diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.50012156576708,
+  "epoch": 1.00024313153416,
   "eval_steps": 500,
-  "global_step": 2057,
+  "global_step": 4114,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -14406,6 +14406,14413 @@
       "learning_rate": 2.158971359258265e-06,
       "loss": 0.4969,
       "step": 2057
+    },
+    {
+      "epoch": 0.50036469730124,
+      "grad_norm": 21.25,
+      "learning_rate": 2.1586395558563363e-06,
+      "loss": 1.3187,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5006078288354,
+      "grad_norm": 18.5,
+      "learning_rate": 2.1583076166425328e-06,
+      "loss": 1.2774,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5008509603695599,
+      "grad_norm": 22.125,
+      "learning_rate": 2.157975541666469e-06,
+      "loss": 1.0634,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5010940919037199,
+      "grad_norm": 18.0,
+      "learning_rate": 2.1576433309777794e-06,
+      "loss": 0.8712,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5013372234378799,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.157310984626118e-06,
+      "loss": 0.6967,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5015803549720399,
+      "grad_norm": 23.375,
+      "learning_rate": 2.1569785026611605e-06,
+      "loss": 1.2955,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5018234865061999,
+      "grad_norm": 21.25,
+      "learning_rate": 2.1566458851326015e-06,
+      "loss": 0.8218,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5020666180403598,
+      "grad_norm": 15.125,
+      "learning_rate": 2.156313132090157e-06,
+      "loss": 0.6786,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5023097495745198,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1559802435835623e-06,
+      "loss": 0.669,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5025528811086798,
+      "grad_norm": 16.375,
+      "learning_rate": 2.1556472196625733e-06,
+      "loss": 0.71,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5027960126428398,
+      "grad_norm": 16.25,
+      "learning_rate": 2.155314060376966e-06,
+      "loss": 0.6031,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5030391441769998,
+      "grad_norm": 19.375,
+      "learning_rate": 2.1549807657765375e-06,
+      "loss": 1.1718,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5032822757111597,
+      "grad_norm": 26.125,
+      "learning_rate": 2.1546473359111037e-06,
+      "loss": 0.8826,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5035254072453197,
+      "grad_norm": 19.625,
+      "learning_rate": 2.154313770830502e-06,
+      "loss": 0.8892,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5037685387794797,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1539800705845886e-06,
+      "loss": 1.0622,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5040116703136397,
+      "grad_norm": 17.375,
+      "learning_rate": 2.1536462352232416e-06,
+      "loss": 0.7742,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5042548018477997,
+      "grad_norm": 16.25,
+      "learning_rate": 2.153312264796359e-06,
+      "loss": 1.0432,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5044979333819597,
+      "grad_norm": 20.75,
+      "learning_rate": 2.152978159353857e-06,
+      "loss": 0.6966,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5047410649161196,
+      "grad_norm": 20.75,
+      "learning_rate": 2.152643918945674e-06,
+      "loss": 0.9008,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5049841964502796,
+      "grad_norm": 15.375,
+      "learning_rate": 2.1523095436217685e-06,
+      "loss": 0.6613,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5052273279844396,
+      "grad_norm": 20.75,
+      "learning_rate": 2.151975033432118e-06,
+      "loss": 1.3671,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5054704595185996,
+      "grad_norm": 19.25,
+      "learning_rate": 2.151640388426721e-06,
+      "loss": 1.0413,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5057135910527596,
+      "grad_norm": 19.125,
+      "learning_rate": 2.151305608655597e-06,
+      "loss": 0.8604,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5059567225869195,
+      "grad_norm": 19.0,
+      "learning_rate": 2.1509706941687824e-06,
+      "loss": 0.794,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5061998541210795,
+      "grad_norm": 17.5,
+      "learning_rate": 2.150635645016338e-06,
+      "loss": 1.0629,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5064429856552395,
+      "grad_norm": 16.625,
+      "learning_rate": 2.150300461248342e-06,
+      "loss": 1.1524,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5066861171893995,
+      "grad_norm": 22.5,
+      "learning_rate": 2.149965142914893e-06,
+      "loss": 1.0185,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5069292487235595,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1496296900661106e-06,
+      "loss": 1.0102,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5071723802577194,
+      "grad_norm": 16.75,
+      "learning_rate": 2.149294102752134e-06,
+      "loss": 0.8522,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5074155117918794,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1489583810231217e-06,
+      "loss": 0.6616,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5076586433260394,
+      "grad_norm": 22.25,
+      "learning_rate": 2.148622524929255e-06,
+      "loss": 1.2406,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5079017748601994,
+      "grad_norm": 25.75,
+      "learning_rate": 2.148286534520731e-06,
+      "loss": 0.9268,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5081449063943594,
+      "grad_norm": 21.0,
+      "learning_rate": 2.147950409847771e-06,
+      "loss": 1.0492,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5083880379285193,
+      "grad_norm": 20.125,
+      "learning_rate": 2.1476141509606144e-06,
+      "loss": 1.084,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5086311694626793,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1472777579095205e-06,
+      "loss": 0.9973,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5088743009968393,
+      "grad_norm": 16.125,
+      "learning_rate": 2.146941230744769e-06,
+      "loss": 0.3655,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5091174325309993,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.14660456951666e-06,
+      "loss": 0.7654,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5093605640651593,
+      "grad_norm": 18.625,
+      "learning_rate": 2.146267774275513e-06,
+      "loss": 0.9237,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5096036955993193,
+      "grad_norm": 17.25,
+      "learning_rate": 2.145930845071668e-06,
+      "loss": 0.7004,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5098468271334792,
+      "grad_norm": 26.5,
+      "learning_rate": 2.145593781955485e-06,
+      "loss": 0.9624,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5100899586676392,
+      "grad_norm": 22.75,
+      "learning_rate": 2.145256584977344e-06,
+      "loss": 0.8197,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5103330902017992,
+      "grad_norm": 22.625,
+      "learning_rate": 2.1449192541876447e-06,
+      "loss": 1.3957,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5105762217359592,
+      "grad_norm": 25.5,
+      "learning_rate": 2.144581789636807e-06,
+      "loss": 0.9053,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5108193532701192,
+      "grad_norm": 17.125,
+      "learning_rate": 2.144244191375271e-06,
+      "loss": 0.8517,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5110624848042791,
+      "grad_norm": 17.125,
+      "learning_rate": 2.143906459453496e-06,
+      "loss": 0.8091,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5113056163384391,
+      "grad_norm": 23.125,
+      "learning_rate": 2.143568593921963e-06,
+      "loss": 0.8385,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5115487478725991,
+      "grad_norm": 20.75,
+      "learning_rate": 2.143230594831171e-06,
+      "loss": 0.9638,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5117918794067591,
+      "grad_norm": 19.25,
+      "learning_rate": 2.1428924622316396e-06,
+      "loss": 0.9676,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5120350109409191,
+      "grad_norm": 19.125,
+      "learning_rate": 2.1425541961739093e-06,
+      "loss": 0.8893,
+      "step": 2106
+    },
+    {
+      "epoch": 0.512278142475079,
+      "grad_norm": 17.375,
+      "learning_rate": 2.1422157967085394e-06,
+      "loss": 0.8004,
+      "step": 2107
+    },
+    {
+      "epoch": 0.512521274009239,
+      "grad_norm": 25.625,
+      "learning_rate": 2.1418772638861095e-06,
+      "loss": 0.8912,
+      "step": 2108
+    },
+    {
+      "epoch": 0.512764405543399,
+      "grad_norm": 18.5,
+      "learning_rate": 2.141538597757219e-06,
+      "loss": 0.6878,
+      "step": 2109
+    },
+    {
+      "epoch": 0.513007537077559,
+      "grad_norm": 20.0,
+      "learning_rate": 2.141199798372488e-06,
+      "loss": 0.8896,
+      "step": 2110
+    },
+    {
+      "epoch": 0.513250668611719,
+      "grad_norm": 25.25,
+      "learning_rate": 2.140860865782556e-06,
+      "loss": 1.0707,
+      "step": 2111
+    },
+    {
+      "epoch": 0.513493800145879,
+      "grad_norm": 31.25,
+      "learning_rate": 2.1405218000380813e-06,
+      "loss": 0.9574,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5137369316800389,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1401826011897436e-06,
+      "loss": 0.8259,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5139800632141989,
+      "grad_norm": 19.5,
+      "learning_rate": 2.1398432692882423e-06,
+      "loss": 1.1377,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5142231947483589,
+      "grad_norm": 17.125,
+      "learning_rate": 2.1395038043842966e-06,
+      "loss": 0.5954,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5144663262825189,
+      "grad_norm": 21.75,
+      "learning_rate": 2.139164206528645e-06,
+      "loss": 0.9733,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5147094578166789,
+      "grad_norm": 16.25,
+      "learning_rate": 2.138824475772046e-06,
+      "loss": 0.6823,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5149525893508388,
+      "grad_norm": 15.375,
+      "learning_rate": 2.138484612165279e-06,
+      "loss": 1.3318,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5151957208849988,
+      "grad_norm": 24.625,
+      "learning_rate": 2.138144615759142e-06,
+      "loss": 1.221,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5154388524191588,
+      "grad_norm": 17.125,
+      "learning_rate": 2.137804486604453e-06,
+      "loss": 0.7901,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5156819839533188,
+      "grad_norm": 16.125,
+      "learning_rate": 2.1374642247520506e-06,
+      "loss": 0.819,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5159251154874788,
+      "grad_norm": 17.375,
+      "learning_rate": 2.137123830252793e-06,
+      "loss": 0.6808,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5161682470216387,
+      "grad_norm": 20.0,
+      "learning_rate": 2.1367833031575576e-06,
+      "loss": 0.6986,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5164113785557987,
+      "grad_norm": 28.625,
+      "learning_rate": 2.1364426435172426e-06,
+      "loss": 0.9064,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5166545100899587,
+      "grad_norm": 26.875,
+      "learning_rate": 2.136101851382765e-06,
+      "loss": 0.8917,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5168976416241187,
+      "grad_norm": 17.375,
+      "learning_rate": 2.1357609268050623e-06,
+      "loss": 0.7116,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5171407731582787,
+      "grad_norm": 23.125,
+      "learning_rate": 2.135419869835091e-06,
+      "loss": 1.0916,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5173839046924386,
+      "grad_norm": 26.25,
+      "learning_rate": 2.1350786805238287e-06,
+      "loss": 1.1954,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5176270362265986,
+      "grad_norm": 28.0,
+      "learning_rate": 2.1347373589222718e-06,
+      "loss": 1.0237,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5178701677607586,
+      "grad_norm": 18.75,
+      "learning_rate": 2.134395905081437e-06,
+      "loss": 0.6715,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5181132992949186,
+      "grad_norm": 18.25,
+      "learning_rate": 2.13405431905236e-06,
+      "loss": 1.0831,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5183564308290786,
+      "grad_norm": 17.25,
+      "learning_rate": 2.1337126008860964e-06,
+      "loss": 0.7293,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5185995623632386,
+      "grad_norm": 20.125,
+      "learning_rate": 2.1333707506337227e-06,
+      "loss": 0.7882,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5188426938973985,
+      "grad_norm": 19.375,
+      "learning_rate": 2.133028768346334e-06,
+      "loss": 0.761,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5190858254315585,
+      "grad_norm": 16.5,
+      "learning_rate": 2.132686654075045e-06,
+      "loss": 0.5887,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5193289569657185,
+      "grad_norm": 22.0,
+      "learning_rate": 2.132344407870992e-06,
+      "loss": 0.9873,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5195720884998785,
+      "grad_norm": 28.25,
+      "learning_rate": 2.1320020297853276e-06,
+      "loss": 1.2165,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5198152200340385,
+      "grad_norm": 18.75,
+      "learning_rate": 2.1316595198692274e-06,
+      "loss": 1.0632,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5200583515681984,
+      "grad_norm": 25.625,
+      "learning_rate": 2.1313168781738855e-06,
+      "loss": 1.2686,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5203014831023584,
+      "grad_norm": 17.125,
+      "learning_rate": 2.130974104750515e-06,
+      "loss": 0.75,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5205446146365184,
+      "grad_norm": 22.0,
+      "learning_rate": 2.13063119965035e-06,
+      "loss": 1.0972,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5207877461706784,
+      "grad_norm": 16.125,
+      "learning_rate": 2.1302881629246426e-06,
+      "loss": 0.5599,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5210308777048384,
+      "grad_norm": 19.875,
+      "learning_rate": 2.1299449946246666e-06,
+      "loss": 1.2064,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5212740092389982,
+      "grad_norm": 21.75,
+      "learning_rate": 2.129601694801714e-06,
+      "loss": 1.2053,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5215171407731582,
+      "grad_norm": 20.375,
+      "learning_rate": 2.1292582635070966e-06,
+      "loss": 0.8525,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5217602723073183,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.128914700792146e-06,
+      "loss": 0.859,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5220034038414783,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.1285710067082147e-06,
+      "loss": 0.4744,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5222465353756383,
+      "grad_norm": 16.125,
+      "learning_rate": 2.128227181306673e-06,
+      "loss": 0.8446,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5224896669097983,
+      "grad_norm": 34.5,
+      "learning_rate": 2.1278832246389116e-06,
+      "loss": 0.8184,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5227327984439581,
+      "grad_norm": 24.375,
+      "learning_rate": 2.1275391367563403e-06,
+      "loss": 1.5662,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5229759299781181,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1271949177103894e-06,
+      "loss": 0.6558,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5232190615122781,
+      "grad_norm": 23.375,
+      "learning_rate": 2.1268505675525084e-06,
+      "loss": 0.8729,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5234621930464382,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1265060863341665e-06,
+      "loss": 1.039,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5237053245805982,
+      "grad_norm": 17.5,
+      "learning_rate": 2.1261614741068522e-06,
+      "loss": 0.7471,
+      "step": 2154
+    },
+    {
+      "epoch": 0.523948456114758,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.1258167309220737e-06,
+      "loss": 0.5395,
+      "step": 2155
+    },
+    {
+      "epoch": 0.524191587648918,
+      "grad_norm": 17.125,
+      "learning_rate": 2.125471856831359e-06,
+      "loss": 0.7584,
+      "step": 2156
+    },
+    {
+      "epoch": 0.524434719183078,
+      "grad_norm": 41.25,
+      "learning_rate": 2.125126851886255e-06,
+      "loss": 0.7686,
+      "step": 2157
+    },
+    {
+      "epoch": 0.524677850717238,
+      "grad_norm": 15.5,
+      "learning_rate": 2.1247817161383295e-06,
+      "loss": 0.4855,
+      "step": 2158
+    },
+    {
+      "epoch": 0.524920982251398,
+      "grad_norm": 16.875,
+      "learning_rate": 2.1244364496391684e-06,
+      "loss": 0.7316,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5251641137855579,
+      "grad_norm": 20.375,
+      "learning_rate": 2.124091052440378e-06,
+      "loss": 1.142,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5254072453197179,
+      "grad_norm": 22.75,
+      "learning_rate": 2.123745524593583e-06,
+      "loss": 0.729,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5256503768538779,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1233998661504297e-06,
+      "loss": 0.9577,
+      "step": 2162
+    },
+    {
+      "epoch": 0.525893508388038,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.1230540771625823e-06,
+      "loss": 0.657,
+      "step": 2163
+    },
+    {
+      "epoch": 0.526136639922198,
+      "grad_norm": 16.75,
+      "learning_rate": 2.122708157681725e-06,
+      "loss": 1.1673,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5263797714563578,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.122362107759561e-06,
+      "loss": 0.7474,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5266229029905178,
+      "grad_norm": 19.625,
+      "learning_rate": 2.1220159274478145e-06,
+      "loss": 0.9294,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5268660345246778,
+      "grad_norm": 17.25,
+      "learning_rate": 2.121669616798227e-06,
+      "loss": 0.7146,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5271091660588378,
+      "grad_norm": 23.25,
+      "learning_rate": 2.1213231758625606e-06,
+      "loss": 0.9958,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5273522975929978,
+      "grad_norm": 18.5,
+      "learning_rate": 2.1209766046925976e-06,
+      "loss": 0.7999,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5275954291271578,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.120629903340139e-06,
+      "loss": 0.6183,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5278385606613177,
+      "grad_norm": 20.75,
+      "learning_rate": 2.120283071857005e-06,
+      "loss": 0.8308,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5280816921954777,
+      "grad_norm": 18.625,
+      "learning_rate": 2.1199361102950357e-06,
+      "loss": 0.6661,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5283248237296377,
+      "grad_norm": 22.375,
+      "learning_rate": 2.11958901870609e-06,
+      "loss": 0.8004,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5285679552637977,
+      "grad_norm": 23.5,
+      "learning_rate": 2.119241797142047e-06,
+      "loss": 1.2852,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5288110867979577,
+      "grad_norm": 27.0,
+      "learning_rate": 2.1188944456548054e-06,
+      "loss": 0.9778,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5290542183321176,
+      "grad_norm": 23.0,
+      "learning_rate": 2.1185469642962826e-06,
+      "loss": 0.8277,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5292973498662776,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1181993531184156e-06,
+      "loss": 0.6125,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5295404814004376,
+      "grad_norm": 20.375,
+      "learning_rate": 2.117851612173161e-06,
+      "loss": 1.027,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5297836129345976,
+      "grad_norm": 17.875,
+      "learning_rate": 2.1175037415124947e-06,
+      "loss": 0.8801,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5300267444687576,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1171557411884116e-06,
+      "loss": 1.0453,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5302698760029175,
+      "grad_norm": 25.875,
+      "learning_rate": 2.116807611252927e-06,
+      "loss": 1.467,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5305130075370775,
+      "grad_norm": 16.875,
+      "learning_rate": 2.1164593517580746e-06,
+      "loss": 0.9979,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5307561390712375,
+      "grad_norm": 15.375,
+      "learning_rate": 2.116110962755908e-06,
+      "loss": 0.6979,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5309992706053975,
+      "grad_norm": 31.0,
+      "learning_rate": 2.1157624442984993e-06,
+      "loss": 1.4906,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5312424021395575,
+      "grad_norm": 19.625,
+      "learning_rate": 2.115413796437941e-06,
+      "loss": 0.9826,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5314855336737175,
+      "grad_norm": 20.625,
+      "learning_rate": 2.115065019226345e-06,
+      "loss": 0.7923,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5317286652078774,
+      "grad_norm": 20.625,
+      "learning_rate": 2.114716112715842e-06,
+      "loss": 1.0459,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5319717967420374,
+      "grad_norm": 17.25,
+      "learning_rate": 2.114367076958581e-06,
+      "loss": 1.155,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5322149282761974,
+      "grad_norm": 18.625,
+      "learning_rate": 2.1140179120067324e-06,
+      "loss": 1.0049,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5324580598103574,
+      "grad_norm": 19.75,
+      "learning_rate": 2.113668617912485e-06,
+      "loss": 0.8863,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5327011913445174,
+      "grad_norm": 15.375,
+      "learning_rate": 2.1133191947280465e-06,
+      "loss": 0.7787,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5329443228786773,
+      "grad_norm": 21.125,
+      "learning_rate": 2.112969642505644e-06,
+      "loss": 0.8467,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5331874544128373,
+      "grad_norm": 22.25,
+      "learning_rate": 2.112619961297525e-06,
+      "loss": 0.7615,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5334305859469973,
+      "grad_norm": 23.625,
+      "learning_rate": 2.1122701511559548e-06,
+      "loss": 0.7331,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5336737174811573,
+      "grad_norm": 17.875,
+      "learning_rate": 2.1119202121332185e-06,
+      "loss": 0.751,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5339168490153173,
+      "grad_norm": 24.625,
+      "learning_rate": 2.11157014428162e-06,
+      "loss": 1.0139,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5341599805494772,
+      "grad_norm": 17.125,
+      "learning_rate": 2.111219947653484e-06,
+      "loss": 0.7356,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5344031120836372,
+      "grad_norm": 23.625,
+      "learning_rate": 2.1108696223011534e-06,
+      "loss": 1.0466,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5346462436177972,
+      "grad_norm": 19.0,
+      "learning_rate": 2.1105191682769895e-06,
+      "loss": 0.7842,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5348893751519572,
+      "grad_norm": 25.125,
+      "learning_rate": 2.1101685856333744e-06,
+      "loss": 1.2724,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5351325066861172,
+      "grad_norm": 20.5,
+      "learning_rate": 2.1098178744227088e-06,
+      "loss": 0.8252,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5353756382202771,
+      "grad_norm": 24.5,
+      "learning_rate": 2.109467034697412e-06,
+      "loss": 0.5705,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5356187697544371,
+      "grad_norm": 22.25,
+      "learning_rate": 2.1091160665099235e-06,
+      "loss": 0.9459,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5358619012885971,
+      "grad_norm": 19.875,
+      "learning_rate": 2.108764969912701e-06,
+      "loss": 1.189,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5361050328227571,
+      "grad_norm": 23.625,
+      "learning_rate": 2.108413744958223e-06,
+      "loss": 1.2873,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5363481643569171,
+      "grad_norm": 17.125,
+      "learning_rate": 2.108062391698985e-06,
+      "loss": 0.7731,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5365912958910771,
+      "grad_norm": 18.625,
+      "learning_rate": 2.1077109101875036e-06,
+      "loss": 0.8861,
+      "step": 2207
+    },
+    {
+      "epoch": 0.536834427425237,
+      "grad_norm": 18.5,
+      "learning_rate": 2.1073593004763134e-06,
+      "loss": 1.0385,
+      "step": 2208
+    },
+    {
+      "epoch": 0.537077558959397,
+      "grad_norm": 24.375,
+      "learning_rate": 2.1070075626179686e-06,
+      "loss": 0.8896,
+      "step": 2209
+    },
+    {
+      "epoch": 0.537320690493557,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.1066556966650427e-06,
+      "loss": 0.6847,
+      "step": 2210
+    },
+    {
+      "epoch": 0.537563822027717,
+      "grad_norm": 18.25,
+      "learning_rate": 2.1063037026701277e-06,
+      "loss": 0.8065,
+      "step": 2211
+    },
+    {
+      "epoch": 0.537806953561877,
+      "grad_norm": 24.75,
+      "learning_rate": 2.1059515806858357e-06,
+      "loss": 0.9747,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5380500850960369,
+      "grad_norm": 22.75,
+      "learning_rate": 2.105599330764797e-06,
+      "loss": 1.2146,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5382932166301969,
+      "grad_norm": 21.0,
+      "learning_rate": 2.105246952959662e-06,
+      "loss": 0.8581,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5385363481643569,
+      "grad_norm": 24.5,
+      "learning_rate": 2.104894447323099e-06,
+      "loss": 1.2318,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5387794796985169,
+      "grad_norm": 22.875,
+      "learning_rate": 2.104541813907796e-06,
+      "loss": 1.3717,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5390226112326769,
+      "grad_norm": 16.375,
+      "learning_rate": 2.104189052766461e-06,
+      "loss": 0.9492,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5392657427668368,
+      "grad_norm": 26.75,
+      "learning_rate": 2.1038361639518195e-06,
+      "loss": 0.9797,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5395088743009968,
+      "grad_norm": 35.5,
+      "learning_rate": 2.1034831475166166e-06,
+      "loss": 0.8712,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5397520058351568,
+      "grad_norm": 25.125,
+      "learning_rate": 2.103130003513618e-06,
+      "loss": 1.3308,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5399951373693168,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.1027767319956055e-06,
+      "loss": 0.3165,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5402382689034768,
+      "grad_norm": 20.375,
+      "learning_rate": 2.1024233330153828e-06,
+      "loss": 0.7538,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5404814004376368,
+      "grad_norm": 20.25,
+      "learning_rate": 2.1020698066257707e-06,
+      "loss": 0.9751,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5407245319717967,
+      "grad_norm": 17.875,
+      "learning_rate": 2.101716152879611e-06,
+      "loss": 0.8774,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5409676635059567,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1013623718297623e-06,
+      "loss": 0.8974,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5412107950401167,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1010084635291036e-06,
+      "loss": 0.8043,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5414539265742767,
+      "grad_norm": 21.75,
+      "learning_rate": 2.1006544280305325e-06,
+      "loss": 0.8488,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5416970581084367,
+      "grad_norm": 22.0,
+      "learning_rate": 2.1003002653869658e-06,
+      "loss": 0.9424,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5419401896425966,
+      "grad_norm": 20.75,
+      "learning_rate": 2.099945975651339e-06,
+      "loss": 0.9605,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5421833211767566,
+      "grad_norm": 24.375,
+      "learning_rate": 2.0995915588766074e-06,
+      "loss": 1.0632,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5424264527109166,
+      "grad_norm": 25.0,
+      "learning_rate": 2.0992370151157444e-06,
+      "loss": 1.0454,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5426695842450766,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0988823444217426e-06,
+      "loss": 0.8084,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5429127157792366,
+      "grad_norm": 17.75,
+      "learning_rate": 2.0985275468476137e-06,
+      "loss": 0.9918,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5431558473133965,
+      "grad_norm": 19.5,
+      "learning_rate": 2.098172622446388e-06,
+      "loss": 0.9706,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5433989788475565,
+      "grad_norm": 23.625,
+      "learning_rate": 2.097817571271116e-06,
+      "loss": 0.8768,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5436421103817165,
+      "grad_norm": 18.75,
+      "learning_rate": 2.0974623933748655e-06,
+      "loss": 0.9007,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5438852419158765,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.097107088810724e-06,
+      "loss": 0.5056,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5441283734500365,
+      "grad_norm": 17.25,
+      "learning_rate": 2.096751657631798e-06,
+      "loss": 1.0783,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5443715049841964,
+      "grad_norm": 22.875,
+      "learning_rate": 2.0963960998912132e-06,
+      "loss": 0.9539,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5446146365183564,
+      "grad_norm": 20.375,
+      "learning_rate": 2.0960404156421133e-06,
+      "loss": 0.9082,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5448577680525164,
+      "grad_norm": 23.375,
+      "learning_rate": 2.095684604937662e-06,
+      "loss": 1.0703,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5451008995866764,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.09532866783104e-06,
+      "loss": 0.6463,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5453440311208364,
+      "grad_norm": 18.5,
+      "learning_rate": 2.09497260437545e-06,
+      "loss": 0.8202,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5455871626549964,
+      "grad_norm": 21.0,
+      "learning_rate": 2.094616414624111e-06,
+      "loss": 0.8521,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5458302941891563,
+      "grad_norm": 33.0,
+      "learning_rate": 2.094260098630262e-06,
+      "loss": 1.0215,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5460734257233163,
+      "grad_norm": 22.375,
+      "learning_rate": 2.09390365644716e-06,
+      "loss": 1.2234,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5463165572574763,
+      "grad_norm": 20.5,
+      "learning_rate": 2.093547088128082e-06,
+      "loss": 0.7184,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5465596887916363,
+      "grad_norm": 34.25,
+      "learning_rate": 2.093190393726323e-06,
+      "loss": 1.2451,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5468028203257963,
+      "grad_norm": 14.375,
+      "learning_rate": 2.0928335732951976e-06,
+      "loss": 0.5257,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5470459518599562,
+      "grad_norm": 21.125,
+      "learning_rate": 2.0924766268880382e-06,
+      "loss": 0.7474,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5472890833941162,
+      "grad_norm": 34.0,
+      "learning_rate": 2.0921195545581967e-06,
+      "loss": 0.8276,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5475322149282762,
+      "grad_norm": 31.625,
+      "learning_rate": 2.091762356359044e-06,
+      "loss": 0.9416,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5477753464624362,
+      "grad_norm": 21.75,
+      "learning_rate": 2.0914050323439695e-06,
+      "loss": 1.4075,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5480184779965962,
+      "grad_norm": 24.5,
+      "learning_rate": 2.0910475825663813e-06,
+      "loss": 1.0821,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5482616095307561,
+      "grad_norm": 15.0,
+      "learning_rate": 2.0906900070797067e-06,
+      "loss": 0.6176,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5485047410649161,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.090332305937391e-06,
+      "loss": 0.4484,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5487478725990761,
+      "grad_norm": 24.375,
+      "learning_rate": 2.089974479192899e-06,
+      "loss": 1.1816,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5489910041332361,
+      "grad_norm": 15.4375,
+      "learning_rate": 2.0896165268997145e-06,
+      "loss": 1.0037,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5492341356673961,
+      "grad_norm": 16.25,
+      "learning_rate": 2.089258449111339e-06,
+      "loss": 0.5961,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5494772672015561,
+      "grad_norm": 27.75,
+      "learning_rate": 2.088900245881294e-06,
+      "loss": 0.9547,
+      "step": 2260
+    },
+    {
+      "epoch": 0.549720398735716,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0885419172631192e-06,
+      "loss": 1.0561,
+      "step": 2261
+    },
+    {
+      "epoch": 0.549963530269876,
+      "grad_norm": 20.0,
+      "learning_rate": 2.088183463310373e-06,
+      "loss": 0.9799,
+      "step": 2262
+    },
+    {
+      "epoch": 0.550206661804036,
+      "grad_norm": 15.625,
+      "learning_rate": 2.0878248840766317e-06,
+      "loss": 0.5904,
+      "step": 2263
+    },
+    {
+      "epoch": 0.550449793338196,
+      "grad_norm": 24.75,
+      "learning_rate": 2.0874661796154916e-06,
+      "loss": 0.7311,
+      "step": 2264
+    },
+    {
+      "epoch": 0.550692924872356,
+      "grad_norm": 18.5,
+      "learning_rate": 2.087107349980568e-06,
+      "loss": 0.7602,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5509360564065159,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.086748395225493e-06,
+      "loss": 0.6092,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5511791879406759,
+      "grad_norm": 18.625,
+      "learning_rate": 2.08638931540392e-06,
+      "loss": 0.7429,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5514223194748359,
+      "grad_norm": 32.5,
+      "learning_rate": 2.086030110569518e-06,
+      "loss": 0.9813,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5516654510089959,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.0856707807759776e-06,
+      "loss": 0.6213,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5519085825431559,
+      "grad_norm": 21.5,
+      "learning_rate": 2.0853113260770063e-06,
+      "loss": 1.0692,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5521517140773158,
+      "grad_norm": 23.0,
+      "learning_rate": 2.0849517465263315e-06,
+      "loss": 1.0135,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5523948456114758,
+      "grad_norm": 18.0,
+      "learning_rate": 2.084592042177698e-06,
+      "loss": 1.0545,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5526379771456358,
+      "grad_norm": 12.3125,
+      "learning_rate": 2.084232213084869e-06,
+      "loss": 0.5815,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5528811086797958,
+      "grad_norm": 21.375,
+      "learning_rate": 2.0838722593016288e-06,
+      "loss": 0.8707,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5531242402139558,
+      "grad_norm": 25.625,
+      "learning_rate": 2.083512180881778e-06,
+      "loss": 0.76,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5533673717481157,
+      "grad_norm": 14.25,
+      "learning_rate": 2.0831519778791363e-06,
+      "loss": 0.4806,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5536105032822757,
+      "grad_norm": 13.75,
+      "learning_rate": 2.0827916503475425e-06,
+      "loss": 0.5682,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5538536348164357,
+      "grad_norm": 26.875,
+      "learning_rate": 2.082431198340854e-06,
+      "loss": 0.9897,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5540967663505957,
+      "grad_norm": 18.75,
+      "learning_rate": 2.082070621912946e-06,
+      "loss": 1.3372,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5543398978847557,
+      "grad_norm": 24.0,
+      "learning_rate": 2.0817099211177137e-06,
+      "loss": 1.2054,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5545830294189157,
+      "grad_norm": 26.5,
+      "learning_rate": 2.081349096009069e-06,
+      "loss": 0.9381,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5548261609530756,
+      "grad_norm": 22.375,
+      "learning_rate": 2.0809881466409444e-06,
+      "loss": 0.9875,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5550692924872356,
+      "grad_norm": 23.375,
+      "learning_rate": 2.080627073067289e-06,
+      "loss": 0.9427,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5553124240213956,
+      "grad_norm": 18.75,
+      "learning_rate": 2.0802658753420728e-06,
+      "loss": 0.8702,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 17.5,
+      "learning_rate": 2.0799045535192817e-06,
+      "loss": 0.5937,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5557986870897156,
+      "grad_norm": 21.75,
+      "learning_rate": 2.0795431076529226e-06,
+      "loss": 1.0784,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5560418186238755,
+      "grad_norm": 21.25,
+      "learning_rate": 2.0791815377970197e-06,
+      "loss": 1.1514,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5562849501580355,
+      "grad_norm": 16.25,
+      "learning_rate": 2.078819844005615e-06,
+      "loss": 0.6522,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5565280816921955,
+      "grad_norm": 17.5,
+      "learning_rate": 2.07845802633277e-06,
+      "loss": 0.9068,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5567712132263555,
+      "grad_norm": 21.0,
+      "learning_rate": 2.078096084832566e-06,
+      "loss": 0.8342,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5570143447605155,
+      "grad_norm": 21.5,
+      "learning_rate": 2.0777340195590996e-06,
+      "loss": 0.6393,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5572574762946754,
+      "grad_norm": 21.375,
+      "learning_rate": 2.0773718305664887e-06,
+      "loss": 0.6652,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5575006078288354,
+      "grad_norm": 22.75,
+      "learning_rate": 2.0770095179088688e-06,
+      "loss": 0.7814,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5577437393629954,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0766470816403935e-06,
+      "loss": 0.9092,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5579868708971554,
+      "grad_norm": 21.5,
+      "learning_rate": 2.076284521815235e-06,
+      "loss": 0.9198,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5582300024313154,
+      "grad_norm": 21.125,
+      "learning_rate": 2.075921838487584e-06,
+      "loss": 0.7984,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5584731339654754,
+      "grad_norm": 17.75,
+      "learning_rate": 2.07555903171165e-06,
+      "loss": 0.8968,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5587162654996353,
+      "grad_norm": 22.0,
+      "learning_rate": 2.0751961015416617e-06,
+      "loss": 1.0454,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5589593970337953,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.0748330480318637e-06,
+      "loss": 0.7281,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5592025285679553,
+      "grad_norm": 15.1875,
+      "learning_rate": 2.0744698712365215e-06,
+      "loss": 0.7593,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5594456601021153,
+      "grad_norm": 22.25,
+      "learning_rate": 2.074106571209918e-06,
+      "loss": 0.8927,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5596887916362753,
+      "grad_norm": 16.75,
+      "learning_rate": 2.073743148006354e-06,
+      "loss": 0.9008,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5599319231704352,
+      "grad_norm": 30.25,
+      "learning_rate": 2.073379601680151e-06,
+      "loss": 0.936,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5601750547045952,
+      "grad_norm": 19.25,
+      "learning_rate": 2.0730159322856454e-06,
+      "loss": 0.7952,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5604181862387552,
+      "grad_norm": 17.5,
+      "learning_rate": 2.0726521398771956e-06,
+      "loss": 0.6307,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5606613177729152,
+      "grad_norm": 19.125,
+      "learning_rate": 2.0722882245091753e-06,
+      "loss": 0.8667,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5609044493070752,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.0719241862359786e-06,
+      "loss": 0.623,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5611475808412351,
+      "grad_norm": 21.125,
+      "learning_rate": 2.0715600251120167e-06,
+      "loss": 1.173,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5613907123753951,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0711957411917207e-06,
+      "loss": 0.8189,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5616338439095551,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0708313345295384e-06,
+      "loss": 0.6672,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5618769754437151,
+      "grad_norm": 19.125,
+      "learning_rate": 2.070466805179937e-06,
+      "loss": 0.7677,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5621201069778751,
+      "grad_norm": 18.625,
+      "learning_rate": 2.0701021531974014e-06,
+      "loss": 0.9986,
+      "step": 2312
+    },
+    {
+      "epoch": 0.562363238512035,
+      "grad_norm": 26.0,
+      "learning_rate": 2.0697373786364357e-06,
+      "loss": 0.9489,
+      "step": 2313
+    },
+    {
+      "epoch": 0.562606370046195,
+      "grad_norm": 42.75,
+      "learning_rate": 2.0693724815515612e-06,
+      "loss": 1.7043,
+      "step": 2314
+    },
+    {
+      "epoch": 0.562849501580355,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0690074619973185e-06,
+      "loss": 0.7949,
+      "step": 2315
+    },
+    {
+      "epoch": 0.563092633114515,
+      "grad_norm": 19.875,
+      "learning_rate": 2.0686423200282652e-06,
+      "loss": 1.1615,
+      "step": 2316
+    },
+    {
+      "epoch": 0.563335764648675,
+      "grad_norm": 16.875,
+      "learning_rate": 2.0682770556989797e-06,
+      "loss": 0.8947,
+      "step": 2317
+    },
+    {
+      "epoch": 0.563578896182835,
+      "grad_norm": 15.375,
+      "learning_rate": 2.0679116690640556e-06,
+      "loss": 0.5107,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5638220277169949,
+      "grad_norm": 17.75,
+      "learning_rate": 2.0675461601781067e-06,
+      "loss": 0.8202,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5640651592511549,
+      "grad_norm": 20.75,
+      "learning_rate": 2.0671805290957646e-06,
+      "loss": 1.0881,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5643082907853149,
+      "grad_norm": 19.125,
+      "learning_rate": 2.0668147758716792e-06,
+      "loss": 0.9051,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5645514223194749,
+      "grad_norm": 28.375,
+      "learning_rate": 2.0664489005605187e-06,
+      "loss": 0.9758,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5647945538536349,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0660829032169695e-06,
+      "loss": 0.8816,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5650376853877948,
+      "grad_norm": 17.875,
+      "learning_rate": 2.0657167838957365e-06,
+      "loss": 0.7318,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5652808169219548,
+      "grad_norm": 17.5,
+      "learning_rate": 2.065350542651542e-06,
+      "loss": 0.6686,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5655239484561148,
+      "grad_norm": 12.125,
+      "learning_rate": 2.064984179539127e-06,
+      "loss": 0.3821,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5657670799902748,
+      "grad_norm": 22.375,
+      "learning_rate": 2.064617694613251e-06,
+      "loss": 1.1379,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5660102115244348,
+      "grad_norm": 17.75,
+      "learning_rate": 2.0642510879286924e-06,
+      "loss": 0.8428,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5662533430585946,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0638843595402456e-06,
+      "loss": 1.07,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5664964745927547,
+      "grad_norm": 27.625,
+      "learning_rate": 2.063517509502725e-06,
+      "loss": 1.1483,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5667396061269147,
+      "grad_norm": 20.25,
+      "learning_rate": 2.063150537870963e-06,
+      "loss": 0.9096,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5669827376610747,
+      "grad_norm": 19.25,
+      "learning_rate": 2.062783444699809e-06,
+      "loss": 0.633,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5672258691952347,
+      "grad_norm": 30.875,
+      "learning_rate": 2.0624162300441327e-06,
+      "loss": 1.0003,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5674690007293947,
+      "grad_norm": 30.375,
+      "learning_rate": 2.062048893958819e-06,
+      "loss": 1.4803,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5677121322635545,
+      "grad_norm": 30.75,
+      "learning_rate": 2.0616814364987738e-06,
+      "loss": 1.3023,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5679552637977145,
+      "grad_norm": 23.375,
+      "learning_rate": 2.0613138577189203e-06,
+      "loss": 0.8766,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5681983953318746,
+      "grad_norm": 21.0,
+      "learning_rate": 2.060946157674198e-06,
+      "loss": 0.7785,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5684415268660346,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0605783364195676e-06,
+      "loss": 0.9711,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5686846584001946,
+      "grad_norm": 17.625,
+      "learning_rate": 2.060210394010005e-06,
+      "loss": 0.9325,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5689277899343544,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0598423305005065e-06,
+      "loss": 0.8241,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5691709214685144,
+      "grad_norm": 17.875,
+      "learning_rate": 2.059474145946086e-06,
+      "loss": 0.7612,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5694140530026744,
+      "grad_norm": 16.0,
+      "learning_rate": 2.0591058404017735e-06,
+      "loss": 0.7962,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5696571845368344,
+      "grad_norm": 18.75,
+      "learning_rate": 2.058737413922619e-06,
+      "loss": 1.0329,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5699003160709945,
+      "grad_norm": 30.375,
+      "learning_rate": 2.0583688665636915e-06,
+      "loss": 1.0044,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5701434476051543,
+      "grad_norm": 21.75,
+      "learning_rate": 2.0580001983800752e-06,
+      "loss": 1.1053,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5703865791393143,
+      "grad_norm": 18.5,
+      "learning_rate": 2.0576314094268753e-06,
+      "loss": 1.1189,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5706297106734743,
+      "grad_norm": 16.375,
+      "learning_rate": 2.0572624997592124e-06,
+      "loss": 0.6924,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5708728422076343,
+      "grad_norm": 26.125,
+      "learning_rate": 2.0568934694322274e-06,
+      "loss": 1.156,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5711159737417943,
+      "grad_norm": 16.625,
+      "learning_rate": 2.0565243185010776e-06,
+      "loss": 0.633,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5713591052759542,
+      "grad_norm": 26.875,
+      "learning_rate": 2.056155047020939e-06,
+      "loss": 0.922,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5716022368101142,
+      "grad_norm": 19.875,
+      "learning_rate": 2.055785655047006e-06,
+      "loss": 0.7393,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5718453683442742,
+      "grad_norm": 16.5,
+      "learning_rate": 2.055416142634491e-06,
+      "loss": 0.7964,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5720884998784342,
+      "grad_norm": 20.5,
+      "learning_rate": 2.055046509838623e-06,
+      "loss": 0.9522,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5723316314125942,
+      "grad_norm": 14.625,
+      "learning_rate": 2.05467675671465e-06,
+      "loss": 0.7326,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5725747629467542,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0543068833178394e-06,
+      "loss": 0.659,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5728178944809141,
+      "grad_norm": 20.125,
+      "learning_rate": 2.053936889703474e-06,
+      "loss": 0.7541,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5730610260150741,
+      "grad_norm": 16.625,
+      "learning_rate": 2.0535667759268556e-06,
+      "loss": 0.9131,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5733041575492341,
+      "grad_norm": 29.125,
+      "learning_rate": 2.0531965420433046e-06,
+      "loss": 1.2867,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5735472890833941,
+      "grad_norm": 21.25,
+      "learning_rate": 2.0528261881081587e-06,
+      "loss": 1.1126,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5737904206175541,
+      "grad_norm": 17.375,
+      "learning_rate": 2.052455714176774e-06,
+      "loss": 0.9647,
+      "step": 2360
+    },
+    {
+      "epoch": 0.574033552151714,
+      "grad_norm": 25.375,
+      "learning_rate": 2.0520851203045236e-06,
+      "loss": 0.9846,
+      "step": 2361
+    },
+    {
+      "epoch": 0.574276683685874,
+      "grad_norm": 20.5,
+      "learning_rate": 2.0517144065467993e-06,
+      "loss": 0.9121,
+      "step": 2362
+    },
+    {
+      "epoch": 0.574519815220034,
+      "grad_norm": 19.25,
+      "learning_rate": 2.0513435729590114e-06,
+      "loss": 0.7062,
+      "step": 2363
+    },
+    {
+      "epoch": 0.574762946754194,
+      "grad_norm": 20.5,
+      "learning_rate": 2.0509726195965863e-06,
+      "loss": 0.8853,
+      "step": 2364
+    },
+    {
+      "epoch": 0.575006078288354,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0506015465149704e-06,
+      "loss": 0.7734,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5752492098225139,
+      "grad_norm": 17.5,
+      "learning_rate": 2.050230353769626e-06,
+      "loss": 0.7197,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5754923413566739,
+      "grad_norm": 20.125,
+      "learning_rate": 2.049859041416035e-06,
+      "loss": 0.8364,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5757354728908339,
+      "grad_norm": 14.5,
+      "learning_rate": 2.0494876095096964e-06,
+      "loss": 0.7339,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5759786044249939,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.049116058106127e-06,
+      "loss": 0.757,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5762217359591539,
+      "grad_norm": 21.75,
+      "learning_rate": 2.0487443872608613e-06,
+      "loss": 1.1455,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5764648674933139,
+      "grad_norm": 21.5,
+      "learning_rate": 2.048372597029452e-06,
+      "loss": 0.7505,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5767079990274738,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.04800068746747e-06,
+      "loss": 0.6198,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5769511305616338,
+      "grad_norm": 18.0,
+      "learning_rate": 2.047628658630503e-06,
+      "loss": 0.9446,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5771942620957938,
+      "grad_norm": 15.5,
+      "learning_rate": 2.0472565105741578e-06,
+      "loss": 1.1734,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5774373936299538,
+      "grad_norm": 19.625,
+      "learning_rate": 2.0468842433540576e-06,
+      "loss": 1.1002,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5776805251641138,
+      "grad_norm": 16.625,
+      "learning_rate": 2.046511857025845e-06,
+      "loss": 1.0567,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5779236566982737,
+      "grad_norm": 17.875,
+      "learning_rate": 2.0461393516451785e-06,
+      "loss": 0.7922,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5781667882324337,
+      "grad_norm": 24.375,
+      "learning_rate": 2.0457667272677366e-06,
+      "loss": 1.0617,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5784099197665937,
+      "grad_norm": 24.375,
+      "learning_rate": 2.0453939839492137e-06,
+      "loss": 0.9944,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5786530513007537,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0450211217453235e-06,
+      "loss": 0.962,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5788961828349137,
+      "grad_norm": 15.5,
+      "learning_rate": 2.0446481407117953e-06,
+      "loss": 0.5756,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5791393143690736,
+      "grad_norm": 17.75,
+      "learning_rate": 2.044275040904379e-06,
+      "loss": 1.2556,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5793824459032336,
+      "grad_norm": 22.875,
+      "learning_rate": 2.0439018223788402e-06,
+      "loss": 0.9041,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5796255774373936,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.043528485190963e-06,
+      "loss": 0.414,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5798687089715536,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0431550293965486e-06,
+      "loss": 0.8813,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5801118405057136,
+      "grad_norm": 21.75,
+      "learning_rate": 2.042781455051417e-06,
+      "loss": 0.7461,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5803549720398735,
+      "grad_norm": 19.75,
+      "learning_rate": 2.042407762211405e-06,
+      "loss": 1.0001,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5805981035740335,
+      "grad_norm": 19.125,
+      "learning_rate": 2.042033950932368e-06,
+      "loss": 1.0402,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5808412351081935,
+      "grad_norm": 20.125,
+      "learning_rate": 2.0416600212701777e-06,
+      "loss": 1.1356,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5810843666423535,
+      "grad_norm": 21.125,
+      "learning_rate": 2.041285973280725e-06,
+      "loss": 0.9779,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5813274981765135,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.0409118070199175e-06,
+      "loss": 0.5775,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5815706297106735,
+      "grad_norm": 19.375,
+      "learning_rate": 2.040537522543681e-06,
+      "loss": 0.9431,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5818137612448334,
+      "grad_norm": 18.625,
+      "learning_rate": 2.0401631199079584e-06,
+      "loss": 0.9297,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5820568927789934,
+      "grad_norm": 25.0,
+      "learning_rate": 2.039788599168711e-06,
+      "loss": 1.3637,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5823000243131534,
+      "grad_norm": 23.75,
+      "learning_rate": 2.0394139603819176e-06,
+      "loss": 1.4877,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5825431558473134,
+      "grad_norm": 24.75,
+      "learning_rate": 2.039039203603574e-06,
+      "loss": 1.1025,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5827862873814734,
+      "grad_norm": 23.625,
+      "learning_rate": 2.0386643288896944e-06,
+      "loss": 1.1648,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5830294189156333,
+      "grad_norm": 20.375,
+      "learning_rate": 2.0382893362963102e-06,
+      "loss": 0.6469,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5832725504497933,
+      "grad_norm": 20.375,
+      "learning_rate": 2.0379142258794703e-06,
+      "loss": 0.8905,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5835156819839533,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.0375389976952416e-06,
+      "loss": 0.4437,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5837588135181133,
+      "grad_norm": 19.625,
+      "learning_rate": 2.0371636517997085e-06,
+      "loss": 1.5071,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5840019450522733,
+      "grad_norm": 18.5,
+      "learning_rate": 2.0367881882489727e-06,
+      "loss": 0.8795,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5842450765864332,
+      "grad_norm": 21.5,
+      "learning_rate": 2.0364126070991543e-06,
+      "loss": 0.9467,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5844882081205932,
+      "grad_norm": 19.75,
+      "learning_rate": 2.036036908406389e-06,
+      "loss": 1.3867,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5847313396547532,
+      "grad_norm": 20.125,
+      "learning_rate": 2.0356610922268335e-06,
+      "loss": 0.7784,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5849744711889132,
+      "grad_norm": 23.125,
+      "learning_rate": 2.035285158616658e-06,
+      "loss": 0.912,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5852176027230732,
+      "grad_norm": 26.125,
+      "learning_rate": 2.034909107632054e-06,
+      "loss": 1.0614,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5854607342572332,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0345329393292272e-06,
+      "loss": 0.643,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5857038657913931,
+      "grad_norm": 19.625,
+      "learning_rate": 2.034156653764404e-06,
+      "loss": 0.7394,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5859469973255531,
+      "grad_norm": 18.625,
+      "learning_rate": 2.033780250993826e-06,
+      "loss": 0.7713,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5861901288597131,
+      "grad_norm": 21.5,
+      "learning_rate": 2.033403731073753e-06,
+      "loss": 1.119,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5864332603938731,
+      "grad_norm": 19.75,
+      "learning_rate": 2.033027094060462e-06,
+      "loss": 1.2968,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5866763919280331,
+      "grad_norm": 15.5,
+      "learning_rate": 2.0326503400102494e-06,
+      "loss": 0.6065,
+      "step": 2413
+    },
+    {
+      "epoch": 0.586919523462193,
+      "grad_norm": 18.0,
+      "learning_rate": 2.0322734689794262e-06,
+      "loss": 0.7435,
+      "step": 2414
+    },
+    {
+      "epoch": 0.587162654996353,
+      "grad_norm": 14.75,
+      "learning_rate": 2.0318964810243224e-06,
+      "loss": 0.4709,
+      "step": 2415
+    },
+    {
+      "epoch": 0.587405786530513,
+      "grad_norm": 27.125,
+      "learning_rate": 2.031519376201286e-06,
+      "loss": 0.9531,
+      "step": 2416
+    },
+    {
+      "epoch": 0.587648918064673,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0311421545666817e-06,
+      "loss": 0.7211,
+      "step": 2417
+    },
+    {
+      "epoch": 0.587892049598833,
+      "grad_norm": 17.625,
+      "learning_rate": 2.0307648161768914e-06,
+      "loss": 0.83,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5881351811329929,
+      "grad_norm": 17.25,
+      "learning_rate": 2.030387361088315e-06,
+      "loss": 0.8202,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5883783126671529,
+      "grad_norm": 13.625,
+      "learning_rate": 2.0300097893573694e-06,
+      "loss": 0.3608,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5886214442013129,
+      "grad_norm": 20.375,
+      "learning_rate": 2.02963210104049e-06,
+      "loss": 0.7299,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5888645757354729,
+      "grad_norm": 17.125,
+      "learning_rate": 2.0292542961941285e-06,
+      "loss": 0.8545,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5891077072696329,
+      "grad_norm": 22.0,
+      "learning_rate": 2.028876374874754e-06,
+      "loss": 1.1379,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5893508388037928,
+      "grad_norm": 18.75,
+      "learning_rate": 2.028498337138853e-06,
+      "loss": 1.0685,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5895939703379528,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.0281201830429316e-06,
+      "loss": 0.9339,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5898371018721128,
+      "grad_norm": 17.75,
+      "learning_rate": 2.02774191264351e-06,
+      "loss": 1.0232,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5900802334062728,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0273635259971268e-06,
+      "loss": 1.0813,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5903233649404328,
+      "grad_norm": 23.625,
+      "learning_rate": 2.0269850231603393e-06,
+      "loss": 1.1019,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5905664964745928,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0266064041897216e-06,
+      "loss": 0.6596,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5908096280087527,
+      "grad_norm": 22.25,
+      "learning_rate": 2.026227669141864e-06,
+      "loss": 1.3176,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5910527595429127,
+      "grad_norm": 19.875,
+      "learning_rate": 2.0258488180733755e-06,
+      "loss": 0.7592,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5912958910770727,
+      "grad_norm": 17.875,
+      "learning_rate": 2.0254698510408815e-06,
+      "loss": 1.0825,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5915390226112327,
+      "grad_norm": 17.0,
+      "learning_rate": 2.0250907681010255e-06,
+      "loss": 1.1475,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5917821541453927,
+      "grad_norm": 17.125,
+      "learning_rate": 2.024711569310468e-06,
+      "loss": 0.8014,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5920252856795526,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0243322547258866e-06,
+      "loss": 1.1512,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5922684172137126,
+      "grad_norm": 23.0,
+      "learning_rate": 2.0239528244039767e-06,
+      "loss": 0.7642,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5925115487478726,
+      "grad_norm": 23.75,
+      "learning_rate": 2.0235732784014507e-06,
+      "loss": 1.2959,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5927546802820326,
+      "grad_norm": 16.875,
+      "learning_rate": 2.0231936167750378e-06,
+      "loss": 0.6246,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5929978118161926,
+      "grad_norm": 24.125,
+      "learning_rate": 2.0228138395814854e-06,
+      "loss": 1.1494,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5932409433503525,
+      "grad_norm": 13.25,
+      "learning_rate": 2.022433946877558e-06,
+      "loss": 0.3006,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5934840748845125,
+      "grad_norm": 37.5,
+      "learning_rate": 2.0220539387200365e-06,
+      "loss": 1.3829,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5937272064186725,
+      "grad_norm": 15.75,
+      "learning_rate": 2.0216738151657208e-06,
+      "loss": 0.7871,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5939703379528325,
+      "grad_norm": 15.25,
+      "learning_rate": 2.0212935762714254e-06,
+      "loss": 0.817,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5942134694869925,
+      "grad_norm": 15.125,
+      "learning_rate": 2.0209132220939845e-06,
+      "loss": 0.5611,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5944566010211525,
+      "grad_norm": 17.375,
+      "learning_rate": 2.0205327526902486e-06,
+      "loss": 0.873,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5946997325553124,
+      "grad_norm": 17.125,
+      "learning_rate": 2.020152168117085e-06,
+      "loss": 1.2222,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5949428640894724,
+      "grad_norm": 21.875,
+      "learning_rate": 2.0197714684313786e-06,
+      "loss": 0.689,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5951859956236324,
+      "grad_norm": 22.5,
+      "learning_rate": 2.019390653690033e-06,
+      "loss": 1.0182,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5954291271577924,
+      "grad_norm": 20.0,
+      "learning_rate": 2.019009723949965e-06,
+      "loss": 1.0079,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5956722586919524,
+      "grad_norm": 23.5,
+      "learning_rate": 2.018628679268113e-06,
+      "loss": 0.9264,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5959153902261123,
+      "grad_norm": 21.625,
+      "learning_rate": 2.0182475197014306e-06,
+      "loss": 0.9782,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5961585217602723,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0178662453068877e-06,
+      "loss": 0.6295,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5964016532944323,
+      "grad_norm": 19.375,
+      "learning_rate": 2.0174848561414734e-06,
+      "loss": 0.6381,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5966447848285923,
+      "grad_norm": 17.625,
+      "learning_rate": 2.017103352262192e-06,
+      "loss": 0.589,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5968879163627523,
+      "grad_norm": 16.875,
+      "learning_rate": 2.0167217337260665e-06,
+      "loss": 0.4894,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5971310478969122,
+      "grad_norm": 17.875,
+      "learning_rate": 2.0163400005901362e-06,
+      "loss": 0.8663,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5973741794310722,
+      "grad_norm": 13.875,
+      "learning_rate": 2.015958152911458e-06,
+      "loss": 0.6678,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5976173109652322,
+      "grad_norm": 16.625,
+      "learning_rate": 2.0155761907471043e-06,
+      "loss": 0.7652,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5978604424993922,
+      "grad_norm": 23.375,
+      "learning_rate": 2.015194114154168e-06,
+      "loss": 0.9417,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5981035740335522,
+      "grad_norm": 18.125,
+      "learning_rate": 2.0148119231897556e-06,
+      "loss": 0.912,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5983467055677121,
+      "grad_norm": 24.0,
+      "learning_rate": 2.0144296179109923e-06,
+      "loss": 0.8892,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5985898371018721,
+      "grad_norm": 26.25,
+      "learning_rate": 2.0140471983750205e-06,
+      "loss": 0.9228,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5988329686360321,
+      "grad_norm": 18.75,
+      "learning_rate": 2.0136646646389996e-06,
+      "loss": 1.0494,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5990761001701921,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.013282016760105e-06,
+      "loss": 0.5649,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5993192317043521,
+      "grad_norm": 39.75,
+      "learning_rate": 2.0128992547955315e-06,
+      "loss": 1.1498,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5995623632385121,
+      "grad_norm": 22.125,
+      "learning_rate": 2.012516378802488e-06,
+      "loss": 1.0061,
+      "step": 2466
+    },
+    {
+      "epoch": 0.599805494772672,
+      "grad_norm": 23.75,
+      "learning_rate": 2.0121333888382032e-06,
+      "loss": 1.245,
+      "step": 2467
+    },
+    {
+      "epoch": 0.600048626306832,
+      "grad_norm": 16.5,
+      "learning_rate": 2.0117502849599204e-06,
+      "loss": 0.6547,
+      "step": 2468
+    },
+    {
+      "epoch": 0.600291757840992,
+      "grad_norm": 19.0,
+      "learning_rate": 2.011367067224902e-06,
+      "loss": 0.6211,
+      "step": 2469
+    },
+    {
+      "epoch": 0.600534889375152,
+      "grad_norm": 19.5,
+      "learning_rate": 2.0109837356904257e-06,
+      "loss": 0.8737,
+      "step": 2470
+    },
+    {
+      "epoch": 0.600778020909312,
+      "grad_norm": 23.375,
+      "learning_rate": 2.0106002904137877e-06,
+      "loss": 0.9821,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6010211524434719,
+      "grad_norm": 18.125,
+      "learning_rate": 2.0102167314523004e-06,
+      "loss": 0.9827,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6012642839776319,
+      "grad_norm": 16.75,
+      "learning_rate": 2.009833058863293e-06,
+      "loss": 0.8096,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6015074155117919,
+      "grad_norm": 24.25,
+      "learning_rate": 2.0094492727041124e-06,
+      "loss": 0.8818,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6017505470459519,
+      "grad_norm": 14.375,
+      "learning_rate": 2.009065373032122e-06,
+      "loss": 0.4699,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6019936785801119,
+      "grad_norm": 14.125,
+      "learning_rate": 2.0086813599047012e-06,
+      "loss": 0.6093,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6022368101142718,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0082972333792496e-06,
+      "loss": 0.6792,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6024799416484318,
+      "grad_norm": 13.5,
+      "learning_rate": 2.007912993513179e-06,
+      "loss": 0.4687,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6027230731825918,
+      "grad_norm": 23.875,
+      "learning_rate": 2.0075286403639226e-06,
+      "loss": 0.792,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6029662047167518,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.0071441739889278e-06,
+      "loss": 0.7047,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6032093362509118,
+      "grad_norm": 25.0,
+      "learning_rate": 2.0067595944456598e-06,
+      "loss": 1.1387,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6034524677850718,
+      "grad_norm": 37.25,
+      "learning_rate": 2.006374901791601e-06,
+      "loss": 1.6589,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6036955993192317,
+      "grad_norm": 16.75,
+      "learning_rate": 2.0059900960842493e-06,
+      "loss": 0.6608,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6039387308533917,
+      "grad_norm": 20.625,
+      "learning_rate": 2.005605177381122e-06,
+      "loss": 0.7116,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6041818623875517,
+      "grad_norm": 32.5,
+      "learning_rate": 2.0052201457397507e-06,
+      "loss": 1.7751,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6044249939217117,
+      "grad_norm": 25.125,
+      "learning_rate": 2.004835001217686e-06,
+      "loss": 1.0185,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6046681254558717,
+      "grad_norm": 15.75,
+      "learning_rate": 2.004449743872494e-06,
+      "loss": 0.5704,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6049112569900316,
+      "grad_norm": 19.375,
+      "learning_rate": 2.0040643737617577e-06,
+      "loss": 0.7855,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6051543885241916,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.0036788909430774e-06,
+      "loss": 0.4763,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6053975200583516,
+      "grad_norm": 20.75,
+      "learning_rate": 2.0032932954740707e-06,
+      "loss": 0.9844,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6056406515925116,
+      "grad_norm": 17.375,
+      "learning_rate": 2.002907587412371e-06,
+      "loss": 0.7559,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6058837831266716,
+      "grad_norm": 20.25,
+      "learning_rate": 2.0025217668156295e-06,
+      "loss": 0.9571,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6061269146608315,
+      "grad_norm": 16.5,
+      "learning_rate": 2.002135833741513e-06,
+      "loss": 0.667,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6063700461949915,
+      "grad_norm": 23.125,
+      "learning_rate": 2.0017497882477068e-06,
+      "loss": 1.0873,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6066131777291515,
+      "grad_norm": 21.375,
+      "learning_rate": 2.001363630391911e-06,
+      "loss": 0.8989,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6068563092633115,
+      "grad_norm": 28.5,
+      "learning_rate": 2.0009773602318444e-06,
+      "loss": 0.8248,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6070994407974715,
+      "grad_norm": 24.125,
+      "learning_rate": 2.0005909778252415e-06,
+      "loss": 0.9701,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6073425723316314,
+      "grad_norm": 15.5,
+      "learning_rate": 2.000204483229854e-06,
+      "loss": 0.7341,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6075857038657914,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9998178765034496e-06,
+      "loss": 0.5449,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6078288353999514,
+      "grad_norm": 19.625,
+      "learning_rate": 1.9994311577038146e-06,
+      "loss": 0.8267,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6080719669341114,
+      "grad_norm": 22.25,
+      "learning_rate": 1.999044326888749e-06,
+      "loss": 0.9458,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6083150984682714,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.9986573841160728e-06,
+      "loss": 0.5654,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6085582300024314,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.9982703294436206e-06,
+      "loss": 0.6877,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6088013615365913,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9978831629292444e-06,
+      "loss": 0.7262,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6090444930707513,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9974958846308136e-06,
+      "loss": 0.7039,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6092876246049113,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9971084946062126e-06,
+      "loss": 0.7864,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6095307561390713,
+      "grad_norm": 20.375,
+      "learning_rate": 1.996720992913345e-06,
+      "loss": 0.7615,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6097738876732313,
+      "grad_norm": 19.625,
+      "learning_rate": 1.9963333796101275e-06,
+      "loss": 0.6926,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6100170192073912,
+      "grad_norm": 18.375,
+      "learning_rate": 1.995945654754497e-06,
+      "loss": 0.777,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6102601507415512,
+      "grad_norm": 27.875,
+      "learning_rate": 1.9955578184044062e-06,
+      "loss": 1.2121,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6105032822757112,
+      "grad_norm": 18.375,
+      "learning_rate": 1.995169870617823e-06,
+      "loss": 0.7759,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6107464138098712,
+      "grad_norm": 19.75,
+      "learning_rate": 1.994781811452733e-06,
+      "loss": 0.672,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6109895453440312,
+      "grad_norm": 23.25,
+      "learning_rate": 1.994393640967138e-06,
+      "loss": 1.1134,
+      "step": 2513
+    },
+    {
+      "epoch": 0.611232676878191,
+      "grad_norm": 26.5,
+      "learning_rate": 1.994005359219058e-06,
+      "loss": 0.9399,
+      "step": 2514
+    },
+    {
+      "epoch": 0.611475808412351,
+      "grad_norm": 23.125,
+      "learning_rate": 1.993616966266527e-06,
+      "loss": 1.2102,
+      "step": 2515
+    },
+    {
+      "epoch": 0.611718939946511,
+      "grad_norm": 16.75,
+      "learning_rate": 1.993228462167598e-06,
+      "loss": 0.6831,
+      "step": 2516
+    },
+    {
+      "epoch": 0.611962071480671,
+      "grad_norm": 16.875,
+      "learning_rate": 1.992839846980339e-06,
+      "loss": 0.5729,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6122052030148311,
+      "grad_norm": 16.625,
+      "learning_rate": 1.992451120762836e-06,
+      "loss": 0.8248,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6124483345489911,
+      "grad_norm": 22.0,
+      "learning_rate": 1.99206228357319e-06,
+      "loss": 1.1109,
+      "step": 2519
+    },
+    {
+      "epoch": 0.612691466083151,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9916733354695204e-06,
+      "loss": 0.9915,
+      "step": 2520
+    },
+    {
+      "epoch": 0.612934597617311,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9912842765099617e-06,
+      "loss": 0.7568,
+      "step": 2521
+    },
+    {
+      "epoch": 0.613177729151471,
+      "grad_norm": 25.875,
+      "learning_rate": 1.990895106752665e-06,
+      "loss": 1.5433,
+      "step": 2522
+    },
+    {
+      "epoch": 0.613420860685631,
+      "grad_norm": 25.125,
+      "learning_rate": 1.9905058262557993e-06,
+      "loss": 0.8675,
+      "step": 2523
+    },
+    {
+      "epoch": 0.613663992219791,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9901164350775482e-06,
+      "loss": 0.8637,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6139071237539508,
+      "grad_norm": 22.625,
+      "learning_rate": 1.9897269332761145e-06,
+      "loss": 0.7131,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6141502552881108,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9893373209097142e-06,
+      "loss": 0.7901,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6143933868222708,
+      "grad_norm": 17.5,
+      "learning_rate": 1.988947598036583e-06,
+      "loss": 0.7823,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6146365183564308,
+      "grad_norm": 17.875,
+      "learning_rate": 1.988557764714971e-06,
+      "loss": 0.8806,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6148796498905909,
+      "grad_norm": 20.5,
+      "learning_rate": 1.9881678210031462e-06,
+      "loss": 0.7563,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6151227814247507,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9877777669593917e-06,
+      "loss": 0.9966,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6153659129589107,
+      "grad_norm": 19.625,
+      "learning_rate": 1.987387602642008e-06,
+      "loss": 0.6233,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6156090444930707,
+      "grad_norm": 18.25,
+      "learning_rate": 1.986997328109312e-06,
+      "loss": 1.0035,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6158521760272307,
+      "grad_norm": 14.0,
+      "learning_rate": 1.9866069434196367e-06,
+      "loss": 0.8368,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6160953075613907,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9862164486313323e-06,
+      "loss": 0.6484,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6163384390955506,
+      "grad_norm": 17.0,
+      "learning_rate": 1.985825843802765e-06,
+      "loss": 0.6463,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6165815706297106,
+      "grad_norm": 19.875,
+      "learning_rate": 1.985435128992317e-06,
+      "loss": 0.7574,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6168247021638706,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9850443042583872e-06,
+      "loss": 0.7423,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6170678336980306,
+      "grad_norm": 15.125,
+      "learning_rate": 1.984653369659392e-06,
+      "loss": 0.6347,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6173109652321906,
+      "grad_norm": 27.125,
+      "learning_rate": 1.9842623252537624e-06,
+      "loss": 1.0904,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6175540967663506,
+      "grad_norm": 21.375,
+      "learning_rate": 1.983871171099947e-06,
+      "loss": 0.8306,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6177972283005105,
+      "grad_norm": 18.75,
+      "learning_rate": 1.983479907256411e-06,
+      "loss": 1.1409,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6180403598346705,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.983088533781635e-06,
+      "loss": 0.5206,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6182834913688305,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9826970507341173e-06,
+      "loss": 0.9233,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6185266229029905,
+      "grad_norm": 27.125,
+      "learning_rate": 1.982305458172371e-06,
+      "loss": 0.7664,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6187697544371505,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9819137561549265e-06,
+      "loss": 1.3842,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6190128859713104,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9815219447403305e-06,
+      "loss": 0.8527,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6192560175054704,
+      "grad_norm": 19.0,
+      "learning_rate": 1.9811300239871463e-06,
+      "loss": 1.0483,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6194991490396304,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9807379939539527e-06,
+      "loss": 0.7207,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6197422805737904,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9803458546993456e-06,
+      "loss": 0.5849,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6199854121079504,
+      "grad_norm": 19.625,
+      "learning_rate": 1.9799536062819376e-06,
+      "loss": 0.6721,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6202285436421103,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9795612487603553e-06,
+      "loss": 0.7343,
+      "step": 2551
+    },
+    {
+      "epoch": 0.6204716751762703,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9791687821932456e-06,
+      "loss": 0.6469,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6207148067104303,
+      "grad_norm": 13.375,
+      "learning_rate": 1.9787762066392675e-06,
+      "loss": 0.4419,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6209579382445903,
+      "grad_norm": 17.25,
+      "learning_rate": 1.978383522157099e-06,
+      "loss": 0.6656,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6212010697787503,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9779907288054332e-06,
+      "loss": 0.5274,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6214442013129103,
+      "grad_norm": 24.625,
+      "learning_rate": 1.977597826642981e-06,
+      "loss": 1.051,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6216873328470702,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9772048157284666e-06,
+      "loss": 0.9632,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6219304643812302,
+      "grad_norm": 19.0,
+      "learning_rate": 1.976811696120634e-06,
+      "loss": 1.0646,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6221735959153902,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9764184678782406e-06,
+      "loss": 0.711,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6224167274495502,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9760251310600614e-06,
+      "loss": 1.0041,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6226598589837102,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9756316857248877e-06,
+      "loss": 0.7743,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6229029905178701,
+      "grad_norm": 18.625,
+      "learning_rate": 1.9752381319315267e-06,
+      "loss": 0.9294,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6231461220520301,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9748444697388008e-06,
+      "loss": 0.631,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6233892535861901,
+      "grad_norm": 17.875,
+      "learning_rate": 1.974450699205551e-06,
+      "loss": 0.8134,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6236323851203501,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9740568203906325e-06,
+      "loss": 0.4797,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6238755166545101,
+      "grad_norm": 26.0,
+      "learning_rate": 1.973662833352917e-06,
+      "loss": 1.2025,
+      "step": 2566
+    },
+    {
+      "epoch": 0.62411864818867,
+      "grad_norm": 17.25,
+      "learning_rate": 1.9732687381512933e-06,
+      "loss": 0.8941,
+      "step": 2567
+    },
+    {
+      "epoch": 0.62436177972283,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9728745348446654e-06,
+      "loss": 1.0143,
+      "step": 2568
+    },
+    {
+      "epoch": 0.62460491125699,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9724802234919535e-06,
+      "loss": 0.9828,
+      "step": 2569
+    },
+    {
+      "epoch": 0.62484804279115,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9720858041520944e-06,
+      "loss": 0.7606,
+      "step": 2570
+    },
+    {
+      "epoch": 0.62509117432531,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9716912768840417e-06,
+      "loss": 0.7916,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6253343058594699,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9712966417467634e-06,
+      "loss": 0.9992,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6255774373936299,
+      "grad_norm": 13.75,
+      "learning_rate": 1.970901898799244e-06,
+      "loss": 0.3656,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6258205689277899,
+      "grad_norm": 19.0,
+      "learning_rate": 1.9705070481004862e-06,
+      "loss": 0.8741,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6260637004619499,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9701120897095063e-06,
+      "loss": 0.6176,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6263068319961099,
+      "grad_norm": 16.25,
+      "learning_rate": 1.969717023685338e-06,
+      "loss": 0.7102,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6265499635302699,
+      "grad_norm": 25.625,
+      "learning_rate": 1.9693218500870303e-06,
+      "loss": 1.0332,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6267930950644298,
+      "grad_norm": 17.375,
+      "learning_rate": 1.968926568973649e-06,
+      "loss": 0.8218,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6270362265985898,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9685311804042756e-06,
+      "loss": 0.8964,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6272793581327498,
+      "grad_norm": 17.75,
+      "learning_rate": 1.968135684438008e-06,
+      "loss": 1.0285,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6275224896669098,
+      "grad_norm": 25.375,
+      "learning_rate": 1.96774008113396e-06,
+      "loss": 1.0416,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6277656212010698,
+      "grad_norm": 19.75,
+      "learning_rate": 1.9673443705512605e-06,
+      "loss": 0.8864,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6280087527352297,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9669485527490563e-06,
+      "loss": 1.0568,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6282518842693897,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9665526277865084e-06,
+      "loss": 0.6917,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6284950158035497,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9661565957227954e-06,
+      "loss": 1.0235,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6287381473377097,
+      "grad_norm": 20.375,
+      "learning_rate": 1.96576045661711e-06,
+      "loss": 0.9047,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6289812788718697,
+      "grad_norm": 21.375,
+      "learning_rate": 1.9653642105286636e-06,
+      "loss": 1.0859,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6292244104060296,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9649678575166808e-06,
+      "loss": 0.7796,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6294675419401896,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9645713976404036e-06,
+      "loss": 0.9683,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6297106734743496,
+      "grad_norm": 17.0,
+      "learning_rate": 1.96417483095909e-06,
+      "loss": 0.3621,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6299538050085096,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.9637781575320138e-06,
+      "loss": 0.7793,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6301969365426696,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9633813774184646e-06,
+      "loss": 0.94,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6304400680768296,
+      "grad_norm": 19.875,
+      "learning_rate": 1.9629844906777483e-06,
+      "loss": 0.7717,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6306831996109895,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9625874973691856e-06,
+      "loss": 0.6891,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6309263311451495,
+      "grad_norm": 19.375,
+      "learning_rate": 1.962190397552115e-06,
+      "loss": 0.9662,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6311694626793095,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9617931912858897e-06,
+      "loss": 0.7517,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6314125942134695,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.9613958786298783e-06,
+      "loss": 0.551,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6316557257476295,
+      "grad_norm": 25.25,
+      "learning_rate": 1.960998459643467e-06,
+      "loss": 0.9711,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6318988572817894,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9606009343860566e-06,
+      "loss": 1.0415,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6321419888159494,
+      "grad_norm": 18.125,
+      "learning_rate": 1.9602033029170637e-06,
+      "loss": 1.009,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6323851203501094,
+      "grad_norm": 19.75,
+      "learning_rate": 1.959805565295922e-06,
+      "loss": 0.7581,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6326282518842694,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9594077215820795e-06,
+      "loss": 0.6777,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6328713834184294,
+      "grad_norm": 20.75,
+      "learning_rate": 1.959009771835001e-06,
+      "loss": 0.7276,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6331145149525893,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9586117161141672e-06,
+      "loss": 1.1352,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6333576464867493,
+      "grad_norm": 45.0,
+      "learning_rate": 1.958213554479074e-06,
+      "loss": 0.9656,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6336007780209093,
+      "grad_norm": 16.75,
+      "learning_rate": 1.957815286989235e-06,
+      "loss": 0.7441,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6338439095550693,
+      "grad_norm": 15.5,
+      "learning_rate": 1.957416913704176e-06,
+      "loss": 0.5765,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6340870410892293,
+      "grad_norm": 24.75,
+      "learning_rate": 1.9570184346834415e-06,
+      "loss": 1.3351,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6343301726233892,
+      "grad_norm": 24.0,
+      "learning_rate": 1.9566198499865917e-06,
+      "loss": 0.9138,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6345733041575492,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9562211596732012e-06,
+      "loss": 0.5859,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6348164356917092,
+      "grad_norm": 17.5,
+      "learning_rate": 1.955822363802862e-06,
+      "loss": 0.6955,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6350595672258692,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9554234624351807e-06,
+      "loss": 0.9278,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6353026987600292,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9550244556297794e-06,
+      "loss": 0.5837,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6355458302941892,
+      "grad_norm": 17.875,
+      "learning_rate": 1.954625343446297e-06,
+      "loss": 0.9824,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6357889618283491,
+      "grad_norm": 23.625,
+      "learning_rate": 1.954226125944388e-06,
+      "loss": 1.0017,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6360320933625091,
+      "grad_norm": 21.125,
+      "learning_rate": 1.953826803183722e-06,
+      "loss": 1.2419,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6362752248966691,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9534273752239844e-06,
+      "loss": 0.44,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6365183564308291,
+      "grad_norm": 18.75,
+      "learning_rate": 1.953027842124878e-06,
+      "loss": 0.7213,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6367614879649891,
+      "grad_norm": 17.25,
+      "learning_rate": 1.9526282039461177e-06,
+      "loss": 0.7985,
+      "step": 2619
+    },
+    {
+      "epoch": 0.637004619499149,
+      "grad_norm": 20.0,
+      "learning_rate": 1.952228460747438e-06,
+      "loss": 0.8595,
+      "step": 2620
+    },
+    {
+      "epoch": 0.637247751033309,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9518286125885872e-06,
+      "loss": 0.7828,
+      "step": 2621
+    },
+    {
+      "epoch": 0.637490882567469,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9514286595293286e-06,
+      "loss": 0.9267,
+      "step": 2622
+    },
+    {
+      "epoch": 0.637734014101629,
+      "grad_norm": 12.125,
+      "learning_rate": 1.9510286016294432e-06,
+      "loss": 0.544,
+      "step": 2623
+    },
+    {
+      "epoch": 0.637977145635789,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9506284389487256e-06,
+      "loss": 0.8828,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6382202771699489,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9502281715469883e-06,
+      "loss": 0.9487,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6384634087041089,
+      "grad_norm": 21.75,
+      "learning_rate": 1.949827799484057e-06,
+      "loss": 0.8824,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6387065402382689,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9494273228197747e-06,
+      "loss": 1.0268,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6389496717724289,
+      "grad_norm": 20.125,
+      "learning_rate": 1.949026741613999e-06,
+      "loss": 0.6952,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6391928033065889,
+      "grad_norm": 14.875,
+      "learning_rate": 1.948626055926605e-06,
+      "loss": 0.6886,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6394359348407489,
+      "grad_norm": 19.875,
+      "learning_rate": 1.948225265817481e-06,
+      "loss": 0.7787,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6396790663749088,
+      "grad_norm": 24.625,
+      "learning_rate": 1.947824371346532e-06,
+      "loss": 1.0134,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6399221979090688,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9474233725736787e-06,
+      "loss": 0.8336,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6401653294432288,
+      "grad_norm": 18.125,
+      "learning_rate": 1.947022269558858e-06,
+      "loss": 0.9251,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6404084609773888,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9466210623620207e-06,
+      "loss": 1.2374,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6406515925115488,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9462197510431346e-06,
+      "loss": 0.5718,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6408947240457087,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9458183356621826e-06,
+      "loss": 0.7142,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6411378555798687,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9454168162791635e-06,
+      "loss": 0.8723,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6413809871140287,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9450151929540908e-06,
+      "loss": 0.6637,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6416241186481887,
+      "grad_norm": 16.375,
+      "learning_rate": 1.944613465746994e-06,
+      "loss": 0.9261,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6418672501823487,
+      "grad_norm": 26.25,
+      "learning_rate": 1.944211634717918e-06,
+      "loss": 0.985,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6421103817165086,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9438096999269243e-06,
+      "loss": 0.9207,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6423535132506686,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9434076614340883e-06,
+      "loss": 0.7383,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6425966447848286,
+      "grad_norm": 15.875,
+      "learning_rate": 1.9430055192995016e-06,
+      "loss": 0.8852,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6428397763189886,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9426032735832717e-06,
+      "loss": 0.6596,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6430829078531486,
+      "grad_norm": 23.25,
+      "learning_rate": 1.94220092434552e-06,
+      "loss": 1.1477,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6433260393873085,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9417984716463868e-06,
+      "loss": 0.555,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6435691709214685,
+      "grad_norm": 19.0,
+      "learning_rate": 1.941395915546024e-06,
+      "loss": 0.7148,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6438123024556285,
+      "grad_norm": 26.125,
+      "learning_rate": 1.9409932561045995e-06,
+      "loss": 1.0506,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6440554339897885,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9405904933823e-06,
+      "loss": 0.7608,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6442985655239485,
+      "grad_norm": 21.25,
+      "learning_rate": 1.940187627439325e-06,
+      "loss": 1.1644,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6445416970581085,
+      "grad_norm": 21.5,
+      "learning_rate": 1.939784658335888e-06,
+      "loss": 0.9409,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6447848285922684,
+      "grad_norm": 17.375,
+      "learning_rate": 1.939381586132221e-06,
+      "loss": 0.9638,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6450279601264284,
+      "grad_norm": 19.375,
+      "learning_rate": 1.93897841088857e-06,
+      "loss": 1.0011,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6452710916605884,
+      "grad_norm": 19.375,
+      "learning_rate": 1.938575132665197e-06,
+      "loss": 0.9863,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6455142231947484,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9381717515223775e-06,
+      "loss": 0.7573,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6457573547289084,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9377682675204053e-06,
+      "loss": 0.6723,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6460004862630683,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9373646807195867e-06,
+      "loss": 1.2054,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6462436177972283,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9369609911802455e-06,
+      "loss": 1.0758,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6464867493313883,
+      "grad_norm": 17.375,
+      "learning_rate": 1.93655719896272e-06,
+      "loss": 0.9359,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6467298808655483,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9361533041273643e-06,
+      "loss": 0.8533,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6469730123997083,
+      "grad_norm": 17.875,
+      "learning_rate": 1.935749306734547e-06,
+      "loss": 0.9048,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6472161439338682,
+      "grad_norm": 16.875,
+      "learning_rate": 1.935345206844652e-06,
+      "loss": 0.7516,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6474592754680282,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9349410045180796e-06,
+      "loss": 0.6485,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6477024070021882,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.9345366998152448e-06,
+      "loss": 0.7834,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6479455385363482,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9341322927965782e-06,
+      "loss": 1.2497,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6481886700705082,
+      "grad_norm": 17.25,
+      "learning_rate": 1.9337277835225248e-06,
+      "loss": 0.7493,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6484318016046682,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9333231720535456e-06,
+      "loss": 1.232,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6486749331388281,
+      "grad_norm": 21.625,
+      "learning_rate": 1.932918458450117e-06,
+      "loss": 1.0351,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6489180646729881,
+      "grad_norm": 22.5,
+      "learning_rate": 1.9325136427727302e-06,
+      "loss": 0.9951,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6491611962071481,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9321087250818927e-06,
+      "loss": 0.8068,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6494043277413081,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9317037054381255e-06,
+      "loss": 0.5792,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6496474592754681,
+      "grad_norm": 13.75,
+      "learning_rate": 1.931298583901966e-06,
+      "loss": 0.6272,
+      "step": 2672
+    },
+    {
+      "epoch": 0.649890590809628,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9308933605339667e-06,
+      "loss": 0.647,
+      "step": 2673
+    },
+    {
+      "epoch": 0.650133722343788,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9304880353946952e-06,
+      "loss": 0.6139,
+      "step": 2674
+    },
+    {
+      "epoch": 0.650376853877948,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.9300826085447345e-06,
+      "loss": 0.671,
+      "step": 2675
+    },
+    {
+      "epoch": 0.650619985412108,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9296770800446825e-06,
+      "loss": 1.1881,
+      "step": 2676
+    },
+    {
+      "epoch": 0.650863116946268,
+      "grad_norm": 29.75,
+      "learning_rate": 1.9292714499551524e-06,
+      "loss": 0.8084,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6511062484804279,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9288657183367725e-06,
+      "loss": 0.8615,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6513493800145879,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9284598852501867e-06,
+      "loss": 1.3256,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6515925115487479,
+      "grad_norm": 16.375,
+      "learning_rate": 1.928053950756054e-06,
+      "loss": 0.7895,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6518356430829079,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9276479149150475e-06,
+      "loss": 0.5394,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6520787746170679,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9272417777878573e-06,
+      "loss": 0.9726,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6523219061512278,
+      "grad_norm": 25.25,
+      "learning_rate": 1.9268355394351862e-06,
+      "loss": 1.1387,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6525650376853878,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9264291999177547e-06,
+      "loss": 1.2903,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6528081692195478,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9260227592962976e-06,
+      "loss": 0.8315,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6530513007537078,
+      "grad_norm": 20.75,
+      "learning_rate": 1.925616217631563e-06,
+      "loss": 1.2539,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6532944322878678,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9252095749843162e-06,
+      "loss": 0.6728,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6535375638220278,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9248028314153383e-06,
+      "loss": 0.796,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6537806953561877,
+      "grad_norm": 15.625,
+      "learning_rate": 1.9243959869854222e-06,
+      "loss": 0.6722,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6540238268903477,
+      "grad_norm": 15.75,
+      "learning_rate": 1.9239890417553786e-06,
+      "loss": 0.8586,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6542669584245077,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9235819957860323e-06,
+      "loss": 0.9895,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6545100899586677,
+      "grad_norm": 20.875,
+      "learning_rate": 1.923174849138224e-06,
+      "loss": 0.9195,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6547532214928277,
+      "grad_norm": 18.0,
+      "learning_rate": 1.9227676018728087e-06,
+      "loss": 1.1034,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6549963530269876,
+      "grad_norm": 20.75,
+      "learning_rate": 1.922360254050655e-06,
+      "loss": 0.9611,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6552394845611476,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9219528057326507e-06,
+      "loss": 0.5477,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6554826160953076,
+      "grad_norm": 17.625,
+      "learning_rate": 1.921545256979694e-06,
+      "loss": 0.7053,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6557257476294676,
+      "grad_norm": 26.75,
+      "learning_rate": 1.9211376078527003e-06,
+      "loss": 0.9475,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6559688791636276,
+      "grad_norm": 23.75,
+      "learning_rate": 1.9207298584126005e-06,
+      "loss": 0.6847,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6562120106977875,
+      "grad_norm": 16.875,
+      "learning_rate": 1.920322008720339e-06,
+      "loss": 0.5618,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6564551422319475,
+      "grad_norm": 16.625,
+      "learning_rate": 1.919914058836877e-06,
+      "loss": 0.8941,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6566982737661075,
+      "grad_norm": 22.25,
+      "learning_rate": 1.919506008823189e-06,
+      "loss": 0.6823,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6569414053002675,
+      "grad_norm": 19.375,
+      "learning_rate": 1.919097858740265e-06,
+      "loss": 0.9736,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6571845368344275,
+      "grad_norm": 16.75,
+      "learning_rate": 1.91868960864911e-06,
+      "loss": 0.8121,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6574276683685875,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9182812586107454e-06,
+      "loss": 0.6385,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6576707999027473,
+      "grad_norm": 18.625,
+      "learning_rate": 1.917872808686204e-06,
+      "loss": 1.0717,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6579139314369074,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9174642589365372e-06,
+      "loss": 0.6511,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6581570629710674,
+      "grad_norm": 25.0,
+      "learning_rate": 1.9170556094228092e-06,
+      "loss": 0.8808,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6584001945052274,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.9166468602061e-06,
+      "loss": 0.5074,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6586433260393874,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9162380113475045e-06,
+      "loss": 0.6399,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6588864575735472,
+      "grad_norm": 21.875,
+      "learning_rate": 1.9158290629081317e-06,
+      "loss": 1.036,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6591295891077072,
+      "grad_norm": 46.75,
+      "learning_rate": 1.915420014949106e-06,
+      "loss": 1.1031,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6593727206418672,
+      "grad_norm": 18.5,
+      "learning_rate": 1.915010867531567e-06,
+      "loss": 0.5675,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6596158521760273,
+      "grad_norm": 26.25,
+      "learning_rate": 1.9146016207166684e-06,
+      "loss": 0.987,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6598589837101873,
+      "grad_norm": 17.5,
+      "learning_rate": 1.91419227456558e-06,
+      "loss": 0.6809,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6601021152443471,
+      "grad_norm": 19.125,
+      "learning_rate": 1.913782829139485e-06,
+      "loss": 0.8617,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6603452467785071,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.9133732844995824e-06,
+      "loss": 0.5928,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6605883783126671,
+      "grad_norm": 20.0,
+      "learning_rate": 1.912963640707085e-06,
+      "loss": 1.1929,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6608315098468271,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.912553897823222e-06,
+      "loss": 0.5935,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6610746413809871,
+      "grad_norm": 23.25,
+      "learning_rate": 1.912144055909237e-06,
+      "loss": 1.0319,
+      "step": 2719
+    },
+    {
+      "epoch": 0.661317772915147,
+      "grad_norm": 25.875,
+      "learning_rate": 1.9117341150263864e-06,
+      "loss": 1.0592,
+      "step": 2720
+    },
+    {
+      "epoch": 0.661560904449307,
+      "grad_norm": 19.875,
+      "learning_rate": 1.911324075235944e-06,
+      "loss": 0.9669,
+      "step": 2721
+    },
+    {
+      "epoch": 0.661804035983467,
+      "grad_norm": 23.375,
+      "learning_rate": 1.910913936599197e-06,
+      "loss": 0.8921,
+      "step": 2722
+    },
+    {
+      "epoch": 0.662047167517627,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9105036991774476e-06,
+      "loss": 0.8377,
+      "step": 2723
+    },
+    {
+      "epoch": 0.662290299051787,
+      "grad_norm": 25.5,
+      "learning_rate": 1.9100933630320135e-06,
+      "loss": 0.8749,
+      "step": 2724
+    },
+    {
+      "epoch": 0.662533430585947,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9096829282242257e-06,
+      "loss": 0.7983,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6627765621201069,
+      "grad_norm": 18.75,
+      "learning_rate": 1.909272394815432e-06,
+      "loss": 0.935,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6630196936542669,
+      "grad_norm": 17.5,
+      "learning_rate": 1.908861762866992e-06,
+      "loss": 0.8205,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6632628251884269,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.908451032440283e-06,
+      "loss": 0.5174,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6635059567225869,
+      "grad_norm": 17.125,
+      "learning_rate": 1.908040203596695e-06,
+      "loss": 0.6204,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6637490882567469,
+      "grad_norm": 22.75,
+      "learning_rate": 1.9076292763976338e-06,
+      "loss": 0.6751,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6639922197909068,
+      "grad_norm": 23.875,
+      "learning_rate": 1.90721825090452e-06,
+      "loss": 0.939,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6642353513250668,
+      "grad_norm": 42.0,
+      "learning_rate": 1.906807127178788e-06,
+      "loss": 1.2166,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6644784828592268,
+      "grad_norm": 17.875,
+      "learning_rate": 1.906395905281887e-06,
+      "loss": 0.433,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6647216143933868,
+      "grad_norm": 30.5,
+      "learning_rate": 1.905984585275282e-06,
+      "loss": 1.1259,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6649647459275468,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9055731672204513e-06,
+      "loss": 1.1893,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6652078774617067,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.9051616511788886e-06,
+      "loss": 0.7084,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6654510089958667,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9047500372121022e-06,
+      "loss": 0.8062,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6656941405300267,
+      "grad_norm": 19.625,
+      "learning_rate": 1.904338325381615e-06,
+      "loss": 0.9368,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6659372720641867,
+      "grad_norm": 21.375,
+      "learning_rate": 1.903926515748964e-06,
+      "loss": 0.8167,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6661804035983467,
+      "grad_norm": 23.75,
+      "learning_rate": 1.9035146083757012e-06,
+      "loss": 1.1495,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6664235351325067,
+      "grad_norm": 20.875,
+      "learning_rate": 1.903102603323394e-06,
+      "loss": 0.652,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9026905006536234e-06,
+      "loss": 0.693,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6669097982008266,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9022783004279852e-06,
+      "loss": 0.8309,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6671529297349866,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9018660027080893e-06,
+      "loss": 0.5793,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6673960612691466,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9014536075555612e-06,
+      "loss": 0.7686,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6676391928033066,
+      "grad_norm": 14.75,
+      "learning_rate": 1.9010411150320408e-06,
+      "loss": 0.7207,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6678823243374665,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9006285251991818e-06,
+      "loss": 0.6781,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6681254558716265,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9002158381186527e-06,
+      "loss": 0.5717,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6683685874057865,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8998030538521373e-06,
+      "loss": 0.7242,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6686117189399465,
+      "grad_norm": 22.125,
+      "learning_rate": 1.8993901724613328e-06,
+      "loss": 0.6692,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6688548504741065,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8989771940079517e-06,
+      "loss": 0.7458,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6690979820082664,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8985641185537207e-06,
+      "loss": 1.3859,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6693411135424264,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8981509461603815e-06,
+      "loss": 0.6103,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6695842450765864,
+      "grad_norm": 17.5,
+      "learning_rate": 1.8977376768896888e-06,
+      "loss": 0.8004,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6698273766107464,
+      "grad_norm": 17.375,
+      "learning_rate": 1.897324310803414e-06,
+      "loss": 0.7593,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6700705081449064,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8969108479633408e-06,
+      "loss": 0.9617,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6703136396790663,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8964972884312694e-06,
+      "loss": 0.6147,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6705567712132263,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8960836322690124e-06,
+      "loss": 1.0357,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6707999027473863,
+      "grad_norm": 24.875,
+      "learning_rate": 1.8956698795383985e-06,
+      "loss": 1.1743,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6710430342815463,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8952560303012702e-06,
+      "loss": 0.5745,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6712861658157063,
+      "grad_norm": 26.375,
+      "learning_rate": 1.8948420846194837e-06,
+      "loss": 1.2181,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6715292973498663,
+      "grad_norm": 17.875,
+      "learning_rate": 1.894428042554911e-06,
+      "loss": 1.0507,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6717724288840262,
+      "grad_norm": 17.0,
+      "learning_rate": 1.8940139041694377e-06,
+      "loss": 0.9185,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6720155604181862,
+      "grad_norm": 25.25,
+      "learning_rate": 1.8935996695249643e-06,
+      "loss": 0.8981,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6722586919523462,
+      "grad_norm": 21.0,
+      "learning_rate": 1.8931853386834047e-06,
+      "loss": 0.8092,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6725018234865062,
+      "grad_norm": 21.375,
+      "learning_rate": 1.8927709117066878e-06,
+      "loss": 0.9859,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6727449550206662,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8923563886567574e-06,
+      "loss": 1.413,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6729880865548261,
+      "grad_norm": 15.625,
+      "learning_rate": 1.8919417695955705e-06,
+      "loss": 0.6552,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6732312180889861,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.8915270545850998e-06,
+      "loss": 0.9794,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6734743496231461,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8911122436873313e-06,
+      "loss": 0.726,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6737174811573061,
+      "grad_norm": 13.5,
+      "learning_rate": 1.890697336964265e-06,
+      "loss": 0.4577,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6739606126914661,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.890282334477917e-06,
+      "loss": 0.637,
+      "step": 2772
+    },
+    {
+      "epoch": 0.674203744225626,
+      "grad_norm": 17.5,
+      "learning_rate": 1.889867236290316e-06,
+      "loss": 0.8888,
+      "step": 2773
+    },
+    {
+      "epoch": 0.674446875759786,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8894520424635055e-06,
+      "loss": 0.9943,
+      "step": 2774
+    },
+    {
+      "epoch": 0.674690007293946,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8890367530595435e-06,
+      "loss": 1.2011,
+      "step": 2775
+    },
+    {
+      "epoch": 0.674933138828106,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8886213681405022e-06,
+      "loss": 0.8127,
+      "step": 2776
+    },
+    {
+      "epoch": 0.675176270362266,
+      "grad_norm": 17.0,
+      "learning_rate": 1.8882058877684684e-06,
+      "loss": 0.8076,
+      "step": 2777
+    },
+    {
+      "epoch": 0.675419401896426,
+      "grad_norm": 18.5,
+      "learning_rate": 1.887790312005542e-06,
+      "loss": 1.0948,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6756625334305859,
+      "grad_norm": 24.75,
+      "learning_rate": 1.887374640913839e-06,
+      "loss": 0.8811,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6759056649647459,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8869588745554874e-06,
+      "loss": 0.9598,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6761487964989059,
+      "grad_norm": 18.5,
+      "learning_rate": 1.8865430129926316e-06,
+      "loss": 0.8612,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6763919280330659,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.8861270562874295e-06,
+      "loss": 0.6872,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6766350595672259,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8857110045020518e-06,
+      "loss": 0.836,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6768781911013858,
+      "grad_norm": 22.125,
+      "learning_rate": 1.885294857698686e-06,
+      "loss": 0.9759,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6771213226355458,
+      "grad_norm": 20.375,
+      "learning_rate": 1.8848786159395317e-06,
+      "loss": 1.1223,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6773644541697058,
+      "grad_norm": 20.375,
+      "learning_rate": 1.884462279286803e-06,
+      "loss": 0.7113,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6776075857038658,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8840458478027296e-06,
+      "loss": 0.818,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6778507172380258,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8836293215495535e-06,
+      "loss": 1.1104,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6780938487721857,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8832127005895325e-06,
+      "loss": 0.6935,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6783369803063457,
+      "grad_norm": 20.0,
+      "learning_rate": 1.882795984984937e-06,
+      "loss": 0.8294,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6785801118405057,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8823791747980535e-06,
+      "loss": 0.7348,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6788232433746657,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8819622700911804e-06,
+      "loss": 0.95,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6790663749088257,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8815452709266314e-06,
+      "loss": 0.8172,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6793095064429856,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8811281773667347e-06,
+      "loss": 0.8765,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6795526379771456,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8807109894738317e-06,
+      "loss": 0.7575,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6797957695113056,
+      "grad_norm": 24.125,
+      "learning_rate": 1.8802937073102796e-06,
+      "loss": 0.8822,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6800389010454656,
+      "grad_norm": 14.875,
+      "learning_rate": 1.8798763309384463e-06,
+      "loss": 0.6037,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6802820325796256,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8794588604207173e-06,
+      "loss": 0.8779,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6805251641137856,
+      "grad_norm": 24.125,
+      "learning_rate": 1.8790412958194903e-06,
+      "loss": 0.6963,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6807682956479455,
+      "grad_norm": 17.375,
+      "learning_rate": 1.878623637197178e-06,
+      "loss": 0.6414,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6810114271821055,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.8782058846162065e-06,
+      "loss": 0.7041,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6812545587162655,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8777880381390157e-06,
+      "loss": 0.4329,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6814976902504255,
+      "grad_norm": 22.75,
+      "learning_rate": 1.8773700978280607e-06,
+      "loss": 1.1243,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6817408217845855,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8769520637458094e-06,
+      "loss": 0.8327,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6819839533187454,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8765339359547441e-06,
+      "loss": 0.6192,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6822270848529054,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8761157145173613e-06,
+      "loss": 1.2173,
+      "step": 2806
+    },
+    {
+      "epoch": 0.6824702163870654,
+      "grad_norm": 18.125,
+      "learning_rate": 1.875697399496172e-06,
+      "loss": 0.9912,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6827133479212254,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8752789909537005e-06,
+      "loss": 0.7923,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6829564794553854,
+      "grad_norm": 31.75,
+      "learning_rate": 1.8748604889524844e-06,
+      "loss": 1.0251,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6831996109895453,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8744418935550764e-06,
+      "loss": 0.625,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6834427425237053,
+      "grad_norm": 17.125,
+      "learning_rate": 1.874023204824043e-06,
+      "loss": 1.0231,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6836858740578653,
+      "grad_norm": 18.0,
+      "learning_rate": 1.8736044228219647e-06,
+      "loss": 0.8958,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6839290055920253,
+      "grad_norm": 22.5,
+      "learning_rate": 1.8731855476114353e-06,
+      "loss": 1.2198,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6841721371261853,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.8727665792550625e-06,
+      "loss": 0.4517,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6844152686603453,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8723475178154693e-06,
+      "loss": 0.9555,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6846584001945052,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8719283633552913e-06,
+      "loss": 0.8075,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6849015317286652,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8715091159371781e-06,
+      "loss": 0.7464,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6851446632628252,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8710897756237939e-06,
+      "loss": 0.9057,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6853877947969852,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8706703424778159e-06,
+      "loss": 0.8518,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6856309263311452,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8702508165619363e-06,
+      "loss": 0.9205,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6858740578653051,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8698311979388594e-06,
+      "loss": 0.6844,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6861171893994651,
+      "grad_norm": 23.5,
+      "learning_rate": 1.8694114866713056e-06,
+      "loss": 1.3278,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6863603209336251,
+      "grad_norm": 30.875,
+      "learning_rate": 1.8689916828220075e-06,
+      "loss": 1.3844,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6866034524677851,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8685717864537116e-06,
+      "loss": 0.4919,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6868465840019451,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8681517976291796e-06,
+      "loss": 0.8494,
+      "step": 2825
+    },
+    {
+      "epoch": 0.687089715536105,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8677317164111856e-06,
+      "loss": 1.1265,
+      "step": 2826
+    },
+    {
+      "epoch": 0.687332847070265,
+      "grad_norm": 18.75,
+      "learning_rate": 1.867311542862518e-06,
+      "loss": 0.5187,
+      "step": 2827
+    },
+    {
+      "epoch": 0.687575978604425,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8668912770459787e-06,
+      "loss": 0.8619,
+      "step": 2828
+    },
+    {
+      "epoch": 0.687819110138585,
+      "grad_norm": 18.625,
+      "learning_rate": 1.866470919024384e-06,
+      "loss": 0.9234,
+      "step": 2829
+    },
+    {
+      "epoch": 0.688062241672745,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8660504688605638e-06,
+      "loss": 0.9266,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6883053732069049,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8656299266173613e-06,
+      "loss": 0.9105,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6885485047410649,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8652092923576342e-06,
+      "loss": 0.8332,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6887916362752249,
+      "grad_norm": 19.875,
+      "learning_rate": 1.864788566144253e-06,
+      "loss": 1.1016,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6890347678093849,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8643677480401032e-06,
+      "loss": 0.8181,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6892778993435449,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8639468381080828e-06,
+      "loss": 0.6619,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6895210308777049,
+      "grad_norm": 15.75,
+      "learning_rate": 1.8635258364111042e-06,
+      "loss": 0.7536,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6897641624118648,
+      "grad_norm": 19.5,
+      "learning_rate": 1.863104743012093e-06,
+      "loss": 0.9637,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6900072939460248,
+      "grad_norm": 20.75,
+      "learning_rate": 1.86268355797399e-06,
+      "loss": 1.1947,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6902504254801848,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8622622813597474e-06,
+      "loss": 0.8352,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6904935570143448,
+      "grad_norm": 22.125,
+      "learning_rate": 1.8618409132323329e-06,
+      "loss": 1.2988,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6907366885485048,
+      "grad_norm": 21.375,
+      "learning_rate": 1.861419453654727e-06,
+      "loss": 0.8299,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6909798200826647,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8609979026899239e-06,
+      "loss": 0.5336,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6912229516168247,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8605762604009323e-06,
+      "loss": 0.6185,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6914660831509847,
+      "grad_norm": 18.5,
+      "learning_rate": 1.8601545268507734e-06,
+      "loss": 0.6208,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6917092146851447,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8597327021024825e-06,
+      "loss": 0.8914,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6919523462193047,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.8593107862191095e-06,
+      "loss": 0.6565,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6921954777534646,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8588887792637158e-06,
+      "loss": 0.9686,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6924386092876246,
+      "grad_norm": 18.75,
+      "learning_rate": 1.858466681299378e-06,
+      "loss": 1.0165,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6926817408217846,
+      "grad_norm": 15.25,
+      "learning_rate": 1.8580444923891865e-06,
+      "loss": 0.6777,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6929248723559446,
+      "grad_norm": 20.75,
+      "learning_rate": 1.8576222125962442e-06,
+      "loss": 1.0999,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6931680038901046,
+      "grad_norm": 28.75,
+      "learning_rate": 1.8571998419836684e-06,
+      "loss": 1.1889,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6934111354242646,
+      "grad_norm": 15.875,
+      "learning_rate": 1.8567773806145892e-06,
+      "loss": 0.5628,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6936542669584245,
+      "grad_norm": 21.375,
+      "learning_rate": 1.8563548285521515e-06,
+      "loss": 1.0427,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6938973984925845,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8559321858595121e-06,
+      "loss": 1.1127,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6941405300267445,
+      "grad_norm": 18.625,
+      "learning_rate": 1.855509452599843e-06,
+      "loss": 1.1072,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6943836615609045,
+      "grad_norm": 14.5,
+      "learning_rate": 1.8550866288363284e-06,
+      "loss": 0.471,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6946267930950645,
+      "grad_norm": 21.875,
+      "learning_rate": 1.8546637146321672e-06,
+      "loss": 0.9184,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6948699246292244,
+      "grad_norm": 122.5,
+      "learning_rate": 1.854240710050571e-06,
+      "loss": 0.9554,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6951130561633844,
+      "grad_norm": 21.125,
+      "learning_rate": 1.853817615154765e-06,
+      "loss": 0.8874,
+      "step": 2859
+    },
+    {
+      "epoch": 0.6953561876975444,
+      "grad_norm": 22.375,
+      "learning_rate": 1.8533944300079876e-06,
+      "loss": 0.8626,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6955993192317044,
+      "grad_norm": 15.25,
+      "learning_rate": 1.8529711546734925e-06,
+      "loss": 0.4943,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6958424507658644,
+      "grad_norm": 18.0,
+      "learning_rate": 1.852547789214544e-06,
+      "loss": 1.1291,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6960855823000243,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8521243336944227e-06,
+      "loss": 0.6409,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6963287138341843,
+      "grad_norm": 20.625,
+      "learning_rate": 1.85170078817642e-06,
+      "loss": 0.7132,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6965718453683443,
+      "grad_norm": 26.75,
+      "learning_rate": 1.8512771527238433e-06,
+      "loss": 0.9868,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6968149769025043,
+      "grad_norm": 12.625,
+      "learning_rate": 1.8508534274000114e-06,
+      "loss": 0.4037,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6970581084366643,
+      "grad_norm": 18.5,
+      "learning_rate": 1.8504296122682578e-06,
+      "loss": 0.9511,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6973012399708242,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8500057073919286e-06,
+      "loss": 1.1812,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6975443715049842,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8495817128343844e-06,
+      "loss": 0.7531,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6977875030391442,
+      "grad_norm": 13.625,
+      "learning_rate": 1.849157628658998e-06,
+      "loss": 0.4609,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6980306345733042,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8487334549291562e-06,
+      "loss": 0.7991,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6982737661074642,
+      "grad_norm": 14.125,
+      "learning_rate": 1.8483091917082586e-06,
+      "loss": 0.2674,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6985168976416242,
+      "grad_norm": 25.875,
+      "learning_rate": 1.8478848390597195e-06,
+      "loss": 1.0291,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6987600291757841,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8474603970469653e-06,
+      "loss": 0.8569,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6990031607099441,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8470358657334363e-06,
+      "loss": 1.0268,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6992462922441041,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.846611245182586e-06,
+      "loss": 0.2849,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6994894237782641,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8461865354578814e-06,
+      "loss": 0.7604,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6997325553124241,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8457617366228027e-06,
+      "loss": 0.8744,
+      "step": 2878
+    },
+    {
+      "epoch": 0.699975686846584,
+      "grad_norm": 24.125,
+      "learning_rate": 1.8453368487408427e-06,
+      "loss": 1.1708,
+      "step": 2879
+    },
+    {
+      "epoch": 0.700218818380744,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8449118718755094e-06,
+      "loss": 0.6322,
+      "step": 2880
+    },
+    {
+      "epoch": 0.700461949914904,
+      "grad_norm": 20.125,
+      "learning_rate": 1.844486806090322e-06,
+      "loss": 0.7541,
+      "step": 2881
+    },
+    {
+      "epoch": 0.700705081449064,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8440616514488146e-06,
+      "loss": 0.7688,
+      "step": 2882
+    },
+    {
+      "epoch": 0.700948212983224,
+      "grad_norm": 28.875,
+      "learning_rate": 1.8436364080145333e-06,
+      "loss": 1.0153,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7011913445173839,
+      "grad_norm": 32.5,
+      "learning_rate": 1.8432110758510386e-06,
+      "loss": 1.027,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7014344760515439,
+      "grad_norm": 24.25,
+      "learning_rate": 1.8427856550219038e-06,
+      "loss": 0.7096,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7016776075857039,
+      "grad_norm": 21.375,
+      "learning_rate": 1.8423601455907145e-06,
+      "loss": 0.7179,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7019207391198639,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8419345476210712e-06,
+      "loss": 0.9237,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7021638706540239,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8415088611765866e-06,
+      "loss": 0.7091,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7024070021881839,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.8410830863208873e-06,
+      "loss": 0.4741,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7026501337223438,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8406572231176124e-06,
+      "loss": 0.789,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7028932652565038,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8402312716304138e-06,
+      "loss": 0.7747,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7031363967906638,
+      "grad_norm": 29.125,
+      "learning_rate": 1.8398052319229586e-06,
+      "loss": 0.9976,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7033795283248238,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8393791040589255e-06,
+      "loss": 1.1398,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7036226598589838,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8389528881020061e-06,
+      "loss": 0.8569,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7038657913931436,
+      "grad_norm": 18.0,
+      "learning_rate": 1.8385265841159056e-06,
+      "loss": 0.8613,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7041089229273036,
+      "grad_norm": 20.375,
+      "learning_rate": 1.8381001921643431e-06,
+      "loss": 0.7865,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7043520544614637,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8376737123110503e-06,
+      "loss": 0.6729,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7045951859956237,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8372471446197716e-06,
+      "loss": 0.7436,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7048383175297837,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8368204891542648e-06,
+      "loss": 0.6284,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7050814490639435,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8363937459783016e-06,
+      "loss": 1.0442,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7053245805981035,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8359669151556652e-06,
+      "loss": 0.8138,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7055677121322635,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8355399967501538e-06,
+      "loss": 1.0141,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7058108436664235,
+      "grad_norm": 20.75,
+      "learning_rate": 1.8351129908255767e-06,
+      "loss": 1.1416,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7060539752005836,
+      "grad_norm": 18.0,
+      "learning_rate": 1.8346858974457585e-06,
+      "loss": 0.8712,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7062971067347436,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8342587166745346e-06,
+      "loss": 0.9932,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7065402382689034,
+      "grad_norm": 22.375,
+      "learning_rate": 1.8338314485757553e-06,
+      "loss": 1.1834,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7067833698030634,
+      "grad_norm": 22.5,
+      "learning_rate": 1.8334040932132825e-06,
+      "loss": 1.0299,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7070265013372234,
+      "grad_norm": 34.0,
+      "learning_rate": 1.8329766506509925e-06,
+      "loss": 1.161,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7072696328713834,
+      "grad_norm": 21.25,
+      "learning_rate": 1.8325491209527737e-06,
+      "loss": 1.3779,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7075127644055434,
+      "grad_norm": 15.875,
+      "learning_rate": 1.8321215041825276e-06,
+      "loss": 0.6067,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7077558959397033,
+      "grad_norm": 26.0,
+      "learning_rate": 1.8316938004041695e-06,
+      "loss": 1.2639,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7079990274738633,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.8312660096816265e-06,
+      "loss": 0.8692,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7082421590080233,
+      "grad_norm": 24.5,
+      "learning_rate": 1.8308381320788397e-06,
+      "loss": 0.9224,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7084852905421833,
+      "grad_norm": 29.5,
+      "learning_rate": 1.8304101676597624e-06,
+      "loss": 1.2138,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7087284220763433,
+      "grad_norm": 28.5,
+      "learning_rate": 1.8299821164883613e-06,
+      "loss": 0.9979,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7089715536105032,
+      "grad_norm": 21.25,
+      "learning_rate": 1.829553978628617e-06,
+      "loss": 1.2982,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7092146851446632,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8291257541445206e-06,
+      "loss": 0.807,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7094578166788232,
+      "grad_norm": 23.75,
+      "learning_rate": 1.828697443100079e-06,
+      "loss": 1.2644,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7097009482129832,
+      "grad_norm": 19.875,
+      "learning_rate": 1.8282690455593096e-06,
+      "loss": 0.8658,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7099440797471432,
+      "grad_norm": 18.125,
+      "learning_rate": 1.8278405615862444e-06,
+      "loss": 0.876,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7101872112813031,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8274119912449279e-06,
+      "loss": 1.1041,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7104303428154631,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8269833345994168e-06,
+      "loss": 0.5185,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7106734743496231,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8265545917137817e-06,
+      "loss": 0.878,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7109166058837831,
+      "grad_norm": 17.375,
+      "learning_rate": 1.826125762652105e-06,
+      "loss": 0.8941,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7111597374179431,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8256968474784835e-06,
+      "loss": 0.6803,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7114028689521031,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8252678462570253e-06,
+      "loss": 1.1147,
+      "step": 2926
+    },
+    {
+      "epoch": 0.711646000486263,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8248387590518522e-06,
+      "loss": 1.1708,
+      "step": 2927
+    },
+    {
+      "epoch": 0.711889132020423,
+      "grad_norm": 19.875,
+      "learning_rate": 1.8244095859270992e-06,
+      "loss": 0.8755,
+      "step": 2928
+    },
+    {
+      "epoch": 0.712132263554583,
+      "grad_norm": 22.0,
+      "learning_rate": 1.8239803269469126e-06,
+      "loss": 0.9856,
+      "step": 2929
+    },
+    {
+      "epoch": 0.712375395088743,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8235509821754532e-06,
+      "loss": 0.7377,
+      "step": 2930
+    },
+    {
+      "epoch": 0.712618526622903,
+      "grad_norm": 20.625,
+      "learning_rate": 1.823121551676894e-06,
+      "loss": 0.9506,
+      "step": 2931
+    },
+    {
+      "epoch": 0.7128616581570629,
+      "grad_norm": 16.375,
+      "learning_rate": 1.822692035515421e-06,
+      "loss": 0.7385,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7131047896912229,
+      "grad_norm": 20.25,
+      "learning_rate": 1.8222624337552325e-06,
+      "loss": 0.9211,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7133479212253829,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8218327464605397e-06,
+      "loss": 1.2839,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7135910527595429,
+      "grad_norm": 22.375,
+      "learning_rate": 1.8214029736955675e-06,
+      "loss": 1.0768,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7138341842937029,
+      "grad_norm": 22.75,
+      "learning_rate": 1.8209731155245523e-06,
+      "loss": 1.1243,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7140773158278628,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8205431720117436e-06,
+      "loss": 0.8342,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7143204473620228,
+      "grad_norm": 15.875,
+      "learning_rate": 1.8201131432214045e-06,
+      "loss": 0.6138,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7145635788961828,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8196830292178097e-06,
+      "loss": 0.805,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7148067104303428,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8192528300652479e-06,
+      "loss": 0.7749,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7150498419645028,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8188225458280187e-06,
+      "loss": 0.7135,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7152929734986628,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8183921765704365e-06,
+      "loss": 0.5857,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7155361050328227,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8179617223568269e-06,
+      "loss": 0.7907,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7157792365669827,
+      "grad_norm": 15.25,
+      "learning_rate": 1.8175311832515289e-06,
+      "loss": 0.4061,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7160223681011427,
+      "grad_norm": 17.0,
+      "learning_rate": 1.8171005593188939e-06,
+      "loss": 0.865,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7162654996353027,
+      "grad_norm": 13.75,
+      "learning_rate": 1.816669850623286e-06,
+      "loss": 0.5594,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7165086311694627,
+      "grad_norm": 29.75,
+      "learning_rate": 1.8162390572290828e-06,
+      "loss": 1.004,
+      "step": 2947
+    },
+    {
+      "epoch": 0.7167517627036226,
+      "grad_norm": 28.75,
+      "learning_rate": 1.8158081792006727e-06,
+      "loss": 1.5714,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7169948942377826,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8153772166024585e-06,
+      "loss": 0.6644,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7172380257719426,
+      "grad_norm": 14.0,
+      "learning_rate": 1.8149461694988548e-06,
+      "loss": 0.4888,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7174811573061026,
+      "grad_norm": 18.625,
+      "learning_rate": 1.814515037954289e-06,
+      "loss": 0.5384,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7177242888402626,
+      "grad_norm": 15.0,
+      "learning_rate": 1.8140838220332019e-06,
+      "loss": 0.6208,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7179674203744225,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8136525218000448e-06,
+      "loss": 0.6364,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7182105519085825,
+      "grad_norm": 17.5,
+      "learning_rate": 1.8132211373192844e-06,
+      "loss": 0.8892,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7184536834427425,
+      "grad_norm": 21.125,
+      "learning_rate": 1.8127896686553973e-06,
+      "loss": 0.8518,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7186968149769025,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8123581158728744e-06,
+      "loss": 0.7045,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7189399465110625,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.811926479036219e-06,
+      "loss": 0.5171,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7191830780452224,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8114947582099466e-06,
+      "loss": 0.8784,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7194262095793824,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8110629534585854e-06,
+      "loss": 0.797,
+      "step": 2959
+    },
+    {
+      "epoch": 0.7196693411135424,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8106310648466754e-06,
+      "loss": 0.7181,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7199124726477024,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8101990924387708e-06,
+      "loss": 1.164,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7201556041818624,
+      "grad_norm": 17.25,
+      "learning_rate": 1.8097670362994368e-06,
+      "loss": 1.1201,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7203987357160224,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8093348964932516e-06,
+      "loss": 1.043,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7206418672501823,
+      "grad_norm": 29.875,
+      "learning_rate": 1.808902673084806e-06,
+      "loss": 1.2297,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7208849987843423,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8084703661387035e-06,
+      "loss": 0.9002,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7211281303185023,
+      "grad_norm": 21.5,
+      "learning_rate": 1.8080379757195597e-06,
+      "loss": 0.912,
+      "step": 2966
+    },
+    {
+      "epoch": 0.7213712618526623,
+      "grad_norm": 15.375,
+      "learning_rate": 1.8076055018920024e-06,
+      "loss": 0.65,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7216143933868223,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8071729447206731e-06,
+      "loss": 0.8992,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7218575249209822,
+      "grad_norm": 19.875,
+      "learning_rate": 1.8067403042702241e-06,
+      "loss": 1.1088,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7221006564551422,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8063075806053219e-06,
+      "loss": 0.909,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7223437879893022,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8058747737906436e-06,
+      "loss": 1.0959,
+      "step": 2971
+    },
+    {
+      "epoch": 0.7225869195234622,
+      "grad_norm": 17.875,
+      "learning_rate": 1.80544188389088e-06,
+      "loss": 1.126,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7228300510576222,
+      "grad_norm": 14.875,
+      "learning_rate": 1.8050089109707345e-06,
+      "loss": 0.4312,
+      "step": 2973
+    },
+    {
+      "epoch": 0.7230731825917821,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8045758550949217e-06,
+      "loss": 0.9033,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7233163141259421,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8041427163281693e-06,
+      "loss": 0.9799,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7235594456601021,
+      "grad_norm": 20.375,
+      "learning_rate": 1.8037094947352177e-06,
+      "loss": 0.8835,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7238025771942621,
+      "grad_norm": 21.625,
+      "learning_rate": 1.8032761903808194e-06,
+      "loss": 1.0635,
+      "step": 2977
+    },
+    {
+      "epoch": 0.7240457087284221,
+      "grad_norm": 17.375,
+      "learning_rate": 1.802842803329739e-06,
+      "loss": 0.768,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7242888402625821,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8024093336467535e-06,
+      "loss": 1.0363,
+      "step": 2979
+    },
+    {
+      "epoch": 0.724531971796742,
+      "grad_norm": 23.875,
+      "learning_rate": 1.8019757813966526e-06,
+      "loss": 0.9208,
+      "step": 2980
+    },
+    {
+      "epoch": 0.724775103330902,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8015421466442385e-06,
+      "loss": 0.7719,
+      "step": 2981
+    },
+    {
+      "epoch": 0.725018234865062,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8011084294543245e-06,
+      "loss": 0.6558,
+      "step": 2982
+    },
+    {
+      "epoch": 0.725261366399222,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8006746298917389e-06,
+      "loss": 0.8556,
+      "step": 2983
+    },
+    {
+      "epoch": 0.725504497933382,
+      "grad_norm": 26.125,
+      "learning_rate": 1.8002407480213183e-06,
+      "loss": 1.2889,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7257476294675419,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7998067839079154e-06,
+      "loss": 0.9437,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7259907610017019,
+      "grad_norm": 23.875,
+      "learning_rate": 1.799372737616393e-06,
+      "loss": 1.1154,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7262338925358619,
+      "grad_norm": 22.5,
+      "learning_rate": 1.798938609211627e-06,
+      "loss": 0.8806,
+      "step": 2987
+    },
+    {
+      "epoch": 0.7264770240700219,
+      "grad_norm": 19.875,
+      "learning_rate": 1.7985043987585054e-06,
+      "loss": 1.0027,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7267201556041819,
+      "grad_norm": 22.875,
+      "learning_rate": 1.7980701063219286e-06,
+      "loss": 1.3771,
+      "step": 2989
+    },
+    {
+      "epoch": 0.7269632871383418,
+      "grad_norm": 24.0,
+      "learning_rate": 1.7976357319668086e-06,
+      "loss": 1.0942,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7272064186725018,
+      "grad_norm": 18.75,
+      "learning_rate": 1.7972012757580703e-06,
+      "loss": 0.6214,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7274495502066618,
+      "grad_norm": 21.125,
+      "learning_rate": 1.7967667377606515e-06,
+      "loss": 1.0108,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7276926817408218,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7963321180395004e-06,
+      "loss": 0.8376,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7279358132749818,
+      "grad_norm": 34.0,
+      "learning_rate": 1.7958974166595788e-06,
+      "loss": 0.9713,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7281789448091417,
+      "grad_norm": 21.5,
+      "learning_rate": 1.7954626336858602e-06,
+      "loss": 1.6373,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7284220763433017,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.7950277691833308e-06,
+      "loss": 0.3456,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7286652078774617,
+      "grad_norm": 22.625,
+      "learning_rate": 1.7945928232169879e-06,
+      "loss": 0.9632,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7289083394116217,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7941577958518424e-06,
+      "loss": 0.7156,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7291514709457817,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.7937226871529162e-06,
+      "loss": 0.9448,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7293946024799417,
+      "grad_norm": 22.625,
+      "learning_rate": 1.7932874971852443e-06,
+      "loss": 1.0137,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7296377340141016,
+      "grad_norm": 21.125,
+      "learning_rate": 1.7928522260138729e-06,
+      "loss": 0.891,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7298808655482616,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7924168737038612e-06,
+      "loss": 0.8415,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7301239970824216,
+      "grad_norm": 21.75,
+      "learning_rate": 1.791981440320279e-06,
+      "loss": 0.641,
+      "step": 3003
+    },
+    {
+      "epoch": 0.7303671286165816,
+      "grad_norm": 15.125,
+      "learning_rate": 1.791545925928211e-06,
+      "loss": 0.6934,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7306102601507416,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7911103305927512e-06,
+      "loss": 0.8781,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7308533916849015,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.7906746543790075e-06,
+      "loss": 0.711,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7310965232190615,
+      "grad_norm": 17.0,
+      "learning_rate": 1.7902388973520987e-06,
+      "loss": 0.7602,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7313396547532215,
+      "grad_norm": 20.5,
+      "learning_rate": 1.7898030595771566e-06,
+      "loss": 0.6901,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7315827862873815,
+      "grad_norm": 19.125,
+      "learning_rate": 1.7893671411193244e-06,
+      "loss": 0.6929,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7318259178215415,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7889311420437578e-06,
+      "loss": 1.0812,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7320690493557014,
+      "grad_norm": 28.25,
+      "learning_rate": 1.7884950624156242e-06,
+      "loss": 1.1491,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7323121808898614,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7880589023001036e-06,
+      "loss": 0.7281,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7325553124240214,
+      "grad_norm": 27.25,
+      "learning_rate": 1.7876226617623874e-06,
+      "loss": 0.9335,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7327984439581814,
+      "grad_norm": 36.25,
+      "learning_rate": 1.7871863408676796e-06,
+      "loss": 1.1777,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7330415754923414,
+      "grad_norm": 23.0,
+      "learning_rate": 1.7867499396811949e-06,
+      "loss": 1.0634,
+      "step": 3015
+    },
+    {
+      "epoch": 0.7332847070265014,
+      "grad_norm": 20.5,
+      "learning_rate": 1.786313458268162e-06,
+      "loss": 0.9238,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7335278385606613,
+      "grad_norm": 16.75,
+      "learning_rate": 1.785876896693821e-06,
+      "loss": 0.9873,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7337709700948213,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7854402550234218e-06,
+      "loss": 0.6296,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7340141016289813,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7850035333222298e-06,
+      "loss": 0.6889,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7342572331631413,
+      "grad_norm": 18.0,
+      "learning_rate": 1.7845667316555198e-06,
+      "loss": 0.4127,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7345003646973013,
+      "grad_norm": 20.0,
+      "learning_rate": 1.7841298500885798e-06,
+      "loss": 0.682,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7347434962314612,
+      "grad_norm": 23.0,
+      "learning_rate": 1.7836928886867082e-06,
+      "loss": 1.0004,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7349866277656212,
+      "grad_norm": 20.75,
+      "learning_rate": 1.783255847515218e-06,
+      "loss": 0.78,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7352297592997812,
+      "grad_norm": 19.625,
+      "learning_rate": 1.7828187266394312e-06,
+      "loss": 0.886,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7354728908339412,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.7823815261246839e-06,
+      "loss": 0.6049,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7357160223681012,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.7819442460363225e-06,
+      "loss": 0.9734,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7359591539022611,
+      "grad_norm": 20.5,
+      "learning_rate": 1.781506886439707e-06,
+      "loss": 0.9641,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7362022854364211,
+      "grad_norm": 37.75,
+      "learning_rate": 1.7810694474002076e-06,
+      "loss": 1.4406,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7364454169705811,
+      "grad_norm": 22.125,
+      "learning_rate": 1.7806319289832078e-06,
+      "loss": 1.0294,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7366885485047411,
+      "grad_norm": 24.0,
+      "learning_rate": 1.7801943312541014e-06,
+      "loss": 0.6694,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7369316800389011,
+      "grad_norm": 18.625,
+      "learning_rate": 1.7797566542782956e-06,
+      "loss": 0.6523,
+      "step": 3031
+    },
+    {
+      "epoch": 0.737174811573061,
+      "grad_norm": 18.75,
+      "learning_rate": 1.779318898121209e-06,
+      "loss": 0.8146,
+      "step": 3032
+    },
+    {
+      "epoch": 0.737417943107221,
+      "grad_norm": 31.875,
+      "learning_rate": 1.7788810628482708e-06,
+      "loss": 0.816,
+      "step": 3033
+    },
+    {
+      "epoch": 0.737661074641381,
+      "grad_norm": 28.125,
+      "learning_rate": 1.778443148524924e-06,
+      "loss": 1.3549,
+      "step": 3034
+    },
+    {
+      "epoch": 0.737904206175541,
+      "grad_norm": 24.0,
+      "learning_rate": 1.778005155216622e-06,
+      "loss": 0.9372,
+      "step": 3035
+    },
+    {
+      "epoch": 0.738147337709701,
+      "grad_norm": 23.125,
+      "learning_rate": 1.7775670829888309e-06,
+      "loss": 0.8605,
+      "step": 3036
+    },
+    {
+      "epoch": 0.738390469243861,
+      "grad_norm": 20.0,
+      "learning_rate": 1.7771289319070276e-06,
+      "loss": 0.9511,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7386336007780209,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7766907020367013e-06,
+      "loss": 0.708,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7388767323121809,
+      "grad_norm": 21.0,
+      "learning_rate": 1.7762523934433538e-06,
+      "loss": 0.8422,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7391198638463409,
+      "grad_norm": 16.875,
+      "learning_rate": 1.7758140061924971e-06,
+      "loss": 0.686,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7393629953805009,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7753755403496564e-06,
+      "loss": 0.73,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7396061269146609,
+      "grad_norm": 17.0,
+      "learning_rate": 1.774936995980367e-06,
+      "loss": 0.6003,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7398492584488208,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7744983731501783e-06,
+      "loss": 0.9744,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7400923899829808,
+      "grad_norm": 20.625,
+      "learning_rate": 1.774059671924649e-06,
+      "loss": 1.174,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7403355215171408,
+      "grad_norm": 17.625,
+      "learning_rate": 1.773620892369351e-06,
+      "loss": 0.8853,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7405786530513008,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.7731820345498672e-06,
+      "loss": 0.5966,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7408217845854608,
+      "grad_norm": 18.625,
+      "learning_rate": 1.7727430985317927e-06,
+      "loss": 1.2801,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7410649161196207,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7723040843807343e-06,
+      "loss": 0.8067,
+      "step": 3048
+    },
+    {
+      "epoch": 0.7413080476537807,
+      "grad_norm": 16.5,
+      "learning_rate": 1.7718649921623097e-06,
+      "loss": 0.635,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7415511791879407,
+      "grad_norm": 16.125,
+      "learning_rate": 1.7714258219421493e-06,
+      "loss": 0.6125,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7417943107221007,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7709865737858945e-06,
+      "loss": 1.0174,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7420374422562607,
+      "grad_norm": 16.625,
+      "learning_rate": 1.7705472477591982e-06,
+      "loss": 0.7565,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7422805737904207,
+      "grad_norm": 23.5,
+      "learning_rate": 1.7701078439277255e-06,
+      "loss": 0.4331,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7425237053245806,
+      "grad_norm": 22.5,
+      "learning_rate": 1.7696683623571533e-06,
+      "loss": 0.9624,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7427668368587406,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.7692288031131694e-06,
+      "loss": 0.4313,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7430099683929006,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7687891662614733e-06,
+      "loss": 0.7108,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7432530999270606,
+      "grad_norm": 21.875,
+      "learning_rate": 1.7683494518677766e-06,
+      "loss": 0.8518,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7434962314612206,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.7679096599978019e-06,
+      "loss": 0.62,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7437393629953805,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7674697907172841e-06,
+      "loss": 1.3389,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7439824945295405,
+      "grad_norm": 22.5,
+      "learning_rate": 1.7670298440919692e-06,
+      "loss": 1.0756,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7442256260637005,
+      "grad_norm": 19.625,
+      "learning_rate": 1.766589820187614e-06,
+      "loss": 0.735,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7444687575978605,
+      "grad_norm": 24.625,
+      "learning_rate": 1.7661497190699894e-06,
+      "loss": 0.9854,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7447118891320205,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7657095408048744e-06,
+      "loss": 0.9337,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7449550206661804,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7652692854580622e-06,
+      "loss": 0.5433,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7451981522003404,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7648289530953561e-06,
+      "loss": 0.8421,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7454412837345004,
+      "grad_norm": 31.875,
+      "learning_rate": 1.7643885437825715e-06,
+      "loss": 1.266,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7456844152686604,
+      "grad_norm": 18.625,
+      "learning_rate": 1.7639480575855356e-06,
+      "loss": 0.7353,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7459275468028204,
+      "grad_norm": 14.0,
+      "learning_rate": 1.7635074945700858e-06,
+      "loss": 0.684,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7461706783369803,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7630668548020726e-06,
+      "loss": 0.6465,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7464138098711403,
+      "grad_norm": 36.5,
+      "learning_rate": 1.762626138347357e-06,
+      "loss": 1.2077,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7466569414053003,
+      "grad_norm": 20.875,
+      "learning_rate": 1.7621853452718115e-06,
+      "loss": 1.0533,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7469000729394603,
+      "grad_norm": 13.875,
+      "learning_rate": 1.7617444756413205e-06,
+      "loss": 0.5045,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7471432044736203,
+      "grad_norm": 16.5,
+      "learning_rate": 1.7613035295217795e-06,
+      "loss": 0.6456,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7473863360077803,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7608625069790959e-06,
+      "loss": 0.8867,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7476294675419402,
+      "grad_norm": 24.0,
+      "learning_rate": 1.760421408079187e-06,
+      "loss": 0.805,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7478725990761002,
+      "grad_norm": 21.5,
+      "learning_rate": 1.759980232887984e-06,
+      "loss": 0.5942,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7481157306102602,
+      "grad_norm": 12.375,
+      "learning_rate": 1.759538981471427e-06,
+      "loss": 0.3936,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7483588621444202,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7590976538954696e-06,
+      "loss": 1.0391,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7486019936785802,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.7586562502260753e-06,
+      "loss": 0.4015,
+      "step": 3079
+    },
+    {
+      "epoch": 0.74884512521274,
+      "grad_norm": 22.0,
+      "learning_rate": 1.7582147705292192e-06,
+      "loss": 0.9547,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7490882567469,
+      "grad_norm": 18.5,
+      "learning_rate": 1.757773214870889e-06,
+      "loss": 0.93,
+      "step": 3081
+    },
+    {
+      "epoch": 0.74933138828106,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7573315833170821e-06,
+      "loss": 0.8952,
+      "step": 3082
+    },
+    {
+      "epoch": 0.74957451981522,
+      "grad_norm": 19.125,
+      "learning_rate": 1.7568898759338082e-06,
+      "loss": 0.8293,
+      "step": 3083
+    },
+    {
+      "epoch": 0.74981765134938,
+      "grad_norm": 18.5,
+      "learning_rate": 1.756448092787088e-06,
+      "loss": 0.8411,
+      "step": 3084
+    },
+    {
+      "epoch": 0.75006078288354,
+      "grad_norm": 13.375,
+      "learning_rate": 1.7560062339429533e-06,
+      "loss": 0.4048,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7503039144177,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7555642994674489e-06,
+      "loss": 1.0634,
+      "step": 3086
+    },
+    {
+      "epoch": 0.75054704595186,
+      "grad_norm": 18.0,
+      "learning_rate": 1.7551222894266278e-06,
+      "loss": 0.7873,
+      "step": 3087
+    },
+    {
+      "epoch": 0.75079017748602,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7546802038865568e-06,
+      "loss": 0.7158,
+      "step": 3088
+    },
+    {
+      "epoch": 0.75103330902018,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.7542380429133133e-06,
+      "loss": 0.3718,
+      "step": 3089
+    },
+    {
+      "epoch": 0.75127644055434,
+      "grad_norm": 12.375,
+      "learning_rate": 1.7537958065729857e-06,
+      "loss": 0.3316,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7515195720884998,
+      "grad_norm": 26.0,
+      "learning_rate": 1.7533534949316745e-06,
+      "loss": 1.5041,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7517627036226598,
+      "grad_norm": 23.25,
+      "learning_rate": 1.7529111080554894e-06,
+      "loss": 0.8524,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7520058351568198,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.7524686460105542e-06,
+      "loss": 0.6641,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7522489666909798,
+      "grad_norm": 22.875,
+      "learning_rate": 1.7520261088630016e-06,
+      "loss": 0.5891,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7524920982251398,
+      "grad_norm": 17.875,
+      "learning_rate": 1.751583496678977e-06,
+      "loss": 0.9094,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7527352297592997,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.751140809524636e-06,
+      "loss": 0.4343,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7529783612934597,
+      "grad_norm": 22.25,
+      "learning_rate": 1.7506980474661462e-06,
+      "loss": 0.7665,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7532214928276197,
+      "grad_norm": 17.0,
+      "learning_rate": 1.750255210569686e-06,
+      "loss": 0.8135,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7534646243617797,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7498122989014443e-06,
+      "loss": 0.927,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7537077558959397,
+      "grad_norm": 18.5,
+      "learning_rate": 1.749369312527623e-06,
+      "loss": 0.7866,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7539508874300996,
+      "grad_norm": 18.5,
+      "learning_rate": 1.7489262515144333e-06,
+      "loss": 0.7273,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7541940189642596,
+      "grad_norm": 18.5,
+      "learning_rate": 1.7484831159280986e-06,
+      "loss": 1.1789,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7544371504984196,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7480399058348529e-06,
+      "loss": 0.8168,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7546802820325796,
+      "grad_norm": 22.625,
+      "learning_rate": 1.747596621300942e-06,
+      "loss": 0.8012,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7549234135667396,
+      "grad_norm": 26.25,
+      "learning_rate": 1.7471532623926227e-06,
+      "loss": 1.0752,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7551665451008995,
+      "grad_norm": 36.25,
+      "learning_rate": 1.7467098291761616e-06,
+      "loss": 0.8673,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7554096766350595,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7462663217178382e-06,
+      "loss": 0.9313,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7556528081692195,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7458227400839422e-06,
+      "loss": 0.8523,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7558959397033795,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7453790843407747e-06,
+      "loss": 0.7026,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7561390712375395,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.7449353545546477e-06,
+      "loss": 0.5233,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7563822027716995,
+      "grad_norm": 14.125,
+      "learning_rate": 1.7444915507918835e-06,
+      "loss": 0.653,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7566253343058594,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7440476731188175e-06,
+      "loss": 1.1768,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7568684658400194,
+      "grad_norm": 17.75,
+      "learning_rate": 1.743603721601794e-06,
+      "loss": 0.5922,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7571115973741794,
+      "grad_norm": 16.375,
+      "learning_rate": 1.7431596963071695e-06,
+      "loss": 0.7568,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7573547289083394,
+      "grad_norm": 15.5,
+      "learning_rate": 1.742715597301311e-06,
+      "loss": 0.6965,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7575978604424994,
+      "grad_norm": 16.0,
+      "learning_rate": 1.7422714246505972e-06,
+      "loss": 0.5604,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7578409919766593,
+      "grad_norm": 24.375,
+      "learning_rate": 1.7418271784214174e-06,
+      "loss": 1.4837,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7580841235108193,
+      "grad_norm": 18.75,
+      "learning_rate": 1.7413828586801713e-06,
+      "loss": 0.9794,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7583272550449793,
+      "grad_norm": 20.125,
+      "learning_rate": 1.7409384654932707e-06,
+      "loss": 1.03,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7585703865791393,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7404939989271374e-06,
+      "loss": 0.923,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7588135181132993,
+      "grad_norm": 19.5,
+      "learning_rate": 1.7400494590482049e-06,
+      "loss": 1.0926,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7590566496474592,
+      "grad_norm": 21.875,
+      "learning_rate": 1.7396048459229175e-06,
+      "loss": 0.6412,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7592997811816192,
+      "grad_norm": 21.375,
+      "learning_rate": 1.73916015961773e-06,
+      "loss": 1.078,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7595429127157792,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7387154001991086e-06,
+      "loss": 0.6388,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7597860442499392,
+      "grad_norm": 17.125,
+      "learning_rate": 1.73827056773353e-06,
+      "loss": 0.6687,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7600291757840992,
+      "grad_norm": 17.5,
+      "learning_rate": 1.7378256622874826e-06,
+      "loss": 0.8569,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7602723073182592,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7373806839274647e-06,
+      "loss": 1.1778,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7605154388524191,
+      "grad_norm": 19.75,
+      "learning_rate": 1.7369356327199862e-06,
+      "loss": 1.0933,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7607585703865791,
+      "grad_norm": 24.375,
+      "learning_rate": 1.736490508731568e-06,
+      "loss": 0.9143,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7610017019207391,
+      "grad_norm": 19.25,
+      "learning_rate": 1.736045312028741e-06,
+      "loss": 0.7533,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7612448334548991,
+      "grad_norm": 18.375,
+      "learning_rate": 1.735600042678048e-06,
+      "loss": 0.9688,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7614879649890591,
+      "grad_norm": 16.875,
+      "learning_rate": 1.735154700746042e-06,
+      "loss": 0.3887,
+      "step": 3132
+    },
+    {
+      "epoch": 0.761731096523219,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7347092862992871e-06,
+      "loss": 0.8986,
+      "step": 3133
+    },
+    {
+      "epoch": 0.761974228057379,
+      "grad_norm": 22.75,
+      "learning_rate": 1.7342637994043582e-06,
+      "loss": 1.1174,
+      "step": 3134
+    },
+    {
+      "epoch": 0.762217359591539,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.733818240127841e-06,
+      "loss": 0.7159,
+      "step": 3135
+    },
+    {
+      "epoch": 0.762460491125699,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.7333726085363317e-06,
+      "loss": 0.5699,
+      "step": 3136
+    },
+    {
+      "epoch": 0.762703622659859,
+      "grad_norm": 19.0,
+      "learning_rate": 1.732926904696438e-06,
+      "loss": 0.7077,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7629467541940189,
+      "grad_norm": 25.25,
+      "learning_rate": 1.7324811286747779e-06,
+      "loss": 0.9547,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7631898857281789,
+      "grad_norm": 16.375,
+      "learning_rate": 1.7320352805379807e-06,
+      "loss": 0.7508,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7634330172623389,
+      "grad_norm": 18.5,
+      "learning_rate": 1.7315893603526857e-06,
+      "loss": 1.1658,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7636761487964989,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7311433681855432e-06,
+      "loss": 1.1308,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7639192803306589,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7306973041032145e-06,
+      "loss": 1.1094,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7641624118648188,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7302511681723721e-06,
+      "loss": 1.0631,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7644055433989788,
+      "grad_norm": 19.75,
+      "learning_rate": 1.729804960459699e-06,
+      "loss": 0.7407,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7646486749331388,
+      "grad_norm": 20.625,
+      "learning_rate": 1.7293586810318872e-06,
+      "loss": 1.0228,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7648918064672988,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.7289123299556419e-06,
+      "loss": 0.5856,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7651349380014588,
+      "grad_norm": 14.0,
+      "learning_rate": 1.7284659072976778e-06,
+      "loss": 0.7226,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7653780695356188,
+      "grad_norm": 17.875,
+      "learning_rate": 1.7280194131247208e-06,
+      "loss": 0.8066,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7656212010697787,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7275728475035063e-06,
+      "loss": 0.6307,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7658643326039387,
+      "grad_norm": 19.125,
+      "learning_rate": 1.727126210500782e-06,
+      "loss": 0.8575,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7661074641380987,
+      "grad_norm": 23.625,
+      "learning_rate": 1.7266795021833052e-06,
+      "loss": 0.8573,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7663505956722587,
+      "grad_norm": 20.875,
+      "learning_rate": 1.7262327226178445e-06,
+      "loss": 1.252,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7665937272064187,
+      "grad_norm": 17.5,
+      "learning_rate": 1.7257858718711784e-06,
+      "loss": 0.9626,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7668368587405786,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7253389500100965e-06,
+      "loss": 0.6853,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7670799902747386,
+      "grad_norm": 16.125,
+      "learning_rate": 1.724891957101399e-06,
+      "loss": 0.7116,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7673231218088986,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7244448932118976e-06,
+      "loss": 0.7574,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7675662533430586,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7239977584084122e-06,
+      "loss": 0.9469,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7678093848772186,
+      "grad_norm": 21.0,
+      "learning_rate": 1.723550552757776e-06,
+      "loss": 1.0499,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7680525164113785,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7231032763268314e-06,
+      "loss": 0.4783,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7682956479455385,
+      "grad_norm": 18.75,
+      "learning_rate": 1.722655929182431e-06,
+      "loss": 1.0957,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7685387794796985,
+      "grad_norm": 17.375,
+      "learning_rate": 1.7222085113914388e-06,
+      "loss": 0.6827,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7687819110138585,
+      "grad_norm": 16.0,
+      "learning_rate": 1.7217610230207294e-06,
+      "loss": 0.682,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7690250425480185,
+      "grad_norm": 42.5,
+      "learning_rate": 1.7213134641371876e-06,
+      "loss": 0.893,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7692681740821785,
+      "grad_norm": 18.75,
+      "learning_rate": 1.7208658348077087e-06,
+      "loss": 0.5515,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7695113056163384,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7204181350991987e-06,
+      "loss": 0.762,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7697544371504984,
+      "grad_norm": 17.5,
+      "learning_rate": 1.7199703650785738e-06,
+      "loss": 0.7527,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7699975686846584,
+      "grad_norm": 14.625,
+      "learning_rate": 1.7195225248127611e-06,
+      "loss": 0.764,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7702407002188184,
+      "grad_norm": 14.875,
+      "learning_rate": 1.7190746143686986e-06,
+      "loss": 0.6568,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7704838317529784,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7186266338133334e-06,
+      "loss": 0.9405,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7707269632871383,
+      "grad_norm": 16.875,
+      "learning_rate": 1.7181785832136245e-06,
+      "loss": 0.9862,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7709700948212983,
+      "grad_norm": 26.0,
+      "learning_rate": 1.7177304626365404e-06,
+      "loss": 1.051,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7712132263554583,
+      "grad_norm": 23.25,
+      "learning_rate": 1.717282272149061e-06,
+      "loss": 1.0199,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7714563578896183,
+      "grad_norm": 25.25,
+      "learning_rate": 1.7168340118181754e-06,
+      "loss": 1.2688,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7716994894237783,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7163856817108845e-06,
+      "loss": 0.8713,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7719426209579382,
+      "grad_norm": 14.125,
+      "learning_rate": 1.715937281894199e-06,
+      "loss": 0.4008,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7721857524920982,
+      "grad_norm": 16.25,
+      "learning_rate": 1.715488812435139e-06,
+      "loss": 0.7364,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7724288840262582,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7150402734007372e-06,
+      "loss": 1.1006,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7726720155604182,
+      "grad_norm": 15.0,
+      "learning_rate": 1.7145916648580345e-06,
+      "loss": 0.5837,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7729151470945782,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.7141429868740843e-06,
+      "loss": 0.6657,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7731582786287381,
+      "grad_norm": 15.25,
+      "learning_rate": 1.7136942395159487e-06,
+      "loss": 0.4304,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7734014101628981,
+      "grad_norm": 16.125,
+      "learning_rate": 1.7132454228507002e-06,
+      "loss": 0.6216,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7736445416970581,
+      "grad_norm": 19.875,
+      "learning_rate": 1.7127965369454233e-06,
+      "loss": 1.0955,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7738876732312181,
+      "grad_norm": 18.25,
+      "learning_rate": 1.7123475818672108e-06,
+      "loss": 0.8218,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7741308047653781,
+      "grad_norm": 22.375,
+      "learning_rate": 1.7118985576831673e-06,
+      "loss": 1.0814,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7743739362995381,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7114494644604072e-06,
+      "loss": 1.1863,
+      "step": 3185
+    },
+    {
+      "epoch": 0.774617067833698,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7110003022660548e-06,
+      "loss": 0.825,
+      "step": 3186
+    },
+    {
+      "epoch": 0.774860199367858,
+      "grad_norm": 25.375,
+      "learning_rate": 1.7105510711672456e-06,
+      "loss": 0.7863,
+      "step": 3187
+    },
+    {
+      "epoch": 0.775103330902018,
+      "grad_norm": 23.5,
+      "learning_rate": 1.710101771231125e-06,
+      "loss": 0.903,
+      "step": 3188
+    },
+    {
+      "epoch": 0.775346462436178,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7096524025248483e-06,
+      "loss": 1.0174,
+      "step": 3189
+    },
+    {
+      "epoch": 0.775589593970338,
+      "grad_norm": 27.0,
+      "learning_rate": 1.7092029651155816e-06,
+      "loss": 1.0477,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7758327255044979,
+      "grad_norm": 19.375,
+      "learning_rate": 1.7087534590705012e-06,
+      "loss": 0.965,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7760758570386579,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7083038844567931e-06,
+      "loss": 1.0624,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7763189885728179,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7078542413416547e-06,
+      "loss": 1.0174,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7765621201069779,
+      "grad_norm": 12.125,
+      "learning_rate": 1.7074045297922924e-06,
+      "loss": 0.6654,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7768052516411379,
+      "grad_norm": 17.375,
+      "learning_rate": 1.7069547498759231e-06,
+      "loss": 0.8047,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7770483831752978,
+      "grad_norm": 18.125,
+      "learning_rate": 1.706504901659775e-06,
+      "loss": 0.7833,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7772915147094578,
+      "grad_norm": 16.625,
+      "learning_rate": 1.706054985211085e-06,
+      "loss": 0.676,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7775346462436178,
+      "grad_norm": 27.125,
+      "learning_rate": 1.7056050005971008e-06,
+      "loss": 1.1959,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 17.875,
+      "learning_rate": 1.7051549478850816e-06,
+      "loss": 0.9621,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7780209093119378,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7047048271422937e-06,
+      "loss": 1.1471,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7782640408460978,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7042546384360171e-06,
+      "loss": 0.7744,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7785071723802577,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.7038043818335389e-06,
+      "loss": 0.691,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7787503039144177,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7033540574021587e-06,
+      "loss": 0.6704,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7789934354485777,
+      "grad_norm": 19.0,
+      "learning_rate": 1.7029036652091846e-06,
+      "loss": 1.4227,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7792365669827377,
+      "grad_norm": 16.625,
+      "learning_rate": 1.7024532053219362e-06,
+      "loss": 0.6215,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7794796985168977,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7020026778077423e-06,
+      "loss": 0.7057,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7797228300510576,
+      "grad_norm": 24.75,
+      "learning_rate": 1.7015520827339415e-06,
+      "loss": 0.8641,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7799659615852176,
+      "grad_norm": 16.125,
+      "learning_rate": 1.7011014201678832e-06,
+      "loss": 1.3396,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7802090931193776,
+      "grad_norm": 17.0,
+      "learning_rate": 1.7006506901769271e-06,
+      "loss": 0.9634,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7804522246535376,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.7001998928284423e-06,
+      "loss": 0.627,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7806953561876976,
+      "grad_norm": 22.5,
+      "learning_rate": 1.6997490281898082e-06,
+      "loss": 0.8008,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7809384877218575,
+      "grad_norm": 19.625,
+      "learning_rate": 1.6992980963284144e-06,
+      "loss": 0.8307,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7811816192560175,
+      "grad_norm": 16.375,
+      "learning_rate": 1.6988470973116611e-06,
+      "loss": 0.7118,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7814247507901775,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6983960312069566e-06,
+      "loss": 0.92,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7816678823243375,
+      "grad_norm": 22.0,
+      "learning_rate": 1.6979448980817212e-06,
+      "loss": 1.061,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7819110138584975,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6974936980033846e-06,
+      "loss": 0.8675,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7821541453926574,
+      "grad_norm": 19.625,
+      "learning_rate": 1.6970424310393863e-06,
+      "loss": 0.705,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7823972769268174,
+      "grad_norm": 18.875,
+      "learning_rate": 1.6965910972571763e-06,
+      "loss": 0.833,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7826404084609774,
+      "grad_norm": 23.5,
+      "learning_rate": 1.6961396967242133e-06,
+      "loss": 1.1624,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7828835399951374,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6956882295079683e-06,
+      "loss": 1.2196,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7831266715292974,
+      "grad_norm": 19.625,
+      "learning_rate": 1.6952366956759195e-06,
+      "loss": 1.1915,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7833698030634574,
+      "grad_norm": 16.875,
+      "learning_rate": 1.6947850952955572e-06,
+      "loss": 0.7576,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7836129345976173,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6943334284343807e-06,
+      "loss": 0.648,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7838560661317773,
+      "grad_norm": 23.875,
+      "learning_rate": 1.6938816951598993e-06,
+      "loss": 0.8941,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7840991976659373,
+      "grad_norm": 29.125,
+      "learning_rate": 1.6934298955396331e-06,
+      "loss": 1.1015,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7843423292000973,
+      "grad_norm": 21.125,
+      "learning_rate": 1.6929780296411099e-06,
+      "loss": 0.9174,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7845854607342573,
+      "grad_norm": 23.25,
+      "learning_rate": 1.69252609753187e-06,
+      "loss": 0.6624,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7848285922684172,
+      "grad_norm": 17.625,
+      "learning_rate": 1.692074099279462e-06,
+      "loss": 0.7137,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7850717238025772,
+      "grad_norm": 21.375,
+      "learning_rate": 1.6916220349514451e-06,
+      "loss": 1.0116,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7853148553367372,
+      "grad_norm": 21.5,
+      "learning_rate": 1.6911699046153884e-06,
+      "loss": 0.8142,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7855579868708972,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6907177083388693e-06,
+      "loss": 0.8175,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7858011184050572,
+      "grad_norm": 21.125,
+      "learning_rate": 1.690265446189478e-06,
+      "loss": 0.7351,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7860442499392171,
+      "grad_norm": 26.875,
+      "learning_rate": 1.6898131182348118e-06,
+      "loss": 1.2597,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7862873814733771,
+      "grad_norm": 22.5,
+      "learning_rate": 1.6893607245424792e-06,
+      "loss": 0.9505,
+      "step": 3234
+    },
+    {
+      "epoch": 0.7865305130075371,
+      "grad_norm": 19.375,
+      "learning_rate": 1.6889082651800983e-06,
+      "loss": 0.994,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7867736445416971,
+      "grad_norm": 41.0,
+      "learning_rate": 1.6884557402152971e-06,
+      "loss": 0.8545,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7870167760758571,
+      "grad_norm": 29.25,
+      "learning_rate": 1.6880031497157133e-06,
+      "loss": 0.4975,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7872599076100171,
+      "grad_norm": 22.75,
+      "learning_rate": 1.687550493748994e-06,
+      "loss": 1.159,
+      "step": 3238
+    },
+    {
+      "epoch": 0.787503039144177,
+      "grad_norm": 28.625,
+      "learning_rate": 1.6870977723827963e-06,
+      "loss": 1.4826,
+      "step": 3239
+    },
+    {
+      "epoch": 0.787746170678337,
+      "grad_norm": 28.125,
+      "learning_rate": 1.686644985684788e-06,
+      "loss": 1.3861,
+      "step": 3240
+    },
+    {
+      "epoch": 0.787989302212497,
+      "grad_norm": 21.25,
+      "learning_rate": 1.6861921337226453e-06,
+      "loss": 1.2006,
+      "step": 3241
+    },
+    {
+      "epoch": 0.788232433746657,
+      "grad_norm": 25.5,
+      "learning_rate": 1.685739216564055e-06,
+      "loss": 1.059,
+      "step": 3242
+    },
+    {
+      "epoch": 0.788475565280817,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.6852862342767132e-06,
+      "loss": 0.6043,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7887186968149769,
+      "grad_norm": 16.75,
+      "learning_rate": 1.6848331869283263e-06,
+      "loss": 0.6907,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7889618283491369,
+      "grad_norm": 17.375,
+      "learning_rate": 1.6843800745866095e-06,
+      "loss": 0.6402,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7892049598832969,
+      "grad_norm": 17.125,
+      "learning_rate": 1.6839268973192888e-06,
+      "loss": 0.814,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7894480914174569,
+      "grad_norm": 18.375,
+      "learning_rate": 1.683473655194099e-06,
+      "loss": 0.4058,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7896912229516169,
+      "grad_norm": 30.75,
+      "learning_rate": 1.683020348278785e-06,
+      "loss": 1.2858,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7899343544857768,
+      "grad_norm": 19.75,
+      "learning_rate": 1.6825669766411015e-06,
+      "loss": 1.006,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7901774860199368,
+      "grad_norm": 22.25,
+      "learning_rate": 1.6821135403488126e-06,
+      "loss": 0.8786,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7904206175540968,
+      "grad_norm": 26.125,
+      "learning_rate": 1.6816600394696924e-06,
+      "loss": 1.0818,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7906637490882568,
+      "grad_norm": 27.75,
+      "learning_rate": 1.681206474071524e-06,
+      "loss": 1.0326,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7909068806224168,
+      "grad_norm": 21.75,
+      "learning_rate": 1.680752844222101e-06,
+      "loss": 1.0456,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7911500121565767,
+      "grad_norm": 23.625,
+      "learning_rate": 1.680299149989226e-06,
+      "loss": 0.6997,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7913931436907367,
+      "grad_norm": 29.25,
+      "learning_rate": 1.6798453914407115e-06,
+      "loss": 0.9938,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7916362752248967,
+      "grad_norm": 22.5,
+      "learning_rate": 1.6793915686443798e-06,
+      "loss": 0.9355,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7918794067590567,
+      "grad_norm": 17.5,
+      "learning_rate": 1.6789376816680622e-06,
+      "loss": 0.5805,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7921225382932167,
+      "grad_norm": 24.25,
+      "learning_rate": 1.6784837305796e-06,
+      "loss": 1.3705,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7923656698273767,
+      "grad_norm": 19.5,
+      "learning_rate": 1.6780297154468442e-06,
+      "loss": 1.0128,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7926088013615366,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.677575636337655e-06,
+      "loss": 0.3846,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7928519328956966,
+      "grad_norm": 25.0,
+      "learning_rate": 1.6771214933199027e-06,
+      "loss": 1.0039,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7930950644298566,
+      "grad_norm": 21.0,
+      "learning_rate": 1.6766672864614658e-06,
+      "loss": 1.1159,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7933381959640166,
+      "grad_norm": 20.5,
+      "learning_rate": 1.676213015830235e-06,
+      "loss": 1.0886,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7935813274981766,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6757586814941078e-06,
+      "loss": 1.0525,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7938244590323364,
+      "grad_norm": 19.0,
+      "learning_rate": 1.6753042835209924e-06,
+      "loss": 1.0886,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7940675905664965,
+      "grad_norm": 16.625,
+      "learning_rate": 1.6748498219788067e-06,
+      "loss": 0.6404,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7943107221006565,
+      "grad_norm": 18.875,
+      "learning_rate": 1.6743952969354777e-06,
+      "loss": 0.972,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7945538536348165,
+      "grad_norm": 16.25,
+      "learning_rate": 1.673940708458942e-06,
+      "loss": 0.5333,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7947969851689765,
+      "grad_norm": 22.25,
+      "learning_rate": 1.6734860566171454e-06,
+      "loss": 0.8507,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7950401167031363,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.6730313414780442e-06,
+      "loss": 0.6042,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7952832482372963,
+      "grad_norm": 16.0,
+      "learning_rate": 1.672576563109603e-06,
+      "loss": 0.6339,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7955263797714563,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.6721217215797961e-06,
+      "loss": 0.7245,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7957695113056164,
+      "grad_norm": 18.125,
+      "learning_rate": 1.6716668169566074e-06,
+      "loss": 0.7123,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7960126428397764,
+      "grad_norm": 19.25,
+      "learning_rate": 1.671211849308031e-06,
+      "loss": 0.8693,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7962557743739364,
+      "grad_norm": 23.25,
+      "learning_rate": 1.670756818702069e-06,
+      "loss": 0.9725,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7964989059080962,
+      "grad_norm": 19.25,
+      "learning_rate": 1.6703017252067336e-06,
+      "loss": 1.1398,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7967420374422562,
+      "grad_norm": 23.625,
+      "learning_rate": 1.6698465688900462e-06,
+      "loss": 0.9758,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7969851689764162,
+      "grad_norm": 21.75,
+      "learning_rate": 1.6693913498200383e-06,
+      "loss": 0.9211,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7972283005105762,
+      "grad_norm": 16.5,
+      "learning_rate": 1.6689360680647504e-06,
+      "loss": 0.589,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7974714320447363,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6684807236922318e-06,
+      "loss": 0.9694,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7977145635788961,
+      "grad_norm": 14.0,
+      "learning_rate": 1.6680253167705409e-06,
+      "loss": 0.4407,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7979576951130561,
+      "grad_norm": 16.0,
+      "learning_rate": 1.6675698473677473e-06,
+      "loss": 0.7154,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7982008266472161,
+      "grad_norm": 14.125,
+      "learning_rate": 1.6671143155519286e-06,
+      "loss": 0.6744,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7984439581813761,
+      "grad_norm": 14.125,
+      "learning_rate": 1.6666587213911714e-06,
+      "loss": 0.4512,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7986870897155361,
+      "grad_norm": 21.25,
+      "learning_rate": 1.6662030649535725e-06,
+      "loss": 1.0703,
+      "step": 3285
+    },
+    {
+      "epoch": 0.798930221249696,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.665747346307237e-06,
+      "loss": 0.6652,
+      "step": 3286
+    },
+    {
+      "epoch": 0.799173352783856,
+      "grad_norm": 18.5,
+      "learning_rate": 1.665291565520281e-06,
+      "loss": 0.7123,
+      "step": 3287
+    },
+    {
+      "epoch": 0.799416484318016,
+      "grad_norm": 18.0,
+      "learning_rate": 1.6648357226608281e-06,
+      "loss": 0.6371,
+      "step": 3288
+    },
+    {
+      "epoch": 0.799659615852176,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6643798177970113e-06,
+      "loss": 0.9583,
+      "step": 3289
+    },
+    {
+      "epoch": 0.799902747386336,
+      "grad_norm": 18.5,
+      "learning_rate": 1.663923850996975e-06,
+      "loss": 1.0215,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8001458789204959,
+      "grad_norm": 19.125,
+      "learning_rate": 1.66346782232887e-06,
+      "loss": 0.9595,
+      "step": 3291
+    },
+    {
+      "epoch": 0.8003890104546559,
+      "grad_norm": 17.875,
+      "learning_rate": 1.6630117318608585e-06,
+      "loss": 0.9341,
+      "step": 3292
+    },
+    {
+      "epoch": 0.8006321419888159,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.6625555796611102e-06,
+      "loss": 0.2916,
+      "step": 3293
+    },
+    {
+      "epoch": 0.8008752735229759,
+      "grad_norm": 19.5,
+      "learning_rate": 1.6620993657978054e-06,
+      "loss": 0.9743,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8011184050571359,
+      "grad_norm": 23.375,
+      "learning_rate": 1.6616430903391337e-06,
+      "loss": 1.0148,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8013615365912959,
+      "grad_norm": 22.875,
+      "learning_rate": 1.6611867533532921e-06,
+      "loss": 1.2657,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8016046681254558,
+      "grad_norm": 20.0,
+      "learning_rate": 1.6607303549084897e-06,
+      "loss": 1.0109,
+      "step": 3297
+    },
+    {
+      "epoch": 0.8018477996596158,
+      "grad_norm": 29.625,
+      "learning_rate": 1.6602738950729417e-06,
+      "loss": 1.0593,
+      "step": 3298
+    },
+    {
+      "epoch": 0.8020909311937758,
+      "grad_norm": 17.625,
+      "learning_rate": 1.6598173739148744e-06,
+      "loss": 0.6808,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8023340627279358,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.6593607915025227e-06,
+      "loss": 0.5933,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8025771942620958,
+      "grad_norm": 14.625,
+      "learning_rate": 1.658904147904131e-06,
+      "loss": 0.5634,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8028203257962557,
+      "grad_norm": 14.75,
+      "learning_rate": 1.6584474431879527e-06,
+      "loss": 0.6136,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8030634573304157,
+      "grad_norm": 18.875,
+      "learning_rate": 1.6579906774222493e-06,
+      "loss": 0.7261,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8033065888645757,
+      "grad_norm": 16.875,
+      "learning_rate": 1.657533850675293e-06,
+      "loss": 0.5733,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8035497203987357,
+      "grad_norm": 23.375,
+      "learning_rate": 1.6570769630153643e-06,
+      "loss": 0.7423,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8037928519328957,
+      "grad_norm": 17.375,
+      "learning_rate": 1.6566200145107534e-06,
+      "loss": 0.5861,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8040359834670556,
+      "grad_norm": 20.625,
+      "learning_rate": 1.6561630052297586e-06,
+      "loss": 0.9351,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8042791150012156,
+      "grad_norm": 23.375,
+      "learning_rate": 1.655705935240688e-06,
+      "loss": 1.3245,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8045222465353756,
+      "grad_norm": 16.75,
+      "learning_rate": 1.6552488046118588e-06,
+      "loss": 0.7572,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8047653780695356,
+      "grad_norm": 15.5,
+      "learning_rate": 1.6547916134115964e-06,
+      "loss": 0.4335,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8050085096036956,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.6543343617082364e-06,
+      "loss": 0.6441,
+      "step": 3311
+    },
+    {
+      "epoch": 0.8052516411378556,
+      "grad_norm": 16.75,
+      "learning_rate": 1.653877049570123e-06,
+      "loss": 0.798,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8054947726720155,
+      "grad_norm": 18.375,
+      "learning_rate": 1.6534196770656097e-06,
+      "loss": 0.6102,
+      "step": 3313
+    },
+    {
+      "epoch": 0.8057379042061755,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6529622442630583e-06,
+      "loss": 1.1213,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8059810357403355,
+      "grad_norm": 20.625,
+      "learning_rate": 1.6525047512308398e-06,
+      "loss": 0.8487,
+      "step": 3315
+    },
+    {
+      "epoch": 0.8062241672744955,
+      "grad_norm": 61.5,
+      "learning_rate": 1.6520471980373348e-06,
+      "loss": 0.8042,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8064672988086555,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6515895847509325e-06,
+      "loss": 0.9871,
+      "step": 3317
+    },
+    {
+      "epoch": 0.8067104303428154,
+      "grad_norm": 17.5,
+      "learning_rate": 1.6511319114400308e-06,
+      "loss": 0.6412,
+      "step": 3318
+    },
+    {
+      "epoch": 0.8069535618769754,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6506741781730379e-06,
+      "loss": 0.6683,
+      "step": 3319
+    },
+    {
+      "epoch": 0.8071966934111354,
+      "grad_norm": 18.0,
+      "learning_rate": 1.6502163850183683e-06,
+      "loss": 0.94,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8074398249452954,
+      "grad_norm": 38.5,
+      "learning_rate": 1.6497585320444487e-06,
+      "loss": 1.5201,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8076829564794554,
+      "grad_norm": 21.625,
+      "learning_rate": 1.649300619319712e-06,
+      "loss": 0.8041,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8079260880136153,
+      "grad_norm": 27.875,
+      "learning_rate": 1.6488426469126017e-06,
+      "loss": 0.9246,
+      "step": 3323
+    },
+    {
+      "epoch": 0.8081692195477753,
+      "grad_norm": 17.125,
+      "learning_rate": 1.6483846148915698e-06,
+      "loss": 0.7319,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8084123510819353,
+      "grad_norm": 19.25,
+      "learning_rate": 1.6479265233250763e-06,
+      "loss": 1.2679,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8086554826160953,
+      "grad_norm": 22.125,
+      "learning_rate": 1.647468372281592e-06,
+      "loss": 0.7942,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8088986141502553,
+      "grad_norm": 25.125,
+      "learning_rate": 1.6470101618295946e-06,
+      "loss": 1.0763,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8091417456844152,
+      "grad_norm": 24.125,
+      "learning_rate": 1.6465518920375723e-06,
+      "loss": 1.0571,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8093848772185752,
+      "grad_norm": 17.625,
+      "learning_rate": 1.6460935629740207e-06,
+      "loss": 0.8949,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8096280087527352,
+      "grad_norm": 20.25,
+      "learning_rate": 1.6456351747074454e-06,
+      "loss": 1.1852,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8098711402868952,
+      "grad_norm": 18.625,
+      "learning_rate": 1.6451767273063605e-06,
+      "loss": 0.7703,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8101142718210552,
+      "grad_norm": 19.0,
+      "learning_rate": 1.6447182208392887e-06,
+      "loss": 1.0443,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8103574033552152,
+      "grad_norm": 16.5,
+      "learning_rate": 1.644259655374762e-06,
+      "loss": 0.7272,
+      "step": 3333
+    },
+    {
+      "epoch": 0.8106005348893751,
+      "grad_norm": 25.125,
+      "learning_rate": 1.6438010309813202e-06,
+      "loss": 1.2095,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8108436664235351,
+      "grad_norm": 16.0,
+      "learning_rate": 1.6433423477275134e-06,
+      "loss": 0.9695,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8110867979576951,
+      "grad_norm": 25.375,
+      "learning_rate": 1.6428836056818995e-06,
+      "loss": 1.247,
+      "step": 3336
+    },
+    {
+      "epoch": 0.8113299294918551,
+      "grad_norm": 18.625,
+      "learning_rate": 1.6424248049130453e-06,
+      "loss": 1.0522,
+      "step": 3337
+    },
+    {
+      "epoch": 0.8115730610260151,
+      "grad_norm": 20.0,
+      "learning_rate": 1.641965945489527e-06,
+      "loss": 0.7952,
+      "step": 3338
+    },
+    {
+      "epoch": 0.811816192560175,
+      "grad_norm": 20.625,
+      "learning_rate": 1.641507027479928e-06,
+      "loss": 0.8989,
+      "step": 3339
+    },
+    {
+      "epoch": 0.812059324094335,
+      "grad_norm": 16.5,
+      "learning_rate": 1.6410480509528427e-06,
+      "loss": 0.5721,
+      "step": 3340
+    },
+    {
+      "epoch": 0.812302455628495,
+      "grad_norm": 21.5,
+      "learning_rate": 1.6405890159768722e-06,
+      "loss": 1.0475,
+      "step": 3341
+    },
+    {
+      "epoch": 0.812545587162655,
+      "grad_norm": 19.125,
+      "learning_rate": 1.640129922620628e-06,
+      "loss": 1.0032,
+      "step": 3342
+    },
+    {
+      "epoch": 0.812788718696815,
+      "grad_norm": 12.5,
+      "learning_rate": 1.6396707709527287e-06,
+      "loss": 0.3883,
+      "step": 3343
+    },
+    {
+      "epoch": 0.8130318502309749,
+      "grad_norm": 20.75,
+      "learning_rate": 1.639211561041803e-06,
+      "loss": 0.6252,
+      "step": 3344
+    },
+    {
+      "epoch": 0.8132749817651349,
+      "grad_norm": 17.25,
+      "learning_rate": 1.6387522929564874e-06,
+      "loss": 0.5771,
+      "step": 3345
+    },
+    {
+      "epoch": 0.8135181132992949,
+      "grad_norm": 40.5,
+      "learning_rate": 1.6382929667654278e-06,
+      "loss": 1.5006,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8137612448334549,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6378335825372786e-06,
+      "loss": 0.7383,
+      "step": 3347
+    },
+    {
+      "epoch": 0.8140043763676149,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6373741403407018e-06,
+      "loss": 1.3033,
+      "step": 3348
+    },
+    {
+      "epoch": 0.8142475079017749,
+      "grad_norm": 16.625,
+      "learning_rate": 1.6369146402443698e-06,
+      "loss": 0.6044,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8144906394359348,
+      "grad_norm": 17.25,
+      "learning_rate": 1.6364550823169625e-06,
+      "loss": 0.704,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8147337709700948,
+      "grad_norm": 19.125,
+      "learning_rate": 1.6359954666271688e-06,
+      "loss": 1.004,
+      "step": 3351
+    },
+    {
+      "epoch": 0.8149769025042548,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.6355357932436863e-06,
+      "loss": 0.4525,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8152200340384148,
+      "grad_norm": 22.5,
+      "learning_rate": 1.635076062235221e-06,
+      "loss": 1.1442,
+      "step": 3353
+    },
+    {
+      "epoch": 0.8154631655725748,
+      "grad_norm": 25.625,
+      "learning_rate": 1.6346162736704878e-06,
+      "loss": 1.3611,
+      "step": 3354
+    },
+    {
+      "epoch": 0.8157062971067347,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.6341564276182097e-06,
+      "loss": 0.8538,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8159494286408947,
+      "grad_norm": 18.125,
+      "learning_rate": 1.6336965241471193e-06,
+      "loss": 0.6753,
+      "step": 3356
+    },
+    {
+      "epoch": 0.8161925601750547,
+      "grad_norm": 14.25,
+      "learning_rate": 1.6332365633259568e-06,
+      "loss": 0.7703,
+      "step": 3357
+    },
+    {
+      "epoch": 0.8164356917092147,
+      "grad_norm": 15.0,
+      "learning_rate": 1.6327765452234706e-06,
+      "loss": 0.4715,
+      "step": 3358
+    },
+    {
+      "epoch": 0.8166788232433747,
+      "grad_norm": 16.5,
+      "learning_rate": 1.6323164699084193e-06,
+      "loss": 0.6636,
+      "step": 3359
+    },
+    {
+      "epoch": 0.8169219547775346,
+      "grad_norm": 25.25,
+      "learning_rate": 1.6318563374495686e-06,
+      "loss": 1.0865,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8171650863116946,
+      "grad_norm": 17.125,
+      "learning_rate": 1.6313961479156935e-06,
+      "loss": 0.5912,
+      "step": 3361
+    },
+    {
+      "epoch": 0.8174082178458546,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.6309359013755772e-06,
+      "loss": 0.2972,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8176513493800146,
+      "grad_norm": 18.875,
+      "learning_rate": 1.630475597898011e-06,
+      "loss": 0.9035,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8178944809141746,
+      "grad_norm": 19.75,
+      "learning_rate": 1.6300152375517964e-06,
+      "loss": 0.9052,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8181376124483345,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.629554820405741e-06,
+      "loss": 0.411,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8183807439824945,
+      "grad_norm": 24.25,
+      "learning_rate": 1.6290943465286623e-06,
+      "loss": 0.6659,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8186238755166545,
+      "grad_norm": 22.5,
+      "learning_rate": 1.6286338159893867e-06,
+      "loss": 0.7502,
+      "step": 3367
+    },
+    {
+      "epoch": 0.8188670070508145,
+      "grad_norm": 22.875,
+      "learning_rate": 1.6281732288567482e-06,
+      "loss": 0.9068,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8191101385849745,
+      "grad_norm": 22.125,
+      "learning_rate": 1.6277125851995892e-06,
+      "loss": 1.0115,
+      "step": 3369
+    },
+    {
+      "epoch": 0.8193532701191345,
+      "grad_norm": 23.0,
+      "learning_rate": 1.6272518850867609e-06,
+      "loss": 0.9395,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8195964016532944,
+      "grad_norm": 23.25,
+      "learning_rate": 1.6267911285871233e-06,
+      "loss": 0.9708,
+      "step": 3371
+    },
+    {
+      "epoch": 0.8198395331874544,
+      "grad_norm": 22.0,
+      "learning_rate": 1.6263303157695438e-06,
+      "loss": 1.1156,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8200826647216144,
+      "grad_norm": 18.625,
+      "learning_rate": 1.625869446702899e-06,
+      "loss": 0.7308,
+      "step": 3373
+    },
+    {
+      "epoch": 0.8203257962557744,
+      "grad_norm": 18.375,
+      "learning_rate": 1.6254085214560743e-06,
+      "loss": 0.7755,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8205689277899344,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6249475400979625e-06,
+      "loss": 0.8015,
+      "step": 3375
+    },
+    {
+      "epoch": 0.8208120593240943,
+      "grad_norm": 18.625,
+      "learning_rate": 1.6244865026974654e-06,
+      "loss": 0.7887,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8210551908582543,
+      "grad_norm": 16.125,
+      "learning_rate": 1.6240254093234925e-06,
+      "loss": 0.7635,
+      "step": 3377
+    },
+    {
+      "epoch": 0.8212983223924143,
+      "grad_norm": 14.75,
+      "learning_rate": 1.6235642600449628e-06,
+      "loss": 0.5193,
+      "step": 3378
+    },
+    {
+      "epoch": 0.8215414539265743,
+      "grad_norm": 21.125,
+      "learning_rate": 1.6231030549308024e-06,
+      "loss": 0.6491,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8217845854607343,
+      "grad_norm": 21.0,
+      "learning_rate": 1.622641794049947e-06,
+      "loss": 0.8781,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8220277169948942,
+      "grad_norm": 17.625,
+      "learning_rate": 1.6221804774713397e-06,
+      "loss": 1.0416,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8222708485290542,
+      "grad_norm": 17.625,
+      "learning_rate": 1.6217191052639323e-06,
+      "loss": 0.6293,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8225139800632142,
+      "grad_norm": 16.875,
+      "learning_rate": 1.6212576774966848e-06,
+      "loss": 0.6833,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8227571115973742,
+      "grad_norm": 22.75,
+      "learning_rate": 1.6207961942385655e-06,
+      "loss": 1.1186,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8230002431315342,
+      "grad_norm": 16.25,
+      "learning_rate": 1.620334655558551e-06,
+      "loss": 0.8758,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8232433746656942,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6198730615256267e-06,
+      "loss": 0.9285,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8234865061998541,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6194114122087852e-06,
+      "loss": 0.6273,
+      "step": 3387
+    },
+    {
+      "epoch": 0.8237296377340141,
+      "grad_norm": 20.25,
+      "learning_rate": 1.6189497076770282e-06,
+      "loss": 0.8995,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8239727692681741,
+      "grad_norm": 19.0,
+      "learning_rate": 1.618487947999365e-06,
+      "loss": 0.6926,
+      "step": 3389
+    },
+    {
+      "epoch": 0.8242159008023341,
+      "grad_norm": 16.875,
+      "learning_rate": 1.6180261332448146e-06,
+      "loss": 0.6152,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8244590323364941,
+      "grad_norm": 22.25,
+      "learning_rate": 1.6175642634824025e-06,
+      "loss": 1.3543,
+      "step": 3391
+    },
+    {
+      "epoch": 0.824702163870654,
+      "grad_norm": 21.25,
+      "learning_rate": 1.6171023387811627e-06,
+      "loss": 1.1115,
+      "step": 3392
+    },
+    {
+      "epoch": 0.824945295404814,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6166403592101384e-06,
+      "loss": 0.8612,
+      "step": 3393
+    },
+    {
+      "epoch": 0.825188426938974,
+      "grad_norm": 16.875,
+      "learning_rate": 1.6161783248383805e-06,
+      "loss": 0.5647,
+      "step": 3394
+    },
+    {
+      "epoch": 0.825431558473134,
+      "grad_norm": 18.5,
+      "learning_rate": 1.6157162357349482e-06,
+      "loss": 0.7841,
+      "step": 3395
+    },
+    {
+      "epoch": 0.825674690007294,
+      "grad_norm": 16.625,
+      "learning_rate": 1.6152540919689077e-06,
+      "loss": 0.5308,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8259178215414539,
+      "grad_norm": 30.5,
+      "learning_rate": 1.6147918936093355e-06,
+      "loss": 1.1225,
+      "step": 3397
+    },
+    {
+      "epoch": 0.8261609530756139,
+      "grad_norm": 21.75,
+      "learning_rate": 1.6143296407253142e-06,
+      "loss": 0.9388,
+      "step": 3398
+    },
+    {
+      "epoch": 0.8264040846097739,
+      "grad_norm": 27.5,
+      "learning_rate": 1.613867333385936e-06,
+      "loss": 1.1364,
+      "step": 3399
+    },
+    {
+      "epoch": 0.8266472161439339,
+      "grad_norm": 18.5,
+      "learning_rate": 1.613404971660301e-06,
+      "loss": 0.8135,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8268903476780939,
+      "grad_norm": 14.625,
+      "learning_rate": 1.612942555617516e-06,
+      "loss": 0.5177,
+      "step": 3401
+    },
+    {
+      "epoch": 0.8271334792122538,
+      "grad_norm": 21.75,
+      "learning_rate": 1.6124800853266986e-06,
+      "loss": 1.094,
+      "step": 3402
+    },
+    {
+      "epoch": 0.8273766107464138,
+      "grad_norm": 24.625,
+      "learning_rate": 1.6120175608569718e-06,
+      "loss": 0.786,
+      "step": 3403
+    },
+    {
+      "epoch": 0.8276197422805738,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6115549822774684e-06,
+      "loss": 0.6718,
+      "step": 3404
+    },
+    {
+      "epoch": 0.8278628738147338,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6110923496573283e-06,
+      "loss": 0.68,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8281060053488938,
+      "grad_norm": 21.0,
+      "learning_rate": 1.6106296630657005e-06,
+      "loss": 1.1493,
+      "step": 3406
+    },
+    {
+      "epoch": 0.8283491368830538,
+      "grad_norm": 20.75,
+      "learning_rate": 1.6101669225717417e-06,
+      "loss": 0.554,
+      "step": 3407
+    },
+    {
+      "epoch": 0.8285922684172137,
+      "grad_norm": 23.25,
+      "learning_rate": 1.6097041282446152e-06,
+      "loss": 1.284,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8288353999513737,
+      "grad_norm": 19.875,
+      "learning_rate": 1.6092412801534949e-06,
+      "loss": 0.8646,
+      "step": 3409
+    },
+    {
+      "epoch": 0.8290785314855337,
+      "grad_norm": 16.75,
+      "learning_rate": 1.6087783783675611e-06,
+      "loss": 0.7641,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8293216630196937,
+      "grad_norm": 24.5,
+      "learning_rate": 1.6083154229560022e-06,
+      "loss": 1.0882,
+      "step": 3411
+    },
+    {
+      "epoch": 0.8295647945538537,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.607852413988015e-06,
+      "loss": 0.4787,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8298079260880136,
+      "grad_norm": 17.625,
+      "learning_rate": 1.607389351532804e-06,
+      "loss": 0.78,
+      "step": 3413
+    },
+    {
+      "epoch": 0.8300510576221736,
+      "grad_norm": 24.5,
+      "learning_rate": 1.6069262356595827e-06,
+      "loss": 0.9327,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8302941891563336,
+      "grad_norm": 23.625,
+      "learning_rate": 1.6064630664375705e-06,
+      "loss": 1.0338,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8305373206904936,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6059998439359967e-06,
+      "loss": 0.9083,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8307804522246536,
+      "grad_norm": 20.375,
+      "learning_rate": 1.6055365682240985e-06,
+      "loss": 1.0087,
+      "step": 3417
+    },
+    {
+      "epoch": 0.8310235837588135,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6050732393711193e-06,
+      "loss": 0.9165,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8312667152929735,
+      "grad_norm": 28.0,
+      "learning_rate": 1.6046098574463126e-06,
+      "loss": 1.3364,
+      "step": 3419
+    },
+    {
+      "epoch": 0.8315098468271335,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.6041464225189376e-06,
+      "loss": 0.5851,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8317529783612935,
+      "grad_norm": 16.25,
+      "learning_rate": 1.603682934658264e-06,
+      "loss": 0.3832,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8319961098954535,
+      "grad_norm": 21.5,
+      "learning_rate": 1.6032193939335676e-06,
+      "loss": 0.7287,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8322392414296135,
+      "grad_norm": 21.875,
+      "learning_rate": 1.6027558004141323e-06,
+      "loss": 0.9834,
+      "step": 3423
+    },
+    {
+      "epoch": 0.8324823729637734,
+      "grad_norm": 19.375,
+      "learning_rate": 1.6022921541692501e-06,
+      "loss": 1.0242,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8327255044979334,
+      "grad_norm": 19.375,
+      "learning_rate": 1.6018284552682215e-06,
+      "loss": 0.9596,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8329686360320934,
+      "grad_norm": 21.125,
+      "learning_rate": 1.6013647037803539e-06,
+      "loss": 0.8035,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8332117675662534,
+      "grad_norm": 20.75,
+      "learning_rate": 1.6009008997749631e-06,
+      "loss": 1.0939,
+      "step": 3427
+    },
+    {
+      "epoch": 0.8334548991004134,
+      "grad_norm": 16.625,
+      "learning_rate": 1.600437043321372e-06,
+      "loss": 0.4163,
+      "step": 3428
+    },
+    {
+      "epoch": 0.8336980306345733,
+      "grad_norm": 23.625,
+      "learning_rate": 1.5999731344889132e-06,
+      "loss": 1.03,
+      "step": 3429
+    },
+    {
+      "epoch": 0.8339411621687333,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.599509173346925e-06,
+      "loss": 0.7969,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8341842937028933,
+      "grad_norm": 17.875,
+      "learning_rate": 1.599045159964755e-06,
+      "loss": 1.0562,
+      "step": 3431
+    },
+    {
+      "epoch": 0.8344274252370533,
+      "grad_norm": 32.5,
+      "learning_rate": 1.598581094411757e-06,
+      "loss": 0.8758,
+      "step": 3432
+    },
+    {
+      "epoch": 0.8346705567712133,
+      "grad_norm": 24.0,
+      "learning_rate": 1.598116976757294e-06,
+      "loss": 1.3562,
+      "step": 3433
+    },
+    {
+      "epoch": 0.8349136883053732,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5976528070707376e-06,
+      "loss": 0.7141,
+      "step": 3434
+    },
+    {
+      "epoch": 0.8351568198395332,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5971885854214642e-06,
+      "loss": 0.9906,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8353999513736932,
+      "grad_norm": 18.75,
+      "learning_rate": 1.596724311878861e-06,
+      "loss": 0.7147,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8356430829078532,
+      "grad_norm": 20.25,
+      "learning_rate": 1.596259986512321e-06,
+      "loss": 0.7191,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8358862144420132,
+      "grad_norm": 22.875,
+      "learning_rate": 1.5957956093912459e-06,
+      "loss": 1.1478,
+      "step": 3438
+    },
+    {
+      "epoch": 0.836129345976173,
+      "grad_norm": 31.5,
+      "learning_rate": 1.5953311805850448e-06,
+      "loss": 0.8481,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8363724775103331,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5948667001631352e-06,
+      "loss": 0.921,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8366156090444931,
+      "grad_norm": 22.375,
+      "learning_rate": 1.594402168194941e-06,
+      "loss": 0.8647,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8368587405786531,
+      "grad_norm": 21.5,
+      "learning_rate": 1.5939375847498944e-06,
+      "loss": 1.1712,
+      "step": 3442
+    },
+    {
+      "epoch": 0.8371018721128131,
+      "grad_norm": 21.625,
+      "learning_rate": 1.5934729498974362e-06,
+      "loss": 0.8327,
+      "step": 3443
+    },
+    {
+      "epoch": 0.8373450036469731,
+      "grad_norm": 23.375,
+      "learning_rate": 1.5930082637070132e-06,
+      "loss": 1.1919,
+      "step": 3444
+    },
+    {
+      "epoch": 0.837588135181133,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5925435262480815e-06,
+      "loss": 0.7159,
+      "step": 3445
+    },
+    {
+      "epoch": 0.837831266715293,
+      "grad_norm": 17.0,
+      "learning_rate": 1.592078737590104e-06,
+      "loss": 0.8424,
+      "step": 3446
+    },
+    {
+      "epoch": 0.838074398249453,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5916138978025509e-06,
+      "loss": 0.6072,
+      "step": 3447
+    },
+    {
+      "epoch": 0.838317529783613,
+      "grad_norm": 19.625,
+      "learning_rate": 1.591149006954901e-06,
+      "loss": 0.6674,
+      "step": 3448
+    },
+    {
+      "epoch": 0.838560661317773,
+      "grad_norm": 31.375,
+      "learning_rate": 1.5906840651166402e-06,
+      "loss": 1.2384,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8388037928519329,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.5902190723572622e-06,
+      "loss": 0.7587,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8390469243860929,
+      "grad_norm": 22.75,
+      "learning_rate": 1.589754028746268e-06,
+      "loss": 1.1475,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8392900559202529,
+      "grad_norm": 16.75,
+      "learning_rate": 1.5892889343531662e-06,
+      "loss": 0.665,
+      "step": 3452
+    },
+    {
+      "epoch": 0.8395331874544129,
+      "grad_norm": 29.5,
+      "learning_rate": 1.588823789247474e-06,
+      "loss": 1.2101,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8397763189885729,
+      "grad_norm": 50.75,
+      "learning_rate": 1.588358593498714e-06,
+      "loss": 0.9499,
+      "step": 3454
+    },
+    {
+      "epoch": 0.8400194505227327,
+      "grad_norm": 15.375,
+      "learning_rate": 1.5878933471764192e-06,
+      "loss": 0.7118,
+      "step": 3455
+    },
+    {
+      "epoch": 0.8402625820568927,
+      "grad_norm": 17.5,
+      "learning_rate": 1.5874280503501278e-06,
+      "loss": 0.7053,
+      "step": 3456
+    },
+    {
+      "epoch": 0.8405057135910527,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5869627030893867e-06,
+      "loss": 1.09,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8407488451252128,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5864973054637504e-06,
+      "loss": 1.0151,
+      "step": 3458
+    },
+    {
+      "epoch": 0.8409919766593728,
+      "grad_norm": 20.625,
+      "learning_rate": 1.5860318575427793e-06,
+      "loss": 0.9082,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8412351081935328,
+      "grad_norm": 15.25,
+      "learning_rate": 1.5855663593960446e-06,
+      "loss": 0.7381,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8414782397276926,
+      "grad_norm": 25.5,
+      "learning_rate": 1.585100811093122e-06,
+      "loss": 1.1652,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8417213712618526,
+      "grad_norm": 21.75,
+      "learning_rate": 1.5846352127035952e-06,
+      "loss": 1.0499,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8419645027960126,
+      "grad_norm": 22.625,
+      "learning_rate": 1.584169564297057e-06,
+      "loss": 0.8066,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8422076343301726,
+      "grad_norm": 17.875,
+      "learning_rate": 1.5837038659431059e-06,
+      "loss": 0.79,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8424507658643327,
+      "grad_norm": 17.375,
+      "learning_rate": 1.583238117711349e-06,
+      "loss": 0.822,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8426938973984925,
+      "grad_norm": 17.375,
+      "learning_rate": 1.5827723196713998e-06,
+      "loss": 0.814,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8429370289326525,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5823064718928807e-06,
+      "loss": 0.6785,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8431801604668125,
+      "grad_norm": 18.75,
+      "learning_rate": 1.58184057444542e-06,
+      "loss": 0.6204,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8434232920009725,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5813746273986541e-06,
+      "loss": 1.026,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8436664235351325,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5809086308222273e-06,
+      "loss": 0.8074,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8439095550692924,
+      "grad_norm": 20.75,
+      "learning_rate": 1.5804425847857908e-06,
+      "loss": 0.9195,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8441526866034524,
+      "grad_norm": 21.25,
+      "learning_rate": 1.5799764893590033e-06,
+      "loss": 1.2829,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8443958181376124,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5795103446115302e-06,
+      "loss": 0.7941,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8446389496717724,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.5790441506130453e-06,
+      "loss": 0.7508,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8448820812059324,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5785779074332292e-06,
+      "loss": 1.1244,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8451252127400923,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5781116151417703e-06,
+      "loss": 0.7926,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8453683442742523,
+      "grad_norm": 24.625,
+      "learning_rate": 1.5776452738083637e-06,
+      "loss": 0.9534,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8456114758084123,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5771788835027122e-06,
+      "loss": 0.9546,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8458546073425723,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5767124442945264e-06,
+      "loss": 0.7319,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8460977388767323,
+      "grad_norm": 15.125,
+      "learning_rate": 1.576245956253523e-06,
+      "loss": 0.7444,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8463408704108923,
+      "grad_norm": 20.0,
+      "learning_rate": 1.575779419449427e-06,
+      "loss": 0.9531,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8465840019450522,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5753128339519702e-06,
+      "loss": 0.9928,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8468271334792122,
+      "grad_norm": 20.5,
+      "learning_rate": 1.574846199830892e-06,
+      "loss": 0.7358,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8470702650133722,
+      "grad_norm": 18.0,
+      "learning_rate": 1.5743795171559392e-06,
+      "loss": 0.78,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8473133965475322,
+      "grad_norm": 17.0,
+      "learning_rate": 1.5739127859968652e-06,
+      "loss": 0.5908,
+      "step": 3485
+    },
+    {
+      "epoch": 0.8475565280816922,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5734460064234314e-06,
+      "loss": 0.9047,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8477996596158521,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.5729791785054056e-06,
+      "loss": 0.5851,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8480427911500121,
+      "grad_norm": 14.75,
+      "learning_rate": 1.5725123023125633e-06,
+      "loss": 0.9931,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8482859226841721,
+      "grad_norm": 17.0,
+      "learning_rate": 1.572045377914688e-06,
+      "loss": 0.7318,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8485290542183321,
+      "grad_norm": 17.875,
+      "learning_rate": 1.5715784053815687e-06,
+      "loss": 0.7543,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8487721857524921,
+      "grad_norm": 17.25,
+      "learning_rate": 1.5711113847830029e-06,
+      "loss": 0.6945,
+      "step": 3491
+    },
+    {
+      "epoch": 0.849015317286652,
+      "grad_norm": 21.5,
+      "learning_rate": 1.5706443161887948e-06,
+      "loss": 0.6763,
+      "step": 3492
+    },
+    {
+      "epoch": 0.849258448820812,
+      "grad_norm": 22.625,
+      "learning_rate": 1.570177199668756e-06,
+      "loss": 1.0924,
+      "step": 3493
+    },
+    {
+      "epoch": 0.849501580354972,
+      "grad_norm": 23.125,
+      "learning_rate": 1.569710035292705e-06,
+      "loss": 0.9922,
+      "step": 3494
+    },
+    {
+      "epoch": 0.849744711889132,
+      "grad_norm": 24.125,
+      "learning_rate": 1.5692428231304676e-06,
+      "loss": 0.8303,
+      "step": 3495
+    },
+    {
+      "epoch": 0.849987843423292,
+      "grad_norm": 25.75,
+      "learning_rate": 1.5687755632518772e-06,
+      "loss": 0.9958,
+      "step": 3496
+    },
+    {
+      "epoch": 0.850230974957452,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.5683082557267728e-06,
+      "loss": 0.3681,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8504741064916119,
+      "grad_norm": 25.25,
+      "learning_rate": 1.567840900625003e-06,
+      "loss": 0.8429,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8507172380257719,
+      "grad_norm": 23.75,
+      "learning_rate": 1.5673734980164204e-06,
+      "loss": 0.8081,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8509603695599319,
+      "grad_norm": 23.0,
+      "learning_rate": 1.5669060479708878e-06,
+      "loss": 0.9996,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8512035010940919,
+      "grad_norm": 19.125,
+      "learning_rate": 1.566438550558273e-06,
+      "loss": 0.906,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8514466326282519,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5659710058484518e-06,
+      "loss": 0.3945,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8516897641624118,
+      "grad_norm": 21.5,
+      "learning_rate": 1.5655034139113072e-06,
+      "loss": 0.812,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8519328956965718,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.5650357748167278e-06,
+      "loss": 0.8199,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8521760272307318,
+      "grad_norm": 23.0,
+      "learning_rate": 1.5645680886346112e-06,
+      "loss": 1.5766,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8524191587648918,
+      "grad_norm": 18.0,
+      "learning_rate": 1.564100355434861e-06,
+      "loss": 0.7244,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8526622902990518,
+      "grad_norm": 15.25,
+      "learning_rate": 1.563632575287388e-06,
+      "loss": 0.6229,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8529054218332117,
+      "grad_norm": 19.75,
+      "learning_rate": 1.56316474826211e-06,
+      "loss": 0.9417,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8531485533673717,
+      "grad_norm": 20.75,
+      "learning_rate": 1.5626968744289516e-06,
+      "loss": 0.8819,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8533916849015317,
+      "grad_norm": 20.0,
+      "learning_rate": 1.5622289538578453e-06,
+      "loss": 0.9293,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8536348164356917,
+      "grad_norm": 18.875,
+      "learning_rate": 1.5617609866187291e-06,
+      "loss": 0.8597,
+      "step": 3511
+    },
+    {
+      "epoch": 0.8538779479698517,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5612929727815494e-06,
+      "loss": 0.6089,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8541210795040116,
+      "grad_norm": 17.5,
+      "learning_rate": 1.5608249124162586e-06,
+      "loss": 0.7872,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8543642110381716,
+      "grad_norm": 19.875,
+      "learning_rate": 1.5603568055928164e-06,
+      "loss": 0.9658,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8546073425723316,
+      "grad_norm": 20.375,
+      "learning_rate": 1.5598886523811898e-06,
+      "loss": 0.8721,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8548504741064916,
+      "grad_norm": 18.625,
+      "learning_rate": 1.559420452851352e-06,
+      "loss": 0.7256,
+      "step": 3516
+    },
+    {
+      "epoch": 0.8550936056406516,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5589522070732838e-06,
+      "loss": 0.8953,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8553367371748116,
+      "grad_norm": 16.875,
+      "learning_rate": 1.558483915116972e-06,
+      "loss": 0.6124,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8555798687089715,
+      "grad_norm": 14.5,
+      "learning_rate": 1.5580155770524119e-06,
+      "loss": 0.4481,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8558230002431315,
+      "grad_norm": 17.25,
+      "learning_rate": 1.557547192949604e-06,
+      "loss": 0.9589,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8560661317772915,
+      "grad_norm": 21.375,
+      "learning_rate": 1.5570787628785563e-06,
+      "loss": 0.9213,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8563092633114515,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5566102869092847e-06,
+      "loss": 0.661,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8565523948456115,
+      "grad_norm": 23.0,
+      "learning_rate": 1.5561417651118098e-06,
+      "loss": 0.9228,
+      "step": 3523
+    },
+    {
+      "epoch": 0.8567955263797714,
+      "grad_norm": 16.625,
+      "learning_rate": 1.5556731975561613e-06,
+      "loss": 0.3625,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8570386579139314,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5552045843123737e-06,
+      "loss": 1.1483,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8572817894480914,
+      "grad_norm": 19.625,
+      "learning_rate": 1.5547359254504903e-06,
+      "loss": 0.6586,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8575249209822514,
+      "grad_norm": 23.125,
+      "learning_rate": 1.5542672210405603e-06,
+      "loss": 0.9238,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8577680525164114,
+      "grad_norm": 21.875,
+      "learning_rate": 1.5537984711526382e-06,
+      "loss": 0.9914,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8580111840505713,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5533296758567884e-06,
+      "loss": 0.9804,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8582543155847313,
+      "grad_norm": 24.875,
+      "learning_rate": 1.5528608352230798e-06,
+      "loss": 1.1042,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8584974471188913,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5523919493215888e-06,
+      "loss": 0.88,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8587405786530513,
+      "grad_norm": 21.75,
+      "learning_rate": 1.5519230182223984e-06,
+      "loss": 0.8651,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8589837101872113,
+      "grad_norm": 19.875,
+      "learning_rate": 1.5514540419955986e-06,
+      "loss": 0.7751,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8592268417213713,
+      "grad_norm": 17.5,
+      "learning_rate": 1.550985020711286e-06,
+      "loss": 0.6555,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8594699732555312,
+      "grad_norm": 18.375,
+      "learning_rate": 1.550515954439564e-06,
+      "loss": 0.829,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8597131047896912,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5500468432505422e-06,
+      "loss": 0.7958,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8599562363238512,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5495776872143379e-06,
+      "loss": 0.8569,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8601993678580112,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5491084864010741e-06,
+      "loss": 0.6053,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8604424993921712,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5486392408808818e-06,
+      "loss": 0.7797,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8606856309263311,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5481699507238965e-06,
+      "loss": 0.7783,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8609287624604911,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5477006160002631e-06,
+      "loss": 0.6464,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8611718939946511,
+      "grad_norm": 20.875,
+      "learning_rate": 1.547231236780131e-06,
+      "loss": 0.7141,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8614150255288111,
+      "grad_norm": 21.875,
+      "learning_rate": 1.546761813133657e-06,
+      "loss": 0.877,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8616581570629711,
+      "grad_norm": 24.125,
+      "learning_rate": 1.5462923451310049e-06,
+      "loss": 1.0813,
+      "step": 3544
+    },
+    {
+      "epoch": 0.861901288597131,
+      "grad_norm": 17.75,
+      "learning_rate": 1.5458228328423447e-06,
+      "loss": 0.6834,
+      "step": 3545
+    },
+    {
+      "epoch": 0.862144420131291,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.545353276337853e-06,
+      "loss": 0.4675,
+      "step": 3546
+    },
+    {
+      "epoch": 0.862387551665451,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5448836756877135e-06,
+      "loss": 0.8901,
+      "step": 3547
+    },
+    {
+      "epoch": 0.862630683199611,
+      "grad_norm": 21.625,
+      "learning_rate": 1.5444140309621153e-06,
+      "loss": 1.0051,
+      "step": 3548
+    },
+    {
+      "epoch": 0.862873814733771,
+      "grad_norm": 19.5,
+      "learning_rate": 1.5439443422312562e-06,
+      "loss": 0.8683,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8631169462679309,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.543474609565338e-06,
+      "loss": 0.7164,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8633600778020909,
+      "grad_norm": 17.75,
+      "learning_rate": 1.5430048330345712e-06,
+      "loss": 0.5809,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8636032093362509,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5425350127091716e-06,
+      "loss": 0.6418,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8638463408704109,
+      "grad_norm": 22.0,
+      "learning_rate": 1.5420651486593624e-06,
+      "loss": 0.9831,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8640894724045709,
+      "grad_norm": 25.25,
+      "learning_rate": 1.5415952409553721e-06,
+      "loss": 0.9321,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8643326039387309,
+      "grad_norm": 25.125,
+      "learning_rate": 1.5411252896674369e-06,
+      "loss": 0.6299,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8645757354728908,
+      "grad_norm": 24.25,
+      "learning_rate": 1.5406552948658e-06,
+      "loss": 1.4371,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8648188670070508,
+      "grad_norm": 23.0,
+      "learning_rate": 1.540185256620709e-06,
+      "loss": 1.1024,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8650619985412108,
+      "grad_norm": 19.125,
+      "learning_rate": 1.53971517500242e-06,
+      "loss": 0.7464,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8653051300753708,
+      "grad_norm": 20.625,
+      "learning_rate": 1.539245050081194e-06,
+      "loss": 0.9117,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8655482616095308,
+      "grad_norm": 22.125,
+      "learning_rate": 1.5387748819273001e-06,
+      "loss": 0.8153,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8657913931436907,
+      "grad_norm": 21.25,
+      "learning_rate": 1.5383046706110133e-06,
+      "loss": 1.0534,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8660345246778507,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.5378344162026137e-06,
+      "loss": 0.3717,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8662776562120107,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5373641187723898e-06,
+      "loss": 0.666,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8665207877461707,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5368937783906352e-06,
+      "loss": 0.7635,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8667639192803307,
+      "grad_norm": 20.75,
+      "learning_rate": 1.5364233951276505e-06,
+      "loss": 0.6799,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8670070508144906,
+      "grad_norm": 82.0,
+      "learning_rate": 1.5359529690537431e-06,
+      "loss": 1.3364,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8672501823486506,
+      "grad_norm": 23.875,
+      "learning_rate": 1.5354825002392254e-06,
+      "loss": 1.0862,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8674933138828106,
+      "grad_norm": 21.5,
+      "learning_rate": 1.535011988754418e-06,
+      "loss": 1.0549,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8677364454169706,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5345414346696463e-06,
+      "loss": 0.6873,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8679795769511306,
+      "grad_norm": 17.125,
+      "learning_rate": 1.5340708380552436e-06,
+      "loss": 0.7122,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8682227084852906,
+      "grad_norm": 17.0,
+      "learning_rate": 1.5336001989815472e-06,
+      "loss": 0.655,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8684658400194505,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5331295175189034e-06,
+      "loss": 0.7531,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8687089715536105,
+      "grad_norm": 17.375,
+      "learning_rate": 1.5326587937376635e-06,
+      "loss": 0.7692,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8689521030877705,
+      "grad_norm": 22.5,
+      "learning_rate": 1.5321880277081852e-06,
+      "loss": 0.8734,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8691952346219305,
+      "grad_norm": 18.875,
+      "learning_rate": 1.5317172195008326e-06,
+      "loss": 0.583,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8694383661560905,
+      "grad_norm": 17.25,
+      "learning_rate": 1.531246369185976e-06,
+      "loss": 0.7171,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8696814976902504,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5307754768339922e-06,
+      "loss": 0.9413,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8699246292244104,
+      "grad_norm": 20.625,
+      "learning_rate": 1.5303045425152643e-06,
+      "loss": 0.9347,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8701677607585704,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.5298335663001814e-06,
+      "loss": 0.7473,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8704108922927304,
+      "grad_norm": 23.25,
+      "learning_rate": 1.5293625482591396e-06,
+      "loss": 1.1312,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8706540238268904,
+      "grad_norm": 18.625,
+      "learning_rate": 1.52889148846254e-06,
+      "loss": 0.5924,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8708971553610503,
+      "grad_norm": 21.75,
+      "learning_rate": 1.5284203869807906e-06,
+      "loss": 0.9332,
+      "step": 3582
+    },
+    {
+      "epoch": 0.8711402868952103,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5279492438843058e-06,
+      "loss": 0.7122,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8713834184293703,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5274780592435064e-06,
+      "loss": 1.2327,
+      "step": 3584
+    },
+    {
+      "epoch": 0.8716265499635303,
+      "grad_norm": 40.75,
+      "learning_rate": 1.527006833128819e-06,
+      "loss": 0.9345,
+      "step": 3585
+    },
+    {
+      "epoch": 0.8718696814976903,
+      "grad_norm": 33.75,
+      "learning_rate": 1.5265355656106757e-06,
+      "loss": 1.377,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8721128130318502,
+      "grad_norm": 18.125,
+      "learning_rate": 1.526064256759517e-06,
+      "loss": 1.1876,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8723559445660102,
+      "grad_norm": 16.75,
+      "learning_rate": 1.5255929066457868e-06,
+      "loss": 1.1715,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8725990761001702,
+      "grad_norm": 17.25,
+      "learning_rate": 1.525121515339937e-06,
+      "loss": 0.5743,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8728422076343302,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5246500829124253e-06,
+      "loss": 0.9771,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8730853391684902,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.5241786094337151e-06,
+      "loss": 0.6867,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8733284707026502,
+      "grad_norm": 16.125,
+      "learning_rate": 1.5237070949742772e-06,
+      "loss": 0.5928,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8735716022368101,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.5232355396045864e-06,
+      "loss": 0.447,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8738147337709701,
+      "grad_norm": 28.375,
+      "learning_rate": 1.5227639433951252e-06,
+      "loss": 0.8632,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8740578653051301,
+      "grad_norm": 20.25,
+      "learning_rate": 1.5222923064163822e-06,
+      "loss": 0.8428,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8743009968392901,
+      "grad_norm": 23.0,
+      "learning_rate": 1.521820628738851e-06,
+      "loss": 1.3935,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8745441283734501,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5213489104330328e-06,
+      "loss": 0.6194,
+      "step": 3597
+    },
+    {
+      "epoch": 0.87478725990761,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5208771515694329e-06,
+      "loss": 0.7012,
+      "step": 3598
+    },
+    {
+      "epoch": 0.87503039144177,
+      "grad_norm": 16.75,
+      "learning_rate": 1.520405352218565e-06,
+      "loss": 0.681,
+      "step": 3599
+    },
+    {
+      "epoch": 0.87527352297593,
+      "grad_norm": 19.5,
+      "learning_rate": 1.519933512450947e-06,
+      "loss": 0.7969,
+      "step": 3600
+    },
+    {
+      "epoch": 0.87551665451009,
+      "grad_norm": 16.5,
+      "learning_rate": 1.5194616323371036e-06,
+      "loss": 0.6389,
+      "step": 3601
+    },
+    {
+      "epoch": 0.87575978604425,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5189897119475654e-06,
+      "loss": 1.2287,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8760029175784099,
+      "grad_norm": 23.875,
+      "learning_rate": 1.5185177513528693e-06,
+      "loss": 1.3549,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8762460491125699,
+      "grad_norm": 18.125,
+      "learning_rate": 1.518045750623558e-06,
+      "loss": 1.167,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8764891806467299,
+      "grad_norm": 24.875,
+      "learning_rate": 1.5175737098301792e-06,
+      "loss": 1.3455,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8767323121808899,
+      "grad_norm": 22.625,
+      "learning_rate": 1.517101629043289e-06,
+      "loss": 0.9083,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8769754437150499,
+      "grad_norm": 17.5,
+      "learning_rate": 1.5166295083334473e-06,
+      "loss": 0.8156,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8772185752492099,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5161573477712205e-06,
+      "loss": 0.5163,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8774617067833698,
+      "grad_norm": 22.375,
+      "learning_rate": 1.5156851474271815e-06,
+      "loss": 1.2113,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8777048383175298,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5152129073719085e-06,
+      "loss": 0.9579,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8779479698516898,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5147406276759865e-06,
+      "loss": 0.4668,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8781911013858498,
+      "grad_norm": 19.125,
+      "learning_rate": 1.514268308410005e-06,
+      "loss": 0.5061,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8784342329200098,
+      "grad_norm": 17.125,
+      "learning_rate": 1.5137959496445612e-06,
+      "loss": 1.0777,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8786773644541697,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5133235514502564e-06,
+      "loss": 0.7726,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8789204959883297,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.5128511138976992e-06,
+      "loss": 0.7334,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8791636275224897,
+      "grad_norm": 16.75,
+      "learning_rate": 1.5123786370575038e-06,
+      "loss": 0.5958,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8794067590566497,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5119061210002892e-06,
+      "loss": 0.8657,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8796498905908097,
+      "grad_norm": 19.0,
+      "learning_rate": 1.5114335657966816e-06,
+      "loss": 1.0357,
+      "step": 3618
+    },
+    {
+      "epoch": 0.8798930221249696,
+      "grad_norm": 20.375,
+      "learning_rate": 1.5109609715173127e-06,
+      "loss": 0.795,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8801361536591296,
+      "grad_norm": 22.875,
+      "learning_rate": 1.5104883382328195e-06,
+      "loss": 0.9602,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8803792851932896,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5100156660138454e-06,
+      "loss": 0.9176,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8806224167274496,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5095429549310392e-06,
+      "loss": 0.5905,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8808655482616096,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5090702050550562e-06,
+      "loss": 0.6135,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8811086797957695,
+      "grad_norm": 23.0,
+      "learning_rate": 1.5085974164565567e-06,
+      "loss": 1.1195,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8813518113299295,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.5081245892062072e-06,
+      "loss": 0.6659,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8815949428640895,
+      "grad_norm": 17.25,
+      "learning_rate": 1.5076517233746796e-06,
+      "loss": 0.8656,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8818380743982495,
+      "grad_norm": 20.625,
+      "learning_rate": 1.5071788190326521e-06,
+      "loss": 0.7428,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8820812059324095,
+      "grad_norm": 27.0,
+      "learning_rate": 1.506705876250809e-06,
+      "loss": 1.2332,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8823243374665695,
+      "grad_norm": 32.75,
+      "learning_rate": 1.5062328950998386e-06,
+      "loss": 0.8189,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8825674690007294,
+      "grad_norm": 30.375,
+      "learning_rate": 1.5057598756504373e-06,
+      "loss": 1.1142,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8828106005348894,
+      "grad_norm": 33.5,
+      "learning_rate": 1.5052868179733054e-06,
+      "loss": 0.918,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8830537320690494,
+      "grad_norm": 26.125,
+      "learning_rate": 1.5048137221391493e-06,
+      "loss": 1.3675,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8832968636032094,
+      "grad_norm": 21.625,
+      "learning_rate": 1.5043405882186819e-06,
+      "loss": 0.8361,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8835399951373694,
+      "grad_norm": 18.875,
+      "learning_rate": 1.5038674162826205e-06,
+      "loss": 0.5911,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8837831266715293,
+      "grad_norm": 18.75,
+      "learning_rate": 1.50339420640169e-06,
+      "loss": 0.7616,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8840262582056893,
+      "grad_norm": 22.875,
+      "learning_rate": 1.5029209586466184e-06,
+      "loss": 1.0386,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8842693897398493,
+      "grad_norm": 23.0,
+      "learning_rate": 1.502447673088142e-06,
+      "loss": 0.999,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8845125212740093,
+      "grad_norm": 25.875,
+      "learning_rate": 1.5019743497970008e-06,
+      "loss": 0.8509,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8847556528081693,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5015009888439408e-06,
+      "loss": 0.5531,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8849987843423291,
+      "grad_norm": 23.375,
+      "learning_rate": 1.5010275902997148e-06,
+      "loss": 0.7159,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8852419158764891,
+      "grad_norm": 12.75,
+      "learning_rate": 1.5005541542350802e-06,
+      "loss": 0.3923,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8854850474106492,
+      "grad_norm": 20.125,
+      "learning_rate": 1.5000806807207999e-06,
+      "loss": 0.7973,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8857281789448092,
+      "grad_norm": 17.0,
+      "learning_rate": 1.499607169827643e-06,
+      "loss": 1.0972,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8859713104789692,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4991336216263833e-06,
+      "loss": 0.7585,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8862144420131292,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4986600361878012e-06,
+      "loss": 1.1922,
+      "step": 3645
+    },
+    {
+      "epoch": 0.886457573547289,
+      "grad_norm": 26.375,
+      "learning_rate": 1.4981864135826823e-06,
+      "loss": 1.1931,
+      "step": 3646
+    },
+    {
+      "epoch": 0.886700705081449,
+      "grad_norm": 19.5,
+      "learning_rate": 1.497712753881818e-06,
+      "loss": 0.8558,
+      "step": 3647
+    },
+    {
+      "epoch": 0.886943836615609,
+      "grad_norm": 16.875,
+      "learning_rate": 1.4972390571560035e-06,
+      "loss": 0.6759,
+      "step": 3648
+    },
+    {
+      "epoch": 0.887186968149769,
+      "grad_norm": 20.0,
+      "learning_rate": 1.496765323476043e-06,
+      "loss": 1.4714,
+      "step": 3649
+    },
+    {
+      "epoch": 0.887430099683929,
+      "grad_norm": 18.25,
+      "learning_rate": 1.4962915529127426e-06,
+      "loss": 0.779,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8876732312180889,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.495817745536916e-06,
+      "loss": 0.7506,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8879163627522489,
+      "grad_norm": 17.75,
+      "learning_rate": 1.495343901419382e-06,
+      "loss": 0.7888,
+      "step": 3652
+    },
+    {
+      "epoch": 0.888159494286409,
+      "grad_norm": 19.75,
+      "learning_rate": 1.4948700206309645e-06,
+      "loss": 0.9544,
+      "step": 3653
+    },
+    {
+      "epoch": 0.888402625820569,
+      "grad_norm": 17.375,
+      "learning_rate": 1.4943961032424939e-06,
+      "loss": 0.877,
+      "step": 3654
+    },
+    {
+      "epoch": 0.888645757354729,
+      "grad_norm": 17.5,
+      "learning_rate": 1.4939221493248043e-06,
+      "loss": 0.6839,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 18.0,
+      "learning_rate": 1.493448158948737e-06,
+      "loss": 0.6713,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8891320204230488,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4929741321851376e-06,
+      "loss": 1.162,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8893751519572088,
+      "grad_norm": 20.75,
+      "learning_rate": 1.492500069104858e-06,
+      "loss": 0.7053,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8896182834913688,
+      "grad_norm": 13.5,
+      "learning_rate": 1.492025969778755e-06,
+      "loss": 0.5691,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8898614150255288,
+      "grad_norm": 20.25,
+      "learning_rate": 1.4915518342776909e-06,
+      "loss": 1.1944,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8901045465596887,
+      "grad_norm": 24.875,
+      "learning_rate": 1.4910776626725336e-06,
+      "loss": 1.1712,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8903476780938487,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.4906034550341559e-06,
+      "loss": 0.7042,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8905908096280087,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4901292114334362e-06,
+      "loss": 0.9218,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8908339411621687,
+      "grad_norm": 23.5,
+      "learning_rate": 1.489654931941259e-06,
+      "loss": 0.8784,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8910770726963287,
+      "grad_norm": 28.625,
+      "learning_rate": 1.4891806166285131e-06,
+      "loss": 1.4625,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8913202042304887,
+      "grad_norm": 17.5,
+      "learning_rate": 1.4887062655660933e-06,
+      "loss": 1.1436,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8915633357646486,
+      "grad_norm": 22.25,
+      "learning_rate": 1.4882318788248996e-06,
+      "loss": 0.7973,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8918064672988086,
+      "grad_norm": 20.125,
+      "learning_rate": 1.487757456475837e-06,
+      "loss": 0.7844,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8920495988329686,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4872829985898167e-06,
+      "loss": 0.9402,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8922927303671286,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4868085052377538e-06,
+      "loss": 0.7525,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8925358619012886,
+      "grad_norm": 25.625,
+      "learning_rate": 1.4863339764905698e-06,
+      "loss": 1.0096,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8927789934354485,
+      "grad_norm": 19.625,
+      "learning_rate": 1.485859412419192e-06,
+      "loss": 0.5815,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8930221249696085,
+      "grad_norm": 23.875,
+      "learning_rate": 1.4853848130945514e-06,
+      "loss": 0.9197,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8932652565037685,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4849101785875853e-06,
+      "loss": 0.8862,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8935083880379285,
+      "grad_norm": 16.625,
+      "learning_rate": 1.4844355089692352e-06,
+      "loss": 0.6091,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8937515195720885,
+      "grad_norm": 20.125,
+      "learning_rate": 1.4839608043104503e-06,
+      "loss": 1.1555,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8939946511062484,
+      "grad_norm": 16.75,
+      "learning_rate": 1.4834860646821825e-06,
+      "loss": 0.6236,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8942377826404084,
+      "grad_norm": 24.25,
+      "learning_rate": 1.4830112901553898e-06,
+      "loss": 1.389,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8944809141745684,
+      "grad_norm": 17.75,
+      "learning_rate": 1.482536480801035e-06,
+      "loss": 0.6193,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8947240457087284,
+      "grad_norm": 21.375,
+      "learning_rate": 1.4820616366900878e-06,
+      "loss": 0.6483,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8949671772428884,
+      "grad_norm": 19.875,
+      "learning_rate": 1.481586757893521e-06,
+      "loss": 0.8801,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8952103087770484,
+      "grad_norm": 21.5,
+      "learning_rate": 1.4811118444823133e-06,
+      "loss": 0.6479,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8954534403112083,
+      "grad_norm": 22.0,
+      "learning_rate": 1.4806368965274492e-06,
+      "loss": 1.0339,
+      "step": 3683
+    },
+    {
+      "epoch": 0.8956965718453683,
+      "grad_norm": 18.25,
+      "learning_rate": 1.4801619140999176e-06,
+      "loss": 0.9632,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8959397033795283,
+      "grad_norm": 15.625,
+      "learning_rate": 1.4796868972707132e-06,
+      "loss": 0.695,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8961828349136883,
+      "grad_norm": 26.5,
+      "learning_rate": 1.479211846110835e-06,
+      "loss": 0.8159,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8964259664478483,
+      "grad_norm": 22.125,
+      "learning_rate": 1.4787367606912872e-06,
+      "loss": 1.1078,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8966690979820082,
+      "grad_norm": 18.375,
+      "learning_rate": 1.478261641083081e-06,
+      "loss": 0.6567,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8969122295161682,
+      "grad_norm": 20.25,
+      "learning_rate": 1.47778648735723e-06,
+      "loss": 0.7557,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8971553610503282,
+      "grad_norm": 16.0,
+      "learning_rate": 1.4773112995847543e-06,
+      "loss": 0.7756,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8973984925844882,
+      "grad_norm": 17.5,
+      "learning_rate": 1.4768360778366791e-06,
+      "loss": 0.5595,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8976416241186482,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4763608221840346e-06,
+      "loss": 1.1363,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8978847556528081,
+      "grad_norm": 18.0,
+      "learning_rate": 1.475885532697856e-06,
+      "loss": 0.909,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8981278871869681,
+      "grad_norm": 21.625,
+      "learning_rate": 1.4754102094491826e-06,
+      "loss": 1.0297,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8983710187211281,
+      "grad_norm": 34.25,
+      "learning_rate": 1.4749348525090611e-06,
+      "loss": 0.9239,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8986141502552881,
+      "grad_norm": 22.0,
+      "learning_rate": 1.4744594619485412e-06,
+      "loss": 1.225,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8988572817894481,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.4739840378386782e-06,
+      "loss": 0.3779,
+      "step": 3697
+    },
+    {
+      "epoch": 0.899100413323608,
+      "grad_norm": 19.875,
+      "learning_rate": 1.473508580250532e-06,
+      "loss": 0.8047,
+      "step": 3698
+    },
+    {
+      "epoch": 0.899343544857768,
+      "grad_norm": 28.625,
+      "learning_rate": 1.4730330892551684e-06,
+      "loss": 0.9387,
+      "step": 3699
+    },
+    {
+      "epoch": 0.899586676391928,
+      "grad_norm": 20.375,
+      "learning_rate": 1.4725575649236578e-06,
+      "loss": 0.5998,
+      "step": 3700
+    },
+    {
+      "epoch": 0.899829807926088,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4720820073270755e-06,
+      "loss": 0.6035,
+      "step": 3701
+    },
+    {
+      "epoch": 0.900072939460248,
+      "grad_norm": 28.0,
+      "learning_rate": 1.4716064165365018e-06,
+      "loss": 0.7062,
+      "step": 3702
+    },
+    {
+      "epoch": 0.900316070994408,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4711307926230216e-06,
+      "loss": 0.7638,
+      "step": 3703
+    },
+    {
+      "epoch": 0.9005592025285679,
+      "grad_norm": 17.625,
+      "learning_rate": 1.4706551356577253e-06,
+      "loss": 0.7957,
+      "step": 3704
+    },
+    {
+      "epoch": 0.9008023340627279,
+      "grad_norm": 17.25,
+      "learning_rate": 1.4701794457117088e-06,
+      "loss": 0.6963,
+      "step": 3705
+    },
+    {
+      "epoch": 0.9010454655968879,
+      "grad_norm": 20.875,
+      "learning_rate": 1.4697037228560706e-06,
+      "loss": 0.5563,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9012885971310479,
+      "grad_norm": 24.375,
+      "learning_rate": 1.4692279671619173e-06,
+      "loss": 1.057,
+      "step": 3707
+    },
+    {
+      "epoch": 0.9015317286652079,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4687521787003577e-06,
+      "loss": 0.8086,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9017748601993678,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4682763575425078e-06,
+      "loss": 0.876,
+      "step": 3709
+    },
+    {
+      "epoch": 0.9020179917335278,
+      "grad_norm": 21.375,
+      "learning_rate": 1.4678005037594853e-06,
+      "loss": 0.7953,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9022611232676878,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4673246174224166e-06,
+      "loss": 0.7017,
+      "step": 3711
+    },
+    {
+      "epoch": 0.9025042548018478,
+      "grad_norm": 17.125,
+      "learning_rate": 1.4668486986024305e-06,
+      "loss": 0.7379,
+      "step": 3712
+    },
+    {
+      "epoch": 0.9027473863360078,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4663727473706612e-06,
+      "loss": 0.6913,
+      "step": 3713
+    },
+    {
+      "epoch": 0.9029905178701677,
+      "grad_norm": 22.0,
+      "learning_rate": 1.465896763798248e-06,
+      "loss": 0.8068,
+      "step": 3714
+    },
+    {
+      "epoch": 0.9032336494043277,
+      "grad_norm": 23.375,
+      "learning_rate": 1.4654207479563349e-06,
+      "loss": 1.0214,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9034767809384877,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4649446999160701e-06,
+      "loss": 1.1374,
+      "step": 3716
+    },
+    {
+      "epoch": 0.9037199124726477,
+      "grad_norm": 17.25,
+      "learning_rate": 1.464468619748608e-06,
+      "loss": 0.6149,
+      "step": 3717
+    },
+    {
+      "epoch": 0.9039630440068077,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4639925075251065e-06,
+      "loss": 0.5648,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9042061755409677,
+      "grad_norm": 16.375,
+      "learning_rate": 1.4635163633167296e-06,
+      "loss": 0.8324,
+      "step": 3719
+    },
+    {
+      "epoch": 0.9044493070751276,
+      "grad_norm": 21.5,
+      "learning_rate": 1.463040187194644e-06,
+      "loss": 0.8693,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9046924386092876,
+      "grad_norm": 23.875,
+      "learning_rate": 1.4625639792300233e-06,
+      "loss": 1.0723,
+      "step": 3721
+    },
+    {
+      "epoch": 0.9049355701434476,
+      "grad_norm": 16.25,
+      "learning_rate": 1.4620877394940447e-06,
+      "loss": 0.7044,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9051787016776076,
+      "grad_norm": 23.75,
+      "learning_rate": 1.4616114680578905e-06,
+      "loss": 0.8572,
+      "step": 3723
+    },
+    {
+      "epoch": 0.9054218332117676,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4611351649927482e-06,
+      "loss": 0.5652,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9056649647459275,
+      "grad_norm": 15.625,
+      "learning_rate": 1.4606588303698082e-06,
+      "loss": 0.3811,
+      "step": 3725
+    },
+    {
+      "epoch": 0.9059080962800875,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4601824642602688e-06,
+      "loss": 1.3879,
+      "step": 3726
+    },
+    {
+      "epoch": 0.9061512278142475,
+      "grad_norm": 21.375,
+      "learning_rate": 1.4597060667353296e-06,
+      "loss": 0.9602,
+      "step": 3727
+    },
+    {
+      "epoch": 0.9063943593484075,
+      "grad_norm": 19.5,
+      "learning_rate": 1.4592296378661968e-06,
+      "loss": 1.0221,
+      "step": 3728
+    },
+    {
+      "epoch": 0.9066374908825675,
+      "grad_norm": 25.125,
+      "learning_rate": 1.4587531777240814e-06,
+      "loss": 1.2204,
+      "step": 3729
+    },
+    {
+      "epoch": 0.9068806224167274,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4582766863801984e-06,
+      "loss": 0.971,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9071237539508874,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4578001639057676e-06,
+      "loss": 0.602,
+      "step": 3731
+    },
+    {
+      "epoch": 0.9073668854850474,
+      "grad_norm": 19.5,
+      "learning_rate": 1.4573236103720132e-06,
+      "loss": 0.8516,
+      "step": 3732
+    },
+    {
+      "epoch": 0.9076100170192074,
+      "grad_norm": 16.5,
+      "learning_rate": 1.4568470258501646e-06,
+      "loss": 0.5341,
+      "step": 3733
+    },
+    {
+      "epoch": 0.9078531485533674,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4563704104114557e-06,
+      "loss": 0.7416,
+      "step": 3734
+    },
+    {
+      "epoch": 0.9080962800875274,
+      "grad_norm": 18.125,
+      "learning_rate": 1.4558937641271248e-06,
+      "loss": 0.6536,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9083394116216873,
+      "grad_norm": 22.75,
+      "learning_rate": 1.455417087068415e-06,
+      "loss": 0.8411,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9085825431558473,
+      "grad_norm": 18.0,
+      "learning_rate": 1.4549403793065737e-06,
+      "loss": 0.8491,
+      "step": 3737
+    },
+    {
+      "epoch": 0.9088256746900073,
+      "grad_norm": 17.875,
+      "learning_rate": 1.4544636409128538e-06,
+      "loss": 0.8147,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9090688062241673,
+      "grad_norm": 18.625,
+      "learning_rate": 1.453986871958511e-06,
+      "loss": 0.6442,
+      "step": 3739
+    },
+    {
+      "epoch": 0.9093119377583273,
+      "grad_norm": 24.625,
+      "learning_rate": 1.4535100725148072e-06,
+      "loss": 0.7537,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9095550692924872,
+      "grad_norm": 19.0,
+      "learning_rate": 1.4530332426530086e-06,
+      "loss": 0.6678,
+      "step": 3741
+    },
+    {
+      "epoch": 0.9097982008266472,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.452556382444385e-06,
+      "loss": 1.1911,
+      "step": 3742
+    },
+    {
+      "epoch": 0.9100413323608072,
+      "grad_norm": 19.75,
+      "learning_rate": 1.4520794919602125e-06,
+      "loss": 1.2636,
+      "step": 3743
+    },
+    {
+      "epoch": 0.9102844638949672,
+      "grad_norm": 24.125,
+      "learning_rate": 1.4516025712717692e-06,
+      "loss": 0.9072,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9105275954291272,
+      "grad_norm": 24.25,
+      "learning_rate": 1.4511256204503403e-06,
+      "loss": 0.6314,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9107707269632871,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4506486395672134e-06,
+      "loss": 0.6908,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9110138584974471,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4501716286936824e-06,
+      "loss": 1.0111,
+      "step": 3747
+    },
+    {
+      "epoch": 0.9112569900316071,
+      "grad_norm": 19.125,
+      "learning_rate": 1.449694587901044e-06,
+      "loss": 0.7819,
+      "step": 3748
+    },
+    {
+      "epoch": 0.9115001215657671,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4492175172606006e-06,
+      "loss": 0.5429,
+      "step": 3749
+    },
+    {
+      "epoch": 0.9117432530999271,
+      "grad_norm": 21.75,
+      "learning_rate": 1.4487404168436593e-06,
+      "loss": 1.2505,
+      "step": 3750
+    },
+    {
+      "epoch": 0.911986384634087,
+      "grad_norm": 22.875,
+      "learning_rate": 1.44826328672153e-06,
+      "loss": 0.6566,
+      "step": 3751
+    },
+    {
+      "epoch": 0.912229516168247,
+      "grad_norm": 20.125,
+      "learning_rate": 1.447786126965528e-06,
+      "loss": 1.0915,
+      "step": 3752
+    },
+    {
+      "epoch": 0.912472647702407,
+      "grad_norm": 21.75,
+      "learning_rate": 1.4473089376469737e-06,
+      "loss": 1.2058,
+      "step": 3753
+    },
+    {
+      "epoch": 0.912715779236567,
+      "grad_norm": 17.375,
+      "learning_rate": 1.446831718837191e-06,
+      "loss": 0.8974,
+      "step": 3754
+    },
+    {
+      "epoch": 0.912958910770727,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.4463544706075088e-06,
+      "loss": 0.5681,
+      "step": 3755
+    },
+    {
+      "epoch": 0.913202042304887,
+      "grad_norm": 23.375,
+      "learning_rate": 1.4458771930292592e-06,
+      "loss": 0.84,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9134451738390469,
+      "grad_norm": 22.875,
+      "learning_rate": 1.4453998861737808e-06,
+      "loss": 1.123,
+      "step": 3757
+    },
+    {
+      "epoch": 0.9136883053732069,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4449225501124146e-06,
+      "loss": 1.0275,
+      "step": 3758
+    },
+    {
+      "epoch": 0.9139314369073669,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4444451849165067e-06,
+      "loss": 0.8495,
+      "step": 3759
+    },
+    {
+      "epoch": 0.9141745684415269,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4439677906574076e-06,
+      "loss": 0.754,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9144176999756869,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4434903674064726e-06,
+      "loss": 0.7724,
+      "step": 3761
+    },
+    {
+      "epoch": 0.9146608315098468,
+      "grad_norm": 17.25,
+      "learning_rate": 1.4430129152350605e-06,
+      "loss": 0.496,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9149039630440068,
+      "grad_norm": 24.0,
+      "learning_rate": 1.4425354342145346e-06,
+      "loss": 0.9878,
+      "step": 3763
+    },
+    {
+      "epoch": 0.9151470945781668,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4420579244162627e-06,
+      "loss": 0.8022,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9153902261123268,
+      "grad_norm": 20.75,
+      "learning_rate": 1.4415803859116173e-06,
+      "loss": 0.7064,
+      "step": 3765
+    },
+    {
+      "epoch": 0.9156333576464868,
+      "grad_norm": 18.5,
+      "learning_rate": 1.441102818771974e-06,
+      "loss": 0.8631,
+      "step": 3766
+    },
+    {
+      "epoch": 0.9158764891806467,
+      "grad_norm": 17.375,
+      "learning_rate": 1.4406252230687148e-06,
+      "loss": 1.3142,
+      "step": 3767
+    },
+    {
+      "epoch": 0.9161196207148067,
+      "grad_norm": 24.125,
+      "learning_rate": 1.440147598873223e-06,
+      "loss": 0.9144,
+      "step": 3768
+    },
+    {
+      "epoch": 0.9163627522489667,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4396699462568894e-06,
+      "loss": 0.5567,
+      "step": 3769
+    },
+    {
+      "epoch": 0.9166058837831267,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.439192265291106e-06,
+      "loss": 0.43,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9168490153172867,
+      "grad_norm": 16.5,
+      "learning_rate": 1.4387145560472712e-06,
+      "loss": 1.0951,
+      "step": 3771
+    },
+    {
+      "epoch": 0.9170921468514467,
+      "grad_norm": 20.75,
+      "learning_rate": 1.4382368185967868e-06,
+      "loss": 0.6864,
+      "step": 3772
+    },
+    {
+      "epoch": 0.9173352783856066,
+      "grad_norm": 21.25,
+      "learning_rate": 1.4377590530110591e-06,
+      "loss": 1.1724,
+      "step": 3773
+    },
+    {
+      "epoch": 0.9175784099197666,
+      "grad_norm": 19.75,
+      "learning_rate": 1.4372812593614983e-06,
+      "loss": 1.2391,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9178215414539266,
+      "grad_norm": 18.625,
+      "learning_rate": 1.4368034377195183e-06,
+      "loss": 0.8745,
+      "step": 3775
+    },
+    {
+      "epoch": 0.9180646729880866,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4363255881565389e-06,
+      "loss": 0.8104,
+      "step": 3776
+    },
+    {
+      "epoch": 0.9183078045222466,
+      "grad_norm": 20.625,
+      "learning_rate": 1.435847710743982e-06,
+      "loss": 0.8274,
+      "step": 3777
+    },
+    {
+      "epoch": 0.9185509360564065,
+      "grad_norm": 14.75,
+      "learning_rate": 1.435369805553275e-06,
+      "loss": 0.7391,
+      "step": 3778
+    },
+    {
+      "epoch": 0.9187940675905665,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4348918726558495e-06,
+      "loss": 1.0393,
+      "step": 3779
+    },
+    {
+      "epoch": 0.9190371991247265,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4344139121231402e-06,
+      "loss": 0.9398,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9192803306588865,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.433935924026587e-06,
+      "loss": 0.7216,
+      "step": 3781
+    },
+    {
+      "epoch": 0.9195234621930465,
+      "grad_norm": 18.0,
+      "learning_rate": 1.433457908437633e-06,
+      "loss": 0.9834,
+      "step": 3782
+    },
+    {
+      "epoch": 0.9197665937272064,
+      "grad_norm": 22.25,
+      "learning_rate": 1.432979865427726e-06,
+      "loss": 1.0621,
+      "step": 3783
+    },
+    {
+      "epoch": 0.9200097252613664,
+      "grad_norm": 18.625,
+      "learning_rate": 1.4325017950683182e-06,
+      "loss": 0.7211,
+      "step": 3784
+    },
+    {
+      "epoch": 0.9202528567955264,
+      "grad_norm": 25.875,
+      "learning_rate": 1.4320236974308652e-06,
+      "loss": 1.1432,
+      "step": 3785
+    },
+    {
+      "epoch": 0.9204959883296864,
+      "grad_norm": 21.125,
+      "learning_rate": 1.431545572586827e-06,
+      "loss": 0.7333,
+      "step": 3786
+    },
+    {
+      "epoch": 0.9207391198638464,
+      "grad_norm": 23.625,
+      "learning_rate": 1.4310674206076675e-06,
+      "loss": 1.1414,
+      "step": 3787
+    },
+    {
+      "epoch": 0.9209822513980063,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4305892415648549e-06,
+      "loss": 0.8259,
+      "step": 3788
+    },
+    {
+      "epoch": 0.9212253829321663,
+      "grad_norm": 19.5,
+      "learning_rate": 1.4301110355298612e-06,
+      "loss": 1.0309,
+      "step": 3789
+    },
+    {
+      "epoch": 0.9214685144663263,
+      "grad_norm": 17.125,
+      "learning_rate": 1.4296328025741626e-06,
+      "loss": 0.8037,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9217116460004863,
+      "grad_norm": 16.5,
+      "learning_rate": 1.4291545427692394e-06,
+      "loss": 1.0646,
+      "step": 3791
+    },
+    {
+      "epoch": 0.9219547775346463,
+      "grad_norm": 20.0,
+      "learning_rate": 1.4286762561865756e-06,
+      "loss": 0.8686,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9221979090688063,
+      "grad_norm": 28.125,
+      "learning_rate": 1.4281979428976594e-06,
+      "loss": 1.1802,
+      "step": 3793
+    },
+    {
+      "epoch": 0.9224410406029662,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.4277196029739831e-06,
+      "loss": 0.7507,
+      "step": 3794
+    },
+    {
+      "epoch": 0.9226841721371262,
+      "grad_norm": 18.5,
+      "learning_rate": 1.427241236487043e-06,
+      "loss": 1.3671,
+      "step": 3795
+    },
+    {
+      "epoch": 0.9229273036712862,
+      "grad_norm": 21.25,
+      "learning_rate": 1.4267628435083388e-06,
+      "loss": 0.7674,
+      "step": 3796
+    },
+    {
+      "epoch": 0.9231704352054462,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4262844241093749e-06,
+      "loss": 1.1878,
+      "step": 3797
+    },
+    {
+      "epoch": 0.9234135667396062,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.4258059783616596e-06,
+      "loss": 0.7436,
+      "step": 3798
+    },
+    {
+      "epoch": 0.9236566982737661,
+      "grad_norm": 20.875,
+      "learning_rate": 1.4253275063367038e-06,
+      "loss": 0.6861,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9238998298079261,
+      "grad_norm": 18.625,
+      "learning_rate": 1.4248490081060248e-06,
+      "loss": 1.0555,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9241429613420861,
+      "grad_norm": 15.25,
+      "learning_rate": 1.4243704837411418e-06,
+      "loss": 0.5213,
+      "step": 3801
+    },
+    {
+      "epoch": 0.9243860928762461,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4238919333135778e-06,
+      "loss": 0.7359,
+      "step": 3802
+    },
+    {
+      "epoch": 0.9246292244104061,
+      "grad_norm": 19.625,
+      "learning_rate": 1.423413356894862e-06,
+      "loss": 0.8722,
+      "step": 3803
+    },
+    {
+      "epoch": 0.924872355944566,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4229347545565248e-06,
+      "loss": 0.9464,
+      "step": 3804
+    },
+    {
+      "epoch": 0.925115487478726,
+      "grad_norm": 18.125,
+      "learning_rate": 1.422456126370102e-06,
+      "loss": 0.777,
+      "step": 3805
+    },
+    {
+      "epoch": 0.925358619012886,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4219774724071322e-06,
+      "loss": 0.8119,
+      "step": 3806
+    },
+    {
+      "epoch": 0.925601750547046,
+      "grad_norm": 16.25,
+      "learning_rate": 1.4214987927391594e-06,
+      "loss": 0.8855,
+      "step": 3807
+    },
+    {
+      "epoch": 0.925844882081206,
+      "grad_norm": 18.625,
+      "learning_rate": 1.42102008743773e-06,
+      "loss": 0.804,
+      "step": 3808
+    },
+    {
+      "epoch": 0.926088013615366,
+      "grad_norm": 16.625,
+      "learning_rate": 1.420541356574395e-06,
+      "loss": 0.4214,
+      "step": 3809
+    },
+    {
+      "epoch": 0.9263311451495259,
+      "grad_norm": 22.25,
+      "learning_rate": 1.4200626002207089e-06,
+      "loss": 0.8928,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9265742766836859,
+      "grad_norm": 36.5,
+      "learning_rate": 1.41958381844823e-06,
+      "loss": 1.06,
+      "step": 3811
+    },
+    {
+      "epoch": 0.9268174082178459,
+      "grad_norm": 23.0,
+      "learning_rate": 1.419105011328521e-06,
+      "loss": 1.194,
+      "step": 3812
+    },
+    {
+      "epoch": 0.9270605397520059,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4186261789331471e-06,
+      "loss": 0.8113,
+      "step": 3813
+    },
+    {
+      "epoch": 0.9273036712861659,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.4181473213336783e-06,
+      "loss": 0.7907,
+      "step": 3814
+    },
+    {
+      "epoch": 0.9275468028203258,
+      "grad_norm": 23.125,
+      "learning_rate": 1.4176684386016886e-06,
+      "loss": 0.9357,
+      "step": 3815
+    },
+    {
+      "epoch": 0.9277899343544858,
+      "grad_norm": 24.25,
+      "learning_rate": 1.417189530808755e-06,
+      "loss": 0.8202,
+      "step": 3816
+    },
+    {
+      "epoch": 0.9280330658886458,
+      "grad_norm": 22.375,
+      "learning_rate": 1.416710598026459e-06,
+      "loss": 0.9124,
+      "step": 3817
+    },
+    {
+      "epoch": 0.9282761974228058,
+      "grad_norm": 19.125,
+      "learning_rate": 1.416231640326384e-06,
+      "loss": 0.8411,
+      "step": 3818
+    },
+    {
+      "epoch": 0.9285193289569658,
+      "grad_norm": 16.5,
+      "learning_rate": 1.41575265778012e-06,
+      "loss": 0.5253,
+      "step": 3819
+    },
+    {
+      "epoch": 0.9287624604911257,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4152736504592587e-06,
+      "loss": 0.9743,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9290055920252857,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4147946184353958e-06,
+      "loss": 0.4579,
+      "step": 3821
+    },
+    {
+      "epoch": 0.9292487235594457,
+      "grad_norm": 23.25,
+      "learning_rate": 1.414315561780131e-06,
+      "loss": 0.8227,
+      "step": 3822
+    },
+    {
+      "epoch": 0.9294918550936057,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4138364805650679e-06,
+      "loss": 0.8655,
+      "step": 3823
+    },
+    {
+      "epoch": 0.9297349866277657,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4133573748618135e-06,
+      "loss": 0.9211,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9299781181619255,
+      "grad_norm": 19.0,
+      "learning_rate": 1.4128782447419775e-06,
+      "loss": 0.7798,
+      "step": 3825
+    },
+    {
+      "epoch": 0.9302212496960856,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4123990902771747e-06,
+      "loss": 1.0926,
+      "step": 3826
+    },
+    {
+      "epoch": 0.9304643812302456,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.411919911539024e-06,
+      "loss": 0.392,
+      "step": 3827
+    },
+    {
+      "epoch": 0.9307075127644056,
+      "grad_norm": 36.0,
+      "learning_rate": 1.4114407085991456e-06,
+      "loss": 0.8685,
+      "step": 3828
+    },
+    {
+      "epoch": 0.9309506442985656,
+      "grad_norm": 24.0,
+      "learning_rate": 1.4109614815291648e-06,
+      "loss": 0.9956,
+      "step": 3829
+    },
+    {
+      "epoch": 0.9311937758327256,
+      "grad_norm": 22.75,
+      "learning_rate": 1.410482230400711e-06,
+      "loss": 1.1699,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9314369073668854,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.410002955285416e-06,
+      "loss": 0.5616,
+      "step": 3831
+    },
+    {
+      "epoch": 0.9316800389010454,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4095236562549167e-06,
+      "loss": 0.9048,
+      "step": 3832
+    },
+    {
+      "epoch": 0.9319231704352055,
+      "grad_norm": 25.0,
+      "learning_rate": 1.409044333380851e-06,
+      "loss": 0.8924,
+      "step": 3833
+    },
+    {
+      "epoch": 0.9321663019693655,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4085649867348635e-06,
+      "loss": 0.9992,
+      "step": 3834
+    },
+    {
+      "epoch": 0.9324094335035255,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4080856163886001e-06,
+      "loss": 0.7677,
+      "step": 3835
+    },
+    {
+      "epoch": 0.9326525650376853,
+      "grad_norm": 21.5,
+      "learning_rate": 1.407606222413711e-06,
+      "loss": 0.9993,
+      "step": 3836
+    },
+    {
+      "epoch": 0.9328956965718453,
+      "grad_norm": 18.125,
+      "learning_rate": 1.4071268048818499e-06,
+      "loss": 0.9306,
+      "step": 3837
+    },
+    {
+      "epoch": 0.9331388281060053,
+      "grad_norm": 22.625,
+      "learning_rate": 1.406647363864674e-06,
+      "loss": 1.0934,
+      "step": 3838
+    },
+    {
+      "epoch": 0.9333819596401653,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4061678994338449e-06,
+      "loss": 0.744,
+      "step": 3839
+    },
+    {
+      "epoch": 0.9336250911743253,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4056884116610255e-06,
+      "loss": 1.2328,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9338682227084852,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4052089006178843e-06,
+      "loss": 0.9657,
+      "step": 3841
+    },
+    {
+      "epoch": 0.9341113542426452,
+      "grad_norm": 20.0,
+      "learning_rate": 1.4047293663760922e-06,
+      "loss": 0.7684,
+      "step": 3842
+    },
+    {
+      "epoch": 0.9343544857768052,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.4042498090073243e-06,
+      "loss": 0.5175,
+      "step": 3843
+    },
+    {
+      "epoch": 0.9345976173109652,
+      "grad_norm": 17.375,
+      "learning_rate": 1.4037702285832586e-06,
+      "loss": 0.9267,
+      "step": 3844
+    },
+    {
+      "epoch": 0.9348407488451252,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.403290625175576e-06,
+      "loss": 0.3068,
+      "step": 3845
+    },
+    {
+      "epoch": 0.9350838803792852,
+      "grad_norm": 21.25,
+      "learning_rate": 1.4028109988559624e-06,
+      "loss": 1.2203,
+      "step": 3846
+    },
+    {
+      "epoch": 0.9353270119134451,
+      "grad_norm": 23.75,
+      "learning_rate": 1.4023313496961059e-06,
+      "loss": 1.0322,
+      "step": 3847
+    },
+    {
+      "epoch": 0.9355701434476051,
+      "grad_norm": 18.875,
+      "learning_rate": 1.401851677767698e-06,
+      "loss": 0.6709,
+      "step": 3848
+    },
+    {
+      "epoch": 0.9358132749817651,
+      "grad_norm": 22.875,
+      "learning_rate": 1.4013719831424347e-06,
+      "loss": 0.9684,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9360564065159251,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4008922658920138e-06,
+      "loss": 0.8002,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9362995380500851,
+      "grad_norm": 20.125,
+      "learning_rate": 1.4004125260881383e-06,
+      "loss": 1.0745,
+      "step": 3851
+    },
+    {
+      "epoch": 0.936542669584245,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.3999327638025128e-06,
+      "loss": 0.4899,
+      "step": 3852
+    },
+    {
+      "epoch": 0.936785801118405,
+      "grad_norm": 17.375,
+      "learning_rate": 1.3994529791068461e-06,
+      "loss": 0.5899,
+      "step": 3853
+    },
+    {
+      "epoch": 0.937028932652565,
+      "grad_norm": 19.125,
+      "learning_rate": 1.398973172072851e-06,
+      "loss": 0.694,
+      "step": 3854
+    },
+    {
+      "epoch": 0.937272064186725,
+      "grad_norm": 24.5,
+      "learning_rate": 1.3984933427722419e-06,
+      "loss": 0.9549,
+      "step": 3855
+    },
+    {
+      "epoch": 0.937515195720885,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3980134912767385e-06,
+      "loss": 0.899,
+      "step": 3856
+    },
+    {
+      "epoch": 0.9377583272550449,
+      "grad_norm": 16.5,
+      "learning_rate": 1.3975336176580625e-06,
+      "loss": 0.9902,
+      "step": 3857
+    },
+    {
+      "epoch": 0.9380014587892049,
+      "grad_norm": 21.25,
+      "learning_rate": 1.3970537219879395e-06,
+      "loss": 0.5704,
+      "step": 3858
+    },
+    {
+      "epoch": 0.9382445903233649,
+      "grad_norm": 17.25,
+      "learning_rate": 1.396573804338098e-06,
+      "loss": 0.6479,
+      "step": 3859
+    },
+    {
+      "epoch": 0.9384877218575249,
+      "grad_norm": 16.625,
+      "learning_rate": 1.3960938647802699e-06,
+      "loss": 1.0088,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9387308533916849,
+      "grad_norm": 16.875,
+      "learning_rate": 1.3956139033861904e-06,
+      "loss": 0.8491,
+      "step": 3861
+    },
+    {
+      "epoch": 0.9389739849258448,
+      "grad_norm": 17.0,
+      "learning_rate": 1.3951339202275984e-06,
+      "loss": 0.7275,
+      "step": 3862
+    },
+    {
+      "epoch": 0.9392171164600048,
+      "grad_norm": 16.625,
+      "learning_rate": 1.394653915376236e-06,
+      "loss": 0.8649,
+      "step": 3863
+    },
+    {
+      "epoch": 0.9394602479941648,
+      "grad_norm": 19.125,
+      "learning_rate": 1.3941738889038467e-06,
+      "loss": 0.7899,
+      "step": 3864
+    },
+    {
+      "epoch": 0.9397033795283248,
+      "grad_norm": 24.875,
+      "learning_rate": 1.3936938408821804e-06,
+      "loss": 1.1052,
+      "step": 3865
+    },
+    {
+      "epoch": 0.9399465110624848,
+      "grad_norm": 16.125,
+      "learning_rate": 1.3932137713829878e-06,
+      "loss": 0.974,
+      "step": 3866
+    },
+    {
+      "epoch": 0.9401896425966448,
+      "grad_norm": 16.25,
+      "learning_rate": 1.3927336804780235e-06,
+      "loss": 0.5647,
+      "step": 3867
+    },
+    {
+      "epoch": 0.9404327741308047,
+      "grad_norm": 15.125,
+      "learning_rate": 1.3922535682390453e-06,
+      "loss": 0.678,
+      "step": 3868
+    },
+    {
+      "epoch": 0.9406759056649647,
+      "grad_norm": 20.625,
+      "learning_rate": 1.391773434737815e-06,
+      "loss": 0.7141,
+      "step": 3869
+    },
+    {
+      "epoch": 0.9409190371991247,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3912932800460965e-06,
+      "loss": 0.9183,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9411621687332847,
+      "grad_norm": 15.625,
+      "learning_rate": 1.3908131042356568e-06,
+      "loss": 0.6303,
+      "step": 3871
+    },
+    {
+      "epoch": 0.9414053002674447,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3903329073782668e-06,
+      "loss": 0.7484,
+      "step": 3872
+    },
+    {
+      "epoch": 0.9416484318016046,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3898526895457e-06,
+      "loss": 0.9081,
+      "step": 3873
+    },
+    {
+      "epoch": 0.9418915633357646,
+      "grad_norm": 20.5,
+      "learning_rate": 1.3893724508097334e-06,
+      "loss": 1.2226,
+      "step": 3874
+    },
+    {
+      "epoch": 0.9421346948699246,
+      "grad_norm": 17.75,
+      "learning_rate": 1.3888921912421473e-06,
+      "loss": 0.7871,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9423778264040846,
+      "grad_norm": 16.875,
+      "learning_rate": 1.388411910914724e-06,
+      "loss": 0.8181,
+      "step": 3876
+    },
+    {
+      "epoch": 0.9426209579382446,
+      "grad_norm": 22.625,
+      "learning_rate": 1.3879316098992507e-06,
+      "loss": 0.9265,
+      "step": 3877
+    },
+    {
+      "epoch": 0.9428640894724045,
+      "grad_norm": 15.875,
+      "learning_rate": 1.3874512882675156e-06,
+      "loss": 0.4322,
+      "step": 3878
+    },
+    {
+      "epoch": 0.9431072210065645,
+      "grad_norm": 17.75,
+      "learning_rate": 1.386970946091312e-06,
+      "loss": 0.6555,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9433503525407245,
+      "grad_norm": 20.125,
+      "learning_rate": 1.3864905834424348e-06,
+      "loss": 0.8817,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9435934840748845,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3860102003926827e-06,
+      "loss": 0.8455,
+      "step": 3881
+    },
+    {
+      "epoch": 0.9438366156090445,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3855297970138571e-06,
+      "loss": 0.6743,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9440797471432045,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3850493733777622e-06,
+      "loss": 1.0024,
+      "step": 3883
+    },
+    {
+      "epoch": 0.9443228786773644,
+      "grad_norm": 20.875,
+      "learning_rate": 1.384568929556207e-06,
+      "loss": 0.8406,
+      "step": 3884
+    },
+    {
+      "epoch": 0.9445660102115244,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3840884656210007e-06,
+      "loss": 1.0866,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9448091417456844,
+      "grad_norm": 17.875,
+      "learning_rate": 1.3836079816439575e-06,
+      "loss": 0.8666,
+      "step": 3886
+    },
+    {
+      "epoch": 0.9450522732798444,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3831274776968936e-06,
+      "loss": 0.9899,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9452954048140044,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3826469538516292e-06,
+      "loss": 1.2078,
+      "step": 3888
+    },
+    {
+      "epoch": 0.9455385363481643,
+      "grad_norm": 17.375,
+      "learning_rate": 1.382166410179987e-06,
+      "loss": 0.8049,
+      "step": 3889
+    },
+    {
+      "epoch": 0.9457816678823243,
+      "grad_norm": 24.125,
+      "learning_rate": 1.381685846753792e-06,
+      "loss": 0.9659,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9460247994164843,
+      "grad_norm": 16.25,
+      "learning_rate": 1.3812052636448728e-06,
+      "loss": 0.4699,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9462679309506443,
+      "grad_norm": 17.625,
+      "learning_rate": 1.380724660925061e-06,
+      "loss": 0.905,
+      "step": 3892
+    },
+    {
+      "epoch": 0.9465110624848043,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3802440386661908e-06,
+      "loss": 0.5628,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9467541940189642,
+      "grad_norm": 16.5,
+      "learning_rate": 1.3797633969401e-06,
+      "loss": 0.7007,
+      "step": 3894
+    },
+    {
+      "epoch": 0.9469973255531242,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3792827358186277e-06,
+      "loss": 1.1727,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9472404570872842,
+      "grad_norm": 16.125,
+      "learning_rate": 1.3788020553736186e-06,
+      "loss": 0.7185,
+      "step": 3896
+    },
+    {
+      "epoch": 0.9474835886214442,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.3783213556769177e-06,
+      "loss": 0.4945,
+      "step": 3897
+    },
+    {
+      "epoch": 0.9477267201556042,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3778406368003735e-06,
+      "loss": 0.5671,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9479698516897641,
+      "grad_norm": 25.25,
+      "learning_rate": 1.3773598988158386e-06,
+      "loss": 1.0576,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9482129832239241,
+      "grad_norm": 22.5,
+      "learning_rate": 1.3768791417951671e-06,
+      "loss": 0.6711,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9484561147580841,
+      "grad_norm": 20.0,
+      "learning_rate": 1.3763983658102168e-06,
+      "loss": 0.7188,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9486992462922441,
+      "grad_norm": 23.375,
+      "learning_rate": 1.3759175709328476e-06,
+      "loss": 1.1597,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9489423778264041,
+      "grad_norm": 18.625,
+      "learning_rate": 1.3754367572349225e-06,
+      "loss": 0.7496,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9491855093605641,
+      "grad_norm": 20.875,
+      "learning_rate": 1.374955924788308e-06,
+      "loss": 0.664,
+      "step": 3904
+    },
+    {
+      "epoch": 0.949428640894724,
+      "grad_norm": 17.25,
+      "learning_rate": 1.3744750736648724e-06,
+      "loss": 0.7426,
+      "step": 3905
+    },
+    {
+      "epoch": 0.949671772428884,
+      "grad_norm": 15.875,
+      "learning_rate": 1.3739942039364876e-06,
+      "loss": 0.5051,
+      "step": 3906
+    },
+    {
+      "epoch": 0.949914903963044,
+      "grad_norm": 22.625,
+      "learning_rate": 1.3735133156750268e-06,
+      "loss": 1.3829,
+      "step": 3907
+    },
+    {
+      "epoch": 0.950158035497204,
+      "grad_norm": 23.125,
+      "learning_rate": 1.3730324089523683e-06,
+      "loss": 0.8434,
+      "step": 3908
+    },
+    {
+      "epoch": 0.950401167031364,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3725514838403914e-06,
+      "loss": 0.5742,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9506442985655239,
+      "grad_norm": 23.0,
+      "learning_rate": 1.3720705404109787e-06,
+      "loss": 1.0885,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9508874300996839,
+      "grad_norm": 25.375,
+      "learning_rate": 1.3715895787360155e-06,
+      "loss": 1.2471,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9511305616338439,
+      "grad_norm": 19.25,
+      "learning_rate": 1.3711085988873898e-06,
+      "loss": 0.6364,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9513736931680039,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3706276009369925e-06,
+      "loss": 0.5563,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9516168247021639,
+      "grad_norm": 17.75,
+      "learning_rate": 1.3701465849567167e-06,
+      "loss": 0.7277,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9518599562363238,
+      "grad_norm": 30.875,
+      "learning_rate": 1.3696655510184592e-06,
+      "loss": 1.1632,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9521030877704838,
+      "grad_norm": 16.375,
+      "learning_rate": 1.369184499194118e-06,
+      "loss": 0.7922,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9523462193046438,
+      "grad_norm": 22.125,
+      "learning_rate": 1.3687034295555951e-06,
+      "loss": 1.055,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9525893508388038,
+      "grad_norm": 21.375,
+      "learning_rate": 1.3682223421747948e-06,
+      "loss": 0.8957,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9528324823729638,
+      "grad_norm": 27.625,
+      "learning_rate": 1.3677412371236232e-06,
+      "loss": 0.9606,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9530756139071238,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.367260114473991e-06,
+      "loss": 0.4122,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9533187454412837,
+      "grad_norm": 17.0,
+      "learning_rate": 1.3667789742978089e-06,
+      "loss": 0.7189,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9535618769754437,
+      "grad_norm": 16.75,
+      "learning_rate": 1.3662978166669924e-06,
+      "loss": 0.6071,
+      "step": 3922
+    },
+    {
+      "epoch": 0.9538050085096037,
+      "grad_norm": 22.0,
+      "learning_rate": 1.3658166416534588e-06,
+      "loss": 0.8416,
+      "step": 3923
+    },
+    {
+      "epoch": 0.9540481400437637,
+      "grad_norm": 20.25,
+      "learning_rate": 1.3653354493291276e-06,
+      "loss": 0.9223,
+      "step": 3924
+    },
+    {
+      "epoch": 0.9542912715779237,
+      "grad_norm": 14.75,
+      "learning_rate": 1.364854239765922e-06,
+      "loss": 0.494,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9545344031120836,
+      "grad_norm": 21.875,
+      "learning_rate": 1.3643730130357662e-06,
+      "loss": 0.9353,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9547775346462436,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3638917692105888e-06,
+      "loss": 0.7108,
+      "step": 3927
+    },
+    {
+      "epoch": 0.9550206661804036,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3634105083623191e-06,
+      "loss": 0.9413,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9552637977145636,
+      "grad_norm": 34.0,
+      "learning_rate": 1.3629292305628905e-06,
+      "loss": 0.3229,
+      "step": 3929
+    },
+    {
+      "epoch": 0.9555069292487236,
+      "grad_norm": 18.5,
+      "learning_rate": 1.362447935884238e-06,
+      "loss": 1.3861,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9557500607828835,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3619666243982993e-06,
+      "loss": 0.6782,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9559931923170435,
+      "grad_norm": 16.875,
+      "learning_rate": 1.361485296177015e-06,
+      "loss": 1.1356,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9562363238512035,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3610039512923278e-06,
+      "loss": 0.9059,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9564794553853635,
+      "grad_norm": 18.25,
+      "learning_rate": 1.3605225898161828e-06,
+      "loss": 0.7756,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9567225869195235,
+      "grad_norm": 17.125,
+      "learning_rate": 1.360041211820528e-06,
+      "loss": 0.6717,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9569657184536834,
+      "grad_norm": 16.0,
+      "learning_rate": 1.3595598173773137e-06,
+      "loss": 0.621,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9572088499878434,
+      "grad_norm": 36.25,
+      "learning_rate": 1.3590784065584927e-06,
+      "loss": 1.2804,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9574519815220034,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3585969794360197e-06,
+      "loss": 0.6779,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9576951130561634,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3581155360818526e-06,
+      "loss": 1.0433,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9579382445903234,
+      "grad_norm": 20.5,
+      "learning_rate": 1.3576340765679516e-06,
+      "loss": 1.1591,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9581813761244834,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.3571526009662784e-06,
+      "loss": 0.6295,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9584245076586433,
+      "grad_norm": 15.625,
+      "learning_rate": 1.356671109348799e-06,
+      "loss": 0.7017,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9586676391928033,
+      "grad_norm": 23.875,
+      "learning_rate": 1.3561896017874799e-06,
+      "loss": 0.8728,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9589107707269633,
+      "grad_norm": 41.75,
+      "learning_rate": 1.355708078354291e-06,
+      "loss": 0.951,
+      "step": 3944
+    },
+    {
+      "epoch": 0.9591539022611233,
+      "grad_norm": 23.125,
+      "learning_rate": 1.3552265391212038e-06,
+      "loss": 1.0707,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9593970337952833,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3547449841601935e-06,
+      "loss": 1.3283,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9596401653294432,
+      "grad_norm": 21.0,
+      "learning_rate": 1.354263413543236e-06,
+      "loss": 0.6235,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9598832968636032,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3537818273423103e-06,
+      "loss": 0.6245,
+      "step": 3948
+    },
+    {
+      "epoch": 0.9601264283977632,
+      "grad_norm": 25.5,
+      "learning_rate": 1.3533002256293987e-06,
+      "loss": 0.8943,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9603695599319232,
+      "grad_norm": 19.75,
+      "learning_rate": 1.352818608476484e-06,
+      "loss": 0.9066,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9606126914660832,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3523369759555526e-06,
+      "loss": 0.6345,
+      "step": 3951
+    },
+    {
+      "epoch": 0.9608558230002431,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3518553281385929e-06,
+      "loss": 1.0054,
+      "step": 3952
+    },
+    {
+      "epoch": 0.9610989545344031,
+      "grad_norm": 18.625,
+      "learning_rate": 1.3513736650975947e-06,
+      "loss": 0.3696,
+      "step": 3953
+    },
+    {
+      "epoch": 0.9613420860685631,
+      "grad_norm": 21.25,
+      "learning_rate": 1.3508919869045522e-06,
+      "loss": 0.9926,
+      "step": 3954
+    },
+    {
+      "epoch": 0.9615852176027231,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3504102936314594e-06,
+      "loss": 1.0307,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9618283491368831,
+      "grad_norm": 25.75,
+      "learning_rate": 1.3499285853503146e-06,
+      "loss": 1.0777,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9620714806710431,
+      "grad_norm": 17.25,
+      "learning_rate": 1.349446862133116e-06,
+      "loss": 0.9713,
+      "step": 3957
+    },
+    {
+      "epoch": 0.962314612205203,
+      "grad_norm": 19.75,
+      "learning_rate": 1.348965124051867e-06,
+      "loss": 0.9016,
+      "step": 3958
+    },
+    {
+      "epoch": 0.962557743739363,
+      "grad_norm": 19.375,
+      "learning_rate": 1.348483371178571e-06,
+      "loss": 0.6962,
+      "step": 3959
+    },
+    {
+      "epoch": 0.962800875273523,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3480016035852342e-06,
+      "loss": 1.2241,
+      "step": 3960
+    },
+    {
+      "epoch": 0.963044006807683,
+      "grad_norm": 18.125,
+      "learning_rate": 1.3475198213438651e-06,
+      "loss": 0.7728,
+      "step": 3961
+    },
+    {
+      "epoch": 0.963287138341843,
+      "grad_norm": 21.375,
+      "learning_rate": 1.3470380245264744e-06,
+      "loss": 0.9239,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9635302698760029,
+      "grad_norm": 26.75,
+      "learning_rate": 1.3465562132050752e-06,
+      "loss": 0.9932,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9637734014101629,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3460743874516823e-06,
+      "loss": 0.8634,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9640165329443229,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3455925473383128e-06,
+      "loss": 1.2086,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9642596644784829,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3451106929369864e-06,
+      "loss": 0.6385,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9645027960126429,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.3446288243197242e-06,
+      "loss": 0.7051,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9647459275468028,
+      "grad_norm": 24.875,
+      "learning_rate": 1.3441469415585501e-06,
+      "loss": 1.1945,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9649890590809628,
+      "grad_norm": 19.5,
+      "learning_rate": 1.3436650447254892e-06,
+      "loss": 0.7287,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9652321906151228,
+      "grad_norm": 16.875,
+      "learning_rate": 1.3431831338925699e-06,
+      "loss": 0.6114,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9654753221492828,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.3427012091318224e-06,
+      "loss": 0.8383,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9657184536834428,
+      "grad_norm": 17.875,
+      "learning_rate": 1.3422192705152773e-06,
+      "loss": 0.697,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9659615852176027,
+      "grad_norm": 26.0,
+      "learning_rate": 1.3417373181149704e-06,
+      "loss": 0.8018,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9662047167517627,
+      "grad_norm": 26.875,
+      "learning_rate": 1.3412553520029365e-06,
+      "loss": 1.0745,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9664478482859227,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3407733722512144e-06,
+      "loss": 1.014,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9666909798200827,
+      "grad_norm": 29.125,
+      "learning_rate": 1.3402913789318436e-06,
+      "loss": 1.9196,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9669341113542427,
+      "grad_norm": 25.875,
+      "learning_rate": 1.3398093721168672e-06,
+      "loss": 0.949,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9671772428884027,
+      "grad_norm": 23.5,
+      "learning_rate": 1.3393273518783292e-06,
+      "loss": 1.3465,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9674203744225626,
+      "grad_norm": 27.5,
+      "learning_rate": 1.3388453182882757e-06,
+      "loss": 0.9278,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9676635059567226,
+      "grad_norm": 40.5,
+      "learning_rate": 1.3383632714187547e-06,
+      "loss": 1.2588,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9679066374908826,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3378812113418168e-06,
+      "loss": 0.7095,
+      "step": 3981
+    },
+    {
+      "epoch": 0.9681497690250426,
+      "grad_norm": 20.125,
+      "learning_rate": 1.3373991381295142e-06,
+      "loss": 1.1938,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9683929005592026,
+      "grad_norm": 19.875,
+      "learning_rate": 1.3369170518539013e-06,
+      "loss": 0.9127,
+      "step": 3983
+    },
+    {
+      "epoch": 0.9686360320933625,
+      "grad_norm": 15.5,
+      "learning_rate": 1.3364349525870332e-06,
+      "loss": 0.596,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9688791636275225,
+      "grad_norm": 18.125,
+      "learning_rate": 1.3359528404009691e-06,
+      "loss": 0.7775,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9691222951616825,
+      "grad_norm": 24.5,
+      "learning_rate": 1.3354707153677685e-06,
+      "loss": 0.8741,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9693654266958425,
+      "grad_norm": 18.25,
+      "learning_rate": 1.334988577559493e-06,
+      "loss": 0.9123,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9696085582300025,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.3345064270482072e-06,
+      "loss": 0.5815,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9698516897641624,
+      "grad_norm": 16.625,
+      "learning_rate": 1.3340242639059764e-06,
+      "loss": 0.8815,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9700948212983224,
+      "grad_norm": 14.625,
+      "learning_rate": 1.3335420882048683e-06,
+      "loss": 0.9273,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9703379528324824,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3330599000169519e-06,
+      "loss": 0.6381,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9705810843666424,
+      "grad_norm": 36.5,
+      "learning_rate": 1.3325776994142991e-06,
+      "loss": 1.6005,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9708242159008024,
+      "grad_norm": 19.75,
+      "learning_rate": 1.3320954864689831e-06,
+      "loss": 0.9846,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9710673474349624,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3316132612530786e-06,
+      "loss": 0.9429,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9713104789691223,
+      "grad_norm": 22.625,
+      "learning_rate": 1.3311310238386626e-06,
+      "loss": 0.8868,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9715536105032823,
+      "grad_norm": 18.875,
+      "learning_rate": 1.3306487742978142e-06,
+      "loss": 0.6734,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9717967420374423,
+      "grad_norm": 21.75,
+      "learning_rate": 1.3301665127026137e-06,
+      "loss": 1.0677,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9720398735716023,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.329684239125143e-06,
+      "loss": 0.7513,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9722830051057623,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3292019536374866e-06,
+      "loss": 1.0038,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9725261366399222,
+      "grad_norm": 16.875,
+      "learning_rate": 1.3287196563117308e-06,
+      "loss": 0.9442,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9727692681740822,
+      "grad_norm": 27.625,
+      "learning_rate": 1.3282373472199623e-06,
+      "loss": 0.8517,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9730123997082422,
+      "grad_norm": 19.75,
+      "learning_rate": 1.3277550264342714e-06,
+      "loss": 1.0273,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9732555312424022,
+      "grad_norm": 18.125,
+      "learning_rate": 1.3272726940267485e-06,
+      "loss": 0.5411,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9734986627765622,
+      "grad_norm": 19.875,
+      "learning_rate": 1.3267903500694875e-06,
+      "loss": 0.998,
+      "step": 4004
+    },
+    {
+      "epoch": 0.973741794310722,
+      "grad_norm": 24.125,
+      "learning_rate": 1.3263079946345822e-06,
+      "loss": 0.9674,
+      "step": 4005
+    },
+    {
+      "epoch": 0.973984925844882,
+      "grad_norm": 21.875,
+      "learning_rate": 1.3258256277941291e-06,
+      "loss": 1.068,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9742280573790421,
+      "grad_norm": 15.875,
+      "learning_rate": 1.3253432496202267e-06,
+      "loss": 0.667,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9744711889132021,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.3248608601849741e-06,
+      "loss": 0.3742,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9747143204473621,
+      "grad_norm": 20.875,
+      "learning_rate": 1.3243784595604733e-06,
+      "loss": 0.6048,
+      "step": 4009
+    },
+    {
+      "epoch": 0.974957451981522,
+      "grad_norm": 22.75,
+      "learning_rate": 1.323896047818827e-06,
+      "loss": 1.2635,
+      "step": 4010
+    },
+    {
+      "epoch": 0.975200583515682,
+      "grad_norm": 17.5,
+      "learning_rate": 1.3234136250321403e-06,
+      "loss": 0.6713,
+      "step": 4011
+    },
+    {
+      "epoch": 0.975443715049842,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3229311912725193e-06,
+      "loss": 0.5918,
+      "step": 4012
+    },
+    {
+      "epoch": 0.975686846584002,
+      "grad_norm": 16.25,
+      "learning_rate": 1.322448746612072e-06,
+      "loss": 0.7757,
+      "step": 4013
+    },
+    {
+      "epoch": 0.975929978118162,
+      "grad_norm": 20.0,
+      "learning_rate": 1.321966291122909e-06,
+      "loss": 0.7835,
+      "step": 4014
+    },
+    {
+      "epoch": 0.976173109652322,
+      "grad_norm": 17.5,
+      "learning_rate": 1.3214838248771396e-06,
+      "loss": 1.0936,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9764162411864818,
+      "grad_norm": 21.0,
+      "learning_rate": 1.3210013479468791e-06,
+      "loss": 0.9025,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9766593727206418,
+      "grad_norm": 23.5,
+      "learning_rate": 1.3205188604042407e-06,
+      "loss": 1.0179,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9769025042548019,
+      "grad_norm": 19.125,
+      "learning_rate": 1.3200363623213406e-06,
+      "loss": 0.9663,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9771456357889619,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3195538537702965e-06,
+      "loss": 0.9284,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9773887673231219,
+      "grad_norm": 16.5,
+      "learning_rate": 1.319071334823228e-06,
+      "loss": 0.912,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9776318988572817,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3185888055522556e-06,
+      "loss": 0.7986,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9778750303914417,
+      "grad_norm": 24.25,
+      "learning_rate": 1.3181062660295013e-06,
+      "loss": 1.0828,
+      "step": 4022
+    },
+    {
+      "epoch": 0.9781181619256017,
+      "grad_norm": 25.0,
+      "learning_rate": 1.3176237163270893e-06,
+      "loss": 0.8634,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9783612934597617,
+      "grad_norm": 14.75,
+      "learning_rate": 1.3171411565171452e-06,
+      "loss": 0.7959,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9786044249939218,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3166585866717953e-06,
+      "loss": 0.7384,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9788475565280816,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3161760068631691e-06,
+      "loss": 0.7498,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9790906880622416,
+      "grad_norm": 22.0,
+      "learning_rate": 1.315693417163395e-06,
+      "loss": 0.7234,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9793338195964016,
+      "grad_norm": 17.0,
+      "learning_rate": 1.315210817644606e-06,
+      "loss": 0.9886,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9795769511305616,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3147282083789337e-06,
+      "loss": 0.9663,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9798200826647216,
+      "grad_norm": 15.5,
+      "learning_rate": 1.3142455894385125e-06,
+      "loss": 0.4486,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9800632141988816,
+      "grad_norm": 15.75,
+      "learning_rate": 1.3137629608954785e-06,
+      "loss": 0.8873,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9803063457330415,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3132803228219688e-06,
+      "loss": 0.794,
+      "step": 4032
+    },
+    {
+      "epoch": 0.9805494772672015,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.3127976752901222e-06,
+      "loss": 0.6404,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9807926088013615,
+      "grad_norm": 17.125,
+      "learning_rate": 1.312315018372078e-06,
+      "loss": 0.9471,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9810357403355215,
+      "grad_norm": 21.0,
+      "learning_rate": 1.3118323521399787e-06,
+      "loss": 1.1323,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9812788718696815,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.3113496766659661e-06,
+      "loss": 0.3732,
+      "step": 4036
+    },
+    {
+      "epoch": 0.9815220034038414,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3108669920221848e-06,
+      "loss": 0.6477,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9817651349380014,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3103842982807802e-06,
+      "loss": 0.7884,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9820082664721614,
+      "grad_norm": 22.25,
+      "learning_rate": 1.3099015955138997e-06,
+      "loss": 0.7812,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9822513980063214,
+      "grad_norm": 17.25,
+      "learning_rate": 1.3094188837936912e-06,
+      "loss": 0.9805,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9824945295404814,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3089361631923043e-06,
+      "loss": 0.8195,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9827376610746413,
+      "grad_norm": 21.75,
+      "learning_rate": 1.3084534337818896e-06,
+      "loss": 1.1097,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9829807926088013,
+      "grad_norm": 28.125,
+      "learning_rate": 1.3079706956345997e-06,
+      "loss": 1.3946,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9832239241429613,
+      "grad_norm": 23.125,
+      "learning_rate": 1.3074879488225883e-06,
+      "loss": 0.8835,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9834670556771213,
+      "grad_norm": 23.125,
+      "learning_rate": 1.3070051934180106e-06,
+      "loss": 0.9037,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9837101872112813,
+      "grad_norm": 17.0,
+      "learning_rate": 1.3065224294930213e-06,
+      "loss": 0.6628,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9839533187454412,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3060396571197794e-06,
+      "loss": 0.8511,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9841964502796012,
+      "grad_norm": 17.75,
+      "learning_rate": 1.3055568763704425e-06,
+      "loss": 0.9845,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9844395818137612,
+      "grad_norm": 17.5,
+      "learning_rate": 1.3050740873171714e-06,
+      "loss": 0.873,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9846827133479212,
+      "grad_norm": 23.375,
+      "learning_rate": 1.3045912900321264e-06,
+      "loss": 1.1688,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9849258448820812,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3041084845874705e-06,
+      "loss": 0.71,
+      "step": 4051
+    },
+    {
+      "epoch": 0.9851689764162412,
+      "grad_norm": 19.5,
+      "learning_rate": 1.303625671055367e-06,
+      "loss": 0.6832,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9854121079504011,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3031428495079807e-06,
+      "loss": 0.6501,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9856552394845611,
+      "grad_norm": 31.5,
+      "learning_rate": 1.302660020017478e-06,
+      "loss": 0.9338,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9858983710187211,
+      "grad_norm": 24.875,
+      "learning_rate": 1.3021771826560256e-06,
+      "loss": 0.9302,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9861415025528811,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.3016943374957922e-06,
+      "loss": 1.0072,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9863846340870411,
+      "grad_norm": 20.875,
+      "learning_rate": 1.301211484608947e-06,
+      "loss": 0.7759,
+      "step": 4057
+    },
+    {
+      "epoch": 0.986627765621201,
+      "grad_norm": 20.5,
+      "learning_rate": 1.3007286240676614e-06,
+      "loss": 0.9501,
+      "step": 4058
+    },
+    {
+      "epoch": 0.986870897155361,
+      "grad_norm": 12.625,
+      "learning_rate": 1.300245755944107e-06,
+      "loss": 0.356,
+      "step": 4059
+    },
+    {
+      "epoch": 0.987114028689521,
+      "grad_norm": 19.75,
+      "learning_rate": 1.2997628803104563e-06,
+      "loss": 0.8057,
+      "step": 4060
+    },
+    {
+      "epoch": 0.987357160223681,
+      "grad_norm": 29.75,
+      "learning_rate": 1.2992799972388836e-06,
+      "loss": 0.9597,
+      "step": 4061
+    },
+    {
+      "epoch": 0.987600291757841,
+      "grad_norm": 18.75,
+      "learning_rate": 1.2987971068015643e-06,
+      "loss": 0.9781,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9878434232920009,
+      "grad_norm": 14.625,
+      "learning_rate": 1.2983142090706744e-06,
+      "loss": 0.5611,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9880865548261609,
+      "grad_norm": 22.125,
+      "learning_rate": 1.297831304118392e-06,
+      "loss": 1.1952,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9883296863603209,
+      "grad_norm": 25.75,
+      "learning_rate": 1.2973483920168948e-06,
+      "loss": 1.1019,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9885728178944809,
+      "grad_norm": 18.375,
+      "learning_rate": 1.2968654728383629e-06,
+      "loss": 0.9886,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9888159494286409,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2963825466549765e-06,
+      "loss": 0.7126,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9890590809628009,
+      "grad_norm": 18.875,
+      "learning_rate": 1.2958996135389174e-06,
+      "loss": 0.6975,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9893022124969608,
+      "grad_norm": 28.125,
+      "learning_rate": 1.2954166735623682e-06,
+      "loss": 1.2519,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9895453440311208,
+      "grad_norm": 16.625,
+      "learning_rate": 1.294933726797513e-06,
+      "loss": 0.7895,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9897884755652808,
+      "grad_norm": 14.5,
+      "learning_rate": 1.2944507733165367e-06,
+      "loss": 0.8475,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9900316070994408,
+      "grad_norm": 63.5,
+      "learning_rate": 1.293967813191624e-06,
+      "loss": 1.0635,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9902747386336008,
+      "grad_norm": 20.875,
+      "learning_rate": 1.2934848464949625e-06,
+      "loss": 0.9305,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9905178701677607,
+      "grad_norm": 20.25,
+      "learning_rate": 1.29300187329874e-06,
+      "loss": 0.7422,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9907610017019207,
+      "grad_norm": 17.5,
+      "learning_rate": 1.2925188936751443e-06,
+      "loss": 1.0064,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9910041332360807,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2920359076963663e-06,
+      "loss": 0.786,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9912472647702407,
+      "grad_norm": 25.125,
+      "learning_rate": 1.291552915434595e-06,
+      "loss": 0.7397,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9914903963044007,
+      "grad_norm": 17.375,
+      "learning_rate": 1.2910699169620235e-06,
+      "loss": 0.6564,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9917335278385606,
+      "grad_norm": 20.375,
+      "learning_rate": 1.2905869123508435e-06,
+      "loss": 0.6919,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9919766593727206,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.290103901673248e-06,
+      "loss": 0.3727,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9922197909068806,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2896208850014325e-06,
+      "loss": 0.8732,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9924629224410406,
+      "grad_norm": 19.625,
+      "learning_rate": 1.2891378624075912e-06,
+      "loss": 0.8371,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9927060539752006,
+      "grad_norm": 17.0,
+      "learning_rate": 1.2886548339639205e-06,
+      "loss": 0.646,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9929491855093605,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.288171799742617e-06,
+      "loss": 0.6926,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9931923170435205,
+      "grad_norm": 18.375,
+      "learning_rate": 1.287688759815879e-06,
+      "loss": 0.4883,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9934354485776805,
+      "grad_norm": 22.5,
+      "learning_rate": 1.2872057142559049e-06,
+      "loss": 0.7492,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9936785801118405,
+      "grad_norm": 18.5,
+      "learning_rate": 1.2867226631348943e-06,
+      "loss": 0.9716,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9939217116460005,
+      "grad_norm": 20.0,
+      "learning_rate": 1.2862396065250473e-06,
+      "loss": 1.0586,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9941648431801605,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.285756544498565e-06,
+      "loss": 0.5721,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9944079747143204,
+      "grad_norm": 22.375,
+      "learning_rate": 1.2852734771276504e-06,
+      "loss": 0.9043,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9946511062484804,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.284790404484505e-06,
+      "loss": 0.4759,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9948942377826404,
+      "grad_norm": 21.0,
+      "learning_rate": 1.2843073266413323e-06,
+      "loss": 1.0459,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9951373693168004,
+      "grad_norm": 20.875,
+      "learning_rate": 1.2838242436703377e-06,
+      "loss": 1.2038,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9953805008509604,
+      "grad_norm": 23.375,
+      "learning_rate": 1.2833411556437255e-06,
+      "loss": 0.9746,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9956236323851203,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.2828580626337024e-06,
+      "loss": 0.5975,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9958667639192803,
+      "grad_norm": 19.875,
+      "learning_rate": 1.2823749647124733e-06,
+      "loss": 0.7218,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9961098954534403,
+      "grad_norm": 16.875,
+      "learning_rate": 1.2818918619522471e-06,
+      "loss": 0.5233,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9963530269876003,
+      "grad_norm": 21.5,
+      "learning_rate": 1.2814087544252316e-06,
+      "loss": 0.8869,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9965961585217603,
+      "grad_norm": 27.875,
+      "learning_rate": 1.2809256422036351e-06,
+      "loss": 0.9395,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9968392900559202,
+      "grad_norm": 20.375,
+      "learning_rate": 1.2804425253596672e-06,
+      "loss": 0.9283,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9970824215900802,
+      "grad_norm": 18.0,
+      "learning_rate": 1.279959403965538e-06,
+      "loss": 0.5885,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9973255531242402,
+      "grad_norm": 21.125,
+      "learning_rate": 1.2794762780934588e-06,
+      "loss": 0.929,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9975686846584002,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2789931478156406e-06,
+      "loss": 0.8388,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9978118161925602,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2785100132042954e-06,
+      "loss": 0.6561,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9980549477267202,
+      "grad_norm": 18.125,
+      "learning_rate": 1.2780268743316369e-06,
+      "loss": 0.4373,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9982980792608801,
+      "grad_norm": 25.625,
+      "learning_rate": 1.2775437312698776e-06,
+      "loss": 1.1276,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9985412107950401,
+      "grad_norm": 23.25,
+      "learning_rate": 1.277060584091232e-06,
+      "loss": 1.1631,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9987843423292001,
+      "grad_norm": 13.0,
+      "learning_rate": 1.2765774328679147e-06,
+      "loss": 0.3814,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9990274738633601,
+      "grad_norm": 21.875,
+      "learning_rate": 1.2760942776721414e-06,
+      "loss": 1.1216,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9992706053975201,
+      "grad_norm": 15.75,
+      "learning_rate": 1.2756111185761277e-06,
+      "loss": 0.5756,
+      "step": 4110
+    },
+    {
+      "epoch": 0.99951373693168,
+      "grad_norm": 17.75,
+      "learning_rate": 1.2751279556520893e-06,
+      "loss": 0.7563,
+      "step": 4111
+    },
+    {
+      "epoch": 0.99975686846584,
+      "grad_norm": 16.375,
+      "learning_rate": 1.2746447889722446e-06,
+      "loss": 0.8834,
+      "step": 4112
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2741616186088103e-06,
+      "loss": 0.9585,
+      "step": 4113
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.1157740354537964,
+      "eval_runtime": 98.4299,
+      "eval_samples_per_second": 5.689,
+      "eval_steps_per_second": 5.689,
+      "step": 4113
+    },
+    {
+      "epoch": 1.00024313153416,
+      "grad_norm": 16.375,
+      "learning_rate": 1.273678444634005e-06,
+      "loss": 0.6687,
+      "step": 4114
     }
   ],
   "logging_steps": 1,
@@ -14425,7 +28832,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.157726307530834e+18,
+  "total_flos": 2.315452615061668e+18,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null