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": 4.871730834654797,
+      "learning_rate": 4.999994831641374e-06,
+      "loss": 1.8478,
+      "step": 1
+    },
+    {
+      "epoch": 0.019417475728155338,
+      "grad_norm": 3.669459516332945,
+      "learning_rate": 4.9999793265868636e-06,
+      "loss": 1.4574,
+      "step": 2
+    },
+    {
+      "epoch": 0.02912621359223301,
+      "grad_norm": 3.321268843922114,
+      "learning_rate": 4.999953484900578e-06,
+      "loss": 1.5713,
+      "step": 3
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 4.757625498508145,
+      "learning_rate": 4.9999173066893655e-06,
+      "loss": 2.1849,
+      "step": 4
+    },
+    {
+      "epoch": 0.04854368932038835,
+      "grad_norm": 4.484235162199009,
+      "learning_rate": 4.9998707921028104e-06,
+      "loss": 2.538,
+      "step": 5
+    },
+    {
+      "epoch": 0.05825242718446602,
+      "grad_norm": 3.0105096661858055,
+      "learning_rate": 4.999813941333237e-06,
+      "loss": 1.5868,
+      "step": 6
+    },
+    {
+      "epoch": 0.06796116504854369,
+      "grad_norm": 2.8411927476391083,
+      "learning_rate": 4.999746754615704e-06,
+      "loss": 1.6472,
+      "step": 7
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 1.506880067190551,
+      "learning_rate": 4.9996692322280085e-06,
+      "loss": 1.0478,
+      "step": 8
+    },
+    {
+      "epoch": 0.08737864077669903,
+      "grad_norm": 4.0008731506353445,
+      "learning_rate": 4.999581374490681e-06,
+      "loss": 2.373,
+      "step": 9
+    },
+    {
+      "epoch": 0.0970873786407767,
+      "grad_norm": 2.4382442959213715,
+      "learning_rate": 4.999483181766986e-06,
+      "loss": 1.7706,
+      "step": 10
+    },
+    {
+      "epoch": 0.10679611650485436,
+      "grad_norm": 1.6052608591178483,
+      "learning_rate": 4.999374654462919e-06,
+      "loss": 1.3104,
+      "step": 11
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 2.7437905836581296,
+      "learning_rate": 4.999255793027207e-06,
+      "loss": 2.048,
+      "step": 12
+    },
+    {
+      "epoch": 0.1262135922330097,
+      "grad_norm": 1.3906335136930252,
+      "learning_rate": 4.999126597951305e-06,
+      "loss": 1.0066,
+      "step": 13
+    },
+    {
+      "epoch": 0.13592233009708737,
+      "grad_norm": 1.3340600696507865,
+      "learning_rate": 4.998987069769394e-06,
+      "loss": 1.0998,
+      "step": 14
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 1.7130835123840018,
+      "learning_rate": 4.998837209058379e-06,
+      "loss": 1.6654,
+      "step": 15
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 1.7208415743567615,
+      "learning_rate": 4.998677016437888e-06,
+      "loss": 1.3388,
+      "step": 16
+    },
+    {
+      "epoch": 0.1650485436893204,
+      "grad_norm": 2.4073143065161937,
+      "learning_rate": 4.998506492570266e-06,
+      "loss": 1.8365,
+      "step": 17
+    },
+    {
+      "epoch": 0.17475728155339806,
+      "grad_norm": 1.4448713906198571,
+      "learning_rate": 4.998325638160576e-06,
+      "loss": 1.4415,
+      "step": 18
+    },
+    {
+      "epoch": 0.18446601941747573,
+      "grad_norm": 1.5964597845558501,
+      "learning_rate": 4.998134453956596e-06,
+      "loss": 1.1629,
+      "step": 19
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.3355139686727637,
+      "learning_rate": 4.997932940748811e-06,
+      "loss": 1.2245,
+      "step": 20
+    },
+    {
+      "epoch": 0.20388349514563106,
+      "grad_norm": 1.2386749477462127,
+      "learning_rate": 4.997721099370416e-06,
+      "loss": 1.2827,
+      "step": 21
+    },
+    {
+      "epoch": 0.21359223300970873,
+      "grad_norm": 1.3520067333735615,
+      "learning_rate": 4.997498930697308e-06,
+      "loss": 1.3963,
+      "step": 22
+    },
+    {
+      "epoch": 0.22330097087378642,
+      "grad_norm": 1.7137484637690432,
+      "learning_rate": 4.997266435648086e-06,
+      "loss": 1.4022,
+      "step": 23
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.9868889107654797,
+      "learning_rate": 4.997023615184044e-06,
+      "loss": 1.7246,
+      "step": 24
+    },
+    {
+      "epoch": 0.24271844660194175,
+      "grad_norm": 1.6032606855000195,
+      "learning_rate": 4.996770470309167e-06,
+      "loss": 1.7975,
+      "step": 25
+    },
+    {
+      "epoch": 0.2524271844660194,
+      "grad_norm": 1.268066436948988,
+      "learning_rate": 4.996507002070131e-06,
+      "loss": 1.3028,
+      "step": 26
+    },
+    {
+      "epoch": 0.2621359223300971,
+      "grad_norm": 1.2303902485186446,
+      "learning_rate": 4.996233211556295e-06,
+      "loss": 1.3613,
+      "step": 27
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 0.9549363286063691,
+      "learning_rate": 4.9959490998996974e-06,
+      "loss": 0.9778,
+      "step": 28
+    },
+    {
+      "epoch": 0.2815533980582524,
+      "grad_norm": 1.2603575097056405,
+      "learning_rate": 4.995654668275049e-06,
+      "loss": 1.4077,
+      "step": 29
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 1.5075543470098034,
+      "learning_rate": 4.995349917899735e-06,
+      "loss": 1.5452,
+      "step": 30
+    },
+    {
+      "epoch": 0.30097087378640774,
+      "grad_norm": 0.9482722379956937,
+      "learning_rate": 4.9950348500338005e-06,
+      "loss": 1.0589,
+      "step": 31
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 0.8370896925167312,
+      "learning_rate": 4.994709465979954e-06,
+      "loss": 0.977,
+      "step": 32
+    },
+    {
+      "epoch": 0.32038834951456313,
+      "grad_norm": 1.204641294852316,
+      "learning_rate": 4.994373767083556e-06,
+      "loss": 1.2391,
+      "step": 33
+    },
+    {
+      "epoch": 0.3300970873786408,
+      "grad_norm": 1.1272285807079887,
+      "learning_rate": 4.994027754732616e-06,
+      "loss": 1.2678,
+      "step": 34
+    },
+    {
+      "epoch": 0.33980582524271846,
+      "grad_norm": 1.0240110484845462,
+      "learning_rate": 4.993671430357788e-06,
+      "loss": 1.1705,
+      "step": 35
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.3831448718921984,
+      "learning_rate": 4.99330479543236e-06,
+      "loss": 1.6517,
+      "step": 36
+    },
+    {
+      "epoch": 0.3592233009708738,
+      "grad_norm": 1.0046287010394588,
+      "learning_rate": 4.992927851472254e-06,
+      "loss": 1.2714,
+      "step": 37
+    },
+    {
+      "epoch": 0.36893203883495146,
+      "grad_norm": 1.3527812004094968,
+      "learning_rate": 4.992540600036014e-06,
+      "loss": 1.6523,
+      "step": 38
+    },
+    {
+      "epoch": 0.3786407766990291,
+      "grad_norm": 1.162751011119127,
+      "learning_rate": 4.992143042724805e-06,
+      "loss": 1.4164,
+      "step": 39
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 1.183212466699315,
+      "learning_rate": 4.991735181182401e-06,
+      "loss": 1.6037,
+      "step": 40
+    },
+    {
+      "epoch": 0.39805825242718446,
+      "grad_norm": 1.0671445151315921,
+      "learning_rate": 4.991317017095182e-06,
+      "loss": 1.2074,
+      "step": 41
+    },
+    {
+      "epoch": 0.4077669902912621,
+      "grad_norm": 1.1374295195195832,
+      "learning_rate": 4.990888552192126e-06,
+      "loss": 1.2052,
+      "step": 42
+    },
+    {
+      "epoch": 0.4174757281553398,
+      "grad_norm": 1.1838914278145132,
+      "learning_rate": 4.9904497882448004e-06,
+      "loss": 1.3772,
+      "step": 43
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.1898023150727208,
+      "learning_rate": 4.990000727067357e-06,
+      "loss": 1.4616,
+      "step": 44
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 1.3045045420389443,
+      "learning_rate": 4.989541370516523e-06,
+      "loss": 1.2714,
+      "step": 45
+    },
+    {
+      "epoch": 0.44660194174757284,
+      "grad_norm": 1.1641307331044313,
+      "learning_rate": 4.989071720491595e-06,
+      "loss": 1.2003,
+      "step": 46
+    },
+    {
+      "epoch": 0.4563106796116505,
+      "grad_norm": 0.6576132196355234,
+      "learning_rate": 4.988591778934428e-06,
+      "loss": 0.791,
+      "step": 47
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 0.8500656663270045,
+      "learning_rate": 4.9881015478294294e-06,
+      "loss": 0.9369,
+      "step": 48
+    },
+    {
+      "epoch": 0.47572815533980584,
+      "grad_norm": 1.4233952816867905,
+      "learning_rate": 4.987601029203553e-06,
+      "loss": 1.4027,
+      "step": 49
+    },
+    {
+      "epoch": 0.4854368932038835,
+      "grad_norm": 1.1258574635846645,
+      "learning_rate": 4.987090225126285e-06,
+      "loss": 1.3511,
+      "step": 50
+    },
+    {
+      "epoch": 0.49514563106796117,
+      "grad_norm": 1.1687633946081721,
+      "learning_rate": 4.98656913770964e-06,
+      "loss": 1.1697,
+      "step": 51
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 1.2317336138068717,
+      "learning_rate": 4.986037769108154e-06,
+      "loss": 1.4503,
+      "step": 52
+    },
+    {
+      "epoch": 0.5145631067961165,
+      "grad_norm": 0.7574877986414429,
+      "learning_rate": 4.9854961215188676e-06,
+      "loss": 1.1225,
+      "step": 53
+    },
+    {
+      "epoch": 0.5242718446601942,
+      "grad_norm": 0.8037978880160027,
+      "learning_rate": 4.984944197181324e-06,
+      "loss": 0.9828,
+      "step": 54
+    },
+    {
+      "epoch": 0.5339805825242718,
+      "grad_norm": 1.1494946115521347,
+      "learning_rate": 4.9843819983775575e-06,
+      "loss": 1.2705,
+      "step": 55
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 1.0404784765183215,
+      "learning_rate": 4.983809527432086e-06,
+      "loss": 1.1189,
+      "step": 56
+    },
+    {
+      "epoch": 0.5533980582524272,
+      "grad_norm": 1.0072279484535855,
+      "learning_rate": 4.983226786711895e-06,
+      "loss": 1.2205,
+      "step": 57
+    },
+    {
+      "epoch": 0.5631067961165048,
+      "grad_norm": 0.9065921594118113,
+      "learning_rate": 4.982633778626437e-06,
+      "loss": 1.07,
+      "step": 58
+    },
+    {
+      "epoch": 0.5728155339805825,
+      "grad_norm": 1.2370148909491399,
+      "learning_rate": 4.982030505627613e-06,
+      "loss": 1.535,
+      "step": 59
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.7862714572091344,
+      "learning_rate": 4.98141697020977e-06,
+      "loss": 1.0275,
+      "step": 60
+    },
+    {
+      "epoch": 0.5922330097087378,
+      "grad_norm": 1.0179109472315906,
+      "learning_rate": 4.9807931749096836e-06,
+      "loss": 1.1225,
+      "step": 61
+    },
+    {
+      "epoch": 0.6019417475728155,
+      "grad_norm": 0.6647718819463199,
+      "learning_rate": 4.980159122306551e-06,
+      "loss": 0.9871,
+      "step": 62
+    },
+    {
+      "epoch": 0.6116504854368932,
+      "grad_norm": 0.7487245734536395,
+      "learning_rate": 4.979514815021984e-06,
+      "loss": 1.2099,
+      "step": 63
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 1.1029648607720772,
+      "learning_rate": 4.978860255719989e-06,
+      "loss": 1.0207,
+      "step": 64
+    },
+    {
+      "epoch": 0.6310679611650486,
+      "grad_norm": 0.7621265063758594,
+      "learning_rate": 4.978195447106965e-06,
+      "loss": 1.1363,
+      "step": 65
+    },
+    {
+      "epoch": 0.6407766990291263,
+      "grad_norm": 2.313544656936807,
+      "learning_rate": 4.9775203919316864e-06,
+      "loss": 0.9453,
+      "step": 66
+    },
+    {
+      "epoch": 0.6504854368932039,
+      "grad_norm": 0.8280847399908062,
+      "learning_rate": 4.976835092985297e-06,
+      "loss": 1.1131,
+      "step": 67
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 0.8927638160250211,
+      "learning_rate": 4.976139553101291e-06,
+      "loss": 1.133,
+      "step": 68
+    },
+    {
+      "epoch": 0.6699029126213593,
+      "grad_norm": 0.9610176243493397,
+      "learning_rate": 4.975433775155509e-06,
+      "loss": 1.1768,
+      "step": 69
+    },
+    {
+      "epoch": 0.6796116504854369,
+      "grad_norm": 1.194498620463663,
+      "learning_rate": 4.974717762066123e-06,
+      "loss": 1.3615,
+      "step": 70
+    },
+    {
+      "epoch": 0.6893203883495146,
+      "grad_norm": 0.8347048742413943,
+      "learning_rate": 4.973991516793621e-06,
+      "loss": 1.2546,
+      "step": 71
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 0.9172122226894214,
+      "learning_rate": 4.973255042340801e-06,
+      "loss": 1.1607,
+      "step": 72
+    },
+    {
+      "epoch": 0.7087378640776699,
+      "grad_norm": 1.03327431376727,
+      "learning_rate": 4.972508341752754e-06,
+      "loss": 1.361,
+      "step": 73
+    },
+    {
+      "epoch": 0.7184466019417476,
+      "grad_norm": 1.3504753816166224,
+      "learning_rate": 4.9717514181168534e-06,
+      "loss": 0.9757,
+      "step": 74
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 1.002604662508283,
+      "learning_rate": 4.970984274562741e-06,
+      "loss": 1.145,
+      "step": 75
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 0.8521212437833093,
+      "learning_rate": 4.970206914262315e-06,
+      "loss": 1.0841,
+      "step": 76
+    },
+    {
+      "epoch": 0.7475728155339806,
+      "grad_norm": 0.7268667125506638,
+      "learning_rate": 4.969419340429717e-06,
+      "loss": 1.0124,
+      "step": 77
+    },
+    {
+      "epoch": 0.7572815533980582,
+      "grad_norm": 1.0230017430471454,
+      "learning_rate": 4.968621556321319e-06,
+      "loss": 1.4082,
+      "step": 78
+    },
+    {
+      "epoch": 0.7669902912621359,
+      "grad_norm": 0.8075925843214921,
+      "learning_rate": 4.967813565235708e-06,
+      "loss": 0.9341,
+      "step": 79
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.7417143914647092,
+      "learning_rate": 4.966995370513675e-06,
+      "loss": 0.9597,
+      "step": 80
+    },
+    {
+      "epoch": 0.7864077669902912,
+      "grad_norm": 0.9927625054735497,
+      "learning_rate": 4.966166975538197e-06,
+      "loss": 1.298,
+      "step": 81
+    },
+    {
+      "epoch": 0.7961165048543689,
+      "grad_norm": 0.6918994370607547,
+      "learning_rate": 4.965328383734429e-06,
+      "loss": 0.9559,
+      "step": 82
+    },
+    {
+      "epoch": 0.8058252427184466,
+      "grad_norm": 0.9084696462282987,
+      "learning_rate": 4.964479598569686e-06,
+      "loss": 1.3896,
+      "step": 83
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.9240522865451553,
+      "learning_rate": 4.963620623553428e-06,
+      "loss": 1.2391,
+      "step": 84
+    },
+    {
+      "epoch": 0.8252427184466019,
+      "grad_norm": 0.9801285757696232,
+      "learning_rate": 4.962751462237248e-06,
+      "loss": 0.9936,
+      "step": 85
+    },
+    {
+      "epoch": 0.8349514563106796,
+      "grad_norm": 0.7316677725651993,
+      "learning_rate": 4.9618721182148564e-06,
+      "loss": 0.8096,
+      "step": 86
+    },
+    {
+      "epoch": 0.8446601941747572,
+      "grad_norm": 0.7986070833021874,
+      "learning_rate": 4.960982595122064e-06,
+      "loss": 0.9936,
+      "step": 87
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 2.991170115347129,
+      "learning_rate": 4.960082896636773e-06,
+      "loss": 1.2219,
+      "step": 88
+    },
+    {
+      "epoch": 0.8640776699029126,
+      "grad_norm": 0.841666379855367,
+      "learning_rate": 4.959173026478952e-06,
+      "loss": 1.0418,
+      "step": 89
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.8276404717553167,
+      "learning_rate": 4.958252988410631e-06,
+      "loss": 0.9154,
+      "step": 90
+    },
+    {
+      "epoch": 0.883495145631068,
+      "grad_norm": 0.9617539349436555,
+      "learning_rate": 4.9573227862358794e-06,
+      "loss": 1.1546,
+      "step": 91
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 0.9410894322853203,
+      "learning_rate": 4.956382423800791e-06,
+      "loss": 1.2453,
+      "step": 92
+    },
+    {
+      "epoch": 0.9029126213592233,
+      "grad_norm": 0.7461574178337397,
+      "learning_rate": 4.955431904993471e-06,
+      "loss": 1.0039,
+      "step": 93
+    },
+    {
+      "epoch": 0.912621359223301,
+      "grad_norm": 0.7787771559636139,
+      "learning_rate": 4.954471233744015e-06,
+      "loss": 0.7834,
+      "step": 94
+    },
+    {
+      "epoch": 0.9223300970873787,
+      "grad_norm": 0.8319061238375747,
+      "learning_rate": 4.9535004140245005e-06,
+      "loss": 0.9597,
+      "step": 95
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 1.241881123747875,
+      "learning_rate": 4.952519449848962e-06,
+      "loss": 1.2155,
+      "step": 96
+    },
+    {
+      "epoch": 0.941747572815534,
+      "grad_norm": 0.9416478821490492,
+      "learning_rate": 4.951528345273379e-06,
+      "loss": 0.9302,
+      "step": 97
+    },
+    {
+      "epoch": 0.9514563106796117,
+      "grad_norm": 0.6462620680684062,
+      "learning_rate": 4.950527104395659e-06,
+      "loss": 0.9483,
+      "step": 98
+    },
+    {
+      "epoch": 0.9611650485436893,
+      "grad_norm": 1.3945318287905581,
+      "learning_rate": 4.9495157313556185e-06,
+      "loss": 0.884,
+      "step": 99
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.6127133377291775,
+      "learning_rate": 4.94849423033497e-06,
+      "loss": 0.6754,
+      "step": 100
+    },
+    {
+      "epoch": 0.9805825242718447,
+      "grad_norm": 0.8204199917952882,
+      "learning_rate": 4.9474626055573e-06,
+      "loss": 1.1493,
+      "step": 101
+    },
+    {
+      "epoch": 0.9902912621359223,
+      "grad_norm": 0.6412217993750089,
+      "learning_rate": 4.946420861288051e-06,
+      "loss": 0.9205,
+      "step": 102
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7254619944280207,
+      "learning_rate": 4.9453690018345144e-06,
+      "loss": 0.9894,
+      "step": 103
+    },
+    {
+      "epoch": 1.0097087378640777,
+      "grad_norm": 0.8838225821340072,
+      "learning_rate": 4.944307031545797e-06,
+      "loss": 1.0139,
+      "step": 104
+    },
+    {
+      "epoch": 1.0194174757281553,
+      "grad_norm": 0.8681724674727661,
+      "learning_rate": 4.943234954812812e-06,
+      "loss": 1.2911,
+      "step": 105
+    },
+    {
+      "epoch": 1.029126213592233,
+      "grad_norm": 0.6707492146981981,
+      "learning_rate": 4.942152776068264e-06,
+      "loss": 0.8712,
+      "step": 106
+    },
+    {
+      "epoch": 1.0388349514563107,
+      "grad_norm": 0.8258075866719828,
+      "learning_rate": 4.941060499786622e-06,
+      "loss": 1.2535,
+      "step": 107
+    },
+    {
+      "epoch": 1.0485436893203883,
+      "grad_norm": 0.9343039880211657,
+      "learning_rate": 4.939958130484106e-06,
+      "loss": 1.0248,
+      "step": 108
+    },
+    {
+      "epoch": 1.058252427184466,
+      "grad_norm": 0.6215772104671746,
+      "learning_rate": 4.938845672718668e-06,
+      "loss": 0.899,
+      "step": 109
+    },
+    {
+      "epoch": 1.0679611650485437,
+      "grad_norm": 0.6106502770977033,
+      "learning_rate": 4.937723131089974e-06,
+      "loss": 0.8534,
+      "step": 110
+    },
+    {
+      "epoch": 1.0776699029126213,
+      "grad_norm": 0.8527770514681675,
+      "learning_rate": 4.93659051023938e-06,
+      "loss": 1.0798,
+      "step": 111
+    },
+    {
+      "epoch": 1.087378640776699,
+      "grad_norm": 0.6013648488199381,
+      "learning_rate": 4.93544781484992e-06,
+      "loss": 0.7126,
+      "step": 112
+    },
+    {
+      "epoch": 1.0970873786407767,
+      "grad_norm": 0.673264040200857,
+      "learning_rate": 4.9342950496462815e-06,
+      "loss": 1.0433,
+      "step": 113
+    },
+    {
+      "epoch": 1.1067961165048543,
+      "grad_norm": 0.6367668762270327,
+      "learning_rate": 4.933132219394786e-06,
+      "loss": 0.8586,
+      "step": 114
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.7326471152183239,
+      "learning_rate": 4.931959328903376e-06,
+      "loss": 1.1661,
+      "step": 115
+    },
+    {
+      "epoch": 1.1262135922330097,
+      "grad_norm": 0.7905138029016624,
+      "learning_rate": 4.930776383021584e-06,
+      "loss": 0.9155,
+      "step": 116
+    },
+    {
+      "epoch": 1.1359223300970873,
+      "grad_norm": 0.7113840452485917,
+      "learning_rate": 4.92958338664052e-06,
+      "loss": 0.7101,
+      "step": 117
+    },
+    {
+      "epoch": 1.145631067961165,
+      "grad_norm": 0.7085441798426322,
+      "learning_rate": 4.928380344692853e-06,
+      "loss": 0.9499,
+      "step": 118
+    },
+    {
+      "epoch": 1.1553398058252426,
+      "grad_norm": 0.6654689711573367,
+      "learning_rate": 4.927167262152784e-06,
+      "loss": 0.9453,
+      "step": 119
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.7093345141752957,
+      "learning_rate": 4.925944144036027e-06,
+      "loss": 1.0183,
+      "step": 120
+    },
+    {
+      "epoch": 1.174757281553398,
+      "grad_norm": 0.7761162990313378,
+      "learning_rate": 4.924710995399796e-06,
+      "loss": 0.9277,
+      "step": 121
+    },
+    {
+      "epoch": 1.1844660194174756,
+      "grad_norm": 0.8690880735001523,
+      "learning_rate": 4.923467821342773e-06,
+      "loss": 1.1022,
+      "step": 122
+    },
+    {
+      "epoch": 1.1941747572815533,
+      "grad_norm": 0.8588202527080991,
+      "learning_rate": 4.922214627005092e-06,
+      "loss": 1.0451,
+      "step": 123
+    },
+    {
+      "epoch": 1.203883495145631,
+      "grad_norm": 0.6415200336267318,
+      "learning_rate": 4.920951417568323e-06,
+      "loss": 1.004,
+      "step": 124
+    },
+    {
+      "epoch": 1.2135922330097086,
+      "grad_norm": 0.796282956613202,
+      "learning_rate": 4.919678198255438e-06,
+      "loss": 0.9759,
+      "step": 125
+    },
+    {
+      "epoch": 1.2233009708737863,
+      "grad_norm": 0.7236053821892523,
+      "learning_rate": 4.918394974330801e-06,
+      "loss": 1.0669,
+      "step": 126
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 0.7465165265947691,
+      "learning_rate": 4.917101751100142e-06,
+      "loss": 1.0007,
+      "step": 127
+    },
+    {
+      "epoch": 1.2427184466019416,
+      "grad_norm": 0.6913202728148214,
+      "learning_rate": 4.915798533910534e-06,
+      "loss": 1.037,
+      "step": 128
+    },
+    {
+      "epoch": 1.2524271844660193,
+      "grad_norm": 0.6269818244555071,
+      "learning_rate": 4.9144853281503715e-06,
+      "loss": 0.7153,
+      "step": 129
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.7329901927771046,
+      "learning_rate": 4.91316213924935e-06,
+      "loss": 1.0037,
+      "step": 130
+    },
+    {
+      "epoch": 1.2718446601941746,
+      "grad_norm": 0.9047110334486224,
+      "learning_rate": 4.911828972678441e-06,
+      "loss": 0.9842,
+      "step": 131
+    },
+    {
+      "epoch": 1.2815533980582523,
+      "grad_norm": 0.7849365731256595,
+      "learning_rate": 4.91048583394987e-06,
+      "loss": 0.8823,
+      "step": 132
+    },
+    {
+      "epoch": 1.29126213592233,
+      "grad_norm": 0.8276108169878443,
+      "learning_rate": 4.909132728617095e-06,
+      "loss": 1.2258,
+      "step": 133
+    },
+    {
+      "epoch": 1.3009708737864076,
+      "grad_norm": 0.6818450934871862,
+      "learning_rate": 4.907769662274785e-06,
+      "loss": 0.9031,
+      "step": 134
+    },
+    {
+      "epoch": 1.3106796116504853,
+      "grad_norm": 0.573255811123994,
+      "learning_rate": 4.90639664055879e-06,
+      "loss": 0.7234,
+      "step": 135
+    },
+    {
+      "epoch": 1.3203883495145632,
+      "grad_norm": 0.6504319050897613,
+      "learning_rate": 4.905013669146127e-06,
+      "loss": 0.8108,
+      "step": 136
+    },
+    {
+      "epoch": 1.3300970873786409,
+      "grad_norm": 0.6160025172155659,
+      "learning_rate": 4.903620753754949e-06,
+      "loss": 0.951,
+      "step": 137
+    },
+    {
+      "epoch": 1.3398058252427185,
+      "grad_norm": 0.7489527548993854,
+      "learning_rate": 4.902217900144524e-06,
+      "loss": 1.1361,
+      "step": 138
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 0.7112858054270321,
+      "learning_rate": 4.900805114115214e-06,
+      "loss": 1.1212,
+      "step": 139
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.6142945035367823,
+      "learning_rate": 4.899382401508446e-06,
+      "loss": 0.6799,
+      "step": 140
+    },
+    {
+      "epoch": 1.3689320388349515,
+      "grad_norm": 0.7192927924067215,
+      "learning_rate": 4.8979497682066916e-06,
+      "loss": 0.9943,
+      "step": 141
+    },
+    {
+      "epoch": 1.3786407766990292,
+      "grad_norm": 0.5629568582132167,
+      "learning_rate": 4.89650722013344e-06,
+      "loss": 0.6788,
+      "step": 142
+    },
+    {
+      "epoch": 1.3883495145631068,
+      "grad_norm": 1.1483194850326925,
+      "learning_rate": 4.895054763253177e-06,
+      "loss": 1.0813,
+      "step": 143
+    },
+    {
+      "epoch": 1.3980582524271845,
+      "grad_norm": 0.7435969328117086,
+      "learning_rate": 4.8935924035713564e-06,
+      "loss": 0.8197,
+      "step": 144
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 1.317031074434914,
+      "learning_rate": 4.892120147134378e-06,
+      "loss": 1.0709,
+      "step": 145
+    },
+    {
+      "epoch": 1.4174757281553398,
+      "grad_norm": 0.5996482737135205,
+      "learning_rate": 4.8906380000295615e-06,
+      "loss": 0.7866,
+      "step": 146
+    },
+    {
+      "epoch": 1.4271844660194175,
+      "grad_norm": 0.7791569768007636,
+      "learning_rate": 4.889145968385121e-06,
+      "loss": 0.8691,
+      "step": 147
+    },
+    {
+      "epoch": 1.4368932038834952,
+      "grad_norm": 0.6501971847860233,
+      "learning_rate": 4.887644058370139e-06,
+      "loss": 0.6976,
+      "step": 148
+    },
+    {
+      "epoch": 1.4466019417475728,
+      "grad_norm": 0.6511216890119076,
+      "learning_rate": 4.886132276194544e-06,
+      "loss": 0.7491,
+      "step": 149
+    },
+    {
+      "epoch": 1.4563106796116505,
+      "grad_norm": 0.612023999998433,
+      "learning_rate": 4.884610628109082e-06,
+      "loss": 0.717,
+      "step": 150
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 0.7773151050365769,
+      "learning_rate": 4.883079120405292e-06,
+      "loss": 1.004,
+      "step": 151
+    },
+    {
+      "epoch": 1.4757281553398058,
+      "grad_norm": 0.6158372837580877,
+      "learning_rate": 4.881537759415478e-06,
+      "loss": 0.6869,
+      "step": 152
+    },
+    {
+      "epoch": 1.4854368932038835,
+      "grad_norm": 0.8120612151123343,
+      "learning_rate": 4.879986551512684e-06,
+      "loss": 0.9272,
+      "step": 153
+    },
+    {
+      "epoch": 1.4951456310679612,
+      "grad_norm": 0.9467728776526185,
+      "learning_rate": 4.878425503110672e-06,
+      "loss": 0.7134,
+      "step": 154
+    },
+    {
+      "epoch": 1.5048543689320388,
+      "grad_norm": 0.7268524122061516,
+      "learning_rate": 4.876854620663887e-06,
+      "loss": 0.9148,
+      "step": 155
+    },
+    {
+      "epoch": 1.5145631067961165,
+      "grad_norm": 0.5995550004105531,
+      "learning_rate": 4.875273910667434e-06,
+      "loss": 0.74,
+      "step": 156
+    },
+    {
+      "epoch": 1.5242718446601942,
+      "grad_norm": 0.8211820446136313,
+      "learning_rate": 4.873683379657057e-06,
+      "loss": 0.7739,
+      "step": 157
+    },
+    {
+      "epoch": 1.5339805825242718,
+      "grad_norm": 0.6823628164842023,
+      "learning_rate": 4.8720830342091015e-06,
+      "loss": 1.1488,
+      "step": 158
+    },
+    {
+      "epoch": 1.5436893203883495,
+      "grad_norm": 0.7014265251368315,
+      "learning_rate": 4.870472880940496e-06,
+      "loss": 0.9573,
+      "step": 159
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 0.8352178057977285,
+      "learning_rate": 4.868852926508721e-06,
+      "loss": 0.9985,
+      "step": 160
+    },
+    {
+      "epoch": 1.5631067961165048,
+      "grad_norm": 0.756687144256477,
+      "learning_rate": 4.867223177611779e-06,
+      "loss": 0.938,
+      "step": 161
+    },
+    {
+      "epoch": 1.5728155339805825,
+      "grad_norm": 0.704162668693477,
+      "learning_rate": 4.865583640988173e-06,
+      "loss": 0.7891,
+      "step": 162
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 0.5937940435523842,
+      "learning_rate": 4.863934323416871e-06,
+      "loss": 0.7824,
+      "step": 163
+    },
+    {
+      "epoch": 1.5922330097087378,
+      "grad_norm": 0.5548908895552046,
+      "learning_rate": 4.862275231717288e-06,
+      "loss": 0.7329,
+      "step": 164
+    },
+    {
+      "epoch": 1.6019417475728155,
+      "grad_norm": 0.8668497443441112,
+      "learning_rate": 4.860606372749247e-06,
+      "loss": 0.7515,
+      "step": 165
+    },
+    {
+      "epoch": 1.6116504854368932,
+      "grad_norm": 0.8678487798318208,
+      "learning_rate": 4.858927753412958e-06,
+      "loss": 0.9059,
+      "step": 166
+    },
+    {
+      "epoch": 1.6213592233009708,
+      "grad_norm": 0.6236208019957972,
+      "learning_rate": 4.857239380648985e-06,
+      "loss": 0.7846,
+      "step": 167
+    },
+    {
+      "epoch": 1.6310679611650487,
+      "grad_norm": 0.6093138120471663,
+      "learning_rate": 4.855541261438223e-06,
+      "loss": 0.8363,
+      "step": 168
+    },
+    {
+      "epoch": 1.6407766990291264,
+      "grad_norm": 0.6297044385352409,
+      "learning_rate": 4.8538334028018605e-06,
+      "loss": 1.0275,
+      "step": 169
+    },
+    {
+      "epoch": 1.650485436893204,
+      "grad_norm": 0.7705428538880952,
+      "learning_rate": 4.8521158118013605e-06,
+      "loss": 0.8634,
+      "step": 170
+    },
+    {
+      "epoch": 1.6601941747572817,
+      "grad_norm": 0.5714628211852456,
+      "learning_rate": 4.850388495538423e-06,
+      "loss": 0.7612,
+      "step": 171
+    },
+    {
+      "epoch": 1.6699029126213594,
+      "grad_norm": 0.6093620351916359,
+      "learning_rate": 4.84865146115496e-06,
+      "loss": 0.6787,
+      "step": 172
+    },
+    {
+      "epoch": 1.679611650485437,
+      "grad_norm": 0.6119153926928719,
+      "learning_rate": 4.846904715833066e-06,
+      "loss": 0.6522,
+      "step": 173
+    },
+    {
+      "epoch": 1.6893203883495147,
+      "grad_norm": 1.4628803952904057,
+      "learning_rate": 4.8451482667949836e-06,
+      "loss": 0.9146,
+      "step": 174
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.6382465121072494,
+      "learning_rate": 4.843382121303082e-06,
+      "loss": 0.7933,
+      "step": 175
+    },
+    {
+      "epoch": 1.70873786407767,
+      "grad_norm": 0.6449697173083225,
+      "learning_rate": 4.841606286659819e-06,
+      "loss": 0.7666,
+      "step": 176
+    },
+    {
+      "epoch": 1.7184466019417477,
+      "grad_norm": 0.5895577438347263,
+      "learning_rate": 4.839820770207714e-06,
+      "loss": 0.6859,
+      "step": 177
+    },
+    {
+      "epoch": 1.7281553398058254,
+      "grad_norm": 0.6104244057063786,
+      "learning_rate": 4.8380255793293195e-06,
+      "loss": 0.691,
+      "step": 178
+    },
+    {
+      "epoch": 1.737864077669903,
+      "grad_norm": 0.6257950437898444,
+      "learning_rate": 4.8362207214471864e-06,
+      "loss": 0.667,
+      "step": 179
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 0.6404597691271853,
+      "learning_rate": 4.83440620402384e-06,
+      "loss": 0.7931,
+      "step": 180
+    },
+    {
+      "epoch": 1.7572815533980584,
+      "grad_norm": 0.554547749031446,
+      "learning_rate": 4.832582034561738e-06,
+      "loss": 0.7692,
+      "step": 181
+    },
+    {
+      "epoch": 1.766990291262136,
+      "grad_norm": 0.8379042242449998,
+      "learning_rate": 4.830748220603251e-06,
+      "loss": 1.0003,
+      "step": 182
+    },
+    {
+      "epoch": 1.7766990291262137,
+      "grad_norm": 1.371164550147673,
+      "learning_rate": 4.828904769730628e-06,
+      "loss": 1.1004,
+      "step": 183
+    },
+    {
+      "epoch": 1.7864077669902914,
+      "grad_norm": 0.663561268446458,
+      "learning_rate": 4.827051689565958e-06,
+      "loss": 0.7922,
+      "step": 184
+    },
+    {
+      "epoch": 1.796116504854369,
+      "grad_norm": 0.6107817143669849,
+      "learning_rate": 4.825188987771149e-06,
+      "loss": 0.9067,
+      "step": 185
+    },
+    {
+      "epoch": 1.8058252427184467,
+      "grad_norm": 0.5868092910997027,
+      "learning_rate": 4.82331667204789e-06,
+      "loss": 0.5892,
+      "step": 186
+    },
+    {
+      "epoch": 1.8155339805825244,
+      "grad_norm": 0.6461919671595425,
+      "learning_rate": 4.821434750137619e-06,
+      "loss": 0.725,
+      "step": 187
+    },
+    {
+      "epoch": 1.825242718446602,
+      "grad_norm": 0.6489841825930465,
+      "learning_rate": 4.819543229821494e-06,
+      "loss": 0.8711,
+      "step": 188
+    },
+    {
+      "epoch": 1.8349514563106797,
+      "grad_norm": 0.5584447600264582,
+      "learning_rate": 4.8176421189203605e-06,
+      "loss": 0.6982,
+      "step": 189
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.5659976253732854,
+      "learning_rate": 4.815731425294716e-06,
+      "loss": 0.7654,
+      "step": 190
+    },
+    {
+      "epoch": 1.854368932038835,
+      "grad_norm": 0.7825930846614281,
+      "learning_rate": 4.813811156844681e-06,
+      "loss": 0.9132,
+      "step": 191
+    },
+    {
+      "epoch": 1.8640776699029127,
+      "grad_norm": 0.662682846727773,
+      "learning_rate": 4.811881321509964e-06,
+      "loss": 0.849,
+      "step": 192
+    },
+    {
+      "epoch": 1.8737864077669903,
+      "grad_norm": 0.610458993645023,
+      "learning_rate": 4.809941927269829e-06,
+      "loss": 1.0337,
+      "step": 193
+    },
+    {
+      "epoch": 1.883495145631068,
+      "grad_norm": 0.6547992835486613,
+      "learning_rate": 4.807992982143064e-06,
+      "loss": 0.7589,
+      "step": 194
+    },
+    {
+      "epoch": 1.8932038834951457,
+      "grad_norm": 0.7542898472365844,
+      "learning_rate": 4.806034494187949e-06,
+      "loss": 1.0626,
+      "step": 195
+    },
+    {
+      "epoch": 1.9029126213592233,
+      "grad_norm": 0.6344494281728731,
+      "learning_rate": 4.804066471502216e-06,
+      "loss": 0.8039,
+      "step": 196
+    },
+    {
+      "epoch": 1.912621359223301,
+      "grad_norm": 0.6136658573521839,
+      "learning_rate": 4.802088922223024e-06,
+      "loss": 0.6425,
+      "step": 197
+    },
+    {
+      "epoch": 1.9223300970873787,
+      "grad_norm": 0.6483657220529678,
+      "learning_rate": 4.80010185452692e-06,
+      "loss": 0.8784,
+      "step": 198
+    },
+    {
+      "epoch": 1.9320388349514563,
+      "grad_norm": 0.5420123808990978,
+      "learning_rate": 4.798105276629806e-06,
+      "loss": 0.6578,
+      "step": 199
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.6687950517102932,
+      "learning_rate": 4.796099196786908e-06,
+      "loss": 0.785,
+      "step": 200
+    },
+    {
+      "epoch": 1.9514563106796117,
+      "grad_norm": 0.6178617434820718,
+      "learning_rate": 4.794083623292737e-06,
+      "loss": 0.9773,
+      "step": 201
+    },
+    {
+      "epoch": 1.9611650485436893,
+      "grad_norm": 0.6711763346625357,
+      "learning_rate": 4.792058564481058e-06,
+      "loss": 0.7296,
+      "step": 202
+    },
+    {
+      "epoch": 1.970873786407767,
+      "grad_norm": 0.6084143889062773,
+      "learning_rate": 4.7900240287248554e-06,
+      "loss": 0.8921,
+      "step": 203
+    },
+    {
+      "epoch": 1.9805825242718447,
+      "grad_norm": 0.7549301985138527,
+      "learning_rate": 4.7879800244362975e-06,
+      "loss": 0.9962,
+      "step": 204
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.6828205600145024,
+      "learning_rate": 4.785926560066703e-06,
+      "loss": 0.9536,
+      "step": 205
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7048432912651719,
+      "learning_rate": 4.783863644106502e-06,
+      "loss": 0.91,
+      "step": 206
+    },
+    {
+      "epoch": 2.0097087378640777,
+      "grad_norm": 0.5925003641609397,
+      "learning_rate": 4.781791285085209e-06,
+      "loss": 0.7574,
+      "step": 207
+    },
+    {
+      "epoch": 2.0194174757281553,
+      "grad_norm": 0.6183118556276791,
+      "learning_rate": 4.779709491571378e-06,
+      "loss": 0.7517,
+      "step": 208
+    },
+    {
+      "epoch": 2.029126213592233,
+      "grad_norm": 0.5513269000902425,
+      "learning_rate": 4.777618272172573e-06,
+      "loss": 0.8483,
+      "step": 209
+    },
+    {
+      "epoch": 2.0388349514563107,
+      "grad_norm": 0.6449371069347959,
+      "learning_rate": 4.775517635535332e-06,
+      "loss": 0.9983,
+      "step": 210
+    },
+    {
+      "epoch": 2.0485436893203883,
+      "grad_norm": 0.6510146500554277,
+      "learning_rate": 4.77340759034513e-06,
+      "loss": 0.943,
+      "step": 211
+    },
+    {
+      "epoch": 2.058252427184466,
+      "grad_norm": 0.5516833176176236,
+      "learning_rate": 4.771288145326343e-06,
+      "loss": 0.8697,
+      "step": 212
+    },
+    {
+      "epoch": 2.0679611650485437,
+      "grad_norm": 0.5372833778405927,
+      "learning_rate": 4.769159309242213e-06,
+      "loss": 0.6181,
+      "step": 213
+    },
+    {
+      "epoch": 2.0776699029126213,
+      "grad_norm": 0.6989081916523086,
+      "learning_rate": 4.767021090894809e-06,
+      "loss": 0.9862,
+      "step": 214
+    },
+    {
+      "epoch": 2.087378640776699,
+      "grad_norm": 0.5906461946301448,
+      "learning_rate": 4.764873499124997e-06,
+      "loss": 0.9271,
+      "step": 215
+    },
+    {
+      "epoch": 2.0970873786407767,
+      "grad_norm": 0.546362671281846,
+      "learning_rate": 4.762716542812395e-06,
+      "loss": 0.7597,
+      "step": 216
+    },
+    {
+      "epoch": 2.1067961165048543,
+      "grad_norm": 0.5766822546336634,
+      "learning_rate": 4.7605502308753415e-06,
+      "loss": 0.7774,
+      "step": 217
+    },
+    {
+      "epoch": 2.116504854368932,
+      "grad_norm": 0.6388672043671925,
+      "learning_rate": 4.758374572270859e-06,
+      "loss": 0.9155,
+      "step": 218
+    },
+    {
+      "epoch": 2.1262135922330097,
+      "grad_norm": 0.5728853545011949,
+      "learning_rate": 4.756189575994614e-06,
+      "loss": 0.6017,
+      "step": 219
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.5366888703406183,
+      "learning_rate": 4.753995251080884e-06,
+      "loss": 0.7425,
+      "step": 220
+    },
+    {
+      "epoch": 2.145631067961165,
+      "grad_norm": 1.3209362433112186,
+      "learning_rate": 4.7517916066025126e-06,
+      "loss": 0.6868,
+      "step": 221
+    },
+    {
+      "epoch": 2.1553398058252426,
+      "grad_norm": 0.6484738703855876,
+      "learning_rate": 4.7495786516708806e-06,
+      "loss": 0.9286,
+      "step": 222
+    },
+    {
+      "epoch": 2.1650485436893203,
+      "grad_norm": 0.6262395253661787,
+      "learning_rate": 4.747356395435865e-06,
+      "loss": 0.6069,
+      "step": 223
+    },
+    {
+      "epoch": 2.174757281553398,
+      "grad_norm": 0.6311571664234611,
+      "learning_rate": 4.745124847085799e-06,
+      "loss": 0.7718,
+      "step": 224
+    },
+    {
+      "epoch": 2.1844660194174756,
+      "grad_norm": 0.613078099794739,
+      "learning_rate": 4.742884015847436e-06,
+      "loss": 0.805,
+      "step": 225
+    },
+    {
+      "epoch": 2.1941747572815533,
+      "grad_norm": 0.5854928160805744,
+      "learning_rate": 4.740633910985911e-06,
+      "loss": 0.8084,
+      "step": 226
+    },
+    {
+      "epoch": 2.203883495145631,
+      "grad_norm": 0.5976089299186353,
+      "learning_rate": 4.738374541804704e-06,
+      "loss": 0.6863,
+      "step": 227
+    },
+    {
+      "epoch": 2.2135922330097086,
+      "grad_norm": 0.6014124521376234,
+      "learning_rate": 4.7361059176456e-06,
+      "loss": 0.7472,
+      "step": 228
+    },
+    {
+      "epoch": 2.2233009708737863,
+      "grad_norm": 0.5685151477620302,
+      "learning_rate": 4.733828047888647e-06,
+      "loss": 0.8648,
+      "step": 229
+    },
+    {
+      "epoch": 2.233009708737864,
+      "grad_norm": 0.5695477126283384,
+      "learning_rate": 4.731540941952126e-06,
+      "loss": 0.5924,
+      "step": 230
+    },
+    {
+      "epoch": 2.2427184466019416,
+      "grad_norm": 0.5651868559249316,
+      "learning_rate": 4.7292446092925016e-06,
+      "loss": 0.6548,
+      "step": 231
+    },
+    {
+      "epoch": 2.2524271844660193,
+      "grad_norm": 0.5710356100550151,
+      "learning_rate": 4.726939059404392e-06,
+      "loss": 0.6689,
+      "step": 232
+    },
+    {
+      "epoch": 2.262135922330097,
+      "grad_norm": 0.5613313610051237,
+      "learning_rate": 4.724624301820524e-06,
+      "loss": 0.673,
+      "step": 233
+    },
+    {
+      "epoch": 2.2718446601941746,
+      "grad_norm": 0.7004250430825069,
+      "learning_rate": 4.722300346111695e-06,
+      "loss": 0.8458,
+      "step": 234
+    },
+    {
+      "epoch": 2.2815533980582523,
+      "grad_norm": 0.6216105948838587,
+      "learning_rate": 4.719967201886734e-06,
+      "loss": 0.7305,
+      "step": 235
+    },
+    {
+      "epoch": 2.29126213592233,
+      "grad_norm": 0.6326118950665,
+      "learning_rate": 4.717624878792461e-06,
+      "loss": 0.9773,
+      "step": 236
+    },
+    {
+      "epoch": 2.3009708737864076,
+      "grad_norm": 0.7022071993385745,
+      "learning_rate": 4.715273386513651e-06,
+      "loss": 0.9996,
+      "step": 237
+    },
+    {
+      "epoch": 2.3106796116504853,
+      "grad_norm": 0.5723457282915828,
+      "learning_rate": 4.712912734772988e-06,
+      "loss": 0.7054,
+      "step": 238
+    },
+    {
+      "epoch": 2.320388349514563,
+      "grad_norm": 0.6945599167433392,
+      "learning_rate": 4.710542933331025e-06,
+      "loss": 0.8439,
+      "step": 239
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.6298659607215348,
+      "learning_rate": 4.708163991986152e-06,
+      "loss": 0.9141,
+      "step": 240
+    },
+    {
+      "epoch": 2.3398058252427183,
+      "grad_norm": 0.5466796831126433,
+      "learning_rate": 4.705775920574546e-06,
+      "loss": 0.8445,
+      "step": 241
+    },
+    {
+      "epoch": 2.349514563106796,
+      "grad_norm": 0.8906548403954916,
+      "learning_rate": 4.703378728970134e-06,
+      "loss": 0.6775,
+      "step": 242
+    },
+    {
+      "epoch": 2.3592233009708736,
+      "grad_norm": 0.5453159831751623,
+      "learning_rate": 4.700972427084551e-06,
+      "loss": 0.8023,
+      "step": 243
+    },
+    {
+      "epoch": 2.3689320388349513,
+      "grad_norm": 0.6086212250840284,
+      "learning_rate": 4.698557024867105e-06,
+      "loss": 0.6731,
+      "step": 244
+    },
+    {
+      "epoch": 2.378640776699029,
+      "grad_norm": 0.5633438402824648,
+      "learning_rate": 4.696132532304727e-06,
+      "loss": 0.8508,
+      "step": 245
+    },
+    {
+      "epoch": 2.3883495145631066,
+      "grad_norm": 0.5713702706197664,
+      "learning_rate": 4.693698959421935e-06,
+      "loss": 0.7985,
+      "step": 246
+    },
+    {
+      "epoch": 2.3980582524271843,
+      "grad_norm": 0.7944485374871474,
+      "learning_rate": 4.691256316280789e-06,
+      "loss": 1.0714,
+      "step": 247
+    },
+    {
+      "epoch": 2.407766990291262,
+      "grad_norm": 0.7146714273823804,
+      "learning_rate": 4.688804612980855e-06,
+      "loss": 0.7486,
+      "step": 248
+    },
+    {
+      "epoch": 2.4174757281553396,
+      "grad_norm": 0.5620249512602807,
+      "learning_rate": 4.686343859659158e-06,
+      "loss": 0.9298,
+      "step": 249
+    },
+    {
+      "epoch": 2.4271844660194173,
+      "grad_norm": 0.573615588141239,
+      "learning_rate": 4.683874066490143e-06,
+      "loss": 0.7426,
+      "step": 250
+    },
+    {
+      "epoch": 2.436893203883495,
+      "grad_norm": 0.5949978059155007,
+      "learning_rate": 4.681395243685631e-06,
+      "loss": 0.7588,
+      "step": 251
+    },
+    {
+      "epoch": 2.4466019417475726,
+      "grad_norm": 0.5548882698312756,
+      "learning_rate": 4.67890740149478e-06,
+      "loss": 0.6808,
+      "step": 252
+    },
+    {
+      "epoch": 2.4563106796116507,
+      "grad_norm": 0.5998026064624129,
+      "learning_rate": 4.676410550204036e-06,
+      "loss": 0.9237,
+      "step": 253
+    },
+    {
+      "epoch": 2.466019417475728,
+      "grad_norm": 0.5343256980470983,
+      "learning_rate": 4.673904700137098e-06,
+      "loss": 0.607,
+      "step": 254
+    },
+    {
+      "epoch": 2.475728155339806,
+      "grad_norm": 0.5129919168354066,
+      "learning_rate": 4.671389861654873e-06,
+      "loss": 0.6515,
+      "step": 255
+    },
+    {
+      "epoch": 2.4854368932038833,
+      "grad_norm": 0.5256394449847143,
+      "learning_rate": 4.668866045155428e-06,
+      "loss": 0.7839,
+      "step": 256
+    },
+    {
+      "epoch": 2.4951456310679614,
+      "grad_norm": 0.6664881549664314,
+      "learning_rate": 4.666333261073956e-06,
+      "loss": 0.8259,
+      "step": 257
+    },
+    {
+      "epoch": 2.5048543689320386,
+      "grad_norm": 0.5878066563376306,
+      "learning_rate": 4.6637915198827265e-06,
+      "loss": 0.8741,
+      "step": 258
+    },
+    {
+      "epoch": 2.5145631067961167,
+      "grad_norm": 0.5762944457977316,
+      "learning_rate": 4.661240832091042e-06,
+      "loss": 0.719,
+      "step": 259
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.635760206889188,
+      "learning_rate": 4.658681208245198e-06,
+      "loss": 0.8797,
+      "step": 260
+    },
+    {
+      "epoch": 2.533980582524272,
+      "grad_norm": 0.5296771268285247,
+      "learning_rate": 4.65611265892844e-06,
+      "loss": 0.7687,
+      "step": 261
+    },
+    {
+      "epoch": 2.5436893203883493,
+      "grad_norm": 0.5975151352963404,
+      "learning_rate": 4.653535194760912e-06,
+      "loss": 0.7371,
+      "step": 262
+    },
+    {
+      "epoch": 2.5533980582524274,
+      "grad_norm": 0.5669500366203135,
+      "learning_rate": 4.650948826399624e-06,
+      "loss": 0.5689,
+      "step": 263
+    },
+    {
+      "epoch": 2.5631067961165046,
+      "grad_norm": 0.5800778206254651,
+      "learning_rate": 4.648353564538397e-06,
+      "loss": 0.7246,
+      "step": 264
+    },
+    {
+      "epoch": 2.5728155339805827,
+      "grad_norm": 0.6321546100365262,
+      "learning_rate": 4.645749419907829e-06,
+      "loss": 0.7167,
+      "step": 265
+    },
+    {
+      "epoch": 2.58252427184466,
+      "grad_norm": 0.5319661367490109,
+      "learning_rate": 4.64313640327524e-06,
+      "loss": 0.5765,
+      "step": 266
+    },
+    {
+      "epoch": 2.592233009708738,
+      "grad_norm": 0.6013985476971264,
+      "learning_rate": 4.640514525444637e-06,
+      "loss": 0.8056,
+      "step": 267
+    },
+    {
+      "epoch": 2.6019417475728153,
+      "grad_norm": 0.6212321355444684,
+      "learning_rate": 4.637883797256663e-06,
+      "loss": 1.0431,
+      "step": 268
+    },
+    {
+      "epoch": 2.6116504854368934,
+      "grad_norm": 0.5534158811339417,
+      "learning_rate": 4.635244229588558e-06,
+      "loss": 0.7036,
+      "step": 269
+    },
+    {
+      "epoch": 2.6213592233009706,
+      "grad_norm": 0.6415745872801252,
+      "learning_rate": 4.632595833354105e-06,
+      "loss": 0.8674,
+      "step": 270
+    },
+    {
+      "epoch": 2.6310679611650487,
+      "grad_norm": 0.6285131287505951,
+      "learning_rate": 4.629938619503593e-06,
+      "loss": 0.7088,
+      "step": 271
+    },
+    {
+      "epoch": 2.6407766990291264,
+      "grad_norm": 0.5291268994493087,
+      "learning_rate": 4.627272599023772e-06,
+      "loss": 0.8775,
+      "step": 272
+    },
+    {
+      "epoch": 2.650485436893204,
+      "grad_norm": 0.5852978169428409,
+      "learning_rate": 4.6245977829378e-06,
+      "loss": 1.0301,
+      "step": 273
+    },
+    {
+      "epoch": 2.6601941747572817,
+      "grad_norm": 0.5619590148291557,
+      "learning_rate": 4.6219141823052035e-06,
+      "loss": 0.655,
+      "step": 274
+    },
+    {
+      "epoch": 2.6699029126213594,
+      "grad_norm": 0.586715896019821,
+      "learning_rate": 4.619221808221833e-06,
+      "loss": 0.9604,
+      "step": 275
+    },
+    {
+      "epoch": 2.679611650485437,
+      "grad_norm": 0.5855513689539318,
+      "learning_rate": 4.616520671819812e-06,
+      "loss": 0.6424,
+      "step": 276
+    },
+    {
+      "epoch": 2.6893203883495147,
+      "grad_norm": 0.647039460728171,
+      "learning_rate": 4.613810784267492e-06,
+      "loss": 0.8329,
+      "step": 277
+    },
+    {
+      "epoch": 2.6990291262135924,
+      "grad_norm": 0.5154279265816364,
+      "learning_rate": 4.61109215676941e-06,
+      "loss": 0.588,
+      "step": 278
+    },
+    {
+      "epoch": 2.70873786407767,
+      "grad_norm": 0.5832083686660431,
+      "learning_rate": 4.608364800566241e-06,
+      "loss": 0.7689,
+      "step": 279
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.5884877738799021,
+      "learning_rate": 4.605628726934747e-06,
+      "loss": 0.8834,
+      "step": 280
+    },
+    {
+      "epoch": 2.7281553398058254,
+      "grad_norm": 0.5126503459961328,
+      "learning_rate": 4.602883947187738e-06,
+      "loss": 0.7756,
+      "step": 281
+    },
+    {
+      "epoch": 2.737864077669903,
+      "grad_norm": 0.5509578989929416,
+      "learning_rate": 4.600130472674017e-06,
+      "loss": 0.63,
+      "step": 282
+    },
+    {
+      "epoch": 2.7475728155339807,
+      "grad_norm": 0.5937802430842546,
+      "learning_rate": 4.5973683147783405e-06,
+      "loss": 0.9692,
+      "step": 283
+    },
+    {
+      "epoch": 2.7572815533980584,
+      "grad_norm": 0.8437616641968025,
+      "learning_rate": 4.594597484921365e-06,
+      "loss": 0.7999,
+      "step": 284
+    },
+    {
+      "epoch": 2.766990291262136,
+      "grad_norm": 0.5815733690670376,
+      "learning_rate": 4.5918179945596055e-06,
+      "loss": 0.7197,
+      "step": 285
+    },
+    {
+      "epoch": 2.7766990291262137,
+      "grad_norm": 0.6766231348468958,
+      "learning_rate": 4.589029855185384e-06,
+      "loss": 0.8009,
+      "step": 286
+    },
+    {
+      "epoch": 2.7864077669902914,
+      "grad_norm": 0.6016607485166134,
+      "learning_rate": 4.586233078326785e-06,
+      "loss": 1.055,
+      "step": 287
+    },
+    {
+      "epoch": 2.796116504854369,
+      "grad_norm": 0.5385528216529567,
+      "learning_rate": 4.583427675547602e-06,
+      "loss": 0.8221,
+      "step": 288
+    },
+    {
+      "epoch": 2.8058252427184467,
+      "grad_norm": 0.5993566648276473,
+      "learning_rate": 4.580613658447301e-06,
+      "loss": 0.938,
+      "step": 289
+    },
+    {
+      "epoch": 2.8155339805825244,
+      "grad_norm": 0.5912484848249645,
+      "learning_rate": 4.577791038660959e-06,
+      "loss": 0.6534,
+      "step": 290
+    },
+    {
+      "epoch": 2.825242718446602,
+      "grad_norm": 0.5558723739620934,
+      "learning_rate": 4.574959827859226e-06,
+      "loss": 0.6328,
+      "step": 291
+    },
+    {
+      "epoch": 2.8349514563106797,
+      "grad_norm": 0.5215220174359476,
+      "learning_rate": 4.572120037748273e-06,
+      "loss": 0.6078,
+      "step": 292
+    },
+    {
+      "epoch": 2.8446601941747574,
+      "grad_norm": 0.5578352792075171,
+      "learning_rate": 4.5692716800697415e-06,
+      "loss": 0.7983,
+      "step": 293
+    },
+    {
+      "epoch": 2.854368932038835,
+      "grad_norm": 0.5631763967068151,
+      "learning_rate": 4.566414766600698e-06,
+      "loss": 0.7932,
+      "step": 294
+    },
+    {
+      "epoch": 2.8640776699029127,
+      "grad_norm": 0.5467170630989897,
+      "learning_rate": 4.563549309153589e-06,
+      "loss": 0.6513,
+      "step": 295
+    },
+    {
+      "epoch": 2.8737864077669903,
+      "grad_norm": 0.6087442326908157,
+      "learning_rate": 4.56067531957618e-06,
+      "loss": 0.7059,
+      "step": 296
+    },
+    {
+      "epoch": 2.883495145631068,
+      "grad_norm": 0.6345964624606057,
+      "learning_rate": 4.557792809751519e-06,
+      "loss": 0.8497,
+      "step": 297
+    },
+    {
+      "epoch": 2.8932038834951457,
+      "grad_norm": 0.6421474826199707,
+      "learning_rate": 4.554901791597883e-06,
+      "loss": 0.754,
+      "step": 298
+    },
+    {
+      "epoch": 2.9029126213592233,
+      "grad_norm": 0.5611662286254485,
+      "learning_rate": 4.552002277068725e-06,
+      "loss": 0.7587,
+      "step": 299
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.6356794525355608,
+      "learning_rate": 4.549094278152631e-06,
+      "loss": 0.8163,
+      "step": 300
+    },
+    {
+      "epoch": 2.9223300970873787,
+      "grad_norm": 0.6336849057165008,
+      "learning_rate": 4.546177806873266e-06,
+      "loss": 0.682,
+      "step": 301
+    },
+    {
+      "epoch": 2.9320388349514563,
+      "grad_norm": 0.6059295762410952,
+      "learning_rate": 4.543252875289326e-06,
+      "loss": 0.6621,
+      "step": 302
+    },
+    {
+      "epoch": 2.941747572815534,
+      "grad_norm": 0.5887024077749718,
+      "learning_rate": 4.540319495494486e-06,
+      "loss": 0.7904,
+      "step": 303
+    },
+    {
+      "epoch": 2.9514563106796117,
+      "grad_norm": 0.6053171571638954,
+      "learning_rate": 4.537377679617353e-06,
+      "loss": 0.8159,
+      "step": 304
+    },
+    {
+      "epoch": 2.9611650485436893,
+      "grad_norm": 0.6162643989939205,
+      "learning_rate": 4.534427439821416e-06,
+      "loss": 0.7054,
+      "step": 305
+    },
+    {
+      "epoch": 2.970873786407767,
+      "grad_norm": 0.5653687306702669,
+      "learning_rate": 4.531468788304992e-06,
+      "loss": 0.6473,
+      "step": 306
+    },
+    {
+      "epoch": 2.9805825242718447,
+      "grad_norm": 0.6147243558106896,
+      "learning_rate": 4.5285017373011784e-06,
+      "loss": 0.6811,
+      "step": 307
+    },
+    {
+      "epoch": 2.9902912621359223,
+      "grad_norm": 0.5023058479913145,
+      "learning_rate": 4.5255262990778024e-06,
+      "loss": 0.5666,
+      "step": 308
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5339774541368152,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.7024,
+      "step": 309
+    },
+    {
+      "epoch": 3.0097087378640777,
+      "grad_norm": 0.5871214604141014,
+      "learning_rate": 4.519550310217013e-06,
+      "loss": 0.792,
+      "step": 310
+    },
+    {
+      "epoch": 3.0194174757281553,
+      "grad_norm": 0.550905273830553,
+      "learning_rate": 4.516549784288442e-06,
+      "loss": 0.6801,
+      "step": 311
+    },
+    {
+      "epoch": 3.029126213592233,
+      "grad_norm": 0.5295468316460928,
+      "learning_rate": 4.513540920557892e-06,
+      "loss": 0.6589,
+      "step": 312
+    },
+    {
+      "epoch": 3.0388349514563107,
+      "grad_norm": 0.5818378095787248,
+      "learning_rate": 4.510523731466072e-06,
+      "loss": 0.9006,
+      "step": 313
+    },
+    {
+      "epoch": 3.0485436893203883,
+      "grad_norm": 0.53839185417949,
+      "learning_rate": 4.507498229488116e-06,
+      "loss": 0.5935,
+      "step": 314
+    },
+    {
+      "epoch": 3.058252427184466,
+      "grad_norm": 0.5162101468182391,
+      "learning_rate": 4.504464427133527e-06,
+      "loss": 0.6615,
+      "step": 315
+    },
+    {
+      "epoch": 3.0679611650485437,
+      "grad_norm": 0.5627252995451077,
+      "learning_rate": 4.501422336946126e-06,
+      "loss": 0.6778,
+      "step": 316
+    },
+    {
+      "epoch": 3.0776699029126213,
+      "grad_norm": 0.5499924181430055,
+      "learning_rate": 4.498371971504005e-06,
+      "loss": 0.6824,
+      "step": 317
+    },
+    {
+      "epoch": 3.087378640776699,
+      "grad_norm": 0.5313000147551927,
+      "learning_rate": 4.49531334341947e-06,
+      "loss": 0.6953,
+      "step": 318
+    },
+    {
+      "epoch": 3.0970873786407767,
+      "grad_norm": 0.5053678892953669,
+      "learning_rate": 4.49224646533899e-06,
+      "loss": 0.624,
+      "step": 319
+    },
+    {
+      "epoch": 3.1067961165048543,
+      "grad_norm": 0.5607765091857444,
+      "learning_rate": 4.489171349943144e-06,
+      "loss": 0.6887,
+      "step": 320
+    },
+    {
+      "epoch": 3.116504854368932,
+      "grad_norm": 0.49971275746543925,
+      "learning_rate": 4.486088009946575e-06,
+      "loss": 0.599,
+      "step": 321
+    },
+    {
+      "epoch": 3.1262135922330097,
+      "grad_norm": 0.5463448565404533,
+      "learning_rate": 4.482996458097926e-06,
+      "loss": 0.7412,
+      "step": 322
+    },
+    {
+      "epoch": 3.1359223300970873,
+      "grad_norm": 0.5577175615372468,
+      "learning_rate": 4.479896707179796e-06,
+      "loss": 0.9504,
+      "step": 323
+    },
+    {
+      "epoch": 3.145631067961165,
+      "grad_norm": 0.739795464222875,
+      "learning_rate": 4.476788770008685e-06,
+      "loss": 0.9296,
+      "step": 324
+    },
+    {
+      "epoch": 3.1553398058252426,
+      "grad_norm": 0.5860377779118467,
+      "learning_rate": 4.473672659434941e-06,
+      "loss": 0.8433,
+      "step": 325
+    },
+    {
+      "epoch": 3.1650485436893203,
+      "grad_norm": 0.5234329633422101,
+      "learning_rate": 4.470548388342704e-06,
+      "loss": 0.7058,
+      "step": 326
+    },
+    {
+      "epoch": 3.174757281553398,
+      "grad_norm": 0.5307383652263797,
+      "learning_rate": 4.467415969649858e-06,
+      "loss": 0.7462,
+      "step": 327
+    },
+    {
+      "epoch": 3.1844660194174756,
+      "grad_norm": 0.5283773909342946,
+      "learning_rate": 4.464275416307973e-06,
+      "loss": 0.7497,
+      "step": 328
+    },
+    {
+      "epoch": 3.1941747572815533,
+      "grad_norm": 0.5706683381161813,
+      "learning_rate": 4.461126741302253e-06,
+      "loss": 0.7677,
+      "step": 329
+    },
+    {
+      "epoch": 3.203883495145631,
+      "grad_norm": 0.568157264873081,
+      "learning_rate": 4.457969957651485e-06,
+      "loss": 0.6759,
+      "step": 330
+    },
+    {
+      "epoch": 3.2135922330097086,
+      "grad_norm": 0.5665376751141662,
+      "learning_rate": 4.454805078407979e-06,
+      "loss": 0.9144,
+      "step": 331
+    },
+    {
+      "epoch": 3.2233009708737863,
+      "grad_norm": 0.5146029291558777,
+      "learning_rate": 4.451632116657521e-06,
+      "loss": 0.6786,
+      "step": 332
+    },
+    {
+      "epoch": 3.233009708737864,
+      "grad_norm": 0.4670736553738834,
+      "learning_rate": 4.448451085519314e-06,
+      "loss": 0.5646,
+      "step": 333
+    },
+    {
+      "epoch": 3.2427184466019416,
+      "grad_norm": 0.5983555089379993,
+      "learning_rate": 4.445261998145927e-06,
+      "loss": 0.7486,
+      "step": 334
+    },
+    {
+      "epoch": 3.2524271844660193,
+      "grad_norm": 0.5247722963016707,
+      "learning_rate": 4.442064867723236e-06,
+      "loss": 0.7228,
+      "step": 335
+    },
+    {
+      "epoch": 3.262135922330097,
+      "grad_norm": 0.5830371706199331,
+      "learning_rate": 4.438859707470376e-06,
+      "loss": 0.7545,
+      "step": 336
+    },
+    {
+      "epoch": 3.2718446601941746,
+      "grad_norm": 0.569317200535919,
+      "learning_rate": 4.435646530639679e-06,
+      "loss": 0.6446,
+      "step": 337
+    },
+    {
+      "epoch": 3.2815533980582523,
+      "grad_norm": 0.5229847762890999,
+      "learning_rate": 4.432425350516627e-06,
+      "loss": 0.7016,
+      "step": 338
+    },
+    {
+      "epoch": 3.29126213592233,
+      "grad_norm": 0.5556853198137727,
+      "learning_rate": 4.42919618041979e-06,
+      "loss": 0.581,
+      "step": 339
+    },
+    {
+      "epoch": 3.3009708737864076,
+      "grad_norm": 0.5504701249347626,
+      "learning_rate": 4.425959033700776e-06,
+      "loss": 0.4669,
+      "step": 340
+    },
+    {
+      "epoch": 3.3106796116504853,
+      "grad_norm": 0.5733436063811748,
+      "learning_rate": 4.422713923744174e-06,
+      "loss": 0.7609,
+      "step": 341
+    },
+    {
+      "epoch": 3.320388349514563,
+      "grad_norm": 0.5712264087934034,
+      "learning_rate": 4.419460863967496e-06,
+      "loss": 0.796,
+      "step": 342
+    },
+    {
+      "epoch": 3.3300970873786406,
+      "grad_norm": 0.5298396914849735,
+      "learning_rate": 4.416199867821126e-06,
+      "loss": 0.7347,
+      "step": 343
+    },
+    {
+      "epoch": 3.3398058252427183,
+      "grad_norm": 0.6167638285377818,
+      "learning_rate": 4.412930948788263e-06,
+      "loss": 0.7213,
+      "step": 344
+    },
+    {
+      "epoch": 3.349514563106796,
+      "grad_norm": 0.6769548528256155,
+      "learning_rate": 4.409654120384863e-06,
+      "loss": 0.9705,
+      "step": 345
+    },
+    {
+      "epoch": 3.3592233009708736,
+      "grad_norm": 0.5558234569754197,
+      "learning_rate": 4.406369396159585e-06,
+      "loss": 0.866,
+      "step": 346
+    },
+    {
+      "epoch": 3.3689320388349513,
+      "grad_norm": 0.5683210417700995,
+      "learning_rate": 4.403076789693735e-06,
+      "loss": 0.8529,
+      "step": 347
+    },
+    {
+      "epoch": 3.378640776699029,
+      "grad_norm": 0.6004152868285685,
+      "learning_rate": 4.399776314601212e-06,
+      "loss": 0.6924,
+      "step": 348
+    },
+    {
+      "epoch": 3.3883495145631066,
+      "grad_norm": 0.5915756102782764,
+      "learning_rate": 4.396467984528445e-06,
+      "loss": 0.809,
+      "step": 349
+    },
+    {
+      "epoch": 3.3980582524271843,
+      "grad_norm": 0.5726590426952238,
+      "learning_rate": 4.393151813154345e-06,
+      "loss": 0.6593,
+      "step": 350
+    },
+    {
+      "epoch": 3.407766990291262,
+      "grad_norm": 0.5071226489196179,
+      "learning_rate": 4.3898278141902396e-06,
+      "loss": 0.6473,
+      "step": 351
+    },
+    {
+      "epoch": 3.4174757281553396,
+      "grad_norm": 0.5733646491001667,
+      "learning_rate": 4.386496001379826e-06,
+      "loss": 0.8021,
+      "step": 352
+    },
+    {
+      "epoch": 3.4271844660194173,
+      "grad_norm": 0.5616644007222712,
+      "learning_rate": 4.383156388499106e-06,
+      "loss": 0.7276,
+      "step": 353
+    },
+    {
+      "epoch": 3.436893203883495,
+      "grad_norm": 0.5270931676906139,
+      "learning_rate": 4.3798089893563335e-06,
+      "loss": 0.6511,
+      "step": 354
+    },
+    {
+      "epoch": 3.4466019417475726,
+      "grad_norm": 0.5956176073578046,
+      "learning_rate": 4.3764538177919555e-06,
+      "loss": 0.5386,
+      "step": 355
+    },
+    {
+      "epoch": 3.4563106796116507,
+      "grad_norm": 0.6087361244099403,
+      "learning_rate": 4.3730908876785574e-06,
+      "loss": 0.7574,
+      "step": 356
+    },
+    {
+      "epoch": 3.466019417475728,
+      "grad_norm": 0.6124986417248809,
+      "learning_rate": 4.3697202129208e-06,
+      "loss": 0.797,
+      "step": 357
+    },
+    {
+      "epoch": 3.475728155339806,
+      "grad_norm": 0.5233985914467231,
+      "learning_rate": 4.36634180745537e-06,
+      "loss": 0.5764,
+      "step": 358
+    },
+    {
+      "epoch": 3.4854368932038833,
+      "grad_norm": 0.5590564532881283,
+      "learning_rate": 4.3629556852509145e-06,
+      "loss": 0.9866,
+      "step": 359
+    },
+    {
+      "epoch": 3.4951456310679614,
+      "grad_norm": 0.5100762691051512,
+      "learning_rate": 4.35956186030799e-06,
+      "loss": 0.6127,
+      "step": 360
+    },
+    {
+      "epoch": 3.5048543689320386,
+      "grad_norm": 0.6118806932618559,
+      "learning_rate": 4.356160346659001e-06,
+      "loss": 0.7548,
+      "step": 361
+    },
+    {
+      "epoch": 3.5145631067961167,
+      "grad_norm": 0.6631159487479685,
+      "learning_rate": 4.3527511583681384e-06,
+      "loss": 1.2605,
+      "step": 362
+    },
+    {
+      "epoch": 3.524271844660194,
+      "grad_norm": 0.6075337136532736,
+      "learning_rate": 4.34933430953133e-06,
+      "loss": 0.7971,
+      "step": 363
+    },
+    {
+      "epoch": 3.533980582524272,
+      "grad_norm": 0.6085963816816855,
+      "learning_rate": 4.345909814276177e-06,
+      "loss": 0.8115,
+      "step": 364
+    },
+    {
+      "epoch": 3.5436893203883493,
+      "grad_norm": 0.6489049958725044,
+      "learning_rate": 4.3424776867618935e-06,
+      "loss": 0.8511,
+      "step": 365
+    },
+    {
+      "epoch": 3.5533980582524274,
+      "grad_norm": 0.5533454922999531,
+      "learning_rate": 4.339037941179253e-06,
+      "loss": 0.7739,
+      "step": 366
+    },
+    {
+      "epoch": 3.5631067961165046,
+      "grad_norm": 0.5324277334095654,
+      "learning_rate": 4.335590591750526e-06,
+      "loss": 0.6825,
+      "step": 367
+    },
+    {
+      "epoch": 3.5728155339805827,
+      "grad_norm": 0.583139527089237,
+      "learning_rate": 4.332135652729423e-06,
+      "loss": 0.6584,
+      "step": 368
+    },
+    {
+      "epoch": 3.58252427184466,
+      "grad_norm": 0.5359360883573963,
+      "learning_rate": 4.328673138401036e-06,
+      "loss": 0.7771,
+      "step": 369
+    },
+    {
+      "epoch": 3.592233009708738,
+      "grad_norm": 0.5374666185272033,
+      "learning_rate": 4.325203063081776e-06,
+      "loss": 0.4958,
+      "step": 370
+    },
+    {
+      "epoch": 3.6019417475728153,
+      "grad_norm": 0.5182456153897488,
+      "learning_rate": 4.32172544111932e-06,
+      "loss": 0.6043,
+      "step": 371
+    },
+    {
+      "epoch": 3.6116504854368934,
+      "grad_norm": 0.5463365868711481,
+      "learning_rate": 4.318240286892544e-06,
+      "loss": 0.67,
+      "step": 372
+    },
+    {
+      "epoch": 3.6213592233009706,
+      "grad_norm": 0.5832479071806481,
+      "learning_rate": 4.314747614811471e-06,
+      "loss": 0.7424,
+      "step": 373
+    },
+    {
+      "epoch": 3.6310679611650487,
+      "grad_norm": 0.5054355197048168,
+      "learning_rate": 4.3112474393172055e-06,
+      "loss": 0.5487,
+      "step": 374
+    },
+    {
+      "epoch": 3.6407766990291264,
+      "grad_norm": 0.5382871622608469,
+      "learning_rate": 4.307739774881878e-06,
+      "loss": 0.6772,
+      "step": 375
+    },
+    {
+      "epoch": 3.650485436893204,
+      "grad_norm": 0.5458477824608565,
+      "learning_rate": 4.304224636008582e-06,
+      "loss": 0.5329,
+      "step": 376
+    },
+    {
+      "epoch": 3.6601941747572817,
+      "grad_norm": 0.5435683453057218,
+      "learning_rate": 4.300702037231318e-06,
+      "loss": 0.8113,
+      "step": 377
+    },
+    {
+      "epoch": 3.6699029126213594,
+      "grad_norm": 0.5383065920679537,
+      "learning_rate": 4.297171993114927e-06,
+      "loss": 0.5284,
+      "step": 378
+    },
+    {
+      "epoch": 3.679611650485437,
+      "grad_norm": 0.5880433935588353,
+      "learning_rate": 4.2936345182550365e-06,
+      "loss": 0.8539,
+      "step": 379
+    },
+    {
+      "epoch": 3.6893203883495147,
+      "grad_norm": 0.5952587100488826,
+      "learning_rate": 4.290089627277998e-06,
+      "loss": 0.5857,
+      "step": 380
+    },
+    {
+      "epoch": 3.6990291262135924,
+      "grad_norm": 0.5286616194694634,
+      "learning_rate": 4.286537334840825e-06,
+      "loss": 0.7462,
+      "step": 381
+    },
+    {
+      "epoch": 3.70873786407767,
+      "grad_norm": 0.6011010828422011,
+      "learning_rate": 4.2829776556311355e-06,
+      "loss": 0.6581,
+      "step": 382
+    },
+    {
+      "epoch": 3.7184466019417477,
+      "grad_norm": 0.59061291418775,
+      "learning_rate": 4.279410604367088e-06,
+      "loss": 0.8607,
+      "step": 383
+    },
+    {
+      "epoch": 3.7281553398058254,
+      "grad_norm": 0.6006128694830518,
+      "learning_rate": 4.275836195797323e-06,
+      "loss": 0.8253,
+      "step": 384
+    },
+    {
+      "epoch": 3.737864077669903,
+      "grad_norm": 0.559500801329847,
+      "learning_rate": 4.2722544447008995e-06,
+      "loss": 0.8451,
+      "step": 385
+    },
+    {
+      "epoch": 3.7475728155339807,
+      "grad_norm": 0.5834501093822622,
+      "learning_rate": 4.268665365887238e-06,
+      "loss": 0.8882,
+      "step": 386
+    },
+    {
+      "epoch": 3.7572815533980584,
+      "grad_norm": 0.5043323998416107,
+      "learning_rate": 4.265068974196056e-06,
+      "loss": 0.5369,
+      "step": 387
+    },
+    {
+      "epoch": 3.766990291262136,
+      "grad_norm": 0.5727251303865297,
+      "learning_rate": 4.261465284497307e-06,
+      "loss": 0.6106,
+      "step": 388
+    },
+    {
+      "epoch": 3.7766990291262137,
+      "grad_norm": 0.5372256278502759,
+      "learning_rate": 4.257854311691118e-06,
+      "loss": 0.7688,
+      "step": 389
+    },
+    {
+      "epoch": 3.7864077669902914,
+      "grad_norm": 0.53361182265545,
+      "learning_rate": 4.254236070707734e-06,
+      "loss": 0.6263,
+      "step": 390
+    },
+    {
+      "epoch": 3.796116504854369,
+      "grad_norm": 0.4761706762756821,
+      "learning_rate": 4.250610576507445e-06,
+      "loss": 0.5494,
+      "step": 391
+    },
+    {
+      "epoch": 3.8058252427184467,
+      "grad_norm": 0.5667551282882679,
+      "learning_rate": 4.246977844080537e-06,
+      "loss": 0.7841,
+      "step": 392
+    },
+    {
+      "epoch": 3.8155339805825244,
+      "grad_norm": 0.5581465960923723,
+      "learning_rate": 4.24333788844722e-06,
+      "loss": 0.727,
+      "step": 393
+    },
+    {
+      "epoch": 3.825242718446602,
+      "grad_norm": 0.641698316355312,
+      "learning_rate": 4.239690724657571e-06,
+      "loss": 0.7415,
+      "step": 394
+    },
+    {
+      "epoch": 3.8349514563106797,
+      "grad_norm": 0.5629852857617874,
+      "learning_rate": 4.236036367791471e-06,
+      "loss": 0.7685,
+      "step": 395
+    },
+    {
+      "epoch": 3.8446601941747574,
+      "grad_norm": 0.5910867032740321,
+      "learning_rate": 4.23237483295854e-06,
+      "loss": 0.797,
+      "step": 396
+    },
+    {
+      "epoch": 3.854368932038835,
+      "grad_norm": 0.5182753315022374,
+      "learning_rate": 4.228706135298081e-06,
+      "loss": 0.5786,
+      "step": 397
+    },
+    {
+      "epoch": 3.8640776699029127,
+      "grad_norm": 0.545028212220101,
+      "learning_rate": 4.225030289979006e-06,
+      "loss": 0.5823,
+      "step": 398
+    },
+    {
+      "epoch": 3.8737864077669903,
+      "grad_norm": 0.5228141064601322,
+      "learning_rate": 4.221347312199788e-06,
+      "loss": 0.5436,
+      "step": 399
+    },
+    {
+      "epoch": 3.883495145631068,
+      "grad_norm": 0.5397693619666596,
+      "learning_rate": 4.2176572171883865e-06,
+      "loss": 0.7937,
+      "step": 400
+    },
+    {
+      "epoch": 3.8932038834951457,
+      "grad_norm": 0.6886419465953945,
+      "learning_rate": 4.213960020202187e-06,
+      "loss": 0.8116,
+      "step": 401
+    },
+    {
+      "epoch": 3.9029126213592233,
+      "grad_norm": 0.5900622648932935,
+      "learning_rate": 4.2102557365279435e-06,
+      "loss": 0.917,
+      "step": 402
+    },
+    {
+      "epoch": 3.912621359223301,
+      "grad_norm": 0.544861080103798,
+      "learning_rate": 4.206544381481708e-06,
+      "loss": 0.8025,
+      "step": 403
+    },
+    {
+      "epoch": 3.9223300970873787,
+      "grad_norm": 0.518875885292913,
+      "learning_rate": 4.202825970408772e-06,
+      "loss": 0.5702,
+      "step": 404
+    },
+    {
+      "epoch": 3.9320388349514563,
+      "grad_norm": 0.656540638196741,
+      "learning_rate": 4.199100518683601e-06,
+      "loss": 0.6321,
+      "step": 405
+    },
+    {
+      "epoch": 3.941747572815534,
+      "grad_norm": 0.5191060858406762,
+      "learning_rate": 4.195368041709772e-06,
+      "loss": 0.7449,
+      "step": 406
+    },
+    {
+      "epoch": 3.9514563106796117,
+      "grad_norm": 0.4946680296221446,
+      "learning_rate": 4.191628554919907e-06,
+      "loss": 0.6497,
+      "step": 407
+    },
+    {
+      "epoch": 3.9611650485436893,
+      "grad_norm": 0.5824785363442938,
+      "learning_rate": 4.187882073775615e-06,
+      "loss": 0.7381,
+      "step": 408
+    },
+    {
+      "epoch": 3.970873786407767,
+      "grad_norm": 0.5193631938328553,
+      "learning_rate": 4.184128613767422e-06,
+      "loss": 0.6982,
+      "step": 409
+    },
+    {
+      "epoch": 3.9805825242718447,
+      "grad_norm": 0.5290930959018318,
+      "learning_rate": 4.18036819041471e-06,
+      "loss": 0.6859,
+      "step": 410
+    },
+    {
+      "epoch": 3.9902912621359223,
+      "grad_norm": 0.5272372038733293,
+      "learning_rate": 4.17660081926565e-06,
+      "loss": 0.5822,
+      "step": 411
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.5630958714803344,
+      "learning_rate": 4.172826515897146e-06,
+      "loss": 0.6599,
+      "step": 412
+    },
+    {
+      "epoch": 4.009708737864078,
+      "grad_norm": 0.5200333011009248,
+      "learning_rate": 4.169045295914757e-06,
+      "loss": 0.7889,
+      "step": 413
+    },
+    {
+      "epoch": 4.019417475728155,
+      "grad_norm": 0.5331918145560857,
+      "learning_rate": 4.165257174952647e-06,
+      "loss": 0.5669,
+      "step": 414
+    },
+    {
+      "epoch": 4.029126213592233,
+      "grad_norm": 0.4838073648864831,
+      "learning_rate": 4.161462168673508e-06,
+      "loss": 0.6616,
+      "step": 415
+    },
+    {
+      "epoch": 4.038834951456311,
+      "grad_norm": 0.5285061889577369,
+      "learning_rate": 4.157660292768502e-06,
+      "loss": 0.7826,
+      "step": 416
+    },
+    {
+      "epoch": 4.048543689320389,
+      "grad_norm": 0.47876226674574307,
+      "learning_rate": 4.1538515629571985e-06,
+      "loss": 0.6324,
+      "step": 417
+    },
+    {
+      "epoch": 4.058252427184466,
+      "grad_norm": 0.4964106466439782,
+      "learning_rate": 4.1500359949875e-06,
+      "loss": 0.7882,
+      "step": 418
+    },
+    {
+      "epoch": 4.067961165048544,
+      "grad_norm": 0.5098927521794874,
+      "learning_rate": 4.1462136046355864e-06,
+      "loss": 0.679,
+      "step": 419
+    },
+    {
+      "epoch": 4.077669902912621,
+      "grad_norm": 0.5261711579674679,
+      "learning_rate": 4.142384407705846e-06,
+      "loss": 0.9621,
+      "step": 420
+    },
+    {
+      "epoch": 4.087378640776699,
+      "grad_norm": 0.4855524472712664,
+      "learning_rate": 4.138548420030808e-06,
+      "loss": 0.5577,
+      "step": 421
+    },
+    {
+      "epoch": 4.097087378640777,
+      "grad_norm": 0.46404425362535634,
+      "learning_rate": 4.13470565747108e-06,
+      "loss": 0.727,
+      "step": 422
+    },
+    {
+      "epoch": 4.106796116504855,
+      "grad_norm": 0.5013144099880142,
+      "learning_rate": 4.130856135915282e-06,
+      "loss": 0.7378,
+      "step": 423
+    },
+    {
+      "epoch": 4.116504854368932,
+      "grad_norm": 0.5316059543260906,
+      "learning_rate": 4.126999871279982e-06,
+      "loss": 0.6906,
+      "step": 424
+    },
+    {
+      "epoch": 4.12621359223301,
+      "grad_norm": 0.4944636869938843,
+      "learning_rate": 4.123136879509626e-06,
+      "loss": 0.6641,
+      "step": 425
+    },
+    {
+      "epoch": 4.135922330097087,
+      "grad_norm": 0.5095570582691661,
+      "learning_rate": 4.119267176576475e-06,
+      "loss": 0.7491,
+      "step": 426
+    },
+    {
+      "epoch": 4.145631067961165,
+      "grad_norm": 0.54005610895594,
+      "learning_rate": 4.11539077848054e-06,
+      "loss": 0.7558,
+      "step": 427
+    },
+    {
+      "epoch": 4.155339805825243,
+      "grad_norm": 0.5923182555591768,
+      "learning_rate": 4.111507701249513e-06,
+      "loss": 0.7929,
+      "step": 428
+    },
+    {
+      "epoch": 4.165048543689321,
+      "grad_norm": 0.6116168581706061,
+      "learning_rate": 4.107617960938702e-06,
+      "loss": 0.5406,
+      "step": 429
+    },
+    {
+      "epoch": 4.174757281553398,
+      "grad_norm": 0.5210538716839132,
+      "learning_rate": 4.103721573630965e-06,
+      "loss": 0.6724,
+      "step": 430
+    },
+    {
+      "epoch": 4.184466019417476,
+      "grad_norm": 0.47356150602269126,
+      "learning_rate": 4.099818555436645e-06,
+      "loss": 0.6737,
+      "step": 431
+    },
+    {
+      "epoch": 4.194174757281553,
+      "grad_norm": 0.6377655423855296,
+      "learning_rate": 4.095908922493499e-06,
+      "loss": 1.024,
+      "step": 432
+    },
+    {
+      "epoch": 4.203883495145631,
+      "grad_norm": 0.5471085748898301,
+      "learning_rate": 4.091992690966636e-06,
+      "loss": 0.5427,
+      "step": 433
+    },
+    {
+      "epoch": 4.213592233009709,
+      "grad_norm": 0.5818794345348097,
+      "learning_rate": 4.088069877048447e-06,
+      "loss": 0.6099,
+      "step": 434
+    },
+    {
+      "epoch": 4.223300970873787,
+      "grad_norm": 0.543595019351133,
+      "learning_rate": 4.084140496958539e-06,
+      "loss": 0.556,
+      "step": 435
+    },
+    {
+      "epoch": 4.233009708737864,
+      "grad_norm": 0.5101773462629047,
+      "learning_rate": 4.080204566943668e-06,
+      "loss": 0.5016,
+      "step": 436
+    },
+    {
+      "epoch": 4.242718446601942,
+      "grad_norm": 0.574864469322262,
+      "learning_rate": 4.076262103277673e-06,
+      "loss": 0.5518,
+      "step": 437
+    },
+    {
+      "epoch": 4.252427184466019,
+      "grad_norm": 0.529410686127538,
+      "learning_rate": 4.072313122261406e-06,
+      "loss": 0.9056,
+      "step": 438
+    },
+    {
+      "epoch": 4.262135922330097,
+      "grad_norm": 0.4724977490535202,
+      "learning_rate": 4.068357640222668e-06,
+      "loss": 0.5444,
+      "step": 439
+    },
+    {
+      "epoch": 4.271844660194175,
+      "grad_norm": 0.5371234765273634,
+      "learning_rate": 4.06439567351614e-06,
+      "loss": 0.5565,
+      "step": 440
+    },
+    {
+      "epoch": 4.281553398058253,
+      "grad_norm": 0.5554052963535526,
+      "learning_rate": 4.0604272385233105e-06,
+      "loss": 0.6978,
+      "step": 441
+    },
+    {
+      "epoch": 4.29126213592233,
+      "grad_norm": 0.5738857111157796,
+      "learning_rate": 4.056452351652418e-06,
+      "loss": 0.5298,
+      "step": 442
+    },
+    {
+      "epoch": 4.300970873786408,
+      "grad_norm": 0.5183938376729829,
+      "learning_rate": 4.052471029338375e-06,
+      "loss": 0.689,
+      "step": 443
+    },
+    {
+      "epoch": 4.310679611650485,
+      "grad_norm": 0.562235068394362,
+      "learning_rate": 4.048483288042703e-06,
+      "loss": 0.5329,
+      "step": 444
+    },
+    {
+      "epoch": 4.320388349514563,
+      "grad_norm": 0.5873359410798857,
+      "learning_rate": 4.0444891442534615e-06,
+      "loss": 0.6088,
+      "step": 445
+    },
+    {
+      "epoch": 4.330097087378641,
+      "grad_norm": 0.5232886853596697,
+      "learning_rate": 4.040488614485187e-06,
+      "loss": 0.6861,
+      "step": 446
+    },
+    {
+      "epoch": 4.339805825242719,
+      "grad_norm": 0.5707575802489561,
+      "learning_rate": 4.036481715278818e-06,
+      "loss": 0.7804,
+      "step": 447
+    },
+    {
+      "epoch": 4.349514563106796,
+      "grad_norm": 0.5743158436311879,
+      "learning_rate": 4.032468463201626e-06,
+      "loss": 0.7531,
+      "step": 448
+    },
+    {
+      "epoch": 4.359223300970874,
+      "grad_norm": 0.5272870524360327,
+      "learning_rate": 4.028448874847152e-06,
+      "loss": 0.4601,
+      "step": 449
+    },
+    {
+      "epoch": 4.368932038834951,
+      "grad_norm": 0.8245819176512555,
+      "learning_rate": 4.024422966835137e-06,
+      "loss": 0.7204,
+      "step": 450
+    },
+    {
+      "epoch": 4.378640776699029,
+      "grad_norm": 0.5818583543824783,
+      "learning_rate": 4.0203907558114475e-06,
+      "loss": 0.6319,
+      "step": 451
+    },
+    {
+      "epoch": 4.388349514563107,
+      "grad_norm": 0.5916617169196222,
+      "learning_rate": 4.016352258448016e-06,
+      "loss": 0.5578,
+      "step": 452
+    },
+    {
+      "epoch": 4.398058252427185,
+      "grad_norm": 0.5247948377375482,
+      "learning_rate": 4.0123074914427635e-06,
+      "loss": 0.5817,
+      "step": 453
+    },
+    {
+      "epoch": 4.407766990291262,
+      "grad_norm": 0.4841360885356806,
+      "learning_rate": 4.008256471519536e-06,
+      "loss": 0.5309,
+      "step": 454
+    },
+    {
+      "epoch": 4.41747572815534,
+      "grad_norm": 0.5386315901899501,
+      "learning_rate": 4.004199215428032e-06,
+      "loss": 0.6876,
+      "step": 455
+    },
+    {
+      "epoch": 4.427184466019417,
+      "grad_norm": 0.6303477188551428,
+      "learning_rate": 4.000135739943735e-06,
+      "loss": 0.7771,
+      "step": 456
+    },
+    {
+      "epoch": 4.436893203883495,
+      "grad_norm": 0.877305177258536,
+      "learning_rate": 3.996066061867844e-06,
+      "loss": 0.7183,
+      "step": 457
+    },
+    {
+      "epoch": 4.446601941747573,
+      "grad_norm": 0.5475353424276045,
+      "learning_rate": 3.991990198027203e-06,
+      "loss": 0.6071,
+      "step": 458
+    },
+    {
+      "epoch": 4.456310679611651,
+      "grad_norm": 0.5369581006252573,
+      "learning_rate": 3.987908165274233e-06,
+      "loss": 0.6663,
+      "step": 459
+    },
+    {
+      "epoch": 4.466019417475728,
+      "grad_norm": 0.5481338983710499,
+      "learning_rate": 3.9838199804868635e-06,
+      "loss": 0.7901,
+      "step": 460
+    },
+    {
+      "epoch": 4.475728155339806,
+      "grad_norm": 0.5673339500133648,
+      "learning_rate": 3.979725660568456e-06,
+      "loss": 0.7221,
+      "step": 461
+    },
+    {
+      "epoch": 4.485436893203883,
+      "grad_norm": 0.5276433542097172,
+      "learning_rate": 3.975625222447742e-06,
+      "loss": 0.5454,
+      "step": 462
+    },
+    {
+      "epoch": 4.495145631067961,
+      "grad_norm": 0.5773222334966291,
+      "learning_rate": 3.97151868307875e-06,
+      "loss": 0.8548,
+      "step": 463
+    },
+    {
+      "epoch": 4.504854368932039,
+      "grad_norm": 0.5288063685496589,
+      "learning_rate": 3.9674060594407345e-06,
+      "loss": 0.6481,
+      "step": 464
+    },
+    {
+      "epoch": 4.514563106796117,
+      "grad_norm": 0.5234908733008361,
+      "learning_rate": 3.963287368538105e-06,
+      "loss": 0.4874,
+      "step": 465
+    },
+    {
+      "epoch": 4.524271844660194,
+      "grad_norm": 0.5548781615871586,
+      "learning_rate": 3.959162627400361e-06,
+      "loss": 0.7774,
+      "step": 466
+    },
+    {
+      "epoch": 4.533980582524272,
+      "grad_norm": 0.6088390104886395,
+      "learning_rate": 3.9550318530820145e-06,
+      "loss": 0.7768,
+      "step": 467
+    },
+    {
+      "epoch": 4.543689320388349,
+      "grad_norm": 0.5362660929991437,
+      "learning_rate": 3.9508950626625244e-06,
+      "loss": 0.5664,
+      "step": 468
+    },
+    {
+      "epoch": 4.553398058252427,
+      "grad_norm": 0.5403037827442222,
+      "learning_rate": 3.946752273246224e-06,
+      "loss": 0.7226,
+      "step": 469
+    },
+    {
+      "epoch": 4.563106796116505,
+      "grad_norm": 0.5160659142912444,
+      "learning_rate": 3.942603501962249e-06,
+      "loss": 0.609,
+      "step": 470
+    },
+    {
+      "epoch": 4.572815533980583,
+      "grad_norm": 0.6197532352354348,
+      "learning_rate": 3.9384487659644716e-06,
+      "loss": 0.5509,
+      "step": 471
+    },
+    {
+      "epoch": 4.58252427184466,
+      "grad_norm": 0.6600383068172292,
+      "learning_rate": 3.934288082431423e-06,
+      "loss": 0.7257,
+      "step": 472
+    },
+    {
+      "epoch": 4.592233009708738,
+      "grad_norm": 0.5641531137481023,
+      "learning_rate": 3.930121468566227e-06,
+      "loss": 0.8232,
+      "step": 473
+    },
+    {
+      "epoch": 4.601941747572815,
+      "grad_norm": 0.5450663648523331,
+      "learning_rate": 3.925948941596528e-06,
+      "loss": 0.6493,
+      "step": 474
+    },
+    {
+      "epoch": 4.611650485436893,
+      "grad_norm": 0.5134107913642527,
+      "learning_rate": 3.92177051877442e-06,
+      "loss": 0.554,
+      "step": 475
+    },
+    {
+      "epoch": 4.621359223300971,
+      "grad_norm": 0.5988443116784232,
+      "learning_rate": 3.917586217376369e-06,
+      "loss": 0.7745,
+      "step": 476
+    },
+    {
+      "epoch": 4.631067961165049,
+      "grad_norm": 0.5138233596193919,
+      "learning_rate": 3.913396054703155e-06,
+      "loss": 0.7129,
+      "step": 477
+    },
+    {
+      "epoch": 4.640776699029126,
+      "grad_norm": 0.5323215028941234,
+      "learning_rate": 3.909200048079786e-06,
+      "loss": 0.7206,
+      "step": 478
+    },
+    {
+      "epoch": 4.650485436893204,
+      "grad_norm": 0.5875971843837734,
+      "learning_rate": 3.9049982148554384e-06,
+      "loss": 0.8862,
+      "step": 479
+    },
+    {
+      "epoch": 4.660194174757281,
+      "grad_norm": 0.5645023952141737,
+      "learning_rate": 3.900790572403376e-06,
+      "loss": 0.6745,
+      "step": 480
+    },
+    {
+      "epoch": 4.669902912621359,
+      "grad_norm": 0.5154008565576635,
+      "learning_rate": 3.896577138120881e-06,
+      "loss": 0.6947,
+      "step": 481
+    },
+    {
+      "epoch": 4.679611650485437,
+      "grad_norm": 0.5394552750418433,
+      "learning_rate": 3.892357929429187e-06,
+      "loss": 0.6791,
+      "step": 482
+    },
+    {
+      "epoch": 4.689320388349515,
+      "grad_norm": 0.5827930464161047,
+      "learning_rate": 3.8881329637734e-06,
+      "loss": 0.5154,
+      "step": 483
+    },
+    {
+      "epoch": 4.699029126213592,
+      "grad_norm": 0.5361181581166996,
+      "learning_rate": 3.883902258622431e-06,
+      "loss": 0.6468,
+      "step": 484
+    },
+    {
+      "epoch": 4.70873786407767,
+      "grad_norm": 0.5711060546895186,
+      "learning_rate": 3.8796658314689205e-06,
+      "loss": 0.7167,
+      "step": 485
+    },
+    {
+      "epoch": 4.718446601941747,
+      "grad_norm": 0.5434222419984399,
+      "learning_rate": 3.875423699829168e-06,
+      "loss": 0.5867,
+      "step": 486
+    },
+    {
+      "epoch": 4.728155339805825,
+      "grad_norm": 0.5096613481897871,
+      "learning_rate": 3.871175881243061e-06,
+      "loss": 0.6356,
+      "step": 487
+    },
+    {
+      "epoch": 4.737864077669903,
+      "grad_norm": 0.542382695866742,
+      "learning_rate": 3.866922393273999e-06,
+      "loss": 0.7997,
+      "step": 488
+    },
+    {
+      "epoch": 4.747572815533981,
+      "grad_norm": 0.5295202287930876,
+      "learning_rate": 3.862663253508822e-06,
+      "loss": 0.6506,
+      "step": 489
+    },
+    {
+      "epoch": 4.757281553398058,
+      "grad_norm": 0.5116429751055901,
+      "learning_rate": 3.858398479557739e-06,
+      "loss": 0.6037,
+      "step": 490
+    },
+    {
+      "epoch": 4.766990291262136,
+      "grad_norm": 0.5283075153391654,
+      "learning_rate": 3.8541280890542565e-06,
+      "loss": 0.6142,
+      "step": 491
+    },
+    {
+      "epoch": 4.776699029126213,
+      "grad_norm": 0.5741236822380017,
+      "learning_rate": 3.849852099655102e-06,
+      "loss": 0.7518,
+      "step": 492
+    },
+    {
+      "epoch": 4.786407766990291,
+      "grad_norm": 0.5062208506690622,
+      "learning_rate": 3.845570529040151e-06,
+      "loss": 0.6338,
+      "step": 493
+    },
+    {
+      "epoch": 4.796116504854369,
+      "grad_norm": 0.5321985087384002,
+      "learning_rate": 3.841283394912361e-06,
+      "loss": 0.6375,
+      "step": 494
+    },
+    {
+      "epoch": 4.805825242718447,
+      "grad_norm": 0.5310703753315527,
+      "learning_rate": 3.836990714997686e-06,
+      "loss": 0.7015,
+      "step": 495
+    },
+    {
+      "epoch": 4.815533980582524,
+      "grad_norm": 0.5620272702074033,
+      "learning_rate": 3.832692507045015e-06,
+      "loss": 0.731,
+      "step": 496
+    },
+    {
+      "epoch": 4.825242718446602,
+      "grad_norm": 0.5060945256107214,
+      "learning_rate": 3.828388788826091e-06,
+      "loss": 0.7914,
+      "step": 497
+    },
+    {
+      "epoch": 4.834951456310679,
+      "grad_norm": 0.5538607078294117,
+      "learning_rate": 3.824079578135442e-06,
+      "loss": 0.6944,
+      "step": 498
+    },
+    {
+      "epoch": 4.844660194174757,
+      "grad_norm": 0.5452852198799888,
+      "learning_rate": 3.819764892790307e-06,
+      "loss": 0.7015,
+      "step": 499
+    },
+    {
+      "epoch": 4.854368932038835,
+      "grad_norm": 0.557748542801134,
+      "learning_rate": 3.815444750630555e-06,
+      "loss": 0.6616,
+      "step": 500
+    },
+    {
+      "epoch": 4.864077669902913,
+      "grad_norm": 0.555226374195002,
+      "learning_rate": 3.811119169518624e-06,
+      "loss": 0.8484,
+      "step": 501
+    },
+    {
+      "epoch": 4.87378640776699,
+      "grad_norm": 0.583053708240317,
+      "learning_rate": 3.8067881673394363e-06,
+      "loss": 1.071,
+      "step": 502
+    },
+    {
+      "epoch": 4.883495145631068,
+      "grad_norm": 0.6078873319472481,
+      "learning_rate": 3.802451762000331e-06,
+      "loss": 0.6256,
+      "step": 503
+    },
+    {
+      "epoch": 4.893203883495145,
+      "grad_norm": 0.539423072387751,
+      "learning_rate": 3.7981099714309856e-06,
+      "loss": 0.5403,
+      "step": 504
+    },
+    {
+      "epoch": 4.902912621359223,
+      "grad_norm": 0.6603698148330428,
+      "learning_rate": 3.7937628135833453e-06,
+      "loss": 0.6906,
+      "step": 505
+    },
+    {
+      "epoch": 4.9126213592233015,
+      "grad_norm": 0.5562121401027199,
+      "learning_rate": 3.7894103064315463e-06,
+      "loss": 0.7039,
+      "step": 506
+    },
+    {
+      "epoch": 4.922330097087379,
+      "grad_norm": 0.5336235803797111,
+      "learning_rate": 3.7850524679718424e-06,
+      "loss": 0.59,
+      "step": 507
+    },
+    {
+      "epoch": 4.932038834951456,
+      "grad_norm": 0.5331917381976123,
+      "learning_rate": 3.7806893162225328e-06,
+      "loss": 0.54,
+      "step": 508
+    },
+    {
+      "epoch": 4.941747572815534,
+      "grad_norm": 0.5058791106293693,
+      "learning_rate": 3.7763208692238818e-06,
+      "loss": 0.6653,
+      "step": 509
+    },
+    {
+      "epoch": 4.951456310679612,
+      "grad_norm": 0.5299074738912896,
+      "learning_rate": 3.7719471450380518e-06,
+      "loss": 0.6194,
+      "step": 510
+    },
+    {
+      "epoch": 4.961165048543689,
+      "grad_norm": 0.5327819159805345,
+      "learning_rate": 3.7675681617490212e-06,
+      "loss": 0.805,
+      "step": 511
+    },
+    {
+      "epoch": 4.970873786407767,
+      "grad_norm": 0.5298660538103491,
+      "learning_rate": 3.7631839374625167e-06,
+      "loss": 0.7929,
+      "step": 512
+    },
+    {
+      "epoch": 4.980582524271845,
+      "grad_norm": 0.4764247185946088,
+      "learning_rate": 3.758794490305932e-06,
+      "loss": 0.611,
+      "step": 513
+    },
+    {
+      "epoch": 4.990291262135923,
+      "grad_norm": 0.5451924892487114,
+      "learning_rate": 3.7543998384282565e-06,
+      "loss": 0.6893,
+      "step": 514
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.5454441184842063,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.7546,
+      "step": 515
+    },
+    {
+      "epoch": 5.009708737864078,
+      "grad_norm": 0.5236580494464181,
+      "learning_rate": 3.745594993213118e-06,
+      "loss": 0.743,
+      "step": 516
+    },
+    {
+      "epoch": 5.019417475728155,
+      "grad_norm": 0.5322157623594023,
+      "learning_rate": 3.7411848362809324e-06,
+      "loss": 0.9126,
+      "step": 517
+    },
+    {
+      "epoch": 5.029126213592233,
+      "grad_norm": 0.4717932076034679,
+      "learning_rate": 3.7367695474380623e-06,
+      "loss": 0.6524,
+      "step": 518
+    },
+    {
+      "epoch": 5.038834951456311,
+      "grad_norm": 0.504345727123636,
+      "learning_rate": 3.7323491449403444e-06,
+      "loss": 0.74,
+      "step": 519
+    },
+    {
+      "epoch": 5.048543689320389,
+      "grad_norm": 0.4527014691294767,
+      "learning_rate": 3.7279236470647593e-06,
+      "loss": 0.5095,
+      "step": 520
+    },
+    {
+      "epoch": 5.058252427184466,
+      "grad_norm": 0.5507904251400323,
+      "learning_rate": 3.723493072109355e-06,
+      "loss": 0.8863,
+      "step": 521
+    },
+    {
+      "epoch": 5.067961165048544,
+      "grad_norm": 0.4965393682876186,
+      "learning_rate": 3.719057438393172e-06,
+      "loss": 0.6478,
+      "step": 522
+    },
+    {
+      "epoch": 5.077669902912621,
+      "grad_norm": 0.5218053726872672,
+      "learning_rate": 3.714616764256166e-06,
+      "loss": 0.8168,
+      "step": 523
+    },
+    {
+      "epoch": 5.087378640776699,
+      "grad_norm": 0.5407179591306065,
+      "learning_rate": 3.7101710680591353e-06,
+      "loss": 0.6817,
+      "step": 524
+    },
+    {
+      "epoch": 5.097087378640777,
+      "grad_norm": 0.5149668318951321,
+      "learning_rate": 3.7057203681836407e-06,
+      "loss": 0.7706,
+      "step": 525
+    },
+    {
+      "epoch": 5.106796116504855,
+      "grad_norm": 0.6260255413647343,
+      "learning_rate": 3.701264683031934e-06,
+      "loss": 0.8372,
+      "step": 526
+    },
+    {
+      "epoch": 5.116504854368932,
+      "grad_norm": 0.5147717662373354,
+      "learning_rate": 3.6968040310268766e-06,
+      "loss": 0.6642,
+      "step": 527
+    },
+    {
+      "epoch": 5.12621359223301,
+      "grad_norm": 0.5229036248727982,
+      "learning_rate": 3.692338430611869e-06,
+      "loss": 0.6629,
+      "step": 528
+    },
+    {
+      "epoch": 5.135922330097087,
+      "grad_norm": 0.5573618752917658,
+      "learning_rate": 3.687867900250771e-06,
+      "loss": 0.6025,
+      "step": 529
+    },
+    {
+      "epoch": 5.145631067961165,
+      "grad_norm": 0.4894399208639397,
+      "learning_rate": 3.683392458427825e-06,
+      "loss": 0.5302,
+      "step": 530
+    },
+    {
+      "epoch": 5.155339805825243,
+      "grad_norm": 0.5637535925415186,
+      "learning_rate": 3.6789121236475818e-06,
+      "loss": 0.7266,
+      "step": 531
+    },
+    {
+      "epoch": 5.165048543689321,
+      "grad_norm": 0.5789728245531511,
+      "learning_rate": 3.674426914434824e-06,
+      "loss": 0.7271,
+      "step": 532
+    },
+    {
+      "epoch": 5.174757281553398,
+      "grad_norm": 0.5466853817465158,
+      "learning_rate": 3.6699368493344856e-06,
+      "loss": 0.7348,
+      "step": 533
+    },
+    {
+      "epoch": 5.184466019417476,
+      "grad_norm": 0.5042069743820944,
+      "learning_rate": 3.665441946911582e-06,
+      "loss": 0.5876,
+      "step": 534
+    },
+    {
+      "epoch": 5.194174757281553,
+      "grad_norm": 0.55765274686579,
+      "learning_rate": 3.660942225751126e-06,
+      "loss": 0.6261,
+      "step": 535
+    },
+    {
+      "epoch": 5.203883495145631,
+      "grad_norm": 0.5585892588024256,
+      "learning_rate": 3.6564377044580558e-06,
+      "loss": 0.6696,
+      "step": 536
+    },
+    {
+      "epoch": 5.213592233009709,
+      "grad_norm": 0.5410505344470596,
+      "learning_rate": 3.6519284016571567e-06,
+      "loss": 0.5517,
+      "step": 537
+    },
+    {
+      "epoch": 5.223300970873787,
+      "grad_norm": 0.5655367086406735,
+      "learning_rate": 3.647414335992985e-06,
+      "loss": 0.8221,
+      "step": 538
+    },
+    {
+      "epoch": 5.233009708737864,
+      "grad_norm": 0.5496575434925196,
+      "learning_rate": 3.642895526129787e-06,
+      "loss": 0.7155,
+      "step": 539
+    },
+    {
+      "epoch": 5.242718446601942,
+      "grad_norm": 0.527282587966023,
+      "learning_rate": 3.638371990751428e-06,
+      "loss": 0.7697,
+      "step": 540
+    },
+    {
+      "epoch": 5.252427184466019,
+      "grad_norm": 0.5886591311740469,
+      "learning_rate": 3.63384374856131e-06,
+      "loss": 0.7738,
+      "step": 541
+    },
+    {
+      "epoch": 5.262135922330097,
+      "grad_norm": 0.547315818665875,
+      "learning_rate": 3.629310818282297e-06,
+      "loss": 0.849,
+      "step": 542
+    },
+    {
+      "epoch": 5.271844660194175,
+      "grad_norm": 0.5281918770795211,
+      "learning_rate": 3.6247732186566365e-06,
+      "loss": 0.6471,
+      "step": 543
+    },
+    {
+      "epoch": 5.281553398058253,
+      "grad_norm": 0.5082451180767762,
+      "learning_rate": 3.6202309684458813e-06,
+      "loss": 0.5249,
+      "step": 544
+    },
+    {
+      "epoch": 5.29126213592233,
+      "grad_norm": 0.5048524465631024,
+      "learning_rate": 3.615684086430815e-06,
+      "loss": 0.5654,
+      "step": 545
+    },
+    {
+      "epoch": 5.300970873786408,
+      "grad_norm": 0.5489198982092451,
+      "learning_rate": 3.61113259141137e-06,
+      "loss": 0.7281,
+      "step": 546
+    },
+    {
+      "epoch": 5.310679611650485,
+      "grad_norm": 0.5248421669868224,
+      "learning_rate": 3.606576502206554e-06,
+      "loss": 0.4695,
+      "step": 547
+    },
+    {
+      "epoch": 5.320388349514563,
+      "grad_norm": 0.5525806088263864,
+      "learning_rate": 3.602015837654369e-06,
+      "loss": 0.5076,
+      "step": 548
+    },
+    {
+      "epoch": 5.330097087378641,
+      "grad_norm": 0.5147558593612964,
+      "learning_rate": 3.5974506166117355e-06,
+      "loss": 0.5361,
+      "step": 549
+    },
+    {
+      "epoch": 5.339805825242719,
+      "grad_norm": 0.5230812119454316,
+      "learning_rate": 3.592880857954413e-06,
+      "loss": 0.79,
+      "step": 550
+    },
+    {
+      "epoch": 5.349514563106796,
+      "grad_norm": 0.47610639093924234,
+      "learning_rate": 3.588306580576922e-06,
+      "loss": 0.5074,
+      "step": 551
+    },
+    {
+      "epoch": 5.359223300970874,
+      "grad_norm": 0.5076655865626652,
+      "learning_rate": 3.583727803392468e-06,
+      "loss": 0.4614,
+      "step": 552
+    },
+    {
+      "epoch": 5.368932038834951,
+      "grad_norm": 0.5034407309618866,
+      "learning_rate": 3.57914454533286e-06,
+      "loss": 0.7016,
+      "step": 553
+    },
+    {
+      "epoch": 5.378640776699029,
+      "grad_norm": 0.5956638435834678,
+      "learning_rate": 3.5745568253484363e-06,
+      "loss": 0.574,
+      "step": 554
+    },
+    {
+      "epoch": 5.388349514563107,
+      "grad_norm": 0.5374526756995527,
+      "learning_rate": 3.5699646624079824e-06,
+      "loss": 0.6492,
+      "step": 555
+    },
+    {
+      "epoch": 5.398058252427185,
+      "grad_norm": 0.5748361523110773,
+      "learning_rate": 3.5653680754986543e-06,
+      "loss": 0.4815,
+      "step": 556
+    },
+    {
+      "epoch": 5.407766990291262,
+      "grad_norm": 0.54567644758393,
+      "learning_rate": 3.560767083625899e-06,
+      "loss": 0.8402,
+      "step": 557
+    },
+    {
+      "epoch": 5.41747572815534,
+      "grad_norm": 0.5659255771617656,
+      "learning_rate": 3.556161705813378e-06,
+      "loss": 0.5835,
+      "step": 558
+    },
+    {
+      "epoch": 5.427184466019417,
+      "grad_norm": 0.6247712187120456,
+      "learning_rate": 3.5515519611028863e-06,
+      "loss": 0.8306,
+      "step": 559
+    },
+    {
+      "epoch": 5.436893203883495,
+      "grad_norm": 0.5173918216113128,
+      "learning_rate": 3.5469378685542742e-06,
+      "loss": 0.733,
+      "step": 560
+    },
+    {
+      "epoch": 5.446601941747573,
+      "grad_norm": 0.5518176200795789,
+      "learning_rate": 3.542319447245372e-06,
+      "loss": 0.9392,
+      "step": 561
+    },
+    {
+      "epoch": 5.456310679611651,
+      "grad_norm": 0.5349757875135974,
+      "learning_rate": 3.537696716271904e-06,
+      "loss": 0.8355,
+      "step": 562
+    },
+    {
+      "epoch": 5.466019417475728,
+      "grad_norm": 0.5211503905687772,
+      "learning_rate": 3.533069694747415e-06,
+      "loss": 0.5956,
+      "step": 563
+    },
+    {
+      "epoch": 5.475728155339806,
+      "grad_norm": 0.4837470952780332,
+      "learning_rate": 3.528438401803192e-06,
+      "loss": 0.5892,
+      "step": 564
+    },
+    {
+      "epoch": 5.485436893203883,
+      "grad_norm": 0.5125612599343026,
+      "learning_rate": 3.52380285658818e-06,
+      "loss": 0.5399,
+      "step": 565
+    },
+    {
+      "epoch": 5.495145631067961,
+      "grad_norm": 0.5410485375898842,
+      "learning_rate": 3.5191630782689074e-06,
+      "loss": 0.6512,
+      "step": 566
+    },
+    {
+      "epoch": 5.504854368932039,
+      "grad_norm": 0.47850608014355017,
+      "learning_rate": 3.5145190860294043e-06,
+      "loss": 0.5441,
+      "step": 567
+    },
+    {
+      "epoch": 5.514563106796117,
+      "grad_norm": 0.49368651452695383,
+      "learning_rate": 3.5098708990711254e-06,
+      "loss": 0.4864,
+      "step": 568
+    },
+    {
+      "epoch": 5.524271844660194,
+      "grad_norm": 0.5210530061794847,
+      "learning_rate": 3.505218536612869e-06,
+      "loss": 0.6663,
+      "step": 569
+    },
+    {
+      "epoch": 5.533980582524272,
+      "grad_norm": 0.5402383645877286,
+      "learning_rate": 3.500562017890695e-06,
+      "loss": 0.5224,
+      "step": 570
+    },
+    {
+      "epoch": 5.543689320388349,
+      "grad_norm": 0.5069706433380721,
+      "learning_rate": 3.495901362157853e-06,
+      "loss": 0.4391,
+      "step": 571
+    },
+    {
+      "epoch": 5.553398058252427,
+      "grad_norm": 0.5370346510366669,
+      "learning_rate": 3.4912365886846934e-06,
+      "loss": 0.8669,
+      "step": 572
+    },
+    {
+      "epoch": 5.563106796116505,
+      "grad_norm": 0.4976926429074089,
+      "learning_rate": 3.4865677167585942e-06,
+      "loss": 0.678,
+      "step": 573
+    },
+    {
+      "epoch": 5.572815533980583,
+      "grad_norm": 0.5118571857656103,
+      "learning_rate": 3.4818947656838796e-06,
+      "loss": 0.5575,
+      "step": 574
+    },
+    {
+      "epoch": 5.58252427184466,
+      "grad_norm": 0.575107709638005,
+      "learning_rate": 3.4772177547817387e-06,
+      "loss": 0.5642,
+      "step": 575
+    },
+    {
+      "epoch": 5.592233009708738,
+      "grad_norm": 0.544943705330846,
+      "learning_rate": 3.472536703390148e-06,
+      "loss": 0.5671,
+      "step": 576
+    },
+    {
+      "epoch": 5.601941747572815,
+      "grad_norm": 0.48590763199396203,
+      "learning_rate": 3.467851630863789e-06,
+      "loss": 0.5398,
+      "step": 577
+    },
+    {
+      "epoch": 5.611650485436893,
+      "grad_norm": 0.48995907325112287,
+      "learning_rate": 3.463162556573969e-06,
+      "loss": 0.6857,
+      "step": 578
+    },
+    {
+      "epoch": 5.621359223300971,
+      "grad_norm": 0.5250045703440803,
+      "learning_rate": 3.4584694999085424e-06,
+      "loss": 0.7885,
+      "step": 579
+    },
+    {
+      "epoch": 5.631067961165049,
+      "grad_norm": 0.5629591675833219,
+      "learning_rate": 3.4537724802718294e-06,
+      "loss": 0.5854,
+      "step": 580
+    },
+    {
+      "epoch": 5.640776699029126,
+      "grad_norm": 0.4846489281273186,
+      "learning_rate": 3.4490715170845356e-06,
+      "loss": 0.5208,
+      "step": 581
+    },
+    {
+      "epoch": 5.650485436893204,
+      "grad_norm": 0.5888541854252343,
+      "learning_rate": 3.4443666297836715e-06,
+      "loss": 0.5512,
+      "step": 582
+    },
+    {
+      "epoch": 5.660194174757281,
+      "grad_norm": 0.6442464798240836,
+      "learning_rate": 3.4396578378224734e-06,
+      "loss": 0.5946,
+      "step": 583
+    },
+    {
+      "epoch": 5.669902912621359,
+      "grad_norm": 0.493758970029001,
+      "learning_rate": 3.4349451606703214e-06,
+      "loss": 0.6526,
+      "step": 584
+    },
+    {
+      "epoch": 5.679611650485437,
+      "grad_norm": 0.581861416552549,
+      "learning_rate": 3.430228617812661e-06,
+      "loss": 0.5385,
+      "step": 585
+    },
+    {
+      "epoch": 5.689320388349515,
+      "grad_norm": 0.496120024238278,
+      "learning_rate": 3.4255082287509183e-06,
+      "loss": 0.4038,
+      "step": 586
+    },
+    {
+      "epoch": 5.699029126213592,
+      "grad_norm": 0.5767006067460728,
+      "learning_rate": 3.420784013002426e-06,
+      "loss": 0.5677,
+      "step": 587
+    },
+    {
+      "epoch": 5.70873786407767,
+      "grad_norm": 0.5158422312441495,
+      "learning_rate": 3.416055990100336e-06,
+      "loss": 0.5685,
+      "step": 588
+    },
+    {
+      "epoch": 5.718446601941747,
+      "grad_norm": 0.5162552308353006,
+      "learning_rate": 3.4113241795935427e-06,
+      "loss": 0.5916,
+      "step": 589
+    },
+    {
+      "epoch": 5.728155339805825,
+      "grad_norm": 0.5431343552918371,
+      "learning_rate": 3.4065886010466014e-06,
+      "loss": 0.6086,
+      "step": 590
+    },
+    {
+      "epoch": 5.737864077669903,
+      "grad_norm": 0.5119365013009938,
+      "learning_rate": 3.401849274039647e-06,
+      "loss": 0.6584,
+      "step": 591
+    },
+    {
+      "epoch": 5.747572815533981,
+      "grad_norm": 0.4947683856841006,
+      "learning_rate": 3.3971062181683117e-06,
+      "loss": 0.5387,
+      "step": 592
+    },
+    {
+      "epoch": 5.757281553398058,
+      "grad_norm": 0.5434375331427597,
+      "learning_rate": 3.3923594530436477e-06,
+      "loss": 0.7175,
+      "step": 593
+    },
+    {
+      "epoch": 5.766990291262136,
+      "grad_norm": 0.5577372563704153,
+      "learning_rate": 3.387608998292041e-06,
+      "loss": 0.6719,
+      "step": 594
+    },
+    {
+      "epoch": 5.776699029126213,
+      "grad_norm": 0.4837142305485574,
+      "learning_rate": 3.382854873555137e-06,
+      "loss": 0.4739,
+      "step": 595
+    },
+    {
+      "epoch": 5.786407766990291,
+      "grad_norm": 0.5306514259002709,
+      "learning_rate": 3.3780970984897504e-06,
+      "loss": 0.5225,
+      "step": 596
+    },
+    {
+      "epoch": 5.796116504854369,
+      "grad_norm": 0.5194892402061854,
+      "learning_rate": 3.373335692767793e-06,
+      "loss": 0.776,
+      "step": 597
+    },
+    {
+      "epoch": 5.805825242718447,
+      "grad_norm": 0.518491658036596,
+      "learning_rate": 3.3685706760761865e-06,
+      "loss": 0.6919,
+      "step": 598
+    },
+    {
+      "epoch": 5.815533980582524,
+      "grad_norm": 0.5938926587571165,
+      "learning_rate": 3.3638020681167827e-06,
+      "loss": 0.5489,
+      "step": 599
+    },
+    {
+      "epoch": 5.825242718446602,
+      "grad_norm": 0.5206196560346433,
+      "learning_rate": 3.3590298886062833e-06,
+      "loss": 0.5532,
+      "step": 600
+    },
+    {
+      "epoch": 5.834951456310679,
+      "grad_norm": 0.4898920438591441,
+      "learning_rate": 3.354254157276155e-06,
+      "loss": 0.7708,
+      "step": 601
+    },
+    {
+      "epoch": 5.844660194174757,
+      "grad_norm": 0.7027567743473708,
+      "learning_rate": 3.3494748938725525e-06,
+      "loss": 0.5813,
+      "step": 602
+    },
+    {
+      "epoch": 5.854368932038835,
+      "grad_norm": 0.5373706241783007,
+      "learning_rate": 3.3446921181562326e-06,
+      "loss": 0.5484,
+      "step": 603
+    },
+    {
+      "epoch": 5.864077669902913,
+      "grad_norm": 0.6287988932770455,
+      "learning_rate": 3.3399058499024767e-06,
+      "loss": 0.4705,
+      "step": 604
+    },
+    {
+      "epoch": 5.87378640776699,
+      "grad_norm": 0.6420488331857355,
+      "learning_rate": 3.3351161089010055e-06,
+      "loss": 0.8357,
+      "step": 605
+    },
+    {
+      "epoch": 5.883495145631068,
+      "grad_norm": 0.483984948258574,
+      "learning_rate": 3.330322914955897e-06,
+      "loss": 0.6679,
+      "step": 606
+    },
+    {
+      "epoch": 5.893203883495145,
+      "grad_norm": 0.5633594210474601,
+      "learning_rate": 3.325526287885509e-06,
+      "loss": 0.6853,
+      "step": 607
+    },
+    {
+      "epoch": 5.902912621359223,
+      "grad_norm": 0.5099944120499966,
+      "learning_rate": 3.3207262475223913e-06,
+      "loss": 0.653,
+      "step": 608
+    },
+    {
+      "epoch": 5.9126213592233015,
+      "grad_norm": 0.6133535314483775,
+      "learning_rate": 3.315922813713209e-06,
+      "loss": 0.8421,
+      "step": 609
+    },
+    {
+      "epoch": 5.922330097087379,
+      "grad_norm": 0.5514042535577254,
+      "learning_rate": 3.3111160063186553e-06,
+      "loss": 0.6114,
+      "step": 610
+    },
+    {
+      "epoch": 5.932038834951456,
+      "grad_norm": 0.515401175634321,
+      "learning_rate": 3.3063058452133756e-06,
+      "loss": 0.7155,
+      "step": 611
+    },
+    {
+      "epoch": 5.941747572815534,
+      "grad_norm": 0.5483222829849357,
+      "learning_rate": 3.301492350285879e-06,
+      "loss": 0.6461,
+      "step": 612
+    },
+    {
+      "epoch": 5.951456310679612,
+      "grad_norm": 0.5102148189624786,
+      "learning_rate": 3.296675541438461e-06,
+      "loss": 0.5192,
+      "step": 613
+    },
+    {
+      "epoch": 5.961165048543689,
+      "grad_norm": 0.5459577941840908,
+      "learning_rate": 3.2918554385871163e-06,
+      "loss": 0.7572,
+      "step": 614
+    },
+    {
+      "epoch": 5.970873786407767,
+      "grad_norm": 0.4986561470940389,
+      "learning_rate": 3.2870320616614626e-06,
+      "loss": 0.4979,
+      "step": 615
+    },
+    {
+      "epoch": 5.980582524271845,
+      "grad_norm": 0.5366419322997734,
+      "learning_rate": 3.282205430604653e-06,
+      "loss": 0.5367,
+      "step": 616
+    },
+    {
+      "epoch": 5.990291262135923,
+      "grad_norm": 0.6234877265968874,
+      "learning_rate": 3.2773755653732954e-06,
+      "loss": 0.6123,
+      "step": 617
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.5139134055855142,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.6528,
+      "step": 618
+    },
+    {
+      "epoch": 6.009708737864078,
+      "grad_norm": 0.5535462185153359,
+      "learning_rate": 3.267706212280146e-06,
+      "loss": 0.5759,
+      "step": 619
+    },
+    {
+      "epoch": 6.019417475728155,
+      "grad_norm": 0.5263133197453993,
+      "learning_rate": 3.2628667643981036e-06,
+      "loss": 0.6369,
+      "step": 620
+    },
+    {
+      "epoch": 6.029126213592233,
+      "grad_norm": 0.5823149773756441,
+      "learning_rate": 3.2580241623008426e-06,
+      "loss": 0.7007,
+      "step": 621
+    },
+    {
+      "epoch": 6.038834951456311,
+      "grad_norm": 0.5698523756699659,
+      "learning_rate": 3.2531784260110067e-06,
+      "loss": 0.6968,
+      "step": 622
+    },
+    {
+      "epoch": 6.048543689320389,
+      "grad_norm": 0.5323935991893491,
+      "learning_rate": 3.2483295755641986e-06,
+      "loss": 0.6614,
+      "step": 623
+    },
+    {
+      "epoch": 6.058252427184466,
+      "grad_norm": 0.5579966093733544,
+      "learning_rate": 3.243477631008897e-06,
+      "loss": 0.7018,
+      "step": 624
+    },
+    {
+      "epoch": 6.067961165048544,
+      "grad_norm": 0.5672146611605858,
+      "learning_rate": 3.238622612406373e-06,
+      "loss": 0.4811,
+      "step": 625
+    },
+    {
+      "epoch": 6.077669902912621,
+      "grad_norm": 0.5487784461526284,
+      "learning_rate": 3.233764539830608e-06,
+      "loss": 0.6582,
+      "step": 626
+    },
+    {
+      "epoch": 6.087378640776699,
+      "grad_norm": 0.4968156455806434,
+      "learning_rate": 3.228903433368212e-06,
+      "loss": 0.4246,
+      "step": 627
+    },
+    {
+      "epoch": 6.097087378640777,
+      "grad_norm": 0.6508059967941122,
+      "learning_rate": 3.224039313118338e-06,
+      "loss": 0.3601,
+      "step": 628
+    },
+    {
+      "epoch": 6.106796116504855,
+      "grad_norm": 0.5223437529762169,
+      "learning_rate": 3.2191721991925993e-06,
+      "loss": 0.5353,
+      "step": 629
+    },
+    {
+      "epoch": 6.116504854368932,
+      "grad_norm": 0.4969256578725589,
+      "learning_rate": 3.21430211171499e-06,
+      "loss": 0.6791,
+      "step": 630
+    },
+    {
+      "epoch": 6.12621359223301,
+      "grad_norm": 0.5033067440842812,
+      "learning_rate": 3.209429070821795e-06,
+      "loss": 0.9006,
+      "step": 631
+    },
+    {
+      "epoch": 6.135922330097087,
+      "grad_norm": 0.5293076970287124,
+      "learning_rate": 3.2045530966615136e-06,
+      "loss": 0.7065,
+      "step": 632
+    },
+    {
+      "epoch": 6.145631067961165,
+      "grad_norm": 0.5232595313548619,
+      "learning_rate": 3.1996742093947724e-06,
+      "loss": 0.5669,
+      "step": 633
+    },
+    {
+      "epoch": 6.155339805825243,
+      "grad_norm": 0.5096836013728645,
+      "learning_rate": 3.1947924291942423e-06,
+      "loss": 0.6983,
+      "step": 634
+    },
+    {
+      "epoch": 6.165048543689321,
+      "grad_norm": 0.5367180425599433,
+      "learning_rate": 3.189907776244556e-06,
+      "loss": 0.5669,
+      "step": 635
+    },
+    {
+      "epoch": 6.174757281553398,
+      "grad_norm": 0.5645144344049964,
+      "learning_rate": 3.185020270742225e-06,
+      "loss": 0.6902,
+      "step": 636
+    },
+    {
+      "epoch": 6.184466019417476,
+      "grad_norm": 0.5634972380711424,
+      "learning_rate": 3.180129932895553e-06,
+      "loss": 0.4903,
+      "step": 637
+    },
+    {
+      "epoch": 6.194174757281553,
+      "grad_norm": 0.5589418053348287,
+      "learning_rate": 3.1752367829245563e-06,
+      "loss": 0.7803,
+      "step": 638
+    },
+    {
+      "epoch": 6.203883495145631,
+      "grad_norm": 0.5558458265176147,
+      "learning_rate": 3.1703408410608777e-06,
+      "loss": 0.8576,
+      "step": 639
+    },
+    {
+      "epoch": 6.213592233009709,
+      "grad_norm": 0.4994812128687224,
+      "learning_rate": 3.1654421275477045e-06,
+      "loss": 0.598,
+      "step": 640
+    },
+    {
+      "epoch": 6.223300970873787,
+      "grad_norm": 0.50112018228175,
+      "learning_rate": 3.1605406626396826e-06,
+      "loss": 0.6546,
+      "step": 641
+    },
+    {
+      "epoch": 6.233009708737864,
+      "grad_norm": 0.48779980049930693,
+      "learning_rate": 3.155636466602836e-06,
+      "loss": 0.629,
+      "step": 642
+    },
+    {
+      "epoch": 6.242718446601942,
+      "grad_norm": 0.5251674835972288,
+      "learning_rate": 3.150729559714478e-06,
+      "loss": 0.7687,
+      "step": 643
+    },
+    {
+      "epoch": 6.252427184466019,
+      "grad_norm": 0.5145810139817877,
+      "learning_rate": 3.145819962263134e-06,
+      "loss": 0.6442,
+      "step": 644
+    },
+    {
+      "epoch": 6.262135922330097,
+      "grad_norm": 0.5189234129866919,
+      "learning_rate": 3.1409076945484513e-06,
+      "loss": 0.536,
+      "step": 645
+    },
+    {
+      "epoch": 6.271844660194175,
+      "grad_norm": 0.5247294425181219,
+      "learning_rate": 3.135992776881119e-06,
+      "loss": 0.4885,
+      "step": 646
+    },
+    {
+      "epoch": 6.281553398058253,
+      "grad_norm": 0.561750334008629,
+      "learning_rate": 3.1310752295827818e-06,
+      "loss": 0.4685,
+      "step": 647
+    },
+    {
+      "epoch": 6.29126213592233,
+      "grad_norm": 0.48682689595750156,
+      "learning_rate": 3.1261550729859602e-06,
+      "loss": 0.6406,
+      "step": 648
+    },
+    {
+      "epoch": 6.300970873786408,
+      "grad_norm": 0.48042953552521844,
+      "learning_rate": 3.12123232743396e-06,
+      "loss": 0.5257,
+      "step": 649
+    },
+    {
+      "epoch": 6.310679611650485,
+      "grad_norm": 0.5265214331531651,
+      "learning_rate": 3.116307013280793e-06,
+      "loss": 0.8245,
+      "step": 650
+    },
+    {
+      "epoch": 6.320388349514563,
+      "grad_norm": 0.5216967258162386,
+      "learning_rate": 3.1113791508910913e-06,
+      "loss": 0.6038,
+      "step": 651
+    },
+    {
+      "epoch": 6.330097087378641,
+      "grad_norm": 0.4684172774395295,
+      "learning_rate": 3.106448760640022e-06,
+      "loss": 0.7096,
+      "step": 652
+    },
+    {
+      "epoch": 6.339805825242719,
+      "grad_norm": 0.5535208394965718,
+      "learning_rate": 3.1015158629132066e-06,
+      "loss": 0.5093,
+      "step": 653
+    },
+    {
+      "epoch": 6.349514563106796,
+      "grad_norm": 0.5230336669127887,
+      "learning_rate": 3.096580478106631e-06,
+      "loss": 0.6788,
+      "step": 654
+    },
+    {
+      "epoch": 6.359223300970874,
+      "grad_norm": 0.5497093072547722,
+      "learning_rate": 3.0916426266265676e-06,
+      "loss": 0.5919,
+      "step": 655
+    },
+    {
+      "epoch": 6.368932038834951,
+      "grad_norm": 0.5462224534842463,
+      "learning_rate": 3.086702328889486e-06,
+      "loss": 0.5686,
+      "step": 656
+    },
+    {
+      "epoch": 6.378640776699029,
+      "grad_norm": 0.5234947411014694,
+      "learning_rate": 3.0817596053219697e-06,
+      "loss": 0.5988,
+      "step": 657
+    },
+    {
+      "epoch": 6.388349514563107,
+      "grad_norm": 0.5588189733664695,
+      "learning_rate": 3.076814476360634e-06,
+      "loss": 0.5349,
+      "step": 658
+    },
+    {
+      "epoch": 6.398058252427185,
+      "grad_norm": 0.5779610115951219,
+      "learning_rate": 3.071866962452038e-06,
+      "loss": 0.3482,
+      "step": 659
+    },
+    {
+      "epoch": 6.407766990291262,
+      "grad_norm": 0.5479800145154113,
+      "learning_rate": 3.066917084052603e-06,
+      "loss": 0.7218,
+      "step": 660
+    },
+    {
+      "epoch": 6.41747572815534,
+      "grad_norm": 0.5397084718954609,
+      "learning_rate": 3.061964861628527e-06,
+      "loss": 0.6851,
+      "step": 661
+    },
+    {
+      "epoch": 6.427184466019417,
+      "grad_norm": 0.5329899425540183,
+      "learning_rate": 3.057010315655698e-06,
+      "loss": 0.4821,
+      "step": 662
+    },
+    {
+      "epoch": 6.436893203883495,
+      "grad_norm": 0.5458422785815938,
+      "learning_rate": 3.0520534666196134e-06,
+      "loss": 0.6327,
+      "step": 663
+    },
+    {
+      "epoch": 6.446601941747573,
+      "grad_norm": 0.47708680903135803,
+      "learning_rate": 3.0470943350152914e-06,
+      "loss": 0.5923,
+      "step": 664
+    },
+    {
+      "epoch": 6.456310679611651,
+      "grad_norm": 0.5144616542269372,
+      "learning_rate": 3.042132941347189e-06,
+      "loss": 0.6899,
+      "step": 665
+    },
+    {
+      "epoch": 6.466019417475728,
+      "grad_norm": 0.5433270155772785,
+      "learning_rate": 3.037169306129115e-06,
+      "loss": 0.6198,
+      "step": 666
+    },
+    {
+      "epoch": 6.475728155339806,
+      "grad_norm": 0.538855795352053,
+      "learning_rate": 3.0322034498841475e-06,
+      "loss": 0.6547,
+      "step": 667
+    },
+    {
+      "epoch": 6.485436893203883,
+      "grad_norm": 0.5708846163744057,
+      "learning_rate": 3.027235393144547e-06,
+      "loss": 0.6127,
+      "step": 668
+    },
+    {
+      "epoch": 6.495145631067961,
+      "grad_norm": 0.5317812805085081,
+      "learning_rate": 3.0222651564516715e-06,
+      "loss": 0.7369,
+      "step": 669
+    },
+    {
+      "epoch": 6.504854368932039,
+      "grad_norm": 0.5550895845970475,
+      "learning_rate": 3.017292760355896e-06,
+      "loss": 0.6136,
+      "step": 670
+    },
+    {
+      "epoch": 6.514563106796117,
+      "grad_norm": 0.553837992106348,
+      "learning_rate": 3.0123182254165194e-06,
+      "loss": 0.4291,
+      "step": 671
+    },
+    {
+      "epoch": 6.524271844660194,
+      "grad_norm": 0.5290319496985271,
+      "learning_rate": 3.0073415722016875e-06,
+      "loss": 0.8316,
+      "step": 672
+    },
+    {
+      "epoch": 6.533980582524272,
+      "grad_norm": 0.5733723058046754,
+      "learning_rate": 3.002362821288302e-06,
+      "loss": 0.5706,
+      "step": 673
+    },
+    {
+      "epoch": 6.543689320388349,
+      "grad_norm": 0.5372610455400619,
+      "learning_rate": 2.9973819932619404e-06,
+      "loss": 0.6444,
+      "step": 674
+    },
+    {
+      "epoch": 6.553398058252427,
+      "grad_norm": 0.5193287209030268,
+      "learning_rate": 2.9923991087167657e-06,
+      "loss": 0.6096,
+      "step": 675
+    },
+    {
+      "epoch": 6.563106796116505,
+      "grad_norm": 0.4949824248830205,
+      "learning_rate": 2.987414188255446e-06,
+      "loss": 0.6912,
+      "step": 676
+    },
+    {
+      "epoch": 6.572815533980583,
+      "grad_norm": 0.5814542413549719,
+      "learning_rate": 2.9824272524890664e-06,
+      "loss": 0.727,
+      "step": 677
+    },
+    {
+      "epoch": 6.58252427184466,
+      "grad_norm": 0.5117293174087842,
+      "learning_rate": 2.977438322037046e-06,
+      "loss": 0.5774,
+      "step": 678
+    },
+    {
+      "epoch": 6.592233009708738,
+      "grad_norm": 0.5409173310731737,
+      "learning_rate": 2.9724474175270485e-06,
+      "loss": 0.5585,
+      "step": 679
+    },
+    {
+      "epoch": 6.601941747572815,
+      "grad_norm": 0.5529086956863897,
+      "learning_rate": 2.967454559594903e-06,
+      "loss": 0.5848,
+      "step": 680
+    },
+    {
+      "epoch": 6.611650485436893,
+      "grad_norm": 0.5142526842665167,
+      "learning_rate": 2.9624597688845126e-06,
+      "loss": 0.5823,
+      "step": 681
+    },
+    {
+      "epoch": 6.621359223300971,
+      "grad_norm": 0.5056072060581505,
+      "learning_rate": 2.957463066047773e-06,
+      "loss": 0.6629,
+      "step": 682
+    },
+    {
+      "epoch": 6.631067961165049,
+      "grad_norm": 0.5648376543089534,
+      "learning_rate": 2.9524644717444866e-06,
+      "loss": 0.4144,
+      "step": 683
+    },
+    {
+      "epoch": 6.640776699029126,
+      "grad_norm": 0.5577514630301723,
+      "learning_rate": 2.9474640066422757e-06,
+      "loss": 0.8292,
+      "step": 684
+    },
+    {
+      "epoch": 6.650485436893204,
+      "grad_norm": 0.6179761336134023,
+      "learning_rate": 2.9424616914164982e-06,
+      "loss": 0.5197,
+      "step": 685
+    },
+    {
+      "epoch": 6.660194174757281,
+      "grad_norm": 0.5762308527579274,
+      "learning_rate": 2.9374575467501605e-06,
+      "loss": 0.5277,
+      "step": 686
+    },
+    {
+      "epoch": 6.669902912621359,
+      "grad_norm": 0.5104768442736158,
+      "learning_rate": 2.9324515933338343e-06,
+      "loss": 0.6059,
+      "step": 687
+    },
+    {
+      "epoch": 6.679611650485437,
+      "grad_norm": 0.5258016358630682,
+      "learning_rate": 2.9274438518655703e-06,
+      "loss": 0.753,
+      "step": 688
+    },
+    {
+      "epoch": 6.689320388349515,
+      "grad_norm": 0.5813275530174288,
+      "learning_rate": 2.9224343430508105e-06,
+      "loss": 0.476,
+      "step": 689
+    },
+    {
+      "epoch": 6.699029126213592,
+      "grad_norm": 0.5889679759396246,
+      "learning_rate": 2.917423087602306e-06,
+      "loss": 0.6838,
+      "step": 690
+    },
+    {
+      "epoch": 6.70873786407767,
+      "grad_norm": 0.5076859976205057,
+      "learning_rate": 2.9124101062400283e-06,
+      "loss": 0.6557,
+      "step": 691
+    },
+    {
+      "epoch": 6.718446601941747,
+      "grad_norm": 0.482464943160344,
+      "learning_rate": 2.907395419691087e-06,
+      "loss": 0.7261,
+      "step": 692
+    },
+    {
+      "epoch": 6.728155339805825,
+      "grad_norm": 0.5332220257567927,
+      "learning_rate": 2.9023790486896404e-06,
+      "loss": 0.5912,
+      "step": 693
+    },
+    {
+      "epoch": 6.737864077669903,
+      "grad_norm": 0.4682638337187069,
+      "learning_rate": 2.8973610139768114e-06,
+      "loss": 0.4614,
+      "step": 694
+    },
+    {
+      "epoch": 6.747572815533981,
+      "grad_norm": 0.558184511949571,
+      "learning_rate": 2.8923413363006038e-06,
+      "loss": 0.7464,
+      "step": 695
+    },
+    {
+      "epoch": 6.757281553398058,
+      "grad_norm": 0.5233023988021802,
+      "learning_rate": 2.887320036415811e-06,
+      "loss": 0.5608,
+      "step": 696
+    },
+    {
+      "epoch": 6.766990291262136,
+      "grad_norm": 0.5834246097072383,
+      "learning_rate": 2.882297135083937e-06,
+      "loss": 0.5184,
+      "step": 697
+    },
+    {
+      "epoch": 6.776699029126213,
+      "grad_norm": 0.5429366118270437,
+      "learning_rate": 2.877272653073107e-06,
+      "loss": 0.4984,
+      "step": 698
+    },
+    {
+      "epoch": 6.786407766990291,
+      "grad_norm": 0.5277264921068271,
+      "learning_rate": 2.87224661115798e-06,
+      "loss": 0.5662,
+      "step": 699
+    },
+    {
+      "epoch": 6.796116504854369,
+      "grad_norm": 0.5382979837388214,
+      "learning_rate": 2.8672190301196655e-06,
+      "loss": 0.6262,
+      "step": 700
+    },
+    {
+      "epoch": 6.805825242718447,
+      "grad_norm": 0.6097486357607171,
+      "learning_rate": 2.8621899307456376e-06,
+      "loss": 0.6926,
+      "step": 701
+    },
+    {
+      "epoch": 6.815533980582524,
+      "grad_norm": 0.5937487238676019,
+      "learning_rate": 2.8571593338296473e-06,
+      "loss": 0.6452,
+      "step": 702
+    },
+    {
+      "epoch": 6.825242718446602,
+      "grad_norm": 0.5132985773419537,
+      "learning_rate": 2.8521272601716376e-06,
+      "loss": 0.3947,
+      "step": 703
+    },
+    {
+      "epoch": 6.834951456310679,
+      "grad_norm": 0.6139505653367293,
+      "learning_rate": 2.8470937305776567e-06,
+      "loss": 0.8517,
+      "step": 704
+    },
+    {
+      "epoch": 6.844660194174757,
+      "grad_norm": 0.5362944543425278,
+      "learning_rate": 2.842058765859776e-06,
+      "loss": 0.5163,
+      "step": 705
+    },
+    {
+      "epoch": 6.854368932038835,
+      "grad_norm": 0.5271174346444056,
+      "learning_rate": 2.837022386835996e-06,
+      "loss": 0.5,
+      "step": 706
+    },
+    {
+      "epoch": 6.864077669902913,
+      "grad_norm": 0.5154656661302897,
+      "learning_rate": 2.8319846143301676e-06,
+      "loss": 0.4827,
+      "step": 707
+    },
+    {
+      "epoch": 6.87378640776699,
+      "grad_norm": 0.46846839589872596,
+      "learning_rate": 2.826945469171903e-06,
+      "loss": 0.5299,
+      "step": 708
+    },
+    {
+      "epoch": 6.883495145631068,
+      "grad_norm": 0.48940397947973197,
+      "learning_rate": 2.82190497219649e-06,
+      "loss": 0.4866,
+      "step": 709
+    },
+    {
+      "epoch": 6.893203883495145,
+      "grad_norm": 0.5913425687153925,
+      "learning_rate": 2.8168631442448046e-06,
+      "loss": 0.6692,
+      "step": 710
+    },
+    {
+      "epoch": 6.902912621359223,
+      "grad_norm": 0.5202999575403233,
+      "learning_rate": 2.8118200061632273e-06,
+      "loss": 0.5154,
+      "step": 711
+    },
+    {
+      "epoch": 6.9126213592233015,
+      "grad_norm": 0.5477222203271073,
+      "learning_rate": 2.8067755788035544e-06,
+      "loss": 0.8153,
+      "step": 712
+    },
+    {
+      "epoch": 6.922330097087379,
+      "grad_norm": 0.5412406643013349,
+      "learning_rate": 2.801729883022915e-06,
+      "loss": 0.5411,
+      "step": 713
+    },
+    {
+      "epoch": 6.932038834951456,
+      "grad_norm": 0.5157277209771828,
+      "learning_rate": 2.7966829396836804e-06,
+      "loss": 0.6641,
+      "step": 714
+    },
+    {
+      "epoch": 6.941747572815534,
+      "grad_norm": 0.585569198546952,
+      "learning_rate": 2.791634769653381e-06,
+      "loss": 0.5722,
+      "step": 715
+    },
+    {
+      "epoch": 6.951456310679612,
+      "grad_norm": 0.518068495388716,
+      "learning_rate": 2.78658539380462e-06,
+      "loss": 0.7203,
+      "step": 716
+    },
+    {
+      "epoch": 6.961165048543689,
+      "grad_norm": 0.5043833753127558,
+      "learning_rate": 2.781534833014985e-06,
+      "loss": 0.5923,
+      "step": 717
+    },
+    {
+      "epoch": 6.970873786407767,
+      "grad_norm": 0.50783508334967,
+      "learning_rate": 2.7764831081669635e-06,
+      "loss": 0.7243,
+      "step": 718
+    },
+    {
+      "epoch": 6.980582524271845,
+      "grad_norm": 0.523109640167941,
+      "learning_rate": 2.771430240147856e-06,
+      "loss": 0.4553,
+      "step": 719
+    },
+    {
+      "epoch": 6.990291262135923,
+      "grad_norm": 0.5501112401053831,
+      "learning_rate": 2.7663762498496905e-06,
+      "loss": 0.6997,
+      "step": 720
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.6013031662632521,
+      "learning_rate": 2.761321158169134e-06,
+      "loss": 0.5361,
+      "step": 721
+    },
+    {
+      "epoch": 7.009708737864078,
+      "grad_norm": 0.5233359625858339,
+      "learning_rate": 2.7562649860074077e-06,
+      "loss": 0.6119,
+      "step": 722
+    },
+    {
+      "epoch": 7.019417475728155,
+      "grad_norm": 0.517203456333739,
+      "learning_rate": 2.7512077542702005e-06,
+      "loss": 0.9122,
+      "step": 723
+    },
+    {
+      "epoch": 7.029126213592233,
+      "grad_norm": 0.5348182520574054,
+      "learning_rate": 2.746149483867582e-06,
+      "loss": 0.5543,
+      "step": 724
+    },
+    {
+      "epoch": 7.038834951456311,
+      "grad_norm": 0.5551672858403643,
+      "learning_rate": 2.741090195713917e-06,
+      "loss": 0.5383,
+      "step": 725
+    },
+    {
+      "epoch": 7.048543689320389,
+      "grad_norm": 0.5589328402666769,
+      "learning_rate": 2.736029910727777e-06,
+      "loss": 0.5495,
+      "step": 726
+    },
+    {
+      "epoch": 7.058252427184466,
+      "grad_norm": 0.9295358484591534,
+      "learning_rate": 2.730968649831858e-06,
+      "loss": 0.4708,
+      "step": 727
+    },
+    {
+      "epoch": 7.067961165048544,
+      "grad_norm": 0.5641431302765963,
+      "learning_rate": 2.7259064339528875e-06,
+      "loss": 0.7656,
+      "step": 728
+    },
+    {
+      "epoch": 7.077669902912621,
+      "grad_norm": 0.5150837177025048,
+      "learning_rate": 2.720843284021543e-06,
+      "loss": 0.5798,
+      "step": 729
+    },
+    {
+      "epoch": 7.087378640776699,
+      "grad_norm": 0.49904621404216537,
+      "learning_rate": 2.7157792209723654e-06,
+      "loss": 0.4392,
+      "step": 730
+    },
+    {
+      "epoch": 7.097087378640777,
+      "grad_norm": 0.5020142101967577,
+      "learning_rate": 2.7107142657436696e-06,
+      "loss": 0.6109,
+      "step": 731
+    },
+    {
+      "epoch": 7.106796116504855,
+      "grad_norm": 0.508200057178077,
+      "learning_rate": 2.705648439277459e-06,
+      "loss": 0.6078,
+      "step": 732
+    },
+    {
+      "epoch": 7.116504854368932,
+      "grad_norm": 0.48671766693455193,
+      "learning_rate": 2.7005817625193398e-06,
+      "loss": 0.4544,
+      "step": 733
+    },
+    {
+      "epoch": 7.12621359223301,
+      "grad_norm": 0.5649840605126365,
+      "learning_rate": 2.695514256418435e-06,
+      "loss": 0.4394,
+      "step": 734
+    },
+    {
+      "epoch": 7.135922330097087,
+      "grad_norm": 0.60875887483297,
+      "learning_rate": 2.6904459419272955e-06,
+      "loss": 0.6065,
+      "step": 735
+    },
+    {
+      "epoch": 7.145631067961165,
+      "grad_norm": 0.5974558558874661,
+      "learning_rate": 2.685376840001814e-06,
+      "loss": 0.6386,
+      "step": 736
+    },
+    {
+      "epoch": 7.155339805825243,
+      "grad_norm": 0.5530398341248075,
+      "learning_rate": 2.6803069716011405e-06,
+      "loss": 0.4432,
+      "step": 737
+    },
+    {
+      "epoch": 7.165048543689321,
+      "grad_norm": 0.5102871169083648,
+      "learning_rate": 2.6752363576875933e-06,
+      "loss": 0.662,
+      "step": 738
+    },
+    {
+      "epoch": 7.174757281553398,
+      "grad_norm": 0.5054879661184567,
+      "learning_rate": 2.6701650192265734e-06,
+      "loss": 0.6198,
+      "step": 739
+    },
+    {
+      "epoch": 7.184466019417476,
+      "grad_norm": 0.5089332252690112,
+      "learning_rate": 2.6650929771864776e-06,
+      "loss": 0.533,
+      "step": 740
+    },
+    {
+      "epoch": 7.194174757281553,
+      "grad_norm": 0.503785749624171,
+      "learning_rate": 2.660020252538611e-06,
+      "loss": 0.7417,
+      "step": 741
+    },
+    {
+      "epoch": 7.203883495145631,
+      "grad_norm": 0.5271479384718044,
+      "learning_rate": 2.6549468662571026e-06,
+      "loss": 0.5881,
+      "step": 742
+    },
+    {
+      "epoch": 7.213592233009709,
+      "grad_norm": 0.5635104879136624,
+      "learning_rate": 2.6498728393188157e-06,
+      "loss": 0.6668,
+      "step": 743
+    },
+    {
+      "epoch": 7.223300970873787,
+      "grad_norm": 0.48666286774042283,
+      "learning_rate": 2.6447981927032634e-06,
+      "loss": 0.5292,
+      "step": 744
+    },
+    {
+      "epoch": 7.233009708737864,
+      "grad_norm": 0.5252410611230779,
+      "learning_rate": 2.639722947392521e-06,
+      "loss": 0.7362,
+      "step": 745
+    },
+    {
+      "epoch": 7.242718446601942,
+      "grad_norm": 0.5055513935431378,
+      "learning_rate": 2.6346471243711376e-06,
+      "loss": 0.6224,
+      "step": 746
+    },
+    {
+      "epoch": 7.252427184466019,
+      "grad_norm": 0.4945880898396502,
+      "learning_rate": 2.629570744626052e-06,
+      "loss": 0.631,
+      "step": 747
+    },
+    {
+      "epoch": 7.262135922330097,
+      "grad_norm": 0.47517967426654084,
+      "learning_rate": 2.624493829146507e-06,
+      "loss": 0.6347,
+      "step": 748
+    },
+    {
+      "epoch": 7.271844660194175,
+      "grad_norm": 0.6161713749712753,
+      "learning_rate": 2.619416398923957e-06,
+      "loss": 0.8685,
+      "step": 749
+    },
+    {
+      "epoch": 7.281553398058253,
+      "grad_norm": 0.6026630303868498,
+      "learning_rate": 2.614338474951987e-06,
+      "loss": 0.4722,
+      "step": 750
+    },
+    {
+      "epoch": 7.29126213592233,
+      "grad_norm": 0.5081080900411881,
+      "learning_rate": 2.6092600782262213e-06,
+      "loss": 0.4761,
+      "step": 751
+    },
+    {
+      "epoch": 7.300970873786408,
+      "grad_norm": 0.4914556776074673,
+      "learning_rate": 2.6041812297442417e-06,
+      "loss": 0.5394,
+      "step": 752
+    },
+    {
+      "epoch": 7.310679611650485,
+      "grad_norm": 0.5174239393590538,
+      "learning_rate": 2.5991019505054965e-06,
+      "loss": 0.5791,
+      "step": 753
+    },
+    {
+      "epoch": 7.320388349514563,
+      "grad_norm": 0.577016310911,
+      "learning_rate": 2.5940222615112143e-06,
+      "loss": 0.4506,
+      "step": 754
+    },
+    {
+      "epoch": 7.330097087378641,
+      "grad_norm": 0.4931993425278694,
+      "learning_rate": 2.5889421837643186e-06,
+      "loss": 0.501,
+      "step": 755
+    },
+    {
+      "epoch": 7.339805825242719,
+      "grad_norm": 0.4880724860087956,
+      "learning_rate": 2.5838617382693415e-06,
+      "loss": 0.6661,
+      "step": 756
+    },
+    {
+      "epoch": 7.349514563106796,
+      "grad_norm": 0.5210458539063239,
+      "learning_rate": 2.5787809460323337e-06,
+      "loss": 0.6234,
+      "step": 757
+    },
+    {
+      "epoch": 7.359223300970874,
+      "grad_norm": 0.5135992991117062,
+      "learning_rate": 2.57369982806078e-06,
+      "loss": 0.641,
+      "step": 758
+    },
+    {
+      "epoch": 7.368932038834951,
+      "grad_norm": 0.5505536500994148,
+      "learning_rate": 2.5686184053635127e-06,
+      "loss": 0.5763,
+      "step": 759
+    },
+    {
+      "epoch": 7.378640776699029,
+      "grad_norm": 0.4778646758208665,
+      "learning_rate": 2.563536698950624e-06,
+      "loss": 0.4859,
+      "step": 760
+    },
+    {
+      "epoch": 7.388349514563107,
+      "grad_norm": 0.540107781162827,
+      "learning_rate": 2.5584547298333772e-06,
+      "loss": 0.4128,
+      "step": 761
+    },
+    {
+      "epoch": 7.398058252427185,
+      "grad_norm": 0.5303028011574588,
+      "learning_rate": 2.5533725190241255e-06,
+      "loss": 0.3898,
+      "step": 762
+    },
+    {
+      "epoch": 7.407766990291262,
+      "grad_norm": 0.5105167298780968,
+      "learning_rate": 2.5482900875362184e-06,
+      "loss": 0.5078,
+      "step": 763
+    },
+    {
+      "epoch": 7.41747572815534,
+      "grad_norm": 0.565588338746564,
+      "learning_rate": 2.543207456383919e-06,
+      "loss": 0.6164,
+      "step": 764
+    },
+    {
+      "epoch": 7.427184466019417,
+      "grad_norm": 0.5525554215167167,
+      "learning_rate": 2.538124646582315e-06,
+      "loss": 0.785,
+      "step": 765
+    },
+    {
+      "epoch": 7.436893203883495,
+      "grad_norm": 0.5695569930727707,
+      "learning_rate": 2.533041679147235e-06,
+      "loss": 0.5288,
+      "step": 766
+    },
+    {
+      "epoch": 7.446601941747573,
+      "grad_norm": 0.5101624379639444,
+      "learning_rate": 2.527958575095157e-06,
+      "loss": 0.5563,
+      "step": 767
+    },
+    {
+      "epoch": 7.456310679611651,
+      "grad_norm": 0.5432188733614521,
+      "learning_rate": 2.522875355443124e-06,
+      "loss": 0.7357,
+      "step": 768
+    },
+    {
+      "epoch": 7.466019417475728,
+      "grad_norm": 0.5464004709948751,
+      "learning_rate": 2.5177920412086586e-06,
+      "loss": 0.7825,
+      "step": 769
+    },
+    {
+      "epoch": 7.475728155339806,
+      "grad_norm": 0.529743426710544,
+      "learning_rate": 2.512708653409674e-06,
+      "loss": 0.6695,
+      "step": 770
+    },
+    {
+      "epoch": 7.485436893203883,
+      "grad_norm": 0.5145423159262458,
+      "learning_rate": 2.507625213064386e-06,
+      "loss": 0.5038,
+      "step": 771
+    },
+    {
+      "epoch": 7.495145631067961,
+      "grad_norm": 0.47269595307770995,
+      "learning_rate": 2.5025417411912307e-06,
+      "loss": 0.4515,
+      "step": 772
+    },
+    {
+      "epoch": 7.504854368932039,
+      "grad_norm": 0.5580818879826553,
+      "learning_rate": 2.4974582588087697e-06,
+      "loss": 0.6909,
+      "step": 773
+    },
+    {
+      "epoch": 7.514563106796117,
+      "grad_norm": 0.5100233581400443,
+      "learning_rate": 2.492374786935614e-06,
+      "loss": 0.6848,
+      "step": 774
+    },
+    {
+      "epoch": 7.524271844660194,
+      "grad_norm": 0.5522174720334939,
+      "learning_rate": 2.487291346590326e-06,
+      "loss": 0.7382,
+      "step": 775
+    },
+    {
+      "epoch": 7.533980582524272,
+      "grad_norm": 0.4890510705027455,
+      "learning_rate": 2.4822079587913414e-06,
+      "loss": 0.5869,
+      "step": 776
+    },
+    {
+      "epoch": 7.543689320388349,
+      "grad_norm": 0.518568061347995,
+      "learning_rate": 2.4771246445568763e-06,
+      "loss": 0.4913,
+      "step": 777
+    },
+    {
+      "epoch": 7.553398058252427,
+      "grad_norm": 0.5727634652449346,
+      "learning_rate": 2.472041424904844e-06,
+      "loss": 0.5031,
+      "step": 778
+    },
+    {
+      "epoch": 7.563106796116505,
+      "grad_norm": 0.49387541758342474,
+      "learning_rate": 2.466958320852766e-06,
+      "loss": 0.5388,
+      "step": 779
+    },
+    {
+      "epoch": 7.572815533980583,
+      "grad_norm": 0.5217948655090912,
+      "learning_rate": 2.4618753534176854e-06,
+      "loss": 0.6361,
+      "step": 780
+    },
+    {
+      "epoch": 7.58252427184466,
+      "grad_norm": 0.5332175891151626,
+      "learning_rate": 2.4567925436160823e-06,
+      "loss": 0.6208,
+      "step": 781
+    },
+    {
+      "epoch": 7.592233009708738,
+      "grad_norm": 0.5318535876006302,
+      "learning_rate": 2.4517099124637824e-06,
+      "loss": 0.7728,
+      "step": 782
+    },
+    {
+      "epoch": 7.601941747572815,
+      "grad_norm": 0.5502888061073115,
+      "learning_rate": 2.4466274809758757e-06,
+      "loss": 0.6001,
+      "step": 783
+    },
+    {
+      "epoch": 7.611650485436893,
+      "grad_norm": 0.4936319038862824,
+      "learning_rate": 2.4415452701666236e-06,
+      "loss": 0.5221,
+      "step": 784
+    },
+    {
+      "epoch": 7.621359223300971,
+      "grad_norm": 0.5022890404063105,
+      "learning_rate": 2.436463301049378e-06,
+      "loss": 0.6604,
+      "step": 785
+    },
+    {
+      "epoch": 7.631067961165049,
+      "grad_norm": 0.5272732094633668,
+      "learning_rate": 2.431381594636488e-06,
+      "loss": 0.6919,
+      "step": 786
+    },
+    {
+      "epoch": 7.640776699029126,
+      "grad_norm": 0.5919653641625273,
+      "learning_rate": 2.42630017193922e-06,
+      "loss": 0.8888,
+      "step": 787
+    },
+    {
+      "epoch": 7.650485436893204,
+      "grad_norm": 0.5004825929075615,
+      "learning_rate": 2.4212190539676667e-06,
+      "loss": 0.3034,
+      "step": 788
+    },
+    {
+      "epoch": 7.660194174757281,
+      "grad_norm": 0.5188376934752825,
+      "learning_rate": 2.4161382617306585e-06,
+      "loss": 0.5967,
+      "step": 789
+    },
+    {
+      "epoch": 7.669902912621359,
+      "grad_norm": 0.50648375774632,
+      "learning_rate": 2.4110578162356814e-06,
+      "loss": 0.5533,
+      "step": 790
+    },
+    {
+      "epoch": 7.679611650485437,
+      "grad_norm": 0.4995699684525013,
+      "learning_rate": 2.405977738488786e-06,
+      "loss": 0.5644,
+      "step": 791
+    },
+    {
+      "epoch": 7.689320388349515,
+      "grad_norm": 0.5177586536034464,
+      "learning_rate": 2.4008980494945044e-06,
+      "loss": 0.6069,
+      "step": 792
+    },
+    {
+      "epoch": 7.699029126213592,
+      "grad_norm": 0.5143304969975345,
+      "learning_rate": 2.3958187702557587e-06,
+      "loss": 0.4916,
+      "step": 793
+    },
+    {
+      "epoch": 7.70873786407767,
+      "grad_norm": 0.5576884681772939,
+      "learning_rate": 2.39073992177378e-06,
+      "loss": 0.4913,
+      "step": 794
+    },
+    {
+      "epoch": 7.718446601941747,
+      "grad_norm": 0.5320452230495075,
+      "learning_rate": 2.385661525048014e-06,
+      "loss": 0.6293,
+      "step": 795
+    },
+    {
+      "epoch": 7.728155339805825,
+      "grad_norm": 0.7020840522130559,
+      "learning_rate": 2.3805836010760435e-06,
+      "loss": 0.6646,
+      "step": 796
+    },
+    {
+      "epoch": 7.737864077669903,
+      "grad_norm": 0.5230839966091113,
+      "learning_rate": 2.375506170853494e-06,
+      "loss": 0.5772,
+      "step": 797
+    },
+    {
+      "epoch": 7.747572815533981,
+      "grad_norm": 0.5734597379091115,
+      "learning_rate": 2.3704292553739487e-06,
+      "loss": 0.4759,
+      "step": 798
+    },
+    {
+      "epoch": 7.757281553398058,
+      "grad_norm": 0.5117329377962575,
+      "learning_rate": 2.3653528756288636e-06,
+      "loss": 0.4828,
+      "step": 799
+    },
+    {
+      "epoch": 7.766990291262136,
+      "grad_norm": 0.47543907618671766,
+      "learning_rate": 2.3602770526074804e-06,
+      "loss": 0.6375,
+      "step": 800
+    },
+    {
+      "epoch": 7.776699029126213,
+      "grad_norm": 0.5109119486237039,
+      "learning_rate": 2.3552018072967375e-06,
+      "loss": 0.6483,
+      "step": 801
+    },
+    {
+      "epoch": 7.786407766990291,
+      "grad_norm": 0.5323210647105892,
+      "learning_rate": 2.3501271606811848e-06,
+      "loss": 0.4874,
+      "step": 802
+    },
+    {
+      "epoch": 7.796116504854369,
+      "grad_norm": 0.5730319631624969,
+      "learning_rate": 2.345053133742898e-06,
+      "loss": 0.7191,
+      "step": 803
+    },
+    {
+      "epoch": 7.805825242718447,
+      "grad_norm": 0.5071812370282966,
+      "learning_rate": 2.3399797474613894e-06,
+      "loss": 0.3434,
+      "step": 804
+    },
+    {
+      "epoch": 7.815533980582524,
+      "grad_norm": 0.5461829256062627,
+      "learning_rate": 2.334907022813523e-06,
+      "loss": 0.485,
+      "step": 805
+    },
+    {
+      "epoch": 7.825242718446602,
+      "grad_norm": 0.5636593344058092,
+      "learning_rate": 2.329834980773427e-06,
+      "loss": 0.5785,
+      "step": 806
+    },
+    {
+      "epoch": 7.834951456310679,
+      "grad_norm": 0.5773502744704323,
+      "learning_rate": 2.324763642312407e-06,
+      "loss": 0.6205,
+      "step": 807
+    },
+    {
+      "epoch": 7.844660194174757,
+      "grad_norm": 0.48736301511057184,
+      "learning_rate": 2.3196930283988603e-06,
+      "loss": 0.6431,
+      "step": 808
+    },
+    {
+      "epoch": 7.854368932038835,
+      "grad_norm": 0.5161473184954976,
+      "learning_rate": 2.3146231599981865e-06,
+      "loss": 0.5869,
+      "step": 809
+    },
+    {
+      "epoch": 7.864077669902913,
+      "grad_norm": 0.5499369220986959,
+      "learning_rate": 2.3095540580727054e-06,
+      "loss": 0.7587,
+      "step": 810
+    },
+    {
+      "epoch": 7.87378640776699,
+      "grad_norm": 0.5625597573563249,
+      "learning_rate": 2.304485743581566e-06,
+      "loss": 0.6425,
+      "step": 811
+    },
+    {
+      "epoch": 7.883495145631068,
+      "grad_norm": 0.5629938236856635,
+      "learning_rate": 2.299418237480661e-06,
+      "loss": 0.4925,
+      "step": 812
+    },
+    {
+      "epoch": 7.893203883495145,
+      "grad_norm": 0.49533812712631564,
+      "learning_rate": 2.294351560722542e-06,
+      "loss": 0.6207,
+      "step": 813
+    },
+    {
+      "epoch": 7.902912621359223,
+      "grad_norm": 0.6428692294617694,
+      "learning_rate": 2.2892857342563316e-06,
+      "loss": 0.4045,
+      "step": 814
+    },
+    {
+      "epoch": 7.9126213592233015,
+      "grad_norm": 0.5474333139164419,
+      "learning_rate": 2.2842207790276355e-06,
+      "loss": 0.6509,
+      "step": 815
+    },
+    {
+      "epoch": 7.922330097087379,
+      "grad_norm": 0.5193991986662039,
+      "learning_rate": 2.279156715978457e-06,
+      "loss": 0.7148,
+      "step": 816
+    },
+    {
+      "epoch": 7.932038834951456,
+      "grad_norm": 0.499433426250342,
+      "learning_rate": 2.274093566047113e-06,
+      "loss": 0.5728,
+      "step": 817
+    },
+    {
+      "epoch": 7.941747572815534,
+      "grad_norm": 0.5512612849904238,
+      "learning_rate": 2.2690313501681426e-06,
+      "loss": 0.5075,
+      "step": 818
+    },
+    {
+      "epoch": 7.951456310679612,
+      "grad_norm": 0.5208798190855127,
+      "learning_rate": 2.263970089272223e-06,
+      "loss": 0.5729,
+      "step": 819
+    },
+    {
+      "epoch": 7.961165048543689,
+      "grad_norm": 0.5298051120860908,
+      "learning_rate": 2.2589098042860838e-06,
+      "loss": 0.5755,
+      "step": 820
+    },
+    {
+      "epoch": 7.970873786407767,
+      "grad_norm": 0.6511172387364248,
+      "learning_rate": 2.2538505161324186e-06,
+      "loss": 0.4285,
+      "step": 821
+    },
+    {
+      "epoch": 7.980582524271845,
+      "grad_norm": 0.5520636005584219,
+      "learning_rate": 2.2487922457298007e-06,
+      "loss": 0.4915,
+      "step": 822
+    },
+    {
+      "epoch": 7.990291262135923,
+      "grad_norm": 0.6008177376268479,
+      "learning_rate": 2.243735013992593e-06,
+      "loss": 0.7512,
+      "step": 823
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.5164857595576907,
+      "learning_rate": 2.238678841830867e-06,
+      "loss": 0.6132,
+      "step": 824
+    },
+    {
+      "epoch": 8.009708737864077,
+      "grad_norm": 0.5927201492916486,
+      "learning_rate": 2.2336237501503103e-06,
+      "loss": 0.4526,
+      "step": 825
+    },
+    {
+      "epoch": 8.019417475728156,
+      "grad_norm": 0.5149776302383433,
+      "learning_rate": 2.2285697598521446e-06,
+      "loss": 0.5165,
+      "step": 826
+    },
+    {
+      "epoch": 8.029126213592233,
+      "grad_norm": 0.4668940317507629,
+      "learning_rate": 2.2235168918330374e-06,
+      "loss": 0.567,
+      "step": 827
+    },
+    {
+      "epoch": 8.03883495145631,
+      "grad_norm": 0.5154594445027514,
+      "learning_rate": 2.2184651669850164e-06,
+      "loss": 0.7516,
+      "step": 828
+    },
+    {
+      "epoch": 8.048543689320388,
+      "grad_norm": 0.4940207401959282,
+      "learning_rate": 2.2134146061953814e-06,
+      "loss": 0.2917,
+      "step": 829
+    },
+    {
+      "epoch": 8.058252427184467,
+      "grad_norm": 0.5178752504635592,
+      "learning_rate": 2.2083652303466196e-06,
+      "loss": 0.8172,
+      "step": 830
+    },
+    {
+      "epoch": 8.067961165048544,
+      "grad_norm": 0.546827519101471,
+      "learning_rate": 2.20331706031632e-06,
+      "loss": 0.7023,
+      "step": 831
+    },
+    {
+      "epoch": 8.077669902912621,
+      "grad_norm": 0.4582055070562166,
+      "learning_rate": 2.1982701169770853e-06,
+      "loss": 0.4083,
+      "step": 832
+    },
+    {
+      "epoch": 8.087378640776699,
+      "grad_norm": 0.4683301177297805,
+      "learning_rate": 2.1932244211964456e-06,
+      "loss": 0.5582,
+      "step": 833
+    },
+    {
+      "epoch": 8.097087378640778,
+      "grad_norm": 0.5178375602465826,
+      "learning_rate": 2.1881799938367735e-06,
+      "loss": 0.356,
+      "step": 834
+    },
+    {
+      "epoch": 8.106796116504855,
+      "grad_norm": 0.504467565940944,
+      "learning_rate": 2.1831368557551962e-06,
+      "loss": 0.4254,
+      "step": 835
+    },
+    {
+      "epoch": 8.116504854368932,
+      "grad_norm": 0.5226565525141085,
+      "learning_rate": 2.1780950278035114e-06,
+      "loss": 0.6666,
+      "step": 836
+    },
+    {
+      "epoch": 8.12621359223301,
+      "grad_norm": 0.587266082983635,
+      "learning_rate": 2.173054530828098e-06,
+      "loss": 0.4792,
+      "step": 837
+    },
+    {
+      "epoch": 8.135922330097088,
+      "grad_norm": 0.5562710685791401,
+      "learning_rate": 2.168015385669833e-06,
+      "loss": 0.4674,
+      "step": 838
+    },
+    {
+      "epoch": 8.145631067961165,
+      "grad_norm": 0.5181070514066362,
+      "learning_rate": 2.162977613164005e-06,
+      "loss": 0.7922,
+      "step": 839
+    },
+    {
+      "epoch": 8.155339805825243,
+      "grad_norm": 0.5332598920234418,
+      "learning_rate": 2.157941234140225e-06,
+      "loss": 0.8334,
+      "step": 840
+    },
+    {
+      "epoch": 8.16504854368932,
+      "grad_norm": 0.508519951977822,
+      "learning_rate": 2.1529062694223437e-06,
+      "loss": 0.562,
+      "step": 841
+    },
+    {
+      "epoch": 8.174757281553399,
+      "grad_norm": 0.5623882061470568,
+      "learning_rate": 2.147872739828364e-06,
+      "loss": 0.3732,
+      "step": 842
+    },
+    {
+      "epoch": 8.184466019417476,
+      "grad_norm": 0.581097080186488,
+      "learning_rate": 2.142840666170354e-06,
+      "loss": 0.3534,
+      "step": 843
+    },
+    {
+      "epoch": 8.194174757281553,
+      "grad_norm": 0.5552223251444132,
+      "learning_rate": 2.1378100692543637e-06,
+      "loss": 0.5417,
+      "step": 844
+    },
+    {
+      "epoch": 8.20388349514563,
+      "grad_norm": 0.6036836009498026,
+      "learning_rate": 2.1327809698803354e-06,
+      "loss": 0.6192,
+      "step": 845
+    },
+    {
+      "epoch": 8.21359223300971,
+      "grad_norm": 0.4918145204654767,
+      "learning_rate": 2.1277533888420203e-06,
+      "loss": 0.5632,
+      "step": 846
+    },
+    {
+      "epoch": 8.223300970873787,
+      "grad_norm": 0.54420380543783,
+      "learning_rate": 2.1227273469268932e-06,
+      "loss": 0.5219,
+      "step": 847
+    },
+    {
+      "epoch": 8.233009708737864,
+      "grad_norm": 0.5919068430953334,
+      "learning_rate": 2.117702864916063e-06,
+      "loss": 0.4869,
+      "step": 848
+    },
+    {
+      "epoch": 8.242718446601941,
+      "grad_norm": 0.5748938854690556,
+      "learning_rate": 2.1126799635841897e-06,
+      "loss": 0.6345,
+      "step": 849
+    },
+    {
+      "epoch": 8.25242718446602,
+      "grad_norm": 0.5669754510278896,
+      "learning_rate": 2.1076586636993975e-06,
+      "loss": 0.6422,
+      "step": 850
+    },
+    {
+      "epoch": 8.262135922330097,
+      "grad_norm": 0.5578707537573963,
+      "learning_rate": 2.102638986023189e-06,
+      "loss": 0.5943,
+      "step": 851
+    },
+    {
+      "epoch": 8.271844660194175,
+      "grad_norm": 0.5244178006438305,
+      "learning_rate": 2.0976209513103604e-06,
+      "loss": 0.5996,
+      "step": 852
+    },
+    {
+      "epoch": 8.281553398058252,
+      "grad_norm": 0.4735653903118842,
+      "learning_rate": 2.0926045803089135e-06,
+      "loss": 0.5118,
+      "step": 853
+    },
+    {
+      "epoch": 8.29126213592233,
+      "grad_norm": 0.5610192165309852,
+      "learning_rate": 2.087589893759972e-06,
+      "loss": 0.6315,
+      "step": 854
+    },
+    {
+      "epoch": 8.300970873786408,
+      "grad_norm": 0.586016198505159,
+      "learning_rate": 2.0825769123976954e-06,
+      "loss": 0.484,
+      "step": 855
+    },
+    {
+      "epoch": 8.310679611650485,
+      "grad_norm": 0.5474063939519695,
+      "learning_rate": 2.077565656949191e-06,
+      "loss": 0.6091,
+      "step": 856
+    },
+    {
+      "epoch": 8.320388349514563,
+      "grad_norm": 0.5292483811678337,
+      "learning_rate": 2.072556148134431e-06,
+      "loss": 0.6521,
+      "step": 857
+    },
+    {
+      "epoch": 8.330097087378642,
+      "grad_norm": 0.4874932647375569,
+      "learning_rate": 2.0675484066661666e-06,
+      "loss": 0.5259,
+      "step": 858
+    },
+    {
+      "epoch": 8.339805825242719,
+      "grad_norm": 0.4851303580845472,
+      "learning_rate": 2.0625424532498407e-06,
+      "loss": 0.4161,
+      "step": 859
+    },
+    {
+      "epoch": 8.349514563106796,
+      "grad_norm": 0.5545793735034753,
+      "learning_rate": 2.057538308583502e-06,
+      "loss": 0.5025,
+      "step": 860
+    },
+    {
+      "epoch": 8.359223300970873,
+      "grad_norm": 0.5080809023551803,
+      "learning_rate": 2.0525359933577243e-06,
+      "loss": 0.8568,
+      "step": 861
+    },
+    {
+      "epoch": 8.368932038834952,
+      "grad_norm": 0.46911329394510076,
+      "learning_rate": 2.047535528255514e-06,
+      "loss": 0.6254,
+      "step": 862
+    },
+    {
+      "epoch": 8.37864077669903,
+      "grad_norm": 0.48403193573667014,
+      "learning_rate": 2.0425369339522276e-06,
+      "loss": 0.3043,
+      "step": 863
+    },
+    {
+      "epoch": 8.388349514563107,
+      "grad_norm": 0.5370398217757195,
+      "learning_rate": 2.0375402311154886e-06,
+      "loss": 0.5867,
+      "step": 864
+    },
+    {
+      "epoch": 8.398058252427184,
+      "grad_norm": 0.6080681817574075,
+      "learning_rate": 2.0325454404050983e-06,
+      "loss": 0.3763,
+      "step": 865
+    },
+    {
+      "epoch": 8.407766990291263,
+      "grad_norm": 0.530337907125514,
+      "learning_rate": 2.0275525824729523e-06,
+      "loss": 0.7515,
+      "step": 866
+    },
+    {
+      "epoch": 8.41747572815534,
+      "grad_norm": 0.4954396437597861,
+      "learning_rate": 2.022561677962955e-06,
+      "loss": 0.5687,
+      "step": 867
+    },
+    {
+      "epoch": 8.427184466019417,
+      "grad_norm": 0.5599603984905624,
+      "learning_rate": 2.017572747510934e-06,
+      "loss": 0.4983,
+      "step": 868
+    },
+    {
+      "epoch": 8.436893203883495,
+      "grad_norm": 0.5593697940698449,
+      "learning_rate": 2.012585811744555e-06,
+      "loss": 0.6475,
+      "step": 869
+    },
+    {
+      "epoch": 8.446601941747574,
+      "grad_norm": 0.5608755353144496,
+      "learning_rate": 2.0076008912832355e-06,
+      "loss": 0.7165,
+      "step": 870
+    },
+    {
+      "epoch": 8.45631067961165,
+      "grad_norm": 0.49618292660157465,
+      "learning_rate": 2.002618006738061e-06,
+      "loss": 0.6876,
+      "step": 871
+    },
+    {
+      "epoch": 8.466019417475728,
+      "grad_norm": 0.52539938762679,
+      "learning_rate": 1.9976371787116992e-06,
+      "loss": 0.6282,
+      "step": 872
+    },
+    {
+      "epoch": 8.475728155339805,
+      "grad_norm": 0.6197757961176346,
+      "learning_rate": 1.9926584277983134e-06,
+      "loss": 0.4668,
+      "step": 873
+    },
+    {
+      "epoch": 8.485436893203884,
+      "grad_norm": 0.5982098576330306,
+      "learning_rate": 1.9876817745834805e-06,
+      "loss": 0.7359,
+      "step": 874
+    },
+    {
+      "epoch": 8.495145631067961,
+      "grad_norm": 0.5184171696351032,
+      "learning_rate": 1.9827072396441044e-06,
+      "loss": 0.629,
+      "step": 875
+    },
+    {
+      "epoch": 8.504854368932039,
+      "grad_norm": 0.5704731732012165,
+      "learning_rate": 1.9777348435483285e-06,
+      "loss": 0.5434,
+      "step": 876
+    },
+    {
+      "epoch": 8.514563106796116,
+      "grad_norm": 0.5914779709255895,
+      "learning_rate": 1.972764606855454e-06,
+      "loss": 0.8657,
+      "step": 877
+    },
+    {
+      "epoch": 8.524271844660195,
+      "grad_norm": 0.4998262061570929,
+      "learning_rate": 1.9677965501158534e-06,
+      "loss": 0.5069,
+      "step": 878
+    },
+    {
+      "epoch": 8.533980582524272,
+      "grad_norm": 0.5060351233776778,
+      "learning_rate": 1.9628306938708857e-06,
+      "loss": 0.5947,
+      "step": 879
+    },
+    {
+      "epoch": 8.54368932038835,
+      "grad_norm": 0.477718552398954,
+      "learning_rate": 1.957867058652812e-06,
+      "loss": 0.5787,
+      "step": 880
+    },
+    {
+      "epoch": 8.553398058252426,
+      "grad_norm": 0.5163268989253287,
+      "learning_rate": 1.952905664984709e-06,
+      "loss": 0.6632,
+      "step": 881
+    },
+    {
+      "epoch": 8.563106796116505,
+      "grad_norm": 0.5530459189954213,
+      "learning_rate": 1.947946533380387e-06,
+      "loss": 0.455,
+      "step": 882
+    },
+    {
+      "epoch": 8.572815533980583,
+      "grad_norm": 0.5116903651673903,
+      "learning_rate": 1.9429896843443025e-06,
+      "loss": 0.4705,
+      "step": 883
+    },
+    {
+      "epoch": 8.58252427184466,
+      "grad_norm": 0.480475135243548,
+      "learning_rate": 1.938035138371474e-06,
+      "loss": 0.7274,
+      "step": 884
+    },
+    {
+      "epoch": 8.592233009708737,
+      "grad_norm": 0.5297320744221382,
+      "learning_rate": 1.933082915947398e-06,
+      "loss": 0.4558,
+      "step": 885
+    },
+    {
+      "epoch": 8.601941747572816,
+      "grad_norm": 0.505128857471151,
+      "learning_rate": 1.928133037547963e-06,
+      "loss": 0.3586,
+      "step": 886
+    },
+    {
+      "epoch": 8.611650485436893,
+      "grad_norm": 0.5928628788232783,
+      "learning_rate": 1.9231855236393677e-06,
+      "loss": 0.6312,
+      "step": 887
+    },
+    {
+      "epoch": 8.62135922330097,
+      "grad_norm": 0.649524612781142,
+      "learning_rate": 1.9182403946780316e-06,
+      "loss": 0.6427,
+      "step": 888
+    },
+    {
+      "epoch": 8.631067961165048,
+      "grad_norm": 0.5876898341916988,
+      "learning_rate": 1.9132976711105146e-06,
+      "loss": 0.6925,
+      "step": 889
+    },
+    {
+      "epoch": 8.640776699029127,
+      "grad_norm": 0.58652849068603,
+      "learning_rate": 1.9083573733734328e-06,
+      "loss": 0.6009,
+      "step": 890
+    },
+    {
+      "epoch": 8.650485436893204,
+      "grad_norm": 0.5556588390992402,
+      "learning_rate": 1.903419521893369e-06,
+      "loss": 0.5995,
+      "step": 891
+    },
+    {
+      "epoch": 8.660194174757281,
+      "grad_norm": 0.5220315347935208,
+      "learning_rate": 1.898484137086794e-06,
+      "loss": 0.5321,
+      "step": 892
+    },
+    {
+      "epoch": 8.669902912621358,
+      "grad_norm": 0.569643001305556,
+      "learning_rate": 1.8935512393599784e-06,
+      "loss": 0.4225,
+      "step": 893
+    },
+    {
+      "epoch": 8.679611650485437,
+      "grad_norm": 0.5373129439323331,
+      "learning_rate": 1.8886208491089095e-06,
+      "loss": 0.5555,
+      "step": 894
+    },
+    {
+      "epoch": 8.689320388349515,
+      "grad_norm": 0.5400895369165816,
+      "learning_rate": 1.8836929867192077e-06,
+      "loss": 0.4617,
+      "step": 895
+    },
+    {
+      "epoch": 8.699029126213592,
+      "grad_norm": 0.487412932412484,
+      "learning_rate": 1.8787676725660405e-06,
+      "loss": 0.4752,
+      "step": 896
+    },
+    {
+      "epoch": 8.70873786407767,
+      "grad_norm": 0.48707203079572564,
+      "learning_rate": 1.8738449270140404e-06,
+      "loss": 0.6907,
+      "step": 897
+    },
+    {
+      "epoch": 8.718446601941748,
+      "grad_norm": 0.48910391480682963,
+      "learning_rate": 1.8689247704172187e-06,
+      "loss": 0.5326,
+      "step": 898
+    },
+    {
+      "epoch": 8.728155339805825,
+      "grad_norm": 0.5147613851501489,
+      "learning_rate": 1.8640072231188825e-06,
+      "loss": 0.5996,
+      "step": 899
+    },
+    {
+      "epoch": 8.737864077669903,
+      "grad_norm": 0.4917113961710611,
+      "learning_rate": 1.8590923054515504e-06,
+      "loss": 0.5886,
+      "step": 900
+    },
+    {
+      "epoch": 8.74757281553398,
+      "grad_norm": 0.5083586615609125,
+      "learning_rate": 1.8541800377368673e-06,
+      "loss": 0.6473,
+      "step": 901
+    },
+    {
+      "epoch": 8.757281553398059,
+      "grad_norm": 0.5294044154978094,
+      "learning_rate": 1.8492704402855229e-06,
+      "loss": 0.6822,
+      "step": 902
+    },
+    {
+      "epoch": 8.766990291262136,
+      "grad_norm": 0.5804348364355945,
+      "learning_rate": 1.8443635333971643e-06,
+      "loss": 0.421,
+      "step": 903
+    },
+    {
+      "epoch": 8.776699029126213,
+      "grad_norm": 0.5203615517682456,
+      "learning_rate": 1.8394593373603173e-06,
+      "loss": 0.6694,
+      "step": 904
+    },
+    {
+      "epoch": 8.78640776699029,
+      "grad_norm": 0.5109166593919562,
+      "learning_rate": 1.8345578724522957e-06,
+      "loss": 0.5472,
+      "step": 905
+    },
+    {
+      "epoch": 8.79611650485437,
+      "grad_norm": 0.560325487440104,
+      "learning_rate": 1.8296591589391227e-06,
+      "loss": 0.5814,
+      "step": 906
+    },
+    {
+      "epoch": 8.805825242718447,
+      "grad_norm": 0.5421411914919385,
+      "learning_rate": 1.8247632170754443e-06,
+      "loss": 0.5985,
+      "step": 907
+    },
+    {
+      "epoch": 8.815533980582524,
+      "grad_norm": 0.5770835806642348,
+      "learning_rate": 1.8198700671044477e-06,
+      "loss": 0.5228,
+      "step": 908
+    },
+    {
+      "epoch": 8.825242718446601,
+      "grad_norm": 0.4895076081734303,
+      "learning_rate": 1.8149797292577757e-06,
+      "loss": 0.6435,
+      "step": 909
+    },
+    {
+      "epoch": 8.83495145631068,
+      "grad_norm": 0.4873199553388694,
+      "learning_rate": 1.8100922237554442e-06,
+      "loss": 0.6997,
+      "step": 910
+    },
+    {
+      "epoch": 8.844660194174757,
+      "grad_norm": 0.5490067414210736,
+      "learning_rate": 1.8052075708057581e-06,
+      "loss": 0.3939,
+      "step": 911
+    },
+    {
+      "epoch": 8.854368932038835,
+      "grad_norm": 0.6321945383667227,
+      "learning_rate": 1.8003257906052284e-06,
+      "loss": 0.4784,
+      "step": 912
+    },
+    {
+      "epoch": 8.864077669902912,
+      "grad_norm": 0.604869131460347,
+      "learning_rate": 1.7954469033384868e-06,
+      "loss": 0.5676,
+      "step": 913
+    },
+    {
+      "epoch": 8.87378640776699,
+      "grad_norm": 0.5447552422923585,
+      "learning_rate": 1.790570929178206e-06,
+      "loss": 0.713,
+      "step": 914
+    },
+    {
+      "epoch": 8.883495145631068,
+      "grad_norm": 0.5302556571289168,
+      "learning_rate": 1.7856978882850112e-06,
+      "loss": 0.5198,
+      "step": 915
+    },
+    {
+      "epoch": 8.893203883495145,
+      "grad_norm": 0.4973051756135799,
+      "learning_rate": 1.780827800807401e-06,
+      "loss": 0.5464,
+      "step": 916
+    },
+    {
+      "epoch": 8.902912621359224,
+      "grad_norm": 0.6496512810930873,
+      "learning_rate": 1.7759606868816623e-06,
+      "loss": 0.6319,
+      "step": 917
+    },
+    {
+      "epoch": 8.912621359223301,
+      "grad_norm": 0.5065859495112356,
+      "learning_rate": 1.771096566631788e-06,
+      "loss": 0.4308,
+      "step": 918
+    },
+    {
+      "epoch": 8.922330097087379,
+      "grad_norm": 0.5437885245827361,
+      "learning_rate": 1.766235460169392e-06,
+      "loss": 0.6901,
+      "step": 919
+    },
+    {
+      "epoch": 8.932038834951456,
+      "grad_norm": 0.6349112557712641,
+      "learning_rate": 1.7613773875936274e-06,
+      "loss": 0.5075,
+      "step": 920
+    },
+    {
+      "epoch": 8.941747572815533,
+      "grad_norm": 0.5653114938530557,
+      "learning_rate": 1.7565223689911038e-06,
+      "loss": 0.5145,
+      "step": 921
+    },
+    {
+      "epoch": 8.951456310679612,
+      "grad_norm": 0.5191163631219877,
+      "learning_rate": 1.7516704244358018e-06,
+      "loss": 0.4828,
+      "step": 922
+    },
+    {
+      "epoch": 8.96116504854369,
+      "grad_norm": 0.49692045826136555,
+      "learning_rate": 1.7468215739889941e-06,
+      "loss": 0.4133,
+      "step": 923
+    },
+    {
+      "epoch": 8.970873786407767,
+      "grad_norm": 0.580247056070986,
+      "learning_rate": 1.741975837699158e-06,
+      "loss": 0.6603,
+      "step": 924
+    },
+    {
+      "epoch": 8.980582524271846,
+      "grad_norm": 0.562012613485115,
+      "learning_rate": 1.7371332356018972e-06,
+      "loss": 0.4871,
+      "step": 925
+    },
+    {
+      "epoch": 8.990291262135923,
+      "grad_norm": 0.598353497274599,
+      "learning_rate": 1.7322937877198545e-06,
+      "loss": 0.606,
+      "step": 926
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.5771508102246277,
+      "learning_rate": 1.7274575140626318e-06,
+      "loss": 0.44,
+      "step": 927
+    },
+    {
+      "epoch": 9.009708737864077,
+      "grad_norm": 0.6340958349337429,
+      "learning_rate": 1.7226244346267063e-06,
+      "loss": 0.5384,
+      "step": 928
+    },
+    {
+      "epoch": 9.019417475728156,
+      "grad_norm": 0.5464524313638762,
+      "learning_rate": 1.7177945693953486e-06,
+      "loss": 0.5522,
+      "step": 929
+    },
+    {
+      "epoch": 9.029126213592233,
+      "grad_norm": 0.509758461452797,
+      "learning_rate": 1.7129679383385384e-06,
+      "loss": 0.5649,
+      "step": 930
+    },
+    {
+      "epoch": 9.03883495145631,
+      "grad_norm": 0.6123362850624466,
+      "learning_rate": 1.7081445614128845e-06,
+      "loss": 0.4895,
+      "step": 931
+    },
+    {
+      "epoch": 9.048543689320388,
+      "grad_norm": 0.5077335227225634,
+      "learning_rate": 1.7033244585615393e-06,
+      "loss": 0.612,
+      "step": 932
+    },
+    {
+      "epoch": 9.058252427184467,
+      "grad_norm": 0.5392346989255187,
+      "learning_rate": 1.698507649714121e-06,
+      "loss": 0.7241,
+      "step": 933
+    },
+    {
+      "epoch": 9.067961165048544,
+      "grad_norm": 0.5141203869994445,
+      "learning_rate": 1.6936941547866248e-06,
+      "loss": 0.4577,
+      "step": 934
+    },
+    {
+      "epoch": 9.077669902912621,
+      "grad_norm": 0.5106376436372138,
+      "learning_rate": 1.688883993681345e-06,
+      "loss": 0.5075,
+      "step": 935
+    },
+    {
+      "epoch": 9.087378640776699,
+      "grad_norm": 0.5179058552152446,
+      "learning_rate": 1.6840771862867922e-06,
+      "loss": 0.5869,
+      "step": 936
+    },
+    {
+      "epoch": 9.097087378640778,
+      "grad_norm": 0.4977753472362273,
+      "learning_rate": 1.6792737524776093e-06,
+      "loss": 0.6939,
+      "step": 937
+    },
+    {
+      "epoch": 9.106796116504855,
+      "grad_norm": 0.5543634504769569,
+      "learning_rate": 1.674473712114492e-06,
+      "loss": 0.4248,
+      "step": 938
+    },
+    {
+      "epoch": 9.116504854368932,
+      "grad_norm": 0.5846797847400803,
+      "learning_rate": 1.6696770850441036e-06,
+      "loss": 0.458,
+      "step": 939
+    },
+    {
+      "epoch": 9.12621359223301,
+      "grad_norm": 0.5457809797121611,
+      "learning_rate": 1.6648838910989955e-06,
+      "loss": 0.7876,
+      "step": 940
+    },
+    {
+      "epoch": 9.135922330097088,
+      "grad_norm": 0.5144940860816298,
+      "learning_rate": 1.6600941500975237e-06,
+      "loss": 0.5822,
+      "step": 941
+    },
+    {
+      "epoch": 9.145631067961165,
+      "grad_norm": 0.5046446642153924,
+      "learning_rate": 1.6553078818437678e-06,
+      "loss": 0.6312,
+      "step": 942
+    },
+    {
+      "epoch": 9.155339805825243,
+      "grad_norm": 0.5401161645711199,
+      "learning_rate": 1.6505251061274492e-06,
+      "loss": 0.7804,
+      "step": 943
+    },
+    {
+      "epoch": 9.16504854368932,
+      "grad_norm": 0.6409458250404014,
+      "learning_rate": 1.6457458427238464e-06,
+      "loss": 0.6905,
+      "step": 944
+    },
+    {
+      "epoch": 9.174757281553399,
+      "grad_norm": 0.5273386933789973,
+      "learning_rate": 1.6409701113937182e-06,
+      "loss": 0.7029,
+      "step": 945
+    },
+    {
+      "epoch": 9.184466019417476,
+      "grad_norm": 0.5109661246472791,
+      "learning_rate": 1.6361979318832173e-06,
+      "loss": 0.4894,
+      "step": 946
+    },
+    {
+      "epoch": 9.194174757281553,
+      "grad_norm": 0.5511532623064545,
+      "learning_rate": 1.6314293239238134e-06,
+      "loss": 0.5167,
+      "step": 947
+    },
+    {
+      "epoch": 9.20388349514563,
+      "grad_norm": 0.6610489196315156,
+      "learning_rate": 1.626664307232207e-06,
+      "loss": 0.5094,
+      "step": 948
+    },
+    {
+      "epoch": 9.21359223300971,
+      "grad_norm": 0.5736305051432473,
+      "learning_rate": 1.62190290151025e-06,
+      "loss": 0.6754,
+      "step": 949
+    },
+    {
+      "epoch": 9.223300970873787,
+      "grad_norm": 0.5279615724891207,
+      "learning_rate": 1.617145126444864e-06,
+      "loss": 0.5783,
+      "step": 950
+    },
+    {
+      "epoch": 9.233009708737864,
+      "grad_norm": 0.5163456645251102,
+      "learning_rate": 1.6123910017079591e-06,
+      "loss": 0.6769,
+      "step": 951
+    },
+    {
+      "epoch": 9.242718446601941,
+      "grad_norm": 0.5833509127515675,
+      "learning_rate": 1.6076405469563533e-06,
+      "loss": 0.3916,
+      "step": 952
+    },
+    {
+      "epoch": 9.25242718446602,
+      "grad_norm": 0.4895578621974047,
+      "learning_rate": 1.6028937818316889e-06,
+      "loss": 0.6144,
+      "step": 953
+    },
+    {
+      "epoch": 9.262135922330097,
+      "grad_norm": 0.5425400198843934,
+      "learning_rate": 1.598150725960354e-06,
+      "loss": 0.2586,
+      "step": 954
+    },
+    {
+      "epoch": 9.271844660194175,
+      "grad_norm": 0.5501148487291081,
+      "learning_rate": 1.5934113989533992e-06,
+      "loss": 0.6158,
+      "step": 955
+    },
+    {
+      "epoch": 9.281553398058252,
+      "grad_norm": 0.5544881707377486,
+      "learning_rate": 1.5886758204064582e-06,
+      "loss": 0.5528,
+      "step": 956
+    },
+    {
+      "epoch": 9.29126213592233,
+      "grad_norm": 0.48248060557065603,
+      "learning_rate": 1.583944009899665e-06,
+      "loss": 0.2886,
+      "step": 957
+    },
+    {
+      "epoch": 9.300970873786408,
+      "grad_norm": 0.6941081918833938,
+      "learning_rate": 1.579215986997575e-06,
+      "loss": 0.4662,
+      "step": 958
+    },
+    {
+      "epoch": 9.310679611650485,
+      "grad_norm": 0.5465640944202089,
+      "learning_rate": 1.5744917712490821e-06,
+      "loss": 0.6941,
+      "step": 959
+    },
+    {
+      "epoch": 9.320388349514563,
+      "grad_norm": 0.5319670471627715,
+      "learning_rate": 1.5697713821873401e-06,
+      "loss": 0.4341,
+      "step": 960
+    },
+    {
+      "epoch": 9.330097087378642,
+      "grad_norm": 0.48885465810429257,
+      "learning_rate": 1.5650548393296788e-06,
+      "loss": 0.4523,
+      "step": 961
+    },
+    {
+      "epoch": 9.339805825242719,
+      "grad_norm": 0.7022004361946983,
+      "learning_rate": 1.5603421621775273e-06,
+      "loss": 0.629,
+      "step": 962
+    },
+    {
+      "epoch": 9.349514563106796,
+      "grad_norm": 0.5632088911699985,
+      "learning_rate": 1.555633370216329e-06,
+      "loss": 0.6666,
+      "step": 963
+    },
+    {
+      "epoch": 9.359223300970873,
+      "grad_norm": 0.49380965838592933,
+      "learning_rate": 1.5509284829154652e-06,
+      "loss": 0.4755,
+      "step": 964
+    },
+    {
+      "epoch": 9.368932038834952,
+      "grad_norm": 0.5020708429363245,
+      "learning_rate": 1.5462275197281717e-06,
+      "loss": 0.5441,
+      "step": 965
+    },
+    {
+      "epoch": 9.37864077669903,
+      "grad_norm": 0.5195307949196531,
+      "learning_rate": 1.5415305000914587e-06,
+      "loss": 0.7319,
+      "step": 966
+    },
+    {
+      "epoch": 9.388349514563107,
+      "grad_norm": 0.48395870668183194,
+      "learning_rate": 1.536837443426032e-06,
+      "loss": 0.3737,
+      "step": 967
+    },
+    {
+      "epoch": 9.398058252427184,
+      "grad_norm": 0.6288833157472449,
+      "learning_rate": 1.5321483691362121e-06,
+      "loss": 0.3642,
+      "step": 968
+    },
+    {
+      "epoch": 9.407766990291263,
+      "grad_norm": 0.7231122831566812,
+      "learning_rate": 1.5274632966098527e-06,
+      "loss": 0.4566,
+      "step": 969
+    },
+    {
+      "epoch": 9.41747572815534,
+      "grad_norm": 0.5043165423307169,
+      "learning_rate": 1.5227822452182617e-06,
+      "loss": 0.4137,
+      "step": 970
+    },
+    {
+      "epoch": 9.427184466019417,
+      "grad_norm": 0.7582603233246429,
+      "learning_rate": 1.5181052343161212e-06,
+      "loss": 0.4781,
+      "step": 971
+    },
+    {
+      "epoch": 9.436893203883495,
+      "grad_norm": 0.5366974044741509,
+      "learning_rate": 1.5134322832414066e-06,
+      "loss": 0.4163,
+      "step": 972
+    },
+    {
+      "epoch": 9.446601941747574,
+      "grad_norm": 0.5799499713204818,
+      "learning_rate": 1.508763411315308e-06,
+      "loss": 0.5431,
+      "step": 973
+    },
+    {
+      "epoch": 9.45631067961165,
+      "grad_norm": 0.5604178353146644,
+      "learning_rate": 1.5040986378421485e-06,
+      "loss": 0.4337,
+      "step": 974
+    },
+    {
+      "epoch": 9.466019417475728,
+      "grad_norm": 0.4513197289000593,
+      "learning_rate": 1.499437982109305e-06,
+      "loss": 0.4195,
+      "step": 975
+    },
+    {
+      "epoch": 9.475728155339805,
+      "grad_norm": 0.6109174978216692,
+      "learning_rate": 1.4947814633871316e-06,
+      "loss": 0.6013,
+      "step": 976
+    },
+    {
+      "epoch": 9.485436893203884,
+      "grad_norm": 0.4822809855329016,
+      "learning_rate": 1.4901291009288748e-06,
+      "loss": 0.3593,
+      "step": 977
+    },
+    {
+      "epoch": 9.495145631067961,
+      "grad_norm": 0.5640776707095488,
+      "learning_rate": 1.4854809139705961e-06,
+      "loss": 0.5642,
+      "step": 978
+    },
+    {
+      "epoch": 9.504854368932039,
+      "grad_norm": 0.5595919599675019,
+      "learning_rate": 1.4808369217310937e-06,
+      "loss": 0.5531,
+      "step": 979
+    },
+    {
+      "epoch": 9.514563106796116,
+      "grad_norm": 0.5302464691788346,
+      "learning_rate": 1.4761971434118207e-06,
+      "loss": 0.6372,
+      "step": 980
+    },
+    {
+      "epoch": 9.524271844660195,
+      "grad_norm": 0.5300165066761687,
+      "learning_rate": 1.4715615981968088e-06,
+      "loss": 0.5707,
+      "step": 981
+    },
+    {
+      "epoch": 9.533980582524272,
+      "grad_norm": 0.4668375811587118,
+      "learning_rate": 1.4669303052525852e-06,
+      "loss": 0.4989,
+      "step": 982
+    },
+    {
+      "epoch": 9.54368932038835,
+      "grad_norm": 0.5560569559935798,
+      "learning_rate": 1.4623032837280971e-06,
+      "loss": 0.3493,
+      "step": 983
+    },
+    {
+      "epoch": 9.553398058252426,
+      "grad_norm": 0.5577549092239438,
+      "learning_rate": 1.4576805527546293e-06,
+      "loss": 0.8323,
+      "step": 984
+    },
+    {
+      "epoch": 9.563106796116505,
+      "grad_norm": 0.5677422426756373,
+      "learning_rate": 1.4530621314457255e-06,
+      "loss": 0.6881,
+      "step": 985
+    },
+    {
+      "epoch": 9.572815533980583,
+      "grad_norm": 0.5279934542850471,
+      "learning_rate": 1.4484480388971141e-06,
+      "loss": 0.7699,
+      "step": 986
+    },
+    {
+      "epoch": 9.58252427184466,
+      "grad_norm": 0.5350476163708038,
+      "learning_rate": 1.4438382941866224e-06,
+      "loss": 0.7338,
+      "step": 987
+    },
+    {
+      "epoch": 9.592233009708737,
+      "grad_norm": 0.5110629848586472,
+      "learning_rate": 1.4392329163741015e-06,
+      "loss": 0.5176,
+      "step": 988
+    },
+    {
+      "epoch": 9.601941747572816,
+      "grad_norm": 0.5687212051187402,
+      "learning_rate": 1.4346319245013463e-06,
+      "loss": 0.5527,
+      "step": 989
+    },
+    {
+      "epoch": 9.611650485436893,
+      "grad_norm": 0.4974392454544525,
+      "learning_rate": 1.430035337592018e-06,
+      "loss": 0.4942,
+      "step": 990
+    },
+    {
+      "epoch": 9.62135922330097,
+      "grad_norm": 0.5629010223404443,
+      "learning_rate": 1.425443174651564e-06,
+      "loss": 0.4756,
+      "step": 991
+    },
+    {
+      "epoch": 9.631067961165048,
+      "grad_norm": 0.5524080467073957,
+      "learning_rate": 1.4208554546671407e-06,
+      "loss": 0.6035,
+      "step": 992
+    },
+    {
+      "epoch": 9.640776699029127,
+      "grad_norm": 0.4804538845483175,
+      "learning_rate": 1.4162721966075323e-06,
+      "loss": 0.7249,
+      "step": 993
+    },
+    {
+      "epoch": 9.650485436893204,
+      "grad_norm": 0.4993730994583247,
+      "learning_rate": 1.411693419423078e-06,
+      "loss": 0.3848,
+      "step": 994
+    },
+    {
+      "epoch": 9.660194174757281,
+      "grad_norm": 0.5354151376861092,
+      "learning_rate": 1.4071191420455873e-06,
+      "loss": 0.5479,
+      "step": 995
+    },
+    {
+      "epoch": 9.669902912621358,
+      "grad_norm": 0.5224179916255741,
+      "learning_rate": 1.4025493833882645e-06,
+      "loss": 0.428,
+      "step": 996
+    },
+    {
+      "epoch": 9.679611650485437,
+      "grad_norm": 0.5441066168724764,
+      "learning_rate": 1.3979841623456309e-06,
+      "loss": 0.485,
+      "step": 997
+    },
+    {
+      "epoch": 9.689320388349515,
+      "grad_norm": 0.5162503314510989,
+      "learning_rate": 1.3934234977934463e-06,
+      "loss": 0.6244,
+      "step": 998
+    },
+    {
+      "epoch": 9.699029126213592,
+      "grad_norm": 0.5088420073204567,
+      "learning_rate": 1.3888674085886302e-06,
+      "loss": 0.5502,
+      "step": 999
+    },
+    {
+      "epoch": 9.70873786407767,
+      "grad_norm": 0.4891658312631955,
+      "learning_rate": 1.3843159135691859e-06,
+      "loss": 0.4974,
+      "step": 1000
+    },
+    {
+      "epoch": 9.718446601941748,
+      "grad_norm": 0.5247239478758943,
+      "learning_rate": 1.3797690315541193e-06,
+      "loss": 0.5293,
+      "step": 1001
+    },
+    {
+      "epoch": 9.728155339805825,
+      "grad_norm": 0.5579036546703239,
+      "learning_rate": 1.3752267813433645e-06,
+      "loss": 0.5561,
+      "step": 1002
+    },
+    {
+      "epoch": 9.737864077669903,
+      "grad_norm": 0.5128549281129826,
+      "learning_rate": 1.3706891817177036e-06,
+      "loss": 0.392,
+      "step": 1003
+    },
+    {
+      "epoch": 9.74757281553398,
+      "grad_norm": 0.5506859233210426,
+      "learning_rate": 1.3661562514386895e-06,
+      "loss": 0.5688,
+      "step": 1004
+    },
+    {
+      "epoch": 9.757281553398059,
+      "grad_norm": 0.5704140162541189,
+      "learning_rate": 1.3616280092485719e-06,
+      "loss": 0.6288,
+      "step": 1005
+    },
+    {
+      "epoch": 9.766990291262136,
+      "grad_norm": 0.5764246997137785,
+      "learning_rate": 1.357104473870213e-06,
+      "loss": 0.655,
+      "step": 1006
+    },
+    {
+      "epoch": 9.776699029126213,
+      "grad_norm": 0.5123137189616399,
+      "learning_rate": 1.3525856640070156e-06,
+      "loss": 0.4107,
+      "step": 1007
+    },
+    {
+      "epoch": 9.78640776699029,
+      "grad_norm": 0.5456395663842477,
+      "learning_rate": 1.3480715983428433e-06,
+      "loss": 0.5157,
+      "step": 1008
+    },
+    {
+      "epoch": 9.79611650485437,
+      "grad_norm": 0.6018889487173135,
+      "learning_rate": 1.3435622955419447e-06,
+      "loss": 0.5223,
+      "step": 1009
+    },
+    {
+      "epoch": 9.805825242718447,
+      "grad_norm": 0.5288879566344424,
+      "learning_rate": 1.3390577742488747e-06,
+      "loss": 0.402,
+      "step": 1010
+    },
+    {
+      "epoch": 9.815533980582524,
+      "grad_norm": 0.5060531265013377,
+      "learning_rate": 1.334558053088419e-06,
+      "loss": 0.5625,
+      "step": 1011
+    },
+    {
+      "epoch": 9.825242718446601,
+      "grad_norm": 0.5983276437902583,
+      "learning_rate": 1.3300631506655148e-06,
+      "loss": 0.7724,
+      "step": 1012
+    },
+    {
+      "epoch": 9.83495145631068,
+      "grad_norm": 0.5421716359030226,
+      "learning_rate": 1.3255730855651772e-06,
+      "loss": 0.6375,
+      "step": 1013
+    },
+    {
+      "epoch": 9.844660194174757,
+      "grad_norm": 0.5137714833067758,
+      "learning_rate": 1.3210878763524186e-06,
+      "loss": 0.565,
+      "step": 1014
+    },
+    {
+      "epoch": 9.854368932038835,
+      "grad_norm": 0.5224752768506955,
+      "learning_rate": 1.3166075415721762e-06,
+      "loss": 0.7022,
+      "step": 1015
+    },
+    {
+      "epoch": 9.864077669902912,
+      "grad_norm": 0.5131878611886128,
+      "learning_rate": 1.3121320997492305e-06,
+      "loss": 0.5082,
+      "step": 1016
+    },
+    {
+      "epoch": 9.87378640776699,
+      "grad_norm": 0.5203915202506901,
+      "learning_rate": 1.307661569388132e-06,
+      "loss": 0.538,
+      "step": 1017
+    },
+    {
+      "epoch": 9.883495145631068,
+      "grad_norm": 0.5259467137163731,
+      "learning_rate": 1.3031959689731236e-06,
+      "loss": 0.4584,
+      "step": 1018
+    },
+    {
+      "epoch": 9.893203883495145,
+      "grad_norm": 0.5245481953235169,
+      "learning_rate": 1.2987353169680667e-06,
+      "loss": 0.6089,
+      "step": 1019
+    },
+    {
+      "epoch": 9.902912621359224,
+      "grad_norm": 0.5303857378086793,
+      "learning_rate": 1.2942796318163595e-06,
+      "loss": 0.5585,
+      "step": 1020
+    },
+    {
+      "epoch": 9.912621359223301,
+      "grad_norm": 0.5041008237960473,
+      "learning_rate": 1.2898289319408653e-06,
+      "loss": 0.5354,
+      "step": 1021
+    },
+    {
+      "epoch": 9.922330097087379,
+      "grad_norm": 0.5552017542716956,
+      "learning_rate": 1.2853832357438346e-06,
+      "loss": 0.6052,
+      "step": 1022
+    },
+    {
+      "epoch": 9.932038834951456,
+      "grad_norm": 0.47893428879532324,
+      "learning_rate": 1.2809425616068288e-06,
+      "loss": 0.5291,
+      "step": 1023
+    },
+    {
+      "epoch": 9.941747572815533,
+      "grad_norm": 0.6660999509501955,
+      "learning_rate": 1.2765069278906456e-06,
+      "loss": 0.6806,
+      "step": 1024
+    },
+    {
+      "epoch": 9.951456310679612,
+      "grad_norm": 0.6474829457761733,
+      "learning_rate": 1.2720763529352415e-06,
+      "loss": 0.5242,
+      "step": 1025
+    },
+    {
+      "epoch": 9.96116504854369,
+      "grad_norm": 0.5121227651736198,
+      "learning_rate": 1.2676508550596562e-06,
+      "loss": 0.4134,
+      "step": 1026
+    },
+    {
+      "epoch": 9.970873786407767,
+      "grad_norm": 0.5377519390532532,
+      "learning_rate": 1.2632304525619388e-06,
+      "loss": 0.429,
+      "step": 1027
+    },
+    {
+      "epoch": 9.980582524271846,
+      "grad_norm": 0.5861629446040602,
+      "learning_rate": 1.2588151637190687e-06,
+      "loss": 0.8707,
+      "step": 1028
+    },
+    {
+      "epoch": 9.990291262135923,
+      "grad_norm": 0.5691854746366819,
+      "learning_rate": 1.2544050067868834e-06,
+      "loss": 0.4264,
+      "step": 1029
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.7478699669714269,
+      "learning_rate": 1.2500000000000007e-06,
+      "loss": 0.5084,
+      "step": 1030
+    },
+    {
+      "epoch": 10.009708737864077,
+      "grad_norm": 0.46151381211002906,
+      "learning_rate": 1.2456001615717445e-06,
+      "loss": 0.5594,
+      "step": 1031
+    },
+    {
+      "epoch": 10.019417475728156,
+      "grad_norm": 0.5102877537048928,
+      "learning_rate": 1.2412055096940692e-06,
+      "loss": 0.4949,
+      "step": 1032
+    },
+    {
+      "epoch": 10.029126213592233,
+      "grad_norm": 0.5822245866471158,
+      "learning_rate": 1.2368160625374835e-06,
+      "loss": 0.797,
+      "step": 1033
+    },
+    {
+      "epoch": 10.03883495145631,
+      "grad_norm": 0.4956969860994467,
+      "learning_rate": 1.2324318382509787e-06,
+      "loss": 0.3893,
+      "step": 1034
+    },
+    {
+      "epoch": 10.048543689320388,
+      "grad_norm": 0.5208191592763981,
+      "learning_rate": 1.2280528549619487e-06,
+      "loss": 0.3821,
+      "step": 1035
+    },
+    {
+      "epoch": 10.058252427184467,
+      "grad_norm": 0.5448550745622244,
+      "learning_rate": 1.2236791307761184e-06,
+      "loss": 0.6886,
+      "step": 1036
+    },
+    {
+      "epoch": 10.067961165048544,
+      "grad_norm": 0.6039482131792457,
+      "learning_rate": 1.2193106837774678e-06,
+      "loss": 0.5576,
+      "step": 1037
+    },
+    {
+      "epoch": 10.077669902912621,
+      "grad_norm": 0.5479538068077612,
+      "learning_rate": 1.2149475320281578e-06,
+      "loss": 0.4806,
+      "step": 1038
+    },
+    {
+      "epoch": 10.087378640776699,
+      "grad_norm": 0.5705314303946492,
+      "learning_rate": 1.2105896935684545e-06,
+      "loss": 0.5369,
+      "step": 1039
+    },
+    {
+      "epoch": 10.097087378640778,
+      "grad_norm": 0.5271533020970816,
+      "learning_rate": 1.2062371864166553e-06,
+      "loss": 0.4633,
+      "step": 1040
+    },
+    {
+      "epoch": 10.106796116504855,
+      "grad_norm": 0.5456418396000027,
+      "learning_rate": 1.2018900285690148e-06,
+      "loss": 0.5174,
+      "step": 1041
+    },
+    {
+      "epoch": 10.116504854368932,
+      "grad_norm": 0.5368574554830884,
+      "learning_rate": 1.1975482379996697e-06,
+      "loss": 0.5409,
+      "step": 1042
+    },
+    {
+      "epoch": 10.12621359223301,
+      "grad_norm": 0.5035113911073187,
+      "learning_rate": 1.1932118326605644e-06,
+      "loss": 0.4153,
+      "step": 1043
+    },
+    {
+      "epoch": 10.135922330097088,
+      "grad_norm": 0.5630447002734048,
+      "learning_rate": 1.188880830481377e-06,
+      "loss": 0.3517,
+      "step": 1044
+    },
+    {
+      "epoch": 10.145631067961165,
+      "grad_norm": 0.46046579136925914,
+      "learning_rate": 1.1845552493694462e-06,
+      "loss": 0.3997,
+      "step": 1045
+    },
+    {
+      "epoch": 10.155339805825243,
+      "grad_norm": 0.5566449442651117,
+      "learning_rate": 1.1802351072096948e-06,
+      "loss": 0.473,
+      "step": 1046
+    },
+    {
+      "epoch": 10.16504854368932,
+      "grad_norm": 0.5372526771927345,
+      "learning_rate": 1.1759204218645577e-06,
+      "loss": 0.4501,
+      "step": 1047
+    },
+    {
+      "epoch": 10.174757281553399,
+      "grad_norm": 0.6565634479247288,
+      "learning_rate": 1.1716112111739095e-06,
+      "loss": 0.4763,
+      "step": 1048
+    },
+    {
+      "epoch": 10.184466019417476,
+      "grad_norm": 0.5550268695193769,
+      "learning_rate": 1.167307492954986e-06,
+      "loss": 0.695,
+      "step": 1049
+    },
+    {
+      "epoch": 10.194174757281553,
+      "grad_norm": 0.5101472665103614,
+      "learning_rate": 1.1630092850023148e-06,
+      "loss": 0.6151,
+      "step": 1050
+    },
+    {
+      "epoch": 10.20388349514563,
+      "grad_norm": 0.5663960308823205,
+      "learning_rate": 1.15871660508764e-06,
+      "loss": 0.7663,
+      "step": 1051
+    },
+    {
+      "epoch": 10.21359223300971,
+      "grad_norm": 0.5189202841990675,
+      "learning_rate": 1.1544294709598491e-06,
+      "loss": 0.7497,
+      "step": 1052
+    },
+    {
+      "epoch": 10.223300970873787,
+      "grad_norm": 0.5286405812389172,
+      "learning_rate": 1.1501479003448992e-06,
+      "loss": 0.6158,
+      "step": 1053
+    },
+    {
+      "epoch": 10.233009708737864,
+      "grad_norm": 0.5004315366131019,
+      "learning_rate": 1.1458719109457445e-06,
+      "loss": 0.4122,
+      "step": 1054
+    },
+    {
+      "epoch": 10.242718446601941,
+      "grad_norm": 0.5071667711625752,
+      "learning_rate": 1.141601520442262e-06,
+      "loss": 0.6746,
+      "step": 1055
+    },
+    {
+      "epoch": 10.25242718446602,
+      "grad_norm": 0.5057112811278313,
+      "learning_rate": 1.1373367464911798e-06,
+      "loss": 0.3907,
+      "step": 1056
+    },
+    {
+      "epoch": 10.262135922330097,
+      "grad_norm": 0.5310657265619477,
+      "learning_rate": 1.1330776067260026e-06,
+      "loss": 0.6539,
+      "step": 1057
+    },
+    {
+      "epoch": 10.271844660194175,
+      "grad_norm": 0.5114027839132974,
+      "learning_rate": 1.12882411875694e-06,
+      "loss": 0.4251,
+      "step": 1058
+    },
+    {
+      "epoch": 10.281553398058252,
+      "grad_norm": 0.5828286358595048,
+      "learning_rate": 1.1245763001708326e-06,
+      "loss": 0.4021,
+      "step": 1059
+    },
+    {
+      "epoch": 10.29126213592233,
+      "grad_norm": 0.4828687846960337,
+      "learning_rate": 1.120334168531081e-06,
+      "loss": 0.5349,
+      "step": 1060
+    },
+    {
+      "epoch": 10.300970873786408,
+      "grad_norm": 0.5241573566112172,
+      "learning_rate": 1.1160977413775704e-06,
+      "loss": 0.6077,
+      "step": 1061
+    },
+    {
+      "epoch": 10.310679611650485,
+      "grad_norm": 0.5066991756324611,
+      "learning_rate": 1.1118670362266003e-06,
+      "loss": 0.4968,
+      "step": 1062
+    },
+    {
+      "epoch": 10.320388349514563,
+      "grad_norm": 0.5118698345971918,
+      "learning_rate": 1.1076420705708137e-06,
+      "loss": 0.4295,
+      "step": 1063
+    },
+    {
+      "epoch": 10.330097087378642,
+      "grad_norm": 0.5509488966267911,
+      "learning_rate": 1.1034228618791197e-06,
+      "loss": 0.3386,
+      "step": 1064
+    },
+    {
+      "epoch": 10.339805825242719,
+      "grad_norm": 0.6303421970689016,
+      "learning_rate": 1.0992094275966256e-06,
+      "loss": 0.3737,
+      "step": 1065
+    },
+    {
+      "epoch": 10.349514563106796,
+      "grad_norm": 0.5369953531447155,
+      "learning_rate": 1.0950017851445624e-06,
+      "loss": 0.6595,
+      "step": 1066
+    },
+    {
+      "epoch": 10.359223300970873,
+      "grad_norm": 0.6048647515454525,
+      "learning_rate": 1.0907999519202142e-06,
+      "loss": 0.5782,
+      "step": 1067
+    },
+    {
+      "epoch": 10.368932038834952,
+      "grad_norm": 0.5011839711044318,
+      "learning_rate": 1.0866039452968464e-06,
+      "loss": 0.4836,
+      "step": 1068
+    },
+    {
+      "epoch": 10.37864077669903,
+      "grad_norm": 0.549369060527332,
+      "learning_rate": 1.0824137826236318e-06,
+      "loss": 0.3571,
+      "step": 1069
+    },
+    {
+      "epoch": 10.388349514563107,
+      "grad_norm": 0.5046590280856795,
+      "learning_rate": 1.078229481225582e-06,
+      "loss": 0.381,
+      "step": 1070
+    },
+    {
+      "epoch": 10.398058252427184,
+      "grad_norm": 0.49366158846755176,
+      "learning_rate": 1.074051058403472e-06,
+      "loss": 0.414,
+      "step": 1071
+    },
+    {
+      "epoch": 10.407766990291263,
+      "grad_norm": 0.566545758141253,
+      "learning_rate": 1.069878531433773e-06,
+      "loss": 0.4003,
+      "step": 1072
+    },
+    {
+      "epoch": 10.41747572815534,
+      "grad_norm": 0.5472254994058976,
+      "learning_rate": 1.0657119175685776e-06,
+      "loss": 0.5429,
+      "step": 1073
+    },
+    {
+      "epoch": 10.427184466019417,
+      "grad_norm": 0.5887149186721222,
+      "learning_rate": 1.061551234035529e-06,
+      "loss": 0.3878,
+      "step": 1074
+    },
+    {
+      "epoch": 10.436893203883495,
+      "grad_norm": 0.5282433190533067,
+      "learning_rate": 1.0573964980377517e-06,
+      "loss": 0.6502,
+      "step": 1075
+    },
+    {
+      "epoch": 10.446601941747574,
+      "grad_norm": 0.44374712602971006,
+      "learning_rate": 1.0532477267537772e-06,
+      "loss": 0.511,
+      "step": 1076
+    },
+    {
+      "epoch": 10.45631067961165,
+      "grad_norm": 0.5490486215222025,
+      "learning_rate": 1.0491049373374762e-06,
+      "loss": 0.6757,
+      "step": 1077
+    },
+    {
+      "epoch": 10.466019417475728,
+      "grad_norm": 0.49442276500022064,
+      "learning_rate": 1.044968146917986e-06,
+      "loss": 0.4764,
+      "step": 1078
+    },
+    {
+      "epoch": 10.475728155339805,
+      "grad_norm": 0.5608038269974015,
+      "learning_rate": 1.0408373725996386e-06,
+      "loss": 0.5028,
+      "step": 1079
+    },
+    {
+      "epoch": 10.485436893203884,
+      "grad_norm": 0.5687457204009948,
+      "learning_rate": 1.0367126314618946e-06,
+      "loss": 0.5785,
+      "step": 1080
+    },
+    {
+      "epoch": 10.495145631067961,
+      "grad_norm": 0.6968670389266779,
+      "learning_rate": 1.0325939405592661e-06,
+      "loss": 0.6075,
+      "step": 1081
+    },
+    {
+      "epoch": 10.504854368932039,
+      "grad_norm": 0.5304379314551411,
+      "learning_rate": 1.0284813169212502e-06,
+      "loss": 0.3914,
+      "step": 1082
+    },
+    {
+      "epoch": 10.514563106796116,
+      "grad_norm": 0.5419260179618722,
+      "learning_rate": 1.024374777552258e-06,
+      "loss": 0.5441,
+      "step": 1083
+    },
+    {
+      "epoch": 10.524271844660195,
+      "grad_norm": 1.194505571995219,
+      "learning_rate": 1.0202743394315444e-06,
+      "loss": 0.5646,
+      "step": 1084
+    },
+    {
+      "epoch": 10.533980582524272,
+      "grad_norm": 0.5014060531536156,
+      "learning_rate": 1.0161800195131372e-06,
+      "loss": 0.5548,
+      "step": 1085
+    },
+    {
+      "epoch": 10.54368932038835,
+      "grad_norm": 0.49937660543182716,
+      "learning_rate": 1.0120918347257669e-06,
+      "loss": 0.5122,
+      "step": 1086
+    },
+    {
+      "epoch": 10.553398058252426,
+      "grad_norm": 0.5462501234068766,
+      "learning_rate": 1.0080098019727979e-06,
+      "loss": 0.452,
+      "step": 1087
+    },
+    {
+      "epoch": 10.563106796116505,
+      "grad_norm": 0.5631895834373033,
+      "learning_rate": 1.0039339381321572e-06,
+      "loss": 0.7399,
+      "step": 1088
+    },
+    {
+      "epoch": 10.572815533980583,
+      "grad_norm": 0.4736886401897993,
+      "learning_rate": 9.998642600562664e-07,
+      "loss": 0.6179,
+      "step": 1089
+    },
+    {
+      "epoch": 10.58252427184466,
+      "grad_norm": 0.46820689865613146,
+      "learning_rate": 9.95800784571969e-07,
+      "loss": 0.6221,
+      "step": 1090
+    },
+    {
+      "epoch": 10.592233009708737,
+      "grad_norm": 0.5195369073586825,
+      "learning_rate": 9.91743528480464e-07,
+      "loss": 0.5183,
+      "step": 1091
+    },
+    {
+      "epoch": 10.601941747572816,
+      "grad_norm": 0.5432189978052531,
+      "learning_rate": 9.876925085572365e-07,
+      "loss": 0.5503,
+      "step": 1092
+    },
+    {
+      "epoch": 10.611650485436893,
+      "grad_norm": 0.5493319095400769,
+      "learning_rate": 9.836477415519843e-07,
+      "loss": 0.5628,
+      "step": 1093
+    },
+    {
+      "epoch": 10.62135922330097,
+      "grad_norm": 0.5150871722296954,
+      "learning_rate": 9.79609244188553e-07,
+      "loss": 0.5485,
+      "step": 1094
+    },
+    {
+      "epoch": 10.631067961165048,
+      "grad_norm": 0.5523462804112262,
+      "learning_rate": 9.755770331648642e-07,
+      "loss": 0.4576,
+      "step": 1095
+    },
+    {
+      "epoch": 10.640776699029127,
+      "grad_norm": 0.5157125198942374,
+      "learning_rate": 9.715511251528486e-07,
+      "loss": 0.5491,
+      "step": 1096
+    },
+    {
+      "epoch": 10.650485436893204,
+      "grad_norm": 0.5227068552137706,
+      "learning_rate": 9.67531536798375e-07,
+      "loss": 0.6811,
+      "step": 1097
+    },
+    {
+      "epoch": 10.660194174757281,
+      "grad_norm": 0.5192500487387761,
+      "learning_rate": 9.635182847211827e-07,
+      "loss": 0.6348,
+      "step": 1098
+    },
+    {
+      "epoch": 10.669902912621358,
+      "grad_norm": 0.5401758469516798,
+      "learning_rate": 9.595113855148128e-07,
+      "loss": 0.5902,
+      "step": 1099
+    },
+    {
+      "epoch": 10.679611650485437,
+      "grad_norm": 0.9322574427952275,
+      "learning_rate": 9.555108557465383e-07,
+      "loss": 0.7462,
+      "step": 1100
+    },
+    {
+      "epoch": 10.689320388349515,
+      "grad_norm": 0.5179604641829436,
+      "learning_rate": 9.51516711957298e-07,
+      "loss": 0.5137,
+      "step": 1101
+    },
+    {
+      "epoch": 10.699029126213592,
+      "grad_norm": 0.5210907828158898,
+      "learning_rate": 9.475289706616256e-07,
+      "loss": 0.5653,
+      "step": 1102
+    },
+    {
+      "epoch": 10.70873786407767,
+      "grad_norm": 0.5610819470304531,
+      "learning_rate": 9.435476483475825e-07,
+      "loss": 0.4509,
+      "step": 1103
+    },
+    {
+      "epoch": 10.718446601941748,
+      "grad_norm": 0.5239244795742548,
+      "learning_rate": 9.395727614766903e-07,
+      "loss": 0.419,
+      "step": 1104
+    },
+    {
+      "epoch": 10.728155339805825,
+      "grad_norm": 0.5043201994579133,
+      "learning_rate": 9.356043264838607e-07,
+      "loss": 0.5223,
+      "step": 1105
+    },
+    {
+      "epoch": 10.737864077669903,
+      "grad_norm": 0.6710199227966344,
+      "learning_rate": 9.316423597773316e-07,
+      "loss": 0.5586,
+      "step": 1106
+    },
+    {
+      "epoch": 10.74757281553398,
+      "grad_norm": 0.5060889411301873,
+      "learning_rate": 9.276868777385942e-07,
+      "loss": 0.5578,
+      "step": 1107
+    },
+    {
+      "epoch": 10.757281553398059,
+      "grad_norm": 0.5180089569639259,
+      "learning_rate": 9.237378967223279e-07,
+      "loss": 0.3907,
+      "step": 1108
+    },
+    {
+      "epoch": 10.766990291262136,
+      "grad_norm": 0.6040964912815993,
+      "learning_rate": 9.197954330563327e-07,
+      "loss": 0.5639,
+      "step": 1109
+    },
+    {
+      "epoch": 10.776699029126213,
+      "grad_norm": 0.5816718750599607,
+      "learning_rate": 9.158595030414621e-07,
+      "loss": 0.5752,
+      "step": 1110
+    },
+    {
+      "epoch": 10.78640776699029,
+      "grad_norm": 0.5417325173104717,
+      "learning_rate": 9.11930122951554e-07,
+      "loss": 0.4933,
+      "step": 1111
+    },
+    {
+      "epoch": 10.79611650485437,
+      "grad_norm": 0.600237968659638,
+      "learning_rate": 9.080073090333646e-07,
+      "loss": 0.4976,
+      "step": 1112
+    },
+    {
+      "epoch": 10.805825242718447,
+      "grad_norm": 0.49463419377226303,
+      "learning_rate": 9.040910775065015e-07,
+      "loss": 0.5993,
+      "step": 1113
+    },
+    {
+      "epoch": 10.815533980582524,
+      "grad_norm": 0.510920823880218,
+      "learning_rate": 9.001814445633558e-07,
+      "loss": 0.4481,
+      "step": 1114
+    },
+    {
+      "epoch": 10.825242718446601,
+      "grad_norm": 0.5307522441544542,
+      "learning_rate": 8.962784263690358e-07,
+      "loss": 0.6389,
+      "step": 1115
+    },
+    {
+      "epoch": 10.83495145631068,
+      "grad_norm": 0.5001773473983797,
+      "learning_rate": 8.923820390612991e-07,
+      "loss": 0.431,
+      "step": 1116
+    },
+    {
+      "epoch": 10.844660194174757,
+      "grad_norm": 0.5452911240935504,
+      "learning_rate": 8.884922987504882e-07,
+      "loss": 0.5698,
+      "step": 1117
+    },
+    {
+      "epoch": 10.854368932038835,
+      "grad_norm": 0.5875035644254444,
+      "learning_rate": 8.846092215194607e-07,
+      "loss": 0.5176,
+      "step": 1118
+    },
+    {
+      "epoch": 10.864077669902912,
+      "grad_norm": 0.593676774427124,
+      "learning_rate": 8.807328234235254e-07,
+      "loss": 0.4337,
+      "step": 1119
+    },
+    {
+      "epoch": 10.87378640776699,
+      "grad_norm": 0.5292460765839232,
+      "learning_rate": 8.768631204903738e-07,
+      "loss": 0.6187,
+      "step": 1120
+    },
+    {
+      "epoch": 10.883495145631068,
+      "grad_norm": 0.49638389592552545,
+      "learning_rate": 8.730001287200177e-07,
+      "loss": 0.7939,
+      "step": 1121
+    },
+    {
+      "epoch": 10.893203883495145,
+      "grad_norm": 0.49414826625658015,
+      "learning_rate": 8.691438640847177e-07,
+      "loss": 0.4413,
+      "step": 1122
+    },
+    {
+      "epoch": 10.902912621359224,
+      "grad_norm": 0.5295200176718633,
+      "learning_rate": 8.652943425289206e-07,
+      "loss": 0.7966,
+      "step": 1123
+    },
+    {
+      "epoch": 10.912621359223301,
+      "grad_norm": 0.5870264475278172,
+      "learning_rate": 8.61451579969193e-07,
+      "loss": 0.4587,
+      "step": 1124
+    },
+    {
+      "epoch": 10.922330097087379,
+      "grad_norm": 0.5685870573511718,
+      "learning_rate": 8.576155922941548e-07,
+      "loss": 0.6506,
+      "step": 1125
+    },
+    {
+      "epoch": 10.932038834951456,
+      "grad_norm": 0.44300049758866367,
+      "learning_rate": 8.537863953644138e-07,
+      "loss": 0.4497,
+      "step": 1126
+    },
+    {
+      "epoch": 10.941747572815533,
+      "grad_norm": 0.5555503272916815,
+      "learning_rate": 8.499640050125007e-07,
+      "loss": 0.6713,
+      "step": 1127
+    },
+    {
+      "epoch": 10.951456310679612,
+      "grad_norm": 0.5182699503371718,
+      "learning_rate": 8.461484370428025e-07,
+      "loss": 0.55,
+      "step": 1128
+    },
+    {
+      "epoch": 10.96116504854369,
+      "grad_norm": 0.5219385783314696,
+      "learning_rate": 8.423397072314985e-07,
+      "loss": 0.6125,
+      "step": 1129
+    },
+    {
+      "epoch": 10.970873786407767,
+      "grad_norm": 0.5394176739754208,
+      "learning_rate": 8.385378313264933e-07,
+      "loss": 0.5786,
+      "step": 1130
+    },
+    {
+      "epoch": 10.980582524271846,
+      "grad_norm": 0.5661425078708068,
+      "learning_rate": 8.347428250473541e-07,
+      "loss": 0.6141,
+      "step": 1131
+    },
+    {
+      "epoch": 10.990291262135923,
+      "grad_norm": 0.7696322655449727,
+      "learning_rate": 8.309547040852434e-07,
+      "loss": 0.4752,
+      "step": 1132
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.47444374346121526,
+      "learning_rate": 8.271734841028553e-07,
+      "loss": 0.6056,
+      "step": 1133
+    },
+    {
+      "epoch": 11.009708737864077,
+      "grad_norm": 0.5035082193387115,
+      "learning_rate": 8.233991807343497e-07,
+      "loss": 0.6257,
+      "step": 1134
+    },
+    {
+      "epoch": 11.019417475728156,
+      "grad_norm": 0.5349228147551298,
+      "learning_rate": 8.196318095852909e-07,
+      "loss": 0.3682,
+      "step": 1135
+    },
+    {
+      "epoch": 11.029126213592233,
+      "grad_norm": 0.560456795350431,
+      "learning_rate": 8.158713862325782e-07,
+      "loss": 0.5356,
+      "step": 1136
+    },
+    {
+      "epoch": 11.03883495145631,
+      "grad_norm": 0.5244174908608809,
+      "learning_rate": 8.12117926224385e-07,
+      "loss": 0.4204,
+      "step": 1137
+    },
+    {
+      "epoch": 11.048543689320388,
+      "grad_norm": 0.5586110284911971,
+      "learning_rate": 8.08371445080093e-07,
+      "loss": 0.4261,
+      "step": 1138
+    },
+    {
+      "epoch": 11.058252427184467,
+      "grad_norm": 0.5836602488957451,
+      "learning_rate": 8.04631958290229e-07,
+      "loss": 0.4734,
+      "step": 1139
+    },
+    {
+      "epoch": 11.067961165048544,
+      "grad_norm": 0.5112021875362739,
+      "learning_rate": 8.008994813163995e-07,
+      "loss": 0.6441,
+      "step": 1140
+    },
+    {
+      "epoch": 11.077669902912621,
+      "grad_norm": 0.5202806365311643,
+      "learning_rate": 7.971740295912289e-07,
+      "loss": 0.4098,
+      "step": 1141
+    },
+    {
+      "epoch": 11.087378640776699,
+      "grad_norm": 0.5160042718236523,
+      "learning_rate": 7.934556185182928e-07,
+      "loss": 0.4041,
+      "step": 1142
+    },
+    {
+      "epoch": 11.097087378640778,
+      "grad_norm": 0.5204797673157634,
+      "learning_rate": 7.897442634720576e-07,
+      "loss": 0.372,
+      "step": 1143
+    },
+    {
+      "epoch": 11.106796116504855,
+      "grad_norm": 0.5067704301983367,
+      "learning_rate": 7.860399797978138e-07,
+      "loss": 0.5713,
+      "step": 1144
+    },
+    {
+      "epoch": 11.116504854368932,
+      "grad_norm": 0.4880515826350417,
+      "learning_rate": 7.823427828116148e-07,
+      "loss": 0.6707,
+      "step": 1145
+    },
+    {
+      "epoch": 11.12621359223301,
+      "grad_norm": 0.5001107410419231,
+      "learning_rate": 7.786526878002126e-07,
+      "loss": 0.4481,
+      "step": 1146
+    },
+    {
+      "epoch": 11.135922330097088,
+      "grad_norm": 0.5033124797385321,
+      "learning_rate": 7.749697100209947e-07,
+      "loss": 0.4927,
+      "step": 1147
+    },
+    {
+      "epoch": 11.145631067961165,
+      "grad_norm": 0.5491536863139684,
+      "learning_rate": 7.7129386470192e-07,
+      "loss": 0.4699,
+      "step": 1148
+    },
+    {
+      "epoch": 11.155339805825243,
+      "grad_norm": 0.5559995427878067,
+      "learning_rate": 7.6762516704146e-07,
+      "loss": 0.4684,
+      "step": 1149
+    },
+    {
+      "epoch": 11.16504854368932,
+      "grad_norm": 0.4642492344086255,
+      "learning_rate": 7.6396363220853e-07,
+      "loss": 0.6739,
+      "step": 1150
+    },
+    {
+      "epoch": 11.174757281553399,
+      "grad_norm": 0.488509927755893,
+      "learning_rate": 7.603092753424298e-07,
+      "loss": 0.5208,
+      "step": 1151
+    },
+    {
+      "epoch": 11.184466019417476,
+      "grad_norm": 0.5224426261949789,
+      "learning_rate": 7.566621115527811e-07,
+      "loss": 0.3606,
+      "step": 1152
+    },
+    {
+      "epoch": 11.194174757281553,
+      "grad_norm": 0.5187700185641912,
+      "learning_rate": 7.530221559194643e-07,
+      "loss": 0.3599,
+      "step": 1153
+    },
+    {
+      "epoch": 11.20388349514563,
+      "grad_norm": 0.46970734175456236,
+      "learning_rate": 7.493894234925558e-07,
+      "loss": 0.4689,
+      "step": 1154
+    },
+    {
+      "epoch": 11.21359223300971,
+      "grad_norm": 0.5340977887603126,
+      "learning_rate": 7.457639292922675e-07,
+      "loss": 0.4931,
+      "step": 1155
+    },
+    {
+      "epoch": 11.223300970873787,
+      "grad_norm": 0.6399746017320636,
+      "learning_rate": 7.421456883088826e-07,
+      "loss": 0.4575,
+      "step": 1156
+    },
+    {
+      "epoch": 11.233009708737864,
+      "grad_norm": 0.5315209210773844,
+      "learning_rate": 7.385347155026934e-07,
+      "loss": 0.6337,
+      "step": 1157
+    },
+    {
+      "epoch": 11.242718446601941,
+      "grad_norm": 0.5118418712712183,
+      "learning_rate": 7.349310258039441e-07,
+      "loss": 0.6285,
+      "step": 1158
+    },
+    {
+      "epoch": 11.25242718446602,
+      "grad_norm": 0.530069361401702,
+      "learning_rate": 7.31334634112762e-07,
+      "loss": 0.7002,
+      "step": 1159
+    },
+    {
+      "epoch": 11.262135922330097,
+      "grad_norm": 0.5172157588149657,
+      "learning_rate": 7.277455552991011e-07,
+      "loss": 0.6858,
+      "step": 1160
+    },
+    {
+      "epoch": 11.271844660194175,
+      "grad_norm": 0.5749812236582913,
+      "learning_rate": 7.241638042026783e-07,
+      "loss": 0.3873,
+      "step": 1161
+    },
+    {
+      "epoch": 11.281553398058252,
+      "grad_norm": 0.5951213032992472,
+      "learning_rate": 7.20589395632913e-07,
+      "loss": 0.4891,
+      "step": 1162
+    },
+    {
+      "epoch": 11.29126213592233,
+      "grad_norm": 0.5134339950348145,
+      "learning_rate": 7.170223443688654e-07,
+      "loss": 0.7757,
+      "step": 1163
+    },
+    {
+      "epoch": 11.300970873786408,
+      "grad_norm": 0.48002921717189445,
+      "learning_rate": 7.134626651591758e-07,
+      "loss": 0.5223,
+      "step": 1164
+    },
+    {
+      "epoch": 11.310679611650485,
+      "grad_norm": 0.5324335789694284,
+      "learning_rate": 7.099103727220024e-07,
+      "loss": 0.267,
+      "step": 1165
+    },
+    {
+      "epoch": 11.320388349514563,
+      "grad_norm": 0.5082511836996648,
+      "learning_rate": 7.063654817449638e-07,
+      "loss": 0.7224,
+      "step": 1166
+    },
+    {
+      "epoch": 11.330097087378642,
+      "grad_norm": 0.4699534351048162,
+      "learning_rate": 7.028280068850734e-07,
+      "loss": 0.3477,
+      "step": 1167
+    },
+    {
+      "epoch": 11.339805825242719,
+      "grad_norm": 0.5290710230891463,
+      "learning_rate": 6.992979627686821e-07,
+      "loss": 0.5673,
+      "step": 1168
+    },
+    {
+      "epoch": 11.349514563106796,
+      "grad_norm": 0.5635904119108134,
+      "learning_rate": 6.957753639914175e-07,
+      "loss": 0.6496,
+      "step": 1169
+    },
+    {
+      "epoch": 11.359223300970873,
+      "grad_norm": 0.4785277786459809,
+      "learning_rate": 6.922602251181221e-07,
+      "loss": 0.5035,
+      "step": 1170
+    },
+    {
+      "epoch": 11.368932038834952,
+      "grad_norm": 0.48631251601303854,
+      "learning_rate": 6.887525606827947e-07,
+      "loss": 0.5194,
+      "step": 1171
+    },
+    {
+      "epoch": 11.37864077669903,
+      "grad_norm": 0.4530281507214442,
+      "learning_rate": 6.852523851885295e-07,
+      "loss": 0.2014,
+      "step": 1172
+    },
+    {
+      "epoch": 11.388349514563107,
+      "grad_norm": 0.5485167968205401,
+      "learning_rate": 6.817597131074566e-07,
+      "loss": 0.5967,
+      "step": 1173
+    },
+    {
+      "epoch": 11.398058252427184,
+      "grad_norm": 0.5659426811727628,
+      "learning_rate": 6.782745588806811e-07,
+      "loss": 0.4703,
+      "step": 1174
+    },
+    {
+      "epoch": 11.407766990291263,
+      "grad_norm": 0.5844594075015748,
+      "learning_rate": 6.747969369182248e-07,
+      "loss": 0.5578,
+      "step": 1175
+    },
+    {
+      "epoch": 11.41747572815534,
+      "grad_norm": 0.6114807287741806,
+      "learning_rate": 6.713268615989654e-07,
+      "loss": 0.6004,
+      "step": 1176
+    },
+    {
+      "epoch": 11.427184466019417,
+      "grad_norm": 0.536472460310527,
+      "learning_rate": 6.678643472705773e-07,
+      "loss": 0.6322,
+      "step": 1177
+    },
+    {
+      "epoch": 11.436893203883495,
+      "grad_norm": 0.5884548398285022,
+      "learning_rate": 6.644094082494746e-07,
+      "loss": 0.4211,
+      "step": 1178
+    },
+    {
+      "epoch": 11.446601941747574,
+      "grad_norm": 0.5293934881901676,
+      "learning_rate": 6.609620588207474e-07,
+      "loss": 0.4222,
+      "step": 1179
+    },
+    {
+      "epoch": 11.45631067961165,
+      "grad_norm": 0.5298417575144992,
+      "learning_rate": 6.575223132381067e-07,
+      "loss": 0.5372,
+      "step": 1180
+    },
+    {
+      "epoch": 11.466019417475728,
+      "grad_norm": 0.5329508454700512,
+      "learning_rate": 6.540901857238233e-07,
+      "loss": 0.3954,
+      "step": 1181
+    },
+    {
+      "epoch": 11.475728155339805,
+      "grad_norm": 0.5330378729803877,
+      "learning_rate": 6.506656904686698e-07,
+      "loss": 0.4937,
+      "step": 1182
+    },
+    {
+      "epoch": 11.485436893203884,
+      "grad_norm": 0.4883863303942178,
+      "learning_rate": 6.472488416318621e-07,
+      "loss": 0.4685,
+      "step": 1183
+    },
+    {
+      "epoch": 11.495145631067961,
+      "grad_norm": 0.5433426004305278,
+      "learning_rate": 6.438396533410002e-07,
+      "loss": 0.5392,
+      "step": 1184
+    },
+    {
+      "epoch": 11.504854368932039,
+      "grad_norm": 0.5155163462176132,
+      "learning_rate": 6.4043813969201e-07,
+      "loss": 0.4761,
+      "step": 1185
+    },
+    {
+      "epoch": 11.514563106796116,
+      "grad_norm": 0.5282705137900666,
+      "learning_rate": 6.370443147490857e-07,
+      "loss": 0.6882,
+      "step": 1186
+    },
+    {
+      "epoch": 11.524271844660195,
+      "grad_norm": 0.5028031458659833,
+      "learning_rate": 6.336581925446309e-07,
+      "loss": 0.4767,
+      "step": 1187
+    },
+    {
+      "epoch": 11.533980582524272,
+      "grad_norm": 0.5977403377061767,
+      "learning_rate": 6.302797870792007e-07,
+      "loss": 0.7029,
+      "step": 1188
+    },
+    {
+      "epoch": 11.54368932038835,
+      "grad_norm": 0.47504759941605157,
+      "learning_rate": 6.269091123214438e-07,
+      "loss": 0.4299,
+      "step": 1189
+    },
+    {
+      "epoch": 11.553398058252426,
+      "grad_norm": 0.5677960420156702,
+      "learning_rate": 6.235461822080449e-07,
+      "loss": 0.7392,
+      "step": 1190
+    },
+    {
+      "epoch": 11.563106796116505,
+      "grad_norm": 0.5148644708372068,
+      "learning_rate": 6.201910106436673e-07,
+      "loss": 0.462,
+      "step": 1191
+    },
+    {
+      "epoch": 11.572815533980583,
+      "grad_norm": 0.5634224658071493,
+      "learning_rate": 6.168436115008941e-07,
+      "loss": 0.6542,
+      "step": 1192
+    },
+    {
+      "epoch": 11.58252427184466,
+      "grad_norm": 0.5270200209649619,
+      "learning_rate": 6.135039986201744e-07,
+      "loss": 0.5779,
+      "step": 1193
+    },
+    {
+      "epoch": 11.592233009708737,
+      "grad_norm": 0.572184122415712,
+      "learning_rate": 6.101721858097606e-07,
+      "loss": 0.5775,
+      "step": 1194
+    },
+    {
+      "epoch": 11.601941747572816,
+      "grad_norm": 0.5022620535938661,
+      "learning_rate": 6.068481868456558e-07,
+      "loss": 0.535,
+      "step": 1195
+    },
+    {
+      "epoch": 11.611650485436893,
+      "grad_norm": 0.5002295261145933,
+      "learning_rate": 6.035320154715549e-07,
+      "loss": 0.4669,
+      "step": 1196
+    },
+    {
+      "epoch": 11.62135922330097,
+      "grad_norm": 0.5684134153610501,
+      "learning_rate": 6.00223685398788e-07,
+      "loss": 0.4175,
+      "step": 1197
+    },
+    {
+      "epoch": 11.631067961165048,
+      "grad_norm": 0.5951925773731871,
+      "learning_rate": 5.969232103062647e-07,
+      "loss": 0.4168,
+      "step": 1198
+    },
+    {
+      "epoch": 11.640776699029127,
+      "grad_norm": 0.5451041361161728,
+      "learning_rate": 5.936306038404158e-07,
+      "loss": 0.3661,
+      "step": 1199
+    },
+    {
+      "epoch": 11.650485436893204,
+      "grad_norm": 0.5600180852134692,
+      "learning_rate": 5.903458796151382e-07,
+      "loss": 0.832,
+      "step": 1200
+    },
+    {
+      "epoch": 11.660194174757281,
+      "grad_norm": 0.5679137831594563,
+      "learning_rate": 5.870690512117377e-07,
+      "loss": 0.4264,
+      "step": 1201
+    },
+    {
+      "epoch": 11.669902912621358,
+      "grad_norm": 0.5512583377764644,
+      "learning_rate": 5.838001321788744e-07,
+      "loss": 0.5387,
+      "step": 1202
+    },
+    {
+      "epoch": 11.679611650485437,
+      "grad_norm": 0.5202492691551494,
+      "learning_rate": 5.80539136032505e-07,
+      "loss": 0.4072,
+      "step": 1203
+    },
+    {
+      "epoch": 11.689320388349515,
+      "grad_norm": 0.5303414222747518,
+      "learning_rate": 5.772860762558269e-07,
+      "loss": 0.7383,
+      "step": 1204
+    },
+    {
+      "epoch": 11.699029126213592,
+      "grad_norm": 0.5747472031996176,
+      "learning_rate": 5.740409662992244e-07,
+      "loss": 0.4772,
+      "step": 1205
+    },
+    {
+      "epoch": 11.70873786407767,
+      "grad_norm": 0.5344538177636208,
+      "learning_rate": 5.708038195802098e-07,
+      "loss": 0.5894,
+      "step": 1206
+    },
+    {
+      "epoch": 11.718446601941748,
+      "grad_norm": 0.6162700091163664,
+      "learning_rate": 5.675746494833733e-07,
+      "loss": 0.6199,
+      "step": 1207
+    },
+    {
+      "epoch": 11.728155339805825,
+      "grad_norm": 0.5212416584936878,
+      "learning_rate": 5.643534693603214e-07,
+      "loss": 0.5544,
+      "step": 1208
+    },
+    {
+      "epoch": 11.737864077669903,
+      "grad_norm": 0.5002178765849649,
+      "learning_rate": 5.61140292529625e-07,
+      "loss": 0.4494,
+      "step": 1209
+    },
+    {
+      "epoch": 11.74757281553398,
+      "grad_norm": 0.49081842853898583,
+      "learning_rate": 5.579351322767643e-07,
+      "loss": 0.4325,
+      "step": 1210
+    },
+    {
+      "epoch": 11.757281553398059,
+      "grad_norm": 0.677193839042659,
+      "learning_rate": 5.547380018540735e-07,
+      "loss": 0.582,
+      "step": 1211
+    },
+    {
+      "epoch": 11.766990291262136,
+      "grad_norm": 0.6177020335391481,
+      "learning_rate": 5.515489144806862e-07,
+      "loss": 0.3849,
+      "step": 1212
+    },
+    {
+      "epoch": 11.776699029126213,
+      "grad_norm": 0.5418788069046763,
+      "learning_rate": 5.483678833424796e-07,
+      "loss": 0.655,
+      "step": 1213
+    },
+    {
+      "epoch": 11.78640776699029,
+      "grad_norm": 0.5041790619256601,
+      "learning_rate": 5.451949215920221e-07,
+      "loss": 0.5462,
+      "step": 1214
+    },
+    {
+      "epoch": 11.79611650485437,
+      "grad_norm": 0.5833392895268761,
+      "learning_rate": 5.420300423485167e-07,
+      "loss": 0.4945,
+      "step": 1215
+    },
+    {
+      "epoch": 11.805825242718447,
+      "grad_norm": 0.49267471870126267,
+      "learning_rate": 5.38873258697748e-07,
+      "loss": 0.5889,
+      "step": 1216
+    },
+    {
+      "epoch": 11.815533980582524,
+      "grad_norm": 0.48439741671139047,
+      "learning_rate": 5.357245836920286e-07,
+      "loss": 0.5955,
+      "step": 1217
+    },
+    {
+      "epoch": 11.825242718446601,
+      "grad_norm": 0.5572093766556138,
+      "learning_rate": 5.325840303501431e-07,
+      "loss": 0.6049,
+      "step": 1218
+    },
+    {
+      "epoch": 11.83495145631068,
+      "grad_norm": 0.5237802353774021,
+      "learning_rate": 5.29451611657297e-07,
+      "loss": 0.5523,
+      "step": 1219
+    },
+    {
+      "epoch": 11.844660194174757,
+      "grad_norm": 0.5624258745574067,
+      "learning_rate": 5.263273405650601e-07,
+      "loss": 0.6399,
+      "step": 1220
+    },
+    {
+      "epoch": 11.854368932038835,
+      "grad_norm": 0.7999810849053399,
+      "learning_rate": 5.232112299913151e-07,
+      "loss": 0.5699,
+      "step": 1221
+    },
+    {
+      "epoch": 11.864077669902912,
+      "grad_norm": 0.4742127157028219,
+      "learning_rate": 5.201032928202043e-07,
+      "loss": 0.4969,
+      "step": 1222
+    },
+    {
+      "epoch": 11.87378640776699,
+      "grad_norm": 0.5137383431300119,
+      "learning_rate": 5.17003541902075e-07,
+      "loss": 0.772,
+      "step": 1223
+    },
+    {
+      "epoch": 11.883495145631068,
+      "grad_norm": 0.46497503447213573,
+      "learning_rate": 5.139119900534259e-07,
+      "loss": 0.5661,
+      "step": 1224
+    },
+    {
+      "epoch": 11.893203883495145,
+      "grad_norm": 0.6079453183093874,
+      "learning_rate": 5.108286500568562e-07,
+      "loss": 0.3941,
+      "step": 1225
+    },
+    {
+      "epoch": 11.902912621359224,
+      "grad_norm": 0.5273197405999158,
+      "learning_rate": 5.077535346610115e-07,
+      "loss": 0.5251,
+      "step": 1226
+    },
+    {
+      "epoch": 11.912621359223301,
+      "grad_norm": 0.48161444879759807,
+      "learning_rate": 5.046866565805311e-07,
+      "loss": 0.3903,
+      "step": 1227
+    },
+    {
+      "epoch": 11.922330097087379,
+      "grad_norm": 0.49154732760101316,
+      "learning_rate": 5.016280284959957e-07,
+      "loss": 0.538,
+      "step": 1228
+    },
+    {
+      "epoch": 11.932038834951456,
+      "grad_norm": 0.4885700992202129,
+      "learning_rate": 4.985776630538746e-07,
+      "loss": 0.6219,
+      "step": 1229
+    },
+    {
+      "epoch": 11.941747572815533,
+      "grad_norm": 0.5594706359230254,
+      "learning_rate": 4.95535572866474e-07,
+      "loss": 0.4817,
+      "step": 1230
+    },
+    {
+      "epoch": 11.951456310679612,
+      "grad_norm": 0.5257629598562028,
+      "learning_rate": 4.925017705118843e-07,
+      "loss": 0.4894,
+      "step": 1231
+    },
+    {
+      "epoch": 11.96116504854369,
+      "grad_norm": 0.5767484443957133,
+      "learning_rate": 4.89476268533928e-07,
+      "loss": 0.7011,
+      "step": 1232
+    },
+    {
+      "epoch": 11.970873786407767,
+      "grad_norm": 0.5950735781630769,
+      "learning_rate": 4.864590794421092e-07,
+      "loss": 0.7114,
+      "step": 1233
+    },
+    {
+      "epoch": 11.980582524271846,
+      "grad_norm": 0.540369700371263,
+      "learning_rate": 4.834502157115597e-07,
+      "loss": 0.4854,
+      "step": 1234
+    },
+    {
+      "epoch": 11.990291262135923,
+      "grad_norm": 0.5818146270683455,
+      "learning_rate": 4.804496897829883e-07,
+      "loss": 0.2926,
+      "step": 1235
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.5089608589552385,
+      "learning_rate": 4.774575140626317e-07,
+      "loss": 0.6658,
+      "step": 1236
+    },
+    {
+      "epoch": 12.009708737864077,
+      "grad_norm": 0.4547147316741764,
+      "learning_rate": 4.744737009221986e-07,
+      "loss": 0.483,
+      "step": 1237
+    },
+    {
+      "epoch": 12.019417475728156,
+      "grad_norm": 0.5829159491473124,
+      "learning_rate": 4.7149826269882294e-07,
+      "loss": 0.616,
+      "step": 1238
+    },
+    {
+      "epoch": 12.029126213592233,
+      "grad_norm": 0.5043098105333009,
+      "learning_rate": 4.6853121169500914e-07,
+      "loss": 0.4629,
+      "step": 1239
+    },
+    {
+      "epoch": 12.03883495145631,
+      "grad_norm": 0.571022287176784,
+      "learning_rate": 4.6557256017858485e-07,
+      "loss": 0.4794,
+      "step": 1240
+    },
+    {
+      "epoch": 12.048543689320388,
+      "grad_norm": 0.5399154525227847,
+      "learning_rate": 4.626223203826477e-07,
+      "loss": 0.5385,
+      "step": 1241
+    },
+    {
+      "epoch": 12.058252427184467,
+      "grad_norm": 0.5198642674251133,
+      "learning_rate": 4.5968050450551527e-07,
+      "loss": 0.3339,
+      "step": 1242
+    },
+    {
+      "epoch": 12.067961165048544,
+      "grad_norm": 0.49500784542537424,
+      "learning_rate": 4.56747124710675e-07,
+      "loss": 0.4985,
+      "step": 1243
+    },
+    {
+      "epoch": 12.077669902912621,
+      "grad_norm": 0.5475203658604212,
+      "learning_rate": 4.5382219312673364e-07,
+      "loss": 0.5487,
+      "step": 1244
+    },
+    {
+      "epoch": 12.087378640776699,
+      "grad_norm": 0.48035513204470476,
+      "learning_rate": 4.5090572184736863e-07,
+      "loss": 0.3491,
+      "step": 1245
+    },
+    {
+      "epoch": 12.097087378640778,
+      "grad_norm": 0.5166017045705675,
+      "learning_rate": 4.4799772293127486e-07,
+      "loss": 0.752,
+      "step": 1246
+    },
+    {
+      "epoch": 12.106796116504855,
+      "grad_norm": 0.5229778700888286,
+      "learning_rate": 4.4509820840211745e-07,
+      "loss": 0.5296,
+      "step": 1247
+    },
+    {
+      "epoch": 12.116504854368932,
+      "grad_norm": 0.5224021014520674,
+      "learning_rate": 4.422071902484812e-07,
+      "loss": 0.4395,
+      "step": 1248
+    },
+    {
+      "epoch": 12.12621359223301,
+      "grad_norm": 0.4859120821940226,
+      "learning_rate": 4.3932468042382075e-07,
+      "loss": 0.3914,
+      "step": 1249
+    },
+    {
+      "epoch": 12.135922330097088,
+      "grad_norm": 0.49206048744899616,
+      "learning_rate": 4.3645069084641195e-07,
+      "loss": 0.5086,
+      "step": 1250
+    },
+    {
+      "epoch": 12.145631067961165,
+      "grad_norm": 0.5507080078693621,
+      "learning_rate": 4.335852333993018e-07,
+      "loss": 0.6545,
+      "step": 1251
+    },
+    {
+      "epoch": 12.155339805825243,
+      "grad_norm": 0.5256375066912108,
+      "learning_rate": 4.3072831993025895e-07,
+      "loss": 0.6475,
+      "step": 1252
+    },
+    {
+      "epoch": 12.16504854368932,
+      "grad_norm": 0.5165162373708071,
+      "learning_rate": 4.278799622517274e-07,
+      "loss": 0.5665,
+      "step": 1253
+    },
+    {
+      "epoch": 12.174757281553399,
+      "grad_norm": 0.48585634241986353,
+      "learning_rate": 4.2504017214077374e-07,
+      "loss": 0.5531,
+      "step": 1254
+    },
+    {
+      "epoch": 12.184466019417476,
+      "grad_norm": 0.5727064553331418,
+      "learning_rate": 4.222089613390412e-07,
+      "loss": 0.4,
+      "step": 1255
+    },
+    {
+      "epoch": 12.194174757281553,
+      "grad_norm": 0.5655692683446615,
+      "learning_rate": 4.1938634155269944e-07,
+      "loss": 0.5369,
+      "step": 1256
+    },
+    {
+      "epoch": 12.20388349514563,
+      "grad_norm": 0.5339261343495418,
+      "learning_rate": 4.165723244523978e-07,
+      "loss": 0.548,
+      "step": 1257
+    },
+    {
+      "epoch": 12.21359223300971,
+      "grad_norm": 0.5995207842725745,
+      "learning_rate": 4.1376692167321626e-07,
+      "loss": 0.7545,
+      "step": 1258
+    },
+    {
+      "epoch": 12.223300970873787,
+      "grad_norm": 0.49946940718627203,
+      "learning_rate": 4.109701448146164e-07,
+      "loss": 0.4537,
+      "step": 1259
+    },
+    {
+      "epoch": 12.233009708737864,
+      "grad_norm": 0.48406505308218556,
+      "learning_rate": 4.0818200544039484e-07,
+      "loss": 0.5457,
+      "step": 1260
+    },
+    {
+      "epoch": 12.242718446601941,
+      "grad_norm": 0.5761404083738131,
+      "learning_rate": 4.054025150786356e-07,
+      "loss": 0.4459,
+      "step": 1261
+    },
+    {
+      "epoch": 12.25242718446602,
+      "grad_norm": 0.5368464804303047,
+      "learning_rate": 4.026316852216605e-07,
+      "loss": 0.5884,
+      "step": 1262
+    },
+    {
+      "epoch": 12.262135922330097,
+      "grad_norm": 0.5455773855091807,
+      "learning_rate": 3.998695273259834e-07,
+      "loss": 0.4718,
+      "step": 1263
+    },
+    {
+      "epoch": 12.271844660194175,
+      "grad_norm": 0.5462489365415362,
+      "learning_rate": 3.971160528122622e-07,
+      "loss": 0.621,
+      "step": 1264
+    },
+    {
+      "epoch": 12.281553398058252,
+      "grad_norm": 0.556552861635233,
+      "learning_rate": 3.9437127306525295e-07,
+      "loss": 0.6744,
+      "step": 1265
+    },
+    {
+      "epoch": 12.29126213592233,
+      "grad_norm": 0.5297638757980944,
+      "learning_rate": 3.9163519943375973e-07,
+      "loss": 0.4573,
+      "step": 1266
+    },
+    {
+      "epoch": 12.300970873786408,
+      "grad_norm": 0.49902627020942136,
+      "learning_rate": 3.889078432305904e-07,
+      "loss": 0.4568,
+      "step": 1267
+    },
+    {
+      "epoch": 12.310679611650485,
+      "grad_norm": 0.5535869111612749,
+      "learning_rate": 3.8618921573250896e-07,
+      "loss": 0.528,
+      "step": 1268
+    },
+    {
+      "epoch": 12.320388349514563,
+      "grad_norm": 0.5204509920890341,
+      "learning_rate": 3.834793281801891e-07,
+      "loss": 0.3842,
+      "step": 1269
+    },
+    {
+      "epoch": 12.330097087378642,
+      "grad_norm": 0.5692383110945788,
+      "learning_rate": 3.8077819177816695e-07,
+      "loss": 0.5053,
+      "step": 1270
+    },
+    {
+      "epoch": 12.339805825242719,
+      "grad_norm": 0.47822476540581443,
+      "learning_rate": 3.780858176947963e-07,
+      "loss": 0.3833,
+      "step": 1271
+    },
+    {
+      "epoch": 12.349514563106796,
+      "grad_norm": 0.5186445853503381,
+      "learning_rate": 3.754022170622007e-07,
+      "loss": 0.3893,
+      "step": 1272
+    },
+    {
+      "epoch": 12.359223300970873,
+      "grad_norm": 0.49920957568724805,
+      "learning_rate": 3.7272740097622884e-07,
+      "loss": 0.6484,
+      "step": 1273
+    },
+    {
+      "epoch": 12.368932038834952,
+      "grad_norm": 0.5579303155302804,
+      "learning_rate": 3.700613804964073e-07,
+      "loss": 0.6875,
+      "step": 1274
+    },
+    {
+      "epoch": 12.37864077669903,
+      "grad_norm": 0.5539922366049171,
+      "learning_rate": 3.6740416664589634e-07,
+      "loss": 0.5997,
+      "step": 1275
+    },
+    {
+      "epoch": 12.388349514563107,
+      "grad_norm": 0.4840847496708577,
+      "learning_rate": 3.6475577041144324e-07,
+      "loss": 0.7178,
+      "step": 1276
+    },
+    {
+      "epoch": 12.398058252427184,
+      "grad_norm": 0.5501337207898733,
+      "learning_rate": 3.6211620274333727e-07,
+      "loss": 0.3739,
+      "step": 1277
+    },
+    {
+      "epoch": 12.407766990291263,
+      "grad_norm": 0.5541786710200026,
+      "learning_rate": 3.594854745553636e-07,
+      "loss": 0.7565,
+      "step": 1278
+    },
+    {
+      "epoch": 12.41747572815534,
+      "grad_norm": 0.5175636279750693,
+      "learning_rate": 3.568635967247605e-07,
+      "loss": 0.7672,
+      "step": 1279
+    },
+    {
+      "epoch": 12.427184466019417,
+      "grad_norm": 0.46450923628601365,
+      "learning_rate": 3.5425058009217193e-07,
+      "loss": 0.5094,
+      "step": 1280
+    },
+    {
+      "epoch": 12.436893203883495,
+      "grad_norm": 0.5786871968641614,
+      "learning_rate": 3.516464354616031e-07,
+      "loss": 0.6925,
+      "step": 1281
+    },
+    {
+      "epoch": 12.446601941747574,
+      "grad_norm": 0.5724996419603651,
+      "learning_rate": 3.4905117360037683e-07,
+      "loss": 0.5977,
+      "step": 1282
+    },
+    {
+      "epoch": 12.45631067961165,
+      "grad_norm": 0.540971499700757,
+      "learning_rate": 3.4646480523908813e-07,
+      "loss": 0.4706,
+      "step": 1283
+    },
+    {
+      "epoch": 12.466019417475728,
+      "grad_norm": 0.5182006459005003,
+      "learning_rate": 3.43887341071561e-07,
+      "loss": 0.6454,
+      "step": 1284
+    },
+    {
+      "epoch": 12.475728155339805,
+      "grad_norm": 0.5425832252884726,
+      "learning_rate": 3.413187917548019e-07,
+      "loss": 0.339,
+      "step": 1285
+    },
+    {
+      "epoch": 12.485436893203884,
+      "grad_norm": 0.539013979648478,
+      "learning_rate": 3.3875916790895883e-07,
+      "loss": 0.828,
+      "step": 1286
+    },
+    {
+      "epoch": 12.495145631067961,
+      "grad_norm": 0.6096747704359214,
+      "learning_rate": 3.3620848011727437e-07,
+      "loss": 0.4499,
+      "step": 1287
+    },
+    {
+      "epoch": 12.504854368932039,
+      "grad_norm": 0.5278542481454789,
+      "learning_rate": 3.336667389260445e-07,
+      "loss": 0.422,
+      "step": 1288
+    },
+    {
+      "epoch": 12.514563106796116,
+      "grad_norm": 0.5403848073760052,
+      "learning_rate": 3.311339548445727e-07,
+      "loss": 0.4152,
+      "step": 1289
+    },
+    {
+      "epoch": 12.524271844660195,
+      "grad_norm": 0.48656658093770444,
+      "learning_rate": 3.2861013834512844e-07,
+      "loss": 0.5419,
+      "step": 1290
+    },
+    {
+      "epoch": 12.533980582524272,
+      "grad_norm": 0.5218058868156833,
+      "learning_rate": 3.2609529986290246e-07,
+      "loss": 0.4006,
+      "step": 1291
+    },
+    {
+      "epoch": 12.54368932038835,
+      "grad_norm": 0.5053568098333768,
+      "learning_rate": 3.235894497959649e-07,
+      "loss": 0.402,
+      "step": 1292
+    },
+    {
+      "epoch": 12.553398058252426,
+      "grad_norm": 0.5028842562515405,
+      "learning_rate": 3.2109259850522045e-07,
+      "loss": 0.5259,
+      "step": 1293
+    },
+    {
+      "epoch": 12.563106796116505,
+      "grad_norm": 0.5127746566227459,
+      "learning_rate": 3.186047563143685e-07,
+      "loss": 0.524,
+      "step": 1294
+    },
+    {
+      "epoch": 12.572815533980583,
+      "grad_norm": 0.502598831483758,
+      "learning_rate": 3.161259335098571e-07,
+      "loss": 0.5442,
+      "step": 1295
+    },
+    {
+      "epoch": 12.58252427184466,
+      "grad_norm": 0.5953762476232632,
+      "learning_rate": 3.1365614034084224e-07,
+      "loss": 0.286,
+      "step": 1296
+    },
+    {
+      "epoch": 12.592233009708737,
+      "grad_norm": 0.5102868867073259,
+      "learning_rate": 3.111953870191459e-07,
+      "loss": 0.447,
+      "step": 1297
+    },
+    {
+      "epoch": 12.601941747572816,
+      "grad_norm": 0.5245488599975241,
+      "learning_rate": 3.087436837192118e-07,
+      "loss": 0.8385,
+      "step": 1298
+    },
+    {
+      "epoch": 12.611650485436893,
+      "grad_norm": 0.5645317531008295,
+      "learning_rate": 3.0630104057806616e-07,
+      "loss": 0.5325,
+      "step": 1299
+    },
+    {
+      "epoch": 12.62135922330097,
+      "grad_norm": 0.5131734216726613,
+      "learning_rate": 3.0386746769527323e-07,
+      "loss": 0.542,
+      "step": 1300
+    },
+    {
+      "epoch": 12.631067961165048,
+      "grad_norm": 0.521794966830633,
+      "learning_rate": 3.0144297513289483e-07,
+      "loss": 0.4628,
+      "step": 1301
+    },
+    {
+      "epoch": 12.640776699029127,
+      "grad_norm": 0.5396439134195453,
+      "learning_rate": 2.9902757291544905e-07,
+      "loss": 0.548,
+      "step": 1302
+    },
+    {
+      "epoch": 12.650485436893204,
+      "grad_norm": 0.5622776596142265,
+      "learning_rate": 2.966212710298674e-07,
+      "loss": 0.371,
+      "step": 1303
+    },
+    {
+      "epoch": 12.660194174757281,
+      "grad_norm": 0.4790177990779847,
+      "learning_rate": 2.94224079425455e-07,
+      "loss": 0.432,
+      "step": 1304
+    },
+    {
+      "epoch": 12.669902912621358,
+      "grad_norm": 0.539797943283852,
+      "learning_rate": 2.9183600801384853e-07,
+      "loss": 0.4002,
+      "step": 1305
+    },
+    {
+      "epoch": 12.679611650485437,
+      "grad_norm": 0.5483551448669917,
+      "learning_rate": 2.8945706666897555e-07,
+      "loss": 0.3479,
+      "step": 1306
+    },
+    {
+      "epoch": 12.689320388349515,
+      "grad_norm": 0.49713629082843547,
+      "learning_rate": 2.870872652270129e-07,
+      "loss": 0.3679,
+      "step": 1307
+    },
+    {
+      "epoch": 12.699029126213592,
+      "grad_norm": 0.4992602508121686,
+      "learning_rate": 2.8472661348634883e-07,
+      "loss": 0.3935,
+      "step": 1308
+    },
+    {
+      "epoch": 12.70873786407767,
+      "grad_norm": 0.5115896966629423,
+      "learning_rate": 2.82375121207539e-07,
+      "loss": 0.4042,
+      "step": 1309
+    },
+    {
+      "epoch": 12.718446601941748,
+      "grad_norm": 0.480675662078779,
+      "learning_rate": 2.8003279811326724e-07,
+      "loss": 0.4149,
+      "step": 1310
+    },
+    {
+      "epoch": 12.728155339805825,
+      "grad_norm": 0.5305606289822804,
+      "learning_rate": 2.776996538883062e-07,
+      "loss": 0.2728,
+      "step": 1311
+    },
+    {
+      "epoch": 12.737864077669903,
+      "grad_norm": 0.49638478140651515,
+      "learning_rate": 2.7537569817947694e-07,
+      "loss": 0.5558,
+      "step": 1312
+    },
+    {
+      "epoch": 12.74757281553398,
+      "grad_norm": 0.5161627991330044,
+      "learning_rate": 2.730609405956083e-07,
+      "loss": 0.4527,
+      "step": 1313
+    },
+    {
+      "epoch": 12.757281553398059,
+      "grad_norm": 0.5053870345750662,
+      "learning_rate": 2.707553907074989e-07,
+      "loss": 0.4976,
+      "step": 1314
+    },
+    {
+      "epoch": 12.766990291262136,
+      "grad_norm": 0.6117896664980851,
+      "learning_rate": 2.684590580478749e-07,
+      "loss": 0.7138,
+      "step": 1315
+    },
+    {
+      "epoch": 12.776699029126213,
+      "grad_norm": 0.5424702681942861,
+      "learning_rate": 2.6617195211135343e-07,
+      "loss": 0.7378,
+      "step": 1316
+    },
+    {
+      "epoch": 12.78640776699029,
+      "grad_norm": 0.45282710489101363,
+      "learning_rate": 2.638940823544012e-07,
+      "loss": 0.4519,
+      "step": 1317
+    },
+    {
+      "epoch": 12.79611650485437,
+      "grad_norm": 0.5634124208174777,
+      "learning_rate": 2.6162545819529624e-07,
+      "loss": 0.5306,
+      "step": 1318
+    },
+    {
+      "epoch": 12.805825242718447,
+      "grad_norm": 0.5317119729976713,
+      "learning_rate": 2.593660890140895e-07,
+      "loss": 0.5186,
+      "step": 1319
+    },
+    {
+      "epoch": 12.815533980582524,
+      "grad_norm": 0.5164784363125361,
+      "learning_rate": 2.57115984152565e-07,
+      "loss": 0.4913,
+      "step": 1320
+    },
+    {
+      "epoch": 12.825242718446601,
+      "grad_norm": 0.583079797245975,
+      "learning_rate": 2.548751529142018e-07,
+      "loss": 0.5579,
+      "step": 1321
+    },
+    {
+      "epoch": 12.83495145631068,
+      "grad_norm": 0.4937883663752087,
+      "learning_rate": 2.526436045641351e-07,
+      "loss": 0.6115,
+      "step": 1322
+    },
+    {
+      "epoch": 12.844660194174757,
+      "grad_norm": 0.4886740713037102,
+      "learning_rate": 2.504213483291193e-07,
+      "loss": 0.5438,
+      "step": 1323
+    },
+    {
+      "epoch": 12.854368932038835,
+      "grad_norm": 0.4742082781225294,
+      "learning_rate": 2.482083933974883e-07,
+      "loss": 0.3571,
+      "step": 1324
+    },
+    {
+      "epoch": 12.864077669902912,
+      "grad_norm": 0.5090813269543996,
+      "learning_rate": 2.4600474891911696e-07,
+      "loss": 0.626,
+      "step": 1325
+    },
+    {
+      "epoch": 12.87378640776699,
+      "grad_norm": 0.5756284766855548,
+      "learning_rate": 2.43810424005386e-07,
+      "loss": 0.4204,
+      "step": 1326
+    },
+    {
+      "epoch": 12.883495145631068,
+      "grad_norm": 0.5302071313339889,
+      "learning_rate": 2.416254277291416e-07,
+      "loss": 0.4967,
+      "step": 1327
+    },
+    {
+      "epoch": 12.893203883495145,
+      "grad_norm": 0.519224974871366,
+      "learning_rate": 2.3944976912465916e-07,
+      "loss": 0.3222,
+      "step": 1328
+    },
+    {
+      "epoch": 12.902912621359224,
+      "grad_norm": 0.5125376672237484,
+      "learning_rate": 2.3728345718760622e-07,
+      "loss": 0.604,
+      "step": 1329
+    },
+    {
+      "epoch": 12.912621359223301,
+      "grad_norm": 0.5025700910016356,
+      "learning_rate": 2.3512650087500338e-07,
+      "loss": 0.6674,
+      "step": 1330
+    },
+    {
+      "epoch": 12.922330097087379,
+      "grad_norm": 0.5472841146691036,
+      "learning_rate": 2.3297890910519093e-07,
+      "loss": 0.4614,
+      "step": 1331
+    },
+    {
+      "epoch": 12.932038834951456,
+      "grad_norm": 0.6618994321490155,
+      "learning_rate": 2.3084069075778758e-07,
+      "loss": 0.3552,
+      "step": 1332
+    },
+    {
+      "epoch": 12.941747572815533,
+      "grad_norm": 0.5653535449986341,
+      "learning_rate": 2.287118546736572e-07,
+      "loss": 0.515,
+      "step": 1333
+    },
+    {
+      "epoch": 12.951456310679612,
+      "grad_norm": 0.5393697804104346,
+      "learning_rate": 2.2659240965487023e-07,
+      "loss": 0.704,
+      "step": 1334
+    },
+    {
+      "epoch": 12.96116504854369,
+      "grad_norm": 0.5516311443510744,
+      "learning_rate": 2.2448236446466847e-07,
+      "loss": 0.7699,
+      "step": 1335
+    },
+    {
+      "epoch": 12.970873786407767,
+      "grad_norm": 0.49351094018608216,
+      "learning_rate": 2.2238172782742763e-07,
+      "loss": 0.4795,
+      "step": 1336
+    },
+    {
+      "epoch": 12.980582524271846,
+      "grad_norm": 0.6006692564530559,
+      "learning_rate": 2.2029050842862277e-07,
+      "loss": 0.5718,
+      "step": 1337
+    },
+    {
+      "epoch": 12.990291262135923,
+      "grad_norm": 0.5026296445519245,
+      "learning_rate": 2.1820871491479102e-07,
+      "loss": 0.5669,
+      "step": 1338
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 0.4662986293212389,
+      "learning_rate": 2.1613635589349756e-07,
+      "loss": 0.6033,
+      "step": 1339
+    },
+    {
+      "epoch": 13.009708737864077,
+      "grad_norm": 0.5762881710629336,
+      "learning_rate": 2.140734399332975e-07,
+      "loss": 0.4637,
+      "step": 1340
+    },
+    {
+      "epoch": 13.019417475728156,
+      "grad_norm": 0.5100299787885914,
+      "learning_rate": 2.1201997556370284e-07,
+      "loss": 0.6593,
+      "step": 1341
+    },
+    {
+      "epoch": 13.029126213592233,
+      "grad_norm": 0.5142901147631517,
+      "learning_rate": 2.0997597127514507e-07,
+      "loss": 0.4614,
+      "step": 1342
+    },
+    {
+      "epoch": 13.03883495145631,
+      "grad_norm": 0.5610525704881512,
+      "learning_rate": 2.079414355189427e-07,
+      "loss": 0.4303,
+      "step": 1343
+    },
+    {
+      "epoch": 13.048543689320388,
+      "grad_norm": 0.565400240725852,
+      "learning_rate": 2.059163767072639e-07,
+      "loss": 0.6574,
+      "step": 1344
+    },
+    {
+      "epoch": 13.058252427184467,
+      "grad_norm": 0.49074149271912093,
+      "learning_rate": 2.0390080321309236e-07,
+      "loss": 0.3981,
+      "step": 1345
+    },
+    {
+      "epoch": 13.067961165048544,
+      "grad_norm": 0.5120110967792576,
+      "learning_rate": 2.01894723370194e-07,
+      "loss": 0.2239,
+      "step": 1346
+    },
+    {
+      "epoch": 13.077669902912621,
+      "grad_norm": 0.5018096971105644,
+      "learning_rate": 1.9989814547308056e-07,
+      "loss": 0.4467,
+      "step": 1347
+    },
+    {
+      "epoch": 13.087378640776699,
+      "grad_norm": 0.5476918123606207,
+      "learning_rate": 1.9791107777697633e-07,
+      "loss": 0.6297,
+      "step": 1348
+    },
+    {
+      "epoch": 13.097087378640778,
+      "grad_norm": 0.572988197563463,
+      "learning_rate": 1.9593352849778453e-07,
+      "loss": 0.6092,
+      "step": 1349
+    },
+    {
+      "epoch": 13.106796116504855,
+      "grad_norm": 0.5261661637028531,
+      "learning_rate": 1.9396550581205208e-07,
+      "loss": 0.4294,
+      "step": 1350
+    },
+    {
+      "epoch": 13.116504854368932,
+      "grad_norm": 0.50494082943497,
+      "learning_rate": 1.920070178569361e-07,
+      "loss": 0.387,
+      "step": 1351
+    },
+    {
+      "epoch": 13.12621359223301,
+      "grad_norm": 0.5111855540996291,
+      "learning_rate": 1.900580727301718e-07,
+      "loss": 0.5553,
+      "step": 1352
+    },
+    {
+      "epoch": 13.135922330097088,
+      "grad_norm": 0.5066506430565386,
+      "learning_rate": 1.8811867849003684e-07,
+      "loss": 0.4342,
+      "step": 1353
+    },
+    {
+      "epoch": 13.145631067961165,
+      "grad_norm": 0.5595564814476269,
+      "learning_rate": 1.8618884315531939e-07,
+      "loss": 0.8207,
+      "step": 1354
+    },
+    {
+      "epoch": 13.155339805825243,
+      "grad_norm": 0.5505727997556397,
+      "learning_rate": 1.8426857470528414e-07,
+      "loss": 0.4033,
+      "step": 1355
+    },
+    {
+      "epoch": 13.16504854368932,
+      "grad_norm": 0.5766520644750025,
+      "learning_rate": 1.8235788107963948e-07,
+      "loss": 0.5713,
+      "step": 1356
+    },
+    {
+      "epoch": 13.174757281553399,
+      "grad_norm": 0.5600331852928223,
+      "learning_rate": 1.8045677017850595e-07,
+      "loss": 0.3118,
+      "step": 1357
+    },
+    {
+      "epoch": 13.184466019417476,
+      "grad_norm": 0.4745484708437102,
+      "learning_rate": 1.785652498623816e-07,
+      "loss": 0.4337,
+      "step": 1358
+    },
+    {
+      "epoch": 13.194174757281553,
+      "grad_norm": 0.527479211214423,
+      "learning_rate": 1.7668332795211074e-07,
+      "loss": 0.6401,
+      "step": 1359
+    },
+    {
+      "epoch": 13.20388349514563,
+      "grad_norm": 0.5317399547389916,
+      "learning_rate": 1.7481101222885126e-07,
+      "loss": 0.4149,
+      "step": 1360
+    },
+    {
+      "epoch": 13.21359223300971,
+      "grad_norm": 0.5417102266482062,
+      "learning_rate": 1.7294831043404264e-07,
+      "loss": 0.4646,
+      "step": 1361
+    },
+    {
+      "epoch": 13.223300970873787,
+      "grad_norm": 0.4964064829393023,
+      "learning_rate": 1.7109523026937302e-07,
+      "loss": 0.4473,
+      "step": 1362
+    },
+    {
+      "epoch": 13.233009708737864,
+      "grad_norm": 0.5687847935424973,
+      "learning_rate": 1.6925177939674936e-07,
+      "loss": 0.6734,
+      "step": 1363
+    },
+    {
+      "epoch": 13.242718446601941,
+      "grad_norm": 0.5204693568021522,
+      "learning_rate": 1.6741796543826321e-07,
+      "loss": 0.5916,
+      "step": 1364
+    },
+    {
+      "epoch": 13.25242718446602,
+      "grad_norm": 0.5230106632722077,
+      "learning_rate": 1.6559379597616136e-07,
+      "loss": 0.4952,
+      "step": 1365
+    },
+    {
+      "epoch": 13.262135922330097,
+      "grad_norm": 0.5125645899558829,
+      "learning_rate": 1.6377927855281362e-07,
+      "loss": 0.5947,
+      "step": 1366
+    },
+    {
+      "epoch": 13.271844660194175,
+      "grad_norm": 0.5122079677737931,
+      "learning_rate": 1.6197442067068136e-07,
+      "loss": 0.561,
+      "step": 1367
+    },
+    {
+      "epoch": 13.281553398058252,
+      "grad_norm": 0.5645728565635226,
+      "learning_rate": 1.6017922979228662e-07,
+      "loss": 0.3307,
+      "step": 1368
+    },
+    {
+      "epoch": 13.29126213592233,
+      "grad_norm": 0.5262239005191955,
+      "learning_rate": 1.5839371334018193e-07,
+      "loss": 0.6279,
+      "step": 1369
+    },
+    {
+      "epoch": 13.300970873786408,
+      "grad_norm": 0.5325025169012865,
+      "learning_rate": 1.5661787869691858e-07,
+      "loss": 0.5475,
+      "step": 1370
+    },
+    {
+      "epoch": 13.310679611650485,
+      "grad_norm": 0.5606899292139784,
+      "learning_rate": 1.5485173320501673e-07,
+      "loss": 0.583,
+      "step": 1371
+    },
+    {
+      "epoch": 13.320388349514563,
+      "grad_norm": 0.5167998308377985,
+      "learning_rate": 1.5309528416693503e-07,
+      "loss": 0.4803,
+      "step": 1372
+    },
+    {
+      "epoch": 13.330097087378642,
+      "grad_norm": 0.5309442940524358,
+      "learning_rate": 1.513485388450403e-07,
+      "loss": 0.5416,
+      "step": 1373
+    },
+    {
+      "epoch": 13.339805825242719,
+      "grad_norm": 0.4917882458055186,
+      "learning_rate": 1.4961150446157759e-07,
+      "loss": 0.5256,
+      "step": 1374
+    },
+    {
+      "epoch": 13.349514563106796,
+      "grad_norm": 0.48576191372876565,
+      "learning_rate": 1.4788418819864037e-07,
+      "loss": 0.4438,
+      "step": 1375
+    },
+    {
+      "epoch": 13.359223300970873,
+      "grad_norm": 0.5502263061660566,
+      "learning_rate": 1.461665971981402e-07,
+      "loss": 0.6946,
+      "step": 1376
+    },
+    {
+      "epoch": 13.368932038834952,
+      "grad_norm": 0.5378410979734453,
+      "learning_rate": 1.444587385617785e-07,
+      "loss": 0.4166,
+      "step": 1377
+    },
+    {
+      "epoch": 13.37864077669903,
+      "grad_norm": 0.49693973418779785,
+      "learning_rate": 1.4276061935101586e-07,
+      "loss": 0.6703,
+      "step": 1378
+    },
+    {
+      "epoch": 13.388349514563107,
+      "grad_norm": 0.6231776665301004,
+      "learning_rate": 1.4107224658704288e-07,
+      "loss": 0.5784,
+      "step": 1379
+    },
+    {
+      "epoch": 13.398058252427184,
+      "grad_norm": 0.5586955074187786,
+      "learning_rate": 1.3939362725075344e-07,
+      "loss": 0.3888,
+      "step": 1380
+    },
+    {
+      "epoch": 13.407766990291263,
+      "grad_norm": 0.5217161899840396,
+      "learning_rate": 1.3772476828271236e-07,
+      "loss": 0.5822,
+      "step": 1381
+    },
+    {
+      "epoch": 13.41747572815534,
+      "grad_norm": 0.5153464617820017,
+      "learning_rate": 1.360656765831289e-07,
+      "loss": 0.5918,
+      "step": 1382
+    },
+    {
+      "epoch": 13.427184466019417,
+      "grad_norm": 0.5657762923904565,
+      "learning_rate": 1.3441635901182803e-07,
+      "loss": 0.5361,
+      "step": 1383
+    },
+    {
+      "epoch": 13.436893203883495,
+      "grad_norm": 0.5323531737775129,
+      "learning_rate": 1.3277682238822142e-07,
+      "loss": 0.5163,
+      "step": 1384
+    },
+    {
+      "epoch": 13.446601941747574,
+      "grad_norm": 0.5177243720403791,
+      "learning_rate": 1.3114707349127954e-07,
+      "loss": 0.5323,
+      "step": 1385
+    },
+    {
+      "epoch": 13.45631067961165,
+      "grad_norm": 0.5406186027574795,
+      "learning_rate": 1.2952711905950377e-07,
+      "loss": 0.4687,
+      "step": 1386
+    },
+    {
+      "epoch": 13.466019417475728,
+      "grad_norm": 0.5716720820723694,
+      "learning_rate": 1.279169657908988e-07,
+      "loss": 0.5792,
+      "step": 1387
+    },
+    {
+      "epoch": 13.475728155339805,
+      "grad_norm": 0.4877829559887591,
+      "learning_rate": 1.263166203429439e-07,
+      "loss": 0.3873,
+      "step": 1388
+    },
+    {
+      "epoch": 13.485436893203884,
+      "grad_norm": 0.5546173120250771,
+      "learning_rate": 1.2472608933256637e-07,
+      "loss": 0.3873,
+      "step": 1389
+    },
+    {
+      "epoch": 13.495145631067961,
+      "grad_norm": 0.48230172753514133,
+      "learning_rate": 1.2314537933611425e-07,
+      "loss": 0.4945,
+      "step": 1390
+    },
+    {
+      "epoch": 13.504854368932039,
+      "grad_norm": 0.4574368189573909,
+      "learning_rate": 1.2157449688932872e-07,
+      "loss": 0.4456,
+      "step": 1391
+    },
+    {
+      "epoch": 13.514563106796116,
+      "grad_norm": 0.5242392231721918,
+      "learning_rate": 1.2001344848731612e-07,
+      "loss": 0.4399,
+      "step": 1392
+    },
+    {
+      "epoch": 13.524271844660195,
+      "grad_norm": 0.5942067255673704,
+      "learning_rate": 1.1846224058452316e-07,
+      "loss": 0.3374,
+      "step": 1393
+    },
+    {
+      "epoch": 13.533980582524272,
+      "grad_norm": 0.5808625655096099,
+      "learning_rate": 1.1692087959470882e-07,
+      "loss": 0.523,
+      "step": 1394
+    },
+    {
+      "epoch": 13.54368932038835,
+      "grad_norm": 0.4945556802549258,
+      "learning_rate": 1.1538937189091825e-07,
+      "loss": 0.4227,
+      "step": 1395
+    },
+    {
+      "epoch": 13.553398058252426,
+      "grad_norm": 0.5403255251924856,
+      "learning_rate": 1.1386772380545669e-07,
+      "loss": 0.4746,
+      "step": 1396
+    },
+    {
+      "epoch": 13.563106796116505,
+      "grad_norm": 0.47061852831174805,
+      "learning_rate": 1.1235594162986168e-07,
+      "loss": 0.5258,
+      "step": 1397
+    },
+    {
+      "epoch": 13.572815533980583,
+      "grad_norm": 0.45606330482815727,
+      "learning_rate": 1.1085403161488012e-07,
+      "loss": 0.5838,
+      "step": 1398
+    },
+    {
+      "epoch": 13.58252427184466,
+      "grad_norm": 0.5701345027363025,
+      "learning_rate": 1.09361999970439e-07,
+      "loss": 0.4101,
+      "step": 1399
+    },
+    {
+      "epoch": 13.592233009708737,
+      "grad_norm": 0.5900234582182388,
+      "learning_rate": 1.0787985286562219e-07,
+      "loss": 0.5734,
+      "step": 1400
+    },
+    {
+      "epoch": 13.601941747572816,
+      "grad_norm": 0.5947961921849346,
+      "learning_rate": 1.0640759642864401e-07,
+      "loss": 0.5931,
+      "step": 1401
+    },
+    {
+      "epoch": 13.611650485436893,
+      "grad_norm": 0.5357137024816893,
+      "learning_rate": 1.0494523674682372e-07,
+      "loss": 0.6879,
+      "step": 1402
+    },
+    {
+      "epoch": 13.62135922330097,
+      "grad_norm": 0.5559096380768329,
+      "learning_rate": 1.0349277986656081e-07,
+      "loss": 0.769,
+      "step": 1403
+    },
+    {
+      "epoch": 13.631067961165048,
+      "grad_norm": 0.5771642974798561,
+      "learning_rate": 1.0205023179330975e-07,
+      "loss": 0.2985,
+      "step": 1404
+    },
+    {
+      "epoch": 13.640776699029127,
+      "grad_norm": 0.5304015816283653,
+      "learning_rate": 1.00617598491555e-07,
+      "loss": 0.481,
+      "step": 1405
+    },
+    {
+      "epoch": 13.650485436893204,
+      "grad_norm": 0.5286982826502439,
+      "learning_rate": 9.919488588478715e-08,
+      "loss": 0.3866,
+      "step": 1406
+    },
+    {
+      "epoch": 13.660194174757281,
+      "grad_norm": 0.5494253997430171,
+      "learning_rate": 9.778209985547682e-08,
+      "loss": 0.7586,
+      "step": 1407
+    },
+    {
+      "epoch": 13.669902912621358,
+      "grad_norm": 0.48745009825917535,
+      "learning_rate": 9.637924624505191e-08,
+      "loss": 0.4863,
+      "step": 1408
+    },
+    {
+      "epoch": 13.679611650485437,
+      "grad_norm": 0.6027383636383639,
+      "learning_rate": 9.498633085387343e-08,
+      "loss": 0.4851,
+      "step": 1409
+    },
+    {
+      "epoch": 13.689320388349515,
+      "grad_norm": 0.6028302352156594,
+      "learning_rate": 9.360335944121029e-08,
+      "loss": 0.502,
+      "step": 1410
+    },
+    {
+      "epoch": 13.699029126213592,
+      "grad_norm": 0.5291390026157702,
+      "learning_rate": 9.223033772521594e-08,
+      "loss": 0.563,
+      "step": 1411
+    },
+    {
+      "epoch": 13.70873786407767,
+      "grad_norm": 0.524461539567668,
+      "learning_rate": 9.086727138290535e-08,
+      "loss": 0.579,
+      "step": 1412
+    },
+    {
+      "epoch": 13.718446601941748,
+      "grad_norm": 0.5531924624277184,
+      "learning_rate": 8.951416605013114e-08,
+      "loss": 0.5283,
+      "step": 1413
+    },
+    {
+      "epoch": 13.728155339805825,
+      "grad_norm": 0.498875465737702,
+      "learning_rate": 8.817102732155996e-08,
+      "loss": 0.493,
+      "step": 1414
+    },
+    {
+      "epoch": 13.737864077669903,
+      "grad_norm": 0.5277705069065116,
+      "learning_rate": 8.683786075065065e-08,
+      "loss": 0.5852,
+      "step": 1415
+    },
+    {
+      "epoch": 13.74757281553398,
+      "grad_norm": 0.48811477594662267,
+      "learning_rate": 8.55146718496283e-08,
+      "loss": 0.5653,
+      "step": 1416
+    },
+    {
+      "epoch": 13.757281553398059,
+      "grad_norm": 0.49370146088269673,
+      "learning_rate": 8.420146608946605e-08,
+      "loss": 0.4668,
+      "step": 1417
+    },
+    {
+      "epoch": 13.766990291262136,
+      "grad_norm": 0.5084353639944768,
+      "learning_rate": 8.28982488998581e-08,
+      "loss": 0.5095,
+      "step": 1418
+    },
+    {
+      "epoch": 13.776699029126213,
+      "grad_norm": 0.46703565271667047,
+      "learning_rate": 8.160502566919942e-08,
+      "loss": 0.4907,
+      "step": 1419
+    },
+    {
+      "epoch": 13.78640776699029,
+      "grad_norm": 0.5068684892515162,
+      "learning_rate": 8.032180174456283e-08,
+      "loss": 0.5642,
+      "step": 1420
+    },
+    {
+      "epoch": 13.79611650485437,
+      "grad_norm": 0.5215588251395539,
+      "learning_rate": 7.904858243167806e-08,
+      "loss": 0.3992,
+      "step": 1421
+    },
+    {
+      "epoch": 13.805825242718447,
+      "grad_norm": 0.48149979973118395,
+      "learning_rate": 7.778537299490796e-08,
+      "loss": 0.4448,
+      "step": 1422
+    },
+    {
+      "epoch": 13.815533980582524,
+      "grad_norm": 0.5316505800654157,
+      "learning_rate": 7.653217865722817e-08,
+      "loss": 0.5612,
+      "step": 1423
+    },
+    {
+      "epoch": 13.825242718446601,
+      "grad_norm": 0.5345196593770813,
+      "learning_rate": 7.528900460020444e-08,
+      "loss": 0.4618,
+      "step": 1424
+    },
+    {
+      "epoch": 13.83495145631068,
+      "grad_norm": 0.5536257142626635,
+      "learning_rate": 7.405585596397314e-08,
+      "loss": 0.4619,
+      "step": 1425
+    },
+    {
+      "epoch": 13.844660194174757,
+      "grad_norm": 0.5307437379543056,
+      "learning_rate": 7.283273784721739e-08,
+      "loss": 0.6937,
+      "step": 1426
+    },
+    {
+      "epoch": 13.854368932038835,
+      "grad_norm": 0.4649068565884641,
+      "learning_rate": 7.161965530714743e-08,
+      "loss": 0.4241,
+      "step": 1427
+    },
+    {
+      "epoch": 13.864077669902912,
+      "grad_norm": 0.4921857856166554,
+      "learning_rate": 7.041661335948024e-08,
+      "loss": 0.4411,
+      "step": 1428
+    },
+    {
+      "epoch": 13.87378640776699,
+      "grad_norm": 0.5207340849219146,
+      "learning_rate": 6.92236169784169e-08,
+      "loss": 0.7874,
+      "step": 1429
+    },
+    {
+      "epoch": 13.883495145631068,
+      "grad_norm": 0.49196505164477133,
+      "learning_rate": 6.804067109662443e-08,
+      "loss": 0.5332,
+      "step": 1430
+    },
+    {
+      "epoch": 13.893203883495145,
+      "grad_norm": 0.5075431393301558,
+      "learning_rate": 6.68677806052137e-08,
+      "loss": 0.5008,
+      "step": 1431
+    },
+    {
+      "epoch": 13.902912621359224,
+      "grad_norm": 0.546839111052632,
+      "learning_rate": 6.57049503537191e-08,
+      "loss": 0.5992,
+      "step": 1432
+    },
+    {
+      "epoch": 13.912621359223301,
+      "grad_norm": 0.48530918747896695,
+      "learning_rate": 6.455218515008049e-08,
+      "loss": 0.5208,
+      "step": 1433
+    },
+    {
+      "epoch": 13.922330097087379,
+      "grad_norm": 0.4639108474015221,
+      "learning_rate": 6.340948976062023e-08,
+      "loss": 0.4323,
+      "step": 1434
+    },
+    {
+      "epoch": 13.932038834951456,
+      "grad_norm": 0.4580616182119806,
+      "learning_rate": 6.227686891002671e-08,
+      "loss": 0.4902,
+      "step": 1435
+    },
+    {
+      "epoch": 13.941747572815533,
+      "grad_norm": 0.5092774475295171,
+      "learning_rate": 6.115432728133198e-08,
+      "loss": 0.679,
+      "step": 1436
+    },
+    {
+      "epoch": 13.951456310679612,
+      "grad_norm": 0.5166759750145309,
+      "learning_rate": 6.004186951589414e-08,
+      "loss": 0.5287,
+      "step": 1437
+    },
+    {
+      "epoch": 13.96116504854369,
+      "grad_norm": 0.6029338550408697,
+      "learning_rate": 5.8939500213378296e-08,
+      "loss": 0.3729,
+      "step": 1438
+    },
+    {
+      "epoch": 13.970873786407767,
+      "grad_norm": 0.4820897061950064,
+      "learning_rate": 5.7847223931735974e-08,
+      "loss": 0.4876,
+      "step": 1439
+    },
+    {
+      "epoch": 13.980582524271846,
+      "grad_norm": 0.5218779280339735,
+      "learning_rate": 5.6765045187187614e-08,
+      "loss": 0.5849,
+      "step": 1440
+    },
+    {
+      "epoch": 13.990291262135923,
+      "grad_norm": 0.5215155120681154,
+      "learning_rate": 5.569296845420375e-08,
+      "loss": 0.7779,
+      "step": 1441
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.5656825108193669,
+      "learning_rate": 5.463099816548578e-08,
+      "loss": 0.4209,
+      "step": 1442
+    },
+    {
+      "epoch": 14.009708737864077,
+      "grad_norm": 0.5218440790701329,
+      "learning_rate": 5.3579138711948587e-08,
+      "loss": 0.4079,
+      "step": 1443
+    },
+    {
+      "epoch": 14.019417475728156,
+      "grad_norm": 0.5184483586663422,
+      "learning_rate": 5.253739444270128e-08,
+      "loss": 0.6546,
+      "step": 1444
+    },
+    {
+      "epoch": 14.029126213592233,
+      "grad_norm": 0.5560840017158974,
+      "learning_rate": 5.150576966503063e-08,
+      "loss": 0.5656,
+      "step": 1445
+    },
+    {
+      "epoch": 14.03883495145631,
+      "grad_norm": 0.5145055616244204,
+      "learning_rate": 5.048426864438183e-08,
+      "loss": 0.736,
+      "step": 1446
+    },
+    {
+      "epoch": 14.048543689320388,
+      "grad_norm": 0.5477079295048156,
+      "learning_rate": 4.9472895604341655e-08,
+      "loss": 0.4869,
+      "step": 1447
+    },
+    {
+      "epoch": 14.058252427184467,
+      "grad_norm": 0.4733588081568726,
+      "learning_rate": 4.8471654726621464e-08,
+      "loss": 0.5382,
+      "step": 1448
+    },
+    {
+      "epoch": 14.067961165048544,
+      "grad_norm": 0.5101422793371593,
+      "learning_rate": 4.7480550151038365e-08,
+      "loss": 0.6127,
+      "step": 1449
+    },
+    {
+      "epoch": 14.077669902912621,
+      "grad_norm": 0.5382006222610713,
+      "learning_rate": 4.649958597549964e-08,
+      "loss": 0.5527,
+      "step": 1450
+    },
+    {
+      "epoch": 14.087378640776699,
+      "grad_norm": 0.5275430010182721,
+      "learning_rate": 4.552876625598501e-08,
+      "loss": 0.7032,
+      "step": 1451
+    },
+    {
+      "epoch": 14.097087378640778,
+      "grad_norm": 0.5458051914647247,
+      "learning_rate": 4.4568095006529975e-08,
+      "loss": 0.4345,
+      "step": 1452
+    },
+    {
+      "epoch": 14.106796116504855,
+      "grad_norm": 0.7051966998510896,
+      "learning_rate": 4.361757619920942e-08,
+      "loss": 0.5567,
+      "step": 1453
+    },
+    {
+      "epoch": 14.116504854368932,
+      "grad_norm": 0.49381508246509925,
+      "learning_rate": 4.2677213764120986e-08,
+      "loss": 0.3299,
+      "step": 1454
+    },
+    {
+      "epoch": 14.12621359223301,
+      "grad_norm": 0.5048558481830228,
+      "learning_rate": 4.174701158936895e-08,
+      "loss": 0.451,
+      "step": 1455
+    },
+    {
+      "epoch": 14.135922330097088,
+      "grad_norm": 0.482626507365039,
+      "learning_rate": 4.082697352104814e-08,
+      "loss": 0.533,
+      "step": 1456
+    },
+    {
+      "epoch": 14.145631067961165,
+      "grad_norm": 0.5382580640297336,
+      "learning_rate": 3.991710336322757e-08,
+      "loss": 0.7137,
+      "step": 1457
+    },
+    {
+      "epoch": 14.155339805825243,
+      "grad_norm": 0.536897740958693,
+      "learning_rate": 3.9017404877935986e-08,
+      "loss": 0.5705,
+      "step": 1458
+    },
+    {
+      "epoch": 14.16504854368932,
+      "grad_norm": 0.5211520199572297,
+      "learning_rate": 3.812788178514437e-08,
+      "loss": 0.7333,
+      "step": 1459
+    },
+    {
+      "epoch": 14.174757281553399,
+      "grad_norm": 0.5267984123710193,
+      "learning_rate": 3.7248537762752666e-08,
+      "loss": 0.5171,
+      "step": 1460
+    },
+    {
+      "epoch": 14.184466019417476,
+      "grad_norm": 0.5584448575561238,
+      "learning_rate": 3.637937644657308e-08,
+      "loss": 0.7025,
+      "step": 1461
+    },
+    {
+      "epoch": 14.194174757281553,
+      "grad_norm": 0.5239791697552872,
+      "learning_rate": 3.55204014303151e-08,
+      "loss": 0.8438,
+      "step": 1462
+    },
+    {
+      "epoch": 14.20388349514563,
+      "grad_norm": 0.5245481621288842,
+      "learning_rate": 3.467161626557164e-08,
+      "loss": 0.6344,
+      "step": 1463
+    },
+    {
+      "epoch": 14.21359223300971,
+      "grad_norm": 0.5069515035185672,
+      "learning_rate": 3.3833024461803756e-08,
+      "loss": 0.4982,
+      "step": 1464
+    },
+    {
+      "epoch": 14.223300970873787,
+      "grad_norm": 0.5281513334975837,
+      "learning_rate": 3.300462948632593e-08,
+      "loss": 0.432,
+      "step": 1465
+    },
+    {
+      "epoch": 14.233009708737864,
+      "grad_norm": 0.5471335526184092,
+      "learning_rate": 3.218643476429167e-08,
+      "loss": 0.5776,
+      "step": 1466
+    },
+    {
+      "epoch": 14.242718446601941,
+      "grad_norm": 0.6037098746834603,
+      "learning_rate": 3.1378443678680706e-08,
+      "loss": 0.4294,
+      "step": 1467
+    },
+    {
+      "epoch": 14.25242718446602,
+      "grad_norm": 0.5288263190310578,
+      "learning_rate": 3.0580659570282886e-08,
+      "loss": 0.5654,
+      "step": 1468
+    },
+    {
+      "epoch": 14.262135922330097,
+      "grad_norm": 0.5553795697688602,
+      "learning_rate": 2.979308573768547e-08,
+      "loss": 0.5792,
+      "step": 1469
+    },
+    {
+      "epoch": 14.271844660194175,
+      "grad_norm": 0.5225773956434618,
+      "learning_rate": 2.9015725437259724e-08,
+      "loss": 0.4668,
+      "step": 1470
+    },
+    {
+      "epoch": 14.281553398058252,
+      "grad_norm": 0.5131128030358816,
+      "learning_rate": 2.8248581883147387e-08,
+      "loss": 0.3315,
+      "step": 1471
+    },
+    {
+      "epoch": 14.29126213592233,
+      "grad_norm": 0.45903451471678286,
+      "learning_rate": 2.7491658247246478e-08,
+      "loss": 0.4191,
+      "step": 1472
+    },
+    {
+      "epoch": 14.300970873786408,
+      "grad_norm": 0.4953151621096038,
+      "learning_rate": 2.6744957659199376e-08,
+      "loss": 0.5154,
+      "step": 1473
+    },
+    {
+      "epoch": 14.310679611650485,
+      "grad_norm": 0.519499457859041,
+      "learning_rate": 2.6008483206379497e-08,
+      "loss": 0.5541,
+      "step": 1474
+    },
+    {
+      "epoch": 14.320388349514563,
+      "grad_norm": 0.5083323776230902,
+      "learning_rate": 2.5282237933877962e-08,
+      "loss": 0.4306,
+      "step": 1475
+    },
+    {
+      "epoch": 14.330097087378642,
+      "grad_norm": 0.5335621363530555,
+      "learning_rate": 2.4566224844491393e-08,
+      "loss": 0.5205,
+      "step": 1476
+    },
+    {
+      "epoch": 14.339805825242719,
+      "grad_norm": 0.501971009726888,
+      "learning_rate": 2.38604468987097e-08,
+      "loss": 0.6609,
+      "step": 1477
+    },
+    {
+      "epoch": 14.349514563106796,
+      "grad_norm": 0.5261429789373724,
+      "learning_rate": 2.316490701470414e-08,
+      "loss": 0.5117,
+      "step": 1478
+    },
+    {
+      "epoch": 14.359223300970873,
+      "grad_norm": 0.5293420600011806,
+      "learning_rate": 2.247960806831373e-08,
+      "loss": 0.5102,
+      "step": 1479
+    },
+    {
+      "epoch": 14.368932038834952,
+      "grad_norm": 0.5026358641196522,
+      "learning_rate": 2.180455289303579e-08,
+      "loss": 0.4931,
+      "step": 1480
+    },
+    {
+      "epoch": 14.37864077669903,
+      "grad_norm": 0.5133562751158053,
+      "learning_rate": 2.113974428001153e-08,
+      "loss": 0.433,
+      "step": 1481
+    },
+    {
+      "epoch": 14.388349514563107,
+      "grad_norm": 0.5265411531668414,
+      "learning_rate": 2.0485184978016604e-08,
+      "loss": 0.4764,
+      "step": 1482
+    },
+    {
+      "epoch": 14.398058252427184,
+      "grad_norm": 0.5367964213460488,
+      "learning_rate": 1.984087769344889e-08,
+      "loss": 0.5355,
+      "step": 1483
+    },
+    {
+      "epoch": 14.407766990291263,
+      "grad_norm": 0.5217857607014399,
+      "learning_rate": 1.9206825090317126e-08,
+      "loss": 0.4859,
+      "step": 1484
+    },
+    {
+      "epoch": 14.41747572815534,
+      "grad_norm": 0.5169259518658883,
+      "learning_rate": 1.8583029790230356e-08,
+      "loss": 0.5476,
+      "step": 1485
+    },
+    {
+      "epoch": 14.427184466019417,
+      "grad_norm": 0.4904849259610178,
+      "learning_rate": 1.796949437238682e-08,
+      "loss": 0.6665,
+      "step": 1486
+    },
+    {
+      "epoch": 14.436893203883495,
+      "grad_norm": 0.49108376586122854,
+      "learning_rate": 1.736622137356342e-08,
+      "loss": 0.551,
+      "step": 1487
+    },
+    {
+      "epoch": 14.446601941747574,
+      "grad_norm": 0.5142950491406901,
+      "learning_rate": 1.677321328810516e-08,
+      "loss": 0.3045,
+      "step": 1488
+    },
+    {
+      "epoch": 14.45631067961165,
+      "grad_norm": 0.503112558064198,
+      "learning_rate": 1.6190472567914617e-08,
+      "loss": 0.5419,
+      "step": 1489
+    },
+    {
+      "epoch": 14.466019417475728,
+      "grad_norm": 0.4622893048563594,
+      "learning_rate": 1.561800162244248e-08,
+      "loss": 0.338,
+      "step": 1490
+    },
+    {
+      "epoch": 14.475728155339805,
+      "grad_norm": 0.4830240169733427,
+      "learning_rate": 1.5055802818676745e-08,
+      "loss": 0.5756,
+      "step": 1491
+    },
+    {
+      "epoch": 14.485436893203884,
+      "grad_norm": 0.5659073116377321,
+      "learning_rate": 1.450387848113327e-08,
+      "loss": 0.4251,
+      "step": 1492
+    },
+    {
+      "epoch": 14.495145631067961,
+      "grad_norm": 0.5067553896210469,
+      "learning_rate": 1.3962230891846618e-08,
+      "loss": 0.5445,
+      "step": 1493
+    },
+    {
+      "epoch": 14.504854368932039,
+      "grad_norm": 0.5077181092700885,
+      "learning_rate": 1.3430862290359781e-08,
+      "loss": 0.6045,
+      "step": 1494
+    },
+    {
+      "epoch": 14.514563106796116,
+      "grad_norm": 0.5136305065251573,
+      "learning_rate": 1.2909774873715585e-08,
+      "loss": 0.6548,
+      "step": 1495
+    },
+    {
+      "epoch": 14.524271844660195,
+      "grad_norm": 0.49226391309275364,
+      "learning_rate": 1.2398970796447807e-08,
+      "loss": 0.5737,
+      "step": 1496
+    },
+    {
+      "epoch": 14.533980582524272,
+      "grad_norm": 0.5238814510404914,
+      "learning_rate": 1.1898452170570618e-08,
+      "loss": 0.4521,
+      "step": 1497
+    },
+    {
+      "epoch": 14.54368932038835,
+      "grad_norm": 0.5053130339074384,
+      "learning_rate": 1.140822106557249e-08,
+      "loss": 0.3477,
+      "step": 1498
+    },
+    {
+      "epoch": 14.553398058252426,
+      "grad_norm": 0.5913005570564529,
+      "learning_rate": 1.0928279508405082e-08,
+      "loss": 0.5268,
+      "step": 1499
+    },
+    {
+      "epoch": 14.563106796116505,
+      "grad_norm": 0.6238009868766736,
+      "learning_rate": 1.0458629483476868e-08,
+      "loss": 0.4591,
+      "step": 1500
+    },
+    {
+      "epoch": 14.572815533980583,
+      "grad_norm": 0.5980143147512443,
+      "learning_rate": 9.999272932643134e-09,
+      "loss": 0.6164,
+      "step": 1501
+    },
+    {
+      "epoch": 14.58252427184466,
+      "grad_norm": 0.5003040722536184,
+      "learning_rate": 9.550211755199879e-09,
+      "loss": 0.5502,
+      "step": 1502
+    },
+    {
+      "epoch": 14.592233009708737,
+      "grad_norm": 0.5932523805715173,
+      "learning_rate": 9.111447807874374e-09,
+      "loss": 0.5563,
+      "step": 1503
+    },
+    {
+      "epoch": 14.601941747572816,
+      "grad_norm": 0.6369284594646255,
+      "learning_rate": 8.682982904817948e-09,
+      "loss": 0.3706,
+      "step": 1504
+    },
+    {
+      "epoch": 14.611650485436893,
+      "grad_norm": 0.4671794754568133,
+      "learning_rate": 8.264818817599052e-09,
+      "loss": 0.6647,
+      "step": 1505
+    },
+    {
+      "epoch": 14.62135922330097,
+      "grad_norm": 0.45435838697605796,
+      "learning_rate": 7.856957275194921e-09,
+      "loss": 0.447,
+      "step": 1506
+    },
+    {
+      "epoch": 14.631067961165048,
+      "grad_norm": 0.49035303412272524,
+      "learning_rate": 7.459399963985758e-09,
+      "loss": 0.4693,
+      "step": 1507
+    },
+    {
+      "epoch": 14.640776699029127,
+      "grad_norm": 0.5547141125879742,
+      "learning_rate": 7.072148527746403e-09,
+      "loss": 0.7602,
+      "step": 1508
+    },
+    {
+      "epoch": 14.650485436893204,
+      "grad_norm": 0.586251716363603,
+      "learning_rate": 6.6952045676405005e-09,
+      "loss": 0.5752,
+      "step": 1509
+    },
+    {
+      "epoch": 14.660194174757281,
+      "grad_norm": 0.5166126228384136,
+      "learning_rate": 6.328569642212734e-09,
+      "loss": 0.5488,
+      "step": 1510
+    },
+    {
+      "epoch": 14.669902912621358,
+      "grad_norm": 0.489603168687867,
+      "learning_rate": 5.972245267384102e-09,
+      "loss": 0.586,
+      "step": 1511
+    },
+    {
+      "epoch": 14.679611650485437,
+      "grad_norm": 0.5365602268760269,
+      "learning_rate": 5.62623291644443e-09,
+      "loss": 0.6064,
+      "step": 1512
+    },
+    {
+      "epoch": 14.689320388349515,
+      "grad_norm": 0.4765356527868554,
+      "learning_rate": 5.290534020046256e-09,
+      "loss": 0.256,
+      "step": 1513
+    },
+    {
+      "epoch": 14.699029126213592,
+      "grad_norm": 0.5282645667952264,
+      "learning_rate": 4.965149966199567e-09,
+      "loss": 0.524,
+      "step": 1514
+    },
+    {
+      "epoch": 14.70873786407767,
+      "grad_norm": 0.5345977975069597,
+      "learning_rate": 4.6500821002654075e-09,
+      "loss": 0.4932,
+      "step": 1515
+    },
+    {
+      "epoch": 14.718446601941748,
+      "grad_norm": 0.5012155423781401,
+      "learning_rate": 4.345331724950885e-09,
+      "loss": 0.5275,
+      "step": 1516
+    },
+    {
+      "epoch": 14.728155339805825,
+      "grad_norm": 0.4761432206250365,
+      "learning_rate": 4.050900100303068e-09,
+      "loss": 0.5482,
+      "step": 1517
+    },
+    {
+      "epoch": 14.737864077669903,
+      "grad_norm": 0.4722854556901931,
+      "learning_rate": 3.766788443705094e-09,
+      "loss": 0.486,
+      "step": 1518
+    },
+    {
+      "epoch": 14.74757281553398,
+      "grad_norm": 0.48685808472088393,
+      "learning_rate": 3.492997929869235e-09,
+      "loss": 0.4182,
+      "step": 1519
+    },
+    {
+      "epoch": 14.757281553398059,
+      "grad_norm": 0.5368190413406362,
+      "learning_rate": 3.2295296908338437e-09,
+      "loss": 0.4503,
+      "step": 1520
+    },
+    {
+      "epoch": 14.766990291262136,
+      "grad_norm": 0.4718713824984552,
+      "learning_rate": 2.976384815957245e-09,
+      "loss": 0.4848,
+      "step": 1521
+    },
+    {
+      "epoch": 14.776699029126213,
+      "grad_norm": 0.5644203137176583,
+      "learning_rate": 2.7335643519144086e-09,
+      "loss": 0.6387,
+      "step": 1522
+    },
+    {
+      "epoch": 14.78640776699029,
+      "grad_norm": 0.4362591665856203,
+      "learning_rate": 2.5010693026922273e-09,
+      "loss": 0.4436,
+      "step": 1523
+    },
+    {
+      "epoch": 14.79611650485437,
+      "grad_norm": 0.5404439955190101,
+      "learning_rate": 2.278900629584524e-09,
+      "loss": 0.3487,
+      "step": 1524
+    },
+    {
+      "epoch": 14.805825242718447,
+      "grad_norm": 0.5112825487443058,
+      "learning_rate": 2.067059251189274e-09,
+      "loss": 0.6591,
+      "step": 1525
+    },
+    {
+      "epoch": 14.815533980582524,
+      "grad_norm": 0.5176623109055024,
+      "learning_rate": 1.8655460434044427e-09,
+      "loss": 0.3372,
+      "step": 1526
+    },
+    {
+      "epoch": 14.825242718446601,
+      "grad_norm": 0.5412558465607696,
+      "learning_rate": 1.6743618394238215e-09,
+      "loss": 0.7489,
+      "step": 1527
+    },
+    {
+      "epoch": 14.83495145631068,
+      "grad_norm": 0.5609397261408766,
+      "learning_rate": 1.493507429734531e-09,
+      "loss": 0.555,
+      "step": 1528
+    },
+    {
+      "epoch": 14.844660194174757,
+      "grad_norm": 0.5233104588853983,
+      "learning_rate": 1.3229835621125786e-09,
+      "loss": 0.4006,
+      "step": 1529
+    },
+    {
+      "epoch": 14.854368932038835,
+      "grad_norm": 0.49863211232575066,
+      "learning_rate": 1.1627909416211947e-09,
+      "loss": 0.3629,
+      "step": 1530
+    },
+    {
+      "epoch": 14.864077669902912,
+      "grad_norm": 0.5077470970611437,
+      "learning_rate": 1.0129302306061128e-09,
+      "loss": 0.4257,
+      "step": 1531
+    },
+    {
+      "epoch": 14.87378640776699,
+      "grad_norm": 0.49886786092314855,
+      "learning_rate": 8.734020486950157e-10,
+      "loss": 0.5751,
+      "step": 1532
+    },
+    {
+      "epoch": 14.883495145631068,
+      "grad_norm": 0.5398797190248956,
+      "learning_rate": 7.442069727930934e-10,
+      "loss": 0.2788,
+      "step": 1533
+    },
+    {
+      "epoch": 14.893203883495145,
+      "grad_norm": 0.5452560970159284,
+      "learning_rate": 6.253455370811012e-10,
+      "loss": 0.5946,
+      "step": 1534
+    },
+    {
+      "epoch": 14.902912621359224,
+      "grad_norm": 0.5511414946543856,
+      "learning_rate": 5.168182330145266e-10,
+      "loss": 0.4084,
+      "step": 1535
+    },
+    {
+      "epoch": 14.912621359223301,
+      "grad_norm": 0.5429125042489947,
+      "learning_rate": 4.186255093194258e-10,
+      "loss": 0.4831,
+      "step": 1536
+    },
+    {
+      "epoch": 14.922330097087379,
+      "grad_norm": 0.5720561401215973,
+      "learning_rate": 3.3076777199186894e-10,
+      "loss": 0.4999,
+      "step": 1537
+    },
+    {
+      "epoch": 14.932038834951456,
+      "grad_norm": 0.5021834442307801,
+      "learning_rate": 2.532453842965521e-10,
+      "loss": 0.366,
+      "step": 1538
+    },
+    {
+      "epoch": 14.941747572815533,
+      "grad_norm": 0.5277645181797495,
+      "learning_rate": 1.8605866676374428e-10,
+      "loss": 0.419,
+      "step": 1539
+    },
+    {
+      "epoch": 14.951456310679612,
+      "grad_norm": 0.5207502460141687,
+      "learning_rate": 1.292078971898425e-10,
+      "loss": 0.427,
+      "step": 1540
+    },
+    {
+      "epoch": 14.96116504854369,
+      "grad_norm": 0.5656367527963858,
+      "learning_rate": 8.269331063459618e-11,
+      "loss": 0.2437,
+      "step": 1541
+    },
+    {
+      "epoch": 14.970873786407767,
+      "grad_norm": 0.5623100607316655,
+      "learning_rate": 4.651509942193988e-11,
+      "loss": 0.5833,
+      "step": 1542
+    },
+    {
+      "epoch": 14.980582524271846,
+      "grad_norm": 0.5028687329515733,
+      "learning_rate": 2.06734131366626e-11,
+      "loss": 0.6296,
+      "step": 1543
+    },
+    {
+      "epoch": 14.990291262135923,
+      "grad_norm": 0.5297985306979774,
+      "learning_rate": 5.168358626628234e-12,
+      "loss": 0.4812,
+      "step": 1544
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.5181687092759953,
+      "learning_rate": 0.0,
+      "loss": 0.4149,
+      "step": 1545
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 74526993309696.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}