diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18529 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 52890,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 103.3125,
+      "learning_rate": 9.996224236844906e-07,
+      "loss": 5.1244,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 133.25,
+      "learning_rate": 9.99244280153685e-07,
+      "loss": 4.3099,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 74.375,
+      "learning_rate": 9.988661366228795e-07,
+      "loss": 3.1455,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 83.875,
+      "learning_rate": 9.98487993092074e-07,
+      "loss": 3.0035,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 106.0625,
+      "learning_rate": 9.981098495612682e-07,
+      "loss": 2.9661,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 55.28125,
+      "learning_rate": 9.977317060304627e-07,
+      "loss": 2.634,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 69.0,
+      "learning_rate": 9.973535624996572e-07,
+      "loss": 2.601,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 64.6875,
+      "learning_rate": 9.969754189688517e-07,
+      "loss": 2.5738,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 99.0,
+      "learning_rate": 9.965972754380461e-07,
+      "loss": 2.6165,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 62.125,
+      "learning_rate": 9.962191319072406e-07,
+      "loss": 2.5547,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 60.625,
+      "learning_rate": 9.95840988376435e-07,
+      "loss": 2.4761,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 59.625,
+      "learning_rate": 9.954628448456296e-07,
+      "loss": 2.4663,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 75.75,
+      "learning_rate": 9.950847013148238e-07,
+      "loss": 2.5917,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 96.0,
+      "learning_rate": 9.947065577840183e-07,
+      "loss": 2.4079,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 85.25,
+      "learning_rate": 9.943284142532128e-07,
+      "loss": 2.4535,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 48.5625,
+      "learning_rate": 9.939502707224072e-07,
+      "loss": 2.4386,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 53.46875,
+      "learning_rate": 9.935721271916017e-07,
+      "loss": 2.4755,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 69.5,
+      "learning_rate": 9.931939836607962e-07,
+      "loss": 2.3306,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 45.09375,
+      "learning_rate": 9.928158401299905e-07,
+      "loss": 2.3735,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 60.90625,
+      "learning_rate": 9.92437696599185e-07,
+      "loss": 2.3466,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 55.34375,
+      "learning_rate": 9.920595530683794e-07,
+      "loss": 2.3971,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 49.375,
+      "learning_rate": 9.916814095375739e-07,
+      "loss": 2.366,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 49.15625,
+      "learning_rate": 9.913032660067684e-07,
+      "loss": 2.3549,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 66.75,
+      "learning_rate": 9.909251224759626e-07,
+      "loss": 2.23,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 66.8125,
+      "learning_rate": 9.905469789451573e-07,
+      "loss": 2.4057,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 60.375,
+      "learning_rate": 9.901688354143518e-07,
+      "loss": 2.339,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 56.65625,
+      "learning_rate": 9.89790691883546e-07,
+      "loss": 2.3804,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 78.375,
+      "learning_rate": 9.894125483527405e-07,
+      "loss": 2.2699,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 65.4375,
+      "learning_rate": 9.89034404821935e-07,
+      "loss": 2.4262,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 55.96875,
+      "learning_rate": 9.886562612911295e-07,
+      "loss": 2.3386,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 71.0625,
+      "learning_rate": 9.88278117760324e-07,
+      "loss": 2.3846,
+      "step": 620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 61.75,
+      "learning_rate": 9.878999742295184e-07,
+      "loss": 2.2632,
+      "step": 640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 47.59375,
+      "learning_rate": 9.875218306987127e-07,
+      "loss": 2.2826,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 68.75,
+      "learning_rate": 9.871436871679071e-07,
+      "loss": 2.3458,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 51.0625,
+      "learning_rate": 9.867655436371016e-07,
+      "loss": 2.2185,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 63.71875,
+      "learning_rate": 9.86387400106296e-07,
+      "loss": 2.2608,
+      "step": 720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 78.375,
+      "learning_rate": 9.860092565754906e-07,
+      "loss": 2.3038,
+      "step": 740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 53.25,
+      "learning_rate": 9.856311130446848e-07,
+      "loss": 2.2971,
+      "step": 760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 61.84375,
+      "learning_rate": 9.852529695138793e-07,
+      "loss": 2.1458,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 58.78125,
+      "learning_rate": 9.848748259830738e-07,
+      "loss": 2.2228,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 48.90625,
+      "learning_rate": 9.844966824522683e-07,
+      "loss": 2.3179,
+      "step": 820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 70.625,
+      "learning_rate": 9.841185389214627e-07,
+      "loss": 2.1868,
+      "step": 840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 64.125,
+      "learning_rate": 9.837403953906572e-07,
+      "loss": 2.2075,
+      "step": 860
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 68.0,
+      "learning_rate": 9.833622518598517e-07,
+      "loss": 2.2674,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 70.3125,
+      "learning_rate": 9.829841083290462e-07,
+      "loss": 2.3387,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 58.5,
+      "learning_rate": 9.826059647982406e-07,
+      "loss": 2.2466,
+      "step": 920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 110.9375,
+      "learning_rate": 9.822278212674349e-07,
+      "loss": 2.1914,
+      "step": 940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 60.96875,
+      "learning_rate": 9.818496777366294e-07,
+      "loss": 2.1054,
+      "step": 960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 73.1875,
+      "learning_rate": 9.814715342058238e-07,
+      "loss": 2.2398,
+      "step": 980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 51.5,
+      "learning_rate": 9.810933906750183e-07,
+      "loss": 2.2347,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 79.125,
+      "learning_rate": 9.807152471442128e-07,
+      "loss": 2.2701,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 63.40625,
+      "learning_rate": 9.803371036134073e-07,
+      "loss": 2.3747,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 71.75,
+      "learning_rate": 9.799589600826015e-07,
+      "loss": 2.3053,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 84.0625,
+      "learning_rate": 9.79580816551796e-07,
+      "loss": 2.2083,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 58.5625,
+      "learning_rate": 9.792026730209905e-07,
+      "loss": 2.301,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 61.59375,
+      "learning_rate": 9.78824529490185e-07,
+      "loss": 2.1949,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 53.0625,
+      "learning_rate": 9.784463859593794e-07,
+      "loss": 2.1185,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 63.4375,
+      "learning_rate": 9.780682424285737e-07,
+      "loss": 2.2003,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 72.4375,
+      "learning_rate": 9.776900988977682e-07,
+      "loss": 2.1994,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 74.375,
+      "learning_rate": 9.773119553669628e-07,
+      "loss": 2.1708,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 47.75,
+      "learning_rate": 9.76933811836157e-07,
+      "loss": 2.1373,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 88.375,
+      "learning_rate": 9.765556683053516e-07,
+      "loss": 2.1869,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 162.75,
+      "learning_rate": 9.76177524774546e-07,
+      "loss": 2.3998,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 77.875,
+      "learning_rate": 9.757993812437405e-07,
+      "loss": 2.1531,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 80.75,
+      "learning_rate": 9.75421237712935e-07,
+      "loss": 2.1803,
+      "step": 1300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 78.25,
+      "learning_rate": 9.750430941821295e-07,
+      "loss": 2.2272,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 62.40625,
+      "learning_rate": 9.746649506513237e-07,
+      "loss": 2.3047,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 92.8125,
+      "learning_rate": 9.742868071205182e-07,
+      "loss": 2.2091,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 77.0625,
+      "learning_rate": 9.739086635897127e-07,
+      "loss": 2.3365,
+      "step": 1380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 67.0,
+      "learning_rate": 9.735305200589072e-07,
+      "loss": 2.1193,
+      "step": 1400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 67.4375,
+      "learning_rate": 9.731523765281016e-07,
+      "loss": 2.1237,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 51.84375,
+      "learning_rate": 9.72774232997296e-07,
+      "loss": 2.1907,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 73.375,
+      "learning_rate": 9.723960894664904e-07,
+      "loss": 2.2086,
+      "step": 1460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 56.625,
+      "learning_rate": 9.720179459356848e-07,
+      "loss": 2.232,
+      "step": 1480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 67.875,
+      "learning_rate": 9.716398024048793e-07,
+      "loss": 2.236,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 85.6875,
+      "learning_rate": 9.712616588740738e-07,
+      "loss": 2.2241,
+      "step": 1520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 64.5,
+      "learning_rate": 9.708835153432683e-07,
+      "loss": 2.1892,
+      "step": 1540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 56.46875,
+      "learning_rate": 9.705053718124627e-07,
+      "loss": 2.2541,
+      "step": 1560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 78.125,
+      "learning_rate": 9.701272282816572e-07,
+      "loss": 2.1645,
+      "step": 1580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 66.5625,
+      "learning_rate": 9.697490847508517e-07,
+      "loss": 2.1083,
+      "step": 1600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 70.375,
+      "learning_rate": 9.69370941220046e-07,
+      "loss": 2.1757,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 55.9375,
+      "learning_rate": 9.689927976892404e-07,
+      "loss": 2.1102,
+      "step": 1640
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 106.4375,
+      "learning_rate": 9.68614654158435e-07,
+      "loss": 2.1047,
+      "step": 1660
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 60.59375,
+      "learning_rate": 9.682365106276294e-07,
+      "loss": 2.2789,
+      "step": 1680
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 63.8125,
+      "learning_rate": 9.678583670968239e-07,
+      "loss": 2.1161,
+      "step": 1700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 80.4375,
+      "learning_rate": 9.674802235660181e-07,
+      "loss": 2.2273,
+      "step": 1720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 58.28125,
+      "learning_rate": 9.671020800352126e-07,
+      "loss": 2.1584,
+      "step": 1740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 60.59375,
+      "learning_rate": 9.66723936504407e-07,
+      "loss": 2.159,
+      "step": 1760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 77.25,
+      "learning_rate": 9.663457929736015e-07,
+      "loss": 2.1554,
+      "step": 1780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 62.125,
+      "learning_rate": 9.65967649442796e-07,
+      "loss": 2.3024,
+      "step": 1800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 47.9375,
+      "learning_rate": 9.655895059119905e-07,
+      "loss": 2.1416,
+      "step": 1820
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 78.875,
+      "learning_rate": 9.652113623811847e-07,
+      "loss": 2.2613,
+      "step": 1840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 71.5,
+      "learning_rate": 9.648332188503792e-07,
+      "loss": 2.1109,
+      "step": 1860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 42.03125,
+      "learning_rate": 9.644550753195737e-07,
+      "loss": 2.0072,
+      "step": 1880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 47.875,
+      "learning_rate": 9.640769317887682e-07,
+      "loss": 2.1403,
+      "step": 1900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 70.25,
+      "learning_rate": 9.636987882579626e-07,
+      "loss": 2.1649,
+      "step": 1920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 52.15625,
+      "learning_rate": 9.633206447271571e-07,
+      "loss": 2.0728,
+      "step": 1940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 65.625,
+      "learning_rate": 9.629425011963516e-07,
+      "loss": 2.2163,
+      "step": 1960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 81.125,
+      "learning_rate": 9.62564357665546e-07,
+      "loss": 2.3386,
+      "step": 1980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 65.0,
+      "learning_rate": 9.621862141347405e-07,
+      "loss": 2.1122,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 62.0,
+      "learning_rate": 9.618080706039348e-07,
+      "loss": 2.201,
+      "step": 2020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 76.9375,
+      "learning_rate": 9.614299270731293e-07,
+      "loss": 2.2634,
+      "step": 2040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 80.9375,
+      "learning_rate": 9.610517835423238e-07,
+      "loss": 2.1005,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 74.5,
+      "learning_rate": 9.606736400115182e-07,
+      "loss": 2.1028,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 76.25,
+      "learning_rate": 9.602954964807127e-07,
+      "loss": 2.2869,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 59.3125,
+      "learning_rate": 9.59917352949907e-07,
+      "loss": 2.1469,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 64.8125,
+      "learning_rate": 9.595392094191014e-07,
+      "loss": 2.084,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 73.125,
+      "learning_rate": 9.59161065888296e-07,
+      "loss": 2.1673,
+      "step": 2160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 161.0,
+      "learning_rate": 9.587829223574904e-07,
+      "loss": 2.0762,
+      "step": 2180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 62.0,
+      "learning_rate": 9.584047788266849e-07,
+      "loss": 2.0542,
+      "step": 2200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 101.0,
+      "learning_rate": 9.580266352958793e-07,
+      "loss": 2.0866,
+      "step": 2220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 66.0,
+      "learning_rate": 9.576484917650736e-07,
+      "loss": 2.1058,
+      "step": 2240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 51.21875,
+      "learning_rate": 9.572703482342683e-07,
+      "loss": 2.18,
+      "step": 2260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 131.125,
+      "learning_rate": 9.568922047034628e-07,
+      "loss": 2.0921,
+      "step": 2280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 68.8125,
+      "learning_rate": 9.56514061172657e-07,
+      "loss": 2.1603,
+      "step": 2300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 106.375,
+      "learning_rate": 9.561359176418515e-07,
+      "loss": 2.1403,
+      "step": 2320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 68.1875,
+      "learning_rate": 9.55757774111046e-07,
+      "loss": 2.2704,
+      "step": 2340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 83.6875,
+      "learning_rate": 9.553796305802404e-07,
+      "loss": 2.2066,
+      "step": 2360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 75.9375,
+      "learning_rate": 9.55001487049435e-07,
+      "loss": 2.1317,
+      "step": 2380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 84.875,
+      "learning_rate": 9.546233435186292e-07,
+      "loss": 2.1652,
+      "step": 2400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 58.65625,
+      "learning_rate": 9.542451999878237e-07,
+      "loss": 2.1021,
+      "step": 2420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 125.0,
+      "learning_rate": 9.538670564570181e-07,
+      "loss": 2.1924,
+      "step": 2440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 56.96875,
+      "learning_rate": 9.534889129262126e-07,
+      "loss": 2.1115,
+      "step": 2460
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 56.03125,
+      "learning_rate": 9.53110769395407e-07,
+      "loss": 2.0641,
+      "step": 2480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 83.625,
+      "learning_rate": 9.527326258646014e-07,
+      "loss": 2.0102,
+      "step": 2500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 66.125,
+      "learning_rate": 9.523544823337959e-07,
+      "loss": 2.0708,
+      "step": 2520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 75.75,
+      "learning_rate": 9.519763388029903e-07,
+      "loss": 2.0818,
+      "step": 2540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 128.0,
+      "learning_rate": 9.515981952721848e-07,
+      "loss": 2.1258,
+      "step": 2560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 75.8125,
+      "learning_rate": 9.512200517413792e-07,
+      "loss": 2.1426,
+      "step": 2580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 63.875,
+      "learning_rate": 9.508419082105737e-07,
+      "loss": 2.1111,
+      "step": 2600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 58.75,
+      "learning_rate": 9.504637646797682e-07,
+      "loss": 2.0416,
+      "step": 2620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 55.0625,
+      "learning_rate": 9.500856211489627e-07,
+      "loss": 2.0279,
+      "step": 2640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 56.96875,
+      "learning_rate": 9.49707477618157e-07,
+      "loss": 2.0627,
+      "step": 2660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 55.75,
+      "learning_rate": 9.493293340873515e-07,
+      "loss": 2.0918,
+      "step": 2680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 62.15625,
+      "learning_rate": 9.48951190556546e-07,
+      "loss": 2.1226,
+      "step": 2700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 75.75,
+      "learning_rate": 9.485730470257403e-07,
+      "loss": 2.0557,
+      "step": 2720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 82.25,
+      "learning_rate": 9.481949034949348e-07,
+      "loss": 2.1832,
+      "step": 2740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 54.5625,
+      "learning_rate": 9.478167599641293e-07,
+      "loss": 2.0379,
+      "step": 2760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 54.21875,
+      "learning_rate": 9.474386164333237e-07,
+      "loss": 2.0732,
+      "step": 2780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 61.03125,
+      "learning_rate": 9.470604729025181e-07,
+      "loss": 2.1374,
+      "step": 2800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 94.0625,
+      "learning_rate": 9.466823293717125e-07,
+      "loss": 2.0421,
+      "step": 2820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 55.40625,
+      "learning_rate": 9.46304185840907e-07,
+      "loss": 2.0783,
+      "step": 2840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 117.6875,
+      "learning_rate": 9.459260423101015e-07,
+      "loss": 2.1041,
+      "step": 2860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 70.875,
+      "learning_rate": 9.455478987792958e-07,
+      "loss": 2.1819,
+      "step": 2880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 90.3125,
+      "learning_rate": 9.451697552484903e-07,
+      "loss": 2.1047,
+      "step": 2900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 51.5625,
+      "learning_rate": 9.447916117176848e-07,
+      "loss": 2.1001,
+      "step": 2920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 77.6875,
+      "learning_rate": 9.444134681868792e-07,
+      "loss": 2.0817,
+      "step": 2940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 101.1875,
+      "learning_rate": 9.440353246560737e-07,
+      "loss": 2.1162,
+      "step": 2960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 83.375,
+      "learning_rate": 9.436571811252682e-07,
+      "loss": 2.1422,
+      "step": 2980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 89.8125,
+      "learning_rate": 9.432790375944626e-07,
+      "loss": 2.1404,
+      "step": 3000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 50.78125,
+      "learning_rate": 9.42900894063657e-07,
+      "loss": 2.0371,
+      "step": 3020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 57.25,
+      "learning_rate": 9.425227505328515e-07,
+      "loss": 2.0327,
+      "step": 3040
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 112.375,
+      "learning_rate": 9.421446070020459e-07,
+      "loss": 2.1609,
+      "step": 3060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 76.3125,
+      "learning_rate": 9.417664634712404e-07,
+      "loss": 2.1167,
+      "step": 3080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 63.6875,
+      "learning_rate": 9.413883199404348e-07,
+      "loss": 2.023,
+      "step": 3100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 46.6875,
+      "learning_rate": 9.410101764096292e-07,
+      "loss": 2.0556,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 84.9375,
+      "learning_rate": 9.406320328788237e-07,
+      "loss": 2.0876,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 107.5,
+      "learning_rate": 9.40253889348018e-07,
+      "loss": 2.087,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 63.4375,
+      "learning_rate": 9.398757458172125e-07,
+      "loss": 2.1845,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 52.9375,
+      "learning_rate": 9.39497602286407e-07,
+      "loss": 2.1219,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 72.1875,
+      "learning_rate": 9.391194587556014e-07,
+      "loss": 2.0844,
+      "step": 3220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 96.875,
+      "learning_rate": 9.387413152247958e-07,
+      "loss": 2.1343,
+      "step": 3240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 64.375,
+      "learning_rate": 9.383631716939903e-07,
+      "loss": 2.0837,
+      "step": 3260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 82.0625,
+      "learning_rate": 9.379850281631847e-07,
+      "loss": 2.1041,
+      "step": 3280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 70.3125,
+      "learning_rate": 9.376068846323792e-07,
+      "loss": 2.1016,
+      "step": 3300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 70.125,
+      "learning_rate": 9.372287411015737e-07,
+      "loss": 2.1038,
+      "step": 3320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 73.6875,
+      "learning_rate": 9.368505975707681e-07,
+      "loss": 2.1319,
+      "step": 3340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 79.4375,
+      "learning_rate": 9.364724540399626e-07,
+      "loss": 1.9609,
+      "step": 3360
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 42.53125,
+      "learning_rate": 9.36094310509157e-07,
+      "loss": 2.1016,
+      "step": 3380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 112.5,
+      "learning_rate": 9.357161669783514e-07,
+      "loss": 2.1199,
+      "step": 3400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 65.625,
+      "learning_rate": 9.353380234475459e-07,
+      "loss": 2.0902,
+      "step": 3420
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 138.875,
+      "learning_rate": 9.349598799167403e-07,
+      "loss": 2.0669,
+      "step": 3440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 83.25,
+      "learning_rate": 9.345817363859347e-07,
+      "loss": 2.1091,
+      "step": 3460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 115.3125,
+      "learning_rate": 9.342035928551292e-07,
+      "loss": 2.0353,
+      "step": 3480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 62.71875,
+      "learning_rate": 9.338254493243236e-07,
+      "loss": 2.0954,
+      "step": 3500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 79.1875,
+      "learning_rate": 9.33447305793518e-07,
+      "loss": 2.096,
+      "step": 3520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 158.0,
+      "learning_rate": 9.330691622627125e-07,
+      "loss": 2.1737,
+      "step": 3540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 57.625,
+      "learning_rate": 9.326910187319069e-07,
+      "loss": 2.1029,
+      "step": 3560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 86.3125,
+      "learning_rate": 9.323128752011014e-07,
+      "loss": 2.1356,
+      "step": 3580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 93.3125,
+      "learning_rate": 9.319347316702958e-07,
+      "loss": 2.1022,
+      "step": 3600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 70.0625,
+      "learning_rate": 9.315565881394902e-07,
+      "loss": 2.0402,
+      "step": 3620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 68.6875,
+      "learning_rate": 9.311784446086848e-07,
+      "loss": 2.1377,
+      "step": 3640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 52.15625,
+      "learning_rate": 9.308003010778793e-07,
+      "loss": 2.0708,
+      "step": 3660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 71.625,
+      "learning_rate": 9.304221575470736e-07,
+      "loss": 2.1001,
+      "step": 3680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 62.78125,
+      "learning_rate": 9.300440140162681e-07,
+      "loss": 2.1044,
+      "step": 3700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 98.375,
+      "learning_rate": 9.296658704854626e-07,
+      "loss": 2.0937,
+      "step": 3720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 80.875,
+      "learning_rate": 9.292877269546569e-07,
+      "loss": 2.1337,
+      "step": 3740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 60.9375,
+      "learning_rate": 9.289095834238514e-07,
+      "loss": 2.066,
+      "step": 3760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 58.8125,
+      "learning_rate": 9.285314398930458e-07,
+      "loss": 2.0683,
+      "step": 3780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 78.125,
+      "learning_rate": 9.281532963622403e-07,
+      "loss": 1.9853,
+      "step": 3800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 79.6875,
+      "learning_rate": 9.277751528314347e-07,
+      "loss": 2.0792,
+      "step": 3820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 71.0625,
+      "learning_rate": 9.273970093006291e-07,
+      "loss": 2.0544,
+      "step": 3840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 45.03125,
+      "learning_rate": 9.270188657698236e-07,
+      "loss": 2.0586,
+      "step": 3860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 50.3125,
+      "learning_rate": 9.26640722239018e-07,
+      "loss": 2.0716,
+      "step": 3880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 67.125,
+      "learning_rate": 9.262625787082124e-07,
+      "loss": 2.0274,
+      "step": 3900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 51.84375,
+      "learning_rate": 9.258844351774069e-07,
+      "loss": 2.02,
+      "step": 3920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 56.71875,
+      "learning_rate": 9.255062916466014e-07,
+      "loss": 2.1464,
+      "step": 3940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 63.34375,
+      "learning_rate": 9.251281481157957e-07,
+      "loss": 2.08,
+      "step": 3960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 98.5625,
+      "learning_rate": 9.247500045849902e-07,
+      "loss": 2.1212,
+      "step": 3980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 72.75,
+      "learning_rate": 9.243718610541848e-07,
+      "loss": 2.1505,
+      "step": 4000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 65.625,
+      "learning_rate": 9.239937175233792e-07,
+      "loss": 2.0742,
+      "step": 4020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 90.625,
+      "learning_rate": 9.236155739925736e-07,
+      "loss": 2.0463,
+      "step": 4040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 78.1875,
+      "learning_rate": 9.23237430461768e-07,
+      "loss": 2.0287,
+      "step": 4060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 51.875,
+      "learning_rate": 9.228592869309625e-07,
+      "loss": 2.0933,
+      "step": 4080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 56.78125,
+      "learning_rate": 9.22481143400157e-07,
+      "loss": 2.001,
+      "step": 4100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 51.90625,
+      "learning_rate": 9.221029998693513e-07,
+      "loss": 1.9934,
+      "step": 4120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 55.78125,
+      "learning_rate": 9.217248563385458e-07,
+      "loss": 2.0693,
+      "step": 4140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 103.0625,
+      "learning_rate": 9.213467128077403e-07,
+      "loss": 2.0205,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 54.375,
+      "learning_rate": 9.209685692769346e-07,
+      "loss": 2.0219,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 65.6875,
+      "learning_rate": 9.205904257461291e-07,
+      "loss": 2.1043,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 70.8125,
+      "learning_rate": 9.202122822153236e-07,
+      "loss": 2.1014,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 73.5625,
+      "learning_rate": 9.19834138684518e-07,
+      "loss": 1.9973,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 61.375,
+      "learning_rate": 9.194559951537124e-07,
+      "loss": 2.0508,
+      "step": 4260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 53.6875,
+      "learning_rate": 9.190778516229069e-07,
+      "loss": 2.0495,
+      "step": 4280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 97.5,
+      "learning_rate": 9.186997080921013e-07,
+      "loss": 2.1024,
+      "step": 4300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 80.3125,
+      "learning_rate": 9.183215645612957e-07,
+      "loss": 2.1573,
+      "step": 4320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 57.875,
+      "learning_rate": 9.179434210304903e-07,
+      "loss": 2.0675,
+      "step": 4340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 68.1875,
+      "learning_rate": 9.175652774996847e-07,
+      "loss": 2.0606,
+      "step": 4360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 79.8125,
+      "learning_rate": 9.171871339688792e-07,
+      "loss": 1.9876,
+      "step": 4380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 79.0,
+      "learning_rate": 9.168089904380735e-07,
+      "loss": 2.0817,
+      "step": 4400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 78.0,
+      "learning_rate": 9.16430846907268e-07,
+      "loss": 2.0678,
+      "step": 4420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 90.5625,
+      "learning_rate": 9.160527033764625e-07,
+      "loss": 2.1116,
+      "step": 4440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 103.4375,
+      "learning_rate": 9.156745598456569e-07,
+      "loss": 2.1465,
+      "step": 4460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 91.9375,
+      "learning_rate": 9.152964163148513e-07,
+      "loss": 2.2426,
+      "step": 4480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 80.125,
+      "learning_rate": 9.149182727840458e-07,
+      "loss": 2.071,
+      "step": 4500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 36.9375,
+      "learning_rate": 9.145401292532402e-07,
+      "loss": 2.0632,
+      "step": 4520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 59.96875,
+      "learning_rate": 9.141619857224346e-07,
+      "loss": 2.0746,
+      "step": 4540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 56.65625,
+      "learning_rate": 9.137838421916291e-07,
+      "loss": 1.9805,
+      "step": 4560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 48.625,
+      "learning_rate": 9.134056986608235e-07,
+      "loss": 2.0773,
+      "step": 4580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 70.6875,
+      "learning_rate": 9.13027555130018e-07,
+      "loss": 2.1277,
+      "step": 4600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 56.0625,
+      "learning_rate": 9.126494115992124e-07,
+      "loss": 2.0332,
+      "step": 4620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 52.90625,
+      "learning_rate": 9.122712680684068e-07,
+      "loss": 2.0042,
+      "step": 4640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 64.6875,
+      "learning_rate": 9.118931245376013e-07,
+      "loss": 2.0929,
+      "step": 4660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 46.0,
+      "learning_rate": 9.115149810067956e-07,
+      "loss": 2.1159,
+      "step": 4680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 61.53125,
+      "learning_rate": 9.111368374759902e-07,
+      "loss": 2.1719,
+      "step": 4700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 69.625,
+      "learning_rate": 9.107586939451847e-07,
+      "loss": 2.0822,
+      "step": 4720
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 83.5,
+      "learning_rate": 9.103805504143791e-07,
+      "loss": 2.0584,
+      "step": 4740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 81.8125,
+      "learning_rate": 9.100024068835735e-07,
+      "loss": 2.0803,
+      "step": 4760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 81.9375,
+      "learning_rate": 9.09624263352768e-07,
+      "loss": 2.0664,
+      "step": 4780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 97.0625,
+      "learning_rate": 9.092461198219624e-07,
+      "loss": 2.0345,
+      "step": 4800
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 63.03125,
+      "learning_rate": 9.088679762911569e-07,
+      "loss": 2.1597,
+      "step": 4820
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 73.4375,
+      "learning_rate": 9.084898327603513e-07,
+      "loss": 2.0724,
+      "step": 4840
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 67.625,
+      "learning_rate": 9.081116892295457e-07,
+      "loss": 1.9974,
+      "step": 4860
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 62.15625,
+      "learning_rate": 9.077335456987402e-07,
+      "loss": 2.0378,
+      "step": 4880
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 78.8125,
+      "learning_rate": 9.073554021679346e-07,
+      "loss": 2.1548,
+      "step": 4900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 52.6875,
+      "learning_rate": 9.06977258637129e-07,
+      "loss": 2.0852,
+      "step": 4920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 53.6875,
+      "learning_rate": 9.065991151063235e-07,
+      "loss": 2.147,
+      "step": 4940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 72.0625,
+      "learning_rate": 9.062209715755179e-07,
+      "loss": 1.944,
+      "step": 4960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 112.125,
+      "learning_rate": 9.058428280447123e-07,
+      "loss": 2.0855,
+      "step": 4980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 91.3125,
+      "learning_rate": 9.054646845139068e-07,
+      "loss": 1.9843,
+      "step": 5000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 101.75,
+      "learning_rate": 9.050865409831012e-07,
+      "loss": 2.1488,
+      "step": 5020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 61.4375,
+      "learning_rate": 9.047083974522958e-07,
+      "loss": 2.0525,
+      "step": 5040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 67.5625,
+      "learning_rate": 9.043302539214902e-07,
+      "loss": 2.105,
+      "step": 5060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 86.25,
+      "learning_rate": 9.039521103906846e-07,
+      "loss": 2.0717,
+      "step": 5080
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 64.3125,
+      "learning_rate": 9.035739668598791e-07,
+      "loss": 2.1015,
+      "step": 5100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 166.625,
+      "learning_rate": 9.031958233290735e-07,
+      "loss": 2.134,
+      "step": 5120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 68.4375,
+      "learning_rate": 9.028176797982679e-07,
+      "loss": 2.0938,
+      "step": 5140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 77.4375,
+      "learning_rate": 9.024395362674624e-07,
+      "loss": 2.0848,
+      "step": 5160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 102.1875,
+      "learning_rate": 9.020613927366569e-07,
+      "loss": 1.9225,
+      "step": 5180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 68.5625,
+      "learning_rate": 9.016832492058512e-07,
+      "loss": 2.0038,
+      "step": 5200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 55.09375,
+      "learning_rate": 9.013051056750457e-07,
+      "loss": 2.1238,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 65.75,
+      "learning_rate": 9.009269621442402e-07,
+      "loss": 2.0999,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 62.28125,
+      "learning_rate": 9.005488186134345e-07,
+      "loss": 2.0663,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 96.3125,
+      "learning_rate": 9.00170675082629e-07,
+      "loss": 2.06,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 82.5625,
+      "learning_rate": 8.997925315518234e-07,
+      "loss": 2.1223,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 116.3125,
+      "learning_rate": 8.994143880210179e-07,
+      "loss": 2.0237,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 63.875,
+      "learning_rate": 8.990362444902123e-07,
+      "loss": 2.0965,
+      "step": 5340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 51.78125,
+      "learning_rate": 8.986581009594067e-07,
+      "loss": 2.0365,
+      "step": 5360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 67.25,
+      "learning_rate": 8.982799574286012e-07,
+      "loss": 2.1012,
+      "step": 5380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 69.3125,
+      "learning_rate": 8.979018138977958e-07,
+      "loss": 1.9336,
+      "step": 5400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 127.125,
+      "learning_rate": 8.975236703669901e-07,
+      "loss": 2.1657,
+      "step": 5420
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 77.4375,
+      "learning_rate": 8.971455268361846e-07,
+      "loss": 1.9824,
+      "step": 5440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 51.28125,
+      "learning_rate": 8.967673833053791e-07,
+      "loss": 2.1134,
+      "step": 5460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 82.4375,
+      "learning_rate": 8.963892397745734e-07,
+      "loss": 2.0806,
+      "step": 5480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 155.875,
+      "learning_rate": 8.960110962437679e-07,
+      "loss": 2.13,
+      "step": 5500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 49.0625,
+      "learning_rate": 8.956329527129624e-07,
+      "loss": 2.1733,
+      "step": 5520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 72.75,
+      "learning_rate": 8.952548091821568e-07,
+      "loss": 2.1312,
+      "step": 5540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 98.0,
+      "learning_rate": 8.948766656513512e-07,
+      "loss": 2.0926,
+      "step": 5560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 47.75,
+      "learning_rate": 8.944985221205457e-07,
+      "loss": 2.0788,
+      "step": 5580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 69.5625,
+      "learning_rate": 8.941203785897401e-07,
+      "loss": 2.1218,
+      "step": 5600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 80.4375,
+      "learning_rate": 8.937422350589346e-07,
+      "loss": 2.1038,
+      "step": 5620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 54.625,
+      "learning_rate": 8.933640915281289e-07,
+      "loss": 2.0546,
+      "step": 5640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 64.375,
+      "learning_rate": 8.929859479973234e-07,
+      "loss": 2.0949,
+      "step": 5660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 69.0,
+      "learning_rate": 8.926078044665179e-07,
+      "loss": 2.037,
+      "step": 5680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 80.75,
+      "learning_rate": 8.922296609357122e-07,
+      "loss": 2.1122,
+      "step": 5700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 81.75,
+      "learning_rate": 8.918515174049067e-07,
+      "loss": 2.0332,
+      "step": 5720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 80.0625,
+      "learning_rate": 8.914733738741013e-07,
+      "loss": 1.9553,
+      "step": 5740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 93.625,
+      "learning_rate": 8.910952303432957e-07,
+      "loss": 1.9639,
+      "step": 5760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 67.875,
+      "learning_rate": 8.907170868124901e-07,
+      "loss": 2.0789,
+      "step": 5780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 71.625,
+      "learning_rate": 8.903389432816846e-07,
+      "loss": 2.2702,
+      "step": 5800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 81.3125,
+      "learning_rate": 8.89960799750879e-07,
+      "loss": 2.124,
+      "step": 5820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 96.6875,
+      "learning_rate": 8.895826562200735e-07,
+      "loss": 2.0788,
+      "step": 5840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 51.375,
+      "learning_rate": 8.892045126892679e-07,
+      "loss": 2.1448,
+      "step": 5860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 69.625,
+      "learning_rate": 8.888263691584623e-07,
+      "loss": 2.0283,
+      "step": 5880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 63.71875,
+      "learning_rate": 8.884482256276568e-07,
+      "loss": 2.0592,
+      "step": 5900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 81.125,
+      "learning_rate": 8.880700820968511e-07,
+      "loss": 2.1446,
+      "step": 5920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 55.84375,
+      "learning_rate": 8.876919385660456e-07,
+      "loss": 2.0625,
+      "step": 5940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 75.375,
+      "learning_rate": 8.873137950352401e-07,
+      "loss": 2.0988,
+      "step": 5960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 68.0,
+      "learning_rate": 8.869356515044345e-07,
+      "loss": 2.007,
+      "step": 5980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 132.875,
+      "learning_rate": 8.865575079736289e-07,
+      "loss": 2.103,
+      "step": 6000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 55.53125,
+      "learning_rate": 8.861793644428234e-07,
+      "loss": 1.9625,
+      "step": 6020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 76.875,
+      "learning_rate": 8.858012209120178e-07,
+      "loss": 2.1369,
+      "step": 6040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 108.0625,
+      "learning_rate": 8.854230773812122e-07,
+      "loss": 2.0546,
+      "step": 6060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 199.625,
+      "learning_rate": 8.850449338504067e-07,
+      "loss": 2.1589,
+      "step": 6080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 58.53125,
+      "learning_rate": 8.846667903196012e-07,
+      "loss": 2.0696,
+      "step": 6100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 52.84375,
+      "learning_rate": 8.842886467887957e-07,
+      "loss": 2.0543,
+      "step": 6120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 76.0625,
+      "learning_rate": 8.839105032579901e-07,
+      "loss": 2.0709,
+      "step": 6140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 86.125,
+      "learning_rate": 8.835323597271845e-07,
+      "loss": 2.1248,
+      "step": 6160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 66.1875,
+      "learning_rate": 8.83154216196379e-07,
+      "loss": 2.0078,
+      "step": 6180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 91.875,
+      "learning_rate": 8.827760726655735e-07,
+      "loss": 2.1086,
+      "step": 6200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 70.75,
+      "learning_rate": 8.823979291347678e-07,
+      "loss": 2.0076,
+      "step": 6220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 58.0,
+      "learning_rate": 8.820197856039623e-07,
+      "loss": 1.9816,
+      "step": 6240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 70.1875,
+      "learning_rate": 8.816416420731567e-07,
+      "loss": 2.0527,
+      "step": 6260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 64.5625,
+      "learning_rate": 8.812634985423511e-07,
+      "loss": 2.0384,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 57.03125,
+      "learning_rate": 8.808853550115456e-07,
+      "loss": 2.091,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 73.0,
+      "learning_rate": 8.8050721148074e-07,
+      "loss": 2.0033,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 61.90625,
+      "learning_rate": 8.801290679499345e-07,
+      "loss": 2.0876,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 56.65625,
+      "learning_rate": 8.797509244191289e-07,
+      "loss": 2.1797,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 72.5625,
+      "learning_rate": 8.793727808883233e-07,
+      "loss": 2.0717,
+      "step": 6380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 149.375,
+      "learning_rate": 8.789946373575178e-07,
+      "loss": 2.0591,
+      "step": 6400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 109.8125,
+      "learning_rate": 8.786164938267123e-07,
+      "loss": 2.0082,
+      "step": 6420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 62.03125,
+      "learning_rate": 8.782383502959067e-07,
+      "loss": 2.0883,
+      "step": 6440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 63.96875,
+      "learning_rate": 8.778602067651012e-07,
+      "loss": 2.0119,
+      "step": 6460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 114.875,
+      "learning_rate": 8.774820632342957e-07,
+      "loss": 2.1563,
+      "step": 6480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 62.15625,
+      "learning_rate": 8.7710391970349e-07,
+      "loss": 2.0736,
+      "step": 6500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 79.125,
+      "learning_rate": 8.767257761726845e-07,
+      "loss": 2.0764,
+      "step": 6520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 63.5625,
+      "learning_rate": 8.76347632641879e-07,
+      "loss": 2.0432,
+      "step": 6540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 55.15625,
+      "learning_rate": 8.759694891110734e-07,
+      "loss": 2.0265,
+      "step": 6560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 66.875,
+      "learning_rate": 8.755913455802678e-07,
+      "loss": 1.9963,
+      "step": 6580
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 82.5625,
+      "learning_rate": 8.752132020494622e-07,
+      "loss": 1.984,
+      "step": 6600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 64.5,
+      "learning_rate": 8.748350585186567e-07,
+      "loss": 2.0488,
+      "step": 6620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 81.4375,
+      "learning_rate": 8.744569149878512e-07,
+      "loss": 2.1111,
+      "step": 6640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 74.6875,
+      "learning_rate": 8.740787714570455e-07,
+      "loss": 2.0399,
+      "step": 6660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 70.3125,
+      "learning_rate": 8.7370062792624e-07,
+      "loss": 1.9512,
+      "step": 6680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 166.0,
+      "learning_rate": 8.733224843954345e-07,
+      "loss": 2.0779,
+      "step": 6700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 71.125,
+      "learning_rate": 8.729443408646288e-07,
+      "loss": 2.0487,
+      "step": 6720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 119.8125,
+      "learning_rate": 8.725661973338233e-07,
+      "loss": 2.1232,
+      "step": 6740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 104.5625,
+      "learning_rate": 8.721880538030178e-07,
+      "loss": 1.9915,
+      "step": 6760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 74.125,
+      "learning_rate": 8.718099102722122e-07,
+      "loss": 2.1389,
+      "step": 6780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 53.90625,
+      "learning_rate": 8.714317667414067e-07,
+      "loss": 2.1114,
+      "step": 6800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 67.125,
+      "learning_rate": 8.710536232106012e-07,
+      "loss": 1.9978,
+      "step": 6820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 69.125,
+      "learning_rate": 8.706754796797956e-07,
+      "loss": 2.1736,
+      "step": 6840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 61.0625,
+      "learning_rate": 8.702973361489901e-07,
+      "loss": 2.1067,
+      "step": 6860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 74.1875,
+      "learning_rate": 8.699191926181844e-07,
+      "loss": 1.9871,
+      "step": 6880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 61.78125,
+      "learning_rate": 8.695410490873789e-07,
+      "loss": 1.9737,
+      "step": 6900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 66.625,
+      "learning_rate": 8.691629055565734e-07,
+      "loss": 2.0105,
+      "step": 6920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 68.6875,
+      "learning_rate": 8.687847620257677e-07,
+      "loss": 2.026,
+      "step": 6940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 49.65625,
+      "learning_rate": 8.684066184949622e-07,
+      "loss": 2.1013,
+      "step": 6960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 83.1875,
+      "learning_rate": 8.680284749641567e-07,
+      "loss": 2.0245,
+      "step": 6980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 86.9375,
+      "learning_rate": 8.676503314333511e-07,
+      "loss": 1.9986,
+      "step": 7000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 56.59375,
+      "learning_rate": 8.672721879025455e-07,
+      "loss": 2.0128,
+      "step": 7020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 63.90625,
+      "learning_rate": 8.6689404437174e-07,
+      "loss": 2.0212,
+      "step": 7040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 76.75,
+      "learning_rate": 8.665159008409344e-07,
+      "loss": 2.077,
+      "step": 7060
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 84.3125,
+      "learning_rate": 8.661377573101288e-07,
+      "loss": 2.075,
+      "step": 7080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 99.0,
+      "learning_rate": 8.657596137793233e-07,
+      "loss": 2.1219,
+      "step": 7100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 112.75,
+      "learning_rate": 8.653814702485177e-07,
+      "loss": 1.8914,
+      "step": 7120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 83.1875,
+      "learning_rate": 8.650033267177123e-07,
+      "loss": 2.091,
+      "step": 7140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 57.65625,
+      "learning_rate": 8.646251831869067e-07,
+      "loss": 2.1047,
+      "step": 7160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 70.0625,
+      "learning_rate": 8.642470396561011e-07,
+      "loss": 2.0086,
+      "step": 7180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 60.40625,
+      "learning_rate": 8.638688961252956e-07,
+      "loss": 2.0193,
+      "step": 7200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 124.125,
+      "learning_rate": 8.6349075259449e-07,
+      "loss": 2.0588,
+      "step": 7220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 49.28125,
+      "learning_rate": 8.631126090636844e-07,
+      "loss": 1.9948,
+      "step": 7240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 71.75,
+      "learning_rate": 8.627344655328789e-07,
+      "loss": 2.0766,
+      "step": 7260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 59.9375,
+      "learning_rate": 8.623563220020733e-07,
+      "loss": 1.9796,
+      "step": 7280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 61.46875,
+      "learning_rate": 8.619781784712677e-07,
+      "loss": 2.213,
+      "step": 7300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 52.40625,
+      "learning_rate": 8.616000349404622e-07,
+      "loss": 2.1209,
+      "step": 7320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 86.25,
+      "learning_rate": 8.612218914096566e-07,
+      "loss": 1.9509,
+      "step": 7340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 78.0,
+      "learning_rate": 8.608437478788511e-07,
+      "loss": 1.9703,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 62.8125,
+      "learning_rate": 8.604656043480455e-07,
+      "loss": 1.9881,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 125.875,
+      "learning_rate": 8.600874608172399e-07,
+      "loss": 1.985,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 73.375,
+      "learning_rate": 8.597093172864344e-07,
+      "loss": 2.0481,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 109.1875,
+      "learning_rate": 8.593311737556289e-07,
+      "loss": 1.9969,
+      "step": 7440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 59.46875,
+      "learning_rate": 8.589530302248232e-07,
+      "loss": 1.9769,
+      "step": 7460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 49.71875,
+      "learning_rate": 8.585748866940178e-07,
+      "loss": 2.0271,
+      "step": 7480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 85.125,
+      "learning_rate": 8.581967431632122e-07,
+      "loss": 2.0354,
+      "step": 7500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 58.8125,
+      "learning_rate": 8.578185996324066e-07,
+      "loss": 2.0471,
+      "step": 7520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 90.25,
+      "learning_rate": 8.574404561016011e-07,
+      "loss": 2.0257,
+      "step": 7540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 68.5,
+      "learning_rate": 8.570623125707955e-07,
+      "loss": 2.0477,
+      "step": 7560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 73.875,
+      "learning_rate": 8.5668416903999e-07,
+      "loss": 2.1175,
+      "step": 7580
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 62.1875,
+      "learning_rate": 8.563060255091844e-07,
+      "loss": 2.0251,
+      "step": 7600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 80.625,
+      "learning_rate": 8.559278819783788e-07,
+      "loss": 2.0128,
+      "step": 7620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 57.25,
+      "learning_rate": 8.555497384475733e-07,
+      "loss": 2.0853,
+      "step": 7640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 95.375,
+      "learning_rate": 8.551715949167678e-07,
+      "loss": 2.0591,
+      "step": 7660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 53.625,
+      "learning_rate": 8.547934513859621e-07,
+      "loss": 2.081,
+      "step": 7680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 74.875,
+      "learning_rate": 8.544153078551566e-07,
+      "loss": 2.0432,
+      "step": 7700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 78.3125,
+      "learning_rate": 8.540371643243511e-07,
+      "loss": 1.9401,
+      "step": 7720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 112.75,
+      "learning_rate": 8.536590207935454e-07,
+      "loss": 2.1571,
+      "step": 7740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 56.65625,
+      "learning_rate": 8.532808772627399e-07,
+      "loss": 2.0862,
+      "step": 7760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 113.1875,
+      "learning_rate": 8.529027337319343e-07,
+      "loss": 2.0632,
+      "step": 7780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 94.3125,
+      "learning_rate": 8.525245902011288e-07,
+      "loss": 2.0372,
+      "step": 7800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 51.125,
+      "learning_rate": 8.521464466703232e-07,
+      "loss": 2.0205,
+      "step": 7820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 54.28125,
+      "learning_rate": 8.517683031395177e-07,
+      "loss": 1.9503,
+      "step": 7840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 60.625,
+      "learning_rate": 8.513901596087122e-07,
+      "loss": 2.1562,
+      "step": 7860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 64.75,
+      "learning_rate": 8.510120160779067e-07,
+      "loss": 2.0641,
+      "step": 7880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 62.59375,
+      "learning_rate": 8.50633872547101e-07,
+      "loss": 2.0897,
+      "step": 7900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 99.5625,
+      "learning_rate": 8.502557290162955e-07,
+      "loss": 2.0611,
+      "step": 7920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 48.28125,
+      "learning_rate": 8.4987758548549e-07,
+      "loss": 2.0871,
+      "step": 7940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 55.5,
+      "learning_rate": 8.494994419546843e-07,
+      "loss": 1.979,
+      "step": 7960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 70.125,
+      "learning_rate": 8.491212984238788e-07,
+      "loss": 2.1194,
+      "step": 7980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 83.3125,
+      "learning_rate": 8.487431548930733e-07,
+      "loss": 2.0838,
+      "step": 8000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 65.25,
+      "learning_rate": 8.483650113622677e-07,
+      "loss": 2.079,
+      "step": 8020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 68.0625,
+      "learning_rate": 8.479868678314621e-07,
+      "loss": 2.0234,
+      "step": 8040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 64.8125,
+      "learning_rate": 8.476087243006566e-07,
+      "loss": 2.0205,
+      "step": 8060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 56.0,
+      "learning_rate": 8.47230580769851e-07,
+      "loss": 1.975,
+      "step": 8080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 81.625,
+      "learning_rate": 8.468524372390454e-07,
+      "loss": 2.1159,
+      "step": 8100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 53.21875,
+      "learning_rate": 8.464742937082398e-07,
+      "loss": 2.0243,
+      "step": 8120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 88.625,
+      "learning_rate": 8.460961501774343e-07,
+      "loss": 1.9559,
+      "step": 8140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 93.3125,
+      "learning_rate": 8.457180066466288e-07,
+      "loss": 2.1196,
+      "step": 8160
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 61.09375,
+      "learning_rate": 8.453398631158232e-07,
+      "loss": 1.896,
+      "step": 8180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 46.9375,
+      "learning_rate": 8.449617195850177e-07,
+      "loss": 2.0552,
+      "step": 8200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 84.125,
+      "learning_rate": 8.445835760542122e-07,
+      "loss": 2.0865,
+      "step": 8220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 48.53125,
+      "learning_rate": 8.442054325234066e-07,
+      "loss": 2.049,
+      "step": 8240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 70.9375,
+      "learning_rate": 8.43827288992601e-07,
+      "loss": 2.0403,
+      "step": 8260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 65.6875,
+      "learning_rate": 8.434491454617955e-07,
+      "loss": 2.1368,
+      "step": 8280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 77.6875,
+      "learning_rate": 8.430710019309899e-07,
+      "loss": 1.979,
+      "step": 8300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 70.0,
+      "learning_rate": 8.426928584001843e-07,
+      "loss": 1.8971,
+      "step": 8320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 168.875,
+      "learning_rate": 8.423147148693788e-07,
+      "loss": 1.999,
+      "step": 8340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 67.8125,
+      "learning_rate": 8.419365713385732e-07,
+      "loss": 1.9871,
+      "step": 8360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 66.6875,
+      "learning_rate": 8.415584278077677e-07,
+      "loss": 1.867,
+      "step": 8380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 63.625,
+      "learning_rate": 8.41180284276962e-07,
+      "loss": 1.9582,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 63.28125,
+      "learning_rate": 8.408021407461565e-07,
+      "loss": 1.9924,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 62.40625,
+      "learning_rate": 8.40423997215351e-07,
+      "loss": 1.9955,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 48.65625,
+      "learning_rate": 8.400458536845453e-07,
+      "loss": 2.0825,
+      "step": 8460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 75.9375,
+      "learning_rate": 8.396677101537398e-07,
+      "loss": 1.9676,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 107.1875,
+      "learning_rate": 8.392895666229343e-07,
+      "loss": 1.9956,
+      "step": 8500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 52.5625,
+      "learning_rate": 8.389114230921287e-07,
+      "loss": 2.0818,
+      "step": 8520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 74.625,
+      "learning_rate": 8.385332795613232e-07,
+      "loss": 1.9427,
+      "step": 8540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 70.125,
+      "learning_rate": 8.381551360305177e-07,
+      "loss": 2.1569,
+      "step": 8560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 64.8125,
+      "learning_rate": 8.377769924997121e-07,
+      "loss": 2.0716,
+      "step": 8580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 59.875,
+      "learning_rate": 8.373988489689066e-07,
+      "loss": 2.06,
+      "step": 8600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 86.1875,
+      "learning_rate": 8.37020705438101e-07,
+      "loss": 2.0076,
+      "step": 8620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 49.8125,
+      "learning_rate": 8.366425619072954e-07,
+      "loss": 2.0739,
+      "step": 8640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 96.1875,
+      "learning_rate": 8.362644183764899e-07,
+      "loss": 2.0958,
+      "step": 8660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 81.9375,
+      "learning_rate": 8.358862748456844e-07,
+      "loss": 2.0505,
+      "step": 8680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 72.4375,
+      "learning_rate": 8.355081313148787e-07,
+      "loss": 2.0513,
+      "step": 8700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 65.125,
+      "learning_rate": 8.351299877840732e-07,
+      "loss": 1.9823,
+      "step": 8720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 60.28125,
+      "learning_rate": 8.347518442532676e-07,
+      "loss": 1.8811,
+      "step": 8740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 57.125,
+      "learning_rate": 8.34373700722462e-07,
+      "loss": 2.0142,
+      "step": 8760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 90.625,
+      "learning_rate": 8.339955571916565e-07,
+      "loss": 1.9925,
+      "step": 8780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 71.4375,
+      "learning_rate": 8.336174136608509e-07,
+      "loss": 2.1022,
+      "step": 8800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 95.5625,
+      "learning_rate": 8.332392701300454e-07,
+      "loss": 1.9648,
+      "step": 8820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 103.9375,
+      "learning_rate": 8.328611265992398e-07,
+      "loss": 2.0242,
+      "step": 8840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 54.96875,
+      "learning_rate": 8.324829830684342e-07,
+      "loss": 1.9736,
+      "step": 8860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 53.0,
+      "learning_rate": 8.321048395376288e-07,
+      "loss": 2.0328,
+      "step": 8880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 61.21875,
+      "learning_rate": 8.317266960068233e-07,
+      "loss": 1.9957,
+      "step": 8900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 51.09375,
+      "learning_rate": 8.313485524760176e-07,
+      "loss": 1.9611,
+      "step": 8920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 48.6875,
+      "learning_rate": 8.309704089452121e-07,
+      "loss": 1.9684,
+      "step": 8940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 61.90625,
+      "learning_rate": 8.305922654144066e-07,
+      "loss": 1.989,
+      "step": 8960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 54.28125,
+      "learning_rate": 8.302141218836009e-07,
+      "loss": 1.9718,
+      "step": 8980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 90.625,
+      "learning_rate": 8.298359783527954e-07,
+      "loss": 2.0448,
+      "step": 9000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 98.4375,
+      "learning_rate": 8.294578348219899e-07,
+      "loss": 1.9535,
+      "step": 9020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 106.625,
+      "learning_rate": 8.290796912911843e-07,
+      "loss": 1.9675,
+      "step": 9040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 67.5625,
+      "learning_rate": 8.287015477603787e-07,
+      "loss": 2.0108,
+      "step": 9060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 47.28125,
+      "learning_rate": 8.283234042295731e-07,
+      "loss": 2.0104,
+      "step": 9080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 57.34375,
+      "learning_rate": 8.279452606987676e-07,
+      "loss": 2.0557,
+      "step": 9100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 95.5,
+      "learning_rate": 8.27567117167962e-07,
+      "loss": 1.9992,
+      "step": 9120
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 56.03125,
+      "learning_rate": 8.271889736371564e-07,
+      "loss": 1.9351,
+      "step": 9140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 121.6875,
+      "learning_rate": 8.268108301063509e-07,
+      "loss": 2.049,
+      "step": 9160
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 65.25,
+      "learning_rate": 8.264326865755454e-07,
+      "loss": 2.0567,
+      "step": 9180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 63.53125,
+      "learning_rate": 8.260545430447397e-07,
+      "loss": 2.1084,
+      "step": 9200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 62.3125,
+      "learning_rate": 8.256763995139342e-07,
+      "loss": 2.0061,
+      "step": 9220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 78.4375,
+      "learning_rate": 8.252982559831288e-07,
+      "loss": 2.0693,
+      "step": 9240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 61.6875,
+      "learning_rate": 8.249201124523232e-07,
+      "loss": 1.9503,
+      "step": 9260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 67.1875,
+      "learning_rate": 8.245419689215176e-07,
+      "loss": 2.061,
+      "step": 9280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 128.5,
+      "learning_rate": 8.241638253907121e-07,
+      "loss": 2.0328,
+      "step": 9300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 61.8125,
+      "learning_rate": 8.237856818599065e-07,
+      "loss": 1.9776,
+      "step": 9320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 79.375,
+      "learning_rate": 8.234075383291009e-07,
+      "loss": 1.9705,
+      "step": 9340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 52.1875,
+      "learning_rate": 8.230293947982953e-07,
+      "loss": 2.1075,
+      "step": 9360
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 54.6875,
+      "learning_rate": 8.226512512674898e-07,
+      "loss": 1.9804,
+      "step": 9380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 60.125,
+      "learning_rate": 8.222731077366843e-07,
+      "loss": 2.0447,
+      "step": 9400
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 57.46875,
+      "learning_rate": 8.218949642058786e-07,
+      "loss": 2.0118,
+      "step": 9420
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 59.78125,
+      "learning_rate": 8.215168206750731e-07,
+      "loss": 2.0065,
+      "step": 9440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 61.6875,
+      "learning_rate": 8.211386771442676e-07,
+      "loss": 1.9687,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 60.84375,
+      "learning_rate": 8.207605336134619e-07,
+      "loss": 2.0927,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 88.25,
+      "learning_rate": 8.203823900826564e-07,
+      "loss": 2.0385,
+      "step": 9500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 58.6875,
+      "learning_rate": 8.200042465518509e-07,
+      "loss": 2.0446,
+      "step": 9520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 65.375,
+      "learning_rate": 8.196261030210453e-07,
+      "loss": 2.1188,
+      "step": 9540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 58.46875,
+      "learning_rate": 8.192479594902397e-07,
+      "loss": 1.9974,
+      "step": 9560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 49.59375,
+      "learning_rate": 8.188698159594343e-07,
+      "loss": 1.9911,
+      "step": 9580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 75.1875,
+      "learning_rate": 8.184916724286287e-07,
+      "loss": 1.9578,
+      "step": 9600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 71.4375,
+      "learning_rate": 8.181135288978232e-07,
+      "loss": 2.0356,
+      "step": 9620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 63.9375,
+      "learning_rate": 8.177353853670176e-07,
+      "loss": 2.0255,
+      "step": 9640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 50.125,
+      "learning_rate": 8.17357241836212e-07,
+      "loss": 1.9512,
+      "step": 9660
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 118.0,
+      "learning_rate": 8.169790983054065e-07,
+      "loss": 2.0679,
+      "step": 9680
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 132.625,
+      "learning_rate": 8.166009547746008e-07,
+      "loss": 2.0188,
+      "step": 9700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 73.0,
+      "learning_rate": 8.162228112437953e-07,
+      "loss": 1.9632,
+      "step": 9720
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 66.125,
+      "learning_rate": 8.158446677129898e-07,
+      "loss": 1.9628,
+      "step": 9740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 62.15625,
+      "learning_rate": 8.154665241821842e-07,
+      "loss": 1.9865,
+      "step": 9760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 56.28125,
+      "learning_rate": 8.150883806513786e-07,
+      "loss": 2.1228,
+      "step": 9780
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 58.5625,
+      "learning_rate": 8.147102371205731e-07,
+      "loss": 1.9204,
+      "step": 9800
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 88.125,
+      "learning_rate": 8.143320935897675e-07,
+      "loss": 1.989,
+      "step": 9820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 100.5625,
+      "learning_rate": 8.13953950058962e-07,
+      "loss": 1.9791,
+      "step": 9840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 66.375,
+      "learning_rate": 8.135758065281564e-07,
+      "loss": 2.0461,
+      "step": 9860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 80.75,
+      "learning_rate": 8.131976629973508e-07,
+      "loss": 2.0083,
+      "step": 9880
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 52.34375,
+      "learning_rate": 8.128195194665453e-07,
+      "loss": 1.9767,
+      "step": 9900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 92.6875,
+      "learning_rate": 8.124413759357397e-07,
+      "loss": 2.0318,
+      "step": 9920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 70.625,
+      "learning_rate": 8.120632324049342e-07,
+      "loss": 2.0566,
+      "step": 9940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 86.9375,
+      "learning_rate": 8.116850888741287e-07,
+      "loss": 2.0286,
+      "step": 9960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 65.875,
+      "learning_rate": 8.113069453433232e-07,
+      "loss": 1.9629,
+      "step": 9980
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 90.6875,
+      "learning_rate": 8.109288018125175e-07,
+      "loss": 1.9711,
+      "step": 10000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 52.125,
+      "learning_rate": 8.10550658281712e-07,
+      "loss": 1.9985,
+      "step": 10020
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 66.25,
+      "learning_rate": 8.101725147509064e-07,
+      "loss": 2.0052,
+      "step": 10040
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 65.375,
+      "learning_rate": 8.097943712201009e-07,
+      "loss": 1.9346,
+      "step": 10060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 53.03125,
+      "learning_rate": 8.094162276892953e-07,
+      "loss": 1.984,
+      "step": 10080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 110.25,
+      "learning_rate": 8.090380841584897e-07,
+      "loss": 2.0204,
+      "step": 10100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 59.03125,
+      "learning_rate": 8.086599406276842e-07,
+      "loss": 2.0792,
+      "step": 10120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 49.8125,
+      "learning_rate": 8.082817970968786e-07,
+      "loss": 2.0051,
+      "step": 10140
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 61.1875,
+      "learning_rate": 8.07903653566073e-07,
+      "loss": 2.0679,
+      "step": 10160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 93.3125,
+      "learning_rate": 8.075255100352675e-07,
+      "loss": 2.099,
+      "step": 10180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 79.6875,
+      "learning_rate": 8.07147366504462e-07,
+      "loss": 2.1754,
+      "step": 10200
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 68.875,
+      "learning_rate": 8.067692229736563e-07,
+      "loss": 2.0048,
+      "step": 10220
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 72.625,
+      "learning_rate": 8.063910794428508e-07,
+      "loss": 1.9806,
+      "step": 10240
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 98.1875,
+      "learning_rate": 8.060129359120452e-07,
+      "loss": 1.9797,
+      "step": 10260
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 67.0625,
+      "learning_rate": 8.056347923812397e-07,
+      "loss": 2.0226,
+      "step": 10280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 55.40625,
+      "learning_rate": 8.052566488504342e-07,
+      "loss": 2.0109,
+      "step": 10300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 107.125,
+      "learning_rate": 8.048785053196286e-07,
+      "loss": 1.8523,
+      "step": 10320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 68.5625,
+      "learning_rate": 8.045003617888231e-07,
+      "loss": 1.9492,
+      "step": 10340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 77.0,
+      "learning_rate": 8.041222182580175e-07,
+      "loss": 1.9863,
+      "step": 10360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 76.5625,
+      "learning_rate": 8.037440747272119e-07,
+      "loss": 1.9792,
+      "step": 10380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 76.0625,
+      "learning_rate": 8.033659311964064e-07,
+      "loss": 2.0448,
+      "step": 10400
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 79.3125,
+      "learning_rate": 8.029877876656009e-07,
+      "loss": 2.0606,
+      "step": 10420
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 53.5625,
+      "learning_rate": 8.026096441347952e-07,
+      "loss": 2.0716,
+      "step": 10440
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 75.0,
+      "learning_rate": 8.022315006039897e-07,
+      "loss": 1.9026,
+      "step": 10460
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 85.0625,
+      "learning_rate": 8.018533570731842e-07,
+      "loss": 1.9879,
+      "step": 10480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 52.46875,
+      "learning_rate": 8.014752135423785e-07,
+      "loss": 2.0018,
+      "step": 10500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 48.4375,
+      "learning_rate": 8.01097070011573e-07,
+      "loss": 1.9966,
+      "step": 10520
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 59.65625,
+      "learning_rate": 8.007189264807675e-07,
+      "loss": 2.0212,
+      "step": 10540
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 49.59375,
+      "learning_rate": 8.003407829499619e-07,
+      "loss": 2.0244,
+      "step": 10560
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 107.625,
+      "learning_rate": 7.999626394191563e-07,
+      "loss": 1.9597,
+      "step": 10580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 67.875,
+      "learning_rate": 7.995844958883507e-07,
+      "loss": 1.7883,
+      "step": 10600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 67.8125,
+      "learning_rate": 7.992063523575452e-07,
+      "loss": 1.7906,
+      "step": 10620
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 60.09375,
+      "learning_rate": 7.988282088267398e-07,
+      "loss": 1.8029,
+      "step": 10640
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 42.25,
+      "learning_rate": 7.984500652959341e-07,
+      "loss": 1.7842,
+      "step": 10660
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 65.6875,
+      "learning_rate": 7.980719217651286e-07,
+      "loss": 1.8204,
+      "step": 10680
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 65.625,
+      "learning_rate": 7.976937782343231e-07,
+      "loss": 1.7802,
+      "step": 10700
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 47.96875,
+      "learning_rate": 7.973156347035174e-07,
+      "loss": 1.7412,
+      "step": 10720
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 78.875,
+      "learning_rate": 7.969374911727119e-07,
+      "loss": 1.8131,
+      "step": 10740
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 56.09375,
+      "learning_rate": 7.965593476419064e-07,
+      "loss": 1.7726,
+      "step": 10760
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 61.1875,
+      "learning_rate": 7.961812041111008e-07,
+      "loss": 1.8361,
+      "step": 10780
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 48.46875,
+      "learning_rate": 7.958030605802952e-07,
+      "loss": 1.6977,
+      "step": 10800
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 74.75,
+      "learning_rate": 7.954249170494897e-07,
+      "loss": 1.7696,
+      "step": 10820
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 72.875,
+      "learning_rate": 7.950467735186841e-07,
+      "loss": 1.7746,
+      "step": 10840
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 47.46875,
+      "learning_rate": 7.946686299878785e-07,
+      "loss": 1.7854,
+      "step": 10860
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 75.0625,
+      "learning_rate": 7.94290486457073e-07,
+      "loss": 1.782,
+      "step": 10880
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 133.25,
+      "learning_rate": 7.939123429262674e-07,
+      "loss": 1.7646,
+      "step": 10900
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 74.5,
+      "learning_rate": 7.935341993954619e-07,
+      "loss": 1.6605,
+      "step": 10920
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 97.1875,
+      "learning_rate": 7.931560558646562e-07,
+      "loss": 1.8058,
+      "step": 10940
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 63.09375,
+      "learning_rate": 7.927779123338507e-07,
+      "loss": 1.8219,
+      "step": 10960
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 72.75,
+      "learning_rate": 7.923997688030453e-07,
+      "loss": 1.7732,
+      "step": 10980
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 93.8125,
+      "learning_rate": 7.920216252722397e-07,
+      "loss": 1.7333,
+      "step": 11000
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 68.375,
+      "learning_rate": 7.916434817414341e-07,
+      "loss": 1.7237,
+      "step": 11020
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 72.1875,
+      "learning_rate": 7.912653382106286e-07,
+      "loss": 1.8252,
+      "step": 11040
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 61.46875,
+      "learning_rate": 7.90887194679823e-07,
+      "loss": 1.688,
+      "step": 11060
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 50.96875,
+      "learning_rate": 7.905090511490174e-07,
+      "loss": 1.7669,
+      "step": 11080
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 49.28125,
+      "learning_rate": 7.901309076182119e-07,
+      "loss": 1.7935,
+      "step": 11100
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 46.09375,
+      "learning_rate": 7.897527640874063e-07,
+      "loss": 1.8401,
+      "step": 11120
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 91.375,
+      "learning_rate": 7.893746205566008e-07,
+      "loss": 1.8535,
+      "step": 11140
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 79.125,
+      "learning_rate": 7.889964770257952e-07,
+      "loss": 1.7106,
+      "step": 11160
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 84.6875,
+      "learning_rate": 7.886183334949896e-07,
+      "loss": 1.7925,
+      "step": 11180
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 60.59375,
+      "learning_rate": 7.882401899641841e-07,
+      "loss": 1.7557,
+      "step": 11200
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 66.5,
+      "learning_rate": 7.878620464333785e-07,
+      "loss": 1.7569,
+      "step": 11220
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 54.125,
+      "learning_rate": 7.874839029025729e-07,
+      "loss": 1.7211,
+      "step": 11240
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 89.9375,
+      "learning_rate": 7.871057593717674e-07,
+      "loss": 1.7641,
+      "step": 11260
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 53.125,
+      "learning_rate": 7.867276158409618e-07,
+      "loss": 1.8629,
+      "step": 11280
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 79.1875,
+      "learning_rate": 7.863494723101562e-07,
+      "loss": 1.8001,
+      "step": 11300
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 59.9375,
+      "learning_rate": 7.859713287793507e-07,
+      "loss": 1.8619,
+      "step": 11320
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 71.75,
+      "learning_rate": 7.855931852485452e-07,
+      "loss": 1.813,
+      "step": 11340
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 99.9375,
+      "learning_rate": 7.852150417177397e-07,
+      "loss": 1.9216,
+      "step": 11360
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 237.25,
+      "learning_rate": 7.848368981869341e-07,
+      "loss": 1.8799,
+      "step": 11380
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 42.0,
+      "learning_rate": 7.844587546561285e-07,
+      "loss": 1.8169,
+      "step": 11400
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 58.875,
+      "learning_rate": 7.84080611125323e-07,
+      "loss": 1.7881,
+      "step": 11420
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 83.875,
+      "learning_rate": 7.837024675945175e-07,
+      "loss": 1.6689,
+      "step": 11440
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 87.5,
+      "learning_rate": 7.833243240637118e-07,
+      "loss": 1.7903,
+      "step": 11460
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 76.0,
+      "learning_rate": 7.829461805329063e-07,
+      "loss": 1.6204,
+      "step": 11480
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 60.65625,
+      "learning_rate": 7.825680370021008e-07,
+      "loss": 1.832,
+      "step": 11500
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 51.875,
+      "learning_rate": 7.821898934712951e-07,
+      "loss": 1.6985,
+      "step": 11520
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 78.125,
+      "learning_rate": 7.818117499404896e-07,
+      "loss": 1.7127,
+      "step": 11540
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 59.40625,
+      "learning_rate": 7.81433606409684e-07,
+      "loss": 1.7354,
+      "step": 11560
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 69.125,
+      "learning_rate": 7.810554628788785e-07,
+      "loss": 1.759,
+      "step": 11580
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 53.875,
+      "learning_rate": 7.806773193480729e-07,
+      "loss": 1.8113,
+      "step": 11600
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 69.0625,
+      "learning_rate": 7.802991758172673e-07,
+      "loss": 1.7437,
+      "step": 11620
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 61.21875,
+      "learning_rate": 7.799210322864618e-07,
+      "loss": 1.8023,
+      "step": 11640
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 53.5625,
+      "learning_rate": 7.795428887556562e-07,
+      "loss": 1.6904,
+      "step": 11660
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 106.25,
+      "learning_rate": 7.791647452248507e-07,
+      "loss": 1.746,
+      "step": 11680
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 55.75,
+      "learning_rate": 7.787866016940452e-07,
+      "loss": 1.7889,
+      "step": 11700
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 49.84375,
+      "learning_rate": 7.784084581632397e-07,
+      "loss": 1.7731,
+      "step": 11720
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 65.6875,
+      "learning_rate": 7.78030314632434e-07,
+      "loss": 1.7623,
+      "step": 11740
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 52.53125,
+      "learning_rate": 7.776521711016285e-07,
+      "loss": 1.8033,
+      "step": 11760
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 77.375,
+      "learning_rate": 7.77274027570823e-07,
+      "loss": 1.8242,
+      "step": 11780
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 75.0,
+      "learning_rate": 7.768958840400174e-07,
+      "loss": 1.6878,
+      "step": 11800
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 68.875,
+      "learning_rate": 7.765177405092118e-07,
+      "loss": 1.7688,
+      "step": 11820
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 70.375,
+      "learning_rate": 7.761395969784063e-07,
+      "loss": 1.7053,
+      "step": 11840
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 50.6875,
+      "learning_rate": 7.757614534476007e-07,
+      "loss": 1.7667,
+      "step": 11860
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 53.84375,
+      "learning_rate": 7.753833099167951e-07,
+      "loss": 1.8148,
+      "step": 11880
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 54.65625,
+      "learning_rate": 7.750051663859895e-07,
+      "loss": 1.758,
+      "step": 11900
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 63.5625,
+      "learning_rate": 7.74627022855184e-07,
+      "loss": 1.8429,
+      "step": 11920
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 58.875,
+      "learning_rate": 7.742488793243785e-07,
+      "loss": 1.7386,
+      "step": 11940
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 48.90625,
+      "learning_rate": 7.738707357935728e-07,
+      "loss": 1.8145,
+      "step": 11960
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 53.1875,
+      "learning_rate": 7.734925922627673e-07,
+      "loss": 1.7081,
+      "step": 11980
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 64.0625,
+      "learning_rate": 7.731144487319618e-07,
+      "loss": 1.7029,
+      "step": 12000
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 52.5625,
+      "learning_rate": 7.727363052011562e-07,
+      "loss": 1.8116,
+      "step": 12020
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 85.9375,
+      "learning_rate": 7.723581616703507e-07,
+      "loss": 1.7725,
+      "step": 12040
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 61.125,
+      "learning_rate": 7.719800181395452e-07,
+      "loss": 1.8037,
+      "step": 12060
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 55.75,
+      "learning_rate": 7.716018746087396e-07,
+      "loss": 1.7618,
+      "step": 12080
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 78.625,
+      "learning_rate": 7.71223731077934e-07,
+      "loss": 1.7472,
+      "step": 12100
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 87.625,
+      "learning_rate": 7.708455875471285e-07,
+      "loss": 1.7094,
+      "step": 12120
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 55.625,
+      "learning_rate": 7.704674440163229e-07,
+      "loss": 1.8509,
+      "step": 12140
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 62.1875,
+      "learning_rate": 7.700893004855174e-07,
+      "loss": 1.7445,
+      "step": 12160
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 68.25,
+      "learning_rate": 7.697111569547117e-07,
+      "loss": 1.8586,
+      "step": 12180
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 73.375,
+      "learning_rate": 7.693330134239062e-07,
+      "loss": 1.7869,
+      "step": 12200
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 63.71875,
+      "learning_rate": 7.689548698931007e-07,
+      "loss": 1.6798,
+      "step": 12220
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 50.625,
+      "learning_rate": 7.68576726362295e-07,
+      "loss": 1.7414,
+      "step": 12240
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 70.125,
+      "learning_rate": 7.681985828314895e-07,
+      "loss": 1.7452,
+      "step": 12260
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 60.8125,
+      "learning_rate": 7.67820439300684e-07,
+      "loss": 1.869,
+      "step": 12280
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 49.375,
+      "learning_rate": 7.674422957698784e-07,
+      "loss": 1.6938,
+      "step": 12300
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 83.0625,
+      "learning_rate": 7.670641522390728e-07,
+      "loss": 1.6603,
+      "step": 12320
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 69.5625,
+      "learning_rate": 7.666860087082673e-07,
+      "loss": 1.7059,
+      "step": 12340
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 51.0,
+      "learning_rate": 7.663078651774617e-07,
+      "loss": 1.7743,
+      "step": 12360
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 58.4375,
+      "learning_rate": 7.659297216466563e-07,
+      "loss": 1.8554,
+      "step": 12380
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 86.625,
+      "learning_rate": 7.655515781158507e-07,
+      "loss": 1.7817,
+      "step": 12400
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 53.125,
+      "learning_rate": 7.651734345850451e-07,
+      "loss": 1.7744,
+      "step": 12420
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 49.21875,
+      "learning_rate": 7.647952910542396e-07,
+      "loss": 1.8232,
+      "step": 12440
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 64.1875,
+      "learning_rate": 7.644171475234341e-07,
+      "loss": 1.7839,
+      "step": 12460
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 77.0625,
+      "learning_rate": 7.640390039926284e-07,
+      "loss": 1.8145,
+      "step": 12480
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 64.375,
+      "learning_rate": 7.636608604618229e-07,
+      "loss": 1.6772,
+      "step": 12500
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 74.5625,
+      "learning_rate": 7.632827169310173e-07,
+      "loss": 1.7555,
+      "step": 12520
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 96.4375,
+      "learning_rate": 7.629045734002117e-07,
+      "loss": 1.6831,
+      "step": 12540
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 55.40625,
+      "learning_rate": 7.625264298694062e-07,
+      "loss": 1.8384,
+      "step": 12560
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 50.25,
+      "learning_rate": 7.621482863386006e-07,
+      "loss": 1.7862,
+      "step": 12580
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 51.46875,
+      "learning_rate": 7.617701428077951e-07,
+      "loss": 1.7822,
+      "step": 12600
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 63.625,
+      "learning_rate": 7.613919992769895e-07,
+      "loss": 1.7557,
+      "step": 12620
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 74.3125,
+      "learning_rate": 7.610138557461839e-07,
+      "loss": 1.8421,
+      "step": 12640
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 58.34375,
+      "learning_rate": 7.606357122153784e-07,
+      "loss": 1.715,
+      "step": 12660
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 65.6875,
+      "learning_rate": 7.602575686845728e-07,
+      "loss": 1.8373,
+      "step": 12680
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 64.3125,
+      "learning_rate": 7.598794251537672e-07,
+      "loss": 1.7039,
+      "step": 12700
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 74.4375,
+      "learning_rate": 7.595012816229618e-07,
+      "loss": 1.7355,
+      "step": 12720
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 73.5,
+      "learning_rate": 7.591231380921563e-07,
+      "loss": 1.7805,
+      "step": 12740
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 71.1875,
+      "learning_rate": 7.587449945613506e-07,
+      "loss": 1.7684,
+      "step": 12760
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 77.8125,
+      "learning_rate": 7.583668510305451e-07,
+      "loss": 1.7547,
+      "step": 12780
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 60.9375,
+      "learning_rate": 7.579887074997395e-07,
+      "loss": 1.7124,
+      "step": 12800
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 68.9375,
+      "learning_rate": 7.57610563968934e-07,
+      "loss": 1.8401,
+      "step": 12820
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 47.625,
+      "learning_rate": 7.572324204381284e-07,
+      "loss": 1.786,
+      "step": 12840
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 56.40625,
+      "learning_rate": 7.568542769073228e-07,
+      "loss": 1.7334,
+      "step": 12860
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 54.21875,
+      "learning_rate": 7.564761333765173e-07,
+      "loss": 1.7812,
+      "step": 12880
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 57.3125,
+      "learning_rate": 7.560979898457117e-07,
+      "loss": 1.7766,
+      "step": 12900
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 48.34375,
+      "learning_rate": 7.557198463149061e-07,
+      "loss": 1.7633,
+      "step": 12920
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 66.5625,
+      "learning_rate": 7.553417027841006e-07,
+      "loss": 1.8936,
+      "step": 12940
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 72.4375,
+      "learning_rate": 7.549635592532951e-07,
+      "loss": 1.8073,
+      "step": 12960
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 57.3125,
+      "learning_rate": 7.545854157224894e-07,
+      "loss": 1.7529,
+      "step": 12980
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 86.0625,
+      "learning_rate": 7.542072721916839e-07,
+      "loss": 1.7475,
+      "step": 13000
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 77.625,
+      "learning_rate": 7.538291286608784e-07,
+      "loss": 1.7339,
+      "step": 13020
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 84.625,
+      "learning_rate": 7.534509851300727e-07,
+      "loss": 1.6812,
+      "step": 13040
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 58.34375,
+      "learning_rate": 7.530728415992672e-07,
+      "loss": 1.8598,
+      "step": 13060
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 47.34375,
+      "learning_rate": 7.526946980684618e-07,
+      "loss": 1.789,
+      "step": 13080
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 64.0,
+      "learning_rate": 7.523165545376562e-07,
+      "loss": 1.7233,
+      "step": 13100
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 71.5625,
+      "learning_rate": 7.519384110068506e-07,
+      "loss": 1.7302,
+      "step": 13120
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 63.90625,
+      "learning_rate": 7.51560267476045e-07,
+      "loss": 1.6977,
+      "step": 13140
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 47.09375,
+      "learning_rate": 7.511821239452395e-07,
+      "loss": 1.8249,
+      "step": 13160
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 52.75,
+      "learning_rate": 7.50803980414434e-07,
+      "loss": 1.7705,
+      "step": 13180
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 67.0,
+      "learning_rate": 7.504258368836283e-07,
+      "loss": 1.6708,
+      "step": 13200
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 82.5,
+      "learning_rate": 7.500476933528228e-07,
+      "loss": 1.8331,
+      "step": 13220
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 96.625,
+      "learning_rate": 7.496695498220173e-07,
+      "loss": 1.8325,
+      "step": 13240
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 56.3125,
+      "learning_rate": 7.492914062912116e-07,
+      "loss": 1.7972,
+      "step": 13260
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 56.53125,
+      "learning_rate": 7.489132627604061e-07,
+      "loss": 1.7719,
+      "step": 13280
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 87.875,
+      "learning_rate": 7.485351192296006e-07,
+      "loss": 1.7606,
+      "step": 13300
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 73.6875,
+      "learning_rate": 7.48156975698795e-07,
+      "loss": 1.7746,
+      "step": 13320
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 80.625,
+      "learning_rate": 7.477788321679894e-07,
+      "loss": 1.767,
+      "step": 13340
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 136.125,
+      "learning_rate": 7.474006886371839e-07,
+      "loss": 1.8342,
+      "step": 13360
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 58.78125,
+      "learning_rate": 7.470225451063783e-07,
+      "loss": 1.739,
+      "step": 13380
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 60.75,
+      "learning_rate": 7.466444015755728e-07,
+      "loss": 1.7303,
+      "step": 13400
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 78.375,
+      "learning_rate": 7.462662580447673e-07,
+      "loss": 1.6749,
+      "step": 13420
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 65.5,
+      "learning_rate": 7.458881145139617e-07,
+      "loss": 1.8015,
+      "step": 13440
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 65.5625,
+      "learning_rate": 7.455099709831562e-07,
+      "loss": 1.7586,
+      "step": 13460
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 48.46875,
+      "learning_rate": 7.451318274523505e-07,
+      "loss": 1.6745,
+      "step": 13480
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 76.75,
+      "learning_rate": 7.44753683921545e-07,
+      "loss": 1.8741,
+      "step": 13500
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 71.625,
+      "learning_rate": 7.443755403907395e-07,
+      "loss": 1.7672,
+      "step": 13520
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 76.9375,
+      "learning_rate": 7.439973968599339e-07,
+      "loss": 1.8212,
+      "step": 13540
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 59.1875,
+      "learning_rate": 7.436192533291283e-07,
+      "loss": 1.775,
+      "step": 13560
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 97.8125,
+      "learning_rate": 7.432411097983228e-07,
+      "loss": 1.8158,
+      "step": 13580
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 65.5,
+      "learning_rate": 7.428629662675172e-07,
+      "loss": 1.7197,
+      "step": 13600
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 71.8125,
+      "learning_rate": 7.424848227367117e-07,
+      "loss": 1.7534,
+      "step": 13620
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 83.3125,
+      "learning_rate": 7.421066792059061e-07,
+      "loss": 1.7694,
+      "step": 13640
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 65.75,
+      "learning_rate": 7.417285356751005e-07,
+      "loss": 1.831,
+      "step": 13660
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 102.5,
+      "learning_rate": 7.41350392144295e-07,
+      "loss": 1.7945,
+      "step": 13680
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 63.1875,
+      "learning_rate": 7.409722486134893e-07,
+      "loss": 1.8187,
+      "step": 13700
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 48.125,
+      "learning_rate": 7.405941050826838e-07,
+      "loss": 1.7061,
+      "step": 13720
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 63.78125,
+      "learning_rate": 7.402159615518783e-07,
+      "loss": 1.7076,
+      "step": 13740
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 44.59375,
+      "learning_rate": 7.398378180210727e-07,
+      "loss": 1.6653,
+      "step": 13760
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 62.8125,
+      "learning_rate": 7.394596744902672e-07,
+      "loss": 1.8329,
+      "step": 13780
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 77.25,
+      "learning_rate": 7.390815309594617e-07,
+      "loss": 1.8275,
+      "step": 13800
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 52.40625,
+      "learning_rate": 7.387033874286561e-07,
+      "loss": 1.7821,
+      "step": 13820
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 54.875,
+      "learning_rate": 7.383252438978506e-07,
+      "loss": 1.8171,
+      "step": 13840
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 68.375,
+      "learning_rate": 7.37947100367045e-07,
+      "loss": 1.7546,
+      "step": 13860
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 66.875,
+      "learning_rate": 7.375689568362394e-07,
+      "loss": 1.79,
+      "step": 13880
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 62.1875,
+      "learning_rate": 7.371908133054339e-07,
+      "loss": 1.88,
+      "step": 13900
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 77.5625,
+      "learning_rate": 7.368126697746283e-07,
+      "loss": 1.7613,
+      "step": 13920
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 58.21875,
+      "learning_rate": 7.364345262438227e-07,
+      "loss": 1.7634,
+      "step": 13940
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 55.0,
+      "learning_rate": 7.360563827130172e-07,
+      "loss": 1.8992,
+      "step": 13960
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 84.0625,
+      "learning_rate": 7.356782391822117e-07,
+      "loss": 1.7488,
+      "step": 13980
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 61.3125,
+      "learning_rate": 7.35300095651406e-07,
+      "loss": 1.7939,
+      "step": 14000
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 54.5625,
+      "learning_rate": 7.349219521206005e-07,
+      "loss": 1.7303,
+      "step": 14020
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 48.6875,
+      "learning_rate": 7.345438085897949e-07,
+      "loss": 1.7983,
+      "step": 14040
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 58.125,
+      "learning_rate": 7.341656650589893e-07,
+      "loss": 1.7608,
+      "step": 14060
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 74.75,
+      "learning_rate": 7.337875215281838e-07,
+      "loss": 1.8493,
+      "step": 14080
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 76.125,
+      "learning_rate": 7.334093779973782e-07,
+      "loss": 1.8708,
+      "step": 14100
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 54.34375,
+      "learning_rate": 7.330312344665728e-07,
+      "loss": 1.7848,
+      "step": 14120
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 82.1875,
+      "learning_rate": 7.326530909357672e-07,
+      "loss": 1.8407,
+      "step": 14140
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 125.0,
+      "learning_rate": 7.322749474049616e-07,
+      "loss": 1.8569,
+      "step": 14160
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 67.875,
+      "learning_rate": 7.318968038741561e-07,
+      "loss": 1.8067,
+      "step": 14180
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 76.375,
+      "learning_rate": 7.315186603433506e-07,
+      "loss": 1.7579,
+      "step": 14200
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 51.34375,
+      "learning_rate": 7.311405168125449e-07,
+      "loss": 1.7588,
+      "step": 14220
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 67.75,
+      "learning_rate": 7.307623732817394e-07,
+      "loss": 1.7899,
+      "step": 14240
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 62.34375,
+      "learning_rate": 7.303842297509339e-07,
+      "loss": 1.755,
+      "step": 14260
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 91.0,
+      "learning_rate": 7.300060862201282e-07,
+      "loss": 1.8114,
+      "step": 14280
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 64.9375,
+      "learning_rate": 7.296279426893227e-07,
+      "loss": 1.7733,
+      "step": 14300
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 99.375,
+      "learning_rate": 7.292497991585172e-07,
+      "loss": 1.8316,
+      "step": 14320
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 65.0,
+      "learning_rate": 7.288716556277116e-07,
+      "loss": 1.7512,
+      "step": 14340
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 76.0625,
+      "learning_rate": 7.28493512096906e-07,
+      "loss": 1.771,
+      "step": 14360
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 52.78125,
+      "learning_rate": 7.281153685661004e-07,
+      "loss": 1.7127,
+      "step": 14380
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 83.8125,
+      "learning_rate": 7.277372250352949e-07,
+      "loss": 1.836,
+      "step": 14400
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 79.6875,
+      "learning_rate": 7.273590815044894e-07,
+      "loss": 1.7365,
+      "step": 14420
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 56.96875,
+      "learning_rate": 7.269809379736837e-07,
+      "loss": 1.7652,
+      "step": 14440
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 56.59375,
+      "learning_rate": 7.266027944428782e-07,
+      "loss": 1.8393,
+      "step": 14460
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 51.40625,
+      "learning_rate": 7.262246509120728e-07,
+      "loss": 1.8095,
+      "step": 14480
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 62.125,
+      "learning_rate": 7.258465073812671e-07,
+      "loss": 1.7587,
+      "step": 14500
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 57.28125,
+      "learning_rate": 7.254683638504616e-07,
+      "loss": 1.6792,
+      "step": 14520
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 61.09375,
+      "learning_rate": 7.250902203196561e-07,
+      "loss": 1.7675,
+      "step": 14540
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 56.40625,
+      "learning_rate": 7.247120767888505e-07,
+      "loss": 1.8372,
+      "step": 14560
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 69.375,
+      "learning_rate": 7.243339332580449e-07,
+      "loss": 1.7053,
+      "step": 14580
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 103.375,
+      "learning_rate": 7.239557897272394e-07,
+      "loss": 1.7466,
+      "step": 14600
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 63.4375,
+      "learning_rate": 7.235776461964338e-07,
+      "loss": 1.7912,
+      "step": 14620
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 74.1875,
+      "learning_rate": 7.231995026656283e-07,
+      "loss": 1.7838,
+      "step": 14640
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 71.25,
+      "learning_rate": 7.228213591348226e-07,
+      "loss": 1.6906,
+      "step": 14660
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 72.75,
+      "learning_rate": 7.224432156040171e-07,
+      "loss": 1.8053,
+      "step": 14680
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 107.3125,
+      "learning_rate": 7.220650720732116e-07,
+      "loss": 1.8471,
+      "step": 14700
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 107.0,
+      "learning_rate": 7.216869285424059e-07,
+      "loss": 1.7873,
+      "step": 14720
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 98.0,
+      "learning_rate": 7.213087850116004e-07,
+      "loss": 1.728,
+      "step": 14740
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 62.3125,
+      "learning_rate": 7.209306414807949e-07,
+      "loss": 1.7931,
+      "step": 14760
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 77.625,
+      "learning_rate": 7.205524979499893e-07,
+      "loss": 1.8416,
+      "step": 14780
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 67.25,
+      "learning_rate": 7.201743544191837e-07,
+      "loss": 1.7702,
+      "step": 14800
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 73.5,
+      "learning_rate": 7.197962108883783e-07,
+      "loss": 1.751,
+      "step": 14820
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 90.6875,
+      "learning_rate": 7.194180673575727e-07,
+      "loss": 1.8144,
+      "step": 14840
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 65.8125,
+      "learning_rate": 7.190399238267672e-07,
+      "loss": 1.7848,
+      "step": 14860
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 75.1875,
+      "learning_rate": 7.186617802959616e-07,
+      "loss": 1.8424,
+      "step": 14880
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 69.5625,
+      "learning_rate": 7.18283636765156e-07,
+      "loss": 1.7273,
+      "step": 14900
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 62.9375,
+      "learning_rate": 7.179054932343505e-07,
+      "loss": 1.7568,
+      "step": 14920
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 49.75,
+      "learning_rate": 7.175273497035449e-07,
+      "loss": 1.8136,
+      "step": 14940
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 75.8125,
+      "learning_rate": 7.171492061727393e-07,
+      "loss": 1.8259,
+      "step": 14960
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 51.125,
+      "learning_rate": 7.167710626419338e-07,
+      "loss": 1.7377,
+      "step": 14980
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 61.8125,
+      "learning_rate": 7.163929191111282e-07,
+      "loss": 1.7435,
+      "step": 15000
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 45.65625,
+      "learning_rate": 7.160147755803226e-07,
+      "loss": 1.7305,
+      "step": 15020
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 48.03125,
+      "learning_rate": 7.156366320495171e-07,
+      "loss": 1.7079,
+      "step": 15040
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 85.0,
+      "learning_rate": 7.152584885187115e-07,
+      "loss": 1.8445,
+      "step": 15060
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 73.5625,
+      "learning_rate": 7.148803449879059e-07,
+      "loss": 1.7944,
+      "step": 15080
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 66.75,
+      "learning_rate": 7.145022014571004e-07,
+      "loss": 1.72,
+      "step": 15100
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 78.4375,
+      "learning_rate": 7.141240579262948e-07,
+      "loss": 1.8119,
+      "step": 15120
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 66.25,
+      "learning_rate": 7.137459143954893e-07,
+      "loss": 1.8503,
+      "step": 15140
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 78.625,
+      "learning_rate": 7.133677708646837e-07,
+      "loss": 1.7444,
+      "step": 15160
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 88.5625,
+      "learning_rate": 7.129896273338782e-07,
+      "loss": 1.7226,
+      "step": 15180
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 52.25,
+      "learning_rate": 7.126114838030727e-07,
+      "loss": 1.6872,
+      "step": 15200
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 129.375,
+      "learning_rate": 7.122333402722672e-07,
+      "loss": 1.7477,
+      "step": 15220
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 60.46875,
+      "learning_rate": 7.118551967414615e-07,
+      "loss": 1.6581,
+      "step": 15240
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 59.1875,
+      "learning_rate": 7.11477053210656e-07,
+      "loss": 1.7731,
+      "step": 15260
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 70.8125,
+      "learning_rate": 7.110989096798505e-07,
+      "loss": 1.7503,
+      "step": 15280
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 82.9375,
+      "learning_rate": 7.107207661490448e-07,
+      "loss": 1.6437,
+      "step": 15300
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 66.375,
+      "learning_rate": 7.103426226182393e-07,
+      "loss": 1.782,
+      "step": 15320
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 120.25,
+      "learning_rate": 7.099644790874337e-07,
+      "loss": 1.9366,
+      "step": 15340
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 71.875,
+      "learning_rate": 7.095863355566282e-07,
+      "loss": 1.8142,
+      "step": 15360
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 63.03125,
+      "learning_rate": 7.092081920258226e-07,
+      "loss": 1.8358,
+      "step": 15380
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 59.0,
+      "learning_rate": 7.08830048495017e-07,
+      "loss": 1.7758,
+      "step": 15400
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 65.0625,
+      "learning_rate": 7.084519049642115e-07,
+      "loss": 1.7326,
+      "step": 15420
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 59.875,
+      "learning_rate": 7.08073761433406e-07,
+      "loss": 1.7505,
+      "step": 15440
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 99.0,
+      "learning_rate": 7.076956179026003e-07,
+      "loss": 1.7379,
+      "step": 15460
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 79.5625,
+      "learning_rate": 7.073174743717948e-07,
+      "loss": 1.8941,
+      "step": 15480
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 70.375,
+      "learning_rate": 7.069393308409893e-07,
+      "loss": 1.8052,
+      "step": 15500
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 93.9375,
+      "learning_rate": 7.065611873101837e-07,
+      "loss": 1.8094,
+      "step": 15520
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 63.5625,
+      "learning_rate": 7.061830437793782e-07,
+      "loss": 1.7806,
+      "step": 15540
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 57.65625,
+      "learning_rate": 7.058049002485727e-07,
+      "loss": 1.7275,
+      "step": 15560
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 58.96875,
+      "learning_rate": 7.054267567177671e-07,
+      "loss": 1.6737,
+      "step": 15580
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 66.75,
+      "learning_rate": 7.050486131869615e-07,
+      "loss": 1.7801,
+      "step": 15600
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 48.59375,
+      "learning_rate": 7.046704696561559e-07,
+      "loss": 1.7297,
+      "step": 15620
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 69.6875,
+      "learning_rate": 7.042923261253504e-07,
+      "loss": 1.7964,
+      "step": 15640
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 49.28125,
+      "learning_rate": 7.039141825945449e-07,
+      "loss": 1.8027,
+      "step": 15660
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 55.46875,
+      "learning_rate": 7.035360390637392e-07,
+      "loss": 1.7836,
+      "step": 15680
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 80.5,
+      "learning_rate": 7.031578955329337e-07,
+      "loss": 1.7742,
+      "step": 15700
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 63.03125,
+      "learning_rate": 7.027797520021282e-07,
+      "loss": 1.7406,
+      "step": 15720
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 51.53125,
+      "learning_rate": 7.024016084713225e-07,
+      "loss": 1.7009,
+      "step": 15740
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 113.25,
+      "learning_rate": 7.02023464940517e-07,
+      "loss": 1.7417,
+      "step": 15760
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 58.96875,
+      "learning_rate": 7.016453214097115e-07,
+      "loss": 1.8331,
+      "step": 15780
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 76.3125,
+      "learning_rate": 7.012671778789059e-07,
+      "loss": 1.702,
+      "step": 15800
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 56.3125,
+      "learning_rate": 7.008890343481003e-07,
+      "loss": 1.8167,
+      "step": 15820
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 56.78125,
+      "learning_rate": 7.005108908172948e-07,
+      "loss": 1.7286,
+      "step": 15840
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 69.6875,
+      "learning_rate": 7.001327472864892e-07,
+      "loss": 1.7007,
+      "step": 15860
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 66.125,
+      "learning_rate": 6.997546037556838e-07,
+      "loss": 1.8052,
+      "step": 15880
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 57.15625,
+      "learning_rate": 6.993764602248782e-07,
+      "loss": 1.7057,
+      "step": 15900
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 69.0,
+      "learning_rate": 6.989983166940726e-07,
+      "loss": 1.8017,
+      "step": 15920
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 61.59375,
+      "learning_rate": 6.986201731632671e-07,
+      "loss": 1.8606,
+      "step": 15940
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 61.96875,
+      "learning_rate": 6.982420296324614e-07,
+      "loss": 1.793,
+      "step": 15960
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 72.0,
+      "learning_rate": 6.978638861016559e-07,
+      "loss": 1.7604,
+      "step": 15980
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 67.625,
+      "learning_rate": 6.974857425708504e-07,
+      "loss": 1.6722,
+      "step": 16000
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 62.40625,
+      "learning_rate": 6.971075990400448e-07,
+      "loss": 1.804,
+      "step": 16020
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 58.84375,
+      "learning_rate": 6.967294555092392e-07,
+      "loss": 1.7244,
+      "step": 16040
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 49.125,
+      "learning_rate": 6.963513119784337e-07,
+      "loss": 1.7763,
+      "step": 16060
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 67.1875,
+      "learning_rate": 6.959731684476281e-07,
+      "loss": 1.7623,
+      "step": 16080
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 62.1875,
+      "learning_rate": 6.955950249168225e-07,
+      "loss": 1.8598,
+      "step": 16100
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 82.6875,
+      "learning_rate": 6.95216881386017e-07,
+      "loss": 1.789,
+      "step": 16120
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 84.0,
+      "learning_rate": 6.948387378552114e-07,
+      "loss": 1.7023,
+      "step": 16140
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 92.4375,
+      "learning_rate": 6.944605943244059e-07,
+      "loss": 1.7511,
+      "step": 16160
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 69.125,
+      "learning_rate": 6.940824507936003e-07,
+      "loss": 1.8223,
+      "step": 16180
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 61.15625,
+      "learning_rate": 6.937043072627947e-07,
+      "loss": 1.8385,
+      "step": 16200
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 71.0,
+      "learning_rate": 6.933261637319893e-07,
+      "loss": 1.6804,
+      "step": 16220
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 57.4375,
+      "learning_rate": 6.929480202011837e-07,
+      "loss": 1.8056,
+      "step": 16240
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 78.875,
+      "learning_rate": 6.925698766703781e-07,
+      "loss": 1.8111,
+      "step": 16260
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 64.125,
+      "learning_rate": 6.921917331395726e-07,
+      "loss": 1.6361,
+      "step": 16280
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 75.125,
+      "learning_rate": 6.91813589608767e-07,
+      "loss": 1.7612,
+      "step": 16300
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 64.25,
+      "learning_rate": 6.914354460779614e-07,
+      "loss": 1.7896,
+      "step": 16320
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 58.96875,
+      "learning_rate": 6.910573025471559e-07,
+      "loss": 1.7171,
+      "step": 16340
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 56.71875,
+      "learning_rate": 6.906791590163503e-07,
+      "loss": 1.7889,
+      "step": 16360
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 53.3125,
+      "learning_rate": 6.903010154855448e-07,
+      "loss": 1.7833,
+      "step": 16380
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 60.9375,
+      "learning_rate": 6.899228719547392e-07,
+      "loss": 1.8044,
+      "step": 16400
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 64.8125,
+      "learning_rate": 6.895447284239336e-07,
+      "loss": 1.8078,
+      "step": 16420
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 69.375,
+      "learning_rate": 6.891665848931281e-07,
+      "loss": 1.69,
+      "step": 16440
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 52.5625,
+      "learning_rate": 6.887884413623226e-07,
+      "loss": 1.7741,
+      "step": 16460
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 50.21875,
+      "learning_rate": 6.884102978315169e-07,
+      "loss": 1.8409,
+      "step": 16480
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 55.34375,
+      "learning_rate": 6.880321543007114e-07,
+      "loss": 1.8224,
+      "step": 16500
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 58.6875,
+      "learning_rate": 6.876540107699058e-07,
+      "loss": 1.7456,
+      "step": 16520
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 52.8125,
+      "learning_rate": 6.872758672391002e-07,
+      "loss": 1.8105,
+      "step": 16540
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 81.6875,
+      "learning_rate": 6.868977237082947e-07,
+      "loss": 1.863,
+      "step": 16560
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 56.0625,
+      "learning_rate": 6.865195801774892e-07,
+      "loss": 1.717,
+      "step": 16580
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 41.25,
+      "learning_rate": 6.861414366466837e-07,
+      "loss": 1.9108,
+      "step": 16600
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 58.375,
+      "learning_rate": 6.857632931158781e-07,
+      "loss": 1.7419,
+      "step": 16620
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 56.625,
+      "learning_rate": 6.853851495850725e-07,
+      "loss": 1.861,
+      "step": 16640
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 62.6875,
+      "learning_rate": 6.85007006054267e-07,
+      "loss": 1.9066,
+      "step": 16660
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 52.15625,
+      "learning_rate": 6.846288625234614e-07,
+      "loss": 1.8078,
+      "step": 16680
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 70.9375,
+      "learning_rate": 6.842507189926558e-07,
+      "loss": 1.7877,
+      "step": 16700
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 71.875,
+      "learning_rate": 6.838725754618503e-07,
+      "loss": 1.7093,
+      "step": 16720
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 75.0,
+      "learning_rate": 6.834944319310448e-07,
+      "loss": 1.8284,
+      "step": 16740
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 73.6875,
+      "learning_rate": 6.831162884002391e-07,
+      "loss": 1.6915,
+      "step": 16760
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 66.625,
+      "learning_rate": 6.827381448694336e-07,
+      "loss": 1.8282,
+      "step": 16780
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 65.4375,
+      "learning_rate": 6.823600013386281e-07,
+      "loss": 1.7428,
+      "step": 16800
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 64.6875,
+      "learning_rate": 6.819818578078225e-07,
+      "loss": 1.7914,
+      "step": 16820
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 94.0,
+      "learning_rate": 6.816037142770169e-07,
+      "loss": 1.7644,
+      "step": 16840
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 61.6875,
+      "learning_rate": 6.812255707462113e-07,
+      "loss": 1.7501,
+      "step": 16860
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 70.625,
+      "learning_rate": 6.808474272154058e-07,
+      "loss": 1.7679,
+      "step": 16880
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 82.1875,
+      "learning_rate": 6.804692836846002e-07,
+      "loss": 1.7524,
+      "step": 16900
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 43.09375,
+      "learning_rate": 6.800911401537947e-07,
+      "loss": 1.7659,
+      "step": 16920
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 53.96875,
+      "learning_rate": 6.797129966229892e-07,
+      "loss": 1.6591,
+      "step": 16940
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 51.46875,
+      "learning_rate": 6.793348530921837e-07,
+      "loss": 1.6777,
+      "step": 16960
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 56.34375,
+      "learning_rate": 6.78956709561378e-07,
+      "loss": 1.7987,
+      "step": 16980
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 86.1875,
+      "learning_rate": 6.785785660305725e-07,
+      "loss": 1.8583,
+      "step": 17000
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 78.0625,
+      "learning_rate": 6.78200422499767e-07,
+      "loss": 1.8726,
+      "step": 17020
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 88.375,
+      "learning_rate": 6.778222789689614e-07,
+      "loss": 1.8256,
+      "step": 17040
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 54.96875,
+      "learning_rate": 6.774441354381558e-07,
+      "loss": 1.8137,
+      "step": 17060
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 55.0,
+      "learning_rate": 6.770659919073503e-07,
+      "loss": 1.6887,
+      "step": 17080
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 55.03125,
+      "learning_rate": 6.766878483765447e-07,
+      "loss": 1.807,
+      "step": 17100
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 67.875,
+      "learning_rate": 6.763097048457391e-07,
+      "loss": 1.7649,
+      "step": 17120
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 58.90625,
+      "learning_rate": 6.759315613149335e-07,
+      "loss": 1.7151,
+      "step": 17140
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 72.8125,
+      "learning_rate": 6.75553417784128e-07,
+      "loss": 1.8062,
+      "step": 17160
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 69.0,
+      "learning_rate": 6.751752742533225e-07,
+      "loss": 1.7687,
+      "step": 17180
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 66.5,
+      "learning_rate": 6.747971307225168e-07,
+      "loss": 1.7421,
+      "step": 17200
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 57.625,
+      "learning_rate": 6.744189871917113e-07,
+      "loss": 1.8331,
+      "step": 17220
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 60.1875,
+      "learning_rate": 6.740408436609058e-07,
+      "loss": 1.7498,
+      "step": 17240
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 51.65625,
+      "learning_rate": 6.736627001301001e-07,
+      "loss": 1.7102,
+      "step": 17260
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 79.625,
+      "learning_rate": 6.732845565992947e-07,
+      "loss": 1.714,
+      "step": 17280
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 52.46875,
+      "learning_rate": 6.729064130684892e-07,
+      "loss": 1.6849,
+      "step": 17300
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 61.3125,
+      "learning_rate": 6.725282695376836e-07,
+      "loss": 1.8046,
+      "step": 17320
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 52.0625,
+      "learning_rate": 6.72150126006878e-07,
+      "loss": 1.6312,
+      "step": 17340
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 54.71875,
+      "learning_rate": 6.717719824760725e-07,
+      "loss": 1.6886,
+      "step": 17360
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 51.375,
+      "learning_rate": 6.713938389452669e-07,
+      "loss": 1.754,
+      "step": 17380
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 55.90625,
+      "learning_rate": 6.710156954144614e-07,
+      "loss": 1.7469,
+      "step": 17400
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 61.78125,
+      "learning_rate": 6.706375518836558e-07,
+      "loss": 1.7955,
+      "step": 17420
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 89.75,
+      "learning_rate": 6.702594083528502e-07,
+      "loss": 1.7665,
+      "step": 17440
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 56.0625,
+      "learning_rate": 6.698812648220447e-07,
+      "loss": 1.8611,
+      "step": 17460
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 63.625,
+      "learning_rate": 6.69503121291239e-07,
+      "loss": 1.7571,
+      "step": 17480
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 70.375,
+      "learning_rate": 6.691249777604335e-07,
+      "loss": 1.8701,
+      "step": 17500
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 91.1875,
+      "learning_rate": 6.68746834229628e-07,
+      "loss": 1.7109,
+      "step": 17520
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 51.1875,
+      "learning_rate": 6.683686906988224e-07,
+      "loss": 1.7953,
+      "step": 17540
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 57.1875,
+      "learning_rate": 6.679905471680168e-07,
+      "loss": 1.6835,
+      "step": 17560
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 62.1875,
+      "learning_rate": 6.676124036372113e-07,
+      "loss": 1.8026,
+      "step": 17580
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 58.125,
+      "learning_rate": 6.672342601064057e-07,
+      "loss": 1.698,
+      "step": 17600
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 53.4375,
+      "learning_rate": 6.668561165756003e-07,
+      "loss": 1.7425,
+      "step": 17620
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 58.4375,
+      "learning_rate": 6.664779730447947e-07,
+      "loss": 1.8302,
+      "step": 17640
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 57.84375,
+      "learning_rate": 6.660998295139891e-07,
+      "loss": 1.8451,
+      "step": 17660
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 60.84375,
+      "learning_rate": 6.657216859831836e-07,
+      "loss": 1.7618,
+      "step": 17680
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 64.25,
+      "learning_rate": 6.65343542452378e-07,
+      "loss": 1.7403,
+      "step": 17700
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 81.0,
+      "learning_rate": 6.649653989215724e-07,
+      "loss": 1.7854,
+      "step": 17720
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 57.125,
+      "learning_rate": 6.645872553907669e-07,
+      "loss": 1.7188,
+      "step": 17740
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 75.1875,
+      "learning_rate": 6.642091118599614e-07,
+      "loss": 1.7599,
+      "step": 17760
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 62.75,
+      "learning_rate": 6.638309683291557e-07,
+      "loss": 1.7381,
+      "step": 17780
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 52.96875,
+      "learning_rate": 6.634528247983502e-07,
+      "loss": 1.7222,
+      "step": 17800
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 54.875,
+      "learning_rate": 6.630746812675446e-07,
+      "loss": 1.808,
+      "step": 17820
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 51.03125,
+      "learning_rate": 6.62696537736739e-07,
+      "loss": 1.599,
+      "step": 17840
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 59.09375,
+      "learning_rate": 6.623183942059335e-07,
+      "loss": 1.808,
+      "step": 17860
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 77.6875,
+      "learning_rate": 6.619402506751279e-07,
+      "loss": 1.7878,
+      "step": 17880
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 63.5625,
+      "learning_rate": 6.615621071443224e-07,
+      "loss": 1.7078,
+      "step": 17900
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 47.6875,
+      "learning_rate": 6.611839636135168e-07,
+      "loss": 1.7639,
+      "step": 17920
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 63.34375,
+      "learning_rate": 6.608058200827112e-07,
+      "loss": 1.8108,
+      "step": 17940
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 52.8125,
+      "learning_rate": 6.604276765519058e-07,
+      "loss": 1.8749,
+      "step": 17960
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 63.28125,
+      "learning_rate": 6.600495330211003e-07,
+      "loss": 1.8105,
+      "step": 17980
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 68.875,
+      "learning_rate": 6.596713894902946e-07,
+      "loss": 1.6754,
+      "step": 18000
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 88.0625,
+      "learning_rate": 6.592932459594891e-07,
+      "loss": 1.7055,
+      "step": 18020
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 69.5625,
+      "learning_rate": 6.589151024286836e-07,
+      "loss": 1.7658,
+      "step": 18040
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 48.65625,
+      "learning_rate": 6.58536958897878e-07,
+      "loss": 1.747,
+      "step": 18060
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 49.3125,
+      "learning_rate": 6.581588153670724e-07,
+      "loss": 1.8633,
+      "step": 18080
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 68.5625,
+      "learning_rate": 6.577806718362668e-07,
+      "loss": 1.7298,
+      "step": 18100
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 56.59375,
+      "learning_rate": 6.574025283054613e-07,
+      "loss": 1.6837,
+      "step": 18120
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 90.5625,
+      "learning_rate": 6.570243847746557e-07,
+      "loss": 1.7123,
+      "step": 18140
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 136.875,
+      "learning_rate": 6.566462412438501e-07,
+      "loss": 1.7742,
+      "step": 18160
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 65.0,
+      "learning_rate": 6.562680977130446e-07,
+      "loss": 1.7315,
+      "step": 18180
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 57.875,
+      "learning_rate": 6.558899541822391e-07,
+      "loss": 1.7881,
+      "step": 18200
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 61.90625,
+      "learning_rate": 6.555118106514334e-07,
+      "loss": 1.6969,
+      "step": 18220
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 65.8125,
+      "learning_rate": 6.551336671206279e-07,
+      "loss": 1.7261,
+      "step": 18240
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 59.03125,
+      "learning_rate": 6.547555235898224e-07,
+      "loss": 1.7779,
+      "step": 18260
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 52.09375,
+      "learning_rate": 6.543773800590167e-07,
+      "loss": 1.7624,
+      "step": 18280
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 79.9375,
+      "learning_rate": 6.539992365282112e-07,
+      "loss": 1.8038,
+      "step": 18300
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 85.0,
+      "learning_rate": 6.536210929974058e-07,
+      "loss": 1.8728,
+      "step": 18320
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 63.71875,
+      "learning_rate": 6.532429494666002e-07,
+      "loss": 1.8022,
+      "step": 18340
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 67.3125,
+      "learning_rate": 6.528648059357946e-07,
+      "loss": 1.7176,
+      "step": 18360
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 69.8125,
+      "learning_rate": 6.524866624049891e-07,
+      "loss": 1.8417,
+      "step": 18380
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 57.25,
+      "learning_rate": 6.521085188741835e-07,
+      "loss": 1.8464,
+      "step": 18400
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 56.71875,
+      "learning_rate": 6.51730375343378e-07,
+      "loss": 1.6192,
+      "step": 18420
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 79.1875,
+      "learning_rate": 6.513522318125723e-07,
+      "loss": 1.8753,
+      "step": 18440
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 53.375,
+      "learning_rate": 6.509740882817668e-07,
+      "loss": 1.6674,
+      "step": 18460
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 63.75,
+      "learning_rate": 6.505959447509613e-07,
+      "loss": 1.7284,
+      "step": 18480
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 72.5,
+      "learning_rate": 6.502178012201556e-07,
+      "loss": 1.8274,
+      "step": 18500
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 51.3125,
+      "learning_rate": 6.498396576893501e-07,
+      "loss": 1.7632,
+      "step": 18520
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 51.65625,
+      "learning_rate": 6.494615141585446e-07,
+      "loss": 1.7839,
+      "step": 18540
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 53.46875,
+      "learning_rate": 6.49083370627739e-07,
+      "loss": 1.6718,
+      "step": 18560
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 71.0625,
+      "learning_rate": 6.487052270969334e-07,
+      "loss": 1.6606,
+      "step": 18580
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 62.9375,
+      "learning_rate": 6.483270835661279e-07,
+      "loss": 1.8147,
+      "step": 18600
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 63.1875,
+      "learning_rate": 6.479489400353223e-07,
+      "loss": 1.8021,
+      "step": 18620
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 54.53125,
+      "learning_rate": 6.475707965045167e-07,
+      "loss": 1.6672,
+      "step": 18640
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 69.0,
+      "learning_rate": 6.471926529737113e-07,
+      "loss": 1.744,
+      "step": 18660
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 69.5625,
+      "learning_rate": 6.468145094429057e-07,
+      "loss": 1.7739,
+      "step": 18680
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 53.875,
+      "learning_rate": 6.464363659121002e-07,
+      "loss": 1.7764,
+      "step": 18700
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 54.0625,
+      "learning_rate": 6.460582223812946e-07,
+      "loss": 1.7612,
+      "step": 18720
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 64.4375,
+      "learning_rate": 6.45680078850489e-07,
+      "loss": 1.632,
+      "step": 18740
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 91.1875,
+      "learning_rate": 6.453019353196835e-07,
+      "loss": 1.8534,
+      "step": 18760
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 95.5625,
+      "learning_rate": 6.449237917888779e-07,
+      "loss": 1.8283,
+      "step": 18780
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 68.5625,
+      "learning_rate": 6.445456482580723e-07,
+      "loss": 1.7274,
+      "step": 18800
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 75.875,
+      "learning_rate": 6.441675047272668e-07,
+      "loss": 1.7398,
+      "step": 18820
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 73.5625,
+      "learning_rate": 6.437893611964612e-07,
+      "loss": 1.717,
+      "step": 18840
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 62.34375,
+      "learning_rate": 6.434112176656556e-07,
+      "loss": 1.7714,
+      "step": 18860
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 86.5625,
+      "learning_rate": 6.430330741348501e-07,
+      "loss": 1.8233,
+      "step": 18880
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 39.03125,
+      "learning_rate": 6.426549306040445e-07,
+      "loss": 1.7017,
+      "step": 18900
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 74.8125,
+      "learning_rate": 6.42276787073239e-07,
+      "loss": 1.7794,
+      "step": 18920
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 58.96875,
+      "learning_rate": 6.418986435424334e-07,
+      "loss": 1.752,
+      "step": 18940
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 56.6875,
+      "learning_rate": 6.415205000116278e-07,
+      "loss": 1.6723,
+      "step": 18960
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 76.3125,
+      "learning_rate": 6.411423564808223e-07,
+      "loss": 1.7992,
+      "step": 18980
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 46.125,
+      "learning_rate": 6.407642129500166e-07,
+      "loss": 1.8077,
+      "step": 19000
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 68.0,
+      "learning_rate": 6.403860694192112e-07,
+      "loss": 1.6902,
+      "step": 19020
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 66.4375,
+      "learning_rate": 6.400079258884057e-07,
+      "loss": 1.7471,
+      "step": 19040
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 72.375,
+      "learning_rate": 6.396297823576001e-07,
+      "loss": 1.7479,
+      "step": 19060
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 109.5625,
+      "learning_rate": 6.392516388267945e-07,
+      "loss": 1.7985,
+      "step": 19080
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 74.625,
+      "learning_rate": 6.38873495295989e-07,
+      "loss": 1.7164,
+      "step": 19100
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 65.875,
+      "learning_rate": 6.384953517651834e-07,
+      "loss": 1.6436,
+      "step": 19120
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 61.3125,
+      "learning_rate": 6.381172082343779e-07,
+      "loss": 1.6285,
+      "step": 19140
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 71.25,
+      "learning_rate": 6.377390647035723e-07,
+      "loss": 1.7548,
+      "step": 19160
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 57.875,
+      "learning_rate": 6.373609211727667e-07,
+      "loss": 1.6347,
+      "step": 19180
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 46.90625,
+      "learning_rate": 6.369827776419612e-07,
+      "loss": 1.7618,
+      "step": 19200
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 54.84375,
+      "learning_rate": 6.366046341111557e-07,
+      "loss": 1.7912,
+      "step": 19220
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 55.0625,
+      "learning_rate": 6.3622649058035e-07,
+      "loss": 1.7301,
+      "step": 19240
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 48.1875,
+      "learning_rate": 6.358483470495445e-07,
+      "loss": 1.7273,
+      "step": 19260
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 76.125,
+      "learning_rate": 6.35470203518739e-07,
+      "loss": 1.7014,
+      "step": 19280
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 49.40625,
+      "learning_rate": 6.350920599879333e-07,
+      "loss": 1.8441,
+      "step": 19300
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 76.75,
+      "learning_rate": 6.347139164571278e-07,
+      "loss": 1.7316,
+      "step": 19320
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 78.25,
+      "learning_rate": 6.343357729263222e-07,
+      "loss": 1.7168,
+      "step": 19340
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 68.0,
+      "learning_rate": 6.339576293955168e-07,
+      "loss": 1.8219,
+      "step": 19360
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 55.75,
+      "learning_rate": 6.335794858647112e-07,
+      "loss": 1.7437,
+      "step": 19380
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 56.90625,
+      "learning_rate": 6.332013423339056e-07,
+      "loss": 1.7404,
+      "step": 19400
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 51.125,
+      "learning_rate": 6.328231988031001e-07,
+      "loss": 1.731,
+      "step": 19420
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 59.0625,
+      "learning_rate": 6.324450552722946e-07,
+      "loss": 1.7489,
+      "step": 19440
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 93.5625,
+      "learning_rate": 6.320669117414889e-07,
+      "loss": 1.7179,
+      "step": 19460
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 77.3125,
+      "learning_rate": 6.316887682106834e-07,
+      "loss": 1.7455,
+      "step": 19480
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 60.21875,
+      "learning_rate": 6.313106246798779e-07,
+      "loss": 1.7477,
+      "step": 19500
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 67.875,
+      "learning_rate": 6.309324811490722e-07,
+      "loss": 1.8577,
+      "step": 19520
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 84.9375,
+      "learning_rate": 6.305543376182667e-07,
+      "loss": 1.8332,
+      "step": 19540
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 69.3125,
+      "learning_rate": 6.301761940874612e-07,
+      "loss": 1.7734,
+      "step": 19560
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 55.9375,
+      "learning_rate": 6.297980505566556e-07,
+      "loss": 1.7441,
+      "step": 19580
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 51.25,
+      "learning_rate": 6.2941990702585e-07,
+      "loss": 1.7063,
+      "step": 19600
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 62.21875,
+      "learning_rate": 6.290417634950445e-07,
+      "loss": 1.8064,
+      "step": 19620
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 82.3125,
+      "learning_rate": 6.286636199642389e-07,
+      "loss": 1.9288,
+      "step": 19640
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 97.4375,
+      "learning_rate": 6.282854764334333e-07,
+      "loss": 1.7964,
+      "step": 19660
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 71.5,
+      "learning_rate": 6.279073329026277e-07,
+      "loss": 1.7302,
+      "step": 19680
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 54.21875,
+      "learning_rate": 6.275291893718222e-07,
+      "loss": 1.7965,
+      "step": 19700
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 68.4375,
+      "learning_rate": 6.271510458410168e-07,
+      "loss": 1.735,
+      "step": 19720
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 65.0625,
+      "learning_rate": 6.267729023102111e-07,
+      "loss": 1.7821,
+      "step": 19740
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 77.1875,
+      "learning_rate": 6.263947587794056e-07,
+      "loss": 1.6467,
+      "step": 19760
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 92.875,
+      "learning_rate": 6.260166152486001e-07,
+      "loss": 1.8485,
+      "step": 19780
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 43.9375,
+      "learning_rate": 6.256384717177945e-07,
+      "loss": 1.7091,
+      "step": 19800
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 53.25,
+      "learning_rate": 6.252603281869889e-07,
+      "loss": 1.8332,
+      "step": 19820
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 48.28125,
+      "learning_rate": 6.248821846561834e-07,
+      "loss": 1.8508,
+      "step": 19840
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 53.46875,
+      "learning_rate": 6.245040411253778e-07,
+      "loss": 1.744,
+      "step": 19860
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 64.9375,
+      "learning_rate": 6.241258975945722e-07,
+      "loss": 1.7944,
+      "step": 19880
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 59.375,
+      "learning_rate": 6.237477540637667e-07,
+      "loss": 1.7726,
+      "step": 19900
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 45.1875,
+      "learning_rate": 6.233696105329611e-07,
+      "loss": 1.778,
+      "step": 19920
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 49.9375,
+      "learning_rate": 6.229914670021556e-07,
+      "loss": 1.7223,
+      "step": 19940
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 69.0625,
+      "learning_rate": 6.226133234713499e-07,
+      "loss": 1.807,
+      "step": 19960
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 45.78125,
+      "learning_rate": 6.222351799405444e-07,
+      "loss": 1.8191,
+      "step": 19980
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 62.15625,
+      "learning_rate": 6.218570364097389e-07,
+      "loss": 1.8354,
+      "step": 20000
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 75.625,
+      "learning_rate": 6.214788928789332e-07,
+      "loss": 1.7356,
+      "step": 20020
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 55.34375,
+      "learning_rate": 6.211007493481277e-07,
+      "loss": 1.7371,
+      "step": 20040
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 57.78125,
+      "learning_rate": 6.207226058173223e-07,
+      "loss": 1.7965,
+      "step": 20060
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 110.1875,
+      "learning_rate": 6.203444622865167e-07,
+      "loss": 1.6851,
+      "step": 20080
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 74.1875,
+      "learning_rate": 6.199663187557111e-07,
+      "loss": 1.7433,
+      "step": 20100
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 54.46875,
+      "learning_rate": 6.195881752249056e-07,
+      "loss": 1.7359,
+      "step": 20120
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 80.375,
+      "learning_rate": 6.192100316941e-07,
+      "loss": 1.7647,
+      "step": 20140
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 67.625,
+      "learning_rate": 6.188318881632945e-07,
+      "loss": 1.6704,
+      "step": 20160
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 78.1875,
+      "learning_rate": 6.184537446324889e-07,
+      "loss": 1.8029,
+      "step": 20180
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 55.59375,
+      "learning_rate": 6.180756011016833e-07,
+      "loss": 1.7328,
+      "step": 20200
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 75.9375,
+      "learning_rate": 6.176974575708778e-07,
+      "loss": 1.7889,
+      "step": 20220
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 51.8125,
+      "learning_rate": 6.173193140400723e-07,
+      "loss": 1.8524,
+      "step": 20240
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 52.75,
+      "learning_rate": 6.169411705092666e-07,
+      "loss": 1.8042,
+      "step": 20260
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 51.3125,
+      "learning_rate": 6.165630269784611e-07,
+      "loss": 1.8007,
+      "step": 20280
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 70.8125,
+      "learning_rate": 6.161848834476555e-07,
+      "loss": 1.7286,
+      "step": 20300
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 61.34375,
+      "learning_rate": 6.158067399168499e-07,
+      "loss": 1.7664,
+      "step": 20320
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 62.375,
+      "learning_rate": 6.154285963860444e-07,
+      "loss": 1.786,
+      "step": 20340
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 81.0625,
+      "learning_rate": 6.150504528552388e-07,
+      "loss": 1.7736,
+      "step": 20360
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 63.65625,
+      "learning_rate": 6.146723093244333e-07,
+      "loss": 1.8188,
+      "step": 20380
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 56.34375,
+      "learning_rate": 6.142941657936277e-07,
+      "loss": 1.8596,
+      "step": 20400
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 59.5625,
+      "learning_rate": 6.139160222628222e-07,
+      "loss": 1.7769,
+      "step": 20420
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 79.875,
+      "learning_rate": 6.135378787320167e-07,
+      "loss": 1.7049,
+      "step": 20440
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 49.78125,
+      "learning_rate": 6.131597352012112e-07,
+      "loss": 1.7867,
+      "step": 20460
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 51.03125,
+      "learning_rate": 6.127815916704055e-07,
+      "loss": 1.7805,
+      "step": 20480
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 65.25,
+      "learning_rate": 6.124034481396e-07,
+      "loss": 1.7941,
+      "step": 20500
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 52.46875,
+      "learning_rate": 6.120253046087945e-07,
+      "loss": 1.7872,
+      "step": 20520
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 94.625,
+      "learning_rate": 6.116471610779888e-07,
+      "loss": 1.8097,
+      "step": 20540
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 50.21875,
+      "learning_rate": 6.112690175471833e-07,
+      "loss": 1.7045,
+      "step": 20560
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 72.25,
+      "learning_rate": 6.108908740163778e-07,
+      "loss": 1.7775,
+      "step": 20580
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 68.625,
+      "learning_rate": 6.105127304855722e-07,
+      "loss": 1.7464,
+      "step": 20600
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 50.8125,
+      "learning_rate": 6.101345869547666e-07,
+      "loss": 1.7768,
+      "step": 20620
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 51.28125,
+      "learning_rate": 6.09756443423961e-07,
+      "loss": 1.7388,
+      "step": 20640
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 101.5,
+      "learning_rate": 6.093782998931555e-07,
+      "loss": 1.7464,
+      "step": 20660
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 77.5625,
+      "learning_rate": 6.090001563623499e-07,
+      "loss": 1.7987,
+      "step": 20680
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 111.0625,
+      "learning_rate": 6.086220128315443e-07,
+      "loss": 1.6151,
+      "step": 20700
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 74.5625,
+      "learning_rate": 6.082438693007388e-07,
+      "loss": 1.8307,
+      "step": 20720
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 98.375,
+      "learning_rate": 6.078657257699333e-07,
+      "loss": 1.8152,
+      "step": 20740
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 60.25,
+      "learning_rate": 6.074875822391277e-07,
+      "loss": 1.753,
+      "step": 20760
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 72.75,
+      "learning_rate": 6.071094387083222e-07,
+      "loss": 1.8042,
+      "step": 20780
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 52.25,
+      "learning_rate": 6.067312951775167e-07,
+      "loss": 1.8273,
+      "step": 20800
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 48.28125,
+      "learning_rate": 6.063531516467111e-07,
+      "loss": 1.7648,
+      "step": 20820
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 59.53125,
+      "learning_rate": 6.059750081159055e-07,
+      "loss": 1.6874,
+      "step": 20840
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 54.3125,
+      "learning_rate": 6.055968645851e-07,
+      "loss": 1.6971,
+      "step": 20860
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 76.0625,
+      "learning_rate": 6.052187210542944e-07,
+      "loss": 1.7653,
+      "step": 20880
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 51.34375,
+      "learning_rate": 6.048405775234888e-07,
+      "loss": 1.7038,
+      "step": 20900
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 58.09375,
+      "learning_rate": 6.044624339926832e-07,
+      "loss": 1.8158,
+      "step": 20920
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 54.375,
+      "learning_rate": 6.040842904618777e-07,
+      "loss": 1.8113,
+      "step": 20940
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 87.3125,
+      "learning_rate": 6.037061469310722e-07,
+      "loss": 1.7534,
+      "step": 20960
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 58.4375,
+      "learning_rate": 6.033280034002665e-07,
+      "loss": 1.7212,
+      "step": 20980
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 94.4375,
+      "learning_rate": 6.02949859869461e-07,
+      "loss": 1.745,
+      "step": 21000
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 49.4375,
+      "learning_rate": 6.025717163386555e-07,
+      "loss": 1.7588,
+      "step": 21020
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 70.8125,
+      "learning_rate": 6.021935728078498e-07,
+      "loss": 1.8333,
+      "step": 21040
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 77.375,
+      "learning_rate": 6.018154292770443e-07,
+      "loss": 1.6461,
+      "step": 21060
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 65.1875,
+      "learning_rate": 6.014372857462388e-07,
+      "loss": 1.7415,
+      "step": 21080
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 56.25,
+      "learning_rate": 6.010591422154332e-07,
+      "loss": 1.7088,
+      "step": 21100
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 60.84375,
+      "learning_rate": 6.006809986846277e-07,
+      "loss": 1.6742,
+      "step": 21120
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 62.03125,
+      "learning_rate": 6.003028551538222e-07,
+      "loss": 1.741,
+      "step": 21140
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 48.9375,
+      "learning_rate": 5.999247116230166e-07,
+      "loss": 1.832,
+      "step": 21160
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 54.46875,
+      "learning_rate": 5.995465680922111e-07,
+      "loss": 1.5626,
+      "step": 21180
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 75.375,
+      "learning_rate": 5.991684245614055e-07,
+      "loss": 1.6111,
+      "step": 21200
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 63.40625,
+      "learning_rate": 5.987902810305999e-07,
+      "loss": 1.6103,
+      "step": 21220
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 44.34375,
+      "learning_rate": 5.984121374997944e-07,
+      "loss": 1.5554,
+      "step": 21240
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 75.3125,
+      "learning_rate": 5.980339939689887e-07,
+      "loss": 1.5105,
+      "step": 21260
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 63.15625,
+      "learning_rate": 5.976558504381832e-07,
+      "loss": 1.5628,
+      "step": 21280
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 59.46875,
+      "learning_rate": 5.972777069073777e-07,
+      "loss": 1.5626,
+      "step": 21300
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 51.15625,
+      "learning_rate": 5.968995633765721e-07,
+      "loss": 1.5411,
+      "step": 21320
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 59.5625,
+      "learning_rate": 5.965214198457665e-07,
+      "loss": 1.613,
+      "step": 21340
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 46.28125,
+      "learning_rate": 5.96143276314961e-07,
+      "loss": 1.5999,
+      "step": 21360
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 53.96875,
+      "learning_rate": 5.957651327841554e-07,
+      "loss": 1.7679,
+      "step": 21380
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 46.75,
+      "learning_rate": 5.953869892533499e-07,
+      "loss": 1.5461,
+      "step": 21400
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 56.6875,
+      "learning_rate": 5.950088457225443e-07,
+      "loss": 1.5189,
+      "step": 21420
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 90.75,
+      "learning_rate": 5.946307021917387e-07,
+      "loss": 1.5582,
+      "step": 21440
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 76.0,
+      "learning_rate": 5.942525586609333e-07,
+      "loss": 1.4733,
+      "step": 21460
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 51.59375,
+      "learning_rate": 5.938744151301278e-07,
+      "loss": 1.6764,
+      "step": 21480
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 116.375,
+      "learning_rate": 5.934962715993221e-07,
+      "loss": 1.5825,
+      "step": 21500
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 57.78125,
+      "learning_rate": 5.931181280685166e-07,
+      "loss": 1.5275,
+      "step": 21520
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 56.53125,
+      "learning_rate": 5.92739984537711e-07,
+      "loss": 1.519,
+      "step": 21540
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 61.125,
+      "learning_rate": 5.923618410069054e-07,
+      "loss": 1.5093,
+      "step": 21560
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 65.625,
+      "learning_rate": 5.919836974760999e-07,
+      "loss": 1.5351,
+      "step": 21580
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 53.6875,
+      "learning_rate": 5.916055539452943e-07,
+      "loss": 1.585,
+      "step": 21600
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 60.75,
+      "learning_rate": 5.912274104144888e-07,
+      "loss": 1.5854,
+      "step": 21620
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 57.4375,
+      "learning_rate": 5.908492668836832e-07,
+      "loss": 1.4904,
+      "step": 21640
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 72.5625,
+      "learning_rate": 5.904711233528776e-07,
+      "loss": 1.5939,
+      "step": 21660
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 73.1875,
+      "learning_rate": 5.900929798220721e-07,
+      "loss": 1.4732,
+      "step": 21680
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 68.0625,
+      "learning_rate": 5.897148362912665e-07,
+      "loss": 1.4965,
+      "step": 21700
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 61.15625,
+      "learning_rate": 5.893366927604609e-07,
+      "loss": 1.6585,
+      "step": 21720
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 58.8125,
+      "learning_rate": 5.889585492296554e-07,
+      "loss": 1.65,
+      "step": 21740
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 50.3125,
+      "learning_rate": 5.885804056988499e-07,
+      "loss": 1.6261,
+      "step": 21760
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 59.09375,
+      "learning_rate": 5.882022621680442e-07,
+      "loss": 1.5723,
+      "step": 21780
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 61.09375,
+      "learning_rate": 5.878241186372387e-07,
+      "loss": 1.567,
+      "step": 21800
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 57.40625,
+      "learning_rate": 5.874459751064333e-07,
+      "loss": 1.4886,
+      "step": 21820
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 82.5625,
+      "learning_rate": 5.870678315756277e-07,
+      "loss": 1.5157,
+      "step": 21840
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 64.875,
+      "learning_rate": 5.866896880448221e-07,
+      "loss": 1.4767,
+      "step": 21860
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 66.6875,
+      "learning_rate": 5.863115445140165e-07,
+      "loss": 1.6021,
+      "step": 21880
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 80.5,
+      "learning_rate": 5.85933400983211e-07,
+      "loss": 1.5284,
+      "step": 21900
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 61.84375,
+      "learning_rate": 5.855552574524054e-07,
+      "loss": 1.5389,
+      "step": 21920
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 69.8125,
+      "learning_rate": 5.851771139215998e-07,
+      "loss": 1.4867,
+      "step": 21940
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 43.875,
+      "learning_rate": 5.847989703907943e-07,
+      "loss": 1.5887,
+      "step": 21960
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 64.5625,
+      "learning_rate": 5.844208268599888e-07,
+      "loss": 1.5928,
+      "step": 21980
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 68.5625,
+      "learning_rate": 5.840426833291831e-07,
+      "loss": 1.6589,
+      "step": 22000
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 71.1875,
+      "learning_rate": 5.836645397983776e-07,
+      "loss": 1.6092,
+      "step": 22020
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 54.75,
+      "learning_rate": 5.832863962675721e-07,
+      "loss": 1.503,
+      "step": 22040
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 67.3125,
+      "learning_rate": 5.829082527367664e-07,
+      "loss": 1.5614,
+      "step": 22060
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 68.0,
+      "learning_rate": 5.825301092059609e-07,
+      "loss": 1.6066,
+      "step": 22080
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 93.25,
+      "learning_rate": 5.821519656751554e-07,
+      "loss": 1.5409,
+      "step": 22100
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 77.6875,
+      "learning_rate": 5.817738221443498e-07,
+      "loss": 1.6052,
+      "step": 22120
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 79.75,
+      "learning_rate": 5.813956786135442e-07,
+      "loss": 1.6158,
+      "step": 22140
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 61.75,
+      "learning_rate": 5.810175350827388e-07,
+      "loss": 1.5924,
+      "step": 22160
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 87.375,
+      "learning_rate": 5.806393915519332e-07,
+      "loss": 1.608,
+      "step": 22180
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 56.46875,
+      "learning_rate": 5.802612480211277e-07,
+      "loss": 1.554,
+      "step": 22200
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 64.375,
+      "learning_rate": 5.79883104490322e-07,
+      "loss": 1.4941,
+      "step": 22220
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 68.625,
+      "learning_rate": 5.795049609595165e-07,
+      "loss": 1.6285,
+      "step": 22240
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 86.6875,
+      "learning_rate": 5.79126817428711e-07,
+      "loss": 1.586,
+      "step": 22260
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 45.625,
+      "learning_rate": 5.787486738979053e-07,
+      "loss": 1.5519,
+      "step": 22280
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 66.1875,
+      "learning_rate": 5.783705303670998e-07,
+      "loss": 1.553,
+      "step": 22300
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 63.15625,
+      "learning_rate": 5.779923868362943e-07,
+      "loss": 1.4016,
+      "step": 22320
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 64.0625,
+      "learning_rate": 5.776142433054887e-07,
+      "loss": 1.5565,
+      "step": 22340
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 118.625,
+      "learning_rate": 5.772360997746831e-07,
+      "loss": 1.5126,
+      "step": 22360
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 61.6875,
+      "learning_rate": 5.768579562438776e-07,
+      "loss": 1.6061,
+      "step": 22380
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 53.6875,
+      "learning_rate": 5.76479812713072e-07,
+      "loss": 1.5049,
+      "step": 22400
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 53.5625,
+      "learning_rate": 5.761016691822665e-07,
+      "loss": 1.5556,
+      "step": 22420
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 84.125,
+      "learning_rate": 5.757235256514608e-07,
+      "loss": 1.5172,
+      "step": 22440
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 57.125,
+      "learning_rate": 5.753453821206553e-07,
+      "loss": 1.5251,
+      "step": 22460
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 52.0,
+      "learning_rate": 5.749672385898498e-07,
+      "loss": 1.4947,
+      "step": 22480
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 71.9375,
+      "learning_rate": 5.745890950590441e-07,
+      "loss": 1.496,
+      "step": 22500
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 56.28125,
+      "learning_rate": 5.742109515282387e-07,
+      "loss": 1.6366,
+      "step": 22520
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 72.5625,
+      "learning_rate": 5.738328079974332e-07,
+      "loss": 1.5916,
+      "step": 22540
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 66.4375,
+      "learning_rate": 5.734546644666276e-07,
+      "loss": 1.572,
+      "step": 22560
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 47.25,
+      "learning_rate": 5.73076520935822e-07,
+      "loss": 1.5181,
+      "step": 22580
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 62.75,
+      "learning_rate": 5.726983774050165e-07,
+      "loss": 1.6111,
+      "step": 22600
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 51.375,
+      "learning_rate": 5.723202338742109e-07,
+      "loss": 1.5621,
+      "step": 22620
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 50.84375,
+      "learning_rate": 5.719420903434054e-07,
+      "loss": 1.5096,
+      "step": 22640
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 61.875,
+      "learning_rate": 5.715639468125998e-07,
+      "loss": 1.5789,
+      "step": 22660
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 81.5,
+      "learning_rate": 5.711858032817942e-07,
+      "loss": 1.5334,
+      "step": 22680
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 68.9375,
+      "learning_rate": 5.708076597509887e-07,
+      "loss": 1.6067,
+      "step": 22700
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 74.8125,
+      "learning_rate": 5.704295162201831e-07,
+      "loss": 1.6537,
+      "step": 22720
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 52.3125,
+      "learning_rate": 5.700513726893775e-07,
+      "loss": 1.4958,
+      "step": 22740
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 57.9375,
+      "learning_rate": 5.69673229158572e-07,
+      "loss": 1.6364,
+      "step": 22760
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 104.6875,
+      "learning_rate": 5.692950856277664e-07,
+      "loss": 1.5181,
+      "step": 22780
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 64.6875,
+      "learning_rate": 5.689169420969608e-07,
+      "loss": 1.6591,
+      "step": 22800
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 91.4375,
+      "learning_rate": 5.685387985661553e-07,
+      "loss": 1.5405,
+      "step": 22820
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 63.3125,
+      "learning_rate": 5.681606550353497e-07,
+      "loss": 1.4792,
+      "step": 22840
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 84.125,
+      "learning_rate": 5.677825115045442e-07,
+      "loss": 1.4895,
+      "step": 22860
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 62.0,
+      "learning_rate": 5.674043679737387e-07,
+      "loss": 1.5591,
+      "step": 22880
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 63.9375,
+      "learning_rate": 5.670262244429331e-07,
+      "loss": 1.599,
+      "step": 22900
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 66.8125,
+      "learning_rate": 5.666480809121276e-07,
+      "loss": 1.6139,
+      "step": 22920
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 109.25,
+      "learning_rate": 5.66269937381322e-07,
+      "loss": 1.5495,
+      "step": 22940
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 84.1875,
+      "learning_rate": 5.658917938505164e-07,
+      "loss": 1.6005,
+      "step": 22960
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 57.125,
+      "learning_rate": 5.655136503197109e-07,
+      "loss": 1.5384,
+      "step": 22980
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 73.25,
+      "learning_rate": 5.651355067889054e-07,
+      "loss": 1.5543,
+      "step": 23000
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 52.125,
+      "learning_rate": 5.647573632580997e-07,
+      "loss": 1.5076,
+      "step": 23020
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 52.6875,
+      "learning_rate": 5.643792197272942e-07,
+      "loss": 1.5428,
+      "step": 23040
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 76.25,
+      "learning_rate": 5.640010761964887e-07,
+      "loss": 1.5638,
+      "step": 23060
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 48.90625,
+      "learning_rate": 5.63622932665683e-07,
+      "loss": 1.5839,
+      "step": 23080
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 74.0,
+      "learning_rate": 5.632447891348775e-07,
+      "loss": 1.5441,
+      "step": 23100
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 61.90625,
+      "learning_rate": 5.628666456040719e-07,
+      "loss": 1.5854,
+      "step": 23120
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 71.25,
+      "learning_rate": 5.624885020732664e-07,
+      "loss": 1.5531,
+      "step": 23140
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 110.75,
+      "learning_rate": 5.621103585424608e-07,
+      "loss": 1.5323,
+      "step": 23160
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 73.375,
+      "learning_rate": 5.617322150116552e-07,
+      "loss": 1.572,
+      "step": 23180
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 88.25,
+      "learning_rate": 5.613540714808498e-07,
+      "loss": 1.4873,
+      "step": 23200
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 70.75,
+      "learning_rate": 5.609759279500443e-07,
+      "loss": 1.545,
+      "step": 23220
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 57.40625,
+      "learning_rate": 5.605977844192386e-07,
+      "loss": 1.5191,
+      "step": 23240
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 65.4375,
+      "learning_rate": 5.602196408884331e-07,
+      "loss": 1.5962,
+      "step": 23260
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 58.25,
+      "learning_rate": 5.598414973576276e-07,
+      "loss": 1.4449,
+      "step": 23280
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 99.125,
+      "learning_rate": 5.594633538268219e-07,
+      "loss": 1.4933,
+      "step": 23300
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 62.875,
+      "learning_rate": 5.590852102960164e-07,
+      "loss": 1.5198,
+      "step": 23320
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 72.0,
+      "learning_rate": 5.587070667652109e-07,
+      "loss": 1.6208,
+      "step": 23340
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 60.0,
+      "learning_rate": 5.583289232344053e-07,
+      "loss": 1.5792,
+      "step": 23360
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 54.84375,
+      "learning_rate": 5.579507797035997e-07,
+      "loss": 1.5194,
+      "step": 23380
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 57.84375,
+      "learning_rate": 5.575726361727941e-07,
+      "loss": 1.5103,
+      "step": 23400
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 52.21875,
+      "learning_rate": 5.571944926419886e-07,
+      "loss": 1.5909,
+      "step": 23420
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 90.5,
+      "learning_rate": 5.56816349111183e-07,
+      "loss": 1.5455,
+      "step": 23440
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 85.3125,
+      "learning_rate": 5.564382055803774e-07,
+      "loss": 1.5711,
+      "step": 23460
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 58.15625,
+      "learning_rate": 5.560600620495719e-07,
+      "loss": 1.5518,
+      "step": 23480
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 62.28125,
+      "learning_rate": 5.556819185187664e-07,
+      "loss": 1.6045,
+      "step": 23500
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 60.875,
+      "learning_rate": 5.553037749879607e-07,
+      "loss": 1.5856,
+      "step": 23520
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 50.75,
+      "learning_rate": 5.549256314571552e-07,
+      "loss": 1.6366,
+      "step": 23540
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 75.0625,
+      "learning_rate": 5.545474879263498e-07,
+      "loss": 1.5648,
+      "step": 23560
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 50.3125,
+      "learning_rate": 5.541693443955442e-07,
+      "loss": 1.565,
+      "step": 23580
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 54.46875,
+      "learning_rate": 5.537912008647386e-07,
+      "loss": 1.5149,
+      "step": 23600
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 92.75,
+      "learning_rate": 5.534130573339331e-07,
+      "loss": 1.4956,
+      "step": 23620
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 54.3125,
+      "learning_rate": 5.530349138031275e-07,
+      "loss": 1.4991,
+      "step": 23640
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 48.15625,
+      "learning_rate": 5.52656770272322e-07,
+      "loss": 1.5886,
+      "step": 23660
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 60.9375,
+      "learning_rate": 5.522786267415164e-07,
+      "loss": 1.5823,
+      "step": 23680
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 68.0625,
+      "learning_rate": 5.519004832107108e-07,
+      "loss": 1.5445,
+      "step": 23700
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 56.25,
+      "learning_rate": 5.515223396799053e-07,
+      "loss": 1.6111,
+      "step": 23720
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 76.6875,
+      "learning_rate": 5.511441961490996e-07,
+      "loss": 1.6356,
+      "step": 23740
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 70.25,
+      "learning_rate": 5.507660526182941e-07,
+      "loss": 1.5651,
+      "step": 23760
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 75.0625,
+      "learning_rate": 5.503879090874886e-07,
+      "loss": 1.4972,
+      "step": 23780
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 54.6875,
+      "learning_rate": 5.50009765556683e-07,
+      "loss": 1.5978,
+      "step": 23800
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 56.28125,
+      "learning_rate": 5.496316220258774e-07,
+      "loss": 1.5318,
+      "step": 23820
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 41.65625,
+      "learning_rate": 5.492534784950719e-07,
+      "loss": 1.468,
+      "step": 23840
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 52.625,
+      "learning_rate": 5.488753349642663e-07,
+      "loss": 1.4403,
+      "step": 23860
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 51.03125,
+      "learning_rate": 5.484971914334607e-07,
+      "loss": 1.5273,
+      "step": 23880
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 118.5,
+      "learning_rate": 5.481190479026553e-07,
+      "loss": 1.5754,
+      "step": 23900
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 53.65625,
+      "learning_rate": 5.477409043718497e-07,
+      "loss": 1.652,
+      "step": 23920
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 55.46875,
+      "learning_rate": 5.473627608410442e-07,
+      "loss": 1.5018,
+      "step": 23940
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 75.875,
+      "learning_rate": 5.469846173102386e-07,
+      "loss": 1.5074,
+      "step": 23960
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 65.375,
+      "learning_rate": 5.46606473779433e-07,
+      "loss": 1.5339,
+      "step": 23980
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 58.84375,
+      "learning_rate": 5.462283302486275e-07,
+      "loss": 1.6258,
+      "step": 24000
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 53.34375,
+      "learning_rate": 5.45850186717822e-07,
+      "loss": 1.5179,
+      "step": 24020
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 94.25,
+      "learning_rate": 5.454720431870163e-07,
+      "loss": 1.6923,
+      "step": 24040
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 59.78125,
+      "learning_rate": 5.450938996562108e-07,
+      "loss": 1.6017,
+      "step": 24060
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 82.5625,
+      "learning_rate": 5.447157561254052e-07,
+      "loss": 1.6311,
+      "step": 24080
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 64.1875,
+      "learning_rate": 5.443376125945996e-07,
+      "loss": 1.6113,
+      "step": 24100
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 62.40625,
+      "learning_rate": 5.439594690637941e-07,
+      "loss": 1.5551,
+      "step": 24120
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 62.3125,
+      "learning_rate": 5.435813255329885e-07,
+      "loss": 1.4569,
+      "step": 24140
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 58.46875,
+      "learning_rate": 5.43203182002183e-07,
+      "loss": 1.5644,
+      "step": 24160
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 92.8125,
+      "learning_rate": 5.428250384713774e-07,
+      "loss": 1.6306,
+      "step": 24180
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 55.125,
+      "learning_rate": 5.424468949405718e-07,
+      "loss": 1.5071,
+      "step": 24200
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 65.5,
+      "learning_rate": 5.420687514097663e-07,
+      "loss": 1.58,
+      "step": 24220
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 49.53125,
+      "learning_rate": 5.416906078789607e-07,
+      "loss": 1.58,
+      "step": 24240
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 59.03125,
+      "learning_rate": 5.413124643481552e-07,
+      "loss": 1.5532,
+      "step": 24260
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 61.5625,
+      "learning_rate": 5.409343208173497e-07,
+      "loss": 1.5801,
+      "step": 24280
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 57.875,
+      "learning_rate": 5.405561772865442e-07,
+      "loss": 1.518,
+      "step": 24300
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 54.40625,
+      "learning_rate": 5.401780337557385e-07,
+      "loss": 1.5373,
+      "step": 24320
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 69.625,
+      "learning_rate": 5.39799890224933e-07,
+      "loss": 1.5509,
+      "step": 24340
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 56.59375,
+      "learning_rate": 5.394217466941274e-07,
+      "loss": 1.603,
+      "step": 24360
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 71.375,
+      "learning_rate": 5.390436031633219e-07,
+      "loss": 1.3879,
+      "step": 24380
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 50.21875,
+      "learning_rate": 5.386654596325163e-07,
+      "loss": 1.5509,
+      "step": 24400
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 69.125,
+      "learning_rate": 5.382873161017107e-07,
+      "loss": 1.5635,
+      "step": 24420
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 67.125,
+      "learning_rate": 5.379091725709052e-07,
+      "loss": 1.4765,
+      "step": 24440
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 98.0,
+      "learning_rate": 5.375310290400996e-07,
+      "loss": 1.5189,
+      "step": 24460
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 58.8125,
+      "learning_rate": 5.37152885509294e-07,
+      "loss": 1.4824,
+      "step": 24480
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 74.6875,
+      "learning_rate": 5.367747419784885e-07,
+      "loss": 1.559,
+      "step": 24500
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 73.875,
+      "learning_rate": 5.36396598447683e-07,
+      "loss": 1.559,
+      "step": 24520
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 54.9375,
+      "learning_rate": 5.360184549168773e-07,
+      "loss": 1.5174,
+      "step": 24540
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 57.5625,
+      "learning_rate": 5.356403113860718e-07,
+      "loss": 1.5646,
+      "step": 24560
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 58.8125,
+      "learning_rate": 5.352621678552663e-07,
+      "loss": 1.4964,
+      "step": 24580
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 54.28125,
+      "learning_rate": 5.348840243244608e-07,
+      "loss": 1.5191,
+      "step": 24600
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 95.9375,
+      "learning_rate": 5.345058807936552e-07,
+      "loss": 1.567,
+      "step": 24620
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 91.5625,
+      "learning_rate": 5.341277372628497e-07,
+      "loss": 1.5026,
+      "step": 24640
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 61.03125,
+      "learning_rate": 5.337495937320441e-07,
+      "loss": 1.6615,
+      "step": 24660
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 53.625,
+      "learning_rate": 5.333714502012385e-07,
+      "loss": 1.5603,
+      "step": 24680
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 59.0625,
+      "learning_rate": 5.329933066704329e-07,
+      "loss": 1.5761,
+      "step": 24700
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 83.4375,
+      "learning_rate": 5.326151631396274e-07,
+      "loss": 1.6023,
+      "step": 24720
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 68.125,
+      "learning_rate": 5.322370196088219e-07,
+      "loss": 1.5586,
+      "step": 24740
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 55.96875,
+      "learning_rate": 5.318588760780162e-07,
+      "loss": 1.5758,
+      "step": 24760
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 53.3125,
+      "learning_rate": 5.314807325472107e-07,
+      "loss": 1.5305,
+      "step": 24780
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 62.0625,
+      "learning_rate": 5.311025890164052e-07,
+      "loss": 1.5119,
+      "step": 24800
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 45.6875,
+      "learning_rate": 5.307244454855995e-07,
+      "loss": 1.5253,
+      "step": 24820
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 72.8125,
+      "learning_rate": 5.30346301954794e-07,
+      "loss": 1.5749,
+      "step": 24840
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 78.875,
+      "learning_rate": 5.299681584239885e-07,
+      "loss": 1.5423,
+      "step": 24860
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 78.4375,
+      "learning_rate": 5.295900148931829e-07,
+      "loss": 1.5867,
+      "step": 24880
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 49.9375,
+      "learning_rate": 5.292118713623773e-07,
+      "loss": 1.6509,
+      "step": 24900
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 68.875,
+      "learning_rate": 5.288337278315718e-07,
+      "loss": 1.5651,
+      "step": 24920
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 55.15625,
+      "learning_rate": 5.284555843007662e-07,
+      "loss": 1.539,
+      "step": 24940
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 62.25,
+      "learning_rate": 5.280774407699608e-07,
+      "loss": 1.584,
+      "step": 24960
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 56.8125,
+      "learning_rate": 5.276992972391551e-07,
+      "loss": 1.5746,
+      "step": 24980
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 71.3125,
+      "learning_rate": 5.273211537083496e-07,
+      "loss": 1.5028,
+      "step": 25000
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 45.90625,
+      "learning_rate": 5.269430101775441e-07,
+      "loss": 1.6168,
+      "step": 25020
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 53.8125,
+      "learning_rate": 5.265648666467384e-07,
+      "loss": 1.5284,
+      "step": 25040
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 55.875,
+      "learning_rate": 5.261867231159329e-07,
+      "loss": 1.5533,
+      "step": 25060
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 62.375,
+      "learning_rate": 5.258085795851274e-07,
+      "loss": 1.5737,
+      "step": 25080
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 54.59375,
+      "learning_rate": 5.254304360543218e-07,
+      "loss": 1.5599,
+      "step": 25100
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 68.1875,
+      "learning_rate": 5.250522925235162e-07,
+      "loss": 1.5021,
+      "step": 25120
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 62.71875,
+      "learning_rate": 5.246741489927107e-07,
+      "loss": 1.44,
+      "step": 25140
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 52.96875,
+      "learning_rate": 5.242960054619051e-07,
+      "loss": 1.4916,
+      "step": 25160
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 54.625,
+      "learning_rate": 5.239178619310996e-07,
+      "loss": 1.5164,
+      "step": 25180
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 52.5625,
+      "learning_rate": 5.23539718400294e-07,
+      "loss": 1.4848,
+      "step": 25200
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 72.0625,
+      "learning_rate": 5.231615748694884e-07,
+      "loss": 1.5058,
+      "step": 25220
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 74.125,
+      "learning_rate": 5.227834313386829e-07,
+      "loss": 1.6166,
+      "step": 25240
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 53.40625,
+      "learning_rate": 5.224052878078772e-07,
+      "loss": 1.5508,
+      "step": 25260
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 57.21875,
+      "learning_rate": 5.220271442770717e-07,
+      "loss": 1.6444,
+      "step": 25280
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 62.5,
+      "learning_rate": 5.216490007462663e-07,
+      "loss": 1.5956,
+      "step": 25300
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 71.0,
+      "learning_rate": 5.212708572154607e-07,
+      "loss": 1.5861,
+      "step": 25320
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 51.8125,
+      "learning_rate": 5.208927136846551e-07,
+      "loss": 1.5077,
+      "step": 25340
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 67.5625,
+      "learning_rate": 5.205145701538496e-07,
+      "loss": 1.592,
+      "step": 25360
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 49.03125,
+      "learning_rate": 5.20136426623044e-07,
+      "loss": 1.6034,
+      "step": 25380
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 49.4375,
+      "learning_rate": 5.197582830922385e-07,
+      "loss": 1.4929,
+      "step": 25400
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 64.875,
+      "learning_rate": 5.193801395614329e-07,
+      "loss": 1.532,
+      "step": 25420
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 57.1875,
+      "learning_rate": 5.190019960306273e-07,
+      "loss": 1.4758,
+      "step": 25440
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 64.0,
+      "learning_rate": 5.186238524998218e-07,
+      "loss": 1.5308,
+      "step": 25460
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 48.34375,
+      "learning_rate": 5.182457089690162e-07,
+      "loss": 1.5379,
+      "step": 25480
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 55.1875,
+      "learning_rate": 5.178675654382106e-07,
+      "loss": 1.5318,
+      "step": 25500
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 67.0,
+      "learning_rate": 5.174894219074051e-07,
+      "loss": 1.5173,
+      "step": 25520
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 44.46875,
+      "learning_rate": 5.171112783765996e-07,
+      "loss": 1.5026,
+      "step": 25540
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 60.875,
+      "learning_rate": 5.167331348457939e-07,
+      "loss": 1.5759,
+      "step": 25560
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 60.75,
+      "learning_rate": 5.163549913149884e-07,
+      "loss": 1.5473,
+      "step": 25580
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 46.84375,
+      "learning_rate": 5.159768477841828e-07,
+      "loss": 1.4403,
+      "step": 25600
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 50.34375,
+      "learning_rate": 5.155987042533772e-07,
+      "loss": 1.5662,
+      "step": 25620
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 62.6875,
+      "learning_rate": 5.152205607225717e-07,
+      "loss": 1.4555,
+      "step": 25640
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 61.59375,
+      "learning_rate": 5.148424171917662e-07,
+      "loss": 1.6103,
+      "step": 25660
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 71.5625,
+      "learning_rate": 5.144642736609607e-07,
+      "loss": 1.5453,
+      "step": 25680
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 68.1875,
+      "learning_rate": 5.140861301301551e-07,
+      "loss": 1.6433,
+      "step": 25700
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 54.34375,
+      "learning_rate": 5.137079865993495e-07,
+      "loss": 1.5406,
+      "step": 25720
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 52.03125,
+      "learning_rate": 5.13329843068544e-07,
+      "loss": 1.518,
+      "step": 25740
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 85.6875,
+      "learning_rate": 5.129516995377385e-07,
+      "loss": 1.6169,
+      "step": 25760
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 68.0625,
+      "learning_rate": 5.125735560069328e-07,
+      "loss": 1.5893,
+      "step": 25780
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 55.71875,
+      "learning_rate": 5.121954124761273e-07,
+      "loss": 1.6479,
+      "step": 25800
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 58.96875,
+      "learning_rate": 5.118172689453218e-07,
+      "loss": 1.6356,
+      "step": 25820
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 57.5625,
+      "learning_rate": 5.114391254145161e-07,
+      "loss": 1.5878,
+      "step": 25840
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 91.875,
+      "learning_rate": 5.110609818837106e-07,
+      "loss": 1.4991,
+      "step": 25860
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 62.8125,
+      "learning_rate": 5.10682838352905e-07,
+      "loss": 1.4639,
+      "step": 25880
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 74.1875,
+      "learning_rate": 5.103046948220995e-07,
+      "loss": 1.6055,
+      "step": 25900
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 69.8125,
+      "learning_rate": 5.099265512912939e-07,
+      "loss": 1.555,
+      "step": 25920
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 65.125,
+      "learning_rate": 5.095484077604883e-07,
+      "loss": 1.625,
+      "step": 25940
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 60.71875,
+      "learning_rate": 5.091702642296828e-07,
+      "loss": 1.5422,
+      "step": 25960
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 56.875,
+      "learning_rate": 5.087921206988773e-07,
+      "loss": 1.6063,
+      "step": 25980
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 58.0625,
+      "learning_rate": 5.084139771680717e-07,
+      "loss": 1.498,
+      "step": 26000
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 56.65625,
+      "learning_rate": 5.080358336372662e-07,
+      "loss": 1.5283,
+      "step": 26020
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 50.3125,
+      "learning_rate": 5.076576901064607e-07,
+      "loss": 1.4989,
+      "step": 26040
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 66.0625,
+      "learning_rate": 5.07279546575655e-07,
+      "loss": 1.4776,
+      "step": 26060
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 78.9375,
+      "learning_rate": 5.069014030448495e-07,
+      "loss": 1.4897,
+      "step": 26080
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 65.0,
+      "learning_rate": 5.06523259514044e-07,
+      "loss": 1.5321,
+      "step": 26100
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 71.0,
+      "learning_rate": 5.061451159832384e-07,
+      "loss": 1.5652,
+      "step": 26120
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 52.8125,
+      "learning_rate": 5.057669724524328e-07,
+      "loss": 1.5619,
+      "step": 26140
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 58.71875,
+      "learning_rate": 5.053888289216273e-07,
+      "loss": 1.6027,
+      "step": 26160
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 62.4375,
+      "learning_rate": 5.050106853908217e-07,
+      "loss": 1.5552,
+      "step": 26180
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 66.1875,
+      "learning_rate": 5.046325418600162e-07,
+      "loss": 1.5539,
+      "step": 26200
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 44.78125,
+      "learning_rate": 5.042543983292105e-07,
+      "loss": 1.4551,
+      "step": 26220
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 87.8125,
+      "learning_rate": 5.03876254798405e-07,
+      "loss": 1.5585,
+      "step": 26240
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 97.8125,
+      "learning_rate": 5.034981112675995e-07,
+      "loss": 1.5294,
+      "step": 26260
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 70.25,
+      "learning_rate": 5.031199677367938e-07,
+      "loss": 1.5808,
+      "step": 26280
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 62.375,
+      "learning_rate": 5.027418242059883e-07,
+      "loss": 1.5392,
+      "step": 26300
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 67.9375,
+      "learning_rate": 5.023636806751828e-07,
+      "loss": 1.53,
+      "step": 26320
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 62.3125,
+      "learning_rate": 5.019855371443772e-07,
+      "loss": 1.5298,
+      "step": 26340
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 64.375,
+      "learning_rate": 5.016073936135717e-07,
+      "loss": 1.493,
+      "step": 26360
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 66.125,
+      "learning_rate": 5.012292500827662e-07,
+      "loss": 1.595,
+      "step": 26380
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 79.0,
+      "learning_rate": 5.008511065519606e-07,
+      "loss": 1.4776,
+      "step": 26400
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 70.125,
+      "learning_rate": 5.004729630211551e-07,
+      "loss": 1.5509,
+      "step": 26420
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 70.4375,
+      "learning_rate": 5.000948194903495e-07,
+      "loss": 1.6068,
+      "step": 26440
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 62.21875,
+      "learning_rate": 4.997166759595439e-07,
+      "loss": 1.553,
+      "step": 26460
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 57.125,
+      "learning_rate": 4.993385324287384e-07,
+      "loss": 1.6032,
+      "step": 26480
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 91.125,
+      "learning_rate": 4.989603888979328e-07,
+      "loss": 1.4595,
+      "step": 26500
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 78.6875,
+      "learning_rate": 4.985822453671272e-07,
+      "loss": 1.5136,
+      "step": 26520
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 57.625,
+      "learning_rate": 4.982041018363217e-07,
+      "loss": 1.592,
+      "step": 26540
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 71.6875,
+      "learning_rate": 4.978259583055161e-07,
+      "loss": 1.5033,
+      "step": 26560
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 67.375,
+      "learning_rate": 4.974478147747105e-07,
+      "loss": 1.5144,
+      "step": 26580
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 53.96875,
+      "learning_rate": 4.97069671243905e-07,
+      "loss": 1.5845,
+      "step": 26600
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 77.5625,
+      "learning_rate": 4.966915277130995e-07,
+      "loss": 1.6229,
+      "step": 26620
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 59.28125,
+      "learning_rate": 4.96313384182294e-07,
+      "loss": 1.6134,
+      "step": 26640
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 43.34375,
+      "learning_rate": 4.959352406514883e-07,
+      "loss": 1.59,
+      "step": 26660
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 76.875,
+      "learning_rate": 4.955570971206828e-07,
+      "loss": 1.5741,
+      "step": 26680
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 70.9375,
+      "learning_rate": 4.951789535898772e-07,
+      "loss": 1.5271,
+      "step": 26700
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 53.28125,
+      "learning_rate": 4.948008100590716e-07,
+      "loss": 1.436,
+      "step": 26720
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 59.03125,
+      "learning_rate": 4.944226665282661e-07,
+      "loss": 1.5151,
+      "step": 26740
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 74.6875,
+      "learning_rate": 4.940445229974605e-07,
+      "loss": 1.4958,
+      "step": 26760
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 55.03125,
+      "learning_rate": 4.936663794666551e-07,
+      "loss": 1.5223,
+      "step": 26780
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 59.1875,
+      "learning_rate": 4.932882359358494e-07,
+      "loss": 1.5894,
+      "step": 26800
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 74.5,
+      "learning_rate": 4.929100924050439e-07,
+      "loss": 1.4906,
+      "step": 26820
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 65.5,
+      "learning_rate": 4.925319488742383e-07,
+      "loss": 1.5399,
+      "step": 26840
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 60.0,
+      "learning_rate": 4.921538053434327e-07,
+      "loss": 1.5463,
+      "step": 26860
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 56.5,
+      "learning_rate": 4.917756618126272e-07,
+      "loss": 1.5375,
+      "step": 26880
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 93.3125,
+      "learning_rate": 4.913975182818216e-07,
+      "loss": 1.5447,
+      "step": 26900
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 88.3125,
+      "learning_rate": 4.910193747510161e-07,
+      "loss": 1.5076,
+      "step": 26920
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 88.9375,
+      "learning_rate": 4.906412312202105e-07,
+      "loss": 1.6689,
+      "step": 26940
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 77.25,
+      "learning_rate": 4.90263087689405e-07,
+      "loss": 1.5952,
+      "step": 26960
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 53.3125,
+      "learning_rate": 4.898849441585995e-07,
+      "loss": 1.4589,
+      "step": 26980
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 61.1875,
+      "learning_rate": 4.895068006277939e-07,
+      "loss": 1.4918,
+      "step": 27000
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 73.375,
+      "learning_rate": 4.891286570969883e-07,
+      "loss": 1.533,
+      "step": 27020
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 75.6875,
+      "learning_rate": 4.887505135661827e-07,
+      "loss": 1.5737,
+      "step": 27040
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 67.3125,
+      "learning_rate": 4.883723700353772e-07,
+      "loss": 1.5193,
+      "step": 27060
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 85.625,
+      "learning_rate": 4.879942265045716e-07,
+      "loss": 1.5908,
+      "step": 27080
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 92.3125,
+      "learning_rate": 4.87616082973766e-07,
+      "loss": 1.612,
+      "step": 27100
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 58.03125,
+      "learning_rate": 4.872379394429605e-07,
+      "loss": 1.5158,
+      "step": 27120
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 62.125,
+      "learning_rate": 4.86859795912155e-07,
+      "loss": 1.5388,
+      "step": 27140
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 53.1875,
+      "learning_rate": 4.864816523813494e-07,
+      "loss": 1.5011,
+      "step": 27160
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 51.40625,
+      "learning_rate": 4.861035088505438e-07,
+      "loss": 1.5886,
+      "step": 27180
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 45.90625,
+      "learning_rate": 4.857253653197383e-07,
+      "loss": 1.5871,
+      "step": 27200
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 67.6875,
+      "learning_rate": 4.853472217889328e-07,
+      "loss": 1.5536,
+      "step": 27220
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 50.34375,
+      "learning_rate": 4.849690782581271e-07,
+      "loss": 1.4826,
+      "step": 27240
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 52.6875,
+      "learning_rate": 4.845909347273216e-07,
+      "loss": 1.599,
+      "step": 27260
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 57.21875,
+      "learning_rate": 4.842127911965161e-07,
+      "loss": 1.5626,
+      "step": 27280
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 68.0625,
+      "learning_rate": 4.838346476657104e-07,
+      "loss": 1.4922,
+      "step": 27300
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 70.5,
+      "learning_rate": 4.834565041349049e-07,
+      "loss": 1.6091,
+      "step": 27320
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 61.34375,
+      "learning_rate": 4.830783606040994e-07,
+      "loss": 1.532,
+      "step": 27340
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 82.0625,
+      "learning_rate": 4.827002170732939e-07,
+      "loss": 1.5448,
+      "step": 27360
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 42.21875,
+      "learning_rate": 4.823220735424882e-07,
+      "loss": 1.4501,
+      "step": 27380
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 64.875,
+      "learning_rate": 4.819439300116827e-07,
+      "loss": 1.6265,
+      "step": 27400
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 60.3125,
+      "learning_rate": 4.815657864808772e-07,
+      "loss": 1.4552,
+      "step": 27420
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 60.0625,
+      "learning_rate": 4.811876429500715e-07,
+      "loss": 1.6434,
+      "step": 27440
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 72.1875,
+      "learning_rate": 4.80809499419266e-07,
+      "loss": 1.5811,
+      "step": 27460
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 61.59375,
+      "learning_rate": 4.804313558884605e-07,
+      "loss": 1.5929,
+      "step": 27480
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 55.90625,
+      "learning_rate": 4.80053212357655e-07,
+      "loss": 1.598,
+      "step": 27500
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 61.5625,
+      "learning_rate": 4.796750688268493e-07,
+      "loss": 1.5171,
+      "step": 27520
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 77.5,
+      "learning_rate": 4.792969252960438e-07,
+      "loss": 1.5044,
+      "step": 27540
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 69.0,
+      "learning_rate": 4.789187817652383e-07,
+      "loss": 1.5655,
+      "step": 27560
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 58.0625,
+      "learning_rate": 4.785406382344327e-07,
+      "loss": 1.4868,
+      "step": 27580
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 65.0,
+      "learning_rate": 4.781624947036271e-07,
+      "loss": 1.5374,
+      "step": 27600
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 72.875,
+      "learning_rate": 4.777843511728216e-07,
+      "loss": 1.5488,
+      "step": 27620
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 58.1875,
+      "learning_rate": 4.77406207642016e-07,
+      "loss": 1.5216,
+      "step": 27640
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 53.96875,
+      "learning_rate": 4.770280641112104e-07,
+      "loss": 1.5225,
+      "step": 27660
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 57.0625,
+      "learning_rate": 4.766499205804049e-07,
+      "loss": 1.5891,
+      "step": 27680
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 99.8125,
+      "learning_rate": 4.7627177704959934e-07,
+      "loss": 1.5102,
+      "step": 27700
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 68.9375,
+      "learning_rate": 4.758936335187938e-07,
+      "loss": 1.5994,
+      "step": 27720
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 58.21875,
+      "learning_rate": 4.7551548998798824e-07,
+      "loss": 1.5237,
+      "step": 27740
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 58.375,
+      "learning_rate": 4.7513734645718266e-07,
+      "loss": 1.5272,
+      "step": 27760
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 59.9375,
+      "learning_rate": 4.7475920292637713e-07,
+      "loss": 1.5752,
+      "step": 27780
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 63.25,
+      "learning_rate": 4.7438105939557155e-07,
+      "loss": 1.5307,
+      "step": 27800
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 67.5,
+      "learning_rate": 4.74002915864766e-07,
+      "loss": 1.5077,
+      "step": 27820
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 72.625,
+      "learning_rate": 4.736247723339605e-07,
+      "loss": 1.4981,
+      "step": 27840
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 54.125,
+      "learning_rate": 4.732466288031549e-07,
+      "loss": 1.62,
+      "step": 27860
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 61.59375,
+      "learning_rate": 4.7286848527234934e-07,
+      "loss": 1.5666,
+      "step": 27880
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 55.15625,
+      "learning_rate": 4.7249034174154377e-07,
+      "loss": 1.5126,
+      "step": 27900
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 58.5,
+      "learning_rate": 4.7211219821073824e-07,
+      "loss": 1.6101,
+      "step": 27920
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 64.125,
+      "learning_rate": 4.7173405467993266e-07,
+      "loss": 1.5902,
+      "step": 27940
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 51.25,
+      "learning_rate": 4.713559111491271e-07,
+      "loss": 1.5703,
+      "step": 27960
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 56.21875,
+      "learning_rate": 4.7097776761832156e-07,
+      "loss": 1.57,
+      "step": 27980
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 117.5625,
+      "learning_rate": 4.70599624087516e-07,
+      "loss": 1.4706,
+      "step": 28000
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 58.75,
+      "learning_rate": 4.7022148055671045e-07,
+      "loss": 1.49,
+      "step": 28020
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 85.1875,
+      "learning_rate": 4.698433370259049e-07,
+      "loss": 1.4867,
+      "step": 28040
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 62.3125,
+      "learning_rate": 4.6946519349509935e-07,
+      "loss": 1.6106,
+      "step": 28060
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 69.8125,
+      "learning_rate": 4.6908704996429377e-07,
+      "loss": 1.5242,
+      "step": 28080
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 73.9375,
+      "learning_rate": 4.687089064334882e-07,
+      "loss": 1.5871,
+      "step": 28100
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 75.6875,
+      "learning_rate": 4.6833076290268266e-07,
+      "loss": 1.5284,
+      "step": 28120
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 60.28125,
+      "learning_rate": 4.679526193718771e-07,
+      "loss": 1.5682,
+      "step": 28140
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 52.8125,
+      "learning_rate": 4.675744758410715e-07,
+      "loss": 1.5701,
+      "step": 28160
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 71.5,
+      "learning_rate": 4.67196332310266e-07,
+      "loss": 1.5179,
+      "step": 28180
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 57.90625,
+      "learning_rate": 4.6681818877946046e-07,
+      "loss": 1.544,
+      "step": 28200
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 81.5625,
+      "learning_rate": 4.664400452486549e-07,
+      "loss": 1.5441,
+      "step": 28220
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 76.0625,
+      "learning_rate": 4.660619017178493e-07,
+      "loss": 1.5856,
+      "step": 28240
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 77.125,
+      "learning_rate": 4.6568375818704377e-07,
+      "loss": 1.5011,
+      "step": 28260
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 67.6875,
+      "learning_rate": 4.653056146562382e-07,
+      "loss": 1.6049,
+      "step": 28280
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 59.4375,
+      "learning_rate": 4.649274711254326e-07,
+      "loss": 1.5911,
+      "step": 28300
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 76.0625,
+      "learning_rate": 4.6454932759462704e-07,
+      "loss": 1.5146,
+      "step": 28320
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 56.34375,
+      "learning_rate": 4.641711840638215e-07,
+      "loss": 1.4633,
+      "step": 28340
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 70.5,
+      "learning_rate": 4.63793040533016e-07,
+      "loss": 1.5427,
+      "step": 28360
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 72.0625,
+      "learning_rate": 4.634148970022104e-07,
+      "loss": 1.5296,
+      "step": 28380
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 63.96875,
+      "learning_rate": 4.630367534714049e-07,
+      "loss": 1.4503,
+      "step": 28400
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 61.09375,
+      "learning_rate": 4.626586099405993e-07,
+      "loss": 1.5682,
+      "step": 28420
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 109.375,
+      "learning_rate": 4.622804664097937e-07,
+      "loss": 1.6351,
+      "step": 28440
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 64.4375,
+      "learning_rate": 4.6190232287898814e-07,
+      "loss": 1.5257,
+      "step": 28460
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 66.3125,
+      "learning_rate": 4.615241793481826e-07,
+      "loss": 1.4647,
+      "step": 28480
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 62.1875,
+      "learning_rate": 4.6114603581737704e-07,
+      "loss": 1.5068,
+      "step": 28500
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 60.46875,
+      "learning_rate": 4.6076789228657146e-07,
+      "loss": 1.5132,
+      "step": 28520
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 63.21875,
+      "learning_rate": 4.60389748755766e-07,
+      "loss": 1.5879,
+      "step": 28540
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 77.9375,
+      "learning_rate": 4.600116052249604e-07,
+      "loss": 1.5218,
+      "step": 28560
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 58.0625,
+      "learning_rate": 4.5963346169415483e-07,
+      "loss": 1.596,
+      "step": 28580
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 48.78125,
+      "learning_rate": 4.592553181633493e-07,
+      "loss": 1.6146,
+      "step": 28600
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 77.8125,
+      "learning_rate": 4.5887717463254373e-07,
+      "loss": 1.4568,
+      "step": 28620
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 155.25,
+      "learning_rate": 4.5849903110173815e-07,
+      "loss": 1.5704,
+      "step": 28640
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 56.4375,
+      "learning_rate": 4.5812088757093257e-07,
+      "loss": 1.5383,
+      "step": 28660
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 98.5,
+      "learning_rate": 4.5774274404012704e-07,
+      "loss": 1.5281,
+      "step": 28680
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 62.25,
+      "learning_rate": 4.5736460050932146e-07,
+      "loss": 1.5042,
+      "step": 28700
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 65.625,
+      "learning_rate": 4.5698645697851594e-07,
+      "loss": 1.5688,
+      "step": 28720
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 54.90625,
+      "learning_rate": 4.566083134477104e-07,
+      "loss": 1.5805,
+      "step": 28740
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 57.90625,
+      "learning_rate": 4.5623016991690483e-07,
+      "loss": 1.5071,
+      "step": 28760
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 63.375,
+      "learning_rate": 4.5585202638609926e-07,
+      "loss": 1.3748,
+      "step": 28780
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 67.6875,
+      "learning_rate": 4.554738828552937e-07,
+      "loss": 1.6716,
+      "step": 28800
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 73.5625,
+      "learning_rate": 4.5509573932448815e-07,
+      "loss": 1.4361,
+      "step": 28820
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 62.0625,
+      "learning_rate": 4.5471759579368257e-07,
+      "loss": 1.6571,
+      "step": 28840
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 58.5,
+      "learning_rate": 4.54339452262877e-07,
+      "loss": 1.5489,
+      "step": 28860
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 53.53125,
+      "learning_rate": 4.539613087320715e-07,
+      "loss": 1.5146,
+      "step": 28880
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 47.5,
+      "learning_rate": 4.5358316520126594e-07,
+      "loss": 1.5828,
+      "step": 28900
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 50.25,
+      "learning_rate": 4.5320502167046036e-07,
+      "loss": 1.4854,
+      "step": 28920
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 73.75,
+      "learning_rate": 4.528268781396548e-07,
+      "loss": 1.5217,
+      "step": 28940
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 56.71875,
+      "learning_rate": 4.5244873460884926e-07,
+      "loss": 1.5262,
+      "step": 28960
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 54.8125,
+      "learning_rate": 4.520705910780437e-07,
+      "loss": 1.5702,
+      "step": 28980
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 80.3125,
+      "learning_rate": 4.516924475472381e-07,
+      "loss": 1.5637,
+      "step": 29000
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 107.75,
+      "learning_rate": 4.513143040164326e-07,
+      "loss": 1.5309,
+      "step": 29020
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 48.09375,
+      "learning_rate": 4.50936160485627e-07,
+      "loss": 1.5284,
+      "step": 29040
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 56.90625,
+      "learning_rate": 4.5055801695482147e-07,
+      "loss": 1.5613,
+      "step": 29060
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 62.375,
+      "learning_rate": 4.5017987342401595e-07,
+      "loss": 1.5028,
+      "step": 29080
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 59.625,
+      "learning_rate": 4.4980172989321037e-07,
+      "loss": 1.6023,
+      "step": 29100
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 70.625,
+      "learning_rate": 4.494235863624048e-07,
+      "loss": 1.5756,
+      "step": 29120
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 66.8125,
+      "learning_rate": 4.490454428315992e-07,
+      "loss": 1.6159,
+      "step": 29140
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 67.8125,
+      "learning_rate": 4.486672993007937e-07,
+      "loss": 1.5633,
+      "step": 29160
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 44.71875,
+      "learning_rate": 4.482891557699881e-07,
+      "loss": 1.5654,
+      "step": 29180
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 58.59375,
+      "learning_rate": 4.4791101223918253e-07,
+      "loss": 1.5233,
+      "step": 29200
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 61.90625,
+      "learning_rate": 4.47532868708377e-07,
+      "loss": 1.5729,
+      "step": 29220
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 76.5625,
+      "learning_rate": 4.471547251775715e-07,
+      "loss": 1.5359,
+      "step": 29240
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 80.875,
+      "learning_rate": 4.467765816467659e-07,
+      "loss": 1.5896,
+      "step": 29260
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 80.5,
+      "learning_rate": 4.463984381159603e-07,
+      "loss": 1.6558,
+      "step": 29280
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 61.0625,
+      "learning_rate": 4.460202945851548e-07,
+      "loss": 1.5273,
+      "step": 29300
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 64.1875,
+      "learning_rate": 4.456421510543492e-07,
+      "loss": 1.5447,
+      "step": 29320
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 69.375,
+      "learning_rate": 4.4526400752354363e-07,
+      "loss": 1.5843,
+      "step": 29340
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 68.125,
+      "learning_rate": 4.448858639927381e-07,
+      "loss": 1.5857,
+      "step": 29360
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 68.0,
+      "learning_rate": 4.4450772046193253e-07,
+      "loss": 1.579,
+      "step": 29380
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 70.6875,
+      "learning_rate": 4.44129576931127e-07,
+      "loss": 1.5822,
+      "step": 29400
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 64.9375,
+      "learning_rate": 4.437514334003214e-07,
+      "loss": 1.6115,
+      "step": 29420
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 101.9375,
+      "learning_rate": 4.433732898695159e-07,
+      "loss": 1.6074,
+      "step": 29440
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 70.6875,
+      "learning_rate": 4.429951463387103e-07,
+      "loss": 1.5953,
+      "step": 29460
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 57.25,
+      "learning_rate": 4.4261700280790474e-07,
+      "loss": 1.492,
+      "step": 29480
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 70.5625,
+      "learning_rate": 4.422388592770992e-07,
+      "loss": 1.6068,
+      "step": 29500
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 57.6875,
+      "learning_rate": 4.4186071574629364e-07,
+      "loss": 1.5683,
+      "step": 29520
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 57.78125,
+      "learning_rate": 4.4148257221548806e-07,
+      "loss": 1.5758,
+      "step": 29540
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 72.125,
+      "learning_rate": 4.411044286846825e-07,
+      "loss": 1.5107,
+      "step": 29560
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 63.46875,
+      "learning_rate": 4.40726285153877e-07,
+      "loss": 1.5422,
+      "step": 29580
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 43.5,
+      "learning_rate": 4.4034814162307143e-07,
+      "loss": 1.5609,
+      "step": 29600
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 57.8125,
+      "learning_rate": 4.3996999809226585e-07,
+      "loss": 1.5686,
+      "step": 29620
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 64.625,
+      "learning_rate": 4.395918545614603e-07,
+      "loss": 1.4721,
+      "step": 29640
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 77.875,
+      "learning_rate": 4.3921371103065475e-07,
+      "loss": 1.5926,
+      "step": 29660
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 66.0,
+      "learning_rate": 4.3883556749984917e-07,
+      "loss": 1.5014,
+      "step": 29680
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 51.5,
+      "learning_rate": 4.3845742396904364e-07,
+      "loss": 1.5162,
+      "step": 29700
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 80.625,
+      "learning_rate": 4.3807928043823806e-07,
+      "loss": 1.5249,
+      "step": 29720
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 75.0,
+      "learning_rate": 4.377011369074325e-07,
+      "loss": 1.5725,
+      "step": 29740
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 79.4375,
+      "learning_rate": 4.3732299337662696e-07,
+      "loss": 1.444,
+      "step": 29760
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 57.4375,
+      "learning_rate": 4.3694484984582143e-07,
+      "loss": 1.5409,
+      "step": 29780
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 79.1875,
+      "learning_rate": 4.3656670631501585e-07,
+      "loss": 1.5865,
+      "step": 29800
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 75.4375,
+      "learning_rate": 4.361885627842103e-07,
+      "loss": 1.4412,
+      "step": 29820
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 58.125,
+      "learning_rate": 4.3581041925340475e-07,
+      "loss": 1.5931,
+      "step": 29840
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 69.4375,
+      "learning_rate": 4.3543227572259917e-07,
+      "loss": 1.6071,
+      "step": 29860
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 60.65625,
+      "learning_rate": 4.350541321917936e-07,
+      "loss": 1.5137,
+      "step": 29880
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 77.8125,
+      "learning_rate": 4.34675988660988e-07,
+      "loss": 1.5608,
+      "step": 29900
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 45.5,
+      "learning_rate": 4.342978451301825e-07,
+      "loss": 1.5166,
+      "step": 29920
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 59.5,
+      "learning_rate": 4.3391970159937696e-07,
+      "loss": 1.6169,
+      "step": 29940
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 61.375,
+      "learning_rate": 4.335415580685714e-07,
+      "loss": 1.4541,
+      "step": 29960
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 60.28125,
+      "learning_rate": 4.3316341453776586e-07,
+      "loss": 1.5651,
+      "step": 29980
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 65.4375,
+      "learning_rate": 4.327852710069603e-07,
+      "loss": 1.6816,
+      "step": 30000
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 60.125,
+      "learning_rate": 4.324071274761547e-07,
+      "loss": 1.502,
+      "step": 30020
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 61.6875,
+      "learning_rate": 4.320289839453491e-07,
+      "loss": 1.4747,
+      "step": 30040
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 47.0,
+      "learning_rate": 4.316508404145436e-07,
+      "loss": 1.5672,
+      "step": 30060
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 63.0625,
+      "learning_rate": 4.31272696883738e-07,
+      "loss": 1.5038,
+      "step": 30080
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 55.0625,
+      "learning_rate": 4.308945533529325e-07,
+      "loss": 1.5884,
+      "step": 30100
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 56.4375,
+      "learning_rate": 4.3051640982212697e-07,
+      "loss": 1.5646,
+      "step": 30120
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 85.8125,
+      "learning_rate": 4.301382662913214e-07,
+      "loss": 1.4855,
+      "step": 30140
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 91.0,
+      "learning_rate": 4.297601227605158e-07,
+      "loss": 1.5274,
+      "step": 30160
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 83.1875,
+      "learning_rate": 4.293819792297103e-07,
+      "loss": 1.5331,
+      "step": 30180
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 57.71875,
+      "learning_rate": 4.290038356989047e-07,
+      "loss": 1.4909,
+      "step": 30200
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 84.5,
+      "learning_rate": 4.286256921680991e-07,
+      "loss": 1.5937,
+      "step": 30220
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 78.5,
+      "learning_rate": 4.2824754863729355e-07,
+      "loss": 1.5401,
+      "step": 30240
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 65.1875,
+      "learning_rate": 4.27869405106488e-07,
+      "loss": 1.5679,
+      "step": 30260
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 66.25,
+      "learning_rate": 4.274912615756825e-07,
+      "loss": 1.4403,
+      "step": 30280
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 56.34375,
+      "learning_rate": 4.271131180448769e-07,
+      "loss": 1.515,
+      "step": 30300
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 50.9375,
+      "learning_rate": 4.267349745140714e-07,
+      "loss": 1.4955,
+      "step": 30320
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 58.6875,
+      "learning_rate": 4.263568309832658e-07,
+      "loss": 1.4947,
+      "step": 30340
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 51.75,
+      "learning_rate": 4.2597868745246023e-07,
+      "loss": 1.578,
+      "step": 30360
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 85.9375,
+      "learning_rate": 4.2560054392165465e-07,
+      "loss": 1.5881,
+      "step": 30380
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 79.25,
+      "learning_rate": 4.2522240039084913e-07,
+      "loss": 1.5744,
+      "step": 30400
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 76.8125,
+      "learning_rate": 4.2484425686004355e-07,
+      "loss": 1.5821,
+      "step": 30420
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 51.375,
+      "learning_rate": 4.2446611332923797e-07,
+      "loss": 1.585,
+      "step": 30440
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 62.59375,
+      "learning_rate": 4.240879697984325e-07,
+      "loss": 1.5283,
+      "step": 30460
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 52.0625,
+      "learning_rate": 4.237098262676269e-07,
+      "loss": 1.4999,
+      "step": 30480
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 51.65625,
+      "learning_rate": 4.2333168273682134e-07,
+      "loss": 1.6095,
+      "step": 30500
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 76.4375,
+      "learning_rate": 4.2295353920601576e-07,
+      "loss": 1.5921,
+      "step": 30520
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 72.1875,
+      "learning_rate": 4.2257539567521024e-07,
+      "loss": 1.6093,
+      "step": 30540
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 67.25,
+      "learning_rate": 4.2219725214440466e-07,
+      "loss": 1.4845,
+      "step": 30560
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 55.6875,
+      "learning_rate": 4.218191086135991e-07,
+      "loss": 1.584,
+      "step": 30580
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 50.96875,
+      "learning_rate": 4.2144096508279355e-07,
+      "loss": 1.4808,
+      "step": 30600
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 100.75,
+      "learning_rate": 4.21062821551988e-07,
+      "loss": 1.5642,
+      "step": 30620
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 53.71875,
+      "learning_rate": 4.2068467802118245e-07,
+      "loss": 1.5242,
+      "step": 30640
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 54.25,
+      "learning_rate": 4.2030653449037687e-07,
+      "loss": 1.5593,
+      "step": 30660
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 61.28125,
+      "learning_rate": 4.1992839095957134e-07,
+      "loss": 1.5166,
+      "step": 30680
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 56.5625,
+      "learning_rate": 4.1955024742876577e-07,
+      "loss": 1.6194,
+      "step": 30700
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 59.03125,
+      "learning_rate": 4.191721038979602e-07,
+      "loss": 1.5371,
+      "step": 30720
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 55.03125,
+      "learning_rate": 4.1879396036715466e-07,
+      "loss": 1.4101,
+      "step": 30740
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 68.125,
+      "learning_rate": 4.184158168363491e-07,
+      "loss": 1.5908,
+      "step": 30760
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 51.59375,
+      "learning_rate": 4.180376733055435e-07,
+      "loss": 1.5034,
+      "step": 30780
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 58.4375,
+      "learning_rate": 4.1765952977473803e-07,
+      "loss": 1.5745,
+      "step": 30800
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 75.4375,
+      "learning_rate": 4.1728138624393245e-07,
+      "loss": 1.5812,
+      "step": 30820
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 66.875,
+      "learning_rate": 4.169032427131269e-07,
+      "loss": 1.5836,
+      "step": 30840
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 53.125,
+      "learning_rate": 4.165250991823213e-07,
+      "loss": 1.534,
+      "step": 30860
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 76.0625,
+      "learning_rate": 4.1614695565151577e-07,
+      "loss": 1.5019,
+      "step": 30880
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 61.9375,
+      "learning_rate": 4.157688121207102e-07,
+      "loss": 1.507,
+      "step": 30900
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 85.25,
+      "learning_rate": 4.153906685899046e-07,
+      "loss": 1.5362,
+      "step": 30920
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 71.625,
+      "learning_rate": 4.150125250590991e-07,
+      "loss": 1.5695,
+      "step": 30940
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 69.25,
+      "learning_rate": 4.146343815282935e-07,
+      "loss": 1.5302,
+      "step": 30960
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 79.0,
+      "learning_rate": 4.14256237997488e-07,
+      "loss": 1.545,
+      "step": 30980
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 68.125,
+      "learning_rate": 4.138780944666824e-07,
+      "loss": 1.5746,
+      "step": 31000
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 56.15625,
+      "learning_rate": 4.134999509358769e-07,
+      "loss": 1.5426,
+      "step": 31020
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 53.6875,
+      "learning_rate": 4.131218074050713e-07,
+      "loss": 1.5182,
+      "step": 31040
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 89.4375,
+      "learning_rate": 4.127436638742657e-07,
+      "loss": 1.4772,
+      "step": 31060
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 66.375,
+      "learning_rate": 4.123655203434602e-07,
+      "loss": 1.5478,
+      "step": 31080
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 65.8125,
+      "learning_rate": 4.119873768126546e-07,
+      "loss": 1.4511,
+      "step": 31100
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 56.5,
+      "learning_rate": 4.1160923328184904e-07,
+      "loss": 1.5412,
+      "step": 31120
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 50.8125,
+      "learning_rate": 4.1123108975104346e-07,
+      "loss": 1.6,
+      "step": 31140
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 71.5625,
+      "learning_rate": 4.10852946220238e-07,
+      "loss": 1.4424,
+      "step": 31160
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 59.28125,
+      "learning_rate": 4.104748026894324e-07,
+      "loss": 1.5589,
+      "step": 31180
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 56.6875,
+      "learning_rate": 4.1009665915862683e-07,
+      "loss": 1.5787,
+      "step": 31200
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 61.6875,
+      "learning_rate": 4.097185156278213e-07,
+      "loss": 1.4297,
+      "step": 31220
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 64.4375,
+      "learning_rate": 4.093403720970157e-07,
+      "loss": 1.557,
+      "step": 31240
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 49.65625,
+      "learning_rate": 4.0896222856621014e-07,
+      "loss": 1.518,
+      "step": 31260
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 59.5,
+      "learning_rate": 4.0858408503540457e-07,
+      "loss": 1.5745,
+      "step": 31280
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 58.25,
+      "learning_rate": 4.0820594150459904e-07,
+      "loss": 1.4711,
+      "step": 31300
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 66.625,
+      "learning_rate": 4.0782779797379346e-07,
+      "loss": 1.5911,
+      "step": 31320
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 78.6875,
+      "learning_rate": 4.0744965444298794e-07,
+      "loss": 1.5814,
+      "step": 31340
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 61.25,
+      "learning_rate": 4.070715109121824e-07,
+      "loss": 1.4427,
+      "step": 31360
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 69.875,
+      "learning_rate": 4.0669336738137683e-07,
+      "loss": 1.4784,
+      "step": 31380
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 50.6875,
+      "learning_rate": 4.0631522385057125e-07,
+      "loss": 1.5007,
+      "step": 31400
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 66.8125,
+      "learning_rate": 4.0593708031976573e-07,
+      "loss": 1.6264,
+      "step": 31420
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 62.65625,
+      "learning_rate": 4.0555893678896015e-07,
+      "loss": 1.5666,
+      "step": 31440
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 71.5,
+      "learning_rate": 4.0518079325815457e-07,
+      "loss": 1.736,
+      "step": 31460
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 69.1875,
+      "learning_rate": 4.04802649727349e-07,
+      "loss": 1.543,
+      "step": 31480
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 70.1875,
+      "learning_rate": 4.044245061965435e-07,
+      "loss": 1.5761,
+      "step": 31500
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 65.75,
+      "learning_rate": 4.0404636266573794e-07,
+      "loss": 1.5896,
+      "step": 31520
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 54.65625,
+      "learning_rate": 4.0366821913493236e-07,
+      "loss": 1.4913,
+      "step": 31540
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 73.0625,
+      "learning_rate": 4.0329007560412683e-07,
+      "loss": 1.5724,
+      "step": 31560
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 66.3125,
+      "learning_rate": 4.0291193207332126e-07,
+      "loss": 1.5511,
+      "step": 31580
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 68.5625,
+      "learning_rate": 4.025337885425157e-07,
+      "loss": 1.5837,
+      "step": 31600
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 62.375,
+      "learning_rate": 4.021556450117101e-07,
+      "loss": 1.5772,
+      "step": 31620
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 69.25,
+      "learning_rate": 4.0177750148090457e-07,
+      "loss": 1.5872,
+      "step": 31640
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 69.0625,
+      "learning_rate": 4.01399357950099e-07,
+      "loss": 1.5467,
+      "step": 31660
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 54.21875,
+      "learning_rate": 4.0102121441929347e-07,
+      "loss": 1.5222,
+      "step": 31680
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 67.6875,
+      "learning_rate": 4.0064307088848794e-07,
+      "loss": 1.5318,
+      "step": 31700
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 77.0625,
+      "learning_rate": 4.0026492735768236e-07,
+      "loss": 1.5116,
+      "step": 31720
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 63.15625,
+      "learning_rate": 3.998867838268768e-07,
+      "loss": 1.4655,
+      "step": 31740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 72.25,
+      "learning_rate": 3.995086402960712e-07,
+      "loss": 1.3648,
+      "step": 31760
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 50.46875,
+      "learning_rate": 3.991304967652657e-07,
+      "loss": 1.3922,
+      "step": 31780
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 52.625,
+      "learning_rate": 3.987523532344601e-07,
+      "loss": 1.4272,
+      "step": 31800
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 68.375,
+      "learning_rate": 3.983742097036545e-07,
+      "loss": 1.3381,
+      "step": 31820
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 61.75,
+      "learning_rate": 3.97996066172849e-07,
+      "loss": 1.3762,
+      "step": 31840
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 64.0,
+      "learning_rate": 3.9761792264204347e-07,
+      "loss": 1.36,
+      "step": 31860
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 54.15625,
+      "learning_rate": 3.972397791112379e-07,
+      "loss": 1.3718,
+      "step": 31880
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 57.75,
+      "learning_rate": 3.9686163558043237e-07,
+      "loss": 1.4966,
+      "step": 31900
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 75.0625,
+      "learning_rate": 3.964834920496268e-07,
+      "loss": 1.4002,
+      "step": 31920
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 53.6875,
+      "learning_rate": 3.961053485188212e-07,
+      "loss": 1.3682,
+      "step": 31940
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 60.3125,
+      "learning_rate": 3.9572720498801563e-07,
+      "loss": 1.4108,
+      "step": 31960
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 100.0625,
+      "learning_rate": 3.953490614572101e-07,
+      "loss": 1.4387,
+      "step": 31980
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 61.40625,
+      "learning_rate": 3.9497091792640453e-07,
+      "loss": 1.3657,
+      "step": 32000
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 57.0625,
+      "learning_rate": 3.94592774395599e-07,
+      "loss": 1.3044,
+      "step": 32020
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 75.625,
+      "learning_rate": 3.942146308647935e-07,
+      "loss": 1.5226,
+      "step": 32040
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 79.125,
+      "learning_rate": 3.938364873339879e-07,
+      "loss": 1.5141,
+      "step": 32060
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 77.125,
+      "learning_rate": 3.934583438031823e-07,
+      "loss": 1.3971,
+      "step": 32080
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 74.6875,
+      "learning_rate": 3.9308020027237674e-07,
+      "loss": 1.3343,
+      "step": 32100
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 58.6875,
+      "learning_rate": 3.927020567415712e-07,
+      "loss": 1.3938,
+      "step": 32120
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 57.375,
+      "learning_rate": 3.9232391321076564e-07,
+      "loss": 1.4423,
+      "step": 32140
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 64.875,
+      "learning_rate": 3.9194576967996006e-07,
+      "loss": 1.4377,
+      "step": 32160
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 57.78125,
+      "learning_rate": 3.9156762614915453e-07,
+      "loss": 1.416,
+      "step": 32180
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 52.8125,
+      "learning_rate": 3.91189482618349e-07,
+      "loss": 1.4056,
+      "step": 32200
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 52.375,
+      "learning_rate": 3.908113390875434e-07,
+      "loss": 1.4797,
+      "step": 32220
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 66.0625,
+      "learning_rate": 3.9043319555673785e-07,
+      "loss": 1.3557,
+      "step": 32240
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 67.5,
+      "learning_rate": 3.900550520259323e-07,
+      "loss": 1.3408,
+      "step": 32260
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 61.5625,
+      "learning_rate": 3.8967690849512674e-07,
+      "loss": 1.4133,
+      "step": 32280
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 70.9375,
+      "learning_rate": 3.8929876496432116e-07,
+      "loss": 1.4408,
+      "step": 32300
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 60.65625,
+      "learning_rate": 3.8892062143351564e-07,
+      "loss": 1.3561,
+      "step": 32320
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 57.78125,
+      "learning_rate": 3.8854247790271006e-07,
+      "loss": 1.4019,
+      "step": 32340
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 54.75,
+      "learning_rate": 3.881643343719045e-07,
+      "loss": 1.2897,
+      "step": 32360
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 55.5,
+      "learning_rate": 3.8778619084109896e-07,
+      "loss": 1.4233,
+      "step": 32380
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 57.90625,
+      "learning_rate": 3.8740804731029343e-07,
+      "loss": 1.3995,
+      "step": 32400
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 60.0625,
+      "learning_rate": 3.8702990377948785e-07,
+      "loss": 1.3652,
+      "step": 32420
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 62.25,
+      "learning_rate": 3.8665176024868227e-07,
+      "loss": 1.4912,
+      "step": 32440
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 69.0625,
+      "learning_rate": 3.8627361671787675e-07,
+      "loss": 1.369,
+      "step": 32460
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 50.65625,
+      "learning_rate": 3.8589547318707117e-07,
+      "loss": 1.426,
+      "step": 32480
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 57.90625,
+      "learning_rate": 3.855173296562656e-07,
+      "loss": 1.2934,
+      "step": 32500
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 70.875,
+      "learning_rate": 3.8513918612546e-07,
+      "loss": 1.4486,
+      "step": 32520
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 70.4375,
+      "learning_rate": 3.847610425946545e-07,
+      "loss": 1.5197,
+      "step": 32540
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 60.625,
+      "learning_rate": 3.8438289906384896e-07,
+      "loss": 1.4181,
+      "step": 32560
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 70.75,
+      "learning_rate": 3.840047555330434e-07,
+      "loss": 1.3959,
+      "step": 32580
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 89.6875,
+      "learning_rate": 3.8362661200223785e-07,
+      "loss": 1.3343,
+      "step": 32600
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 55.75,
+      "learning_rate": 3.832484684714323e-07,
+      "loss": 1.4992,
+      "step": 32620
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 72.875,
+      "learning_rate": 3.828703249406267e-07,
+      "loss": 1.4326,
+      "step": 32640
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 59.40625,
+      "learning_rate": 3.8249218140982117e-07,
+      "loss": 1.4861,
+      "step": 32660
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 50.625,
+      "learning_rate": 3.821140378790156e-07,
+      "loss": 1.4226,
+      "step": 32680
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 56.3125,
+      "learning_rate": 3.8173589434821e-07,
+      "loss": 1.383,
+      "step": 32700
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 56.09375,
+      "learning_rate": 3.813577508174045e-07,
+      "loss": 1.5258,
+      "step": 32720
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 68.4375,
+      "learning_rate": 3.8097960728659896e-07,
+      "loss": 1.3827,
+      "step": 32740
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 52.09375,
+      "learning_rate": 3.806014637557934e-07,
+      "loss": 1.4179,
+      "step": 32760
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 55.5625,
+      "learning_rate": 3.802233202249878e-07,
+      "loss": 1.3972,
+      "step": 32780
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 51.625,
+      "learning_rate": 3.798451766941823e-07,
+      "loss": 1.4764,
+      "step": 32800
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 75.3125,
+      "learning_rate": 3.794670331633767e-07,
+      "loss": 1.4087,
+      "step": 32820
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 67.8125,
+      "learning_rate": 3.790888896325711e-07,
+      "loss": 1.4917,
+      "step": 32840
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 59.4375,
+      "learning_rate": 3.7871074610176554e-07,
+      "loss": 1.3819,
+      "step": 32860
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 56.0625,
+      "learning_rate": 3.7833260257096e-07,
+      "loss": 1.3787,
+      "step": 32880
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 64.0625,
+      "learning_rate": 3.779544590401545e-07,
+      "loss": 1.4065,
+      "step": 32900
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 59.375,
+      "learning_rate": 3.775763155093489e-07,
+      "loss": 1.427,
+      "step": 32920
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 61.09375,
+      "learning_rate": 3.771981719785434e-07,
+      "loss": 1.3911,
+      "step": 32940
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 65.9375,
+      "learning_rate": 3.768200284477378e-07,
+      "loss": 1.3648,
+      "step": 32960
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 66.6875,
+      "learning_rate": 3.7644188491693223e-07,
+      "loss": 1.4081,
+      "step": 32980
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 55.40625,
+      "learning_rate": 3.7606374138612665e-07,
+      "loss": 1.4276,
+      "step": 33000
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 54.34375,
+      "learning_rate": 3.756855978553211e-07,
+      "loss": 1.4641,
+      "step": 33020
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 64.9375,
+      "learning_rate": 3.7530745432451555e-07,
+      "loss": 1.35,
+      "step": 33040
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 46.28125,
+      "learning_rate": 3.7492931079370997e-07,
+      "loss": 1.4257,
+      "step": 33060
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 57.6875,
+      "learning_rate": 3.745511672629045e-07,
+      "loss": 1.4374,
+      "step": 33080
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 63.09375,
+      "learning_rate": 3.741730237320989e-07,
+      "loss": 1.4295,
+      "step": 33100
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 69.4375,
+      "learning_rate": 3.7379488020129334e-07,
+      "loss": 1.3691,
+      "step": 33120
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 63.34375,
+      "learning_rate": 3.734167366704878e-07,
+      "loss": 1.4291,
+      "step": 33140
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 52.3125,
+      "learning_rate": 3.7303859313968223e-07,
+      "loss": 1.4837,
+      "step": 33160
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 81.8125,
+      "learning_rate": 3.7266044960887665e-07,
+      "loss": 1.4837,
+      "step": 33180
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 57.09375,
+      "learning_rate": 3.722823060780711e-07,
+      "loss": 1.3974,
+      "step": 33200
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 65.1875,
+      "learning_rate": 3.7190416254726555e-07,
+      "loss": 1.3418,
+      "step": 33220
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 70.25,
+      "learning_rate": 3.7152601901645997e-07,
+      "loss": 1.4303,
+      "step": 33240
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 54.75,
+      "learning_rate": 3.7114787548565445e-07,
+      "loss": 1.4754,
+      "step": 33260
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 65.8125,
+      "learning_rate": 3.707697319548489e-07,
+      "loss": 1.4474,
+      "step": 33280
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 61.4375,
+      "learning_rate": 3.7039158842404334e-07,
+      "loss": 1.3412,
+      "step": 33300
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 80.0,
+      "learning_rate": 3.7001344489323776e-07,
+      "loss": 1.3211,
+      "step": 33320
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 62.0,
+      "learning_rate": 3.696353013624322e-07,
+      "loss": 1.3295,
+      "step": 33340
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 70.375,
+      "learning_rate": 3.6925715783162666e-07,
+      "loss": 1.4678,
+      "step": 33360
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 59.125,
+      "learning_rate": 3.688790143008211e-07,
+      "loss": 1.3333,
+      "step": 33380
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 63.5625,
+      "learning_rate": 3.685008707700155e-07,
+      "loss": 1.3726,
+      "step": 33400
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 64.625,
+      "learning_rate": 3.6812272723921003e-07,
+      "loss": 1.4712,
+      "step": 33420
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 64.9375,
+      "learning_rate": 3.6774458370840445e-07,
+      "loss": 1.3021,
+      "step": 33440
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 56.5625,
+      "learning_rate": 3.6736644017759887e-07,
+      "loss": 1.3805,
+      "step": 33460
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 59.1875,
+      "learning_rate": 3.669882966467933e-07,
+      "loss": 1.4557,
+      "step": 33480
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 64.1875,
+      "learning_rate": 3.6661015311598777e-07,
+      "loss": 1.4616,
+      "step": 33500
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 68.6875,
+      "learning_rate": 3.662320095851822e-07,
+      "loss": 1.3874,
+      "step": 33520
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 56.59375,
+      "learning_rate": 3.658538660543766e-07,
+      "loss": 1.4573,
+      "step": 33540
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 70.4375,
+      "learning_rate": 3.654757225235711e-07,
+      "loss": 1.3524,
+      "step": 33560
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 69.75,
+      "learning_rate": 3.650975789927655e-07,
+      "loss": 1.3711,
+      "step": 33580
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 68.75,
+      "learning_rate": 3.6471943546196e-07,
+      "loss": 1.3215,
+      "step": 33600
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 51.875,
+      "learning_rate": 3.6434129193115445e-07,
+      "loss": 1.4378,
+      "step": 33620
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 53.5,
+      "learning_rate": 3.639631484003489e-07,
+      "loss": 1.4243,
+      "step": 33640
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 118.8125,
+      "learning_rate": 3.635850048695433e-07,
+      "loss": 1.3906,
+      "step": 33660
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 49.21875,
+      "learning_rate": 3.632068613387377e-07,
+      "loss": 1.3534,
+      "step": 33680
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 61.1875,
+      "learning_rate": 3.628287178079322e-07,
+      "loss": 1.423,
+      "step": 33700
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 72.3125,
+      "learning_rate": 3.624505742771266e-07,
+      "loss": 1.4193,
+      "step": 33720
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 66.625,
+      "learning_rate": 3.6207243074632103e-07,
+      "loss": 1.4438,
+      "step": 33740
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 70.9375,
+      "learning_rate": 3.6169428721551545e-07,
+      "loss": 1.4021,
+      "step": 33760
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 64.4375,
+      "learning_rate": 3.6131614368471e-07,
+      "loss": 1.3662,
+      "step": 33780
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 74.6875,
+      "learning_rate": 3.609380001539044e-07,
+      "loss": 1.4504,
+      "step": 33800
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 83.25,
+      "learning_rate": 3.605598566230988e-07,
+      "loss": 1.3239,
+      "step": 33820
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 78.375,
+      "learning_rate": 3.601817130922933e-07,
+      "loss": 1.4084,
+      "step": 33840
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 78.5625,
+      "learning_rate": 3.598035695614877e-07,
+      "loss": 1.4374,
+      "step": 33860
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 67.5,
+      "learning_rate": 3.5942542603068214e-07,
+      "loss": 1.3577,
+      "step": 33880
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 54.09375,
+      "learning_rate": 3.590472824998766e-07,
+      "loss": 1.3353,
+      "step": 33900
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 53.34375,
+      "learning_rate": 3.5866913896907104e-07,
+      "loss": 1.3715,
+      "step": 33920
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 56.6875,
+      "learning_rate": 3.582909954382655e-07,
+      "loss": 1.3706,
+      "step": 33940
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 84.9375,
+      "learning_rate": 3.5791285190745993e-07,
+      "loss": 1.384,
+      "step": 33960
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 57.84375,
+      "learning_rate": 3.575347083766544e-07,
+      "loss": 1.3953,
+      "step": 33980
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 78.625,
+      "learning_rate": 3.5715656484584883e-07,
+      "loss": 1.3413,
+      "step": 34000
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 82.25,
+      "learning_rate": 3.5677842131504325e-07,
+      "loss": 1.3603,
+      "step": 34020
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 62.34375,
+      "learning_rate": 3.564002777842377e-07,
+      "loss": 1.3919,
+      "step": 34040
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 57.0,
+      "learning_rate": 3.5602213425343214e-07,
+      "loss": 1.4083,
+      "step": 34060
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 77.875,
+      "learning_rate": 3.5564399072262657e-07,
+      "loss": 1.4505,
+      "step": 34080
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 77.3125,
+      "learning_rate": 3.55265847191821e-07,
+      "loss": 1.3939,
+      "step": 34100
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 70.3125,
+      "learning_rate": 3.548877036610155e-07,
+      "loss": 1.4308,
+      "step": 34120
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 59.0,
+      "learning_rate": 3.5450956013020994e-07,
+      "loss": 1.3746,
+      "step": 34140
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 68.375,
+      "learning_rate": 3.5413141659940436e-07,
+      "loss": 1.3084,
+      "step": 34160
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 59.75,
+      "learning_rate": 3.5375327306859883e-07,
+      "loss": 1.4063,
+      "step": 34180
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 62.4375,
+      "learning_rate": 3.5337512953779325e-07,
+      "loss": 1.4903,
+      "step": 34200
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 63.875,
+      "learning_rate": 3.529969860069877e-07,
+      "loss": 1.4406,
+      "step": 34220
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 61.71875,
+      "learning_rate": 3.526188424761821e-07,
+      "loss": 1.51,
+      "step": 34240
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 73.1875,
+      "learning_rate": 3.5224069894537657e-07,
+      "loss": 1.375,
+      "step": 34260
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 60.40625,
+      "learning_rate": 3.51862555414571e-07,
+      "loss": 1.3797,
+      "step": 34280
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 55.25,
+      "learning_rate": 3.5148441188376547e-07,
+      "loss": 1.3388,
+      "step": 34300
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 96.0,
+      "learning_rate": 3.5110626835295994e-07,
+      "loss": 1.4355,
+      "step": 34320
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 65.125,
+      "learning_rate": 3.5072812482215436e-07,
+      "loss": 1.3663,
+      "step": 34340
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 56.59375,
+      "learning_rate": 3.503499812913488e-07,
+      "loss": 1.3666,
+      "step": 34360
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 62.875,
+      "learning_rate": 3.4997183776054326e-07,
+      "loss": 1.3237,
+      "step": 34380
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 47.59375,
+      "learning_rate": 3.495936942297377e-07,
+      "loss": 1.5043,
+      "step": 34400
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 64.75,
+      "learning_rate": 3.492155506989321e-07,
+      "loss": 1.3668,
+      "step": 34420
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 76.0,
+      "learning_rate": 3.488374071681265e-07,
+      "loss": 1.4768,
+      "step": 34440
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 55.625,
+      "learning_rate": 3.48459263637321e-07,
+      "loss": 1.3829,
+      "step": 34460
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 55.375,
+      "learning_rate": 3.4808112010651547e-07,
+      "loss": 1.3135,
+      "step": 34480
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 57.8125,
+      "learning_rate": 3.477029765757099e-07,
+      "loss": 1.3666,
+      "step": 34500
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 62.34375,
+      "learning_rate": 3.4732483304490436e-07,
+      "loss": 1.3976,
+      "step": 34520
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 56.21875,
+      "learning_rate": 3.469466895140988e-07,
+      "loss": 1.4472,
+      "step": 34540
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 79.125,
+      "learning_rate": 3.465685459832932e-07,
+      "loss": 1.3524,
+      "step": 34560
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 82.375,
+      "learning_rate": 3.4619040245248763e-07,
+      "loss": 1.3981,
+      "step": 34580
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 57.21875,
+      "learning_rate": 3.458122589216821e-07,
+      "loss": 1.3506,
+      "step": 34600
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 73.8125,
+      "learning_rate": 3.454341153908765e-07,
+      "loss": 1.4456,
+      "step": 34620
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 97.1875,
+      "learning_rate": 3.45055971860071e-07,
+      "loss": 1.3992,
+      "step": 34640
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 81.5625,
+      "learning_rate": 3.4467782832926547e-07,
+      "loss": 1.4132,
+      "step": 34660
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 103.8125,
+      "learning_rate": 3.442996847984599e-07,
+      "loss": 1.3113,
+      "step": 34680
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 63.34375,
+      "learning_rate": 3.439215412676543e-07,
+      "loss": 1.4059,
+      "step": 34700
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 114.6875,
+      "learning_rate": 3.4354339773684874e-07,
+      "loss": 1.4594,
+      "step": 34720
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 73.5,
+      "learning_rate": 3.431652542060432e-07,
+      "loss": 1.269,
+      "step": 34740
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 67.5625,
+      "learning_rate": 3.4278711067523763e-07,
+      "loss": 1.3647,
+      "step": 34760
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 79.75,
+      "learning_rate": 3.4240896714443205e-07,
+      "loss": 1.4436,
+      "step": 34780
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 53.625,
+      "learning_rate": 3.4203082361362653e-07,
+      "loss": 1.3997,
+      "step": 34800
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 66.125,
+      "learning_rate": 3.41652680082821e-07,
+      "loss": 1.3917,
+      "step": 34820
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 60.1875,
+      "learning_rate": 3.412745365520154e-07,
+      "loss": 1.3912,
+      "step": 34840
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 65.5,
+      "learning_rate": 3.408963930212099e-07,
+      "loss": 1.468,
+      "step": 34860
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 81.25,
+      "learning_rate": 3.405182494904043e-07,
+      "loss": 1.4517,
+      "step": 34880
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 74.0625,
+      "learning_rate": 3.4014010595959874e-07,
+      "loss": 1.5081,
+      "step": 34900
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 70.125,
+      "learning_rate": 3.3976196242879316e-07,
+      "loss": 1.2948,
+      "step": 34920
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 49.84375,
+      "learning_rate": 3.3938381889798764e-07,
+      "loss": 1.2734,
+      "step": 34940
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 60.5,
+      "learning_rate": 3.3900567536718206e-07,
+      "loss": 1.5046,
+      "step": 34960
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 104.375,
+      "learning_rate": 3.386275318363765e-07,
+      "loss": 1.4961,
+      "step": 34980
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 61.78125,
+      "learning_rate": 3.38249388305571e-07,
+      "loss": 1.4419,
+      "step": 35000
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 60.03125,
+      "learning_rate": 3.378712447747654e-07,
+      "loss": 1.4329,
+      "step": 35020
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 56.90625,
+      "learning_rate": 3.3749310124395985e-07,
+      "loss": 1.4915,
+      "step": 35040
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 83.3125,
+      "learning_rate": 3.3711495771315427e-07,
+      "loss": 1.4082,
+      "step": 35060
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 57.0625,
+      "learning_rate": 3.3673681418234874e-07,
+      "loss": 1.4659,
+      "step": 35080
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 68.75,
+      "learning_rate": 3.3635867065154316e-07,
+      "loss": 1.4333,
+      "step": 35100
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 84.375,
+      "learning_rate": 3.359805271207376e-07,
+      "loss": 1.4449,
+      "step": 35120
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 62.96875,
+      "learning_rate": 3.3560238358993206e-07,
+      "loss": 1.4052,
+      "step": 35140
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 58.46875,
+      "learning_rate": 3.352242400591265e-07,
+      "loss": 1.4209,
+      "step": 35160
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 94.875,
+      "learning_rate": 3.3484609652832096e-07,
+      "loss": 1.4097,
+      "step": 35180
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 62.84375,
+      "learning_rate": 3.344679529975154e-07,
+      "loss": 1.4691,
+      "step": 35200
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 50.25,
+      "learning_rate": 3.3408980946670985e-07,
+      "loss": 1.3639,
+      "step": 35220
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 71.0,
+      "learning_rate": 3.3371166593590427e-07,
+      "loss": 1.3215,
+      "step": 35240
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 82.5,
+      "learning_rate": 3.333335224050987e-07,
+      "loss": 1.428,
+      "step": 35260
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 80.4375,
+      "learning_rate": 3.3295537887429317e-07,
+      "loss": 1.4102,
+      "step": 35280
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 62.8125,
+      "learning_rate": 3.325772353434876e-07,
+      "loss": 1.3712,
+      "step": 35300
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 65.9375,
+      "learning_rate": 3.32199091812682e-07,
+      "loss": 1.4205,
+      "step": 35320
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 72.0,
+      "learning_rate": 3.3182094828187654e-07,
+      "loss": 1.3905,
+      "step": 35340
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 75.0,
+      "learning_rate": 3.3144280475107096e-07,
+      "loss": 1.4007,
+      "step": 35360
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 69.125,
+      "learning_rate": 3.310646612202654e-07,
+      "loss": 1.4627,
+      "step": 35380
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 92.25,
+      "learning_rate": 3.306865176894598e-07,
+      "loss": 1.3625,
+      "step": 35400
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 56.5,
+      "learning_rate": 3.303083741586543e-07,
+      "loss": 1.4248,
+      "step": 35420
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 56.34375,
+      "learning_rate": 3.299302306278487e-07,
+      "loss": 1.384,
+      "step": 35440
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 60.875,
+      "learning_rate": 3.295520870970431e-07,
+      "loss": 1.4455,
+      "step": 35460
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 58.90625,
+      "learning_rate": 3.2917394356623754e-07,
+      "loss": 1.4998,
+      "step": 35480
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 64.375,
+      "learning_rate": 3.28795800035432e-07,
+      "loss": 1.4044,
+      "step": 35500
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 65.0,
+      "learning_rate": 3.284176565046265e-07,
+      "loss": 1.3424,
+      "step": 35520
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 78.1875,
+      "learning_rate": 3.280395129738209e-07,
+      "loss": 1.3932,
+      "step": 35540
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 115.625,
+      "learning_rate": 3.276613694430154e-07,
+      "loss": 1.4185,
+      "step": 35560
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 51.8125,
+      "learning_rate": 3.272832259122098e-07,
+      "loss": 1.357,
+      "step": 35580
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 81.0,
+      "learning_rate": 3.2690508238140423e-07,
+      "loss": 1.3221,
+      "step": 35600
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 58.84375,
+      "learning_rate": 3.265269388505987e-07,
+      "loss": 1.3425,
+      "step": 35620
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 79.8125,
+      "learning_rate": 3.261487953197931e-07,
+      "loss": 1.5156,
+      "step": 35640
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 99.625,
+      "learning_rate": 3.2577065178898754e-07,
+      "loss": 1.4331,
+      "step": 35660
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 64.4375,
+      "learning_rate": 3.2539250825818196e-07,
+      "loss": 1.4773,
+      "step": 35680
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 57.9375,
+      "learning_rate": 3.250143647273765e-07,
+      "loss": 1.386,
+      "step": 35700
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 67.6875,
+      "learning_rate": 3.246362211965709e-07,
+      "loss": 1.4126,
+      "step": 35720
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 79.6875,
+      "learning_rate": 3.2425807766576533e-07,
+      "loss": 1.4665,
+      "step": 35740
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 53.125,
+      "learning_rate": 3.238799341349598e-07,
+      "loss": 1.3703,
+      "step": 35760
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 53.0,
+      "learning_rate": 3.2350179060415423e-07,
+      "loss": 1.4106,
+      "step": 35780
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 54.09375,
+      "learning_rate": 3.2312364707334865e-07,
+      "loss": 1.3638,
+      "step": 35800
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 86.25,
+      "learning_rate": 3.2274550354254307e-07,
+      "loss": 1.4043,
+      "step": 35820
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 70.25,
+      "learning_rate": 3.2236736001173755e-07,
+      "loss": 1.4668,
+      "step": 35840
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 46.53125,
+      "learning_rate": 3.2198921648093197e-07,
+      "loss": 1.4028,
+      "step": 35860
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 62.8125,
+      "learning_rate": 3.2161107295012644e-07,
+      "loss": 1.4119,
+      "step": 35880
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 62.8125,
+      "learning_rate": 3.212329294193209e-07,
+      "loss": 1.3546,
+      "step": 35900
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 59.78125,
+      "learning_rate": 3.2085478588851534e-07,
+      "loss": 1.4402,
+      "step": 35920
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 63.1875,
+      "learning_rate": 3.2047664235770976e-07,
+      "loss": 1.3611,
+      "step": 35940
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 55.875,
+      "learning_rate": 3.200984988269042e-07,
+      "loss": 1.4106,
+      "step": 35960
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 61.53125,
+      "learning_rate": 3.1972035529609865e-07,
+      "loss": 1.4203,
+      "step": 35980
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 56.03125,
+      "learning_rate": 3.193422117652931e-07,
+      "loss": 1.4757,
+      "step": 36000
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 62.4375,
+      "learning_rate": 3.189640682344875e-07,
+      "loss": 1.4443,
+      "step": 36020
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 74.125,
+      "learning_rate": 3.18585924703682e-07,
+      "loss": 1.3875,
+      "step": 36040
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 72.4375,
+      "learning_rate": 3.1820778117287645e-07,
+      "loss": 1.3827,
+      "step": 36060
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 63.84375,
+      "learning_rate": 3.1782963764207087e-07,
+      "loss": 1.4127,
+      "step": 36080
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 58.0,
+      "learning_rate": 3.1745149411126534e-07,
+      "loss": 1.4644,
+      "step": 36100
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 67.3125,
+      "learning_rate": 3.1707335058045976e-07,
+      "loss": 1.3745,
+      "step": 36120
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 73.25,
+      "learning_rate": 3.166952070496542e-07,
+      "loss": 1.4179,
+      "step": 36140
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 58.1875,
+      "learning_rate": 3.163170635188486e-07,
+      "loss": 1.4096,
+      "step": 36160
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 70.125,
+      "learning_rate": 3.159389199880431e-07,
+      "loss": 1.4569,
+      "step": 36180
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 55.6875,
+      "learning_rate": 3.155607764572375e-07,
+      "loss": 1.338,
+      "step": 36200
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 70.5,
+      "learning_rate": 3.15182632926432e-07,
+      "loss": 1.4023,
+      "step": 36220
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 351.5,
+      "learning_rate": 3.1480448939562645e-07,
+      "loss": 1.4474,
+      "step": 36240
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 69.4375,
+      "learning_rate": 3.1442634586482087e-07,
+      "loss": 1.3545,
+      "step": 36260
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 69.875,
+      "learning_rate": 3.140482023340153e-07,
+      "loss": 1.3057,
+      "step": 36280
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 48.0625,
+      "learning_rate": 3.136700588032097e-07,
+      "loss": 1.4568,
+      "step": 36300
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 81.625,
+      "learning_rate": 3.132919152724042e-07,
+      "loss": 1.3281,
+      "step": 36320
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 52.8125,
+      "learning_rate": 3.129137717415986e-07,
+      "loss": 1.4704,
+      "step": 36340
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 65.1875,
+      "learning_rate": 3.1253562821079303e-07,
+      "loss": 1.4349,
+      "step": 36360
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 76.125,
+      "learning_rate": 3.121574846799875e-07,
+      "loss": 1.4236,
+      "step": 36380
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 50.34375,
+      "learning_rate": 3.11779341149182e-07,
+      "loss": 1.2999,
+      "step": 36400
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 61.28125,
+      "learning_rate": 3.114011976183764e-07,
+      "loss": 1.3776,
+      "step": 36420
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 71.1875,
+      "learning_rate": 3.110230540875708e-07,
+      "loss": 1.3691,
+      "step": 36440
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 51.34375,
+      "learning_rate": 3.106449105567653e-07,
+      "loss": 1.335,
+      "step": 36460
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 60.625,
+      "learning_rate": 3.102667670259597e-07,
+      "loss": 1.2949,
+      "step": 36480
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 55.96875,
+      "learning_rate": 3.0988862349515414e-07,
+      "loss": 1.402,
+      "step": 36500
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 76.25,
+      "learning_rate": 3.095104799643486e-07,
+      "loss": 1.4453,
+      "step": 36520
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 64.8125,
+      "learning_rate": 3.0913233643354303e-07,
+      "loss": 1.3819,
+      "step": 36540
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 54.96875,
+      "learning_rate": 3.087541929027375e-07,
+      "loss": 1.3811,
+      "step": 36560
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 57.84375,
+      "learning_rate": 3.08376049371932e-07,
+      "loss": 1.4251,
+      "step": 36580
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 57.8125,
+      "learning_rate": 3.079979058411264e-07,
+      "loss": 1.4553,
+      "step": 36600
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 53.875,
+      "learning_rate": 3.076197623103208e-07,
+      "loss": 1.3474,
+      "step": 36620
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 63.6875,
+      "learning_rate": 3.0724161877951525e-07,
+      "loss": 1.4288,
+      "step": 36640
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 55.78125,
+      "learning_rate": 3.068634752487097e-07,
+      "loss": 1.431,
+      "step": 36660
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 63.46875,
+      "learning_rate": 3.0648533171790414e-07,
+      "loss": 1.3994,
+      "step": 36680
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 90.375,
+      "learning_rate": 3.0610718818709856e-07,
+      "loss": 1.4356,
+      "step": 36700
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 71.0625,
+      "learning_rate": 3.0572904465629304e-07,
+      "loss": 1.3515,
+      "step": 36720
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 67.1875,
+      "learning_rate": 3.053509011254875e-07,
+      "loss": 1.3501,
+      "step": 36740
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 66.25,
+      "learning_rate": 3.0497275759468193e-07,
+      "loss": 1.4016,
+      "step": 36760
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 69.4375,
+      "learning_rate": 3.0459461406387635e-07,
+      "loss": 1.3546,
+      "step": 36780
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 51.0625,
+      "learning_rate": 3.0421647053307083e-07,
+      "loss": 1.4189,
+      "step": 36800
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 72.0625,
+      "learning_rate": 3.0383832700226525e-07,
+      "loss": 1.4171,
+      "step": 36820
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 54.59375,
+      "learning_rate": 3.0346018347145967e-07,
+      "loss": 1.4448,
+      "step": 36840
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 49.0625,
+      "learning_rate": 3.0308203994065415e-07,
+      "loss": 1.4652,
+      "step": 36860
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 72.4375,
+      "learning_rate": 3.0270389640984857e-07,
+      "loss": 1.41,
+      "step": 36880
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 67.1875,
+      "learning_rate": 3.02325752879043e-07,
+      "loss": 1.3746,
+      "step": 36900
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 86.5,
+      "learning_rate": 3.0194760934823746e-07,
+      "loss": 1.3318,
+      "step": 36920
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 61.90625,
+      "learning_rate": 3.0156946581743194e-07,
+      "loss": 1.3764,
+      "step": 36940
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 63.375,
+      "learning_rate": 3.0119132228662636e-07,
+      "loss": 1.4998,
+      "step": 36960
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 52.8125,
+      "learning_rate": 3.008131787558208e-07,
+      "loss": 1.4612,
+      "step": 36980
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 62.125,
+      "learning_rate": 3.0043503522501525e-07,
+      "loss": 1.3757,
+      "step": 37000
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 56.5625,
+      "learning_rate": 3.000568916942097e-07,
+      "loss": 1.3478,
+      "step": 37020
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 87.3125,
+      "learning_rate": 2.996787481634041e-07,
+      "loss": 1.3444,
+      "step": 37040
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 59.375,
+      "learning_rate": 2.993006046325985e-07,
+      "loss": 1.4231,
+      "step": 37060
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 64.625,
+      "learning_rate": 2.98922461101793e-07,
+      "loss": 1.4394,
+      "step": 37080
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 58.21875,
+      "learning_rate": 2.9854431757098747e-07,
+      "loss": 1.3933,
+      "step": 37100
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 87.5625,
+      "learning_rate": 2.981661740401819e-07,
+      "loss": 1.4413,
+      "step": 37120
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 63.03125,
+      "learning_rate": 2.9778803050937636e-07,
+      "loss": 1.513,
+      "step": 37140
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 87.375,
+      "learning_rate": 2.974098869785708e-07,
+      "loss": 1.3841,
+      "step": 37160
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 69.6875,
+      "learning_rate": 2.970317434477652e-07,
+      "loss": 1.4671,
+      "step": 37180
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 63.5625,
+      "learning_rate": 2.966535999169596e-07,
+      "loss": 1.3005,
+      "step": 37200
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 58.96875,
+      "learning_rate": 2.962754563861541e-07,
+      "loss": 1.3706,
+      "step": 37220
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 59.0625,
+      "learning_rate": 2.958973128553485e-07,
+      "loss": 1.3605,
+      "step": 37240
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 53.15625,
+      "learning_rate": 2.95519169324543e-07,
+      "loss": 1.4114,
+      "step": 37260
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 76.8125,
+      "learning_rate": 2.9514102579373747e-07,
+      "loss": 1.3111,
+      "step": 37280
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 66.9375,
+      "learning_rate": 2.947628822629319e-07,
+      "loss": 1.3771,
+      "step": 37300
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 72.5625,
+      "learning_rate": 2.943847387321263e-07,
+      "loss": 1.3847,
+      "step": 37320
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 55.4375,
+      "learning_rate": 2.940065952013208e-07,
+      "loss": 1.3796,
+      "step": 37340
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 60.9375,
+      "learning_rate": 2.936284516705152e-07,
+      "loss": 1.4556,
+      "step": 37360
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 77.6875,
+      "learning_rate": 2.9325030813970963e-07,
+      "loss": 1.334,
+      "step": 37380
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 77.5625,
+      "learning_rate": 2.9287216460890405e-07,
+      "loss": 1.4496,
+      "step": 37400
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 62.6875,
+      "learning_rate": 2.924940210780985e-07,
+      "loss": 1.3992,
+      "step": 37420
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 75.125,
+      "learning_rate": 2.92115877547293e-07,
+      "loss": 1.3938,
+      "step": 37440
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 117.625,
+      "learning_rate": 2.917377340164874e-07,
+      "loss": 1.4084,
+      "step": 37460
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 81.9375,
+      "learning_rate": 2.913595904856819e-07,
+      "loss": 1.4994,
+      "step": 37480
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 87.6875,
+      "learning_rate": 2.909814469548763e-07,
+      "loss": 1.383,
+      "step": 37500
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 59.1875,
+      "learning_rate": 2.9060330342407074e-07,
+      "loss": 1.3683,
+      "step": 37520
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 71.5,
+      "learning_rate": 2.9022515989326516e-07,
+      "loss": 1.4564,
+      "step": 37540
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 59.53125,
+      "learning_rate": 2.8984701636245963e-07,
+      "loss": 1.3854,
+      "step": 37560
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 66.8125,
+      "learning_rate": 2.8946887283165405e-07,
+      "loss": 1.4842,
+      "step": 37580
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 60.84375,
+      "learning_rate": 2.890907293008485e-07,
+      "loss": 1.4141,
+      "step": 37600
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 53.40625,
+      "learning_rate": 2.88712585770043e-07,
+      "loss": 1.4447,
+      "step": 37620
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 64.0625,
+      "learning_rate": 2.883344422392374e-07,
+      "loss": 1.3315,
+      "step": 37640
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 77.8125,
+      "learning_rate": 2.8795629870843184e-07,
+      "loss": 1.4563,
+      "step": 37660
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 74.1875,
+      "learning_rate": 2.8757815517762627e-07,
+      "loss": 1.3781,
+      "step": 37680
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 68.625,
+      "learning_rate": 2.8720001164682074e-07,
+      "loss": 1.3988,
+      "step": 37700
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 76.0625,
+      "learning_rate": 2.8682186811601516e-07,
+      "loss": 1.4011,
+      "step": 37720
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 76.0625,
+      "learning_rate": 2.864437245852096e-07,
+      "loss": 1.371,
+      "step": 37740
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 67.6875,
+      "learning_rate": 2.8606558105440406e-07,
+      "loss": 1.3667,
+      "step": 37760
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 51.25,
+      "learning_rate": 2.856874375235985e-07,
+      "loss": 1.4467,
+      "step": 37780
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 50.96875,
+      "learning_rate": 2.8530929399279295e-07,
+      "loss": 1.3751,
+      "step": 37800
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 61.90625,
+      "learning_rate": 2.8493115046198743e-07,
+      "loss": 1.3476,
+      "step": 37820
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 76.5625,
+      "learning_rate": 2.8455300693118185e-07,
+      "loss": 1.2849,
+      "step": 37840
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 49.28125,
+      "learning_rate": 2.8417486340037627e-07,
+      "loss": 1.4695,
+      "step": 37860
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 66.0625,
+      "learning_rate": 2.837967198695707e-07,
+      "loss": 1.373,
+      "step": 37880
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 61.21875,
+      "learning_rate": 2.8341857633876516e-07,
+      "loss": 1.365,
+      "step": 37900
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 77.9375,
+      "learning_rate": 2.830404328079596e-07,
+      "loss": 1.3943,
+      "step": 37920
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 71.875,
+      "learning_rate": 2.82662289277154e-07,
+      "loss": 1.4403,
+      "step": 37940
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 64.875,
+      "learning_rate": 2.8228414574634853e-07,
+      "loss": 1.4911,
+      "step": 37960
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 62.4375,
+      "learning_rate": 2.8190600221554296e-07,
+      "loss": 1.4561,
+      "step": 37980
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 63.9375,
+      "learning_rate": 2.815278586847374e-07,
+      "loss": 1.2846,
+      "step": 38000
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 61.875,
+      "learning_rate": 2.811497151539318e-07,
+      "loss": 1.4566,
+      "step": 38020
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 67.625,
+      "learning_rate": 2.8077157162312627e-07,
+      "loss": 1.4428,
+      "step": 38040
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 68.6875,
+      "learning_rate": 2.803934280923207e-07,
+      "loss": 1.4209,
+      "step": 38060
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 60.71875,
+      "learning_rate": 2.800152845615151e-07,
+      "loss": 1.2564,
+      "step": 38080
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 64.5,
+      "learning_rate": 2.796371410307096e-07,
+      "loss": 1.5052,
+      "step": 38100
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 65.75,
+      "learning_rate": 2.79258997499904e-07,
+      "loss": 1.3543,
+      "step": 38120
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 64.25,
+      "learning_rate": 2.788808539690985e-07,
+      "loss": 1.3216,
+      "step": 38140
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 64.4375,
+      "learning_rate": 2.785027104382929e-07,
+      "loss": 1.3808,
+      "step": 38160
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 61.875,
+      "learning_rate": 2.781245669074874e-07,
+      "loss": 1.4452,
+      "step": 38180
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 73.0,
+      "learning_rate": 2.777464233766818e-07,
+      "loss": 1.3897,
+      "step": 38200
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 51.90625,
+      "learning_rate": 2.773682798458762e-07,
+      "loss": 1.4239,
+      "step": 38220
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 63.875,
+      "learning_rate": 2.769901363150707e-07,
+      "loss": 1.2993,
+      "step": 38240
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 60.1875,
+      "learning_rate": 2.766119927842651e-07,
+      "loss": 1.4217,
+      "step": 38260
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 77.9375,
+      "learning_rate": 2.7623384925345954e-07,
+      "loss": 1.2893,
+      "step": 38280
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 59.9375,
+      "learning_rate": 2.7585570572265396e-07,
+      "loss": 1.4078,
+      "step": 38300
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 74.75,
+      "learning_rate": 2.754775621918485e-07,
+      "loss": 1.4683,
+      "step": 38320
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 53.9375,
+      "learning_rate": 2.750994186610429e-07,
+      "loss": 1.4265,
+      "step": 38340
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 51.625,
+      "learning_rate": 2.7472127513023733e-07,
+      "loss": 1.3032,
+      "step": 38360
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 60.71875,
+      "learning_rate": 2.743431315994318e-07,
+      "loss": 1.4044,
+      "step": 38380
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 57.46875,
+      "learning_rate": 2.7396498806862623e-07,
+      "loss": 1.2896,
+      "step": 38400
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 60.75,
+      "learning_rate": 2.7358684453782065e-07,
+      "loss": 1.4532,
+      "step": 38420
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 115.375,
+      "learning_rate": 2.732087010070151e-07,
+      "loss": 1.4785,
+      "step": 38440
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 71.5,
+      "learning_rate": 2.7283055747620954e-07,
+      "loss": 1.3397,
+      "step": 38460
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 48.15625,
+      "learning_rate": 2.7245241394540396e-07,
+      "loss": 1.3967,
+      "step": 38480
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 59.96875,
+      "learning_rate": 2.7207427041459844e-07,
+      "loss": 1.4357,
+      "step": 38500
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 118.1875,
+      "learning_rate": 2.716961268837929e-07,
+      "loss": 1.3167,
+      "step": 38520
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 77.375,
+      "learning_rate": 2.7131798335298733e-07,
+      "loss": 1.4363,
+      "step": 38540
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 57.375,
+      "learning_rate": 2.7093983982218176e-07,
+      "loss": 1.3934,
+      "step": 38560
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 66.9375,
+      "learning_rate": 2.7056169629137623e-07,
+      "loss": 1.3952,
+      "step": 38580
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 53.09375,
+      "learning_rate": 2.7018355276057065e-07,
+      "loss": 1.3954,
+      "step": 38600
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 80.3125,
+      "learning_rate": 2.6980540922976507e-07,
+      "loss": 1.4784,
+      "step": 38620
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 65.5,
+      "learning_rate": 2.694272656989595e-07,
+      "loss": 1.3971,
+      "step": 38640
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 53.59375,
+      "learning_rate": 2.69049122168154e-07,
+      "loss": 1.412,
+      "step": 38660
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 81.3125,
+      "learning_rate": 2.6867097863734844e-07,
+      "loss": 1.4045,
+      "step": 38680
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 68.875,
+      "learning_rate": 2.6829283510654286e-07,
+      "loss": 1.3473,
+      "step": 38700
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 70.5625,
+      "learning_rate": 2.6791469157573734e-07,
+      "loss": 1.4147,
+      "step": 38720
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 76.8125,
+      "learning_rate": 2.6753654804493176e-07,
+      "loss": 1.438,
+      "step": 38740
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 65.375,
+      "learning_rate": 2.671584045141262e-07,
+      "loss": 1.4028,
+      "step": 38760
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 51.65625,
+      "learning_rate": 2.667802609833206e-07,
+      "loss": 1.343,
+      "step": 38780
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 68.0625,
+      "learning_rate": 2.664021174525151e-07,
+      "loss": 1.3459,
+      "step": 38800
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 60.15625,
+      "learning_rate": 2.660239739217095e-07,
+      "loss": 1.3992,
+      "step": 38820
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 84.75,
+      "learning_rate": 2.6564583039090397e-07,
+      "loss": 1.3889,
+      "step": 38840
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 80.25,
+      "learning_rate": 2.6526768686009845e-07,
+      "loss": 1.3914,
+      "step": 38860
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 63.0,
+      "learning_rate": 2.6488954332929287e-07,
+      "loss": 1.2944,
+      "step": 38880
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 90.6875,
+      "learning_rate": 2.645113997984873e-07,
+      "loss": 1.4389,
+      "step": 38900
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 82.75,
+      "learning_rate": 2.6413325626768176e-07,
+      "loss": 1.3992,
+      "step": 38920
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 76.0,
+      "learning_rate": 2.637551127368762e-07,
+      "loss": 1.4458,
+      "step": 38940
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 72.875,
+      "learning_rate": 2.633769692060706e-07,
+      "loss": 1.4511,
+      "step": 38960
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 78.4375,
+      "learning_rate": 2.6299882567526503e-07,
+      "loss": 1.3879,
+      "step": 38980
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 65.0625,
+      "learning_rate": 2.626206821444595e-07,
+      "loss": 1.3725,
+      "step": 39000
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 65.5625,
+      "learning_rate": 2.62242538613654e-07,
+      "loss": 1.3031,
+      "step": 39020
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 68.9375,
+      "learning_rate": 2.618643950828484e-07,
+      "loss": 1.4757,
+      "step": 39040
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 64.375,
+      "learning_rate": 2.6148625155204287e-07,
+      "loss": 1.3408,
+      "step": 39060
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 57.4375,
+      "learning_rate": 2.611081080212373e-07,
+      "loss": 1.359,
+      "step": 39080
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 59.78125,
+      "learning_rate": 2.607299644904317e-07,
+      "loss": 1.4038,
+      "step": 39100
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 65.375,
+      "learning_rate": 2.6035182095962613e-07,
+      "loss": 1.4142,
+      "step": 39120
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 72.25,
+      "learning_rate": 2.599736774288206e-07,
+      "loss": 1.4517,
+      "step": 39140
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 57.125,
+      "learning_rate": 2.5959553389801503e-07,
+      "loss": 1.4371,
+      "step": 39160
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 82.1875,
+      "learning_rate": 2.592173903672095e-07,
+      "loss": 1.3103,
+      "step": 39180
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 73.125,
+      "learning_rate": 2.58839246836404e-07,
+      "loss": 1.3995,
+      "step": 39200
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 72.375,
+      "learning_rate": 2.584611033055984e-07,
+      "loss": 1.4843,
+      "step": 39220
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 56.1875,
+      "learning_rate": 2.580829597747928e-07,
+      "loss": 1.4293,
+      "step": 39240
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 73.4375,
+      "learning_rate": 2.5770481624398724e-07,
+      "loss": 1.4464,
+      "step": 39260
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 64.875,
+      "learning_rate": 2.573266727131817e-07,
+      "loss": 1.3824,
+      "step": 39280
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 47.34375,
+      "learning_rate": 2.5694852918237614e-07,
+      "loss": 1.4179,
+      "step": 39300
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 53.34375,
+      "learning_rate": 2.5657038565157056e-07,
+      "loss": 1.4673,
+      "step": 39320
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 57.21875,
+      "learning_rate": 2.5619224212076503e-07,
+      "loss": 1.3636,
+      "step": 39340
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 75.0,
+      "learning_rate": 2.558140985899595e-07,
+      "loss": 1.4275,
+      "step": 39360
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 62.1875,
+      "learning_rate": 2.5543595505915393e-07,
+      "loss": 1.4423,
+      "step": 39380
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 57.0625,
+      "learning_rate": 2.5505781152834835e-07,
+      "loss": 1.3202,
+      "step": 39400
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 74.9375,
+      "learning_rate": 2.546796679975428e-07,
+      "loss": 1.4716,
+      "step": 39420
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 62.03125,
+      "learning_rate": 2.5430152446673725e-07,
+      "loss": 1.3571,
+      "step": 39440
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 65.25,
+      "learning_rate": 2.5392338093593167e-07,
+      "loss": 1.44,
+      "step": 39460
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 48.59375,
+      "learning_rate": 2.5354523740512614e-07,
+      "loss": 1.4017,
+      "step": 39480
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 58.78125,
+      "learning_rate": 2.5316709387432056e-07,
+      "loss": 1.4145,
+      "step": 39500
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 108.5,
+      "learning_rate": 2.52788950343515e-07,
+      "loss": 1.3431,
+      "step": 39520
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 66.0,
+      "learning_rate": 2.524108068127095e-07,
+      "loss": 1.4277,
+      "step": 39540
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 72.625,
+      "learning_rate": 2.5203266328190393e-07,
+      "loss": 1.4133,
+      "step": 39560
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 68.75,
+      "learning_rate": 2.5165451975109835e-07,
+      "loss": 1.3857,
+      "step": 39580
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 57.0,
+      "learning_rate": 2.512763762202928e-07,
+      "loss": 1.4572,
+      "step": 39600
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 71.5625,
+      "learning_rate": 2.5089823268948725e-07,
+      "loss": 1.3965,
+      "step": 39620
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 66.75,
+      "learning_rate": 2.5052008915868167e-07,
+      "loss": 1.3597,
+      "step": 39640
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 69.0625,
+      "learning_rate": 2.501419456278761e-07,
+      "loss": 1.3411,
+      "step": 39660
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 54.0625,
+      "learning_rate": 2.4976380209707057e-07,
+      "loss": 1.4024,
+      "step": 39680
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 75.4375,
+      "learning_rate": 2.49385658566265e-07,
+      "loss": 1.4341,
+      "step": 39700
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 60.875,
+      "learning_rate": 2.4900751503545946e-07,
+      "loss": 1.4068,
+      "step": 39720
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 61.0,
+      "learning_rate": 2.486293715046539e-07,
+      "loss": 1.3496,
+      "step": 39740
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 59.875,
+      "learning_rate": 2.4825122797384836e-07,
+      "loss": 1.39,
+      "step": 39760
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 71.4375,
+      "learning_rate": 2.478730844430428e-07,
+      "loss": 1.3572,
+      "step": 39780
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 77.4375,
+      "learning_rate": 2.474949409122372e-07,
+      "loss": 1.3969,
+      "step": 39800
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 52.34375,
+      "learning_rate": 2.471167973814317e-07,
+      "loss": 1.4315,
+      "step": 39820
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 102.9375,
+      "learning_rate": 2.467386538506261e-07,
+      "loss": 1.3345,
+      "step": 39840
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 68.6875,
+      "learning_rate": 2.4636051031982057e-07,
+      "loss": 1.3399,
+      "step": 39860
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 54.78125,
+      "learning_rate": 2.45982366789015e-07,
+      "loss": 1.4267,
+      "step": 39880
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 70.875,
+      "learning_rate": 2.456042232582094e-07,
+      "loss": 1.4507,
+      "step": 39900
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 56.40625,
+      "learning_rate": 2.452260797274039e-07,
+      "loss": 1.4029,
+      "step": 39920
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 71.25,
+      "learning_rate": 2.448479361965983e-07,
+      "loss": 1.4685,
+      "step": 39940
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 62.96875,
+      "learning_rate": 2.444697926657928e-07,
+      "loss": 1.4023,
+      "step": 39960
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 69.875,
+      "learning_rate": 2.440916491349872e-07,
+      "loss": 1.4639,
+      "step": 39980
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 65.6875,
+      "learning_rate": 2.437135056041816e-07,
+      "loss": 1.3749,
+      "step": 40000
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 69.6875,
+      "learning_rate": 2.433353620733761e-07,
+      "loss": 1.41,
+      "step": 40020
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 67.1875,
+      "learning_rate": 2.429572185425705e-07,
+      "loss": 1.4743,
+      "step": 40040
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 71.375,
+      "learning_rate": 2.42579075011765e-07,
+      "loss": 1.3925,
+      "step": 40060
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 63.84375,
+      "learning_rate": 2.422009314809594e-07,
+      "loss": 1.4514,
+      "step": 40080
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 101.9375,
+      "learning_rate": 2.418227879501539e-07,
+      "loss": 1.4082,
+      "step": 40100
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 72.375,
+      "learning_rate": 2.414446444193483e-07,
+      "loss": 1.3906,
+      "step": 40120
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 54.625,
+      "learning_rate": 2.4106650088854273e-07,
+      "loss": 1.3851,
+      "step": 40140
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 75.1875,
+      "learning_rate": 2.406883573577372e-07,
+      "loss": 1.3869,
+      "step": 40160
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 65.5625,
+      "learning_rate": 2.4031021382693163e-07,
+      "loss": 1.3966,
+      "step": 40180
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 61.78125,
+      "learning_rate": 2.399320702961261e-07,
+      "loss": 1.3518,
+      "step": 40200
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 64.125,
+      "learning_rate": 2.395539267653205e-07,
+      "loss": 1.4587,
+      "step": 40220
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 64.25,
+      "learning_rate": 2.3917578323451495e-07,
+      "loss": 1.4854,
+      "step": 40240
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 68.75,
+      "learning_rate": 2.3879763970370937e-07,
+      "loss": 1.3629,
+      "step": 40260
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 93.6875,
+      "learning_rate": 2.3841949617290384e-07,
+      "loss": 1.4779,
+      "step": 40280
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 52.40625,
+      "learning_rate": 2.380413526420983e-07,
+      "loss": 1.3754,
+      "step": 40300
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 61.46875,
+      "learning_rate": 2.3766320911129274e-07,
+      "loss": 1.395,
+      "step": 40320
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 72.3125,
+      "learning_rate": 2.3728506558048716e-07,
+      "loss": 1.3755,
+      "step": 40340
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 57.53125,
+      "learning_rate": 2.3690692204968163e-07,
+      "loss": 1.3909,
+      "step": 40360
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 66.6875,
+      "learning_rate": 2.3652877851887605e-07,
+      "loss": 1.385,
+      "step": 40380
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 59.40625,
+      "learning_rate": 2.361506349880705e-07,
+      "loss": 1.4461,
+      "step": 40400
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 70.75,
+      "learning_rate": 2.3577249145726495e-07,
+      "loss": 1.379,
+      "step": 40420
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 73.0,
+      "learning_rate": 2.3539434792645937e-07,
+      "loss": 1.3794,
+      "step": 40440
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 58.6875,
+      "learning_rate": 2.3501620439565384e-07,
+      "loss": 1.4292,
+      "step": 40460
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 67.25,
+      "learning_rate": 2.3463806086484827e-07,
+      "loss": 1.4327,
+      "step": 40480
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 88.4375,
+      "learning_rate": 2.3425991733404271e-07,
+      "loss": 1.5276,
+      "step": 40500
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 65.9375,
+      "learning_rate": 2.3388177380323714e-07,
+      "loss": 1.3107,
+      "step": 40520
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 65.25,
+      "learning_rate": 2.335036302724316e-07,
+      "loss": 1.3638,
+      "step": 40540
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 83.625,
+      "learning_rate": 2.3312548674162606e-07,
+      "loss": 1.3733,
+      "step": 40560
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 66.6875,
+      "learning_rate": 2.3274734321082048e-07,
+      "loss": 1.4898,
+      "step": 40580
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 64.5,
+      "learning_rate": 2.3236919968001493e-07,
+      "loss": 1.4436,
+      "step": 40600
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 67.5625,
+      "learning_rate": 2.3199105614920935e-07,
+      "loss": 1.4491,
+      "step": 40620
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 64.1875,
+      "learning_rate": 2.3161291261840382e-07,
+      "loss": 1.3816,
+      "step": 40640
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 59.03125,
+      "learning_rate": 2.3123476908759827e-07,
+      "loss": 1.4964,
+      "step": 40660
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 59.28125,
+      "learning_rate": 2.308566255567927e-07,
+      "loss": 1.364,
+      "step": 40680
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 62.15625,
+      "learning_rate": 2.3047848202598714e-07,
+      "loss": 1.3505,
+      "step": 40700
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 64.9375,
+      "learning_rate": 2.3010033849518159e-07,
+      "loss": 1.3727,
+      "step": 40720
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 74.375,
+      "learning_rate": 2.2972219496437603e-07,
+      "loss": 1.4421,
+      "step": 40740
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 88.5625,
+      "learning_rate": 2.2934405143357046e-07,
+      "loss": 1.4352,
+      "step": 40760
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 59.125,
+      "learning_rate": 2.289659079027649e-07,
+      "loss": 1.3484,
+      "step": 40780
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 61.25,
+      "learning_rate": 2.2858776437195938e-07,
+      "loss": 1.404,
+      "step": 40800
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 79.5,
+      "learning_rate": 2.282096208411538e-07,
+      "loss": 1.4222,
+      "step": 40820
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 55.6875,
+      "learning_rate": 2.2783147731034825e-07,
+      "loss": 1.3627,
+      "step": 40840
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 71.25,
+      "learning_rate": 2.2745333377954267e-07,
+      "loss": 1.4019,
+      "step": 40860
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 75.0,
+      "learning_rate": 2.2707519024873712e-07,
+      "loss": 1.3863,
+      "step": 40880
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 58.78125,
+      "learning_rate": 2.266970467179316e-07,
+      "loss": 1.3336,
+      "step": 40900
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 72.5625,
+      "learning_rate": 2.26318903187126e-07,
+      "loss": 1.437,
+      "step": 40920
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 53.90625,
+      "learning_rate": 2.2594075965632046e-07,
+      "loss": 1.3679,
+      "step": 40940
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 62.96875,
+      "learning_rate": 2.2556261612551488e-07,
+      "loss": 1.4215,
+      "step": 40960
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 62.03125,
+      "learning_rate": 2.2518447259470935e-07,
+      "loss": 1.4289,
+      "step": 40980
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 60.78125,
+      "learning_rate": 2.2480632906390378e-07,
+      "loss": 1.52,
+      "step": 41000
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 54.03125,
+      "learning_rate": 2.2442818553309822e-07,
+      "loss": 1.3679,
+      "step": 41020
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 61.6875,
+      "learning_rate": 2.2405004200229267e-07,
+      "loss": 1.5041,
+      "step": 41040
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 58.40625,
+      "learning_rate": 2.2367189847148712e-07,
+      "loss": 1.4264,
+      "step": 41060
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 58.03125,
+      "learning_rate": 2.2329375494068157e-07,
+      "loss": 1.3305,
+      "step": 41080
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 62.5625,
+      "learning_rate": 2.22915611409876e-07,
+      "loss": 1.3939,
+      "step": 41100
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 62.5,
+      "learning_rate": 2.2253746787907044e-07,
+      "loss": 1.3357,
+      "step": 41120
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 61.03125,
+      "learning_rate": 2.2215932434826486e-07,
+      "loss": 1.3863,
+      "step": 41140
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 71.5625,
+      "learning_rate": 2.2178118081745933e-07,
+      "loss": 1.5696,
+      "step": 41160
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 56.25,
+      "learning_rate": 2.2140303728665378e-07,
+      "loss": 1.3852,
+      "step": 41180
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 56.0625,
+      "learning_rate": 2.210248937558482e-07,
+      "loss": 1.3793,
+      "step": 41200
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 68.8125,
+      "learning_rate": 2.2064675022504265e-07,
+      "loss": 1.3969,
+      "step": 41220
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 86.3125,
+      "learning_rate": 2.202686066942371e-07,
+      "loss": 1.4984,
+      "step": 41240
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 73.5,
+      "learning_rate": 2.1989046316343154e-07,
+      "loss": 1.4178,
+      "step": 41260
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 82.875,
+      "learning_rate": 2.19512319632626e-07,
+      "loss": 1.4235,
+      "step": 41280
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 57.90625,
+      "learning_rate": 2.191341761018204e-07,
+      "loss": 1.3364,
+      "step": 41300
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 60.28125,
+      "learning_rate": 2.187560325710149e-07,
+      "loss": 1.3978,
+      "step": 41320
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 147.0,
+      "learning_rate": 2.183778890402093e-07,
+      "loss": 1.3867,
+      "step": 41340
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 106.375,
+      "learning_rate": 2.1799974550940376e-07,
+      "loss": 1.4255,
+      "step": 41360
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 75.1875,
+      "learning_rate": 2.1762160197859818e-07,
+      "loss": 1.4822,
+      "step": 41380
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 60.625,
+      "learning_rate": 2.1724345844779263e-07,
+      "loss": 1.4114,
+      "step": 41400
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 66.0,
+      "learning_rate": 2.168653149169871e-07,
+      "loss": 1.4062,
+      "step": 41420
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 61.8125,
+      "learning_rate": 2.1648717138618152e-07,
+      "loss": 1.4739,
+      "step": 41440
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 79.8125,
+      "learning_rate": 2.1610902785537597e-07,
+      "loss": 1.4277,
+      "step": 41460
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 55.84375,
+      "learning_rate": 2.157308843245704e-07,
+      "loss": 1.4047,
+      "step": 41480
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 84.0,
+      "learning_rate": 2.1535274079376486e-07,
+      "loss": 1.3995,
+      "step": 41500
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 66.125,
+      "learning_rate": 2.149745972629593e-07,
+      "loss": 1.497,
+      "step": 41520
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 75.0625,
+      "learning_rate": 2.1459645373215373e-07,
+      "loss": 1.4269,
+      "step": 41540
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 78.0625,
+      "learning_rate": 2.1421831020134818e-07,
+      "loss": 1.4038,
+      "step": 41560
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 64.3125,
+      "learning_rate": 2.138401666705426e-07,
+      "loss": 1.3915,
+      "step": 41580
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 53.21875,
+      "learning_rate": 2.1346202313973708e-07,
+      "loss": 1.385,
+      "step": 41600
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 64.75,
+      "learning_rate": 2.130838796089315e-07,
+      "loss": 1.4048,
+      "step": 41620
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 69.9375,
+      "learning_rate": 2.1270573607812595e-07,
+      "loss": 1.383,
+      "step": 41640
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 55.25,
+      "learning_rate": 2.123275925473204e-07,
+      "loss": 1.4297,
+      "step": 41660
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 62.84375,
+      "learning_rate": 2.1194944901651484e-07,
+      "loss": 1.4653,
+      "step": 41680
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 63.0625,
+      "learning_rate": 2.115713054857093e-07,
+      "loss": 1.3194,
+      "step": 41700
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 93.0,
+      "learning_rate": 2.111931619549037e-07,
+      "loss": 1.3567,
+      "step": 41720
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 73.375,
+      "learning_rate": 2.1081501842409816e-07,
+      "loss": 1.3192,
+      "step": 41740
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 60.53125,
+      "learning_rate": 2.1043687489329263e-07,
+      "loss": 1.3712,
+      "step": 41760
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 55.0,
+      "learning_rate": 2.1005873136248705e-07,
+      "loss": 1.4137,
+      "step": 41780
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 62.53125,
+      "learning_rate": 2.096805878316815e-07,
+      "loss": 1.4085,
+      "step": 41800
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 61.84375,
+      "learning_rate": 2.0930244430087592e-07,
+      "loss": 1.3769,
+      "step": 41820
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 71.75,
+      "learning_rate": 2.0892430077007037e-07,
+      "loss": 1.3457,
+      "step": 41840
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 70.9375,
+      "learning_rate": 2.0854615723926482e-07,
+      "loss": 1.3542,
+      "step": 41860
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 78.5,
+      "learning_rate": 2.0816801370845927e-07,
+      "loss": 1.438,
+      "step": 41880
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 78.1875,
+      "learning_rate": 2.0778987017765371e-07,
+      "loss": 1.4064,
+      "step": 41900
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 73.8125,
+      "learning_rate": 2.0741172664684814e-07,
+      "loss": 1.4502,
+      "step": 41920
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 48.5,
+      "learning_rate": 2.070335831160426e-07,
+      "loss": 1.3534,
+      "step": 41940
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 99.125,
+      "learning_rate": 2.0665543958523703e-07,
+      "loss": 1.3896,
+      "step": 41960
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 66.1875,
+      "learning_rate": 2.0627729605443148e-07,
+      "loss": 1.3242,
+      "step": 41980
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 83.0,
+      "learning_rate": 2.058991525236259e-07,
+      "loss": 1.49,
+      "step": 42000
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 74.3125,
+      "learning_rate": 2.0552100899282037e-07,
+      "loss": 1.451,
+      "step": 42020
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 65.125,
+      "learning_rate": 2.0514286546201482e-07,
+      "loss": 1.307,
+      "step": 42040
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 67.75,
+      "learning_rate": 2.0476472193120924e-07,
+      "loss": 1.5365,
+      "step": 42060
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 56.3125,
+      "learning_rate": 2.043865784004037e-07,
+      "loss": 1.3086,
+      "step": 42080
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 72.8125,
+      "learning_rate": 2.040084348695981e-07,
+      "loss": 1.3316,
+      "step": 42100
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 68.75,
+      "learning_rate": 2.0363029133879259e-07,
+      "loss": 1.4897,
+      "step": 42120
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 63.0625,
+      "learning_rate": 2.0325214780798703e-07,
+      "loss": 1.3649,
+      "step": 42140
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 63.75,
+      "learning_rate": 2.0287400427718146e-07,
+      "loss": 1.4543,
+      "step": 42160
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 59.21875,
+      "learning_rate": 2.024958607463759e-07,
+      "loss": 1.4654,
+      "step": 42180
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 63.5,
+      "learning_rate": 2.0211771721557035e-07,
+      "loss": 1.4143,
+      "step": 42200
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 74.875,
+      "learning_rate": 2.017395736847648e-07,
+      "loss": 1.3129,
+      "step": 42220
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 74.375,
+      "learning_rate": 2.0136143015395922e-07,
+      "loss": 1.4688,
+      "step": 42240
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 78.1875,
+      "learning_rate": 2.0098328662315367e-07,
+      "loss": 1.4214,
+      "step": 42260
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 76.6875,
+      "learning_rate": 2.0060514309234814e-07,
+      "loss": 1.3917,
+      "step": 42280
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 59.3125,
+      "learning_rate": 2.0022699956154256e-07,
+      "loss": 1.3565,
+      "step": 42300
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 71.4375,
+      "learning_rate": 1.99848856030737e-07,
+      "loss": 1.3224,
+      "step": 42320
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 79.0,
+      "learning_rate": 1.9947071249993143e-07,
+      "loss": 1.3764,
+      "step": 42340
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 67.6875,
+      "learning_rate": 1.9909256896912588e-07,
+      "loss": 1.2843,
+      "step": 42360
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 63.21875,
+      "learning_rate": 1.9871442543832035e-07,
+      "loss": 1.3439,
+      "step": 42380
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 52.71875,
+      "learning_rate": 1.9833628190751478e-07,
+      "loss": 1.2762,
+      "step": 42400
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 50.59375,
+      "learning_rate": 1.9795813837670922e-07,
+      "loss": 1.3074,
+      "step": 42420
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 65.4375,
+      "learning_rate": 1.9757999484590365e-07,
+      "loss": 1.451,
+      "step": 42440
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 79.9375,
+      "learning_rate": 1.9720185131509812e-07,
+      "loss": 1.337,
+      "step": 42460
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 80.875,
+      "learning_rate": 1.9682370778429254e-07,
+      "loss": 1.316,
+      "step": 42480
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 73.875,
+      "learning_rate": 1.96445564253487e-07,
+      "loss": 1.252,
+      "step": 42500
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 59.90625,
+      "learning_rate": 1.9606742072268144e-07,
+      "loss": 1.3394,
+      "step": 42520
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 61.0625,
+      "learning_rate": 1.9568927719187586e-07,
+      "loss": 1.4067,
+      "step": 42540
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 73.3125,
+      "learning_rate": 1.9531113366107033e-07,
+      "loss": 1.3229,
+      "step": 42560
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 99.625,
+      "learning_rate": 1.9493299013026475e-07,
+      "loss": 1.3768,
+      "step": 42580
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 57.78125,
+      "learning_rate": 1.945548465994592e-07,
+      "loss": 1.3839,
+      "step": 42600
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 61.0,
+      "learning_rate": 1.9417670306865365e-07,
+      "loss": 1.3187,
+      "step": 42620
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 56.5625,
+      "learning_rate": 1.937985595378481e-07,
+      "loss": 1.3785,
+      "step": 42640
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 52.09375,
+      "learning_rate": 1.9342041600704254e-07,
+      "loss": 1.2809,
+      "step": 42660
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 60.65625,
+      "learning_rate": 1.9304227247623697e-07,
+      "loss": 1.3787,
+      "step": 42680
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 58.15625,
+      "learning_rate": 1.926641289454314e-07,
+      "loss": 1.2968,
+      "step": 42700
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 76.875,
+      "learning_rate": 1.9228598541462586e-07,
+      "loss": 1.3426,
+      "step": 42720
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 64.125,
+      "learning_rate": 1.919078418838203e-07,
+      "loss": 1.389,
+      "step": 42740
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 74.875,
+      "learning_rate": 1.9152969835301476e-07,
+      "loss": 1.3012,
+      "step": 42760
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 68.3125,
+      "learning_rate": 1.9115155482220918e-07,
+      "loss": 1.3064,
+      "step": 42780
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 67.375,
+      "learning_rate": 1.9077341129140363e-07,
+      "loss": 1.2949,
+      "step": 42800
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 63.84375,
+      "learning_rate": 1.9039526776059807e-07,
+      "loss": 1.3852,
+      "step": 42820
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 60.4375,
+      "learning_rate": 1.9001712422979252e-07,
+      "loss": 1.3849,
+      "step": 42840
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 56.71875,
+      "learning_rate": 1.8963898069898694e-07,
+      "loss": 1.2859,
+      "step": 42860
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 60.90625,
+      "learning_rate": 1.892608371681814e-07,
+      "loss": 1.3057,
+      "step": 42880
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 78.1875,
+      "learning_rate": 1.8888269363737586e-07,
+      "loss": 1.3186,
+      "step": 42900
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 56.84375,
+      "learning_rate": 1.8850455010657029e-07,
+      "loss": 1.36,
+      "step": 42920
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 63.03125,
+      "learning_rate": 1.8812640657576473e-07,
+      "loss": 1.2948,
+      "step": 42940
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 57.90625,
+      "learning_rate": 1.8774826304495915e-07,
+      "loss": 1.434,
+      "step": 42960
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 58.1875,
+      "learning_rate": 1.8737011951415363e-07,
+      "loss": 1.3389,
+      "step": 42980
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 82.6875,
+      "learning_rate": 1.8699197598334808e-07,
+      "loss": 1.3851,
+      "step": 43000
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 59.65625,
+      "learning_rate": 1.866138324525425e-07,
+      "loss": 1.3092,
+      "step": 43020
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 83.875,
+      "learning_rate": 1.8623568892173695e-07,
+      "loss": 1.2534,
+      "step": 43040
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 56.9375,
+      "learning_rate": 1.8585754539093137e-07,
+      "loss": 1.3689,
+      "step": 43060
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 65.9375,
+      "learning_rate": 1.8547940186012584e-07,
+      "loss": 1.3127,
+      "step": 43080
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 72.9375,
+      "learning_rate": 1.8510125832932026e-07,
+      "loss": 1.4009,
+      "step": 43100
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 57.40625,
+      "learning_rate": 1.847231147985147e-07,
+      "loss": 1.3522,
+      "step": 43120
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 60.40625,
+      "learning_rate": 1.8434497126770916e-07,
+      "loss": 1.3938,
+      "step": 43140
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 62.71875,
+      "learning_rate": 1.839668277369036e-07,
+      "loss": 1.3835,
+      "step": 43160
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 53.5625,
+      "learning_rate": 1.8358868420609805e-07,
+      "loss": 1.3114,
+      "step": 43180
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 71.3125,
+      "learning_rate": 1.8321054067529248e-07,
+      "loss": 1.3608,
+      "step": 43200
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 109.1875,
+      "learning_rate": 1.8283239714448692e-07,
+      "loss": 1.3388,
+      "step": 43220
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 64.8125,
+      "learning_rate": 1.8245425361368137e-07,
+      "loss": 1.2882,
+      "step": 43240
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 61.8125,
+      "learning_rate": 1.8207611008287582e-07,
+      "loss": 1.3305,
+      "step": 43260
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 63.46875,
+      "learning_rate": 1.8169796655207027e-07,
+      "loss": 1.3216,
+      "step": 43280
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 54.28125,
+      "learning_rate": 1.813198230212647e-07,
+      "loss": 1.3642,
+      "step": 43300
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 56.78125,
+      "learning_rate": 1.8094167949045914e-07,
+      "loss": 1.2574,
+      "step": 43320
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 56.28125,
+      "learning_rate": 1.8056353595965358e-07,
+      "loss": 1.2901,
+      "step": 43340
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 59.375,
+      "learning_rate": 1.8018539242884803e-07,
+      "loss": 1.3838,
+      "step": 43360
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 57.25,
+      "learning_rate": 1.7980724889804248e-07,
+      "loss": 1.307,
+      "step": 43380
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 67.4375,
+      "learning_rate": 1.794291053672369e-07,
+      "loss": 1.2821,
+      "step": 43400
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 89.375,
+      "learning_rate": 1.7905096183643137e-07,
+      "loss": 1.337,
+      "step": 43420
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 104.4375,
+      "learning_rate": 1.786728183056258e-07,
+      "loss": 1.3838,
+      "step": 43440
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 63.78125,
+      "learning_rate": 1.7829467477482024e-07,
+      "loss": 1.3542,
+      "step": 43460
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 64.1875,
+      "learning_rate": 1.779165312440147e-07,
+      "loss": 1.3268,
+      "step": 43480
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 60.1875,
+      "learning_rate": 1.775383877132091e-07,
+      "loss": 1.2482,
+      "step": 43500
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 72.0,
+      "learning_rate": 1.7716024418240359e-07,
+      "loss": 1.3675,
+      "step": 43520
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 51.375,
+      "learning_rate": 1.76782100651598e-07,
+      "loss": 1.3264,
+      "step": 43540
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 79.0625,
+      "learning_rate": 1.7640395712079246e-07,
+      "loss": 1.2761,
+      "step": 43560
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 59.5625,
+      "learning_rate": 1.7602581358998688e-07,
+      "loss": 1.3908,
+      "step": 43580
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 67.3125,
+      "learning_rate": 1.7564767005918135e-07,
+      "loss": 1.3082,
+      "step": 43600
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 79.625,
+      "learning_rate": 1.752695265283758e-07,
+      "loss": 1.3252,
+      "step": 43620
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 82.0,
+      "learning_rate": 1.7489138299757022e-07,
+      "loss": 1.3506,
+      "step": 43640
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 75.6875,
+      "learning_rate": 1.7451323946676467e-07,
+      "loss": 1.2852,
+      "step": 43660
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 62.15625,
+      "learning_rate": 1.7413509593595912e-07,
+      "loss": 1.2953,
+      "step": 43680
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 70.0625,
+      "learning_rate": 1.7375695240515356e-07,
+      "loss": 1.3503,
+      "step": 43700
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 69.625,
+      "learning_rate": 1.73378808874348e-07,
+      "loss": 1.3355,
+      "step": 43720
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 82.3125,
+      "learning_rate": 1.7300066534354243e-07,
+      "loss": 1.3057,
+      "step": 43740
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 71.3125,
+      "learning_rate": 1.7262252181273688e-07,
+      "loss": 1.2385,
+      "step": 43760
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 64.8125,
+      "learning_rate": 1.7224437828193133e-07,
+      "loss": 1.3327,
+      "step": 43780
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 64.9375,
+      "learning_rate": 1.7186623475112578e-07,
+      "loss": 1.2779,
+      "step": 43800
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 75.875,
+      "learning_rate": 1.714880912203202e-07,
+      "loss": 1.3603,
+      "step": 43820
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 47.15625,
+      "learning_rate": 1.7110994768951465e-07,
+      "loss": 1.2305,
+      "step": 43840
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 69.1875,
+      "learning_rate": 1.7073180415870912e-07,
+      "loss": 1.2893,
+      "step": 43860
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 61.5,
+      "learning_rate": 1.7035366062790354e-07,
+      "loss": 1.3423,
+      "step": 43880
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 58.71875,
+      "learning_rate": 1.69975517097098e-07,
+      "loss": 1.2352,
+      "step": 43900
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 51.1875,
+      "learning_rate": 1.695973735662924e-07,
+      "loss": 1.314,
+      "step": 43920
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 60.59375,
+      "learning_rate": 1.6921923003548688e-07,
+      "loss": 1.2465,
+      "step": 43940
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 87.125,
+      "learning_rate": 1.688410865046813e-07,
+      "loss": 1.3573,
+      "step": 43960
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 78.6875,
+      "learning_rate": 1.6846294297387575e-07,
+      "loss": 1.3748,
+      "step": 43980
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 64.75,
+      "learning_rate": 1.680847994430702e-07,
+      "loss": 1.4336,
+      "step": 44000
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 69.0625,
+      "learning_rate": 1.6770665591226462e-07,
+      "loss": 1.3579,
+      "step": 44020
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 65.125,
+      "learning_rate": 1.673285123814591e-07,
+      "loss": 1.3783,
+      "step": 44040
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 57.1875,
+      "learning_rate": 1.6695036885065352e-07,
+      "loss": 1.277,
+      "step": 44060
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 63.78125,
+      "learning_rate": 1.6657222531984797e-07,
+      "loss": 1.3084,
+      "step": 44080
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 62.75,
+      "learning_rate": 1.6619408178904241e-07,
+      "loss": 1.4371,
+      "step": 44100
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 60.6875,
+      "learning_rate": 1.6581593825823686e-07,
+      "loss": 1.3106,
+      "step": 44120
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 55.90625,
+      "learning_rate": 1.654377947274313e-07,
+      "loss": 1.3069,
+      "step": 44140
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 73.625,
+      "learning_rate": 1.6505965119662573e-07,
+      "loss": 1.3339,
+      "step": 44160
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 56.03125,
+      "learning_rate": 1.6468150766582018e-07,
+      "loss": 1.1934,
+      "step": 44180
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 59.59375,
+      "learning_rate": 1.643033641350146e-07,
+      "loss": 1.2749,
+      "step": 44200
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 52.15625,
+      "learning_rate": 1.6392522060420907e-07,
+      "loss": 1.2481,
+      "step": 44220
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 74.25,
+      "learning_rate": 1.6354707707340352e-07,
+      "loss": 1.3556,
+      "step": 44240
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 54.09375,
+      "learning_rate": 1.6316893354259794e-07,
+      "loss": 1.3386,
+      "step": 44260
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 60.78125,
+      "learning_rate": 1.627907900117924e-07,
+      "loss": 1.4361,
+      "step": 44280
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 67.5,
+      "learning_rate": 1.6241264648098684e-07,
+      "loss": 1.3713,
+      "step": 44300
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 65.875,
+      "learning_rate": 1.6203450295018129e-07,
+      "loss": 1.3499,
+      "step": 44320
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 54.84375,
+      "learning_rate": 1.6165635941937573e-07,
+      "loss": 1.3358,
+      "step": 44340
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 54.78125,
+      "learning_rate": 1.6127821588857015e-07,
+      "loss": 1.32,
+      "step": 44360
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 72.0,
+      "learning_rate": 1.6090007235776463e-07,
+      "loss": 1.3005,
+      "step": 44380
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 58.6875,
+      "learning_rate": 1.6052192882695905e-07,
+      "loss": 1.2939,
+      "step": 44400
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 64.875,
+      "learning_rate": 1.601437852961535e-07,
+      "loss": 1.4002,
+      "step": 44420
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 61.25,
+      "learning_rate": 1.5976564176534792e-07,
+      "loss": 1.2259,
+      "step": 44440
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 71.1875,
+      "learning_rate": 1.5938749823454237e-07,
+      "loss": 1.432,
+      "step": 44460
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 73.25,
+      "learning_rate": 1.5900935470373684e-07,
+      "loss": 1.316,
+      "step": 44480
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 93.3125,
+      "learning_rate": 1.5863121117293126e-07,
+      "loss": 1.3683,
+      "step": 44500
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 65.9375,
+      "learning_rate": 1.582530676421257e-07,
+      "loss": 1.2994,
+      "step": 44520
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 76.5,
+      "learning_rate": 1.5787492411132013e-07,
+      "loss": 1.3194,
+      "step": 44540
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 63.0625,
+      "learning_rate": 1.574967805805146e-07,
+      "loss": 1.2961,
+      "step": 44560
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 60.8125,
+      "learning_rate": 1.5711863704970905e-07,
+      "loss": 1.4499,
+      "step": 44580
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 76.3125,
+      "learning_rate": 1.5674049351890348e-07,
+      "loss": 1.3242,
+      "step": 44600
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 70.25,
+      "learning_rate": 1.5636234998809792e-07,
+      "loss": 1.3083,
+      "step": 44620
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 68.1875,
+      "learning_rate": 1.5598420645729237e-07,
+      "loss": 1.3901,
+      "step": 44640
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 53.3125,
+      "learning_rate": 1.5560606292648682e-07,
+      "loss": 1.3179,
+      "step": 44660
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 92.125,
+      "learning_rate": 1.5522791939568124e-07,
+      "loss": 1.2245,
+      "step": 44680
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 77.0625,
+      "learning_rate": 1.548497758648757e-07,
+      "loss": 1.3041,
+      "step": 44700
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 56.09375,
+      "learning_rate": 1.5447163233407014e-07,
+      "loss": 1.3729,
+      "step": 44720
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 68.625,
+      "learning_rate": 1.5409348880326458e-07,
+      "loss": 1.314,
+      "step": 44740
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 65.625,
+      "learning_rate": 1.5371534527245903e-07,
+      "loss": 1.3265,
+      "step": 44760
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 60.09375,
+      "learning_rate": 1.5333720174165345e-07,
+      "loss": 1.257,
+      "step": 44780
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 57.375,
+      "learning_rate": 1.529590582108479e-07,
+      "loss": 1.3232,
+      "step": 44800
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 79.1875,
+      "learning_rate": 1.5258091468004235e-07,
+      "loss": 1.3386,
+      "step": 44820
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 80.5,
+      "learning_rate": 1.522027711492368e-07,
+      "loss": 1.3517,
+      "step": 44840
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 59.4375,
+      "learning_rate": 1.5182462761843124e-07,
+      "loss": 1.2794,
+      "step": 44860
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 69.375,
+      "learning_rate": 1.5144648408762566e-07,
+      "loss": 1.2523,
+      "step": 44880
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 98.875,
+      "learning_rate": 1.5106834055682014e-07,
+      "loss": 1.3374,
+      "step": 44900
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 77.8125,
+      "learning_rate": 1.5069019702601456e-07,
+      "loss": 1.3555,
+      "step": 44920
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 59.78125,
+      "learning_rate": 1.50312053495209e-07,
+      "loss": 1.3583,
+      "step": 44940
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 56.75,
+      "learning_rate": 1.4993390996440346e-07,
+      "loss": 1.3552,
+      "step": 44960
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 51.96875,
+      "learning_rate": 1.4955576643359788e-07,
+      "loss": 1.3089,
+      "step": 44980
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 56.9375,
+      "learning_rate": 1.4917762290279235e-07,
+      "loss": 1.3219,
+      "step": 45000
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 71.5,
+      "learning_rate": 1.4879947937198677e-07,
+      "loss": 1.3258,
+      "step": 45020
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 84.1875,
+      "learning_rate": 1.4842133584118122e-07,
+      "loss": 1.3392,
+      "step": 45040
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 66.625,
+      "learning_rate": 1.4804319231037564e-07,
+      "loss": 1.2842,
+      "step": 45060
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 72.8125,
+      "learning_rate": 1.4766504877957012e-07,
+      "loss": 1.32,
+      "step": 45080
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 59.25,
+      "learning_rate": 1.4728690524876456e-07,
+      "loss": 1.334,
+      "step": 45100
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 32.84375,
+      "learning_rate": 1.4690876171795899e-07,
+      "loss": 1.2431,
+      "step": 45120
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 47.5,
+      "learning_rate": 1.4653061818715343e-07,
+      "loss": 1.3683,
+      "step": 45140
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 60.78125,
+      "learning_rate": 1.4615247465634785e-07,
+      "loss": 1.4307,
+      "step": 45160
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 74.0,
+      "learning_rate": 1.4577433112554233e-07,
+      "loss": 1.3282,
+      "step": 45180
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 59.28125,
+      "learning_rate": 1.4539618759473678e-07,
+      "loss": 1.3151,
+      "step": 45200
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 60.0,
+      "learning_rate": 1.450180440639312e-07,
+      "loss": 1.2706,
+      "step": 45220
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 62.0625,
+      "learning_rate": 1.4463990053312565e-07,
+      "loss": 1.2947,
+      "step": 45240
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 62.03125,
+      "learning_rate": 1.442617570023201e-07,
+      "loss": 1.3164,
+      "step": 45260
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 66.5,
+      "learning_rate": 1.4388361347151454e-07,
+      "loss": 1.2706,
+      "step": 45280
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 61.75,
+      "learning_rate": 1.4350546994070896e-07,
+      "loss": 1.3473,
+      "step": 45300
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 51.75,
+      "learning_rate": 1.431273264099034e-07,
+      "loss": 1.2948,
+      "step": 45320
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 59.3125,
+      "learning_rate": 1.4274918287909788e-07,
+      "loss": 1.262,
+      "step": 45340
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 101.625,
+      "learning_rate": 1.423710393482923e-07,
+      "loss": 1.3206,
+      "step": 45360
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 61.96875,
+      "learning_rate": 1.4199289581748675e-07,
+      "loss": 1.3655,
+      "step": 45380
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 56.375,
+      "learning_rate": 1.4161475228668117e-07,
+      "loss": 1.3355,
+      "step": 45400
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 61.6875,
+      "learning_rate": 1.4123660875587562e-07,
+      "loss": 1.3372,
+      "step": 45420
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 55.9375,
+      "learning_rate": 1.408584652250701e-07,
+      "loss": 1.2686,
+      "step": 45440
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 64.75,
+      "learning_rate": 1.4048032169426452e-07,
+      "loss": 1.2195,
+      "step": 45460
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 57.5625,
+      "learning_rate": 1.4010217816345897e-07,
+      "loss": 1.3275,
+      "step": 45480
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 62.40625,
+      "learning_rate": 1.397240346326534e-07,
+      "loss": 1.2764,
+      "step": 45500
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 54.5,
+      "learning_rate": 1.3934589110184786e-07,
+      "loss": 1.2676,
+      "step": 45520
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 69.625,
+      "learning_rate": 1.3896774757104228e-07,
+      "loss": 1.4004,
+      "step": 45540
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 61.78125,
+      "learning_rate": 1.3858960404023673e-07,
+      "loss": 1.3371,
+      "step": 45560
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 67.125,
+      "learning_rate": 1.3821146050943118e-07,
+      "loss": 1.4369,
+      "step": 45580
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 49.09375,
+      "learning_rate": 1.3783331697862563e-07,
+      "loss": 1.229,
+      "step": 45600
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 96.4375,
+      "learning_rate": 1.3745517344782007e-07,
+      "loss": 1.3351,
+      "step": 45620
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 69.5625,
+      "learning_rate": 1.370770299170145e-07,
+      "loss": 1.3794,
+      "step": 45640
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 50.8125,
+      "learning_rate": 1.3669888638620894e-07,
+      "loss": 1.3705,
+      "step": 45660
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 62.1875,
+      "learning_rate": 1.3632074285540336e-07,
+      "loss": 1.3257,
+      "step": 45680
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 54.1875,
+      "learning_rate": 1.3594259932459784e-07,
+      "loss": 1.2859,
+      "step": 45700
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 56.84375,
+      "learning_rate": 1.3556445579379229e-07,
+      "loss": 1.3751,
+      "step": 45720
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 51.90625,
+      "learning_rate": 1.351863122629867e-07,
+      "loss": 1.3439,
+      "step": 45740
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 58.6875,
+      "learning_rate": 1.3480816873218116e-07,
+      "loss": 1.3596,
+      "step": 45760
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 62.71875,
+      "learning_rate": 1.344300252013756e-07,
+      "loss": 1.3372,
+      "step": 45780
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 62.5625,
+      "learning_rate": 1.3405188167057005e-07,
+      "loss": 1.2702,
+      "step": 45800
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 85.125,
+      "learning_rate": 1.336737381397645e-07,
+      "loss": 1.3388,
+      "step": 45820
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 86.875,
+      "learning_rate": 1.3329559460895892e-07,
+      "loss": 1.2513,
+      "step": 45840
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 70.8125,
+      "learning_rate": 1.3291745107815337e-07,
+      "loss": 1.3841,
+      "step": 45860
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 64.8125,
+      "learning_rate": 1.3253930754734782e-07,
+      "loss": 1.2639,
+      "step": 45880
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 64.5,
+      "learning_rate": 1.3216116401654226e-07,
+      "loss": 1.4525,
+      "step": 45900
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 76.0,
+      "learning_rate": 1.3178302048573668e-07,
+      "loss": 1.3358,
+      "step": 45920
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 88.0625,
+      "learning_rate": 1.3140487695493113e-07,
+      "loss": 1.3377,
+      "step": 45940
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 56.9375,
+      "learning_rate": 1.310267334241256e-07,
+      "loss": 1.4271,
+      "step": 45960
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 67.4375,
+      "learning_rate": 1.3064858989332003e-07,
+      "loss": 1.3373,
+      "step": 45980
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 69.8125,
+      "learning_rate": 1.3027044636251448e-07,
+      "loss": 1.3895,
+      "step": 46000
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 66.125,
+      "learning_rate": 1.298923028317089e-07,
+      "loss": 1.354,
+      "step": 46020
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 49.0625,
+      "learning_rate": 1.2951415930090337e-07,
+      "loss": 1.3607,
+      "step": 46040
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 66.8125,
+      "learning_rate": 1.2913601577009782e-07,
+      "loss": 1.2385,
+      "step": 46060
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 79.0,
+      "learning_rate": 1.2875787223929224e-07,
+      "loss": 1.2652,
+      "step": 46080
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 58.09375,
+      "learning_rate": 1.283797287084867e-07,
+      "loss": 1.2984,
+      "step": 46100
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 64.75,
+      "learning_rate": 1.280015851776811e-07,
+      "loss": 1.2636,
+      "step": 46120
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 52.625,
+      "learning_rate": 1.2762344164687558e-07,
+      "loss": 1.4068,
+      "step": 46140
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 53.9375,
+      "learning_rate": 1.2724529811607e-07,
+      "loss": 1.3923,
+      "step": 46160
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 55.40625,
+      "learning_rate": 1.2686715458526445e-07,
+      "loss": 1.3176,
+      "step": 46180
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 70.4375,
+      "learning_rate": 1.264890110544589e-07,
+      "loss": 1.34,
+      "step": 46200
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 75.875,
+      "learning_rate": 1.2611086752365335e-07,
+      "loss": 1.3322,
+      "step": 46220
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 66.75,
+      "learning_rate": 1.257327239928478e-07,
+      "loss": 1.3074,
+      "step": 46240
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 77.4375,
+      "learning_rate": 1.2535458046204222e-07,
+      "loss": 1.2949,
+      "step": 46260
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 59.625,
+      "learning_rate": 1.2497643693123666e-07,
+      "loss": 1.2506,
+      "step": 46280
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 57.75,
+      "learning_rate": 1.245982934004311e-07,
+      "loss": 1.3649,
+      "step": 46300
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 74.375,
+      "learning_rate": 1.2422014986962556e-07,
+      "loss": 1.3469,
+      "step": 46320
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 65.5625,
+      "learning_rate": 1.2384200633882e-07,
+      "loss": 1.4123,
+      "step": 46340
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 76.875,
+      "learning_rate": 1.2346386280801443e-07,
+      "loss": 1.3209,
+      "step": 46360
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 64.75,
+      "learning_rate": 1.2308571927720888e-07,
+      "loss": 1.3176,
+      "step": 46380
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 74.8125,
+      "learning_rate": 1.2270757574640333e-07,
+      "loss": 1.3496,
+      "step": 46400
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 219.875,
+      "learning_rate": 1.2232943221559777e-07,
+      "loss": 1.3346,
+      "step": 46420
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 53.90625,
+      "learning_rate": 1.2195128868479222e-07,
+      "loss": 1.3611,
+      "step": 46440
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 66.375,
+      "learning_rate": 1.2157314515398667e-07,
+      "loss": 1.3481,
+      "step": 46460
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 82.375,
+      "learning_rate": 1.211950016231811e-07,
+      "loss": 1.4615,
+      "step": 46480
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 66.9375,
+      "learning_rate": 1.2081685809237554e-07,
+      "loss": 1.2017,
+      "step": 46500
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 53.75,
+      "learning_rate": 1.2043871456156999e-07,
+      "loss": 1.3885,
+      "step": 46520
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 74.75,
+      "learning_rate": 1.2006057103076443e-07,
+      "loss": 1.3945,
+      "step": 46540
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 53.8125,
+      "learning_rate": 1.1968242749995888e-07,
+      "loss": 1.4098,
+      "step": 46560
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 54.5625,
+      "learning_rate": 1.193042839691533e-07,
+      "loss": 1.2309,
+      "step": 46580
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 76.375,
+      "learning_rate": 1.1892614043834775e-07,
+      "loss": 1.3407,
+      "step": 46600
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 69.875,
+      "learning_rate": 1.185479969075422e-07,
+      "loss": 1.3216,
+      "step": 46620
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 89.9375,
+      "learning_rate": 1.1816985337673665e-07,
+      "loss": 1.4638,
+      "step": 46640
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 68.625,
+      "learning_rate": 1.1779170984593108e-07,
+      "loss": 1.3594,
+      "step": 46660
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 65.1875,
+      "learning_rate": 1.1741356631512553e-07,
+      "loss": 1.279,
+      "step": 46680
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 83.6875,
+      "learning_rate": 1.1703542278431996e-07,
+      "loss": 1.2234,
+      "step": 46700
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 65.4375,
+      "learning_rate": 1.1665727925351441e-07,
+      "loss": 1.2726,
+      "step": 46720
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 71.375,
+      "learning_rate": 1.1627913572270886e-07,
+      "loss": 1.3769,
+      "step": 46740
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 77.5625,
+      "learning_rate": 1.159009921919033e-07,
+      "loss": 1.3536,
+      "step": 46760
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 67.0,
+      "learning_rate": 1.1552284866109774e-07,
+      "loss": 1.2303,
+      "step": 46780
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 63.15625,
+      "learning_rate": 1.1514470513029219e-07,
+      "loss": 1.3474,
+      "step": 46800
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 57.59375,
+      "learning_rate": 1.1476656159948662e-07,
+      "loss": 1.3236,
+      "step": 46820
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 73.0625,
+      "learning_rate": 1.1438841806868106e-07,
+      "loss": 1.2705,
+      "step": 46840
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 69.5625,
+      "learning_rate": 1.1401027453787552e-07,
+      "loss": 1.3801,
+      "step": 46860
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 58.5625,
+      "learning_rate": 1.1363213100706995e-07,
+      "loss": 1.3132,
+      "step": 46880
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 69.75,
+      "learning_rate": 1.132539874762644e-07,
+      "loss": 1.3228,
+      "step": 46900
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 56.625,
+      "learning_rate": 1.1287584394545883e-07,
+      "loss": 1.3565,
+      "step": 46920
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 92.1875,
+      "learning_rate": 1.1249770041465328e-07,
+      "loss": 1.4107,
+      "step": 46940
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 75.375,
+      "learning_rate": 1.1211955688384772e-07,
+      "loss": 1.359,
+      "step": 46960
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 79.875,
+      "learning_rate": 1.1174141335304218e-07,
+      "loss": 1.3833,
+      "step": 46980
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 68.0625,
+      "learning_rate": 1.1136326982223661e-07,
+      "loss": 1.3504,
+      "step": 47000
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 83.9375,
+      "learning_rate": 1.1098512629143106e-07,
+      "loss": 1.3007,
+      "step": 47020
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 65.6875,
+      "learning_rate": 1.106069827606255e-07,
+      "loss": 1.3499,
+      "step": 47040
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 63.0,
+      "learning_rate": 1.1022883922981993e-07,
+      "loss": 1.3585,
+      "step": 47060
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 85.0,
+      "learning_rate": 1.0985069569901438e-07,
+      "loss": 1.3445,
+      "step": 47080
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 54.78125,
+      "learning_rate": 1.0947255216820881e-07,
+      "loss": 1.2835,
+      "step": 47100
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 61.59375,
+      "learning_rate": 1.0909440863740327e-07,
+      "loss": 1.2405,
+      "step": 47120
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 90.25,
+      "learning_rate": 1.0871626510659771e-07,
+      "loss": 1.2945,
+      "step": 47140
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 56.28125,
+      "learning_rate": 1.0833812157579216e-07,
+      "loss": 1.3271,
+      "step": 47160
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 56.03125,
+      "learning_rate": 1.0795997804498659e-07,
+      "loss": 1.3141,
+      "step": 47180
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 50.59375,
+      "learning_rate": 1.0758183451418104e-07,
+      "loss": 1.351,
+      "step": 47200
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 65.5,
+      "learning_rate": 1.0720369098337547e-07,
+      "loss": 1.3218,
+      "step": 47220
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 64.6875,
+      "learning_rate": 1.0682554745256993e-07,
+      "loss": 1.3122,
+      "step": 47240
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 51.875,
+      "learning_rate": 1.0644740392176437e-07,
+      "loss": 1.3517,
+      "step": 47260
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 78.9375,
+      "learning_rate": 1.0606926039095882e-07,
+      "loss": 1.4208,
+      "step": 47280
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 51.4375,
+      "learning_rate": 1.0569111686015325e-07,
+      "loss": 1.336,
+      "step": 47300
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 86.375,
+      "learning_rate": 1.0531297332934768e-07,
+      "loss": 1.3902,
+      "step": 47320
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 88.625,
+      "learning_rate": 1.0493482979854213e-07,
+      "loss": 1.3191,
+      "step": 47340
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 86.0625,
+      "learning_rate": 1.0455668626773658e-07,
+      "loss": 1.307,
+      "step": 47360
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 67.75,
+      "learning_rate": 1.0417854273693103e-07,
+      "loss": 1.3867,
+      "step": 47380
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 61.125,
+      "learning_rate": 1.0380039920612546e-07,
+      "loss": 1.3503,
+      "step": 47400
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 62.9375,
+      "learning_rate": 1.0342225567531991e-07,
+      "loss": 1.271,
+      "step": 47420
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 60.4375,
+      "learning_rate": 1.0304411214451434e-07,
+      "loss": 1.3358,
+      "step": 47440
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 54.875,
+      "learning_rate": 1.0266596861370879e-07,
+      "loss": 1.3072,
+      "step": 47460
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 60.0625,
+      "learning_rate": 1.0228782508290324e-07,
+      "loss": 1.4395,
+      "step": 47480
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 55.0625,
+      "learning_rate": 1.0190968155209769e-07,
+      "loss": 1.2708,
+      "step": 47500
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 55.59375,
+      "learning_rate": 1.0153153802129212e-07,
+      "loss": 1.3655,
+      "step": 47520
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 70.8125,
+      "learning_rate": 1.0115339449048656e-07,
+      "loss": 1.3221,
+      "step": 47540
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 58.125,
+      "learning_rate": 1.00775250959681e-07,
+      "loss": 1.3146,
+      "step": 47560
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 82.375,
+      "learning_rate": 1.0039710742887544e-07,
+      "loss": 1.4091,
+      "step": 47580
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 81.6875,
+      "learning_rate": 1.000189638980699e-07,
+      "loss": 1.4975,
+      "step": 47600
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 64.625,
+      "learning_rate": 9.964082036726433e-08,
+      "loss": 1.2658,
+      "step": 47620
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 66.5,
+      "learning_rate": 9.926267683645878e-08,
+      "loss": 1.343,
+      "step": 47640
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 70.0,
+      "learning_rate": 9.888453330565322e-08,
+      "loss": 1.2768,
+      "step": 47660
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 71.9375,
+      "learning_rate": 9.850638977484766e-08,
+      "loss": 1.2412,
+      "step": 47680
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 61.0,
+      "learning_rate": 9.81282462440421e-08,
+      "loss": 1.3141,
+      "step": 47700
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 81.125,
+      "learning_rate": 9.775010271323656e-08,
+      "loss": 1.3319,
+      "step": 47720
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 70.6875,
+      "learning_rate": 9.7371959182431e-08,
+      "loss": 1.3988,
+      "step": 47740
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 61.25,
+      "learning_rate": 9.699381565162544e-08,
+      "loss": 1.2966,
+      "step": 47760
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 53.59375,
+      "learning_rate": 9.661567212081988e-08,
+      "loss": 1.3256,
+      "step": 47780
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 63.375,
+      "learning_rate": 9.623752859001431e-08,
+      "loss": 1.3215,
+      "step": 47800
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 111.6875,
+      "learning_rate": 9.585938505920876e-08,
+      "loss": 1.3497,
+      "step": 47820
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 64.625,
+      "learning_rate": 9.54812415284032e-08,
+      "loss": 1.3466,
+      "step": 47840
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 71.125,
+      "learning_rate": 9.510309799759766e-08,
+      "loss": 1.3301,
+      "step": 47860
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 64.0625,
+      "learning_rate": 9.472495446679209e-08,
+      "loss": 1.4499,
+      "step": 47880
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 75.0625,
+      "learning_rate": 9.434681093598654e-08,
+      "loss": 1.4131,
+      "step": 47900
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 57.90625,
+      "learning_rate": 9.396866740518097e-08,
+      "loss": 1.3384,
+      "step": 47920
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 60.09375,
+      "learning_rate": 9.359052387437542e-08,
+      "loss": 1.3501,
+      "step": 47940
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 54.53125,
+      "learning_rate": 9.321238034356985e-08,
+      "loss": 1.3278,
+      "step": 47960
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 58.0,
+      "learning_rate": 9.283423681276432e-08,
+      "loss": 1.3404,
+      "step": 47980
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 63.53125,
+      "learning_rate": 9.245609328195875e-08,
+      "loss": 1.3284,
+      "step": 48000
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 79.8125,
+      "learning_rate": 9.207794975115318e-08,
+      "loss": 1.3528,
+      "step": 48020
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 68.3125,
+      "learning_rate": 9.169980622034763e-08,
+      "loss": 1.4362,
+      "step": 48040
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 71.4375,
+      "learning_rate": 9.132166268954207e-08,
+      "loss": 1.4823,
+      "step": 48060
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 95.5,
+      "learning_rate": 9.094351915873651e-08,
+      "loss": 1.3332,
+      "step": 48080
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 70.9375,
+      "learning_rate": 9.056537562793096e-08,
+      "loss": 1.3638,
+      "step": 48100
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 70.5625,
+      "learning_rate": 9.018723209712541e-08,
+      "loss": 1.3538,
+      "step": 48120
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 54.09375,
+      "learning_rate": 8.980908856631984e-08,
+      "loss": 1.3523,
+      "step": 48140
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 84.375,
+      "learning_rate": 8.943094503551429e-08,
+      "loss": 1.3176,
+      "step": 48160
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 64.25,
+      "learning_rate": 8.905280150470873e-08,
+      "loss": 1.4906,
+      "step": 48180
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 52.375,
+      "learning_rate": 8.867465797390317e-08,
+      "loss": 1.3775,
+      "step": 48200
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 75.9375,
+      "learning_rate": 8.829651444309762e-08,
+      "loss": 1.2116,
+      "step": 48220
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 62.75,
+      "learning_rate": 8.791837091229207e-08,
+      "loss": 1.2772,
+      "step": 48240
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 61.21875,
+      "learning_rate": 8.75402273814865e-08,
+      "loss": 1.4184,
+      "step": 48260
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 67.25,
+      "learning_rate": 8.716208385068094e-08,
+      "loss": 1.3029,
+      "step": 48280
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 62.09375,
+      "learning_rate": 8.678394031987539e-08,
+      "loss": 1.3634,
+      "step": 48300
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 79.5,
+      "learning_rate": 8.640579678906982e-08,
+      "loss": 1.3963,
+      "step": 48320
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 92.5625,
+      "learning_rate": 8.602765325826428e-08,
+      "loss": 1.3531,
+      "step": 48340
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 78.75,
+      "learning_rate": 8.564950972745872e-08,
+      "loss": 1.3542,
+      "step": 48360
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 79.8125,
+      "learning_rate": 8.527136619665317e-08,
+      "loss": 1.3637,
+      "step": 48380
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 83.0625,
+      "learning_rate": 8.48932226658476e-08,
+      "loss": 1.3512,
+      "step": 48400
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 45.3125,
+      "learning_rate": 8.451507913504205e-08,
+      "loss": 1.242,
+      "step": 48420
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 78.125,
+      "learning_rate": 8.413693560423648e-08,
+      "loss": 1.3456,
+      "step": 48440
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 67.5,
+      "learning_rate": 8.375879207343094e-08,
+      "loss": 1.2762,
+      "step": 48460
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 59.4375,
+      "learning_rate": 8.338064854262538e-08,
+      "loss": 1.3527,
+      "step": 48480
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 75.75,
+      "learning_rate": 8.300250501181981e-08,
+      "loss": 1.3539,
+      "step": 48500
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 71.625,
+      "learning_rate": 8.262436148101426e-08,
+      "loss": 1.2919,
+      "step": 48520
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 77.4375,
+      "learning_rate": 8.22462179502087e-08,
+      "loss": 1.2577,
+      "step": 48540
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 56.90625,
+      "learning_rate": 8.186807441940314e-08,
+      "loss": 1.286,
+      "step": 48560
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 60.75,
+      "learning_rate": 8.148993088859758e-08,
+      "loss": 1.4127,
+      "step": 48580
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 53.875,
+      "learning_rate": 8.111178735779204e-08,
+      "loss": 1.2502,
+      "step": 48600
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 62.5,
+      "learning_rate": 8.073364382698647e-08,
+      "loss": 1.2631,
+      "step": 48620
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 70.3125,
+      "learning_rate": 8.035550029618092e-08,
+      "loss": 1.2837,
+      "step": 48640
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 49.3125,
+      "learning_rate": 7.997735676537535e-08,
+      "loss": 1.3133,
+      "step": 48660
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 60.4375,
+      "learning_rate": 7.95992132345698e-08,
+      "loss": 1.4197,
+      "step": 48680
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 60.8125,
+      "learning_rate": 7.922106970376424e-08,
+      "loss": 1.3261,
+      "step": 48700
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 77.0,
+      "learning_rate": 7.884292617295868e-08,
+      "loss": 1.2969,
+      "step": 48720
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 58.09375,
+      "learning_rate": 7.846478264215313e-08,
+      "loss": 1.2446,
+      "step": 48740
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 52.125,
+      "learning_rate": 7.808663911134757e-08,
+      "loss": 1.2991,
+      "step": 48760
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 62.4375,
+      "learning_rate": 7.770849558054201e-08,
+      "loss": 1.2746,
+      "step": 48780
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 53.84375,
+      "learning_rate": 7.733035204973645e-08,
+      "loss": 1.3841,
+      "step": 48800
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 66.625,
+      "learning_rate": 7.69522085189309e-08,
+      "loss": 1.3453,
+      "step": 48820
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 84.25,
+      "learning_rate": 7.657406498812534e-08,
+      "loss": 1.2658,
+      "step": 48840
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 76.75,
+      "learning_rate": 7.619592145731979e-08,
+      "loss": 1.2577,
+      "step": 48860
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 78.125,
+      "learning_rate": 7.581777792651423e-08,
+      "loss": 1.3535,
+      "step": 48880
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 65.375,
+      "learning_rate": 7.543963439570867e-08,
+      "loss": 1.2434,
+      "step": 48900
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 67.3125,
+      "learning_rate": 7.506149086490311e-08,
+      "loss": 1.2902,
+      "step": 48920
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 67.9375,
+      "learning_rate": 7.468334733409756e-08,
+      "loss": 1.4072,
+      "step": 48940
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 68.8125,
+      "learning_rate": 7.4305203803292e-08,
+      "loss": 1.281,
+      "step": 48960
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 75.0625,
+      "learning_rate": 7.392706027248644e-08,
+      "loss": 1.2644,
+      "step": 48980
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 82.875,
+      "learning_rate": 7.354891674168089e-08,
+      "loss": 1.4248,
+      "step": 49000
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 63.3125,
+      "learning_rate": 7.317077321087532e-08,
+      "loss": 1.4421,
+      "step": 49020
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 79.1875,
+      "learning_rate": 7.279262968006977e-08,
+      "loss": 1.3399,
+      "step": 49040
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 65.3125,
+      "learning_rate": 7.24144861492642e-08,
+      "loss": 1.3482,
+      "step": 49060
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 74.5,
+      "learning_rate": 7.203634261845867e-08,
+      "loss": 1.3647,
+      "step": 49080
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 72.25,
+      "learning_rate": 7.16581990876531e-08,
+      "loss": 1.2468,
+      "step": 49100
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 61.5,
+      "learning_rate": 7.128005555684755e-08,
+      "loss": 1.3302,
+      "step": 49120
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 70.875,
+      "learning_rate": 7.090191202604198e-08,
+      "loss": 1.3486,
+      "step": 49140
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 69.625,
+      "learning_rate": 7.052376849523643e-08,
+      "loss": 1.2402,
+      "step": 49160
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 71.25,
+      "learning_rate": 7.014562496443086e-08,
+      "loss": 1.3408,
+      "step": 49180
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 61.40625,
+      "learning_rate": 6.97674814336253e-08,
+      "loss": 1.3726,
+      "step": 49200
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 45.6875,
+      "learning_rate": 6.938933790281976e-08,
+      "loss": 1.2908,
+      "step": 49220
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 66.125,
+      "learning_rate": 6.90111943720142e-08,
+      "loss": 1.2688,
+      "step": 49240
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 44.46875,
+      "learning_rate": 6.863305084120864e-08,
+      "loss": 1.2785,
+      "step": 49260
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 76.25,
+      "learning_rate": 6.825490731040308e-08,
+      "loss": 1.2984,
+      "step": 49280
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 76.9375,
+      "learning_rate": 6.787676377959752e-08,
+      "loss": 1.394,
+      "step": 49300
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 62.21875,
+      "learning_rate": 6.749862024879196e-08,
+      "loss": 1.336,
+      "step": 49320
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 67.875,
+      "learning_rate": 6.712047671798642e-08,
+      "loss": 1.3415,
+      "step": 49340
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 57.78125,
+      "learning_rate": 6.674233318718085e-08,
+      "loss": 1.428,
+      "step": 49360
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 74.8125,
+      "learning_rate": 6.63641896563753e-08,
+      "loss": 1.2926,
+      "step": 49380
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 58.21875,
+      "learning_rate": 6.598604612556974e-08,
+      "loss": 1.3439,
+      "step": 49400
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 52.34375,
+      "learning_rate": 6.560790259476418e-08,
+      "loss": 1.2582,
+      "step": 49420
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 63.09375,
+      "learning_rate": 6.522975906395862e-08,
+      "loss": 1.385,
+      "step": 49440
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 64.4375,
+      "learning_rate": 6.485161553315307e-08,
+      "loss": 1.3645,
+      "step": 49460
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 74.875,
+      "learning_rate": 6.447347200234751e-08,
+      "loss": 1.3553,
+      "step": 49480
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 76.5,
+      "learning_rate": 6.409532847154195e-08,
+      "loss": 1.3469,
+      "step": 49500
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 62.90625,
+      "learning_rate": 6.37171849407364e-08,
+      "loss": 1.2842,
+      "step": 49520
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 57.90625,
+      "learning_rate": 6.333904140993083e-08,
+      "loss": 1.2475,
+      "step": 49540
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 51.21875,
+      "learning_rate": 6.296089787912528e-08,
+      "loss": 1.4054,
+      "step": 49560
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 63.75,
+      "learning_rate": 6.258275434831973e-08,
+      "loss": 1.2913,
+      "step": 49580
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 62.875,
+      "learning_rate": 6.220461081751416e-08,
+      "loss": 1.2777,
+      "step": 49600
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 67.8125,
+      "learning_rate": 6.182646728670861e-08,
+      "loss": 1.4151,
+      "step": 49620
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 68.8125,
+      "learning_rate": 6.144832375590306e-08,
+      "loss": 1.3384,
+      "step": 49640
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 93.875,
+      "learning_rate": 6.107018022509749e-08,
+      "loss": 1.3058,
+      "step": 49660
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 106.4375,
+      "learning_rate": 6.069203669429194e-08,
+      "loss": 1.358,
+      "step": 49680
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 59.34375,
+      "learning_rate": 6.031389316348639e-08,
+      "loss": 1.3103,
+      "step": 49700
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 67.625,
+      "learning_rate": 5.993574963268082e-08,
+      "loss": 1.2339,
+      "step": 49720
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 74.5,
+      "learning_rate": 5.955760610187527e-08,
+      "loss": 1.3031,
+      "step": 49740
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 79.8125,
+      "learning_rate": 5.917946257106971e-08,
+      "loss": 1.2387,
+      "step": 49760
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 60.5,
+      "learning_rate": 5.880131904026415e-08,
+      "loss": 1.2539,
+      "step": 49780
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 79.75,
+      "learning_rate": 5.842317550945859e-08,
+      "loss": 1.3769,
+      "step": 49800
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 86.0625,
+      "learning_rate": 5.8045031978653034e-08,
+      "loss": 1.3498,
+      "step": 49820
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 66.875,
+      "learning_rate": 5.7666888447847475e-08,
+      "loss": 1.345,
+      "step": 49840
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 61.03125,
+      "learning_rate": 5.728874491704192e-08,
+      "loss": 1.3413,
+      "step": 49860
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 69.1875,
+      "learning_rate": 5.6910601386236364e-08,
+      "loss": 1.3266,
+      "step": 49880
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 115.125,
+      "learning_rate": 5.6532457855430805e-08,
+      "loss": 1.3336,
+      "step": 49900
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 59.78125,
+      "learning_rate": 5.615431432462525e-08,
+      "loss": 1.4154,
+      "step": 49920
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 62.8125,
+      "learning_rate": 5.5776170793819694e-08,
+      "loss": 1.4303,
+      "step": 49940
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 88.4375,
+      "learning_rate": 5.5398027263014136e-08,
+      "loss": 1.4158,
+      "step": 49960
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 75.25,
+      "learning_rate": 5.5019883732208583e-08,
+      "loss": 1.2602,
+      "step": 49980
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 69.8125,
+      "learning_rate": 5.4641740201403024e-08,
+      "loss": 1.3976,
+      "step": 50000
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 69.875,
+      "learning_rate": 5.4263596670597466e-08,
+      "loss": 1.304,
+      "step": 50020
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 69.125,
+      "learning_rate": 5.388545313979191e-08,
+      "loss": 1.2925,
+      "step": 50040
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 64.0,
+      "learning_rate": 5.350730960898635e-08,
+      "loss": 1.3776,
+      "step": 50060
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 53.0,
+      "learning_rate": 5.312916607818079e-08,
+      "loss": 1.4188,
+      "step": 50080
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 66.375,
+      "learning_rate": 5.275102254737524e-08,
+      "loss": 1.2999,
+      "step": 50100
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 64.8125,
+      "learning_rate": 5.237287901656968e-08,
+      "loss": 1.4259,
+      "step": 50120
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 58.84375,
+      "learning_rate": 5.199473548576412e-08,
+      "loss": 1.4155,
+      "step": 50140
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 85.25,
+      "learning_rate": 5.161659195495857e-08,
+      "loss": 1.2742,
+      "step": 50160
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 71.5625,
+      "learning_rate": 5.123844842415301e-08,
+      "loss": 1.3077,
+      "step": 50180
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 70.5625,
+      "learning_rate": 5.086030489334745e-08,
+      "loss": 1.4397,
+      "step": 50200
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 70.125,
+      "learning_rate": 5.04821613625419e-08,
+      "loss": 1.2983,
+      "step": 50220
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 74.375,
+      "learning_rate": 5.010401783173634e-08,
+      "loss": 1.3589,
+      "step": 50240
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 67.875,
+      "learning_rate": 4.972587430093078e-08,
+      "loss": 1.3045,
+      "step": 50260
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 64.5,
+      "learning_rate": 4.934773077012522e-08,
+      "loss": 1.3912,
+      "step": 50280
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 63.46875,
+      "learning_rate": 4.896958723931966e-08,
+      "loss": 1.3652,
+      "step": 50300
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 72.75,
+      "learning_rate": 4.85914437085141e-08,
+      "loss": 1.3422,
+      "step": 50320
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 53.46875,
+      "learning_rate": 4.821330017770855e-08,
+      "loss": 1.2532,
+      "step": 50340
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 56.9375,
+      "learning_rate": 4.783515664690299e-08,
+      "loss": 1.4502,
+      "step": 50360
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 51.40625,
+      "learning_rate": 4.745701311609743e-08,
+      "loss": 1.3256,
+      "step": 50380
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 88.25,
+      "learning_rate": 4.707886958529188e-08,
+      "loss": 1.2456,
+      "step": 50400
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 64.9375,
+      "learning_rate": 4.670072605448632e-08,
+      "loss": 1.3789,
+      "step": 50420
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 69.6875,
+      "learning_rate": 4.632258252368076e-08,
+      "loss": 1.3511,
+      "step": 50440
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 60.84375,
+      "learning_rate": 4.594443899287521e-08,
+      "loss": 1.3829,
+      "step": 50460
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 65.75,
+      "learning_rate": 4.556629546206965e-08,
+      "loss": 1.316,
+      "step": 50480
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 60.9375,
+      "learning_rate": 4.518815193126409e-08,
+      "loss": 1.3926,
+      "step": 50500
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 68.9375,
+      "learning_rate": 4.481000840045853e-08,
+      "loss": 1.3642,
+      "step": 50520
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 52.03125,
+      "learning_rate": 4.4431864869652975e-08,
+      "loss": 1.3064,
+      "step": 50540
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 64.8125,
+      "learning_rate": 4.4053721338847417e-08,
+      "loss": 1.3323,
+      "step": 50560
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 67.25,
+      "learning_rate": 4.367557780804186e-08,
+      "loss": 1.2735,
+      "step": 50580
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 55.1875,
+      "learning_rate": 4.3297434277236306e-08,
+      "loss": 1.3655,
+      "step": 50600
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 71.1875,
+      "learning_rate": 4.2919290746430747e-08,
+      "loss": 1.2776,
+      "step": 50620
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 62.9375,
+      "learning_rate": 4.254114721562519e-08,
+      "loss": 1.2622,
+      "step": 50640
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 79.0,
+      "learning_rate": 4.2163003684819636e-08,
+      "loss": 1.4572,
+      "step": 50660
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 70.625,
+      "learning_rate": 4.178486015401408e-08,
+      "loss": 1.304,
+      "step": 50680
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 68.5,
+      "learning_rate": 4.140671662320852e-08,
+      "loss": 1.3719,
+      "step": 50700
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 65.0625,
+      "learning_rate": 4.1028573092402966e-08,
+      "loss": 1.3088,
+      "step": 50720
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 63.25,
+      "learning_rate": 4.065042956159741e-08,
+      "loss": 1.3505,
+      "step": 50740
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 61.8125,
+      "learning_rate": 4.027228603079184e-08,
+      "loss": 1.2913,
+      "step": 50760
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 54.9375,
+      "learning_rate": 3.989414249998629e-08,
+      "loss": 1.3873,
+      "step": 50780
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 62.9375,
+      "learning_rate": 3.951599896918073e-08,
+      "loss": 1.4092,
+      "step": 50800
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 67.5,
+      "learning_rate": 3.913785543837517e-08,
+      "loss": 1.2893,
+      "step": 50820
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 59.625,
+      "learning_rate": 3.875971190756962e-08,
+      "loss": 1.354,
+      "step": 50840
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 54.25,
+      "learning_rate": 3.838156837676406e-08,
+      "loss": 1.3941,
+      "step": 50860
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 72.8125,
+      "learning_rate": 3.80034248459585e-08,
+      "loss": 1.3033,
+      "step": 50880
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 69.625,
+      "learning_rate": 3.762528131515295e-08,
+      "loss": 1.2482,
+      "step": 50900
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 53.9375,
+      "learning_rate": 3.724713778434739e-08,
+      "loss": 1.3942,
+      "step": 50920
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 89.1875,
+      "learning_rate": 3.686899425354183e-08,
+      "loss": 1.3397,
+      "step": 50940
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 58.75,
+      "learning_rate": 3.649085072273628e-08,
+      "loss": 1.2749,
+      "step": 50960
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 71.3125,
+      "learning_rate": 3.611270719193072e-08,
+      "loss": 1.3422,
+      "step": 50980
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 54.34375,
+      "learning_rate": 3.5734563661125155e-08,
+      "loss": 1.2571,
+      "step": 51000
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 58.84375,
+      "learning_rate": 3.53564201303196e-08,
+      "loss": 1.2357,
+      "step": 51020
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 61.78125,
+      "learning_rate": 3.4978276599514044e-08,
+      "loss": 1.3384,
+      "step": 51040
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 61.40625,
+      "learning_rate": 3.4600133068708485e-08,
+      "loss": 1.1818,
+      "step": 51060
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 67.25,
+      "learning_rate": 3.422198953790293e-08,
+      "loss": 1.3936,
+      "step": 51080
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 70.8125,
+      "learning_rate": 3.3843846007097374e-08,
+      "loss": 1.2922,
+      "step": 51100
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 62.75,
+      "learning_rate": 3.3465702476291815e-08,
+      "loss": 1.4717,
+      "step": 51120
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 55.375,
+      "learning_rate": 3.308755894548626e-08,
+      "loss": 1.342,
+      "step": 51140
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 84.625,
+      "learning_rate": 3.2709415414680704e-08,
+      "loss": 1.3032,
+      "step": 51160
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 67.625,
+      "learning_rate": 3.2331271883875145e-08,
+      "loss": 1.366,
+      "step": 51180
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 67.625,
+      "learning_rate": 3.195312835306959e-08,
+      "loss": 1.3584,
+      "step": 51200
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 65.25,
+      "learning_rate": 3.1574984822264034e-08,
+      "loss": 1.3087,
+      "step": 51220
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 61.5,
+      "learning_rate": 3.1196841291458475e-08,
+      "loss": 1.3388,
+      "step": 51240
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 63.78125,
+      "learning_rate": 3.0818697760652917e-08,
+      "loss": 1.2657,
+      "step": 51260
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 68.75,
+      "learning_rate": 3.044055422984736e-08,
+      "loss": 1.3103,
+      "step": 51280
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 56.4375,
+      "learning_rate": 3.00624106990418e-08,
+      "loss": 1.4243,
+      "step": 51300
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 83.875,
+      "learning_rate": 2.9684267168236243e-08,
+      "loss": 1.2968,
+      "step": 51320
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 64.3125,
+      "learning_rate": 2.9306123637430688e-08,
+      "loss": 1.313,
+      "step": 51340
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 78.875,
+      "learning_rate": 2.892798010662513e-08,
+      "loss": 1.3133,
+      "step": 51360
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 68.0625,
+      "learning_rate": 2.8549836575819574e-08,
+      "loss": 1.3116,
+      "step": 51380
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 72.0625,
+      "learning_rate": 2.8171693045014018e-08,
+      "loss": 1.3143,
+      "step": 51400
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 56.96875,
+      "learning_rate": 2.7793549514208456e-08,
+      "loss": 1.3107,
+      "step": 51420
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 57.75,
+      "learning_rate": 2.74154059834029e-08,
+      "loss": 1.3432,
+      "step": 51440
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 75.25,
+      "learning_rate": 2.7037262452597345e-08,
+      "loss": 1.3119,
+      "step": 51460
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 81.875,
+      "learning_rate": 2.6659118921791786e-08,
+      "loss": 1.2609,
+      "step": 51480
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 70.25,
+      "learning_rate": 2.628097539098623e-08,
+      "loss": 1.2975,
+      "step": 51500
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 63.9375,
+      "learning_rate": 2.5902831860180668e-08,
+      "loss": 1.3698,
+      "step": 51520
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 72.125,
+      "learning_rate": 2.5524688329375113e-08,
+      "loss": 1.3457,
+      "step": 51540
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 62.4375,
+      "learning_rate": 2.5146544798569557e-08,
+      "loss": 1.3434,
+      "step": 51560
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 92.9375,
+      "learning_rate": 2.4768401267763998e-08,
+      "loss": 1.3211,
+      "step": 51580
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 59.6875,
+      "learning_rate": 2.4390257736958443e-08,
+      "loss": 1.3081,
+      "step": 51600
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 62.09375,
+      "learning_rate": 2.4012114206152887e-08,
+      "loss": 1.333,
+      "step": 51620
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 54.59375,
+      "learning_rate": 2.3633970675347325e-08,
+      "loss": 1.316,
+      "step": 51640
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 70.9375,
+      "learning_rate": 2.325582714454177e-08,
+      "loss": 1.3066,
+      "step": 51660
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 74.5625,
+      "learning_rate": 2.2877683613736214e-08,
+      "loss": 1.3551,
+      "step": 51680
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 72.0,
+      "learning_rate": 2.2499540082930655e-08,
+      "loss": 1.3349,
+      "step": 51700
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 69.0,
+      "learning_rate": 2.21213965521251e-08,
+      "loss": 1.2775,
+      "step": 51720
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 74.4375,
+      "learning_rate": 2.1743253021319544e-08,
+      "loss": 1.3279,
+      "step": 51740
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 79.25,
+      "learning_rate": 2.1365109490513982e-08,
+      "loss": 1.4207,
+      "step": 51760
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 66.4375,
+      "learning_rate": 2.0986965959708426e-08,
+      "loss": 1.3499,
+      "step": 51780
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 64.1875,
+      "learning_rate": 2.060882242890287e-08,
+      "loss": 1.3024,
+      "step": 51800
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 53.90625,
+      "learning_rate": 2.0230678898097312e-08,
+      "loss": 1.2509,
+      "step": 51820
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 58.78125,
+      "learning_rate": 1.9852535367291757e-08,
+      "loss": 1.3702,
+      "step": 51840
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 57.5625,
+      "learning_rate": 1.94743918364862e-08,
+      "loss": 1.3503,
+      "step": 51860
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 58.75,
+      "learning_rate": 1.909624830568064e-08,
+      "loss": 1.2938,
+      "step": 51880
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 53.6875,
+      "learning_rate": 1.8718104774875083e-08,
+      "loss": 1.384,
+      "step": 51900
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 69.6875,
+      "learning_rate": 1.8339961244069524e-08,
+      "loss": 1.3979,
+      "step": 51920
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 62.40625,
+      "learning_rate": 1.796181771326397e-08,
+      "loss": 1.3146,
+      "step": 51940
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 65.6875,
+      "learning_rate": 1.7583674182458413e-08,
+      "loss": 1.2778,
+      "step": 51960
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 57.375,
+      "learning_rate": 1.7205530651652855e-08,
+      "loss": 1.2466,
+      "step": 51980
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 77.25,
+      "learning_rate": 1.6827387120847296e-08,
+      "loss": 1.3475,
+      "step": 52000
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 56.28125,
+      "learning_rate": 1.644924359004174e-08,
+      "loss": 1.2942,
+      "step": 52020
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 68.5,
+      "learning_rate": 1.607110005923618e-08,
+      "loss": 1.3799,
+      "step": 52040
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 57.90625,
+      "learning_rate": 1.5692956528430626e-08,
+      "loss": 1.366,
+      "step": 52060
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 70.5625,
+      "learning_rate": 1.531481299762507e-08,
+      "loss": 1.3255,
+      "step": 52080
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 59.0625,
+      "learning_rate": 1.493666946681951e-08,
+      "loss": 1.3114,
+      "step": 52100
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 61.46875,
+      "learning_rate": 1.4558525936013954e-08,
+      "loss": 1.2898,
+      "step": 52120
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 57.6875,
+      "learning_rate": 1.4180382405208397e-08,
+      "loss": 1.3028,
+      "step": 52140
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 61.96875,
+      "learning_rate": 1.380223887440284e-08,
+      "loss": 1.3921,
+      "step": 52160
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 61.53125,
+      "learning_rate": 1.3424095343597283e-08,
+      "loss": 1.3036,
+      "step": 52180
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 52.0625,
+      "learning_rate": 1.3045951812791725e-08,
+      "loss": 1.2947,
+      "step": 52200
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 63.03125,
+      "learning_rate": 1.2667808281986167e-08,
+      "loss": 1.3312,
+      "step": 52220
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 68.375,
+      "learning_rate": 1.2289664751180611e-08,
+      "loss": 1.2572,
+      "step": 52240
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 65.9375,
+      "learning_rate": 1.1911521220375054e-08,
+      "loss": 1.346,
+      "step": 52260
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 60.625,
+      "learning_rate": 1.1533377689569495e-08,
+      "loss": 1.2104,
+      "step": 52280
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 89.4375,
+      "learning_rate": 1.115523415876394e-08,
+      "loss": 1.3119,
+      "step": 52300
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 72.875,
+      "learning_rate": 1.0777090627958382e-08,
+      "loss": 1.4323,
+      "step": 52320
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 59.21875,
+      "learning_rate": 1.0398947097152824e-08,
+      "loss": 1.2751,
+      "step": 52340
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 63.96875,
+      "learning_rate": 1.0020803566347268e-08,
+      "loss": 1.3833,
+      "step": 52360
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 91.8125,
+      "learning_rate": 9.64266003554171e-09,
+      "loss": 1.4989,
+      "step": 52380
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 57.59375,
+      "learning_rate": 9.264516504736152e-09,
+      "loss": 1.3041,
+      "step": 52400
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 62.28125,
+      "learning_rate": 8.886372973930595e-09,
+      "loss": 1.2397,
+      "step": 52420
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 51.78125,
+      "learning_rate": 8.50822944312504e-09,
+      "loss": 1.3625,
+      "step": 52440
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 75.5,
+      "learning_rate": 8.13008591231948e-09,
+      "loss": 1.3415,
+      "step": 52460
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 53.59375,
+      "learning_rate": 7.751942381513923e-09,
+      "loss": 1.232,
+      "step": 52480
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 67.1875,
+      "learning_rate": 7.373798850708366e-09,
+      "loss": 1.2606,
+      "step": 52500
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 62.0,
+      "learning_rate": 6.995655319902809e-09,
+      "loss": 1.3394,
+      "step": 52520
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 62.6875,
+      "learning_rate": 6.6175117890972525e-09,
+      "loss": 1.2363,
+      "step": 52540
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 62.625,
+      "learning_rate": 6.2393682582916944e-09,
+      "loss": 1.2659,
+      "step": 52560
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 82.5625,
+      "learning_rate": 5.861224727486137e-09,
+      "loss": 1.3677,
+      "step": 52580
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 54.65625,
+      "learning_rate": 5.48308119668058e-09,
+      "loss": 1.2904,
+      "step": 52600
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 76.4375,
+      "learning_rate": 5.104937665875023e-09,
+      "loss": 1.313,
+      "step": 52620
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 67.8125,
+      "learning_rate": 4.726794135069465e-09,
+      "loss": 1.3009,
+      "step": 52640
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 75.75,
+      "learning_rate": 4.3486506042639085e-09,
+      "loss": 1.3895,
+      "step": 52660
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 87.25,
+      "learning_rate": 3.970507073458351e-09,
+      "loss": 1.3273,
+      "step": 52680
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 59.78125,
+      "learning_rate": 3.592363542652794e-09,
+      "loss": 1.3326,
+      "step": 52700
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 73.0,
+      "learning_rate": 3.2142200118472365e-09,
+      "loss": 1.2185,
+      "step": 52720
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 61.0625,
+      "learning_rate": 2.8360764810416793e-09,
+      "loss": 1.4331,
+      "step": 52740
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 66.8125,
+      "learning_rate": 2.457932950236122e-09,
+      "loss": 1.2308,
+      "step": 52760
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 60.71875,
+      "learning_rate": 2.0797894194305645e-09,
+      "loss": 1.4168,
+      "step": 52780
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 56.5625,
+      "learning_rate": 1.7016458886250076e-09,
+      "loss": 1.3373,
+      "step": 52800
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 53.40625,
+      "learning_rate": 1.3235023578194504e-09,
+      "loss": 1.3291,
+      "step": 52820
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 54.625,
+      "learning_rate": 9.45358827013893e-10,
+      "loss": 1.38,
+      "step": 52840
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 84.875,
+      "learning_rate": 5.672152962083359e-10,
+      "loss": 1.2983,
+      "step": 52860
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 135.625,
+      "learning_rate": 1.8907176540277862e-10,
+      "loss": 1.4019,
+      "step": 52880
+    }
+  ],
+  "logging_steps": 20,
+  "max_steps": 52890,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "total_flos": 1.7722335641506468e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}