diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6608 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 29.891151091309005,
+  "global_step": 530000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 7.9054,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.928493793261255e-05,
+      "loss": 6.3313,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.85698758652251e-05,
+      "loss": 5.9425,
+      "step": 1500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.785481379783766e-05,
+      "loss": 5.7223,
+      "step": 2000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7139751730450204e-05,
+      "loss": 5.5629,
+      "step": 2500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.642468966306275e-05,
+      "loss": 5.442,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.570962759567531e-05,
+      "loss": 5.334,
+      "step": 3500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.499456552828786e-05,
+      "loss": 5.2403,
+      "step": 4000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.4279503460900405e-05,
+      "loss": 5.1663,
+      "step": 4500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.356444139351296e-05,
+      "loss": 5.0992,
+      "step": 5000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.284937932612551e-05,
+      "loss": 5.0405,
+      "step": 5500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.213431725873806e-05,
+      "loss": 4.986,
+      "step": 6000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.1419255191350606e-05,
+      "loss": 4.9365,
+      "step": 6500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.070419312396317e-05,
+      "loss": 4.8932,
+      "step": 7000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.9989131056575714e-05,
+      "loss": 4.8477,
+      "step": 7500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 3.927406898918826e-05,
+      "loss": 4.8121,
+      "step": 8000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 3.8559006921800814e-05,
+      "loss": 4.7778,
+      "step": 8500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 3.784394485441337e-05,
+      "loss": 4.7411,
+      "step": 9000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 3.7128882787025915e-05,
+      "loss": 4.7129,
+      "step": 9500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 3.641382071963846e-05,
+      "loss": 4.6799,
+      "step": 10000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.432817196982588e-05,
+      "loss": 4.6588,
+      "step": 10500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.404458056831717e-05,
+      "loss": 4.6292,
+      "step": 11000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.376098916680846e-05,
+      "loss": 4.6063,
+      "step": 11500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.347739776529976e-05,
+      "loss": 4.5796,
+      "step": 12000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.319380636379105e-05,
+      "loss": 4.5554,
+      "step": 12500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.2910214962282343e-05,
+      "loss": 4.5335,
+      "step": 13000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.2626623560773635e-05,
+      "loss": 4.5173,
+      "step": 13500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.2343032159264934e-05,
+      "loss": 4.4928,
+      "step": 14000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.2059440757756225e-05,
+      "loss": 4.4717,
+      "step": 14500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.1775849356247524e-05,
+      "loss": 4.4479,
+      "step": 15000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.1492257954738816e-05,
+      "loss": 4.437,
+      "step": 15500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.120866655323011e-05,
+      "loss": 4.4213,
+      "step": 16000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.0925075151721406e-05,
+      "loss": 4.4059,
+      "step": 16500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.06414837502127e-05,
+      "loss": 4.3865,
+      "step": 17000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.035789234870399e-05,
+      "loss": 4.3713,
+      "step": 17500
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 4.288460731506348,
+      "eval_runtime": 860.8156,
+      "eval_samples_per_second": 264.664,
+      "eval_steps_per_second": 33.084,
+      "step": 17731
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007430094719528e-05,
+      "loss": 4.3433,
+      "step": 18000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.979070954568658e-05,
+      "loss": 4.3132,
+      "step": 18500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.950711814417787e-05,
+      "loss": 4.3063,
+      "step": 19000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.922352674266916e-05,
+      "loss": 4.2905,
+      "step": 19500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.893993534116046e-05,
+      "loss": 4.2872,
+      "step": 20000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.8656343939651754e-05,
+      "loss": 4.2758,
+      "step": 20500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.8372752538143045e-05,
+      "loss": 4.2675,
+      "step": 21000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.808916113663434e-05,
+      "loss": 4.2582,
+      "step": 21500
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.7805569735125636e-05,
+      "loss": 4.2483,
+      "step": 22000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.752197833361693e-05,
+      "loss": 4.2384,
+      "step": 22500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.723838693210822e-05,
+      "loss": 4.2275,
+      "step": 23000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 3.695479553059951e-05,
+      "loss": 4.2214,
+      "step": 23500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 3.667120412909081e-05,
+      "loss": 4.2135,
+      "step": 24000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 3.63876127275821e-05,
+      "loss": 4.2036,
+      "step": 24500
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 3.610402132607339e-05,
+      "loss": 4.1925,
+      "step": 25000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 3.582042992456469e-05,
+      "loss": 4.1852,
+      "step": 25500
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 3.553683852305598e-05,
+      "loss": 4.1817,
+      "step": 26000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 3.5253247121547275e-05,
+      "loss": 4.1713,
+      "step": 26500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 3.4969655720038567e-05,
+      "loss": 4.1722,
+      "step": 27000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 3.4686064318529865e-05,
+      "loss": 4.1611,
+      "step": 27500
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 3.440247291702116e-05,
+      "loss": 4.1498,
+      "step": 28000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 3.4118881515512455e-05,
+      "loss": 4.1434,
+      "step": 28500
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.383529011400374e-05,
+      "loss": 4.1377,
+      "step": 29000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 3.355169871249504e-05,
+      "loss": 4.1337,
+      "step": 29500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 3.326810731098634e-05,
+      "loss": 4.1265,
+      "step": 30000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 3.298451590947763e-05,
+      "loss": 4.1284,
+      "step": 30500
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 3.270092450796892e-05,
+      "loss": 4.1102,
+      "step": 31000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 3.241733310646021e-05,
+      "loss": 4.1166,
+      "step": 31500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 3.213374170495151e-05,
+      "loss": 4.1047,
+      "step": 32000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 3.18501503034428e-05,
+      "loss": 4.0991,
+      "step": 32500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 3.1566558901934095e-05,
+      "loss": 4.0919,
+      "step": 33000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 3.1282967500425386e-05,
+      "loss": 4.0908,
+      "step": 33500
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 3.0999376098916685e-05,
+      "loss": 4.081,
+      "step": 34000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 3.0715784697407977e-05,
+      "loss": 4.0792,
+      "step": 34500
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 3.043219329589927e-05,
+      "loss": 4.0755,
+      "step": 35000
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 4.004979610443115,
+      "eval_runtime": 860.9555,
+      "eval_samples_per_second": 264.621,
+      "eval_steps_per_second": 33.078,
+      "step": 35462
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.0148601894390567e-05,
+      "loss": 4.0625,
+      "step": 35500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.9865010492881855e-05,
+      "loss": 4.0114,
+      "step": 36000
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.9581419091373154e-05,
+      "loss": 4.0113,
+      "step": 36500
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.9297827689864442e-05,
+      "loss": 4.0112,
+      "step": 37000
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.901423628835574e-05,
+      "loss": 4.002,
+      "step": 37500
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.873064488684703e-05,
+      "loss": 4.005,
+      "step": 38000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.8447053485338327e-05,
+      "loss": 4.0107,
+      "step": 38500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.8163462083829616e-05,
+      "loss": 3.9993,
+      "step": 39000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.7879870682320914e-05,
+      "loss": 4.0016,
+      "step": 39500
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.759627928081221e-05,
+      "loss": 3.9919,
+      "step": 40000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.73126878793035e-05,
+      "loss": 3.9919,
+      "step": 40500
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.7029096477794796e-05,
+      "loss": 3.9848,
+      "step": 41000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.6745505076286088e-05,
+      "loss": 3.9896,
+      "step": 41500
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.6461913674777383e-05,
+      "loss": 3.9849,
+      "step": 42000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.6178322273268675e-05,
+      "loss": 3.9784,
+      "step": 42500
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.589473087175997e-05,
+      "loss": 3.9746,
+      "step": 43000
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.5611139470251262e-05,
+      "loss": 3.9651,
+      "step": 43500
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.5327548068742557e-05,
+      "loss": 3.9689,
+      "step": 44000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.5043956667233852e-05,
+      "loss": 3.9698,
+      "step": 44500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.4760365265725144e-05,
+      "loss": 3.9631,
+      "step": 45000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.447677386421644e-05,
+      "loss": 3.9595,
+      "step": 45500
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.419318246270773e-05,
+      "loss": 3.9582,
+      "step": 46000
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.3909591061199026e-05,
+      "loss": 3.9564,
+      "step": 46500
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.362599965969032e-05,
+      "loss": 3.9534,
+      "step": 47000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.3342408258181613e-05,
+      "loss": 3.95,
+      "step": 47500
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.3058816856672908e-05,
+      "loss": 3.9472,
+      "step": 48000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.27752254551642e-05,
+      "loss": 3.9486,
+      "step": 48500
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.2491634053655495e-05,
+      "loss": 3.9404,
+      "step": 49000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.2208042652146787e-05,
+      "loss": 3.9339,
+      "step": 49500
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.192445125063808e-05,
+      "loss": 3.9353,
+      "step": 50000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.1640859849129373e-05,
+      "loss": 3.9354,
+      "step": 50500
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.135726844762067e-05,
+      "loss": 3.9301,
+      "step": 51000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.1073677046111964e-05,
+      "loss": 3.9274,
+      "step": 51500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.079008564460326e-05,
+      "loss": 3.9231,
+      "step": 52000
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.050649424309455e-05,
+      "loss": 3.9214,
+      "step": 52500
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.0222902841585846e-05,
+      "loss": 3.9244,
+      "step": 53000
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 3.8807497024536133,
+      "eval_runtime": 859.6671,
+      "eval_samples_per_second": 265.018,
+      "eval_steps_per_second": 33.128,
+      "step": 53193
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.9939311440077137e-05,
+      "loss": 3.8879,
+      "step": 53500
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.9655720038568433e-05,
+      "loss": 3.8667,
+      "step": 54000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.9372128637059724e-05,
+      "loss": 3.8687,
+      "step": 54500
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.908853723555102e-05,
+      "loss": 3.8726,
+      "step": 55000
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.880494583404231e-05,
+      "loss": 3.8683,
+      "step": 55500
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.8521354432533606e-05,
+      "loss": 3.8649,
+      "step": 56000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.82377630310249e-05,
+      "loss": 3.8723,
+      "step": 56500
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.7954171629516197e-05,
+      "loss": 3.867,
+      "step": 57000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.767058022800749e-05,
+      "loss": 3.8646,
+      "step": 57500
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.7386988826498783e-05,
+      "loss": 3.8663,
+      "step": 58000
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.7103397424990075e-05,
+      "loss": 3.8549,
+      "step": 58500
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.681980602348137e-05,
+      "loss": 3.8581,
+      "step": 59000
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.6536214621972662e-05,
+      "loss": 3.8602,
+      "step": 59500
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.6252623220463957e-05,
+      "loss": 3.8606,
+      "step": 60000
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.596903181895525e-05,
+      "loss": 3.8573,
+      "step": 60500
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.5685440417446544e-05,
+      "loss": 3.8579,
+      "step": 61000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.540184901593784e-05,
+      "loss": 3.853,
+      "step": 61500
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.5118257614429133e-05,
+      "loss": 3.8544,
+      "step": 62000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.4834666212920426e-05,
+      "loss": 3.847,
+      "step": 62500
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1.455107481141172e-05,
+      "loss": 3.8492,
+      "step": 63000
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 1.4267483409903013e-05,
+      "loss": 3.8466,
+      "step": 63500
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.3983892008394306e-05,
+      "loss": 3.8439,
+      "step": 64000
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.37003006068856e-05,
+      "loss": 3.8457,
+      "step": 64500
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.3416709205376893e-05,
+      "loss": 3.8434,
+      "step": 65000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.3133117803868187e-05,
+      "loss": 3.8368,
+      "step": 65500
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 1.2849526402359482e-05,
+      "loss": 3.841,
+      "step": 66000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.2565935000850775e-05,
+      "loss": 3.8426,
+      "step": 66500
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.2282343599342069e-05,
+      "loss": 3.8318,
+      "step": 67000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.1998752197833362e-05,
+      "loss": 3.8418,
+      "step": 67500
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.1715160796324656e-05,
+      "loss": 3.834,
+      "step": 68000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.1431569394815949e-05,
+      "loss": 3.8356,
+      "step": 68500
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.1147977993307242e-05,
+      "loss": 3.8393,
+      "step": 69000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 1.0864386591798536e-05,
+      "loss": 3.8317,
+      "step": 69500
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.0580795190289831e-05,
+      "loss": 3.8317,
+      "step": 70000
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.0297203788781125e-05,
+      "loss": 3.8276,
+      "step": 70500
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 3.812286853790283,
+      "eval_runtime": 859.7712,
+      "eval_samples_per_second": 264.986,
+      "eval_steps_per_second": 33.124,
+      "step": 70924
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.0013612387272418e-05,
+      "loss": 3.8235,
+      "step": 71000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.730020985763711e-06,
+      "loss": 3.7872,
+      "step": 71500
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 9.446429584255005e-06,
+      "loss": 3.7813,
+      "step": 72000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 9.1628381827463e-06,
+      "loss": 3.7885,
+      "step": 72500
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.879246781237593e-06,
+      "loss": 3.7837,
+      "step": 73000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.595655379728887e-06,
+      "loss": 3.7849,
+      "step": 73500
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.31206397822018e-06,
+      "loss": 3.7867,
+      "step": 74000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.028472576711475e-06,
+      "loss": 3.7838,
+      "step": 74500
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 7.744881175202769e-06,
+      "loss": 3.7853,
+      "step": 75000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 7.461289773694062e-06,
+      "loss": 3.7855,
+      "step": 75500
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 7.177698372185356e-06,
+      "loss": 3.7883,
+      "step": 76000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 6.894106970676649e-06,
+      "loss": 3.7792,
+      "step": 76500
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 6.6105155691679434e-06,
+      "loss": 3.7845,
+      "step": 77000
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 6.326924167659237e-06,
+      "loss": 3.7818,
+      "step": 77500
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 6.04333276615053e-06,
+      "loss": 3.7807,
+      "step": 78000
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 5.759741364641824e-06,
+      "loss": 3.7801,
+      "step": 78500
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 5.476149963133118e-06,
+      "loss": 3.7759,
+      "step": 79000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 5.1925585616244115e-06,
+      "loss": 3.7825,
+      "step": 79500
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 4.908967160115706e-06,
+      "loss": 3.7858,
+      "step": 80000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 2.737684520106329e-05,
+      "loss": 3.8083,
+      "step": 80500
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 2.7235450483569934e-05,
+      "loss": 3.8201,
+      "step": 81000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 2.7094055766076583e-05,
+      "loss": 3.8216,
+      "step": 81500
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 2.695266104858323e-05,
+      "loss": 3.8234,
+      "step": 82000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.6811266331089874e-05,
+      "loss": 3.8227,
+      "step": 82500
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 2.666987161359652e-05,
+      "loss": 3.8237,
+      "step": 83000
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 2.652847689610316e-05,
+      "loss": 3.8267,
+      "step": 83500
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 2.6387082178609807e-05,
+      "loss": 3.8183,
+      "step": 84000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 2.6245687461116452e-05,
+      "loss": 3.8273,
+      "step": 84500
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 2.6104292743623097e-05,
+      "loss": 3.827,
+      "step": 85000
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 2.5962898026129746e-05,
+      "loss": 3.8238,
+      "step": 85500
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.582150330863639e-05,
+      "loss": 3.8235,
+      "step": 86000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.5680108591143037e-05,
+      "loss": 3.8168,
+      "step": 86500
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 2.5538713873649683e-05,
+      "loss": 3.8192,
+      "step": 87000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 2.5397319156156328e-05,
+      "loss": 3.821,
+      "step": 87500
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.525592443866297e-05,
+      "loss": 3.8137,
+      "step": 88000
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 2.5114529721169615e-05,
+      "loss": 3.8183,
+      "step": 88500
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 3.806265354156494,
+      "eval_runtime": 860.47,
+      "eval_samples_per_second": 264.77,
+      "eval_steps_per_second": 33.097,
+      "step": 88655
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 2.4973135003676264e-05,
+      "loss": 3.787,
+      "step": 89000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.483174028618291e-05,
+      "loss": 3.7753,
+      "step": 89500
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 2.4690345568689555e-05,
+      "loss": 3.781,
+      "step": 90000
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 2.45489508511962e-05,
+      "loss": 3.781,
+      "step": 90500
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 2.4407556133702846e-05,
+      "loss": 3.7791,
+      "step": 91000
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 2.426616141620949e-05,
+      "loss": 3.7782,
+      "step": 91500
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.4124766698716137e-05,
+      "loss": 3.7754,
+      "step": 92000
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 2.3983371981222782e-05,
+      "loss": 3.7808,
+      "step": 92500
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.3841977263729427e-05,
+      "loss": 3.7761,
+      "step": 93000
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 2.3700582546236076e-05,
+      "loss": 3.7795,
+      "step": 93500
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 2.3559187828742718e-05,
+      "loss": 3.7861,
+      "step": 94000
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 2.3417793111249364e-05,
+      "loss": 3.7788,
+      "step": 94500
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 2.327639839375601e-05,
+      "loss": 3.7763,
+      "step": 95000
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 2.3135003676262658e-05,
+      "loss": 3.7804,
+      "step": 95500
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 2.2993608958769303e-05,
+      "loss": 3.7721,
+      "step": 96000
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 2.2852214241275945e-05,
+      "loss": 3.7698,
+      "step": 96500
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 2.271081952378259e-05,
+      "loss": 3.775,
+      "step": 97000
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 2.256942480628924e-05,
+      "loss": 3.772,
+      "step": 97500
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 2.2428030088795885e-05,
+      "loss": 3.7743,
+      "step": 98000
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.228663537130253e-05,
+      "loss": 3.7746,
+      "step": 98500
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 2.2145240653809172e-05,
+      "loss": 3.7706,
+      "step": 99000
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 2.200384593631582e-05,
+      "loss": 3.7669,
+      "step": 99500
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 2.1862451218822467e-05,
+      "loss": 3.7704,
+      "step": 100000
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 2.1721056501329112e-05,
+      "loss": 3.7677,
+      "step": 100500
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 2.1579661783835754e-05,
+      "loss": 3.7705,
+      "step": 101000
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 2.1438267066342403e-05,
+      "loss": 3.7676,
+      "step": 101500
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 2.1296872348849048e-05,
+      "loss": 3.763,
+      "step": 102000
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 2.1155477631355694e-05,
+      "loss": 3.7676,
+      "step": 102500
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 2.101408291386234e-05,
+      "loss": 3.7668,
+      "step": 103000
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 2.0872688196368985e-05,
+      "loss": 3.763,
+      "step": 103500
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 2.073129347887563e-05,
+      "loss": 3.762,
+      "step": 104000
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 2.0589898761382275e-05,
+      "loss": 3.7595,
+      "step": 104500
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 2.044850404388892e-05,
+      "loss": 3.7625,
+      "step": 105000
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 2.030710932639557e-05,
+      "loss": 3.7623,
+      "step": 105500
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 2.016571460890221e-05,
+      "loss": 3.7612,
+      "step": 106000
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 3.7527620792388916,
+      "eval_runtime": 859.2059,
+      "eval_samples_per_second": 265.16,
+      "eval_steps_per_second": 33.146,
+      "step": 106386
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.0024319891408857e-05,
+      "loss": 3.744,
+      "step": 106500
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 1.9882925173915502e-05,
+      "loss": 3.7041,
+      "step": 107000
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.974153045642215e-05,
+      "loss": 3.708,
+      "step": 107500
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 1.9600135738928793e-05,
+      "loss": 3.7088,
+      "step": 108000
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 1.945874102143544e-05,
+      "loss": 3.709,
+      "step": 108500
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1.9317346303942084e-05,
+      "loss": 3.7106,
+      "step": 109000
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 1.9175951586448733e-05,
+      "loss": 3.7105,
+      "step": 109500
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1.9034556868955378e-05,
+      "loss": 3.7095,
+      "step": 110000
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1.889316215146202e-05,
+      "loss": 3.7096,
+      "step": 110500
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1.8751767433968666e-05,
+      "loss": 3.7127,
+      "step": 111000
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 1.8610372716475315e-05,
+      "loss": 3.7108,
+      "step": 111500
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 1.846897799898196e-05,
+      "loss": 3.7165,
+      "step": 112000
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.8327583281488605e-05,
+      "loss": 3.7106,
+      "step": 112500
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1.8186188563995247e-05,
+      "loss": 3.7119,
+      "step": 113000
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1.8044793846501896e-05,
+      "loss": 3.7073,
+      "step": 113500
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.790339912900854e-05,
+      "loss": 3.7135,
+      "step": 114000
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.7762004411515187e-05,
+      "loss": 3.712,
+      "step": 114500
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 1.7620609694021832e-05,
+      "loss": 3.7126,
+      "step": 115000
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1.7479214976528478e-05,
+      "loss": 3.7101,
+      "step": 115500
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1.7337820259035123e-05,
+      "loss": 3.7074,
+      "step": 116000
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1.719642554154177e-05,
+      "loss": 3.7061,
+      "step": 116500
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1.7055030824048414e-05,
+      "loss": 3.7018,
+      "step": 117000
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.691363610655506e-05,
+      "loss": 3.7063,
+      "step": 117500
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1.6772241389061705e-05,
+      "loss": 3.7086,
+      "step": 118000
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 1.663084667156835e-05,
+      "loss": 3.7064,
+      "step": 118500
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.6489451954074996e-05,
+      "loss": 3.7049,
+      "step": 119000
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1.6348057236581645e-05,
+      "loss": 3.7019,
+      "step": 119500
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 1.6206662519088287e-05,
+      "loss": 3.7057,
+      "step": 120000
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1.6065267801594932e-05,
+      "loss": 3.7046,
+      "step": 120500
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 1.5923873084101577e-05,
+      "loss": 3.7094,
+      "step": 121000
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1.5782478366608226e-05,
+      "loss": 3.7016,
+      "step": 121500
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 1.564108364911487e-05,
+      "loss": 3.6999,
+      "step": 122000
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1.5499688931621514e-05,
+      "loss": 3.7064,
+      "step": 122500
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1.535829421412816e-05,
+      "loss": 3.7024,
+      "step": 123000
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1.5216899496634806e-05,
+      "loss": 3.702,
+      "step": 123500
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1.5075504779141453e-05,
+      "loss": 3.6976,
+      "step": 124000
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 3.7144248485565186,
+      "eval_runtime": 859.1145,
+      "eval_samples_per_second": 265.188,
+      "eval_steps_per_second": 33.149,
+      "step": 124117
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 1.4934110061648099e-05,
+      "loss": 3.6596,
+      "step": 124500
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 1.4792715344154742e-05,
+      "loss": 3.657,
+      "step": 125000
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 1.4651320626661388e-05,
+      "loss": 3.6579,
+      "step": 125500
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 1.4509925909168035e-05,
+      "loss": 3.6603,
+      "step": 126000
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 1.436853119167468e-05,
+      "loss": 3.6607,
+      "step": 126500
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 1.4227136474181324e-05,
+      "loss": 3.6531,
+      "step": 127000
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 1.408574175668797e-05,
+      "loss": 3.6601,
+      "step": 127500
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 1.3944347039194617e-05,
+      "loss": 3.6541,
+      "step": 128000
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 1.3802952321701262e-05,
+      "loss": 3.6614,
+      "step": 128500
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 1.3661557604207909e-05,
+      "loss": 3.6619,
+      "step": 129000
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 1.3520162886714551e-05,
+      "loss": 3.6576,
+      "step": 129500
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 1.3378768169221198e-05,
+      "loss": 3.6563,
+      "step": 130000
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 1.3237373451727844e-05,
+      "loss": 3.6527,
+      "step": 130500
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 1.309597873423449e-05,
+      "loss": 3.6587,
+      "step": 131000
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 1.2954584016741136e-05,
+      "loss": 3.6628,
+      "step": 131500
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 1.281318929924778e-05,
+      "loss": 3.6589,
+      "step": 132000
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 1.2671794581754425e-05,
+      "loss": 3.6568,
+      "step": 132500
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 1.2530399864261072e-05,
+      "loss": 3.6583,
+      "step": 133000
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 1.2389005146767716e-05,
+      "loss": 3.6586,
+      "step": 133500
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 1.2247610429274363e-05,
+      "loss": 3.6534,
+      "step": 134000
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 1.2106215711781009e-05,
+      "loss": 3.6598,
+      "step": 134500
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 1.1964820994287654e-05,
+      "loss": 3.6616,
+      "step": 135000
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 1.18234262767943e-05,
+      "loss": 3.6594,
+      "step": 135500
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 1.1682031559300945e-05,
+      "loss": 3.6565,
+      "step": 136000
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 1.154063684180759e-05,
+      "loss": 3.6628,
+      "step": 136500
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 1.1399242124314236e-05,
+      "loss": 3.6568,
+      "step": 137000
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 1.1257847406820881e-05,
+      "loss": 3.6521,
+      "step": 137500
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 1.1116452689327528e-05,
+      "loss": 3.6545,
+      "step": 138000
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 1.0975057971834172e-05,
+      "loss": 3.6551,
+      "step": 138500
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 1.0833663254340819e-05,
+      "loss": 3.6546,
+      "step": 139000
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 1.0692268536847463e-05,
+      "loss": 3.6545,
+      "step": 139500
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 1.055087381935411e-05,
+      "loss": 3.6546,
+      "step": 140000
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 1.0409479101860754e-05,
+      "loss": 3.6538,
+      "step": 140500
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 1.02680843843674e-05,
+      "loss": 3.6578,
+      "step": 141000
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 1.0126689666874046e-05,
+      "loss": 3.6553,
+      "step": 141500
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 3.688018560409546,
+      "eval_runtime": 860.525,
+      "eval_samples_per_second": 264.753,
+      "eval_steps_per_second": 33.095,
+      "step": 141848
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 9.985294949380691e-06,
+      "loss": 3.6382,
+      "step": 142000
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 9.843900231887337e-06,
+      "loss": 3.6156,
+      "step": 142500
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 9.702505514393982e-06,
+      "loss": 3.6166,
+      "step": 143000
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 9.561110796900628e-06,
+      "loss": 3.6137,
+      "step": 143500
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 9.419716079407275e-06,
+      "loss": 3.618,
+      "step": 144000
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 9.278321361913919e-06,
+      "loss": 3.6158,
+      "step": 144500
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 9.136926644420566e-06,
+      "loss": 3.6158,
+      "step": 145000
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 8.99553192692721e-06,
+      "loss": 3.6178,
+      "step": 145500
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 8.854137209433856e-06,
+      "loss": 3.6262,
+      "step": 146000
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 8.7127424919405e-06,
+      "loss": 3.6197,
+      "step": 146500
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 8.571347774447147e-06,
+      "loss": 3.6185,
+      "step": 147000
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 8.429953056953793e-06,
+      "loss": 3.617,
+      "step": 147500
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 8.288558339460438e-06,
+      "loss": 3.6157,
+      "step": 148000
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 8.147163621967084e-06,
+      "loss": 3.6202,
+      "step": 148500
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 8.005768904473729e-06,
+      "loss": 3.6187,
+      "step": 149000
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 7.864374186980374e-06,
+      "loss": 3.6189,
+      "step": 149500
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 7.72297946948702e-06,
+      "loss": 3.6212,
+      "step": 150000
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 7.581584751993666e-06,
+      "loss": 3.6188,
+      "step": 150500
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 7.440190034500312e-06,
+      "loss": 3.6146,
+      "step": 151000
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 7.298795317006957e-06,
+      "loss": 3.6246,
+      "step": 151500
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 7.157400599513603e-06,
+      "loss": 3.6147,
+      "step": 152000
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 7.016005882020248e-06,
+      "loss": 3.6233,
+      "step": 152500
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 6.874611164526894e-06,
+      "loss": 3.621,
+      "step": 153000
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 6.7332164470335385e-06,
+      "loss": 3.6209,
+      "step": 153500
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 6.591821729540185e-06,
+      "loss": 3.6155,
+      "step": 154000
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 6.450427012046831e-06,
+      "loss": 3.6176,
+      "step": 154500
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 6.309032294553476e-06,
+      "loss": 3.6138,
+      "step": 155000
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 6.167637577060122e-06,
+      "loss": 3.6194,
+      "step": 155500
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 6.026242859566767e-06,
+      "loss": 3.6161,
+      "step": 156000
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 5.884848142073413e-06,
+      "loss": 3.6121,
+      "step": 156500
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 5.743453424580058e-06,
+      "loss": 3.6169,
+      "step": 157000
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 5.6020587070867035e-06,
+      "loss": 3.6173,
+      "step": 157500
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 5.460663989593349e-06,
+      "loss": 3.6137,
+      "step": 158000
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 5.319269272099994e-06,
+      "loss": 3.614,
+      "step": 158500
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 5.177874554606641e-06,
+      "loss": 3.6115,
+      "step": 159000
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 5.036479837113286e-06,
+      "loss": 3.6107,
+      "step": 159500
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 3.6684656143188477,
+      "eval_runtime": 860.3576,
+      "eval_samples_per_second": 264.805,
+      "eval_steps_per_second": 33.101,
+      "step": 159579
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 4.895085119619931e-06,
+      "loss": 3.5905,
+      "step": 160000
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 4.753690402126577e-06,
+      "loss": 3.5857,
+      "step": 160500
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 4.612295684633222e-06,
+      "loss": 3.584,
+      "step": 161000
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 4.470900967139868e-06,
+      "loss": 3.5892,
+      "step": 161500
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 4.329506249646514e-06,
+      "loss": 3.5847,
+      "step": 162000
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 4.188111532153159e-06,
+      "loss": 3.5864,
+      "step": 162500
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 4.046716814659805e-06,
+      "loss": 3.5882,
+      "step": 163000
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 3.90532209716645e-06,
+      "loss": 3.5864,
+      "step": 163500
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 3.7639273796730956e-06,
+      "loss": 3.5862,
+      "step": 164000
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 3.622532662179741e-06,
+      "loss": 3.5871,
+      "step": 164500
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 3.4811379446863864e-06,
+      "loss": 3.5843,
+      "step": 165000
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 3.3397432271930326e-06,
+      "loss": 3.5877,
+      "step": 165500
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 3.198348509699678e-06,
+      "loss": 3.5847,
+      "step": 166000
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 3.0569537922063235e-06,
+      "loss": 3.5853,
+      "step": 166500
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 2.915559074712969e-06,
+      "loss": 3.5876,
+      "step": 167000
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 2.7741643572196143e-06,
+      "loss": 3.5896,
+      "step": 167500
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 2.63276963972626e-06,
+      "loss": 3.5886,
+      "step": 168000
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 2.4913749222329055e-06,
+      "loss": 3.5863,
+      "step": 168500
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 2.349980204739551e-06,
+      "loss": 3.5921,
+      "step": 169000
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 2.208585487246197e-06,
+      "loss": 3.5952,
+      "step": 169500
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 2.067190769752842e-06,
+      "loss": 3.5875,
+      "step": 170000
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 2.5996837230317406e-05,
+      "loss": 3.6184,
+      "step": 170500
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 2.592623969275952e-05,
+      "loss": 3.6347,
+      "step": 171000
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 2.5855642155201625e-05,
+      "loss": 3.6381,
+      "step": 171500
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 2.5785044617643738e-05,
+      "loss": 3.6428,
+      "step": 172000
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 2.5714447080085845e-05,
+      "loss": 3.6468,
+      "step": 172500
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 2.5643849542527958e-05,
+      "loss": 3.6507,
+      "step": 173000
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 2.5573252004970067e-05,
+      "loss": 3.6533,
+      "step": 173500
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 2.550265446741218e-05,
+      "loss": 3.646,
+      "step": 174000
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 2.5432056929854287e-05,
+      "loss": 3.6513,
+      "step": 174500
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 2.53614593922964e-05,
+      "loss": 3.6528,
+      "step": 175000
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 2.5290861854738506e-05,
+      "loss": 3.656,
+      "step": 175500
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 2.522026431718062e-05,
+      "loss": 3.651,
+      "step": 176000
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 2.5149666779622726e-05,
+      "loss": 3.6527,
+      "step": 176500
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 2.507906924206484e-05,
+      "loss": 3.6576,
+      "step": 177000
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 3.6973445415496826,
+      "eval_runtime": 860.6149,
+      "eval_samples_per_second": 264.726,
+      "eval_steps_per_second": 33.091,
+      "step": 177310
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 2.500847170450695e-05,
+      "loss": 3.6368,
+      "step": 177500
+    },
+    {
+      "epoch": 10.04,
+      "learning_rate": 2.4937874166949058e-05,
+      "loss": 3.6203,
+      "step": 178000
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 2.4867276629391168e-05,
+      "loss": 3.6227,
+      "step": 178500
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 2.4796679091833278e-05,
+      "loss": 3.6261,
+      "step": 179000
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 2.4726081554275387e-05,
+      "loss": 3.6253,
+      "step": 179500
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 2.4655484016717497e-05,
+      "loss": 3.6296,
+      "step": 180000
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 2.4584886479159607e-05,
+      "loss": 3.6292,
+      "step": 180500
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 2.4514288941601717e-05,
+      "loss": 3.635,
+      "step": 181000
+    },
+    {
+      "epoch": 10.24,
+      "learning_rate": 2.444369140404383e-05,
+      "loss": 3.6267,
+      "step": 181500
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 2.437309386648594e-05,
+      "loss": 3.637,
+      "step": 182000
+    },
+    {
+      "epoch": 10.29,
+      "learning_rate": 2.430249632892805e-05,
+      "loss": 3.6339,
+      "step": 182500
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 2.423189879137016e-05,
+      "loss": 3.6369,
+      "step": 183000
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 2.416130125381227e-05,
+      "loss": 3.6363,
+      "step": 183500
+    },
+    {
+      "epoch": 10.38,
+      "learning_rate": 2.409070371625438e-05,
+      "loss": 3.6373,
+      "step": 184000
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 2.4020106178696488e-05,
+      "loss": 3.6365,
+      "step": 184500
+    },
+    {
+      "epoch": 10.43,
+      "learning_rate": 2.3949508641138598e-05,
+      "loss": 3.6401,
+      "step": 185000
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 2.3878911103580707e-05,
+      "loss": 3.633,
+      "step": 185500
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 2.380831356602282e-05,
+      "loss": 3.6379,
+      "step": 186000
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 2.373771602846493e-05,
+      "loss": 3.6306,
+      "step": 186500
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 2.366711849090704e-05,
+      "loss": 3.6375,
+      "step": 187000
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 2.359652095334915e-05,
+      "loss": 3.6401,
+      "step": 187500
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 2.352592341579126e-05,
+      "loss": 3.6425,
+      "step": 188000
+    },
+    {
+      "epoch": 10.63,
+      "learning_rate": 2.345532587823337e-05,
+      "loss": 3.6464,
+      "step": 188500
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 2.338472834067548e-05,
+      "loss": 3.6418,
+      "step": 189000
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 2.331413080311759e-05,
+      "loss": 3.6419,
+      "step": 189500
+    },
+    {
+      "epoch": 10.72,
+      "learning_rate": 2.32435332655597e-05,
+      "loss": 3.6386,
+      "step": 190000
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 2.3172935728001808e-05,
+      "loss": 3.6403,
+      "step": 190500
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 2.3102338190443918e-05,
+      "loss": 3.639,
+      "step": 191000
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 2.3031740652886028e-05,
+      "loss": 3.6391,
+      "step": 191500
+    },
+    {
+      "epoch": 10.83,
+      "learning_rate": 2.2961143115328137e-05,
+      "loss": 3.6352,
+      "step": 192000
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 2.2890545577770247e-05,
+      "loss": 3.6367,
+      "step": 192500
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 2.2819948040212357e-05,
+      "loss": 3.641,
+      "step": 193000
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 2.2749350502654466e-05,
+      "loss": 3.642,
+      "step": 193500
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 2.267875296509658e-05,
+      "loss": 3.6384,
+      "step": 194000
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 2.260815542753869e-05,
+      "loss": 3.6334,
+      "step": 194500
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 2.25375578899808e-05,
+      "loss": 3.6441,
+      "step": 195000
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 3.6708991527557373,
+      "eval_runtime": 860.0592,
+      "eval_samples_per_second": 264.897,
+      "eval_steps_per_second": 33.113,
+      "step": 195041
+    },
+    {
+      "epoch": 11.03,
+      "learning_rate": 2.246696035242291e-05,
+      "loss": 3.5829,
+      "step": 195500
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 2.239636281486502e-05,
+      "loss": 3.5793,
+      "step": 196000
+    },
+    {
+      "epoch": 11.08,
+      "learning_rate": 2.2325765277307128e-05,
+      "loss": 3.5869,
+      "step": 196500
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 2.2255167739749238e-05,
+      "loss": 3.5855,
+      "step": 197000
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 2.2184570202191348e-05,
+      "loss": 3.587,
+      "step": 197500
+    },
+    {
+      "epoch": 11.17,
+      "learning_rate": 2.2113972664633457e-05,
+      "loss": 3.5912,
+      "step": 198000
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 2.204337512707557e-05,
+      "loss": 3.5869,
+      "step": 198500
+    },
+    {
+      "epoch": 11.22,
+      "learning_rate": 2.197277758951768e-05,
+      "loss": 3.5893,
+      "step": 199000
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 2.190218005195979e-05,
+      "loss": 3.598,
+      "step": 199500
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 2.18315825144019e-05,
+      "loss": 3.5969,
+      "step": 200000
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 2.176098497684401e-05,
+      "loss": 3.5992,
+      "step": 200500
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 2.169038743928612e-05,
+      "loss": 3.5917,
+      "step": 201000
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 2.161978990172823e-05,
+      "loss": 3.5978,
+      "step": 201500
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 2.154919236417034e-05,
+      "loss": 3.5915,
+      "step": 202000
+    },
+    {
+      "epoch": 11.42,
+      "learning_rate": 2.1478594826612448e-05,
+      "loss": 3.5994,
+      "step": 202500
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 2.140799728905456e-05,
+      "loss": 3.5927,
+      "step": 203000
+    },
+    {
+      "epoch": 11.48,
+      "learning_rate": 2.133739975149667e-05,
+      "loss": 3.5964,
+      "step": 203500
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 2.126680221393878e-05,
+      "loss": 3.6008,
+      "step": 204000
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 2.1196204676380887e-05,
+      "loss": 3.5938,
+      "step": 204500
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 2.1125607138822997e-05,
+      "loss": 3.5922,
+      "step": 205000
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 2.1055009601265106e-05,
+      "loss": 3.6034,
+      "step": 205500
+    },
+    {
+      "epoch": 11.62,
+      "learning_rate": 2.0984412063707216e-05,
+      "loss": 3.5964,
+      "step": 206000
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 2.091381452614933e-05,
+      "loss": 3.5963,
+      "step": 206500
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 2.084321698859144e-05,
+      "loss": 3.5982,
+      "step": 207000
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 2.077261945103355e-05,
+      "loss": 3.6001,
+      "step": 207500
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 2.070202191347566e-05,
+      "loss": 3.5998,
+      "step": 208000
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 2.0631424375917768e-05,
+      "loss": 3.6043,
+      "step": 208500
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 2.0560826838359878e-05,
+      "loss": 3.6034,
+      "step": 209000
+    },
+    {
+      "epoch": 11.82,
+      "learning_rate": 2.0490229300801988e-05,
+      "loss": 3.6005,
+      "step": 209500
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 2.0419631763244097e-05,
+      "loss": 3.5998,
+      "step": 210000
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 2.0349034225686207e-05,
+      "loss": 3.6008,
+      "step": 210500
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 2.027843668812832e-05,
+      "loss": 3.5994,
+      "step": 211000
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 2.020783915057043e-05,
+      "loss": 3.5982,
+      "step": 211500
+    },
+    {
+      "epoch": 11.96,
+      "learning_rate": 2.013724161301254e-05,
+      "loss": 3.5993,
+      "step": 212000
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 2.006664407545465e-05,
+      "loss": 3.602,
+      "step": 212500
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 3.646425247192383,
+      "eval_runtime": 859.6242,
+      "eval_samples_per_second": 265.031,
+      "eval_steps_per_second": 33.13,
+      "step": 212772
+    },
+    {
+      "epoch": 12.01,
+      "learning_rate": 1.999604653789676e-05,
+      "loss": 3.5684,
+      "step": 213000
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 1.992544900033887e-05,
+      "loss": 3.54,
+      "step": 213500
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 1.985485146278098e-05,
+      "loss": 3.5453,
+      "step": 214000
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 1.9784253925223088e-05,
+      "loss": 3.5457,
+      "step": 214500
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 1.97136563876652e-05,
+      "loss": 3.5534,
+      "step": 215000
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 1.964305885010731e-05,
+      "loss": 3.552,
+      "step": 215500
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 1.957246131254942e-05,
+      "loss": 3.5531,
+      "step": 216000
+    },
+    {
+      "epoch": 12.21,
+      "learning_rate": 1.950186377499153e-05,
+      "loss": 3.5527,
+      "step": 216500
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 1.943126623743364e-05,
+      "loss": 3.5559,
+      "step": 217000
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 1.936066869987575e-05,
+      "loss": 3.556,
+      "step": 217500
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 1.929007116231786e-05,
+      "loss": 3.5563,
+      "step": 218000
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 1.921947362475997e-05,
+      "loss": 3.5622,
+      "step": 218500
+    },
+    {
+      "epoch": 12.35,
+      "learning_rate": 1.914887608720208e-05,
+      "loss": 3.5628,
+      "step": 219000
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 1.9078278549644192e-05,
+      "loss": 3.563,
+      "step": 219500
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 1.90076810120863e-05,
+      "loss": 3.5585,
+      "step": 220000
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 1.8937083474528408e-05,
+      "loss": 3.5627,
+      "step": 220500
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 1.8866485936970518e-05,
+      "loss": 3.56,
+      "step": 221000
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 1.8795888399412628e-05,
+      "loss": 3.5653,
+      "step": 221500
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 1.8725290861854737e-05,
+      "loss": 3.5669,
+      "step": 222000
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 1.8654693324296847e-05,
+      "loss": 3.5644,
+      "step": 222500
+    },
+    {
+      "epoch": 12.58,
+      "learning_rate": 1.858409578673896e-05,
+      "loss": 3.5678,
+      "step": 223000
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 1.851349824918107e-05,
+      "loss": 3.562,
+      "step": 223500
+    },
+    {
+      "epoch": 12.63,
+      "learning_rate": 1.844290071162318e-05,
+      "loss": 3.5674,
+      "step": 224000
+    },
+    {
+      "epoch": 12.66,
+      "learning_rate": 1.837230317406529e-05,
+      "loss": 3.5661,
+      "step": 224500
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 1.83017056365074e-05,
+      "loss": 3.5652,
+      "step": 225000
+    },
+    {
+      "epoch": 12.72,
+      "learning_rate": 1.823110809894951e-05,
+      "loss": 3.567,
+      "step": 225500
+    },
+    {
+      "epoch": 12.75,
+      "learning_rate": 1.816051056139162e-05,
+      "loss": 3.5663,
+      "step": 226000
+    },
+    {
+      "epoch": 12.77,
+      "learning_rate": 1.808991302383373e-05,
+      "loss": 3.5649,
+      "step": 226500
+    },
+    {
+      "epoch": 12.8,
+      "learning_rate": 1.8019315486275838e-05,
+      "loss": 3.5655,
+      "step": 227000
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 3.5664,
+      "step": 227500
+    },
+    {
+      "epoch": 12.86,
+      "learning_rate": 1.787812041116006e-05,
+      "loss": 3.5617,
+      "step": 228000
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 1.780752287360217e-05,
+      "loss": 3.5643,
+      "step": 228500
+    },
+    {
+      "epoch": 12.92,
+      "learning_rate": 1.773692533604428e-05,
+      "loss": 3.5652,
+      "step": 229000
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 1.766632779848639e-05,
+      "loss": 3.5637,
+      "step": 229500
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 1.75957302609285e-05,
+      "loss": 3.5626,
+      "step": 230000
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 1.752513272337061e-05,
+      "loss": 3.5671,
+      "step": 230500
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 3.6284356117248535,
+      "eval_runtime": 859.8678,
+      "eval_samples_per_second": 264.956,
+      "eval_steps_per_second": 33.12,
+      "step": 230503
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 1.745453518581272e-05,
+      "loss": 3.5115,
+      "step": 231000
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 1.738393764825483e-05,
+      "loss": 3.5145,
+      "step": 231500
+    },
+    {
+      "epoch": 13.08,
+      "learning_rate": 1.7313340110696942e-05,
+      "loss": 3.5134,
+      "step": 232000
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 1.7242742573139052e-05,
+      "loss": 3.5196,
+      "step": 232500
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 1.717214503558116e-05,
+      "loss": 3.5224,
+      "step": 233000
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 1.710154749802327e-05,
+      "loss": 3.5222,
+      "step": 233500
+    },
+    {
+      "epoch": 13.2,
+      "learning_rate": 1.703094996046538e-05,
+      "loss": 3.5284,
+      "step": 234000
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 1.696035242290749e-05,
+      "loss": 3.5227,
+      "step": 234500
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 1.6889754885349597e-05,
+      "loss": 3.5228,
+      "step": 235000
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 1.681915734779171e-05,
+      "loss": 3.5313,
+      "step": 235500
+    },
+    {
+      "epoch": 13.31,
+      "learning_rate": 1.674855981023382e-05,
+      "loss": 3.5317,
+      "step": 236000
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 1.667796227267593e-05,
+      "loss": 3.5257,
+      "step": 236500
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 1.660736473511804e-05,
+      "loss": 3.5308,
+      "step": 237000
+    },
+    {
+      "epoch": 13.39,
+      "learning_rate": 1.653676719756015e-05,
+      "loss": 3.5269,
+      "step": 237500
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 1.646616966000226e-05,
+      "loss": 3.5294,
+      "step": 238000
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 1.639557212244437e-05,
+      "loss": 3.5298,
+      "step": 238500
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 1.6324974584886478e-05,
+      "loss": 3.5276,
+      "step": 239000
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 1.6254377047328588e-05,
+      "loss": 3.53,
+      "step": 239500
+    },
+    {
+      "epoch": 13.54,
+      "learning_rate": 1.61837795097707e-05,
+      "loss": 3.5307,
+      "step": 240000
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 1.611318197221281e-05,
+      "loss": 3.5359,
+      "step": 240500
+    },
+    {
+      "epoch": 13.59,
+      "learning_rate": 1.604258443465492e-05,
+      "loss": 3.5333,
+      "step": 241000
+    },
+    {
+      "epoch": 13.62,
+      "learning_rate": 1.597198689709703e-05,
+      "loss": 3.5384,
+      "step": 241500
+    },
+    {
+      "epoch": 13.65,
+      "learning_rate": 1.590138935953914e-05,
+      "loss": 3.5369,
+      "step": 242000
+    },
+    {
+      "epoch": 13.68,
+      "learning_rate": 1.583079182198125e-05,
+      "loss": 3.5339,
+      "step": 242500
+    },
+    {
+      "epoch": 13.7,
+      "learning_rate": 1.576019428442336e-05,
+      "loss": 3.5339,
+      "step": 243000
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 1.568959674686547e-05,
+      "loss": 3.5378,
+      "step": 243500
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 1.5618999209307582e-05,
+      "loss": 3.5342,
+      "step": 244000
+    },
+    {
+      "epoch": 13.79,
+      "learning_rate": 1.5548401671749692e-05,
+      "loss": 3.5358,
+      "step": 244500
+    },
+    {
+      "epoch": 13.82,
+      "learning_rate": 1.54778041341918e-05,
+      "loss": 3.5333,
+      "step": 245000
+    },
+    {
+      "epoch": 13.85,
+      "learning_rate": 1.540720659663391e-05,
+      "loss": 3.5336,
+      "step": 245500
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 1.533660905907602e-05,
+      "loss": 3.5374,
+      "step": 246000
+    },
+    {
+      "epoch": 13.9,
+      "learning_rate": 1.526601152151813e-05,
+      "loss": 3.5273,
+      "step": 246500
+    },
+    {
+      "epoch": 13.93,
+      "learning_rate": 1.519541398396024e-05,
+      "loss": 3.5332,
+      "step": 247000
+    },
+    {
+      "epoch": 13.96,
+      "learning_rate": 1.5124816446402352e-05,
+      "loss": 3.5335,
+      "step": 247500
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 1.5054218908844462e-05,
+      "loss": 3.5329,
+      "step": 248000
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 3.613527297973633,
+      "eval_runtime": 859.4422,
+      "eval_samples_per_second": 265.087,
+      "eval_steps_per_second": 33.137,
+      "step": 248234
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 1.4983621371286571e-05,
+      "loss": 3.5101,
+      "step": 248500
+    },
+    {
+      "epoch": 14.04,
+      "learning_rate": 1.4913023833728681e-05,
+      "loss": 3.488,
+      "step": 249000
+    },
+    {
+      "epoch": 14.07,
+      "learning_rate": 1.4842426296170792e-05,
+      "loss": 3.4851,
+      "step": 249500
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 1.4771828758612899e-05,
+      "loss": 3.4952,
+      "step": 250000
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 1.4701231221055008e-05,
+      "loss": 3.4913,
+      "step": 250500
+    },
+    {
+      "epoch": 14.16,
+      "learning_rate": 1.463063368349712e-05,
+      "loss": 3.491,
+      "step": 251000
+    },
+    {
+      "epoch": 14.18,
+      "learning_rate": 1.456003614593923e-05,
+      "loss": 3.4961,
+      "step": 251500
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 1.448943860838134e-05,
+      "loss": 3.4923,
+      "step": 252000
+    },
+    {
+      "epoch": 14.24,
+      "learning_rate": 1.4418841070823449e-05,
+      "loss": 3.5026,
+      "step": 252500
+    },
+    {
+      "epoch": 14.27,
+      "learning_rate": 1.4348243533265559e-05,
+      "loss": 3.4991,
+      "step": 253000
+    },
+    {
+      "epoch": 14.3,
+      "learning_rate": 1.427764599570767e-05,
+      "loss": 3.5018,
+      "step": 253500
+    },
+    {
+      "epoch": 14.33,
+      "learning_rate": 1.420704845814978e-05,
+      "loss": 3.4944,
+      "step": 254000
+    },
+    {
+      "epoch": 14.35,
+      "learning_rate": 1.413645092059189e-05,
+      "loss": 3.4994,
+      "step": 254500
+    },
+    {
+      "epoch": 14.38,
+      "learning_rate": 1.4065853383034e-05,
+      "loss": 3.5039,
+      "step": 255000
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 1.399525584547611e-05,
+      "loss": 3.5029,
+      "step": 255500
+    },
+    {
+      "epoch": 14.44,
+      "learning_rate": 1.392465830791822e-05,
+      "loss": 3.5038,
+      "step": 256000
+    },
+    {
+      "epoch": 14.47,
+      "learning_rate": 1.385406077036033e-05,
+      "loss": 3.5012,
+      "step": 256500
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 1.378346323280244e-05,
+      "loss": 3.5093,
+      "step": 257000
+    },
+    {
+      "epoch": 14.52,
+      "learning_rate": 1.3712865695244551e-05,
+      "loss": 3.5047,
+      "step": 257500
+    },
+    {
+      "epoch": 14.55,
+      "learning_rate": 1.3642268157686661e-05,
+      "loss": 3.5092,
+      "step": 258000
+    },
+    {
+      "epoch": 14.58,
+      "learning_rate": 1.357167062012877e-05,
+      "loss": 3.5083,
+      "step": 258500
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 1.350107308257088e-05,
+      "loss": 3.5023,
+      "step": 259000
+    },
+    {
+      "epoch": 14.64,
+      "learning_rate": 1.343047554501299e-05,
+      "loss": 3.5037,
+      "step": 259500
+    },
+    {
+      "epoch": 14.66,
+      "learning_rate": 1.3359878007455102e-05,
+      "loss": 3.507,
+      "step": 260000
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 1.3289280469897211e-05,
+      "loss": 3.5005,
+      "step": 260500
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 1.3218682932339321e-05,
+      "loss": 3.5107,
+      "step": 261000
+    },
+    {
+      "epoch": 14.75,
+      "learning_rate": 1.314808539478143e-05,
+      "loss": 3.5042,
+      "step": 261500
+    },
+    {
+      "epoch": 14.78,
+      "learning_rate": 1.3077487857223542e-05,
+      "loss": 3.5057,
+      "step": 262000
+    },
+    {
+      "epoch": 14.8,
+      "learning_rate": 1.3006890319665652e-05,
+      "loss": 3.5072,
+      "step": 262500
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 1.2936292782107762e-05,
+      "loss": 3.5061,
+      "step": 263000
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 1.2865695244549871e-05,
+      "loss": 3.5102,
+      "step": 263500
+    },
+    {
+      "epoch": 14.89,
+      "learning_rate": 1.2795097706991983e-05,
+      "loss": 3.5129,
+      "step": 264000
+    },
+    {
+      "epoch": 14.92,
+      "learning_rate": 1.2724500169434093e-05,
+      "loss": 3.5113,
+      "step": 264500
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 1.2653902631876199e-05,
+      "loss": 3.5039,
+      "step": 265000
+    },
+    {
+      "epoch": 14.97,
+      "learning_rate": 1.258330509431831e-05,
+      "loss": 3.5122,
+      "step": 265500
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 3.6014883518218994,
+      "eval_runtime": 858.9229,
+      "eval_samples_per_second": 265.247,
+      "eval_steps_per_second": 33.157,
+      "step": 265965
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 1.251270755676042e-05,
+      "loss": 3.5016,
+      "step": 266000
+    },
+    {
+      "epoch": 15.03,
+      "learning_rate": 1.2442110019202531e-05,
+      "loss": 3.4598,
+      "step": 266500
+    },
+    {
+      "epoch": 15.06,
+      "learning_rate": 1.2371512481644641e-05,
+      "loss": 3.4628,
+      "step": 267000
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 1.230091494408675e-05,
+      "loss": 3.4712,
+      "step": 267500
+    },
+    {
+      "epoch": 15.11,
+      "learning_rate": 1.2230317406528862e-05,
+      "loss": 3.4672,
+      "step": 268000
+    },
+    {
+      "epoch": 15.14,
+      "learning_rate": 1.2159719868970972e-05,
+      "loss": 3.4665,
+      "step": 268500
+    },
+    {
+      "epoch": 15.17,
+      "learning_rate": 1.208912233141308e-05,
+      "loss": 3.4665,
+      "step": 269000
+    },
+    {
+      "epoch": 15.2,
+      "learning_rate": 1.201852479385519e-05,
+      "loss": 3.4733,
+      "step": 269500
+    },
+    {
+      "epoch": 15.23,
+      "learning_rate": 1.1947927256297301e-05,
+      "loss": 3.4735,
+      "step": 270000
+    },
+    {
+      "epoch": 15.26,
+      "learning_rate": 1.187732971873941e-05,
+      "loss": 3.4755,
+      "step": 270500
+    },
+    {
+      "epoch": 15.28,
+      "learning_rate": 1.180673218118152e-05,
+      "loss": 3.4743,
+      "step": 271000
+    },
+    {
+      "epoch": 15.31,
+      "learning_rate": 1.173613464362363e-05,
+      "loss": 3.4799,
+      "step": 271500
+    },
+    {
+      "epoch": 15.34,
+      "learning_rate": 1.1665537106065742e-05,
+      "loss": 3.4805,
+      "step": 272000
+    },
+    {
+      "epoch": 15.37,
+      "learning_rate": 1.1594939568507851e-05,
+      "loss": 3.4771,
+      "step": 272500
+    },
+    {
+      "epoch": 15.4,
+      "learning_rate": 1.1524342030949961e-05,
+      "loss": 3.482,
+      "step": 273000
+    },
+    {
+      "epoch": 15.42,
+      "learning_rate": 1.1453744493392071e-05,
+      "loss": 3.4788,
+      "step": 273500
+    },
+    {
+      "epoch": 15.45,
+      "learning_rate": 1.138314695583418e-05,
+      "loss": 3.4807,
+      "step": 274000
+    },
+    {
+      "epoch": 15.48,
+      "learning_rate": 1.1312549418276292e-05,
+      "loss": 3.477,
+      "step": 274500
+    },
+    {
+      "epoch": 15.51,
+      "learning_rate": 1.1241951880718402e-05,
+      "loss": 3.4829,
+      "step": 275000
+    },
+    {
+      "epoch": 15.54,
+      "learning_rate": 1.1171354343160511e-05,
+      "loss": 3.4778,
+      "step": 275500
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 1.1100756805602621e-05,
+      "loss": 3.4838,
+      "step": 276000
+    },
+    {
+      "epoch": 15.59,
+      "learning_rate": 1.1030159268044731e-05,
+      "loss": 3.4746,
+      "step": 276500
+    },
+    {
+      "epoch": 15.62,
+      "learning_rate": 1.095956173048684e-05,
+      "loss": 3.4845,
+      "step": 277000
+    },
+    {
+      "epoch": 15.65,
+      "learning_rate": 1.088896419292895e-05,
+      "loss": 3.4892,
+      "step": 277500
+    },
+    {
+      "epoch": 15.68,
+      "learning_rate": 1.081836665537106e-05,
+      "loss": 3.4802,
+      "step": 278000
+    },
+    {
+      "epoch": 15.71,
+      "learning_rate": 1.0747769117813171e-05,
+      "loss": 3.4791,
+      "step": 278500
+    },
+    {
+      "epoch": 15.74,
+      "learning_rate": 1.0677171580255281e-05,
+      "loss": 3.4818,
+      "step": 279000
+    },
+    {
+      "epoch": 15.76,
+      "learning_rate": 1.0606574042697391e-05,
+      "loss": 3.478,
+      "step": 279500
+    },
+    {
+      "epoch": 15.79,
+      "learning_rate": 1.05359765051395e-05,
+      "loss": 3.479,
+      "step": 280000
+    },
+    {
+      "epoch": 15.82,
+      "learning_rate": 1.0465378967581612e-05,
+      "loss": 3.4834,
+      "step": 280500
+    },
+    {
+      "epoch": 15.85,
+      "learning_rate": 1.0394781430023722e-05,
+      "loss": 3.4799,
+      "step": 281000
+    },
+    {
+      "epoch": 15.88,
+      "learning_rate": 1.0324183892465831e-05,
+      "loss": 3.4798,
+      "step": 281500
+    },
+    {
+      "epoch": 15.9,
+      "learning_rate": 1.0253586354907941e-05,
+      "loss": 3.4855,
+      "step": 282000
+    },
+    {
+      "epoch": 15.93,
+      "learning_rate": 1.0182988817350053e-05,
+      "loss": 3.4856,
+      "step": 282500
+    },
+    {
+      "epoch": 15.96,
+      "learning_rate": 1.0112391279792162e-05,
+      "loss": 3.4834,
+      "step": 283000
+    },
+    {
+      "epoch": 15.99,
+      "learning_rate": 1.0041793742234272e-05,
+      "loss": 3.487,
+      "step": 283500
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 3.5896410942077637,
+      "eval_runtime": 859.6502,
+      "eval_samples_per_second": 265.023,
+      "eval_steps_per_second": 33.129,
+      "step": 283696
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 9.97119620467638e-06,
+      "loss": 3.46,
+      "step": 284000
+    },
+    {
+      "epoch": 16.05,
+      "learning_rate": 9.900598667118491e-06,
+      "loss": 3.4381,
+      "step": 284500
+    },
+    {
+      "epoch": 16.07,
+      "learning_rate": 9.830001129560601e-06,
+      "loss": 3.4432,
+      "step": 285000
+    },
+    {
+      "epoch": 16.1,
+      "learning_rate": 9.759403592002711e-06,
+      "loss": 3.4443,
+      "step": 285500
+    },
+    {
+      "epoch": 16.13,
+      "learning_rate": 9.68880605444482e-06,
+      "loss": 3.4481,
+      "step": 286000
+    },
+    {
+      "epoch": 16.16,
+      "learning_rate": 9.618208516886932e-06,
+      "loss": 3.4494,
+      "step": 286500
+    },
+    {
+      "epoch": 16.19,
+      "learning_rate": 9.547610979329042e-06,
+      "loss": 3.4531,
+      "step": 287000
+    },
+    {
+      "epoch": 16.21,
+      "learning_rate": 9.477013441771152e-06,
+      "loss": 3.4528,
+      "step": 287500
+    },
+    {
+      "epoch": 16.24,
+      "learning_rate": 9.406415904213261e-06,
+      "loss": 3.4534,
+      "step": 288000
+    },
+    {
+      "epoch": 16.27,
+      "learning_rate": 9.335818366655371e-06,
+      "loss": 3.4524,
+      "step": 288500
+    },
+    {
+      "epoch": 16.3,
+      "learning_rate": 9.265220829097482e-06,
+      "loss": 3.4561,
+      "step": 289000
+    },
+    {
+      "epoch": 16.33,
+      "learning_rate": 9.194623291539592e-06,
+      "loss": 3.4514,
+      "step": 289500
+    },
+    {
+      "epoch": 16.36,
+      "learning_rate": 9.124025753981702e-06,
+      "loss": 3.4528,
+      "step": 290000
+    },
+    {
+      "epoch": 16.38,
+      "learning_rate": 9.053428216423812e-06,
+      "loss": 3.4565,
+      "step": 290500
+    },
+    {
+      "epoch": 16.41,
+      "learning_rate": 8.982830678865923e-06,
+      "loss": 3.4601,
+      "step": 291000
+    },
+    {
+      "epoch": 16.44,
+      "learning_rate": 8.912233141308031e-06,
+      "loss": 3.4571,
+      "step": 291500
+    },
+    {
+      "epoch": 16.47,
+      "learning_rate": 8.84163560375014e-06,
+      "loss": 3.4532,
+      "step": 292000
+    },
+    {
+      "epoch": 16.5,
+      "learning_rate": 8.77103806619225e-06,
+      "loss": 3.4607,
+      "step": 292500
+    },
+    {
+      "epoch": 16.52,
+      "learning_rate": 8.700440528634362e-06,
+      "loss": 3.4616,
+      "step": 293000
+    },
+    {
+      "epoch": 16.55,
+      "learning_rate": 8.629842991076472e-06,
+      "loss": 3.4629,
+      "step": 293500
+    },
+    {
+      "epoch": 16.58,
+      "learning_rate": 8.559245453518581e-06,
+      "loss": 3.4557,
+      "step": 294000
+    },
+    {
+      "epoch": 16.61,
+      "learning_rate": 8.488647915960691e-06,
+      "loss": 3.4593,
+      "step": 294500
+    },
+    {
+      "epoch": 16.64,
+      "learning_rate": 8.418050378402802e-06,
+      "loss": 3.4591,
+      "step": 295000
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 8.347452840844912e-06,
+      "loss": 3.4619,
+      "step": 295500
+    },
+    {
+      "epoch": 16.69,
+      "learning_rate": 8.276855303287022e-06,
+      "loss": 3.4594,
+      "step": 296000
+    },
+    {
+      "epoch": 16.72,
+      "learning_rate": 8.206257765729132e-06,
+      "loss": 3.4588,
+      "step": 296500
+    },
+    {
+      "epoch": 16.75,
+      "learning_rate": 8.135660228171241e-06,
+      "loss": 3.4623,
+      "step": 297000
+    },
+    {
+      "epoch": 16.78,
+      "learning_rate": 8.065062690613353e-06,
+      "loss": 3.4601,
+      "step": 297500
+    },
+    {
+      "epoch": 16.81,
+      "learning_rate": 7.994465153055462e-06,
+      "loss": 3.4614,
+      "step": 298000
+    },
+    {
+      "epoch": 16.83,
+      "learning_rate": 7.923867615497572e-06,
+      "loss": 3.46,
+      "step": 298500
+    },
+    {
+      "epoch": 16.86,
+      "learning_rate": 7.853270077939682e-06,
+      "loss": 3.4612,
+      "step": 299000
+    },
+    {
+      "epoch": 16.89,
+      "learning_rate": 7.782672540381792e-06,
+      "loss": 3.4629,
+      "step": 299500
+    },
+    {
+      "epoch": 16.92,
+      "learning_rate": 7.712075002823901e-06,
+      "loss": 3.4621,
+      "step": 300000
+    },
+    {
+      "epoch": 16.95,
+      "learning_rate": 7.641477465266011e-06,
+      "loss": 3.4602,
+      "step": 300500
+    },
+    {
+      "epoch": 16.98,
+      "learning_rate": 7.570879927708122e-06,
+      "loss": 3.4656,
+      "step": 301000
+    },
+    {
+      "epoch": 17.0,
+      "eval_loss": 3.5812108516693115,
+      "eval_runtime": 859.9572,
+      "eval_samples_per_second": 264.928,
+      "eval_steps_per_second": 33.117,
+      "step": 301427
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 7.500282390150231e-06,
+      "loss": 3.4562,
+      "step": 301500
+    },
+    {
+      "epoch": 17.03,
+      "learning_rate": 7.429684852592342e-06,
+      "loss": 3.4315,
+      "step": 302000
+    },
+    {
+      "epoch": 17.06,
+      "learning_rate": 7.359087315034452e-06,
+      "loss": 3.4326,
+      "step": 302500
+    },
+    {
+      "epoch": 17.09,
+      "learning_rate": 7.288489777476562e-06,
+      "loss": 3.432,
+      "step": 303000
+    },
+    {
+      "epoch": 17.12,
+      "learning_rate": 7.217892239918672e-06,
+      "loss": 3.435,
+      "step": 303500
+    },
+    {
+      "epoch": 17.15,
+      "learning_rate": 7.1472947023607825e-06,
+      "loss": 3.4312,
+      "step": 304000
+    },
+    {
+      "epoch": 17.17,
+      "learning_rate": 7.076697164802892e-06,
+      "loss": 3.4332,
+      "step": 304500
+    },
+    {
+      "epoch": 17.2,
+      "learning_rate": 7.006099627245003e-06,
+      "loss": 3.433,
+      "step": 305000
+    },
+    {
+      "epoch": 17.23,
+      "learning_rate": 6.9355020896871125e-06,
+      "loss": 3.4289,
+      "step": 305500
+    },
+    {
+      "epoch": 17.26,
+      "learning_rate": 6.864904552129223e-06,
+      "loss": 3.4376,
+      "step": 306000
+    },
+    {
+      "epoch": 17.29,
+      "learning_rate": 6.794307014571331e-06,
+      "loss": 3.4358,
+      "step": 306500
+    },
+    {
+      "epoch": 17.31,
+      "learning_rate": 6.723709477013442e-06,
+      "loss": 3.434,
+      "step": 307000
+    },
+    {
+      "epoch": 17.34,
+      "learning_rate": 6.653111939455551e-06,
+      "loss": 3.4351,
+      "step": 307500
+    },
+    {
+      "epoch": 17.37,
+      "learning_rate": 6.582514401897662e-06,
+      "loss": 3.4357,
+      "step": 308000
+    },
+    {
+      "epoch": 17.4,
+      "learning_rate": 6.511916864339772e-06,
+      "loss": 3.4315,
+      "step": 308500
+    },
+    {
+      "epoch": 17.43,
+      "learning_rate": 6.441319326781882e-06,
+      "loss": 3.4405,
+      "step": 309000
+    },
+    {
+      "epoch": 17.46,
+      "learning_rate": 6.370721789223992e-06,
+      "loss": 3.4362,
+      "step": 309500
+    },
+    {
+      "epoch": 17.48,
+      "learning_rate": 6.3001242516661025e-06,
+      "loss": 3.4362,
+      "step": 310000
+    },
+    {
+      "epoch": 17.51,
+      "learning_rate": 6.229526714108212e-06,
+      "loss": 3.436,
+      "step": 310500
+    },
+    {
+      "epoch": 17.54,
+      "learning_rate": 6.158929176550323e-06,
+      "loss": 3.4355,
+      "step": 311000
+    },
+    {
+      "epoch": 17.57,
+      "learning_rate": 6.0883316389924325e-06,
+      "loss": 3.4379,
+      "step": 311500
+    },
+    {
+      "epoch": 17.6,
+      "learning_rate": 6.017734101434542e-06,
+      "loss": 3.437,
+      "step": 312000
+    },
+    {
+      "epoch": 17.62,
+      "learning_rate": 5.947136563876652e-06,
+      "loss": 3.4371,
+      "step": 312500
+    },
+    {
+      "epoch": 17.65,
+      "learning_rate": 5.8765390263187625e-06,
+      "loss": 3.4417,
+      "step": 313000
+    },
+    {
+      "epoch": 17.68,
+      "learning_rate": 5.805941488760872e-06,
+      "loss": 3.4445,
+      "step": 313500
+    },
+    {
+      "epoch": 17.71,
+      "learning_rate": 5.735343951202982e-06,
+      "loss": 3.4358,
+      "step": 314000
+    },
+    {
+      "epoch": 17.74,
+      "learning_rate": 5.6647464136450925e-06,
+      "loss": 3.44,
+      "step": 314500
+    },
+    {
+      "epoch": 17.77,
+      "learning_rate": 5.594148876087202e-06,
+      "loss": 3.4435,
+      "step": 315000
+    },
+    {
+      "epoch": 17.79,
+      "learning_rate": 5.523551338529313e-06,
+      "loss": 3.4419,
+      "step": 315500
+    },
+    {
+      "epoch": 17.82,
+      "learning_rate": 5.452953800971422e-06,
+      "loss": 3.4375,
+      "step": 316000
+    },
+    {
+      "epoch": 17.85,
+      "learning_rate": 5.382356263413532e-06,
+      "loss": 3.4409,
+      "step": 316500
+    },
+    {
+      "epoch": 17.88,
+      "learning_rate": 5.311758725855642e-06,
+      "loss": 3.4402,
+      "step": 317000
+    },
+    {
+      "epoch": 17.91,
+      "learning_rate": 5.2411611882977525e-06,
+      "loss": 3.4439,
+      "step": 317500
+    },
+    {
+      "epoch": 17.93,
+      "learning_rate": 5.170563650739862e-06,
+      "loss": 3.4414,
+      "step": 318000
+    },
+    {
+      "epoch": 17.96,
+      "learning_rate": 5.099966113181973e-06,
+      "loss": 3.4368,
+      "step": 318500
+    },
+    {
+      "epoch": 17.99,
+      "learning_rate": 5.0293685756240826e-06,
+      "loss": 3.4438,
+      "step": 319000
+    },
+    {
+      "epoch": 18.0,
+      "eval_loss": 3.5748109817504883,
+      "eval_runtime": 860.2197,
+      "eval_samples_per_second": 264.847,
+      "eval_steps_per_second": 33.107,
+      "step": 319158
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 4.958771038066193e-06,
+      "loss": 3.4245,
+      "step": 319500
+    },
+    {
+      "epoch": 18.05,
+      "learning_rate": 4.888173500508302e-06,
+      "loss": 3.4111,
+      "step": 320000
+    },
+    {
+      "epoch": 18.08,
+      "learning_rate": 4.8175759629504126e-06,
+      "loss": 3.4159,
+      "step": 320500
+    },
+    {
+      "epoch": 18.1,
+      "learning_rate": 4.746978425392522e-06,
+      "loss": 3.4101,
+      "step": 321000
+    },
+    {
+      "epoch": 18.13,
+      "learning_rate": 4.676380887834633e-06,
+      "loss": 3.416,
+      "step": 321500
+    },
+    {
+      "epoch": 18.16,
+      "learning_rate": 4.6057833502767426e-06,
+      "loss": 3.414,
+      "step": 322000
+    },
+    {
+      "epoch": 18.19,
+      "learning_rate": 4.535185812718853e-06,
+      "loss": 3.4111,
+      "step": 322500
+    },
+    {
+      "epoch": 18.22,
+      "learning_rate": 4.464588275160963e-06,
+      "loss": 3.4234,
+      "step": 323000
+    },
+    {
+      "epoch": 18.24,
+      "learning_rate": 4.393990737603073e-06,
+      "loss": 3.4169,
+      "step": 323500
+    },
+    {
+      "epoch": 18.27,
+      "learning_rate": 4.323393200045182e-06,
+      "loss": 3.4224,
+      "step": 324000
+    },
+    {
+      "epoch": 18.3,
+      "learning_rate": 4.252795662487293e-06,
+      "loss": 3.4242,
+      "step": 324500
+    },
+    {
+      "epoch": 18.33,
+      "learning_rate": 4.182198124929403e-06,
+      "loss": 3.4177,
+      "step": 325000
+    },
+    {
+      "epoch": 18.36,
+      "learning_rate": 4.111600587371512e-06,
+      "loss": 3.4214,
+      "step": 325500
+    },
+    {
+      "epoch": 18.39,
+      "learning_rate": 4.041003049813623e-06,
+      "loss": 3.4213,
+      "step": 326000
+    },
+    {
+      "epoch": 18.41,
+      "learning_rate": 3.970405512255733e-06,
+      "loss": 3.4261,
+      "step": 326500
+    },
+    {
+      "epoch": 18.44,
+      "learning_rate": 3.899807974697843e-06,
+      "loss": 3.4181,
+      "step": 327000
+    },
+    {
+      "epoch": 18.47,
+      "learning_rate": 3.829210437139952e-06,
+      "loss": 3.4202,
+      "step": 327500
+    },
+    {
+      "epoch": 18.5,
+      "learning_rate": 3.7586128995820626e-06,
+      "loss": 3.4213,
+      "step": 328000
+    },
+    {
+      "epoch": 18.53,
+      "learning_rate": 3.6880153620241728e-06,
+      "loss": 3.4289,
+      "step": 328500
+    },
+    {
+      "epoch": 18.56,
+      "learning_rate": 3.617417824466283e-06,
+      "loss": 3.4254,
+      "step": 329000
+    },
+    {
+      "epoch": 18.58,
+      "learning_rate": 3.546820286908393e-06,
+      "loss": 3.4221,
+      "step": 329500
+    },
+    {
+      "epoch": 18.61,
+      "learning_rate": 3.476222749350503e-06,
+      "loss": 3.4201,
+      "step": 330000
+    },
+    {
+      "epoch": 18.64,
+      "learning_rate": 3.405625211792613e-06,
+      "loss": 3.4238,
+      "step": 330500
+    },
+    {
+      "epoch": 18.67,
+      "learning_rate": 3.3350276742347226e-06,
+      "loss": 3.4235,
+      "step": 331000
+    },
+    {
+      "epoch": 18.7,
+      "learning_rate": 3.2644301366768328e-06,
+      "loss": 3.4176,
+      "step": 331500
+    },
+    {
+      "epoch": 18.72,
+      "learning_rate": 3.193832599118943e-06,
+      "loss": 3.4208,
+      "step": 332000
+    },
+    {
+      "epoch": 18.75,
+      "learning_rate": 3.1232350615610526e-06,
+      "loss": 3.423,
+      "step": 332500
+    },
+    {
+      "epoch": 18.78,
+      "learning_rate": 3.0526375240031628e-06,
+      "loss": 3.4212,
+      "step": 333000
+    },
+    {
+      "epoch": 18.81,
+      "learning_rate": 2.982039986445273e-06,
+      "loss": 3.4249,
+      "step": 333500
+    },
+    {
+      "epoch": 18.84,
+      "learning_rate": 2.9114424488873826e-06,
+      "loss": 3.4181,
+      "step": 334000
+    },
+    {
+      "epoch": 18.87,
+      "learning_rate": 2.840844911329493e-06,
+      "loss": 3.4228,
+      "step": 334500
+    },
+    {
+      "epoch": 18.89,
+      "learning_rate": 2.770247373771603e-06,
+      "loss": 3.4243,
+      "step": 335000
+    },
+    {
+      "epoch": 18.92,
+      "learning_rate": 2.699649836213713e-06,
+      "loss": 3.4253,
+      "step": 335500
+    },
+    {
+      "epoch": 18.95,
+      "learning_rate": 2.629052298655823e-06,
+      "loss": 3.4184,
+      "step": 336000
+    },
+    {
+      "epoch": 18.98,
+      "learning_rate": 2.558454761097933e-06,
+      "loss": 3.427,
+      "step": 336500
+    },
+    {
+      "epoch": 19.0,
+      "eval_loss": 3.569897413253784,
+      "eval_runtime": 860.0705,
+      "eval_samples_per_second": 264.893,
+      "eval_steps_per_second": 33.112,
+      "step": 336889
+    },
+    {
+      "epoch": 19.01,
+      "learning_rate": 2.487857223540043e-06,
+      "loss": 3.4203,
+      "step": 337000
+    },
+    {
+      "epoch": 19.03,
+      "learning_rate": 2.4172596859821532e-06,
+      "loss": 3.4046,
+      "step": 337500
+    },
+    {
+      "epoch": 19.06,
+      "learning_rate": 2.346662148424263e-06,
+      "loss": 3.4027,
+      "step": 338000
+    },
+    {
+      "epoch": 19.09,
+      "learning_rate": 2.276064610866373e-06,
+      "loss": 3.4071,
+      "step": 338500
+    },
+    {
+      "epoch": 19.12,
+      "learning_rate": 2.2054670733084832e-06,
+      "loss": 3.4082,
+      "step": 339000
+    },
+    {
+      "epoch": 19.15,
+      "learning_rate": 2.134869535750593e-06,
+      "loss": 3.4071,
+      "step": 339500
+    },
+    {
+      "epoch": 19.18,
+      "learning_rate": 2.064271998192703e-06,
+      "loss": 3.4052,
+      "step": 340000
+    },
+    {
+      "epoch": 19.2,
+      "learning_rate": 1.9936744606348133e-06,
+      "loss": 3.4072,
+      "step": 340500
+    },
+    {
+      "epoch": 19.23,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 3.4036,
+      "step": 341000
+    },
+    {
+      "epoch": 19.26,
+      "learning_rate": 1.8524793855190331e-06,
+      "loss": 3.4015,
+      "step": 341500
+    },
+    {
+      "epoch": 19.29,
+      "learning_rate": 1.781881847961143e-06,
+      "loss": 3.4058,
+      "step": 342000
+    },
+    {
+      "epoch": 19.32,
+      "learning_rate": 1.7112843104032532e-06,
+      "loss": 3.4041,
+      "step": 342500
+    },
+    {
+      "epoch": 19.34,
+      "learning_rate": 1.6406867728453633e-06,
+      "loss": 3.4061,
+      "step": 343000
+    },
+    {
+      "epoch": 19.37,
+      "learning_rate": 1.570089235287473e-06,
+      "loss": 3.405,
+      "step": 343500
+    },
+    {
+      "epoch": 19.4,
+      "learning_rate": 1.4994916977295832e-06,
+      "loss": 3.4084,
+      "step": 344000
+    },
+    {
+      "epoch": 19.43,
+      "learning_rate": 1.4288941601716933e-06,
+      "loss": 3.4033,
+      "step": 344500
+    },
+    {
+      "epoch": 19.46,
+      "learning_rate": 1.3582966226138033e-06,
+      "loss": 3.4063,
+      "step": 345000
+    },
+    {
+      "epoch": 19.49,
+      "learning_rate": 1.2876990850559134e-06,
+      "loss": 3.4113,
+      "step": 345500
+    },
+    {
+      "epoch": 19.51,
+      "learning_rate": 1.2171015474980234e-06,
+      "loss": 3.408,
+      "step": 346000
+    },
+    {
+      "epoch": 19.54,
+      "learning_rate": 1.1465040099401335e-06,
+      "loss": 3.4107,
+      "step": 346500
+    },
+    {
+      "epoch": 19.57,
+      "learning_rate": 1.0759064723822434e-06,
+      "loss": 3.4072,
+      "step": 347000
+    },
+    {
+      "epoch": 19.6,
+      "learning_rate": 1.0053089348243534e-06,
+      "loss": 3.4045,
+      "step": 347500
+    },
+    {
+      "epoch": 19.63,
+      "learning_rate": 9.347113972664634e-07,
+      "loss": 3.4063,
+      "step": 348000
+    },
+    {
+      "epoch": 19.65,
+      "learning_rate": 8.641138597085733e-07,
+      "loss": 3.4104,
+      "step": 348500
+    },
+    {
+      "epoch": 19.68,
+      "learning_rate": 7.935163221506835e-07,
+      "loss": 3.4067,
+      "step": 349000
+    },
+    {
+      "epoch": 19.71,
+      "learning_rate": 7.229187845927934e-07,
+      "loss": 3.4046,
+      "step": 349500
+    },
+    {
+      "epoch": 19.74,
+      "learning_rate": 6.523212470349035e-07,
+      "loss": 3.403,
+      "step": 350000
+    },
+    {
+      "epoch": 19.77,
+      "learning_rate": 1.7069980994674747e-05,
+      "loss": 3.4261,
+      "step": 350500
+    },
+    {
+      "epoch": 19.8,
+      "learning_rate": 1.7022938110381427e-05,
+      "loss": 3.4388,
+      "step": 351000
+    },
+    {
+      "epoch": 19.82,
+      "learning_rate": 1.6975895226088104e-05,
+      "loss": 3.4412,
+      "step": 351500
+    },
+    {
+      "epoch": 19.85,
+      "learning_rate": 1.692885234179478e-05,
+      "loss": 3.4422,
+      "step": 352000
+    },
+    {
+      "epoch": 19.88,
+      "learning_rate": 1.688180945750146e-05,
+      "loss": 3.4486,
+      "step": 352500
+    },
+    {
+      "epoch": 19.91,
+      "learning_rate": 1.6834766573208138e-05,
+      "loss": 3.4484,
+      "step": 353000
+    },
+    {
+      "epoch": 19.94,
+      "learning_rate": 1.6787723688914815e-05,
+      "loss": 3.453,
+      "step": 353500
+    },
+    {
+      "epoch": 19.97,
+      "learning_rate": 1.6740680804621495e-05,
+      "loss": 3.4555,
+      "step": 354000
+    },
+    {
+      "epoch": 19.99,
+      "learning_rate": 1.6693637920328172e-05,
+      "loss": 3.4618,
+      "step": 354500
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 3.5916237831115723,
+      "eval_runtime": 859.4544,
+      "eval_samples_per_second": 265.083,
+      "eval_steps_per_second": 33.136,
+      "step": 354620
+    },
+    {
+      "epoch": 20.02,
+      "learning_rate": 1.6646595036034852e-05,
+      "loss": 3.4385,
+      "step": 355000
+    },
+    {
+      "epoch": 20.05,
+      "learning_rate": 1.659955215174153e-05,
+      "loss": 3.4409,
+      "step": 355500
+    },
+    {
+      "epoch": 20.08,
+      "learning_rate": 1.6552509267448206e-05,
+      "loss": 3.4432,
+      "step": 356000
+    },
+    {
+      "epoch": 20.11,
+      "learning_rate": 1.6505466383154886e-05,
+      "loss": 3.4478,
+      "step": 356500
+    },
+    {
+      "epoch": 20.13,
+      "learning_rate": 1.6458423498861563e-05,
+      "loss": 3.4471,
+      "step": 357000
+    },
+    {
+      "epoch": 20.16,
+      "learning_rate": 1.6411380614568243e-05,
+      "loss": 3.4482,
+      "step": 357500
+    },
+    {
+      "epoch": 20.19,
+      "learning_rate": 1.636433773027492e-05,
+      "loss": 3.4443,
+      "step": 358000
+    },
+    {
+      "epoch": 20.22,
+      "learning_rate": 1.6317294845981597e-05,
+      "loss": 3.4487,
+      "step": 358500
+    },
+    {
+      "epoch": 20.25,
+      "learning_rate": 1.6270251961688277e-05,
+      "loss": 3.4514,
+      "step": 359000
+    },
+    {
+      "epoch": 20.28,
+      "learning_rate": 1.6223209077394954e-05,
+      "loss": 3.4481,
+      "step": 359500
+    },
+    {
+      "epoch": 20.3,
+      "learning_rate": 1.6176166193101635e-05,
+      "loss": 3.4513,
+      "step": 360000
+    },
+    {
+      "epoch": 20.33,
+      "learning_rate": 1.612912330880831e-05,
+      "loss": 3.4527,
+      "step": 360500
+    },
+    {
+      "epoch": 20.36,
+      "learning_rate": 1.6082080424514988e-05,
+      "loss": 3.4545,
+      "step": 361000
+    },
+    {
+      "epoch": 20.39,
+      "learning_rate": 1.603503754022167e-05,
+      "loss": 3.4538,
+      "step": 361500
+    },
+    {
+      "epoch": 20.42,
+      "learning_rate": 1.5987994655928345e-05,
+      "loss": 3.4566,
+      "step": 362000
+    },
+    {
+      "epoch": 20.44,
+      "learning_rate": 1.5940951771635022e-05,
+      "loss": 3.462,
+      "step": 362500
+    },
+    {
+      "epoch": 20.47,
+      "learning_rate": 1.5893908887341703e-05,
+      "loss": 3.454,
+      "step": 363000
+    },
+    {
+      "epoch": 20.5,
+      "learning_rate": 1.584686600304838e-05,
+      "loss": 3.4643,
+      "step": 363500
+    },
+    {
+      "epoch": 20.53,
+      "learning_rate": 1.579982311875506e-05,
+      "loss": 3.4671,
+      "step": 364000
+    },
+    {
+      "epoch": 20.56,
+      "learning_rate": 1.5752780234461737e-05,
+      "loss": 3.464,
+      "step": 364500
+    },
+    {
+      "epoch": 20.59,
+      "learning_rate": 1.5705737350168413e-05,
+      "loss": 3.4596,
+      "step": 365000
+    },
+    {
+      "epoch": 20.61,
+      "learning_rate": 1.5658694465875094e-05,
+      "loss": 3.464,
+      "step": 365500
+    },
+    {
+      "epoch": 20.64,
+      "learning_rate": 1.561165158158177e-05,
+      "loss": 3.4623,
+      "step": 366000
+    },
+    {
+      "epoch": 20.67,
+      "learning_rate": 1.556460869728845e-05,
+      "loss": 3.4662,
+      "step": 366500
+    },
+    {
+      "epoch": 20.7,
+      "learning_rate": 1.5517565812995128e-05,
+      "loss": 3.4654,
+      "step": 367000
+    },
+    {
+      "epoch": 20.73,
+      "learning_rate": 1.5470522928701805e-05,
+      "loss": 3.4657,
+      "step": 367500
+    },
+    {
+      "epoch": 20.75,
+      "learning_rate": 1.5423480044408485e-05,
+      "loss": 3.4667,
+      "step": 368000
+    },
+    {
+      "epoch": 20.78,
+      "learning_rate": 1.5376437160115162e-05,
+      "loss": 3.4694,
+      "step": 368500
+    },
+    {
+      "epoch": 20.81,
+      "learning_rate": 1.532939427582184e-05,
+      "loss": 3.466,
+      "step": 369000
+    },
+    {
+      "epoch": 20.84,
+      "learning_rate": 1.528235139152852e-05,
+      "loss": 3.4692,
+      "step": 369500
+    },
+    {
+      "epoch": 20.87,
+      "learning_rate": 1.5235308507235196e-05,
+      "loss": 3.4678,
+      "step": 370000
+    },
+    {
+      "epoch": 20.9,
+      "learning_rate": 1.5188265622941874e-05,
+      "loss": 3.471,
+      "step": 370500
+    },
+    {
+      "epoch": 20.92,
+      "learning_rate": 1.5141222738648553e-05,
+      "loss": 3.4701,
+      "step": 371000
+    },
+    {
+      "epoch": 20.95,
+      "learning_rate": 1.5094179854355231e-05,
+      "loss": 3.4677,
+      "step": 371500
+    },
+    {
+      "epoch": 20.98,
+      "learning_rate": 1.5047136970061908e-05,
+      "loss": 3.4688,
+      "step": 372000
+    },
+    {
+      "epoch": 21.0,
+      "eval_loss": 3.582139015197754,
+      "eval_runtime": 860.6638,
+      "eval_samples_per_second": 264.711,
+      "eval_steps_per_second": 33.09,
+      "step": 372351
+    },
+    {
+      "epoch": 21.01,
+      "learning_rate": 1.5000094085768587e-05,
+      "loss": 3.4558,
+      "step": 372500
+    },
+    {
+      "epoch": 21.04,
+      "learning_rate": 1.4953051201475265e-05,
+      "loss": 3.4203,
+      "step": 373000
+    },
+    {
+      "epoch": 21.06,
+      "learning_rate": 1.4906008317181944e-05,
+      "loss": 3.4192,
+      "step": 373500
+    },
+    {
+      "epoch": 21.09,
+      "learning_rate": 1.4858965432888623e-05,
+      "loss": 3.4208,
+      "step": 374000
+    },
+    {
+      "epoch": 21.12,
+      "learning_rate": 1.48119225485953e-05,
+      "loss": 3.4236,
+      "step": 374500
+    },
+    {
+      "epoch": 21.15,
+      "learning_rate": 1.4764879664301978e-05,
+      "loss": 3.4278,
+      "step": 375000
+    },
+    {
+      "epoch": 21.18,
+      "learning_rate": 1.4717836780008657e-05,
+      "loss": 3.4287,
+      "step": 375500
+    },
+    {
+      "epoch": 21.21,
+      "learning_rate": 1.4670793895715335e-05,
+      "loss": 3.4333,
+      "step": 376000
+    },
+    {
+      "epoch": 21.23,
+      "learning_rate": 1.4623751011422012e-05,
+      "loss": 3.4288,
+      "step": 376500
+    },
+    {
+      "epoch": 21.26,
+      "learning_rate": 1.457670812712869e-05,
+      "loss": 3.4313,
+      "step": 377000
+    },
+    {
+      "epoch": 21.29,
+      "learning_rate": 1.452966524283537e-05,
+      "loss": 3.4357,
+      "step": 377500
+    },
+    {
+      "epoch": 21.32,
+      "learning_rate": 1.4482622358542048e-05,
+      "loss": 3.4356,
+      "step": 378000
+    },
+    {
+      "epoch": 21.35,
+      "learning_rate": 1.4435579474248725e-05,
+      "loss": 3.4386,
+      "step": 378500
+    },
+    {
+      "epoch": 21.37,
+      "learning_rate": 1.4388536589955403e-05,
+      "loss": 3.4449,
+      "step": 379000
+    },
+    {
+      "epoch": 21.4,
+      "learning_rate": 1.4341493705662082e-05,
+      "loss": 3.4411,
+      "step": 379500
+    },
+    {
+      "epoch": 21.43,
+      "learning_rate": 1.429445082136876e-05,
+      "loss": 3.441,
+      "step": 380000
+    },
+    {
+      "epoch": 21.46,
+      "learning_rate": 1.4247407937075439e-05,
+      "loss": 3.4424,
+      "step": 380500
+    },
+    {
+      "epoch": 21.49,
+      "learning_rate": 1.4200365052782116e-05,
+      "loss": 3.4444,
+      "step": 381000
+    },
+    {
+      "epoch": 21.52,
+      "learning_rate": 1.4153322168488794e-05,
+      "loss": 3.4467,
+      "step": 381500
+    },
+    {
+      "epoch": 21.54,
+      "learning_rate": 1.4106279284195473e-05,
+      "loss": 3.4413,
+      "step": 382000
+    },
+    {
+      "epoch": 21.57,
+      "learning_rate": 1.4059236399902152e-05,
+      "loss": 3.4435,
+      "step": 382500
+    },
+    {
+      "epoch": 21.6,
+      "learning_rate": 1.4012193515608828e-05,
+      "loss": 3.4398,
+      "step": 383000
+    },
+    {
+      "epoch": 21.63,
+      "learning_rate": 1.3965150631315507e-05,
+      "loss": 3.4444,
+      "step": 383500
+    },
+    {
+      "epoch": 21.66,
+      "learning_rate": 1.3918107747022186e-05,
+      "loss": 3.45,
+      "step": 384000
+    },
+    {
+      "epoch": 21.69,
+      "learning_rate": 1.3871064862728864e-05,
+      "loss": 3.4467,
+      "step": 384500
+    },
+    {
+      "epoch": 21.71,
+      "learning_rate": 1.3824021978435541e-05,
+      "loss": 3.4433,
+      "step": 385000
+    },
+    {
+      "epoch": 21.74,
+      "learning_rate": 1.377697909414222e-05,
+      "loss": 3.4472,
+      "step": 385500
+    },
+    {
+      "epoch": 21.77,
+      "learning_rate": 1.3729936209848898e-05,
+      "loss": 3.446,
+      "step": 386000
+    },
+    {
+      "epoch": 21.8,
+      "learning_rate": 1.3682893325555577e-05,
+      "loss": 3.4479,
+      "step": 386500
+    },
+    {
+      "epoch": 21.83,
+      "learning_rate": 1.3635850441262255e-05,
+      "loss": 3.4484,
+      "step": 387000
+    },
+    {
+      "epoch": 21.85,
+      "learning_rate": 1.3588807556968932e-05,
+      "loss": 3.4534,
+      "step": 387500
+    },
+    {
+      "epoch": 21.88,
+      "learning_rate": 1.354176467267561e-05,
+      "loss": 3.4485,
+      "step": 388000
+    },
+    {
+      "epoch": 21.91,
+      "learning_rate": 1.349472178838229e-05,
+      "loss": 3.4465,
+      "step": 388500
+    },
+    {
+      "epoch": 21.94,
+      "learning_rate": 1.3447678904088968e-05,
+      "loss": 3.4475,
+      "step": 389000
+    },
+    {
+      "epoch": 21.97,
+      "learning_rate": 1.3400636019795645e-05,
+      "loss": 3.4553,
+      "step": 389500
+    },
+    {
+      "epoch": 22.0,
+      "learning_rate": 1.3353593135502323e-05,
+      "loss": 3.4502,
+      "step": 390000
+    },
+    {
+      "epoch": 22.0,
+      "eval_loss": 3.5741803646087646,
+      "eval_runtime": 861.811,
+      "eval_samples_per_second": 264.358,
+      "eval_steps_per_second": 33.046,
+      "step": 390082
+    },
+    {
+      "epoch": 22.02,
+      "learning_rate": 1.3306550251209002e-05,
+      "loss": 3.4106,
+      "step": 390500
+    },
+    {
+      "epoch": 22.05,
+      "learning_rate": 1.325950736691568e-05,
+      "loss": 3.4081,
+      "step": 391000
+    },
+    {
+      "epoch": 22.08,
+      "learning_rate": 1.3212464482622359e-05,
+      "loss": 3.4116,
+      "step": 391500
+    },
+    {
+      "epoch": 22.11,
+      "learning_rate": 1.3165421598329036e-05,
+      "loss": 3.4073,
+      "step": 392000
+    },
+    {
+      "epoch": 22.14,
+      "learning_rate": 1.3118378714035714e-05,
+      "loss": 3.4114,
+      "step": 392500
+    },
+    {
+      "epoch": 22.16,
+      "learning_rate": 1.3071335829742393e-05,
+      "loss": 3.412,
+      "step": 393000
+    },
+    {
+      "epoch": 22.19,
+      "learning_rate": 1.3024292945449072e-05,
+      "loss": 3.4129,
+      "step": 393500
+    },
+    {
+      "epoch": 22.22,
+      "learning_rate": 1.2977250061155748e-05,
+      "loss": 3.4169,
+      "step": 394000
+    },
+    {
+      "epoch": 22.25,
+      "learning_rate": 1.2930207176862427e-05,
+      "loss": 3.4106,
+      "step": 394500
+    },
+    {
+      "epoch": 22.28,
+      "learning_rate": 1.2883164292569106e-05,
+      "loss": 3.4163,
+      "step": 395000
+    },
+    {
+      "epoch": 22.31,
+      "learning_rate": 1.2836121408275784e-05,
+      "loss": 3.4197,
+      "step": 395500
+    },
+    {
+      "epoch": 22.33,
+      "learning_rate": 1.2789078523982461e-05,
+      "loss": 3.4195,
+      "step": 396000
+    },
+    {
+      "epoch": 22.36,
+      "learning_rate": 1.274203563968914e-05,
+      "loss": 3.4128,
+      "step": 396500
+    },
+    {
+      "epoch": 22.39,
+      "learning_rate": 1.2694992755395818e-05,
+      "loss": 3.4224,
+      "step": 397000
+    },
+    {
+      "epoch": 22.42,
+      "learning_rate": 1.2647949871102497e-05,
+      "loss": 3.422,
+      "step": 397500
+    },
+    {
+      "epoch": 22.45,
+      "learning_rate": 1.2600906986809175e-05,
+      "loss": 3.424,
+      "step": 398000
+    },
+    {
+      "epoch": 22.47,
+      "learning_rate": 1.2553864102515856e-05,
+      "loss": 3.424,
+      "step": 398500
+    },
+    {
+      "epoch": 22.5,
+      "learning_rate": 1.2506821218222534e-05,
+      "loss": 3.4251,
+      "step": 399000
+    },
+    {
+      "epoch": 22.53,
+      "learning_rate": 1.2459778333929211e-05,
+      "loss": 3.4241,
+      "step": 399500
+    },
+    {
+      "epoch": 22.56,
+      "learning_rate": 1.2412735449635888e-05,
+      "loss": 3.4282,
+      "step": 400000
+    },
+    {
+      "epoch": 22.59,
+      "learning_rate": 1.2365692565342566e-05,
+      "loss": 3.4269,
+      "step": 400500
+    },
+    {
+      "epoch": 22.62,
+      "learning_rate": 1.2318649681049245e-05,
+      "loss": 3.4258,
+      "step": 401000
+    },
+    {
+      "epoch": 22.64,
+      "learning_rate": 1.2271606796755924e-05,
+      "loss": 3.4238,
+      "step": 401500
+    },
+    {
+      "epoch": 22.67,
+      "learning_rate": 1.22245639124626e-05,
+      "loss": 3.4293,
+      "step": 402000
+    },
+    {
+      "epoch": 22.7,
+      "learning_rate": 1.2177521028169279e-05,
+      "loss": 3.4269,
+      "step": 402500
+    },
+    {
+      "epoch": 22.73,
+      "learning_rate": 1.2130478143875958e-05,
+      "loss": 3.424,
+      "step": 403000
+    },
+    {
+      "epoch": 22.76,
+      "learning_rate": 1.2083435259582636e-05,
+      "loss": 3.432,
+      "step": 403500
+    },
+    {
+      "epoch": 22.78,
+      "learning_rate": 1.2036392375289315e-05,
+      "loss": 3.4304,
+      "step": 404000
+    },
+    {
+      "epoch": 22.81,
+      "learning_rate": 1.1989349490995992e-05,
+      "loss": 3.4302,
+      "step": 404500
+    },
+    {
+      "epoch": 22.84,
+      "learning_rate": 1.194230660670267e-05,
+      "loss": 3.4301,
+      "step": 405000
+    },
+    {
+      "epoch": 22.87,
+      "learning_rate": 1.1895263722409349e-05,
+      "loss": 3.4315,
+      "step": 405500
+    },
+    {
+      "epoch": 22.9,
+      "learning_rate": 1.1848220838116027e-05,
+      "loss": 3.4274,
+      "step": 406000
+    },
+    {
+      "epoch": 22.93,
+      "learning_rate": 1.1801177953822704e-05,
+      "loss": 3.4335,
+      "step": 406500
+    },
+    {
+      "epoch": 22.95,
+      "learning_rate": 1.1754135069529383e-05,
+      "loss": 3.4301,
+      "step": 407000
+    },
+    {
+      "epoch": 22.98,
+      "learning_rate": 1.1707092185236061e-05,
+      "loss": 3.4303,
+      "step": 407500
+    },
+    {
+      "epoch": 23.0,
+      "eval_loss": 3.566288709640503,
+      "eval_runtime": 859.4108,
+      "eval_samples_per_second": 265.097,
+      "eval_steps_per_second": 33.138,
+      "step": 407813
+    },
+    {
+      "epoch": 23.01,
+      "learning_rate": 1.166004930094274e-05,
+      "loss": 3.4179,
+      "step": 408000
+    },
+    {
+      "epoch": 23.04,
+      "learning_rate": 1.1613006416649418e-05,
+      "loss": 3.3839,
+      "step": 408500
+    },
+    {
+      "epoch": 23.07,
+      "learning_rate": 1.1565963532356095e-05,
+      "loss": 3.3893,
+      "step": 409000
+    },
+    {
+      "epoch": 23.1,
+      "learning_rate": 1.1518920648062774e-05,
+      "loss": 3.3931,
+      "step": 409500
+    },
+    {
+      "epoch": 23.12,
+      "learning_rate": 1.1471877763769452e-05,
+      "loss": 3.3943,
+      "step": 410000
+    },
+    {
+      "epoch": 23.15,
+      "learning_rate": 1.1424834879476131e-05,
+      "loss": 3.396,
+      "step": 410500
+    },
+    {
+      "epoch": 23.18,
+      "learning_rate": 1.1377791995182808e-05,
+      "loss": 3.3988,
+      "step": 411000
+    },
+    {
+      "epoch": 23.21,
+      "learning_rate": 1.1330749110889487e-05,
+      "loss": 3.3979,
+      "step": 411500
+    },
+    {
+      "epoch": 23.24,
+      "learning_rate": 1.1283706226596167e-05,
+      "loss": 3.4025,
+      "step": 412000
+    },
+    {
+      "epoch": 23.26,
+      "learning_rate": 1.1236663342302844e-05,
+      "loss": 3.4016,
+      "step": 412500
+    },
+    {
+      "epoch": 23.29,
+      "learning_rate": 1.1189620458009522e-05,
+      "loss": 3.4051,
+      "step": 413000
+    },
+    {
+      "epoch": 23.32,
+      "learning_rate": 1.11425775737162e-05,
+      "loss": 3.4066,
+      "step": 413500
+    },
+    {
+      "epoch": 23.35,
+      "learning_rate": 1.109553468942288e-05,
+      "loss": 3.4032,
+      "step": 414000
+    },
+    {
+      "epoch": 23.38,
+      "learning_rate": 1.1048491805129558e-05,
+      "loss": 3.401,
+      "step": 414500
+    },
+    {
+      "epoch": 23.41,
+      "learning_rate": 1.1001448920836235e-05,
+      "loss": 3.4062,
+      "step": 415000
+    },
+    {
+      "epoch": 23.43,
+      "learning_rate": 1.0954406036542913e-05,
+      "loss": 3.4046,
+      "step": 415500
+    },
+    {
+      "epoch": 23.46,
+      "learning_rate": 1.0907363152249592e-05,
+      "loss": 3.403,
+      "step": 416000
+    },
+    {
+      "epoch": 23.49,
+      "learning_rate": 1.086032026795627e-05,
+      "loss": 3.4082,
+      "step": 416500
+    },
+    {
+      "epoch": 23.52,
+      "learning_rate": 1.0813277383662947e-05,
+      "loss": 3.4062,
+      "step": 417000
+    },
+    {
+      "epoch": 23.55,
+      "learning_rate": 1.0766234499369626e-05,
+      "loss": 3.4059,
+      "step": 417500
+    },
+    {
+      "epoch": 23.57,
+      "learning_rate": 1.0719191615076305e-05,
+      "loss": 3.4051,
+      "step": 418000
+    },
+    {
+      "epoch": 23.6,
+      "learning_rate": 1.0672148730782983e-05,
+      "loss": 3.4084,
+      "step": 418500
+    },
+    {
+      "epoch": 23.63,
+      "learning_rate": 1.062510584648966e-05,
+      "loss": 3.4098,
+      "step": 419000
+    },
+    {
+      "epoch": 23.66,
+      "learning_rate": 1.0578062962196339e-05,
+      "loss": 3.4129,
+      "step": 419500
+    },
+    {
+      "epoch": 23.69,
+      "learning_rate": 1.0531020077903017e-05,
+      "loss": 3.4123,
+      "step": 420000
+    },
+    {
+      "epoch": 23.72,
+      "learning_rate": 1.0483977193609696e-05,
+      "loss": 3.4118,
+      "step": 420500
+    },
+    {
+      "epoch": 23.74,
+      "learning_rate": 1.0436934309316374e-05,
+      "loss": 3.4061,
+      "step": 421000
+    },
+    {
+      "epoch": 23.77,
+      "learning_rate": 1.0389891425023051e-05,
+      "loss": 3.4157,
+      "step": 421500
+    },
+    {
+      "epoch": 23.8,
+      "learning_rate": 1.034284854072973e-05,
+      "loss": 3.4088,
+      "step": 422000
+    },
+    {
+      "epoch": 23.83,
+      "learning_rate": 1.0295805656436408e-05,
+      "loss": 3.4107,
+      "step": 422500
+    },
+    {
+      "epoch": 23.86,
+      "learning_rate": 1.0248762772143087e-05,
+      "loss": 3.414,
+      "step": 423000
+    },
+    {
+      "epoch": 23.88,
+      "learning_rate": 1.0201719887849764e-05,
+      "loss": 3.4169,
+      "step": 423500
+    },
+    {
+      "epoch": 23.91,
+      "learning_rate": 1.0154677003556442e-05,
+      "loss": 3.4168,
+      "step": 424000
+    },
+    {
+      "epoch": 23.94,
+      "learning_rate": 1.010763411926312e-05,
+      "loss": 3.4131,
+      "step": 424500
+    },
+    {
+      "epoch": 23.97,
+      "learning_rate": 1.00605912349698e-05,
+      "loss": 3.4162,
+      "step": 425000
+    },
+    {
+      "epoch": 24.0,
+      "learning_rate": 1.0013548350676476e-05,
+      "loss": 3.4126,
+      "step": 425500
+    },
+    {
+      "epoch": 24.0,
+      "eval_loss": 3.560105323791504,
+      "eval_runtime": 859.9342,
+      "eval_samples_per_second": 264.935,
+      "eval_steps_per_second": 33.118,
+      "step": 425544
+    },
+    {
+      "epoch": 24.03,
+      "learning_rate": 9.966505466383155e-06,
+      "loss": 3.3813,
+      "step": 426000
+    },
+    {
+      "epoch": 24.05,
+      "learning_rate": 9.919462582089833e-06,
+      "loss": 3.3754,
+      "step": 426500
+    },
+    {
+      "epoch": 24.08,
+      "learning_rate": 9.872419697796512e-06,
+      "loss": 3.3793,
+      "step": 427000
+    },
+    {
+      "epoch": 24.11,
+      "learning_rate": 9.82537681350319e-06,
+      "loss": 3.38,
+      "step": 427500
+    },
+    {
+      "epoch": 24.14,
+      "learning_rate": 9.778333929209867e-06,
+      "loss": 3.3797,
+      "step": 428000
+    },
+    {
+      "epoch": 24.17,
+      "learning_rate": 9.731291044916546e-06,
+      "loss": 3.3844,
+      "step": 428500
+    },
+    {
+      "epoch": 24.19,
+      "learning_rate": 9.684248160623225e-06,
+      "loss": 3.3853,
+      "step": 429000
+    },
+    {
+      "epoch": 24.22,
+      "learning_rate": 9.637205276329903e-06,
+      "loss": 3.3808,
+      "step": 429500
+    },
+    {
+      "epoch": 24.25,
+      "learning_rate": 9.59016239203658e-06,
+      "loss": 3.3847,
+      "step": 430000
+    },
+    {
+      "epoch": 24.28,
+      "learning_rate": 9.543119507743259e-06,
+      "loss": 3.3879,
+      "step": 430500
+    },
+    {
+      "epoch": 24.31,
+      "learning_rate": 9.496076623449937e-06,
+      "loss": 3.3847,
+      "step": 431000
+    },
+    {
+      "epoch": 24.34,
+      "learning_rate": 9.449033739156616e-06,
+      "loss": 3.3837,
+      "step": 431500
+    },
+    {
+      "epoch": 24.36,
+      "learning_rate": 9.401990854863294e-06,
+      "loss": 3.3863,
+      "step": 432000
+    },
+    {
+      "epoch": 24.39,
+      "learning_rate": 9.354947970569971e-06,
+      "loss": 3.3862,
+      "step": 432500
+    },
+    {
+      "epoch": 24.42,
+      "learning_rate": 9.30790508627665e-06,
+      "loss": 3.3882,
+      "step": 433000
+    },
+    {
+      "epoch": 24.45,
+      "learning_rate": 9.260862201983328e-06,
+      "loss": 3.3904,
+      "step": 433500
+    },
+    {
+      "epoch": 24.48,
+      "learning_rate": 9.213819317690007e-06,
+      "loss": 3.3907,
+      "step": 434000
+    },
+    {
+      "epoch": 24.51,
+      "learning_rate": 9.166776433396684e-06,
+      "loss": 3.3888,
+      "step": 434500
+    },
+    {
+      "epoch": 24.53,
+      "learning_rate": 9.119733549103362e-06,
+      "loss": 3.3882,
+      "step": 435000
+    },
+    {
+      "epoch": 24.56,
+      "learning_rate": 9.072690664810041e-06,
+      "loss": 3.3944,
+      "step": 435500
+    },
+    {
+      "epoch": 24.59,
+      "learning_rate": 9.02564778051672e-06,
+      "loss": 3.4002,
+      "step": 436000
+    },
+    {
+      "epoch": 24.62,
+      "learning_rate": 8.978604896223396e-06,
+      "loss": 3.3936,
+      "step": 436500
+    },
+    {
+      "epoch": 24.65,
+      "learning_rate": 8.931562011930075e-06,
+      "loss": 3.3935,
+      "step": 437000
+    },
+    {
+      "epoch": 24.67,
+      "learning_rate": 8.884519127636753e-06,
+      "loss": 3.3914,
+      "step": 437500
+    },
+    {
+      "epoch": 24.7,
+      "learning_rate": 8.837476243343432e-06,
+      "loss": 3.3971,
+      "step": 438000
+    },
+    {
+      "epoch": 24.73,
+      "learning_rate": 8.79043335905011e-06,
+      "loss": 3.3956,
+      "step": 438500
+    },
+    {
+      "epoch": 24.76,
+      "learning_rate": 8.74339047475679e-06,
+      "loss": 3.3923,
+      "step": 439000
+    },
+    {
+      "epoch": 24.79,
+      "learning_rate": 8.696347590463468e-06,
+      "loss": 3.393,
+      "step": 439500
+    },
+    {
+      "epoch": 24.82,
+      "learning_rate": 8.649304706170146e-06,
+      "loss": 3.3975,
+      "step": 440000
+    },
+    {
+      "epoch": 24.84,
+      "learning_rate": 8.602261821876823e-06,
+      "loss": 3.3979,
+      "step": 440500
+    },
+    {
+      "epoch": 24.87,
+      "learning_rate": 8.555218937583502e-06,
+      "loss": 3.3958,
+      "step": 441000
+    },
+    {
+      "epoch": 24.9,
+      "learning_rate": 8.50817605329018e-06,
+      "loss": 3.4036,
+      "step": 441500
+    },
+    {
+      "epoch": 24.93,
+      "learning_rate": 8.461133168996859e-06,
+      "loss": 3.4009,
+      "step": 442000
+    },
+    {
+      "epoch": 24.96,
+      "learning_rate": 8.414090284703536e-06,
+      "loss": 3.3963,
+      "step": 442500
+    },
+    {
+      "epoch": 24.98,
+      "learning_rate": 8.367047400410214e-06,
+      "loss": 3.3982,
+      "step": 443000
+    },
+    {
+      "epoch": 25.0,
+      "eval_loss": 3.5556538105010986,
+      "eval_runtime": 860.2845,
+      "eval_samples_per_second": 264.828,
+      "eval_steps_per_second": 33.104,
+      "step": 443275
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 8.320004516116893e-06,
+      "loss": 3.3834,
+      "step": 443500
+    },
+    {
+      "epoch": 25.04,
+      "learning_rate": 8.272961631823571e-06,
+      "loss": 3.365,
+      "step": 444000
+    },
+    {
+      "epoch": 25.07,
+      "learning_rate": 8.22591874753025e-06,
+      "loss": 3.3641,
+      "step": 444500
+    },
+    {
+      "epoch": 25.1,
+      "learning_rate": 8.178875863236927e-06,
+      "loss": 3.3643,
+      "step": 445000
+    },
+    {
+      "epoch": 25.13,
+      "learning_rate": 8.131832978943605e-06,
+      "loss": 3.3723,
+      "step": 445500
+    },
+    {
+      "epoch": 25.15,
+      "learning_rate": 8.084790094650284e-06,
+      "loss": 3.3711,
+      "step": 446000
+    },
+    {
+      "epoch": 25.18,
+      "learning_rate": 8.037747210356963e-06,
+      "loss": 3.3663,
+      "step": 446500
+    },
+    {
+      "epoch": 25.21,
+      "learning_rate": 7.99070432606364e-06,
+      "loss": 3.368,
+      "step": 447000
+    },
+    {
+      "epoch": 25.24,
+      "learning_rate": 7.943661441770318e-06,
+      "loss": 3.3691,
+      "step": 447500
+    },
+    {
+      "epoch": 25.27,
+      "learning_rate": 7.896618557476997e-06,
+      "loss": 3.3733,
+      "step": 448000
+    },
+    {
+      "epoch": 25.29,
+      "learning_rate": 7.849575673183675e-06,
+      "loss": 3.3713,
+      "step": 448500
+    },
+    {
+      "epoch": 25.32,
+      "learning_rate": 7.802532788890352e-06,
+      "loss": 3.3695,
+      "step": 449000
+    },
+    {
+      "epoch": 25.35,
+      "learning_rate": 7.75548990459703e-06,
+      "loss": 3.3751,
+      "step": 449500
+    },
+    {
+      "epoch": 25.38,
+      "learning_rate": 7.70844702030371e-06,
+      "loss": 3.3769,
+      "step": 450000
+    },
+    {
+      "epoch": 25.41,
+      "learning_rate": 7.661404136010388e-06,
+      "loss": 3.3783,
+      "step": 450500
+    },
+    {
+      "epoch": 25.44,
+      "learning_rate": 7.6143612517170655e-06,
+      "loss": 3.3743,
+      "step": 451000
+    },
+    {
+      "epoch": 25.46,
+      "learning_rate": 7.567318367423744e-06,
+      "loss": 3.375,
+      "step": 451500
+    },
+    {
+      "epoch": 25.49,
+      "learning_rate": 7.520275483130422e-06,
+      "loss": 3.3749,
+      "step": 452000
+    },
+    {
+      "epoch": 25.52,
+      "learning_rate": 7.4732325988371e-06,
+      "loss": 3.377,
+      "step": 452500
+    },
+    {
+      "epoch": 25.55,
+      "learning_rate": 7.426189714543778e-06,
+      "loss": 3.3709,
+      "step": 453000
+    },
+    {
+      "epoch": 25.58,
+      "learning_rate": 7.379146830250457e-06,
+      "loss": 3.3807,
+      "step": 453500
+    },
+    {
+      "epoch": 25.6,
+      "learning_rate": 7.332103945957134e-06,
+      "loss": 3.3785,
+      "step": 454000
+    },
+    {
+      "epoch": 25.63,
+      "learning_rate": 7.285061061663813e-06,
+      "loss": 3.3809,
+      "step": 454500
+    },
+    {
+      "epoch": 25.66,
+      "learning_rate": 7.238018177370491e-06,
+      "loss": 3.379,
+      "step": 455000
+    },
+    {
+      "epoch": 25.69,
+      "learning_rate": 7.190975293077169e-06,
+      "loss": 3.3808,
+      "step": 455500
+    },
+    {
+      "epoch": 25.72,
+      "learning_rate": 7.143932408783848e-06,
+      "loss": 3.3849,
+      "step": 456000
+    },
+    {
+      "epoch": 25.75,
+      "learning_rate": 7.0968895244905256e-06,
+      "loss": 3.3845,
+      "step": 456500
+    },
+    {
+      "epoch": 25.77,
+      "learning_rate": 7.049846640197204e-06,
+      "loss": 3.3793,
+      "step": 457000
+    },
+    {
+      "epoch": 25.8,
+      "learning_rate": 7.002803755903882e-06,
+      "loss": 3.3838,
+      "step": 457500
+    },
+    {
+      "epoch": 25.83,
+      "learning_rate": 6.95576087161056e-06,
+      "loss": 3.3796,
+      "step": 458000
+    },
+    {
+      "epoch": 25.86,
+      "learning_rate": 6.908717987317238e-06,
+      "loss": 3.3828,
+      "step": 458500
+    },
+    {
+      "epoch": 25.89,
+      "learning_rate": 6.861675103023917e-06,
+      "loss": 3.3845,
+      "step": 459000
+    },
+    {
+      "epoch": 25.92,
+      "learning_rate": 6.8146322187305944e-06,
+      "loss": 3.382,
+      "step": 459500
+    },
+    {
+      "epoch": 25.94,
+      "learning_rate": 6.767589334437273e-06,
+      "loss": 3.3828,
+      "step": 460000
+    },
+    {
+      "epoch": 25.97,
+      "learning_rate": 6.720546450143951e-06,
+      "loss": 3.3774,
+      "step": 460500
+    },
+    {
+      "epoch": 26.0,
+      "learning_rate": 6.673503565850629e-06,
+      "loss": 3.3832,
+      "step": 461000
+    },
+    {
+      "epoch": 26.0,
+      "eval_loss": 3.550107717514038,
+      "eval_runtime": 859.3358,
+      "eval_samples_per_second": 265.12,
+      "eval_steps_per_second": 33.141,
+      "step": 461006
+    },
+    {
+      "epoch": 26.03,
+      "learning_rate": 6.626460681557307e-06,
+      "loss": 3.3516,
+      "step": 461500
+    },
+    {
+      "epoch": 26.06,
+      "learning_rate": 6.579417797263986e-06,
+      "loss": 3.3521,
+      "step": 462000
+    },
+    {
+      "epoch": 26.08,
+      "learning_rate": 6.532374912970664e-06,
+      "loss": 3.3533,
+      "step": 462500
+    },
+    {
+      "epoch": 26.11,
+      "learning_rate": 6.485332028677342e-06,
+      "loss": 3.3493,
+      "step": 463000
+    },
+    {
+      "epoch": 26.14,
+      "learning_rate": 6.4382891443840204e-06,
+      "loss": 3.3576,
+      "step": 463500
+    },
+    {
+      "epoch": 26.17,
+      "learning_rate": 6.391246260090698e-06,
+      "loss": 3.3533,
+      "step": 464000
+    },
+    {
+      "epoch": 26.2,
+      "learning_rate": 6.344203375797377e-06,
+      "loss": 3.3592,
+      "step": 464500
+    },
+    {
+      "epoch": 26.23,
+      "learning_rate": 6.297160491504056e-06,
+      "loss": 3.3543,
+      "step": 465000
+    },
+    {
+      "epoch": 26.25,
+      "learning_rate": 6.250117607210734e-06,
+      "loss": 3.3571,
+      "step": 465500
+    },
+    {
+      "epoch": 26.28,
+      "learning_rate": 6.203074722917412e-06,
+      "loss": 3.361,
+      "step": 466000
+    },
+    {
+      "epoch": 26.31,
+      "learning_rate": 6.15603183862409e-06,
+      "loss": 3.358,
+      "step": 466500
+    },
+    {
+      "epoch": 26.34,
+      "learning_rate": 6.108988954330768e-06,
+      "loss": 3.3632,
+      "step": 467000
+    },
+    {
+      "epoch": 26.37,
+      "learning_rate": 6.0619460700374465e-06,
+      "loss": 3.3561,
+      "step": 467500
+    },
+    {
+      "epoch": 26.39,
+      "learning_rate": 6.014903185744124e-06,
+      "loss": 3.3647,
+      "step": 468000
+    },
+    {
+      "epoch": 26.42,
+      "learning_rate": 5.967860301450803e-06,
+      "loss": 3.3581,
+      "step": 468500
+    },
+    {
+      "epoch": 26.45,
+      "learning_rate": 5.9208174171574805e-06,
+      "loss": 3.3651,
+      "step": 469000
+    },
+    {
+      "epoch": 26.48,
+      "learning_rate": 5.873774532864159e-06,
+      "loss": 3.364,
+      "step": 469500
+    },
+    {
+      "epoch": 26.51,
+      "learning_rate": 5.826731648570837e-06,
+      "loss": 3.3621,
+      "step": 470000
+    },
+    {
+      "epoch": 26.54,
+      "learning_rate": 5.779688764277515e-06,
+      "loss": 3.3669,
+      "step": 470500
+    },
+    {
+      "epoch": 26.56,
+      "learning_rate": 5.732645879984194e-06,
+      "loss": 3.3607,
+      "step": 471000
+    },
+    {
+      "epoch": 26.59,
+      "learning_rate": 5.6856029956908725e-06,
+      "loss": 3.3613,
+      "step": 471500
+    },
+    {
+      "epoch": 26.62,
+      "learning_rate": 5.63856011139755e-06,
+      "loss": 3.3672,
+      "step": 472000
+    },
+    {
+      "epoch": 26.65,
+      "learning_rate": 5.591517227104229e-06,
+      "loss": 3.3666,
+      "step": 472500
+    },
+    {
+      "epoch": 26.68,
+      "learning_rate": 5.544474342810907e-06,
+      "loss": 3.3653,
+      "step": 473000
+    },
+    {
+      "epoch": 26.7,
+      "learning_rate": 5.497431458517585e-06,
+      "loss": 3.3717,
+      "step": 473500
+    },
+    {
+      "epoch": 26.73,
+      "learning_rate": 5.450388574224264e-06,
+      "loss": 3.3701,
+      "step": 474000
+    },
+    {
+      "epoch": 26.76,
+      "learning_rate": 5.403345689930941e-06,
+      "loss": 3.364,
+      "step": 474500
+    },
+    {
+      "epoch": 26.79,
+      "learning_rate": 5.35630280563762e-06,
+      "loss": 3.3687,
+      "step": 475000
+    },
+    {
+      "epoch": 26.82,
+      "learning_rate": 5.309259921344298e-06,
+      "loss": 3.365,
+      "step": 475500
+    },
+    {
+      "epoch": 26.85,
+      "learning_rate": 5.262217037050976e-06,
+      "loss": 3.3683,
+      "step": 476000
+    },
+    {
+      "epoch": 26.87,
+      "learning_rate": 5.215174152757654e-06,
+      "loss": 3.3723,
+      "step": 476500
+    },
+    {
+      "epoch": 26.9,
+      "learning_rate": 5.1681312684643325e-06,
+      "loss": 3.3718,
+      "step": 477000
+    },
+    {
+      "epoch": 26.93,
+      "learning_rate": 5.12108838417101e-06,
+      "loss": 3.3684,
+      "step": 477500
+    },
+    {
+      "epoch": 26.96,
+      "learning_rate": 5.074045499877689e-06,
+      "loss": 3.3694,
+      "step": 478000
+    },
+    {
+      "epoch": 26.99,
+      "learning_rate": 5.0270026155843665e-06,
+      "loss": 3.3695,
+      "step": 478500
+    },
+    {
+      "epoch": 27.0,
+      "eval_loss": 3.5464906692504883,
+      "eval_runtime": 861.0727,
+      "eval_samples_per_second": 264.585,
+      "eval_steps_per_second": 33.074,
+      "step": 478737
+    },
+    {
+      "epoch": 27.01,
+      "learning_rate": 4.979959731291045e-06,
+      "loss": 3.357,
+      "step": 479000
+    },
+    {
+      "epoch": 27.04,
+      "learning_rate": 4.932916846997724e-06,
+      "loss": 3.3454,
+      "step": 479500
+    },
+    {
+      "epoch": 27.07,
+      "learning_rate": 4.885873962704401e-06,
+      "loss": 3.3438,
+      "step": 480000
+    },
+    {
+      "epoch": 27.1,
+      "learning_rate": 4.83883107841108e-06,
+      "loss": 3.3425,
+      "step": 480500
+    },
+    {
+      "epoch": 27.13,
+      "learning_rate": 4.791788194117758e-06,
+      "loss": 3.3451,
+      "step": 481000
+    },
+    {
+      "epoch": 27.16,
+      "learning_rate": 4.744745309824436e-06,
+      "loss": 3.3429,
+      "step": 481500
+    },
+    {
+      "epoch": 27.18,
+      "learning_rate": 4.697702425531114e-06,
+      "loss": 3.3475,
+      "step": 482000
+    },
+    {
+      "epoch": 27.21,
+      "learning_rate": 4.6506595412377925e-06,
+      "loss": 3.3515,
+      "step": 482500
+    },
+    {
+      "epoch": 27.24,
+      "learning_rate": 4.60361665694447e-06,
+      "loss": 3.3481,
+      "step": 483000
+    },
+    {
+      "epoch": 27.27,
+      "learning_rate": 4.556573772651149e-06,
+      "loss": 3.3485,
+      "step": 483500
+    },
+    {
+      "epoch": 27.3,
+      "learning_rate": 4.5095308883578265e-06,
+      "loss": 3.3448,
+      "step": 484000
+    },
+    {
+      "epoch": 27.33,
+      "learning_rate": 4.462488004064506e-06,
+      "loss": 3.347,
+      "step": 484500
+    },
+    {
+      "epoch": 27.35,
+      "learning_rate": 4.415445119771184e-06,
+      "loss": 3.3505,
+      "step": 485000
+    },
+    {
+      "epoch": 27.38,
+      "learning_rate": 4.368402235477862e-06,
+      "loss": 3.3507,
+      "step": 485500
+    },
+    {
+      "epoch": 27.41,
+      "learning_rate": 4.32135935118454e-06,
+      "loss": 3.3483,
+      "step": 486000
+    },
+    {
+      "epoch": 27.44,
+      "learning_rate": 4.2743164668912185e-06,
+      "loss": 3.3499,
+      "step": 486500
+    },
+    {
+      "epoch": 27.47,
+      "learning_rate": 4.227273582597896e-06,
+      "loss": 3.3531,
+      "step": 487000
+    },
+    {
+      "epoch": 27.49,
+      "learning_rate": 4.180230698304575e-06,
+      "loss": 3.3495,
+      "step": 487500
+    },
+    {
+      "epoch": 27.52,
+      "learning_rate": 4.133187814011253e-06,
+      "loss": 3.3527,
+      "step": 488000
+    },
+    {
+      "epoch": 27.55,
+      "learning_rate": 4.086144929717931e-06,
+      "loss": 3.3534,
+      "step": 488500
+    },
+    {
+      "epoch": 27.58,
+      "learning_rate": 4.03910204542461e-06,
+      "loss": 3.3518,
+      "step": 489000
+    },
+    {
+      "epoch": 27.61,
+      "learning_rate": 3.992059161131287e-06,
+      "loss": 3.352,
+      "step": 489500
+    },
+    {
+      "epoch": 27.64,
+      "learning_rate": 3.945016276837966e-06,
+      "loss": 3.3569,
+      "step": 490000
+    },
+    {
+      "epoch": 27.66,
+      "learning_rate": 3.897973392544644e-06,
+      "loss": 3.3542,
+      "step": 490500
+    },
+    {
+      "epoch": 27.69,
+      "learning_rate": 3.850930508251322e-06,
+      "loss": 3.352,
+      "step": 491000
+    },
+    {
+      "epoch": 27.72,
+      "learning_rate": 3.8038876239580004e-06,
+      "loss": 3.3476,
+      "step": 491500
+    },
+    {
+      "epoch": 27.75,
+      "learning_rate": 3.7568447396646786e-06,
+      "loss": 3.3581,
+      "step": 492000
+    },
+    {
+      "epoch": 27.78,
+      "learning_rate": 3.7098018553713567e-06,
+      "loss": 3.3593,
+      "step": 492500
+    },
+    {
+      "epoch": 27.8,
+      "learning_rate": 3.662758971078035e-06,
+      "loss": 3.3519,
+      "step": 493000
+    },
+    {
+      "epoch": 27.83,
+      "learning_rate": 3.615716086784713e-06,
+      "loss": 3.3515,
+      "step": 493500
+    },
+    {
+      "epoch": 27.86,
+      "learning_rate": 3.568673202491391e-06,
+      "loss": 3.3579,
+      "step": 494000
+    },
+    {
+      "epoch": 27.89,
+      "learning_rate": 3.5216303181980693e-06,
+      "loss": 3.3567,
+      "step": 494500
+    },
+    {
+      "epoch": 27.92,
+      "learning_rate": 3.4745874339047474e-06,
+      "loss": 3.3516,
+      "step": 495000
+    },
+    {
+      "epoch": 27.95,
+      "learning_rate": 3.4275445496114256e-06,
+      "loss": 3.3578,
+      "step": 495500
+    },
+    {
+      "epoch": 27.97,
+      "learning_rate": 3.3805016653181037e-06,
+      "loss": 3.3517,
+      "step": 496000
+    },
+    {
+      "epoch": 28.0,
+      "eval_loss": 3.542843818664551,
+      "eval_runtime": 859.162,
+      "eval_samples_per_second": 265.174,
+      "eval_steps_per_second": 33.147,
+      "step": 496468
+    },
+    {
+      "epoch": 28.0,
+      "learning_rate": 3.333458781024782e-06,
+      "loss": 3.3555,
+      "step": 496500
+    },
+    {
+      "epoch": 28.03,
+      "learning_rate": 3.2864158967314604e-06,
+      "loss": 3.3329,
+      "step": 497000
+    },
+    {
+      "epoch": 28.06,
+      "learning_rate": 3.2393730124381386e-06,
+      "loss": 3.3364,
+      "step": 497500
+    },
+    {
+      "epoch": 28.09,
+      "learning_rate": 3.192330128144817e-06,
+      "loss": 3.3337,
+      "step": 498000
+    },
+    {
+      "epoch": 28.11,
+      "learning_rate": 3.1452872438514953e-06,
+      "loss": 3.34,
+      "step": 498500
+    },
+    {
+      "epoch": 28.14,
+      "learning_rate": 3.0982443595581734e-06,
+      "loss": 3.3379,
+      "step": 499000
+    },
+    {
+      "epoch": 28.17,
+      "learning_rate": 3.0512014752648516e-06,
+      "loss": 3.3386,
+      "step": 499500
+    },
+    {
+      "epoch": 28.2,
+      "learning_rate": 3.0041585909715297e-06,
+      "loss": 3.335,
+      "step": 500000
+    },
+    {
+      "epoch": 28.23,
+      "learning_rate": 2.957115706678208e-06,
+      "loss": 3.3351,
+      "step": 500500
+    },
+    {
+      "epoch": 28.26,
+      "learning_rate": 2.910072822384886e-06,
+      "loss": 3.3398,
+      "step": 501000
+    },
+    {
+      "epoch": 28.28,
+      "learning_rate": 2.8630299380915646e-06,
+      "loss": 3.3323,
+      "step": 501500
+    },
+    {
+      "epoch": 28.31,
+      "learning_rate": 2.8159870537982427e-06,
+      "loss": 3.339,
+      "step": 502000
+    },
+    {
+      "epoch": 28.34,
+      "learning_rate": 2.768944169504921e-06,
+      "loss": 3.3374,
+      "step": 502500
+    },
+    {
+      "epoch": 28.37,
+      "learning_rate": 2.721901285211599e-06,
+      "loss": 3.3367,
+      "step": 503000
+    },
+    {
+      "epoch": 28.4,
+      "learning_rate": 2.674858400918277e-06,
+      "loss": 3.3337,
+      "step": 503500
+    },
+    {
+      "epoch": 28.42,
+      "learning_rate": 2.6278155166249553e-06,
+      "loss": 3.3371,
+      "step": 504000
+    },
+    {
+      "epoch": 28.45,
+      "learning_rate": 2.5807726323316335e-06,
+      "loss": 3.3464,
+      "step": 504500
+    },
+    {
+      "epoch": 28.48,
+      "learning_rate": 2.5337297480383116e-06,
+      "loss": 3.3459,
+      "step": 505000
+    },
+    {
+      "epoch": 28.51,
+      "learning_rate": 2.48668686374499e-06,
+      "loss": 3.344,
+      "step": 505500
+    },
+    {
+      "epoch": 28.54,
+      "learning_rate": 2.4396439794516683e-06,
+      "loss": 3.3355,
+      "step": 506000
+    },
+    {
+      "epoch": 28.57,
+      "learning_rate": 2.3926010951583465e-06,
+      "loss": 3.3448,
+      "step": 506500
+    },
+    {
+      "epoch": 28.59,
+      "learning_rate": 2.3455582108650246e-06,
+      "loss": 3.343,
+      "step": 507000
+    },
+    {
+      "epoch": 28.62,
+      "learning_rate": 2.2985153265717028e-06,
+      "loss": 3.3461,
+      "step": 507500
+    },
+    {
+      "epoch": 28.65,
+      "learning_rate": 2.2514724422783813e-06,
+      "loss": 3.3407,
+      "step": 508000
+    },
+    {
+      "epoch": 28.68,
+      "learning_rate": 2.2044295579850595e-06,
+      "loss": 3.3378,
+      "step": 508500
+    },
+    {
+      "epoch": 28.71,
+      "learning_rate": 2.1573866736917376e-06,
+      "loss": 3.3437,
+      "step": 509000
+    },
+    {
+      "epoch": 28.73,
+      "learning_rate": 2.1103437893984158e-06,
+      "loss": 3.3464,
+      "step": 509500
+    },
+    {
+      "epoch": 28.76,
+      "learning_rate": 2.063300905105094e-06,
+      "loss": 3.3434,
+      "step": 510000
+    },
+    {
+      "epoch": 28.79,
+      "learning_rate": 2.016258020811772e-06,
+      "loss": 3.3374,
+      "step": 510500
+    },
+    {
+      "epoch": 28.82,
+      "learning_rate": 1.9692151365184502e-06,
+      "loss": 3.3488,
+      "step": 511000
+    },
+    {
+      "epoch": 28.85,
+      "learning_rate": 1.9221722522251284e-06,
+      "loss": 3.3462,
+      "step": 511500
+    },
+    {
+      "epoch": 28.88,
+      "learning_rate": 1.8751293679318067e-06,
+      "loss": 3.3398,
+      "step": 512000
+    },
+    {
+      "epoch": 28.9,
+      "learning_rate": 1.8280864836384849e-06,
+      "loss": 3.3429,
+      "step": 512500
+    },
+    {
+      "epoch": 28.93,
+      "learning_rate": 1.781043599345163e-06,
+      "loss": 3.3382,
+      "step": 513000
+    },
+    {
+      "epoch": 28.96,
+      "learning_rate": 1.7340007150518412e-06,
+      "loss": 3.3414,
+      "step": 513500
+    },
+    {
+      "epoch": 28.99,
+      "learning_rate": 1.6869578307585193e-06,
+      "loss": 3.3375,
+      "step": 514000
+    },
+    {
+      "epoch": 29.0,
+      "eval_loss": 3.540179967880249,
+      "eval_runtime": 860.1608,
+      "eval_samples_per_second": 264.866,
+      "eval_steps_per_second": 33.109,
+      "step": 514199
+    },
+    {
+      "epoch": 29.02,
+      "learning_rate": 1.6399149464651979e-06,
+      "loss": 3.336,
+      "step": 514500
+    },
+    {
+      "epoch": 29.05,
+      "learning_rate": 1.592872062171876e-06,
+      "loss": 3.3316,
+      "step": 515000
+    },
+    {
+      "epoch": 29.07,
+      "learning_rate": 1.5458291778785542e-06,
+      "loss": 3.3263,
+      "step": 515500
+    },
+    {
+      "epoch": 29.1,
+      "learning_rate": 1.4987862935852325e-06,
+      "loss": 3.3269,
+      "step": 516000
+    },
+    {
+      "epoch": 29.13,
+      "learning_rate": 1.4517434092919107e-06,
+      "loss": 3.3329,
+      "step": 516500
+    },
+    {
+      "epoch": 29.16,
+      "learning_rate": 1.4047005249985888e-06,
+      "loss": 3.3321,
+      "step": 517000
+    },
+    {
+      "epoch": 29.19,
+      "learning_rate": 1.357657640705267e-06,
+      "loss": 3.3296,
+      "step": 517500
+    },
+    {
+      "epoch": 29.21,
+      "learning_rate": 1.310614756411945e-06,
+      "loss": 3.3256,
+      "step": 518000
+    },
+    {
+      "epoch": 29.24,
+      "learning_rate": 1.2635718721186233e-06,
+      "loss": 3.3269,
+      "step": 518500
+    },
+    {
+      "epoch": 29.27,
+      "learning_rate": 1.2165289878253016e-06,
+      "loss": 3.33,
+      "step": 519000
+    },
+    {
+      "epoch": 29.3,
+      "learning_rate": 1.16948610353198e-06,
+      "loss": 3.3332,
+      "step": 519500
+    },
+    {
+      "epoch": 29.33,
+      "learning_rate": 1.1224432192386581e-06,
+      "loss": 3.3308,
+      "step": 520000
+    },
+    {
+      "epoch": 29.36,
+      "learning_rate": 1.0754003349453363e-06,
+      "loss": 3.3305,
+      "step": 520500
+    },
+    {
+      "epoch": 29.38,
+      "learning_rate": 1.0283574506520144e-06,
+      "loss": 3.3304,
+      "step": 521000
+    },
+    {
+      "epoch": 29.41,
+      "learning_rate": 9.813145663586925e-07,
+      "loss": 3.3361,
+      "step": 521500
+    },
+    {
+      "epoch": 29.44,
+      "learning_rate": 9.342716820653708e-07,
+      "loss": 3.3296,
+      "step": 522000
+    },
+    {
+      "epoch": 29.47,
+      "learning_rate": 8.872287977720489e-07,
+      "loss": 3.3292,
+      "step": 522500
+    },
+    {
+      "epoch": 29.5,
+      "learning_rate": 8.401859134787273e-07,
+      "loss": 3.329,
+      "step": 523000
+    },
+    {
+      "epoch": 29.52,
+      "learning_rate": 7.931430291854055e-07,
+      "loss": 3.3334,
+      "step": 523500
+    },
+    {
+      "epoch": 29.55,
+      "learning_rate": 7.461001448920837e-07,
+      "loss": 3.3317,
+      "step": 524000
+    },
+    {
+      "epoch": 29.58,
+      "learning_rate": 6.990572605987618e-07,
+      "loss": 3.3288,
+      "step": 524500
+    },
+    {
+      "epoch": 29.61,
+      "learning_rate": 6.5201437630544e-07,
+      "loss": 3.3311,
+      "step": 525000
+    },
+    {
+      "epoch": 29.64,
+      "learning_rate": 6.049714920121182e-07,
+      "loss": 3.3321,
+      "step": 525500
+    },
+    {
+      "epoch": 29.67,
+      "learning_rate": 5.579286077187965e-07,
+      "loss": 3.3318,
+      "step": 526000
+    },
+    {
+      "epoch": 29.69,
+      "learning_rate": 5.108857234254746e-07,
+      "loss": 3.3331,
+      "step": 526500
+    },
+    {
+      "epoch": 29.72,
+      "learning_rate": 4.6384283913215295e-07,
+      "loss": 3.3286,
+      "step": 527000
+    },
+    {
+      "epoch": 29.75,
+      "learning_rate": 4.167999548388311e-07,
+      "loss": 3.3293,
+      "step": 527500
+    },
+    {
+      "epoch": 29.78,
+      "learning_rate": 3.6975707054550934e-07,
+      "loss": 3.3342,
+      "step": 528000
+    },
+    {
+      "epoch": 29.81,
+      "learning_rate": 3.227141862521875e-07,
+      "loss": 3.3296,
+      "step": 528500
+    },
+    {
+      "epoch": 29.83,
+      "learning_rate": 2.7567130195886574e-07,
+      "loss": 3.3293,
+      "step": 529000
+    },
+    {
+      "epoch": 29.86,
+      "learning_rate": 2.2862841766554391e-07,
+      "loss": 3.3345,
+      "step": 529500
+    },
+    {
+      "epoch": 29.89,
+      "learning_rate": 1.8158553337222214e-07,
+      "loss": 3.329,
+      "step": 530000
+    }
+  ],
+  "max_steps": 531930,
+  "num_train_epochs": 30,
+  "total_flos": 2.215731804585984e+18,
+  "trial_name": null,
+  "trial_params": null
+}