-06,
+      "loss": 0.1086,
+      "step": 886
+    },
+    {
+      "epoch": 8.611650485436893,
+      "grad_norm": 0.7543183263678916,
+      "learning_rate": 1.9231855236393677e-06,
+      "loss": 0.1329,
+      "step": 887
+    },
+    {
+      "epoch": 8.62135922330097,
+      "grad_norm": 0.6446977874615103,
+      "learning_rate": 1.9182403946780316e-06,
+      "loss": 0.1087,
+      "step": 888
+    },
+    {
+      "epoch": 8.631067961165048,
+      "grad_norm": 0.79165417035794,
+      "learning_rate": 1.9132976711105146e-06,
+      "loss": 0.2162,
+      "step": 889
+    },
+    {
+      "epoch": 8.640776699029127,
+      "grad_norm": 0.6474212890207318,
+      "learning_rate": 1.9083573733734328e-06,
+      "loss": 0.1667,
+      "step": 890
+    },
+    {
+      "epoch": 8.650485436893204,
+      "grad_norm": 0.938194675661121,
+      "learning_rate": 1.903419521893369e-06,
+      "loss": 0.1331,
+      "step": 891
+    },
+    {
+      "epoch": 8.660194174757281,
+      "grad_norm": 0.7444633617064056,
+      "learning_rate": 1.898484137086794e-06,
+      "loss": 0.1258,
+      "step": 892
+    },
+    {
+      "epoch": 8.669902912621358,
+      "grad_norm": 0.7271455868832031,
+      "learning_rate": 1.8935512393599784e-06,
+      "loss": 0.0347,
+      "step": 893
+    },
+    {
+      "epoch": 8.679611650485437,
+      "grad_norm": 1.0122920027200322,
+      "learning_rate": 1.8886208491089095e-06,
+      "loss": 0.0897,
+      "step": 894
+    },
+    {
+      "epoch": 8.689320388349515,
+      "grad_norm": 0.610405675841603,
+      "learning_rate": 1.8836929867192077e-06,
+      "loss": 0.0614,
+      "step": 895
+    },
+    {
+      "epoch": 8.699029126213592,
+      "grad_norm": 0.7295850031852491,
+      "learning_rate": 1.8787676725660405e-06,
+      "loss": 0.0909,
+      "step": 896
+    },
+    {
+      "epoch": 8.70873786407767,
+      "grad_norm": 0.832667949302075,
+      "learning_rate": 1.8738449270140404e-06,
+      "loss": 0.2159,
+      "step": 897
+    },
+    {
+      "epoch": 8.718446601941748,
+      "grad_norm": 0.6861487310656976,
+      "learning_rate": 1.8689247704172187e-06,
+      "loss": 0.1567,
+      "step": 898
+    },
+    {
+      "epoch": 8.728155339805825,
+      "grad_norm": 0.7588351556503447,
+      "learning_rate": 1.8640072231188825e-06,
+      "loss": 0.1879,
+      "step": 899
+    },
+    {
+      "epoch": 8.737864077669903,
+      "grad_norm": 0.7201360467114077,
+      "learning_rate": 1.8590923054515504e-06,
+      "loss": 0.1627,
+      "step": 900
+    },
+    {
+      "epoch": 8.74757281553398,
+      "grad_norm": 0.7254000831754608,
+      "learning_rate": 1.8541800377368673e-06,
+      "loss": 0.1887,
+      "step": 901
+    },
+    {
+      "epoch": 8.757281553398059,
+      "grad_norm": 0.8622921617060856,
+      "learning_rate": 1.8492704402855229e-06,
+      "loss": 0.1916,
+      "step": 902
+    },
+    {
+      "epoch": 8.766990291262136,
+      "grad_norm": 0.693010331212834,
+      "learning_rate": 1.8443635333971643e-06,
+      "loss": 0.0764,
+      "step": 903
+    },
+    {
+      "epoch": 8.776699029126213,
+      "grad_norm": 0.9849662506213444,
+      "learning_rate": 1.8394593373603173e-06,
+      "loss": 0.1249,
+      "step": 904
+    },
+    {
+      "epoch": 8.78640776699029,
+      "grad_norm": 0.603276080839406,
+      "learning_rate": 1.8345578724522957e-06,
+      "loss": 0.1143,
+      "step": 905
+    },
+    {
+      "epoch": 8.79611650485437,
+      "grad_norm": 0.8844794167233913,
+      "learning_rate": 1.8296591589391227e-06,
+      "loss": 0.1003,
+      "step": 906
+    },
+    {
+      "epoch": 8.805825242718447,
+      "grad_norm": 0.6706572895742536,
+      "learning_rate": 1.8247632170754443e-06,
+      "loss": 0.1094,
+      "step": 907
+    },
+    {
+      "epoch": 8.815533980582524,
+      "grad_norm": 0.7048555162130203,
+      "learning_rate": 1.8198700671044477e-06,
+      "loss": 0.0615,
+      "step": 908
+    },
+    {
+      "epoch": 8.825242718446601,
+      "grad_norm": 0.7951541866582822,
+      "learning_rate": 1.8149797292577757e-06,
+      "loss": 0.1881,
+      "step": 909
+    },
+    {
+      "epoch": 8.83495145631068,
+      "grad_norm": 1.0779476607088172,
+      "learning_rate": 1.8100922237554442e-06,
+      "loss": 0.2077,
+      "step": 910
+    },
+    {
+      "epoch": 8.844660194174757,
+      "grad_norm": 0.5376687871944611,
+      "learning_rate": 1.8052075708057581e-06,
+      "loss": 0.0418,
+      "step": 911
+    },
+    {
+      "epoch": 8.854368932038835,
+      "grad_norm": 0.5817902698395092,
+      "learning_rate": 1.8003257906052284e-06,
+      "loss": 0.0272,
+      "step": 912
+    },
+    {
+      "epoch": 8.864077669902912,
+      "grad_norm": 0.9047000568519542,
+      "learning_rate": 1.7954469033384868e-06,
+      "loss": 0.1319,
+      "step": 913
+    },
+    {
+      "epoch": 8.87378640776699,
+      "grad_norm": 0.7285928029702656,
+      "learning_rate": 1.790570929178206e-06,
+      "loss": 0.1669,
+      "step": 914
+    },
+    {
+      "epoch": 8.883495145631068,
+      "grad_norm": 0.6615354057172695,
+      "learning_rate": 1.7856978882850112e-06,
+      "loss": 0.1078,
+      "step": 915
+    },
+    {
+      "epoch": 8.893203883495145,
+      "grad_norm": 0.684988992100525,
+      "learning_rate": 1.780827800807401e-06,
+      "loss": 0.1277,
+      "step": 916
+    },
+    {
+      "epoch": 8.902912621359224,
+      "grad_norm": 1.0329628068804662,
+      "learning_rate": 1.7759606868816623e-06,
+      "loss": 0.1635,
+      "step": 917
+    },
+    {
+      "epoch": 8.912621359223301,
+      "grad_norm": 0.519382288792405,
+      "learning_rate": 1.771096566631788e-06,
+      "loss": 0.0937,
+      "step": 918
+    },
+    {
+      "epoch": 8.922330097087379,
+      "grad_norm": 0.8421607606627751,
+      "learning_rate": 1.766235460169392e-06,
+      "loss": 0.1776,
+      "step": 919
+    },
+    {
+      "epoch": 8.932038834951456,
+      "grad_norm": 1.1862589863385231,
+      "learning_rate": 1.7613773875936274e-06,
+      "loss": 0.053,
+      "step": 920
+    },
+    {
+      "epoch": 8.941747572815533,
+      "grad_norm": 0.8864371486994366,
+      "learning_rate": 1.7565223689911038e-06,
+      "loss": 0.0515,
+      "step": 921
+    },
+    {
+      "epoch": 8.951456310679612,
+      "grad_norm": 0.5967041565663332,
+      "learning_rate": 1.7516704244358018e-06,
+      "loss": 0.1032,
+      "step": 922
+    },
+    {
+      "epoch": 8.96116504854369,
+      "grad_norm": 0.6128809380818757,
+      "learning_rate": 1.7468215739889941e-06,
+      "loss": 0.0798,
+      "step": 923
+    },
+    {
+      "epoch": 8.970873786407767,
+      "grad_norm": 0.9034318595552594,
+      "learning_rate": 1.741975837699158e-06,
+      "loss": 0.2227,
+      "step": 924
+    },
+    {
+      "epoch": 8.980582524271846,
+      "grad_norm": 0.5952541990579747,
+      "learning_rate": 1.7371332356018972e-06,
+      "loss": 0.1644,
+      "step": 925
+    },
+    {
+      "epoch": 8.990291262135923,
+      "grad_norm": 0.8057774859794358,
+      "learning_rate": 1.7322937877198545e-06,
+      "loss": 0.1789,
+      "step": 926
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.5973155402909465,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.0662,
+      "step": 927
+    },
+    {
+      "epoch": 9.009708737864077,
+      "grad_norm": 0.46982841443298656,
+      "learning_rate": 1.7226244346267063e-06,
+      "loss": 0.1126,
+      "step": 928
+    },
+    {
+      "epoch": 9.019417475728156,
+      "grad_norm": 0.6395071722994249,
+      "learning_rate": 1.7177945693953486e-06,
+      "loss": 0.1169,
+      "step": 929
+    },
+    {
+      "epoch": 9.029126213592233,
+      "grad_norm": 0.6000833565715967,
+      "learning_rate": 1.7129679383385384e-06,
+      "loss": 0.1407,
+      "step": 930
+    },
+    {
+      "epoch": 9.03883495145631,
+      "grad_norm": 0.5328547699863273,
+      "learning_rate": 1.7081445614128845e-06,
+      "loss": 0.0644,
+      "step": 931
+    },
+    {
+      "epoch": 9.048543689320388,
+      "grad_norm": 0.698674577006318,
+      "learning_rate": 1.7033244585615393e-06,
+      "loss": 0.2139,
+      "step": 932
+    },
+    {
+      "epoch": 9.058252427184467,
+      "grad_norm": 0.9070203866078601,
+      "learning_rate": 1.698507649714121e-06,
+      "loss": 0.1712,
+      "step": 933
+    },
+    {
+      "epoch": 9.067961165048544,
+      "grad_norm": 0.5121865723375197,
+      "learning_rate": 1.6936941547866248e-06,
+      "loss": 0.0602,
+      "step": 934
+    },
+    {
+      "epoch": 9.077669902912621,
+      "grad_norm": 0.6197111730604409,
+      "learning_rate": 1.688883993681345e-06,
+      "loss": 0.1063,
+      "step": 935
+    },
+    {
+      "epoch": 9.087378640776699,
+      "grad_norm": 0.7859446210738993,
+      "learning_rate": 1.6840771862867922e-06,
+      "loss": 0.0857,
+      "step": 936
+    },
+    {
+      "epoch": 9.097087378640778,
+      "grad_norm": 1.1642241783373757,
+      "learning_rate": 1.6792737524776093e-06,
+      "loss": 0.1344,
+      "step": 937
+    },
+    {
+      "epoch": 9.106796116504855,
+      "grad_norm": 0.4302682219192299,
+      "learning_rate": 1.674473712114492e-06,
+      "loss": 0.0379,
+      "step": 938
+    },
+    {
+      "epoch": 9.116504854368932,
+      "grad_norm": 0.621250635600905,
+      "learning_rate": 1.6696770850441036e-06,
+      "loss": 0.0401,
+      "step": 939
+    },
+    {
+      "epoch": 9.12621359223301,
+      "grad_norm": 0.9740610618310332,
+      "learning_rate": 1.6648838910989955e-06,
+      "loss": 0.2171,
+      "step": 940
+    },
+    {
+      "epoch": 9.135922330097088,
+      "grad_norm": 0.8966783169812266,
+      "learning_rate": 1.6600941500975237e-06,
+      "loss": 0.0965,
+      "step": 941
+    },
+    {
+      "epoch": 9.145631067961165,
+      "grad_norm": 0.7400396070349328,
+      "learning_rate": 1.6553078818437678e-06,
+      "loss": 0.1748,
+      "step": 942
+    },
+    {
+      "epoch": 9.155339805825243,
+      "grad_norm": 0.8321559134472875,
+      "learning_rate": 1.6505251061274492e-06,
+      "loss": 0.1614,
+      "step": 943
+    },
+    {
+      "epoch": 9.16504854368932,
+      "grad_norm": 0.8790366855460469,
+      "learning_rate": 1.6457458427238464e-06,
+      "loss": 0.2559,
+      "step": 944
+    },
+    {
+      "epoch": 9.174757281553399,
+      "grad_norm": 0.8881109198480825,
+      "learning_rate": 1.6409701113937182e-06,
+      "loss": 0.0951,
+      "step": 945
+    },
+    {
+      "epoch": 9.184466019417476,
+      "grad_norm": 0.8440100111279388,
+      "learning_rate": 1.6361979318832173e-06,
+      "loss": 0.0804,
+      "step": 946
+    },
+    {
+      "epoch": 9.194174757281553,
+      "grad_norm": 0.6968010735521492,
+      "learning_rate": 1.6314293239238134e-06,
+      "loss": 0.1062,
+      "step": 947
+    },
+    {
+      "epoch": 9.20388349514563,
+      "grad_norm": 0.5054689570887443,
+      "learning_rate": 1.626664307232207e-06,
+      "loss": 0.0801,
+      "step": 948
+    },
+    {
+      "epoch": 9.21359223300971,
+      "grad_norm": 0.7768640631942998,
+      "learning_rate": 1.62190290151025e-06,
+      "loss": 0.1664,
+      "step": 949
+    },
+    {
+      "epoch": 9.223300970873787,
+      "grad_norm": 0.5024211551473686,
+      "learning_rate": 1.617145126444864e-06,
+      "loss": 0.0903,
+      "step": 950
+    },
+    {
+      "epoch": 9.233009708737864,
+      "grad_norm": 0.6883186340673217,
+      "learning_rate": 1.6123910017079591e-06,
+      "loss": 0.1663,
+      "step": 951
+    },
+    {
+      "epoch": 9.242718446601941,
+      "grad_norm": 0.4722355032472164,
+      "learning_rate": 1.6076405469563533e-06,
+      "loss": 0.0675,
+      "step": 952
+    },
+    {
+      "epoch": 9.25242718446602,
+      "grad_norm": 0.6836772775774017,
+      "learning_rate": 1.6028937818316889e-06,
+      "loss": 0.1429,
+      "step": 953
+    },
+    {
+      "epoch": 9.262135922330097,
+      "grad_norm": 0.5257068429565035,
+      "learning_rate": 1.598150725960354e-06,
+      "loss": 0.0209,
+      "step": 954
+    },
+    {
+      "epoch": 9.271844660194175,
+      "grad_norm": 0.7877874466415006,
+      "learning_rate": 1.5934113989533992e-06,
+      "loss": 0.0966,
+      "step": 955
+    },
+    {
+      "epoch": 9.281553398058252,
+      "grad_norm": 1.044912299647016,
+      "learning_rate": 1.5886758204064582e-06,
+      "loss": 0.0712,
+      "step": 956
+    },
+    {
+      "epoch": 9.29126213592233,
+      "grad_norm": 0.33787344891092924,
+      "learning_rate": 1.583944009899665e-06,
+      "loss": 0.0273,
+      "step": 957
+    },
+    {
+      "epoch": 9.300970873786408,
+      "grad_norm": 0.7332496066271933,
+      "learning_rate": 1.579215986997575e-06,
+      "loss": 0.0734,
+      "step": 958
+    },
+    {
+      "epoch": 9.310679611650485,
+      "grad_norm": 0.9742544872879907,
+      "learning_rate": 1.5744917712490821e-06,
+      "loss": 0.1911,
+      "step": 959
+    },
+    {
+      "epoch": 9.320388349514563,
+      "grad_norm": 0.421873246338127,
+      "learning_rate": 1.5697713821873401e-06,
+      "loss": 0.0244,
+      "step": 960
+    },
+    {
+      "epoch": 9.330097087378642,
+      "grad_norm": 0.5145290760788898,
+      "learning_rate": 1.5650548393296788e-06,
+      "loss": 0.0457,
+      "step": 961
+    },
+    {
+      "epoch": 9.339805825242719,
+      "grad_norm": 0.7881806168450413,
+      "learning_rate": 1.5603421621775273e-06,
+      "loss": 0.1336,
+      "step": 962
+    },
+    {
+      "epoch": 9.349514563106796,
+      "grad_norm": 0.7928543632419307,
+      "learning_rate": 1.555633370216329e-06,
+      "loss": 0.0845,
+      "step": 963
+    },
+    {
+      "epoch": 9.359223300970873,
+      "grad_norm": 0.5771454392471547,
+      "learning_rate": 1.5509284829154652e-06,
+      "loss": 0.1106,
+      "step": 964
+    },
+    {
+      "epoch": 9.368932038834952,
+      "grad_norm": 0.6558019594502399,
+      "learning_rate": 1.5462275197281717e-06,
+      "loss": 0.0633,
+      "step": 965
+    },
+    {
+      "epoch": 9.37864077669903,
+      "grad_norm": 1.1584092476213597,
+      "learning_rate": 1.5415305000914587e-06,
+      "loss": 0.1725,
+      "step": 966
+    },
+    {
+      "epoch": 9.388349514563107,
+      "grad_norm": 0.5452649508861512,
+      "learning_rate": 1.536837443426032e-06,
+      "loss": 0.0297,
+      "step": 967
+    },
+    {
+      "epoch": 9.398058252427184,
+      "grad_norm": 0.4314536564363454,
+      "learning_rate": 1.5321483691362121e-06,
+      "loss": 0.0797,
+      "step": 968
+    },
+    {
+      "epoch": 9.407766990291263,
+      "grad_norm": 0.6324718727678328,
+      "learning_rate": 1.5274632966098527e-06,
+      "loss": 0.0332,
+      "step": 969
+    },
+    {
+      "epoch": 9.41747572815534,
+      "grad_norm": 0.4549599787302605,
+      "learning_rate": 1.5227822452182617e-06,
+      "loss": 0.0556,
+      "step": 970
+    },
+    {
+      "epoch": 9.427184466019417,
+      "grad_norm": 0.7287131128875295,
+      "learning_rate": 1.5181052343161212e-06,
+      "loss": 0.1135,
+      "step": 971
+    },
+    {
+      "epoch": 9.436893203883495,
+      "grad_norm": 0.3758449427962483,
+      "learning_rate": 1.5134322832414066e-06,
+      "loss": 0.0328,
+      "step": 972
+    },
+    {
+      "epoch": 9.446601941747574,
+      "grad_norm": 0.629439947427426,
+      "learning_rate": 1.508763411315308e-06,
+      "loss": 0.1158,
+      "step": 973
+    },
+    {
+      "epoch": 9.45631067961165,
+      "grad_norm": 0.5492590601077479,
+      "learning_rate": 1.5040986378421485e-06,
+      "loss": 0.0621,
+      "step": 974
+    },
+    {
+      "epoch": 9.466019417475728,
+      "grad_norm": 0.47227042867230984,
+      "learning_rate": 1.499437982109305e-06,
+      "loss": 0.0823,
+      "step": 975
+    },
+    {
+      "epoch": 9.475728155339805,
+      "grad_norm": 0.8874051976671528,
+      "learning_rate": 1.4947814633871316e-06,
+      "loss": 0.0789,
+      "step": 976
+    },
+    {
+      "epoch": 9.485436893203884,
+      "grad_norm": 0.5313571637739412,
+      "learning_rate": 1.4901291009288748e-06,
+      "loss": 0.048,
+      "step": 977
+    },
+    {
+      "epoch": 9.495145631067961,
+      "grad_norm": 0.6634950979669227,
+      "learning_rate": 1.4854809139705961e-06,
+      "loss": 0.129,
+      "step": 978
+    },
+    {
+      "epoch": 9.504854368932039,
+      "grad_norm": 0.7211690230550263,
+      "learning_rate": 1.4808369217310937e-06,
+      "loss": 0.1095,
+      "step": 979
+    },
+    {
+      "epoch": 9.514563106796116,
+      "grad_norm": 0.8721245712399799,
+      "learning_rate": 1.4761971434118207e-06,
+      "loss": 0.1667,
+      "step": 980
+    },
+    {
+      "epoch": 9.524271844660195,
+      "grad_norm": 0.7538810617432105,
+      "learning_rate": 1.4715615981968088e-06,
+      "loss": 0.082,
+      "step": 981
+    },
+    {
+      "epoch": 9.533980582524272,
+      "grad_norm": 0.6613308701670603,
+      "learning_rate": 1.4669303052525852e-06,
+      "loss": 0.12,
+      "step": 982
+    },
+    {
+      "epoch": 9.54368932038835,
+      "grad_norm": 0.40582479242542924,
+      "learning_rate": 1.4623032837280971e-06,
+      "loss": 0.012,
+      "step": 983
+    },
+    {
+      "epoch": 9.553398058252426,
+      "grad_norm": 0.9306885236611109,
+      "learning_rate": 1.4576805527546293e-06,
+      "loss": 0.1777,
+      "step": 984
+    },
+    {
+      "epoch": 9.563106796116505,
+      "grad_norm": 1.0023614165143215,
+      "learning_rate": 1.4530621314457255e-06,
+      "loss": 0.0978,
+      "step": 985
+    },
+    {
+      "epoch": 9.572815533980583,
+      "grad_norm": 0.9617385737325488,
+      "learning_rate": 1.4484480388971141e-06,
+      "loss": 0.2374,
+      "step": 986
+    },
+    {
+      "epoch": 9.58252427184466,
+      "grad_norm": 0.9212729138833889,
+      "learning_rate": 1.4438382941866224e-06,
+      "loss": 0.1527,
+      "step": 987
+    },
+    {
+      "epoch": 9.592233009708737,
+      "grad_norm": 0.6306564260185002,
+      "learning_rate": 1.4392329163741015e-06,
+      "loss": 0.1009,
+      "step": 988
+    },
+    {
+      "epoch": 9.601941747572816,
+      "grad_norm": 0.5734759604323216,
+      "learning_rate": 1.4346319245013463e-06,
+      "loss": 0.1903,
+      "step": 989
+    },
+    {
+      "epoch": 9.611650485436893,
+      "grad_norm": 0.9291804725028308,
+      "learning_rate": 1.430035337592018e-06,
+      "loss": 0.0629,
+      "step": 990
+    },
+    {
+      "epoch": 9.62135922330097,
+      "grad_norm": 0.5720126721481944,
+      "learning_rate": 1.425443174651564e-06,
+      "loss": 0.0699,
+      "step": 991
+    },
+    {
+      "epoch": 9.631067961165048,
+      "grad_norm": 0.6827408883331223,
+      "learning_rate": 1.4208554546671407e-06,
+      "loss": 0.1166,
+      "step": 992
+    },
+    {
+      "epoch": 9.640776699029127,
+      "grad_norm": 0.9468184547358578,
+      "learning_rate": 1.4162721966075323e-06,
+      "loss": 0.1508,
+      "step": 993
+    },
+    {
+      "epoch": 9.650485436893204,
+      "grad_norm": 0.5185651892093444,
+      "learning_rate": 1.411693419423078e-06,
+      "loss": 0.0747,
+      "step": 994
+    },
+    {
+      "epoch": 9.660194174757281,
+      "grad_norm": 0.7182852834221573,
+      "learning_rate": 1.4071191420455873e-06,
+      "loss": 0.0868,
+      "step": 995
+    },
+    {
+      "epoch": 9.669902912621358,
+      "grad_norm": 0.5124393893502427,
+      "learning_rate": 1.4025493833882645e-06,
+      "loss": 0.0544,
+      "step": 996
+    },
+    {
+      "epoch": 9.679611650485437,
+      "grad_norm": 0.4519760606619813,
+      "learning_rate": 1.3979841623456309e-06,
+      "loss": 0.0298,
+      "step": 997
+    },
+    {
+      "epoch": 9.689320388349515,
+      "grad_norm": 0.7723673765353795,
+      "learning_rate": 1.3934234977934463e-06,
+      "loss": 0.1117,
+      "step": 998
+    },
+    {
+      "epoch": 9.699029126213592,
+      "grad_norm": 0.5591540435943934,
+      "learning_rate": 1.3888674085886302e-06,
+      "loss": 0.0856,
+      "step": 999
+    },
+    {
+      "epoch": 9.70873786407767,
+      "grad_norm": 0.5572779776561083,
+      "learning_rate": 1.3843159135691859e-06,
+      "loss": 0.1138,
+      "step": 1000
+    },
+    {
+      "epoch": 9.718446601941748,
+      "grad_norm": 0.579120584588862,
+      "learning_rate": 1.3797690315541193e-06,
+      "loss": 0.091,
+      "step": 1001
+    },
+    {
+      "epoch": 9.728155339805825,
+      "grad_norm": 0.7351017325830426,
+      "learning_rate": 1.3752267813433645e-06,
+      "loss": 0.1019,
+      "step": 1002
+    },
+    {
+      "epoch": 9.737864077669903,
+      "grad_norm": 0.44019226102987263,
+      "learning_rate": 1.3706891817177036e-06,
+      "loss": 0.0632,
+      "step": 1003
+    },
+    {
+      "epoch": 9.74757281553398,
+      "grad_norm": 0.6462349810299945,
+      "learning_rate": 1.3661562514386895e-06,
+      "loss": 0.1481,
+      "step": 1004
+    },
+    {
+      "epoch": 9.757281553398059,
+      "grad_norm": 0.8965376719847379,
+      "learning_rate": 1.3616280092485719e-06,
+      "loss": 0.1183,
+      "step": 1005
+    },
+    {
+      "epoch": 9.766990291262136,
+      "grad_norm": 0.7108013520860924,
+      "learning_rate": 1.357104473870213e-06,
+      "loss": 0.168,
+      "step": 1006
+    },
+    {
+      "epoch": 9.776699029126213,
+      "grad_norm": 0.6653262047892645,
+      "learning_rate": 1.3525856640070156e-06,
+      "loss": 0.0664,
+      "step": 1007
+    },
+    {
+      "epoch": 9.78640776699029,
+      "grad_norm": 0.5630299546987264,
+      "learning_rate": 1.3480715983428433e-06,
+      "loss": 0.07,
+      "step": 1008
+    },
+    {
+      "epoch": 9.79611650485437,
+      "grad_norm": 0.5795784482057218,
+      "learning_rate": 1.3435622955419447e-06,
+      "loss": 0.1229,
+      "step": 1009
+    },
+    {
+      "epoch": 9.805825242718447,
+      "grad_norm": 0.45648098304255846,
+      "learning_rate": 1.3390577742488747e-06,
+      "loss": 0.0891,
+      "step": 1010
+    },
+    {
+      "epoch": 9.815533980582524,
+      "grad_norm": 0.5330321441449665,
+      "learning_rate": 1.334558053088419e-06,
+      "loss": 0.0854,
+      "step": 1011
+    },
+    {
+      "epoch": 9.825242718446601,
+      "grad_norm": 0.8383784644182197,
+      "learning_rate": 1.3300631506655148e-06,
+      "loss": 0.1271,
+      "step": 1012
+    },
+    {
+      "epoch": 9.83495145631068,
+      "grad_norm": 0.7994557312933233,
+      "learning_rate": 1.3255730855651772e-06,
+      "loss": 0.154,
+      "step": 1013
+    },
+    {
+      "epoch": 9.844660194174757,
+      "grad_norm": 0.6121104973654957,
+      "learning_rate": 1.3210878763524186e-06,
+      "loss": 0.1244,
+      "step": 1014
+    },
+    {
+      "epoch": 9.854368932038835,
+      "grad_norm": 0.8704001110194098,
+      "learning_rate": 1.3166075415721762e-06,
+      "loss": 0.1757,
+      "step": 1015
+    },
+    {
+      "epoch": 9.864077669902912,
+      "grad_norm": 0.5125263843444212,
+      "learning_rate": 1.3121320997492305e-06,
+      "loss": 0.0905,
+      "step": 1016
+    },
+    {
+      "epoch": 9.87378640776699,
+      "grad_norm": 0.6892296957249096,
+      "learning_rate": 1.307661569388132e-06,
+      "loss": 0.0987,
+      "step": 1017
+    },
+    {
+      "epoch": 9.883495145631068,
+      "grad_norm": 0.7269487551833638,
+      "learning_rate": 1.3031959689731236e-06,
+      "loss": 0.0787,
+      "step": 1018
+    },
+    {
+      "epoch": 9.893203883495145,
+      "grad_norm": 0.6176952721891387,
+      "learning_rate": 1.2987353169680667e-06,
+      "loss": 0.1318,
+      "step": 1019
+    },
+    {
+      "epoch": 9.902912621359224,
+      "grad_norm": 0.6021673456770801,
+      "learning_rate": 1.2942796318163595e-06,
+      "loss": 0.0492,
+      "step": 1020
+    },
+    {
+      "epoch": 9.912621359223301,
+      "grad_norm": 0.6496396475871585,
+      "learning_rate": 1.2898289319408653e-06,
+      "loss": 0.0956,
+      "step": 1021
+    },
+    {
+      "epoch": 9.922330097087379,
+      "grad_norm": 0.9117599759177222,
+      "learning_rate": 1.2853832357438346e-06,
+      "loss": 0.0815,
+      "step": 1022
+    },
+    {
+      "epoch": 9.932038834951456,
+      "grad_norm": 0.6343515859693896,
+      "learning_rate": 1.2809425616068288e-06,
+      "loss": 0.1379,
+      "step": 1023
+    },
+    {
+      "epoch": 9.941747572815533,
+      "grad_norm": 0.8668978005843916,
+      "learning_rate": 1.2765069278906456e-06,
+      "loss": 0.1431,
+      "step": 1024
+    },
+    {
+      "epoch": 9.951456310679612,
+      "grad_norm": 0.5567206920970875,
+      "learning_rate": 1.2720763529352415e-06,
+      "loss": 0.0251,
+      "step": 1025
+    },
+    {
+      "epoch": 9.96116504854369,
+      "grad_norm": 0.7044628563483055,
+      "learning_rate": 1.2676508550596562e-06,
+      "loss": 0.0398,
+      "step": 1026
+    },
+    {
+      "epoch": 9.970873786407767,
+      "grad_norm": 0.7421745735260384,
+      "learning_rate": 1.2632304525619388e-06,
+      "loss": 0.0626,
+      "step": 1027
+    },
+    {
+      "epoch": 9.980582524271846,
+      "grad_norm": 1.1608591744618904,
+      "learning_rate": 1.2588151637190687e-06,
+      "loss": 0.2441,
+      "step": 1028
+    },
+    {
+      "epoch": 9.990291262135923,
+      "grad_norm": 0.5136678107573567,
+      "learning_rate": 1.2544050067868834e-06,
+      "loss": 0.0588,
+      "step": 1029
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.44097012341303254,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.098,
+      "step": 1030
+    },
+    {
+      "epoch": 10.009708737864077,
+      "grad_norm": 0.7139105592958839,
+      "learning_rate": 1.2456001615717445e-06,
+      "loss": 0.0817,
+      "step": 1031
+    },
+    {
+      "epoch": 10.019417475728156,
+      "grad_norm": 0.5817012295383788,
+      "learning_rate": 1.2412055096940692e-06,
+      "loss": 0.0855,
+      "step": 1032
+    },
+    {
+      "epoch": 10.029126213592233,
+      "grad_norm": 0.9943377613107669,
+      "learning_rate": 1.2368160625374835e-06,
+      "loss": 0.139,
+      "step": 1033
+    },
+    {
+      "epoch": 10.03883495145631,
+      "grad_norm": 0.3686137208464406,
+      "learning_rate": 1.2324318382509787e-06,
+      "loss": 0.0559,
+      "step": 1034
+    },
+    {
+      "epoch": 10.048543689320388,
+      "grad_norm": 0.38577477799819054,
+      "learning_rate": 1.2280528549619487e-06,
+      "loss": 0.0288,
+      "step": 1035
+    },
+    {
+      "epoch": 10.058252427184467,
+      "grad_norm": 0.6582129917764945,
+      "learning_rate": 1.2236791307761184e-06,
+      "loss": 0.1647,
+      "step": 1036
+    },
+    {
+      "epoch": 10.067961165048544,
+      "grad_norm": 0.7512464872879958,
+      "learning_rate": 1.2193106837774678e-06,
+      "loss": 0.0545,
+      "step": 1037
+    },
+    {
+      "epoch": 10.077669902912621,
+      "grad_norm": 0.5425451526026376,
+      "learning_rate": 1.2149475320281578e-06,
+      "loss": 0.0398,
+      "step": 1038
+    },
+    {
+      "epoch": 10.087378640776699,
+      "grad_norm": 0.7960650124670342,
+      "learning_rate": 1.2105896935684545e-06,
+      "loss": 0.082,
+      "step": 1039
+    },
+    {
+      "epoch": 10.097087378640778,
+      "grad_norm": 0.5338834907010502,
+      "learning_rate": 1.2062371864166553e-06,
+      "loss": 0.082,
+      "step": 1040
+    },
+    {
+      "epoch": 10.106796116504855,
+      "grad_norm": 0.5672377297057393,
+      "learning_rate": 1.2018900285690148e-06,
+      "loss": 0.0588,
+      "step": 1041
+    },
+    {
+      "epoch": 10.116504854368932,
+      "grad_norm": 0.5383809127022262,
+      "learning_rate": 1.1975482379996697e-06,
+      "loss": 0.0963,
+      "step": 1042
+    },
+    {
+      "epoch": 10.12621359223301,
+      "grad_norm": 0.554544672201788,
+      "learning_rate": 1.1932118326605644e-06,
+      "loss": 0.0802,
+      "step": 1043
+    },
+    {
+      "epoch": 10.135922330097088,
+      "grad_norm": 0.40501491454406263,
+      "learning_rate": 1.188880830481377e-06,
+      "loss": 0.0155,
+      "step": 1044
+    },
+    {
+      "epoch": 10.145631067961165,
+      "grad_norm": 0.4982365542620892,
+      "learning_rate": 1.1845552493694462e-06,
+      "loss": 0.0624,
+      "step": 1045
+    },
+    {
+      "epoch": 10.155339805825243,
+      "grad_norm": 0.44284165252569085,
+      "learning_rate": 1.1802351072096948e-06,
+      "loss": 0.0389,
+      "step": 1046
+    },
+    {
+      "epoch": 10.16504854368932,
+      "grad_norm": 0.43688194002614283,
+      "learning_rate": 1.1759204218645577e-06,
+      "loss": 0.0331,
+      "step": 1047
+    },
+    {
+      "epoch": 10.174757281553399,
+      "grad_norm": 0.9610954492065917,
+      "learning_rate": 1.1716112111739095e-06,
+      "loss": 0.057,
+      "step": 1048
+    },
+    {
+      "epoch": 10.184466019417476,
+      "grad_norm": 0.7894221279495568,
+      "learning_rate": 1.167307492954986e-06,
+      "loss": 0.1321,
+      "step": 1049
+    },
+    {
+      "epoch": 10.194174757281553,
+      "grad_norm": 0.7105486043207161,
+      "learning_rate": 1.1630092850023148e-06,
+      "loss": 0.117,
+      "step": 1050
+    },
+    {
+      "epoch": 10.20388349514563,
+      "grad_norm": 0.8709137289348258,
+      "learning_rate": 1.15871660508764e-06,
+      "loss": 0.1173,
+      "step": 1051
+    },
+    {
+      "epoch": 10.21359223300971,
+      "grad_norm": 0.7568948585913298,
+      "learning_rate": 1.1544294709598491e-06,
+      "loss": 0.1474,
+      "step": 1052
+    },
+    {
+      "epoch": 10.223300970873787,
+      "grad_norm": 0.6724147094520684,
+      "learning_rate": 1.1501479003448992e-06,
+      "loss": 0.0697,
+      "step": 1053
+    },
+    {
+      "epoch": 10.233009708737864,
+      "grad_norm": 0.4949654633371321,
+      "learning_rate": 1.1458719109457445e-06,
+      "loss": 0.0578,
+      "step": 1054
+    },
+    {
+      "epoch": 10.242718446601941,
+      "grad_norm": 0.8068088269521827,
+      "learning_rate": 1.141601520442262e-06,
+      "loss": 0.1246,
+      "step": 1055
+    },
+    {
+      "epoch": 10.25242718446602,
+      "grad_norm": 0.43316042424228557,
+      "learning_rate": 1.1373367464911798e-06,
+      "loss": 0.0454,
+      "step": 1056
+    },
+    {
+      "epoch": 10.262135922330097,
+      "grad_norm": 0.6652579848799286,
+      "learning_rate": 1.1330776067260026e-06,
+      "loss": 0.1283,
+      "step": 1057
+    },
+    {
+      "epoch": 10.271844660194175,
+      "grad_norm": 0.4042774504841769,
+      "learning_rate": 1.12882411875694e-06,
+      "loss": 0.0252,
+      "step": 1058
+    },
+    {
+      "epoch": 10.281553398058252,
+      "grad_norm": 0.5385885477070604,
+      "learning_rate": 1.1245763001708326e-06,
+      "loss": 0.0388,
+      "step": 1059
+    },
+    {
+      "epoch": 10.29126213592233,
+      "grad_norm": 0.6052457558453441,
+      "learning_rate": 1.120334168531081e-06,
+      "loss": 0.0734,
+      "step": 1060
+    },
+    {
+      "epoch": 10.300970873786408,
+      "grad_norm": 0.5293111516776242,
+      "learning_rate": 1.1160977413775704e-06,
+      "loss": 0.1185,
+      "step": 1061
+    },
+    {
+      "epoch": 10.310679611650485,
+      "grad_norm": 0.43792304060112175,
+      "learning_rate": 1.1118670362266003e-06,
+      "loss": 0.0504,
+      "step": 1062
+    },
+    {
+      "epoch": 10.320388349514563,
+      "grad_norm": 0.5208009686668422,
+      "learning_rate": 1.1076420705708137e-06,
+      "loss": 0.049,
+      "step": 1063
+    },
+    {
+      "epoch": 10.330097087378642,
+      "grad_norm": 0.32151178708408873,
+      "learning_rate": 1.1034228618791197e-06,
+      "loss": 0.0048,
+      "step": 1064
+    },
+    {
+      "epoch": 10.339805825242719,
+      "grad_norm": 0.40783412915581874,
+      "learning_rate": 1.0992094275966256e-06,
+      "loss": 0.0396,
+      "step": 1065
+    },
+    {
+      "epoch": 10.349514563106796,
+      "grad_norm": 0.8301501331889736,
+      "learning_rate": 1.0950017851445624e-06,
+      "loss": 0.0771,
+      "step": 1066
+    },
+    {
+      "epoch": 10.359223300970873,
+      "grad_norm": 0.656158454275523,
+      "learning_rate": 1.0907999519202142e-06,
+      "loss": 0.1427,
+      "step": 1067
+    },
+    {
+      "epoch": 10.368932038834952,
+      "grad_norm": 0.6226103543328013,
+      "learning_rate": 1.0866039452968464e-06,
+      "loss": 0.0849,
+      "step": 1068
+    },
+    {
+      "epoch": 10.37864077669903,
+      "grad_norm": 0.30026497224539017,
+      "learning_rate": 1.0824137826236318e-06,
+      "loss": 0.0053,
+      "step": 1069
+    },
+    {
+      "epoch": 10.388349514563107,
+      "grad_norm": 0.33866413080844127,
+      "learning_rate": 1.078229481225582e-06,
+      "loss": 0.0183,
+      "step": 1070
+    },
+    {
+      "epoch": 10.398058252427184,
+      "grad_norm": 0.5566432505296565,
+      "learning_rate": 1.074051058403472e-06,
+      "loss": 0.0759,
+      "step": 1071
+    },
+    {
+      "epoch": 10.407766990291263,
+      "grad_norm": 0.49851774518490277,
+      "learning_rate": 1.069878531433773e-06,
+      "loss": 0.0528,
+      "step": 1072
+    },
+    {
+      "epoch": 10.41747572815534,
+      "grad_norm": 0.4860505947440488,
+      "learning_rate": 1.0657119175685776e-06,
+      "loss": 0.0484,
+      "step": 1073
+    },
+    {
+      "epoch": 10.427184466019417,
+      "grad_norm": 0.3145673776783807,
+      "learning_rate": 1.061551234035529e-06,
+      "loss": 0.0153,
+      "step": 1074
+    },
+    {
+      "epoch": 10.436893203883495,
+      "grad_norm": 0.6816619685019172,
+      "learning_rate": 1.0573964980377517e-06,
+      "loss": 0.1271,
+      "step": 1075
+    },
+    {
+      "epoch": 10.446601941747574,
+      "grad_norm": 0.5845770900591787,
+      "learning_rate": 1.0532477267537772e-06,
+      "loss": 0.1469,
+      "step": 1076
+    },
+    {
+      "epoch": 10.45631067961165,
+      "grad_norm": 0.7233633949979615,
+      "learning_rate": 1.0491049373374762e-06,
+      "loss": 0.2243,
+      "step": 1077
+    },
+    {
+      "epoch": 10.466019417475728,
+      "grad_norm": 0.5464884216114049,
+      "learning_rate": 1.044968146917986e-06,
+      "loss": 0.0812,
+      "step": 1078
+    },
+    {
+      "epoch": 10.475728155339805,
+      "grad_norm": 0.3391211976990657,
+      "learning_rate": 1.0408373725996386e-06,
+      "loss": 0.0188,
+      "step": 1079
+    },
+    {
+      "epoch": 10.485436893203884,
+      "grad_norm": 0.6029803423460801,
+      "learning_rate": 1.0367126314618946e-06,
+      "loss": 0.0608,
+      "step": 1080
+    },
+    {
+      "epoch": 10.495145631067961,
+      "grad_norm": 0.8359118873064847,
+      "learning_rate": 1.0325939405592661e-06,
+      "loss": 0.1468,
+      "step": 1081
+    },
+    {
+      "epoch": 10.504854368932039,
+      "grad_norm": 0.4613807440413381,
+      "learning_rate": 1.0284813169212502e-06,
+      "loss": 0.0419,
+      "step": 1082
+    },
+    {
+      "epoch": 10.514563106796116,
+      "grad_norm": 0.7117820265098361,
+      "learning_rate": 1.024374777552258e-06,
+      "loss": 0.0602,
+      "step": 1083
+    },
+    {
+      "epoch": 10.524271844660195,
+      "grad_norm": 0.673361722656393,
+      "learning_rate": 1.0202743394315444e-06,
+      "loss": 0.0572,
+      "step": 1084
+    },
+    {
+      "epoch": 10.533980582524272,
+      "grad_norm": 0.718842425297297,
+      "learning_rate": 1.0161800195131372e-06,
+      "loss": 0.0709,
+      "step": 1085
+    },
+    {
+      "epoch": 10.54368932038835,
+      "grad_norm": 0.519489444086323,
+      "learning_rate": 1.0120918347257669e-06,
+      "loss": 0.0652,
+      "step": 1086
+    },
+    {
+      "epoch": 10.553398058252426,
+      "grad_norm": 0.531214341067294,
+      "learning_rate": 1.0080098019727979e-06,
+      "loss": 0.0405,
+      "step": 1087
+    },
+    {
+      "epoch": 10.563106796116505,
+      "grad_norm": 0.8237012418593747,
+      "learning_rate": 1.0039339381321572e-06,
+      "loss": 0.1696,
+      "step": 1088
+    },
+    {
+      "epoch": 10.572815533980583,
+      "grad_norm": 0.6702768856937782,
+      "learning_rate": 9.998642600562664e-07,
+      "loss": 0.1737,
+      "step": 1089
+    },
+    {
+      "epoch": 10.58252427184466,
+      "grad_norm": 0.6967567433213261,
+      "learning_rate": 9.95800784571969e-07,
+      "loss": 0.1259,
+      "step": 1090
+    },
+    {
+      "epoch": 10.592233009708737,
+      "grad_norm": 0.7311688346400493,
+      "learning_rate": 9.91743528480464e-07,
+      "loss": 0.1139,
+      "step": 1091
+    },
+    {
+      "epoch": 10.601941747572816,
+      "grad_norm": 0.8124930399760492,
+      "learning_rate": 9.876925085572365e-07,
+      "loss": 0.0526,
+      "step": 1092
+    },
+    {
+      "epoch": 10.611650485436893,
+      "grad_norm": 0.6821859039678345,
+      "learning_rate": 9.836477415519843e-07,
+      "loss": 0.0798,
+      "step": 1093
+    },
+    {
+      "epoch": 10.62135922330097,
+      "grad_norm": 0.5388868436932894,
+      "learning_rate": 9.79609244188553e-07,
+      "loss": 0.0595,
+      "step": 1094
+    },
+    {
+      "epoch": 10.631067961165048,
+      "grad_norm": 0.46754323131853315,
+      "learning_rate": 9.755770331648642e-07,
+      "loss": 0.0724,
+      "step": 1095
+    },
+    {
+      "epoch": 10.640776699029127,
+      "grad_norm": 0.6883957401661763,
+      "learning_rate": 9.715511251528486e-07,
+      "loss": 0.1118,
+      "step": 1096
+    },
+    {
+      "epoch": 10.650485436893204,
+      "grad_norm": 0.6882504391474683,
+      "learning_rate": 9.67531536798375e-07,
+      "loss": 0.1209,
+      "step": 1097
+    },
+    {
+      "epoch": 10.660194174757281,
+      "grad_norm": 0.5983944209437926,
+      "learning_rate": 9.635182847211827e-07,
+      "loss": 0.1064,
+      "step": 1098
+    },
+    {
+      "epoch": 10.669902912621358,
+      "grad_norm": 0.6811521493632275,
+      "learning_rate": 9.595113855148128e-07,
+      "loss": 0.0937,
+      "step": 1099
+    },
+    {
+      "epoch": 10.679611650485437,
+      "grad_norm": 0.6779393838141433,
+      "learning_rate": 9.555108557465383e-07,
+      "loss": 0.1937,
+      "step": 1100
+    },
+    {
+      "epoch": 10.689320388349515,
+      "grad_norm": 0.6498698531590674,
+      "learning_rate": 9.51516711957298e-07,
+      "loss": 0.1226,
+      "step": 1101
+    },
+    {
+      "epoch": 10.699029126213592,
+      "grad_norm": 0.530378508456918,
+      "learning_rate": 9.475289706616256e-07,
+      "loss": 0.0828,
+      "step": 1102
+    },
+    {
+      "epoch": 10.70873786407767,
+      "grad_norm": 0.7058714411604254,
+      "learning_rate": 9.435476483475825e-07,
+      "loss": 0.0793,
+      "step": 1103
+    },
+    {
+      "epoch": 10.718446601941748,
+      "grad_norm": 0.45348419126683187,
+      "learning_rate": 9.395727614766903e-07,
+      "loss": 0.0799,
+      "step": 1104
+    },
+    {
+      "epoch": 10.728155339805825,
+      "grad_norm": 0.6579676572313675,
+      "learning_rate": 9.356043264838607e-07,
+      "loss": 0.0686,
+      "step": 1105
+    },
+    {
+      "epoch": 10.737864077669903,
+      "grad_norm": 0.5367393390599688,
+      "learning_rate": 9.316423597773316e-07,
+      "loss": 0.0398,
+      "step": 1106
+    },
+    {
+      "epoch": 10.74757281553398,
+      "grad_norm": 0.6485846910490585,
+      "learning_rate": 9.276868777385942e-07,
+      "loss": 0.1605,
+      "step": 1107
+    },
+    {
+      "epoch": 10.757281553398059,
+      "grad_norm": 0.4379559794587696,
+      "learning_rate": 9.237378967223279e-07,
+      "loss": 0.0705,
+      "step": 1108
+    },
+    {
+      "epoch": 10.766990291262136,
+      "grad_norm": 0.5300271333343034,
+      "learning_rate": 9.197954330563327e-07,
+      "loss": 0.0732,
+      "step": 1109
+    },
+    {
+      "epoch": 10.776699029126213,
+      "grad_norm": 0.43694024379416824,
+      "learning_rate": 9.158595030414621e-07,
+      "loss": 0.0752,
+      "step": 1110
+    },
+    {
+      "epoch": 10.78640776699029,
+      "grad_norm": 0.5706126142150842,
+      "learning_rate": 9.11930122951554e-07,
+      "loss": 0.052,
+      "step": 1111
+    },
+    {
+      "epoch": 10.79611650485437,
+      "grad_norm": 0.4968147443898269,
+      "learning_rate": 9.080073090333646e-07,
+      "loss": 0.0618,
+      "step": 1112
+    },
+    {
+      "epoch": 10.805825242718447,
+      "grad_norm": 0.7309923946940867,
+      "learning_rate": 9.040910775065015e-07,
+      "loss": 0.1172,
+      "step": 1113
+    },
+    {
+      "epoch": 10.815533980582524,
+      "grad_norm": 0.4875038724359515,
+      "learning_rate": 9.001814445633558e-07,
+      "loss": 0.03,
+      "step": 1114
+    },
+    {
+      "epoch": 10.825242718446601,
+      "grad_norm": 0.8623721100042515,
+      "learning_rate": 8.962784263690358e-07,
+      "loss": 0.13,
+      "step": 1115
+    },
+    {
+      "epoch": 10.83495145631068,
+      "grad_norm": 0.6094686102538099,
+      "learning_rate": 8.923820390612991e-07,
+      "loss": 0.0742,
+      "step": 1116
+    },
+    {
+      "epoch": 10.844660194174757,
+      "grad_norm": 0.492347526443207,
+      "learning_rate": 8.884922987504882e-07,
+      "loss": 0.0564,
+      "step": 1117
+    },
+    {
+      "epoch": 10.854368932038835,
+      "grad_norm": 0.6685413110043293,
+      "learning_rate": 8.846092215194607e-07,
+      "loss": 0.1173,
+      "step": 1118
+    },
+    {
+      "epoch": 10.864077669902912,
+      "grad_norm": 0.34671831539541903,
+      "learning_rate": 8.807328234235254e-07,
+      "loss": 0.0343,
+      "step": 1119
+    },
+    {
+      "epoch": 10.87378640776699,
+      "grad_norm": 0.8551886410623927,
+      "learning_rate": 8.768631204903738e-07,
+      "loss": 0.1335,
+      "step": 1120
+    },
+    {
+      "epoch": 10.883495145631068,
+      "grad_norm": 0.8845986465738103,
+      "learning_rate": 8.730001287200177e-07,
+      "loss": 0.1715,
+      "step": 1121
+    },
+    {
+      "epoch": 10.893203883495145,
+      "grad_norm": 0.4515371919319129,
+      "learning_rate": 8.691438640847177e-07,
+      "loss": 0.0566,
+      "step": 1122
+    },
+    {
+      "epoch": 10.902912621359224,
+      "grad_norm": 0.8524620005123392,
+      "learning_rate": 8.652943425289206e-07,
+      "loss": 0.1265,
+      "step": 1123
+    },
+    {
+      "epoch": 10.912621359223301,
+      "grad_norm": 0.45736310100320554,
+      "learning_rate": 8.61451579969193e-07,
+      "loss": 0.0841,
+      "step": 1124
+    },
+    {
+      "epoch": 10.922330097087379,
+      "grad_norm": 0.528927869401848,
+      "learning_rate": 8.576155922941548e-07,
+      "loss": 0.0996,
+      "step": 1125
+    },
+    {
+      "epoch": 10.932038834951456,
+      "grad_norm": 0.5638408786011265,
+      "learning_rate": 8.537863953644138e-07,
+      "loss": 0.0894,
+      "step": 1126
+    },
+    {
+      "epoch": 10.941747572815533,
+      "grad_norm": 0.706509560577999,
+      "learning_rate": 8.499640050125007e-07,
+      "loss": 0.0952,
+      "step": 1127
+    },
+    {
+      "epoch": 10.951456310679612,
+      "grad_norm": 0.5818787000984781,
+      "learning_rate": 8.461484370428025e-07,
+      "loss": 0.0874,
+      "step": 1128
+    },
+    {
+      "epoch": 10.96116504854369,
+      "grad_norm": 0.5492742060252098,
+      "learning_rate": 8.423397072314985e-07,
+      "loss": 0.0736,
+      "step": 1129
+    },
+    {
+      "epoch": 10.970873786407767,
+      "grad_norm": 0.6223368740910458,
+      "learning_rate": 8.385378313264933e-07,
+      "loss": 0.0789,
+      "step": 1130
+    },
+    {
+      "epoch": 10.980582524271846,
+      "grad_norm": 0.7563283702541171,
+      "learning_rate": 8.347428250473541e-07,
+      "loss": 0.0494,
+      "step": 1131
+    },
+    {
+      "epoch": 10.990291262135923,
+      "grad_norm": 0.8246887369704925,
+      "learning_rate": 8.309547040852434e-07,
+      "loss": 0.0263,
+      "step": 1132
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.6209593471191105,
+      "learning_rate": 8.271734841028553e-07,
+      "loss": 0.0908,
+      "step": 1133
+    },
+    {
+      "epoch": 11.009708737864077,
+      "grad_norm": 0.6181167641935551,
+      "learning_rate": 8.233991807343497e-07,
+      "loss": 0.101,
+      "step": 1134
+    },
+    {
+      "epoch": 11.019417475728156,
+      "grad_norm": 0.3294479029172337,
+      "learning_rate": 8.196318095852909e-07,
+      "loss": 0.0329,
+      "step": 1135
+    },
+    {
+      "epoch": 11.029126213592233,
+      "grad_norm": 0.5181029223146717,
+      "learning_rate": 8.158713862325782e-07,
+      "loss": 0.0914,
+      "step": 1136
+    },
+    {
+      "epoch": 11.03883495145631,
+      "grad_norm": 0.38511569448960764,
+      "learning_rate": 8.12117926224385e-07,
+      "loss": 0.0414,
+      "step": 1137
+    },
+    {
+      "epoch": 11.048543689320388,
+      "grad_norm": 0.3120588007580457,
+      "learning_rate": 8.08371445080093e-07,
+      "loss": 0.0287,
+      "step": 1138
+    },
+    {
+      "epoch": 11.058252427184467,
+      "grad_norm": 0.35699123657831555,
+      "learning_rate": 8.04631958290229e-07,
+      "loss": 0.0224,
+      "step": 1139
+    },
+    {
+      "epoch": 11.067961165048544,
+      "grad_norm": 0.6874946805767921,
+      "learning_rate": 8.008994813163995e-07,
+      "loss": 0.142,
+      "step": 1140
+    },
+    {
+      "epoch": 11.077669902912621,
+      "grad_norm": 0.3712955806485938,
+      "learning_rate": 7.971740295912289e-07,
+      "loss": 0.0484,
+      "step": 1141
+    },
+    {
+      "epoch": 11.087378640776699,
+      "grad_norm": 0.4014946057982372,
+      "learning_rate": 7.934556185182928e-07,
+      "loss": 0.0498,
+      "step": 1142
+    },
+    {
+      "epoch": 11.097087378640778,
+      "grad_norm": 0.3599712596998677,
+      "learning_rate": 7.897442634720576e-07,
+      "loss": 0.0478,
+      "step": 1143
+    },
+    {
+      "epoch": 11.106796116504855,
+      "grad_norm": 0.6865673878538356,
+      "learning_rate": 7.860399797978138e-07,
+      "loss": 0.19,
+      "step": 1144
+    },
+    {
+      "epoch": 11.116504854368932,
+      "grad_norm": 0.6849682241558045,
+      "learning_rate": 7.823427828116148e-07,
+      "loss": 0.1168,
+      "step": 1145
+    },
+    {
+      "epoch": 11.12621359223301,
+      "grad_norm": 0.33934326981903534,
+      "learning_rate": 7.786526878002126e-07,
+      "loss": 0.0274,
+      "step": 1146
+    },
+    {
+      "epoch": 11.135922330097088,
+      "grad_norm": 0.5005017045058894,
+      "learning_rate": 7.749697100209947e-07,
+      "loss": 0.0795,
+      "step": 1147
+    },
+    {
+      "epoch": 11.145631067961165,
+      "grad_norm": 0.5020843300505056,
+      "learning_rate": 7.7129386470192e-07,
+      "loss": 0.0436,
+      "step": 1148
+    },
+    {
+      "epoch": 11.155339805825243,
+      "grad_norm": 0.4480201862020612,
+      "learning_rate": 7.6762516704146e-07,
+      "loss": 0.0674,
+      "step": 1149
+    },
+    {
+      "epoch": 11.16504854368932,
+      "grad_norm": 1.0455476424219226,
+      "learning_rate": 7.6396363220853e-07,
+      "loss": 0.1638,
+      "step": 1150
+    },
+    {
+      "epoch": 11.174757281553399,
+      "grad_norm": 0.576435965844432,
+      "learning_rate": 7.603092753424298e-07,
+      "loss": 0.0766,
+      "step": 1151
+    },
+    {
+      "epoch": 11.184466019417476,
+      "grad_norm": 0.32921646457176373,
+      "learning_rate": 7.566621115527811e-07,
+      "loss": 0.0348,
+      "step": 1152
+    },
+    {
+      "epoch": 11.194174757281553,
+      "grad_norm": 0.3367784456418941,
+      "learning_rate": 7.530221559194643e-07,
+      "loss": 0.0243,
+      "step": 1153
+    },
+    {
+      "epoch": 11.20388349514563,
+      "grad_norm": 0.38545867844467285,
+      "learning_rate": 7.493894234925558e-07,
+      "loss": 0.039,
+      "step": 1154
+    },
+    {
+      "epoch": 11.21359223300971,
+      "grad_norm": 0.45936186805463514,
+      "learning_rate": 7.457639292922675e-07,
+      "loss": 0.0729,
+      "step": 1155
+    },
+    {
+      "epoch": 11.223300970873787,
+      "grad_norm": 0.4905506201117723,
+      "learning_rate": 7.421456883088826e-07,
+      "loss": 0.0171,
+      "step": 1156
+    },
+    {
+      "epoch": 11.233009708737864,
+      "grad_norm": 0.6665752104633995,
+      "learning_rate": 7.385347155026934e-07,
+      "loss": 0.1204,
+      "step": 1157
+    },
+    {
+      "epoch": 11.242718446601941,
+      "grad_norm": 0.6519589690169562,
+      "learning_rate": 7.349310258039441e-07,
+      "loss": 0.1025,
+      "step": 1158
+    },
+    {
+      "epoch": 11.25242718446602,
+      "grad_norm": 0.6646274396055756,
+      "learning_rate": 7.31334634112762e-07,
+      "loss": 0.1342,
+      "step": 1159
+    },
+    {
+      "epoch": 11.262135922330097,
+      "grad_norm": 0.6617933047697685,
+      "learning_rate": 7.277455552991011e-07,
+      "loss": 0.1056,
+      "step": 1160
+    },
+    {
+      "epoch": 11.271844660194175,
+      "grad_norm": 0.28557786336015395,
+      "learning_rate": 7.241638042026783e-07,
+      "loss": 0.033,
+      "step": 1161
+    },
+    {
+      "epoch": 11.281553398058252,
+      "grad_norm": 0.3674990070012525,
+      "learning_rate": 7.20589395632913e-07,
+      "loss": 0.047,
+      "step": 1162
+    },
+    {
+      "epoch": 11.29126213592233,
+      "grad_norm": 0.7990437129788919,
+      "learning_rate": 7.170223443688654e-07,
+      "loss": 0.1382,
+      "step": 1163
+    },
+    {
+      "epoch": 11.300970873786408,
+      "grad_norm": 0.519463264590884,
+      "learning_rate": 7.134626651591758e-07,
+      "loss": 0.1124,
+      "step": 1164
+    },
+    {
+      "epoch": 11.310679611650485,
+      "grad_norm": 0.36441443966860854,
+      "learning_rate": 7.099103727220024e-07,
+      "loss": 0.0158,
+      "step": 1165
+    },
+    {
+      "epoch": 11.320388349514563,
+      "grad_norm": 1.0816424946810301,
+      "learning_rate": 7.063654817449638e-07,
+      "loss": 0.1172,
+      "step": 1166
+    },
+    {
+      "epoch": 11.330097087378642,
+      "grad_norm": 0.2813942915433234,
+      "learning_rate": 7.028280068850734e-07,
+      "loss": 0.0191,
+      "step": 1167
+    },
+    {
+      "epoch": 11.339805825242719,
+      "grad_norm": 0.5559575454325191,
+      "learning_rate": 6.992979627686821e-07,
+      "loss": 0.071,
+      "step": 1168
+    },
+    {
+      "epoch": 11.349514563106796,
+      "grad_norm": 0.6406408723152291,
+      "learning_rate": 6.957753639914175e-07,
+      "loss": 0.1288,
+      "step": 1169
+    },
+    {
+      "epoch": 11.359223300970873,
+      "grad_norm": 0.5616743245613947,
+      "learning_rate": 6.922602251181221e-07,
+      "loss": 0.0625,
+      "step": 1170
+    },
+    {
+      "epoch": 11.368932038834952,
+      "grad_norm": 0.6733872123620726,
+      "learning_rate": 6.887525606827947e-07,
+      "loss": 0.0958,
+      "step": 1171
+    },
+    {
+      "epoch": 11.37864077669903,
+      "grad_norm": 0.34814787820573256,
+      "learning_rate": 6.852523851885295e-07,
+      "loss": 0.0136,
+      "step": 1172
+    },
+    {
+      "epoch": 11.388349514563107,
+      "grad_norm": 0.45047065458077723,
+      "learning_rate": 6.817597131074566e-07,
+      "loss": 0.0339,
+      "step": 1173
+    },
+    {
+      "epoch": 11.398058252427184,
+      "grad_norm": 0.6006060313215769,
+      "learning_rate": 6.782745588806811e-07,
+      "loss": 0.0393,
+      "step": 1174
+    },
+    {
+      "epoch": 11.407766990291263,
+      "grad_norm": 0.4142872047428767,
+      "learning_rate": 6.747969369182248e-07,
+      "loss": 0.0461,
+      "step": 1175
+    },
+    {
+      "epoch": 11.41747572815534,
+      "grad_norm": 0.6117430165933601,
+      "learning_rate": 6.713268615989654e-07,
+      "loss": 0.08,
+      "step": 1176
+    },
+    {
+      "epoch": 11.427184466019417,
+      "grad_norm": 0.5949255359920863,
+      "learning_rate": 6.678643472705773e-07,
+      "loss": 0.1192,
+      "step": 1177
+    },
+    {
+      "epoch": 11.436893203883495,
+      "grad_norm": 0.46402567424048835,
+      "learning_rate": 6.644094082494746e-07,
+      "loss": 0.0223,
+      "step": 1178
+    },
+    {
+      "epoch": 11.446601941747574,
+      "grad_norm": 0.47547059564197874,
+      "learning_rate": 6.609620588207474e-07,
+      "loss": 0.0596,
+      "step": 1179
+    },
+    {
+      "epoch": 11.45631067961165,
+      "grad_norm": 0.5177539890623492,
+      "learning_rate": 6.575223132381067e-07,
+      "loss": 0.0997,
+      "step": 1180
+    },
+    {
+      "epoch": 11.466019417475728,
+      "grad_norm": 0.3197005241753569,
+      "learning_rate": 6.540901857238233e-07,
+      "loss": 0.03,
+      "step": 1181
+    },
+    {
+      "epoch": 11.475728155339805,
+      "grad_norm": 0.45523175279424344,
+      "learning_rate": 6.506656904686698e-07,
+      "loss": 0.0629,
+      "step": 1182
+    },
+    {
+      "epoch": 11.485436893203884,
+      "grad_norm": 0.4295816991757006,
+      "learning_rate": 6.472488416318621e-07,
+      "loss": 0.0689,
+      "step": 1183
+    },
+    {
+      "epoch": 11.495145631067961,
+      "grad_norm": 0.5455657421626049,
+      "learning_rate": 6.438396533410002e-07,
+      "loss": 0.069,
+      "step": 1184
+    },
+    {
+      "epoch": 11.504854368932039,
+      "grad_norm": 0.5744593180389926,
+      "learning_rate": 6.4043813969201e-07,
+      "loss": 0.034,
+      "step": 1185
+    },
+    {
+      "epoch": 11.514563106796116,
+      "grad_norm": 0.9974821460316966,
+      "learning_rate": 6.370443147490857e-07,
+      "loss": 0.0938,
+      "step": 1186
+    },
+    {
+      "epoch": 11.524271844660195,
+      "grad_norm": 0.37997934488556223,
+      "learning_rate": 6.336581925446309e-07,
+      "loss": 0.052,
+      "step": 1187
+    },
+    {
+      "epoch": 11.533980582524272,
+      "grad_norm": 0.5888434151251862,
+      "learning_rate": 6.302797870792007e-07,
+      "loss": 0.0605,
+      "step": 1188
+    },
+    {
+      "epoch": 11.54368932038835,
+      "grad_norm": 0.4110534807034854,
+      "learning_rate": 6.269091123214438e-07,
+      "loss": 0.0465,
+      "step": 1189
+    },
+    {
+      "epoch": 11.553398058252426,
+      "grad_norm": 0.6312639822083126,
+      "learning_rate": 6.235461822080449e-07,
+      "loss": 0.0984,
+      "step": 1190
+    },
+    {
+      "epoch": 11.563106796116505,
+      "grad_norm": 0.4184886778766774,
+      "learning_rate": 6.201910106436673e-07,
+      "loss": 0.0488,
+      "step": 1191
+    },
+    {
+      "epoch": 11.572815533980583,
+      "grad_norm": 0.5720300965783884,
+      "learning_rate": 6.168436115008941e-07,
+      "loss": 0.1142,
+      "step": 1192
+    },
+    {
+      "epoch": 11.58252427184466,
+      "grad_norm": 0.34247681006710967,
+      "learning_rate": 6.135039986201744e-07,
+      "loss": 0.0212,
+      "step": 1193
+    },
+    {
+      "epoch": 11.592233009708737,
+      "grad_norm": 0.8696585733772886,
+      "learning_rate": 6.101721858097606e-07,
+      "loss": 0.1344,
+      "step": 1194
+    },
+    {
+      "epoch": 11.601941747572816,
+      "grad_norm": 0.5733831310709949,
+      "learning_rate": 6.068481868456558e-07,
+      "loss": 0.1033,
+      "step": 1195
+    },
+    {
+      "epoch": 11.611650485436893,
+      "grad_norm": 0.3726614583489184,
+      "learning_rate": 6.035320154715549e-07,
+      "loss": 0.0509,
+      "step": 1196
+    },
+    {
+      "epoch": 11.62135922330097,
+      "grad_norm": 0.4551557088282536,
+      "learning_rate": 6.00223685398788e-07,
+      "loss": 0.027,
+      "step": 1197
+    },
+    {
+      "epoch": 11.631067961165048,
+      "grad_norm": 0.3944998808707215,
+      "learning_rate": 5.969232103062647e-07,
+      "loss": 0.0715,
+      "step": 1198
+    },
+    {
+      "epoch": 11.640776699029127,
+      "grad_norm": 0.12933923216594015,
+      "learning_rate": 5.936306038404158e-07,
+      "loss": 0.0041,
+      "step": 1199
+    },
+    {
+      "epoch": 11.650485436893204,
+      "grad_norm": 0.7601718028278395,
+      "learning_rate": 5.903458796151382e-07,
+      "loss": 0.2222,
+      "step": 1200
+    },
+    {
+      "epoch": 11.660194174757281,
+      "grad_norm": 0.33435340501447186,
+      "learning_rate": 5.870690512117377e-07,
+      "loss": 0.0368,
+      "step": 1201
+    },
+    {
+      "epoch": 11.669902912621358,
+      "grad_norm": 0.4520959141289102,
+      "learning_rate": 5.838001321788744e-07,
+      "loss": 0.0293,
+      "step": 1202
+    },
+    {
+      "epoch": 11.679611650485437,
+      "grad_norm": 0.3901390181712294,
+      "learning_rate": 5.80539136032505e-07,
+      "loss": 0.0468,
+      "step": 1203
+    },
+    {
+      "epoch": 11.689320388349515,
+      "grad_norm": 0.675514183694151,
+      "learning_rate": 5.772860762558269e-07,
+      "loss": 0.1079,
+      "step": 1204
+    },
+    {
+      "epoch": 11.699029126213592,
+      "grad_norm": 0.4486143842474179,
+      "learning_rate": 5.740409662992244e-07,
+      "loss": 0.0563,
+      "step": 1205
+    },
+    {
+      "epoch": 11.70873786407767,
+      "grad_norm": 0.543813104820745,
+      "learning_rate": 5.708038195802098e-07,
+      "loss": 0.0363,
+      "step": 1206
+    },
+    {
+      "epoch": 11.718446601941748,
+      "grad_norm": 0.6554627134569425,
+      "learning_rate": 5.675746494833733e-07,
+      "loss": 0.0514,
+      "step": 1207
+    },
+    {
+      "epoch": 11.728155339805825,
+      "grad_norm": 0.4505479059998464,
+      "learning_rate": 5.643534693603214e-07,
+      "loss": 0.027,
+      "step": 1208
+    },
+    {
+      "epoch": 11.737864077669903,
+      "grad_norm": 0.5220937013677023,
+      "learning_rate": 5.61140292529625e-07,
+      "loss": 0.0843,
+      "step": 1209
+    },
+    {
+      "epoch": 11.74757281553398,
+      "grad_norm": 0.3609261316739773,
+      "learning_rate": 5.579351322767643e-07,
+      "loss": 0.0334,
+      "step": 1210
+    },
+    {
+      "epoch": 11.757281553398059,
+      "grad_norm": 0.5351464600044358,
+      "learning_rate": 5.547380018540735e-07,
+      "loss": 0.1097,
+      "step": 1211
+    },
+    {
+      "epoch": 11.766990291262136,
+      "grad_norm": 0.277458376852476,
+      "learning_rate": 5.515489144806862e-07,
+      "loss": 0.0319,
+      "step": 1212
+    },
+    {
+      "epoch": 11.776699029126213,
+      "grad_norm": 0.8927290578987768,
+      "learning_rate": 5.483678833424796e-07,
+      "loss": 0.1248,
+      "step": 1213
+    },
+    {
+      "epoch": 11.78640776699029,
+      "grad_norm": 0.48393508803090785,
+      "learning_rate": 5.451949215920221e-07,
+      "loss": 0.0574,
+      "step": 1214
+    },
+    {
+      "epoch": 11.79611650485437,
+      "grad_norm": 0.5362198231627286,
+      "learning_rate": 5.420300423485167e-07,
+      "loss": 0.0403,
+      "step": 1215
+    },
+    {
+      "epoch": 11.805825242718447,
+      "grad_norm": 0.6831834341558954,
+      "learning_rate": 5.38873258697748e-07,
+      "loss": 0.0537,
+      "step": 1216
+    },
+    {
+      "epoch": 11.815533980582524,
+      "grad_norm": 0.7235561973898877,
+      "learning_rate": 5.357245836920286e-07,
+      "loss": 0.0887,
+      "step": 1217
+    },
+    {
+      "epoch": 11.825242718446601,
+      "grad_norm": 0.5011302176239788,
+      "learning_rate": 5.325840303501431e-07,
+      "loss": 0.0639,
+      "step": 1218
+    },
+    {
+      "epoch": 11.83495145631068,
+      "grad_norm": 0.48804341236318166,
+      "learning_rate": 5.29451611657297e-07,
+      "loss": 0.0798,
+      "step": 1219
+    },
+    {
+      "epoch": 11.844660194174757,
+      "grad_norm": 0.5695494446583874,
+      "learning_rate": 5.263273405650601e-07,
+      "loss": 0.0955,
+      "step": 1220
+    },
+    {
+      "epoch": 11.854368932038835,
+      "grad_norm": 0.5618369954982938,
+      "learning_rate": 5.232112299913151e-07,
+      "loss": 0.0821,
+      "step": 1221
+    },
+    {
+      "epoch": 11.864077669902912,
+      "grad_norm": 0.4770004457987746,
+      "learning_rate": 5.201032928202043e-07,
+      "loss": 0.0906,
+      "step": 1222
+    },
+    {
+      "epoch": 11.87378640776699,
+      "grad_norm": 0.7242922898548877,
+      "learning_rate": 5.17003541902075e-07,
+      "loss": 0.1936,
+      "step": 1223
+    },
+    {
+      "epoch": 11.883495145631068,
+      "grad_norm": 0.50876950009869,
+      "learning_rate": 5.139119900534259e-07,
+      "loss": 0.1009,
+      "step": 1224
+    },
+    {
+      "epoch": 11.893203883495145,
+      "grad_norm": 0.4007899567960424,
+      "learning_rate": 5.108286500568562e-07,
+      "loss": 0.0333,
+      "step": 1225
+    },
+    {
+      "epoch": 11.902912621359224,
+      "grad_norm": 0.5343766514104662,
+      "learning_rate": 5.077535346610115e-07,
+      "loss": 0.0554,
+      "step": 1226
+    },
+    {
+      "epoch": 11.912621359223301,
+      "grad_norm": 0.535666198151551,
+      "learning_rate": 5.046866565805311e-07,
+      "loss": 0.0292,
+      "step": 1227
+    },
+    {
+      "epoch": 11.922330097087379,
+      "grad_norm": 0.4527597562080739,
+      "learning_rate": 5.016280284959957e-07,
+      "loss": 0.043,
+      "step": 1228
+    },
+    {
+      "epoch": 11.932038834951456,
+      "grad_norm": 0.6442865775092774,
+      "learning_rate": 4.985776630538746e-07,
+      "loss": 0.0828,
+      "step": 1229
+    },
+    {
+      "epoch": 11.941747572815533,
+      "grad_norm": 0.3977555677630664,
+      "learning_rate": 4.95535572866474e-07,
+      "loss": 0.0296,
+      "step": 1230
+    },
+    {
+      "epoch": 11.951456310679612,
+      "grad_norm": 0.27529002957180204,
+      "learning_rate": 4.925017705118843e-07,
+      "loss": 0.0157,
+      "step": 1231
+    },
+    {
+      "epoch": 11.96116504854369,
+      "grad_norm": 0.7429628409554702,
+      "learning_rate": 4.89476268533928e-07,
+      "loss": 0.0759,
+      "step": 1232
+    },
+    {
+      "epoch": 11.970873786407767,
+      "grad_norm": 0.6138000975659442,
+      "learning_rate": 4.864590794421092e-07,
+      "loss": 0.0529,
+      "step": 1233
+    },
+    {
+      "epoch": 11.980582524271846,
+      "grad_norm": 0.6182825584807163,
+      "learning_rate": 4.834502157115597e-07,
+      "loss": 0.092,
+      "step": 1234
+    },
+    {
+      "epoch": 11.990291262135923,
+      "grad_norm": 0.24435544743564855,
+      "learning_rate": 4.804496897829883e-07,
+      "loss": 0.0125,
+      "step": 1235
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.8424484319909388,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.1078,
+      "step": 1236
+    },
+    {
+      "epoch": 12.009708737864077,
+      "grad_norm": 0.5129391571852897,
+      "learning_rate": 4.744737009221986e-07,
+      "loss": 0.0749,
+      "step": 1237
+    },
+    {
+      "epoch": 12.019417475728156,
+      "grad_norm": 0.5919150233331142,
+      "learning_rate": 4.7149826269882294e-07,
+      "loss": 0.0702,
+      "step": 1238
+    },
+    {
+      "epoch": 12.029126213592233,
+      "grad_norm": 0.44284697506749743,
+      "learning_rate": 4.6853121169500914e-07,
+      "loss": 0.0745,
+      "step": 1239
+    },
+    {
+      "epoch": 12.03883495145631,
+      "grad_norm": 0.2865977034088059,
+      "learning_rate": 4.6557256017858485e-07,
+      "loss": 0.0215,
+      "step": 1240
+    },
+    {
+      "epoch": 12.048543689320388,
+      "grad_norm": 0.4160486540359403,
+      "learning_rate": 4.626223203826477e-07,
+      "loss": 0.0407,
+      "step": 1241
+    },
+    {
+      "epoch": 12.058252427184467,
+      "grad_norm": 0.2749010600950671,
+      "learning_rate": 4.5968050450551527e-07,
+      "loss": 0.0199,
+      "step": 1242
+    },
+    {
+      "epoch": 12.067961165048544,
+      "grad_norm": 0.45897302010048824,
+      "learning_rate": 4.56747124710675e-07,
+      "loss": 0.0625,
+      "step": 1243
+    },
+    {
+      "epoch": 12.077669902912621,
+      "grad_norm": 0.5477002655080845,
+      "learning_rate": 4.5382219312673364e-07,
+      "loss": 0.0808,
+      "step": 1244
+    },
+    {
+      "epoch": 12.087378640776699,
+      "grad_norm": 0.2698283023459079,
+      "learning_rate": 4.5090572184736863e-07,
+      "loss": 0.0212,
+      "step": 1245
+    },
+    {
+      "epoch": 12.097087378640778,
+      "grad_norm": 0.7558827780843274,
+      "learning_rate": 4.4799772293127486e-07,
+      "loss": 0.1129,
+      "step": 1246
+    },
+    {
+      "epoch": 12.106796116504855,
+      "grad_norm": 0.5006893727164596,
+      "learning_rate": 4.4509820840211745e-07,
+      "loss": 0.0552,
+      "step": 1247
+    },
+    {
+      "epoch": 12.116504854368932,
+      "grad_norm": 0.41471815721736377,
+      "learning_rate": 4.422071902484812e-07,
+      "loss": 0.0522,
+      "step": 1248
+    },
+    {
+      "epoch": 12.12621359223301,
+      "grad_norm": 0.34642625579244757,
+      "learning_rate": 4.3932468042382075e-07,
+      "loss": 0.0473,
+      "step": 1249
+    },
+    {
+      "epoch": 12.135922330097088,
+      "grad_norm": 0.4724063216096118,
+      "learning_rate": 4.3645069084641195e-07,
+      "loss": 0.0693,
+      "step": 1250
+    },
+    {
+      "epoch": 12.145631067961165,
+      "grad_norm": 0.5018916517574505,
+      "learning_rate": 4.335852333993018e-07,
+      "loss": 0.0828,
+      "step": 1251
+    },
+    {
+      "epoch": 12.155339805825243,
+      "grad_norm": 0.47914801086777986,
+      "learning_rate": 4.3072831993025895e-07,
+      "loss": 0.0661,
+      "step": 1252
+    },
+    {
+      "epoch": 12.16504854368932,
+      "grad_norm": 0.47409307088022173,
+      "learning_rate": 4.278799622517274e-07,
+      "loss": 0.0635,
+      "step": 1253
+    },
+    {
+      "epoch": 12.174757281553399,
+      "grad_norm": 0.4932946423501068,
+      "learning_rate": 4.2504017214077374e-07,
+      "loss": 0.0719,
+      "step": 1254
+    },
+    {
+      "epoch": 12.184466019417476,
+      "grad_norm": 0.2552904501885662,
+      "learning_rate": 4.222089613390412e-07,
+      "loss": 0.0125,
+      "step": 1255
+    },
+    {
+      "epoch": 12.194174757281553,
+      "grad_norm": 0.3765642214308931,
+      "learning_rate": 4.1938634155269944e-07,
+      "loss": 0.0429,
+      "step": 1256
+    },
+    {
+      "epoch": 12.20388349514563,
+      "grad_norm": 0.47304984449216975,
+      "learning_rate": 4.165723244523978e-07,
+      "loss": 0.0658,
+      "step": 1257
+    },
+    {
+      "epoch": 12.21359223300971,
+      "grad_norm": 0.5467448402302595,
+      "learning_rate": 4.1376692167321626e-07,
+      "loss": 0.082,
+      "step": 1258
+    },
+    {
+      "epoch": 12.223300970873787,
+      "grad_norm": 0.4153879006735531,
+      "learning_rate": 4.109701448146164e-07,
+      "loss": 0.0567,
+      "step": 1259
+    },
+    {
+      "epoch": 12.233009708737864,
+      "grad_norm": 0.5149132605681401,
+      "learning_rate": 4.0818200544039484e-07,
+      "loss": 0.0642,
+      "step": 1260
+    },
+    {
+      "epoch": 12.242718446601941,
+      "grad_norm": 0.3664582297237133,
+      "learning_rate": 4.054025150786356e-07,
+      "loss": 0.0207,
+      "step": 1261
+    },
+    {
+      "epoch": 12.25242718446602,
+      "grad_norm": 0.4704662984536062,
+      "learning_rate": 4.026316852216605e-07,
+      "loss": 0.0571,
+      "step": 1262
+    },
+    {
+      "epoch": 12.262135922330097,
+      "grad_norm": 0.4316859066370346,
+      "learning_rate": 3.998695273259834e-07,
+      "loss": 0.0687,
+      "step": 1263
+    },
+    {
+      "epoch": 12.271844660194175,
+      "grad_norm": 0.5327787979259686,
+      "learning_rate": 3.971160528122622e-07,
+      "loss": 0.0856,
+      "step": 1264
+    },
+    {
+      "epoch": 12.281553398058252,
+      "grad_norm": 0.5018036756970643,
+      "learning_rate": 3.9437127306525295e-07,
+      "loss": 0.1052,
+      "step": 1265
+    },
+    {
+      "epoch": 12.29126213592233,
+      "grad_norm": 0.3177685723026178,
+      "learning_rate": 3.9163519943375973e-07,
+      "loss": 0.0283,
+      "step": 1266
+    },
+    {
+      "epoch": 12.300970873786408,
+      "grad_norm": 0.5165156442833573,
+      "learning_rate": 3.889078432305904e-07,
+      "loss": 0.118,
+      "step": 1267
+    },
+    {
+      "epoch": 12.310679611650485,
+      "grad_norm": 0.5068137023886546,
+      "learning_rate": 3.8618921573250896e-07,
+      "loss": 0.0775,
+      "step": 1268
+    },
+    {
+      "epoch": 12.320388349514563,
+      "grad_norm": 0.48634984455464964,
+      "learning_rate": 3.834793281801891e-07,
+      "loss": 0.0344,
+      "step": 1269
+    },
+    {
+      "epoch": 12.330097087378642,
+      "grad_norm": 0.5145224870057581,
+      "learning_rate": 3.8077819177816695e-07,
+      "loss": 0.1188,
+      "step": 1270
+    },
+    {
+      "epoch": 12.339805825242719,
+      "grad_norm": 0.7335439939955951,
+      "learning_rate": 3.780858176947963e-07,
+      "loss": 0.0506,
+      "step": 1271
+    },
+    {
+      "epoch": 12.349514563106796,
+      "grad_norm": 0.2890103308208032,
+      "learning_rate": 3.754022170622007e-07,
+      "loss": 0.0195,
+      "step": 1272
+    },
+    {
+      "epoch": 12.359223300970873,
+      "grad_norm": 0.5274645907133224,
+      "learning_rate": 3.7272740097622884e-07,
+      "loss": 0.1081,
+      "step": 1273
+    },
+    {
+      "epoch": 12.368932038834952,
+      "grad_norm": 0.5229710405073345,
+      "learning_rate": 3.700613804964073e-07,
+      "loss": 0.0967,
+      "step": 1274
+    },
+    {
+      "epoch": 12.37864077669903,
+      "grad_norm": 0.5402500562513454,
+      "learning_rate": 3.6740416664589634e-07,
+      "loss": 0.0806,
+      "step": 1275
+    },
+    {
+      "epoch": 12.388349514563107,
+      "grad_norm": 0.6932722573513473,
+      "learning_rate": 3.6475577041144324e-07,
+      "loss": 0.1607,
+      "step": 1276
+    },
+    {
+      "epoch": 12.398058252427184,
+      "grad_norm": 0.13364800561868725,
+      "learning_rate": 3.6211620274333727e-07,
+      "loss": 0.0039,
+      "step": 1277
+    },
+    {
+      "epoch": 12.407766990291263,
+      "grad_norm": 0.5635217291309577,
+      "learning_rate": 3.594854745553636e-07,
+      "loss": 0.089,
+      "step": 1278
+    },
+    {
+      "epoch": 12.41747572815534,
+      "grad_norm": 0.8039998796146036,
+      "learning_rate": 3.568635967247605e-07,
+      "loss": 0.1989,
+      "step": 1279
+    },
+    {
+      "epoch": 12.427184466019417,
+      "grad_norm": 0.44452718171739464,
+      "learning_rate": 3.5425058009217193e-07,
+      "loss": 0.0505,
+      "step": 1280
+    },
+    {
+      "epoch": 12.436893203883495,
+      "grad_norm": 0.4779630622559031,
+      "learning_rate": 3.516464354616031e-07,
+      "loss": 0.0523,
+      "step": 1281
+    },
+    {
+      "epoch": 12.446601941747574,
+      "grad_norm": 0.5081805906166073,
+      "learning_rate": 3.4905117360037683e-07,
+      "loss": 0.0823,
+      "step": 1282
+    },
+    {
+      "epoch": 12.45631067961165,
+      "grad_norm": 0.3588244687459001,
+      "learning_rate": 3.4646480523908813e-07,
+      "loss": 0.0347,
+      "step": 1283
+    },
+    {
+      "epoch": 12.466019417475728,
+      "grad_norm": 0.6863092097683883,
+      "learning_rate": 3.43887341071561e-07,
+      "loss": 0.1559,
+      "step": 1284
+    },
+    {
+      "epoch": 12.475728155339805,
+      "grad_norm": 0.20582305371798984,
+      "learning_rate": 3.413187917548019e-07,
+      "loss": 0.0101,
+      "step": 1285
+    },
+    {
+      "epoch": 12.485436893203884,
+      "grad_norm": 0.6067333614215648,
+      "learning_rate": 3.3875916790895883e-07,
+      "loss": 0.1478,
+      "step": 1286
+    },
+    {
+      "epoch": 12.495145631067961,
+      "grad_norm": 0.3103903276686346,
+      "learning_rate": 3.3620848011727437e-07,
+      "loss": 0.0157,
+      "step": 1287
+    },
+    {
+      "epoch": 12.504854368932039,
+      "grad_norm": 0.2398685670975744,
+      "learning_rate": 3.336667389260445e-07,
+      "loss": 0.0115,
+      "step": 1288
+    },
+    {
+      "epoch": 12.514563106796116,
+      "grad_norm": 0.34028815868720313,
+      "learning_rate": 3.311339548445727e-07,
+      "loss": 0.0188,
+      "step": 1289
+    },
+    {
+      "epoch": 12.524271844660195,
+      "grad_norm": 0.5281529937098564,
+      "learning_rate": 3.2861013834512844e-07,
+      "loss": 0.0705,
+      "step": 1290
+    },
+    {
+      "epoch": 12.533980582524272,
+      "grad_norm": 0.44922395870868254,
+      "learning_rate": 3.2609529986290246e-07,
+      "loss": 0.0594,
+      "step": 1291
+    },
+    {
+      "epoch": 12.54368932038835,
+      "grad_norm": 0.41450177685889456,
+      "learning_rate": 3.235894497959649e-07,
+      "loss": 0.0323,
+      "step": 1292
+    },
+    {
+      "epoch": 12.553398058252426,
+      "grad_norm": 0.4860873761111312,
+      "learning_rate": 3.2109259850522045e-07,
+      "loss": 0.0639,
+      "step": 1293
+    },
+    {
+      "epoch": 12.563106796116505,
+      "grad_norm": 0.5025904279404237,
+      "learning_rate": 3.186047563143685e-07,
+      "loss": 0.0705,
+      "step": 1294
+    },
+    {
+      "epoch": 12.572815533980583,
+      "grad_norm": 0.375714081361163,
+      "learning_rate": 3.161259335098571e-07,
+      "loss": 0.035,
+      "step": 1295
+    },
+    {
+      "epoch": 12.58252427184466,
+      "grad_norm": 0.2256329375444976,
+      "learning_rate": 3.1365614034084224e-07,
+      "loss": 0.012,
+      "step": 1296
+    },
+    {
+      "epoch": 12.592233009708737,
+      "grad_norm": 0.2881991746431682,
+      "learning_rate": 3.111953870191459e-07,
+      "loss": 0.0255,
+      "step": 1297
+    },
+    {
+      "epoch": 12.601941747572816,
+      "grad_norm": 0.6335108729127409,
+      "learning_rate": 3.087436837192118e-07,
+      "loss": 0.1165,
+      "step": 1298
+    },
+    {
+      "epoch": 12.611650485436893,
+      "grad_norm": 0.47028695989481234,
+      "learning_rate": 3.0630104057806616e-07,
+      "loss": 0.1114,
+      "step": 1299
+    },
+    {
+      "epoch": 12.62135922330097,
+      "grad_norm": 0.5121407246325066,
+      "learning_rate": 3.0386746769527323e-07,
+      "loss": 0.0878,
+      "step": 1300
+    },
+    {
+      "epoch": 12.631067961165048,
+      "grad_norm": 0.31821815756972943,
+      "learning_rate": 3.0144297513289483e-07,
+      "loss": 0.0224,
+      "step": 1301
+    },
+    {
+      "epoch": 12.640776699029127,
+      "grad_norm": 0.36131410103922523,
+      "learning_rate": 2.9902757291544905e-07,
+      "loss": 0.0338,
+      "step": 1302
+    },
+    {
+      "epoch": 12.650485436893204,
+      "grad_norm": 0.18546100236750176,
+      "learning_rate": 2.966212710298674e-07,
+      "loss": 0.0077,
+      "step": 1303
+    },
+    {
+      "epoch": 12.660194174757281,
+      "grad_norm": 0.4403213642112429,
+      "learning_rate": 2.94224079425455e-07,
+      "loss": 0.0444,
+      "step": 1304
+    },
+    {
+      "epoch": 12.669902912621358,
+      "grad_norm": 0.30588085097166257,
+      "learning_rate": 2.9183600801384853e-07,
+      "loss": 0.0166,
+      "step": 1305
+    },
+    {
+      "epoch": 12.679611650485437,
+      "grad_norm": 0.19655227972846973,
+      "learning_rate": 2.8945706666897555e-07,
+      "loss": 0.0067,
+      "step": 1306
+    },
+    {
+      "epoch": 12.689320388349515,
+      "grad_norm": 0.4034524699481664,
+      "learning_rate": 2.870872652270129e-07,
+      "loss": 0.0485,
+      "step": 1307
+    },
+    {
+      "epoch": 12.699029126213592,
+      "grad_norm": 0.3342688517721637,
+      "learning_rate": 2.8472661348634883e-07,
+      "loss": 0.0224,
+      "step": 1308
+    },
+    {
+      "epoch": 12.70873786407767,
+      "grad_norm": 0.35443206447047615,
+      "learning_rate": 2.82375121207539e-07,
+      "loss": 0.051,
+      "step": 1309
+    },
+    {
+      "epoch": 12.718446601941748,
+      "grad_norm": 0.413534023109782,
+      "learning_rate": 2.8003279811326724e-07,
+      "loss": 0.0508,
+      "step": 1310
+    },
+    {
+      "epoch": 12.728155339805825,
+      "grad_norm": 0.19635527159953645,
+      "learning_rate": 2.776996538883062e-07,
+      "loss": 0.0085,
+      "step": 1311
+    },
+    {
+      "epoch": 12.737864077669903,
+      "grad_norm": 0.5416519429825231,
+      "learning_rate": 2.7537569817947694e-07,
+      "loss": 0.0942,
+      "step": 1312
+    },
+    {
+      "epoch": 12.74757281553398,
+      "grad_norm": 0.40253806362920896,
+      "learning_rate": 2.730609405956083e-07,
+      "loss": 0.0523,
+      "step": 1313
+    },
+    {
+      "epoch": 12.757281553398059,
+      "grad_norm": 0.49734033038996395,
+      "learning_rate": 2.707553907074989e-07,
+      "loss": 0.0739,
+      "step": 1314
+    },
+    {
+      "epoch": 12.766990291262136,
+      "grad_norm": 0.5385383009087314,
+      "learning_rate": 2.684590580478749e-07,
+      "loss": 0.0546,
+      "step": 1315
+    },
+    {
+      "epoch": 12.776699029126213,
+      "grad_norm": 0.6286671447796734,
+      "learning_rate": 2.6617195211135343e-07,
+      "loss": 0.0952,
+      "step": 1316
+    },
+    {
+      "epoch": 12.78640776699029,
+      "grad_norm": 0.5855989296835651,
+      "learning_rate": 2.638940823544012e-07,
+      "loss": 0.0667,
+      "step": 1317
+    },
+    {
+      "epoch": 12.79611650485437,
+      "grad_norm": 0.33409793915536806,
+      "learning_rate": 2.6162545819529624e-07,
+      "loss": 0.036,
+      "step": 1318
+    },
+    {
+      "epoch": 12.805825242718447,
+      "grad_norm": 0.5781144234706445,
+      "learning_rate": 2.593660890140895e-07,
+      "loss": 0.0608,
+      "step": 1319
+    },
+    {
+      "epoch": 12.815533980582524,
+      "grad_norm": 0.5362480167302143,
+      "learning_rate": 2.57115984152565e-07,
+      "loss": 0.0719,
+      "step": 1320
+    },
+    {
+      "epoch": 12.825242718446601,
+      "grad_norm": 0.4294606474451055,
+      "learning_rate": 2.548751529142018e-07,
+      "loss": 0.0856,
+      "step": 1321
+    },
+    {
+      "epoch": 12.83495145631068,
+      "grad_norm": 0.6234632237849053,
+      "learning_rate": 2.526436045641351e-07,
+      "loss": 0.1454,
+      "step": 1322
+    },
+    {
+      "epoch": 12.844660194174757,
+      "grad_norm": 0.4658449714656103,
+      "learning_rate": 2.504213483291193e-07,
+      "loss": 0.0475,
+      "step": 1323
+    },
+    {
+      "epoch": 12.854368932038835,
+      "grad_norm": 0.23326161655035757,
+      "learning_rate": 2.482083933974883e-07,
+      "loss": 0.0105,
+      "step": 1324
+    },
+    {
+      "epoch": 12.864077669902912,
+      "grad_norm": 0.5872406960043914,
+      "learning_rate": 2.4600474891911696e-07,
+      "loss": 0.0925,
+      "step": 1325
+    },
+    {
+      "epoch": 12.87378640776699,
+      "grad_norm": 0.2975400185848686,
+      "learning_rate": 2.43810424005386e-07,
+      "loss": 0.0382,
+      "step": 1326
+    },
+    {
+      "epoch": 12.883495145631068,
+      "grad_norm": 0.40439938966679095,
+      "learning_rate": 2.416254277291416e-07,
+      "loss": 0.0424,
+      "step": 1327
+    },
+    {
+      "epoch": 12.893203883495145,
+      "grad_norm": 0.14455878990664323,
+      "learning_rate": 2.3944976912465916e-07,
+      "loss": 0.0043,
+      "step": 1328
+    },
+    {
+      "epoch": 12.902912621359224,
+      "grad_norm": 0.5413563478617743,
+      "learning_rate": 2.3728345718760622e-07,
+      "loss": 0.0858,
+      "step": 1329
+    },
+    {
+      "epoch": 12.912621359223301,
+      "grad_norm": 0.6114421008718739,
+      "learning_rate": 2.3512650087500338e-07,
+      "loss": 0.1262,
+      "step": 1330
+    },
+    {
+      "epoch": 12.922330097087379,
+      "grad_norm": 0.40434143898326247,
+      "learning_rate": 2.3297890910519093e-07,
+      "loss": 0.0438,
+      "step": 1331
+    },
+    {
+      "epoch": 12.932038834951456,
+      "grad_norm": 0.15674358972140642,
+      "learning_rate": 2.3084069075778758e-07,
+      "loss": 0.0042,
+      "step": 1332
+    },
+    {
+      "epoch": 12.941747572815533,
+      "grad_norm": 0.4385566852023257,
+      "learning_rate": 2.287118546736572e-07,
+      "loss": 0.0527,
+      "step": 1333
+    },
+    {
+      "epoch": 12.951456310679612,
+      "grad_norm": 0.726154070219994,
+      "learning_rate": 2.2659240965487023e-07,
+      "loss": 0.0914,
+      "step": 1334
+    },
+    {
+      "epoch": 12.96116504854369,
+      "grad_norm": 0.6620758511943872,
+      "learning_rate": 2.2448236446466847e-07,
+      "loss": 0.0509,
+      "step": 1335
+    },
+    {
+      "epoch": 12.970873786407767,
+      "grad_norm": 0.6769976034708189,
+      "learning_rate": 2.2238172782742763e-07,
+      "loss": 0.1015,
+      "step": 1336
+    },
+    {
+      "epoch": 12.980582524271846,
+      "grad_norm": 0.24635745023823036,
+      "learning_rate": 2.2029050842862277e-07,
+      "loss": 0.0112,
+      "step": 1337
+    },
+    {
+      "epoch": 12.990291262135923,
+      "grad_norm": 0.5211584464733092,
+      "learning_rate": 2.1820871491479102e-07,
+      "loss": 0.0667,
+      "step": 1338
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.5913539447878209,
+      "learning_rate": 2.1613635589349756e-07,
+      "loss": 0.0904,
+      "step": 1339
+    },
+    {
+      "epoch": 13.009708737864077,
+      "grad_norm": 0.3101489157097074,
+      "learning_rate": 2.140734399332975e-07,
+      "loss": 0.0542,
+      "step": 1340
+    },
+    {
+      "epoch": 13.019417475728156,
+      "grad_norm": 0.5351936690061951,
+      "learning_rate": 2.1201997556370284e-07,
+      "loss": 0.0846,
+      "step": 1341
+    },
+    {
+      "epoch": 13.029126213592233,
+      "grad_norm": 0.3577628010330391,
+      "learning_rate": 2.0997597127514507e-07,
+      "loss": 0.0232,
+      "step": 1342
+    },
+    {
+      "epoch": 13.03883495145631,
+      "grad_norm": 0.4025795427933299,
+      "learning_rate": 2.079414355189427e-07,
+      "loss": 0.0669,
+      "step": 1343
+    },
+    {
+      "epoch": 13.048543689320388,
+      "grad_norm": 0.48650418922975114,
+      "learning_rate": 2.059163767072639e-07,
+      "loss": 0.0605,
+      "step": 1344
+    },
+    {
+      "epoch": 13.058252427184467,
+      "grad_norm": 0.3135425232894008,
+      "learning_rate": 2.0390080321309236e-07,
+      "loss": 0.0269,
+      "step": 1345
+    },
+    {
+      "epoch": 13.067961165048544,
+      "grad_norm": 0.08376709705022621,
+      "learning_rate": 2.01894723370194e-07,
+      "loss": 0.0022,
+      "step": 1346
+    },
+    {
+      "epoch": 13.077669902912621,
+      "grad_norm": 0.33362664095446987,
+      "learning_rate": 1.9989814547308056e-07,
+      "loss": 0.037,
+      "step": 1347
+    },
+    {
+      "epoch": 13.087378640776699,
+      "grad_norm": 0.6004311575355316,
+      "learning_rate": 1.9791107777697633e-07,
+      "loss": 0.1411,
+      "step": 1348
+    },
+    {
+      "epoch": 13.097087378640778,
+      "grad_norm": 0.34278443971860567,
+      "learning_rate": 1.9593352849778453e-07,
+      "loss": 0.0254,
+      "step": 1349
+    },
+    {
+      "epoch": 13.106796116504855,
+      "grad_norm": 0.3286684875471459,
+      "learning_rate": 1.9396550581205208e-07,
+      "loss": 0.0228,
+      "step": 1350
+    },
+    {
+      "epoch": 13.116504854368932,
+      "grad_norm": 0.34666070981368075,
+      "learning_rate": 1.920070178569361e-07,
+      "loss": 0.0311,
+      "step": 1351
+    },
+    {
+      "epoch": 13.12621359223301,
+      "grad_norm": 0.49387258406448203,
+      "learning_rate": 1.900580727301718e-07,
+      "loss": 0.0555,
+      "step": 1352
+    },
+    {
+      "epoch": 13.135922330097088,
+      "grad_norm": 0.5181298337375579,
+      "learning_rate": 1.8811867849003684e-07,
+      "loss": 0.0442,
+      "step": 1353
+    },
+    {
+      "epoch": 13.145631067961165,
+      "grad_norm": 0.5372907594532009,
+      "learning_rate": 1.8618884315531939e-07,
+      "loss": 0.0758,
+      "step": 1354
+    },
+    {
+      "epoch": 13.155339805825243,
+      "grad_norm": 0.3637821275916941,
+      "learning_rate": 1.8426857470528414e-07,
+      "loss": 0.0646,
+      "step": 1355
+    },
+    {
+      "epoch": 13.16504854368932,
+      "grad_norm": 0.5236428238911527,
+      "learning_rate": 1.8235788107963948e-07,
+      "loss": 0.1017,
+      "step": 1356
+    },
+    {
+      "epoch": 13.174757281553399,
+      "grad_norm": 0.16816534480055276,
+      "learning_rate": 1.8045677017850595e-07,
+      "loss": 0.0062,
+      "step": 1357
+    },
+    {
+      "epoch": 13.184466019417476,
+      "grad_norm": 0.31536745363842367,
+      "learning_rate": 1.785652498623816e-07,
+      "loss": 0.0448,
+      "step": 1358
+    },
+    {
+      "epoch": 13.194174757281553,
+      "grad_norm": 0.5779882605598496,
+      "learning_rate": 1.7668332795211074e-07,
+      "loss": 0.0749,
+      "step": 1359
+    },
+    {
+      "epoch": 13.20388349514563,
+      "grad_norm": 0.2541991656984927,
+      "learning_rate": 1.7481101222885126e-07,
+      "loss": 0.0157,
+      "step": 1360
+    },
+    {
+      "epoch": 13.21359223300971,
+      "grad_norm": 0.3365028596139971,
+      "learning_rate": 1.7294831043404264e-07,
+      "loss": 0.0388,
+      "step": 1361
+    },
+    {
+      "epoch": 13.223300970873787,
+      "grad_norm": 0.48615199324448427,
+      "learning_rate": 1.7109523026937302e-07,
+      "loss": 0.0977,
+      "step": 1362
+    },
+    {
+      "epoch": 13.233009708737864,
+      "grad_norm": 0.4739890157105656,
+      "learning_rate": 1.6925177939674936e-07,
+      "loss": 0.0525,
+      "step": 1363
+    },
+    {
+      "epoch": 13.242718446601941,
+      "grad_norm": 0.5274812968875578,
+      "learning_rate": 1.6741796543826321e-07,
+      "loss": 0.0833,
+      "step": 1364
+    },
+    {
+      "epoch": 13.25242718446602,
+      "grad_norm": 0.5190394218310534,
+      "learning_rate": 1.6559379597616136e-07,
+      "loss": 0.0458,
+      "step": 1365
+    },
+    {
+      "epoch": 13.262135922330097,
+      "grad_norm": 0.5907235640627251,
+      "learning_rate": 1.6377927855281362e-07,
+      "loss": 0.1027,
+      "step": 1366
+    },
+    {
+      "epoch": 13.271844660194175,
+      "grad_norm": 0.5009182999514356,
+      "learning_rate": 1.6197442067068136e-07,
+      "loss": 0.0922,
+      "step": 1367
+    },
+    {
+      "epoch": 13.281553398058252,
+      "grad_norm": 0.2994761075980326,
+      "learning_rate": 1.6017922979228662e-07,
+      "loss": 0.0286,
+      "step": 1368
+    },
+    {
+      "epoch": 13.29126213592233,
+      "grad_norm": 0.4167528295855181,
+      "learning_rate": 1.5839371334018193e-07,
+      "loss": 0.048,
+      "step": 1369
+    },
+    {
+      "epoch": 13.300970873786408,
+      "grad_norm": 0.568993358454377,
+      "learning_rate": 1.5661787869691858e-07,
+      "loss": 0.0872,
+      "step": 1370
+    },
+    {
+      "epoch": 13.310679611650485,
+      "grad_norm": 0.6080611595176447,
+      "learning_rate": 1.5485173320501673e-07,
+      "loss": 0.0818,
+      "step": 1371
+    },
+    {
+      "epoch": 13.320388349514563,
+      "grad_norm": 0.4176513366927112,
+      "learning_rate": 1.5309528416693503e-07,
+      "loss": 0.0689,
+      "step": 1372
+    },
+    {
+      "epoch": 13.330097087378642,
+      "grad_norm": 0.38487885624551355,
+      "learning_rate": 1.513485388450403e-07,
+      "loss": 0.0369,
+      "step": 1373
+    },
+    {
+      "epoch": 13.339805825242719,
+      "grad_norm": 0.3808252984699954,
+      "learning_rate": 1.4961150446157759e-07,
+      "loss": 0.0523,
+      "step": 1374
+    },
+    {
+      "epoch": 13.349514563106796,
+      "grad_norm": 0.43465823014950594,
+      "learning_rate": 1.4788418819864037e-07,
+      "loss": 0.0417,
+      "step": 1375
+    },
+    {
+      "epoch": 13.359223300970873,
+      "grad_norm": 0.6908558471861086,
+      "learning_rate": 1.461665971981402e-07,
+      "loss": 0.1031,
+      "step": 1376
+    },
+    {
+      "epoch": 13.368932038834952,
+      "grad_norm": 0.22220724572717085,
+      "learning_rate": 1.444587385617785e-07,
+      "loss": 0.0108,
+      "step": 1377
+    },
+    {
+      "epoch": 13.37864077669903,
+      "grad_norm": 0.6187017283324233,
+      "learning_rate": 1.4276061935101586e-07,
+      "loss": 0.0797,
+      "step": 1378
+    },
+    {
+      "epoch": 13.388349514563107,
+      "grad_norm": 0.35673209366579156,
+      "learning_rate": 1.4107224658704288e-07,
+      "loss": 0.0278,
+      "step": 1379
+    },
+    {
+      "epoch": 13.398058252427184,
+      "grad_norm": 0.2088978860127839,
+      "learning_rate": 1.3939362725075344e-07,
+      "loss": 0.0096,
+      "step": 1380
+    },
+    {
+      "epoch": 13.407766990291263,
+      "grad_norm": 0.5737431049471658,
+      "learning_rate": 1.3772476828271236e-07,
+      "loss": 0.146,
+      "step": 1381
+    },
+    {
+      "epoch": 13.41747572815534,
+      "grad_norm": 0.5013268774854202,
+      "learning_rate": 1.360656765831289e-07,
+      "loss": 0.0851,
+      "step": 1382
+    },
+    {
+      "epoch": 13.427184466019417,
+      "grad_norm": 0.3552375372327559,
+      "learning_rate": 1.3441635901182803e-07,
+      "loss": 0.0332,
+      "step": 1383
+    },
+    {
+      "epoch": 13.436893203883495,
+      "grad_norm": 0.4105442051073008,
+      "learning_rate": 1.3277682238822142e-07,
+      "loss": 0.0427,
+      "step": 1384
+    },
+    {
+      "epoch": 13.446601941747574,
+      "grad_norm": 0.43977774703292793,
+      "learning_rate": 1.3114707349127954e-07,
+      "loss": 0.0495,
+      "step": 1385
+    },
+    {
+      "epoch": 13.45631067961165,
+      "grad_norm": 0.42583520614554343,
+      "learning_rate": 1.2952711905950377e-07,
+      "loss": 0.046,
+      "step": 1386
+    },
+    {
+      "epoch": 13.466019417475728,
+      "grad_norm": 0.4654975107942523,
+      "learning_rate": 1.279169657908988e-07,
+      "loss": 0.0662,
+      "step": 1387
+    },
+    {
+      "epoch": 13.475728155339805,
+      "grad_norm": 0.316054598646831,
+      "learning_rate": 1.263166203429439e-07,
+      "loss": 0.0362,
+      "step": 1388
+    },
+    {
+      "epoch": 13.485436893203884,
+      "grad_norm": 0.44344850299453137,
+      "learning_rate": 1.2472608933256637e-07,
+      "loss": 0.0266,
+      "step": 1389
+    },
+    {
+      "epoch": 13.495145631067961,
+      "grad_norm": 0.5686958760472558,
+      "learning_rate": 1.2314537933611425e-07,
+      "loss": 0.0526,
+      "step": 1390
+    },
+    {
+      "epoch": 13.504854368932039,
+      "grad_norm": 0.3714309782597695,
+      "learning_rate": 1.2157449688932872e-07,
+      "loss": 0.051,
+      "step": 1391
+    },
+    {
+      "epoch": 13.514563106796116,
+      "grad_norm": 0.43841752239263004,
+      "learning_rate": 1.2001344848731612e-07,
+      "loss": 0.0416,
+      "step": 1392
+    },
+    {
+      "epoch": 13.524271844660195,
+      "grad_norm": 0.27187015113253077,
+      "learning_rate": 1.1846224058452316e-07,
+      "loss": 0.0286,
+      "step": 1393
+    },
+    {
+      "epoch": 13.533980582524272,
+      "grad_norm": 0.36456437143328213,
+      "learning_rate": 1.1692087959470882e-07,
+      "loss": 0.0362,
+      "step": 1394
+    },
+    {
+      "epoch": 13.54368932038835,
+      "grad_norm": 0.5055952692825199,
+      "learning_rate": 1.1538937189091825e-07,
+      "loss": 0.0755,
+      "step": 1395
+    },
+    {
+      "epoch": 13.553398058252426,
+      "grad_norm": 0.3439908449532863,
+      "learning_rate": 1.1386772380545669e-07,
+      "loss": 0.0434,
+      "step": 1396
+    },
+    {
+      "epoch": 13.563106796116505,
+      "grad_norm": 0.3739766241535124,
+      "learning_rate": 1.1235594162986168e-07,
+      "loss": 0.0494,
+      "step": 1397
+    },
+    {
+      "epoch": 13.572815533980583,
+      "grad_norm": 0.568010631241099,
+      "learning_rate": 1.1085403161488012e-07,
+      "loss": 0.1044,
+      "step": 1398
+    },
+    {
+      "epoch": 13.58252427184466,
+      "grad_norm": 0.17116162913172464,
+      "learning_rate": 1.09361999970439e-07,
+      "loss": 0.0067,
+      "step": 1399
+    },
+    {
+      "epoch": 13.592233009708737,
+      "grad_norm": 0.3321360792703964,
+      "learning_rate": 1.0787985286562219e-07,
+      "loss": 0.0363,
+      "step": 1400
+    },
+    {
+      "epoch": 13.601941747572816,
+      "grad_norm": 0.47803315897026527,
+      "learning_rate": 1.0640759642864401e-07,
+      "loss": 0.0494,
+      "step": 1401
+    },
+    {
+      "epoch": 13.611650485436893,
+      "grad_norm": 0.5771243658845532,
+      "learning_rate": 1.0494523674682372e-07,
+      "loss": 0.0916,
+      "step": 1402
+    },
+    {
+      "epoch": 13.62135922330097,
+      "grad_norm": 0.5059429937647754,
+      "learning_rate": 1.0349277986656081e-07,
+      "loss": 0.0721,
+      "step": 1403
+    },
+    {
+      "epoch": 13.631067961165048,
+      "grad_norm": 0.24550599941178855,
+      "learning_rate": 1.0205023179330975e-07,
+      "loss": 0.0214,
+      "step": 1404
+    },
+    {
+      "epoch": 13.640776699029127,
+      "grad_norm": 0.6067893059454117,
+      "learning_rate": 1.00617598491555e-07,
+      "loss": 0.1443,
+      "step": 1405
+    },
+    {
+      "epoch": 13.650485436893204,
+      "grad_norm": 0.3674398742489763,
+      "learning_rate": 9.919488588478715e-08,
+      "loss": 0.0268,
+      "step": 1406
+    },
+    {
+      "epoch": 13.660194174757281,
+      "grad_norm": 0.5471451347636245,
+      "learning_rate": 9.778209985547682e-08,
+      "loss": 0.0937,
+      "step": 1407
+    },
+    {
+      "epoch": 13.669902912621358,
+      "grad_norm": 0.40045155813855254,
+      "learning_rate": 9.637924624505191e-08,
+      "loss": 0.0511,
+      "step": 1408
+    },
+    {
+      "epoch": 13.679611650485437,
+      "grad_norm": 0.3763845426376623,
+      "learning_rate": 9.498633085387343e-08,
+      "loss": 0.0407,
+      "step": 1409
+    },
+    {
+      "epoch": 13.689320388349515,
+      "grad_norm": 0.3101617655771819,
+      "learning_rate": 9.360335944121029e-08,
+      "loss": 0.0322,
+      "step": 1410
+    },
+    {
+      "epoch": 13.699029126213592,
+      "grad_norm": 0.3418703008016437,
+      "learning_rate": 9.223033772521594e-08,
+      "loss": 0.0257,
+      "step": 1411
+    },
+    {
+      "epoch": 13.70873786407767,
+      "grad_norm": 0.44907257872282763,
+      "learning_rate": 9.086727138290535e-08,
+      "loss": 0.0514,
+      "step": 1412
+    },
+    {
+      "epoch": 13.718446601941748,
+      "grad_norm": 0.3842567901939408,
+      "learning_rate": 8.951416605013114e-08,
+      "loss": 0.0429,
+      "step": 1413
+    },
+    {
+      "epoch": 13.728155339805825,
+      "grad_norm": 0.2966384701948387,
+      "learning_rate": 8.817102732155996e-08,
+      "loss": 0.0245,
+      "step": 1414
+    },
+    {
+      "epoch": 13.737864077669903,
+      "grad_norm": 0.5763576401759796,
+      "learning_rate": 8.683786075065065e-08,
+      "loss": 0.0675,
+      "step": 1415
+    },
+    {
+      "epoch": 13.74757281553398,
+      "grad_norm": 0.5002820834330115,
+      "learning_rate": 8.55146718496283e-08,
+      "loss": 0.0994,
+      "step": 1416
+    },
+    {
+      "epoch": 13.757281553398059,
+      "grad_norm": 0.42837675868305464,
+      "learning_rate": 8.420146608946605e-08,
+      "loss": 0.0633,
+      "step": 1417
+    },
+    {
+      "epoch": 13.766990291262136,
+      "grad_norm": 0.35423588956574104,
+      "learning_rate": 8.28982488998581e-08,
+      "loss": 0.0263,
+      "step": 1418
+    },
+    {
+      "epoch": 13.776699029126213,
+      "grad_norm": 0.5133064138909689,
+      "learning_rate": 8.160502566919942e-08,
+      "loss": 0.08,
+      "step": 1419
+    },
+    {
+      "epoch": 13.78640776699029,
+      "grad_norm": 0.42893161447189443,
+      "learning_rate": 8.032180174456283e-08,
+      "loss": 0.0423,
+      "step": 1420
+    },
+    {
+      "epoch": 13.79611650485437,
+      "grad_norm": 0.3124843126356028,
+      "learning_rate": 7.904858243167806e-08,
+      "loss": 0.0408,
+      "step": 1421
+    },
+    {
+      "epoch": 13.805825242718447,
+      "grad_norm": 0.4065666684492401,
+      "learning_rate": 7.778537299490796e-08,
+      "loss": 0.0637,
+      "step": 1422
+    },
+    {
+      "epoch": 13.815533980582524,
+      "grad_norm": 0.4457701399914273,
+      "learning_rate": 7.653217865722817e-08,
+      "loss": 0.0678,
+      "step": 1423
+    },
+    {
+      "epoch": 13.825242718446601,
+      "grad_norm": 0.3569532201060994,
+      "learning_rate": 7.528900460020444e-08,
+      "loss": 0.0383,
+      "step": 1424
+    },
+    {
+      "epoch": 13.83495145631068,
+      "grad_norm": 0.46545111674165585,
+      "learning_rate": 7.405585596397314e-08,
+      "loss": 0.0559,
+      "step": 1425
+    },
+    {
+      "epoch": 13.844660194174757,
+      "grad_norm": 0.5259399357202312,
+      "learning_rate": 7.283273784721739e-08,
+      "loss": 0.1231,
+      "step": 1426
+    },
+    {
+      "epoch": 13.854368932038835,
+      "grad_norm": 0.3144082807257404,
+      "learning_rate": 7.161965530714743e-08,
+      "loss": 0.0257,
+      "step": 1427
+    },
+    {
+      "epoch": 13.864077669902912,
+      "grad_norm": 0.4046497084629934,
+      "learning_rate": 7.041661335948024e-08,
+      "loss": 0.037,
+      "step": 1428
+    },
+    {
+      "epoch": 13.87378640776699,
+      "grad_norm": 0.6242338725168409,
+      "learning_rate": 6.92236169784169e-08,
+      "loss": 0.0617,
+      "step": 1429
+    },
+    {
+      "epoch": 13.883495145631068,
+      "grad_norm": 0.4927064101341263,
+      "learning_rate": 6.804067109662443e-08,
+      "loss": 0.0615,
+      "step": 1430
+    },
+    {
+      "epoch": 13.893203883495145,
+      "grad_norm": 0.4815006421883216,
+      "learning_rate": 6.68677806052137e-08,
+      "loss": 0.0757,
+      "step": 1431
+    },
+    {
+      "epoch": 13.902912621359224,
+      "grad_norm": 0.38866205557135447,
+      "learning_rate": 6.57049503537191e-08,
+      "loss": 0.0523,
+      "step": 1432
+    },
+    {
+      "epoch": 13.912621359223301,
+      "grad_norm": 0.45644868532380084,
+      "learning_rate": 6.455218515008049e-08,
+      "loss": 0.0538,
+      "step": 1433
+    },
+    {
+      "epoch": 13.922330097087379,
+      "grad_norm": 0.6472398646291064,
+      "learning_rate": 6.340948976062023e-08,
+      "loss": 0.0763,
+      "step": 1434
+    },
+    {
+      "epoch": 13.932038834951456,
+      "grad_norm": 0.6919248428459206,
+      "learning_rate": 6.227686891002671e-08,
+      "loss": 0.1206,
+      "step": 1435
+    },
+    {
+      "epoch": 13.941747572815533,
+      "grad_norm": 0.5771683769037051,
+      "learning_rate": 6.115432728133198e-08,
+      "loss": 0.0981,
+      "step": 1436
+    },
+    {
+      "epoch": 13.951456310679612,
+      "grad_norm": 0.4918068717506031,
+      "learning_rate": 6.004186951589414e-08,
+      "loss": 0.0886,
+      "step": 1437
+    },
+    {
+      "epoch": 13.96116504854369,
+      "grad_norm": 0.18476223002745681,
+      "learning_rate": 5.8939500213378296e-08,
+      "loss": 0.0061,
+      "step": 1438
+    },
+    {
+      "epoch": 13.970873786407767,
+      "grad_norm": 0.5073076174865154,
+      "learning_rate": 5.7847223931735974e-08,
+      "loss": 0.0667,
+      "step": 1439
+    },
+    {
+      "epoch": 13.980582524271846,
+      "grad_norm": 0.4571146557509417,
+      "learning_rate": 5.6765045187187614e-08,
+      "loss": 0.0726,
+      "step": 1440
+    },
+    {
+      "epoch": 13.990291262135923,
+      "grad_norm": 0.5620996547576711,
+      "learning_rate": 5.569296845420375e-08,
+      "loss": 0.0942,
+      "step": 1441
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.37701892224712136,
+      "learning_rate": 5.463099816548578e-08,
+      "loss": 0.0525,
+      "step": 1442
+    },
+    {
+      "epoch": 14.009708737864077,
+      "grad_norm": 0.27852037202811797,
+      "learning_rate": 5.3579138711948587e-08,
+      "loss": 0.0142,
+      "step": 1443
+    },
+    {
+      "epoch": 14.019417475728156,
+      "grad_norm": 0.5002120793072816,
+      "learning_rate": 5.253739444270128e-08,
+      "loss": 0.0811,
+      "step": 1444
+    },
+    {
+      "epoch": 14.029126213592233,
+      "grad_norm": 0.4954344423768613,
+      "learning_rate": 5.150576966503063e-08,
+      "loss": 0.0648,
+      "step": 1445
+    },
+    {
+      "epoch": 14.03883495145631,
+      "grad_norm": 0.8070824057658714,
+      "learning_rate": 5.048426864438183e-08,
+      "loss": 0.1415,
+      "step": 1446
+    },
+    {
+      "epoch": 14.048543689320388,
+      "grad_norm": 0.2485266305298399,
+      "learning_rate": 4.9472895604341655e-08,
+      "loss": 0.0173,
+      "step": 1447
+    },
+    {
+      "epoch": 14.058252427184467,
+      "grad_norm": 0.5002219465557189,
+      "learning_rate": 4.8471654726621464e-08,
+      "loss": 0.1107,
+      "step": 1448
+    },
+    {
+      "epoch": 14.067961165048544,
+      "grad_norm": 0.48859206182415127,
+      "learning_rate": 4.7480550151038365e-08,
+      "loss": 0.0607,
+      "step": 1449
+    },
+    {
+      "epoch": 14.077669902912621,
+      "grad_norm": 0.4401242445739508,
+      "learning_rate": 4.649958597549964e-08,
+      "loss": 0.0418,
+      "step": 1450
+    },
+    {
+      "epoch": 14.087378640776699,
+      "grad_norm": 0.5063356236988111,
+      "learning_rate": 4.552876625598501e-08,
+      "loss": 0.0795,
+      "step": 1451
+    },
+    {
+      "epoch": 14.097087378640778,
+      "grad_norm": 0.46075177267272516,
+      "learning_rate": 4.4568095006529975e-08,
+      "loss": 0.0687,
+      "step": 1452
+    },
+    {
+      "epoch": 14.106796116504855,
+      "grad_norm": 0.2271681561635824,
+      "learning_rate": 4.361757619920942e-08,
+      "loss": 0.0095,
+      "step": 1453
+    },
+    {
+      "epoch": 14.116504854368932,
+      "grad_norm": 0.22316895376169477,
+      "learning_rate": 4.2677213764120986e-08,
+      "loss": 0.0104,
+      "step": 1454
+    },
+    {
+      "epoch": 14.12621359223301,
+      "grad_norm": 0.40179318592167806,
+      "learning_rate": 4.174701158936895e-08,
+      "loss": 0.0273,
+      "step": 1455
+    },
+    {
+      "epoch": 14.135922330097088,
+      "grad_norm": 0.4980320711294089,
+      "learning_rate": 4.082697352104814e-08,
+      "loss": 0.0628,
+      "step": 1456
+    },
+    {
+      "epoch": 14.145631067961165,
+      "grad_norm": 0.443797410910131,
+      "learning_rate": 3.991710336322757e-08,
+      "loss": 0.0529,
+      "step": 1457
+    },
+    {
+      "epoch": 14.155339805825243,
+      "grad_norm": 0.4147350657878042,
+      "learning_rate": 3.9017404877935986e-08,
+      "loss": 0.0422,
+      "step": 1458
+    },
+    {
+      "epoch": 14.16504854368932,
+      "grad_norm": 0.6320743323439815,
+      "learning_rate": 3.812788178514437e-08,
+      "loss": 0.1207,
+      "step": 1459
+    },
+    {
+      "epoch": 14.174757281553399,
+      "grad_norm": 0.2851768686756557,
+      "learning_rate": 3.7248537762752666e-08,
+      "loss": 0.0156,
+      "step": 1460
+    },
+    {
+      "epoch": 14.184466019417476,
+      "grad_norm": 0.4898600075169955,
+      "learning_rate": 3.637937644657308e-08,
+      "loss": 0.0609,
+      "step": 1461
+    },
+    {
+      "epoch": 14.194174757281553,
+      "grad_norm": 0.6112591760337616,
+      "learning_rate": 3.55204014303151e-08,
+      "loss": 0.1028,
+      "step": 1462
+    },
+    {
+      "epoch": 14.20388349514563,
+      "grad_norm": 0.5361184519943515,
+      "learning_rate": 3.467161626557164e-08,
+      "loss": 0.1553,
+      "step": 1463
+    },
+    {
+      "epoch": 14.21359223300971,
+      "grad_norm": 0.3749624794722601,
+      "learning_rate": 3.3833024461803756e-08,
+      "loss": 0.0425,
+      "step": 1464
+    },
+    {
+      "epoch": 14.223300970873787,
+      "grad_norm": 0.3687071096568528,
+      "learning_rate": 3.300462948632593e-08,
+      "loss": 0.0528,
+      "step": 1465
+    },
+    {
+      "epoch": 14.233009708737864,
+      "grad_norm": 0.5065834596804688,
+      "learning_rate": 3.218643476429167e-08,
+      "loss": 0.1048,
+      "step": 1466
+    },
+    {
+      "epoch": 14.242718446601941,
+      "grad_norm": 0.2498819248286375,
+      "learning_rate": 3.1378443678680706e-08,
+      "loss": 0.0256,
+      "step": 1467
+    },
+    {
+      "epoch": 14.25242718446602,
+      "grad_norm": 0.4454945892622417,
+      "learning_rate": 3.0580659570282886e-08,
+      "loss": 0.0552,
+      "step": 1468
+    },
+    {
+      "epoch": 14.262135922330097,
+      "grad_norm": 0.46476529726448046,
+      "learning_rate": 2.979308573768547e-08,
+      "loss": 0.0736,
+      "step": 1469
+    },
+    {
+      "epoch": 14.271844660194175,
+      "grad_norm": 0.3614514478408299,
+      "learning_rate": 2.9015725437259724e-08,
+      "loss": 0.0444,
+      "step": 1470
+    },
+    {
+      "epoch": 14.281553398058252,
+      "grad_norm": 0.23856721195733988,
+      "learning_rate": 2.8248581883147387e-08,
+      "loss": 0.0134,
+      "step": 1471
+    },
+    {
+      "epoch": 14.29126213592233,
+      "grad_norm": 0.4462682057295648,
+      "learning_rate": 2.7491658247246478e-08,
+      "loss": 0.0674,
+      "step": 1472
+    },
+    {
+      "epoch": 14.300970873786408,
+      "grad_norm": 0.4735946001772513,
+      "learning_rate": 2.6744957659199376e-08,
+      "loss": 0.0662,
+      "step": 1473
+    },
+    {
+      "epoch": 14.310679611650485,
+      "grad_norm": 0.401590281873634,
+      "learning_rate": 2.6008483206379497e-08,
+      "loss": 0.0656,
+      "step": 1474
+    },
+    {
+      "epoch": 14.320388349514563,
+      "grad_norm": 0.5538260679883152,
+      "learning_rate": 2.5282237933877962e-08,
+      "loss": 0.0609,
+      "step": 1475
+    },
+    {
+      "epoch": 14.330097087378642,
+      "grad_norm": 0.36741254932703427,
+      "learning_rate": 2.4566224844491393e-08,
+      "loss": 0.0389,
+      "step": 1476
+    },
+    {
+      "epoch": 14.339805825242719,
+      "grad_norm": 0.47615607691212797,
+      "learning_rate": 2.38604468987097e-08,
+      "loss": 0.0685,
+      "step": 1477
+    },
+    {
+      "epoch": 14.349514563106796,
+      "grad_norm": 0.2774257621570567,
+      "learning_rate": 2.316490701470414e-08,
+      "loss": 0.0174,
+      "step": 1478
+    },
+    {
+      "epoch": 14.359223300970873,
+      "grad_norm": 0.4715962885247446,
+      "learning_rate": 2.247960806831373e-08,
+      "loss": 0.0725,
+      "step": 1479
+    },
+    {
+      "epoch": 14.368932038834952,
+      "grad_norm": 0.41298032350196945,
+      "learning_rate": 2.180455289303579e-08,
+      "loss": 0.0657,
+      "step": 1480
+    },
+    {
+      "epoch": 14.37864077669903,
+      "grad_norm": 0.4392490725579392,
+      "learning_rate": 2.113974428001153e-08,
+      "loss": 0.0629,
+      "step": 1481
+    },
+    {
+      "epoch": 14.388349514563107,
+      "grad_norm": 0.33456818961908225,
+      "learning_rate": 2.0485184978016604e-08,
+      "loss": 0.0247,
+      "step": 1482
+    },
+    {
+      "epoch": 14.398058252427184,
+      "grad_norm": 0.4210642043933347,
+      "learning_rate": 1.984087769344889e-08,
+      "loss": 0.0578,
+      "step": 1483
+    },
+    {
+      "epoch": 14.407766990291263,
+      "grad_norm": 0.4547431665479732,
+      "learning_rate": 1.9206825090317126e-08,
+      "loss": 0.0644,
+      "step": 1484
+    },
+    {
+      "epoch": 14.41747572815534,
+      "grad_norm": 0.4137452093457974,
+      "learning_rate": 1.8583029790230356e-08,
+      "loss": 0.0684,
+      "step": 1485
+    },
+    {
+      "epoch": 14.427184466019417,
+      "grad_norm": 0.5768594770321711,
+      "learning_rate": 1.796949437238682e-08,
+      "loss": 0.096,
+      "step": 1486
+    },
+    {
+      "epoch": 14.436893203883495,
+      "grad_norm": 0.5188349954764778,
+      "learning_rate": 1.736622137356342e-08,
+      "loss": 0.0814,
+      "step": 1487
+    },
+    {
+      "epoch": 14.446601941747574,
+      "grad_norm": 0.3045994084525141,
+      "learning_rate": 1.677321328810516e-08,
+      "loss": 0.0114,
+      "step": 1488
+    },
+    {
+      "epoch": 14.45631067961165,
+      "grad_norm": 0.42995889830896394,
+      "learning_rate": 1.6190472567914617e-08,
+      "loss": 0.0794,
+      "step": 1489
+    },
+    {
+      "epoch": 14.466019417475728,
+      "grad_norm": 0.44410933161737604,
+      "learning_rate": 1.561800162244248e-08,
+      "loss": 0.0538,
+      "step": 1490
+    },
+    {
+      "epoch": 14.475728155339805,
+      "grad_norm": 0.5574574290398137,
+      "learning_rate": 1.5055802818676745e-08,
+      "loss": 0.1046,
+      "step": 1491
+    },
+    {
+      "epoch": 14.485436893203884,
+      "grad_norm": 0.33854090541431436,
+      "learning_rate": 1.450387848113327e-08,
+      "loss": 0.0422,
+      "step": 1492
+    },
+    {
+      "epoch": 14.495145631067961,
+      "grad_norm": 0.3988422398212394,
+      "learning_rate": 1.3962230891846618e-08,
+      "loss": 0.043,
+      "step": 1493
+    },
+    {
+      "epoch": 14.504854368932039,
+      "grad_norm": 0.49523592862180527,
+      "learning_rate": 1.3430862290359781e-08,
+      "loss": 0.0687,
+      "step": 1494
+    },
+    {
+      "epoch": 14.514563106796116,
+      "grad_norm": 0.5839454104438518,
+      "learning_rate": 1.2909774873715585e-08,
+      "loss": 0.1464,
+      "step": 1495
+    },
+    {
+      "epoch": 14.524271844660195,
+      "grad_norm": 0.5374831629152403,
+      "learning_rate": 1.2398970796447807e-08,
+      "loss": 0.0745,
+      "step": 1496
+    },
+    {
+      "epoch": 14.533980582524272,
+      "grad_norm": 0.3781775887894092,
+      "learning_rate": 1.1898452170570618e-08,
+      "loss": 0.072,
+      "step": 1497
+    },
+    {
+      "epoch": 14.54368932038835,
+      "grad_norm": 0.2968854268436536,
+      "learning_rate": 1.140822106557249e-08,
+      "loss": 0.0205,
+      "step": 1498
+    },
+    {
+      "epoch": 14.553398058252426,
+      "grad_norm": 0.23045972300528747,
+      "learning_rate": 1.0928279508405082e-08,
+      "loss": 0.011,
+      "step": 1499
+    },
+    {
+      "epoch": 14.563106796116505,
+      "grad_norm": 0.4716850829610772,
+      "learning_rate": 1.0458629483476868e-08,
+      "loss": 0.0827,
+      "step": 1500
+    },
+    {
+      "epoch": 14.572815533980583,
+      "grad_norm": 0.4419550519402443,
+      "learning_rate": 9.999272932643134e-09,
+      "loss": 0.0346,
+      "step": 1501
+    },
+    {
+      "epoch": 14.58252427184466,
+      "grad_norm": 0.8818612325287888,
+      "learning_rate": 9.550211755199879e-09,
+      "loss": 0.0911,
+      "step": 1502
+    },
+    {
+      "epoch": 14.592233009708737,
+      "grad_norm": 0.5946985929189472,
+      "learning_rate": 9.111447807874374e-09,
+      "loss": 0.0219,
+      "step": 1503
+    },
+    {
+      "epoch": 14.601941747572816,
+      "grad_norm": 0.10701016268942182,
+      "learning_rate": 8.682982904817948e-09,
+      "loss": 0.0028,
+      "step": 1504
+    },
+    {
+      "epoch": 14.611650485436893,
+      "grad_norm": 0.6231038615943904,
+      "learning_rate": 8.264818817599052e-09,
+      "loss": 0.1465,
+      "step": 1505
+    },
+    {
+      "epoch": 14.62135922330097,
+      "grad_norm": 0.32493446734913417,
+      "learning_rate": 7.856957275194921e-09,
+      "loss": 0.0239,
+      "step": 1506
+    },
+    {
+      "epoch": 14.631067961165048,
+      "grad_norm": 0.30621128115752955,
+      "learning_rate": 7.459399963985758e-09,
+      "loss": 0.0219,
+      "step": 1507
+    },
+    {
+      "epoch": 14.640776699029127,
+      "grad_norm": 0.5681331507606439,
+      "learning_rate": 7.072148527746403e-09,
+      "loss": 0.0612,
+      "step": 1508
+    },
+    {
+      "epoch": 14.650485436893204,
+      "grad_norm": 0.4678809532098972,
+      "learning_rate": 6.6952045676405005e-09,
+      "loss": 0.0734,
+      "step": 1509
+    },
+    {
+      "epoch": 14.660194174757281,
+      "grad_norm": 0.42191270347256216,
+      "learning_rate": 6.328569642212734e-09,
+      "loss": 0.042,
+      "step": 1510
+    },
+    {
+      "epoch": 14.669902912621358,
+      "grad_norm": 0.6089098976587644,
+      "learning_rate": 5.972245267384102e-09,
+      "loss": 0.1034,
+      "step": 1511
+    },
+    {
+      "epoch": 14.679611650485437,
+      "grad_norm": 0.6347427788645178,
+      "learning_rate": 5.62623291644443e-09,
+      "loss": 0.0419,
+      "step": 1512
+    },
+    {
+      "epoch": 14.689320388349515,
+      "grad_norm": 0.21563629066961087,
+      "learning_rate": 5.290534020046256e-09,
+      "loss": 0.0139,
+      "step": 1513
+    },
+    {
+      "epoch": 14.699029126213592,
+      "grad_norm": 0.5066555604026627,
+      "learning_rate": 4.965149966199567e-09,
+      "loss": 0.046,
+      "step": 1514
+    },
+    {
+      "epoch": 14.70873786407767,
+      "grad_norm": 0.40413935811547114,
+      "learning_rate": 4.6500821002654075e-09,
+      "loss": 0.0652,
+      "step": 1515
+    },
+    {
+      "epoch": 14.718446601941748,
+      "grad_norm": 0.5986384312203215,
+      "learning_rate": 4.345331724950885e-09,
+      "loss": 0.0543,
+      "step": 1516
+    },
+    {
+      "epoch": 14.728155339805825,
+      "grad_norm": 0.4450958848994586,
+      "learning_rate": 4.050900100303068e-09,
+      "loss": 0.0459,
+      "step": 1517
+    },
+    {
+      "epoch": 14.737864077669903,
+      "grad_norm": 0.3924298698415564,
+      "learning_rate": 3.766788443705094e-09,
+      "loss": 0.0361,
+      "step": 1518
+    },
+    {
+      "epoch": 14.74757281553398,
+      "grad_norm": 0.35509867963290737,
+      "learning_rate": 3.492997929869235e-09,
+      "loss": 0.0375,
+      "step": 1519
+    },
+    {
+      "epoch": 14.757281553398059,
+      "grad_norm": 0.32188506207739537,
+      "learning_rate": 3.2295296908338437e-09,
+      "loss": 0.0254,
+      "step": 1520
+    },
+    {
+      "epoch": 14.766990291262136,
+      "grad_norm": 0.3947000336245188,
+      "learning_rate": 2.976384815957245e-09,
+      "loss": 0.03,
+      "step": 1521
+    },
+    {
+      "epoch": 14.776699029126213,
+      "grad_norm": 0.4224276282074339,
+      "learning_rate": 2.7335643519144086e-09,
+      "loss": 0.0383,
+      "step": 1522
+    },
+    {
+      "epoch": 14.78640776699029,
+      "grad_norm": 0.5089503895148835,
+      "learning_rate": 2.5010693026922273e-09,
+      "loss": 0.1007,
+      "step": 1523
+    },
+    {
+      "epoch": 14.79611650485437,
+      "grad_norm": 0.18020715011955366,
+      "learning_rate": 2.278900629584524e-09,
+      "loss": 0.0066,
+      "step": 1524
+    },
+    {
+      "epoch": 14.805825242718447,
+      "grad_norm": 0.6643910364446401,
+      "learning_rate": 2.067059251189274e-09,
+      "loss": 0.1364,
+      "step": 1525
+    },
+    {
+      "epoch": 14.815533980582524,
+      "grad_norm": 0.14798425026090903,
+      "learning_rate": 1.8655460434044427e-09,
+      "loss": 0.0044,
+      "step": 1526
+    },
+    {
+      "epoch": 14.825242718446601,
+      "grad_norm": 0.5631198394596968,
+      "learning_rate": 1.6743618394238215e-09,
+      "loss": 0.0981,
+      "step": 1527
+    },
+    {
+      "epoch": 14.83495145631068,
+      "grad_norm": 0.45409258209552955,
+      "learning_rate": 1.493507429734531e-09,
+      "loss": 0.0465,
+      "step": 1528
+    },
+    {
+      "epoch": 14.844660194174757,
+      "grad_norm": 0.42087852066764153,
+      "learning_rate": 1.3229835621125786e-09,
+      "loss": 0.0783,
+      "step": 1529
+    },
+    {
+      "epoch": 14.854368932038835,
+      "grad_norm": 0.36289372361154304,
+      "learning_rate": 1.1627909416211947e-09,
+      "loss": 0.0504,
+      "step": 1530
+    },
+    {
+      "epoch": 14.864077669902912,
+      "grad_norm": 0.35522890887727715,
+      "learning_rate": 1.0129302306061128e-09,
+      "loss": 0.0414,
+      "step": 1531
+    },
+    {
+      "epoch": 14.87378640776699,
+      "grad_norm": 0.4579376664607033,
+      "learning_rate": 8.734020486950157e-10,
+      "loss": 0.0432,
+      "step": 1532
+    },
+    {
+      "epoch": 14.883495145631068,
+      "grad_norm": 0.13707830522982126,
+      "learning_rate": 7.442069727930934e-10,
+      "loss": 0.0048,
+      "step": 1533
+    },
+    {
+      "epoch": 14.893203883495145,
+      "grad_norm": 0.4002252720770533,
+      "learning_rate": 6.253455370811012e-10,
+      "loss": 0.0371,
+      "step": 1534
+    },
+    {
+      "epoch": 14.902912621359224,
+      "grad_norm": 0.3147483793059573,
+      "learning_rate": 5.168182330145266e-10,
+      "loss": 0.0295,
+      "step": 1535
+    },
+    {
+      "epoch": 14.912621359223301,
+      "grad_norm": 0.3919323958341479,
+      "learning_rate": 4.186255093194258e-10,
+      "loss": 0.0575,
+      "step": 1536
+    },
+    {
+      "epoch": 14.922330097087379,
+      "grad_norm": 0.39816059669182213,
+      "learning_rate": 3.3076777199186894e-10,
+      "loss": 0.0455,
+      "step": 1537
+    },
+    {
+      "epoch": 14.932038834951456,
+      "grad_norm": 0.3752851863862166,
+      "learning_rate": 2.532453842965521e-10,
+      "loss": 0.0503,
+      "step": 1538
+    },
+    {
+      "epoch": 14.941747572815533,
+      "grad_norm": 0.35630558512507376,
+      "learning_rate": 1.8605866676374428e-10,
+      "loss": 0.0323,
+      "step": 1539
+    },
+    {
+      "epoch": 14.951456310679612,
+      "grad_norm": 0.39981192379176855,
+      "learning_rate": 1.292078971898425e-10,
+      "loss": 0.0875,
+      "step": 1540
+    },
+    {
+      "epoch": 14.96116504854369,
+      "grad_norm": 0.06618155040773933,
+      "learning_rate": 8.269331063459618e-11,
+      "loss": 0.0014,
+      "step": 1541
+    },
+    {
+      "epoch": 14.970873786407767,
+      "grad_norm": 0.6044643732379602,
+      "learning_rate": 4.651509942193988e-11,
+      "loss": 0.0481,
+      "step": 1542
+    },
+    {
+      "epoch": 14.980582524271846,
+      "grad_norm": 0.4479730592528947,
+      "learning_rate": 2.06734131366626e-11,
+      "loss": 0.0538,
+      "step": 1543
+    },
+    {
+      "epoch": 14.990291262135923,
+      "grad_norm": 0.3639756177331702,
+      "learning_rate": 5.168358626628234e-12,
+      "loss": 0.037,
+      "step": 1544
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.26971244188279725,
+      "learning_rate": 0.0,
+      "loss": 0.0139,
+      "step": 1545
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 171438419165184.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}