diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,26454 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 7512,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007987220447284345,
+      "grad_norm": 63.611212563977595,
+      "learning_rate": 2.6595744680851065e-08,
+      "loss": 1.2894,
+      "step": 2
+    },
+    {
+      "epoch": 0.001597444089456869,
+      "grad_norm": 105.73905369340336,
+      "learning_rate": 5.319148936170213e-08,
+      "loss": 1.1515,
+      "step": 4
+    },
+    {
+      "epoch": 0.0023961661341853034,
+      "grad_norm": 97.48653856616731,
+      "learning_rate": 7.978723404255319e-08,
+      "loss": 1.1725,
+      "step": 6
+    },
+    {
+      "epoch": 0.003194888178913738,
+      "grad_norm": 113.1691721809818,
+      "learning_rate": 1.0638297872340426e-07,
+      "loss": 1.2009,
+      "step": 8
+    },
+    {
+      "epoch": 0.003993610223642172,
+      "grad_norm": 234.6545065934056,
+      "learning_rate": 1.3297872340425533e-07,
+      "loss": 1.2124,
+      "step": 10
+    },
+    {
+      "epoch": 0.004792332268370607,
+      "grad_norm": 56.50421981052306,
+      "learning_rate": 1.5957446808510638e-07,
+      "loss": 1.2292,
+      "step": 12
+    },
+    {
+      "epoch": 0.005591054313099041,
+      "grad_norm": 67.76208027590963,
+      "learning_rate": 1.8617021276595745e-07,
+      "loss": 1.1731,
+      "step": 14
+    },
+    {
+      "epoch": 0.006389776357827476,
+      "grad_norm": 19.773545305324223,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 1.1682,
+      "step": 16
+    },
+    {
+      "epoch": 0.00718849840255591,
+      "grad_norm": 108.48956883104074,
+      "learning_rate": 2.393617021276596e-07,
+      "loss": 1.1885,
+      "step": 18
+    },
+    {
+      "epoch": 0.007987220447284345,
+      "grad_norm": 74.594447029578,
+      "learning_rate": 2.6595744680851066e-07,
+      "loss": 1.1143,
+      "step": 20
+    },
+    {
+      "epoch": 0.00878594249201278,
+      "grad_norm": 193.76064469208063,
+      "learning_rate": 2.9255319148936174e-07,
+      "loss": 1.1366,
+      "step": 22
+    },
+    {
+      "epoch": 0.009584664536741214,
+      "grad_norm": 165.84272450497056,
+      "learning_rate": 3.1914893617021275e-07,
+      "loss": 1.1753,
+      "step": 24
+    },
+    {
+      "epoch": 0.010383386581469648,
+      "grad_norm": 150.10218395823307,
+      "learning_rate": 3.457446808510639e-07,
+      "loss": 1.1228,
+      "step": 26
+    },
+    {
+      "epoch": 0.011182108626198083,
+      "grad_norm": 25.354463533053462,
+      "learning_rate": 3.723404255319149e-07,
+      "loss": 1.138,
+      "step": 28
+    },
+    {
+      "epoch": 0.011980830670926517,
+      "grad_norm": 106.72660838009219,
+      "learning_rate": 3.98936170212766e-07,
+      "loss": 1.1287,
+      "step": 30
+    },
+    {
+      "epoch": 0.012779552715654952,
+      "grad_norm": 83.97368303722932,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 1.0818,
+      "step": 32
+    },
+    {
+      "epoch": 0.013578274760383386,
+      "grad_norm": 72.42974677624225,
+      "learning_rate": 4.5212765957446816e-07,
+      "loss": 0.974,
+      "step": 34
+    },
+    {
+      "epoch": 0.01437699680511182,
+      "grad_norm": 98.949142462491,
+      "learning_rate": 4.787234042553192e-07,
+      "loss": 1.0153,
+      "step": 36
+    },
+    {
+      "epoch": 0.015175718849840255,
+      "grad_norm": 81.09462884756049,
+      "learning_rate": 5.053191489361702e-07,
+      "loss": 1.0437,
+      "step": 38
+    },
+    {
+      "epoch": 0.01597444089456869,
+      "grad_norm": 71.5460678669271,
+      "learning_rate": 5.319148936170213e-07,
+      "loss": 0.9027,
+      "step": 40
+    },
+    {
+      "epoch": 0.016773162939297124,
+      "grad_norm": 97.98883013825395,
+      "learning_rate": 5.585106382978723e-07,
+      "loss": 0.8252,
+      "step": 42
+    },
+    {
+      "epoch": 0.01757188498402556,
+      "grad_norm": 130.2877391749827,
+      "learning_rate": 5.851063829787235e-07,
+      "loss": 0.7532,
+      "step": 44
+    },
+    {
+      "epoch": 0.018370607028753993,
+      "grad_norm": 98.81399161595317,
+      "learning_rate": 6.117021276595745e-07,
+      "loss": 0.7013,
+      "step": 46
+    },
+    {
+      "epoch": 0.019169329073482427,
+      "grad_norm": 22.295409656345377,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 0.6467,
+      "step": 48
+    },
+    {
+      "epoch": 0.019968051118210862,
+      "grad_norm": 34.582920133699176,
+      "learning_rate": 6.648936170212766e-07,
+      "loss": 0.6351,
+      "step": 50
+    },
+    {
+      "epoch": 0.020766773162939296,
+      "grad_norm": 50.861520008988386,
+      "learning_rate": 6.914893617021278e-07,
+      "loss": 0.5912,
+      "step": 52
+    },
+    {
+      "epoch": 0.02156549520766773,
+      "grad_norm": 49.12926837947303,
+      "learning_rate": 7.180851063829789e-07,
+      "loss": 0.5413,
+      "step": 54
+    },
+    {
+      "epoch": 0.022364217252396165,
+      "grad_norm": 20.811523309346935,
+      "learning_rate": 7.446808510638298e-07,
+      "loss": 0.479,
+      "step": 56
+    },
+    {
+      "epoch": 0.0231629392971246,
+      "grad_norm": 12.332350980787394,
+      "learning_rate": 7.712765957446809e-07,
+      "loss": 0.4714,
+      "step": 58
+    },
+    {
+      "epoch": 0.023961661341853034,
+      "grad_norm": 15.193345562611738,
+      "learning_rate": 7.97872340425532e-07,
+      "loss": 0.4622,
+      "step": 60
+    },
+    {
+      "epoch": 0.02476038338658147,
+      "grad_norm": 5.637530351411472,
+      "learning_rate": 8.24468085106383e-07,
+      "loss": 0.4079,
+      "step": 62
+    },
+    {
+      "epoch": 0.025559105431309903,
+      "grad_norm": 11.657766039541796,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 0.3716,
+      "step": 64
+    },
+    {
+      "epoch": 0.026357827476038338,
+      "grad_norm": 8.003062600948699,
+      "learning_rate": 8.776595744680852e-07,
+      "loss": 0.4022,
+      "step": 66
+    },
+    {
+      "epoch": 0.027156549520766772,
+      "grad_norm": 9.863778660497967,
+      "learning_rate": 9.042553191489363e-07,
+      "loss": 0.4061,
+      "step": 68
+    },
+    {
+      "epoch": 0.027955271565495207,
+      "grad_norm": 26.781259322239684,
+      "learning_rate": 9.308510638297872e-07,
+      "loss": 0.3638,
+      "step": 70
+    },
+    {
+      "epoch": 0.02875399361022364,
+      "grad_norm": 5.445843816912823,
+      "learning_rate": 9.574468085106384e-07,
+      "loss": 0.3581,
+      "step": 72
+    },
+    {
+      "epoch": 0.029552715654952075,
+      "grad_norm": 9.213363331667637,
+      "learning_rate": 9.840425531914895e-07,
+      "loss": 0.3702,
+      "step": 74
+    },
+    {
+      "epoch": 0.03035143769968051,
+      "grad_norm": 4.727611360243214,
+      "learning_rate": 1.0106382978723404e-06,
+      "loss": 0.3487,
+      "step": 76
+    },
+    {
+      "epoch": 0.031150159744408944,
+      "grad_norm": 18.19406307616779,
+      "learning_rate": 1.0372340425531915e-06,
+      "loss": 0.3222,
+      "step": 78
+    },
+    {
+      "epoch": 0.03194888178913738,
+      "grad_norm": 9.595831277572767,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 0.3258,
+      "step": 80
+    },
+    {
+      "epoch": 0.03274760383386582,
+      "grad_norm": 35.38120873285774,
+      "learning_rate": 1.0904255319148938e-06,
+      "loss": 0.3055,
+      "step": 82
+    },
+    {
+      "epoch": 0.03354632587859425,
+      "grad_norm": 5.084938749131369,
+      "learning_rate": 1.1170212765957447e-06,
+      "loss": 0.3174,
+      "step": 84
+    },
+    {
+      "epoch": 0.034345047923322686,
+      "grad_norm": 7.402514872898367,
+      "learning_rate": 1.1436170212765958e-06,
+      "loss": 0.3097,
+      "step": 86
+    },
+    {
+      "epoch": 0.03514376996805112,
+      "grad_norm": 3.949621968931367,
+      "learning_rate": 1.170212765957447e-06,
+      "loss": 0.3152,
+      "step": 88
+    },
+    {
+      "epoch": 0.035942492012779555,
+      "grad_norm": 10.814942895226139,
+      "learning_rate": 1.196808510638298e-06,
+      "loss": 0.3037,
+      "step": 90
+    },
+    {
+      "epoch": 0.036741214057507986,
+      "grad_norm": 14.354678286624155,
+      "learning_rate": 1.223404255319149e-06,
+      "loss": 0.3261,
+      "step": 92
+    },
+    {
+      "epoch": 0.037539936102236424,
+      "grad_norm": 6.168542351180327,
+      "learning_rate": 1.25e-06,
+      "loss": 0.31,
+      "step": 94
+    },
+    {
+      "epoch": 0.038338658146964855,
+      "grad_norm": 11.539898124756117,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 0.311,
+      "step": 96
+    },
+    {
+      "epoch": 0.03913738019169329,
+      "grad_norm": 4.013221758787171,
+      "learning_rate": 1.3031914893617024e-06,
+      "loss": 0.2996,
+      "step": 98
+    },
+    {
+      "epoch": 0.039936102236421724,
+      "grad_norm": 3.860547140813227,
+      "learning_rate": 1.3297872340425533e-06,
+      "loss": 0.2763,
+      "step": 100
+    },
+    {
+      "epoch": 0.04073482428115016,
+      "grad_norm": 17.85966416524752,
+      "learning_rate": 1.3563829787234042e-06,
+      "loss": 0.2904,
+      "step": 102
+    },
+    {
+      "epoch": 0.04153354632587859,
+      "grad_norm": 16.697164957447292,
+      "learning_rate": 1.3829787234042555e-06,
+      "loss": 0.2832,
+      "step": 104
+    },
+    {
+      "epoch": 0.04233226837060703,
+      "grad_norm": 25.51601683379472,
+      "learning_rate": 1.4095744680851064e-06,
+      "loss": 0.2861,
+      "step": 106
+    },
+    {
+      "epoch": 0.04313099041533546,
+      "grad_norm": 43.70792394022941,
+      "learning_rate": 1.4361702127659578e-06,
+      "loss": 0.2882,
+      "step": 108
+    },
+    {
+      "epoch": 0.0439297124600639,
+      "grad_norm": 21.39234312256687,
+      "learning_rate": 1.4627659574468087e-06,
+      "loss": 0.2868,
+      "step": 110
+    },
+    {
+      "epoch": 0.04472843450479233,
+      "grad_norm": 4.299149147995996,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 0.2802,
+      "step": 112
+    },
+    {
+      "epoch": 0.04552715654952077,
+      "grad_norm": 4.853881749446152,
+      "learning_rate": 1.515957446808511e-06,
+      "loss": 0.2766,
+      "step": 114
+    },
+    {
+      "epoch": 0.0463258785942492,
+      "grad_norm": 14.252871564540007,
+      "learning_rate": 1.5425531914893618e-06,
+      "loss": 0.2696,
+      "step": 116
+    },
+    {
+      "epoch": 0.04712460063897764,
+      "grad_norm": 4.330757958493644,
+      "learning_rate": 1.5691489361702128e-06,
+      "loss": 0.2868,
+      "step": 118
+    },
+    {
+      "epoch": 0.04792332268370607,
+      "grad_norm": 5.144959414872726,
+      "learning_rate": 1.595744680851064e-06,
+      "loss": 0.2985,
+      "step": 120
+    },
+    {
+      "epoch": 0.048722044728434506,
+      "grad_norm": 13.510588438287128,
+      "learning_rate": 1.622340425531915e-06,
+      "loss": 0.2903,
+      "step": 122
+    },
+    {
+      "epoch": 0.04952076677316294,
+      "grad_norm": 9.984785828599879,
+      "learning_rate": 1.648936170212766e-06,
+      "loss": 0.3019,
+      "step": 124
+    },
+    {
+      "epoch": 0.050319488817891375,
+      "grad_norm": 11.150832562553619,
+      "learning_rate": 1.6755319148936172e-06,
+      "loss": 0.271,
+      "step": 126
+    },
+    {
+      "epoch": 0.051118210862619806,
+      "grad_norm": 14.058578339810987,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 0.273,
+      "step": 128
+    },
+    {
+      "epoch": 0.051916932907348244,
+      "grad_norm": 5.00151816005258,
+      "learning_rate": 1.7287234042553195e-06,
+      "loss": 0.2607,
+      "step": 130
+    },
+    {
+      "epoch": 0.052715654952076675,
+      "grad_norm": 15.004005157673818,
+      "learning_rate": 1.7553191489361704e-06,
+      "loss": 0.2654,
+      "step": 132
+    },
+    {
+      "epoch": 0.05351437699680511,
+      "grad_norm": 7.002483899819058,
+      "learning_rate": 1.7819148936170213e-06,
+      "loss": 0.2623,
+      "step": 134
+    },
+    {
+      "epoch": 0.054313099041533544,
+      "grad_norm": 3.9841361364612062,
+      "learning_rate": 1.8085106382978727e-06,
+      "loss": 0.2934,
+      "step": 136
+    },
+    {
+      "epoch": 0.05511182108626198,
+      "grad_norm": 3.8494401246050054,
+      "learning_rate": 1.8351063829787236e-06,
+      "loss": 0.274,
+      "step": 138
+    },
+    {
+      "epoch": 0.05591054313099041,
+      "grad_norm": 5.497525399265219,
+      "learning_rate": 1.8617021276595745e-06,
+      "loss": 0.2647,
+      "step": 140
+    },
+    {
+      "epoch": 0.05670926517571885,
+      "grad_norm": 4.980587322498712,
+      "learning_rate": 1.8882978723404258e-06,
+      "loss": 0.273,
+      "step": 142
+    },
+    {
+      "epoch": 0.05750798722044728,
+      "grad_norm": 3.7891589019522653,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 0.2517,
+      "step": 144
+    },
+    {
+      "epoch": 0.05830670926517572,
+      "grad_norm": 4.92828775714686,
+      "learning_rate": 1.941489361702128e-06,
+      "loss": 0.2704,
+      "step": 146
+    },
+    {
+      "epoch": 0.05910543130990415,
+      "grad_norm": 4.103755865538166,
+      "learning_rate": 1.968085106382979e-06,
+      "loss": 0.2591,
+      "step": 148
+    },
+    {
+      "epoch": 0.05990415335463259,
+      "grad_norm": 3.435413585690327,
+      "learning_rate": 1.99468085106383e-06,
+      "loss": 0.2509,
+      "step": 150
+    },
+    {
+      "epoch": 0.06070287539936102,
+      "grad_norm": 4.277721208929287,
+      "learning_rate": 2.021276595744681e-06,
+      "loss": 0.2584,
+      "step": 152
+    },
+    {
+      "epoch": 0.06150159744408946,
+      "grad_norm": 3.799249236289544,
+      "learning_rate": 2.047872340425532e-06,
+      "loss": 0.2589,
+      "step": 154
+    },
+    {
+      "epoch": 0.06230031948881789,
+      "grad_norm": 3.7296853670347825,
+      "learning_rate": 2.074468085106383e-06,
+      "loss": 0.2567,
+      "step": 156
+    },
+    {
+      "epoch": 0.06309904153354633,
+      "grad_norm": 4.512870491032634,
+      "learning_rate": 2.101063829787234e-06,
+      "loss": 0.2481,
+      "step": 158
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 9.719375176261458,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.24,
+      "step": 160
+    },
+    {
+      "epoch": 0.06469648562300319,
+      "grad_norm": 5.994224591000894,
+      "learning_rate": 2.1542553191489364e-06,
+      "loss": 0.245,
+      "step": 162
+    },
+    {
+      "epoch": 0.06549520766773163,
+      "grad_norm": 5.4415526514767025,
+      "learning_rate": 2.1808510638297876e-06,
+      "loss": 0.2653,
+      "step": 164
+    },
+    {
+      "epoch": 0.06629392971246006,
+      "grad_norm": 5.939080958451638,
+      "learning_rate": 2.2074468085106387e-06,
+      "loss": 0.2472,
+      "step": 166
+    },
+    {
+      "epoch": 0.0670926517571885,
+      "grad_norm": 5.160493886252985,
+      "learning_rate": 2.2340425531914894e-06,
+      "loss": 0.2471,
+      "step": 168
+    },
+    {
+      "epoch": 0.06789137380191693,
+      "grad_norm": 3.652646803021308,
+      "learning_rate": 2.2606382978723405e-06,
+      "loss": 0.2324,
+      "step": 170
+    },
+    {
+      "epoch": 0.06869009584664537,
+      "grad_norm": 3.8466095536871343,
+      "learning_rate": 2.2872340425531916e-06,
+      "loss": 0.2452,
+      "step": 172
+    },
+    {
+      "epoch": 0.0694888178913738,
+      "grad_norm": 3.3542675856152537,
+      "learning_rate": 2.3138297872340428e-06,
+      "loss": 0.2419,
+      "step": 174
+    },
+    {
+      "epoch": 0.07028753993610223,
+      "grad_norm": 4.481843921006337,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 0.2451,
+      "step": 176
+    },
+    {
+      "epoch": 0.07108626198083066,
+      "grad_norm": 3.03610682279202,
+      "learning_rate": 2.367021276595745e-06,
+      "loss": 0.2183,
+      "step": 178
+    },
+    {
+      "epoch": 0.07188498402555911,
+      "grad_norm": 3.1175806494594482,
+      "learning_rate": 2.393617021276596e-06,
+      "loss": 0.2349,
+      "step": 180
+    },
+    {
+      "epoch": 0.07268370607028754,
+      "grad_norm": 3.4236470265390033,
+      "learning_rate": 2.420212765957447e-06,
+      "loss": 0.2406,
+      "step": 182
+    },
+    {
+      "epoch": 0.07348242811501597,
+      "grad_norm": 3.604740428294787,
+      "learning_rate": 2.446808510638298e-06,
+      "loss": 0.2349,
+      "step": 184
+    },
+    {
+      "epoch": 0.0742811501597444,
+      "grad_norm": 3.255303266074429,
+      "learning_rate": 2.473404255319149e-06,
+      "loss": 0.2402,
+      "step": 186
+    },
+    {
+      "epoch": 0.07507987220447285,
+      "grad_norm": 3.030059130242592,
+      "learning_rate": 2.5e-06,
+      "loss": 0.2233,
+      "step": 188
+    },
+    {
+      "epoch": 0.07587859424920128,
+      "grad_norm": 5.396885260460128,
+      "learning_rate": 2.5265957446808513e-06,
+      "loss": 0.2264,
+      "step": 190
+    },
+    {
+      "epoch": 0.07667731629392971,
+      "grad_norm": 3.4037128475892384,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 0.2266,
+      "step": 192
+    },
+    {
+      "epoch": 0.07747603833865814,
+      "grad_norm": 3.540697871576238,
+      "learning_rate": 2.5797872340425536e-06,
+      "loss": 0.2302,
+      "step": 194
+    },
+    {
+      "epoch": 0.07827476038338659,
+      "grad_norm": 2.8155061709596616,
+      "learning_rate": 2.6063829787234047e-06,
+      "loss": 0.2442,
+      "step": 196
+    },
+    {
+      "epoch": 0.07907348242811502,
+      "grad_norm": 2.9188721708463885,
+      "learning_rate": 2.6329787234042554e-06,
+      "loss": 0.2261,
+      "step": 198
+    },
+    {
+      "epoch": 0.07987220447284345,
+      "grad_norm": 2.9114746100952598,
+      "learning_rate": 2.6595744680851065e-06,
+      "loss": 0.2287,
+      "step": 200
+    },
+    {
+      "epoch": 0.08067092651757188,
+      "grad_norm": 3.2180252275052257,
+      "learning_rate": 2.6861702127659577e-06,
+      "loss": 0.2187,
+      "step": 202
+    },
+    {
+      "epoch": 0.08146964856230032,
+      "grad_norm": 2.8873639429163314,
+      "learning_rate": 2.7127659574468084e-06,
+      "loss": 0.2158,
+      "step": 204
+    },
+    {
+      "epoch": 0.08226837060702875,
+      "grad_norm": 3.489228955679207,
+      "learning_rate": 2.73936170212766e-06,
+      "loss": 0.2258,
+      "step": 206
+    },
+    {
+      "epoch": 0.08306709265175719,
+      "grad_norm": 2.981624517929962,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 0.2108,
+      "step": 208
+    },
+    {
+      "epoch": 0.08386581469648563,
+      "grad_norm": 3.21155859005065,
+      "learning_rate": 2.7925531914893617e-06,
+      "loss": 0.216,
+      "step": 210
+    },
+    {
+      "epoch": 0.08466453674121406,
+      "grad_norm": 3.0096285120757935,
+      "learning_rate": 2.819148936170213e-06,
+      "loss": 0.2248,
+      "step": 212
+    },
+    {
+      "epoch": 0.08546325878594249,
+      "grad_norm": 2.9629033272723735,
+      "learning_rate": 2.845744680851064e-06,
+      "loss": 0.2319,
+      "step": 214
+    },
+    {
+      "epoch": 0.08626198083067092,
+      "grad_norm": 3.1407766370777646,
+      "learning_rate": 2.8723404255319155e-06,
+      "loss": 0.2083,
+      "step": 216
+    },
+    {
+      "epoch": 0.08706070287539937,
+      "grad_norm": 2.987122511363789,
+      "learning_rate": 2.8989361702127662e-06,
+      "loss": 0.2169,
+      "step": 218
+    },
+    {
+      "epoch": 0.0878594249201278,
+      "grad_norm": 3.0880793923727063,
+      "learning_rate": 2.9255319148936174e-06,
+      "loss": 0.2103,
+      "step": 220
+    },
+    {
+      "epoch": 0.08865814696485623,
+      "grad_norm": 2.7280690541697075,
+      "learning_rate": 2.9521276595744685e-06,
+      "loss": 0.1997,
+      "step": 222
+    },
+    {
+      "epoch": 0.08945686900958466,
+      "grad_norm": 4.457072529565402,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 0.212,
+      "step": 224
+    },
+    {
+      "epoch": 0.0902555910543131,
+      "grad_norm": 2.6269741757814176,
+      "learning_rate": 3.0053191489361703e-06,
+      "loss": 0.2092,
+      "step": 226
+    },
+    {
+      "epoch": 0.09105431309904154,
+      "grad_norm": 6.22452274606642,
+      "learning_rate": 3.031914893617022e-06,
+      "loss": 0.2049,
+      "step": 228
+    },
+    {
+      "epoch": 0.09185303514376997,
+      "grad_norm": 2.877291306172186,
+      "learning_rate": 3.0585106382978726e-06,
+      "loss": 0.2083,
+      "step": 230
+    },
+    {
+      "epoch": 0.0926517571884984,
+      "grad_norm": 5.1194444563249935,
+      "learning_rate": 3.0851063829787237e-06,
+      "loss": 0.2054,
+      "step": 232
+    },
+    {
+      "epoch": 0.09345047923322684,
+      "grad_norm": 6.2843799815917745,
+      "learning_rate": 3.111702127659575e-06,
+      "loss": 0.2194,
+      "step": 234
+    },
+    {
+      "epoch": 0.09424920127795527,
+      "grad_norm": 5.861687233187676,
+      "learning_rate": 3.1382978723404255e-06,
+      "loss": 0.2284,
+      "step": 236
+    },
+    {
+      "epoch": 0.0950479233226837,
+      "grad_norm": 3.213879356151013,
+      "learning_rate": 3.164893617021277e-06,
+      "loss": 0.2207,
+      "step": 238
+    },
+    {
+      "epoch": 0.09584664536741214,
+      "grad_norm": 3.1630850214772255,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 0.2066,
+      "step": 240
+    },
+    {
+      "epoch": 0.09664536741214058,
+      "grad_norm": 3.148239375316538,
+      "learning_rate": 3.218085106382979e-06,
+      "loss": 0.2212,
+      "step": 242
+    },
+    {
+      "epoch": 0.09744408945686901,
+      "grad_norm": 2.508025815693479,
+      "learning_rate": 3.24468085106383e-06,
+      "loss": 0.2054,
+      "step": 244
+    },
+    {
+      "epoch": 0.09824281150159744,
+      "grad_norm": 2.909768854428482,
+      "learning_rate": 3.271276595744681e-06,
+      "loss": 0.2186,
+      "step": 246
+    },
+    {
+      "epoch": 0.09904153354632587,
+      "grad_norm": 7.665173774329105,
+      "learning_rate": 3.297872340425532e-06,
+      "loss": 0.2228,
+      "step": 248
+    },
+    {
+      "epoch": 0.09984025559105432,
+      "grad_norm": 2.253729407128243,
+      "learning_rate": 3.3244680851063834e-06,
+      "loss": 0.2049,
+      "step": 250
+    },
+    {
+      "epoch": 0.10063897763578275,
+      "grad_norm": 3.8706889996285336,
+      "learning_rate": 3.3510638297872345e-06,
+      "loss": 0.2259,
+      "step": 252
+    },
+    {
+      "epoch": 0.10143769968051118,
+      "grad_norm": 3.5646371999914357,
+      "learning_rate": 3.377659574468085e-06,
+      "loss": 0.1859,
+      "step": 254
+    },
+    {
+      "epoch": 0.10223642172523961,
+      "grad_norm": 2.7753062530577597,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 0.2166,
+      "step": 256
+    },
+    {
+      "epoch": 0.10303514376996806,
+      "grad_norm": 3.0863522747098022,
+      "learning_rate": 3.4308510638297874e-06,
+      "loss": 0.2346,
+      "step": 258
+    },
+    {
+      "epoch": 0.10383386581469649,
+      "grad_norm": 2.7019568800833897,
+      "learning_rate": 3.457446808510639e-06,
+      "loss": 0.2031,
+      "step": 260
+    },
+    {
+      "epoch": 0.10463258785942492,
+      "grad_norm": 2.297946892414017,
+      "learning_rate": 3.4840425531914897e-06,
+      "loss": 0.1934,
+      "step": 262
+    },
+    {
+      "epoch": 0.10543130990415335,
+      "grad_norm": 2.714689444342462,
+      "learning_rate": 3.510638297872341e-06,
+      "loss": 0.2191,
+      "step": 264
+    },
+    {
+      "epoch": 0.1062300319488818,
+      "grad_norm": 2.2968600529073098,
+      "learning_rate": 3.537234042553192e-06,
+      "loss": 0.2078,
+      "step": 266
+    },
+    {
+      "epoch": 0.10702875399361023,
+      "grad_norm": 2.740039735249897,
+      "learning_rate": 3.5638297872340426e-06,
+      "loss": 0.2196,
+      "step": 268
+    },
+    {
+      "epoch": 0.10782747603833866,
+      "grad_norm": 2.4769261724935423,
+      "learning_rate": 3.5904255319148938e-06,
+      "loss": 0.2217,
+      "step": 270
+    },
+    {
+      "epoch": 0.10862619808306709,
+      "grad_norm": 2.3980460329324713,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 0.2174,
+      "step": 272
+    },
+    {
+      "epoch": 0.10942492012779553,
+      "grad_norm": 2.350186883116406,
+      "learning_rate": 3.643617021276596e-06,
+      "loss": 0.2164,
+      "step": 274
+    },
+    {
+      "epoch": 0.11022364217252396,
+      "grad_norm": 2.3070632701092,
+      "learning_rate": 3.670212765957447e-06,
+      "loss": 0.2058,
+      "step": 276
+    },
+    {
+      "epoch": 0.1110223642172524,
+      "grad_norm": 2.5876090442770914,
+      "learning_rate": 3.6968085106382983e-06,
+      "loss": 0.2102,
+      "step": 278
+    },
+    {
+      "epoch": 0.11182108626198083,
+      "grad_norm": 2.527498495500664,
+      "learning_rate": 3.723404255319149e-06,
+      "loss": 0.1878,
+      "step": 280
+    },
+    {
+      "epoch": 0.11261980830670927,
+      "grad_norm": 3.1408931185668734,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.2028,
+      "step": 282
+    },
+    {
+      "epoch": 0.1134185303514377,
+      "grad_norm": 2.438660009024771,
+      "learning_rate": 3.7765957446808516e-06,
+      "loss": 0.198,
+      "step": 284
+    },
+    {
+      "epoch": 0.11421725239616613,
+      "grad_norm": 3.163227332011823,
+      "learning_rate": 3.8031914893617023e-06,
+      "loss": 0.2143,
+      "step": 286
+    },
+    {
+      "epoch": 0.11501597444089456,
+      "grad_norm": 2.62548659237651,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 0.2129,
+      "step": 288
+    },
+    {
+      "epoch": 0.11581469648562301,
+      "grad_norm": 4.68249530525775,
+      "learning_rate": 3.856382978723405e-06,
+      "loss": 0.211,
+      "step": 290
+    },
+    {
+      "epoch": 0.11661341853035144,
+      "grad_norm": 2.5096724167776387,
+      "learning_rate": 3.882978723404256e-06,
+      "loss": 0.1901,
+      "step": 292
+    },
+    {
+      "epoch": 0.11741214057507987,
+      "grad_norm": 2.172894520215211,
+      "learning_rate": 3.909574468085106e-06,
+      "loss": 0.1936,
+      "step": 294
+    },
+    {
+      "epoch": 0.1182108626198083,
+      "grad_norm": 5.3196819233421495,
+      "learning_rate": 3.936170212765958e-06,
+      "loss": 0.2104,
+      "step": 296
+    },
+    {
+      "epoch": 0.11900958466453675,
+      "grad_norm": 2.4110105936218904,
+      "learning_rate": 3.962765957446809e-06,
+      "loss": 0.2143,
+      "step": 298
+    },
+    {
+      "epoch": 0.11980830670926518,
+      "grad_norm": 2.4033618282750724,
+      "learning_rate": 3.98936170212766e-06,
+      "loss": 0.2164,
+      "step": 300
+    },
+    {
+      "epoch": 0.12060702875399361,
+      "grad_norm": 3.457118213267144,
+      "learning_rate": 4.015957446808511e-06,
+      "loss": 0.2032,
+      "step": 302
+    },
+    {
+      "epoch": 0.12140575079872204,
+      "grad_norm": 2.408758592662987,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 0.2115,
+      "step": 304
+    },
+    {
+      "epoch": 0.12220447284345048,
+      "grad_norm": 2.59506801145671,
+      "learning_rate": 4.069148936170213e-06,
+      "loss": 0.2214,
+      "step": 306
+    },
+    {
+      "epoch": 0.12300319488817892,
+      "grad_norm": 3.0414661937486933,
+      "learning_rate": 4.095744680851064e-06,
+      "loss": 0.2088,
+      "step": 308
+    },
+    {
+      "epoch": 0.12380191693290735,
+      "grad_norm": 3.647758244704823,
+      "learning_rate": 4.122340425531915e-06,
+      "loss": 0.2198,
+      "step": 310
+    },
+    {
+      "epoch": 0.12460063897763578,
+      "grad_norm": 2.67698633677868,
+      "learning_rate": 4.148936170212766e-06,
+      "loss": 0.2252,
+      "step": 312
+    },
+    {
+      "epoch": 0.1253993610223642,
+      "grad_norm": 2.898397178237324,
+      "learning_rate": 4.175531914893618e-06,
+      "loss": 0.2181,
+      "step": 314
+    },
+    {
+      "epoch": 0.12619808306709265,
+      "grad_norm": 2.6840679225522526,
+      "learning_rate": 4.202127659574468e-06,
+      "loss": 0.1992,
+      "step": 316
+    },
+    {
+      "epoch": 0.1269968051118211,
+      "grad_norm": 3.1572391222603886,
+      "learning_rate": 4.228723404255319e-06,
+      "loss": 0.1952,
+      "step": 318
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 3.5678297945143873,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 0.2083,
+      "step": 320
+    },
+    {
+      "epoch": 0.12859424920127796,
+      "grad_norm": 2.645675379463557,
+      "learning_rate": 4.281914893617022e-06,
+      "loss": 0.2196,
+      "step": 322
+    },
+    {
+      "epoch": 0.12939297124600638,
+      "grad_norm": 3.2806373380609672,
+      "learning_rate": 4.308510638297873e-06,
+      "loss": 0.2044,
+      "step": 324
+    },
+    {
+      "epoch": 0.13019169329073482,
+      "grad_norm": 3.877339337802895,
+      "learning_rate": 4.3351063829787236e-06,
+      "loss": 0.2188,
+      "step": 326
+    },
+    {
+      "epoch": 0.13099041533546327,
+      "grad_norm": 2.855122134235383,
+      "learning_rate": 4.361702127659575e-06,
+      "loss": 0.2178,
+      "step": 328
+    },
+    {
+      "epoch": 0.13178913738019168,
+      "grad_norm": 2.2492003663786764,
+      "learning_rate": 4.388297872340426e-06,
+      "loss": 0.2113,
+      "step": 330
+    },
+    {
+      "epoch": 0.13258785942492013,
+      "grad_norm": 2.3120507711054166,
+      "learning_rate": 4.414893617021277e-06,
+      "loss": 0.2162,
+      "step": 332
+    },
+    {
+      "epoch": 0.13338658146964857,
+      "grad_norm": 2.306824843021891,
+      "learning_rate": 4.441489361702128e-06,
+      "loss": 0.1951,
+      "step": 334
+    },
+    {
+      "epoch": 0.134185303514377,
+      "grad_norm": 2.4562848392795424,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 0.2091,
+      "step": 336
+    },
+    {
+      "epoch": 0.13498402555910544,
+      "grad_norm": 2.327058395343108,
+      "learning_rate": 4.49468085106383e-06,
+      "loss": 0.1936,
+      "step": 338
+    },
+    {
+      "epoch": 0.13578274760383385,
+      "grad_norm": 3.2365595394106377,
+      "learning_rate": 4.521276595744681e-06,
+      "loss": 0.2076,
+      "step": 340
+    },
+    {
+      "epoch": 0.1365814696485623,
+      "grad_norm": 2.3773207333074797,
+      "learning_rate": 4.547872340425532e-06,
+      "loss": 0.215,
+      "step": 342
+    },
+    {
+      "epoch": 0.13738019169329074,
+      "grad_norm": 2.588667676878435,
+      "learning_rate": 4.574468085106383e-06,
+      "loss": 0.2142,
+      "step": 344
+    },
+    {
+      "epoch": 0.13817891373801916,
+      "grad_norm": 1.9922157018157862,
+      "learning_rate": 4.601063829787235e-06,
+      "loss": 0.188,
+      "step": 346
+    },
+    {
+      "epoch": 0.1389776357827476,
+      "grad_norm": 1.9986841603432328,
+      "learning_rate": 4.6276595744680855e-06,
+      "loss": 0.2123,
+      "step": 348
+    },
+    {
+      "epoch": 0.13977635782747605,
+      "grad_norm": 3.8191377619872346,
+      "learning_rate": 4.654255319148936e-06,
+      "loss": 0.1909,
+      "step": 350
+    },
+    {
+      "epoch": 0.14057507987220447,
+      "grad_norm": 2.707801265704851,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 0.1804,
+      "step": 352
+    },
+    {
+      "epoch": 0.1413738019169329,
+      "grad_norm": 2.3382091629044264,
+      "learning_rate": 4.707446808510639e-06,
+      "loss": 0.2268,
+      "step": 354
+    },
+    {
+      "epoch": 0.14217252396166133,
+      "grad_norm": 2.4107325950799963,
+      "learning_rate": 4.73404255319149e-06,
+      "loss": 0.204,
+      "step": 356
+    },
+    {
+      "epoch": 0.14297124600638977,
+      "grad_norm": 2.1678512268790944,
+      "learning_rate": 4.760638297872341e-06,
+      "loss": 0.2078,
+      "step": 358
+    },
+    {
+      "epoch": 0.14376996805111822,
+      "grad_norm": 2.028736424354992,
+      "learning_rate": 4.787234042553192e-06,
+      "loss": 0.2091,
+      "step": 360
+    },
+    {
+      "epoch": 0.14456869009584664,
+      "grad_norm": 2.396579988473481,
+      "learning_rate": 4.813829787234043e-06,
+      "loss": 0.2045,
+      "step": 362
+    },
+    {
+      "epoch": 0.14536741214057508,
+      "grad_norm": 2.3056223101352162,
+      "learning_rate": 4.840425531914894e-06,
+      "loss": 0.2012,
+      "step": 364
+    },
+    {
+      "epoch": 0.14616613418530353,
+      "grad_norm": 2.1325793827510964,
+      "learning_rate": 4.867021276595745e-06,
+      "loss": 0.1997,
+      "step": 366
+    },
+    {
+      "epoch": 0.14696485623003194,
+      "grad_norm": 2.2181283979891213,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 0.2216,
+      "step": 368
+    },
+    {
+      "epoch": 0.1477635782747604,
+      "grad_norm": 2.2520136230085988,
+      "learning_rate": 4.9202127659574475e-06,
+      "loss": 0.2093,
+      "step": 370
+    },
+    {
+      "epoch": 0.1485623003194888,
+      "grad_norm": 2.3640436529911817,
+      "learning_rate": 4.946808510638298e-06,
+      "loss": 0.2169,
+      "step": 372
+    },
+    {
+      "epoch": 0.14936102236421725,
+      "grad_norm": 2.3535553790924952,
+      "learning_rate": 4.973404255319149e-06,
+      "loss": 0.2098,
+      "step": 374
+    },
+    {
+      "epoch": 0.1501597444089457,
+      "grad_norm": 1.8756114301592517,
+      "learning_rate": 5e-06,
+      "loss": 0.1986,
+      "step": 376
+    },
+    {
+      "epoch": 0.1509584664536741,
+      "grad_norm": 2.292659355977137,
+      "learning_rate": 5.026595744680851e-06,
+      "loss": 0.2046,
+      "step": 378
+    },
+    {
+      "epoch": 0.15175718849840256,
+      "grad_norm": 2.6775913216468994,
+      "learning_rate": 5.053191489361703e-06,
+      "loss": 0.2376,
+      "step": 380
+    },
+    {
+      "epoch": 0.152555910543131,
+      "grad_norm": 2.140742604355663,
+      "learning_rate": 5.079787234042553e-06,
+      "loss": 0.2076,
+      "step": 382
+    },
+    {
+      "epoch": 0.15335463258785942,
+      "grad_norm": 2.2577220617248983,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 0.1857,
+      "step": 384
+    },
+    {
+      "epoch": 0.15415335463258786,
+      "grad_norm": 2.322529114683112,
+      "learning_rate": 5.1329787234042565e-06,
+      "loss": 0.2252,
+      "step": 386
+    },
+    {
+      "epoch": 0.15495207667731628,
+      "grad_norm": 2.174339578110931,
+      "learning_rate": 5.159574468085107e-06,
+      "loss": 0.1859,
+      "step": 388
+    },
+    {
+      "epoch": 0.15575079872204473,
+      "grad_norm": 2.6754457483199996,
+      "learning_rate": 5.186170212765958e-06,
+      "loss": 0.2134,
+      "step": 390
+    },
+    {
+      "epoch": 0.15654952076677317,
+      "grad_norm": 2.5869207840676287,
+      "learning_rate": 5.212765957446809e-06,
+      "loss": 0.2313,
+      "step": 392
+    },
+    {
+      "epoch": 0.1573482428115016,
+      "grad_norm": 2.2594133874239555,
+      "learning_rate": 5.23936170212766e-06,
+      "loss": 0.2063,
+      "step": 394
+    },
+    {
+      "epoch": 0.15814696485623003,
+      "grad_norm": 2.3216671883603532,
+      "learning_rate": 5.265957446808511e-06,
+      "loss": 0.2068,
+      "step": 396
+    },
+    {
+      "epoch": 0.15894568690095848,
+      "grad_norm": 2.2695053056973014,
+      "learning_rate": 5.292553191489362e-06,
+      "loss": 0.1911,
+      "step": 398
+    },
+    {
+      "epoch": 0.1597444089456869,
+      "grad_norm": 2.2393947266924474,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 0.2134,
+      "step": 400
+    },
+    {
+      "epoch": 0.16054313099041534,
+      "grad_norm": 2.3780064899990223,
+      "learning_rate": 5.345744680851064e-06,
+      "loss": 0.1861,
+      "step": 402
+    },
+    {
+      "epoch": 0.16134185303514376,
+      "grad_norm": 2.0639877092575456,
+      "learning_rate": 5.372340425531915e-06,
+      "loss": 0.1934,
+      "step": 404
+    },
+    {
+      "epoch": 0.1621405750798722,
+      "grad_norm": 2.3834910258697883,
+      "learning_rate": 5.398936170212766e-06,
+      "loss": 0.2268,
+      "step": 406
+    },
+    {
+      "epoch": 0.16293929712460065,
+      "grad_norm": 2.1443375139668253,
+      "learning_rate": 5.425531914893617e-06,
+      "loss": 0.1974,
+      "step": 408
+    },
+    {
+      "epoch": 0.16373801916932906,
+      "grad_norm": 2.1691183471396323,
+      "learning_rate": 5.452127659574469e-06,
+      "loss": 0.1985,
+      "step": 410
+    },
+    {
+      "epoch": 0.1645367412140575,
+      "grad_norm": 2.4848574597065904,
+      "learning_rate": 5.47872340425532e-06,
+      "loss": 0.2127,
+      "step": 412
+    },
+    {
+      "epoch": 0.16533546325878595,
+      "grad_norm": 2.3579622949897723,
+      "learning_rate": 5.5053191489361705e-06,
+      "loss": 0.2174,
+      "step": 414
+    },
+    {
+      "epoch": 0.16613418530351437,
+      "grad_norm": 2.070124617271312,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 0.1862,
+      "step": 416
+    },
+    {
+      "epoch": 0.16693290734824281,
+      "grad_norm": 2.1267517696375213,
+      "learning_rate": 5.558510638297873e-06,
+      "loss": 0.2201,
+      "step": 418
+    },
+    {
+      "epoch": 0.16773162939297126,
+      "grad_norm": 2.7521587042179134,
+      "learning_rate": 5.5851063829787235e-06,
+      "loss": 0.1829,
+      "step": 420
+    },
+    {
+      "epoch": 0.16853035143769968,
+      "grad_norm": 2.4493518149473106,
+      "learning_rate": 5.611702127659575e-06,
+      "loss": 0.2229,
+      "step": 422
+    },
+    {
+      "epoch": 0.16932907348242812,
+      "grad_norm": 2.248345922944704,
+      "learning_rate": 5.638297872340426e-06,
+      "loss": 0.1946,
+      "step": 424
+    },
+    {
+      "epoch": 0.17012779552715654,
+      "grad_norm": 2.0487939416257763,
+      "learning_rate": 5.664893617021277e-06,
+      "loss": 0.2045,
+      "step": 426
+    },
+    {
+      "epoch": 0.17092651757188498,
+      "grad_norm": 2.2646018633860696,
+      "learning_rate": 5.691489361702128e-06,
+      "loss": 0.2146,
+      "step": 428
+    },
+    {
+      "epoch": 0.17172523961661343,
+      "grad_norm": 1.8960573449253009,
+      "learning_rate": 5.718085106382979e-06,
+      "loss": 0.2079,
+      "step": 430
+    },
+    {
+      "epoch": 0.17252396166134185,
+      "grad_norm": 2.176740909319673,
+      "learning_rate": 5.744680851063831e-06,
+      "loss": 0.21,
+      "step": 432
+    },
+    {
+      "epoch": 0.1733226837060703,
+      "grad_norm": 2.1430161168845987,
+      "learning_rate": 5.771276595744682e-06,
+      "loss": 0.2119,
+      "step": 434
+    },
+    {
+      "epoch": 0.17412140575079874,
+      "grad_norm": 1.9672794318914277,
+      "learning_rate": 5.7978723404255325e-06,
+      "loss": 0.2103,
+      "step": 436
+    },
+    {
+      "epoch": 0.17492012779552715,
+      "grad_norm": 2.1725129238295904,
+      "learning_rate": 5.824468085106384e-06,
+      "loss": 0.2097,
+      "step": 438
+    },
+    {
+      "epoch": 0.1757188498402556,
+      "grad_norm": 1.9320076776263606,
+      "learning_rate": 5.851063829787235e-06,
+      "loss": 0.2141,
+      "step": 440
+    },
+    {
+      "epoch": 0.17651757188498401,
+      "grad_norm": 2.0107708483969544,
+      "learning_rate": 5.877659574468085e-06,
+      "loss": 0.2138,
+      "step": 442
+    },
+    {
+      "epoch": 0.17731629392971246,
+      "grad_norm": 1.9692652691440204,
+      "learning_rate": 5.904255319148937e-06,
+      "loss": 0.1874,
+      "step": 444
+    },
+    {
+      "epoch": 0.1781150159744409,
+      "grad_norm": 2.1222862804558305,
+      "learning_rate": 5.930851063829788e-06,
+      "loss": 0.1875,
+      "step": 446
+    },
+    {
+      "epoch": 0.17891373801916932,
+      "grad_norm": 2.0073449512509414,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 0.2073,
+      "step": 448
+    },
+    {
+      "epoch": 0.17971246006389777,
+      "grad_norm": 2.0189150793812027,
+      "learning_rate": 5.98404255319149e-06,
+      "loss": 0.1886,
+      "step": 450
+    },
+    {
+      "epoch": 0.1805111821086262,
+      "grad_norm": 2.327500057066851,
+      "learning_rate": 6.010638297872341e-06,
+      "loss": 0.2026,
+      "step": 452
+    },
+    {
+      "epoch": 0.18130990415335463,
+      "grad_norm": 2.2501232755144183,
+      "learning_rate": 6.037234042553191e-06,
+      "loss": 0.2132,
+      "step": 454
+    },
+    {
+      "epoch": 0.18210862619808307,
+      "grad_norm": 2.317281078409067,
+      "learning_rate": 6.063829787234044e-06,
+      "loss": 0.2112,
+      "step": 456
+    },
+    {
+      "epoch": 0.1829073482428115,
+      "grad_norm": 1.9924372480322023,
+      "learning_rate": 6.090425531914894e-06,
+      "loss": 0.2092,
+      "step": 458
+    },
+    {
+      "epoch": 0.18370607028753994,
+      "grad_norm": 2.2346624173451963,
+      "learning_rate": 6.117021276595745e-06,
+      "loss": 0.2144,
+      "step": 460
+    },
+    {
+      "epoch": 0.18450479233226838,
+      "grad_norm": 2.175934261679526,
+      "learning_rate": 6.143617021276597e-06,
+      "loss": 0.1849,
+      "step": 462
+    },
+    {
+      "epoch": 0.1853035143769968,
+      "grad_norm": 2.266384026702202,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 0.2113,
+      "step": 464
+    },
+    {
+      "epoch": 0.18610223642172524,
+      "grad_norm": 1.9034790451109018,
+      "learning_rate": 6.196808510638298e-06,
+      "loss": 0.2115,
+      "step": 466
+    },
+    {
+      "epoch": 0.1869009584664537,
+      "grad_norm": 2.2484524626124087,
+      "learning_rate": 6.22340425531915e-06,
+      "loss": 0.2069,
+      "step": 468
+    },
+    {
+      "epoch": 0.1876996805111821,
+      "grad_norm": 2.0092685848128102,
+      "learning_rate": 6.25e-06,
+      "loss": 0.2031,
+      "step": 470
+    },
+    {
+      "epoch": 0.18849840255591055,
+      "grad_norm": 1.870835955521368,
+      "learning_rate": 6.276595744680851e-06,
+      "loss": 0.1903,
+      "step": 472
+    },
+    {
+      "epoch": 0.18929712460063897,
+      "grad_norm": 1.9836945986391599,
+      "learning_rate": 6.303191489361703e-06,
+      "loss": 0.2256,
+      "step": 474
+    },
+    {
+      "epoch": 0.1900958466453674,
+      "grad_norm": 1.8732225695197424,
+      "learning_rate": 6.329787234042554e-06,
+      "loss": 0.2,
+      "step": 476
+    },
+    {
+      "epoch": 0.19089456869009586,
+      "grad_norm": 2.0073497811760856,
+      "learning_rate": 6.356382978723404e-06,
+      "loss": 0.1996,
+      "step": 478
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 2.1089028896634785,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 0.1705,
+      "step": 480
+    },
+    {
+      "epoch": 0.19249201277955272,
+      "grad_norm": 2.3296857896927032,
+      "learning_rate": 6.409574468085107e-06,
+      "loss": 0.2127,
+      "step": 482
+    },
+    {
+      "epoch": 0.19329073482428116,
+      "grad_norm": 1.9240064591019737,
+      "learning_rate": 6.436170212765958e-06,
+      "loss": 0.2044,
+      "step": 484
+    },
+    {
+      "epoch": 0.19408945686900958,
+      "grad_norm": 2.0124517600701757,
+      "learning_rate": 6.462765957446809e-06,
+      "loss": 0.2149,
+      "step": 486
+    },
+    {
+      "epoch": 0.19488817891373802,
+      "grad_norm": 2.1954071243853,
+      "learning_rate": 6.48936170212766e-06,
+      "loss": 0.2202,
+      "step": 488
+    },
+    {
+      "epoch": 0.19568690095846644,
+      "grad_norm": 1.9656201249731728,
+      "learning_rate": 6.515957446808511e-06,
+      "loss": 0.1946,
+      "step": 490
+    },
+    {
+      "epoch": 0.1964856230031949,
+      "grad_norm": 1.989066037021515,
+      "learning_rate": 6.542553191489362e-06,
+      "loss": 0.2004,
+      "step": 492
+    },
+    {
+      "epoch": 0.19728434504792333,
+      "grad_norm": 1.872072162374712,
+      "learning_rate": 6.569148936170213e-06,
+      "loss": 0.2077,
+      "step": 494
+    },
+    {
+      "epoch": 0.19808306709265175,
+      "grad_norm": 1.9921049486714641,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 0.1874,
+      "step": 496
+    },
+    {
+      "epoch": 0.1988817891373802,
+      "grad_norm": 1.7219513015277055,
+      "learning_rate": 6.622340425531916e-06,
+      "loss": 0.1821,
+      "step": 498
+    },
+    {
+      "epoch": 0.19968051118210864,
+      "grad_norm": 2.0749672908831616,
+      "learning_rate": 6.648936170212767e-06,
+      "loss": 0.2053,
+      "step": 500
+    },
+    {
+      "epoch": 0.19968051118210864,
+      "eval_loss": 0.19043219089508057,
+      "eval_runtime": 420.1871,
+      "eval_samples_per_second": 42.379,
+      "eval_steps_per_second": 5.298,
+      "step": 500
+    },
+    {
+      "epoch": 0.20047923322683706,
+      "grad_norm": 2.037306801330803,
+      "learning_rate": 6.6755319148936174e-06,
+      "loss": 0.2176,
+      "step": 502
+    },
+    {
+      "epoch": 0.2012779552715655,
+      "grad_norm": 1.9203371381017114,
+      "learning_rate": 6.702127659574469e-06,
+      "loss": 0.2125,
+      "step": 504
+    },
+    {
+      "epoch": 0.20207667731629392,
+      "grad_norm": 1.985111227444487,
+      "learning_rate": 6.72872340425532e-06,
+      "loss": 0.1973,
+      "step": 506
+    },
+    {
+      "epoch": 0.20287539936102236,
+      "grad_norm": 1.8951174304875817,
+      "learning_rate": 6.75531914893617e-06,
+      "loss": 0.2149,
+      "step": 508
+    },
+    {
+      "epoch": 0.2036741214057508,
+      "grad_norm": 2.0497514964752748,
+      "learning_rate": 6.781914893617022e-06,
+      "loss": 0.1978,
+      "step": 510
+    },
+    {
+      "epoch": 0.20447284345047922,
+      "grad_norm": 2.0766753850394175,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 0.2375,
+      "step": 512
+    },
+    {
+      "epoch": 0.20527156549520767,
+      "grad_norm": 2.346429408581163,
+      "learning_rate": 6.835106382978723e-06,
+      "loss": 0.2194,
+      "step": 514
+    },
+    {
+      "epoch": 0.20607028753993611,
+      "grad_norm": 2.0826174062149647,
+      "learning_rate": 6.861702127659575e-06,
+      "loss": 0.2056,
+      "step": 516
+    },
+    {
+      "epoch": 0.20686900958466453,
+      "grad_norm": 1.99670732186987,
+      "learning_rate": 6.888297872340426e-06,
+      "loss": 0.221,
+      "step": 518
+    },
+    {
+      "epoch": 0.20766773162939298,
+      "grad_norm": 2.036868735781395,
+      "learning_rate": 6.914893617021278e-06,
+      "loss": 0.2165,
+      "step": 520
+    },
+    {
+      "epoch": 0.2084664536741214,
+      "grad_norm": 1.8657555178307903,
+      "learning_rate": 6.941489361702129e-06,
+      "loss": 0.2146,
+      "step": 522
+    },
+    {
+      "epoch": 0.20926517571884984,
+      "grad_norm": 1.9998386354651805,
+      "learning_rate": 6.968085106382979e-06,
+      "loss": 0.2165,
+      "step": 524
+    },
+    {
+      "epoch": 0.21006389776357828,
+      "grad_norm": 1.890648331838507,
+      "learning_rate": 6.994680851063831e-06,
+      "loss": 0.2215,
+      "step": 526
+    },
+    {
+      "epoch": 0.2108626198083067,
+      "grad_norm": 1.8453249030357137,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 0.227,
+      "step": 528
+    },
+    {
+      "epoch": 0.21166134185303515,
+      "grad_norm": 1.8696171981456626,
+      "learning_rate": 7.047872340425532e-06,
+      "loss": 0.202,
+      "step": 530
+    },
+    {
+      "epoch": 0.2124600638977636,
+      "grad_norm": 1.9343681296268649,
+      "learning_rate": 7.074468085106384e-06,
+      "loss": 0.2207,
+      "step": 532
+    },
+    {
+      "epoch": 0.213258785942492,
+      "grad_norm": 2.004530964431762,
+      "learning_rate": 7.101063829787235e-06,
+      "loss": 0.2024,
+      "step": 534
+    },
+    {
+      "epoch": 0.21405750798722045,
+      "grad_norm": 1.9649699331241137,
+      "learning_rate": 7.127659574468085e-06,
+      "loss": 0.201,
+      "step": 536
+    },
+    {
+      "epoch": 0.21485623003194887,
+      "grad_norm": 1.998147156306178,
+      "learning_rate": 7.154255319148937e-06,
+      "loss": 0.1925,
+      "step": 538
+    },
+    {
+      "epoch": 0.21565495207667731,
+      "grad_norm": 1.872474740663219,
+      "learning_rate": 7.1808510638297875e-06,
+      "loss": 0.1979,
+      "step": 540
+    },
+    {
+      "epoch": 0.21645367412140576,
+      "grad_norm": 2.0711681755729634,
+      "learning_rate": 7.207446808510638e-06,
+      "loss": 0.2468,
+      "step": 542
+    },
+    {
+      "epoch": 0.21725239616613418,
+      "grad_norm": 1.9701565676595827,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 0.2092,
+      "step": 544
+    },
+    {
+      "epoch": 0.21805111821086262,
+      "grad_norm": 2.0764550423720927,
+      "learning_rate": 7.260638297872341e-06,
+      "loss": 0.2148,
+      "step": 546
+    },
+    {
+      "epoch": 0.21884984025559107,
+      "grad_norm": 3.213288011368614,
+      "learning_rate": 7.287234042553192e-06,
+      "loss": 0.2287,
+      "step": 548
+    },
+    {
+      "epoch": 0.21964856230031948,
+      "grad_norm": 1.8177331310953289,
+      "learning_rate": 7.313829787234044e-06,
+      "loss": 0.207,
+      "step": 550
+    },
+    {
+      "epoch": 0.22044728434504793,
+      "grad_norm": 1.7655366549601414,
+      "learning_rate": 7.340425531914894e-06,
+      "loss": 0.2064,
+      "step": 552
+    },
+    {
+      "epoch": 0.22124600638977635,
+      "grad_norm": 1.9594454321675279,
+      "learning_rate": 7.367021276595745e-06,
+      "loss": 0.1922,
+      "step": 554
+    },
+    {
+      "epoch": 0.2220447284345048,
+      "grad_norm": 2.1733754411582407,
+      "learning_rate": 7.3936170212765965e-06,
+      "loss": 0.2266,
+      "step": 556
+    },
+    {
+      "epoch": 0.22284345047923323,
+      "grad_norm": 1.8248091216623237,
+      "learning_rate": 7.420212765957447e-06,
+      "loss": 0.2221,
+      "step": 558
+    },
+    {
+      "epoch": 0.22364217252396165,
+      "grad_norm": 2.058872918004827,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 0.2225,
+      "step": 560
+    },
+    {
+      "epoch": 0.2244408945686901,
+      "grad_norm": 1.909102118780366,
+      "learning_rate": 7.47340425531915e-06,
+      "loss": 0.1983,
+      "step": 562
+    },
+    {
+      "epoch": 0.22523961661341854,
+      "grad_norm": 2.0278745632782584,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2076,
+      "step": 564
+    },
+    {
+      "epoch": 0.22603833865814696,
+      "grad_norm": 1.8175943014340494,
+      "learning_rate": 7.526595744680851e-06,
+      "loss": 0.1953,
+      "step": 566
+    },
+    {
+      "epoch": 0.2268370607028754,
+      "grad_norm": 2.025268083189413,
+      "learning_rate": 7.553191489361703e-06,
+      "loss": 0.2338,
+      "step": 568
+    },
+    {
+      "epoch": 0.22763578274760382,
+      "grad_norm": 1.8856154811157113,
+      "learning_rate": 7.579787234042554e-06,
+      "loss": 0.2077,
+      "step": 570
+    },
+    {
+      "epoch": 0.22843450479233227,
+      "grad_norm": 2.59965148545968,
+      "learning_rate": 7.606382978723405e-06,
+      "loss": 0.2057,
+      "step": 572
+    },
+    {
+      "epoch": 0.2292332268370607,
+      "grad_norm": 1.977278774789199,
+      "learning_rate": 7.632978723404256e-06,
+      "loss": 0.2021,
+      "step": 574
+    },
+    {
+      "epoch": 0.23003194888178913,
+      "grad_norm": 2.1109133002264735,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 0.2156,
+      "step": 576
+    },
+    {
+      "epoch": 0.23083067092651757,
+      "grad_norm": 2.017717675854075,
+      "learning_rate": 7.686170212765958e-06,
+      "loss": 0.2235,
+      "step": 578
+    },
+    {
+      "epoch": 0.23162939297124602,
+      "grad_norm": 1.8309604789907734,
+      "learning_rate": 7.71276595744681e-06,
+      "loss": 0.1801,
+      "step": 580
+    },
+    {
+      "epoch": 0.23242811501597443,
+      "grad_norm": 1.7699335955775852,
+      "learning_rate": 7.73936170212766e-06,
+      "loss": 0.1943,
+      "step": 582
+    },
+    {
+      "epoch": 0.23322683706070288,
+      "grad_norm": 1.9194267663124567,
+      "learning_rate": 7.765957446808511e-06,
+      "loss": 0.2153,
+      "step": 584
+    },
+    {
+      "epoch": 0.2340255591054313,
+      "grad_norm": 2.097345382192139,
+      "learning_rate": 7.792553191489362e-06,
+      "loss": 0.2338,
+      "step": 586
+    },
+    {
+      "epoch": 0.23482428115015974,
+      "grad_norm": 1.8940063065960397,
+      "learning_rate": 7.819148936170213e-06,
+      "loss": 0.2031,
+      "step": 588
+    },
+    {
+      "epoch": 0.2356230031948882,
+      "grad_norm": 1.9690164598346551,
+      "learning_rate": 7.845744680851064e-06,
+      "loss": 0.2118,
+      "step": 590
+    },
+    {
+      "epoch": 0.2364217252396166,
+      "grad_norm": 3.228651225314339,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 0.2039,
+      "step": 592
+    },
+    {
+      "epoch": 0.23722044728434505,
+      "grad_norm": 1.9156226006051282,
+      "learning_rate": 7.898936170212767e-06,
+      "loss": 0.2252,
+      "step": 594
+    },
+    {
+      "epoch": 0.2380191693290735,
+      "grad_norm": 1.7768150211461646,
+      "learning_rate": 7.925531914893617e-06,
+      "loss": 0.225,
+      "step": 596
+    },
+    {
+      "epoch": 0.2388178913738019,
+      "grad_norm": 1.9652050222591573,
+      "learning_rate": 7.95212765957447e-06,
+      "loss": 0.1965,
+      "step": 598
+    },
+    {
+      "epoch": 0.23961661341853036,
+      "grad_norm": 2.0572778537541323,
+      "learning_rate": 7.97872340425532e-06,
+      "loss": 0.203,
+      "step": 600
+    },
+    {
+      "epoch": 0.24041533546325877,
+      "grad_norm": 1.9770278124409975,
+      "learning_rate": 8.005319148936171e-06,
+      "loss": 0.2106,
+      "step": 602
+    },
+    {
+      "epoch": 0.24121405750798722,
+      "grad_norm": 1.8557771165334138,
+      "learning_rate": 8.031914893617022e-06,
+      "loss": 0.2075,
+      "step": 604
+    },
+    {
+      "epoch": 0.24201277955271566,
+      "grad_norm": 1.770990487744795,
+      "learning_rate": 8.058510638297873e-06,
+      "loss": 0.2138,
+      "step": 606
+    },
+    {
+      "epoch": 0.24281150159744408,
+      "grad_norm": 1.6450172966793275,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 0.1858,
+      "step": 608
+    },
+    {
+      "epoch": 0.24361022364217252,
+      "grad_norm": 1.6531257153455905,
+      "learning_rate": 8.111702127659576e-06,
+      "loss": 0.2181,
+      "step": 610
+    },
+    {
+      "epoch": 0.24440894568690097,
+      "grad_norm": 1.7953441983194227,
+      "learning_rate": 8.138297872340426e-06,
+      "loss": 0.2168,
+      "step": 612
+    },
+    {
+      "epoch": 0.2452076677316294,
+      "grad_norm": 1.9624853486169551,
+      "learning_rate": 8.164893617021277e-06,
+      "loss": 0.2025,
+      "step": 614
+    },
+    {
+      "epoch": 0.24600638977635783,
+      "grad_norm": 1.9202333243211456,
+      "learning_rate": 8.191489361702128e-06,
+      "loss": 0.2094,
+      "step": 616
+    },
+    {
+      "epoch": 0.24680511182108625,
+      "grad_norm": 1.9054794785418179,
+      "learning_rate": 8.218085106382978e-06,
+      "loss": 0.2287,
+      "step": 618
+    },
+    {
+      "epoch": 0.2476038338658147,
+      "grad_norm": 1.6738673715101495,
+      "learning_rate": 8.24468085106383e-06,
+      "loss": 0.2113,
+      "step": 620
+    },
+    {
+      "epoch": 0.24840255591054314,
+      "grad_norm": 1.6070602027792467,
+      "learning_rate": 8.271276595744682e-06,
+      "loss": 0.1973,
+      "step": 622
+    },
+    {
+      "epoch": 0.24920127795527156,
+      "grad_norm": 2.0785351155835134,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 0.2166,
+      "step": 624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7980372006314775,
+      "learning_rate": 8.324468085106385e-06,
+      "loss": 0.2113,
+      "step": 626
+    },
+    {
+      "epoch": 0.2507987220447284,
+      "grad_norm": 2.0745759191840096,
+      "learning_rate": 8.351063829787235e-06,
+      "loss": 0.2359,
+      "step": 628
+    },
+    {
+      "epoch": 0.2515974440894569,
+      "grad_norm": 1.9864922897804826,
+      "learning_rate": 8.377659574468086e-06,
+      "loss": 0.2062,
+      "step": 630
+    },
+    {
+      "epoch": 0.2523961661341853,
+      "grad_norm": 2.030788677329014,
+      "learning_rate": 8.404255319148937e-06,
+      "loss": 0.2217,
+      "step": 632
+    },
+    {
+      "epoch": 0.2531948881789137,
+      "grad_norm": 1.7547740269944636,
+      "learning_rate": 8.430851063829787e-06,
+      "loss": 0.2203,
+      "step": 634
+    },
+    {
+      "epoch": 0.2539936102236422,
+      "grad_norm": 1.7889762015226203,
+      "learning_rate": 8.457446808510638e-06,
+      "loss": 0.2165,
+      "step": 636
+    },
+    {
+      "epoch": 0.2547923322683706,
+      "grad_norm": 1.6288012322878784,
+      "learning_rate": 8.48404255319149e-06,
+      "loss": 0.1979,
+      "step": 638
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 1.7348141744007637,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 0.1996,
+      "step": 640
+    },
+    {
+      "epoch": 0.2563897763578275,
+      "grad_norm": 1.7262432044292573,
+      "learning_rate": 8.537234042553192e-06,
+      "loss": 0.1901,
+      "step": 642
+    },
+    {
+      "epoch": 0.2571884984025559,
+      "grad_norm": 1.9327837811354087,
+      "learning_rate": 8.563829787234044e-06,
+      "loss": 0.2001,
+      "step": 644
+    },
+    {
+      "epoch": 0.25798722044728434,
+      "grad_norm": 2.211386389180962,
+      "learning_rate": 8.590425531914895e-06,
+      "loss": 0.2213,
+      "step": 646
+    },
+    {
+      "epoch": 0.25878594249201275,
+      "grad_norm": 1.8243601975607768,
+      "learning_rate": 8.617021276595746e-06,
+      "loss": 0.224,
+      "step": 648
+    },
+    {
+      "epoch": 0.2595846645367412,
+      "grad_norm": 1.932439055129622,
+      "learning_rate": 8.643617021276596e-06,
+      "loss": 0.1924,
+      "step": 650
+    },
+    {
+      "epoch": 0.26038338658146964,
+      "grad_norm": 1.746238022604233,
+      "learning_rate": 8.670212765957447e-06,
+      "loss": 0.2149,
+      "step": 652
+    },
+    {
+      "epoch": 0.26118210862619806,
+      "grad_norm": 1.7613718087296124,
+      "learning_rate": 8.696808510638298e-06,
+      "loss": 0.2085,
+      "step": 654
+    },
+    {
+      "epoch": 0.26198083067092653,
+      "grad_norm": 1.7993776636313457,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 0.2287,
+      "step": 656
+    },
+    {
+      "epoch": 0.26277955271565495,
+      "grad_norm": 2.6631359329733972,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.1972,
+      "step": 658
+    },
+    {
+      "epoch": 0.26357827476038337,
+      "grad_norm": 2.0244414755737146,
+      "learning_rate": 8.776595744680852e-06,
+      "loss": 0.2113,
+      "step": 660
+    },
+    {
+      "epoch": 0.26437699680511184,
+      "grad_norm": 2.635052210402534,
+      "learning_rate": 8.803191489361704e-06,
+      "loss": 0.1862,
+      "step": 662
+    },
+    {
+      "epoch": 0.26517571884984026,
+      "grad_norm": 5.482936732062825,
+      "learning_rate": 8.829787234042555e-06,
+      "loss": 0.2109,
+      "step": 664
+    },
+    {
+      "epoch": 0.2659744408945687,
+      "grad_norm": 1.8743456888792152,
+      "learning_rate": 8.856382978723404e-06,
+      "loss": 0.2045,
+      "step": 666
+    },
+    {
+      "epoch": 0.26677316293929715,
+      "grad_norm": 1.4845513209949641,
+      "learning_rate": 8.882978723404256e-06,
+      "loss": 0.1779,
+      "step": 668
+    },
+    {
+      "epoch": 0.26757188498402557,
+      "grad_norm": 1.7833087776873404,
+      "learning_rate": 8.909574468085107e-06,
+      "loss": 0.1949,
+      "step": 670
+    },
+    {
+      "epoch": 0.268370607028754,
+      "grad_norm": 1.7237851857753832,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 0.2309,
+      "step": 672
+    },
+    {
+      "epoch": 0.26916932907348246,
+      "grad_norm": 1.7943976011736698,
+      "learning_rate": 8.96276595744681e-06,
+      "loss": 0.2246,
+      "step": 674
+    },
+    {
+      "epoch": 0.26996805111821087,
+      "grad_norm": 1.6645946104715852,
+      "learning_rate": 8.98936170212766e-06,
+      "loss": 0.213,
+      "step": 676
+    },
+    {
+      "epoch": 0.2707667731629393,
+      "grad_norm": 1.6878179409944127,
+      "learning_rate": 9.015957446808511e-06,
+      "loss": 0.2177,
+      "step": 678
+    },
+    {
+      "epoch": 0.2715654952076677,
+      "grad_norm": 1.6507645045148651,
+      "learning_rate": 9.042553191489362e-06,
+      "loss": 0.1924,
+      "step": 680
+    },
+    {
+      "epoch": 0.2723642172523962,
+      "grad_norm": 1.9095635971119895,
+      "learning_rate": 9.069148936170213e-06,
+      "loss": 0.2165,
+      "step": 682
+    },
+    {
+      "epoch": 0.2731629392971246,
+      "grad_norm": 1.630380756355513,
+      "learning_rate": 9.095744680851063e-06,
+      "loss": 0.2158,
+      "step": 684
+    },
+    {
+      "epoch": 0.273961661341853,
+      "grad_norm": 1.7077429176964645,
+      "learning_rate": 9.122340425531916e-06,
+      "loss": 0.2023,
+      "step": 686
+    },
+    {
+      "epoch": 0.2747603833865815,
+      "grad_norm": 1.7776755211491289,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 0.2179,
+      "step": 688
+    },
+    {
+      "epoch": 0.2755591054313099,
+      "grad_norm": 1.6954241240154826,
+      "learning_rate": 9.175531914893617e-06,
+      "loss": 0.2137,
+      "step": 690
+    },
+    {
+      "epoch": 0.2763578274760383,
+      "grad_norm": 1.7891304751555133,
+      "learning_rate": 9.20212765957447e-06,
+      "loss": 0.2117,
+      "step": 692
+    },
+    {
+      "epoch": 0.2771565495207668,
+      "grad_norm": 1.9639021184922467,
+      "learning_rate": 9.22872340425532e-06,
+      "loss": 0.2372,
+      "step": 694
+    },
+    {
+      "epoch": 0.2779552715654952,
+      "grad_norm": 1.5797174817631339,
+      "learning_rate": 9.255319148936171e-06,
+      "loss": 0.2063,
+      "step": 696
+    },
+    {
+      "epoch": 0.2787539936102236,
+      "grad_norm": 1.8249112869928803,
+      "learning_rate": 9.281914893617022e-06,
+      "loss": 0.2095,
+      "step": 698
+    },
+    {
+      "epoch": 0.2795527156549521,
+      "grad_norm": 1.7706586811506164,
+      "learning_rate": 9.308510638297872e-06,
+      "loss": 0.2073,
+      "step": 700
+    },
+    {
+      "epoch": 0.2803514376996805,
+      "grad_norm": 1.7271382999203275,
+      "learning_rate": 9.335106382978723e-06,
+      "loss": 0.2099,
+      "step": 702
+    },
+    {
+      "epoch": 0.28115015974440893,
+      "grad_norm": 1.6484304946170987,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 0.2133,
+      "step": 704
+    },
+    {
+      "epoch": 0.2819488817891374,
+      "grad_norm": 1.724151909486214,
+      "learning_rate": 9.388297872340426e-06,
+      "loss": 0.212,
+      "step": 706
+    },
+    {
+      "epoch": 0.2827476038338658,
+      "grad_norm": 1.5897126511168378,
+      "learning_rate": 9.414893617021279e-06,
+      "loss": 0.2,
+      "step": 708
+    },
+    {
+      "epoch": 0.28354632587859424,
+      "grad_norm": 1.7835429984815048,
+      "learning_rate": 9.44148936170213e-06,
+      "loss": 0.2318,
+      "step": 710
+    },
+    {
+      "epoch": 0.28434504792332266,
+      "grad_norm": 1.7751330791980955,
+      "learning_rate": 9.46808510638298e-06,
+      "loss": 0.2316,
+      "step": 712
+    },
+    {
+      "epoch": 0.28514376996805113,
+      "grad_norm": 1.8933530962247043,
+      "learning_rate": 9.49468085106383e-06,
+      "loss": 0.2265,
+      "step": 714
+    },
+    {
+      "epoch": 0.28594249201277955,
+      "grad_norm": 1.561209719510769,
+      "learning_rate": 9.521276595744681e-06,
+      "loss": 0.199,
+      "step": 716
+    },
+    {
+      "epoch": 0.28674121405750796,
+      "grad_norm": 1.8127384648554525,
+      "learning_rate": 9.547872340425532e-06,
+      "loss": 0.213,
+      "step": 718
+    },
+    {
+      "epoch": 0.28753993610223644,
+      "grad_norm": 3.645955534528657,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 0.2213,
+      "step": 720
+    },
+    {
+      "epoch": 0.28833865814696485,
+      "grad_norm": 2.2672006586352476,
+      "learning_rate": 9.601063829787235e-06,
+      "loss": 0.2139,
+      "step": 722
+    },
+    {
+      "epoch": 0.28913738019169327,
+      "grad_norm": 1.8519024796984893,
+      "learning_rate": 9.627659574468086e-06,
+      "loss": 0.2153,
+      "step": 724
+    },
+    {
+      "epoch": 0.28993610223642174,
+      "grad_norm": 1.560446824326613,
+      "learning_rate": 9.654255319148937e-06,
+      "loss": 0.2023,
+      "step": 726
+    },
+    {
+      "epoch": 0.29073482428115016,
+      "grad_norm": 1.7699204671345776,
+      "learning_rate": 9.680851063829787e-06,
+      "loss": 0.2238,
+      "step": 728
+    },
+    {
+      "epoch": 0.2915335463258786,
+      "grad_norm": 1.6865554967271952,
+      "learning_rate": 9.707446808510638e-06,
+      "loss": 0.2029,
+      "step": 730
+    },
+    {
+      "epoch": 0.29233226837060705,
+      "grad_norm": 1.6482995869085524,
+      "learning_rate": 9.73404255319149e-06,
+      "loss": 0.189,
+      "step": 732
+    },
+    {
+      "epoch": 0.29313099041533547,
+      "grad_norm": 1.7007866856749894,
+      "learning_rate": 9.760638297872341e-06,
+      "loss": 0.2203,
+      "step": 734
+    },
+    {
+      "epoch": 0.2939297124600639,
+      "grad_norm": 2.0612744928368585,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 0.2011,
+      "step": 736
+    },
+    {
+      "epoch": 0.29472843450479236,
+      "grad_norm": 1.6349743417964269,
+      "learning_rate": 9.813829787234044e-06,
+      "loss": 0.217,
+      "step": 738
+    },
+    {
+      "epoch": 0.2955271565495208,
+      "grad_norm": 1.6647666932854406,
+      "learning_rate": 9.840425531914895e-06,
+      "loss": 0.2083,
+      "step": 740
+    },
+    {
+      "epoch": 0.2963258785942492,
+      "grad_norm": 1.6704980371536418,
+      "learning_rate": 9.867021276595746e-06,
+      "loss": 0.1969,
+      "step": 742
+    },
+    {
+      "epoch": 0.2971246006389776,
+      "grad_norm": 1.6928260652626874,
+      "learning_rate": 9.893617021276596e-06,
+      "loss": 0.1927,
+      "step": 744
+    },
+    {
+      "epoch": 0.2979233226837061,
+      "grad_norm": 1.6889079212558915,
+      "learning_rate": 9.920212765957447e-06,
+      "loss": 0.2108,
+      "step": 746
+    },
+    {
+      "epoch": 0.2987220447284345,
+      "grad_norm": 1.6008195623226404,
+      "learning_rate": 9.946808510638298e-06,
+      "loss": 0.2079,
+      "step": 748
+    },
+    {
+      "epoch": 0.2995207667731629,
+      "grad_norm": 1.6391899308564928,
+      "learning_rate": 9.97340425531915e-06,
+      "loss": 0.224,
+      "step": 750
+    },
+    {
+      "epoch": 0.3003194888178914,
+      "grad_norm": 1.6579897287838672,
+      "learning_rate": 1e-05,
+      "loss": 0.2007,
+      "step": 752
+    },
+    {
+      "epoch": 0.3011182108626198,
+      "grad_norm": 1.6060266288421388,
+      "learning_rate": 9.99999784023588e-06,
+      "loss": 0.1831,
+      "step": 754
+    },
+    {
+      "epoch": 0.3019169329073482,
+      "grad_norm": 1.5755540572254816,
+      "learning_rate": 9.999991360945382e-06,
+      "loss": 0.2366,
+      "step": 756
+    },
+    {
+      "epoch": 0.3027156549520767,
+      "grad_norm": 1.7099166011335876,
+      "learning_rate": 9.999980562134104e-06,
+      "loss": 0.2231,
+      "step": 758
+    },
+    {
+      "epoch": 0.3035143769968051,
+      "grad_norm": 1.5689663504063338,
+      "learning_rate": 9.999965443811378e-06,
+      "loss": 0.1979,
+      "step": 760
+    },
+    {
+      "epoch": 0.30431309904153353,
+      "grad_norm": 2.1289064244731657,
+      "learning_rate": 9.999946005990262e-06,
+      "loss": 0.2284,
+      "step": 762
+    },
+    {
+      "epoch": 0.305111821086262,
+      "grad_norm": 1.530947535684416,
+      "learning_rate": 9.99992224868755e-06,
+      "loss": 0.2067,
+      "step": 764
+    },
+    {
+      "epoch": 0.3059105431309904,
+      "grad_norm": 1.666888953841592,
+      "learning_rate": 9.999894171923764e-06,
+      "loss": 0.2061,
+      "step": 766
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 1.5905369733593537,
+      "learning_rate": 9.999861775723162e-06,
+      "loss": 0.2047,
+      "step": 768
+    },
+    {
+      "epoch": 0.3075079872204473,
+      "grad_norm": 1.9351094836921137,
+      "learning_rate": 9.99982506011373e-06,
+      "loss": 0.2111,
+      "step": 770
+    },
+    {
+      "epoch": 0.3083067092651757,
+      "grad_norm": 1.6775809786308205,
+      "learning_rate": 9.999784025127187e-06,
+      "loss": 0.2009,
+      "step": 772
+    },
+    {
+      "epoch": 0.30910543130990414,
+      "grad_norm": 1.6337987957663225,
+      "learning_rate": 9.999738670798983e-06,
+      "loss": 0.1965,
+      "step": 774
+    },
+    {
+      "epoch": 0.30990415335463256,
+      "grad_norm": 1.6853294588250392,
+      "learning_rate": 9.999688997168301e-06,
+      "loss": 0.2094,
+      "step": 776
+    },
+    {
+      "epoch": 0.31070287539936103,
+      "grad_norm": 2.380951431702259,
+      "learning_rate": 9.999635004278054e-06,
+      "loss": 0.2131,
+      "step": 778
+    },
+    {
+      "epoch": 0.31150159744408945,
+      "grad_norm": 1.6821758607705601,
+      "learning_rate": 9.999576692174887e-06,
+      "loss": 0.2148,
+      "step": 780
+    },
+    {
+      "epoch": 0.31230031948881787,
+      "grad_norm": 1.5362646502228101,
+      "learning_rate": 9.999514060909175e-06,
+      "loss": 0.2348,
+      "step": 782
+    },
+    {
+      "epoch": 0.31309904153354634,
+      "grad_norm": 1.6226459616330424,
+      "learning_rate": 9.999447110535026e-06,
+      "loss": 0.2031,
+      "step": 784
+    },
+    {
+      "epoch": 0.31389776357827476,
+      "grad_norm": 1.494547748123698,
+      "learning_rate": 9.999375841110277e-06,
+      "loss": 0.2004,
+      "step": 786
+    },
+    {
+      "epoch": 0.3146964856230032,
+      "grad_norm": 1.800819253841243,
+      "learning_rate": 9.999300252696502e-06,
+      "loss": 0.2446,
+      "step": 788
+    },
+    {
+      "epoch": 0.31549520766773165,
+      "grad_norm": 1.60492245047213,
+      "learning_rate": 9.999220345359e-06,
+      "loss": 0.2044,
+      "step": 790
+    },
+    {
+      "epoch": 0.31629392971246006,
+      "grad_norm": 1.8934326067851825,
+      "learning_rate": 9.999136119166803e-06,
+      "loss": 0.2043,
+      "step": 792
+    },
+    {
+      "epoch": 0.3170926517571885,
+      "grad_norm": 1.5648068073113917,
+      "learning_rate": 9.999047574192677e-06,
+      "loss": 0.2089,
+      "step": 794
+    },
+    {
+      "epoch": 0.31789137380191695,
+      "grad_norm": 1.6312196131131216,
+      "learning_rate": 9.998954710513113e-06,
+      "loss": 0.1983,
+      "step": 796
+    },
+    {
+      "epoch": 0.31869009584664537,
+      "grad_norm": 1.4933485598542207,
+      "learning_rate": 9.998857528208337e-06,
+      "loss": 0.2149,
+      "step": 798
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 1.7284519491739558,
+      "learning_rate": 9.998756027362308e-06,
+      "loss": 0.2049,
+      "step": 800
+    },
+    {
+      "epoch": 0.32028753993610226,
+      "grad_norm": 1.6062264896600003,
+      "learning_rate": 9.998650208062713e-06,
+      "loss": 0.2302,
+      "step": 802
+    },
+    {
+      "epoch": 0.3210862619808307,
+      "grad_norm": 1.541668593359599,
+      "learning_rate": 9.998540070400966e-06,
+      "loss": 0.214,
+      "step": 804
+    },
+    {
+      "epoch": 0.3218849840255591,
+      "grad_norm": 1.5744953156084605,
+      "learning_rate": 9.998425614472217e-06,
+      "loss": 0.2209,
+      "step": 806
+    },
+    {
+      "epoch": 0.3226837060702875,
+      "grad_norm": 1.5946944416094957,
+      "learning_rate": 9.99830684037535e-06,
+      "loss": 0.2209,
+      "step": 808
+    },
+    {
+      "epoch": 0.323482428115016,
+      "grad_norm": 1.6547243881770919,
+      "learning_rate": 9.998183748212968e-06,
+      "loss": 0.2291,
+      "step": 810
+    },
+    {
+      "epoch": 0.3242811501597444,
+      "grad_norm": 1.5020701296905943,
+      "learning_rate": 9.998056338091415e-06,
+      "loss": 0.1947,
+      "step": 812
+    },
+    {
+      "epoch": 0.3250798722044728,
+      "grad_norm": 1.775646858056913,
+      "learning_rate": 9.997924610120758e-06,
+      "loss": 0.2122,
+      "step": 814
+    },
+    {
+      "epoch": 0.3258785942492013,
+      "grad_norm": 1.9643173710094557,
+      "learning_rate": 9.9977885644148e-06,
+      "loss": 0.2304,
+      "step": 816
+    },
+    {
+      "epoch": 0.3266773162939297,
+      "grad_norm": 2.2694487421961305,
+      "learning_rate": 9.997648201091073e-06,
+      "loss": 0.1922,
+      "step": 818
+    },
+    {
+      "epoch": 0.3274760383386581,
+      "grad_norm": 1.6657139292549983,
+      "learning_rate": 9.997503520270837e-06,
+      "loss": 0.2166,
+      "step": 820
+    },
+    {
+      "epoch": 0.3282747603833866,
+      "grad_norm": 1.419957033481622,
+      "learning_rate": 9.997354522079078e-06,
+      "loss": 0.2025,
+      "step": 822
+    },
+    {
+      "epoch": 0.329073482428115,
+      "grad_norm": 1.499108476748773,
+      "learning_rate": 9.997201206644522e-06,
+      "loss": 0.2159,
+      "step": 824
+    },
+    {
+      "epoch": 0.32987220447284343,
+      "grad_norm": 1.624603961236652,
+      "learning_rate": 9.997043574099616e-06,
+      "loss": 0.2147,
+      "step": 826
+    },
+    {
+      "epoch": 0.3306709265175719,
+      "grad_norm": 1.5313961974893107,
+      "learning_rate": 9.996881624580542e-06,
+      "loss": 0.2063,
+      "step": 828
+    },
+    {
+      "epoch": 0.3314696485623003,
+      "grad_norm": 1.5963165181678973,
+      "learning_rate": 9.996715358227208e-06,
+      "loss": 0.2163,
+      "step": 830
+    },
+    {
+      "epoch": 0.33226837060702874,
+      "grad_norm": 2.5507679267995793,
+      "learning_rate": 9.99654477518325e-06,
+      "loss": 0.1961,
+      "step": 832
+    },
+    {
+      "epoch": 0.3330670926517572,
+      "grad_norm": 1.729152523099962,
+      "learning_rate": 9.99636987559604e-06,
+      "loss": 0.2024,
+      "step": 834
+    },
+    {
+      "epoch": 0.33386581469648563,
+      "grad_norm": 1.7210992603515631,
+      "learning_rate": 9.99619065961667e-06,
+      "loss": 0.2252,
+      "step": 836
+    },
+    {
+      "epoch": 0.33466453674121405,
+      "grad_norm": 1.560277624031283,
+      "learning_rate": 9.99600712739997e-06,
+      "loss": 0.2267,
+      "step": 838
+    },
+    {
+      "epoch": 0.3354632587859425,
+      "grad_norm": 1.4591188372864774,
+      "learning_rate": 9.995819279104494e-06,
+      "loss": 0.2011,
+      "step": 840
+    },
+    {
+      "epoch": 0.33626198083067094,
+      "grad_norm": 1.5041613555668587,
+      "learning_rate": 9.995627114892522e-06,
+      "loss": 0.2274,
+      "step": 842
+    },
+    {
+      "epoch": 0.33706070287539935,
+      "grad_norm": 1.6583699889454577,
+      "learning_rate": 9.995430634930068e-06,
+      "loss": 0.1911,
+      "step": 844
+    },
+    {
+      "epoch": 0.33785942492012777,
+      "grad_norm": 1.4587897948416573,
+      "learning_rate": 9.99522983938687e-06,
+      "loss": 0.2237,
+      "step": 846
+    },
+    {
+      "epoch": 0.33865814696485624,
+      "grad_norm": 1.503331285695136,
+      "learning_rate": 9.995024728436402e-06,
+      "loss": 0.2094,
+      "step": 848
+    },
+    {
+      "epoch": 0.33945686900958466,
+      "grad_norm": 1.535943980164657,
+      "learning_rate": 9.994815302255854e-06,
+      "loss": 0.2166,
+      "step": 850
+    },
+    {
+      "epoch": 0.3402555910543131,
+      "grad_norm": 1.6504775369785527,
+      "learning_rate": 9.994601561026156e-06,
+      "loss": 0.2002,
+      "step": 852
+    },
+    {
+      "epoch": 0.34105431309904155,
+      "grad_norm": 1.4302740240626306,
+      "learning_rate": 9.994383504931955e-06,
+      "loss": 0.2082,
+      "step": 854
+    },
+    {
+      "epoch": 0.34185303514376997,
+      "grad_norm": 1.593656408563002,
+      "learning_rate": 9.994161134161635e-06,
+      "loss": 0.2094,
+      "step": 856
+    },
+    {
+      "epoch": 0.3426517571884984,
+      "grad_norm": 1.3695207582393358,
+      "learning_rate": 9.9939344489073e-06,
+      "loss": 0.1919,
+      "step": 858
+    },
+    {
+      "epoch": 0.34345047923322686,
+      "grad_norm": 1.6060661840602912,
+      "learning_rate": 9.993703449364787e-06,
+      "loss": 0.206,
+      "step": 860
+    },
+    {
+      "epoch": 0.3442492012779553,
+      "grad_norm": 1.4773210324662711,
+      "learning_rate": 9.993468135733658e-06,
+      "loss": 0.2192,
+      "step": 862
+    },
+    {
+      "epoch": 0.3450479233226837,
+      "grad_norm": 1.406665980225023,
+      "learning_rate": 9.993228508217201e-06,
+      "loss": 0.1907,
+      "step": 864
+    },
+    {
+      "epoch": 0.34584664536741216,
+      "grad_norm": 1.5457381119375846,
+      "learning_rate": 9.99298456702243e-06,
+      "loss": 0.188,
+      "step": 866
+    },
+    {
+      "epoch": 0.3466453674121406,
+      "grad_norm": 1.392632813313601,
+      "learning_rate": 9.99273631236009e-06,
+      "loss": 0.1814,
+      "step": 868
+    },
+    {
+      "epoch": 0.347444089456869,
+      "grad_norm": 1.541510486431838,
+      "learning_rate": 9.992483744444648e-06,
+      "loss": 0.202,
+      "step": 870
+    },
+    {
+      "epoch": 0.34824281150159747,
+      "grad_norm": 1.416164840329261,
+      "learning_rate": 9.9922268634943e-06,
+      "loss": 0.2034,
+      "step": 872
+    },
+    {
+      "epoch": 0.3490415335463259,
+      "grad_norm": 1.52233433092975,
+      "learning_rate": 9.991965669730965e-06,
+      "loss": 0.2186,
+      "step": 874
+    },
+    {
+      "epoch": 0.3498402555910543,
+      "grad_norm": 1.59873194009716,
+      "learning_rate": 9.991700163380292e-06,
+      "loss": 0.2353,
+      "step": 876
+    },
+    {
+      "epoch": 0.3506389776357827,
+      "grad_norm": 1.7195541732829867,
+      "learning_rate": 9.991430344671653e-06,
+      "loss": 0.2121,
+      "step": 878
+    },
+    {
+      "epoch": 0.3514376996805112,
+      "grad_norm": 1.6835655527767803,
+      "learning_rate": 9.991156213838143e-06,
+      "loss": 0.2048,
+      "step": 880
+    },
+    {
+      "epoch": 0.3522364217252396,
+      "grad_norm": 1.4417934095264713,
+      "learning_rate": 9.990877771116588e-06,
+      "loss": 0.1967,
+      "step": 882
+    },
+    {
+      "epoch": 0.35303514376996803,
+      "grad_norm": 1.6176474246414694,
+      "learning_rate": 9.990595016747536e-06,
+      "loss": 0.203,
+      "step": 884
+    },
+    {
+      "epoch": 0.3538338658146965,
+      "grad_norm": 1.5007198389997798,
+      "learning_rate": 9.99030795097526e-06,
+      "loss": 0.2201,
+      "step": 886
+    },
+    {
+      "epoch": 0.3546325878594249,
+      "grad_norm": 1.628516285198201,
+      "learning_rate": 9.990016574047757e-06,
+      "loss": 0.208,
+      "step": 888
+    },
+    {
+      "epoch": 0.35543130990415334,
+      "grad_norm": 1.5813428579068682,
+      "learning_rate": 9.989720886216749e-06,
+      "loss": 0.2196,
+      "step": 890
+    },
+    {
+      "epoch": 0.3562300319488818,
+      "grad_norm": 1.5508584713469626,
+      "learning_rate": 9.989420887737684e-06,
+      "loss": 0.2145,
+      "step": 892
+    },
+    {
+      "epoch": 0.3570287539936102,
+      "grad_norm": 1.7163590462789873,
+      "learning_rate": 9.989116578869732e-06,
+      "loss": 0.2393,
+      "step": 894
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 1.361569590542389,
+      "learning_rate": 9.988807959875785e-06,
+      "loss": 0.2064,
+      "step": 896
+    },
+    {
+      "epoch": 0.3586261980830671,
+      "grad_norm": 1.451447508537209,
+      "learning_rate": 9.988495031022465e-06,
+      "loss": 0.204,
+      "step": 898
+    },
+    {
+      "epoch": 0.35942492012779553,
+      "grad_norm": 1.4487071973493906,
+      "learning_rate": 9.988177792580107e-06,
+      "loss": 0.2009,
+      "step": 900
+    },
+    {
+      "epoch": 0.36022364217252395,
+      "grad_norm": 1.419806864642736,
+      "learning_rate": 9.98785624482278e-06,
+      "loss": 0.2024,
+      "step": 902
+    },
+    {
+      "epoch": 0.3610223642172524,
+      "grad_norm": 1.578737310761318,
+      "learning_rate": 9.987530388028269e-06,
+      "loss": 0.2092,
+      "step": 904
+    },
+    {
+      "epoch": 0.36182108626198084,
+      "grad_norm": 1.6200056976952275,
+      "learning_rate": 9.987200222478084e-06,
+      "loss": 0.209,
+      "step": 906
+    },
+    {
+      "epoch": 0.36261980830670926,
+      "grad_norm": 1.6251607361656684,
+      "learning_rate": 9.986865748457457e-06,
+      "loss": 0.2102,
+      "step": 908
+    },
+    {
+      "epoch": 0.3634185303514377,
+      "grad_norm": 1.60473197937151,
+      "learning_rate": 9.986526966255341e-06,
+      "loss": 0.2188,
+      "step": 910
+    },
+    {
+      "epoch": 0.36421725239616615,
+      "grad_norm": 1.4451319185877254,
+      "learning_rate": 9.986183876164412e-06,
+      "loss": 0.211,
+      "step": 912
+    },
+    {
+      "epoch": 0.36501597444089456,
+      "grad_norm": 1.7222368755921156,
+      "learning_rate": 9.985836478481069e-06,
+      "loss": 0.2093,
+      "step": 914
+    },
+    {
+      "epoch": 0.365814696485623,
+      "grad_norm": 1.5858620117161173,
+      "learning_rate": 9.98548477350543e-06,
+      "loss": 0.2128,
+      "step": 916
+    },
+    {
+      "epoch": 0.36661341853035145,
+      "grad_norm": 1.5923894300594037,
+      "learning_rate": 9.985128761541334e-06,
+      "loss": 0.2185,
+      "step": 918
+    },
+    {
+      "epoch": 0.36741214057507987,
+      "grad_norm": 1.5815229483720405,
+      "learning_rate": 9.984768442896342e-06,
+      "loss": 0.232,
+      "step": 920
+    },
+    {
+      "epoch": 0.3682108626198083,
+      "grad_norm": 1.3902654600389392,
+      "learning_rate": 9.984403817881736e-06,
+      "loss": 0.1875,
+      "step": 922
+    },
+    {
+      "epoch": 0.36900958466453676,
+      "grad_norm": 1.434425802911243,
+      "learning_rate": 9.984034886812519e-06,
+      "loss": 0.1891,
+      "step": 924
+    },
+    {
+      "epoch": 0.3698083067092652,
+      "grad_norm": 1.4965650327837223,
+      "learning_rate": 9.98366165000741e-06,
+      "loss": 0.212,
+      "step": 926
+    },
+    {
+      "epoch": 0.3706070287539936,
+      "grad_norm": 1.5934149301565348,
+      "learning_rate": 9.983284107788852e-06,
+      "loss": 0.2081,
+      "step": 928
+    },
+    {
+      "epoch": 0.37140575079872207,
+      "grad_norm": 1.5649065798161081,
+      "learning_rate": 9.982902260483003e-06,
+      "loss": 0.2148,
+      "step": 930
+    },
+    {
+      "epoch": 0.3722044728434505,
+      "grad_norm": 1.5919056980501831,
+      "learning_rate": 9.982516108419746e-06,
+      "loss": 0.2088,
+      "step": 932
+    },
+    {
+      "epoch": 0.3730031948881789,
+      "grad_norm": 1.3802136089446853,
+      "learning_rate": 9.982125651932681e-06,
+      "loss": 0.1826,
+      "step": 934
+    },
+    {
+      "epoch": 0.3738019169329074,
+      "grad_norm": 1.4882986051122178,
+      "learning_rate": 9.981730891359123e-06,
+      "loss": 0.2278,
+      "step": 936
+    },
+    {
+      "epoch": 0.3746006389776358,
+      "grad_norm": 1.377902572686802,
+      "learning_rate": 9.981331827040109e-06,
+      "loss": 0.2011,
+      "step": 938
+    },
+    {
+      "epoch": 0.3753993610223642,
+      "grad_norm": 1.574021407706636,
+      "learning_rate": 9.980928459320393e-06,
+      "loss": 0.2101,
+      "step": 940
+    },
+    {
+      "epoch": 0.3761980830670926,
+      "grad_norm": 1.4461584644500969,
+      "learning_rate": 9.980520788548445e-06,
+      "loss": 0.1986,
+      "step": 942
+    },
+    {
+      "epoch": 0.3769968051118211,
+      "grad_norm": 1.5726540162240885,
+      "learning_rate": 9.980108815076456e-06,
+      "loss": 0.2079,
+      "step": 944
+    },
+    {
+      "epoch": 0.3777955271565495,
+      "grad_norm": 1.631975663438154,
+      "learning_rate": 9.979692539260331e-06,
+      "loss": 0.2071,
+      "step": 946
+    },
+    {
+      "epoch": 0.37859424920127793,
+      "grad_norm": 1.3678329735675085,
+      "learning_rate": 9.979271961459696e-06,
+      "loss": 0.2109,
+      "step": 948
+    },
+    {
+      "epoch": 0.3793929712460064,
+      "grad_norm": 1.641531450859251,
+      "learning_rate": 9.978847082037886e-06,
+      "loss": 0.1935,
+      "step": 950
+    },
+    {
+      "epoch": 0.3801916932907348,
+      "grad_norm": 1.5492312533670054,
+      "learning_rate": 9.978417901361958e-06,
+      "loss": 0.2046,
+      "step": 952
+    },
+    {
+      "epoch": 0.38099041533546324,
+      "grad_norm": 1.7625259362995305,
+      "learning_rate": 9.977984419802686e-06,
+      "loss": 0.2085,
+      "step": 954
+    },
+    {
+      "epoch": 0.3817891373801917,
+      "grad_norm": 1.3629858192513753,
+      "learning_rate": 9.977546637734557e-06,
+      "loss": 0.2026,
+      "step": 956
+    },
+    {
+      "epoch": 0.38258785942492013,
+      "grad_norm": 1.6553817872981966,
+      "learning_rate": 9.97710455553577e-06,
+      "loss": 0.1986,
+      "step": 958
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 1.6709973514366674,
+      "learning_rate": 9.976658173588244e-06,
+      "loss": 0.1885,
+      "step": 960
+    },
+    {
+      "epoch": 0.384185303514377,
+      "grad_norm": 1.5781155572362893,
+      "learning_rate": 9.976207492277612e-06,
+      "loss": 0.1978,
+      "step": 962
+    },
+    {
+      "epoch": 0.38498402555910544,
+      "grad_norm": 1.5471644306914103,
+      "learning_rate": 9.97575251199322e-06,
+      "loss": 0.2062,
+      "step": 964
+    },
+    {
+      "epoch": 0.38578274760383385,
+      "grad_norm": 1.6210596494320104,
+      "learning_rate": 9.975293233128128e-06,
+      "loss": 0.211,
+      "step": 966
+    },
+    {
+      "epoch": 0.3865814696485623,
+      "grad_norm": 1.4763420559956302,
+      "learning_rate": 9.974829656079106e-06,
+      "loss": 0.2018,
+      "step": 968
+    },
+    {
+      "epoch": 0.38738019169329074,
+      "grad_norm": 1.6383352001216414,
+      "learning_rate": 9.974361781246647e-06,
+      "loss": 0.2094,
+      "step": 970
+    },
+    {
+      "epoch": 0.38817891373801916,
+      "grad_norm": 1.7933627710025113,
+      "learning_rate": 9.973889609034945e-06,
+      "loss": 0.2163,
+      "step": 972
+    },
+    {
+      "epoch": 0.3889776357827476,
+      "grad_norm": 1.5400540527111912,
+      "learning_rate": 9.973413139851918e-06,
+      "loss": 0.1863,
+      "step": 974
+    },
+    {
+      "epoch": 0.38977635782747605,
+      "grad_norm": 1.3560103646758506,
+      "learning_rate": 9.972932374109184e-06,
+      "loss": 0.1959,
+      "step": 976
+    },
+    {
+      "epoch": 0.39057507987220447,
+      "grad_norm": 1.9026289683010842,
+      "learning_rate": 9.972447312222084e-06,
+      "loss": 0.1825,
+      "step": 978
+    },
+    {
+      "epoch": 0.3913738019169329,
+      "grad_norm": 1.5573821099296685,
+      "learning_rate": 9.971957954609663e-06,
+      "loss": 0.206,
+      "step": 980
+    },
+    {
+      "epoch": 0.39217252396166136,
+      "grad_norm": 1.463153557338603,
+      "learning_rate": 9.971464301694683e-06,
+      "loss": 0.2139,
+      "step": 982
+    },
+    {
+      "epoch": 0.3929712460063898,
+      "grad_norm": 1.4361640616504048,
+      "learning_rate": 9.97096635390361e-06,
+      "loss": 0.2039,
+      "step": 984
+    },
+    {
+      "epoch": 0.3937699680511182,
+      "grad_norm": 1.3944298003112567,
+      "learning_rate": 9.970464111666627e-06,
+      "loss": 0.2078,
+      "step": 986
+    },
+    {
+      "epoch": 0.39456869009584666,
+      "grad_norm": 1.427105562837558,
+      "learning_rate": 9.969957575417621e-06,
+      "loss": 0.1837,
+      "step": 988
+    },
+    {
+      "epoch": 0.3953674121405751,
+      "grad_norm": 1.520588415638903,
+      "learning_rate": 9.969446745594193e-06,
+      "loss": 0.2239,
+      "step": 990
+    },
+    {
+      "epoch": 0.3961661341853035,
+      "grad_norm": 2.079367024655261,
+      "learning_rate": 9.968931622637652e-06,
+      "loss": 0.1997,
+      "step": 992
+    },
+    {
+      "epoch": 0.39696485623003197,
+      "grad_norm": 1.5129459484138021,
+      "learning_rate": 9.968412206993015e-06,
+      "loss": 0.1953,
+      "step": 994
+    },
+    {
+      "epoch": 0.3977635782747604,
+      "grad_norm": 1.478216675491551,
+      "learning_rate": 9.967888499109008e-06,
+      "loss": 0.2081,
+      "step": 996
+    },
+    {
+      "epoch": 0.3985623003194888,
+      "grad_norm": 1.5236800584109207,
+      "learning_rate": 9.967360499438067e-06,
+      "loss": 0.2043,
+      "step": 998
+    },
+    {
+      "epoch": 0.3993610223642173,
+      "grad_norm": 1.4639393772316662,
+      "learning_rate": 9.966828208436332e-06,
+      "loss": 0.2035,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3993610223642173,
+      "eval_loss": 0.1855674833059311,
+      "eval_runtime": 418.9403,
+      "eval_samples_per_second": 42.505,
+      "eval_steps_per_second": 5.313,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4001597444089457,
+      "grad_norm": 1.925447602175885,
+      "learning_rate": 9.966291626563651e-06,
+      "loss": 0.2141,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4009584664536741,
+      "grad_norm": 1.5342974238522853,
+      "learning_rate": 9.965750754283583e-06,
+      "loss": 0.1967,
+      "step": 1004
+    },
+    {
+      "epoch": 0.40175718849840253,
+      "grad_norm": 1.365762174319038,
+      "learning_rate": 9.96520559206339e-06,
+      "loss": 0.1787,
+      "step": 1006
+    },
+    {
+      "epoch": 0.402555910543131,
+      "grad_norm": 1.6415643245446048,
+      "learning_rate": 9.96465614037404e-06,
+      "loss": 0.2094,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4033546325878594,
+      "grad_norm": 1.487881148105119,
+      "learning_rate": 9.964102399690206e-06,
+      "loss": 0.2296,
+      "step": 1010
+    },
+    {
+      "epoch": 0.40415335463258784,
+      "grad_norm": 1.3885469945101532,
+      "learning_rate": 9.96354437049027e-06,
+      "loss": 0.1953,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4049520766773163,
+      "grad_norm": 1.5376182595234218,
+      "learning_rate": 9.962982053256317e-06,
+      "loss": 0.2067,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4057507987220447,
+      "grad_norm": 1.5925253053184176,
+      "learning_rate": 9.962415448474134e-06,
+      "loss": 0.2034,
+      "step": 1016
+    },
+    {
+      "epoch": 0.40654952076677314,
+      "grad_norm": 1.5707593570512013,
+      "learning_rate": 9.961844556633216e-06,
+      "loss": 0.2107,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4073482428115016,
+      "grad_norm": 1.538254717520436,
+      "learning_rate": 9.961269378226756e-06,
+      "loss": 0.2014,
+      "step": 1020
+    },
+    {
+      "epoch": 0.40814696485623003,
+      "grad_norm": 1.5245585221099645,
+      "learning_rate": 9.960689913751658e-06,
+      "loss": 0.2002,
+      "step": 1022
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 1.4248623479875928,
+      "learning_rate": 9.960106163708522e-06,
+      "loss": 0.1883,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4097444089456869,
+      "grad_norm": 1.7412362384789994,
+      "learning_rate": 9.959518128601657e-06,
+      "loss": 0.218,
+      "step": 1026
+    },
+    {
+      "epoch": 0.41054313099041534,
+      "grad_norm": 1.4925072692951855,
+      "learning_rate": 9.958925808939066e-06,
+      "loss": 0.2111,
+      "step": 1028
+    },
+    {
+      "epoch": 0.41134185303514376,
+      "grad_norm": 1.4973115296085746,
+      "learning_rate": 9.958329205232456e-06,
+      "loss": 0.2059,
+      "step": 1030
+    },
+    {
+      "epoch": 0.41214057507987223,
+      "grad_norm": 1.468856647822638,
+      "learning_rate": 9.95772831799724e-06,
+      "loss": 0.2093,
+      "step": 1032
+    },
+    {
+      "epoch": 0.41293929712460065,
+      "grad_norm": 1.4085028520516685,
+      "learning_rate": 9.957123147752527e-06,
+      "loss": 0.2116,
+      "step": 1034
+    },
+    {
+      "epoch": 0.41373801916932906,
+      "grad_norm": 1.3478745438364825,
+      "learning_rate": 9.956513695021126e-06,
+      "loss": 0.2144,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4145367412140575,
+      "grad_norm": 1.414787678483826,
+      "learning_rate": 9.955899960329546e-06,
+      "loss": 0.2114,
+      "step": 1038
+    },
+    {
+      "epoch": 0.41533546325878595,
+      "grad_norm": 1.4758641440286802,
+      "learning_rate": 9.955281944207998e-06,
+      "loss": 0.2114,
+      "step": 1040
+    },
+    {
+      "epoch": 0.41613418530351437,
+      "grad_norm": 1.3808751418519412,
+      "learning_rate": 9.95465964719039e-06,
+      "loss": 0.1996,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4169329073482428,
+      "grad_norm": 1.3552203407425432,
+      "learning_rate": 9.954033069814324e-06,
+      "loss": 0.2197,
+      "step": 1044
+    },
+    {
+      "epoch": 0.41773162939297126,
+      "grad_norm": 1.7431939971291783,
+      "learning_rate": 9.953402212621107e-06,
+      "loss": 0.1766,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4185303514376997,
+      "grad_norm": 1.2999138749816357,
+      "learning_rate": 9.95276707615574e-06,
+      "loss": 0.1742,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4193290734824281,
+      "grad_norm": 1.4945842460673373,
+      "learning_rate": 9.952127660966919e-06,
+      "loss": 0.2246,
+      "step": 1050
+    },
+    {
+      "epoch": 0.42012779552715657,
+      "grad_norm": 1.5152322746280356,
+      "learning_rate": 9.95148396760704e-06,
+      "loss": 0.2192,
+      "step": 1052
+    },
+    {
+      "epoch": 0.420926517571885,
+      "grad_norm": 1.4230040236709558,
+      "learning_rate": 9.950835996632193e-06,
+      "loss": 0.1982,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4217252396166134,
+      "grad_norm": 1.3823627265224154,
+      "learning_rate": 9.950183748602164e-06,
+      "loss": 0.2143,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4225239616613419,
+      "grad_norm": 1.6379038373994437,
+      "learning_rate": 9.949527224080434e-06,
+      "loss": 0.2006,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4233226837060703,
+      "grad_norm": 1.4280283780098375,
+      "learning_rate": 9.948866423634178e-06,
+      "loss": 0.1841,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4241214057507987,
+      "grad_norm": 1.417933520198746,
+      "learning_rate": 9.948201347834265e-06,
+      "loss": 0.2032,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4249201277955272,
+      "grad_norm": 1.3378469760405771,
+      "learning_rate": 9.947531997255256e-06,
+      "loss": 0.1926,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4257188498402556,
+      "grad_norm": 1.5459222264184573,
+      "learning_rate": 9.94685837247541e-06,
+      "loss": 0.2052,
+      "step": 1066
+    },
+    {
+      "epoch": 0.426517571884984,
+      "grad_norm": 1.501440385232651,
+      "learning_rate": 9.946180474076675e-06,
+      "loss": 0.2052,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4273162939297125,
+      "grad_norm": 1.5358561082091893,
+      "learning_rate": 9.945498302644687e-06,
+      "loss": 0.2034,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4281150159744409,
+      "grad_norm": 1.6969353624664514,
+      "learning_rate": 9.944811858768782e-06,
+      "loss": 0.2141,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4289137380191693,
+      "grad_norm": 1.4361329940984802,
+      "learning_rate": 9.944121143041982e-06,
+      "loss": 0.1955,
+      "step": 1074
+    },
+    {
+      "epoch": 0.42971246006389774,
+      "grad_norm": 1.4944093870984845,
+      "learning_rate": 9.943426156061e-06,
+      "loss": 0.1968,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4305111821086262,
+      "grad_norm": 1.265035184804792,
+      "learning_rate": 9.942726898426238e-06,
+      "loss": 0.2113,
+      "step": 1078
+    },
+    {
+      "epoch": 0.43130990415335463,
+      "grad_norm": 1.4364726173362738,
+      "learning_rate": 9.94202337074179e-06,
+      "loss": 0.2034,
+      "step": 1080
+    },
+    {
+      "epoch": 0.43210862619808305,
+      "grad_norm": 1.4034504271470043,
+      "learning_rate": 9.941315573615437e-06,
+      "loss": 0.1783,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4329073482428115,
+      "grad_norm": 1.4068580877161003,
+      "learning_rate": 9.940603507658649e-06,
+      "loss": 0.2071,
+      "step": 1084
+    },
+    {
+      "epoch": 0.43370607028753994,
+      "grad_norm": 1.4617297171517643,
+      "learning_rate": 9.939887173486583e-06,
+      "loss": 0.2086,
+      "step": 1086
+    },
+    {
+      "epoch": 0.43450479233226835,
+      "grad_norm": 1.4870587568723221,
+      "learning_rate": 9.939166571718086e-06,
+      "loss": 0.2191,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4353035143769968,
+      "grad_norm": 1.3330964975081798,
+      "learning_rate": 9.938441702975689e-06,
+      "loss": 0.2105,
+      "step": 1090
+    },
+    {
+      "epoch": 0.43610223642172524,
+      "grad_norm": 1.4071934617523967,
+      "learning_rate": 9.93771256788561e-06,
+      "loss": 0.1986,
+      "step": 1092
+    },
+    {
+      "epoch": 0.43690095846645366,
+      "grad_norm": 1.4956330891164953,
+      "learning_rate": 9.936979167077754e-06,
+      "loss": 0.2242,
+      "step": 1094
+    },
+    {
+      "epoch": 0.43769968051118213,
+      "grad_norm": 1.500694287898403,
+      "learning_rate": 9.936241501185706e-06,
+      "loss": 0.213,
+      "step": 1096
+    },
+    {
+      "epoch": 0.43849840255591055,
+      "grad_norm": 1.431648588389163,
+      "learning_rate": 9.935499570846746e-06,
+      "loss": 0.1994,
+      "step": 1098
+    },
+    {
+      "epoch": 0.43929712460063897,
+      "grad_norm": 1.3419204436535992,
+      "learning_rate": 9.934753376701827e-06,
+      "loss": 0.1917,
+      "step": 1100
+    },
+    {
+      "epoch": 0.44009584664536744,
+      "grad_norm": 1.2901653329852891,
+      "learning_rate": 9.934002919395593e-06,
+      "loss": 0.1923,
+      "step": 1102
+    },
+    {
+      "epoch": 0.44089456869009586,
+      "grad_norm": 1.6677348530796035,
+      "learning_rate": 9.933248199576366e-06,
+      "loss": 0.2037,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4416932907348243,
+      "grad_norm": 1.380349679623337,
+      "learning_rate": 9.932489217896154e-06,
+      "loss": 0.1909,
+      "step": 1106
+    },
+    {
+      "epoch": 0.4424920127795527,
+      "grad_norm": 1.4203485606081072,
+      "learning_rate": 9.931725975010647e-06,
+      "loss": 0.1934,
+      "step": 1108
+    },
+    {
+      "epoch": 0.44329073482428116,
+      "grad_norm": 1.4039574653298539,
+      "learning_rate": 9.930958471579212e-06,
+      "loss": 0.2137,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4440894568690096,
+      "grad_norm": 1.4554296915268317,
+      "learning_rate": 9.930186708264902e-06,
+      "loss": 0.2081,
+      "step": 1112
+    },
+    {
+      "epoch": 0.444888178913738,
+      "grad_norm": 1.5862456773052678,
+      "learning_rate": 9.929410685734446e-06,
+      "loss": 0.2074,
+      "step": 1114
+    },
+    {
+      "epoch": 0.44568690095846647,
+      "grad_norm": 1.371290557948402,
+      "learning_rate": 9.928630404658255e-06,
+      "loss": 0.1934,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4464856230031949,
+      "grad_norm": 1.5523688713967079,
+      "learning_rate": 9.92784586571042e-06,
+      "loss": 0.2114,
+      "step": 1118
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 1.5913885443768627,
+      "learning_rate": 9.927057069568704e-06,
+      "loss": 0.197,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4480830670926518,
+      "grad_norm": 1.376586499393768,
+      "learning_rate": 9.926264016914555e-06,
+      "loss": 0.2114,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4488817891373802,
+      "grad_norm": 1.3604215439899947,
+      "learning_rate": 9.925466708433097e-06,
+      "loss": 0.1949,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4496805111821086,
+      "grad_norm": 1.4064368128907085,
+      "learning_rate": 9.924665144813128e-06,
+      "loss": 0.1876,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4504792332268371,
+      "grad_norm": 1.4493555350796723,
+      "learning_rate": 9.923859326747125e-06,
+      "loss": 0.166,
+      "step": 1128
+    },
+    {
+      "epoch": 0.4512779552715655,
+      "grad_norm": 1.3005030779658915,
+      "learning_rate": 9.923049254931235e-06,
+      "loss": 0.179,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4520766773162939,
+      "grad_norm": 1.301603097338869,
+      "learning_rate": 9.922234930065286e-06,
+      "loss": 0.1828,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4528753993610224,
+      "grad_norm": 1.424132465075419,
+      "learning_rate": 9.921416352852779e-06,
+      "loss": 0.2027,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4536741214057508,
+      "grad_norm": 1.5184245441350142,
+      "learning_rate": 9.920593524000887e-06,
+      "loss": 0.2163,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4544728434504792,
+      "grad_norm": 1.2649380282890057,
+      "learning_rate": 9.919766444220454e-06,
+      "loss": 0.1952,
+      "step": 1138
+    },
+    {
+      "epoch": 0.45527156549520764,
+      "grad_norm": 1.3192846529035582,
+      "learning_rate": 9.918935114226001e-06,
+      "loss": 0.2147,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4560702875399361,
+      "grad_norm": 1.6014388375199928,
+      "learning_rate": 9.91809953473572e-06,
+      "loss": 0.209,
+      "step": 1142
+    },
+    {
+      "epoch": 0.45686900958466453,
+      "grad_norm": 1.2763233267206437,
+      "learning_rate": 9.917259706471469e-06,
+      "loss": 0.1894,
+      "step": 1144
+    },
+    {
+      "epoch": 0.45766773162939295,
+      "grad_norm": 1.355011355722016,
+      "learning_rate": 9.916415630158782e-06,
+      "loss": 0.2019,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4584664536741214,
+      "grad_norm": 1.420940628734508,
+      "learning_rate": 9.915567306526863e-06,
+      "loss": 0.2176,
+      "step": 1148
+    },
+    {
+      "epoch": 0.45926517571884984,
+      "grad_norm": 1.3488193076357031,
+      "learning_rate": 9.914714736308582e-06,
+      "loss": 0.2032,
+      "step": 1150
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 1.288799255340392,
+      "learning_rate": 9.913857920240481e-06,
+      "loss": 0.2077,
+      "step": 1152
+    },
+    {
+      "epoch": 0.46086261980830673,
+      "grad_norm": 1.3786756620906035,
+      "learning_rate": 9.912996859062764e-06,
+      "loss": 0.2113,
+      "step": 1154
+    },
+    {
+      "epoch": 0.46166134185303515,
+      "grad_norm": 1.50469580343903,
+      "learning_rate": 9.912131553519313e-06,
+      "loss": 0.2073,
+      "step": 1156
+    },
+    {
+      "epoch": 0.46246006389776356,
+      "grad_norm": 1.4700159822947163,
+      "learning_rate": 9.911262004357665e-06,
+      "loss": 0.2053,
+      "step": 1158
+    },
+    {
+      "epoch": 0.46325878594249204,
+      "grad_norm": 1.519205743263198,
+      "learning_rate": 9.91038821232903e-06,
+      "loss": 0.2217,
+      "step": 1160
+    },
+    {
+      "epoch": 0.46405750798722045,
+      "grad_norm": 1.3708099691931215,
+      "learning_rate": 9.909510178188281e-06,
+      "loss": 0.1891,
+      "step": 1162
+    },
+    {
+      "epoch": 0.46485623003194887,
+      "grad_norm": 1.2711919877902211,
+      "learning_rate": 9.90862790269396e-06,
+      "loss": 0.2003,
+      "step": 1164
+    },
+    {
+      "epoch": 0.46565495207667734,
+      "grad_norm": 1.296959216697176,
+      "learning_rate": 9.907741386608267e-06,
+      "loss": 0.1895,
+      "step": 1166
+    },
+    {
+      "epoch": 0.46645367412140576,
+      "grad_norm": 1.4988072467804332,
+      "learning_rate": 9.906850630697068e-06,
+      "loss": 0.229,
+      "step": 1168
+    },
+    {
+      "epoch": 0.4672523961661342,
+      "grad_norm": 1.357985179422065,
+      "learning_rate": 9.905955635729894e-06,
+      "loss": 0.208,
+      "step": 1170
+    },
+    {
+      "epoch": 0.4680511182108626,
+      "grad_norm": 1.3438998637885513,
+      "learning_rate": 9.905056402479933e-06,
+      "loss": 0.1809,
+      "step": 1172
+    },
+    {
+      "epoch": 0.46884984025559107,
+      "grad_norm": 1.2383109161845713,
+      "learning_rate": 9.904152931724043e-06,
+      "loss": 0.1971,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4696485623003195,
+      "grad_norm": 1.5560174102928908,
+      "learning_rate": 9.903245224242732e-06,
+      "loss": 0.2032,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4704472843450479,
+      "grad_norm": 1.38347824637013,
+      "learning_rate": 9.902333280820176e-06,
+      "loss": 0.198,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4712460063897764,
+      "grad_norm": 1.4051889306019656,
+      "learning_rate": 9.901417102244208e-06,
+      "loss": 0.1987,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4720447284345048,
+      "grad_norm": 1.6159750486401823,
+      "learning_rate": 9.90049668930632e-06,
+      "loss": 0.2029,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4728434504792332,
+      "grad_norm": 1.5547809497795395,
+      "learning_rate": 9.899572042801662e-06,
+      "loss": 0.1937,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4736421725239617,
+      "grad_norm": 1.392059461645592,
+      "learning_rate": 9.898643163529041e-06,
+      "loss": 0.1783,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4744408945686901,
+      "grad_norm": 1.4227792657414569,
+      "learning_rate": 9.89771005229092e-06,
+      "loss": 0.1942,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4752396166134185,
+      "grad_norm": 1.525020829545986,
+      "learning_rate": 9.89677270989342e-06,
+      "loss": 0.2135,
+      "step": 1190
+    },
+    {
+      "epoch": 0.476038338658147,
+      "grad_norm": 1.7374484802274244,
+      "learning_rate": 9.895831137146319e-06,
+      "loss": 0.2081,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4768370607028754,
+      "grad_norm": 1.3661453654617517,
+      "learning_rate": 9.894885334863044e-06,
+      "loss": 0.2055,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4776357827476038,
+      "grad_norm": 1.5602720293300305,
+      "learning_rate": 9.893935303860677e-06,
+      "loss": 0.2171,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4784345047923323,
+      "grad_norm": 3.1025231784672216,
+      "learning_rate": 9.892981044959961e-06,
+      "loss": 0.193,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4792332268370607,
+      "grad_norm": 1.262885402820455,
+      "learning_rate": 9.89202255898528e-06,
+      "loss": 0.1916,
+      "step": 1200
+    },
+    {
+      "epoch": 0.48003194888178913,
+      "grad_norm": 1.381018034591307,
+      "learning_rate": 9.891059846764679e-06,
+      "loss": 0.1878,
+      "step": 1202
+    },
+    {
+      "epoch": 0.48083067092651754,
+      "grad_norm": 1.5768051172665962,
+      "learning_rate": 9.89009290912985e-06,
+      "loss": 0.1967,
+      "step": 1204
+    },
+    {
+      "epoch": 0.481629392971246,
+      "grad_norm": 1.463151710058678,
+      "learning_rate": 9.889121746916132e-06,
+      "loss": 0.2269,
+      "step": 1206
+    },
+    {
+      "epoch": 0.48242811501597443,
+      "grad_norm": 1.34065053745843,
+      "learning_rate": 9.888146360962523e-06,
+      "loss": 0.1865,
+      "step": 1208
+    },
+    {
+      "epoch": 0.48322683706070285,
+      "grad_norm": 1.6723543690494453,
+      "learning_rate": 9.887166752111663e-06,
+      "loss": 0.2129,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4840255591054313,
+      "grad_norm": 1.5794398784564423,
+      "learning_rate": 9.88618292120984e-06,
+      "loss": 0.2224,
+      "step": 1212
+    },
+    {
+      "epoch": 0.48482428115015974,
+      "grad_norm": 1.450388456580997,
+      "learning_rate": 9.88519486910699e-06,
+      "loss": 0.1966,
+      "step": 1214
+    },
+    {
+      "epoch": 0.48562300319488816,
+      "grad_norm": 1.7814239798658509,
+      "learning_rate": 9.8842025966567e-06,
+      "loss": 0.1904,
+      "step": 1216
+    },
+    {
+      "epoch": 0.48642172523961663,
+      "grad_norm": 1.7391567788714228,
+      "learning_rate": 9.883206104716198e-06,
+      "loss": 0.2133,
+      "step": 1218
+    },
+    {
+      "epoch": 0.48722044728434505,
+      "grad_norm": 1.442571651767735,
+      "learning_rate": 9.882205394146362e-06,
+      "loss": 0.1989,
+      "step": 1220
+    },
+    {
+      "epoch": 0.48801916932907347,
+      "grad_norm": 1.4088522689829923,
+      "learning_rate": 9.881200465811706e-06,
+      "loss": 0.2175,
+      "step": 1222
+    },
+    {
+      "epoch": 0.48881789137380194,
+      "grad_norm": 1.2413815231676442,
+      "learning_rate": 9.880191320580396e-06,
+      "loss": 0.1705,
+      "step": 1224
+    },
+    {
+      "epoch": 0.48961661341853036,
+      "grad_norm": 1.4154117503900958,
+      "learning_rate": 9.87917795932424e-06,
+      "loss": 0.1817,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4904153354632588,
+      "grad_norm": 1.3393213535614317,
+      "learning_rate": 9.878160382918685e-06,
+      "loss": 0.2009,
+      "step": 1228
+    },
+    {
+      "epoch": 0.49121405750798725,
+      "grad_norm": 1.5366573602302467,
+      "learning_rate": 9.87713859224282e-06,
+      "loss": 0.2339,
+      "step": 1230
+    },
+    {
+      "epoch": 0.49201277955271566,
+      "grad_norm": 1.2826491297833846,
+      "learning_rate": 9.876112588179378e-06,
+      "loss": 0.1837,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4928115015974441,
+      "grad_norm": 1.4232335556523423,
+      "learning_rate": 9.875082371614728e-06,
+      "loss": 0.1985,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4936102236421725,
+      "grad_norm": 1.4944078372691827,
+      "learning_rate": 9.874047943438879e-06,
+      "loss": 0.1777,
+      "step": 1236
+    },
+    {
+      "epoch": 0.49440894568690097,
+      "grad_norm": 1.3263567228958557,
+      "learning_rate": 9.873009304545482e-06,
+      "loss": 0.2043,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4952076677316294,
+      "grad_norm": 1.491665028254892,
+      "learning_rate": 9.87196645583182e-06,
+      "loss": 0.1995,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4960063897763578,
+      "grad_norm": 1.4711866133685396,
+      "learning_rate": 9.870919398198819e-06,
+      "loss": 0.2151,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4968051118210863,
+      "grad_norm": 1.3471791986545025,
+      "learning_rate": 9.869868132551037e-06,
+      "loss": 0.2301,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4976038338658147,
+      "grad_norm": 1.30305885026678,
+      "learning_rate": 9.868812659796669e-06,
+      "loss": 0.1955,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4984025559105431,
+      "grad_norm": 1.3754860324491927,
+      "learning_rate": 9.86775298084754e-06,
+      "loss": 0.2245,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4992012779552716,
+      "grad_norm": 1.2464730265182025,
+      "learning_rate": 9.866689096619117e-06,
+      "loss": 0.1811,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3277366908866204,
+      "learning_rate": 9.865621008030492e-06,
+      "loss": 0.1916,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5007987220447284,
+      "grad_norm": 1.4195649241005022,
+      "learning_rate": 9.864548716004399e-06,
+      "loss": 0.1932,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5015974440894568,
+      "grad_norm": 1.4925132735569981,
+      "learning_rate": 9.863472221467189e-06,
+      "loss": 0.1938,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5023961661341853,
+      "grad_norm": 1.2979288287029658,
+      "learning_rate": 9.862391525348856e-06,
+      "loss": 0.1877,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5031948881789138,
+      "grad_norm": 1.3864119857396864,
+      "learning_rate": 9.861306628583021e-06,
+      "loss": 0.1958,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5039936102236422,
+      "grad_norm": 1.5118691478458204,
+      "learning_rate": 9.86021753210693e-06,
+      "loss": 0.2145,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5047923322683706,
+      "grad_norm": 1.3753508468091324,
+      "learning_rate": 9.85912423686146e-06,
+      "loss": 0.202,
+      "step": 1264
+    },
+    {
+      "epoch": 0.505591054313099,
+      "grad_norm": 1.4184249679671,
+      "learning_rate": 9.858026743791114e-06,
+      "loss": 0.2002,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5063897763578274,
+      "grad_norm": 1.423875732058238,
+      "learning_rate": 9.856925053844025e-06,
+      "loss": 0.2118,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5071884984025559,
+      "grad_norm": 1.4222433141825608,
+      "learning_rate": 9.855819167971946e-06,
+      "loss": 0.1782,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5079872204472844,
+      "grad_norm": 1.3607705696064714,
+      "learning_rate": 9.854709087130261e-06,
+      "loss": 0.2206,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5087859424920128,
+      "grad_norm": 1.4114769617291796,
+      "learning_rate": 9.853594812277973e-06,
+      "loss": 0.214,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5095846645367412,
+      "grad_norm": 1.5913052917009076,
+      "learning_rate": 9.852476344377713e-06,
+      "loss": 0.2109,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5103833865814696,
+      "grad_norm": 1.3774331396577875,
+      "learning_rate": 9.851353684395728e-06,
+      "loss": 0.2027,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 1.2773430420906078,
+      "learning_rate": 9.850226833301893e-06,
+      "loss": 0.1985,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5119808306709265,
+      "grad_norm": 1.6292303107969242,
+      "learning_rate": 9.849095792069701e-06,
+      "loss": 0.2125,
+      "step": 1282
+    },
+    {
+      "epoch": 0.512779552715655,
+      "grad_norm": 1.4215039291593559,
+      "learning_rate": 9.847960561676263e-06,
+      "loss": 0.2037,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5135782747603834,
+      "grad_norm": 1.3067667391370512,
+      "learning_rate": 9.846821143102313e-06,
+      "loss": 0.1677,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5143769968051118,
+      "grad_norm": 1.2923555649063807,
+      "learning_rate": 9.8456775373322e-06,
+      "loss": 0.1866,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5151757188498403,
+      "grad_norm": 1.4943462008704147,
+      "learning_rate": 9.844529745353892e-06,
+      "loss": 0.2126,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5159744408945687,
+      "grad_norm": 1.5684625006318347,
+      "learning_rate": 9.843377768158972e-06,
+      "loss": 0.2084,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5167731629392971,
+      "grad_norm": 1.4914103834569956,
+      "learning_rate": 9.84222160674264e-06,
+      "loss": 0.2153,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5175718849840255,
+      "grad_norm": 1.328585368359251,
+      "learning_rate": 9.841061262103713e-06,
+      "loss": 0.1919,
+      "step": 1296
+    },
+    {
+      "epoch": 0.518370607028754,
+      "grad_norm": 1.3701342649025663,
+      "learning_rate": 9.839896735244615e-06,
+      "loss": 0.1981,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5191693290734825,
+      "grad_norm": 1.5701845518180864,
+      "learning_rate": 9.83872802717139e-06,
+      "loss": 0.1849,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5199680511182109,
+      "grad_norm": 1.3075421535239862,
+      "learning_rate": 9.83755513889369e-06,
+      "loss": 0.1918,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5207667731629393,
+      "grad_norm": 1.5057769635536875,
+      "learning_rate": 9.836378071424782e-06,
+      "loss": 0.2218,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5215654952076677,
+      "grad_norm": 1.3361043392598404,
+      "learning_rate": 9.835196825781539e-06,
+      "loss": 0.1998,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5223642172523961,
+      "grad_norm": 1.3294074227646568,
+      "learning_rate": 9.834011402984447e-06,
+      "loss": 0.1885,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5231629392971247,
+      "grad_norm": 1.5233983111587897,
+      "learning_rate": 9.8328218040576e-06,
+      "loss": 0.2051,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5239616613418531,
+      "grad_norm": 1.263265485856387,
+      "learning_rate": 9.831628030028698e-06,
+      "loss": 0.1898,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5247603833865815,
+      "grad_norm": 1.3267088643171874,
+      "learning_rate": 9.830430081929047e-06,
+      "loss": 0.1816,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5255591054313099,
+      "grad_norm": 1.4118242503209621,
+      "learning_rate": 9.829227960793566e-06,
+      "loss": 0.2206,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5263578274760383,
+      "grad_norm": 1.2848755780312735,
+      "learning_rate": 9.82802166766077e-06,
+      "loss": 0.191,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5271565495207667,
+      "grad_norm": 1.6025918801894552,
+      "learning_rate": 9.826811203572785e-06,
+      "loss": 0.2231,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5279552715654952,
+      "grad_norm": 1.3694781938576936,
+      "learning_rate": 9.82559656957534e-06,
+      "loss": 0.2133,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5287539936102237,
+      "grad_norm": 1.5702346185266045,
+      "learning_rate": 9.824377766717758e-06,
+      "loss": 0.204,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5295527156549521,
+      "grad_norm": 1.4119518722483408,
+      "learning_rate": 9.823154796052974e-06,
+      "loss": 0.1944,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5303514376996805,
+      "grad_norm": 1.2510934475746587,
+      "learning_rate": 9.821927658637518e-06,
+      "loss": 0.1901,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5311501597444089,
+      "grad_norm": 1.23475379515121,
+      "learning_rate": 9.82069635553152e-06,
+      "loss": 0.179,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5319488817891374,
+      "grad_norm": 1.3831688689107673,
+      "learning_rate": 9.819460887798714e-06,
+      "loss": 0.1854,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5327476038338658,
+      "grad_norm": 1.495164379191231,
+      "learning_rate": 9.818221256506421e-06,
+      "loss": 0.2067,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5335463258785943,
+      "grad_norm": 1.4669136010120607,
+      "learning_rate": 9.81697746272557e-06,
+      "loss": 0.2182,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5343450479233227,
+      "grad_norm": 1.3093260674635532,
+      "learning_rate": 9.81572950753068e-06,
+      "loss": 0.2077,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5351437699680511,
+      "grad_norm": 1.4203515470016688,
+      "learning_rate": 9.814477391999868e-06,
+      "loss": 0.1832,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5359424920127795,
+      "grad_norm": 1.2863505009713285,
+      "learning_rate": 9.813221117214842e-06,
+      "loss": 0.2022,
+      "step": 1342
+    },
+    {
+      "epoch": 0.536741214057508,
+      "grad_norm": 1.2084177020870481,
+      "learning_rate": 9.811960684260907e-06,
+      "loss": 0.1963,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5375399361022364,
+      "grad_norm": 1.3072664824588964,
+      "learning_rate": 9.810696094226952e-06,
+      "loss": 0.2205,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5383386581469649,
+      "grad_norm": 1.1614780192043668,
+      "learning_rate": 9.809427348205472e-06,
+      "loss": 0.1875,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5391373801916933,
+      "grad_norm": 1.3175352444594506,
+      "learning_rate": 9.808154447292539e-06,
+      "loss": 0.185,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5399361022364217,
+      "grad_norm": 1.2834953987740787,
+      "learning_rate": 9.80687739258782e-06,
+      "loss": 0.1924,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5407348242811502,
+      "grad_norm": 1.3409325351626424,
+      "learning_rate": 9.805596185194571e-06,
+      "loss": 0.206,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5415335463258786,
+      "grad_norm": 1.367209463101786,
+      "learning_rate": 9.804310826219633e-06,
+      "loss": 0.2099,
+      "step": 1356
+    },
+    {
+      "epoch": 0.542332268370607,
+      "grad_norm": 1.2819870163156417,
+      "learning_rate": 9.803021316773434e-06,
+      "loss": 0.1881,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5431309904153354,
+      "grad_norm": 1.3214326903766342,
+      "learning_rate": 9.801727657969988e-06,
+      "loss": 0.1964,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5439297124600639,
+      "grad_norm": 1.1833472018926297,
+      "learning_rate": 9.800429850926898e-06,
+      "loss": 0.1836,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5447284345047924,
+      "grad_norm": 1.4748511681523404,
+      "learning_rate": 9.799127896765346e-06,
+      "loss": 0.1772,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5455271565495208,
+      "grad_norm": 1.2243926123531605,
+      "learning_rate": 9.797821796610094e-06,
+      "loss": 0.2206,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5463258785942492,
+      "grad_norm": 1.292732092241605,
+      "learning_rate": 9.796511551589492e-06,
+      "loss": 0.1888,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5471246006389776,
+      "grad_norm": 1.348610309652153,
+      "learning_rate": 9.795197162835468e-06,
+      "loss": 0.2045,
+      "step": 1370
+    },
+    {
+      "epoch": 0.547923322683706,
+      "grad_norm": 1.2828128744633993,
+      "learning_rate": 9.79387863148353e-06,
+      "loss": 0.1785,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5487220447284346,
+      "grad_norm": 1.369286359401732,
+      "learning_rate": 9.792555958672762e-06,
+      "loss": 0.1839,
+      "step": 1374
+    },
+    {
+      "epoch": 0.549520766773163,
+      "grad_norm": 1.4014364965242936,
+      "learning_rate": 9.791229145545832e-06,
+      "loss": 0.1972,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5503194888178914,
+      "grad_norm": 1.465996725744811,
+      "learning_rate": 9.789898193248978e-06,
+      "loss": 0.1957,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5511182108626198,
+      "grad_norm": 1.5171558530092926,
+      "learning_rate": 9.788563102932023e-06,
+      "loss": 0.1966,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5519169329073482,
+      "grad_norm": 1.3311267145287,
+      "learning_rate": 9.787223875748353e-06,
+      "loss": 0.2064,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5527156549520766,
+      "grad_norm": 1.3066784419896205,
+      "learning_rate": 9.785880512854937e-06,
+      "loss": 0.1853,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5535143769968051,
+      "grad_norm": 1.2390174296827288,
+      "learning_rate": 9.784533015412311e-06,
+      "loss": 0.1991,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5543130990415336,
+      "grad_norm": 1.306300904073364,
+      "learning_rate": 9.78318138458459e-06,
+      "loss": 0.1946,
+      "step": 1388
+    },
+    {
+      "epoch": 0.555111821086262,
+      "grad_norm": 1.2680623158657511,
+      "learning_rate": 9.781825621539451e-06,
+      "loss": 0.1961,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5559105431309904,
+      "grad_norm": 1.2795990600690346,
+      "learning_rate": 9.78046572744815e-06,
+      "loss": 0.1852,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5567092651757188,
+      "grad_norm": 1.4140574609807046,
+      "learning_rate": 9.779101703485503e-06,
+      "loss": 0.1915,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5575079872204473,
+      "grad_norm": 1.1983648174489345,
+      "learning_rate": 9.7777335508299e-06,
+      "loss": 0.1812,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5583067092651757,
+      "grad_norm": 1.2309594478518777,
+      "learning_rate": 9.776361270663295e-06,
+      "loss": 0.1861,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5591054313099042,
+      "grad_norm": 1.1982821972463766,
+      "learning_rate": 9.77498486417121e-06,
+      "loss": 0.1713,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5599041533546326,
+      "grad_norm": 1.4024839662731965,
+      "learning_rate": 9.77360433254273e-06,
+      "loss": 0.2106,
+      "step": 1402
+    },
+    {
+      "epoch": 0.560702875399361,
+      "grad_norm": 1.2213412954218998,
+      "learning_rate": 9.772219676970502e-06,
+      "loss": 0.1693,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5615015974440895,
+      "grad_norm": 1.4583671258013307,
+      "learning_rate": 9.770830898650739e-06,
+      "loss": 0.2006,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 1.2563510565936002,
+      "learning_rate": 9.769437998783216e-06,
+      "loss": 0.1862,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5630990415335463,
+      "grad_norm": 1.314486055157245,
+      "learning_rate": 9.768040978571265e-06,
+      "loss": 0.1755,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5638977635782748,
+      "grad_norm": 1.3853554706786548,
+      "learning_rate": 9.76663983922178e-06,
+      "loss": 0.2031,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5646964856230032,
+      "grad_norm": 1.2878686521894571,
+      "learning_rate": 9.765234581945215e-06,
+      "loss": 0.2123,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5654952076677316,
+      "grad_norm": 1.252378097297008,
+      "learning_rate": 9.763825207955577e-06,
+      "loss": 0.1935,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5662939297124601,
+      "grad_norm": 1.213815586126145,
+      "learning_rate": 9.762411718470434e-06,
+      "loss": 0.1658,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5670926517571885,
+      "grad_norm": 1.3949685159556597,
+      "learning_rate": 9.760994114710906e-06,
+      "loss": 0.2114,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5678913738019169,
+      "grad_norm": 1.2817914821838783,
+      "learning_rate": 9.759572397901671e-06,
+      "loss": 0.1767,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5686900958466453,
+      "grad_norm": 1.4145245969611204,
+      "learning_rate": 9.758146569270957e-06,
+      "loss": 0.1986,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5694888178913738,
+      "grad_norm": 1.2826653603852296,
+      "learning_rate": 9.756716630050546e-06,
+      "loss": 0.1854,
+      "step": 1426
+    },
+    {
+      "epoch": 0.5702875399361023,
+      "grad_norm": 1.2441705422611444,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.1923,
+      "step": 1428
+    },
+    {
+      "epoch": 0.5710862619808307,
+      "grad_norm": 1.4151560859787753,
+      "learning_rate": 9.75384442478551e-06,
+      "loss": 0.2145,
+      "step": 1430
+    },
+    {
+      "epoch": 0.5718849840255591,
+      "grad_norm": 1.2387933950813024,
+      "learning_rate": 9.7524021612222e-06,
+      "loss": 0.1804,
+      "step": 1432
+    },
+    {
+      "epoch": 0.5726837060702875,
+      "grad_norm": 1.395550176337113,
+      "learning_rate": 9.75095579203182e-06,
+      "loss": 0.203,
+      "step": 1434
+    },
+    {
+      "epoch": 0.5734824281150159,
+      "grad_norm": 1.4116044575628213,
+      "learning_rate": 9.749505318463894e-06,
+      "loss": 0.2111,
+      "step": 1436
+    },
+    {
+      "epoch": 0.5742811501597445,
+      "grad_norm": 1.2825403647392892,
+      "learning_rate": 9.748050741771498e-06,
+      "loss": 0.1891,
+      "step": 1438
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 1.3389504388761955,
+      "learning_rate": 9.746592063211247e-06,
+      "loss": 0.1989,
+      "step": 1440
+    },
+    {
+      "epoch": 0.5758785942492013,
+      "grad_norm": 1.2944657988503898,
+      "learning_rate": 9.7451292840433e-06,
+      "loss": 0.1986,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5766773162939297,
+      "grad_norm": 1.373583553210306,
+      "learning_rate": 9.743662405531361e-06,
+      "loss": 0.1896,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5774760383386581,
+      "grad_norm": 1.3216412289151287,
+      "learning_rate": 9.742191428942677e-06,
+      "loss": 0.1841,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5782747603833865,
+      "grad_norm": 1.4646317495074423,
+      "learning_rate": 9.74071635554803e-06,
+      "loss": 0.1987,
+      "step": 1448
+    },
+    {
+      "epoch": 0.579073482428115,
+      "grad_norm": 1.3041923646858569,
+      "learning_rate": 9.739237186621747e-06,
+      "loss": 0.1825,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5798722044728435,
+      "grad_norm": 1.3593275898591164,
+      "learning_rate": 9.737753923441689e-06,
+      "loss": 0.1859,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5806709265175719,
+      "grad_norm": 1.2580533184275235,
+      "learning_rate": 9.736266567289255e-06,
+      "loss": 0.1969,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5814696485623003,
+      "grad_norm": 1.325474938146526,
+      "learning_rate": 9.73477511944938e-06,
+      "loss": 0.2043,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5822683706070287,
+      "grad_norm": 1.3507712058043326,
+      "learning_rate": 9.733279581210535e-06,
+      "loss": 0.1979,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5830670926517572,
+      "grad_norm": 1.2160048011358895,
+      "learning_rate": 9.731779953864725e-06,
+      "loss": 0.197,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5838658146964856,
+      "grad_norm": 1.3331470604472404,
+      "learning_rate": 9.730276238707486e-06,
+      "loss": 0.1808,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5846645367412141,
+      "grad_norm": 1.2875278440764244,
+      "learning_rate": 9.728768437037882e-06,
+      "loss": 0.1728,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5854632587859425,
+      "grad_norm": 1.3527080702439602,
+      "learning_rate": 9.72725655015852e-06,
+      "loss": 0.1979,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5862619808306709,
+      "grad_norm": 1.2097235904416872,
+      "learning_rate": 9.725740579375518e-06,
+      "loss": 0.1863,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5870607028753994,
+      "grad_norm": 1.3817835609057318,
+      "learning_rate": 9.724220525998538e-06,
+      "loss": 0.1966,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5878594249201278,
+      "grad_norm": 1.374598787526082,
+      "learning_rate": 9.722696391340762e-06,
+      "loss": 0.2084,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5886581469648562,
+      "grad_norm": 1.1682542939498242,
+      "learning_rate": 9.721168176718896e-06,
+      "loss": 0.1819,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5894568690095847,
+      "grad_norm": 1.3405797462449764,
+      "learning_rate": 9.719635883453175e-06,
+      "loss": 0.2133,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5902555910543131,
+      "grad_norm": 1.3482283589788477,
+      "learning_rate": 9.718099512867355e-06,
+      "loss": 0.1737,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5910543130990416,
+      "grad_norm": 1.3206877696300103,
+      "learning_rate": 9.716559066288716e-06,
+      "loss": 0.1913,
+      "step": 1480
+    },
+    {
+      "epoch": 0.59185303514377,
+      "grad_norm": 1.3089066804848661,
+      "learning_rate": 9.715014545048059e-06,
+      "loss": 0.189,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5926517571884984,
+      "grad_norm": 1.1772602990825523,
+      "learning_rate": 9.713465950479704e-06,
+      "loss": 0.1634,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5934504792332268,
+      "grad_norm": 1.6415957840833397,
+      "learning_rate": 9.711913283921488e-06,
+      "loss": 0.1986,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5942492012779552,
+      "grad_norm": 1.3340911692430844,
+      "learning_rate": 9.710356546714774e-06,
+      "loss": 0.1762,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5950479233226837,
+      "grad_norm": 1.19131607954315,
+      "learning_rate": 9.708795740204431e-06,
+      "loss": 0.1807,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5958466453674122,
+      "grad_norm": 1.377988455402913,
+      "learning_rate": 9.70723086573885e-06,
+      "loss": 0.1939,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5966453674121406,
+      "grad_norm": 1.25358131516604,
+      "learning_rate": 9.705661924669937e-06,
+      "loss": 0.1792,
+      "step": 1494
+    },
+    {
+      "epoch": 0.597444089456869,
+      "grad_norm": 1.2404137677794904,
+      "learning_rate": 9.704088918353108e-06,
+      "loss": 0.1687,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5982428115015974,
+      "grad_norm": 1.2604619975409577,
+      "learning_rate": 9.70251184814729e-06,
+      "loss": 0.1939,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5990415335463258,
+      "grad_norm": 1.363988707372948,
+      "learning_rate": 9.700930715414923e-06,
+      "loss": 0.1931,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5990415335463258,
+      "eval_loss": 0.1750628650188446,
+      "eval_runtime": 416.5473,
+      "eval_samples_per_second": 42.749,
+      "eval_steps_per_second": 5.344,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5998402555910544,
+      "grad_norm": 1.5877690610897348,
+      "learning_rate": 9.69934552152196e-06,
+      "loss": 0.1771,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6006389776357828,
+      "grad_norm": 1.3591700844899557,
+      "learning_rate": 9.697756267837856e-06,
+      "loss": 0.2061,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6014376996805112,
+      "grad_norm": 1.7943521151008524,
+      "learning_rate": 9.696162955735577e-06,
+      "loss": 0.1918,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6022364217252396,
+      "grad_norm": 1.3757909830144823,
+      "learning_rate": 9.694565586591595e-06,
+      "loss": 0.1972,
+      "step": 1508
+    },
+    {
+      "epoch": 0.603035143769968,
+      "grad_norm": 1.3887040239577928,
+      "learning_rate": 9.692964161785885e-06,
+      "loss": 0.1861,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6038338658146964,
+      "grad_norm": 1.1809219692093975,
+      "learning_rate": 9.691358682701927e-06,
+      "loss": 0.1718,
+      "step": 1512
+    },
+    {
+      "epoch": 0.604632587859425,
+      "grad_norm": 1.3982270520799365,
+      "learning_rate": 9.689749150726705e-06,
+      "loss": 0.1747,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6054313099041534,
+      "grad_norm": 1.3379960842286045,
+      "learning_rate": 9.688135567250701e-06,
+      "loss": 0.1977,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6062300319488818,
+      "grad_norm": 1.3650657458967188,
+      "learning_rate": 9.6865179336679e-06,
+      "loss": 0.234,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6070287539936102,
+      "grad_norm": 1.220639149821231,
+      "learning_rate": 9.684896251375784e-06,
+      "loss": 0.1832,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6078274760383386,
+      "grad_norm": 1.3205908355697185,
+      "learning_rate": 9.683270521775334e-06,
+      "loss": 0.1877,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6086261980830671,
+      "grad_norm": 1.1383918655715137,
+      "learning_rate": 9.681640746271026e-06,
+      "loss": 0.1849,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6094249201277955,
+      "grad_norm": 1.2899856170845478,
+      "learning_rate": 9.680006926270833e-06,
+      "loss": 0.1879,
+      "step": 1526
+    },
+    {
+      "epoch": 0.610223642172524,
+      "grad_norm": 1.3318559141238482,
+      "learning_rate": 9.678369063186224e-06,
+      "loss": 0.197,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6110223642172524,
+      "grad_norm": 1.1812228890278815,
+      "learning_rate": 9.676727158432153e-06,
+      "loss": 0.1913,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6118210862619808,
+      "grad_norm": 1.4859051313262472,
+      "learning_rate": 9.675081213427076e-06,
+      "loss": 0.2113,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6126198083067093,
+      "grad_norm": 1.297766937816727,
+      "learning_rate": 9.673431229592928e-06,
+      "loss": 0.1929,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 1.2317306535256243,
+      "learning_rate": 9.671777208355146e-06,
+      "loss": 0.2108,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6142172523961661,
+      "grad_norm": 1.2840356619037692,
+      "learning_rate": 9.670119151142644e-06,
+      "loss": 0.1853,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6150159744408946,
+      "grad_norm": 1.2450276467812322,
+      "learning_rate": 9.668457059387828e-06,
+      "loss": 0.2088,
+      "step": 1540
+    },
+    {
+      "epoch": 0.615814696485623,
+      "grad_norm": 1.3269269430845416,
+      "learning_rate": 9.66679093452659e-06,
+      "loss": 0.2082,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6166134185303515,
+      "grad_norm": 1.1360832482053789,
+      "learning_rate": 9.665120777998303e-06,
+      "loss": 0.1941,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6174121405750799,
+      "grad_norm": 1.2899083518509165,
+      "learning_rate": 9.663446591245825e-06,
+      "loss": 0.1761,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6182108626198083,
+      "grad_norm": 1.352813774180764,
+      "learning_rate": 9.661768375715493e-06,
+      "loss": 0.1922,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6190095846645367,
+      "grad_norm": 1.2861565795443108,
+      "learning_rate": 9.660086132857132e-06,
+      "loss": 0.2052,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6198083067092651,
+      "grad_norm": 1.5903917159408412,
+      "learning_rate": 9.658399864124037e-06,
+      "loss": 0.2057,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6206070287539937,
+      "grad_norm": 1.30967345968437,
+      "learning_rate": 9.656709570972987e-06,
+      "loss": 0.1925,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6214057507987221,
+      "grad_norm": 1.328489492989945,
+      "learning_rate": 9.655015254864236e-06,
+      "loss": 0.204,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6222044728434505,
+      "grad_norm": 1.6209499481494345,
+      "learning_rate": 9.653316917261511e-06,
+      "loss": 0.2084,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6230031948881789,
+      "grad_norm": 1.40907634938526,
+      "learning_rate": 9.65161455963202e-06,
+      "loss": 0.2028,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6238019169329073,
+      "grad_norm": 1.1340460294386678,
+      "learning_rate": 9.649908183446432e-06,
+      "loss": 0.1703,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6246006389776357,
+      "grad_norm": 1.2652621838230012,
+      "learning_rate": 9.648197790178902e-06,
+      "loss": 0.1775,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6253993610223643,
+      "grad_norm": 1.1637259399366224,
+      "learning_rate": 9.646483381307047e-06,
+      "loss": 0.1786,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6261980830670927,
+      "grad_norm": 1.3590258772922685,
+      "learning_rate": 9.64476495831195e-06,
+      "loss": 0.1957,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6269968051118211,
+      "grad_norm": 1.5133394068843713,
+      "learning_rate": 9.643042522678172e-06,
+      "loss": 0.2086,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6277955271565495,
+      "grad_norm": 1.5150111988589272,
+      "learning_rate": 9.641316075893731e-06,
+      "loss": 0.2025,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6285942492012779,
+      "grad_norm": 1.2465479067467933,
+      "learning_rate": 9.639585619450116e-06,
+      "loss": 0.1887,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6293929712460063,
+      "grad_norm": 1.2608367845246782,
+      "learning_rate": 9.637851154842279e-06,
+      "loss": 0.1908,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6301916932907349,
+      "grad_norm": 1.228578420129268,
+      "learning_rate": 9.636112683568633e-06,
+      "loss": 0.182,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6309904153354633,
+      "grad_norm": 1.2794906854021242,
+      "learning_rate": 9.63437020713105e-06,
+      "loss": 0.1891,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6317891373801917,
+      "grad_norm": 1.2079726286437402,
+      "learning_rate": 9.632623727034868e-06,
+      "loss": 0.1918,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6325878594249201,
+      "grad_norm": 1.4507684848320332,
+      "learning_rate": 9.630873244788884e-06,
+      "loss": 0.1928,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6333865814696485,
+      "grad_norm": 1.3354253708604187,
+      "learning_rate": 9.629118761905343e-06,
+      "loss": 0.2098,
+      "step": 1586
+    },
+    {
+      "epoch": 0.634185303514377,
+      "grad_norm": 1.2516888814265885,
+      "learning_rate": 9.627360279899958e-06,
+      "loss": 0.1941,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6349840255591054,
+      "grad_norm": 1.193887849956848,
+      "learning_rate": 9.62559780029189e-06,
+      "loss": 0.1811,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6357827476038339,
+      "grad_norm": 1.2324593768790522,
+      "learning_rate": 9.623831324603755e-06,
+      "loss": 0.1896,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6365814696485623,
+      "grad_norm": 1.252327759159766,
+      "learning_rate": 9.62206085436162e-06,
+      "loss": 0.2004,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6373801916932907,
+      "grad_norm": 1.375448278876851,
+      "learning_rate": 9.620286391095004e-06,
+      "loss": 0.2213,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6381789137380192,
+      "grad_norm": 1.2928127323852008,
+      "learning_rate": 9.618507936336878e-06,
+      "loss": 0.184,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 1.292316256258317,
+      "learning_rate": 9.61672549162366e-06,
+      "loss": 0.1974,
+      "step": 1600
+    },
+    {
+      "epoch": 0.639776357827476,
+      "grad_norm": 1.2009672868034973,
+      "learning_rate": 9.61493905849521e-06,
+      "loss": 0.1926,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6405750798722045,
+      "grad_norm": 1.1528776664831937,
+      "learning_rate": 9.61314863849484e-06,
+      "loss": 0.1781,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6413738019169329,
+      "grad_norm": 1.228036340284955,
+      "learning_rate": 9.611354233169305e-06,
+      "loss": 0.1712,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6421725239616614,
+      "grad_norm": 1.2207441407251327,
+      "learning_rate": 9.6095558440688e-06,
+      "loss": 0.187,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6429712460063898,
+      "grad_norm": 1.2828242186591854,
+      "learning_rate": 9.607753472746967e-06,
+      "loss": 0.1847,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6437699680511182,
+      "grad_norm": 1.147124947259627,
+      "learning_rate": 9.605947120760878e-06,
+      "loss": 0.1555,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6445686900958466,
+      "grad_norm": 1.1939260367535258,
+      "learning_rate": 9.604136789671056e-06,
+      "loss": 0.2005,
+      "step": 1614
+    },
+    {
+      "epoch": 0.645367412140575,
+      "grad_norm": 1.20685907752979,
+      "learning_rate": 9.602322481041457e-06,
+      "loss": 0.1927,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6461661341853036,
+      "grad_norm": 1.2690850828036355,
+      "learning_rate": 9.600504196439468e-06,
+      "loss": 0.1789,
+      "step": 1618
+    },
+    {
+      "epoch": 0.646964856230032,
+      "grad_norm": 1.3543205528313849,
+      "learning_rate": 9.59868193743592e-06,
+      "loss": 0.22,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6477635782747604,
+      "grad_norm": 1.3669603475853995,
+      "learning_rate": 9.596855705605069e-06,
+      "loss": 0.1948,
+      "step": 1622
+    },
+    {
+      "epoch": 0.6485623003194888,
+      "grad_norm": 1.3838109830343859,
+      "learning_rate": 9.595025502524609e-06,
+      "loss": 0.1992,
+      "step": 1624
+    },
+    {
+      "epoch": 0.6493610223642172,
+      "grad_norm": 1.1502765370043713,
+      "learning_rate": 9.593191329775663e-06,
+      "loss": 0.1778,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6501597444089456,
+      "grad_norm": 1.2318516996914193,
+      "learning_rate": 9.591353188942782e-06,
+      "loss": 0.2073,
+      "step": 1628
+    },
+    {
+      "epoch": 0.6509584664536742,
+      "grad_norm": 1.6179688640460086,
+      "learning_rate": 9.589511081613947e-06,
+      "loss": 0.2033,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6517571884984026,
+      "grad_norm": 1.1413033654758473,
+      "learning_rate": 9.587665009380565e-06,
+      "loss": 0.1859,
+      "step": 1632
+    },
+    {
+      "epoch": 0.652555910543131,
+      "grad_norm": 1.2785775485264639,
+      "learning_rate": 9.585814973837468e-06,
+      "loss": 0.1959,
+      "step": 1634
+    },
+    {
+      "epoch": 0.6533546325878594,
+      "grad_norm": 1.2901165952115163,
+      "learning_rate": 9.583960976582914e-06,
+      "loss": 0.1962,
+      "step": 1636
+    },
+    {
+      "epoch": 0.6541533546325878,
+      "grad_norm": 1.2679330444702808,
+      "learning_rate": 9.582103019218577e-06,
+      "loss": 0.1907,
+      "step": 1638
+    },
+    {
+      "epoch": 0.6549520766773163,
+      "grad_norm": 1.246571134661727,
+      "learning_rate": 9.580241103349562e-06,
+      "loss": 0.1712,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6557507987220448,
+      "grad_norm": 1.342636088102212,
+      "learning_rate": 9.578375230584384e-06,
+      "loss": 0.1789,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6565495207667732,
+      "grad_norm": 1.173038253799244,
+      "learning_rate": 9.576505402534984e-06,
+      "loss": 0.1717,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6573482428115016,
+      "grad_norm": 1.2394488250110955,
+      "learning_rate": 9.574631620816718e-06,
+      "loss": 0.186,
+      "step": 1646
+    },
+    {
+      "epoch": 0.65814696485623,
+      "grad_norm": 1.4829342645126746,
+      "learning_rate": 9.572753887048353e-06,
+      "loss": 0.1965,
+      "step": 1648
+    },
+    {
+      "epoch": 0.6589456869009584,
+      "grad_norm": 1.2666969584636023,
+      "learning_rate": 9.570872202852077e-06,
+      "loss": 0.181,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6597444089456869,
+      "grad_norm": 1.2265218244462477,
+      "learning_rate": 9.568986569853487e-06,
+      "loss": 0.1811,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6605431309904153,
+      "grad_norm": 1.391040746962543,
+      "learning_rate": 9.56709698968159e-06,
+      "loss": 0.1955,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6613418530351438,
+      "grad_norm": 1.2540103131383575,
+      "learning_rate": 9.565203463968808e-06,
+      "loss": 0.2158,
+      "step": 1656
+    },
+    {
+      "epoch": 0.6621405750798722,
+      "grad_norm": 1.218580930627416,
+      "learning_rate": 9.563305994350966e-06,
+      "loss": 0.2075,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6629392971246006,
+      "grad_norm": 1.2969445843723761,
+      "learning_rate": 9.5614045824673e-06,
+      "loss": 0.174,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6637380191693291,
+      "grad_norm": 1.2080482907459895,
+      "learning_rate": 9.55949922996045e-06,
+      "loss": 0.1718,
+      "step": 1662
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 1.3088014021283991,
+      "learning_rate": 9.557589938476462e-06,
+      "loss": 0.1958,
+      "step": 1664
+    },
+    {
+      "epoch": 0.6653354632587859,
+      "grad_norm": 1.2483532241185216,
+      "learning_rate": 9.555676709664783e-06,
+      "loss": 0.1667,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6661341853035144,
+      "grad_norm": 1.4228507269979516,
+      "learning_rate": 9.55375954517826e-06,
+      "loss": 0.1887,
+      "step": 1668
+    },
+    {
+      "epoch": 0.6669329073482428,
+      "grad_norm": 1.426968884001837,
+      "learning_rate": 9.551838446673144e-06,
+      "loss": 0.1733,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6677316293929713,
+      "grad_norm": 1.238164625762186,
+      "learning_rate": 9.549913415809084e-06,
+      "loss": 0.1909,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6685303514376997,
+      "grad_norm": 1.1947296500798315,
+      "learning_rate": 9.547984454249125e-06,
+      "loss": 0.19,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6693290734824281,
+      "grad_norm": 1.3096866424518478,
+      "learning_rate": 9.546051563659704e-06,
+      "loss": 0.1838,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6701277955271565,
+      "grad_norm": 1.2325979346576457,
+      "learning_rate": 9.54411474571066e-06,
+      "loss": 0.213,
+      "step": 1678
+    },
+    {
+      "epoch": 0.670926517571885,
+      "grad_norm": 1.1989676471273152,
+      "learning_rate": 9.542174002075221e-06,
+      "loss": 0.1815,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6717252396166135,
+      "grad_norm": 1.3573812014117548,
+      "learning_rate": 9.540229334430005e-06,
+      "loss": 0.1897,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6725239616613419,
+      "grad_norm": 1.2474833071465166,
+      "learning_rate": 9.53828074445502e-06,
+      "loss": 0.1846,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6733226837060703,
+      "grad_norm": 1.2370853532394706,
+      "learning_rate": 9.536328233833668e-06,
+      "loss": 0.1926,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6741214057507987,
+      "grad_norm": 1.2326249171384345,
+      "learning_rate": 9.534371804252727e-06,
+      "loss": 0.1971,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6749201277955271,
+      "grad_norm": 1.2299339970845358,
+      "learning_rate": 9.532411457402374e-06,
+      "loss": 0.1806,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6757188498402555,
+      "grad_norm": 1.4027305493123479,
+      "learning_rate": 9.530447194976164e-06,
+      "loss": 0.1939,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6765175718849841,
+      "grad_norm": 1.7551361158569634,
+      "learning_rate": 9.52847901867103e-06,
+      "loss": 0.2109,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6773162939297125,
+      "grad_norm": 1.2968839868364699,
+      "learning_rate": 9.526506930187294e-06,
+      "loss": 0.1987,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6781150159744409,
+      "grad_norm": 1.4078086349831407,
+      "learning_rate": 9.524530931228653e-06,
+      "loss": 0.1989,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6789137380191693,
+      "grad_norm": 1.231887491149701,
+      "learning_rate": 9.522551023502183e-06,
+      "loss": 0.1718,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6797124600638977,
+      "grad_norm": 1.4088570515681005,
+      "learning_rate": 9.520567208718337e-06,
+      "loss": 0.1791,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6805111821086262,
+      "grad_norm": 1.4104721082557625,
+      "learning_rate": 9.518579488590947e-06,
+      "loss": 0.2051,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6813099041533547,
+      "grad_norm": 1.4204595182317563,
+      "learning_rate": 9.516587864837213e-06,
+      "loss": 0.2222,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6821086261980831,
+      "grad_norm": 1.3446005177751752,
+      "learning_rate": 9.51459233917771e-06,
+      "loss": 0.1923,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6829073482428115,
+      "grad_norm": 1.3177179723391306,
+      "learning_rate": 9.512592913336385e-06,
+      "loss": 0.1911,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6837060702875399,
+      "grad_norm": 1.3007833533147979,
+      "learning_rate": 9.510589589040554e-06,
+      "loss": 0.2003,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6845047923322684,
+      "grad_norm": 1.3084903604314289,
+      "learning_rate": 9.508582368020897e-06,
+      "loss": 0.2068,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6853035143769968,
+      "grad_norm": 1.2332676487219392,
+      "learning_rate": 9.506571252011467e-06,
+      "loss": 0.1809,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6861022364217252,
+      "grad_norm": 1.4206295803871527,
+      "learning_rate": 9.504556242749677e-06,
+      "loss": 0.1879,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6869009584664537,
+      "grad_norm": 1.4363115148587187,
+      "learning_rate": 9.502537341976305e-06,
+      "loss": 0.1955,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6876996805111821,
+      "grad_norm": 1.3048263552086317,
+      "learning_rate": 9.500514551435491e-06,
+      "loss": 0.2062,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6884984025559105,
+      "grad_norm": 1.4533257081032092,
+      "learning_rate": 9.498487872874735e-06,
+      "loss": 0.2065,
+      "step": 1724
+    },
+    {
+      "epoch": 0.689297124600639,
+      "grad_norm": 1.2029574969239287,
+      "learning_rate": 9.496457308044895e-06,
+      "loss": 0.1848,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6900958466453674,
+      "grad_norm": 1.2074365689053979,
+      "learning_rate": 9.494422858700188e-06,
+      "loss": 0.1966,
+      "step": 1728
+    },
+    {
+      "epoch": 0.6908945686900958,
+      "grad_norm": 1.2900526030749346,
+      "learning_rate": 9.492384526598188e-06,
+      "loss": 0.1982,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6916932907348243,
+      "grad_norm": 1.2246323016242,
+      "learning_rate": 9.49034231349982e-06,
+      "loss": 0.1981,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6924920127795527,
+      "grad_norm": 1.2646452794419145,
+      "learning_rate": 9.488296221169363e-06,
+      "loss": 0.1748,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6932907348242812,
+      "grad_norm": 1.2901816109550333,
+      "learning_rate": 9.48624625137445e-06,
+      "loss": 0.2003,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6940894568690096,
+      "grad_norm": 1.236520057555359,
+      "learning_rate": 9.484192405886058e-06,
+      "loss": 0.1943,
+      "step": 1738
+    },
+    {
+      "epoch": 0.694888178913738,
+      "grad_norm": 1.3841072722578425,
+      "learning_rate": 9.48213468647852e-06,
+      "loss": 0.2006,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6956869009584664,
+      "grad_norm": 1.2637297258797011,
+      "learning_rate": 9.480073094929507e-06,
+      "loss": 0.1508,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6964856230031949,
+      "grad_norm": 1.3190153343742,
+      "learning_rate": 9.478007633020043e-06,
+      "loss": 0.1791,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6972843450479234,
+      "grad_norm": 1.2377929664861391,
+      "learning_rate": 9.47593830253449e-06,
+      "loss": 0.1919,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6980830670926518,
+      "grad_norm": 1.2951377520887053,
+      "learning_rate": 9.473865105260556e-06,
+      "loss": 0.1931,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6988817891373802,
+      "grad_norm": 1.203360302375175,
+      "learning_rate": 9.471788042989285e-06,
+      "loss": 0.1628,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6996805111821086,
+      "grad_norm": 1.3053684785835005,
+      "learning_rate": 9.469707117515068e-06,
+      "loss": 0.197,
+      "step": 1752
+    },
+    {
+      "epoch": 0.700479233226837,
+      "grad_norm": 1.1639750404721112,
+      "learning_rate": 9.467622330635622e-06,
+      "loss": 0.1811,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7012779552715654,
+      "grad_norm": 1.3446647639072014,
+      "learning_rate": 9.465533684152011e-06,
+      "loss": 0.1931,
+      "step": 1756
+    },
+    {
+      "epoch": 0.702076677316294,
+      "grad_norm": 1.336079168507431,
+      "learning_rate": 9.463441179868626e-06,
+      "loss": 0.1804,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 1.210643436064521,
+      "learning_rate": 9.461344819593194e-06,
+      "loss": 0.1948,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7036741214057508,
+      "grad_norm": 1.2291459699844245,
+      "learning_rate": 9.459244605136775e-06,
+      "loss": 0.1854,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7044728434504792,
+      "grad_norm": 1.5941109492311214,
+      "learning_rate": 9.45714053831375e-06,
+      "loss": 0.2128,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7052715654952076,
+      "grad_norm": 1.1893124518881097,
+      "learning_rate": 9.45503262094184e-06,
+      "loss": 0.1735,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7060702875399361,
+      "grad_norm": 1.244259188061574,
+      "learning_rate": 9.452920854842085e-06,
+      "loss": 0.203,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7068690095846646,
+      "grad_norm": 1.1591101687705918,
+      "learning_rate": 9.45080524183885e-06,
+      "loss": 0.1931,
+      "step": 1770
+    },
+    {
+      "epoch": 0.707667731629393,
+      "grad_norm": 1.185305488469298,
+      "learning_rate": 9.448685783759825e-06,
+      "loss": 0.1826,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7084664536741214,
+      "grad_norm": 1.3935670621941678,
+      "learning_rate": 9.446562482436026e-06,
+      "loss": 0.1882,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7092651757188498,
+      "grad_norm": 1.3143460322695941,
+      "learning_rate": 9.44443533970178e-06,
+      "loss": 0.1815,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7100638977635783,
+      "grad_norm": 1.3400068262690805,
+      "learning_rate": 9.442304357394741e-06,
+      "loss": 0.1968,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7108626198083067,
+      "grad_norm": 1.2627252602067203,
+      "learning_rate": 9.440169537355874e-06,
+      "loss": 0.1837,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7116613418530351,
+      "grad_norm": 1.1054955641627056,
+      "learning_rate": 9.438030881429465e-06,
+      "loss": 0.1603,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7124600638977636,
+      "grad_norm": 1.3680070238884574,
+      "learning_rate": 9.435888391463108e-06,
+      "loss": 0.1996,
+      "step": 1784
+    },
+    {
+      "epoch": 0.713258785942492,
+      "grad_norm": 1.297397045088257,
+      "learning_rate": 9.433742069307714e-06,
+      "loss": 0.1987,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7140575079872205,
+      "grad_norm": 1.2477413014257834,
+      "learning_rate": 9.431591916817503e-06,
+      "loss": 0.1821,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7148562300319489,
+      "grad_norm": 1.3162258234912159,
+      "learning_rate": 9.429437935850003e-06,
+      "loss": 0.2119,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 1.215098930469111,
+      "learning_rate": 9.427280128266049e-06,
+      "loss": 0.1842,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7164536741214057,
+      "grad_norm": 1.2140728281775635,
+      "learning_rate": 9.425118495929788e-06,
+      "loss": 0.1813,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7172523961661342,
+      "grad_norm": 1.0941868316938037,
+      "learning_rate": 9.422953040708662e-06,
+      "loss": 0.1836,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7180511182108626,
+      "grad_norm": 1.2213745174193513,
+      "learning_rate": 9.420783764473418e-06,
+      "loss": 0.1902,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7188498402555911,
+      "grad_norm": 1.1225238944421074,
+      "learning_rate": 9.418610669098114e-06,
+      "loss": 0.1924,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7196485623003195,
+      "grad_norm": 1.194219981117358,
+      "learning_rate": 9.41643375646009e-06,
+      "loss": 0.1901,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7204472843450479,
+      "grad_norm": 1.405501209107951,
+      "learning_rate": 9.41425302844e-06,
+      "loss": 0.1962,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7212460063897763,
+      "grad_norm": 1.2754008408470467,
+      "learning_rate": 9.412068486921786e-06,
+      "loss": 0.1981,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7220447284345048,
+      "grad_norm": 1.153012769564869,
+      "learning_rate": 9.409880133792684e-06,
+      "loss": 0.1601,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7228434504792333,
+      "grad_norm": 1.412330054040255,
+      "learning_rate": 9.407687970943223e-06,
+      "loss": 0.2015,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7236421725239617,
+      "grad_norm": 1.2512273615084581,
+      "learning_rate": 9.405492000267228e-06,
+      "loss": 0.162,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7244408945686901,
+      "grad_norm": 1.3404549541893849,
+      "learning_rate": 9.403292223661811e-06,
+      "loss": 0.1858,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7252396166134185,
+      "grad_norm": 1.2882725353355315,
+      "learning_rate": 9.40108864302737e-06,
+      "loss": 0.1918,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7260383386581469,
+      "grad_norm": 1.2513310169482343,
+      "learning_rate": 9.398881260267589e-06,
+      "loss": 0.1998,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7268370607028753,
+      "grad_norm": 1.337658331922668,
+      "learning_rate": 9.396670077289443e-06,
+      "loss": 0.1932,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7276357827476039,
+      "grad_norm": 1.2367717186372091,
+      "learning_rate": 9.394455096003182e-06,
+      "loss": 0.2068,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7284345047923323,
+      "grad_norm": 1.1564610765311962,
+      "learning_rate": 9.392236318322339e-06,
+      "loss": 0.1801,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7292332268370607,
+      "grad_norm": 1.2112589486368208,
+      "learning_rate": 9.390013746163733e-06,
+      "loss": 0.1837,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7300319488817891,
+      "grad_norm": 1.254583211491624,
+      "learning_rate": 9.387787381447455e-06,
+      "loss": 0.2105,
+      "step": 1828
+    },
+    {
+      "epoch": 0.7308306709265175,
+      "grad_norm": 1.1680394605252384,
+      "learning_rate": 9.385557226096873e-06,
+      "loss": 0.1933,
+      "step": 1830
+    },
+    {
+      "epoch": 0.731629392971246,
+      "grad_norm": 1.1210837547747836,
+      "learning_rate": 9.383323282038632e-06,
+      "loss": 0.1847,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7324281150159745,
+      "grad_norm": 1.2069663644377342,
+      "learning_rate": 9.381085551202648e-06,
+      "loss": 0.1968,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7332268370607029,
+      "grad_norm": 1.2372918873896177,
+      "learning_rate": 9.378844035522112e-06,
+      "loss": 0.1969,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7340255591054313,
+      "grad_norm": 1.1846330324034746,
+      "learning_rate": 9.376598736933478e-06,
+      "loss": 0.191,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7348242811501597,
+      "grad_norm": 1.1755331433603171,
+      "learning_rate": 9.374349657376473e-06,
+      "loss": 0.1647,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7356230031948882,
+      "grad_norm": 1.2996956651624354,
+      "learning_rate": 9.372096798794093e-06,
+      "loss": 0.182,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7364217252396166,
+      "grad_norm": 1.2224015710817062,
+      "learning_rate": 9.36984016313259e-06,
+      "loss": 0.1732,
+      "step": 1844
+    },
+    {
+      "epoch": 0.737220447284345,
+      "grad_norm": 1.196928762655955,
+      "learning_rate": 9.367579752341488e-06,
+      "loss": 0.1863,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7380191693290735,
+      "grad_norm": 1.2553608491551194,
+      "learning_rate": 9.365315568373569e-06,
+      "loss": 0.1825,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7388178913738019,
+      "grad_norm": 1.1728882445017346,
+      "learning_rate": 9.363047613184872e-06,
+      "loss": 0.1767,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7396166134185304,
+      "grad_norm": 1.179558407608193,
+      "learning_rate": 9.360775888734699e-06,
+      "loss": 0.1862,
+      "step": 1852
+    },
+    {
+      "epoch": 0.7404153354632588,
+      "grad_norm": 1.1401251188622028,
+      "learning_rate": 9.358500396985603e-06,
+      "loss": 0.1836,
+      "step": 1854
+    },
+    {
+      "epoch": 0.7412140575079872,
+      "grad_norm": 1.1844349399914762,
+      "learning_rate": 9.356221139903395e-06,
+      "loss": 0.1885,
+      "step": 1856
+    },
+    {
+      "epoch": 0.7420127795527156,
+      "grad_norm": 1.2044582641333985,
+      "learning_rate": 9.353938119457137e-06,
+      "loss": 0.1865,
+      "step": 1858
+    },
+    {
+      "epoch": 0.7428115015974441,
+      "grad_norm": 1.2685669411269742,
+      "learning_rate": 9.351651337619145e-06,
+      "loss": 0.1754,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7436102236421726,
+      "grad_norm": 1.4085265142769048,
+      "learning_rate": 9.349360796364984e-06,
+      "loss": 0.1972,
+      "step": 1862
+    },
+    {
+      "epoch": 0.744408945686901,
+      "grad_norm": 1.3358324906099626,
+      "learning_rate": 9.347066497673462e-06,
+      "loss": 0.1934,
+      "step": 1864
+    },
+    {
+      "epoch": 0.7452076677316294,
+      "grad_norm": 1.1473591638995988,
+      "learning_rate": 9.34476844352664e-06,
+      "loss": 0.1756,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7460063897763578,
+      "grad_norm": 1.1732315132551707,
+      "learning_rate": 9.342466635909815e-06,
+      "loss": 0.1937,
+      "step": 1868
+    },
+    {
+      "epoch": 0.7468051118210862,
+      "grad_norm": 1.0907285192894114,
+      "learning_rate": 9.340161076811539e-06,
+      "loss": 0.1769,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7476038338658147,
+      "grad_norm": 1.3140788804501988,
+      "learning_rate": 9.337851768223589e-06,
+      "loss": 0.2101,
+      "step": 1872
+    },
+    {
+      "epoch": 0.7484025559105432,
+      "grad_norm": 1.2720702400836585,
+      "learning_rate": 9.335538712140997e-06,
+      "loss": 0.1755,
+      "step": 1874
+    },
+    {
+      "epoch": 0.7492012779552716,
+      "grad_norm": 1.1203884095157366,
+      "learning_rate": 9.333221910562022e-06,
+      "loss": 0.1645,
+      "step": 1876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3590498055976206,
+      "learning_rate": 9.330901365488163e-06,
+      "loss": 0.1898,
+      "step": 1878
+    },
+    {
+      "epoch": 0.7507987220447284,
+      "grad_norm": 1.2561659399741483,
+      "learning_rate": 9.328577078924151e-06,
+      "loss": 0.2081,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7515974440894568,
+      "grad_norm": 1.5567261987394814,
+      "learning_rate": 9.326249052877949e-06,
+      "loss": 0.1899,
+      "step": 1882
+    },
+    {
+      "epoch": 0.7523961661341853,
+      "grad_norm": 1.10938481898858,
+      "learning_rate": 9.323917289360755e-06,
+      "loss": 0.1822,
+      "step": 1884
+    },
+    {
+      "epoch": 0.7531948881789138,
+      "grad_norm": 1.199525742926509,
+      "learning_rate": 9.321581790386989e-06,
+      "loss": 0.1838,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7539936102236422,
+      "grad_norm": 1.2183809794152496,
+      "learning_rate": 9.319242557974306e-06,
+      "loss": 0.1813,
+      "step": 1888
+    },
+    {
+      "epoch": 0.7547923322683706,
+      "grad_norm": 1.3361557842251508,
+      "learning_rate": 9.316899594143581e-06,
+      "loss": 0.1923,
+      "step": 1890
+    },
+    {
+      "epoch": 0.755591054313099,
+      "grad_norm": 1.147280166061991,
+      "learning_rate": 9.31455290091891e-06,
+      "loss": 0.1681,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7563897763578274,
+      "grad_norm": 1.2035996373756572,
+      "learning_rate": 9.31220248032762e-06,
+      "loss": 0.1752,
+      "step": 1894
+    },
+    {
+      "epoch": 0.7571884984025559,
+      "grad_norm": 1.2112026467710946,
+      "learning_rate": 9.309848334400247e-06,
+      "loss": 0.1862,
+      "step": 1896
+    },
+    {
+      "epoch": 0.7579872204472844,
+      "grad_norm": 1.345396574129663,
+      "learning_rate": 9.307490465170555e-06,
+      "loss": 0.2152,
+      "step": 1898
+    },
+    {
+      "epoch": 0.7587859424920128,
+      "grad_norm": 1.1731055253334703,
+      "learning_rate": 9.30512887467552e-06,
+      "loss": 0.1845,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7595846645367412,
+      "grad_norm": 1.2248021789089425,
+      "learning_rate": 9.302763564955332e-06,
+      "loss": 0.1896,
+      "step": 1902
+    },
+    {
+      "epoch": 0.7603833865814696,
+      "grad_norm": 1.2159843127230634,
+      "learning_rate": 9.300394538053395e-06,
+      "loss": 0.2073,
+      "step": 1904
+    },
+    {
+      "epoch": 0.7611821086261981,
+      "grad_norm": 1.0782210847536804,
+      "learning_rate": 9.298021796016328e-06,
+      "loss": 0.1965,
+      "step": 1906
+    },
+    {
+      "epoch": 0.7619808306709265,
+      "grad_norm": 1.0674705155900905,
+      "learning_rate": 9.295645340893954e-06,
+      "loss": 0.1828,
+      "step": 1908
+    },
+    {
+      "epoch": 0.762779552715655,
+      "grad_norm": 1.242030763933389,
+      "learning_rate": 9.293265174739304e-06,
+      "loss": 0.185,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7635782747603834,
+      "grad_norm": 1.2869300880171743,
+      "learning_rate": 9.29088129960862e-06,
+      "loss": 0.2072,
+      "step": 1912
+    },
+    {
+      "epoch": 0.7643769968051118,
+      "grad_norm": 1.2987640167915715,
+      "learning_rate": 9.288493717561346e-06,
+      "loss": 0.1797,
+      "step": 1914
+    },
+    {
+      "epoch": 0.7651757188498403,
+      "grad_norm": 1.360583270135005,
+      "learning_rate": 9.286102430660124e-06,
+      "loss": 0.1915,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7659744408945687,
+      "grad_norm": 1.1308395049372801,
+      "learning_rate": 9.283707440970804e-06,
+      "loss": 0.1732,
+      "step": 1918
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 1.1740116097009,
+      "learning_rate": 9.281308750562426e-06,
+      "loss": 0.1954,
+      "step": 1920
+    },
+    {
+      "epoch": 0.7675718849840255,
+      "grad_norm": 1.2021810735396423,
+      "learning_rate": 9.278906361507238e-06,
+      "loss": 0.1941,
+      "step": 1922
+    },
+    {
+      "epoch": 0.768370607028754,
+      "grad_norm": 1.3054645933982225,
+      "learning_rate": 9.276500275880676e-06,
+      "loss": 0.2025,
+      "step": 1924
+    },
+    {
+      "epoch": 0.7691693290734825,
+      "grad_norm": 1.2735690354152704,
+      "learning_rate": 9.274090495761368e-06,
+      "loss": 0.1977,
+      "step": 1926
+    },
+    {
+      "epoch": 0.7699680511182109,
+      "grad_norm": 1.0912094717996301,
+      "learning_rate": 9.271677023231137e-06,
+      "loss": 0.1918,
+      "step": 1928
+    },
+    {
+      "epoch": 0.7707667731629393,
+      "grad_norm": 1.2385601982012775,
+      "learning_rate": 9.269259860375001e-06,
+      "loss": 0.1853,
+      "step": 1930
+    },
+    {
+      "epoch": 0.7715654952076677,
+      "grad_norm": 1.2615706597096068,
+      "learning_rate": 9.266839009281154e-06,
+      "loss": 0.1775,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7723642172523961,
+      "grad_norm": 1.2657440936466628,
+      "learning_rate": 9.264414472040986e-06,
+      "loss": 0.2122,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7731629392971247,
+      "grad_norm": 1.2434750494740934,
+      "learning_rate": 9.261986250749068e-06,
+      "loss": 0.1642,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7739616613418531,
+      "grad_norm": 1.2150467984561792,
+      "learning_rate": 9.259554347503157e-06,
+      "loss": 0.179,
+      "step": 1938
+    },
+    {
+      "epoch": 0.7747603833865815,
+      "grad_norm": 1.0460909251749075,
+      "learning_rate": 9.257118764404183e-06,
+      "loss": 0.1766,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7755591054313099,
+      "grad_norm": 1.2648684414877398,
+      "learning_rate": 9.254679503556261e-06,
+      "loss": 0.1873,
+      "step": 1942
+    },
+    {
+      "epoch": 0.7763578274760383,
+      "grad_norm": 1.0629676893025566,
+      "learning_rate": 9.252236567066686e-06,
+      "loss": 0.1765,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7771565495207667,
+      "grad_norm": 1.1476890935760726,
+      "learning_rate": 9.249789957045921e-06,
+      "loss": 0.1828,
+      "step": 1946
+    },
+    {
+      "epoch": 0.7779552715654952,
+      "grad_norm": 1.255044114071295,
+      "learning_rate": 9.247339675607606e-06,
+      "loss": 0.1962,
+      "step": 1948
+    },
+    {
+      "epoch": 0.7787539936102237,
+      "grad_norm": 1.1772586864322332,
+      "learning_rate": 9.244885724868556e-06,
+      "loss": 0.1948,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7795527156549521,
+      "grad_norm": 1.134211924482269,
+      "learning_rate": 9.242428106948748e-06,
+      "loss": 0.1797,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7803514376996805,
+      "grad_norm": 1.1580694174114012,
+      "learning_rate": 9.239966823971339e-06,
+      "loss": 0.1818,
+      "step": 1954
+    },
+    {
+      "epoch": 0.7811501597444089,
+      "grad_norm": 1.3621884578061603,
+      "learning_rate": 9.23750187806264e-06,
+      "loss": 0.2012,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7819488817891374,
+      "grad_norm": 1.107597745587062,
+      "learning_rate": 9.235033271352132e-06,
+      "loss": 0.1816,
+      "step": 1958
+    },
+    {
+      "epoch": 0.7827476038338658,
+      "grad_norm": 1.1104284440920194,
+      "learning_rate": 9.23256100597246e-06,
+      "loss": 0.1748,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7835463258785943,
+      "grad_norm": 1.0772833846504462,
+      "learning_rate": 9.230085084059428e-06,
+      "loss": 0.1577,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7843450479233227,
+      "grad_norm": 1.5348516969940666,
+      "learning_rate": 9.227605507751998e-06,
+      "loss": 0.1797,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7851437699680511,
+      "grad_norm": 1.1556035368347786,
+      "learning_rate": 9.22512227919229e-06,
+      "loss": 0.1904,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7859424920127795,
+      "grad_norm": 1.1541070739348673,
+      "learning_rate": 9.22263540052558e-06,
+      "loss": 0.1629,
+      "step": 1968
+    },
+    {
+      "epoch": 0.786741214057508,
+      "grad_norm": 1.1977607582272027,
+      "learning_rate": 9.220144873900294e-06,
+      "loss": 0.1852,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7875399361022364,
+      "grad_norm": 1.2010602313619996,
+      "learning_rate": 9.217650701468016e-06,
+      "loss": 0.167,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7883386581469649,
+      "grad_norm": 1.2177543309257506,
+      "learning_rate": 9.215152885383473e-06,
+      "loss": 0.1785,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7891373801916933,
+      "grad_norm": 1.3217713389862689,
+      "learning_rate": 9.212651427804544e-06,
+      "loss": 0.2005,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7899361022364217,
+      "grad_norm": 1.2318521295090967,
+      "learning_rate": 9.210146330892251e-06,
+      "loss": 0.1804,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7907348242811502,
+      "grad_norm": 1.2625115330731185,
+      "learning_rate": 9.20763759681076e-06,
+      "loss": 0.1648,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7915335463258786,
+      "grad_norm": 1.1952075781633953,
+      "learning_rate": 9.205125227727386e-06,
+      "loss": 0.1753,
+      "step": 1982
+    },
+    {
+      "epoch": 0.792332268370607,
+      "grad_norm": 1.2767418757560534,
+      "learning_rate": 9.202609225812572e-06,
+      "loss": 0.1694,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7931309904153354,
+      "grad_norm": 1.2452947191655441,
+      "learning_rate": 9.200089593239911e-06,
+      "loss": 0.2021,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7939297124600639,
+      "grad_norm": 1.2365293461357185,
+      "learning_rate": 9.197566332186125e-06,
+      "loss": 0.1858,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7947284345047924,
+      "grad_norm": 1.2036375025069872,
+      "learning_rate": 9.195039444831076e-06,
+      "loss": 0.1957,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7955271565495208,
+      "grad_norm": 1.2642281456239493,
+      "learning_rate": 9.192508933357753e-06,
+      "loss": 0.1764,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7963258785942492,
+      "grad_norm": 1.1075926832298097,
+      "learning_rate": 9.189974799952283e-06,
+      "loss": 0.1696,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7971246006389776,
+      "grad_norm": 1.1441930453048488,
+      "learning_rate": 9.187437046803916e-06,
+      "loss": 0.1882,
+      "step": 1996
+    },
+    {
+      "epoch": 0.797923322683706,
+      "grad_norm": 1.1610488689052545,
+      "learning_rate": 9.184895676105033e-06,
+      "loss": 0.1844,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7987220447284346,
+      "grad_norm": 1.2457291872936878,
+      "learning_rate": 9.182350690051134e-06,
+      "loss": 0.187,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7987220447284346,
+      "eval_loss": 0.16647948324680328,
+      "eval_runtime": 417.9228,
+      "eval_samples_per_second": 42.608,
+      "eval_steps_per_second": 5.326,
+      "step": 2000
+    },
+    {
+      "epoch": 0.799520766773163,
+      "grad_norm": 1.1216999386193873,
+      "learning_rate": 9.179802090840852e-06,
+      "loss": 0.1554,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8003194888178914,
+      "grad_norm": 1.2327864947476759,
+      "learning_rate": 9.177249880675934e-06,
+      "loss": 0.19,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8011182108626198,
+      "grad_norm": 1.1765118852291059,
+      "learning_rate": 9.174694061761249e-06,
+      "loss": 0.1752,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8019169329073482,
+      "grad_norm": 1.3929323793396107,
+      "learning_rate": 9.172134636304783e-06,
+      "loss": 0.197,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8027156549520766,
+      "grad_norm": 1.1116097957166795,
+      "learning_rate": 9.169571606517637e-06,
+      "loss": 0.1821,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8035143769968051,
+      "grad_norm": 1.1758254974846156,
+      "learning_rate": 9.16700497461403e-06,
+      "loss": 0.1801,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8043130990415336,
+      "grad_norm": 1.2306371246505314,
+      "learning_rate": 9.164434742811287e-06,
+      "loss": 0.1888,
+      "step": 2014
+    },
+    {
+      "epoch": 0.805111821086262,
+      "grad_norm": 8.822301554752972,
+      "learning_rate": 9.16186091332985e-06,
+      "loss": 0.2123,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8059105431309904,
+      "grad_norm": 3.5201171269663365,
+      "learning_rate": 9.159283488393259e-06,
+      "loss": 0.1897,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8067092651757188,
+      "grad_norm": 1.5889480805348766,
+      "learning_rate": 9.15670247022817e-06,
+      "loss": 0.1948,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8075079872204473,
+      "grad_norm": 1.2452616237787566,
+      "learning_rate": 9.154117861064337e-06,
+      "loss": 0.1736,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8083067092651757,
+      "grad_norm": 1.1779378285841855,
+      "learning_rate": 9.15152966313462e-06,
+      "loss": 0.1949,
+      "step": 2024
+    },
+    {
+      "epoch": 0.8091054313099042,
+      "grad_norm": 1.1749802960279865,
+      "learning_rate": 9.148937878674975e-06,
+      "loss": 0.1931,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8099041533546326,
+      "grad_norm": 1.1577483816469516,
+      "learning_rate": 9.146342509924464e-06,
+      "loss": 0.1761,
+      "step": 2028
+    },
+    {
+      "epoch": 0.810702875399361,
+      "grad_norm": 1.1493255720503592,
+      "learning_rate": 9.143743559125238e-06,
+      "loss": 0.2112,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8115015974440895,
+      "grad_norm": 1.1867113395597348,
+      "learning_rate": 9.141141028522544e-06,
+      "loss": 0.1871,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8123003194888179,
+      "grad_norm": 1.2097756605247039,
+      "learning_rate": 9.138534920364725e-06,
+      "loss": 0.1895,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8130990415335463,
+      "grad_norm": 1.2656565196236773,
+      "learning_rate": 9.135925236903213e-06,
+      "loss": 0.1931,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8138977635782748,
+      "grad_norm": 1.1498073348190923,
+      "learning_rate": 9.133311980392525e-06,
+      "loss": 0.1622,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8146964856230032,
+      "grad_norm": 1.1313101324027541,
+      "learning_rate": 9.130695153090272e-06,
+      "loss": 0.1909,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8154952076677316,
+      "grad_norm": 1.2312685403966457,
+      "learning_rate": 9.128074757257142e-06,
+      "loss": 0.2096,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8162939297124601,
+      "grad_norm": 1.1681617587331703,
+      "learning_rate": 9.125450795156913e-06,
+      "loss": 0.168,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8170926517571885,
+      "grad_norm": 1.1768723977340287,
+      "learning_rate": 9.12282326905644e-06,
+      "loss": 0.1889,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 1.0800467032397891,
+      "learning_rate": 9.120192181225658e-06,
+      "loss": 0.1948,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8186900958466453,
+      "grad_norm": 1.1247061859219583,
+      "learning_rate": 9.117557533937575e-06,
+      "loss": 0.1713,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8194888178913738,
+      "grad_norm": 1.172438969339138,
+      "learning_rate": 9.114919329468283e-06,
+      "loss": 0.162,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8202875399361023,
+      "grad_norm": 1.2278975843705813,
+      "learning_rate": 9.112277570096938e-06,
+      "loss": 0.1865,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8210862619808307,
+      "grad_norm": 1.2992348324926213,
+      "learning_rate": 9.109632258105771e-06,
+      "loss": 0.2055,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8218849840255591,
+      "grad_norm": 1.251504510998654,
+      "learning_rate": 9.106983395780086e-06,
+      "loss": 0.1843,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8226837060702875,
+      "grad_norm": 1.276208869165608,
+      "learning_rate": 9.104330985408245e-06,
+      "loss": 0.1886,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8234824281150159,
+      "grad_norm": 1.356357122339326,
+      "learning_rate": 9.101675029281683e-06,
+      "loss": 0.1981,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8242811501597445,
+      "grad_norm": 1.2399736943434136,
+      "learning_rate": 9.099015529694894e-06,
+      "loss": 0.1897,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8250798722044729,
+      "grad_norm": 1.2876220526804583,
+      "learning_rate": 9.096352488945437e-06,
+      "loss": 0.1796,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8258785942492013,
+      "grad_norm": 1.2532320467708493,
+      "learning_rate": 9.093685909333926e-06,
+      "loss": 0.1788,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8266773162939297,
+      "grad_norm": 1.2586518550619221,
+      "learning_rate": 9.091015793164035e-06,
+      "loss": 0.1768,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8274760383386581,
+      "grad_norm": 1.1394590760675556,
+      "learning_rate": 9.088342142742493e-06,
+      "loss": 0.1741,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8282747603833865,
+      "grad_norm": 1.09902223120881,
+      "learning_rate": 9.08566496037908e-06,
+      "loss": 0.1749,
+      "step": 2074
+    },
+    {
+      "epoch": 0.829073482428115,
+      "grad_norm": 1.094791391856433,
+      "learning_rate": 9.08298424838663e-06,
+      "loss": 0.1785,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8298722044728435,
+      "grad_norm": 1.3016717542317755,
+      "learning_rate": 9.080300009081025e-06,
+      "loss": 0.2058,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 1.2437203627775957,
+      "learning_rate": 9.077612244781196e-06,
+      "loss": 0.18,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8314696485623003,
+      "grad_norm": 1.1494190321851014,
+      "learning_rate": 9.074920957809115e-06,
+      "loss": 0.1993,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8322683706070287,
+      "grad_norm": 1.197846354885409,
+      "learning_rate": 9.0722261504898e-06,
+      "loss": 0.1924,
+      "step": 2084
+    },
+    {
+      "epoch": 0.8330670926517572,
+      "grad_norm": 1.2305171514008204,
+      "learning_rate": 9.069527825151314e-06,
+      "loss": 0.2045,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8338658146964856,
+      "grad_norm": 1.1363998111576086,
+      "learning_rate": 9.066825984124751e-06,
+      "loss": 0.1878,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8346645367412141,
+      "grad_norm": 1.2416863327291112,
+      "learning_rate": 9.064120629744253e-06,
+      "loss": 0.187,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8354632587859425,
+      "grad_norm": 1.1003807898549165,
+      "learning_rate": 9.061411764346983e-06,
+      "loss": 0.1727,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8362619808306709,
+      "grad_norm": 1.1391607571000586,
+      "learning_rate": 9.05869939027315e-06,
+      "loss": 0.1792,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8370607028753994,
+      "grad_norm": 1.191740022915474,
+      "learning_rate": 9.055983509865988e-06,
+      "loss": 0.1852,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8378594249201278,
+      "grad_norm": 1.263318830688807,
+      "learning_rate": 9.053264125471763e-06,
+      "loss": 0.1794,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8386581469648562,
+      "grad_norm": 1.2749000778835136,
+      "learning_rate": 9.050541239439764e-06,
+      "loss": 0.1683,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8394568690095847,
+      "grad_norm": 1.264926687502538,
+      "learning_rate": 9.04781485412231e-06,
+      "loss": 0.1811,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8402555910543131,
+      "grad_norm": 1.218176101735685,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.1826,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8410543130990416,
+      "grad_norm": 1.1339821842224997,
+      "learning_rate": 9.04235159505541e-06,
+      "loss": 0.1748,
+      "step": 2106
+    },
+    {
+      "epoch": 0.84185303514377,
+      "grad_norm": 1.228556794787821,
+      "learning_rate": 9.039614726025708e-06,
+      "loss": 0.1816,
+      "step": 2108
+    },
+    {
+      "epoch": 0.8426517571884984,
+      "grad_norm": 1.1590750565644556,
+      "learning_rate": 9.036874367150024e-06,
+      "loss": 0.1919,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8434504792332268,
+      "grad_norm": 1.1992254533267062,
+      "learning_rate": 9.034130520795774e-06,
+      "loss": 0.1786,
+      "step": 2112
+    },
+    {
+      "epoch": 0.8442492012779552,
+      "grad_norm": 1.2866570327279305,
+      "learning_rate": 9.03138318933338e-06,
+      "loss": 0.1779,
+      "step": 2114
+    },
+    {
+      "epoch": 0.8450479233226837,
+      "grad_norm": 1.1304382720588986,
+      "learning_rate": 9.028632375136277e-06,
+      "loss": 0.1916,
+      "step": 2116
+    },
+    {
+      "epoch": 0.8458466453674122,
+      "grad_norm": 1.2026241441378607,
+      "learning_rate": 9.025878080580908e-06,
+      "loss": 0.1865,
+      "step": 2118
+    },
+    {
+      "epoch": 0.8466453674121406,
+      "grad_norm": 1.2239810455485836,
+      "learning_rate": 9.023120308046726e-06,
+      "loss": 0.2069,
+      "step": 2120
+    },
+    {
+      "epoch": 0.847444089456869,
+      "grad_norm": 1.1593857182783416,
+      "learning_rate": 9.020359059916189e-06,
+      "loss": 0.1802,
+      "step": 2122
+    },
+    {
+      "epoch": 0.8482428115015974,
+      "grad_norm": 1.1851101442311125,
+      "learning_rate": 9.017594338574746e-06,
+      "loss": 0.1718,
+      "step": 2124
+    },
+    {
+      "epoch": 0.8490415335463258,
+      "grad_norm": 1.3424377790206583,
+      "learning_rate": 9.014826146410863e-06,
+      "loss": 0.187,
+      "step": 2126
+    },
+    {
+      "epoch": 0.8498402555910544,
+      "grad_norm": 1.2211268818753362,
+      "learning_rate": 9.012054485815995e-06,
+      "loss": 0.2054,
+      "step": 2128
+    },
+    {
+      "epoch": 0.8506389776357828,
+      "grad_norm": 1.2093558338280859,
+      "learning_rate": 9.009279359184594e-06,
+      "loss": 0.1853,
+      "step": 2130
+    },
+    {
+      "epoch": 0.8514376996805112,
+      "grad_norm": 1.3142251663193079,
+      "learning_rate": 9.006500768914106e-06,
+      "loss": 0.182,
+      "step": 2132
+    },
+    {
+      "epoch": 0.8522364217252396,
+      "grad_norm": 1.1390165214451813,
+      "learning_rate": 9.003718717404977e-06,
+      "loss": 0.1714,
+      "step": 2134
+    },
+    {
+      "epoch": 0.853035143769968,
+      "grad_norm": 1.2833538623348046,
+      "learning_rate": 9.00093320706063e-06,
+      "loss": 0.1829,
+      "step": 2136
+    },
+    {
+      "epoch": 0.8538338658146964,
+      "grad_norm": 1.201198063387828,
+      "learning_rate": 8.998144240287487e-06,
+      "loss": 0.1836,
+      "step": 2138
+    },
+    {
+      "epoch": 0.854632587859425,
+      "grad_norm": 1.26048990415032,
+      "learning_rate": 8.995351819494954e-06,
+      "loss": 0.1823,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8554313099041534,
+      "grad_norm": 1.1311889575435519,
+      "learning_rate": 8.992555947095414e-06,
+      "loss": 0.1692,
+      "step": 2142
+    },
+    {
+      "epoch": 0.8562300319488818,
+      "grad_norm": 1.351907102192122,
+      "learning_rate": 8.989756625504237e-06,
+      "loss": 0.1979,
+      "step": 2144
+    },
+    {
+      "epoch": 0.8570287539936102,
+      "grad_norm": 1.3577115649569347,
+      "learning_rate": 8.98695385713978e-06,
+      "loss": 0.1888,
+      "step": 2146
+    },
+    {
+      "epoch": 0.8578274760383386,
+      "grad_norm": 1.1332848107317308,
+      "learning_rate": 8.984147644423362e-06,
+      "loss": 0.1737,
+      "step": 2148
+    },
+    {
+      "epoch": 0.8586261980830671,
+      "grad_norm": 1.276000255217281,
+      "learning_rate": 8.981337989779291e-06,
+      "loss": 0.1904,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8594249201277955,
+      "grad_norm": 1.2196944395263352,
+      "learning_rate": 8.978524895634842e-06,
+      "loss": 0.1976,
+      "step": 2152
+    },
+    {
+      "epoch": 0.860223642172524,
+      "grad_norm": 1.0522481436377091,
+      "learning_rate": 8.975708364420264e-06,
+      "loss": 0.1529,
+      "step": 2154
+    },
+    {
+      "epoch": 0.8610223642172524,
+      "grad_norm": 1.2862973570919587,
+      "learning_rate": 8.972888398568772e-06,
+      "loss": 0.1941,
+      "step": 2156
+    },
+    {
+      "epoch": 0.8618210862619808,
+      "grad_norm": 1.1472291142227702,
+      "learning_rate": 8.970065000516553e-06,
+      "loss": 0.1884,
+      "step": 2158
+    },
+    {
+      "epoch": 0.8626198083067093,
+      "grad_norm": 1.2425852591025766,
+      "learning_rate": 8.967238172702754e-06,
+      "loss": 0.1758,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8634185303514377,
+      "grad_norm": 1.1523178078837133,
+      "learning_rate": 8.964407917569488e-06,
+      "loss": 0.1773,
+      "step": 2162
+    },
+    {
+      "epoch": 0.8642172523961661,
+      "grad_norm": 1.0544506917154455,
+      "learning_rate": 8.96157423756183e-06,
+      "loss": 0.1782,
+      "step": 2164
+    },
+    {
+      "epoch": 0.8650159744408946,
+      "grad_norm": 1.398137403531195,
+      "learning_rate": 8.958737135127812e-06,
+      "loss": 0.1888,
+      "step": 2166
+    },
+    {
+      "epoch": 0.865814696485623,
+      "grad_norm": 1.1006386353076576,
+      "learning_rate": 8.95589661271842e-06,
+      "loss": 0.1613,
+      "step": 2168
+    },
+    {
+      "epoch": 0.8666134185303515,
+      "grad_norm": 1.1816019658476726,
+      "learning_rate": 8.953052672787602e-06,
+      "loss": 0.1828,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8674121405750799,
+      "grad_norm": 1.160479928900505,
+      "learning_rate": 8.95020531779225e-06,
+      "loss": 0.1816,
+      "step": 2172
+    },
+    {
+      "epoch": 0.8682108626198083,
+      "grad_norm": 1.1134898684750454,
+      "learning_rate": 8.94735455019221e-06,
+      "loss": 0.1806,
+      "step": 2174
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 1.1953510040643958,
+      "learning_rate": 8.94450037245028e-06,
+      "loss": 0.1759,
+      "step": 2176
+    },
+    {
+      "epoch": 0.8698083067092651,
+      "grad_norm": 1.2112582987106082,
+      "learning_rate": 8.941642787032197e-06,
+      "loss": 0.1684,
+      "step": 2178
+    },
+    {
+      "epoch": 0.8706070287539937,
+      "grad_norm": 1.3031673371016932,
+      "learning_rate": 8.938781796406646e-06,
+      "loss": 0.1846,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8714057507987221,
+      "grad_norm": 1.247581011761579,
+      "learning_rate": 8.935917403045251e-06,
+      "loss": 0.1849,
+      "step": 2182
+    },
+    {
+      "epoch": 0.8722044728434505,
+      "grad_norm": 1.2073797176687306,
+      "learning_rate": 8.933049609422582e-06,
+      "loss": 0.17,
+      "step": 2184
+    },
+    {
+      "epoch": 0.8730031948881789,
+      "grad_norm": 1.1534987114988338,
+      "learning_rate": 8.930178418016138e-06,
+      "loss": 0.185,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8738019169329073,
+      "grad_norm": 1.1820635621691253,
+      "learning_rate": 8.92730383130636e-06,
+      "loss": 0.1632,
+      "step": 2188
+    },
+    {
+      "epoch": 0.8746006389776357,
+      "grad_norm": 1.1918298750579541,
+      "learning_rate": 8.924425851776619e-06,
+      "loss": 0.1789,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8753993610223643,
+      "grad_norm": 1.2772477046804016,
+      "learning_rate": 8.921544481913218e-06,
+      "loss": 0.1932,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8761980830670927,
+      "grad_norm": 1.1382431039382104,
+      "learning_rate": 8.918659724205387e-06,
+      "loss": 0.1829,
+      "step": 2194
+    },
+    {
+      "epoch": 0.8769968051118211,
+      "grad_norm": 1.124034930363523,
+      "learning_rate": 8.915771581145286e-06,
+      "loss": 0.157,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8777955271565495,
+      "grad_norm": 1.1938500101023577,
+      "learning_rate": 8.912880055227998e-06,
+      "loss": 0.1873,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8785942492012779,
+      "grad_norm": 1.247886644845545,
+      "learning_rate": 8.909985148951528e-06,
+      "loss": 0.1893,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8793929712460063,
+      "grad_norm": 1.2882483087755097,
+      "learning_rate": 8.907086864816804e-06,
+      "loss": 0.1986,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8801916932907349,
+      "grad_norm": 1.1916502807322977,
+      "learning_rate": 8.904185205327667e-06,
+      "loss": 0.1585,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8809904153354633,
+      "grad_norm": 1.1609680993035214,
+      "learning_rate": 8.901280172990878e-06,
+      "loss": 0.1756,
+      "step": 2206
+    },
+    {
+      "epoch": 0.8817891373801917,
+      "grad_norm": 1.0458575482646773,
+      "learning_rate": 8.898371770316113e-06,
+      "loss": 0.1545,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8825878594249201,
+      "grad_norm": 1.3354091660753635,
+      "learning_rate": 8.895459999815954e-06,
+      "loss": 0.1997,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8833865814696485,
+      "grad_norm": 1.1262191292997419,
+      "learning_rate": 8.892544864005899e-06,
+      "loss": 0.1743,
+      "step": 2212
+    },
+    {
+      "epoch": 0.884185303514377,
+      "grad_norm": 1.3566110237290778,
+      "learning_rate": 8.889626365404348e-06,
+      "loss": 0.1922,
+      "step": 2214
+    },
+    {
+      "epoch": 0.8849840255591054,
+      "grad_norm": 1.376910187006041,
+      "learning_rate": 8.886704506532611e-06,
+      "loss": 0.1913,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8857827476038339,
+      "grad_norm": 1.2397186314872872,
+      "learning_rate": 8.883779289914894e-06,
+      "loss": 0.1732,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8865814696485623,
+      "grad_norm": 1.2887051577849917,
+      "learning_rate": 8.880850718078313e-06,
+      "loss": 0.2123,
+      "step": 2220
+    },
+    {
+      "epoch": 0.8873801916932907,
+      "grad_norm": 1.2786244843973318,
+      "learning_rate": 8.877918793552875e-06,
+      "loss": 0.1828,
+      "step": 2222
+    },
+    {
+      "epoch": 0.8881789137380192,
+      "grad_norm": 1.0958110453050904,
+      "learning_rate": 8.874983518871488e-06,
+      "loss": 0.1746,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8889776357827476,
+      "grad_norm": 1.2263512727265042,
+      "learning_rate": 8.87204489656995e-06,
+      "loss": 0.2076,
+      "step": 2226
+    },
+    {
+      "epoch": 0.889776357827476,
+      "grad_norm": 1.2566470262879932,
+      "learning_rate": 8.869102929186954e-06,
+      "loss": 0.1942,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8905750798722045,
+      "grad_norm": 1.313910249745457,
+      "learning_rate": 8.866157619264086e-06,
+      "loss": 0.2064,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8913738019169329,
+      "grad_norm": 1.2130369145657738,
+      "learning_rate": 8.86320896934581e-06,
+      "loss": 0.176,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8921725239616614,
+      "grad_norm": 1.1816578044938435,
+      "learning_rate": 8.860256981979485e-06,
+      "loss": 0.1932,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8929712460063898,
+      "grad_norm": 1.095754287469498,
+      "learning_rate": 8.857301659715348e-06,
+      "loss": 0.1808,
+      "step": 2236
+    },
+    {
+      "epoch": 0.8937699680511182,
+      "grad_norm": 1.2405828782550468,
+      "learning_rate": 8.854343005106521e-06,
+      "loss": 0.1975,
+      "step": 2238
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 1.139128509869952,
+      "learning_rate": 8.851381020709e-06,
+      "loss": 0.1638,
+      "step": 2240
+    },
+    {
+      "epoch": 0.895367412140575,
+      "grad_norm": 1.1346192386483716,
+      "learning_rate": 8.848415709081659e-06,
+      "loss": 0.1621,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8961661341853036,
+      "grad_norm": 1.2827068653023486,
+      "learning_rate": 8.845447072786251e-06,
+      "loss": 0.19,
+      "step": 2244
+    },
+    {
+      "epoch": 0.896964856230032,
+      "grad_norm": 1.1215489197999844,
+      "learning_rate": 8.842475114387394e-06,
+      "loss": 0.1732,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8977635782747604,
+      "grad_norm": 1.1479343886524802,
+      "learning_rate": 8.839499836452584e-06,
+      "loss": 0.1727,
+      "step": 2248
+    },
+    {
+      "epoch": 0.8985623003194888,
+      "grad_norm": 1.1534484096320634,
+      "learning_rate": 8.836521241552177e-06,
+      "loss": 0.1885,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8993610223642172,
+      "grad_norm": 1.234578241869321,
+      "learning_rate": 8.833539332259398e-06,
+      "loss": 0.1672,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9001597444089456,
+      "grad_norm": 1.2103918137414391,
+      "learning_rate": 8.830554111150337e-06,
+      "loss": 0.1825,
+      "step": 2254
+    },
+    {
+      "epoch": 0.9009584664536742,
+      "grad_norm": 1.0351256185130608,
+      "learning_rate": 8.827565580803944e-06,
+      "loss": 0.1624,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9017571884984026,
+      "grad_norm": 1.1562824946118846,
+      "learning_rate": 8.824573743802023e-06,
+      "loss": 0.1724,
+      "step": 2258
+    },
+    {
+      "epoch": 0.902555910543131,
+      "grad_norm": 1.352880063666687,
+      "learning_rate": 8.821578602729242e-06,
+      "loss": 0.1709,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9033546325878594,
+      "grad_norm": 1.2118411341735884,
+      "learning_rate": 8.81858016017312e-06,
+      "loss": 0.172,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9041533546325878,
+      "grad_norm": 1.2657731749208432,
+      "learning_rate": 8.815578418724031e-06,
+      "loss": 0.1944,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9049520766773163,
+      "grad_norm": 1.2696811826921244,
+      "learning_rate": 8.812573380975191e-06,
+      "loss": 0.1945,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9057507987220448,
+      "grad_norm": 1.260789859526146,
+      "learning_rate": 8.809565049522673e-06,
+      "loss": 0.187,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9065495207667732,
+      "grad_norm": 1.1561924099805676,
+      "learning_rate": 8.806553426965391e-06,
+      "loss": 0.184,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9073482428115016,
+      "grad_norm": 1.2184507876835764,
+      "learning_rate": 8.803538515905102e-06,
+      "loss": 0.181,
+      "step": 2272
+    },
+    {
+      "epoch": 0.90814696485623,
+      "grad_norm": 1.1489816284226748,
+      "learning_rate": 8.800520318946404e-06,
+      "loss": 0.1601,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9089456869009584,
+      "grad_norm": 1.150382635417185,
+      "learning_rate": 8.797498838696737e-06,
+      "loss": 0.173,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9097444089456869,
+      "grad_norm": 1.2396571356927486,
+      "learning_rate": 8.79447407776637e-06,
+      "loss": 0.1775,
+      "step": 2278
+    },
+    {
+      "epoch": 0.9105431309904153,
+      "grad_norm": 1.1258691868543116,
+      "learning_rate": 8.791446038768416e-06,
+      "loss": 0.1796,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9113418530351438,
+      "grad_norm": 1.1837850225751487,
+      "learning_rate": 8.788414724318814e-06,
+      "loss": 0.186,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9121405750798722,
+      "grad_norm": 1.1771823649393418,
+      "learning_rate": 8.785380137036332e-06,
+      "loss": 0.1774,
+      "step": 2284
+    },
+    {
+      "epoch": 0.9129392971246006,
+      "grad_norm": 1.195664396595768,
+      "learning_rate": 8.782342279542569e-06,
+      "loss": 0.1715,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9137380191693291,
+      "grad_norm": 1.394493043461634,
+      "learning_rate": 8.779301154461945e-06,
+      "loss": 0.1954,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9145367412140575,
+      "grad_norm": 1.484773906372258,
+      "learning_rate": 8.776256764421706e-06,
+      "loss": 0.1649,
+      "step": 2290
+    },
+    {
+      "epoch": 0.9153354632587859,
+      "grad_norm": 1.1713356392013368,
+      "learning_rate": 8.773209112051919e-06,
+      "loss": 0.1676,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9161341853035144,
+      "grad_norm": 1.2492572619724682,
+      "learning_rate": 8.770158199985466e-06,
+      "loss": 0.203,
+      "step": 2294
+    },
+    {
+      "epoch": 0.9169329073482428,
+      "grad_norm": 1.112250308815286,
+      "learning_rate": 8.76710403085805e-06,
+      "loss": 0.1675,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9177316293929713,
+      "grad_norm": 1.2119980335045586,
+      "learning_rate": 8.764046607308183e-06,
+      "loss": 0.1822,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9185303514376997,
+      "grad_norm": 1.2003429295335617,
+      "learning_rate": 8.760985931977191e-06,
+      "loss": 0.1648,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9193290734824281,
+      "grad_norm": 1.2923004205685706,
+      "learning_rate": 8.757922007509208e-06,
+      "loss": 0.1672,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 1.2602401218318184,
+      "learning_rate": 8.754854836551174e-06,
+      "loss": 0.1697,
+      "step": 2304
+    },
+    {
+      "epoch": 0.920926517571885,
+      "grad_norm": 1.3225061494713877,
+      "learning_rate": 8.75178442175284e-06,
+      "loss": 0.2173,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9217252396166135,
+      "grad_norm": 1.236225325668187,
+      "learning_rate": 8.748710765766752e-06,
+      "loss": 0.1748,
+      "step": 2308
+    },
+    {
+      "epoch": 0.9225239616613419,
+      "grad_norm": 1.5668493173725866,
+      "learning_rate": 8.745633871248257e-06,
+      "loss": 0.1792,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9233226837060703,
+      "grad_norm": 1.2818950876816324,
+      "learning_rate": 8.742553740855507e-06,
+      "loss": 0.1701,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9241214057507987,
+      "grad_norm": 1.0563021739777534,
+      "learning_rate": 8.739470377249436e-06,
+      "loss": 0.1673,
+      "step": 2314
+    },
+    {
+      "epoch": 0.9249201277955271,
+      "grad_norm": 1.1015536336953309,
+      "learning_rate": 8.736383783093788e-06,
+      "loss": 0.1768,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9257188498402555,
+      "grad_norm": 1.2209081010826341,
+      "learning_rate": 8.733293961055082e-06,
+      "loss": 0.1893,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9265175718849841,
+      "grad_norm": 1.1147003675746905,
+      "learning_rate": 8.730200913802638e-06,
+      "loss": 0.1727,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9273162939297125,
+      "grad_norm": 1.1772795182600777,
+      "learning_rate": 8.727104644008553e-06,
+      "loss": 0.1777,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9281150159744409,
+      "grad_norm": 1.106532804822251,
+      "learning_rate": 8.724005154347714e-06,
+      "loss": 0.1781,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9289137380191693,
+      "grad_norm": 1.15574880894598,
+      "learning_rate": 8.720902447497788e-06,
+      "loss": 0.1777,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9297124600638977,
+      "grad_norm": 1.2312851048264675,
+      "learning_rate": 8.717796526139218e-06,
+      "loss": 0.204,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9305111821086262,
+      "grad_norm": 1.261983053853919,
+      "learning_rate": 8.71468739295523e-06,
+      "loss": 0.1993,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9313099041533547,
+      "grad_norm": 1.1982722648050632,
+      "learning_rate": 8.711575050631823e-06,
+      "loss": 0.1968,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9321086261980831,
+      "grad_norm": 1.1624739320562267,
+      "learning_rate": 8.708459501857762e-06,
+      "loss": 0.1864,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9329073482428115,
+      "grad_norm": 1.027366394484867,
+      "learning_rate": 8.70534074932459e-06,
+      "loss": 0.1687,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9337060702875399,
+      "grad_norm": 1.1021491198370645,
+      "learning_rate": 8.702218795726619e-06,
+      "loss": 0.1684,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9345047923322684,
+      "grad_norm": 1.1040575428181885,
+      "learning_rate": 8.699093643760914e-06,
+      "loss": 0.1739,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9353035143769968,
+      "grad_norm": 1.195647306811607,
+      "learning_rate": 8.695965296127318e-06,
+      "loss": 0.1816,
+      "step": 2342
+    },
+    {
+      "epoch": 0.9361022364217252,
+      "grad_norm": 1.0812569337947433,
+      "learning_rate": 8.692833755528426e-06,
+      "loss": 0.1766,
+      "step": 2344
+    },
+    {
+      "epoch": 0.9369009584664537,
+      "grad_norm": 1.3202981154225082,
+      "learning_rate": 8.689699024669594e-06,
+      "loss": 0.1879,
+      "step": 2346
+    },
+    {
+      "epoch": 0.9376996805111821,
+      "grad_norm": 1.1970245625047584,
+      "learning_rate": 8.686561106258932e-06,
+      "loss": 0.1803,
+      "step": 2348
+    },
+    {
+      "epoch": 0.9384984025559105,
+      "grad_norm": 1.2213363903695773,
+      "learning_rate": 8.683420003007308e-06,
+      "loss": 0.1947,
+      "step": 2350
+    },
+    {
+      "epoch": 0.939297124600639,
+      "grad_norm": 1.168137561259312,
+      "learning_rate": 8.680275717628336e-06,
+      "loss": 0.1683,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9400958466453674,
+      "grad_norm": 1.1140547555618086,
+      "learning_rate": 8.677128252838386e-06,
+      "loss": 0.1808,
+      "step": 2354
+    },
+    {
+      "epoch": 0.9408945686900958,
+      "grad_norm": 1.0913344127324964,
+      "learning_rate": 8.673977611356567e-06,
+      "loss": 0.1768,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9416932907348243,
+      "grad_norm": 1.2470605938854136,
+      "learning_rate": 8.670823795904737e-06,
+      "loss": 0.1881,
+      "step": 2358
+    },
+    {
+      "epoch": 0.9424920127795527,
+      "grad_norm": 1.1432903930033393,
+      "learning_rate": 8.667666809207495e-06,
+      "loss": 0.1814,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9432907348242812,
+      "grad_norm": 1.1254326884247754,
+      "learning_rate": 8.664506653992181e-06,
+      "loss": 0.1721,
+      "step": 2362
+    },
+    {
+      "epoch": 0.9440894568690096,
+      "grad_norm": 1.1774199109472463,
+      "learning_rate": 8.661343332988869e-06,
+      "loss": 0.1644,
+      "step": 2364
+    },
+    {
+      "epoch": 0.944888178913738,
+      "grad_norm": 1.1966632218098605,
+      "learning_rate": 8.65817684893037e-06,
+      "loss": 0.1735,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9456869009584664,
+      "grad_norm": 1.0407093351327132,
+      "learning_rate": 8.655007204552228e-06,
+      "loss": 0.1544,
+      "step": 2368
+    },
+    {
+      "epoch": 0.9464856230031949,
+      "grad_norm": 1.2274155096905306,
+      "learning_rate": 8.651834402592719e-06,
+      "loss": 0.1798,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9472843450479234,
+      "grad_norm": 1.267776362466128,
+      "learning_rate": 8.64865844579284e-06,
+      "loss": 0.1751,
+      "step": 2372
+    },
+    {
+      "epoch": 0.9480830670926518,
+      "grad_norm": 1.1768732812780864,
+      "learning_rate": 8.64547933689632e-06,
+      "loss": 0.1785,
+      "step": 2374
+    },
+    {
+      "epoch": 0.9488817891373802,
+      "grad_norm": 1.1577676779075388,
+      "learning_rate": 8.64229707864961e-06,
+      "loss": 0.1947,
+      "step": 2376
+    },
+    {
+      "epoch": 0.9496805111821086,
+      "grad_norm": 1.1591103988078189,
+      "learning_rate": 8.63911167380188e-06,
+      "loss": 0.1788,
+      "step": 2378
+    },
+    {
+      "epoch": 0.950479233226837,
+      "grad_norm": 1.11481900810481,
+      "learning_rate": 8.635923125105019e-06,
+      "loss": 0.1676,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9512779552715654,
+      "grad_norm": 1.2948111421923072,
+      "learning_rate": 8.632731435313634e-06,
+      "loss": 0.1645,
+      "step": 2382
+    },
+    {
+      "epoch": 0.952076677316294,
+      "grad_norm": 1.1733292169857867,
+      "learning_rate": 8.629536607185042e-06,
+      "loss": 0.1701,
+      "step": 2384
+    },
+    {
+      "epoch": 0.9528753993610224,
+      "grad_norm": 1.1636636504635067,
+      "learning_rate": 8.626338643479275e-06,
+      "loss": 0.17,
+      "step": 2386
+    },
+    {
+      "epoch": 0.9536741214057508,
+      "grad_norm": 1.2570201463967705,
+      "learning_rate": 8.62313754695907e-06,
+      "loss": 0.1738,
+      "step": 2388
+    },
+    {
+      "epoch": 0.9544728434504792,
+      "grad_norm": 1.0902131985821306,
+      "learning_rate": 8.619933320389872e-06,
+      "loss": 0.1723,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9552715654952076,
+      "grad_norm": 1.2111908413145964,
+      "learning_rate": 8.616725966539831e-06,
+      "loss": 0.1926,
+      "step": 2392
+    },
+    {
+      "epoch": 0.9560702875399361,
+      "grad_norm": 1.1942013221000287,
+      "learning_rate": 8.6135154881798e-06,
+      "loss": 0.1888,
+      "step": 2394
+    },
+    {
+      "epoch": 0.9568690095846646,
+      "grad_norm": 1.1699831896669461,
+      "learning_rate": 8.610301888083327e-06,
+      "loss": 0.2023,
+      "step": 2396
+    },
+    {
+      "epoch": 0.957667731629393,
+      "grad_norm": 1.1882929391966546,
+      "learning_rate": 8.607085169026661e-06,
+      "loss": 0.1852,
+      "step": 2398
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 1.1905600662435856,
+      "learning_rate": 8.60386533378874e-06,
+      "loss": 0.1919,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9592651757188498,
+      "grad_norm": 1.2472441081594776,
+      "learning_rate": 8.600642385151206e-06,
+      "loss": 0.1923,
+      "step": 2402
+    },
+    {
+      "epoch": 0.9600638977635783,
+      "grad_norm": 1.0759279066292031,
+      "learning_rate": 8.597416325898373e-06,
+      "loss": 0.1685,
+      "step": 2404
+    },
+    {
+      "epoch": 0.9608626198083067,
+      "grad_norm": 1.1583330331881445,
+      "learning_rate": 8.594187158817257e-06,
+      "loss": 0.1749,
+      "step": 2406
+    },
+    {
+      "epoch": 0.9616613418530351,
+      "grad_norm": 1.2089521342785554,
+      "learning_rate": 8.590954886697554e-06,
+      "loss": 0.1939,
+      "step": 2408
+    },
+    {
+      "epoch": 0.9624600638977636,
+      "grad_norm": 1.2614329419101098,
+      "learning_rate": 8.58771951233164e-06,
+      "loss": 0.2026,
+      "step": 2410
+    },
+    {
+      "epoch": 0.963258785942492,
+      "grad_norm": 1.1366241513550654,
+      "learning_rate": 8.584481038514573e-06,
+      "loss": 0.1639,
+      "step": 2412
+    },
+    {
+      "epoch": 0.9640575079872205,
+      "grad_norm": 1.3624993080110601,
+      "learning_rate": 8.581239468044093e-06,
+      "loss": 0.185,
+      "step": 2414
+    },
+    {
+      "epoch": 0.9648562300319489,
+      "grad_norm": 1.0567057031046825,
+      "learning_rate": 8.577994803720605e-06,
+      "loss": 0.1661,
+      "step": 2416
+    },
+    {
+      "epoch": 0.9656549520766773,
+      "grad_norm": 1.1539180273643035,
+      "learning_rate": 8.574747048347199e-06,
+      "loss": 0.1664,
+      "step": 2418
+    },
+    {
+      "epoch": 0.9664536741214057,
+      "grad_norm": 1.236448054190876,
+      "learning_rate": 8.571496204729623e-06,
+      "loss": 0.1818,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9672523961661342,
+      "grad_norm": 1.0633439918324645,
+      "learning_rate": 8.568242275676304e-06,
+      "loss": 0.1735,
+      "step": 2422
+    },
+    {
+      "epoch": 0.9680511182108626,
+      "grad_norm": 1.1561922874116346,
+      "learning_rate": 8.564985263998327e-06,
+      "loss": 0.1651,
+      "step": 2424
+    },
+    {
+      "epoch": 0.9688498402555911,
+      "grad_norm": 1.1398950472118397,
+      "learning_rate": 8.561725172509444e-06,
+      "loss": 0.163,
+      "step": 2426
+    },
+    {
+      "epoch": 0.9696485623003195,
+      "grad_norm": 1.1653225900272541,
+      "learning_rate": 8.558462004026065e-06,
+      "loss": 0.173,
+      "step": 2428
+    },
+    {
+      "epoch": 0.9704472843450479,
+      "grad_norm": 1.3275231148352606,
+      "learning_rate": 8.555195761367263e-06,
+      "loss": 0.1819,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 1.0555861455966542,
+      "learning_rate": 8.551926447354759e-06,
+      "loss": 0.1699,
+      "step": 2432
+    },
+    {
+      "epoch": 0.9720447284345048,
+      "grad_norm": 1.2506959109380367,
+      "learning_rate": 8.548654064812934e-06,
+      "loss": 0.1792,
+      "step": 2434
+    },
+    {
+      "epoch": 0.9728434504792333,
+      "grad_norm": 1.1541294441379601,
+      "learning_rate": 8.54537861656882e-06,
+      "loss": 0.1796,
+      "step": 2436
+    },
+    {
+      "epoch": 0.9736421725239617,
+      "grad_norm": 1.3106265759711455,
+      "learning_rate": 8.542100105452093e-06,
+      "loss": 0.1823,
+      "step": 2438
+    },
+    {
+      "epoch": 0.9744408945686901,
+      "grad_norm": 1.2700540275591943,
+      "learning_rate": 8.538818534295076e-06,
+      "loss": 0.183,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9752396166134185,
+      "grad_norm": 1.4250215144969003,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.1865,
+      "step": 2442
+    },
+    {
+      "epoch": 0.9760383386581469,
+      "grad_norm": 1.1840076787843286,
+      "learning_rate": 8.532246223202689e-06,
+      "loss": 0.1773,
+      "step": 2444
+    },
+    {
+      "epoch": 0.9768370607028753,
+      "grad_norm": 1.0800090683302819,
+      "learning_rate": 8.528955488945177e-06,
+      "loss": 0.1762,
+      "step": 2446
+    },
+    {
+      "epoch": 0.9776357827476039,
+      "grad_norm": 1.056788808510418,
+      "learning_rate": 8.525661706003083e-06,
+      "loss": 0.1752,
+      "step": 2448
+    },
+    {
+      "epoch": 0.9784345047923323,
+      "grad_norm": 1.1120115317215207,
+      "learning_rate": 8.522364877221926e-06,
+      "loss": 0.1827,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9792332268370607,
+      "grad_norm": 1.2484051846053978,
+      "learning_rate": 8.519065005449858e-06,
+      "loss": 0.1744,
+      "step": 2452
+    },
+    {
+      "epoch": 0.9800319488817891,
+      "grad_norm": 1.2862407811422434,
+      "learning_rate": 8.515762093537654e-06,
+      "loss": 0.1816,
+      "step": 2454
+    },
+    {
+      "epoch": 0.9808306709265175,
+      "grad_norm": 1.1176065762239906,
+      "learning_rate": 8.512456144338717e-06,
+      "loss": 0.1779,
+      "step": 2456
+    },
+    {
+      "epoch": 0.981629392971246,
+      "grad_norm": 1.1181750039143228,
+      "learning_rate": 8.509147160709079e-06,
+      "loss": 0.1791,
+      "step": 2458
+    },
+    {
+      "epoch": 0.9824281150159745,
+      "grad_norm": 1.1233000113043177,
+      "learning_rate": 8.505835145507387e-06,
+      "loss": 0.1801,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9832268370607029,
+      "grad_norm": 1.149372977172484,
+      "learning_rate": 8.502520101594909e-06,
+      "loss": 0.1621,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9840255591054313,
+      "grad_norm": 1.1858307935499464,
+      "learning_rate": 8.499202031835532e-06,
+      "loss": 0.1832,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9848242811501597,
+      "grad_norm": 1.290001462169469,
+      "learning_rate": 8.495880939095754e-06,
+      "loss": 0.1894,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9856230031948882,
+      "grad_norm": 1.3581022131665044,
+      "learning_rate": 8.492556826244687e-06,
+      "loss": 0.2,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9864217252396166,
+      "grad_norm": 1.1056676733741198,
+      "learning_rate": 8.489229696154049e-06,
+      "loss": 0.1805,
+      "step": 2470
+    },
+    {
+      "epoch": 0.987220447284345,
+      "grad_norm": 1.1250903817623754,
+      "learning_rate": 8.485899551698166e-06,
+      "loss": 0.177,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9880191693290735,
+      "grad_norm": 1.148392440744409,
+      "learning_rate": 8.482566395753975e-06,
+      "loss": 0.1867,
+      "step": 2474
+    },
+    {
+      "epoch": 0.9888178913738019,
+      "grad_norm": 1.2558757559301705,
+      "learning_rate": 8.479230231201001e-06,
+      "loss": 0.1964,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9896166134185304,
+      "grad_norm": 1.2579913217677217,
+      "learning_rate": 8.475891060921378e-06,
+      "loss": 0.1829,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9904153354632588,
+      "grad_norm": 1.316707228945154,
+      "learning_rate": 8.472548887799833e-06,
+      "loss": 0.1696,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9912140575079872,
+      "grad_norm": 1.2049638106018488,
+      "learning_rate": 8.46920371472369e-06,
+      "loss": 0.1887,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9920127795527156,
+      "grad_norm": 1.2377015394107287,
+      "learning_rate": 8.465855544582862e-06,
+      "loss": 0.2014,
+      "step": 2484
+    },
+    {
+      "epoch": 0.9928115015974441,
+      "grad_norm": 1.1268024008456778,
+      "learning_rate": 8.462504380269853e-06,
+      "loss": 0.179,
+      "step": 2486
+    },
+    {
+      "epoch": 0.9936102236421726,
+      "grad_norm": 1.446057306444501,
+      "learning_rate": 8.45915022467975e-06,
+      "loss": 0.1918,
+      "step": 2488
+    },
+    {
+      "epoch": 0.994408945686901,
+      "grad_norm": 1.186023464310668,
+      "learning_rate": 8.455793080710231e-06,
+      "loss": 0.1706,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9952076677316294,
+      "grad_norm": 1.2273194439241009,
+      "learning_rate": 8.452432951261549e-06,
+      "loss": 0.1857,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9960063897763578,
+      "grad_norm": 1.0809657844755842,
+      "learning_rate": 8.44906983923654e-06,
+      "loss": 0.1403,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9968051118210862,
+      "grad_norm": 1.1417636929002628,
+      "learning_rate": 8.445703747540614e-06,
+      "loss": 0.1727,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9976038338658147,
+      "grad_norm": 1.2435186121479649,
+      "learning_rate": 8.442334679081757e-06,
+      "loss": 0.1651,
+      "step": 2498
+    },
+    {
+      "epoch": 0.9984025559105432,
+      "grad_norm": 1.261179386652203,
+      "learning_rate": 8.438962636770528e-06,
+      "loss": 0.1916,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9984025559105432,
+      "eval_loss": 0.1609409898519516,
+      "eval_runtime": 418.1212,
+      "eval_samples_per_second": 42.588,
+      "eval_steps_per_second": 5.324,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9992012779552716,
+      "grad_norm": 1.2177062771646232,
+      "learning_rate": 8.43558762352005e-06,
+      "loss": 0.1956,
+      "step": 2502
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.129270940930756,
+      "learning_rate": 8.43220964224602e-06,
+      "loss": 0.1687,
+      "step": 2504
+    },
+    {
+      "epoch": 1.0007987220447285,
+      "grad_norm": 0.827421022738481,
+      "learning_rate": 8.428828695866694e-06,
+      "loss": 0.1127,
+      "step": 2506
+    },
+    {
+      "epoch": 1.0015974440894568,
+      "grad_norm": 0.8532426168374652,
+      "learning_rate": 8.425444787302887e-06,
+      "loss": 0.1097,
+      "step": 2508
+    },
+    {
+      "epoch": 1.0023961661341854,
+      "grad_norm": 1.0387260637251956,
+      "learning_rate": 8.422057919477984e-06,
+      "loss": 0.1209,
+      "step": 2510
+    },
+    {
+      "epoch": 1.0031948881789137,
+      "grad_norm": 0.9532301555359793,
+      "learning_rate": 8.418668095317912e-06,
+      "loss": 0.1124,
+      "step": 2512
+    },
+    {
+      "epoch": 1.0039936102236422,
+      "grad_norm": 0.9749882933863255,
+      "learning_rate": 8.415275317751163e-06,
+      "loss": 0.1033,
+      "step": 2514
+    },
+    {
+      "epoch": 1.0047923322683705,
+      "grad_norm": 1.215365298510418,
+      "learning_rate": 8.411879589708775e-06,
+      "loss": 0.1136,
+      "step": 2516
+    },
+    {
+      "epoch": 1.005591054313099,
+      "grad_norm": 1.0938388152428302,
+      "learning_rate": 8.408480914124338e-06,
+      "loss": 0.1126,
+      "step": 2518
+    },
+    {
+      "epoch": 1.0063897763578276,
+      "grad_norm": 1.1091853030022452,
+      "learning_rate": 8.405079293933986e-06,
+      "loss": 0.0985,
+      "step": 2520
+    },
+    {
+      "epoch": 1.0071884984025559,
+      "grad_norm": 1.1309767039321081,
+      "learning_rate": 8.401674732076399e-06,
+      "loss": 0.1121,
+      "step": 2522
+    },
+    {
+      "epoch": 1.0079872204472844,
+      "grad_norm": 1.2204646216277761,
+      "learning_rate": 8.398267231492797e-06,
+      "loss": 0.1276,
+      "step": 2524
+    },
+    {
+      "epoch": 1.0087859424920127,
+      "grad_norm": 1.380795556523787,
+      "learning_rate": 8.394856795126937e-06,
+      "loss": 0.1039,
+      "step": 2526
+    },
+    {
+      "epoch": 1.0095846645367412,
+      "grad_norm": 1.0922220217499492,
+      "learning_rate": 8.391443425925118e-06,
+      "loss": 0.1126,
+      "step": 2528
+    },
+    {
+      "epoch": 1.0103833865814698,
+      "grad_norm": 0.990994836569366,
+      "learning_rate": 8.388027126836168e-06,
+      "loss": 0.1109,
+      "step": 2530
+    },
+    {
+      "epoch": 1.011182108626198,
+      "grad_norm": 1.160502599618174,
+      "learning_rate": 8.384607900811442e-06,
+      "loss": 0.11,
+      "step": 2532
+    },
+    {
+      "epoch": 1.0119808306709266,
+      "grad_norm": 1.0016564935445926,
+      "learning_rate": 8.381185750804835e-06,
+      "loss": 0.1007,
+      "step": 2534
+    },
+    {
+      "epoch": 1.012779552715655,
+      "grad_norm": 1.0190577969031376,
+      "learning_rate": 8.37776067977276e-06,
+      "loss": 0.1095,
+      "step": 2536
+    },
+    {
+      "epoch": 1.0135782747603834,
+      "grad_norm": 1.0372954792800193,
+      "learning_rate": 8.374332690674151e-06,
+      "loss": 0.1087,
+      "step": 2538
+    },
+    {
+      "epoch": 1.0143769968051117,
+      "grad_norm": 1.0326525572009215,
+      "learning_rate": 8.370901786470475e-06,
+      "loss": 0.1023,
+      "step": 2540
+    },
+    {
+      "epoch": 1.0151757188498403,
+      "grad_norm": 0.9829529420381988,
+      "learning_rate": 8.367467970125702e-06,
+      "loss": 0.1104,
+      "step": 2542
+    },
+    {
+      "epoch": 1.0159744408945688,
+      "grad_norm": 1.1974618667877126,
+      "learning_rate": 8.36403124460633e-06,
+      "loss": 0.1052,
+      "step": 2544
+    },
+    {
+      "epoch": 1.016773162939297,
+      "grad_norm": 1.13559909487521,
+      "learning_rate": 8.360591612881363e-06,
+      "loss": 0.1076,
+      "step": 2546
+    },
+    {
+      "epoch": 1.0175718849840256,
+      "grad_norm": 1.1337615805939383,
+      "learning_rate": 8.357149077922322e-06,
+      "loss": 0.1179,
+      "step": 2548
+    },
+    {
+      "epoch": 1.018370607028754,
+      "grad_norm": 1.1381630091831911,
+      "learning_rate": 8.353703642703228e-06,
+      "loss": 0.1082,
+      "step": 2550
+    },
+    {
+      "epoch": 1.0191693290734825,
+      "grad_norm": 1.1081002212560531,
+      "learning_rate": 8.350255310200611e-06,
+      "loss": 0.1133,
+      "step": 2552
+    },
+    {
+      "epoch": 1.0199680511182108,
+      "grad_norm": 1.012013673101353,
+      "learning_rate": 8.346804083393511e-06,
+      "loss": 0.1133,
+      "step": 2554
+    },
+    {
+      "epoch": 1.0207667731629393,
+      "grad_norm": 1.0949479826249693,
+      "learning_rate": 8.343349965263459e-06,
+      "loss": 0.109,
+      "step": 2556
+    },
+    {
+      "epoch": 1.0215654952076678,
+      "grad_norm": 1.4697029214526798,
+      "learning_rate": 8.339892958794487e-06,
+      "loss": 0.1223,
+      "step": 2558
+    },
+    {
+      "epoch": 1.0223642172523961,
+      "grad_norm": 1.1256837730078384,
+      "learning_rate": 8.336433066973122e-06,
+      "loss": 0.1106,
+      "step": 2560
+    },
+    {
+      "epoch": 1.0231629392971247,
+      "grad_norm": 1.0111334067086826,
+      "learning_rate": 8.332970292788384e-06,
+      "loss": 0.0992,
+      "step": 2562
+    },
+    {
+      "epoch": 1.023961661341853,
+      "grad_norm": 1.1457290491628267,
+      "learning_rate": 8.329504639231784e-06,
+      "loss": 0.1055,
+      "step": 2564
+    },
+    {
+      "epoch": 1.0247603833865815,
+      "grad_norm": 1.094415886169632,
+      "learning_rate": 8.32603610929732e-06,
+      "loss": 0.1233,
+      "step": 2566
+    },
+    {
+      "epoch": 1.0255591054313098,
+      "grad_norm": 1.115090103670034,
+      "learning_rate": 8.322564705981476e-06,
+      "loss": 0.1027,
+      "step": 2568
+    },
+    {
+      "epoch": 1.0263578274760383,
+      "grad_norm": 1.2411016808966997,
+      "learning_rate": 8.319090432283213e-06,
+      "loss": 0.1251,
+      "step": 2570
+    },
+    {
+      "epoch": 1.0271565495207668,
+      "grad_norm": 1.161818122354369,
+      "learning_rate": 8.315613291203977e-06,
+      "loss": 0.1136,
+      "step": 2572
+    },
+    {
+      "epoch": 1.0279552715654952,
+      "grad_norm": 1.010109815757735,
+      "learning_rate": 8.312133285747693e-06,
+      "loss": 0.1116,
+      "step": 2574
+    },
+    {
+      "epoch": 1.0287539936102237,
+      "grad_norm": 1.0352536009528577,
+      "learning_rate": 8.308650418920751e-06,
+      "loss": 0.0985,
+      "step": 2576
+    },
+    {
+      "epoch": 1.029552715654952,
+      "grad_norm": 1.0385517544202987,
+      "learning_rate": 8.305164693732026e-06,
+      "loss": 0.1085,
+      "step": 2578
+    },
+    {
+      "epoch": 1.0303514376996805,
+      "grad_norm": 1.0075596705684045,
+      "learning_rate": 8.301676113192853e-06,
+      "loss": 0.1092,
+      "step": 2580
+    },
+    {
+      "epoch": 1.031150159744409,
+      "grad_norm": 1.0700338942744492,
+      "learning_rate": 8.298184680317035e-06,
+      "loss": 0.1124,
+      "step": 2582
+    },
+    {
+      "epoch": 1.0319488817891374,
+      "grad_norm": 1.0149219991197933,
+      "learning_rate": 8.294690398120843e-06,
+      "loss": 0.1121,
+      "step": 2584
+    },
+    {
+      "epoch": 1.0327476038338659,
+      "grad_norm": 1.2039242543481947,
+      "learning_rate": 8.291193269623005e-06,
+      "loss": 0.1222,
+      "step": 2586
+    },
+    {
+      "epoch": 1.0335463258785942,
+      "grad_norm": 1.2377862042067174,
+      "learning_rate": 8.287693297844712e-06,
+      "loss": 0.1145,
+      "step": 2588
+    },
+    {
+      "epoch": 1.0343450479233227,
+      "grad_norm": 1.091238797708431,
+      "learning_rate": 8.28419048580961e-06,
+      "loss": 0.1155,
+      "step": 2590
+    },
+    {
+      "epoch": 1.035143769968051,
+      "grad_norm": 1.078587618975172,
+      "learning_rate": 8.280684836543794e-06,
+      "loss": 0.1082,
+      "step": 2592
+    },
+    {
+      "epoch": 1.0359424920127795,
+      "grad_norm": 1.0454759141708203,
+      "learning_rate": 8.277176353075818e-06,
+      "loss": 0.1056,
+      "step": 2594
+    },
+    {
+      "epoch": 1.036741214057508,
+      "grad_norm": 1.168733586560627,
+      "learning_rate": 8.27366503843668e-06,
+      "loss": 0.1159,
+      "step": 2596
+    },
+    {
+      "epoch": 1.0375399361022364,
+      "grad_norm": 0.9639822523537996,
+      "learning_rate": 8.270150895659824e-06,
+      "loss": 0.1141,
+      "step": 2598
+    },
+    {
+      "epoch": 1.038338658146965,
+      "grad_norm": 1.080334562027545,
+      "learning_rate": 8.266633927781135e-06,
+      "loss": 0.097,
+      "step": 2600
+    },
+    {
+      "epoch": 1.0391373801916932,
+      "grad_norm": 1.0212322282727473,
+      "learning_rate": 8.263114137838946e-06,
+      "loss": 0.1141,
+      "step": 2602
+    },
+    {
+      "epoch": 1.0399361022364217,
+      "grad_norm": 1.0394948988809494,
+      "learning_rate": 8.259591528874023e-06,
+      "loss": 0.1101,
+      "step": 2604
+    },
+    {
+      "epoch": 1.04073482428115,
+      "grad_norm": 1.3623022308744146,
+      "learning_rate": 8.256066103929566e-06,
+      "loss": 0.1273,
+      "step": 2606
+    },
+    {
+      "epoch": 1.0415335463258786,
+      "grad_norm": 1.022587558449341,
+      "learning_rate": 8.25253786605121e-06,
+      "loss": 0.1124,
+      "step": 2608
+    },
+    {
+      "epoch": 1.042332268370607,
+      "grad_norm": 0.986241541472163,
+      "learning_rate": 8.249006818287017e-06,
+      "loss": 0.1017,
+      "step": 2610
+    },
+    {
+      "epoch": 1.0431309904153354,
+      "grad_norm": 1.1218445133529082,
+      "learning_rate": 8.245472963687484e-06,
+      "loss": 0.1051,
+      "step": 2612
+    },
+    {
+      "epoch": 1.043929712460064,
+      "grad_norm": 0.9853007519672737,
+      "learning_rate": 8.241936305305526e-06,
+      "loss": 0.0933,
+      "step": 2614
+    },
+    {
+      "epoch": 1.0447284345047922,
+      "grad_norm": 0.9238979931931091,
+      "learning_rate": 8.238396846196483e-06,
+      "loss": 0.0964,
+      "step": 2616
+    },
+    {
+      "epoch": 1.0455271565495208,
+      "grad_norm": 1.145209666797251,
+      "learning_rate": 8.23485458941811e-06,
+      "loss": 0.1178,
+      "step": 2618
+    },
+    {
+      "epoch": 1.0463258785942493,
+      "grad_norm": 1.128182977350629,
+      "learning_rate": 8.231309538030586e-06,
+      "loss": 0.1141,
+      "step": 2620
+    },
+    {
+      "epoch": 1.0471246006389776,
+      "grad_norm": 0.9914083611784468,
+      "learning_rate": 8.2277616950965e-06,
+      "loss": 0.0989,
+      "step": 2622
+    },
+    {
+      "epoch": 1.0479233226837061,
+      "grad_norm": 1.1620950439007605,
+      "learning_rate": 8.224211063680854e-06,
+      "loss": 0.1215,
+      "step": 2624
+    },
+    {
+      "epoch": 1.0487220447284344,
+      "grad_norm": 1.1508342917411967,
+      "learning_rate": 8.220657646851059e-06,
+      "loss": 0.1205,
+      "step": 2626
+    },
+    {
+      "epoch": 1.049520766773163,
+      "grad_norm": 1.2150567330620372,
+      "learning_rate": 8.217101447676929e-06,
+      "loss": 0.106,
+      "step": 2628
+    },
+    {
+      "epoch": 1.0503194888178913,
+      "grad_norm": 1.1152102769811438,
+      "learning_rate": 8.213542469230688e-06,
+      "loss": 0.1223,
+      "step": 2630
+    },
+    {
+      "epoch": 1.0511182108626198,
+      "grad_norm": 1.1807098190243748,
+      "learning_rate": 8.209980714586955e-06,
+      "loss": 0.1122,
+      "step": 2632
+    },
+    {
+      "epoch": 1.0519169329073483,
+      "grad_norm": 1.1286829181729654,
+      "learning_rate": 8.206416186822753e-06,
+      "loss": 0.1101,
+      "step": 2634
+    },
+    {
+      "epoch": 1.0527156549520766,
+      "grad_norm": 1.0697286341076933,
+      "learning_rate": 8.202848889017494e-06,
+      "loss": 0.1112,
+      "step": 2636
+    },
+    {
+      "epoch": 1.0535143769968052,
+      "grad_norm": 1.0611961961878171,
+      "learning_rate": 8.19927882425299e-06,
+      "loss": 0.1079,
+      "step": 2638
+    },
+    {
+      "epoch": 1.0543130990415335,
+      "grad_norm": 1.0124139526923805,
+      "learning_rate": 8.195705995613436e-06,
+      "loss": 0.0996,
+      "step": 2640
+    },
+    {
+      "epoch": 1.055111821086262,
+      "grad_norm": 1.108195697612016,
+      "learning_rate": 8.192130406185425e-06,
+      "loss": 0.1107,
+      "step": 2642
+    },
+    {
+      "epoch": 1.0559105431309903,
+      "grad_norm": 1.1111577119822706,
+      "learning_rate": 8.188552059057924e-06,
+      "loss": 0.116,
+      "step": 2644
+    },
+    {
+      "epoch": 1.0567092651757188,
+      "grad_norm": 1.0697354896619498,
+      "learning_rate": 8.184970957322287e-06,
+      "loss": 0.1026,
+      "step": 2646
+    },
+    {
+      "epoch": 1.0575079872204474,
+      "grad_norm": 1.1184255850296823,
+      "learning_rate": 8.181387104072252e-06,
+      "loss": 0.1008,
+      "step": 2648
+    },
+    {
+      "epoch": 1.0583067092651757,
+      "grad_norm": 1.123703636746383,
+      "learning_rate": 8.177800502403928e-06,
+      "loss": 0.1234,
+      "step": 2650
+    },
+    {
+      "epoch": 1.0591054313099042,
+      "grad_norm": 1.1145099095973394,
+      "learning_rate": 8.1742111554158e-06,
+      "loss": 0.1038,
+      "step": 2652
+    },
+    {
+      "epoch": 1.0599041533546325,
+      "grad_norm": 1.0924462066716556,
+      "learning_rate": 8.170619066208723e-06,
+      "loss": 0.1026,
+      "step": 2654
+    },
+    {
+      "epoch": 1.060702875399361,
+      "grad_norm": 1.275295631090924,
+      "learning_rate": 8.167024237885927e-06,
+      "loss": 0.1202,
+      "step": 2656
+    },
+    {
+      "epoch": 1.0615015974440896,
+      "grad_norm": 1.001711382132048,
+      "learning_rate": 8.163426673553002e-06,
+      "loss": 0.1028,
+      "step": 2658
+    },
+    {
+      "epoch": 1.0623003194888179,
+      "grad_norm": 1.090098993511894,
+      "learning_rate": 8.159826376317906e-06,
+      "loss": 0.1074,
+      "step": 2660
+    },
+    {
+      "epoch": 1.0630990415335464,
+      "grad_norm": 1.0661454669618209,
+      "learning_rate": 8.156223349290957e-06,
+      "loss": 0.1101,
+      "step": 2662
+    },
+    {
+      "epoch": 1.0638977635782747,
+      "grad_norm": 1.256680601891414,
+      "learning_rate": 8.152617595584827e-06,
+      "loss": 0.1085,
+      "step": 2664
+    },
+    {
+      "epoch": 1.0646964856230032,
+      "grad_norm": 1.0814675764007797,
+      "learning_rate": 8.149009118314549e-06,
+      "loss": 0.1041,
+      "step": 2666
+    },
+    {
+      "epoch": 1.0654952076677315,
+      "grad_norm": 1.055396058607972,
+      "learning_rate": 8.145397920597505e-06,
+      "loss": 0.1012,
+      "step": 2668
+    },
+    {
+      "epoch": 1.06629392971246,
+      "grad_norm": 1.0902362525938407,
+      "learning_rate": 8.141784005553433e-06,
+      "loss": 0.1129,
+      "step": 2670
+    },
+    {
+      "epoch": 1.0670926517571886,
+      "grad_norm": 1.040946925167894,
+      "learning_rate": 8.138167376304411e-06,
+      "loss": 0.1032,
+      "step": 2672
+    },
+    {
+      "epoch": 1.067891373801917,
+      "grad_norm": 1.1293507367739193,
+      "learning_rate": 8.134548035974868e-06,
+      "loss": 0.1093,
+      "step": 2674
+    },
+    {
+      "epoch": 1.0686900958466454,
+      "grad_norm": 1.1475487980987538,
+      "learning_rate": 8.13092598769157e-06,
+      "loss": 0.1168,
+      "step": 2676
+    },
+    {
+      "epoch": 1.0694888178913737,
+      "grad_norm": 1.0688926041290217,
+      "learning_rate": 8.127301234583627e-06,
+      "loss": 0.1209,
+      "step": 2678
+    },
+    {
+      "epoch": 1.0702875399361023,
+      "grad_norm": 1.1130827659156695,
+      "learning_rate": 8.123673779782481e-06,
+      "loss": 0.1112,
+      "step": 2680
+    },
+    {
+      "epoch": 1.0710862619808306,
+      "grad_norm": 0.999458562659492,
+      "learning_rate": 8.120043626421915e-06,
+      "loss": 0.1036,
+      "step": 2682
+    },
+    {
+      "epoch": 1.071884984025559,
+      "grad_norm": 1.1445057181311435,
+      "learning_rate": 8.116410777638035e-06,
+      "loss": 0.1088,
+      "step": 2684
+    },
+    {
+      "epoch": 1.0726837060702876,
+      "grad_norm": 1.060266912369435,
+      "learning_rate": 8.112775236569282e-06,
+      "loss": 0.1024,
+      "step": 2686
+    },
+    {
+      "epoch": 1.073482428115016,
+      "grad_norm": 1.0993996765043779,
+      "learning_rate": 8.10913700635642e-06,
+      "loss": 0.1037,
+      "step": 2688
+    },
+    {
+      "epoch": 1.0742811501597445,
+      "grad_norm": 1.039132296487209,
+      "learning_rate": 8.105496090142535e-06,
+      "loss": 0.1075,
+      "step": 2690
+    },
+    {
+      "epoch": 1.0750798722044728,
+      "grad_norm": 1.1082486440070727,
+      "learning_rate": 8.101852491073036e-06,
+      "loss": 0.105,
+      "step": 2692
+    },
+    {
+      "epoch": 1.0758785942492013,
+      "grad_norm": 1.1651294916681243,
+      "learning_rate": 8.098206212295652e-06,
+      "loss": 0.1154,
+      "step": 2694
+    },
+    {
+      "epoch": 1.0766773162939298,
+      "grad_norm": 1.172460162145474,
+      "learning_rate": 8.094557256960419e-06,
+      "loss": 0.116,
+      "step": 2696
+    },
+    {
+      "epoch": 1.0774760383386581,
+      "grad_norm": 0.9459826583314704,
+      "learning_rate": 8.090905628219693e-06,
+      "loss": 0.1068,
+      "step": 2698
+    },
+    {
+      "epoch": 1.0782747603833867,
+      "grad_norm": 1.0709333935677605,
+      "learning_rate": 8.087251329228136e-06,
+      "loss": 0.0917,
+      "step": 2700
+    },
+    {
+      "epoch": 1.079073482428115,
+      "grad_norm": 1.1016086096260114,
+      "learning_rate": 8.083594363142717e-06,
+      "loss": 0.1144,
+      "step": 2702
+    },
+    {
+      "epoch": 1.0798722044728435,
+      "grad_norm": 1.0018060470975783,
+      "learning_rate": 8.079934733122708e-06,
+      "loss": 0.1071,
+      "step": 2704
+    },
+    {
+      "epoch": 1.0806709265175718,
+      "grad_norm": 1.145863183921058,
+      "learning_rate": 8.07627244232969e-06,
+      "loss": 0.1177,
+      "step": 2706
+    },
+    {
+      "epoch": 1.0814696485623003,
+      "grad_norm": 1.0861499113619286,
+      "learning_rate": 8.07260749392753e-06,
+      "loss": 0.116,
+      "step": 2708
+    },
+    {
+      "epoch": 1.0822683706070289,
+      "grad_norm": 1.0911649754162966,
+      "learning_rate": 8.068939891082401e-06,
+      "loss": 0.1092,
+      "step": 2710
+    },
+    {
+      "epoch": 1.0830670926517572,
+      "grad_norm": 1.1624721618006877,
+      "learning_rate": 8.065269636962765e-06,
+      "loss": 0.1035,
+      "step": 2712
+    },
+    {
+      "epoch": 1.0838658146964857,
+      "grad_norm": 1.0820597055920262,
+      "learning_rate": 8.061596734739377e-06,
+      "loss": 0.1005,
+      "step": 2714
+    },
+    {
+      "epoch": 1.084664536741214,
+      "grad_norm": 1.1521338536239825,
+      "learning_rate": 8.057921187585274e-06,
+      "loss": 0.1021,
+      "step": 2716
+    },
+    {
+      "epoch": 1.0854632587859425,
+      "grad_norm": 1.088468478053058,
+      "learning_rate": 8.054242998675787e-06,
+      "loss": 0.1133,
+      "step": 2718
+    },
+    {
+      "epoch": 1.0862619808306708,
+      "grad_norm": 1.139917756342621,
+      "learning_rate": 8.05056217118852e-06,
+      "loss": 0.108,
+      "step": 2720
+    },
+    {
+      "epoch": 1.0870607028753994,
+      "grad_norm": 1.128084026950435,
+      "learning_rate": 8.046878708303363e-06,
+      "loss": 0.1026,
+      "step": 2722
+    },
+    {
+      "epoch": 1.0878594249201279,
+      "grad_norm": 1.1291139425698837,
+      "learning_rate": 8.04319261320248e-06,
+      "loss": 0.1095,
+      "step": 2724
+    },
+    {
+      "epoch": 1.0886581469648562,
+      "grad_norm": 1.3447212908276018,
+      "learning_rate": 8.039503889070306e-06,
+      "loss": 0.1194,
+      "step": 2726
+    },
+    {
+      "epoch": 1.0894568690095847,
+      "grad_norm": 1.0745254740908237,
+      "learning_rate": 8.035812539093557e-06,
+      "loss": 0.1093,
+      "step": 2728
+    },
+    {
+      "epoch": 1.090255591054313,
+      "grad_norm": 1.0828953637676413,
+      "learning_rate": 8.032118566461206e-06,
+      "loss": 0.1108,
+      "step": 2730
+    },
+    {
+      "epoch": 1.0910543130990416,
+      "grad_norm": 1.0548110887259554,
+      "learning_rate": 8.0284219743645e-06,
+      "loss": 0.1062,
+      "step": 2732
+    },
+    {
+      "epoch": 1.09185303514377,
+      "grad_norm": 1.1071922672380121,
+      "learning_rate": 8.024722765996944e-06,
+      "loss": 0.1037,
+      "step": 2734
+    },
+    {
+      "epoch": 1.0926517571884984,
+      "grad_norm": 1.230092777293863,
+      "learning_rate": 8.021020944554305e-06,
+      "loss": 0.1154,
+      "step": 2736
+    },
+    {
+      "epoch": 1.093450479233227,
+      "grad_norm": 1.0630731902655604,
+      "learning_rate": 8.017316513234607e-06,
+      "loss": 0.1152,
+      "step": 2738
+    },
+    {
+      "epoch": 1.0942492012779552,
+      "grad_norm": 1.1292429935560224,
+      "learning_rate": 8.013609475238132e-06,
+      "loss": 0.11,
+      "step": 2740
+    },
+    {
+      "epoch": 1.0950479233226837,
+      "grad_norm": 1.1249867737292523,
+      "learning_rate": 8.009899833767407e-06,
+      "loss": 0.1032,
+      "step": 2742
+    },
+    {
+      "epoch": 1.095846645367412,
+      "grad_norm": 1.430619873698458,
+      "learning_rate": 8.006187592027215e-06,
+      "loss": 0.1178,
+      "step": 2744
+    },
+    {
+      "epoch": 1.0966453674121406,
+      "grad_norm": 1.110036751210643,
+      "learning_rate": 8.002472753224582e-06,
+      "loss": 0.1145,
+      "step": 2746
+    },
+    {
+      "epoch": 1.097444089456869,
+      "grad_norm": 0.957513352594371,
+      "learning_rate": 7.998755320568778e-06,
+      "loss": 0.1018,
+      "step": 2748
+    },
+    {
+      "epoch": 1.0982428115015974,
+      "grad_norm": 1.042054355253303,
+      "learning_rate": 7.995035297271313e-06,
+      "loss": 0.1013,
+      "step": 2750
+    },
+    {
+      "epoch": 1.099041533546326,
+      "grad_norm": 1.1532599431868966,
+      "learning_rate": 7.991312686545939e-06,
+      "loss": 0.1173,
+      "step": 2752
+    },
+    {
+      "epoch": 1.0998402555910542,
+      "grad_norm": 1.2036914344298868,
+      "learning_rate": 7.987587491608636e-06,
+      "loss": 0.1186,
+      "step": 2754
+    },
+    {
+      "epoch": 1.1006389776357828,
+      "grad_norm": 1.082895054115626,
+      "learning_rate": 7.983859715677627e-06,
+      "loss": 0.1006,
+      "step": 2756
+    },
+    {
+      "epoch": 1.101437699680511,
+      "grad_norm": 1.1041877219800416,
+      "learning_rate": 7.980129361973355e-06,
+      "loss": 0.1061,
+      "step": 2758
+    },
+    {
+      "epoch": 1.1022364217252396,
+      "grad_norm": 1.0649445718677086,
+      "learning_rate": 7.976396433718492e-06,
+      "loss": 0.1034,
+      "step": 2760
+    },
+    {
+      "epoch": 1.1030351437699681,
+      "grad_norm": 1.1926141701225625,
+      "learning_rate": 7.97266093413794e-06,
+      "loss": 0.1158,
+      "step": 2762
+    },
+    {
+      "epoch": 1.1038338658146964,
+      "grad_norm": 1.2011616729379744,
+      "learning_rate": 7.968922866458812e-06,
+      "loss": 0.1166,
+      "step": 2764
+    },
+    {
+      "epoch": 1.104632587859425,
+      "grad_norm": 1.1032014477011283,
+      "learning_rate": 7.965182233910453e-06,
+      "loss": 0.1079,
+      "step": 2766
+    },
+    {
+      "epoch": 1.1054313099041533,
+      "grad_norm": 1.0491357869845093,
+      "learning_rate": 7.961439039724413e-06,
+      "loss": 0.1185,
+      "step": 2768
+    },
+    {
+      "epoch": 1.1062300319488818,
+      "grad_norm": 1.1311466555151601,
+      "learning_rate": 7.95769328713446e-06,
+      "loss": 0.1182,
+      "step": 2770
+    },
+    {
+      "epoch": 1.1070287539936103,
+      "grad_norm": 1.2237403558053657,
+      "learning_rate": 7.953944979376567e-06,
+      "loss": 0.1097,
+      "step": 2772
+    },
+    {
+      "epoch": 1.1078274760383386,
+      "grad_norm": 1.2431901555713765,
+      "learning_rate": 7.950194119688922e-06,
+      "loss": 0.1329,
+      "step": 2774
+    },
+    {
+      "epoch": 1.1086261980830672,
+      "grad_norm": 1.0880599521982486,
+      "learning_rate": 7.946440711311913e-06,
+      "loss": 0.1162,
+      "step": 2776
+    },
+    {
+      "epoch": 1.1094249201277955,
+      "grad_norm": 1.1073519932980254,
+      "learning_rate": 7.942684757488133e-06,
+      "loss": 0.1027,
+      "step": 2778
+    },
+    {
+      "epoch": 1.110223642172524,
+      "grad_norm": 0.9960392472144209,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.1002,
+      "step": 2780
+    },
+    {
+      "epoch": 1.1110223642172523,
+      "grad_norm": 1.1223224889936787,
+      "learning_rate": 7.935165226481603e-06,
+      "loss": 0.1138,
+      "step": 2782
+    },
+    {
+      "epoch": 1.1118210862619808,
+      "grad_norm": 1.0435684073325897,
+      "learning_rate": 7.931401655795021e-06,
+      "loss": 0.1074,
+      "step": 2784
+    },
+    {
+      "epoch": 1.1126198083067094,
+      "grad_norm": 1.0106236333934397,
+      "learning_rate": 7.92763555265399e-06,
+      "loss": 0.0992,
+      "step": 2786
+    },
+    {
+      "epoch": 1.1134185303514377,
+      "grad_norm": 1.2351704038049207,
+      "learning_rate": 7.923866920312069e-06,
+      "loss": 0.1091,
+      "step": 2788
+    },
+    {
+      "epoch": 1.1142172523961662,
+      "grad_norm": 1.2422399561551012,
+      "learning_rate": 7.920095762025e-06,
+      "loss": 0.1189,
+      "step": 2790
+    },
+    {
+      "epoch": 1.1150159744408945,
+      "grad_norm": 1.0941653259345385,
+      "learning_rate": 7.916322081050708e-06,
+      "loss": 0.1003,
+      "step": 2792
+    },
+    {
+      "epoch": 1.115814696485623,
+      "grad_norm": 1.1760490829802384,
+      "learning_rate": 7.912545880649299e-06,
+      "loss": 0.1267,
+      "step": 2794
+    },
+    {
+      "epoch": 1.1166134185303513,
+      "grad_norm": 1.158226709859037,
+      "learning_rate": 7.90876716408305e-06,
+      "loss": 0.0968,
+      "step": 2796
+    },
+    {
+      "epoch": 1.1174121405750799,
+      "grad_norm": 1.1375503575879282,
+      "learning_rate": 7.904985934616419e-06,
+      "loss": 0.1077,
+      "step": 2798
+    },
+    {
+      "epoch": 1.1182108626198084,
+      "grad_norm": 0.9183058494090953,
+      "learning_rate": 7.90120219551603e-06,
+      "loss": 0.1032,
+      "step": 2800
+    },
+    {
+      "epoch": 1.1190095846645367,
+      "grad_norm": 1.0610420482924088,
+      "learning_rate": 7.897415950050676e-06,
+      "loss": 0.1111,
+      "step": 2802
+    },
+    {
+      "epoch": 1.1198083067092652,
+      "grad_norm": 1.1233765060725236,
+      "learning_rate": 7.893627201491319e-06,
+      "loss": 0.0985,
+      "step": 2804
+    },
+    {
+      "epoch": 1.1206070287539935,
+      "grad_norm": 1.0398157065300724,
+      "learning_rate": 7.889835953111075e-06,
+      "loss": 0.1164,
+      "step": 2806
+    },
+    {
+      "epoch": 1.121405750798722,
+      "grad_norm": 1.0997762546624286,
+      "learning_rate": 7.88604220818523e-06,
+      "loss": 0.1097,
+      "step": 2808
+    },
+    {
+      "epoch": 1.1222044728434506,
+      "grad_norm": 1.0435817236551026,
+      "learning_rate": 7.88224596999122e-06,
+      "loss": 0.1046,
+      "step": 2810
+    },
+    {
+      "epoch": 1.123003194888179,
+      "grad_norm": 1.2695278909995464,
+      "learning_rate": 7.878447241808634e-06,
+      "loss": 0.1102,
+      "step": 2812
+    },
+    {
+      "epoch": 1.1238019169329074,
+      "grad_norm": 1.081882759084368,
+      "learning_rate": 7.874646026919219e-06,
+      "loss": 0.114,
+      "step": 2814
+    },
+    {
+      "epoch": 1.1246006389776357,
+      "grad_norm": 1.2277213100870545,
+      "learning_rate": 7.870842328606863e-06,
+      "loss": 0.118,
+      "step": 2816
+    },
+    {
+      "epoch": 1.1253993610223643,
+      "grad_norm": 1.0784722478718025,
+      "learning_rate": 7.867036150157603e-06,
+      "loss": 0.1132,
+      "step": 2818
+    },
+    {
+      "epoch": 1.1261980830670926,
+      "grad_norm": 1.1255206572769305,
+      "learning_rate": 7.86322749485962e-06,
+      "loss": 0.1101,
+      "step": 2820
+    },
+    {
+      "epoch": 1.126996805111821,
+      "grad_norm": 1.162832702222223,
+      "learning_rate": 7.85941636600323e-06,
+      "loss": 0.1161,
+      "step": 2822
+    },
+    {
+      "epoch": 1.1277955271565494,
+      "grad_norm": 1.1447345276638636,
+      "learning_rate": 7.85560276688089e-06,
+      "loss": 0.1162,
+      "step": 2824
+    },
+    {
+      "epoch": 1.128594249201278,
+      "grad_norm": 1.187515171501569,
+      "learning_rate": 7.851786700787191e-06,
+      "loss": 0.1046,
+      "step": 2826
+    },
+    {
+      "epoch": 1.1293929712460065,
+      "grad_norm": 1.1412349222191334,
+      "learning_rate": 7.847968171018851e-06,
+      "loss": 0.1097,
+      "step": 2828
+    },
+    {
+      "epoch": 1.1301916932907348,
+      "grad_norm": 0.9891747869008474,
+      "learning_rate": 7.844147180874723e-06,
+      "loss": 0.0948,
+      "step": 2830
+    },
+    {
+      "epoch": 1.1309904153354633,
+      "grad_norm": 1.140395014058789,
+      "learning_rate": 7.84032373365578e-06,
+      "loss": 0.1038,
+      "step": 2832
+    },
+    {
+      "epoch": 1.1317891373801916,
+      "grad_norm": 1.0508986553012427,
+      "learning_rate": 7.836497832665119e-06,
+      "loss": 0.1091,
+      "step": 2834
+    },
+    {
+      "epoch": 1.1325878594249201,
+      "grad_norm": 1.016036817043171,
+      "learning_rate": 7.83266948120796e-06,
+      "loss": 0.0973,
+      "step": 2836
+    },
+    {
+      "epoch": 1.1333865814696487,
+      "grad_norm": 1.130667401388963,
+      "learning_rate": 7.828838682591635e-06,
+      "loss": 0.1207,
+      "step": 2838
+    },
+    {
+      "epoch": 1.134185303514377,
+      "grad_norm": 1.039182888114013,
+      "learning_rate": 7.825005440125595e-06,
+      "loss": 0.106,
+      "step": 2840
+    },
+    {
+      "epoch": 1.1349840255591055,
+      "grad_norm": 1.1559863889230508,
+      "learning_rate": 7.821169757121395e-06,
+      "loss": 0.1113,
+      "step": 2842
+    },
+    {
+      "epoch": 1.1357827476038338,
+      "grad_norm": 1.0074524350682563,
+      "learning_rate": 7.817331636892708e-06,
+      "loss": 0.1171,
+      "step": 2844
+    },
+    {
+      "epoch": 1.1365814696485623,
+      "grad_norm": 1.1004441537401717,
+      "learning_rate": 7.813491082755306e-06,
+      "loss": 0.0954,
+      "step": 2846
+    },
+    {
+      "epoch": 1.1373801916932909,
+      "grad_norm": 1.0467946843112212,
+      "learning_rate": 7.809648098027067e-06,
+      "loss": 0.1098,
+      "step": 2848
+    },
+    {
+      "epoch": 1.1381789137380192,
+      "grad_norm": 1.1629386911794455,
+      "learning_rate": 7.805802686027965e-06,
+      "loss": 0.1186,
+      "step": 2850
+    },
+    {
+      "epoch": 1.1389776357827477,
+      "grad_norm": 1.0232023211944392,
+      "learning_rate": 7.801954850080075e-06,
+      "loss": 0.0925,
+      "step": 2852
+    },
+    {
+      "epoch": 1.139776357827476,
+      "grad_norm": 1.2284293666842674,
+      "learning_rate": 7.798104593507562e-06,
+      "loss": 0.1137,
+      "step": 2854
+    },
+    {
+      "epoch": 1.1405750798722045,
+      "grad_norm": 1.0439656778352415,
+      "learning_rate": 7.794251919636687e-06,
+      "loss": 0.1086,
+      "step": 2856
+    },
+    {
+      "epoch": 1.1413738019169328,
+      "grad_norm": 1.038970880098722,
+      "learning_rate": 7.790396831795792e-06,
+      "loss": 0.101,
+      "step": 2858
+    },
+    {
+      "epoch": 1.1421725239616614,
+      "grad_norm": 1.0250967712229946,
+      "learning_rate": 7.786539333315316e-06,
+      "loss": 0.0997,
+      "step": 2860
+    },
+    {
+      "epoch": 1.1429712460063897,
+      "grad_norm": 1.2328372564490289,
+      "learning_rate": 7.782679427527768e-06,
+      "loss": 0.1113,
+      "step": 2862
+    },
+    {
+      "epoch": 1.1437699680511182,
+      "grad_norm": 1.0417991972771192,
+      "learning_rate": 7.778817117767748e-06,
+      "loss": 0.1005,
+      "step": 2864
+    },
+    {
+      "epoch": 1.1445686900958467,
+      "grad_norm": 1.0282768724124047,
+      "learning_rate": 7.77495240737192e-06,
+      "loss": 0.1025,
+      "step": 2866
+    },
+    {
+      "epoch": 1.145367412140575,
+      "grad_norm": 1.0115415567641788,
+      "learning_rate": 7.771085299679033e-06,
+      "loss": 0.0994,
+      "step": 2868
+    },
+    {
+      "epoch": 1.1461661341853036,
+      "grad_norm": 1.122551661496836,
+      "learning_rate": 7.767215798029906e-06,
+      "loss": 0.1009,
+      "step": 2870
+    },
+    {
+      "epoch": 1.1469648562300319,
+      "grad_norm": 1.2689706066461852,
+      "learning_rate": 7.76334390576742e-06,
+      "loss": 0.117,
+      "step": 2872
+    },
+    {
+      "epoch": 1.1477635782747604,
+      "grad_norm": 1.1324363654209109,
+      "learning_rate": 7.759469626236523e-06,
+      "loss": 0.1068,
+      "step": 2874
+    },
+    {
+      "epoch": 1.148562300319489,
+      "grad_norm": 1.162099881311204,
+      "learning_rate": 7.75559296278423e-06,
+      "loss": 0.1149,
+      "step": 2876
+    },
+    {
+      "epoch": 1.1493610223642172,
+      "grad_norm": 1.1144526194493574,
+      "learning_rate": 7.751713918759612e-06,
+      "loss": 0.1189,
+      "step": 2878
+    },
+    {
+      "epoch": 1.1501597444089458,
+      "grad_norm": 1.0366214429764016,
+      "learning_rate": 7.747832497513797e-06,
+      "loss": 0.1143,
+      "step": 2880
+    },
+    {
+      "epoch": 1.150958466453674,
+      "grad_norm": 1.033901501799641,
+      "learning_rate": 7.743948702399966e-06,
+      "loss": 0.1038,
+      "step": 2882
+    },
+    {
+      "epoch": 1.1517571884984026,
+      "grad_norm": 1.091707818538779,
+      "learning_rate": 7.740062536773352e-06,
+      "loss": 0.1018,
+      "step": 2884
+    },
+    {
+      "epoch": 1.1525559105431311,
+      "grad_norm": 1.086352546060504,
+      "learning_rate": 7.736174003991234e-06,
+      "loss": 0.113,
+      "step": 2886
+    },
+    {
+      "epoch": 1.1533546325878594,
+      "grad_norm": 1.0324848912394144,
+      "learning_rate": 7.732283107412938e-06,
+      "loss": 0.1012,
+      "step": 2888
+    },
+    {
+      "epoch": 1.154153354632588,
+      "grad_norm": 1.180351977551197,
+      "learning_rate": 7.728389850399834e-06,
+      "loss": 0.107,
+      "step": 2890
+    },
+    {
+      "epoch": 1.1549520766773163,
+      "grad_norm": 1.1213498513689995,
+      "learning_rate": 7.724494236315327e-06,
+      "loss": 0.118,
+      "step": 2892
+    },
+    {
+      "epoch": 1.1557507987220448,
+      "grad_norm": 1.1297732537347598,
+      "learning_rate": 7.72059626852486e-06,
+      "loss": 0.1052,
+      "step": 2894
+    },
+    {
+      "epoch": 1.156549520766773,
+      "grad_norm": 1.1260273445244882,
+      "learning_rate": 7.71669595039591e-06,
+      "loss": 0.1231,
+      "step": 2896
+    },
+    {
+      "epoch": 1.1573482428115016,
+      "grad_norm": 1.0804349790876964,
+      "learning_rate": 7.712793285297982e-06,
+      "loss": 0.102,
+      "step": 2898
+    },
+    {
+      "epoch": 1.15814696485623,
+      "grad_norm": 1.0887409006629178,
+      "learning_rate": 7.70888827660261e-06,
+      "loss": 0.1038,
+      "step": 2900
+    },
+    {
+      "epoch": 1.1589456869009584,
+      "grad_norm": 1.3253348724046994,
+      "learning_rate": 7.704980927683359e-06,
+      "loss": 0.1085,
+      "step": 2902
+    },
+    {
+      "epoch": 1.159744408945687,
+      "grad_norm": 1.2297063221974824,
+      "learning_rate": 7.701071241915804e-06,
+      "loss": 0.1114,
+      "step": 2904
+    },
+    {
+      "epoch": 1.1605431309904153,
+      "grad_norm": 1.1163333185767002,
+      "learning_rate": 7.697159222677544e-06,
+      "loss": 0.1065,
+      "step": 2906
+    },
+    {
+      "epoch": 1.1613418530351438,
+      "grad_norm": 1.108863651299088,
+      "learning_rate": 7.693244873348197e-06,
+      "loss": 0.1128,
+      "step": 2908
+    },
+    {
+      "epoch": 1.1621405750798721,
+      "grad_norm": 1.0477128725454634,
+      "learning_rate": 7.689328197309394e-06,
+      "loss": 0.1014,
+      "step": 2910
+    },
+    {
+      "epoch": 1.1629392971246006,
+      "grad_norm": 1.1667151644208233,
+      "learning_rate": 7.685409197944768e-06,
+      "loss": 0.0998,
+      "step": 2912
+    },
+    {
+      "epoch": 1.1637380191693292,
+      "grad_norm": 1.2632234446787278,
+      "learning_rate": 7.681487878639968e-06,
+      "loss": 0.1067,
+      "step": 2914
+    },
+    {
+      "epoch": 1.1645367412140575,
+      "grad_norm": 1.288703128597724,
+      "learning_rate": 7.677564242782645e-06,
+      "loss": 0.1093,
+      "step": 2916
+    },
+    {
+      "epoch": 1.165335463258786,
+      "grad_norm": 1.151244739656633,
+      "learning_rate": 7.673638293762447e-06,
+      "loss": 0.1086,
+      "step": 2918
+    },
+    {
+      "epoch": 1.1661341853035143,
+      "grad_norm": 1.1676995737767575,
+      "learning_rate": 7.669710034971025e-06,
+      "loss": 0.0907,
+      "step": 2920
+    },
+    {
+      "epoch": 1.1669329073482428,
+      "grad_norm": 1.0099832716415298,
+      "learning_rate": 7.665779469802024e-06,
+      "loss": 0.1039,
+      "step": 2922
+    },
+    {
+      "epoch": 1.1677316293929714,
+      "grad_norm": 1.2160567960293138,
+      "learning_rate": 7.661846601651081e-06,
+      "loss": 0.1054,
+      "step": 2924
+    },
+    {
+      "epoch": 1.1685303514376997,
+      "grad_norm": 1.2002285628920948,
+      "learning_rate": 7.657911433915824e-06,
+      "loss": 0.1236,
+      "step": 2926
+    },
+    {
+      "epoch": 1.1693290734824282,
+      "grad_norm": 1.1011958816133838,
+      "learning_rate": 7.653973969995866e-06,
+      "loss": 0.1043,
+      "step": 2928
+    },
+    {
+      "epoch": 1.1701277955271565,
+      "grad_norm": 1.1222330810291938,
+      "learning_rate": 7.650034213292804e-06,
+      "loss": 0.0964,
+      "step": 2930
+    },
+    {
+      "epoch": 1.170926517571885,
+      "grad_norm": 1.191923146958323,
+      "learning_rate": 7.646092167210217e-06,
+      "loss": 0.1247,
+      "step": 2932
+    },
+    {
+      "epoch": 1.1717252396166133,
+      "grad_norm": 1.1057104845426344,
+      "learning_rate": 7.642147835153659e-06,
+      "loss": 0.1151,
+      "step": 2934
+    },
+    {
+      "epoch": 1.1725239616613419,
+      "grad_norm": 1.1144118547772344,
+      "learning_rate": 7.638201220530664e-06,
+      "loss": 0.106,
+      "step": 2936
+    },
+    {
+      "epoch": 1.1733226837060702,
+      "grad_norm": 1.03554140015377,
+      "learning_rate": 7.634252326750733e-06,
+      "loss": 0.1093,
+      "step": 2938
+    },
+    {
+      "epoch": 1.1741214057507987,
+      "grad_norm": 1.067085791357077,
+      "learning_rate": 7.630301157225336e-06,
+      "loss": 0.106,
+      "step": 2940
+    },
+    {
+      "epoch": 1.1749201277955272,
+      "grad_norm": 1.0936909170630784,
+      "learning_rate": 7.626347715367912e-06,
+      "loss": 0.1063,
+      "step": 2942
+    },
+    {
+      "epoch": 1.1757188498402555,
+      "grad_norm": 1.3484514118083,
+      "learning_rate": 7.622392004593862e-06,
+      "loss": 0.0992,
+      "step": 2944
+    },
+    {
+      "epoch": 1.176517571884984,
+      "grad_norm": 1.169969792758724,
+      "learning_rate": 7.618434028320547e-06,
+      "loss": 0.1107,
+      "step": 2946
+    },
+    {
+      "epoch": 1.1773162939297124,
+      "grad_norm": 1.0742999979059649,
+      "learning_rate": 7.614473789967285e-06,
+      "loss": 0.1039,
+      "step": 2948
+    },
+    {
+      "epoch": 1.178115015974441,
+      "grad_norm": 1.1077968599944528,
+      "learning_rate": 7.610511292955347e-06,
+      "loss": 0.1006,
+      "step": 2950
+    },
+    {
+      "epoch": 1.1789137380191694,
+      "grad_norm": 1.1624715070814888,
+      "learning_rate": 7.60654654070796e-06,
+      "loss": 0.0994,
+      "step": 2952
+    },
+    {
+      "epoch": 1.1797124600638977,
+      "grad_norm": 1.0968086205411158,
+      "learning_rate": 7.602579536650292e-06,
+      "loss": 0.1119,
+      "step": 2954
+    },
+    {
+      "epoch": 1.1805111821086263,
+      "grad_norm": 1.168735576631901,
+      "learning_rate": 7.59861028420946e-06,
+      "loss": 0.1136,
+      "step": 2956
+    },
+    {
+      "epoch": 1.1813099041533546,
+      "grad_norm": 1.2384061026226614,
+      "learning_rate": 7.594638786814526e-06,
+      "loss": 0.1176,
+      "step": 2958
+    },
+    {
+      "epoch": 1.182108626198083,
+      "grad_norm": 1.241137963433562,
+      "learning_rate": 7.59066504789649e-06,
+      "loss": 0.12,
+      "step": 2960
+    },
+    {
+      "epoch": 1.1829073482428114,
+      "grad_norm": 1.1850341035866552,
+      "learning_rate": 7.586689070888284e-06,
+      "loss": 0.119,
+      "step": 2962
+    },
+    {
+      "epoch": 1.18370607028754,
+      "grad_norm": 1.1220698860625842,
+      "learning_rate": 7.58271085922478e-06,
+      "loss": 0.1121,
+      "step": 2964
+    },
+    {
+      "epoch": 1.1845047923322685,
+      "grad_norm": 1.037375641928384,
+      "learning_rate": 7.578730416342777e-06,
+      "loss": 0.1102,
+      "step": 2966
+    },
+    {
+      "epoch": 1.1853035143769968,
+      "grad_norm": 1.2171636076950145,
+      "learning_rate": 7.574747745681e-06,
+      "loss": 0.1163,
+      "step": 2968
+    },
+    {
+      "epoch": 1.1861022364217253,
+      "grad_norm": 1.1555647145995425,
+      "learning_rate": 7.5707628506801015e-06,
+      "loss": 0.1033,
+      "step": 2970
+    },
+    {
+      "epoch": 1.1869009584664536,
+      "grad_norm": 1.3403187676618649,
+      "learning_rate": 7.566775734782656e-06,
+      "loss": 0.1066,
+      "step": 2972
+    },
+    {
+      "epoch": 1.1876996805111821,
+      "grad_norm": 1.1035130819507677,
+      "learning_rate": 7.562786401433156e-06,
+      "loss": 0.1073,
+      "step": 2974
+    },
+    {
+      "epoch": 1.1884984025559104,
+      "grad_norm": 1.0839124042096482,
+      "learning_rate": 7.558794854078006e-06,
+      "loss": 0.1101,
+      "step": 2976
+    },
+    {
+      "epoch": 1.189297124600639,
+      "grad_norm": 1.019947468283778,
+      "learning_rate": 7.5548010961655295e-06,
+      "loss": 0.1033,
+      "step": 2978
+    },
+    {
+      "epoch": 1.1900958466453675,
+      "grad_norm": 1.0753891042248978,
+      "learning_rate": 7.5508051311459555e-06,
+      "loss": 0.109,
+      "step": 2980
+    },
+    {
+      "epoch": 1.1908945686900958,
+      "grad_norm": 1.0714175737276472,
+      "learning_rate": 7.546806962471419e-06,
+      "loss": 0.1017,
+      "step": 2982
+    },
+    {
+      "epoch": 1.1916932907348243,
+      "grad_norm": 1.0997902745143622,
+      "learning_rate": 7.542806593595963e-06,
+      "loss": 0.106,
+      "step": 2984
+    },
+    {
+      "epoch": 1.1924920127795526,
+      "grad_norm": 1.0225175483161848,
+      "learning_rate": 7.538804027975528e-06,
+      "loss": 0.1092,
+      "step": 2986
+    },
+    {
+      "epoch": 1.1932907348242812,
+      "grad_norm": 1.107891680578376,
+      "learning_rate": 7.534799269067952e-06,
+      "loss": 0.1091,
+      "step": 2988
+    },
+    {
+      "epoch": 1.1940894568690097,
+      "grad_norm": 1.0551747572162382,
+      "learning_rate": 7.530792320332971e-06,
+      "loss": 0.1088,
+      "step": 2990
+    },
+    {
+      "epoch": 1.194888178913738,
+      "grad_norm": 1.0629259318761108,
+      "learning_rate": 7.526783185232208e-06,
+      "loss": 0.0937,
+      "step": 2992
+    },
+    {
+      "epoch": 1.1956869009584665,
+      "grad_norm": 1.1933108946432194,
+      "learning_rate": 7.522771867229179e-06,
+      "loss": 0.108,
+      "step": 2994
+    },
+    {
+      "epoch": 1.1964856230031948,
+      "grad_norm": 1.157609849649065,
+      "learning_rate": 7.518758369789286e-06,
+      "loss": 0.1192,
+      "step": 2996
+    },
+    {
+      "epoch": 1.1972843450479234,
+      "grad_norm": 1.167506126596324,
+      "learning_rate": 7.514742696379809e-06,
+      "loss": 0.1096,
+      "step": 2998
+    },
+    {
+      "epoch": 1.1980830670926517,
+      "grad_norm": 1.0333315229582538,
+      "learning_rate": 7.51072485046991e-06,
+      "loss": 0.1085,
+      "step": 3000
+    },
+    {
+      "epoch": 1.1980830670926517,
+      "eval_loss": 0.16306817531585693,
+      "eval_runtime": 417.6568,
+      "eval_samples_per_second": 42.635,
+      "eval_steps_per_second": 5.33,
+      "step": 3000
+    },
+    {
+      "epoch": 1.1988817891373802,
+      "grad_norm": 1.192881477161516,
+      "learning_rate": 7.5067048355306334e-06,
+      "loss": 0.1082,
+      "step": 3002
+    },
+    {
+      "epoch": 1.1996805111821087,
+      "grad_norm": 1.2452028368788162,
+      "learning_rate": 7.50268265503489e-06,
+      "loss": 0.1156,
+      "step": 3004
+    },
+    {
+      "epoch": 1.200479233226837,
+      "grad_norm": 1.1510650440674393,
+      "learning_rate": 7.498658312457464e-06,
+      "loss": 0.1133,
+      "step": 3006
+    },
+    {
+      "epoch": 1.2012779552715656,
+      "grad_norm": 1.0520484152103042,
+      "learning_rate": 7.494631811275008e-06,
+      "loss": 0.1118,
+      "step": 3008
+    },
+    {
+      "epoch": 1.2020766773162939,
+      "grad_norm": 1.0970981862348477,
+      "learning_rate": 7.49060315496604e-06,
+      "loss": 0.1072,
+      "step": 3010
+    },
+    {
+      "epoch": 1.2028753993610224,
+      "grad_norm": 1.1102889095542705,
+      "learning_rate": 7.486572347010937e-06,
+      "loss": 0.1056,
+      "step": 3012
+    },
+    {
+      "epoch": 1.2036741214057507,
+      "grad_norm": 1.178700112165599,
+      "learning_rate": 7.482539390891941e-06,
+      "loss": 0.1153,
+      "step": 3014
+    },
+    {
+      "epoch": 1.2044728434504792,
+      "grad_norm": 1.0483204125787329,
+      "learning_rate": 7.478504290093138e-06,
+      "loss": 0.0957,
+      "step": 3016
+    },
+    {
+      "epoch": 1.2052715654952078,
+      "grad_norm": 1.1855590333372437,
+      "learning_rate": 7.474467048100484e-06,
+      "loss": 0.1165,
+      "step": 3018
+    },
+    {
+      "epoch": 1.206070287539936,
+      "grad_norm": 1.132537958694911,
+      "learning_rate": 7.470427668401766e-06,
+      "loss": 0.1087,
+      "step": 3020
+    },
+    {
+      "epoch": 1.2068690095846646,
+      "grad_norm": 1.2170415750175325,
+      "learning_rate": 7.466386154486634e-06,
+      "loss": 0.1147,
+      "step": 3022
+    },
+    {
+      "epoch": 1.207667731629393,
+      "grad_norm": 1.1082959534882166,
+      "learning_rate": 7.462342509846571e-06,
+      "loss": 0.1048,
+      "step": 3024
+    },
+    {
+      "epoch": 1.2084664536741214,
+      "grad_norm": 1.1141000146197995,
+      "learning_rate": 7.458296737974905e-06,
+      "loss": 0.1132,
+      "step": 3026
+    },
+    {
+      "epoch": 1.20926517571885,
+      "grad_norm": 1.0298899347940416,
+      "learning_rate": 7.4542488423668005e-06,
+      "loss": 0.1064,
+      "step": 3028
+    },
+    {
+      "epoch": 1.2100638977635783,
+      "grad_norm": 1.083163833346684,
+      "learning_rate": 7.450198826519259e-06,
+      "loss": 0.1123,
+      "step": 3030
+    },
+    {
+      "epoch": 1.2108626198083068,
+      "grad_norm": 1.0762705695988668,
+      "learning_rate": 7.446146693931111e-06,
+      "loss": 0.1008,
+      "step": 3032
+    },
+    {
+      "epoch": 1.211661341853035,
+      "grad_norm": 1.1867201378486008,
+      "learning_rate": 7.442092448103019e-06,
+      "loss": 0.1256,
+      "step": 3034
+    },
+    {
+      "epoch": 1.2124600638977636,
+      "grad_norm": 1.1714410675076814,
+      "learning_rate": 7.438036092537465e-06,
+      "loss": 0.1039,
+      "step": 3036
+    },
+    {
+      "epoch": 1.213258785942492,
+      "grad_norm": 1.0370675696738088,
+      "learning_rate": 7.43397763073876e-06,
+      "loss": 0.1039,
+      "step": 3038
+    },
+    {
+      "epoch": 1.2140575079872205,
+      "grad_norm": 1.097720581265046,
+      "learning_rate": 7.42991706621303e-06,
+      "loss": 0.1143,
+      "step": 3040
+    },
+    {
+      "epoch": 1.2148562300319488,
+      "grad_norm": 0.9975002092372904,
+      "learning_rate": 7.4258544024682245e-06,
+      "loss": 0.1039,
+      "step": 3042
+    },
+    {
+      "epoch": 1.2156549520766773,
+      "grad_norm": 1.032389232375386,
+      "learning_rate": 7.421789643014096e-06,
+      "loss": 0.1029,
+      "step": 3044
+    },
+    {
+      "epoch": 1.2164536741214058,
+      "grad_norm": 1.159406643141239,
+      "learning_rate": 7.417722791362216e-06,
+      "loss": 0.121,
+      "step": 3046
+    },
+    {
+      "epoch": 1.2172523961661341,
+      "grad_norm": 1.075616311885778,
+      "learning_rate": 7.413653851025959e-06,
+      "loss": 0.1027,
+      "step": 3048
+    },
+    {
+      "epoch": 1.2180511182108626,
+      "grad_norm": 1.0677123331182898,
+      "learning_rate": 7.4095828255205085e-06,
+      "loss": 0.0933,
+      "step": 3050
+    },
+    {
+      "epoch": 1.218849840255591,
+      "grad_norm": 1.1351445749769558,
+      "learning_rate": 7.405509718362842e-06,
+      "loss": 0.1028,
+      "step": 3052
+    },
+    {
+      "epoch": 1.2196485623003195,
+      "grad_norm": 1.218634180761916,
+      "learning_rate": 7.401434533071745e-06,
+      "loss": 0.1146,
+      "step": 3054
+    },
+    {
+      "epoch": 1.220447284345048,
+      "grad_norm": 1.1682846848481725,
+      "learning_rate": 7.397357273167789e-06,
+      "loss": 0.1014,
+      "step": 3056
+    },
+    {
+      "epoch": 1.2212460063897763,
+      "grad_norm": 1.1641572787088017,
+      "learning_rate": 7.393277942173345e-06,
+      "loss": 0.1065,
+      "step": 3058
+    },
+    {
+      "epoch": 1.2220447284345048,
+      "grad_norm": 1.1514146815699455,
+      "learning_rate": 7.389196543612567e-06,
+      "loss": 0.1014,
+      "step": 3060
+    },
+    {
+      "epoch": 1.2228434504792332,
+      "grad_norm": 1.330134394264511,
+      "learning_rate": 7.3851130810113995e-06,
+      "loss": 0.118,
+      "step": 3062
+    },
+    {
+      "epoch": 1.2236421725239617,
+      "grad_norm": 1.1828426669934808,
+      "learning_rate": 7.381027557897568e-06,
+      "loss": 0.1054,
+      "step": 3064
+    },
+    {
+      "epoch": 1.2244408945686902,
+      "grad_norm": 1.1440466876810054,
+      "learning_rate": 7.376939977800581e-06,
+      "loss": 0.1157,
+      "step": 3066
+    },
+    {
+      "epoch": 1.2252396166134185,
+      "grad_norm": 1.1022139516494007,
+      "learning_rate": 7.372850344251722e-06,
+      "loss": 0.0968,
+      "step": 3068
+    },
+    {
+      "epoch": 1.226038338658147,
+      "grad_norm": 1.069906203066803,
+      "learning_rate": 7.368758660784048e-06,
+      "loss": 0.1154,
+      "step": 3070
+    },
+    {
+      "epoch": 1.2268370607028753,
+      "grad_norm": 1.066501858996855,
+      "learning_rate": 7.364664930932385e-06,
+      "loss": 0.1114,
+      "step": 3072
+    },
+    {
+      "epoch": 1.2276357827476039,
+      "grad_norm": 1.090807997716351,
+      "learning_rate": 7.360569158233332e-06,
+      "loss": 0.1058,
+      "step": 3074
+    },
+    {
+      "epoch": 1.2284345047923322,
+      "grad_norm": 1.1653540017717812,
+      "learning_rate": 7.356471346225249e-06,
+      "loss": 0.1141,
+      "step": 3076
+    },
+    {
+      "epoch": 1.2292332268370607,
+      "grad_norm": 1.2183346386496956,
+      "learning_rate": 7.352371498448261e-06,
+      "loss": 0.1086,
+      "step": 3078
+    },
+    {
+      "epoch": 1.230031948881789,
+      "grad_norm": 1.1495635595373217,
+      "learning_rate": 7.348269618444248e-06,
+      "loss": 0.1128,
+      "step": 3080
+    },
+    {
+      "epoch": 1.2308306709265175,
+      "grad_norm": 1.0657147313743496,
+      "learning_rate": 7.344165709756847e-06,
+      "loss": 0.1087,
+      "step": 3082
+    },
+    {
+      "epoch": 1.231629392971246,
+      "grad_norm": 1.1350112419460496,
+      "learning_rate": 7.340059775931449e-06,
+      "loss": 0.1121,
+      "step": 3084
+    },
+    {
+      "epoch": 1.2324281150159744,
+      "grad_norm": 1.1125316251538973,
+      "learning_rate": 7.33595182051519e-06,
+      "loss": 0.1112,
+      "step": 3086
+    },
+    {
+      "epoch": 1.233226837060703,
+      "grad_norm": 1.1893400382987338,
+      "learning_rate": 7.331841847056962e-06,
+      "loss": 0.1047,
+      "step": 3088
+    },
+    {
+      "epoch": 1.2340255591054312,
+      "grad_norm": 1.2177476387914874,
+      "learning_rate": 7.3277298591073895e-06,
+      "loss": 0.1136,
+      "step": 3090
+    },
+    {
+      "epoch": 1.2348242811501597,
+      "grad_norm": 1.0631740360903161,
+      "learning_rate": 7.323615860218844e-06,
+      "loss": 0.1144,
+      "step": 3092
+    },
+    {
+      "epoch": 1.2356230031948883,
+      "grad_norm": 1.092461962405361,
+      "learning_rate": 7.319499853945431e-06,
+      "loss": 0.0995,
+      "step": 3094
+    },
+    {
+      "epoch": 1.2364217252396166,
+      "grad_norm": 1.0890967378843055,
+      "learning_rate": 7.315381843842995e-06,
+      "loss": 0.1033,
+      "step": 3096
+    },
+    {
+      "epoch": 1.237220447284345,
+      "grad_norm": 1.2131460595478127,
+      "learning_rate": 7.3112618334691035e-06,
+      "loss": 0.1159,
+      "step": 3098
+    },
+    {
+      "epoch": 1.2380191693290734,
+      "grad_norm": 1.2188104551185952,
+      "learning_rate": 7.307139826383058e-06,
+      "loss": 0.1053,
+      "step": 3100
+    },
+    {
+      "epoch": 1.238817891373802,
+      "grad_norm": 1.0889591234624394,
+      "learning_rate": 7.303015826145886e-06,
+      "loss": 0.1012,
+      "step": 3102
+    },
+    {
+      "epoch": 1.2396166134185305,
+      "grad_norm": 1.1040750999295241,
+      "learning_rate": 7.298889836320334e-06,
+      "loss": 0.1013,
+      "step": 3104
+    },
+    {
+      "epoch": 1.2404153354632588,
+      "grad_norm": 1.100669836825261,
+      "learning_rate": 7.294761860470866e-06,
+      "loss": 0.1083,
+      "step": 3106
+    },
+    {
+      "epoch": 1.2412140575079873,
+      "grad_norm": 1.1773345051128925,
+      "learning_rate": 7.290631902163665e-06,
+      "loss": 0.11,
+      "step": 3108
+    },
+    {
+      "epoch": 1.2420127795527156,
+      "grad_norm": 1.2401290593387904,
+      "learning_rate": 7.286499964966625e-06,
+      "loss": 0.1203,
+      "step": 3110
+    },
+    {
+      "epoch": 1.2428115015974441,
+      "grad_norm": 1.1482156063703826,
+      "learning_rate": 7.282366052449351e-06,
+      "loss": 0.1122,
+      "step": 3112
+    },
+    {
+      "epoch": 1.2436102236421724,
+      "grad_norm": 1.4107571214662287,
+      "learning_rate": 7.278230168183152e-06,
+      "loss": 0.1183,
+      "step": 3114
+    },
+    {
+      "epoch": 1.244408945686901,
+      "grad_norm": 1.0325520625453315,
+      "learning_rate": 7.274092315741042e-06,
+      "loss": 0.1019,
+      "step": 3116
+    },
+    {
+      "epoch": 1.2452076677316293,
+      "grad_norm": 1.1107905673766356,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.1051,
+      "step": 3118
+    },
+    {
+      "epoch": 1.2460063897763578,
+      "grad_norm": 1.1281556579434084,
+      "learning_rate": 7.265810720629643e-06,
+      "loss": 0.1187,
+      "step": 3120
+    },
+    {
+      "epoch": 1.2468051118210863,
+      "grad_norm": 1.097863022792409,
+      "learning_rate": 7.261666985114871e-06,
+      "loss": 0.1172,
+      "step": 3122
+    },
+    {
+      "epoch": 1.2476038338658146,
+      "grad_norm": 1.1043922887663058,
+      "learning_rate": 7.257521295733214e-06,
+      "loss": 0.0985,
+      "step": 3124
+    },
+    {
+      "epoch": 1.2484025559105432,
+      "grad_norm": 1.1286717401317095,
+      "learning_rate": 7.253373656066159e-06,
+      "loss": 0.1152,
+      "step": 3126
+    },
+    {
+      "epoch": 1.2492012779552715,
+      "grad_norm": 1.1389798339479362,
+      "learning_rate": 7.249224069696876e-06,
+      "loss": 0.1098,
+      "step": 3128
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1580209218201918,
+      "learning_rate": 7.245072540210213e-06,
+      "loss": 0.1199,
+      "step": 3130
+    },
+    {
+      "epoch": 1.2507987220447285,
+      "grad_norm": 1.019767242278647,
+      "learning_rate": 7.2409190711927015e-06,
+      "loss": 0.0945,
+      "step": 3132
+    },
+    {
+      "epoch": 1.2515974440894568,
+      "grad_norm": 1.0864399680532384,
+      "learning_rate": 7.236763666232546e-06,
+      "loss": 0.1001,
+      "step": 3134
+    },
+    {
+      "epoch": 1.2523961661341854,
+      "grad_norm": 1.261597038834277,
+      "learning_rate": 7.232606328919627e-06,
+      "loss": 0.1046,
+      "step": 3136
+    },
+    {
+      "epoch": 1.2531948881789137,
+      "grad_norm": 1.1146593566415728,
+      "learning_rate": 7.228447062845487e-06,
+      "loss": 0.1201,
+      "step": 3138
+    },
+    {
+      "epoch": 1.2539936102236422,
+      "grad_norm": 1.2339789575226763,
+      "learning_rate": 7.224285871603344e-06,
+      "loss": 0.1071,
+      "step": 3140
+    },
+    {
+      "epoch": 1.2547923322683707,
+      "grad_norm": 1.1413529281364962,
+      "learning_rate": 7.2201227587880704e-06,
+      "loss": 0.1157,
+      "step": 3142
+    },
+    {
+      "epoch": 1.255591054313099,
+      "grad_norm": 1.1321943139020432,
+      "learning_rate": 7.215957727996208e-06,
+      "loss": 0.1136,
+      "step": 3144
+    },
+    {
+      "epoch": 1.2563897763578276,
+      "grad_norm": 1.1035881377949015,
+      "learning_rate": 7.211790782825945e-06,
+      "loss": 0.112,
+      "step": 3146
+    },
+    {
+      "epoch": 1.2571884984025559,
+      "grad_norm": 1.1454443734559139,
+      "learning_rate": 7.207621926877133e-06,
+      "loss": 0.107,
+      "step": 3148
+    },
+    {
+      "epoch": 1.2579872204472844,
+      "grad_norm": 1.0366769144913708,
+      "learning_rate": 7.203451163751268e-06,
+      "loss": 0.1025,
+      "step": 3150
+    },
+    {
+      "epoch": 1.2587859424920127,
+      "grad_norm": 1.1952040379853797,
+      "learning_rate": 7.199278497051498e-06,
+      "loss": 0.1259,
+      "step": 3152
+    },
+    {
+      "epoch": 1.2595846645367412,
+      "grad_norm": 1.1719876875826358,
+      "learning_rate": 7.195103930382609e-06,
+      "loss": 0.1102,
+      "step": 3154
+    },
+    {
+      "epoch": 1.2603833865814695,
+      "grad_norm": 1.2526525796976853,
+      "learning_rate": 7.190927467351037e-06,
+      "loss": 0.1219,
+      "step": 3156
+    },
+    {
+      "epoch": 1.261182108626198,
+      "grad_norm": 1.047570651725447,
+      "learning_rate": 7.186749111564852e-06,
+      "loss": 0.099,
+      "step": 3158
+    },
+    {
+      "epoch": 1.2619808306709266,
+      "grad_norm": 1.0251728815129595,
+      "learning_rate": 7.182568866633757e-06,
+      "loss": 0.1129,
+      "step": 3160
+    },
+    {
+      "epoch": 1.262779552715655,
+      "grad_norm": 1.0528626625673922,
+      "learning_rate": 7.178386736169087e-06,
+      "loss": 0.0972,
+      "step": 3162
+    },
+    {
+      "epoch": 1.2635782747603834,
+      "grad_norm": 1.1434985715089727,
+      "learning_rate": 7.174202723783815e-06,
+      "loss": 0.1108,
+      "step": 3164
+    },
+    {
+      "epoch": 1.2643769968051117,
+      "grad_norm": 1.053836775418805,
+      "learning_rate": 7.170016833092526e-06,
+      "loss": 0.111,
+      "step": 3166
+    },
+    {
+      "epoch": 1.2651757188498403,
+      "grad_norm": 1.1472854273249073,
+      "learning_rate": 7.16582906771144e-06,
+      "loss": 0.1101,
+      "step": 3168
+    },
+    {
+      "epoch": 1.2659744408945688,
+      "grad_norm": 1.019394721627375,
+      "learning_rate": 7.161639431258387e-06,
+      "loss": 0.1082,
+      "step": 3170
+    },
+    {
+      "epoch": 1.266773162939297,
+      "grad_norm": 1.1497720973804415,
+      "learning_rate": 7.157447927352821e-06,
+      "loss": 0.1038,
+      "step": 3172
+    },
+    {
+      "epoch": 1.2675718849840256,
+      "grad_norm": 0.931060363409149,
+      "learning_rate": 7.153254559615802e-06,
+      "loss": 0.0934,
+      "step": 3174
+    },
+    {
+      "epoch": 1.268370607028754,
+      "grad_norm": 1.1003773790175626,
+      "learning_rate": 7.149059331670009e-06,
+      "loss": 0.1116,
+      "step": 3176
+    },
+    {
+      "epoch": 1.2691693290734825,
+      "grad_norm": 0.9906365881640852,
+      "learning_rate": 7.144862247139716e-06,
+      "loss": 0.1048,
+      "step": 3178
+    },
+    {
+      "epoch": 1.269968051118211,
+      "grad_norm": 1.1404557162820228,
+      "learning_rate": 7.140663309650817e-06,
+      "loss": 0.1132,
+      "step": 3180
+    },
+    {
+      "epoch": 1.2707667731629393,
+      "grad_norm": 1.2239606071851836,
+      "learning_rate": 7.1364625228307915e-06,
+      "loss": 0.1124,
+      "step": 3182
+    },
+    {
+      "epoch": 1.2715654952076676,
+      "grad_norm": 1.1186497681482894,
+      "learning_rate": 7.132259890308726e-06,
+      "loss": 0.1072,
+      "step": 3184
+    },
+    {
+      "epoch": 1.2723642172523961,
+      "grad_norm": 1.1465574716494185,
+      "learning_rate": 7.128055415715295e-06,
+      "loss": 0.1034,
+      "step": 3186
+    },
+    {
+      "epoch": 1.2731629392971247,
+      "grad_norm": 1.1631071186812576,
+      "learning_rate": 7.123849102682772e-06,
+      "loss": 0.1104,
+      "step": 3188
+    },
+    {
+      "epoch": 1.273961661341853,
+      "grad_norm": 0.9901243870652381,
+      "learning_rate": 7.119640954845011e-06,
+      "loss": 0.0994,
+      "step": 3190
+    },
+    {
+      "epoch": 1.2747603833865815,
+      "grad_norm": 1.085814944090195,
+      "learning_rate": 7.115430975837457e-06,
+      "loss": 0.1105,
+      "step": 3192
+    },
+    {
+      "epoch": 1.2755591054313098,
+      "grad_norm": 1.2215853050594954,
+      "learning_rate": 7.111219169297134e-06,
+      "loss": 0.1054,
+      "step": 3194
+    },
+    {
+      "epoch": 1.2763578274760383,
+      "grad_norm": 1.062258920330732,
+      "learning_rate": 7.107005538862647e-06,
+      "loss": 0.1087,
+      "step": 3196
+    },
+    {
+      "epoch": 1.2771565495207668,
+      "grad_norm": 1.143873543200049,
+      "learning_rate": 7.102790088174172e-06,
+      "loss": 0.1098,
+      "step": 3198
+    },
+    {
+      "epoch": 1.2779552715654952,
+      "grad_norm": 0.8890153264482474,
+      "learning_rate": 7.098572820873461e-06,
+      "loss": 0.096,
+      "step": 3200
+    },
+    {
+      "epoch": 1.2787539936102237,
+      "grad_norm": 1.0763979750774153,
+      "learning_rate": 7.0943537406038385e-06,
+      "loss": 0.104,
+      "step": 3202
+    },
+    {
+      "epoch": 1.279552715654952,
+      "grad_norm": 1.005382468063005,
+      "learning_rate": 7.09013285101019e-06,
+      "loss": 0.0956,
+      "step": 3204
+    },
+    {
+      "epoch": 1.2803514376996805,
+      "grad_norm": 1.0144396802490765,
+      "learning_rate": 7.085910155738964e-06,
+      "loss": 0.0964,
+      "step": 3206
+    },
+    {
+      "epoch": 1.281150159744409,
+      "grad_norm": 1.1536744431519097,
+      "learning_rate": 7.081685658438173e-06,
+      "loss": 0.127,
+      "step": 3208
+    },
+    {
+      "epoch": 1.2819488817891374,
+      "grad_norm": 1.1286638022231128,
+      "learning_rate": 7.0774593627573815e-06,
+      "loss": 0.1083,
+      "step": 3210
+    },
+    {
+      "epoch": 1.2827476038338659,
+      "grad_norm": 1.0529407867194136,
+      "learning_rate": 7.073231272347714e-06,
+      "loss": 0.1067,
+      "step": 3212
+    },
+    {
+      "epoch": 1.2835463258785942,
+      "grad_norm": 1.0573840208402359,
+      "learning_rate": 7.069001390861838e-06,
+      "loss": 0.0897,
+      "step": 3214
+    },
+    {
+      "epoch": 1.2843450479233227,
+      "grad_norm": 1.1598664588481753,
+      "learning_rate": 7.064769721953975e-06,
+      "loss": 0.1103,
+      "step": 3216
+    },
+    {
+      "epoch": 1.2851437699680512,
+      "grad_norm": 1.2632526669395345,
+      "learning_rate": 7.060536269279887e-06,
+      "loss": 0.119,
+      "step": 3218
+    },
+    {
+      "epoch": 1.2859424920127795,
+      "grad_norm": 1.0466455925316702,
+      "learning_rate": 7.056301036496875e-06,
+      "loss": 0.0984,
+      "step": 3220
+    },
+    {
+      "epoch": 1.2867412140575079,
+      "grad_norm": 1.0228282332371574,
+      "learning_rate": 7.052064027263785e-06,
+      "loss": 0.0896,
+      "step": 3222
+    },
+    {
+      "epoch": 1.2875399361022364,
+      "grad_norm": 1.1687594953534266,
+      "learning_rate": 7.047825245240989e-06,
+      "loss": 0.1064,
+      "step": 3224
+    },
+    {
+      "epoch": 1.288338658146965,
+      "grad_norm": 1.1265735374171337,
+      "learning_rate": 7.0435846940903974e-06,
+      "loss": 0.1104,
+      "step": 3226
+    },
+    {
+      "epoch": 1.2891373801916932,
+      "grad_norm": 1.070111795962378,
+      "learning_rate": 7.039342377475444e-06,
+      "loss": 0.1072,
+      "step": 3228
+    },
+    {
+      "epoch": 1.2899361022364217,
+      "grad_norm": 1.2165890989586463,
+      "learning_rate": 7.035098299061094e-06,
+      "loss": 0.1192,
+      "step": 3230
+    },
+    {
+      "epoch": 1.29073482428115,
+      "grad_norm": 1.1082530581997079,
+      "learning_rate": 7.030852462513827e-06,
+      "loss": 0.1065,
+      "step": 3232
+    },
+    {
+      "epoch": 1.2915335463258786,
+      "grad_norm": 1.124650039405341,
+      "learning_rate": 7.026604871501647e-06,
+      "loss": 0.1045,
+      "step": 3234
+    },
+    {
+      "epoch": 1.292332268370607,
+      "grad_norm": 1.2186986975548504,
+      "learning_rate": 7.02235552969407e-06,
+      "loss": 0.1142,
+      "step": 3236
+    },
+    {
+      "epoch": 1.2931309904153354,
+      "grad_norm": 0.991464353367416,
+      "learning_rate": 7.018104440762128e-06,
+      "loss": 0.0919,
+      "step": 3238
+    },
+    {
+      "epoch": 1.293929712460064,
+      "grad_norm": 1.020705207223098,
+      "learning_rate": 7.013851608378359e-06,
+      "loss": 0.0978,
+      "step": 3240
+    },
+    {
+      "epoch": 1.2947284345047922,
+      "grad_norm": 1.0373213189747952,
+      "learning_rate": 7.009597036216813e-06,
+      "loss": 0.1088,
+      "step": 3242
+    },
+    {
+      "epoch": 1.2955271565495208,
+      "grad_norm": 0.9464480336921676,
+      "learning_rate": 7.005340727953035e-06,
+      "loss": 0.0953,
+      "step": 3244
+    },
+    {
+      "epoch": 1.2963258785942493,
+      "grad_norm": 1.0939229933892893,
+      "learning_rate": 7.001082687264075e-06,
+      "loss": 0.1079,
+      "step": 3246
+    },
+    {
+      "epoch": 1.2971246006389776,
+      "grad_norm": 1.2560358384718262,
+      "learning_rate": 6.9968229178284775e-06,
+      "loss": 0.1165,
+      "step": 3248
+    },
+    {
+      "epoch": 1.2979233226837061,
+      "grad_norm": 1.2448345478431648,
+      "learning_rate": 6.992561423326284e-06,
+      "loss": 0.1002,
+      "step": 3250
+    },
+    {
+      "epoch": 1.2987220447284344,
+      "grad_norm": 1.1712943475606261,
+      "learning_rate": 6.988298207439022e-06,
+      "loss": 0.0989,
+      "step": 3252
+    },
+    {
+      "epoch": 1.299520766773163,
+      "grad_norm": 1.1773991585755725,
+      "learning_rate": 6.9840332738497065e-06,
+      "loss": 0.1133,
+      "step": 3254
+    },
+    {
+      "epoch": 1.3003194888178915,
+      "grad_norm": 1.1116582831985906,
+      "learning_rate": 6.979766626242839e-06,
+      "loss": 0.1142,
+      "step": 3256
+    },
+    {
+      "epoch": 1.3011182108626198,
+      "grad_norm": 1.1586326246366871,
+      "learning_rate": 6.975498268304401e-06,
+      "loss": 0.1125,
+      "step": 3258
+    },
+    {
+      "epoch": 1.3019169329073481,
+      "grad_norm": 1.0830604429290178,
+      "learning_rate": 6.971228203721849e-06,
+      "loss": 0.1086,
+      "step": 3260
+    },
+    {
+      "epoch": 1.3027156549520766,
+      "grad_norm": 1.1206997334831217,
+      "learning_rate": 6.96695643618412e-06,
+      "loss": 0.1124,
+      "step": 3262
+    },
+    {
+      "epoch": 1.3035143769968052,
+      "grad_norm": 0.967514162516545,
+      "learning_rate": 6.9626829693816135e-06,
+      "loss": 0.1125,
+      "step": 3264
+    },
+    {
+      "epoch": 1.3043130990415335,
+      "grad_norm": 0.97337749177029,
+      "learning_rate": 6.958407807006205e-06,
+      "loss": 0.1037,
+      "step": 3266
+    },
+    {
+      "epoch": 1.305111821086262,
+      "grad_norm": 1.1764350961941878,
+      "learning_rate": 6.954130952751228e-06,
+      "loss": 0.1131,
+      "step": 3268
+    },
+    {
+      "epoch": 1.3059105431309903,
+      "grad_norm": 0.9892258025188561,
+      "learning_rate": 6.949852410311484e-06,
+      "loss": 0.1034,
+      "step": 3270
+    },
+    {
+      "epoch": 1.3067092651757188,
+      "grad_norm": 1.0360171336900805,
+      "learning_rate": 6.945572183383229e-06,
+      "loss": 0.1067,
+      "step": 3272
+    },
+    {
+      "epoch": 1.3075079872204474,
+      "grad_norm": 1.2410869037730936,
+      "learning_rate": 6.941290275664175e-06,
+      "loss": 0.1053,
+      "step": 3274
+    },
+    {
+      "epoch": 1.3083067092651757,
+      "grad_norm": 1.0698597118994493,
+      "learning_rate": 6.9370066908534875e-06,
+      "loss": 0.1002,
+      "step": 3276
+    },
+    {
+      "epoch": 1.3091054313099042,
+      "grad_norm": 1.0720761674526624,
+      "learning_rate": 6.932721432651779e-06,
+      "loss": 0.1093,
+      "step": 3278
+    },
+    {
+      "epoch": 1.3099041533546325,
+      "grad_norm": 1.1243594373271324,
+      "learning_rate": 6.928434504761106e-06,
+      "loss": 0.1219,
+      "step": 3280
+    },
+    {
+      "epoch": 1.310702875399361,
+      "grad_norm": 1.1859750342924074,
+      "learning_rate": 6.924145910884972e-06,
+      "loss": 0.109,
+      "step": 3282
+    },
+    {
+      "epoch": 1.3115015974440896,
+      "grad_norm": 1.0930001670477891,
+      "learning_rate": 6.919855654728317e-06,
+      "loss": 0.108,
+      "step": 3284
+    },
+    {
+      "epoch": 1.3123003194888179,
+      "grad_norm": 0.9985564810729535,
+      "learning_rate": 6.9155637399975196e-06,
+      "loss": 0.0973,
+      "step": 3286
+    },
+    {
+      "epoch": 1.3130990415335464,
+      "grad_norm": 1.056704078140189,
+      "learning_rate": 6.911270170400385e-06,
+      "loss": 0.0985,
+      "step": 3288
+    },
+    {
+      "epoch": 1.3138977635782747,
+      "grad_norm": 1.2319666270989271,
+      "learning_rate": 6.9069749496461555e-06,
+      "loss": 0.112,
+      "step": 3290
+    },
+    {
+      "epoch": 1.3146964856230032,
+      "grad_norm": 1.1376694239915568,
+      "learning_rate": 6.902678081445495e-06,
+      "loss": 0.1048,
+      "step": 3292
+    },
+    {
+      "epoch": 1.3154952076677318,
+      "grad_norm": 1.169749984507156,
+      "learning_rate": 6.898379569510491e-06,
+      "loss": 0.1021,
+      "step": 3294
+    },
+    {
+      "epoch": 1.31629392971246,
+      "grad_norm": 1.223585035436731,
+      "learning_rate": 6.894079417554657e-06,
+      "loss": 0.1194,
+      "step": 3296
+    },
+    {
+      "epoch": 1.3170926517571884,
+      "grad_norm": 1.19066097682282,
+      "learning_rate": 6.889777629292914e-06,
+      "loss": 0.1184,
+      "step": 3298
+    },
+    {
+      "epoch": 1.317891373801917,
+      "grad_norm": 1.0965120097047536,
+      "learning_rate": 6.885474208441602e-06,
+      "loss": 0.0979,
+      "step": 3300
+    },
+    {
+      "epoch": 1.3186900958466454,
+      "grad_norm": 1.0211040534374387,
+      "learning_rate": 6.881169158718474e-06,
+      "loss": 0.1081,
+      "step": 3302
+    },
+    {
+      "epoch": 1.3194888178913737,
+      "grad_norm": 1.017263749768144,
+      "learning_rate": 6.8768624838426815e-06,
+      "loss": 0.0988,
+      "step": 3304
+    },
+    {
+      "epoch": 1.3202875399361023,
+      "grad_norm": 1.0524568085425485,
+      "learning_rate": 6.872554187534788e-06,
+      "loss": 0.1032,
+      "step": 3306
+    },
+    {
+      "epoch": 1.3210862619808306,
+      "grad_norm": 1.0173821412579143,
+      "learning_rate": 6.868244273516755e-06,
+      "loss": 0.1045,
+      "step": 3308
+    },
+    {
+      "epoch": 1.321884984025559,
+      "grad_norm": 1.092162092574483,
+      "learning_rate": 6.863932745511942e-06,
+      "loss": 0.1127,
+      "step": 3310
+    },
+    {
+      "epoch": 1.3226837060702876,
+      "grad_norm": 1.1738686449026956,
+      "learning_rate": 6.859619607245102e-06,
+      "loss": 0.1048,
+      "step": 3312
+    },
+    {
+      "epoch": 1.323482428115016,
+      "grad_norm": 1.1698036650478054,
+      "learning_rate": 6.855304862442379e-06,
+      "loss": 0.109,
+      "step": 3314
+    },
+    {
+      "epoch": 1.3242811501597445,
+      "grad_norm": 1.0611253385279913,
+      "learning_rate": 6.850988514831304e-06,
+      "loss": 0.1036,
+      "step": 3316
+    },
+    {
+      "epoch": 1.3250798722044728,
+      "grad_norm": 1.180905968110651,
+      "learning_rate": 6.846670568140797e-06,
+      "loss": 0.1082,
+      "step": 3318
+    },
+    {
+      "epoch": 1.3258785942492013,
+      "grad_norm": 1.054233295900519,
+      "learning_rate": 6.842351026101155e-06,
+      "loss": 0.1059,
+      "step": 3320
+    },
+    {
+      "epoch": 1.3266773162939298,
+      "grad_norm": 1.080658997418189,
+      "learning_rate": 6.838029892444056e-06,
+      "loss": 0.1072,
+      "step": 3322
+    },
+    {
+      "epoch": 1.3274760383386581,
+      "grad_norm": 1.0388885899845177,
+      "learning_rate": 6.833707170902551e-06,
+      "loss": 0.1042,
+      "step": 3324
+    },
+    {
+      "epoch": 1.3282747603833867,
+      "grad_norm": 1.0681215628181655,
+      "learning_rate": 6.829382865211063e-06,
+      "loss": 0.1012,
+      "step": 3326
+    },
+    {
+      "epoch": 1.329073482428115,
+      "grad_norm": 0.9880264033407431,
+      "learning_rate": 6.825056979105382e-06,
+      "loss": 0.1094,
+      "step": 3328
+    },
+    {
+      "epoch": 1.3298722044728435,
+      "grad_norm": 1.0627754491783672,
+      "learning_rate": 6.820729516322671e-06,
+      "loss": 0.1045,
+      "step": 3330
+    },
+    {
+      "epoch": 1.330670926517572,
+      "grad_norm": 0.9621593864476723,
+      "learning_rate": 6.816400480601445e-06,
+      "loss": 0.0988,
+      "step": 3332
+    },
+    {
+      "epoch": 1.3314696485623003,
+      "grad_norm": 1.0320888666137535,
+      "learning_rate": 6.812069875681585e-06,
+      "loss": 0.0948,
+      "step": 3334
+    },
+    {
+      "epoch": 1.3322683706070286,
+      "grad_norm": 1.0860752731846834,
+      "learning_rate": 6.807737705304324e-06,
+      "loss": 0.0979,
+      "step": 3336
+    },
+    {
+      "epoch": 1.3330670926517572,
+      "grad_norm": 1.1173806530253718,
+      "learning_rate": 6.803403973212247e-06,
+      "loss": 0.1084,
+      "step": 3338
+    },
+    {
+      "epoch": 1.3338658146964857,
+      "grad_norm": 1.211178041971597,
+      "learning_rate": 6.799068683149291e-06,
+      "loss": 0.1026,
+      "step": 3340
+    },
+    {
+      "epoch": 1.334664536741214,
+      "grad_norm": 1.324084252679483,
+      "learning_rate": 6.79473183886074e-06,
+      "loss": 0.106,
+      "step": 3342
+    },
+    {
+      "epoch": 1.3354632587859425,
+      "grad_norm": 1.190661411969892,
+      "learning_rate": 6.790393444093214e-06,
+      "loss": 0.1103,
+      "step": 3344
+    },
+    {
+      "epoch": 1.3362619808306708,
+      "grad_norm": 1.0769935421724353,
+      "learning_rate": 6.786053502594679e-06,
+      "loss": 0.0949,
+      "step": 3346
+    },
+    {
+      "epoch": 1.3370607028753994,
+      "grad_norm": 1.0884920315893003,
+      "learning_rate": 6.781712018114435e-06,
+      "loss": 0.1041,
+      "step": 3348
+    },
+    {
+      "epoch": 1.3378594249201279,
+      "grad_norm": 1.451297567535448,
+      "learning_rate": 6.777368994403113e-06,
+      "loss": 0.1101,
+      "step": 3350
+    },
+    {
+      "epoch": 1.3386581469648562,
+      "grad_norm": 1.0225728615227607,
+      "learning_rate": 6.773024435212678e-06,
+      "loss": 0.0971,
+      "step": 3352
+    },
+    {
+      "epoch": 1.3394568690095847,
+      "grad_norm": 1.0713734854451504,
+      "learning_rate": 6.7686783442964195e-06,
+      "loss": 0.1011,
+      "step": 3354
+    },
+    {
+      "epoch": 1.340255591054313,
+      "grad_norm": 1.0477177218409457,
+      "learning_rate": 6.7643307254089485e-06,
+      "loss": 0.107,
+      "step": 3356
+    },
+    {
+      "epoch": 1.3410543130990416,
+      "grad_norm": 1.0932469970665097,
+      "learning_rate": 6.759981582306197e-06,
+      "loss": 0.1081,
+      "step": 3358
+    },
+    {
+      "epoch": 1.34185303514377,
+      "grad_norm": 1.1498681000315574,
+      "learning_rate": 6.7556309187454185e-06,
+      "loss": 0.1036,
+      "step": 3360
+    },
+    {
+      "epoch": 1.3426517571884984,
+      "grad_norm": 1.1306486794158652,
+      "learning_rate": 6.751278738485169e-06,
+      "loss": 0.117,
+      "step": 3362
+    },
+    {
+      "epoch": 1.343450479233227,
+      "grad_norm": 1.0798075050613067,
+      "learning_rate": 6.746925045285327e-06,
+      "loss": 0.1075,
+      "step": 3364
+    },
+    {
+      "epoch": 1.3442492012779552,
+      "grad_norm": 1.1185887548421796,
+      "learning_rate": 6.742569842907071e-06,
+      "loss": 0.1177,
+      "step": 3366
+    },
+    {
+      "epoch": 1.3450479233226837,
+      "grad_norm": 1.185572779424059,
+      "learning_rate": 6.738213135112884e-06,
+      "loss": 0.0999,
+      "step": 3368
+    },
+    {
+      "epoch": 1.3458466453674123,
+      "grad_norm": 1.113143717088373,
+      "learning_rate": 6.733854925666552e-06,
+      "loss": 0.111,
+      "step": 3370
+    },
+    {
+      "epoch": 1.3466453674121406,
+      "grad_norm": 1.0888946989797392,
+      "learning_rate": 6.729495218333157e-06,
+      "loss": 0.0965,
+      "step": 3372
+    },
+    {
+      "epoch": 1.3474440894568689,
+      "grad_norm": 1.0800532165447028,
+      "learning_rate": 6.725134016879071e-06,
+      "loss": 0.1294,
+      "step": 3374
+    },
+    {
+      "epoch": 1.3482428115015974,
+      "grad_norm": 1.0293169314360555,
+      "learning_rate": 6.720771325071965e-06,
+      "loss": 0.1005,
+      "step": 3376
+    },
+    {
+      "epoch": 1.349041533546326,
+      "grad_norm": 1.1266946956354467,
+      "learning_rate": 6.716407146680793e-06,
+      "loss": 0.1034,
+      "step": 3378
+    },
+    {
+      "epoch": 1.3498402555910542,
+      "grad_norm": 1.0175557765865235,
+      "learning_rate": 6.71204148547579e-06,
+      "loss": 0.1032,
+      "step": 3380
+    },
+    {
+      "epoch": 1.3506389776357828,
+      "grad_norm": 1.129124738760774,
+      "learning_rate": 6.7076743452284776e-06,
+      "loss": 0.1155,
+      "step": 3382
+    },
+    {
+      "epoch": 1.351437699680511,
+      "grad_norm": 1.1223728907114148,
+      "learning_rate": 6.703305729711653e-06,
+      "loss": 0.1065,
+      "step": 3384
+    },
+    {
+      "epoch": 1.3522364217252396,
+      "grad_norm": 1.0545042346249254,
+      "learning_rate": 6.698935642699386e-06,
+      "loss": 0.0987,
+      "step": 3386
+    },
+    {
+      "epoch": 1.3530351437699681,
+      "grad_norm": 1.0675803055716808,
+      "learning_rate": 6.694564087967023e-06,
+      "loss": 0.1139,
+      "step": 3388
+    },
+    {
+      "epoch": 1.3538338658146964,
+      "grad_norm": 1.173914284307619,
+      "learning_rate": 6.6901910692911706e-06,
+      "loss": 0.1187,
+      "step": 3390
+    },
+    {
+      "epoch": 1.354632587859425,
+      "grad_norm": 1.0709511536838934,
+      "learning_rate": 6.685816590449708e-06,
+      "loss": 0.1068,
+      "step": 3392
+    },
+    {
+      "epoch": 1.3554313099041533,
+      "grad_norm": 0.9466690009096309,
+      "learning_rate": 6.68144065522177e-06,
+      "loss": 0.0983,
+      "step": 3394
+    },
+    {
+      "epoch": 1.3562300319488818,
+      "grad_norm": 1.114009958699587,
+      "learning_rate": 6.677063267387754e-06,
+      "loss": 0.1099,
+      "step": 3396
+    },
+    {
+      "epoch": 1.3570287539936103,
+      "grad_norm": 1.1077866780843397,
+      "learning_rate": 6.672684430729305e-06,
+      "loss": 0.1036,
+      "step": 3398
+    },
+    {
+      "epoch": 1.3578274760383386,
+      "grad_norm": 1.0412230391242072,
+      "learning_rate": 6.668304149029331e-06,
+      "loss": 0.1001,
+      "step": 3400
+    },
+    {
+      "epoch": 1.3586261980830672,
+      "grad_norm": 1.1010636435498742,
+      "learning_rate": 6.663922426071978e-06,
+      "loss": 0.1045,
+      "step": 3402
+    },
+    {
+      "epoch": 1.3594249201277955,
+      "grad_norm": 1.1984509482714603,
+      "learning_rate": 6.659539265642643e-06,
+      "loss": 0.106,
+      "step": 3404
+    },
+    {
+      "epoch": 1.360223642172524,
+      "grad_norm": 0.9863537625149396,
+      "learning_rate": 6.655154671527962e-06,
+      "loss": 0.1033,
+      "step": 3406
+    },
+    {
+      "epoch": 1.3610223642172525,
+      "grad_norm": 1.0898694676194725,
+      "learning_rate": 6.650768647515813e-06,
+      "loss": 0.1068,
+      "step": 3408
+    },
+    {
+      "epoch": 1.3618210862619808,
+      "grad_norm": 1.0711686739516757,
+      "learning_rate": 6.646381197395302e-06,
+      "loss": 0.1094,
+      "step": 3410
+    },
+    {
+      "epoch": 1.3626198083067091,
+      "grad_norm": 0.9804227369687718,
+      "learning_rate": 6.641992324956776e-06,
+      "loss": 0.1105,
+      "step": 3412
+    },
+    {
+      "epoch": 1.3634185303514377,
+      "grad_norm": 1.1535558172179992,
+      "learning_rate": 6.637602033991807e-06,
+      "loss": 0.1053,
+      "step": 3414
+    },
+    {
+      "epoch": 1.3642172523961662,
+      "grad_norm": 1.0241900868101663,
+      "learning_rate": 6.63321032829319e-06,
+      "loss": 0.1094,
+      "step": 3416
+    },
+    {
+      "epoch": 1.3650159744408945,
+      "grad_norm": 1.1116602107856417,
+      "learning_rate": 6.628817211654945e-06,
+      "loss": 0.1079,
+      "step": 3418
+    },
+    {
+      "epoch": 1.365814696485623,
+      "grad_norm": 1.099401650328653,
+      "learning_rate": 6.624422687872312e-06,
+      "loss": 0.1147,
+      "step": 3420
+    },
+    {
+      "epoch": 1.3666134185303513,
+      "grad_norm": 1.1672020742722258,
+      "learning_rate": 6.6200267607417415e-06,
+      "loss": 0.1112,
+      "step": 3422
+    },
+    {
+      "epoch": 1.3674121405750799,
+      "grad_norm": 1.0307138522820616,
+      "learning_rate": 6.615629434060903e-06,
+      "loss": 0.116,
+      "step": 3424
+    },
+    {
+      "epoch": 1.3682108626198084,
+      "grad_norm": 1.2489824104367433,
+      "learning_rate": 6.611230711628669e-06,
+      "loss": 0.0957,
+      "step": 3426
+    },
+    {
+      "epoch": 1.3690095846645367,
+      "grad_norm": 0.9719925332909197,
+      "learning_rate": 6.6068305972451245e-06,
+      "loss": 0.106,
+      "step": 3428
+    },
+    {
+      "epoch": 1.3698083067092652,
+      "grad_norm": 1.058537905484153,
+      "learning_rate": 6.602429094711549e-06,
+      "loss": 0.1027,
+      "step": 3430
+    },
+    {
+      "epoch": 1.3706070287539935,
+      "grad_norm": 1.0235718542609205,
+      "learning_rate": 6.598026207830428e-06,
+      "loss": 0.0964,
+      "step": 3432
+    },
+    {
+      "epoch": 1.371405750798722,
+      "grad_norm": 1.0786959847501962,
+      "learning_rate": 6.593621940405439e-06,
+      "loss": 0.1014,
+      "step": 3434
+    },
+    {
+      "epoch": 1.3722044728434506,
+      "grad_norm": 1.137212886630325,
+      "learning_rate": 6.589216296241455e-06,
+      "loss": 0.1168,
+      "step": 3436
+    },
+    {
+      "epoch": 1.373003194888179,
+      "grad_norm": 1.0030513413696003,
+      "learning_rate": 6.584809279144535e-06,
+      "loss": 0.0999,
+      "step": 3438
+    },
+    {
+      "epoch": 1.3738019169329074,
+      "grad_norm": 1.1517827771669715,
+      "learning_rate": 6.5804008929219284e-06,
+      "loss": 0.1177,
+      "step": 3440
+    },
+    {
+      "epoch": 1.3746006389776357,
+      "grad_norm": 1.1512991234134653,
+      "learning_rate": 6.575991141382063e-06,
+      "loss": 0.0966,
+      "step": 3442
+    },
+    {
+      "epoch": 1.3753993610223643,
+      "grad_norm": 1.1000379035510455,
+      "learning_rate": 6.571580028334547e-06,
+      "loss": 0.1053,
+      "step": 3444
+    },
+    {
+      "epoch": 1.3761980830670926,
+      "grad_norm": 1.0218098747432192,
+      "learning_rate": 6.56716755759017e-06,
+      "loss": 0.0912,
+      "step": 3446
+    },
+    {
+      "epoch": 1.376996805111821,
+      "grad_norm": 1.1551745154641282,
+      "learning_rate": 6.562753732960887e-06,
+      "loss": 0.1119,
+      "step": 3448
+    },
+    {
+      "epoch": 1.3777955271565494,
+      "grad_norm": 1.2194490422535142,
+      "learning_rate": 6.5583385582598255e-06,
+      "loss": 0.099,
+      "step": 3450
+    },
+    {
+      "epoch": 1.378594249201278,
+      "grad_norm": 1.4172435672608823,
+      "learning_rate": 6.553922037301283e-06,
+      "loss": 0.0978,
+      "step": 3452
+    },
+    {
+      "epoch": 1.3793929712460065,
+      "grad_norm": 1.0695393339059773,
+      "learning_rate": 6.549504173900715e-06,
+      "loss": 0.0917,
+      "step": 3454
+    },
+    {
+      "epoch": 1.3801916932907348,
+      "grad_norm": 1.196402708368458,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.1113,
+      "step": 3456
+    },
+    {
+      "epoch": 1.3809904153354633,
+      "grad_norm": 1.0433584799967095,
+      "learning_rate": 6.540664435041127e-06,
+      "loss": 0.104,
+      "step": 3458
+    },
+    {
+      "epoch": 1.3817891373801916,
+      "grad_norm": 1.149354927046393,
+      "learning_rate": 6.536242567218808e-06,
+      "loss": 0.1031,
+      "step": 3460
+    },
+    {
+      "epoch": 1.3825878594249201,
+      "grad_norm": 1.1148556785558263,
+      "learning_rate": 6.531819372227856e-06,
+      "loss": 0.1032,
+      "step": 3462
+    },
+    {
+      "epoch": 1.3833865814696487,
+      "grad_norm": 1.071711948179689,
+      "learning_rate": 6.527394853889499e-06,
+      "loss": 0.1085,
+      "step": 3464
+    },
+    {
+      "epoch": 1.384185303514377,
+      "grad_norm": 1.082175681158764,
+      "learning_rate": 6.522969016026099e-06,
+      "loss": 0.11,
+      "step": 3466
+    },
+    {
+      "epoch": 1.3849840255591055,
+      "grad_norm": 1.1369109201663514,
+      "learning_rate": 6.518541862461163e-06,
+      "loss": 0.1169,
+      "step": 3468
+    },
+    {
+      "epoch": 1.3857827476038338,
+      "grad_norm": 1.2011247923954287,
+      "learning_rate": 6.514113397019335e-06,
+      "loss": 0.1081,
+      "step": 3470
+    },
+    {
+      "epoch": 1.3865814696485623,
+      "grad_norm": 1.1792397645094768,
+      "learning_rate": 6.5096836235263904e-06,
+      "loss": 0.1202,
+      "step": 3472
+    },
+    {
+      "epoch": 1.3873801916932909,
+      "grad_norm": 1.0727123114625765,
+      "learning_rate": 6.505252545809238e-06,
+      "loss": 0.0962,
+      "step": 3474
+    },
+    {
+      "epoch": 1.3881789137380192,
+      "grad_norm": 1.0276014422922428,
+      "learning_rate": 6.500820167695906e-06,
+      "loss": 0.0996,
+      "step": 3476
+    },
+    {
+      "epoch": 1.3889776357827475,
+      "grad_norm": 0.9724065635327491,
+      "learning_rate": 6.496386493015554e-06,
+      "loss": 0.0987,
+      "step": 3478
+    },
+    {
+      "epoch": 1.389776357827476,
+      "grad_norm": 1.1292315012262628,
+      "learning_rate": 6.491951525598461e-06,
+      "loss": 0.0999,
+      "step": 3480
+    },
+    {
+      "epoch": 1.3905750798722045,
+      "grad_norm": 1.1716250389490488,
+      "learning_rate": 6.487515269276015e-06,
+      "loss": 0.1015,
+      "step": 3482
+    },
+    {
+      "epoch": 1.3913738019169328,
+      "grad_norm": 1.036997684705096,
+      "learning_rate": 6.483077727880726e-06,
+      "loss": 0.0863,
+      "step": 3484
+    },
+    {
+      "epoch": 1.3921725239616614,
+      "grad_norm": 1.199757124869139,
+      "learning_rate": 6.478638905246213e-06,
+      "loss": 0.1116,
+      "step": 3486
+    },
+    {
+      "epoch": 1.3929712460063897,
+      "grad_norm": 1.142726930599777,
+      "learning_rate": 6.4741988052071965e-06,
+      "loss": 0.1098,
+      "step": 3488
+    },
+    {
+      "epoch": 1.3937699680511182,
+      "grad_norm": 0.9651746600463053,
+      "learning_rate": 6.469757431599503e-06,
+      "loss": 0.1038,
+      "step": 3490
+    },
+    {
+      "epoch": 1.3945686900958467,
+      "grad_norm": 1.1474675700025958,
+      "learning_rate": 6.465314788260067e-06,
+      "loss": 0.1154,
+      "step": 3492
+    },
+    {
+      "epoch": 1.395367412140575,
+      "grad_norm": 1.0346444553697798,
+      "learning_rate": 6.460870879026906e-06,
+      "loss": 0.0961,
+      "step": 3494
+    },
+    {
+      "epoch": 1.3961661341853036,
+      "grad_norm": 1.032379620946616,
+      "learning_rate": 6.45642570773914e-06,
+      "loss": 0.0997,
+      "step": 3496
+    },
+    {
+      "epoch": 1.3969648562300319,
+      "grad_norm": 1.016064340983906,
+      "learning_rate": 6.451979278236979e-06,
+      "loss": 0.0931,
+      "step": 3498
+    },
+    {
+      "epoch": 1.3977635782747604,
+      "grad_norm": 1.178554041440438,
+      "learning_rate": 6.447531594361719e-06,
+      "loss": 0.1153,
+      "step": 3500
+    },
+    {
+      "epoch": 1.3977635782747604,
+      "eval_loss": 0.1600114107131958,
+      "eval_runtime": 418.1465,
+      "eval_samples_per_second": 42.586,
+      "eval_steps_per_second": 5.323,
+      "step": 3500
+    },
+    {
+      "epoch": 1.398562300319489,
+      "grad_norm": 1.1584355046290076,
+      "learning_rate": 6.443082659955737e-06,
+      "loss": 0.1037,
+      "step": 3502
+    },
+    {
+      "epoch": 1.3993610223642172,
+      "grad_norm": 1.0340861375288486,
+      "learning_rate": 6.438632478862495e-06,
+      "loss": 0.1011,
+      "step": 3504
+    },
+    {
+      "epoch": 1.4001597444089458,
+      "grad_norm": 1.1875264702825228,
+      "learning_rate": 6.434181054926528e-06,
+      "loss": 0.1125,
+      "step": 3506
+    },
+    {
+      "epoch": 1.400958466453674,
+      "grad_norm": 1.0588069817372803,
+      "learning_rate": 6.429728391993446e-06,
+      "loss": 0.107,
+      "step": 3508
+    },
+    {
+      "epoch": 1.4017571884984026,
+      "grad_norm": 1.1098486981661322,
+      "learning_rate": 6.425274493909932e-06,
+      "loss": 0.1132,
+      "step": 3510
+    },
+    {
+      "epoch": 1.4025559105431311,
+      "grad_norm": 0.9823134309813667,
+      "learning_rate": 6.4208193645237314e-06,
+      "loss": 0.0941,
+      "step": 3512
+    },
+    {
+      "epoch": 1.4033546325878594,
+      "grad_norm": 1.0327788368934439,
+      "learning_rate": 6.416363007683656e-06,
+      "loss": 0.1105,
+      "step": 3514
+    },
+    {
+      "epoch": 1.4041533546325877,
+      "grad_norm": 1.1821001072193735,
+      "learning_rate": 6.411905427239577e-06,
+      "loss": 0.1084,
+      "step": 3516
+    },
+    {
+      "epoch": 1.4049520766773163,
+      "grad_norm": 1.1379784770115198,
+      "learning_rate": 6.407446627042426e-06,
+      "loss": 0.1135,
+      "step": 3518
+    },
+    {
+      "epoch": 1.4057507987220448,
+      "grad_norm": 1.1147819119237963,
+      "learning_rate": 6.402986610944183e-06,
+      "loss": 0.1076,
+      "step": 3520
+    },
+    {
+      "epoch": 1.406549520766773,
+      "grad_norm": 1.1200339092988951,
+      "learning_rate": 6.398525382797884e-06,
+      "loss": 0.1063,
+      "step": 3522
+    },
+    {
+      "epoch": 1.4073482428115016,
+      "grad_norm": 1.1312798096131238,
+      "learning_rate": 6.394062946457604e-06,
+      "loss": 0.1002,
+      "step": 3524
+    },
+    {
+      "epoch": 1.40814696485623,
+      "grad_norm": 1.199479655032073,
+      "learning_rate": 6.389599305778471e-06,
+      "loss": 0.108,
+      "step": 3526
+    },
+    {
+      "epoch": 1.4089456869009584,
+      "grad_norm": 1.218231109858983,
+      "learning_rate": 6.385134464616649e-06,
+      "loss": 0.1069,
+      "step": 3528
+    },
+    {
+      "epoch": 1.409744408945687,
+      "grad_norm": 1.1995058184118474,
+      "learning_rate": 6.38066842682934e-06,
+      "loss": 0.1103,
+      "step": 3530
+    },
+    {
+      "epoch": 1.4105431309904153,
+      "grad_norm": 1.1046534115214985,
+      "learning_rate": 6.376201196274778e-06,
+      "loss": 0.1142,
+      "step": 3532
+    },
+    {
+      "epoch": 1.4113418530351438,
+      "grad_norm": 1.0585983588946046,
+      "learning_rate": 6.37173277681223e-06,
+      "loss": 0.1114,
+      "step": 3534
+    },
+    {
+      "epoch": 1.4121405750798721,
+      "grad_norm": 1.1029224104867834,
+      "learning_rate": 6.367263172301985e-06,
+      "loss": 0.1113,
+      "step": 3536
+    },
+    {
+      "epoch": 1.4129392971246006,
+      "grad_norm": 1.1060776788940239,
+      "learning_rate": 6.3627923866053656e-06,
+      "loss": 0.0969,
+      "step": 3538
+    },
+    {
+      "epoch": 1.4137380191693292,
+      "grad_norm": 1.1032132052448507,
+      "learning_rate": 6.358320423584704e-06,
+      "loss": 0.0979,
+      "step": 3540
+    },
+    {
+      "epoch": 1.4145367412140575,
+      "grad_norm": 1.029503416268003,
+      "learning_rate": 6.353847287103356e-06,
+      "loss": 0.0953,
+      "step": 3542
+    },
+    {
+      "epoch": 1.415335463258786,
+      "grad_norm": 1.1352827365353346,
+      "learning_rate": 6.3493729810256895e-06,
+      "loss": 0.1118,
+      "step": 3544
+    },
+    {
+      "epoch": 1.4161341853035143,
+      "grad_norm": 1.119812994569396,
+      "learning_rate": 6.344897509217085e-06,
+      "loss": 0.1114,
+      "step": 3546
+    },
+    {
+      "epoch": 1.4169329073482428,
+      "grad_norm": 1.076940990174831,
+      "learning_rate": 6.340420875543922e-06,
+      "loss": 0.107,
+      "step": 3548
+    },
+    {
+      "epoch": 1.4177316293929714,
+      "grad_norm": 1.1517799889594393,
+      "learning_rate": 6.335943083873596e-06,
+      "loss": 0.0933,
+      "step": 3550
+    },
+    {
+      "epoch": 1.4185303514376997,
+      "grad_norm": 1.1617154802397773,
+      "learning_rate": 6.331464138074493e-06,
+      "loss": 0.105,
+      "step": 3552
+    },
+    {
+      "epoch": 1.419329073482428,
+      "grad_norm": 0.9494287593142456,
+      "learning_rate": 6.326984042016e-06,
+      "loss": 0.0921,
+      "step": 3554
+    },
+    {
+      "epoch": 1.4201277955271565,
+      "grad_norm": 1.1141280255887196,
+      "learning_rate": 6.322502799568498e-06,
+      "loss": 0.1061,
+      "step": 3556
+    },
+    {
+      "epoch": 1.420926517571885,
+      "grad_norm": 1.0414217088512951,
+      "learning_rate": 6.3180204146033586e-06,
+      "loss": 0.1029,
+      "step": 3558
+    },
+    {
+      "epoch": 1.4217252396166133,
+      "grad_norm": 1.0724759251125648,
+      "learning_rate": 6.313536890992935e-06,
+      "loss": 0.1053,
+      "step": 3560
+    },
+    {
+      "epoch": 1.4225239616613419,
+      "grad_norm": 1.0754047728283922,
+      "learning_rate": 6.309052232610574e-06,
+      "loss": 0.1096,
+      "step": 3562
+    },
+    {
+      "epoch": 1.4233226837060702,
+      "grad_norm": 1.1017909660001841,
+      "learning_rate": 6.3045664433305945e-06,
+      "loss": 0.1092,
+      "step": 3564
+    },
+    {
+      "epoch": 1.4241214057507987,
+      "grad_norm": 1.4387268439796592,
+      "learning_rate": 6.300079527028297e-06,
+      "loss": 0.1133,
+      "step": 3566
+    },
+    {
+      "epoch": 1.4249201277955272,
+      "grad_norm": 1.1303501334206185,
+      "learning_rate": 6.29559148757995e-06,
+      "loss": 0.1065,
+      "step": 3568
+    },
+    {
+      "epoch": 1.4257188498402555,
+      "grad_norm": 0.9640525041341275,
+      "learning_rate": 6.291102328862801e-06,
+      "loss": 0.0988,
+      "step": 3570
+    },
+    {
+      "epoch": 1.426517571884984,
+      "grad_norm": 1.0585869277703588,
+      "learning_rate": 6.286612054755056e-06,
+      "loss": 0.1022,
+      "step": 3572
+    },
+    {
+      "epoch": 1.4273162939297124,
+      "grad_norm": 1.0056038546914692,
+      "learning_rate": 6.282120669135892e-06,
+      "loss": 0.099,
+      "step": 3574
+    },
+    {
+      "epoch": 1.428115015974441,
+      "grad_norm": 1.1206641244335471,
+      "learning_rate": 6.277628175885437e-06,
+      "loss": 0.1167,
+      "step": 3576
+    },
+    {
+      "epoch": 1.4289137380191694,
+      "grad_norm": 1.1096458600950112,
+      "learning_rate": 6.273134578884785e-06,
+      "loss": 0.1058,
+      "step": 3578
+    },
+    {
+      "epoch": 1.4297124600638977,
+      "grad_norm": 0.9857461120602693,
+      "learning_rate": 6.2686398820159785e-06,
+      "loss": 0.0947,
+      "step": 3580
+    },
+    {
+      "epoch": 1.4305111821086263,
+      "grad_norm": 1.0203378985341014,
+      "learning_rate": 6.2641440891620146e-06,
+      "loss": 0.0914,
+      "step": 3582
+    },
+    {
+      "epoch": 1.4313099041533546,
+      "grad_norm": 1.0724408105973864,
+      "learning_rate": 6.2596472042068275e-06,
+      "loss": 0.0934,
+      "step": 3584
+    },
+    {
+      "epoch": 1.432108626198083,
+      "grad_norm": 1.2414438842066344,
+      "learning_rate": 6.2551492310353094e-06,
+      "loss": 0.1149,
+      "step": 3586
+    },
+    {
+      "epoch": 1.4329073482428116,
+      "grad_norm": 1.101684816522059,
+      "learning_rate": 6.250650173533279e-06,
+      "loss": 0.1142,
+      "step": 3588
+    },
+    {
+      "epoch": 1.43370607028754,
+      "grad_norm": 1.1753348478737677,
+      "learning_rate": 6.2461500355875e-06,
+      "loss": 0.116,
+      "step": 3590
+    },
+    {
+      "epoch": 1.4345047923322682,
+      "grad_norm": 1.1874416062991624,
+      "learning_rate": 6.241648821085666e-06,
+      "loss": 0.1073,
+      "step": 3592
+    },
+    {
+      "epoch": 1.4353035143769968,
+      "grad_norm": 1.0601160321918857,
+      "learning_rate": 6.237146533916402e-06,
+      "loss": 0.1013,
+      "step": 3594
+    },
+    {
+      "epoch": 1.4361022364217253,
+      "grad_norm": 1.108271170806747,
+      "learning_rate": 6.232643177969259e-06,
+      "loss": 0.0952,
+      "step": 3596
+    },
+    {
+      "epoch": 1.4369009584664536,
+      "grad_norm": 1.030322538054295,
+      "learning_rate": 6.2281387571347126e-06,
+      "loss": 0.1006,
+      "step": 3598
+    },
+    {
+      "epoch": 1.4376996805111821,
+      "grad_norm": 1.1028346810795997,
+      "learning_rate": 6.223633275304157e-06,
+      "loss": 0.111,
+      "step": 3600
+    },
+    {
+      "epoch": 1.4384984025559104,
+      "grad_norm": 1.0355367948281153,
+      "learning_rate": 6.2191267363699026e-06,
+      "loss": 0.1017,
+      "step": 3602
+    },
+    {
+      "epoch": 1.439297124600639,
+      "grad_norm": 1.2163948163518867,
+      "learning_rate": 6.214619144225176e-06,
+      "loss": 0.1077,
+      "step": 3604
+    },
+    {
+      "epoch": 1.4400958466453675,
+      "grad_norm": 1.1518703062495919,
+      "learning_rate": 6.210110502764107e-06,
+      "loss": 0.1155,
+      "step": 3606
+    },
+    {
+      "epoch": 1.4408945686900958,
+      "grad_norm": 1.1725592196005963,
+      "learning_rate": 6.205600815881741e-06,
+      "loss": 0.0979,
+      "step": 3608
+    },
+    {
+      "epoch": 1.4416932907348243,
+      "grad_norm": 1.0475879501811074,
+      "learning_rate": 6.2010900874740225e-06,
+      "loss": 0.106,
+      "step": 3610
+    },
+    {
+      "epoch": 1.4424920127795526,
+      "grad_norm": 1.07813779648937,
+      "learning_rate": 6.1965783214377895e-06,
+      "loss": 0.1022,
+      "step": 3612
+    },
+    {
+      "epoch": 1.4432907348242812,
+      "grad_norm": 1.0886593706957781,
+      "learning_rate": 6.192065521670787e-06,
+      "loss": 0.1051,
+      "step": 3614
+    },
+    {
+      "epoch": 1.4440894568690097,
+      "grad_norm": 1.1378082559578835,
+      "learning_rate": 6.187551692071648e-06,
+      "loss": 0.1084,
+      "step": 3616
+    },
+    {
+      "epoch": 1.444888178913738,
+      "grad_norm": 1.2011389042931888,
+      "learning_rate": 6.183036836539893e-06,
+      "loss": 0.1092,
+      "step": 3618
+    },
+    {
+      "epoch": 1.4456869009584665,
+      "grad_norm": 1.1251988986799606,
+      "learning_rate": 6.178520958975933e-06,
+      "loss": 0.1068,
+      "step": 3620
+    },
+    {
+      "epoch": 1.4464856230031948,
+      "grad_norm": 1.0150796615451205,
+      "learning_rate": 6.17400406328106e-06,
+      "loss": 0.1075,
+      "step": 3622
+    },
+    {
+      "epoch": 1.4472843450479234,
+      "grad_norm": 1.0885442213663745,
+      "learning_rate": 6.1694861533574445e-06,
+      "loss": 0.1127,
+      "step": 3624
+    },
+    {
+      "epoch": 1.4480830670926519,
+      "grad_norm": 1.064305780213124,
+      "learning_rate": 6.164967233108137e-06,
+      "loss": 0.0966,
+      "step": 3626
+    },
+    {
+      "epoch": 1.4488817891373802,
+      "grad_norm": 1.0844139564754607,
+      "learning_rate": 6.160447306437055e-06,
+      "loss": 0.1076,
+      "step": 3628
+    },
+    {
+      "epoch": 1.4496805111821085,
+      "grad_norm": 1.034026504245696,
+      "learning_rate": 6.1559263772489905e-06,
+      "loss": 0.096,
+      "step": 3630
+    },
+    {
+      "epoch": 1.450479233226837,
+      "grad_norm": 1.001104101700096,
+      "learning_rate": 6.1514044494496e-06,
+      "loss": 0.0994,
+      "step": 3632
+    },
+    {
+      "epoch": 1.4512779552715656,
+      "grad_norm": 1.1773822780221779,
+      "learning_rate": 6.146881526945401e-06,
+      "loss": 0.1076,
+      "step": 3634
+    },
+    {
+      "epoch": 1.4520766773162939,
+      "grad_norm": 1.0588404307223935,
+      "learning_rate": 6.142357613643773e-06,
+      "loss": 0.0929,
+      "step": 3636
+    },
+    {
+      "epoch": 1.4528753993610224,
+      "grad_norm": 1.2341260840634936,
+      "learning_rate": 6.13783271345295e-06,
+      "loss": 0.1156,
+      "step": 3638
+    },
+    {
+      "epoch": 1.4536741214057507,
+      "grad_norm": 1.2029429887737177,
+      "learning_rate": 6.133306830282021e-06,
+      "loss": 0.1131,
+      "step": 3640
+    },
+    {
+      "epoch": 1.4544728434504792,
+      "grad_norm": 0.9829158742595866,
+      "learning_rate": 6.128779968040917e-06,
+      "loss": 0.0977,
+      "step": 3642
+    },
+    {
+      "epoch": 1.4552715654952078,
+      "grad_norm": 1.0321776789284078,
+      "learning_rate": 6.1242521306404236e-06,
+      "loss": 0.0928,
+      "step": 3644
+    },
+    {
+      "epoch": 1.456070287539936,
+      "grad_norm": 1.0207766757489227,
+      "learning_rate": 6.119723321992164e-06,
+      "loss": 0.0956,
+      "step": 3646
+    },
+    {
+      "epoch": 1.4568690095846646,
+      "grad_norm": 1.2565358842354744,
+      "learning_rate": 6.115193546008602e-06,
+      "loss": 0.1075,
+      "step": 3648
+    },
+    {
+      "epoch": 1.457667731629393,
+      "grad_norm": 1.0795577230748994,
+      "learning_rate": 6.110662806603036e-06,
+      "loss": 0.1073,
+      "step": 3650
+    },
+    {
+      "epoch": 1.4584664536741214,
+      "grad_norm": 1.1116179183482724,
+      "learning_rate": 6.106131107689599e-06,
+      "loss": 0.1091,
+      "step": 3652
+    },
+    {
+      "epoch": 1.45926517571885,
+      "grad_norm": 1.1185062550976665,
+      "learning_rate": 6.101598453183248e-06,
+      "loss": 0.0981,
+      "step": 3654
+    },
+    {
+      "epoch": 1.4600638977635783,
+      "grad_norm": 1.1929060063575452,
+      "learning_rate": 6.097064846999774e-06,
+      "loss": 0.1114,
+      "step": 3656
+    },
+    {
+      "epoch": 1.4608626198083068,
+      "grad_norm": 1.1672964109989814,
+      "learning_rate": 6.09253029305578e-06,
+      "loss": 0.1048,
+      "step": 3658
+    },
+    {
+      "epoch": 1.461661341853035,
+      "grad_norm": 1.3459483840304294,
+      "learning_rate": 6.087994795268696e-06,
+      "loss": 0.1175,
+      "step": 3660
+    },
+    {
+      "epoch": 1.4624600638977636,
+      "grad_norm": 1.1453927282681513,
+      "learning_rate": 6.0834583575567606e-06,
+      "loss": 0.1024,
+      "step": 3662
+    },
+    {
+      "epoch": 1.4632587859424921,
+      "grad_norm": 1.0361766854922676,
+      "learning_rate": 6.078920983839032e-06,
+      "loss": 0.0971,
+      "step": 3664
+    },
+    {
+      "epoch": 1.4640575079872205,
+      "grad_norm": 1.0777649114365144,
+      "learning_rate": 6.07438267803537e-06,
+      "loss": 0.1028,
+      "step": 3666
+    },
+    {
+      "epoch": 1.4648562300319488,
+      "grad_norm": 1.1306654064542798,
+      "learning_rate": 6.069843444066444e-06,
+      "loss": 0.1095,
+      "step": 3668
+    },
+    {
+      "epoch": 1.4656549520766773,
+      "grad_norm": 1.0195281420397648,
+      "learning_rate": 6.065303285853724e-06,
+      "loss": 0.1075,
+      "step": 3670
+    },
+    {
+      "epoch": 1.4664536741214058,
+      "grad_norm": 1.023251007369703,
+      "learning_rate": 6.060762207319479e-06,
+      "loss": 0.1002,
+      "step": 3672
+    },
+    {
+      "epoch": 1.4672523961661341,
+      "grad_norm": 1.0920777216930633,
+      "learning_rate": 6.056220212386769e-06,
+      "loss": 0.1051,
+      "step": 3674
+    },
+    {
+      "epoch": 1.4680511182108626,
+      "grad_norm": 1.1039024614877324,
+      "learning_rate": 6.0516773049794545e-06,
+      "loss": 0.098,
+      "step": 3676
+    },
+    {
+      "epoch": 1.468849840255591,
+      "grad_norm": 0.9870962605725515,
+      "learning_rate": 6.0471334890221735e-06,
+      "loss": 0.1052,
+      "step": 3678
+    },
+    {
+      "epoch": 1.4696485623003195,
+      "grad_norm": 1.0307618080780037,
+      "learning_rate": 6.042588768440358e-06,
+      "loss": 0.0995,
+      "step": 3680
+    },
+    {
+      "epoch": 1.470447284345048,
+      "grad_norm": 1.0513694638233513,
+      "learning_rate": 6.038043147160215e-06,
+      "loss": 0.0954,
+      "step": 3682
+    },
+    {
+      "epoch": 1.4712460063897763,
+      "grad_norm": 1.0657839501710005,
+      "learning_rate": 6.033496629108736e-06,
+      "loss": 0.0995,
+      "step": 3684
+    },
+    {
+      "epoch": 1.4720447284345048,
+      "grad_norm": 1.1159795703718869,
+      "learning_rate": 6.02894921821368e-06,
+      "loss": 0.0966,
+      "step": 3686
+    },
+    {
+      "epoch": 1.4728434504792332,
+      "grad_norm": 0.9598839228144777,
+      "learning_rate": 6.024400918403581e-06,
+      "loss": 0.0968,
+      "step": 3688
+    },
+    {
+      "epoch": 1.4736421725239617,
+      "grad_norm": 1.0384010675372382,
+      "learning_rate": 6.019851733607744e-06,
+      "loss": 0.0955,
+      "step": 3690
+    },
+    {
+      "epoch": 1.4744408945686902,
+      "grad_norm": 1.1078140257609426,
+      "learning_rate": 6.015301667756234e-06,
+      "loss": 0.0942,
+      "step": 3692
+    },
+    {
+      "epoch": 1.4752396166134185,
+      "grad_norm": 1.2701615597150369,
+      "learning_rate": 6.0107507247798765e-06,
+      "loss": 0.0995,
+      "step": 3694
+    },
+    {
+      "epoch": 1.476038338658147,
+      "grad_norm": 1.2102770076457041,
+      "learning_rate": 6.006198908610261e-06,
+      "loss": 0.1202,
+      "step": 3696
+    },
+    {
+      "epoch": 1.4768370607028753,
+      "grad_norm": 1.145740181813383,
+      "learning_rate": 6.0016462231797225e-06,
+      "loss": 0.1117,
+      "step": 3698
+    },
+    {
+      "epoch": 1.4776357827476039,
+      "grad_norm": 1.0833428968981065,
+      "learning_rate": 5.997092672421356e-06,
+      "loss": 0.1037,
+      "step": 3700
+    },
+    {
+      "epoch": 1.4784345047923324,
+      "grad_norm": 1.0801478244071052,
+      "learning_rate": 5.9925382602689974e-06,
+      "loss": 0.1019,
+      "step": 3702
+    },
+    {
+      "epoch": 1.4792332268370607,
+      "grad_norm": 1.2134338336543735,
+      "learning_rate": 5.987982990657229e-06,
+      "loss": 0.1139,
+      "step": 3704
+    },
+    {
+      "epoch": 1.480031948881789,
+      "grad_norm": 1.0064298338790032,
+      "learning_rate": 5.9834268675213745e-06,
+      "loss": 0.1045,
+      "step": 3706
+    },
+    {
+      "epoch": 1.4808306709265175,
+      "grad_norm": 1.1699403671691384,
+      "learning_rate": 5.978869894797494e-06,
+      "loss": 0.1054,
+      "step": 3708
+    },
+    {
+      "epoch": 1.481629392971246,
+      "grad_norm": 0.9388532857646246,
+      "learning_rate": 5.974312076422381e-06,
+      "loss": 0.098,
+      "step": 3710
+    },
+    {
+      "epoch": 1.4824281150159744,
+      "grad_norm": 1.0345840078184048,
+      "learning_rate": 5.9697534163335645e-06,
+      "loss": 0.1077,
+      "step": 3712
+    },
+    {
+      "epoch": 1.483226837060703,
+      "grad_norm": 1.0824548650493346,
+      "learning_rate": 5.965193918469292e-06,
+      "loss": 0.099,
+      "step": 3714
+    },
+    {
+      "epoch": 1.4840255591054312,
+      "grad_norm": 1.1507346874731688,
+      "learning_rate": 5.9606335867685424e-06,
+      "loss": 0.1013,
+      "step": 3716
+    },
+    {
+      "epoch": 1.4848242811501597,
+      "grad_norm": 0.9628660977174563,
+      "learning_rate": 5.9560724251710116e-06,
+      "loss": 0.0943,
+      "step": 3718
+    },
+    {
+      "epoch": 1.4856230031948883,
+      "grad_norm": 1.0718435481145705,
+      "learning_rate": 5.95151043761711e-06,
+      "loss": 0.1014,
+      "step": 3720
+    },
+    {
+      "epoch": 1.4864217252396166,
+      "grad_norm": 0.929707333376811,
+      "learning_rate": 5.9469476280479685e-06,
+      "loss": 0.0907,
+      "step": 3722
+    },
+    {
+      "epoch": 1.487220447284345,
+      "grad_norm": 1.0656203252614662,
+      "learning_rate": 5.9423840004054235e-06,
+      "loss": 0.1024,
+      "step": 3724
+    },
+    {
+      "epoch": 1.4880191693290734,
+      "grad_norm": 1.1218591644267362,
+      "learning_rate": 5.9378195586320155e-06,
+      "loss": 0.1046,
+      "step": 3726
+    },
+    {
+      "epoch": 1.488817891373802,
+      "grad_norm": 1.171155440361567,
+      "learning_rate": 5.933254306670995e-06,
+      "loss": 0.1007,
+      "step": 3728
+    },
+    {
+      "epoch": 1.4896166134185305,
+      "grad_norm": 1.0258926148385281,
+      "learning_rate": 5.9286882484663054e-06,
+      "loss": 0.0955,
+      "step": 3730
+    },
+    {
+      "epoch": 1.4904153354632588,
+      "grad_norm": 1.133084775825884,
+      "learning_rate": 5.924121387962594e-06,
+      "loss": 0.1002,
+      "step": 3732
+    },
+    {
+      "epoch": 1.4912140575079873,
+      "grad_norm": 1.1855051630642663,
+      "learning_rate": 5.919553729105194e-06,
+      "loss": 0.1012,
+      "step": 3734
+    },
+    {
+      "epoch": 1.4920127795527156,
+      "grad_norm": 1.1146676203875627,
+      "learning_rate": 5.914985275840135e-06,
+      "loss": 0.1081,
+      "step": 3736
+    },
+    {
+      "epoch": 1.4928115015974441,
+      "grad_norm": 1.1041278385511843,
+      "learning_rate": 5.910416032114128e-06,
+      "loss": 0.1026,
+      "step": 3738
+    },
+    {
+      "epoch": 1.4936102236421724,
+      "grad_norm": 0.9898027730365901,
+      "learning_rate": 5.905846001874566e-06,
+      "loss": 0.0888,
+      "step": 3740
+    },
+    {
+      "epoch": 1.494408945686901,
+      "grad_norm": 1.1585515500268038,
+      "learning_rate": 5.90127518906953e-06,
+      "loss": 0.1052,
+      "step": 3742
+    },
+    {
+      "epoch": 1.4952076677316293,
+      "grad_norm": 1.0347303041203528,
+      "learning_rate": 5.896703597647765e-06,
+      "loss": 0.1058,
+      "step": 3744
+    },
+    {
+      "epoch": 1.4960063897763578,
+      "grad_norm": 0.9935016480185027,
+      "learning_rate": 5.892131231558696e-06,
+      "loss": 0.1115,
+      "step": 3746
+    },
+    {
+      "epoch": 1.4968051118210863,
+      "grad_norm": 1.0501951176001236,
+      "learning_rate": 5.88755809475242e-06,
+      "loss": 0.1025,
+      "step": 3748
+    },
+    {
+      "epoch": 1.4976038338658146,
+      "grad_norm": 1.0008810336543572,
+      "learning_rate": 5.882984191179691e-06,
+      "loss": 0.0984,
+      "step": 3750
+    },
+    {
+      "epoch": 1.4984025559105432,
+      "grad_norm": 1.2054742114396801,
+      "learning_rate": 5.878409524791931e-06,
+      "loss": 0.1101,
+      "step": 3752
+    },
+    {
+      "epoch": 1.4992012779552715,
+      "grad_norm": 1.094413921625223,
+      "learning_rate": 5.8738340995412216e-06,
+      "loss": 0.1055,
+      "step": 3754
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0505101513083803,
+      "learning_rate": 5.869257919380298e-06,
+      "loss": 0.1026,
+      "step": 3756
+    },
+    {
+      "epoch": 1.5007987220447285,
+      "grad_norm": 0.970894387236391,
+      "learning_rate": 5.864680988262546e-06,
+      "loss": 0.0935,
+      "step": 3758
+    },
+    {
+      "epoch": 1.5015974440894568,
+      "grad_norm": 1.1554311867796712,
+      "learning_rate": 5.8601033101420055e-06,
+      "loss": 0.1188,
+      "step": 3760
+    },
+    {
+      "epoch": 1.5023961661341851,
+      "grad_norm": 1.0830901347700692,
+      "learning_rate": 5.855524888973358e-06,
+      "loss": 0.1106,
+      "step": 3762
+    },
+    {
+      "epoch": 1.5031948881789137,
+      "grad_norm": 1.057272294501186,
+      "learning_rate": 5.850945728711925e-06,
+      "loss": 0.1028,
+      "step": 3764
+    },
+    {
+      "epoch": 1.5039936102236422,
+      "grad_norm": 1.0336589663794098,
+      "learning_rate": 5.846365833313672e-06,
+      "loss": 0.1019,
+      "step": 3766
+    },
+    {
+      "epoch": 1.5047923322683707,
+      "grad_norm": 1.11484447745884,
+      "learning_rate": 5.841785206735192e-06,
+      "loss": 0.1009,
+      "step": 3768
+    },
+    {
+      "epoch": 1.505591054313099,
+      "grad_norm": 1.1990777552155663,
+      "learning_rate": 5.837203852933721e-06,
+      "loss": 0.1078,
+      "step": 3770
+    },
+    {
+      "epoch": 1.5063897763578273,
+      "grad_norm": 1.2341825295207347,
+      "learning_rate": 5.83262177586711e-06,
+      "loss": 0.1103,
+      "step": 3772
+    },
+    {
+      "epoch": 1.5071884984025559,
+      "grad_norm": 1.372034044865105,
+      "learning_rate": 5.828038979493844e-06,
+      "loss": 0.1056,
+      "step": 3774
+    },
+    {
+      "epoch": 1.5079872204472844,
+      "grad_norm": 1.1741674239048192,
+      "learning_rate": 5.823455467773027e-06,
+      "loss": 0.1079,
+      "step": 3776
+    },
+    {
+      "epoch": 1.508785942492013,
+      "grad_norm": 1.0252917230334802,
+      "learning_rate": 5.81887124466438e-06,
+      "loss": 0.0914,
+      "step": 3778
+    },
+    {
+      "epoch": 1.5095846645367412,
+      "grad_norm": 1.1697427527897497,
+      "learning_rate": 5.814286314128239e-06,
+      "loss": 0.0991,
+      "step": 3780
+    },
+    {
+      "epoch": 1.5103833865814695,
+      "grad_norm": 1.042769204041001,
+      "learning_rate": 5.809700680125552e-06,
+      "loss": 0.1059,
+      "step": 3782
+    },
+    {
+      "epoch": 1.511182108626198,
+      "grad_norm": 1.0296503251696472,
+      "learning_rate": 5.805114346617874e-06,
+      "loss": 0.0946,
+      "step": 3784
+    },
+    {
+      "epoch": 1.5119808306709266,
+      "grad_norm": 1.009763042669415,
+      "learning_rate": 5.800527317567365e-06,
+      "loss": 0.0968,
+      "step": 3786
+    },
+    {
+      "epoch": 1.5127795527156551,
+      "grad_norm": 1.0866525391827386,
+      "learning_rate": 5.795939596936783e-06,
+      "loss": 0.1014,
+      "step": 3788
+    },
+    {
+      "epoch": 1.5135782747603834,
+      "grad_norm": 1.1514100776695269,
+      "learning_rate": 5.791351188689489e-06,
+      "loss": 0.1016,
+      "step": 3790
+    },
+    {
+      "epoch": 1.5143769968051117,
+      "grad_norm": 1.2403482979785827,
+      "learning_rate": 5.786762096789431e-06,
+      "loss": 0.1046,
+      "step": 3792
+    },
+    {
+      "epoch": 1.5151757188498403,
+      "grad_norm": 1.1732506896242403,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.1127,
+      "step": 3794
+    },
+    {
+      "epoch": 1.5159744408945688,
+      "grad_norm": 1.0881019594599013,
+      "learning_rate": 5.777581877889788e-06,
+      "loss": 0.0894,
+      "step": 3796
+    },
+    {
+      "epoch": 1.516773162939297,
+      "grad_norm": 1.1933368615094555,
+      "learning_rate": 5.772990758821046e-06,
+      "loss": 0.1078,
+      "step": 3798
+    },
+    {
+      "epoch": 1.5175718849840254,
+      "grad_norm": 0.9420287917337775,
+      "learning_rate": 5.768398971961221e-06,
+      "loss": 0.0926,
+      "step": 3800
+    },
+    {
+      "epoch": 1.518370607028754,
+      "grad_norm": 1.0673845641768755,
+      "learning_rate": 5.763806521277184e-06,
+      "loss": 0.0958,
+      "step": 3802
+    },
+    {
+      "epoch": 1.5191693290734825,
+      "grad_norm": 1.0756212442333353,
+      "learning_rate": 5.759213410736377e-06,
+      "loss": 0.0984,
+      "step": 3804
+    },
+    {
+      "epoch": 1.519968051118211,
+      "grad_norm": 1.0303021520595748,
+      "learning_rate": 5.7546196443068195e-06,
+      "loss": 0.099,
+      "step": 3806
+    },
+    {
+      "epoch": 1.5207667731629393,
+      "grad_norm": 1.1602092966550535,
+      "learning_rate": 5.750025225957086e-06,
+      "loss": 0.1052,
+      "step": 3808
+    },
+    {
+      "epoch": 1.5215654952076676,
+      "grad_norm": 1.1125067715645671,
+      "learning_rate": 5.745430159656324e-06,
+      "loss": 0.0988,
+      "step": 3810
+    },
+    {
+      "epoch": 1.5223642172523961,
+      "grad_norm": 1.0076518432605175,
+      "learning_rate": 5.740834449374237e-06,
+      "loss": 0.1006,
+      "step": 3812
+    },
+    {
+      "epoch": 1.5231629392971247,
+      "grad_norm": 1.1636274304360332,
+      "learning_rate": 5.7362380990810836e-06,
+      "loss": 0.0982,
+      "step": 3814
+    },
+    {
+      "epoch": 1.5239616613418532,
+      "grad_norm": 1.589543826243573,
+      "learning_rate": 5.731641112747679e-06,
+      "loss": 0.1033,
+      "step": 3816
+    },
+    {
+      "epoch": 1.5247603833865815,
+      "grad_norm": 1.130992557679762,
+      "learning_rate": 5.7270434943453844e-06,
+      "loss": 0.098,
+      "step": 3818
+    },
+    {
+      "epoch": 1.5255591054313098,
+      "grad_norm": 0.9432691902452544,
+      "learning_rate": 5.722445247846107e-06,
+      "loss": 0.0888,
+      "step": 3820
+    },
+    {
+      "epoch": 1.5263578274760383,
+      "grad_norm": 1.106347020699136,
+      "learning_rate": 5.717846377222302e-06,
+      "loss": 0.0936,
+      "step": 3822
+    },
+    {
+      "epoch": 1.5271565495207668,
+      "grad_norm": 1.1052796082215754,
+      "learning_rate": 5.713246886446954e-06,
+      "loss": 0.1007,
+      "step": 3824
+    },
+    {
+      "epoch": 1.5279552715654952,
+      "grad_norm": 1.1941612947080527,
+      "learning_rate": 5.708646779493592e-06,
+      "loss": 0.1107,
+      "step": 3826
+    },
+    {
+      "epoch": 1.5287539936102237,
+      "grad_norm": 1.0917404470546062,
+      "learning_rate": 5.704046060336276e-06,
+      "loss": 0.0985,
+      "step": 3828
+    },
+    {
+      "epoch": 1.529552715654952,
+      "grad_norm": 1.026940535614031,
+      "learning_rate": 5.699444732949592e-06,
+      "loss": 0.1008,
+      "step": 3830
+    },
+    {
+      "epoch": 1.5303514376996805,
+      "grad_norm": 1.1825164845911036,
+      "learning_rate": 5.694842801308651e-06,
+      "loss": 0.1039,
+      "step": 3832
+    },
+    {
+      "epoch": 1.531150159744409,
+      "grad_norm": 1.2523857121893918,
+      "learning_rate": 5.69024026938909e-06,
+      "loss": 0.1073,
+      "step": 3834
+    },
+    {
+      "epoch": 1.5319488817891374,
+      "grad_norm": 0.9618590795563883,
+      "learning_rate": 5.6856371411670605e-06,
+      "loss": 0.0895,
+      "step": 3836
+    },
+    {
+      "epoch": 1.5327476038338657,
+      "grad_norm": 1.087295231368687,
+      "learning_rate": 5.681033420619233e-06,
+      "loss": 0.1079,
+      "step": 3838
+    },
+    {
+      "epoch": 1.5335463258785942,
+      "grad_norm": 1.0751240176712655,
+      "learning_rate": 5.676429111722786e-06,
+      "loss": 0.119,
+      "step": 3840
+    },
+    {
+      "epoch": 1.5343450479233227,
+      "grad_norm": 0.9761262443774417,
+      "learning_rate": 5.67182421845541e-06,
+      "loss": 0.0985,
+      "step": 3842
+    },
+    {
+      "epoch": 1.5351437699680512,
+      "grad_norm": 1.064750803984951,
+      "learning_rate": 5.6672187447952944e-06,
+      "loss": 0.1164,
+      "step": 3844
+    },
+    {
+      "epoch": 1.5359424920127795,
+      "grad_norm": 1.0336803651600035,
+      "learning_rate": 5.662612694721139e-06,
+      "loss": 0.0933,
+      "step": 3846
+    },
+    {
+      "epoch": 1.5367412140575079,
+      "grad_norm": 1.216380707446393,
+      "learning_rate": 5.6580060722121325e-06,
+      "loss": 0.0985,
+      "step": 3848
+    },
+    {
+      "epoch": 1.5375399361022364,
+      "grad_norm": 1.0486840682788419,
+      "learning_rate": 5.6533988812479626e-06,
+      "loss": 0.0884,
+      "step": 3850
+    },
+    {
+      "epoch": 1.538338658146965,
+      "grad_norm": 1.0017531938240458,
+      "learning_rate": 5.648791125808809e-06,
+      "loss": 0.0971,
+      "step": 3852
+    },
+    {
+      "epoch": 1.5391373801916934,
+      "grad_norm": 0.9711321520409679,
+      "learning_rate": 5.644182809875338e-06,
+      "loss": 0.0967,
+      "step": 3854
+    },
+    {
+      "epoch": 1.5399361022364217,
+      "grad_norm": 1.0444214456541656,
+      "learning_rate": 5.639573937428699e-06,
+      "loss": 0.1005,
+      "step": 3856
+    },
+    {
+      "epoch": 1.54073482428115,
+      "grad_norm": 1.1384563164764703,
+      "learning_rate": 5.634964512450522e-06,
+      "loss": 0.0937,
+      "step": 3858
+    },
+    {
+      "epoch": 1.5415335463258786,
+      "grad_norm": 1.1704875948472755,
+      "learning_rate": 5.630354538922916e-06,
+      "loss": 0.1069,
+      "step": 3860
+    },
+    {
+      "epoch": 1.542332268370607,
+      "grad_norm": 1.153542545533936,
+      "learning_rate": 5.6257440208284645e-06,
+      "loss": 0.0944,
+      "step": 3862
+    },
+    {
+      "epoch": 1.5431309904153354,
+      "grad_norm": 1.2017126948215129,
+      "learning_rate": 5.621132962150216e-06,
+      "loss": 0.1056,
+      "step": 3864
+    },
+    {
+      "epoch": 1.543929712460064,
+      "grad_norm": 1.1096553295923908,
+      "learning_rate": 5.616521366871697e-06,
+      "loss": 0.1084,
+      "step": 3866
+    },
+    {
+      "epoch": 1.5447284345047922,
+      "grad_norm": 1.2932559709159934,
+      "learning_rate": 5.611909238976885e-06,
+      "loss": 0.1077,
+      "step": 3868
+    },
+    {
+      "epoch": 1.5455271565495208,
+      "grad_norm": 1.1009979044862985,
+      "learning_rate": 5.607296582450224e-06,
+      "loss": 0.0946,
+      "step": 3870
+    },
+    {
+      "epoch": 1.5463258785942493,
+      "grad_norm": 1.1753388398369289,
+      "learning_rate": 5.6026834012766155e-06,
+      "loss": 0.1056,
+      "step": 3872
+    },
+    {
+      "epoch": 1.5471246006389776,
+      "grad_norm": 1.1300535782520567,
+      "learning_rate": 5.598069699441414e-06,
+      "loss": 0.1072,
+      "step": 3874
+    },
+    {
+      "epoch": 1.547923322683706,
+      "grad_norm": 1.1361407815794689,
+      "learning_rate": 5.5934554809304184e-06,
+      "loss": 0.1057,
+      "step": 3876
+    },
+    {
+      "epoch": 1.5487220447284344,
+      "grad_norm": 1.0774923728003276,
+      "learning_rate": 5.5888407497298824e-06,
+      "loss": 0.0978,
+      "step": 3878
+    },
+    {
+      "epoch": 1.549520766773163,
+      "grad_norm": 1.1078465378909663,
+      "learning_rate": 5.584225509826497e-06,
+      "loss": 0.0847,
+      "step": 3880
+    },
+    {
+      "epoch": 1.5503194888178915,
+      "grad_norm": 1.117154173356503,
+      "learning_rate": 5.579609765207393e-06,
+      "loss": 0.1053,
+      "step": 3882
+    },
+    {
+      "epoch": 1.5511182108626198,
+      "grad_norm": 1.1442174998289312,
+      "learning_rate": 5.574993519860139e-06,
+      "loss": 0.1024,
+      "step": 3884
+    },
+    {
+      "epoch": 1.5519169329073481,
+      "grad_norm": 1.1475487693664175,
+      "learning_rate": 5.5703767777727354e-06,
+      "loss": 0.1073,
+      "step": 3886
+    },
+    {
+      "epoch": 1.5527156549520766,
+      "grad_norm": 1.1044103730150632,
+      "learning_rate": 5.565759542933612e-06,
+      "loss": 0.1073,
+      "step": 3888
+    },
+    {
+      "epoch": 1.5535143769968052,
+      "grad_norm": 1.1544497802356082,
+      "learning_rate": 5.561141819331624e-06,
+      "loss": 0.1055,
+      "step": 3890
+    },
+    {
+      "epoch": 1.5543130990415337,
+      "grad_norm": 1.13696637094894,
+      "learning_rate": 5.556523610956049e-06,
+      "loss": 0.1032,
+      "step": 3892
+    },
+    {
+      "epoch": 1.555111821086262,
+      "grad_norm": 1.0623393891516841,
+      "learning_rate": 5.55190492179658e-06,
+      "loss": 0.1139,
+      "step": 3894
+    },
+    {
+      "epoch": 1.5559105431309903,
+      "grad_norm": 1.0054614979615821,
+      "learning_rate": 5.547285755843334e-06,
+      "loss": 0.0888,
+      "step": 3896
+    },
+    {
+      "epoch": 1.5567092651757188,
+      "grad_norm": 0.9923971125933808,
+      "learning_rate": 5.542666117086832e-06,
+      "loss": 0.0884,
+      "step": 3898
+    },
+    {
+      "epoch": 1.5575079872204474,
+      "grad_norm": 1.1566532715523212,
+      "learning_rate": 5.538046009518007e-06,
+      "loss": 0.1053,
+      "step": 3900
+    },
+    {
+      "epoch": 1.5583067092651757,
+      "grad_norm": 1.0988640530351912,
+      "learning_rate": 5.5334254371281934e-06,
+      "loss": 0.0987,
+      "step": 3902
+    },
+    {
+      "epoch": 1.5591054313099042,
+      "grad_norm": 1.1389960010587739,
+      "learning_rate": 5.5288044039091335e-06,
+      "loss": 0.1075,
+      "step": 3904
+    },
+    {
+      "epoch": 1.5599041533546325,
+      "grad_norm": 1.2172334905611906,
+      "learning_rate": 5.524182913852961e-06,
+      "loss": 0.1017,
+      "step": 3906
+    },
+    {
+      "epoch": 1.560702875399361,
+      "grad_norm": 1.0961769840109552,
+      "learning_rate": 5.519560970952208e-06,
+      "loss": 0.1074,
+      "step": 3908
+    },
+    {
+      "epoch": 1.5615015974440896,
+      "grad_norm": 0.9984956732002762,
+      "learning_rate": 5.514938579199798e-06,
+      "loss": 0.1151,
+      "step": 3910
+    },
+    {
+      "epoch": 1.5623003194888179,
+      "grad_norm": 1.0275799079683399,
+      "learning_rate": 5.510315742589042e-06,
+      "loss": 0.1024,
+      "step": 3912
+    },
+    {
+      "epoch": 1.5630990415335462,
+      "grad_norm": 1.11512381478294,
+      "learning_rate": 5.505692465113633e-06,
+      "loss": 0.1132,
+      "step": 3914
+    },
+    {
+      "epoch": 1.5638977635782747,
+      "grad_norm": 1.0178232850189146,
+      "learning_rate": 5.5010687507676466e-06,
+      "loss": 0.1075,
+      "step": 3916
+    },
+    {
+      "epoch": 1.5646964856230032,
+      "grad_norm": 1.0142014809665243,
+      "learning_rate": 5.496444603545535e-06,
+      "loss": 0.0962,
+      "step": 3918
+    },
+    {
+      "epoch": 1.5654952076677318,
+      "grad_norm": 1.046870424333617,
+      "learning_rate": 5.491820027442126e-06,
+      "loss": 0.1043,
+      "step": 3920
+    },
+    {
+      "epoch": 1.56629392971246,
+      "grad_norm": 1.1510480151229603,
+      "learning_rate": 5.487195026452619e-06,
+      "loss": 0.102,
+      "step": 3922
+    },
+    {
+      "epoch": 1.5670926517571884,
+      "grad_norm": 1.27959847253193,
+      "learning_rate": 5.482569604572577e-06,
+      "loss": 0.1021,
+      "step": 3924
+    },
+    {
+      "epoch": 1.567891373801917,
+      "grad_norm": 0.9918976439707066,
+      "learning_rate": 5.477943765797926e-06,
+      "loss": 0.1,
+      "step": 3926
+    },
+    {
+      "epoch": 1.5686900958466454,
+      "grad_norm": 1.080408951439636,
+      "learning_rate": 5.473317514124958e-06,
+      "loss": 0.1054,
+      "step": 3928
+    },
+    {
+      "epoch": 1.569488817891374,
+      "grad_norm": 1.081550022627091,
+      "learning_rate": 5.4686908535503135e-06,
+      "loss": 0.1019,
+      "step": 3930
+    },
+    {
+      "epoch": 1.5702875399361023,
+      "grad_norm": 1.0811985668125383,
+      "learning_rate": 5.464063788070996e-06,
+      "loss": 0.0936,
+      "step": 3932
+    },
+    {
+      "epoch": 1.5710862619808306,
+      "grad_norm": 0.9821534674944753,
+      "learning_rate": 5.459436321684348e-06,
+      "loss": 0.0933,
+      "step": 3934
+    },
+    {
+      "epoch": 1.571884984025559,
+      "grad_norm": 1.0914512560209013,
+      "learning_rate": 5.454808458388069e-06,
+      "loss": 0.1148,
+      "step": 3936
+    },
+    {
+      "epoch": 1.5726837060702876,
+      "grad_norm": 1.164283845889731,
+      "learning_rate": 5.4501802021801935e-06,
+      "loss": 0.1018,
+      "step": 3938
+    },
+    {
+      "epoch": 1.573482428115016,
+      "grad_norm": 0.970687102295011,
+      "learning_rate": 5.445551557059098e-06,
+      "loss": 0.0936,
+      "step": 3940
+    },
+    {
+      "epoch": 1.5742811501597445,
+      "grad_norm": 1.1042280103745048,
+      "learning_rate": 5.440922527023494e-06,
+      "loss": 0.093,
+      "step": 3942
+    },
+    {
+      "epoch": 1.5750798722044728,
+      "grad_norm": 1.2131440637336284,
+      "learning_rate": 5.436293116072431e-06,
+      "loss": 0.1136,
+      "step": 3944
+    },
+    {
+      "epoch": 1.5758785942492013,
+      "grad_norm": 1.055324845043691,
+      "learning_rate": 5.431663328205279e-06,
+      "loss": 0.1043,
+      "step": 3946
+    },
+    {
+      "epoch": 1.5766773162939298,
+      "grad_norm": 0.9811618715390004,
+      "learning_rate": 5.42703316742174e-06,
+      "loss": 0.093,
+      "step": 3948
+    },
+    {
+      "epoch": 1.5774760383386581,
+      "grad_norm": 1.2173508502121098,
+      "learning_rate": 5.4224026377218365e-06,
+      "loss": 0.1045,
+      "step": 3950
+    },
+    {
+      "epoch": 1.5782747603833864,
+      "grad_norm": 1.1661185719259703,
+      "learning_rate": 5.417771743105908e-06,
+      "loss": 0.1098,
+      "step": 3952
+    },
+    {
+      "epoch": 1.579073482428115,
+      "grad_norm": 1.0058471682242138,
+      "learning_rate": 5.413140487574608e-06,
+      "loss": 0.1106,
+      "step": 3954
+    },
+    {
+      "epoch": 1.5798722044728435,
+      "grad_norm": 1.0394406333548927,
+      "learning_rate": 5.408508875128911e-06,
+      "loss": 0.0929,
+      "step": 3956
+    },
+    {
+      "epoch": 1.580670926517572,
+      "grad_norm": 1.1297610914562695,
+      "learning_rate": 5.403876909770087e-06,
+      "loss": 0.101,
+      "step": 3958
+    },
+    {
+      "epoch": 1.5814696485623003,
+      "grad_norm": 1.1543975422216644,
+      "learning_rate": 5.399244595499721e-06,
+      "loss": 0.1121,
+      "step": 3960
+    },
+    {
+      "epoch": 1.5822683706070286,
+      "grad_norm": 1.1306410767900936,
+      "learning_rate": 5.394611936319692e-06,
+      "loss": 0.1016,
+      "step": 3962
+    },
+    {
+      "epoch": 1.5830670926517572,
+      "grad_norm": 0.9448150345192816,
+      "learning_rate": 5.389978936232185e-06,
+      "loss": 0.0893,
+      "step": 3964
+    },
+    {
+      "epoch": 1.5838658146964857,
+      "grad_norm": 0.9463251751033753,
+      "learning_rate": 5.385345599239669e-06,
+      "loss": 0.0946,
+      "step": 3966
+    },
+    {
+      "epoch": 1.5846645367412142,
+      "grad_norm": 1.0596847272661667,
+      "learning_rate": 5.380711929344915e-06,
+      "loss": 0.1151,
+      "step": 3968
+    },
+    {
+      "epoch": 1.5854632587859425,
+      "grad_norm": 1.0752597682928209,
+      "learning_rate": 5.376077930550973e-06,
+      "loss": 0.107,
+      "step": 3970
+    },
+    {
+      "epoch": 1.5862619808306708,
+      "grad_norm": 1.116596755502733,
+      "learning_rate": 5.371443606861186e-06,
+      "loss": 0.1019,
+      "step": 3972
+    },
+    {
+      "epoch": 1.5870607028753994,
+      "grad_norm": 1.0051885963957903,
+      "learning_rate": 5.366808962279166e-06,
+      "loss": 0.0962,
+      "step": 3974
+    },
+    {
+      "epoch": 1.5878594249201279,
+      "grad_norm": 1.027781803302719,
+      "learning_rate": 5.362174000808813e-06,
+      "loss": 0.0971,
+      "step": 3976
+    },
+    {
+      "epoch": 1.5886581469648562,
+      "grad_norm": 0.8827091346960793,
+      "learning_rate": 5.3575387264542934e-06,
+      "loss": 0.0855,
+      "step": 3978
+    },
+    {
+      "epoch": 1.5894568690095847,
+      "grad_norm": 1.072987187940347,
+      "learning_rate": 5.352903143220051e-06,
+      "loss": 0.0967,
+      "step": 3980
+    },
+    {
+      "epoch": 1.590255591054313,
+      "grad_norm": 1.0526867412563083,
+      "learning_rate": 5.348267255110787e-06,
+      "loss": 0.092,
+      "step": 3982
+    },
+    {
+      "epoch": 1.5910543130990416,
+      "grad_norm": 1.1489504347454111,
+      "learning_rate": 5.343631066131476e-06,
+      "loss": 0.1105,
+      "step": 3984
+    },
+    {
+      "epoch": 1.59185303514377,
+      "grad_norm": 1.1540840295187986,
+      "learning_rate": 5.338994580287345e-06,
+      "loss": 0.1011,
+      "step": 3986
+    },
+    {
+      "epoch": 1.5926517571884984,
+      "grad_norm": 1.1394245571406438,
+      "learning_rate": 5.334357801583882e-06,
+      "loss": 0.0996,
+      "step": 3988
+    },
+    {
+      "epoch": 1.5934504792332267,
+      "grad_norm": 1.0287583891620165,
+      "learning_rate": 5.329720734026824e-06,
+      "loss": 0.0957,
+      "step": 3990
+    },
+    {
+      "epoch": 1.5942492012779552,
+      "grad_norm": 1.0779004754661263,
+      "learning_rate": 5.325083381622165e-06,
+      "loss": 0.0979,
+      "step": 3992
+    },
+    {
+      "epoch": 1.5950479233226837,
+      "grad_norm": 1.200252910830259,
+      "learning_rate": 5.320445748376133e-06,
+      "loss": 0.0968,
+      "step": 3994
+    },
+    {
+      "epoch": 1.5958466453674123,
+      "grad_norm": 1.2695805654962276,
+      "learning_rate": 5.3158078382952095e-06,
+      "loss": 0.1137,
+      "step": 3996
+    },
+    {
+      "epoch": 1.5966453674121406,
+      "grad_norm": 1.0386371478246852,
+      "learning_rate": 5.311169655386112e-06,
+      "loss": 0.0981,
+      "step": 3998
+    },
+    {
+      "epoch": 1.5974440894568689,
+      "grad_norm": 1.1857255254365702,
+      "learning_rate": 5.30653120365579e-06,
+      "loss": 0.1205,
+      "step": 4000
+    },
+    {
+      "epoch": 1.5974440894568689,
+      "eval_loss": 0.15445688366889954,
+      "eval_runtime": 417.6616,
+      "eval_samples_per_second": 42.635,
+      "eval_steps_per_second": 5.33,
+      "step": 4000
+    },
+    {
+      "epoch": 1.5982428115015974,
+      "grad_norm": 1.1097474658690198,
+      "learning_rate": 5.301892487111431e-06,
+      "loss": 0.1025,
+      "step": 4002
+    },
+    {
+      "epoch": 1.599041533546326,
+      "grad_norm": 1.1919685218314882,
+      "learning_rate": 5.2972535097604474e-06,
+      "loss": 0.112,
+      "step": 4004
+    },
+    {
+      "epoch": 1.5998402555910545,
+      "grad_norm": 0.9947826284264967,
+      "learning_rate": 5.292614275610476e-06,
+      "loss": 0.1021,
+      "step": 4006
+    },
+    {
+      "epoch": 1.6006389776357828,
+      "grad_norm": 0.9427973792484328,
+      "learning_rate": 5.28797478866938e-06,
+      "loss": 0.091,
+      "step": 4008
+    },
+    {
+      "epoch": 1.601437699680511,
+      "grad_norm": 1.1411139408840176,
+      "learning_rate": 5.283335052945238e-06,
+      "loss": 0.0975,
+      "step": 4010
+    },
+    {
+      "epoch": 1.6022364217252396,
+      "grad_norm": 1.1012457970159555,
+      "learning_rate": 5.278695072446342e-06,
+      "loss": 0.0951,
+      "step": 4012
+    },
+    {
+      "epoch": 1.6030351437699681,
+      "grad_norm": 1.1626019144814381,
+      "learning_rate": 5.2740548511812e-06,
+      "loss": 0.0998,
+      "step": 4014
+    },
+    {
+      "epoch": 1.6038338658146964,
+      "grad_norm": 1.1201103649096615,
+      "learning_rate": 5.269414393158523e-06,
+      "loss": 0.0987,
+      "step": 4016
+    },
+    {
+      "epoch": 1.604632587859425,
+      "grad_norm": 1.1630697149501579,
+      "learning_rate": 5.264773702387232e-06,
+      "loss": 0.1022,
+      "step": 4018
+    },
+    {
+      "epoch": 1.6054313099041533,
+      "grad_norm": 1.2522557555471332,
+      "learning_rate": 5.2601327828764415e-06,
+      "loss": 0.1042,
+      "step": 4020
+    },
+    {
+      "epoch": 1.6062300319488818,
+      "grad_norm": 1.1002664728813976,
+      "learning_rate": 5.255491638635472e-06,
+      "loss": 0.1131,
+      "step": 4022
+    },
+    {
+      "epoch": 1.6070287539936103,
+      "grad_norm": 1.1476460289094692,
+      "learning_rate": 5.250850273673831e-06,
+      "loss": 0.1041,
+      "step": 4024
+    },
+    {
+      "epoch": 1.6078274760383386,
+      "grad_norm": 1.2997227467476715,
+      "learning_rate": 5.246208692001224e-06,
+      "loss": 0.109,
+      "step": 4026
+    },
+    {
+      "epoch": 1.608626198083067,
+      "grad_norm": 1.094935343872026,
+      "learning_rate": 5.241566897627536e-06,
+      "loss": 0.1007,
+      "step": 4028
+    },
+    {
+      "epoch": 1.6094249201277955,
+      "grad_norm": 1.083296893505705,
+      "learning_rate": 5.236924894562841e-06,
+      "loss": 0.102,
+      "step": 4030
+    },
+    {
+      "epoch": 1.610223642172524,
+      "grad_norm": 1.0723795202774216,
+      "learning_rate": 5.232282686817392e-06,
+      "loss": 0.0996,
+      "step": 4032
+    },
+    {
+      "epoch": 1.6110223642172525,
+      "grad_norm": 1.078393017722415,
+      "learning_rate": 5.227640278401616e-06,
+      "loss": 0.1047,
+      "step": 4034
+    },
+    {
+      "epoch": 1.6118210862619808,
+      "grad_norm": 1.1241219264829283,
+      "learning_rate": 5.222997673326118e-06,
+      "loss": 0.1125,
+      "step": 4036
+    },
+    {
+      "epoch": 1.6126198083067091,
+      "grad_norm": 1.1494086764740628,
+      "learning_rate": 5.218354875601672e-06,
+      "loss": 0.1077,
+      "step": 4038
+    },
+    {
+      "epoch": 1.6134185303514377,
+      "grad_norm": 1.0313683486241818,
+      "learning_rate": 5.213711889239214e-06,
+      "loss": 0.1056,
+      "step": 4040
+    },
+    {
+      "epoch": 1.6142172523961662,
+      "grad_norm": 1.0785913080866496,
+      "learning_rate": 5.209068718249849e-06,
+      "loss": 0.1026,
+      "step": 4042
+    },
+    {
+      "epoch": 1.6150159744408947,
+      "grad_norm": 1.1093921293966087,
+      "learning_rate": 5.2044253666448364e-06,
+      "loss": 0.1019,
+      "step": 4044
+    },
+    {
+      "epoch": 1.615814696485623,
+      "grad_norm": 1.048516005384363,
+      "learning_rate": 5.1997818384355945e-06,
+      "loss": 0.1198,
+      "step": 4046
+    },
+    {
+      "epoch": 1.6166134185303513,
+      "grad_norm": 1.0667431272462546,
+      "learning_rate": 5.195138137633695e-06,
+      "loss": 0.1,
+      "step": 4048
+    },
+    {
+      "epoch": 1.6174121405750799,
+      "grad_norm": 1.162801685632335,
+      "learning_rate": 5.190494268250856e-06,
+      "loss": 0.108,
+      "step": 4050
+    },
+    {
+      "epoch": 1.6182108626198084,
+      "grad_norm": 1.0168007347140544,
+      "learning_rate": 5.185850234298943e-06,
+      "loss": 0.0927,
+      "step": 4052
+    },
+    {
+      "epoch": 1.6190095846645367,
+      "grad_norm": 1.0037984889528644,
+      "learning_rate": 5.1812060397899624e-06,
+      "loss": 0.0876,
+      "step": 4054
+    },
+    {
+      "epoch": 1.619808306709265,
+      "grad_norm": 1.0382499080256122,
+      "learning_rate": 5.17656168873606e-06,
+      "loss": 0.107,
+      "step": 4056
+    },
+    {
+      "epoch": 1.6206070287539935,
+      "grad_norm": 1.1185835251899559,
+      "learning_rate": 5.171917185149518e-06,
+      "loss": 0.1044,
+      "step": 4058
+    },
+    {
+      "epoch": 1.621405750798722,
+      "grad_norm": 1.0657788685206113,
+      "learning_rate": 5.167272533042748e-06,
+      "loss": 0.097,
+      "step": 4060
+    },
+    {
+      "epoch": 1.6222044728434506,
+      "grad_norm": 1.0542383484498254,
+      "learning_rate": 5.162627736428293e-06,
+      "loss": 0.0946,
+      "step": 4062
+    },
+    {
+      "epoch": 1.623003194888179,
+      "grad_norm": 1.1196684728128978,
+      "learning_rate": 5.157982799318817e-06,
+      "loss": 0.1071,
+      "step": 4064
+    },
+    {
+      "epoch": 1.6238019169329072,
+      "grad_norm": 1.0973928801710329,
+      "learning_rate": 5.153337725727109e-06,
+      "loss": 0.0965,
+      "step": 4066
+    },
+    {
+      "epoch": 1.6246006389776357,
+      "grad_norm": 1.085899130068226,
+      "learning_rate": 5.148692519666072e-06,
+      "loss": 0.087,
+      "step": 4068
+    },
+    {
+      "epoch": 1.6253993610223643,
+      "grad_norm": 1.0337553281158798,
+      "learning_rate": 5.1440471851487286e-06,
+      "loss": 0.1031,
+      "step": 4070
+    },
+    {
+      "epoch": 1.6261980830670928,
+      "grad_norm": 1.0455524042989242,
+      "learning_rate": 5.139401726188208e-06,
+      "loss": 0.1012,
+      "step": 4072
+    },
+    {
+      "epoch": 1.626996805111821,
+      "grad_norm": 1.1232283088505088,
+      "learning_rate": 5.1347561467977495e-06,
+      "loss": 0.1048,
+      "step": 4074
+    },
+    {
+      "epoch": 1.6277955271565494,
+      "grad_norm": 1.1377273856980297,
+      "learning_rate": 5.130110450990694e-06,
+      "loss": 0.1015,
+      "step": 4076
+    },
+    {
+      "epoch": 1.628594249201278,
+      "grad_norm": 1.0376776381456139,
+      "learning_rate": 5.1254646427804855e-06,
+      "loss": 0.0892,
+      "step": 4078
+    },
+    {
+      "epoch": 1.6293929712460065,
+      "grad_norm": 1.0639780774252217,
+      "learning_rate": 5.120818726180662e-06,
+      "loss": 0.0909,
+      "step": 4080
+    },
+    {
+      "epoch": 1.630191693290735,
+      "grad_norm": 1.0527532913896227,
+      "learning_rate": 5.116172705204859e-06,
+      "loss": 0.092,
+      "step": 4082
+    },
+    {
+      "epoch": 1.6309904153354633,
+      "grad_norm": 1.129982374301854,
+      "learning_rate": 5.111526583866801e-06,
+      "loss": 0.1016,
+      "step": 4084
+    },
+    {
+      "epoch": 1.6317891373801916,
+      "grad_norm": 1.0125124329212987,
+      "learning_rate": 5.106880366180297e-06,
+      "loss": 0.0933,
+      "step": 4086
+    },
+    {
+      "epoch": 1.6325878594249201,
+      "grad_norm": 1.0905837841114607,
+      "learning_rate": 5.1022340561592396e-06,
+      "loss": 0.1028,
+      "step": 4088
+    },
+    {
+      "epoch": 1.6333865814696487,
+      "grad_norm": 1.058617488709683,
+      "learning_rate": 5.097587657817605e-06,
+      "loss": 0.0988,
+      "step": 4090
+    },
+    {
+      "epoch": 1.634185303514377,
+      "grad_norm": 1.1990334757384018,
+      "learning_rate": 5.09294117516944e-06,
+      "loss": 0.1045,
+      "step": 4092
+    },
+    {
+      "epoch": 1.6349840255591053,
+      "grad_norm": 1.1978345768548917,
+      "learning_rate": 5.08829461222887e-06,
+      "loss": 0.1134,
+      "step": 4094
+    },
+    {
+      "epoch": 1.6357827476038338,
+      "grad_norm": 1.082413217695199,
+      "learning_rate": 5.083647973010085e-06,
+      "loss": 0.1002,
+      "step": 4096
+    },
+    {
+      "epoch": 1.6365814696485623,
+      "grad_norm": 1.1386432805954145,
+      "learning_rate": 5.079001261527345e-06,
+      "loss": 0.1029,
+      "step": 4098
+    },
+    {
+      "epoch": 1.6373801916932909,
+      "grad_norm": 1.1557604620641984,
+      "learning_rate": 5.074354481794969e-06,
+      "loss": 0.0951,
+      "step": 4100
+    },
+    {
+      "epoch": 1.6381789137380192,
+      "grad_norm": 1.2243404366242359,
+      "learning_rate": 5.069707637827336e-06,
+      "loss": 0.1121,
+      "step": 4102
+    },
+    {
+      "epoch": 1.6389776357827475,
+      "grad_norm": 1.157977041885688,
+      "learning_rate": 5.065060733638878e-06,
+      "loss": 0.1104,
+      "step": 4104
+    },
+    {
+      "epoch": 1.639776357827476,
+      "grad_norm": 1.094920449521866,
+      "learning_rate": 5.0604137732440875e-06,
+      "loss": 0.0898,
+      "step": 4106
+    },
+    {
+      "epoch": 1.6405750798722045,
+      "grad_norm": 0.9795046872308669,
+      "learning_rate": 5.055766760657497e-06,
+      "loss": 0.0933,
+      "step": 4108
+    },
+    {
+      "epoch": 1.641373801916933,
+      "grad_norm": 1.2246816953116304,
+      "learning_rate": 5.051119699893686e-06,
+      "loss": 0.1047,
+      "step": 4110
+    },
+    {
+      "epoch": 1.6421725239616614,
+      "grad_norm": 0.9474742886147536,
+      "learning_rate": 5.046472594967279e-06,
+      "loss": 0.0825,
+      "step": 4112
+    },
+    {
+      "epoch": 1.6429712460063897,
+      "grad_norm": 1.050959142356909,
+      "learning_rate": 5.041825449892933e-06,
+      "loss": 0.102,
+      "step": 4114
+    },
+    {
+      "epoch": 1.6437699680511182,
+      "grad_norm": 1.0798115551123924,
+      "learning_rate": 5.037178268685345e-06,
+      "loss": 0.0943,
+      "step": 4116
+    },
+    {
+      "epoch": 1.6445686900958467,
+      "grad_norm": 0.9443315270439328,
+      "learning_rate": 5.032531055359241e-06,
+      "loss": 0.094,
+      "step": 4118
+    },
+    {
+      "epoch": 1.645367412140575,
+      "grad_norm": 1.0292007756942556,
+      "learning_rate": 5.027883813929374e-06,
+      "loss": 0.0981,
+      "step": 4120
+    },
+    {
+      "epoch": 1.6461661341853036,
+      "grad_norm": 1.036373592711972,
+      "learning_rate": 5.0232365484105235e-06,
+      "loss": 0.0955,
+      "step": 4122
+    },
+    {
+      "epoch": 1.6469648562300319,
+      "grad_norm": 1.0940288914430132,
+      "learning_rate": 5.018589262817488e-06,
+      "loss": 0.106,
+      "step": 4124
+    },
+    {
+      "epoch": 1.6477635782747604,
+      "grad_norm": 1.1951722585755589,
+      "learning_rate": 5.013941961165082e-06,
+      "loss": 0.1127,
+      "step": 4126
+    },
+    {
+      "epoch": 1.648562300319489,
+      "grad_norm": 1.0463562631972685,
+      "learning_rate": 5.009294647468137e-06,
+      "loss": 0.0902,
+      "step": 4128
+    },
+    {
+      "epoch": 1.6493610223642172,
+      "grad_norm": 1.1257302379959406,
+      "learning_rate": 5.004647325741495e-06,
+      "loss": 0.0911,
+      "step": 4130
+    },
+    {
+      "epoch": 1.6501597444089455,
+      "grad_norm": 0.9285397218586545,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 4132
+    },
+    {
+      "epoch": 1.650958466453674,
+      "grad_norm": 1.13221857976679,
+      "learning_rate": 4.9953526742585065e-06,
+      "loss": 0.093,
+      "step": 4134
+    },
+    {
+      "epoch": 1.6517571884984026,
+      "grad_norm": 1.119100390728916,
+      "learning_rate": 4.990705352531864e-06,
+      "loss": 0.096,
+      "step": 4136
+    },
+    {
+      "epoch": 1.6525559105431311,
+      "grad_norm": 1.0890152144340473,
+      "learning_rate": 4.9860580388349196e-06,
+      "loss": 0.0968,
+      "step": 4138
+    },
+    {
+      "epoch": 1.6533546325878594,
+      "grad_norm": 1.0415531324033633,
+      "learning_rate": 4.981410737182515e-06,
+      "loss": 0.1018,
+      "step": 4140
+    },
+    {
+      "epoch": 1.6541533546325877,
+      "grad_norm": 1.1279757348197161,
+      "learning_rate": 4.976763451589478e-06,
+      "loss": 0.0989,
+      "step": 4142
+    },
+    {
+      "epoch": 1.6549520766773163,
+      "grad_norm": 1.1174158609629532,
+      "learning_rate": 4.972116186070626e-06,
+      "loss": 0.0981,
+      "step": 4144
+    },
+    {
+      "epoch": 1.6557507987220448,
+      "grad_norm": 0.978788514984628,
+      "learning_rate": 4.96746894464076e-06,
+      "loss": 0.086,
+      "step": 4146
+    },
+    {
+      "epoch": 1.6565495207667733,
+      "grad_norm": 1.0391532776923562,
+      "learning_rate": 4.962821731314656e-06,
+      "loss": 0.0893,
+      "step": 4148
+    },
+    {
+      "epoch": 1.6573482428115016,
+      "grad_norm": 1.1408365847624822,
+      "learning_rate": 4.958174550107069e-06,
+      "loss": 0.0926,
+      "step": 4150
+    },
+    {
+      "epoch": 1.65814696485623,
+      "grad_norm": 1.0976480883499857,
+      "learning_rate": 4.953527405032723e-06,
+      "loss": 0.0969,
+      "step": 4152
+    },
+    {
+      "epoch": 1.6589456869009584,
+      "grad_norm": 1.099197170300517,
+      "learning_rate": 4.948880300106315e-06,
+      "loss": 0.0889,
+      "step": 4154
+    },
+    {
+      "epoch": 1.659744408945687,
+      "grad_norm": 0.9962207801035776,
+      "learning_rate": 4.944233239342505e-06,
+      "loss": 0.0828,
+      "step": 4156
+    },
+    {
+      "epoch": 1.6605431309904153,
+      "grad_norm": 1.164367259050149,
+      "learning_rate": 4.939586226755913e-06,
+      "loss": 0.0993,
+      "step": 4158
+    },
+    {
+      "epoch": 1.6613418530351438,
+      "grad_norm": 1.0915643883560848,
+      "learning_rate": 4.934939266361123e-06,
+      "loss": 0.0958,
+      "step": 4160
+    },
+    {
+      "epoch": 1.6621405750798721,
+      "grad_norm": 1.0384971354394401,
+      "learning_rate": 4.930292362172667e-06,
+      "loss": 0.0987,
+      "step": 4162
+    },
+    {
+      "epoch": 1.6629392971246006,
+      "grad_norm": 0.9714525373847863,
+      "learning_rate": 4.9256455182050345e-06,
+      "loss": 0.0985,
+      "step": 4164
+    },
+    {
+      "epoch": 1.6637380191693292,
+      "grad_norm": 1.0774781447464505,
+      "learning_rate": 4.920998738472657e-06,
+      "loss": 0.0956,
+      "step": 4166
+    },
+    {
+      "epoch": 1.6645367412140575,
+      "grad_norm": 1.1101370347081296,
+      "learning_rate": 4.916352026989914e-06,
+      "loss": 0.1085,
+      "step": 4168
+    },
+    {
+      "epoch": 1.6653354632587858,
+      "grad_norm": 1.1576145184604405,
+      "learning_rate": 4.911705387771131e-06,
+      "loss": 0.0998,
+      "step": 4170
+    },
+    {
+      "epoch": 1.6661341853035143,
+      "grad_norm": 1.0183758786516857,
+      "learning_rate": 4.90705882483056e-06,
+      "loss": 0.0856,
+      "step": 4172
+    },
+    {
+      "epoch": 1.6669329073482428,
+      "grad_norm": 1.110373708674366,
+      "learning_rate": 4.902412342182396e-06,
+      "loss": 0.0922,
+      "step": 4174
+    },
+    {
+      "epoch": 1.6677316293929714,
+      "grad_norm": 1.0395390430208176,
+      "learning_rate": 4.897765943840761e-06,
+      "loss": 0.0986,
+      "step": 4176
+    },
+    {
+      "epoch": 1.6685303514376997,
+      "grad_norm": 1.0700900909956106,
+      "learning_rate": 4.8931196338197045e-06,
+      "loss": 0.1094,
+      "step": 4178
+    },
+    {
+      "epoch": 1.669329073482428,
+      "grad_norm": 1.0744452627922665,
+      "learning_rate": 4.888473416133201e-06,
+      "loss": 0.0945,
+      "step": 4180
+    },
+    {
+      "epoch": 1.6701277955271565,
+      "grad_norm": 1.1251668177944376,
+      "learning_rate": 4.883827294795142e-06,
+      "loss": 0.0973,
+      "step": 4182
+    },
+    {
+      "epoch": 1.670926517571885,
+      "grad_norm": 0.9568523532004309,
+      "learning_rate": 4.87918127381934e-06,
+      "loss": 0.0924,
+      "step": 4184
+    },
+    {
+      "epoch": 1.6717252396166136,
+      "grad_norm": 1.0605161422920366,
+      "learning_rate": 4.874535357219517e-06,
+      "loss": 0.0969,
+      "step": 4186
+    },
+    {
+      "epoch": 1.6725239616613419,
+      "grad_norm": 1.0949585639802137,
+      "learning_rate": 4.869889549009309e-06,
+      "loss": 0.1023,
+      "step": 4188
+    },
+    {
+      "epoch": 1.6733226837060702,
+      "grad_norm": 1.0200170497583934,
+      "learning_rate": 4.8652438532022505e-06,
+      "loss": 0.0968,
+      "step": 4190
+    },
+    {
+      "epoch": 1.6741214057507987,
+      "grad_norm": 1.1240991571874712,
+      "learning_rate": 4.860598273811793e-06,
+      "loss": 0.1056,
+      "step": 4192
+    },
+    {
+      "epoch": 1.6749201277955272,
+      "grad_norm": 1.1355929086530765,
+      "learning_rate": 4.855952814851272e-06,
+      "loss": 0.096,
+      "step": 4194
+    },
+    {
+      "epoch": 1.6757188498402555,
+      "grad_norm": 0.9333875889896245,
+      "learning_rate": 4.851307480333929e-06,
+      "loss": 0.0939,
+      "step": 4196
+    },
+    {
+      "epoch": 1.676517571884984,
+      "grad_norm": 1.093435156403032,
+      "learning_rate": 4.846662274272893e-06,
+      "loss": 0.0945,
+      "step": 4198
+    },
+    {
+      "epoch": 1.6773162939297124,
+      "grad_norm": 1.0883095662785185,
+      "learning_rate": 4.842017200681185e-06,
+      "loss": 0.0916,
+      "step": 4200
+    },
+    {
+      "epoch": 1.678115015974441,
+      "grad_norm": 1.0711389939013298,
+      "learning_rate": 4.8373722635717095e-06,
+      "loss": 0.1045,
+      "step": 4202
+    },
+    {
+      "epoch": 1.6789137380191694,
+      "grad_norm": 0.9565453405933899,
+      "learning_rate": 4.832727466957254e-06,
+      "loss": 0.084,
+      "step": 4204
+    },
+    {
+      "epoch": 1.6797124600638977,
+      "grad_norm": 0.9827521143045396,
+      "learning_rate": 4.828082814850484e-06,
+      "loss": 0.0895,
+      "step": 4206
+    },
+    {
+      "epoch": 1.680511182108626,
+      "grad_norm": 1.0715525135749726,
+      "learning_rate": 4.823438311263943e-06,
+      "loss": 0.0879,
+      "step": 4208
+    },
+    {
+      "epoch": 1.6813099041533546,
+      "grad_norm": 1.1041262630985615,
+      "learning_rate": 4.81879396021004e-06,
+      "loss": 0.1067,
+      "step": 4210
+    },
+    {
+      "epoch": 1.682108626198083,
+      "grad_norm": 1.0586655240128,
+      "learning_rate": 4.814149765701059e-06,
+      "loss": 0.1022,
+      "step": 4212
+    },
+    {
+      "epoch": 1.6829073482428116,
+      "grad_norm": 1.186055020520042,
+      "learning_rate": 4.809505731749144e-06,
+      "loss": 0.1139,
+      "step": 4214
+    },
+    {
+      "epoch": 1.68370607028754,
+      "grad_norm": 1.1122466690968793,
+      "learning_rate": 4.804861862366306e-06,
+      "loss": 0.1051,
+      "step": 4216
+    },
+    {
+      "epoch": 1.6845047923322682,
+      "grad_norm": 1.0754569180945435,
+      "learning_rate": 4.8002181615644055e-06,
+      "loss": 0.0961,
+      "step": 4218
+    },
+    {
+      "epoch": 1.6853035143769968,
+      "grad_norm": 1.0878287200397418,
+      "learning_rate": 4.795574633355165e-06,
+      "loss": 0.0885,
+      "step": 4220
+    },
+    {
+      "epoch": 1.6861022364217253,
+      "grad_norm": 1.1364082104476718,
+      "learning_rate": 4.790931281750152e-06,
+      "loss": 0.1,
+      "step": 4222
+    },
+    {
+      "epoch": 1.6869009584664538,
+      "grad_norm": 1.0119111534228,
+      "learning_rate": 4.786288110760787e-06,
+      "loss": 0.09,
+      "step": 4224
+    },
+    {
+      "epoch": 1.6876996805111821,
+      "grad_norm": 1.072752741199254,
+      "learning_rate": 4.78164512439833e-06,
+      "loss": 0.1017,
+      "step": 4226
+    },
+    {
+      "epoch": 1.6884984025559104,
+      "grad_norm": 1.27263253817292,
+      "learning_rate": 4.777002326673884e-06,
+      "loss": 0.1046,
+      "step": 4228
+    },
+    {
+      "epoch": 1.689297124600639,
+      "grad_norm": 1.0255757488823853,
+      "learning_rate": 4.772359721598386e-06,
+      "loss": 0.1052,
+      "step": 4230
+    },
+    {
+      "epoch": 1.6900958466453675,
+      "grad_norm": 1.1943745213053019,
+      "learning_rate": 4.767717313182611e-06,
+      "loss": 0.0992,
+      "step": 4232
+    },
+    {
+      "epoch": 1.6908945686900958,
+      "grad_norm": 1.1636475166052027,
+      "learning_rate": 4.763075105437161e-06,
+      "loss": 0.0991,
+      "step": 4234
+    },
+    {
+      "epoch": 1.6916932907348243,
+      "grad_norm": 1.0844729963688502,
+      "learning_rate": 4.758433102372466e-06,
+      "loss": 0.1026,
+      "step": 4236
+    },
+    {
+      "epoch": 1.6924920127795526,
+      "grad_norm": 1.1114480349576357,
+      "learning_rate": 4.753791307998776e-06,
+      "loss": 0.093,
+      "step": 4238
+    },
+    {
+      "epoch": 1.6932907348242812,
+      "grad_norm": 0.9636290087080969,
+      "learning_rate": 4.74914972632617e-06,
+      "loss": 0.0865,
+      "step": 4240
+    },
+    {
+      "epoch": 1.6940894568690097,
+      "grad_norm": 1.0002038172553764,
+      "learning_rate": 4.744508361364529e-06,
+      "loss": 0.0907,
+      "step": 4242
+    },
+    {
+      "epoch": 1.694888178913738,
+      "grad_norm": 1.0275565003608014,
+      "learning_rate": 4.73986721712356e-06,
+      "loss": 0.0944,
+      "step": 4244
+    },
+    {
+      "epoch": 1.6956869009584663,
+      "grad_norm": 1.0458584389637045,
+      "learning_rate": 4.73522629761277e-06,
+      "loss": 0.0982,
+      "step": 4246
+    },
+    {
+      "epoch": 1.6964856230031948,
+      "grad_norm": 1.0999887308140721,
+      "learning_rate": 4.730585606841479e-06,
+      "loss": 0.0994,
+      "step": 4248
+    },
+    {
+      "epoch": 1.6972843450479234,
+      "grad_norm": 1.0805343621335457,
+      "learning_rate": 4.725945148818801e-06,
+      "loss": 0.0922,
+      "step": 4250
+    },
+    {
+      "epoch": 1.6980830670926519,
+      "grad_norm": 1.1009086534316133,
+      "learning_rate": 4.721304927553659e-06,
+      "loss": 0.0978,
+      "step": 4252
+    },
+    {
+      "epoch": 1.6988817891373802,
+      "grad_norm": 1.0686039798032159,
+      "learning_rate": 4.716664947054764e-06,
+      "loss": 0.0926,
+      "step": 4254
+    },
+    {
+      "epoch": 1.6996805111821085,
+      "grad_norm": 1.2264200698539793,
+      "learning_rate": 4.7120252113306216e-06,
+      "loss": 0.0997,
+      "step": 4256
+    },
+    {
+      "epoch": 1.700479233226837,
+      "grad_norm": 1.0512823258442452,
+      "learning_rate": 4.707385724389526e-06,
+      "loss": 0.1054,
+      "step": 4258
+    },
+    {
+      "epoch": 1.7012779552715656,
+      "grad_norm": 1.2065999780404362,
+      "learning_rate": 4.702746490239554e-06,
+      "loss": 0.1031,
+      "step": 4260
+    },
+    {
+      "epoch": 1.702076677316294,
+      "grad_norm": 1.190627424392971,
+      "learning_rate": 4.69810751288857e-06,
+      "loss": 0.0996,
+      "step": 4262
+    },
+    {
+      "epoch": 1.7028753993610224,
+      "grad_norm": 1.136415340105126,
+      "learning_rate": 4.693468796344211e-06,
+      "loss": 0.1046,
+      "step": 4264
+    },
+    {
+      "epoch": 1.7036741214057507,
+      "grad_norm": 1.056631284731312,
+      "learning_rate": 4.6888303446138895e-06,
+      "loss": 0.0916,
+      "step": 4266
+    },
+    {
+      "epoch": 1.7044728434504792,
+      "grad_norm": 1.0261838244170962,
+      "learning_rate": 4.684192161704792e-06,
+      "loss": 0.0876,
+      "step": 4268
+    },
+    {
+      "epoch": 1.7052715654952078,
+      "grad_norm": 0.9913120866257363,
+      "learning_rate": 4.679554251623869e-06,
+      "loss": 0.0975,
+      "step": 4270
+    },
+    {
+      "epoch": 1.706070287539936,
+      "grad_norm": 0.9909781303643188,
+      "learning_rate": 4.6749166183778375e-06,
+      "loss": 0.0899,
+      "step": 4272
+    },
+    {
+      "epoch": 1.7068690095846646,
+      "grad_norm": 1.0133453073875816,
+      "learning_rate": 4.670279265973177e-06,
+      "loss": 0.0944,
+      "step": 4274
+    },
+    {
+      "epoch": 1.707667731629393,
+      "grad_norm": 1.084241919717283,
+      "learning_rate": 4.665642198416119e-06,
+      "loss": 0.1014,
+      "step": 4276
+    },
+    {
+      "epoch": 1.7084664536741214,
+      "grad_norm": 1.1144409380935076,
+      "learning_rate": 4.661005419712657e-06,
+      "loss": 0.1041,
+      "step": 4278
+    },
+    {
+      "epoch": 1.70926517571885,
+      "grad_norm": 1.0934629080398204,
+      "learning_rate": 4.656368933868525e-06,
+      "loss": 0.0912,
+      "step": 4280
+    },
+    {
+      "epoch": 1.7100638977635783,
+      "grad_norm": 1.1463371146812387,
+      "learning_rate": 4.651732744889215e-06,
+      "loss": 0.1037,
+      "step": 4282
+    },
+    {
+      "epoch": 1.7108626198083066,
+      "grad_norm": 1.1729562022224107,
+      "learning_rate": 4.647096856779951e-06,
+      "loss": 0.0965,
+      "step": 4284
+    },
+    {
+      "epoch": 1.711661341853035,
+      "grad_norm": 1.1118256334824752,
+      "learning_rate": 4.642461273545707e-06,
+      "loss": 0.1034,
+      "step": 4286
+    },
+    {
+      "epoch": 1.7124600638977636,
+      "grad_norm": 1.132712399437204,
+      "learning_rate": 4.637825999191189e-06,
+      "loss": 0.0982,
+      "step": 4288
+    },
+    {
+      "epoch": 1.7132587859424921,
+      "grad_norm": 1.0848347886256517,
+      "learning_rate": 4.633191037720835e-06,
+      "loss": 0.1018,
+      "step": 4290
+    },
+    {
+      "epoch": 1.7140575079872205,
+      "grad_norm": 1.0658858217343385,
+      "learning_rate": 4.628556393138816e-06,
+      "loss": 0.097,
+      "step": 4292
+    },
+    {
+      "epoch": 1.7148562300319488,
+      "grad_norm": 1.0923087477574933,
+      "learning_rate": 4.623922069449028e-06,
+      "loss": 0.1016,
+      "step": 4294
+    },
+    {
+      "epoch": 1.7156549520766773,
+      "grad_norm": 1.161421995581851,
+      "learning_rate": 4.619288070655086e-06,
+      "loss": 0.0925,
+      "step": 4296
+    },
+    {
+      "epoch": 1.7164536741214058,
+      "grad_norm": 0.9470108447713622,
+      "learning_rate": 4.614654400760333e-06,
+      "loss": 0.0989,
+      "step": 4298
+    },
+    {
+      "epoch": 1.7172523961661343,
+      "grad_norm": 1.0516808788040757,
+      "learning_rate": 4.610021063767818e-06,
+      "loss": 0.1055,
+      "step": 4300
+    },
+    {
+      "epoch": 1.7180511182108626,
+      "grad_norm": 1.1453407398766986,
+      "learning_rate": 4.60538806368031e-06,
+      "loss": 0.0986,
+      "step": 4302
+    },
+    {
+      "epoch": 1.718849840255591,
+      "grad_norm": 0.9890278332037488,
+      "learning_rate": 4.600755404500281e-06,
+      "loss": 0.0956,
+      "step": 4304
+    },
+    {
+      "epoch": 1.7196485623003195,
+      "grad_norm": 1.0910031893484988,
+      "learning_rate": 4.596123090229913e-06,
+      "loss": 0.0959,
+      "step": 4306
+    },
+    {
+      "epoch": 1.720447284345048,
+      "grad_norm": 0.954845783952346,
+      "learning_rate": 4.59149112487109e-06,
+      "loss": 0.0866,
+      "step": 4308
+    },
+    {
+      "epoch": 1.7212460063897763,
+      "grad_norm": 1.0693628763645768,
+      "learning_rate": 4.5868595124253915e-06,
+      "loss": 0.0984,
+      "step": 4310
+    },
+    {
+      "epoch": 1.7220447284345048,
+      "grad_norm": 1.0048364707538442,
+      "learning_rate": 4.582228256894093e-06,
+      "loss": 0.0939,
+      "step": 4312
+    },
+    {
+      "epoch": 1.7228434504792332,
+      "grad_norm": 1.1603565075058264,
+      "learning_rate": 4.577597362278165e-06,
+      "loss": 0.1026,
+      "step": 4314
+    },
+    {
+      "epoch": 1.7236421725239617,
+      "grad_norm": 1.1312773137686531,
+      "learning_rate": 4.572966832578261e-06,
+      "loss": 0.1076,
+      "step": 4316
+    },
+    {
+      "epoch": 1.7244408945686902,
+      "grad_norm": 1.0962659717799779,
+      "learning_rate": 4.568336671794722e-06,
+      "loss": 0.0956,
+      "step": 4318
+    },
+    {
+      "epoch": 1.7252396166134185,
+      "grad_norm": 1.0220568178816403,
+      "learning_rate": 4.56370688392757e-06,
+      "loss": 0.0904,
+      "step": 4320
+    },
+    {
+      "epoch": 1.7260383386581468,
+      "grad_norm": 0.9834388109812503,
+      "learning_rate": 4.5590774729765076e-06,
+      "loss": 0.0939,
+      "step": 4322
+    },
+    {
+      "epoch": 1.7268370607028753,
+      "grad_norm": 1.2111951372748109,
+      "learning_rate": 4.554448442940905e-06,
+      "loss": 0.0968,
+      "step": 4324
+    },
+    {
+      "epoch": 1.7276357827476039,
+      "grad_norm": 1.1926580241309344,
+      "learning_rate": 4.549819797819809e-06,
+      "loss": 0.1017,
+      "step": 4326
+    },
+    {
+      "epoch": 1.7284345047923324,
+      "grad_norm": 1.0572382689425925,
+      "learning_rate": 4.545191541611933e-06,
+      "loss": 0.0871,
+      "step": 4328
+    },
+    {
+      "epoch": 1.7292332268370607,
+      "grad_norm": 1.0126727217605194,
+      "learning_rate": 4.540563678315652e-06,
+      "loss": 0.0895,
+      "step": 4330
+    },
+    {
+      "epoch": 1.730031948881789,
+      "grad_norm": 1.0365109166295632,
+      "learning_rate": 4.535936211929005e-06,
+      "loss": 0.0971,
+      "step": 4332
+    },
+    {
+      "epoch": 1.7308306709265175,
+      "grad_norm": 1.2600986522602922,
+      "learning_rate": 4.5313091464496865e-06,
+      "loss": 0.1067,
+      "step": 4334
+    },
+    {
+      "epoch": 1.731629392971246,
+      "grad_norm": 1.1349358517047685,
+      "learning_rate": 4.526682485875044e-06,
+      "loss": 0.0954,
+      "step": 4336
+    },
+    {
+      "epoch": 1.7324281150159746,
+      "grad_norm": 1.0404286721064964,
+      "learning_rate": 4.5220562342020755e-06,
+      "loss": 0.1014,
+      "step": 4338
+    },
+    {
+      "epoch": 1.733226837060703,
+      "grad_norm": 1.2958712155282752,
+      "learning_rate": 4.517430395427424e-06,
+      "loss": 0.1121,
+      "step": 4340
+    },
+    {
+      "epoch": 1.7340255591054312,
+      "grad_norm": 1.1652339074566052,
+      "learning_rate": 4.512804973547383e-06,
+      "loss": 0.0985,
+      "step": 4342
+    },
+    {
+      "epoch": 1.7348242811501597,
+      "grad_norm": 1.204403964641989,
+      "learning_rate": 4.508179972557875e-06,
+      "loss": 0.1071,
+      "step": 4344
+    },
+    {
+      "epoch": 1.7356230031948883,
+      "grad_norm": 1.0146957490804427,
+      "learning_rate": 4.503555396454468e-06,
+      "loss": 0.089,
+      "step": 4346
+    },
+    {
+      "epoch": 1.7364217252396166,
+      "grad_norm": 1.038473290463769,
+      "learning_rate": 4.498931249232357e-06,
+      "loss": 0.1022,
+      "step": 4348
+    },
+    {
+      "epoch": 1.7372204472843449,
+      "grad_norm": 1.0396114915923156,
+      "learning_rate": 4.49430753488637e-06,
+      "loss": 0.0958,
+      "step": 4350
+    },
+    {
+      "epoch": 1.7380191693290734,
+      "grad_norm": 0.9976185951916329,
+      "learning_rate": 4.489684257410959e-06,
+      "loss": 0.0872,
+      "step": 4352
+    },
+    {
+      "epoch": 1.738817891373802,
+      "grad_norm": 0.8940691455776141,
+      "learning_rate": 4.485061420800202e-06,
+      "loss": 0.0736,
+      "step": 4354
+    },
+    {
+      "epoch": 1.7396166134185305,
+      "grad_norm": 1.0562465504348986,
+      "learning_rate": 4.480439029047792e-06,
+      "loss": 0.0928,
+      "step": 4356
+    },
+    {
+      "epoch": 1.7404153354632588,
+      "grad_norm": 1.3220699498489616,
+      "learning_rate": 4.47581708614704e-06,
+      "loss": 0.1032,
+      "step": 4358
+    },
+    {
+      "epoch": 1.741214057507987,
+      "grad_norm": 1.044321957199521,
+      "learning_rate": 4.471195596090867e-06,
+      "loss": 0.0893,
+      "step": 4360
+    },
+    {
+      "epoch": 1.7420127795527156,
+      "grad_norm": 1.1907505200191142,
+      "learning_rate": 4.466574562871807e-06,
+      "loss": 0.0955,
+      "step": 4362
+    },
+    {
+      "epoch": 1.7428115015974441,
+      "grad_norm": 1.2092564945972575,
+      "learning_rate": 4.461953990481995e-06,
+      "loss": 0.1037,
+      "step": 4364
+    },
+    {
+      "epoch": 1.7436102236421727,
+      "grad_norm": 1.0977545829393272,
+      "learning_rate": 4.45733388291317e-06,
+      "loss": 0.0854,
+      "step": 4366
+    },
+    {
+      "epoch": 1.744408945686901,
+      "grad_norm": 1.0941242092439605,
+      "learning_rate": 4.452714244156667e-06,
+      "loss": 0.1029,
+      "step": 4368
+    },
+    {
+      "epoch": 1.7452076677316293,
+      "grad_norm": 1.1568260987190029,
+      "learning_rate": 4.448095078203421e-06,
+      "loss": 0.0954,
+      "step": 4370
+    },
+    {
+      "epoch": 1.7460063897763578,
+      "grad_norm": 1.021261678557611,
+      "learning_rate": 4.443476389043955e-06,
+      "loss": 0.1055,
+      "step": 4372
+    },
+    {
+      "epoch": 1.7468051118210863,
+      "grad_norm": 1.1554032258458062,
+      "learning_rate": 4.438858180668379e-06,
+      "loss": 0.0907,
+      "step": 4374
+    },
+    {
+      "epoch": 1.7476038338658149,
+      "grad_norm": 0.9759468899388195,
+      "learning_rate": 4.434240457066388e-06,
+      "loss": 0.0893,
+      "step": 4376
+    },
+    {
+      "epoch": 1.7484025559105432,
+      "grad_norm": 1.1342860014921987,
+      "learning_rate": 4.429623222227265e-06,
+      "loss": 0.0965,
+      "step": 4378
+    },
+    {
+      "epoch": 1.7492012779552715,
+      "grad_norm": 1.0507089455768055,
+      "learning_rate": 4.425006480139861e-06,
+      "loss": 0.0925,
+      "step": 4380
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.044726162269031,
+      "learning_rate": 4.420390234792608e-06,
+      "loss": 0.1003,
+      "step": 4382
+    },
+    {
+      "epoch": 1.7507987220447285,
+      "grad_norm": 1.1039342031586883,
+      "learning_rate": 4.415774490173504e-06,
+      "loss": 0.0893,
+      "step": 4384
+    },
+    {
+      "epoch": 1.7515974440894568,
+      "grad_norm": 1.075578365679244,
+      "learning_rate": 4.411159250270119e-06,
+      "loss": 0.0948,
+      "step": 4386
+    },
+    {
+      "epoch": 1.7523961661341851,
+      "grad_norm": 1.0755507149516828,
+      "learning_rate": 4.406544519069582e-06,
+      "loss": 0.0943,
+      "step": 4388
+    },
+    {
+      "epoch": 1.7531948881789137,
+      "grad_norm": 1.2422536273434488,
+      "learning_rate": 4.401930300558588e-06,
+      "loss": 0.1042,
+      "step": 4390
+    },
+    {
+      "epoch": 1.7539936102236422,
+      "grad_norm": 1.125409271455376,
+      "learning_rate": 4.397316598723385e-06,
+      "loss": 0.1008,
+      "step": 4392
+    },
+    {
+      "epoch": 1.7547923322683707,
+      "grad_norm": 0.9912325306555418,
+      "learning_rate": 4.392703417549777e-06,
+      "loss": 0.0893,
+      "step": 4394
+    },
+    {
+      "epoch": 1.755591054313099,
+      "grad_norm": 1.1298243639433942,
+      "learning_rate": 4.388090761023118e-06,
+      "loss": 0.0968,
+      "step": 4396
+    },
+    {
+      "epoch": 1.7563897763578273,
+      "grad_norm": 1.03503405875718,
+      "learning_rate": 4.3834786331283055e-06,
+      "loss": 0.0925,
+      "step": 4398
+    },
+    {
+      "epoch": 1.7571884984025559,
+      "grad_norm": 1.1423391405624483,
+      "learning_rate": 4.3788670378497836e-06,
+      "loss": 0.0989,
+      "step": 4400
+    },
+    {
+      "epoch": 1.7579872204472844,
+      "grad_norm": 1.0654118903753726,
+      "learning_rate": 4.374255979171538e-06,
+      "loss": 0.0993,
+      "step": 4402
+    },
+    {
+      "epoch": 1.758785942492013,
+      "grad_norm": 1.1141658088401454,
+      "learning_rate": 4.369645461077085e-06,
+      "loss": 0.1005,
+      "step": 4404
+    },
+    {
+      "epoch": 1.7595846645367412,
+      "grad_norm": 1.0554546089720755,
+      "learning_rate": 4.365035487549481e-06,
+      "loss": 0.0894,
+      "step": 4406
+    },
+    {
+      "epoch": 1.7603833865814695,
+      "grad_norm": 1.0768929412416388,
+      "learning_rate": 4.360426062571303e-06,
+      "loss": 0.0984,
+      "step": 4408
+    },
+    {
+      "epoch": 1.761182108626198,
+      "grad_norm": 1.0579464025983196,
+      "learning_rate": 4.3558171901246635e-06,
+      "loss": 0.0989,
+      "step": 4410
+    },
+    {
+      "epoch": 1.7619808306709266,
+      "grad_norm": 1.0630018800627314,
+      "learning_rate": 4.351208874191192e-06,
+      "loss": 0.1076,
+      "step": 4412
+    },
+    {
+      "epoch": 1.7627795527156551,
+      "grad_norm": 1.0546218890926364,
+      "learning_rate": 4.346601118752039e-06,
+      "loss": 0.0952,
+      "step": 4414
+    },
+    {
+      "epoch": 1.7635782747603834,
+      "grad_norm": 0.9956147311105759,
+      "learning_rate": 4.341993927787871e-06,
+      "loss": 0.0947,
+      "step": 4416
+    },
+    {
+      "epoch": 1.7643769968051117,
+      "grad_norm": 1.0747196506285701,
+      "learning_rate": 4.337387305278864e-06,
+      "loss": 0.0928,
+      "step": 4418
+    },
+    {
+      "epoch": 1.7651757188498403,
+      "grad_norm": 1.121082140094148,
+      "learning_rate": 4.332781255204708e-06,
+      "loss": 0.0919,
+      "step": 4420
+    },
+    {
+      "epoch": 1.7659744408945688,
+      "grad_norm": 1.1615527137887067,
+      "learning_rate": 4.328175781544593e-06,
+      "loss": 0.108,
+      "step": 4422
+    },
+    {
+      "epoch": 1.766773162939297,
+      "grad_norm": 0.9564313983946037,
+      "learning_rate": 4.323570888277215e-06,
+      "loss": 0.0932,
+      "step": 4424
+    },
+    {
+      "epoch": 1.7675718849840254,
+      "grad_norm": 1.0463601569604801,
+      "learning_rate": 4.318966579380768e-06,
+      "loss": 0.0895,
+      "step": 4426
+    },
+    {
+      "epoch": 1.768370607028754,
+      "grad_norm": 1.163800342765404,
+      "learning_rate": 4.31436285883294e-06,
+      "loss": 0.1033,
+      "step": 4428
+    },
+    {
+      "epoch": 1.7691693290734825,
+      "grad_norm": 1.039700123732381,
+      "learning_rate": 4.3097597306109115e-06,
+      "loss": 0.0958,
+      "step": 4430
+    },
+    {
+      "epoch": 1.769968051118211,
+      "grad_norm": 1.1932778000690207,
+      "learning_rate": 4.305157198691351e-06,
+      "loss": 0.0972,
+      "step": 4432
+    },
+    {
+      "epoch": 1.7707667731629393,
+      "grad_norm": 1.0406189324731365,
+      "learning_rate": 4.30055526705041e-06,
+      "loss": 0.0978,
+      "step": 4434
+    },
+    {
+      "epoch": 1.7715654952076676,
+      "grad_norm": 1.1896973942703604,
+      "learning_rate": 4.2959539396637265e-06,
+      "loss": 0.1038,
+      "step": 4436
+    },
+    {
+      "epoch": 1.7723642172523961,
+      "grad_norm": 1.1009904452623798,
+      "learning_rate": 4.291353220506409e-06,
+      "loss": 0.0933,
+      "step": 4438
+    },
+    {
+      "epoch": 1.7731629392971247,
+      "grad_norm": 1.0740791345413632,
+      "learning_rate": 4.286753113553049e-06,
+      "loss": 0.0901,
+      "step": 4440
+    },
+    {
+      "epoch": 1.7739616613418532,
+      "grad_norm": 1.0663374997120347,
+      "learning_rate": 4.2821536227777016e-06,
+      "loss": 0.0899,
+      "step": 4442
+    },
+    {
+      "epoch": 1.7747603833865815,
+      "grad_norm": 0.9560869355538776,
+      "learning_rate": 4.277554752153895e-06,
+      "loss": 0.0862,
+      "step": 4444
+    },
+    {
+      "epoch": 1.7755591054313098,
+      "grad_norm": 1.1931319093376327,
+      "learning_rate": 4.272956505654616e-06,
+      "loss": 0.1013,
+      "step": 4446
+    },
+    {
+      "epoch": 1.7763578274760383,
+      "grad_norm": 1.1029308559048692,
+      "learning_rate": 4.268358887252322e-06,
+      "loss": 0.0971,
+      "step": 4448
+    },
+    {
+      "epoch": 1.7771565495207668,
+      "grad_norm": 1.0214787393703604,
+      "learning_rate": 4.263761900918916e-06,
+      "loss": 0.0896,
+      "step": 4450
+    },
+    {
+      "epoch": 1.7779552715654952,
+      "grad_norm": 1.1538006513738512,
+      "learning_rate": 4.259165550625765e-06,
+      "loss": 0.1008,
+      "step": 4452
+    },
+    {
+      "epoch": 1.7787539936102237,
+      "grad_norm": 1.1069143300956876,
+      "learning_rate": 4.254569840343677e-06,
+      "loss": 0.0904,
+      "step": 4454
+    },
+    {
+      "epoch": 1.779552715654952,
+      "grad_norm": 1.2455527249765532,
+      "learning_rate": 4.249974774042915e-06,
+      "loss": 0.109,
+      "step": 4456
+    },
+    {
+      "epoch": 1.7803514376996805,
+      "grad_norm": 1.0281866104050703,
+      "learning_rate": 4.245380355693183e-06,
+      "loss": 0.0902,
+      "step": 4458
+    },
+    {
+      "epoch": 1.781150159744409,
+      "grad_norm": 1.0141761099764164,
+      "learning_rate": 4.240786589263623e-06,
+      "loss": 0.0941,
+      "step": 4460
+    },
+    {
+      "epoch": 1.7819488817891374,
+      "grad_norm": 1.0607373591565843,
+      "learning_rate": 4.236193478722818e-06,
+      "loss": 0.0991,
+      "step": 4462
+    },
+    {
+      "epoch": 1.7827476038338657,
+      "grad_norm": 0.9907340835532562,
+      "learning_rate": 4.231601028038781e-06,
+      "loss": 0.0797,
+      "step": 4464
+    },
+    {
+      "epoch": 1.7835463258785942,
+      "grad_norm": 1.112591753336717,
+      "learning_rate": 4.2270092411789555e-06,
+      "loss": 0.0888,
+      "step": 4466
+    },
+    {
+      "epoch": 1.7843450479233227,
+      "grad_norm": 1.2313881450954993,
+      "learning_rate": 4.222418122110212e-06,
+      "loss": 0.1055,
+      "step": 4468
+    },
+    {
+      "epoch": 1.7851437699680512,
+      "grad_norm": 1.0862948713868201,
+      "learning_rate": 4.217827674798845e-06,
+      "loss": 0.098,
+      "step": 4470
+    },
+    {
+      "epoch": 1.7859424920127795,
+      "grad_norm": 1.060155034044239,
+      "learning_rate": 4.2132379032105695e-06,
+      "loss": 0.0996,
+      "step": 4472
+    },
+    {
+      "epoch": 1.7867412140575079,
+      "grad_norm": 1.1021893535744396,
+      "learning_rate": 4.208648811310513e-06,
+      "loss": 0.1053,
+      "step": 4474
+    },
+    {
+      "epoch": 1.7875399361022364,
+      "grad_norm": 1.188590840385766,
+      "learning_rate": 4.204060403063218e-06,
+      "loss": 0.1067,
+      "step": 4476
+    },
+    {
+      "epoch": 1.788338658146965,
+      "grad_norm": 1.1601988700851704,
+      "learning_rate": 4.199472682432637e-06,
+      "loss": 0.1025,
+      "step": 4478
+    },
+    {
+      "epoch": 1.7891373801916934,
+      "grad_norm": 1.2008156142165938,
+      "learning_rate": 4.194885653382128e-06,
+      "loss": 0.1031,
+      "step": 4480
+    },
+    {
+      "epoch": 1.7899361022364217,
+      "grad_norm": 1.1590922201903628,
+      "learning_rate": 4.190299319874449e-06,
+      "loss": 0.0935,
+      "step": 4482
+    },
+    {
+      "epoch": 1.79073482428115,
+      "grad_norm": 1.1080943430270318,
+      "learning_rate": 4.185713685871763e-06,
+      "loss": 0.1121,
+      "step": 4484
+    },
+    {
+      "epoch": 1.7915335463258786,
+      "grad_norm": 1.0680082395002684,
+      "learning_rate": 4.1811287553356214e-06,
+      "loss": 0.0997,
+      "step": 4486
+    },
+    {
+      "epoch": 1.792332268370607,
+      "grad_norm": 1.169219691370761,
+      "learning_rate": 4.176544532226974e-06,
+      "loss": 0.106,
+      "step": 4488
+    },
+    {
+      "epoch": 1.7931309904153354,
+      "grad_norm": 1.0745747283017117,
+      "learning_rate": 4.171961020506158e-06,
+      "loss": 0.0981,
+      "step": 4490
+    },
+    {
+      "epoch": 1.793929712460064,
+      "grad_norm": 1.3455597906264747,
+      "learning_rate": 4.167378224132891e-06,
+      "loss": 0.0982,
+      "step": 4492
+    },
+    {
+      "epoch": 1.7947284345047922,
+      "grad_norm": 1.1972510456665748,
+      "learning_rate": 4.162796147066279e-06,
+      "loss": 0.0978,
+      "step": 4494
+    },
+    {
+      "epoch": 1.7955271565495208,
+      "grad_norm": 1.1197171631008844,
+      "learning_rate": 4.158214793264808e-06,
+      "loss": 0.0837,
+      "step": 4496
+    },
+    {
+      "epoch": 1.7963258785942493,
+      "grad_norm": 0.9880917577387588,
+      "learning_rate": 4.15363416668633e-06,
+      "loss": 0.0884,
+      "step": 4498
+    },
+    {
+      "epoch": 1.7971246006389776,
+      "grad_norm": 1.080269566035693,
+      "learning_rate": 4.149054271288076e-06,
+      "loss": 0.102,
+      "step": 4500
+    },
+    {
+      "epoch": 1.7971246006389776,
+      "eval_loss": 0.14962869882583618,
+      "eval_runtime": 418.7656,
+      "eval_samples_per_second": 42.523,
+      "eval_steps_per_second": 5.316,
+      "step": 4500
+    },
+    {
+      "epoch": 1.797923322683706,
+      "grad_norm": 1.067144176619292,
+      "learning_rate": 4.144475111026643e-06,
+      "loss": 0.104,
+      "step": 4502
+    },
+    {
+      "epoch": 1.7987220447284344,
+      "grad_norm": 1.060342189299496,
+      "learning_rate": 4.139896689857995e-06,
+      "loss": 0.0939,
+      "step": 4504
+    },
+    {
+      "epoch": 1.799520766773163,
+      "grad_norm": 1.059142250496446,
+      "learning_rate": 4.1353190117374545e-06,
+      "loss": 0.0987,
+      "step": 4506
+    },
+    {
+      "epoch": 1.8003194888178915,
+      "grad_norm": 1.0930753497438312,
+      "learning_rate": 4.130742080619704e-06,
+      "loss": 0.0907,
+      "step": 4508
+    },
+    {
+      "epoch": 1.8011182108626198,
+      "grad_norm": 1.0270924676421964,
+      "learning_rate": 4.126165900458781e-06,
+      "loss": 0.09,
+      "step": 4510
+    },
+    {
+      "epoch": 1.8019169329073481,
+      "grad_norm": 1.1110244428747031,
+      "learning_rate": 4.121590475208071e-06,
+      "loss": 0.1013,
+      "step": 4512
+    },
+    {
+      "epoch": 1.8027156549520766,
+      "grad_norm": 1.0236243613153262,
+      "learning_rate": 4.11701580882031e-06,
+      "loss": 0.0976,
+      "step": 4514
+    },
+    {
+      "epoch": 1.8035143769968052,
+      "grad_norm": 1.06524813010489,
+      "learning_rate": 4.1124419052475815e-06,
+      "loss": 0.1,
+      "step": 4516
+    },
+    {
+      "epoch": 1.8043130990415337,
+      "grad_norm": 1.1330875679173593,
+      "learning_rate": 4.107868768441304e-06,
+      "loss": 0.1166,
+      "step": 4518
+    },
+    {
+      "epoch": 1.805111821086262,
+      "grad_norm": 1.4531489997318303,
+      "learning_rate": 4.1032964023522366e-06,
+      "loss": 0.0858,
+      "step": 4520
+    },
+    {
+      "epoch": 1.8059105431309903,
+      "grad_norm": 1.1107724952996303,
+      "learning_rate": 4.098724810930472e-06,
+      "loss": 0.1035,
+      "step": 4522
+    },
+    {
+      "epoch": 1.8067092651757188,
+      "grad_norm": 1.0926573298123923,
+      "learning_rate": 4.0941539981254345e-06,
+      "loss": 0.1026,
+      "step": 4524
+    },
+    {
+      "epoch": 1.8075079872204474,
+      "grad_norm": 0.9721623108703124,
+      "learning_rate": 4.089583967885874e-06,
+      "loss": 0.1025,
+      "step": 4526
+    },
+    {
+      "epoch": 1.8083067092651757,
+      "grad_norm": 1.0370754238710584,
+      "learning_rate": 4.085014724159866e-06,
+      "loss": 0.0942,
+      "step": 4528
+    },
+    {
+      "epoch": 1.8091054313099042,
+      "grad_norm": 1.0940664318814297,
+      "learning_rate": 4.0804462708948076e-06,
+      "loss": 0.0988,
+      "step": 4530
+    },
+    {
+      "epoch": 1.8099041533546325,
+      "grad_norm": 1.0244857172089716,
+      "learning_rate": 4.075878612037408e-06,
+      "loss": 0.089,
+      "step": 4532
+    },
+    {
+      "epoch": 1.810702875399361,
+      "grad_norm": 1.0706540589086802,
+      "learning_rate": 4.071311751533696e-06,
+      "loss": 0.0996,
+      "step": 4534
+    },
+    {
+      "epoch": 1.8115015974440896,
+      "grad_norm": 1.1352354117693362,
+      "learning_rate": 4.066745693329008e-06,
+      "loss": 0.0924,
+      "step": 4536
+    },
+    {
+      "epoch": 1.8123003194888179,
+      "grad_norm": 1.0659040735499372,
+      "learning_rate": 4.062180441367985e-06,
+      "loss": 0.0939,
+      "step": 4538
+    },
+    {
+      "epoch": 1.8130990415335462,
+      "grad_norm": 1.1129307554195655,
+      "learning_rate": 4.057615999594578e-06,
+      "loss": 0.1015,
+      "step": 4540
+    },
+    {
+      "epoch": 1.8138977635782747,
+      "grad_norm": 1.0090716344766517,
+      "learning_rate": 4.053052371952032e-06,
+      "loss": 0.1119,
+      "step": 4542
+    },
+    {
+      "epoch": 1.8146964856230032,
+      "grad_norm": 1.1454677517831415,
+      "learning_rate": 4.0484895623828906e-06,
+      "loss": 0.0984,
+      "step": 4544
+    },
+    {
+      "epoch": 1.8154952076677318,
+      "grad_norm": 1.1153321341980642,
+      "learning_rate": 4.04392757482899e-06,
+      "loss": 0.0977,
+      "step": 4546
+    },
+    {
+      "epoch": 1.81629392971246,
+      "grad_norm": 1.0654841000292614,
+      "learning_rate": 4.039366413231458e-06,
+      "loss": 0.0746,
+      "step": 4548
+    },
+    {
+      "epoch": 1.8170926517571884,
+      "grad_norm": 1.0311164679367486,
+      "learning_rate": 4.034806081530709e-06,
+      "loss": 0.0932,
+      "step": 4550
+    },
+    {
+      "epoch": 1.817891373801917,
+      "grad_norm": 1.044082427416449,
+      "learning_rate": 4.030246583666437e-06,
+      "loss": 0.0967,
+      "step": 4552
+    },
+    {
+      "epoch": 1.8186900958466454,
+      "grad_norm": 1.1455338692396,
+      "learning_rate": 4.0256879235776195e-06,
+      "loss": 0.097,
+      "step": 4554
+    },
+    {
+      "epoch": 1.819488817891374,
+      "grad_norm": 1.1193697932818554,
+      "learning_rate": 4.0211301052025075e-06,
+      "loss": 0.1049,
+      "step": 4556
+    },
+    {
+      "epoch": 1.8202875399361023,
+      "grad_norm": 0.9888093223647104,
+      "learning_rate": 4.016573132478628e-06,
+      "loss": 0.0877,
+      "step": 4558
+    },
+    {
+      "epoch": 1.8210862619808306,
+      "grad_norm": 0.9046844324915236,
+      "learning_rate": 4.012017009342773e-06,
+      "loss": 0.0953,
+      "step": 4560
+    },
+    {
+      "epoch": 1.821884984025559,
+      "grad_norm": 0.8970891044543646,
+      "learning_rate": 4.007461739731003e-06,
+      "loss": 0.0782,
+      "step": 4562
+    },
+    {
+      "epoch": 1.8226837060702876,
+      "grad_norm": 1.0504566189159226,
+      "learning_rate": 4.002907327578644e-06,
+      "loss": 0.096,
+      "step": 4564
+    },
+    {
+      "epoch": 1.823482428115016,
+      "grad_norm": 1.0396603169543621,
+      "learning_rate": 3.9983537768202775e-06,
+      "loss": 0.0966,
+      "step": 4566
+    },
+    {
+      "epoch": 1.8242811501597445,
+      "grad_norm": 1.0690123100147384,
+      "learning_rate": 3.99380109138974e-06,
+      "loss": 0.0781,
+      "step": 4568
+    },
+    {
+      "epoch": 1.8250798722044728,
+      "grad_norm": 1.0129487553480776,
+      "learning_rate": 3.989249275220124e-06,
+      "loss": 0.0884,
+      "step": 4570
+    },
+    {
+      "epoch": 1.8258785942492013,
+      "grad_norm": 1.2215632281871855,
+      "learning_rate": 3.984698332243767e-06,
+      "loss": 0.1092,
+      "step": 4572
+    },
+    {
+      "epoch": 1.8266773162939298,
+      "grad_norm": 1.0991702997365314,
+      "learning_rate": 3.980148266392257e-06,
+      "loss": 0.0998,
+      "step": 4574
+    },
+    {
+      "epoch": 1.8274760383386581,
+      "grad_norm": 1.0590928610197108,
+      "learning_rate": 3.97559908159642e-06,
+      "loss": 0.0884,
+      "step": 4576
+    },
+    {
+      "epoch": 1.8282747603833864,
+      "grad_norm": 1.1608407892475257,
+      "learning_rate": 3.971050781786323e-06,
+      "loss": 0.0974,
+      "step": 4578
+    },
+    {
+      "epoch": 1.829073482428115,
+      "grad_norm": 1.0916397944105056,
+      "learning_rate": 3.966503370891266e-06,
+      "loss": 0.0906,
+      "step": 4580
+    },
+    {
+      "epoch": 1.8298722044728435,
+      "grad_norm": 1.0395162456922145,
+      "learning_rate": 3.961956852839787e-06,
+      "loss": 0.0923,
+      "step": 4582
+    },
+    {
+      "epoch": 1.830670926517572,
+      "grad_norm": 1.068914817127223,
+      "learning_rate": 3.9574112315596425e-06,
+      "loss": 0.1025,
+      "step": 4584
+    },
+    {
+      "epoch": 1.8314696485623003,
+      "grad_norm": 1.0785020160593977,
+      "learning_rate": 3.952866510977827e-06,
+      "loss": 0.0863,
+      "step": 4586
+    },
+    {
+      "epoch": 1.8322683706070286,
+      "grad_norm": 1.1371133744493924,
+      "learning_rate": 3.948322695020546e-06,
+      "loss": 0.0911,
+      "step": 4588
+    },
+    {
+      "epoch": 1.8330670926517572,
+      "grad_norm": 0.9829285073485866,
+      "learning_rate": 3.943779787613231e-06,
+      "loss": 0.0857,
+      "step": 4590
+    },
+    {
+      "epoch": 1.8338658146964857,
+      "grad_norm": 1.5286458187672483,
+      "learning_rate": 3.9392377926805226e-06,
+      "loss": 0.0919,
+      "step": 4592
+    },
+    {
+      "epoch": 1.8346645367412142,
+      "grad_norm": 1.0894667065959676,
+      "learning_rate": 3.934696714146277e-06,
+      "loss": 0.0965,
+      "step": 4594
+    },
+    {
+      "epoch": 1.8354632587859425,
+      "grad_norm": 1.147015209954055,
+      "learning_rate": 3.930156555933557e-06,
+      "loss": 0.0959,
+      "step": 4596
+    },
+    {
+      "epoch": 1.8362619808306708,
+      "grad_norm": 1.0015372010863606,
+      "learning_rate": 3.925617321964632e-06,
+      "loss": 0.0912,
+      "step": 4598
+    },
+    {
+      "epoch": 1.8370607028753994,
+      "grad_norm": 1.0413881872588078,
+      "learning_rate": 3.92107901616097e-06,
+      "loss": 0.1001,
+      "step": 4600
+    },
+    {
+      "epoch": 1.8378594249201279,
+      "grad_norm": 1.1162230686005519,
+      "learning_rate": 3.916541642443242e-06,
+      "loss": 0.0963,
+      "step": 4602
+    },
+    {
+      "epoch": 1.8386581469648562,
+      "grad_norm": 1.0840497131856408,
+      "learning_rate": 3.912005204731307e-06,
+      "loss": 0.1011,
+      "step": 4604
+    },
+    {
+      "epoch": 1.8394568690095847,
+      "grad_norm": 1.0114188708406382,
+      "learning_rate": 3.907469706944222e-06,
+      "loss": 0.0894,
+      "step": 4606
+    },
+    {
+      "epoch": 1.840255591054313,
+      "grad_norm": 1.0694376480162955,
+      "learning_rate": 3.9029351530002264e-06,
+      "loss": 0.086,
+      "step": 4608
+    },
+    {
+      "epoch": 1.8410543130990416,
+      "grad_norm": 1.3957274955440564,
+      "learning_rate": 3.898401546816752e-06,
+      "loss": 0.0927,
+      "step": 4610
+    },
+    {
+      "epoch": 1.84185303514377,
+      "grad_norm": 0.9916548350535773,
+      "learning_rate": 3.8938688923104015e-06,
+      "loss": 0.0877,
+      "step": 4612
+    },
+    {
+      "epoch": 1.8426517571884984,
+      "grad_norm": 1.0958198194521465,
+      "learning_rate": 3.8893371933969644e-06,
+      "loss": 0.0921,
+      "step": 4614
+    },
+    {
+      "epoch": 1.8434504792332267,
+      "grad_norm": 0.9890953898234616,
+      "learning_rate": 3.884806453991399e-06,
+      "loss": 0.0898,
+      "step": 4616
+    },
+    {
+      "epoch": 1.8442492012779552,
+      "grad_norm": 1.099760617918534,
+      "learning_rate": 3.880276678007838e-06,
+      "loss": 0.0967,
+      "step": 4618
+    },
+    {
+      "epoch": 1.8450479233226837,
+      "grad_norm": 1.1341946755721188,
+      "learning_rate": 3.875747869359578e-06,
+      "loss": 0.0911,
+      "step": 4620
+    },
+    {
+      "epoch": 1.8458466453674123,
+      "grad_norm": 1.0924073653891035,
+      "learning_rate": 3.871220031959085e-06,
+      "loss": 0.0991,
+      "step": 4622
+    },
+    {
+      "epoch": 1.8466453674121406,
+      "grad_norm": 1.1994386671625121,
+      "learning_rate": 3.866693169717982e-06,
+      "loss": 0.1012,
+      "step": 4624
+    },
+    {
+      "epoch": 1.8474440894568689,
+      "grad_norm": 1.1087200452695056,
+      "learning_rate": 3.8621672865470505e-06,
+      "loss": 0.0901,
+      "step": 4626
+    },
+    {
+      "epoch": 1.8482428115015974,
+      "grad_norm": 1.1622238598835266,
+      "learning_rate": 3.8576423863562285e-06,
+      "loss": 0.0949,
+      "step": 4628
+    },
+    {
+      "epoch": 1.849041533546326,
+      "grad_norm": 1.031896254949219,
+      "learning_rate": 3.8531184730546e-06,
+      "loss": 0.0986,
+      "step": 4630
+    },
+    {
+      "epoch": 1.8498402555910545,
+      "grad_norm": 1.1267405061140343,
+      "learning_rate": 3.848595550550401e-06,
+      "loss": 0.0963,
+      "step": 4632
+    },
+    {
+      "epoch": 1.8506389776357828,
+      "grad_norm": 1.091392917402568,
+      "learning_rate": 3.84407362275101e-06,
+      "loss": 0.0952,
+      "step": 4634
+    },
+    {
+      "epoch": 1.851437699680511,
+      "grad_norm": 1.07201414860687,
+      "learning_rate": 3.839552693562946e-06,
+      "loss": 0.1002,
+      "step": 4636
+    },
+    {
+      "epoch": 1.8522364217252396,
+      "grad_norm": 1.0286376655218412,
+      "learning_rate": 3.835032766891865e-06,
+      "loss": 0.0948,
+      "step": 4638
+    },
+    {
+      "epoch": 1.8530351437699681,
+      "grad_norm": 1.0789695524482392,
+      "learning_rate": 3.830513846642556e-06,
+      "loss": 0.0967,
+      "step": 4640
+    },
+    {
+      "epoch": 1.8538338658146964,
+      "grad_norm": 1.1844906017376726,
+      "learning_rate": 3.825995936718942e-06,
+      "loss": 0.0951,
+      "step": 4642
+    },
+    {
+      "epoch": 1.854632587859425,
+      "grad_norm": 0.9957944115590945,
+      "learning_rate": 3.821479041024069e-06,
+      "loss": 0.0902,
+      "step": 4644
+    },
+    {
+      "epoch": 1.8554313099041533,
+      "grad_norm": 1.0579821850330315,
+      "learning_rate": 3.816963163460109e-06,
+      "loss": 0.1,
+      "step": 4646
+    },
+    {
+      "epoch": 1.8562300319488818,
+      "grad_norm": 1.080547600595392,
+      "learning_rate": 3.8124483079283546e-06,
+      "loss": 0.1039,
+      "step": 4648
+    },
+    {
+      "epoch": 1.8570287539936103,
+      "grad_norm": 1.0913158139520798,
+      "learning_rate": 3.8079344783292145e-06,
+      "loss": 0.1021,
+      "step": 4650
+    },
+    {
+      "epoch": 1.8578274760383386,
+      "grad_norm": 1.1039317269281381,
+      "learning_rate": 3.803421678562213e-06,
+      "loss": 0.096,
+      "step": 4652
+    },
+    {
+      "epoch": 1.858626198083067,
+      "grad_norm": 1.0837414701686472,
+      "learning_rate": 3.79890991252598e-06,
+      "loss": 0.0931,
+      "step": 4654
+    },
+    {
+      "epoch": 1.8594249201277955,
+      "grad_norm": 0.9817886186922399,
+      "learning_rate": 3.7943991841182586e-06,
+      "loss": 0.0918,
+      "step": 4656
+    },
+    {
+      "epoch": 1.860223642172524,
+      "grad_norm": 1.2753358547680838,
+      "learning_rate": 3.7898894972358934e-06,
+      "loss": 0.0879,
+      "step": 4658
+    },
+    {
+      "epoch": 1.8610223642172525,
+      "grad_norm": 1.0132549906162278,
+      "learning_rate": 3.7853808557748263e-06,
+      "loss": 0.0982,
+      "step": 4660
+    },
+    {
+      "epoch": 1.8618210862619808,
+      "grad_norm": 1.0604493491043667,
+      "learning_rate": 3.7808732636300987e-06,
+      "loss": 0.0909,
+      "step": 4662
+    },
+    {
+      "epoch": 1.8626198083067091,
+      "grad_norm": 1.0110270014624545,
+      "learning_rate": 3.7763667246958447e-06,
+      "loss": 0.0913,
+      "step": 4664
+    },
+    {
+      "epoch": 1.8634185303514377,
+      "grad_norm": 1.1816528766134167,
+      "learning_rate": 3.771861242865288e-06,
+      "loss": 0.0987,
+      "step": 4666
+    },
+    {
+      "epoch": 1.8642172523961662,
+      "grad_norm": 1.1276543836075537,
+      "learning_rate": 3.767356822030742e-06,
+      "loss": 0.1006,
+      "step": 4668
+    },
+    {
+      "epoch": 1.8650159744408947,
+      "grad_norm": 1.0029943146012643,
+      "learning_rate": 3.7628534660835996e-06,
+      "loss": 0.0925,
+      "step": 4670
+    },
+    {
+      "epoch": 1.865814696485623,
+      "grad_norm": 1.1636739447782507,
+      "learning_rate": 3.758351178914336e-06,
+      "loss": 0.0946,
+      "step": 4672
+    },
+    {
+      "epoch": 1.8666134185303513,
+      "grad_norm": 1.0912114279896796,
+      "learning_rate": 3.753849964412502e-06,
+      "loss": 0.09,
+      "step": 4674
+    },
+    {
+      "epoch": 1.8674121405750799,
+      "grad_norm": 1.2020359973073858,
+      "learning_rate": 3.749349826466724e-06,
+      "loss": 0.1037,
+      "step": 4676
+    },
+    {
+      "epoch": 1.8682108626198084,
+      "grad_norm": 0.9661948745934541,
+      "learning_rate": 3.744850768964692e-06,
+      "loss": 0.0862,
+      "step": 4678
+    },
+    {
+      "epoch": 1.8690095846645367,
+      "grad_norm": 1.0588016583262083,
+      "learning_rate": 3.7403527957931716e-06,
+      "loss": 0.0929,
+      "step": 4680
+    },
+    {
+      "epoch": 1.869808306709265,
+      "grad_norm": 1.095002175972765,
+      "learning_rate": 3.7358559108379867e-06,
+      "loss": 0.0982,
+      "step": 4682
+    },
+    {
+      "epoch": 1.8706070287539935,
+      "grad_norm": 1.0767223792716407,
+      "learning_rate": 3.731360117984022e-06,
+      "loss": 0.0911,
+      "step": 4684
+    },
+    {
+      "epoch": 1.871405750798722,
+      "grad_norm": 1.0769547783245055,
+      "learning_rate": 3.7268654211152156e-06,
+      "loss": 0.0931,
+      "step": 4686
+    },
+    {
+      "epoch": 1.8722044728434506,
+      "grad_norm": 1.1230982099429845,
+      "learning_rate": 3.7223718241145646e-06,
+      "loss": 0.0946,
+      "step": 4688
+    },
+    {
+      "epoch": 1.873003194888179,
+      "grad_norm": 1.0232620319916697,
+      "learning_rate": 3.71787933086411e-06,
+      "loss": 0.0863,
+      "step": 4690
+    },
+    {
+      "epoch": 1.8738019169329072,
+      "grad_norm": 1.1155033054935346,
+      "learning_rate": 3.713387945244945e-06,
+      "loss": 0.1052,
+      "step": 4692
+    },
+    {
+      "epoch": 1.8746006389776357,
+      "grad_norm": 1.086463915995651,
+      "learning_rate": 3.7088976711372006e-06,
+      "loss": 0.0894,
+      "step": 4694
+    },
+    {
+      "epoch": 1.8753993610223643,
+      "grad_norm": 1.1158786507370728,
+      "learning_rate": 3.7044085124200517e-06,
+      "loss": 0.0897,
+      "step": 4696
+    },
+    {
+      "epoch": 1.8761980830670928,
+      "grad_norm": 1.1044666274555124,
+      "learning_rate": 3.6999204729717057e-06,
+      "loss": 0.0935,
+      "step": 4698
+    },
+    {
+      "epoch": 1.876996805111821,
+      "grad_norm": 1.0364347122571194,
+      "learning_rate": 3.695433556669406e-06,
+      "loss": 0.0881,
+      "step": 4700
+    },
+    {
+      "epoch": 1.8777955271565494,
+      "grad_norm": 1.179238289944297,
+      "learning_rate": 3.690947767389426e-06,
+      "loss": 0.0903,
+      "step": 4702
+    },
+    {
+      "epoch": 1.878594249201278,
+      "grad_norm": 1.2284443916952599,
+      "learning_rate": 3.6864631090070656e-06,
+      "loss": 0.1027,
+      "step": 4704
+    },
+    {
+      "epoch": 1.8793929712460065,
+      "grad_norm": 1.1165613439165785,
+      "learning_rate": 3.6819795853966435e-06,
+      "loss": 0.0908,
+      "step": 4706
+    },
+    {
+      "epoch": 1.880191693290735,
+      "grad_norm": 1.1221181152137276,
+      "learning_rate": 3.6774972004315035e-06,
+      "loss": 0.0994,
+      "step": 4708
+    },
+    {
+      "epoch": 1.8809904153354633,
+      "grad_norm": 1.120777637697723,
+      "learning_rate": 3.6730159579840007e-06,
+      "loss": 0.1016,
+      "step": 4710
+    },
+    {
+      "epoch": 1.8817891373801916,
+      "grad_norm": 1.0499328599769773,
+      "learning_rate": 3.668535861925509e-06,
+      "loss": 0.1007,
+      "step": 4712
+    },
+    {
+      "epoch": 1.8825878594249201,
+      "grad_norm": 1.0952348015619684,
+      "learning_rate": 3.6640569161264055e-06,
+      "loss": 0.1008,
+      "step": 4714
+    },
+    {
+      "epoch": 1.8833865814696487,
+      "grad_norm": 1.0556886162538723,
+      "learning_rate": 3.6595791244560795e-06,
+      "loss": 0.0909,
+      "step": 4716
+    },
+    {
+      "epoch": 1.884185303514377,
+      "grad_norm": 1.0779044498229438,
+      "learning_rate": 3.655102490782918e-06,
+      "loss": 0.1072,
+      "step": 4718
+    },
+    {
+      "epoch": 1.8849840255591053,
+      "grad_norm": 1.0237827262029766,
+      "learning_rate": 3.650627018974312e-06,
+      "loss": 0.0964,
+      "step": 4720
+    },
+    {
+      "epoch": 1.8857827476038338,
+      "grad_norm": 0.9712789192282018,
+      "learning_rate": 3.6461527128966457e-06,
+      "loss": 0.0942,
+      "step": 4722
+    },
+    {
+      "epoch": 1.8865814696485623,
+      "grad_norm": 1.1258507713014763,
+      "learning_rate": 3.6416795764152967e-06,
+      "loss": 0.104,
+      "step": 4724
+    },
+    {
+      "epoch": 1.8873801916932909,
+      "grad_norm": 0.9821860855241218,
+      "learning_rate": 3.6372076133946353e-06,
+      "loss": 0.087,
+      "step": 4726
+    },
+    {
+      "epoch": 1.8881789137380192,
+      "grad_norm": 1.0700005208247971,
+      "learning_rate": 3.632736827698015e-06,
+      "loss": 0.0883,
+      "step": 4728
+    },
+    {
+      "epoch": 1.8889776357827475,
+      "grad_norm": 0.9731935760176809,
+      "learning_rate": 3.6282672231877714e-06,
+      "loss": 0.0883,
+      "step": 4730
+    },
+    {
+      "epoch": 1.889776357827476,
+      "grad_norm": 1.045015911757986,
+      "learning_rate": 3.623798803725223e-06,
+      "loss": 0.097,
+      "step": 4732
+    },
+    {
+      "epoch": 1.8905750798722045,
+      "grad_norm": 1.218717767375124,
+      "learning_rate": 3.619331573170661e-06,
+      "loss": 0.0921,
+      "step": 4734
+    },
+    {
+      "epoch": 1.891373801916933,
+      "grad_norm": 1.058626369915082,
+      "learning_rate": 3.6148655353833518e-06,
+      "loss": 0.0856,
+      "step": 4736
+    },
+    {
+      "epoch": 1.8921725239616614,
+      "grad_norm": 1.0773766259028001,
+      "learning_rate": 3.6104006942215296e-06,
+      "loss": 0.0909,
+      "step": 4738
+    },
+    {
+      "epoch": 1.8929712460063897,
+      "grad_norm": 1.0619367090404177,
+      "learning_rate": 3.605937053542398e-06,
+      "loss": 0.0854,
+      "step": 4740
+    },
+    {
+      "epoch": 1.8937699680511182,
+      "grad_norm": 1.0878885631245991,
+      "learning_rate": 3.6014746172021197e-06,
+      "loss": 0.096,
+      "step": 4742
+    },
+    {
+      "epoch": 1.8945686900958467,
+      "grad_norm": 1.0235847877959412,
+      "learning_rate": 3.5970133890558184e-06,
+      "loss": 0.0826,
+      "step": 4744
+    },
+    {
+      "epoch": 1.895367412140575,
+      "grad_norm": 1.1515886076002215,
+      "learning_rate": 3.5925533729575745e-06,
+      "loss": 0.0891,
+      "step": 4746
+    },
+    {
+      "epoch": 1.8961661341853036,
+      "grad_norm": 1.0787838461269919,
+      "learning_rate": 3.588094572760423e-06,
+      "loss": 0.0947,
+      "step": 4748
+    },
+    {
+      "epoch": 1.8969648562300319,
+      "grad_norm": 1.1230844413759828,
+      "learning_rate": 3.583636992316345e-06,
+      "loss": 0.0902,
+      "step": 4750
+    },
+    {
+      "epoch": 1.8977635782747604,
+      "grad_norm": 1.0936152470611447,
+      "learning_rate": 3.5791806354762702e-06,
+      "loss": 0.0998,
+      "step": 4752
+    },
+    {
+      "epoch": 1.898562300319489,
+      "grad_norm": 1.2224698612193565,
+      "learning_rate": 3.5747255060900687e-06,
+      "loss": 0.1032,
+      "step": 4754
+    },
+    {
+      "epoch": 1.8993610223642172,
+      "grad_norm": 1.0690769327967091,
+      "learning_rate": 3.5702716080065546e-06,
+      "loss": 0.0856,
+      "step": 4756
+    },
+    {
+      "epoch": 1.9001597444089455,
+      "grad_norm": 1.0519872936846355,
+      "learning_rate": 3.5658189450734727e-06,
+      "loss": 0.088,
+      "step": 4758
+    },
+    {
+      "epoch": 1.900958466453674,
+      "grad_norm": 1.1511736996930644,
+      "learning_rate": 3.5613675211375066e-06,
+      "loss": 0.093,
+      "step": 4760
+    },
+    {
+      "epoch": 1.9017571884984026,
+      "grad_norm": 1.1254396775195248,
+      "learning_rate": 3.5569173400442634e-06,
+      "loss": 0.086,
+      "step": 4762
+    },
+    {
+      "epoch": 1.9025559105431311,
+      "grad_norm": 1.2048144094026239,
+      "learning_rate": 3.5524684056382824e-06,
+      "loss": 0.096,
+      "step": 4764
+    },
+    {
+      "epoch": 1.9033546325878594,
+      "grad_norm": 1.0379587437323166,
+      "learning_rate": 3.5480207217630224e-06,
+      "loss": 0.0929,
+      "step": 4766
+    },
+    {
+      "epoch": 1.9041533546325877,
+      "grad_norm": 1.0861096082948125,
+      "learning_rate": 3.5435742922608618e-06,
+      "loss": 0.0843,
+      "step": 4768
+    },
+    {
+      "epoch": 1.9049520766773163,
+      "grad_norm": 1.094770765692993,
+      "learning_rate": 3.539129120973095e-06,
+      "loss": 0.0913,
+      "step": 4770
+    },
+    {
+      "epoch": 1.9057507987220448,
+      "grad_norm": 1.0808303484556492,
+      "learning_rate": 3.534685211739935e-06,
+      "loss": 0.084,
+      "step": 4772
+    },
+    {
+      "epoch": 1.9065495207667733,
+      "grad_norm": 1.0531941918931411,
+      "learning_rate": 3.5302425684004957e-06,
+      "loss": 0.0837,
+      "step": 4774
+    },
+    {
+      "epoch": 1.9073482428115016,
+      "grad_norm": 1.0972839581319511,
+      "learning_rate": 3.525801194792805e-06,
+      "loss": 0.0969,
+      "step": 4776
+    },
+    {
+      "epoch": 1.90814696485623,
+      "grad_norm": 1.1088742467231305,
+      "learning_rate": 3.521361094753788e-06,
+      "loss": 0.09,
+      "step": 4778
+    },
+    {
+      "epoch": 1.9089456869009584,
+      "grad_norm": 1.062941595145544,
+      "learning_rate": 3.516922272119274e-06,
+      "loss": 0.0826,
+      "step": 4780
+    },
+    {
+      "epoch": 1.909744408945687,
+      "grad_norm": 1.01692209679147,
+      "learning_rate": 3.5124847307239863e-06,
+      "loss": 0.0857,
+      "step": 4782
+    },
+    {
+      "epoch": 1.9105431309904153,
+      "grad_norm": 1.2055133548353663,
+      "learning_rate": 3.508048474401541e-06,
+      "loss": 0.096,
+      "step": 4784
+    },
+    {
+      "epoch": 1.9113418530351438,
+      "grad_norm": 1.001313631088606,
+      "learning_rate": 3.503613506984447e-06,
+      "loss": 0.0884,
+      "step": 4786
+    },
+    {
+      "epoch": 1.9121405750798721,
+      "grad_norm": 1.1432556607777271,
+      "learning_rate": 3.499179832304096e-06,
+      "loss": 0.0909,
+      "step": 4788
+    },
+    {
+      "epoch": 1.9129392971246006,
+      "grad_norm": 1.1768617881519154,
+      "learning_rate": 3.4947474541907655e-06,
+      "loss": 0.0895,
+      "step": 4790
+    },
+    {
+      "epoch": 1.9137380191693292,
+      "grad_norm": 1.054267796873096,
+      "learning_rate": 3.4903163764736104e-06,
+      "loss": 0.0944,
+      "step": 4792
+    },
+    {
+      "epoch": 1.9145367412140575,
+      "grad_norm": 1.1032448450072194,
+      "learning_rate": 3.4858866029806658e-06,
+      "loss": 0.0956,
+      "step": 4794
+    },
+    {
+      "epoch": 1.9153354632587858,
+      "grad_norm": 0.9183906781681136,
+      "learning_rate": 3.4814581375388384e-06,
+      "loss": 0.0838,
+      "step": 4796
+    },
+    {
+      "epoch": 1.9161341853035143,
+      "grad_norm": 1.1483868020594459,
+      "learning_rate": 3.4770309839739026e-06,
+      "loss": 0.0913,
+      "step": 4798
+    },
+    {
+      "epoch": 1.9169329073482428,
+      "grad_norm": 1.0510791282744207,
+      "learning_rate": 3.4726051461105016e-06,
+      "loss": 0.0895,
+      "step": 4800
+    },
+    {
+      "epoch": 1.9177316293929714,
+      "grad_norm": 1.0681148740204216,
+      "learning_rate": 3.468180627772144e-06,
+      "loss": 0.0901,
+      "step": 4802
+    },
+    {
+      "epoch": 1.9185303514376997,
+      "grad_norm": 1.056725620741662,
+      "learning_rate": 3.4637574327811934e-06,
+      "loss": 0.087,
+      "step": 4804
+    },
+    {
+      "epoch": 1.919329073482428,
+      "grad_norm": 1.1165517293384117,
+      "learning_rate": 3.459335564958875e-06,
+      "loss": 0.0949,
+      "step": 4806
+    },
+    {
+      "epoch": 1.9201277955271565,
+      "grad_norm": 1.1714069822717896,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.1003,
+      "step": 4808
+    },
+    {
+      "epoch": 1.920926517571885,
+      "grad_norm": 1.0857128594280534,
+      "learning_rate": 3.4504958260992877e-06,
+      "loss": 0.0917,
+      "step": 4810
+    },
+    {
+      "epoch": 1.9217252396166136,
+      "grad_norm": 0.9514303291715913,
+      "learning_rate": 3.4460779626987186e-06,
+      "loss": 0.081,
+      "step": 4812
+    },
+    {
+      "epoch": 1.9225239616613419,
+      "grad_norm": 1.1178021513245053,
+      "learning_rate": 3.441661441740176e-06,
+      "loss": 0.0945,
+      "step": 4814
+    },
+    {
+      "epoch": 1.9233226837060702,
+      "grad_norm": 1.0436166022295423,
+      "learning_rate": 3.437246267039115e-06,
+      "loss": 0.0939,
+      "step": 4816
+    },
+    {
+      "epoch": 1.9241214057507987,
+      "grad_norm": 0.9558565581069985,
+      "learning_rate": 3.4328324424098315e-06,
+      "loss": 0.0866,
+      "step": 4818
+    },
+    {
+      "epoch": 1.9249201277955272,
+      "grad_norm": 1.0690618672851848,
+      "learning_rate": 3.4284199716654526e-06,
+      "loss": 0.0971,
+      "step": 4820
+    },
+    {
+      "epoch": 1.9257188498402555,
+      "grad_norm": 1.0330544019396488,
+      "learning_rate": 3.424008858617939e-06,
+      "loss": 0.0877,
+      "step": 4822
+    },
+    {
+      "epoch": 1.926517571884984,
+      "grad_norm": 1.090972476610282,
+      "learning_rate": 3.419599107078073e-06,
+      "loss": 0.0933,
+      "step": 4824
+    },
+    {
+      "epoch": 1.9273162939297124,
+      "grad_norm": 1.1389205515436118,
+      "learning_rate": 3.4151907208554657e-06,
+      "loss": 0.0914,
+      "step": 4826
+    },
+    {
+      "epoch": 1.928115015974441,
+      "grad_norm": 1.0521512673067093,
+      "learning_rate": 3.4107837037585463e-06,
+      "loss": 0.0956,
+      "step": 4828
+    },
+    {
+      "epoch": 1.9289137380191694,
+      "grad_norm": 0.9943276734942584,
+      "learning_rate": 3.4063780595945627e-06,
+      "loss": 0.0971,
+      "step": 4830
+    },
+    {
+      "epoch": 1.9297124600638977,
+      "grad_norm": 1.1979073104114557,
+      "learning_rate": 3.401973792169574e-06,
+      "loss": 0.0929,
+      "step": 4832
+    },
+    {
+      "epoch": 1.930511182108626,
+      "grad_norm": 0.9579355850140132,
+      "learning_rate": 3.397570905288453e-06,
+      "loss": 0.0895,
+      "step": 4834
+    },
+    {
+      "epoch": 1.9313099041533546,
+      "grad_norm": 1.0976423710464434,
+      "learning_rate": 3.393169402754878e-06,
+      "loss": 0.0844,
+      "step": 4836
+    },
+    {
+      "epoch": 1.932108626198083,
+      "grad_norm": 1.1417293656435412,
+      "learning_rate": 3.388769288371333e-06,
+      "loss": 0.0954,
+      "step": 4838
+    },
+    {
+      "epoch": 1.9329073482428116,
+      "grad_norm": 0.9657207657700554,
+      "learning_rate": 3.384370565939098e-06,
+      "loss": 0.08,
+      "step": 4840
+    },
+    {
+      "epoch": 1.93370607028754,
+      "grad_norm": 1.122410966064945,
+      "learning_rate": 3.3799732392582598e-06,
+      "loss": 0.1073,
+      "step": 4842
+    },
+    {
+      "epoch": 1.9345047923322682,
+      "grad_norm": 0.9499390486261202,
+      "learning_rate": 3.375577312127689e-06,
+      "loss": 0.081,
+      "step": 4844
+    },
+    {
+      "epoch": 1.9353035143769968,
+      "grad_norm": 1.089004347076352,
+      "learning_rate": 3.3711827883450552e-06,
+      "loss": 0.0916,
+      "step": 4846
+    },
+    {
+      "epoch": 1.9361022364217253,
+      "grad_norm": 1.121477411735401,
+      "learning_rate": 3.3667896717068105e-06,
+      "loss": 0.0944,
+      "step": 4848
+    },
+    {
+      "epoch": 1.9369009584664538,
+      "grad_norm": 1.137024861471035,
+      "learning_rate": 3.3623979660081944e-06,
+      "loss": 0.0898,
+      "step": 4850
+    },
+    {
+      "epoch": 1.9376996805111821,
+      "grad_norm": 1.036107942325039,
+      "learning_rate": 3.3580076750432244e-06,
+      "loss": 0.1024,
+      "step": 4852
+    },
+    {
+      "epoch": 1.9384984025559104,
+      "grad_norm": 1.0079391443696057,
+      "learning_rate": 3.3536188026047e-06,
+      "loss": 0.0806,
+      "step": 4854
+    },
+    {
+      "epoch": 1.939297124600639,
+      "grad_norm": 1.0869459285245229,
+      "learning_rate": 3.34923135248419e-06,
+      "loss": 0.0918,
+      "step": 4856
+    },
+    {
+      "epoch": 1.9400958466453675,
+      "grad_norm": 1.1976937501317675,
+      "learning_rate": 3.3448453284720407e-06,
+      "loss": 0.0992,
+      "step": 4858
+    },
+    {
+      "epoch": 1.9408945686900958,
+      "grad_norm": 1.1088825356732732,
+      "learning_rate": 3.340460734357359e-06,
+      "loss": 0.0986,
+      "step": 4860
+    },
+    {
+      "epoch": 1.9416932907348243,
+      "grad_norm": 1.03132569544229,
+      "learning_rate": 3.336077573928023e-06,
+      "loss": 0.0864,
+      "step": 4862
+    },
+    {
+      "epoch": 1.9424920127795526,
+      "grad_norm": 1.029156495025623,
+      "learning_rate": 3.33169585097067e-06,
+      "loss": 0.0865,
+      "step": 4864
+    },
+    {
+      "epoch": 1.9432907348242812,
+      "grad_norm": 1.1169233865438082,
+      "learning_rate": 3.3273155692706956e-06,
+      "loss": 0.087,
+      "step": 4866
+    },
+    {
+      "epoch": 1.9440894568690097,
+      "grad_norm": 1.0237220064497436,
+      "learning_rate": 3.3229367326122475e-06,
+      "loss": 0.088,
+      "step": 4868
+    },
+    {
+      "epoch": 1.944888178913738,
+      "grad_norm": 1.0716607652719201,
+      "learning_rate": 3.318559344778231e-06,
+      "loss": 0.094,
+      "step": 4870
+    },
+    {
+      "epoch": 1.9456869009584663,
+      "grad_norm": 0.9965570911044683,
+      "learning_rate": 3.314183409550293e-06,
+      "loss": 0.0913,
+      "step": 4872
+    },
+    {
+      "epoch": 1.9464856230031948,
+      "grad_norm": 1.0605235496788143,
+      "learning_rate": 3.3098089307088307e-06,
+      "loss": 0.088,
+      "step": 4874
+    },
+    {
+      "epoch": 1.9472843450479234,
+      "grad_norm": 1.0509076016295873,
+      "learning_rate": 3.3054359120329788e-06,
+      "loss": 0.0967,
+      "step": 4876
+    },
+    {
+      "epoch": 1.9480830670926519,
+      "grad_norm": 1.0048257649363173,
+      "learning_rate": 3.301064357300615e-06,
+      "loss": 0.0924,
+      "step": 4878
+    },
+    {
+      "epoch": 1.9488817891373802,
+      "grad_norm": 1.1009166155154577,
+      "learning_rate": 3.2966942702883494e-06,
+      "loss": 0.0979,
+      "step": 4880
+    },
+    {
+      "epoch": 1.9496805111821085,
+      "grad_norm": 1.123335138251925,
+      "learning_rate": 3.2923256547715245e-06,
+      "loss": 0.0952,
+      "step": 4882
+    },
+    {
+      "epoch": 1.950479233226837,
+      "grad_norm": 1.0429249357254708,
+      "learning_rate": 3.287958514524212e-06,
+      "loss": 0.0868,
+      "step": 4884
+    },
+    {
+      "epoch": 1.9512779552715656,
+      "grad_norm": 1.0420461671724415,
+      "learning_rate": 3.2835928533192086e-06,
+      "loss": 0.091,
+      "step": 4886
+    },
+    {
+      "epoch": 1.952076677316294,
+      "grad_norm": 1.066480483177895,
+      "learning_rate": 3.279228674928035e-06,
+      "loss": 0.0911,
+      "step": 4888
+    },
+    {
+      "epoch": 1.9528753993610224,
+      "grad_norm": 1.034614202531424,
+      "learning_rate": 3.2748659831209293e-06,
+      "loss": 0.0925,
+      "step": 4890
+    },
+    {
+      "epoch": 1.9536741214057507,
+      "grad_norm": 1.010806358143463,
+      "learning_rate": 3.270504781666845e-06,
+      "loss": 0.0948,
+      "step": 4892
+    },
+    {
+      "epoch": 1.9544728434504792,
+      "grad_norm": 0.9676259211028123,
+      "learning_rate": 3.2661450743334495e-06,
+      "loss": 0.0838,
+      "step": 4894
+    },
+    {
+      "epoch": 1.9552715654952078,
+      "grad_norm": 1.9272872858361931,
+      "learning_rate": 3.261786864887117e-06,
+      "loss": 0.0941,
+      "step": 4896
+    },
+    {
+      "epoch": 1.956070287539936,
+      "grad_norm": 1.1661864224993708,
+      "learning_rate": 3.2574301570929313e-06,
+      "loss": 0.0946,
+      "step": 4898
+    },
+    {
+      "epoch": 1.9568690095846646,
+      "grad_norm": 1.0658970850145277,
+      "learning_rate": 3.2530749547146745e-06,
+      "loss": 0.0846,
+      "step": 4900
+    },
+    {
+      "epoch": 1.957667731629393,
+      "grad_norm": 1.1441635938072476,
+      "learning_rate": 3.2487212615148316e-06,
+      "loss": 0.0957,
+      "step": 4902
+    },
+    {
+      "epoch": 1.9584664536741214,
+      "grad_norm": 1.0929232430584095,
+      "learning_rate": 3.244369081254585e-06,
+      "loss": 0.0881,
+      "step": 4904
+    },
+    {
+      "epoch": 1.95926517571885,
+      "grad_norm": 0.9707906667424272,
+      "learning_rate": 3.240018417693803e-06,
+      "loss": 0.0854,
+      "step": 4906
+    },
+    {
+      "epoch": 1.9600638977635783,
+      "grad_norm": 1.166565929217826,
+      "learning_rate": 3.235669274591051e-06,
+      "loss": 0.0952,
+      "step": 4908
+    },
+    {
+      "epoch": 1.9608626198083066,
+      "grad_norm": 0.9958821823082478,
+      "learning_rate": 3.231321655703581e-06,
+      "loss": 0.0896,
+      "step": 4910
+    },
+    {
+      "epoch": 1.961661341853035,
+      "grad_norm": 1.16345333204102,
+      "learning_rate": 3.226975564787322e-06,
+      "loss": 0.0967,
+      "step": 4912
+    },
+    {
+      "epoch": 1.9624600638977636,
+      "grad_norm": 1.0194684910483922,
+      "learning_rate": 3.222631005596888e-06,
+      "loss": 0.0792,
+      "step": 4914
+    },
+    {
+      "epoch": 1.9632587859424921,
+      "grad_norm": 1.1421002598232535,
+      "learning_rate": 3.218287981885567e-06,
+      "loss": 0.0998,
+      "step": 4916
+    },
+    {
+      "epoch": 1.9640575079872205,
+      "grad_norm": 1.0778986286246897,
+      "learning_rate": 3.2139464974053225e-06,
+      "loss": 0.0979,
+      "step": 4918
+    },
+    {
+      "epoch": 1.9648562300319488,
+      "grad_norm": 1.071349946349977,
+      "learning_rate": 3.209606555906788e-06,
+      "loss": 0.0792,
+      "step": 4920
+    },
+    {
+      "epoch": 1.9656549520766773,
+      "grad_norm": 0.9608643299126582,
+      "learning_rate": 3.2052681611392616e-06,
+      "loss": 0.0946,
+      "step": 4922
+    },
+    {
+      "epoch": 1.9664536741214058,
+      "grad_norm": 1.0944937155116632,
+      "learning_rate": 3.20093131685071e-06,
+      "loss": 0.094,
+      "step": 4924
+    },
+    {
+      "epoch": 1.9672523961661343,
+      "grad_norm": 1.045526176848772,
+      "learning_rate": 3.1965960267877544e-06,
+      "loss": 0.0986,
+      "step": 4926
+    },
+    {
+      "epoch": 1.9680511182108626,
+      "grad_norm": 1.2126133137340172,
+      "learning_rate": 3.192262294695679e-06,
+      "loss": 0.1038,
+      "step": 4928
+    },
+    {
+      "epoch": 1.968849840255591,
+      "grad_norm": 1.1192306783949446,
+      "learning_rate": 3.187930124318417e-06,
+      "loss": 0.0928,
+      "step": 4930
+    },
+    {
+      "epoch": 1.9696485623003195,
+      "grad_norm": 1.1586123177119687,
+      "learning_rate": 3.1835995193985548e-06,
+      "loss": 0.0973,
+      "step": 4932
+    },
+    {
+      "epoch": 1.970447284345048,
+      "grad_norm": 1.1103102164856757,
+      "learning_rate": 3.1792704836773303e-06,
+      "loss": 0.0953,
+      "step": 4934
+    },
+    {
+      "epoch": 1.9712460063897763,
+      "grad_norm": 1.165044371843411,
+      "learning_rate": 3.174943020894618e-06,
+      "loss": 0.0933,
+      "step": 4936
+    },
+    {
+      "epoch": 1.9720447284345048,
+      "grad_norm": 0.9945637263764828,
+      "learning_rate": 3.170617134788939e-06,
+      "loss": 0.0801,
+      "step": 4938
+    },
+    {
+      "epoch": 1.9728434504792332,
+      "grad_norm": 1.0071008319354997,
+      "learning_rate": 3.1662928290974514e-06,
+      "loss": 0.0916,
+      "step": 4940
+    },
+    {
+      "epoch": 1.9736421725239617,
+      "grad_norm": 1.1375941744992564,
+      "learning_rate": 3.161970107555945e-06,
+      "loss": 0.1005,
+      "step": 4942
+    },
+    {
+      "epoch": 1.9744408945686902,
+      "grad_norm": 1.0807791951129158,
+      "learning_rate": 3.1576489738988457e-06,
+      "loss": 0.0919,
+      "step": 4944
+    },
+    {
+      "epoch": 1.9752396166134185,
+      "grad_norm": 1.0609866630309928,
+      "learning_rate": 3.153329431859204e-06,
+      "loss": 0.0915,
+      "step": 4946
+    },
+    {
+      "epoch": 1.9760383386581468,
+      "grad_norm": 1.1417443819482096,
+      "learning_rate": 3.1490114851686984e-06,
+      "loss": 0.0931,
+      "step": 4948
+    },
+    {
+      "epoch": 1.9768370607028753,
+      "grad_norm": 0.9100647284833052,
+      "learning_rate": 3.144695137557624e-06,
+      "loss": 0.0772,
+      "step": 4950
+    },
+    {
+      "epoch": 1.9776357827476039,
+      "grad_norm": 0.9703032312651547,
+      "learning_rate": 3.140380392754901e-06,
+      "loss": 0.098,
+      "step": 4952
+    },
+    {
+      "epoch": 1.9784345047923324,
+      "grad_norm": 1.0592857088516412,
+      "learning_rate": 3.1360672544880586e-06,
+      "loss": 0.0962,
+      "step": 4954
+    },
+    {
+      "epoch": 1.9792332268370607,
+      "grad_norm": 1.0580116375862416,
+      "learning_rate": 3.1317557264832454e-06,
+      "loss": 0.0841,
+      "step": 4956
+    },
+    {
+      "epoch": 1.980031948881789,
+      "grad_norm": 1.0265861095673052,
+      "learning_rate": 3.1274458124652117e-06,
+      "loss": 0.0901,
+      "step": 4958
+    },
+    {
+      "epoch": 1.9808306709265175,
+      "grad_norm": 1.0372575193954332,
+      "learning_rate": 3.12313751615732e-06,
+      "loss": 0.0736,
+      "step": 4960
+    },
+    {
+      "epoch": 1.981629392971246,
+      "grad_norm": 0.9602359888946597,
+      "learning_rate": 3.1188308412815276e-06,
+      "loss": 0.087,
+      "step": 4962
+    },
+    {
+      "epoch": 1.9824281150159746,
+      "grad_norm": 1.1199736388647585,
+      "learning_rate": 3.114525791558398e-06,
+      "loss": 0.0867,
+      "step": 4964
+    },
+    {
+      "epoch": 1.983226837060703,
+      "grad_norm": 1.0792261618908379,
+      "learning_rate": 3.1102223707070865e-06,
+      "loss": 0.0887,
+      "step": 4966
+    },
+    {
+      "epoch": 1.9840255591054312,
+      "grad_norm": 1.1181889010486792,
+      "learning_rate": 3.1059205824453446e-06,
+      "loss": 0.0929,
+      "step": 4968
+    },
+    {
+      "epoch": 1.9848242811501597,
+      "grad_norm": 1.0142291902192897,
+      "learning_rate": 3.101620430489509e-06,
+      "loss": 0.0876,
+      "step": 4970
+    },
+    {
+      "epoch": 1.9856230031948883,
+      "grad_norm": 1.2214112942901962,
+      "learning_rate": 3.0973219185545077e-06,
+      "loss": 0.1051,
+      "step": 4972
+    },
+    {
+      "epoch": 1.9864217252396166,
+      "grad_norm": 1.1881886135246487,
+      "learning_rate": 3.093025050353847e-06,
+      "loss": 0.0861,
+      "step": 4974
+    },
+    {
+      "epoch": 1.9872204472843449,
+      "grad_norm": 1.0284769527949214,
+      "learning_rate": 3.0887298295996183e-06,
+      "loss": 0.0848,
+      "step": 4976
+    },
+    {
+      "epoch": 1.9880191693290734,
+      "grad_norm": 1.0067298405880323,
+      "learning_rate": 3.0844362600024813e-06,
+      "loss": 0.0859,
+      "step": 4978
+    },
+    {
+      "epoch": 1.988817891373802,
+      "grad_norm": 1.0990861865733983,
+      "learning_rate": 3.0801443452716835e-06,
+      "loss": 0.0938,
+      "step": 4980
+    },
+    {
+      "epoch": 1.9896166134185305,
+      "grad_norm": 0.914893405123526,
+      "learning_rate": 3.0758540891150286e-06,
+      "loss": 0.0785,
+      "step": 4982
+    },
+    {
+      "epoch": 1.9904153354632588,
+      "grad_norm": 1.04304535262742,
+      "learning_rate": 3.0715654952388957e-06,
+      "loss": 0.0926,
+      "step": 4984
+    },
+    {
+      "epoch": 1.991214057507987,
+      "grad_norm": 1.1088676583366113,
+      "learning_rate": 3.067278567348223e-06,
+      "loss": 0.0885,
+      "step": 4986
+    },
+    {
+      "epoch": 1.9920127795527156,
+      "grad_norm": 1.0903427682959805,
+      "learning_rate": 3.062993309146514e-06,
+      "loss": 0.083,
+      "step": 4988
+    },
+    {
+      "epoch": 1.9928115015974441,
+      "grad_norm": 1.1545810549870772,
+      "learning_rate": 3.0587097243358254e-06,
+      "loss": 0.0835,
+      "step": 4990
+    },
+    {
+      "epoch": 1.9936102236421727,
+      "grad_norm": 1.080925053565247,
+      "learning_rate": 3.054427816616773e-06,
+      "loss": 0.0955,
+      "step": 4992
+    },
+    {
+      "epoch": 1.994408945686901,
+      "grad_norm": 1.0715872751652953,
+      "learning_rate": 3.0501475896885175e-06,
+      "loss": 0.0842,
+      "step": 4994
+    },
+    {
+      "epoch": 1.9952076677316293,
+      "grad_norm": 1.1378173558732894,
+      "learning_rate": 3.045869047248774e-06,
+      "loss": 0.0974,
+      "step": 4996
+    },
+    {
+      "epoch": 1.9960063897763578,
+      "grad_norm": 1.110967676470196,
+      "learning_rate": 3.041592192993798e-06,
+      "loss": 0.0928,
+      "step": 4998
+    },
+    {
+      "epoch": 1.9968051118210863,
+      "grad_norm": 0.9911386613472772,
+      "learning_rate": 3.0373170306183885e-06,
+      "loss": 0.0737,
+      "step": 5000
+    },
+    {
+      "epoch": 1.9968051118210863,
+      "eval_loss": 0.1455243080854416,
+      "eval_runtime": 417.8708,
+      "eval_samples_per_second": 42.614,
+      "eval_steps_per_second": 5.327,
+      "step": 5000
+    },
+    {
+      "epoch": 1.9976038338658149,
+      "grad_norm": 1.124610364572997,
+      "learning_rate": 3.0330435638158805e-06,
+      "loss": 0.0935,
+      "step": 5002
+    },
+    {
+      "epoch": 1.9984025559105432,
+      "grad_norm": 1.053117611083023,
+      "learning_rate": 3.028771796278151e-06,
+      "loss": 0.1017,
+      "step": 5004
+    },
+    {
+      "epoch": 1.9992012779552715,
+      "grad_norm": 1.151547544509955,
+      "learning_rate": 3.0245017316956e-06,
+      "loss": 0.0902,
+      "step": 5006
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1612757379061158,
+      "learning_rate": 3.020233373757162e-06,
+      "loss": 0.0987,
+      "step": 5008
+    },
+    {
+      "epoch": 2.0007987220447285,
+      "grad_norm": 0.6298174764155958,
+      "learning_rate": 3.0159667261502944e-06,
+      "loss": 0.0395,
+      "step": 5010
+    },
+    {
+      "epoch": 2.001597444089457,
+      "grad_norm": 0.625623431879438,
+      "learning_rate": 3.0117017925609802e-06,
+      "loss": 0.0417,
+      "step": 5012
+    },
+    {
+      "epoch": 2.002396166134185,
+      "grad_norm": 0.6897871271662769,
+      "learning_rate": 3.007438576673717e-06,
+      "loss": 0.0421,
+      "step": 5014
+    },
+    {
+      "epoch": 2.0031948881789137,
+      "grad_norm": 0.6744101045802077,
+      "learning_rate": 3.0031770821715233e-06,
+      "loss": 0.0378,
+      "step": 5016
+    },
+    {
+      "epoch": 2.003993610223642,
+      "grad_norm": 0.7756444837427113,
+      "learning_rate": 2.9989173127359267e-06,
+      "loss": 0.04,
+      "step": 5018
+    },
+    {
+      "epoch": 2.0047923322683707,
+      "grad_norm": 0.7859449397270831,
+      "learning_rate": 2.9946592720469662e-06,
+      "loss": 0.037,
+      "step": 5020
+    },
+    {
+      "epoch": 2.0055910543130993,
+      "grad_norm": 0.6704787926555024,
+      "learning_rate": 2.9904029637831887e-06,
+      "loss": 0.0347,
+      "step": 5022
+    },
+    {
+      "epoch": 2.0063897763578273,
+      "grad_norm": 0.751662927080756,
+      "learning_rate": 2.9861483916216404e-06,
+      "loss": 0.0389,
+      "step": 5024
+    },
+    {
+      "epoch": 2.007188498402556,
+      "grad_norm": 0.9221469194496961,
+      "learning_rate": 2.981895559237873e-06,
+      "loss": 0.0447,
+      "step": 5026
+    },
+    {
+      "epoch": 2.0079872204472844,
+      "grad_norm": 0.80221078116781,
+      "learning_rate": 2.9776444703059316e-06,
+      "loss": 0.0345,
+      "step": 5028
+    },
+    {
+      "epoch": 2.008785942492013,
+      "grad_norm": 0.8059814098019393,
+      "learning_rate": 2.9733951284983555e-06,
+      "loss": 0.0373,
+      "step": 5030
+    },
+    {
+      "epoch": 2.009584664536741,
+      "grad_norm": 0.9358045993876437,
+      "learning_rate": 2.969147537486175e-06,
+      "loss": 0.0408,
+      "step": 5032
+    },
+    {
+      "epoch": 2.0103833865814695,
+      "grad_norm": 0.8885370742724994,
+      "learning_rate": 2.9649017009389077e-06,
+      "loss": 0.0483,
+      "step": 5034
+    },
+    {
+      "epoch": 2.011182108626198,
+      "grad_norm": 1.057708534412327,
+      "learning_rate": 2.9606576225245566e-06,
+      "loss": 0.0379,
+      "step": 5036
+    },
+    {
+      "epoch": 2.0119808306709266,
+      "grad_norm": 0.8731366761911997,
+      "learning_rate": 2.9564153059096047e-06,
+      "loss": 0.0375,
+      "step": 5038
+    },
+    {
+      "epoch": 2.012779552715655,
+      "grad_norm": 0.8694849953429314,
+      "learning_rate": 2.952174754759012e-06,
+      "loss": 0.0359,
+      "step": 5040
+    },
+    {
+      "epoch": 2.013578274760383,
+      "grad_norm": 0.8842840721862356,
+      "learning_rate": 2.947935972736217e-06,
+      "loss": 0.0351,
+      "step": 5042
+    },
+    {
+      "epoch": 2.0143769968051117,
+      "grad_norm": 0.8362117175566706,
+      "learning_rate": 2.9436989635031253e-06,
+      "loss": 0.0353,
+      "step": 5044
+    },
+    {
+      "epoch": 2.0151757188498403,
+      "grad_norm": 0.8610101010994339,
+      "learning_rate": 2.9394637307201156e-06,
+      "loss": 0.0365,
+      "step": 5046
+    },
+    {
+      "epoch": 2.015974440894569,
+      "grad_norm": 0.8646570046386132,
+      "learning_rate": 2.935230278046025e-06,
+      "loss": 0.0376,
+      "step": 5048
+    },
+    {
+      "epoch": 2.0167731629392973,
+      "grad_norm": 0.8838043984613749,
+      "learning_rate": 2.9309986091381616e-06,
+      "loss": 0.0331,
+      "step": 5050
+    },
+    {
+      "epoch": 2.0175718849840254,
+      "grad_norm": 0.8681957347092829,
+      "learning_rate": 2.9267687276522876e-06,
+      "loss": 0.0355,
+      "step": 5052
+    },
+    {
+      "epoch": 2.018370607028754,
+      "grad_norm": 0.9197550042905789,
+      "learning_rate": 2.922540637242619e-06,
+      "loss": 0.0366,
+      "step": 5054
+    },
+    {
+      "epoch": 2.0191693290734825,
+      "grad_norm": 0.9123850396028426,
+      "learning_rate": 2.9183143415618297e-06,
+      "loss": 0.035,
+      "step": 5056
+    },
+    {
+      "epoch": 2.019968051118211,
+      "grad_norm": 0.8949092568287165,
+      "learning_rate": 2.9140898442610375e-06,
+      "loss": 0.0383,
+      "step": 5058
+    },
+    {
+      "epoch": 2.0207667731629395,
+      "grad_norm": 1.0371718273633521,
+      "learning_rate": 2.909867148989812e-06,
+      "loss": 0.0365,
+      "step": 5060
+    },
+    {
+      "epoch": 2.0215654952076676,
+      "grad_norm": 0.7653356631917099,
+      "learning_rate": 2.905646259396162e-06,
+      "loss": 0.0358,
+      "step": 5062
+    },
+    {
+      "epoch": 2.022364217252396,
+      "grad_norm": 0.7314418159084973,
+      "learning_rate": 2.9014271791265403e-06,
+      "loss": 0.031,
+      "step": 5064
+    },
+    {
+      "epoch": 2.0231629392971247,
+      "grad_norm": 0.8465101892289698,
+      "learning_rate": 2.8972099118258305e-06,
+      "loss": 0.0313,
+      "step": 5066
+    },
+    {
+      "epoch": 2.023961661341853,
+      "grad_norm": 0.8580294083980158,
+      "learning_rate": 2.8929944611373555e-06,
+      "loss": 0.0349,
+      "step": 5068
+    },
+    {
+      "epoch": 2.0247603833865813,
+      "grad_norm": 0.820776207639085,
+      "learning_rate": 2.888780830702867e-06,
+      "loss": 0.0337,
+      "step": 5070
+    },
+    {
+      "epoch": 2.02555910543131,
+      "grad_norm": 0.9528932671606859,
+      "learning_rate": 2.8845690241625437e-06,
+      "loss": 0.0396,
+      "step": 5072
+    },
+    {
+      "epoch": 2.0263578274760383,
+      "grad_norm": 1.0061064855225768,
+      "learning_rate": 2.88035904515499e-06,
+      "loss": 0.0397,
+      "step": 5074
+    },
+    {
+      "epoch": 2.027156549520767,
+      "grad_norm": 0.8452096897085726,
+      "learning_rate": 2.8761508973172293e-06,
+      "loss": 0.0346,
+      "step": 5076
+    },
+    {
+      "epoch": 2.0279552715654954,
+      "grad_norm": 0.7634443097067318,
+      "learning_rate": 2.871944584284705e-06,
+      "loss": 0.0322,
+      "step": 5078
+    },
+    {
+      "epoch": 2.0287539936102235,
+      "grad_norm": 1.0425744751303125,
+      "learning_rate": 2.867740109691277e-06,
+      "loss": 0.0355,
+      "step": 5080
+    },
+    {
+      "epoch": 2.029552715654952,
+      "grad_norm": 1.0565258664896207,
+      "learning_rate": 2.86353747716921e-06,
+      "loss": 0.0378,
+      "step": 5082
+    },
+    {
+      "epoch": 2.0303514376996805,
+      "grad_norm": 1.0498591946749731,
+      "learning_rate": 2.859336690349185e-06,
+      "loss": 0.0407,
+      "step": 5084
+    },
+    {
+      "epoch": 2.031150159744409,
+      "grad_norm": 0.7704667558370644,
+      "learning_rate": 2.8551377528602836e-06,
+      "loss": 0.0325,
+      "step": 5086
+    },
+    {
+      "epoch": 2.0319488817891376,
+      "grad_norm": 0.9433544175624843,
+      "learning_rate": 2.850940668329996e-06,
+      "loss": 0.034,
+      "step": 5088
+    },
+    {
+      "epoch": 2.0327476038338657,
+      "grad_norm": 0.8469629469335119,
+      "learning_rate": 2.8467454403842005e-06,
+      "loss": 0.0341,
+      "step": 5090
+    },
+    {
+      "epoch": 2.033546325878594,
+      "grad_norm": 0.7628386290158324,
+      "learning_rate": 2.842552072647182e-06,
+      "loss": 0.0306,
+      "step": 5092
+    },
+    {
+      "epoch": 2.0343450479233227,
+      "grad_norm": 0.9647679056914126,
+      "learning_rate": 2.838360568741613e-06,
+      "loss": 0.0419,
+      "step": 5094
+    },
+    {
+      "epoch": 2.0351437699680512,
+      "grad_norm": 0.8818509646802964,
+      "learning_rate": 2.8341709322885624e-06,
+      "loss": 0.0322,
+      "step": 5096
+    },
+    {
+      "epoch": 2.0359424920127798,
+      "grad_norm": 0.9443959085201566,
+      "learning_rate": 2.8299831669074744e-06,
+      "loss": 0.04,
+      "step": 5098
+    },
+    {
+      "epoch": 2.036741214057508,
+      "grad_norm": 1.0294482005419765,
+      "learning_rate": 2.8257972762161865e-06,
+      "loss": 0.0371,
+      "step": 5100
+    },
+    {
+      "epoch": 2.0375399361022364,
+      "grad_norm": 1.0211095104000223,
+      "learning_rate": 2.8216132638309124e-06,
+      "loss": 0.0365,
+      "step": 5102
+    },
+    {
+      "epoch": 2.038338658146965,
+      "grad_norm": 1.9713879821461935,
+      "learning_rate": 2.817431133366246e-06,
+      "loss": 0.0437,
+      "step": 5104
+    },
+    {
+      "epoch": 2.0391373801916934,
+      "grad_norm": 0.9823140675823749,
+      "learning_rate": 2.8132508884351504e-06,
+      "loss": 0.0379,
+      "step": 5106
+    },
+    {
+      "epoch": 2.0399361022364215,
+      "grad_norm": 0.8967790877119935,
+      "learning_rate": 2.809072532648963e-06,
+      "loss": 0.0335,
+      "step": 5108
+    },
+    {
+      "epoch": 2.04073482428115,
+      "grad_norm": 0.9950222932994464,
+      "learning_rate": 2.804896069617391e-06,
+      "loss": 0.0414,
+      "step": 5110
+    },
+    {
+      "epoch": 2.0415335463258786,
+      "grad_norm": 1.0503584997850632,
+      "learning_rate": 2.800721502948506e-06,
+      "loss": 0.0371,
+      "step": 5112
+    },
+    {
+      "epoch": 2.042332268370607,
+      "grad_norm": 0.8863498646029845,
+      "learning_rate": 2.7965488362487337e-06,
+      "loss": 0.037,
+      "step": 5114
+    },
+    {
+      "epoch": 2.0431309904153356,
+      "grad_norm": 0.9129952865118284,
+      "learning_rate": 2.7923780731228665e-06,
+      "loss": 0.0379,
+      "step": 5116
+    },
+    {
+      "epoch": 2.0439297124600637,
+      "grad_norm": 0.9360258175259091,
+      "learning_rate": 2.7882092171740544e-06,
+      "loss": 0.0386,
+      "step": 5118
+    },
+    {
+      "epoch": 2.0447284345047922,
+      "grad_norm": 0.9038391882536216,
+      "learning_rate": 2.7840422720037943e-06,
+      "loss": 0.0336,
+      "step": 5120
+    },
+    {
+      "epoch": 2.0455271565495208,
+      "grad_norm": 0.861340429852411,
+      "learning_rate": 2.77987724121193e-06,
+      "loss": 0.0396,
+      "step": 5122
+    },
+    {
+      "epoch": 2.0463258785942493,
+      "grad_norm": 0.9598176808433139,
+      "learning_rate": 2.775714128396658e-06,
+      "loss": 0.0353,
+      "step": 5124
+    },
+    {
+      "epoch": 2.047124600638978,
+      "grad_norm": 0.8776533817537266,
+      "learning_rate": 2.7715529371545138e-06,
+      "loss": 0.0357,
+      "step": 5126
+    },
+    {
+      "epoch": 2.047923322683706,
+      "grad_norm": 0.9993457119073764,
+      "learning_rate": 2.767393671080376e-06,
+      "loss": 0.0357,
+      "step": 5128
+    },
+    {
+      "epoch": 2.0487220447284344,
+      "grad_norm": 0.9982845161626676,
+      "learning_rate": 2.763236333767455e-06,
+      "loss": 0.0329,
+      "step": 5130
+    },
+    {
+      "epoch": 2.049520766773163,
+      "grad_norm": 0.9890197900955416,
+      "learning_rate": 2.7590809288073e-06,
+      "loss": 0.0334,
+      "step": 5132
+    },
+    {
+      "epoch": 2.0503194888178915,
+      "grad_norm": 0.838107029970905,
+      "learning_rate": 2.7549274597897878e-06,
+      "loss": 0.0349,
+      "step": 5134
+    },
+    {
+      "epoch": 2.0511182108626196,
+      "grad_norm": 0.8727714025662038,
+      "learning_rate": 2.7507759303031257e-06,
+      "loss": 0.0358,
+      "step": 5136
+    },
+    {
+      "epoch": 2.051916932907348,
+      "grad_norm": 0.9757184296946524,
+      "learning_rate": 2.7466263439338424e-06,
+      "loss": 0.0388,
+      "step": 5138
+    },
+    {
+      "epoch": 2.0527156549520766,
+      "grad_norm": 0.9770465103402337,
+      "learning_rate": 2.7424787042667856e-06,
+      "loss": 0.0423,
+      "step": 5140
+    },
+    {
+      "epoch": 2.053514376996805,
+      "grad_norm": 0.8990138413398377,
+      "learning_rate": 2.7383330148851293e-06,
+      "loss": 0.0315,
+      "step": 5142
+    },
+    {
+      "epoch": 2.0543130990415337,
+      "grad_norm": 0.8588787728847247,
+      "learning_rate": 2.7341892793703594e-06,
+      "loss": 0.0315,
+      "step": 5144
+    },
+    {
+      "epoch": 2.055111821086262,
+      "grad_norm": 0.9632194247423946,
+      "learning_rate": 2.7300475013022666e-06,
+      "loss": 0.0335,
+      "step": 5146
+    },
+    {
+      "epoch": 2.0559105431309903,
+      "grad_norm": 0.730091684485515,
+      "learning_rate": 2.7259076842589595e-06,
+      "loss": 0.0316,
+      "step": 5148
+    },
+    {
+      "epoch": 2.056709265175719,
+      "grad_norm": 0.8722727704441476,
+      "learning_rate": 2.721769831816849e-06,
+      "loss": 0.0385,
+      "step": 5150
+    },
+    {
+      "epoch": 2.0575079872204474,
+      "grad_norm": 0.9033151795461563,
+      "learning_rate": 2.7176339475506515e-06,
+      "loss": 0.0369,
+      "step": 5152
+    },
+    {
+      "epoch": 2.058306709265176,
+      "grad_norm": 0.9092639705567416,
+      "learning_rate": 2.7135000350333762e-06,
+      "loss": 0.0371,
+      "step": 5154
+    },
+    {
+      "epoch": 2.059105431309904,
+      "grad_norm": 1.0234924384548971,
+      "learning_rate": 2.7093680978363367e-06,
+      "loss": 0.0379,
+      "step": 5156
+    },
+    {
+      "epoch": 2.0599041533546325,
+      "grad_norm": 0.9677538639597134,
+      "learning_rate": 2.7052381395291355e-06,
+      "loss": 0.0308,
+      "step": 5158
+    },
+    {
+      "epoch": 2.060702875399361,
+      "grad_norm": 1.0727950843621823,
+      "learning_rate": 2.7011101636796677e-06,
+      "loss": 0.0454,
+      "step": 5160
+    },
+    {
+      "epoch": 2.0615015974440896,
+      "grad_norm": 1.02920375202306,
+      "learning_rate": 2.6969841738541165e-06,
+      "loss": 0.0421,
+      "step": 5162
+    },
+    {
+      "epoch": 2.062300319488818,
+      "grad_norm": 1.1885181849696531,
+      "learning_rate": 2.6928601736169423e-06,
+      "loss": 0.0444,
+      "step": 5164
+    },
+    {
+      "epoch": 2.063099041533546,
+      "grad_norm": 1.0717318540862604,
+      "learning_rate": 2.6887381665308977e-06,
+      "loss": 0.0385,
+      "step": 5166
+    },
+    {
+      "epoch": 2.0638977635782747,
+      "grad_norm": 0.7332635219694293,
+      "learning_rate": 2.6846181561570085e-06,
+      "loss": 0.0322,
+      "step": 5168
+    },
+    {
+      "epoch": 2.0646964856230032,
+      "grad_norm": 0.854254385375307,
+      "learning_rate": 2.68050014605457e-06,
+      "loss": 0.037,
+      "step": 5170
+    },
+    {
+      "epoch": 2.0654952076677318,
+      "grad_norm": 0.8081983560893325,
+      "learning_rate": 2.6763841397811576e-06,
+      "loss": 0.034,
+      "step": 5172
+    },
+    {
+      "epoch": 2.06629392971246,
+      "grad_norm": 0.9741434482130531,
+      "learning_rate": 2.6722701408926117e-06,
+      "loss": 0.0356,
+      "step": 5174
+    },
+    {
+      "epoch": 2.0670926517571884,
+      "grad_norm": 0.8836539172435507,
+      "learning_rate": 2.668158152943039e-06,
+      "loss": 0.0402,
+      "step": 5176
+    },
+    {
+      "epoch": 2.067891373801917,
+      "grad_norm": 0.9171861393472643,
+      "learning_rate": 2.664048179484812e-06,
+      "loss": 0.0354,
+      "step": 5178
+    },
+    {
+      "epoch": 2.0686900958466454,
+      "grad_norm": 0.9101919637114287,
+      "learning_rate": 2.6599402240685546e-06,
+      "loss": 0.0354,
+      "step": 5180
+    },
+    {
+      "epoch": 2.069488817891374,
+      "grad_norm": 0.9747609472900693,
+      "learning_rate": 2.6558342902431553e-06,
+      "loss": 0.0365,
+      "step": 5182
+    },
+    {
+      "epoch": 2.070287539936102,
+      "grad_norm": 0.8324045055842462,
+      "learning_rate": 2.651730381555754e-06,
+      "loss": 0.0386,
+      "step": 5184
+    },
+    {
+      "epoch": 2.0710862619808306,
+      "grad_norm": 0.7931217955187458,
+      "learning_rate": 2.64762850155174e-06,
+      "loss": 0.0329,
+      "step": 5186
+    },
+    {
+      "epoch": 2.071884984025559,
+      "grad_norm": 0.797923260607255,
+      "learning_rate": 2.6435286537747512e-06,
+      "loss": 0.0336,
+      "step": 5188
+    },
+    {
+      "epoch": 2.0726837060702876,
+      "grad_norm": 0.958305531859498,
+      "learning_rate": 2.6394308417666686e-06,
+      "loss": 0.0376,
+      "step": 5190
+    },
+    {
+      "epoch": 2.073482428115016,
+      "grad_norm": 0.9482044419754099,
+      "learning_rate": 2.635335069067617e-06,
+      "loss": 0.0393,
+      "step": 5192
+    },
+    {
+      "epoch": 2.0742811501597442,
+      "grad_norm": 1.0868323089043634,
+      "learning_rate": 2.6312413392159553e-06,
+      "loss": 0.0379,
+      "step": 5194
+    },
+    {
+      "epoch": 2.0750798722044728,
+      "grad_norm": 0.8207739576375747,
+      "learning_rate": 2.6271496557482795e-06,
+      "loss": 0.0343,
+      "step": 5196
+    },
+    {
+      "epoch": 2.0758785942492013,
+      "grad_norm": 0.8539413020355823,
+      "learning_rate": 2.6230600221994195e-06,
+      "loss": 0.0397,
+      "step": 5198
+    },
+    {
+      "epoch": 2.07667731629393,
+      "grad_norm": 0.8965919210232974,
+      "learning_rate": 2.618972442102432e-06,
+      "loss": 0.0381,
+      "step": 5200
+    },
+    {
+      "epoch": 2.0774760383386583,
+      "grad_norm": 0.8973056412608038,
+      "learning_rate": 2.614886918988604e-06,
+      "loss": 0.0342,
+      "step": 5202
+    },
+    {
+      "epoch": 2.0782747603833864,
+      "grad_norm": 0.9313693805488207,
+      "learning_rate": 2.610803456387436e-06,
+      "loss": 0.0358,
+      "step": 5204
+    },
+    {
+      "epoch": 2.079073482428115,
+      "grad_norm": 0.7917376380998945,
+      "learning_rate": 2.6067220578266574e-06,
+      "loss": 0.0329,
+      "step": 5206
+    },
+    {
+      "epoch": 2.0798722044728435,
+      "grad_norm": 0.961444379029632,
+      "learning_rate": 2.602642726832212e-06,
+      "loss": 0.0361,
+      "step": 5208
+    },
+    {
+      "epoch": 2.080670926517572,
+      "grad_norm": 0.8429511197266044,
+      "learning_rate": 2.5985654669282556e-06,
+      "loss": 0.0334,
+      "step": 5210
+    },
+    {
+      "epoch": 2.0814696485623,
+      "grad_norm": 0.9355752641552857,
+      "learning_rate": 2.5944902816371573e-06,
+      "loss": 0.0395,
+      "step": 5212
+    },
+    {
+      "epoch": 2.0822683706070286,
+      "grad_norm": 0.8965794716087095,
+      "learning_rate": 2.5904171744794927e-06,
+      "loss": 0.031,
+      "step": 5214
+    },
+    {
+      "epoch": 2.083067092651757,
+      "grad_norm": 0.906394944166487,
+      "learning_rate": 2.5863461489740403e-06,
+      "loss": 0.0362,
+      "step": 5216
+    },
+    {
+      "epoch": 2.0838658146964857,
+      "grad_norm": 0.916184630737243,
+      "learning_rate": 2.5822772086377863e-06,
+      "loss": 0.035,
+      "step": 5218
+    },
+    {
+      "epoch": 2.084664536741214,
+      "grad_norm": 0.9777881509465214,
+      "learning_rate": 2.5782103569859057e-06,
+      "loss": 0.0388,
+      "step": 5220
+    },
+    {
+      "epoch": 2.0854632587859423,
+      "grad_norm": 1.0639934309744312,
+      "learning_rate": 2.5741455975317776e-06,
+      "loss": 0.0379,
+      "step": 5222
+    },
+    {
+      "epoch": 2.086261980830671,
+      "grad_norm": 0.8855617385185786,
+      "learning_rate": 2.57008293378697e-06,
+      "loss": 0.0366,
+      "step": 5224
+    },
+    {
+      "epoch": 2.0870607028753994,
+      "grad_norm": 0.9693916259214078,
+      "learning_rate": 2.566022369261243e-06,
+      "loss": 0.0384,
+      "step": 5226
+    },
+    {
+      "epoch": 2.087859424920128,
+      "grad_norm": 0.9541503121738648,
+      "learning_rate": 2.5619639074625374e-06,
+      "loss": 0.0367,
+      "step": 5228
+    },
+    {
+      "epoch": 2.0886581469648564,
+      "grad_norm": 1.0451416066763384,
+      "learning_rate": 2.557907551896984e-06,
+      "loss": 0.0352,
+      "step": 5230
+    },
+    {
+      "epoch": 2.0894568690095845,
+      "grad_norm": 0.9406486155465448,
+      "learning_rate": 2.553853306068888e-06,
+      "loss": 0.0326,
+      "step": 5232
+    },
+    {
+      "epoch": 2.090255591054313,
+      "grad_norm": 0.97097723247182,
+      "learning_rate": 2.549801173480742e-06,
+      "loss": 0.0348,
+      "step": 5234
+    },
+    {
+      "epoch": 2.0910543130990416,
+      "grad_norm": 1.0076436948344818,
+      "learning_rate": 2.5457511576332008e-06,
+      "loss": 0.0423,
+      "step": 5236
+    },
+    {
+      "epoch": 2.09185303514377,
+      "grad_norm": 1.1036965373373038,
+      "learning_rate": 2.5417032620250962e-06,
+      "loss": 0.0392,
+      "step": 5238
+    },
+    {
+      "epoch": 2.0926517571884986,
+      "grad_norm": 0.9846203021426592,
+      "learning_rate": 2.5376574901534303e-06,
+      "loss": 0.0382,
+      "step": 5240
+    },
+    {
+      "epoch": 2.0934504792332267,
+      "grad_norm": 1.346629151831746,
+      "learning_rate": 2.5336138455133684e-06,
+      "loss": 0.0379,
+      "step": 5242
+    },
+    {
+      "epoch": 2.094249201277955,
+      "grad_norm": 0.822209398062977,
+      "learning_rate": 2.5295723315982344e-06,
+      "loss": 0.0308,
+      "step": 5244
+    },
+    {
+      "epoch": 2.0950479233226837,
+      "grad_norm": 0.912394753377447,
+      "learning_rate": 2.5255329518995185e-06,
+      "loss": 0.0325,
+      "step": 5246
+    },
+    {
+      "epoch": 2.0958466453674123,
+      "grad_norm": 0.9775498681194755,
+      "learning_rate": 2.5214957099068613e-06,
+      "loss": 0.0384,
+      "step": 5248
+    },
+    {
+      "epoch": 2.0966453674121404,
+      "grad_norm": 0.810870250386815,
+      "learning_rate": 2.517460609108063e-06,
+      "loss": 0.0362,
+      "step": 5250
+    },
+    {
+      "epoch": 2.097444089456869,
+      "grad_norm": 1.3581064330087438,
+      "learning_rate": 2.5134276529890646e-06,
+      "loss": 0.0388,
+      "step": 5252
+    },
+    {
+      "epoch": 2.0982428115015974,
+      "grad_norm": 0.9668476880810652,
+      "learning_rate": 2.509396845033962e-06,
+      "loss": 0.036,
+      "step": 5254
+    },
+    {
+      "epoch": 2.099041533546326,
+      "grad_norm": 0.9552646055277381,
+      "learning_rate": 2.5053681887249916e-06,
+      "loss": 0.0354,
+      "step": 5256
+    },
+    {
+      "epoch": 2.0998402555910545,
+      "grad_norm": 0.962167773860268,
+      "learning_rate": 2.501341687542538e-06,
+      "loss": 0.0378,
+      "step": 5258
+    },
+    {
+      "epoch": 2.1006389776357826,
+      "grad_norm": 0.8997107031650471,
+      "learning_rate": 2.497317344965111e-06,
+      "loss": 0.0378,
+      "step": 5260
+    },
+    {
+      "epoch": 2.101437699680511,
+      "grad_norm": 0.9369128058048221,
+      "learning_rate": 2.493295164469367e-06,
+      "loss": 0.0352,
+      "step": 5262
+    },
+    {
+      "epoch": 2.1022364217252396,
+      "grad_norm": 0.8376692502141213,
+      "learning_rate": 2.4892751495300893e-06,
+      "loss": 0.0319,
+      "step": 5264
+    },
+    {
+      "epoch": 2.103035143769968,
+      "grad_norm": 0.9548206607669597,
+      "learning_rate": 2.4852573036201937e-06,
+      "loss": 0.0395,
+      "step": 5266
+    },
+    {
+      "epoch": 2.1038338658146967,
+      "grad_norm": 0.9387516661598057,
+      "learning_rate": 2.481241630210716e-06,
+      "loss": 0.0349,
+      "step": 5268
+    },
+    {
+      "epoch": 2.1046325878594248,
+      "grad_norm": 0.874167713863454,
+      "learning_rate": 2.4772281327708213e-06,
+      "loss": 0.0335,
+      "step": 5270
+    },
+    {
+      "epoch": 2.1054313099041533,
+      "grad_norm": 1.0185097250661654,
+      "learning_rate": 2.4732168147677927e-06,
+      "loss": 0.0357,
+      "step": 5272
+    },
+    {
+      "epoch": 2.106230031948882,
+      "grad_norm": 1.043882328934825,
+      "learning_rate": 2.46920767966703e-06,
+      "loss": 0.0347,
+      "step": 5274
+    },
+    {
+      "epoch": 2.1070287539936103,
+      "grad_norm": 1.0232740029306155,
+      "learning_rate": 2.4652007309320497e-06,
+      "loss": 0.0412,
+      "step": 5276
+    },
+    {
+      "epoch": 2.107827476038339,
+      "grad_norm": 0.884509400992361,
+      "learning_rate": 2.461195972024472e-06,
+      "loss": 0.0345,
+      "step": 5278
+    },
+    {
+      "epoch": 2.108626198083067,
+      "grad_norm": 0.9621954527543679,
+      "learning_rate": 2.4571934064040364e-06,
+      "loss": 0.0392,
+      "step": 5280
+    },
+    {
+      "epoch": 2.1094249201277955,
+      "grad_norm": 0.9325329293080564,
+      "learning_rate": 2.453193037528582e-06,
+      "loss": 0.0372,
+      "step": 5282
+    },
+    {
+      "epoch": 2.110223642172524,
+      "grad_norm": 0.77944405137395,
+      "learning_rate": 2.449194868854046e-06,
+      "loss": 0.0326,
+      "step": 5284
+    },
+    {
+      "epoch": 2.1110223642172525,
+      "grad_norm": 0.8761868440249553,
+      "learning_rate": 2.4451989038344713e-06,
+      "loss": 0.0341,
+      "step": 5286
+    },
+    {
+      "epoch": 2.1118210862619806,
+      "grad_norm": 0.8178367213212951,
+      "learning_rate": 2.4412051459219945e-06,
+      "loss": 0.0306,
+      "step": 5288
+    },
+    {
+      "epoch": 2.112619808306709,
+      "grad_norm": 0.8640047451958446,
+      "learning_rate": 2.4372135985668473e-06,
+      "loss": 0.0332,
+      "step": 5290
+    },
+    {
+      "epoch": 2.1134185303514377,
+      "grad_norm": 1.0879914849221164,
+      "learning_rate": 2.433224265217346e-06,
+      "loss": 0.0374,
+      "step": 5292
+    },
+    {
+      "epoch": 2.114217252396166,
+      "grad_norm": 0.9388437125454063,
+      "learning_rate": 2.4292371493199e-06,
+      "loss": 0.039,
+      "step": 5294
+    },
+    {
+      "epoch": 2.1150159744408947,
+      "grad_norm": 0.9016588347052378,
+      "learning_rate": 2.425252254319002e-06,
+      "loss": 0.0298,
+      "step": 5296
+    },
+    {
+      "epoch": 2.115814696485623,
+      "grad_norm": 0.9448880094558859,
+      "learning_rate": 2.4212695836572255e-06,
+      "loss": 0.0348,
+      "step": 5298
+    },
+    {
+      "epoch": 2.1166134185303513,
+      "grad_norm": 1.058137286496039,
+      "learning_rate": 2.4172891407752225e-06,
+      "loss": 0.0407,
+      "step": 5300
+    },
+    {
+      "epoch": 2.11741214057508,
+      "grad_norm": 1.0908676673345152,
+      "learning_rate": 2.4133109291117156e-06,
+      "loss": 0.0337,
+      "step": 5302
+    },
+    {
+      "epoch": 2.1182108626198084,
+      "grad_norm": 1.0100591708294226,
+      "learning_rate": 2.4093349521035105e-06,
+      "loss": 0.0357,
+      "step": 5304
+    },
+    {
+      "epoch": 2.119009584664537,
+      "grad_norm": 0.9113240938208252,
+      "learning_rate": 2.405361213185475e-06,
+      "loss": 0.0395,
+      "step": 5306
+    },
+    {
+      "epoch": 2.119808306709265,
+      "grad_norm": 1.0238016688467406,
+      "learning_rate": 2.4013897157905414e-06,
+      "loss": 0.0414,
+      "step": 5308
+    },
+    {
+      "epoch": 2.1206070287539935,
+      "grad_norm": 0.9249144986993447,
+      "learning_rate": 2.39742046334971e-06,
+      "loss": 0.038,
+      "step": 5310
+    },
+    {
+      "epoch": 2.121405750798722,
+      "grad_norm": 0.8853057009535151,
+      "learning_rate": 2.3934534592920416e-06,
+      "loss": 0.0381,
+      "step": 5312
+    },
+    {
+      "epoch": 2.1222044728434506,
+      "grad_norm": 0.9382408504614896,
+      "learning_rate": 2.3894887070446526e-06,
+      "loss": 0.0327,
+      "step": 5314
+    },
+    {
+      "epoch": 2.123003194888179,
+      "grad_norm": 0.9682097923317222,
+      "learning_rate": 2.385526210032717e-06,
+      "loss": 0.0345,
+      "step": 5316
+    },
+    {
+      "epoch": 2.123801916932907,
+      "grad_norm": 1.007959850484497,
+      "learning_rate": 2.3815659716794544e-06,
+      "loss": 0.0362,
+      "step": 5318
+    },
+    {
+      "epoch": 2.1246006389776357,
+      "grad_norm": 0.87498608229541,
+      "learning_rate": 2.3776079954061385e-06,
+      "loss": 0.0314,
+      "step": 5320
+    },
+    {
+      "epoch": 2.1253993610223643,
+      "grad_norm": 0.8230237293988656,
+      "learning_rate": 2.3736522846320894e-06,
+      "loss": 0.0329,
+      "step": 5322
+    },
+    {
+      "epoch": 2.126198083067093,
+      "grad_norm": 0.893770656802708,
+      "learning_rate": 2.369698842774667e-06,
+      "loss": 0.0379,
+      "step": 5324
+    },
+    {
+      "epoch": 2.126996805111821,
+      "grad_norm": 0.9516447751453979,
+      "learning_rate": 2.365747673249268e-06,
+      "loss": 0.0398,
+      "step": 5326
+    },
+    {
+      "epoch": 2.1277955271565494,
+      "grad_norm": 0.8303186005151683,
+      "learning_rate": 2.3617987794693358e-06,
+      "loss": 0.0291,
+      "step": 5328
+    },
+    {
+      "epoch": 2.128594249201278,
+      "grad_norm": 0.9512354460837069,
+      "learning_rate": 2.3578521648463414e-06,
+      "loss": 0.0326,
+      "step": 5330
+    },
+    {
+      "epoch": 2.1293929712460065,
+      "grad_norm": 0.9842083956952964,
+      "learning_rate": 2.3539078327897846e-06,
+      "loss": 0.036,
+      "step": 5332
+    },
+    {
+      "epoch": 2.130191693290735,
+      "grad_norm": 0.8378931235387028,
+      "learning_rate": 2.3499657867071973e-06,
+      "loss": 0.0289,
+      "step": 5334
+    },
+    {
+      "epoch": 2.130990415335463,
+      "grad_norm": 0.9684963805154867,
+      "learning_rate": 2.3460260300041355e-06,
+      "loss": 0.0319,
+      "step": 5336
+    },
+    {
+      "epoch": 2.1317891373801916,
+      "grad_norm": 0.9200136986577663,
+      "learning_rate": 2.342088566084177e-06,
+      "loss": 0.0345,
+      "step": 5338
+    },
+    {
+      "epoch": 2.13258785942492,
+      "grad_norm": 0.9761484143353245,
+      "learning_rate": 2.3381533983489213e-06,
+      "loss": 0.0296,
+      "step": 5340
+    },
+    {
+      "epoch": 2.1333865814696487,
+      "grad_norm": 0.8882033903773601,
+      "learning_rate": 2.334220530197979e-06,
+      "loss": 0.0305,
+      "step": 5342
+    },
+    {
+      "epoch": 2.134185303514377,
+      "grad_norm": 0.8114312897596324,
+      "learning_rate": 2.3302899650289773e-06,
+      "loss": 0.0317,
+      "step": 5344
+    },
+    {
+      "epoch": 2.1349840255591053,
+      "grad_norm": 0.9453542646442153,
+      "learning_rate": 2.3263617062375556e-06,
+      "loss": 0.0381,
+      "step": 5346
+    },
+    {
+      "epoch": 2.135782747603834,
+      "grad_norm": 1.0338865546370732,
+      "learning_rate": 2.322435757217357e-06,
+      "loss": 0.0361,
+      "step": 5348
+    },
+    {
+      "epoch": 2.1365814696485623,
+      "grad_norm": 0.9901385988788762,
+      "learning_rate": 2.3185121213600328e-06,
+      "loss": 0.0391,
+      "step": 5350
+    },
+    {
+      "epoch": 2.137380191693291,
+      "grad_norm": 0.9091912381004705,
+      "learning_rate": 2.314590802055232e-06,
+      "loss": 0.0316,
+      "step": 5352
+    },
+    {
+      "epoch": 2.1381789137380194,
+      "grad_norm": 0.9808804213459485,
+      "learning_rate": 2.3106718026906073e-06,
+      "loss": 0.0383,
+      "step": 5354
+    },
+    {
+      "epoch": 2.1389776357827475,
+      "grad_norm": 0.8455611288247308,
+      "learning_rate": 2.306755126651804e-06,
+      "loss": 0.0345,
+      "step": 5356
+    },
+    {
+      "epoch": 2.139776357827476,
+      "grad_norm": 0.8955449128754519,
+      "learning_rate": 2.3028407773224576e-06,
+      "loss": 0.0349,
+      "step": 5358
+    },
+    {
+      "epoch": 2.1405750798722045,
+      "grad_norm": 0.9436017969146995,
+      "learning_rate": 2.2989287580841985e-06,
+      "loss": 0.0399,
+      "step": 5360
+    },
+    {
+      "epoch": 2.141373801916933,
+      "grad_norm": 0.980164453106363,
+      "learning_rate": 2.2950190723166427e-06,
+      "loss": 0.0412,
+      "step": 5362
+    },
+    {
+      "epoch": 2.142172523961661,
+      "grad_norm": 0.8957004905788968,
+      "learning_rate": 2.291111723397391e-06,
+      "loss": 0.0327,
+      "step": 5364
+    },
+    {
+      "epoch": 2.1429712460063897,
+      "grad_norm": 0.8853186556177322,
+      "learning_rate": 2.2872067147020204e-06,
+      "loss": 0.0368,
+      "step": 5366
+    },
+    {
+      "epoch": 2.143769968051118,
+      "grad_norm": 0.9295930702988132,
+      "learning_rate": 2.2833040496040925e-06,
+      "loss": 0.0375,
+      "step": 5368
+    },
+    {
+      "epoch": 2.1445686900958467,
+      "grad_norm": 0.8269626905882466,
+      "learning_rate": 2.2794037314751412e-06,
+      "loss": 0.0377,
+      "step": 5370
+    },
+    {
+      "epoch": 2.1453674121405752,
+      "grad_norm": 0.9022026161705567,
+      "learning_rate": 2.275505763684674e-06,
+      "loss": 0.0377,
+      "step": 5372
+    },
+    {
+      "epoch": 2.1461661341853033,
+      "grad_norm": 0.867448073493281,
+      "learning_rate": 2.2716101496001663e-06,
+      "loss": 0.0385,
+      "step": 5374
+    },
+    {
+      "epoch": 2.146964856230032,
+      "grad_norm": 0.8312737054421381,
+      "learning_rate": 2.267716892587062e-06,
+      "loss": 0.0307,
+      "step": 5376
+    },
+    {
+      "epoch": 2.1477635782747604,
+      "grad_norm": 0.881602657629581,
+      "learning_rate": 2.2638259960087665e-06,
+      "loss": 0.0307,
+      "step": 5378
+    },
+    {
+      "epoch": 2.148562300319489,
+      "grad_norm": 0.9967389956242162,
+      "learning_rate": 2.2599374632266514e-06,
+      "loss": 0.0356,
+      "step": 5380
+    },
+    {
+      "epoch": 2.1493610223642174,
+      "grad_norm": 0.8461757104444423,
+      "learning_rate": 2.2560512976000366e-06,
+      "loss": 0.0347,
+      "step": 5382
+    },
+    {
+      "epoch": 2.1501597444089455,
+      "grad_norm": 1.0004876530323552,
+      "learning_rate": 2.252167502486205e-06,
+      "loss": 0.0353,
+      "step": 5384
+    },
+    {
+      "epoch": 2.150958466453674,
+      "grad_norm": 0.9652946295733036,
+      "learning_rate": 2.2482860812403887e-06,
+      "loss": 0.0322,
+      "step": 5386
+    },
+    {
+      "epoch": 2.1517571884984026,
+      "grad_norm": 0.8292517773055327,
+      "learning_rate": 2.2444070372157724e-06,
+      "loss": 0.0339,
+      "step": 5388
+    },
+    {
+      "epoch": 2.152555910543131,
+      "grad_norm": 0.8955686140212332,
+      "learning_rate": 2.2405303737634794e-06,
+      "loss": 0.0339,
+      "step": 5390
+    },
+    {
+      "epoch": 2.1533546325878596,
+      "grad_norm": 0.8707856960045441,
+      "learning_rate": 2.2366560942325833e-06,
+      "loss": 0.0378,
+      "step": 5392
+    },
+    {
+      "epoch": 2.1541533546325877,
+      "grad_norm": 0.9343304509307997,
+      "learning_rate": 2.232784201970094e-06,
+      "loss": 0.0329,
+      "step": 5394
+    },
+    {
+      "epoch": 2.1549520766773163,
+      "grad_norm": 0.8050623246924522,
+      "learning_rate": 2.228914700320967e-06,
+      "loss": 0.0316,
+      "step": 5396
+    },
+    {
+      "epoch": 2.155750798722045,
+      "grad_norm": 0.9979211632823265,
+      "learning_rate": 2.2250475926280814e-06,
+      "loss": 0.039,
+      "step": 5398
+    },
+    {
+      "epoch": 2.1565495207667733,
+      "grad_norm": 1.1270330722228508,
+      "learning_rate": 2.2211828822322547e-06,
+      "loss": 0.0391,
+      "step": 5400
+    },
+    {
+      "epoch": 2.1573482428115014,
+      "grad_norm": 1.1140624044951437,
+      "learning_rate": 2.217320572472232e-06,
+      "loss": 0.0357,
+      "step": 5402
+    },
+    {
+      "epoch": 2.15814696485623,
+      "grad_norm": 1.0128344684936457,
+      "learning_rate": 2.2134606666846863e-06,
+      "loss": 0.0402,
+      "step": 5404
+    },
+    {
+      "epoch": 2.1589456869009584,
+      "grad_norm": 0.8869317328324887,
+      "learning_rate": 2.209603168204209e-06,
+      "loss": 0.0301,
+      "step": 5406
+    },
+    {
+      "epoch": 2.159744408945687,
+      "grad_norm": 0.8719598409358753,
+      "learning_rate": 2.205748080363316e-06,
+      "loss": 0.0327,
+      "step": 5408
+    },
+    {
+      "epoch": 2.1605431309904155,
+      "grad_norm": 0.8425272637471012,
+      "learning_rate": 2.2018954064924392e-06,
+      "loss": 0.0317,
+      "step": 5410
+    },
+    {
+      "epoch": 2.1613418530351436,
+      "grad_norm": 0.9849794944821841,
+      "learning_rate": 2.1980451499199262e-06,
+      "loss": 0.0352,
+      "step": 5412
+    },
+    {
+      "epoch": 2.162140575079872,
+      "grad_norm": 1.0062068145536687,
+      "learning_rate": 2.1941973139720368e-06,
+      "loss": 0.034,
+      "step": 5414
+    },
+    {
+      "epoch": 2.1629392971246006,
+      "grad_norm": 1.025042092696479,
+      "learning_rate": 2.190351901972935e-06,
+      "loss": 0.0364,
+      "step": 5416
+    },
+    {
+      "epoch": 2.163738019169329,
+      "grad_norm": 0.9038868061338211,
+      "learning_rate": 2.1865089172446928e-06,
+      "loss": 0.0342,
+      "step": 5418
+    },
+    {
+      "epoch": 2.1645367412140577,
+      "grad_norm": 0.9811870087519612,
+      "learning_rate": 2.1826683631072932e-06,
+      "loss": 0.0354,
+      "step": 5420
+    },
+    {
+      "epoch": 2.165335463258786,
+      "grad_norm": 0.9159807996481562,
+      "learning_rate": 2.1788302428786057e-06,
+      "loss": 0.0343,
+      "step": 5422
+    },
+    {
+      "epoch": 2.1661341853035143,
+      "grad_norm": 1.065699305014696,
+      "learning_rate": 2.1749945598744076e-06,
+      "loss": 0.0384,
+      "step": 5424
+    },
+    {
+      "epoch": 2.166932907348243,
+      "grad_norm": 0.9974570265355018,
+      "learning_rate": 2.171161317408366e-06,
+      "loss": 0.0333,
+      "step": 5426
+    },
+    {
+      "epoch": 2.1677316293929714,
+      "grad_norm": 1.0782436820025358,
+      "learning_rate": 2.1673305187920422e-06,
+      "loss": 0.036,
+      "step": 5428
+    },
+    {
+      "epoch": 2.1685303514377,
+      "grad_norm": 1.0684005095258051,
+      "learning_rate": 2.163502167334882e-06,
+      "loss": 0.0406,
+      "step": 5430
+    },
+    {
+      "epoch": 2.169329073482428,
+      "grad_norm": 0.933155992724429,
+      "learning_rate": 2.159676266344222e-06,
+      "loss": 0.0385,
+      "step": 5432
+    },
+    {
+      "epoch": 2.1701277955271565,
+      "grad_norm": 0.9059016726528943,
+      "learning_rate": 2.155852819125278e-06,
+      "loss": 0.0304,
+      "step": 5434
+    },
+    {
+      "epoch": 2.170926517571885,
+      "grad_norm": 0.8433132037155296,
+      "learning_rate": 2.1520318289811493e-06,
+      "loss": 0.0311,
+      "step": 5436
+    },
+    {
+      "epoch": 2.1717252396166136,
+      "grad_norm": 0.8527052161750599,
+      "learning_rate": 2.1482132992128125e-06,
+      "loss": 0.0325,
+      "step": 5438
+    },
+    {
+      "epoch": 2.1725239616613417,
+      "grad_norm": 0.9111474137493859,
+      "learning_rate": 2.144397233119112e-06,
+      "loss": 0.0326,
+      "step": 5440
+    },
+    {
+      "epoch": 2.17332268370607,
+      "grad_norm": 0.8278663259091216,
+      "learning_rate": 2.1405836339967707e-06,
+      "loss": 0.0316,
+      "step": 5442
+    },
+    {
+      "epoch": 2.1741214057507987,
+      "grad_norm": 0.8126851634140201,
+      "learning_rate": 2.136772505140382e-06,
+      "loss": 0.0297,
+      "step": 5444
+    },
+    {
+      "epoch": 2.1749201277955272,
+      "grad_norm": 0.8533464184624501,
+      "learning_rate": 2.1329638498423978e-06,
+      "loss": 0.0311,
+      "step": 5446
+    },
+    {
+      "epoch": 2.1757188498402558,
+      "grad_norm": 0.853939722117267,
+      "learning_rate": 2.1291576713931382e-06,
+      "loss": 0.0367,
+      "step": 5448
+    },
+    {
+      "epoch": 2.176517571884984,
+      "grad_norm": 1.1560454061905685,
+      "learning_rate": 2.125353973080782e-06,
+      "loss": 0.0383,
+      "step": 5450
+    },
+    {
+      "epoch": 2.1773162939297124,
+      "grad_norm": 1.013070043627522,
+      "learning_rate": 2.121552758191366e-06,
+      "loss": 0.0323,
+      "step": 5452
+    },
+    {
+      "epoch": 2.178115015974441,
+      "grad_norm": 1.011960788493202,
+      "learning_rate": 2.117754030008783e-06,
+      "loss": 0.0364,
+      "step": 5454
+    },
+    {
+      "epoch": 2.1789137380191694,
+      "grad_norm": 0.8655045405461433,
+      "learning_rate": 2.1139577918147715e-06,
+      "loss": 0.0333,
+      "step": 5456
+    },
+    {
+      "epoch": 2.179712460063898,
+      "grad_norm": 0.8272557585170817,
+      "learning_rate": 2.1101640468889255e-06,
+      "loss": 0.0354,
+      "step": 5458
+    },
+    {
+      "epoch": 2.180511182108626,
+      "grad_norm": 1.0574986102956363,
+      "learning_rate": 2.1063727985086827e-06,
+      "loss": 0.0333,
+      "step": 5460
+    },
+    {
+      "epoch": 2.1813099041533546,
+      "grad_norm": 1.0764551576276693,
+      "learning_rate": 2.102584049949326e-06,
+      "loss": 0.0307,
+      "step": 5462
+    },
+    {
+      "epoch": 2.182108626198083,
+      "grad_norm": 0.9248131636959693,
+      "learning_rate": 2.0987978044839707e-06,
+      "loss": 0.0331,
+      "step": 5464
+    },
+    {
+      "epoch": 2.1829073482428116,
+      "grad_norm": 0.9555477735825861,
+      "learning_rate": 2.0950140653835814e-06,
+      "loss": 0.032,
+      "step": 5466
+    },
+    {
+      "epoch": 2.18370607028754,
+      "grad_norm": 1.0306732674470207,
+      "learning_rate": 2.0912328359169498e-06,
+      "loss": 0.0356,
+      "step": 5468
+    },
+    {
+      "epoch": 2.1845047923322682,
+      "grad_norm": 1.0189431478737734,
+      "learning_rate": 2.087454119350703e-06,
+      "loss": 0.0317,
+      "step": 5470
+    },
+    {
+      "epoch": 2.1853035143769968,
+      "grad_norm": 0.8526696540231634,
+      "learning_rate": 2.0836779189492925e-06,
+      "loss": 0.034,
+      "step": 5472
+    },
+    {
+      "epoch": 2.1861022364217253,
+      "grad_norm": 0.8741979901728469,
+      "learning_rate": 2.079904237975e-06,
+      "loss": 0.0313,
+      "step": 5474
+    },
+    {
+      "epoch": 2.186900958466454,
+      "grad_norm": 0.9921256355991658,
+      "learning_rate": 2.0761330796879307e-06,
+      "loss": 0.0356,
+      "step": 5476
+    },
+    {
+      "epoch": 2.187699680511182,
+      "grad_norm": 1.0651481599446444,
+      "learning_rate": 2.0723644473460114e-06,
+      "loss": 0.0376,
+      "step": 5478
+    },
+    {
+      "epoch": 2.1884984025559104,
+      "grad_norm": 0.9968183069089577,
+      "learning_rate": 2.068598344204981e-06,
+      "loss": 0.0317,
+      "step": 5480
+    },
+    {
+      "epoch": 2.189297124600639,
+      "grad_norm": 1.0274581864731451,
+      "learning_rate": 2.064834773518399e-06,
+      "loss": 0.0384,
+      "step": 5482
+    },
+    {
+      "epoch": 2.1900958466453675,
+      "grad_norm": 0.7933724201731249,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.0335,
+      "step": 5484
+    },
+    {
+      "epoch": 2.190894568690096,
+      "grad_norm": 0.8535934658038107,
+      "learning_rate": 2.0573152425118703e-06,
+      "loss": 0.0366,
+      "step": 5486
+    },
+    {
+      "epoch": 2.191693290734824,
+      "grad_norm": 0.8643572127104457,
+      "learning_rate": 2.053559288688086e-06,
+      "loss": 0.0308,
+      "step": 5488
+    },
+    {
+      "epoch": 2.1924920127795526,
+      "grad_norm": 0.9310105796168394,
+      "learning_rate": 2.0498058803110775e-06,
+      "loss": 0.0327,
+      "step": 5490
+    },
+    {
+      "epoch": 2.193290734824281,
+      "grad_norm": 0.9425151428347762,
+      "learning_rate": 2.0460550206234324e-06,
+      "loss": 0.0367,
+      "step": 5492
+    },
+    {
+      "epoch": 2.1940894568690097,
+      "grad_norm": 0.8903710371632394,
+      "learning_rate": 2.042306712865543e-06,
+      "loss": 0.0332,
+      "step": 5494
+    },
+    {
+      "epoch": 2.194888178913738,
+      "grad_norm": 0.9619998941867759,
+      "learning_rate": 2.0385609602755878e-06,
+      "loss": 0.0342,
+      "step": 5496
+    },
+    {
+      "epoch": 2.1956869009584663,
+      "grad_norm": 0.8331065075000925,
+      "learning_rate": 2.0348177660895473e-06,
+      "loss": 0.0268,
+      "step": 5498
+    },
+    {
+      "epoch": 2.196485623003195,
+      "grad_norm": 0.7812360489536925,
+      "learning_rate": 2.031077133541188e-06,
+      "loss": 0.0261,
+      "step": 5500
+    },
+    {
+      "epoch": 2.196485623003195,
+      "eval_loss": 0.17987604439258575,
+      "eval_runtime": 417.2456,
+      "eval_samples_per_second": 42.678,
+      "eval_steps_per_second": 5.335,
+      "step": 5500
+    },
+    {
+      "epoch": 2.1972843450479234,
+      "grad_norm": 1.0053912019916695,
+      "learning_rate": 2.027339065862064e-06,
+      "loss": 0.0331,
+      "step": 5502
+    },
+    {
+      "epoch": 2.198083067092652,
+      "grad_norm": 0.9961342952209272,
+      "learning_rate": 2.02360356628151e-06,
+      "loss": 0.0321,
+      "step": 5504
+    },
+    {
+      "epoch": 2.1988817891373804,
+      "grad_norm": 1.1371974183191713,
+      "learning_rate": 2.019870638026648e-06,
+      "loss": 0.0357,
+      "step": 5506
+    },
+    {
+      "epoch": 2.1996805111821085,
+      "grad_norm": 1.0401230360921,
+      "learning_rate": 2.016140284322375e-06,
+      "loss": 0.0346,
+      "step": 5508
+    },
+    {
+      "epoch": 2.200479233226837,
+      "grad_norm": 0.8835551794696535,
+      "learning_rate": 2.0124125083913636e-06,
+      "loss": 0.0334,
+      "step": 5510
+    },
+    {
+      "epoch": 2.2012779552715656,
+      "grad_norm": 0.8819429072389985,
+      "learning_rate": 2.0086873134540626e-06,
+      "loss": 0.0336,
+      "step": 5512
+    },
+    {
+      "epoch": 2.202076677316294,
+      "grad_norm": 0.87779251719051,
+      "learning_rate": 2.004964702728688e-06,
+      "loss": 0.0328,
+      "step": 5514
+    },
+    {
+      "epoch": 2.202875399361022,
+      "grad_norm": 0.8894119758104073,
+      "learning_rate": 2.0012446794312236e-06,
+      "loss": 0.0284,
+      "step": 5516
+    },
+    {
+      "epoch": 2.2036741214057507,
+      "grad_norm": 1.1793198313761428,
+      "learning_rate": 1.997527246775421e-06,
+      "loss": 0.0375,
+      "step": 5518
+    },
+    {
+      "epoch": 2.2044728434504792,
+      "grad_norm": 0.9932553166480073,
+      "learning_rate": 1.9938124079727874e-06,
+      "loss": 0.0345,
+      "step": 5520
+    },
+    {
+      "epoch": 2.2052715654952078,
+      "grad_norm": 1.0489913796282613,
+      "learning_rate": 1.9901001662325946e-06,
+      "loss": 0.0339,
+      "step": 5522
+    },
+    {
+      "epoch": 2.2060702875399363,
+      "grad_norm": 0.9036884868208198,
+      "learning_rate": 1.9863905247618702e-06,
+      "loss": 0.0318,
+      "step": 5524
+    },
+    {
+      "epoch": 2.2068690095846644,
+      "grad_norm": 1.0039658882019877,
+      "learning_rate": 1.9826834867653956e-06,
+      "loss": 0.0378,
+      "step": 5526
+    },
+    {
+      "epoch": 2.207667731629393,
+      "grad_norm": 0.9584370299885461,
+      "learning_rate": 1.9789790554456977e-06,
+      "loss": 0.0351,
+      "step": 5528
+    },
+    {
+      "epoch": 2.2084664536741214,
+      "grad_norm": 1.0693519590795688,
+      "learning_rate": 1.9752772340030584e-06,
+      "loss": 0.033,
+      "step": 5530
+    },
+    {
+      "epoch": 2.20926517571885,
+      "grad_norm": 0.9311182383118676,
+      "learning_rate": 1.9715780256355014e-06,
+      "loss": 0.0315,
+      "step": 5532
+    },
+    {
+      "epoch": 2.2100638977635785,
+      "grad_norm": 0.886315430031371,
+      "learning_rate": 1.967881433538795e-06,
+      "loss": 0.0321,
+      "step": 5534
+    },
+    {
+      "epoch": 2.2108626198083066,
+      "grad_norm": 0.9667837860549522,
+      "learning_rate": 1.9641874609064443e-06,
+      "loss": 0.0298,
+      "step": 5536
+    },
+    {
+      "epoch": 2.211661341853035,
+      "grad_norm": 0.9696824443915598,
+      "learning_rate": 1.960496110929694e-06,
+      "loss": 0.0348,
+      "step": 5538
+    },
+    {
+      "epoch": 2.2124600638977636,
+      "grad_norm": 0.9405160870011481,
+      "learning_rate": 1.9568073867975217e-06,
+      "loss": 0.0307,
+      "step": 5540
+    },
+    {
+      "epoch": 2.213258785942492,
+      "grad_norm": 0.9078061671374033,
+      "learning_rate": 1.9531212916966395e-06,
+      "loss": 0.0337,
+      "step": 5542
+    },
+    {
+      "epoch": 2.2140575079872207,
+      "grad_norm": 0.8927963016211171,
+      "learning_rate": 1.9494378288114816e-06,
+      "loss": 0.0318,
+      "step": 5544
+    },
+    {
+      "epoch": 2.2148562300319488,
+      "grad_norm": 0.8652473614478016,
+      "learning_rate": 1.945757001324215e-06,
+      "loss": 0.029,
+      "step": 5546
+    },
+    {
+      "epoch": 2.2156549520766773,
+      "grad_norm": 1.0041756596183897,
+      "learning_rate": 1.9420788124147266e-06,
+      "loss": 0.0353,
+      "step": 5548
+    },
+    {
+      "epoch": 2.216453674121406,
+      "grad_norm": 0.8814650902754025,
+      "learning_rate": 1.938403265260625e-06,
+      "loss": 0.0325,
+      "step": 5550
+    },
+    {
+      "epoch": 2.2172523961661343,
+      "grad_norm": 0.9335249441231224,
+      "learning_rate": 1.9347303630372373e-06,
+      "loss": 0.0349,
+      "step": 5552
+    },
+    {
+      "epoch": 2.2180511182108624,
+      "grad_norm": 0.889201823881265,
+      "learning_rate": 1.931060108917601e-06,
+      "loss": 0.0344,
+      "step": 5554
+    },
+    {
+      "epoch": 2.218849840255591,
+      "grad_norm": 0.8902317631970373,
+      "learning_rate": 1.92739250607247e-06,
+      "loss": 0.0342,
+      "step": 5556
+    },
+    {
+      "epoch": 2.2196485623003195,
+      "grad_norm": 0.8614037453548934,
+      "learning_rate": 1.9237275576703125e-06,
+      "loss": 0.0346,
+      "step": 5558
+    },
+    {
+      "epoch": 2.220447284345048,
+      "grad_norm": 1.121859600487758,
+      "learning_rate": 1.9200652668772924e-06,
+      "loss": 0.0357,
+      "step": 5560
+    },
+    {
+      "epoch": 2.2212460063897765,
+      "grad_norm": 0.9204383985383575,
+      "learning_rate": 1.9164056368572847e-06,
+      "loss": 0.0327,
+      "step": 5562
+    },
+    {
+      "epoch": 2.2220447284345046,
+      "grad_norm": 1.069244881184746,
+      "learning_rate": 1.912748670771865e-06,
+      "loss": 0.035,
+      "step": 5564
+    },
+    {
+      "epoch": 2.222843450479233,
+      "grad_norm": 0.9412541774316218,
+      "learning_rate": 1.909094371780309e-06,
+      "loss": 0.034,
+      "step": 5566
+    },
+    {
+      "epoch": 2.2236421725239617,
+      "grad_norm": 1.1270763500131213,
+      "learning_rate": 1.9054427430395828e-06,
+      "loss": 0.0376,
+      "step": 5568
+    },
+    {
+      "epoch": 2.22444089456869,
+      "grad_norm": 1.0152263236453896,
+      "learning_rate": 1.9017937877043496e-06,
+      "loss": 0.035,
+      "step": 5570
+    },
+    {
+      "epoch": 2.2252396166134187,
+      "grad_norm": 1.0576930750635203,
+      "learning_rate": 1.8981475089269641e-06,
+      "loss": 0.0348,
+      "step": 5572
+    },
+    {
+      "epoch": 2.226038338658147,
+      "grad_norm": 0.9331195909365617,
+      "learning_rate": 1.8945039098574658e-06,
+      "loss": 0.0322,
+      "step": 5574
+    },
+    {
+      "epoch": 2.2268370607028753,
+      "grad_norm": 0.9101266572569631,
+      "learning_rate": 1.890862993643583e-06,
+      "loss": 0.0367,
+      "step": 5576
+    },
+    {
+      "epoch": 2.227635782747604,
+      "grad_norm": 0.958445615782898,
+      "learning_rate": 1.8872247634307205e-06,
+      "loss": 0.0357,
+      "step": 5578
+    },
+    {
+      "epoch": 2.2284345047923324,
+      "grad_norm": 0.9925711413975752,
+      "learning_rate": 1.883589222361965e-06,
+      "loss": 0.0369,
+      "step": 5580
+    },
+    {
+      "epoch": 2.229233226837061,
+      "grad_norm": 1.0786281449106017,
+      "learning_rate": 1.8799563735780873e-06,
+      "loss": 0.0329,
+      "step": 5582
+    },
+    {
+      "epoch": 2.230031948881789,
+      "grad_norm": 0.924349636618338,
+      "learning_rate": 1.8763262202175204e-06,
+      "loss": 0.031,
+      "step": 5584
+    },
+    {
+      "epoch": 2.2308306709265175,
+      "grad_norm": 0.8786179873798488,
+      "learning_rate": 1.8726987654163753e-06,
+      "loss": 0.0284,
+      "step": 5586
+    },
+    {
+      "epoch": 2.231629392971246,
+      "grad_norm": 0.8699008504990049,
+      "learning_rate": 1.8690740123084316e-06,
+      "loss": 0.0329,
+      "step": 5588
+    },
+    {
+      "epoch": 2.2324281150159746,
+      "grad_norm": 0.9317629329233429,
+      "learning_rate": 1.8654519640251334e-06,
+      "loss": 0.0334,
+      "step": 5590
+    },
+    {
+      "epoch": 2.2332268370607027,
+      "grad_norm": 0.858865932937765,
+      "learning_rate": 1.8618326236955908e-06,
+      "loss": 0.032,
+      "step": 5592
+    },
+    {
+      "epoch": 2.234025559105431,
+      "grad_norm": 0.9990448382210804,
+      "learning_rate": 1.858215994446569e-06,
+      "loss": 0.0358,
+      "step": 5594
+    },
+    {
+      "epoch": 2.2348242811501597,
+      "grad_norm": 0.9366419998033949,
+      "learning_rate": 1.8546020794024955e-06,
+      "loss": 0.032,
+      "step": 5596
+    },
+    {
+      "epoch": 2.2356230031948883,
+      "grad_norm": 0.9425282547585793,
+      "learning_rate": 1.8509908816854527e-06,
+      "loss": 0.0348,
+      "step": 5598
+    },
+    {
+      "epoch": 2.236421725239617,
+      "grad_norm": 0.9345193697622752,
+      "learning_rate": 1.8473824044151762e-06,
+      "loss": 0.0348,
+      "step": 5600
+    },
+    {
+      "epoch": 2.237220447284345,
+      "grad_norm": 0.9048364378406388,
+      "learning_rate": 1.843776650709046e-06,
+      "loss": 0.0292,
+      "step": 5602
+    },
+    {
+      "epoch": 2.2380191693290734,
+      "grad_norm": 1.0024652573185826,
+      "learning_rate": 1.8401736236820933e-06,
+      "loss": 0.0287,
+      "step": 5604
+    },
+    {
+      "epoch": 2.238817891373802,
+      "grad_norm": 1.141968574912131,
+      "learning_rate": 1.836573326446997e-06,
+      "loss": 0.0359,
+      "step": 5606
+    },
+    {
+      "epoch": 2.2396166134185305,
+      "grad_norm": 1.00331976516427,
+      "learning_rate": 1.8329757621140748e-06,
+      "loss": 0.0341,
+      "step": 5608
+    },
+    {
+      "epoch": 2.2404153354632586,
+      "grad_norm": 0.9747936880200038,
+      "learning_rate": 1.8293809337912789e-06,
+      "loss": 0.0379,
+      "step": 5610
+    },
+    {
+      "epoch": 2.241214057507987,
+      "grad_norm": 0.9903639773982147,
+      "learning_rate": 1.8257888445842026e-06,
+      "loss": 0.0335,
+      "step": 5612
+    },
+    {
+      "epoch": 2.2420127795527156,
+      "grad_norm": 0.9391122747126337,
+      "learning_rate": 1.8221994975960739e-06,
+      "loss": 0.028,
+      "step": 5614
+    },
+    {
+      "epoch": 2.242811501597444,
+      "grad_norm": 0.9322142611011831,
+      "learning_rate": 1.81861289592775e-06,
+      "loss": 0.0332,
+      "step": 5616
+    },
+    {
+      "epoch": 2.2436102236421727,
+      "grad_norm": 0.926413396097988,
+      "learning_rate": 1.815029042677714e-06,
+      "loss": 0.0363,
+      "step": 5618
+    },
+    {
+      "epoch": 2.244408945686901,
+      "grad_norm": 0.9401248489200065,
+      "learning_rate": 1.8114479409420783e-06,
+      "loss": 0.0349,
+      "step": 5620
+    },
+    {
+      "epoch": 2.2452076677316293,
+      "grad_norm": 0.9208704620021309,
+      "learning_rate": 1.8078695938145768e-06,
+      "loss": 0.0315,
+      "step": 5622
+    },
+    {
+      "epoch": 2.246006389776358,
+      "grad_norm": 0.9916844141557866,
+      "learning_rate": 1.8042940043865658e-06,
+      "loss": 0.0314,
+      "step": 5624
+    },
+    {
+      "epoch": 2.2468051118210863,
+      "grad_norm": 1.0110707106239305,
+      "learning_rate": 1.8007211757470117e-06,
+      "loss": 0.0371,
+      "step": 5626
+    },
+    {
+      "epoch": 2.247603833865815,
+      "grad_norm": 1.1436090733823392,
+      "learning_rate": 1.7971511109825064e-06,
+      "loss": 0.0347,
+      "step": 5628
+    },
+    {
+      "epoch": 2.248402555910543,
+      "grad_norm": 0.9259112088416946,
+      "learning_rate": 1.7935838131772481e-06,
+      "loss": 0.0338,
+      "step": 5630
+    },
+    {
+      "epoch": 2.2492012779552715,
+      "grad_norm": 0.9880736506122585,
+      "learning_rate": 1.7900192854130465e-06,
+      "loss": 0.035,
+      "step": 5632
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.8024555556579349,
+      "learning_rate": 1.786457530769314e-06,
+      "loss": 0.0328,
+      "step": 5634
+    },
+    {
+      "epoch": 2.2507987220447285,
+      "grad_norm": 0.8899527470220918,
+      "learning_rate": 1.7828985523230725e-06,
+      "loss": 0.0334,
+      "step": 5636
+    },
+    {
+      "epoch": 2.251597444089457,
+      "grad_norm": 0.9363903078372263,
+      "learning_rate": 1.779342353148943e-06,
+      "loss": 0.0348,
+      "step": 5638
+    },
+    {
+      "epoch": 2.252396166134185,
+      "grad_norm": 0.9060811105083338,
+      "learning_rate": 1.7757889363191484e-06,
+      "loss": 0.034,
+      "step": 5640
+    },
+    {
+      "epoch": 2.2531948881789137,
+      "grad_norm": 1.025351909976564,
+      "learning_rate": 1.7722383049035019e-06,
+      "loss": 0.0391,
+      "step": 5642
+    },
+    {
+      "epoch": 2.253993610223642,
+      "grad_norm": 1.0426242945439066,
+      "learning_rate": 1.7686904619694156e-06,
+      "loss": 0.0356,
+      "step": 5644
+    },
+    {
+      "epoch": 2.2547923322683707,
+      "grad_norm": 0.8919813927820728,
+      "learning_rate": 1.7651454105818915e-06,
+      "loss": 0.0306,
+      "step": 5646
+    },
+    {
+      "epoch": 2.255591054313099,
+      "grad_norm": 0.956484596890347,
+      "learning_rate": 1.7616031538035189e-06,
+      "loss": 0.0354,
+      "step": 5648
+    },
+    {
+      "epoch": 2.2563897763578273,
+      "grad_norm": 0.90756103740582,
+      "learning_rate": 1.758063694694474e-06,
+      "loss": 0.0336,
+      "step": 5650
+    },
+    {
+      "epoch": 2.257188498402556,
+      "grad_norm": 0.8099489111484716,
+      "learning_rate": 1.7545270363125155e-06,
+      "loss": 0.029,
+      "step": 5652
+    },
+    {
+      "epoch": 2.2579872204472844,
+      "grad_norm": 1.0364198349070215,
+      "learning_rate": 1.7509931817129821e-06,
+      "loss": 0.0343,
+      "step": 5654
+    },
+    {
+      "epoch": 2.258785942492013,
+      "grad_norm": 0.87676918992303,
+      "learning_rate": 1.7474621339487925e-06,
+      "loss": 0.032,
+      "step": 5656
+    },
+    {
+      "epoch": 2.2595846645367414,
+      "grad_norm": 0.968068180478562,
+      "learning_rate": 1.7439338960704355e-06,
+      "loss": 0.0318,
+      "step": 5658
+    },
+    {
+      "epoch": 2.2603833865814695,
+      "grad_norm": 0.9143786360310998,
+      "learning_rate": 1.7404084711259777e-06,
+      "loss": 0.0306,
+      "step": 5660
+    },
+    {
+      "epoch": 2.261182108626198,
+      "grad_norm": 0.983712060636582,
+      "learning_rate": 1.736885862161054e-06,
+      "loss": 0.0334,
+      "step": 5662
+    },
+    {
+      "epoch": 2.2619808306709266,
+      "grad_norm": 0.9626625864043248,
+      "learning_rate": 1.7333660722188667e-06,
+      "loss": 0.034,
+      "step": 5664
+    },
+    {
+      "epoch": 2.262779552715655,
+      "grad_norm": 0.9609884877774749,
+      "learning_rate": 1.7298491043401794e-06,
+      "loss": 0.0297,
+      "step": 5666
+    },
+    {
+      "epoch": 2.263578274760383,
+      "grad_norm": 0.9396750240874573,
+      "learning_rate": 1.7263349615633228e-06,
+      "loss": 0.0332,
+      "step": 5668
+    },
+    {
+      "epoch": 2.2643769968051117,
+      "grad_norm": 0.9290418992667302,
+      "learning_rate": 1.7228236469241837e-06,
+      "loss": 0.0298,
+      "step": 5670
+    },
+    {
+      "epoch": 2.2651757188498403,
+      "grad_norm": 0.9095375069866244,
+      "learning_rate": 1.7193151634562071e-06,
+      "loss": 0.0304,
+      "step": 5672
+    },
+    {
+      "epoch": 2.265974440894569,
+      "grad_norm": 1.010681342107459,
+      "learning_rate": 1.715809514190392e-06,
+      "loss": 0.0307,
+      "step": 5674
+    },
+    {
+      "epoch": 2.2667731629392973,
+      "grad_norm": 1.0154601770035847,
+      "learning_rate": 1.712306702155288e-06,
+      "loss": 0.0306,
+      "step": 5676
+    },
+    {
+      "epoch": 2.2675718849840254,
+      "grad_norm": 1.0108750470875814,
+      "learning_rate": 1.7088067303769946e-06,
+      "loss": 0.0325,
+      "step": 5678
+    },
+    {
+      "epoch": 2.268370607028754,
+      "grad_norm": 1.0636729615862845,
+      "learning_rate": 1.7053096018791588e-06,
+      "loss": 0.0409,
+      "step": 5680
+    },
+    {
+      "epoch": 2.2691693290734825,
+      "grad_norm": 0.9202813920561608,
+      "learning_rate": 1.7018153196829662e-06,
+      "loss": 0.0332,
+      "step": 5682
+    },
+    {
+      "epoch": 2.269968051118211,
+      "grad_norm": 1.214698931969398,
+      "learning_rate": 1.6983238868071489e-06,
+      "loss": 0.0338,
+      "step": 5684
+    },
+    {
+      "epoch": 2.270766773162939,
+      "grad_norm": 0.8519680135270461,
+      "learning_rate": 1.6948353062679752e-06,
+      "loss": 0.0309,
+      "step": 5686
+    },
+    {
+      "epoch": 2.2715654952076676,
+      "grad_norm": 0.8609963780131156,
+      "learning_rate": 1.691349581079249e-06,
+      "loss": 0.0316,
+      "step": 5688
+    },
+    {
+      "epoch": 2.272364217252396,
+      "grad_norm": 1.03764079744846,
+      "learning_rate": 1.687866714252311e-06,
+      "loss": 0.035,
+      "step": 5690
+    },
+    {
+      "epoch": 2.2731629392971247,
+      "grad_norm": 0.9903945590247691,
+      "learning_rate": 1.6843867087960252e-06,
+      "loss": 0.0338,
+      "step": 5692
+    },
+    {
+      "epoch": 2.273961661341853,
+      "grad_norm": 0.9835600544969014,
+      "learning_rate": 1.6809095677167897e-06,
+      "loss": 0.033,
+      "step": 5694
+    },
+    {
+      "epoch": 2.2747603833865817,
+      "grad_norm": 0.9107299972526909,
+      "learning_rate": 1.6774352940185269e-06,
+      "loss": 0.0329,
+      "step": 5696
+    },
+    {
+      "epoch": 2.27555910543131,
+      "grad_norm": 1.0179931131158924,
+      "learning_rate": 1.6739638907026806e-06,
+      "loss": 0.0324,
+      "step": 5698
+    },
+    {
+      "epoch": 2.2763578274760383,
+      "grad_norm": 0.8564341030978901,
+      "learning_rate": 1.6704953607682161e-06,
+      "loss": 0.0371,
+      "step": 5700
+    },
+    {
+      "epoch": 2.277156549520767,
+      "grad_norm": 0.9606560664042086,
+      "learning_rate": 1.6670297072116165e-06,
+      "loss": 0.0357,
+      "step": 5702
+    },
+    {
+      "epoch": 2.2779552715654954,
+      "grad_norm": 0.970530456375879,
+      "learning_rate": 1.663566933026879e-06,
+      "loss": 0.0368,
+      "step": 5704
+    },
+    {
+      "epoch": 2.2787539936102235,
+      "grad_norm": 0.823503415647808,
+      "learning_rate": 1.6601070412055154e-06,
+      "loss": 0.0308,
+      "step": 5706
+    },
+    {
+      "epoch": 2.279552715654952,
+      "grad_norm": 0.8849782867422896,
+      "learning_rate": 1.6566500347365421e-06,
+      "loss": 0.0339,
+      "step": 5708
+    },
+    {
+      "epoch": 2.2803514376996805,
+      "grad_norm": 0.9443290922040906,
+      "learning_rate": 1.6531959166064893e-06,
+      "loss": 0.0382,
+      "step": 5710
+    },
+    {
+      "epoch": 2.281150159744409,
+      "grad_norm": 0.9814437843516525,
+      "learning_rate": 1.6497446897993885e-06,
+      "loss": 0.0314,
+      "step": 5712
+    },
+    {
+      "epoch": 2.2819488817891376,
+      "grad_norm": 0.9443631962649643,
+      "learning_rate": 1.6462963572967756e-06,
+      "loss": 0.0329,
+      "step": 5714
+    },
+    {
+      "epoch": 2.2827476038338657,
+      "grad_norm": 0.9206008196409816,
+      "learning_rate": 1.6428509220776812e-06,
+      "loss": 0.0327,
+      "step": 5716
+    },
+    {
+      "epoch": 2.283546325878594,
+      "grad_norm": 0.9303152959167551,
+      "learning_rate": 1.6394083871186362e-06,
+      "loss": 0.0326,
+      "step": 5718
+    },
+    {
+      "epoch": 2.2843450479233227,
+      "grad_norm": 0.9131910544498667,
+      "learning_rate": 1.6359687553936714e-06,
+      "loss": 0.0326,
+      "step": 5720
+    },
+    {
+      "epoch": 2.2851437699680512,
+      "grad_norm": 1.0967870027836903,
+      "learning_rate": 1.6325320298742986e-06,
+      "loss": 0.0395,
+      "step": 5722
+    },
+    {
+      "epoch": 2.2859424920127793,
+      "grad_norm": 0.8427528083071858,
+      "learning_rate": 1.6290982135295269e-06,
+      "loss": 0.0331,
+      "step": 5724
+    },
+    {
+      "epoch": 2.286741214057508,
+      "grad_norm": 0.8644733263443983,
+      "learning_rate": 1.6256673093258485e-06,
+      "loss": 0.0345,
+      "step": 5726
+    },
+    {
+      "epoch": 2.2875399361022364,
+      "grad_norm": 0.9392972386925895,
+      "learning_rate": 1.6222393202272414e-06,
+      "loss": 0.0304,
+      "step": 5728
+    },
+    {
+      "epoch": 2.288338658146965,
+      "grad_norm": 0.9017990006181082,
+      "learning_rate": 1.618814249195167e-06,
+      "loss": 0.0314,
+      "step": 5730
+    },
+    {
+      "epoch": 2.2891373801916934,
+      "grad_norm": 0.7993416720434479,
+      "learning_rate": 1.6153920991885591e-06,
+      "loss": 0.0286,
+      "step": 5732
+    },
+    {
+      "epoch": 2.289936102236422,
+      "grad_norm": 0.9176139011413571,
+      "learning_rate": 1.6119728731638345e-06,
+      "loss": 0.0369,
+      "step": 5734
+    },
+    {
+      "epoch": 2.29073482428115,
+      "grad_norm": 0.874001701788666,
+      "learning_rate": 1.6085565740748825e-06,
+      "loss": 0.033,
+      "step": 5736
+    },
+    {
+      "epoch": 2.2915335463258786,
+      "grad_norm": 0.9359227443696032,
+      "learning_rate": 1.605143204873064e-06,
+      "loss": 0.0377,
+      "step": 5738
+    },
+    {
+      "epoch": 2.292332268370607,
+      "grad_norm": 0.8690038620673686,
+      "learning_rate": 1.6017327685072047e-06,
+      "loss": 0.031,
+      "step": 5740
+    },
+    {
+      "epoch": 2.2931309904153356,
+      "grad_norm": 1.055903086127581,
+      "learning_rate": 1.5983252679236006e-06,
+      "loss": 0.0355,
+      "step": 5742
+    },
+    {
+      "epoch": 2.2939297124600637,
+      "grad_norm": 1.1058038934128054,
+      "learning_rate": 1.5949207060660138e-06,
+      "loss": 0.0328,
+      "step": 5744
+    },
+    {
+      "epoch": 2.2947284345047922,
+      "grad_norm": 0.9784683752210179,
+      "learning_rate": 1.5915190858756635e-06,
+      "loss": 0.0334,
+      "step": 5746
+    },
+    {
+      "epoch": 2.2955271565495208,
+      "grad_norm": 0.9248930707330709,
+      "learning_rate": 1.5881204102912262e-06,
+      "loss": 0.0334,
+      "step": 5748
+    },
+    {
+      "epoch": 2.2963258785942493,
+      "grad_norm": 1.028126750218338,
+      "learning_rate": 1.5847246822488388e-06,
+      "loss": 0.0344,
+      "step": 5750
+    },
+    {
+      "epoch": 2.297124600638978,
+      "grad_norm": 0.8999392063400926,
+      "learning_rate": 1.581331904682089e-06,
+      "loss": 0.0293,
+      "step": 5752
+    },
+    {
+      "epoch": 2.297923322683706,
+      "grad_norm": 0.96073417050456,
+      "learning_rate": 1.5779420805220185e-06,
+      "loss": 0.0332,
+      "step": 5754
+    },
+    {
+      "epoch": 2.2987220447284344,
+      "grad_norm": 0.9826008607081279,
+      "learning_rate": 1.574555212697113e-06,
+      "loss": 0.0338,
+      "step": 5756
+    },
+    {
+      "epoch": 2.299520766773163,
+      "grad_norm": 0.9259548407344154,
+      "learning_rate": 1.5711713041333077e-06,
+      "loss": 0.0355,
+      "step": 5758
+    },
+    {
+      "epoch": 2.3003194888178915,
+      "grad_norm": 0.999628943497661,
+      "learning_rate": 1.5677903577539806e-06,
+      "loss": 0.0353,
+      "step": 5760
+    },
+    {
+      "epoch": 2.3011182108626196,
+      "grad_norm": 1.004302579075343,
+      "learning_rate": 1.5644123764799517e-06,
+      "loss": 0.0289,
+      "step": 5762
+    },
+    {
+      "epoch": 2.301916932907348,
+      "grad_norm": 1.0927223195156304,
+      "learning_rate": 1.561037363229475e-06,
+      "loss": 0.0378,
+      "step": 5764
+    },
+    {
+      "epoch": 2.3027156549520766,
+      "grad_norm": 0.9410698542998797,
+      "learning_rate": 1.5576653209182436e-06,
+      "loss": 0.0331,
+      "step": 5766
+    },
+    {
+      "epoch": 2.303514376996805,
+      "grad_norm": 0.8643954286211598,
+      "learning_rate": 1.5542962524593869e-06,
+      "loss": 0.0317,
+      "step": 5768
+    },
+    {
+      "epoch": 2.3043130990415337,
+      "grad_norm": 1.0178268360718439,
+      "learning_rate": 1.550930160763462e-06,
+      "loss": 0.0333,
+      "step": 5770
+    },
+    {
+      "epoch": 2.3051118210862622,
+      "grad_norm": 1.0794526119781374,
+      "learning_rate": 1.547567048738452e-06,
+      "loss": 0.037,
+      "step": 5772
+    },
+    {
+      "epoch": 2.3059105431309903,
+      "grad_norm": 0.9306102687878339,
+      "learning_rate": 1.5442069192897695e-06,
+      "loss": 0.0319,
+      "step": 5774
+    },
+    {
+      "epoch": 2.306709265175719,
+      "grad_norm": 0.9959647755317999,
+      "learning_rate": 1.54084977532025e-06,
+      "loss": 0.0304,
+      "step": 5776
+    },
+    {
+      "epoch": 2.3075079872204474,
+      "grad_norm": 1.178192165622762,
+      "learning_rate": 1.5374956197301494e-06,
+      "loss": 0.0359,
+      "step": 5778
+    },
+    {
+      "epoch": 2.308306709265176,
+      "grad_norm": 0.970629316470673,
+      "learning_rate": 1.5341444554171397e-06,
+      "loss": 0.0323,
+      "step": 5780
+    },
+    {
+      "epoch": 2.309105431309904,
+      "grad_norm": 0.9995642863840597,
+      "learning_rate": 1.5307962852763115e-06,
+      "loss": 0.0314,
+      "step": 5782
+    },
+    {
+      "epoch": 2.3099041533546325,
+      "grad_norm": 0.8635977267044507,
+      "learning_rate": 1.5274511122001684e-06,
+      "loss": 0.0292,
+      "step": 5784
+    },
+    {
+      "epoch": 2.310702875399361,
+      "grad_norm": 0.941234774919912,
+      "learning_rate": 1.524108939078624e-06,
+      "loss": 0.0335,
+      "step": 5786
+    },
+    {
+      "epoch": 2.3115015974440896,
+      "grad_norm": 0.9670030899801637,
+      "learning_rate": 1.5207697687990004e-06,
+      "loss": 0.0355,
+      "step": 5788
+    },
+    {
+      "epoch": 2.312300319488818,
+      "grad_norm": 1.0863404667557413,
+      "learning_rate": 1.5174336042460264e-06,
+      "loss": 0.0359,
+      "step": 5790
+    },
+    {
+      "epoch": 2.313099041533546,
+      "grad_norm": 0.9487702530128579,
+      "learning_rate": 1.5141004483018323e-06,
+      "loss": 0.0303,
+      "step": 5792
+    },
+    {
+      "epoch": 2.3138977635782747,
+      "grad_norm": 0.8581962375915709,
+      "learning_rate": 1.5107703038459531e-06,
+      "loss": 0.0329,
+      "step": 5794
+    },
+    {
+      "epoch": 2.3146964856230032,
+      "grad_norm": 0.9463896237746494,
+      "learning_rate": 1.5074431737553158e-06,
+      "loss": 0.0344,
+      "step": 5796
+    },
+    {
+      "epoch": 2.3154952076677318,
+      "grad_norm": 1.0247492412011727,
+      "learning_rate": 1.5041190609042477e-06,
+      "loss": 0.0322,
+      "step": 5798
+    },
+    {
+      "epoch": 2.31629392971246,
+      "grad_norm": 0.8881026553630956,
+      "learning_rate": 1.5007979681644696e-06,
+      "loss": 0.033,
+      "step": 5800
+    },
+    {
+      "epoch": 2.3170926517571884,
+      "grad_norm": 0.9632852088358282,
+      "learning_rate": 1.4974798984050941e-06,
+      "loss": 0.0348,
+      "step": 5802
+    },
+    {
+      "epoch": 2.317891373801917,
+      "grad_norm": 0.8335404581898861,
+      "learning_rate": 1.4941648544926164e-06,
+      "loss": 0.0312,
+      "step": 5804
+    },
+    {
+      "epoch": 2.3186900958466454,
+      "grad_norm": 0.9511476448808138,
+      "learning_rate": 1.4908528392909233e-06,
+      "loss": 0.0343,
+      "step": 5806
+    },
+    {
+      "epoch": 2.319488817891374,
+      "grad_norm": 0.9453004464927225,
+      "learning_rate": 1.4875438556612836e-06,
+      "loss": 0.0305,
+      "step": 5808
+    },
+    {
+      "epoch": 2.3202875399361025,
+      "grad_norm": 0.9770824243294399,
+      "learning_rate": 1.4842379064623474e-06,
+      "loss": 0.0316,
+      "step": 5810
+    },
+    {
+      "epoch": 2.3210862619808306,
+      "grad_norm": 0.9354726575321289,
+      "learning_rate": 1.4809349945501422e-06,
+      "loss": 0.03,
+      "step": 5812
+    },
+    {
+      "epoch": 2.321884984025559,
+      "grad_norm": 0.9809930908311559,
+      "learning_rate": 1.4776351227780732e-06,
+      "loss": 0.0327,
+      "step": 5814
+    },
+    {
+      "epoch": 2.3226837060702876,
+      "grad_norm": 0.9812165421073059,
+      "learning_rate": 1.474338293996917e-06,
+      "loss": 0.0291,
+      "step": 5816
+    },
+    {
+      "epoch": 2.323482428115016,
+      "grad_norm": 1.0065291780758312,
+      "learning_rate": 1.4710445110548255e-06,
+      "loss": 0.0349,
+      "step": 5818
+    },
+    {
+      "epoch": 2.3242811501597442,
+      "grad_norm": 1.0293319955302287,
+      "learning_rate": 1.467753776797312e-06,
+      "loss": 0.0331,
+      "step": 5820
+    },
+    {
+      "epoch": 2.3250798722044728,
+      "grad_norm": 0.9835892465385753,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.0328,
+      "step": 5822
+    },
+    {
+      "epoch": 2.3258785942492013,
+      "grad_norm": 0.8998383284814865,
+      "learning_rate": 1.4611814657049257e-06,
+      "loss": 0.0329,
+      "step": 5824
+    },
+    {
+      "epoch": 2.32667731629393,
+      "grad_norm": 0.9572453991906558,
+      "learning_rate": 1.4578998945479084e-06,
+      "loss": 0.0328,
+      "step": 5826
+    },
+    {
+      "epoch": 2.3274760383386583,
+      "grad_norm": 1.0608318737189215,
+      "learning_rate": 1.4546213834311823e-06,
+      "loss": 0.0362,
+      "step": 5828
+    },
+    {
+      "epoch": 2.3282747603833864,
+      "grad_norm": 0.9194197731835649,
+      "learning_rate": 1.4513459351870669e-06,
+      "loss": 0.0313,
+      "step": 5830
+    },
+    {
+      "epoch": 2.329073482428115,
+      "grad_norm": 0.9174893410490872,
+      "learning_rate": 1.4480735526452427e-06,
+      "loss": 0.0325,
+      "step": 5832
+    },
+    {
+      "epoch": 2.3298722044728435,
+      "grad_norm": 0.8639598370634627,
+      "learning_rate": 1.4448042386327394e-06,
+      "loss": 0.0309,
+      "step": 5834
+    },
+    {
+      "epoch": 2.330670926517572,
+      "grad_norm": 1.0738100152461907,
+      "learning_rate": 1.4415379959739356e-06,
+      "loss": 0.0376,
+      "step": 5836
+    },
+    {
+      "epoch": 2.3314696485623,
+      "grad_norm": 0.8696561197568158,
+      "learning_rate": 1.4382748274905573e-06,
+      "loss": 0.0296,
+      "step": 5838
+    },
+    {
+      "epoch": 2.3322683706070286,
+      "grad_norm": 1.0538751425130783,
+      "learning_rate": 1.4350147360016743e-06,
+      "loss": 0.0368,
+      "step": 5840
+    },
+    {
+      "epoch": 2.333067092651757,
+      "grad_norm": 1.0257476904138636,
+      "learning_rate": 1.4317577243236968e-06,
+      "loss": 0.0365,
+      "step": 5842
+    },
+    {
+      "epoch": 2.3338658146964857,
+      "grad_norm": 0.8938198309680814,
+      "learning_rate": 1.4285037952703785e-06,
+      "loss": 0.0357,
+      "step": 5844
+    },
+    {
+      "epoch": 2.334664536741214,
+      "grad_norm": 0.9464804274403688,
+      "learning_rate": 1.425252951652803e-06,
+      "loss": 0.0293,
+      "step": 5846
+    },
+    {
+      "epoch": 2.3354632587859427,
+      "grad_norm": 0.9175488309023649,
+      "learning_rate": 1.4220051962793952e-06,
+      "loss": 0.038,
+      "step": 5848
+    },
+    {
+      "epoch": 2.336261980830671,
+      "grad_norm": 0.88533327061595,
+      "learning_rate": 1.4187605319559078e-06,
+      "loss": 0.0301,
+      "step": 5850
+    },
+    {
+      "epoch": 2.3370607028753994,
+      "grad_norm": 0.9127303239271386,
+      "learning_rate": 1.4155189614854275e-06,
+      "loss": 0.0319,
+      "step": 5852
+    },
+    {
+      "epoch": 2.337859424920128,
+      "grad_norm": 0.9168807782532327,
+      "learning_rate": 1.4122804876683616e-06,
+      "loss": 0.0308,
+      "step": 5854
+    },
+    {
+      "epoch": 2.3386581469648564,
+      "grad_norm": 1.1257679223953572,
+      "learning_rate": 1.4090451133024473e-06,
+      "loss": 0.0388,
+      "step": 5856
+    },
+    {
+      "epoch": 2.3394568690095845,
+      "grad_norm": 0.8696474054620578,
+      "learning_rate": 1.4058128411827432e-06,
+      "loss": 0.0355,
+      "step": 5858
+    },
+    {
+      "epoch": 2.340255591054313,
+      "grad_norm": 0.9000167630617952,
+      "learning_rate": 1.4025836741016274e-06,
+      "loss": 0.0341,
+      "step": 5860
+    },
+    {
+      "epoch": 2.3410543130990416,
+      "grad_norm": 0.9240511974633537,
+      "learning_rate": 1.399357614848796e-06,
+      "loss": 0.0307,
+      "step": 5862
+    },
+    {
+      "epoch": 2.34185303514377,
+      "grad_norm": 0.8909434883498069,
+      "learning_rate": 1.3961346662112585e-06,
+      "loss": 0.0341,
+      "step": 5864
+    },
+    {
+      "epoch": 2.3426517571884986,
+      "grad_norm": 1.0135872043495942,
+      "learning_rate": 1.3929148309733392e-06,
+      "loss": 0.0286,
+      "step": 5866
+    },
+    {
+      "epoch": 2.3434504792332267,
+      "grad_norm": 0.9580825595304057,
+      "learning_rate": 1.3896981119166741e-06,
+      "loss": 0.0346,
+      "step": 5868
+    },
+    {
+      "epoch": 2.344249201277955,
+      "grad_norm": 0.8615986543919892,
+      "learning_rate": 1.3864845118202013e-06,
+      "loss": 0.034,
+      "step": 5870
+    },
+    {
+      "epoch": 2.3450479233226837,
+      "grad_norm": 1.1873836274839265,
+      "learning_rate": 1.3832740334601692e-06,
+      "loss": 0.0317,
+      "step": 5872
+    },
+    {
+      "epoch": 2.3458466453674123,
+      "grad_norm": 0.8711721095623617,
+      "learning_rate": 1.3800666796101291e-06,
+      "loss": 0.0331,
+      "step": 5874
+    },
+    {
+      "epoch": 2.3466453674121404,
+      "grad_norm": 1.010832936665381,
+      "learning_rate": 1.3768624530409324e-06,
+      "loss": 0.0353,
+      "step": 5876
+    },
+    {
+      "epoch": 2.347444089456869,
+      "grad_norm": 0.9937261665959263,
+      "learning_rate": 1.373661356520727e-06,
+      "loss": 0.0362,
+      "step": 5878
+    },
+    {
+      "epoch": 2.3482428115015974,
+      "grad_norm": 0.9184012307526624,
+      "learning_rate": 1.3704633928149575e-06,
+      "loss": 0.0287,
+      "step": 5880
+    },
+    {
+      "epoch": 2.349041533546326,
+      "grad_norm": 0.9982535009466615,
+      "learning_rate": 1.3672685646863653e-06,
+      "loss": 0.033,
+      "step": 5882
+    },
+    {
+      "epoch": 2.3498402555910545,
+      "grad_norm": 0.9164440397933561,
+      "learning_rate": 1.3640768748949811e-06,
+      "loss": 0.0293,
+      "step": 5884
+    },
+    {
+      "epoch": 2.3506389776357826,
+      "grad_norm": 1.002185223742666,
+      "learning_rate": 1.360888326198121e-06,
+      "loss": 0.0341,
+      "step": 5886
+    },
+    {
+      "epoch": 2.351437699680511,
+      "grad_norm": 0.9987751966058985,
+      "learning_rate": 1.3577029213503911e-06,
+      "loss": 0.0321,
+      "step": 5888
+    },
+    {
+      "epoch": 2.3522364217252396,
+      "grad_norm": 0.9634859351636175,
+      "learning_rate": 1.354520663103681e-06,
+      "loss": 0.033,
+      "step": 5890
+    },
+    {
+      "epoch": 2.353035143769968,
+      "grad_norm": 1.0094199782296815,
+      "learning_rate": 1.351341554207163e-06,
+      "loss": 0.0356,
+      "step": 5892
+    },
+    {
+      "epoch": 2.3538338658146967,
+      "grad_norm": 0.953311272783066,
+      "learning_rate": 1.3481655974072845e-06,
+      "loss": 0.0333,
+      "step": 5894
+    },
+    {
+      "epoch": 2.3546325878594248,
+      "grad_norm": 0.9161467503492303,
+      "learning_rate": 1.3449927954477732e-06,
+      "loss": 0.0305,
+      "step": 5896
+    },
+    {
+      "epoch": 2.3554313099041533,
+      "grad_norm": 1.0625770728133241,
+      "learning_rate": 1.3418231510696312e-06,
+      "loss": 0.0335,
+      "step": 5898
+    },
+    {
+      "epoch": 2.356230031948882,
+      "grad_norm": 0.9368140797679172,
+      "learning_rate": 1.3386566670111339e-06,
+      "loss": 0.0351,
+      "step": 5900
+    },
+    {
+      "epoch": 2.3570287539936103,
+      "grad_norm": 0.9236886045000889,
+      "learning_rate": 1.3354933460078217e-06,
+      "loss": 0.0314,
+      "step": 5902
+    },
+    {
+      "epoch": 2.357827476038339,
+      "grad_norm": 1.137543124830593,
+      "learning_rate": 1.3323331907925046e-06,
+      "loss": 0.038,
+      "step": 5904
+    },
+    {
+      "epoch": 2.358626198083067,
+      "grad_norm": 0.9008576498883231,
+      "learning_rate": 1.3291762040952626e-06,
+      "loss": 0.0273,
+      "step": 5906
+    },
+    {
+      "epoch": 2.3594249201277955,
+      "grad_norm": 0.9264640068095965,
+      "learning_rate": 1.3260223886434342e-06,
+      "loss": 0.0284,
+      "step": 5908
+    },
+    {
+      "epoch": 2.360223642172524,
+      "grad_norm": 1.0330425296416088,
+      "learning_rate": 1.3228717471616153e-06,
+      "loss": 0.0359,
+      "step": 5910
+    },
+    {
+      "epoch": 2.3610223642172525,
+      "grad_norm": 0.8006806552776662,
+      "learning_rate": 1.319724282371664e-06,
+      "loss": 0.0328,
+      "step": 5912
+    },
+    {
+      "epoch": 2.3618210862619806,
+      "grad_norm": 0.9719478594390628,
+      "learning_rate": 1.3165799969926928e-06,
+      "loss": 0.0359,
+      "step": 5914
+    },
+    {
+      "epoch": 2.362619808306709,
+      "grad_norm": 0.9089803281939999,
+      "learning_rate": 1.3134388937410697e-06,
+      "loss": 0.0296,
+      "step": 5916
+    },
+    {
+      "epoch": 2.3634185303514377,
+      "grad_norm": 1.0859836987195648,
+      "learning_rate": 1.3103009753304085e-06,
+      "loss": 0.0289,
+      "step": 5918
+    },
+    {
+      "epoch": 2.364217252396166,
+      "grad_norm": 0.9338786842895157,
+      "learning_rate": 1.307166244471576e-06,
+      "loss": 0.0303,
+      "step": 5920
+    },
+    {
+      "epoch": 2.3650159744408947,
+      "grad_norm": 0.9352879035371544,
+      "learning_rate": 1.3040347038726831e-06,
+      "loss": 0.0313,
+      "step": 5922
+    },
+    {
+      "epoch": 2.365814696485623,
+      "grad_norm": 0.9193300328510401,
+      "learning_rate": 1.3009063562390866e-06,
+      "loss": 0.0308,
+      "step": 5924
+    },
+    {
+      "epoch": 2.3666134185303513,
+      "grad_norm": 0.8330378897517992,
+      "learning_rate": 1.297781204273385e-06,
+      "loss": 0.0277,
+      "step": 5926
+    },
+    {
+      "epoch": 2.36741214057508,
+      "grad_norm": 1.0098776366020665,
+      "learning_rate": 1.2946592506754097e-06,
+      "loss": 0.0292,
+      "step": 5928
+    },
+    {
+      "epoch": 2.3682108626198084,
+      "grad_norm": 0.9244538147214044,
+      "learning_rate": 1.2915404981422386e-06,
+      "loss": 0.0358,
+      "step": 5930
+    },
+    {
+      "epoch": 2.369009584664537,
+      "grad_norm": 1.009009426826755,
+      "learning_rate": 1.28842494936818e-06,
+      "loss": 0.0307,
+      "step": 5932
+    },
+    {
+      "epoch": 2.369808306709265,
+      "grad_norm": 0.8599683395682384,
+      "learning_rate": 1.2853126070447709e-06,
+      "loss": 0.0312,
+      "step": 5934
+    },
+    {
+      "epoch": 2.3706070287539935,
+      "grad_norm": 1.2504223376409314,
+      "learning_rate": 1.282203473860783e-06,
+      "loss": 0.0324,
+      "step": 5936
+    },
+    {
+      "epoch": 2.371405750798722,
+      "grad_norm": 0.9600194803305533,
+      "learning_rate": 1.2790975525022136e-06,
+      "loss": 0.0332,
+      "step": 5938
+    },
+    {
+      "epoch": 2.3722044728434506,
+      "grad_norm": 0.8579303834114889,
+      "learning_rate": 1.275994845652288e-06,
+      "loss": 0.028,
+      "step": 5940
+    },
+    {
+      "epoch": 2.373003194888179,
+      "grad_norm": 0.9767263492376159,
+      "learning_rate": 1.2728953559914486e-06,
+      "loss": 0.0343,
+      "step": 5942
+    },
+    {
+      "epoch": 2.373801916932907,
+      "grad_norm": 0.9810633979289892,
+      "learning_rate": 1.2697990861973635e-06,
+      "loss": 0.0283,
+      "step": 5944
+    },
+    {
+      "epoch": 2.3746006389776357,
+      "grad_norm": 0.8914061929408517,
+      "learning_rate": 1.2667060389449182e-06,
+      "loss": 0.0318,
+      "step": 5946
+    },
+    {
+      "epoch": 2.3753993610223643,
+      "grad_norm": 1.1915603298866675,
+      "learning_rate": 1.2636162169062133e-06,
+      "loss": 0.0402,
+      "step": 5948
+    },
+    {
+      "epoch": 2.376198083067093,
+      "grad_norm": 0.836740461461381,
+      "learning_rate": 1.260529622750563e-06,
+      "loss": 0.0311,
+      "step": 5950
+    },
+    {
+      "epoch": 2.376996805111821,
+      "grad_norm": 0.9210823534934908,
+      "learning_rate": 1.257446259144494e-06,
+      "loss": 0.0314,
+      "step": 5952
+    },
+    {
+      "epoch": 2.3777955271565494,
+      "grad_norm": 0.9674053346660237,
+      "learning_rate": 1.2543661287517423e-06,
+      "loss": 0.0301,
+      "step": 5954
+    },
+    {
+      "epoch": 2.378594249201278,
+      "grad_norm": 0.9378146372928756,
+      "learning_rate": 1.25128923423325e-06,
+      "loss": 0.028,
+      "step": 5956
+    },
+    {
+      "epoch": 2.3793929712460065,
+      "grad_norm": 0.882404097137147,
+      "learning_rate": 1.2482155782471612e-06,
+      "loss": 0.0289,
+      "step": 5958
+    },
+    {
+      "epoch": 2.380191693290735,
+      "grad_norm": 1.0244194815400358,
+      "learning_rate": 1.2451451634488264e-06,
+      "loss": 0.0319,
+      "step": 5960
+    },
+    {
+      "epoch": 2.380990415335463,
+      "grad_norm": 0.9513392453303519,
+      "learning_rate": 1.242077992490794e-06,
+      "loss": 0.0291,
+      "step": 5962
+    },
+    {
+      "epoch": 2.3817891373801916,
+      "grad_norm": 0.8933825432617372,
+      "learning_rate": 1.2390140680228107e-06,
+      "loss": 0.0324,
+      "step": 5964
+    },
+    {
+      "epoch": 2.38258785942492,
+      "grad_norm": 0.9739451587254204,
+      "learning_rate": 1.2359533926918193e-06,
+      "loss": 0.0325,
+      "step": 5966
+    },
+    {
+      "epoch": 2.3833865814696487,
+      "grad_norm": 1.0284624062364278,
+      "learning_rate": 1.2328959691419517e-06,
+      "loss": 0.0377,
+      "step": 5968
+    },
+    {
+      "epoch": 2.384185303514377,
+      "grad_norm": 0.9808325264631864,
+      "learning_rate": 1.2298418000145345e-06,
+      "loss": 0.0311,
+      "step": 5970
+    },
+    {
+      "epoch": 2.3849840255591053,
+      "grad_norm": 0.9709468792899423,
+      "learning_rate": 1.2267908879480822e-06,
+      "loss": 0.0333,
+      "step": 5972
+    },
+    {
+      "epoch": 2.385782747603834,
+      "grad_norm": 1.033875589705149,
+      "learning_rate": 1.2237432355782947e-06,
+      "loss": 0.0374,
+      "step": 5974
+    },
+    {
+      "epoch": 2.3865814696485623,
+      "grad_norm": 1.078337957714353,
+      "learning_rate": 1.2206988455380558e-06,
+      "loss": 0.0326,
+      "step": 5976
+    },
+    {
+      "epoch": 2.387380191693291,
+      "grad_norm": 1.0758642600387078,
+      "learning_rate": 1.2176577204574318e-06,
+      "loss": 0.0313,
+      "step": 5978
+    },
+    {
+      "epoch": 2.3881789137380194,
+      "grad_norm": 0.901933168392248,
+      "learning_rate": 1.214619862963668e-06,
+      "loss": 0.0319,
+      "step": 5980
+    },
+    {
+      "epoch": 2.3889776357827475,
+      "grad_norm": 1.0229488021571116,
+      "learning_rate": 1.2115852756811875e-06,
+      "loss": 0.0367,
+      "step": 5982
+    },
+    {
+      "epoch": 2.389776357827476,
+      "grad_norm": 0.8928180495678354,
+      "learning_rate": 1.2085539612315844e-06,
+      "loss": 0.0307,
+      "step": 5984
+    },
+    {
+      "epoch": 2.3905750798722045,
+      "grad_norm": 0.9981047761041587,
+      "learning_rate": 1.2055259222336303e-06,
+      "loss": 0.0352,
+      "step": 5986
+    },
+    {
+      "epoch": 2.391373801916933,
+      "grad_norm": 0.7912725756162461,
+      "learning_rate": 1.202501161303265e-06,
+      "loss": 0.0272,
+      "step": 5988
+    },
+    {
+      "epoch": 2.392172523961661,
+      "grad_norm": 0.9543956236269545,
+      "learning_rate": 1.1994796810535981e-06,
+      "loss": 0.0337,
+      "step": 5990
+    },
+    {
+      "epoch": 2.3929712460063897,
+      "grad_norm": 0.8444542368501661,
+      "learning_rate": 1.1964614840949002e-06,
+      "loss": 0.028,
+      "step": 5992
+    },
+    {
+      "epoch": 2.393769968051118,
+      "grad_norm": 0.9428028723910442,
+      "learning_rate": 1.1934465730346106e-06,
+      "loss": 0.0373,
+      "step": 5994
+    },
+    {
+      "epoch": 2.3945686900958467,
+      "grad_norm": 0.8624521055571825,
+      "learning_rate": 1.1904349504773276e-06,
+      "loss": 0.0317,
+      "step": 5996
+    },
+    {
+      "epoch": 2.3953674121405752,
+      "grad_norm": 0.9861824702284993,
+      "learning_rate": 1.1874266190248095e-06,
+      "loss": 0.0334,
+      "step": 5998
+    },
+    {
+      "epoch": 2.3961661341853033,
+      "grad_norm": 1.09155387358314,
+      "learning_rate": 1.1844215812759708e-06,
+      "loss": 0.0349,
+      "step": 6000
+    },
+    {
+      "epoch": 2.3961661341853033,
+      "eval_loss": 0.1782146841287613,
+      "eval_runtime": 418.6183,
+      "eval_samples_per_second": 42.538,
+      "eval_steps_per_second": 5.317,
+      "step": 6000
+    },
+    {
+      "epoch": 2.396964856230032,
+      "grad_norm": 0.8415845821620274,
+      "learning_rate": 1.1814198398268794e-06,
+      "loss": 0.0284,
+      "step": 6002
+    },
+    {
+      "epoch": 2.3977635782747604,
+      "grad_norm": 0.8937417948243025,
+      "learning_rate": 1.1784213972707581e-06,
+      "loss": 0.0294,
+      "step": 6004
+    },
+    {
+      "epoch": 2.398562300319489,
+      "grad_norm": 0.9462060632845546,
+      "learning_rate": 1.175426256197979e-06,
+      "loss": 0.0339,
+      "step": 6006
+    },
+    {
+      "epoch": 2.3993610223642174,
+      "grad_norm": 0.9979933172942926,
+      "learning_rate": 1.1724344191960591e-06,
+      "loss": 0.0329,
+      "step": 6008
+    },
+    {
+      "epoch": 2.4001597444089455,
+      "grad_norm": 0.8510571914760195,
+      "learning_rate": 1.169445888849664e-06,
+      "loss": 0.0285,
+      "step": 6010
+    },
+    {
+      "epoch": 2.400958466453674,
+      "grad_norm": 0.8796317934735823,
+      "learning_rate": 1.1664606677406025e-06,
+      "loss": 0.0285,
+      "step": 6012
+    },
+    {
+      "epoch": 2.4017571884984026,
+      "grad_norm": 1.1059376491146813,
+      "learning_rate": 1.1634787584478257e-06,
+      "loss": 0.0286,
+      "step": 6014
+    },
+    {
+      "epoch": 2.402555910543131,
+      "grad_norm": 1.0839859430790064,
+      "learning_rate": 1.1605001635474183e-06,
+      "loss": 0.031,
+      "step": 6016
+    },
+    {
+      "epoch": 2.4033546325878596,
+      "grad_norm": 1.0568594830906386,
+      "learning_rate": 1.157524885612607e-06,
+      "loss": 0.0336,
+      "step": 6018
+    },
+    {
+      "epoch": 2.4041533546325877,
+      "grad_norm": 0.8544876905209444,
+      "learning_rate": 1.1545529272137496e-06,
+      "loss": 0.0293,
+      "step": 6020
+    },
+    {
+      "epoch": 2.4049520766773163,
+      "grad_norm": 0.9202306605617924,
+      "learning_rate": 1.1515842909183422e-06,
+      "loss": 0.0305,
+      "step": 6022
+    },
+    {
+      "epoch": 2.405750798722045,
+      "grad_norm": 1.2779812066659912,
+      "learning_rate": 1.1486189792910024e-06,
+      "loss": 0.0327,
+      "step": 6024
+    },
+    {
+      "epoch": 2.4065495207667733,
+      "grad_norm": 0.8423188795687913,
+      "learning_rate": 1.1456569948934804e-06,
+      "loss": 0.0323,
+      "step": 6026
+    },
+    {
+      "epoch": 2.4073482428115014,
+      "grad_norm": 0.9219710260036055,
+      "learning_rate": 1.142698340284652e-06,
+      "loss": 0.0367,
+      "step": 6028
+    },
+    {
+      "epoch": 2.40814696485623,
+      "grad_norm": 0.8836691940601481,
+      "learning_rate": 1.139743018020517e-06,
+      "loss": 0.031,
+      "step": 6030
+    },
+    {
+      "epoch": 2.4089456869009584,
+      "grad_norm": 1.009462941023179,
+      "learning_rate": 1.1367910306541918e-06,
+      "loss": 0.0313,
+      "step": 6032
+    },
+    {
+      "epoch": 2.409744408945687,
+      "grad_norm": 0.9333397367826699,
+      "learning_rate": 1.133842380735916e-06,
+      "loss": 0.0289,
+      "step": 6034
+    },
+    {
+      "epoch": 2.4105431309904155,
+      "grad_norm": 0.9518139468601376,
+      "learning_rate": 1.1308970708130458e-06,
+      "loss": 0.0329,
+      "step": 6036
+    },
+    {
+      "epoch": 2.4113418530351436,
+      "grad_norm": 1.0093135425496413,
+      "learning_rate": 1.1279551034300523e-06,
+      "loss": 0.0293,
+      "step": 6038
+    },
+    {
+      "epoch": 2.412140575079872,
+      "grad_norm": 0.8766410436250559,
+      "learning_rate": 1.1250164811285148e-06,
+      "loss": 0.0314,
+      "step": 6040
+    },
+    {
+      "epoch": 2.4129392971246006,
+      "grad_norm": 0.7888244567724153,
+      "learning_rate": 1.1220812064471248e-06,
+      "loss": 0.0302,
+      "step": 6042
+    },
+    {
+      "epoch": 2.413738019169329,
+      "grad_norm": 0.9752687497406552,
+      "learning_rate": 1.119149281921687e-06,
+      "loss": 0.0338,
+      "step": 6044
+    },
+    {
+      "epoch": 2.4145367412140573,
+      "grad_norm": 1.015706721424374,
+      "learning_rate": 1.1162207100851069e-06,
+      "loss": 0.0316,
+      "step": 6046
+    },
+    {
+      "epoch": 2.415335463258786,
+      "grad_norm": 0.9186099679113181,
+      "learning_rate": 1.1132954934673911e-06,
+      "loss": 0.0348,
+      "step": 6048
+    },
+    {
+      "epoch": 2.4161341853035143,
+      "grad_norm": 0.9014882003540872,
+      "learning_rate": 1.110373634595653e-06,
+      "loss": 0.0279,
+      "step": 6050
+    },
+    {
+      "epoch": 2.416932907348243,
+      "grad_norm": 0.8554968617937979,
+      "learning_rate": 1.1074551359941022e-06,
+      "loss": 0.0274,
+      "step": 6052
+    },
+    {
+      "epoch": 2.4177316293929714,
+      "grad_norm": 1.023254004503027,
+      "learning_rate": 1.1045400001840474e-06,
+      "loss": 0.0346,
+      "step": 6054
+    },
+    {
+      "epoch": 2.4185303514377,
+      "grad_norm": 1.0199229730833819,
+      "learning_rate": 1.1016282296838887e-06,
+      "loss": 0.0372,
+      "step": 6056
+    },
+    {
+      "epoch": 2.419329073482428,
+      "grad_norm": 0.8141953630986705,
+      "learning_rate": 1.0987198270091225e-06,
+      "loss": 0.0297,
+      "step": 6058
+    },
+    {
+      "epoch": 2.4201277955271565,
+      "grad_norm": 0.9824196187903327,
+      "learning_rate": 1.0958147946723341e-06,
+      "loss": 0.0295,
+      "step": 6060
+    },
+    {
+      "epoch": 2.420926517571885,
+      "grad_norm": 1.116417391297416,
+      "learning_rate": 1.0929131351831974e-06,
+      "loss": 0.0339,
+      "step": 6062
+    },
+    {
+      "epoch": 2.4217252396166136,
+      "grad_norm": 0.9588644560235544,
+      "learning_rate": 1.090014851048473e-06,
+      "loss": 0.0322,
+      "step": 6064
+    },
+    {
+      "epoch": 2.4225239616613417,
+      "grad_norm": 0.876473284974182,
+      "learning_rate": 1.0871199447720022e-06,
+      "loss": 0.0337,
+      "step": 6066
+    },
+    {
+      "epoch": 2.42332268370607,
+      "grad_norm": 0.8818096541129938,
+      "learning_rate": 1.0842284188547142e-06,
+      "loss": 0.0299,
+      "step": 6068
+    },
+    {
+      "epoch": 2.4241214057507987,
+      "grad_norm": 0.9093312218826864,
+      "learning_rate": 1.0813402757946145e-06,
+      "loss": 0.0326,
+      "step": 6070
+    },
+    {
+      "epoch": 2.4249201277955272,
+      "grad_norm": 0.8911104729900233,
+      "learning_rate": 1.078455518086784e-06,
+      "loss": 0.0311,
+      "step": 6072
+    },
+    {
+      "epoch": 2.4257188498402558,
+      "grad_norm": 0.9634405906730059,
+      "learning_rate": 1.0755741482233822e-06,
+      "loss": 0.0321,
+      "step": 6074
+    },
+    {
+      "epoch": 2.426517571884984,
+      "grad_norm": 0.8619867335268869,
+      "learning_rate": 1.0726961686936406e-06,
+      "loss": 0.0274,
+      "step": 6076
+    },
+    {
+      "epoch": 2.4273162939297124,
+      "grad_norm": 0.8467036432543594,
+      "learning_rate": 1.069821581983862e-06,
+      "loss": 0.0308,
+      "step": 6078
+    },
+    {
+      "epoch": 2.428115015974441,
+      "grad_norm": 0.9553299696013161,
+      "learning_rate": 1.0669503905774198e-06,
+      "loss": 0.0305,
+      "step": 6080
+    },
+    {
+      "epoch": 2.4289137380191694,
+      "grad_norm": 0.9434220839422669,
+      "learning_rate": 1.0640825969547498e-06,
+      "loss": 0.032,
+      "step": 6082
+    },
+    {
+      "epoch": 2.4297124600638975,
+      "grad_norm": 0.907723226631111,
+      "learning_rate": 1.061218203593356e-06,
+      "loss": 0.0292,
+      "step": 6084
+    },
+    {
+      "epoch": 2.430511182108626,
+      "grad_norm": 1.0082338096229233,
+      "learning_rate": 1.0583572129678043e-06,
+      "loss": 0.0345,
+      "step": 6086
+    },
+    {
+      "epoch": 2.4313099041533546,
+      "grad_norm": 0.9128983473815063,
+      "learning_rate": 1.055499627549722e-06,
+      "loss": 0.0312,
+      "step": 6088
+    },
+    {
+      "epoch": 2.432108626198083,
+      "grad_norm": 0.9047232760560915,
+      "learning_rate": 1.0526454498077892e-06,
+      "loss": 0.0333,
+      "step": 6090
+    },
+    {
+      "epoch": 2.4329073482428116,
+      "grad_norm": 0.9748694127638722,
+      "learning_rate": 1.0497946822077504e-06,
+      "loss": 0.0301,
+      "step": 6092
+    },
+    {
+      "epoch": 2.43370607028754,
+      "grad_norm": 0.9900482589709648,
+      "learning_rate": 1.0469473272123998e-06,
+      "loss": 0.0312,
+      "step": 6094
+    },
+    {
+      "epoch": 2.4345047923322682,
+      "grad_norm": 1.0325102686584386,
+      "learning_rate": 1.0441033872815804e-06,
+      "loss": 0.0334,
+      "step": 6096
+    },
+    {
+      "epoch": 2.4353035143769968,
+      "grad_norm": 0.8638115070891277,
+      "learning_rate": 1.0412628648721895e-06,
+      "loss": 0.0296,
+      "step": 6098
+    },
+    {
+      "epoch": 2.4361022364217253,
+      "grad_norm": 1.1255119248973784,
+      "learning_rate": 1.0384257624381705e-06,
+      "loss": 0.0305,
+      "step": 6100
+    },
+    {
+      "epoch": 2.436900958466454,
+      "grad_norm": 0.9446843609413517,
+      "learning_rate": 1.0355920824305127e-06,
+      "loss": 0.0323,
+      "step": 6102
+    },
+    {
+      "epoch": 2.437699680511182,
+      "grad_norm": 0.9313934818835989,
+      "learning_rate": 1.0327618272972484e-06,
+      "loss": 0.0322,
+      "step": 6104
+    },
+    {
+      "epoch": 2.4384984025559104,
+      "grad_norm": 1.1986399439946025,
+      "learning_rate": 1.0299349994834497e-06,
+      "loss": 0.0369,
+      "step": 6106
+    },
+    {
+      "epoch": 2.439297124600639,
+      "grad_norm": 1.0849811906791489,
+      "learning_rate": 1.0271116014312293e-06,
+      "loss": 0.0274,
+      "step": 6108
+    },
+    {
+      "epoch": 2.4400958466453675,
+      "grad_norm": 0.9885919632765612,
+      "learning_rate": 1.0242916355797372e-06,
+      "loss": 0.0294,
+      "step": 6110
+    },
+    {
+      "epoch": 2.440894568690096,
+      "grad_norm": 1.1666414296370116,
+      "learning_rate": 1.0214751043651582e-06,
+      "loss": 0.0327,
+      "step": 6112
+    },
+    {
+      "epoch": 2.441693290734824,
+      "grad_norm": 1.0373974615204293,
+      "learning_rate": 1.018662010220709e-06,
+      "loss": 0.0345,
+      "step": 6114
+    },
+    {
+      "epoch": 2.4424920127795526,
+      "grad_norm": 0.8806182333785598,
+      "learning_rate": 1.0158523555766375e-06,
+      "loss": 0.03,
+      "step": 6116
+    },
+    {
+      "epoch": 2.443290734824281,
+      "grad_norm": 0.8377676265174371,
+      "learning_rate": 1.0130461428602206e-06,
+      "loss": 0.0286,
+      "step": 6118
+    },
+    {
+      "epoch": 2.4440894568690097,
+      "grad_norm": 0.9542688751670757,
+      "learning_rate": 1.010243374495763e-06,
+      "loss": 0.0296,
+      "step": 6120
+    },
+    {
+      "epoch": 2.4448881789137378,
+      "grad_norm": 0.845828630003871,
+      "learning_rate": 1.0074440529045882e-06,
+      "loss": 0.0299,
+      "step": 6122
+    },
+    {
+      "epoch": 2.4456869009584663,
+      "grad_norm": 0.838659411317932,
+      "learning_rate": 1.0046481805050484e-06,
+      "loss": 0.0286,
+      "step": 6124
+    },
+    {
+      "epoch": 2.446485623003195,
+      "grad_norm": 0.964366696594686,
+      "learning_rate": 1.001855759712513e-06,
+      "loss": 0.0309,
+      "step": 6126
+    },
+    {
+      "epoch": 2.4472843450479234,
+      "grad_norm": 1.0550865306282715,
+      "learning_rate": 9.990667929393715e-07,
+      "loss": 0.0357,
+      "step": 6128
+    },
+    {
+      "epoch": 2.448083067092652,
+      "grad_norm": 0.9665555146057448,
+      "learning_rate": 9.962812825950252e-07,
+      "loss": 0.0334,
+      "step": 6130
+    },
+    {
+      "epoch": 2.4488817891373804,
+      "grad_norm": 1.1013658647610531,
+      "learning_rate": 9.934992310858944e-07,
+      "loss": 0.0358,
+      "step": 6132
+    },
+    {
+      "epoch": 2.4496805111821085,
+      "grad_norm": 0.9451008760428278,
+      "learning_rate": 9.90720640815408e-07,
+      "loss": 0.0319,
+      "step": 6134
+    },
+    {
+      "epoch": 2.450479233226837,
+      "grad_norm": 0.9344422669847151,
+      "learning_rate": 9.879455141840067e-07,
+      "loss": 0.0298,
+      "step": 6136
+    },
+    {
+      "epoch": 2.4512779552715656,
+      "grad_norm": 0.8508316348278806,
+      "learning_rate": 9.851738535891375e-07,
+      "loss": 0.0266,
+      "step": 6138
+    },
+    {
+      "epoch": 2.452076677316294,
+      "grad_norm": 0.9862824857730245,
+      "learning_rate": 9.824056614252542e-07,
+      "loss": 0.0314,
+      "step": 6140
+    },
+    {
+      "epoch": 2.452875399361022,
+      "grad_norm": 0.8320616360939587,
+      "learning_rate": 9.79640940083813e-07,
+      "loss": 0.0299,
+      "step": 6142
+    },
+    {
+      "epoch": 2.4536741214057507,
+      "grad_norm": 0.9792614234160211,
+      "learning_rate": 9.768796919532742e-07,
+      "loss": 0.0314,
+      "step": 6144
+    },
+    {
+      "epoch": 2.4544728434504792,
+      "grad_norm": 0.9293003315650749,
+      "learning_rate": 9.741219194190925e-07,
+      "loss": 0.0338,
+      "step": 6146
+    },
+    {
+      "epoch": 2.4552715654952078,
+      "grad_norm": 0.9682041377572445,
+      "learning_rate": 9.71367624863725e-07,
+      "loss": 0.0329,
+      "step": 6148
+    },
+    {
+      "epoch": 2.4560702875399363,
+      "grad_norm": 0.8325832618656944,
+      "learning_rate": 9.686168106666216e-07,
+      "loss": 0.031,
+      "step": 6150
+    },
+    {
+      "epoch": 2.4568690095846644,
+      "grad_norm": 0.9350572549104323,
+      "learning_rate": 9.658694792042284e-07,
+      "loss": 0.0288,
+      "step": 6152
+    },
+    {
+      "epoch": 2.457667731629393,
+      "grad_norm": 0.8277288114049663,
+      "learning_rate": 9.631256328499772e-07,
+      "loss": 0.031,
+      "step": 6154
+    },
+    {
+      "epoch": 2.4584664536741214,
+      "grad_norm": 0.9213528535489943,
+      "learning_rate": 9.603852739742941e-07,
+      "loss": 0.0284,
+      "step": 6156
+    },
+    {
+      "epoch": 2.45926517571885,
+      "grad_norm": 0.8009301927639629,
+      "learning_rate": 9.576484049445895e-07,
+      "loss": 0.029,
+      "step": 6158
+    },
+    {
+      "epoch": 2.460063897763578,
+      "grad_norm": 0.8921062677918278,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.0354,
+      "step": 6160
+    },
+    {
+      "epoch": 2.4608626198083066,
+      "grad_norm": 0.8850346677550753,
+      "learning_rate": 9.521851458776915e-07,
+      "loss": 0.0297,
+      "step": 6162
+    },
+    {
+      "epoch": 2.461661341853035,
+      "grad_norm": 0.9779735465184336,
+      "learning_rate": 9.494587605602368e-07,
+      "loss": 0.0289,
+      "step": 6164
+    },
+    {
+      "epoch": 2.4624600638977636,
+      "grad_norm": 0.9498573454203734,
+      "learning_rate": 9.467358745282379e-07,
+      "loss": 0.0279,
+      "step": 6166
+    },
+    {
+      "epoch": 2.463258785942492,
+      "grad_norm": 0.9196211437633159,
+      "learning_rate": 9.440164901340127e-07,
+      "loss": 0.0311,
+      "step": 6168
+    },
+    {
+      "epoch": 2.4640575079872207,
+      "grad_norm": 0.8765231457695842,
+      "learning_rate": 9.413006097268512e-07,
+      "loss": 0.0243,
+      "step": 6170
+    },
+    {
+      "epoch": 2.4648562300319488,
+      "grad_norm": 0.853315794604095,
+      "learning_rate": 9.385882356530179e-07,
+      "loss": 0.026,
+      "step": 6172
+    },
+    {
+      "epoch": 2.4656549520766773,
+      "grad_norm": 0.9844044641457295,
+      "learning_rate": 9.358793702557489e-07,
+      "loss": 0.0293,
+      "step": 6174
+    },
+    {
+      "epoch": 2.466453674121406,
+      "grad_norm": 0.9636119419245541,
+      "learning_rate": 9.331740158752495e-07,
+      "loss": 0.0334,
+      "step": 6176
+    },
+    {
+      "epoch": 2.4672523961661343,
+      "grad_norm": 1.1029243201298184,
+      "learning_rate": 9.304721748486878e-07,
+      "loss": 0.0344,
+      "step": 6178
+    },
+    {
+      "epoch": 2.4680511182108624,
+      "grad_norm": 0.9591374043106669,
+      "learning_rate": 9.277738495102012e-07,
+      "loss": 0.0308,
+      "step": 6180
+    },
+    {
+      "epoch": 2.468849840255591,
+      "grad_norm": 0.9656785120322073,
+      "learning_rate": 9.250790421908862e-07,
+      "loss": 0.0261,
+      "step": 6182
+    },
+    {
+      "epoch": 2.4696485623003195,
+      "grad_norm": 1.0686970243259972,
+      "learning_rate": 9.223877552188065e-07,
+      "loss": 0.0325,
+      "step": 6184
+    },
+    {
+      "epoch": 2.470447284345048,
+      "grad_norm": 0.9422767101505445,
+      "learning_rate": 9.196999909189764e-07,
+      "loss": 0.0293,
+      "step": 6186
+    },
+    {
+      "epoch": 2.4712460063897765,
+      "grad_norm": 0.9128608441727557,
+      "learning_rate": 9.17015751613371e-07,
+      "loss": 0.0278,
+      "step": 6188
+    },
+    {
+      "epoch": 2.4720447284345046,
+      "grad_norm": 1.0182227621831177,
+      "learning_rate": 9.14335039620921e-07,
+      "loss": 0.0291,
+      "step": 6190
+    },
+    {
+      "epoch": 2.472843450479233,
+      "grad_norm": 1.0323032210841876,
+      "learning_rate": 9.116578572575091e-07,
+      "loss": 0.0381,
+      "step": 6192
+    },
+    {
+      "epoch": 2.4736421725239617,
+      "grad_norm": 1.007977038062956,
+      "learning_rate": 9.089842068359661e-07,
+      "loss": 0.0328,
+      "step": 6194
+    },
+    {
+      "epoch": 2.47444089456869,
+      "grad_norm": 1.0276724023266843,
+      "learning_rate": 9.06314090666075e-07,
+      "loss": 0.0353,
+      "step": 6196
+    },
+    {
+      "epoch": 2.4752396166134183,
+      "grad_norm": 1.033358203994531,
+      "learning_rate": 9.03647511054564e-07,
+      "loss": 0.0291,
+      "step": 6198
+    },
+    {
+      "epoch": 2.476038338658147,
+      "grad_norm": 0.8075474991970624,
+      "learning_rate": 9.009844703051063e-07,
+      "loss": 0.0276,
+      "step": 6200
+    },
+    {
+      "epoch": 2.4768370607028753,
+      "grad_norm": 0.9186532720625925,
+      "learning_rate": 8.98324970718319e-07,
+      "loss": 0.0283,
+      "step": 6202
+    },
+    {
+      "epoch": 2.477635782747604,
+      "grad_norm": 0.9745581052474273,
+      "learning_rate": 8.956690145917557e-07,
+      "loss": 0.0337,
+      "step": 6204
+    },
+    {
+      "epoch": 2.4784345047923324,
+      "grad_norm": 0.8345103534692212,
+      "learning_rate": 8.930166042199146e-07,
+      "loss": 0.0256,
+      "step": 6206
+    },
+    {
+      "epoch": 2.479233226837061,
+      "grad_norm": 0.9005598945512714,
+      "learning_rate": 8.903677418942292e-07,
+      "loss": 0.0285,
+      "step": 6208
+    },
+    {
+      "epoch": 2.480031948881789,
+      "grad_norm": 1.0271916802595848,
+      "learning_rate": 8.877224299030629e-07,
+      "loss": 0.0317,
+      "step": 6210
+    },
+    {
+      "epoch": 2.4808306709265175,
+      "grad_norm": 1.075693558366458,
+      "learning_rate": 8.850806705317183e-07,
+      "loss": 0.0368,
+      "step": 6212
+    },
+    {
+      "epoch": 2.481629392971246,
+      "grad_norm": 0.8722330378555192,
+      "learning_rate": 8.824424660624247e-07,
+      "loss": 0.029,
+      "step": 6214
+    },
+    {
+      "epoch": 2.4824281150159746,
+      "grad_norm": 1.0601895552621947,
+      "learning_rate": 8.79807818774343e-07,
+      "loss": 0.0293,
+      "step": 6216
+    },
+    {
+      "epoch": 2.4832268370607027,
+      "grad_norm": 0.8707900195092941,
+      "learning_rate": 8.771767309435614e-07,
+      "loss": 0.0304,
+      "step": 6218
+    },
+    {
+      "epoch": 2.484025559105431,
+      "grad_norm": 0.9733820547813553,
+      "learning_rate": 8.745492048430876e-07,
+      "loss": 0.0343,
+      "step": 6220
+    },
+    {
+      "epoch": 2.4848242811501597,
+      "grad_norm": 0.9317695338486248,
+      "learning_rate": 8.719252427428582e-07,
+      "loss": 0.0303,
+      "step": 6222
+    },
+    {
+      "epoch": 2.4856230031948883,
+      "grad_norm": 1.033404563883472,
+      "learning_rate": 8.693048469097293e-07,
+      "loss": 0.0325,
+      "step": 6224
+    },
+    {
+      "epoch": 2.486421725239617,
+      "grad_norm": 0.9136265608554973,
+      "learning_rate": 8.666880196074767e-07,
+      "loss": 0.0308,
+      "step": 6226
+    },
+    {
+      "epoch": 2.487220447284345,
+      "grad_norm": 1.1596479673050444,
+      "learning_rate": 8.640747630967883e-07,
+      "loss": 0.0294,
+      "step": 6228
+    },
+    {
+      "epoch": 2.4880191693290734,
+      "grad_norm": 1.0426530374304792,
+      "learning_rate": 8.614650796352747e-07,
+      "loss": 0.0306,
+      "step": 6230
+    },
+    {
+      "epoch": 2.488817891373802,
+      "grad_norm": 0.9242364015680927,
+      "learning_rate": 8.58858971477457e-07,
+      "loss": 0.0327,
+      "step": 6232
+    },
+    {
+      "epoch": 2.4896166134185305,
+      "grad_norm": 0.9313580110161238,
+      "learning_rate": 8.562564408747637e-07,
+      "loss": 0.0306,
+      "step": 6234
+    },
+    {
+      "epoch": 2.4904153354632586,
+      "grad_norm": 1.067403189080394,
+      "learning_rate": 8.536574900755367e-07,
+      "loss": 0.0317,
+      "step": 6236
+    },
+    {
+      "epoch": 2.491214057507987,
+      "grad_norm": 0.9285372176981973,
+      "learning_rate": 8.510621213250248e-07,
+      "loss": 0.0305,
+      "step": 6238
+    },
+    {
+      "epoch": 2.4920127795527156,
+      "grad_norm": 0.8346074025633845,
+      "learning_rate": 8.484703368653812e-07,
+      "loss": 0.0273,
+      "step": 6240
+    },
+    {
+      "epoch": 2.492811501597444,
+      "grad_norm": 0.9152125909961968,
+      "learning_rate": 8.458821389356647e-07,
+      "loss": 0.0303,
+      "step": 6242
+    },
+    {
+      "epoch": 2.4936102236421727,
+      "grad_norm": 0.9838042565841876,
+      "learning_rate": 8.432975297718321e-07,
+      "loss": 0.0299,
+      "step": 6244
+    },
+    {
+      "epoch": 2.494408945686901,
+      "grad_norm": 0.9237463657264794,
+      "learning_rate": 8.407165116067423e-07,
+      "loss": 0.0304,
+      "step": 6246
+    },
+    {
+      "epoch": 2.4952076677316293,
+      "grad_norm": 1.1889649482379565,
+      "learning_rate": 8.381390866701517e-07,
+      "loss": 0.0354,
+      "step": 6248
+    },
+    {
+      "epoch": 2.496006389776358,
+      "grad_norm": 0.9830687823209033,
+      "learning_rate": 8.355652571887135e-07,
+      "loss": 0.0317,
+      "step": 6250
+    },
+    {
+      "epoch": 2.4968051118210863,
+      "grad_norm": 0.8992861344610856,
+      "learning_rate": 8.329950253859703e-07,
+      "loss": 0.0299,
+      "step": 6252
+    },
+    {
+      "epoch": 2.497603833865815,
+      "grad_norm": 1.0201169404321446,
+      "learning_rate": 8.304283934823626e-07,
+      "loss": 0.0337,
+      "step": 6254
+    },
+    {
+      "epoch": 2.498402555910543,
+      "grad_norm": 0.996273309697261,
+      "learning_rate": 8.278653636952177e-07,
+      "loss": 0.0309,
+      "step": 6256
+    },
+    {
+      "epoch": 2.4992012779552715,
+      "grad_norm": 0.8898734510036184,
+      "learning_rate": 8.25305938238753e-07,
+      "loss": 0.031,
+      "step": 6258
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.8809017767144187,
+      "learning_rate": 8.227501193240673e-07,
+      "loss": 0.0344,
+      "step": 6260
+    },
+    {
+      "epoch": 2.5007987220447285,
+      "grad_norm": 0.9543528237861113,
+      "learning_rate": 8.201979091591488e-07,
+      "loss": 0.0282,
+      "step": 6262
+    },
+    {
+      "epoch": 2.501597444089457,
+      "grad_norm": 1.111241667863821,
+      "learning_rate": 8.176493099488664e-07,
+      "loss": 0.0358,
+      "step": 6264
+    },
+    {
+      "epoch": 2.502396166134185,
+      "grad_norm": 0.869787895901821,
+      "learning_rate": 8.151043238949697e-07,
+      "loss": 0.0272,
+      "step": 6266
+    },
+    {
+      "epoch": 2.5031948881789137,
+      "grad_norm": 0.9762765697943674,
+      "learning_rate": 8.125629531960849e-07,
+      "loss": 0.0299,
+      "step": 6268
+    },
+    {
+      "epoch": 2.503993610223642,
+      "grad_norm": 0.9008988645129132,
+      "learning_rate": 8.100252000477177e-07,
+      "loss": 0.0295,
+      "step": 6270
+    },
+    {
+      "epoch": 2.5047923322683707,
+      "grad_norm": 0.9858370192539189,
+      "learning_rate": 8.074910666422475e-07,
+      "loss": 0.0316,
+      "step": 6272
+    },
+    {
+      "epoch": 2.505591054313099,
+      "grad_norm": 0.9950028382350193,
+      "learning_rate": 8.049605551689255e-07,
+      "loss": 0.0314,
+      "step": 6274
+    },
+    {
+      "epoch": 2.5063897763578273,
+      "grad_norm": 0.8706385388900013,
+      "learning_rate": 8.024336678138761e-07,
+      "loss": 0.03,
+      "step": 6276
+    },
+    {
+      "epoch": 2.507188498402556,
+      "grad_norm": 0.9317584954844236,
+      "learning_rate": 7.999104067600904e-07,
+      "loss": 0.0319,
+      "step": 6278
+    },
+    {
+      "epoch": 2.5079872204472844,
+      "grad_norm": 0.9028386865531697,
+      "learning_rate": 7.973907741874287e-07,
+      "loss": 0.0319,
+      "step": 6280
+    },
+    {
+      "epoch": 2.508785942492013,
+      "grad_norm": 1.0429573535057794,
+      "learning_rate": 7.948747722726169e-07,
+      "loss": 0.0362,
+      "step": 6282
+    },
+    {
+      "epoch": 2.5095846645367414,
+      "grad_norm": 0.9075149519955287,
+      "learning_rate": 7.923624031892402e-07,
+      "loss": 0.0297,
+      "step": 6284
+    },
+    {
+      "epoch": 2.5103833865814695,
+      "grad_norm": 0.9244441640851673,
+      "learning_rate": 7.898536691077508e-07,
+      "loss": 0.0356,
+      "step": 6286
+    },
+    {
+      "epoch": 2.511182108626198,
+      "grad_norm": 0.9718862084825739,
+      "learning_rate": 7.873485721954572e-07,
+      "loss": 0.0327,
+      "step": 6288
+    },
+    {
+      "epoch": 2.5119808306709266,
+      "grad_norm": 0.8694408724126972,
+      "learning_rate": 7.848471146165287e-07,
+      "loss": 0.0297,
+      "step": 6290
+    },
+    {
+      "epoch": 2.512779552715655,
+      "grad_norm": 0.8938047550855578,
+      "learning_rate": 7.823492985319858e-07,
+      "loss": 0.0275,
+      "step": 6292
+    },
+    {
+      "epoch": 2.513578274760383,
+      "grad_norm": 0.8065785132438869,
+      "learning_rate": 7.798551260997067e-07,
+      "loss": 0.0312,
+      "step": 6294
+    },
+    {
+      "epoch": 2.5143769968051117,
+      "grad_norm": 0.929264264045834,
+      "learning_rate": 7.773645994744222e-07,
+      "loss": 0.0256,
+      "step": 6296
+    },
+    {
+      "epoch": 2.5151757188498403,
+      "grad_norm": 1.3557337175166702,
+      "learning_rate": 7.748777208077118e-07,
+      "loss": 0.0288,
+      "step": 6298
+    },
+    {
+      "epoch": 2.515974440894569,
+      "grad_norm": 1.1432470693302257,
+      "learning_rate": 7.723944922480037e-07,
+      "loss": 0.0343,
+      "step": 6300
+    },
+    {
+      "epoch": 2.5167731629392973,
+      "grad_norm": 1.1285169628245737,
+      "learning_rate": 7.699149159405734e-07,
+      "loss": 0.0344,
+      "step": 6302
+    },
+    {
+      "epoch": 2.5175718849840254,
+      "grad_norm": 0.9264387543909826,
+      "learning_rate": 7.674389940275406e-07,
+      "loss": 0.0299,
+      "step": 6304
+    },
+    {
+      "epoch": 2.518370607028754,
+      "grad_norm": 0.9885710066492656,
+      "learning_rate": 7.649667286478696e-07,
+      "loss": 0.0349,
+      "step": 6306
+    },
+    {
+      "epoch": 2.5191693290734825,
+      "grad_norm": 0.8115719655821809,
+      "learning_rate": 7.624981219373623e-07,
+      "loss": 0.0301,
+      "step": 6308
+    },
+    {
+      "epoch": 2.519968051118211,
+      "grad_norm": 0.8265082284522801,
+      "learning_rate": 7.600331760286627e-07,
+      "loss": 0.0291,
+      "step": 6310
+    },
+    {
+      "epoch": 2.520766773162939,
+      "grad_norm": 0.8510973944883193,
+      "learning_rate": 7.575718930512516e-07,
+      "loss": 0.0259,
+      "step": 6312
+    },
+    {
+      "epoch": 2.5215654952076676,
+      "grad_norm": 0.9955720237715131,
+      "learning_rate": 7.551142751314455e-07,
+      "loss": 0.0309,
+      "step": 6314
+    },
+    {
+      "epoch": 2.522364217252396,
+      "grad_norm": 0.94544787132954,
+      "learning_rate": 7.526603243923958e-07,
+      "loss": 0.0278,
+      "step": 6316
+    },
+    {
+      "epoch": 2.5231629392971247,
+      "grad_norm": 0.7689821541255585,
+      "learning_rate": 7.502100429540815e-07,
+      "loss": 0.0283,
+      "step": 6318
+    },
+    {
+      "epoch": 2.523961661341853,
+      "grad_norm": 0.8326325372584937,
+      "learning_rate": 7.47763432933315e-07,
+      "loss": 0.0288,
+      "step": 6320
+    },
+    {
+      "epoch": 2.5247603833865817,
+      "grad_norm": 0.8468951802399899,
+      "learning_rate": 7.453204964437394e-07,
+      "loss": 0.0261,
+      "step": 6322
+    },
+    {
+      "epoch": 2.52555910543131,
+      "grad_norm": 0.9023769818457569,
+      "learning_rate": 7.428812355958181e-07,
+      "loss": 0.0316,
+      "step": 6324
+    },
+    {
+      "epoch": 2.5263578274760383,
+      "grad_norm": 0.7923001953812997,
+      "learning_rate": 7.404456524968445e-07,
+      "loss": 0.0246,
+      "step": 6326
+    },
+    {
+      "epoch": 2.527156549520767,
+      "grad_norm": 0.8829690105277411,
+      "learning_rate": 7.380137492509309e-07,
+      "loss": 0.0262,
+      "step": 6328
+    },
+    {
+      "epoch": 2.527955271565495,
+      "grad_norm": 0.7804684297568552,
+      "learning_rate": 7.355855279590146e-07,
+      "loss": 0.0243,
+      "step": 6330
+    },
+    {
+      "epoch": 2.5287539936102235,
+      "grad_norm": 1.0507641034180348,
+      "learning_rate": 7.33160990718847e-07,
+      "loss": 0.0318,
+      "step": 6332
+    },
+    {
+      "epoch": 2.529552715654952,
+      "grad_norm": 0.9578888020234547,
+      "learning_rate": 7.307401396250008e-07,
+      "loss": 0.0255,
+      "step": 6334
+    },
+    {
+      "epoch": 2.5303514376996805,
+      "grad_norm": 0.9269807171184239,
+      "learning_rate": 7.283229767688627e-07,
+      "loss": 0.0308,
+      "step": 6336
+    },
+    {
+      "epoch": 2.531150159744409,
+      "grad_norm": 0.8790871744926378,
+      "learning_rate": 7.259095042386338e-07,
+      "loss": 0.0301,
+      "step": 6338
+    },
+    {
+      "epoch": 2.5319488817891376,
+      "grad_norm": 0.9895066251185489,
+      "learning_rate": 7.23499724119327e-07,
+      "loss": 0.0291,
+      "step": 6340
+    },
+    {
+      "epoch": 2.5327476038338657,
+      "grad_norm": 0.8899644518567821,
+      "learning_rate": 7.210936384927631e-07,
+      "loss": 0.0311,
+      "step": 6342
+    },
+    {
+      "epoch": 2.533546325878594,
+      "grad_norm": 1.0693235257658584,
+      "learning_rate": 7.186912494375736e-07,
+      "loss": 0.0323,
+      "step": 6344
+    },
+    {
+      "epoch": 2.5343450479233227,
+      "grad_norm": 1.044274627458564,
+      "learning_rate": 7.162925590291986e-07,
+      "loss": 0.0294,
+      "step": 6346
+    },
+    {
+      "epoch": 2.5351437699680512,
+      "grad_norm": 0.9703071429232554,
+      "learning_rate": 7.13897569339877e-07,
+      "loss": 0.0292,
+      "step": 6348
+    },
+    {
+      "epoch": 2.5359424920127793,
+      "grad_norm": 0.9567409636353577,
+      "learning_rate": 7.115062824386554e-07,
+      "loss": 0.0262,
+      "step": 6350
+    },
+    {
+      "epoch": 2.536741214057508,
+      "grad_norm": 0.8461463315662133,
+      "learning_rate": 7.091187003913802e-07,
+      "loss": 0.0298,
+      "step": 6352
+    },
+    {
+      "epoch": 2.5375399361022364,
+      "grad_norm": 0.9718174385536262,
+      "learning_rate": 7.067348252606965e-07,
+      "loss": 0.031,
+      "step": 6354
+    },
+    {
+      "epoch": 2.538338658146965,
+      "grad_norm": 0.8730259562016701,
+      "learning_rate": 7.043546591060485e-07,
+      "loss": 0.0299,
+      "step": 6356
+    },
+    {
+      "epoch": 2.5391373801916934,
+      "grad_norm": 0.9333674604962541,
+      "learning_rate": 7.019782039836737e-07,
+      "loss": 0.0295,
+      "step": 6358
+    },
+    {
+      "epoch": 2.539936102236422,
+      "grad_norm": 0.888933835128418,
+      "learning_rate": 6.996054619466053e-07,
+      "loss": 0.0322,
+      "step": 6360
+    },
+    {
+      "epoch": 2.54073482428115,
+      "grad_norm": 0.9530139786909172,
+      "learning_rate": 6.972364350446698e-07,
+      "loss": 0.032,
+      "step": 6362
+    },
+    {
+      "epoch": 2.5415335463258786,
+      "grad_norm": 0.8887252902742203,
+      "learning_rate": 6.948711253244827e-07,
+      "loss": 0.0312,
+      "step": 6364
+    },
+    {
+      "epoch": 2.542332268370607,
+      "grad_norm": 0.8352734988241693,
+      "learning_rate": 6.92509534829447e-07,
+      "loss": 0.0272,
+      "step": 6366
+    },
+    {
+      "epoch": 2.543130990415335,
+      "grad_norm": 0.9214285326384319,
+      "learning_rate": 6.901516655997536e-07,
+      "loss": 0.0279,
+      "step": 6368
+    },
+    {
+      "epoch": 2.5439297124600637,
+      "grad_norm": 1.0207519839454122,
+      "learning_rate": 6.877975196723824e-07,
+      "loss": 0.0295,
+      "step": 6370
+    },
+    {
+      "epoch": 2.5447284345047922,
+      "grad_norm": 1.0846736644854258,
+      "learning_rate": 6.854470990810907e-07,
+      "loss": 0.0371,
+      "step": 6372
+    },
+    {
+      "epoch": 2.5455271565495208,
+      "grad_norm": 0.929449666083272,
+      "learning_rate": 6.831004058564211e-07,
+      "loss": 0.03,
+      "step": 6374
+    },
+    {
+      "epoch": 2.5463258785942493,
+      "grad_norm": 0.9865284947555943,
+      "learning_rate": 6.80757442025694e-07,
+      "loss": 0.0311,
+      "step": 6376
+    },
+    {
+      "epoch": 2.547124600638978,
+      "grad_norm": 0.9826417414313666,
+      "learning_rate": 6.784182096130104e-07,
+      "loss": 0.0285,
+      "step": 6378
+    },
+    {
+      "epoch": 2.547923322683706,
+      "grad_norm": 0.9343640996748256,
+      "learning_rate": 6.76082710639247e-07,
+      "loss": 0.0292,
+      "step": 6380
+    },
+    {
+      "epoch": 2.5487220447284344,
+      "grad_norm": 0.9782586764772694,
+      "learning_rate": 6.737509471220527e-07,
+      "loss": 0.0327,
+      "step": 6382
+    },
+    {
+      "epoch": 2.549520766773163,
+      "grad_norm": 0.9736964982154661,
+      "learning_rate": 6.714229210758516e-07,
+      "loss": 0.0292,
+      "step": 6384
+    },
+    {
+      "epoch": 2.5503194888178915,
+      "grad_norm": 0.9366260036987168,
+      "learning_rate": 6.690986345118389e-07,
+      "loss": 0.0308,
+      "step": 6386
+    },
+    {
+      "epoch": 2.5511182108626196,
+      "grad_norm": 1.052050189264594,
+      "learning_rate": 6.667780894379799e-07,
+      "loss": 0.0308,
+      "step": 6388
+    },
+    {
+      "epoch": 2.551916932907348,
+      "grad_norm": 0.9827812854065209,
+      "learning_rate": 6.644612878590034e-07,
+      "loss": 0.0316,
+      "step": 6390
+    },
+    {
+      "epoch": 2.5527156549520766,
+      "grad_norm": 0.8386586055394432,
+      "learning_rate": 6.621482317764105e-07,
+      "loss": 0.0305,
+      "step": 6392
+    },
+    {
+      "epoch": 2.553514376996805,
+      "grad_norm": 0.9894645955986272,
+      "learning_rate": 6.598389231884628e-07,
+      "loss": 0.0297,
+      "step": 6394
+    },
+    {
+      "epoch": 2.5543130990415337,
+      "grad_norm": 0.8839682059943562,
+      "learning_rate": 6.575333640901855e-07,
+      "loss": 0.0285,
+      "step": 6396
+    },
+    {
+      "epoch": 2.5551118210862622,
+      "grad_norm": 0.8887779152215409,
+      "learning_rate": 6.552315564733625e-07,
+      "loss": 0.0285,
+      "step": 6398
+    },
+    {
+      "epoch": 2.5559105431309903,
+      "grad_norm": 1.0490578369091348,
+      "learning_rate": 6.529335023265387e-07,
+      "loss": 0.0289,
+      "step": 6400
+    },
+    {
+      "epoch": 2.556709265175719,
+      "grad_norm": 1.0894646563356374,
+      "learning_rate": 6.506392036350168e-07,
+      "loss": 0.0322,
+      "step": 6402
+    },
+    {
+      "epoch": 2.5575079872204474,
+      "grad_norm": 0.9316306126563935,
+      "learning_rate": 6.483486623808555e-07,
+      "loss": 0.0282,
+      "step": 6404
+    },
+    {
+      "epoch": 2.5583067092651754,
+      "grad_norm": 1.0138875975414274,
+      "learning_rate": 6.460618805428637e-07,
+      "loss": 0.029,
+      "step": 6406
+    },
+    {
+      "epoch": 2.559105431309904,
+      "grad_norm": 0.9409011790958446,
+      "learning_rate": 6.437788600966066e-07,
+      "loss": 0.0275,
+      "step": 6408
+    },
+    {
+      "epoch": 2.5599041533546325,
+      "grad_norm": 1.0950975759175423,
+      "learning_rate": 6.414996030143982e-07,
+      "loss": 0.026,
+      "step": 6410
+    },
+    {
+      "epoch": 2.560702875399361,
+      "grad_norm": 0.8621907379188767,
+      "learning_rate": 6.392241112653031e-07,
+      "loss": 0.0263,
+      "step": 6412
+    },
+    {
+      "epoch": 2.5615015974440896,
+      "grad_norm": 1.0987324235841187,
+      "learning_rate": 6.369523868151278e-07,
+      "loss": 0.03,
+      "step": 6414
+    },
+    {
+      "epoch": 2.562300319488818,
+      "grad_norm": 1.0180249667885073,
+      "learning_rate": 6.346844316264312e-07,
+      "loss": 0.0337,
+      "step": 6416
+    },
+    {
+      "epoch": 2.563099041533546,
+      "grad_norm": 0.918284409010965,
+      "learning_rate": 6.324202476585112e-07,
+      "loss": 0.0293,
+      "step": 6418
+    },
+    {
+      "epoch": 2.5638977635782747,
+      "grad_norm": 0.8900922129741325,
+      "learning_rate": 6.301598368674106e-07,
+      "loss": 0.0262,
+      "step": 6420
+    },
+    {
+      "epoch": 2.5646964856230032,
+      "grad_norm": 0.8698502363272664,
+      "learning_rate": 6.279032012059089e-07,
+      "loss": 0.0279,
+      "step": 6422
+    },
+    {
+      "epoch": 2.5654952076677318,
+      "grad_norm": 0.7647567338632194,
+      "learning_rate": 6.256503426235277e-07,
+      "loss": 0.0259,
+      "step": 6424
+    },
+    {
+      "epoch": 2.56629392971246,
+      "grad_norm": 0.9236140855263162,
+      "learning_rate": 6.234012630665237e-07,
+      "loss": 0.0309,
+      "step": 6426
+    },
+    {
+      "epoch": 2.5670926517571884,
+      "grad_norm": 0.9418851113359151,
+      "learning_rate": 6.211559644778908e-07,
+      "loss": 0.0287,
+      "step": 6428
+    },
+    {
+      "epoch": 2.567891373801917,
+      "grad_norm": 0.9410493586581912,
+      "learning_rate": 6.189144487973531e-07,
+      "loss": 0.0245,
+      "step": 6430
+    },
+    {
+      "epoch": 2.5686900958466454,
+      "grad_norm": 0.8715415469545029,
+      "learning_rate": 6.166767179613691e-07,
+      "loss": 0.0272,
+      "step": 6432
+    },
+    {
+      "epoch": 2.569488817891374,
+      "grad_norm": 1.0011250141948451,
+      "learning_rate": 6.144427739031284e-07,
+      "loss": 0.0336,
+      "step": 6434
+    },
+    {
+      "epoch": 2.5702875399361025,
+      "grad_norm": 0.8930124959505643,
+      "learning_rate": 6.122126185525462e-07,
+      "loss": 0.0287,
+      "step": 6436
+    },
+    {
+      "epoch": 2.5710862619808306,
+      "grad_norm": 0.8345328277096289,
+      "learning_rate": 6.099862538362678e-07,
+      "loss": 0.0274,
+      "step": 6438
+    },
+    {
+      "epoch": 2.571884984025559,
+      "grad_norm": 0.8429043604050549,
+      "learning_rate": 6.077636816776611e-07,
+      "loss": 0.0259,
+      "step": 6440
+    },
+    {
+      "epoch": 2.5726837060702876,
+      "grad_norm": 0.9282997534416909,
+      "learning_rate": 6.055449039968197e-07,
+      "loss": 0.0308,
+      "step": 6442
+    },
+    {
+      "epoch": 2.5734824281150157,
+      "grad_norm": 1.078810276103343,
+      "learning_rate": 6.033299227105588e-07,
+      "loss": 0.03,
+      "step": 6444
+    },
+    {
+      "epoch": 2.5742811501597442,
+      "grad_norm": 0.9006283103160382,
+      "learning_rate": 6.011187397324114e-07,
+      "loss": 0.0277,
+      "step": 6446
+    },
+    {
+      "epoch": 2.5750798722044728,
+      "grad_norm": 1.0725617589003342,
+      "learning_rate": 5.989113569726312e-07,
+      "loss": 0.0312,
+      "step": 6448
+    },
+    {
+      "epoch": 2.5758785942492013,
+      "grad_norm": 1.0044942875803178,
+      "learning_rate": 5.967077763381895e-07,
+      "loss": 0.0269,
+      "step": 6450
+    },
+    {
+      "epoch": 2.57667731629393,
+      "grad_norm": 1.0483372526983028,
+      "learning_rate": 5.945079997327713e-07,
+      "loss": 0.0303,
+      "step": 6452
+    },
+    {
+      "epoch": 2.5774760383386583,
+      "grad_norm": 0.885373336335105,
+      "learning_rate": 5.923120290567779e-07,
+      "loss": 0.0292,
+      "step": 6454
+    },
+    {
+      "epoch": 2.5782747603833864,
+      "grad_norm": 0.9919843795418669,
+      "learning_rate": 5.901198662073188e-07,
+      "loss": 0.0372,
+      "step": 6456
+    },
+    {
+      "epoch": 2.579073482428115,
+      "grad_norm": 1.300703442861457,
+      "learning_rate": 5.87931513078216e-07,
+      "loss": 0.0246,
+      "step": 6458
+    },
+    {
+      "epoch": 2.5798722044728435,
+      "grad_norm": 1.0888147527555645,
+      "learning_rate": 5.8574697156e-07,
+      "loss": 0.0368,
+      "step": 6460
+    },
+    {
+      "epoch": 2.580670926517572,
+      "grad_norm": 1.183957007842984,
+      "learning_rate": 5.835662435399098e-07,
+      "loss": 0.0345,
+      "step": 6462
+    },
+    {
+      "epoch": 2.5814696485623,
+      "grad_norm": 0.9387448812321908,
+      "learning_rate": 5.813893309018881e-07,
+      "loss": 0.026,
+      "step": 6464
+    },
+    {
+      "epoch": 2.5822683706070286,
+      "grad_norm": 0.9875605986563711,
+      "learning_rate": 5.792162355265812e-07,
+      "loss": 0.0323,
+      "step": 6466
+    },
+    {
+      "epoch": 2.583067092651757,
+      "grad_norm": 0.8843843609089413,
+      "learning_rate": 5.770469592913408e-07,
+      "loss": 0.029,
+      "step": 6468
+    },
+    {
+      "epoch": 2.5838658146964857,
+      "grad_norm": 0.9754923443275886,
+      "learning_rate": 5.748815040702138e-07,
+      "loss": 0.0307,
+      "step": 6470
+    },
+    {
+      "epoch": 2.584664536741214,
+      "grad_norm": 1.005644213615307,
+      "learning_rate": 5.727198717339511e-07,
+      "loss": 0.0324,
+      "step": 6472
+    },
+    {
+      "epoch": 2.5854632587859427,
+      "grad_norm": 0.9932008626732668,
+      "learning_rate": 5.705620641499981e-07,
+      "loss": 0.0306,
+      "step": 6474
+    },
+    {
+      "epoch": 2.586261980830671,
+      "grad_norm": 0.9873244519751645,
+      "learning_rate": 5.684080831824978e-07,
+      "loss": 0.0331,
+      "step": 6476
+    },
+    {
+      "epoch": 2.5870607028753994,
+      "grad_norm": 0.9241878530410826,
+      "learning_rate": 5.662579306922872e-07,
+      "loss": 0.0322,
+      "step": 6478
+    },
+    {
+      "epoch": 2.587859424920128,
+      "grad_norm": 1.0995821963687364,
+      "learning_rate": 5.641116085368931e-07,
+      "loss": 0.0308,
+      "step": 6480
+    },
+    {
+      "epoch": 2.588658146964856,
+      "grad_norm": 0.8378812796834532,
+      "learning_rate": 5.619691185705356e-07,
+      "loss": 0.0285,
+      "step": 6482
+    },
+    {
+      "epoch": 2.5894568690095845,
+      "grad_norm": 0.8605425287906566,
+      "learning_rate": 5.598304626441264e-07,
+      "loss": 0.0251,
+      "step": 6484
+    },
+    {
+      "epoch": 2.590255591054313,
+      "grad_norm": 1.0050366580290326,
+      "learning_rate": 5.576956426052605e-07,
+      "loss": 0.0287,
+      "step": 6486
+    },
+    {
+      "epoch": 2.5910543130990416,
+      "grad_norm": 1.065340889189702,
+      "learning_rate": 5.555646602982207e-07,
+      "loss": 0.0336,
+      "step": 6488
+    },
+    {
+      "epoch": 2.59185303514377,
+      "grad_norm": 0.9304491062820247,
+      "learning_rate": 5.53437517563975e-07,
+      "loss": 0.0282,
+      "step": 6490
+    },
+    {
+      "epoch": 2.5926517571884986,
+      "grad_norm": 0.9396575473059782,
+      "learning_rate": 5.513142162401746e-07,
+      "loss": 0.0258,
+      "step": 6492
+    },
+    {
+      "epoch": 2.5934504792332267,
+      "grad_norm": 0.9393766036210395,
+      "learning_rate": 5.491947581611517e-07,
+      "loss": 0.0266,
+      "step": 6494
+    },
+    {
+      "epoch": 2.594249201277955,
+      "grad_norm": 0.9435384740988189,
+      "learning_rate": 5.470791451579172e-07,
+      "loss": 0.0317,
+      "step": 6496
+    },
+    {
+      "epoch": 2.5950479233226837,
+      "grad_norm": 0.9321052021346842,
+      "learning_rate": 5.449673790581611e-07,
+      "loss": 0.0299,
+      "step": 6498
+    },
+    {
+      "epoch": 2.5958466453674123,
+      "grad_norm": 0.9112996056101765,
+      "learning_rate": 5.428594616862504e-07,
+      "loss": 0.0269,
+      "step": 6500
+    },
+    {
+      "epoch": 2.5958466453674123,
+      "eval_loss": 0.18101496994495392,
+      "eval_runtime": 417.6423,
+      "eval_samples_per_second": 42.637,
+      "eval_steps_per_second": 5.33,
+      "step": 6500
+    },
+    {
+      "epoch": 2.5966453674121404,
+      "grad_norm": 0.9226194319100446,
+      "learning_rate": 5.407553948632277e-07,
+      "loss": 0.0304,
+      "step": 6502
+    },
+    {
+      "epoch": 2.597444089456869,
+      "grad_norm": 0.9405971080377886,
+      "learning_rate": 5.386551804068063e-07,
+      "loss": 0.0314,
+      "step": 6504
+    },
+    {
+      "epoch": 2.5982428115015974,
+      "grad_norm": 0.9385959964411323,
+      "learning_rate": 5.365588201313737e-07,
+      "loss": 0.0297,
+      "step": 6506
+    },
+    {
+      "epoch": 2.599041533546326,
+      "grad_norm": 0.84541282801553,
+      "learning_rate": 5.344663158479901e-07,
+      "loss": 0.0268,
+      "step": 6508
+    },
+    {
+      "epoch": 2.5998402555910545,
+      "grad_norm": 0.9713712244776086,
+      "learning_rate": 5.323776693643784e-07,
+      "loss": 0.0306,
+      "step": 6510
+    },
+    {
+      "epoch": 2.600638977635783,
+      "grad_norm": 0.8383820244075577,
+      "learning_rate": 5.302928824849335e-07,
+      "loss": 0.0265,
+      "step": 6512
+    },
+    {
+      "epoch": 2.601437699680511,
+      "grad_norm": 1.124289996932428,
+      "learning_rate": 5.282119570107147e-07,
+      "loss": 0.0322,
+      "step": 6514
+    },
+    {
+      "epoch": 2.6022364217252396,
+      "grad_norm": 0.9423889108993881,
+      "learning_rate": 5.261348947394451e-07,
+      "loss": 0.0281,
+      "step": 6516
+    },
+    {
+      "epoch": 2.603035143769968,
+      "grad_norm": 0.9129093506596754,
+      "learning_rate": 5.240616974655116e-07,
+      "loss": 0.0279,
+      "step": 6518
+    },
+    {
+      "epoch": 2.6038338658146962,
+      "grad_norm": 0.7842673666707356,
+      "learning_rate": 5.219923669799587e-07,
+      "loss": 0.0241,
+      "step": 6520
+    },
+    {
+      "epoch": 2.6046325878594248,
+      "grad_norm": 1.1436169228498474,
+      "learning_rate": 5.199269050704935e-07,
+      "loss": 0.0325,
+      "step": 6522
+    },
+    {
+      "epoch": 2.6054313099041533,
+      "grad_norm": 0.8690653883300302,
+      "learning_rate": 5.178653135214811e-07,
+      "loss": 0.0298,
+      "step": 6524
+    },
+    {
+      "epoch": 2.606230031948882,
+      "grad_norm": 0.8819467380201907,
+      "learning_rate": 5.158075941139429e-07,
+      "loss": 0.0291,
+      "step": 6526
+    },
+    {
+      "epoch": 2.6070287539936103,
+      "grad_norm": 0.9898669074800155,
+      "learning_rate": 5.137537486255517e-07,
+      "loss": 0.0306,
+      "step": 6528
+    },
+    {
+      "epoch": 2.607827476038339,
+      "grad_norm": 1.129969353766448,
+      "learning_rate": 5.117037788306367e-07,
+      "loss": 0.029,
+      "step": 6530
+    },
+    {
+      "epoch": 2.608626198083067,
+      "grad_norm": 1.18791037882053,
+      "learning_rate": 5.096576865001802e-07,
+      "loss": 0.0308,
+      "step": 6532
+    },
+    {
+      "epoch": 2.6094249201277955,
+      "grad_norm": 0.9477100539787796,
+      "learning_rate": 5.07615473401813e-07,
+      "loss": 0.0277,
+      "step": 6534
+    },
+    {
+      "epoch": 2.610223642172524,
+      "grad_norm": 0.9068341807452427,
+      "learning_rate": 5.055771412998122e-07,
+      "loss": 0.0289,
+      "step": 6536
+    },
+    {
+      "epoch": 2.6110223642172525,
+      "grad_norm": 0.895582110683765,
+      "learning_rate": 5.035426919551062e-07,
+      "loss": 0.0268,
+      "step": 6538
+    },
+    {
+      "epoch": 2.6118210862619806,
+      "grad_norm": 0.9258207040673696,
+      "learning_rate": 5.015121271252659e-07,
+      "loss": 0.0287,
+      "step": 6540
+    },
+    {
+      "epoch": 2.612619808306709,
+      "grad_norm": 0.8949264605397048,
+      "learning_rate": 4.994854485645106e-07,
+      "loss": 0.0291,
+      "step": 6542
+    },
+    {
+      "epoch": 2.6134185303514377,
+      "grad_norm": 1.0785382270595747,
+      "learning_rate": 4.974626580236957e-07,
+      "loss": 0.0302,
+      "step": 6544
+    },
+    {
+      "epoch": 2.614217252396166,
+      "grad_norm": 1.0483244361602544,
+      "learning_rate": 4.954437572503235e-07,
+      "loss": 0.031,
+      "step": 6546
+    },
+    {
+      "epoch": 2.6150159744408947,
+      "grad_norm": 1.028765681064945,
+      "learning_rate": 4.934287479885336e-07,
+      "loss": 0.0303,
+      "step": 6548
+    },
+    {
+      "epoch": 2.6158146964856233,
+      "grad_norm": 0.9557621074130701,
+      "learning_rate": 4.914176319791037e-07,
+      "loss": 0.025,
+      "step": 6550
+    },
+    {
+      "epoch": 2.6166134185303513,
+      "grad_norm": 0.8620869517168456,
+      "learning_rate": 4.894104109594466e-07,
+      "loss": 0.0275,
+      "step": 6552
+    },
+    {
+      "epoch": 2.61741214057508,
+      "grad_norm": 0.8437067964701872,
+      "learning_rate": 4.874070866636149e-07,
+      "loss": 0.0274,
+      "step": 6554
+    },
+    {
+      "epoch": 2.6182108626198084,
+      "grad_norm": 0.8006419089172357,
+      "learning_rate": 4.854076608222901e-07,
+      "loss": 0.0303,
+      "step": 6556
+    },
+    {
+      "epoch": 2.6190095846645365,
+      "grad_norm": 0.8823397281540788,
+      "learning_rate": 4.834121351627885e-07,
+      "loss": 0.0254,
+      "step": 6558
+    },
+    {
+      "epoch": 2.619808306709265,
+      "grad_norm": 0.8499615378338659,
+      "learning_rate": 4.814205114090543e-07,
+      "loss": 0.0239,
+      "step": 6560
+    },
+    {
+      "epoch": 2.6206070287539935,
+      "grad_norm": 1.03836923861178,
+      "learning_rate": 4.794327912816637e-07,
+      "loss": 0.0337,
+      "step": 6562
+    },
+    {
+      "epoch": 2.621405750798722,
+      "grad_norm": 1.064067294335829,
+      "learning_rate": 4.774489764978185e-07,
+      "loss": 0.0267,
+      "step": 6564
+    },
+    {
+      "epoch": 2.6222044728434506,
+      "grad_norm": 0.9357645207898694,
+      "learning_rate": 4.754690687713498e-07,
+      "loss": 0.0281,
+      "step": 6566
+    },
+    {
+      "epoch": 2.623003194888179,
+      "grad_norm": 0.9735104123969658,
+      "learning_rate": 4.734930698127077e-07,
+      "loss": 0.0294,
+      "step": 6568
+    },
+    {
+      "epoch": 2.623801916932907,
+      "grad_norm": 1.0247460222068252,
+      "learning_rate": 4.715209813289706e-07,
+      "loss": 0.0307,
+      "step": 6570
+    },
+    {
+      "epoch": 2.6246006389776357,
+      "grad_norm": 0.9357520689720981,
+      "learning_rate": 4.695528050238368e-07,
+      "loss": 0.031,
+      "step": 6572
+    },
+    {
+      "epoch": 2.6253993610223643,
+      "grad_norm": 0.9787665294252466,
+      "learning_rate": 4.675885425976251e-07,
+      "loss": 0.031,
+      "step": 6574
+    },
+    {
+      "epoch": 2.626198083067093,
+      "grad_norm": 1.0180864802210927,
+      "learning_rate": 4.6562819574727304e-07,
+      "loss": 0.0281,
+      "step": 6576
+    },
+    {
+      "epoch": 2.626996805111821,
+      "grad_norm": 0.8926200488656815,
+      "learning_rate": 4.6367176616633426e-07,
+      "loss": 0.0295,
+      "step": 6578
+    },
+    {
+      "epoch": 2.6277955271565494,
+      "grad_norm": 1.0500600630048418,
+      "learning_rate": 4.6171925554498066e-07,
+      "loss": 0.0333,
+      "step": 6580
+    },
+    {
+      "epoch": 2.628594249201278,
+      "grad_norm": 1.182205096455108,
+      "learning_rate": 4.597706655699974e-07,
+      "loss": 0.0343,
+      "step": 6582
+    },
+    {
+      "epoch": 2.6293929712460065,
+      "grad_norm": 0.8380383559912178,
+      "learning_rate": 4.578259979247801e-07,
+      "loss": 0.0249,
+      "step": 6584
+    },
+    {
+      "epoch": 2.630191693290735,
+      "grad_norm": 0.9217249683454195,
+      "learning_rate": 4.558852542893405e-07,
+      "loss": 0.0274,
+      "step": 6586
+    },
+    {
+      "epoch": 2.6309904153354635,
+      "grad_norm": 0.9654985885991184,
+      "learning_rate": 4.539484363402963e-07,
+      "loss": 0.0313,
+      "step": 6588
+    },
+    {
+      "epoch": 2.6317891373801916,
+      "grad_norm": 0.9416775486966591,
+      "learning_rate": 4.520155457508768e-07,
+      "loss": 0.0278,
+      "step": 6590
+    },
+    {
+      "epoch": 2.63258785942492,
+      "grad_norm": 1.0149290675646854,
+      "learning_rate": 4.500865841909169e-07,
+      "loss": 0.0283,
+      "step": 6592
+    },
+    {
+      "epoch": 2.6333865814696487,
+      "grad_norm": 1.0066112830843752,
+      "learning_rate": 4.4816155332685687e-07,
+      "loss": 0.0321,
+      "step": 6594
+    },
+    {
+      "epoch": 2.6341853035143767,
+      "grad_norm": 0.9399055283517556,
+      "learning_rate": 4.462404548217414e-07,
+      "loss": 0.0274,
+      "step": 6596
+    },
+    {
+      "epoch": 2.6349840255591053,
+      "grad_norm": 0.9221492425778504,
+      "learning_rate": 4.4432329033521903e-07,
+      "loss": 0.0295,
+      "step": 6598
+    },
+    {
+      "epoch": 2.635782747603834,
+      "grad_norm": 0.890695263755349,
+      "learning_rate": 4.4241006152353885e-07,
+      "loss": 0.0253,
+      "step": 6600
+    },
+    {
+      "epoch": 2.6365814696485623,
+      "grad_norm": 0.9178158597505024,
+      "learning_rate": 4.405007700395497e-07,
+      "loss": 0.0281,
+      "step": 6602
+    },
+    {
+      "epoch": 2.637380191693291,
+      "grad_norm": 1.0639158204341026,
+      "learning_rate": 4.385954175326995e-07,
+      "loss": 0.0264,
+      "step": 6604
+    },
+    {
+      "epoch": 2.6381789137380194,
+      "grad_norm": 1.012323859666103,
+      "learning_rate": 4.366940056490343e-07,
+      "loss": 0.0308,
+      "step": 6606
+    },
+    {
+      "epoch": 2.6389776357827475,
+      "grad_norm": 0.9727641969282357,
+      "learning_rate": 4.3479653603119287e-07,
+      "loss": 0.0271,
+      "step": 6608
+    },
+    {
+      "epoch": 2.639776357827476,
+      "grad_norm": 0.8772595568384365,
+      "learning_rate": 4.329030103184095e-07,
+      "loss": 0.0259,
+      "step": 6610
+    },
+    {
+      "epoch": 2.6405750798722045,
+      "grad_norm": 0.8227589976875681,
+      "learning_rate": 4.3101343014651356e-07,
+      "loss": 0.0277,
+      "step": 6612
+    },
+    {
+      "epoch": 2.641373801916933,
+      "grad_norm": 0.9510948759982237,
+      "learning_rate": 4.2912779714792296e-07,
+      "loss": 0.0265,
+      "step": 6614
+    },
+    {
+      "epoch": 2.642172523961661,
+      "grad_norm": 1.0097839001611448,
+      "learning_rate": 4.2724611295164755e-07,
+      "loss": 0.026,
+      "step": 6616
+    },
+    {
+      "epoch": 2.6429712460063897,
+      "grad_norm": 0.9098862544268895,
+      "learning_rate": 4.2536837918328353e-07,
+      "loss": 0.0306,
+      "step": 6618
+    },
+    {
+      "epoch": 2.643769968051118,
+      "grad_norm": 0.8802136378624498,
+      "learning_rate": 4.2349459746501674e-07,
+      "loss": 0.0249,
+      "step": 6620
+    },
+    {
+      "epoch": 2.6445686900958467,
+      "grad_norm": 0.9041537710345738,
+      "learning_rate": 4.2162476941561723e-07,
+      "loss": 0.0291,
+      "step": 6622
+    },
+    {
+      "epoch": 2.6453674121405752,
+      "grad_norm": 1.007668717045686,
+      "learning_rate": 4.197588966504401e-07,
+      "loss": 0.0269,
+      "step": 6624
+    },
+    {
+      "epoch": 2.6461661341853038,
+      "grad_norm": 0.9464297822234001,
+      "learning_rate": 4.178969807814237e-07,
+      "loss": 0.0292,
+      "step": 6626
+    },
+    {
+      "epoch": 2.646964856230032,
+      "grad_norm": 0.9545384276330572,
+      "learning_rate": 4.1603902341708804e-07,
+      "loss": 0.0315,
+      "step": 6628
+    },
+    {
+      "epoch": 2.6477635782747604,
+      "grad_norm": 0.9924287706034195,
+      "learning_rate": 4.1418502616253185e-07,
+      "loss": 0.0293,
+      "step": 6630
+    },
+    {
+      "epoch": 2.648562300319489,
+      "grad_norm": 0.8072318680535587,
+      "learning_rate": 4.123349906194357e-07,
+      "loss": 0.0281,
+      "step": 6632
+    },
+    {
+      "epoch": 2.649361022364217,
+      "grad_norm": 0.8755579059821392,
+      "learning_rate": 4.1048891838605386e-07,
+      "loss": 0.0243,
+      "step": 6634
+    },
+    {
+      "epoch": 2.6501597444089455,
+      "grad_norm": 1.1069033803045907,
+      "learning_rate": 4.0864681105721895e-07,
+      "loss": 0.031,
+      "step": 6636
+    },
+    {
+      "epoch": 2.650958466453674,
+      "grad_norm": 0.9327831704930124,
+      "learning_rate": 4.068086702243379e-07,
+      "loss": 0.0283,
+      "step": 6638
+    },
+    {
+      "epoch": 2.6517571884984026,
+      "grad_norm": 1.0816491366183623,
+      "learning_rate": 4.0497449747539217e-07,
+      "loss": 0.0324,
+      "step": 6640
+    },
+    {
+      "epoch": 2.652555910543131,
+      "grad_norm": 0.9653480667623733,
+      "learning_rate": 4.031442943949321e-07,
+      "loss": 0.0271,
+      "step": 6642
+    },
+    {
+      "epoch": 2.6533546325878596,
+      "grad_norm": 1.025462257232587,
+      "learning_rate": 4.013180625640811e-07,
+      "loss": 0.0268,
+      "step": 6644
+    },
+    {
+      "epoch": 2.6541533546325877,
+      "grad_norm": 1.0073308408827422,
+      "learning_rate": 3.994958035605323e-07,
+      "loss": 0.0335,
+      "step": 6646
+    },
+    {
+      "epoch": 2.6549520766773163,
+      "grad_norm": 0.7759996809586879,
+      "learning_rate": 3.9767751895854467e-07,
+      "loss": 0.0278,
+      "step": 6648
+    },
+    {
+      "epoch": 2.655750798722045,
+      "grad_norm": 0.9648450344658905,
+      "learning_rate": 3.958632103289439e-07,
+      "loss": 0.0323,
+      "step": 6650
+    },
+    {
+      "epoch": 2.6565495207667733,
+      "grad_norm": 0.8621409106150557,
+      "learning_rate": 3.940528792391224e-07,
+      "loss": 0.0223,
+      "step": 6652
+    },
+    {
+      "epoch": 2.6573482428115014,
+      "grad_norm": 0.8191488899403426,
+      "learning_rate": 3.9224652725303514e-07,
+      "loss": 0.0286,
+      "step": 6654
+    },
+    {
+      "epoch": 2.65814696485623,
+      "grad_norm": 1.0056039386415438,
+      "learning_rate": 3.904441559312006e-07,
+      "loss": 0.0311,
+      "step": 6656
+    },
+    {
+      "epoch": 2.6589456869009584,
+      "grad_norm": 1.0006415392629169,
+      "learning_rate": 3.886457668306959e-07,
+      "loss": 0.031,
+      "step": 6658
+    },
+    {
+      "epoch": 2.659744408945687,
+      "grad_norm": 1.01436188125872,
+      "learning_rate": 3.8685136150516056e-07,
+      "loss": 0.032,
+      "step": 6660
+    },
+    {
+      "epoch": 2.6605431309904155,
+      "grad_norm": 0.9807096116942686,
+      "learning_rate": 3.8506094150479125e-07,
+      "loss": 0.0266,
+      "step": 6662
+    },
+    {
+      "epoch": 2.661341853035144,
+      "grad_norm": 1.0204227532844772,
+      "learning_rate": 3.8327450837634284e-07,
+      "loss": 0.0283,
+      "step": 6664
+    },
+    {
+      "epoch": 2.662140575079872,
+      "grad_norm": 0.8193622993500249,
+      "learning_rate": 3.8149206366312365e-07,
+      "loss": 0.0269,
+      "step": 6666
+    },
+    {
+      "epoch": 2.6629392971246006,
+      "grad_norm": 0.9734485918188198,
+      "learning_rate": 3.7971360890499686e-07,
+      "loss": 0.0266,
+      "step": 6668
+    },
+    {
+      "epoch": 2.663738019169329,
+      "grad_norm": 0.8642796045293297,
+      "learning_rate": 3.7793914563838187e-07,
+      "loss": 0.0251,
+      "step": 6670
+    },
+    {
+      "epoch": 2.6645367412140573,
+      "grad_norm": 1.033909020021237,
+      "learning_rate": 3.7616867539624733e-07,
+      "loss": 0.0352,
+      "step": 6672
+    },
+    {
+      "epoch": 2.665335463258786,
+      "grad_norm": 0.9998346566813847,
+      "learning_rate": 3.7440219970811155e-07,
+      "loss": 0.0321,
+      "step": 6674
+    },
+    {
+      "epoch": 2.6661341853035143,
+      "grad_norm": 1.1423057215989467,
+      "learning_rate": 3.7263972010004256e-07,
+      "loss": 0.0324,
+      "step": 6676
+    },
+    {
+      "epoch": 2.666932907348243,
+      "grad_norm": 0.8695749763174104,
+      "learning_rate": 3.708812380946569e-07,
+      "loss": 0.0256,
+      "step": 6678
+    },
+    {
+      "epoch": 2.6677316293929714,
+      "grad_norm": 0.9934106793723311,
+      "learning_rate": 3.691267552111183e-07,
+      "loss": 0.0302,
+      "step": 6680
+    },
+    {
+      "epoch": 2.6685303514377,
+      "grad_norm": 1.0381623580396713,
+      "learning_rate": 3.67376272965132e-07,
+      "loss": 0.0281,
+      "step": 6682
+    },
+    {
+      "epoch": 2.669329073482428,
+      "grad_norm": 0.8905369484200444,
+      "learning_rate": 3.6562979286895115e-07,
+      "loss": 0.0255,
+      "step": 6684
+    },
+    {
+      "epoch": 2.6701277955271565,
+      "grad_norm": 0.8948659099592012,
+      "learning_rate": 3.6388731643136944e-07,
+      "loss": 0.0252,
+      "step": 6686
+    },
+    {
+      "epoch": 2.670926517571885,
+      "grad_norm": 0.9124920896558937,
+      "learning_rate": 3.621488451577221e-07,
+      "loss": 0.0293,
+      "step": 6688
+    },
+    {
+      "epoch": 2.6717252396166136,
+      "grad_norm": 0.7777444783097159,
+      "learning_rate": 3.60414380549885e-07,
+      "loss": 0.0251,
+      "step": 6690
+    },
+    {
+      "epoch": 2.6725239616613417,
+      "grad_norm": 0.9000009491244071,
+      "learning_rate": 3.586839241062695e-07,
+      "loss": 0.0254,
+      "step": 6692
+    },
+    {
+      "epoch": 2.67332268370607,
+      "grad_norm": 0.9763457977314538,
+      "learning_rate": 3.5695747732182873e-07,
+      "loss": 0.0307,
+      "step": 6694
+    },
+    {
+      "epoch": 2.6741214057507987,
+      "grad_norm": 1.1307665134525087,
+      "learning_rate": 3.552350416880507e-07,
+      "loss": 0.0297,
+      "step": 6696
+    },
+    {
+      "epoch": 2.6749201277955272,
+      "grad_norm": 0.864801567907232,
+      "learning_rate": 3.535166186929556e-07,
+      "loss": 0.0296,
+      "step": 6698
+    },
+    {
+      "epoch": 2.6757188498402558,
+      "grad_norm": 0.8870307933812397,
+      "learning_rate": 3.518022098210988e-07,
+      "loss": 0.0292,
+      "step": 6700
+    },
+    {
+      "epoch": 2.6765175718849843,
+      "grad_norm": 1.0409321048802573,
+      "learning_rate": 3.500918165535683e-07,
+      "loss": 0.0345,
+      "step": 6702
+    },
+    {
+      "epoch": 2.6773162939297124,
+      "grad_norm": 0.9716769842609737,
+      "learning_rate": 3.483854403679832e-07,
+      "loss": 0.0296,
+      "step": 6704
+    },
+    {
+      "epoch": 2.678115015974441,
+      "grad_norm": 1.257329352898907,
+      "learning_rate": 3.4668308273848985e-07,
+      "loss": 0.0267,
+      "step": 6706
+    },
+    {
+      "epoch": 2.6789137380191694,
+      "grad_norm": 1.0441757060258385,
+      "learning_rate": 3.4498474513576574e-07,
+      "loss": 0.0317,
+      "step": 6708
+    },
+    {
+      "epoch": 2.6797124600638975,
+      "grad_norm": 1.1129495022310048,
+      "learning_rate": 3.432904290270139e-07,
+      "loss": 0.0286,
+      "step": 6710
+    },
+    {
+      "epoch": 2.680511182108626,
+      "grad_norm": 0.9584352042603548,
+      "learning_rate": 3.416001358759635e-07,
+      "loss": 0.028,
+      "step": 6712
+    },
+    {
+      "epoch": 2.6813099041533546,
+      "grad_norm": 1.1251450657365984,
+      "learning_rate": 3.3991386714286924e-07,
+      "loss": 0.0278,
+      "step": 6714
+    },
+    {
+      "epoch": 2.682108626198083,
+      "grad_norm": 0.9969919886738752,
+      "learning_rate": 3.382316242845074e-07,
+      "loss": 0.0328,
+      "step": 6716
+    },
+    {
+      "epoch": 2.6829073482428116,
+      "grad_norm": 1.1126055490955078,
+      "learning_rate": 3.365534087541772e-07,
+      "loss": 0.0277,
+      "step": 6718
+    },
+    {
+      "epoch": 2.68370607028754,
+      "grad_norm": 1.0455467801276128,
+      "learning_rate": 3.3487922200169944e-07,
+      "loss": 0.0314,
+      "step": 6720
+    },
+    {
+      "epoch": 2.6845047923322682,
+      "grad_norm": 0.9840651799523558,
+      "learning_rate": 3.332090654734116e-07,
+      "loss": 0.0325,
+      "step": 6722
+    },
+    {
+      "epoch": 2.6853035143769968,
+      "grad_norm": 0.9993440868747537,
+      "learning_rate": 3.315429406121723e-07,
+      "loss": 0.0285,
+      "step": 6724
+    },
+    {
+      "epoch": 2.6861022364217253,
+      "grad_norm": 1.041727821587971,
+      "learning_rate": 3.2988084885735684e-07,
+      "loss": 0.0351,
+      "step": 6726
+    },
+    {
+      "epoch": 2.686900958466454,
+      "grad_norm": 0.979266522959519,
+      "learning_rate": 3.2822279164485494e-07,
+      "loss": 0.0267,
+      "step": 6728
+    },
+    {
+      "epoch": 2.687699680511182,
+      "grad_norm": 0.8894076061044695,
+      "learning_rate": 3.2656877040707247e-07,
+      "loss": 0.0306,
+      "step": 6730
+    },
+    {
+      "epoch": 2.6884984025559104,
+      "grad_norm": 1.1099161112289966,
+      "learning_rate": 3.2491878657292643e-07,
+      "loss": 0.029,
+      "step": 6732
+    },
+    {
+      "epoch": 2.689297124600639,
+      "grad_norm": 0.7905372930822995,
+      "learning_rate": 3.2327284156784765e-07,
+      "loss": 0.0248,
+      "step": 6734
+    },
+    {
+      "epoch": 2.6900958466453675,
+      "grad_norm": 0.8972437845079343,
+      "learning_rate": 3.2163093681377765e-07,
+      "loss": 0.0322,
+      "step": 6736
+    },
+    {
+      "epoch": 2.690894568690096,
+      "grad_norm": 0.7811198103953365,
+      "learning_rate": 3.1999307372916675e-07,
+      "loss": 0.0211,
+      "step": 6738
+    },
+    {
+      "epoch": 2.6916932907348246,
+      "grad_norm": 0.8481986944549397,
+      "learning_rate": 3.183592537289748e-07,
+      "loss": 0.0258,
+      "step": 6740
+    },
+    {
+      "epoch": 2.6924920127795526,
+      "grad_norm": 0.8974957564028379,
+      "learning_rate": 3.1672947822466714e-07,
+      "loss": 0.0307,
+      "step": 6742
+    },
+    {
+      "epoch": 2.693290734824281,
+      "grad_norm": 0.8069628364489961,
+      "learning_rate": 3.151037486242181e-07,
+      "loss": 0.0253,
+      "step": 6744
+    },
+    {
+      "epoch": 2.6940894568690097,
+      "grad_norm": 1.0619428207925592,
+      "learning_rate": 3.13482066332102e-07,
+      "loss": 0.0328,
+      "step": 6746
+    },
+    {
+      "epoch": 2.6948881789137378,
+      "grad_norm": 1.0230784961867097,
+      "learning_rate": 3.1186443274930035e-07,
+      "loss": 0.033,
+      "step": 6748
+    },
+    {
+      "epoch": 2.6956869009584663,
+      "grad_norm": 0.9466613888969522,
+      "learning_rate": 3.102508492732964e-07,
+      "loss": 0.0271,
+      "step": 6750
+    },
+    {
+      "epoch": 2.696485623003195,
+      "grad_norm": 0.9853478032894241,
+      "learning_rate": 3.08641317298074e-07,
+      "loss": 0.0321,
+      "step": 6752
+    },
+    {
+      "epoch": 2.6972843450479234,
+      "grad_norm": 1.0460709064794675,
+      "learning_rate": 3.07035838214117e-07,
+      "loss": 0.0354,
+      "step": 6754
+    },
+    {
+      "epoch": 2.698083067092652,
+      "grad_norm": 0.9909587345397065,
+      "learning_rate": 3.0543441340840696e-07,
+      "loss": 0.0324,
+      "step": 6756
+    },
+    {
+      "epoch": 2.6988817891373804,
+      "grad_norm": 1.0147873081466183,
+      "learning_rate": 3.0383704426442396e-07,
+      "loss": 0.0325,
+      "step": 6758
+    },
+    {
+      "epoch": 2.6996805111821085,
+      "grad_norm": 0.9477745651098728,
+      "learning_rate": 3.022437321621452e-07,
+      "loss": 0.0307,
+      "step": 6760
+    },
+    {
+      "epoch": 2.700479233226837,
+      "grad_norm": 0.8576252217343889,
+      "learning_rate": 3.006544784780413e-07,
+      "loss": 0.0261,
+      "step": 6762
+    },
+    {
+      "epoch": 2.7012779552715656,
+      "grad_norm": 0.929114631572117,
+      "learning_rate": 2.9906928458507735e-07,
+      "loss": 0.0298,
+      "step": 6764
+    },
+    {
+      "epoch": 2.702076677316294,
+      "grad_norm": 0.8245393411728102,
+      "learning_rate": 2.9748815185271174e-07,
+      "loss": 0.0273,
+      "step": 6766
+    },
+    {
+      "epoch": 2.702875399361022,
+      "grad_norm": 1.0143221518300005,
+      "learning_rate": 2.959110816468935e-07,
+      "loss": 0.0294,
+      "step": 6768
+    },
+    {
+      "epoch": 2.7036741214057507,
+      "grad_norm": 1.1519123024674627,
+      "learning_rate": 2.94338075330064e-07,
+      "loss": 0.0287,
+      "step": 6770
+    },
+    {
+      "epoch": 2.7044728434504792,
+      "grad_norm": 0.9244500535568194,
+      "learning_rate": 2.927691342611505e-07,
+      "loss": 0.032,
+      "step": 6772
+    },
+    {
+      "epoch": 2.7052715654952078,
+      "grad_norm": 0.8604757808842944,
+      "learning_rate": 2.9120425979557e-07,
+      "loss": 0.0244,
+      "step": 6774
+    },
+    {
+      "epoch": 2.7060702875399363,
+      "grad_norm": 0.8443290403852196,
+      "learning_rate": 2.896434532852277e-07,
+      "loss": 0.0266,
+      "step": 6776
+    },
+    {
+      "epoch": 2.706869009584665,
+      "grad_norm": 0.9923984530744449,
+      "learning_rate": 2.880867160785128e-07,
+      "loss": 0.0299,
+      "step": 6778
+    },
+    {
+      "epoch": 2.707667731629393,
+      "grad_norm": 0.9535512491511288,
+      "learning_rate": 2.865340495202984e-07,
+      "loss": 0.0289,
+      "step": 6780
+    },
+    {
+      "epoch": 2.7084664536741214,
+      "grad_norm": 0.9510201264153861,
+      "learning_rate": 2.849854549519426e-07,
+      "loss": 0.0294,
+      "step": 6782
+    },
+    {
+      "epoch": 2.70926517571885,
+      "grad_norm": 0.9033130404655279,
+      "learning_rate": 2.834409337112842e-07,
+      "loss": 0.0273,
+      "step": 6784
+    },
+    {
+      "epoch": 2.710063897763578,
+      "grad_norm": 1.0548477729978563,
+      "learning_rate": 2.8190048713264586e-07,
+      "loss": 0.0311,
+      "step": 6786
+    },
+    {
+      "epoch": 2.7108626198083066,
+      "grad_norm": 0.9948422314656845,
+      "learning_rate": 2.8036411654682627e-07,
+      "loss": 0.0304,
+      "step": 6788
+    },
+    {
+      "epoch": 2.711661341853035,
+      "grad_norm": 0.8822930619200026,
+      "learning_rate": 2.7883182328110494e-07,
+      "loss": 0.0281,
+      "step": 6790
+    },
+    {
+      "epoch": 2.7124600638977636,
+      "grad_norm": 1.0705668513390736,
+      "learning_rate": 2.7730360865923954e-07,
+      "loss": 0.0291,
+      "step": 6792
+    },
+    {
+      "epoch": 2.713258785942492,
+      "grad_norm": 0.9950549300332445,
+      "learning_rate": 2.75779474001463e-07,
+      "loss": 0.0306,
+      "step": 6794
+    },
+    {
+      "epoch": 2.7140575079872207,
+      "grad_norm": 0.9852425718587328,
+      "learning_rate": 2.7425942062448254e-07,
+      "loss": 0.0291,
+      "step": 6796
+    },
+    {
+      "epoch": 2.7148562300319488,
+      "grad_norm": 0.9530301493388222,
+      "learning_rate": 2.727434498414827e-07,
+      "loss": 0.0305,
+      "step": 6798
+    },
+    {
+      "epoch": 2.7156549520766773,
+      "grad_norm": 0.9617585154543935,
+      "learning_rate": 2.712315629621176e-07,
+      "loss": 0.0304,
+      "step": 6800
+    },
+    {
+      "epoch": 2.716453674121406,
+      "grad_norm": 0.8160240654532492,
+      "learning_rate": 2.697237612925169e-07,
+      "loss": 0.0254,
+      "step": 6802
+    },
+    {
+      "epoch": 2.7172523961661343,
+      "grad_norm": 1.0732849620331193,
+      "learning_rate": 2.682200461352763e-07,
+      "loss": 0.029,
+      "step": 6804
+    },
+    {
+      "epoch": 2.7180511182108624,
+      "grad_norm": 1.1148345917930245,
+      "learning_rate": 2.6672041878946507e-07,
+      "loss": 0.0364,
+      "step": 6806
+    },
+    {
+      "epoch": 2.718849840255591,
+      "grad_norm": 0.9845373546577002,
+      "learning_rate": 2.6522488055062076e-07,
+      "loss": 0.0323,
+      "step": 6808
+    },
+    {
+      "epoch": 2.7196485623003195,
+      "grad_norm": 0.9694597420527851,
+      "learning_rate": 2.6373343271074657e-07,
+      "loss": 0.0305,
+      "step": 6810
+    },
+    {
+      "epoch": 2.720447284345048,
+      "grad_norm": 0.8865423882611635,
+      "learning_rate": 2.6224607655831236e-07,
+      "loss": 0.0301,
+      "step": 6812
+    },
+    {
+      "epoch": 2.7212460063897765,
+      "grad_norm": 1.007509492035804,
+      "learning_rate": 2.607628133782536e-07,
+      "loss": 0.0306,
+      "step": 6814
+    },
+    {
+      "epoch": 2.722044728434505,
+      "grad_norm": 0.8870706562836438,
+      "learning_rate": 2.5928364445196975e-07,
+      "loss": 0.0294,
+      "step": 6816
+    },
+    {
+      "epoch": 2.722843450479233,
+      "grad_norm": 0.9708567758541311,
+      "learning_rate": 2.578085710573247e-07,
+      "loss": 0.028,
+      "step": 6818
+    },
+    {
+      "epoch": 2.7236421725239617,
+      "grad_norm": 0.8934830296672568,
+      "learning_rate": 2.563375944686397e-07,
+      "loss": 0.0306,
+      "step": 6820
+    },
+    {
+      "epoch": 2.72444089456869,
+      "grad_norm": 1.072610973986754,
+      "learning_rate": 2.548707159567021e-07,
+      "loss": 0.0288,
+      "step": 6822
+    },
+    {
+      "epoch": 2.7252396166134183,
+      "grad_norm": 0.8222963836412455,
+      "learning_rate": 2.534079367887549e-07,
+      "loss": 0.0278,
+      "step": 6824
+    },
+    {
+      "epoch": 2.726038338658147,
+      "grad_norm": 0.9085074348725313,
+      "learning_rate": 2.519492582285027e-07,
+      "loss": 0.0304,
+      "step": 6826
+    },
+    {
+      "epoch": 2.7268370607028753,
+      "grad_norm": 0.9177610792802552,
+      "learning_rate": 2.504946815361065e-07,
+      "loss": 0.0331,
+      "step": 6828
+    },
+    {
+      "epoch": 2.727635782747604,
+      "grad_norm": 1.094130825800624,
+      "learning_rate": 2.4904420796818097e-07,
+      "loss": 0.0303,
+      "step": 6830
+    },
+    {
+      "epoch": 2.7284345047923324,
+      "grad_norm": 1.0338426208891864,
+      "learning_rate": 2.475978387778e-07,
+      "loss": 0.0311,
+      "step": 6832
+    },
+    {
+      "epoch": 2.729233226837061,
+      "grad_norm": 0.8384072672060133,
+      "learning_rate": 2.461555752144912e-07,
+      "loss": 0.0292,
+      "step": 6834
+    },
+    {
+      "epoch": 2.730031948881789,
+      "grad_norm": 1.015906477269788,
+      "learning_rate": 2.447174185242324e-07,
+      "loss": 0.0277,
+      "step": 6836
+    },
+    {
+      "epoch": 2.7308306709265175,
+      "grad_norm": 0.9215903866822917,
+      "learning_rate": 2.432833699494558e-07,
+      "loss": 0.0277,
+      "step": 6838
+    },
+    {
+      "epoch": 2.731629392971246,
+      "grad_norm": 0.913267265555896,
+      "learning_rate": 2.4185343072904376e-07,
+      "loss": 0.0248,
+      "step": 6840
+    },
+    {
+      "epoch": 2.7324281150159746,
+      "grad_norm": 0.9333269872247636,
+      "learning_rate": 2.404276020983304e-07,
+      "loss": 0.0292,
+      "step": 6842
+    },
+    {
+      "epoch": 2.7332268370607027,
+      "grad_norm": 1.0187672549108133,
+      "learning_rate": 2.3900588528909475e-07,
+      "loss": 0.0344,
+      "step": 6844
+    },
+    {
+      "epoch": 2.734025559105431,
+      "grad_norm": 0.9419705337564647,
+      "learning_rate": 2.375882815295677e-07,
+      "loss": 0.0286,
+      "step": 6846
+    },
+    {
+      "epoch": 2.7348242811501597,
+      "grad_norm": 0.918071642194504,
+      "learning_rate": 2.3617479204442462e-07,
+      "loss": 0.028,
+      "step": 6848
+    },
+    {
+      "epoch": 2.7356230031948883,
+      "grad_norm": 1.0220943478986433,
+      "learning_rate": 2.3476541805478647e-07,
+      "loss": 0.0287,
+      "step": 6850
+    },
+    {
+      "epoch": 2.736421725239617,
+      "grad_norm": 1.0904072022439066,
+      "learning_rate": 2.3336016077822154e-07,
+      "loss": 0.0296,
+      "step": 6852
+    },
+    {
+      "epoch": 2.737220447284345,
+      "grad_norm": 0.9166980890980327,
+      "learning_rate": 2.3195902142873593e-07,
+      "loss": 0.0269,
+      "step": 6854
+    },
+    {
+      "epoch": 2.7380191693290734,
+      "grad_norm": 0.9240202525664635,
+      "learning_rate": 2.305620012167853e-07,
+      "loss": 0.028,
+      "step": 6856
+    },
+    {
+      "epoch": 2.738817891373802,
+      "grad_norm": 0.9499828140577764,
+      "learning_rate": 2.2916910134926197e-07,
+      "loss": 0.0273,
+      "step": 6858
+    },
+    {
+      "epoch": 2.7396166134185305,
+      "grad_norm": 1.0046665742939103,
+      "learning_rate": 2.2778032302949948e-07,
+      "loss": 0.0299,
+      "step": 6860
+    },
+    {
+      "epoch": 2.7404153354632586,
+      "grad_norm": 0.9346433823459767,
+      "learning_rate": 2.2639566745727203e-07,
+      "loss": 0.0301,
+      "step": 6862
+    },
+    {
+      "epoch": 2.741214057507987,
+      "grad_norm": 1.0474839299661065,
+      "learning_rate": 2.2501513582879108e-07,
+      "loss": 0.035,
+      "step": 6864
+    },
+    {
+      "epoch": 2.7420127795527156,
+      "grad_norm": 0.9326384987743596,
+      "learning_rate": 2.236387293367054e-07,
+      "loss": 0.0267,
+      "step": 6866
+    },
+    {
+      "epoch": 2.742811501597444,
+      "grad_norm": 0.9513219854071373,
+      "learning_rate": 2.2226644917010153e-07,
+      "loss": 0.0299,
+      "step": 6868
+    },
+    {
+      "epoch": 2.7436102236421727,
+      "grad_norm": 0.8426342593529366,
+      "learning_rate": 2.208982965144979e-07,
+      "loss": 0.0245,
+      "step": 6870
+    },
+    {
+      "epoch": 2.744408945686901,
+      "grad_norm": 0.967982836150597,
+      "learning_rate": 2.1953427255185122e-07,
+      "loss": 0.0333,
+      "step": 6872
+    },
+    {
+      "epoch": 2.7452076677316293,
+      "grad_norm": 0.9628141955997789,
+      "learning_rate": 2.1817437846054889e-07,
+      "loss": 0.032,
+      "step": 6874
+    },
+    {
+      "epoch": 2.746006389776358,
+      "grad_norm": 0.9095860682951128,
+      "learning_rate": 2.1681861541541117e-07,
+      "loss": 0.027,
+      "step": 6876
+    },
+    {
+      "epoch": 2.7468051118210863,
+      "grad_norm": 0.9999018913590277,
+      "learning_rate": 2.1546698458768888e-07,
+      "loss": 0.0266,
+      "step": 6878
+    },
+    {
+      "epoch": 2.747603833865815,
+      "grad_norm": 0.9155500721532669,
+      "learning_rate": 2.1411948714506414e-07,
+      "loss": 0.0248,
+      "step": 6880
+    },
+    {
+      "epoch": 2.748402555910543,
+      "grad_norm": 0.899083057917597,
+      "learning_rate": 2.1277612425164796e-07,
+      "loss": 0.0291,
+      "step": 6882
+    },
+    {
+      "epoch": 2.7492012779552715,
+      "grad_norm": 1.0507417274391886,
+      "learning_rate": 2.1143689706797809e-07,
+      "loss": 0.0304,
+      "step": 6884
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.9237015927908853,
+      "learning_rate": 2.101018067510213e-07,
+      "loss": 0.0288,
+      "step": 6886
+    },
+    {
+      "epoch": 2.7507987220447285,
+      "grad_norm": 0.9396213034157446,
+      "learning_rate": 2.0877085445416889e-07,
+      "loss": 0.0313,
+      "step": 6888
+    },
+    {
+      "epoch": 2.751597444089457,
+      "grad_norm": 0.893230887911447,
+      "learning_rate": 2.0744404132723882e-07,
+      "loss": 0.0271,
+      "step": 6890
+    },
+    {
+      "epoch": 2.752396166134185,
+      "grad_norm": 1.1227883203425046,
+      "learning_rate": 2.0612136851647258e-07,
+      "loss": 0.028,
+      "step": 6892
+    },
+    {
+      "epoch": 2.7531948881789137,
+      "grad_norm": 1.1000266312956004,
+      "learning_rate": 2.0480283716453388e-07,
+      "loss": 0.0316,
+      "step": 6894
+    },
+    {
+      "epoch": 2.753993610223642,
+      "grad_norm": 1.0036859264242648,
+      "learning_rate": 2.034884484105093e-07,
+      "loss": 0.032,
+      "step": 6896
+    },
+    {
+      "epoch": 2.7547923322683707,
+      "grad_norm": 0.8664572975169675,
+      "learning_rate": 2.0217820338990723e-07,
+      "loss": 0.0293,
+      "step": 6898
+    },
+    {
+      "epoch": 2.755591054313099,
+      "grad_norm": 1.152106392290739,
+      "learning_rate": 2.0087210323465555e-07,
+      "loss": 0.0271,
+      "step": 6900
+    },
+    {
+      "epoch": 2.7563897763578273,
+      "grad_norm": 0.9072929731158895,
+      "learning_rate": 1.9957014907310224e-07,
+      "loss": 0.0297,
+      "step": 6902
+    },
+    {
+      "epoch": 2.757188498402556,
+      "grad_norm": 0.899746462852363,
+      "learning_rate": 1.98272342030012e-07,
+      "loss": 0.0269,
+      "step": 6904
+    },
+    {
+      "epoch": 2.7579872204472844,
+      "grad_norm": 0.9831117034435982,
+      "learning_rate": 1.96978683226568e-07,
+      "loss": 0.0286,
+      "step": 6906
+    },
+    {
+      "epoch": 2.758785942492013,
+      "grad_norm": 0.9654610196040313,
+      "learning_rate": 1.9568917378037012e-07,
+      "loss": 0.0286,
+      "step": 6908
+    },
+    {
+      "epoch": 2.7595846645367414,
+      "grad_norm": 0.8702316945781633,
+      "learning_rate": 1.9440381480543115e-07,
+      "loss": 0.0292,
+      "step": 6910
+    },
+    {
+      "epoch": 2.7603833865814695,
+      "grad_norm": 0.8701781601681666,
+      "learning_rate": 1.9312260741218114e-07,
+      "loss": 0.0296,
+      "step": 6912
+    },
+    {
+      "epoch": 2.761182108626198,
+      "grad_norm": 1.0906874136348748,
+      "learning_rate": 1.9184555270746198e-07,
+      "loss": 0.032,
+      "step": 6914
+    },
+    {
+      "epoch": 2.7619808306709266,
+      "grad_norm": 0.8789894864802608,
+      "learning_rate": 1.9057265179452945e-07,
+      "loss": 0.0283,
+      "step": 6916
+    },
+    {
+      "epoch": 2.762779552715655,
+      "grad_norm": 0.8289933518344175,
+      "learning_rate": 1.8930390577304836e-07,
+      "loss": 0.0254,
+      "step": 6918
+    },
+    {
+      "epoch": 2.763578274760383,
+      "grad_norm": 1.109173584946323,
+      "learning_rate": 1.8803931573909584e-07,
+      "loss": 0.0325,
+      "step": 6920
+    },
+    {
+      "epoch": 2.7643769968051117,
+      "grad_norm": 0.8437516857169298,
+      "learning_rate": 1.8677888278515854e-07,
+      "loss": 0.0275,
+      "step": 6922
+    },
+    {
+      "epoch": 2.7651757188498403,
+      "grad_norm": 0.957579586236523,
+      "learning_rate": 1.8552260800013266e-07,
+      "loss": 0.0324,
+      "step": 6924
+    },
+    {
+      "epoch": 2.765974440894569,
+      "grad_norm": 1.133247804875847,
+      "learning_rate": 1.8427049246932005e-07,
+      "loss": 0.0303,
+      "step": 6926
+    },
+    {
+      "epoch": 2.7667731629392973,
+      "grad_norm": 0.9042200415102326,
+      "learning_rate": 1.8302253727443041e-07,
+      "loss": 0.0261,
+      "step": 6928
+    },
+    {
+      "epoch": 2.7675718849840254,
+      "grad_norm": 1.0109916146672329,
+      "learning_rate": 1.817787434935797e-07,
+      "loss": 0.0317,
+      "step": 6930
+    },
+    {
+      "epoch": 2.768370607028754,
+      "grad_norm": 0.8886867588028137,
+      "learning_rate": 1.805391122012884e-07,
+      "loss": 0.0307,
+      "step": 6932
+    },
+    {
+      "epoch": 2.7691693290734825,
+      "grad_norm": 0.9881453864615531,
+      "learning_rate": 1.7930364446848035e-07,
+      "loss": 0.0282,
+      "step": 6934
+    },
+    {
+      "epoch": 2.769968051118211,
+      "grad_norm": 0.9753621627206346,
+      "learning_rate": 1.7807234136248296e-07,
+      "loss": 0.0316,
+      "step": 6936
+    },
+    {
+      "epoch": 2.770766773162939,
+      "grad_norm": 0.9729298132394265,
+      "learning_rate": 1.7684520394702697e-07,
+      "loss": 0.0322,
+      "step": 6938
+    },
+    {
+      "epoch": 2.7715654952076676,
+      "grad_norm": 1.0067163359277183,
+      "learning_rate": 1.7562223328224327e-07,
+      "loss": 0.0273,
+      "step": 6940
+    },
+    {
+      "epoch": 2.772364217252396,
+      "grad_norm": 1.014978444523247,
+      "learning_rate": 1.7440343042466225e-07,
+      "loss": 0.0222,
+      "step": 6942
+    },
+    {
+      "epoch": 2.7731629392971247,
+      "grad_norm": 0.9727814177226888,
+      "learning_rate": 1.731887964272144e-07,
+      "loss": 0.0284,
+      "step": 6944
+    },
+    {
+      "epoch": 2.773961661341853,
+      "grad_norm": 0.8674877180013327,
+      "learning_rate": 1.7197833233922933e-07,
+      "loss": 0.0266,
+      "step": 6946
+    },
+    {
+      "epoch": 2.7747603833865817,
+      "grad_norm": 0.92093939870497,
+      "learning_rate": 1.7077203920643548e-07,
+      "loss": 0.0277,
+      "step": 6948
+    },
+    {
+      "epoch": 2.77555910543131,
+      "grad_norm": 1.0354340580208805,
+      "learning_rate": 1.695699180709537e-07,
+      "loss": 0.0318,
+      "step": 6950
+    },
+    {
+      "epoch": 2.7763578274760383,
+      "grad_norm": 0.8959582446725475,
+      "learning_rate": 1.6837196997130434e-07,
+      "loss": 0.0273,
+      "step": 6952
+    },
+    {
+      "epoch": 2.777156549520767,
+      "grad_norm": 0.9082092205680304,
+      "learning_rate": 1.671781959424018e-07,
+      "loss": 0.0283,
+      "step": 6954
+    },
+    {
+      "epoch": 2.777955271565495,
+      "grad_norm": 0.8526839728174191,
+      "learning_rate": 1.6598859701555448e-07,
+      "loss": 0.0279,
+      "step": 6956
+    },
+    {
+      "epoch": 2.7787539936102235,
+      "grad_norm": 0.9183824299916817,
+      "learning_rate": 1.648031742184619e-07,
+      "loss": 0.0221,
+      "step": 6958
+    },
+    {
+      "epoch": 2.779552715654952,
+      "grad_norm": 0.8785486078394016,
+      "learning_rate": 1.6362192857521942e-07,
+      "loss": 0.028,
+      "step": 6960
+    },
+    {
+      "epoch": 2.7803514376996805,
+      "grad_norm": 0.984747118195936,
+      "learning_rate": 1.6244486110631062e-07,
+      "loss": 0.0343,
+      "step": 6962
+    },
+    {
+      "epoch": 2.781150159744409,
+      "grad_norm": 0.8833396654406691,
+      "learning_rate": 1.6127197282861106e-07,
+      "loss": 0.0275,
+      "step": 6964
+    },
+    {
+      "epoch": 2.7819488817891376,
+      "grad_norm": 0.8586218999503099,
+      "learning_rate": 1.6010326475538628e-07,
+      "loss": 0.0267,
+      "step": 6966
+    },
+    {
+      "epoch": 2.7827476038338657,
+      "grad_norm": 1.1132610814922481,
+      "learning_rate": 1.5893873789628812e-07,
+      "loss": 0.0357,
+      "step": 6968
+    },
+    {
+      "epoch": 2.783546325878594,
+      "grad_norm": 1.0377705902735281,
+      "learning_rate": 1.5777839325735955e-07,
+      "loss": 0.0348,
+      "step": 6970
+    },
+    {
+      "epoch": 2.7843450479233227,
+      "grad_norm": 0.8553477275906,
+      "learning_rate": 1.5662223184102876e-07,
+      "loss": 0.0282,
+      "step": 6972
+    },
+    {
+      "epoch": 2.7851437699680512,
+      "grad_norm": 0.9887534929150167,
+      "learning_rate": 1.55470254646109e-07,
+      "loss": 0.0339,
+      "step": 6974
+    },
+    {
+      "epoch": 2.7859424920127793,
+      "grad_norm": 0.8463180888040372,
+      "learning_rate": 1.5432246266780083e-07,
+      "loss": 0.0265,
+      "step": 6976
+    },
+    {
+      "epoch": 2.786741214057508,
+      "grad_norm": 0.9279956838643539,
+      "learning_rate": 1.5317885689768775e-07,
+      "loss": 0.028,
+      "step": 6978
+    },
+    {
+      "epoch": 2.7875399361022364,
+      "grad_norm": 0.898362447794585,
+      "learning_rate": 1.520394383237378e-07,
+      "loss": 0.0254,
+      "step": 6980
+    },
+    {
+      "epoch": 2.788338658146965,
+      "grad_norm": 0.999242247390689,
+      "learning_rate": 1.5090420793030025e-07,
+      "loss": 0.026,
+      "step": 6982
+    },
+    {
+      "epoch": 2.7891373801916934,
+      "grad_norm": 1.0069528441967504,
+      "learning_rate": 1.4977316669810782e-07,
+      "loss": 0.0299,
+      "step": 6984
+    },
+    {
+      "epoch": 2.789936102236422,
+      "grad_norm": 0.9597592645048728,
+      "learning_rate": 1.4864631560427277e-07,
+      "loss": 0.0286,
+      "step": 6986
+    },
+    {
+      "epoch": 2.79073482428115,
+      "grad_norm": 0.8133798258927686,
+      "learning_rate": 1.4752365562228865e-07,
+      "loss": 0.0256,
+      "step": 6988
+    },
+    {
+      "epoch": 2.7915335463258786,
+      "grad_norm": 1.0299858930198784,
+      "learning_rate": 1.4640518772202794e-07,
+      "loss": 0.0287,
+      "step": 6990
+    },
+    {
+      "epoch": 2.792332268370607,
+      "grad_norm": 0.8756875752154023,
+      "learning_rate": 1.4529091286973994e-07,
+      "loss": 0.0285,
+      "step": 6992
+    },
+    {
+      "epoch": 2.793130990415335,
+      "grad_norm": 0.9409132612586905,
+      "learning_rate": 1.4418083202805467e-07,
+      "loss": 0.031,
+      "step": 6994
+    },
+    {
+      "epoch": 2.7939297124600637,
+      "grad_norm": 0.9200301064033846,
+      "learning_rate": 1.4307494615597716e-07,
+      "loss": 0.0311,
+      "step": 6996
+    },
+    {
+      "epoch": 2.7947284345047922,
+      "grad_norm": 0.8145829376320072,
+      "learning_rate": 1.4197325620888714e-07,
+      "loss": 0.0246,
+      "step": 6998
+    },
+    {
+      "epoch": 2.7955271565495208,
+      "grad_norm": 0.9865390880493666,
+      "learning_rate": 1.4087576313854212e-07,
+      "loss": 0.0296,
+      "step": 7000
+    },
+    {
+      "epoch": 2.7955271565495208,
+      "eval_loss": 0.1819150149822235,
+      "eval_runtime": 417.4017,
+      "eval_samples_per_second": 42.662,
+      "eval_steps_per_second": 5.333,
+      "step": 7000
+    },
+    {
+      "epoch": 2.7963258785942493,
+      "grad_norm": 1.0878362509592698,
+      "learning_rate": 1.397824678930715e-07,
+      "loss": 0.0269,
+      "step": 7002
+    },
+    {
+      "epoch": 2.797124600638978,
+      "grad_norm": 0.8833659039797074,
+      "learning_rate": 1.386933714169797e-07,
+      "loss": 0.0295,
+      "step": 7004
+    },
+    {
+      "epoch": 2.797923322683706,
+      "grad_norm": 1.198555362781041,
+      "learning_rate": 1.3760847465114413e-07,
+      "loss": 0.0231,
+      "step": 7006
+    },
+    {
+      "epoch": 2.7987220447284344,
+      "grad_norm": 0.9766022877177375,
+      "learning_rate": 1.365277785328123e-07,
+      "loss": 0.0316,
+      "step": 7008
+    },
+    {
+      "epoch": 2.799520766773163,
+      "grad_norm": 0.9931479598265185,
+      "learning_rate": 1.3545128399560349e-07,
+      "loss": 0.0307,
+      "step": 7010
+    },
+    {
+      "epoch": 2.8003194888178915,
+      "grad_norm": 0.9470485404305614,
+      "learning_rate": 1.3437899196950765e-07,
+      "loss": 0.03,
+      "step": 7012
+    },
+    {
+      "epoch": 2.8011182108626196,
+      "grad_norm": 1.0425827226683049,
+      "learning_rate": 1.3331090338088437e-07,
+      "loss": 0.0291,
+      "step": 7014
+    },
+    {
+      "epoch": 2.801916932907348,
+      "grad_norm": 0.9479599981229176,
+      "learning_rate": 1.3224701915246053e-07,
+      "loss": 0.033,
+      "step": 7016
+    },
+    {
+      "epoch": 2.8027156549520766,
+      "grad_norm": 1.0722352133209674,
+      "learning_rate": 1.3118734020333257e-07,
+      "loss": 0.0305,
+      "step": 7018
+    },
+    {
+      "epoch": 2.803514376996805,
+      "grad_norm": 1.1283019296777128,
+      "learning_rate": 1.3013186744896323e-07,
+      "loss": 0.0322,
+      "step": 7020
+    },
+    {
+      "epoch": 2.8043130990415337,
+      "grad_norm": 1.0443979108148196,
+      "learning_rate": 1.2908060180118088e-07,
+      "loss": 0.0279,
+      "step": 7022
+    },
+    {
+      "epoch": 2.8051118210862622,
+      "grad_norm": 0.9204477323838135,
+      "learning_rate": 1.280335441681796e-07,
+      "loss": 0.026,
+      "step": 7024
+    },
+    {
+      "epoch": 2.8059105431309903,
+      "grad_norm": 1.091376167660413,
+      "learning_rate": 1.2699069545451858e-07,
+      "loss": 0.0303,
+      "step": 7026
+    },
+    {
+      "epoch": 2.806709265175719,
+      "grad_norm": 0.8186223224524324,
+      "learning_rate": 1.2595205656112164e-07,
+      "loss": 0.0251,
+      "step": 7028
+    },
+    {
+      "epoch": 2.8075079872204474,
+      "grad_norm": 0.9426669872933783,
+      "learning_rate": 1.2491762838527376e-07,
+      "loss": 0.0278,
+      "step": 7030
+    },
+    {
+      "epoch": 2.8083067092651754,
+      "grad_norm": 0.8562334740689089,
+      "learning_rate": 1.2388741182062348e-07,
+      "loss": 0.0254,
+      "step": 7032
+    },
+    {
+      "epoch": 2.809105431309904,
+      "grad_norm": 1.0003749268195539,
+      "learning_rate": 1.2286140775718048e-07,
+      "loss": 0.0254,
+      "step": 7034
+    },
+    {
+      "epoch": 2.8099041533546325,
+      "grad_norm": 0.9287476537267818,
+      "learning_rate": 1.2183961708131574e-07,
+      "loss": 0.0331,
+      "step": 7036
+    },
+    {
+      "epoch": 2.810702875399361,
+      "grad_norm": 1.0100009275296533,
+      "learning_rate": 1.2082204067576043e-07,
+      "loss": 0.0339,
+      "step": 7038
+    },
+    {
+      "epoch": 2.8115015974440896,
+      "grad_norm": 1.0019967241849932,
+      "learning_rate": 1.198086794196035e-07,
+      "loss": 0.034,
+      "step": 7040
+    },
+    {
+      "epoch": 2.812300319488818,
+      "grad_norm": 1.0274376107046983,
+      "learning_rate": 1.187995341882947e-07,
+      "loss": 0.0296,
+      "step": 7042
+    },
+    {
+      "epoch": 2.813099041533546,
+      "grad_norm": 0.9758492633679886,
+      "learning_rate": 1.1779460585363945e-07,
+      "loss": 0.0304,
+      "step": 7044
+    },
+    {
+      "epoch": 2.8138977635782747,
+      "grad_norm": 1.0055929014699148,
+      "learning_rate": 1.1679389528380159e-07,
+      "loss": 0.0285,
+      "step": 7046
+    },
+    {
+      "epoch": 2.8146964856230032,
+      "grad_norm": 0.9825352483369219,
+      "learning_rate": 1.1579740334330014e-07,
+      "loss": 0.0279,
+      "step": 7048
+    },
+    {
+      "epoch": 2.8154952076677318,
+      "grad_norm": 0.9663886663279524,
+      "learning_rate": 1.1480513089301037e-07,
+      "loss": 0.0238,
+      "step": 7050
+    },
+    {
+      "epoch": 2.81629392971246,
+      "grad_norm": 0.7632311003499087,
+      "learning_rate": 1.1381707879016158e-07,
+      "loss": 0.0228,
+      "step": 7052
+    },
+    {
+      "epoch": 2.8170926517571884,
+      "grad_norm": 0.8411185807947373,
+      "learning_rate": 1.1283324788833872e-07,
+      "loss": 0.0261,
+      "step": 7054
+    },
+    {
+      "epoch": 2.817891373801917,
+      "grad_norm": 1.0106383088967026,
+      "learning_rate": 1.1185363903747748e-07,
+      "loss": 0.0249,
+      "step": 7056
+    },
+    {
+      "epoch": 2.8186900958466454,
+      "grad_norm": 1.0568473872652466,
+      "learning_rate": 1.1087825308386812e-07,
+      "loss": 0.0328,
+      "step": 7058
+    },
+    {
+      "epoch": 2.819488817891374,
+      "grad_norm": 1.0600837943890258,
+      "learning_rate": 1.0990709087015217e-07,
+      "loss": 0.0288,
+      "step": 7060
+    },
+    {
+      "epoch": 2.8202875399361025,
+      "grad_norm": 0.9659386588248836,
+      "learning_rate": 1.0894015323532181e-07,
+      "loss": 0.0294,
+      "step": 7062
+    },
+    {
+      "epoch": 2.8210862619808306,
+      "grad_norm": 0.9822944918742355,
+      "learning_rate": 1.0797744101472052e-07,
+      "loss": 0.0298,
+      "step": 7064
+    },
+    {
+      "epoch": 2.821884984025559,
+      "grad_norm": 0.9411392883906924,
+      "learning_rate": 1.0701895504004022e-07,
+      "loss": 0.0268,
+      "step": 7066
+    },
+    {
+      "epoch": 2.8226837060702876,
+      "grad_norm": 0.8205930442237354,
+      "learning_rate": 1.0606469613932247e-07,
+      "loss": 0.0285,
+      "step": 7068
+    },
+    {
+      "epoch": 2.8234824281150157,
+      "grad_norm": 0.9585034609165559,
+      "learning_rate": 1.0511466513695778e-07,
+      "loss": 0.0295,
+      "step": 7070
+    },
+    {
+      "epoch": 2.8242811501597442,
+      "grad_norm": 1.017339428173637,
+      "learning_rate": 1.0416886285368188e-07,
+      "loss": 0.032,
+      "step": 7072
+    },
+    {
+      "epoch": 2.8250798722044728,
+      "grad_norm": 0.9480053562427759,
+      "learning_rate": 1.032272901065795e-07,
+      "loss": 0.0276,
+      "step": 7074
+    },
+    {
+      "epoch": 2.8258785942492013,
+      "grad_norm": 0.9241183013217804,
+      "learning_rate": 1.0228994770908052e-07,
+      "loss": 0.0313,
+      "step": 7076
+    },
+    {
+      "epoch": 2.82667731629393,
+      "grad_norm": 0.8600565904245382,
+      "learning_rate": 1.0135683647096107e-07,
+      "loss": 0.0312,
+      "step": 7078
+    },
+    {
+      "epoch": 2.8274760383386583,
+      "grad_norm": 0.857256000234177,
+      "learning_rate": 1.0042795719833964e-07,
+      "loss": 0.0304,
+      "step": 7080
+    },
+    {
+      "epoch": 2.8282747603833864,
+      "grad_norm": 0.9257782290418327,
+      "learning_rate": 9.950331069368102e-08,
+      "loss": 0.029,
+      "step": 7082
+    },
+    {
+      "epoch": 2.829073482428115,
+      "grad_norm": 0.9014008656073322,
+      "learning_rate": 9.858289775579289e-08,
+      "loss": 0.0277,
+      "step": 7084
+    },
+    {
+      "epoch": 2.8298722044728435,
+      "grad_norm": 0.9335945085748736,
+      "learning_rate": 9.766671917982529e-08,
+      "loss": 0.0258,
+      "step": 7086
+    },
+    {
+      "epoch": 2.830670926517572,
+      "grad_norm": 0.893406276113889,
+      "learning_rate": 9.675477575726954e-08,
+      "loss": 0.0286,
+      "step": 7088
+    },
+    {
+      "epoch": 2.8314696485623,
+      "grad_norm": 0.9135616915468692,
+      "learning_rate": 9.58470682759588e-08,
+      "loss": 0.0267,
+      "step": 7090
+    },
+    {
+      "epoch": 2.8322683706070286,
+      "grad_norm": 0.8914005786007674,
+      "learning_rate": 9.494359752006687e-08,
+      "loss": 0.0267,
+      "step": 7092
+    },
+    {
+      "epoch": 2.833067092651757,
+      "grad_norm": 0.9868182622898214,
+      "learning_rate": 9.404436427010777e-08,
+      "loss": 0.0259,
+      "step": 7094
+    },
+    {
+      "epoch": 2.8338658146964857,
+      "grad_norm": 0.8736365212101689,
+      "learning_rate": 9.314936930293283e-08,
+      "loss": 0.0238,
+      "step": 7096
+    },
+    {
+      "epoch": 2.834664536741214,
+      "grad_norm": 0.8341663421385362,
+      "learning_rate": 9.225861339173415e-08,
+      "loss": 0.0296,
+      "step": 7098
+    },
+    {
+      "epoch": 2.8354632587859427,
+      "grad_norm": 0.9404555958593751,
+      "learning_rate": 9.137209730604113e-08,
+      "loss": 0.029,
+      "step": 7100
+    },
+    {
+      "epoch": 2.836261980830671,
+      "grad_norm": 0.9105942982702757,
+      "learning_rate": 9.048982181171895e-08,
+      "loss": 0.0289,
+      "step": 7102
+    },
+    {
+      "epoch": 2.8370607028753994,
+      "grad_norm": 1.0814449020759576,
+      "learning_rate": 8.961178767097178e-08,
+      "loss": 0.0308,
+      "step": 7104
+    },
+    {
+      "epoch": 2.837859424920128,
+      "grad_norm": 0.8932338308675243,
+      "learning_rate": 8.873799564233676e-08,
+      "loss": 0.0301,
+      "step": 7106
+    },
+    {
+      "epoch": 2.838658146964856,
+      "grad_norm": 1.012162689411644,
+      "learning_rate": 8.786844648068837e-08,
+      "loss": 0.0313,
+      "step": 7108
+    },
+    {
+      "epoch": 2.8394568690095845,
+      "grad_norm": 1.0902329946901537,
+      "learning_rate": 8.700314093723572e-08,
+      "loss": 0.0315,
+      "step": 7110
+    },
+    {
+      "epoch": 2.840255591054313,
+      "grad_norm": 0.9095047254271645,
+      "learning_rate": 8.614207975952083e-08,
+      "loss": 0.0311,
+      "step": 7112
+    },
+    {
+      "epoch": 2.8410543130990416,
+      "grad_norm": 0.9218237866635856,
+      "learning_rate": 8.528526369141809e-08,
+      "loss": 0.0302,
+      "step": 7114
+    },
+    {
+      "epoch": 2.84185303514377,
+      "grad_norm": 0.8660219136797084,
+      "learning_rate": 8.443269347313765e-08,
+      "loss": 0.0265,
+      "step": 7116
+    },
+    {
+      "epoch": 2.8426517571884986,
+      "grad_norm": 0.9364401145220109,
+      "learning_rate": 8.358436984121865e-08,
+      "loss": 0.0273,
+      "step": 7118
+    },
+    {
+      "epoch": 2.8434504792332267,
+      "grad_norm": 1.023124638752514,
+      "learning_rate": 8.274029352853264e-08,
+      "loss": 0.0276,
+      "step": 7120
+    },
+    {
+      "epoch": 2.844249201277955,
+      "grad_norm": 0.8561164210287661,
+      "learning_rate": 8.190046526428241e-08,
+      "loss": 0.0291,
+      "step": 7122
+    },
+    {
+      "epoch": 2.8450479233226837,
+      "grad_norm": 0.8438795870319646,
+      "learning_rate": 8.106488577399985e-08,
+      "loss": 0.0264,
+      "step": 7124
+    },
+    {
+      "epoch": 2.8458466453674123,
+      "grad_norm": 0.9601387385173464,
+      "learning_rate": 8.02335557795464e-08,
+      "loss": 0.0268,
+      "step": 7126
+    },
+    {
+      "epoch": 2.8466453674121404,
+      "grad_norm": 1.0416344807221194,
+      "learning_rate": 7.940647599911477e-08,
+      "loss": 0.0274,
+      "step": 7128
+    },
+    {
+      "epoch": 2.847444089456869,
+      "grad_norm": 1.030804565632412,
+      "learning_rate": 7.858364714722122e-08,
+      "loss": 0.0297,
+      "step": 7130
+    },
+    {
+      "epoch": 2.8482428115015974,
+      "grad_norm": 0.848403943487444,
+      "learning_rate": 7.776506993471323e-08,
+      "loss": 0.0262,
+      "step": 7132
+    },
+    {
+      "epoch": 2.849041533546326,
+      "grad_norm": 0.9248426395931749,
+      "learning_rate": 7.695074506876566e-08,
+      "loss": 0.0275,
+      "step": 7134
+    },
+    {
+      "epoch": 2.8498402555910545,
+      "grad_norm": 1.0380277597027192,
+      "learning_rate": 7.614067325287632e-08,
+      "loss": 0.029,
+      "step": 7136
+    },
+    {
+      "epoch": 2.850638977635783,
+      "grad_norm": 1.0174264329449134,
+      "learning_rate": 7.533485518687211e-08,
+      "loss": 0.0262,
+      "step": 7138
+    },
+    {
+      "epoch": 2.851437699680511,
+      "grad_norm": 0.9035723830180614,
+      "learning_rate": 7.453329156690337e-08,
+      "loss": 0.0312,
+      "step": 7140
+    },
+    {
+      "epoch": 2.8522364217252396,
+      "grad_norm": 0.8948154240149504,
+      "learning_rate": 7.373598308544505e-08,
+      "loss": 0.0286,
+      "step": 7142
+    },
+    {
+      "epoch": 2.853035143769968,
+      "grad_norm": 0.8767549377755169,
+      "learning_rate": 7.294293043129785e-08,
+      "loss": 0.0271,
+      "step": 7144
+    },
+    {
+      "epoch": 2.8538338658146962,
+      "grad_norm": 1.0359127327227582,
+      "learning_rate": 7.215413428958263e-08,
+      "loss": 0.0299,
+      "step": 7146
+    },
+    {
+      "epoch": 2.8546325878594248,
+      "grad_norm": 1.054388868239098,
+      "learning_rate": 7.136959534174592e-08,
+      "loss": 0.0295,
+      "step": 7148
+    },
+    {
+      "epoch": 2.8554313099041533,
+      "grad_norm": 0.9105886167898716,
+      "learning_rate": 7.058931426555449e-08,
+      "loss": 0.0278,
+      "step": 7150
+    },
+    {
+      "epoch": 2.856230031948882,
+      "grad_norm": 0.9273333490997623,
+      "learning_rate": 6.981329173509909e-08,
+      "loss": 0.0287,
+      "step": 7152
+    },
+    {
+      "epoch": 2.8570287539936103,
+      "grad_norm": 0.8079326225611323,
+      "learning_rate": 6.904152842078848e-08,
+      "loss": 0.0269,
+      "step": 7154
+    },
+    {
+      "epoch": 2.857827476038339,
+      "grad_norm": 0.9892975697892658,
+      "learning_rate": 6.827402498935377e-08,
+      "loss": 0.0336,
+      "step": 7156
+    },
+    {
+      "epoch": 2.858626198083067,
+      "grad_norm": 0.9535750532619426,
+      "learning_rate": 6.75107821038462e-08,
+      "loss": 0.029,
+      "step": 7158
+    },
+    {
+      "epoch": 2.8594249201277955,
+      "grad_norm": 0.8407797191652399,
+      "learning_rate": 6.675180042363505e-08,
+      "loss": 0.0254,
+      "step": 7160
+    },
+    {
+      "epoch": 2.860223642172524,
+      "grad_norm": 1.0055696204335296,
+      "learning_rate": 6.599708060440857e-08,
+      "loss": 0.0291,
+      "step": 7162
+    },
+    {
+      "epoch": 2.8610223642172525,
+      "grad_norm": 0.860800590976241,
+      "learning_rate": 6.524662329817411e-08,
+      "loss": 0.0244,
+      "step": 7164
+    },
+    {
+      "epoch": 2.8618210862619806,
+      "grad_norm": 0.9470630865172524,
+      "learning_rate": 6.450042915325527e-08,
+      "loss": 0.0288,
+      "step": 7166
+    },
+    {
+      "epoch": 2.862619808306709,
+      "grad_norm": 0.917809030684146,
+      "learning_rate": 6.375849881429418e-08,
+      "loss": 0.0278,
+      "step": 7168
+    },
+    {
+      "epoch": 2.8634185303514377,
+      "grad_norm": 0.9384143679170448,
+      "learning_rate": 6.302083292224814e-08,
+      "loss": 0.0312,
+      "step": 7170
+    },
+    {
+      "epoch": 2.864217252396166,
+      "grad_norm": 0.8768613951600329,
+      "learning_rate": 6.22874321143907e-08,
+      "loss": 0.0249,
+      "step": 7172
+    },
+    {
+      "epoch": 2.8650159744408947,
+      "grad_norm": 0.7174846049452049,
+      "learning_rate": 6.15582970243117e-08,
+      "loss": 0.0223,
+      "step": 7174
+    },
+    {
+      "epoch": 2.8658146964856233,
+      "grad_norm": 1.0161897086560874,
+      "learning_rate": 6.083342828191453e-08,
+      "loss": 0.0269,
+      "step": 7176
+    },
+    {
+      "epoch": 2.8666134185303513,
+      "grad_norm": 0.9854394127230478,
+      "learning_rate": 6.011282651341655e-08,
+      "loss": 0.0288,
+      "step": 7178
+    },
+    {
+      "epoch": 2.86741214057508,
+      "grad_norm": 0.901696634188199,
+      "learning_rate": 5.9396492341351475e-08,
+      "loss": 0.0276,
+      "step": 7180
+    },
+    {
+      "epoch": 2.8682108626198084,
+      "grad_norm": 0.9841680067526405,
+      "learning_rate": 5.868442638456373e-08,
+      "loss": 0.0288,
+      "step": 7182
+    },
+    {
+      "epoch": 2.8690095846645365,
+      "grad_norm": 0.9134522070907273,
+      "learning_rate": 5.797662925821068e-08,
+      "loss": 0.03,
+      "step": 7184
+    },
+    {
+      "epoch": 2.869808306709265,
+      "grad_norm": 1.0297579858038646,
+      "learning_rate": 5.7273101573762644e-08,
+      "loss": 0.0297,
+      "step": 7186
+    },
+    {
+      "epoch": 2.8706070287539935,
+      "grad_norm": 1.002668230554201,
+      "learning_rate": 5.6573843939001224e-08,
+      "loss": 0.0321,
+      "step": 7188
+    },
+    {
+      "epoch": 2.871405750798722,
+      "grad_norm": 0.8448908330163806,
+      "learning_rate": 5.5878856958018755e-08,
+      "loss": 0.0275,
+      "step": 7190
+    },
+    {
+      "epoch": 2.8722044728434506,
+      "grad_norm": 0.8951090877529749,
+      "learning_rate": 5.518814123121885e-08,
+      "loss": 0.028,
+      "step": 7192
+    },
+    {
+      "epoch": 2.873003194888179,
+      "grad_norm": 0.8887400334865615,
+      "learning_rate": 5.450169735531419e-08,
+      "loss": 0.0266,
+      "step": 7194
+    },
+    {
+      "epoch": 2.873801916932907,
+      "grad_norm": 0.9527424771255477,
+      "learning_rate": 5.381952592332762e-08,
+      "loss": 0.0292,
+      "step": 7196
+    },
+    {
+      "epoch": 2.8746006389776357,
+      "grad_norm": 0.9236990470777641,
+      "learning_rate": 5.3141627524591066e-08,
+      "loss": 0.0268,
+      "step": 7198
+    },
+    {
+      "epoch": 2.8753993610223643,
+      "grad_norm": 0.99547523531349,
+      "learning_rate": 5.246800274474439e-08,
+      "loss": 0.0287,
+      "step": 7200
+    },
+    {
+      "epoch": 2.876198083067093,
+      "grad_norm": 1.0395745492495505,
+      "learning_rate": 5.179865216573654e-08,
+      "loss": 0.0272,
+      "step": 7202
+    },
+    {
+      "epoch": 2.876996805111821,
+      "grad_norm": 0.9239013292846188,
+      "learning_rate": 5.1133576365823277e-08,
+      "loss": 0.03,
+      "step": 7204
+    },
+    {
+      "epoch": 2.8777955271565494,
+      "grad_norm": 0.8510701957695702,
+      "learning_rate": 5.047277591956668e-08,
+      "loss": 0.0237,
+      "step": 7206
+    },
+    {
+      "epoch": 2.878594249201278,
+      "grad_norm": 1.0550009694911764,
+      "learning_rate": 4.981625139783619e-08,
+      "loss": 0.0264,
+      "step": 7208
+    },
+    {
+      "epoch": 2.8793929712460065,
+      "grad_norm": 0.8651740589602106,
+      "learning_rate": 4.916400336780758e-08,
+      "loss": 0.0289,
+      "step": 7210
+    },
+    {
+      "epoch": 2.880191693290735,
+      "grad_norm": 0.9833410681522953,
+      "learning_rate": 4.851603239296065e-08,
+      "loss": 0.0263,
+      "step": 7212
+    },
+    {
+      "epoch": 2.8809904153354635,
+      "grad_norm": 0.9156057914045268,
+      "learning_rate": 4.787233903308208e-08,
+      "loss": 0.0266,
+      "step": 7214
+    },
+    {
+      "epoch": 2.8817891373801916,
+      "grad_norm": 0.9881205371437639,
+      "learning_rate": 4.723292384426203e-08,
+      "loss": 0.0295,
+      "step": 7216
+    },
+    {
+      "epoch": 2.88258785942492,
+      "grad_norm": 0.9124860856419786,
+      "learning_rate": 4.65977873788942e-08,
+      "loss": 0.0287,
+      "step": 7218
+    },
+    {
+      "epoch": 2.8833865814696487,
+      "grad_norm": 0.9969073107367281,
+      "learning_rate": 4.596693018567744e-08,
+      "loss": 0.0288,
+      "step": 7220
+    },
+    {
+      "epoch": 2.8841853035143767,
+      "grad_norm": 0.8841157317264496,
+      "learning_rate": 4.534035280961191e-08,
+      "loss": 0.0309,
+      "step": 7222
+    },
+    {
+      "epoch": 2.8849840255591053,
+      "grad_norm": 1.0078494252596897,
+      "learning_rate": 4.471805579200239e-08,
+      "loss": 0.0312,
+      "step": 7224
+    },
+    {
+      "epoch": 2.885782747603834,
+      "grad_norm": 0.9932395040774775,
+      "learning_rate": 4.41000396704544e-08,
+      "loss": 0.0283,
+      "step": 7226
+    },
+    {
+      "epoch": 2.8865814696485623,
+      "grad_norm": 0.9777222302672404,
+      "learning_rate": 4.3486304978875294e-08,
+      "loss": 0.0286,
+      "step": 7228
+    },
+    {
+      "epoch": 2.887380191693291,
+      "grad_norm": 0.9488138150582978,
+      "learning_rate": 4.287685224747373e-08,
+      "loss": 0.0287,
+      "step": 7230
+    },
+    {
+      "epoch": 2.8881789137380194,
+      "grad_norm": 0.9515551976834303,
+      "learning_rate": 4.227168200276077e-08,
+      "loss": 0.0301,
+      "step": 7232
+    },
+    {
+      "epoch": 2.8889776357827475,
+      "grad_norm": 1.0020004748739546,
+      "learning_rate": 4.167079476754432e-08,
+      "loss": 0.0303,
+      "step": 7234
+    },
+    {
+      "epoch": 2.889776357827476,
+      "grad_norm": 0.9940040887434322,
+      "learning_rate": 4.1074191060935794e-08,
+      "loss": 0.0314,
+      "step": 7236
+    },
+    {
+      "epoch": 2.8905750798722045,
+      "grad_norm": 0.9574123979978048,
+      "learning_rate": 4.048187139834403e-08,
+      "loss": 0.0266,
+      "step": 7238
+    },
+    {
+      "epoch": 2.891373801916933,
+      "grad_norm": 0.9872954533945656,
+      "learning_rate": 3.989383629147747e-08,
+      "loss": 0.0305,
+      "step": 7240
+    },
+    {
+      "epoch": 2.892172523961661,
+      "grad_norm": 1.093299647415339,
+      "learning_rate": 3.9310086248342536e-08,
+      "loss": 0.0309,
+      "step": 7242
+    },
+    {
+      "epoch": 2.8929712460063897,
+      "grad_norm": 0.9207625460483227,
+      "learning_rate": 3.873062177324472e-08,
+      "loss": 0.0293,
+      "step": 7244
+    },
+    {
+      "epoch": 2.893769968051118,
+      "grad_norm": 0.8613042189991472,
+      "learning_rate": 3.8155443366785786e-08,
+      "loss": 0.0278,
+      "step": 7246
+    },
+    {
+      "epoch": 2.8945686900958467,
+      "grad_norm": 0.9093281736899324,
+      "learning_rate": 3.758455152586715e-08,
+      "loss": 0.0273,
+      "step": 7248
+    },
+    {
+      "epoch": 2.8953674121405752,
+      "grad_norm": 0.8976668145886221,
+      "learning_rate": 3.7017946743683754e-08,
+      "loss": 0.0279,
+      "step": 7250
+    },
+    {
+      "epoch": 2.8961661341853038,
+      "grad_norm": 0.9438489705868893,
+      "learning_rate": 3.645562950973014e-08,
+      "loss": 0.0288,
+      "step": 7252
+    },
+    {
+      "epoch": 2.896964856230032,
+      "grad_norm": 0.9039303228109454,
+      "learning_rate": 3.589760030979439e-08,
+      "loss": 0.029,
+      "step": 7254
+    },
+    {
+      "epoch": 2.8977635782747604,
+      "grad_norm": 0.8665135816874094,
+      "learning_rate": 3.534385962596143e-08,
+      "loss": 0.0257,
+      "step": 7256
+    },
+    {
+      "epoch": 2.898562300319489,
+      "grad_norm": 0.9885112089994623,
+      "learning_rate": 3.479440793661082e-08,
+      "loss": 0.0287,
+      "step": 7258
+    },
+    {
+      "epoch": 2.899361022364217,
+      "grad_norm": 0.9313125684039982,
+      "learning_rate": 3.4249245716417303e-08,
+      "loss": 0.0311,
+      "step": 7260
+    },
+    {
+      "epoch": 2.9001597444089455,
+      "grad_norm": 1.0211354686014047,
+      "learning_rate": 3.370837343634914e-08,
+      "loss": 0.0281,
+      "step": 7262
+    },
+    {
+      "epoch": 2.900958466453674,
+      "grad_norm": 0.9261447629517585,
+      "learning_rate": 3.3171791563669785e-08,
+      "loss": 0.0285,
+      "step": 7264
+    },
+    {
+      "epoch": 2.9017571884984026,
+      "grad_norm": 0.9543015423065228,
+      "learning_rate": 3.263950056193455e-08,
+      "loss": 0.0285,
+      "step": 7266
+    },
+    {
+      "epoch": 2.902555910543131,
+      "grad_norm": 0.876741688352455,
+      "learning_rate": 3.211150089099224e-08,
+      "loss": 0.0261,
+      "step": 7268
+    },
+    {
+      "epoch": 2.9033546325878596,
+      "grad_norm": 0.9732009390231579,
+      "learning_rate": 3.1587793006985224e-08,
+      "loss": 0.0308,
+      "step": 7270
+    },
+    {
+      "epoch": 2.9041533546325877,
+      "grad_norm": 0.924537711179532,
+      "learning_rate": 3.10683773623488e-08,
+      "loss": 0.028,
+      "step": 7272
+    },
+    {
+      "epoch": 2.9049520766773163,
+      "grad_norm": 0.8328671231098053,
+      "learning_rate": 3.055325440580736e-08,
+      "loss": 0.0231,
+      "step": 7274
+    },
+    {
+      "epoch": 2.905750798722045,
+      "grad_norm": 0.8481571211750596,
+      "learning_rate": 3.004242458237994e-08,
+      "loss": 0.029,
+      "step": 7276
+    },
+    {
+      "epoch": 2.9065495207667733,
+      "grad_norm": 1.0264649887625952,
+      "learning_rate": 2.9535888333374064e-08,
+      "loss": 0.0316,
+      "step": 7278
+    },
+    {
+      "epoch": 2.9073482428115014,
+      "grad_norm": 0.9904821324548735,
+      "learning_rate": 2.9033646096390255e-08,
+      "loss": 0.0317,
+      "step": 7280
+    },
+    {
+      "epoch": 2.90814696485623,
+      "grad_norm": 0.9529343344181515,
+      "learning_rate": 2.853569830531755e-08,
+      "loss": 0.0294,
+      "step": 7282
+    },
+    {
+      "epoch": 2.9089456869009584,
+      "grad_norm": 0.8950285806155511,
+      "learning_rate": 2.8042045390336835e-08,
+      "loss": 0.027,
+      "step": 7284
+    },
+    {
+      "epoch": 2.909744408945687,
+      "grad_norm": 0.9915085372273561,
+      "learning_rate": 2.7552687777916976e-08,
+      "loss": 0.029,
+      "step": 7286
+    },
+    {
+      "epoch": 2.9105431309904155,
+      "grad_norm": 1.0843075888116502,
+      "learning_rate": 2.706762589081646e-08,
+      "loss": 0.0329,
+      "step": 7288
+    },
+    {
+      "epoch": 2.911341853035144,
+      "grad_norm": 1.0723208662350674,
+      "learning_rate": 2.6586860148084537e-08,
+      "loss": 0.0314,
+      "step": 7290
+    },
+    {
+      "epoch": 2.912140575079872,
+      "grad_norm": 0.9855410494694304,
+      "learning_rate": 2.6110390965055632e-08,
+      "loss": 0.0332,
+      "step": 7292
+    },
+    {
+      "epoch": 2.9129392971246006,
+      "grad_norm": 0.9456950838662169,
+      "learning_rate": 2.563821875335437e-08,
+      "loss": 0.0306,
+      "step": 7294
+    },
+    {
+      "epoch": 2.913738019169329,
+      "grad_norm": 0.9431752923514247,
+      "learning_rate": 2.517034392089446e-08,
+      "loss": 0.0242,
+      "step": 7296
+    },
+    {
+      "epoch": 2.9145367412140573,
+      "grad_norm": 0.9335986279046339,
+      "learning_rate": 2.4706766871874232e-08,
+      "loss": 0.029,
+      "step": 7298
+    },
+    {
+      "epoch": 2.915335463258786,
+      "grad_norm": 0.772170384658901,
+      "learning_rate": 2.4247488006781116e-08,
+      "loss": 0.0259,
+      "step": 7300
+    },
+    {
+      "epoch": 2.9161341853035143,
+      "grad_norm": 0.9779269003823648,
+      "learning_rate": 2.3792507722388835e-08,
+      "loss": 0.0301,
+      "step": 7302
+    },
+    {
+      "epoch": 2.916932907348243,
+      "grad_norm": 0.9611794936794353,
+      "learning_rate": 2.3341826411756863e-08,
+      "loss": 0.0305,
+      "step": 7304
+    },
+    {
+      "epoch": 2.9177316293929714,
+      "grad_norm": 0.8596768995810088,
+      "learning_rate": 2.2895444464232087e-08,
+      "loss": 0.0279,
+      "step": 7306
+    },
+    {
+      "epoch": 2.9185303514377,
+      "grad_norm": 1.050550890425119,
+      "learning_rate": 2.2453362265445477e-08,
+      "loss": 0.0266,
+      "step": 7308
+    },
+    {
+      "epoch": 2.919329073482428,
+      "grad_norm": 1.0172145059261652,
+      "learning_rate": 2.2015580197314868e-08,
+      "loss": 0.0295,
+      "step": 7310
+    },
+    {
+      "epoch": 2.9201277955271565,
+      "grad_norm": 0.8631988032830935,
+      "learning_rate": 2.158209863804217e-08,
+      "loss": 0.0258,
+      "step": 7312
+    },
+    {
+      "epoch": 2.920926517571885,
+      "grad_norm": 1.0172194609395517,
+      "learning_rate": 2.1152917962115606e-08,
+      "loss": 0.0311,
+      "step": 7314
+    },
+    {
+      "epoch": 2.9217252396166136,
+      "grad_norm": 0.8656946266189468,
+      "learning_rate": 2.0728038540305807e-08,
+      "loss": 0.025,
+      "step": 7316
+    },
+    {
+      "epoch": 2.9225239616613417,
+      "grad_norm": 1.2040154733832662,
+      "learning_rate": 2.030746073966916e-08,
+      "loss": 0.0299,
+      "step": 7318
+    },
+    {
+      "epoch": 2.92332268370607,
+      "grad_norm": 0.981108582856345,
+      "learning_rate": 1.9891184923544472e-08,
+      "loss": 0.0312,
+      "step": 7320
+    },
+    {
+      "epoch": 2.9241214057507987,
+      "grad_norm": 1.0050547681748903,
+      "learning_rate": 1.9479211451555735e-08,
+      "loss": 0.0306,
+      "step": 7322
+    },
+    {
+      "epoch": 2.9249201277955272,
+      "grad_norm": 0.9723436269119015,
+      "learning_rate": 1.9071540679608815e-08,
+      "loss": 0.0274,
+      "step": 7324
+    },
+    {
+      "epoch": 2.9257188498402558,
+      "grad_norm": 0.890152909346874,
+      "learning_rate": 1.8668172959891985e-08,
+      "loss": 0.0286,
+      "step": 7326
+    },
+    {
+      "epoch": 2.9265175718849843,
+      "grad_norm": 1.0167534719731643,
+      "learning_rate": 1.826910864087761e-08,
+      "loss": 0.0293,
+      "step": 7328
+    },
+    {
+      "epoch": 2.9273162939297124,
+      "grad_norm": 0.9875484380951924,
+      "learning_rate": 1.7874348067319912e-08,
+      "loss": 0.031,
+      "step": 7330
+    },
+    {
+      "epoch": 2.928115015974441,
+      "grad_norm": 1.1309464473632505,
+      "learning_rate": 1.7483891580253877e-08,
+      "loss": 0.0304,
+      "step": 7332
+    },
+    {
+      "epoch": 2.9289137380191694,
+      "grad_norm": 0.8279929400769935,
+      "learning_rate": 1.7097739516997447e-08,
+      "loss": 0.0268,
+      "step": 7334
+    },
+    {
+      "epoch": 2.9297124600638975,
+      "grad_norm": 0.9993434381726024,
+      "learning_rate": 1.6715892211150442e-08,
+      "loss": 0.0286,
+      "step": 7336
+    },
+    {
+      "epoch": 2.930511182108626,
+      "grad_norm": 1.007539806056888,
+      "learning_rate": 1.6338349992591763e-08,
+      "loss": 0.0294,
+      "step": 7338
+    },
+    {
+      "epoch": 2.9313099041533546,
+      "grad_norm": 1.0452749714129563,
+      "learning_rate": 1.5965113187482174e-08,
+      "loss": 0.0328,
+      "step": 7340
+    },
+    {
+      "epoch": 2.932108626198083,
+      "grad_norm": 0.8474586458371822,
+      "learning_rate": 1.5596182118264303e-08,
+      "loss": 0.0256,
+      "step": 7342
+    },
+    {
+      "epoch": 2.9329073482428116,
+      "grad_norm": 0.9287019035012101,
+      "learning_rate": 1.5231557103658755e-08,
+      "loss": 0.0286,
+      "step": 7344
+    },
+    {
+      "epoch": 2.93370607028754,
+      "grad_norm": 1.0230975340136683,
+      "learning_rate": 1.4871238458667447e-08,
+      "loss": 0.0308,
+      "step": 7346
+    },
+    {
+      "epoch": 2.9345047923322682,
+      "grad_norm": 0.9679221553199847,
+      "learning_rate": 1.4515226494571376e-08,
+      "loss": 0.0279,
+      "step": 7348
+    },
+    {
+      "epoch": 2.9353035143769968,
+      "grad_norm": 1.0129002341137234,
+      "learning_rate": 1.41635215189323e-08,
+      "loss": 0.0243,
+      "step": 7350
+    },
+    {
+      "epoch": 2.9361022364217253,
+      "grad_norm": 1.159004848542131,
+      "learning_rate": 1.3816123835588835e-08,
+      "loss": 0.0309,
+      "step": 7352
+    },
+    {
+      "epoch": 2.936900958466454,
+      "grad_norm": 1.0452986083737144,
+      "learning_rate": 1.3473033744660358e-08,
+      "loss": 0.0297,
+      "step": 7354
+    },
+    {
+      "epoch": 2.937699680511182,
+      "grad_norm": 0.944187394021172,
+      "learning_rate": 1.3134251542544774e-08,
+      "loss": 0.0255,
+      "step": 7356
+    },
+    {
+      "epoch": 2.9384984025559104,
+      "grad_norm": 1.1423229379322075,
+      "learning_rate": 1.2799777521916856e-08,
+      "loss": 0.0306,
+      "step": 7358
+    },
+    {
+      "epoch": 2.939297124600639,
+      "grad_norm": 0.9042757983900201,
+      "learning_rate": 1.2469611971731576e-08,
+      "loss": 0.028,
+      "step": 7360
+    },
+    {
+      "epoch": 2.9400958466453675,
+      "grad_norm": 1.219554574679699,
+      "learning_rate": 1.2143755177220774e-08,
+      "loss": 0.0292,
+      "step": 7362
+    },
+    {
+      "epoch": 2.940894568690096,
+      "grad_norm": 1.0160171362099817,
+      "learning_rate": 1.1822207419893151e-08,
+      "loss": 0.0316,
+      "step": 7364
+    },
+    {
+      "epoch": 2.9416932907348246,
+      "grad_norm": 0.9410768507733369,
+      "learning_rate": 1.1504968977536502e-08,
+      "loss": 0.0311,
+      "step": 7366
+    },
+    {
+      "epoch": 2.9424920127795526,
+      "grad_norm": 0.9541185215999987,
+      "learning_rate": 1.1192040124214931e-08,
+      "loss": 0.0251,
+      "step": 7368
+    },
+    {
+      "epoch": 2.943290734824281,
+      "grad_norm": 1.1401434841727258,
+      "learning_rate": 1.0883421130268857e-08,
+      "loss": 0.0389,
+      "step": 7370
+    },
+    {
+      "epoch": 2.9440894568690097,
+      "grad_norm": 0.9102452503551742,
+      "learning_rate": 1.0579112262316116e-08,
+      "loss": 0.0299,
+      "step": 7372
+    },
+    {
+      "epoch": 2.9448881789137378,
+      "grad_norm": 0.8297271919791628,
+      "learning_rate": 1.027911378325086e-08,
+      "loss": 0.0266,
+      "step": 7374
+    },
+    {
+      "epoch": 2.9456869009584663,
+      "grad_norm": 0.977055359117076,
+      "learning_rate": 9.983425952243552e-09,
+      "loss": 0.0275,
+      "step": 7376
+    },
+    {
+      "epoch": 2.946485623003195,
+      "grad_norm": 0.8856220307196366,
+      "learning_rate": 9.692049024740968e-09,
+      "loss": 0.0273,
+      "step": 7378
+    },
+    {
+      "epoch": 2.9472843450479234,
+      "grad_norm": 1.0185133421121422,
+      "learning_rate": 9.404983252464528e-09,
+      "loss": 0.0308,
+      "step": 7380
+    },
+    {
+      "epoch": 2.948083067092652,
+      "grad_norm": 0.9226722623416709,
+      "learning_rate": 9.12222888341252e-09,
+      "loss": 0.0278,
+      "step": 7382
+    },
+    {
+      "epoch": 2.9488817891373804,
+      "grad_norm": 0.8681408122208755,
+      "learning_rate": 8.84378616185788e-09,
+      "loss": 0.0289,
+      "step": 7384
+    },
+    {
+      "epoch": 2.9496805111821085,
+      "grad_norm": 0.9820066250150342,
+      "learning_rate": 8.569655328349302e-09,
+      "loss": 0.0306,
+      "step": 7386
+    },
+    {
+      "epoch": 2.950479233226837,
+      "grad_norm": 0.9224388253014746,
+      "learning_rate": 8.299836619709011e-09,
+      "loss": 0.0252,
+      "step": 7388
+    },
+    {
+      "epoch": 2.9512779552715656,
+      "grad_norm": 1.078281742555588,
+      "learning_rate": 8.034330269034995e-09,
+      "loss": 0.0339,
+      "step": 7390
+    },
+    {
+      "epoch": 2.952076677316294,
+      "grad_norm": 1.1121638039606885,
+      "learning_rate": 7.773136505700995e-09,
+      "loss": 0.0315,
+      "step": 7392
+    },
+    {
+      "epoch": 2.952875399361022,
+      "grad_norm": 0.8527091048044436,
+      "learning_rate": 7.516255555352069e-09,
+      "loss": 0.0277,
+      "step": 7394
+    },
+    {
+      "epoch": 2.9536741214057507,
+      "grad_norm": 0.9156141320301706,
+      "learning_rate": 7.2636876399107e-09,
+      "loss": 0.0269,
+      "step": 7396
+    },
+    {
+      "epoch": 2.9544728434504792,
+      "grad_norm": 1.3378705563366167,
+      "learning_rate": 7.015432977570679e-09,
+      "loss": 0.0277,
+      "step": 7398
+    },
+    {
+      "epoch": 2.9552715654952078,
+      "grad_norm": 0.9035834422975698,
+      "learning_rate": 6.7714917828004545e-09,
+      "loss": 0.0281,
+      "step": 7400
+    },
+    {
+      "epoch": 2.9560702875399363,
+      "grad_norm": 1.0788469833239718,
+      "learning_rate": 6.531864266343113e-09,
+      "loss": 0.0311,
+      "step": 7402
+    },
+    {
+      "epoch": 2.956869009584665,
+      "grad_norm": 1.0030626593801413,
+      "learning_rate": 6.296550635213616e-09,
+      "loss": 0.0282,
+      "step": 7404
+    },
+    {
+      "epoch": 2.957667731629393,
+      "grad_norm": 0.9736720908524453,
+      "learning_rate": 6.0655510927010165e-09,
+      "loss": 0.0309,
+      "step": 7406
+    },
+    {
+      "epoch": 2.9584664536741214,
+      "grad_norm": 0.976953889137446,
+      "learning_rate": 5.838865838366792e-09,
+      "loss": 0.03,
+      "step": 7408
+    },
+    {
+      "epoch": 2.95926517571885,
+      "grad_norm": 0.96316953344491,
+      "learning_rate": 5.616495068046513e-09,
+      "loss": 0.0306,
+      "step": 7410
+    },
+    {
+      "epoch": 2.960063897763578,
+      "grad_norm": 0.8882325546800076,
+      "learning_rate": 5.398438973845954e-09,
+      "loss": 0.0328,
+      "step": 7412
+    },
+    {
+      "epoch": 2.9608626198083066,
+      "grad_norm": 1.0434465347465718,
+      "learning_rate": 5.184697744146094e-09,
+      "loss": 0.0296,
+      "step": 7414
+    },
+    {
+      "epoch": 2.961661341853035,
+      "grad_norm": 1.1676463120664913,
+      "learning_rate": 4.975271563599227e-09,
+      "loss": 0.0342,
+      "step": 7416
+    },
+    {
+      "epoch": 2.9624600638977636,
+      "grad_norm": 1.125598033675333,
+      "learning_rate": 4.770160613129515e-09,
+      "loss": 0.0305,
+      "step": 7418
+    },
+    {
+      "epoch": 2.963258785942492,
+      "grad_norm": 0.7862908310858415,
+      "learning_rate": 4.569365069933551e-09,
+      "loss": 0.0256,
+      "step": 7420
+    },
+    {
+      "epoch": 2.9640575079872207,
+      "grad_norm": 0.9443569000718535,
+      "learning_rate": 4.372885107479796e-09,
+      "loss": 0.0306,
+      "step": 7422
+    },
+    {
+      "epoch": 2.9648562300319488,
+      "grad_norm": 0.800637449098191,
+      "learning_rate": 4.180720895508028e-09,
+      "loss": 0.0293,
+      "step": 7424
+    },
+    {
+      "epoch": 2.9656549520766773,
+      "grad_norm": 0.9430328576301563,
+      "learning_rate": 3.992872600030451e-09,
+      "loss": 0.0271,
+      "step": 7426
+    },
+    {
+      "epoch": 2.966453674121406,
+      "grad_norm": 0.9058280169647789,
+      "learning_rate": 3.809340383330584e-09,
+      "loss": 0.0265,
+      "step": 7428
+    },
+    {
+      "epoch": 2.9672523961661343,
+      "grad_norm": 0.976846038410627,
+      "learning_rate": 3.630124403961599e-09,
+      "loss": 0.0308,
+      "step": 7430
+    },
+    {
+      "epoch": 2.9680511182108624,
+      "grad_norm": 0.9428661317946987,
+      "learning_rate": 3.4552248167507576e-09,
+      "loss": 0.0271,
+      "step": 7432
+    },
+    {
+      "epoch": 2.968849840255591,
+      "grad_norm": 0.9860305014063514,
+      "learning_rate": 3.284641772793862e-09,
+      "loss": 0.0299,
+      "step": 7434
+    },
+    {
+      "epoch": 2.9696485623003195,
+      "grad_norm": 0.8472915574856456,
+      "learning_rate": 3.118375419458586e-09,
+      "loss": 0.0231,
+      "step": 7436
+    },
+    {
+      "epoch": 2.970447284345048,
+      "grad_norm": 0.9416785407994653,
+      "learning_rate": 2.956425900383919e-09,
+      "loss": 0.0333,
+      "step": 7438
+    },
+    {
+      "epoch": 2.9712460063897765,
+      "grad_norm": 0.7967253694302693,
+      "learning_rate": 2.798793355478502e-09,
+      "loss": 0.03,
+      "step": 7440
+    },
+    {
+      "epoch": 2.972044728434505,
+      "grad_norm": 0.9101321750296127,
+      "learning_rate": 2.6454779209217353e-09,
+      "loss": 0.0288,
+      "step": 7442
+    },
+    {
+      "epoch": 2.972843450479233,
+      "grad_norm": 0.960448163452539,
+      "learning_rate": 2.496479729164891e-09,
+      "loss": 0.0307,
+      "step": 7444
+    },
+    {
+      "epoch": 2.9736421725239617,
+      "grad_norm": 0.9953358583136821,
+      "learning_rate": 2.3517989089272274e-09,
+      "loss": 0.027,
+      "step": 7446
+    },
+    {
+      "epoch": 2.97444089456869,
+      "grad_norm": 0.9262672112649574,
+      "learning_rate": 2.2114355851993175e-09,
+      "loss": 0.0252,
+      "step": 7448
+    },
+    {
+      "epoch": 2.9752396166134183,
+      "grad_norm": 1.0161598052132237,
+      "learning_rate": 2.0753898792424954e-09,
+      "loss": 0.03,
+      "step": 7450
+    },
+    {
+      "epoch": 2.976038338658147,
+      "grad_norm": 0.8238758679662858,
+      "learning_rate": 1.943661908586636e-09,
+      "loss": 0.0226,
+      "step": 7452
+    },
+    {
+      "epoch": 2.9768370607028753,
+      "grad_norm": 1.1294195598874084,
+      "learning_rate": 1.81625178703293e-09,
+      "loss": 0.0314,
+      "step": 7454
+    },
+    {
+      "epoch": 2.977635782747604,
+      "grad_norm": 0.7948759771645315,
+      "learning_rate": 1.6931596246516636e-09,
+      "loss": 0.0237,
+      "step": 7456
+    },
+    {
+      "epoch": 2.9784345047923324,
+      "grad_norm": 0.9603638334498134,
+      "learning_rate": 1.5743855277822185e-09,
+      "loss": 0.0271,
+      "step": 7458
+    },
+    {
+      "epoch": 2.979233226837061,
+      "grad_norm": 1.004902038270927,
+      "learning_rate": 1.4599295990352924e-09,
+      "loss": 0.0274,
+      "step": 7460
+    },
+    {
+      "epoch": 2.980031948881789,
+      "grad_norm": 1.3727183415621365,
+      "learning_rate": 1.3497919372890135e-09,
+      "loss": 0.0299,
+      "step": 7462
+    },
+    {
+      "epoch": 2.9808306709265175,
+      "grad_norm": 1.0113889053741916,
+      "learning_rate": 1.24397263769227e-09,
+      "loss": 0.0334,
+      "step": 7464
+    },
+    {
+      "epoch": 2.981629392971246,
+      "grad_norm": 1.058301608314339,
+      "learning_rate": 1.1424717916630468e-09,
+      "loss": 0.0314,
+      "step": 7466
+    },
+    {
+      "epoch": 2.9824281150159746,
+      "grad_norm": 1.0685945832080852,
+      "learning_rate": 1.0452894868884235e-09,
+      "loss": 0.0287,
+      "step": 7468
+    },
+    {
+      "epoch": 2.9832268370607027,
+      "grad_norm": 0.9439988444316789,
+      "learning_rate": 9.52425807324575e-10,
+      "loss": 0.0279,
+      "step": 7470
+    },
+    {
+      "epoch": 2.984025559105431,
+      "grad_norm": 1.0539650853101212,
+      "learning_rate": 8.638808331973281e-10,
+      "loss": 0.0306,
+      "step": 7472
+    },
+    {
+      "epoch": 2.9848242811501597,
+      "grad_norm": 0.8544922347586041,
+      "learning_rate": 7.796546410004934e-10,
+      "loss": 0.0302,
+      "step": 7474
+    },
+    {
+      "epoch": 2.9856230031948883,
+      "grad_norm": 0.9303223740142429,
+      "learning_rate": 6.997473034986435e-10,
+      "loss": 0.0281,
+      "step": 7476
+    },
+    {
+      "epoch": 2.986421725239617,
+      "grad_norm": 0.9400919377169601,
+      "learning_rate": 6.241588897232253e-10,
+      "loss": 0.0276,
+      "step": 7478
+    },
+    {
+      "epoch": 2.987220447284345,
+      "grad_norm": 0.9420254495627837,
+      "learning_rate": 5.528894649758921e-10,
+      "loss": 0.0252,
+      "step": 7480
+    },
+    {
+      "epoch": 2.9880191693290734,
+      "grad_norm": 0.8844462225772538,
+      "learning_rate": 4.859390908268369e-10,
+      "loss": 0.0271,
+      "step": 7482
+    },
+    {
+      "epoch": 2.988817891373802,
+      "grad_norm": 0.9000038358376025,
+      "learning_rate": 4.2330782511423865e-10,
+      "loss": 0.0291,
+      "step": 7484
+    },
+    {
+      "epoch": 2.9896166134185305,
+      "grad_norm": 1.0116302989209616,
+      "learning_rate": 3.649957219464817e-10,
+      "loss": 0.0312,
+      "step": 7486
+    },
+    {
+      "epoch": 2.9904153354632586,
+      "grad_norm": 1.0168619061466087,
+      "learning_rate": 3.1100283169938074e-10,
+      "loss": 0.0257,
+      "step": 7488
+    },
+    {
+      "epoch": 2.991214057507987,
+      "grad_norm": 0.9317368624862982,
+      "learning_rate": 2.613292010172908e-10,
+      "loss": 0.0326,
+      "step": 7490
+    },
+    {
+      "epoch": 2.9920127795527156,
+      "grad_norm": 1.0576362124675651,
+      "learning_rate": 2.1597487281366236e-10,
+      "loss": 0.0279,
+      "step": 7492
+    },
+    {
+      "epoch": 2.992811501597444,
+      "grad_norm": 0.9509971386881086,
+      "learning_rate": 1.7493988627104164e-10,
+      "loss": 0.0318,
+      "step": 7494
+    },
+    {
+      "epoch": 2.9936102236421727,
+      "grad_norm": 0.8370706486653864,
+      "learning_rate": 1.3822427683884975e-10,
+      "loss": 0.0273,
+      "step": 7496
+    },
+    {
+      "epoch": 2.994408945686901,
+      "grad_norm": 1.0612898069636711,
+      "learning_rate": 1.0582807623671365e-10,
+      "loss": 0.0352,
+      "step": 7498
+    },
+    {
+      "epoch": 2.9952076677316293,
+      "grad_norm": 0.805238580681461,
+      "learning_rate": 7.775131245169044e-11,
+      "loss": 0.0258,
+      "step": 7500
+    },
+    {
+      "epoch": 2.9952076677316293,
+      "eval_loss": 0.18183551728725433,
+      "eval_runtime": 419.4635,
+      "eval_samples_per_second": 42.452,
+      "eval_steps_per_second": 5.307,
+      "step": 7500
+    },
+    {
+      "epoch": 2.996006389776358,
+      "grad_norm": 0.8616048488279973,
+      "learning_rate": 5.399400973882252e-11,
+      "loss": 0.0255,
+      "step": 7502
+    },
+    {
+      "epoch": 2.9968051118210863,
+      "grad_norm": 0.8828142582638174,
+      "learning_rate": 3.4556188622802964e-11,
+      "loss": 0.0298,
+      "step": 7504
+    },
+    {
+      "epoch": 2.997603833865815,
+      "grad_norm": 1.0530041672485366,
+      "learning_rate": 1.9437865895755027e-11,
+      "loss": 0.0281,
+      "step": 7506
+    },
+    {
+      "epoch": 2.998402555910543,
+      "grad_norm": 0.8820484624604644,
+      "learning_rate": 8.639054618897468e-12,
+      "loss": 0.0258,
+      "step": 7508
+    },
+    {
+      "epoch": 2.9992012779552715,
+      "grad_norm": 1.0521391566214802,
+      "learning_rate": 2.1597641214343714e-12,
+      "loss": 0.028,
+      "step": 7510
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.0637471844248951,
+      "learning_rate": 0.0,
+      "loss": 0.0258,
+      "step": 7512
+    },
+    {
+      "epoch": 3.0,
+      "step": 7512,
+      "total_flos": 730578500321280.0,
+      "train_loss": 0.11755168595261023,
+      "train_runtime": 63778.2421,
+      "train_samples_per_second": 7.538,
+      "train_steps_per_second": 0.118
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 7512,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 730578500321280.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}