diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6216 @@
+{
+  "best_metric": 2.1569273471832275,
+  "best_model_checkpoint": "/netscratch/gutsche/data/LOLTwitchBertTraining_da66d5ea-8560-4a50-b7a3-fb9e339ee1af/TwitchLeagueBert/checkpoint-500000",
+  "epoch": 7.430303750817333,
+  "global_step": 500000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.998142424062296e-05,
+      "loss": 8.231,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9962848481245916e-05,
+      "loss": 7.4239,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.994427272186887e-05,
+      "loss": 7.1438,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.992569696249183e-05,
+      "loss": 6.9853,
+      "step": 2000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.990712120311479e-05,
+      "loss": 6.8714,
+      "step": 2500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9888545443737744e-05,
+      "loss": 6.7786,
+      "step": 3000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.98699696843607e-05,
+      "loss": 6.7338,
+      "step": 3500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.985139392498366e-05,
+      "loss": 6.6649,
+      "step": 4000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9832818165606614e-05,
+      "loss": 6.6232,
+      "step": 4500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9814242406229564e-05,
+      "loss": 6.5875,
+      "step": 5000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.979566664685252e-05,
+      "loss": 6.5195,
+      "step": 5500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.977709088747548e-05,
+      "loss": 6.4647,
+      "step": 6000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9758515128098435e-05,
+      "loss": 6.2629,
+      "step": 6500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.973993936872139e-05,
+      "loss": 6.0753,
+      "step": 7000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.972136360934435e-05,
+      "loss": 5.9132,
+      "step": 7500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9702787849967306e-05,
+      "loss": 5.7549,
+      "step": 8000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.968421209059026e-05,
+      "loss": 5.5722,
+      "step": 8500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.966563633121322e-05,
+      "loss": 5.3825,
+      "step": 9000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9647060571836176e-05,
+      "loss": 5.2093,
+      "step": 9500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.962848481245913e-05,
+      "loss": 5.065,
+      "step": 10000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.960990905308209e-05,
+      "loss": 4.9503,
+      "step": 10500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.959133329370505e-05,
+      "loss": 4.8124,
+      "step": 11000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.957275753432801e-05,
+      "loss": 4.6793,
+      "step": 11500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.955418177495097e-05,
+      "loss": 4.5688,
+      "step": 12000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9535606015573924e-05,
+      "loss": 4.441,
+      "step": 12500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.951703025619688e-05,
+      "loss": 4.2956,
+      "step": 13000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.949845449681984e-05,
+      "loss": 4.221,
+      "step": 13500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.947987873744279e-05,
+      "loss": 4.129,
+      "step": 14000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9461302978065745e-05,
+      "loss": 4.0381,
+      "step": 14500
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.94427272186887e-05,
+      "loss": 3.9632,
+      "step": 15000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.942415145931166e-05,
+      "loss": 3.9009,
+      "step": 15500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9405575699934615e-05,
+      "loss": 3.8547,
+      "step": 16000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.938699994055757e-05,
+      "loss": 3.787,
+      "step": 16500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.936842418118053e-05,
+      "loss": 3.7538,
+      "step": 17000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.9349848421803486e-05,
+      "loss": 3.71,
+      "step": 17500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.933127266242644e-05,
+      "loss": 3.6738,
+      "step": 18000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.93126969030494e-05,
+      "loss": 3.6138,
+      "step": 18500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.929412114367236e-05,
+      "loss": 3.5978,
+      "step": 19000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.9275545384295314e-05,
+      "loss": 3.5557,
+      "step": 19500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.925696962491827e-05,
+      "loss": 3.5383,
+      "step": 20000
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 3.430026054382324,
+      "eval_runtime": 1172.1185,
+      "eval_samples_per_second": 408.251,
+      "eval_steps_per_second": 51.032,
+      "step": 20000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.923839386554123e-05,
+      "loss": 3.5163,
+      "step": 20500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.9219818106164184e-05,
+      "loss": 3.4927,
+      "step": 21000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.920124234678714e-05,
+      "loss": 3.4945,
+      "step": 21500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.91826665874101e-05,
+      "loss": 3.4371,
+      "step": 22000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.9164090828033055e-05,
+      "loss": 3.4193,
+      "step": 22500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.914551506865601e-05,
+      "loss": 3.4072,
+      "step": 23000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.912693930927896e-05,
+      "loss": 3.3675,
+      "step": 23500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.910836354990192e-05,
+      "loss": 3.3547,
+      "step": 24000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.9089787790524875e-05,
+      "loss": 3.3442,
+      "step": 24500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.907121203114783e-05,
+      "loss": 3.3343,
+      "step": 25000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.905263627177079e-05,
+      "loss": 3.3113,
+      "step": 25500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.9034060512393746e-05,
+      "loss": 3.2771,
+      "step": 26000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.90154847530167e-05,
+      "loss": 3.273,
+      "step": 26500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.899690899363966e-05,
+      "loss": 3.2727,
+      "step": 27000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.897833323426262e-05,
+      "loss": 3.2305,
+      "step": 27500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.8959757474885574e-05,
+      "loss": 3.2263,
+      "step": 28000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.894118171550853e-05,
+      "loss": 3.2133,
+      "step": 28500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.892260595613149e-05,
+      "loss": 3.2353,
+      "step": 29000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.8904030196754444e-05,
+      "loss": 3.1918,
+      "step": 29500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.88854544373774e-05,
+      "loss": 3.1688,
+      "step": 30000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.886687867800036e-05,
+      "loss": 3.1584,
+      "step": 30500
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.8848302918623315e-05,
+      "loss": 3.1597,
+      "step": 31000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.882972715924627e-05,
+      "loss": 3.1416,
+      "step": 31500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.881115139986923e-05,
+      "loss": 3.1551,
+      "step": 32000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.8792575640492185e-05,
+      "loss": 3.1232,
+      "step": 32500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.877399988111514e-05,
+      "loss": 3.1159,
+      "step": 33000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.87554241217381e-05,
+      "loss": 3.0914,
+      "step": 33500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.8736848362361056e-05,
+      "loss": 3.0866,
+      "step": 34000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.871827260298401e-05,
+      "loss": 3.0784,
+      "step": 34500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.869969684360697e-05,
+      "loss": 3.0815,
+      "step": 35000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.8681121084229927e-05,
+      "loss": 3.0667,
+      "step": 35500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.8662545324852883e-05,
+      "loss": 3.0564,
+      "step": 36000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.864396956547584e-05,
+      "loss": 3.0549,
+      "step": 36500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.86253938060988e-05,
+      "loss": 3.0333,
+      "step": 37000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.8606818046721754e-05,
+      "loss": 3.0262,
+      "step": 37500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.858824228734471e-05,
+      "loss": 3.037,
+      "step": 38000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.856966652796767e-05,
+      "loss": 3.0241,
+      "step": 38500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.8551090768590625e-05,
+      "loss": 3.0134,
+      "step": 39000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.853251500921358e-05,
+      "loss": 3.0102,
+      "step": 39500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.851393924983654e-05,
+      "loss": 2.9842,
+      "step": 40000
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 2.925679922103882,
+      "eval_runtime": 1173.0278,
+      "eval_samples_per_second": 407.935,
+      "eval_steps_per_second": 50.992,
+      "step": 40000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.8495363490459495e-05,
+      "loss": 2.9908,
+      "step": 40500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.847678773108245e-05,
+      "loss": 2.9797,
+      "step": 41000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.845821197170541e-05,
+      "loss": 2.981,
+      "step": 41500
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.843963621232836e-05,
+      "loss": 2.9777,
+      "step": 42000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.8421060452951316e-05,
+      "loss": 2.9611,
+      "step": 42500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.840248469357427e-05,
+      "loss": 2.9599,
+      "step": 43000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.838390893419723e-05,
+      "loss": 2.9528,
+      "step": 43500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.8365333174820187e-05,
+      "loss": 2.948,
+      "step": 44000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.8346757415443143e-05,
+      "loss": 2.946,
+      "step": 44500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.83281816560661e-05,
+      "loss": 2.949,
+      "step": 45000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.830960589668906e-05,
+      "loss": 2.9323,
+      "step": 45500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.8291030137312014e-05,
+      "loss": 2.9262,
+      "step": 46000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.827245437793497e-05,
+      "loss": 2.9044,
+      "step": 46500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.825387861855793e-05,
+      "loss": 2.9031,
+      "step": 47000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.8235302859180885e-05,
+      "loss": 2.9183,
+      "step": 47500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.821672709980384e-05,
+      "loss": 2.8872,
+      "step": 48000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.81981513404268e-05,
+      "loss": 2.8891,
+      "step": 48500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.8179575581049755e-05,
+      "loss": 2.8871,
+      "step": 49000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.816099982167271e-05,
+      "loss": 2.8634,
+      "step": 49500
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.814242406229567e-05,
+      "loss": 2.9043,
+      "step": 50000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.8123848302918626e-05,
+      "loss": 2.8807,
+      "step": 50500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.810527254354158e-05,
+      "loss": 2.8617,
+      "step": 51000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.808669678416454e-05,
+      "loss": 2.8533,
+      "step": 51500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.8068121024787496e-05,
+      "loss": 2.8709,
+      "step": 52000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.804954526541045e-05,
+      "loss": 2.8524,
+      "step": 52500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.803096950603341e-05,
+      "loss": 2.8494,
+      "step": 53000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.801239374665637e-05,
+      "loss": 2.847,
+      "step": 53500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.7993817987279324e-05,
+      "loss": 2.8241,
+      "step": 54000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.797524222790228e-05,
+      "loss": 2.8502,
+      "step": 54500
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.795666646852524e-05,
+      "loss": 2.8225,
+      "step": 55000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.7938090709148195e-05,
+      "loss": 2.8416,
+      "step": 55500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.791951494977115e-05,
+      "loss": 2.7997,
+      "step": 56000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.790093919039411e-05,
+      "loss": 2.8175,
+      "step": 56500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.7882363431017065e-05,
+      "loss": 2.8017,
+      "step": 57000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.786378767164002e-05,
+      "loss": 2.8122,
+      "step": 57500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.784521191226298e-05,
+      "loss": 2.7994,
+      "step": 58000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.7826636152885936e-05,
+      "loss": 2.7929,
+      "step": 58500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.780806039350889e-05,
+      "loss": 2.7802,
+      "step": 59000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.778948463413185e-05,
+      "loss": 2.7921,
+      "step": 59500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.7770908874754806e-05,
+      "loss": 2.7801,
+      "step": 60000
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 2.7286553382873535,
+      "eval_runtime": 1174.8336,
+      "eval_samples_per_second": 407.308,
+      "eval_steps_per_second": 50.914,
+      "step": 60000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.7752333115377756e-05,
+      "loss": 2.7903,
+      "step": 60500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.773375735600071e-05,
+      "loss": 2.783,
+      "step": 61000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.771518159662367e-05,
+      "loss": 2.7792,
+      "step": 61500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 4.769660583724663e-05,
+      "loss": 2.7564,
+      "step": 62000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.7678030077869584e-05,
+      "loss": 2.7652,
+      "step": 62500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.765945431849254e-05,
+      "loss": 2.777,
+      "step": 63000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.76408785591155e-05,
+      "loss": 2.7674,
+      "step": 63500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 4.7622302799738455e-05,
+      "loss": 2.7754,
+      "step": 64000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 4.760372704036141e-05,
+      "loss": 2.7536,
+      "step": 64500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.758515128098437e-05,
+      "loss": 2.7451,
+      "step": 65000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.7566575521607325e-05,
+      "loss": 2.7382,
+      "step": 65500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.754799976223028e-05,
+      "loss": 2.7419,
+      "step": 66000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.752942400285324e-05,
+      "loss": 2.7564,
+      "step": 66500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7510848243476196e-05,
+      "loss": 2.7373,
+      "step": 67000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.749227248409915e-05,
+      "loss": 2.7353,
+      "step": 67500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.747369672472211e-05,
+      "loss": 2.7355,
+      "step": 68000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7455120965345066e-05,
+      "loss": 2.725,
+      "step": 68500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.743654520596802e-05,
+      "loss": 2.7095,
+      "step": 69000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 4.741796944659097e-05,
+      "loss": 2.7259,
+      "step": 69500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 4.739939368721393e-05,
+      "loss": 2.7142,
+      "step": 70000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 4.738081792783689e-05,
+      "loss": 2.7066,
+      "step": 70500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.7362242168459844e-05,
+      "loss": 2.7199,
+      "step": 71000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 4.73436664090828e-05,
+      "loss": 2.7113,
+      "step": 71500
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 4.7325090649705764e-05,
+      "loss": 2.7058,
+      "step": 72000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.730651489032872e-05,
+      "loss": 2.709,
+      "step": 72500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 4.728793913095168e-05,
+      "loss": 2.6836,
+      "step": 73000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.7269363371574635e-05,
+      "loss": 2.6948,
+      "step": 73500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 4.725078761219759e-05,
+      "loss": 2.694,
+      "step": 74000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.723221185282055e-05,
+      "loss": 2.6896,
+      "step": 74500
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 4.7213636093443506e-05,
+      "loss": 2.7087,
+      "step": 75000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 4.719506033406646e-05,
+      "loss": 2.6669,
+      "step": 75500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 4.717648457468942e-05,
+      "loss": 2.6623,
+      "step": 76000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.7157908815312376e-05,
+      "loss": 2.6854,
+      "step": 76500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 4.713933305593533e-05,
+      "loss": 2.6837,
+      "step": 77000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 4.712075729655829e-05,
+      "loss": 2.6628,
+      "step": 77500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 4.710218153718125e-05,
+      "loss": 2.6616,
+      "step": 78000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.70836057778042e-05,
+      "loss": 2.6765,
+      "step": 78500
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 4.7065030018427154e-05,
+      "loss": 2.6744,
+      "step": 79000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 4.704645425905011e-05,
+      "loss": 2.6777,
+      "step": 79500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.702787849967307e-05,
+      "loss": 2.6607,
+      "step": 80000
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 2.6154634952545166,
+      "eval_runtime": 1176.1132,
+      "eval_samples_per_second": 406.865,
+      "eval_steps_per_second": 50.858,
+      "step": 80000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.7009302740296025e-05,
+      "loss": 2.6604,
+      "step": 80500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 4.699072698091898e-05,
+      "loss": 2.6638,
+      "step": 81000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 4.697215122154194e-05,
+      "loss": 2.6532,
+      "step": 81500
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 4.6953575462164895e-05,
+      "loss": 2.6609,
+      "step": 82000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.693499970278785e-05,
+      "loss": 2.6634,
+      "step": 82500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 4.691642394341081e-05,
+      "loss": 2.6548,
+      "step": 83000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 4.6897848184033766e-05,
+      "loss": 2.6581,
+      "step": 83500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 4.687927242465672e-05,
+      "loss": 2.6587,
+      "step": 84000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.686069666527968e-05,
+      "loss": 2.6481,
+      "step": 84500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 4.6842120905902636e-05,
+      "loss": 2.6438,
+      "step": 85000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 4.682354514652559e-05,
+      "loss": 2.6372,
+      "step": 85500
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 4.680496938714855e-05,
+      "loss": 2.6396,
+      "step": 86000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.678639362777151e-05,
+      "loss": 2.6389,
+      "step": 86500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.6767817868394464e-05,
+      "loss": 2.6416,
+      "step": 87000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 4.674924210901742e-05,
+      "loss": 2.6479,
+      "step": 87500
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 4.673066634964037e-05,
+      "loss": 2.6172,
+      "step": 88000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.671209059026333e-05,
+      "loss": 2.612,
+      "step": 88500
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 4.6693514830886285e-05,
+      "loss": 2.6318,
+      "step": 89000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 4.667493907150924e-05,
+      "loss": 2.6373,
+      "step": 89500
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.66563633121322e-05,
+      "loss": 2.6444,
+      "step": 90000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 4.6637787552755155e-05,
+      "loss": 2.6158,
+      "step": 90500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 4.661921179337811e-05,
+      "loss": 2.6245,
+      "step": 91000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 4.660063603400107e-05,
+      "loss": 2.6326,
+      "step": 91500
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.6582060274624026e-05,
+      "loss": 2.6318,
+      "step": 92000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 4.656348451524698e-05,
+      "loss": 2.6283,
+      "step": 92500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 4.654490875586994e-05,
+      "loss": 2.6029,
+      "step": 93000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.6526332996492896e-05,
+      "loss": 2.6183,
+      "step": 93500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.650775723711585e-05,
+      "loss": 2.6165,
+      "step": 94000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 4.648918147773882e-05,
+      "loss": 2.6081,
+      "step": 94500
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 4.6470605718361774e-05,
+      "loss": 2.6337,
+      "step": 95000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 4.645202995898473e-05,
+      "loss": 2.5937,
+      "step": 95500
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.643345419960769e-05,
+      "loss": 2.6143,
+      "step": 96000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 4.6414878440230644e-05,
+      "loss": 2.6044,
+      "step": 96500
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 4.6396302680853594e-05,
+      "loss": 2.608,
+      "step": 97000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 4.637772692147655e-05,
+      "loss": 2.6047,
+      "step": 97500
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.635915116209951e-05,
+      "loss": 2.5922,
+      "step": 98000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 4.6340575402722465e-05,
+      "loss": 2.6051,
+      "step": 98500
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 4.632199964334542e-05,
+      "loss": 2.5986,
+      "step": 99000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.630342388396838e-05,
+      "loss": 2.5852,
+      "step": 99500
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.6284848124591336e-05,
+      "loss": 2.5837,
+      "step": 100000
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 2.5408413410186768,
+      "eval_runtime": 1176.2652,
+      "eval_samples_per_second": 406.812,
+      "eval_steps_per_second": 50.852,
+      "step": 100000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 4.626627236521429e-05,
+      "loss": 2.583,
+      "step": 100500
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 4.624769660583725e-05,
+      "loss": 2.586,
+      "step": 101000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 4.6229120846460206e-05,
+      "loss": 2.5929,
+      "step": 101500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.621054508708316e-05,
+      "loss": 2.5938,
+      "step": 102000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 4.619196932770612e-05,
+      "loss": 2.5759,
+      "step": 102500
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 4.617339356832908e-05,
+      "loss": 2.5853,
+      "step": 103000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 4.6154817808952034e-05,
+      "loss": 2.5807,
+      "step": 103500
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.613624204957499e-05,
+      "loss": 2.5606,
+      "step": 104000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 4.611766629019795e-05,
+      "loss": 2.5749,
+      "step": 104500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 4.6099090530820904e-05,
+      "loss": 2.5578,
+      "step": 105000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 4.608051477144386e-05,
+      "loss": 2.5818,
+      "step": 105500
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.606193901206682e-05,
+      "loss": 2.5786,
+      "step": 106000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.604336325268977e-05,
+      "loss": 2.5885,
+      "step": 106500
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.6024787493312725e-05,
+      "loss": 2.567,
+      "step": 107000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.600621173393568e-05,
+      "loss": 2.5774,
+      "step": 107500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 4.598763597455864e-05,
+      "loss": 2.5623,
+      "step": 108000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 4.5969060215181596e-05,
+      "loss": 2.5703,
+      "step": 108500
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 4.595048445580455e-05,
+      "loss": 2.5592,
+      "step": 109000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.593190869642751e-05,
+      "loss": 2.5846,
+      "step": 109500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 4.5913332937050466e-05,
+      "loss": 2.5889,
+      "step": 110000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 4.589475717767342e-05,
+      "loss": 2.5598,
+      "step": 110500
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 4.587618141829638e-05,
+      "loss": 2.5651,
+      "step": 111000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.585760565891934e-05,
+      "loss": 2.5668,
+      "step": 111500
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 4.5839029899542294e-05,
+      "loss": 2.5509,
+      "step": 112000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 4.582045414016525e-05,
+      "loss": 2.5583,
+      "step": 112500
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.580187838078821e-05,
+      "loss": 2.5532,
+      "step": 113000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.5783302621411164e-05,
+      "loss": 2.5517,
+      "step": 113500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 4.576472686203412e-05,
+      "loss": 2.5641,
+      "step": 114000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 4.574615110265708e-05,
+      "loss": 2.5515,
+      "step": 114500
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 4.5727575343280035e-05,
+      "loss": 2.5724,
+      "step": 115000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.570899958390299e-05,
+      "loss": 2.5452,
+      "step": 115500
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 4.569042382452595e-05,
+      "loss": 2.5574,
+      "step": 116000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 4.5671848065148906e-05,
+      "loss": 2.5637,
+      "step": 116500
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 4.565327230577186e-05,
+      "loss": 2.5539,
+      "step": 117000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.563469654639482e-05,
+      "loss": 2.5437,
+      "step": 117500
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 4.5616120787017776e-05,
+      "loss": 2.5361,
+      "step": 118000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 4.559754502764073e-05,
+      "loss": 2.5522,
+      "step": 118500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 4.557896926826369e-05,
+      "loss": 2.5208,
+      "step": 119000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.556039350888665e-05,
+      "loss": 2.5325,
+      "step": 119500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.5541817749509604e-05,
+      "loss": 2.548,
+      "step": 120000
+    },
+    {
+      "epoch": 1.78,
+      "eval_loss": 2.4832136631011963,
+      "eval_runtime": 1173.7957,
+      "eval_samples_per_second": 407.668,
+      "eval_steps_per_second": 50.959,
+      "step": 120000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 4.552324199013256e-05,
+      "loss": 2.5339,
+      "step": 120500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 4.550466623075552e-05,
+      "loss": 2.5376,
+      "step": 121000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.5486090471378474e-05,
+      "loss": 2.5418,
+      "step": 121500
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 4.546751471200143e-05,
+      "loss": 2.5264,
+      "step": 122000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 4.544893895262439e-05,
+      "loss": 2.5246,
+      "step": 122500
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 4.5430363193247345e-05,
+      "loss": 2.5272,
+      "step": 123000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.54117874338703e-05,
+      "loss": 2.5343,
+      "step": 123500
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 4.539321167449326e-05,
+      "loss": 2.5353,
+      "step": 124000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 4.5374635915116215e-05,
+      "loss": 2.5169,
+      "step": 124500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 4.5356060155739166e-05,
+      "loss": 2.5234,
+      "step": 125000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 4.533748439636212e-05,
+      "loss": 2.5327,
+      "step": 125500
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 4.531890863698508e-05,
+      "loss": 2.5133,
+      "step": 126000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.5300332877608036e-05,
+      "loss": 2.511,
+      "step": 126500
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.528175711823099e-05,
+      "loss": 2.5247,
+      "step": 127000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 4.526318135885395e-05,
+      "loss": 2.5223,
+      "step": 127500
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 4.524460559947691e-05,
+      "loss": 2.519,
+      "step": 128000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 4.5226029840099864e-05,
+      "loss": 2.5239,
+      "step": 128500
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 4.520745408072282e-05,
+      "loss": 2.5309,
+      "step": 129000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 4.518887832134578e-05,
+      "loss": 2.5285,
+      "step": 129500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 4.5170302561968734e-05,
+      "loss": 2.5085,
+      "step": 130000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 4.515172680259169e-05,
+      "loss": 2.5184,
+      "step": 130500
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.513315104321465e-05,
+      "loss": 2.5042,
+      "step": 131000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 4.5114575283837605e-05,
+      "loss": 2.5227,
+      "step": 131500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 4.509599952446056e-05,
+      "loss": 2.5163,
+      "step": 132000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 4.507742376508352e-05,
+      "loss": 2.516,
+      "step": 132500
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.5058848005706475e-05,
+      "loss": 2.509,
+      "step": 133000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.504027224632943e-05,
+      "loss": 2.4914,
+      "step": 133500
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 4.502169648695239e-05,
+      "loss": 2.4957,
+      "step": 134000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.5003120727575346e-05,
+      "loss": 2.5123,
+      "step": 134500
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.49845449681983e-05,
+      "loss": 2.4971,
+      "step": 135000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.496596920882126e-05,
+      "loss": 2.5011,
+      "step": 135500
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.494739344944422e-05,
+      "loss": 2.4925,
+      "step": 136000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.4928817690067174e-05,
+      "loss": 2.4903,
+      "step": 136500
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.491024193069013e-05,
+      "loss": 2.4814,
+      "step": 137000
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.489166617131309e-05,
+      "loss": 2.4838,
+      "step": 137500
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.4873090411936044e-05,
+      "loss": 2.4835,
+      "step": 138000
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.4854514652559e-05,
+      "loss": 2.4766,
+      "step": 138500
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.483593889318196e-05,
+      "loss": 2.5004,
+      "step": 139000
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.4817363133804915e-05,
+      "loss": 2.4843,
+      "step": 139500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.479878737442787e-05,
+      "loss": 2.484,
+      "step": 140000
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 2.4386236667633057,
+      "eval_runtime": 1167.064,
+      "eval_samples_per_second": 410.02,
+      "eval_steps_per_second": 51.253,
+      "step": 140000
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.478021161505083e-05,
+      "loss": 2.4955,
+      "step": 140500
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.4761635855673785e-05,
+      "loss": 2.4862,
+      "step": 141000
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.474306009629674e-05,
+      "loss": 2.4922,
+      "step": 141500
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.47244843369197e-05,
+      "loss": 2.4734,
+      "step": 142000
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.4705908577542656e-05,
+      "loss": 2.4911,
+      "step": 142500
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.468733281816561e-05,
+      "loss": 2.4899,
+      "step": 143000
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.466875705878856e-05,
+      "loss": 2.4889,
+      "step": 143500
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.465018129941152e-05,
+      "loss": 2.4845,
+      "step": 144000
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.463160554003448e-05,
+      "loss": 2.4755,
+      "step": 144500
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 4.4613029780657434e-05,
+      "loss": 2.4836,
+      "step": 145000
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 4.459445402128039e-05,
+      "loss": 2.4606,
+      "step": 145500
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 4.457587826190335e-05,
+      "loss": 2.4963,
+      "step": 146000
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.4557302502526304e-05,
+      "loss": 2.4754,
+      "step": 146500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 4.453872674314926e-05,
+      "loss": 2.4714,
+      "step": 147000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 4.452015098377222e-05,
+      "loss": 2.466,
+      "step": 147500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 4.4501575224395175e-05,
+      "loss": 2.4702,
+      "step": 148000
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 4.448299946501813e-05,
+      "loss": 2.4735,
+      "step": 148500
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 4.446442370564109e-05,
+      "loss": 2.4641,
+      "step": 149000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 4.4445847946264045e-05,
+      "loss": 2.4772,
+      "step": 149500
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 4.4427272186887e-05,
+      "loss": 2.4615,
+      "step": 150000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 4.440869642750996e-05,
+      "loss": 2.4603,
+      "step": 150500
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 4.4390120668132916e-05,
+      "loss": 2.4679,
+      "step": 151000
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 4.437154490875587e-05,
+      "loss": 2.4712,
+      "step": 151500
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 4.435296914937883e-05,
+      "loss": 2.4766,
+      "step": 152000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.433439339000178e-05,
+      "loss": 2.4638,
+      "step": 152500
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 4.431581763062474e-05,
+      "loss": 2.4565,
+      "step": 153000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 4.4297241871247694e-05,
+      "loss": 2.4677,
+      "step": 153500
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 4.427866611187065e-05,
+      "loss": 2.4417,
+      "step": 154000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.426009035249361e-05,
+      "loss": 2.4655,
+      "step": 154500
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 4.4241514593116564e-05,
+      "loss": 2.4513,
+      "step": 155000
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 4.422293883373953e-05,
+      "loss": 2.442,
+      "step": 155500
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 4.4204363074362485e-05,
+      "loss": 2.4576,
+      "step": 156000
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.418578731498544e-05,
+      "loss": 2.4634,
+      "step": 156500
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 4.41672115556084e-05,
+      "loss": 2.4598,
+      "step": 157000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 4.4148635796231355e-05,
+      "loss": 2.4593,
+      "step": 157500
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 4.413006003685431e-05,
+      "loss": 2.458,
+      "step": 158000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.411148427747727e-05,
+      "loss": 2.4646,
+      "step": 158500
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 4.4092908518100226e-05,
+      "loss": 2.4422,
+      "step": 159000
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 4.407433275872318e-05,
+      "loss": 2.457,
+      "step": 159500
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 4.405575699934614e-05,
+      "loss": 2.4523,
+      "step": 160000
+    },
+    {
+      "epoch": 2.38,
+      "eval_loss": 2.4072976112365723,
+      "eval_runtime": 1169.4217,
+      "eval_samples_per_second": 409.193,
+      "eval_steps_per_second": 51.149,
+      "step": 160000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.4037181239969097e-05,
+      "loss": 2.4538,
+      "step": 160500
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 4.401860548059205e-05,
+      "loss": 2.4586,
+      "step": 161000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 4.400002972121501e-05,
+      "loss": 2.4356,
+      "step": 161500
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 4.398145396183796e-05,
+      "loss": 2.4499,
+      "step": 162000
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 4.396287820246092e-05,
+      "loss": 2.4595,
+      "step": 162500
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 4.3944302443083874e-05,
+      "loss": 2.4467,
+      "step": 163000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 4.392572668370683e-05,
+      "loss": 2.4433,
+      "step": 163500
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 4.390715092432979e-05,
+      "loss": 2.4355,
+      "step": 164000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 4.3888575164952745e-05,
+      "loss": 2.4455,
+      "step": 164500
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 4.38699994055757e-05,
+      "loss": 2.4448,
+      "step": 165000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 4.385142364619866e-05,
+      "loss": 2.4265,
+      "step": 165500
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 4.3832847886821615e-05,
+      "loss": 2.4565,
+      "step": 166000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 4.381427212744457e-05,
+      "loss": 2.4497,
+      "step": 166500
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 4.379569636806753e-05,
+      "loss": 2.4439,
+      "step": 167000
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 4.3777120608690486e-05,
+      "loss": 2.4371,
+      "step": 167500
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 4.375854484931344e-05,
+      "loss": 2.4391,
+      "step": 168000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 4.37399690899364e-05,
+      "loss": 2.4305,
+      "step": 168500
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 4.3721393330559357e-05,
+      "loss": 2.4374,
+      "step": 169000
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 4.3702817571182313e-05,
+      "loss": 2.4512,
+      "step": 169500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 4.368424181180527e-05,
+      "loss": 2.4332,
+      "step": 170000
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 4.366566605242823e-05,
+      "loss": 2.4393,
+      "step": 170500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 4.364709029305118e-05,
+      "loss": 2.441,
+      "step": 171000
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 4.3628514533674134e-05,
+      "loss": 2.4285,
+      "step": 171500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 4.360993877429709e-05,
+      "loss": 2.4414,
+      "step": 172000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 4.359136301492005e-05,
+      "loss": 2.4428,
+      "step": 172500
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 4.3572787255543005e-05,
+      "loss": 2.4217,
+      "step": 173000
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 4.355421149616596e-05,
+      "loss": 2.4316,
+      "step": 173500
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 4.353563573678892e-05,
+      "loss": 2.436,
+      "step": 174000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 4.3517059977411875e-05,
+      "loss": 2.434,
+      "step": 174500
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 4.349848421803483e-05,
+      "loss": 2.446,
+      "step": 175000
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 4.347990845865779e-05,
+      "loss": 2.4214,
+      "step": 175500
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 4.3461332699280746e-05,
+      "loss": 2.4192,
+      "step": 176000
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 4.34427569399037e-05,
+      "loss": 2.4253,
+      "step": 176500
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 4.342418118052666e-05,
+      "loss": 2.4183,
+      "step": 177000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 4.340560542114962e-05,
+      "loss": 2.4335,
+      "step": 177500
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 4.338702966177258e-05,
+      "loss": 2.4218,
+      "step": 178000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 4.336845390239554e-05,
+      "loss": 2.4254,
+      "step": 178500
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 4.3349878143018494e-05,
+      "loss": 2.4325,
+      "step": 179000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 4.333130238364145e-05,
+      "loss": 2.4248,
+      "step": 179500
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 4.331272662426441e-05,
+      "loss": 2.4108,
+      "step": 180000
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 2.3746867179870605,
+      "eval_runtime": 1167.3479,
+      "eval_samples_per_second": 409.92,
+      "eval_steps_per_second": 51.24,
+      "step": 180000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 4.329415086488736e-05,
+      "loss": 2.4213,
+      "step": 180500
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 4.3275575105510315e-05,
+      "loss": 2.4314,
+      "step": 181000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.325699934613327e-05,
+      "loss": 2.4087,
+      "step": 181500
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 4.323842358675623e-05,
+      "loss": 2.4095,
+      "step": 182000
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.3219847827379185e-05,
+      "loss": 2.4348,
+      "step": 182500
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.320127206800214e-05,
+      "loss": 2.4169,
+      "step": 183000
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.31826963086251e-05,
+      "loss": 2.4208,
+      "step": 183500
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.3164120549248056e-05,
+      "loss": 2.4124,
+      "step": 184000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.314554478987101e-05,
+      "loss": 2.4248,
+      "step": 184500
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 4.312696903049397e-05,
+      "loss": 2.4132,
+      "step": 185000
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.3108393271116926e-05,
+      "loss": 2.4087,
+      "step": 185500
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 4.308981751173988e-05,
+      "loss": 2.4107,
+      "step": 186000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 4.307124175236284e-05,
+      "loss": 2.4198,
+      "step": 186500
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 4.30526659929858e-05,
+      "loss": 2.4287,
+      "step": 187000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 4.3034090233608754e-05,
+      "loss": 2.4077,
+      "step": 187500
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 4.301551447423171e-05,
+      "loss": 2.4149,
+      "step": 188000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 4.299693871485467e-05,
+      "loss": 2.4122,
+      "step": 188500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 4.2978362955477625e-05,
+      "loss": 2.4147,
+      "step": 189000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 4.2959787196100575e-05,
+      "loss": 2.4117,
+      "step": 189500
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 4.294121143672353e-05,
+      "loss": 2.3958,
+      "step": 190000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 4.292263567734649e-05,
+      "loss": 2.4063,
+      "step": 190500
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 4.2904059917969445e-05,
+      "loss": 2.4023,
+      "step": 191000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 4.28854841585924e-05,
+      "loss": 2.4059,
+      "step": 191500
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 4.286690839921536e-05,
+      "loss": 2.4122,
+      "step": 192000
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 4.2848332639838316e-05,
+      "loss": 2.3919,
+      "step": 192500
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 4.282975688046127e-05,
+      "loss": 2.4151,
+      "step": 193000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 4.281118112108423e-05,
+      "loss": 2.4087,
+      "step": 193500
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 4.2792605361707186e-05,
+      "loss": 2.4097,
+      "step": 194000
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 4.277402960233014e-05,
+      "loss": 2.4022,
+      "step": 194500
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 4.27554538429531e-05,
+      "loss": 2.4093,
+      "step": 195000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.273687808357606e-05,
+      "loss": 2.3997,
+      "step": 195500
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.2718302324199014e-05,
+      "loss": 2.3854,
+      "step": 196000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.269972656482197e-05,
+      "loss": 2.4188,
+      "step": 196500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.268115080544493e-05,
+      "loss": 2.4187,
+      "step": 197000
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 4.2662575046067885e-05,
+      "loss": 2.4008,
+      "step": 197500
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 4.264399928669084e-05,
+      "loss": 2.4086,
+      "step": 198000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 4.26254235273138e-05,
+      "loss": 2.4052,
+      "step": 198500
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 4.2606847767936755e-05,
+      "loss": 2.3966,
+      "step": 199000
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 4.258827200855971e-05,
+      "loss": 2.3967,
+      "step": 199500
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.256969624918267e-05,
+      "loss": 2.4206,
+      "step": 200000
+    },
+    {
+      "epoch": 2.97,
+      "eval_loss": 2.3452727794647217,
+      "eval_runtime": 1168.756,
+      "eval_samples_per_second": 409.426,
+      "eval_steps_per_second": 51.178,
+      "step": 200000
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 4.2551120489805626e-05,
+      "loss": 2.3994,
+      "step": 200500
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.253254473042858e-05,
+      "loss": 2.4004,
+      "step": 201000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.251396897105154e-05,
+      "loss": 2.3882,
+      "step": 201500
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.2495393211674496e-05,
+      "loss": 2.3894,
+      "step": 202000
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 4.247681745229745e-05,
+      "loss": 2.3908,
+      "step": 202500
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 4.245824169292041e-05,
+      "loss": 2.3832,
+      "step": 203000
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 4.243966593354337e-05,
+      "loss": 2.3835,
+      "step": 203500
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 4.2421090174166324e-05,
+      "loss": 2.374,
+      "step": 204000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 4.240251441478928e-05,
+      "loss": 2.3812,
+      "step": 204500
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 4.238393865541224e-05,
+      "loss": 2.3771,
+      "step": 205000
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 4.2365362896035194e-05,
+      "loss": 2.3816,
+      "step": 205500
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 4.234678713665815e-05,
+      "loss": 2.3849,
+      "step": 206000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 4.232821137728111e-05,
+      "loss": 2.3945,
+      "step": 206500
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 4.2309635617904065e-05,
+      "loss": 2.3823,
+      "step": 207000
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 4.229105985852702e-05,
+      "loss": 2.3742,
+      "step": 207500
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 4.227248409914997e-05,
+      "loss": 2.3822,
+      "step": 208000
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 4.225390833977293e-05,
+      "loss": 2.3919,
+      "step": 208500
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 4.2235332580395886e-05,
+      "loss": 2.3726,
+      "step": 209000
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 4.221675682101884e-05,
+      "loss": 2.3727,
+      "step": 209500
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 4.21981810616418e-05,
+      "loss": 2.3765,
+      "step": 210000
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 4.2179605302264756e-05,
+      "loss": 2.3831,
+      "step": 210500
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.216102954288771e-05,
+      "loss": 2.3838,
+      "step": 211000
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 4.214245378351067e-05,
+      "loss": 2.3788,
+      "step": 211500
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 4.212387802413363e-05,
+      "loss": 2.3722,
+      "step": 212000
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 4.2105302264756584e-05,
+      "loss": 2.3849,
+      "step": 212500
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 4.208672650537954e-05,
+      "loss": 2.3757,
+      "step": 213000
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 4.20681507460025e-05,
+      "loss": 2.3825,
+      "step": 213500
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 4.2049574986625454e-05,
+      "loss": 2.3783,
+      "step": 214000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 4.203099922724841e-05,
+      "loss": 2.3588,
+      "step": 214500
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 4.201242346787137e-05,
+      "loss": 2.3763,
+      "step": 215000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 4.1993847708494325e-05,
+      "loss": 2.3748,
+      "step": 215500
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 4.197527194911728e-05,
+      "loss": 2.3613,
+      "step": 216000
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 4.195669618974024e-05,
+      "loss": 2.385,
+      "step": 216500
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 4.1938120430363196e-05,
+      "loss": 2.3727,
+      "step": 217000
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 4.191954467098615e-05,
+      "loss": 2.3911,
+      "step": 217500
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 4.190096891160911e-05,
+      "loss": 2.358,
+      "step": 218000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 4.1882393152232066e-05,
+      "loss": 2.3718,
+      "step": 218500
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 4.186381739285502e-05,
+      "loss": 2.3734,
+      "step": 219000
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 4.184524163347798e-05,
+      "loss": 2.3699,
+      "step": 219500
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 4.182666587410094e-05,
+      "loss": 2.362,
+      "step": 220000
+    },
+    {
+      "epoch": 3.27,
+      "eval_loss": 2.3277173042297363,
+      "eval_runtime": 1172.0653,
+      "eval_samples_per_second": 408.27,
+      "eval_steps_per_second": 51.034,
+      "step": 220000
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 4.1808090114723894e-05,
+      "loss": 2.378,
+      "step": 220500
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 4.178951435534685e-05,
+      "loss": 2.3702,
+      "step": 221000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 4.177093859596981e-05,
+      "loss": 2.3787,
+      "step": 221500
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 4.1752362836592764e-05,
+      "loss": 2.3755,
+      "step": 222000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 4.173378707721572e-05,
+      "loss": 2.3809,
+      "step": 222500
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 4.171521131783868e-05,
+      "loss": 2.3747,
+      "step": 223000
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 4.1696635558461635e-05,
+      "loss": 2.3726,
+      "step": 223500
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 4.167805979908459e-05,
+      "loss": 2.3743,
+      "step": 224000
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 4.165948403970755e-05,
+      "loss": 2.3715,
+      "step": 224500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 4.1640908280330506e-05,
+      "loss": 2.3734,
+      "step": 225000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 4.162233252095346e-05,
+      "loss": 2.3728,
+      "step": 225500
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 4.160375676157642e-05,
+      "loss": 2.3459,
+      "step": 226000
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 4.158518100219937e-05,
+      "loss": 2.3441,
+      "step": 226500
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 4.1566605242822326e-05,
+      "loss": 2.3704,
+      "step": 227000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 4.154802948344528e-05,
+      "loss": 2.3557,
+      "step": 227500
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 4.152945372406824e-05,
+      "loss": 2.3613,
+      "step": 228000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 4.15108779646912e-05,
+      "loss": 2.3701,
+      "step": 228500
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 4.1492302205314154e-05,
+      "loss": 2.3622,
+      "step": 229000
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 4.147372644593711e-05,
+      "loss": 2.3577,
+      "step": 229500
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 4.145515068656007e-05,
+      "loss": 2.3664,
+      "step": 230000
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 4.1436574927183024e-05,
+      "loss": 2.3697,
+      "step": 230500
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 4.141799916780598e-05,
+      "loss": 2.3642,
+      "step": 231000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 4.139942340842894e-05,
+      "loss": 2.3529,
+      "step": 231500
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 4.1380847649051895e-05,
+      "loss": 2.358,
+      "step": 232000
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 4.136227188967485e-05,
+      "loss": 2.3583,
+      "step": 232500
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 4.134369613029781e-05,
+      "loss": 2.3387,
+      "step": 233000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 4.1325120370920766e-05,
+      "loss": 2.3523,
+      "step": 233500
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 4.130654461154372e-05,
+      "loss": 2.357,
+      "step": 234000
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 4.128796885216668e-05,
+      "loss": 2.3451,
+      "step": 234500
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 4.1269393092789636e-05,
+      "loss": 2.3608,
+      "step": 235000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 4.1250817333412586e-05,
+      "loss": 2.3598,
+      "step": 235500
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.123224157403554e-05,
+      "loss": 2.3555,
+      "step": 236000
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 4.12136658146585e-05,
+      "loss": 2.3693,
+      "step": 236500
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 4.119509005528146e-05,
+      "loss": 2.3603,
+      "step": 237000
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 4.1176514295904414e-05,
+      "loss": 2.3474,
+      "step": 237500
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 4.115793853652737e-05,
+      "loss": 2.3459,
+      "step": 238000
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 4.1139362777150334e-05,
+      "loss": 2.3358,
+      "step": 238500
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 4.112078701777329e-05,
+      "loss": 2.3561,
+      "step": 239000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 4.110221125839625e-05,
+      "loss": 2.3444,
+      "step": 239500
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 4.1083635499019205e-05,
+      "loss": 2.3487,
+      "step": 240000
+    },
+    {
+      "epoch": 3.57,
+      "eval_loss": 2.3072495460510254,
+      "eval_runtime": 1169.8184,
+      "eval_samples_per_second": 409.054,
+      "eval_steps_per_second": 51.132,
+      "step": 240000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 4.106505973964216e-05,
+      "loss": 2.3609,
+      "step": 240500
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 4.104648398026512e-05,
+      "loss": 2.3558,
+      "step": 241000
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 4.1027908220888075e-05,
+      "loss": 2.3436,
+      "step": 241500
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 4.100933246151103e-05,
+      "loss": 2.3591,
+      "step": 242000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 4.099075670213399e-05,
+      "loss": 2.351,
+      "step": 242500
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 4.0972180942756946e-05,
+      "loss": 2.35,
+      "step": 243000
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 4.09536051833799e-05,
+      "loss": 2.3409,
+      "step": 243500
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 4.093502942400286e-05,
+      "loss": 2.3378,
+      "step": 244000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 4.091645366462582e-05,
+      "loss": 2.3332,
+      "step": 244500
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 4.089787790524877e-05,
+      "loss": 2.3397,
+      "step": 245000
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 4.0879302145871724e-05,
+      "loss": 2.3522,
+      "step": 245500
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 4.086072638649468e-05,
+      "loss": 2.337,
+      "step": 246000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 4.084215062711764e-05,
+      "loss": 2.338,
+      "step": 246500
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 4.0823574867740594e-05,
+      "loss": 2.3575,
+      "step": 247000
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 4.080499910836355e-05,
+      "loss": 2.3475,
+      "step": 247500
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 4.078642334898651e-05,
+      "loss": 2.342,
+      "step": 248000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 4.0767847589609465e-05,
+      "loss": 2.3267,
+      "step": 248500
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 4.074927183023242e-05,
+      "loss": 2.3542,
+      "step": 249000
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 4.073069607085538e-05,
+      "loss": 2.3408,
+      "step": 249500
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 4.0712120311478335e-05,
+      "loss": 2.3391,
+      "step": 250000
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 4.069354455210129e-05,
+      "loss": 2.3637,
+      "step": 250500
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 4.067496879272425e-05,
+      "loss": 2.3387,
+      "step": 251000
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 4.0656393033347206e-05,
+      "loss": 2.343,
+      "step": 251500
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 4.063781727397016e-05,
+      "loss": 2.3342,
+      "step": 252000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 4.061924151459312e-05,
+      "loss": 2.3491,
+      "step": 252500
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 4.060066575521608e-05,
+      "loss": 2.3389,
+      "step": 253000
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 4.0582089995839034e-05,
+      "loss": 2.3436,
+      "step": 253500
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 4.0563514236461984e-05,
+      "loss": 2.3422,
+      "step": 254000
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 4.054493847708494e-05,
+      "loss": 2.3417,
+      "step": 254500
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 4.05263627177079e-05,
+      "loss": 2.3427,
+      "step": 255000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 4.0507786958330854e-05,
+      "loss": 2.3573,
+      "step": 255500
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 4.048921119895381e-05,
+      "loss": 2.3288,
+      "step": 256000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 4.047063543957677e-05,
+      "loss": 2.3442,
+      "step": 256500
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 4.0452059680199725e-05,
+      "loss": 2.3388,
+      "step": 257000
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 4.043348392082268e-05,
+      "loss": 2.3437,
+      "step": 257500
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 4.041490816144564e-05,
+      "loss": 2.3443,
+      "step": 258000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 4.0396332402068596e-05,
+      "loss": 2.3411,
+      "step": 258500
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 4.037775664269155e-05,
+      "loss": 2.3433,
+      "step": 259000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 4.035918088331451e-05,
+      "loss": 2.3394,
+      "step": 259500
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 4.0340605123937466e-05,
+      "loss": 2.3362,
+      "step": 260000
+    },
+    {
+      "epoch": 3.86,
+      "eval_loss": 2.28617525100708,
+      "eval_runtime": 1173.3291,
+      "eval_samples_per_second": 407.83,
+      "eval_steps_per_second": 50.979,
+      "step": 260000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 4.032202936456042e-05,
+      "loss": 2.3356,
+      "step": 260500
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 4.030345360518339e-05,
+      "loss": 2.3292,
+      "step": 261000
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 4.0284877845806343e-05,
+      "loss": 2.3394,
+      "step": 261500
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 4.02663020864293e-05,
+      "loss": 2.3251,
+      "step": 262000
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 4.024772632705226e-05,
+      "loss": 2.3336,
+      "step": 262500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 4.0229150567675214e-05,
+      "loss": 2.3347,
+      "step": 263000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 4.0210574808298164e-05,
+      "loss": 2.322,
+      "step": 263500
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 4.019199904892112e-05,
+      "loss": 2.3369,
+      "step": 264000
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 4.017342328954408e-05,
+      "loss": 2.3512,
+      "step": 264500
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 4.0154847530167035e-05,
+      "loss": 2.3244,
+      "step": 265000
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 4.013627177078999e-05,
+      "loss": 2.3438,
+      "step": 265500
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 4.011769601141295e-05,
+      "loss": 2.3308,
+      "step": 266000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 4.0099120252035905e-05,
+      "loss": 2.3381,
+      "step": 266500
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 4.008054449265886e-05,
+      "loss": 2.3419,
+      "step": 267000
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 4.006196873328182e-05,
+      "loss": 2.32,
+      "step": 267500
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 4.0043392973904776e-05,
+      "loss": 2.3292,
+      "step": 268000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 4.002481721452773e-05,
+      "loss": 2.3299,
+      "step": 268500
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.000624145515069e-05,
+      "loss": 2.3196,
+      "step": 269000
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 3.998766569577365e-05,
+      "loss": 2.3343,
+      "step": 269500
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 3.9969089936396604e-05,
+      "loss": 2.3204,
+      "step": 270000
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 3.995051417701956e-05,
+      "loss": 2.3187,
+      "step": 270500
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.993193841764252e-05,
+      "loss": 2.3179,
+      "step": 271000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 3.9913362658265474e-05,
+      "loss": 2.3186,
+      "step": 271500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 3.989478689888843e-05,
+      "loss": 2.3148,
+      "step": 272000
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 3.987621113951138e-05,
+      "loss": 2.3124,
+      "step": 272500
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.985763538013434e-05,
+      "loss": 2.3076,
+      "step": 273000
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.9839059620757295e-05,
+      "loss": 2.3126,
+      "step": 273500
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 3.982048386138025e-05,
+      "loss": 2.3095,
+      "step": 274000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 3.980190810200321e-05,
+      "loss": 2.3168,
+      "step": 274500
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.9783332342626165e-05,
+      "loss": 2.3255,
+      "step": 275000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 3.976475658324912e-05,
+      "loss": 2.3307,
+      "step": 275500
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 3.974618082387208e-05,
+      "loss": 2.3287,
+      "step": 276000
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 3.9727605064495036e-05,
+      "loss": 2.3289,
+      "step": 276500
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 3.970902930511799e-05,
+      "loss": 2.3022,
+      "step": 277000
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 3.969045354574095e-05,
+      "loss": 2.3183,
+      "step": 277500
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.967187778636391e-05,
+      "loss": 2.3131,
+      "step": 278000
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 3.9653302026986864e-05,
+      "loss": 2.3048,
+      "step": 278500
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 3.963472626760982e-05,
+      "loss": 2.3163,
+      "step": 279000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 3.961615050823278e-05,
+      "loss": 2.2998,
+      "step": 279500
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 3.9597574748855734e-05,
+      "loss": 2.3124,
+      "step": 280000
+    },
+    {
+      "epoch": 4.16,
+      "eval_loss": 2.2729697227478027,
+      "eval_runtime": 1170.2787,
+      "eval_samples_per_second": 408.893,
+      "eval_steps_per_second": 51.112,
+      "step": 280000
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 3.957899898947869e-05,
+      "loss": 2.3128,
+      "step": 280500
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 3.956042323010165e-05,
+      "loss": 2.3044,
+      "step": 281000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 3.9541847470724605e-05,
+      "loss": 2.2981,
+      "step": 281500
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 3.952327171134756e-05,
+      "loss": 2.3192,
+      "step": 282000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 3.950469595197052e-05,
+      "loss": 2.3131,
+      "step": 282500
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 3.9486120192593475e-05,
+      "loss": 2.3026,
+      "step": 283000
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 3.946754443321643e-05,
+      "loss": 2.3117,
+      "step": 283500
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 3.944896867383939e-05,
+      "loss": 2.3112,
+      "step": 284000
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 3.9430392914462346e-05,
+      "loss": 2.2928,
+      "step": 284500
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 3.94118171550853e-05,
+      "loss": 2.3177,
+      "step": 285000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 3.939324139570826e-05,
+      "loss": 2.2941,
+      "step": 285500
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 3.9374665636331217e-05,
+      "loss": 2.3002,
+      "step": 286000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 3.9356089876954173e-05,
+      "loss": 2.313,
+      "step": 286500
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 3.933751411757713e-05,
+      "loss": 2.3148,
+      "step": 287000
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 3.931893835820009e-05,
+      "loss": 2.2997,
+      "step": 287500
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 3.9300362598823044e-05,
+      "loss": 2.3193,
+      "step": 288000
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 3.9281786839446e-05,
+      "loss": 2.2981,
+      "step": 288500
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 3.926321108006896e-05,
+      "loss": 2.3109,
+      "step": 289000
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 3.9244635320691915e-05,
+      "loss": 2.3225,
+      "step": 289500
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 3.922605956131487e-05,
+      "loss": 2.3158,
+      "step": 290000
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 3.920748380193783e-05,
+      "loss": 2.3111,
+      "step": 290500
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 3.918890804256078e-05,
+      "loss": 2.3073,
+      "step": 291000
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 3.9170332283183735e-05,
+      "loss": 2.31,
+      "step": 291500
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 3.915175652380669e-05,
+      "loss": 2.3014,
+      "step": 292000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 3.913318076442965e-05,
+      "loss": 2.3115,
+      "step": 292500
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 3.9114605005052606e-05,
+      "loss": 2.3114,
+      "step": 293000
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 3.909602924567556e-05,
+      "loss": 2.2953,
+      "step": 293500
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 3.907745348629852e-05,
+      "loss": 2.3095,
+      "step": 294000
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 3.9058877726921477e-05,
+      "loss": 2.306,
+      "step": 294500
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 3.9040301967544433e-05,
+      "loss": 2.2984,
+      "step": 295000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 3.902172620816739e-05,
+      "loss": 2.3158,
+      "step": 295500
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 3.900315044879035e-05,
+      "loss": 2.3099,
+      "step": 296000
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.8984574689413304e-05,
+      "loss": 2.3073,
+      "step": 296500
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 3.896599893003626e-05,
+      "loss": 2.2892,
+      "step": 297000
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 3.894742317065922e-05,
+      "loss": 2.2994,
+      "step": 297500
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.8928847411282175e-05,
+      "loss": 2.3005,
+      "step": 298000
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.891027165190513e-05,
+      "loss": 2.2941,
+      "step": 298500
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 3.889169589252809e-05,
+      "loss": 2.2916,
+      "step": 299000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 3.8873120133151045e-05,
+      "loss": 2.2974,
+      "step": 299500
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 3.8854544373774e-05,
+      "loss": 2.3023,
+      "step": 300000
+    },
+    {
+      "epoch": 4.46,
+      "eval_loss": 2.2589497566223145,
+      "eval_runtime": 1170.5911,
+      "eval_samples_per_second": 408.784,
+      "eval_steps_per_second": 51.098,
+      "step": 300000
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.883596861439696e-05,
+      "loss": 2.3098,
+      "step": 300500
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 3.8817392855019916e-05,
+      "loss": 2.3006,
+      "step": 301000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 3.879881709564287e-05,
+      "loss": 2.3057,
+      "step": 301500
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 3.878024133626583e-05,
+      "loss": 2.2969,
+      "step": 302000
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.8761665576888786e-05,
+      "loss": 2.2979,
+      "step": 302500
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 3.874308981751174e-05,
+      "loss": 2.3112,
+      "step": 303000
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.87245140581347e-05,
+      "loss": 2.301,
+      "step": 303500
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 3.870593829875766e-05,
+      "loss": 2.3021,
+      "step": 304000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.8687362539380614e-05,
+      "loss": 2.2979,
+      "step": 304500
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 3.866878678000357e-05,
+      "loss": 2.2935,
+      "step": 305000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 3.865021102062653e-05,
+      "loss": 2.3147,
+      "step": 305500
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 3.8631635261249485e-05,
+      "loss": 2.2913,
+      "step": 306000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 3.861305950187244e-05,
+      "loss": 2.2993,
+      "step": 306500
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 3.85944837424954e-05,
+      "loss": 2.2963,
+      "step": 307000
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 3.8575907983118355e-05,
+      "loss": 2.3011,
+      "step": 307500
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 3.855733222374131e-05,
+      "loss": 2.299,
+      "step": 308000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 3.853875646436427e-05,
+      "loss": 2.2991,
+      "step": 308500
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 3.8520180704987226e-05,
+      "loss": 2.2924,
+      "step": 309000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 3.8501604945610176e-05,
+      "loss": 2.3082,
+      "step": 309500
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.848302918623313e-05,
+      "loss": 2.2975,
+      "step": 310000
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 3.846445342685609e-05,
+      "loss": 2.2793,
+      "step": 310500
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 3.8445877667479046e-05,
+      "loss": 2.2974,
+      "step": 311000
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 3.8427301908102e-05,
+      "loss": 2.2725,
+      "step": 311500
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 3.840872614872496e-05,
+      "loss": 2.2981,
+      "step": 312000
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 3.839015038934792e-05,
+      "loss": 2.2918,
+      "step": 312500
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 3.8371574629970874e-05,
+      "loss": 2.3004,
+      "step": 313000
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 3.835299887059383e-05,
+      "loss": 2.2885,
+      "step": 313500
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 3.833442311121679e-05,
+      "loss": 2.2865,
+      "step": 314000
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 3.8315847351839745e-05,
+      "loss": 2.2962,
+      "step": 314500
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 3.82972715924627e-05,
+      "loss": 2.2958,
+      "step": 315000
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 3.827869583308566e-05,
+      "loss": 2.2916,
+      "step": 315500
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 3.8260120073708615e-05,
+      "loss": 2.286,
+      "step": 316000
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 3.824154431433157e-05,
+      "loss": 2.2902,
+      "step": 316500
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 3.822296855495453e-05,
+      "loss": 2.285,
+      "step": 317000
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 3.8204392795577486e-05,
+      "loss": 2.2856,
+      "step": 317500
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 3.818581703620044e-05,
+      "loss": 2.2944,
+      "step": 318000
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 3.81672412768234e-05,
+      "loss": 2.2838,
+      "step": 318500
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 3.814866551744635e-05,
+      "loss": 2.2844,
+      "step": 319000
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 3.8130089758069306e-05,
+      "loss": 2.2976,
+      "step": 319500
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 3.811151399869226e-05,
+      "loss": 2.2939,
+      "step": 320000
+    },
+    {
+      "epoch": 4.76,
+      "eval_loss": 2.23964524269104,
+      "eval_runtime": 1167.8582,
+      "eval_samples_per_second": 409.741,
+      "eval_steps_per_second": 51.218,
+      "step": 320000
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 3.809293823931522e-05,
+      "loss": 2.2839,
+      "step": 320500
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 3.807436247993818e-05,
+      "loss": 2.2969,
+      "step": 321000
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 3.805578672056114e-05,
+      "loss": 2.2928,
+      "step": 321500
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 3.80372109611841e-05,
+      "loss": 2.2835,
+      "step": 322000
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 3.8018635201807054e-05,
+      "loss": 2.2863,
+      "step": 322500
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 3.800005944243001e-05,
+      "loss": 2.2792,
+      "step": 323000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 3.798148368305297e-05,
+      "loss": 2.2975,
+      "step": 323500
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 3.7962907923675925e-05,
+      "loss": 2.2984,
+      "step": 324000
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 3.794433216429888e-05,
+      "loss": 2.2908,
+      "step": 324500
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 3.792575640492184e-05,
+      "loss": 2.2916,
+      "step": 325000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 3.7907180645544796e-05,
+      "loss": 2.2923,
+      "step": 325500
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 3.788860488616775e-05,
+      "loss": 2.2853,
+      "step": 326000
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 3.787002912679071e-05,
+      "loss": 2.273,
+      "step": 326500
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 3.7851453367413666e-05,
+      "loss": 2.29,
+      "step": 327000
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 3.783287760803662e-05,
+      "loss": 2.2899,
+      "step": 327500
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 3.781430184865957e-05,
+      "loss": 2.2808,
+      "step": 328000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 3.779572608928253e-05,
+      "loss": 2.2711,
+      "step": 328500
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 3.777715032990549e-05,
+      "loss": 2.2725,
+      "step": 329000
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 3.7758574570528444e-05,
+      "loss": 2.2751,
+      "step": 329500
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 3.77399988111514e-05,
+      "loss": 2.2781,
+      "step": 330000
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 3.772142305177436e-05,
+      "loss": 2.272,
+      "step": 330500
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 3.7702847292397314e-05,
+      "loss": 2.2925,
+      "step": 331000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 3.768427153302027e-05,
+      "loss": 2.2816,
+      "step": 331500
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 3.766569577364323e-05,
+      "loss": 2.2754,
+      "step": 332000
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 3.7647120014266185e-05,
+      "loss": 2.2754,
+      "step": 332500
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 3.762854425488914e-05,
+      "loss": 2.2822,
+      "step": 333000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 3.76099684955121e-05,
+      "loss": 2.296,
+      "step": 333500
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 3.7591392736135056e-05,
+      "loss": 2.2843,
+      "step": 334000
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 3.757281697675801e-05,
+      "loss": 2.286,
+      "step": 334500
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 3.755424121738097e-05,
+      "loss": 2.2677,
+      "step": 335000
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 3.7535665458003926e-05,
+      "loss": 2.2799,
+      "step": 335500
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 3.751708969862688e-05,
+      "loss": 2.2951,
+      "step": 336000
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 3.749851393924984e-05,
+      "loss": 2.275,
+      "step": 336500
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 3.74799381798728e-05,
+      "loss": 2.2563,
+      "step": 337000
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 3.746136242049575e-05,
+      "loss": 2.2742,
+      "step": 337500
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 3.7442786661118704e-05,
+      "loss": 2.2554,
+      "step": 338000
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 3.742421090174166e-05,
+      "loss": 2.2593,
+      "step": 338500
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 3.740563514236462e-05,
+      "loss": 2.2618,
+      "step": 339000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 3.7387059382987574e-05,
+      "loss": 2.2789,
+      "step": 339500
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 3.736848362361053e-05,
+      "loss": 2.259,
+      "step": 340000
+    },
+    {
+      "epoch": 5.05,
+      "eval_loss": 2.2331488132476807,
+      "eval_runtime": 1167.7161,
+      "eval_samples_per_second": 409.791,
+      "eval_steps_per_second": 51.224,
+      "step": 340000
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 3.734990786423349e-05,
+      "loss": 2.2633,
+      "step": 340500
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 3.7331332104856445e-05,
+      "loss": 2.2652,
+      "step": 341000
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 3.73127563454794e-05,
+      "loss": 2.265,
+      "step": 341500
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 3.729418058610236e-05,
+      "loss": 2.257,
+      "step": 342000
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 3.7275604826725316e-05,
+      "loss": 2.2673,
+      "step": 342500
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 3.725702906734827e-05,
+      "loss": 2.2638,
+      "step": 343000
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 3.723845330797123e-05,
+      "loss": 2.2633,
+      "step": 343500
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 3.721987754859419e-05,
+      "loss": 2.2703,
+      "step": 344000
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 3.720130178921715e-05,
+      "loss": 2.2733,
+      "step": 344500
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 3.718272602984011e-05,
+      "loss": 2.2554,
+      "step": 345000
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 3.7164150270463064e-05,
+      "loss": 2.2698,
+      "step": 345500
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 3.714557451108602e-05,
+      "loss": 2.2485,
+      "step": 346000
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 3.712699875170897e-05,
+      "loss": 2.2506,
+      "step": 346500
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 3.710842299233193e-05,
+      "loss": 2.2747,
+      "step": 347000
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 3.7089847232954884e-05,
+      "loss": 2.2738,
+      "step": 347500
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 3.707127147357784e-05,
+      "loss": 2.2735,
+      "step": 348000
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 3.70526957142008e-05,
+      "loss": 2.2552,
+      "step": 348500
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 3.7034119954823755e-05,
+      "loss": 2.2401,
+      "step": 349000
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 3.701554419544671e-05,
+      "loss": 2.2791,
+      "step": 349500
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 3.699696843606967e-05,
+      "loss": 2.2582,
+      "step": 350000
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 3.6978392676692626e-05,
+      "loss": 2.2702,
+      "step": 350500
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 3.695981691731558e-05,
+      "loss": 2.2713,
+      "step": 351000
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 3.694124115793854e-05,
+      "loss": 2.2739,
+      "step": 351500
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 3.6922665398561496e-05,
+      "loss": 2.2761,
+      "step": 352000
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 3.690408963918445e-05,
+      "loss": 2.2729,
+      "step": 352500
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 3.688551387980741e-05,
+      "loss": 2.2559,
+      "step": 353000
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 3.686693812043037e-05,
+      "loss": 2.2592,
+      "step": 353500
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 3.6848362361053324e-05,
+      "loss": 2.2654,
+      "step": 354000
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 3.682978660167628e-05,
+      "loss": 2.2688,
+      "step": 354500
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 3.681121084229924e-05,
+      "loss": 2.2707,
+      "step": 355000
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 3.679263508292219e-05,
+      "loss": 2.2546,
+      "step": 355500
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 3.6774059323545144e-05,
+      "loss": 2.252,
+      "step": 356000
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 3.67554835641681e-05,
+      "loss": 2.2726,
+      "step": 356500
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 3.673690780479106e-05,
+      "loss": 2.2653,
+      "step": 357000
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 3.6718332045414015e-05,
+      "loss": 2.2718,
+      "step": 357500
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 3.669975628603697e-05,
+      "loss": 2.2759,
+      "step": 358000
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 3.668118052665993e-05,
+      "loss": 2.2726,
+      "step": 358500
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 3.6662604767282886e-05,
+      "loss": 2.2477,
+      "step": 359000
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 3.664402900790584e-05,
+      "loss": 2.2541,
+      "step": 359500
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 3.66254532485288e-05,
+      "loss": 2.2599,
+      "step": 360000
+    },
+    {
+      "epoch": 5.35,
+      "eval_loss": 2.2197089195251465,
+      "eval_runtime": 1171.2972,
+      "eval_samples_per_second": 408.538,
+      "eval_steps_per_second": 51.067,
+      "step": 360000
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 3.6606877489151756e-05,
+      "loss": 2.2723,
+      "step": 360500
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 3.658830172977471e-05,
+      "loss": 2.2763,
+      "step": 361000
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 3.656972597039767e-05,
+      "loss": 2.2848,
+      "step": 361500
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 3.655115021102063e-05,
+      "loss": 2.2703,
+      "step": 362000
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 3.6532574451643584e-05,
+      "loss": 2.2587,
+      "step": 362500
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 3.651399869226654e-05,
+      "loss": 2.2558,
+      "step": 363000
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 3.64954229328895e-05,
+      "loss": 2.2648,
+      "step": 363500
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 3.6476847173512454e-05,
+      "loss": 2.2641,
+      "step": 364000
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 3.645827141413541e-05,
+      "loss": 2.27,
+      "step": 364500
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 3.643969565475837e-05,
+      "loss": 2.2674,
+      "step": 365000
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 3.6421119895381325e-05,
+      "loss": 2.2521,
+      "step": 365500
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 3.640254413600428e-05,
+      "loss": 2.2581,
+      "step": 366000
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 3.638396837662724e-05,
+      "loss": 2.2676,
+      "step": 366500
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 3.6365392617250196e-05,
+      "loss": 2.2647,
+      "step": 367000
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 3.634681685787315e-05,
+      "loss": 2.2673,
+      "step": 367500
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 3.632824109849611e-05,
+      "loss": 2.2454,
+      "step": 368000
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 3.6309665339119066e-05,
+      "loss": 2.2705,
+      "step": 368500
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 3.629108957974202e-05,
+      "loss": 2.261,
+      "step": 369000
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 3.627251382036498e-05,
+      "loss": 2.2623,
+      "step": 369500
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 3.625393806098794e-05,
+      "loss": 2.2501,
+      "step": 370000
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 3.6235362301610894e-05,
+      "loss": 2.2412,
+      "step": 370500
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 3.621678654223385e-05,
+      "loss": 2.2633,
+      "step": 371000
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 3.619821078285681e-05,
+      "loss": 2.2673,
+      "step": 371500
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 3.6179635023479764e-05,
+      "loss": 2.254,
+      "step": 372000
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 3.616105926410272e-05,
+      "loss": 2.2498,
+      "step": 372500
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 3.614248350472568e-05,
+      "loss": 2.2674,
+      "step": 373000
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 3.6123907745348635e-05,
+      "loss": 2.2512,
+      "step": 373500
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 3.6105331985971585e-05,
+      "loss": 2.2631,
+      "step": 374000
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 3.608675622659454e-05,
+      "loss": 2.2676,
+      "step": 374500
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 3.60681804672175e-05,
+      "loss": 2.2579,
+      "step": 375000
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 3.6049604707840456e-05,
+      "loss": 2.2494,
+      "step": 375500
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 3.603102894846341e-05,
+      "loss": 2.2611,
+      "step": 376000
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 3.601245318908637e-05,
+      "loss": 2.2579,
+      "step": 376500
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 3.5993877429709326e-05,
+      "loss": 2.2426,
+      "step": 377000
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 3.597530167033228e-05,
+      "loss": 2.2619,
+      "step": 377500
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 3.595672591095524e-05,
+      "loss": 2.2522,
+      "step": 378000
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 3.59381501515782e-05,
+      "loss": 2.2407,
+      "step": 378500
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 3.5919574392201154e-05,
+      "loss": 2.246,
+      "step": 379000
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 3.590099863282411e-05,
+      "loss": 2.2502,
+      "step": 379500
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 3.588242287344707e-05,
+      "loss": 2.2592,
+      "step": 380000
+    },
+    {
+      "epoch": 5.65,
+      "eval_loss": 2.2073557376861572,
+      "eval_runtime": 1169.5113,
+      "eval_samples_per_second": 409.161,
+      "eval_steps_per_second": 51.145,
+      "step": 380000
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 3.5863847114070024e-05,
+      "loss": 2.2533,
+      "step": 380500
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 3.584527135469298e-05,
+      "loss": 2.2621,
+      "step": 381000
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 3.582669559531594e-05,
+      "loss": 2.271,
+      "step": 381500
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 3.5808119835938895e-05,
+      "loss": 2.2483,
+      "step": 382000
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 3.578954407656185e-05,
+      "loss": 2.2463,
+      "step": 382500
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 3.577096831718481e-05,
+      "loss": 2.2378,
+      "step": 383000
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 3.5752392557807765e-05,
+      "loss": 2.2455,
+      "step": 383500
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 3.573381679843072e-05,
+      "loss": 2.2523,
+      "step": 384000
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 3.571524103905368e-05,
+      "loss": 2.2671,
+      "step": 384500
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 3.5696665279676636e-05,
+      "loss": 2.2539,
+      "step": 385000
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 3.567808952029959e-05,
+      "loss": 2.2363,
+      "step": 385500
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 3.565951376092255e-05,
+      "loss": 2.2596,
+      "step": 386000
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 3.564093800154551e-05,
+      "loss": 2.2579,
+      "step": 386500
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 3.5622362242168464e-05,
+      "loss": 2.2377,
+      "step": 387000
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 3.560378648279142e-05,
+      "loss": 2.2446,
+      "step": 387500
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 3.558521072341438e-05,
+      "loss": 2.2591,
+      "step": 388000
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 3.5566634964037334e-05,
+      "loss": 2.2553,
+      "step": 388500
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 3.554805920466029e-05,
+      "loss": 2.2583,
+      "step": 389000
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 3.552948344528325e-05,
+      "loss": 2.2662,
+      "step": 389500
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 3.5510907685906205e-05,
+      "loss": 2.2561,
+      "step": 390000
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 3.549233192652916e-05,
+      "loss": 2.2549,
+      "step": 390500
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 3.547375616715212e-05,
+      "loss": 2.2479,
+      "step": 391000
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 3.5455180407775075e-05,
+      "loss": 2.252,
+      "step": 391500
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 3.543660464839803e-05,
+      "loss": 2.2437,
+      "step": 392000
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 3.541802888902098e-05,
+      "loss": 2.2543,
+      "step": 392500
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 3.539945312964394e-05,
+      "loss": 2.2443,
+      "step": 393000
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 3.5380877370266896e-05,
+      "loss": 2.2419,
+      "step": 393500
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 3.536230161088985e-05,
+      "loss": 2.246,
+      "step": 394000
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 3.534372585151281e-05,
+      "loss": 2.2477,
+      "step": 394500
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 3.532515009213577e-05,
+      "loss": 2.2246,
+      "step": 395000
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 3.5306574332758724e-05,
+      "loss": 2.239,
+      "step": 395500
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 3.528799857338168e-05,
+      "loss": 2.2489,
+      "step": 396000
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 3.526942281400464e-05,
+      "loss": 2.258,
+      "step": 396500
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 3.5250847054627594e-05,
+      "loss": 2.2508,
+      "step": 397000
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 3.523227129525055e-05,
+      "loss": 2.2514,
+      "step": 397500
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 3.521369553587351e-05,
+      "loss": 2.2521,
+      "step": 398000
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 3.5195119776496465e-05,
+      "loss": 2.2354,
+      "step": 398500
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 3.517654401711942e-05,
+      "loss": 2.2356,
+      "step": 399000
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 3.515796825774238e-05,
+      "loss": 2.2398,
+      "step": 399500
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 3.5139392498365335e-05,
+      "loss": 2.2509,
+      "step": 400000
+    },
+    {
+      "epoch": 5.94,
+      "eval_loss": 2.1972875595092773,
+      "eval_runtime": 1180.0595,
+      "eval_samples_per_second": 405.504,
+      "eval_steps_per_second": 50.688,
+      "step": 400000
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 3.512081673898829e-05,
+      "loss": 2.2406,
+      "step": 400500
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 3.510224097961125e-05,
+      "loss": 2.248,
+      "step": 401000
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 3.5083665220234206e-05,
+      "loss": 2.2454,
+      "step": 401500
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 3.5065089460857156e-05,
+      "loss": 2.2378,
+      "step": 402000
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 3.504651370148011e-05,
+      "loss": 2.2527,
+      "step": 402500
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 3.502793794210307e-05,
+      "loss": 2.2499,
+      "step": 403000
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.500936218272603e-05,
+      "loss": 2.2522,
+      "step": 403500
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.4990786423348984e-05,
+      "loss": 2.2534,
+      "step": 404000
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 3.497221066397194e-05,
+      "loss": 2.2333,
+      "step": 404500
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 3.4953634904594904e-05,
+      "loss": 2.2299,
+      "step": 405000
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 3.493505914521786e-05,
+      "loss": 2.2253,
+      "step": 405500
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 3.491648338584082e-05,
+      "loss": 2.2195,
+      "step": 406000
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 3.4897907626463775e-05,
+      "loss": 2.2268,
+      "step": 406500
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 3.487933186708673e-05,
+      "loss": 2.2463,
+      "step": 407000
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 3.486075610770969e-05,
+      "loss": 2.2348,
+      "step": 407500
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 3.4842180348332645e-05,
+      "loss": 2.2322,
+      "step": 408000
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 3.48236045889556e-05,
+      "loss": 2.2329,
+      "step": 408500
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 3.480502882957856e-05,
+      "loss": 2.2468,
+      "step": 409000
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 3.4786453070201516e-05,
+      "loss": 2.2346,
+      "step": 409500
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 3.476787731082447e-05,
+      "loss": 2.2336,
+      "step": 410000
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 3.474930155144743e-05,
+      "loss": 2.2469,
+      "step": 410500
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 3.473072579207038e-05,
+      "loss": 2.2323,
+      "step": 411000
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 3.4712150032693337e-05,
+      "loss": 2.2362,
+      "step": 411500
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 3.4693574273316293e-05,
+      "loss": 2.2301,
+      "step": 412000
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 3.467499851393925e-05,
+      "loss": 2.2322,
+      "step": 412500
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 3.465642275456221e-05,
+      "loss": 2.2366,
+      "step": 413000
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 3.4637846995185164e-05,
+      "loss": 2.2288,
+      "step": 413500
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 3.461927123580812e-05,
+      "loss": 2.2249,
+      "step": 414000
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 3.460069547643108e-05,
+      "loss": 2.223,
+      "step": 414500
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 3.4582119717054035e-05,
+      "loss": 2.2451,
+      "step": 415000
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 3.456354395767699e-05,
+      "loss": 2.2193,
+      "step": 415500
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 3.454496819829995e-05,
+      "loss": 2.2199,
+      "step": 416000
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 3.4526392438922905e-05,
+      "loss": 2.2209,
+      "step": 416500
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 3.450781667954586e-05,
+      "loss": 2.2412,
+      "step": 417000
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 3.448924092016882e-05,
+      "loss": 2.2233,
+      "step": 417500
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 3.4470665160791776e-05,
+      "loss": 2.2238,
+      "step": 418000
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 3.445208940141473e-05,
+      "loss": 2.2278,
+      "step": 418500
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 3.443351364203769e-05,
+      "loss": 2.2463,
+      "step": 419000
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 3.4414937882660646e-05,
+      "loss": 2.2274,
+      "step": 419500
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 3.43963621232836e-05,
+      "loss": 2.2398,
+      "step": 420000
+    },
+    {
+      "epoch": 6.24,
+      "eval_loss": 2.1873366832733154,
+      "eval_runtime": 1175.9686,
+      "eval_samples_per_second": 406.915,
+      "eval_steps_per_second": 50.864,
+      "step": 420000
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 3.4377786363906553e-05,
+      "loss": 2.2302,
+      "step": 420500
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 3.435921060452951e-05,
+      "loss": 2.2177,
+      "step": 421000
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 3.434063484515247e-05,
+      "loss": 2.2148,
+      "step": 421500
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 3.4322059085775424e-05,
+      "loss": 2.2423,
+      "step": 422000
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 3.430348332639838e-05,
+      "loss": 2.2232,
+      "step": 422500
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 3.428490756702134e-05,
+      "loss": 2.2277,
+      "step": 423000
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 3.4266331807644295e-05,
+      "loss": 2.2159,
+      "step": 423500
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 3.424775604826725e-05,
+      "loss": 2.2154,
+      "step": 424000
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 3.422918028889021e-05,
+      "loss": 2.2239,
+      "step": 424500
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 3.4210604529513165e-05,
+      "loss": 2.2272,
+      "step": 425000
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 3.419202877013612e-05,
+      "loss": 2.2287,
+      "step": 425500
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 3.417345301075908e-05,
+      "loss": 2.2073,
+      "step": 426000
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 3.4154877251382036e-05,
+      "loss": 2.2283,
+      "step": 426500
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 3.4136301492005e-05,
+      "loss": 2.2245,
+      "step": 427000
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 3.4117725732627956e-05,
+      "loss": 2.2176,
+      "step": 427500
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 3.409914997325091e-05,
+      "loss": 2.2198,
+      "step": 428000
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 3.408057421387387e-05,
+      "loss": 2.2204,
+      "step": 428500
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 3.406199845449683e-05,
+      "loss": 2.228,
+      "step": 429000
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 3.404342269511978e-05,
+      "loss": 2.2358,
+      "step": 429500
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 3.4024846935742734e-05,
+      "loss": 2.2181,
+      "step": 430000
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 3.400627117636569e-05,
+      "loss": 2.2349,
+      "step": 430500
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 3.398769541698865e-05,
+      "loss": 2.2303,
+      "step": 431000
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 3.3969119657611605e-05,
+      "loss": 2.2166,
+      "step": 431500
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 3.395054389823456e-05,
+      "loss": 2.2314,
+      "step": 432000
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 3.393196813885752e-05,
+      "loss": 2.223,
+      "step": 432500
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 3.3913392379480475e-05,
+      "loss": 2.2243,
+      "step": 433000
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 3.389481662010343e-05,
+      "loss": 2.2128,
+      "step": 433500
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 3.387624086072639e-05,
+      "loss": 2.2173,
+      "step": 434000
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 3.3857665101349346e-05,
+      "loss": 2.2156,
+      "step": 434500
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 3.38390893419723e-05,
+      "loss": 2.2206,
+      "step": 435000
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 3.382051358259526e-05,
+      "loss": 2.2302,
+      "step": 435500
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 3.3801937823218216e-05,
+      "loss": 2.2321,
+      "step": 436000
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 3.378336206384117e-05,
+      "loss": 2.2234,
+      "step": 436500
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 3.376478630446413e-05,
+      "loss": 2.2253,
+      "step": 437000
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 3.374621054508709e-05,
+      "loss": 2.2317,
+      "step": 437500
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 3.3727634785710044e-05,
+      "loss": 2.2242,
+      "step": 438000
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 3.3709059026333e-05,
+      "loss": 2.2215,
+      "step": 438500
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 3.369048326695595e-05,
+      "loss": 2.2171,
+      "step": 439000
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 3.367190750757891e-05,
+      "loss": 2.2373,
+      "step": 439500
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 3.3653331748201865e-05,
+      "loss": 2.2186,
+      "step": 440000
+    },
+    {
+      "epoch": 6.54,
+      "eval_loss": 2.1790640354156494,
+      "eval_runtime": 1174.122,
+      "eval_samples_per_second": 407.555,
+      "eval_steps_per_second": 50.944,
+      "step": 440000
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 3.363475598882482e-05,
+      "loss": 2.2281,
+      "step": 440500
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 3.361618022944778e-05,
+      "loss": 2.22,
+      "step": 441000
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 3.3597604470070735e-05,
+      "loss": 2.2231,
+      "step": 441500
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 3.357902871069369e-05,
+      "loss": 2.2214,
+      "step": 442000
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 3.356045295131665e-05,
+      "loss": 2.2324,
+      "step": 442500
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 3.3541877191939606e-05,
+      "loss": 2.2181,
+      "step": 443000
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 3.352330143256256e-05,
+      "loss": 2.2236,
+      "step": 443500
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 3.350472567318552e-05,
+      "loss": 2.2106,
+      "step": 444000
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 3.3486149913808476e-05,
+      "loss": 2.2127,
+      "step": 444500
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 3.346757415443143e-05,
+      "loss": 2.2115,
+      "step": 445000
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 3.344899839505439e-05,
+      "loss": 2.2187,
+      "step": 445500
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 3.343042263567735e-05,
+      "loss": 2.2218,
+      "step": 446000
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 3.3411846876300304e-05,
+      "loss": 2.2248,
+      "step": 446500
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 3.339327111692326e-05,
+      "loss": 2.232,
+      "step": 447000
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 3.337469535754622e-05,
+      "loss": 2.212,
+      "step": 447500
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 3.3356119598169175e-05,
+      "loss": 2.2106,
+      "step": 448000
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 3.333754383879213e-05,
+      "loss": 2.2225,
+      "step": 448500
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 3.331896807941509e-05,
+      "loss": 2.2321,
+      "step": 449000
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 3.3300392320038045e-05,
+      "loss": 2.2221,
+      "step": 449500
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 3.3281816560661e-05,
+      "loss": 2.2185,
+      "step": 450000
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 3.326324080128396e-05,
+      "loss": 2.2134,
+      "step": 450500
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 3.3244665041906916e-05,
+      "loss": 2.2039,
+      "step": 451000
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 3.322608928252987e-05,
+      "loss": 2.2098,
+      "step": 451500
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 3.320751352315283e-05,
+      "loss": 2.2225,
+      "step": 452000
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 3.3188937763775786e-05,
+      "loss": 2.2206,
+      "step": 452500
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 3.317036200439874e-05,
+      "loss": 2.2199,
+      "step": 453000
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 3.31517862450217e-05,
+      "loss": 2.2166,
+      "step": 453500
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 3.313321048564466e-05,
+      "loss": 2.2239,
+      "step": 454000
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 3.3114634726267614e-05,
+      "loss": 2.2111,
+      "step": 454500
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 3.309605896689057e-05,
+      "loss": 2.2194,
+      "step": 455000
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 3.307748320751353e-05,
+      "loss": 2.2126,
+      "step": 455500
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 3.3058907448136484e-05,
+      "loss": 2.217,
+      "step": 456000
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 3.304033168875944e-05,
+      "loss": 2.2248,
+      "step": 456500
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 3.30217559293824e-05,
+      "loss": 2.2098,
+      "step": 457000
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 3.300318017000535e-05,
+      "loss": 2.2119,
+      "step": 457500
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 3.2984604410628305e-05,
+      "loss": 2.219,
+      "step": 458000
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 3.296602865125126e-05,
+      "loss": 2.2106,
+      "step": 458500
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 3.294745289187422e-05,
+      "loss": 2.2204,
+      "step": 459000
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 3.2928877132497176e-05,
+      "loss": 2.2149,
+      "step": 459500
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 3.291030137312013e-05,
+      "loss": 2.2104,
+      "step": 460000
+    },
+    {
+      "epoch": 6.84,
+      "eval_loss": 2.169421672821045,
+      "eval_runtime": 1168.4614,
+      "eval_samples_per_second": 409.529,
+      "eval_steps_per_second": 51.191,
+      "step": 460000
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 3.289172561374309e-05,
+      "loss": 2.2138,
+      "step": 460500
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 3.2873149854366046e-05,
+      "loss": 2.2321,
+      "step": 461000
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 3.2854574094989e-05,
+      "loss": 2.2211,
+      "step": 461500
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 3.283599833561196e-05,
+      "loss": 2.2242,
+      "step": 462000
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 3.281742257623492e-05,
+      "loss": 2.2124,
+      "step": 462500
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 3.2798846816857874e-05,
+      "loss": 2.1997,
+      "step": 463000
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 3.278027105748083e-05,
+      "loss": 2.1998,
+      "step": 463500
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 3.276169529810379e-05,
+      "loss": 2.2351,
+      "step": 464000
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 3.2743119538726744e-05,
+      "loss": 2.212,
+      "step": 464500
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 3.27245437793497e-05,
+      "loss": 2.2183,
+      "step": 465000
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 3.270596801997266e-05,
+      "loss": 2.2212,
+      "step": 465500
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 3.2687392260595615e-05,
+      "loss": 2.213,
+      "step": 466000
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 3.266881650121857e-05,
+      "loss": 2.2161,
+      "step": 466500
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 3.265024074184153e-05,
+      "loss": 2.2158,
+      "step": 467000
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 3.2631664982464486e-05,
+      "loss": 2.2063,
+      "step": 467500
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 3.261308922308744e-05,
+      "loss": 2.2006,
+      "step": 468000
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 3.25945134637104e-05,
+      "loss": 2.2047,
+      "step": 468500
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 3.2575937704333356e-05,
+      "loss": 2.21,
+      "step": 469000
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 3.255736194495631e-05,
+      "loss": 2.2152,
+      "step": 469500
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 3.253878618557927e-05,
+      "loss": 2.2111,
+      "step": 470000
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 3.252021042620223e-05,
+      "loss": 2.212,
+      "step": 470500
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 3.2501634666825184e-05,
+      "loss": 2.2223,
+      "step": 471000
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 3.248305890744814e-05,
+      "loss": 2.2091,
+      "step": 471500
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 3.24644831480711e-05,
+      "loss": 2.1951,
+      "step": 472000
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 3.2445907388694054e-05,
+      "loss": 2.2172,
+      "step": 472500
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 3.242733162931701e-05,
+      "loss": 2.2028,
+      "step": 473000
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 3.240875586993997e-05,
+      "loss": 2.2088,
+      "step": 473500
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 3.2390180110562925e-05,
+      "loss": 2.1987,
+      "step": 474000
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 3.237160435118588e-05,
+      "loss": 2.2137,
+      "step": 474500
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 3.235302859180884e-05,
+      "loss": 2.192,
+      "step": 475000
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 3.2334452832431796e-05,
+      "loss": 2.2041,
+      "step": 475500
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 3.2315877073054746e-05,
+      "loss": 2.2021,
+      "step": 476000
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 3.22973013136777e-05,
+      "loss": 2.2162,
+      "step": 476500
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 3.227872555430066e-05,
+      "loss": 2.2022,
+      "step": 477000
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 3.2260149794923616e-05,
+      "loss": 2.2008,
+      "step": 477500
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 3.224157403554657e-05,
+      "loss": 2.19,
+      "step": 478000
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 3.222299827616953e-05,
+      "loss": 2.191,
+      "step": 478500
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 3.220442251679249e-05,
+      "loss": 2.2063,
+      "step": 479000
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 3.2185846757415444e-05,
+      "loss": 2.2085,
+      "step": 479500
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 3.21672709980384e-05,
+      "loss": 2.2138,
+      "step": 480000
+    },
+    {
+      "epoch": 7.13,
+      "eval_loss": 2.1657702922821045,
+      "eval_runtime": 1171.0463,
+      "eval_samples_per_second": 408.625,
+      "eval_steps_per_second": 51.078,
+      "step": 480000
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 3.214869523866136e-05,
+      "loss": 2.1954,
+      "step": 480500
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 3.2130119479284314e-05,
+      "loss": 2.2031,
+      "step": 481000
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 3.211154371990727e-05,
+      "loss": 2.2035,
+      "step": 481500
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 3.209296796053023e-05,
+      "loss": 2.1875,
+      "step": 482000
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 3.2074392201153185e-05,
+      "loss": 2.1987,
+      "step": 482500
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 3.205581644177614e-05,
+      "loss": 2.2026,
+      "step": 483000
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 3.20372406823991e-05,
+      "loss": 2.2065,
+      "step": 483500
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 3.2018664923022056e-05,
+      "loss": 2.2025,
+      "step": 484000
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 3.200008916364501e-05,
+      "loss": 2.1838,
+      "step": 484500
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 3.198151340426796e-05,
+      "loss": 2.208,
+      "step": 485000
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 3.196293764489092e-05,
+      "loss": 2.196,
+      "step": 485500
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 3.1944361885513876e-05,
+      "loss": 2.2111,
+      "step": 486000
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 3.192578612613683e-05,
+      "loss": 2.1992,
+      "step": 486500
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 3.190721036675979e-05,
+      "loss": 2.1927,
+      "step": 487000
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 3.188863460738275e-05,
+      "loss": 2.2105,
+      "step": 487500
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 3.187005884800571e-05,
+      "loss": 2.1957,
+      "step": 488000
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 3.185148308862867e-05,
+      "loss": 2.2054,
+      "step": 488500
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 3.1832907329251624e-05,
+      "loss": 2.1903,
+      "step": 489000
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 3.181433156987458e-05,
+      "loss": 2.2005,
+      "step": 489500
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 3.179575581049754e-05,
+      "loss": 2.2015,
+      "step": 490000
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 3.1777180051120495e-05,
+      "loss": 2.1928,
+      "step": 490500
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 3.175860429174345e-05,
+      "loss": 2.2067,
+      "step": 491000
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 3.174002853236641e-05,
+      "loss": 2.1882,
+      "step": 491500
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 3.1721452772989365e-05,
+      "loss": 2.1994,
+      "step": 492000
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 3.170287701361232e-05,
+      "loss": 2.2026,
+      "step": 492500
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 3.168430125423528e-05,
+      "loss": 2.1945,
+      "step": 493000
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 3.1665725494858236e-05,
+      "loss": 2.2011,
+      "step": 493500
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 3.1647149735481186e-05,
+      "loss": 2.1971,
+      "step": 494000
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 3.162857397610414e-05,
+      "loss": 2.1957,
+      "step": 494500
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 3.16099982167271e-05,
+      "loss": 2.1941,
+      "step": 495000
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 3.159142245735006e-05,
+      "loss": 2.1832,
+      "step": 495500
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 3.1572846697973014e-05,
+      "loss": 2.2005,
+      "step": 496000
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 3.155427093859597e-05,
+      "loss": 2.1913,
+      "step": 496500
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 3.153569517921893e-05,
+      "loss": 2.2032,
+      "step": 497000
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 3.1517119419841884e-05,
+      "loss": 2.2084,
+      "step": 497500
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 3.149854366046484e-05,
+      "loss": 2.2094,
+      "step": 498000
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 3.14799679010878e-05,
+      "loss": 2.177,
+      "step": 498500
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 3.1461392141710755e-05,
+      "loss": 2.1894,
+      "step": 499000
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 3.144281638233371e-05,
+      "loss": 2.1928,
+      "step": 499500
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 3.142424062295667e-05,
+      "loss": 2.1995,
+      "step": 500000
+    },
+    {
+      "epoch": 7.43,
+      "eval_loss": 2.1569273471832275,
+      "eval_runtime": 1174.6025,
+      "eval_samples_per_second": 407.388,
+      "eval_steps_per_second": 50.924,
+      "step": 500000
+    }
+  ],
+  "max_steps": 1345840,
+  "num_train_epochs": 20,
+  "total_flos": 1.0609639946214728e+18,
+  "trial_name": null,
+  "trial_params": null
+}