diff --git "a/checkpoint-686880/trainer_state.json" "b/checkpoint-686880/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-686880/trainer_state.json"
@@ -0,0 +1,8143 @@
+{
+  "best_metric": 3.848841428756714,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/binding-case/transformer/0/checkpoints/checkpoint-686880",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 686880,
+  "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": 10.9819,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 6.8109,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 6.191,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 5.9683,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 5.803,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 5.6962,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 5.5926,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 5.52,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 5.4384,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 5.3919,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 5.3319,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 5.2903,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 5.2461,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 5.2002,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 5.1624,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 5.1206,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 5.0928,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.0621,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98490857024181e-05,
+      "loss": 5.035,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984069975490758e-05,
+      "loss": 5.0055,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 4.9923,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 4.9647,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 4.9352,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 4.9223,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 4.9113,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 4.8755,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 4.8571,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 4.846,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 4.8307,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 4.8077,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 4.7903,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 4.7884,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 4.7724,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97233456261715e-05,
+      "loss": 4.7664,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.971495967866098e-05,
+      "loss": 4.7516,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 4.7332,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 4.7271,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968981821493315e-05,
+      "loss": 4.6934,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968143226742263e-05,
+      "loss": 4.7001,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967304631991211e-05,
+      "loss": 4.6722,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966466037240159e-05,
+      "loss": 4.6729,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965627442489107e-05,
+      "loss": 4.6598,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 4.6628,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 4.6413,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 4.6367,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962274701365272e-05,
+      "loss": 4.63,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 4.6121,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 4.623,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 4.5882,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9589219602414374e-05,
+      "loss": 4.5808,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958083365490385e-05,
+      "loss": 4.5924,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 4.5887,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 4.565,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 4.5567,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954732262246923e-05,
+      "loss": 4.5574,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953893667495871e-05,
+      "loss": 4.5389,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953055072744819e-05,
+      "loss": 4.5611,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952216477993767e-05,
+      "loss": 4.5212,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951377883242715e-05,
+      "loss": 4.5329,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.950539288491663e-05,
+      "loss": 4.5215,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949700693740611e-05,
+      "loss": 4.5065,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948862098989559e-05,
+      "loss": 4.5152,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94802514211888e-05,
+      "loss": 4.4986,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947186547367828e-05,
+      "loss": 4.4946,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946347952616776e-05,
+      "loss": 4.4843,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 4.4986,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 4.4717,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943833806243993e-05,
+      "loss": 4.4756,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942995211492941e-05,
+      "loss": 4.4531,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942156616741889e-05,
+      "loss": 4.4628,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941318021990837e-05,
+      "loss": 4.4738,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404810651201585e-05,
+      "loss": 4.4646,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396424703691065e-05,
+      "loss": 4.447,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 4.4592,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 4.4597,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371283239963234e-05,
+      "loss": 4.434,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936291367125644e-05,
+      "loss": 4.4407,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935452772374592e-05,
+      "loss": 4.4226,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93461417762354e-05,
+      "loss": 4.4148,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933775582872488e-05,
+      "loss": 4.4117,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932936988121436e-05,
+      "loss": 4.4124,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932098393370384e-05,
+      "loss": 4.4179,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931259798619332e-05,
+      "loss": 4.4238,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930422841748654e-05,
+      "loss": 4.4144,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929584246997602e-05,
+      "loss": 4.3893,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92874565224655e-05,
+      "loss": 4.3948,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927907057495498e-05,
+      "loss": 4.3976,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927070100624819e-05,
+      "loss": 4.3892,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.926231505873767e-05,
+      "loss": 4.3798,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925394549003088e-05,
+      "loss": 4.3773,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924555954252036e-05,
+      "loss": 4.3723,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923717359500984e-05,
+      "loss": 4.3802,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9228804026303046e-05,
+      "loss": 4.3674,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922041807879253e-05,
+      "loss": 4.3732,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921203213128201e-05,
+      "loss": 4.3643,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920364618377149e-05,
+      "loss": 4.3622,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919526023626097e-05,
+      "loss": 4.3465,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918687428875045e-05,
+      "loss": 4.3496,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917848834123993e-05,
+      "loss": 4.3468,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917010239372941e-05,
+      "loss": 4.3516,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916171644621889e-05,
+      "loss": 4.3518,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915333049870837e-05,
+      "loss": 4.3292,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914494455119785e-05,
+      "loss": 4.336,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136558603687325e-05,
+      "loss": 4.3284,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912818903498054e-05,
+      "loss": 4.3328,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911980308747002e-05,
+      "loss": 4.3189,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911143351876323e-05,
+      "loss": 4.3165,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910304757125272e-05,
+      "loss": 4.3241,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90946616237422e-05,
+      "loss": 4.3138,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908627567623168e-05,
+      "loss": 4.3106,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907788972872115e-05,
+      "loss": 4.303,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906950378121063e-05,
+      "loss": 4.3008,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9061134212503846e-05,
+      "loss": 4.3136,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9052748264993326e-05,
+      "loss": 4.3015,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90443623174828e-05,
+      "loss": 4.3016,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903597636997228e-05,
+      "loss": 4.3044,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027606801265495e-05,
+      "loss": 4.3093,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019220853754975e-05,
+      "loss": 4.2939,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010834906244455e-05,
+      "loss": 4.2821,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002448958733935e-05,
+      "loss": 4.289,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994063011223415e-05,
+      "loss": 4.2938,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985693442516624e-05,
+      "loss": 4.2806,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977307495006104e-05,
+      "loss": 4.2796,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968921547495584e-05,
+      "loss": 4.277,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960535599985064e-05,
+      "loss": 4.2791,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952149652474544e-05,
+      "loss": 4.2731,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943763704964024e-05,
+      "loss": 4.2757,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935377757453504e-05,
+      "loss": 4.2673,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892700818874671e-05,
+      "loss": 4.2693,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891862224123619e-05,
+      "loss": 4.2617,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891023629372567e-05,
+      "loss": 4.2794,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890185034621515e-05,
+      "loss": 4.2627,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889346439870464e-05,
+      "loss": 4.2554,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888507845119412e-05,
+      "loss": 4.2516,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.2607,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.2549,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885993698746629e-05,
+      "loss": 4.2604,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515674187595e-05,
+      "loss": 4.247,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884318147124898e-05,
+      "loss": 4.2529,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883479552373846e-05,
+      "loss": 4.2473,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882640957622794e-05,
+      "loss": 4.2471,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881802362871742e-05,
+      "loss": 4.2375,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88096376812069e-05,
+      "loss": 4.2478,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880125173369638e-05,
+      "loss": 4.2354,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8792882164989586e-05,
+      "loss": 4.2258,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878449621747907e-05,
+      "loss": 4.2472,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877611026996855e-05,
+      "loss": 4.2313,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876772432245803e-05,
+      "loss": 4.2294,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875933837494751e-05,
+      "loss": 4.2391,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8750952427436986e-05,
+      "loss": 4.2202,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1929707527160645,
+      "eval_runtime": 303.6122,
+      "eval_samples_per_second": 1256.837,
+      "eval_steps_per_second": 39.277,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.21,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.2134,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.872581096370916e-05,
+      "loss": 4.2211,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717425016198635e-05,
+      "loss": 4.2135,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709039068688115e-05,
+      "loss": 4.2212,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.2058,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.869228355247081e-05,
+      "loss": 4.209,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.868389760496029e-05,
+      "loss": 4.1986,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675528036253507e-05,
+      "loss": 4.2129,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667142088742986e-05,
+      "loss": 4.1954,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865875614123246e-05,
+      "loss": 4.1943,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865037019372194e-05,
+      "loss": 4.1977,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8642000625015155e-05,
+      "loss": 4.1939,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8633614677504635e-05,
+      "loss": 4.1855,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862522872999411e-05,
+      "loss": 4.1868,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861684278248359e-05,
+      "loss": 4.1708,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.1818,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.1743,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.1688,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.1941,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574929423734724e-05,
+      "loss": 4.1726,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566543476224204e-05,
+      "loss": 4.1747,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558157528713684e-05,
+      "loss": 4.1715,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549771581203164e-05,
+      "loss": 4.1866,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854140201249637e-05,
+      "loss": 4.1592,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853301606498585e-05,
+      "loss": 4.1627,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852463011747533e-05,
+      "loss": 4.167,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851624416996481e-05,
+      "loss": 4.1656,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850785822245429e-05,
+      "loss": 4.1547,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84994886537475e-05,
+      "loss": 4.154,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849110270623698e-05,
+      "loss": 4.1615,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848271675872646e-05,
+      "loss": 4.1619,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847433081121595e-05,
+      "loss": 4.1629,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846594486370543e-05,
+      "loss": 4.1552,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845757529499864e-05,
+      "loss": 4.1487,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844918934748812e-05,
+      "loss": 4.1564,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84408033999776e-05,
+      "loss": 4.1319,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843241745246708e-05,
+      "loss": 4.1513,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842404788376029e-05,
+      "loss": 4.1307,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841566193624977e-05,
+      "loss": 4.1399,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840727598873925e-05,
+      "loss": 4.1361,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839889004122873e-05,
+      "loss": 4.1457,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8390520472521936e-05,
+      "loss": 4.1379,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8382134525011416e-05,
+      "loss": 4.1314,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83737485775009e-05,
+      "loss": 4.1365,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836536262999038e-05,
+      "loss": 4.1282,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835699306128359e-05,
+      "loss": 4.1406,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834860711377307e-05,
+      "loss": 4.1134,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834022116626255e-05,
+      "loss": 4.1139,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833183521875203e-05,
+      "loss": 4.131,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832346565004524e-05,
+      "loss": 4.1334,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.831507970253472e-05,
+      "loss": 4.1175,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83066937550242e-05,
+      "loss": 4.1132,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829830780751368e-05,
+      "loss": 4.121,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828993823880689e-05,
+      "loss": 4.1037,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828155229129637e-05,
+      "loss": 4.1295,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273166343785856e-05,
+      "loss": 4.1088,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264780396275336e-05,
+      "loss": 4.1174,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256410827568545e-05,
+      "loss": 4.1124,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8248024880058025e-05,
+      "loss": 4.0996,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239638932547505e-05,
+      "loss": 4.1148,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8231252985036985e-05,
+      "loss": 4.1013,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222867037526465e-05,
+      "loss": 4.0995,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214481090015945e-05,
+      "loss": 4.0965,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8206111521309154e-05,
+      "loss": 4.1139,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8197725573798634e-05,
+      "loss": 4.0967,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189339626288114e-05,
+      "loss": 4.0963,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180953678777594e-05,
+      "loss": 4.087,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817258411007081e-05,
+      "loss": 4.0938,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816419816256029e-05,
+      "loss": 4.1099,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815581221504977e-05,
+      "loss": 4.1053,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814742626753924e-05,
+      "loss": 4.0935,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813905669883246e-05,
+      "loss": 4.1029,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813067075132194e-05,
+      "loss": 4.1067,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812228480381141e-05,
+      "loss": 4.0916,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811389885630089e-05,
+      "loss": 4.0973,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810552928759411e-05,
+      "loss": 4.0873,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809714334008359e-05,
+      "loss": 4.0784,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808875739257306e-05,
+      "loss": 4.0801,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808037144506255e-05,
+      "loss": 4.0843,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807200187635576e-05,
+      "loss": 4.0874,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063615928845243e-05,
+      "loss": 4.1034,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055229981334717e-05,
+      "loss": 4.0908,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046844033824197e-05,
+      "loss": 4.0694,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803847446511741e-05,
+      "loss": 4.0781,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8030088517606886e-05,
+      "loss": 4.0839,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8021702570096366e-05,
+      "loss": 4.0813,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8013316622585846e-05,
+      "loss": 4.0713,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800494705387906e-05,
+      "loss": 4.0733,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7996561106368535e-05,
+      "loss": 4.0716,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7988175158858015e-05,
+      "loss": 4.0769,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7979789211347494e-05,
+      "loss": 4.0684,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797141964264071e-05,
+      "loss": 4.0775,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796303369513019e-05,
+      "loss": 4.0719,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795464774761967e-05,
+      "loss": 4.0736,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794626180010915e-05,
+      "loss": 4.056,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793789223140236e-05,
+      "loss": 4.0674,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792950628389184e-05,
+      "loss": 4.0622,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792112033638132e-05,
+      "loss": 4.0683,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79127343888708e-05,
+      "loss": 4.0675,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790436482016401e-05,
+      "loss": 4.0582,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789597887265349e-05,
+      "loss": 4.06,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788759292514297e-05,
+      "loss": 4.0584,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787920697763245e-05,
+      "loss": 4.0615,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870837408925664e-05,
+      "loss": 4.0484,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786246784021888e-05,
+      "loss": 4.0508,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.785408189270836e-05,
+      "loss": 4.06,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784569594519783e-05,
+      "loss": 4.0503,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783730999768731e-05,
+      "loss": 4.0502,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782892405017679e-05,
+      "loss": 4.044,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782053810266627e-05,
+      "loss": 4.0421,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781215215515575e-05,
+      "loss": 4.0629,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780376620764523e-05,
+      "loss": 4.045,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779539663893844e-05,
+      "loss": 4.0533,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778701069142792e-05,
+      "loss": 4.0513,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77786247439174e-05,
+      "loss": 4.0634,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777025517521062e-05,
+      "loss": 4.0496,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77618692277001e-05,
+      "loss": 4.0355,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775348328018958e-05,
+      "loss": 4.044,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774509733267906e-05,
+      "loss": 4.0511,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773671138516854e-05,
+      "loss": 4.036,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7728341816461747e-05,
+      "loss": 4.0403,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719955868951227e-05,
+      "loss": 4.0369,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7711569921440707e-05,
+      "loss": 4.0474,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7703183973930186e-05,
+      "loss": 4.0344,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7694798026419666e-05,
+      "loss": 4.0439,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686412078909146e-05,
+      "loss": 4.036,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7678026131398626e-05,
+      "loss": 4.0382,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766965656269184e-05,
+      "loss": 4.0298,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766127061518132e-05,
+      "loss": 4.0513,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76528846676708e-05,
+      "loss": 4.0418,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764449872016028e-05,
+      "loss": 4.0265,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763612915145349e-05,
+      "loss": 4.0289,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762774320394297e-05,
+      "loss": 4.0373,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761935725643245e-05,
+      "loss": 4.0364,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761097130892193e-05,
+      "loss": 4.0444,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760260174021514e-05,
+      "loss": 4.0279,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759421579270462e-05,
+      "loss": 4.0357,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75858298451941e-05,
+      "loss": 4.0293,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757744389768358e-05,
+      "loss": 4.0357,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7569074328976796e-05,
+      "loss": 4.0265,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560688381466276e-05,
+      "loss": 4.0314,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552302433955756e-05,
+      "loss": 4.0237,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543916486445236e-05,
+      "loss": 4.018,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535546917738445e-05,
+      "loss": 4.0388,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527160970227925e-05,
+      "loss": 4.0253,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518775022717405e-05,
+      "loss": 4.022,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510389075206885e-05,
+      "loss": 4.0264,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7502003127696365e-05,
+      "loss": 4.0177,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.026247978210449,
+      "eval_runtime": 304.5864,
+      "eval_samples_per_second": 1252.817,
+      "eval_steps_per_second": 39.151,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493617180185845e-05,
+      "loss": 4.0071,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748523123267532e-05,
+      "loss": 4.0095,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74768452851648e-05,
+      "loss": 4.0252,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746845933765428e-05,
+      "loss": 4.0124,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7460073390143764e-05,
+      "loss": 4.0264,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7451687442633244e-05,
+      "loss": 4.0096,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7443301495122724e-05,
+      "loss": 4.012,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7434915547612204e-05,
+      "loss": 4.0067,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426529600101684e-05,
+      "loss": 4.0207,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7418143652591164e-05,
+      "loss": 4.0062,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409757705080644e-05,
+      "loss": 4.0046,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740138813637385e-05,
+      "loss": 4.0032,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739303494647079e-05,
+      "loss": 4.0055,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.738464899896027e-05,
+      "loss": 3.9975,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737626305144975e-05,
+      "loss": 4.0014,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736787710393923e-05,
+      "loss": 3.9905,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735949115642871e-05,
+      "loss": 3.9945,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73511052089182e-05,
+      "loss": 3.9904,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734271926140768e-05,
+      "loss": 3.9857,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733433331389716e-05,
+      "loss": 4.0199,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732594736638664e-05,
+      "loss": 3.9938,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731756141887612e-05,
+      "loss": 3.9901,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73091754713656e-05,
+      "loss": 3.997,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730080590265881e-05,
+      "loss": 4.0043,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292436333952016e-05,
+      "loss": 3.9825,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7284050386441496e-05,
+      "loss": 3.991,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275664438930976e-05,
+      "loss": 3.9905,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267278491420456e-05,
+      "loss": 3.9943,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258892543909936e-05,
+      "loss": 3.9799,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250506596399416e-05,
+      "loss": 3.9855,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7242120648888896e-05,
+      "loss": 3.9876,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723373470137838e-05,
+      "loss": 3.9938,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722536513267159e-05,
+      "loss": 3.9991,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721697918516107e-05,
+      "loss": 3.9877,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720859323765055e-05,
+      "loss": 3.9806,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720020729014003e-05,
+      "loss": 3.9915,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719183772143324e-05,
+      "loss": 3.9727,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718346815272645e-05,
+      "loss": 3.9854,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717508220521593e-05,
+      "loss": 3.9675,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716669625770541e-05,
+      "loss": 3.9742,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715831031019489e-05,
+      "loss": 3.978,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714992436268437e-05,
+      "loss": 3.9854,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714153841517385e-05,
+      "loss": 3.9804,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713315246766333e-05,
+      "loss": 3.9731,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712476652015281e-05,
+      "loss": 3.9779,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116396951446025e-05,
+      "loss": 3.9731,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7108011003935505e-05,
+      "loss": 3.9817,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709962505642498e-05,
+      "loss": 3.9597,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091255487718194e-05,
+      "loss": 3.9587,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082869540207674e-05,
+      "loss": 3.9771,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074483592697154e-05,
+      "loss": 3.9757,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706609764518663e-05,
+      "loss": 3.9645,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705771169767611e-05,
+      "loss": 3.9632,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704932575016559e-05,
+      "loss": 3.9681,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70409561814588e-05,
+      "loss": 3.9525,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703257023394828e-05,
+      "loss": 3.9745,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702418428643776e-05,
+      "loss": 3.9614,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701579833892724e-05,
+      "loss": 3.9693,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700741239141672e-05,
+      "loss": 3.9634,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69990264439062e-05,
+      "loss": 3.9589,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699064049639568e-05,
+      "loss": 3.9596,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698225454888516e-05,
+      "loss": 3.9595,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697388498017837e-05,
+      "loss": 3.9533,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696549903266785e-05,
+      "loss": 3.9506,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695711308515733e-05,
+      "loss": 3.9668,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694872713764681e-05,
+      "loss": 3.9545,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694035756894003e-05,
+      "loss": 3.9553,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693197162142951e-05,
+      "loss": 3.9445,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692358567391899e-05,
+      "loss": 3.9511,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691519972640847e-05,
+      "loss": 3.9683,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906846536505406e-05,
+      "loss": 3.9639,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898460588994886e-05,
+      "loss": 3.9555,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6890074641484366e-05,
+      "loss": 3.9634,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881688693973846e-05,
+      "loss": 3.9675,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873302746463326e-05,
+      "loss": 3.9498,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864916798952806e-05,
+      "loss": 3.9577,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856547230246015e-05,
+      "loss": 3.9514,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6848161282735495e-05,
+      "loss": 3.9438,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683977533522498e-05,
+      "loss": 3.9509,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683138938771446e-05,
+      "loss": 3.944,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682300344020394e-05,
+      "loss": 3.9527,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681461749269342e-05,
+      "loss": 3.9631,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68062315451829e-05,
+      "loss": 3.9597,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679786197647611e-05,
+      "loss": 3.9321,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678947602896559e-05,
+      "loss": 3.9492,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678109008145507e-05,
+      "loss": 3.9501,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677270413394455e-05,
+      "loss": 3.9493,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676433456523776e-05,
+      "loss": 3.9429,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675594861772724e-05,
+      "loss": 3.9371,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674756267021672e-05,
+      "loss": 3.9463,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.67391767227062e-05,
+      "loss": 3.9422,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730807153999415e-05,
+      "loss": 3.9384,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722421206488895e-05,
+      "loss": 3.949,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6714035258978375e-05,
+      "loss": 3.9409,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705649311467855e-05,
+      "loss": 3.9483,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697263363957335e-05,
+      "loss": 3.9321,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688893795250544e-05,
+      "loss": 3.9365,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680507847740024e-05,
+      "loss": 3.9374,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6672121900229504e-05,
+      "loss": 3.9457,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663735952718984e-05,
+      "loss": 3.9416,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665536638401219e-05,
+      "loss": 3.933,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664698043650167e-05,
+      "loss": 3.9362,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663859448899115e-05,
+      "loss": 3.9332,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663020854148063e-05,
+      "loss": 3.9374,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662182259397011e-05,
+      "loss": 3.9222,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661345302526333e-05,
+      "loss": 3.9255,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660506707775281e-05,
+      "loss": 3.9428,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659668113024229e-05,
+      "loss": 3.9317,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658829518273176e-05,
+      "loss": 3.9253,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657992561402498e-05,
+      "loss": 3.9232,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657153966651446e-05,
+      "loss": 3.9233,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656315371900393e-05,
+      "loss": 3.9391,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655476777149341e-05,
+      "loss": 3.9298,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654638182398289e-05,
+      "loss": 3.9282,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6538012255276106e-05,
+      "loss": 3.9304,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529626307765586e-05,
+      "loss": 3.9459,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521240360255066e-05,
+      "loss": 3.9325,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512854412744546e-05,
+      "loss": 3.9187,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650448484403776e-05,
+      "loss": 3.9286,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6496098896527235e-05,
+      "loss": 3.9302,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487712949016715e-05,
+      "loss": 3.9214,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479327001506195e-05,
+      "loss": 3.9239,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6470941053995675e-05,
+      "loss": 3.9207,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6462571485288884e-05,
+      "loss": 3.9315,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6454185537778364e-05,
+      "loss": 3.9177,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6445799590267844e-05,
+      "loss": 3.932,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437413642757324e-05,
+      "loss": 3.9149,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6429027695246804e-05,
+      "loss": 3.9267,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642065812654002e-05,
+      "loss": 3.9149,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64122721790295e-05,
+      "loss": 3.9377,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640388623151898e-05,
+      "loss": 3.9297,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639550028400846e-05,
+      "loss": 3.9194,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638713071530167e-05,
+      "loss": 3.9122,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637874476779115e-05,
+      "loss": 3.9253,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637035882028063e-05,
+      "loss": 3.9225,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636197287277011e-05,
+      "loss": 3.9323,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635358692525959e-05,
+      "loss": 3.915,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63452173565528e-05,
+      "loss": 3.9273,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633683140904228e-05,
+      "loss": 3.92,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632844546153176e-05,
+      "loss": 3.9309,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6320059514021244e-05,
+      "loss": 3.9137,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311673566510724e-05,
+      "loss": 3.9227,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303303997803934e-05,
+      "loss": 3.9161,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294918050293413e-05,
+      "loss": 3.9093,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286532102782893e-05,
+      "loss": 3.9275,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6278146155272373e-05,
+      "loss": 3.9229,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626976020776185e-05,
+      "loss": 3.91,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626137426025133e-05,
+      "loss": 3.9211,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625298831274081e-05,
+      "loss": 3.9112,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.953259229660034,
+      "eval_runtime": 312.5733,
+      "eval_samples_per_second": 1220.805,
+      "eval_steps_per_second": 38.151,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624461874403402e-05,
+      "loss": 3.9069,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.62362327965235e-05,
+      "loss": 3.9035,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622784684901298e-05,
+      "loss": 3.9145,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621946090150246e-05,
+      "loss": 3.9111,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621109133279568e-05,
+      "loss": 3.9243,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620270538528516e-05,
+      "loss": 3.9013,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619431943777464e-05,
+      "loss": 3.9091,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618593349026412e-05,
+      "loss": 3.9022,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617754754275359e-05,
+      "loss": 3.921,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616916159524307e-05,
+      "loss": 3.9018,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616079202653629e-05,
+      "loss": 3.8978,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615240607902577e-05,
+      "loss": 3.9052,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6144036510318976e-05,
+      "loss": 3.901,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135650562808456e-05,
+      "loss": 3.8982,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127264615297936e-05,
+      "loss": 3.8991,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6118878667787416e-05,
+      "loss": 3.8923,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110492720276896e-05,
+      "loss": 3.8901,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6102106772766376e-05,
+      "loss": 3.8921,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093720825255856e-05,
+      "loss": 3.8879,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085334877745336e-05,
+      "loss": 3.919,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076965309038545e-05,
+      "loss": 3.8928,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068579361528025e-05,
+      "loss": 3.8933,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6060193414017505e-05,
+      "loss": 3.8986,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051807466506985e-05,
+      "loss": 3.9037,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6043437897800194e-05,
+      "loss": 3.8894,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6035051950289674e-05,
+      "loss": 3.8957,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6026666002779154e-05,
+      "loss": 3.8906,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018280055268634e-05,
+      "loss": 3.8953,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600991048656185e-05,
+      "loss": 3.885,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600152453905133e-05,
+      "loss": 3.8884,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599313859154081e-05,
+      "loss": 3.8939,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598475264403029e-05,
+      "loss": 3.8965,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597636669651977e-05,
+      "loss": 3.9021,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596799712781298e-05,
+      "loss": 3.8945,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595961118030246e-05,
+      "loss": 3.8846,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595122523279194e-05,
+      "loss": 3.8946,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594283928528142e-05,
+      "loss": 3.8812,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593446971657463e-05,
+      "loss": 3.8886,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592608376906411e-05,
+      "loss": 3.8744,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591769782155359e-05,
+      "loss": 3.8805,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590931187404307e-05,
+      "loss": 3.8808,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590094230533628e-05,
+      "loss": 3.8946,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589255635782576e-05,
+      "loss": 3.8889,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588417041031524e-05,
+      "loss": 3.8808,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587578446280472e-05,
+      "loss": 3.8842,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58673985152942e-05,
+      "loss": 3.8822,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585902894658741e-05,
+      "loss": 3.8892,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585064299907689e-05,
+      "loss": 3.8715,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584225705156637e-05,
+      "loss": 3.8672,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583387110405585e-05,
+      "loss": 3.8874,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582550153534906e-05,
+      "loss": 3.8856,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581711558783854e-05,
+      "loss": 3.8694,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580872964032802e-05,
+      "loss": 3.8769,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580034369281751e-05,
+      "loss": 3.8801,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579197412411072e-05,
+      "loss": 3.8646,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.57835881766002e-05,
+      "loss": 3.8824,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577520222908968e-05,
+      "loss": 3.87,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576681628157916e-05,
+      "loss": 3.8845,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758446712872366e-05,
+      "loss": 3.8699,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5750060765361846e-05,
+      "loss": 3.879,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741674817851326e-05,
+      "loss": 3.8674,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733288870340806e-05,
+      "loss": 3.8727,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724902922830286e-05,
+      "loss": 3.8613,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716516975319765e-05,
+      "loss": 3.8649,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5708147406612975e-05,
+      "loss": 3.8832,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569976145910246e-05,
+      "loss": 3.8643,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569137551159194e-05,
+      "loss": 3.8701,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568298956408142e-05,
+      "loss": 3.8572,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567461999537463e-05,
+      "loss": 3.8676,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566623404786411e-05,
+      "loss": 3.8817,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565784810035359e-05,
+      "loss": 3.8769,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564946215284307e-05,
+      "loss": 3.8738,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564109258413628e-05,
+      "loss": 3.8793,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563270663662576e-05,
+      "loss": 3.8779,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562432068911524e-05,
+      "loss": 3.8676,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561593474160472e-05,
+      "loss": 3.8702,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560756517289793e-05,
+      "loss": 3.8705,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5599179225387415e-05,
+      "loss": 3.862,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5590793277876895e-05,
+      "loss": 3.8631,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5582407330366375e-05,
+      "loss": 3.8619,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5574037761659584e-05,
+      "loss": 3.867,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5565651814149064e-05,
+      "loss": 3.8786,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5557265866638544e-05,
+      "loss": 3.8748,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5548879919128024e-05,
+      "loss": 3.8482,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55404939716175e-05,
+      "loss": 3.8694,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553212440291071e-05,
+      "loss": 3.8662,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552373845540019e-05,
+      "loss": 3.8653,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515352507889666e-05,
+      "loss": 3.8609,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550696656037915e-05,
+      "loss": 3.8521,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549859699167237e-05,
+      "loss": 3.8665,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549021104416185e-05,
+      "loss": 3.8597,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548182509665132e-05,
+      "loss": 3.8594,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54734391491408e-05,
+      "loss": 3.8659,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546506958043402e-05,
+      "loss": 3.8612,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54566836329235e-05,
+      "loss": 3.8701,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544829768541297e-05,
+      "loss": 3.8539,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543991173790245e-05,
+      "loss": 3.8529,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5431542169195667e-05,
+      "loss": 3.8558,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542315622168514e-05,
+      "loss": 3.8692,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541477027417462e-05,
+      "loss": 3.8589,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406384326664106e-05,
+      "loss": 3.8536,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397998379153586e-05,
+      "loss": 3.8529,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389628810446796e-05,
+      "loss": 3.8564,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5381242862936275e-05,
+      "loss": 3.8647,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372856915425755e-05,
+      "loss": 3.8479,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364470967915235e-05,
+      "loss": 3.8461,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5356101399208444e-05,
+      "loss": 3.8624,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347715451697924e-05,
+      "loss": 3.8514,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5339329504187404e-05,
+      "loss": 3.8475,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5330943556676884e-05,
+      "loss": 3.8429,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5322573987970093e-05,
+      "loss": 3.8481,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5314188040459573e-05,
+      "loss": 3.8629,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530580209294906e-05,
+      "loss": 3.8462,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529741614543854e-05,
+      "loss": 3.851,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528904657673175e-05,
+      "loss": 3.8565,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528066062922123e-05,
+      "loss": 3.8667,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527227468171071e-05,
+      "loss": 3.8612,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526388873420019e-05,
+      "loss": 3.841,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52555191654934e-05,
+      "loss": 3.853,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524713321798288e-05,
+      "loss": 3.8499,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523874727047236e-05,
+      "loss": 3.8482,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523037770176557e-05,
+      "loss": 3.8496,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522199175425505e-05,
+      "loss": 3.8417,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521360580674453e-05,
+      "loss": 3.8555,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520521985923401e-05,
+      "loss": 3.844,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196833911723494e-05,
+      "loss": 3.8532,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188447964212974e-05,
+      "loss": 3.8427,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5180062016702454e-05,
+      "loss": 3.8509,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5171676069191934e-05,
+      "loss": 3.8389,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516330650048514e-05,
+      "loss": 3.8669,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515492055297462e-05,
+      "loss": 3.8475,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51465346054641e-05,
+      "loss": 3.8495,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513814865795358e-05,
+      "loss": 3.8381,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512977908924679e-05,
+      "loss": 3.8544,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512139314173627e-05,
+      "loss": 3.8478,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511300719422575e-05,
+      "loss": 3.855,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510462124671523e-05,
+      "loss": 3.8435,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509625167800845e-05,
+      "loss": 3.8521,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508786573049793e-05,
+      "loss": 3.8471,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507947978298741e-05,
+      "loss": 3.8602,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507109383547689e-05,
+      "loss": 3.8374,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062724266770096e-05,
+      "loss": 3.8512,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054338319259576e-05,
+      "loss": 3.8433,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5045952371749056e-05,
+      "loss": 3.8406,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5037566424238536e-05,
+      "loss": 3.8502,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5029180476728016e-05,
+      "loss": 3.8499,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020810908021225e-05,
+      "loss": 3.8395,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012424960510705e-05,
+      "loss": 3.8435,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5004039013000185e-05,
+      "loss": 3.8429,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.911320447921753,
+      "eval_runtime": 357.6326,
+      "eval_samples_per_second": 1066.992,
+      "eval_steps_per_second": 33.344,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995653065489665e-05,
+      "loss": 3.8338,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4987267117979145e-05,
+      "loss": 3.8298,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.497888117046863e-05,
+      "loss": 3.8431,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970495222958105e-05,
+      "loss": 3.8421,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4962109275447585e-05,
+      "loss": 3.8532,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953723327937065e-05,
+      "loss": 3.8351,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945337380426545e-05,
+      "loss": 3.8361,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936951432916025e-05,
+      "loss": 3.8326,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928581864209234e-05,
+      "loss": 3.8485,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4920195916698714e-05,
+      "loss": 3.8376,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911809969188194e-05,
+      "loss": 3.8276,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903424021677674e-05,
+      "loss": 3.8321,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489507083177462e-05,
+      "loss": 3.8307,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48866848842641e-05,
+      "loss": 3.8277,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487829893675358e-05,
+      "loss": 3.8356,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486991298924306e-05,
+      "loss": 3.819,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486152704173254e-05,
+      "loss": 3.825,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485314109422202e-05,
+      "loss": 3.8227,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48447551467115e-05,
+      "loss": 3.8199,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483636919920098e-05,
+      "loss": 3.8465,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482799963049419e-05,
+      "loss": 3.8298,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481961368298367e-05,
+      "loss": 3.8258,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481122773547315e-05,
+      "loss": 3.829,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480284178796263e-05,
+      "loss": 3.8382,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4794472219255837e-05,
+      "loss": 3.821,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478608627174532e-05,
+      "loss": 3.8293,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47777003242348e-05,
+      "loss": 3.8231,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476931437672428e-05,
+      "loss": 3.823,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476094480801749e-05,
+      "loss": 3.8182,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475255886050697e-05,
+      "loss": 3.8249,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474417291299645e-05,
+      "loss": 3.8268,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473578696548593e-05,
+      "loss": 3.8307,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472741739677914e-05,
+      "loss": 3.8346,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471903144926862e-05,
+      "loss": 3.829,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.47106455017581e-05,
+      "loss": 3.8172,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470225955424758e-05,
+      "loss": 3.8267,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469388998554079e-05,
+      "loss": 3.8197,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468550403803028e-05,
+      "loss": 3.8199,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467711809051976e-05,
+      "loss": 3.8098,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466873214300924e-05,
+      "loss": 3.8174,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660362574302446e-05,
+      "loss": 3.8189,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651976626791926e-05,
+      "loss": 3.8278,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643590679281406e-05,
+      "loss": 3.8227,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4635204731770886e-05,
+      "loss": 3.8184,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626835163064095e-05,
+      "loss": 3.819,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618449215553575e-05,
+      "loss": 3.8182,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4610063268043055e-05,
+      "loss": 3.8235,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601677320532535e-05,
+      "loss": 3.8079,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4593307751825744e-05,
+      "loss": 3.8003,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584921804315224e-05,
+      "loss": 3.8222,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457653585680471e-05,
+      "loss": 3.8211,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456814990929419e-05,
+      "loss": 3.8045,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45597803405874e-05,
+      "loss": 3.8132,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455139439307688e-05,
+      "loss": 3.8186,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454300844556636e-05,
+      "loss": 3.8002,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453462249805584e-05,
+      "loss": 3.8191,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452625292934905e-05,
+      "loss": 3.8098,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451786698183853e-05,
+      "loss": 3.8181,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450948103432801e-05,
+      "loss": 3.8081,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450109508681749e-05,
+      "loss": 3.8149,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44927255181107e-05,
+      "loss": 3.8064,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448433957060018e-05,
+      "loss": 3.8138,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475953623089664e-05,
+      "loss": 3.7988,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467567675579144e-05,
+      "loss": 3.8078,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445919810687235e-05,
+      "loss": 3.8174,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445081215936183e-05,
+      "loss": 3.8067,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444242621185131e-05,
+      "loss": 3.8051,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443404026434079e-05,
+      "loss": 3.7973,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425670695634e-05,
+      "loss": 3.8032,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441728474812348e-05,
+      "loss": 3.8141,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440889880061296e-05,
+      "loss": 3.8182,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440051285310244e-05,
+      "loss": 3.8123,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439214328439565e-05,
+      "loss": 3.8189,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438375733688513e-05,
+      "loss": 3.8192,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437537138937462e-05,
+      "loss": 3.8045,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43669854418641e-05,
+      "loss": 3.8074,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435861587315731e-05,
+      "loss": 3.8125,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435022992564679e-05,
+      "loss": 3.7998,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434184397813627e-05,
+      "loss": 3.801,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433345803062575e-05,
+      "loss": 3.8058,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4325088461918956e-05,
+      "loss": 3.8044,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316702514408436e-05,
+      "loss": 3.8202,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4308316566897916e-05,
+      "loss": 3.8157,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299930619387396e-05,
+      "loss": 3.7903,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291561050680605e-05,
+      "loss": 3.8059,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4283175103170085e-05,
+      "loss": 3.8083,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427478915565957e-05,
+      "loss": 3.8079,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.426640320814905e-05,
+      "loss": 3.8034,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425803363944226e-05,
+      "loss": 3.7914,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424964769193174e-05,
+      "loss": 3.8046,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424126174442122e-05,
+      "loss": 3.8014,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42328757969107e-05,
+      "loss": 3.8014,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422450622820391e-05,
+      "loss": 3.8019,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.421612028069339e-05,
+      "loss": 3.8064,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420773433318287e-05,
+      "loss": 3.8069,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419934838567235e-05,
+      "loss": 3.8044,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419097881696556e-05,
+      "loss": 3.7898,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418259286945504e-05,
+      "loss": 3.802,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4174206921944525e-05,
+      "loss": 3.8058,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4165820974434005e-05,
+      "loss": 3.8066,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4157451405727214e-05,
+      "loss": 3.7938,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4149065458216694e-05,
+      "loss": 3.7979,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4140679510706174e-05,
+      "loss": 3.7983,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4132293563195654e-05,
+      "loss": 3.8054,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412392399448886e-05,
+      "loss": 3.7925,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411553804697834e-05,
+      "loss": 3.7911,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410715209946782e-05,
+      "loss": 3.8002,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4098766151957296e-05,
+      "loss": 3.7964,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409039658325051e-05,
+      "loss": 3.7893,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408201063573999e-05,
+      "loss": 3.7893,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407362468822948e-05,
+      "loss": 3.7896,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406523874071895e-05,
+      "loss": 3.8049,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405686917201217e-05,
+      "loss": 3.7934,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404848322450165e-05,
+      "loss": 3.7913,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404009727699113e-05,
+      "loss": 3.7995,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40317113294806e-05,
+      "loss": 3.8112,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402334176077382e-05,
+      "loss": 3.8013,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40149558132633e-05,
+      "loss": 3.7856,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400656986575277e-05,
+      "loss": 3.793,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399818391824225e-05,
+      "loss": 3.7928,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3989814349535466e-05,
+      "loss": 3.7897,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398144478082868e-05,
+      "loss": 3.7946,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397305883331816e-05,
+      "loss": 3.7857,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396467288580764e-05,
+      "loss": 3.7975,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395628693829712e-05,
+      "loss": 3.7935,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39479009907866e-05,
+      "loss": 3.7977,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3939515043276075e-05,
+      "loss": 3.785,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3931129095765555e-05,
+      "loss": 3.7929,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3922743148255035e-05,
+      "loss": 3.7912,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3914373579548244e-05,
+      "loss": 3.8042,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3905987632037724e-05,
+      "loss": 3.7912,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3897601684527204e-05,
+      "loss": 3.7968,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3889215737016684e-05,
+      "loss": 3.7838,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38808461683099e-05,
+      "loss": 3.8005,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387246022079938e-05,
+      "loss": 3.7957,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386407427328886e-05,
+      "loss": 3.7979,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.385568832577834e-05,
+      "loss": 3.7861,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.384731875707155e-05,
+      "loss": 3.802,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383893280956103e-05,
+      "loss": 3.7893,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.383054686205051e-05,
+      "loss": 3.8066,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.382216091453999e-05,
+      "loss": 3.7844,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.38137913458332e-05,
+      "loss": 3.797,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.380540539832268e-05,
+      "loss": 3.7889,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379701945081216e-05,
+      "loss": 3.786,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378863350330164e-05,
+      "loss": 3.7989,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378026393459485e-05,
+      "loss": 3.7951,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377187798708433e-05,
+      "loss": 3.7873,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376349203957381e-05,
+      "loss": 3.7868,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375510609206329e-05,
+      "loss": 3.7931,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8865561485290527,
+      "eval_runtime": 305.1088,
+      "eval_samples_per_second": 1250.672,
+      "eval_steps_per_second": 39.084,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.37467365233565e-05,
+      "loss": 3.7839,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373835057584598e-05,
+      "loss": 3.778,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372996462833546e-05,
+      "loss": 3.7897,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372157868082494e-05,
+      "loss": 3.7874,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371320911211815e-05,
+      "loss": 3.7968,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370482316460763e-05,
+      "loss": 3.7823,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369643721709711e-05,
+      "loss": 3.7815,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368805126958659e-05,
+      "loss": 3.7811,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367968170087981e-05,
+      "loss": 3.7967,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367129575336929e-05,
+      "loss": 3.7819,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366290980585877e-05,
+      "loss": 3.7772,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365452385834825e-05,
+      "loss": 3.7791,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364617066844519e-05,
+      "loss": 3.7811,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637784720934665e-05,
+      "loss": 3.7763,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3629398773424145e-05,
+      "loss": 3.7837,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3621012825913625e-05,
+      "loss": 3.7683,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612626878403105e-05,
+      "loss": 3.7705,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3604240930892585e-05,
+      "loss": 3.7742,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3595854983382065e-05,
+      "loss": 3.7669,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3587469035871545e-05,
+      "loss": 3.793,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357909946716476e-05,
+      "loss": 3.7816,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357071351965424e-05,
+      "loss": 3.7762,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356232757214372e-05,
+      "loss": 3.7732,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35539416246332e-05,
+      "loss": 3.7843,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354557205592641e-05,
+      "loss": 3.7708,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353718610841589e-05,
+      "loss": 3.7802,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352880016090537e-05,
+      "loss": 3.7694,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352041421339485e-05,
+      "loss": 3.7753,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351204464468806e-05,
+      "loss": 3.7704,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350365869717754e-05,
+      "loss": 3.7726,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349527274966702e-05,
+      "loss": 3.7765,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34868868021565e-05,
+      "loss": 3.7793,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478517233449714e-05,
+      "loss": 3.7818,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3470131285939194e-05,
+      "loss": 3.7809,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461745338428674e-05,
+      "loss": 3.7657,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3453359390918154e-05,
+      "loss": 3.7789,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344498982221136e-05,
+      "loss": 3.7722,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343660387470084e-05,
+      "loss": 3.7636,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342821792719032e-05,
+      "loss": 3.7659,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34198319796798e-05,
+      "loss": 3.765,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341146241097301e-05,
+      "loss": 3.7685,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340307646346249e-05,
+      "loss": 3.7796,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339469051595197e-05,
+      "loss": 3.77,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.338630456844145e-05,
+      "loss": 3.7682,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.337793499973467e-05,
+      "loss": 3.7744,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336954905222415e-05,
+      "loss": 3.7687,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336116310471363e-05,
+      "loss": 3.7708,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335277715720311e-05,
+      "loss": 3.7596,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334440758849632e-05,
+      "loss": 3.7506,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33360216409858e-05,
+      "loss": 3.7747,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332763569347528e-05,
+      "loss": 3.7685,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331924974596476e-05,
+      "loss": 3.7605,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3310880177257966e-05,
+      "loss": 3.763,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3302494229747446e-05,
+      "loss": 3.767,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3294108282236926e-05,
+      "loss": 3.7519,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328573871353014e-05,
+      "loss": 3.7684,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327735276601962e-05,
+      "loss": 3.7662,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32689668185091e-05,
+      "loss": 3.7684,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326058087099858e-05,
+      "loss": 3.7625,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325219492348806e-05,
+      "loss": 3.7639,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324382535478127e-05,
+      "loss": 3.7554,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323543940727075e-05,
+      "loss": 3.7678,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322705345976023e-05,
+      "loss": 3.7537,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321866751224971e-05,
+      "loss": 3.7577,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321029794354292e-05,
+      "loss": 3.7702,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32019119960324e-05,
+      "loss": 3.7598,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319352604852188e-05,
+      "loss": 3.7557,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318514010101136e-05,
+      "loss": 3.7539,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3176770532304575e-05,
+      "loss": 3.755,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3168384584794055e-05,
+      "loss": 3.7652,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159998637283535e-05,
+      "loss": 3.7713,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151612689773015e-05,
+      "loss": 3.7632,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3143243121066224e-05,
+      "loss": 3.7717,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3134857173555704e-05,
+      "loss": 3.7692,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3126471226045184e-05,
+      "loss": 3.7654,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3118085278534664e-05,
+      "loss": 3.7582,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310971570982787e-05,
+      "loss": 3.7602,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310132976231735e-05,
+      "loss": 3.7592,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309294381480683e-05,
+      "loss": 3.7514,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308455786729631e-05,
+      "loss": 3.7649,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307618829858953e-05,
+      "loss": 3.7543,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306780235107901e-05,
+      "loss": 3.7709,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305941640356849e-05,
+      "loss": 3.7722,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305103045605797e-05,
+      "loss": 3.7432,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304266088735118e-05,
+      "loss": 3.7626,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303427493984066e-05,
+      "loss": 3.7549,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302588899233014e-05,
+      "loss": 3.7624,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301751942362335e-05,
+      "loss": 3.7573,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300913347611283e-05,
+      "loss": 3.748,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300074752860231e-05,
+      "loss": 3.7597,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299236158109179e-05,
+      "loss": 3.7546,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983992012385e-05,
+      "loss": 3.7593,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297560606487448e-05,
+      "loss": 3.7529,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296722011736396e-05,
+      "loss": 3.7621,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295883416985344e-05,
+      "loss": 3.761,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295046460114665e-05,
+      "loss": 3.7587,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294207865363613e-05,
+      "loss": 3.7428,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293369270612561e-05,
+      "loss": 3.757,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292530675861509e-05,
+      "loss": 3.7564,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29169371899083e-05,
+      "loss": 3.7623,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290855124239778e-05,
+      "loss": 3.7486,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290016529488726e-05,
+      "loss": 3.754,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289177934737674e-05,
+      "loss": 3.7485,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2883409778669956e-05,
+      "loss": 3.7619,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2875023831159436e-05,
+      "loss": 3.7448,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866637883648916e-05,
+      "loss": 3.7483,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2858251936138396e-05,
+      "loss": 3.7529,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849882367431605e-05,
+      "loss": 3.7513,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2841496419921085e-05,
+      "loss": 3.7431,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2833110472410565e-05,
+      "loss": 3.7483,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282472452490004e-05,
+      "loss": 3.7454,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2816354956193254e-05,
+      "loss": 3.7546,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807969008682734e-05,
+      "loss": 3.7561,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799583061172214e-05,
+      "loss": 3.7438,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2791197113661694e-05,
+      "loss": 3.7574,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27828275449549e-05,
+      "loss": 3.7672,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277444159744439e-05,
+      "loss": 3.7587,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276605564993387e-05,
+      "loss": 3.7395,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275766970242334e-05,
+      "loss": 3.7514,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274930013371656e-05,
+      "loss": 3.7478,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274091418620604e-05,
+      "loss": 3.7441,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273252823869551e-05,
+      "loss": 3.7497,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272414229118499e-05,
+      "loss": 3.7398,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271577272247821e-05,
+      "loss": 3.7535,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270738677496769e-05,
+      "loss": 3.7476,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269900082745716e-05,
+      "loss": 3.7559,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269061487994665e-05,
+      "loss": 3.7438,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268224531123986e-05,
+      "loss": 3.7448,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2673859363729344e-05,
+      "loss": 3.7522,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266547341621882e-05,
+      "loss": 3.7585,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26570874687083e-05,
+      "loss": 3.7483,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264871790000151e-05,
+      "loss": 3.7528,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2640331952490986e-05,
+      "loss": 3.741,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631946004980466e-05,
+      "loss": 3.7573,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623560057469946e-05,
+      "loss": 3.7498,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261519048876316e-05,
+      "loss": 3.7574,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606804541252635e-05,
+      "loss": 3.7454,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2598418593742115e-05,
+      "loss": 3.7564,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2590032646231595e-05,
+      "loss": 3.7451,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258166307752481e-05,
+      "loss": 3.7663,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257327713001429e-05,
+      "loss": 3.7398,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256489118250377e-05,
+      "loss": 3.7563,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255650523499325e-05,
+      "loss": 3.7431,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254813566628646e-05,
+      "loss": 3.744,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253974971877594e-05,
+      "loss": 3.751,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253136377126542e-05,
+      "loss": 3.7499,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25229778237549e-05,
+      "loss": 3.7515,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251460825504811e-05,
+      "loss": 3.7394,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250622230753759e-05,
+      "loss": 3.7526,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.8701553344726562,
+      "eval_runtime": 306.7651,
+      "eval_samples_per_second": 1243.919,
+      "eval_steps_per_second": 38.873,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249783636002707e-05,
+      "loss": 3.7459,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248945041251655e-05,
+      "loss": 3.7344,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2481064465006035e-05,
+      "loss": 3.744,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472678517495515e-05,
+      "loss": 3.7485,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2464292569984995e-05,
+      "loss": 3.7525,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455906622474475e-05,
+      "loss": 3.7435,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2447520674963955e-05,
+      "loss": 3.7336,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2439134727453435e-05,
+      "loss": 3.741,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2430765158746644e-05,
+      "loss": 3.7508,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2422379211236124e-05,
+      "loss": 3.7413,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2413993263725604e-05,
+      "loss": 3.7363,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2405607316215084e-05,
+      "loss": 3.7348,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239725412631202e-05,
+      "loss": 3.7426,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23888681788015e-05,
+      "loss": 3.732,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238048223129099e-05,
+      "loss": 3.7418,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237209628378047e-05,
+      "loss": 3.7263,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236371033626995e-05,
+      "loss": 3.7277,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235532438875943e-05,
+      "loss": 3.73,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234693844124891e-05,
+      "loss": 3.7271,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233855249373839e-05,
+      "loss": 3.7506,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23301829250316e-05,
+      "loss": 3.7438,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232179697752108e-05,
+      "loss": 3.7355,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231341103001056e-05,
+      "loss": 3.7315,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230502508250004e-05,
+      "loss": 3.7409,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296655513793247e-05,
+      "loss": 3.7298,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2288269566282726e-05,
+      "loss": 3.7403,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279883618772206e-05,
+      "loss": 3.7286,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2271497671261686e-05,
+      "loss": 3.7349,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22631281025549e-05,
+      "loss": 3.7313,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225474215504438e-05,
+      "loss": 3.7294,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224635620753386e-05,
+      "loss": 3.7354,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223797026002334e-05,
+      "loss": 3.7423,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222960069131655e-05,
+      "loss": 3.7364,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222121474380603e-05,
+      "loss": 3.7389,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221282879629551e-05,
+      "loss": 3.7252,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220444284878499e-05,
+      "loss": 3.7433,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21960732800782e-05,
+      "loss": 3.7312,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218768733256768e-05,
+      "loss": 3.723,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217930138505716e-05,
+      "loss": 3.7261,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170931816350376e-05,
+      "loss": 3.7258,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2162545868839856e-05,
+      "loss": 3.7287,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2154159921329336e-05,
+      "loss": 3.7347,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145773973818816e-05,
+      "loss": 3.7299,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2137404405112025e-05,
+      "loss": 3.7308,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2129018457601505e-05,
+      "loss": 3.7348,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2120632510090985e-05,
+      "loss": 3.7307,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2112246562580465e-05,
+      "loss": 3.7307,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103876993873674e-05,
+      "loss": 3.718,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2095491046363154e-05,
+      "loss": 3.7089,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2087105098852634e-05,
+      "loss": 3.7348,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2078719151342114e-05,
+      "loss": 3.7339,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207034958263533e-05,
+      "loss": 3.7189,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206196363512481e-05,
+      "loss": 3.7287,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205357768761429e-05,
+      "loss": 3.7239,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204519174010377e-05,
+      "loss": 3.714,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203682217139698e-05,
+      "loss": 3.7245,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202843622388646e-05,
+      "loss": 3.7287,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202005027637594e-05,
+      "loss": 3.7288,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201166432886542e-05,
+      "loss": 3.7222,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200329476015863e-05,
+      "loss": 3.7278,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199490881264811e-05,
+      "loss": 3.7168,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198652286513759e-05,
+      "loss": 3.7301,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197813691762707e-05,
+      "loss": 3.7156,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196976734892028e-05,
+      "loss": 3.7183,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196138140140976e-05,
+      "loss": 3.7306,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195299545389924e-05,
+      "loss": 3.722,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1944609506388716e-05,
+      "loss": 3.7189,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193623993768193e-05,
+      "loss": 3.7157,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192785399017141e-05,
+      "loss": 3.7145,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191946804266089e-05,
+      "loss": 3.721,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1911082095150365e-05,
+      "loss": 3.7343,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190271252644358e-05,
+      "loss": 3.7226,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189432657893306e-05,
+      "loss": 3.7354,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188594063142254e-05,
+      "loss": 3.7297,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187755468391202e-05,
+      "loss": 3.731,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186918511520524e-05,
+      "loss": 3.7208,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186079916769472e-05,
+      "loss": 3.7222,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185241322018419e-05,
+      "loss": 3.7211,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184402727267367e-05,
+      "loss": 3.7113,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835657703966886e-05,
+      "loss": 3.7262,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1827271756456366e-05,
+      "loss": 3.7192,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181888580894584e-05,
+      "loss": 3.7323,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181049986143532e-05,
+      "loss": 3.7307,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1802130292728535e-05,
+      "loss": 3.7053,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179374434521801e-05,
+      "loss": 3.7279,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178535839770749e-05,
+      "loss": 3.7168,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776972450196975e-05,
+      "loss": 3.7238,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176860288149019e-05,
+      "loss": 3.7184,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1760216933979664e-05,
+      "loss": 3.7121,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751830986469144e-05,
+      "loss": 3.7217,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1743445038958624e-05,
+      "loss": 3.7162,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173507547025183e-05,
+      "loss": 3.7199,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172668952274131e-05,
+      "loss": 3.718,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171830357523079e-05,
+      "loss": 3.7246,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170991762772027e-05,
+      "loss": 3.7289,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170154805901348e-05,
+      "loss": 3.7145,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169316211150296e-05,
+      "loss": 3.7103,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168477616399244e-05,
+      "loss": 3.7183,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167639021648193e-05,
+      "loss": 3.7184,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166802064777514e-05,
+      "loss": 3.7281,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165963470026462e-05,
+      "loss": 3.7125,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16512487527541e-05,
+      "loss": 3.7162,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642879184047307e-05,
+      "loss": 3.7132,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1634493236536786e-05,
+      "loss": 3.7254,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1626107289026266e-05,
+      "loss": 3.7068,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617721341515746e-05,
+      "loss": 3.713,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1609351772808955e-05,
+      "loss": 3.7129,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1600965825298435e-05,
+      "loss": 3.7158,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1592579877787915e-05,
+      "loss": 3.7058,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1584193930277395e-05,
+      "loss": 3.7127,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157582436157061e-05,
+      "loss": 3.7057,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156743841406009e-05,
+      "loss": 3.7203,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155905246654957e-05,
+      "loss": 3.7187,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155066651903905e-05,
+      "loss": 3.7104,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154229695033226e-05,
+      "loss": 3.7169,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153391100282174e-05,
+      "loss": 3.7317,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152552505531122e-05,
+      "loss": 3.7269,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15171391078007e-05,
+      "loss": 3.7049,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508785917897645e-05,
+      "loss": 3.7121,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1500399970387125e-05,
+      "loss": 3.7129,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1492014022876605e-05,
+      "loss": 3.7031,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1483628075366085e-05,
+      "loss": 3.7161,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1475242127855565e-05,
+      "loss": 3.6987,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466856180345045e-05,
+      "loss": 3.7205,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1458470232834525e-05,
+      "loss": 3.7068,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1450084285324005e-05,
+      "loss": 3.7211,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1441714716617214e-05,
+      "loss": 3.7098,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1433328769106694e-05,
+      "loss": 3.7098,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424942821596174e-05,
+      "loss": 3.7132,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1416556874085654e-05,
+      "loss": 3.7226,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140818730537886e-05,
+      "loss": 3.7104,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139980135786834e-05,
+      "loss": 3.718,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139141541035782e-05,
+      "loss": 3.7036,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13830294628473e-05,
+      "loss": 3.7227,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137465989414052e-05,
+      "loss": 3.7126,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136627394663e-05,
+      "loss": 3.7163,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135788799911948e-05,
+      "loss": 3.7119,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134950205160896e-05,
+      "loss": 3.7193,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134113248290217e-05,
+      "loss": 3.7115,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133274653539165e-05,
+      "loss": 3.7281,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132436058788113e-05,
+      "loss": 3.7058,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131597464037061e-05,
+      "loss": 3.7196,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1307605071663816e-05,
+      "loss": 3.7117,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1299219124153296e-05,
+      "loss": 3.7125,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1290833176642776e-05,
+      "loss": 3.71,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1282447229132256e-05,
+      "loss": 3.7137,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127407766042547e-05,
+      "loss": 3.7162,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126569171291495e-05,
+      "loss": 3.7047,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125730576540443e-05,
+      "loss": 3.7213,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.8593947887420654,
+      "eval_runtime": 308.7788,
+      "eval_samples_per_second": 1235.807,
+      "eval_steps_per_second": 38.62,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124891981789391e-05,
+      "loss": 3.7047,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124055024918712e-05,
+      "loss": 3.7008,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12321643016766e-05,
+      "loss": 3.7091,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122377835416608e-05,
+      "loss": 3.7105,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121540878545929e-05,
+      "loss": 3.7173,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120702283794877e-05,
+      "loss": 3.7105,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119863689043825e-05,
+      "loss": 3.7034,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119025094292773e-05,
+      "loss": 3.7034,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1181881374220946e-05,
+      "loss": 3.7161,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1173495426710426e-05,
+      "loss": 3.7094,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1165109479199906e-05,
+      "loss": 3.7013,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1156723531689386e-05,
+      "loss": 3.6989,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1148370341786324e-05,
+      "loss": 3.7062,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1139984394275804e-05,
+      "loss": 3.6988,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1131598446765284e-05,
+      "loss": 3.7065,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1123212499254764e-05,
+      "loss": 3.6901,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1114826551744244e-05,
+      "loss": 3.6957,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1106440604233724e-05,
+      "loss": 3.6971,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1098054656723204e-05,
+      "loss": 3.6935,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1089668709212684e-05,
+      "loss": 3.7127,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1081282761702164e-05,
+      "loss": 3.709,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107289681419165e-05,
+      "loss": 3.702,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106451086668113e-05,
+      "loss": 3.6968,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105612491917061e-05,
+      "loss": 3.7064,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104775535046382e-05,
+      "loss": 3.6974,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10393694029533e-05,
+      "loss": 3.7063,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103098345544278e-05,
+      "loss": 3.693,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102259750793226e-05,
+      "loss": 3.705,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101422793922547e-05,
+      "loss": 3.7017,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100584199171495e-05,
+      "loss": 3.6929,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099745604420443e-05,
+      "loss": 3.6986,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098907009669391e-05,
+      "loss": 3.7076,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098070052798712e-05,
+      "loss": 3.7018,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09723145804766e-05,
+      "loss": 3.7118,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0963928632966084e-05,
+      "loss": 3.6911,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0955542685455564e-05,
+      "loss": 3.7034,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094717311674877e-05,
+      "loss": 3.7005,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093878716923825e-05,
+      "loss": 3.6912,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093040122172773e-05,
+      "loss": 3.6926,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092203165302094e-05,
+      "loss": 3.6914,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091364570551042e-05,
+      "loss": 3.6933,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09052597579999e-05,
+      "loss": 3.7035,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089687381048938e-05,
+      "loss": 3.6971,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088850424178259e-05,
+      "loss": 3.698,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088011829427207e-05,
+      "loss": 3.7028,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087173234676155e-05,
+      "loss": 3.6963,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086334639925104e-05,
+      "loss": 3.6944,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085497683054425e-05,
+      "loss": 3.6907,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084659088303373e-05,
+      "loss": 3.6699,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083820493552321e-05,
+      "loss": 3.7068,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082981898801269e-05,
+      "loss": 3.6935,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0821449419305896e-05,
+      "loss": 3.6937,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0813063471795376e-05,
+      "loss": 3.6918,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0804677524284856e-05,
+      "loss": 3.6917,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079629157677433e-05,
+      "loss": 3.6805,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787922008067545e-05,
+      "loss": 3.6901,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0779536060557025e-05,
+      "loss": 3.6968,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0771150113046505e-05,
+      "loss": 3.694,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762764165535985e-05,
+      "loss": 3.6963,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07543945968292e-05,
+      "loss": 3.6878,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074600864931868e-05,
+      "loss": 3.6844,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073762270180816e-05,
+      "loss": 3.7008,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0729236754297634e-05,
+      "loss": 3.6791,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072086718559085e-05,
+      "loss": 3.6863,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071248123808033e-05,
+      "loss": 3.6962,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07040952905698e-05,
+      "loss": 3.6909,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069570934305928e-05,
+      "loss": 3.6828,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06873397743525e-05,
+      "loss": 3.6864,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067895382684198e-05,
+      "loss": 3.6787,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067056787933146e-05,
+      "loss": 3.6883,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066218193182094e-05,
+      "loss": 3.7014,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0653812363114154e-05,
+      "loss": 3.6906,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0645426415603634e-05,
+      "loss": 3.7023,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.063704046809311e-05,
+      "loss": 3.6959,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062865452058259e-05,
+      "loss": 3.704,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06202849518758e-05,
+      "loss": 3.6862,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611899004365276e-05,
+      "loss": 3.694,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0603513056854756e-05,
+      "loss": 3.6905,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0595127109344236e-05,
+      "loss": 3.676,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058675754063745e-05,
+      "loss": 3.6946,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.057837159312693e-05,
+      "loss": 3.6872,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056998564561641e-05,
+      "loss": 3.6985,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056159969810589e-05,
+      "loss": 3.7012,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055323012939911e-05,
+      "loss": 3.6765,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054484418188858e-05,
+      "loss": 3.6915,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053645823437806e-05,
+      "loss": 3.6854,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052808866567128e-05,
+      "loss": 3.6944,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051970271816075e-05,
+      "loss": 3.6849,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051131677065023e-05,
+      "loss": 3.6814,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050293082313971e-05,
+      "loss": 3.6864,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049454487562919e-05,
+      "loss": 3.6855,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048615892811867e-05,
+      "loss": 3.6884,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047777298060815e-05,
+      "loss": 3.6869,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0469387033097637e-05,
+      "loss": 3.6965,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0461017464390846e-05,
+      "loss": 3.6948,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0452631516880326e-05,
+      "loss": 3.6831,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0444245569369806e-05,
+      "loss": 3.6775,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435859621859285e-05,
+      "loss": 3.6859,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0427490053152495e-05,
+      "loss": 3.6917,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0419104105641975e-05,
+      "loss": 3.6921,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0410718158131454e-05,
+      "loss": 3.6854,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0402332210620934e-05,
+      "loss": 3.6831,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0393962641914144e-05,
+      "loss": 3.6792,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0385576694403623e-05,
+      "loss": 3.6967,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0377190746893103e-05,
+      "loss": 3.672,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036880479938259e-05,
+      "loss": 3.6827,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03604352306758e-05,
+      "loss": 3.6781,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035204928316528e-05,
+      "loss": 3.6865,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034366333565476e-05,
+      "loss": 3.6752,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033527738814424e-05,
+      "loss": 3.687,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032690781943745e-05,
+      "loss": 3.6736,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031852187192693e-05,
+      "loss": 3.6875,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031013592441641e-05,
+      "loss": 3.6874,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030176635570962e-05,
+      "loss": 3.6782,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02933804081991e-05,
+      "loss": 3.686,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028499446068858e-05,
+      "loss": 3.7022,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027660851317806e-05,
+      "loss": 3.6854,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026823894447127e-05,
+      "loss": 3.6841,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025986937576448e-05,
+      "loss": 3.6746,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025148342825396e-05,
+      "loss": 3.6816,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024309748074344e-05,
+      "loss": 3.6792,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023471153323292e-05,
+      "loss": 3.6823,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02263255857224e-05,
+      "loss": 3.6673,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021793963821188e-05,
+      "loss": 3.6886,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020955369070136e-05,
+      "loss": 3.6781,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020116774319084e-05,
+      "loss": 3.6873,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019279817448405e-05,
+      "loss": 3.6786,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018441222697353e-05,
+      "loss": 3.6824,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017602627946301e-05,
+      "loss": 3.6797,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016764033195249e-05,
+      "loss": 3.6942,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0159270763245707e-05,
+      "loss": 3.6805,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0150884815735187e-05,
+      "loss": 3.6885,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0142498868224667e-05,
+      "loss": 3.6707,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0134112920714146e-05,
+      "loss": 3.6925,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0125743352007356e-05,
+      "loss": 3.6827,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0117357404496836e-05,
+      "loss": 3.6863,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0108971456986315e-05,
+      "loss": 3.6813,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0100585509475795e-05,
+      "loss": 3.6885,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0092215940769005e-05,
+      "loss": 3.6774,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0083829993258484e-05,
+      "loss": 3.6963,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0075444045747964e-05,
+      "loss": 3.6758,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0067058098237444e-05,
+      "loss": 3.6889,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005868852953066e-05,
+      "loss": 3.6825,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005030258202014e-05,
+      "loss": 3.6809,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004191663450962e-05,
+      "loss": 3.6773,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00335306869991e-05,
+      "loss": 3.6862,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002516111829231e-05,
+      "loss": 3.6867,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001677517078179e-05,
+      "loss": 3.6757,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000838922327127e-05,
+      "loss": 3.6897,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.853278398513794,
+      "eval_runtime": 313.1401,
+      "eval_samples_per_second": 1218.595,
+      "eval_steps_per_second": 38.082,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.000000327576075e-05,
+      "loss": 3.6777,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999161732825023e-05,
+      "loss": 3.6712,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998323138073971e-05,
+      "loss": 3.6763,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997484543322918e-05,
+      "loss": 3.6819,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996645948571867e-05,
+      "loss": 3.687,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995807353820815e-05,
+      "loss": 3.6781,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994968759069763e-05,
+      "loss": 3.6789,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994130164318711e-05,
+      "loss": 3.6741,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993293207448032e-05,
+      "loss": 3.683,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99245461269698e-05,
+      "loss": 3.6831,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991616017945928e-05,
+      "loss": 3.6701,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990777423194876e-05,
+      "loss": 3.6677,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98994210420457e-05,
+      "loss": 3.6821,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989103509453518e-05,
+      "loss": 3.668,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9882649147024656e-05,
+      "loss": 3.6715,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9874263199514136e-05,
+      "loss": 3.6644,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986587725200362e-05,
+      "loss": 3.6651,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.98574913044931e-05,
+      "loss": 3.6717,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984910535698258e-05,
+      "loss": 3.6638,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984071940947206e-05,
+      "loss": 3.6824,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983234984076527e-05,
+      "loss": 3.68,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982396389325475e-05,
+      "loss": 3.6762,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981557794574423e-05,
+      "loss": 3.6681,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980719199823371e-05,
+      "loss": 3.6736,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979882242952692e-05,
+      "loss": 3.6706,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97904364820164e-05,
+      "loss": 3.674,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978205053450588e-05,
+      "loss": 3.6668,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977366458699536e-05,
+      "loss": 3.676,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9765295018288576e-05,
+      "loss": 3.6665,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9756909070778056e-05,
+      "loss": 3.667,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9748523123267536e-05,
+      "loss": 3.6652,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9740137175757016e-05,
+      "loss": 3.681,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9731767607050225e-05,
+      "loss": 3.677,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9723398038343434e-05,
+      "loss": 3.6806,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9715012090832914e-05,
+      "loss": 3.6615,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9706626143322394e-05,
+      "loss": 3.6683,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9698240195811874e-05,
+      "loss": 3.6781,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9689854248301354e-05,
+      "loss": 3.6617,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9681468300790834e-05,
+      "loss": 3.6615,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9673082353280314e-05,
+      "loss": 3.662,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9664696405769794e-05,
+      "loss": 3.6652,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965632683706301e-05,
+      "loss": 3.6679,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964794088955249e-05,
+      "loss": 3.676,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963955494204197e-05,
+      "loss": 3.6715,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963116899453145e-05,
+      "loss": 3.6674,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962279942582466e-05,
+      "loss": 3.671,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961441347831414e-05,
+      "loss": 3.6623,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960602753080362e-05,
+      "loss": 3.6635,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95976415832931e-05,
+      "loss": 3.6399,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958927201458631e-05,
+      "loss": 3.6789,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958088606707579e-05,
+      "loss": 3.6638,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957250011956527e-05,
+      "loss": 3.671,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956411417205475e-05,
+      "loss": 3.6612,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9555744603347964e-05,
+      "loss": 3.6618,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9547358655837444e-05,
+      "loss": 3.6533,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953898908713065e-05,
+      "loss": 3.6611,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953060313962013e-05,
+      "loss": 3.671,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952221719210961e-05,
+      "loss": 3.6638,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951383124459909e-05,
+      "loss": 3.6723,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95054616758923e-05,
+      "loss": 3.661,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949707572838178e-05,
+      "loss": 3.6528,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948868978087126e-05,
+      "loss": 3.6742,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948030383336074e-05,
+      "loss": 3.6499,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947191788585022e-05,
+      "loss": 3.6544,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94635319383397e-05,
+      "loss": 3.6692,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945514599082918e-05,
+      "loss": 3.6671,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.944676004331866e-05,
+      "loss": 3.6503,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943839047461188e-05,
+      "loss": 3.6641,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.943000452710136e-05,
+      "loss": 3.6479,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.942161857959084e-05,
+      "loss": 3.6635,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941323263208032e-05,
+      "loss": 3.6681,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9404863063373526e-05,
+      "loss": 3.6682,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9396477115863006e-05,
+      "loss": 3.6669,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9388091168352486e-05,
+      "loss": 3.6686,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937970522084196e-05,
+      "loss": 3.6781,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937131927333144e-05,
+      "loss": 3.6564,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362949704624655e-05,
+      "loss": 3.6675,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9354563757114135e-05,
+      "loss": 3.662,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9346177809603615e-05,
+      "loss": 3.6437,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9337791862093095e-05,
+      "loss": 3.668,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9329405914582575e-05,
+      "loss": 3.6618,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9321019967072055e-05,
+      "loss": 3.6664,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9312634019561535e-05,
+      "loss": 3.6774,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9304248072051015e-05,
+      "loss": 3.6508,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9295878503344224e-05,
+      "loss": 3.6567,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9287492555833704e-05,
+      "loss": 3.6587,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927912298712691e-05,
+      "loss": 3.6674,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927073703961639e-05,
+      "loss": 3.6544,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926235109210587e-05,
+      "loss": 3.6563,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925396514459535e-05,
+      "loss": 3.659,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924557919708484e-05,
+      "loss": 3.6607,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923719324957432e-05,
+      "loss": 3.6573,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92288073020638e-05,
+      "loss": 3.6609,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922042135455328e-05,
+      "loss": 3.6685,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921205178584649e-05,
+      "loss": 3.6616,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920366583833597e-05,
+      "loss": 3.6635,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919527989082545e-05,
+      "loss": 3.6477,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918691032211866e-05,
+      "loss": 3.6587,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917852437460814e-05,
+      "loss": 3.6612,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917013842709762e-05,
+      "loss": 3.6686,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91617524795871e-05,
+      "loss": 3.6574,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915336653207658e-05,
+      "loss": 3.654,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914498058456606e-05,
+      "loss": 3.6585,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913659463705554e-05,
+      "loss": 3.6634,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9128208689545024e-05,
+      "loss": 3.6476,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911983912083823e-05,
+      "loss": 3.6557,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911145317332771e-05,
+      "loss": 3.6462,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910306722581719e-05,
+      "loss": 3.6652,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909468127830667e-05,
+      "loss": 3.6478,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908631170959988e-05,
+      "loss": 3.656,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907792576208936e-05,
+      "loss": 3.6479,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906953981457884e-05,
+      "loss": 3.6584,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906117024587205e-05,
+      "loss": 3.6629,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905278429836153e-05,
+      "loss": 3.6515,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904439835085101e-05,
+      "loss": 3.6569,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903601240334049e-05,
+      "loss": 3.6706,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902762645582998e-05,
+      "loss": 3.6637,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901924050831945e-05,
+      "loss": 3.6523,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9010870939612667e-05,
+      "loss": 3.6471,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9002484992102147e-05,
+      "loss": 3.6528,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899409904459162e-05,
+      "loss": 3.658,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89857130970811e-05,
+      "loss": 3.6493,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897732714957058e-05,
+      "loss": 3.644,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896894120206006e-05,
+      "loss": 3.6606,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896055525454954e-05,
+      "loss": 3.6535,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895218568584275e-05,
+      "loss": 3.6615,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894379973833223e-05,
+      "loss": 3.6492,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8935413790821715e-05,
+      "loss": 3.6555,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8927027843311195e-05,
+      "loss": 3.6503,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8918641895800675e-05,
+      "loss": 3.6649,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8910255948290155e-05,
+      "loss": 3.6575,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8901886379583364e-05,
+      "loss": 3.6589,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8893500432072844e-05,
+      "loss": 3.6455,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8885114484562324e-05,
+      "loss": 3.6689,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8876728537051804e-05,
+      "loss": 3.6563,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886835896834501e-05,
+      "loss": 3.6617,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885997302083449e-05,
+      "loss": 3.6539,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885158707332397e-05,
+      "loss": 3.6565,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884320112581345e-05,
+      "loss": 3.6543,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883481517830293e-05,
+      "loss": 3.6703,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882642923079241e-05,
+      "loss": 3.6498,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881804328328189e-05,
+      "loss": 3.6642,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880965733577138e-05,
+      "loss": 3.6564,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880128776706459e-05,
+      "loss": 3.6524,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879290181955407e-05,
+      "loss": 3.6529,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878451587204355e-05,
+      "loss": 3.6606,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877612992453303e-05,
+      "loss": 3.6574,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876776035582624e-05,
+      "loss": 3.6547,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875937440831572e-05,
+      "loss": 3.6622,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.848841428756714,
+      "eval_runtime": 303.5435,
+      "eval_samples_per_second": 1257.121,
+      "eval_steps_per_second": 39.286,
+      "step": 686880
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.747449239165768e+17,
+  "trial_name": null,
+  "trial_params": null
+}