diff --git "a/checkpoint-1068480/trainer_state.json" "b/checkpoint-1068480/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1068480/trainer_state.json"
@@ -0,0 +1,12653 @@
+{
+  "best_metric": 4.030486583709717,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-sent-neg/lstm/4/checkpoints/checkpoint-1068480",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1068480,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8194,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5555,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0629,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9948,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9484,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9278,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7608,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6496,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5525,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4704,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.4073,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3344,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.264,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.1874,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.145,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0865,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.046,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 6.0009,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9637,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.9208,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8886,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8482,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.815,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7832,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7639,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7265,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7032,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6747,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6451,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6353,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6097,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6019,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5645,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5545,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714926921053515e-05,
+      "loss": 5.543,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5257,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5043,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.4693,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9681399509815164e-05,
+      "loss": 5.4704,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4375,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4397,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4237,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4131,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.3931,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963113296116324e-05,
+      "loss": 5.3855,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 5.3868,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3685,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3637,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3361,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 5.3214,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 5.3302,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3169,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3086,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 5.2772,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 5.2927,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.2669,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2899,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2405,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2513,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2317,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2205,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 5.2213,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2083,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.1865,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.1881,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.2083,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1823,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1697,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1589,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1448,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 5.1595,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1502,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1441,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1498,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 5.133,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 5.1283,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1067,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1071,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.0975,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 5.0885,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 5.09,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 5.0854,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 5.0892,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930421203868281e-05,
+      "loss": 5.0788,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 5.0532,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0626,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.0632,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.0658,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0438,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.047,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92455267849129e-05,
+      "loss": 5.0392,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923714083740238e-05,
+      "loss": 5.0301,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0257,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0213,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921199937367455e-05,
+      "loss": 5.0103,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 5.0184,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919522747865351e-05,
+      "loss": 4.9993,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9913,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.9878,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9927,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9871,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9716,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9728,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.9606,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.9649,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9415,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9502,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910303119244898e-05,
+      "loss": 4.9441,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909464524493846e-05,
+      "loss": 4.9394,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.945,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.9373,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.9158,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906111783370011e-05,
+      "loss": 4.9256,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905273188618959e-05,
+      "loss": 4.9286,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.9201,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9189,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.9105,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9065,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901081852744072e-05,
+      "loss": 4.8963,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90024325799302e-05,
+      "loss": 4.903,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994046632419686e-05,
+      "loss": 4.8981,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.8815,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.884,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.8807,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.8879,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.8639,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8696,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.8574,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926991809942984e-05,
+      "loss": 4.8635,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918605862432463e-05,
+      "loss": 4.8539,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.868,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.8553,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8486,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8468,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8417,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.849,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.8482,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.8386,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.8341,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.8345,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.8244,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.8123,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.8194,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.8107,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.8027,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.8034,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.802,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.8046,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.8055,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7935,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.759880065917969,
+      "eval_runtime": 288.7522,
+      "eval_samples_per_second": 1321.517,
+      "eval_steps_per_second": 41.298,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.7832,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.78,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.7903,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.7835,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7866,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.7688,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7701,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7637,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7685,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7684,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7641,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7716,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.7404,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.75,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.7438,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.7366,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860844045616934e-05,
+      "loss": 4.7479,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.7386,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.7363,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858328261363778e-05,
+      "loss": 4.7561,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.7368,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.7311,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7315,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.745,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7232,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.7195,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7189,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7161,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.7149,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.7084,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.7274,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.7084,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.7063,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.7086,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.7068,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7016,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6822,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6885,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6758,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.692,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6806,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.6865,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6758,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838208538860022e-05,
+      "loss": 4.6808,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6842,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.6768,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.6789,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.6673,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834015565104763e-05,
+      "loss": 4.6592,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833176970353711e-05,
+      "loss": 4.6713,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832338375602659e-05,
+      "loss": 4.6646,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831499780851606e-05,
+      "loss": 4.6707,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6418,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.6645,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.6422,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.6748,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.639,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.6544,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.6413,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247942986039366e-05,
+      "loss": 4.6347,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239557038528846e-05,
+      "loss": 4.6404,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6396,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6207,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6273,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206029627290495e-05,
+      "loss": 4.6431,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.6295,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.6252,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.6219,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8172502216052144e-05,
+      "loss": 4.611,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8164116268541624e-05,
+      "loss": 4.6308,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8155730321031104e-05,
+      "loss": 4.6249,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147344373520584e-05,
+      "loss": 4.6255,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138958426010064e-05,
+      "loss": 4.6324,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813057247849955e-05,
+      "loss": 4.626,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812218653098903e-05,
+      "loss": 4.6253,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811381696228224e-05,
+      "loss": 4.6037,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810543101477172e-05,
+      "loss": 4.6155,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80970450672612e-05,
+      "loss": 4.6094,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808865911975068e-05,
+      "loss": 4.5979,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808027317224016e-05,
+      "loss": 4.6072,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807190360353337e-05,
+      "loss": 4.6086,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.6179,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.6098,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.5887,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.5977,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.6053,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.6099,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.5981,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800484878105667e-05,
+      "loss": 4.6026,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799646283354615e-05,
+      "loss": 4.5991,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798807688603563e-05,
+      "loss": 4.5874,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797969093852511e-05,
+      "loss": 4.5933,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797130499101459e-05,
+      "loss": 4.5928,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796291904350407e-05,
+      "loss": 4.5811,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795454947479728e-05,
+      "loss": 4.5972,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794616352728676e-05,
+      "loss": 4.5794,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793777757977624e-05,
+      "loss": 4.5767,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792939163226572e-05,
+      "loss": 4.5797,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5849,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5841,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.5715,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789586422102737e-05,
+      "loss": 4.571,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788747827351685e-05,
+      "loss": 4.566,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787909232600633e-05,
+      "loss": 4.576,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5517,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786233680978902e-05,
+      "loss": 4.5601,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78539508622785e-05,
+      "loss": 4.5677,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784556491476798e-05,
+      "loss": 4.5615,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783717896725746e-05,
+      "loss": 4.5632,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782879301974694e-05,
+      "loss": 4.5613,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782040707223642e-05,
+      "loss": 4.5502,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812037503529635e-05,
+      "loss": 4.5603,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803651556019115e-05,
+      "loss": 4.5575,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795265608508595e-05,
+      "loss": 4.5573,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786896039801804e-05,
+      "loss": 4.5603,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778510092291284e-05,
+      "loss": 4.561,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777014052358449e-05,
+      "loss": 4.5512,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776175457607397e-05,
+      "loss": 4.5452,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775336862856345e-05,
+      "loss": 4.5537,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774498268105293e-05,
+      "loss": 4.5489,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773659673354241e-05,
+      "loss": 4.5363,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772821078603189e-05,
+      "loss": 4.5463,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771982483852137e-05,
+      "loss": 4.5401,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771143889101086e-05,
+      "loss": 4.5505,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770305294350034e-05,
+      "loss": 4.5322,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769466699598982e-05,
+      "loss": 4.5409,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5198,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767791147977251e-05,
+      "loss": 4.5419,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766952553226199e-05,
+      "loss": 4.5292,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766113958475147e-05,
+      "loss": 4.5462,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765275363724095e-05,
+      "loss": 4.5375,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764436768973043e-05,
+      "loss": 4.5346,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76359817422199e-05,
+      "loss": 4.5253,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762759579470938e-05,
+      "loss": 4.5271,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761920984719886e-05,
+      "loss": 4.5408,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761082389968834e-05,
+      "loss": 4.5423,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760243795217783e-05,
+      "loss": 4.531,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759406838347104e-05,
+      "loss": 4.5288,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758568243596052e-05,
+      "loss": 4.5248,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757729648845e-05,
+      "loss": 4.5235,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756891054093948e-05,
+      "loss": 4.5167,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756052459342896e-05,
+      "loss": 4.523,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552155024722166e-05,
+      "loss": 4.5166,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543769077211646e-05,
+      "loss": 4.511,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535383129701126e-05,
+      "loss": 4.5155,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7526997182190606e-05,
+      "loss": 4.5132,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.516,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510241665973295e-05,
+      "loss": 4.5166,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750185571846278e-05,
+      "loss": 4.5123,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.480078220367432,
+      "eval_runtime": 293.8973,
+      "eval_samples_per_second": 1298.382,
+      "eval_steps_per_second": 40.575,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749346977095226e-05,
+      "loss": 4.4999,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748508382344174e-05,
+      "loss": 4.5012,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747669787593122e-05,
+      "loss": 4.5111,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.5087,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.5131,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745154003339966e-05,
+      "loss": 4.491,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744315408588914e-05,
+      "loss": 4.5024,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743476813837862e-05,
+      "loss": 4.4928,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426382190868095e-05,
+      "loss": 4.503,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417996243357575e-05,
+      "loss": 4.501,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409610295847055e-05,
+      "loss": 4.5012,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401224348336535e-05,
+      "loss": 4.5044,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392838400826015e-05,
+      "loss": 4.4784,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384452453315495e-05,
+      "loss": 4.4877,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376066505804975e-05,
+      "loss": 4.4891,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736769693709819e-05,
+      "loss": 4.481,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735931098958767e-05,
+      "loss": 4.4907,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735092504207715e-05,
+      "loss": 4.4839,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734253909456663e-05,
+      "loss": 4.4846,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733416952585984e-05,
+      "loss": 4.5107,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732578357834932e-05,
+      "loss": 4.4839,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73173976308388e-05,
+      "loss": 4.4864,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730901168332828e-05,
+      "loss": 4.4832,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730062573581776e-05,
+      "loss": 4.5003,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729223978830724e-05,
+      "loss": 4.4769,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728385384079672e-05,
+      "loss": 4.4766,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.72754678932862e-05,
+      "loss": 4.4794,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726708194577568e-05,
+      "loss": 4.4796,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725869599826516e-05,
+      "loss": 4.4702,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725031005075464e-05,
+      "loss": 4.4744,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724192410324412e-05,
+      "loss": 4.4883,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233554534537335e-05,
+      "loss": 4.4762,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.473,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216799018320024e-05,
+      "loss": 4.4788,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208413070809504e-05,
+      "loss": 4.4739,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7200027123298984e-05,
+      "loss": 4.473,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719164117578846e-05,
+      "loss": 4.4539,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718325522827794e-05,
+      "loss": 4.4694,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717486928076742e-05,
+      "loss": 4.4522,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71664833332569e-05,
+      "loss": 4.4638,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715811376455011e-05,
+      "loss": 4.4591,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714972781703959e-05,
+      "loss": 4.4654,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714134186952907e-05,
+      "loss": 4.4572,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713297230082228e-05,
+      "loss": 4.4586,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712458635331176e-05,
+      "loss": 4.4646,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711620040580124e-05,
+      "loss": 4.461,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710781445829072e-05,
+      "loss": 4.4642,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70994285107802e-05,
+      "loss": 4.4502,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709104256326968e-05,
+      "loss": 4.4419,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708265661575916e-05,
+      "loss": 4.4615,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707427066824864e-05,
+      "loss": 4.4524,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706590109954185e-05,
+      "loss": 4.456,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057531530835066e-05,
+      "loss": 4.436,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049145583324546e-05,
+      "loss": 4.4601,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040759635814026e-05,
+      "loss": 4.4338,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032373688303506e-05,
+      "loss": 4.4647,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7023987740792986e-05,
+      "loss": 4.4263,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4501,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4392,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4375,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.4361,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982074382044115e-05,
+      "loss": 4.4392,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4219,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4283,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695694929712002e-05,
+      "loss": 4.4393,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.4355,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4275,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693179145458846e-05,
+      "loss": 4.4275,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692340550707794e-05,
+      "loss": 4.422,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691501955956742e-05,
+      "loss": 4.4306,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69066336120569e-05,
+      "loss": 4.4322,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689824766454638e-05,
+      "loss": 4.431,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688987809583959e-05,
+      "loss": 4.4411,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688149214832907e-05,
+      "loss": 4.4415,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687310620081855e-05,
+      "loss": 4.4346,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686472025330803e-05,
+      "loss": 4.4154,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685633430579751e-05,
+      "loss": 4.4257,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847964737090725e-05,
+      "loss": 4.4296,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839578789580204e-05,
+      "loss": 4.4112,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831192842069684e-05,
+      "loss": 4.4243,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822823273362894e-05,
+      "loss": 4.4227,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814437325852373e-05,
+      "loss": 4.4338,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806051378341853e-05,
+      "loss": 4.4263,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6797665430831333e-05,
+      "loss": 4.4071,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678927948332081e-05,
+      "loss": 4.4194,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6780893535810287e-05,
+      "loss": 4.4231,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6772507588299766e-05,
+      "loss": 4.4282,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764121640789246e-05,
+      "loss": 4.4195,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755735693278726e-05,
+      "loss": 4.4247,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747349745768206e-05,
+      "loss": 4.4231,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673896379825769e-05,
+      "loss": 4.4074,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67305942295509e-05,
+      "loss": 4.4211,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672220828204038e-05,
+      "loss": 4.415,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671383871333359e-05,
+      "loss": 4.406,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670545276582307e-05,
+      "loss": 4.4267,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669706681831255e-05,
+      "loss": 4.4087,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668868087080203e-05,
+      "loss": 4.4014,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668029492329151e-05,
+      "loss": 4.4094,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667190897578099e-05,
+      "loss": 4.415,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666352302827047e-05,
+      "loss": 4.4123,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665513708075995e-05,
+      "loss": 4.4033,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664676751205316e-05,
+      "loss": 4.399,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663838156454265e-05,
+      "loss": 4.3981,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662999561703213e-05,
+      "loss": 4.4104,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662160966952161e-05,
+      "loss": 4.3852,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613240100814816e-05,
+      "loss": 4.3977,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604854153304296e-05,
+      "loss": 4.4034,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596468205793776e-05,
+      "loss": 4.396,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588082258283256e-05,
+      "loss": 4.3973,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579712689576465e-05,
+      "loss": 4.4002,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571326742065945e-05,
+      "loss": 4.3905,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562940794555425e-05,
+      "loss": 4.3972,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554554847044905e-05,
+      "loss": 4.3957,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546185278338114e-05,
+      "loss": 4.3951,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65377993308276e-05,
+      "loss": 4.4017,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652941338331708e-05,
+      "loss": 4.4006,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652102743580656e-05,
+      "loss": 4.3979,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651264148829604e-05,
+      "loss": 4.3855,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650425554078552e-05,
+      "loss": 4.3925,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495869593275e-05,
+      "loss": 4.3896,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648748364576447e-05,
+      "loss": 4.3782,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647909769825395e-05,
+      "loss": 4.395,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64707445083509e-05,
+      "loss": 4.38,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646235856084038e-05,
+      "loss": 4.3947,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645397261332986e-05,
+      "loss": 4.3772,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.3878,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.3712,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64288147707983e-05,
+      "loss": 4.3888,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642042882328778e-05,
+      "loss": 4.375,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641204287577726e-05,
+      "loss": 4.3969,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640365692826674e-05,
+      "loss": 4.3902,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639527098075622e-05,
+      "loss": 4.3834,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63868850332457e-05,
+      "loss": 4.3727,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637849908573518e-05,
+      "loss": 4.3781,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637014589583212e-05,
+      "loss": 4.3909,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.3954,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353374000811076e-05,
+      "loss": 4.3807,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6344988053300556e-05,
+      "loss": 4.3839,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633661848459377e-05,
+      "loss": 4.3791,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632823253708325e-05,
+      "loss": 4.3786,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631984658957273e-05,
+      "loss": 4.3708,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631146064206221e-05,
+      "loss": 4.3751,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630307469455169e-05,
+      "loss": 4.3764,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629468874704117e-05,
+      "loss": 4.3685,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628630279953065e-05,
+      "loss": 4.3698,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627791685202013e-05,
+      "loss": 4.3756,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626954728331334e-05,
+      "loss": 4.3685,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626116133580282e-05,
+      "loss": 4.374,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62527753882923e-05,
+      "loss": 4.3681,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.345285892486572,
+      "eval_runtime": 293.9389,
+      "eval_samples_per_second": 1298.198,
+      "eval_steps_per_second": 40.57,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624438944078178e-05,
+      "loss": 4.3662,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.3587,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.3754,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219247977053956e-05,
+      "loss": 4.3659,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210862029543436e-05,
+      "loss": 4.3744,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.3529,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194122892129854e-05,
+      "loss": 4.3627,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3526,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177350997108814e-05,
+      "loss": 4.37,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.3637,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160579102087774e-05,
+      "loss": 4.3657,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152193154577254e-05,
+      "loss": 4.3678,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614382358587046e-05,
+      "loss": 4.3434,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613543763835994e-05,
+      "loss": 4.3554,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612706806965316e-05,
+      "loss": 4.3542,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611868212214264e-05,
+      "loss": 4.3512,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611029617463212e-05,
+      "loss": 4.3548,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61019102271216e-05,
+      "loss": 4.3555,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609352427961108e-05,
+      "loss": 4.3487,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608513833210056e-05,
+      "loss": 4.3831,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607675238459004e-05,
+      "loss": 4.3514,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606836643707952e-05,
+      "loss": 4.3545,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059980489569e-05,
+      "loss": 4.3548,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605161092086221e-05,
+      "loss": 4.3657,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604322497335169e-05,
+      "loss": 4.3469,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603483902584117e-05,
+      "loss": 4.3541,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602645307833065e-05,
+      "loss": 4.3477,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018083509623864e-05,
+      "loss": 4.3498,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009697562113344e-05,
+      "loss": 4.3406,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001311614602824e-05,
+      "loss": 4.3496,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5992925667092303e-05,
+      "loss": 4.3594,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984539719581783e-05,
+      "loss": 4.3503,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976153772071257e-05,
+      "loss": 4.3513,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967767824560737e-05,
+      "loss": 4.3501,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595939825585395e-05,
+      "loss": 4.346,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595101230834343e-05,
+      "loss": 4.351,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942626360832906e-05,
+      "loss": 4.3295,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934240413322386e-05,
+      "loss": 4.3448,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.327,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3432,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3392,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590071300208404e-05,
+      "loss": 4.3435,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589232705457352e-05,
+      "loss": 4.3358,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883941107063e-05,
+      "loss": 4.3352,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587555515955248e-05,
+      "loss": 4.3432,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586716921204196e-05,
+      "loss": 4.3402,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3418,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585043007462838e-05,
+      "loss": 4.3358,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584204412711786e-05,
+      "loss": 4.318,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583365817960734e-05,
+      "loss": 4.3406,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582527223209682e-05,
+      "loss": 4.3347,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58168862845863e-05,
+      "loss": 4.3357,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808500337075786e-05,
+      "loss": 4.3213,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.3447,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3117,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783375252151684e-05,
+      "loss": 4.3487,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774989304641164e-05,
+      "loss": 4.312,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766603357130644e-05,
+      "loss": 4.3285,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758217409620124e-05,
+      "loss": 4.3224,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749831462109604e-05,
+      "loss": 4.3225,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3185,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573307594589229e-05,
+      "loss": 4.3292,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572468999838177e-05,
+      "loss": 4.3041,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571630405087125e-05,
+      "loss": 4.312,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570791810336074e-05,
+      "loss": 4.3268,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569953215585022e-05,
+      "loss": 4.3208,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56911462083397e-05,
+      "loss": 4.3155,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568276026082918e-05,
+      "loss": 4.3158,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567437431331866e-05,
+      "loss": 4.3036,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566598836580814e-05,
+      "loss": 4.3183,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565761879710135e-05,
+      "loss": 4.322,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564923284959083e-05,
+      "loss": 4.3211,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564086328088404e-05,
+      "loss": 4.3311,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563247733337352e-05,
+      "loss": 4.3291,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624091385863e-05,
+      "loss": 4.3259,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561570543835248e-05,
+      "loss": 4.3047,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560731949084196e-05,
+      "loss": 4.3143,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559894992213517e-05,
+      "loss": 4.3225,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559056397462465e-05,
+      "loss": 4.3025,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558217802711413e-05,
+      "loss": 4.3181,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557379207960361e-05,
+      "loss": 4.3117,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556540613209309e-05,
+      "loss": 4.3221,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557020184582566e-05,
+      "loss": 4.3188,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548634237072046e-05,
+      "loss": 4.3009,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540248289561526e-05,
+      "loss": 4.3135,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.3136,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.321,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515106825833695e-05,
+      "loss": 4.3139,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506720878323175e-05,
+      "loss": 4.3165,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498334930812655e-05,
+      "loss": 4.3202,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548994898330214e-05,
+      "loss": 4.3006,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548156303579162e-05,
+      "loss": 4.3157,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547319346708483e-05,
+      "loss": 4.307,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546482389837804e-05,
+      "loss": 4.3043,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545643795086752e-05,
+      "loss": 4.3191,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448052003357e-05,
+      "loss": 4.3062,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543966605584648e-05,
+      "loss": 4.2952,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543128010833596e-05,
+      "loss": 4.3052,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542289416082544e-05,
+      "loss": 4.312,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541450821331492e-05,
+      "loss": 4.3111,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54061222658044e-05,
+      "loss": 4.2996,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539773631829388e-05,
+      "loss": 4.2952,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538935037078336e-05,
+      "loss": 4.2954,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538096442327284e-05,
+      "loss": 4.3096,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372594854566055e-05,
+      "loss": 4.282,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364225285859264e-05,
+      "loss": 4.2961,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355839338348744e-05,
+      "loss": 4.3011,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347453390838224e-05,
+      "loss": 4.2982,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339067443327704e-05,
+      "loss": 4.2927,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533069787462091e-05,
+      "loss": 4.3021,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532231192711039e-05,
+      "loss": 4.2926,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531392597959987e-05,
+      "loss": 4.2931,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530554003208935e-05,
+      "loss": 4.2976,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529715408457883e-05,
+      "loss": 4.2944,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528878451587205e-05,
+      "loss": 4.3064,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528039856836153e-05,
+      "loss": 4.2974,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527202899965474e-05,
+      "loss": 4.3005,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526364305214422e-05,
+      "loss": 4.2866,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52552571046337e-05,
+      "loss": 4.2949,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524687115712318e-05,
+      "loss": 4.2932,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523848520961266e-05,
+      "loss": 4.2821,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523009926210214e-05,
+      "loss": 4.2965,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522171331459162e-05,
+      "loss": 4.2855,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52133273670811e-05,
+      "loss": 4.2954,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520494141957057e-05,
+      "loss": 4.2809,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519655547206005e-05,
+      "loss": 4.2928,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518816952454953e-05,
+      "loss": 4.2676,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517979995584275e-05,
+      "loss": 4.2969,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517141400833223e-05,
+      "loss": 4.2819,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516302806082171e-05,
+      "loss": 4.2954,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154642113311187e-05,
+      "loss": 4.2934,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146256165800667e-05,
+      "loss": 4.2922,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137870218290147e-05,
+      "loss": 4.2756,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129484270779626e-05,
+      "loss": 4.2892,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121114702072836e-05,
+      "loss": 4.2941,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112728754562316e-05,
+      "loss": 4.2991,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104342807051795e-05,
+      "loss": 4.2883,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095956859541275e-05,
+      "loss": 4.2886,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087570912030755e-05,
+      "loss": 4.2866,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079184964520235e-05,
+      "loss": 4.2885,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070799017009715e-05,
+      "loss": 4.2745,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506242944830293e-05,
+      "loss": 4.2827,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505404350079241e-05,
+      "loss": 4.2836,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504565755328189e-05,
+      "loss": 4.2777,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503727160577137e-05,
+      "loss": 4.2765,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502888565826085e-05,
+      "loss": 4.2873,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502049971075033e-05,
+      "loss": 4.2744,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501213014204354e-05,
+      "loss": 4.2792,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500374419453302e-05,
+      "loss": 4.2854,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.261732578277588,
+      "eval_runtime": 291.0958,
+      "eval_samples_per_second": 1310.878,
+      "eval_steps_per_second": 40.966,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49953582470225e-05,
+      "loss": 4.2664,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498697229951198e-05,
+      "loss": 4.2698,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497858635200146e-05,
+      "loss": 4.2812,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497020040449093e-05,
+      "loss": 4.2745,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496181445698042e-05,
+      "loss": 4.2878,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49534285094699e-05,
+      "loss": 4.2646,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494504256195938e-05,
+      "loss": 4.2722,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493665661444886e-05,
+      "loss": 4.2636,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492827066693834e-05,
+      "loss": 4.2794,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491988471942782e-05,
+      "loss": 4.278,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.49114987719173e-05,
+      "loss": 4.272,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490311282440678e-05,
+      "loss": 4.2846,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489472687689626e-05,
+      "loss": 4.2535,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.488634092938574e-05,
+      "loss": 4.2687,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487795498187522e-05,
+      "loss": 4.2707,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48695690343647e-05,
+      "loss": 4.2611,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486119946565791e-05,
+      "loss": 4.2662,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485281351814739e-05,
+      "loss": 4.2695,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484442757063687e-05,
+      "loss": 4.2643,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483604162312635e-05,
+      "loss": 4.2938,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482765567561583e-05,
+      "loss": 4.2667,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481926972810531e-05,
+      "loss": 4.2654,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481088378059479e-05,
+      "loss": 4.2734,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480249783308427e-05,
+      "loss": 4.2799,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479412826437748e-05,
+      "loss": 4.2604,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478574231686696e-05,
+      "loss": 4.2694,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477737274816017e-05,
+      "loss": 4.2615,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476898680064965e-05,
+      "loss": 4.2643,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4760600853139126e-05,
+      "loss": 4.2565,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752214905628606e-05,
+      "loss": 4.2653,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4743828958118086e-05,
+      "loss": 4.2762,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47354593894113e-05,
+      "loss": 4.2657,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472707344190078e-05,
+      "loss": 4.267,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718703873194e-05,
+      "loss": 4.268,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471031792568348e-05,
+      "loss": 4.2643,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470193197817295e-05,
+      "loss": 4.2669,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469354603066243e-05,
+      "loss": 4.2476,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468516008315191e-05,
+      "loss": 4.2595,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467677413564139e-05,
+      "loss": 4.2487,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466838818813087e-05,
+      "loss": 4.2548,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466000224062035e-05,
+      "loss": 4.2583,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465161629310983e-05,
+      "loss": 4.2635,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464323034559931e-05,
+      "loss": 4.2534,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463484439808879e-05,
+      "loss": 4.2572,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462645845057827e-05,
+      "loss": 4.2615,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461807250306775e-05,
+      "loss": 4.2585,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609702934360966e-05,
+      "loss": 4.2612,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601316986850446e-05,
+      "loss": 4.2542,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4592947418143655e-05,
+      "loss": 4.2391,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584561470633135e-05,
+      "loss": 4.2618,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576191901926344e-05,
+      "loss": 4.2498,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567805954415824e-05,
+      "loss": 4.2571,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559420006905304e-05,
+      "loss": 4.2457,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551034059394784e-05,
+      "loss": 4.2586,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542648111884264e-05,
+      "loss": 4.236,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534262164373744e-05,
+      "loss": 4.2668,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525876216863224e-05,
+      "loss": 4.2328,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451750664815644e-05,
+      "loss": 4.246,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450912070064592e-05,
+      "loss": 4.2507,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45007347531354e-05,
+      "loss": 4.2425,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449234880562488e-05,
+      "loss": 4.2431,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448396285811436e-05,
+      "loss": 4.2475,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447557691060384e-05,
+      "loss": 4.2351,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446719096309331e-05,
+      "loss": 4.2313,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445880501558279e-05,
+      "loss": 4.2467,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445041906807227e-05,
+      "loss": 4.2475,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444204949936548e-05,
+      "loss": 4.2359,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443366355185496e-05,
+      "loss": 4.2436,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442527760434444e-05,
+      "loss": 4.2276,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441689165683393e-05,
+      "loss": 4.2424,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440850570932341e-05,
+      "loss": 4.2458,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440011976181289e-05,
+      "loss": 4.2428,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439173381430237e-05,
+      "loss": 4.2518,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438334786679185e-05,
+      "loss": 4.2512,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437499467688879e-05,
+      "loss": 4.2556,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436660872937827e-05,
+      "loss": 4.2298,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4358222781867747e-05,
+      "loss": 4.2396,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4349836834357227e-05,
+      "loss": 4.2475,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4341450886846707e-05,
+      "loss": 4.2238,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333081318139916e-05,
+      "loss": 4.245,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432471174943313e-05,
+      "loss": 4.2395,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431632580192261e-05,
+      "loss": 4.2462,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430793985441209e-05,
+      "loss": 4.2473,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429955390690157e-05,
+      "loss": 4.2263,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429116795939105e-05,
+      "loss": 4.241,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428278201188053e-05,
+      "loss": 4.2372,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427439606437001e-05,
+      "loss": 4.245,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426601011685949e-05,
+      "loss": 4.2383,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425762416934897e-05,
+      "loss": 4.2467,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424923822183845e-05,
+      "loss": 4.2439,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424085227432793e-05,
+      "loss": 4.23,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.423246632681741e-05,
+      "loss": 4.2418,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422409675811062e-05,
+      "loss": 4.2309,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215727189403836e-05,
+      "loss": 4.2379,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207341241893316e-05,
+      "loss": 4.2469,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198955294382796e-05,
+      "loss": 4.2306,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190569346872276e-05,
+      "loss": 4.2275,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182199778165485e-05,
+      "loss": 4.2306,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4173830209458694e-05,
+      "loss": 4.2389,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165444261948174e-05,
+      "loss": 4.2453,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157058314437654e-05,
+      "loss": 4.2268,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4148672366927134e-05,
+      "loss": 4.2265,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140286419416614e-05,
+      "loss": 4.2228,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4131900471906094e-05,
+      "loss": 4.2385,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4123514524395574e-05,
+      "loss": 4.2119,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115128576885054e-05,
+      "loss": 4.2249,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106742629374534e-05,
+      "loss": 4.2325,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409835668186402e-05,
+      "loss": 4.226,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899707343535e-05,
+      "loss": 4.2244,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081584786842974e-05,
+      "loss": 4.2361,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407321521813619e-05,
+      "loss": 4.2227,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40648456494294e-05,
+      "loss": 4.2194,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405645970191888e-05,
+      "loss": 4.2308,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404807375440836e-05,
+      "loss": 4.2274,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403968780689784e-05,
+      "loss": 4.2344,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403130185938732e-05,
+      "loss": 4.2312,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402291591187679e-05,
+      "loss": 4.2302,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401452996436627e-05,
+      "loss": 4.2179,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400614401685576e-05,
+      "loss": 4.2266,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399775806934524e-05,
+      "loss": 4.2238,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398937212183472e-05,
+      "loss": 4.2126,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39809861743242e-05,
+      "loss": 4.2328,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397260022681368e-05,
+      "loss": 4.2136,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396423065810689e-05,
+      "loss": 4.228,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395584471059637e-05,
+      "loss": 4.2203,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394745876308585e-05,
+      "loss": 4.2244,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393907281557533e-05,
+      "loss": 4.2031,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930703246868536e-05,
+      "loss": 4.2232,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922333678161745e-05,
+      "loss": 4.2186,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3913947730651225e-05,
+      "loss": 4.2252,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390557816194444e-05,
+      "loss": 4.2245,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389719221443392e-05,
+      "loss": 4.2286,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38888062669234e-05,
+      "loss": 4.207,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388042031941288e-05,
+      "loss": 4.2227,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387203437190236e-05,
+      "loss": 4.2284,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386364842439184e-05,
+      "loss": 4.228,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385526247688132e-05,
+      "loss": 4.2232,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38468765293708e-05,
+      "loss": 4.2274,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383849058186028e-05,
+      "loss": 4.2176,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383010463434976e-05,
+      "loss": 4.2245,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382171868683924e-05,
+      "loss": 4.2065,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.381333273932872e-05,
+      "loss": 4.2201,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380496317062193e-05,
+      "loss": 4.2136,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379657722311141e-05,
+      "loss": 4.2142,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378819127560089e-05,
+      "loss": 4.2149,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779805328090376e-05,
+      "loss": 4.2144,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771435759383585e-05,
+      "loss": 4.2155,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763066190676794e-05,
+      "loss": 4.2121,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3754696621970004e-05,
+      "loss": 4.2246,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.204609394073486,
+      "eval_runtime": 289.5427,
+      "eval_samples_per_second": 1317.909,
+      "eval_steps_per_second": 41.186,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3746310674459483e-05,
+      "loss": 4.209,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3737924726948963e-05,
+      "loss": 4.203,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3729538779438443e-05,
+      "loss": 4.217,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372115283192792e-05,
+      "loss": 4.2146,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37127668844174e-05,
+      "loss": 4.2172,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370438093690688e-05,
+      "loss": 4.2064,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369599498939636e-05,
+      "loss": 4.2034,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368760904188584e-05,
+      "loss": 4.2055,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367922309437533e-05,
+      "loss": 4.2119,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367083714686481e-05,
+      "loss": 4.2186,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366245119935428e-05,
+      "loss": 4.2073,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36540816306475e-05,
+      "loss": 4.2191,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364569568313698e-05,
+      "loss": 4.1961,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363732611443019e-05,
+      "loss": 4.2029,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362894016691967e-05,
+      "loss": 4.2097,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362055421940915e-05,
+      "loss": 4.1935,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361216827189863e-05,
+      "loss": 4.2031,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36037823243881e-05,
+      "loss": 4.2068,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359539637687758e-05,
+      "loss": 4.1998,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358701042936707e-05,
+      "loss": 4.2285,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357862448185655e-05,
+      "loss": 4.2103,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357025491314976e-05,
+      "loss": 4.2061,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356188534444297e-05,
+      "loss": 4.2097,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355349939693245e-05,
+      "loss": 4.2186,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545113449421926e-05,
+      "loss": 4.1973,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3536727501911406e-05,
+      "loss": 4.207,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3528341554400886e-05,
+      "loss": 4.2019,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3519955606890366e-05,
+      "loss": 4.1993,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3511569659379846e-05,
+      "loss": 4.1921,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3503200090673055e-05,
+      "loss": 4.2072,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3494814143162535e-05,
+      "loss": 4.2163,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348642819565202e-05,
+      "loss": 4.2046,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34780422481415e-05,
+      "loss": 4.2101,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346967267943471e-05,
+      "loss": 4.204,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461303110727926e-05,
+      "loss": 4.1998,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34529171632174e-05,
+      "loss": 4.2054,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344453121570688e-05,
+      "loss": 4.1937,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343614526819636e-05,
+      "loss": 4.1944,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342775932068584e-05,
+      "loss": 4.1897,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341937337317532e-05,
+      "loss": 4.1952,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34109874256648e-05,
+      "loss": 4.1999,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340260147815428e-05,
+      "loss": 4.2034,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339421553064376e-05,
+      "loss": 4.1924,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338582958313324e-05,
+      "loss": 4.1984,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337744363562272e-05,
+      "loss": 4.2046,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33690576881122e-05,
+      "loss": 4.1977,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360688119405415e-05,
+      "loss": 4.2018,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352302171894895e-05,
+      "loss": 4.1956,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3343916224384375e-05,
+      "loss": 4.1798,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3335530276873855e-05,
+      "loss": 4.2058,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327160708167064e-05,
+      "loss": 4.1915,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3318774760656544e-05,
+      "loss": 4.1999,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3310388813146024e-05,
+      "loss": 4.1869,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3302002865635504e-05,
+      "loss": 4.197,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329363329692871e-05,
+      "loss": 4.1784,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328524734941819e-05,
+      "loss": 4.2044,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327686140190767e-05,
+      "loss": 4.1765,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326847545439715e-05,
+      "loss": 4.1919,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326008950688664e-05,
+      "loss": 4.1941,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325170355937611e-05,
+      "loss": 4.1822,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324333399066933e-05,
+      "loss": 4.1837,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323494804315881e-05,
+      "loss": 4.1961,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322656209564829e-05,
+      "loss": 4.175,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321817614813776e-05,
+      "loss": 4.1728,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320979020062724e-05,
+      "loss": 4.1901,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320140425311672e-05,
+      "loss": 4.1922,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31930183056062e-05,
+      "loss": 4.1774,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184665115703146e-05,
+      "loss": 4.1904,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176279168192626e-05,
+      "loss": 4.1684,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167893220682106e-05,
+      "loss": 4.1824,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159507273171586e-05,
+      "loss": 4.1896,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151121325661066e-05,
+      "loss": 4.1892,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3142735378150546e-05,
+      "loss": 4.192,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3134349430640026e-05,
+      "loss": 4.1938,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3125963483129506e-05,
+      "loss": 4.2028,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3117577535618986e-05,
+      "loss": 4.1722,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109207966912195e-05,
+      "loss": 4.1871,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3100822019401675e-05,
+      "loss": 4.193,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3092436071891155e-05,
+      "loss": 4.1656,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3084066503184364e-05,
+      "loss": 4.1888,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075680555673844e-05,
+      "loss": 4.1841,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306729460816333e-05,
+      "loss": 4.1899,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305890866065281e-05,
+      "loss": 4.1909,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305052271314229e-05,
+      "loss": 4.175,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304213676563177e-05,
+      "loss": 4.1808,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303375081812125e-05,
+      "loss": 4.1823,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302536487061073e-05,
+      "loss": 4.1915,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301697892310021e-05,
+      "loss": 4.1793,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300859297558969e-05,
+      "loss": 4.1922,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.30002234068829e-05,
+      "loss": 4.1924,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299183745937238e-05,
+      "loss": 4.1776,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298345151186186e-05,
+      "loss": 4.183,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297506556435134e-05,
+      "loss": 4.1753,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296667961684082e-05,
+      "loss": 4.1847,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295831004813403e-05,
+      "loss": 4.1941,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949924100623515e-05,
+      "loss": 4.1723,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941538153112995e-05,
+      "loss": 4.1772,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933152205602475e-05,
+      "loss": 4.1726,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292476625809195e-05,
+      "loss": 4.1852,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291638031058143e-05,
+      "loss": 4.1936,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290799436307091e-05,
+      "loss": 4.1754,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2899624794364124e-05,
+      "loss": 4.1727,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28912388468536e-05,
+      "loss": 4.1652,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288285289934308e-05,
+      "loss": 4.1881,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287446695183256e-05,
+      "loss": 4.1562,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866097383125766e-05,
+      "loss": 4.1749,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285771143561525e-05,
+      "loss": 4.1749,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284932548810473e-05,
+      "loss": 4.1763,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284093954059421e-05,
+      "loss": 4.1726,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283255359308369e-05,
+      "loss": 4.1801,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282416764557317e-05,
+      "loss": 4.1736,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281578169806265e-05,
+      "loss": 4.1658,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280739575055213e-05,
+      "loss": 4.1782,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279904256064907e-05,
+      "loss": 4.175,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279065661313855e-05,
+      "loss": 4.1823,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278227066562803e-05,
+      "loss": 4.1818,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277388471811751e-05,
+      "loss": 4.1747,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276549877060699e-05,
+      "loss": 4.1667,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275711282309647e-05,
+      "loss": 4.1729,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274874325438969e-05,
+      "loss": 4.176,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274035730687917e-05,
+      "loss": 4.163,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273197135936865e-05,
+      "loss": 4.1762,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2723585411858127e-05,
+      "loss": 4.1633,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715215843151336e-05,
+      "loss": 4.1761,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2706829895640816e-05,
+      "loss": 4.1692,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698460326934025e-05,
+      "loss": 4.1707,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690074379423505e-05,
+      "loss": 4.1572,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2681688431912985e-05,
+      "loss": 4.1697,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2673302484402465e-05,
+      "loss": 4.1696,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2664916536891945e-05,
+      "loss": 4.1763,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2656530589381425e-05,
+      "loss": 4.1701,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2648144641870904e-05,
+      "loss": 4.1774,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263977507316412e-05,
+      "loss": 4.1593,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26313891256536e-05,
+      "loss": 4.172,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262300317814308e-05,
+      "loss": 4.18,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261461723063256e-05,
+      "loss": 4.1772,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260623128312204e-05,
+      "loss": 4.1731,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259784533561152e-05,
+      "loss": 4.1812,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589459388101e-05,
+      "loss": 4.1663,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258107344059048e-05,
+      "loss": 4.1755,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257268749307996e-05,
+      "loss": 4.1566,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256430154556943e-05,
+      "loss": 4.1693,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255591559805891e-05,
+      "loss": 4.1644,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254752965054839e-05,
+      "loss": 4.1633,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253917646064534e-05,
+      "loss": 4.166,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2530790513134825e-05,
+      "loss": 4.1619,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522404565624305e-05,
+      "loss": 4.1736,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514018618113785e-05,
+      "loss": 4.1549,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2505649049406994e-05,
+      "loss": 4.1788,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.163358211517334,
+      "eval_runtime": 290.0071,
+      "eval_samples_per_second": 1315.799,
+      "eval_steps_per_second": 41.12,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497263101896474e-05,
+      "loss": 4.1572,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488877154385954e-05,
+      "loss": 4.1527,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480491206875434e-05,
+      "loss": 4.168,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247210525936491e-05,
+      "loss": 4.1653,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246371931185439e-05,
+      "loss": 4.1672,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245533336434387e-05,
+      "loss": 4.1587,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244694741683335e-05,
+      "loss": 4.1496,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243856146932283e-05,
+      "loss": 4.1591,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243017552181231e-05,
+      "loss": 4.1627,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2421789574301793e-05,
+      "loss": 4.1718,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241340362679127e-05,
+      "loss": 4.1616,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240501767928075e-05,
+      "loss": 4.1668,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239663173177023e-05,
+      "loss": 4.1524,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238824578425971e-05,
+      "loss": 4.1543,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237985983674919e-05,
+      "loss": 4.1615,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237147388923867e-05,
+      "loss": 4.148,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363087941728146e-05,
+      "loss": 4.155,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354701994217626e-05,
+      "loss": 4.1605,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346316046707106e-05,
+      "loss": 4.1533,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337946478000315e-05,
+      "loss": 4.177,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329560530489795e-05,
+      "loss": 4.1672,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321174582979275e-05,
+      "loss": 4.1558,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231278863546876e-05,
+      "loss": 4.161,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230440268795824e-05,
+      "loss": 4.173,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229603311925145e-05,
+      "loss": 4.1503,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228764717174093e-05,
+      "loss": 4.1616,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227926122423041e-05,
+      "loss": 4.1522,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227087527671989e-05,
+      "loss": 4.1534,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226248932920937e-05,
+      "loss": 4.148,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225411976050258e-05,
+      "loss": 4.16,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224573381299206e-05,
+      "loss": 4.1667,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223734786548154e-05,
+      "loss": 4.1594,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222896191797102e-05,
+      "loss": 4.1644,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22205759704605e-05,
+      "loss": 4.1603,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221220640175371e-05,
+      "loss": 4.1514,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203820454243196e-05,
+      "loss": 4.1579,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195450885536405e-05,
+      "loss": 4.1497,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187064938025885e-05,
+      "loss": 4.1478,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178678990515365e-05,
+      "loss": 4.1459,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170293043004845e-05,
+      "loss": 4.1448,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2161907095494325e-05,
+      "loss": 4.1534,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2153521147983804e-05,
+      "loss": 4.1572,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145135200473284e-05,
+      "loss": 4.1491,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2136749252962764e-05,
+      "loss": 4.1498,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128363305452244e-05,
+      "loss": 4.1597,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2119977357941724e-05,
+      "loss": 4.1516,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111591410431204e-05,
+      "loss": 4.1581,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210320546292068e-05,
+      "loss": 4.1514,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209483589421389e-05,
+      "loss": 4.1314,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208644994670338e-05,
+      "loss": 4.1619,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207806399919286e-05,
+      "loss": 4.1464,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206967805168233e-05,
+      "loss": 4.1558,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206130848297555e-05,
+      "loss": 4.1458,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205292253546503e-05,
+      "loss": 4.1463,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204453658795451e-05,
+      "loss": 4.1404,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203615064044398e-05,
+      "loss": 4.1531,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20277810717372e-05,
+      "loss": 4.135,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201939512422668e-05,
+      "loss": 4.1452,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201100917671615e-05,
+      "loss": 4.154,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200262322920563e-05,
+      "loss": 4.1339,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199423728169512e-05,
+      "loss": 4.1373,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19858513341846e-05,
+      "loss": 4.1556,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197746538667408e-05,
+      "loss": 4.1272,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196907943916356e-05,
+      "loss": 4.1299,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196070987045677e-05,
+      "loss": 4.1423,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195232392294625e-05,
+      "loss": 4.1466,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194393797543573e-05,
+      "loss": 4.1368,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193555202792521e-05,
+      "loss": 4.1485,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927166080414687e-05,
+      "loss": 4.1245,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918780132904167e-05,
+      "loss": 4.1361,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910410564197376e-05,
+      "loss": 4.1473,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902024616686856e-05,
+      "loss": 4.1445,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1893638669176336e-05,
+      "loss": 4.1492,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885252721665816e-05,
+      "loss": 4.1504,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18768667741553e-05,
+      "loss": 4.1562,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186848082664478e-05,
+      "loss": 4.1296,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186009487913426e-05,
+      "loss": 4.1407,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185170893162374e-05,
+      "loss": 4.1468,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184332298411322e-05,
+      "loss": 4.1248,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183495341540643e-05,
+      "loss": 4.1444,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182658384669964e-05,
+      "loss": 4.1436,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181819789918912e-05,
+      "loss": 4.1486,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18098119516786e-05,
+      "loss": 4.147,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180142600416808e-05,
+      "loss": 4.1348,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179305643546129e-05,
+      "loss": 4.1371,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178467048795077e-05,
+      "loss": 4.141,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177628454044025e-05,
+      "loss": 4.1446,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767914971733465e-05,
+      "loss": 4.1409,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1759529024222945e-05,
+      "loss": 4.1482,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751143076712425e-05,
+      "loss": 4.1479,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1742757129201905e-05,
+      "loss": 4.1377,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734371181691385e-05,
+      "loss": 4.1392,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1725985234180865e-05,
+      "loss": 4.1308,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171759928667034e-05,
+      "loss": 4.1478,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170921333915982e-05,
+      "loss": 4.1485,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17008273916493e-05,
+      "loss": 4.1361,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169244144413878e-05,
+      "loss": 4.1336,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168405549662826e-05,
+      "loss": 4.1286,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167566954911774e-05,
+      "loss": 4.1436,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667299980410954e-05,
+      "loss": 4.1507,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1658914032900434e-05,
+      "loss": 4.1329,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1650528085389914e-05,
+      "loss": 4.1324,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642142137879393e-05,
+      "loss": 4.1275,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16337725691726e-05,
+      "loss": 4.1414,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162538662166208e-05,
+      "loss": 4.1205,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161700067415156e-05,
+      "loss": 4.1333,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160861472664104e-05,
+      "loss": 4.1269,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160022877913052e-05,
+      "loss": 4.1386,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159185921042373e-05,
+      "loss": 4.1311,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158348964171694e-05,
+      "loss": 4.1383,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157510369420643e-05,
+      "loss": 4.1331,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156671774669591e-05,
+      "loss": 4.1238,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155833179918539e-05,
+      "loss": 4.1384,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154994585167487e-05,
+      "loss": 4.1348,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1541576282968076e-05,
+      "loss": 4.1394,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533190335457556e-05,
+      "loss": 4.1402,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524804387947036e-05,
+      "loss": 4.1288,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516418440436516e-05,
+      "loss": 4.134,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508032492925996e-05,
+      "loss": 4.1263,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1499646545415476e-05,
+      "loss": 4.1355,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1491260597904956e-05,
+      "loss": 4.1314,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482874650394436e-05,
+      "loss": 4.1319,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474505081687645e-05,
+      "loss": 4.1181,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466119134177125e-05,
+      "loss": 4.1359,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145773318666661e-05,
+      "loss": 4.1245,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144936361795982e-05,
+      "loss": 4.1325,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14409776704493e-05,
+      "loss": 4.116,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143259172293878e-05,
+      "loss": 4.1318,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142420577542826e-05,
+      "loss": 4.1265,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141581982791774e-05,
+      "loss": 4.1391,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140745025921095e-05,
+      "loss": 4.1292,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139906431170043e-05,
+      "loss": 4.1355,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139067836418991e-05,
+      "loss": 4.1181,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138229241667939e-05,
+      "loss": 4.1328,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137390646916887e-05,
+      "loss": 4.1391,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136552052165835e-05,
+      "loss": 4.1358,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135713457414783e-05,
+      "loss": 4.1334,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134874862663731e-05,
+      "loss": 4.1391,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340379057930525e-05,
+      "loss": 4.1255,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1331993110420005e-05,
+      "loss": 4.1381,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323623541713214e-05,
+      "loss": 4.1179,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315237594202694e-05,
+      "loss": 4.1307,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306851646692174e-05,
+      "loss": 4.1248,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129846569918165e-05,
+      "loss": 4.1266,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129007975167113e-05,
+      "loss": 4.1227,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128169380416061e-05,
+      "loss": 4.1204,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127330785665009e-05,
+      "loss": 4.1308,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12649382879433e-05,
+      "loss": 4.1189,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125655234043278e-05,
+      "loss": 4.141,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.131695747375488,
+      "eval_runtime": 305.7931,
+      "eval_samples_per_second": 1247.873,
+      "eval_steps_per_second": 38.997,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124816639292226e-05,
+      "loss": 4.1185,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123978044541174e-05,
+      "loss": 4.1132,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123139449790122e-05,
+      "loss": 4.1269,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12230085503907e-05,
+      "loss": 4.1259,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121462260288018e-05,
+      "loss": 4.125,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120623665536966e-05,
+      "loss": 4.1245,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119786708666287e-05,
+      "loss": 4.1163,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118948113915235e-05,
+      "loss": 4.1171,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118109519164183e-05,
+      "loss": 4.1255,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117272562293504e-05,
+      "loss": 4.1313,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116433967542452e-05,
+      "loss": 4.1248,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1155953727914e-05,
+      "loss": 4.123,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114756778040348e-05,
+      "loss": 4.1192,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113918183289297e-05,
+      "loss": 4.114,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113081226418618e-05,
+      "loss": 4.1219,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112242631667566e-05,
+      "loss": 4.11,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111404036916514e-05,
+      "loss": 4.1193,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105654421654617e-05,
+      "loss": 4.1238,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1097284852947826e-05,
+      "loss": 4.1135,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1088898905437306e-05,
+      "loss": 4.1372,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1080529336730515e-05,
+      "loss": 4.1331,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072143389219995e-05,
+      "loss": 4.1191,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063757441709475e-05,
+      "loss": 4.1227,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1055371494198955e-05,
+      "loss": 4.1342,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1046985546688435e-05,
+      "loss": 4.1121,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103859959917792e-05,
+      "loss": 4.1246,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10302136516674e-05,
+      "loss": 4.1184,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102182770415688e-05,
+      "loss": 4.1145,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101345813545009e-05,
+      "loss": 4.1106,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100507218793957e-05,
+      "loss": 4.1247,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099668624042905e-05,
+      "loss": 4.1269,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098830029291853e-05,
+      "loss": 4.1247,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097993072421174e-05,
+      "loss": 4.1283,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097154477670122e-05,
+      "loss": 4.1176,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09631588291907e-05,
+      "loss": 4.1166,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095477288168018e-05,
+      "loss": 4.1212,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094638693416966e-05,
+      "loss": 4.1158,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093800098665914e-05,
+      "loss": 4.1055,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092961503914862e-05,
+      "loss": 4.1087,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921245470441835e-05,
+      "loss": 4.1104,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091285952293131e-05,
+      "loss": 4.115,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090447357542079e-05,
+      "loss": 4.1222,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089608762791027e-05,
+      "loss": 4.1136,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088770168039975e-05,
+      "loss": 4.1106,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087931573288923e-05,
+      "loss": 4.1231,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087092978537871e-05,
+      "loss": 4.117,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086254383786819e-05,
+      "loss": 4.1208,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08541742691614e-05,
+      "loss": 4.1171,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084578832165088e-05,
+      "loss": 4.0901,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083740237414036e-05,
+      "loss": 4.131,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829016426629844e-05,
+      "loss": 4.1079,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082064685792305e-05,
+      "loss": 4.121,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081227728921626e-05,
+      "loss": 4.108,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080389134170574e-05,
+      "loss": 4.1122,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079550539419522e-05,
+      "loss": 4.1029,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07871194466847e-05,
+      "loss": 4.1151,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077873349917418e-05,
+      "loss": 4.1013,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077036393046739e-05,
+      "loss": 4.1067,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076197798295687e-05,
+      "loss": 4.1208,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075359203544635e-05,
+      "loss": 4.1012,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074520608793583e-05,
+      "loss": 4.0996,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073682014042531e-05,
+      "loss": 4.1181,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07284341929148e-05,
+      "loss": 4.0937,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072004824540428e-05,
+      "loss": 4.0922,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071166229789376e-05,
+      "loss": 4.1093,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0703292729186966e-05,
+      "loss": 4.1105,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0694923160480175e-05,
+      "loss": 4.1028,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0686537212969655e-05,
+      "loss": 4.1119,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678151265459135e-05,
+      "loss": 4.0891,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0669765317948615e-05,
+      "loss": 4.1036,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0661379370438095e-05,
+      "loss": 4.1126,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0652993422927575e-05,
+      "loss": 4.1071,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0644623854220784e-05,
+      "loss": 4.1169,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636237906710264e-05,
+      "loss": 4.1135,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062785195919975e-05,
+      "loss": 4.1198,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061946601168923e-05,
+      "loss": 4.0946,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061108006417871e-05,
+      "loss": 4.1093,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060269411666819e-05,
+      "loss": 4.111,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059430816915767e-05,
+      "loss": 4.0933,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0585922221647144e-05,
+      "loss": 4.108,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057755265294036e-05,
+      "loss": 4.1094,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056916670542984e-05,
+      "loss": 4.1125,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056078075791932e-05,
+      "loss": 4.1161,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055241118921253e-05,
+      "loss": 4.0973,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054402524170201e-05,
+      "loss": 4.1027,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053563929419149e-05,
+      "loss": 4.1076,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052725334668097e-05,
+      "loss": 4.1086,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0518883777974184e-05,
+      "loss": 4.1088,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0510497830463664e-05,
+      "loss": 4.1107,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0502111882953144e-05,
+      "loss": 4.1132,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049372593544262e-05,
+      "loss": 4.1069,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04853399879321e-05,
+      "loss": 4.1041,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047695404042158e-05,
+      "loss": 4.0975,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046856809291106e-05,
+      "loss": 4.1126,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046018214540054e-05,
+      "loss": 4.1133,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045179619789002e-05,
+      "loss": 4.1022,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443426629183226e-05,
+      "loss": 4.0981,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435040681672706e-05,
+      "loss": 4.0953,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0426654734162186e-05,
+      "loss": 4.1082,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418268786651666e-05,
+      "loss": 4.1177,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040989921794488e-05,
+      "loss": 4.0991,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040151327043436e-05,
+      "loss": 4.1002,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039314370172757e-05,
+      "loss": 4.0984,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038475775421705e-05,
+      "loss": 4.1013,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037637180670653e-05,
+      "loss": 4.0879,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036798585919601e-05,
+      "loss": 4.1008,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035959991168549e-05,
+      "loss": 4.0891,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035121396417497e-05,
+      "loss": 4.1051,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034282801666445e-05,
+      "loss": 4.0993,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033444206915393e-05,
+      "loss": 4.1011,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032605612164341e-05,
+      "loss": 4.0984,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031768655293662e-05,
+      "loss": 4.0977,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309316984229836e-05,
+      "loss": 4.1008,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0300931036719316e-05,
+      "loss": 4.1017,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0292545089208796e-05,
+      "loss": 4.1033,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0284159141698276e-05,
+      "loss": 4.1098,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0275773194187756e-05,
+      "loss": 4.0966,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267387246677236e-05,
+      "loss": 4.1022,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259001299166716e-05,
+      "loss": 4.0939,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0250615351656196e-05,
+      "loss": 4.1003,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242229404145675e-05,
+      "loss": 4.0999,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0233843456635155e-05,
+      "loss": 4.1002,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022545750912463e-05,
+      "loss": 4.0864,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021707156161411e-05,
+      "loss": 4.0989,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0208701992907324e-05,
+      "loss": 4.096,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200316045396804e-05,
+      "loss": 4.0964,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019194647669002e-05,
+      "loss": 4.0798,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01835605291795e-05,
+      "loss": 4.1041,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017517458166898e-05,
+      "loss": 4.0919,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016678863415845e-05,
+      "loss": 4.1063,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015840268664793e-05,
+      "loss": 4.098,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015001673913741e-05,
+      "loss": 4.1028,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014163079162689e-05,
+      "loss": 4.0864,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013324484411637e-05,
+      "loss": 4.1036,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012487527540958e-05,
+      "loss": 4.1072,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01165057067028e-05,
+      "loss": 4.1028,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010811975919228e-05,
+      "loss": 4.1042,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009973381168176e-05,
+      "loss": 4.1034,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009134786417124e-05,
+      "loss": 4.0948,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008296191666072e-05,
+      "loss": 4.104,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00745759691502e-05,
+      "loss": 4.087,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006619002163968e-05,
+      "loss": 4.1019,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005782045293289e-05,
+      "loss": 4.0901,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004943450542237e-05,
+      "loss": 4.0977,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004104855791185e-05,
+      "loss": 4.0864,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003266261040133e-05,
+      "loss": 4.0899,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002427666289081e-05,
+      "loss": 4.0944,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0015907094184016e-05,
+      "loss": 4.0926,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000753752547723e-05,
+      "loss": 4.1095,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.107858657836914,
+      "eval_runtime": 312.1407,
+      "eval_samples_per_second": 1222.497,
+      "eval_steps_per_second": 38.204,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999915157796671e-05,
+      "loss": 4.0826,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999076563045619e-05,
+      "loss": 4.0857,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998237968294567e-05,
+      "loss": 4.0909,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997399373543515e-05,
+      "loss": 4.0941,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996560778792463e-05,
+      "loss": 4.0969,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995722184041411e-05,
+      "loss": 4.0893,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994883589290359e-05,
+      "loss": 4.0927,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994044994539307e-05,
+      "loss": 4.0866,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993206399788255e-05,
+      "loss": 4.0898,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992367805037203e-05,
+      "loss": 4.1033,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991529210286151e-05,
+      "loss": 4.0912,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9906906155350984e-05,
+      "loss": 4.0919,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9898520207840464e-05,
+      "loss": 4.0871,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890134260329944e-05,
+      "loss": 4.081,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988174831281943e-05,
+      "loss": 4.0941,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987336236530891e-05,
+      "loss": 4.077,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986499279660212e-05,
+      "loss": 4.0874,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98566068490916e-05,
+      "loss": 4.0931,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984822090158108e-05,
+      "loss": 4.0795,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983983495407056e-05,
+      "loss": 4.1044,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983144900656004e-05,
+      "loss": 4.1034,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982306305904952e-05,
+      "loss": 4.0906,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981469349034273e-05,
+      "loss": 4.0884,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980630754283221e-05,
+      "loss": 4.1033,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979792159532169e-05,
+      "loss": 4.0883,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978953564781117e-05,
+      "loss": 4.0886,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978114970030065e-05,
+      "loss": 4.0879,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9772780131593865e-05,
+      "loss": 4.0846,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9764394184083345e-05,
+      "loss": 4.085,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9756008236572825e-05,
+      "loss": 4.0895,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9747622289062305e-05,
+      "loss": 4.0928,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739252720355514e-05,
+      "loss": 4.0964,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9730866772844994e-05,
+      "loss": 4.0962,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9722480825334474e-05,
+      "loss": 4.0898,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9714094877823954e-05,
+      "loss": 4.0817,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9705708930313434e-05,
+      "loss": 4.0945,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697322982802913e-05,
+      "loss": 4.0867,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968895341409612e-05,
+      "loss": 4.0762,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96805674665856e-05,
+      "loss": 4.0795,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967218151907508e-05,
+      "loss": 4.0762,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96638119503683e-05,
+      "loss": 4.0836,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965542600285778e-05,
+      "loss": 4.0897,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964704005534726e-05,
+      "loss": 4.0871,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963865410783674e-05,
+      "loss": 4.0836,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963026816032622e-05,
+      "loss": 4.0859,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96218822128157e-05,
+      "loss": 4.0884,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961349626530518e-05,
+      "loss": 4.087,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960511031779465e-05,
+      "loss": 4.0903,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959674074908787e-05,
+      "loss": 4.0613,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958835480157735e-05,
+      "loss": 4.0993,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957996885406682e-05,
+      "loss": 4.0784,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957158290655631e-05,
+      "loss": 4.0942,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956319695904579e-05,
+      "loss": 4.0772,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955481101153527e-05,
+      "loss": 4.0807,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954642506402475e-05,
+      "loss": 4.075,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9538055495317956e-05,
+      "loss": 4.083,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9529669547807436e-05,
+      "loss": 4.075,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952129997910065e-05,
+      "loss": 4.0772,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9512914031590125e-05,
+      "loss": 4.0904,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9504528084079605e-05,
+      "loss": 4.0756,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9496142136569085e-05,
+      "loss": 4.0671,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9487756189058565e-05,
+      "loss": 4.0924,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479370241548045e-05,
+      "loss": 4.0651,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9470984294037525e-05,
+      "loss": 4.0627,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462598346527005e-05,
+      "loss": 4.0797,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454212399016485e-05,
+      "loss": 4.0809,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944582645150597e-05,
+      "loss": 4.0744,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943744050399545e-05,
+      "loss": 4.0804,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942907093528866e-05,
+      "loss": 4.0611,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942070136658187e-05,
+      "loss": 4.0737,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941231541907135e-05,
+      "loss": 4.083,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940392947156083e-05,
+      "loss": 4.078,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939554352405031e-05,
+      "loss": 4.0887,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938715757653979e-05,
+      "loss": 4.0798,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937877162902927e-05,
+      "loss": 4.0927,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937038568151875e-05,
+      "loss": 4.0665,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936199973400823e-05,
+      "loss": 4.0801,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9353646544105174e-05,
+      "loss": 4.0788,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9345260596594654e-05,
+      "loss": 4.0682,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9336874649084134e-05,
+      "loss": 4.0779,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932850508037734e-05,
+      "loss": 4.0844,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932011913286682e-05,
+      "loss": 4.0805,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93117331853563e-05,
+      "loss": 4.0904,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930334723784578e-05,
+      "loss": 4.0666,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929496129033526e-05,
+      "loss": 4.0694,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928657534282474e-05,
+      "loss": 4.0823,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927818939531422e-05,
+      "loss": 4.0789,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92698034478037e-05,
+      "loss": 4.0798,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926141750029318e-05,
+      "loss": 4.0856,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925303155278266e-05,
+      "loss": 4.0796,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924464560527214e-05,
+      "loss": 4.081,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923625965776162e-05,
+      "loss": 4.0746,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92278737102511e-05,
+      "loss": 4.0692,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921950414154431e-05,
+      "loss": 4.0861,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921111819403379e-05,
+      "loss": 4.0794,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920273224652327e-05,
+      "loss": 4.077,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919434629901275e-05,
+      "loss": 4.0724,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918597673030596e-05,
+      "loss": 4.0672,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917759078279544e-05,
+      "loss": 4.076,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916920483528492e-05,
+      "loss": 4.0887,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91608188877744e-05,
+      "loss": 4.0735,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9152449319067617e-05,
+      "loss": 4.0713,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914407975036083e-05,
+      "loss": 4.0682,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913569380285031e-05,
+      "loss": 4.0695,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127307855339786e-05,
+      "loss": 4.0676,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9118921907829266e-05,
+      "loss": 4.0679,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9110535960318745e-05,
+      "loss": 4.0623,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102150012808225e-05,
+      "loss": 4.078,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9093764065297705e-05,
+      "loss": 4.0743,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9085378117787185e-05,
+      "loss": 4.071,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077008549080394e-05,
+      "loss": 4.0643,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068638980373604e-05,
+      "loss": 4.0742,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060253032863083e-05,
+      "loss": 4.0731,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90518834641563e-05,
+      "loss": 4.0733,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9043497516645786e-05,
+      "loss": 4.0774,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903511156913526e-05,
+      "loss": 4.0811,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902672562162474e-05,
+      "loss": 4.0655,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901833967411422e-05,
+      "loss": 4.0787,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90099537266037e-05,
+      "loss": 4.0619,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900156777909318e-05,
+      "loss": 4.072,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899318183158266e-05,
+      "loss": 4.0719,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898479588407214e-05,
+      "loss": 4.0673,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897642631536535e-05,
+      "loss": 4.0639,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896804036785483e-05,
+      "loss": 4.0729,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895965442034431e-05,
+      "loss": 4.0715,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8951284851637524e-05,
+      "loss": 4.0703,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8942898904127004e-05,
+      "loss": 4.0529,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8934512956616484e-05,
+      "loss": 4.0735,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926127009105964e-05,
+      "loss": 4.0655,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8917741061595444e-05,
+      "loss": 4.0799,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8909355114084924e-05,
+      "loss": 4.0727,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8900969166574404e-05,
+      "loss": 4.0684,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892583219063884e-05,
+      "loss": 4.0607,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884197271553364e-05,
+      "loss": 4.0755,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887582770284657e-05,
+      "loss": 4.0778,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886744175533605e-05,
+      "loss": 4.0763,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885905580782553e-05,
+      "loss": 4.0766,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885066986031501e-05,
+      "loss": 4.0749,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884230029160822e-05,
+      "loss": 4.0696,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88339143440977e-05,
+      "loss": 4.076,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882552839658719e-05,
+      "loss": 4.0584,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88171588278804e-05,
+      "loss": 4.0743,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880877288036988e-05,
+      "loss": 4.0658,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880038693285936e-05,
+      "loss": 4.0689,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879200098534884e-05,
+      "loss": 4.0587,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8783631416642046e-05,
+      "loss": 4.0614,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8775245469131526e-05,
+      "loss": 4.0694,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8766859521621006e-05,
+      "loss": 4.0686,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8758473574110486e-05,
+      "loss": 4.0812,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.087856292724609,
+      "eval_runtime": 289.4911,
+      "eval_samples_per_second": 1318.144,
+      "eval_steps_per_second": 41.193,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750087626599966e-05,
+      "loss": 4.0527,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874170167908944e-05,
+      "loss": 4.0592,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8733332110382655e-05,
+      "loss": 4.0601,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872494616287214e-05,
+      "loss": 4.072,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8716560215361615e-05,
+      "loss": 4.069,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8708174267851095e-05,
+      "loss": 4.0604,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8699788320340575e-05,
+      "loss": 4.0659,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8691402372830055e-05,
+      "loss": 4.0592,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683016425319535e-05,
+      "loss": 4.0607,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8674630477809015e-05,
+      "loss": 4.0781,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8666244530298495e-05,
+      "loss": 4.0644,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8657858582787975e-05,
+      "loss": 4.063,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8649472635277455e-05,
+      "loss": 4.0699,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641086687766935e-05,
+      "loss": 4.047,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8632717119060144e-05,
+      "loss": 4.0649,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8624331171549624e-05,
+      "loss": 4.0531,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861594522403911e-05,
+      "loss": 4.0626,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860755927652859e-05,
+      "loss": 4.0656,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859917332901807e-05,
+      "loss": 4.0525,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859078738150755e-05,
+      "loss": 4.0744,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858240143399703e-05,
+      "loss": 4.0782,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857403186529024e-05,
+      "loss": 4.0654,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856564591777972e-05,
+      "loss": 4.0665,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85572599702692e-05,
+      "loss": 4.0702,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854887402275868e-05,
+      "loss": 4.0666,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854048807524815e-05,
+      "loss": 4.059,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853210212773763e-05,
+      "loss": 4.0656,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852373255903085e-05,
+      "loss": 4.0607,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851534661152033e-05,
+      "loss": 4.0565,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850696066400981e-05,
+      "loss": 4.0623,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849857471649929e-05,
+      "loss": 4.0648,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849018876898877e-05,
+      "loss": 4.0732,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848180282147825e-05,
+      "loss": 4.0697,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847341687396773e-05,
+      "loss": 4.0671,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846503092645721e-05,
+      "loss": 4.0566,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845666135775042e-05,
+      "loss": 4.0619,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84482754102399e-05,
+      "loss": 4.0655,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843988946272938e-05,
+      "loss": 4.0526,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843150351521886e-05,
+      "loss": 4.0508,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423133946512066e-05,
+      "loss": 4.0513,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414747999001546e-05,
+      "loss": 4.0603,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840636205149103e-05,
+      "loss": 4.06,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839797610398051e-05,
+      "loss": 4.0614,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838959015646999e-05,
+      "loss": 4.06,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83812205877632e-05,
+      "loss": 4.0628,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837283464025268e-05,
+      "loss": 4.0583,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836444869274216e-05,
+      "loss": 4.0614,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835606274523164e-05,
+      "loss": 4.0688,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834767679772112e-05,
+      "loss": 4.0335,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83392908502106e-05,
+      "loss": 4.0674,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833090490270008e-05,
+      "loss": 4.055,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832253533399329e-05,
+      "loss": 4.0713,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831414938648277e-05,
+      "loss": 4.0519,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830576343897225e-05,
+      "loss": 4.0549,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829737749146173e-05,
+      "loss": 4.0469,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828899154395121e-05,
+      "loss": 4.059,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82806055964407e-05,
+      "loss": 4.0548,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8272236027733906e-05,
+      "loss": 4.0479,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8263850080223386e-05,
+      "loss": 4.0642,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8255464132712866e-05,
+      "loss": 4.0538,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824707818520234e-05,
+      "loss": 4.0357,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823869223769182e-05,
+      "loss": 4.0656,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82303062901813e-05,
+      "loss": 4.0443,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822192034267078e-05,
+      "loss": 4.0368,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821353439516026e-05,
+      "loss": 4.0555,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820516482645347e-05,
+      "loss": 4.0541,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819677887894295e-05,
+      "loss": 4.0457,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8188409310236164e-05,
+      "loss": 4.0605,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8180023362725644e-05,
+      "loss": 4.0368,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8171637415215124e-05,
+      "loss": 4.0463,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8163251467704604e-05,
+      "loss": 4.0593,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8154865520194084e-05,
+      "loss": 4.052,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8146479572683564e-05,
+      "loss": 4.0655,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8138093625173044e-05,
+      "loss": 4.0522,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8129707677662524e-05,
+      "loss": 4.0667,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812135448775946e-05,
+      "loss": 4.0449,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811298491905268e-05,
+      "loss": 4.0581,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810459897154216e-05,
+      "loss": 4.0492,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809621302403164e-05,
+      "loss": 4.0498,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808782707652112e-05,
+      "loss": 4.0486,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80794411290106e-05,
+      "loss": 4.0565,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807105518150008e-05,
+      "loss": 4.0562,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806266923398956e-05,
+      "loss": 4.0676,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805428328647904e-05,
+      "loss": 4.0417,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804589733896852e-05,
+      "loss": 4.0459,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037511391458e-05,
+      "loss": 4.0581,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802912544394748e-05,
+      "loss": 4.056,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8020755875240687e-05,
+      "loss": 4.0532,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8012386306533896e-05,
+      "loss": 4.063,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8004000359023376e-05,
+      "loss": 4.055,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7995614411512856e-05,
+      "loss": 4.0586,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798722846400234e-05,
+      "loss": 4.0539,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797884251649182e-05,
+      "loss": 4.0449,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79704565689813e-05,
+      "loss": 4.0586,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796207062147078e-05,
+      "loss": 4.0579,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795368467396026e-05,
+      "loss": 4.0524,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794529872644974e-05,
+      "loss": 4.0513,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793692915774295e-05,
+      "loss": 4.0388,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792854321023243e-05,
+      "loss": 4.0528,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792015726272191e-05,
+      "loss": 4.0635,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791177131521139e-05,
+      "loss": 4.0496,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79034017465046e-05,
+      "loss": 4.0498,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789501579899408e-05,
+      "loss": 4.0457,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788662985148356e-05,
+      "loss": 4.0413,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787824390397304e-05,
+      "loss": 4.047,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786985795646253e-05,
+      "loss": 4.0423,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7861472008952e-05,
+      "loss": 4.0402,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7853102440245216e-05,
+      "loss": 4.0544,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7844716492734696e-05,
+      "loss": 4.0545,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836330545224176e-05,
+      "loss": 4.0441,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782794459771365e-05,
+      "loss": 4.0451,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781955865020313e-05,
+      "loss": 4.0468,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781117270269261e-05,
+      "loss": 4.0546,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780278675518209e-05,
+      "loss": 4.0476,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779440080767157e-05,
+      "loss": 4.0574,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778603123896478e-05,
+      "loss": 4.058,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7777645291454264e-05,
+      "loss": 4.0436,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7769259343943744e-05,
+      "loss": 4.0518,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7760873396433224e-05,
+      "loss": 4.042,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7752487448922704e-05,
+      "loss": 4.0469,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744101501412184e-05,
+      "loss": 4.0521,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7735715553901664e-05,
+      "loss": 4.0433,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772734598519487e-05,
+      "loss": 4.0394,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771896003768435e-05,
+      "loss": 4.0519,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771057409017383e-05,
+      "loss": 4.0479,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770218814266331e-05,
+      "loss": 4.0484,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769380219515279e-05,
+      "loss": 4.0317,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768541624764227e-05,
+      "loss": 4.0481,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767703030013175e-05,
+      "loss": 4.042,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766864435262123e-05,
+      "loss": 4.0577,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766027478391444e-05,
+      "loss": 4.052,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765188883640393e-05,
+      "loss": 4.0491,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764350288889341e-05,
+      "loss": 4.034,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763513332018662e-05,
+      "loss": 4.058,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76267473726761e-05,
+      "loss": 4.051,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761836142516558e-05,
+      "loss": 4.0584,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760997547765506e-05,
+      "loss": 4.0494,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760160590894827e-05,
+      "loss": 4.0519,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759321996143775e-05,
+      "loss": 4.0467,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758483401392723e-05,
+      "loss": 4.0585,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757644806641671e-05,
+      "loss": 4.0307,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756806211890618e-05,
+      "loss": 4.0537,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755967617139567e-05,
+      "loss": 4.0416,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755129022388515e-05,
+      "loss": 4.0462,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7542904276374627e-05,
+      "loss": 4.0389,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7534534707667836e-05,
+      "loss": 4.0407,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7526148760157316e-05,
+      "loss": 4.0437,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7517762812646796e-05,
+      "loss": 4.0479,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.750939324394001e-05,
+      "loss": 4.0542,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.072641849517822,
+      "eval_runtime": 290.1913,
+      "eval_samples_per_second": 1314.963,
+      "eval_steps_per_second": 41.094,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7501007296429485e-05,
+      "loss": 4.0288,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7492621348918965e-05,
+      "loss": 4.038,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7484235401408445e-05,
+      "loss": 4.038,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7475849453897924e-05,
+      "loss": 4.0496,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467463506387404e-05,
+      "loss": 4.0506,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459077558876884e-05,
+      "loss": 4.0368,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7450691611366364e-05,
+      "loss": 4.0406,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744230566385585e-05,
+      "loss": 4.0373,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743391971634533e-05,
+      "loss": 4.0414,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742553376883481e-05,
+      "loss": 4.0507,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741714782132429e-05,
+      "loss": 4.0435,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740876187381377e-05,
+      "loss": 4.0431,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740037592630325e-05,
+      "loss": 4.0488,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7391989978792724e-05,
+      "loss": 4.0279,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7383604031282204e-05,
+      "loss": 4.0411,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7375218083771684e-05,
+      "loss": 4.0292,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7366832136261164e-05,
+      "loss": 4.0402,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7358446188750644e-05,
+      "loss": 4.0454,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7350060241240124e-05,
+      "loss": 4.0313,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734169067253333e-05,
+      "loss": 4.0499,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733330472502282e-05,
+      "loss": 4.0563,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73249187775123e-05,
+      "loss": 4.0436,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731653283000178e-05,
+      "loss": 4.0474,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730814688249126e-05,
+      "loss": 4.0466,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729976093498074e-05,
+      "loss": 4.0467,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729139136627395e-05,
+      "loss": 4.0376,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728300541876343e-05,
+      "loss": 4.0381,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727463585005664e-05,
+      "loss": 4.044,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726624990254612e-05,
+      "loss": 4.0348,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.72578639550356e-05,
+      "loss": 4.0408,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724947800752508e-05,
+      "loss": 4.0456,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724109206001456e-05,
+      "loss": 4.0429,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723270611250404e-05,
+      "loss": 4.0526,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722432016499352e-05,
+      "loss": 4.0442,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721595059628673e-05,
+      "loss": 4.0354,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720756464877621e-05,
+      "loss": 4.0409,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719917870126569e-05,
+      "loss": 4.045,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719079275375517e-05,
+      "loss": 4.0291,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718240680624465e-05,
+      "loss": 4.0296,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717402085873413e-05,
+      "loss": 4.0347,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716565129002734e-05,
+      "loss": 4.0334,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715726534251682e-05,
+      "loss": 4.0414,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71488793950063e-05,
+      "loss": 4.0377,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714049344749578e-05,
+      "loss": 4.04,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.713210749998526e-05,
+      "loss": 4.0356,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7123721552474735e-05,
+      "loss": 4.0421,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711535198376795e-05,
+      "loss": 4.0389,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710696603625744e-05,
+      "loss": 4.0479,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709858008874691e-05,
+      "loss": 4.0139,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709019414123639e-05,
+      "loss": 4.0435,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708180819372587e-05,
+      "loss": 4.0358,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707342224621535e-05,
+      "loss": 4.0478,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706503629870483e-05,
+      "loss": 4.0312,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705665035119431e-05,
+      "loss": 4.032,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7048297161291256e-05,
+      "loss": 4.0292,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703991121378073e-05,
+      "loss": 4.0344,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703152526627021e-05,
+      "loss": 4.0399,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702313931875969e-05,
+      "loss": 4.0211,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7014753371249176e-05,
+      "loss": 4.0478,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7006383802542385e-05,
+      "loss": 4.0329,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6997997855031865e-05,
+      "loss": 4.0111,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6989611907521345e-05,
+      "loss": 4.0433,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6981225960010824e-05,
+      "loss": 4.0262,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6972840012500304e-05,
+      "loss": 4.0141,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6964454064989784e-05,
+      "loss": 4.0354,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6956084496282993e-05,
+      "loss": 4.034,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6947698548772473e-05,
+      "loss": 4.0222,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693932898006568e-05,
+      "loss": 4.0421,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693094303255516e-05,
+      "loss": 4.0133,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692255708504464e-05,
+      "loss": 4.0247,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691417113753413e-05,
+      "loss": 4.0354,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690578519002361e-05,
+      "loss": 4.0354,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.689739924251309e-05,
+      "loss": 4.0405,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688901329500257e-05,
+      "loss": 4.0319,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688062734749205e-05,
+      "loss": 4.0464,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687224139998153e-05,
+      "loss": 4.0244,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686387183127474e-05,
+      "loss": 4.039,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685548588376422e-05,
+      "loss": 4.024,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68470999362537e-05,
+      "loss": 4.028,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683871398874318e-05,
+      "loss": 4.0303,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683032804123266e-05,
+      "loss": 4.0302,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6821974851329596e-05,
+      "loss": 4.0394,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681358890381908e-05,
+      "loss": 4.0449,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680520295630856e-05,
+      "loss": 4.0282,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679681700879804e-05,
+      "loss": 4.026,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678843106128752e-05,
+      "loss": 4.033,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6780045113777e-05,
+      "loss": 4.0386,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677165916626648e-05,
+      "loss": 4.0283,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676327321875596e-05,
+      "loss": 4.0402,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675488727124544e-05,
+      "loss": 4.0372,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674650132373492e-05,
+      "loss": 4.0381,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673813175502813e-05,
+      "loss": 4.0318,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672974580751761e-05,
+      "loss": 4.0249,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672135986000709e-05,
+      "loss": 4.037,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6712973912496565e-05,
+      "loss": 4.0367,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704587964986045e-05,
+      "loss": 4.0353,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669620201747553e-05,
+      "loss": 4.0285,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668781606996501e-05,
+      "loss": 4.019,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667944650125822e-05,
+      "loss": 4.0314,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66710605537477e-05,
+      "loss": 4.0447,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666267460623718e-05,
+      "loss": 4.0313,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665428865872666e-05,
+      "loss": 4.029,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664590271121614e-05,
+      "loss": 4.0281,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663751676370562e-05,
+      "loss": 4.0225,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66291308161951e-05,
+      "loss": 4.0312,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662074486868458e-05,
+      "loss": 4.0168,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661239167878152e-05,
+      "loss": 4.0234,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6604005731271e-05,
+      "loss": 4.0301,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6595619783760485e-05,
+      "loss": 4.0326,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6587233836249965e-05,
+      "loss": 4.0277,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6578847888739445e-05,
+      "loss": 4.0213,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6570478320032654e-05,
+      "loss": 4.0233,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6562092372522134e-05,
+      "loss": 4.0361,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655372280381534e-05,
+      "loss": 4.0259,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654533685630482e-05,
+      "loss": 4.039,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65369509087943e-05,
+      "loss": 4.0332,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652856496128378e-05,
+      "loss": 4.0273,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652017901377326e-05,
+      "loss": 4.0286,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651179306626274e-05,
+      "loss": 4.0254,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650340711875222e-05,
+      "loss": 4.0271,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64950211712417e-05,
+      "loss": 4.0375,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648663522373118e-05,
+      "loss": 4.0191,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.647824927622067e-05,
+      "loss": 4.019,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646986332871015e-05,
+      "loss": 4.0299,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646147738119963e-05,
+      "loss": 4.028,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.645310781249284e-05,
+      "loss": 4.0252,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644473824378605e-05,
+      "loss": 4.0158,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643635229627553e-05,
+      "loss": 4.0291,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642796634876501e-05,
+      "loss": 4.0169,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6419596780058217e-05,
+      "loss": 4.0398,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6411210832547697e-05,
+      "loss": 4.0275,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6402824885037176e-05,
+      "loss": 4.0296,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6394438937526656e-05,
+      "loss": 4.0144,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63860857476236e-05,
+      "loss": 4.0374,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637769980011308e-05,
+      "loss": 4.0302,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636931385260256e-05,
+      "loss": 4.036,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636092790509204e-05,
+      "loss": 4.0355,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635254195758152e-05,
+      "loss": 4.0302,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6344156010071e-05,
+      "loss": 4.023,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633577006256048e-05,
+      "loss": 4.0421,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632738411504996e-05,
+      "loss": 4.0131,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631899816753944e-05,
+      "loss": 4.0336,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631061222002892e-05,
+      "loss": 4.0225,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63022262725184e-05,
+      "loss": 4.026,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6293840325007874e-05,
+      "loss": 4.0193,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628545437749736e-05,
+      "loss": 4.0189,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6277101187594306e-05,
+      "loss": 4.0221,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6268715240083786e-05,
+      "loss": 4.0307,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6260329292573266e-05,
+      "loss": 4.0325,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.059254169464111,
+      "eval_runtime": 292.9685,
+      "eval_samples_per_second": 1302.498,
+      "eval_steps_per_second": 40.704,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6251943345062746e-05,
+      "loss": 4.0189,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6243557397552226e-05,
+      "loss": 4.0164,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62351714500417e-05,
+      "loss": 4.016,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622678550253118e-05,
+      "loss": 4.0302,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621839955502066e-05,
+      "loss": 4.029,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621001360751014e-05,
+      "loss": 4.02,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620162765999962e-05,
+      "loss": 4.0229,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61932417124891e-05,
+      "loss": 4.0205,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6184872143782315e-05,
+      "loss": 4.0215,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6176486196271795e-05,
+      "loss": 4.0272,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6168100248761275e-05,
+      "loss": 4.0266,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6159714301250754e-05,
+      "loss": 4.022,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6151328353740234e-05,
+      "loss": 4.0311,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6142942406229714e-05,
+      "loss": 4.0079,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6134572837522923e-05,
+      "loss": 4.0214,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6126186890012403e-05,
+      "loss": 4.0088,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611780094250188e-05,
+      "loss": 4.0235,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610941499499136e-05,
+      "loss": 4.0285,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610104542628457e-05,
+      "loss": 4.0063,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609265947877405e-05,
+      "loss": 4.0292,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608427353126353e-05,
+      "loss": 4.0366,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607588758375301e-05,
+      "loss": 4.0252,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606750163624249e-05,
+      "loss": 4.0287,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605911568873198e-05,
+      "loss": 4.0295,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605072974122146e-05,
+      "loss": 4.0255,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.604234379371094e-05,
+      "loss": 4.0202,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603395784620042e-05,
+      "loss": 4.0157,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602557189868989e-05,
+      "loss": 4.0201,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601721870878684e-05,
+      "loss": 4.0179,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600883276127632e-05,
+      "loss": 4.0229,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.60004468137658e-05,
+      "loss": 4.0264,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599206086625528e-05,
+      "loss": 4.0296,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598367491874476e-05,
+      "loss": 4.0246,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597528897123424e-05,
+      "loss": 4.0313,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5966919402527446e-05,
+      "loss": 4.0128,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595854983382066e-05,
+      "loss": 4.0224,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595016388631014e-05,
+      "loss": 4.0274,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594177793879962e-05,
+      "loss": 4.0057,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59333919912891e-05,
+      "loss": 4.0141,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592500604377858e-05,
+      "loss": 4.0154,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591662009626806e-05,
+      "loss": 4.0122,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5908234148757535e-05,
+      "loss": 4.0203,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5899848201247015e-05,
+      "loss": 4.0209,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5891462253736495e-05,
+      "loss": 4.0181,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588309268502971e-05,
+      "loss": 4.0197,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5874706737519184e-05,
+      "loss": 4.0204,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586632079000867e-05,
+      "loss": 4.0158,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585793484249815e-05,
+      "loss": 4.0269,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584954889498763e-05,
+      "loss": 4.0005,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584116294747711e-05,
+      "loss": 4.0221,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583277699996659e-05,
+      "loss": 4.0127,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58244074312598e-05,
+      "loss": 4.028,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581602148374928e-05,
+      "loss": 4.0135,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580763553623876e-05,
+      "loss": 4.0098,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579924958872824e-05,
+      "loss": 4.0108,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579086364121772e-05,
+      "loss": 4.015,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57824776937072e-05,
+      "loss": 4.0219,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577410812500041e-05,
+      "loss": 4.0009,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576572217748989e-05,
+      "loss": 4.0288,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575733622997937e-05,
+      "loss": 4.0176,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5748950282468855e-05,
+      "loss": 3.9928,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5740564334958335e-05,
+      "loss": 4.0234,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5732194766251544e-05,
+      "loss": 4.0073,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5723808818741024e-05,
+      "loss": 3.9983,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5715422871230504e-05,
+      "loss": 4.0128,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5707036923719984e-05,
+      "loss": 4.0165,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5698650976209464e-05,
+      "loss": 4.0076,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5690265028698944e-05,
+      "loss": 4.0238,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5681879081188424e-05,
+      "loss": 3.9929,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567350951248163e-05,
+      "loss": 4.0035,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566512356497111e-05,
+      "loss": 4.0187,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565673761746059e-05,
+      "loss": 4.0181,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564835166995007e-05,
+      "loss": 4.019,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563996572243955e-05,
+      "loss": 4.016,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563157977492903e-05,
+      "loss": 4.0226,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562319382741851e-05,
+      "loss": 4.009,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561482425871172e-05,
+      "loss": 4.0213,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56064383112012e-05,
+      "loss": 4.0058,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559805236369068e-05,
+      "loss": 4.0085,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558966641618016e-05,
+      "loss": 4.0084,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558128046866964e-05,
+      "loss": 4.0151,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557291089996285e-05,
+      "loss": 4.0197,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556452495245233e-05,
+      "loss": 4.0253,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555613900494181e-05,
+      "loss": 4.0127,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5547769436235026e-05,
+      "loss": 4.0022,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5539383488724506e-05,
+      "loss": 4.0149,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5530997541213986e-05,
+      "loss": 4.0226,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5522611593703466e-05,
+      "loss": 4.0108,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5514225646192946e-05,
+      "loss": 4.0211,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5505839698682426e-05,
+      "loss": 4.0194,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5497453751171906e-05,
+      "loss": 4.0189,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5489067803661386e-05,
+      "loss": 4.0121,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5480681856150866e-05,
+      "loss": 4.0032,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5472312287444075e-05,
+      "loss": 4.0211,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5463926339933555e-05,
+      "loss": 4.0194,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5455540392423035e-05,
+      "loss": 4.0135,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5447154444912515e-05,
+      "loss": 4.0133,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5438784876205724e-05,
+      "loss": 3.9988,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543039892869521e-05,
+      "loss": 4.0135,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542201298118469e-05,
+      "loss": 4.0235,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541362703367417e-05,
+      "loss": 4.0126,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540524108616365e-05,
+      "loss": 4.0092,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539685513865313e-05,
+      "loss": 4.0092,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538846919114261e-05,
+      "loss": 4.0059,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5380083243632084e-05,
+      "loss": 4.0079,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5371697296121564e-05,
+      "loss": 4.0041,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536334410621851e-05,
+      "loss": 4.0034,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535495815870799e-05,
+      "loss": 4.0119,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534657221119747e-05,
+      "loss": 4.0134,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533820264249068e-05,
+      "loss": 4.0094,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5329816694980164e-05,
+      "loss": 4.0069,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5321430747469644e-05,
+      "loss": 4.003,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5313044799959124e-05,
+      "loss": 4.0156,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5304658852448604e-05,
+      "loss": 4.0092,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296272904938084e-05,
+      "loss": 4.0177,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528788695742756e-05,
+      "loss": 4.0147,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527950100991704e-05,
+      "loss": 4.0093,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527113144121025e-05,
+      "loss": 4.013,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526274549369973e-05,
+      "loss": 4.0068,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5254359546189206e-05,
+      "loss": 4.0079,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5245973598678686e-05,
+      "loss": 4.0152,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5237587651168166e-05,
+      "loss": 4.0041,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5229201703657646e-05,
+      "loss": 4.0031,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522083213495086e-05,
+      "loss": 4.0131,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521244618744034e-05,
+      "loss": 4.0089,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520406023992982e-05,
+      "loss": 4.0041,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51956742924193e-05,
+      "loss": 4.0,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518728834490878e-05,
+      "loss": 4.0064,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517891877620199e-05,
+      "loss": 4.0035,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517053282869147e-05,
+      "loss": 4.0225,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516214688118095e-05,
+      "loss": 4.0073,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515376093367043e-05,
+      "loss": 4.0142,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514539136496364e-05,
+      "loss": 3.9984,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513700541745312e-05,
+      "loss": 4.0139,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.51286194699426e-05,
+      "loss": 4.0128,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512023352243209e-05,
+      "loss": 4.0171,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5111847574921567e-05,
+      "loss": 4.0195,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5103461627411047e-05,
+      "loss": 4.0137,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5095075679900526e-05,
+      "loss": 4.009,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5086689732390006e-05,
+      "loss": 4.0169,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5078320163683216e-05,
+      "loss": 4.0006,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5069934216172696e-05,
+      "loss": 4.0127,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5061548268662175e-05,
+      "loss": 4.0065,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5053162321151655e-05,
+      "loss": 4.0068,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5044792752444865e-05,
+      "loss": 4.0006,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5036423183738074e-05,
+      "loss": 4.0056,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5028037236227554e-05,
+      "loss": 4.0046,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501965128871704e-05,
+      "loss": 4.0068,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501126534120652e-05,
+      "loss": 4.0159,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.048269748687744,
+      "eval_runtime": 291.3423,
+      "eval_samples_per_second": 1309.769,
+      "eval_steps_per_second": 40.931,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5002879393696e-05,
+      "loss": 4.0094,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499449344618548e-05,
+      "loss": 3.9962,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498610749867496e-05,
+      "loss": 4.0004,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497772155116444e-05,
+      "loss": 4.0124,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496933560365392e-05,
+      "loss": 4.0092,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496094965614339e-05,
+      "loss": 4.0051,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495256370863287e-05,
+      "loss": 4.0106,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494417776112235e-05,
+      "loss": 4.0007,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493579181361183e-05,
+      "loss": 4.0026,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492740586610131e-05,
+      "loss": 4.0072,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491901991859079e-05,
+      "loss": 4.0134,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491063397108027e-05,
+      "loss": 4.0031,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490224802356975e-05,
+      "loss": 4.016,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489386207605923e-05,
+      "loss": 3.9873,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488547612854872e-05,
+      "loss": 4.0059,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48770901810382e-05,
+      "loss": 3.9918,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486870423352768e-05,
+      "loss": 4.0056,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486031828601716e-05,
+      "loss": 4.0088,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485194871731037e-05,
+      "loss": 3.9914,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484356276979985e-05,
+      "loss": 4.0095,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483517682228933e-05,
+      "loss": 4.0215,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482679087477881e-05,
+      "loss": 4.0059,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481840492726828e-05,
+      "loss": 4.0127,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481001897975776e-05,
+      "loss": 4.0112,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480164941105098e-05,
+      "loss": 4.0106,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479326346354046e-05,
+      "loss": 4.0026,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478487751602994e-05,
+      "loss": 3.9968,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477649156851942e-05,
+      "loss": 4.005,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.47681056210089e-05,
+      "loss": 3.9999,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475971967349838e-05,
+      "loss": 4.0037,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475133372598786e-05,
+      "loss": 4.0131,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474294777847734e-05,
+      "loss": 4.008,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4734578209770546e-05,
+      "loss": 4.0051,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726192262260026e-05,
+      "loss": 4.0163,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717822693553235e-05,
+      "loss": 3.9999,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4709436746042715e-05,
+      "loss": 4.0058,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4701050798532195e-05,
+      "loss": 4.0052,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4692664851021675e-05,
+      "loss": 3.9904,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4684278903511155e-05,
+      "loss": 3.9993,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467589295600064e-05,
+      "loss": 3.9979,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466750700849012e-05,
+      "loss": 3.9941,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465913743978333e-05,
+      "loss": 4.0036,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465075149227281e-05,
+      "loss": 4.0046,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464236554476229e-05,
+      "loss": 4.0023,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463397959725177e-05,
+      "loss": 4.0036,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462559364974125e-05,
+      "loss": 4.008,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461722408103446e-05,
+      "loss": 3.9967,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460883813352394e-05,
+      "loss": 4.0078,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460045218601342e-05,
+      "loss": 3.987,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45920662385029e-05,
+      "loss": 4.0028,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458368029099238e-05,
+      "loss": 3.9982,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575310722285595e-05,
+      "loss": 4.0129,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4566924774775075e-05,
+      "loss": 3.9977,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4558538827264555e-05,
+      "loss": 3.9918,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550152879754035e-05,
+      "loss": 3.9976,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4541783311047244e-05,
+      "loss": 3.996,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533413742340454e-05,
+      "loss": 4.0075,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4525027794829933e-05,
+      "loss": 3.9787,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516641847319413e-05,
+      "loss": 4.0142,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508255899808893e-05,
+      "loss": 4.0026,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449986995229837e-05,
+      "loss": 3.9764,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449148400478785e-05,
+      "loss": 4.0075,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448309805727733e-05,
+      "loss": 3.9892,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447471210976681e-05,
+      "loss": 3.9825,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446634254106003e-05,
+      "loss": 3.9909,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445795659354951e-05,
+      "loss": 4.0017,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444958702484272e-05,
+      "loss": 3.9949,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444121745613593e-05,
+      "loss": 3.9993,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443283150862541e-05,
+      "loss": 3.9813,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442444556111489e-05,
+      "loss": 3.9867,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441605961360437e-05,
+      "loss": 4.0014,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440767366609385e-05,
+      "loss": 4.005,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4399304097387056e-05,
+      "loss": 4.0036,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4390918149876536e-05,
+      "loss": 3.9992,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4382532202366016e-05,
+      "loss": 4.0064,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4374146254855496e-05,
+      "loss": 3.9936,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436576030734498e-05,
+      "loss": 4.0002,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435737435983446e-05,
+      "loss": 3.9932,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434898841232394e-05,
+      "loss": 3.9924,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340602464813416e-05,
+      "loss": 3.9963,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4332216517302896e-05,
+      "loss": 3.9904,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4323830569792376e-05,
+      "loss": 4.0078,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315444622281856e-05,
+      "loss": 4.0094,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307058674771336e-05,
+      "loss": 4.0001,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429870548486828e-05,
+      "loss": 3.9814,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429031953735776e-05,
+      "loss": 3.9991,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4281933589847234e-05,
+      "loss": 4.0037,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427354764233672e-05,
+      "loss": 3.9964,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42651616948262e-05,
+      "loss": 4.0043,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425677574731568e-05,
+      "loss": 4.006,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424838979980516e-05,
+      "loss": 3.999,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424000385229464e-05,
+      "loss": 3.999,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423161790478412e-05,
+      "loss": 3.9887,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4223264714881065e-05,
+      "loss": 4.0055,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421487876737054e-05,
+      "loss": 3.9999,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420649281986002e-05,
+      "loss": 4.0006,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41981068723495e-05,
+      "loss": 3.9965,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418972092483898e-05,
+      "loss": 3.9834,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418133497732846e-05,
+      "loss": 3.9968,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4172965408621674e-05,
+      "loss": 4.0081,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4164579461111154e-05,
+      "loss": 4.0013,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156193513600634e-05,
+      "loss": 3.9864,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147807566090114e-05,
+      "loss": 3.9975,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139421618579594e-05,
+      "loss": 3.9898,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131035671069074e-05,
+      "loss": 3.9944,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122649723558554e-05,
+      "loss": 3.9918,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411428015485176e-05,
+      "loss": 3.9875,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410589420734124e-05,
+      "loss": 3.9926,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409750825983072e-05,
+      "loss": 3.9965,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40891223123202e-05,
+      "loss": 3.9957,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408073636480968e-05,
+      "loss": 3.9881,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407235041729916e-05,
+      "loss": 3.9908,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406396446978864e-05,
+      "loss": 3.9992,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405557852227812e-05,
+      "loss": 3.9953,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404720895357134e-05,
+      "loss": 3.9986,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403882300606082e-05,
+      "loss": 3.9973,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40304370585503e-05,
+      "loss": 3.9985,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402205111103978e-05,
+      "loss": 3.9968,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401368154233299e-05,
+      "loss": 3.9908,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400529559482247e-05,
+      "loss": 3.9944,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399690964731195e-05,
+      "loss": 3.9982,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398852369980142e-05,
+      "loss": 3.9847,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3980154131094637e-05,
+      "loss": 3.9936,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3971768183584117e-05,
+      "loss": 3.9915,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3963382236073596e-05,
+      "loss": 3.995,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3954996288563076e-05,
+      "loss": 3.9916,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3946610341052556e-05,
+      "loss": 3.9841,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3938224393542036e-05,
+      "loss": 3.9856,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929838446031516e-05,
+      "loss": 3.9885,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921452498520996e-05,
+      "loss": 4.0071,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913082929814205e-05,
+      "loss": 3.9952,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904696982303685e-05,
+      "loss": 3.9944,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896327413596894e-05,
+      "loss": 3.9863,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3887941466086374e-05,
+      "loss": 3.997,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3879555518575854e-05,
+      "loss": 3.999,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871169571065334e-05,
+      "loss": 3.9992,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3862783623554814e-05,
+      "loss": 4.0012,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854397676044294e-05,
+      "loss": 3.9962,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3846011728533774e-05,
+      "loss": 3.9984,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383764215982699e-05,
+      "loss": 4.0032,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382925621231647e-05,
+      "loss": 3.9843,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382087026480595e-05,
+      "loss": 3.9968,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381248431729543e-05,
+      "loss": 3.9911,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380411474858864e-05,
+      "loss": 3.9932,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379572880107812e-05,
+      "loss": 3.9857,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37873428535676e-05,
+      "loss": 3.9892,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377895690605708e-05,
+      "loss": 3.9873,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377057095854656e-05,
+      "loss": 3.9917,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376220138983977e-05,
+      "loss": 3.9968,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.038536548614502,
+      "eval_runtime": 289.49,
+      "eval_samples_per_second": 1318.149,
+      "eval_steps_per_second": 41.193,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375381544232925e-05,
+      "loss": 3.995,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.374542949481873e-05,
+      "loss": 3.9797,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3737043547308215e-05,
+      "loss": 3.9858,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3728657599797694e-05,
+      "loss": 3.9963,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3720288031090904e-05,
+      "loss": 3.9917,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3711902083580384e-05,
+      "loss": 3.9918,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3703516136069863e-05,
+      "loss": 3.9918,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3695130188559343e-05,
+      "loss": 3.9853,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3686744241048823e-05,
+      "loss": 3.986,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36783582935383e-05,
+      "loss": 3.9952,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366997234602778e-05,
+      "loss": 3.9948,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366160277732099e-05,
+      "loss": 3.9895,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365321682981047e-05,
+      "loss": 4.0029,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364483088229995e-05,
+      "loss": 3.9671,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363644493478943e-05,
+      "loss": 3.9894,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362805898727891e-05,
+      "loss": 3.9773,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361968941857213e-05,
+      "loss": 3.9932,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361130347106161e-05,
+      "loss": 3.9907,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360293390235482e-05,
+      "loss": 3.9811,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35945479548443e-05,
+      "loss": 3.9861,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358616200733378e-05,
+      "loss": 4.0091,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357777605982326e-05,
+      "loss": 3.9887,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356939011231273e-05,
+      "loss": 3.9948,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356100416480221e-05,
+      "loss": 3.9983,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355261821729169e-05,
+      "loss": 3.9961,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354423226978117e-05,
+      "loss": 3.9867,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353584632227065e-05,
+      "loss": 3.9822,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3527476753563866e-05,
+      "loss": 3.9898,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351910718485708e-05,
+      "loss": 3.9841,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351072123734656e-05,
+      "loss": 3.9867,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3502335289836035e-05,
+      "loss": 3.9971,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3493949342325515e-05,
+      "loss": 3.9937,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3485563394814995e-05,
+      "loss": 3.9941,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3477177447304475e-05,
+      "loss": 3.9982,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468791499793955e-05,
+      "loss": 3.9841,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3460405552283435e-05,
+      "loss": 3.9917,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3452019604772915e-05,
+      "loss": 3.9909,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3443650036066124e-05,
+      "loss": 3.9736,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3435264088555604e-05,
+      "loss": 3.9894,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342687814104509e-05,
+      "loss": 3.9812,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341849219353457e-05,
+      "loss": 3.9773,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341010624602405e-05,
+      "loss": 3.9891,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340172029851353e-05,
+      "loss": 3.9951,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339335072980674e-05,
+      "loss": 3.9811,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338496478229622e-05,
+      "loss": 3.9884,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33765788347857e-05,
+      "loss": 3.9914,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336819288727518e-05,
+      "loss": 3.9815,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335980693976466e-05,
+      "loss": 3.9953,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335143737105787e-05,
+      "loss": 3.9736,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334305142354735e-05,
+      "loss": 3.9848,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333466547603683e-05,
+      "loss": 3.9862,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332627952852631e-05,
+      "loss": 3.996,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331789358101579e-05,
+      "loss": 3.983,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309524012309004e-05,
+      "loss": 3.9758,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3301138064798484e-05,
+      "loss": 3.9806,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3292752117287964e-05,
+      "loss": 3.9801,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284366169777444e-05,
+      "loss": 3.9955,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.327598022226692e-05,
+      "loss": 3.9673,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32675942747564e-05,
+      "loss": 3.9944,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325922470604961e-05,
+      "loss": 3.9845,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325083875853909e-05,
+      "loss": 3.9676,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3242452811028566e-05,
+      "loss": 3.9877,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3234066863518046e-05,
+      "loss": 3.9778,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3225680916007526e-05,
+      "loss": 3.9686,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321731134730074e-05,
+      "loss": 3.9747,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320892539979022e-05,
+      "loss": 3.9865,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32005394522797e-05,
+      "loss": 3.9823,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319215350476918e-05,
+      "loss": 3.9772,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318376755725866e-05,
+      "loss": 3.9719,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317538160974814e-05,
+      "loss": 3.9722,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316699566223762e-05,
+      "loss": 3.9839,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.31586097147271e-05,
+      "loss": 3.9885,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315025652482404e-05,
+      "loss": 3.9861,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314187057731352e-05,
+      "loss": 3.9918,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3133484629803e-05,
+      "loss": 3.9884,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312509868229248e-05,
+      "loss": 3.9786,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311671273478196e-05,
+      "loss": 3.9851,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3108343166075175e-05,
+      "loss": 3.9802,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3099957218564655e-05,
+      "loss": 3.9802,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3091571271054135e-05,
+      "loss": 3.9754,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3083185323543615e-05,
+      "loss": 3.9787,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3074799376033095e-05,
+      "loss": 3.9945,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3066413428522575e-05,
+      "loss": 3.9928,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3058027481012055e-05,
+      "loss": 3.9885,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3049641533501535e-05,
+      "loss": 3.9679,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3041271964794744e-05,
+      "loss": 3.9846,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3032886017284224e-05,
+      "loss": 3.9889,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302451644857743e-05,
+      "loss": 3.9842,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301613050106691e-05,
+      "loss": 3.9843,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30077445535564e-05,
+      "loss": 3.9952,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299935860604588e-05,
+      "loss": 3.9844,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299097265853536e-05,
+      "loss": 3.9825,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298258671102484e-05,
+      "loss": 3.9767,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297420076351432e-05,
+      "loss": 3.9881,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296583119480753e-05,
+      "loss": 3.9863,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295744524729701e-05,
+      "loss": 3.9893,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294905929978649e-05,
+      "loss": 3.9768,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294067335227597e-05,
+      "loss": 3.9741,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293228740476545e-05,
+      "loss": 3.9805,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292391783605866e-05,
+      "loss": 3.9943,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.291553188854814e-05,
+      "loss": 3.9855,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2907162319841354e-05,
+      "loss": 3.9749,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2898776372330834e-05,
+      "loss": 3.9822,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2890390424820314e-05,
+      "loss": 3.9725,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2882004477309793e-05,
+      "loss": 3.9808,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2873634908603e-05,
+      "loss": 3.9761,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286524896109248e-05,
+      "loss": 3.9745,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.285686301358196e-05,
+      "loss": 3.9823,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284847706607144e-05,
+      "loss": 3.9806,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.284010749736465e-05,
+      "loss": 3.9777,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.283172154985413e-05,
+      "loss": 3.9757,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.282333560234361e-05,
+      "loss": 3.977,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281494965483309e-05,
+      "loss": 3.9839,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280656370732257e-05,
+      "loss": 3.9816,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279817775981205e-05,
+      "loss": 3.9856,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278979181230153e-05,
+      "loss": 3.9796,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278140586479101e-05,
+      "loss": 3.9871,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277301991728049e-05,
+      "loss": 3.9809,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276463396976997e-05,
+      "loss": 3.9756,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275624802225945e-05,
+      "loss": 3.9832,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274787845355266e-05,
+      "loss": 3.9795,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273949250604214e-05,
+      "loss": 3.9721,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273110655853162e-05,
+      "loss": 3.9769,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27227206110211e-05,
+      "loss": 3.9797,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271435104231431e-05,
+      "loss": 3.9792,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270596509480379e-05,
+      "loss": 3.9715,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269757914729327e-05,
+      "loss": 3.9737,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2689193199782756e-05,
+      "loss": 3.9735,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2680807252272236e-05,
+      "loss": 3.9758,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2672437683565445e-05,
+      "loss": 3.9919,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2664051736054925e-05,
+      "loss": 3.9809,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2655665788544405e-05,
+      "loss": 3.9823,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2647279841033885e-05,
+      "loss": 3.9694,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2638893893523365e-05,
+      "loss": 3.9848,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2630507946012845e-05,
+      "loss": 3.9846,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2622121998502325e-05,
+      "loss": 3.9861,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2613736050991805e-05,
+      "loss": 3.9826,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2605350103481284e-05,
+      "loss": 3.9835,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2596980534774494e-05,
+      "loss": 3.9879,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258861096606771e-05,
+      "loss": 3.9886,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258022501855719e-05,
+      "loss": 3.9718,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257183907104667e-05,
+      "loss": 3.9803,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.256345312353615e-05,
+      "loss": 3.9784,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255506717602563e-05,
+      "loss": 3.9786,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254668122851511e-05,
+      "loss": 3.9719,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253829528100459e-05,
+      "loss": 3.9766,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252990933349406e-05,
+      "loss": 3.9703,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252153976478728e-05,
+      "loss": 3.983,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251315381727676e-05,
+      "loss": 3.977,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.030486583709717,
+      "eval_runtime": 297.7712,
+      "eval_samples_per_second": 1281.491,
+      "eval_steps_per_second": 40.048,
+      "step": 1068480
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.385857090665577e+17,
+  "trial_name": null,
+  "trial_params": null
+}