diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7803 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 16.0,
+  "global_step": 637840,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9984322087043773e-05,
+      "loss": 2.8529,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.996864417408755e-05,
+      "loss": 2.3821,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.995296626113132e-05,
+      "loss": 2.282,
+      "step": 1500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9937288348175092e-05,
+      "loss": 2.1631,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9921610435218864e-05,
+      "loss": 2.093,
+      "step": 2500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.990593252226264e-05,
+      "loss": 2.09,
+      "step": 3000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.989025460930641e-05,
+      "loss": 2.0326,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9874576696350182e-05,
+      "loss": 2.0064,
+      "step": 4000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9858898783393958e-05,
+      "loss": 2.0513,
+      "step": 4500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.984322087043773e-05,
+      "loss": 1.955,
+      "step": 5000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.98275429574815e-05,
+      "loss": 1.9279,
+      "step": 5500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9811865044525273e-05,
+      "loss": 1.8879,
+      "step": 6000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9796187131569048e-05,
+      "loss": 1.9094,
+      "step": 6500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.978050921861282e-05,
+      "loss": 1.8483,
+      "step": 7000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.976483130565659e-05,
+      "loss": 1.7955,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9749153392700363e-05,
+      "loss": 1.7358,
+      "step": 8000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.973347547974414e-05,
+      "loss": 1.7735,
+      "step": 8500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.971779756678791e-05,
+      "loss": 1.7997,
+      "step": 9000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9702119653831685e-05,
+      "loss": 1.6992,
+      "step": 9500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9686441740875457e-05,
+      "loss": 1.7818,
+      "step": 10000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.967076382791923e-05,
+      "loss": 1.7817,
+      "step": 10500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9655085914963e-05,
+      "loss": 1.7968,
+      "step": 11000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9639408002006776e-05,
+      "loss": 1.7211,
+      "step": 11500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9623730089050548e-05,
+      "loss": 1.7302,
+      "step": 12000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.960805217609432e-05,
+      "loss": 1.8177,
+      "step": 12500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.959237426313809e-05,
+      "loss": 1.6991,
+      "step": 13000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9576696350181866e-05,
+      "loss": 1.7491,
+      "step": 13500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9561018437225638e-05,
+      "loss": 1.7191,
+      "step": 14000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9545340524269413e-05,
+      "loss": 1.7411,
+      "step": 14500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.952966261131318e-05,
+      "loss": 1.6523,
+      "step": 15000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9513984698356957e-05,
+      "loss": 1.6879,
+      "step": 15500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.949830678540073e-05,
+      "loss": 1.6819,
+      "step": 16000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9482628872444504e-05,
+      "loss": 1.7571,
+      "step": 16500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9466950959488275e-05,
+      "loss": 1.6614,
+      "step": 17000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9451273046532047e-05,
+      "loss": 1.6589,
+      "step": 17500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.943559513357582e-05,
+      "loss": 1.6081,
+      "step": 18000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9419917220619594e-05,
+      "loss": 1.5725,
+      "step": 18500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9404239307663366e-05,
+      "loss": 1.6776,
+      "step": 19000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9388561394707138e-05,
+      "loss": 1.6364,
+      "step": 19500
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.937288348175091e-05,
+      "loss": 1.6534,
+      "step": 20000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9357205568794684e-05,
+      "loss": 1.5984,
+      "step": 20500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.9341527655838456e-05,
+      "loss": 1.6573,
+      "step": 21000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.932584974288223e-05,
+      "loss": 1.734,
+      "step": 21500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.9310171829926e-05,
+      "loss": 1.603,
+      "step": 22000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.9294493916969775e-05,
+      "loss": 1.6416,
+      "step": 22500
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.9278816004013547e-05,
+      "loss": 1.5898,
+      "step": 23000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.9263138091057322e-05,
+      "loss": 1.6236,
+      "step": 23500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.924746017810109e-05,
+      "loss": 1.5564,
+      "step": 24000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.9231782265144865e-05,
+      "loss": 1.5903,
+      "step": 24500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.9216104352188637e-05,
+      "loss": 1.6229,
+      "step": 25000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.9200426439232412e-05,
+      "loss": 1.5936,
+      "step": 25500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.9184748526276184e-05,
+      "loss": 1.6015,
+      "step": 26000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.9169070613319956e-05,
+      "loss": 1.5906,
+      "step": 26500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.9153392700363727e-05,
+      "loss": 1.6663,
+      "step": 27000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.9137714787407503e-05,
+      "loss": 1.5678,
+      "step": 27500
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.9122036874451274e-05,
+      "loss": 1.621,
+      "step": 28000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.910635896149505e-05,
+      "loss": 1.5988,
+      "step": 28500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.9090681048538818e-05,
+      "loss": 1.6418,
+      "step": 29000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.9075003135582593e-05,
+      "loss": 1.5442,
+      "step": 29500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.9059325222626365e-05,
+      "loss": 1.6137,
+      "step": 30000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.904364730967014e-05,
+      "loss": 1.5937,
+      "step": 30500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.902796939671391e-05,
+      "loss": 1.551,
+      "step": 31000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.9012291483757683e-05,
+      "loss": 1.5752,
+      "step": 31500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.8996613570801455e-05,
+      "loss": 1.6542,
+      "step": 32000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.898093565784523e-05,
+      "loss": 1.5128,
+      "step": 32500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.8965257744889002e-05,
+      "loss": 1.5094,
+      "step": 33000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.8949579831932774e-05,
+      "loss": 1.5501,
+      "step": 33500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.8933901918976546e-05,
+      "loss": 1.5973,
+      "step": 34000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.891822400602032e-05,
+      "loss": 1.534,
+      "step": 34500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.8902546093064093e-05,
+      "loss": 1.5083,
+      "step": 35000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.8886868180107864e-05,
+      "loss": 1.612,
+      "step": 35500
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.887119026715164e-05,
+      "loss": 1.5596,
+      "step": 36000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.885551235419541e-05,
+      "loss": 1.5497,
+      "step": 36500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.8839834441239183e-05,
+      "loss": 1.6015,
+      "step": 37000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8824156528282958e-05,
+      "loss": 1.5206,
+      "step": 37500
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.880847861532673e-05,
+      "loss": 1.5523,
+      "step": 38000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.87928007023705e-05,
+      "loss": 1.5119,
+      "step": 38500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.8777122789414273e-05,
+      "loss": 1.5777,
+      "step": 39000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.876144487645805e-05,
+      "loss": 1.5688,
+      "step": 39500
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.4448622465133667,
+      "eval_runtime": 128.1797,
+      "eval_samples_per_second": 276.456,
+      "eval_steps_per_second": 34.561,
+      "step": 39865
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.874576696350182e-05,
+      "loss": 1.522,
+      "step": 40000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8730089050545592e-05,
+      "loss": 1.5006,
+      "step": 40500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.8714411137589367e-05,
+      "loss": 1.5478,
+      "step": 41000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.869873322463314e-05,
+      "loss": 1.4551,
+      "step": 41500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.868305531167691e-05,
+      "loss": 1.5168,
+      "step": 42000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.8667377398720682e-05,
+      "loss": 1.4736,
+      "step": 42500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.8651699485764458e-05,
+      "loss": 1.4973,
+      "step": 43000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.863602157280823e-05,
+      "loss": 1.4207,
+      "step": 43500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.8620343659852e-05,
+      "loss": 1.4869,
+      "step": 44000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.8604665746895773e-05,
+      "loss": 1.4581,
+      "step": 44500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.8588987833939548e-05,
+      "loss": 1.4674,
+      "step": 45000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.857330992098332e-05,
+      "loss": 1.4434,
+      "step": 45500
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.8557632008027095e-05,
+      "loss": 1.4838,
+      "step": 46000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.8541954095070867e-05,
+      "loss": 1.4814,
+      "step": 46500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.852627618211464e-05,
+      "loss": 1.4879,
+      "step": 47000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.851059826915841e-05,
+      "loss": 1.4201,
+      "step": 47500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.8494920356202185e-05,
+      "loss": 1.5135,
+      "step": 48000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.8479242443245957e-05,
+      "loss": 1.4917,
+      "step": 48500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.846356453028973e-05,
+      "loss": 1.4462,
+      "step": 49000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.84478866173335e-05,
+      "loss": 1.456,
+      "step": 49500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.8432208704377276e-05,
+      "loss": 1.4534,
+      "step": 50000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.8416530791421048e-05,
+      "loss": 1.4424,
+      "step": 50500
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.8400852878464823e-05,
+      "loss": 1.4568,
+      "step": 51000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.838517496550859e-05,
+      "loss": 1.4868,
+      "step": 51500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.8369497052552366e-05,
+      "loss": 1.5445,
+      "step": 52000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.8353819139596138e-05,
+      "loss": 1.4009,
+      "step": 52500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.8338141226639913e-05,
+      "loss": 1.4949,
+      "step": 53000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.8322463313683685e-05,
+      "loss": 1.4977,
+      "step": 53500
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.8306785400727457e-05,
+      "loss": 1.4251,
+      "step": 54000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.829110748777123e-05,
+      "loss": 1.4671,
+      "step": 54500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.8275429574815004e-05,
+      "loss": 1.4663,
+      "step": 55000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.8259751661858775e-05,
+      "loss": 1.474,
+      "step": 55500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.8244073748902547e-05,
+      "loss": 1.4399,
+      "step": 56000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.822839583594632e-05,
+      "loss": 1.4721,
+      "step": 56500
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.8212717922990094e-05,
+      "loss": 1.4915,
+      "step": 57000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.8197040010033866e-05,
+      "loss": 1.4602,
+      "step": 57500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.818136209707764e-05,
+      "loss": 1.4274,
+      "step": 58000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.816568418412141e-05,
+      "loss": 1.4175,
+      "step": 58500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.8150006271165184e-05,
+      "loss": 1.4513,
+      "step": 59000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.8134328358208956e-05,
+      "loss": 1.4499,
+      "step": 59500
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.811865044525273e-05,
+      "loss": 1.4535,
+      "step": 60000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.81029725322965e-05,
+      "loss": 1.4235,
+      "step": 60500
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.8087294619340275e-05,
+      "loss": 1.4936,
+      "step": 61000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.8071616706384047e-05,
+      "loss": 1.4303,
+      "step": 61500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.8055938793427822e-05,
+      "loss": 1.4397,
+      "step": 62000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.8040260880471593e-05,
+      "loss": 1.4463,
+      "step": 62500
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.8024582967515365e-05,
+      "loss": 1.4258,
+      "step": 63000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.8008905054559137e-05,
+      "loss": 1.462,
+      "step": 63500
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.7993227141602912e-05,
+      "loss": 1.4396,
+      "step": 64000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.7977549228646684e-05,
+      "loss": 1.4277,
+      "step": 64500
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.796187131569046e-05,
+      "loss": 1.4342,
+      "step": 65000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.7946193402734227e-05,
+      "loss": 1.4458,
+      "step": 65500
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.7930515489778003e-05,
+      "loss": 1.4409,
+      "step": 66000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.7914837576821774e-05,
+      "loss": 1.4699,
+      "step": 66500
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.789915966386555e-05,
+      "loss": 1.4831,
+      "step": 67000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.7883481750909318e-05,
+      "loss": 1.3923,
+      "step": 67500
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.7867803837953093e-05,
+      "loss": 1.4465,
+      "step": 68000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.7852125924996865e-05,
+      "loss": 1.4096,
+      "step": 68500
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.783644801204064e-05,
+      "loss": 1.4535,
+      "step": 69000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.782077009908441e-05,
+      "loss": 1.4147,
+      "step": 69500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.7805092186128183e-05,
+      "loss": 1.3796,
+      "step": 70000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.7789414273171955e-05,
+      "loss": 1.3972,
+      "step": 70500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.777373636021573e-05,
+      "loss": 1.4675,
+      "step": 71000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.7758058447259502e-05,
+      "loss": 1.46,
+      "step": 71500
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.7742380534303274e-05,
+      "loss": 1.4559,
+      "step": 72000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.7726702621347046e-05,
+      "loss": 1.4631,
+      "step": 72500
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.771102470839082e-05,
+      "loss": 1.408,
+      "step": 73000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.7695346795434593e-05,
+      "loss": 1.4728,
+      "step": 73500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.7679668882478368e-05,
+      "loss": 1.4401,
+      "step": 74000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.766399096952214e-05,
+      "loss": 1.4551,
+      "step": 74500
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.764831305656591e-05,
+      "loss": 1.4492,
+      "step": 75000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.7632635143609683e-05,
+      "loss": 1.4464,
+      "step": 75500
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.7616957230653458e-05,
+      "loss": 1.4745,
+      "step": 76000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.760127931769723e-05,
+      "loss": 1.411,
+      "step": 76500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.7585601404741e-05,
+      "loss": 1.3978,
+      "step": 77000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.7569923491784773e-05,
+      "loss": 1.4891,
+      "step": 77500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.755424557882855e-05,
+      "loss": 1.4735,
+      "step": 78000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.753856766587232e-05,
+      "loss": 1.487,
+      "step": 78500
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.7522889752916092e-05,
+      "loss": 1.4213,
+      "step": 79000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.7507211839959867e-05,
+      "loss": 1.4332,
+      "step": 79500
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.3822311162948608,
+      "eval_runtime": 127.8424,
+      "eval_samples_per_second": 277.185,
+      "eval_steps_per_second": 34.652,
+      "step": 79730
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.749153392700364e-05,
+      "loss": 1.3539,
+      "step": 80000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.747585601404741e-05,
+      "loss": 1.3973,
+      "step": 80500
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.7460178101091186e-05,
+      "loss": 1.3728,
+      "step": 81000
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.7444500188134958e-05,
+      "loss": 1.3788,
+      "step": 81500
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.742882227517873e-05,
+      "loss": 1.4604,
+      "step": 82000
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.74131443622225e-05,
+      "loss": 1.3327,
+      "step": 82500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.7397466449266276e-05,
+      "loss": 1.3296,
+      "step": 83000
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.7381788536310048e-05,
+      "loss": 1.3417,
+      "step": 83500
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.736611062335382e-05,
+      "loss": 1.4489,
+      "step": 84000
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.7350432710397595e-05,
+      "loss": 1.3476,
+      "step": 84500
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.7334754797441367e-05,
+      "loss": 1.422,
+      "step": 85000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.731907688448514e-05,
+      "loss": 1.3592,
+      "step": 85500
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.730339897152891e-05,
+      "loss": 1.41,
+      "step": 86000
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.7287721058572685e-05,
+      "loss": 1.3894,
+      "step": 86500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.7272043145616457e-05,
+      "loss": 1.3512,
+      "step": 87000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.725636523266023e-05,
+      "loss": 1.3567,
+      "step": 87500
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.7240687319704e-05,
+      "loss": 1.4404,
+      "step": 88000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.7225009406747776e-05,
+      "loss": 1.3535,
+      "step": 88500
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.7209331493791548e-05,
+      "loss": 1.4144,
+      "step": 89000
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.7193653580835323e-05,
+      "loss": 1.432,
+      "step": 89500
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.7177975667879094e-05,
+      "loss": 1.3745,
+      "step": 90000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.7162297754922866e-05,
+      "loss": 1.375,
+      "step": 90500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.7146619841966638e-05,
+      "loss": 1.3853,
+      "step": 91000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.7130941929010413e-05,
+      "loss": 1.3474,
+      "step": 91500
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.7115264016054185e-05,
+      "loss": 1.3424,
+      "step": 92000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.7099586103097957e-05,
+      "loss": 1.3599,
+      "step": 92500
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.708390819014173e-05,
+      "loss": 1.4519,
+      "step": 93000
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.7068230277185504e-05,
+      "loss": 1.4344,
+      "step": 93500
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.7052552364229275e-05,
+      "loss": 1.3346,
+      "step": 94000
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.703687445127305e-05,
+      "loss": 1.3485,
+      "step": 94500
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.702119653831682e-05,
+      "loss": 1.3506,
+      "step": 95000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.7005518625360594e-05,
+      "loss": 1.3044,
+      "step": 95500
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.6989840712404366e-05,
+      "loss": 1.3287,
+      "step": 96000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.697416279944814e-05,
+      "loss": 1.362,
+      "step": 96500
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.695848488649191e-05,
+      "loss": 1.378,
+      "step": 97000
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6942806973535684e-05,
+      "loss": 1.4017,
+      "step": 97500
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6927129060579456e-05,
+      "loss": 1.3816,
+      "step": 98000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.691145114762323e-05,
+      "loss": 1.3901,
+      "step": 98500
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.6895773234667003e-05,
+      "loss": 1.3633,
+      "step": 99000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.6880095321710775e-05,
+      "loss": 1.3504,
+      "step": 99500
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.6864417408754547e-05,
+      "loss": 1.336,
+      "step": 100000
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.6848739495798322e-05,
+      "loss": 1.4251,
+      "step": 100500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.6833061582842093e-05,
+      "loss": 1.3247,
+      "step": 101000
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.681738366988587e-05,
+      "loss": 1.381,
+      "step": 101500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.6801705756929637e-05,
+      "loss": 1.3824,
+      "step": 102000
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.6786027843973412e-05,
+      "loss": 1.4192,
+      "step": 102500
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.6770349931017184e-05,
+      "loss": 1.355,
+      "step": 103000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.675467201806096e-05,
+      "loss": 1.3567,
+      "step": 103500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.6738994105104727e-05,
+      "loss": 1.3636,
+      "step": 104000
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.6723316192148503e-05,
+      "loss": 1.3647,
+      "step": 104500
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.6707638279192274e-05,
+      "loss": 1.3875,
+      "step": 105000
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.669196036623605e-05,
+      "loss": 1.3918,
+      "step": 105500
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.667628245327982e-05,
+      "loss": 1.3435,
+      "step": 106000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.6660604540323593e-05,
+      "loss": 1.3854,
+      "step": 106500
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.6644926627367365e-05,
+      "loss": 1.3943,
+      "step": 107000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.662924871441114e-05,
+      "loss": 1.375,
+      "step": 107500
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.661357080145491e-05,
+      "loss": 1.3497,
+      "step": 108000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.6597892888498683e-05,
+      "loss": 1.3838,
+      "step": 108500
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.6582214975542455e-05,
+      "loss": 1.3096,
+      "step": 109000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.656653706258623e-05,
+      "loss": 1.3499,
+      "step": 109500
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.6550859149630002e-05,
+      "loss": 1.3998,
+      "step": 110000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.6535181236673777e-05,
+      "loss": 1.3906,
+      "step": 110500
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.6519503323717546e-05,
+      "loss": 1.2859,
+      "step": 111000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.650382541076132e-05,
+      "loss": 1.3169,
+      "step": 111500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.6488147497805093e-05,
+      "loss": 1.3822,
+      "step": 112000
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.6472469584848868e-05,
+      "loss": 1.4047,
+      "step": 112500
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.645679167189264e-05,
+      "loss": 1.3801,
+      "step": 113000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.644111375893641e-05,
+      "loss": 1.3764,
+      "step": 113500
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.6425435845980183e-05,
+      "loss": 1.33,
+      "step": 114000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.6409757933023958e-05,
+      "loss": 1.3235,
+      "step": 114500
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.639408002006773e-05,
+      "loss": 1.3344,
+      "step": 115000
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.63784021071115e-05,
+      "loss": 1.3674,
+      "step": 115500
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.6362724194155273e-05,
+      "loss": 1.3511,
+      "step": 116000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.634704628119905e-05,
+      "loss": 1.3671,
+      "step": 116500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.633136836824282e-05,
+      "loss": 1.3888,
+      "step": 117000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.6315690455286595e-05,
+      "loss": 1.3589,
+      "step": 117500
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.6300012542330367e-05,
+      "loss": 1.3533,
+      "step": 118000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.628433462937414e-05,
+      "loss": 1.2987,
+      "step": 118500
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.626865671641791e-05,
+      "loss": 1.3448,
+      "step": 119000
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.6252978803461686e-05,
+      "loss": 1.277,
+      "step": 119500
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.338686227798462,
+      "eval_runtime": 128.3357,
+      "eval_samples_per_second": 276.12,
+      "eval_steps_per_second": 34.519,
+      "step": 119595
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.6237300890505458e-05,
+      "loss": 1.3447,
+      "step": 120000
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.622162297754923e-05,
+      "loss": 1.2503,
+      "step": 120500
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.6205945064593e-05,
+      "loss": 1.2342,
+      "step": 121000
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.6190267151636776e-05,
+      "loss": 1.3394,
+      "step": 121500
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.6174589238680548e-05,
+      "loss": 1.2895,
+      "step": 122000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.615891132572432e-05,
+      "loss": 1.4003,
+      "step": 122500
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.6143233412768095e-05,
+      "loss": 1.3084,
+      "step": 123000
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.6127555499811867e-05,
+      "loss": 1.2751,
+      "step": 123500
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.611187758685564e-05,
+      "loss": 1.3414,
+      "step": 124000
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.609619967389941e-05,
+      "loss": 1.3392,
+      "step": 124500
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.6080521760943185e-05,
+      "loss": 1.3551,
+      "step": 125000
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.6064843847986957e-05,
+      "loss": 1.2876,
+      "step": 125500
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.604916593503073e-05,
+      "loss": 1.3126,
+      "step": 126000
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.6033488022074504e-05,
+      "loss": 1.2531,
+      "step": 126500
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.6017810109118276e-05,
+      "loss": 1.3762,
+      "step": 127000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.6002132196162048e-05,
+      "loss": 1.3384,
+      "step": 127500
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.5986454283205823e-05,
+      "loss": 1.3136,
+      "step": 128000
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.5970776370249594e-05,
+      "loss": 1.3164,
+      "step": 128500
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.5955098457293366e-05,
+      "loss": 1.2976,
+      "step": 129000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.5939420544337138e-05,
+      "loss": 1.3044,
+      "step": 129500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.5923742631380913e-05,
+      "loss": 1.2693,
+      "step": 130000
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.5908064718424685e-05,
+      "loss": 1.2781,
+      "step": 130500
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.5892386805468457e-05,
+      "loss": 1.2856,
+      "step": 131000
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.587670889251223e-05,
+      "loss": 1.2876,
+      "step": 131500
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.5861030979556004e-05,
+      "loss": 1.3101,
+      "step": 132000
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.5845353066599775e-05,
+      "loss": 1.3436,
+      "step": 132500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.582967515364355e-05,
+      "loss": 1.295,
+      "step": 133000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.581399724068732e-05,
+      "loss": 1.3083,
+      "step": 133500
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.5798319327731094e-05,
+      "loss": 1.3935,
+      "step": 134000
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.5782641414774866e-05,
+      "loss": 1.3764,
+      "step": 134500
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.576696350181864e-05,
+      "loss": 1.2953,
+      "step": 135000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.5751285588862413e-05,
+      "loss": 1.3373,
+      "step": 135500
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.5735607675906184e-05,
+      "loss": 1.367,
+      "step": 136000
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.5719929762949956e-05,
+      "loss": 1.3062,
+      "step": 136500
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.570425184999373e-05,
+      "loss": 1.276,
+      "step": 137000
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.5688573937037503e-05,
+      "loss": 1.339,
+      "step": 137500
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.5672896024081278e-05,
+      "loss": 1.3375,
+      "step": 138000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 1.5657218111125047e-05,
+      "loss": 1.3677,
+      "step": 138500
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 1.5641540198168822e-05,
+      "loss": 1.2983,
+      "step": 139000
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 1.5625862285212593e-05,
+      "loss": 1.3559,
+      "step": 139500
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 1.561018437225637e-05,
+      "loss": 1.3124,
+      "step": 140000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 1.5594506459300137e-05,
+      "loss": 1.373,
+      "step": 140500
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 1.5578828546343912e-05,
+      "loss": 1.2698,
+      "step": 141000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 1.5563150633387684e-05,
+      "loss": 1.2757,
+      "step": 141500
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.554747272043146e-05,
+      "loss": 1.3387,
+      "step": 142000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 1.553179480747523e-05,
+      "loss": 1.3115,
+      "step": 142500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 1.5516116894519003e-05,
+      "loss": 1.3385,
+      "step": 143000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 1.5500438981562774e-05,
+      "loss": 1.3294,
+      "step": 143500
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 1.548476106860655e-05,
+      "loss": 1.3153,
+      "step": 144000
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 1.546908315565032e-05,
+      "loss": 1.2989,
+      "step": 144500
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 1.5453405242694093e-05,
+      "loss": 1.2992,
+      "step": 145000
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 1.5437727329737865e-05,
+      "loss": 1.3188,
+      "step": 145500
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 1.542204941678164e-05,
+      "loss": 1.3685,
+      "step": 146000
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 1.540637150382541e-05,
+      "loss": 1.3352,
+      "step": 146500
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 1.5390693590869187e-05,
+      "loss": 1.3314,
+      "step": 147000
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 1.5375015677912955e-05,
+      "loss": 1.3009,
+      "step": 147500
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 1.535933776495673e-05,
+      "loss": 1.2473,
+      "step": 148000
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 1.5343659852000502e-05,
+      "loss": 1.3388,
+      "step": 148500
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 1.5327981939044277e-05,
+      "loss": 1.3062,
+      "step": 149000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 1.531230402608805e-05,
+      "loss": 1.3164,
+      "step": 149500
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 1.529662611313182e-05,
+      "loss": 1.2968,
+      "step": 150000
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 1.5280948200175593e-05,
+      "loss": 1.2951,
+      "step": 150500
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 1.5265270287219368e-05,
+      "loss": 1.3349,
+      "step": 151000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 1.524959237426314e-05,
+      "loss": 1.3453,
+      "step": 151500
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 1.5233914461306911e-05,
+      "loss": 1.3503,
+      "step": 152000
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 1.5218236548350685e-05,
+      "loss": 1.2881,
+      "step": 152500
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 1.5202558635394458e-05,
+      "loss": 1.265,
+      "step": 153000
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 1.5186880722438232e-05,
+      "loss": 1.2804,
+      "step": 153500
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 1.5171202809482003e-05,
+      "loss": 1.3297,
+      "step": 154000
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 1.5155524896525775e-05,
+      "loss": 1.2666,
+      "step": 154500
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 1.5139846983569548e-05,
+      "loss": 1.2965,
+      "step": 155000
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 1.5124169070613322e-05,
+      "loss": 1.3152,
+      "step": 155500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 1.5108491157657095e-05,
+      "loss": 1.3266,
+      "step": 156000
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 1.5092813244700865e-05,
+      "loss": 1.3224,
+      "step": 156500
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 1.5077135331744639e-05,
+      "loss": 1.3181,
+      "step": 157000
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.5061457418788412e-05,
+      "loss": 1.2806,
+      "step": 157500
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.5045779505832186e-05,
+      "loss": 1.3513,
+      "step": 158000
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.503010159287596e-05,
+      "loss": 1.3295,
+      "step": 158500
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.501442367991973e-05,
+      "loss": 1.3483,
+      "step": 159000
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.283326506614685,
+      "eval_runtime": 128.3046,
+      "eval_samples_per_second": 276.187,
+      "eval_steps_per_second": 34.527,
+      "step": 159460
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.4998745766963503e-05,
+      "loss": 1.2967,
+      "step": 159500
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.4983067854007276e-05,
+      "loss": 1.3119,
+      "step": 160000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1.496738994105105e-05,
+      "loss": 1.2953,
+      "step": 160500
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1.495171202809482e-05,
+      "loss": 1.2804,
+      "step": 161000
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 1.4936034115138593e-05,
+      "loss": 1.2275,
+      "step": 161500
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1.4920356202182367e-05,
+      "loss": 1.2728,
+      "step": 162000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1.490467828922614e-05,
+      "loss": 1.2205,
+      "step": 162500
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1.4889000376269914e-05,
+      "loss": 1.2487,
+      "step": 163000
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.4873322463313684e-05,
+      "loss": 1.2796,
+      "step": 163500
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1.4857644550357457e-05,
+      "loss": 1.2288,
+      "step": 164000
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.484196663740123e-05,
+      "loss": 1.3309,
+      "step": 164500
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.4826288724445004e-05,
+      "loss": 1.2506,
+      "step": 165000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.4810610811488774e-05,
+      "loss": 1.2754,
+      "step": 165500
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.4794932898532548e-05,
+      "loss": 1.2578,
+      "step": 166000
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.4779254985576321e-05,
+      "loss": 1.2982,
+      "step": 166500
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.4763577072620094e-05,
+      "loss": 1.2595,
+      "step": 167000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.4747899159663868e-05,
+      "loss": 1.2515,
+      "step": 167500
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.4732221246707638e-05,
+      "loss": 1.2115,
+      "step": 168000
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1.4716543333751411e-05,
+      "loss": 1.2686,
+      "step": 168500
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.4700865420795185e-05,
+      "loss": 1.3288,
+      "step": 169000
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.4685187507838958e-05,
+      "loss": 1.2665,
+      "step": 169500
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1.4669509594882732e-05,
+      "loss": 1.2856,
+      "step": 170000
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.4653831681926502e-05,
+      "loss": 1.216,
+      "step": 170500
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.4638153768970275e-05,
+      "loss": 1.2685,
+      "step": 171000
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.4622475856014049e-05,
+      "loss": 1.2446,
+      "step": 171500
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.4606797943057822e-05,
+      "loss": 1.2396,
+      "step": 172000
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.4591120030101594e-05,
+      "loss": 1.3141,
+      "step": 172500
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.4575442117145366e-05,
+      "loss": 1.2931,
+      "step": 173000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.455976420418914e-05,
+      "loss": 1.2463,
+      "step": 173500
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1.4544086291232913e-05,
+      "loss": 1.2987,
+      "step": 174000
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1.4528408378276686e-05,
+      "loss": 1.2794,
+      "step": 174500
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.4512730465320458e-05,
+      "loss": 1.3229,
+      "step": 175000
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1.449705255236423e-05,
+      "loss": 1.3247,
+      "step": 175500
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1.4481374639408003e-05,
+      "loss": 1.2707,
+      "step": 176000
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 1.4465696726451776e-05,
+      "loss": 1.3131,
+      "step": 176500
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1.4450018813495548e-05,
+      "loss": 1.2587,
+      "step": 177000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1.4434340900539322e-05,
+      "loss": 1.2548,
+      "step": 177500
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1.4418662987583093e-05,
+      "loss": 1.2543,
+      "step": 178000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1.4402985074626867e-05,
+      "loss": 1.2267,
+      "step": 178500
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1.438730716167064e-05,
+      "loss": 1.2485,
+      "step": 179000
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.4371629248714412e-05,
+      "loss": 1.2508,
+      "step": 179500
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1.4355951335758186e-05,
+      "loss": 1.2642,
+      "step": 180000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.4340273422801957e-05,
+      "loss": 1.2739,
+      "step": 180500
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 1.432459550984573e-05,
+      "loss": 1.2474,
+      "step": 181000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 1.4308917596889503e-05,
+      "loss": 1.3511,
+      "step": 181500
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.4293239683933276e-05,
+      "loss": 1.2219,
+      "step": 182000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1.427756177097705e-05,
+      "loss": 1.2916,
+      "step": 182500
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.4261883858020821e-05,
+      "loss": 1.2512,
+      "step": 183000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.4246205945064595e-05,
+      "loss": 1.2476,
+      "step": 183500
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.4230528032108366e-05,
+      "loss": 1.2676,
+      "step": 184000
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.421485011915214e-05,
+      "loss": 1.2402,
+      "step": 184500
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.4199172206195913e-05,
+      "loss": 1.2829,
+      "step": 185000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.4183494293239685e-05,
+      "loss": 1.2755,
+      "step": 185500
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.4167816380283457e-05,
+      "loss": 1.2779,
+      "step": 186000
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.415213846732723e-05,
+      "loss": 1.2341,
+      "step": 186500
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.4136460554371004e-05,
+      "loss": 1.2414,
+      "step": 187000
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.4120782641414777e-05,
+      "loss": 1.2824,
+      "step": 187500
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.4105104728458549e-05,
+      "loss": 1.2999,
+      "step": 188000
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.408942681550232e-05,
+      "loss": 1.2051,
+      "step": 188500
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.4073748902546094e-05,
+      "loss": 1.2794,
+      "step": 189000
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.4058070989589868e-05,
+      "loss": 1.2951,
+      "step": 189500
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.4042393076633641e-05,
+      "loss": 1.261,
+      "step": 190000
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.4026715163677413e-05,
+      "loss": 1.2629,
+      "step": 190500
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.4011037250721185e-05,
+      "loss": 1.2497,
+      "step": 191000
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.3995359337764958e-05,
+      "loss": 1.2706,
+      "step": 191500
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.3979681424808732e-05,
+      "loss": 1.343,
+      "step": 192000
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.3964003511852505e-05,
+      "loss": 1.2364,
+      "step": 192500
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.3948325598896275e-05,
+      "loss": 1.3043,
+      "step": 193000
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.3932647685940048e-05,
+      "loss": 1.3173,
+      "step": 193500
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.3916969772983822e-05,
+      "loss": 1.2572,
+      "step": 194000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.3901291860027595e-05,
+      "loss": 1.2196,
+      "step": 194500
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.3885613947071369e-05,
+      "loss": 1.2879,
+      "step": 195000
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.3869936034115139e-05,
+      "loss": 1.2211,
+      "step": 195500
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.3854258121158912e-05,
+      "loss": 1.2857,
+      "step": 196000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.3838580208202686e-05,
+      "loss": 1.2577,
+      "step": 196500
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.382290229524646e-05,
+      "loss": 1.2711,
+      "step": 197000
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.380722438229023e-05,
+      "loss": 1.2587,
+      "step": 197500
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.3791546469334003e-05,
+      "loss": 1.242,
+      "step": 198000
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.3775868556377776e-05,
+      "loss": 1.28,
+      "step": 198500
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.376019064342155e-05,
+      "loss": 1.2675,
+      "step": 199000
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": NaN,
+      "eval_runtime": 128.227,
+      "eval_samples_per_second": 276.354,
+      "eval_steps_per_second": 34.548,
+      "step": 199325
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.3744512730465323e-05,
+      "loss": 1.2284,
+      "step": 199500
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.3728834817509093e-05,
+      "loss": 1.2201,
+      "step": 200000
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1.3713156904552867e-05,
+      "loss": 1.2137,
+      "step": 200500
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.369747899159664e-05,
+      "loss": 1.2371,
+      "step": 201000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1.3681801078640414e-05,
+      "loss": 1.2993,
+      "step": 201500
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.3666123165684184e-05,
+      "loss": 1.2606,
+      "step": 202000
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1.3650445252727957e-05,
+      "loss": 1.27,
+      "step": 202500
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1.363476733977173e-05,
+      "loss": 1.2601,
+      "step": 203000
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1.3619089426815504e-05,
+      "loss": 1.2119,
+      "step": 203500
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1.3603411513859277e-05,
+      "loss": 1.2293,
+      "step": 204000
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 1.3587733600903048e-05,
+      "loss": 1.1723,
+      "step": 204500
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1.3572055687946821e-05,
+      "loss": 1.2246,
+      "step": 205000
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 1.3556377774990594e-05,
+      "loss": 1.1749,
+      "step": 205500
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 1.3540699862034368e-05,
+      "loss": 1.2344,
+      "step": 206000
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 1.3525021949078141e-05,
+      "loss": 1.2247,
+      "step": 206500
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 1.3509344036121911e-05,
+      "loss": 1.2209,
+      "step": 207000
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 1.3493666123165685e-05,
+      "loss": 1.252,
+      "step": 207500
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 1.3477988210209458e-05,
+      "loss": 1.1713,
+      "step": 208000
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 1.3462310297253232e-05,
+      "loss": 1.2387,
+      "step": 208500
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 1.3446632384297002e-05,
+      "loss": 1.2507,
+      "step": 209000
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 1.3430954471340775e-05,
+      "loss": 1.2971,
+      "step": 209500
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 1.3415276558384549e-05,
+      "loss": 1.2168,
+      "step": 210000
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 1.3399598645428322e-05,
+      "loss": 1.2556,
+      "step": 210500
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 1.3383920732472096e-05,
+      "loss": 1.2302,
+      "step": 211000
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 1.3368242819515866e-05,
+      "loss": 1.261,
+      "step": 211500
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 1.3352564906559639e-05,
+      "loss": 1.1894,
+      "step": 212000
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 1.3336886993603413e-05,
+      "loss": 1.2134,
+      "step": 212500
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 1.3321209080647186e-05,
+      "loss": 1.273,
+      "step": 213000
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 1.3305531167690958e-05,
+      "loss": 1.3024,
+      "step": 213500
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 1.328985325473473e-05,
+      "loss": 1.2547,
+      "step": 214000
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 1.3274175341778503e-05,
+      "loss": 1.2567,
+      "step": 214500
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 1.3258497428822276e-05,
+      "loss": 1.2505,
+      "step": 215000
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 1.324281951586605e-05,
+      "loss": 1.3005,
+      "step": 215500
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 1.3227141602909822e-05,
+      "loss": 1.2019,
+      "step": 216000
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 1.3211463689953593e-05,
+      "loss": 1.228,
+      "step": 216500
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1.3195785776997367e-05,
+      "loss": 1.2223,
+      "step": 217000
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 1.318010786404114e-05,
+      "loss": 1.2821,
+      "step": 217500
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 1.3164429951084912e-05,
+      "loss": 1.2201,
+      "step": 218000
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 1.3148752038128686e-05,
+      "loss": 1.2231,
+      "step": 218500
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 1.3133074125172457e-05,
+      "loss": 1.2294,
+      "step": 219000
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 1.311739621221623e-05,
+      "loss": 1.2198,
+      "step": 219500
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 1.3101718299260004e-05,
+      "loss": 1.2677,
+      "step": 220000
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 1.3086040386303776e-05,
+      "loss": 1.218,
+      "step": 220500
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 1.307036247334755e-05,
+      "loss": 1.2484,
+      "step": 221000
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 1.3054684560391321e-05,
+      "loss": 1.1905,
+      "step": 221500
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 1.3039006647435095e-05,
+      "loss": 1.2688,
+      "step": 222000
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 1.3023328734478866e-05,
+      "loss": 1.248,
+      "step": 222500
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 1.300765082152264e-05,
+      "loss": 1.2758,
+      "step": 223000
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 1.2991972908566413e-05,
+      "loss": 1.2298,
+      "step": 223500
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 1.2976294995610185e-05,
+      "loss": 1.2808,
+      "step": 224000
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 1.2960617082653959e-05,
+      "loss": 1.2008,
+      "step": 224500
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 1.294493916969773e-05,
+      "loss": 1.2582,
+      "step": 225000
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 1.2929261256741504e-05,
+      "loss": 1.209,
+      "step": 225500
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 1.2913583343785277e-05,
+      "loss": 1.2933,
+      "step": 226000
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1.2897905430829049e-05,
+      "loss": 1.2576,
+      "step": 226500
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 1.2882227517872822e-05,
+      "loss": 1.2619,
+      "step": 227000
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 1.2866549604916594e-05,
+      "loss": 1.1988,
+      "step": 227500
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 1.2850871691960368e-05,
+      "loss": 1.2358,
+      "step": 228000
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1.2835193779004141e-05,
+      "loss": 1.2377,
+      "step": 228500
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 1.2819515866047913e-05,
+      "loss": 1.2357,
+      "step": 229000
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 1.2803837953091685e-05,
+      "loss": 1.2496,
+      "step": 229500
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 1.2788160040135458e-05,
+      "loss": 1.2092,
+      "step": 230000
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 1.2772482127179232e-05,
+      "loss": 1.267,
+      "step": 230500
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 1.2756804214223005e-05,
+      "loss": 1.2355,
+      "step": 231000
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 1.2741126301266777e-05,
+      "loss": 1.2225,
+      "step": 231500
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 1.2725448388310548e-05,
+      "loss": 1.2578,
+      "step": 232000
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 1.2709770475354322e-05,
+      "loss": 1.2468,
+      "step": 232500
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 1.2694092562398095e-05,
+      "loss": 1.2151,
+      "step": 233000
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 1.2678414649441869e-05,
+      "loss": 1.1998,
+      "step": 233500
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 1.2662736736485639e-05,
+      "loss": 1.2087,
+      "step": 234000
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 1.2647058823529412e-05,
+      "loss": 1.2726,
+      "step": 234500
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 1.2631380910573186e-05,
+      "loss": 1.2146,
+      "step": 235000
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1.261570299761696e-05,
+      "loss": 1.2373,
+      "step": 235500
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 1.2600025084660733e-05,
+      "loss": 1.264,
+      "step": 236000
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 1.2584347171704503e-05,
+      "loss": 1.2405,
+      "step": 236500
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 1.2568669258748276e-05,
+      "loss": 1.2153,
+      "step": 237000
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 1.255299134579205e-05,
+      "loss": 1.2191,
+      "step": 237500
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 1.2537313432835823e-05,
+      "loss": 1.3123,
+      "step": 238000
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 1.2521635519879593e-05,
+      "loss": 1.2528,
+      "step": 238500
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 1.2505957606923367e-05,
+      "loss": 1.2381,
+      "step": 239000
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 1.254626989364624,
+      "eval_runtime": 127.3598,
+      "eval_samples_per_second": 278.235,
+      "eval_steps_per_second": 34.783,
+      "step": 239190
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 1.249027969396714e-05,
+      "loss": 1.2663,
+      "step": 239500
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 1.2474601781010914e-05,
+      "loss": 1.2548,
+      "step": 240000
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 1.2458923868054687e-05,
+      "loss": 1.2001,
+      "step": 240500
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 1.2443245955098457e-05,
+      "loss": 1.2247,
+      "step": 241000
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.242756804214223e-05,
+      "loss": 1.2295,
+      "step": 241500
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1.2411890129186004e-05,
+      "loss": 1.1823,
+      "step": 242000
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 1.2396212216229777e-05,
+      "loss": 1.1923,
+      "step": 242500
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 1.2380534303273551e-05,
+      "loss": 1.1905,
+      "step": 243000
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 1.2364856390317321e-05,
+      "loss": 1.1896,
+      "step": 243500
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 1.2349178477361094e-05,
+      "loss": 1.1912,
+      "step": 244000
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 1.2333500564404868e-05,
+      "loss": 1.235,
+      "step": 244500
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 1.2317822651448641e-05,
+      "loss": 1.185,
+      "step": 245000
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 1.2302144738492411e-05,
+      "loss": 1.2781,
+      "step": 245500
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 1.2286466825536185e-05,
+      "loss": 1.2342,
+      "step": 246000
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 1.2270788912579958e-05,
+      "loss": 1.2257,
+      "step": 246500
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 1.2255110999623732e-05,
+      "loss": 1.2575,
+      "step": 247000
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 1.2239433086667505e-05,
+      "loss": 1.2028,
+      "step": 247500
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 1.2223755173711275e-05,
+      "loss": 1.2228,
+      "step": 248000
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 1.2208077260755049e-05,
+      "loss": 1.1886,
+      "step": 248500
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 1.2192399347798822e-05,
+      "loss": 1.1669,
+      "step": 249000
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 1.2176721434842596e-05,
+      "loss": 1.1943,
+      "step": 249500
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 1.2161043521886366e-05,
+      "loss": 1.2077,
+      "step": 250000
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 1.2145365608930139e-05,
+      "loss": 1.2227,
+      "step": 250500
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 1.2129687695973913e-05,
+      "loss": 1.2227,
+      "step": 251000
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 1.2114009783017686e-05,
+      "loss": 1.1362,
+      "step": 251500
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 1.209833187006146e-05,
+      "loss": 1.2037,
+      "step": 252000
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 1.208265395710523e-05,
+      "loss": 1.2609,
+      "step": 252500
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 1.2066976044149003e-05,
+      "loss": 1.1977,
+      "step": 253000
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 1.2051298131192776e-05,
+      "loss": 1.1819,
+      "step": 253500
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 1.203562021823655e-05,
+      "loss": 1.1479,
+      "step": 254000
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 1.2019942305280322e-05,
+      "loss": 1.256,
+      "step": 254500
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 1.2004264392324093e-05,
+      "loss": 1.1129,
+      "step": 255000
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 1.1988586479367867e-05,
+      "loss": 1.2096,
+      "step": 255500
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 1.197290856641164e-05,
+      "loss": 1.2323,
+      "step": 256000
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.1957230653455414e-05,
+      "loss": 1.2027,
+      "step": 256500
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 1.1941552740499186e-05,
+      "loss": 1.2175,
+      "step": 257000
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.1925874827542957e-05,
+      "loss": 1.193,
+      "step": 257500
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 1.191019691458673e-05,
+      "loss": 1.1808,
+      "step": 258000
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 1.1894519001630504e-05,
+      "loss": 1.1994,
+      "step": 258500
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 1.1878841088674276e-05,
+      "loss": 1.1686,
+      "step": 259000
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 1.186316317571805e-05,
+      "loss": 1.1599,
+      "step": 259500
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 1.1847485262761821e-05,
+      "loss": 1.2177,
+      "step": 260000
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1.1831807349805595e-05,
+      "loss": 1.2177,
+      "step": 260500
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 1.1816129436849368e-05,
+      "loss": 1.1605,
+      "step": 261000
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 1.180045152389314e-05,
+      "loss": 1.2487,
+      "step": 261500
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 1.1784773610936913e-05,
+      "loss": 1.2641,
+      "step": 262000
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 1.1769095697980685e-05,
+      "loss": 1.1597,
+      "step": 262500
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 1.1753417785024459e-05,
+      "loss": 1.1846,
+      "step": 263000
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1.1737739872068232e-05,
+      "loss": 1.2127,
+      "step": 263500
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 1.1722061959112004e-05,
+      "loss": 1.1935,
+      "step": 264000
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.1706384046155777e-05,
+      "loss": 1.2079,
+      "step": 264500
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 1.1690706133199549e-05,
+      "loss": 1.2393,
+      "step": 265000
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 1.1675028220243322e-05,
+      "loss": 1.2295,
+      "step": 265500
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 1.1659350307287094e-05,
+      "loss": 1.2593,
+      "step": 266000
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1.1643672394330868e-05,
+      "loss": 1.2002,
+      "step": 266500
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 1.1627994481374641e-05,
+      "loss": 1.1717,
+      "step": 267000
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 1.1612316568418413e-05,
+      "loss": 1.2141,
+      "step": 267500
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 1.1596638655462186e-05,
+      "loss": 1.171,
+      "step": 268000
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 1.1580960742505958e-05,
+      "loss": 1.1938,
+      "step": 268500
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 1.1565282829549731e-05,
+      "loss": 1.21,
+      "step": 269000
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1.1549604916593505e-05,
+      "loss": 1.2357,
+      "step": 269500
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 1.1533927003637277e-05,
+      "loss": 1.1811,
+      "step": 270000
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 1.1518249090681048e-05,
+      "loss": 1.2406,
+      "step": 270500
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 1.1502571177724822e-05,
+      "loss": 1.2038,
+      "step": 271000
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 1.1486893264768595e-05,
+      "loss": 1.204,
+      "step": 271500
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 1.1471215351812369e-05,
+      "loss": 1.207,
+      "step": 272000
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1.145553743885614e-05,
+      "loss": 1.1973,
+      "step": 272500
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 1.1439859525899912e-05,
+      "loss": 1.2101,
+      "step": 273000
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 1.1424181612943686e-05,
+      "loss": 1.1494,
+      "step": 273500
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 1.140850369998746e-05,
+      "loss": 1.2214,
+      "step": 274000
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 1.1392825787031233e-05,
+      "loss": 1.1631,
+      "step": 274500
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 1.1377147874075003e-05,
+      "loss": 1.2273,
+      "step": 275000
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1.1361469961118776e-05,
+      "loss": 1.238,
+      "step": 275500
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 1.134579204816255e-05,
+      "loss": 1.2387,
+      "step": 276000
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 1.1330114135206323e-05,
+      "loss": 1.2499,
+      "step": 276500
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 1.1314436222250097e-05,
+      "loss": 1.18,
+      "step": 277000
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 1.1298758309293867e-05,
+      "loss": 1.2259,
+      "step": 277500
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 1.128308039633764e-05,
+      "loss": 1.2246,
+      "step": 278000
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1.1267402483381414e-05,
+      "loss": 1.1923,
+      "step": 278500
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 1.1251724570425187e-05,
+      "loss": 1.1841,
+      "step": 279000
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 1.2204227447509766,
+      "eval_runtime": 127.6995,
+      "eval_samples_per_second": 277.495,
+      "eval_steps_per_second": 34.691,
+      "step": 279055
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 1.123604665746896e-05,
+      "loss": 1.1389,
+      "step": 279500
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 1.122036874451273e-05,
+      "loss": 1.1898,
+      "step": 280000
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 1.1204690831556504e-05,
+      "loss": 1.1146,
+      "step": 280500
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 1.1189012918600277e-05,
+      "loss": 1.1572,
+      "step": 281000
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 1.1173335005644051e-05,
+      "loss": 1.1553,
+      "step": 281500
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 1.1157657092687821e-05,
+      "loss": 1.2099,
+      "step": 282000
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 1.1141979179731594e-05,
+      "loss": 1.1562,
+      "step": 282500
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 1.1126301266775368e-05,
+      "loss": 1.1754,
+      "step": 283000
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 1.1110623353819141e-05,
+      "loss": 1.2016,
+      "step": 283500
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 1.1094945440862915e-05,
+      "loss": 1.177,
+      "step": 284000
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 1.1079267527906685e-05,
+      "loss": 1.2008,
+      "step": 284500
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 1.1063589614950458e-05,
+      "loss": 1.2134,
+      "step": 285000
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 1.1047911701994232e-05,
+      "loss": 1.1965,
+      "step": 285500
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 1.1032233789038005e-05,
+      "loss": 1.1614,
+      "step": 286000
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 1.1016555876081775e-05,
+      "loss": 1.2116,
+      "step": 286500
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 1.1000877963125549e-05,
+      "loss": 1.1696,
+      "step": 287000
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 1.0985200050169322e-05,
+      "loss": 1.192,
+      "step": 287500
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 1.0969522137213096e-05,
+      "loss": 1.1865,
+      "step": 288000
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 1.0953844224256869e-05,
+      "loss": 1.1716,
+      "step": 288500
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 1.0938166311300639e-05,
+      "loss": 1.1395,
+      "step": 289000
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 1.0922488398344413e-05,
+      "loss": 1.1886,
+      "step": 289500
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 1.0906810485388186e-05,
+      "loss": 1.1592,
+      "step": 290000
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 1.089113257243196e-05,
+      "loss": 1.1792,
+      "step": 290500
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 1.0875454659475731e-05,
+      "loss": 1.2241,
+      "step": 291000
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 1.0859776746519503e-05,
+      "loss": 1.2109,
+      "step": 291500
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 1.0844098833563276e-05,
+      "loss": 1.1347,
+      "step": 292000
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 1.082842092060705e-05,
+      "loss": 1.1454,
+      "step": 292500
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 1.0812743007650823e-05,
+      "loss": 1.1888,
+      "step": 293000
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 1.0797065094694595e-05,
+      "loss": 1.1764,
+      "step": 293500
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 1.0781387181738367e-05,
+      "loss": 1.2194,
+      "step": 294000
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 1.076570926878214e-05,
+      "loss": 1.1219,
+      "step": 294500
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 1.0750031355825914e-05,
+      "loss": 1.2311,
+      "step": 295000
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 1.0734353442869687e-05,
+      "loss": 1.2233,
+      "step": 295500
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 1.0718675529913459e-05,
+      "loss": 1.2083,
+      "step": 296000
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 1.070299761695723e-05,
+      "loss": 1.1935,
+      "step": 296500
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 1.0687319704001004e-05,
+      "loss": 1.1264,
+      "step": 297000
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 1.0671641791044778e-05,
+      "loss": 1.1662,
+      "step": 297500
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 1.065596387808855e-05,
+      "loss": 1.159,
+      "step": 298000
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 1.0640285965132323e-05,
+      "loss": 1.1865,
+      "step": 298500
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 1.0624608052176095e-05,
+      "loss": 1.2219,
+      "step": 299000
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 1.0608930139219868e-05,
+      "loss": 1.1702,
+      "step": 299500
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 1.0593252226263642e-05,
+      "loss": 1.1198,
+      "step": 300000
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 1.0577574313307413e-05,
+      "loss": 1.1777,
+      "step": 300500
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 1.0561896400351187e-05,
+      "loss": 1.1786,
+      "step": 301000
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 1.0546218487394959e-05,
+      "loss": 1.2348,
+      "step": 301500
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 1.0530540574438732e-05,
+      "loss": 1.2165,
+      "step": 302000
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 1.0514862661482504e-05,
+      "loss": 1.2002,
+      "step": 302500
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 1.0499184748526277e-05,
+      "loss": 1.1752,
+      "step": 303000
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 1.0483506835570049e-05,
+      "loss": 1.1828,
+      "step": 303500
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 1.0467828922613822e-05,
+      "loss": 1.1647,
+      "step": 304000
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 1.0452151009657596e-05,
+      "loss": 1.1328,
+      "step": 304500
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 1.0436473096701368e-05,
+      "loss": 1.1921,
+      "step": 305000
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 1.0420795183745141e-05,
+      "loss": 1.1367,
+      "step": 305500
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 1.0405117270788913e-05,
+      "loss": 1.2171,
+      "step": 306000
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 1.0389439357832686e-05,
+      "loss": 1.1139,
+      "step": 306500
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 1.0373761444876458e-05,
+      "loss": 1.1636,
+      "step": 307000
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 1.0358083531920231e-05,
+      "loss": 1.1886,
+      "step": 307500
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 1.0342405618964005e-05,
+      "loss": 1.1854,
+      "step": 308000
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 1.0326727706007777e-05,
+      "loss": 1.1162,
+      "step": 308500
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 1.031104979305155e-05,
+      "loss": 1.1896,
+      "step": 309000
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 1.0295371880095322e-05,
+      "loss": 1.181,
+      "step": 309500
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 1.0279693967139095e-05,
+      "loss": 1.2068,
+      "step": 310000
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 1.0264016054182869e-05,
+      "loss": 1.1469,
+      "step": 310500
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 1.024833814122664e-05,
+      "loss": 1.1962,
+      "step": 311000
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 1.0232660228270412e-05,
+      "loss": 1.2163,
+      "step": 311500
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 1.0216982315314186e-05,
+      "loss": 1.1355,
+      "step": 312000
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 1.020130440235796e-05,
+      "loss": 1.177,
+      "step": 312500
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 1.0185626489401733e-05,
+      "loss": 1.1534,
+      "step": 313000
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 1.0169948576445504e-05,
+      "loss": 1.152,
+      "step": 313500
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 1.0154270663489276e-05,
+      "loss": 1.1655,
+      "step": 314000
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 1.013859275053305e-05,
+      "loss": 1.2176,
+      "step": 314500
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 1.0122914837576823e-05,
+      "loss": 1.1287,
+      "step": 315000
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 1.0107236924620597e-05,
+      "loss": 1.1938,
+      "step": 315500
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 1.0091559011664368e-05,
+      "loss": 1.1679,
+      "step": 316000
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 1.007588109870814e-05,
+      "loss": 1.2031,
+      "step": 316500
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 1.0060203185751914e-05,
+      "loss": 1.1444,
+      "step": 317000
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 1.0044525272795687e-05,
+      "loss": 1.1229,
+      "step": 317500
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 1.002884735983946e-05,
+      "loss": 1.1438,
+      "step": 318000
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 1.001316944688323e-05,
+      "loss": 1.1608,
+      "step": 318500
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 1.2094546556472778,
+      "eval_runtime": 127.5825,
+      "eval_samples_per_second": 277.75,
+      "eval_steps_per_second": 34.723,
+      "step": 318920
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 9.997491533927004e-06,
+      "loss": 1.1374,
+      "step": 319000
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 9.981813620970777e-06,
+      "loss": 1.1764,
+      "step": 319500
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 9.966135708014551e-06,
+      "loss": 1.1906,
+      "step": 320000
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 9.950457795058323e-06,
+      "loss": 1.1977,
+      "step": 320500
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 9.934779882102096e-06,
+      "loss": 1.1281,
+      "step": 321000
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 9.919101969145868e-06,
+      "loss": 1.119,
+      "step": 321500
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 9.903424056189641e-06,
+      "loss": 1.1616,
+      "step": 322000
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 9.887746143233413e-06,
+      "loss": 1.1467,
+      "step": 322500
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 9.872068230277187e-06,
+      "loss": 1.1215,
+      "step": 323000
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 9.856390317320958e-06,
+      "loss": 1.1575,
+      "step": 323500
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 9.840712404364732e-06,
+      "loss": 1.139,
+      "step": 324000
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 9.825034491408505e-06,
+      "loss": 1.1644,
+      "step": 324500
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 9.809356578452277e-06,
+      "loss": 1.1423,
+      "step": 325000
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 9.79367866549605e-06,
+      "loss": 1.1717,
+      "step": 325500
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 9.778000752539822e-06,
+      "loss": 1.1742,
+      "step": 326000
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 9.762322839583596e-06,
+      "loss": 1.1472,
+      "step": 326500
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 9.746644926627367e-06,
+      "loss": 1.1373,
+      "step": 327000
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 9.73096701367114e-06,
+      "loss": 1.1726,
+      "step": 327500
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 9.715289100714914e-06,
+      "loss": 1.0968,
+      "step": 328000
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 9.699611187758686e-06,
+      "loss": 1.1652,
+      "step": 328500
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 9.68393327480246e-06,
+      "loss": 1.1436,
+      "step": 329000
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 9.668255361846231e-06,
+      "loss": 1.079,
+      "step": 329500
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 9.652577448890005e-06,
+      "loss": 1.1103,
+      "step": 330000
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 9.636899535933776e-06,
+      "loss": 1.1668,
+      "step": 330500
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 9.62122162297755e-06,
+      "loss": 1.1915,
+      "step": 331000
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 9.605543710021322e-06,
+      "loss": 1.1656,
+      "step": 331500
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 9.589865797065095e-06,
+      "loss": 1.1738,
+      "step": 332000
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 9.574187884108869e-06,
+      "loss": 1.1476,
+      "step": 332500
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 9.55850997115264e-06,
+      "loss": 1.1956,
+      "step": 333000
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 9.542832058196414e-06,
+      "loss": 1.0962,
+      "step": 333500
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 9.527154145240186e-06,
+      "loss": 1.1355,
+      "step": 334000
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 9.511476232283959e-06,
+      "loss": 1.1325,
+      "step": 334500
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 9.49579831932773e-06,
+      "loss": 1.1264,
+      "step": 335000
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 9.480120406371504e-06,
+      "loss": 1.1675,
+      "step": 335500
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 9.464442493415278e-06,
+      "loss": 1.2012,
+      "step": 336000
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 9.44876458045905e-06,
+      "loss": 1.1367,
+      "step": 336500
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 9.433086667502823e-06,
+      "loss": 1.1672,
+      "step": 337000
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 9.417408754546595e-06,
+      "loss": 1.2382,
+      "step": 337500
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 9.401730841590368e-06,
+      "loss": 1.0987,
+      "step": 338000
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 9.38605292863414e-06,
+      "loss": 1.1812,
+      "step": 338500
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 9.370375015677913e-06,
+      "loss": 1.1735,
+      "step": 339000
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 9.354697102721687e-06,
+      "loss": 1.239,
+      "step": 339500
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 9.339019189765458e-06,
+      "loss": 1.0647,
+      "step": 340000
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 9.323341276809232e-06,
+      "loss": 1.116,
+      "step": 340500
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 9.307663363853004e-06,
+      "loss": 1.1161,
+      "step": 341000
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 9.291985450896777e-06,
+      "loss": 1.1656,
+      "step": 341500
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 9.27630753794055e-06,
+      "loss": 1.0908,
+      "step": 342000
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 9.260629624984322e-06,
+      "loss": 1.1451,
+      "step": 342500
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 9.244951712028096e-06,
+      "loss": 1.1246,
+      "step": 343000
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 9.229273799071868e-06,
+      "loss": 1.2057,
+      "step": 343500
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 9.213595886115641e-06,
+      "loss": 1.1393,
+      "step": 344000
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 9.197917973159414e-06,
+      "loss": 1.1479,
+      "step": 344500
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 9.182240060203186e-06,
+      "loss": 1.1118,
+      "step": 345000
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 9.16656214724696e-06,
+      "loss": 1.1616,
+      "step": 345500
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 9.150884234290731e-06,
+      "loss": 1.135,
+      "step": 346000
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 9.135206321334505e-06,
+      "loss": 1.1225,
+      "step": 346500
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 9.119528408378278e-06,
+      "loss": 1.1603,
+      "step": 347000
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 9.10385049542205e-06,
+      "loss": 1.1703,
+      "step": 347500
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 9.088172582465824e-06,
+      "loss": 1.2262,
+      "step": 348000
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 9.072494669509595e-06,
+      "loss": 1.06,
+      "step": 348500
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 9.056816756553369e-06,
+      "loss": 1.1594,
+      "step": 349000
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 9.041138843597142e-06,
+      "loss": 1.2018,
+      "step": 349500
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 9.025460930640914e-06,
+      "loss": 1.1478,
+      "step": 350000
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 9.009783017684687e-06,
+      "loss": 1.145,
+      "step": 350500
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 8.99410510472846e-06,
+      "loss": 1.1299,
+      "step": 351000
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 8.978427191772233e-06,
+      "loss": 1.1853,
+      "step": 351500
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 8.962749278816004e-06,
+      "loss": 1.1524,
+      "step": 352000
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 8.947071365859778e-06,
+      "loss": 1.2222,
+      "step": 352500
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 8.931393452903551e-06,
+      "loss": 1.1312,
+      "step": 353000
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 8.915715539947323e-06,
+      "loss": 1.171,
+      "step": 353500
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 8.900037626991097e-06,
+      "loss": 1.1357,
+      "step": 354000
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 8.884359714034868e-06,
+      "loss": 1.1231,
+      "step": 354500
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 8.868681801078642e-06,
+      "loss": 1.1771,
+      "step": 355000
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 8.853003888122414e-06,
+      "loss": 1.1985,
+      "step": 355500
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 8.837325975166187e-06,
+      "loss": 1.1563,
+      "step": 356000
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 8.82164806220996e-06,
+      "loss": 1.1187,
+      "step": 356500
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 8.805970149253732e-06,
+      "loss": 1.1113,
+      "step": 357000
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 8.790292236297506e-06,
+      "loss": 1.1305,
+      "step": 357500
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 8.774614323341277e-06,
+      "loss": 1.1053,
+      "step": 358000
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 8.758936410385051e-06,
+      "loss": 1.2045,
+      "step": 358500
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 1.181683897972107,
+      "eval_runtime": 127.1715,
+      "eval_samples_per_second": 278.647,
+      "eval_steps_per_second": 34.835,
+      "step": 358785
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 8.743258497428823e-06,
+      "loss": 1.1085,
+      "step": 359000
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 8.727580584472596e-06,
+      "loss": 1.0804,
+      "step": 359500
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 8.711902671516368e-06,
+      "loss": 1.1311,
+      "step": 360000
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 8.696224758560141e-06,
+      "loss": 1.159,
+      "step": 360500
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 8.680546845603915e-06,
+      "loss": 1.1409,
+      "step": 361000
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 8.664868932647686e-06,
+      "loss": 1.1436,
+      "step": 361500
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 8.64919101969146e-06,
+      "loss": 1.1535,
+      "step": 362000
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 8.633513106735232e-06,
+      "loss": 1.0645,
+      "step": 362500
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 8.617835193779005e-06,
+      "loss": 1.1221,
+      "step": 363000
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 8.602157280822777e-06,
+      "loss": 1.1127,
+      "step": 363500
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 8.58647936786655e-06,
+      "loss": 1.1139,
+      "step": 364000
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 8.570801454910324e-06,
+      "loss": 1.1128,
+      "step": 364500
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 8.555123541954096e-06,
+      "loss": 1.1384,
+      "step": 365000
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 8.539445628997869e-06,
+      "loss": 1.0713,
+      "step": 365500
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 8.52376771604164e-06,
+      "loss": 1.1902,
+      "step": 366000
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 8.508089803085414e-06,
+      "loss": 1.1502,
+      "step": 366500
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 8.492411890129186e-06,
+      "loss": 1.136,
+      "step": 367000
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 8.47673397717296e-06,
+      "loss": 1.0885,
+      "step": 367500
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 8.461056064216731e-06,
+      "loss": 1.1144,
+      "step": 368000
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 8.445378151260505e-06,
+      "loss": 1.1733,
+      "step": 368500
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 8.429700238304278e-06,
+      "loss": 1.1403,
+      "step": 369000
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 8.41402232534805e-06,
+      "loss": 1.145,
+      "step": 369500
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 8.398344412391823e-06,
+      "loss": 1.0408,
+      "step": 370000
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 8.382666499435595e-06,
+      "loss": 1.1128,
+      "step": 370500
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 8.366988586479369e-06,
+      "loss": 1.1214,
+      "step": 371000
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 8.35131067352314e-06,
+      "loss": 1.1047,
+      "step": 371500
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 8.335632760566914e-06,
+      "loss": 1.1368,
+      "step": 372000
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 8.319954847610687e-06,
+      "loss": 1.1855,
+      "step": 372500
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 8.304276934654459e-06,
+      "loss": 1.1038,
+      "step": 373000
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 8.288599021698232e-06,
+      "loss": 1.172,
+      "step": 373500
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 8.272921108742004e-06,
+      "loss": 1.1118,
+      "step": 374000
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 8.257243195785778e-06,
+      "loss": 1.0809,
+      "step": 374500
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 8.24156528282955e-06,
+      "loss": 1.1414,
+      "step": 375000
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 8.225887369873323e-06,
+      "loss": 1.1448,
+      "step": 375500
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 8.210209456917095e-06,
+      "loss": 1.1143,
+      "step": 376000
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 8.194531543960868e-06,
+      "loss": 1.1027,
+      "step": 376500
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 8.178853631004642e-06,
+      "loss": 1.1401,
+      "step": 377000
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 8.163175718048413e-06,
+      "loss": 1.131,
+      "step": 377500
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 8.147497805092187e-06,
+      "loss": 1.1049,
+      "step": 378000
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 8.131819892135958e-06,
+      "loss": 1.1312,
+      "step": 378500
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 8.116141979179732e-06,
+      "loss": 1.1163,
+      "step": 379000
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 8.100464066223505e-06,
+      "loss": 1.105,
+      "step": 379500
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 8.084786153267277e-06,
+      "loss": 1.1223,
+      "step": 380000
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 8.06910824031105e-06,
+      "loss": 1.159,
+      "step": 380500
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 8.053430327354822e-06,
+      "loss": 1.0807,
+      "step": 381000
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 8.037752414398596e-06,
+      "loss": 1.1768,
+      "step": 381500
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 8.02207450144237e-06,
+      "loss": 1.1399,
+      "step": 382000
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 8.006396588486141e-06,
+      "loss": 1.1426,
+      "step": 382500
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 7.990718675529914e-06,
+      "loss": 1.066,
+      "step": 383000
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 7.975040762573686e-06,
+      "loss": 1.1044,
+      "step": 383500
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 7.95936284961746e-06,
+      "loss": 1.1479,
+      "step": 384000
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 7.943684936661233e-06,
+      "loss": 1.0977,
+      "step": 384500
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 7.928007023705005e-06,
+      "loss": 1.1236,
+      "step": 385000
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 7.912329110748778e-06,
+      "loss": 1.152,
+      "step": 385500
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 7.89665119779255e-06,
+      "loss": 1.1036,
+      "step": 386000
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 7.880973284836324e-06,
+      "loss": 1.1227,
+      "step": 386500
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 7.865295371880095e-06,
+      "loss": 1.0814,
+      "step": 387000
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 7.849617458923869e-06,
+      "loss": 1.158,
+      "step": 387500
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 7.833939545967642e-06,
+      "loss": 1.1237,
+      "step": 388000
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 7.818261633011414e-06,
+      "loss": 1.173,
+      "step": 388500
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 7.802583720055187e-06,
+      "loss": 1.1304,
+      "step": 389000
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 7.78690580709896e-06,
+      "loss": 1.0871,
+      "step": 389500
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 7.771227894142733e-06,
+      "loss": 1.1715,
+      "step": 390000
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 7.755549981186506e-06,
+      "loss": 1.132,
+      "step": 390500
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 7.739872068230278e-06,
+      "loss": 1.1601,
+      "step": 391000
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 7.724194155274051e-06,
+      "loss": 1.1632,
+      "step": 391500
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 7.708516242317823e-06,
+      "loss": 1.1204,
+      "step": 392000
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 7.692838329361597e-06,
+      "loss": 1.0789,
+      "step": 392500
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 7.67716041640537e-06,
+      "loss": 1.1613,
+      "step": 393000
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 7.661482503449142e-06,
+      "loss": 1.1628,
+      "step": 393500
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 7.645804590492915e-06,
+      "loss": 1.0951,
+      "step": 394000
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 7.630126677536687e-06,
+      "loss": 1.1478,
+      "step": 394500
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 7.6144487645804596e-06,
+      "loss": 1.1075,
+      "step": 395000
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 7.598770851624232e-06,
+      "loss": 1.1255,
+      "step": 395500
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 7.583092938668006e-06,
+      "loss": 1.0771,
+      "step": 396000
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 7.567415025711777e-06,
+      "loss": 1.1575,
+      "step": 396500
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 7.551737112755551e-06,
+      "loss": 1.1018,
+      "step": 397000
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 7.5360591997993235e-06,
+      "loss": 1.1161,
+      "step": 397500
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 7.520381286843096e-06,
+      "loss": 1.1226,
+      "step": 398000
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 7.5047033738868695e-06,
+      "loss": 1.1401,
+      "step": 398500
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 1.1810191869735718,
+      "eval_runtime": 128.7944,
+      "eval_samples_per_second": 275.136,
+      "eval_steps_per_second": 34.396,
+      "step": 398650
+    },
+    {
+      "epoch": 10.01,
+      "learning_rate": 7.489025460930641e-06,
+      "loss": 1.0806,
+      "step": 399000
+    },
+    {
+      "epoch": 10.02,
+      "learning_rate": 7.473347547974415e-06,
+      "loss": 1.1156,
+      "step": 399500
+    },
+    {
+      "epoch": 10.03,
+      "learning_rate": 7.4576696350181865e-06,
+      "loss": 1.1052,
+      "step": 400000
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 7.44199172206196e-06,
+      "loss": 1.0788,
+      "step": 400500
+    },
+    {
+      "epoch": 10.06,
+      "learning_rate": 7.4263138091057325e-06,
+      "loss": 1.0814,
+      "step": 401000
+    },
+    {
+      "epoch": 10.07,
+      "learning_rate": 7.410635896149505e-06,
+      "loss": 1.0861,
+      "step": 401500
+    },
+    {
+      "epoch": 10.08,
+      "learning_rate": 7.394957983193279e-06,
+      "loss": 1.0579,
+      "step": 402000
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 7.37928007023705e-06,
+      "loss": 1.0808,
+      "step": 402500
+    },
+    {
+      "epoch": 10.11,
+      "learning_rate": 7.363602157280824e-06,
+      "loss": 1.0664,
+      "step": 403000
+    },
+    {
+      "epoch": 10.12,
+      "learning_rate": 7.3479242443245956e-06,
+      "loss": 1.1089,
+      "step": 403500
+    },
+    {
+      "epoch": 10.13,
+      "learning_rate": 7.332246331368369e-06,
+      "loss": 1.041,
+      "step": 404000
+    },
+    {
+      "epoch": 10.15,
+      "learning_rate": 7.316568418412141e-06,
+      "loss": 1.1319,
+      "step": 404500
+    },
+    {
+      "epoch": 10.16,
+      "learning_rate": 7.300890505455914e-06,
+      "loss": 1.0706,
+      "step": 405000
+    },
+    {
+      "epoch": 10.17,
+      "learning_rate": 7.285212592499688e-06,
+      "loss": 1.1228,
+      "step": 405500
+    },
+    {
+      "epoch": 10.18,
+      "learning_rate": 7.2695346795434594e-06,
+      "loss": 1.0994,
+      "step": 406000
+    },
+    {
+      "epoch": 10.2,
+      "learning_rate": 7.253856766587233e-06,
+      "loss": 1.1089,
+      "step": 406500
+    },
+    {
+      "epoch": 10.21,
+      "learning_rate": 7.238178853631005e-06,
+      "loss": 1.0791,
+      "step": 407000
+    },
+    {
+      "epoch": 10.22,
+      "learning_rate": 7.222500940674778e-06,
+      "loss": 1.0507,
+      "step": 407500
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 7.20682302771855e-06,
+      "loss": 1.0938,
+      "step": 408000
+    },
+    {
+      "epoch": 10.25,
+      "learning_rate": 7.191145114762323e-06,
+      "loss": 1.1371,
+      "step": 408500
+    },
+    {
+      "epoch": 10.26,
+      "learning_rate": 7.175467201806097e-06,
+      "loss": 1.1441,
+      "step": 409000
+    },
+    {
+      "epoch": 10.27,
+      "learning_rate": 7.1597892888498685e-06,
+      "loss": 1.1253,
+      "step": 409500
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 7.144111375893642e-06,
+      "loss": 1.0862,
+      "step": 410000
+    },
+    {
+      "epoch": 10.3,
+      "learning_rate": 7.128433462937414e-06,
+      "loss": 1.0671,
+      "step": 410500
+    },
+    {
+      "epoch": 10.31,
+      "learning_rate": 7.112755549981187e-06,
+      "loss": 1.1025,
+      "step": 411000
+    },
+    {
+      "epoch": 10.32,
+      "learning_rate": 7.09707763702496e-06,
+      "loss": 1.112,
+      "step": 411500
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 7.081399724068732e-06,
+      "loss": 1.1957,
+      "step": 412000
+    },
+    {
+      "epoch": 10.35,
+      "learning_rate": 7.065721811112505e-06,
+      "loss": 1.1165,
+      "step": 412500
+    },
+    {
+      "epoch": 10.36,
+      "learning_rate": 7.050043898156278e-06,
+      "loss": 1.0835,
+      "step": 413000
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 7.034365985200051e-06,
+      "loss": 1.0938,
+      "step": 413500
+    },
+    {
+      "epoch": 10.39,
+      "learning_rate": 7.018688072243824e-06,
+      "loss": 1.1519,
+      "step": 414000
+    },
+    {
+      "epoch": 10.4,
+      "learning_rate": 7.003010159287596e-06,
+      "loss": 1.1585,
+      "step": 414500
+    },
+    {
+      "epoch": 10.41,
+      "learning_rate": 6.987332246331369e-06,
+      "loss": 1.1207,
+      "step": 415000
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 6.9716543333751415e-06,
+      "loss": 1.031,
+      "step": 415500
+    },
+    {
+      "epoch": 10.44,
+      "learning_rate": 6.955976420418914e-06,
+      "loss": 1.1214,
+      "step": 416000
+    },
+    {
+      "epoch": 10.45,
+      "learning_rate": 6.9402985074626876e-06,
+      "loss": 1.0939,
+      "step": 416500
+    },
+    {
+      "epoch": 10.46,
+      "learning_rate": 6.92462059450646e-06,
+      "loss": 1.1134,
+      "step": 417000
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 6.908942681550233e-06,
+      "loss": 1.1446,
+      "step": 417500
+    },
+    {
+      "epoch": 10.49,
+      "learning_rate": 6.893264768594005e-06,
+      "loss": 1.0843,
+      "step": 418000
+    },
+    {
+      "epoch": 10.5,
+      "learning_rate": 6.877586855637778e-06,
+      "loss": 1.1155,
+      "step": 418500
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 6.8619089426815514e-06,
+      "loss": 1.0844,
+      "step": 419000
+    },
+    {
+      "epoch": 10.52,
+      "learning_rate": 6.846231029725323e-06,
+      "loss": 1.1019,
+      "step": 419500
+    },
+    {
+      "epoch": 10.54,
+      "learning_rate": 6.830553116769097e-06,
+      "loss": 1.1204,
+      "step": 420000
+    },
+    {
+      "epoch": 10.55,
+      "learning_rate": 6.814875203812868e-06,
+      "loss": 1.1081,
+      "step": 420500
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 6.799197290856642e-06,
+      "loss": 1.1415,
+      "step": 421000
+    },
+    {
+      "epoch": 10.57,
+      "learning_rate": 6.783519377900415e-06,
+      "loss": 1.0147,
+      "step": 421500
+    },
+    {
+      "epoch": 10.59,
+      "learning_rate": 6.767841464944187e-06,
+      "loss": 1.1238,
+      "step": 422000
+    },
+    {
+      "epoch": 10.6,
+      "learning_rate": 6.7521635519879605e-06,
+      "loss": 1.0905,
+      "step": 422500
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 6.736485639031732e-06,
+      "loss": 1.0703,
+      "step": 423000
+    },
+    {
+      "epoch": 10.62,
+      "learning_rate": 6.720807726075506e-06,
+      "loss": 1.1379,
+      "step": 423500
+    },
+    {
+      "epoch": 10.64,
+      "learning_rate": 6.7051298131192775e-06,
+      "loss": 1.0753,
+      "step": 424000
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 6.689451900163051e-06,
+      "loss": 1.0929,
+      "step": 424500
+    },
+    {
+      "epoch": 10.66,
+      "learning_rate": 6.673773987206824e-06,
+      "loss": 1.1311,
+      "step": 425000
+    },
+    {
+      "epoch": 10.67,
+      "learning_rate": 6.658096074250596e-06,
+      "loss": 1.1616,
+      "step": 425500
+    },
+    {
+      "epoch": 10.69,
+      "learning_rate": 6.64241816129437e-06,
+      "loss": 1.0553,
+      "step": 426000
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 6.626740248338141e-06,
+      "loss": 1.126,
+      "step": 426500
+    },
+    {
+      "epoch": 10.71,
+      "learning_rate": 6.611062335381915e-06,
+      "loss": 1.1379,
+      "step": 427000
+    },
+    {
+      "epoch": 10.72,
+      "learning_rate": 6.595384422425687e-06,
+      "loss": 1.1454,
+      "step": 427500
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 6.57970650946946e-06,
+      "loss": 1.135,
+      "step": 428000
+    },
+    {
+      "epoch": 10.75,
+      "learning_rate": 6.564028596513232e-06,
+      "loss": 1.1442,
+      "step": 428500
+    },
+    {
+      "epoch": 10.76,
+      "learning_rate": 6.548350683557005e-06,
+      "loss": 1.051,
+      "step": 429000
+    },
+    {
+      "epoch": 10.77,
+      "learning_rate": 6.532672770600779e-06,
+      "loss": 1.0752,
+      "step": 429500
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 6.5169948576445505e-06,
+      "loss": 1.117,
+      "step": 430000
+    },
+    {
+      "epoch": 10.8,
+      "learning_rate": 6.501316944688324e-06,
+      "loss": 1.1297,
+      "step": 430500
+    },
+    {
+      "epoch": 10.81,
+      "learning_rate": 6.485639031732096e-06,
+      "loss": 1.0782,
+      "step": 431000
+    },
+    {
+      "epoch": 10.82,
+      "learning_rate": 6.469961118775869e-06,
+      "loss": 1.0986,
+      "step": 431500
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 6.454283205819642e-06,
+      "loss": 1.1466,
+      "step": 432000
+    },
+    {
+      "epoch": 10.85,
+      "learning_rate": 6.438605292863414e-06,
+      "loss": 1.0771,
+      "step": 432500
+    },
+    {
+      "epoch": 10.86,
+      "learning_rate": 6.422927379907187e-06,
+      "loss": 1.0713,
+      "step": 433000
+    },
+    {
+      "epoch": 10.87,
+      "learning_rate": 6.4072494669509596e-06,
+      "loss": 1.0749,
+      "step": 433500
+    },
+    {
+      "epoch": 10.89,
+      "learning_rate": 6.391571553994733e-06,
+      "loss": 1.1106,
+      "step": 434000
+    },
+    {
+      "epoch": 10.9,
+      "learning_rate": 6.375893641038506e-06,
+      "loss": 1.0764,
+      "step": 434500
+    },
+    {
+      "epoch": 10.91,
+      "learning_rate": 6.360215728082278e-06,
+      "loss": 1.1313,
+      "step": 435000
+    },
+    {
+      "epoch": 10.92,
+      "learning_rate": 6.344537815126051e-06,
+      "loss": 1.0613,
+      "step": 435500
+    },
+    {
+      "epoch": 10.94,
+      "learning_rate": 6.3288599021698234e-06,
+      "loss": 1.0666,
+      "step": 436000
+    },
+    {
+      "epoch": 10.95,
+      "learning_rate": 6.313181989213596e-06,
+      "loss": 1.1066,
+      "step": 436500
+    },
+    {
+      "epoch": 10.96,
+      "learning_rate": 6.2975040762573695e-06,
+      "loss": 1.1211,
+      "step": 437000
+    },
+    {
+      "epoch": 10.97,
+      "learning_rate": 6.281826163301142e-06,
+      "loss": 1.1016,
+      "step": 437500
+    },
+    {
+      "epoch": 10.99,
+      "learning_rate": 6.266148250344915e-06,
+      "loss": 1.1414,
+      "step": 438000
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 6.250470337388687e-06,
+      "loss": 1.0748,
+      "step": 438500
+    },
+    {
+      "epoch": 11.0,
+      "eval_loss": 1.1498711109161377,
+      "eval_runtime": 127.548,
+      "eval_samples_per_second": 277.825,
+      "eval_steps_per_second": 34.732,
+      "step": 438515
+    },
+    {
+      "epoch": 11.01,
+      "learning_rate": 6.23479242443246e-06,
+      "loss": 1.083,
+      "step": 439000
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 6.219114511476233e-06,
+      "loss": 1.0758,
+      "step": 439500
+    },
+    {
+      "epoch": 11.04,
+      "learning_rate": 6.203436598520005e-06,
+      "loss": 1.0222,
+      "step": 440000
+    },
+    {
+      "epoch": 11.05,
+      "learning_rate": 6.187758685563779e-06,
+      "loss": 1.0369,
+      "step": 440500
+    },
+    {
+      "epoch": 11.06,
+      "learning_rate": 6.17208077260755e-06,
+      "loss": 1.0377,
+      "step": 441000
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 6.156402859651324e-06,
+      "loss": 1.0744,
+      "step": 441500
+    },
+    {
+      "epoch": 11.09,
+      "learning_rate": 6.140724946695097e-06,
+      "loss": 1.0682,
+      "step": 442000
+    },
+    {
+      "epoch": 11.1,
+      "learning_rate": 6.125047033738869e-06,
+      "loss": 1.0682,
+      "step": 442500
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 6.1093691207826425e-06,
+      "loss": 1.0586,
+      "step": 443000
+    },
+    {
+      "epoch": 11.13,
+      "learning_rate": 6.093691207826414e-06,
+      "loss": 1.1158,
+      "step": 443500
+    },
+    {
+      "epoch": 11.14,
+      "learning_rate": 6.078013294870188e-06,
+      "loss": 1.1443,
+      "step": 444000
+    },
+    {
+      "epoch": 11.15,
+      "learning_rate": 6.0623353819139594e-06,
+      "loss": 1.0765,
+      "step": 444500
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 6.046657468957733e-06,
+      "loss": 1.0657,
+      "step": 445000
+    },
+    {
+      "epoch": 11.18,
+      "learning_rate": 6.030979556001506e-06,
+      "loss": 1.0981,
+      "step": 445500
+    },
+    {
+      "epoch": 11.19,
+      "learning_rate": 6.015301643045278e-06,
+      "loss": 1.0796,
+      "step": 446000
+    },
+    {
+      "epoch": 11.2,
+      "learning_rate": 5.9996237300890516e-06,
+      "loss": 1.075,
+      "step": 446500
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 5.983945817132823e-06,
+      "loss": 1.0764,
+      "step": 447000
+    },
+    {
+      "epoch": 11.23,
+      "learning_rate": 5.968267904176597e-06,
+      "loss": 1.0833,
+      "step": 447500
+    },
+    {
+      "epoch": 11.24,
+      "learning_rate": 5.9525899912203685e-06,
+      "loss": 1.0742,
+      "step": 448000
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 5.936912078264142e-06,
+      "loss": 1.0643,
+      "step": 448500
+    },
+    {
+      "epoch": 11.26,
+      "learning_rate": 5.921234165307915e-06,
+      "loss": 1.0914,
+      "step": 449000
+    },
+    {
+      "epoch": 11.28,
+      "learning_rate": 5.905556252351687e-06,
+      "loss": 1.1228,
+      "step": 449500
+    },
+    {
+      "epoch": 11.29,
+      "learning_rate": 5.889878339395461e-06,
+      "loss": 1.0618,
+      "step": 450000
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 5.874200426439232e-06,
+      "loss": 1.067,
+      "step": 450500
+    },
+    {
+      "epoch": 11.31,
+      "learning_rate": 5.858522513483006e-06,
+      "loss": 1.0584,
+      "step": 451000
+    },
+    {
+      "epoch": 11.33,
+      "learning_rate": 5.8428446005267785e-06,
+      "loss": 1.0778,
+      "step": 451500
+    },
+    {
+      "epoch": 11.34,
+      "learning_rate": 5.827166687570551e-06,
+      "loss": 1.1007,
+      "step": 452000
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 5.811488774614324e-06,
+      "loss": 1.1034,
+      "step": 452500
+    },
+    {
+      "epoch": 11.36,
+      "learning_rate": 5.795810861658096e-06,
+      "loss": 1.098,
+      "step": 453000
+    },
+    {
+      "epoch": 11.38,
+      "learning_rate": 5.78013294870187e-06,
+      "loss": 1.0907,
+      "step": 453500
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 5.7644550357456415e-06,
+      "loss": 1.0783,
+      "step": 454000
+    },
+    {
+      "epoch": 11.4,
+      "learning_rate": 5.748777122789415e-06,
+      "loss": 1.0586,
+      "step": 454500
+    },
+    {
+      "epoch": 11.41,
+      "learning_rate": 5.7330992098331876e-06,
+      "loss": 1.0824,
+      "step": 455000
+    },
+    {
+      "epoch": 11.43,
+      "learning_rate": 5.71742129687696e-06,
+      "loss": 1.0782,
+      "step": 455500
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 5.701743383920733e-06,
+      "loss": 1.1143,
+      "step": 456000
+    },
+    {
+      "epoch": 11.45,
+      "learning_rate": 5.686065470964505e-06,
+      "loss": 1.0858,
+      "step": 456500
+    },
+    {
+      "epoch": 11.46,
+      "learning_rate": 5.670387558008278e-06,
+      "loss": 1.0773,
+      "step": 457000
+    },
+    {
+      "epoch": 11.48,
+      "learning_rate": 5.6547096450520514e-06,
+      "loss": 1.0885,
+      "step": 457500
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 5.639031732095824e-06,
+      "loss": 1.0955,
+      "step": 458000
+    },
+    {
+      "epoch": 11.5,
+      "learning_rate": 5.623353819139597e-06,
+      "loss": 1.0341,
+      "step": 458500
+    },
+    {
+      "epoch": 11.51,
+      "learning_rate": 5.607675906183369e-06,
+      "loss": 1.0371,
+      "step": 459000
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 5.591997993227142e-06,
+      "loss": 1.0617,
+      "step": 459500
+    },
+    {
+      "epoch": 11.54,
+      "learning_rate": 5.576320080270915e-06,
+      "loss": 1.1171,
+      "step": 460000
+    },
+    {
+      "epoch": 11.55,
+      "learning_rate": 5.560642167314687e-06,
+      "loss": 1.0636,
+      "step": 460500
+    },
+    {
+      "epoch": 11.56,
+      "learning_rate": 5.5449642543584605e-06,
+      "loss": 1.1014,
+      "step": 461000
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 5.529286341402233e-06,
+      "loss": 1.0767,
+      "step": 461500
+    },
+    {
+      "epoch": 11.59,
+      "learning_rate": 5.513608428446006e-06,
+      "loss": 1.0649,
+      "step": 462000
+    },
+    {
+      "epoch": 11.6,
+      "learning_rate": 5.497930515489779e-06,
+      "loss": 1.0812,
+      "step": 462500
+    },
+    {
+      "epoch": 11.61,
+      "learning_rate": 5.482252602533551e-06,
+      "loss": 1.0922,
+      "step": 463000
+    },
+    {
+      "epoch": 11.63,
+      "learning_rate": 5.466574689577324e-06,
+      "loss": 1.1091,
+      "step": 463500
+    },
+    {
+      "epoch": 11.64,
+      "learning_rate": 5.450896776621096e-06,
+      "loss": 1.0437,
+      "step": 464000
+    },
+    {
+      "epoch": 11.65,
+      "learning_rate": 5.43521886366487e-06,
+      "loss": 1.0618,
+      "step": 464500
+    },
+    {
+      "epoch": 11.66,
+      "learning_rate": 5.419540950708641e-06,
+      "loss": 1.1011,
+      "step": 465000
+    },
+    {
+      "epoch": 11.68,
+      "learning_rate": 5.403863037752415e-06,
+      "loss": 1.1289,
+      "step": 465500
+    },
+    {
+      "epoch": 11.69,
+      "learning_rate": 5.388185124796188e-06,
+      "loss": 1.0708,
+      "step": 466000
+    },
+    {
+      "epoch": 11.7,
+      "learning_rate": 5.37250721183996e-06,
+      "loss": 1.1337,
+      "step": 466500
+    },
+    {
+      "epoch": 11.71,
+      "learning_rate": 5.3568292988837335e-06,
+      "loss": 1.0979,
+      "step": 467000
+    },
+    {
+      "epoch": 11.73,
+      "learning_rate": 5.341151385927505e-06,
+      "loss": 1.0715,
+      "step": 467500
+    },
+    {
+      "epoch": 11.74,
+      "learning_rate": 5.325473472971279e-06,
+      "loss": 1.067,
+      "step": 468000
+    },
+    {
+      "epoch": 11.75,
+      "learning_rate": 5.3097955600150505e-06,
+      "loss": 1.1121,
+      "step": 468500
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 1.101,
+      "step": 469000
+    },
+    {
+      "epoch": 11.78,
+      "learning_rate": 5.278439734102597e-06,
+      "loss": 1.0775,
+      "step": 469500
+    },
+    {
+      "epoch": 11.79,
+      "learning_rate": 5.262761821146369e-06,
+      "loss": 1.0705,
+      "step": 470000
+    },
+    {
+      "epoch": 11.8,
+      "learning_rate": 5.247083908190143e-06,
+      "loss": 1.0401,
+      "step": 470500
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 5.231405995233914e-06,
+      "loss": 1.0716,
+      "step": 471000
+    },
+    {
+      "epoch": 11.83,
+      "learning_rate": 5.215728082277688e-06,
+      "loss": 1.0586,
+      "step": 471500
+    },
+    {
+      "epoch": 11.84,
+      "learning_rate": 5.20005016932146e-06,
+      "loss": 1.0127,
+      "step": 472000
+    },
+    {
+      "epoch": 11.85,
+      "learning_rate": 5.184372256365233e-06,
+      "loss": 1.0542,
+      "step": 472500
+    },
+    {
+      "epoch": 11.87,
+      "learning_rate": 5.168694343409006e-06,
+      "loss": 1.1456,
+      "step": 473000
+    },
+    {
+      "epoch": 11.88,
+      "learning_rate": 5.153016430452778e-06,
+      "loss": 1.1196,
+      "step": 473500
+    },
+    {
+      "epoch": 11.89,
+      "learning_rate": 5.137338517496552e-06,
+      "loss": 1.0219,
+      "step": 474000
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 5.121660604540324e-06,
+      "loss": 1.0733,
+      "step": 474500
+    },
+    {
+      "epoch": 11.92,
+      "learning_rate": 5.105982691584097e-06,
+      "loss": 1.0688,
+      "step": 475000
+    },
+    {
+      "epoch": 11.93,
+      "learning_rate": 5.0903047786278695e-06,
+      "loss": 1.1228,
+      "step": 475500
+    },
+    {
+      "epoch": 11.94,
+      "learning_rate": 5.074626865671642e-06,
+      "loss": 1.1206,
+      "step": 476000
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 5.058948952715415e-06,
+      "loss": 1.0533,
+      "step": 476500
+    },
+    {
+      "epoch": 11.97,
+      "learning_rate": 5.043271039759188e-06,
+      "loss": 1.0524,
+      "step": 477000
+    },
+    {
+      "epoch": 11.98,
+      "learning_rate": 5.02759312680296e-06,
+      "loss": 1.0691,
+      "step": 477500
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 5.011915213846733e-06,
+      "loss": 1.0857,
+      "step": 478000
+    },
+    {
+      "epoch": 12.0,
+      "eval_loss": 1.1345129013061523,
+      "eval_runtime": 128.1857,
+      "eval_samples_per_second": 276.443,
+      "eval_steps_per_second": 34.559,
+      "step": 478380
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 4.996237300890506e-06,
+      "loss": 1.085,
+      "step": 478500
+    },
+    {
+      "epoch": 12.02,
+      "learning_rate": 4.9805593879342786e-06,
+      "loss": 1.0517,
+      "step": 479000
+    },
+    {
+      "epoch": 12.03,
+      "learning_rate": 4.964881474978051e-06,
+      "loss": 1.0369,
+      "step": 479500
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 4.949203562021825e-06,
+      "loss": 1.0236,
+      "step": 480000
+    },
+    {
+      "epoch": 12.05,
+      "learning_rate": 4.933525649065597e-06,
+      "loss": 1.1322,
+      "step": 480500
+    },
+    {
+      "epoch": 12.07,
+      "learning_rate": 4.91784773610937e-06,
+      "loss": 1.0517,
+      "step": 481000
+    },
+    {
+      "epoch": 12.08,
+      "learning_rate": 4.9021698231531425e-06,
+      "loss": 1.0192,
+      "step": 481500
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 4.886491910196915e-06,
+      "loss": 1.0468,
+      "step": 482000
+    },
+    {
+      "epoch": 12.1,
+      "learning_rate": 4.870813997240688e-06,
+      "loss": 1.0142,
+      "step": 482500
+    },
+    {
+      "epoch": 12.12,
+      "learning_rate": 4.85513608428446e-06,
+      "loss": 1.0511,
+      "step": 483000
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 4.839458171328233e-06,
+      "loss": 1.0379,
+      "step": 483500
+    },
+    {
+      "epoch": 12.14,
+      "learning_rate": 4.823780258372006e-06,
+      "loss": 1.0645,
+      "step": 484000
+    },
+    {
+      "epoch": 12.15,
+      "learning_rate": 4.808102345415779e-06,
+      "loss": 1.0567,
+      "step": 484500
+    },
+    {
+      "epoch": 12.17,
+      "learning_rate": 4.7924244324595515e-06,
+      "loss": 1.0486,
+      "step": 485000
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 4.776746519503324e-06,
+      "loss": 1.0741,
+      "step": 485500
+    },
+    {
+      "epoch": 12.19,
+      "learning_rate": 4.761068606547097e-06,
+      "loss": 1.0203,
+      "step": 486000
+    },
+    {
+      "epoch": 12.2,
+      "learning_rate": 4.745390693590869e-06,
+      "loss": 1.0649,
+      "step": 486500
+    },
+    {
+      "epoch": 12.22,
+      "learning_rate": 4.729712780634642e-06,
+      "loss": 1.0491,
+      "step": 487000
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 4.7140348676784146e-06,
+      "loss": 1.085,
+      "step": 487500
+    },
+    {
+      "epoch": 12.24,
+      "learning_rate": 4.698356954722188e-06,
+      "loss": 1.042,
+      "step": 488000
+    },
+    {
+      "epoch": 12.25,
+      "learning_rate": 4.682679041765961e-06,
+      "loss": 1.0425,
+      "step": 488500
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 4.667001128809733e-06,
+      "loss": 1.0799,
+      "step": 489000
+    },
+    {
+      "epoch": 12.28,
+      "learning_rate": 4.651323215853506e-06,
+      "loss": 1.0631,
+      "step": 489500
+    },
+    {
+      "epoch": 12.29,
+      "learning_rate": 4.6356453028972785e-06,
+      "loss": 1.102,
+      "step": 490000
+    },
+    {
+      "epoch": 12.3,
+      "learning_rate": 4.619967389941051e-06,
+      "loss": 1.0461,
+      "step": 490500
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 4.604289476984824e-06,
+      "loss": 1.0657,
+      "step": 491000
+    },
+    {
+      "epoch": 12.33,
+      "learning_rate": 4.588611564028596e-06,
+      "loss": 1.0682,
+      "step": 491500
+    },
+    {
+      "epoch": 12.34,
+      "learning_rate": 4.57293365107237e-06,
+      "loss": 1.0744,
+      "step": 492000
+    },
+    {
+      "epoch": 12.35,
+      "learning_rate": 4.557255738116142e-06,
+      "loss": 1.0515,
+      "step": 492500
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 4.541577825159915e-06,
+      "loss": 1.069,
+      "step": 493000
+    },
+    {
+      "epoch": 12.38,
+      "learning_rate": 4.5258999122036875e-06,
+      "loss": 1.0799,
+      "step": 493500
+    },
+    {
+      "epoch": 12.39,
+      "learning_rate": 4.51022199924746e-06,
+      "loss": 1.0657,
+      "step": 494000
+    },
+    {
+      "epoch": 12.4,
+      "learning_rate": 4.494544086291233e-06,
+      "loss": 1.0746,
+      "step": 494500
+    },
+    {
+      "epoch": 12.42,
+      "learning_rate": 4.478866173335006e-06,
+      "loss": 1.0562,
+      "step": 495000
+    },
+    {
+      "epoch": 12.43,
+      "learning_rate": 4.463188260378779e-06,
+      "loss": 1.0672,
+      "step": 495500
+    },
+    {
+      "epoch": 12.44,
+      "learning_rate": 4.447510347422551e-06,
+      "loss": 1.0975,
+      "step": 496000
+    },
+    {
+      "epoch": 12.45,
+      "learning_rate": 4.431832434466324e-06,
+      "loss": 1.0556,
+      "step": 496500
+    },
+    {
+      "epoch": 12.47,
+      "learning_rate": 4.416154521510097e-06,
+      "loss": 1.114,
+      "step": 497000
+    },
+    {
+      "epoch": 12.48,
+      "learning_rate": 4.40047660855387e-06,
+      "loss": 1.0487,
+      "step": 497500
+    },
+    {
+      "epoch": 12.49,
+      "learning_rate": 4.384798695597643e-06,
+      "loss": 1.0882,
+      "step": 498000
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 4.369120782641415e-06,
+      "loss": 1.0642,
+      "step": 498500
+    },
+    {
+      "epoch": 12.52,
+      "learning_rate": 4.353442869685188e-06,
+      "loss": 1.0573,
+      "step": 499000
+    },
+    {
+      "epoch": 12.53,
+      "learning_rate": 4.3377649567289605e-06,
+      "loss": 1.0843,
+      "step": 499500
+    },
+    {
+      "epoch": 12.54,
+      "learning_rate": 4.322087043772734e-06,
+      "loss": 1.0389,
+      "step": 500000
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 4.3064091308165066e-06,
+      "loss": 1.0181,
+      "step": 500500
+    },
+    {
+      "epoch": 12.57,
+      "learning_rate": 4.290731217860279e-06,
+      "loss": 1.066,
+      "step": 501000
+    },
+    {
+      "epoch": 12.58,
+      "learning_rate": 4.275053304904052e-06,
+      "loss": 1.0766,
+      "step": 501500
+    },
+    {
+      "epoch": 12.59,
+      "learning_rate": 4.259375391947824e-06,
+      "loss": 1.0449,
+      "step": 502000
+    },
+    {
+      "epoch": 12.61,
+      "learning_rate": 4.243697478991597e-06,
+      "loss": 1.0254,
+      "step": 502500
+    },
+    {
+      "epoch": 12.62,
+      "learning_rate": 4.22801956603537e-06,
+      "loss": 1.0405,
+      "step": 503000
+    },
+    {
+      "epoch": 12.63,
+      "learning_rate": 4.212341653079142e-06,
+      "loss": 1.0537,
+      "step": 503500
+    },
+    {
+      "epoch": 12.64,
+      "learning_rate": 4.196663740122916e-06,
+      "loss": 1.0986,
+      "step": 504000
+    },
+    {
+      "epoch": 12.66,
+      "learning_rate": 4.180985827166688e-06,
+      "loss": 1.0777,
+      "step": 504500
+    },
+    {
+      "epoch": 12.67,
+      "learning_rate": 4.165307914210461e-06,
+      "loss": 1.0084,
+      "step": 505000
+    },
+    {
+      "epoch": 12.68,
+      "learning_rate": 4.1496300012542335e-06,
+      "loss": 1.0684,
+      "step": 505500
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 4.133952088298006e-06,
+      "loss": 1.0387,
+      "step": 506000
+    },
+    {
+      "epoch": 12.71,
+      "learning_rate": 4.118274175341779e-06,
+      "loss": 1.0547,
+      "step": 506500
+    },
+    {
+      "epoch": 12.72,
+      "learning_rate": 4.102596262385551e-06,
+      "loss": 1.051,
+      "step": 507000
+    },
+    {
+      "epoch": 12.73,
+      "learning_rate": 4.086918349429324e-06,
+      "loss": 1.066,
+      "step": 507500
+    },
+    {
+      "epoch": 12.74,
+      "learning_rate": 4.071240436473097e-06,
+      "loss": 1.0669,
+      "step": 508000
+    },
+    {
+      "epoch": 12.76,
+      "learning_rate": 4.05556252351687e-06,
+      "loss": 1.0481,
+      "step": 508500
+    },
+    {
+      "epoch": 12.77,
+      "learning_rate": 4.0398846105606426e-06,
+      "loss": 1.0656,
+      "step": 509000
+    },
+    {
+      "epoch": 12.78,
+      "learning_rate": 4.024206697604415e-06,
+      "loss": 1.0604,
+      "step": 509500
+    },
+    {
+      "epoch": 12.79,
+      "learning_rate": 4.008528784648188e-06,
+      "loss": 1.067,
+      "step": 510000
+    },
+    {
+      "epoch": 12.81,
+      "learning_rate": 3.99285087169196e-06,
+      "loss": 1.0644,
+      "step": 510500
+    },
+    {
+      "epoch": 12.82,
+      "learning_rate": 3.977172958735733e-06,
+      "loss": 1.042,
+      "step": 511000
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 3.961495045779506e-06,
+      "loss": 1.0498,
+      "step": 511500
+    },
+    {
+      "epoch": 12.84,
+      "learning_rate": 3.945817132823279e-06,
+      "loss": 1.0805,
+      "step": 512000
+    },
+    {
+      "epoch": 12.86,
+      "learning_rate": 3.930139219867052e-06,
+      "loss": 1.0669,
+      "step": 512500
+    },
+    {
+      "epoch": 12.87,
+      "learning_rate": 3.914461306910824e-06,
+      "loss": 1.0549,
+      "step": 513000
+    },
+    {
+      "epoch": 12.88,
+      "learning_rate": 3.898783393954597e-06,
+      "loss": 1.0469,
+      "step": 513500
+    },
+    {
+      "epoch": 12.89,
+      "learning_rate": 3.8831054809983695e-06,
+      "loss": 1.0471,
+      "step": 514000
+    },
+    {
+      "epoch": 12.91,
+      "learning_rate": 3.867427568042142e-06,
+      "loss": 1.0423,
+      "step": 514500
+    },
+    {
+      "epoch": 12.92,
+      "learning_rate": 3.8517496550859155e-06,
+      "loss": 1.0984,
+      "step": 515000
+    },
+    {
+      "epoch": 12.93,
+      "learning_rate": 3.836071742129688e-06,
+      "loss": 1.0397,
+      "step": 515500
+    },
+    {
+      "epoch": 12.94,
+      "learning_rate": 3.820393829173461e-06,
+      "loss": 1.0844,
+      "step": 516000
+    },
+    {
+      "epoch": 12.96,
+      "learning_rate": 3.8047159162172338e-06,
+      "loss": 1.0314,
+      "step": 516500
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 3.7890380032610064e-06,
+      "loss": 1.078,
+      "step": 517000
+    },
+    {
+      "epoch": 12.98,
+      "learning_rate": 3.773360090304779e-06,
+      "loss": 1.0066,
+      "step": 517500
+    },
+    {
+      "epoch": 12.99,
+      "learning_rate": 3.7576821773485516e-06,
+      "loss": 1.0561,
+      "step": 518000
+    },
+    {
+      "epoch": 13.0,
+      "eval_loss": 1.1331528425216675,
+      "eval_runtime": 128.581,
+      "eval_samples_per_second": 275.593,
+      "eval_steps_per_second": 34.453,
+      "step": 518245
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 3.742004264392324e-06,
+      "loss": 1.0781,
+      "step": 518500
+    },
+    {
+      "epoch": 13.02,
+      "learning_rate": 3.726326351436097e-06,
+      "loss": 1.0371,
+      "step": 519000
+    },
+    {
+      "epoch": 13.03,
+      "learning_rate": 3.71064843847987e-06,
+      "loss": 1.0217,
+      "step": 519500
+    },
+    {
+      "epoch": 13.04,
+      "learning_rate": 3.6949705255236424e-06,
+      "loss": 1.0137,
+      "step": 520000
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 3.6792926125674155e-06,
+      "loss": 1.0569,
+      "step": 520500
+    },
+    {
+      "epoch": 13.07,
+      "learning_rate": 3.663614699611188e-06,
+      "loss": 1.0482,
+      "step": 521000
+    },
+    {
+      "epoch": 13.08,
+      "learning_rate": 3.6479367866549607e-06,
+      "loss": 1.0459,
+      "step": 521500
+    },
+    {
+      "epoch": 13.09,
+      "learning_rate": 3.6322588736987337e-06,
+      "loss": 0.9999,
+      "step": 522000
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 3.6165809607425063e-06,
+      "loss": 1.0628,
+      "step": 522500
+    },
+    {
+      "epoch": 13.12,
+      "learning_rate": 3.600903047786279e-06,
+      "loss": 1.0165,
+      "step": 523000
+    },
+    {
+      "epoch": 13.13,
+      "learning_rate": 3.5852251348300515e-06,
+      "loss": 1.0406,
+      "step": 523500
+    },
+    {
+      "epoch": 13.14,
+      "learning_rate": 3.569547221873824e-06,
+      "loss": 1.0782,
+      "step": 524000
+    },
+    {
+      "epoch": 13.16,
+      "learning_rate": 3.5538693089175976e-06,
+      "loss": 1.0609,
+      "step": 524500
+    },
+    {
+      "epoch": 13.17,
+      "learning_rate": 3.53819139596137e-06,
+      "loss": 1.0281,
+      "step": 525000
+    },
+    {
+      "epoch": 13.18,
+      "learning_rate": 3.522513483005143e-06,
+      "loss": 1.0175,
+      "step": 525500
+    },
+    {
+      "epoch": 13.19,
+      "learning_rate": 3.5068355700489154e-06,
+      "loss": 1.0347,
+      "step": 526000
+    },
+    {
+      "epoch": 13.21,
+      "learning_rate": 3.491157657092688e-06,
+      "loss": 1.0351,
+      "step": 526500
+    },
+    {
+      "epoch": 13.22,
+      "learning_rate": 3.4754797441364606e-06,
+      "loss": 1.0858,
+      "step": 527000
+    },
+    {
+      "epoch": 13.23,
+      "learning_rate": 3.4598018311802332e-06,
+      "loss": 1.0591,
+      "step": 527500
+    },
+    {
+      "epoch": 13.24,
+      "learning_rate": 3.4441239182240063e-06,
+      "loss": 1.0785,
+      "step": 528000
+    },
+    {
+      "epoch": 13.26,
+      "learning_rate": 3.4284460052677793e-06,
+      "loss": 1.065,
+      "step": 528500
+    },
+    {
+      "epoch": 13.27,
+      "learning_rate": 3.412768092311552e-06,
+      "loss": 1.0888,
+      "step": 529000
+    },
+    {
+      "epoch": 13.28,
+      "learning_rate": 3.3970901793553245e-06,
+      "loss": 1.029,
+      "step": 529500
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 3.381412266399097e-06,
+      "loss": 1.0326,
+      "step": 530000
+    },
+    {
+      "epoch": 13.31,
+      "learning_rate": 3.3657343534428697e-06,
+      "loss": 1.0316,
+      "step": 530500
+    },
+    {
+      "epoch": 13.32,
+      "learning_rate": 3.3500564404866427e-06,
+      "loss": 1.0677,
+      "step": 531000
+    },
+    {
+      "epoch": 13.33,
+      "learning_rate": 3.3343785275304153e-06,
+      "loss": 1.007,
+      "step": 531500
+    },
+    {
+      "epoch": 13.35,
+      "learning_rate": 3.318700614574188e-06,
+      "loss": 1.0707,
+      "step": 532000
+    },
+    {
+      "epoch": 13.36,
+      "learning_rate": 3.303022701617961e-06,
+      "loss": 1.047,
+      "step": 532500
+    },
+    {
+      "epoch": 13.37,
+      "learning_rate": 3.2873447886617336e-06,
+      "loss": 1.0405,
+      "step": 533000
+    },
+    {
+      "epoch": 13.38,
+      "learning_rate": 3.2716668757055066e-06,
+      "loss": 1.0474,
+      "step": 533500
+    },
+    {
+      "epoch": 13.4,
+      "learning_rate": 3.2559889627492792e-06,
+      "loss": 1.056,
+      "step": 534000
+    },
+    {
+      "epoch": 13.41,
+      "learning_rate": 3.240311049793052e-06,
+      "loss": 1.062,
+      "step": 534500
+    },
+    {
+      "epoch": 13.42,
+      "learning_rate": 3.2246331368368244e-06,
+      "loss": 1.03,
+      "step": 535000
+    },
+    {
+      "epoch": 13.43,
+      "learning_rate": 3.208955223880597e-06,
+      "loss": 1.0735,
+      "step": 535500
+    },
+    {
+      "epoch": 13.45,
+      "learning_rate": 3.1932773109243696e-06,
+      "loss": 1.0494,
+      "step": 536000
+    },
+    {
+      "epoch": 13.46,
+      "learning_rate": 3.177599397968143e-06,
+      "loss": 1.0271,
+      "step": 536500
+    },
+    {
+      "epoch": 13.47,
+      "learning_rate": 3.1619214850119157e-06,
+      "loss": 1.0283,
+      "step": 537000
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 3.1462435720556883e-06,
+      "loss": 1.0483,
+      "step": 537500
+    },
+    {
+      "epoch": 13.5,
+      "learning_rate": 3.130565659099461e-06,
+      "loss": 1.0449,
+      "step": 538000
+    },
+    {
+      "epoch": 13.51,
+      "learning_rate": 3.1148877461432335e-06,
+      "loss": 1.0583,
+      "step": 538500
+    },
+    {
+      "epoch": 13.52,
+      "learning_rate": 3.099209833187006e-06,
+      "loss": 1.035,
+      "step": 539000
+    },
+    {
+      "epoch": 13.53,
+      "learning_rate": 3.083531920230779e-06,
+      "loss": 1.0174,
+      "step": 539500
+    },
+    {
+      "epoch": 13.55,
+      "learning_rate": 3.0678540072745518e-06,
+      "loss": 1.0895,
+      "step": 540000
+    },
+    {
+      "epoch": 13.56,
+      "learning_rate": 3.052176094318325e-06,
+      "loss": 1.1153,
+      "step": 540500
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 3.0364981813620974e-06,
+      "loss": 1.0644,
+      "step": 541000
+    },
+    {
+      "epoch": 13.58,
+      "learning_rate": 3.02082026840587e-06,
+      "loss": 0.9903,
+      "step": 541500
+    },
+    {
+      "epoch": 13.6,
+      "learning_rate": 3.005142355449643e-06,
+      "loss": 1.0423,
+      "step": 542000
+    },
+    {
+      "epoch": 13.61,
+      "learning_rate": 2.9894644424934156e-06,
+      "loss": 1.068,
+      "step": 542500
+    },
+    {
+      "epoch": 13.62,
+      "learning_rate": 2.9737865295371883e-06,
+      "loss": 1.062,
+      "step": 543000
+    },
+    {
+      "epoch": 13.63,
+      "learning_rate": 2.958108616580961e-06,
+      "loss": 1.0714,
+      "step": 543500
+    },
+    {
+      "epoch": 13.65,
+      "learning_rate": 2.9424307036247335e-06,
+      "loss": 1.0567,
+      "step": 544000
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 2.9267527906685065e-06,
+      "loss": 1.0153,
+      "step": 544500
+    },
+    {
+      "epoch": 13.67,
+      "learning_rate": 2.9110748777122795e-06,
+      "loss": 0.9978,
+      "step": 545000
+    },
+    {
+      "epoch": 13.68,
+      "learning_rate": 2.895396964756052e-06,
+      "loss": 1.0683,
+      "step": 545500
+    },
+    {
+      "epoch": 13.7,
+      "learning_rate": 2.8797190517998247e-06,
+      "loss": 1.0301,
+      "step": 546000
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 2.8640411388435973e-06,
+      "loss": 1.0082,
+      "step": 546500
+    },
+    {
+      "epoch": 13.72,
+      "learning_rate": 2.84836322588737e-06,
+      "loss": 1.0835,
+      "step": 547000
+    },
+    {
+      "epoch": 13.73,
+      "learning_rate": 2.8326853129311426e-06,
+      "loss": 1.0267,
+      "step": 547500
+    },
+    {
+      "epoch": 13.75,
+      "learning_rate": 2.817007399974915e-06,
+      "loss": 0.9884,
+      "step": 548000
+    },
+    {
+      "epoch": 13.76,
+      "learning_rate": 2.8013294870186886e-06,
+      "loss": 1.0167,
+      "step": 548500
+    },
+    {
+      "epoch": 13.77,
+      "learning_rate": 2.7856515740624612e-06,
+      "loss": 0.9803,
+      "step": 549000
+    },
+    {
+      "epoch": 13.78,
+      "learning_rate": 2.769973661106234e-06,
+      "loss": 1.0467,
+      "step": 549500
+    },
+    {
+      "epoch": 13.8,
+      "learning_rate": 2.7542957481500064e-06,
+      "loss": 1.0797,
+      "step": 550000
+    },
+    {
+      "epoch": 13.81,
+      "learning_rate": 2.738617835193779e-06,
+      "loss": 1.0044,
+      "step": 550500
+    },
+    {
+      "epoch": 13.82,
+      "learning_rate": 2.722939922237552e-06,
+      "loss": 1.0395,
+      "step": 551000
+    },
+    {
+      "epoch": 13.83,
+      "learning_rate": 2.7072620092813247e-06,
+      "loss": 1.0939,
+      "step": 551500
+    },
+    {
+      "epoch": 13.85,
+      "learning_rate": 2.6915840963250973e-06,
+      "loss": 1.0077,
+      "step": 552000
+    },
+    {
+      "epoch": 13.86,
+      "learning_rate": 2.6759061833688703e-06,
+      "loss": 1.0878,
+      "step": 552500
+    },
+    {
+      "epoch": 13.87,
+      "learning_rate": 2.660228270412643e-06,
+      "loss": 1.0205,
+      "step": 553000
+    },
+    {
+      "epoch": 13.88,
+      "learning_rate": 2.644550357456416e-06,
+      "loss": 0.993,
+      "step": 553500
+    },
+    {
+      "epoch": 13.9,
+      "learning_rate": 2.6288724445001886e-06,
+      "loss": 1.0401,
+      "step": 554000
+    },
+    {
+      "epoch": 13.91,
+      "learning_rate": 2.613194531543961e-06,
+      "loss": 1.0286,
+      "step": 554500
+    },
+    {
+      "epoch": 13.92,
+      "learning_rate": 2.5975166185877338e-06,
+      "loss": 1.0799,
+      "step": 555000
+    },
+    {
+      "epoch": 13.93,
+      "learning_rate": 2.5818387056315064e-06,
+      "loss": 1.002,
+      "step": 555500
+    },
+    {
+      "epoch": 13.95,
+      "learning_rate": 2.566160792675279e-06,
+      "loss": 1.009,
+      "step": 556000
+    },
+    {
+      "epoch": 13.96,
+      "learning_rate": 2.5504828797190524e-06,
+      "loss": 1.029,
+      "step": 556500
+    },
+    {
+      "epoch": 13.97,
+      "learning_rate": 2.534804966762825e-06,
+      "loss": 1.0592,
+      "step": 557000
+    },
+    {
+      "epoch": 13.98,
+      "learning_rate": 2.5191270538065976e-06,
+      "loss": 1.0336,
+      "step": 557500
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 2.5034491408503702e-06,
+      "loss": 1.0431,
+      "step": 558000
+    },
+    {
+      "epoch": 14.0,
+      "eval_loss": 1.1127265691757202,
+      "eval_runtime": 127.2849,
+      "eval_samples_per_second": 278.399,
+      "eval_steps_per_second": 34.804,
+      "step": 558110
+    },
+    {
+      "epoch": 14.01,
+      "learning_rate": 2.487771227894143e-06,
+      "loss": 1.0293,
+      "step": 558500
+    },
+    {
+      "epoch": 14.02,
+      "learning_rate": 2.4720933149379155e-06,
+      "loss": 1.0318,
+      "step": 559000
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 2.456415401981688e-06,
+      "loss": 1.0244,
+      "step": 559500
+    },
+    {
+      "epoch": 14.05,
+      "learning_rate": 2.440737489025461e-06,
+      "loss": 1.0643,
+      "step": 560000
+    },
+    {
+      "epoch": 14.06,
+      "learning_rate": 2.4250595760692337e-06,
+      "loss": 1.0464,
+      "step": 560500
+    },
+    {
+      "epoch": 14.07,
+      "learning_rate": 2.4093816631130067e-06,
+      "loss": 1.0078,
+      "step": 561000
+    },
+    {
+      "epoch": 14.09,
+      "learning_rate": 2.3937037501567793e-06,
+      "loss": 1.0639,
+      "step": 561500
+    },
+    {
+      "epoch": 14.1,
+      "learning_rate": 2.378025837200552e-06,
+      "loss": 1.0297,
+      "step": 562000
+    },
+    {
+      "epoch": 14.11,
+      "learning_rate": 2.362347924244325e-06,
+      "loss": 1.0035,
+      "step": 562500
+    },
+    {
+      "epoch": 14.12,
+      "learning_rate": 2.3466700112880976e-06,
+      "loss": 1.0322,
+      "step": 563000
+    },
+    {
+      "epoch": 14.14,
+      "learning_rate": 2.33099209833187e-06,
+      "loss": 1.038,
+      "step": 563500
+    },
+    {
+      "epoch": 14.15,
+      "learning_rate": 2.3153141853756432e-06,
+      "loss": 0.9761,
+      "step": 564000
+    },
+    {
+      "epoch": 14.16,
+      "learning_rate": 2.299636272419416e-06,
+      "loss": 1.0188,
+      "step": 564500
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 2.2839583594631884e-06,
+      "loss": 1.0077,
+      "step": 565000
+    },
+    {
+      "epoch": 14.19,
+      "learning_rate": 2.268280446506961e-06,
+      "loss": 1.0246,
+      "step": 565500
+    },
+    {
+      "epoch": 14.2,
+      "learning_rate": 2.252602533550734e-06,
+      "loss": 1.0695,
+      "step": 566000
+    },
+    {
+      "epoch": 14.21,
+      "learning_rate": 2.2369246205945067e-06,
+      "loss": 1.0196,
+      "step": 566500
+    },
+    {
+      "epoch": 14.22,
+      "learning_rate": 2.2212467076382793e-06,
+      "loss": 1.0,
+      "step": 567000
+    },
+    {
+      "epoch": 14.24,
+      "learning_rate": 2.205568794682052e-06,
+      "loss": 1.0327,
+      "step": 567500
+    },
+    {
+      "epoch": 14.25,
+      "learning_rate": 2.189890881725825e-06,
+      "loss": 0.9988,
+      "step": 568000
+    },
+    {
+      "epoch": 14.26,
+      "learning_rate": 2.1742129687695975e-06,
+      "loss": 1.0017,
+      "step": 568500
+    },
+    {
+      "epoch": 14.27,
+      "learning_rate": 2.15853505581337e-06,
+      "loss": 1.0132,
+      "step": 569000
+    },
+    {
+      "epoch": 14.29,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 1.0325,
+      "step": 569500
+    },
+    {
+      "epoch": 14.3,
+      "learning_rate": 2.1271792299009158e-06,
+      "loss": 1.0412,
+      "step": 570000
+    },
+    {
+      "epoch": 14.31,
+      "learning_rate": 2.1115013169446884e-06,
+      "loss": 0.9906,
+      "step": 570500
+    },
+    {
+      "epoch": 14.32,
+      "learning_rate": 2.095823403988461e-06,
+      "loss": 0.9947,
+      "step": 571000
+    },
+    {
+      "epoch": 14.34,
+      "learning_rate": 2.080145491032234e-06,
+      "loss": 1.0673,
+      "step": 571500
+    },
+    {
+      "epoch": 14.35,
+      "learning_rate": 2.0644675780760066e-06,
+      "loss": 1.0175,
+      "step": 572000
+    },
+    {
+      "epoch": 14.36,
+      "learning_rate": 2.0487896651197796e-06,
+      "loss": 1.0722,
+      "step": 572500
+    },
+    {
+      "epoch": 14.37,
+      "learning_rate": 2.0331117521635522e-06,
+      "loss": 1.0239,
+      "step": 573000
+    },
+    {
+      "epoch": 14.39,
+      "learning_rate": 2.017433839207325e-06,
+      "loss": 1.0321,
+      "step": 573500
+    },
+    {
+      "epoch": 14.4,
+      "learning_rate": 2.001755926251098e-06,
+      "loss": 1.0703,
+      "step": 574000
+    },
+    {
+      "epoch": 14.41,
+      "learning_rate": 1.9860780132948705e-06,
+      "loss": 1.0156,
+      "step": 574500
+    },
+    {
+      "epoch": 14.42,
+      "learning_rate": 1.970400100338643e-06,
+      "loss": 1.0085,
+      "step": 575000
+    },
+    {
+      "epoch": 14.44,
+      "learning_rate": 1.9547221873824157e-06,
+      "loss": 1.0225,
+      "step": 575500
+    },
+    {
+      "epoch": 14.45,
+      "learning_rate": 1.9390442744261887e-06,
+      "loss": 1.0439,
+      "step": 576000
+    },
+    {
+      "epoch": 14.46,
+      "learning_rate": 1.9233663614699613e-06,
+      "loss": 1.032,
+      "step": 576500
+    },
+    {
+      "epoch": 14.47,
+      "learning_rate": 1.907688448513734e-06,
+      "loss": 1.0493,
+      "step": 577000
+    },
+    {
+      "epoch": 14.49,
+      "learning_rate": 1.8920105355575065e-06,
+      "loss": 1.0107,
+      "step": 577500
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 1.8763326226012796e-06,
+      "loss": 1.0462,
+      "step": 578000
+    },
+    {
+      "epoch": 14.51,
+      "learning_rate": 1.8606547096450522e-06,
+      "loss": 1.0023,
+      "step": 578500
+    },
+    {
+      "epoch": 14.52,
+      "learning_rate": 1.844976796688825e-06,
+      "loss": 1.0987,
+      "step": 579000
+    },
+    {
+      "epoch": 14.54,
+      "learning_rate": 1.8292988837325976e-06,
+      "loss": 1.0454,
+      "step": 579500
+    },
+    {
+      "epoch": 14.55,
+      "learning_rate": 1.8136209707763704e-06,
+      "loss": 1.0101,
+      "step": 580000
+    },
+    {
+      "epoch": 14.56,
+      "learning_rate": 1.7979430578201432e-06,
+      "loss": 1.0031,
+      "step": 580500
+    },
+    {
+      "epoch": 14.57,
+      "learning_rate": 1.7822651448639158e-06,
+      "loss": 1.0342,
+      "step": 581000
+    },
+    {
+      "epoch": 14.59,
+      "learning_rate": 1.7665872319076885e-06,
+      "loss": 1.0407,
+      "step": 581500
+    },
+    {
+      "epoch": 14.6,
+      "learning_rate": 1.7509093189514615e-06,
+      "loss": 1.005,
+      "step": 582000
+    },
+    {
+      "epoch": 14.61,
+      "learning_rate": 1.735231405995234e-06,
+      "loss": 1.0042,
+      "step": 582500
+    },
+    {
+      "epoch": 14.62,
+      "learning_rate": 1.7195534930390067e-06,
+      "loss": 1.0799,
+      "step": 583000
+    },
+    {
+      "epoch": 14.64,
+      "learning_rate": 1.7038755800827795e-06,
+      "loss": 1.018,
+      "step": 583500
+    },
+    {
+      "epoch": 14.65,
+      "learning_rate": 1.6881976671265523e-06,
+      "loss": 0.9983,
+      "step": 584000
+    },
+    {
+      "epoch": 14.66,
+      "learning_rate": 1.672519754170325e-06,
+      "loss": 1.0174,
+      "step": 584500
+    },
+    {
+      "epoch": 14.67,
+      "learning_rate": 1.6568418412140978e-06,
+      "loss": 1.0593,
+      "step": 585000
+    },
+    {
+      "epoch": 14.69,
+      "learning_rate": 1.6411639282578704e-06,
+      "loss": 1.0357,
+      "step": 585500
+    },
+    {
+      "epoch": 14.7,
+      "learning_rate": 1.6254860153016434e-06,
+      "loss": 1.0329,
+      "step": 586000
+    },
+    {
+      "epoch": 14.71,
+      "learning_rate": 1.609808102345416e-06,
+      "loss": 0.9979,
+      "step": 586500
+    },
+    {
+      "epoch": 14.72,
+      "learning_rate": 1.5941301893891886e-06,
+      "loss": 1.0639,
+      "step": 587000
+    },
+    {
+      "epoch": 14.74,
+      "learning_rate": 1.5784522764329612e-06,
+      "loss": 1.0678,
+      "step": 587500
+    },
+    {
+      "epoch": 14.75,
+      "learning_rate": 1.5627743634767342e-06,
+      "loss": 0.9992,
+      "step": 588000
+    },
+    {
+      "epoch": 14.76,
+      "learning_rate": 1.5470964505205068e-06,
+      "loss": 1.0042,
+      "step": 588500
+    },
+    {
+      "epoch": 14.77,
+      "learning_rate": 1.5314185375642795e-06,
+      "loss": 1.002,
+      "step": 589000
+    },
+    {
+      "epoch": 14.79,
+      "learning_rate": 1.5157406246080523e-06,
+      "loss": 1.0495,
+      "step": 589500
+    },
+    {
+      "epoch": 14.8,
+      "learning_rate": 1.500062711651825e-06,
+      "loss": 1.0445,
+      "step": 590000
+    },
+    {
+      "epoch": 14.81,
+      "learning_rate": 1.484384798695598e-06,
+      "loss": 1.0455,
+      "step": 590500
+    },
+    {
+      "epoch": 14.83,
+      "learning_rate": 1.4687068857393705e-06,
+      "loss": 1.0491,
+      "step": 591000
+    },
+    {
+      "epoch": 14.84,
+      "learning_rate": 1.4530289727831431e-06,
+      "loss": 1.0236,
+      "step": 591500
+    },
+    {
+      "epoch": 14.85,
+      "learning_rate": 1.4373510598269161e-06,
+      "loss": 1.0846,
+      "step": 592000
+    },
+    {
+      "epoch": 14.86,
+      "learning_rate": 1.4216731468706888e-06,
+      "loss": 1.0061,
+      "step": 592500
+    },
+    {
+      "epoch": 14.88,
+      "learning_rate": 1.4059952339144614e-06,
+      "loss": 0.9879,
+      "step": 593000
+    },
+    {
+      "epoch": 14.89,
+      "learning_rate": 1.3903173209582342e-06,
+      "loss": 1.1006,
+      "step": 593500
+    },
+    {
+      "epoch": 14.9,
+      "learning_rate": 1.374639408002007e-06,
+      "loss": 1.0208,
+      "step": 594000
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 1.3589614950457796e-06,
+      "loss": 0.9906,
+      "step": 594500
+    },
+    {
+      "epoch": 14.93,
+      "learning_rate": 1.3432835820895524e-06,
+      "loss": 1.0184,
+      "step": 595000
+    },
+    {
+      "epoch": 14.94,
+      "learning_rate": 1.327605669133325e-06,
+      "loss": 1.0255,
+      "step": 595500
+    },
+    {
+      "epoch": 14.95,
+      "learning_rate": 1.3119277561770978e-06,
+      "loss": 0.9888,
+      "step": 596000
+    },
+    {
+      "epoch": 14.96,
+      "learning_rate": 1.2962498432208707e-06,
+      "loss": 0.9886,
+      "step": 596500
+    },
+    {
+      "epoch": 14.98,
+      "learning_rate": 1.2805719302646433e-06,
+      "loss": 1.0307,
+      "step": 597000
+    },
+    {
+      "epoch": 14.99,
+      "learning_rate": 1.2648940173084159e-06,
+      "loss": 1.0273,
+      "step": 597500
+    },
+    {
+      "epoch": 15.0,
+      "eval_loss": 1.109073281288147,
+      "eval_runtime": 127.2061,
+      "eval_samples_per_second": 278.572,
+      "eval_steps_per_second": 34.825,
+      "step": 597975
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 1.2492161043521887e-06,
+      "loss": 1.0451,
+      "step": 598000
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 1.2335381913959615e-06,
+      "loss": 1.029,
+      "step": 598500
+    },
+    {
+      "epoch": 15.03,
+      "learning_rate": 1.2178602784397341e-06,
+      "loss": 0.9851,
+      "step": 599000
+    },
+    {
+      "epoch": 15.04,
+      "learning_rate": 1.202182365483507e-06,
+      "loss": 1.041,
+      "step": 599500
+    },
+    {
+      "epoch": 15.05,
+      "learning_rate": 1.1865044525272797e-06,
+      "loss": 1.0834,
+      "step": 600000
+    },
+    {
+      "epoch": 15.06,
+      "learning_rate": 1.1708265395710524e-06,
+      "loss": 1.0252,
+      "step": 600500
+    },
+    {
+      "epoch": 15.08,
+      "learning_rate": 1.1551486266148252e-06,
+      "loss": 0.9628,
+      "step": 601000
+    },
+    {
+      "epoch": 15.09,
+      "learning_rate": 1.1394707136585978e-06,
+      "loss": 0.9933,
+      "step": 601500
+    },
+    {
+      "epoch": 15.1,
+      "learning_rate": 1.1237928007023706e-06,
+      "loss": 0.9848,
+      "step": 602000
+    },
+    {
+      "epoch": 15.11,
+      "learning_rate": 1.1081148877461432e-06,
+      "loss": 1.027,
+      "step": 602500
+    },
+    {
+      "epoch": 15.13,
+      "learning_rate": 1.092436974789916e-06,
+      "loss": 0.9849,
+      "step": 603000
+    },
+    {
+      "epoch": 15.14,
+      "learning_rate": 1.0767590618336886e-06,
+      "loss": 0.9837,
+      "step": 603500
+    },
+    {
+      "epoch": 15.15,
+      "learning_rate": 1.0610811488774614e-06,
+      "loss": 1.0149,
+      "step": 604000
+    },
+    {
+      "epoch": 15.16,
+      "learning_rate": 1.0454032359212343e-06,
+      "loss": 1.0314,
+      "step": 604500
+    },
+    {
+      "epoch": 15.18,
+      "learning_rate": 1.029725322965007e-06,
+      "loss": 1.0298,
+      "step": 605000
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 1.0140474100087797e-06,
+      "loss": 1.0303,
+      "step": 605500
+    },
+    {
+      "epoch": 15.2,
+      "learning_rate": 9.983694970525525e-07,
+      "loss": 0.989,
+      "step": 606000
+    },
+    {
+      "epoch": 15.21,
+      "learning_rate": 9.826915840963251e-07,
+      "loss": 1.0632,
+      "step": 606500
+    },
+    {
+      "epoch": 15.23,
+      "learning_rate": 9.67013671140098e-07,
+      "loss": 0.9958,
+      "step": 607000
+    },
+    {
+      "epoch": 15.24,
+      "learning_rate": 9.513357581838705e-07,
+      "loss": 0.9956,
+      "step": 607500
+    },
+    {
+      "epoch": 15.25,
+      "learning_rate": 9.356578452276434e-07,
+      "loss": 1.017,
+      "step": 608000
+    },
+    {
+      "epoch": 15.26,
+      "learning_rate": 9.199799322714161e-07,
+      "loss": 1.0128,
+      "step": 608500
+    },
+    {
+      "epoch": 15.28,
+      "learning_rate": 9.043020193151889e-07,
+      "loss": 0.9813,
+      "step": 609000
+    },
+    {
+      "epoch": 15.29,
+      "learning_rate": 8.886241063589615e-07,
+      "loss": 0.998,
+      "step": 609500
+    },
+    {
+      "epoch": 15.3,
+      "learning_rate": 8.729461934027343e-07,
+      "loss": 0.9707,
+      "step": 610000
+    },
+    {
+      "epoch": 15.31,
+      "learning_rate": 8.57268280446507e-07,
+      "loss": 1.0107,
+      "step": 610500
+    },
+    {
+      "epoch": 15.33,
+      "learning_rate": 8.415903674902797e-07,
+      "loss": 1.0159,
+      "step": 611000
+    },
+    {
+      "epoch": 15.34,
+      "learning_rate": 8.259124545340524e-07,
+      "loss": 1.0025,
+      "step": 611500
+    },
+    {
+      "epoch": 15.35,
+      "learning_rate": 8.102345415778253e-07,
+      "loss": 0.9583,
+      "step": 612000
+    },
+    {
+      "epoch": 15.36,
+      "learning_rate": 7.945566286215979e-07,
+      "loss": 1.0236,
+      "step": 612500
+    },
+    {
+      "epoch": 15.38,
+      "learning_rate": 7.788787156653707e-07,
+      "loss": 1.0134,
+      "step": 613000
+    },
+    {
+      "epoch": 15.39,
+      "learning_rate": 7.632008027091434e-07,
+      "loss": 0.9746,
+      "step": 613500
+    },
+    {
+      "epoch": 15.4,
+      "learning_rate": 7.475228897529162e-07,
+      "loss": 0.9889,
+      "step": 614000
+    },
+    {
+      "epoch": 15.41,
+      "learning_rate": 7.318449767966888e-07,
+      "loss": 1.0125,
+      "step": 614500
+    },
+    {
+      "epoch": 15.43,
+      "learning_rate": 7.161670638404616e-07,
+      "loss": 1.0084,
+      "step": 615000
+    },
+    {
+      "epoch": 15.44,
+      "learning_rate": 7.004891508842344e-07,
+      "loss": 1.043,
+      "step": 615500
+    },
+    {
+      "epoch": 15.45,
+      "learning_rate": 6.848112379280071e-07,
+      "loss": 0.9813,
+      "step": 616000
+    },
+    {
+      "epoch": 15.46,
+      "learning_rate": 6.691333249717798e-07,
+      "loss": 1.0617,
+      "step": 616500
+    },
+    {
+      "epoch": 15.48,
+      "learning_rate": 6.534554120155526e-07,
+      "loss": 1.0439,
+      "step": 617000
+    },
+    {
+      "epoch": 15.49,
+      "learning_rate": 6.377774990593252e-07,
+      "loss": 0.9883,
+      "step": 617500
+    },
+    {
+      "epoch": 15.5,
+      "learning_rate": 6.22099586103098e-07,
+      "loss": 1.0232,
+      "step": 618000
+    },
+    {
+      "epoch": 15.51,
+      "learning_rate": 6.064216731468707e-07,
+      "loss": 1.086,
+      "step": 618500
+    },
+    {
+      "epoch": 15.53,
+      "learning_rate": 5.907437601906435e-07,
+      "loss": 1.0029,
+      "step": 619000
+    },
+    {
+      "epoch": 15.54,
+      "learning_rate": 5.750658472344163e-07,
+      "loss": 0.9831,
+      "step": 619500
+    },
+    {
+      "epoch": 15.55,
+      "learning_rate": 5.59387934278189e-07,
+      "loss": 1.0251,
+      "step": 620000
+    },
+    {
+      "epoch": 15.57,
+      "learning_rate": 5.437100213219617e-07,
+      "loss": 0.9849,
+      "step": 620500
+    },
+    {
+      "epoch": 15.58,
+      "learning_rate": 5.280321083657344e-07,
+      "loss": 0.9952,
+      "step": 621000
+    },
+    {
+      "epoch": 15.59,
+      "learning_rate": 5.123541954095072e-07,
+      "loss": 1.0433,
+      "step": 621500
+    },
+    {
+      "epoch": 15.6,
+      "learning_rate": 4.966762824532798e-07,
+      "loss": 1.0284,
+      "step": 622000
+    },
+    {
+      "epoch": 15.62,
+      "learning_rate": 4.809983694970525e-07,
+      "loss": 1.0445,
+      "step": 622500
+    },
+    {
+      "epoch": 15.63,
+      "learning_rate": 4.653204565408253e-07,
+      "loss": 0.9913,
+      "step": 623000
+    },
+    {
+      "epoch": 15.64,
+      "learning_rate": 4.49642543584598e-07,
+      "loss": 1.0422,
+      "step": 623500
+    },
+    {
+      "epoch": 15.65,
+      "learning_rate": 4.3396463062837077e-07,
+      "loss": 0.9829,
+      "step": 624000
+    },
+    {
+      "epoch": 15.67,
+      "learning_rate": 4.182867176721435e-07,
+      "loss": 1.0181,
+      "step": 624500
+    },
+    {
+      "epoch": 15.68,
+      "learning_rate": 4.026088047159162e-07,
+      "loss": 0.985,
+      "step": 625000
+    },
+    {
+      "epoch": 15.69,
+      "learning_rate": 3.8693089175968896e-07,
+      "loss": 0.9835,
+      "step": 625500
+    },
+    {
+      "epoch": 15.7,
+      "learning_rate": 3.712529788034617e-07,
+      "loss": 1.0481,
+      "step": 626000
+    },
+    {
+      "epoch": 15.72,
+      "learning_rate": 3.5557506584723444e-07,
+      "loss": 1.002,
+      "step": 626500
+    },
+    {
+      "epoch": 15.73,
+      "learning_rate": 3.3989715289100715e-07,
+      "loss": 1.0452,
+      "step": 627000
+    },
+    {
+      "epoch": 15.74,
+      "learning_rate": 3.2421923993477986e-07,
+      "loss": 1.0042,
+      "step": 627500
+    },
+    {
+      "epoch": 15.75,
+      "learning_rate": 3.0854132697855263e-07,
+      "loss": 0.9587,
+      "step": 628000
+    },
+    {
+      "epoch": 15.77,
+      "learning_rate": 2.928634140223254e-07,
+      "loss": 1.0267,
+      "step": 628500
+    },
+    {
+      "epoch": 15.78,
+      "learning_rate": 2.771855010660981e-07,
+      "loss": 0.9982,
+      "step": 629000
+    },
+    {
+      "epoch": 15.79,
+      "learning_rate": 2.6150758810987087e-07,
+      "loss": 1.0544,
+      "step": 629500
+    },
+    {
+      "epoch": 15.8,
+      "learning_rate": 2.458296751536436e-07,
+      "loss": 1.0435,
+      "step": 630000
+    },
+    {
+      "epoch": 15.82,
+      "learning_rate": 2.3015176219741632e-07,
+      "loss": 1.0338,
+      "step": 630500
+    },
+    {
+      "epoch": 15.83,
+      "learning_rate": 2.1447384924118906e-07,
+      "loss": 1.0075,
+      "step": 631000
+    },
+    {
+      "epoch": 15.84,
+      "learning_rate": 1.9879593628496177e-07,
+      "loss": 1.0073,
+      "step": 631500
+    },
+    {
+      "epoch": 15.85,
+      "learning_rate": 1.831180233287345e-07,
+      "loss": 0.9963,
+      "step": 632000
+    },
+    {
+      "epoch": 15.87,
+      "learning_rate": 1.6744011037250722e-07,
+      "loss": 1.0126,
+      "step": 632500
+    },
+    {
+      "epoch": 15.88,
+      "learning_rate": 1.5176219741627996e-07,
+      "loss": 1.034,
+      "step": 633000
+    },
+    {
+      "epoch": 15.89,
+      "learning_rate": 1.360842844600527e-07,
+      "loss": 1.0527,
+      "step": 633500
+    },
+    {
+      "epoch": 15.9,
+      "learning_rate": 1.2040637150382544e-07,
+      "loss": 1.0447,
+      "step": 634000
+    },
+    {
+      "epoch": 15.92,
+      "learning_rate": 1.0472845854759816e-07,
+      "loss": 0.9632,
+      "step": 634500
+    },
+    {
+      "epoch": 15.93,
+      "learning_rate": 8.905054559137089e-08,
+      "loss": 1.0167,
+      "step": 635000
+    },
+    {
+      "epoch": 15.94,
+      "learning_rate": 7.337263263514361e-08,
+      "loss": 1.0249,
+      "step": 635500
+    },
+    {
+      "epoch": 15.95,
+      "learning_rate": 5.7694719678916344e-08,
+      "loss": 0.9865,
+      "step": 636000
+    },
+    {
+      "epoch": 15.97,
+      "learning_rate": 4.2016806722689076e-08,
+      "loss": 0.982,
+      "step": 636500
+    },
+    {
+      "epoch": 15.98,
+      "learning_rate": 2.633889376646181e-08,
+      "loss": 1.0187,
+      "step": 637000
+    },
+    {
+      "epoch": 15.99,
+      "learning_rate": 1.0660980810234543e-08,
+      "loss": 1.0275,
+      "step": 637500
+    },
+    {
+      "epoch": 16.0,
+      "eval_loss": 1.1107752323150635,
+      "eval_runtime": 128.1911,
+      "eval_samples_per_second": 276.431,
+      "eval_steps_per_second": 34.558,
+      "step": 637840
+    },
+    {
+      "epoch": 16.0,
+      "step": 637840,
+      "total_flos": 1.0362059176376341e+17,
+      "train_loss": 1.2091509097166406,
+      "train_runtime": 77632.0272,
+      "train_samples_per_second": 65.729,
+      "train_steps_per_second": 8.216
+    }
+  ],
+  "max_steps": 637840,
+  "num_train_epochs": 16,
+  "total_flos": 1.0362059176376341e+17,
+  "trial_name": null,
+  "trial_params": null
+}