diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6025 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 8.510633952051473,
+  "global_step": 500000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00014999965456413008,
+      "loss": 0.6484,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001499986182599296,
+      "loss": 0.5436,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00014999689109762647,
+      "loss": 0.5149,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00014999447309426713,
+      "loss": 0.4996,
+      "step": 2000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00014999136427371623,
+      "loss": 0.4888,
+      "step": 2500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0001499875646666566,
+      "loss": 0.4797,
+      "step": 3000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00014998307431058887,
+      "loss": 0.4711,
+      "step": 3500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00014997789324983099,
+      "loss": 0.463,
+      "step": 4000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00014997202153551792,
+      "loss": 0.4545,
+      "step": 4500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001499654592256012,
+      "loss": 0.4466,
+      "step": 5000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001499582063848481,
+      "loss": 0.4398,
+      "step": 5500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00014995026308484122,
+      "loss": 0.4337,
+      "step": 6000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00014994162940397778,
+      "loss": 0.4286,
+      "step": 6500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00014993230542746872,
+      "loss": 0.4242,
+      "step": 7000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00014992229124733788,
+      "loss": 0.4203,
+      "step": 7500
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0001499115869624212,
+      "loss": 0.4173,
+      "step": 8000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00014990019267836565,
+      "loss": 0.4141,
+      "step": 8500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00014988810850762824,
+      "loss": 0.4116,
+      "step": 9000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00014987533456947482,
+      "loss": 0.4095,
+      "step": 9500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000149861870989979,
+      "loss": 0.4075,
+      "step": 10000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0001498477758876907,
+      "loss": 0.4057,
+      "step": 10500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00014983293618814342,
+      "loss": 0.4038,
+      "step": 11000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00014981740726570863,
+      "loss": 0.4021,
+      "step": 11500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00014980118927365057,
+      "loss": 0.4005,
+      "step": 12000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00014978428237203426,
+      "loss": 0.3991,
+      "step": 12500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00014976668672772396,
+      "loss": 0.3977,
+      "step": 13000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00014974843976988143,
+      "loss": 0.3968,
+      "step": 13500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00014972946854455738,
+      "loss": 0.3953,
+      "step": 14000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00014970980911752974,
+      "loss": 0.3945,
+      "step": 14500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001496894616828291,
+      "loss": 0.3931,
+      "step": 15000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001496684264412764,
+      "loss": 0.3921,
+      "step": 15500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014964670360048104,
+      "loss": 0.3909,
+      "step": 16000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00014962429337483864,
+      "loss": 0.3903,
+      "step": 16500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00014960119598552907,
+      "loss": 0.3892,
+      "step": 17000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00014957745991457127,
+      "loss": 0.3885,
+      "step": 17500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014955303988623036,
+      "loss": 0.3877,
+      "step": 18000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00014952788514615529,
+      "loss": 0.3871,
+      "step": 18500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0001495020441939234,
+      "loss": 0.3858,
+      "step": 19000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0001494755172845744,
+      "loss": 0.3854,
+      "step": 19500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00014944830467991823,
+      "loss": 0.3847,
+      "step": 20000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00014942046312846905,
+      "loss": 0.3839,
+      "step": 20500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001493918813157191,
+      "loss": 0.3837,
+      "step": 21000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001493626146331152,
+      "loss": 0.383,
+      "step": 21500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00014933266336950772,
+      "loss": 0.3821,
+      "step": 22000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00014930208977431846,
+      "loss": 0.382,
+      "step": 22500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00014927077160993785,
+      "loss": 0.381,
+      "step": 23000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0001492387697710072,
+      "loss": 0.3804,
+      "step": 23500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001492060845733717,
+      "loss": 0.38,
+      "step": 24000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00014917278375754084,
+      "loss": 0.3793,
+      "step": 24500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00014913880296235724,
+      "loss": 0.3791,
+      "step": 25000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00014910407237991193,
+      "loss": 0.3787,
+      "step": 25500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00014906865976817066,
+      "loss": 0.3779,
+      "step": 26000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00014903256547664165,
+      "loss": 0.3778,
+      "step": 26500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0001489958640925119,
+      "loss": 0.377,
+      "step": 27000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00014895840887839445,
+      "loss": 0.3768,
+      "step": 27500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.000148920273072621,
+      "loss": 0.3761,
+      "step": 28000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00014888145705157658,
+      "loss": 0.3759,
+      "step": 28500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0001488420408682806,
+      "loss": 0.3755,
+      "step": 29000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00014880186693119052,
+      "loss": 0.3749,
+      "step": 29500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00014876101394745023,
+      "loss": 0.3747,
+      "step": 30000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00014871948232026222,
+      "loss": 0.3741,
+      "step": 30500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00014867727245952692,
+      "loss": 0.3739,
+      "step": 31000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00014863447123337502,
+      "loss": 0.3736,
+      "step": 31500
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00014859090751637814,
+      "loss": 0.3733,
+      "step": 32000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00014854666683481534,
+      "loss": 0.3728,
+      "step": 32500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001485017496253243,
+      "loss": 0.3725,
+      "step": 33000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00014845624819224238,
+      "loss": 0.3722,
+      "step": 33500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00014840998061432743,
+      "loss": 0.3719,
+      "step": 34000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00014836313241655734,
+      "loss": 0.3717,
+      "step": 34500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014831551629313194,
+      "loss": 0.3711,
+      "step": 35000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014826722592314076,
+      "loss": 0.3709,
+      "step": 35500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014821826178319038,
+      "loss": 0.3709,
+      "step": 36000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001481687243030059,
+      "loss": 0.3703,
+      "step": 36500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00014811841542465117,
+      "loss": 0.3698,
+      "step": 37000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00014806743424503673,
+      "loss": 0.3697,
+      "step": 37500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001480157812673262,
+      "loss": 0.3692,
+      "step": 38000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001479634570013137,
+      "loss": 0.369,
+      "step": 38500
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00014791046196341849,
+      "loss": 0.3689,
+      "step": 39000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001478567966766803,
+      "loss": 0.3685,
+      "step": 39500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001478024616707538,
+      "loss": 0.3683,
+      "step": 40000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00014774756815776608,
+      "loss": 0.368,
+      "step": 40500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00014769189666559521,
+      "loss": 0.3679,
+      "step": 41000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00014763555708173254,
+      "loss": 0.3676,
+      "step": 41500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001475785499622271,
+      "loss": 0.3669,
+      "step": 42000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00014752087586971605,
+      "loss": 0.3673,
+      "step": 42500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0001474626527191014,
+      "loss": 0.3668,
+      "step": 43000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00014740364772589337,
+      "loss": 0.3666,
+      "step": 43500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00014734397748589426,
+      "loss": 0.3665,
+      "step": 44000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0001472836425880252,
+      "loss": 0.366,
+      "step": 44500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00014722264362776731,
+      "loss": 0.3657,
+      "step": 45000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0001471609812071557,
+      "loss": 0.3657,
+      "step": 45500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00014709865593477356,
+      "loss": 0.3653,
+      "step": 46000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00014703566842574615,
+      "loss": 0.3652,
+      "step": 46500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0001469721472598581,
+      "loss": 0.3645,
+      "step": 47000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00014690783847039622,
+      "loss": 0.3647,
+      "step": 47500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.00014684286932758046,
+      "loss": 0.3643,
+      "step": 48000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00014677724047262996,
+      "loss": 0.3644,
+      "step": 48500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0001467110857864263,
+      "loss": 0.3641,
+      "step": 49000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00014664414077306508,
+      "loss": 0.3636,
+      "step": 49500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0001465765380089398,
+      "loss": 0.3635,
+      "step": 50000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.00014650841533628042,
+      "loss": 0.3635,
+      "step": 50500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00014643950039089107,
+      "loss": 0.3631,
+      "step": 51000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001463699297144556,
+      "loss": 0.3628,
+      "step": 51500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014629970399360846,
+      "loss": 0.363,
+      "step": 52000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00014622882392144917,
+      "loss": 0.3624,
+      "step": 52500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.00014615729019753547,
+      "loss": 0.3623,
+      "step": 53000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00014608510352787625,
+      "loss": 0.3623,
+      "step": 53500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001460122646249249,
+      "loss": 0.3621,
+      "step": 54000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00014593892183813694,
+      "loss": 0.3617,
+      "step": 54500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00014586478193255307,
+      "loss": 0.3617,
+      "step": 55000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00014578999196816224,
+      "loss": 0.3616,
+      "step": 55500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00014571455268311119,
+      "loss": 0.361,
+      "step": 56000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00014563861764445836,
+      "loss": 0.3613,
+      "step": 56500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00014556188325305048,
+      "loss": 0.3607,
+      "step": 57000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00014548450179232327,
+      "loss": 0.3606,
+      "step": 57500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00014540647402600055,
+      "loss": 0.3607,
+      "step": 58000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00014532780072418487,
+      "loss": 0.3602,
+      "step": 58500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00014524864194242183,
+      "loss": 0.3607,
+      "step": 59000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00014516868119257174,
+      "loss": 0.3599,
+      "step": 59500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00014508823910339927,
+      "loss": 0.3599,
+      "step": 60000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00014500699405591802,
+      "loss": 0.3599,
+      "step": 60500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00014492510741564985,
+      "loss": 0.3595,
+      "step": 61000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.00014484257999078278,
+      "loss": 0.3593,
+      "step": 61500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00014475941259582925,
+      "loss": 0.3595,
+      "step": 62000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0001446756060516178,
+      "loss": 0.359,
+      "step": 62500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.00014459116118528524,
+      "loss": 0.3589,
+      "step": 63000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00014450607883026827,
+      "loss": 0.3586,
+      "step": 63500
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.00014442053189911774,
+      "loss": 0.3587,
+      "step": 64000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00014433417836295833,
+      "loss": 0.3583,
+      "step": 64500
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014424718987443127,
+      "loss": 0.3583,
+      "step": 65000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00014415956729207775,
+      "loss": 0.3582,
+      "step": 65500
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00014407131148069734,
+      "loss": 0.3579,
+      "step": 66000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00014398260171818916,
+      "loss": 0.3578,
+      "step": 66500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00014389308333022545,
+      "loss": 0.3576,
+      "step": 67000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00014380311527004513,
+      "loss": 0.3577,
+      "step": 67500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00014371233783246623,
+      "loss": 0.3579,
+      "step": 68000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00014362093157983466,
+      "loss": 0.3572,
+      "step": 68500
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0001435288974142932,
+      "loss": 0.3575,
+      "step": 69000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00014343623624418192,
+      "loss": 0.3571,
+      "step": 69500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00014334294898402916,
+      "loss": 0.3567,
+      "step": 70000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0001432490365545425,
+      "loss": 0.3567,
+      "step": 70500
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00014315449988259974,
+      "loss": 0.3566,
+      "step": 71000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001430593399012396,
+      "loss": 0.3564,
+      "step": 71500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00014296374973485408,
+      "loss": 0.3561,
+      "step": 72000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0001428675406248976,
+      "loss": 0.356,
+      "step": 72500
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00014277051885498018,
+      "loss": 0.3562,
+      "step": 73000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00014267287756538312,
+      "loss": 0.3557,
+      "step": 73500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00014257461771978656,
+      "loss": 0.3557,
+      "step": 74000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0001424757402879755,
+      "loss": 0.3556,
+      "step": 74500
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0001423762462458303,
+      "loss": 0.3552,
+      "step": 75000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00014227613657531694,
+      "loss": 0.3552,
+      "step": 75500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.00014217541226447747,
+      "loss": 0.3551,
+      "step": 76000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00014207427759508945,
+      "loss": 0.3552,
+      "step": 76500
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.000141972328216269,
+      "loss": 0.355,
+      "step": 77000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0001418697671955883,
+      "loss": 0.3546,
+      "step": 77500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00014176659554528318,
+      "loss": 0.3546,
+      "step": 78000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00014166302245385192,
+      "loss": 0.3544,
+      "step": 78500
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00014155863382125022,
+      "loss": 0.3545,
+      "step": 79000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00014145363762978493,
+      "loss": 0.3544,
+      "step": 79500
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00014134803491572606,
+      "loss": 0.3542,
+      "step": 80000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.00014124203974129223,
+      "loss": 0.3539,
+      "step": 80500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00014113522832260466,
+      "loss": 0.3541,
+      "step": 81000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00014102781352389485,
+      "loss": 0.3537,
+      "step": 81500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00014091979640530345,
+      "loss": 0.3535,
+      "step": 82000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00014081139586900036,
+      "loss": 0.3533,
+      "step": 82500
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0001407021785141258,
+      "loss": 0.3535,
+      "step": 83000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00014059236205325574,
+      "loss": 0.3536,
+      "step": 83500
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00014048194757023435,
+      "loss": 0.353,
+      "step": 84000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00014037115877265014,
+      "loss": 0.353,
+      "step": 84500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00014025955271103284,
+      "loss": 0.3529,
+      "step": 85000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00014014735191195807,
+      "loss": 0.3528,
+      "step": 85500
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00014003478366336457,
+      "loss": 0.3527,
+      "step": 86000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00013992253468820287,
+      "loss": 0.3559,
+      "step": 86500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00013980856343455598,
+      "loss": 0.3566,
+      "step": 87000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0001396940018945337,
+      "loss": 0.3554,
+      "step": 87500
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00013957885119881218,
+      "loss": 0.3552,
+      "step": 88000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00013946311248388228,
+      "loss": 0.3548,
+      "step": 88500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00013934678689203836,
+      "loss": 0.3547,
+      "step": 89000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0001392298755713671,
+      "loss": 0.3547,
+      "step": 89500
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00013911237967573605,
+      "loss": 0.3541,
+      "step": 90000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0001389943003647822,
+      "loss": 0.3539,
+      "step": 90500
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00013887563880390073,
+      "loss": 0.3536,
+      "step": 91000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0001387563961642333,
+      "loss": 0.3537,
+      "step": 91500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00013863657362265668,
+      "loss": 0.3549,
+      "step": 92000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0001385161723617709,
+      "loss": 0.3539,
+      "step": 92500
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00013839519356988788,
+      "loss": 0.3535,
+      "step": 93000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0001382736384410194,
+      "loss": 0.3537,
+      "step": 93500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00013815150817486554,
+      "loss": 0.3527,
+      "step": 94000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0001380288039768027,
+      "loss": 0.3529,
+      "step": 94500
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00013790552705787178,
+      "loss": 0.353,
+      "step": 95000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0001377816786347662,
+      "loss": 0.3526,
+      "step": 95500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.00013765725992981992,
+      "loss": 0.3527,
+      "step": 96000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00013753227217099532,
+      "loss": 0.3528,
+      "step": 96500
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0001374067165918711,
+      "loss": 0.3524,
+      "step": 97000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00013728059443163013,
+      "loss": 0.352,
+      "step": 97500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00013715390693504724,
+      "loss": 0.3519,
+      "step": 98000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00013702665535247687,
+      "loss": 0.3519,
+      "step": 98500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00013689884093984076,
+      "loss": 0.3519,
+      "step": 99000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00013677072227018223,
+      "loss": 0.3526,
+      "step": 99500
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00013664178710672204,
+      "loss": 0.3517,
+      "step": 100000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00013651229291169033,
+      "loss": 0.3516,
+      "step": 100500
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00013638250162282539,
+      "loss": 0.3516,
+      "step": 101000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00013625189431597695,
+      "loss": 0.3515,
+      "step": 101500
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00013612073182564145,
+      "loss": 0.3512,
+      "step": 102000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00013599007137178395,
+      "loss": 0.3527,
+      "step": 102500
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0001358578068190425,
+      "loss": 0.352,
+      "step": 103000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00013572499097229843,
+      "loss": 0.3512,
+      "step": 103500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00013559162514239058,
+      "loss": 0.3509,
+      "step": 104000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0001354577106455858,
+      "loss": 0.3504,
+      "step": 104500
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00013532324880356615,
+      "loss": 0.3507,
+      "step": 105000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0001351882409434157,
+      "loss": 0.3504,
+      "step": 105500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00013505268839760753,
+      "loss": 0.3501,
+      "step": 106000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00013491659250399054,
+      "loss": 0.3503,
+      "step": 106500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00013477995460577628,
+      "loss": 0.3505,
+      "step": 107000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00013464277605152562,
+      "loss": 0.3502,
+      "step": 107500
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0001345050581951355,
+      "loss": 0.3503,
+      "step": 108000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00013436680239582559,
+      "loss": 0.35,
+      "step": 108500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00013422801001812477,
+      "loss": 0.3506,
+      "step": 109000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00013408868243185783,
+      "loss": 0.3504,
+      "step": 109500
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00013394882101213176,
+      "loss": 0.3497,
+      "step": 110000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00013380842713932233,
+      "loss": 0.3495,
+      "step": 110500
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.00013366750219906035,
+      "loss": 0.3499,
+      "step": 111000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00013352604758221816,
+      "loss": 0.3494,
+      "step": 111500
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.00013338406468489563,
+      "loss": 0.3499,
+      "step": 112000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0001332415549084067,
+      "loss": 0.3496,
+      "step": 112500
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0001330985196592653,
+      "loss": 0.349,
+      "step": 113000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.00013295496034917162,
+      "loss": 0.3491,
+      "step": 113500
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00013281087839499807,
+      "loss": 0.3489,
+      "step": 114000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00013266627521877535,
+      "loss": 0.3485,
+      "step": 114500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0001325211522476785,
+      "loss": 0.3487,
+      "step": 115000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.00013237551091401257,
+      "loss": 0.349,
+      "step": 115500
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0001322299383160388,
+      "loss": 0.3494,
+      "step": 116000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0001320832666336504,
+      "loss": 0.3493,
+      "step": 116500
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00013193637579394162,
+      "loss": 0.3491,
+      "step": 117000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00013178867850631044,
+      "loss": 0.3497,
+      "step": 117500
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00013164047008533037,
+      "loss": 0.3489,
+      "step": 118000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00013149175199375862,
+      "loss": 0.3489,
+      "step": 118500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.00013134252569938276,
+      "loss": 0.3479,
+      "step": 119000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0001311927926750061,
+      "loss": 0.3482,
+      "step": 119500
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00013104255439843306,
+      "loss": 0.3479,
+      "step": 120000
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00013089181235245484,
+      "loss": 0.3476,
+      "step": 120500
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00013074056802483454,
+      "loss": 0.348,
+      "step": 121000
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00013058882290829265,
+      "loss": 0.3481,
+      "step": 121500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.00013043688348660256,
+      "loss": 0.3482,
+      "step": 122000
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.00013028414228420884,
+      "loss": 0.3474,
+      "step": 122500
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0001301309047976313,
+      "loss": 0.3478,
+      "step": 123000
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.00012997717253926207,
+      "loss": 0.3479,
+      "step": 123500
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00012982294702637649,
+      "loss": 0.3473,
+      "step": 124000
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0001296682297811181,
+      "loss": 0.3472,
+      "step": 124500
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.00012951302233048365,
+      "loss": 0.3475,
+      "step": 125000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00012935732620630798,
+      "loss": 0.3473,
+      "step": 125500
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00012920145579691002,
+      "loss": 0.3473,
+      "step": 126000
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.00012904478791008278,
+      "loss": 0.3476,
+      "step": 126500
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.00012888763597099904,
+      "loss": 0.3474,
+      "step": 127000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.000128730001530685,
+      "loss": 0.3473,
+      "step": 127500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00012857188614492896,
+      "loss": 0.3471,
+      "step": 128000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0001284136090411938,
+      "loss": 0.3471,
+      "step": 128500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00012825453740496415,
+      "loss": 0.3467,
+      "step": 129000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00012809498951593147,
+      "loss": 0.3467,
+      "step": 129500
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.000127934966948769,
+      "loss": 0.3466,
+      "step": 130000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00012777447128283487,
+      "loss": 0.3464,
+      "step": 130500
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0001276135041021565,
+      "loss": 0.3462,
+      "step": 131000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00012745239033755645,
+      "loss": 0.3467,
+      "step": 131500
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0001272908101084899,
+      "loss": 0.3464,
+      "step": 132000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0001271284397979516,
+      "loss": 0.3463,
+      "step": 132500
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.000126965604348739,
+      "loss": 0.3463,
+      "step": 133000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00012680230536797228,
+      "loss": 0.3461,
+      "step": 133500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0001266385444673464,
+      "loss": 0.3458,
+      "step": 134000
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00012647432326311537,
+      "loss": 0.3459,
+      "step": 134500
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00012630964337607617,
+      "loss": 0.3459,
+      "step": 135000
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.00012614483716050272,
+      "loss": 0.3456,
+      "step": 135500
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0001259792456975559,
+      "loss": 0.3459,
+      "step": 136000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00012581320043801664,
+      "loss": 0.3456,
+      "step": 136500
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00012564670302068462,
+      "loss": 0.3457,
+      "step": 137000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00012547975508882204,
+      "loss": 0.3452,
+      "step": 137500
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00012531235829013772,
+      "loss": 0.3452,
+      "step": 138000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00012514451427677039,
+      "loss": 0.3451,
+      "step": 138500
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.0001249762247052727,
+      "loss": 0.3446,
+      "step": 139000
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00012480749123659483,
+      "loss": 0.3445,
+      "step": 139500
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.00012463865432770735,
+      "loss": 0.3455,
+      "step": 140000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00012446903894448242,
+      "loss": 0.345,
+      "step": 140500
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00012429898466979599,
+      "loss": 0.3447,
+      "step": 141000
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0001241284931820151,
+      "loss": 0.3445,
+      "step": 141500
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.00012395756616382189,
+      "loss": 0.3446,
+      "step": 142000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0001237865484557726,
+      "loss": 0.3444,
+      "step": 142500
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00012361475630459222,
+      "loss": 0.3441,
+      "step": 143000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0001234428785670676,
+      "loss": 0.3443,
+      "step": 143500
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00012327022805140274,
+      "loss": 0.3438,
+      "step": 144000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00012309715047605592,
+      "loss": 0.3437,
+      "step": 144500
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.00012292364754923295,
+      "loss": 0.3444,
+      "step": 145000
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0001227497209833376,
+      "loss": 0.3441,
+      "step": 145500
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.00012257572161186733,
+      "loss": 0.3439,
+      "step": 146000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.00012240095376042992,
+      "loss": 0.3436,
+      "step": 146500
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00012222576742869685,
+      "loss": 0.3434,
+      "step": 147000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.00012205016434568649,
+      "loss": 0.3434,
+      "step": 147500
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00012187414624453037,
+      "loss": 0.3433,
+      "step": 148000
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00012169771486245609,
+      "loss": 0.3429,
+      "step": 148500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00012152087194077019,
+      "loss": 0.3429,
+      "step": 149000
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00012134361922484087,
+      "loss": 0.3435,
+      "step": 149500
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.0001211659584640809,
+      "loss": 0.3427,
+      "step": 150000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00012098789141193026,
+      "loss": 0.3423,
+      "step": 150500
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00012080941982583888,
+      "loss": 0.3425,
+      "step": 151000
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.00012063054546724922,
+      "loss": 0.3425,
+      "step": 151500
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00012045127010157905,
+      "loss": 0.3426,
+      "step": 152000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00012027195524467175,
+      "loss": 0.3422,
+      "step": 152500
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00012009188397006361,
+      "loss": 0.342,
+      "step": 153000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00011991141700474647,
+      "loss": 0.3422,
+      "step": 153500
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00011973055612985635,
+      "loss": 0.3417,
+      "step": 154000
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00011954930313041712,
+      "loss": 0.3416,
+      "step": 154500
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.00011936802347035692,
+      "loss": 0.3417,
+      "step": 155000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00011918599236764778,
+      "loss": 0.3415,
+      "step": 155500
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00011900357451501401,
+      "loss": 0.3417,
+      "step": 156000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00011882077171284612,
+      "loss": 0.3421,
+      "step": 156500
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.00011863795251840704,
+      "loss": 0.3414,
+      "step": 157000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00011845512101771738,
+      "loss": 0.3419,
+      "step": 157500
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.00011827117647893832,
+      "loss": 0.3414,
+      "step": 158000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.0001180868542190981,
+      "loss": 0.3423,
+      "step": 158500
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00011790215605738305,
+      "loss": 0.3417,
+      "step": 159000
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00011771708381668941,
+      "loss": 0.3412,
+      "step": 159500
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00011753163932360551,
+      "loss": 0.3408,
+      "step": 160000
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.00011734582440839364,
+      "loss": 0.3411,
+      "step": 160500
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00011715964090497195,
+      "loss": 0.3406,
+      "step": 161000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.00011697309065089649,
+      "loss": 0.3412,
+      "step": 161500
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.00011678654968062415,
+      "loss": 0.3411,
+      "step": 162000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00011659964709277792,
+      "loss": 0.3414,
+      "step": 162500
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00011641200908936181,
+      "loss": 0.3414,
+      "step": 163000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00011622401171411614,
+      "loss": 0.3412,
+      "step": 163500
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.00011603565682249906,
+      "loss": 0.3411,
+      "step": 164000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00011584732404830404,
+      "loss": 0.3408,
+      "step": 164500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0001156586388892709,
+      "loss": 0.3405,
+      "step": 165000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.00011546922402047642,
+      "loss": 0.3403,
+      "step": 165500
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.00011527945908476035,
+      "loss": 0.3409,
+      "step": 166000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.000115089345955026,
+      "loss": 0.3406,
+      "step": 166500
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00011489926777088424,
+      "loss": 0.3406,
+      "step": 167000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00011470846457254845,
+      "loss": 0.3403,
+      "step": 167500
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.00011451731881567971,
+      "loss": 0.3402,
+      "step": 168000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.00011432583238680949,
+      "loss": 0.3405,
+      "step": 168500
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00011413400717583146,
+      "loss": 0.3401,
+      "step": 169000
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00011394184507598311,
+      "loss": 0.3398,
+      "step": 169500
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00011374934798382673,
+      "loss": 0.3396,
+      "step": 170000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00011355651779923088,
+      "loss": 0.3393,
+      "step": 170500
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00011336335642535168,
+      "loss": 0.3391,
+      "step": 171000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.00011316986576861393,
+      "loss": 0.3394,
+      "step": 171500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.00011297604773869221,
+      "loss": 0.3395,
+      "step": 172000
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00011278229285900902,
+      "loss": 0.3397,
+      "step": 172500
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.00011258782646982266,
+      "loss": 0.3396,
+      "step": 173000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.0001123930384519453,
+      "loss": 0.3392,
+      "step": 173500
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00011219793072785603,
+      "loss": 0.3394,
+      "step": 174000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00011200250522318928,
+      "loss": 0.3391,
+      "step": 174500
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.00011180676386671593,
+      "loss": 0.3389,
+      "step": 175000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00011161070859032411,
+      "loss": 0.3387,
+      "step": 175500
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.00011141434132900025,
+      "loss": 0.3385,
+      "step": 176000
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.00011121805768356343,
+      "loss": 0.339,
+      "step": 176500
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00011102107288390457,
+      "loss": 0.3386,
+      "step": 177000
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00011082378191878007,
+      "loss": 0.3383,
+      "step": 177500
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.000110626582228053,
+      "loss": 0.3385,
+      "step": 178000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00011042908147320174,
+      "loss": 0.3391,
+      "step": 178500
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00011023128159568977,
+      "loss": 0.339,
+      "step": 179000
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00011003278725476702,
+      "loss": 0.3386,
+      "step": 179500
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00010983399650235882,
+      "loss": 0.3384,
+      "step": 180000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00010963491130044959,
+      "loss": 0.3385,
+      "step": 180500
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00010943553361392991,
+      "loss": 0.339,
+      "step": 181000
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00010923586541057708,
+      "loss": 0.3388,
+      "step": 181500
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00010903590866103563,
+      "loss": 0.3386,
+      "step": 182000
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00010883606611012889,
+      "loss": 0.3387,
+      "step": 182500
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00010863553875873429,
+      "loss": 0.3383,
+      "step": 183000
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00010843472878613276,
+      "loss": 0.3387,
+      "step": 183500
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.0001082336381742377,
+      "loss": 0.3381,
+      "step": 184000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00010803226890773223,
+      "loss": 0.3381,
+      "step": 184500
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00010783062297404974,
+      "loss": 0.338,
+      "step": 185000
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00010762870236335414,
+      "loss": 0.3379,
+      "step": 185500
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.0001074265090685204,
+      "loss": 0.338,
+      "step": 186000
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00010722404508511468,
+      "loss": 0.3377,
+      "step": 186500
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00010702131241137474,
+      "loss": 0.3376,
+      "step": 187000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00010681831304819025,
+      "loss": 0.3378,
+      "step": 187500
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00010661504899908298,
+      "loss": 0.3377,
+      "step": 188000
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00010641192958446453,
+      "loss": 0.3384,
+      "step": 188500
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00010620814270384223,
+      "loss": 0.3378,
+      "step": 189000
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00010600450550732924,
+      "loss": 0.3378,
+      "step": 189500
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00010580020382426679,
+      "loss": 0.3376,
+      "step": 190000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00010559564750360856,
+      "loss": 0.3375,
+      "step": 190500
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00010539083856424286,
+      "loss": 0.3372,
+      "step": 191000
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.00010518618939537595,
+      "loss": 0.337,
+      "step": 191500
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00010498088178033872,
+      "loss": 0.3367,
+      "step": 192000
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.0001047753276140835,
+      "loss": 0.3368,
+      "step": 192500
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00010456952892534694,
+      "loss": 0.3367,
+      "step": 193000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00010436348774527898,
+      "loss": 0.3364,
+      "step": 193500
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00010415720610742296,
+      "loss": 0.3366,
+      "step": 194000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00010395068604769528,
+      "loss": 0.3365,
+      "step": 194500
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00010374392960436561,
+      "loss": 0.3369,
+      "step": 195000
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00010353693881803657,
+      "loss": 0.3364,
+      "step": 195500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00010332971573162363,
+      "loss": 0.3363,
+      "step": 196000
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00010312226239033498,
+      "loss": 0.336,
+      "step": 196500
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.0001029145808416514,
+      "loss": 0.336,
+      "step": 197000
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.00010270667313530591,
+      "loss": 0.3357,
+      "step": 197500
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.0001024985413232636,
+      "loss": 0.3357,
+      "step": 198000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.0001022901874597015,
+      "loss": 0.3354,
+      "step": 198500
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00010208161360098805,
+      "loss": 0.3355,
+      "step": 199000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.00010187282180566307,
+      "loss": 0.3354,
+      "step": 199500
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.00010166506882009926,
+      "loss": 0.3356,
+      "step": 200000
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.00010145626726491797,
+      "loss": 0.3363,
+      "step": 200500
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00010124683571819103,
+      "loss": 0.3363,
+      "step": 201000
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.00010103719447377321,
+      "loss": 0.3363,
+      "step": 201500
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00010082734560073903,
+      "loss": 0.3358,
+      "step": 202000
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00010061729117021209,
+      "loss": 0.336,
+      "step": 202500
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 0.00010040703325534484,
+      "loss": 0.336,
+      "step": 203000
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 0.00010019657393129802,
+      "loss": 0.3359,
+      "step": 203500
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.998591527522016e-05,
+      "loss": 0.3357,
+      "step": 204000
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.977505936622718e-05,
+      "loss": 0.3356,
+      "step": 204500
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.956400828538178e-05,
+      "loss": 0.3358,
+      "step": 205000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.935318679532955e-05,
+      "loss": 0.3353,
+      "step": 205500
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.914217506040476e-05,
+      "loss": 0.3356,
+      "step": 206000
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.893055172104751e-05,
+      "loss": 0.3354,
+      "step": 206500
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.871874154478737e-05,
+      "loss": 0.3352,
+      "step": 207000
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.850674662210523e-05,
+      "loss": 0.3349,
+      "step": 207500
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.829456904530539e-05,
+      "loss": 0.3349,
+      "step": 208000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.808263580357464e-05,
+      "loss": 0.3351,
+      "step": 208500
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.787052480668757e-05,
+      "loss": 0.3352,
+      "step": 209000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.76578125405694e-05,
+      "loss": 0.3355,
+      "step": 209500
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.744535194458411e-05,
+      "loss": 0.3354,
+      "step": 210000
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.723229357085885e-05,
+      "loss": 0.3354,
+      "step": 210500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.701906512135297e-05,
+      "loss": 0.3355,
+      "step": 211000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.680566870054522e-05,
+      "loss": 0.3348,
+      "step": 211500
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.659210641457217e-05,
+      "loss": 0.3349,
+      "step": 212000
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.637838037120726e-05,
+      "loss": 0.3346,
+      "step": 212500
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.616449267984031e-05,
+      "loss": 0.3346,
+      "step": 213000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.59504454514565e-05,
+      "loss": 0.3343,
+      "step": 213500
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.573666936362591e-05,
+      "loss": 0.3341,
+      "step": 214000
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.5522309708951e-05,
+      "loss": 0.3344,
+      "step": 214500
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 9.53077968553459e-05,
+      "loss": 0.3339,
+      "step": 215000
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 9.509313291996582e-05,
+      "loss": 0.3342,
+      "step": 215500
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 9.487832002145718e-05,
+      "loss": 0.334,
+      "step": 216000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.466336027993655e-05,
+      "loss": 0.3339,
+      "step": 216500
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.44482558169698e-05,
+      "loss": 0.3339,
+      "step": 217000
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 9.423300875555111e-05,
+      "loss": 0.3339,
+      "step": 217500
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 9.401762122008212e-05,
+      "loss": 0.3334,
+      "step": 218000
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 9.380209533635084e-05,
+      "loss": 0.3341,
+      "step": 218500
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 9.35864332315107e-05,
+      "loss": 0.3341,
+      "step": 219000
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.337063703405964e-05,
+      "loss": 0.3338,
+      "step": 219500
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.315514086042207e-05,
+      "loss": 0.3338,
+      "step": 220000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 9.293994761279615e-05,
+      "loss": 0.3336,
+      "step": 220500
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.272376268146326e-05,
+      "loss": 0.3336,
+      "step": 221000
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 9.250745217172946e-05,
+      "loss": 0.3336,
+      "step": 221500
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 9.229101821849224e-05,
+      "loss": 0.3334,
+      "step": 222000
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 9.207446295786727e-05,
+      "loss": 0.3335,
+      "step": 222500
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.185778852716756e-05,
+      "loss": 0.3331,
+      "step": 223000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.164099706488229e-05,
+      "loss": 0.3333,
+      "step": 223500
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.14245246366035e-05,
+      "loss": 0.3334,
+      "step": 224000
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 9.12075057545787e-05,
+      "loss": 0.3332,
+      "step": 224500
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.099037625899677e-05,
+      "loss": 0.3333,
+      "step": 225000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 9.077313829283815e-05,
+      "loss": 0.3331,
+      "step": 225500
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 9.055579400015386e-05,
+      "loss": 0.3331,
+      "step": 226000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.033834552604434e-05,
+      "loss": 0.3331,
+      "step": 226500
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.012079501663827e-05,
+      "loss": 0.3329,
+      "step": 227000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 8.990314461907129e-05,
+      "loss": 0.3329,
+      "step": 227500
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 8.968583207385645e-05,
+      "loss": 0.3324,
+      "step": 228000
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 8.946798853433352e-05,
+      "loss": 0.3328,
+      "step": 228500
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 8.925005154958616e-05,
+      "loss": 0.3326,
+      "step": 229000
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 8.903202327056435e-05,
+      "loss": 0.3324,
+      "step": 229500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 8.881390584911924e-05,
+      "loss": 0.3322,
+      "step": 230000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 8.859570143798174e-05,
+      "loss": 0.3325,
+      "step": 230500
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 8.837741219074127e-05,
+      "loss": 0.3324,
+      "step": 231000
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 8.815904026182456e-05,
+      "loss": 0.3322,
+      "step": 231500
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 8.794058780647441e-05,
+      "loss": 0.3319,
+      "step": 232000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 8.772205698072836e-05,
+      "loss": 0.3324,
+      "step": 232500
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 8.750432451851294e-05,
+      "loss": 0.3322,
+      "step": 233000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 8.728564371508559e-05,
+      "loss": 0.332,
+      "step": 233500
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 8.706732858104335e-05,
+      "loss": 0.3322,
+      "step": 234000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 8.684850626123223e-05,
+      "loss": 0.3323,
+      "step": 234500
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 8.66296163494292e-05,
+      "loss": 0.3321,
+      "step": 235000
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 8.641066100598926e-05,
+      "loss": 0.332,
+      "step": 235500
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 8.619164239191329e-05,
+      "loss": 0.3319,
+      "step": 236000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 8.59725626688266e-05,
+      "loss": 0.3319,
+      "step": 236500
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 8.575342399895754e-05,
+      "loss": 0.3315,
+      "step": 237000
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.55342285451164e-05,
+      "loss": 0.3316,
+      "step": 237500
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.531497847067375e-05,
+      "loss": 0.3317,
+      "step": 238000
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.509567593953936e-05,
+      "loss": 0.3314,
+      "step": 238500
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.487632311614066e-05,
+      "loss": 0.3315,
+      "step": 239000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.465692216540142e-05,
+      "loss": 0.3315,
+      "step": 239500
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.443791419097638e-05,
+      "loss": 0.3312,
+      "step": 240000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.421842356763658e-05,
+      "loss": 0.3314,
+      "step": 240500
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.399889131015923e-05,
+      "loss": 0.331,
+      "step": 241000
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.377931958523902e-05,
+      "loss": 0.3312,
+      "step": 241500
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.356014981379603e-05,
+      "loss": 0.3313,
+      "step": 242000
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.334050572371366e-05,
+      "loss": 0.3309,
+      "step": 242500
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.31208286641884e-05,
+      "loss": 0.3309,
+      "step": 243000
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.290112080334417e-05,
+      "loss": 0.3309,
+      "step": 243500
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.268138430960883e-05,
+      "loss": 0.3308,
+      "step": 244000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 8.246206090257837e-05,
+      "loss": 0.3308,
+      "step": 244500
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.224271329310167e-05,
+      "loss": 0.3306,
+      "step": 245000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.20229039981619e-05,
+      "loss": 0.3306,
+      "step": 245500
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.180307473797635e-05,
+      "loss": 0.3308,
+      "step": 246000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.158322768217107e-05,
+      "loss": 0.3303,
+      "step": 246500
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.136336500054776e-05,
+      "loss": 0.3302,
+      "step": 247000
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 8.114348886306229e-05,
+      "loss": 0.3305,
+      "step": 247500
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 8.092360143980338e-05,
+      "loss": 0.3302,
+      "step": 248000
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 8.070414470170364e-05,
+      "loss": 0.3305,
+      "step": 248500
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 8.048424122931269e-05,
+      "loss": 0.3303,
+      "step": 249000
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 8.02643329776563e-05,
+      "loss": 0.3301,
+      "step": 249500
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 8.004442211714013e-05,
+      "loss": 0.3298,
+      "step": 250000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 7.982451081819563e-05,
+      "loss": 0.33,
+      "step": 250500
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 7.960460125125849e-05,
+      "loss": 0.33,
+      "step": 251000
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 7.938469558674736e-05,
+      "loss": 0.3302,
+      "step": 251500
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 7.916479599504238e-05,
+      "loss": 0.3301,
+      "step": 252000
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 7.894534441949177e-05,
+      "loss": 0.3301,
+      "step": 252500
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 7.872546346128575e-05,
+      "loss": 0.3303,
+      "step": 253000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 7.850603480500256e-05,
+      "loss": 0.33,
+      "step": 253500
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 7.828618114346693e-05,
+      "loss": 0.3299,
+      "step": 254000
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 7.806634439663151e-05,
+      "loss": 0.3301,
+      "step": 254500
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 7.784652673419626e-05,
+      "loss": 0.33,
+      "step": 255000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 7.762673032567275e-05,
+      "loss": 0.3298,
+      "step": 255500
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 7.74069573403628e-05,
+      "loss": 0.33,
+      "step": 256000
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 7.718720994733705e-05,
+      "loss": 0.3297,
+      "step": 256500
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 7.696749031541353e-05,
+      "loss": 0.3298,
+      "step": 257000
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 7.674823996122986e-05,
+      "loss": 0.3296,
+      "step": 257500
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 7.652858229048815e-05,
+      "loss": 0.3297,
+      "step": 258000
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 7.630895888123776e-05,
+      "loss": 0.3294,
+      "step": 258500
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 7.60893719010731e-05,
+      "loss": 0.3295,
+      "step": 259000
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 7.587026257403724e-05,
+      "loss": 0.3292,
+      "step": 259500
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.565119384297032e-05,
+      "loss": 0.3292,
+      "step": 260000
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 7.54317289758932e-05,
+      "loss": 0.3294,
+      "step": 260500
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 7.521230919580681e-05,
+      "loss": 0.329,
+      "step": 261000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 7.499293666829577e-05,
+      "loss": 0.3291,
+      "step": 261500
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 7.47736135584784e-05,
+      "loss": 0.3289,
+      "step": 262000
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 7.455478052112257e-05,
+      "loss": 0.3292,
+      "step": 262500
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 7.43355626304227e-05,
+      "loss": 0.329,
+      "step": 263000
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 7.411640064543287e-05,
+      "loss": 0.3294,
+      "step": 263500
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 7.389773487763577e-05,
+      "loss": 0.329,
+      "step": 264000
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 7.367869106999357e-05,
+      "loss": 0.3288,
+      "step": 264500
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 7.345970965111851e-05,
+      "loss": 0.3289,
+      "step": 265000
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 7.324079278226878e-05,
+      "loss": 0.3287,
+      "step": 265500
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 7.302194262406548e-05,
+      "loss": 0.3288,
+      "step": 266000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 7.280316133647132e-05,
+      "loss": 0.3288,
+      "step": 266500
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 7.258488842696201e-05,
+      "loss": 0.3286,
+      "step": 267000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 7.236625120920332e-05,
+      "loss": 0.3288,
+      "step": 267500
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 7.214768933346332e-05,
+      "loss": 0.3288,
+      "step": 268000
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 7.192920495685949e-05,
+      "loss": 0.3285,
+      "step": 268500
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 7.17112369642576e-05,
+      "loss": 0.3284,
+      "step": 269000
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 7.149335045150111e-05,
+      "loss": 0.3286,
+      "step": 269500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 7.127511116710097e-05,
+      "loss": 0.3286,
+      "step": 270000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 7.10573942128076e-05,
+      "loss": 0.3286,
+      "step": 270500
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 7.083932912507537e-05,
+      "loss": 0.3284,
+      "step": 271000
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.062135444946636e-05,
+      "loss": 0.3283,
+      "step": 271500
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.040347233730262e-05,
+      "loss": 0.3281,
+      "step": 272000
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 7.018568493899267e-05,
+      "loss": 0.3284,
+      "step": 272500
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 6.996799440401013e-05,
+      "loss": 0.3281,
+      "step": 273000
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 6.975040288087275e-05,
+      "loss": 0.3281,
+      "step": 273500
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 6.9532912517121e-05,
+      "loss": 0.3281,
+      "step": 274000
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 6.931552545929699e-05,
+      "loss": 0.3279,
+      "step": 274500
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 6.909824385292317e-05,
+      "loss": 0.3281,
+      "step": 275000
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 6.888106984248133e-05,
+      "loss": 0.3279,
+      "step": 275500
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 6.866400557139124e-05,
+      "loss": 0.328,
+      "step": 276000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 6.844705318198965e-05,
+      "loss": 0.3276,
+      "step": 276500
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 6.823064837702448e-05,
+      "loss": 0.3279,
+      "step": 277000
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 6.801392593911085e-05,
+      "loss": 0.3278,
+      "step": 277500
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 6.779732179890923e-05,
+      "loss": 0.3277,
+      "step": 278000
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 6.758083809421504e-05,
+      "loss": 0.3277,
+      "step": 278500
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 6.736447696163502e-05,
+      "loss": 0.3276,
+      "step": 279000
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 6.714824053656624e-05,
+      "loss": 0.3273,
+      "step": 279500
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 6.693213095317489e-05,
+      "loss": 0.3274,
+      "step": 280000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 6.671615034437533e-05,
+      "loss": 0.3272,
+      "step": 280500
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 6.650030084180895e-05,
+      "loss": 0.3272,
+      "step": 281000
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 6.628458457582322e-05,
+      "loss": 0.3273,
+      "step": 281500
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 6.606900367545062e-05,
+      "loss": 0.3272,
+      "step": 282000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 6.585356026838752e-05,
+      "loss": 0.327,
+      "step": 282500
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 6.563825648097345e-05,
+      "loss": 0.3268,
+      "step": 283000
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 6.542309443816984e-05,
+      "loss": 0.3271,
+      "step": 283500
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 6.520807626353919e-05,
+      "loss": 0.3268,
+      "step": 284000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 6.499363367648454e-05,
+      "loss": 0.3268,
+      "step": 284500
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 6.477890930484919e-05,
+      "loss": 0.3269,
+      "step": 285000
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 6.456433515923394e-05,
+      "loss": 0.3271,
+      "step": 285500
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 6.434991335739906e-05,
+      "loss": 0.3268,
+      "step": 286000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 6.413564601560114e-05,
+      "loss": 0.3267,
+      "step": 286500
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 6.392196331243996e-05,
+      "loss": 0.3266,
+      "step": 287000
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 6.370801091388329e-05,
+      "loss": 0.3265,
+      "step": 287500
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 6.349421931068155e-05,
+      "loss": 0.3265,
+      "step": 288000
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 6.328059061287157e-05,
+      "loss": 0.3263,
+      "step": 288500
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 6.306712692888235e-05,
+      "loss": 0.3263,
+      "step": 289000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 6.285383036551425e-05,
+      "loss": 0.3262,
+      "step": 289500
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 6.264070302791827e-05,
+      "loss": 0.3263,
+      "step": 290000
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 6.242774701957516e-05,
+      "loss": 0.3263,
+      "step": 290500
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 6.221496444227476e-05,
+      "loss": 0.3261,
+      "step": 291000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 6.200235739609522e-05,
+      "loss": 0.3261,
+      "step": 291500
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 6.179035265954763e-05,
+      "loss": 0.326,
+      "step": 292000
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 6.157810260735043e-05,
+      "loss": 0.3261,
+      "step": 292500
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 6.136603437184328e-05,
+      "loss": 0.3259,
+      "step": 293000
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 6.115457362977345e-05,
+      "loss": 0.3258,
+      "step": 293500
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 6.0942874930827204e-05,
+      "loss": 0.326,
+      "step": 294000
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 6.0731364318011724e-05,
+      "loss": 0.3256,
+      "step": 294500
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 6.052004387885136e-05,
+      "loss": 0.3258,
+      "step": 295000
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 6.030891569899353e-05,
+      "loss": 0.3259,
+      "step": 295500
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 6.009798186218815e-05,
+      "loss": 0.3256,
+      "step": 296000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 5.988724445026701e-05,
+      "loss": 0.3255,
+      "step": 296500
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 5.96767055431233e-05,
+      "loss": 0.3255,
+      "step": 297000
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 5.946636721869104e-05,
+      "loss": 0.3255,
+      "step": 297500
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 5.925665162062394e-05,
+      "loss": 0.3254,
+      "step": 298000
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 5.904672027594373e-05,
+      "loss": 0.3254,
+      "step": 298500
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 5.8836995731675224e-05,
+      "loss": 0.3254,
+      "step": 299000
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 5.862748005771498e-05,
+      "loss": 0.3254,
+      "step": 299500
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 5.841817532189814e-05,
+      "loss": 0.3252,
+      "step": 300000
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 5.820950155949188e-05,
+      "loss": 0.3256,
+      "step": 300500
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 5.80006244629261e-05,
+      "loss": 0.3257,
+      "step": 301000
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 5.779196449131535e-05,
+      "loss": 0.3254,
+      "step": 301500
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 5.758352370404932e-05,
+      "loss": 0.3253,
+      "step": 302000
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 5.7375304158354426e-05,
+      "loss": 0.3251,
+      "step": 302500
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 5.716730790927354e-05,
+      "loss": 0.3251,
+      "step": 303000
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 5.695953700964566e-05,
+      "loss": 0.3251,
+      "step": 303500
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 5.675199351008569e-05,
+      "loss": 0.3253,
+      "step": 304000
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 5.654550825539481e-05,
+      "loss": 0.3253,
+      "step": 304500
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 5.633842476876674e-05,
+      "loss": 0.3248,
+      "step": 305000
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 5.6131574812333625e-05,
+      "loss": 0.3251,
+      "step": 305500
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 5.592496042762104e-05,
+      "loss": 0.3251,
+      "step": 306000
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 5.571858365382955e-05,
+      "loss": 0.3251,
+      "step": 306500
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 5.5512446527814595e-05,
+      "loss": 0.325,
+      "step": 307000
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 5.530655108406638e-05,
+      "loss": 0.3248,
+      "step": 307500
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 5.5100899354689826e-05,
+      "loss": 0.3249,
+      "step": 308000
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 5.4895903934754626e-05,
+      "loss": 0.3249,
+      "step": 308500
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 5.469074522323032e-05,
+      "loss": 0.3249,
+      "step": 309000
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 5.448624587103454e-05,
+      "loss": 0.325,
+      "step": 309500
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 5.428158826048664e-05,
+      "loss": 0.3249,
+      "step": 310000
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 5.407718448027968e-05,
+      "loss": 0.3248,
+      "step": 310500
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 5.387303654779643e-05,
+      "loss": 0.3245,
+      "step": 311000
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 5.3669146477894576e-05,
+      "loss": 0.3245,
+      "step": 311500
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 5.3465516282886794e-05,
+      "loss": 0.3245,
+      "step": 312000
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 5.32621479725209e-05,
+      "loss": 0.3245,
+      "step": 312500
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 5.305904355396001e-05,
+      "loss": 0.3244,
+      "step": 313000
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 5.285620503176271e-05,
+      "loss": 0.3243,
+      "step": 313500
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 5.265363440786338e-05,
+      "loss": 0.3244,
+      "step": 314000
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 5.2451333681552255e-05,
+      "loss": 0.3241,
+      "step": 314500
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 5.2249304849455853e-05,
+      "loss": 0.3243,
+      "step": 315000
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 5.204754990551717e-05,
+      "loss": 0.3244,
+      "step": 315500
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 5.184647352245537e-05,
+      "loss": 0.3241,
+      "step": 316000
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 5.164527176811031e-05,
+      "loss": 0.3245,
+      "step": 316500
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 5.144434986348564e-05,
+      "loss": 0.3242,
+      "step": 317000
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 5.1243709791599435e-05,
+      "loss": 0.3243,
+      "step": 317500
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 5.1043753960645855e-05,
+      "loss": 0.324,
+      "step": 318000
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 5.0843682918592574e-05,
+      "loss": 0.3241,
+      "step": 318500
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 5.064389963761796e-05,
+      "loss": 0.3242,
+      "step": 319000
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 5.044440608950243e-05,
+      "loss": 0.3239,
+      "step": 319500
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 5.024520424316668e-05,
+      "loss": 0.3239,
+      "step": 320000
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 5.004629606465258e-05,
+      "loss": 0.3239,
+      "step": 320500
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 4.9847683517103505e-05,
+      "loss": 0.324,
+      "step": 321000
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 4.9649368560745154e-05,
+      "loss": 0.3238,
+      "step": 321500
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 4.945135315286604e-05,
+      "loss": 0.3238,
+      "step": 322000
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 4.925363924779834e-05,
+      "loss": 0.3239,
+      "step": 322500
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 4.905622879689845e-05,
+      "loss": 0.3238,
+      "step": 323000
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 4.8859517652539344e-05,
+      "loss": 0.3234,
+      "step": 323500
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 4.8663112624023495e-05,
+      "loss": 0.3238,
+      "step": 324000
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 4.846662297269393e-05,
+      "loss": 0.3235,
+      "step": 324500
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 4.827044454306512e-05,
+      "loss": 0.3237,
+      "step": 325000
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 4.807457927133891e-05,
+      "loss": 0.3234,
+      "step": 325500
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 4.7879419875243765e-05,
+      "loss": 0.3236,
+      "step": 326000
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 4.768418607958038e-05,
+      "loss": 0.3236,
+      "step": 326500
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 4.748927122795397e-05,
+      "loss": 0.3234,
+      "step": 327000
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 4.729467724409539e-05,
+      "loss": 0.3233,
+      "step": 327500
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 4.710040604856878e-05,
+      "loss": 0.3236,
+      "step": 328000
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 4.690723469535981e-05,
+      "loss": 0.3235,
+      "step": 328500
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 4.6713613515138837e-05,
+      "loss": 0.3234,
+      "step": 329000
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 4.652032085811418e-05,
+      "loss": 0.3234,
+      "step": 329500
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 4.63273586320064e-05,
+      "loss": 0.3235,
+      "step": 330000
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 4.613472874127472e-05,
+      "loss": 0.3233,
+      "step": 330500
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 4.594243308709837e-05,
+      "loss": 0.3231,
+      "step": 331000
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 4.575047356735788e-05,
+      "loss": 0.3231,
+      "step": 331500
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 4.5558852076616174e-05,
+      "loss": 0.3234,
+      "step": 332000
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 4.536757050610003e-05,
+      "loss": 0.3231,
+      "step": 332500
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 4.5176630743681286e-05,
+      "loss": 0.323,
+      "step": 333000
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 4.498603467385838e-05,
+      "loss": 0.3229,
+      "step": 333500
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 4.4795784177737496e-05,
+      "loss": 0.323,
+      "step": 334000
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 4.460588113301429e-05,
+      "loss": 0.3232,
+      "step": 334500
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 4.4416327413955116e-05,
+      "loss": 0.3231,
+      "step": 335000
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 4.4227502944687974e-05,
+      "loss": 0.3231,
+      "step": 335500
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 4.403865277795774e-05,
+      "loss": 0.3229,
+      "step": 336000
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 4.385015753520643e-05,
+      "loss": 0.3231,
+      "step": 336500
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 4.366201907680594e-05,
+      "loss": 0.3229,
+      "step": 337000
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 4.3474614460084984e-05,
+      "loss": 0.3228,
+      "step": 337500
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 4.328719441456219e-05,
+      "loss": 0.3227,
+      "step": 338000
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 4.310013670960897e-05,
+      "loss": 0.323,
+      "step": 338500
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 4.291344319140932e-05,
+      "loss": 0.323,
+      "step": 339000
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 4.272711570255294e-05,
+      "loss": 0.3229,
+      "step": 339500
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 4.254115608201689e-05,
+      "loss": 0.3228,
+      "step": 340000
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 4.2355566165147584e-05,
+      "loss": 0.3225,
+      "step": 340500
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 4.217034778364253e-05,
+      "loss": 0.3224,
+      "step": 341000
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 4.1985872081738785e-05,
+      "loss": 0.3225,
+      "step": 341500
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 4.1801770064696236e-05,
+      "loss": 0.3226,
+      "step": 342000
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 4.1617675731054795e-05,
+      "loss": 0.3225,
+      "step": 342500
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 4.143396021545834e-05,
+      "loss": 0.3226,
+      "step": 343000
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 4.12506253311048e-05,
+      "loss": 0.3224,
+      "step": 343500
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 4.1067672887435514e-05,
+      "loss": 0.3228,
+      "step": 344000
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 4.0885104690117184e-05,
+      "loss": 0.3225,
+      "step": 344500
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 4.0702922541024256e-05,
+      "loss": 0.3223,
+      "step": 345000
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 4.052149143856278e-05,
+      "loss": 0.3225,
+      "step": 345500
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 4.034008599521639e-05,
+      "loss": 0.3224,
+      "step": 346000
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 4.015907197920985e-05,
+      "loss": 0.3225,
+      "step": 346500
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 3.997881202506971e-05,
+      "loss": 0.3221,
+      "step": 347000
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 3.979894548550711e-05,
+      "loss": 0.3225,
+      "step": 347500
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 3.961911486455148e-05,
+      "loss": 0.3226,
+      "step": 348000
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 3.9439682786629685e-05,
+      "loss": 0.3224,
+      "step": 348500
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 3.9260651022663946e-05,
+      "loss": 0.3223,
+      "step": 349000
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 3.908202133962548e-05,
+      "loss": 0.3223,
+      "step": 349500
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 3.890379550051711e-05,
+      "loss": 0.3224,
+      "step": 350000
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 3.872633049886935e-05,
+      "loss": 0.3222,
+      "step": 350500
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 3.8548916804204586e-05,
+      "loss": 0.3222,
+      "step": 351000
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 3.837191221499646e-05,
+      "loss": 0.3221,
+      "step": 351500
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 3.819531847820887e-05,
+      "loss": 0.3221,
+      "step": 352000
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.8019489286107e-05,
+      "loss": 0.3223,
+      "step": 352500
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 3.7843721648413935e-05,
+      "loss": 0.3221,
+      "step": 353000
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 3.766837007616901e-05,
+      "loss": 0.3221,
+      "step": 353500
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 3.7493436300021477e-05,
+      "loss": 0.3221,
+      "step": 354000
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 3.7318922046497047e-05,
+      "loss": 0.322,
+      "step": 354500
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 3.7145176802452735e-05,
+      "loss": 0.3223,
+      "step": 355000
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 3.697150590953425e-05,
+      "loss": 0.322,
+      "step": 355500
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 3.679860575795753e-05,
+      "loss": 0.3219,
+      "step": 356000
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 3.6625785068168896e-05,
+      "loss": 0.3218,
+      "step": 356500
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 3.645339246437073e-05,
+      "loss": 0.3218,
+      "step": 357000
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 3.6281429648008496e-05,
+      "loss": 0.3216,
+      "step": 357500
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 3.610989831628571e-05,
+      "loss": 0.322,
+      "step": 358000
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 3.593880016214741e-05,
+      "loss": 0.3217,
+      "step": 358500
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 3.576813687426327e-05,
+      "loss": 0.3218,
+      "step": 359000
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 3.5598250153690636e-05,
+      "loss": 0.3217,
+      "step": 359500
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 3.5428460769004264e-05,
+      "loss": 0.3216,
+      "step": 360000
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 3.525911128741588e-05,
+      "loss": 0.3217,
+      "step": 360500
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 3.5090203380336435e-05,
+      "loss": 0.3216,
+      "step": 361000
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 3.4921738714818835e-05,
+      "loss": 0.3216,
+      "step": 361500
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 3.4753718953541264e-05,
+      "loss": 0.3216,
+      "step": 362000
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 3.458614575479098e-05,
+      "loss": 0.3216,
+      "step": 362500
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 3.441902077244776e-05,
+      "loss": 0.3218,
+      "step": 363000
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 3.425234565596775e-05,
+      "loss": 0.3216,
+      "step": 363500
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 3.4086454045879215e-05,
+      "loss": 0.3217,
+      "step": 364000
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 3.39206826837805e-05,
+      "loss": 0.3217,
+      "step": 364500
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 3.37553661059407e-05,
+      "loss": 0.3218,
+      "step": 365000
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 3.359050594396775e-05,
+      "loss": 0.3216,
+      "step": 365500
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 3.342610382496483e-05,
+      "loss": 0.3217,
+      "step": 366000
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 3.326216137151454e-05,
+      "loss": 0.3217,
+      "step": 366500
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 3.309868020166266e-05,
+      "loss": 0.3215,
+      "step": 367000
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 3.293566192890239e-05,
+      "loss": 0.3215,
+      "step": 367500
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 3.277310816215822e-05,
+      "loss": 0.3215,
+      "step": 368000
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 3.261134421484081e-05,
+      "loss": 0.3215,
+      "step": 368500
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 3.244972333153504e-05,
+      "loss": 0.3213,
+      "step": 369000
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 3.228857175026319e-05,
+      "loss": 0.3213,
+      "step": 369500
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 3.212789106152626e-05,
+      "loss": 0.3215,
+      "step": 370000
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 3.1967682851177825e-05,
+      "loss": 0.3212,
+      "step": 370500
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 3.180794870040819e-05,
+      "loss": 0.3216,
+      "step": 371000
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 3.1649008227028015e-05,
+      "loss": 0.3214,
+      "step": 371500
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 3.149022596427511e-05,
+      "loss": 0.3211,
+      "step": 372000
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 3.133223860153585e-05,
+      "loss": 0.3214,
+      "step": 372500
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 3.117441448272091e-05,
+      "loss": 0.3214,
+      "step": 373000
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 3.1017072252723306e-05,
+      "loss": 0.3214,
+      "step": 373500
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 3.086021346444738e-05,
+      "loss": 0.3212,
+      "step": 374000
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 3.070415192857639e-05,
+      "loss": 0.3212,
+      "step": 374500
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 3.054826368875238e-05,
+      "loss": 0.3212,
+      "step": 375000
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 3.0392863517601306e-05,
+      "loss": 0.3212,
+      "step": 375500
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 3.0237952948860104e-05,
+      "loss": 0.3208,
+      "step": 376000
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 3.0083841859145522e-05,
+      "loss": 0.3209,
+      "step": 376500
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 2.9929914090262566e-05,
+      "loss": 0.3209,
+      "step": 377000
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 2.977648049291346e-05,
+      "loss": 0.3209,
+      "step": 377500
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 2.9623542581425794e-05,
+      "loss": 0.321,
+      "step": 378000
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 2.947110186523508e-05,
+      "loss": 0.3209,
+      "step": 378500
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 2.9319159848869626e-05,
+      "loss": 0.3209,
+      "step": 379000
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 2.9167718031935745e-05,
+      "loss": 0.3207,
+      "step": 379500
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 2.9016777909103095e-05,
+      "loss": 0.3207,
+      "step": 380000
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 2.88663409700897e-05,
+      "loss": 0.3209,
+      "step": 380500
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 2.8716408699647456e-05,
+      "loss": 0.3207,
+      "step": 381000
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 2.85669825775473e-05,
+      "loss": 0.3207,
+      "step": 381500
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 2.8418064078564776e-05,
+      "loss": 0.3208,
+      "step": 382000
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 2.8269950982228437e-05,
+      "loss": 0.3208,
+      "step": 382500
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 2.8122051111179246e-05,
+      "loss": 0.3207,
+      "step": 383000
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 2.7974663254541644e-05,
+      "loss": 0.3208,
+      "step": 383500
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 2.782778886697424e-05,
+      "loss": 0.3208,
+      "step": 384000
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 2.768172160215594e-05,
+      "loss": 0.3205,
+      "step": 384500
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 2.753587746225503e-05,
+      "loss": 0.3208,
+      "step": 385000
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 2.7390551122062965e-05,
+      "loss": 0.3208,
+      "step": 385500
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 2.7245744015892008e-05,
+      "loss": 0.3206,
+      "step": 386000
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 2.7101457572929877e-05,
+      "loss": 0.3206,
+      "step": 386500
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 2.6957693217225504e-05,
+      "loss": 0.3207,
+      "step": 387000
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 2.6814452367675008e-05,
+      "loss": 0.3205,
+      "step": 387500
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 2.6671736438007784e-05,
+      "loss": 0.3206,
+      "step": 388000
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 2.6529830689765492e-05,
+      "loss": 0.3206,
+      "step": 388500
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 2.6388167763455287e-05,
+      "loss": 0.3204,
+      "step": 389000
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 2.62470339642853e-05,
+      "loss": 0.3204,
+      "step": 389500
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 2.6106711361365064e-05,
+      "loss": 0.3206,
+      "step": 390000
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 2.5966638924843332e-05,
+      "loss": 0.3204,
+      "step": 390500
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 2.5827099775781274e-05,
+      "loss": 0.3204,
+      "step": 391000
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 2.5688095291373918e-05,
+      "loss": 0.3203,
+      "step": 391500
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 2.554962684353946e-05,
+      "loss": 0.3203,
+      "step": 392000
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 2.5411695798905557e-05,
+      "loss": 0.3205,
+      "step": 392500
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 2.5274303518795908e-05,
+      "loss": 0.3206,
+      "step": 393000
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 2.513745135921688e-05,
+      "loss": 0.3203,
+      "step": 393500
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 2.5001412750937412e-05,
+      "loss": 0.3205,
+      "step": 394000
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 2.4865643792130036e-05,
+      "loss": 0.3204,
+      "step": 394500
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 2.473041898715981e-05,
+      "loss": 0.3201,
+      "step": 395000
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 2.459600848399211e-05,
+      "loss": 0.3203,
+      "step": 395500
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 2.4462142610782094e-05,
+      "loss": 0.3203,
+      "step": 396000
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 2.4328556058264563e-05,
+      "loss": 0.3203,
+      "step": 396500
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 2.419551896042324e-05,
+      "loss": 0.3201,
+      "step": 397000
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 2.4063032630280625e-05,
+      "loss": 0.3201,
+      "step": 397500
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 2.3931098375423278e-05,
+      "loss": 0.3203,
+      "step": 398000
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 2.3799717497989015e-05,
+      "loss": 0.3202,
+      "step": 398500
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 2.3668891294654094e-05,
+      "loss": 0.3199,
+      "step": 399000
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 2.3538621056620283e-05,
+      "loss": 0.3202,
+      "step": 399500
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 2.340890806960229e-05,
+      "loss": 0.3201,
+      "step": 400000
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 2.327975361381488e-05,
+      "loss": 0.3202,
+      "step": 400500
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 2.3151158963960426e-05,
+      "loss": 0.3202,
+      "step": 401000
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 2.302312538921618e-05,
+      "loss": 0.3201,
+      "step": 401500
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 2.2895654153221863e-05,
+      "loss": 0.32,
+      "step": 402000
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 2.276874651406709e-05,
+      "loss": 0.3197,
+      "step": 402500
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 2.2642655845309204e-05,
+      "loss": 0.3199,
+      "step": 403000
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 2.2516878018406796e-05,
+      "loss": 0.32,
+      "step": 403500
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 2.2391667526711484e-05,
+      "loss": 0.3202,
+      "step": 404000
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 2.226702560600025e-05,
+      "loss": 0.3198,
+      "step": 404500
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 2.2142953486438546e-05,
+      "loss": 0.32,
+      "step": 405000
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 2.201945239256808e-05,
+      "loss": 0.3199,
+      "step": 405500
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 2.1896523543294826e-05,
+      "loss": 0.3198,
+      "step": 406000
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 2.177465642951213e-05,
+      "loss": 0.3199,
+      "step": 406500
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 2.165287340248795e-05,
+      "loss": 0.3197,
+      "step": 407000
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 2.1531666238047615e-05,
+      "loss": 0.32,
+      "step": 407500
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 2.141103613245686e-05,
+      "loss": 0.32,
+      "step": 408000
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 2.129098427628619e-05,
+      "loss": 0.3199,
+      "step": 408500
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 2.1171511854398904e-05,
+      "loss": 0.3197,
+      "step": 409000
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 2.1052620045939608e-05,
+      "loss": 0.32,
+      "step": 409500
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 2.0934546062965312e-05,
+      "loss": 0.32,
+      "step": 410000
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 2.0816817828791883e-05,
+      "loss": 0.3197,
+      "step": 410500
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.069990741325806e-05,
+      "loss": 0.3195,
+      "step": 411000
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.0583347391788827e-05,
+      "loss": 0.3199,
+      "step": 411500
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 2.046737378869338e-05,
+      "loss": 0.3197,
+      "step": 412000
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 2.0351987748584336e-05,
+      "loss": 0.3196,
+      "step": 412500
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 2.0237190410275346e-05,
+      "loss": 0.3197,
+      "step": 413000
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 2.0122982906769767e-05,
+      "loss": 0.3195,
+      "step": 413500
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 2.000936636524959e-05,
+      "loss": 0.3195,
+      "step": 414000
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 1.9896567364338962e-05,
+      "loss": 0.3197,
+      "step": 414500
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 1.978413491748704e-05,
+      "loss": 0.3197,
+      "step": 415000
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 1.9672296776913344e-05,
+      "loss": 0.3197,
+      "step": 415500
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 1.956105404641519e-05,
+      "loss": 0.3195,
+      "step": 416000
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 1.945040782391339e-05,
+      "loss": 0.3196,
+      "step": 416500
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 1.9340359201441538e-05,
+      "loss": 0.3195,
+      "step": 417000
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 1.923090926513507e-05,
+      "loss": 0.3193,
+      "step": 417500
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 1.912205909522066e-05,
+      "loss": 0.3196,
+      "step": 418000
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 1.9013809766005588e-05,
+      "loss": 0.3196,
+      "step": 418500
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 1.8906162345866986e-05,
+      "loss": 0.3195,
+      "step": 419000
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 1.879911789724147e-05,
+      "loss": 0.3196,
+      "step": 419500
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 1.8692889753936567e-05,
+      "loss": 0.3195,
+      "step": 420000
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 1.858705320063063e-05,
+      "loss": 0.3193,
+      "step": 420500
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 1.84818227683162e-05,
+      "loss": 0.3197,
+      "step": 421000
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 1.837719949557515e-05,
+      "loss": 0.3195,
+      "step": 421500
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 1.8273391837498994e-05,
+      "loss": 0.3193,
+      "step": 422000
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 1.816998475621245e-05,
+      "loss": 0.3193,
+      "step": 422500
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 1.806718791221458e-05,
+      "loss": 0.3195,
+      "step": 423000
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 1.796500232006872e-05,
+      "loss": 0.3195,
+      "step": 423500
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 1.786342898830544e-05,
+      "loss": 0.3195,
+      "step": 424000
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 1.776246891941248e-05,
+      "loss": 0.3192,
+      "step": 424500
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 1.7662523268144196e-05,
+      "loss": 0.3195,
+      "step": 425000
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 1.7562790245271302e-05,
+      "loss": 0.3194,
+      "step": 425500
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 1.746367345245176e-05,
+      "loss": 0.3194,
+      "step": 426000
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 1.7365173867928312e-05,
+      "loss": 0.3194,
+      "step": 426500
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 1.726729246385207e-05,
+      "loss": 0.3194,
+      "step": 427000
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 1.7170224112240894e-05,
+      "loss": 0.3192,
+      "step": 427500
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 1.707358071993139e-05,
+      "loss": 0.3192,
+      "step": 428000
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 1.6977558385975733e-05,
+      "loss": 0.3192,
+      "step": 428500
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 1.6882158058075643e-05,
+      "loss": 0.3193,
+      "step": 429000
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 1.6787569610229864e-05,
+      "loss": 0.3193,
+      "step": 429500
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 1.6693414864285253e-05,
+      "loss": 0.3193,
+      "step": 430000
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 1.6599884928778023e-05,
+      "loss": 0.3191,
+      "step": 430500
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 1.650698072681089e-05,
+      "loss": 0.3191,
+      "step": 431000
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 1.6414703175310808e-05,
+      "loss": 0.3194,
+      "step": 431500
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 1.632305318501997e-05,
+      "loss": 0.3191,
+      "step": 432000
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 1.623221307572887e-05,
+      "loss": 0.3194,
+      "step": 432500
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 1.6141819655678393e-05,
+      "loss": 0.3191,
+      "step": 433000
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 1.6052056490087305e-05,
+      "loss": 0.3192,
+      "step": 433500
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 1.5962924464881844e-05,
+      "loss": 0.3192,
+      "step": 434000
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 1.5874424459759086e-05,
+      "loss": 0.3194,
+      "step": 434500
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 1.578655734817838e-05,
+      "loss": 0.3192,
+      "step": 435000
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 1.5699323997352622e-05,
+      "loss": 0.3192,
+      "step": 435500
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 1.5612725268239792e-05,
+      "loss": 0.3193,
+      "step": 436000
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 1.5526762015534355e-05,
+      "loss": 0.3192,
+      "step": 436500
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 1.5441435087658935e-05,
+      "loss": 0.3192,
+      "step": 437000
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 1.53569140698284e-05,
+      "loss": 0.3193,
+      "step": 437500
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 1.5273028503707987e-05,
+      "loss": 0.3191,
+      "step": 438000
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 1.518961302104034e-05,
+      "loss": 0.3191,
+      "step": 438500
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 1.5106837190727158e-05,
+      "loss": 0.3192,
+      "step": 439000
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 1.5024701829732482e-05,
+      "loss": 0.3196,
+      "step": 439500
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 1.4943207748699148e-05,
+      "loss": 0.3192,
+      "step": 440000
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 1.4862516814602821e-05,
+      "loss": 0.3191,
+      "step": 440500
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 1.478230641353945e-05,
+      "loss": 0.3189,
+      "step": 441000
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 1.4702739684782337e-05,
+      "loss": 0.319,
+      "step": 441500
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 1.4623817413622956e-05,
+      "loss": 0.319,
+      "step": 442000
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 1.4545696288600928e-05,
+      "loss": 0.3193,
+      "step": 442500
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 1.4468063970276709e-05,
+      "loss": 0.3191,
+      "step": 443000
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 1.4391078425704797e-05,
+      "loss": 0.319,
+      "step": 443500
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 1.4314740414701418e-05,
+      "loss": 0.319,
+      "step": 444000
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 1.423905069069194e-05,
+      "loss": 0.3191,
+      "step": 444500
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 1.4164010000703365e-05,
+      "loss": 0.319,
+      "step": 445000
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 1.4089619085357035e-05,
+      "loss": 0.3189,
+      "step": 445500
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 1.4015878678861224e-05,
+      "loss": 0.319,
+      "step": 446000
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 1.3942789509003995e-05,
+      "loss": 0.3191,
+      "step": 446500
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 1.3870640746343366e-05,
+      "loss": 0.319,
+      "step": 447000
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 1.3798853595303208e-05,
+      "loss": 0.3189,
+      "step": 447500
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 1.3727719822851586e-05,
+      "loss": 0.3191,
+      "step": 448000
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 1.365724013105011e-05,
+      "loss": 0.3191,
+      "step": 448500
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 1.3587415215504886e-05,
+      "loss": 0.3193,
+      "step": 449000
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 1.3518245765359632e-05,
+      "loss": 0.3188,
+      "step": 449500
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 1.3449732463288913e-05,
+      "loss": 0.3191,
+      "step": 450000
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 1.3381875985491348e-05,
+      "loss": 0.3188,
+      "step": 450500
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 1.3314677001682987e-05,
+      "loss": 0.3186,
+      "step": 451000
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 1.324813617509066e-05,
+      "loss": 0.3191,
+      "step": 451500
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 1.3182254162445461e-05,
+      "loss": 0.3189,
+      "step": 452000
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 1.3117031613976277e-05,
+      "loss": 0.3188,
+      "step": 452500
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 1.30524691734033e-05,
+      "loss": 0.3189,
+      "step": 453000
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 1.2988567477931785e-05,
+      "loss": 0.3188,
+      "step": 453500
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 1.2925327158245641e-05,
+      "loss": 0.319,
+      "step": 454000
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 1.286274883850132e-05,
+      "loss": 0.3187,
+      "step": 454500
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 1.2800833136321545e-05,
+      "loss": 0.3189,
+      "step": 455000
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 1.2739702505431287e-05,
+      "loss": 0.319,
+      "step": 455500
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 1.2679112536818441e-05,
+      "loss": 0.319,
+      "step": 456000
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 1.2619186998186322e-05,
+      "loss": 0.319,
+      "step": 456500
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 1.2559926480975791e-05,
+      "loss": 0.3189,
+      "step": 457000
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 1.250144809522549e-05,
+      "loss": 0.3188,
+      "step": 457500
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 1.2443633230865312e-05,
+      "loss": 0.3188,
+      "step": 458000
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 1.2386368592745543e-05,
+      "loss": 0.3189,
+      "step": 458500
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 1.232977127387103e-05,
+      "loss": 0.319,
+      "step": 459000
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 1.2273841832834476e-05,
+      "loss": 0.3189,
+      "step": 459500
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 1.2218580821636874e-05,
+      "loss": 0.319,
+      "step": 460000
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 1.2163988785682091e-05,
+      "loss": 0.3187,
+      "step": 460500
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 1.2110066263771492e-05,
+      "loss": 0.3188,
+      "step": 461000
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 1.2056813788098594e-05,
+      "loss": 0.3188,
+      "step": 461500
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 1.2004231884243836e-05,
+      "loss": 0.3188,
+      "step": 462000
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 1.195242422270528e-05,
+      "loss": 0.3186,
+      "step": 462500
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 1.1901285479547943e-05,
+      "loss": 0.3187,
+      "step": 463000
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 1.1850715688988035e-05,
+      "loss": 0.3187,
+      "step": 463500
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 1.1800818504351289e-05,
+      "loss": 0.3186,
+      "step": 464000
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 1.175159441810279e-05,
+      "loss": 0.3187,
+      "step": 464500
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 1.1703043916064376e-05,
+      "loss": 0.3188,
+      "step": 465000
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 1.1655167477409934e-05,
+      "loss": 0.3187,
+      "step": 465500
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 1.1607965574660545e-05,
+      "loss": 0.3186,
+      "step": 466000
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 1.1561438673679967e-05,
+      "loss": 0.3187,
+      "step": 466500
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 1.1515587233669903e-05,
+      "loss": 0.3187,
+      "step": 467000
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 1.1470411707165581e-05,
+      "loss": 0.3188,
+      "step": 467500
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 1.142600086306578e-05,
+      "loss": 0.3186,
+      "step": 468000
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 1.1382177140459145e-05,
+      "loss": 0.3187,
+      "step": 468500
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 1.1339030648061997e-05,
+      "loss": 0.3186,
+      "step": 469000
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 1.1296561811712808e-05,
+      "loss": 0.3188,
+      "step": 469500
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 1.1254771050561826e-05,
+      "loss": 0.3186,
+      "step": 470000
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 1.1213740324212508e-05,
+      "loss": 0.3186,
+      "step": 470500
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 1.117330558594806e-05,
+      "loss": 0.3189,
+      "step": 471000
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 1.1133550139371026e-05,
+      "loss": 0.3185,
+      "step": 471500
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 1.1094474376851596e-05,
+      "loss": 0.3187,
+      "step": 472000
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 1.1056154796474798e-05,
+      "loss": 0.3186,
+      "step": 472500
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 1.1018438191073521e-05,
+      "loss": 0.3187,
+      "step": 473000
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 1.0981475797706187e-05,
+      "loss": 0.3186,
+      "step": 473500
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 1.0945191867295903e-05,
+      "loss": 0.3186,
+      "step": 474000
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 1.0909516085370731e-05,
+      "loss": 0.3186,
+      "step": 474500
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 1.0874522198628832e-05,
+      "loss": 0.3185,
+      "step": 475000
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 1.0840210552445753e-05,
+      "loss": 0.3185,
+      "step": 475500
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 1.080664806216178e-05,
+      "loss": 0.3185,
+      "step": 476000
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 1.0773700540136868e-05,
+      "loss": 0.3186,
+      "step": 476500
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 1.0741436253739831e-05,
+      "loss": 0.3186,
+      "step": 477000
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 1.0709855521406153e-05,
+      "loss": 0.3184,
+      "step": 477500
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 1.0678958654824923e-05,
+      "loss": 0.3185,
+      "step": 478000
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 1.0648745958935727e-05,
+      "loss": 0.3184,
+      "step": 478500
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 1.0619217731925679e-05,
+      "loss": 0.3182,
+      "step": 479000
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 1.059037426522648e-05,
+      "loss": 0.3183,
+      "step": 479500
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 1.0562215843511462e-05,
+      "loss": 0.3186,
+      "step": 480000
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 1.0534742744692915e-05,
+      "loss": 0.3186,
+      "step": 480500
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 1.0507955239919215e-05,
+      "loss": 0.3185,
+      "step": 481000
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 1.0481853593572226e-05,
+      "loss": 0.3184,
+      "step": 481500
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 1.0456488209413605e-05,
+      "loss": 0.3185,
+      "step": 482000
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 1.0431757673006884e-05,
+      "loss": 0.3184,
+      "step": 482500
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 1.0407713747066088e-05,
+      "loss": 0.3184,
+      "step": 483000
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 1.038435666889507e-05,
+      "loss": 0.3186,
+      "step": 483500
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 1.036168666901875e-05,
+      "loss": 0.3183,
+      "step": 484000
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 1.0339747250503798e-05,
+      "loss": 0.3185,
+      "step": 484500
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 1.0318450696414725e-05,
+      "loss": 0.3186,
+      "step": 485000
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 1.0297841871085698e-05,
+      "loss": 0.3183,
+      "step": 485500
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 1.0277920977917513e-05,
+      "loss": 0.3184,
+      "step": 486000
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 1.0258763773572561e-05,
+      "loss": 0.3186,
+      "step": 486500
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 1.0240216574124515e-05,
+      "loss": 0.3183,
+      "step": 487000
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 1.022235787557572e-05,
+      "loss": 0.3183,
+      "step": 487500
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 1.0205187854184327e-05,
+      "loss": 0.3185,
+      "step": 488000
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 1.0188706679411501e-05,
+      "loss": 0.3182,
+      "step": 488500
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 1.0172914513919786e-05,
+      "loss": 0.3185,
+      "step": 489000
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 1.0157811513571486e-05,
+      "loss": 0.3183,
+      "step": 489500
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 1.0143397827427121e-05,
+      "loss": 0.3183,
+      "step": 490000
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 1.0129673597743943e-05,
+      "loss": 0.318,
+      "step": 490500
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 1.0116638959974567e-05,
+      "loss": 0.3183,
+      "step": 491000
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 1.0104294042765599e-05,
+      "loss": 0.3184,
+      "step": 491500
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 1.009263896795639e-05,
+      "loss": 0.3185,
+      "step": 492000
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 1.008167385057782e-05,
+      "loss": 0.3184,
+      "step": 492500
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 1.007139879885118e-05,
+      "loss": 0.3184,
+      "step": 493000
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 1.0061832395105667e-05,
+      "loss": 0.3182,
+      "step": 493500
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 1.0052936391489836e-05,
+      "loss": 0.3184,
+      "step": 494000
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 1.0044730737153053e-05,
+      "loss": 0.3181,
+      "step": 494500
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 1.0037215513081807e-05,
+      "loss": 0.3182,
+      "step": 495000
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 1.003039079344833e-05,
+      "loss": 0.3185,
+      "step": 495500
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 1.0024256645609842e-05,
+      "loss": 0.3182,
+      "step": 496000
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 1.0018813130107917e-05,
+      "loss": 0.3183,
+      "step": 496500
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 1.0014069116989727e-05,
+      "loss": 0.3181,
+      "step": 497000
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 1.0010005639011795e-05,
+      "loss": 0.3182,
+      "step": 497500
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 1.000663293402211e-05,
+      "loss": 0.3183,
+      "step": 498000
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 1.0003951035307893e-05,
+      "loss": 0.3181,
+      "step": 498500
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 1.0001959969338421e-05,
+      "loss": 0.3181,
+      "step": 499000
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 1.0000659755764688e-05,
+      "loss": 0.3181,
+      "step": 499500
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 1.000005040741929e-05,
+      "loss": 0.3182,
+      "step": 500000
+    },
+    {
+      "epoch": 8.51,
+      "step": 500000,
+      "total_flos": 0.0,
+      "train_loss": 0.33808115518188475,
+      "train_runtime": 215644.7567,
+      "train_samples_per_second": 593.569,
+      "train_steps_per_second": 2.319
+    }
+  ],
+  "max_steps": 500000,
+  "num_train_epochs": 9,
+  "total_flos": 0.0,
+  "trial_name": null,
+  "trial_params": null
+}