diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4750 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 8.330377362483862,
+  "eval_steps": 1000000,
+  "global_step": 393602,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 10.2978,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 9.7725,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5e-06,
+      "loss": 9.3054,
+      "step": 1500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 8.8548,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.5e-06,
+      "loss": 8.3696,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 3e-06,
+      "loss": 7.8962,
+      "step": 3000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3.5e-06,
+      "loss": 7.5047,
+      "step": 3500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 7.2423,
+      "step": 4000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.5e-06,
+      "loss": 7.076,
+      "step": 4500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-06,
+      "loss": 6.9674,
+      "step": 5000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 6.8806,
+      "step": 5500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 6e-06,
+      "loss": 6.8081,
+      "step": 6000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 6.7556,
+      "step": 6500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 7e-06,
+      "loss": 6.7068,
+      "step": 7000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 6.6621,
+      "step": 7500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 6.6238,
+      "step": 8000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 8.5e-06,
+      "loss": 6.5916,
+      "step": 8500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 9e-06,
+      "loss": 6.5633,
+      "step": 9000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.5e-06,
+      "loss": 6.5335,
+      "step": 9500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1e-05,
+      "loss": 6.5168,
+      "step": 10000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.998939532121573e-06,
+      "loss": 6.4904,
+      "step": 10500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.997879064243145e-06,
+      "loss": 6.4703,
+      "step": 11000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 9.996818596364717e-06,
+      "loss": 6.4553,
+      "step": 11500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.995758128486289e-06,
+      "loss": 6.4361,
+      "step": 12000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 9.994697660607861e-06,
+      "loss": 6.4251,
+      "step": 12500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.993637192729433e-06,
+      "loss": 6.4036,
+      "step": 13000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.992576724851005e-06,
+      "loss": 6.3981,
+      "step": 13500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.991516256972577e-06,
+      "loss": 6.3899,
+      "step": 14000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.99045578909415e-06,
+      "loss": 6.3779,
+      "step": 14500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.989395321215721e-06,
+      "loss": 6.3661,
+      "step": 15000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.988334853337293e-06,
+      "loss": 6.3581,
+      "step": 15500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.987274385458866e-06,
+      "loss": 6.3458,
+      "step": 16000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.986213917580438e-06,
+      "loss": 6.3374,
+      "step": 16500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.98515344970201e-06,
+      "loss": 6.333,
+      "step": 17000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.984092981823582e-06,
+      "loss": 6.3215,
+      "step": 17500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.983032513945154e-06,
+      "loss": 6.315,
+      "step": 18000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.981972046066726e-06,
+      "loss": 6.3127,
+      "step": 18500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.980911578188298e-06,
+      "loss": 6.3007,
+      "step": 19000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.97985111030987e-06,
+      "loss": 6.2972,
+      "step": 19500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.978790642431442e-06,
+      "loss": 6.2885,
+      "step": 20000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.977730174553012e-06,
+      "loss": 6.2814,
+      "step": 20500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.976669706674586e-06,
+      "loss": 6.2801,
+      "step": 21000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.975609238796158e-06,
+      "loss": 6.2696,
+      "step": 21500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.97454877091773e-06,
+      "loss": 6.2687,
+      "step": 22000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.973488303039302e-06,
+      "loss": 6.2638,
+      "step": 22500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.972427835160874e-06,
+      "loss": 6.2573,
+      "step": 23000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.971367367282446e-06,
+      "loss": 6.254,
+      "step": 23500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.970306899404017e-06,
+      "loss": 6.2491,
+      "step": 24000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.96924643152559e-06,
+      "loss": 6.2393,
+      "step": 24500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.968185963647163e-06,
+      "loss": 6.2382,
+      "step": 25000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.967125495768733e-06,
+      "loss": 6.2407,
+      "step": 25500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.966065027890307e-06,
+      "loss": 6.2316,
+      "step": 26000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.965004560011879e-06,
+      "loss": 6.2249,
+      "step": 26500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.963944092133449e-06,
+      "loss": 6.222,
+      "step": 27000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.962883624255023e-06,
+      "loss": 6.2207,
+      "step": 27500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.961823156376595e-06,
+      "loss": 6.2161,
+      "step": 28000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.960762688498167e-06,
+      "loss": 6.2149,
+      "step": 28500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.959702220619737e-06,
+      "loss": 6.2107,
+      "step": 29000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.958641752741311e-06,
+      "loss": 6.2089,
+      "step": 29500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.957581284862883e-06,
+      "loss": 6.2017,
+      "step": 30000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.956520816984453e-06,
+      "loss": 6.1967,
+      "step": 30500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.955460349106027e-06,
+      "loss": 6.1991,
+      "step": 31000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.9543998812276e-06,
+      "loss": 6.191,
+      "step": 31500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.95333941334917e-06,
+      "loss": 6.1908,
+      "step": 32000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.952278945470742e-06,
+      "loss": 6.1873,
+      "step": 32500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.951218477592315e-06,
+      "loss": 6.1829,
+      "step": 33000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.950158009713886e-06,
+      "loss": 6.1761,
+      "step": 33500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.949097541835458e-06,
+      "loss": 6.1754,
+      "step": 34000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.948037073957032e-06,
+      "loss": 6.1774,
+      "step": 34500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.946976606078602e-06,
+      "loss": 6.1757,
+      "step": 35000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.945916138200174e-06,
+      "loss": 6.1684,
+      "step": 35500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.944855670321746e-06,
+      "loss": 6.1678,
+      "step": 36000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.94379520244332e-06,
+      "loss": 6.167,
+      "step": 36500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.94273473456489e-06,
+      "loss": 6.1603,
+      "step": 37000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.941674266686462e-06,
+      "loss": 6.1562,
+      "step": 37500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.940613798808036e-06,
+      "loss": 6.1562,
+      "step": 38000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.939553330929606e-06,
+      "loss": 6.156,
+      "step": 38500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.938492863051178e-06,
+      "loss": 6.1525,
+      "step": 39000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.937432395172752e-06,
+      "loss": 6.151,
+      "step": 39500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.936371927294323e-06,
+      "loss": 6.1504,
+      "step": 40000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.935311459415895e-06,
+      "loss": 6.1462,
+      "step": 40500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.934250991537467e-06,
+      "loss": 6.1416,
+      "step": 41000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 9.933190523659039e-06,
+      "loss": 6.1474,
+      "step": 41500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 9.93213005578061e-06,
+      "loss": 6.1385,
+      "step": 42000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.931069587902183e-06,
+      "loss": 6.1392,
+      "step": 42500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.930009120023757e-06,
+      "loss": 6.1357,
+      "step": 43000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 9.928948652145327e-06,
+      "loss": 6.1363,
+      "step": 43500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 9.927888184266899e-06,
+      "loss": 6.1341,
+      "step": 44000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 9.926827716388471e-06,
+      "loss": 6.1311,
+      "step": 44500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.925767248510043e-06,
+      "loss": 6.1296,
+      "step": 45000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.924706780631615e-06,
+      "loss": 6.1284,
+      "step": 45500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.923646312753187e-06,
+      "loss": 6.1253,
+      "step": 46000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.92258584487476e-06,
+      "loss": 6.1242,
+      "step": 46500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.921525376996331e-06,
+      "loss": 6.1189,
+      "step": 47000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 9.920464909117903e-06,
+      "loss": 6.1176,
+      "step": 47500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.919404441239475e-06,
+      "loss": 6.1162,
+      "step": 48000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.918343973361048e-06,
+      "loss": 6.1126,
+      "step": 48500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 9.91728350548262e-06,
+      "loss": 6.1149,
+      "step": 49000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.916223037604192e-06,
+      "loss": 6.1067,
+      "step": 49500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 9.915162569725764e-06,
+      "loss": 6.1088,
+      "step": 50000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.914102101847336e-06,
+      "loss": 6.1069,
+      "step": 50500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.913041633968908e-06,
+      "loss": 6.1065,
+      "step": 51000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 9.91198116609048e-06,
+      "loss": 6.1041,
+      "step": 51500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.910920698212052e-06,
+      "loss": 6.1033,
+      "step": 52000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 9.909860230333624e-06,
+      "loss": 6.097,
+      "step": 52500
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.908799762455196e-06,
+      "loss": 6.0966,
+      "step": 53000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.907739294576768e-06,
+      "loss": 6.0961,
+      "step": 53500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 9.90667882669834e-06,
+      "loss": 6.0952,
+      "step": 54000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.905618358819912e-06,
+      "loss": 6.0898,
+      "step": 54500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 9.904557890941484e-06,
+      "loss": 6.0857,
+      "step": 55000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 9.903497423063056e-06,
+      "loss": 6.0871,
+      "step": 55500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 9.902436955184628e-06,
+      "loss": 6.0812,
+      "step": 56000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.9013764873062e-06,
+      "loss": 6.0773,
+      "step": 56500
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 9.900316019427772e-06,
+      "loss": 6.0753,
+      "step": 57000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.899255551549345e-06,
+      "loss": 6.0707,
+      "step": 57500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.898195083670917e-06,
+      "loss": 6.0688,
+      "step": 58000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 9.897134615792489e-06,
+      "loss": 6.0624,
+      "step": 58500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.89607414791406e-06,
+      "loss": 6.0634,
+      "step": 59000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 9.895013680035633e-06,
+      "loss": 6.0618,
+      "step": 59500
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.893953212157205e-06,
+      "loss": 6.0569,
+      "step": 60000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.892892744278777e-06,
+      "loss": 6.0542,
+      "step": 60500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 9.891832276400349e-06,
+      "loss": 6.054,
+      "step": 61000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 9.890771808521921e-06,
+      "loss": 6.0444,
+      "step": 61500
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 9.889711340643493e-06,
+      "loss": 6.0469,
+      "step": 62000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 9.888650872765065e-06,
+      "loss": 6.0368,
+      "step": 62500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9.887590404886637e-06,
+      "loss": 6.0409,
+      "step": 63000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 9.88652993700821e-06,
+      "loss": 6.0348,
+      "step": 63500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 9.88546946912978e-06,
+      "loss": 6.029,
+      "step": 64000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 9.884409001251353e-06,
+      "loss": 6.0302,
+      "step": 64500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 9.883348533372925e-06,
+      "loss": 6.0234,
+      "step": 65000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 9.882288065494497e-06,
+      "loss": 6.0239,
+      "step": 65500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 9.88122759761607e-06,
+      "loss": 6.0268,
+      "step": 66000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 9.880167129737642e-06,
+      "loss": 6.0203,
+      "step": 66500
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 9.879106661859214e-06,
+      "loss": 6.0194,
+      "step": 67000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 9.878046193980784e-06,
+      "loss": 6.0128,
+      "step": 67500
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 9.876985726102358e-06,
+      "loss": 6.0186,
+      "step": 68000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.87592525822393e-06,
+      "loss": 6.0154,
+      "step": 68500
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 9.8748647903455e-06,
+      "loss": 6.0099,
+      "step": 69000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.873804322467074e-06,
+      "loss": 6.0056,
+      "step": 69500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.872743854588646e-06,
+      "loss": 6.0067,
+      "step": 70000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 9.871683386710216e-06,
+      "loss": 6.001,
+      "step": 70500
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.87062291883179e-06,
+      "loss": 5.9986,
+      "step": 71000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 9.869562450953362e-06,
+      "loss": 5.997,
+      "step": 71500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.868501983074934e-06,
+      "loss": 5.994,
+      "step": 72000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.867441515196505e-06,
+      "loss": 5.9933,
+      "step": 72500
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.866381047318078e-06,
+      "loss": 5.9978,
+      "step": 73000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.86532057943965e-06,
+      "loss": 5.9953,
+      "step": 73500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.86426011156122e-06,
+      "loss": 5.988,
+      "step": 74000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.863199643682794e-06,
+      "loss": 5.9881,
+      "step": 74500
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.862139175804367e-06,
+      "loss": 5.9847,
+      "step": 75000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.861078707925937e-06,
+      "loss": 5.9851,
+      "step": 75500
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.860018240047509e-06,
+      "loss": 5.981,
+      "step": 76000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.858957772169083e-06,
+      "loss": 5.9789,
+      "step": 76500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.857897304290653e-06,
+      "loss": 5.9752,
+      "step": 77000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.856836836412225e-06,
+      "loss": 5.9759,
+      "step": 77500
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.855776368533799e-06,
+      "loss": 5.9721,
+      "step": 78000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 9.85471590065537e-06,
+      "loss": 5.9684,
+      "step": 78500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.853655432776941e-06,
+      "loss": 5.9667,
+      "step": 79000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.852594964898515e-06,
+      "loss": 5.9664,
+      "step": 79500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 9.851534497020087e-06,
+      "loss": 5.9612,
+      "step": 80000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 9.850474029141657e-06,
+      "loss": 5.9578,
+      "step": 80500
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 9.84941356126323e-06,
+      "loss": 5.9617,
+      "step": 81000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.848353093384803e-06,
+      "loss": 5.9555,
+      "step": 81500
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 9.847292625506374e-06,
+      "loss": 5.9548,
+      "step": 82000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.846232157627946e-06,
+      "loss": 5.9499,
+      "step": 82500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 9.84517168974952e-06,
+      "loss": 5.9488,
+      "step": 83000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 9.84411122187109e-06,
+      "loss": 5.9505,
+      "step": 83500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 9.843050753992662e-06,
+      "loss": 5.9429,
+      "step": 84000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 9.841990286114234e-06,
+      "loss": 5.941,
+      "step": 84500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 9.840929818235806e-06,
+      "loss": 5.9386,
+      "step": 85000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 9.839869350357378e-06,
+      "loss": 5.9369,
+      "step": 85500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 9.83880888247895e-06,
+      "loss": 5.9384,
+      "step": 86000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.837748414600524e-06,
+      "loss": 5.9346,
+      "step": 86500
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 9.836687946722094e-06,
+      "loss": 5.9332,
+      "step": 87000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.835627478843666e-06,
+      "loss": 5.9326,
+      "step": 87500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 9.834567010965238e-06,
+      "loss": 5.9322,
+      "step": 88000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.83350654308681e-06,
+      "loss": 5.9284,
+      "step": 88500
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.832446075208382e-06,
+      "loss": 5.9273,
+      "step": 89000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 9.831385607329954e-06,
+      "loss": 5.9193,
+      "step": 89500
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 9.830325139451527e-06,
+      "loss": 5.9252,
+      "step": 90000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.829264671573099e-06,
+      "loss": 5.9202,
+      "step": 90500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.82820420369467e-06,
+      "loss": 5.918,
+      "step": 91000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 9.827143735816243e-06,
+      "loss": 5.9166,
+      "step": 91500
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 9.826083267937815e-06,
+      "loss": 5.914,
+      "step": 92000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 9.825022800059387e-06,
+      "loss": 5.9153,
+      "step": 92500
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.823962332180959e-06,
+      "loss": 5.9117,
+      "step": 93000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 9.822901864302531e-06,
+      "loss": 5.9068,
+      "step": 93500
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 9.821841396424103e-06,
+      "loss": 5.9057,
+      "step": 94000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.820780928545675e-06,
+      "loss": 5.9074,
+      "step": 94500
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 9.819720460667247e-06,
+      "loss": 5.9047,
+      "step": 95000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.81865999278882e-06,
+      "loss": 5.8999,
+      "step": 95500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.817599524910391e-06,
+      "loss": 5.8976,
+      "step": 96000
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 9.816539057031963e-06,
+      "loss": 5.8995,
+      "step": 96500
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 9.815478589153535e-06,
+      "loss": 5.8921,
+      "step": 97000
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 9.814418121275107e-06,
+      "loss": 5.8938,
+      "step": 97500
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 9.81335765339668e-06,
+      "loss": 5.8907,
+      "step": 98000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.812297185518252e-06,
+      "loss": 5.8881,
+      "step": 98500
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.811236717639824e-06,
+      "loss": 5.8904,
+      "step": 99000
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 9.810176249761396e-06,
+      "loss": 5.8861,
+      "step": 99500
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.809115781882968e-06,
+      "loss": 5.8827,
+      "step": 100000
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 9.80805531400454e-06,
+      "loss": 5.8818,
+      "step": 100500
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 9.806994846126112e-06,
+      "loss": 5.88,
+      "step": 101000
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 9.805934378247684e-06,
+      "loss": 5.8762,
+      "step": 101500
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 9.804873910369256e-06,
+      "loss": 5.8756,
+      "step": 102000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 9.803813442490828e-06,
+      "loss": 5.8718,
+      "step": 102500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 9.8027529746124e-06,
+      "loss": 5.8696,
+      "step": 103000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 9.801692506733972e-06,
+      "loss": 5.8676,
+      "step": 103500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 9.800632038855544e-06,
+      "loss": 5.8667,
+      "step": 104000
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 9.799571570977116e-06,
+      "loss": 5.8648,
+      "step": 104500
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 9.798511103098688e-06,
+      "loss": 5.8673,
+      "step": 105000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 9.79745063522026e-06,
+      "loss": 5.8635,
+      "step": 105500
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 9.796390167341832e-06,
+      "loss": 5.859,
+      "step": 106000
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 9.795329699463404e-06,
+      "loss": 5.8543,
+      "step": 106500
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 9.794269231584976e-06,
+      "loss": 5.8519,
+      "step": 107000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 9.793208763706547e-06,
+      "loss": 5.8538,
+      "step": 107500
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 9.79214829582812e-06,
+      "loss": 5.8513,
+      "step": 108000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 9.791087827949693e-06,
+      "loss": 5.8498,
+      "step": 108500
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 9.790027360071265e-06,
+      "loss": 5.8461,
+      "step": 109000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 9.788966892192837e-06,
+      "loss": 5.8421,
+      "step": 109500
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 9.787906424314409e-06,
+      "loss": 5.8423,
+      "step": 110000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 9.786845956435981e-06,
+      "loss": 5.8457,
+      "step": 110500
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 9.785785488557551e-06,
+      "loss": 5.8397,
+      "step": 111000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 9.784725020679125e-06,
+      "loss": 5.8343,
+      "step": 111500
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 9.783664552800697e-06,
+      "loss": 5.8308,
+      "step": 112000
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 9.782604084922267e-06,
+      "loss": 5.8312,
+      "step": 112500
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 9.781543617043841e-06,
+      "loss": 5.8282,
+      "step": 113000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 9.780483149165413e-06,
+      "loss": 5.8263,
+      "step": 113500
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 9.779422681286984e-06,
+      "loss": 5.8257,
+      "step": 114000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 9.778362213408557e-06,
+      "loss": 5.8228,
+      "step": 114500
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 9.77730174553013e-06,
+      "loss": 5.8187,
+      "step": 115000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 9.776241277651701e-06,
+      "loss": 5.8178,
+      "step": 115500
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 9.775180809773272e-06,
+      "loss": 5.8174,
+      "step": 116000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 9.774120341894846e-06,
+      "loss": 5.812,
+      "step": 116500
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 9.773059874016418e-06,
+      "loss": 5.812,
+      "step": 117000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 9.771999406137988e-06,
+      "loss": 5.8087,
+      "step": 117500
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 9.770938938259562e-06,
+      "loss": 5.81,
+      "step": 118000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 9.769878470381134e-06,
+      "loss": 5.8033,
+      "step": 118500
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 9.768818002502704e-06,
+      "loss": 5.7976,
+      "step": 119000
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 9.767757534624276e-06,
+      "loss": 5.7953,
+      "step": 119500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 9.76669706674585e-06,
+      "loss": 5.7956,
+      "step": 120000
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 9.76563659886742e-06,
+      "loss": 5.793,
+      "step": 120500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 9.764576130988992e-06,
+      "loss": 5.7913,
+      "step": 121000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 9.763515663110566e-06,
+      "loss": 5.7906,
+      "step": 121500
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.762455195232136e-06,
+      "loss": 5.784,
+      "step": 122000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.761394727353709e-06,
+      "loss": 5.7842,
+      "step": 122500
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.760334259475282e-06,
+      "loss": 5.7773,
+      "step": 123000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 9.759273791596854e-06,
+      "loss": 5.7758,
+      "step": 123500
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 9.758213323718425e-06,
+      "loss": 5.7719,
+      "step": 124000
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 9.757152855839997e-06,
+      "loss": 5.7707,
+      "step": 124500
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 9.75609238796157e-06,
+      "loss": 5.7625,
+      "step": 125000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 9.755031920083141e-06,
+      "loss": 5.7589,
+      "step": 125500
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 9.753971452204713e-06,
+      "loss": 5.7565,
+      "step": 126000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 9.752910984326287e-06,
+      "loss": 5.7524,
+      "step": 126500
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 9.751850516447857e-06,
+      "loss": 5.7533,
+      "step": 127000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 9.750790048569429e-06,
+      "loss": 5.744,
+      "step": 127500
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 9.749729580691001e-06,
+      "loss": 5.7375,
+      "step": 128000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 9.748669112812573e-06,
+      "loss": 5.733,
+      "step": 128500
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 9.747608644934145e-06,
+      "loss": 5.7267,
+      "step": 129000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 9.746548177055717e-06,
+      "loss": 5.7186,
+      "step": 129500
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 9.745487709177291e-06,
+      "loss": 5.711,
+      "step": 130000
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 9.744427241298861e-06,
+      "loss": 5.7026,
+      "step": 130500
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 9.743366773420434e-06,
+      "loss": 5.6917,
+      "step": 131000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 9.742306305542006e-06,
+      "loss": 5.6811,
+      "step": 131500
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 9.741245837663578e-06,
+      "loss": 5.6735,
+      "step": 132000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 9.74018536978515e-06,
+      "loss": 5.6597,
+      "step": 132500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 9.739124901906722e-06,
+      "loss": 5.6461,
+      "step": 133000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 9.738064434028294e-06,
+      "loss": 5.6332,
+      "step": 133500
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 9.737003966149866e-06,
+      "loss": 5.6161,
+      "step": 134000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 9.735943498271438e-06,
+      "loss": 5.5956,
+      "step": 134500
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 9.73488303039301e-06,
+      "loss": 5.5786,
+      "step": 135000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.733822562514582e-06,
+      "loss": 5.5567,
+      "step": 135500
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 9.732762094636154e-06,
+      "loss": 5.5356,
+      "step": 136000
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 9.731701626757726e-06,
+      "loss": 5.5121,
+      "step": 136500
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 9.730641158879298e-06,
+      "loss": 5.4913,
+      "step": 137000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 9.72958069100087e-06,
+      "loss": 5.4667,
+      "step": 137500
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 9.728520223122442e-06,
+      "loss": 5.444,
+      "step": 138000
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 9.727459755244014e-06,
+      "loss": 5.4253,
+      "step": 138500
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 9.726399287365586e-06,
+      "loss": 5.3978,
+      "step": 139000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 9.725338819487158e-06,
+      "loss": 5.371,
+      "step": 139500
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.72427835160873e-06,
+      "loss": 5.3497,
+      "step": 140000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 9.723217883730303e-06,
+      "loss": 5.3215,
+      "step": 140500
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 9.722157415851875e-06,
+      "loss": 5.3026,
+      "step": 141000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.721096947973447e-06,
+      "loss": 5.2778,
+      "step": 141500
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 9.720036480095019e-06,
+      "loss": 5.2511,
+      "step": 142000
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 9.71897601221659e-06,
+      "loss": 5.2249,
+      "step": 142500
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 9.717915544338163e-06,
+      "loss": 5.1973,
+      "step": 143000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 9.716855076459735e-06,
+      "loss": 5.1697,
+      "step": 143500
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 9.715794608581307e-06,
+      "loss": 5.1364,
+      "step": 144000
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 9.714734140702879e-06,
+      "loss": 5.1041,
+      "step": 144500
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 9.713673672824451e-06,
+      "loss": 5.07,
+      "step": 145000
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 9.712613204946023e-06,
+      "loss": 5.0388,
+      "step": 145500
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 9.711552737067595e-06,
+      "loss": 4.997,
+      "step": 146000
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 9.710492269189167e-06,
+      "loss": 4.9617,
+      "step": 146500
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 9.70943180131074e-06,
+      "loss": 4.9256,
+      "step": 147000
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 9.708371333432311e-06,
+      "loss": 4.8889,
+      "step": 147500
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 9.707310865553883e-06,
+      "loss": 4.8512,
+      "step": 148000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 9.706250397675456e-06,
+      "loss": 4.8101,
+      "step": 148500
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 9.705189929797028e-06,
+      "loss": 4.7805,
+      "step": 149000
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 9.7041294619186e-06,
+      "loss": 4.7506,
+      "step": 149500
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 9.703068994040172e-06,
+      "loss": 4.7209,
+      "step": 150000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 9.702008526161744e-06,
+      "loss": 4.6953,
+      "step": 150500
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 9.700948058283314e-06,
+      "loss": 4.671,
+      "step": 151000
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 9.699887590404888e-06,
+      "loss": 4.6463,
+      "step": 151500
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 9.69882712252646e-06,
+      "loss": 4.6211,
+      "step": 152000
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 9.697766654648032e-06,
+      "loss": 4.5955,
+      "step": 152500
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 9.696706186769604e-06,
+      "loss": 4.5699,
+      "step": 153000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 9.695645718891176e-06,
+      "loss": 4.5515,
+      "step": 153500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 9.694585251012748e-06,
+      "loss": 4.5269,
+      "step": 154000
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 9.693524783134318e-06,
+      "loss": 4.5072,
+      "step": 154500
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 9.692464315255892e-06,
+      "loss": 4.4891,
+      "step": 155000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 9.691403847377464e-06,
+      "loss": 4.4732,
+      "step": 155500
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 9.690343379499035e-06,
+      "loss": 4.4603,
+      "step": 156000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 9.689282911620608e-06,
+      "loss": 4.4384,
+      "step": 156500
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 9.68822244374218e-06,
+      "loss": 4.4243,
+      "step": 157000
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 9.68716197586375e-06,
+      "loss": 4.4107,
+      "step": 157500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 9.686101507985325e-06,
+      "loss": 4.3883,
+      "step": 158000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 9.685041040106897e-06,
+      "loss": 4.3754,
+      "step": 158500
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 9.683980572228469e-06,
+      "loss": 4.3579,
+      "step": 159000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 9.682920104350039e-06,
+      "loss": 4.3508,
+      "step": 159500
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 9.681859636471613e-06,
+      "loss": 4.3345,
+      "step": 160000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 9.680799168593185e-06,
+      "loss": 4.3243,
+      "step": 160500
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 9.679738700714755e-06,
+      "loss": 4.3109,
+      "step": 161000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 9.678678232836329e-06,
+      "loss": 4.2968,
+      "step": 161500
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 9.677617764957901e-06,
+      "loss": 4.2919,
+      "step": 162000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 9.676557297079471e-06,
+      "loss": 4.2791,
+      "step": 162500
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 9.675496829201043e-06,
+      "loss": 4.2732,
+      "step": 163000
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.674436361322617e-06,
+      "loss": 4.2592,
+      "step": 163500
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.673375893444188e-06,
+      "loss": 4.2478,
+      "step": 164000
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.67231542556576e-06,
+      "loss": 4.2355,
+      "step": 164500
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.671254957687333e-06,
+      "loss": 4.2362,
+      "step": 165000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.670194489808904e-06,
+      "loss": 4.2221,
+      "step": 165500
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.669134021930476e-06,
+      "loss": 4.2126,
+      "step": 166000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.66807355405205e-06,
+      "loss": 4.2076,
+      "step": 166500
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.667013086173622e-06,
+      "loss": 4.2008,
+      "step": 167000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.665952618295192e-06,
+      "loss": 4.1944,
+      "step": 167500
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.664892150416764e-06,
+      "loss": 4.1811,
+      "step": 168000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.663831682538338e-06,
+      "loss": 4.171,
+      "step": 168500
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.662771214659908e-06,
+      "loss": 4.1648,
+      "step": 169000
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.66171074678148e-06,
+      "loss": 4.1578,
+      "step": 169500
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.660650278903054e-06,
+      "loss": 4.1465,
+      "step": 170000
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.659589811024624e-06,
+      "loss": 4.1465,
+      "step": 170500
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.658529343146196e-06,
+      "loss": 4.135,
+      "step": 171000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.657468875267768e-06,
+      "loss": 4.1276,
+      "step": 171500
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.65640840738934e-06,
+      "loss": 4.1281,
+      "step": 172000
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.655347939510913e-06,
+      "loss": 4.1178,
+      "step": 172500
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 9.654287471632485e-06,
+      "loss": 4.1146,
+      "step": 173000
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 9.653227003754057e-06,
+      "loss": 4.1086,
+      "step": 173500
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 9.652166535875629e-06,
+      "loss": 4.1017,
+      "step": 174000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.6511060679972e-06,
+      "loss": 4.0986,
+      "step": 174500
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 9.650045600118773e-06,
+      "loss": 4.0941,
+      "step": 175000
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 9.648985132240345e-06,
+      "loss": 4.0866,
+      "step": 175500
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 9.647924664361917e-06,
+      "loss": 4.082,
+      "step": 176000
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.646864196483489e-06,
+      "loss": 4.0786,
+      "step": 176500
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 9.645803728605061e-06,
+      "loss": 4.0701,
+      "step": 177000
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.644743260726633e-06,
+      "loss": 4.0672,
+      "step": 177500
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 9.643682792848205e-06,
+      "loss": 4.0611,
+      "step": 178000
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 9.642622324969777e-06,
+      "loss": 4.0576,
+      "step": 178500
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 9.64156185709135e-06,
+      "loss": 4.0546,
+      "step": 179000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.640501389212921e-06,
+      "loss": 4.0456,
+      "step": 179500
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.639440921334493e-06,
+      "loss": 4.0446,
+      "step": 180000
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 9.638380453456065e-06,
+      "loss": 4.0406,
+      "step": 180500
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.637319985577638e-06,
+      "loss": 4.0355,
+      "step": 181000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 9.63625951769921e-06,
+      "loss": 4.0315,
+      "step": 181500
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 9.635199049820782e-06,
+      "loss": 4.0271,
+      "step": 182000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.634138581942354e-06,
+      "loss": 4.0272,
+      "step": 182500
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 9.633078114063926e-06,
+      "loss": 4.0211,
+      "step": 183000
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 9.632017646185498e-06,
+      "loss": 4.0115,
+      "step": 183500
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 9.63095717830707e-06,
+      "loss": 4.0089,
+      "step": 184000
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 9.629896710428642e-06,
+      "loss": 4.0004,
+      "step": 184500
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 9.628836242550214e-06,
+      "loss": 4.0004,
+      "step": 185000
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 9.627775774671786e-06,
+      "loss": 3.9982,
+      "step": 185500
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 9.626715306793358e-06,
+      "loss": 3.9956,
+      "step": 186000
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 9.62565483891493e-06,
+      "loss": 3.9913,
+      "step": 186500
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 9.624594371036502e-06,
+      "loss": 3.9897,
+      "step": 187000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 9.623533903158074e-06,
+      "loss": 3.9808,
+      "step": 187500
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 9.622473435279646e-06,
+      "loss": 3.9768,
+      "step": 188000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 9.621412967401218e-06,
+      "loss": 3.974,
+      "step": 188500
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.62035249952279e-06,
+      "loss": 3.9686,
+      "step": 189000
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.619292031644362e-06,
+      "loss": 3.9639,
+      "step": 189500
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.618231563765935e-06,
+      "loss": 3.961,
+      "step": 190000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.617171095887507e-06,
+      "loss": 3.9583,
+      "step": 190500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.616110628009079e-06,
+      "loss": 3.9557,
+      "step": 191000
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 9.61505016013065e-06,
+      "loss": 3.9514,
+      "step": 191500
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 9.613989692252223e-06,
+      "loss": 3.9442,
+      "step": 192000
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 9.612929224373795e-06,
+      "loss": 3.9443,
+      "step": 192500
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 9.611868756495367e-06,
+      "loss": 3.9393,
+      "step": 193000
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 9.610808288616939e-06,
+      "loss": 3.9406,
+      "step": 193500
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 9.609747820738511e-06,
+      "loss": 3.9306,
+      "step": 194000
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 9.608687352860081e-06,
+      "loss": 3.9312,
+      "step": 194500
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 9.607626884981655e-06,
+      "loss": 3.9263,
+      "step": 195000
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 9.606566417103227e-06,
+      "loss": 3.9233,
+      "step": 195500
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 9.6055059492248e-06,
+      "loss": 3.9235,
+      "step": 196000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 9.604445481346371e-06,
+      "loss": 3.9183,
+      "step": 196500
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 9.603385013467943e-06,
+      "loss": 3.9129,
+      "step": 197000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 9.602324545589515e-06,
+      "loss": 3.9096,
+      "step": 197500
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 9.601264077711086e-06,
+      "loss": 3.9053,
+      "step": 198000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 9.60020360983266e-06,
+      "loss": 3.9058,
+      "step": 198500
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 9.599143141954232e-06,
+      "loss": 3.8997,
+      "step": 199000
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 9.598082674075802e-06,
+      "loss": 3.8997,
+      "step": 199500
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 9.597022206197376e-06,
+      "loss": 3.8978,
+      "step": 200000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 9.595961738318948e-06,
+      "loss": 3.8954,
+      "step": 200500
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 9.594901270440518e-06,
+      "loss": 3.8907,
+      "step": 201000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 9.593840802562092e-06,
+      "loss": 3.8884,
+      "step": 201500
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 9.592780334683664e-06,
+      "loss": 3.8837,
+      "step": 202000
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 9.591719866805234e-06,
+      "loss": 3.8801,
+      "step": 202500
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 9.590659398926806e-06,
+      "loss": 3.8807,
+      "step": 203000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 9.58959893104838e-06,
+      "loss": 3.8753,
+      "step": 203500
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 9.588538463169952e-06,
+      "loss": 3.8759,
+      "step": 204000
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 9.587477995291522e-06,
+      "loss": 3.873,
+      "step": 204500
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 9.586417527413096e-06,
+      "loss": 3.8664,
+      "step": 205000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 9.585357059534668e-06,
+      "loss": 3.8635,
+      "step": 205500
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 9.584296591656239e-06,
+      "loss": 3.8651,
+      "step": 206000
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 9.58323612377781e-06,
+      "loss": 3.8572,
+      "step": 206500
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 9.582175655899384e-06,
+      "loss": 3.858,
+      "step": 207000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 9.581115188020955e-06,
+      "loss": 3.8526,
+      "step": 207500
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 9.580054720142527e-06,
+      "loss": 3.8523,
+      "step": 208000
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 9.5789942522641e-06,
+      "loss": 3.8488,
+      "step": 208500
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 9.577933784385671e-06,
+      "loss": 3.8488,
+      "step": 209000
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 9.576873316507243e-06,
+      "loss": 3.8397,
+      "step": 209500
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 9.575812848628817e-06,
+      "loss": 3.8399,
+      "step": 210000
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 9.574752380750389e-06,
+      "loss": 3.84,
+      "step": 210500
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 9.57369191287196e-06,
+      "loss": 3.8327,
+      "step": 211000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 9.572631444993531e-06,
+      "loss": 3.8315,
+      "step": 211500
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 9.571570977115105e-06,
+      "loss": 3.8314,
+      "step": 212000
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 9.570510509236675e-06,
+      "loss": 3.8259,
+      "step": 212500
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 9.569450041358247e-06,
+      "loss": 3.8266,
+      "step": 213000
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 9.568389573479821e-06,
+      "loss": 3.8228,
+      "step": 213500
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 9.567329105601392e-06,
+      "loss": 3.8155,
+      "step": 214000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 9.566268637722964e-06,
+      "loss": 3.8178,
+      "step": 214500
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 9.565208169844536e-06,
+      "loss": 3.8154,
+      "step": 215000
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 9.564147701966108e-06,
+      "loss": 3.8108,
+      "step": 215500
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 9.56308723408768e-06,
+      "loss": 3.8081,
+      "step": 216000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 9.562026766209252e-06,
+      "loss": 3.8033,
+      "step": 216500
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 9.560966298330824e-06,
+      "loss": 3.8055,
+      "step": 217000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 9.559905830452396e-06,
+      "loss": 3.8058,
+      "step": 217500
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 9.558845362573968e-06,
+      "loss": 3.8,
+      "step": 218000
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 9.557784894695542e-06,
+      "loss": 3.7926,
+      "step": 218500
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 9.556724426817112e-06,
+      "loss": 3.794,
+      "step": 219000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 9.555663958938684e-06,
+      "loss": 3.7948,
+      "step": 219500
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 9.554603491060256e-06,
+      "loss": 3.7903,
+      "step": 220000
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 9.553543023181828e-06,
+      "loss": 3.7873,
+      "step": 220500
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 9.5524825553034e-06,
+      "loss": 3.781,
+      "step": 221000
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 9.551422087424972e-06,
+      "loss": 3.778,
+      "step": 221500
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 9.550361619546544e-06,
+      "loss": 3.776,
+      "step": 222000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 9.549301151668117e-06,
+      "loss": 3.7793,
+      "step": 222500
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 9.548240683789689e-06,
+      "loss": 3.7759,
+      "step": 223000
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 9.54718021591126e-06,
+      "loss": 3.7702,
+      "step": 223500
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 9.546119748032833e-06,
+      "loss": 3.769,
+      "step": 224000
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 9.545059280154405e-06,
+      "loss": 3.765,
+      "step": 224500
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 9.543998812275977e-06,
+      "loss": 3.7651,
+      "step": 225000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 9.542938344397549e-06,
+      "loss": 3.7583,
+      "step": 225500
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 9.541877876519121e-06,
+      "loss": 3.7576,
+      "step": 226000
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 9.540817408640693e-06,
+      "loss": 3.7607,
+      "step": 226500
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 9.539756940762265e-06,
+      "loss": 3.7528,
+      "step": 227000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 9.538696472883837e-06,
+      "loss": 3.757,
+      "step": 227500
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 9.53763600500541e-06,
+      "loss": 3.7447,
+      "step": 228000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 9.536575537126981e-06,
+      "loss": 3.7489,
+      "step": 228500
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 9.535515069248553e-06,
+      "loss": 3.7448,
+      "step": 229000
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 9.534454601370125e-06,
+      "loss": 3.7422,
+      "step": 229500
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 9.533394133491697e-06,
+      "loss": 3.7387,
+      "step": 230000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 9.53233366561327e-06,
+      "loss": 3.7366,
+      "step": 230500
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 9.531273197734842e-06,
+      "loss": 3.729,
+      "step": 231000
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 9.530212729856414e-06,
+      "loss": 3.7324,
+      "step": 231500
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 9.529152261977986e-06,
+      "loss": 3.729,
+      "step": 232000
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 9.528091794099558e-06,
+      "loss": 3.7213,
+      "step": 232500
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 9.52703132622113e-06,
+      "loss": 3.7219,
+      "step": 233000
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 9.525970858342702e-06,
+      "loss": 3.7196,
+      "step": 233500
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 9.524910390464274e-06,
+      "loss": 3.7215,
+      "step": 234000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 9.523849922585846e-06,
+      "loss": 3.7179,
+      "step": 234500
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 9.522789454707418e-06,
+      "loss": 3.7128,
+      "step": 235000
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 9.52172898682899e-06,
+      "loss": 3.7145,
+      "step": 235500
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 9.520668518950562e-06,
+      "loss": 3.7166,
+      "step": 236000
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 9.519608051072134e-06,
+      "loss": 3.7055,
+      "step": 236500
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 9.518547583193706e-06,
+      "loss": 3.6995,
+      "step": 237000
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 9.517487115315278e-06,
+      "loss": 3.704,
+      "step": 237500
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 9.516426647436849e-06,
+      "loss": 3.699,
+      "step": 238000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 9.515366179558422e-06,
+      "loss": 3.6973,
+      "step": 238500
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 9.514305711679994e-06,
+      "loss": 3.6919,
+      "step": 239000
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 9.513245243801566e-06,
+      "loss": 3.6888,
+      "step": 239500
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 9.512184775923139e-06,
+      "loss": 3.6869,
+      "step": 240000
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 9.51112430804471e-06,
+      "loss": 3.6846,
+      "step": 240500
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 9.510063840166283e-06,
+      "loss": 3.6845,
+      "step": 241000
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 9.509003372287855e-06,
+      "loss": 3.6802,
+      "step": 241500
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 9.507942904409427e-06,
+      "loss": 3.6773,
+      "step": 242000
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 9.506882436530999e-06,
+      "loss": 3.6766,
+      "step": 242500
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 9.50582196865257e-06,
+      "loss": 3.6725,
+      "step": 243000
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 9.504761500774143e-06,
+      "loss": 3.6757,
+      "step": 243500
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 9.503701032895715e-06,
+      "loss": 3.6714,
+      "step": 244000
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 9.502640565017285e-06,
+      "loss": 3.6655,
+      "step": 244500
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 9.501580097138859e-06,
+      "loss": 3.663,
+      "step": 245000
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 9.500519629260431e-06,
+      "loss": 3.661,
+      "step": 245500
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 9.499459161382002e-06,
+      "loss": 3.6603,
+      "step": 246000
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 9.498398693503574e-06,
+      "loss": 3.6558,
+      "step": 246500
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 9.497338225625147e-06,
+      "loss": 3.6516,
+      "step": 247000
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 9.49627775774672e-06,
+      "loss": 3.6473,
+      "step": 247500
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 9.49521728986829e-06,
+      "loss": 3.6462,
+      "step": 248000
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 9.494156821989864e-06,
+      "loss": 3.6451,
+      "step": 248500
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 9.493096354111436e-06,
+      "loss": 3.6426,
+      "step": 249000
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 9.492035886233006e-06,
+      "loss": 3.6415,
+      "step": 249500
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 9.490975418354578e-06,
+      "loss": 3.6416,
+      "step": 250000
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 9.489914950476152e-06,
+      "loss": 3.6379,
+      "step": 250500
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 9.488854482597722e-06,
+      "loss": 3.6336,
+      "step": 251000
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 9.487794014719294e-06,
+      "loss": 3.6349,
+      "step": 251500
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 9.486733546840868e-06,
+      "loss": 3.6346,
+      "step": 252000
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 9.485673078962438e-06,
+      "loss": 3.6296,
+      "step": 252500
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 9.48461261108401e-06,
+      "loss": 3.6285,
+      "step": 253000
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 9.483552143205584e-06,
+      "loss": 3.6202,
+      "step": 253500
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 9.482491675327156e-06,
+      "loss": 3.6227,
+      "step": 254000
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 9.481431207448726e-06,
+      "loss": 3.6183,
+      "step": 254500
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 9.480370739570299e-06,
+      "loss": 3.6164,
+      "step": 255000
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 9.479310271691872e-06,
+      "loss": 3.6182,
+      "step": 255500
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 9.478249803813443e-06,
+      "loss": 3.6094,
+      "step": 256000
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 9.477189335935015e-06,
+      "loss": 3.6059,
+      "step": 256500
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 9.476128868056588e-06,
+      "loss": 3.6092,
+      "step": 257000
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 9.475068400178159e-06,
+      "loss": 3.6027,
+      "step": 257500
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 9.474007932299731e-06,
+      "loss": 3.6035,
+      "step": 258000
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 9.472947464421303e-06,
+      "loss": 3.6048,
+      "step": 258500
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 9.471886996542875e-06,
+      "loss": 3.5984,
+      "step": 259000
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 9.470826528664447e-06,
+      "loss": 3.599,
+      "step": 259500
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 9.469766060786019e-06,
+      "loss": 3.5953,
+      "step": 260000
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 9.468705592907591e-06,
+      "loss": 3.5915,
+      "step": 260500
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 9.467645125029163e-06,
+      "loss": 3.5924,
+      "step": 261000
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 9.466584657150735e-06,
+      "loss": 3.5899,
+      "step": 261500
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 9.465524189272309e-06,
+      "loss": 3.5848,
+      "step": 262000
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 9.46446372139388e-06,
+      "loss": 3.5842,
+      "step": 262500
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 9.463403253515451e-06,
+      "loss": 3.5813,
+      "step": 263000
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 9.462342785637024e-06,
+      "loss": 3.5788,
+      "step": 263500
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 9.461282317758596e-06,
+      "loss": 3.5807,
+      "step": 264000
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 9.460221849880168e-06,
+      "loss": 3.5835,
+      "step": 264500
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 9.45916138200174e-06,
+      "loss": 3.5747,
+      "step": 265000
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 9.458100914123312e-06,
+      "loss": 3.5752,
+      "step": 265500
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 9.457040446244884e-06,
+      "loss": 3.5651,
+      "step": 266000
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 9.455979978366456e-06,
+      "loss": 3.5679,
+      "step": 266500
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 9.454919510488028e-06,
+      "loss": 3.5685,
+      "step": 267000
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 9.4538590426096e-06,
+      "loss": 3.5684,
+      "step": 267500
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 9.452798574731172e-06,
+      "loss": 3.565,
+      "step": 268000
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 9.451738106852744e-06,
+      "loss": 3.5604,
+      "step": 268500
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 9.450677638974316e-06,
+      "loss": 3.557,
+      "step": 269000
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 9.449617171095888e-06,
+      "loss": 3.5568,
+      "step": 269500
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 9.44855670321746e-06,
+      "loss": 3.558,
+      "step": 270000
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 9.447496235339032e-06,
+      "loss": 3.554,
+      "step": 270500
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 9.446435767460604e-06,
+      "loss": 3.5498,
+      "step": 271000
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 9.445375299582176e-06,
+      "loss": 3.5487,
+      "step": 271500
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 9.444314831703748e-06,
+      "loss": 3.5492,
+      "step": 272000
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 9.44325436382532e-06,
+      "loss": 3.5505,
+      "step": 272500
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 9.442193895946893e-06,
+      "loss": 3.5454,
+      "step": 273000
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 9.441133428068465e-06,
+      "loss": 3.5382,
+      "step": 273500
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 9.440072960190037e-06,
+      "loss": 3.5435,
+      "step": 274000
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 9.439012492311609e-06,
+      "loss": 3.5352,
+      "step": 274500
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 9.43795202443318e-06,
+      "loss": 3.5396,
+      "step": 275000
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 9.436891556554753e-06,
+      "loss": 3.5385,
+      "step": 275500
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 9.435831088676325e-06,
+      "loss": 3.5345,
+      "step": 276000
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 9.434770620797897e-06,
+      "loss": 3.536,
+      "step": 276500
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 9.433710152919469e-06,
+      "loss": 3.5328,
+      "step": 277000
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 9.432649685041041e-06,
+      "loss": 3.5295,
+      "step": 277500
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 9.431589217162613e-06,
+      "loss": 3.523,
+      "step": 278000
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 9.430528749284185e-06,
+      "loss": 3.5285,
+      "step": 278500
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 9.429468281405757e-06,
+      "loss": 3.5246,
+      "step": 279000
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 9.42840781352733e-06,
+      "loss": 3.5232,
+      "step": 279500
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 9.427347345648901e-06,
+      "loss": 3.5244,
+      "step": 280000
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 9.426286877770473e-06,
+      "loss": 3.5188,
+      "step": 280500
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 9.425226409892046e-06,
+      "loss": 3.515,
+      "step": 281000
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 9.424165942013616e-06,
+      "loss": 3.5144,
+      "step": 281500
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 9.42310547413519e-06,
+      "loss": 3.516,
+      "step": 282000
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 9.422045006256762e-06,
+      "loss": 3.5101,
+      "step": 282500
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 9.420984538378334e-06,
+      "loss": 3.5159,
+      "step": 283000
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 9.419924070499906e-06,
+      "loss": 3.5143,
+      "step": 283500
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 9.418863602621478e-06,
+      "loss": 3.5109,
+      "step": 284000
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 9.41780313474305e-06,
+      "loss": 3.5023,
+      "step": 284500
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 9.416742666864622e-06,
+      "loss": 3.5023,
+      "step": 285000
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 9.415682198986194e-06,
+      "loss": 3.5007,
+      "step": 285500
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 9.414621731107766e-06,
+      "loss": 3.5021,
+      "step": 286000
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 9.413561263229336e-06,
+      "loss": 3.5046,
+      "step": 286500
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 9.41250079535091e-06,
+      "loss": 3.4983,
+      "step": 287000
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 9.411440327472482e-06,
+      "loss": 3.4976,
+      "step": 287500
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 9.410379859594053e-06,
+      "loss": 3.497,
+      "step": 288000
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 9.409319391715626e-06,
+      "loss": 3.4954,
+      "step": 288500
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 9.408258923837198e-06,
+      "loss": 3.4934,
+      "step": 289000
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 9.407198455958769e-06,
+      "loss": 3.4946,
+      "step": 289500
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 9.40613798808034e-06,
+      "loss": 3.4891,
+      "step": 290000
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 9.405077520201915e-06,
+      "loss": 3.4827,
+      "step": 290500
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 9.404017052323487e-06,
+      "loss": 3.4887,
+      "step": 291000
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 9.402956584445057e-06,
+      "loss": 3.4853,
+      "step": 291500
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 9.40189611656663e-06,
+      "loss": 3.4865,
+      "step": 292000
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 9.400835648688203e-06,
+      "loss": 3.4813,
+      "step": 292500
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 9.399775180809773e-06,
+      "loss": 3.4875,
+      "step": 293000
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 9.398714712931345e-06,
+      "loss": 3.4763,
+      "step": 293500
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 9.397654245052919e-06,
+      "loss": 3.48,
+      "step": 294000
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 9.39659377717449e-06,
+      "loss": 3.477,
+      "step": 294500
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 9.395533309296061e-06,
+      "loss": 3.4744,
+      "step": 295000
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 9.394472841417635e-06,
+      "loss": 3.4738,
+      "step": 295500
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 9.393412373539206e-06,
+      "loss": 3.475,
+      "step": 296000
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 9.392351905660778e-06,
+      "loss": 3.4727,
+      "step": 296500
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 9.391291437782351e-06,
+      "loss": 3.4728,
+      "step": 297000
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 9.390230969903923e-06,
+      "loss": 3.4692,
+      "step": 297500
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 9.389170502025494e-06,
+      "loss": 3.4672,
+      "step": 298000
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 9.388110034147066e-06,
+      "loss": 3.4681,
+      "step": 298500
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 9.38704956626864e-06,
+      "loss": 3.4627,
+      "step": 299000
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 9.38598909839021e-06,
+      "loss": 3.4635,
+      "step": 299500
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 9.384928630511782e-06,
+      "loss": 3.468,
+      "step": 300000
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 9.383868162633356e-06,
+      "loss": 3.46,
+      "step": 300500
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 9.382807694754926e-06,
+      "loss": 3.4592,
+      "step": 301000
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 9.381747226876498e-06,
+      "loss": 3.4574,
+      "step": 301500
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 9.38068675899807e-06,
+      "loss": 3.4546,
+      "step": 302000
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 9.379626291119642e-06,
+      "loss": 3.4538,
+      "step": 302500
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 9.378565823241214e-06,
+      "loss": 3.4551,
+      "step": 303000
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 9.377505355362786e-06,
+      "loss": 3.4501,
+      "step": 303500
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 9.376444887484358e-06,
+      "loss": 3.4497,
+      "step": 304000
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 9.37538441960593e-06,
+      "loss": 3.4535,
+      "step": 304500
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 9.374323951727503e-06,
+      "loss": 3.4458,
+      "step": 305000
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 9.373263483849076e-06,
+      "loss": 3.4496,
+      "step": 305500
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 9.372203015970647e-06,
+      "loss": 3.4443,
+      "step": 306000
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 9.371142548092219e-06,
+      "loss": 3.4426,
+      "step": 306500
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 9.37008208021379e-06,
+      "loss": 3.4463,
+      "step": 307000
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 9.369021612335363e-06,
+      "loss": 3.4414,
+      "step": 307500
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 9.367961144456935e-06,
+      "loss": 3.4439,
+      "step": 308000
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 9.366900676578507e-06,
+      "loss": 3.4361,
+      "step": 308500
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 9.365840208700079e-06,
+      "loss": 3.4364,
+      "step": 309000
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 9.364779740821651e-06,
+      "loss": 3.4359,
+      "step": 309500
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 9.363719272943223e-06,
+      "loss": 3.4376,
+      "step": 310000
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 9.362658805064795e-06,
+      "loss": 3.4365,
+      "step": 310500
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 9.361598337186367e-06,
+      "loss": 3.4334,
+      "step": 311000
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 9.36053786930794e-06,
+      "loss": 3.4312,
+      "step": 311500
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 9.359477401429511e-06,
+      "loss": 3.4309,
+      "step": 312000
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 9.358416933551083e-06,
+      "loss": 3.4294,
+      "step": 312500
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 9.357356465672655e-06,
+      "loss": 3.4315,
+      "step": 313000
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 9.356295997794228e-06,
+      "loss": 3.4292,
+      "step": 313500
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 9.3552355299158e-06,
+      "loss": 3.4235,
+      "step": 314000
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 9.354175062037372e-06,
+      "loss": 3.4239,
+      "step": 314500
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 9.353114594158944e-06,
+      "loss": 3.4271,
+      "step": 315000
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 9.352054126280516e-06,
+      "loss": 3.425,
+      "step": 315500
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 9.350993658402088e-06,
+      "loss": 3.4216,
+      "step": 316000
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 9.34993319052366e-06,
+      "loss": 3.4231,
+      "step": 316500
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 9.348872722645232e-06,
+      "loss": 3.4182,
+      "step": 317000
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 9.347812254766804e-06,
+      "loss": 3.4201,
+      "step": 317500
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 9.346751786888376e-06,
+      "loss": 3.4149,
+      "step": 318000
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 9.345691319009948e-06,
+      "loss": 3.4201,
+      "step": 318500
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 9.34463085113152e-06,
+      "loss": 3.4136,
+      "step": 319000
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 9.343570383253092e-06,
+      "loss": 3.416,
+      "step": 319500
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 9.342509915374664e-06,
+      "loss": 3.4113,
+      "step": 320000
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 9.341449447496236e-06,
+      "loss": 3.4144,
+      "step": 320500
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 9.340388979617808e-06,
+      "loss": 3.4132,
+      "step": 321000
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 9.33932851173938e-06,
+      "loss": 3.4094,
+      "step": 321500
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 9.338268043860952e-06,
+      "loss": 3.4109,
+      "step": 322000
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 9.337207575982525e-06,
+      "loss": 3.4106,
+      "step": 322500
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 9.336147108104097e-06,
+      "loss": 3.4075,
+      "step": 323000
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 9.335086640225669e-06,
+      "loss": 3.4041,
+      "step": 323500
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 9.33402617234724e-06,
+      "loss": 3.4064,
+      "step": 324000
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 9.332965704468813e-06,
+      "loss": 3.4027,
+      "step": 324500
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 9.331905236590383e-06,
+      "loss": 3.4003,
+      "step": 325000
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 9.330844768711957e-06,
+      "loss": 3.3966,
+      "step": 325500
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 9.329784300833529e-06,
+      "loss": 3.395,
+      "step": 326000
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 9.328723832955101e-06,
+      "loss": 3.3988,
+      "step": 326500
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 9.327663365076673e-06,
+      "loss": 3.4019,
+      "step": 327000
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 9.326602897198245e-06,
+      "loss": 3.394,
+      "step": 327500
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 9.325542429319817e-06,
+      "loss": 3.395,
+      "step": 328000
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 9.32448196144139e-06,
+      "loss": 3.3954,
+      "step": 328500
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 9.323421493562961e-06,
+      "loss": 3.3921,
+      "step": 329000
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 9.322361025684533e-06,
+      "loss": 3.3951,
+      "step": 329500
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 9.321300557806104e-06,
+      "loss": 3.3895,
+      "step": 330000
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 9.320240089927677e-06,
+      "loss": 3.3886,
+      "step": 330500
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 9.31917962204925e-06,
+      "loss": 3.3855,
+      "step": 331000
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 9.31811915417082e-06,
+      "loss": 3.3858,
+      "step": 331500
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 9.317058686292394e-06,
+      "loss": 3.3864,
+      "step": 332000
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 9.315998218413966e-06,
+      "loss": 3.3834,
+      "step": 332500
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 9.314937750535536e-06,
+      "loss": 3.3897,
+      "step": 333000
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 9.313877282657108e-06,
+      "loss": 3.3874,
+      "step": 333500
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 9.312816814778682e-06,
+      "loss": 3.3837,
+      "step": 334000
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 9.311756346900254e-06,
+      "loss": 3.3831,
+      "step": 334500
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 9.310695879021824e-06,
+      "loss": 3.3797,
+      "step": 335000
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 9.309635411143398e-06,
+      "loss": 3.3791,
+      "step": 335500
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 9.30857494326497e-06,
+      "loss": 3.3778,
+      "step": 336000
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 9.30751447538654e-06,
+      "loss": 3.3813,
+      "step": 336500
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 9.306454007508113e-06,
+      "loss": 3.3786,
+      "step": 337000
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 9.305393539629686e-06,
+      "loss": 3.375,
+      "step": 337500
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 9.304333071751257e-06,
+      "loss": 3.3786,
+      "step": 338000
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 9.303272603872829e-06,
+      "loss": 3.374,
+      "step": 338500
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 9.302212135994402e-06,
+      "loss": 3.3721,
+      "step": 339000
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 9.301151668115973e-06,
+      "loss": 3.3742,
+      "step": 339500
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 9.300091200237545e-06,
+      "loss": 3.3734,
+      "step": 340000
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 9.299030732359119e-06,
+      "loss": 3.365,
+      "step": 340500
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 9.29797026448069e-06,
+      "loss": 3.3726,
+      "step": 341000
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 9.296909796602261e-06,
+      "loss": 3.3665,
+      "step": 341500
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 9.295849328723833e-06,
+      "loss": 3.3622,
+      "step": 342000
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 9.294788860845407e-06,
+      "loss": 3.3664,
+      "step": 342500
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 9.293728392966977e-06,
+      "loss": 3.3667,
+      "step": 343000
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 9.29266792508855e-06,
+      "loss": 3.3641,
+      "step": 343500
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 9.291607457210123e-06,
+      "loss": 3.3619,
+      "step": 344000
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 9.290546989331693e-06,
+      "loss": 3.364,
+      "step": 344500
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 9.289486521453265e-06,
+      "loss": 3.3629,
+      "step": 345000
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 9.288426053574837e-06,
+      "loss": 3.3597,
+      "step": 345500
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 9.28736558569641e-06,
+      "loss": 3.3541,
+      "step": 346000
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 9.286305117817982e-06,
+      "loss": 3.3643,
+      "step": 346500
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 9.285244649939554e-06,
+      "loss": 3.3558,
+      "step": 347000
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 9.284184182061126e-06,
+      "loss": 3.3574,
+      "step": 347500
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 9.283123714182698e-06,
+      "loss": 3.3549,
+      "step": 348000
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 9.28206324630427e-06,
+      "loss": 3.3562,
+      "step": 348500
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 9.281002778425844e-06,
+      "loss": 3.3544,
+      "step": 349000
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 9.279942310547414e-06,
+      "loss": 3.3534,
+      "step": 349500
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 9.278881842668986e-06,
+      "loss": 3.3485,
+      "step": 350000
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 9.277821374790558e-06,
+      "loss": 3.3504,
+      "step": 350500
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 9.27676090691213e-06,
+      "loss": 3.3514,
+      "step": 351000
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 9.275700439033702e-06,
+      "loss": 3.3484,
+      "step": 351500
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 9.274639971155274e-06,
+      "loss": 3.3512,
+      "step": 352000
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 9.273579503276846e-06,
+      "loss": 3.3507,
+      "step": 352500
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 9.272519035398418e-06,
+      "loss": 3.3474,
+      "step": 353000
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 9.27145856751999e-06,
+      "loss": 3.3468,
+      "step": 353500
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 9.270398099641562e-06,
+      "loss": 3.3491,
+      "step": 354000
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 9.269337631763134e-06,
+      "loss": 3.3434,
+      "step": 354500
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 9.268277163884707e-06,
+      "loss": 3.3381,
+      "step": 355000
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 9.267216696006279e-06,
+      "loss": 3.3459,
+      "step": 355500
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 9.26615622812785e-06,
+      "loss": 3.3419,
+      "step": 356000
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 9.265095760249423e-06,
+      "loss": 3.3419,
+      "step": 356500
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 9.264035292370995e-06,
+      "loss": 3.343,
+      "step": 357000
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 9.262974824492567e-06,
+      "loss": 3.3416,
+      "step": 357500
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 9.261914356614139e-06,
+      "loss": 3.3389,
+      "step": 358000
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 9.260853888735711e-06,
+      "loss": 3.3432,
+      "step": 358500
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 9.259793420857283e-06,
+      "loss": 3.3366,
+      "step": 359000
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 9.258732952978855e-06,
+      "loss": 3.3379,
+      "step": 359500
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 9.257672485100427e-06,
+      "loss": 3.3335,
+      "step": 360000
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 9.256612017222e-06,
+      "loss": 3.3285,
+      "step": 360500
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 9.255551549343571e-06,
+      "loss": 3.3366,
+      "step": 361000
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 9.254491081465143e-06,
+      "loss": 3.3306,
+      "step": 361500
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 9.253430613586715e-06,
+      "loss": 3.3308,
+      "step": 362000
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 9.252370145708287e-06,
+      "loss": 3.3286,
+      "step": 362500
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 9.25130967782986e-06,
+      "loss": 3.3287,
+      "step": 363000
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 9.250249209951432e-06,
+      "loss": 3.3302,
+      "step": 363500
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 9.249188742073004e-06,
+      "loss": 3.3306,
+      "step": 364000
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 9.248128274194576e-06,
+      "loss": 3.3285,
+      "step": 364500
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 9.247067806316148e-06,
+      "loss": 3.3257,
+      "step": 365000
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 9.24600733843772e-06,
+      "loss": 3.3244,
+      "step": 365500
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 9.244946870559292e-06,
+      "loss": 3.3262,
+      "step": 366000
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 9.243886402680864e-06,
+      "loss": 3.3265,
+      "step": 366500
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 9.242825934802436e-06,
+      "loss": 3.3231,
+      "step": 367000
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 9.241765466924008e-06,
+      "loss": 3.3228,
+      "step": 367500
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 9.24070499904558e-06,
+      "loss": 3.3233,
+      "step": 368000
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 9.23964453116715e-06,
+      "loss": 3.3209,
+      "step": 368500
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 9.238584063288724e-06,
+      "loss": 3.3206,
+      "step": 369000
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 9.237523595410296e-06,
+      "loss": 3.3184,
+      "step": 369500
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 9.236463127531868e-06,
+      "loss": 3.3204,
+      "step": 370000
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 9.23540265965344e-06,
+      "loss": 3.3166,
+      "step": 370500
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 9.234342191775012e-06,
+      "loss": 3.3192,
+      "step": 371000
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 9.233281723896584e-06,
+      "loss": 3.3152,
+      "step": 371500
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 9.232221256018156e-06,
+      "loss": 3.3141,
+      "step": 372000
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 9.231160788139729e-06,
+      "loss": 3.3126,
+      "step": 372500
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 9.2301003202613e-06,
+      "loss": 3.3103,
+      "step": 373000
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 9.229039852382871e-06,
+      "loss": 3.3135,
+      "step": 373500
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 9.227979384504445e-06,
+      "loss": 3.3107,
+      "step": 374000
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 9.226918916626017e-06,
+      "loss": 3.3091,
+      "step": 374500
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 9.225858448747587e-06,
+      "loss": 3.3088,
+      "step": 375000
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 9.224797980869161e-06,
+      "loss": 3.3099,
+      "step": 375500
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 9.223737512990733e-06,
+      "loss": 3.3115,
+      "step": 376000
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 9.222677045112303e-06,
+      "loss": 3.3089,
+      "step": 376500
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 9.221616577233875e-06,
+      "loss": 3.3086,
+      "step": 377000
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 9.220556109355449e-06,
+      "loss": 3.3074,
+      "step": 377500
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 9.219495641477021e-06,
+      "loss": 3.3125,
+      "step": 378000
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 9.218435173598592e-06,
+      "loss": 3.3036,
+      "step": 378500
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 9.217374705720165e-06,
+      "loss": 3.303,
+      "step": 379000
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 9.216314237841737e-06,
+      "loss": 3.305,
+      "step": 379500
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 9.215253769963308e-06,
+      "loss": 3.3005,
+      "step": 380000
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 9.214193302084881e-06,
+      "loss": 3.2986,
+      "step": 380500
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 9.213132834206454e-06,
+      "loss": 3.2993,
+      "step": 381000
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 9.212072366328024e-06,
+      "loss": 3.2999,
+      "step": 381500
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 9.211011898449596e-06,
+      "loss": 3.3021,
+      "step": 382000
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 9.20995143057117e-06,
+      "loss": 3.2982,
+      "step": 382500
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 9.20889096269274e-06,
+      "loss": 3.2981,
+      "step": 383000
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 9.207830494814312e-06,
+      "loss": 3.2974,
+      "step": 383500
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 9.206770026935886e-06,
+      "loss": 3.2924,
+      "step": 384000
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 9.205709559057458e-06,
+      "loss": 3.2967,
+      "step": 384500
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 9.204649091179028e-06,
+      "loss": 3.2964,
+      "step": 385000
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 9.2035886233006e-06,
+      "loss": 3.2937,
+      "step": 385500
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 9.202528155422174e-06,
+      "loss": 3.2978,
+      "step": 386000
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 9.201467687543744e-06,
+      "loss": 3.2905,
+      "step": 386500
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 9.200407219665316e-06,
+      "loss": 3.2955,
+      "step": 387000
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 9.19934675178689e-06,
+      "loss": 3.2918,
+      "step": 387500
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 9.19828628390846e-06,
+      "loss": 3.2904,
+      "step": 388000
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 9.197225816030033e-06,
+      "loss": 3.2932,
+      "step": 388500
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 9.196165348151605e-06,
+      "loss": 3.2922,
+      "step": 389000
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 9.195104880273177e-06,
+      "loss": 3.2901,
+      "step": 389500
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 9.194044412394749e-06,
+      "loss": 3.293,
+      "step": 390000
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 9.192983944516321e-06,
+      "loss": 3.2875,
+      "step": 390500
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 9.191923476637893e-06,
+      "loss": 3.2888,
+      "step": 391000
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 9.190863008759465e-06,
+      "loss": 3.2872,
+      "step": 391500
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 9.189802540881037e-06,
+      "loss": 3.2897,
+      "step": 392000
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 9.18874207300261e-06,
+      "loss": 3.2908,
+      "step": 392500
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 9.187681605124181e-06,
+      "loss": 3.2869,
+      "step": 393000
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 9.186621137245753e-06,
+      "loss": 3.2837,
+      "step": 393500
+    },
+    {
+      "epoch": 8.33,
+      "step": 393602,
+      "total_flos": 1.1093766621573939e+18,
+      "train_loss": 4.550724849134638,
+      "train_runtime": 197998.9321,
+      "train_samples_per_second": 2958.985,
+      "train_steps_per_second": 23.863
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 4724900,
+  "num_train_epochs": 100,
+  "save_steps": 1000000,
+  "total_flos": 1.1093766621573939e+18,
+  "trial_name": null,
+  "trial_params": null
+}