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": 4.52614688873291,
+      "learning_rate": 4.999994831641374e-06,
+      "loss": 0.8521,
+      "step": 1
+    },
+    {
+      "epoch": 0.019417475728155338,
+      "grad_norm": 3.6663663387298584,
+      "learning_rate": 4.9999793265868636e-06,
+      "loss": 0.6915,
+      "step": 2
+    },
+    {
+      "epoch": 0.02912621359223301,
+      "grad_norm": 2.7436909675598145,
+      "learning_rate": 4.999953484900578e-06,
+      "loss": 0.8366,
+      "step": 3
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 2.088653326034546,
+      "learning_rate": 4.9999173066893655e-06,
+      "loss": 0.7118,
+      "step": 4
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 2.199723958969116,
+      "learning_rate": 4.9998707921028104e-06,
+      "loss": 0.7858,
+      "step": 5
+    },
+    {
+      "epoch": 0.05825242718446602,
+      "grad_norm": 2.0426740646362305,
+      "learning_rate": 4.999813941333237e-06,
+      "loss": 0.6171,
+      "step": 6
+    },
+    {
+      "epoch": 0.06796116504854369,
+      "grad_norm": 1.950857400894165,
+      "learning_rate": 4.999746754615704e-06,
+      "loss": 0.7283,
+      "step": 7
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.5054230690002441,
+      "learning_rate": 4.9996692322280085e-06,
+      "loss": 0.6389,
+      "step": 8
+    },
+    {
+      "epoch": 0.08737864077669903,
+      "grad_norm": 1.304260015487671,
+      "learning_rate": 4.999581374490681e-06,
+      "loss": 0.6234,
+      "step": 9
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 1.4494102001190186,
+      "learning_rate": 4.999483181766986e-06,
+      "loss": 0.8027,
+      "step": 10
+    },
+    {
+      "epoch": 0.10679611650485436,
+      "grad_norm": 1.3682693243026733,
+      "learning_rate": 4.999374654462919e-06,
+      "loss": 0.7446,
+      "step": 11
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.1007122993469238,
+      "learning_rate": 4.999255793027207e-06,
+      "loss": 0.7415,
+      "step": 12
+    },
+    {
+      "epoch": 0.1262135922330097,
+      "grad_norm": 1.0817819833755493,
+      "learning_rate": 4.999126597951305e-06,
+      "loss": 0.5366,
+      "step": 13
+    },
+    {
+      "epoch": 0.13592233009708737,
+      "grad_norm": 1.2209179401397705,
+      "learning_rate": 4.998987069769394e-06,
+      "loss": 0.5151,
+      "step": 14
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 1.3821749687194824,
+      "learning_rate": 4.998837209058379e-06,
+      "loss": 0.6989,
+      "step": 15
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 1.2181178331375122,
+      "learning_rate": 4.998677016437888e-06,
+      "loss": 0.5743,
+      "step": 16
+    },
+    {
+      "epoch": 0.1650485436893204,
+      "grad_norm": 1.045208215713501,
+      "learning_rate": 4.998506492570266e-06,
+      "loss": 0.4859,
+      "step": 17
+    },
+    {
+      "epoch": 0.17475728155339806,
+      "grad_norm": 1.3972703218460083,
+      "learning_rate": 4.998325638160576e-06,
+      "loss": 0.7708,
+      "step": 18
+    },
+    {
+      "epoch": 0.18446601941747573,
+      "grad_norm": 1.2529926300048828,
+      "learning_rate": 4.998134453956596e-06,
+      "loss": 0.5568,
+      "step": 19
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.3563785552978516,
+      "learning_rate": 4.997932940748811e-06,
+      "loss": 0.6435,
+      "step": 20
+    },
+    {
+      "epoch": 0.20388349514563106,
+      "grad_norm": 1.1624324321746826,
+      "learning_rate": 4.997721099370416e-06,
+      "loss": 0.6125,
+      "step": 21
+    },
+    {
+      "epoch": 0.21359223300970873,
+      "grad_norm": 1.2726877927780151,
+      "learning_rate": 4.997498930697308e-06,
+      "loss": 0.5921,
+      "step": 22
+    },
+    {
+      "epoch": 0.22330097087378642,
+      "grad_norm": 1.0864430665969849,
+      "learning_rate": 4.997266435648086e-06,
+      "loss": 0.5279,
+      "step": 23
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.2319843769073486,
+      "learning_rate": 4.997023615184044e-06,
+      "loss": 0.799,
+      "step": 24
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 1.12433922290802,
+      "learning_rate": 4.996770470309167e-06,
+      "loss": 0.7428,
+      "step": 25
+    },
+    {
+      "epoch": 0.2524271844660194,
+      "grad_norm": 1.1892931461334229,
+      "learning_rate": 4.996507002070131e-06,
+      "loss": 0.6064,
+      "step": 26
+    },
+    {
+      "epoch": 0.2621359223300971,
+      "grad_norm": 0.9294205904006958,
+      "learning_rate": 4.996233211556295e-06,
+      "loss": 0.6833,
+      "step": 27
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.1600544452667236,
+      "learning_rate": 4.9959490998996974e-06,
+      "loss": 0.4822,
+      "step": 28
+    },
+    {
+      "epoch": 0.2815533980582524,
+      "grad_norm": 0.8306522965431213,
+      "learning_rate": 4.995654668275049e-06,
+      "loss": 0.5528,
+      "step": 29
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.9367732405662537,
+      "learning_rate": 4.995349917899735e-06,
+      "loss": 0.4968,
+      "step": 30
+    },
+    {
+      "epoch": 0.30097087378640774,
+      "grad_norm": 0.9846139550209045,
+      "learning_rate": 4.9950348500338005e-06,
+      "loss": 0.4915,
+      "step": 31
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.137281894683838,
+      "learning_rate": 4.994709465979954e-06,
+      "loss": 0.6108,
+      "step": 32
+    },
+    {
+      "epoch": 0.32038834951456313,
+      "grad_norm": 0.9563445448875427,
+      "learning_rate": 4.994373767083556e-06,
+      "loss": 0.5569,
+      "step": 33
+    },
+    {
+      "epoch": 0.3300970873786408,
+      "grad_norm": 1.1258145570755005,
+      "learning_rate": 4.994027754732616e-06,
+      "loss": 0.6422,
+      "step": 34
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 1.071385145187378,
+      "learning_rate": 4.993671430357788e-06,
+      "loss": 0.5001,
+      "step": 35
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.0511752367019653,
+      "learning_rate": 4.99330479543236e-06,
+      "loss": 0.5523,
+      "step": 36
+    },
+    {
+      "epoch": 0.3592233009708738,
+      "grad_norm": 1.044432282447815,
+      "learning_rate": 4.992927851472254e-06,
+      "loss": 0.6826,
+      "step": 37
+    },
+    {
+      "epoch": 0.36893203883495146,
+      "grad_norm": 0.8882154226303101,
+      "learning_rate": 4.992540600036014e-06,
+      "loss": 0.7096,
+      "step": 38
+    },
+    {
+      "epoch": 0.3786407766990291,
+      "grad_norm": 1.0710127353668213,
+      "learning_rate": 4.992143042724805e-06,
+      "loss": 0.5487,
+      "step": 39
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 2.931645154953003,
+      "learning_rate": 4.991735181182401e-06,
+      "loss": 0.819,
+      "step": 40
+    },
+    {
+      "epoch": 0.39805825242718446,
+      "grad_norm": 1.0483241081237793,
+      "learning_rate": 4.991317017095182e-06,
+      "loss": 0.5693,
+      "step": 41
+    },
+    {
+      "epoch": 0.4077669902912621,
+      "grad_norm": 1.897265911102295,
+      "learning_rate": 4.990888552192126e-06,
+      "loss": 0.5371,
+      "step": 42
+    },
+    {
+      "epoch": 0.4174757281553398,
+      "grad_norm": 0.9810941815376282,
+      "learning_rate": 4.9904497882448004e-06,
+      "loss": 0.6299,
+      "step": 43
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 0.7764018774032593,
+      "learning_rate": 4.990000727067357e-06,
+      "loss": 0.5914,
+      "step": 44
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 0.7854965925216675,
+      "learning_rate": 4.989541370516523e-06,
+      "loss": 0.5044,
+      "step": 45
+    },
+    {
+      "epoch": 0.44660194174757284,
+      "grad_norm": 0.8319512009620667,
+      "learning_rate": 4.989071720491595e-06,
+      "loss": 0.4736,
+      "step": 46
+    },
+    {
+      "epoch": 0.4563106796116505,
+      "grad_norm": 0.9436545372009277,
+      "learning_rate": 4.988591778934428e-06,
+      "loss": 0.476,
+      "step": 47
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.3040553331375122,
+      "learning_rate": 4.9881015478294294e-06,
+      "loss": 0.5776,
+      "step": 48
+    },
+    {
+      "epoch": 0.47572815533980584,
+      "grad_norm": 0.933722198009491,
+      "learning_rate": 4.987601029203553e-06,
+      "loss": 0.4888,
+      "step": 49
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 0.858695387840271,
+      "learning_rate": 4.987090225126285e-06,
+      "loss": 0.599,
+      "step": 50
+    },
+    {
+      "epoch": 0.49514563106796117,
+      "grad_norm": 0.8849862217903137,
+      "learning_rate": 4.98656913770964e-06,
+      "loss": 0.5087,
+      "step": 51
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 0.8477684855461121,
+      "learning_rate": 4.986037769108154e-06,
+      "loss": 0.5507,
+      "step": 52
+    },
+    {
+      "epoch": 0.5145631067961165,
+      "grad_norm": 1.0618562698364258,
+      "learning_rate": 4.9854961215188676e-06,
+      "loss": 0.7117,
+      "step": 53
+    },
+    {
+      "epoch": 0.5242718446601942,
+      "grad_norm": 0.9216580390930176,
+      "learning_rate": 4.984944197181324e-06,
+      "loss": 0.4642,
+      "step": 54
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 0.9276666641235352,
+      "learning_rate": 4.9843819983775575e-06,
+      "loss": 0.4878,
+      "step": 55
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 0.9354626536369324,
+      "learning_rate": 4.983809527432086e-06,
+      "loss": 0.4856,
+      "step": 56
+    },
+    {
+      "epoch": 0.5533980582524272,
+      "grad_norm": 1.0111173391342163,
+      "learning_rate": 4.983226786711895e-06,
+      "loss": 0.5896,
+      "step": 57
+    },
+    {
+      "epoch": 0.5631067961165048,
+      "grad_norm": 0.8906674385070801,
+      "learning_rate": 4.982633778626437e-06,
+      "loss": 0.5374,
+      "step": 58
+    },
+    {
+      "epoch": 0.5728155339805825,
+      "grad_norm": 0.7241902947425842,
+      "learning_rate": 4.982030505627613e-06,
+      "loss": 0.6278,
+      "step": 59
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.9223883748054504,
+      "learning_rate": 4.98141697020977e-06,
+      "loss": 0.5967,
+      "step": 60
+    },
+    {
+      "epoch": 0.5922330097087378,
+      "grad_norm": 0.9214841723442078,
+      "learning_rate": 4.9807931749096836e-06,
+      "loss": 0.4895,
+      "step": 61
+    },
+    {
+      "epoch": 0.6019417475728155,
+      "grad_norm": 1.0902329683303833,
+      "learning_rate": 4.980159122306551e-06,
+      "loss": 0.6806,
+      "step": 62
+    },
+    {
+      "epoch": 0.6116504854368932,
+      "grad_norm": 0.9778655171394348,
+      "learning_rate": 4.979514815021984e-06,
+      "loss": 0.7513,
+      "step": 63
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 0.9169423580169678,
+      "learning_rate": 4.978860255719989e-06,
+      "loss": 0.5485,
+      "step": 64
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.8950898051261902,
+      "learning_rate": 4.978195447106965e-06,
+      "loss": 0.69,
+      "step": 65
+    },
+    {
+      "epoch": 0.6407766990291263,
+      "grad_norm": 0.9813127517700195,
+      "learning_rate": 4.9775203919316864e-06,
+      "loss": 0.5714,
+      "step": 66
+    },
+    {
+      "epoch": 0.6504854368932039,
+      "grad_norm": 1.1932649612426758,
+      "learning_rate": 4.976835092985297e-06,
+      "loss": 0.7081,
+      "step": 67
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 0.8608429431915283,
+      "learning_rate": 4.976139553101291e-06,
+      "loss": 0.6091,
+      "step": 68
+    },
+    {
+      "epoch": 0.6699029126213593,
+      "grad_norm": 0.9685009717941284,
+      "learning_rate": 4.975433775155509e-06,
+      "loss": 0.5879,
+      "step": 69
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 0.72447669506073,
+      "learning_rate": 4.974717762066123e-06,
+      "loss": 0.5094,
+      "step": 70
+    },
+    {
+      "epoch": 0.6893203883495146,
+      "grad_norm": 1.0234659910202026,
+      "learning_rate": 4.973991516793621e-06,
+      "loss": 0.7861,
+      "step": 71
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 0.8732417225837708,
+      "learning_rate": 4.973255042340801e-06,
+      "loss": 0.5652,
+      "step": 72
+    },
+    {
+      "epoch": 0.7087378640776699,
+      "grad_norm": 0.9726948142051697,
+      "learning_rate": 4.972508341752754e-06,
+      "loss": 0.7384,
+      "step": 73
+    },
+    {
+      "epoch": 0.7184466019417476,
+      "grad_norm": 0.9769085645675659,
+      "learning_rate": 4.9717514181168534e-06,
+      "loss": 0.5375,
+      "step": 74
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.8254098296165466,
+      "learning_rate": 4.970984274562741e-06,
+      "loss": 0.448,
+      "step": 75
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 1.040366530418396,
+      "learning_rate": 4.970206914262315e-06,
+      "loss": 0.6201,
+      "step": 76
+    },
+    {
+      "epoch": 0.7475728155339806,
+      "grad_norm": 0.928465723991394,
+      "learning_rate": 4.969419340429717e-06,
+      "loss": 0.5439,
+      "step": 77
+    },
+    {
+      "epoch": 0.7572815533980582,
+      "grad_norm": 0.8173317909240723,
+      "learning_rate": 4.968621556321319e-06,
+      "loss": 0.5596,
+      "step": 78
+    },
+    {
+      "epoch": 0.7669902912621359,
+      "grad_norm": 1.272718906402588,
+      "learning_rate": 4.967813565235708e-06,
+      "loss": 0.653,
+      "step": 79
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.8894286751747131,
+      "learning_rate": 4.966995370513675e-06,
+      "loss": 0.4878,
+      "step": 80
+    },
+    {
+      "epoch": 0.7864077669902912,
+      "grad_norm": 1.0618493556976318,
+      "learning_rate": 4.966166975538197e-06,
+      "loss": 0.7623,
+      "step": 81
+    },
+    {
+      "epoch": 0.7961165048543689,
+      "grad_norm": 0.9374388456344604,
+      "learning_rate": 4.965328383734429e-06,
+      "loss": 0.5505,
+      "step": 82
+    },
+    {
+      "epoch": 0.8058252427184466,
+      "grad_norm": 1.0057796239852905,
+      "learning_rate": 4.964479598569686e-06,
+      "loss": 0.7241,
+      "step": 83
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.9318037033081055,
+      "learning_rate": 4.963620623553428e-06,
+      "loss": 0.6787,
+      "step": 84
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.8819167017936707,
+      "learning_rate": 4.962751462237248e-06,
+      "loss": 0.5103,
+      "step": 85
+    },
+    {
+      "epoch": 0.8349514563106796,
+      "grad_norm": 1.0126394033432007,
+      "learning_rate": 4.9618721182148564e-06,
+      "loss": 0.4538,
+      "step": 86
+    },
+    {
+      "epoch": 0.8446601941747572,
+      "grad_norm": 0.9401935338973999,
+      "learning_rate": 4.960982595122064e-06,
+      "loss": 0.4892,
+      "step": 87
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 0.875956118106842,
+      "learning_rate": 4.960082896636773e-06,
+      "loss": 0.6734,
+      "step": 88
+    },
+    {
+      "epoch": 0.8640776699029126,
+      "grad_norm": 0.9892510175704956,
+      "learning_rate": 4.959173026478952e-06,
+      "loss": 0.5805,
+      "step": 89
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.8165844082832336,
+      "learning_rate": 4.958252988410631e-06,
+      "loss": 0.4292,
+      "step": 90
+    },
+    {
+      "epoch": 0.883495145631068,
+      "grad_norm": 1.2144172191619873,
+      "learning_rate": 4.9573227862358794e-06,
+      "loss": 0.5701,
+      "step": 91
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 0.8784387111663818,
+      "learning_rate": 4.956382423800791e-06,
+      "loss": 0.5766,
+      "step": 92
+    },
+    {
+      "epoch": 0.9029126213592233,
+      "grad_norm": 0.8603959679603577,
+      "learning_rate": 4.955431904993471e-06,
+      "loss": 0.6135,
+      "step": 93
+    },
+    {
+      "epoch": 0.912621359223301,
+      "grad_norm": 0.9591665863990784,
+      "learning_rate": 4.954471233744015e-06,
+      "loss": 0.4231,
+      "step": 94
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.9828278422355652,
+      "learning_rate": 4.9535004140245005e-06,
+      "loss": 0.6176,
+      "step": 95
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 0.9605256915092468,
+      "learning_rate": 4.952519449848962e-06,
+      "loss": 0.5469,
+      "step": 96
+    },
+    {
+      "epoch": 0.941747572815534,
+      "grad_norm": 1.0379314422607422,
+      "learning_rate": 4.951528345273379e-06,
+      "loss": 0.5562,
+      "step": 97
+    },
+    {
+      "epoch": 0.9514563106796117,
+      "grad_norm": 0.9734949469566345,
+      "learning_rate": 4.950527104395659e-06,
+      "loss": 0.6546,
+      "step": 98
+    },
+    {
+      "epoch": 0.9611650485436893,
+      "grad_norm": 1.0504772663116455,
+      "learning_rate": 4.9495157313556185e-06,
+      "loss": 0.4437,
+      "step": 99
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.8973089456558228,
+      "learning_rate": 4.94849423033497e-06,
+      "loss": 0.3843,
+      "step": 100
+    },
+    {
+      "epoch": 0.9805825242718447,
+      "grad_norm": 0.9390032887458801,
+      "learning_rate": 4.9474626055573e-06,
+      "loss": 0.7418,
+      "step": 101
+    },
+    {
+      "epoch": 0.9902912621359223,
+      "grad_norm": 0.830512523651123,
+      "learning_rate": 4.946420861288051e-06,
+      "loss": 0.5576,
+      "step": 102
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9454949498176575,
+      "learning_rate": 4.9453690018345144e-06,
+      "loss": 0.5272,
+      "step": 103
+    },
+    {
+      "epoch": 1.0097087378640777,
+      "grad_norm": 1.0955694913864136,
+      "learning_rate": 4.944307031545797e-06,
+      "loss": 0.4085,
+      "step": 104
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 1.0182733535766602,
+      "learning_rate": 4.943234954812812e-06,
+      "loss": 0.6185,
+      "step": 105
+    },
+    {
+      "epoch": 1.029126213592233,
+      "grad_norm": 1.0509144067764282,
+      "learning_rate": 4.942152776068264e-06,
+      "loss": 0.4346,
+      "step": 106
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 1.108575701713562,
+      "learning_rate": 4.941060499786622e-06,
+      "loss": 0.6482,
+      "step": 107
+    },
+    {
+      "epoch": 1.0485436893203883,
+      "grad_norm": 1.0695301294326782,
+      "learning_rate": 4.939958130484106e-06,
+      "loss": 0.4688,
+      "step": 108
+    },
+    {
+      "epoch": 1.058252427184466,
+      "grad_norm": 0.8110963106155396,
+      "learning_rate": 4.938845672718668e-06,
+      "loss": 0.4636,
+      "step": 109
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 1.0377169847488403,
+      "learning_rate": 4.937723131089974e-06,
+      "loss": 0.5277,
+      "step": 110
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 1.179370641708374,
+      "learning_rate": 4.93659051023938e-06,
+      "loss": 0.5084,
+      "step": 111
+    },
+    {
+      "epoch": 1.087378640776699,
+      "grad_norm": 0.7915495038032532,
+      "learning_rate": 4.93544781484992e-06,
+      "loss": 0.3111,
+      "step": 112
+    },
+    {
+      "epoch": 1.0970873786407767,
+      "grad_norm": 0.8695587515830994,
+      "learning_rate": 4.9342950496462815e-06,
+      "loss": 0.5161,
+      "step": 113
+    },
+    {
+      "epoch": 1.1067961165048543,
+      "grad_norm": 1.138805866241455,
+      "learning_rate": 4.933132219394786e-06,
+      "loss": 0.5075,
+      "step": 114
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.8836705088615417,
+      "learning_rate": 4.931959328903376e-06,
+      "loss": 0.639,
+      "step": 115
+    },
+    {
+      "epoch": 1.1262135922330097,
+      "grad_norm": 0.8428300023078918,
+      "learning_rate": 4.930776383021584e-06,
+      "loss": 0.4566,
+      "step": 116
+    },
+    {
+      "epoch": 1.1359223300970873,
+      "grad_norm": 1.0511326789855957,
+      "learning_rate": 4.92958338664052e-06,
+      "loss": 0.3212,
+      "step": 117
+    },
+    {
+      "epoch": 1.145631067961165,
+      "grad_norm": 0.7950966954231262,
+      "learning_rate": 4.928380344692853e-06,
+      "loss": 0.3876,
+      "step": 118
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 0.8359726071357727,
+      "learning_rate": 4.927167262152784e-06,
+      "loss": 0.491,
+      "step": 119
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.8725942969322205,
+      "learning_rate": 4.925944144036027e-06,
+      "loss": 0.4993,
+      "step": 120
+    },
+    {
+      "epoch": 1.174757281553398,
+      "grad_norm": 0.8538187742233276,
+      "learning_rate": 4.924710995399796e-06,
+      "loss": 0.4621,
+      "step": 121
+    },
+    {
+      "epoch": 1.1844660194174756,
+      "grad_norm": 0.8445570468902588,
+      "learning_rate": 4.923467821342773e-06,
+      "loss": 0.4652,
+      "step": 122
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 0.8798047304153442,
+      "learning_rate": 4.922214627005092e-06,
+      "loss": 0.5119,
+      "step": 123
+    },
+    {
+      "epoch": 1.203883495145631,
+      "grad_norm": 0.9884085059165955,
+      "learning_rate": 4.920951417568323e-06,
+      "loss": 0.5963,
+      "step": 124
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.7976876497268677,
+      "learning_rate": 4.919678198255438e-06,
+      "loss": 0.409,
+      "step": 125
+    },
+    {
+      "epoch": 1.2233009708737863,
+      "grad_norm": 0.8497043251991272,
+      "learning_rate": 4.918394974330801e-06,
+      "loss": 0.5069,
+      "step": 126
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 0.9168796539306641,
+      "learning_rate": 4.917101751100142e-06,
+      "loss": 0.4699,
+      "step": 127
+    },
+    {
+      "epoch": 1.2427184466019416,
+      "grad_norm": 0.9214611053466797,
+      "learning_rate": 4.915798533910534e-06,
+      "loss": 0.4605,
+      "step": 128
+    },
+    {
+      "epoch": 1.2524271844660193,
+      "grad_norm": 0.916888415813446,
+      "learning_rate": 4.9144853281503715e-06,
+      "loss": 0.3928,
+      "step": 129
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 1.035357117652893,
+      "learning_rate": 4.91316213924935e-06,
+      "loss": 0.5672,
+      "step": 130
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 0.8245042562484741,
+      "learning_rate": 4.911828972678441e-06,
+      "loss": 0.3609,
+      "step": 131
+    },
+    {
+      "epoch": 1.2815533980582523,
+      "grad_norm": 0.9270608425140381,
+      "learning_rate": 4.91048583394987e-06,
+      "loss": 0.4302,
+      "step": 132
+    },
+    {
+      "epoch": 1.29126213592233,
+      "grad_norm": 0.8770717978477478,
+      "learning_rate": 4.909132728617095e-06,
+      "loss": 0.7062,
+      "step": 133
+    },
+    {
+      "epoch": 1.3009708737864076,
+      "grad_norm": 0.9357777833938599,
+      "learning_rate": 4.907769662274785e-06,
+      "loss": 0.4277,
+      "step": 134
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.8497308492660522,
+      "learning_rate": 4.90639664055879e-06,
+      "loss": 0.3775,
+      "step": 135
+    },
+    {
+      "epoch": 1.3203883495145632,
+      "grad_norm": 0.8059908151626587,
+      "learning_rate": 4.905013669146127e-06,
+      "loss": 0.4194,
+      "step": 136
+    },
+    {
+      "epoch": 1.3300970873786409,
+      "grad_norm": 0.9206295609474182,
+      "learning_rate": 4.903620753754949e-06,
+      "loss": 0.5536,
+      "step": 137
+    },
+    {
+      "epoch": 1.3398058252427185,
+      "grad_norm": 1.0086661577224731,
+      "learning_rate": 4.902217900144524e-06,
+      "loss": 0.7059,
+      "step": 138
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 0.829765260219574,
+      "learning_rate": 4.900805114115214e-06,
+      "loss": 0.6443,
+      "step": 139
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 1.0548204183578491,
+      "learning_rate": 4.899382401508446e-06,
+      "loss": 0.3785,
+      "step": 140
+    },
+    {
+      "epoch": 1.3689320388349515,
+      "grad_norm": 0.8123111128807068,
+      "learning_rate": 4.8979497682066916e-06,
+      "loss": 0.5454,
+      "step": 141
+    },
+    {
+      "epoch": 1.3786407766990292,
+      "grad_norm": 0.7998213171958923,
+      "learning_rate": 4.89650722013344e-06,
+      "loss": 0.3798,
+      "step": 142
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 0.7438819408416748,
+      "learning_rate": 4.895054763253177e-06,
+      "loss": 0.4741,
+      "step": 143
+    },
+    {
+      "epoch": 1.3980582524271845,
+      "grad_norm": 0.9747439622879028,
+      "learning_rate": 4.8935924035713564e-06,
+      "loss": 0.398,
+      "step": 144
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.8595287799835205,
+      "learning_rate": 4.892120147134378e-06,
+      "loss": 0.4111,
+      "step": 145
+    },
+    {
+      "epoch": 1.4174757281553398,
+      "grad_norm": 1.0879790782928467,
+      "learning_rate": 4.8906380000295615e-06,
+      "loss": 0.5177,
+      "step": 146
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 0.8619869947433472,
+      "learning_rate": 4.889145968385121e-06,
+      "loss": 0.4312,
+      "step": 147
+    },
+    {
+      "epoch": 1.4368932038834952,
+      "grad_norm": 0.8725226521492004,
+      "learning_rate": 4.887644058370139e-06,
+      "loss": 0.3597,
+      "step": 148
+    },
+    {
+      "epoch": 1.4466019417475728,
+      "grad_norm": 0.913958728313446,
+      "learning_rate": 4.886132276194544e-06,
+      "loss": 0.3836,
+      "step": 149
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.9644704461097717,
+      "learning_rate": 4.884610628109082e-06,
+      "loss": 0.3876,
+      "step": 150
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 0.7304561138153076,
+      "learning_rate": 4.883079120405292e-06,
+      "loss": 0.401,
+      "step": 151
+    },
+    {
+      "epoch": 1.4757281553398058,
+      "grad_norm": 0.9006849527359009,
+      "learning_rate": 4.881537759415478e-06,
+      "loss": 0.3514,
+      "step": 152
+    },
+    {
+      "epoch": 1.4854368932038835,
+      "grad_norm": 0.9095068573951721,
+      "learning_rate": 4.879986551512684e-06,
+      "loss": 0.5258,
+      "step": 153
+    },
+    {
+      "epoch": 1.4951456310679612,
+      "grad_norm": 0.8439248204231262,
+      "learning_rate": 4.878425503110672e-06,
+      "loss": 0.4044,
+      "step": 154
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.9092299938201904,
+      "learning_rate": 4.876854620663887e-06,
+      "loss": 0.4692,
+      "step": 155
+    },
+    {
+      "epoch": 1.5145631067961165,
+      "grad_norm": 1.0393599271774292,
+      "learning_rate": 4.875273910667434e-06,
+      "loss": 0.4612,
+      "step": 156
+    },
+    {
+      "epoch": 1.5242718446601942,
+      "grad_norm": 0.9718154072761536,
+      "learning_rate": 4.873683379657057e-06,
+      "loss": 0.3271,
+      "step": 157
+    },
+    {
+      "epoch": 1.5339805825242718,
+      "grad_norm": 0.8826969265937805,
+      "learning_rate": 4.8720830342091015e-06,
+      "loss": 0.719,
+      "step": 158
+    },
+    {
+      "epoch": 1.5436893203883495,
+      "grad_norm": 0.8973050713539124,
+      "learning_rate": 4.870472880940496e-06,
+      "loss": 0.5879,
+      "step": 159
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 1.1856048107147217,
+      "learning_rate": 4.868852926508721e-06,
+      "loss": 0.6377,
+      "step": 160
+    },
+    {
+      "epoch": 1.5631067961165048,
+      "grad_norm": 1.0391895771026611,
+      "learning_rate": 4.867223177611779e-06,
+      "loss": 0.5097,
+      "step": 161
+    },
+    {
+      "epoch": 1.5728155339805825,
+      "grad_norm": 1.1057053804397583,
+      "learning_rate": 4.865583640988173e-06,
+      "loss": 0.3714,
+      "step": 162
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 0.9357965588569641,
+      "learning_rate": 4.863934323416871e-06,
+      "loss": 0.4876,
+      "step": 163
+    },
+    {
+      "epoch": 1.5922330097087378,
+      "grad_norm": 0.8261438012123108,
+      "learning_rate": 4.862275231717288e-06,
+      "loss": 0.4145,
+      "step": 164
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 1.001351237297058,
+      "learning_rate": 4.860606372749247e-06,
+      "loss": 0.3615,
+      "step": 165
+    },
+    {
+      "epoch": 1.6116504854368932,
+      "grad_norm": 0.8855088353157043,
+      "learning_rate": 4.858927753412958e-06,
+      "loss": 0.4166,
+      "step": 166
+    },
+    {
+      "epoch": 1.6213592233009708,
+      "grad_norm": 0.9390084743499756,
+      "learning_rate": 4.857239380648985e-06,
+      "loss": 0.4242,
+      "step": 167
+    },
+    {
+      "epoch": 1.6310679611650487,
+      "grad_norm": 0.86589115858078,
+      "learning_rate": 4.855541261438223e-06,
+      "loss": 0.5007,
+      "step": 168
+    },
+    {
+      "epoch": 1.6407766990291264,
+      "grad_norm": 0.8739226460456848,
+      "learning_rate": 4.8538334028018605e-06,
+      "loss": 0.5952,
+      "step": 169
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.759211540222168,
+      "learning_rate": 4.8521158118013605e-06,
+      "loss": 0.3322,
+      "step": 170
+    },
+    {
+      "epoch": 1.6601941747572817,
+      "grad_norm": 1.1307138204574585,
+      "learning_rate": 4.850388495538423e-06,
+      "loss": 0.4473,
+      "step": 171
+    },
+    {
+      "epoch": 1.6699029126213594,
+      "grad_norm": 1.050327181816101,
+      "learning_rate": 4.84865146115496e-06,
+      "loss": 0.4032,
+      "step": 172
+    },
+    {
+      "epoch": 1.679611650485437,
+      "grad_norm": 1.0714014768600464,
+      "learning_rate": 4.846904715833066e-06,
+      "loss": 0.3714,
+      "step": 173
+    },
+    {
+      "epoch": 1.6893203883495147,
+      "grad_norm": 0.7853191494941711,
+      "learning_rate": 4.8451482667949836e-06,
+      "loss": 0.5307,
+      "step": 174
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.9957652688026428,
+      "learning_rate": 4.843382121303082e-06,
+      "loss": 0.4776,
+      "step": 175
+    },
+    {
+      "epoch": 1.70873786407767,
+      "grad_norm": 0.888883650302887,
+      "learning_rate": 4.841606286659819e-06,
+      "loss": 0.3613,
+      "step": 176
+    },
+    {
+      "epoch": 1.7184466019417477,
+      "grad_norm": 0.9413757920265198,
+      "learning_rate": 4.839820770207714e-06,
+      "loss": 0.3656,
+      "step": 177
+    },
+    {
+      "epoch": 1.7281553398058254,
+      "grad_norm": 0.9671488404273987,
+      "learning_rate": 4.8380255793293195e-06,
+      "loss": 0.4002,
+      "step": 178
+    },
+    {
+      "epoch": 1.737864077669903,
+      "grad_norm": 1.2052301168441772,
+      "learning_rate": 4.8362207214471864e-06,
+      "loss": 0.3936,
+      "step": 179
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.9285777807235718,
+      "learning_rate": 4.83440620402384e-06,
+      "loss": 0.469,
+      "step": 180
+    },
+    {
+      "epoch": 1.7572815533980584,
+      "grad_norm": 0.8492085933685303,
+      "learning_rate": 4.832582034561738e-06,
+      "loss": 0.4749,
+      "step": 181
+    },
+    {
+      "epoch": 1.766990291262136,
+      "grad_norm": 0.8664201498031616,
+      "learning_rate": 4.830748220603251e-06,
+      "loss": 0.4987,
+      "step": 182
+    },
+    {
+      "epoch": 1.7766990291262137,
+      "grad_norm": 0.7961488366127014,
+      "learning_rate": 4.828904769730628e-06,
+      "loss": 0.5023,
+      "step": 183
+    },
+    {
+      "epoch": 1.7864077669902914,
+      "grad_norm": 0.9886943101882935,
+      "learning_rate": 4.827051689565958e-06,
+      "loss": 0.4413,
+      "step": 184
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.8723036646842957,
+      "learning_rate": 4.825188987771149e-06,
+      "loss": 0.5691,
+      "step": 185
+    },
+    {
+      "epoch": 1.8058252427184467,
+      "grad_norm": 0.9434221982955933,
+      "learning_rate": 4.82331667204789e-06,
+      "loss": 0.3126,
+      "step": 186
+    },
+    {
+      "epoch": 1.8155339805825244,
+      "grad_norm": 0.9722500443458557,
+      "learning_rate": 4.821434750137619e-06,
+      "loss": 0.4777,
+      "step": 187
+    },
+    {
+      "epoch": 1.825242718446602,
+      "grad_norm": 0.9892464876174927,
+      "learning_rate": 4.819543229821494e-06,
+      "loss": 0.535,
+      "step": 188
+    },
+    {
+      "epoch": 1.8349514563106797,
+      "grad_norm": 0.8184608221054077,
+      "learning_rate": 4.8176421189203605e-06,
+      "loss": 0.3786,
+      "step": 189
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.9847081899642944,
+      "learning_rate": 4.815731425294716e-06,
+      "loss": 0.495,
+      "step": 190
+    },
+    {
+      "epoch": 1.854368932038835,
+      "grad_norm": 0.8743010759353638,
+      "learning_rate": 4.813811156844681e-06,
+      "loss": 0.5254,
+      "step": 191
+    },
+    {
+      "epoch": 1.8640776699029127,
+      "grad_norm": 1.206464409828186,
+      "learning_rate": 4.811881321509964e-06,
+      "loss": 0.521,
+      "step": 192
+    },
+    {
+      "epoch": 1.8737864077669903,
+      "grad_norm": 0.8556103110313416,
+      "learning_rate": 4.809941927269829e-06,
+      "loss": 0.6493,
+      "step": 193
+    },
+    {
+      "epoch": 1.883495145631068,
+      "grad_norm": 1.0219848155975342,
+      "learning_rate": 4.807992982143064e-06,
+      "loss": 0.468,
+      "step": 194
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 1.0638519525527954,
+      "learning_rate": 4.806034494187949e-06,
+      "loss": 0.6301,
+      "step": 195
+    },
+    {
+      "epoch": 1.9029126213592233,
+      "grad_norm": 0.9077790379524231,
+      "learning_rate": 4.804066471502216e-06,
+      "loss": 0.419,
+      "step": 196
+    },
+    {
+      "epoch": 1.912621359223301,
+      "grad_norm": 0.9752828478813171,
+      "learning_rate": 4.802088922223024e-06,
+      "loss": 0.3641,
+      "step": 197
+    },
+    {
+      "epoch": 1.9223300970873787,
+      "grad_norm": 0.8168025016784668,
+      "learning_rate": 4.80010185452692e-06,
+      "loss": 0.4678,
+      "step": 198
+    },
+    {
+      "epoch": 1.9320388349514563,
+      "grad_norm": 0.9176221489906311,
+      "learning_rate": 4.798105276629806e-06,
+      "loss": 0.3585,
+      "step": 199
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.9507747888565063,
+      "learning_rate": 4.796099196786908e-06,
+      "loss": 0.3969,
+      "step": 200
+    },
+    {
+      "epoch": 1.9514563106796117,
+      "grad_norm": 0.7999218106269836,
+      "learning_rate": 4.794083623292737e-06,
+      "loss": 0.5478,
+      "step": 201
+    },
+    {
+      "epoch": 1.9611650485436893,
+      "grad_norm": 0.9281819462776184,
+      "learning_rate": 4.792058564481058e-06,
+      "loss": 0.3663,
+      "step": 202
+    },
+    {
+      "epoch": 1.970873786407767,
+      "grad_norm": 0.9587502479553223,
+      "learning_rate": 4.7900240287248554e-06,
+      "loss": 0.5232,
+      "step": 203
+    },
+    {
+      "epoch": 1.9805825242718447,
+      "grad_norm": 0.8954132795333862,
+      "learning_rate": 4.7879800244362975e-06,
+      "loss": 0.5567,
+      "step": 204
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.9513785243034363,
+      "learning_rate": 4.785926560066703e-06,
+      "loss": 0.555,
+      "step": 205
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8883495330810547,
+      "learning_rate": 4.783863644106502e-06,
+      "loss": 0.4269,
+      "step": 206
+    },
+    {
+      "epoch": 2.0097087378640777,
+      "grad_norm": 1.053060531616211,
+      "learning_rate": 4.781791285085209e-06,
+      "loss": 0.3394,
+      "step": 207
+    },
+    {
+      "epoch": 2.0194174757281553,
+      "grad_norm": 0.967287003993988,
+      "learning_rate": 4.779709491571378e-06,
+      "loss": 0.2951,
+      "step": 208
+    },
+    {
+      "epoch": 2.029126213592233,
+      "grad_norm": 0.8971940279006958,
+      "learning_rate": 4.777618272172573e-06,
+      "loss": 0.4388,
+      "step": 209
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.9578713178634644,
+      "learning_rate": 4.775517635535332e-06,
+      "loss": 0.4951,
+      "step": 210
+    },
+    {
+      "epoch": 2.0485436893203883,
+      "grad_norm": 1.1447572708129883,
+      "learning_rate": 4.77340759034513e-06,
+      "loss": 0.4791,
+      "step": 211
+    },
+    {
+      "epoch": 2.058252427184466,
+      "grad_norm": 1.332855224609375,
+      "learning_rate": 4.771288145326343e-06,
+      "loss": 0.4653,
+      "step": 212
+    },
+    {
+      "epoch": 2.0679611650485437,
+      "grad_norm": 1.055438756942749,
+      "learning_rate": 4.769159309242213e-06,
+      "loss": 0.2767,
+      "step": 213
+    },
+    {
+      "epoch": 2.0776699029126213,
+      "grad_norm": 1.3705360889434814,
+      "learning_rate": 4.767021090894809e-06,
+      "loss": 0.4402,
+      "step": 214
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 1.3517563343048096,
+      "learning_rate": 4.764873499124997e-06,
+      "loss": 0.4773,
+      "step": 215
+    },
+    {
+      "epoch": 2.0970873786407767,
+      "grad_norm": 1.0658684968948364,
+      "learning_rate": 4.762716542812395e-06,
+      "loss": 0.3238,
+      "step": 216
+    },
+    {
+      "epoch": 2.1067961165048543,
+      "grad_norm": 0.9824750423431396,
+      "learning_rate": 4.7605502308753415e-06,
+      "loss": 0.3833,
+      "step": 217
+    },
+    {
+      "epoch": 2.116504854368932,
+      "grad_norm": 1.1418805122375488,
+      "learning_rate": 4.758374572270859e-06,
+      "loss": 0.4576,
+      "step": 218
+    },
+    {
+      "epoch": 2.1262135922330097,
+      "grad_norm": 0.9111520051956177,
+      "learning_rate": 4.756189575994614e-06,
+      "loss": 0.1972,
+      "step": 219
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 1.0712579488754272,
+      "learning_rate": 4.753995251080884e-06,
+      "loss": 0.3815,
+      "step": 220
+    },
+    {
+      "epoch": 2.145631067961165,
+      "grad_norm": 0.7820471525192261,
+      "learning_rate": 4.7517916066025126e-06,
+      "loss": 0.2703,
+      "step": 221
+    },
+    {
+      "epoch": 2.1553398058252426,
+      "grad_norm": 0.9843001961708069,
+      "learning_rate": 4.7495786516708806e-06,
+      "loss": 0.4997,
+      "step": 222
+    },
+    {
+      "epoch": 2.1650485436893203,
+      "grad_norm": 1.0501731634140015,
+      "learning_rate": 4.747356395435865e-06,
+      "loss": 0.2681,
+      "step": 223
+    },
+    {
+      "epoch": 2.174757281553398,
+      "grad_norm": 1.0242283344268799,
+      "learning_rate": 4.745124847085799e-06,
+      "loss": 0.3817,
+      "step": 224
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 1.1368554830551147,
+      "learning_rate": 4.742884015847436e-06,
+      "loss": 0.3516,
+      "step": 225
+    },
+    {
+      "epoch": 2.1941747572815533,
+      "grad_norm": 1.0896131992340088,
+      "learning_rate": 4.740633910985911e-06,
+      "loss": 0.3989,
+      "step": 226
+    },
+    {
+      "epoch": 2.203883495145631,
+      "grad_norm": 0.8043588399887085,
+      "learning_rate": 4.738374541804704e-06,
+      "loss": 0.2405,
+      "step": 227
+    },
+    {
+      "epoch": 2.2135922330097086,
+      "grad_norm": 0.9075763821601868,
+      "learning_rate": 4.7361059176456e-06,
+      "loss": 0.3208,
+      "step": 228
+    },
+    {
+      "epoch": 2.2233009708737863,
+      "grad_norm": 0.9826246500015259,
+      "learning_rate": 4.733828047888647e-06,
+      "loss": 0.4524,
+      "step": 229
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 1.135237693786621,
+      "learning_rate": 4.731540941952126e-06,
+      "loss": 0.2596,
+      "step": 230
+    },
+    {
+      "epoch": 2.2427184466019416,
+      "grad_norm": 0.924315869808197,
+      "learning_rate": 4.7292446092925016e-06,
+      "loss": 0.3294,
+      "step": 231
+    },
+    {
+      "epoch": 2.2524271844660193,
+      "grad_norm": 0.9574539065361023,
+      "learning_rate": 4.726939059404392e-06,
+      "loss": 0.2563,
+      "step": 232
+    },
+    {
+      "epoch": 2.262135922330097,
+      "grad_norm": 0.8917362689971924,
+      "learning_rate": 4.724624301820524e-06,
+      "loss": 0.2961,
+      "step": 233
+    },
+    {
+      "epoch": 2.2718446601941746,
+      "grad_norm": 1.1570966243743896,
+      "learning_rate": 4.722300346111695e-06,
+      "loss": 0.4054,
+      "step": 234
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 1.1190950870513916,
+      "learning_rate": 4.719967201886734e-06,
+      "loss": 0.348,
+      "step": 235
+    },
+    {
+      "epoch": 2.29126213592233,
+      "grad_norm": 0.9346051216125488,
+      "learning_rate": 4.717624878792461e-06,
+      "loss": 0.516,
+      "step": 236
+    },
+    {
+      "epoch": 2.3009708737864076,
+      "grad_norm": 0.9378127455711365,
+      "learning_rate": 4.715273386513651e-06,
+      "loss": 0.4853,
+      "step": 237
+    },
+    {
+      "epoch": 2.3106796116504853,
+      "grad_norm": 1.0024279356002808,
+      "learning_rate": 4.712912734772988e-06,
+      "loss": 0.3455,
+      "step": 238
+    },
+    {
+      "epoch": 2.320388349514563,
+      "grad_norm": 0.7446038722991943,
+      "learning_rate": 4.710542933331025e-06,
+      "loss": 0.3764,
+      "step": 239
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.8997549414634705,
+      "learning_rate": 4.708163991986152e-06,
+      "loss": 0.4377,
+      "step": 240
+    },
+    {
+      "epoch": 2.3398058252427183,
+      "grad_norm": 0.877932071685791,
+      "learning_rate": 4.705775920574546e-06,
+      "loss": 0.4254,
+      "step": 241
+    },
+    {
+      "epoch": 2.349514563106796,
+      "grad_norm": 1.1586366891860962,
+      "learning_rate": 4.703378728970134e-06,
+      "loss": 0.2999,
+      "step": 242
+    },
+    {
+      "epoch": 2.3592233009708736,
+      "grad_norm": 0.8719814419746399,
+      "learning_rate": 4.700972427084551e-06,
+      "loss": 0.4169,
+      "step": 243
+    },
+    {
+      "epoch": 2.3689320388349513,
+      "grad_norm": 1.0403155088424683,
+      "learning_rate": 4.698557024867105e-06,
+      "loss": 0.2651,
+      "step": 244
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 1.027769684791565,
+      "learning_rate": 4.696132532304727e-06,
+      "loss": 0.4728,
+      "step": 245
+    },
+    {
+      "epoch": 2.3883495145631066,
+      "grad_norm": 1.0018792152404785,
+      "learning_rate": 4.693698959421935e-06,
+      "loss": 0.349,
+      "step": 246
+    },
+    {
+      "epoch": 2.3980582524271843,
+      "grad_norm": 1.0755680799484253,
+      "learning_rate": 4.691256316280789e-06,
+      "loss": 0.5781,
+      "step": 247
+    },
+    {
+      "epoch": 2.407766990291262,
+      "grad_norm": 0.8174302577972412,
+      "learning_rate": 4.688804612980855e-06,
+      "loss": 0.2627,
+      "step": 248
+    },
+    {
+      "epoch": 2.4174757281553396,
+      "grad_norm": 0.8010028600692749,
+      "learning_rate": 4.686343859659158e-06,
+      "loss": 0.4919,
+      "step": 249
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.9900104403495789,
+      "learning_rate": 4.683874066490143e-06,
+      "loss": 0.3611,
+      "step": 250
+    },
+    {
+      "epoch": 2.436893203883495,
+      "grad_norm": 1.2603962421417236,
+      "learning_rate": 4.681395243685631e-06,
+      "loss": 0.3744,
+      "step": 251
+    },
+    {
+      "epoch": 2.4466019417475726,
+      "grad_norm": 0.9518654346466064,
+      "learning_rate": 4.67890740149478e-06,
+      "loss": 0.345,
+      "step": 252
+    },
+    {
+      "epoch": 2.4563106796116507,
+      "grad_norm": 0.819825291633606,
+      "learning_rate": 4.676410550204036e-06,
+      "loss": 0.4169,
+      "step": 253
+    },
+    {
+      "epoch": 2.466019417475728,
+      "grad_norm": 0.8665376305580139,
+      "learning_rate": 4.673904700137098e-06,
+      "loss": 0.2292,
+      "step": 254
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.8893418908119202,
+      "learning_rate": 4.671389861654873e-06,
+      "loss": 0.304,
+      "step": 255
+    },
+    {
+      "epoch": 2.4854368932038833,
+      "grad_norm": 0.915643572807312,
+      "learning_rate": 4.668866045155428e-06,
+      "loss": 0.3982,
+      "step": 256
+    },
+    {
+      "epoch": 2.4951456310679614,
+      "grad_norm": 1.0549052953720093,
+      "learning_rate": 4.666333261073956e-06,
+      "loss": 0.4209,
+      "step": 257
+    },
+    {
+      "epoch": 2.5048543689320386,
+      "grad_norm": 1.2616932392120361,
+      "learning_rate": 4.6637915198827265e-06,
+      "loss": 0.4782,
+      "step": 258
+    },
+    {
+      "epoch": 2.5145631067961167,
+      "grad_norm": 1.1108514070510864,
+      "learning_rate": 4.661240832091042e-06,
+      "loss": 0.3456,
+      "step": 259
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.8557240962982178,
+      "learning_rate": 4.658681208245198e-06,
+      "loss": 0.386,
+      "step": 260
+    },
+    {
+      "epoch": 2.533980582524272,
+      "grad_norm": 0.8849290609359741,
+      "learning_rate": 4.65611265892844e-06,
+      "loss": 0.3983,
+      "step": 261
+    },
+    {
+      "epoch": 2.5436893203883493,
+      "grad_norm": 0.9907529950141907,
+      "learning_rate": 4.653535194760912e-06,
+      "loss": 0.366,
+      "step": 262
+    },
+    {
+      "epoch": 2.5533980582524274,
+      "grad_norm": 0.9792741537094116,
+      "learning_rate": 4.650948826399624e-06,
+      "loss": 0.2161,
+      "step": 263
+    },
+    {
+      "epoch": 2.5631067961165046,
+      "grad_norm": 0.9505898356437683,
+      "learning_rate": 4.648353564538397e-06,
+      "loss": 0.3018,
+      "step": 264
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 1.1814008951187134,
+      "learning_rate": 4.645749419907829e-06,
+      "loss": 0.3034,
+      "step": 265
+    },
+    {
+      "epoch": 2.58252427184466,
+      "grad_norm": 1.0968157052993774,
+      "learning_rate": 4.64313640327524e-06,
+      "loss": 0.2418,
+      "step": 266
+    },
+    {
+      "epoch": 2.592233009708738,
+      "grad_norm": 0.9502717852592468,
+      "learning_rate": 4.640514525444637e-06,
+      "loss": 0.4245,
+      "step": 267
+    },
+    {
+      "epoch": 2.6019417475728153,
+      "grad_norm": 1.1883552074432373,
+      "learning_rate": 4.637883797256663e-06,
+      "loss": 0.6322,
+      "step": 268
+    },
+    {
+      "epoch": 2.6116504854368934,
+      "grad_norm": 0.9291763305664062,
+      "learning_rate": 4.635244229588558e-06,
+      "loss": 0.3531,
+      "step": 269
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.68864905834198,
+      "learning_rate": 4.632595833354105e-06,
+      "loss": 0.3944,
+      "step": 270
+    },
+    {
+      "epoch": 2.6310679611650487,
+      "grad_norm": 1.281590223312378,
+      "learning_rate": 4.629938619503593e-06,
+      "loss": 0.3025,
+      "step": 271
+    },
+    {
+      "epoch": 2.6407766990291264,
+      "grad_norm": 1.1126227378845215,
+      "learning_rate": 4.627272599023772e-06,
+      "loss": 0.4944,
+      "step": 272
+    },
+    {
+      "epoch": 2.650485436893204,
+      "grad_norm": 0.91072678565979,
+      "learning_rate": 4.6245977829378e-06,
+      "loss": 0.5801,
+      "step": 273
+    },
+    {
+      "epoch": 2.6601941747572817,
+      "grad_norm": 0.9558573365211487,
+      "learning_rate": 4.6219141823052035e-06,
+      "loss": 0.2808,
+      "step": 274
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.9114917516708374,
+      "learning_rate": 4.619221808221833e-06,
+      "loss": 0.5259,
+      "step": 275
+    },
+    {
+      "epoch": 2.679611650485437,
+      "grad_norm": 0.9031375050544739,
+      "learning_rate": 4.616520671819812e-06,
+      "loss": 0.2877,
+      "step": 276
+    },
+    {
+      "epoch": 2.6893203883495147,
+      "grad_norm": 1.0576499700546265,
+      "learning_rate": 4.613810784267492e-06,
+      "loss": 0.42,
+      "step": 277
+    },
+    {
+      "epoch": 2.6990291262135924,
+      "grad_norm": 0.8092033267021179,
+      "learning_rate": 4.61109215676941e-06,
+      "loss": 0.2962,
+      "step": 278
+    },
+    {
+      "epoch": 2.70873786407767,
+      "grad_norm": 0.9271148443222046,
+      "learning_rate": 4.608364800566241e-06,
+      "loss": 0.4057,
+      "step": 279
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.7499092817306519,
+      "learning_rate": 4.605628726934747e-06,
+      "loss": 0.4567,
+      "step": 280
+    },
+    {
+      "epoch": 2.7281553398058254,
+      "grad_norm": 0.8146029710769653,
+      "learning_rate": 4.602883947187738e-06,
+      "loss": 0.4149,
+      "step": 281
+    },
+    {
+      "epoch": 2.737864077669903,
+      "grad_norm": 0.888672947883606,
+      "learning_rate": 4.600130472674017e-06,
+      "loss": 0.2856,
+      "step": 282
+    },
+    {
+      "epoch": 2.7475728155339807,
+      "grad_norm": 0.9286831021308899,
+      "learning_rate": 4.5973683147783405e-06,
+      "loss": 0.5477,
+      "step": 283
+    },
+    {
+      "epoch": 2.7572815533980584,
+      "grad_norm": 1.3980002403259277,
+      "learning_rate": 4.594597484921365e-06,
+      "loss": 0.3656,
+      "step": 284
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 1.0174404382705688,
+      "learning_rate": 4.5918179945596055e-06,
+      "loss": 0.3542,
+      "step": 285
+    },
+    {
+      "epoch": 2.7766990291262137,
+      "grad_norm": 0.9232505559921265,
+      "learning_rate": 4.589029855185384e-06,
+      "loss": 0.3179,
+      "step": 286
+    },
+    {
+      "epoch": 2.7864077669902914,
+      "grad_norm": 1.0175881385803223,
+      "learning_rate": 4.586233078326785e-06,
+      "loss": 0.6136,
+      "step": 287
+    },
+    {
+      "epoch": 2.796116504854369,
+      "grad_norm": 0.9283950328826904,
+      "learning_rate": 4.583427675547602e-06,
+      "loss": 0.4247,
+      "step": 288
+    },
+    {
+      "epoch": 2.8058252427184467,
+      "grad_norm": 0.9678206443786621,
+      "learning_rate": 4.580613658447301e-06,
+      "loss": 0.502,
+      "step": 289
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.9418540000915527,
+      "learning_rate": 4.577791038660959e-06,
+      "loss": 0.3092,
+      "step": 290
+    },
+    {
+      "epoch": 2.825242718446602,
+      "grad_norm": 0.958787202835083,
+      "learning_rate": 4.574959827859226e-06,
+      "loss": 0.2833,
+      "step": 291
+    },
+    {
+      "epoch": 2.8349514563106797,
+      "grad_norm": 0.9094011783599854,
+      "learning_rate": 4.572120037748273e-06,
+      "loss": 0.3078,
+      "step": 292
+    },
+    {
+      "epoch": 2.8446601941747574,
+      "grad_norm": 0.8375478982925415,
+      "learning_rate": 4.5692716800697415e-06,
+      "loss": 0.4403,
+      "step": 293
+    },
+    {
+      "epoch": 2.854368932038835,
+      "grad_norm": 0.8053695559501648,
+      "learning_rate": 4.566414766600698e-06,
+      "loss": 0.3585,
+      "step": 294
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.871597945690155,
+      "learning_rate": 4.563549309153589e-06,
+      "loss": 0.238,
+      "step": 295
+    },
+    {
+      "epoch": 2.8737864077669903,
+      "grad_norm": 0.8860502243041992,
+      "learning_rate": 4.56067531957618e-06,
+      "loss": 0.2631,
+      "step": 296
+    },
+    {
+      "epoch": 2.883495145631068,
+      "grad_norm": 0.7494659423828125,
+      "learning_rate": 4.557792809751519e-06,
+      "loss": 0.3956,
+      "step": 297
+    },
+    {
+      "epoch": 2.8932038834951457,
+      "grad_norm": 0.8635668158531189,
+      "learning_rate": 4.554901791597883e-06,
+      "loss": 0.3569,
+      "step": 298
+    },
+    {
+      "epoch": 2.9029126213592233,
+      "grad_norm": 1.0071423053741455,
+      "learning_rate": 4.552002277068725e-06,
+      "loss": 0.3698,
+      "step": 299
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 1.0618561506271362,
+      "learning_rate": 4.549094278152631e-06,
+      "loss": 0.4215,
+      "step": 300
+    },
+    {
+      "epoch": 2.9223300970873787,
+      "grad_norm": 1.3928142786026,
+      "learning_rate": 4.546177806873266e-06,
+      "loss": 0.3282,
+      "step": 301
+    },
+    {
+      "epoch": 2.9320388349514563,
+      "grad_norm": 1.0209264755249023,
+      "learning_rate": 4.543252875289326e-06,
+      "loss": 0.3716,
+      "step": 302
+    },
+    {
+      "epoch": 2.941747572815534,
+      "grad_norm": 0.9945548176765442,
+      "learning_rate": 4.540319495494486e-06,
+      "loss": 0.3973,
+      "step": 303
+    },
+    {
+      "epoch": 2.9514563106796117,
+      "grad_norm": 1.065015435218811,
+      "learning_rate": 4.537377679617353e-06,
+      "loss": 0.3996,
+      "step": 304
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 1.1069443225860596,
+      "learning_rate": 4.534427439821416e-06,
+      "loss": 0.2877,
+      "step": 305
+    },
+    {
+      "epoch": 2.970873786407767,
+      "grad_norm": 1.0089571475982666,
+      "learning_rate": 4.531468788304992e-06,
+      "loss": 0.2799,
+      "step": 306
+    },
+    {
+      "epoch": 2.9805825242718447,
+      "grad_norm": 1.0506935119628906,
+      "learning_rate": 4.5285017373011784e-06,
+      "loss": 0.3221,
+      "step": 307
+    },
+    {
+      "epoch": 2.9902912621359223,
+      "grad_norm": 0.7255403995513916,
+      "learning_rate": 4.5255262990778024e-06,
+      "loss": 0.2756,
+      "step": 308
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8445290923118591,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.2747,
+      "step": 309
+    },
+    {
+      "epoch": 3.0097087378640777,
+      "grad_norm": 0.9877444505691528,
+      "learning_rate": 4.519550310217013e-06,
+      "loss": 0.3152,
+      "step": 310
+    },
+    {
+      "epoch": 3.0194174757281553,
+      "grad_norm": 1.0058015584945679,
+      "learning_rate": 4.516549784288442e-06,
+      "loss": 0.2565,
+      "step": 311
+    },
+    {
+      "epoch": 3.029126213592233,
+      "grad_norm": 0.9974623322486877,
+      "learning_rate": 4.513540920557892e-06,
+      "loss": 0.2612,
+      "step": 312
+    },
+    {
+      "epoch": 3.0388349514563107,
+      "grad_norm": 0.9780676960945129,
+      "learning_rate": 4.510523731466072e-06,
+      "loss": 0.431,
+      "step": 313
+    },
+    {
+      "epoch": 3.0485436893203883,
+      "grad_norm": 1.2715204954147339,
+      "learning_rate": 4.507498229488116e-06,
+      "loss": 0.1685,
+      "step": 314
+    },
+    {
+      "epoch": 3.058252427184466,
+      "grad_norm": 2.212153673171997,
+      "learning_rate": 4.504464427133527e-06,
+      "loss": 0.2729,
+      "step": 315
+    },
+    {
+      "epoch": 3.0679611650485437,
+      "grad_norm": 1.3692691326141357,
+      "learning_rate": 4.501422336946126e-06,
+      "loss": 0.2697,
+      "step": 316
+    },
+    {
+      "epoch": 3.0776699029126213,
+      "grad_norm": 1.1158126592636108,
+      "learning_rate": 4.498371971504005e-06,
+      "loss": 0.2113,
+      "step": 317
+    },
+    {
+      "epoch": 3.087378640776699,
+      "grad_norm": 0.8955278396606445,
+      "learning_rate": 4.49531334341947e-06,
+      "loss": 0.2378,
+      "step": 318
+    },
+    {
+      "epoch": 3.0970873786407767,
+      "grad_norm": 0.9961569309234619,
+      "learning_rate": 4.49224646533899e-06,
+      "loss": 0.2678,
+      "step": 319
+    },
+    {
+      "epoch": 3.1067961165048543,
+      "grad_norm": 0.8884671926498413,
+      "learning_rate": 4.489171349943144e-06,
+      "loss": 0.2445,
+      "step": 320
+    },
+    {
+      "epoch": 3.116504854368932,
+      "grad_norm": 1.1398262977600098,
+      "learning_rate": 4.486088009946575e-06,
+      "loss": 0.1827,
+      "step": 321
+    },
+    {
+      "epoch": 3.1262135922330097,
+      "grad_norm": 1.2452175617218018,
+      "learning_rate": 4.482996458097926e-06,
+      "loss": 0.3053,
+      "step": 322
+    },
+    {
+      "epoch": 3.1359223300970873,
+      "grad_norm": 1.0263898372650146,
+      "learning_rate": 4.479896707179796e-06,
+      "loss": 0.462,
+      "step": 323
+    },
+    {
+      "epoch": 3.145631067961165,
+      "grad_norm": 1.0178437232971191,
+      "learning_rate": 4.476788770008685e-06,
+      "loss": 0.4596,
+      "step": 324
+    },
+    {
+      "epoch": 3.1553398058252426,
+      "grad_norm": 1.0502902269363403,
+      "learning_rate": 4.473672659434941e-06,
+      "loss": 0.3861,
+      "step": 325
+    },
+    {
+      "epoch": 3.1650485436893203,
+      "grad_norm": 1.2341052293777466,
+      "learning_rate": 4.470548388342704e-06,
+      "loss": 0.296,
+      "step": 326
+    },
+    {
+      "epoch": 3.174757281553398,
+      "grad_norm": 1.025456190109253,
+      "learning_rate": 4.467415969649858e-06,
+      "loss": 0.2583,
+      "step": 327
+    },
+    {
+      "epoch": 3.1844660194174756,
+      "grad_norm": 1.0375516414642334,
+      "learning_rate": 4.464275416307973e-06,
+      "loss": 0.3423,
+      "step": 328
+    },
+    {
+      "epoch": 3.1941747572815533,
+      "grad_norm": 1.2566360235214233,
+      "learning_rate": 4.461126741302253e-06,
+      "loss": 0.3176,
+      "step": 329
+    },
+    {
+      "epoch": 3.203883495145631,
+      "grad_norm": 1.177641749382019,
+      "learning_rate": 4.457969957651485e-06,
+      "loss": 0.2558,
+      "step": 330
+    },
+    {
+      "epoch": 3.2135922330097086,
+      "grad_norm": 1.0258479118347168,
+      "learning_rate": 4.454805078407979e-06,
+      "loss": 0.4843,
+      "step": 331
+    },
+    {
+      "epoch": 3.2233009708737863,
+      "grad_norm": 0.9031769037246704,
+      "learning_rate": 4.451632116657521e-06,
+      "loss": 0.2528,
+      "step": 332
+    },
+    {
+      "epoch": 3.233009708737864,
+      "grad_norm": 0.9155856370925903,
+      "learning_rate": 4.448451085519314e-06,
+      "loss": 0.2014,
+      "step": 333
+    },
+    {
+      "epoch": 3.2427184466019416,
+      "grad_norm": 1.1005407571792603,
+      "learning_rate": 4.445261998145927e-06,
+      "loss": 0.3033,
+      "step": 334
+    },
+    {
+      "epoch": 3.2524271844660193,
+      "grad_norm": 0.9126582741737366,
+      "learning_rate": 4.442064867723236e-06,
+      "loss": 0.2785,
+      "step": 335
+    },
+    {
+      "epoch": 3.262135922330097,
+      "grad_norm": 1.0484691858291626,
+      "learning_rate": 4.438859707470376e-06,
+      "loss": 0.293,
+      "step": 336
+    },
+    {
+      "epoch": 3.2718446601941746,
+      "grad_norm": 0.9439131021499634,
+      "learning_rate": 4.435646530639679e-06,
+      "loss": 0.1952,
+      "step": 337
+    },
+    {
+      "epoch": 3.2815533980582523,
+      "grad_norm": 1.2514874935150146,
+      "learning_rate": 4.432425350516627e-06,
+      "loss": 0.2916,
+      "step": 338
+    },
+    {
+      "epoch": 3.29126213592233,
+      "grad_norm": 1.256081223487854,
+      "learning_rate": 4.42919618041979e-06,
+      "loss": 0.1883,
+      "step": 339
+    },
+    {
+      "epoch": 3.3009708737864076,
+      "grad_norm": 0.9661153554916382,
+      "learning_rate": 4.425959033700776e-06,
+      "loss": 0.1276,
+      "step": 340
+    },
+    {
+      "epoch": 3.3106796116504853,
+      "grad_norm": 0.9738348126411438,
+      "learning_rate": 4.422713923744174e-06,
+      "loss": 0.348,
+      "step": 341
+    },
+    {
+      "epoch": 3.320388349514563,
+      "grad_norm": 1.2630469799041748,
+      "learning_rate": 4.419460863967496e-06,
+      "loss": 0.375,
+      "step": 342
+    },
+    {
+      "epoch": 3.3300970873786406,
+      "grad_norm": 0.866310179233551,
+      "learning_rate": 4.416199867821126e-06,
+      "loss": 0.3097,
+      "step": 343
+    },
+    {
+      "epoch": 3.3398058252427183,
+      "grad_norm": 1.4241950511932373,
+      "learning_rate": 4.412930948788263e-06,
+      "loss": 0.2579,
+      "step": 344
+    },
+    {
+      "epoch": 3.349514563106796,
+      "grad_norm": 0.9571874737739563,
+      "learning_rate": 4.409654120384863e-06,
+      "loss": 0.436,
+      "step": 345
+    },
+    {
+      "epoch": 3.3592233009708736,
+      "grad_norm": 1.0651613473892212,
+      "learning_rate": 4.406369396159585e-06,
+      "loss": 0.3728,
+      "step": 346
+    },
+    {
+      "epoch": 3.3689320388349513,
+      "grad_norm": 1.2672362327575684,
+      "learning_rate": 4.403076789693735e-06,
+      "loss": 0.448,
+      "step": 347
+    },
+    {
+      "epoch": 3.378640776699029,
+      "grad_norm": 1.0908823013305664,
+      "learning_rate": 4.399776314601212e-06,
+      "loss": 0.2399,
+      "step": 348
+    },
+    {
+      "epoch": 3.3883495145631066,
+      "grad_norm": 1.2599388360977173,
+      "learning_rate": 4.396467984528445e-06,
+      "loss": 0.3701,
+      "step": 349
+    },
+    {
+      "epoch": 3.3980582524271843,
+      "grad_norm": 1.1919026374816895,
+      "learning_rate": 4.393151813154345e-06,
+      "loss": 0.2457,
+      "step": 350
+    },
+    {
+      "epoch": 3.407766990291262,
+      "grad_norm": 1.035890817642212,
+      "learning_rate": 4.3898278141902396e-06,
+      "loss": 0.2621,
+      "step": 351
+    },
+    {
+      "epoch": 3.4174757281553396,
+      "grad_norm": 1.1830110549926758,
+      "learning_rate": 4.386496001379826e-06,
+      "loss": 0.3427,
+      "step": 352
+    },
+    {
+      "epoch": 3.4271844660194173,
+      "grad_norm": 0.8301703929901123,
+      "learning_rate": 4.383156388499106e-06,
+      "loss": 0.3473,
+      "step": 353
+    },
+    {
+      "epoch": 3.436893203883495,
+      "grad_norm": 0.8561685085296631,
+      "learning_rate": 4.3798089893563335e-06,
+      "loss": 0.2333,
+      "step": 354
+    },
+    {
+      "epoch": 3.4466019417475726,
+      "grad_norm": 1.0540318489074707,
+      "learning_rate": 4.3764538177919555e-06,
+      "loss": 0.1516,
+      "step": 355
+    },
+    {
+      "epoch": 3.4563106796116507,
+      "grad_norm": 1.0419858694076538,
+      "learning_rate": 4.3730908876785574e-06,
+      "loss": 0.2708,
+      "step": 356
+    },
+    {
+      "epoch": 3.466019417475728,
+      "grad_norm": 1.171985149383545,
+      "learning_rate": 4.3697202129208e-06,
+      "loss": 0.3133,
+      "step": 357
+    },
+    {
+      "epoch": 3.475728155339806,
+      "grad_norm": 1.0638673305511475,
+      "learning_rate": 4.36634180745537e-06,
+      "loss": 0.1745,
+      "step": 358
+    },
+    {
+      "epoch": 3.4854368932038833,
+      "grad_norm": 0.8694558143615723,
+      "learning_rate": 4.3629556852509145e-06,
+      "loss": 0.4621,
+      "step": 359
+    },
+    {
+      "epoch": 3.4951456310679614,
+      "grad_norm": 0.932741105556488,
+      "learning_rate": 4.35956186030799e-06,
+      "loss": 0.2332,
+      "step": 360
+    },
+    {
+      "epoch": 3.5048543689320386,
+      "grad_norm": 1.1963168382644653,
+      "learning_rate": 4.356160346659001e-06,
+      "loss": 0.3011,
+      "step": 361
+    },
+    {
+      "epoch": 3.5145631067961167,
+      "grad_norm": 1.1899361610412598,
+      "learning_rate": 4.3527511583681384e-06,
+      "loss": 0.6784,
+      "step": 362
+    },
+    {
+      "epoch": 3.524271844660194,
+      "grad_norm": 1.1308809518814087,
+      "learning_rate": 4.34933430953133e-06,
+      "loss": 0.3338,
+      "step": 363
+    },
+    {
+      "epoch": 3.533980582524272,
+      "grad_norm": 1.0777356624603271,
+      "learning_rate": 4.345909814276177e-06,
+      "loss": 0.362,
+      "step": 364
+    },
+    {
+      "epoch": 3.5436893203883493,
+      "grad_norm": 0.9632010459899902,
+      "learning_rate": 4.3424776867618935e-06,
+      "loss": 0.3633,
+      "step": 365
+    },
+    {
+      "epoch": 3.5533980582524274,
+      "grad_norm": 1.0557489395141602,
+      "learning_rate": 4.339037941179253e-06,
+      "loss": 0.3633,
+      "step": 366
+    },
+    {
+      "epoch": 3.5631067961165046,
+      "grad_norm": 0.9937204122543335,
+      "learning_rate": 4.335590591750526e-06,
+      "loss": 0.2539,
+      "step": 367
+    },
+    {
+      "epoch": 3.5728155339805827,
+      "grad_norm": 1.128485083580017,
+      "learning_rate": 4.332135652729423e-06,
+      "loss": 0.2688,
+      "step": 368
+    },
+    {
+      "epoch": 3.58252427184466,
+      "grad_norm": 0.9485536217689514,
+      "learning_rate": 4.328673138401036e-06,
+      "loss": 0.3675,
+      "step": 369
+    },
+    {
+      "epoch": 3.592233009708738,
+      "grad_norm": 1.0004498958587646,
+      "learning_rate": 4.325203063081776e-06,
+      "loss": 0.1809,
+      "step": 370
+    },
+    {
+      "epoch": 3.6019417475728153,
+      "grad_norm": 1.2935380935668945,
+      "learning_rate": 4.32172544111932e-06,
+      "loss": 0.2047,
+      "step": 371
+    },
+    {
+      "epoch": 3.6116504854368934,
+      "grad_norm": 0.8594838976860046,
+      "learning_rate": 4.318240286892544e-06,
+      "loss": 0.2426,
+      "step": 372
+    },
+    {
+      "epoch": 3.6213592233009706,
+      "grad_norm": 0.865341305732727,
+      "learning_rate": 4.314747614811471e-06,
+      "loss": 0.269,
+      "step": 373
+    },
+    {
+      "epoch": 3.6310679611650487,
+      "grad_norm": 0.975472629070282,
+      "learning_rate": 4.3112474393172055e-06,
+      "loss": 0.2057,
+      "step": 374
+    },
+    {
+      "epoch": 3.6407766990291264,
+      "grad_norm": 0.9958301782608032,
+      "learning_rate": 4.307739774881878e-06,
+      "loss": 0.2612,
+      "step": 375
+    },
+    {
+      "epoch": 3.650485436893204,
+      "grad_norm": 1.216472864151001,
+      "learning_rate": 4.304224636008582e-06,
+      "loss": 0.1837,
+      "step": 376
+    },
+    {
+      "epoch": 3.6601941747572817,
+      "grad_norm": 1.432723879814148,
+      "learning_rate": 4.300702037231318e-06,
+      "loss": 0.4011,
+      "step": 377
+    },
+    {
+      "epoch": 3.6699029126213594,
+      "grad_norm": 0.9345541596412659,
+      "learning_rate": 4.297171993114927e-06,
+      "loss": 0.1405,
+      "step": 378
+    },
+    {
+      "epoch": 3.679611650485437,
+      "grad_norm": 0.8612492084503174,
+      "learning_rate": 4.2936345182550365e-06,
+      "loss": 0.4025,
+      "step": 379
+    },
+    {
+      "epoch": 3.6893203883495147,
+      "grad_norm": 0.9304858446121216,
+      "learning_rate": 4.290089627277998e-06,
+      "loss": 0.2106,
+      "step": 380
+    },
+    {
+      "epoch": 3.6990291262135924,
+      "grad_norm": 0.8536747694015503,
+      "learning_rate": 4.286537334840825e-06,
+      "loss": 0.3325,
+      "step": 381
+    },
+    {
+      "epoch": 3.70873786407767,
+      "grad_norm": 1.4801360368728638,
+      "learning_rate": 4.2829776556311355e-06,
+      "loss": 0.2465,
+      "step": 382
+    },
+    {
+      "epoch": 3.7184466019417477,
+      "grad_norm": 1.1944482326507568,
+      "learning_rate": 4.279410604367088e-06,
+      "loss": 0.4483,
+      "step": 383
+    },
+    {
+      "epoch": 3.7281553398058254,
+      "grad_norm": 1.1957206726074219,
+      "learning_rate": 4.275836195797323e-06,
+      "loss": 0.3758,
+      "step": 384
+    },
+    {
+      "epoch": 3.737864077669903,
+      "grad_norm": 0.9933880567550659,
+      "learning_rate": 4.2722544447008995e-06,
+      "loss": 0.4136,
+      "step": 385
+    },
+    {
+      "epoch": 3.7475728155339807,
+      "grad_norm": 1.062076210975647,
+      "learning_rate": 4.268665365887238e-06,
+      "loss": 0.441,
+      "step": 386
+    },
+    {
+      "epoch": 3.7572815533980584,
+      "grad_norm": 0.8356176018714905,
+      "learning_rate": 4.265068974196056e-06,
+      "loss": 0.1671,
+      "step": 387
+    },
+    {
+      "epoch": 3.766990291262136,
+      "grad_norm": 1.0670630931854248,
+      "learning_rate": 4.261465284497307e-06,
+      "loss": 0.2107,
+      "step": 388
+    },
+    {
+      "epoch": 3.7766990291262137,
+      "grad_norm": 0.9141836762428284,
+      "learning_rate": 4.257854311691118e-06,
+      "loss": 0.3743,
+      "step": 389
+    },
+    {
+      "epoch": 3.7864077669902914,
+      "grad_norm": 1.2973644733428955,
+      "learning_rate": 4.254236070707734e-06,
+      "loss": 0.2203,
+      "step": 390
+    },
+    {
+      "epoch": 3.796116504854369,
+      "grad_norm": 0.9033822417259216,
+      "learning_rate": 4.250610576507445e-06,
+      "loss": 0.2088,
+      "step": 391
+    },
+    {
+      "epoch": 3.8058252427184467,
+      "grad_norm": 1.1606167554855347,
+      "learning_rate": 4.246977844080537e-06,
+      "loss": 0.3493,
+      "step": 392
+    },
+    {
+      "epoch": 3.8155339805825244,
+      "grad_norm": 1.1568549871444702,
+      "learning_rate": 4.24333788844722e-06,
+      "loss": 0.2675,
+      "step": 393
+    },
+    {
+      "epoch": 3.825242718446602,
+      "grad_norm": 1.2251572608947754,
+      "learning_rate": 4.239690724657571e-06,
+      "loss": 0.2949,
+      "step": 394
+    },
+    {
+      "epoch": 3.8349514563106797,
+      "grad_norm": 0.7754173278808594,
+      "learning_rate": 4.236036367791471e-06,
+      "loss": 0.3149,
+      "step": 395
+    },
+    {
+      "epoch": 3.8446601941747574,
+      "grad_norm": 1.0242606401443481,
+      "learning_rate": 4.23237483295854e-06,
+      "loss": 0.4014,
+      "step": 396
+    },
+    {
+      "epoch": 3.854368932038835,
+      "grad_norm": 1.0819196701049805,
+      "learning_rate": 4.228706135298081e-06,
+      "loss": 0.2485,
+      "step": 397
+    },
+    {
+      "epoch": 3.8640776699029127,
+      "grad_norm": 1.0384984016418457,
+      "learning_rate": 4.225030289979006e-06,
+      "loss": 0.2049,
+      "step": 398
+    },
+    {
+      "epoch": 3.8737864077669903,
+      "grad_norm": 0.9361140131950378,
+      "learning_rate": 4.221347312199788e-06,
+      "loss": 0.2095,
+      "step": 399
+    },
+    {
+      "epoch": 3.883495145631068,
+      "grad_norm": 0.7564887404441833,
+      "learning_rate": 4.2176572171883865e-06,
+      "loss": 0.3075,
+      "step": 400
+    },
+    {
+      "epoch": 3.8932038834951457,
+      "grad_norm": 1.040762186050415,
+      "learning_rate": 4.213960020202187e-06,
+      "loss": 0.3612,
+      "step": 401
+    },
+    {
+      "epoch": 3.9029126213592233,
+      "grad_norm": 1.23052179813385,
+      "learning_rate": 4.2102557365279435e-06,
+      "loss": 0.4987,
+      "step": 402
+    },
+    {
+      "epoch": 3.912621359223301,
+      "grad_norm": 1.1888084411621094,
+      "learning_rate": 4.206544381481708e-06,
+      "loss": 0.3761,
+      "step": 403
+    },
+    {
+      "epoch": 3.9223300970873787,
+      "grad_norm": 1.0521200895309448,
+      "learning_rate": 4.202825970408772e-06,
+      "loss": 0.1774,
+      "step": 404
+    },
+    {
+      "epoch": 3.9320388349514563,
+      "grad_norm": 1.4193365573883057,
+      "learning_rate": 4.199100518683601e-06,
+      "loss": 0.2208,
+      "step": 405
+    },
+    {
+      "epoch": 3.941747572815534,
+      "grad_norm": 0.8504016399383545,
+      "learning_rate": 4.195368041709772e-06,
+      "loss": 0.3786,
+      "step": 406
+    },
+    {
+      "epoch": 3.9514563106796117,
+      "grad_norm": 0.9185003042221069,
+      "learning_rate": 4.191628554919907e-06,
+      "loss": 0.2445,
+      "step": 407
+    },
+    {
+      "epoch": 3.9611650485436893,
+      "grad_norm": 1.0671206712722778,
+      "learning_rate": 4.187882073775615e-06,
+      "loss": 0.2674,
+      "step": 408
+    },
+    {
+      "epoch": 3.970873786407767,
+      "grad_norm": 0.9746595025062561,
+      "learning_rate": 4.184128613767422e-06,
+      "loss": 0.2695,
+      "step": 409
+    },
+    {
+      "epoch": 3.9805825242718447,
+      "grad_norm": 0.9572174549102783,
+      "learning_rate": 4.18036819041471e-06,
+      "loss": 0.3031,
+      "step": 410
+    },
+    {
+      "epoch": 3.9902912621359223,
+      "grad_norm": 1.1593903303146362,
+      "learning_rate": 4.17660081926565e-06,
+      "loss": 0.1894,
+      "step": 411
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.0100287199020386,
+      "learning_rate": 4.172826515897146e-06,
+      "loss": 0.1792,
+      "step": 412
+    },
+    {
+      "epoch": 4.009708737864078,
+      "grad_norm": 0.8554271459579468,
+      "learning_rate": 4.169045295914757e-06,
+      "loss": 0.3282,
+      "step": 413
+    },
+    {
+      "epoch": 4.019417475728155,
+      "grad_norm": 1.0497305393218994,
+      "learning_rate": 4.165257174952647e-06,
+      "loss": 0.1456,
+      "step": 414
+    },
+    {
+      "epoch": 4.029126213592233,
+      "grad_norm": 0.9351025223731995,
+      "learning_rate": 4.161462168673508e-06,
+      "loss": 0.2288,
+      "step": 415
+    },
+    {
+      "epoch": 4.038834951456311,
+      "grad_norm": 1.175843596458435,
+      "learning_rate": 4.157660292768502e-06,
+      "loss": 0.3184,
+      "step": 416
+    },
+    {
+      "epoch": 4.048543689320389,
+      "grad_norm": 1.3104645013809204,
+      "learning_rate": 4.1538515629571985e-06,
+      "loss": 0.2433,
+      "step": 417
+    },
+    {
+      "epoch": 4.058252427184466,
+      "grad_norm": 1.4812334775924683,
+      "learning_rate": 4.1500359949875e-06,
+      "loss": 0.2837,
+      "step": 418
+    },
+    {
+      "epoch": 4.067961165048544,
+      "grad_norm": 1.3549306392669678,
+      "learning_rate": 4.1462136046355864e-06,
+      "loss": 0.2224,
+      "step": 419
+    },
+    {
+      "epoch": 4.077669902912621,
+      "grad_norm": 1.1158041954040527,
+      "learning_rate": 4.142384407705846e-06,
+      "loss": 0.4027,
+      "step": 420
+    },
+    {
+      "epoch": 4.087378640776699,
+      "grad_norm": 1.0280787944793701,
+      "learning_rate": 4.138548420030808e-06,
+      "loss": 0.1437,
+      "step": 421
+    },
+    {
+      "epoch": 4.097087378640777,
+      "grad_norm": 1.043931245803833,
+      "learning_rate": 4.13470565747108e-06,
+      "loss": 0.2594,
+      "step": 422
+    },
+    {
+      "epoch": 4.106796116504855,
+      "grad_norm": 0.8889212608337402,
+      "learning_rate": 4.130856135915282e-06,
+      "loss": 0.1997,
+      "step": 423
+    },
+    {
+      "epoch": 4.116504854368932,
+      "grad_norm": 1.3302550315856934,
+      "learning_rate": 4.126999871279982e-06,
+      "loss": 0.2012,
+      "step": 424
+    },
+    {
+      "epoch": 4.12621359223301,
+      "grad_norm": 1.0227890014648438,
+      "learning_rate": 4.123136879509626e-06,
+      "loss": 0.2964,
+      "step": 425
+    },
+    {
+      "epoch": 4.135922330097087,
+      "grad_norm": 1.2447762489318848,
+      "learning_rate": 4.119267176576475e-06,
+      "loss": 0.2678,
+      "step": 426
+    },
+    {
+      "epoch": 4.145631067961165,
+      "grad_norm": 0.8202763795852661,
+      "learning_rate": 4.11539077848054e-06,
+      "loss": 0.3348,
+      "step": 427
+    },
+    {
+      "epoch": 4.155339805825243,
+      "grad_norm": 1.2533812522888184,
+      "learning_rate": 4.111507701249513e-06,
+      "loss": 0.2985,
+      "step": 428
+    },
+    {
+      "epoch": 4.165048543689321,
+      "grad_norm": 1.6550928354263306,
+      "learning_rate": 4.107617960938702e-06,
+      "loss": 0.1144,
+      "step": 429
+    },
+    {
+      "epoch": 4.174757281553398,
+      "grad_norm": 1.470839023590088,
+      "learning_rate": 4.103721573630965e-06,
+      "loss": 0.2227,
+      "step": 430
+    },
+    {
+      "epoch": 4.184466019417476,
+      "grad_norm": 1.1045992374420166,
+      "learning_rate": 4.099818555436645e-06,
+      "loss": 0.2515,
+      "step": 431
+    },
+    {
+      "epoch": 4.194174757281553,
+      "grad_norm": 1.691807746887207,
+      "learning_rate": 4.095908922493499e-06,
+      "loss": 0.463,
+      "step": 432
+    },
+    {
+      "epoch": 4.203883495145631,
+      "grad_norm": 0.9316760897636414,
+      "learning_rate": 4.091992690966636e-06,
+      "loss": 0.0993,
+      "step": 433
+    },
+    {
+      "epoch": 4.213592233009709,
+      "grad_norm": 1.1786795854568481,
+      "learning_rate": 4.088069877048447e-06,
+      "loss": 0.1766,
+      "step": 434
+    },
+    {
+      "epoch": 4.223300970873787,
+      "grad_norm": 0.8992094397544861,
+      "learning_rate": 4.084140496958539e-06,
+      "loss": 0.1756,
+      "step": 435
+    },
+    {
+      "epoch": 4.233009708737864,
+      "grad_norm": 0.9905827045440674,
+      "learning_rate": 4.080204566943668e-06,
+      "loss": 0.1459,
+      "step": 436
+    },
+    {
+      "epoch": 4.242718446601942,
+      "grad_norm": 0.9543220400810242,
+      "learning_rate": 4.076262103277673e-06,
+      "loss": 0.1493,
+      "step": 437
+    },
+    {
+      "epoch": 4.252427184466019,
+      "grad_norm": 0.9142147302627563,
+      "learning_rate": 4.072313122261406e-06,
+      "loss": 0.4272,
+      "step": 438
+    },
+    {
+      "epoch": 4.262135922330097,
+      "grad_norm": 0.8952410817146301,
+      "learning_rate": 4.068357640222668e-06,
+      "loss": 0.121,
+      "step": 439
+    },
+    {
+      "epoch": 4.271844660194175,
+      "grad_norm": 0.8736719489097595,
+      "learning_rate": 4.06439567351614e-06,
+      "loss": 0.1525,
+      "step": 440
+    },
+    {
+      "epoch": 4.281553398058253,
+      "grad_norm": 0.9316339492797852,
+      "learning_rate": 4.0604272385233105e-06,
+      "loss": 0.2146,
+      "step": 441
+    },
+    {
+      "epoch": 4.29126213592233,
+      "grad_norm": 1.1137328147888184,
+      "learning_rate": 4.056452351652418e-06,
+      "loss": 0.1259,
+      "step": 442
+    },
+    {
+      "epoch": 4.300970873786408,
+      "grad_norm": 1.1264760494232178,
+      "learning_rate": 4.052471029338375e-06,
+      "loss": 0.2629,
+      "step": 443
+    },
+    {
+      "epoch": 4.310679611650485,
+      "grad_norm": 0.9680051803588867,
+      "learning_rate": 4.048483288042703e-06,
+      "loss": 0.1495,
+      "step": 444
+    },
+    {
+      "epoch": 4.320388349514563,
+      "grad_norm": 1.0331041812896729,
+      "learning_rate": 4.0444891442534615e-06,
+      "loss": 0.1509,
+      "step": 445
+    },
+    {
+      "epoch": 4.330097087378641,
+      "grad_norm": 0.9411431550979614,
+      "learning_rate": 4.040488614485187e-06,
+      "loss": 0.2739,
+      "step": 446
+    },
+    {
+      "epoch": 4.339805825242719,
+      "grad_norm": 1.2953507900238037,
+      "learning_rate": 4.036481715278818e-06,
+      "loss": 0.2616,
+      "step": 447
+    },
+    {
+      "epoch": 4.349514563106796,
+      "grad_norm": 0.8018221855163574,
+      "learning_rate": 4.032468463201626e-06,
+      "loss": 0.2696,
+      "step": 448
+    },
+    {
+      "epoch": 4.359223300970874,
+      "grad_norm": 0.9439960718154907,
+      "learning_rate": 4.028448874847152e-06,
+      "loss": 0.0845,
+      "step": 449
+    },
+    {
+      "epoch": 4.368932038834951,
+      "grad_norm": 1.1227456331253052,
+      "learning_rate": 4.024422966835137e-06,
+      "loss": 0.2061,
+      "step": 450
+    },
+    {
+      "epoch": 4.378640776699029,
+      "grad_norm": 1.0585674047470093,
+      "learning_rate": 4.0203907558114475e-06,
+      "loss": 0.1585,
+      "step": 451
+    },
+    {
+      "epoch": 4.388349514563107,
+      "grad_norm": 0.9007663130760193,
+      "learning_rate": 4.016352258448016e-06,
+      "loss": 0.1067,
+      "step": 452
+    },
+    {
+      "epoch": 4.398058252427185,
+      "grad_norm": 1.3700408935546875,
+      "learning_rate": 4.0123074914427635e-06,
+      "loss": 0.148,
+      "step": 453
+    },
+    {
+      "epoch": 4.407766990291262,
+      "grad_norm": 0.8373147249221802,
+      "learning_rate": 4.008256471519536e-06,
+      "loss": 0.1144,
+      "step": 454
+    },
+    {
+      "epoch": 4.41747572815534,
+      "grad_norm": 1.0520269870758057,
+      "learning_rate": 4.004199215428032e-06,
+      "loss": 0.2349,
+      "step": 455
+    },
+    {
+      "epoch": 4.427184466019417,
+      "grad_norm": 1.3709717988967896,
+      "learning_rate": 4.000135739943735e-06,
+      "loss": 0.3069,
+      "step": 456
+    },
+    {
+      "epoch": 4.436893203883495,
+      "grad_norm": 1.0186606645584106,
+      "learning_rate": 3.996066061867844e-06,
+      "loss": 0.2285,
+      "step": 457
+    },
+    {
+      "epoch": 4.446601941747573,
+      "grad_norm": 1.0011564493179321,
+      "learning_rate": 3.991990198027203e-06,
+      "loss": 0.1904,
+      "step": 458
+    },
+    {
+      "epoch": 4.456310679611651,
+      "grad_norm": 1.174538493156433,
+      "learning_rate": 3.987908165274233e-06,
+      "loss": 0.2158,
+      "step": 459
+    },
+    {
+      "epoch": 4.466019417475728,
+      "grad_norm": 0.9717496037483215,
+      "learning_rate": 3.9838199804868635e-06,
+      "loss": 0.3015,
+      "step": 460
+    },
+    {
+      "epoch": 4.475728155339806,
+      "grad_norm": 1.2071607112884521,
+      "learning_rate": 3.979725660568456e-06,
+      "loss": 0.2892,
+      "step": 461
+    },
+    {
+      "epoch": 4.485436893203883,
+      "grad_norm": 0.8957129120826721,
+      "learning_rate": 3.975625222447742e-06,
+      "loss": 0.1432,
+      "step": 462
+    },
+    {
+      "epoch": 4.495145631067961,
+      "grad_norm": 1.402980923652649,
+      "learning_rate": 3.97151868307875e-06,
+      "loss": 0.2913,
+      "step": 463
+    },
+    {
+      "epoch": 4.504854368932039,
+      "grad_norm": 1.0455483198165894,
+      "learning_rate": 3.9674060594407345e-06,
+      "loss": 0.1864,
+      "step": 464
+    },
+    {
+      "epoch": 4.514563106796117,
+      "grad_norm": 0.9044820666313171,
+      "learning_rate": 3.963287368538105e-06,
+      "loss": 0.0945,
+      "step": 465
+    },
+    {
+      "epoch": 4.524271844660194,
+      "grad_norm": 1.2196074724197388,
+      "learning_rate": 3.959162627400361e-06,
+      "loss": 0.3229,
+      "step": 466
+    },
+    {
+      "epoch": 4.533980582524272,
+      "grad_norm": 1.430355191230774,
+      "learning_rate": 3.9550318530820145e-06,
+      "loss": 0.2709,
+      "step": 467
+    },
+    {
+      "epoch": 4.543689320388349,
+      "grad_norm": 1.0211271047592163,
+      "learning_rate": 3.9508950626625244e-06,
+      "loss": 0.1399,
+      "step": 468
+    },
+    {
+      "epoch": 4.553398058252427,
+      "grad_norm": 1.0876238346099854,
+      "learning_rate": 3.946752273246224e-06,
+      "loss": 0.2654,
+      "step": 469
+    },
+    {
+      "epoch": 4.563106796116505,
+      "grad_norm": 0.975965678691864,
+      "learning_rate": 3.942603501962249e-06,
+      "loss": 0.1921,
+      "step": 470
+    },
+    {
+      "epoch": 4.572815533980583,
+      "grad_norm": 1.0028088092803955,
+      "learning_rate": 3.9384487659644716e-06,
+      "loss": 0.1228,
+      "step": 471
+    },
+    {
+      "epoch": 4.58252427184466,
+      "grad_norm": 1.1247432231903076,
+      "learning_rate": 3.934288082431423e-06,
+      "loss": 0.2308,
+      "step": 472
+    },
+    {
+      "epoch": 4.592233009708738,
+      "grad_norm": 0.99698805809021,
+      "learning_rate": 3.930121468566227e-06,
+      "loss": 0.3171,
+      "step": 473
+    },
+    {
+      "epoch": 4.601941747572815,
+      "grad_norm": 1.0016257762908936,
+      "learning_rate": 3.925948941596528e-06,
+      "loss": 0.1961,
+      "step": 474
+    },
+    {
+      "epoch": 4.611650485436893,
+      "grad_norm": 0.7484853863716125,
+      "learning_rate": 3.92177051877442e-06,
+      "loss": 0.1746,
+      "step": 475
+    },
+    {
+      "epoch": 4.621359223300971,
+      "grad_norm": 0.9711058139801025,
+      "learning_rate": 3.917586217376369e-06,
+      "loss": 0.2399,
+      "step": 476
+    },
+    {
+      "epoch": 4.631067961165049,
+      "grad_norm": 0.9540738463401794,
+      "learning_rate": 3.913396054703155e-06,
+      "loss": 0.2647,
+      "step": 477
+    },
+    {
+      "epoch": 4.640776699029126,
+      "grad_norm": 0.9963274002075195,
+      "learning_rate": 3.909200048079786e-06,
+      "loss": 0.2405,
+      "step": 478
+    },
+    {
+      "epoch": 4.650485436893204,
+      "grad_norm": 1.0638726949691772,
+      "learning_rate": 3.9049982148554384e-06,
+      "loss": 0.3632,
+      "step": 479
+    },
+    {
+      "epoch": 4.660194174757281,
+      "grad_norm": 1.0812081098556519,
+      "learning_rate": 3.900790572403376e-06,
+      "loss": 0.2,
+      "step": 480
+    },
+    {
+      "epoch": 4.669902912621359,
+      "grad_norm": 0.8857947587966919,
+      "learning_rate": 3.896577138120881e-06,
+      "loss": 0.2758,
+      "step": 481
+    },
+    {
+      "epoch": 4.679611650485437,
+      "grad_norm": 1.2224204540252686,
+      "learning_rate": 3.892357929429187e-06,
+      "loss": 0.1827,
+      "step": 482
+    },
+    {
+      "epoch": 4.689320388349515,
+      "grad_norm": 1.1760519742965698,
+      "learning_rate": 3.8881329637734e-06,
+      "loss": 0.1317,
+      "step": 483
+    },
+    {
+      "epoch": 4.699029126213592,
+      "grad_norm": 0.9335840344429016,
+      "learning_rate": 3.883902258622431e-06,
+      "loss": 0.2218,
+      "step": 484
+    },
+    {
+      "epoch": 4.70873786407767,
+      "grad_norm": 1.1263084411621094,
+      "learning_rate": 3.8796658314689205e-06,
+      "loss": 0.2378,
+      "step": 485
+    },
+    {
+      "epoch": 4.718446601941747,
+      "grad_norm": 1.0587648153305054,
+      "learning_rate": 3.875423699829168e-06,
+      "loss": 0.1328,
+      "step": 486
+    },
+    {
+      "epoch": 4.728155339805825,
+      "grad_norm": 0.9542792439460754,
+      "learning_rate": 3.871175881243061e-06,
+      "loss": 0.1703,
+      "step": 487
+    },
+    {
+      "epoch": 4.737864077669903,
+      "grad_norm": 0.9771331548690796,
+      "learning_rate": 3.866922393273999e-06,
+      "loss": 0.3161,
+      "step": 488
+    },
+    {
+      "epoch": 4.747572815533981,
+      "grad_norm": 1.167136788368225,
+      "learning_rate": 3.862663253508822e-06,
+      "loss": 0.1772,
+      "step": 489
+    },
+    {
+      "epoch": 4.757281553398058,
+      "grad_norm": 0.9686987400054932,
+      "learning_rate": 3.858398479557739e-06,
+      "loss": 0.1755,
+      "step": 490
+    },
+    {
+      "epoch": 4.766990291262136,
+      "grad_norm": 1.1323494911193848,
+      "learning_rate": 3.8541280890542565e-06,
+      "loss": 0.2034,
+      "step": 491
+    },
+    {
+      "epoch": 4.776699029126213,
+      "grad_norm": 1.456445574760437,
+      "learning_rate": 3.849852099655102e-06,
+      "loss": 0.3151,
+      "step": 492
+    },
+    {
+      "epoch": 4.786407766990291,
+      "grad_norm": 0.9582480788230896,
+      "learning_rate": 3.845570529040151e-06,
+      "loss": 0.1992,
+      "step": 493
+    },
+    {
+      "epoch": 4.796116504854369,
+      "grad_norm": 0.9113503098487854,
+      "learning_rate": 3.841283394912361e-06,
+      "loss": 0.1958,
+      "step": 494
+    },
+    {
+      "epoch": 4.805825242718447,
+      "grad_norm": 1.0523179769515991,
+      "learning_rate": 3.836990714997686e-06,
+      "loss": 0.2865,
+      "step": 495
+    },
+    {
+      "epoch": 4.815533980582524,
+      "grad_norm": 1.2636815309524536,
+      "learning_rate": 3.832692507045015e-06,
+      "loss": 0.2601,
+      "step": 496
+    },
+    {
+      "epoch": 4.825242718446602,
+      "grad_norm": 0.8661506175994873,
+      "learning_rate": 3.828388788826091e-06,
+      "loss": 0.312,
+      "step": 497
+    },
+    {
+      "epoch": 4.834951456310679,
+      "grad_norm": 1.0353248119354248,
+      "learning_rate": 3.824079578135442e-06,
+      "loss": 0.2472,
+      "step": 498
+    },
+    {
+      "epoch": 4.844660194174757,
+      "grad_norm": 1.523149847984314,
+      "learning_rate": 3.819764892790307e-06,
+      "loss": 0.2525,
+      "step": 499
+    },
+    {
+      "epoch": 4.854368932038835,
+      "grad_norm": 1.2362757921218872,
+      "learning_rate": 3.815444750630555e-06,
+      "loss": 0.2148,
+      "step": 500
+    },
+    {
+      "epoch": 4.864077669902913,
+      "grad_norm": 1.7284575700759888,
+      "learning_rate": 3.811119169518624e-06,
+      "loss": 0.3737,
+      "step": 501
+    },
+    {
+      "epoch": 4.87378640776699,
+      "grad_norm": 1.220207929611206,
+      "learning_rate": 3.8067881673394363e-06,
+      "loss": 0.4219,
+      "step": 502
+    },
+    {
+      "epoch": 4.883495145631068,
+      "grad_norm": 1.0464245080947876,
+      "learning_rate": 3.802451762000331e-06,
+      "loss": 0.1622,
+      "step": 503
+    },
+    {
+      "epoch": 4.893203883495145,
+      "grad_norm": 1.5916672945022583,
+      "learning_rate": 3.7981099714309856e-06,
+      "loss": 0.149,
+      "step": 504
+    },
+    {
+      "epoch": 4.902912621359223,
+      "grad_norm": 1.0251778364181519,
+      "learning_rate": 3.7937628135833453e-06,
+      "loss": 0.2037,
+      "step": 505
+    },
+    {
+      "epoch": 4.9126213592233015,
+      "grad_norm": 1.2007070779800415,
+      "learning_rate": 3.7894103064315463e-06,
+      "loss": 0.197,
+      "step": 506
+    },
+    {
+      "epoch": 4.922330097087379,
+      "grad_norm": 1.1666109561920166,
+      "learning_rate": 3.7850524679718424e-06,
+      "loss": 0.155,
+      "step": 507
+    },
+    {
+      "epoch": 4.932038834951456,
+      "grad_norm": 0.9731155633926392,
+      "learning_rate": 3.7806893162225328e-06,
+      "loss": 0.1791,
+      "step": 508
+    },
+    {
+      "epoch": 4.941747572815534,
+      "grad_norm": 1.1099647283554077,
+      "learning_rate": 3.7763208692238818e-06,
+      "loss": 0.2402,
+      "step": 509
+    },
+    {
+      "epoch": 4.951456310679612,
+      "grad_norm": 1.1254994869232178,
+      "learning_rate": 3.7719471450380518e-06,
+      "loss": 0.1703,
+      "step": 510
+    },
+    {
+      "epoch": 4.961165048543689,
+      "grad_norm": 1.1321799755096436,
+      "learning_rate": 3.7675681617490212e-06,
+      "loss": 0.3277,
+      "step": 511
+    },
+    {
+      "epoch": 4.970873786407767,
+      "grad_norm": 0.9587062001228333,
+      "learning_rate": 3.7631839374625167e-06,
+      "loss": 0.3316,
+      "step": 512
+    },
+    {
+      "epoch": 4.980582524271845,
+      "grad_norm": 0.9437461495399475,
+      "learning_rate": 3.758794490305932e-06,
+      "loss": 0.166,
+      "step": 513
+    },
+    {
+      "epoch": 4.990291262135923,
+      "grad_norm": 1.0000038146972656,
+      "learning_rate": 3.7543998384282565e-06,
+      "loss": 0.2137,
+      "step": 514
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.8058333396911621,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.2744,
+      "step": 515
+    },
+    {
+      "epoch": 5.009708737864078,
+      "grad_norm": 1.1363271474838257,
+      "learning_rate": 3.745594993213118e-06,
+      "loss": 0.2244,
+      "step": 516
+    },
+    {
+      "epoch": 5.019417475728155,
+      "grad_norm": 1.0915714502334595,
+      "learning_rate": 3.7411848362809324e-06,
+      "loss": 0.3713,
+      "step": 517
+    },
+    {
+      "epoch": 5.029126213592233,
+      "grad_norm": 1.0293701887130737,
+      "learning_rate": 3.7367695474380623e-06,
+      "loss": 0.2063,
+      "step": 518
+    },
+    {
+      "epoch": 5.038834951456311,
+      "grad_norm": 1.7663441896438599,
+      "learning_rate": 3.7323491449403444e-06,
+      "loss": 0.1835,
+      "step": 519
+    },
+    {
+      "epoch": 5.048543689320389,
+      "grad_norm": 1.2588750123977661,
+      "learning_rate": 3.7279236470647593e-06,
+      "loss": 0.0947,
+      "step": 520
+    },
+    {
+      "epoch": 5.058252427184466,
+      "grad_norm": 1.9125629663467407,
+      "learning_rate": 3.723493072109355e-06,
+      "loss": 0.3472,
+      "step": 521
+    },
+    {
+      "epoch": 5.067961165048544,
+      "grad_norm": 1.6187901496887207,
+      "learning_rate": 3.719057438393172e-06,
+      "loss": 0.1696,
+      "step": 522
+    },
+    {
+      "epoch": 5.077669902912621,
+      "grad_norm": 1.0210967063903809,
+      "learning_rate": 3.714616764256166e-06,
+      "loss": 0.2502,
+      "step": 523
+    },
+    {
+      "epoch": 5.087378640776699,
+      "grad_norm": 1.0009822845458984,
+      "learning_rate": 3.7101710680591353e-06,
+      "loss": 0.1797,
+      "step": 524
+    },
+    {
+      "epoch": 5.097087378640777,
+      "grad_norm": 1.2591980695724487,
+      "learning_rate": 3.7057203681836407e-06,
+      "loss": 0.2277,
+      "step": 525
+    },
+    {
+      "epoch": 5.106796116504855,
+      "grad_norm": 1.310861349105835,
+      "learning_rate": 3.701264683031934e-06,
+      "loss": 0.2296,
+      "step": 526
+    },
+    {
+      "epoch": 5.116504854368932,
+      "grad_norm": 1.2342560291290283,
+      "learning_rate": 3.6968040310268766e-06,
+      "loss": 0.1746,
+      "step": 527
+    },
+    {
+      "epoch": 5.12621359223301,
+      "grad_norm": 1.3253822326660156,
+      "learning_rate": 3.692338430611869e-06,
+      "loss": 0.2296,
+      "step": 528
+    },
+    {
+      "epoch": 5.135922330097087,
+      "grad_norm": 1.0516308546066284,
+      "learning_rate": 3.687867900250771e-06,
+      "loss": 0.1245,
+      "step": 529
+    },
+    {
+      "epoch": 5.145631067961165,
+      "grad_norm": 1.3177831172943115,
+      "learning_rate": 3.683392458427825e-06,
+      "loss": 0.1418,
+      "step": 530
+    },
+    {
+      "epoch": 5.155339805825243,
+      "grad_norm": 1.109287142753601,
+      "learning_rate": 3.6789121236475818e-06,
+      "loss": 0.2233,
+      "step": 531
+    },
+    {
+      "epoch": 5.165048543689321,
+      "grad_norm": 1.1260926723480225,
+      "learning_rate": 3.674426914434824e-06,
+      "loss": 0.2019,
+      "step": 532
+    },
+    {
+      "epoch": 5.174757281553398,
+      "grad_norm": 1.3299827575683594,
+      "learning_rate": 3.6699368493344856e-06,
+      "loss": 0.2099,
+      "step": 533
+    },
+    {
+      "epoch": 5.184466019417476,
+      "grad_norm": 0.9847913980484009,
+      "learning_rate": 3.665441946911582e-06,
+      "loss": 0.0921,
+      "step": 534
+    },
+    {
+      "epoch": 5.194174757281553,
+      "grad_norm": 1.4914610385894775,
+      "learning_rate": 3.660942225751126e-06,
+      "loss": 0.1286,
+      "step": 535
+    },
+    {
+      "epoch": 5.203883495145631,
+      "grad_norm": 1.0563063621520996,
+      "learning_rate": 3.6564377044580558e-06,
+      "loss": 0.1488,
+      "step": 536
+    },
+    {
+      "epoch": 5.213592233009709,
+      "grad_norm": 1.1781753301620483,
+      "learning_rate": 3.6519284016571567e-06,
+      "loss": 0.1075,
+      "step": 537
+    },
+    {
+      "epoch": 5.223300970873787,
+      "grad_norm": 1.2290960550308228,
+      "learning_rate": 3.647414335992985e-06,
+      "loss": 0.2092,
+      "step": 538
+    },
+    {
+      "epoch": 5.233009708737864,
+      "grad_norm": 1.2922903299331665,
+      "learning_rate": 3.642895526129787e-06,
+      "loss": 0.216,
+      "step": 539
+    },
+    {
+      "epoch": 5.242718446601942,
+      "grad_norm": 1.2318174839019775,
+      "learning_rate": 3.638371990751428e-06,
+      "loss": 0.1835,
+      "step": 540
+    },
+    {
+      "epoch": 5.252427184466019,
+      "grad_norm": 1.2748051881790161,
+      "learning_rate": 3.63384374856131e-06,
+      "loss": 0.1683,
+      "step": 541
+    },
+    {
+      "epoch": 5.262135922330097,
+      "grad_norm": 1.7720893621444702,
+      "learning_rate": 3.629310818282297e-06,
+      "loss": 0.3087,
+      "step": 542
+    },
+    {
+      "epoch": 5.271844660194175,
+      "grad_norm": 0.8520978093147278,
+      "learning_rate": 3.6247732186566365e-06,
+      "loss": 0.1854,
+      "step": 543
+    },
+    {
+      "epoch": 5.281553398058253,
+      "grad_norm": 0.9303215742111206,
+      "learning_rate": 3.6202309684458813e-06,
+      "loss": 0.117,
+      "step": 544
+    },
+    {
+      "epoch": 5.29126213592233,
+      "grad_norm": 1.3285412788391113,
+      "learning_rate": 3.615684086430815e-06,
+      "loss": 0.1257,
+      "step": 545
+    },
+    {
+      "epoch": 5.300970873786408,
+      "grad_norm": 1.0170793533325195,
+      "learning_rate": 3.61113259141137e-06,
+      "loss": 0.1576,
+      "step": 546
+    },
+    {
+      "epoch": 5.310679611650485,
+      "grad_norm": 0.8529905080795288,
+      "learning_rate": 3.606576502206554e-06,
+      "loss": 0.0613,
+      "step": 547
+    },
+    {
+      "epoch": 5.320388349514563,
+      "grad_norm": 0.7129078507423401,
+      "learning_rate": 3.602015837654369e-06,
+      "loss": 0.0713,
+      "step": 548
+    },
+    {
+      "epoch": 5.330097087378641,
+      "grad_norm": 0.9281150102615356,
+      "learning_rate": 3.5974506166117355e-06,
+      "loss": 0.0702,
+      "step": 549
+    },
+    {
+      "epoch": 5.339805825242719,
+      "grad_norm": 1.1607550382614136,
+      "learning_rate": 3.592880857954413e-06,
+      "loss": 0.2438,
+      "step": 550
+    },
+    {
+      "epoch": 5.349514563106796,
+      "grad_norm": 2.1823811531066895,
+      "learning_rate": 3.588306580576922e-06,
+      "loss": 0.0838,
+      "step": 551
+    },
+    {
+      "epoch": 5.359223300970874,
+      "grad_norm": 0.9428465366363525,
+      "learning_rate": 3.583727803392468e-06,
+      "loss": 0.0444,
+      "step": 552
+    },
+    {
+      "epoch": 5.368932038834951,
+      "grad_norm": 0.8492387533187866,
+      "learning_rate": 3.57914454533286e-06,
+      "loss": 0.1829,
+      "step": 553
+    },
+    {
+      "epoch": 5.378640776699029,
+      "grad_norm": 1.1316030025482178,
+      "learning_rate": 3.5745568253484363e-06,
+      "loss": 0.1218,
+      "step": 554
+    },
+    {
+      "epoch": 5.388349514563107,
+      "grad_norm": 0.8767411708831787,
+      "learning_rate": 3.5699646624079824e-06,
+      "loss": 0.1396,
+      "step": 555
+    },
+    {
+      "epoch": 5.398058252427185,
+      "grad_norm": 1.06037437915802,
+      "learning_rate": 3.5653680754986543e-06,
+      "loss": 0.105,
+      "step": 556
+    },
+    {
+      "epoch": 5.407766990291262,
+      "grad_norm": 1.111646056175232,
+      "learning_rate": 3.560767083625899e-06,
+      "loss": 0.3099,
+      "step": 557
+    },
+    {
+      "epoch": 5.41747572815534,
+      "grad_norm": 1.120338797569275,
+      "learning_rate": 3.556161705813378e-06,
+      "loss": 0.1221,
+      "step": 558
+    },
+    {
+      "epoch": 5.427184466019417,
+      "grad_norm": 1.3789657354354858,
+      "learning_rate": 3.5515519611028863e-06,
+      "loss": 0.2341,
+      "step": 559
+    },
+    {
+      "epoch": 5.436893203883495,
+      "grad_norm": 0.9921037554740906,
+      "learning_rate": 3.5469378685542742e-06,
+      "loss": 0.2036,
+      "step": 560
+    },
+    {
+      "epoch": 5.446601941747573,
+      "grad_norm": 1.2706873416900635,
+      "learning_rate": 3.542319447245372e-06,
+      "loss": 0.3167,
+      "step": 561
+    },
+    {
+      "epoch": 5.456310679611651,
+      "grad_norm": 1.579001784324646,
+      "learning_rate": 3.537696716271904e-06,
+      "loss": 0.2752,
+      "step": 562
+    },
+    {
+      "epoch": 5.466019417475728,
+      "grad_norm": 1.1241706609725952,
+      "learning_rate": 3.533069694747415e-06,
+      "loss": 0.1022,
+      "step": 563
+    },
+    {
+      "epoch": 5.475728155339806,
+      "grad_norm": 0.8345101475715637,
+      "learning_rate": 3.528438401803192e-06,
+      "loss": 0.1361,
+      "step": 564
+    },
+    {
+      "epoch": 5.485436893203883,
+      "grad_norm": 0.9768244028091431,
+      "learning_rate": 3.52380285658818e-06,
+      "loss": 0.1032,
+      "step": 565
+    },
+    {
+      "epoch": 5.495145631067961,
+      "grad_norm": 1.464938998222351,
+      "learning_rate": 3.5191630782689074e-06,
+      "loss": 0.1876,
+      "step": 566
+    },
+    {
+      "epoch": 5.504854368932039,
+      "grad_norm": 0.9765818119049072,
+      "learning_rate": 3.5145190860294043e-06,
+      "loss": 0.1057,
+      "step": 567
+    },
+    {
+      "epoch": 5.514563106796117,
+      "grad_norm": 0.9651658535003662,
+      "learning_rate": 3.5098708990711254e-06,
+      "loss": 0.0653,
+      "step": 568
+    },
+    {
+      "epoch": 5.524271844660194,
+      "grad_norm": 1.2055721282958984,
+      "learning_rate": 3.505218536612869e-06,
+      "loss": 0.1895,
+      "step": 569
+    },
+    {
+      "epoch": 5.533980582524272,
+      "grad_norm": 1.0697849988937378,
+      "learning_rate": 3.500562017890695e-06,
+      "loss": 0.0845,
+      "step": 570
+    },
+    {
+      "epoch": 5.543689320388349,
+      "grad_norm": 0.6971541047096252,
+      "learning_rate": 3.495901362157853e-06,
+      "loss": 0.0609,
+      "step": 571
+    },
+    {
+      "epoch": 5.553398058252427,
+      "grad_norm": 1.18365478515625,
+      "learning_rate": 3.4912365886846934e-06,
+      "loss": 0.3142,
+      "step": 572
+    },
+    {
+      "epoch": 5.563106796116505,
+      "grad_norm": 1.1030703783035278,
+      "learning_rate": 3.4865677167585942e-06,
+      "loss": 0.1735,
+      "step": 573
+    },
+    {
+      "epoch": 5.572815533980583,
+      "grad_norm": 1.0149526596069336,
+      "learning_rate": 3.4818947656838796e-06,
+      "loss": 0.1038,
+      "step": 574
+    },
+    {
+      "epoch": 5.58252427184466,
+      "grad_norm": 0.9888204336166382,
+      "learning_rate": 3.4772177547817387e-06,
+      "loss": 0.0588,
+      "step": 575
+    },
+    {
+      "epoch": 5.592233009708738,
+      "grad_norm": 1.2715967893600464,
+      "learning_rate": 3.472536703390148e-06,
+      "loss": 0.108,
+      "step": 576
+    },
+    {
+      "epoch": 5.601941747572815,
+      "grad_norm": 0.8117223381996155,
+      "learning_rate": 3.467851630863789e-06,
+      "loss": 0.0809,
+      "step": 577
+    },
+    {
+      "epoch": 5.611650485436893,
+      "grad_norm": 0.9409249424934387,
+      "learning_rate": 3.463162556573969e-06,
+      "loss": 0.2129,
+      "step": 578
+    },
+    {
+      "epoch": 5.621359223300971,
+      "grad_norm": 1.2371786832809448,
+      "learning_rate": 3.4584694999085424e-06,
+      "loss": 0.2876,
+      "step": 579
+    },
+    {
+      "epoch": 5.631067961165049,
+      "grad_norm": 0.9954182505607605,
+      "learning_rate": 3.4537724802718294e-06,
+      "loss": 0.1089,
+      "step": 580
+    },
+    {
+      "epoch": 5.640776699029126,
+      "grad_norm": 0.8118783831596375,
+      "learning_rate": 3.4490715170845356e-06,
+      "loss": 0.1262,
+      "step": 581
+    },
+    {
+      "epoch": 5.650485436893204,
+      "grad_norm": 0.7283351421356201,
+      "learning_rate": 3.4443666297836715e-06,
+      "loss": 0.1363,
+      "step": 582
+    },
+    {
+      "epoch": 5.660194174757281,
+      "grad_norm": 1.0266672372817993,
+      "learning_rate": 3.4396578378224734e-06,
+      "loss": 0.1708,
+      "step": 583
+    },
+    {
+      "epoch": 5.669902912621359,
+      "grad_norm": 0.8701723217964172,
+      "learning_rate": 3.4349451606703214e-06,
+      "loss": 0.1945,
+      "step": 584
+    },
+    {
+      "epoch": 5.679611650485437,
+      "grad_norm": 1.1535297632217407,
+      "learning_rate": 3.430228617812661e-06,
+      "loss": 0.1021,
+      "step": 585
+    },
+    {
+      "epoch": 5.689320388349515,
+      "grad_norm": 0.7845752239227295,
+      "learning_rate": 3.4255082287509183e-06,
+      "loss": 0.0567,
+      "step": 586
+    },
+    {
+      "epoch": 5.699029126213592,
+      "grad_norm": 1.0981193780899048,
+      "learning_rate": 3.420784013002426e-06,
+      "loss": 0.1126,
+      "step": 587
+    },
+    {
+      "epoch": 5.70873786407767,
+      "grad_norm": 1.113115906715393,
+      "learning_rate": 3.416055990100336e-06,
+      "loss": 0.0997,
+      "step": 588
+    },
+    {
+      "epoch": 5.718446601941747,
+      "grad_norm": 1.045980453491211,
+      "learning_rate": 3.4113241795935427e-06,
+      "loss": 0.1004,
+      "step": 589
+    },
+    {
+      "epoch": 5.728155339805825,
+      "grad_norm": 0.9982129335403442,
+      "learning_rate": 3.4065886010466014e-06,
+      "loss": 0.1795,
+      "step": 590
+    },
+    {
+      "epoch": 5.737864077669903,
+      "grad_norm": 1.106252670288086,
+      "learning_rate": 3.401849274039647e-06,
+      "loss": 0.1923,
+      "step": 591
+    },
+    {
+      "epoch": 5.747572815533981,
+      "grad_norm": 0.9183070063591003,
+      "learning_rate": 3.3971062181683117e-06,
+      "loss": 0.1105,
+      "step": 592
+    },
+    {
+      "epoch": 5.757281553398058,
+      "grad_norm": 1.006274938583374,
+      "learning_rate": 3.3923594530436477e-06,
+      "loss": 0.198,
+      "step": 593
+    },
+    {
+      "epoch": 5.766990291262136,
+      "grad_norm": 1.1224919557571411,
+      "learning_rate": 3.387608998292041e-06,
+      "loss": 0.1351,
+      "step": 594
+    },
+    {
+      "epoch": 5.776699029126213,
+      "grad_norm": 0.8881707191467285,
+      "learning_rate": 3.382854873555137e-06,
+      "loss": 0.0772,
+      "step": 595
+    },
+    {
+      "epoch": 5.786407766990291,
+      "grad_norm": 1.3855921030044556,
+      "learning_rate": 3.3780970984897504e-06,
+      "loss": 0.0927,
+      "step": 596
+    },
+    {
+      "epoch": 5.796116504854369,
+      "grad_norm": 1.5147404670715332,
+      "learning_rate": 3.373335692767793e-06,
+      "loss": 0.2161,
+      "step": 597
+    },
+    {
+      "epoch": 5.805825242718447,
+      "grad_norm": 1.1337159872055054,
+      "learning_rate": 3.3685706760761865e-06,
+      "loss": 0.1655,
+      "step": 598
+    },
+    {
+      "epoch": 5.815533980582524,
+      "grad_norm": 1.1046392917633057,
+      "learning_rate": 3.3638020681167827e-06,
+      "loss": 0.1096,
+      "step": 599
+    },
+    {
+      "epoch": 5.825242718446602,
+      "grad_norm": 0.7068084478378296,
+      "learning_rate": 3.3590298886062833e-06,
+      "loss": 0.1372,
+      "step": 600
+    },
+    {
+      "epoch": 5.834951456310679,
+      "grad_norm": 1.148061990737915,
+      "learning_rate": 3.354254157276155e-06,
+      "loss": 0.2396,
+      "step": 601
+    },
+    {
+      "epoch": 5.844660194174757,
+      "grad_norm": 0.9790175557136536,
+      "learning_rate": 3.3494748938725525e-06,
+      "loss": 0.1436,
+      "step": 602
+    },
+    {
+      "epoch": 5.854368932038835,
+      "grad_norm": 1.1377002000808716,
+      "learning_rate": 3.3446921181562326e-06,
+      "loss": 0.1633,
+      "step": 603
+    },
+    {
+      "epoch": 5.864077669902913,
+      "grad_norm": 0.8733611702919006,
+      "learning_rate": 3.3399058499024767e-06,
+      "loss": 0.0667,
+      "step": 604
+    },
+    {
+      "epoch": 5.87378640776699,
+      "grad_norm": 1.4806839227676392,
+      "learning_rate": 3.3351161089010055e-06,
+      "loss": 0.186,
+      "step": 605
+    },
+    {
+      "epoch": 5.883495145631068,
+      "grad_norm": 1.0615733861923218,
+      "learning_rate": 3.330322914955897e-06,
+      "loss": 0.1802,
+      "step": 606
+    },
+    {
+      "epoch": 5.893203883495145,
+      "grad_norm": 1.1709858179092407,
+      "learning_rate": 3.325526287885509e-06,
+      "loss": 0.1916,
+      "step": 607
+    },
+    {
+      "epoch": 5.902912621359223,
+      "grad_norm": 1.0924482345581055,
+      "learning_rate": 3.3207262475223913e-06,
+      "loss": 0.1862,
+      "step": 608
+    },
+    {
+      "epoch": 5.9126213592233015,
+      "grad_norm": 1.733832836151123,
+      "learning_rate": 3.315922813713209e-06,
+      "loss": 0.3237,
+      "step": 609
+    },
+    {
+      "epoch": 5.922330097087379,
+      "grad_norm": 0.9331390857696533,
+      "learning_rate": 3.3111160063186553e-06,
+      "loss": 0.1006,
+      "step": 610
+    },
+    {
+      "epoch": 5.932038834951456,
+      "grad_norm": 1.0888906717300415,
+      "learning_rate": 3.3063058452133756e-06,
+      "loss": 0.2272,
+      "step": 611
+    },
+    {
+      "epoch": 5.941747572815534,
+      "grad_norm": 0.9720425605773926,
+      "learning_rate": 3.301492350285879e-06,
+      "loss": 0.1531,
+      "step": 612
+    },
+    {
+      "epoch": 5.951456310679612,
+      "grad_norm": 1.0783334970474243,
+      "learning_rate": 3.296675541438461e-06,
+      "loss": 0.0999,
+      "step": 613
+    },
+    {
+      "epoch": 5.961165048543689,
+      "grad_norm": 1.107725739479065,
+      "learning_rate": 3.2918554385871163e-06,
+      "loss": 0.2651,
+      "step": 614
+    },
+    {
+      "epoch": 5.970873786407767,
+      "grad_norm": 0.7212538719177246,
+      "learning_rate": 3.2870320616614626e-06,
+      "loss": 0.0738,
+      "step": 615
+    },
+    {
+      "epoch": 5.980582524271845,
+      "grad_norm": 0.9390976428985596,
+      "learning_rate": 3.282205430604653e-06,
+      "loss": 0.1179,
+      "step": 616
+    },
+    {
+      "epoch": 5.990291262135923,
+      "grad_norm": 1.0487676858901978,
+      "learning_rate": 3.2773755653732954e-06,
+      "loss": 0.1506,
+      "step": 617
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.952416718006134,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.1571,
+      "step": 618
+    },
+    {
+      "epoch": 6.009708737864078,
+      "grad_norm": 1.1496444940567017,
+      "learning_rate": 3.267706212280146e-06,
+      "loss": 0.0913,
+      "step": 619
+    },
+    {
+      "epoch": 6.019417475728155,
+      "grad_norm": 1.0084459781646729,
+      "learning_rate": 3.2628667643981036e-06,
+      "loss": 0.0861,
+      "step": 620
+    },
+    {
+      "epoch": 6.029126213592233,
+      "grad_norm": 1.3622573614120483,
+      "learning_rate": 3.2580241623008426e-06,
+      "loss": 0.1241,
+      "step": 621
+    },
+    {
+      "epoch": 6.038834951456311,
+      "grad_norm": 1.3345288038253784,
+      "learning_rate": 3.2531784260110067e-06,
+      "loss": 0.173,
+      "step": 622
+    },
+    {
+      "epoch": 6.048543689320389,
+      "grad_norm": 1.3314571380615234,
+      "learning_rate": 3.2483295755641986e-06,
+      "loss": 0.1099,
+      "step": 623
+    },
+    {
+      "epoch": 6.058252427184466,
+      "grad_norm": 1.4012583494186401,
+      "learning_rate": 3.243477631008897e-06,
+      "loss": 0.1808,
+      "step": 624
+    },
+    {
+      "epoch": 6.067961165048544,
+      "grad_norm": 1.1909350156784058,
+      "learning_rate": 3.238622612406373e-06,
+      "loss": 0.0395,
+      "step": 625
+    },
+    {
+      "epoch": 6.077669902912621,
+      "grad_norm": 1.5543804168701172,
+      "learning_rate": 3.233764539830608e-06,
+      "loss": 0.1112,
+      "step": 626
+    },
+    {
+      "epoch": 6.087378640776699,
+      "grad_norm": 0.7781279683113098,
+      "learning_rate": 3.228903433368212e-06,
+      "loss": 0.0314,
+      "step": 627
+    },
+    {
+      "epoch": 6.097087378640777,
+      "grad_norm": 0.5457009077072144,
+      "learning_rate": 3.224039313118338e-06,
+      "loss": 0.0346,
+      "step": 628
+    },
+    {
+      "epoch": 6.106796116504855,
+      "grad_norm": 0.9449105262756348,
+      "learning_rate": 3.2191721991925993e-06,
+      "loss": 0.0701,
+      "step": 629
+    },
+    {
+      "epoch": 6.116504854368932,
+      "grad_norm": 1.0498558282852173,
+      "learning_rate": 3.21430211171499e-06,
+      "loss": 0.1421,
+      "step": 630
+    },
+    {
+      "epoch": 6.12621359223301,
+      "grad_norm": 1.1856399774551392,
+      "learning_rate": 3.209429070821795e-06,
+      "loss": 0.2414,
+      "step": 631
+    },
+    {
+      "epoch": 6.135922330097087,
+      "grad_norm": 1.2752889394760132,
+      "learning_rate": 3.2045530966615136e-06,
+      "loss": 0.1537,
+      "step": 632
+    },
+    {
+      "epoch": 6.145631067961165,
+      "grad_norm": 0.9089634418487549,
+      "learning_rate": 3.1996742093947724e-06,
+      "loss": 0.1174,
+      "step": 633
+    },
+    {
+      "epoch": 6.155339805825243,
+      "grad_norm": 48.404693603515625,
+      "learning_rate": 3.1947924291942423e-06,
+      "loss": 0.2074,
+      "step": 634
+    },
+    {
+      "epoch": 6.165048543689321,
+      "grad_norm": 1.4127225875854492,
+      "learning_rate": 3.189907776244556e-06,
+      "loss": 0.0639,
+      "step": 635
+    },
+    {
+      "epoch": 6.174757281553398,
+      "grad_norm": 1.9454190731048584,
+      "learning_rate": 3.185020270742225e-06,
+      "loss": 0.145,
+      "step": 636
+    },
+    {
+      "epoch": 6.184466019417476,
+      "grad_norm": 0.8144808411598206,
+      "learning_rate": 3.180129932895553e-06,
+      "loss": 0.0605,
+      "step": 637
+    },
+    {
+      "epoch": 6.194174757281553,
+      "grad_norm": 1.178910732269287,
+      "learning_rate": 3.1752367829245563e-06,
+      "loss": 0.2024,
+      "step": 638
+    },
+    {
+      "epoch": 6.203883495145631,
+      "grad_norm": 1.5138559341430664,
+      "learning_rate": 3.1703408410608777e-06,
+      "loss": 0.2836,
+      "step": 639
+    },
+    {
+      "epoch": 6.213592233009709,
+      "grad_norm": 1.081778883934021,
+      "learning_rate": 3.1654421275477045e-06,
+      "loss": 0.0874,
+      "step": 640
+    },
+    {
+      "epoch": 6.223300970873787,
+      "grad_norm": 1.0488624572753906,
+      "learning_rate": 3.1605406626396826e-06,
+      "loss": 0.081,
+      "step": 641
+    },
+    {
+      "epoch": 6.233009708737864,
+      "grad_norm": 1.043975591659546,
+      "learning_rate": 3.155636466602836e-06,
+      "loss": 0.1343,
+      "step": 642
+    },
+    {
+      "epoch": 6.242718446601942,
+      "grad_norm": 1.2351950407028198,
+      "learning_rate": 3.150729559714478e-06,
+      "loss": 0.1853,
+      "step": 643
+    },
+    {
+      "epoch": 6.252427184466019,
+      "grad_norm": 1.028315544128418,
+      "learning_rate": 3.145819962263134e-06,
+      "loss": 0.1328,
+      "step": 644
+    },
+    {
+      "epoch": 6.262135922330097,
+      "grad_norm": 0.966011643409729,
+      "learning_rate": 3.1409076945484513e-06,
+      "loss": 0.0671,
+      "step": 645
+    },
+    {
+      "epoch": 6.271844660194175,
+      "grad_norm": 1.0541956424713135,
+      "learning_rate": 3.135992776881119e-06,
+      "loss": 0.0451,
+      "step": 646
+    },
+    {
+      "epoch": 6.281553398058253,
+      "grad_norm": 1.1905815601348877,
+      "learning_rate": 3.1310752295827818e-06,
+      "loss": 0.0375,
+      "step": 647
+    },
+    {
+      "epoch": 6.29126213592233,
+      "grad_norm": 1.232605218887329,
+      "learning_rate": 3.1261550729859602e-06,
+      "loss": 0.1158,
+      "step": 648
+    },
+    {
+      "epoch": 6.300970873786408,
+      "grad_norm": 1.013516902923584,
+      "learning_rate": 3.12123232743396e-06,
+      "loss": 0.0596,
+      "step": 649
+    },
+    {
+      "epoch": 6.310679611650485,
+      "grad_norm": 1.258267879486084,
+      "learning_rate": 3.116307013280793e-06,
+      "loss": 0.2079,
+      "step": 650
+    },
+    {
+      "epoch": 6.320388349514563,
+      "grad_norm": 0.8476516008377075,
+      "learning_rate": 3.1113791508910913e-06,
+      "loss": 0.1106,
+      "step": 651
+    },
+    {
+      "epoch": 6.330097087378641,
+      "grad_norm": 1.2680771350860596,
+      "learning_rate": 3.106448760640022e-06,
+      "loss": 0.1778,
+      "step": 652
+    },
+    {
+      "epoch": 6.339805825242719,
+      "grad_norm": 0.8958115577697754,
+      "learning_rate": 3.1015158629132066e-06,
+      "loss": 0.0379,
+      "step": 653
+    },
+    {
+      "epoch": 6.349514563106796,
+      "grad_norm": 1.1448755264282227,
+      "learning_rate": 3.096580478106631e-06,
+      "loss": 0.1273,
+      "step": 654
+    },
+    {
+      "epoch": 6.359223300970874,
+      "grad_norm": 0.9504626393318176,
+      "learning_rate": 3.0916426266265676e-06,
+      "loss": 0.077,
+      "step": 655
+    },
+    {
+      "epoch": 6.368932038834951,
+      "grad_norm": 1.1011178493499756,
+      "learning_rate": 3.086702328889486e-06,
+      "loss": 0.0519,
+      "step": 656
+    },
+    {
+      "epoch": 6.378640776699029,
+      "grad_norm": 0.9548404216766357,
+      "learning_rate": 3.0817596053219697e-06,
+      "loss": 0.0942,
+      "step": 657
+    },
+    {
+      "epoch": 6.388349514563107,
+      "grad_norm": 0.8062845468521118,
+      "learning_rate": 3.076814476360634e-06,
+      "loss": 0.0985,
+      "step": 658
+    },
+    {
+      "epoch": 6.398058252427185,
+      "grad_norm": 0.608669102191925,
+      "learning_rate": 3.071866962452038e-06,
+      "loss": 0.0501,
+      "step": 659
+    },
+    {
+      "epoch": 6.407766990291262,
+      "grad_norm": 1.308223009109497,
+      "learning_rate": 3.066917084052603e-06,
+      "loss": 0.1855,
+      "step": 660
+    },
+    {
+      "epoch": 6.41747572815534,
+      "grad_norm": 1.656266450881958,
+      "learning_rate": 3.061964861628527e-06,
+      "loss": 0.2386,
+      "step": 661
+    },
+    {
+      "epoch": 6.427184466019417,
+      "grad_norm": 0.7621467709541321,
+      "learning_rate": 3.057010315655698e-06,
+      "loss": 0.0692,
+      "step": 662
+    },
+    {
+      "epoch": 6.436893203883495,
+      "grad_norm": 1.2241672277450562,
+      "learning_rate": 3.0520534666196134e-06,
+      "loss": 0.1118,
+      "step": 663
+    },
+    {
+      "epoch": 6.446601941747573,
+      "grad_norm": 0.9063080549240112,
+      "learning_rate": 3.0470943350152914e-06,
+      "loss": 0.1127,
+      "step": 664
+    },
+    {
+      "epoch": 6.456310679611651,
+      "grad_norm": 0.880898654460907,
+      "learning_rate": 3.042132941347189e-06,
+      "loss": 0.189,
+      "step": 665
+    },
+    {
+      "epoch": 6.466019417475728,
+      "grad_norm": 0.8484269380569458,
+      "learning_rate": 3.037169306129115e-06,
+      "loss": 0.0735,
+      "step": 666
+    },
+    {
+      "epoch": 6.475728155339806,
+      "grad_norm": 0.9555432200431824,
+      "learning_rate": 3.0322034498841475e-06,
+      "loss": 0.122,
+      "step": 667
+    },
+    {
+      "epoch": 6.485436893203883,
+      "grad_norm": 0.981014609336853,
+      "learning_rate": 3.027235393144547e-06,
+      "loss": 0.125,
+      "step": 668
+    },
+    {
+      "epoch": 6.495145631067961,
+      "grad_norm": 1.0492215156555176,
+      "learning_rate": 3.0222651564516715e-06,
+      "loss": 0.1992,
+      "step": 669
+    },
+    {
+      "epoch": 6.504854368932039,
+      "grad_norm": 1.2064902782440186,
+      "learning_rate": 3.017292760355896e-06,
+      "loss": 0.0663,
+      "step": 670
+    },
+    {
+      "epoch": 6.514563106796117,
+      "grad_norm": 1.1765069961547852,
+      "learning_rate": 3.0123182254165194e-06,
+      "loss": 0.0401,
+      "step": 671
+    },
+    {
+      "epoch": 6.524271844660194,
+      "grad_norm": 1.5717885494232178,
+      "learning_rate": 3.0073415722016875e-06,
+      "loss": 0.2454,
+      "step": 672
+    },
+    {
+      "epoch": 6.533980582524272,
+      "grad_norm": 1.1897467374801636,
+      "learning_rate": 3.002362821288302e-06,
+      "loss": 0.061,
+      "step": 673
+    },
+    {
+      "epoch": 6.543689320388349,
+      "grad_norm": 1.0062282085418701,
+      "learning_rate": 2.9973819932619404e-06,
+      "loss": 0.1018,
+      "step": 674
+    },
+    {
+      "epoch": 6.553398058252427,
+      "grad_norm": 1.0144904851913452,
+      "learning_rate": 2.9923991087167657e-06,
+      "loss": 0.1027,
+      "step": 675
+    },
+    {
+      "epoch": 6.563106796116505,
+      "grad_norm": 0.9750237464904785,
+      "learning_rate": 2.987414188255446e-06,
+      "loss": 0.1482,
+      "step": 676
+    },
+    {
+      "epoch": 6.572815533980583,
+      "grad_norm": 1.158136010169983,
+      "learning_rate": 2.9824272524890664e-06,
+      "loss": 0.1345,
+      "step": 677
+    },
+    {
+      "epoch": 6.58252427184466,
+      "grad_norm": 1.1052918434143066,
+      "learning_rate": 2.977438322037046e-06,
+      "loss": 0.0791,
+      "step": 678
+    },
+    {
+      "epoch": 6.592233009708738,
+      "grad_norm": 0.8905234932899475,
+      "learning_rate": 2.9724474175270485e-06,
+      "loss": 0.0684,
+      "step": 679
+    },
+    {
+      "epoch": 6.601941747572815,
+      "grad_norm": 1.0031498670578003,
+      "learning_rate": 2.967454559594903e-06,
+      "loss": 0.0628,
+      "step": 680
+    },
+    {
+      "epoch": 6.611650485436893,
+      "grad_norm": 2.124293327331543,
+      "learning_rate": 2.9624597688845126e-06,
+      "loss": 0.1071,
+      "step": 681
+    },
+    {
+      "epoch": 6.621359223300971,
+      "grad_norm": 1.101279854774475,
+      "learning_rate": 2.957463066047773e-06,
+      "loss": 0.0892,
+      "step": 682
+    },
+    {
+      "epoch": 6.631067961165049,
+      "grad_norm": 0.6379441618919373,
+      "learning_rate": 2.9524644717444866e-06,
+      "loss": 0.0469,
+      "step": 683
+    },
+    {
+      "epoch": 6.640776699029126,
+      "grad_norm": 1.4411451816558838,
+      "learning_rate": 2.9474640066422757e-06,
+      "loss": 0.2329,
+      "step": 684
+    },
+    {
+      "epoch": 6.650485436893204,
+      "grad_norm": 1.02129065990448,
+      "learning_rate": 2.9424616914164982e-06,
+      "loss": 0.0619,
+      "step": 685
+    },
+    {
+      "epoch": 6.660194174757281,
+      "grad_norm": 0.7449476718902588,
+      "learning_rate": 2.9374575467501605e-06,
+      "loss": 0.063,
+      "step": 686
+    },
+    {
+      "epoch": 6.669902912621359,
+      "grad_norm": 1.1670392751693726,
+      "learning_rate": 2.9324515933338343e-06,
+      "loss": 0.0926,
+      "step": 687
+    },
+    {
+      "epoch": 6.679611650485437,
+      "grad_norm": 1.2292739152908325,
+      "learning_rate": 2.9274438518655703e-06,
+      "loss": 0.1585,
+      "step": 688
+    },
+    {
+      "epoch": 6.689320388349515,
+      "grad_norm": 0.9029463529586792,
+      "learning_rate": 2.9224343430508105e-06,
+      "loss": 0.0726,
+      "step": 689
+    },
+    {
+      "epoch": 6.699029126213592,
+      "grad_norm": 0.9147337079048157,
+      "learning_rate": 2.917423087602306e-06,
+      "loss": 0.1295,
+      "step": 690
+    },
+    {
+      "epoch": 6.70873786407767,
+      "grad_norm": 0.9737975597381592,
+      "learning_rate": 2.9124101062400283e-06,
+      "loss": 0.153,
+      "step": 691
+    },
+    {
+      "epoch": 6.718446601941747,
+      "grad_norm": 1.2561050653457642,
+      "learning_rate": 2.907395419691087e-06,
+      "loss": 0.178,
+      "step": 692
+    },
+    {
+      "epoch": 6.728155339805825,
+      "grad_norm": 1.113938808441162,
+      "learning_rate": 2.9023790486896404e-06,
+      "loss": 0.081,
+      "step": 693
+    },
+    {
+      "epoch": 6.737864077669903,
+      "grad_norm": 0.82021164894104,
+      "learning_rate": 2.8973610139768114e-06,
+      "loss": 0.0724,
+      "step": 694
+    },
+    {
+      "epoch": 6.747572815533981,
+      "grad_norm": 9.422320365905762,
+      "learning_rate": 2.8923413363006038e-06,
+      "loss": 0.1952,
+      "step": 695
+    },
+    {
+      "epoch": 6.757281553398058,
+      "grad_norm": 0.9243801236152649,
+      "learning_rate": 2.887320036415811e-06,
+      "loss": 0.0708,
+      "step": 696
+    },
+    {
+      "epoch": 6.766990291262136,
+      "grad_norm": 1.160498857498169,
+      "learning_rate": 2.882297135083937e-06,
+      "loss": 0.0429,
+      "step": 697
+    },
+    {
+      "epoch": 6.776699029126213,
+      "grad_norm": 0.9941173791885376,
+      "learning_rate": 2.877272653073107e-06,
+      "loss": 0.0608,
+      "step": 698
+    },
+    {
+      "epoch": 6.786407766990291,
+      "grad_norm": 1.100027084350586,
+      "learning_rate": 2.87224661115798e-06,
+      "loss": 0.096,
+      "step": 699
+    },
+    {
+      "epoch": 6.796116504854369,
+      "grad_norm": 0.8699636459350586,
+      "learning_rate": 2.8672190301196655e-06,
+      "loss": 0.1668,
+      "step": 700
+    },
+    {
+      "epoch": 6.805825242718447,
+      "grad_norm": 1.1691793203353882,
+      "learning_rate": 2.8621899307456376e-06,
+      "loss": 0.1234,
+      "step": 701
+    },
+    {
+      "epoch": 6.815533980582524,
+      "grad_norm": 1.030217170715332,
+      "learning_rate": 2.8571593338296473e-06,
+      "loss": 0.09,
+      "step": 702
+    },
+    {
+      "epoch": 6.825242718446602,
+      "grad_norm": 0.5228599905967712,
+      "learning_rate": 2.8521272601716376e-06,
+      "loss": 0.0317,
+      "step": 703
+    },
+    {
+      "epoch": 6.834951456310679,
+      "grad_norm": 1.4643433094024658,
+      "learning_rate": 2.8470937305776567e-06,
+      "loss": 0.2609,
+      "step": 704
+    },
+    {
+      "epoch": 6.844660194174757,
+      "grad_norm": 0.9849478006362915,
+      "learning_rate": 2.842058765859776e-06,
+      "loss": 0.0848,
+      "step": 705
+    },
+    {
+      "epoch": 6.854368932038835,
+      "grad_norm": 0.9210445284843445,
+      "learning_rate": 2.837022386835996e-06,
+      "loss": 0.0809,
+      "step": 706
+    },
+    {
+      "epoch": 6.864077669902913,
+      "grad_norm": 0.8688995242118835,
+      "learning_rate": 2.8319846143301676e-06,
+      "loss": 0.064,
+      "step": 707
+    },
+    {
+      "epoch": 6.87378640776699,
+      "grad_norm": 0.8066009879112244,
+      "learning_rate": 2.826945469171903e-06,
+      "loss": 0.0467,
+      "step": 708
+    },
+    {
+      "epoch": 6.883495145631068,
+      "grad_norm": 0.6212665438652039,
+      "learning_rate": 2.82190497219649e-06,
+      "loss": 0.0793,
+      "step": 709
+    },
+    {
+      "epoch": 6.893203883495145,
+      "grad_norm": 1.0815949440002441,
+      "learning_rate": 2.8168631442448046e-06,
+      "loss": 0.1624,
+      "step": 710
+    },
+    {
+      "epoch": 6.902912621359223,
+      "grad_norm": 0.7521616816520691,
+      "learning_rate": 2.8118200061632273e-06,
+      "loss": 0.0374,
+      "step": 711
+    },
+    {
+      "epoch": 6.9126213592233015,
+      "grad_norm": 1.0690211057662964,
+      "learning_rate": 2.8067755788035544e-06,
+      "loss": 0.1545,
+      "step": 712
+    },
+    {
+      "epoch": 6.922330097087379,
+      "grad_norm": 1.1906397342681885,
+      "learning_rate": 2.801729883022915e-06,
+      "loss": 0.0859,
+      "step": 713
+    },
+    {
+      "epoch": 6.932038834951456,
+      "grad_norm": 0.871376633644104,
+      "learning_rate": 2.7966829396836804e-06,
+      "loss": 0.1442,
+      "step": 714
+    },
+    {
+      "epoch": 6.941747572815534,
+      "grad_norm": 1.062356948852539,
+      "learning_rate": 2.791634769653381e-06,
+      "loss": 0.0734,
+      "step": 715
+    },
+    {
+      "epoch": 6.951456310679612,
+      "grad_norm": 0.9127151966094971,
+      "learning_rate": 2.78658539380462e-06,
+      "loss": 0.2102,
+      "step": 716
+    },
+    {
+      "epoch": 6.961165048543689,
+      "grad_norm": 0.8193743228912354,
+      "learning_rate": 2.781534833014985e-06,
+      "loss": 0.1019,
+      "step": 717
+    },
+    {
+      "epoch": 6.970873786407767,
+      "grad_norm": 0.989223301410675,
+      "learning_rate": 2.7764831081669635e-06,
+      "loss": 0.1714,
+      "step": 718
+    },
+    {
+      "epoch": 6.980582524271845,
+      "grad_norm": 0.6328715085983276,
+      "learning_rate": 2.771430240147856e-06,
+      "loss": 0.0497,
+      "step": 719
+    },
+    {
+      "epoch": 6.990291262135923,
+      "grad_norm": 1.1424680948257446,
+      "learning_rate": 2.7663762498496905e-06,
+      "loss": 0.1421,
+      "step": 720
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 1.1053463220596313,
+      "learning_rate": 2.761321158169134e-06,
+      "loss": 0.0579,
+      "step": 721
+    },
+    {
+      "epoch": 7.009708737864078,
+      "grad_norm": 0.916090726852417,
+      "learning_rate": 2.7562649860074077e-06,
+      "loss": 0.051,
+      "step": 722
+    },
+    {
+      "epoch": 7.019417475728155,
+      "grad_norm": 1.599184513092041,
+      "learning_rate": 2.7512077542702005e-06,
+      "loss": 0.2136,
+      "step": 723
+    },
+    {
+      "epoch": 7.029126213592233,
+      "grad_norm": 0.8245609998703003,
+      "learning_rate": 2.746149483867582e-06,
+      "loss": 0.0343,
+      "step": 724
+    },
+    {
+      "epoch": 7.038834951456311,
+      "grad_norm": 0.9597436785697937,
+      "learning_rate": 2.741090195713917e-06,
+      "loss": 0.0577,
+      "step": 725
+    },
+    {
+      "epoch": 7.048543689320389,
+      "grad_norm": 1.0133090019226074,
+      "learning_rate": 2.736029910727777e-06,
+      "loss": 0.0816,
+      "step": 726
+    },
+    {
+      "epoch": 7.058252427184466,
+      "grad_norm": 1.2042464017868042,
+      "learning_rate": 2.730968649831858e-06,
+      "loss": 0.064,
+      "step": 727
+    },
+    {
+      "epoch": 7.067961165048544,
+      "grad_norm": 1.2887951135635376,
+      "learning_rate": 2.7259064339528875e-06,
+      "loss": 0.1278,
+      "step": 728
+    },
+    {
+      "epoch": 7.077669902912621,
+      "grad_norm": 1.1621922254562378,
+      "learning_rate": 2.720843284021543e-06,
+      "loss": 0.107,
+      "step": 729
+    },
+    {
+      "epoch": 7.087378640776699,
+      "grad_norm": 0.6943169832229614,
+      "learning_rate": 2.7157792209723654e-06,
+      "loss": 0.0554,
+      "step": 730
+    },
+    {
+      "epoch": 7.097087378640777,
+      "grad_norm": 0.9973244667053223,
+      "learning_rate": 2.7107142657436696e-06,
+      "loss": 0.0963,
+      "step": 731
+    },
+    {
+      "epoch": 7.106796116504855,
+      "grad_norm": 0.9302243590354919,
+      "learning_rate": 2.705648439277459e-06,
+      "loss": 0.043,
+      "step": 732
+    },
+    {
+      "epoch": 7.116504854368932,
+      "grad_norm": 0.6210493445396423,
+      "learning_rate": 2.7005817625193398e-06,
+      "loss": 0.0524,
+      "step": 733
+    },
+    {
+      "epoch": 7.12621359223301,
+      "grad_norm": 0.792597770690918,
+      "learning_rate": 2.695514256418435e-06,
+      "loss": 0.0359,
+      "step": 734
+    },
+    {
+      "epoch": 7.135922330097087,
+      "grad_norm": 1.1926745176315308,
+      "learning_rate": 2.6904459419272955e-06,
+      "loss": 0.0771,
+      "step": 735
+    },
+    {
+      "epoch": 7.145631067961165,
+      "grad_norm": 1.0876519680023193,
+      "learning_rate": 2.685376840001814e-06,
+      "loss": 0.0669,
+      "step": 736
+    },
+    {
+      "epoch": 7.155339805825243,
+      "grad_norm": 1.0764493942260742,
+      "learning_rate": 2.6803069716011405e-06,
+      "loss": 0.0242,
+      "step": 737
+    },
+    {
+      "epoch": 7.165048543689321,
+      "grad_norm": 1.308888554573059,
+      "learning_rate": 2.6752363576875933e-06,
+      "loss": 0.0749,
+      "step": 738
+    },
+    {
+      "epoch": 7.174757281553398,
+      "grad_norm": 1.2870591878890991,
+      "learning_rate": 2.6701650192265734e-06,
+      "loss": 0.0776,
+      "step": 739
+    },
+    {
+      "epoch": 7.184466019417476,
+      "grad_norm": 1.0227237939834595,
+      "learning_rate": 2.6650929771864776e-06,
+      "loss": 0.0491,
+      "step": 740
+    },
+    {
+      "epoch": 7.194174757281553,
+      "grad_norm": 1.2830699682235718,
+      "learning_rate": 2.660020252538611e-06,
+      "loss": 0.121,
+      "step": 741
+    },
+    {
+      "epoch": 7.203883495145631,
+      "grad_norm": 0.7571777701377869,
+      "learning_rate": 2.6549468662571026e-06,
+      "loss": 0.0598,
+      "step": 742
+    },
+    {
+      "epoch": 7.213592233009709,
+      "grad_norm": 1.2842772006988525,
+      "learning_rate": 2.6498728393188157e-06,
+      "loss": 0.1016,
+      "step": 743
+    },
+    {
+      "epoch": 7.223300970873787,
+      "grad_norm": 0.9173499345779419,
+      "learning_rate": 2.6447981927032634e-06,
+      "loss": 0.0308,
+      "step": 744
+    },
+    {
+      "epoch": 7.233009708737864,
+      "grad_norm": 1.0887157917022705,
+      "learning_rate": 2.639722947392521e-06,
+      "loss": 0.1625,
+      "step": 745
+    },
+    {
+      "epoch": 7.242718446601942,
+      "grad_norm": 0.8496439456939697,
+      "learning_rate": 2.6346471243711376e-06,
+      "loss": 0.061,
+      "step": 746
+    },
+    {
+      "epoch": 7.252427184466019,
+      "grad_norm": 0.7860584855079651,
+      "learning_rate": 2.629570744626052e-06,
+      "loss": 0.1061,
+      "step": 747
+    },
+    {
+      "epoch": 7.262135922330097,
+      "grad_norm": 0.8548651933670044,
+      "learning_rate": 2.624493829146507e-06,
+      "loss": 0.0929,
+      "step": 748
+    },
+    {
+      "epoch": 7.271844660194175,
+      "grad_norm": 1.2701689004898071,
+      "learning_rate": 2.619416398923957e-06,
+      "loss": 0.1547,
+      "step": 749
+    },
+    {
+      "epoch": 7.281553398058253,
+      "grad_norm": 0.8744873404502869,
+      "learning_rate": 2.614338474951987e-06,
+      "loss": 0.022,
+      "step": 750
+    },
+    {
+      "epoch": 7.29126213592233,
+      "grad_norm": 0.7233145236968994,
+      "learning_rate": 2.6092600782262213e-06,
+      "loss": 0.0523,
+      "step": 751
+    },
+    {
+      "epoch": 7.300970873786408,
+      "grad_norm": 0.7879371643066406,
+      "learning_rate": 2.6041812297442417e-06,
+      "loss": 0.044,
+      "step": 752
+    },
+    {
+      "epoch": 7.310679611650485,
+      "grad_norm": 1.221299409866333,
+      "learning_rate": 2.5991019505054965e-06,
+      "loss": 0.0541,
+      "step": 753
+    },
+    {
+      "epoch": 7.320388349514563,
+      "grad_norm": 0.7391517758369446,
+      "learning_rate": 2.5940222615112143e-06,
+      "loss": 0.0332,
+      "step": 754
+    },
+    {
+      "epoch": 7.330097087378641,
+      "grad_norm": 0.7787492871284485,
+      "learning_rate": 2.5889421837643186e-06,
+      "loss": 0.0446,
+      "step": 755
+    },
+    {
+      "epoch": 7.339805825242719,
+      "grad_norm": 6.488763332366943,
+      "learning_rate": 2.5838617382693415e-06,
+      "loss": 0.1187,
+      "step": 756
+    },
+    {
+      "epoch": 7.349514563106796,
+      "grad_norm": 1.3808194398880005,
+      "learning_rate": 2.5787809460323337e-06,
+      "loss": 0.0555,
+      "step": 757
+    },
+    {
+      "epoch": 7.359223300970874,
+      "grad_norm": 1.1066080331802368,
+      "learning_rate": 2.57369982806078e-06,
+      "loss": 0.105,
+      "step": 758
+    },
+    {
+      "epoch": 7.368932038834951,
+      "grad_norm": 0.9817208647727966,
+      "learning_rate": 2.5686184053635127e-06,
+      "loss": 0.0499,
+      "step": 759
+    },
+    {
+      "epoch": 7.378640776699029,
+      "grad_norm": 0.6625164747238159,
+      "learning_rate": 2.563536698950624e-06,
+      "loss": 0.0348,
+      "step": 760
+    },
+    {
+      "epoch": 7.388349514563107,
+      "grad_norm": 0.5409085750579834,
+      "learning_rate": 2.5584547298333772e-06,
+      "loss": 0.0123,
+      "step": 761
+    },
+    {
+      "epoch": 7.398058252427185,
+      "grad_norm": 0.6516930460929871,
+      "learning_rate": 2.5533725190241255e-06,
+      "loss": 0.0101,
+      "step": 762
+    },
+    {
+      "epoch": 7.407766990291262,
+      "grad_norm": 0.7117874026298523,
+      "learning_rate": 2.5482900875362184e-06,
+      "loss": 0.0352,
+      "step": 763
+    },
+    {
+      "epoch": 7.41747572815534,
+      "grad_norm": 1.0977354049682617,
+      "learning_rate": 2.543207456383919e-06,
+      "loss": 0.0468,
+      "step": 764
+    },
+    {
+      "epoch": 7.427184466019417,
+      "grad_norm": 1.0205830335617065,
+      "learning_rate": 2.538124646582315e-06,
+      "loss": 0.1731,
+      "step": 765
+    },
+    {
+      "epoch": 7.436893203883495,
+      "grad_norm": 0.7433698177337646,
+      "learning_rate": 2.533041679147235e-06,
+      "loss": 0.0431,
+      "step": 766
+    },
+    {
+      "epoch": 7.446601941747573,
+      "grad_norm": 0.7881643176078796,
+      "learning_rate": 2.527958575095157e-06,
+      "loss": 0.0361,
+      "step": 767
+    },
+    {
+      "epoch": 7.456310679611651,
+      "grad_norm": 1.1216914653778076,
+      "learning_rate": 2.522875355443124e-06,
+      "loss": 0.1596,
+      "step": 768
+    },
+    {
+      "epoch": 7.466019417475728,
+      "grad_norm": 1.3237402439117432,
+      "learning_rate": 2.5177920412086586e-06,
+      "loss": 0.1079,
+      "step": 769
+    },
+    {
+      "epoch": 7.475728155339806,
+      "grad_norm": 1.0378979444503784,
+      "learning_rate": 2.512708653409674e-06,
+      "loss": 0.0652,
+      "step": 770
+    },
+    {
+      "epoch": 7.485436893203883,
+      "grad_norm": 1.159886121749878,
+      "learning_rate": 2.507625213064386e-06,
+      "loss": 0.0465,
+      "step": 771
+    },
+    {
+      "epoch": 7.495145631067961,
+      "grad_norm": 0.673026978969574,
+      "learning_rate": 2.5025417411912307e-06,
+      "loss": 0.034,
+      "step": 772
+    },
+    {
+      "epoch": 7.504854368932039,
+      "grad_norm": 1.011098861694336,
+      "learning_rate": 2.4974582588087697e-06,
+      "loss": 0.0951,
+      "step": 773
+    },
+    {
+      "epoch": 7.514563106796117,
+      "grad_norm": 1.035995602607727,
+      "learning_rate": 2.492374786935614e-06,
+      "loss": 0.0827,
+      "step": 774
+    },
+    {
+      "epoch": 7.524271844660194,
+      "grad_norm": 1.1238621473312378,
+      "learning_rate": 2.487291346590326e-06,
+      "loss": 0.1215,
+      "step": 775
+    },
+    {
+      "epoch": 7.533980582524272,
+      "grad_norm": 0.7399255037307739,
+      "learning_rate": 2.4822079587913414e-06,
+      "loss": 0.0792,
+      "step": 776
+    },
+    {
+      "epoch": 7.543689320388349,
+      "grad_norm": 0.7715346217155457,
+      "learning_rate": 2.4771246445568763e-06,
+      "loss": 0.0311,
+      "step": 777
+    },
+    {
+      "epoch": 7.553398058252427,
+      "grad_norm": 0.9662187099456787,
+      "learning_rate": 2.472041424904844e-06,
+      "loss": 0.0511,
+      "step": 778
+    },
+    {
+      "epoch": 7.563106796116505,
+      "grad_norm": 1.0473203659057617,
+      "learning_rate": 2.466958320852766e-06,
+      "loss": 0.0905,
+      "step": 779
+    },
+    {
+      "epoch": 7.572815533980583,
+      "grad_norm": 1.532143473625183,
+      "learning_rate": 2.4618753534176854e-06,
+      "loss": 0.0664,
+      "step": 780
+    },
+    {
+      "epoch": 7.58252427184466,
+      "grad_norm": 1.3228340148925781,
+      "learning_rate": 2.4567925436160823e-06,
+      "loss": 0.0745,
+      "step": 781
+    },
+    {
+      "epoch": 7.592233009708738,
+      "grad_norm": 1.191823124885559,
+      "learning_rate": 2.4517099124637824e-06,
+      "loss": 0.1207,
+      "step": 782
+    },
+    {
+      "epoch": 7.601941747572815,
+      "grad_norm": 1.2542591094970703,
+      "learning_rate": 2.4466274809758757e-06,
+      "loss": 0.0542,
+      "step": 783
+    },
+    {
+      "epoch": 7.611650485436893,
+      "grad_norm": 0.9405433535575867,
+      "learning_rate": 2.4415452701666236e-06,
+      "loss": 0.0919,
+      "step": 784
+    },
+    {
+      "epoch": 7.621359223300971,
+      "grad_norm": 0.9204425811767578,
+      "learning_rate": 2.436463301049378e-06,
+      "loss": 0.0929,
+      "step": 785
+    },
+    {
+      "epoch": 7.631067961165049,
+      "grad_norm": 1.5501075983047485,
+      "learning_rate": 2.431381594636488e-06,
+      "loss": 0.1076,
+      "step": 786
+    },
+    {
+      "epoch": 7.640776699029126,
+      "grad_norm": 1.5708661079406738,
+      "learning_rate": 2.42630017193922e-06,
+      "loss": 0.2274,
+      "step": 787
+    },
+    {
+      "epoch": 7.650485436893204,
+      "grad_norm": 0.5069414377212524,
+      "learning_rate": 2.4212190539676667e-06,
+      "loss": 0.0152,
+      "step": 788
+    },
+    {
+      "epoch": 7.660194174757281,
+      "grad_norm": 1.0907994508743286,
+      "learning_rate": 2.4161382617306585e-06,
+      "loss": 0.0799,
+      "step": 789
+    },
+    {
+      "epoch": 7.669902912621359,
+      "grad_norm": 1.0139005184173584,
+      "learning_rate": 2.4110578162356814e-06,
+      "loss": 0.0385,
+      "step": 790
+    },
+    {
+      "epoch": 7.679611650485437,
+      "grad_norm": 0.8794927597045898,
+      "learning_rate": 2.405977738488786e-06,
+      "loss": 0.0673,
+      "step": 791
+    },
+    {
+      "epoch": 7.689320388349515,
+      "grad_norm": 0.8962538242340088,
+      "learning_rate": 2.4008980494945044e-06,
+      "loss": 0.0576,
+      "step": 792
+    },
+    {
+      "epoch": 7.699029126213592,
+      "grad_norm": 0.9313005805015564,
+      "learning_rate": 2.3958187702557587e-06,
+      "loss": 0.0263,
+      "step": 793
+    },
+    {
+      "epoch": 7.70873786407767,
+      "grad_norm": 0.6752762198448181,
+      "learning_rate": 2.39073992177378e-06,
+      "loss": 0.0334,
+      "step": 794
+    },
+    {
+      "epoch": 7.718446601941747,
+      "grad_norm": 1.2299002408981323,
+      "learning_rate": 2.385661525048014e-06,
+      "loss": 0.0686,
+      "step": 795
+    },
+    {
+      "epoch": 7.728155339805825,
+      "grad_norm": 1.4804145097732544,
+      "learning_rate": 2.3805836010760435e-06,
+      "loss": 0.1083,
+      "step": 796
+    },
+    {
+      "epoch": 7.737864077669903,
+      "grad_norm": 0.7202876210212708,
+      "learning_rate": 2.375506170853494e-06,
+      "loss": 0.0239,
+      "step": 797
+    },
+    {
+      "epoch": 7.747572815533981,
+      "grad_norm": 0.6486926674842834,
+      "learning_rate": 2.3704292553739487e-06,
+      "loss": 0.0279,
+      "step": 798
+    },
+    {
+      "epoch": 7.757281553398058,
+      "grad_norm": 0.7923670411109924,
+      "learning_rate": 2.3653528756288636e-06,
+      "loss": 0.075,
+      "step": 799
+    },
+    {
+      "epoch": 7.766990291262136,
+      "grad_norm": 1.2188405990600586,
+      "learning_rate": 2.3602770526074804e-06,
+      "loss": 0.1341,
+      "step": 800
+    },
+    {
+      "epoch": 7.776699029126213,
+      "grad_norm": 1.213319182395935,
+      "learning_rate": 2.3552018072967375e-06,
+      "loss": 0.0744,
+      "step": 801
+    },
+    {
+      "epoch": 7.786407766990291,
+      "grad_norm": 0.9601984620094299,
+      "learning_rate": 2.3501271606811848e-06,
+      "loss": 0.0378,
+      "step": 802
+    },
+    {
+      "epoch": 7.796116504854369,
+      "grad_norm": 1.3879088163375854,
+      "learning_rate": 2.345053133742898e-06,
+      "loss": 0.1409,
+      "step": 803
+    },
+    {
+      "epoch": 7.805825242718447,
+      "grad_norm": 0.746889054775238,
+      "learning_rate": 2.3399797474613894e-06,
+      "loss": 0.0254,
+      "step": 804
+    },
+    {
+      "epoch": 7.815533980582524,
+      "grad_norm": 1.1101627349853516,
+      "learning_rate": 2.334907022813523e-06,
+      "loss": 0.0435,
+      "step": 805
+    },
+    {
+      "epoch": 7.825242718446602,
+      "grad_norm": 0.9770099520683289,
+      "learning_rate": 2.329834980773427e-06,
+      "loss": 0.1422,
+      "step": 806
+    },
+    {
+      "epoch": 7.834951456310679,
+      "grad_norm": 1.0562567710876465,
+      "learning_rate": 2.324763642312407e-06,
+      "loss": 0.065,
+      "step": 807
+    },
+    {
+      "epoch": 7.844660194174757,
+      "grad_norm": 1.1696139574050903,
+      "learning_rate": 2.3196930283988603e-06,
+      "loss": 0.0962,
+      "step": 808
+    },
+    {
+      "epoch": 7.854368932038835,
+      "grad_norm": 0.7921373248100281,
+      "learning_rate": 2.3146231599981865e-06,
+      "loss": 0.1024,
+      "step": 809
+    },
+    {
+      "epoch": 7.864077669902913,
+      "grad_norm": 1.2344034910202026,
+      "learning_rate": 2.3095540580727054e-06,
+      "loss": 0.1732,
+      "step": 810
+    },
+    {
+      "epoch": 7.87378640776699,
+      "grad_norm": 1.0766966342926025,
+      "learning_rate": 2.304485743581566e-06,
+      "loss": 0.1065,
+      "step": 811
+    },
+    {
+      "epoch": 7.883495145631068,
+      "grad_norm": 1.139400839805603,
+      "learning_rate": 2.299418237480661e-06,
+      "loss": 0.0175,
+      "step": 812
+    },
+    {
+      "epoch": 7.893203883495145,
+      "grad_norm": 0.9371456503868103,
+      "learning_rate": 2.294351560722542e-06,
+      "loss": 0.0914,
+      "step": 813
+    },
+    {
+      "epoch": 7.902912621359223,
+      "grad_norm": 0.5872753262519836,
+      "learning_rate": 2.2892857342563316e-06,
+      "loss": 0.0351,
+      "step": 814
+    },
+    {
+      "epoch": 7.9126213592233015,
+      "grad_norm": 1.1175590753555298,
+      "learning_rate": 2.2842207790276355e-06,
+      "loss": 0.0764,
+      "step": 815
+    },
+    {
+      "epoch": 7.922330097087379,
+      "grad_norm": 1.2251633405685425,
+      "learning_rate": 2.279156715978457e-06,
+      "loss": 0.1079,
+      "step": 816
+    },
+    {
+      "epoch": 7.932038834951456,
+      "grad_norm": 1.0800000429153442,
+      "learning_rate": 2.274093566047113e-06,
+      "loss": 0.0533,
+      "step": 817
+    },
+    {
+      "epoch": 7.941747572815534,
+      "grad_norm": 0.7505971789360046,
+      "learning_rate": 2.2690313501681426e-06,
+      "loss": 0.0411,
+      "step": 818
+    },
+    {
+      "epoch": 7.951456310679612,
+      "grad_norm": 0.9410497546195984,
+      "learning_rate": 2.263970089272223e-06,
+      "loss": 0.0541,
+      "step": 819
+    },
+    {
+      "epoch": 7.961165048543689,
+      "grad_norm": 0.8549265265464783,
+      "learning_rate": 2.2589098042860838e-06,
+      "loss": 0.1001,
+      "step": 820
+    },
+    {
+      "epoch": 7.970873786407767,
+      "grad_norm": 0.736650824546814,
+      "learning_rate": 2.2538505161324186e-06,
+      "loss": 0.0271,
+      "step": 821
+    },
+    {
+      "epoch": 7.980582524271845,
+      "grad_norm": 0.8051927089691162,
+      "learning_rate": 2.2487922457298007e-06,
+      "loss": 0.0299,
+      "step": 822
+    },
+    {
+      "epoch": 7.990291262135923,
+      "grad_norm": 1.251429557800293,
+      "learning_rate": 2.243735013992593e-06,
+      "loss": 0.1019,
+      "step": 823
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.8267395496368408,
+      "learning_rate": 2.238678841830867e-06,
+      "loss": 0.0444,
+      "step": 824
+    },
+    {
+      "epoch": 8.009708737864077,
+      "grad_norm": 1.0311859846115112,
+      "learning_rate": 2.2336237501503103e-06,
+      "loss": 0.0393,
+      "step": 825
+    },
+    {
+      "epoch": 8.019417475728156,
+      "grad_norm": 0.6791521906852722,
+      "learning_rate": 2.2285697598521446e-06,
+      "loss": 0.0173,
+      "step": 826
+    },
+    {
+      "epoch": 8.029126213592233,
+      "grad_norm": 0.6706322431564331,
+      "learning_rate": 2.2235168918330374e-06,
+      "loss": 0.036,
+      "step": 827
+    },
+    {
+      "epoch": 8.03883495145631,
+      "grad_norm": 0.9080128073692322,
+      "learning_rate": 2.2184651669850164e-06,
+      "loss": 0.0482,
+      "step": 828
+    },
+    {
+      "epoch": 8.048543689320388,
+      "grad_norm": 0.49124234914779663,
+      "learning_rate": 2.2134146061953814e-06,
+      "loss": 0.0033,
+      "step": 829
+    },
+    {
+      "epoch": 8.058252427184467,
+      "grad_norm": 1.4356873035430908,
+      "learning_rate": 2.2083652303466196e-06,
+      "loss": 0.1166,
+      "step": 830
+    },
+    {
+      "epoch": 8.067961165048544,
+      "grad_norm": 1.9703857898712158,
+      "learning_rate": 2.20331706031632e-06,
+      "loss": 0.1,
+      "step": 831
+    },
+    {
+      "epoch": 8.077669902912621,
+      "grad_norm": 0.6531624794006348,
+      "learning_rate": 2.1982701169770853e-06,
+      "loss": 0.0243,
+      "step": 832
+    },
+    {
+      "epoch": 8.087378640776699,
+      "grad_norm": 0.9504621624946594,
+      "learning_rate": 2.1932244211964456e-06,
+      "loss": 0.0462,
+      "step": 833
+    },
+    {
+      "epoch": 8.097087378640778,
+      "grad_norm": 0.9275246858596802,
+      "learning_rate": 2.1881799938367735e-06,
+      "loss": 0.0403,
+      "step": 834
+    },
+    {
+      "epoch": 8.106796116504855,
+      "grad_norm": 0.7024449706077576,
+      "learning_rate": 2.1831368557551962e-06,
+      "loss": 0.0266,
+      "step": 835
+    },
+    {
+      "epoch": 8.116504854368932,
+      "grad_norm": 1.1542211771011353,
+      "learning_rate": 2.1780950278035114e-06,
+      "loss": 0.062,
+      "step": 836
+    },
+    {
+      "epoch": 8.12621359223301,
+      "grad_norm": 0.7213531136512756,
+      "learning_rate": 2.173054530828098e-06,
+      "loss": 0.0134,
+      "step": 837
+    },
+    {
+      "epoch": 8.135922330097088,
+      "grad_norm": 0.6973257660865784,
+      "learning_rate": 2.168015385669833e-06,
+      "loss": 0.0167,
+      "step": 838
+    },
+    {
+      "epoch": 8.145631067961165,
+      "grad_norm": 1.4313327074050903,
+      "learning_rate": 2.162977613164005e-06,
+      "loss": 0.1217,
+      "step": 839
+    },
+    {
+      "epoch": 8.155339805825243,
+      "grad_norm": 1.4857988357543945,
+      "learning_rate": 2.157941234140225e-06,
+      "loss": 0.1104,
+      "step": 840
+    },
+    {
+      "epoch": 8.16504854368932,
+      "grad_norm": 1.080141544342041,
+      "learning_rate": 2.1529062694223437e-06,
+      "loss": 0.0566,
+      "step": 841
+    },
+    {
+      "epoch": 8.174757281553399,
+      "grad_norm": 0.6205718517303467,
+      "learning_rate": 2.147872739828364e-06,
+      "loss": 0.012,
+      "step": 842
+    },
+    {
+      "epoch": 8.184466019417476,
+      "grad_norm": 0.6710229516029358,
+      "learning_rate": 2.142840666170354e-06,
+      "loss": 0.0163,
+      "step": 843
+    },
+    {
+      "epoch": 8.194174757281553,
+      "grad_norm": 0.6269445419311523,
+      "learning_rate": 2.1378100692543637e-06,
+      "loss": 0.017,
+      "step": 844
+    },
+    {
+      "epoch": 8.20388349514563,
+      "grad_norm": 0.7391650676727295,
+      "learning_rate": 2.1327809698803354e-06,
+      "loss": 0.0296,
+      "step": 845
+    },
+    {
+      "epoch": 8.21359223300971,
+      "grad_norm": 0.861961305141449,
+      "learning_rate": 2.1277533888420203e-06,
+      "loss": 0.0403,
+      "step": 846
+    },
+    {
+      "epoch": 8.223300970873787,
+      "grad_norm": 0.6323930025100708,
+      "learning_rate": 2.1227273469268932e-06,
+      "loss": 0.0148,
+      "step": 847
+    },
+    {
+      "epoch": 8.233009708737864,
+      "grad_norm": 0.9467630386352539,
+      "learning_rate": 2.117702864916063e-06,
+      "loss": 0.0272,
+      "step": 848
+    },
+    {
+      "epoch": 8.242718446601941,
+      "grad_norm": 1.1341761350631714,
+      "learning_rate": 2.1126799635841897e-06,
+      "loss": 0.0373,
+      "step": 849
+    },
+    {
+      "epoch": 8.25242718446602,
+      "grad_norm": 0.8145177960395813,
+      "learning_rate": 2.1076586636993975e-06,
+      "loss": 0.039,
+      "step": 850
+    },
+    {
+      "epoch": 8.262135922330097,
+      "grad_norm": 0.7213461995124817,
+      "learning_rate": 2.102638986023189e-06,
+      "loss": 0.023,
+      "step": 851
+    },
+    {
+      "epoch": 8.271844660194175,
+      "grad_norm": 1.1217619180679321,
+      "learning_rate": 2.0976209513103604e-06,
+      "loss": 0.067,
+      "step": 852
+    },
+    {
+      "epoch": 8.281553398058252,
+      "grad_norm": 0.49746865034103394,
+      "learning_rate": 2.0926045803089135e-06,
+      "loss": 0.0148,
+      "step": 853
+    },
+    {
+      "epoch": 8.29126213592233,
+      "grad_norm": 0.917957067489624,
+      "learning_rate": 2.087589893759972e-06,
+      "loss": 0.0287,
+      "step": 854
+    },
+    {
+      "epoch": 8.300970873786408,
+      "grad_norm": 0.7017801403999329,
+      "learning_rate": 2.0825769123976954e-06,
+      "loss": 0.0574,
+      "step": 855
+    },
+    {
+      "epoch": 8.310679611650485,
+      "grad_norm": 0.9924259781837463,
+      "learning_rate": 2.077565656949191e-06,
+      "loss": 0.0996,
+      "step": 856
+    },
+    {
+      "epoch": 8.320388349514563,
+      "grad_norm": 0.8891076445579529,
+      "learning_rate": 2.072556148134431e-06,
+      "loss": 0.0578,
+      "step": 857
+    },
+    {
+      "epoch": 8.330097087378642,
+      "grad_norm": 0.8946473598480225,
+      "learning_rate": 2.0675484066661666e-06,
+      "loss": 0.0427,
+      "step": 858
+    },
+    {
+      "epoch": 8.339805825242719,
+      "grad_norm": 0.6834955215454102,
+      "learning_rate": 2.0625424532498407e-06,
+      "loss": 0.0176,
+      "step": 859
+    },
+    {
+      "epoch": 8.349514563106796,
+      "grad_norm": 0.9527811408042908,
+      "learning_rate": 2.057538308583502e-06,
+      "loss": 0.0226,
+      "step": 860
+    },
+    {
+      "epoch": 8.359223300970873,
+      "grad_norm": 1.418888807296753,
+      "learning_rate": 2.0525359933577243e-06,
+      "loss": 0.0945,
+      "step": 861
+    },
+    {
+      "epoch": 8.368932038834952,
+      "grad_norm": 0.8171473741531372,
+      "learning_rate": 2.047535528255514e-06,
+      "loss": 0.0348,
+      "step": 862
+    },
+    {
+      "epoch": 8.37864077669903,
+      "grad_norm": 0.38823753595352173,
+      "learning_rate": 2.0425369339522276e-06,
+      "loss": 0.0034,
+      "step": 863
+    },
+    {
+      "epoch": 8.388349514563107,
+      "grad_norm": 0.68483966588974,
+      "learning_rate": 2.0375402311154886e-06,
+      "loss": 0.0425,
+      "step": 864
+    },
+    {
+      "epoch": 8.398058252427184,
+      "grad_norm": 0.6909242272377014,
+      "learning_rate": 2.0325454404050983e-06,
+      "loss": 0.0043,
+      "step": 865
+    },
+    {
+      "epoch": 8.407766990291263,
+      "grad_norm": 0.9698901772499084,
+      "learning_rate": 2.0275525824729523e-06,
+      "loss": 0.0846,
+      "step": 866
+    },
+    {
+      "epoch": 8.41747572815534,
+      "grad_norm": 0.7008579969406128,
+      "learning_rate": 2.022561677962955e-06,
+      "loss": 0.0503,
+      "step": 867
+    },
+    {
+      "epoch": 8.427184466019417,
+      "grad_norm": 0.823596179485321,
+      "learning_rate": 2.017572747510934e-06,
+      "loss": 0.0309,
+      "step": 868
+    },
+    {
+      "epoch": 8.436893203883495,
+      "grad_norm": 1.0459940433502197,
+      "learning_rate": 2.012585811744555e-06,
+      "loss": 0.0496,
+      "step": 869
+    },
+    {
+      "epoch": 8.446601941747574,
+      "grad_norm": 0.9439486861228943,
+      "learning_rate": 2.0076008912832355e-06,
+      "loss": 0.0703,
+      "step": 870
+    },
+    {
+      "epoch": 8.45631067961165,
+      "grad_norm": 1.2154759168624878,
+      "learning_rate": 2.002618006738061e-06,
+      "loss": 0.0716,
+      "step": 871
+    },
+    {
+      "epoch": 8.466019417475728,
+      "grad_norm": 0.9697406888008118,
+      "learning_rate": 1.9976371787116992e-06,
+      "loss": 0.037,
+      "step": 872
+    },
+    {
+      "epoch": 8.475728155339805,
+      "grad_norm": 0.6543058753013611,
+      "learning_rate": 1.9926584277983134e-06,
+      "loss": 0.0158,
+      "step": 873
+    },
+    {
+      "epoch": 8.485436893203884,
+      "grad_norm": 1.0881898403167725,
+      "learning_rate": 1.9876817745834805e-06,
+      "loss": 0.0691,
+      "step": 874
+    },
+    {
+      "epoch": 8.495145631067961,
+      "grad_norm": 1.221283197402954,
+      "learning_rate": 1.9827072396441044e-06,
+      "loss": 0.1064,
+      "step": 875
+    },
+    {
+      "epoch": 8.504854368932039,
+      "grad_norm": 1.0721032619476318,
+      "learning_rate": 1.9777348435483285e-06,
+      "loss": 0.0776,
+      "step": 876
+    },
+    {
+      "epoch": 8.514563106796116,
+      "grad_norm": 1.2138664722442627,
+      "learning_rate": 1.972764606855454e-06,
+      "loss": 0.0738,
+      "step": 877
+    },
+    {
+      "epoch": 8.524271844660195,
+      "grad_norm": 0.8295838236808777,
+      "learning_rate": 1.9677965501158534e-06,
+      "loss": 0.0355,
+      "step": 878
+    },
+    {
+      "epoch": 8.533980582524272,
+      "grad_norm": 0.9322443008422852,
+      "learning_rate": 1.9628306938708857e-06,
+      "loss": 0.0639,
+      "step": 879
+    },
+    {
+      "epoch": 8.54368932038835,
+      "grad_norm": 0.7268901467323303,
+      "learning_rate": 1.957867058652812e-06,
+      "loss": 0.0355,
+      "step": 880
+    },
+    {
+      "epoch": 8.553398058252426,
+      "grad_norm": 0.9112371802330017,
+      "learning_rate": 1.952905664984709e-06,
+      "loss": 0.1148,
+      "step": 881
+    },
+    {
+      "epoch": 8.563106796116505,
+      "grad_norm": 0.8290453553199768,
+      "learning_rate": 1.947946533380387e-06,
+      "loss": 0.0247,
+      "step": 882
+    },
+    {
+      "epoch": 8.572815533980583,
+      "grad_norm": 1.105477213859558,
+      "learning_rate": 1.9429896843443025e-06,
+      "loss": 0.0115,
+      "step": 883
+    },
+    {
+      "epoch": 8.58252427184466,
+      "grad_norm": 0.9778973460197449,
+      "learning_rate": 1.938035138371474e-06,
+      "loss": 0.1297,
+      "step": 884
+    },
+    {
+      "epoch": 8.592233009708737,
+      "grad_norm": 0.5014190673828125,
+      "learning_rate": 1.933082915947398e-06,
+      "loss": 0.0712,
+      "step": 885
+    },
+    {
+      "epoch": 8.601941747572816,
+      "grad_norm": 0.7944493293762207,
+      "learning_rate": 1.928133037547963e-06,
+      "loss": 0.0222,
+      "step": 886
+    },
+    {
+      "epoch": 8.611650485436893,
+      "grad_norm": 0.9558936953544617,
+      "learning_rate": 1.9231855236393677e-06,
+      "loss": 0.0311,
+      "step": 887
+    },
+    {
+      "epoch": 8.62135922330097,
+      "grad_norm": 0.9489585757255554,
+      "learning_rate": 1.9182403946780316e-06,
+      "loss": 0.0544,
+      "step": 888
+    },
+    {
+      "epoch": 8.631067961165048,
+      "grad_norm": 1.0101840496063232,
+      "learning_rate": 1.9132976711105146e-06,
+      "loss": 0.0858,
+      "step": 889
+    },
+    {
+      "epoch": 8.640776699029127,
+      "grad_norm": 0.6758891940116882,
+      "learning_rate": 1.9083573733734328e-06,
+      "loss": 0.0452,
+      "step": 890
+    },
+    {
+      "epoch": 8.650485436893204,
+      "grad_norm": 0.9936819076538086,
+      "learning_rate": 1.903419521893369e-06,
+      "loss": 0.0409,
+      "step": 891
+    },
+    {
+      "epoch": 8.660194174757281,
+      "grad_norm": 0.8402771949768066,
+      "learning_rate": 1.898484137086794e-06,
+      "loss": 0.0413,
+      "step": 892
+    },
+    {
+      "epoch": 8.669902912621358,
+      "grad_norm": 0.6089380979537964,
+      "learning_rate": 1.8935512393599784e-06,
+      "loss": 0.0082,
+      "step": 893
+    },
+    {
+      "epoch": 8.679611650485437,
+      "grad_norm": 0.9442456960678101,
+      "learning_rate": 1.8886208491089095e-06,
+      "loss": 0.0168,
+      "step": 894
+    },
+    {
+      "epoch": 8.689320388349515,
+      "grad_norm": 0.9080673456192017,
+      "learning_rate": 1.8836929867192077e-06,
+      "loss": 0.0168,
+      "step": 895
+    },
+    {
+      "epoch": 8.699029126213592,
+      "grad_norm": 0.7783305048942566,
+      "learning_rate": 1.8787676725660405e-06,
+      "loss": 0.0228,
+      "step": 896
+    },
+    {
+      "epoch": 8.70873786407767,
+      "grad_norm": 1.021910548210144,
+      "learning_rate": 1.8738449270140404e-06,
+      "loss": 0.1082,
+      "step": 897
+    },
+    {
+      "epoch": 8.718446601941748,
+      "grad_norm": 0.9895579218864441,
+      "learning_rate": 1.8689247704172187e-06,
+      "loss": 0.0542,
+      "step": 898
+    },
+    {
+      "epoch": 8.728155339805825,
+      "grad_norm": 0.8246378898620605,
+      "learning_rate": 1.8640072231188825e-06,
+      "loss": 0.0667,
+      "step": 899
+    },
+    {
+      "epoch": 8.737864077669903,
+      "grad_norm": 0.9075983762741089,
+      "learning_rate": 1.8590923054515504e-06,
+      "loss": 0.0458,
+      "step": 900
+    },
+    {
+      "epoch": 8.74757281553398,
+      "grad_norm": 0.8279749751091003,
+      "learning_rate": 1.8541800377368673e-06,
+      "loss": 0.052,
+      "step": 901
+    },
+    {
+      "epoch": 8.757281553398059,
+      "grad_norm": 0.9606509804725647,
+      "learning_rate": 1.8492704402855229e-06,
+      "loss": 0.0584,
+      "step": 902
+    },
+    {
+      "epoch": 8.766990291262136,
+      "grad_norm": 0.42115524411201477,
+      "learning_rate": 1.8443635333971643e-06,
+      "loss": 0.0156,
+      "step": 903
+    },
+    {
+      "epoch": 8.776699029126213,
+      "grad_norm": 1.1302638053894043,
+      "learning_rate": 1.8394593373603173e-06,
+      "loss": 0.0397,
+      "step": 904
+    },
+    {
+      "epoch": 8.78640776699029,
+      "grad_norm": 1.0244662761688232,
+      "learning_rate": 1.8345578724522957e-06,
+      "loss": 0.0385,
+      "step": 905
+    },
+    {
+      "epoch": 8.79611650485437,
+      "grad_norm": 1.0088191032409668,
+      "learning_rate": 1.8296591589391227e-06,
+      "loss": 0.0297,
+      "step": 906
+    },
+    {
+      "epoch": 8.805825242718447,
+      "grad_norm": 0.9084392189979553,
+      "learning_rate": 1.8247632170754443e-06,
+      "loss": 0.0368,
+      "step": 907
+    },
+    {
+      "epoch": 8.815533980582524,
+      "grad_norm": 0.488815575838089,
+      "learning_rate": 1.8198700671044477e-06,
+      "loss": 0.0094,
+      "step": 908
+    },
+    {
+      "epoch": 8.825242718446601,
+      "grad_norm": 1.1691783666610718,
+      "learning_rate": 1.8149797292577757e-06,
+      "loss": 0.0653,
+      "step": 909
+    },
+    {
+      "epoch": 8.83495145631068,
+      "grad_norm": 1.1218113899230957,
+      "learning_rate": 1.8100922237554442e-06,
+      "loss": 0.087,
+      "step": 910
+    },
+    {
+      "epoch": 8.844660194174757,
+      "grad_norm": 0.9231531023979187,
+      "learning_rate": 1.8052075708057581e-06,
+      "loss": 0.0097,
+      "step": 911
+    },
+    {
+      "epoch": 8.854368932038835,
+      "grad_norm": 0.40473511815071106,
+      "learning_rate": 1.8003257906052284e-06,
+      "loss": 0.0056,
+      "step": 912
+    },
+    {
+      "epoch": 8.864077669902912,
+      "grad_norm": 1.0644712448120117,
+      "learning_rate": 1.7954469033384868e-06,
+      "loss": 0.057,
+      "step": 913
+    },
+    {
+      "epoch": 8.87378640776699,
+      "grad_norm": 0.9894399642944336,
+      "learning_rate": 1.790570929178206e-06,
+      "loss": 0.0567,
+      "step": 914
+    },
+    {
+      "epoch": 8.883495145631068,
+      "grad_norm": 0.7185762524604797,
+      "learning_rate": 1.7856978882850112e-06,
+      "loss": 0.0388,
+      "step": 915
+    },
+    {
+      "epoch": 8.893203883495145,
+      "grad_norm": 0.9258529543876648,
+      "learning_rate": 1.780827800807401e-06,
+      "loss": 0.0512,
+      "step": 916
+    },
+    {
+      "epoch": 8.902912621359224,
+      "grad_norm": 1.1820943355560303,
+      "learning_rate": 1.7759606868816623e-06,
+      "loss": 0.0576,
+      "step": 917
+    },
+    {
+      "epoch": 8.912621359223301,
+      "grad_norm": 0.7165997624397278,
+      "learning_rate": 1.771096566631788e-06,
+      "loss": 0.0326,
+      "step": 918
+    },
+    {
+      "epoch": 8.922330097087379,
+      "grad_norm": 1.1988775730133057,
+      "learning_rate": 1.766235460169392e-06,
+      "loss": 0.0602,
+      "step": 919
+    },
+    {
+      "epoch": 8.932038834951456,
+      "grad_norm": 0.915666937828064,
+      "learning_rate": 1.7613773875936274e-06,
+      "loss": 0.0139,
+      "step": 920
+    },
+    {
+      "epoch": 8.941747572815533,
+      "grad_norm": 0.922701895236969,
+      "learning_rate": 1.7565223689911038e-06,
+      "loss": 0.0131,
+      "step": 921
+    },
+    {
+      "epoch": 8.951456310679612,
+      "grad_norm": 0.5915096402168274,
+      "learning_rate": 1.7516704244358018e-06,
+      "loss": 0.0354,
+      "step": 922
+    },
+    {
+      "epoch": 8.96116504854369,
+      "grad_norm": 0.5728499293327332,
+      "learning_rate": 1.7468215739889941e-06,
+      "loss": 0.0277,
+      "step": 923
+    },
+    {
+      "epoch": 8.970873786407767,
+      "grad_norm": 1.0376454591751099,
+      "learning_rate": 1.741975837699158e-06,
+      "loss": 0.0891,
+      "step": 924
+    },
+    {
+      "epoch": 8.980582524271846,
+      "grad_norm": 0.5719737410545349,
+      "learning_rate": 1.7371332356018972e-06,
+      "loss": 0.0702,
+      "step": 925
+    },
+    {
+      "epoch": 8.990291262135923,
+      "grad_norm": 0.9259313344955444,
+      "learning_rate": 1.7322937877198545e-06,
+      "loss": 0.0859,
+      "step": 926
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.8504927754402161,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.0235,
+      "step": 927
+    },
+    {
+      "epoch": 9.009708737864077,
+      "grad_norm": 0.8123329281806946,
+      "learning_rate": 1.7226244346267063e-06,
+      "loss": 0.0469,
+      "step": 928
+    },
+    {
+      "epoch": 9.019417475728156,
+      "grad_norm": 0.7285428047180176,
+      "learning_rate": 1.7177945693953486e-06,
+      "loss": 0.0227,
+      "step": 929
+    },
+    {
+      "epoch": 9.029126213592233,
+      "grad_norm": 0.857140064239502,
+      "learning_rate": 1.7129679383385384e-06,
+      "loss": 0.0319,
+      "step": 930
+    },
+    {
+      "epoch": 9.03883495145631,
+      "grad_norm": 0.5479164123535156,
+      "learning_rate": 1.7081445614128845e-06,
+      "loss": 0.0118,
+      "step": 931
+    },
+    {
+      "epoch": 9.048543689320388,
+      "grad_norm": 0.8059089183807373,
+      "learning_rate": 1.7033244585615393e-06,
+      "loss": 0.0663,
+      "step": 932
+    },
+    {
+      "epoch": 9.058252427184467,
+      "grad_norm": 1.0029854774475098,
+      "learning_rate": 1.698507649714121e-06,
+      "loss": 0.0417,
+      "step": 933
+    },
+    {
+      "epoch": 9.067961165048544,
+      "grad_norm": 0.69701087474823,
+      "learning_rate": 1.6936941547866248e-06,
+      "loss": 0.0103,
+      "step": 934
+    },
+    {
+      "epoch": 9.077669902912621,
+      "grad_norm": 0.5342279076576233,
+      "learning_rate": 1.688883993681345e-06,
+      "loss": 0.0278,
+      "step": 935
+    },
+    {
+      "epoch": 9.087378640776699,
+      "grad_norm": 0.7861754298210144,
+      "learning_rate": 1.6840771862867922e-06,
+      "loss": 0.0186,
+      "step": 936
+    },
+    {
+      "epoch": 9.097087378640778,
+      "grad_norm": 1.3431063890457153,
+      "learning_rate": 1.6792737524776093e-06,
+      "loss": 0.0341,
+      "step": 937
+    },
+    {
+      "epoch": 9.106796116504855,
+      "grad_norm": 0.44472604990005493,
+      "learning_rate": 1.674473712114492e-06,
+      "loss": 0.0082,
+      "step": 938
+    },
+    {
+      "epoch": 9.116504854368932,
+      "grad_norm": 0.5097460746765137,
+      "learning_rate": 1.6696770850441036e-06,
+      "loss": 0.0043,
+      "step": 939
+    },
+    {
+      "epoch": 9.12621359223301,
+      "grad_norm": 1.295851707458496,
+      "learning_rate": 1.6648838910989955e-06,
+      "loss": 0.0904,
+      "step": 940
+    },
+    {
+      "epoch": 9.135922330097088,
+      "grad_norm": 1.0630472898483276,
+      "learning_rate": 1.6600941500975237e-06,
+      "loss": 0.0222,
+      "step": 941
+    },
+    {
+      "epoch": 9.145631067961165,
+      "grad_norm": 1.1637173891067505,
+      "learning_rate": 1.6553078818437678e-06,
+      "loss": 0.0509,
+      "step": 942
+    },
+    {
+      "epoch": 9.155339805825243,
+      "grad_norm": 0.9822437763214111,
+      "learning_rate": 1.6505251061274492e-06,
+      "loss": 0.0379,
+      "step": 943
+    },
+    {
+      "epoch": 9.16504854368932,
+      "grad_norm": 0.8363843560218811,
+      "learning_rate": 1.6457458427238464e-06,
+      "loss": 0.107,
+      "step": 944
+    },
+    {
+      "epoch": 9.174757281553399,
+      "grad_norm": 1.046858549118042,
+      "learning_rate": 1.6409701113937182e-06,
+      "loss": 0.0228,
+      "step": 945
+    },
+    {
+      "epoch": 9.184466019417476,
+      "grad_norm": 0.7185032367706299,
+      "learning_rate": 1.6361979318832173e-06,
+      "loss": 0.0168,
+      "step": 946
+    },
+    {
+      "epoch": 9.194174757281553,
+      "grad_norm": 0.6683576703071594,
+      "learning_rate": 1.6314293239238134e-06,
+      "loss": 0.0401,
+      "step": 947
+    },
+    {
+      "epoch": 9.20388349514563,
+      "grad_norm": 0.619415283203125,
+      "learning_rate": 1.626664307232207e-06,
+      "loss": 0.0252,
+      "step": 948
+    },
+    {
+      "epoch": 9.21359223300971,
+      "grad_norm": 0.7388278841972351,
+      "learning_rate": 1.62190290151025e-06,
+      "loss": 0.0444,
+      "step": 949
+    },
+    {
+      "epoch": 9.223300970873787,
+      "grad_norm": 0.7596797943115234,
+      "learning_rate": 1.617145126444864e-06,
+      "loss": 0.0315,
+      "step": 950
+    },
+    {
+      "epoch": 9.233009708737864,
+      "grad_norm": 0.7415765523910522,
+      "learning_rate": 1.6123910017079591e-06,
+      "loss": 0.0363,
+      "step": 951
+    },
+    {
+      "epoch": 9.242718446601941,
+      "grad_norm": 0.38191917538642883,
+      "learning_rate": 1.6076405469563533e-06,
+      "loss": 0.0189,
+      "step": 952
+    },
+    {
+      "epoch": 9.25242718446602,
+      "grad_norm": 0.7858003973960876,
+      "learning_rate": 1.6028937818316889e-06,
+      "loss": 0.0297,
+      "step": 953
+    },
+    {
+      "epoch": 9.262135922330097,
+      "grad_norm": 0.45921990275382996,
+      "learning_rate": 1.598150725960354e-06,
+      "loss": 0.0047,
+      "step": 954
+    },
+    {
+      "epoch": 9.271844660194175,
+      "grad_norm": 0.7559367418289185,
+      "learning_rate": 1.5934113989533992e-06,
+      "loss": 0.0199,
+      "step": 955
+    },
+    {
+      "epoch": 9.281553398058252,
+      "grad_norm": 0.7790589928627014,
+      "learning_rate": 1.5886758204064582e-06,
+      "loss": 0.0174,
+      "step": 956
+    },
+    {
+      "epoch": 9.29126213592233,
+      "grad_norm": 0.3947029411792755,
+      "learning_rate": 1.583944009899665e-06,
+      "loss": 0.0047,
+      "step": 957
+    },
+    {
+      "epoch": 9.300970873786408,
+      "grad_norm": 0.7276475429534912,
+      "learning_rate": 1.579215986997575e-06,
+      "loss": 0.0155,
+      "step": 958
+    },
+    {
+      "epoch": 9.310679611650485,
+      "grad_norm": 1.0900495052337646,
+      "learning_rate": 1.5744917712490821e-06,
+      "loss": 0.0772,
+      "step": 959
+    },
+    {
+      "epoch": 9.320388349514563,
+      "grad_norm": 0.44895508885383606,
+      "learning_rate": 1.5697713821873401e-06,
+      "loss": 0.0038,
+      "step": 960
+    },
+    {
+      "epoch": 9.330097087378642,
+      "grad_norm": 0.6020745635032654,
+      "learning_rate": 1.5650548393296788e-06,
+      "loss": 0.0069,
+      "step": 961
+    },
+    {
+      "epoch": 9.339805825242719,
+      "grad_norm": 1.2439451217651367,
+      "learning_rate": 1.5603421621775273e-06,
+      "loss": 0.0321,
+      "step": 962
+    },
+    {
+      "epoch": 9.349514563106796,
+      "grad_norm": 0.9518343806266785,
+      "learning_rate": 1.555633370216329e-06,
+      "loss": 0.0215,
+      "step": 963
+    },
+    {
+      "epoch": 9.359223300970873,
+      "grad_norm": 0.6604970693588257,
+      "learning_rate": 1.5509284829154652e-06,
+      "loss": 0.0341,
+      "step": 964
+    },
+    {
+      "epoch": 9.368932038834952,
+      "grad_norm": 0.6232668161392212,
+      "learning_rate": 1.5462275197281717e-06,
+      "loss": 0.0087,
+      "step": 965
+    },
+    {
+      "epoch": 9.37864077669903,
+      "grad_norm": 1.5144613981246948,
+      "learning_rate": 1.5415305000914587e-06,
+      "loss": 0.063,
+      "step": 966
+    },
+    {
+      "epoch": 9.388349514563107,
+      "grad_norm": 0.38772061467170715,
+      "learning_rate": 1.536837443426032e-06,
+      "loss": 0.0041,
+      "step": 967
+    },
+    {
+      "epoch": 9.398058252427184,
+      "grad_norm": 0.34827372431755066,
+      "learning_rate": 1.5321483691362121e-06,
+      "loss": 0.0166,
+      "step": 968
+    },
+    {
+      "epoch": 9.407766990291263,
+      "grad_norm": 0.5997104644775391,
+      "learning_rate": 1.5274632966098527e-06,
+      "loss": 0.0051,
+      "step": 969
+    },
+    {
+      "epoch": 9.41747572815534,
+      "grad_norm": 0.39720046520233154,
+      "learning_rate": 1.5227822452182617e-06,
+      "loss": 0.0072,
+      "step": 970
+    },
+    {
+      "epoch": 9.427184466019417,
+      "grad_norm": 0.4819827079772949,
+      "learning_rate": 1.5181052343161212e-06,
+      "loss": 0.0221,
+      "step": 971
+    },
+    {
+      "epoch": 9.436893203883495,
+      "grad_norm": 0.3216429650783539,
+      "learning_rate": 1.5134322832414066e-06,
+      "loss": 0.0051,
+      "step": 972
+    },
+    {
+      "epoch": 9.446601941747574,
+      "grad_norm": 0.6433607339859009,
+      "learning_rate": 1.508763411315308e-06,
+      "loss": 0.0217,
+      "step": 973
+    },
+    {
+      "epoch": 9.45631067961165,
+      "grad_norm": 0.44555774331092834,
+      "learning_rate": 1.5040986378421485e-06,
+      "loss": 0.0143,
+      "step": 974
+    },
+    {
+      "epoch": 9.466019417475728,
+      "grad_norm": 0.37716561555862427,
+      "learning_rate": 1.499437982109305e-06,
+      "loss": 0.0132,
+      "step": 975
+    },
+    {
+      "epoch": 9.475728155339805,
+      "grad_norm": 0.930448591709137,
+      "learning_rate": 1.4947814633871316e-06,
+      "loss": 0.0136,
+      "step": 976
+    },
+    {
+      "epoch": 9.485436893203884,
+      "grad_norm": 0.3260258436203003,
+      "learning_rate": 1.4901291009288748e-06,
+      "loss": 0.0093,
+      "step": 977
+    },
+    {
+      "epoch": 9.495145631067961,
+      "grad_norm": 0.9126110672950745,
+      "learning_rate": 1.4854809139705961e-06,
+      "loss": 0.0547,
+      "step": 978
+    },
+    {
+      "epoch": 9.504854368932039,
+      "grad_norm": 0.5943132638931274,
+      "learning_rate": 1.4808369217310937e-06,
+      "loss": 0.0202,
+      "step": 979
+    },
+    {
+      "epoch": 9.514563106796116,
+      "grad_norm": 1.0365543365478516,
+      "learning_rate": 1.4761971434118207e-06,
+      "loss": 0.045,
+      "step": 980
+    },
+    {
+      "epoch": 9.524271844660195,
+      "grad_norm": 0.8611214756965637,
+      "learning_rate": 1.4715615981968088e-06,
+      "loss": 0.0206,
+      "step": 981
+    },
+    {
+      "epoch": 9.533980582524272,
+      "grad_norm": 0.49241188168525696,
+      "learning_rate": 1.4669303052525852e-06,
+      "loss": 0.0222,
+      "step": 982
+    },
+    {
+      "epoch": 9.54368932038835,
+      "grad_norm": 0.34316757321357727,
+      "learning_rate": 1.4623032837280971e-06,
+      "loss": 0.0018,
+      "step": 983
+    },
+    {
+      "epoch": 9.553398058252426,
+      "grad_norm": 1.0957579612731934,
+      "learning_rate": 1.4576805527546293e-06,
+      "loss": 0.0436,
+      "step": 984
+    },
+    {
+      "epoch": 9.563106796116505,
+      "grad_norm": 1.0365829467773438,
+      "learning_rate": 1.4530621314457255e-06,
+      "loss": 0.0305,
+      "step": 985
+    },
+    {
+      "epoch": 9.572815533980583,
+      "grad_norm": 1.2111074924468994,
+      "learning_rate": 1.4484480388971141e-06,
+      "loss": 0.0607,
+      "step": 986
+    },
+    {
+      "epoch": 9.58252427184466,
+      "grad_norm": 1.0916733741760254,
+      "learning_rate": 1.4438382941866224e-06,
+      "loss": 0.0365,
+      "step": 987
+    },
+    {
+      "epoch": 9.592233009708737,
+      "grad_norm": 0.630837082862854,
+      "learning_rate": 1.4392329163741015e-06,
+      "loss": 0.0192,
+      "step": 988
+    },
+    {
+      "epoch": 9.601941747572816,
+      "grad_norm": 0.5758682489395142,
+      "learning_rate": 1.4346319245013463e-06,
+      "loss": 0.0894,
+      "step": 989
+    },
+    {
+      "epoch": 9.611650485436893,
+      "grad_norm": 0.761637806892395,
+      "learning_rate": 1.430035337592018e-06,
+      "loss": 0.0084,
+      "step": 990
+    },
+    {
+      "epoch": 9.62135922330097,
+      "grad_norm": 0.431892067193985,
+      "learning_rate": 1.425443174651564e-06,
+      "loss": 0.0118,
+      "step": 991
+    },
+    {
+      "epoch": 9.631067961165048,
+      "grad_norm": 0.5555681586265564,
+      "learning_rate": 1.4208554546671407e-06,
+      "loss": 0.0249,
+      "step": 992
+    },
+    {
+      "epoch": 9.640776699029127,
+      "grad_norm": 0.8118094205856323,
+      "learning_rate": 1.4162721966075323e-06,
+      "loss": 0.0362,
+      "step": 993
+    },
+    {
+      "epoch": 9.650485436893204,
+      "grad_norm": 0.3614587187767029,
+      "learning_rate": 1.411693419423078e-06,
+      "loss": 0.0181,
+      "step": 994
+    },
+    {
+      "epoch": 9.660194174757281,
+      "grad_norm": 0.6472867131233215,
+      "learning_rate": 1.4071191420455873e-06,
+      "loss": 0.0171,
+      "step": 995
+    },
+    {
+      "epoch": 9.669902912621358,
+      "grad_norm": 0.46341660618782043,
+      "learning_rate": 1.4025493833882645e-06,
+      "loss": 0.0078,
+      "step": 996
+    },
+    {
+      "epoch": 9.679611650485437,
+      "grad_norm": 0.46548253297805786,
+      "learning_rate": 1.3979841623456309e-06,
+      "loss": 0.0041,
+      "step": 997
+    },
+    {
+      "epoch": 9.689320388349515,
+      "grad_norm": 1.1735632419586182,
+      "learning_rate": 1.3934234977934463e-06,
+      "loss": 0.0297,
+      "step": 998
+    },
+    {
+      "epoch": 9.699029126213592,
+      "grad_norm": 0.49554625153541565,
+      "learning_rate": 1.3888674085886302e-06,
+      "loss": 0.0159,
+      "step": 999
+    },
+    {
+      "epoch": 9.70873786407767,
+      "grad_norm": 0.47762471437454224,
+      "learning_rate": 1.3843159135691859e-06,
+      "loss": 0.0232,
+      "step": 1000
+    },
+    {
+      "epoch": 9.718446601941748,
+      "grad_norm": 0.4060843586921692,
+      "learning_rate": 1.3797690315541193e-06,
+      "loss": 0.0168,
+      "step": 1001
+    },
+    {
+      "epoch": 9.728155339805825,
+      "grad_norm": 0.7846500873565674,
+      "learning_rate": 1.3752267813433645e-06,
+      "loss": 0.0289,
+      "step": 1002
+    },
+    {
+      "epoch": 9.737864077669903,
+      "grad_norm": 0.39978328347206116,
+      "learning_rate": 1.3706891817177036e-06,
+      "loss": 0.0081,
+      "step": 1003
+    },
+    {
+      "epoch": 9.74757281553398,
+      "grad_norm": 0.6862725019454956,
+      "learning_rate": 1.3661562514386895e-06,
+      "loss": 0.061,
+      "step": 1004
+    },
+    {
+      "epoch": 9.757281553398059,
+      "grad_norm": 0.9297090768814087,
+      "learning_rate": 1.3616280092485719e-06,
+      "loss": 0.0315,
+      "step": 1005
+    },
+    {
+      "epoch": 9.766990291262136,
+      "grad_norm": 1.188409686088562,
+      "learning_rate": 1.357104473870213e-06,
+      "loss": 0.0807,
+      "step": 1006
+    },
+    {
+      "epoch": 9.776699029126213,
+      "grad_norm": 0.6750743985176086,
+      "learning_rate": 1.3525856640070156e-06,
+      "loss": 0.0165,
+      "step": 1007
+    },
+    {
+      "epoch": 9.78640776699029,
+      "grad_norm": 0.6655329465866089,
+      "learning_rate": 1.3480715983428433e-06,
+      "loss": 0.0164,
+      "step": 1008
+    },
+    {
+      "epoch": 9.79611650485437,
+      "grad_norm": 0.8620071411132812,
+      "learning_rate": 1.3435622955419447e-06,
+      "loss": 0.0428,
+      "step": 1009
+    },
+    {
+      "epoch": 9.805825242718447,
+      "grad_norm": 0.4294649064540863,
+      "learning_rate": 1.3390577742488747e-06,
+      "loss": 0.0133,
+      "step": 1010
+    },
+    {
+      "epoch": 9.815533980582524,
+      "grad_norm": 1.072709321975708,
+      "learning_rate": 1.334558053088419e-06,
+      "loss": 0.0211,
+      "step": 1011
+    },
+    {
+      "epoch": 9.825242718446601,
+      "grad_norm": 0.9287241101264954,
+      "learning_rate": 1.3300631506655148e-06,
+      "loss": 0.0255,
+      "step": 1012
+    },
+    {
+      "epoch": 9.83495145631068,
+      "grad_norm": 1.010530948638916,
+      "learning_rate": 1.3255730855651772e-06,
+      "loss": 0.0495,
+      "step": 1013
+    },
+    {
+      "epoch": 9.844660194174757,
+      "grad_norm": 0.4488737881183624,
+      "learning_rate": 1.3210878763524186e-06,
+      "loss": 0.0232,
+      "step": 1014
+    },
+    {
+      "epoch": 9.854368932038835,
+      "grad_norm": 1.137284755706787,
+      "learning_rate": 1.3166075415721762e-06,
+      "loss": 0.0491,
+      "step": 1015
+    },
+    {
+      "epoch": 9.864077669902912,
+      "grad_norm": 0.5317333936691284,
+      "learning_rate": 1.3121320997492305e-06,
+      "loss": 0.014,
+      "step": 1016
+    },
+    {
+      "epoch": 9.87378640776699,
+      "grad_norm": 0.8078583478927612,
+      "learning_rate": 1.307661569388132e-06,
+      "loss": 0.0242,
+      "step": 1017
+    },
+    {
+      "epoch": 9.883495145631068,
+      "grad_norm": 0.6492327451705933,
+      "learning_rate": 1.3031959689731236e-06,
+      "loss": 0.0162,
+      "step": 1018
+    },
+    {
+      "epoch": 9.893203883495145,
+      "grad_norm": 0.6473954916000366,
+      "learning_rate": 1.2987353169680667e-06,
+      "loss": 0.0526,
+      "step": 1019
+    },
+    {
+      "epoch": 9.902912621359224,
+      "grad_norm": 0.710628092288971,
+      "learning_rate": 1.2942796318163595e-06,
+      "loss": 0.015,
+      "step": 1020
+    },
+    {
+      "epoch": 9.912621359223301,
+      "grad_norm": 0.4777820408344269,
+      "learning_rate": 1.2898289319408653e-06,
+      "loss": 0.0178,
+      "step": 1021
+    },
+    {
+      "epoch": 9.922330097087379,
+      "grad_norm": 0.9042239189147949,
+      "learning_rate": 1.2853832357438346e-06,
+      "loss": 0.0242,
+      "step": 1022
+    },
+    {
+      "epoch": 9.932038834951456,
+      "grad_norm": 0.5531700849533081,
+      "learning_rate": 1.2809425616068288e-06,
+      "loss": 0.0339,
+      "step": 1023
+    },
+    {
+      "epoch": 9.941747572815533,
+      "grad_norm": 1.1424311399459839,
+      "learning_rate": 1.2765069278906456e-06,
+      "loss": 0.0472,
+      "step": 1024
+    },
+    {
+      "epoch": 9.951456310679612,
+      "grad_norm": 0.8695238828659058,
+      "learning_rate": 1.2720763529352415e-06,
+      "loss": 0.0071,
+      "step": 1025
+    },
+    {
+      "epoch": 9.96116504854369,
+      "grad_norm": 0.4688761234283447,
+      "learning_rate": 1.2676508550596562e-06,
+      "loss": 0.0061,
+      "step": 1026
+    },
+    {
+      "epoch": 9.970873786407767,
+      "grad_norm": 0.5627487301826477,
+      "learning_rate": 1.2632304525619388e-06,
+      "loss": 0.011,
+      "step": 1027
+    },
+    {
+      "epoch": 9.980582524271846,
+      "grad_norm": 1.3065028190612793,
+      "learning_rate": 1.2588151637190687e-06,
+      "loss": 0.0695,
+      "step": 1028
+    },
+    {
+      "epoch": 9.990291262135923,
+      "grad_norm": 0.5717766284942627,
+      "learning_rate": 1.2544050067868834e-06,
+      "loss": 0.01,
+      "step": 1029
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.7852147817611694,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.0306,
+      "step": 1030
+    },
+    {
+      "epoch": 10.009708737864077,
+      "grad_norm": 0.40971171855926514,
+      "learning_rate": 1.2456001615717445e-06,
+      "loss": 0.0079,
+      "step": 1031
+    },
+    {
+      "epoch": 10.019417475728156,
+      "grad_norm": 0.39056044816970825,
+      "learning_rate": 1.2412055096940692e-06,
+      "loss": 0.0129,
+      "step": 1032
+    },
+    {
+      "epoch": 10.029126213592233,
+      "grad_norm": 0.9858739376068115,
+      "learning_rate": 1.2368160625374835e-06,
+      "loss": 0.0261,
+      "step": 1033
+    },
+    {
+      "epoch": 10.03883495145631,
+      "grad_norm": 0.2877596616744995,
+      "learning_rate": 1.2324318382509787e-06,
+      "loss": 0.0103,
+      "step": 1034
+    },
+    {
+      "epoch": 10.048543689320388,
+      "grad_norm": 0.29269400238990784,
+      "learning_rate": 1.2280528549619487e-06,
+      "loss": 0.0041,
+      "step": 1035
+    },
+    {
+      "epoch": 10.058252427184467,
+      "grad_norm": 0.5697181820869446,
+      "learning_rate": 1.2236791307761184e-06,
+      "loss": 0.0243,
+      "step": 1036
+    },
+    {
+      "epoch": 10.067961165048544,
+      "grad_norm": 0.6563681364059448,
+      "learning_rate": 1.2193106837774678e-06,
+      "loss": 0.0055,
+      "step": 1037
+    },
+    {
+      "epoch": 10.077669902912621,
+      "grad_norm": 0.33351874351501465,
+      "learning_rate": 1.2149475320281578e-06,
+      "loss": 0.0033,
+      "step": 1038
+    },
+    {
+      "epoch": 10.087378640776699,
+      "grad_norm": 0.906010091304779,
+      "learning_rate": 1.2105896935684545e-06,
+      "loss": 0.016,
+      "step": 1039
+    },
+    {
+      "epoch": 10.097087378640778,
+      "grad_norm": 0.5395269393920898,
+      "learning_rate": 1.2062371864166553e-06,
+      "loss": 0.0265,
+      "step": 1040
+    },
+    {
+      "epoch": 10.106796116504855,
+      "grad_norm": 0.5314574837684631,
+      "learning_rate": 1.2018900285690148e-06,
+      "loss": 0.0065,
+      "step": 1041
+    },
+    {
+      "epoch": 10.116504854368932,
+      "grad_norm": 0.41846874356269836,
+      "learning_rate": 1.1975482379996697e-06,
+      "loss": 0.0175,
+      "step": 1042
+    },
+    {
+      "epoch": 10.12621359223301,
+      "grad_norm": 0.6140009760856628,
+      "learning_rate": 1.1932118326605644e-06,
+      "loss": 0.0279,
+      "step": 1043
+    },
+    {
+      "epoch": 10.135922330097088,
+      "grad_norm": 0.260271281003952,
+      "learning_rate": 1.188880830481377e-06,
+      "loss": 0.0017,
+      "step": 1044
+    },
+    {
+      "epoch": 10.145631067961165,
+      "grad_norm": 0.30545926094055176,
+      "learning_rate": 1.1845552493694462e-06,
+      "loss": 0.0069,
+      "step": 1045
+    },
+    {
+      "epoch": 10.155339805825243,
+      "grad_norm": 0.42265552282333374,
+      "learning_rate": 1.1802351072096948e-06,
+      "loss": 0.0043,
+      "step": 1046
+    },
+    {
+      "epoch": 10.16504854368932,
+      "grad_norm": 0.24939106404781342,
+      "learning_rate": 1.1759204218645577e-06,
+      "loss": 0.0037,
+      "step": 1047
+    },
+    {
+      "epoch": 10.174757281553399,
+      "grad_norm": 0.4412282705307007,
+      "learning_rate": 1.1716112111739095e-06,
+      "loss": 0.0057,
+      "step": 1048
+    },
+    {
+      "epoch": 10.184466019417476,
+      "grad_norm": 1.1326662302017212,
+      "learning_rate": 1.167307492954986e-06,
+      "loss": 0.036,
+      "step": 1049
+    },
+    {
+      "epoch": 10.194174757281553,
+      "grad_norm": 0.5459797382354736,
+      "learning_rate": 1.1630092850023148e-06,
+      "loss": 0.0164,
+      "step": 1050
+    },
+    {
+      "epoch": 10.20388349514563,
+      "grad_norm": 1.0390825271606445,
+      "learning_rate": 1.15871660508764e-06,
+      "loss": 0.0239,
+      "step": 1051
+    },
+    {
+      "epoch": 10.21359223300971,
+      "grad_norm": 0.6237937211990356,
+      "learning_rate": 1.1544294709598491e-06,
+      "loss": 0.0182,
+      "step": 1052
+    },
+    {
+      "epoch": 10.223300970873787,
+      "grad_norm": 0.4918019473552704,
+      "learning_rate": 1.1501479003448992e-06,
+      "loss": 0.0145,
+      "step": 1053
+    },
+    {
+      "epoch": 10.233009708737864,
+      "grad_norm": 0.3002013564109802,
+      "learning_rate": 1.1458719109457445e-06,
+      "loss": 0.0049,
+      "step": 1054
+    },
+    {
+      "epoch": 10.242718446601941,
+      "grad_norm": 0.6791167259216309,
+      "learning_rate": 1.141601520442262e-06,
+      "loss": 0.0175,
+      "step": 1055
+    },
+    {
+      "epoch": 10.25242718446602,
+      "grad_norm": 0.25156375765800476,
+      "learning_rate": 1.1373367464911798e-06,
+      "loss": 0.0045,
+      "step": 1056
+    },
+    {
+      "epoch": 10.262135922330097,
+      "grad_norm": 0.5771428346633911,
+      "learning_rate": 1.1330776067260026e-06,
+      "loss": 0.0454,
+      "step": 1057
+    },
+    {
+      "epoch": 10.271844660194175,
+      "grad_norm": 0.40346211194992065,
+      "learning_rate": 1.12882411875694e-06,
+      "loss": 0.0025,
+      "step": 1058
+    },
+    {
+      "epoch": 10.281553398058252,
+      "grad_norm": 0.2417290210723877,
+      "learning_rate": 1.1245763001708326e-06,
+      "loss": 0.0032,
+      "step": 1059
+    },
+    {
+      "epoch": 10.29126213592233,
+      "grad_norm": 0.37248948216438293,
+      "learning_rate": 1.120334168531081e-06,
+      "loss": 0.0069,
+      "step": 1060
+    },
+    {
+      "epoch": 10.300970873786408,
+      "grad_norm": 0.6066130995750427,
+      "learning_rate": 1.1160977413775704e-06,
+      "loss": 0.0322,
+      "step": 1061
+    },
+    {
+      "epoch": 10.310679611650485,
+      "grad_norm": 0.22785338759422302,
+      "learning_rate": 1.1118670362266003e-06,
+      "loss": 0.0068,
+      "step": 1062
+    },
+    {
+      "epoch": 10.320388349514563,
+      "grad_norm": 0.4489402174949646,
+      "learning_rate": 1.1076420705708137e-06,
+      "loss": 0.0047,
+      "step": 1063
+    },
+    {
+      "epoch": 10.330097087378642,
+      "grad_norm": 0.37608322501182556,
+      "learning_rate": 1.1034228618791197e-06,
+      "loss": 0.0007,
+      "step": 1064
+    },
+    {
+      "epoch": 10.339805825242719,
+      "grad_norm": 0.34204986691474915,
+      "learning_rate": 1.0992094275966256e-06,
+      "loss": 0.0058,
+      "step": 1065
+    },
+    {
+      "epoch": 10.349514563106796,
+      "grad_norm": 0.7375351190567017,
+      "learning_rate": 1.0950017851445624e-06,
+      "loss": 0.0098,
+      "step": 1066
+    },
+    {
+      "epoch": 10.359223300970873,
+      "grad_norm": 0.7323979735374451,
+      "learning_rate": 1.0907999519202142e-06,
+      "loss": 0.0252,
+      "step": 1067
+    },
+    {
+      "epoch": 10.368932038834952,
+      "grad_norm": 0.549560546875,
+      "learning_rate": 1.0866039452968464e-06,
+      "loss": 0.0105,
+      "step": 1068
+    },
+    {
+      "epoch": 10.37864077669903,
+      "grad_norm": 0.565077006816864,
+      "learning_rate": 1.0824137826236318e-06,
+      "loss": 0.0008,
+      "step": 1069
+    },
+    {
+      "epoch": 10.388349514563107,
+      "grad_norm": 0.17279475927352905,
+      "learning_rate": 1.078229481225582e-06,
+      "loss": 0.0021,
+      "step": 1070
+    },
+    {
+      "epoch": 10.398058252427184,
+      "grad_norm": 0.5564970374107361,
+      "learning_rate": 1.074051058403472e-06,
+      "loss": 0.0166,
+      "step": 1071
+    },
+    {
+      "epoch": 10.407766990291263,
+      "grad_norm": 0.49482810497283936,
+      "learning_rate": 1.069878531433773e-06,
+      "loss": 0.0057,
+      "step": 1072
+    },
+    {
+      "epoch": 10.41747572815534,
+      "grad_norm": 0.3255504369735718,
+      "learning_rate": 1.0657119175685776e-06,
+      "loss": 0.0053,
+      "step": 1073
+    },
+    {
+      "epoch": 10.427184466019417,
+      "grad_norm": 0.18100124597549438,
+      "learning_rate": 1.061551234035529e-06,
+      "loss": 0.0015,
+      "step": 1074
+    },
+    {
+      "epoch": 10.436893203883495,
+      "grad_norm": 0.6678428649902344,
+      "learning_rate": 1.0573964980377517e-06,
+      "loss": 0.0254,
+      "step": 1075
+    },
+    {
+      "epoch": 10.446601941747574,
+      "grad_norm": 0.5514142513275146,
+      "learning_rate": 1.0532477267537772e-06,
+      "loss": 0.0571,
+      "step": 1076
+    },
+    {
+      "epoch": 10.45631067961165,
+      "grad_norm": 0.6988746523857117,
+      "learning_rate": 1.0491049373374762e-06,
+      "loss": 0.0575,
+      "step": 1077
+    },
+    {
+      "epoch": 10.466019417475728,
+      "grad_norm": 0.6181220412254333,
+      "learning_rate": 1.044968146917986e-06,
+      "loss": 0.0136,
+      "step": 1078
+    },
+    {
+      "epoch": 10.475728155339805,
+      "grad_norm": 0.2651762068271637,
+      "learning_rate": 1.0408373725996386e-06,
+      "loss": 0.0024,
+      "step": 1079
+    },
+    {
+      "epoch": 10.485436893203884,
+      "grad_norm": 0.5839412212371826,
+      "learning_rate": 1.0367126314618946e-06,
+      "loss": 0.0071,
+      "step": 1080
+    },
+    {
+      "epoch": 10.495145631067961,
+      "grad_norm": 0.629555881023407,
+      "learning_rate": 1.0325939405592661e-06,
+      "loss": 0.023,
+      "step": 1081
+    },
+    {
+      "epoch": 10.504854368932039,
+      "grad_norm": 0.2746008038520813,
+      "learning_rate": 1.0284813169212502e-06,
+      "loss": 0.0038,
+      "step": 1082
+    },
+    {
+      "epoch": 10.514563106796116,
+      "grad_norm": 0.5150997042655945,
+      "learning_rate": 1.024374777552258e-06,
+      "loss": 0.0062,
+      "step": 1083
+    },
+    {
+      "epoch": 10.524271844660195,
+      "grad_norm": 0.5501676201820374,
+      "learning_rate": 1.0202743394315444e-06,
+      "loss": 0.0069,
+      "step": 1084
+    },
+    {
+      "epoch": 10.533980582524272,
+      "grad_norm": 0.5312268733978271,
+      "learning_rate": 1.0161800195131372e-06,
+      "loss": 0.0074,
+      "step": 1085
+    },
+    {
+      "epoch": 10.54368932038835,
+      "grad_norm": 0.4340658187866211,
+      "learning_rate": 1.0120918347257669e-06,
+      "loss": 0.0101,
+      "step": 1086
+    },
+    {
+      "epoch": 10.553398058252426,
+      "grad_norm": 0.27946993708610535,
+      "learning_rate": 1.0080098019727979e-06,
+      "loss": 0.0071,
+      "step": 1087
+    },
+    {
+      "epoch": 10.563106796116505,
+      "grad_norm": 1.1418503522872925,
+      "learning_rate": 1.0039339381321572e-06,
+      "loss": 0.0417,
+      "step": 1088
+    },
+    {
+      "epoch": 10.572815533980583,
+      "grad_norm": 0.98741614818573,
+      "learning_rate": 9.998642600562664e-07,
+      "loss": 0.0409,
+      "step": 1089
+    },
+    {
+      "epoch": 10.58252427184466,
+      "grad_norm": 0.7151482105255127,
+      "learning_rate": 9.95800784571969e-07,
+      "loss": 0.0205,
+      "step": 1090
+    },
+    {
+      "epoch": 10.592233009708737,
+      "grad_norm": 0.4826299250125885,
+      "learning_rate": 9.91743528480464e-07,
+      "loss": 0.0229,
+      "step": 1091
+    },
+    {
+      "epoch": 10.601941747572816,
+      "grad_norm": 0.7173331379890442,
+      "learning_rate": 9.876925085572365e-07,
+      "loss": 0.0098,
+      "step": 1092
+    },
+    {
+      "epoch": 10.611650485436893,
+      "grad_norm": 0.4329473376274109,
+      "learning_rate": 9.836477415519843e-07,
+      "loss": 0.0136,
+      "step": 1093
+    },
+    {
+      "epoch": 10.62135922330097,
+      "grad_norm": 0.4952574074268341,
+      "learning_rate": 9.79609244188553e-07,
+      "loss": 0.0101,
+      "step": 1094
+    },
+    {
+      "epoch": 10.631067961165048,
+      "grad_norm": 0.3168066442012787,
+      "learning_rate": 9.755770331648642e-07,
+      "loss": 0.0132,
+      "step": 1095
+    },
+    {
+      "epoch": 10.640776699029127,
+      "grad_norm": 0.4660727381706238,
+      "learning_rate": 9.715511251528486e-07,
+      "loss": 0.0275,
+      "step": 1096
+    },
+    {
+      "epoch": 10.650485436893204,
+      "grad_norm": 0.6679536700248718,
+      "learning_rate": 9.67531536798375e-07,
+      "loss": 0.0236,
+      "step": 1097
+    },
+    {
+      "epoch": 10.660194174757281,
+      "grad_norm": 0.6162199378013611,
+      "learning_rate": 9.635182847211827e-07,
+      "loss": 0.0219,
+      "step": 1098
+    },
+    {
+      "epoch": 10.669902912621358,
+      "grad_norm": 0.9909278154373169,
+      "learning_rate": 9.595113855148128e-07,
+      "loss": 0.0132,
+      "step": 1099
+    },
+    {
+      "epoch": 10.679611650485437,
+      "grad_norm": 1.3139898777008057,
+      "learning_rate": 9.555108557465383e-07,
+      "loss": 0.0774,
+      "step": 1100
+    },
+    {
+      "epoch": 10.689320388349515,
+      "grad_norm": 0.6796632409095764,
+      "learning_rate": 9.51516711957298e-07,
+      "loss": 0.0432,
+      "step": 1101
+    },
+    {
+      "epoch": 10.699029126213592,
+      "grad_norm": 0.4316011965274811,
+      "learning_rate": 9.475289706616256e-07,
+      "loss": 0.0149,
+      "step": 1102
+    },
+    {
+      "epoch": 10.70873786407767,
+      "grad_norm": 0.5164605975151062,
+      "learning_rate": 9.435476483475825e-07,
+      "loss": 0.0199,
+      "step": 1103
+    },
+    {
+      "epoch": 10.718446601941748,
+      "grad_norm": 0.45947888493537903,
+      "learning_rate": 9.395727614766903e-07,
+      "loss": 0.0142,
+      "step": 1104
+    },
+    {
+      "epoch": 10.728155339805825,
+      "grad_norm": 0.651980459690094,
+      "learning_rate": 9.356043264838607e-07,
+      "loss": 0.0112,
+      "step": 1105
+    },
+    {
+      "epoch": 10.737864077669903,
+      "grad_norm": 0.3955351710319519,
+      "learning_rate": 9.316423597773316e-07,
+      "loss": 0.004,
+      "step": 1106
+    },
+    {
+      "epoch": 10.74757281553398,
+      "grad_norm": 0.6277558207511902,
+      "learning_rate": 9.276868777385942e-07,
+      "loss": 0.0408,
+      "step": 1107
+    },
+    {
+      "epoch": 10.757281553398059,
+      "grad_norm": 0.4933006465435028,
+      "learning_rate": 9.237378967223279e-07,
+      "loss": 0.0151,
+      "step": 1108
+    },
+    {
+      "epoch": 10.766990291262136,
+      "grad_norm": 0.590498149394989,
+      "learning_rate": 9.197954330563327e-07,
+      "loss": 0.0178,
+      "step": 1109
+    },
+    {
+      "epoch": 10.776699029126213,
+      "grad_norm": 0.2939163148403168,
+      "learning_rate": 9.158595030414621e-07,
+      "loss": 0.0114,
+      "step": 1110
+    },
+    {
+      "epoch": 10.78640776699029,
+      "grad_norm": 0.6207811832427979,
+      "learning_rate": 9.11930122951554e-07,
+      "loss": 0.0099,
+      "step": 1111
+    },
+    {
+      "epoch": 10.79611650485437,
+      "grad_norm": 0.6594740152359009,
+      "learning_rate": 9.080073090333646e-07,
+      "loss": 0.0104,
+      "step": 1112
+    },
+    {
+      "epoch": 10.805825242718447,
+      "grad_norm": 0.7800750732421875,
+      "learning_rate": 9.040910775065015e-07,
+      "loss": 0.0211,
+      "step": 1113
+    },
+    {
+      "epoch": 10.815533980582524,
+      "grad_norm": 0.402415007352829,
+      "learning_rate": 9.001814445633558e-07,
+      "loss": 0.004,
+      "step": 1114
+    },
+    {
+      "epoch": 10.825242718446601,
+      "grad_norm": 0.7148779630661011,
+      "learning_rate": 8.962784263690358e-07,
+      "loss": 0.0198,
+      "step": 1115
+    },
+    {
+      "epoch": 10.83495145631068,
+      "grad_norm": 0.647210419178009,
+      "learning_rate": 8.923820390612991e-07,
+      "loss": 0.0121,
+      "step": 1116
+    },
+    {
+      "epoch": 10.844660194174757,
+      "grad_norm": 0.512809157371521,
+      "learning_rate": 8.884922987504882e-07,
+      "loss": 0.0097,
+      "step": 1117
+    },
+    {
+      "epoch": 10.854368932038835,
+      "grad_norm": 0.6944689750671387,
+      "learning_rate": 8.846092215194607e-07,
+      "loss": 0.0308,
+      "step": 1118
+    },
+    {
+      "epoch": 10.864077669902912,
+      "grad_norm": 0.5818324685096741,
+      "learning_rate": 8.807328234235254e-07,
+      "loss": 0.0058,
+      "step": 1119
+    },
+    {
+      "epoch": 10.87378640776699,
+      "grad_norm": 0.8093370795249939,
+      "learning_rate": 8.768631204903738e-07,
+      "loss": 0.024,
+      "step": 1120
+    },
+    {
+      "epoch": 10.883495145631068,
+      "grad_norm": 0.8604028820991516,
+      "learning_rate": 8.730001287200177e-07,
+      "loss": 0.0256,
+      "step": 1121
+    },
+    {
+      "epoch": 10.893203883495145,
+      "grad_norm": 0.33591437339782715,
+      "learning_rate": 8.691438640847177e-07,
+      "loss": 0.005,
+      "step": 1122
+    },
+    {
+      "epoch": 10.902912621359224,
+      "grad_norm": 1.100368618965149,
+      "learning_rate": 8.652943425289206e-07,
+      "loss": 0.026,
+      "step": 1123
+    },
+    {
+      "epoch": 10.912621359223301,
+      "grad_norm": 0.29850178956985474,
+      "learning_rate": 8.61451579969193e-07,
+      "loss": 0.0122,
+      "step": 1124
+    },
+    {
+      "epoch": 10.922330097087379,
+      "grad_norm": 0.5824030041694641,
+      "learning_rate": 8.576155922941548e-07,
+      "loss": 0.0179,
+      "step": 1125
+    },
+    {
+      "epoch": 10.932038834951456,
+      "grad_norm": 0.3939349949359894,
+      "learning_rate": 8.537863953644138e-07,
+      "loss": 0.0134,
+      "step": 1126
+    },
+    {
+      "epoch": 10.941747572815533,
+      "grad_norm": 0.5187537670135498,
+      "learning_rate": 8.499640050125007e-07,
+      "loss": 0.0127,
+      "step": 1127
+    },
+    {
+      "epoch": 10.951456310679612,
+      "grad_norm": 0.5530067086219788,
+      "learning_rate": 8.461484370428025e-07,
+      "loss": 0.0273,
+      "step": 1128
+    },
+    {
+      "epoch": 10.96116504854369,
+      "grad_norm": 0.8861692547798157,
+      "learning_rate": 8.423397072314985e-07,
+      "loss": 0.0151,
+      "step": 1129
+    },
+    {
+      "epoch": 10.970873786407767,
+      "grad_norm": 0.39885714650154114,
+      "learning_rate": 8.385378313264933e-07,
+      "loss": 0.0088,
+      "step": 1130
+    },
+    {
+      "epoch": 10.980582524271846,
+      "grad_norm": 0.6531423330307007,
+      "learning_rate": 8.347428250473541e-07,
+      "loss": 0.007,
+      "step": 1131
+    },
+    {
+      "epoch": 10.990291262135923,
+      "grad_norm": 0.36868107318878174,
+      "learning_rate": 8.309547040852434e-07,
+      "loss": 0.0031,
+      "step": 1132
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.3329010307788849,
+      "learning_rate": 8.271734841028553e-07,
+      "loss": 0.0075,
+      "step": 1133
+    },
+    {
+      "epoch": 11.009708737864077,
+      "grad_norm": 0.30606314539909363,
+      "learning_rate": 8.233991807343497e-07,
+      "loss": 0.0086,
+      "step": 1134
+    },
+    {
+      "epoch": 11.019417475728156,
+      "grad_norm": 0.2135513722896576,
+      "learning_rate": 8.196318095852909e-07,
+      "loss": 0.0026,
+      "step": 1135
+    },
+    {
+      "epoch": 11.029126213592233,
+      "grad_norm": 0.2945385277271271,
+      "learning_rate": 8.158713862325782e-07,
+      "loss": 0.0131,
+      "step": 1136
+    },
+    {
+      "epoch": 11.03883495145631,
+      "grad_norm": 0.2576274275779724,
+      "learning_rate": 8.12117926224385e-07,
+      "loss": 0.0038,
+      "step": 1137
+    },
+    {
+      "epoch": 11.048543689320388,
+      "grad_norm": 0.1427377313375473,
+      "learning_rate": 8.08371445080093e-07,
+      "loss": 0.0027,
+      "step": 1138
+    },
+    {
+      "epoch": 11.058252427184467,
+      "grad_norm": 0.20165517926216125,
+      "learning_rate": 8.04631958290229e-07,
+      "loss": 0.0017,
+      "step": 1139
+    },
+    {
+      "epoch": 11.067961165048544,
+      "grad_norm": 0.5164278149604797,
+      "learning_rate": 8.008994813163995e-07,
+      "loss": 0.0131,
+      "step": 1140
+    },
+    {
+      "epoch": 11.077669902912621,
+      "grad_norm": 0.18127082288265228,
+      "learning_rate": 7.971740295912289e-07,
+      "loss": 0.0053,
+      "step": 1141
+    },
+    {
+      "epoch": 11.087378640776699,
+      "grad_norm": 0.21812501549720764,
+      "learning_rate": 7.934556185182928e-07,
+      "loss": 0.0052,
+      "step": 1142
+    },
+    {
+      "epoch": 11.097087378640778,
+      "grad_norm": 0.2931056320667267,
+      "learning_rate": 7.897442634720576e-07,
+      "loss": 0.0077,
+      "step": 1143
+    },
+    {
+      "epoch": 11.106796116504855,
+      "grad_norm": 0.7248398065567017,
+      "learning_rate": 7.860399797978138e-07,
+      "loss": 0.0627,
+      "step": 1144
+    },
+    {
+      "epoch": 11.116504854368932,
+      "grad_norm": 0.37893545627593994,
+      "learning_rate": 7.823427828116148e-07,
+      "loss": 0.0108,
+      "step": 1145
+    },
+    {
+      "epoch": 11.12621359223301,
+      "grad_norm": 0.2130487859249115,
+      "learning_rate": 7.786526878002126e-07,
+      "loss": 0.0019,
+      "step": 1146
+    },
+    {
+      "epoch": 11.135922330097088,
+      "grad_norm": 0.3752915859222412,
+      "learning_rate": 7.749697100209947e-07,
+      "loss": 0.0082,
+      "step": 1147
+    },
+    {
+      "epoch": 11.145631067961165,
+      "grad_norm": 0.27648448944091797,
+      "learning_rate": 7.7129386470192e-07,
+      "loss": 0.0031,
+      "step": 1148
+    },
+    {
+      "epoch": 11.155339805825243,
+      "grad_norm": 0.3961423933506012,
+      "learning_rate": 7.6762516704146e-07,
+      "loss": 0.0107,
+      "step": 1149
+    },
+    {
+      "epoch": 11.16504854368932,
+      "grad_norm": 0.5809851288795471,
+      "learning_rate": 7.6396363220853e-07,
+      "loss": 0.0158,
+      "step": 1150
+    },
+    {
+      "epoch": 11.174757281553399,
+      "grad_norm": 0.3201679587364197,
+      "learning_rate": 7.603092753424298e-07,
+      "loss": 0.0087,
+      "step": 1151
+    },
+    {
+      "epoch": 11.184466019417476,
+      "grad_norm": 0.388967365026474,
+      "learning_rate": 7.566621115527811e-07,
+      "loss": 0.0027,
+      "step": 1152
+    },
+    {
+      "epoch": 11.194174757281553,
+      "grad_norm": 0.22992132604122162,
+      "learning_rate": 7.530221559194643e-07,
+      "loss": 0.0016,
+      "step": 1153
+    },
+    {
+      "epoch": 11.20388349514563,
+      "grad_norm": 0.5471656322479248,
+      "learning_rate": 7.493894234925558e-07,
+      "loss": 0.0028,
+      "step": 1154
+    },
+    {
+      "epoch": 11.21359223300971,
+      "grad_norm": 0.40595000982284546,
+      "learning_rate": 7.457639292922675e-07,
+      "loss": 0.0142,
+      "step": 1155
+    },
+    {
+      "epoch": 11.223300970873787,
+      "grad_norm": 0.19051553308963776,
+      "learning_rate": 7.421456883088826e-07,
+      "loss": 0.0008,
+      "step": 1156
+    },
+    {
+      "epoch": 11.233009708737864,
+      "grad_norm": 0.3705807030200958,
+      "learning_rate": 7.385347155026934e-07,
+      "loss": 0.0096,
+      "step": 1157
+    },
+    {
+      "epoch": 11.242718446601941,
+      "grad_norm": 0.5367113947868347,
+      "learning_rate": 7.349310258039441e-07,
+      "loss": 0.0106,
+      "step": 1158
+    },
+    {
+      "epoch": 11.25242718446602,
+      "grad_norm": 0.7418363094329834,
+      "learning_rate": 7.31334634112762e-07,
+      "loss": 0.0365,
+      "step": 1159
+    },
+    {
+      "epoch": 11.262135922330097,
+      "grad_norm": 0.6008816957473755,
+      "learning_rate": 7.277455552991011e-07,
+      "loss": 0.0241,
+      "step": 1160
+    },
+    {
+      "epoch": 11.271844660194175,
+      "grad_norm": 0.24267537891864777,
+      "learning_rate": 7.241638042026783e-07,
+      "loss": 0.0045,
+      "step": 1161
+    },
+    {
+      "epoch": 11.281553398058252,
+      "grad_norm": 0.17620030045509338,
+      "learning_rate": 7.20589395632913e-07,
+      "loss": 0.0041,
+      "step": 1162
+    },
+    {
+      "epoch": 11.29126213592233,
+      "grad_norm": 0.7443981170654297,
+      "learning_rate": 7.170223443688654e-07,
+      "loss": 0.0167,
+      "step": 1163
+    },
+    {
+      "epoch": 11.300970873786408,
+      "grad_norm": 0.4085586369037628,
+      "learning_rate": 7.134626651591758e-07,
+      "loss": 0.0103,
+      "step": 1164
+    },
+    {
+      "epoch": 11.310679611650485,
+      "grad_norm": 0.10864268243312836,
+      "learning_rate": 7.099103727220024e-07,
+      "loss": 0.0017,
+      "step": 1165
+    },
+    {
+      "epoch": 11.320388349514563,
+      "grad_norm": 0.9929929375648499,
+      "learning_rate": 7.063654817449638e-07,
+      "loss": 0.0271,
+      "step": 1166
+    },
+    {
+      "epoch": 11.330097087378642,
+      "grad_norm": 0.14670035243034363,
+      "learning_rate": 7.028280068850734e-07,
+      "loss": 0.0018,
+      "step": 1167
+    },
+    {
+      "epoch": 11.339805825242719,
+      "grad_norm": 0.8203486204147339,
+      "learning_rate": 6.992979627686821e-07,
+      "loss": 0.0103,
+      "step": 1168
+    },
+    {
+      "epoch": 11.349514563106796,
+      "grad_norm": 0.5324729681015015,
+      "learning_rate": 6.957753639914175e-07,
+      "loss": 0.0137,
+      "step": 1169
+    },
+    {
+      "epoch": 11.359223300970873,
+      "grad_norm": 0.31972795724868774,
+      "learning_rate": 6.922602251181221e-07,
+      "loss": 0.0054,
+      "step": 1170
+    },
+    {
+      "epoch": 11.368932038834952,
+      "grad_norm": 0.8650928139686584,
+      "learning_rate": 6.887525606827947e-07,
+      "loss": 0.0189,
+      "step": 1171
+    },
+    {
+      "epoch": 11.37864077669903,
+      "grad_norm": 0.09976081550121307,
+      "learning_rate": 6.852523851885295e-07,
+      "loss": 0.0013,
+      "step": 1172
+    },
+    {
+      "epoch": 11.388349514563107,
+      "grad_norm": 0.3876700699329376,
+      "learning_rate": 6.817597131074566e-07,
+      "loss": 0.0032,
+      "step": 1173
+    },
+    {
+      "epoch": 11.398058252427184,
+      "grad_norm": 0.2802983820438385,
+      "learning_rate": 6.782745588806811e-07,
+      "loss": 0.0039,
+      "step": 1174
+    },
+    {
+      "epoch": 11.407766990291263,
+      "grad_norm": 0.2554693818092346,
+      "learning_rate": 6.747969369182248e-07,
+      "loss": 0.0071,
+      "step": 1175
+    },
+    {
+      "epoch": 11.41747572815534,
+      "grad_norm": 0.6343085169792175,
+      "learning_rate": 6.713268615989654e-07,
+      "loss": 0.0075,
+      "step": 1176
+    },
+    {
+      "epoch": 11.427184466019417,
+      "grad_norm": 0.46486198902130127,
+      "learning_rate": 6.678643472705773e-07,
+      "loss": 0.0209,
+      "step": 1177
+    },
+    {
+      "epoch": 11.436893203883495,
+      "grad_norm": 0.12777316570281982,
+      "learning_rate": 6.644094082494746e-07,
+      "loss": 0.0014,
+      "step": 1178
+    },
+    {
+      "epoch": 11.446601941747574,
+      "grad_norm": 0.43883708119392395,
+      "learning_rate": 6.609620588207474e-07,
+      "loss": 0.0191,
+      "step": 1179
+    },
+    {
+      "epoch": 11.45631067961165,
+      "grad_norm": 0.40817007422447205,
+      "learning_rate": 6.575223132381067e-07,
+      "loss": 0.0101,
+      "step": 1180
+    },
+    {
+      "epoch": 11.466019417475728,
+      "grad_norm": 0.35821834206581116,
+      "learning_rate": 6.540901857238233e-07,
+      "loss": 0.0087,
+      "step": 1181
+    },
+    {
+      "epoch": 11.475728155339805,
+      "grad_norm": 0.4091426432132721,
+      "learning_rate": 6.506656904686698e-07,
+      "loss": 0.0107,
+      "step": 1182
+    },
+    {
+      "epoch": 11.485436893203884,
+      "grad_norm": 0.40708640217781067,
+      "learning_rate": 6.472488416318621e-07,
+      "loss": 0.0104,
+      "step": 1183
+    },
+    {
+      "epoch": 11.495145631067961,
+      "grad_norm": 0.9649088978767395,
+      "learning_rate": 6.438396533410002e-07,
+      "loss": 0.0066,
+      "step": 1184
+    },
+    {
+      "epoch": 11.504854368932039,
+      "grad_norm": 0.22220736742019653,
+      "learning_rate": 6.4043813969201e-07,
+      "loss": 0.0018,
+      "step": 1185
+    },
+    {
+      "epoch": 11.514563106796116,
+      "grad_norm": 0.9987980723381042,
+      "learning_rate": 6.370443147490857e-07,
+      "loss": 0.0144,
+      "step": 1186
+    },
+    {
+      "epoch": 11.524271844660195,
+      "grad_norm": 0.24521130323410034,
+      "learning_rate": 6.336581925446309e-07,
+      "loss": 0.0067,
+      "step": 1187
+    },
+    {
+      "epoch": 11.533980582524272,
+      "grad_norm": 0.4199708104133606,
+      "learning_rate": 6.302797870792007e-07,
+      "loss": 0.0094,
+      "step": 1188
+    },
+    {
+      "epoch": 11.54368932038835,
+      "grad_norm": 0.2658780813217163,
+      "learning_rate": 6.269091123214438e-07,
+      "loss": 0.0038,
+      "step": 1189
+    },
+    {
+      "epoch": 11.553398058252426,
+      "grad_norm": 0.5286270380020142,
+      "learning_rate": 6.235461822080449e-07,
+      "loss": 0.0121,
+      "step": 1190
+    },
+    {
+      "epoch": 11.563106796116505,
+      "grad_norm": 0.3407003581523895,
+      "learning_rate": 6.201910106436673e-07,
+      "loss": 0.0075,
+      "step": 1191
+    },
+    {
+      "epoch": 11.572815533980583,
+      "grad_norm": 0.36705833673477173,
+      "learning_rate": 6.168436115008941e-07,
+      "loss": 0.0117,
+      "step": 1192
+    },
+    {
+      "epoch": 11.58252427184466,
+      "grad_norm": 0.31512248516082764,
+      "learning_rate": 6.135039986201744e-07,
+      "loss": 0.0025,
+      "step": 1193
+    },
+    {
+      "epoch": 11.592233009708737,
+      "grad_norm": 0.6591824889183044,
+      "learning_rate": 6.101721858097606e-07,
+      "loss": 0.0171,
+      "step": 1194
+    },
+    {
+      "epoch": 11.601941747572816,
+      "grad_norm": 0.27943211793899536,
+      "learning_rate": 6.068481868456558e-07,
+      "loss": 0.0107,
+      "step": 1195
+    },
+    {
+      "epoch": 11.611650485436893,
+      "grad_norm": 0.17899493873119354,
+      "learning_rate": 6.035320154715549e-07,
+      "loss": 0.0048,
+      "step": 1196
+    },
+    {
+      "epoch": 11.62135922330097,
+      "grad_norm": 0.17442212998867035,
+      "learning_rate": 6.00223685398788e-07,
+      "loss": 0.0036,
+      "step": 1197
+    },
+    {
+      "epoch": 11.631067961165048,
+      "grad_norm": 0.47512590885162354,
+      "learning_rate": 5.969232103062647e-07,
+      "loss": 0.0191,
+      "step": 1198
+    },
+    {
+      "epoch": 11.640776699029127,
+      "grad_norm": 0.12684416770935059,
+      "learning_rate": 5.936306038404158e-07,
+      "loss": 0.0005,
+      "step": 1199
+    },
+    {
+      "epoch": 11.650485436893204,
+      "grad_norm": 0.7837037444114685,
+      "learning_rate": 5.903458796151382e-07,
+      "loss": 0.046,
+      "step": 1200
+    },
+    {
+      "epoch": 11.660194174757281,
+      "grad_norm": 0.18797621130943298,
+      "learning_rate": 5.870690512117377e-07,
+      "loss": 0.004,
+      "step": 1201
+    },
+    {
+      "epoch": 11.669902912621358,
+      "grad_norm": 0.367638498544693,
+      "learning_rate": 5.838001321788744e-07,
+      "loss": 0.0032,
+      "step": 1202
+    },
+    {
+      "epoch": 11.679611650485437,
+      "grad_norm": 0.3879675269126892,
+      "learning_rate": 5.80539136032505e-07,
+      "loss": 0.0078,
+      "step": 1203
+    },
+    {
+      "epoch": 11.689320388349515,
+      "grad_norm": 0.648831307888031,
+      "learning_rate": 5.772860762558269e-07,
+      "loss": 0.0085,
+      "step": 1204
+    },
+    {
+      "epoch": 11.699029126213592,
+      "grad_norm": 0.20488865673542023,
+      "learning_rate": 5.740409662992244e-07,
+      "loss": 0.0043,
+      "step": 1205
+    },
+    {
+      "epoch": 11.70873786407767,
+      "grad_norm": 0.39734843373298645,
+      "learning_rate": 5.708038195802098e-07,
+      "loss": 0.0017,
+      "step": 1206
+    },
+    {
+      "epoch": 11.718446601941748,
+      "grad_norm": 0.42358076572418213,
+      "learning_rate": 5.675746494833733e-07,
+      "loss": 0.0043,
+      "step": 1207
+    },
+    {
+      "epoch": 11.728155339805825,
+      "grad_norm": 0.15531939268112183,
+      "learning_rate": 5.643534693603214e-07,
+      "loss": 0.0019,
+      "step": 1208
+    },
+    {
+      "epoch": 11.737864077669903,
+      "grad_norm": 0.4669681489467621,
+      "learning_rate": 5.61140292529625e-07,
+      "loss": 0.0082,
+      "step": 1209
+    },
+    {
+      "epoch": 11.74757281553398,
+      "grad_norm": 0.3049868047237396,
+      "learning_rate": 5.579351322767643e-07,
+      "loss": 0.0045,
+      "step": 1210
+    },
+    {
+      "epoch": 11.757281553398059,
+      "grad_norm": 0.4822055399417877,
+      "learning_rate": 5.547380018540735e-07,
+      "loss": 0.0268,
+      "step": 1211
+    },
+    {
+      "epoch": 11.766990291262136,
+      "grad_norm": 0.1961973011493683,
+      "learning_rate": 5.515489144806862e-07,
+      "loss": 0.0029,
+      "step": 1212
+    },
+    {
+      "epoch": 11.776699029126213,
+      "grad_norm": 0.39638909697532654,
+      "learning_rate": 5.483678833424796e-07,
+      "loss": 0.0104,
+      "step": 1213
+    },
+    {
+      "epoch": 11.78640776699029,
+      "grad_norm": 0.2661190330982208,
+      "learning_rate": 5.451949215920221e-07,
+      "loss": 0.0062,
+      "step": 1214
+    },
+    {
+      "epoch": 11.79611650485437,
+      "grad_norm": 0.3237748444080353,
+      "learning_rate": 5.420300423485167e-07,
+      "loss": 0.0022,
+      "step": 1215
+    },
+    {
+      "epoch": 11.805825242718447,
+      "grad_norm": 0.37730419635772705,
+      "learning_rate": 5.38873258697748e-07,
+      "loss": 0.003,
+      "step": 1216
+    },
+    {
+      "epoch": 11.815533980582524,
+      "grad_norm": 0.4794592559337616,
+      "learning_rate": 5.357245836920286e-07,
+      "loss": 0.0081,
+      "step": 1217
+    },
+    {
+      "epoch": 11.825242718446601,
+      "grad_norm": 0.3878485858440399,
+      "learning_rate": 5.325840303501431e-07,
+      "loss": 0.0062,
+      "step": 1218
+    },
+    {
+      "epoch": 11.83495145631068,
+      "grad_norm": 0.3394526243209839,
+      "learning_rate": 5.29451611657297e-07,
+      "loss": 0.0075,
+      "step": 1219
+    },
+    {
+      "epoch": 11.844660194174757,
+      "grad_norm": 0.46053069829940796,
+      "learning_rate": 5.263273405650601e-07,
+      "loss": 0.0339,
+      "step": 1220
+    },
+    {
+      "epoch": 11.854368932038835,
+      "grad_norm": 0.3216699957847595,
+      "learning_rate": 5.232112299913151e-07,
+      "loss": 0.0056,
+      "step": 1221
+    },
+    {
+      "epoch": 11.864077669902912,
+      "grad_norm": 0.24092155694961548,
+      "learning_rate": 5.201032928202043e-07,
+      "loss": 0.0095,
+      "step": 1222
+    },
+    {
+      "epoch": 11.87378640776699,
+      "grad_norm": 0.9538231492042542,
+      "learning_rate": 5.17003541902075e-07,
+      "loss": 0.0429,
+      "step": 1223
+    },
+    {
+      "epoch": 11.883495145631068,
+      "grad_norm": 0.4639842212200165,
+      "learning_rate": 5.139119900534259e-07,
+      "loss": 0.021,
+      "step": 1224
+    },
+    {
+      "epoch": 11.893203883495145,
+      "grad_norm": 0.3017670810222626,
+      "learning_rate": 5.108286500568562e-07,
+      "loss": 0.0019,
+      "step": 1225
+    },
+    {
+      "epoch": 11.902912621359224,
+      "grad_norm": 0.4440242350101471,
+      "learning_rate": 5.077535346610115e-07,
+      "loss": 0.0061,
+      "step": 1226
+    },
+    {
+      "epoch": 11.912621359223301,
+      "grad_norm": 0.16742651164531708,
+      "learning_rate": 5.046866565805311e-07,
+      "loss": 0.0014,
+      "step": 1227
+    },
+    {
+      "epoch": 11.922330097087379,
+      "grad_norm": 0.22309471666812897,
+      "learning_rate": 5.016280284959957e-07,
+      "loss": 0.0027,
+      "step": 1228
+    },
+    {
+      "epoch": 11.932038834951456,
+      "grad_norm": 0.39898568391799927,
+      "learning_rate": 4.985776630538746e-07,
+      "loss": 0.008,
+      "step": 1229
+    },
+    {
+      "epoch": 11.941747572815533,
+      "grad_norm": 0.2526729702949524,
+      "learning_rate": 4.95535572866474e-07,
+      "loss": 0.0013,
+      "step": 1230
+    },
+    {
+      "epoch": 11.951456310679612,
+      "grad_norm": 0.1417032927274704,
+      "learning_rate": 4.925017705118843e-07,
+      "loss": 0.0014,
+      "step": 1231
+    },
+    {
+      "epoch": 11.96116504854369,
+      "grad_norm": 0.39468225836753845,
+      "learning_rate": 4.89476268533928e-07,
+      "loss": 0.0049,
+      "step": 1232
+    },
+    {
+      "epoch": 11.970873786407767,
+      "grad_norm": 0.33007320761680603,
+      "learning_rate": 4.864590794421092e-07,
+      "loss": 0.0044,
+      "step": 1233
+    },
+    {
+      "epoch": 11.980582524271846,
+      "grad_norm": 0.4248305857181549,
+      "learning_rate": 4.834502157115597e-07,
+      "loss": 0.0226,
+      "step": 1234
+    },
+    {
+      "epoch": 11.990291262135923,
+      "grad_norm": 0.09050054848194122,
+      "learning_rate": 4.804496897829883e-07,
+      "loss": 0.0009,
+      "step": 1235
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.3121471107006073,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.0067,
+      "step": 1236
+    },
+    {
+      "epoch": 12.009708737864077,
+      "grad_norm": 0.20965978503227234,
+      "learning_rate": 4.744737009221986e-07,
+      "loss": 0.0045,
+      "step": 1237
+    },
+    {
+      "epoch": 12.019417475728156,
+      "grad_norm": 0.2554788887500763,
+      "learning_rate": 4.7149826269882294e-07,
+      "loss": 0.0044,
+      "step": 1238
+    },
+    {
+      "epoch": 12.029126213592233,
+      "grad_norm": 0.22696569561958313,
+      "learning_rate": 4.6853121169500914e-07,
+      "loss": 0.0076,
+      "step": 1239
+    },
+    {
+      "epoch": 12.03883495145631,
+      "grad_norm": 0.2505325973033905,
+      "learning_rate": 4.6557256017858485e-07,
+      "loss": 0.0019,
+      "step": 1240
+    },
+    {
+      "epoch": 12.048543689320388,
+      "grad_norm": 0.13231754302978516,
+      "learning_rate": 4.626223203826477e-07,
+      "loss": 0.0019,
+      "step": 1241
+    },
+    {
+      "epoch": 12.058252427184467,
+      "grad_norm": 0.15132735669612885,
+      "learning_rate": 4.5968050450551527e-07,
+      "loss": 0.0017,
+      "step": 1242
+    },
+    {
+      "epoch": 12.067961165048544,
+      "grad_norm": 0.16848385334014893,
+      "learning_rate": 4.56747124710675e-07,
+      "loss": 0.0042,
+      "step": 1243
+    },
+    {
+      "epoch": 12.077669902912621,
+      "grad_norm": 0.2351788580417633,
+      "learning_rate": 4.5382219312673364e-07,
+      "loss": 0.0087,
+      "step": 1244
+    },
+    {
+      "epoch": 12.087378640776699,
+      "grad_norm": 0.08185896277427673,
+      "learning_rate": 4.5090572184736863e-07,
+      "loss": 0.0013,
+      "step": 1245
+    },
+    {
+      "epoch": 12.097087378640778,
+      "grad_norm": 0.33229878544807434,
+      "learning_rate": 4.4799772293127486e-07,
+      "loss": 0.0087,
+      "step": 1246
+    },
+    {
+      "epoch": 12.106796116504855,
+      "grad_norm": 0.5459170937538147,
+      "learning_rate": 4.4509820840211745e-07,
+      "loss": 0.0075,
+      "step": 1247
+    },
+    {
+      "epoch": 12.116504854368932,
+      "grad_norm": 0.22844095528125763,
+      "learning_rate": 4.422071902484812e-07,
+      "loss": 0.0059,
+      "step": 1248
+    },
+    {
+      "epoch": 12.12621359223301,
+      "grad_norm": 0.1771807223558426,
+      "learning_rate": 4.3932468042382075e-07,
+      "loss": 0.0051,
+      "step": 1249
+    },
+    {
+      "epoch": 12.135922330097088,
+      "grad_norm": 0.3462653160095215,
+      "learning_rate": 4.3645069084641195e-07,
+      "loss": 0.0053,
+      "step": 1250
+    },
+    {
+      "epoch": 12.145631067961165,
+      "grad_norm": 0.23488548398017883,
+      "learning_rate": 4.335852333993018e-07,
+      "loss": 0.0107,
+      "step": 1251
+    },
+    {
+      "epoch": 12.155339805825243,
+      "grad_norm": 0.20178791880607605,
+      "learning_rate": 4.3072831993025895e-07,
+      "loss": 0.0024,
+      "step": 1252
+    },
+    {
+      "epoch": 12.16504854368932,
+      "grad_norm": 0.28149035573005676,
+      "learning_rate": 4.278799622517274e-07,
+      "loss": 0.0042,
+      "step": 1253
+    },
+    {
+      "epoch": 12.174757281553399,
+      "grad_norm": 0.22097015380859375,
+      "learning_rate": 4.2504017214077374e-07,
+      "loss": 0.0053,
+      "step": 1254
+    },
+    {
+      "epoch": 12.184466019417476,
+      "grad_norm": 0.3525868356227875,
+      "learning_rate": 4.222089613390412e-07,
+      "loss": 0.0009,
+      "step": 1255
+    },
+    {
+      "epoch": 12.194174757281553,
+      "grad_norm": 0.2622714936733246,
+      "learning_rate": 4.1938634155269944e-07,
+      "loss": 0.0049,
+      "step": 1256
+    },
+    {
+      "epoch": 12.20388349514563,
+      "grad_norm": 0.19608217477798462,
+      "learning_rate": 4.165723244523978e-07,
+      "loss": 0.0071,
+      "step": 1257
+    },
+    {
+      "epoch": 12.21359223300971,
+      "grad_norm": 0.22802308201789856,
+      "learning_rate": 4.1376692167321626e-07,
+      "loss": 0.0059,
+      "step": 1258
+    },
+    {
+      "epoch": 12.223300970873787,
+      "grad_norm": 0.20340679585933685,
+      "learning_rate": 4.109701448146164e-07,
+      "loss": 0.0042,
+      "step": 1259
+    },
+    {
+      "epoch": 12.233009708737864,
+      "grad_norm": 0.22457940876483917,
+      "learning_rate": 4.0818200544039484e-07,
+      "loss": 0.0045,
+      "step": 1260
+    },
+    {
+      "epoch": 12.242718446601941,
+      "grad_norm": 0.09476789087057114,
+      "learning_rate": 4.054025150786356e-07,
+      "loss": 0.0009,
+      "step": 1261
+    },
+    {
+      "epoch": 12.25242718446602,
+      "grad_norm": 0.16748803853988647,
+      "learning_rate": 4.026316852216605e-07,
+      "loss": 0.0043,
+      "step": 1262
+    },
+    {
+      "epoch": 12.262135922330097,
+      "grad_norm": 0.1957675963640213,
+      "learning_rate": 3.998695273259834e-07,
+      "loss": 0.004,
+      "step": 1263
+    },
+    {
+      "epoch": 12.271844660194175,
+      "grad_norm": 0.35359010100364685,
+      "learning_rate": 3.971160528122622e-07,
+      "loss": 0.0053,
+      "step": 1264
+    },
+    {
+      "epoch": 12.281553398058252,
+      "grad_norm": 0.36833491921424866,
+      "learning_rate": 3.9437127306525295e-07,
+      "loss": 0.0243,
+      "step": 1265
+    },
+    {
+      "epoch": 12.29126213592233,
+      "grad_norm": 0.13272319734096527,
+      "learning_rate": 3.9163519943375973e-07,
+      "loss": 0.003,
+      "step": 1266
+    },
+    {
+      "epoch": 12.300970873786408,
+      "grad_norm": 0.4329506456851959,
+      "learning_rate": 3.889078432305904e-07,
+      "loss": 0.0333,
+      "step": 1267
+    },
+    {
+      "epoch": 12.310679611650485,
+      "grad_norm": 0.2429269254207611,
+      "learning_rate": 3.8618921573250896e-07,
+      "loss": 0.0083,
+      "step": 1268
+    },
+    {
+      "epoch": 12.320388349514563,
+      "grad_norm": 0.11397936940193176,
+      "learning_rate": 3.834793281801891e-07,
+      "loss": 0.0017,
+      "step": 1269
+    },
+    {
+      "epoch": 12.330097087378642,
+      "grad_norm": 0.6022169589996338,
+      "learning_rate": 3.8077819177816695e-07,
+      "loss": 0.0253,
+      "step": 1270
+    },
+    {
+      "epoch": 12.339805825242719,
+      "grad_norm": 0.1681753695011139,
+      "learning_rate": 3.780858176947963e-07,
+      "loss": 0.0033,
+      "step": 1271
+    },
+    {
+      "epoch": 12.349514563106796,
+      "grad_norm": 0.1316511482000351,
+      "learning_rate": 3.754022170622007e-07,
+      "loss": 0.0007,
+      "step": 1272
+    },
+    {
+      "epoch": 12.359223300970873,
+      "grad_norm": 0.33542078733444214,
+      "learning_rate": 3.7272740097622884e-07,
+      "loss": 0.0119,
+      "step": 1273
+    },
+    {
+      "epoch": 12.368932038834952,
+      "grad_norm": 0.38612493872642517,
+      "learning_rate": 3.700613804964073e-07,
+      "loss": 0.0106,
+      "step": 1274
+    },
+    {
+      "epoch": 12.37864077669903,
+      "grad_norm": 0.5517392754554749,
+      "learning_rate": 3.6740416664589634e-07,
+      "loss": 0.0111,
+      "step": 1275
+    },
+    {
+      "epoch": 12.388349514563107,
+      "grad_norm": 0.3845731317996979,
+      "learning_rate": 3.6475577041144324e-07,
+      "loss": 0.0117,
+      "step": 1276
+    },
+    {
+      "epoch": 12.398058252427184,
+      "grad_norm": 0.04387803003191948,
+      "learning_rate": 3.6211620274333727e-07,
+      "loss": 0.0004,
+      "step": 1277
+    },
+    {
+      "epoch": 12.407766990291263,
+      "grad_norm": 0.5418988466262817,
+      "learning_rate": 3.594854745553636e-07,
+      "loss": 0.0065,
+      "step": 1278
+    },
+    {
+      "epoch": 12.41747572815534,
+      "grad_norm": 0.5132270455360413,
+      "learning_rate": 3.568635967247605e-07,
+      "loss": 0.0273,
+      "step": 1279
+    },
+    {
+      "epoch": 12.427184466019417,
+      "grad_norm": 0.11886625736951828,
+      "learning_rate": 3.5425058009217193e-07,
+      "loss": 0.0019,
+      "step": 1280
+    },
+    {
+      "epoch": 12.436893203883495,
+      "grad_norm": 0.19105517864227295,
+      "learning_rate": 3.516464354616031e-07,
+      "loss": 0.0058,
+      "step": 1281
+    },
+    {
+      "epoch": 12.446601941747574,
+      "grad_norm": 0.457914263010025,
+      "learning_rate": 3.4905117360037683e-07,
+      "loss": 0.0115,
+      "step": 1282
+    },
+    {
+      "epoch": 12.45631067961165,
+      "grad_norm": 0.20160461962223053,
+      "learning_rate": 3.4646480523908813e-07,
+      "loss": 0.002,
+      "step": 1283
+    },
+    {
+      "epoch": 12.466019417475728,
+      "grad_norm": 0.4353931248188019,
+      "learning_rate": 3.43887341071561e-07,
+      "loss": 0.039,
+      "step": 1284
+    },
+    {
+      "epoch": 12.475728155339805,
+      "grad_norm": 0.07777180522680283,
+      "learning_rate": 3.413187917548019e-07,
+      "loss": 0.0008,
+      "step": 1285
+    },
+    {
+      "epoch": 12.485436893203884,
+      "grad_norm": 0.7818756699562073,
+      "learning_rate": 3.3875916790895883e-07,
+      "loss": 0.0242,
+      "step": 1286
+    },
+    {
+      "epoch": 12.495145631067961,
+      "grad_norm": 0.10199937224388123,
+      "learning_rate": 3.3620848011727437e-07,
+      "loss": 0.0009,
+      "step": 1287
+    },
+    {
+      "epoch": 12.504854368932039,
+      "grad_norm": 0.06977365165948868,
+      "learning_rate": 3.336667389260445e-07,
+      "loss": 0.0007,
+      "step": 1288
+    },
+    {
+      "epoch": 12.514563106796116,
+      "grad_norm": 0.11607404798269272,
+      "learning_rate": 3.311339548445727e-07,
+      "loss": 0.001,
+      "step": 1289
+    },
+    {
+      "epoch": 12.524271844660195,
+      "grad_norm": 0.18200449645519257,
+      "learning_rate": 3.2861013834512844e-07,
+      "loss": 0.0034,
+      "step": 1290
+    },
+    {
+      "epoch": 12.533980582524272,
+      "grad_norm": 0.34803321957588196,
+      "learning_rate": 3.2609529986290246e-07,
+      "loss": 0.0148,
+      "step": 1291
+    },
+    {
+      "epoch": 12.54368932038835,
+      "grad_norm": 0.20186111330986023,
+      "learning_rate": 3.235894497959649e-07,
+      "loss": 0.002,
+      "step": 1292
+    },
+    {
+      "epoch": 12.553398058252426,
+      "grad_norm": 0.3056010603904724,
+      "learning_rate": 3.2109259850522045e-07,
+      "loss": 0.0032,
+      "step": 1293
+    },
+    {
+      "epoch": 12.563106796116505,
+      "grad_norm": 0.35836365818977356,
+      "learning_rate": 3.186047563143685e-07,
+      "loss": 0.0064,
+      "step": 1294
+    },
+    {
+      "epoch": 12.572815533980583,
+      "grad_norm": 0.17417111992835999,
+      "learning_rate": 3.161259335098571e-07,
+      "loss": 0.0023,
+      "step": 1295
+    },
+    {
+      "epoch": 12.58252427184466,
+      "grad_norm": 0.06176523491740227,
+      "learning_rate": 3.1365614034084224e-07,
+      "loss": 0.0007,
+      "step": 1296
+    },
+    {
+      "epoch": 12.592233009708737,
+      "grad_norm": 0.1570029854774475,
+      "learning_rate": 3.111953870191459e-07,
+      "loss": 0.0025,
+      "step": 1297
+    },
+    {
+      "epoch": 12.601941747572816,
+      "grad_norm": 0.40762364864349365,
+      "learning_rate": 3.087436837192118e-07,
+      "loss": 0.0094,
+      "step": 1298
+    },
+    {
+      "epoch": 12.611650485436893,
+      "grad_norm": 0.6000169515609741,
+      "learning_rate": 3.0630104057806616e-07,
+      "loss": 0.0307,
+      "step": 1299
+    },
+    {
+      "epoch": 12.62135922330097,
+      "grad_norm": 0.22923287749290466,
+      "learning_rate": 3.0386746769527323e-07,
+      "loss": 0.0072,
+      "step": 1300
+    },
+    {
+      "epoch": 12.631067961165048,
+      "grad_norm": 0.09936048835515976,
+      "learning_rate": 3.0144297513289483e-07,
+      "loss": 0.0013,
+      "step": 1301
+    },
+    {
+      "epoch": 12.640776699029127,
+      "grad_norm": 0.17220628261566162,
+      "learning_rate": 2.9902757291544905e-07,
+      "loss": 0.0019,
+      "step": 1302
+    },
+    {
+      "epoch": 12.650485436893204,
+      "grad_norm": 0.0635104775428772,
+      "learning_rate": 2.966212710298674e-07,
+      "loss": 0.0006,
+      "step": 1303
+    },
+    {
+      "epoch": 12.660194174757281,
+      "grad_norm": 0.25115472078323364,
+      "learning_rate": 2.94224079425455e-07,
+      "loss": 0.0026,
+      "step": 1304
+    },
+    {
+      "epoch": 12.669902912621358,
+      "grad_norm": 0.4713001847267151,
+      "learning_rate": 2.9183600801384853e-07,
+      "loss": 0.0014,
+      "step": 1305
+    },
+    {
+      "epoch": 12.679611650485437,
+      "grad_norm": 0.0625031441450119,
+      "learning_rate": 2.8945706666897555e-07,
+      "loss": 0.0005,
+      "step": 1306
+    },
+    {
+      "epoch": 12.689320388349515,
+      "grad_norm": 0.16515658795833588,
+      "learning_rate": 2.870872652270129e-07,
+      "loss": 0.0047,
+      "step": 1307
+    },
+    {
+      "epoch": 12.699029126213592,
+      "grad_norm": 0.08110034465789795,
+      "learning_rate": 2.8472661348634883e-07,
+      "loss": 0.0011,
+      "step": 1308
+    },
+    {
+      "epoch": 12.70873786407767,
+      "grad_norm": 0.17738069593906403,
+      "learning_rate": 2.82375121207539e-07,
+      "loss": 0.0053,
+      "step": 1309
+    },
+    {
+      "epoch": 12.718446601941748,
+      "grad_norm": 0.18028077483177185,
+      "learning_rate": 2.8003279811326724e-07,
+      "loss": 0.0046,
+      "step": 1310
+    },
+    {
+      "epoch": 12.728155339805825,
+      "grad_norm": 0.07242338359355927,
+      "learning_rate": 2.776996538883062e-07,
+      "loss": 0.0005,
+      "step": 1311
+    },
+    {
+      "epoch": 12.737864077669903,
+      "grad_norm": 0.2609570026397705,
+      "learning_rate": 2.7537569817947694e-07,
+      "loss": 0.006,
+      "step": 1312
+    },
+    {
+      "epoch": 12.74757281553398,
+      "grad_norm": 0.19686202704906464,
+      "learning_rate": 2.730609405956083e-07,
+      "loss": 0.0068,
+      "step": 1313
+    },
+    {
+      "epoch": 12.757281553398059,
+      "grad_norm": 0.20468905568122864,
+      "learning_rate": 2.707553907074989e-07,
+      "loss": 0.0063,
+      "step": 1314
+    },
+    {
+      "epoch": 12.766990291262136,
+      "grad_norm": 0.3390023708343506,
+      "learning_rate": 2.684590580478749e-07,
+      "loss": 0.0037,
+      "step": 1315
+    },
+    {
+      "epoch": 12.776699029126213,
+      "grad_norm": 0.5263711214065552,
+      "learning_rate": 2.6617195211135343e-07,
+      "loss": 0.0063,
+      "step": 1316
+    },
+    {
+      "epoch": 12.78640776699029,
+      "grad_norm": 0.23772577941417694,
+      "learning_rate": 2.638940823544012e-07,
+      "loss": 0.0053,
+      "step": 1317
+    },
+    {
+      "epoch": 12.79611650485437,
+      "grad_norm": 0.12934847176074982,
+      "learning_rate": 2.6162545819529624e-07,
+      "loss": 0.0027,
+      "step": 1318
+    },
+    {
+      "epoch": 12.805825242718447,
+      "grad_norm": 0.19521230459213257,
+      "learning_rate": 2.593660890140895e-07,
+      "loss": 0.0051,
+      "step": 1319
+    },
+    {
+      "epoch": 12.815533980582524,
+      "grad_norm": 0.2085026204586029,
+      "learning_rate": 2.57115984152565e-07,
+      "loss": 0.0051,
+      "step": 1320
+    },
+    {
+      "epoch": 12.825242718446601,
+      "grad_norm": 0.4785703122615814,
+      "learning_rate": 2.548751529142018e-07,
+      "loss": 0.0142,
+      "step": 1321
+    },
+    {
+      "epoch": 12.83495145631068,
+      "grad_norm": 0.47134339809417725,
+      "learning_rate": 2.526436045641351e-07,
+      "loss": 0.0128,
+      "step": 1322
+    },
+    {
+      "epoch": 12.844660194174757,
+      "grad_norm": 0.15493355691432953,
+      "learning_rate": 2.504213483291193e-07,
+      "loss": 0.0024,
+      "step": 1323
+    },
+    {
+      "epoch": 12.854368932038835,
+      "grad_norm": 0.06153065711259842,
+      "learning_rate": 2.482083933974883e-07,
+      "loss": 0.0007,
+      "step": 1324
+    },
+    {
+      "epoch": 12.864077669902912,
+      "grad_norm": 0.2635805606842041,
+      "learning_rate": 2.4600474891911696e-07,
+      "loss": 0.0055,
+      "step": 1325
+    },
+    {
+      "epoch": 12.87378640776699,
+      "grad_norm": 0.16463325917720795,
+      "learning_rate": 2.43810424005386e-07,
+      "loss": 0.0043,
+      "step": 1326
+    },
+    {
+      "epoch": 12.883495145631068,
+      "grad_norm": 0.2017255276441574,
+      "learning_rate": 2.416254277291416e-07,
+      "loss": 0.0071,
+      "step": 1327
+    },
+    {
+      "epoch": 12.893203883495145,
+      "grad_norm": 0.039516631513834,
+      "learning_rate": 2.3944976912465916e-07,
+      "loss": 0.0004,
+      "step": 1328
+    },
+    {
+      "epoch": 12.902912621359224,
+      "grad_norm": 0.19528639316558838,
+      "learning_rate": 2.3728345718760622e-07,
+      "loss": 0.0057,
+      "step": 1329
+    },
+    {
+      "epoch": 12.912621359223301,
+      "grad_norm": 0.38023102283477783,
+      "learning_rate": 2.3512650087500338e-07,
+      "loss": 0.018,
+      "step": 1330
+    },
+    {
+      "epoch": 12.922330097087379,
+      "grad_norm": 0.13987840712070465,
+      "learning_rate": 2.3297890910519093e-07,
+      "loss": 0.0019,
+      "step": 1331
+    },
+    {
+      "epoch": 12.932038834951456,
+      "grad_norm": 0.05007342994213104,
+      "learning_rate": 2.3084069075778758e-07,
+      "loss": 0.0004,
+      "step": 1332
+    },
+    {
+      "epoch": 12.941747572815533,
+      "grad_norm": 0.19607791304588318,
+      "learning_rate": 2.287118546736572e-07,
+      "loss": 0.0036,
+      "step": 1333
+    },
+    {
+      "epoch": 12.951456310679612,
+      "grad_norm": 0.4152860641479492,
+      "learning_rate": 2.2659240965487023e-07,
+      "loss": 0.005,
+      "step": 1334
+    },
+    {
+      "epoch": 12.96116504854369,
+      "grad_norm": 0.31173020601272583,
+      "learning_rate": 2.2448236446466847e-07,
+      "loss": 0.0028,
+      "step": 1335
+    },
+    {
+      "epoch": 12.970873786407767,
+      "grad_norm": 0.31397274136543274,
+      "learning_rate": 2.2238172782742763e-07,
+      "loss": 0.014,
+      "step": 1336
+    },
+    {
+      "epoch": 12.980582524271846,
+      "grad_norm": 0.06522321701049805,
+      "learning_rate": 2.2029050842862277e-07,
+      "loss": 0.0008,
+      "step": 1337
+    },
+    {
+      "epoch": 12.990291262135923,
+      "grad_norm": 0.15653946995735168,
+      "learning_rate": 2.1820871491479102e-07,
+      "loss": 0.0027,
+      "step": 1338
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.2695692479610443,
+      "learning_rate": 2.1613635589349756e-07,
+      "loss": 0.0042,
+      "step": 1339
+    },
+    {
+      "epoch": 13.009708737864077,
+      "grad_norm": 0.3138914108276367,
+      "learning_rate": 2.140734399332975e-07,
+      "loss": 0.0115,
+      "step": 1340
+    },
+    {
+      "epoch": 13.019417475728156,
+      "grad_norm": 0.2193198800086975,
+      "learning_rate": 2.1201997556370284e-07,
+      "loss": 0.0048,
+      "step": 1341
+    },
+    {
+      "epoch": 13.029126213592233,
+      "grad_norm": 0.08152951300144196,
+      "learning_rate": 2.0997597127514507e-07,
+      "loss": 0.0012,
+      "step": 1342
+    },
+    {
+      "epoch": 13.03883495145631,
+      "grad_norm": 0.37986671924591064,
+      "learning_rate": 2.079414355189427e-07,
+      "loss": 0.007,
+      "step": 1343
+    },
+    {
+      "epoch": 13.048543689320388,
+      "grad_norm": 0.15214113891124725,
+      "learning_rate": 2.059163767072639e-07,
+      "loss": 0.0032,
+      "step": 1344
+    },
+    {
+      "epoch": 13.058252427184467,
+      "grad_norm": 0.09272062033414841,
+      "learning_rate": 2.0390080321309236e-07,
+      "loss": 0.0013,
+      "step": 1345
+    },
+    {
+      "epoch": 13.067961165048544,
+      "grad_norm": 0.02976040169596672,
+      "learning_rate": 2.01894723370194e-07,
+      "loss": 0.0003,
+      "step": 1346
+    },
+    {
+      "epoch": 13.077669902912621,
+      "grad_norm": 0.12307222932577133,
+      "learning_rate": 1.9989814547308056e-07,
+      "loss": 0.0024,
+      "step": 1347
+    },
+    {
+      "epoch": 13.087378640776699,
+      "grad_norm": 0.3986649513244629,
+      "learning_rate": 1.9791107777697633e-07,
+      "loss": 0.0171,
+      "step": 1348
+    },
+    {
+      "epoch": 13.097087378640778,
+      "grad_norm": 0.11371457576751709,
+      "learning_rate": 1.9593352849778453e-07,
+      "loss": 0.0014,
+      "step": 1349
+    },
+    {
+      "epoch": 13.106796116504855,
+      "grad_norm": 0.08083812892436981,
+      "learning_rate": 1.9396550581205208e-07,
+      "loss": 0.0011,
+      "step": 1350
+    },
+    {
+      "epoch": 13.116504854368932,
+      "grad_norm": 0.19208478927612305,
+      "learning_rate": 1.920070178569361e-07,
+      "loss": 0.0028,
+      "step": 1351
+    },
+    {
+      "epoch": 13.12621359223301,
+      "grad_norm": 0.23043976724147797,
+      "learning_rate": 1.900580727301718e-07,
+      "loss": 0.0024,
+      "step": 1352
+    },
+    {
+      "epoch": 13.135922330097088,
+      "grad_norm": 0.1643299162387848,
+      "learning_rate": 1.8811867849003684e-07,
+      "loss": 0.0046,
+      "step": 1353
+    },
+    {
+      "epoch": 13.145631067961165,
+      "grad_norm": 0.18977408111095428,
+      "learning_rate": 1.8618884315531939e-07,
+      "loss": 0.0044,
+      "step": 1354
+    },
+    {
+      "epoch": 13.155339805825243,
+      "grad_norm": 0.39201587438583374,
+      "learning_rate": 1.8426857470528414e-07,
+      "loss": 0.025,
+      "step": 1355
+    },
+    {
+      "epoch": 13.16504854368932,
+      "grad_norm": 0.41091713309288025,
+      "learning_rate": 1.8235788107963948e-07,
+      "loss": 0.0184,
+      "step": 1356
+    },
+    {
+      "epoch": 13.174757281553399,
+      "grad_norm": 0.04530488699674606,
+      "learning_rate": 1.8045677017850595e-07,
+      "loss": 0.0004,
+      "step": 1357
+    },
+    {
+      "epoch": 13.184466019417476,
+      "grad_norm": 0.1892017126083374,
+      "learning_rate": 1.785652498623816e-07,
+      "loss": 0.006,
+      "step": 1358
+    },
+    {
+      "epoch": 13.194174757281553,
+      "grad_norm": 0.26920121908187866,
+      "learning_rate": 1.7668332795211074e-07,
+      "loss": 0.0055,
+      "step": 1359
+    },
+    {
+      "epoch": 13.20388349514563,
+      "grad_norm": 0.08278663456439972,
+      "learning_rate": 1.7481101222885126e-07,
+      "loss": 0.0012,
+      "step": 1360
+    },
+    {
+      "epoch": 13.21359223300971,
+      "grad_norm": 0.17943482100963593,
+      "learning_rate": 1.7294831043404264e-07,
+      "loss": 0.0046,
+      "step": 1361
+    },
+    {
+      "epoch": 13.223300970873787,
+      "grad_norm": 0.28155428171157837,
+      "learning_rate": 1.7109523026937302e-07,
+      "loss": 0.0108,
+      "step": 1362
+    },
+    {
+      "epoch": 13.233009708737864,
+      "grad_norm": 0.14628250896930695,
+      "learning_rate": 1.6925177939674936e-07,
+      "loss": 0.0042,
+      "step": 1363
+    },
+    {
+      "epoch": 13.242718446601941,
+      "grad_norm": 0.21149447560310364,
+      "learning_rate": 1.6741796543826321e-07,
+      "loss": 0.0083,
+      "step": 1364
+    },
+    {
+      "epoch": 13.25242718446602,
+      "grad_norm": 0.17508278787136078,
+      "learning_rate": 1.6559379597616136e-07,
+      "loss": 0.0041,
+      "step": 1365
+    },
+    {
+      "epoch": 13.262135922330097,
+      "grad_norm": 0.2160256952047348,
+      "learning_rate": 1.6377927855281362e-07,
+      "loss": 0.0049,
+      "step": 1366
+    },
+    {
+      "epoch": 13.271844660194175,
+      "grad_norm": 0.7788318991661072,
+      "learning_rate": 1.6197442067068136e-07,
+      "loss": 0.0106,
+      "step": 1367
+    },
+    {
+      "epoch": 13.281553398058252,
+      "grad_norm": 0.1593695729970932,
+      "learning_rate": 1.6017922979228662e-07,
+      "loss": 0.0042,
+      "step": 1368
+    },
+    {
+      "epoch": 13.29126213592233,
+      "grad_norm": 0.13514795899391174,
+      "learning_rate": 1.5839371334018193e-07,
+      "loss": 0.0026,
+      "step": 1369
+    },
+    {
+      "epoch": 13.300970873786408,
+      "grad_norm": 0.19250337779521942,
+      "learning_rate": 1.5661787869691858e-07,
+      "loss": 0.0043,
+      "step": 1370
+    },
+    {
+      "epoch": 13.310679611650485,
+      "grad_norm": 0.18198005855083466,
+      "learning_rate": 1.5485173320501673e-07,
+      "loss": 0.0037,
+      "step": 1371
+    },
+    {
+      "epoch": 13.320388349514563,
+      "grad_norm": 0.18506045639514923,
+      "learning_rate": 1.5309528416693503e-07,
+      "loss": 0.0039,
+      "step": 1372
+    },
+    {
+      "epoch": 13.330097087378642,
+      "grad_norm": 0.09922155737876892,
+      "learning_rate": 1.513485388450403e-07,
+      "loss": 0.0013,
+      "step": 1373
+    },
+    {
+      "epoch": 13.339805825242719,
+      "grad_norm": 0.4548637270927429,
+      "learning_rate": 1.4961150446157759e-07,
+      "loss": 0.0123,
+      "step": 1374
+    },
+    {
+      "epoch": 13.349514563106796,
+      "grad_norm": 0.1478271484375,
+      "learning_rate": 1.4788418819864037e-07,
+      "loss": 0.0025,
+      "step": 1375
+    },
+    {
+      "epoch": 13.359223300970873,
+      "grad_norm": 0.2919103801250458,
+      "learning_rate": 1.461665971981402e-07,
+      "loss": 0.0069,
+      "step": 1376
+    },
+    {
+      "epoch": 13.368932038834952,
+      "grad_norm": 0.0701352208852768,
+      "learning_rate": 1.444587385617785e-07,
+      "loss": 0.0008,
+      "step": 1377
+    },
+    {
+      "epoch": 13.37864077669903,
+      "grad_norm": 0.15089794993400574,
+      "learning_rate": 1.4276061935101586e-07,
+      "loss": 0.0031,
+      "step": 1378
+    },
+    {
+      "epoch": 13.388349514563107,
+      "grad_norm": 0.08464773744344711,
+      "learning_rate": 1.4107224658704288e-07,
+      "loss": 0.0009,
+      "step": 1379
+    },
+    {
+      "epoch": 13.398058252427184,
+      "grad_norm": 0.0651252344250679,
+      "learning_rate": 1.3939362725075344e-07,
+      "loss": 0.0006,
+      "step": 1380
+    },
+    {
+      "epoch": 13.407766990291263,
+      "grad_norm": 0.43149158358573914,
+      "learning_rate": 1.3772476828271236e-07,
+      "loss": 0.038,
+      "step": 1381
+    },
+    {
+      "epoch": 13.41747572815534,
+      "grad_norm": 0.2249184548854828,
+      "learning_rate": 1.360656765831289e-07,
+      "loss": 0.0051,
+      "step": 1382
+    },
+    {
+      "epoch": 13.427184466019417,
+      "grad_norm": 0.13268357515335083,
+      "learning_rate": 1.3441635901182803e-07,
+      "loss": 0.0016,
+      "step": 1383
+    },
+    {
+      "epoch": 13.436893203883495,
+      "grad_norm": 0.10658839344978333,
+      "learning_rate": 1.3277682238822142e-07,
+      "loss": 0.0015,
+      "step": 1384
+    },
+    {
+      "epoch": 13.446601941747574,
+      "grad_norm": 0.1615607887506485,
+      "learning_rate": 1.3114707349127954e-07,
+      "loss": 0.0026,
+      "step": 1385
+    },
+    {
+      "epoch": 13.45631067961165,
+      "grad_norm": 0.1557530164718628,
+      "learning_rate": 1.2952711905950377e-07,
+      "loss": 0.003,
+      "step": 1386
+    },
+    {
+      "epoch": 13.466019417475728,
+      "grad_norm": 0.286628782749176,
+      "learning_rate": 1.279169657908988e-07,
+      "loss": 0.0054,
+      "step": 1387
+    },
+    {
+      "epoch": 13.475728155339805,
+      "grad_norm": 0.12180517613887787,
+      "learning_rate": 1.263166203429439e-07,
+      "loss": 0.0018,
+      "step": 1388
+    },
+    {
+      "epoch": 13.485436893203884,
+      "grad_norm": 0.08528687059879303,
+      "learning_rate": 1.2472608933256637e-07,
+      "loss": 0.001,
+      "step": 1389
+    },
+    {
+      "epoch": 13.495145631067961,
+      "grad_norm": 0.12740404903888702,
+      "learning_rate": 1.2314537933611425e-07,
+      "loss": 0.0028,
+      "step": 1390
+    },
+    {
+      "epoch": 13.504854368932039,
+      "grad_norm": 0.14788556098937988,
+      "learning_rate": 1.2157449688932872e-07,
+      "loss": 0.0038,
+      "step": 1391
+    },
+    {
+      "epoch": 13.514563106796116,
+      "grad_norm": 0.15698152780532837,
+      "learning_rate": 1.2001344848731612e-07,
+      "loss": 0.0018,
+      "step": 1392
+    },
+    {
+      "epoch": 13.524271844660195,
+      "grad_norm": 0.13510096073150635,
+      "learning_rate": 1.1846224058452316e-07,
+      "loss": 0.0026,
+      "step": 1393
+    },
+    {
+      "epoch": 13.533980582524272,
+      "grad_norm": 0.13202345371246338,
+      "learning_rate": 1.1692087959470882e-07,
+      "loss": 0.0019,
+      "step": 1394
+    },
+    {
+      "epoch": 13.54368932038835,
+      "grad_norm": 0.27190133929252625,
+      "learning_rate": 1.1538937189091825e-07,
+      "loss": 0.007,
+      "step": 1395
+    },
+    {
+      "epoch": 13.553398058252426,
+      "grad_norm": 0.1908790022134781,
+      "learning_rate": 1.1386772380545669e-07,
+      "loss": 0.0046,
+      "step": 1396
+    },
+    {
+      "epoch": 13.563106796116505,
+      "grad_norm": 0.17838077247142792,
+      "learning_rate": 1.1235594162986168e-07,
+      "loss": 0.0045,
+      "step": 1397
+    },
+    {
+      "epoch": 13.572815533980583,
+      "grad_norm": 0.23380354046821594,
+      "learning_rate": 1.1085403161488012e-07,
+      "loss": 0.0043,
+      "step": 1398
+    },
+    {
+      "epoch": 13.58252427184466,
+      "grad_norm": 0.04911648854613304,
+      "learning_rate": 1.09361999970439e-07,
+      "loss": 0.0005,
+      "step": 1399
+    },
+    {
+      "epoch": 13.592233009708737,
+      "grad_norm": 0.12357534468173981,
+      "learning_rate": 1.0787985286562219e-07,
+      "loss": 0.0023,
+      "step": 1400
+    },
+    {
+      "epoch": 13.601941747572816,
+      "grad_norm": 0.11849277466535568,
+      "learning_rate": 1.0640759642864401e-07,
+      "loss": 0.0015,
+      "step": 1401
+    },
+    {
+      "epoch": 13.611650485436893,
+      "grad_norm": 0.20324498414993286,
+      "learning_rate": 1.0494523674682372e-07,
+      "loss": 0.0062,
+      "step": 1402
+    },
+    {
+      "epoch": 13.62135922330097,
+      "grad_norm": 0.16543932259082794,
+      "learning_rate": 1.0349277986656081e-07,
+      "loss": 0.0041,
+      "step": 1403
+    },
+    {
+      "epoch": 13.631067961165048,
+      "grad_norm": 0.14420188963413239,
+      "learning_rate": 1.0205023179330975e-07,
+      "loss": 0.0028,
+      "step": 1404
+    },
+    {
+      "epoch": 13.640776699029127,
+      "grad_norm": 0.47798407077789307,
+      "learning_rate": 1.00617598491555e-07,
+      "loss": 0.0376,
+      "step": 1405
+    },
+    {
+      "epoch": 13.650485436893204,
+      "grad_norm": 0.09702909737825394,
+      "learning_rate": 9.919488588478715e-08,
+      "loss": 0.0011,
+      "step": 1406
+    },
+    {
+      "epoch": 13.660194174757281,
+      "grad_norm": 0.24692678451538086,
+      "learning_rate": 9.778209985547682e-08,
+      "loss": 0.0054,
+      "step": 1407
+    },
+    {
+      "epoch": 13.669902912621358,
+      "grad_norm": 0.17881527543067932,
+      "learning_rate": 9.637924624505191e-08,
+      "loss": 0.0035,
+      "step": 1408
+    },
+    {
+      "epoch": 13.679611650485437,
+      "grad_norm": 0.1385125368833542,
+      "learning_rate": 9.498633085387343e-08,
+      "loss": 0.0031,
+      "step": 1409
+    },
+    {
+      "epoch": 13.689320388349515,
+      "grad_norm": 0.15435045957565308,
+      "learning_rate": 9.360335944121029e-08,
+      "loss": 0.0035,
+      "step": 1410
+    },
+    {
+      "epoch": 13.699029126213592,
+      "grad_norm": 0.09265675395727158,
+      "learning_rate": 9.223033772521594e-08,
+      "loss": 0.0014,
+      "step": 1411
+    },
+    {
+      "epoch": 13.70873786407767,
+      "grad_norm": 0.19740736484527588,
+      "learning_rate": 9.086727138290535e-08,
+      "loss": 0.0074,
+      "step": 1412
+    },
+    {
+      "epoch": 13.718446601941748,
+      "grad_norm": 0.14353950321674347,
+      "learning_rate": 8.951416605013114e-08,
+      "loss": 0.0033,
+      "step": 1413
+    },
+    {
+      "epoch": 13.728155339805825,
+      "grad_norm": 0.11986155807971954,
+      "learning_rate": 8.817102732155996e-08,
+      "loss": 0.0017,
+      "step": 1414
+    },
+    {
+      "epoch": 13.737864077669903,
+      "grad_norm": 0.16766305267810822,
+      "learning_rate": 8.683786075065065e-08,
+      "loss": 0.0031,
+      "step": 1415
+    },
+    {
+      "epoch": 13.74757281553398,
+      "grad_norm": 0.27344322204589844,
+      "learning_rate": 8.55146718496283e-08,
+      "loss": 0.0132,
+      "step": 1416
+    },
+    {
+      "epoch": 13.757281553398059,
+      "grad_norm": 0.21480204164981842,
+      "learning_rate": 8.420146608946605e-08,
+      "loss": 0.0055,
+      "step": 1417
+    },
+    {
+      "epoch": 13.766990291262136,
+      "grad_norm": 0.09167057275772095,
+      "learning_rate": 8.28982488998581e-08,
+      "loss": 0.0015,
+      "step": 1418
+    },
+    {
+      "epoch": 13.776699029126213,
+      "grad_norm": 0.25587520003318787,
+      "learning_rate": 8.160502566919942e-08,
+      "loss": 0.0044,
+      "step": 1419
+    },
+    {
+      "epoch": 13.78640776699029,
+      "grad_norm": 0.10709410905838013,
+      "learning_rate": 8.032180174456283e-08,
+      "loss": 0.0014,
+      "step": 1420
+    },
+    {
+      "epoch": 13.79611650485437,
+      "grad_norm": 0.15303583443164825,
+      "learning_rate": 7.904858243167806e-08,
+      "loss": 0.0043,
+      "step": 1421
+    },
+    {
+      "epoch": 13.805825242718447,
+      "grad_norm": 0.43564650416374207,
+      "learning_rate": 7.778537299490796e-08,
+      "loss": 0.0134,
+      "step": 1422
+    },
+    {
+      "epoch": 13.815533980582524,
+      "grad_norm": 0.16968989372253418,
+      "learning_rate": 7.653217865722817e-08,
+      "loss": 0.0037,
+      "step": 1423
+    },
+    {
+      "epoch": 13.825242718446601,
+      "grad_norm": 0.15621624886989594,
+      "learning_rate": 7.528900460020444e-08,
+      "loss": 0.0029,
+      "step": 1424
+    },
+    {
+      "epoch": 13.83495145631068,
+      "grad_norm": 0.25654706358909607,
+      "learning_rate": 7.405585596397314e-08,
+      "loss": 0.0023,
+      "step": 1425
+    },
+    {
+      "epoch": 13.844660194174757,
+      "grad_norm": 0.6071222424507141,
+      "learning_rate": 7.283273784721739e-08,
+      "loss": 0.0169,
+      "step": 1426
+    },
+    {
+      "epoch": 13.854368932038835,
+      "grad_norm": 0.09629572927951813,
+      "learning_rate": 7.161965530714743e-08,
+      "loss": 0.0012,
+      "step": 1427
+    },
+    {
+      "epoch": 13.864077669902912,
+      "grad_norm": 0.13124921917915344,
+      "learning_rate": 7.041661335948024e-08,
+      "loss": 0.002,
+      "step": 1428
+    },
+    {
+      "epoch": 13.87378640776699,
+      "grad_norm": 0.14983834326267242,
+      "learning_rate": 6.92236169784169e-08,
+      "loss": 0.0027,
+      "step": 1429
+    },
+    {
+      "epoch": 13.883495145631068,
+      "grad_norm": 0.16207829117774963,
+      "learning_rate": 6.804067109662443e-08,
+      "loss": 0.0036,
+      "step": 1430
+    },
+    {
+      "epoch": 13.893203883495145,
+      "grad_norm": 0.23562447726726532,
+      "learning_rate": 6.68677806052137e-08,
+      "loss": 0.0079,
+      "step": 1431
+    },
+    {
+      "epoch": 13.902912621359224,
+      "grad_norm": 0.2869737148284912,
+      "learning_rate": 6.57049503537191e-08,
+      "loss": 0.0069,
+      "step": 1432
+    },
+    {
+      "epoch": 13.912621359223301,
+      "grad_norm": 0.13367785513401031,
+      "learning_rate": 6.455218515008049e-08,
+      "loss": 0.0022,
+      "step": 1433
+    },
+    {
+      "epoch": 13.922330097087379,
+      "grad_norm": 0.21501076221466064,
+      "learning_rate": 6.340948976062023e-08,
+      "loss": 0.0055,
+      "step": 1434
+    },
+    {
+      "epoch": 13.932038834951456,
+      "grad_norm": 0.2994862198829651,
+      "learning_rate": 6.227686891002671e-08,
+      "loss": 0.0078,
+      "step": 1435
+    },
+    {
+      "epoch": 13.941747572815533,
+      "grad_norm": 0.1983334869146347,
+      "learning_rate": 6.115432728133198e-08,
+      "loss": 0.004,
+      "step": 1436
+    },
+    {
+      "epoch": 13.951456310679612,
+      "grad_norm": 0.2330927550792694,
+      "learning_rate": 6.004186951589414e-08,
+      "loss": 0.0093,
+      "step": 1437
+    },
+    {
+      "epoch": 13.96116504854369,
+      "grad_norm": 0.04621034115552902,
+      "learning_rate": 5.8939500213378296e-08,
+      "loss": 0.0004,
+      "step": 1438
+    },
+    {
+      "epoch": 13.970873786407767,
+      "grad_norm": 0.15329445898532867,
+      "learning_rate": 5.7847223931735974e-08,
+      "loss": 0.0033,
+      "step": 1439
+    },
+    {
+      "epoch": 13.980582524271846,
+      "grad_norm": 0.18739864230155945,
+      "learning_rate": 5.6765045187187614e-08,
+      "loss": 0.0033,
+      "step": 1440
+    },
+    {
+      "epoch": 13.990291262135923,
+      "grad_norm": 0.2754325270652771,
+      "learning_rate": 5.569296845420375e-08,
+      "loss": 0.0069,
+      "step": 1441
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.28285136818885803,
+      "learning_rate": 5.463099816548578e-08,
+      "loss": 0.0098,
+      "step": 1442
+    },
+    {
+      "epoch": 14.009708737864077,
+      "grad_norm": 0.0704316720366478,
+      "learning_rate": 5.3579138711948587e-08,
+      "loss": 0.0007,
+      "step": 1443
+    },
+    {
+      "epoch": 14.019417475728156,
+      "grad_norm": 0.22070623934268951,
+      "learning_rate": 5.253739444270128e-08,
+      "loss": 0.0049,
+      "step": 1444
+    },
+    {
+      "epoch": 14.029126213592233,
+      "grad_norm": 0.25098851323127747,
+      "learning_rate": 5.150576966503063e-08,
+      "loss": 0.005,
+      "step": 1445
+    },
+    {
+      "epoch": 14.03883495145631,
+      "grad_norm": 0.21593202650547028,
+      "learning_rate": 5.048426864438183e-08,
+      "loss": 0.0048,
+      "step": 1446
+    },
+    {
+      "epoch": 14.048543689320388,
+      "grad_norm": 0.07581067085266113,
+      "learning_rate": 4.9472895604341655e-08,
+      "loss": 0.001,
+      "step": 1447
+    },
+    {
+      "epoch": 14.058252427184467,
+      "grad_norm": 0.3260473310947418,
+      "learning_rate": 4.8471654726621464e-08,
+      "loss": 0.0166,
+      "step": 1448
+    },
+    {
+      "epoch": 14.067961165048544,
+      "grad_norm": 0.12206926196813583,
+      "learning_rate": 4.7480550151038365e-08,
+      "loss": 0.0029,
+      "step": 1449
+    },
+    {
+      "epoch": 14.077669902912621,
+      "grad_norm": 0.08410502225160599,
+      "learning_rate": 4.649958597549964e-08,
+      "loss": 0.0012,
+      "step": 1450
+    },
+    {
+      "epoch": 14.087378640776699,
+      "grad_norm": 0.16693179309368134,
+      "learning_rate": 4.552876625598501e-08,
+      "loss": 0.0047,
+      "step": 1451
+    },
+    {
+      "epoch": 14.097087378640778,
+      "grad_norm": 0.1822391152381897,
+      "learning_rate": 4.4568095006529975e-08,
+      "loss": 0.0054,
+      "step": 1452
+    },
+    {
+      "epoch": 14.106796116504855,
+      "grad_norm": 0.05579308047890663,
+      "learning_rate": 4.361757619920942e-08,
+      "loss": 0.0006,
+      "step": 1453
+    },
+    {
+      "epoch": 14.116504854368932,
+      "grad_norm": 0.07022775709629059,
+      "learning_rate": 4.2677213764120986e-08,
+      "loss": 0.0007,
+      "step": 1454
+    },
+    {
+      "epoch": 14.12621359223301,
+      "grad_norm": 0.08001091331243515,
+      "learning_rate": 4.174701158936895e-08,
+      "loss": 0.0011,
+      "step": 1455
+    },
+    {
+      "epoch": 14.135922330097088,
+      "grad_norm": 0.20097486674785614,
+      "learning_rate": 4.082697352104814e-08,
+      "loss": 0.007,
+      "step": 1456
+    },
+    {
+      "epoch": 14.145631067961165,
+      "grad_norm": 0.1303526610136032,
+      "learning_rate": 3.991710336322757e-08,
+      "loss": 0.0029,
+      "step": 1457
+    },
+    {
+      "epoch": 14.155339805825243,
+      "grad_norm": 0.10144368559122086,
+      "learning_rate": 3.9017404877935986e-08,
+      "loss": 0.0014,
+      "step": 1458
+    },
+    {
+      "epoch": 14.16504854368932,
+      "grad_norm": 0.29990750551223755,
+      "learning_rate": 3.812788178514437e-08,
+      "loss": 0.0089,
+      "step": 1459
+    },
+    {
+      "epoch": 14.174757281553399,
+      "grad_norm": 0.08991169184446335,
+      "learning_rate": 3.7248537762752666e-08,
+      "loss": 0.001,
+      "step": 1460
+    },
+    {
+      "epoch": 14.184466019417476,
+      "grad_norm": 0.12679709494113922,
+      "learning_rate": 3.637937644657308e-08,
+      "loss": 0.0026,
+      "step": 1461
+    },
+    {
+      "epoch": 14.194174757281553,
+      "grad_norm": 0.21683473885059357,
+      "learning_rate": 3.55204014303151e-08,
+      "loss": 0.006,
+      "step": 1462
+    },
+    {
+      "epoch": 14.20388349514563,
+      "grad_norm": 0.5879408121109009,
+      "learning_rate": 3.467161626557164e-08,
+      "loss": 0.0465,
+      "step": 1463
+    },
+    {
+      "epoch": 14.21359223300971,
+      "grad_norm": 0.12856099009513855,
+      "learning_rate": 3.3833024461803756e-08,
+      "loss": 0.0026,
+      "step": 1464
+    },
+    {
+      "epoch": 14.223300970873787,
+      "grad_norm": 0.22118137776851654,
+      "learning_rate": 3.300462948632593e-08,
+      "loss": 0.0089,
+      "step": 1465
+    },
+    {
+      "epoch": 14.233009708737864,
+      "grad_norm": 0.3074754774570465,
+      "learning_rate": 3.218643476429167e-08,
+      "loss": 0.0165,
+      "step": 1466
+    },
+    {
+      "epoch": 14.242718446601941,
+      "grad_norm": 0.12004687637090683,
+      "learning_rate": 3.1378443678680706e-08,
+      "loss": 0.0026,
+      "step": 1467
+    },
+    {
+      "epoch": 14.25242718446602,
+      "grad_norm": 0.14797070622444153,
+      "learning_rate": 3.0580659570282886e-08,
+      "loss": 0.0034,
+      "step": 1468
+    },
+    {
+      "epoch": 14.262135922330097,
+      "grad_norm": 0.2764511704444885,
+      "learning_rate": 2.979308573768547e-08,
+      "loss": 0.0077,
+      "step": 1469
+    },
+    {
+      "epoch": 14.271844660194175,
+      "grad_norm": 0.15794546902179718,
+      "learning_rate": 2.9015725437259724e-08,
+      "loss": 0.0036,
+      "step": 1470
+    },
+    {
+      "epoch": 14.281553398058252,
+      "grad_norm": 0.056184958666563034,
+      "learning_rate": 2.8248581883147387e-08,
+      "loss": 0.0005,
+      "step": 1471
+    },
+    {
+      "epoch": 14.29126213592233,
+      "grad_norm": 0.20542171597480774,
+      "learning_rate": 2.7491658247246478e-08,
+      "loss": 0.0058,
+      "step": 1472
+    },
+    {
+      "epoch": 14.300970873786408,
+      "grad_norm": 0.1507173627614975,
+      "learning_rate": 2.6744957659199376e-08,
+      "loss": 0.003,
+      "step": 1473
+    },
+    {
+      "epoch": 14.310679611650485,
+      "grad_norm": 0.1602557897567749,
+      "learning_rate": 2.6008483206379497e-08,
+      "loss": 0.0041,
+      "step": 1474
+    },
+    {
+      "epoch": 14.320388349514563,
+      "grad_norm": 0.1496906727552414,
+      "learning_rate": 2.5282237933877962e-08,
+      "loss": 0.0022,
+      "step": 1475
+    },
+    {
+      "epoch": 14.330097087378642,
+      "grad_norm": 0.1473391056060791,
+      "learning_rate": 2.4566224844491393e-08,
+      "loss": 0.003,
+      "step": 1476
+    },
+    {
+      "epoch": 14.339805825242719,
+      "grad_norm": 0.16431382298469543,
+      "learning_rate": 2.38604468987097e-08,
+      "loss": 0.0042,
+      "step": 1477
+    },
+    {
+      "epoch": 14.349514563106796,
+      "grad_norm": 0.06767686456441879,
+      "learning_rate": 2.316490701470414e-08,
+      "loss": 0.0009,
+      "step": 1478
+    },
+    {
+      "epoch": 14.359223300970873,
+      "grad_norm": 0.27724364399909973,
+      "learning_rate": 2.247960806831373e-08,
+      "loss": 0.0035,
+      "step": 1479
+    },
+    {
+      "epoch": 14.368932038834952,
+      "grad_norm": 0.30783212184906006,
+      "learning_rate": 2.180455289303579e-08,
+      "loss": 0.0127,
+      "step": 1480
+    },
+    {
+      "epoch": 14.37864077669903,
+      "grad_norm": 0.15318472683429718,
+      "learning_rate": 2.113974428001153e-08,
+      "loss": 0.004,
+      "step": 1481
+    },
+    {
+      "epoch": 14.388349514563107,
+      "grad_norm": 0.06371254473924637,
+      "learning_rate": 2.0485184978016604e-08,
+      "loss": 0.0008,
+      "step": 1482
+    },
+    {
+      "epoch": 14.398058252427184,
+      "grad_norm": 0.14508719742298126,
+      "learning_rate": 1.984087769344889e-08,
+      "loss": 0.0029,
+      "step": 1483
+    },
+    {
+      "epoch": 14.407766990291263,
+      "grad_norm": 0.2267545610666275,
+      "learning_rate": 1.9206825090317126e-08,
+      "loss": 0.0051,
+      "step": 1484
+    },
+    {
+      "epoch": 14.41747572815534,
+      "grad_norm": 0.27744215726852417,
+      "learning_rate": 1.8583029790230356e-08,
+      "loss": 0.0085,
+      "step": 1485
+    },
+    {
+      "epoch": 14.427184466019417,
+      "grad_norm": 0.24904626607894897,
+      "learning_rate": 1.796949437238682e-08,
+      "loss": 0.0063,
+      "step": 1486
+    },
+    {
+      "epoch": 14.436893203883495,
+      "grad_norm": 0.19574686884880066,
+      "learning_rate": 1.736622137356342e-08,
+      "loss": 0.0062,
+      "step": 1487
+    },
+    {
+      "epoch": 14.446601941747574,
+      "grad_norm": 0.06552054733037949,
+      "learning_rate": 1.677321328810516e-08,
+      "loss": 0.0007,
+      "step": 1488
+    },
+    {
+      "epoch": 14.45631067961165,
+      "grad_norm": 0.23957927525043488,
+      "learning_rate": 1.6190472567914617e-08,
+      "loss": 0.0097,
+      "step": 1489
+    },
+    {
+      "epoch": 14.466019417475728,
+      "grad_norm": 0.10903111845254898,
+      "learning_rate": 1.561800162244248e-08,
+      "loss": 0.0018,
+      "step": 1490
+    },
+    {
+      "epoch": 14.475728155339805,
+      "grad_norm": 0.24251382052898407,
+      "learning_rate": 1.5055802818676745e-08,
+      "loss": 0.0086,
+      "step": 1491
+    },
+    {
+      "epoch": 14.485436893203884,
+      "grad_norm": 0.15319575369358063,
+      "learning_rate": 1.450387848113327e-08,
+      "loss": 0.0045,
+      "step": 1492
+    },
+    {
+      "epoch": 14.495145631067961,
+      "grad_norm": 0.10059983283281326,
+      "learning_rate": 1.3962230891846618e-08,
+      "loss": 0.0017,
+      "step": 1493
+    },
+    {
+      "epoch": 14.504854368932039,
+      "grad_norm": 0.1764296442270279,
+      "learning_rate": 1.3430862290359781e-08,
+      "loss": 0.0033,
+      "step": 1494
+    },
+    {
+      "epoch": 14.514563106796116,
+      "grad_norm": 0.4749487042427063,
+      "learning_rate": 1.2909774873715585e-08,
+      "loss": 0.0158,
+      "step": 1495
+    },
+    {
+      "epoch": 14.524271844660195,
+      "grad_norm": 0.18315227329730988,
+      "learning_rate": 1.2398970796447807e-08,
+      "loss": 0.004,
+      "step": 1496
+    },
+    {
+      "epoch": 14.533980582524272,
+      "grad_norm": 0.36333203315734863,
+      "learning_rate": 1.1898452170570618e-08,
+      "loss": 0.0238,
+      "step": 1497
+    },
+    {
+      "epoch": 14.54368932038835,
+      "grad_norm": 0.06321419030427933,
+      "learning_rate": 1.140822106557249e-08,
+      "loss": 0.0005,
+      "step": 1498
+    },
+    {
+      "epoch": 14.553398058252426,
+      "grad_norm": 0.0857374295592308,
+      "learning_rate": 1.0928279508405082e-08,
+      "loss": 0.0006,
+      "step": 1499
+    },
+    {
+      "epoch": 14.563106796116505,
+      "grad_norm": 0.2555052638053894,
+      "learning_rate": 1.0458629483476868e-08,
+      "loss": 0.0063,
+      "step": 1500
+    },
+    {
+      "epoch": 14.572815533980583,
+      "grad_norm": 0.10225080698728561,
+      "learning_rate": 9.999272932643134e-09,
+      "loss": 0.0013,
+      "step": 1501
+    },
+    {
+      "epoch": 14.58252427184466,
+      "grad_norm": 0.19144372642040253,
+      "learning_rate": 9.550211755199879e-09,
+      "loss": 0.0038,
+      "step": 1502
+    },
+    {
+      "epoch": 14.592233009708737,
+      "grad_norm": 0.08165448158979416,
+      "learning_rate": 9.111447807874374e-09,
+      "loss": 0.0009,
+      "step": 1503
+    },
+    {
+      "epoch": 14.601941747572816,
+      "grad_norm": 0.04350195825099945,
+      "learning_rate": 8.682982904817948e-09,
+      "loss": 0.0003,
+      "step": 1504
+    },
+    {
+      "epoch": 14.611650485436893,
+      "grad_norm": 0.3616931736469269,
+      "learning_rate": 8.264818817599052e-09,
+      "loss": 0.0147,
+      "step": 1505
+    },
+    {
+      "epoch": 14.62135922330097,
+      "grad_norm": 0.09430540353059769,
+      "learning_rate": 7.856957275194921e-09,
+      "loss": 0.0015,
+      "step": 1506
+    },
+    {
+      "epoch": 14.631067961165048,
+      "grad_norm": 0.09992361813783646,
+      "learning_rate": 7.459399963985758e-09,
+      "loss": 0.0014,
+      "step": 1507
+    },
+    {
+      "epoch": 14.640776699029127,
+      "grad_norm": 0.21097907423973083,
+      "learning_rate": 7.072148527746403e-09,
+      "loss": 0.0067,
+      "step": 1508
+    },
+    {
+      "epoch": 14.650485436893204,
+      "grad_norm": 0.19425953924655914,
+      "learning_rate": 6.6952045676405005e-09,
+      "loss": 0.0035,
+      "step": 1509
+    },
+    {
+      "epoch": 14.660194174757281,
+      "grad_norm": 0.12785759568214417,
+      "learning_rate": 6.328569642212734e-09,
+      "loss": 0.0021,
+      "step": 1510
+    },
+    {
+      "epoch": 14.669902912621358,
+      "grad_norm": 0.2563770115375519,
+      "learning_rate": 5.972245267384102e-09,
+      "loss": 0.0085,
+      "step": 1511
+    },
+    {
+      "epoch": 14.679611650485437,
+      "grad_norm": 0.12719255685806274,
+      "learning_rate": 5.62623291644443e-09,
+      "loss": 0.002,
+      "step": 1512
+    },
+    {
+      "epoch": 14.689320388349515,
+      "grad_norm": 0.05718519538640976,
+      "learning_rate": 5.290534020046256e-09,
+      "loss": 0.0008,
+      "step": 1513
+    },
+    {
+      "epoch": 14.699029126213592,
+      "grad_norm": 0.11268595606088638,
+      "learning_rate": 4.965149966199567e-09,
+      "loss": 0.0018,
+      "step": 1514
+    },
+    {
+      "epoch": 14.70873786407767,
+      "grad_norm": 0.1993200033903122,
+      "learning_rate": 4.6500821002654075e-09,
+      "loss": 0.0077,
+      "step": 1515
+    },
+    {
+      "epoch": 14.718446601941748,
+      "grad_norm": 0.1494385302066803,
+      "learning_rate": 4.345331724950885e-09,
+      "loss": 0.0035,
+      "step": 1516
+    },
+    {
+      "epoch": 14.728155339805825,
+      "grad_norm": 0.13596144318580627,
+      "learning_rate": 4.050900100303068e-09,
+      "loss": 0.003,
+      "step": 1517
+    },
+    {
+      "epoch": 14.737864077669903,
+      "grad_norm": 0.09926627576351166,
+      "learning_rate": 3.766788443705094e-09,
+      "loss": 0.0017,
+      "step": 1518
+    },
+    {
+      "epoch": 14.74757281553398,
+      "grad_norm": 0.10796511173248291,
+      "learning_rate": 3.492997929869235e-09,
+      "loss": 0.0021,
+      "step": 1519
+    },
+    {
+      "epoch": 14.757281553398059,
+      "grad_norm": 0.12152813374996185,
+      "learning_rate": 3.2295296908338437e-09,
+      "loss": 0.0014,
+      "step": 1520
+    },
+    {
+      "epoch": 14.766990291262136,
+      "grad_norm": 0.12858039140701294,
+      "learning_rate": 2.976384815957245e-09,
+      "loss": 0.0015,
+      "step": 1521
+    },
+    {
+      "epoch": 14.776699029126213,
+      "grad_norm": 0.12139680236577988,
+      "learning_rate": 2.7335643519144086e-09,
+      "loss": 0.0015,
+      "step": 1522
+    },
+    {
+      "epoch": 14.78640776699029,
+      "grad_norm": 0.22163856029510498,
+      "learning_rate": 2.5010693026922273e-09,
+      "loss": 0.0066,
+      "step": 1523
+    },
+    {
+      "epoch": 14.79611650485437,
+      "grad_norm": 0.04654043912887573,
+      "learning_rate": 2.278900629584524e-09,
+      "loss": 0.0005,
+      "step": 1524
+    },
+    {
+      "epoch": 14.805825242718447,
+      "grad_norm": 0.33974435925483704,
+      "learning_rate": 2.067059251189274e-09,
+      "loss": 0.0097,
+      "step": 1525
+    },
+    {
+      "epoch": 14.815533980582524,
+      "grad_norm": 0.04179006814956665,
+      "learning_rate": 1.8655460434044427e-09,
+      "loss": 0.0004,
+      "step": 1526
+    },
+    {
+      "epoch": 14.825242718446601,
+      "grad_norm": 0.22010868787765503,
+      "learning_rate": 1.6743618394238215e-09,
+      "loss": 0.0069,
+      "step": 1527
+    },
+    {
+      "epoch": 14.83495145631068,
+      "grad_norm": 1.0385078191757202,
+      "learning_rate": 1.493507429734531e-09,
+      "loss": 0.0043,
+      "step": 1528
+    },
+    {
+      "epoch": 14.844660194174757,
+      "grad_norm": 0.4172058403491974,
+      "learning_rate": 1.3229835621125786e-09,
+      "loss": 0.0129,
+      "step": 1529
+    },
+    {
+      "epoch": 14.854368932038835,
+      "grad_norm": 0.17302429676055908,
+      "learning_rate": 1.1627909416211947e-09,
+      "loss": 0.0037,
+      "step": 1530
+    },
+    {
+      "epoch": 14.864077669902912,
+      "grad_norm": 0.1551245003938675,
+      "learning_rate": 1.0129302306061128e-09,
+      "loss": 0.004,
+      "step": 1531
+    },
+    {
+      "epoch": 14.87378640776699,
+      "grad_norm": 0.12023022770881653,
+      "learning_rate": 8.734020486950157e-10,
+      "loss": 0.0019,
+      "step": 1532
+    },
+    {
+      "epoch": 14.883495145631068,
+      "grad_norm": 0.04651810601353645,
+      "learning_rate": 7.442069727930934e-10,
+      "loss": 0.0004,
+      "step": 1533
+    },
+    {
+      "epoch": 14.893203883495145,
+      "grad_norm": 0.09399501979351044,
+      "learning_rate": 6.253455370811012e-10,
+      "loss": 0.0013,
+      "step": 1534
+    },
+    {
+      "epoch": 14.902912621359224,
+      "grad_norm": 0.12654393911361694,
+      "learning_rate": 5.168182330145266e-10,
+      "loss": 0.0024,
+      "step": 1535
+    },
+    {
+      "epoch": 14.912621359223301,
+      "grad_norm": 0.21424488723278046,
+      "learning_rate": 4.186255093194258e-10,
+      "loss": 0.0069,
+      "step": 1536
+    },
+    {
+      "epoch": 14.922330097087379,
+      "grad_norm": 0.20085856318473816,
+      "learning_rate": 3.3076777199186894e-10,
+      "loss": 0.0023,
+      "step": 1537
+    },
+    {
+      "epoch": 14.932038834951456,
+      "grad_norm": 0.14510715007781982,
+      "learning_rate": 2.532453842965521e-10,
+      "loss": 0.0045,
+      "step": 1538
+    },
+    {
+      "epoch": 14.941747572815533,
+      "grad_norm": 0.10943800210952759,
+      "learning_rate": 1.8605866676374428e-10,
+      "loss": 0.0011,
+      "step": 1539
+    },
+    {
+      "epoch": 14.951456310679612,
+      "grad_norm": 0.3463171422481537,
+      "learning_rate": 1.292078971898425e-10,
+      "loss": 0.0215,
+      "step": 1540
+    },
+    {
+      "epoch": 14.96116504854369,
+      "grad_norm": 0.033019255846738815,
+      "learning_rate": 8.269331063459618e-11,
+      "loss": 0.0002,
+      "step": 1541
+    },
+    {
+      "epoch": 14.970873786407767,
+      "grad_norm": 0.12394634634256363,
+      "learning_rate": 4.651509942193988e-11,
+      "loss": 0.0021,
+      "step": 1542
+    },
+    {
+      "epoch": 14.980582524271846,
+      "grad_norm": 0.13520504534244537,
+      "learning_rate": 2.06734131366626e-11,
+      "loss": 0.0027,
+      "step": 1543
+    },
+    {
+      "epoch": 14.990291262135923,
+      "grad_norm": 0.11344866454601288,
+      "learning_rate": 5.168358626628234e-12,
+      "loss": 0.002,
+      "step": 1544
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.062394145876169205,
+      "learning_rate": 0.0,
+      "loss": 0.0007,
+      "step": 1545
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 515,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 432414568808448.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}