diff --git "a/checkpoint-1602720/trainer_state.json" "b/checkpoint-1602720/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1602720/trainer_state.json"
@@ -0,0 +1,18973 @@
+{
+  "best_metric": 4.008484840393066,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/superlative-quantifier/lstm/1/checkpoints/checkpoint-1602720",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1602720,
+  "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": 10.8209,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.557,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0547,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9882,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9429,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8951,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994970069374061e-05,
+      "loss": 6.7146,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994131474623009e-05,
+      "loss": 6.6081,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5181,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4469,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3851,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3131,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2423,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1805,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.123,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0739,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0293,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9925,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.943,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9161,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8796,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8333,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8037,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7688,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7471,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7153,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6855,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6566,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6513,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6118,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.5972,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5894,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5598,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5408,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714926921053515e-05,
+      "loss": 5.5362,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5146,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.5045,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.4656,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9681399509815164e-05,
+      "loss": 5.4595,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4311,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 5.4277,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 5.4151,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4104,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.3795,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3789,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.365,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3665,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.358,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3299,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.316,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3244,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3189,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.2938,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.2848,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2809,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2632,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2712,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2366,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.24,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2305,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.229,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2061,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2129,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.1864,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.1922,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.1918,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.1842,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1634,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1478,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1405,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1551,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1436,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1286,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1376,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1374,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.1211,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1104,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.1002,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.0968,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.082,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932932074480317e-05,
+      "loss": 5.0842,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0732,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0849,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0713,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0493,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0523,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0511,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.0486,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0441,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0264,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0252,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0211,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0072,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 4.9976,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 4.9969,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0045,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.9924,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9819,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9724,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.9749,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.9763,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.9693,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914489541478666e-05,
+      "loss": 4.9624,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913650946727614e-05,
+      "loss": 4.9507,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9608,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9384,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9429,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9356,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.927,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9237,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9282,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9161,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9142,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9108,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9062,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9093,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.903,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.8986,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.8964,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.8898,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.8909,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8778,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8732,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8721,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8707,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.865,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8625,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.861,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.848,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8601,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8557,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8375,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8241,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8308,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8406,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8466,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8287,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8281,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8281,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8269,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8123,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881797449230622e-05,
+      "loss": 4.8077,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88095885447957e-05,
+      "loss": 4.8145,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880120259728518e-05,
+      "loss": 4.8058,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879281664977466e-05,
+      "loss": 4.8026,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.8055,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.7899,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.8026,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.7962,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.7763,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.761977195739746,
+      "eval_runtime": 292.7746,
+      "eval_samples_per_second": 1303.361,
+      "eval_steps_per_second": 40.731,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.7889,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.7787,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.79,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7771,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7778,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7637,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7638,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7445,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7717,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7626,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7587,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7638,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864196786740769e-05,
+      "loss": 4.7474,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863358191989717e-05,
+      "loss": 4.7456,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862519597238665e-05,
+      "loss": 4.74,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861681002487613e-05,
+      "loss": 4.7389,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860842407736561e-05,
+      "loss": 4.7394,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.731,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7419,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7439,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.731,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7228,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.7201,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.7265,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7183,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7051,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.7063,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7197,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.6935,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849943951733631e-05,
+      "loss": 4.6993,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849105356982579e-05,
+      "loss": 4.7083,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.7049,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.7036,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.7021,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.6988,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.6995,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6776,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6871,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6715,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6762,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6769,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8398873662425e-05,
+      "loss": 4.6767,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.6728,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6698,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.6686,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.6736,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6761,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6586,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6502,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.6651,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323416513634046e-05,
+      "loss": 4.6696,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315030566123526e-05,
+      "loss": 4.655,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306644618613006e-05,
+      "loss": 4.6493,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.65,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.6439,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281503154885175e-05,
+      "loss": 4.653,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.6323,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264731259864135e-05,
+      "loss": 4.6442,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.6409,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6418,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823958979613631e-05,
+      "loss": 4.6222,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823120384862579e-05,
+      "loss": 4.644,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822281790111527e-05,
+      "loss": 4.6171,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6338,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6344,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819767643738744e-05,
+      "loss": 4.6327,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.6175,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6122,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.6102,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6219,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.623,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814737713112805e-05,
+      "loss": 4.615,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.6217,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.6273,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.6232,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.6044,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.6116,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809707782486866e-05,
+      "loss": 4.6124,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.5902,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808032230865135e-05,
+      "loss": 4.6105,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8071936361140827e-05,
+      "loss": 4.5939,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6141,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.6044,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046794897413e-05,
+      "loss": 4.5821,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803840894990248e-05,
+      "loss": 4.5998,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803002300239196e-05,
+      "loss": 4.589,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.6028,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.5945,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800488153866413e-05,
+      "loss": 4.5906,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.5872,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.5834,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.5768,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797133774862205e-05,
+      "loss": 4.5713,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796295180111153e-05,
+      "loss": 4.5752,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.5851,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5756,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793781033738371e-05,
+      "loss": 4.5709,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792942438987319e-05,
+      "loss": 4.565,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.5723,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.5783,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904282926145356e-05,
+      "loss": 4.5701,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895896978634836e-05,
+      "loss": 4.5656,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887511031124316e-05,
+      "loss": 4.5653,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879125083613796e-05,
+      "loss": 4.5693,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870755514907005e-05,
+      "loss": 4.5519,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862385946200214e-05,
+      "loss": 4.5614,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785401637749343e-05,
+      "loss": 4.5582,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784563042998291e-05,
+      "loss": 4.5506,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783724448247239e-05,
+      "loss": 4.5513,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782885853496187e-05,
+      "loss": 4.5558,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782047258745135e-05,
+      "loss": 4.5502,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781208663994083e-05,
+      "loss": 4.5469,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780370069243031e-05,
+      "loss": 4.5482,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779531474491979e-05,
+      "loss": 4.545,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.5566,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777855922870248e-05,
+      "loss": 4.5514,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777017328119196e-05,
+      "loss": 4.5469,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776178733368144e-05,
+      "loss": 4.5472,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775340138617092e-05,
+      "loss": 4.5468,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7745031817464134e-05,
+      "loss": 4.545,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7736645869953614e-05,
+      "loss": 4.5364,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728259922443094e-05,
+      "loss": 4.5342,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719873974932574e-05,
+      "loss": 4.5381,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711488027422054e-05,
+      "loss": 4.5387,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770311845871526e-05,
+      "loss": 4.535,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769473251120474e-05,
+      "loss": 4.536,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768634656369422e-05,
+      "loss": 4.5302,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76779606161837e-05,
+      "loss": 4.5317,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766957466867318e-05,
+      "loss": 4.5318,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766120509996639e-05,
+      "loss": 4.5389,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765281915245587e-05,
+      "loss": 4.5205,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764443320494535e-05,
+      "loss": 4.5106,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763604725743483e-05,
+      "loss": 4.5165,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762766130992431e-05,
+      "loss": 4.5301,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761927536241379e-05,
+      "loss": 4.5331,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761088941490327e-05,
+      "loss": 4.5252,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760250346739275e-05,
+      "loss": 4.5226,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759413389868596e-05,
+      "loss": 4.5264,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758576432997918e-05,
+      "loss": 4.5255,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757737838246866e-05,
+      "loss": 4.5155,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7568992434958137e-05,
+      "loss": 4.5131,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756060648744761e-05,
+      "loss": 4.5189,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755222053993709e-05,
+      "loss": 4.5139,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754383459242657e-05,
+      "loss": 4.509,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753544864491605e-05,
+      "loss": 4.5209,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527062697405536e-05,
+      "loss": 4.5029,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518693128698745e-05,
+      "loss": 4.5161,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510307181188225e-05,
+      "loss": 4.5105,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501921233677705e-05,
+      "loss": 4.4989,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.4853105545043945,
+      "eval_runtime": 294.3628,
+      "eval_samples_per_second": 1296.329,
+      "eval_steps_per_second": 40.511,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493535286167185e-05,
+      "loss": 4.51,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485149338656665e-05,
+      "loss": 4.5002,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476763391146145e-05,
+      "loss": 4.5093,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468377443635625e-05,
+      "loss": 4.502,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459991496125105e-05,
+      "loss": 4.507,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451605548614585e-05,
+      "loss": 4.494,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443219601104065e-05,
+      "loss": 4.491,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434833653593545e-05,
+      "loss": 4.479,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426447706083025e-05,
+      "loss": 4.5096,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418061758572505e-05,
+      "loss": 4.4995,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409675811061985e-05,
+      "loss": 4.4911,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401289863551465e-05,
+      "loss": 4.5075,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739292029484468e-05,
+      "loss": 4.4844,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384534347334154e-05,
+      "loss": 4.4904,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376148399823634e-05,
+      "loss": 4.4828,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367762452313114e-05,
+      "loss": 4.4799,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359376504802594e-05,
+      "loss": 4.4882,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73510069360958e-05,
+      "loss": 4.4769,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734262098858528e-05,
+      "loss": 4.4909,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733423504107476e-05,
+      "loss": 4.4923,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732584909356424e-05,
+      "loss": 4.4842,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731746314605372e-05,
+      "loss": 4.4785,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73090771985432e-05,
+      "loss": 4.4775,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730069125103269e-05,
+      "loss": 4.4838,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729230530352217e-05,
+      "loss": 4.4711,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728393573481538e-05,
+      "loss": 4.4705,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727554978730486e-05,
+      "loss": 4.4609,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726716383979434e-05,
+      "loss": 4.4818,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725877789228382e-05,
+      "loss": 4.4558,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72503919447733e-05,
+      "loss": 4.4659,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724202237606651e-05,
+      "loss": 4.4717,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723363642855599e-05,
+      "loss": 4.4709,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722525048104547e-05,
+      "loss": 4.47,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721686453353495e-05,
+      "loss": 4.472,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7208494964828156e-05,
+      "loss": 4.4668,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720012539612137e-05,
+      "loss": 4.4723,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719173944861085e-05,
+      "loss": 4.4542,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718335350110033e-05,
+      "loss": 4.4599,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717496755358981e-05,
+      "loss": 4.4447,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716658160607929e-05,
+      "loss": 4.4536,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715819565856877e-05,
+      "loss": 4.4509,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714980971105825e-05,
+      "loss": 4.4603,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714142376354773e-05,
+      "loss": 4.4508,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713303781603721e-05,
+      "loss": 4.4521,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124651868526685e-05,
+      "loss": 4.4536,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116265921016165e-05,
+      "loss": 4.4554,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710789635230938e-05,
+      "loss": 4.4594,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709951040479886e-05,
+      "loss": 4.4439,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709112445728834e-05,
+      "loss": 4.4333,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708273850977782e-05,
+      "loss": 4.456,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70743525622673e-05,
+      "loss": 4.4543,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7065982993560517e-05,
+      "loss": 4.4451,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057613424853726e-05,
+      "loss": 4.4437,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049227477343206e-05,
+      "loss": 4.437,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040841529832686e-05,
+      "loss": 4.4398,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703245558232216e-05,
+      "loss": 4.4458,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702406963481164e-05,
+      "loss": 4.4285,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701568368730112e-05,
+      "loss": 4.4403,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70072977397906e-05,
+      "loss": 4.4364,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699891179228008e-05,
+      "loss": 4.4354,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990542223573294e-05,
+      "loss": 4.4237,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698217265486651e-05,
+      "loss": 4.4447,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973786707355983e-05,
+      "loss": 4.4215,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965400759845463e-05,
+      "loss": 4.4342,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6957014812334943e-05,
+      "loss": 4.4319,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694862886482442e-05,
+      "loss": 4.4388,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694025929611763e-05,
+      "loss": 4.4205,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693187334860711e-05,
+      "loss": 4.4216,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692348740109659e-05,
+      "loss": 4.4112,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691510145358607e-05,
+      "loss": 4.4307,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690671550607555e-05,
+      "loss": 4.4322,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689832955856503e-05,
+      "loss": 4.4251,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688994361105451e-05,
+      "loss": 4.429,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881557663544e-05,
+      "loss": 4.4376,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687318809483721e-05,
+      "loss": 4.4363,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686483490493415e-05,
+      "loss": 4.4133,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685644895742363e-05,
+      "loss": 4.4227,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848063009913106e-05,
+      "loss": 4.4287,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839677062402586e-05,
+      "loss": 4.4071,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831291114892066e-05,
+      "loss": 4.4251,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682292154618528e-05,
+      "loss": 4.4082,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814535598674755e-05,
+      "loss": 4.432,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806149651164235e-05,
+      "loss": 4.4238,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679776370365372e-05,
+      "loss": 4.4018,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67893777561432e-05,
+      "loss": 4.4137,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678099180863268e-05,
+      "loss": 4.4119,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677260586112216e-05,
+      "loss": 4.4209,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676421991361164e-05,
+      "loss": 4.4167,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675583396610112e-05,
+      "loss": 4.413,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674746439739433e-05,
+      "loss": 4.4098,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673907844988381e-05,
+      "loss": 4.4057,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673069250237329e-05,
+      "loss": 4.4022,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672230655486277e-05,
+      "loss": 4.3913,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671392060735225e-05,
+      "loss": 4.4031,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670555103864546e-05,
+      "loss": 4.4123,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669716509113494e-05,
+      "loss": 4.404,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668877914362442e-05,
+      "loss": 4.3984,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680393196113906e-05,
+      "loss": 4.3939,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672007248603386e-05,
+      "loss": 4.3985,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663621301092866e-05,
+      "loss": 4.4121,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655235353582346e-05,
+      "loss": 4.4003,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664684940607182e-05,
+      "loss": 4.3985,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66384634585613e-05,
+      "loss": 4.3923,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630093889854515e-05,
+      "loss": 4.4051,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621707942343995e-05,
+      "loss": 4.3836,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661332199483347e-05,
+      "loss": 4.3997,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604952426126684e-05,
+      "loss": 4.39,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659658285741989e-05,
+      "loss": 4.3919,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658819690990937e-05,
+      "loss": 4.3865,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657981096239886e-05,
+      "loss": 4.3909,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657142501488834e-05,
+      "loss": 4.3898,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656303906737782e-05,
+      "loss": 4.3834,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655465311986729e-05,
+      "loss": 4.391,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654626717235677e-05,
+      "loss": 4.3852,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653788122484625e-05,
+      "loss": 4.3933,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652949527733573e-05,
+      "loss": 4.3924,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652112570862894e-05,
+      "loss": 4.3897,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651273976111842e-05,
+      "loss": 4.3902,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65043538136079e-05,
+      "loss": 4.3861,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649596786609738e-05,
+      "loss": 4.3829,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648758191858686e-05,
+      "loss": 4.3815,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647919597107634e-05,
+      "loss": 4.3809,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647081002356582e-05,
+      "loss": 4.3786,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646244045485904e-05,
+      "loss": 4.3873,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645405450734852e-05,
+      "loss": 4.3788,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445668559838e-05,
+      "loss": 4.3847,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643728261232748e-05,
+      "loss": 4.3789,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642889666481696e-05,
+      "loss": 4.3784,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642051071730644e-05,
+      "loss": 4.3759,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641212476979592e-05,
+      "loss": 4.3901,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64037388222854e-05,
+      "loss": 4.3704,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6395369253578606e-05,
+      "loss": 4.3659,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6386983306068086e-05,
+      "loss": 4.3633,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6378613737361295e-05,
+      "loss": 4.3783,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6370227789850775e-05,
+      "loss": 4.3838,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636184184234026e-05,
+      "loss": 4.3812,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635345589482974e-05,
+      "loss": 4.3732,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634506994731922e-05,
+      "loss": 4.3809,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63366839998087e-05,
+      "loss": 4.3752,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632829805229818e-05,
+      "loss": 4.3754,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6319912104787655e-05,
+      "loss": 4.3647,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631154253608087e-05,
+      "loss": 4.3757,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630317296737408e-05,
+      "loss": 4.3704,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629478701986356e-05,
+      "loss": 4.3616,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628640107235304e-05,
+      "loss": 4.3764,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627801512484252e-05,
+      "loss": 4.3659,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269629177332e-05,
+      "loss": 4.3678,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626124322982148e-05,
+      "loss": 4.3691,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625285728231096e-05,
+      "loss": 4.3556,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.3511128425598145,
+      "eval_runtime": 293.5882,
+      "eval_samples_per_second": 1299.749,
+      "eval_steps_per_second": 40.618,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624447133480044e-05,
+      "loss": 4.3691,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623608538728992e-05,
+      "loss": 4.3603,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622771581858313e-05,
+      "loss": 4.368,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621932987107261e-05,
+      "loss": 4.3606,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621094392356209e-05,
+      "loss": 4.3711,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202574354855305e-05,
+      "loss": 4.3562,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619418840734478e-05,
+      "loss": 4.3524,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618580245983426e-05,
+      "loss": 4.3442,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617741651232374e-05,
+      "loss": 4.3698,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616903056481322e-05,
+      "loss": 4.3668,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160660996106434e-05,
+      "loss": 4.351,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152275048595913e-05,
+      "loss": 4.3688,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614390547988913e-05,
+      "loss": 4.3514,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61355195323786e-05,
+      "loss": 4.3544,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612713358486808e-05,
+      "loss": 4.3519,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611874763735756e-05,
+      "loss": 4.3446,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611037806865078e-05,
+      "loss": 4.3528,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610199212114025e-05,
+      "loss": 4.3462,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609360617362973e-05,
+      "loss": 4.3577,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608522022611921e-05,
+      "loss": 4.3576,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607683427860869e-05,
+      "loss": 4.3574,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606844833109817e-05,
+      "loss": 4.3482,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606006238358765e-05,
+      "loss": 4.3436,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605169281488087e-05,
+      "loss": 4.3564,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604330686737035e-05,
+      "loss": 4.3361,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6034937298663556e-05,
+      "loss": 4.3432,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026551351153036e-05,
+      "loss": 4.3342,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018165403642516e-05,
+      "loss": 4.3516,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009779456131996e-05,
+      "loss": 4.3283,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001393508621476e-05,
+      "loss": 4.3384,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5993007561110956e-05,
+      "loss": 4.3469,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5984621613600436e-05,
+      "loss": 4.3407,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5976235666089916e-05,
+      "loss": 4.3431,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5967849718579396e-05,
+      "loss": 4.3444,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959480149872605e-05,
+      "loss": 4.342,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595109420236209e-05,
+      "loss": 4.346,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594270825485157e-05,
+      "loss": 4.3341,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593432230734105e-05,
+      "loss": 4.3271,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592593635983053e-05,
+      "loss": 4.3251,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591755041232001e-05,
+      "loss": 4.3298,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590916446480949e-05,
+      "loss": 4.3268,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900778517298965e-05,
+      "loss": 4.3387,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892392569788445e-05,
+      "loss": 4.3273,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588403937988539e-05,
+      "loss": 4.3304,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587565343237487e-05,
+      "loss": 4.3347,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586726748486435e-05,
+      "loss": 4.3286,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585888153735383e-05,
+      "loss": 4.3416,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585049558984331e-05,
+      "loss": 4.324,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5842126021136525e-05,
+      "loss": 4.3089,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833740073626005e-05,
+      "loss": 4.3405,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825354126115485e-05,
+      "loss": 4.3321,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816968178604965e-05,
+      "loss": 4.3318,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580858223109444e-05,
+      "loss": 4.3221,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580019628358392e-05,
+      "loss": 4.3206,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57918103360734e-05,
+      "loss": 4.3219,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578342438856288e-05,
+      "loss": 4.3258,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577505481985609e-05,
+      "loss": 4.3137,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576666887234557e-05,
+      "loss": 4.3181,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575828292483505e-05,
+      "loss": 4.3261,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574989697732453e-05,
+      "loss": 4.3174,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574152740861774e-05,
+      "loss": 4.3068,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573315783991096e-05,
+      "loss": 4.3309,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572477189240044e-05,
+      "loss": 4.306,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571638594488991e-05,
+      "loss": 4.3143,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570799999737939e-05,
+      "loss": 4.3228,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569961404986887e-05,
+      "loss": 4.3256,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569124448116208e-05,
+      "loss": 4.3077,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568285853365156e-05,
+      "loss": 4.3107,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567447258614104e-05,
+      "loss": 4.2946,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566608663863052e-05,
+      "loss": 4.3229,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565770069112e-05,
+      "loss": 4.3165,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564931474360948e-05,
+      "loss": 4.3155,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564092879609896e-05,
+      "loss": 4.3129,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563254284858845e-05,
+      "loss": 4.3262,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624173279881657e-05,
+      "loss": 4.3274,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615803711174866e-05,
+      "loss": 4.3031,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5607417763664346e-05,
+      "loss": 4.312,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599031816153826e-05,
+      "loss": 4.3192,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590645868643306e-05,
+      "loss": 4.2956,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582259921132786e-05,
+      "loss": 4.3169,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5573873973622265e-05,
+      "loss": 4.3018,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565488026111745e-05,
+      "loss": 4.3183,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557102078601225e-05,
+      "loss": 4.3171,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548732509894434e-05,
+      "loss": 4.2962,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5540346562383914e-05,
+      "loss": 4.3076,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55319606148734e-05,
+      "loss": 4.3029,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552357466736288e-05,
+      "loss": 4.3165,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551520509865609e-05,
+      "loss": 4.3063,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550681915114557e-05,
+      "loss": 4.3053,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549843320363505e-05,
+      "loss": 4.3028,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549004725612453e-05,
+      "loss": 4.3016,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548167768741774e-05,
+      "loss": 4.298,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547329173990722e-05,
+      "loss": 4.2865,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54649057923967e-05,
+      "loss": 4.2994,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545651984488618e-05,
+      "loss": 4.3065,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544813389737566e-05,
+      "loss": 4.3055,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543976432866887e-05,
+      "loss": 4.2869,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431378381158355e-05,
+      "loss": 4.2931,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422992433647835e-05,
+      "loss": 4.2951,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414622864941044e-05,
+      "loss": 4.3073,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406236917430524e-05,
+      "loss": 4.2966,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539786734872373e-05,
+      "loss": 4.2978,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538948140121321e-05,
+      "loss": 4.2895,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538109545370269e-05,
+      "loss": 4.3054,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53727258849959e-05,
+      "loss": 4.2802,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536433993748538e-05,
+      "loss": 4.2977,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535595398997486e-05,
+      "loss": 4.2886,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534756804246434e-05,
+      "loss": 4.2913,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533918209495382e-05,
+      "loss": 4.2839,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533079614744331e-05,
+      "loss": 4.2975,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532241019993279e-05,
+      "loss": 4.2851,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531402425242227e-05,
+      "loss": 4.2843,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530563830491175e-05,
+      "loss": 4.2897,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529726873620496e-05,
+      "loss": 4.2879,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528888278869444e-05,
+      "loss": 4.2972,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528049684118392e-05,
+      "loss": 4.2952,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5272127272477126e-05,
+      "loss": 4.2916,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263741324966606e-05,
+      "loss": 4.2964,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255355377456086e-05,
+      "loss": 4.2885,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246985808749295e-05,
+      "loss": 4.2831,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238599861238775e-05,
+      "loss": 4.2825,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230213913728255e-05,
+      "loss": 4.2829,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522182796621774e-05,
+      "loss": 4.2856,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521344201870722e-05,
+      "loss": 4.2921,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5205056071196695e-05,
+      "loss": 4.281,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196670123686175e-05,
+      "loss": 4.2906,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188284176175655e-05,
+      "loss": 4.2794,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5179898228665135e-05,
+      "loss": 4.2832,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171512281154615e-05,
+      "loss": 4.2806,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5163126333644095e-05,
+      "loss": 4.2947,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5154740386133575e-05,
+      "loss": 4.2778,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5146354438623055e-05,
+      "loss": 4.2706,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5137968491112535e-05,
+      "loss": 4.2659,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5129598922405744e-05,
+      "loss": 4.2836,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5121212974895224e-05,
+      "loss": 4.2888,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511282702738471e-05,
+      "loss": 4.2888,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510444107987419e-05,
+      "loss": 4.283,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509605513236367e-05,
+      "loss": 4.2877,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508766918485315e-05,
+      "loss": 4.2817,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507928323734263e-05,
+      "loss": 4.2792,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070897289832104e-05,
+      "loss": 4.2747,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506252772112532e-05,
+      "loss": 4.2823,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50541417736148e-05,
+      "loss": 4.2791,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504575582610428e-05,
+      "loss": 4.2672,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503736987859375e-05,
+      "loss": 4.2856,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502898393108323e-05,
+      "loss": 4.2784,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502061436237645e-05,
+      "loss": 4.2715,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501222841486593e-05,
+      "loss": 4.2762,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500384246735541e-05,
+      "loss": 4.2721,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.268988132476807,
+      "eval_runtime": 290.8232,
+      "eval_samples_per_second": 1312.107,
+      "eval_steps_per_second": 41.004,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499545651984489e-05,
+      "loss": 4.2769,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498707057233437e-05,
+      "loss": 4.2721,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497868462482385e-05,
+      "loss": 4.2804,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497029867731333e-05,
+      "loss": 4.2681,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.496191272980281e-05,
+      "loss": 4.2811,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.495352678229229e-05,
+      "loss": 4.2708,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.494514083478177e-05,
+      "loss": 4.2591,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.493675488727125e-05,
+      "loss": 4.2602,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492836893976073e-05,
+      "loss": 4.2753,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491998299225021e-05,
+      "loss": 4.2759,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491159704473969e-05,
+      "loss": 4.2658,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490321109722917e-05,
+      "loss": 4.2782,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489484152852238e-05,
+      "loss": 4.2673,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886455581011864e-05,
+      "loss": 4.268,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4878069633501344e-05,
+      "loss": 4.2637,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869683685990824e-05,
+      "loss": 4.2592,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48612977384803e-05,
+      "loss": 4.2634,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485291179096978e-05,
+      "loss": 4.2589,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484452584345926e-05,
+      "loss": 4.2729,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483613989594874e-05,
+      "loss": 4.2712,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4827770327241946e-05,
+      "loss": 4.2728,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4819384379731426e-05,
+      "loss": 4.2638,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4810998432220906e-05,
+      "loss": 4.2571,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4802612484710386e-05,
+      "loss": 4.2683,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4794226537199866e-05,
+      "loss": 4.256,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4785840589689346e-05,
+      "loss": 4.2579,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4777454642178825e-05,
+      "loss": 4.248,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476908507347204e-05,
+      "loss": 4.2614,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476069912596152e-05,
+      "loss": 4.25,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4752313178451e-05,
+      "loss": 4.2496,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474392723094048e-05,
+      "loss": 4.2621,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473554128342996e-05,
+      "loss": 4.2562,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472715533591944e-05,
+      "loss": 4.2642,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471876938840892e-05,
+      "loss": 4.2638,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471039981970213e-05,
+      "loss": 4.2539,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470201387219161e-05,
+      "loss": 4.2654,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469362792468109e-05,
+      "loss": 4.2553,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468524197717057e-05,
+      "loss": 4.2429,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467685602966005e-05,
+      "loss": 4.2403,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466847008214953e-05,
+      "loss": 4.2473,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466008413463901e-05,
+      "loss": 4.2472,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465169818712849e-05,
+      "loss": 4.2586,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643328618421706e-05,
+      "loss": 4.2458,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463494267091118e-05,
+      "loss": 4.2472,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462655672340066e-05,
+      "loss": 4.2553,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461817077589014e-05,
+      "loss": 4.2509,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460978482837962e-05,
+      "loss": 4.2581,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460141525967283e-05,
+      "loss": 4.242,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459302931216231e-05,
+      "loss": 4.2251,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458464336465179e-05,
+      "loss": 4.2585,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457625741714127e-05,
+      "loss": 4.2572,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567887848434484e-05,
+      "loss": 4.2518,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559501900923964e-05,
+      "loss": 4.2446,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551115953413444e-05,
+      "loss": 4.2403,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4542730005902923e-05,
+      "loss": 4.2426,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453436043719613e-05,
+      "loss": 4.2439,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452597448968561e-05,
+      "loss": 4.2361,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451760492097883e-05,
+      "loss": 4.2398,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45092189734683e-05,
+      "loss": 4.2489,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450083302595778e-05,
+      "loss": 4.2415,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449244707844726e-05,
+      "loss": 4.2299,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448406113093674e-05,
+      "loss": 4.2533,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447567518342622e-05,
+      "loss": 4.2288,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44672892359157e-05,
+      "loss": 4.231,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445890328840519e-05,
+      "loss": 4.2481,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44505337196984e-05,
+      "loss": 4.2502,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444214777218788e-05,
+      "loss": 4.2283,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443376182467736e-05,
+      "loss": 4.2351,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442537587716684e-05,
+      "loss": 4.2165,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4417006308460046e-05,
+      "loss": 4.2445,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408620360949526e-05,
+      "loss": 4.244,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400234413439006e-05,
+      "loss": 4.2349,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391848465928486e-05,
+      "loss": 4.2404,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4383495276025424e-05,
+      "loss": 4.247,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437510932851491e-05,
+      "loss": 4.2561,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436672338100439e-05,
+      "loss": 4.2259,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435833743349387e-05,
+      "loss": 4.24,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434995148598335e-05,
+      "loss": 4.2435,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434156553847283e-05,
+      "loss": 4.2198,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433319596976604e-05,
+      "loss": 4.2444,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432481002225552e-05,
+      "loss": 4.2296,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316424074745e-05,
+      "loss": 4.2426,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430803812723448e-05,
+      "loss": 4.2423,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429966855852769e-05,
+      "loss": 4.2239,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429128261101717e-05,
+      "loss": 4.2288,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428289666350665e-05,
+      "loss": 4.2284,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427451071599613e-05,
+      "loss": 4.2451,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426612476848561e-05,
+      "loss": 4.2327,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257738820975095e-05,
+      "loss": 4.2308,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249352873464575e-05,
+      "loss": 4.2271,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240966925954055e-05,
+      "loss": 4.2315,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232597357247264e-05,
+      "loss": 4.2263,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224211409736744e-05,
+      "loss": 4.2128,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215841841029954e-05,
+      "loss": 4.2291,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207455893519433e-05,
+      "loss": 4.2345,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4199069946008913e-05,
+      "loss": 4.2339,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419068399849839e-05,
+      "loss": 4.2152,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418229805098787e-05,
+      "loss": 4.2196,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417391210347735e-05,
+      "loss": 4.2249,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416554253477056e-05,
+      "loss": 4.2318,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415717296606378e-05,
+      "loss": 4.2298,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414878701855326e-05,
+      "loss": 4.2252,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414040107104274e-05,
+      "loss": 4.2208,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413201512353222e-05,
+      "loss": 4.2328,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41236291760217e-05,
+      "loss": 4.209,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411524322851118e-05,
+      "loss": 4.227,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410685728100066e-05,
+      "loss": 4.2141,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409847133349014e-05,
+      "loss": 4.2256,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409008538597961e-05,
+      "loss": 4.21,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408171581727283e-05,
+      "loss": 4.228,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407332986976231e-05,
+      "loss": 4.2179,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406494392225179e-05,
+      "loss": 4.2153,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405655797474127e-05,
+      "loss": 4.2188,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404818840603448e-05,
+      "loss": 4.2182,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403980245852396e-05,
+      "loss": 4.2251,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031416511013436e-05,
+      "loss": 4.231,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402304694230665e-05,
+      "loss": 4.2242,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401466099479613e-05,
+      "loss": 4.2252,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400627504728561e-05,
+      "loss": 4.2195,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3997889099775085e-05,
+      "loss": 4.2151,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3989503152264565e-05,
+      "loss": 4.2113,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3981117204754045e-05,
+      "loss": 4.2221,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3972731257243525e-05,
+      "loss": 4.2141,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3964345309733005e-05,
+      "loss": 4.2227,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3955959362222485e-05,
+      "loss": 4.2151,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3947573414711965e-05,
+      "loss": 4.2233,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393920384600518e-05,
+      "loss": 4.2126,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393081789849466e-05,
+      "loss": 4.2166,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392243195098414e-05,
+      "loss": 4.212,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391404600347362e-05,
+      "loss": 4.2251,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39056600559631e-05,
+      "loss": 4.2145,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389727410845258e-05,
+      "loss": 4.2064,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388888816094206e-05,
+      "loss": 4.1986,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388050221343154e-05,
+      "loss": 4.2164,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387213264472475e-05,
+      "loss": 4.2236,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386374669721423e-05,
+      "loss": 4.2204,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385536074970371e-05,
+      "loss": 4.2196,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384697480219319e-05,
+      "loss": 4.2182,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383858885468267e-05,
+      "loss": 4.2152,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830219285975885e-05,
+      "loss": 4.2193,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821833338465365e-05,
+      "loss": 4.2063,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813447390954845e-05,
+      "loss": 4.2168,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3805061443444325e-05,
+      "loss": 4.2115,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796691874737534e-05,
+      "loss": 4.2049,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788305927227014e-05,
+      "loss": 4.2189,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377993635852022e-05,
+      "loss": 4.2136,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.37715504110097e-05,
+      "loss": 4.2093,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376316446349918e-05,
+      "loss": 4.2096,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375477851598866e-05,
+      "loss": 4.2111,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.212958812713623,
+      "eval_runtime": 302.3652,
+      "eval_samples_per_second": 1262.02,
+      "eval_steps_per_second": 39.439,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374639256847814e-05,
+      "loss": 4.2084,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373800662096762e-05,
+      "loss": 4.2086,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37296206734571e-05,
+      "loss": 4.2155,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372123472594658e-05,
+      "loss": 4.2045,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37128651572398e-05,
+      "loss": 4.2173,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370447920972927e-05,
+      "loss": 4.2079,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369609326221875e-05,
+      "loss": 4.1978,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368770731470823e-05,
+      "loss": 4.195,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367933774600145e-05,
+      "loss": 4.2125,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670968177294657e-05,
+      "loss": 4.2132,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662582229784137e-05,
+      "loss": 4.2051,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654196282273616e-05,
+      "loss": 4.2136,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645826713566826e-05,
+      "loss": 4.2063,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363744076605631e-05,
+      "loss": 4.206,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362905481854579e-05,
+      "loss": 4.2012,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362066887103527e-05,
+      "loss": 4.1943,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612282923524745e-05,
+      "loss": 4.2026,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3603896976014225e-05,
+      "loss": 4.1959,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595511028503705e-05,
+      "loss": 4.2099,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3587125080993185e-05,
+      "loss": 4.2108,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3578739133482665e-05,
+      "loss": 4.2145,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3570369564775874e-05,
+      "loss": 4.2018,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3561983617265354e-05,
+      "loss": 4.1985,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3553597669754834e-05,
+      "loss": 4.2037,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545211722244314e-05,
+      "loss": 4.194,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353684215353753e-05,
+      "loss": 4.198,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352845620602701e-05,
+      "loss": 4.1832,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352007025851649e-05,
+      "loss": 4.2036,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351168431100597e-05,
+      "loss": 4.1929,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350329836349545e-05,
+      "loss": 4.19,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349491241598493e-05,
+      "loss": 4.1983,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348652646847441e-05,
+      "loss": 4.2015,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347814052096389e-05,
+      "loss": 4.2054,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34697709522571e-05,
+      "loss": 4.2009,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346138500474658e-05,
+      "loss": 4.1943,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345299905723606e-05,
+      "loss": 4.2038,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344461310972554e-05,
+      "loss": 4.1996,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343622716221502e-05,
+      "loss": 4.1825,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34278412147045e-05,
+      "loss": 4.1835,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419471645997714e-05,
+      "loss": 4.1837,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411085698487194e-05,
+      "loss": 4.1891,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402699750976674e-05,
+      "loss": 4.1989,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394313803466154e-05,
+      "loss": 4.1898,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385927855955634e-05,
+      "loss": 4.1855,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337754190844511e-05,
+      "loss": 4.1988,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336915596093459e-05,
+      "loss": 4.1894,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336077001342407e-05,
+      "loss": 4.2044,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352400444717276e-05,
+      "loss": 4.1841,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3344014497206756e-05,
+      "loss": 4.1645,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3335628549696236e-05,
+      "loss": 4.2034,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3327242602185716e-05,
+      "loss": 4.1954,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331887303347893e-05,
+      "loss": 4.1949,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331048708596841e-05,
+      "loss": 4.1895,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330210113845789e-05,
+      "loss": 4.1836,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329373156975111e-05,
+      "loss": 4.1834,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328534562224058e-05,
+      "loss": 4.1862,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327695967473006e-05,
+      "loss": 4.1797,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326857372721954e-05,
+      "loss": 4.1831,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326018777970902e-05,
+      "loss": 4.1889,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32518018321985e-05,
+      "loss": 4.1852,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324341588468798e-05,
+      "loss": 4.1754,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323504631598119e-05,
+      "loss": 4.1944,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322666036847067e-05,
+      "loss": 4.1757,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321827442096015e-05,
+      "loss": 4.1764,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320988847344964e-05,
+      "loss": 4.1904,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320150252593912e-05,
+      "loss": 4.1907,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31931165784286e-05,
+      "loss": 4.1722,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184730630918077e-05,
+      "loss": 4.1812,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176344683407557e-05,
+      "loss": 4.1628,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167975114700766e-05,
+      "loss": 4.1867,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159589167190246e-05,
+      "loss": 4.1859,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151203219679726e-05,
+      "loss": 4.1848,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3142817272169205e-05,
+      "loss": 4.1807,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3134447703462415e-05,
+      "loss": 4.1927,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3126061755951895e-05,
+      "loss": 4.1966,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3117692187245104e-05,
+      "loss": 4.1704,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310930623973459e-05,
+      "loss": 4.1865,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310092029222407e-05,
+      "loss": 4.1869,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309253434471355e-05,
+      "loss": 4.1631,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308414839720303e-05,
+      "loss": 4.1899,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307576244969251e-05,
+      "loss": 4.1741,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306737650218199e-05,
+      "loss": 4.1892,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305899055467147e-05,
+      "loss": 4.1867,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305060460716094e-05,
+      "loss": 4.1708,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304225141725789e-05,
+      "loss": 4.1708,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303386546974737e-05,
+      "loss": 4.1753,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302547952223685e-05,
+      "loss": 4.1921,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301709357472633e-05,
+      "loss": 4.1764,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300870762721581e-05,
+      "loss": 4.177,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300032167970529e-05,
+      "loss": 4.1728,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991952110998504e-05,
+      "loss": 4.1782,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983566163487984e-05,
+      "loss": 4.1724,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975180215977464e-05,
+      "loss": 4.1566,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966794268466944e-05,
+      "loss": 4.1744,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295840832095642e-05,
+      "loss": 4.1817,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29500223734459e-05,
+      "loss": 4.18,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294165280473911e-05,
+      "loss": 4.1649,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933266857228586e-05,
+      "loss": 4.1661,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29248972885218e-05,
+      "loss": 4.1707,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291651134101128e-05,
+      "loss": 4.1754,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290812539350076e-05,
+      "loss": 4.1799,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289973944599024e-05,
+      "loss": 4.1661,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289135349847972e-05,
+      "loss": 4.1718,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28829675509692e-05,
+      "loss": 4.1804,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287458160345868e-05,
+      "loss": 4.1591,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286619565594816e-05,
+      "loss": 4.171,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285780970843764e-05,
+      "loss": 4.162,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284942376092712e-05,
+      "loss": 4.1745,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.28410378134166e-05,
+      "loss": 4.1607,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283265186590608e-05,
+      "loss": 4.1744,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282428229719929e-05,
+      "loss": 4.1654,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281589634968877e-05,
+      "loss": 4.1648,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280751040217825e-05,
+      "loss": 4.1645,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279912445466773e-05,
+      "loss": 4.1682,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790754885960946e-05,
+      "loss": 4.1671,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782368938450426e-05,
+      "loss": 4.183,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2773982990939906e-05,
+      "loss": 4.1734,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765597043429386e-05,
+      "loss": 4.1745,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757211095918866e-05,
+      "loss": 4.1625,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748825148408346e-05,
+      "loss": 4.1695,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2740439200897826e-05,
+      "loss": 4.1604,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27320532533873e-05,
+      "loss": 4.1658,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2723683684680515e-05,
+      "loss": 4.1657,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2715297737169995e-05,
+      "loss": 4.1697,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2706928168463204e-05,
+      "loss": 4.1674,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2698542220952684e-05,
+      "loss": 4.1664,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2690156273442164e-05,
+      "loss": 4.163,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2681770325931644e-05,
+      "loss": 4.1708,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267338437842113e-05,
+      "loss": 4.1621,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2664998430910604e-05,
+      "loss": 4.1707,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2656612483400084e-05,
+      "loss": 4.1644,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26482429146933e-05,
+      "loss": 4.1597,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263985696718277e-05,
+      "loss": 4.145,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263147101967225e-05,
+      "loss": 4.1685,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262308507216173e-05,
+      "loss": 4.1712,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261469912465121e-05,
+      "loss": 4.1718,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260631317714069e-05,
+      "loss": 4.1669,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259792722963017e-05,
+      "loss": 4.168,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258954128211965e-05,
+      "loss": 4.1625,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258117171341287e-05,
+      "loss": 4.174,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257278576590235e-05,
+      "loss": 4.1558,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256439981839183e-05,
+      "loss": 4.1639,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255601387088131e-05,
+      "loss": 4.1611,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254764430217452e-05,
+      "loss": 4.1585,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539258354664e-05,
+      "loss": 4.1727,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253087240715348e-05,
+      "loss": 4.1588,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522502838446686e-05,
+      "loss": 4.1658,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2514116890936166e-05,
+      "loss": 4.1571,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2505730943425646e-05,
+      "loss": 4.1619,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.172065734863281,
+      "eval_runtime": 305.6246,
+      "eval_samples_per_second": 1248.561,
+      "eval_steps_per_second": 39.018,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2497344995915126e-05,
+      "loss": 4.1668,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488959048404606e-05,
+      "loss": 4.1602,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480573100894086e-05,
+      "loss": 4.1649,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472187153383566e-05,
+      "loss": 4.1548,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2463801205873046e-05,
+      "loss": 4.1697,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245541525836253e-05,
+      "loss": 4.1584,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244702931085201e-05,
+      "loss": 4.1512,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243864336334149e-05,
+      "loss": 4.1458,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430257415830966e-05,
+      "loss": 4.1639,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2421871468320446e-05,
+      "loss": 4.1641,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413485520809926e-05,
+      "loss": 4.1562,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405099573299406e-05,
+      "loss": 4.165,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2396730004592615e-05,
+      "loss": 4.1575,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2388344057082095e-05,
+      "loss": 4.1581,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379958109571575e-05,
+      "loss": 4.1499,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2371572162061055e-05,
+      "loss": 4.147,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2363186214550535e-05,
+      "loss": 4.1549,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354800267040015e-05,
+      "loss": 4.152,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23464143195295e-05,
+      "loss": 4.1585,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233802837201898e-05,
+      "loss": 4.1629,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232965880331219e-05,
+      "loss": 4.1682,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232127285580167e-05,
+      "loss": 4.1562,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231288690829115e-05,
+      "loss": 4.1522,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230450096078063e-05,
+      "loss": 4.154,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229613139207384e-05,
+      "loss": 4.1511,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228774544456332e-05,
+      "loss": 4.1512,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22793594970528e-05,
+      "loss": 4.1336,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227097354954228e-05,
+      "loss": 4.1587,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226260398083549e-05,
+      "loss": 4.149,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225421803332497e-05,
+      "loss": 4.1414,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245832085814455e-05,
+      "loss": 4.1507,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237446138303935e-05,
+      "loss": 4.155,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229060190793415e-05,
+      "loss": 4.1521,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220674243282895e-05,
+      "loss": 4.1563,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212288295772375e-05,
+      "loss": 4.1463,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220390234826185e-05,
+      "loss": 4.1584,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219551640075133e-05,
+      "loss": 4.1537,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218713045324081e-05,
+      "loss": 4.1359,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217874450573029e-05,
+      "loss": 4.1388,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217035855821977e-05,
+      "loss": 4.1374,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216198898951298e-05,
+      "loss": 4.1441,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215360304200246e-05,
+      "loss": 4.1544,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214523347329567e-05,
+      "loss": 4.1425,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213684752578515e-05,
+      "loss": 4.1411,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212846157827463e-05,
+      "loss": 4.1485,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212007563076411e-05,
+      "loss": 4.1432,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211168968325359e-05,
+      "loss": 4.1583,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21033201145468e-05,
+      "loss": 4.1414,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209493416703628e-05,
+      "loss": 4.1171,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208654821952576e-05,
+      "loss": 4.1566,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207816227201524e-05,
+      "loss": 4.1513,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206977632450472e-05,
+      "loss": 4.1499,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20613903769942e-05,
+      "loss": 4.1423,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205302080828741e-05,
+      "loss": 4.1392,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204463486077689e-05,
+      "loss": 4.1372,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036265292070106e-05,
+      "loss": 4.1428,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027879344559586e-05,
+      "loss": 4.1362,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019493397049066e-05,
+      "loss": 4.1353,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011107449538546e-05,
+      "loss": 4.1465,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2002721502028026e-05,
+      "loss": 4.1427,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994335554517506e-05,
+      "loss": 4.1305,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1985965985810715e-05,
+      "loss": 4.1503,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1977580038300195e-05,
+      "loss": 4.1285,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969194090789675e-05,
+      "loss": 4.1336,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1960808143279155e-05,
+      "loss": 4.1471,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1952422195768635e-05,
+      "loss": 4.1468,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1944036248258115e-05,
+      "loss": 4.1278,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1935650300747595e-05,
+      "loss": 4.137,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927264353237075e-05,
+      "loss": 4.118,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918878405726555e-05,
+      "loss": 4.1428,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191049245821604e-05,
+      "loss": 4.1422,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190212288950925e-05,
+      "loss": 4.1423,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189373694199873e-05,
+      "loss": 4.1374,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188535099448821e-05,
+      "loss": 4.1463,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876965046977684e-05,
+      "loss": 4.1548,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868579099467164e-05,
+      "loss": 4.1298,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860193151956644e-05,
+      "loss": 4.1419,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851807204446124e-05,
+      "loss": 4.1436,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843421256935604e-05,
+      "loss": 4.1243,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183505168822881e-05,
+      "loss": 4.1421,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182668211952203e-05,
+      "loss": 4.1372,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181829617201151e-05,
+      "loss": 4.1416,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180991022450099e-05,
+      "loss": 4.1452,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180152427699047e-05,
+      "loss": 4.1296,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1793154708283684e-05,
+      "loss": 4.1292,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178476876077316e-05,
+      "loss": 4.133,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177638281326264e-05,
+      "loss": 4.1429,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176799686575212e-05,
+      "loss": 4.136,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175962729704533e-05,
+      "loss": 4.1341,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175125772833854e-05,
+      "loss": 4.1329,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174287178082802e-05,
+      "loss": 4.1383,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17344858333175e-05,
+      "loss": 4.1311,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172609988580698e-05,
+      "loss": 4.114,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17177303171002e-05,
+      "loss": 4.1342,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170934436958968e-05,
+      "loss": 4.1359,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170095842207916e-05,
+      "loss": 4.1379,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169257247456863e-05,
+      "loss": 4.1257,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168418652705811e-05,
+      "loss": 4.1232,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167580057954759e-05,
+      "loss": 4.1321,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166741463203707e-05,
+      "loss": 4.1324,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165902868452655e-05,
+      "loss": 4.1418,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165065911581976e-05,
+      "loss": 4.125,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164227316830924e-05,
+      "loss": 4.131,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163388722079872e-05,
+      "loss": 4.1362,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16255012732882e-05,
+      "loss": 4.1178,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161711532577769e-05,
+      "loss": 4.1311,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160872937826717e-05,
+      "loss": 4.1163,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160034343075665e-05,
+      "loss": 4.1366,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159195748324613e-05,
+      "loss": 4.1172,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158357153573561e-05,
+      "loss": 4.1326,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1575185588225087e-05,
+      "loss": 4.1233,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1566799640714567e-05,
+      "loss": 4.1224,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558413693204047e-05,
+      "loss": 4.1257,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1550044124497256e-05,
+      "loss": 4.1285,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1541674555790465e-05,
+      "loss": 4.127,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533288608279945e-05,
+      "loss": 4.1412,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524902660769425e-05,
+      "loss": 4.1268,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516516713258905e-05,
+      "loss": 4.141,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508130765748385e-05,
+      "loss": 4.1242,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1499744818237864e-05,
+      "loss": 4.1246,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149137524953108e-05,
+      "loss": 4.1255,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148298930202056e-05,
+      "loss": 4.1236,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147460335451004e-05,
+      "loss": 4.1234,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146621740699952e-05,
+      "loss": 4.1302,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145784783829273e-05,
+      "loss": 4.1263,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144946189078221e-05,
+      "loss": 4.1289,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144109232207542e-05,
+      "loss": 4.1237,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14327063745649e-05,
+      "loss": 4.1283,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142432042705438e-05,
+      "loss": 4.1239,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141593447954386e-05,
+      "loss": 4.1251,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140754853203334e-05,
+      "loss": 4.1221,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139916258452282e-05,
+      "loss": 4.1237,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13907766370123e-05,
+      "loss": 4.1035,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138239068950178e-05,
+      "loss": 4.13,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1374021120794994e-05,
+      "loss": 4.1301,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136563517328447e-05,
+      "loss": 4.1327,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135724922577395e-05,
+      "loss": 4.124,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134886327826343e-05,
+      "loss": 4.1309,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134047733075291e-05,
+      "loss": 4.1251,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332107762046116e-05,
+      "loss": 4.1344,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323721814535596e-05,
+      "loss": 4.115,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315335867025076e-05,
+      "loss": 4.1217,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306949919514556e-05,
+      "loss": 4.1244,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129856397200404e-05,
+      "loss": 4.1217,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129017802449352e-05,
+      "loss": 4.1271,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1281792076983e-05,
+      "loss": 4.1235,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127340612947248e-05,
+      "loss": 4.1253,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126505293956942e-05,
+      "loss": 4.119,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12566669920589e-05,
+      "loss": 4.1268,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.141274452209473,
+      "eval_runtime": 290.3953,
+      "eval_samples_per_second": 1314.04,
+      "eval_steps_per_second": 41.065,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124828104454838e-05,
+      "loss": 4.1175,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123989509703786e-05,
+      "loss": 4.1262,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123150914952734e-05,
+      "loss": 4.1216,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122313958082055e-05,
+      "loss": 4.1176,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121475363331003e-05,
+      "loss": 4.1287,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120636768579951e-05,
+      "loss": 4.1238,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1197981738288996e-05,
+      "loss": 4.1136,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1189595790778476e-05,
+      "loss": 4.1114,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1181209843267956e-05,
+      "loss": 4.1193,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1172840274561165e-05,
+      "loss": 4.1292,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1164454327050645e-05,
+      "loss": 4.1157,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156068379540125e-05,
+      "loss": 4.1255,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1147698810833334e-05,
+      "loss": 4.1265,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1139312863322814e-05,
+      "loss": 4.1158,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1130926915812294e-05,
+      "loss": 4.1136,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122540968301774e-05,
+      "loss": 4.1102,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114155020791254e-05,
+      "loss": 4.1161,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105769073280734e-05,
+      "loss": 4.1121,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1097383125770214e-05,
+      "loss": 4.1176,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1088997178259694e-05,
+      "loss": 4.1293,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108061123074918e-05,
+      "loss": 4.1295,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072225283238654e-05,
+      "loss": 4.1197,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063839335728134e-05,
+      "loss": 4.1141,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105546976702135e-05,
+      "loss": 4.1159,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104708381951083e-05,
+      "loss": 4.1144,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10386978720003e-05,
+      "loss": 4.1125,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103031192448978e-05,
+      "loss": 4.0996,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021942355783e-05,
+      "loss": 4.1206,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101355640827247e-05,
+      "loss": 4.1131,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100517046076195e-05,
+      "loss": 4.1052,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099678451325143e-05,
+      "loss": 4.1073,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098839856574092e-05,
+      "loss": 4.1215,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098002899703413e-05,
+      "loss": 4.1186,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097164304952361e-05,
+      "loss": 4.1176,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096325710201309e-05,
+      "loss": 4.1113,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0954887533306303e-05,
+      "loss": 4.1187,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094650158579578e-05,
+      "loss": 4.1194,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0938115638285257e-05,
+      "loss": 4.0994,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0929729690774737e-05,
+      "loss": 4.1011,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0921360122067946e-05,
+      "loss": 4.1003,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0912974174557426e-05,
+      "loss": 4.1074,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0904588227046906e-05,
+      "loss": 4.1139,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0896202279536386e-05,
+      "loss": 4.1116,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088781633202587e-05,
+      "loss": 4.1048,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087943038451535e-05,
+      "loss": 4.1109,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087104443700483e-05,
+      "loss": 4.1064,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086265848949431e-05,
+      "loss": 4.1182,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085427254198379e-05,
+      "loss": 4.1096,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084588659447327e-05,
+      "loss": 4.0818,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083750064696275e-05,
+      "loss": 4.1194,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082911469945223e-05,
+      "loss": 4.1145,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082074513074544e-05,
+      "loss": 4.1191,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081235918323492e-05,
+      "loss": 4.1044,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08039732357244e-05,
+      "loss": 4.1006,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079558728821388e-05,
+      "loss": 4.105,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787234098310826e-05,
+      "loss": 4.107,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0778848150800306e-05,
+      "loss": 4.106,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077049496089725e-05,
+      "loss": 4.096,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762109013386724e-05,
+      "loss": 4.1103,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0753723065876204e-05,
+      "loss": 4.1078,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0745337118365684e-05,
+      "loss": 4.0914,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0736951170855164e-05,
+      "loss": 4.1137,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0728565223344644e-05,
+      "loss": 4.0944,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0720179275834124e-05,
+      "loss": 4.0964,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0711793328323604e-05,
+      "loss": 4.1099,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0703407380813084e-05,
+      "loss": 4.1113,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0695021433302564e-05,
+      "loss": 4.0953,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0686635485792044e-05,
+      "loss": 4.1009,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678249538281524e-05,
+      "loss": 4.0836,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066987996957474e-05,
+      "loss": 4.1063,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066149402206422e-05,
+      "loss": 4.1084,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06531080745537e-05,
+      "loss": 4.1071,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064472212704318e-05,
+      "loss": 4.1006,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063633617953266e-05,
+      "loss": 4.108,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062795023202214e-05,
+      "loss": 4.1223,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061956428451161e-05,
+      "loss": 4.0895,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.061117833700109e-05,
+      "loss": 4.1075,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060279238949057e-05,
+      "loss": 4.1075,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059442282078378e-05,
+      "loss": 4.0913,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058603687327326e-05,
+      "loss": 4.1065,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057765092576274e-05,
+      "loss": 4.1024,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056926497825223e-05,
+      "loss": 4.1048,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056087903074171e-05,
+      "loss": 4.1134,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055249308323119e-05,
+      "loss": 4.0898,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05441235145244e-05,
+      "loss": 4.0948,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053573756701388e-05,
+      "loss": 4.0982,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052735161950336e-05,
+      "loss": 4.1082,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051896567199284e-05,
+      "loss": 4.0991,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051057972448232e-05,
+      "loss": 4.1043,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.05021937769718e-05,
+      "loss": 4.1013,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049380782946128e-05,
+      "loss": 4.0971,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0485438260754486e-05,
+      "loss": 4.0983,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0477052313243966e-05,
+      "loss": 4.0803,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0468666365733446e-05,
+      "loss": 4.0979,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460280418222926e-05,
+      "loss": 4.1009,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045191084951614e-05,
+      "loss": 4.1046,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044352490200562e-05,
+      "loss": 4.0917,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04351389544951e-05,
+      "loss": 4.0868,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042675300698458e-05,
+      "loss": 4.0963,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041836705947406e-05,
+      "loss": 4.0986,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040999749076727e-05,
+      "loss": 4.1024,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040161154325675e-05,
+      "loss": 4.0965,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039322559574623e-05,
+      "loss": 4.0988,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038483964823571e-05,
+      "loss": 4.0965,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.037647007952892e-05,
+      "loss": 4.0882,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03680841320184e-05,
+      "loss": 4.0927,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035969818450788e-05,
+      "loss": 4.0832,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351312236997366e-05,
+      "loss": 4.1042,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342926289486846e-05,
+      "loss": 4.0847,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0334556720780055e-05,
+      "loss": 4.0975,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326170773269535e-05,
+      "loss": 4.09,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0317784825759015e-05,
+      "loss": 4.0953,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0309398878248495e-05,
+      "loss": 4.0912,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030101293073797e-05,
+      "loss": 4.0931,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029262698322745e-05,
+      "loss": 4.0937,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028424103571693e-05,
+      "loss": 4.1087,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0275871467010144e-05,
+      "loss": 4.0906,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026748551949962e-05,
+      "loss": 4.106,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259099571989104e-05,
+      "loss": 4.0893,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0250713624478584e-05,
+      "loss": 4.0911,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242327676968064e-05,
+      "loss": 4.0912,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023395810826127e-05,
+      "loss": 4.0929,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022557216075075e-05,
+      "loss": 4.0865,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021718621324023e-05,
+      "loss": 4.0965,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020880026572971e-05,
+      "loss": 4.0891,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020043069702292e-05,
+      "loss": 4.0947,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019206112831614e-05,
+      "loss": 4.0941,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018367518080562e-05,
+      "loss": 4.0908,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017528923329509e-05,
+      "loss": 4.094,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016690328578457e-05,
+      "loss": 4.0972,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015851733827406e-05,
+      "loss": 4.0837,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015013139076354e-05,
+      "loss": 4.0914,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014176182205675e-05,
+      "loss": 4.0692,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013337587454623e-05,
+      "loss": 4.0971,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012498992703571e-05,
+      "loss": 4.0978,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0116603979525187e-05,
+      "loss": 4.1005,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108234410818396e-05,
+      "loss": 4.0914,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0099848463307876e-05,
+      "loss": 4.097,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0091462515797356e-05,
+      "loss": 4.0939,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0083076568286836e-05,
+      "loss": 4.0995,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0074706999580045e-05,
+      "loss": 4.0828,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0066321052069525e-05,
+      "loss": 4.0918,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0057935104559005e-05,
+      "loss": 4.0904,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004954915704849e-05,
+      "loss": 4.0902,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004116320953797e-05,
+      "loss": 4.0928,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003277726202745e-05,
+      "loss": 4.086,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002439131451693e-05,
+      "loss": 4.0977,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001600536700641e-05,
+      "loss": 4.0839,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000765217710335e-05,
+      "loss": 4.0969,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.116448879241943,
+      "eval_runtime": 288.6273,
+      "eval_samples_per_second": 1322.089,
+      "eval_steps_per_second": 41.316,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999926622959283e-05,
+      "loss": 4.0873,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999088028208231e-05,
+      "loss": 4.0897,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998249433457179e-05,
+      "loss": 4.092,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997410838706127e-05,
+      "loss": 4.0885,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996572243955075e-05,
+      "loss": 4.0968,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995733649204023e-05,
+      "loss": 4.0915,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994895054452971e-05,
+      "loss": 4.0809,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994056459701919e-05,
+      "loss": 4.0788,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9932178649508676e-05,
+      "loss": 4.0869,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9923792701998156e-05,
+      "loss": 4.0972,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9915406754487636e-05,
+      "loss": 4.0886,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990702080697711e-05,
+      "loss": 4.0912,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9898651238270325e-05,
+      "loss": 4.0934,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890265290759805e-05,
+      "loss": 4.0836,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988187934324928e-05,
+      "loss": 4.0796,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987349339573876e-05,
+      "loss": 4.0765,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986510744822824e-05,
+      "loss": 4.0867,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985672150071772e-05,
+      "loss": 4.0812,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98483355532072e-05,
+      "loss": 4.0867,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983994960569668e-05,
+      "loss": 4.0959,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983156365818616e-05,
+      "loss": 4.0968,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823177710675644e-05,
+      "loss": 4.0841,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9814791763165124e-05,
+      "loss": 4.0856,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9806405815654604e-05,
+      "loss": 4.085,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979803624694781e-05,
+      "loss": 4.0822,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978966667824102e-05,
+      "loss": 4.0822,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97812807307305e-05,
+      "loss": 4.0691,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977289478321998e-05,
+      "loss": 4.09,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976450883570946e-05,
+      "loss": 4.0825,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975612288819894e-05,
+      "loss": 4.0714,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974775331949215e-05,
+      "loss": 4.0773,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973936737198163e-05,
+      "loss": 4.0896,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973098142447111e-05,
+      "loss": 4.0904,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97225954769606e-05,
+      "loss": 4.087,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971420952945008e-05,
+      "loss": 4.0809,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970583996074329e-05,
+      "loss": 4.0799,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969745401323277e-05,
+      "loss": 4.0937,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968906806572225e-05,
+      "loss": 4.0703,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968068211821173e-05,
+      "loss": 4.069,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672312549504936e-05,
+      "loss": 4.0733,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9663926601994416e-05,
+      "loss": 4.0732,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9655540654483896e-05,
+      "loss": 4.0826,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9647154706973376e-05,
+      "loss": 4.0737,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9638768759462856e-05,
+      "loss": 4.0815,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9630382811952336e-05,
+      "loss": 4.0765,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9621996864441816e-05,
+      "loss": 4.0739,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613610916931296e-05,
+      "loss": 4.0941,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960524134822451e-05,
+      "loss": 4.0753,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959685540071399e-05,
+      "loss": 4.0566,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9588469453203465e-05,
+      "loss": 4.0866,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958009988449668e-05,
+      "loss": 4.0802,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957171393698616e-05,
+      "loss": 4.0902,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956332798947564e-05,
+      "loss": 4.0766,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9554942041965114e-05,
+      "loss": 4.0707,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9546556094454594e-05,
+      "loss": 4.0746,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953818652574781e-05,
+      "loss": 4.0718,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952980057823729e-05,
+      "loss": 4.0789,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952141463072677e-05,
+      "loss": 4.0633,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951302868321625e-05,
+      "loss": 4.0837,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950464273570573e-05,
+      "loss": 4.0774,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949625678819521e-05,
+      "loss": 4.064,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948788721948842e-05,
+      "loss": 4.0821,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94795012719779e-05,
+      "loss": 4.0628,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947111532446738e-05,
+      "loss": 4.0655,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946272937695686e-05,
+      "loss": 4.079,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945434342944634e-05,
+      "loss": 4.0842,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944595748193582e-05,
+      "loss": 4.0592,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94375715344253e-05,
+      "loss": 4.0773,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942918558691478e-05,
+      "loss": 4.0542,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942081601820799e-05,
+      "loss": 4.0766,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941243007069747e-05,
+      "loss": 4.0806,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9404044123186954e-05,
+      "loss": 4.0725,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939567455448016e-05,
+      "loss": 4.0762,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938728860696964e-05,
+      "loss": 4.0775,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937890265945912e-05,
+      "loss": 4.094,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.93705167119486e-05,
+      "loss": 4.0597,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936213076443808e-05,
+      "loss": 4.0793,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935374481692756e-05,
+      "loss": 4.075,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934535886941704e-05,
+      "loss": 4.0679,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933698930071025e-05,
+      "loss": 4.0767,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932860335319973e-05,
+      "loss": 4.0711,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932021740568921e-05,
+      "loss": 4.0775,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931183145817869e-05,
+      "loss": 4.0827,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930344551066817e-05,
+      "loss": 4.0603,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929507594196139e-05,
+      "loss": 4.0684,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928668999445087e-05,
+      "loss": 4.0696,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927830404694035e-05,
+      "loss": 4.0832,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926991809942983e-05,
+      "loss": 4.0653,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9261548530723036e-05,
+      "loss": 4.0753,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9253162583212516e-05,
+      "loss": 4.0749,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9244776635701996e-05,
+      "loss": 4.0705,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236390688191476e-05,
+      "loss": 4.0683,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922800474068095e-05,
+      "loss": 4.0516,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921961879317043e-05,
+      "loss": 4.0676,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211249224463645e-05,
+      "loss": 4.0719,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9202863276953125e-05,
+      "loss": 4.0791,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9194477329442605e-05,
+      "loss": 4.0627,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9186091381932085e-05,
+      "loss": 4.0585,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9177705434421565e-05,
+      "loss": 4.0667,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9169319486911045e-05,
+      "loss": 4.0748,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9160933539400525e-05,
+      "loss": 4.0736,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9152547591890005e-05,
+      "loss": 4.0686,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144178023183214e-05,
+      "loss": 4.074,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9135792075672694e-05,
+      "loss": 4.066,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127406128162174e-05,
+      "loss": 4.0656,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9119020180651654e-05,
+      "loss": 4.0605,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911065061194486e-05,
+      "loss": 4.0604,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910228104323808e-05,
+      "loss": 4.0759,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909389509572756e-05,
+      "loss": 4.0564,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908550914821704e-05,
+      "loss": 4.0668,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907712320070652e-05,
+      "loss": 4.058,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068737253196e-05,
+      "loss": 4.0704,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906035130568548e-05,
+      "loss": 4.0638,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905196535817496e-05,
+      "loss": 4.0653,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904359578946817e-05,
+      "loss": 4.0665,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903520984195765e-05,
+      "loss": 4.0783,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902682389444713e-05,
+      "loss": 4.0668,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901843794693661e-05,
+      "loss": 4.0778,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901005199942609e-05,
+      "loss": 4.0626,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900166605191557e-05,
+      "loss": 4.0655,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899328010440505e-05,
+      "loss": 4.0654,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898489415689453e-05,
+      "loss": 4.063,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897650820938401e-05,
+      "loss": 4.0596,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896812226187349e-05,
+      "loss": 4.069,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895973631436297e-05,
+      "loss": 4.0642,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895135036685245e-05,
+      "loss": 4.0676,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894298079814566e-05,
+      "loss": 4.0641,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8934594850635136e-05,
+      "loss": 4.0676,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926208903124616e-05,
+      "loss": 4.0641,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8917822955614096e-05,
+      "loss": 4.0745,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890945338690731e-05,
+      "loss": 4.0563,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8901067439396785e-05,
+      "loss": 4.0654,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892681491886265e-05,
+      "loss": 4.0395,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884295544375745e-05,
+      "loss": 4.0729,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887590959686523e-05,
+      "loss": 4.0705,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886754002815844e-05,
+      "loss": 4.0764,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885915408064792e-05,
+      "loss": 4.0625,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88507681331374e-05,
+      "loss": 4.0715,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884238218562688e-05,
+      "loss": 4.0665,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883399623811636e-05,
+      "loss": 4.0762,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882561029060584e-05,
+      "loss": 4.056,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881722434309532e-05,
+      "loss": 4.0631,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88088383955848e-05,
+      "loss": 4.0611,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880046882687801e-05,
+      "loss": 4.0696,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879208287936749e-05,
+      "loss": 4.0609,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878369693185697e-05,
+      "loss": 4.0601,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877531098434645e-05,
+      "loss": 4.0671,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8766941415639666e-05,
+      "loss": 4.0621,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8758555468129145e-05,
+      "loss": 4.0684,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.097685813903809,
+      "eval_runtime": 296.8978,
+      "eval_samples_per_second": 1285.261,
+      "eval_steps_per_second": 40.165,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8750169520618625e-05,
+      "loss": 4.0535,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8741783573108105e-05,
+      "loss": 4.0682,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8733414004401314e-05,
+      "loss": 4.0601,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8725044435694524e-05,
+      "loss": 4.0633,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871667486698773e-05,
+      "loss": 4.0734,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870828891947721e-05,
+      "loss": 4.0632,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869990297196669e-05,
+      "loss": 4.0543,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869151702445617e-05,
+      "loss": 4.0526,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868313107694565e-05,
+      "loss": 4.0603,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.867474512943514e-05,
+      "loss": 4.0693,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866635918192462e-05,
+      "loss": 4.0599,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86579732344141e-05,
+      "loss": 4.0647,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864960366570731e-05,
+      "loss": 4.0677,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.864121771819679e-05,
+      "loss": 4.0573,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863283177068627e-05,
+      "loss": 4.0544,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862444582317575e-05,
+      "loss": 4.0503,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861605987566523e-05,
+      "loss": 4.0637,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860767392815471e-05,
+      "loss": 4.0552,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859928798064419e-05,
+      "loss": 4.0546,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859090203313367e-05,
+      "loss": 4.0673,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858253246442688e-05,
+      "loss": 4.0714,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857414651691636e-05,
+      "loss": 4.0611,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856576056940584e-05,
+      "loss": 4.0593,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855739100069905e-05,
+      "loss": 4.0582,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854900505318853e-05,
+      "loss": 4.0577,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854061910567801e-05,
+      "loss": 4.0562,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853223315816749e-05,
+      "loss": 4.0397,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852384721065697e-05,
+      "loss": 4.0629,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8515461263146446e-05,
+      "loss": 4.0622,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850709169443966e-05,
+      "loss": 4.0446,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849870574692914e-05,
+      "loss": 4.0538,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8490319799418615e-05,
+      "loss": 4.0585,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8481933851908095e-05,
+      "loss": 4.0635,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8473547904397575e-05,
+      "loss": 4.0631,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846517833569079e-05,
+      "loss": 4.0597,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845679238818027e-05,
+      "loss": 4.0513,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844840644066975e-05,
+      "loss": 4.0716,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844002049315923e-05,
+      "loss": 4.0434,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843163454564871e-05,
+      "loss": 4.0435,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842326497694192e-05,
+      "loss": 4.047,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84148790294314e-05,
+      "loss": 4.0503,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840649308192088e-05,
+      "loss": 4.0559,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839810713441036e-05,
+      "loss": 4.0491,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838972118689984e-05,
+      "loss": 4.0541,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838133523938932e-05,
+      "loss": 4.0508,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83729492918788e-05,
+      "loss": 4.0489,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836456334436828e-05,
+      "loss": 4.0624,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8356193775661495e-05,
+      "loss": 4.0563,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8347807828150975e-05,
+      "loss": 4.0316,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8339421880640455e-05,
+      "loss": 4.0582,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8331052311933664e-05,
+      "loss": 4.0582,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8322666364423144e-05,
+      "loss": 4.0653,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8314280416912624e-05,
+      "loss": 4.0501,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8305894469402104e-05,
+      "loss": 4.0472,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8297508521891584e-05,
+      "loss": 4.0452,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828913895318479e-05,
+      "loss": 4.0487,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828075300567427e-05,
+      "loss": 4.0522,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827238343696748e-05,
+      "loss": 4.0382,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826399748945696e-05,
+      "loss": 4.0556,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825561154194645e-05,
+      "loss": 4.0533,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824722559443593e-05,
+      "loss": 4.0425,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823883964692541e-05,
+      "loss": 4.0537,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823045369941489e-05,
+      "loss": 4.0409,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822206775190437e-05,
+      "loss": 4.0422,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821368180439385e-05,
+      "loss": 4.0481,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820531223568706e-05,
+      "loss": 4.058,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819692628817654e-05,
+      "loss": 4.0394,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818854034066602e-05,
+      "loss": 4.053,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81801543931555e-05,
+      "loss": 4.0251,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817176844564498e-05,
+      "loss": 4.0524,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8163398876938187e-05,
+      "loss": 4.0557,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8155012929427666e-05,
+      "loss": 4.0524,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8146626981917146e-05,
+      "loss": 4.0466,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813824103440663e-05,
+      "loss": 4.0541,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8129855086896106e-05,
+      "loss": 4.0683,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8121469139385586e-05,
+      "loss": 4.0388,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8113083191875066e-05,
+      "loss": 4.0553,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810471362316828e-05,
+      "loss": 4.0454,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809634405446149e-05,
+      "loss": 4.0474,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808795810695097e-05,
+      "loss": 4.0508,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807957215944045e-05,
+      "loss": 4.045,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8071186211929924e-05,
+      "loss": 4.0541,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8062800264419404e-05,
+      "loss": 4.0613,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8054414316908884e-05,
+      "loss": 4.0355,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804602836939837e-05,
+      "loss": 4.0459,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803764242188785e-05,
+      "loss": 4.0456,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802927285318106e-05,
+      "loss": 4.0573,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802088690567054e-05,
+      "loss": 4.0411,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801250095816002e-05,
+      "loss": 4.0546,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.80041150106495e-05,
+      "loss": 4.0474,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799572906313898e-05,
+      "loss": 4.0444,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798734311562846e-05,
+      "loss": 4.0436,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797895716811794e-05,
+      "loss": 4.0288,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797057122060742e-05,
+      "loss": 4.0434,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796220165190063e-05,
+      "loss": 4.051,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795381570439011e-05,
+      "loss": 4.0561,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794542975687959e-05,
+      "loss": 4.0358,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793704380936907e-05,
+      "loss": 4.038,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792865786185855e-05,
+      "loss": 4.0438,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7920288293151764e-05,
+      "loss": 4.047,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7911902345641244e-05,
+      "loss": 4.0506,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7903516398130724e-05,
+      "loss": 4.042,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7895130450620204e-05,
+      "loss": 4.0479,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7886744503109684e-05,
+      "loss": 4.0435,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787839131320662e-05,
+      "loss": 4.0442,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78700053656961e-05,
+      "loss": 4.034,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786161941818558e-05,
+      "loss": 4.0377,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785323347067506e-05,
+      "loss": 4.0499,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784484752316454e-05,
+      "loss": 4.0372,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783647795445776e-05,
+      "loss": 4.0425,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782809200694724e-05,
+      "loss": 4.033,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781970605943672e-05,
+      "loss": 4.0431,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78113201119262e-05,
+      "loss": 4.0397,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780293416441568e-05,
+      "loss": 4.0414,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779456459570889e-05,
+      "loss": 4.0457,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778617864819837e-05,
+      "loss": 4.0577,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777779270068785e-05,
+      "loss": 4.0434,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776940675317733e-05,
+      "loss": 4.0484,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776102080566681e-05,
+      "loss": 4.0448,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7752651236960016e-05,
+      "loss": 4.037,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744265289449496e-05,
+      "loss": 4.0475,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7735879341938976e-05,
+      "loss": 4.0352,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7727493394428456e-05,
+      "loss": 4.0376,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771910744691794e-05,
+      "loss": 4.0444,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7710721499407416e-05,
+      "loss": 4.0433,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770235193070063e-05,
+      "loss": 4.0441,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769398236199384e-05,
+      "loss": 4.0384,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768559641448332e-05,
+      "loss": 4.0493,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76772104669728e-05,
+      "loss": 4.0362,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766882451946228e-05,
+      "loss": 4.0514,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766043857195176e-05,
+      "loss": 4.0371,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7652052624441234e-05,
+      "loss": 4.0401,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7643666676930714e-05,
+      "loss": 4.0156,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7635280729420194e-05,
+      "loss": 4.0491,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762689478190968e-05,
+      "loss": 4.0451,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761850883439916e-05,
+      "loss": 4.054,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761012288688864e-05,
+      "loss": 4.04,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760173693937812e-05,
+      "loss": 4.0467,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759336737067133e-05,
+      "loss": 4.041,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758499780196454e-05,
+      "loss": 4.0578,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757661185445402e-05,
+      "loss": 4.0322,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75682259069435e-05,
+      "loss": 4.0375,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755983995943298e-05,
+      "loss": 4.0391,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755145401192246e-05,
+      "loss": 4.0459,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754306806441194e-05,
+      "loss": 4.0388,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753468211690142e-05,
+      "loss": 4.0358,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7526312548194634e-05,
+      "loss": 4.0425,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7517926600684114e-05,
+      "loss": 4.0432,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7509540653173594e-05,
+      "loss": 4.0424,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.081941604614258,
+      "eval_runtime": 296.7323,
+      "eval_samples_per_second": 1285.977,
+      "eval_steps_per_second": 40.188,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7501154705663074e-05,
+      "loss": 4.0299,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7492768758152554e-05,
+      "loss": 4.0455,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7484382810642034e-05,
+      "loss": 4.0388,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7475996863131514e-05,
+      "loss": 4.04,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467610915620994e-05,
+      "loss": 4.0489,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459224968110474e-05,
+      "loss": 4.04,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745083902059995e-05,
+      "loss": 4.0365,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744245307308943e-05,
+      "loss": 4.03,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.743406712557891e-05,
+      "loss": 4.0357,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.742568117806839e-05,
+      "loss": 4.0451,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.741729523055787e-05,
+      "loss": 4.0431,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740890928304735e-05,
+      "loss": 4.0434,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740053971434056e-05,
+      "loss": 4.0475,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739215376683004e-05,
+      "loss": 4.035,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738376781931952e-05,
+      "loss": 4.0335,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7375381871809e-05,
+      "loss": 4.0261,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736699592429848e-05,
+      "loss": 4.0437,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735860997678796e-05,
+      "loss": 4.0331,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735022402927744e-05,
+      "loss": 4.0316,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734185446057065e-05,
+      "loss": 4.0442,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733346851306013e-05,
+      "loss": 4.0502,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732508256554961e-05,
+      "loss": 4.0388,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731669661803909e-05,
+      "loss": 4.0376,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730831067052857e-05,
+      "loss": 4.0372,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729992472301805e-05,
+      "loss": 4.0351,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729153877550753e-05,
+      "loss": 4.0355,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728315282799701e-05,
+      "loss": 4.0175,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727476688048649e-05,
+      "loss": 4.039,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726638093297597e-05,
+      "loss": 4.0396,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725799498546545e-05,
+      "loss": 4.0244,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724962541675866e-05,
+      "loss": 4.0312,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724123946924814e-05,
+      "loss": 4.0376,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723285352173762e-05,
+      "loss": 4.0366,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72244675742271e-05,
+      "loss": 4.0448,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721608162671658e-05,
+      "loss": 4.0353,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720769567920606e-05,
+      "loss": 4.0298,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719930973169554e-05,
+      "loss": 4.0521,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719092378418502e-05,
+      "loss": 4.0158,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71825378366745e-05,
+      "loss": 4.023,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7174184646771445e-05,
+      "loss": 4.0275,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7165798699260925e-05,
+      "loss": 4.0266,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157412751750405e-05,
+      "loss": 4.0331,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7149026804239885e-05,
+      "loss": 4.0316,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140640856729365e-05,
+      "loss": 4.0326,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7132254909218845e-05,
+      "loss": 4.0314,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7123868961708324e-05,
+      "loss": 4.0272,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7115499393001534e-05,
+      "loss": 4.0367,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7107113445491014e-05,
+      "loss": 4.0348,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7098727497980494e-05,
+      "loss": 4.0128,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70903579292737e-05,
+      "loss": 4.0357,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708197198176319e-05,
+      "loss": 4.0357,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707358603425267e-05,
+      "loss": 4.0445,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706520008674215e-05,
+      "loss": 4.0287,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705681413923163e-05,
+      "loss": 4.0279,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704842819172111e-05,
+      "loss": 4.0237,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704004224421059e-05,
+      "loss": 4.024,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703165629670007e-05,
+      "loss": 4.0366,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702327034918955e-05,
+      "loss": 4.0176,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701488440167902e-05,
+      "loss": 4.0344,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.70064984541685e-05,
+      "loss": 4.0306,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699811250665798e-05,
+      "loss": 4.0257,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69897429379512e-05,
+      "loss": 4.0304,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698135699044067e-05,
+      "loss": 4.0182,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697298742173389e-05,
+      "loss": 4.0234,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6964601474223374e-05,
+      "loss": 4.0261,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6956215526712854e-05,
+      "loss": 4.0372,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694782957920233e-05,
+      "loss": 4.0177,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693944363169181e-05,
+      "loss": 4.0333,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693107406298502e-05,
+      "loss": 4.0059,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922688115474496e-05,
+      "loss": 4.0294,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6914302167963976e-05,
+      "loss": 4.0314,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6905916220453456e-05,
+      "loss": 4.035,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6897530272942936e-05,
+      "loss": 4.0246,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6889144325432416e-05,
+      "loss": 4.0282,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6880758377921896e-05,
+      "loss": 4.047,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6872372430411376e-05,
+      "loss": 4.0176,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686400286170459e-05,
+      "loss": 4.038,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685561691419407e-05,
+      "loss": 4.0215,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684723096668355e-05,
+      "loss": 4.0243,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683884501917303e-05,
+      "loss": 4.0312,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683045907166251e-05,
+      "loss": 4.0229,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682208950295572e-05,
+      "loss": 4.0353,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.68137035554452e-05,
+      "loss": 4.0384,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680531760793468e-05,
+      "loss": 4.0216,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679693166042416e-05,
+      "loss": 4.0211,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678854571291364e-05,
+      "loss": 4.0235,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678015976540312e-05,
+      "loss": 4.0356,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.67717738178926e-05,
+      "loss": 4.0207,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676340424918581e-05,
+      "loss": 4.0334,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675501830167529e-05,
+      "loss": 4.0273,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6746632354164776e-05,
+      "loss": 4.0259,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738246406654256e-05,
+      "loss": 4.019,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729860459143736e-05,
+      "loss": 4.0106,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6721490890436945e-05,
+      "loss": 4.0199,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6713121321730154e-05,
+      "loss": 4.0303,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704735374219634e-05,
+      "loss": 4.0323,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696349426709114e-05,
+      "loss": 4.0168,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668797985800232e-05,
+      "loss": 4.0156,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66795939104918e-05,
+      "loss": 4.0227,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667120796298128e-05,
+      "loss": 4.0282,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666282201547076e-05,
+      "loss": 4.0326,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665443606796024e-05,
+      "loss": 4.0177,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664605012044973e-05,
+      "loss": 4.0294,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663766417293921e-05,
+      "loss": 4.0217,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662927822542868e-05,
+      "loss": 4.0198,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662089227791816e-05,
+      "loss": 4.0164,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661250633040764e-05,
+      "loss": 4.0207,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660412038289712e-05,
+      "loss": 4.0251,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65957344353866e-05,
+      "loss": 4.017,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658736486667981e-05,
+      "loss": 4.0209,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657897891916929e-05,
+      "loss": 4.0145,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657060935046251e-05,
+      "loss": 4.0211,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656222340295198e-05,
+      "loss": 4.0236,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655383745544147e-05,
+      "loss": 4.017,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654545150793095e-05,
+      "loss": 4.0248,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653706556042043e-05,
+      "loss": 4.0377,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652867961290991e-05,
+      "loss": 4.0257,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652029366539939e-05,
+      "loss": 4.0275,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651190771788887e-05,
+      "loss": 4.0229,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650352177037835e-05,
+      "loss": 4.0157,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649513582286783e-05,
+      "loss": 4.0319,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6486766254161036e-05,
+      "loss": 4.0117,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6478380306650516e-05,
+      "loss": 4.0181,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6469994359139996e-05,
+      "loss": 4.0247,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6461608411629476e-05,
+      "loss": 4.0229,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6453222464118956e-05,
+      "loss": 4.0215,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6444852895412165e-05,
+      "loss": 4.0195,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643646694790165e-05,
+      "loss": 4.025,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642808100039113e-05,
+      "loss": 4.0197,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641969505288061e-05,
+      "loss": 4.0269,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641130910537009e-05,
+      "loss": 4.0208,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640292315785957e-05,
+      "loss": 4.0162,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6394537210349045e-05,
+      "loss": 3.9976,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6386151262838525e-05,
+      "loss": 4.0295,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6377765315328005e-05,
+      "loss": 4.0217,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6369379367817485e-05,
+      "loss": 4.0348,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6361009799110694e-05,
+      "loss": 4.0189,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6352623851600174e-05,
+      "loss": 4.0237,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634425428289339e-05,
+      "loss": 4.0231,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633586833538287e-05,
+      "loss": 4.0383,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632748238787235e-05,
+      "loss": 4.0097,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631909644036183e-05,
+      "loss": 4.0197,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631071049285131e-05,
+      "loss": 4.0213,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630232454534079e-05,
+      "loss": 4.025,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629393859783027e-05,
+      "loss": 4.0174,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628555265031975e-05,
+      "loss": 4.0181,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627718308161296e-05,
+      "loss": 4.0189,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626879713410244e-05,
+      "loss": 4.0269,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626041118659192e-05,
+      "loss": 4.0213,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.068659782409668,
+      "eval_runtime": 293.169,
+      "eval_samples_per_second": 1301.608,
+      "eval_steps_per_second": 40.676,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62520252390814e-05,
+      "loss": 4.0137,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6243672049178337e-05,
+      "loss": 4.0224,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623528610166782e-05,
+      "loss": 4.0195,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62269001541573e-05,
+      "loss": 4.0208,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621851420664678e-05,
+      "loss": 4.027,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621012825913626e-05,
+      "loss": 4.0195,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620174231162574e-05,
+      "loss": 4.0201,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619337274291895e-05,
+      "loss": 4.0122,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618498679540843e-05,
+      "loss": 4.0125,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617660084789791e-05,
+      "loss": 4.0235,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616821490038739e-05,
+      "loss": 4.0244,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615982895287687e-05,
+      "loss": 4.0203,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615145938417008e-05,
+      "loss": 4.0326,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614307343665956e-05,
+      "loss": 4.0107,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613468748914904e-05,
+      "loss": 4.0139,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612630154163852e-05,
+      "loss": 4.0064,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611793197293174e-05,
+      "loss": 4.0228,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610954602542122e-05,
+      "loss": 4.0127,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61011600779107e-05,
+      "loss": 4.0149,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609277413040018e-05,
+      "loss": 4.0218,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608438818288966e-05,
+      "loss": 4.0303,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607600223537914e-05,
+      "loss": 4.0206,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6067616287868617e-05,
+      "loss": 4.0158,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6059230340358097e-05,
+      "loss": 4.0187,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6050860771651306e-05,
+      "loss": 4.018,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6042474824140786e-05,
+      "loss": 4.0155,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6034105255433995e-05,
+      "loss": 4.0008,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6025719307923475e-05,
+      "loss": 4.0162,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601733336041296e-05,
+      "loss": 4.0172,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600894741290244e-05,
+      "loss": 4.0015,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600056146539192e-05,
+      "loss": 4.0145,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59921755178814e-05,
+      "loss": 4.0203,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598380594917461e-05,
+      "loss": 4.0154,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597542000166409e-05,
+      "loss": 4.0259,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596703405415357e-05,
+      "loss": 4.0173,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595864810664305e-05,
+      "loss": 4.0118,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595026215913253e-05,
+      "loss": 4.0351,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5941876211622e-05,
+      "loss": 3.9941,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593349026411148e-05,
+      "loss": 4.005,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592510431660096e-05,
+      "loss": 4.0128,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591671836909044e-05,
+      "loss": 4.0066,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590834880038366e-05,
+      "loss": 4.013,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589996285287314e-05,
+      "loss": 4.0174,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589157690536262e-05,
+      "loss": 4.0106,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588320733665583e-05,
+      "loss": 4.0135,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5874837767949044e-05,
+      "loss": 4.0071,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5866451820438524e-05,
+      "loss": 4.0172,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5858065872928004e-05,
+      "loss": 4.0179,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584967992541748e-05,
+      "loss": 3.9957,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584131035671069e-05,
+      "loss": 4.0148,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583292440920017e-05,
+      "loss": 4.0179,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582453846168965e-05,
+      "loss": 4.0211,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581615251417913e-05,
+      "loss": 4.0095,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580776656666861e-05,
+      "loss": 4.0075,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579938061915809e-05,
+      "loss": 4.0094,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57910110504513e-05,
+      "loss": 4.0022,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578262510294078e-05,
+      "loss": 4.0185,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577423915543026e-05,
+      "loss": 3.9976,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576585320791974e-05,
+      "loss": 4.0148,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575746726040922e-05,
+      "loss": 4.0099,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57490813128987e-05,
+      "loss": 4.0087,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574069536538818e-05,
+      "loss": 4.0072,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573232579668139e-05,
+      "loss": 4.0057,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572393984917087e-05,
+      "loss": 4.006,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571555390166035e-05,
+      "loss": 3.9995,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570716795414983e-05,
+      "loss": 4.0221,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569878200663932e-05,
+      "loss": 4.0038,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56903960591288e-05,
+      "loss": 4.0052,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568201011161828e-05,
+      "loss": 3.991,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567362416410776e-05,
+      "loss": 4.0075,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566523821659724e-05,
+      "loss": 4.0126,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5656868647890446e-05,
+      "loss": 4.0166,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5648482700379926e-05,
+      "loss": 4.0062,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5640096752869406e-05,
+      "loss": 4.0121,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5631710805358886e-05,
+      "loss": 4.0241,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562332485784836e-05,
+      "loss": 4.0034,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561493891033784e-05,
+      "loss": 4.0199,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560655296282732e-05,
+      "loss": 4.0032,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55981670153168e-05,
+      "loss": 4.0049,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5589797446610015e-05,
+      "loss": 4.01,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5581411499099495e-05,
+      "loss": 4.0041,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557304193039271e-05,
+      "loss": 4.0169,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556465598288219e-05,
+      "loss": 4.0236,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5556270035371664e-05,
+      "loss": 4.0076,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554790046666488e-05,
+      "loss": 3.995,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553951451915436e-05,
+      "loss": 4.0051,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553112857164383e-05,
+      "loss": 4.0148,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552274262413331e-05,
+      "loss": 4.0039,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551435667662279e-05,
+      "loss": 4.0123,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550597072911227e-05,
+      "loss": 4.0082,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549758478160175e-05,
+      "loss": 4.0098,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548919883409124e-05,
+      "loss": 4.0025,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548081288658072e-05,
+      "loss": 3.9923,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54724269390702e-05,
+      "loss": 4.0015,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546404099155968e-05,
+      "loss": 4.0059,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545565504404916e-05,
+      "loss": 4.0211,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544728547534237e-05,
+      "loss": 3.9945,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543891590663558e-05,
+      "loss": 3.9996,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543052995912506e-05,
+      "loss": 4.0073,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5422160390418267e-05,
+      "loss": 4.0086,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5413774442907746e-05,
+      "loss": 4.0127,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5405388495397226e-05,
+      "loss": 4.0021,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5397002547886706e-05,
+      "loss": 4.0143,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538861660037619e-05,
+      "loss": 4.0029,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538023065286567e-05,
+      "loss": 4.0017,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537184470535515e-05,
+      "loss": 3.9991,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536345875784463e-05,
+      "loss": 4.0022,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535507281033411e-05,
+      "loss": 4.0084,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534668686282359e-05,
+      "loss": 3.9983,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533830091531307e-05,
+      "loss": 4.0033,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532991496780255e-05,
+      "loss": 3.9942,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5321529020292026e-05,
+      "loss": 4.0012,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.531315945158524e-05,
+      "loss": 4.0062,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.530477350407472e-05,
+      "loss": 4.005,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296387556564195e-05,
+      "loss": 4.0047,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5288001609053675e-05,
+      "loss": 4.0121,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527961566154316e-05,
+      "loss": 4.0102,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527122971403264e-05,
+      "loss": 4.0125,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526284376652212e-05,
+      "loss": 4.0073,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52544578190116e-05,
+      "loss": 4.002,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524608825030481e-05,
+      "loss": 4.0093,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5237718681598027e-05,
+      "loss": 3.9953,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52293327340875e-05,
+      "loss": 4.0013,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522094678657698e-05,
+      "loss": 4.008,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521256083906646e-05,
+      "loss": 4.0057,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520417489155594e-05,
+      "loss": 4.0048,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519578894404542e-05,
+      "loss": 4.0011,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518741937533863e-05,
+      "loss": 4.0089,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517903342782811e-05,
+      "loss": 4.0012,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5170647480317595e-05,
+      "loss": 4.0111,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5162261532807075e-05,
+      "loss": 4.0019,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5153875585296555e-05,
+      "loss": 4.0017,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145506016589764e-05,
+      "loss": 3.9745,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137120069079244e-05,
+      "loss": 4.0119,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5128734121568724e-05,
+      "loss": 4.0033,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5120348174058204e-05,
+      "loss": 4.0191,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5111962226547684e-05,
+      "loss": 4.0059,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5103576279037164e-05,
+      "loss": 4.005,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5095190331526644e-05,
+      "loss": 4.0081,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5086804384016124e-05,
+      "loss": 4.0178,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5078418436505604e-05,
+      "loss": 3.9926,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507004886779881e-05,
+      "loss": 4.006,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506167929909203e-05,
+      "loss": 4.0025,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505329335158151e-05,
+      "loss": 4.0057,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504490740407099e-05,
+      "loss": 4.0035,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.503652145656047e-05,
+      "loss": 4.0018,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502813550904995e-05,
+      "loss": 4.0007,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501976594034316e-05,
+      "loss": 4.0095,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501137999283264e-05,
+      "loss": 4.0022,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.057714939117432,
+      "eval_runtime": 310.8979,
+      "eval_samples_per_second": 1227.384,
+      "eval_steps_per_second": 38.357,
+      "step": 915840
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.500299404532212e-05,
+      "loss": 3.9981,
+      "step": 915968
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.49946080978116e-05,
+      "loss": 4.0057,
+      "step": 916480
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.498622215030108e-05,
+      "loss": 3.9997,
+      "step": 916992
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.497783620279056e-05,
+      "loss": 4.0069,
+      "step": 917504
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.496945025528003e-05,
+      "loss": 4.0035,
+      "step": 918016
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.496106430776952e-05,
+      "loss": 4.0082,
+      "step": 918528
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.4952694739062733e-05,
+      "loss": 4.0006,
+      "step": 919040
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.494432517035594e-05,
+      "loss": 3.9923,
+      "step": 919552
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.493593922284542e-05,
+      "loss": 3.9952,
+      "step": 920064
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.49275532753349e-05,
+      "loss": 4.0146,
+      "step": 920576
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.491916732782438e-05,
+      "loss": 4.0045,
+      "step": 921088
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.4910781380313856e-05,
+      "loss": 3.9987,
+      "step": 921600
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.49024281904108e-05,
+      "loss": 4.0182,
+      "step": 922112
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.489404224290028e-05,
+      "loss": 3.9919,
+      "step": 922624
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.488565629538976e-05,
+      "loss": 3.9975,
+      "step": 923136
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.487727034787924e-05,
+      "loss": 3.9908,
+      "step": 923648
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.486888440036872e-05,
+      "loss": 4.0038,
+      "step": 924160
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.48604984528582e-05,
+      "loss": 3.9983,
+      "step": 924672
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.485211250534769e-05,
+      "loss": 4.0003,
+      "step": 925184
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.4843742936640896e-05,
+      "loss": 4.0052,
+      "step": 925696
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.4835356989130376e-05,
+      "loss": 4.0092,
+      "step": 926208
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.4826971041619856e-05,
+      "loss": 4.008,
+      "step": 926720
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.481858509410933e-05,
+      "loss": 3.9952,
+      "step": 927232
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.481019914659881e-05,
+      "loss": 4.0032,
+      "step": 927744
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.480181319908829e-05,
+      "loss": 4.0037,
+      "step": 928256
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.479342725157777e-05,
+      "loss": 3.9983,
+      "step": 928768
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.478504130406725e-05,
+      "loss": 3.9841,
+      "step": 929280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.477665535655673e-05,
+      "loss": 3.9957,
+      "step": 929792
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.476826940904621e-05,
+      "loss": 4.0029,
+      "step": 930304
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.475988346153569e-05,
+      "loss": 3.9863,
+      "step": 930816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.475149751402517e-05,
+      "loss": 3.9965,
+      "step": 931328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4743127945318385e-05,
+      "loss": 3.998,
+      "step": 931840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4734758376611594e-05,
+      "loss": 4.0029,
+      "step": 932352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4726372429101074e-05,
+      "loss": 4.0052,
+      "step": 932864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4717986481590554e-05,
+      "loss": 3.9991,
+      "step": 933376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.470961691288376e-05,
+      "loss": 4.0003,
+      "step": 933888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.470123096537324e-05,
+      "loss": 4.016,
+      "step": 934400
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.469284501786272e-05,
+      "loss": 3.9771,
+      "step": 934912
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.46844590703522e-05,
+      "loss": 3.9904,
+      "step": 935424
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.467607312284168e-05,
+      "loss": 3.9969,
+      "step": 935936
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.466768717533116e-05,
+      "loss": 3.9862,
+      "step": 936448
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.465930122782064e-05,
+      "loss": 3.9988,
+      "step": 936960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.465091528031012e-05,
+      "loss": 3.9989,
+      "step": 937472
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.46425293327996e-05,
+      "loss": 3.9941,
+      "step": 937984
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.463414338528909e-05,
+      "loss": 3.9963,
+      "step": 938496
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.462575743777857e-05,
+      "loss": 3.9915,
+      "step": 939008
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.461737149026805e-05,
+      "loss": 3.9974,
+      "step": 939520
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.460900192156126e-05,
+      "loss": 4.008,
+      "step": 940032
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.460061597405074e-05,
+      "loss": 3.9731,
+      "step": 940544
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.459223002654022e-05,
+      "loss": 3.9957,
+      "step": 941056
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.458384407902969e-05,
+      "loss": 4.0037,
+      "step": 941568
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.457547451032291e-05,
+      "loss": 4.0051,
+      "step": 942080
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4567104941616116e-05,
+      "loss": 3.9975,
+      "step": 942592
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4558718994105596e-05,
+      "loss": 3.9883,
+      "step": 943104
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4550333046595076e-05,
+      "loss": 3.995,
+      "step": 943616
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4541947099084556e-05,
+      "loss": 3.9842,
+      "step": 944128
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.453356115157404e-05,
+      "loss": 4.0001,
+      "step": 944640
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.452519158286725e-05,
+      "loss": 3.9821,
+      "step": 945152
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.451680563535673e-05,
+      "loss": 4.003,
+      "step": 945664
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.450843606664994e-05,
+      "loss": 3.9915,
+      "step": 946176
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.450005011913942e-05,
+      "loss": 3.9942,
+      "step": 946688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.44916641716289e-05,
+      "loss": 3.9872,
+      "step": 947200
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.448327822411838e-05,
+      "loss": 3.9928,
+      "step": 947712
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.447489227660786e-05,
+      "loss": 3.9881,
+      "step": 948224
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.446652270790107e-05,
+      "loss": 3.9862,
+      "step": 948736
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.445813676039055e-05,
+      "loss": 4.0027,
+      "step": 949248
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.444975081288003e-05,
+      "loss": 3.9944,
+      "step": 949760
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.444136486536951e-05,
+      "loss": 3.985,
+      "step": 950272
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.443297891785899e-05,
+      "loss": 3.9761,
+      "step": 950784
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.442459297034847e-05,
+      "loss": 3.9954,
+      "step": 951296
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4416223401641686e-05,
+      "loss": 3.9917,
+      "step": 951808
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4407837454131166e-05,
+      "loss": 4.0049,
+      "step": 952320
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.439945150662064e-05,
+      "loss": 3.9865,
+      "step": 952832
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.439106555911012e-05,
+      "loss": 3.999,
+      "step": 953344
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.43826796115996e-05,
+      "loss": 4.0058,
+      "step": 953856
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.437429366408908e-05,
+      "loss": 3.9911,
+      "step": 954368
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.436592409538229e-05,
+      "loss": 4.0002,
+      "step": 954880
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.435753814787177e-05,
+      "loss": 3.9899,
+      "step": 955392
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.434915220036125e-05,
+      "loss": 3.9883,
+      "step": 955904
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4340766252850734e-05,
+      "loss": 3.9972,
+      "step": 956416
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4332380305340214e-05,
+      "loss": 3.988,
+      "step": 956928
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4323994357829694e-05,
+      "loss": 3.9932,
+      "step": 957440
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4315608410319174e-05,
+      "loss": 4.0096,
+      "step": 957952
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.4307222462808654e-05,
+      "loss": 3.995,
+      "step": 958464
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.429885289410186e-05,
+      "loss": 3.9777,
+      "step": 958976
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.429046694659134e-05,
+      "loss": 3.9906,
+      "step": 959488
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.428208099908082e-05,
+      "loss": 3.9968,
+      "step": 960000
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.42736950515703e-05,
+      "loss": 3.994,
+      "step": 960512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.426530910405978e-05,
+      "loss": 3.9944,
+      "step": 961024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.425693953535299e-05,
+      "loss": 3.9941,
+      "step": 961536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.424855358784247e-05,
+      "loss": 3.9933,
+      "step": 962048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.424016764033195e-05,
+      "loss": 3.9841,
+      "step": 962560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.423178169282143e-05,
+      "loss": 3.9741,
+      "step": 963072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.422339574531092e-05,
+      "loss": 3.9861,
+      "step": 963584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.42150097978004e-05,
+      "loss": 3.9883,
+      "step": 964096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.420664022909361e-05,
+      "loss": 4.0095,
+      "step": 964608
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.419825428158309e-05,
+      "loss": 3.9788,
+      "step": 965120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.418986833407257e-05,
+      "loss": 3.9826,
+      "step": 965632
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.418148238656205e-05,
+      "loss": 3.9911,
+      "step": 966144
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.417309643905153e-05,
+      "loss": 3.9926,
+      "step": 966656
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4164710491541e-05,
+      "loss": 3.9984,
+      "step": 967168
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.415632454403048e-05,
+      "loss": 3.9879,
+      "step": 967680
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.414793859651996e-05,
+      "loss": 3.9912,
+      "step": 968192
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.413956902781318e-05,
+      "loss": 3.9907,
+      "step": 968704
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4131183080302657e-05,
+      "loss": 3.9847,
+      "step": 969216
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.412281351159587e-05,
+      "loss": 3.9854,
+      "step": 969728
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.411442756408535e-05,
+      "loss": 3.9893,
+      "step": 970240
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4106041616574826e-05,
+      "loss": 3.9931,
+      "step": 970752
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4097655669064306e-05,
+      "loss": 3.9829,
+      "step": 971264
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4089269721553786e-05,
+      "loss": 3.9849,
+      "step": 971776
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4080900152847e-05,
+      "loss": 3.9805,
+      "step": 972288
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4072514205336475e-05,
+      "loss": 3.9867,
+      "step": 972800
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4064128257825955e-05,
+      "loss": 3.9905,
+      "step": 973312
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4055742310315434e-05,
+      "loss": 3.9875,
+      "step": 973824
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4047356362804914e-05,
+      "loss": 3.9888,
+      "step": 974336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.4038970415294394e-05,
+      "loss": 3.9943,
+      "step": 974848
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.403060084658761e-05,
+      "loss": 3.9987,
+      "step": 975360
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.402223127788082e-05,
+      "loss": 3.9956,
+      "step": 975872
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.40138453303703e-05,
+      "loss": 3.9915,
+      "step": 976384
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.400545938285978e-05,
+      "loss": 3.9881,
+      "step": 976896
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.399707343534926e-05,
+      "loss": 3.9903,
+      "step": 977408
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.398868748783874e-05,
+      "loss": 3.98,
+      "step": 977920
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.398030154032822e-05,
+      "loss": 3.9875,
+      "step": 978432
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.39719155928177e-05,
+      "loss": 3.9901,
+      "step": 978944
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.396352964530718e-05,
+      "loss": 3.9908,
+      "step": 979456
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.395514369779666e-05,
+      "loss": 3.992,
+      "step": 979968
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.394677412908987e-05,
+      "loss": 3.985,
+      "step": 980480
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.393838818157935e-05,
+      "loss": 3.9887,
+      "step": 980992
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.393000223406883e-05,
+      "loss": 3.9849,
+      "step": 981504
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.392161628655831e-05,
+      "loss": 3.9972,
+      "step": 982016
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.391323033904779e-05,
+      "loss": 3.9872,
+      "step": 982528
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.3904860770341004e-05,
+      "loss": 3.9865,
+      "step": 983040
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.3896474822830484e-05,
+      "loss": 3.9636,
+      "step": 983552
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.3888088875319964e-05,
+      "loss": 3.9913,
+      "step": 984064
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.3879702927809444e-05,
+      "loss": 3.9921,
+      "step": 984576
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.3871316980298924e-05,
+      "loss": 4.0018,
+      "step": 985088
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.3862931032788404e-05,
+      "loss": 3.9922,
+      "step": 985600
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.3854545085277884e-05,
+      "loss": 3.9856,
+      "step": 986112
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.3846159137767363e-05,
+      "loss": 3.9938,
+      "step": 986624
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.383777319025684e-05,
+      "loss": 4.0018,
+      "step": 987136
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.382940362155005e-05,
+      "loss": 3.9792,
+      "step": 987648
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.382101767403953e-05,
+      "loss": 3.9936,
+      "step": 988160
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.381264810533274e-05,
+      "loss": 3.9866,
+      "step": 988672
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.380426215782223e-05,
+      "loss": 3.9927,
+      "step": 989184
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.379587621031171e-05,
+      "loss": 3.9889,
+      "step": 989696
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.378749026280119e-05,
+      "loss": 3.9858,
+      "step": 990208
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.37791206940944e-05,
+      "loss": 3.9852,
+      "step": 990720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.377073474658388e-05,
+      "loss": 3.9925,
+      "step": 991232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.376234879907336e-05,
+      "loss": 3.9921,
+      "step": 991744
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.049084663391113,
+      "eval_runtime": 290.0513,
+      "eval_samples_per_second": 1315.599,
+      "eval_steps_per_second": 41.113,
+      "step": 992160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375396285156284e-05,
+      "loss": 3.9905,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374557690405231e-05,
+      "loss": 3.9913,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373719095654179e-05,
+      "loss": 3.9847,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372880500903127e-05,
+      "loss": 3.9936,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372041906152075e-05,
+      "loss": 3.9926,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371203311401023e-05,
+      "loss": 3.9882,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370364716649971e-05,
+      "loss": 3.9871,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36952612189892e-05,
+      "loss": 3.9771,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.368687527147868e-05,
+      "loss": 3.9831,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3678505702771886e-05,
+      "loss": 3.9957,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3670119755261366e-05,
+      "loss": 3.9894,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3661733807750846e-05,
+      "loss": 3.9908,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3653347860240326e-05,
+      "loss": 4.0006,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3644961912729806e-05,
+      "loss": 3.9809,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3636575965219286e-05,
+      "loss": 3.9819,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3628190017708766e-05,
+      "loss": 3.9771,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3619804070198246e-05,
+      "loss": 3.989,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3611418122687726e-05,
+      "loss": 3.9828,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36030321751772e-05,
+      "loss": 3.9833,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.359464622766668e-05,
+      "loss": 3.9893,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3586276658959895e-05,
+      "loss": 3.9974,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3577890711449375e-05,
+      "loss": 3.9915,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3569504763938855e-05,
+      "loss": 3.9841,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3561118816428334e-05,
+      "loss": 3.9845,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3552732868917814e-05,
+      "loss": 3.9927,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3544346921407294e-05,
+      "loss": 3.9832,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3535960973896774e-05,
+      "loss": 3.9691,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3527575026386254e-05,
+      "loss": 3.9837,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3519189078875734e-05,
+      "loss": 3.9825,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3510803131365214e-05,
+      "loss": 3.9742,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3502417183854694e-05,
+      "loss": 3.9836,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3494031236344174e-05,
+      "loss": 3.9823,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3485645288833654e-05,
+      "loss": 3.9889,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3477259341323134e-05,
+      "loss": 3.9919,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346888977261634e-05,
+      "loss": 3.9874,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346052020390956e-05,
+      "loss": 3.9869,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345213425639904e-05,
+      "loss": 3.9998,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344374830888852e-05,
+      "loss": 3.9654,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3435362361378e-05,
+      "loss": 3.9791,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342697641386748e-05,
+      "loss": 3.9794,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341859046635696e-05,
+      "loss": 3.9684,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341020451884644e-05,
+      "loss": 3.9861,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340181857133591e-05,
+      "loss": 3.9865,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.339344900262913e-05,
+      "loss": 3.9797,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.338506305511861e-05,
+      "loss": 3.9778,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.337667710760808e-05,
+      "loss": 3.9826,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.336829116009757e-05,
+      "loss": 3.9816,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335993797019451e-05,
+      "loss": 3.9922,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.335155202268399e-05,
+      "loss": 3.9616,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.334316607517347e-05,
+      "loss": 3.9752,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.333478012766295e-05,
+      "loss": 3.9944,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.332639418015243e-05,
+      "loss": 3.9901,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.331802461144564e-05,
+      "loss": 3.9812,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.330963866393512e-05,
+      "loss": 3.9718,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.33012527164246e-05,
+      "loss": 3.9793,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.329286676891408e-05,
+      "loss": 3.9687,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3284480821403555e-05,
+      "loss": 3.9871,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.327611125269677e-05,
+      "loss": 3.969,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.326772530518625e-05,
+      "loss": 3.9842,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325933935767574e-05,
+      "loss": 3.9806,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325095341016521e-05,
+      "loss": 3.9827,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324256746265469e-05,
+      "loss": 3.9729,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3234197893947906e-05,
+      "loss": 3.9799,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3225811946437386e-05,
+      "loss": 3.9715,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321742599892686e-05,
+      "loss": 3.9695,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320904005141634e-05,
+      "loss": 3.9879,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320065410390582e-05,
+      "loss": 3.9815,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31922681563953e-05,
+      "loss": 3.9726,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318389858768851e-05,
+      "loss": 3.9628,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317551264017799e-05,
+      "loss": 3.9775,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3167126692667475e-05,
+      "loss": 3.9784,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3158740745156955e-05,
+      "loss": 3.9942,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3150354797646435e-05,
+      "loss": 3.9716,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3141968850135915e-05,
+      "loss": 3.9849,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3133582902625395e-05,
+      "loss": 3.993,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3125196955114875e-05,
+      "loss": 3.9738,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3116827386408084e-05,
+      "loss": 3.9805,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3108441438897564e-05,
+      "loss": 3.9792,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3100055491387044e-05,
+      "loss": 3.9776,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309168592268025e-05,
+      "loss": 3.978,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308329997516973e-05,
+      "loss": 3.9756,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307491402765921e-05,
+      "loss": 3.9824,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306652808014869e-05,
+      "loss": 3.9945,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.305814213263817e-05,
+      "loss": 3.9804,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304975618512766e-05,
+      "loss": 3.963,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304137023761714e-05,
+      "loss": 3.9773,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303298429010662e-05,
+      "loss": 3.9833,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30245983425961e-05,
+      "loss": 3.9792,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.301622877388931e-05,
+      "loss": 3.9836,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300785920518252e-05,
+      "loss": 3.9782,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2999473257672e-05,
+      "loss": 3.9824,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299108731016148e-05,
+      "loss": 3.9674,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298270136265096e-05,
+      "loss": 3.9618,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297431541514044e-05,
+      "loss": 3.9732,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296592946762992e-05,
+      "loss": 3.9703,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29575435201194e-05,
+      "loss": 3.9892,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294915757260888e-05,
+      "loss": 3.9732,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.294077162509836e-05,
+      "loss": 3.9653,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.293238567758784e-05,
+      "loss": 3.9795,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.292399973007732e-05,
+      "loss": 3.9743,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2915630161370526e-05,
+      "loss": 3.9813,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2907244213860006e-05,
+      "loss": 3.9758,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2898858266349486e-05,
+      "loss": 3.9755,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2890472318838966e-05,
+      "loss": 3.9743,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2882086371328446e-05,
+      "loss": 3.9749,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2873700423817926e-05,
+      "loss": 3.9701,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2865314476307406e-05,
+      "loss": 3.9718,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2856928528796886e-05,
+      "loss": 3.9809,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2848558960090095e-05,
+      "loss": 3.9705,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2840173012579575e-05,
+      "loss": 3.9706,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283178706506906e-05,
+      "loss": 3.9674,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.282341749636227e-05,
+      "loss": 3.9731,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281503154885175e-05,
+      "loss": 3.9754,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280664560134123e-05,
+      "loss": 3.9741,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279825965383071e-05,
+      "loss": 3.9738,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278987370632019e-05,
+      "loss": 3.9829,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.27815041376134e-05,
+      "loss": 3.9828,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.277311819010288e-05,
+      "loss": 3.9817,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276473224259236e-05,
+      "loss": 3.9809,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.275634629508184e-05,
+      "loss": 3.9726,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274796034757132e-05,
+      "loss": 3.9766,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273959077886453e-05,
+      "loss": 3.9665,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2731221210157744e-05,
+      "loss": 3.9761,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2722835262647224e-05,
+      "loss": 3.9698,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2714449315136704e-05,
+      "loss": 3.9786,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2706063367626184e-05,
+      "loss": 3.9806,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2697677420115664e-05,
+      "loss": 3.9753,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268930785140887e-05,
+      "loss": 3.9735,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268092190389835e-05,
+      "loss": 3.9701,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267253595638783e-05,
+      "loss": 3.985,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266415000887731e-05,
+      "loss": 3.9758,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265576406136679e-05,
+      "loss": 3.9715,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264737811385627e-05,
+      "loss": 3.9507,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263899216634575e-05,
+      "loss": 3.9749,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263060621883523e-05,
+      "loss": 3.9804,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262223665012845e-05,
+      "loss": 3.9874,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261385070261793e-05,
+      "loss": 3.9779,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260546475510741e-05,
+      "loss": 3.969,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259707880759688e-05,
+      "loss": 3.9766,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258869286008636e-05,
+      "loss": 3.9907,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258030691257584e-05,
+      "loss": 3.9674,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.257193734386905e-05,
+      "loss": 3.9796,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.256355139635853e-05,
+      "loss": 3.9749,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255516544884801e-05,
+      "loss": 3.9751,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254677950133749e-05,
+      "loss": 3.9767,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253839355382697e-05,
+      "loss": 3.9734,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253000760631645e-05,
+      "loss": 3.9688,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.252162165880594e-05,
+      "loss": 3.9814,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2513252090099147e-05,
+      "loss": 3.9758,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.041431903839111,
+      "eval_runtime": 288.5451,
+      "eval_samples_per_second": 1322.466,
+      "eval_steps_per_second": 41.328,
+      "step": 1068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2504866142588627e-05,
+      "loss": 3.964,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2496480195078107e-05,
+      "loss": 3.9794,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2488094247567586e-05,
+      "loss": 3.9709,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2479708300057066e-05,
+      "loss": 3.9792,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2471338731350276e-05,
+      "loss": 3.9777,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2462952783839755e-05,
+      "loss": 3.9817,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2454566836329235e-05,
+      "loss": 3.9709,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2446180888818715e-05,
+      "loss": 3.9616,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2437794941308195e-05,
+      "loss": 3.9673,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2429408993797675e-05,
+      "loss": 3.984,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2421023046287155e-05,
+      "loss": 3.9731,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.241265347758037e-05,
+      "loss": 3.9738,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.240428390887358e-05,
+      "loss": 3.9873,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239589796136306e-05,
+      "loss": 3.9694,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.238751201385254e-05,
+      "loss": 3.9692,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237912606634202e-05,
+      "loss": 3.9616,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23707401188315e-05,
+      "loss": 3.9727,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236235417132098e-05,
+      "loss": 3.9697,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235396822381046e-05,
+      "loss": 3.9695,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.234558227629994e-05,
+      "loss": 3.9737,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233719632878941e-05,
+      "loss": 3.987,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232882676008263e-05,
+      "loss": 3.9753,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232044081257211e-05,
+      "loss": 3.9728,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.231205486506159e-05,
+      "loss": 3.972,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.230366891755107e-05,
+      "loss": 3.9739,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.229528297004055e-05,
+      "loss": 3.9736,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2286913401333765e-05,
+      "loss": 3.9568,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2278527453823245e-05,
+      "loss": 3.9656,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.227014150631272e-05,
+      "loss": 3.9694,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.22617555588022e-05,
+      "loss": 3.9627,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.225336961129168e-05,
+      "loss": 3.9647,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.224500004258489e-05,
+      "loss": 3.9759,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.223661409507437e-05,
+      "loss": 3.9704,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.222822814756385e-05,
+      "loss": 3.9787,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221984220005333e-05,
+      "loss": 3.9727,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221147263134654e-05,
+      "loss": 3.9706,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.220308668383602e-05,
+      "loss": 3.9825,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.21947007363255e-05,
+      "loss": 3.9565,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.218631478881498e-05,
+      "loss": 3.9686,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217792884130446e-05,
+      "loss": 3.9629,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216954289379394e-05,
+      "loss": 3.9578,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216115694628342e-05,
+      "loss": 3.9722,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.21527709987729e-05,
+      "loss": 3.9736,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.214438505126238e-05,
+      "loss": 3.9646,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.213601548255559e-05,
+      "loss": 3.9677,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.21276459138488e-05,
+      "loss": 3.9676,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211925996633828e-05,
+      "loss": 3.9723,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211087401882776e-05,
+      "loss": 3.9736,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210248807131725e-05,
+      "loss": 3.9498,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209410212380673e-05,
+      "loss": 3.9619,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2085732555099936e-05,
+      "loss": 3.9796,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2077346607589416e-05,
+      "loss": 3.9756,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2068960660078896e-05,
+      "loss": 3.9698,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2060574712568376e-05,
+      "loss": 3.9593,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2052205143861585e-05,
+      "loss": 3.9685,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2043819196351065e-05,
+      "loss": 3.9582,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2035433248840545e-05,
+      "loss": 3.9691,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2027047301330025e-05,
+      "loss": 3.9567,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2018661353819505e-05,
+      "loss": 3.9733,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2010275406308985e-05,
+      "loss": 3.9689,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.20019058376022e-05,
+      "loss": 3.9669,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.199351989009168e-05,
+      "loss": 3.9576,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.198513394258116e-05,
+      "loss": 3.9678,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.197674799507064e-05,
+      "loss": 3.9557,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.196836204756012e-05,
+      "loss": 3.9599,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.195999247885333e-05,
+      "loss": 3.9709,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.195160653134281e-05,
+      "loss": 3.9706,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.194322058383229e-05,
+      "loss": 3.9612,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.193483463632177e-05,
+      "loss": 3.9482,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.192644868881125e-05,
+      "loss": 3.965,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191806274130072e-05,
+      "loss": 3.9654,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19096767937902e-05,
+      "loss": 3.9775,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190130722508342e-05,
+      "loss": 3.9577,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.18929212775729e-05,
+      "loss": 3.9731,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188453533006238e-05,
+      "loss": 3.9721,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187614938255186e-05,
+      "loss": 3.9658,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1867779813845074e-05,
+      "loss": 3.968,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185939386633455e-05,
+      "loss": 3.9645,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.185100791882403e-05,
+      "loss": 3.9684,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.184262197131351e-05,
+      "loss": 3.9618,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.183423602380299e-05,
+      "loss": 3.9642,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1825866455096196e-05,
+      "loss": 3.9681,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1817480507585676e-05,
+      "loss": 3.979,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1809094560075156e-05,
+      "loss": 3.9667,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1800708612564636e-05,
+      "loss": 3.9499,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1792322665054116e-05,
+      "loss": 3.9658,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.17839367175436e-05,
+      "loss": 3.9675,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.177555077003308e-05,
+      "loss": 3.9669,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176716482252256e-05,
+      "loss": 3.97,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.17588116326195e-05,
+      "loss": 3.9645,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.175042568510898e-05,
+      "loss": 3.9655,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174203973759846e-05,
+      "loss": 3.9542,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173365379008794e-05,
+      "loss": 3.952,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172526784257742e-05,
+      "loss": 3.9537,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.17168818950669e-05,
+      "loss": 3.9589,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170851232636011e-05,
+      "loss": 3.9766,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170012637884959e-05,
+      "loss": 3.9579,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169174043133907e-05,
+      "loss": 3.9528,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1683354483828557e-05,
+      "loss": 3.9708,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1674968536318036e-05,
+      "loss": 3.9626,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1666598967611246e-05,
+      "loss": 3.971,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1658213020100726e-05,
+      "loss": 3.963,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1649827072590205e-05,
+      "loss": 3.96,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1641441125079685e-05,
+      "loss": 3.9636,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1633055177569165e-05,
+      "loss": 3.9624,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1624669230058645e-05,
+      "loss": 3.9541,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1616283282548125e-05,
+      "loss": 3.9631,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1607897335037605e-05,
+      "loss": 3.9619,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1599511387527085e-05,
+      "loss": 3.9591,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159112544001656e-05,
+      "loss": 3.9556,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.158273949250604e-05,
+      "loss": 3.9577,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1574369923799254e-05,
+      "loss": 3.956,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156600035509247e-05,
+      "loss": 3.9613,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155761440758195e-05,
+      "loss": 3.9627,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154922846007143e-05,
+      "loss": 3.9582,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154084251256091e-05,
+      "loss": 3.9711,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153245656505038e-05,
+      "loss": 3.9734,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152407061753986e-05,
+      "loss": 3.9668,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.151568467002934e-05,
+      "loss": 3.9636,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150729872251882e-05,
+      "loss": 3.9627,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14989127750083e-05,
+      "loss": 3.9626,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149054320630151e-05,
+      "loss": 3.9518,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.148215725879099e-05,
+      "loss": 3.9635,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.147377131128048e-05,
+      "loss": 3.9601,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.146538536376996e-05,
+      "loss": 3.9645,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.145701579506317e-05,
+      "loss": 3.9625,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144862984755265e-05,
+      "loss": 3.9644,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144026027884586e-05,
+      "loss": 3.9593,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.143187433133534e-05,
+      "loss": 3.9561,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.142348838382482e-05,
+      "loss": 3.9732,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14151024363143e-05,
+      "loss": 3.9645,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.140671648880378e-05,
+      "loss": 3.9569,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.139833054129326e-05,
+      "loss": 3.9394,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1389944593782737e-05,
+      "loss": 3.9577,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1381558646272217e-05,
+      "loss": 3.9672,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1373172698761697e-05,
+      "loss": 3.9744,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1364786751251176e-05,
+      "loss": 3.9648,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135641718254439e-05,
+      "loss": 3.9564,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.134803123503387e-05,
+      "loss": 3.9642,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133964528752335e-05,
+      "loss": 3.9778,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133125934001283e-05,
+      "loss": 3.9578,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.132288977130604e-05,
+      "loss": 3.9609,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.131450382379552e-05,
+      "loss": 3.9645,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1306117876285e-05,
+      "loss": 3.962,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129773192877448e-05,
+      "loss": 3.9614,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128934598126396e-05,
+      "loss": 3.9605,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128097641255717e-05,
+      "loss": 3.9533,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127259046504665e-05,
+      "loss": 3.9723,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126420451753613e-05,
+      "loss": 3.9621,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.034172534942627,
+      "eval_runtime": 289.1222,
+      "eval_samples_per_second": 1319.826,
+      "eval_steps_per_second": 41.246,
+      "step": 1144800
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.125581857002561e-05,
+      "loss": 3.9553,
+      "step": 1144832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.12474326225151e-05,
+      "loss": 3.9657,
+      "step": 1145344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.123904667500457e-05,
+      "loss": 3.9559,
+      "step": 1145856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.123066072749405e-05,
+      "loss": 3.9699,
+      "step": 1146368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1222291158787266e-05,
+      "loss": 3.9644,
+      "step": 1146880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1213905211276746e-05,
+      "loss": 3.9705,
+      "step": 1147392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.120551926376622e-05,
+      "loss": 3.9572,
+      "step": 1147904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.11971333162557e-05,
+      "loss": 3.95,
+      "step": 1148416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.118874736874518e-05,
+      "loss": 3.952,
+      "step": 1148928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1180377800038395e-05,
+      "loss": 3.9775,
+      "step": 1149440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.117199185252787e-05,
+      "loss": 3.9594,
+      "step": 1149952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.116360590501735e-05,
+      "loss": 3.9647,
+      "step": 1150464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1155219957506835e-05,
+      "loss": 3.9707,
+      "step": 1150976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1146850388800044e-05,
+      "loss": 3.9619,
+      "step": 1151488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1138464441289524e-05,
+      "loss": 3.9532,
+      "step": 1152000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1130078493779004e-05,
+      "loss": 3.9532,
+      "step": 1152512
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1121692546268484e-05,
+      "loss": 3.9545,
+      "step": 1153024
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1113306598757964e-05,
+      "loss": 3.958,
+      "step": 1153536
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1104920651247443e-05,
+      "loss": 3.9585,
+      "step": 1154048
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1096534703736923e-05,
+      "loss": 3.964,
+      "step": 1154560
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.1088148756226403e-05,
+      "loss": 3.9766,
+      "step": 1155072
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.107976280871588e-05,
+      "loss": 3.9629,
+      "step": 1155584
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.107137686120536e-05,
+      "loss": 3.9606,
+      "step": 1156096
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.106299091369484e-05,
+      "loss": 3.9576,
+      "step": 1156608
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.105460496618432e-05,
+      "loss": 3.9645,
+      "step": 1157120
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.104623539747753e-05,
+      "loss": 3.9563,
+      "step": 1157632
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.103786582877075e-05,
+      "loss": 3.9466,
+      "step": 1158144
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.102947988126023e-05,
+      "loss": 3.9522,
+      "step": 1158656
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.102109393374971e-05,
+      "loss": 3.9571,
+      "step": 1159168
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 3.101270798623919e-05,
+      "loss": 3.9506,
+      "step": 1159680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.100432203872867e-05,
+      "loss": 3.9512,
+      "step": 1160192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.099593609121815e-05,
+      "loss": 3.9648,
+      "step": 1160704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.098755014370763e-05,
+      "loss": 3.9595,
+      "step": 1161216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.097918057500084e-05,
+      "loss": 3.9606,
+      "step": 1161728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.097079462749032e-05,
+      "loss": 3.9676,
+      "step": 1162240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0962441437587255e-05,
+      "loss": 3.9583,
+      "step": 1162752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.095405549007674e-05,
+      "loss": 3.9647,
+      "step": 1163264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.094566954256622e-05,
+      "loss": 3.9474,
+      "step": 1163776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.09372835950557e-05,
+      "loss": 3.9556,
+      "step": 1164288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.092889764754518e-05,
+      "loss": 3.9494,
+      "step": 1164800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.092051170003466e-05,
+      "loss": 3.9516,
+      "step": 1165312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.091212575252414e-05,
+      "loss": 3.9549,
+      "step": 1165824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.090373980501362e-05,
+      "loss": 3.9647,
+      "step": 1166336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.08953538575031e-05,
+      "loss": 3.9476,
+      "step": 1166848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.088696790999258e-05,
+      "loss": 3.9579,
+      "step": 1167360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0878581962482055e-05,
+      "loss": 3.9531,
+      "step": 1167872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0870196014971535e-05,
+      "loss": 3.9584,
+      "step": 1168384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0861810067461015e-05,
+      "loss": 3.962,
+      "step": 1168896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0853424119950495e-05,
+      "loss": 3.9404,
+      "step": 1169408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0845038172439975e-05,
+      "loss": 3.9473,
+      "step": 1169920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.083666860373319e-05,
+      "loss": 3.9656,
+      "step": 1170432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.082828265622267e-05,
+      "loss": 3.9654,
+      "step": 1170944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.081989670871215e-05,
+      "loss": 3.9522,
+      "step": 1171456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.081151076120163e-05,
+      "loss": 3.9515,
+      "step": 1171968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.080312481369111e-05,
+      "loss": 3.9577,
+      "step": 1172480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.079473886618059e-05,
+      "loss": 3.9437,
+      "step": 1172992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.07863692974738e-05,
+      "loss": 3.9575,
+      "step": 1173504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.077798334996328e-05,
+      "loss": 3.9444,
+      "step": 1174016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.076959740245276e-05,
+      "loss": 3.9609,
+      "step": 1174528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.076121145494224e-05,
+      "loss": 3.9533,
+      "step": 1175040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.075284188623545e-05,
+      "loss": 3.9557,
+      "step": 1175552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.074445593872493e-05,
+      "loss": 3.945,
+      "step": 1176064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.073606999121441e-05,
+      "loss": 3.9588,
+      "step": 1176576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.072768404370389e-05,
+      "loss": 3.9391,
+      "step": 1177088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0719298096193375e-05,
+      "loss": 3.9535,
+      "step": 1177600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0710912148682855e-05,
+      "loss": 3.9591,
+      "step": 1178112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0702542579976064e-05,
+      "loss": 3.9583,
+      "step": 1178624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0694156632465544e-05,
+      "loss": 3.9458,
+      "step": 1179136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0685770684955024e-05,
+      "loss": 3.9392,
+      "step": 1179648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0677384737444504e-05,
+      "loss": 3.9493,
+      "step": 1180160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.066901516873771e-05,
+      "loss": 3.9535,
+      "step": 1180672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.066062922122719e-05,
+      "loss": 3.9667,
+      "step": 1181184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.065224327371667e-05,
+      "loss": 3.9462,
+      "step": 1181696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.064385732620615e-05,
+      "loss": 3.9618,
+      "step": 1182208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.063547137869563e-05,
+      "loss": 3.9593,
+      "step": 1182720
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.062710180998884e-05,
+      "loss": 3.9562,
+      "step": 1183232
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.061871586247833e-05,
+      "loss": 3.9559,
+      "step": 1183744
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.061034629377154e-05,
+      "loss": 3.9547,
+      "step": 1184256
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.060196034626102e-05,
+      "loss": 3.9563,
+      "step": 1184768
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.05935743987505e-05,
+      "loss": 3.9494,
+      "step": 1185280
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.058518845123998e-05,
+      "loss": 3.9518,
+      "step": 1185792
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.057680250372946e-05,
+      "loss": 3.9538,
+      "step": 1186304
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.056841655621894e-05,
+      "loss": 3.9683,
+      "step": 1186816
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.056003060870842e-05,
+      "loss": 3.9535,
+      "step": 1187328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0551661040001626e-05,
+      "loss": 3.9477,
+      "step": 1187840
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0543275092491106e-05,
+      "loss": 3.9466,
+      "step": 1188352
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0534889144980586e-05,
+      "loss": 3.9567,
+      "step": 1188864
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0526503197470066e-05,
+      "loss": 3.9541,
+      "step": 1189376
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0518117249959546e-05,
+      "loss": 3.9616,
+      "step": 1189888
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 3.0509731302449023e-05,
+      "loss": 3.9511,
+      "step": 1190400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0501345354938503e-05,
+      "loss": 3.9553,
+      "step": 1190912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0492959407427986e-05,
+      "loss": 3.9422,
+      "step": 1191424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0484573459917466e-05,
+      "loss": 3.9406,
+      "step": 1191936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0476187512406946e-05,
+      "loss": 3.9425,
+      "step": 1192448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0467817943700155e-05,
+      "loss": 3.9462,
+      "step": 1192960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0459431996189635e-05,
+      "loss": 3.9637,
+      "step": 1193472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0451046048679115e-05,
+      "loss": 3.9545,
+      "step": 1193984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0442660101168595e-05,
+      "loss": 3.9425,
+      "step": 1194496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0434274153658075e-05,
+      "loss": 3.954,
+      "step": 1195008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.042588820614756e-05,
+      "loss": 3.9543,
+      "step": 1195520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.041750225863704e-05,
+      "loss": 3.9576,
+      "step": 1196032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0409116311126518e-05,
+      "loss": 3.9522,
+      "step": 1196544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0400746742419727e-05,
+      "loss": 3.9476,
+      "step": 1197056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0392377173712936e-05,
+      "loss": 3.9508,
+      "step": 1197568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.038399122620242e-05,
+      "loss": 3.9558,
+      "step": 1198080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.03756052786919e-05,
+      "loss": 3.9382,
+      "step": 1198592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.036723570998511e-05,
+      "loss": 3.9491,
+      "step": 1199104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.035884976247459e-05,
+      "loss": 3.9566,
+      "step": 1199616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.035046381496407e-05,
+      "loss": 3.9451,
+      "step": 1200128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.034207786745355e-05,
+      "loss": 3.9472,
+      "step": 1200640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.033369191994303e-05,
+      "loss": 3.9492,
+      "step": 1201152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0325305972432512e-05,
+      "loss": 3.9436,
+      "step": 1201664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0316920024921992e-05,
+      "loss": 3.953,
+      "step": 1202176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0308534077411472e-05,
+      "loss": 3.9517,
+      "step": 1202688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.030016450870468e-05,
+      "loss": 3.9432,
+      "step": 1203200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.029177856119416e-05,
+      "loss": 3.963,
+      "step": 1203712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.028339261368364e-05,
+      "loss": 3.9572,
+      "step": 1204224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.027500666617312e-05,
+      "loss": 3.9563,
+      "step": 1204736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0266620718662604e-05,
+      "loss": 3.9559,
+      "step": 1205248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0258234771152077e-05,
+      "loss": 3.9506,
+      "step": 1205760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0249848823641557e-05,
+      "loss": 3.9505,
+      "step": 1206272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0241462876131037e-05,
+      "loss": 3.9453,
+      "step": 1206784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0233076928620517e-05,
+      "loss": 3.9465,
+      "step": 1207296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0224690981109997e-05,
+      "loss": 3.9526,
+      "step": 1207808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.021632141240321e-05,
+      "loss": 3.9507,
+      "step": 1208320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.020793546489269e-05,
+      "loss": 3.9483,
+      "step": 1208832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.019954951738217e-05,
+      "loss": 3.955,
+      "step": 1209344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.019116356987165e-05,
+      "loss": 3.9491,
+      "step": 1209856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.018279400116486e-05,
+      "loss": 3.9441,
+      "step": 1210368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0174408053654342e-05,
+      "loss": 3.9615,
+      "step": 1210880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0166022106143822e-05,
+      "loss": 3.9543,
+      "step": 1211392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0157636158633302e-05,
+      "loss": 3.9461,
+      "step": 1211904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0149250211122782e-05,
+      "loss": 3.9334,
+      "step": 1212416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0140864263612262e-05,
+      "loss": 3.9461,
+      "step": 1212928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0132478316101742e-05,
+      "loss": 3.9585,
+      "step": 1213440
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0124092368591222e-05,
+      "loss": 3.962,
+      "step": 1213952
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0115706421080702e-05,
+      "loss": 3.9519,
+      "step": 1214464
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0107336852373914e-05,
+      "loss": 3.9464,
+      "step": 1214976
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0098967283667123e-05,
+      "loss": 3.9553,
+      "step": 1215488
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0090581336156603e-05,
+      "loss": 3.9655,
+      "step": 1216000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0082195388646083e-05,
+      "loss": 3.9469,
+      "step": 1216512
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0073809441135563e-05,
+      "loss": 3.9524,
+      "step": 1217024
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0065423493625043e-05,
+      "loss": 3.9502,
+      "step": 1217536
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0057037546114526e-05,
+      "loss": 3.9513,
+      "step": 1218048
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0048651598604006e-05,
+      "loss": 3.95,
+      "step": 1218560
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0040265651093486e-05,
+      "loss": 3.9555,
+      "step": 1219072
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0031912461190425e-05,
+      "loss": 3.9412,
+      "step": 1219584
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0023526513679905e-05,
+      "loss": 3.9583,
+      "step": 1220096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 3.0015140566169388e-05,
+      "loss": 3.9496,
+      "step": 1220608
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 3.0006754618658868e-05,
+      "loss": 3.9449,
+      "step": 1221120
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.028400421142578,
+      "eval_runtime": 289.1133,
+      "eval_samples_per_second": 1319.867,
+      "eval_steps_per_second": 41.247,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998368671148348e-05,
+      "loss": 3.955,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989982723637828e-05,
+      "loss": 3.9473,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981596776127308e-05,
+      "loss": 3.9573,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9973210828616788e-05,
+      "loss": 3.9526,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9964824881106264e-05,
+      "loss": 3.9593,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9956438933595744e-05,
+      "loss": 3.9463,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9948052986085224e-05,
+      "loss": 3.9376,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939667038574704e-05,
+      "loss": 3.9407,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931281091064184e-05,
+      "loss": 3.9653,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922895143553664e-05,
+      "loss": 3.9502,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914509196043144e-05,
+      "loss": 3.9476,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9906123248532624e-05,
+      "loss": 3.9584,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9897753679825836e-05,
+      "loss": 3.9515,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889367732315316e-05,
+      "loss": 3.9434,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9880981784804796e-05,
+      "loss": 3.946,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9872595837294276e-05,
+      "loss": 3.9396,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9864209889783756e-05,
+      "loss": 3.9487,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9855823942273236e-05,
+      "loss": 3.9431,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9847437994762716e-05,
+      "loss": 3.9559,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9839052047252196e-05,
+      "loss": 3.9672,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983068247854541e-05,
+      "loss": 3.9501,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.982229653103489e-05,
+      "loss": 3.9448,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.981391058352437e-05,
+      "loss": 3.9493,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.980552463601385e-05,
+      "loss": 3.9548,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.979713868850332e-05,
+      "loss": 3.9442,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788752740992805e-05,
+      "loss": 3.9379,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978038317228602e-05,
+      "loss": 3.9394,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.97719972247755e-05,
+      "loss": 3.9544,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763611277264974e-05,
+      "loss": 3.9369,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755225329754454e-05,
+      "loss": 3.9393,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9746839382243934e-05,
+      "loss": 3.9524,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9738453434733414e-05,
+      "loss": 3.9514,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9730083866026626e-05,
+      "loss": 3.9504,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9721697918516106e-05,
+      "loss": 3.9547,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9713328349809322e-05,
+      "loss": 3.9468,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9704942402298795e-05,
+      "loss": 3.9561,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9696556454788275e-05,
+      "loss": 3.9366,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.968817050727776e-05,
+      "loss": 3.9457,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967978455976724e-05,
+      "loss": 3.9399,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.967139861225672e-05,
+      "loss": 3.9389,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.96630126647462e-05,
+      "loss": 3.9451,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.965462671723568e-05,
+      "loss": 3.9532,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9646257148528888e-05,
+      "loss": 3.9379,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9637871201018368e-05,
+      "loss": 3.9432,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962948525350785e-05,
+      "loss": 3.9436,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962109930599733e-05,
+      "loss": 3.9469,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961271335848681e-05,
+      "loss": 3.9526,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.960432741097629e-05,
+      "loss": 3.935,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.959594146346577e-05,
+      "loss": 3.9349,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.958755551595525e-05,
+      "loss": 3.9531,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957918594724846e-05,
+      "loss": 3.957,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9570816378541672e-05,
+      "loss": 3.94,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562430431031152e-05,
+      "loss": 3.9433,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9554044483520632e-05,
+      "loss": 3.9475,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545658536010112e-05,
+      "loss": 3.9318,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9537272588499592e-05,
+      "loss": 3.945,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528886640989072e-05,
+      "loss": 3.9343,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520500693478552e-05,
+      "loss": 3.9498,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9512114745968032e-05,
+      "loss": 3.9452,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503745177261244e-05,
+      "loss": 3.9464,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9495359229750724e-05,
+      "loss": 3.929,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486973282240204e-05,
+      "loss": 3.9503,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478587334729684e-05,
+      "loss": 3.9292,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.947020138721916e-05,
+      "loss": 3.9423,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.946181543970864e-05,
+      "loss": 3.9481,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.945342949219812e-05,
+      "loss": 3.9499,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.94450435446876e-05,
+      "loss": 3.9352,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.943667397598081e-05,
+      "loss": 3.9317,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942828802847029e-05,
+      "loss": 3.9345,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9419934838567238e-05,
+      "loss": 3.9451,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411548891056718e-05,
+      "loss": 3.95,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9403162943546198e-05,
+      "loss": 3.9393,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394776996035678e-05,
+      "loss": 3.9484,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386391048525158e-05,
+      "loss": 3.954,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9378005101014635e-05,
+      "loss": 3.9439,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369619153504115e-05,
+      "loss": 3.9445,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9361233205993594e-05,
+      "loss": 3.9437,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.935286363728681e-05,
+      "loss": 3.9485,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344477689776284e-05,
+      "loss": 3.9364,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9336091742265763e-05,
+      "loss": 3.9428,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9327705794755243e-05,
+      "loss": 3.9413,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9319319847244723e-05,
+      "loss": 3.959,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310933899734207e-05,
+      "loss": 3.9452,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302547952223687e-05,
+      "loss": 3.9318,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9294178383516896e-05,
+      "loss": 3.9397,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9285792436006376e-05,
+      "loss": 3.9421,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9277406488495856e-05,
+      "loss": 3.9484,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9269020540985336e-05,
+      "loss": 3.9495,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260650972278548e-05,
+      "loss": 3.9413,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252265024768028e-05,
+      "loss": 3.9418,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243879077257508e-05,
+      "loss": 3.9361,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235493129746988e-05,
+      "loss": 3.9314,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9227107182236468e-05,
+      "loss": 3.9309,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9218721234725948e-05,
+      "loss": 3.9418,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921035166601916e-05,
+      "loss": 3.9509,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.920196571850864e-05,
+      "loss": 3.9425,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.919357977099812e-05,
+      "loss": 3.9332,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91851938234876e-05,
+      "loss": 3.9383,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.917682425478081e-05,
+      "loss": 3.9451,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916843830727029e-05,
+      "loss": 3.947,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.916005235975977e-05,
+      "loss": 3.9423,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9151682791052982e-05,
+      "loss": 3.9389,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143296843542462e-05,
+      "loss": 3.9382,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.913491089603194e-05,
+      "loss": 3.9471,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.912652494852142e-05,
+      "loss": 3.9297,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91181390010109e-05,
+      "loss": 3.9344,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910975305350038e-05,
+      "loss": 3.9453,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.910136710598986e-05,
+      "loss": 3.9373,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092997537283074e-05,
+      "loss": 3.9352,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9084611589772554e-05,
+      "loss": 3.9393,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076225642262034e-05,
+      "loss": 3.9353,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067839694751514e-05,
+      "loss": 3.9395,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059453747240994e-05,
+      "loss": 3.9369,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.905106779973047e-05,
+      "loss": 3.9346,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904268185221995e-05,
+      "loss": 3.9496,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.903429590470943e-05,
+      "loss": 3.9491,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025926336002646e-05,
+      "loss": 3.9438,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901754038849212e-05,
+      "loss": 3.9511,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.90091544409816e-05,
+      "loss": 3.9401,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000768493471083e-05,
+      "loss": 3.9393,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992382545960563e-05,
+      "loss": 3.9321,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8983996598450043e-05,
+      "loss": 3.9377,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8975610650939523e-05,
+      "loss": 3.9409,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8967224703429002e-05,
+      "loss": 3.9411,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958838755918482e-05,
+      "loss": 3.9406,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.895046918721169e-05,
+      "loss": 3.9411,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8942083239701175e-05,
+      "loss": 3.9422,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933697292190655e-05,
+      "loss": 3.9324,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925311344680135e-05,
+      "loss": 3.9466,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916958154777073e-05,
+      "loss": 3.9447,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908572207266553e-05,
+      "loss": 3.9334,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8900186259756036e-05,
+      "loss": 3.9247,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891800312245516e-05,
+      "loss": 3.9351,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883414364734996e-05,
+      "loss": 3.9476,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8875028417224476e-05,
+      "loss": 3.953,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8866658848517685e-05,
+      "loss": 3.9359,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8858272901007165e-05,
+      "loss": 3.9378,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849886953496645e-05,
+      "loss": 3.9493,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841501005986125e-05,
+      "loss": 3.9512,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.883311505847561e-05,
+      "loss": 3.9402,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882472911096509e-05,
+      "loss": 3.9395,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881634316345457e-05,
+      "loss": 3.9423,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880795721594405e-05,
+      "loss": 3.9424,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8799587647237257e-05,
+      "loss": 3.9358,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791201699726737e-05,
+      "loss": 3.9473,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8782815752216217e-05,
+      "loss": 3.9347,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.87744298047057e-05,
+      "loss": 3.9493,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.876606023599891e-05,
+      "loss": 3.9428,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.875767428848839e-05,
+      "loss": 3.931,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.023680210113525,
+      "eval_runtime": 285.1017,
+      "eval_samples_per_second": 1338.438,
+      "eval_steps_per_second": 41.827,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874928834097787e-05,
+      "loss": 3.9453,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874090239346735e-05,
+      "loss": 3.9377,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.873251644595683e-05,
+      "loss": 3.9476,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8724130498446306e-05,
+      "loss": 3.9424,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715744550935786e-05,
+      "loss": 3.9495,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707358603425266e-05,
+      "loss": 3.9372,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8698972655914746e-05,
+      "loss": 3.9303,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8690603087207955e-05,
+      "loss": 3.9237,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.868221713969744e-05,
+      "loss": 3.955,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.867383119218692e-05,
+      "loss": 3.944,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.86654452446764e-05,
+      "loss": 3.9377,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.865705929716588e-05,
+      "loss": 3.948,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8648706107262823e-05,
+      "loss": 3.9366,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640320159752303e-05,
+      "loss": 3.9365,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.863193421224178e-05,
+      "loss": 3.9333,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.862354826473126e-05,
+      "loss": 3.9342,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.861516231722074e-05,
+      "loss": 3.9361,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.860677636971022e-05,
+      "loss": 3.9322,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.85983904221997e-05,
+      "loss": 3.945,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.859000447468918e-05,
+      "loss": 3.9568,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8581634905982392e-05,
+      "loss": 3.9403,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8573248958471872e-05,
+      "loss": 3.9346,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564863010961352e-05,
+      "loss": 3.9387,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8556477063450832e-05,
+      "loss": 3.9425,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8548091115940312e-05,
+      "loss": 3.9363,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539705168429792e-05,
+      "loss": 3.9278,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531319220919272e-05,
+      "loss": 3.9301,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8522949652212484e-05,
+      "loss": 3.9453,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8514580083505693e-05,
+      "loss": 3.924,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8506194135995173e-05,
+      "loss": 3.9289,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497808188484653e-05,
+      "loss": 3.9405,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489422240974133e-05,
+      "loss": 3.9427,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8481036293463613e-05,
+      "loss": 3.9447,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472650345953093e-05,
+      "loss": 3.9397,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464280777246306e-05,
+      "loss": 3.9372,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455894829735786e-05,
+      "loss": 3.945,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447508882225266e-05,
+      "loss": 3.9245,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8439122934714746e-05,
+      "loss": 3.9385,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8430736987204226e-05,
+      "loss": 3.9258,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8422351039693706e-05,
+      "loss": 3.9291,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8413965092183185e-05,
+      "loss": 3.9341,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.840557914467267e-05,
+      "loss": 3.9404,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8397193197162142e-05,
+      "loss": 3.9355,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8388807249651622e-05,
+      "loss": 3.9294,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8380421302141102e-05,
+      "loss": 3.9359,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8372035354630582e-05,
+      "loss": 3.9373,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8363649407120062e-05,
+      "loss": 3.9406,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8355279838413274e-05,
+      "loss": 3.9278,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8346893890902754e-05,
+      "loss": 3.9239,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8338524322195963e-05,
+      "loss": 3.9397,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8330138374685443e-05,
+      "loss": 3.9454,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8321752427174923e-05,
+      "loss": 3.9312,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8313366479664407e-05,
+      "loss": 3.9328,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304980532153887e-05,
+      "loss": 3.9326,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296610963447096e-05,
+      "loss": 3.9262,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8288225015936576e-05,
+      "loss": 3.9337,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8279839068426056e-05,
+      "loss": 3.923,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271453120915536e-05,
+      "loss": 3.9387,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8263067173405015e-05,
+      "loss": 3.9365,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8254681225894495e-05,
+      "loss": 3.9365,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.824629527838398e-05,
+      "loss": 3.9201,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.823790933087346e-05,
+      "loss": 3.9413,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8229539762166668e-05,
+      "loss": 3.9174,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8221153814656148e-05,
+      "loss": 3.9337,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212767867145628e-05,
+      "loss": 3.9391,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8204381919635108e-05,
+      "loss": 3.9376,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8195995972124588e-05,
+      "loss": 3.9265,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.818761002461407e-05,
+      "loss": 3.9241,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817922407710355e-05,
+      "loss": 3.9235,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817083812959303e-05,
+      "loss": 3.9334,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.816246856088624e-05,
+      "loss": 3.9383,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.815408261337572e-05,
+      "loss": 3.9343,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8145713044668932e-05,
+      "loss": 3.9383,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8137327097158412e-05,
+      "loss": 3.9418,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128941149647892e-05,
+      "loss": 3.9379,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.81205715809411e-05,
+      "loss": 3.9247,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.811218563343058e-05,
+      "loss": 3.9385,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.810379968592006e-05,
+      "loss": 3.9394,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.809541373840954e-05,
+      "loss": 3.9224,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8087027790899025e-05,
+      "loss": 3.9413,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8078641843388498e-05,
+      "loss": 3.927,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070255895877978e-05,
+      "loss": 3.9497,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8061869948367458e-05,
+      "loss": 3.9366,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8053484000856938e-05,
+      "loss": 3.9178,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.804511443215015e-05,
+      "loss": 3.9369,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.803672848463963e-05,
+      "loss": 3.9273,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8028358915932846e-05,
+      "loss": 3.9418,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8019972968422326e-05,
+      "loss": 3.935,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.80115870209118e-05,
+      "loss": 3.9338,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.800320107340128e-05,
+      "loss": 3.9336,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7994815125890762e-05,
+      "loss": 3.9257,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7986429178380242e-05,
+      "loss": 3.9208,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7978043230869722e-05,
+      "loss": 3.9204,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7969657283359202e-05,
+      "loss": 3.9309,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7961271335848682e-05,
+      "loss": 3.9409,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7952885388338162e-05,
+      "loss": 3.9314,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.794451581963137e-05,
+      "loss": 3.9247,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7936129872120855e-05,
+      "loss": 3.9275,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7927743924610335e-05,
+      "loss": 3.9345,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919357977099815e-05,
+      "loss": 3.9387,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7910972029589295e-05,
+      "loss": 3.9306,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7902602460882504e-05,
+      "loss": 3.9285,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7894216513371984e-05,
+      "loss": 3.931,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885830565861464e-05,
+      "loss": 3.937,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877444618350947e-05,
+      "loss": 3.9183,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7869058670840427e-05,
+      "loss": 3.9289,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7860672723329907e-05,
+      "loss": 3.935,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7852286775819387e-05,
+      "loss": 3.9283,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7843917207112596e-05,
+      "loss": 3.9239,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7835531259602076e-05,
+      "loss": 3.9292,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7827145312091556e-05,
+      "loss": 3.9293,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7818759364581036e-05,
+      "loss": 3.9278,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810389795874248e-05,
+      "loss": 3.9305,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7802003848363728e-05,
+      "loss": 3.9246,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7793617900853208e-05,
+      "loss": 3.94,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7785231953342688e-05,
+      "loss": 3.9414,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7776846005832165e-05,
+      "loss": 3.9346,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768460058321645e-05,
+      "loss": 3.9354,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.776009048961486e-05,
+      "loss": 3.9331,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7751704542104334e-05,
+      "loss": 3.9317,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7743318594593814e-05,
+      "loss": 3.9242,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734932647083294e-05,
+      "loss": 3.9243,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726546699572774e-05,
+      "loss": 3.9321,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7718160752062257e-05,
+      "loss": 3.9337,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7709774804551737e-05,
+      "loss": 3.9314,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7701388857041217e-05,
+      "loss": 3.931,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7693002909530697e-05,
+      "loss": 3.9276,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7684633340823906e-05,
+      "loss": 3.9318,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7676247393313386e-05,
+      "loss": 3.9307,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667861445802866e-05,
+      "loss": 3.9387,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765947549829235e-05,
+      "loss": 3.9261,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765108955078183e-05,
+      "loss": 3.9163,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7642719982075038e-05,
+      "loss": 3.9213,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634334034564518e-05,
+      "loss": 3.9386,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7625948087053998e-05,
+      "loss": 3.9407,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7617562139543478e-05,
+      "loss": 3.932,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7609176192032958e-05,
+      "loss": 3.9286,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.760079024452244e-05,
+      "loss": 3.9397,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.759240429701192e-05,
+      "loss": 3.9381,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.758403472830513e-05,
+      "loss": 3.9328,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.757564878079461e-05,
+      "loss": 3.9289,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.756726283328409e-05,
+      "loss": 3.93,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755887688577357e-05,
+      "loss": 3.9312,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7550507317066783e-05,
+      "loss": 3.9242,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7542121369556263e-05,
+      "loss": 3.9408,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7533735422045743e-05,
+      "loss": 3.9257,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7525365853338952e-05,
+      "loss": 3.9375,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.751697990582843e-05,
+      "loss": 3.9328,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.750859395831791e-05,
+      "loss": 3.9215,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.01902961730957,
+      "eval_runtime": 287.6478,
+      "eval_samples_per_second": 1326.591,
+      "eval_steps_per_second": 41.457,
+      "step": 1373760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7500208010807395e-05,
+      "loss": 3.9358,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491822063296875e-05,
+      "loss": 3.9289,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483436115786348e-05,
+      "loss": 3.9349,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7475050168275828e-05,
+      "loss": 3.9318,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466664220765308e-05,
+      "loss": 3.9401,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458278273254788e-05,
+      "loss": 3.9283,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744989232574427e-05,
+      "loss": 3.9218,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744150637823375e-05,
+      "loss": 3.9141,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743312043072323e-05,
+      "loss": 3.945,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.742473448321271e-05,
+      "loss": 3.9334,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741634853570219e-05,
+      "loss": 3.9255,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.740796258819167e-05,
+      "loss": 3.9426,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739959301948488e-05,
+      "loss": 3.9251,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7391207071974364e-05,
+      "loss": 3.9311,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7382821124463843e-05,
+      "loss": 3.9201,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7374435176953323e-05,
+      "loss": 3.9235,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7366049229442803e-05,
+      "loss": 3.9324,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7357663281932283e-05,
+      "loss": 3.9189,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7349277334421763e-05,
+      "loss": 3.9356,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734089138691124e-05,
+      "loss": 3.9451,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332521818204452e-05,
+      "loss": 3.9333,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7324135870693936e-05,
+      "loss": 3.9275,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.731574992318341e-05,
+      "loss": 3.9308,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.730736397567289e-05,
+      "loss": 3.9333,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.729897802816237e-05,
+      "loss": 3.9246,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.729059208065185e-05,
+      "loss": 3.9223,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.728220613314133e-05,
+      "loss": 3.9164,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7273820185630812e-05,
+      "loss": 3.9374,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.726545061692402e-05,
+      "loss": 3.9121,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.72570646694135e-05,
+      "loss": 3.9239,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724869510070671e-05,
+      "loss": 3.9303,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.724030915319619e-05,
+      "loss": 3.9342,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7231939584489406e-05,
+      "loss": 3.9318,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223553636978883e-05,
+      "loss": 3.9314,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7215167689468363e-05,
+      "loss": 3.9276,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7206781741957842e-05,
+      "loss": 3.9373,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7198395794447322e-05,
+      "loss": 3.9199,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7190009846936802e-05,
+      "loss": 3.9285,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7181623899426282e-05,
+      "loss": 3.9132,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7173237951915766e-05,
+      "loss": 3.923,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7164852004405246e-05,
+      "loss": 3.9207,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7156482435698455e-05,
+      "loss": 3.9352,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7148096488187935e-05,
+      "loss": 3.9246,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139710540677415e-05,
+      "loss": 3.9209,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131324593166895e-05,
+      "loss": 3.9279,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122938645656375e-05,
+      "loss": 3.9284,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7114552698145858e-05,
+      "loss": 3.9308,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7106166750635338e-05,
+      "loss": 3.9171,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7097780803124818e-05,
+      "loss": 3.9112,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089411234418027e-05,
+      "loss": 3.935,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7081025286907507e-05,
+      "loss": 3.9337,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7072639339396987e-05,
+      "loss": 3.9256,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7064253391886467e-05,
+      "loss": 3.9251,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.705586744437595e-05,
+      "loss": 3.9196,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.704749787566916e-05,
+      "loss": 3.92,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703912830696237e-05,
+      "loss": 3.9257,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7030742359451848e-05,
+      "loss": 3.9145,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7022356411941328e-05,
+      "loss": 3.9289,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.701397046443081e-05,
+      "loss": 3.9261,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.700558451692029e-05,
+      "loss": 3.9223,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.69972149482135e-05,
+      "loss": 3.9127,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698884537950671e-05,
+      "loss": 3.9331,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698045943199619e-05,
+      "loss": 3.9118,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6972073484485673e-05,
+      "loss": 3.9241,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6963687536975153e-05,
+      "loss": 3.9269,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6955301589464633e-05,
+      "loss": 3.9305,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6946915641954113e-05,
+      "loss": 3.9153,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6938529694443593e-05,
+      "loss": 3.918,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6930143746933066e-05,
+      "loss": 3.9084,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921774178226282e-05,
+      "loss": 3.9258,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913388230715765e-05,
+      "loss": 3.9307,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6905002283205245e-05,
+      "loss": 3.9263,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.689661633569472e-05,
+      "loss": 3.9279,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.68882303881842e-05,
+      "loss": 3.9306,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.687984444067368e-05,
+      "loss": 3.9353,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871474871966894e-05,
+      "loss": 3.9144,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.686308892445637e-05,
+      "loss": 3.9282,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6854719355749587e-05,
+      "loss": 3.9336,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6846333408239067e-05,
+      "loss": 3.9145,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.683794746072854e-05,
+      "loss": 3.9301,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682956151321802e-05,
+      "loss": 3.9167,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6821175565707503e-05,
+      "loss": 3.9414,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6812789618196983e-05,
+      "loss": 3.9288,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804420049490192e-05,
+      "loss": 3.9082,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6796034101979672e-05,
+      "loss": 3.9224,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6787648154469152e-05,
+      "loss": 3.9216,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6779262206958632e-05,
+      "loss": 3.9301,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770876259448112e-05,
+      "loss": 3.9278,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6762490311937595e-05,
+      "loss": 3.9241,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6754120743230804e-05,
+      "loss": 3.9235,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745734795720284e-05,
+      "loss": 3.9189,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737348848209764e-05,
+      "loss": 3.9147,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728962900699244e-05,
+      "loss": 3.907,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720576953188724e-05,
+      "loss": 3.9234,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6712191005678204e-05,
+      "loss": 3.932,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703805058167684e-05,
+      "loss": 3.9237,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695419110657167e-05,
+      "loss": 3.9156,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6687033163146647e-05,
+      "loss": 3.9189,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6678647215636127e-05,
+      "loss": 3.9223,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670261268125607e-05,
+      "loss": 3.9327,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.666187532061508e-05,
+      "loss": 3.923,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653505751908296e-05,
+      "loss": 3.9256,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6645119804397776e-05,
+      "loss": 3.9181,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.663673385688726e-05,
+      "loss": 3.9281,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.662836428818047e-05,
+      "loss": 3.9076,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661997834066995e-05,
+      "loss": 3.924,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.661159239315943e-05,
+      "loss": 3.921,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6603206445648905e-05,
+      "loss": 3.9228,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6594820498138385e-05,
+      "loss": 3.9144,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6586434550627865e-05,
+      "loss": 3.9217,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6578048603117345e-05,
+      "loss": 3.9179,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6569662655606825e-05,
+      "loss": 3.9176,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6561293086900034e-05,
+      "loss": 3.9225,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6552907139389514e-05,
+      "loss": 3.9201,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6544521191878997e-05,
+      "loss": 3.9286,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6536135244368477e-05,
+      "loss": 3.9315,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6527765675661687e-05,
+      "loss": 3.9283,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519379728151166e-05,
+      "loss": 3.9282,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6510993780640646e-05,
+      "loss": 3.923,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6502607833130126e-05,
+      "loss": 3.9168,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.649423826442334e-05,
+      "loss": 3.9177,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.648585231691282e-05,
+      "loss": 3.9195,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64774663694023e-05,
+      "loss": 3.918,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.646908042189178e-05,
+      "loss": 3.9266,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460710853184988e-05,
+      "loss": 3.9206,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452324905674468e-05,
+      "loss": 3.9254,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.644393895816395e-05,
+      "loss": 3.9205,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643555301065343e-05,
+      "loss": 3.9206,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.642716706314291e-05,
+      "loss": 3.9195,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.641879749443612e-05,
+      "loss": 3.9346,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64104115469256e-05,
+      "loss": 3.9141,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.640202559941508e-05,
+      "loss": 3.9143,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639363965190456e-05,
+      "loss": 3.9104,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385270083197772e-05,
+      "loss": 3.925,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6376884135687252e-05,
+      "loss": 3.9321,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.636851456698046e-05,
+      "loss": 3.9265,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.636012861946994e-05,
+      "loss": 3.9193,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.635174267195942e-05,
+      "loss": 3.933,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343356724448905e-05,
+      "loss": 3.9255,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334970776938385e-05,
+      "loss": 3.9269,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6326584829427865e-05,
+      "loss": 3.92,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6318198881917345e-05,
+      "loss": 3.9244,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309812934406825e-05,
+      "loss": 3.9233,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301443365700034e-05,
+      "loss": 3.9135,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6293057418189514e-05,
+      "loss": 3.9308,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6284687849482726e-05,
+      "loss": 3.9221,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276301901972206e-05,
+      "loss": 3.9256,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6267915954461686e-05,
+      "loss": 3.9235,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259530006951166e-05,
+      "loss": 3.9109,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.015507221221924,
+      "eval_runtime": 293.8182,
+      "eval_samples_per_second": 1298.732,
+      "eval_steps_per_second": 40.586,
+      "step": 1450080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6251144059440646e-05,
+      "loss": 3.9259,
+      "step": 1450496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6242758111930126e-05,
+      "loss": 3.9232,
+      "step": 1451008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6234372164419606e-05,
+      "loss": 3.926,
+      "step": 1451520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6225986216909086e-05,
+      "loss": 3.9224,
+      "step": 1452032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6217600269398562e-05,
+      "loss": 3.9352,
+      "step": 1452544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6209214321888042e-05,
+      "loss": 3.9213,
+      "step": 1453056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6200828374377522e-05,
+      "loss": 3.9133,
+      "step": 1453568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6192442426867002e-05,
+      "loss": 3.9088,
+      "step": 1454080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6184056479356482e-05,
+      "loss": 3.9318,
+      "step": 1454592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6175670531845962e-05,
+      "loss": 3.9283,
+      "step": 1455104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6167284584335446e-05,
+      "loss": 3.9122,
+      "step": 1455616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6158898636824925e-05,
+      "loss": 3.9319,
+      "step": 1456128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6150529068118135e-05,
+      "loss": 3.9191,
+      "step": 1456640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6142143120607615e-05,
+      "loss": 3.9211,
+      "step": 1457152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6133757173097094e-05,
+      "loss": 3.9178,
+      "step": 1457664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6125371225586574e-05,
+      "loss": 3.9142,
+      "step": 1458176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6116985278076054e-05,
+      "loss": 3.9194,
+      "step": 1458688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6108599330565538e-05,
+      "loss": 3.9119,
+      "step": 1459200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6100213383055018e-05,
+      "loss": 3.927,
+      "step": 1459712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6091827435544498e-05,
+      "loss": 3.9331,
+      "step": 1460224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6083457866837707e-05,
+      "loss": 3.929,
+      "step": 1460736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6075071919327187e-05,
+      "loss": 3.9207,
+      "step": 1461248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6066685971816667e-05,
+      "loss": 3.9191,
+      "step": 1461760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6058300024306147e-05,
+      "loss": 3.9302,
+      "step": 1462272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.604991407679563e-05,
+      "loss": 3.9141,
+      "step": 1462784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.604154450808884e-05,
+      "loss": 3.9151,
+      "step": 1463296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.603315856057832e-05,
+      "loss": 3.9086,
+      "step": 1463808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.60247726130678e-05,
+      "loss": 3.9264,
+      "step": 1464320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.601638666555728e-05,
+      "loss": 3.9065,
+      "step": 1464832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6008000718046755e-05,
+      "loss": 3.9153,
+      "step": 1465344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5999614770536235e-05,
+      "loss": 3.9231,
+      "step": 1465856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5991228823025715e-05,
+      "loss": 3.9227,
+      "step": 1466368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5982842875515195e-05,
+      "loss": 3.9238,
+      "step": 1466880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5974473306808404e-05,
+      "loss": 3.9256,
+      "step": 1467392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.596610373810162e-05,
+      "loss": 3.9195,
+      "step": 1467904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.59577177905911e-05,
+      "loss": 3.9287,
+      "step": 1468416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5949331843080577e-05,
+      "loss": 3.9153,
+      "step": 1468928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5940945895570057e-05,
+      "loss": 3.9113,
+      "step": 1469440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5932576326863273e-05,
+      "loss": 3.9083,
+      "step": 1469952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5924190379352753e-05,
+      "loss": 3.9146,
+      "step": 1470464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.591580443184223e-05,
+      "loss": 3.9122,
+      "step": 1470976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.590741848433171e-05,
+      "loss": 3.9286,
+      "step": 1471488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.589903253682119e-05,
+      "loss": 3.9147,
+      "step": 1472000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.589064658931067e-05,
+      "loss": 3.914,
+      "step": 1472512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.588226064180015e-05,
+      "loss": 3.924,
+      "step": 1473024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.587387469428963e-05,
+      "loss": 3.9135,
+      "step": 1473536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5865505125582838e-05,
+      "loss": 3.9279,
+      "step": 1474048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.585713555687605e-05,
+      "loss": 3.911,
+      "step": 1474560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.584874960936553e-05,
+      "loss": 3.8974,
+      "step": 1475072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.584036366185501e-05,
+      "loss": 3.9326,
+      "step": 1475584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.583197771434449e-05,
+      "loss": 3.9236,
+      "step": 1476096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.582359176683397e-05,
+      "loss": 3.9218,
+      "step": 1476608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.581520581932345e-05,
+      "loss": 3.9137,
+      "step": 1477120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.580681987181293e-05,
+      "loss": 3.9121,
+      "step": 1477632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5798433924302414e-05,
+      "loss": 3.9114,
+      "step": 1478144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5790064355595623e-05,
+      "loss": 3.9162,
+      "step": 1478656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5781678408085103e-05,
+      "loss": 3.9076,
+      "step": 1479168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5773292460574583e-05,
+      "loss": 3.9154,
+      "step": 1479680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5764906513064063e-05,
+      "loss": 3.9244,
+      "step": 1480192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5756536944357275e-05,
+      "loss": 3.9128,
+      "step": 1480704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5748167375650484e-05,
+      "loss": 3.9051,
+      "step": 1481216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5739781428139964e-05,
+      "loss": 3.9272,
+      "step": 1481728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5731395480629444e-05,
+      "loss": 3.904,
+      "step": 1482240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5723009533118924e-05,
+      "loss": 3.9097,
+      "step": 1482752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5714623585608404e-05,
+      "loss": 3.9213,
+      "step": 1483264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5706237638097884e-05,
+      "loss": 3.9217,
+      "step": 1483776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5697851690587367e-05,
+      "loss": 3.9044,
+      "step": 1484288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5689465743076847e-05,
+      "loss": 3.9113,
+      "step": 1484800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5681079795566327e-05,
+      "loss": 3.8968,
+      "step": 1485312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5672710226859536e-05,
+      "loss": 3.9205,
+      "step": 1485824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5664324279349016e-05,
+      "loss": 3.9211,
+      "step": 1486336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5655938331838496e-05,
+      "loss": 3.9183,
+      "step": 1486848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5647552384327976e-05,
+      "loss": 3.9164,
+      "step": 1487360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5639166436817456e-05,
+      "loss": 3.9237,
+      "step": 1487872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.563079686811067e-05,
+      "loss": 3.9311,
+      "step": 1488384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5622427299403878e-05,
+      "loss": 3.9065,
+      "step": 1488896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5614041351893358e-05,
+      "loss": 3.9196,
+      "step": 1489408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5605655404382838e-05,
+      "loss": 3.9245,
+      "step": 1489920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5597269456872318e-05,
+      "loss": 3.9042,
+      "step": 1490432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.55888835093618e-05,
+      "loss": 3.9235,
+      "step": 1490944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.558049756185128e-05,
+      "loss": 3.912,
+      "step": 1491456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.557211161434076e-05,
+      "loss": 3.9288,
+      "step": 1491968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5563725666830234e-05,
+      "loss": 3.9217,
+      "step": 1492480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5555339719319714e-05,
+      "loss": 3.9037,
+      "step": 1492992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5546953771809194e-05,
+      "loss": 3.9157,
+      "step": 1493504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5538567824298677e-05,
+      "loss": 3.9106,
+      "step": 1494016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5530181876788157e-05,
+      "loss": 3.9248,
+      "step": 1494528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5521812308081366e-05,
+      "loss": 3.916,
+      "step": 1495040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5513442739374582e-05,
+      "loss": 3.9137,
+      "step": 1495552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5505056791864055e-05,
+      "loss": 3.9155,
+      "step": 1496064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.549667084435354e-05,
+      "loss": 3.9153,
+      "step": 1496576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.548828489684302e-05,
+      "loss": 3.9095,
+      "step": 1497088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.54798989493325e-05,
+      "loss": 3.8992,
+      "step": 1497600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.547151300182198e-05,
+      "loss": 3.9149,
+      "step": 1498112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5463143433115188e-05,
+      "loss": 3.9253,
+      "step": 1498624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5454757485604668e-05,
+      "loss": 3.92,
+      "step": 1499136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5446371538094148e-05,
+      "loss": 3.9016,
+      "step": 1499648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.543798559058363e-05,
+      "loss": 3.9131,
+      "step": 1500160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.542959964307311e-05,
+      "loss": 3.9123,
+      "step": 1500672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.542121369556259e-05,
+      "loss": 3.9222,
+      "step": 1501184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.541282774805207e-05,
+      "loss": 3.9135,
+      "step": 1501696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.540444180054155e-05,
+      "loss": 3.9164,
+      "step": 1502208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.539607223183476e-05,
+      "loss": 3.9089,
+      "step": 1502720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.538768628432424e-05,
+      "loss": 3.9222,
+      "step": 1503232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5379300336813723e-05,
+      "loss": 3.9,
+      "step": 1503744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5370914389303203e-05,
+      "loss": 3.9159,
+      "step": 1504256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5362544820596412e-05,
+      "loss": 3.912,
+      "step": 1504768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5354158873085892e-05,
+      "loss": 3.916,
+      "step": 1505280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5345772925575372e-05,
+      "loss": 3.9061,
+      "step": 1505792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5337386978064852e-05,
+      "loss": 3.9168,
+      "step": 1506304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5329001030554332e-05,
+      "loss": 3.9064,
+      "step": 1506816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5320631461847545e-05,
+      "loss": 3.9109,
+      "step": 1507328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5312245514337024e-05,
+      "loss": 3.9141,
+      "step": 1507840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5303859566826504e-05,
+      "loss": 3.9105,
+      "step": 1508352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5295473619315984e-05,
+      "loss": 3.9223,
+      "step": 1508864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5287087671805464e-05,
+      "loss": 3.9234,
+      "step": 1509376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5278718103098677e-05,
+      "loss": 3.9187,
+      "step": 1509888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5270332155588157e-05,
+      "loss": 3.9212,
+      "step": 1510400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5261962586881366e-05,
+      "loss": 3.916,
+      "step": 1510912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5253576639370846e-05,
+      "loss": 3.9083,
+      "step": 1511424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5245190691860326e-05,
+      "loss": 3.9087,
+      "step": 1511936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5236804744349806e-05,
+      "loss": 3.911,
+      "step": 1512448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5228418796839286e-05,
+      "loss": 3.9137,
+      "step": 1512960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.522003284932877e-05,
+      "loss": 3.9208,
+      "step": 1513472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.521164690181825e-05,
+      "loss": 3.9108,
+      "step": 1513984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5203260954307722e-05,
+      "loss": 3.9181,
+      "step": 1514496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5194891385600938e-05,
+      "loss": 3.9095,
+      "step": 1515008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5186521816894147e-05,
+      "loss": 3.9105,
+      "step": 1515520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.517813586938363e-05,
+      "loss": 3.9123,
+      "step": 1516032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.516974992187311e-05,
+      "loss": 3.9252,
+      "step": 1516544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.516136397436259e-05,
+      "loss": 3.91,
+      "step": 1517056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.515297802685207e-05,
+      "loss": 3.9053,
+      "step": 1517568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5144592079341544e-05,
+      "loss": 3.8997,
+      "step": 1518080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5136206131831023e-05,
+      "loss": 3.9154,
+      "step": 1518592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5127820184320507e-05,
+      "loss": 3.923,
+      "step": 1519104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5119434236809987e-05,
+      "loss": 3.9202,
+      "step": 1519616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5111048289299467e-05,
+      "loss": 3.9148,
+      "step": 1520128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5102662341788947e-05,
+      "loss": 3.9247,
+      "step": 1520640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5094276394278427e-05,
+      "loss": 3.9169,
+      "step": 1521152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5085906825571636e-05,
+      "loss": 3.9169,
+      "step": 1521664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5077520878061116e-05,
+      "loss": 3.9141,
+      "step": 1522176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.50691349305506e-05,
+      "loss": 3.9146,
+      "step": 1522688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.506074898304008e-05,
+      "loss": 3.9171,
+      "step": 1523200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.505236303552956e-05,
+      "loss": 3.9032,
+      "step": 1523712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5043993466822768e-05,
+      "loss": 3.9216,
+      "step": 1524224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5035623898115977e-05,
+      "loss": 3.9182,
+      "step": 1524736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.502723795060546e-05,
+      "loss": 3.9132,
+      "step": 1525248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.501885200309494e-05,
+      "loss": 3.9135,
+      "step": 1525760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.501046605558442e-05,
+      "loss": 3.9095,
+      "step": 1526272
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.012427806854248,
+      "eval_runtime": 297.812,
+      "eval_samples_per_second": 1281.315,
+      "eval_steps_per_second": 40.042,
+      "step": 1526400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.50020801080739e-05,
+      "loss": 3.9158,
+      "step": 1526784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.499369416056338e-05,
+      "loss": 3.9172,
+      "step": 1527296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.498530821305286e-05,
+      "loss": 3.9196,
+      "step": 1527808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.497692226554234e-05,
+      "loss": 3.9128,
+      "step": 1528320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4968536318031817e-05,
+      "loss": 3.9259,
+      "step": 1528832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.49601503705213e-05,
+      "loss": 3.915,
+      "step": 1529344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4951780801814513e-05,
+      "loss": 3.9029,
+      "step": 1529856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4943411233107722e-05,
+      "loss": 3.9036,
+      "step": 1530368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.49350252855972e-05,
+      "loss": 3.9193,
+      "step": 1530880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.492663933808668e-05,
+      "loss": 3.9203,
+      "step": 1531392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.491825339057616e-05,
+      "loss": 3.9092,
+      "step": 1531904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.490986744306564e-05,
+      "loss": 3.9236,
+      "step": 1532416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4901497874358854e-05,
+      "loss": 3.913,
+      "step": 1532928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4893111926848334e-05,
+      "loss": 3.9135,
+      "step": 1533440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4884725979337814e-05,
+      "loss": 3.9085,
+      "step": 1533952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.487634003182729e-05,
+      "loss": 3.9068,
+      "step": 1534464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.486795408431677e-05,
+      "loss": 3.9095,
+      "step": 1534976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4859568136806254e-05,
+      "loss": 3.9061,
+      "step": 1535488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4851198568099463e-05,
+      "loss": 3.9197,
+      "step": 1536000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4842812620588943e-05,
+      "loss": 3.9229,
+      "step": 1536512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4834426673078423e-05,
+      "loss": 3.9205,
+      "step": 1537024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4826040725567903e-05,
+      "loss": 3.9152,
+      "step": 1537536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4817654778057383e-05,
+      "loss": 3.9092,
+      "step": 1538048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4809285209350595e-05,
+      "loss": 3.9188,
+      "step": 1538560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4800899261840075e-05,
+      "loss": 3.9095,
+      "step": 1539072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4792513314329555e-05,
+      "loss": 3.9079,
+      "step": 1539584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4784127366819035e-05,
+      "loss": 3.9007,
+      "step": 1540096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4775741419308515e-05,
+      "loss": 3.9145,
+      "step": 1540608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.4767355471797995e-05,
+      "loss": 3.9033,
+      "step": 1541120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.475896952428747e-05,
+      "loss": 3.9029,
+      "step": 1541632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4750583576776955e-05,
+      "loss": 3.9183,
+      "step": 1542144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4742214008070167e-05,
+      "loss": 3.9129,
+      "step": 1542656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4733828060559644e-05,
+      "loss": 3.9193,
+      "step": 1543168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4725442113049124e-05,
+      "loss": 3.9179,
+      "step": 1543680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.471708892314607e-05,
+      "loss": 3.909,
+      "step": 1544192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.470870297563555e-05,
+      "loss": 3.9209,
+      "step": 1544704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.470031702812503e-05,
+      "loss": 3.9107,
+      "step": 1545216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.469193108061451e-05,
+      "loss": 3.9023,
+      "step": 1545728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.468354513310399e-05,
+      "loss": 3.8983,
+      "step": 1546240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.467515918559347e-05,
+      "loss": 3.9056,
+      "step": 1546752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4666773238082945e-05,
+      "loss": 3.9056,
+      "step": 1547264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4658387290572425e-05,
+      "loss": 3.9205,
+      "step": 1547776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.465000134306191e-05,
+      "loss": 3.9085,
+      "step": 1548288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.464161539555139e-05,
+      "loss": 3.9055,
+      "step": 1548800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.463322944804087e-05,
+      "loss": 3.917,
+      "step": 1549312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.462484350053035e-05,
+      "loss": 3.9096,
+      "step": 1549824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4616457553019825e-05,
+      "loss": 3.9166,
+      "step": 1550336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4608087984313037e-05,
+      "loss": 3.9014,
+      "step": 1550848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4599702036802517e-05,
+      "loss": 3.8865,
+      "step": 1551360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.459133246809573e-05,
+      "loss": 3.9224,
+      "step": 1551872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.458294652058521e-05,
+      "loss": 3.9202,
+      "step": 1552384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.457456057307469e-05,
+      "loss": 3.9116,
+      "step": 1552896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.456617462556417e-05,
+      "loss": 3.9062,
+      "step": 1553408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4557788678053646e-05,
+      "loss": 3.9047,
+      "step": 1553920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4549402730543126e-05,
+      "loss": 3.9017,
+      "step": 1554432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4541033161836342e-05,
+      "loss": 3.9075,
+      "step": 1554944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4532647214325822e-05,
+      "loss": 3.9035,
+      "step": 1555456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.45242612668153e-05,
+      "loss": 3.906,
+      "step": 1555968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.451587531930478e-05,
+      "loss": 3.9148,
+      "step": 1556480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.450748937179426e-05,
+      "loss": 3.9094,
+      "step": 1556992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.449910342428374e-05,
+      "loss": 3.8978,
+      "step": 1557504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.449073385557695e-05,
+      "loss": 3.9189,
+      "step": 1558016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.448234790806643e-05,
+      "loss": 3.8957,
+      "step": 1558528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.447396196055591e-05,
+      "loss": 3.8984,
+      "step": 1559040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.446557601304539e-05,
+      "loss": 3.9155,
+      "step": 1559552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.445719006553487e-05,
+      "loss": 3.9171,
+      "step": 1560064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.444880411802435e-05,
+      "loss": 3.8962,
+      "step": 1560576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4440434549317563e-05,
+      "loss": 3.9043,
+      "step": 1561088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4432048601807043e-05,
+      "loss": 3.8876,
+      "step": 1561600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4423662654296523e-05,
+      "loss": 3.9123,
+      "step": 1562112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4415293085589732e-05,
+      "loss": 3.9144,
+      "step": 1562624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4406907138079212e-05,
+      "loss": 3.9063,
+      "step": 1563136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4398521190568692e-05,
+      "loss": 3.9104,
+      "step": 1563648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4390135243058172e-05,
+      "loss": 3.9129,
+      "step": 1564160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4381749295547656e-05,
+      "loss": 3.9263,
+      "step": 1564672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4373363348037132e-05,
+      "loss": 3.8963,
+      "step": 1565184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4364977400526612e-05,
+      "loss": 3.9147,
+      "step": 1565696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4356591453016092e-05,
+      "loss": 3.9142,
+      "step": 1566208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4348205505505572e-05,
+      "loss": 3.8933,
+      "step": 1566720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.433983593679878e-05,
+      "loss": 3.9179,
+      "step": 1567232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4331449989288264e-05,
+      "loss": 3.9048,
+      "step": 1567744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4323064041777744e-05,
+      "loss": 3.9179,
+      "step": 1568256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4314678094267224e-05,
+      "loss": 3.9139,
+      "step": 1568768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4306292146756704e-05,
+      "loss": 3.897,
+      "step": 1569280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4297922578049913e-05,
+      "loss": 3.9033,
+      "step": 1569792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4289536630539393e-05,
+      "loss": 3.9041,
+      "step": 1570304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4281150683028873e-05,
+      "loss": 3.9202,
+      "step": 1570816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4272764735518357e-05,
+      "loss": 3.9063,
+      "step": 1571328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4264395166811566e-05,
+      "loss": 3.9065,
+      "step": 1571840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4256009219301046e-05,
+      "loss": 3.9052,
+      "step": 1572352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4247623271790526e-05,
+      "loss": 3.9074,
+      "step": 1572864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4239237324280006e-05,
+      "loss": 3.9013,
+      "step": 1573376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4230851376769486e-05,
+      "loss": 3.8883,
+      "step": 1573888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4222465429258966e-05,
+      "loss": 3.9088,
+      "step": 1574400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4214079481748445e-05,
+      "loss": 3.9143,
+      "step": 1574912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4205709913041658e-05,
+      "loss": 3.9129,
+      "step": 1575424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4197323965531135e-05,
+      "loss": 3.8951,
+      "step": 1575936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4188938018020614e-05,
+      "loss": 3.9029,
+      "step": 1576448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4180552070510094e-05,
+      "loss": 3.9058,
+      "step": 1576960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4172166122999574e-05,
+      "loss": 3.9119,
+      "step": 1577472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4163780175489058e-05,
+      "loss": 3.9094,
+      "step": 1577984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4155410606782267e-05,
+      "loss": 3.9071,
+      "step": 1578496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4147024659271747e-05,
+      "loss": 3.9034,
+      "step": 1579008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4138638711761227e-05,
+      "loss": 3.9156,
+      "step": 1579520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4130252764250707e-05,
+      "loss": 3.8918,
+      "step": 1580032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4121866816740187e-05,
+      "loss": 3.9074,
+      "step": 1580544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4113480869229667e-05,
+      "loss": 3.8988,
+      "step": 1581056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4105094921719147e-05,
+      "loss": 3.9104,
+      "step": 1581568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4096708974208627e-05,
+      "loss": 3.8938,
+      "step": 1582080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.408833940550184e-05,
+      "loss": 3.9116,
+      "step": 1582592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4079953457991316e-05,
+      "loss": 3.9015,
+      "step": 1583104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4071567510480796e-05,
+      "loss": 3.9026,
+      "step": 1583616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.406318156297028e-05,
+      "loss": 3.9036,
+      "step": 1584128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.405479561545976e-05,
+      "loss": 3.9026,
+      "step": 1584640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4046426046752968e-05,
+      "loss": 3.9108,
+      "step": 1585152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4038040099242448e-05,
+      "loss": 3.9186,
+      "step": 1585664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.4029654151731928e-05,
+      "loss": 3.9102,
+      "step": 1586176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.402128458302514e-05,
+      "loss": 3.9113,
+      "step": 1586688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.401289863551462e-05,
+      "loss": 3.9086,
+      "step": 1587200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.40045126880041e-05,
+      "loss": 3.9024,
+      "step": 1587712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.399612674049358e-05,
+      "loss": 3.8962,
+      "step": 1588224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.398774079298306e-05,
+      "loss": 3.9096,
+      "step": 1588736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.397935484547254e-05,
+      "loss": 3.9027,
+      "step": 1589248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.397096889796202e-05,
+      "loss": 3.9112,
+      "step": 1589760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3962582950451497e-05,
+      "loss": 3.9033,
+      "step": 1590272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.395419700294098e-05,
+      "loss": 3.9082,
+      "step": 1590784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.394581105543046e-05,
+      "loss": 3.9032,
+      "step": 1591296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.393742510791994e-05,
+      "loss": 3.9063,
+      "step": 1591808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.392905553921315e-05,
+      "loss": 3.9014,
+      "step": 1592320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.392066959170263e-05,
+      "loss": 3.9156,
+      "step": 1592832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.391230002299584e-05,
+      "loss": 3.9058,
+      "step": 1593344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.390391407548532e-05,
+      "loss": 3.8979,
+      "step": 1593856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.38955281279748e-05,
+      "loss": 3.8926,
+      "step": 1594368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.388714218046428e-05,
+      "loss": 3.9088,
+      "step": 1594880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.387875623295376e-05,
+      "loss": 3.9144,
+      "step": 1595392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.387037028544324e-05,
+      "loss": 3.9128,
+      "step": 1595904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.386198433793272e-05,
+      "loss": 3.9085,
+      "step": 1596416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.38535983904222e-05,
+      "loss": 3.9141,
+      "step": 1596928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3845228821715414e-05,
+      "loss": 3.9091,
+      "step": 1597440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3836842874204894e-05,
+      "loss": 3.9135,
+      "step": 1597952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3828456926694374e-05,
+      "loss": 3.9029,
+      "step": 1598464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.382007097918385e-05,
+      "loss": 3.9087,
+      "step": 1598976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3811701410477063e-05,
+      "loss": 3.905,
+      "step": 1599488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3803315462966543e-05,
+      "loss": 3.9006,
+      "step": 1600000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3794929515456026e-05,
+      "loss": 3.9128,
+      "step": 1600512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3786543567945502e-05,
+      "loss": 3.9109,
+      "step": 1601024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3778173999238715e-05,
+      "loss": 3.9068,
+      "step": 1601536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3769788051728195e-05,
+      "loss": 3.9044,
+      "step": 1602048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.3761402104217675e-05,
+      "loss": 3.9049,
+      "step": 1602560
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.008484840393066,
+      "eval_runtime": 301.7379,
+      "eval_samples_per_second": 1264.644,
+      "eval_steps_per_second": 39.521,
+      "step": 1602720
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.355064878507324e+17,
+  "trial_name": null,
+  "trial_params": null
+}